key: cord-0041936-j58tdxhn
authors: nan
title: Poster Session Abstracts
date: 2018-09-07
journal: Pediatr Pulmonol
DOI: 10.1002/ppul.24152
sha: 1cfbfd8f93ad8e1c48eb3c4448c594c85cdb0d84
doc_id: 41936
cord_uid: j58tdxhn

nan

Most cases of cystic fibrosis (CF) are caused by loss of CFTR expression due to genetic mutations that decrease protein folding efficiency. While the efficiency of NBD1 folding is a limiting step in CFTR biosynthesis, and NBD1 does not reversibly fold in vitro, little is known about how CF-causing mutations affect the de-novo folding landscape as CFTR is synthesized in the cell. To address this question, we examined 10 CF-causing, missense mutations to determine their effect on NBD1 folding intermediates using FRET between an N-terminal CFP donor and a small acceptor dye that were cotranslationally incorporated at specific locations within NBD1 (Khushoo A, et al. Mol Cell. 2011; 41:682-92; and Kim SJ, et al. Science. 2015; 348:444-8) . This approach enabled us to compare compaction of wild-type and mutant nascent polypeptides at defined stages of synthesis and identify mutations that cotranslationally perturb folding of NBD1 nascent polypeptides as they emerge from the ribosome. Interestingly, two mutations, A455E and L558S, located within different subdomains of NBD1 showed similar effect in disrupting folding of both the NBD1 α-subdomain and its β-sheet core. In addition, A455E and L558S folding defects are transiently observed only during a brief window of synthesis, namely as the α-subdomain and the β-sheet core are synthesized, whereas the mutation-induced folding defect was not observed in the full-length domain. Thus, these two mutations appear to affect folding by perturbing the timing of sequential subdomain compactions. Surprisingly, this transient folding defect was observed as the nascent polypeptide emerged from the ribosome, and was not observed if the nascent protein was released from the ribosome into solution. Direct effects of the ribosome on the growing polypeptide chain therefore participate in cotranslational misfolding: first by destabilizing the misfolded state, and subsequently by facilitating eventual compaction of the domain. Remarkably, two suppressor mutations (S492P and I539T) eliminated the observed cotranslational folding defect for A455E NBD1 and partially corrected trafficking of full-length A455E CFTR. Our result indicates that cotranslational folding intermediates are targets for disruption by disease-causing mutations and that mutations which disrupt folding intermediates can have major impact on a final folding outcome in disease. Thus, drugs that target the cotranslational folding pathway may provide an important therapeutic strategy for treating certain misfolding diseases.

Oliver, K.E. 1, 2 ; Rauscher, R. 3 ; Ali, H. 1 ; Icyuz, M. 4, 5 ; Rab, A. 1,2 ; Hong, J.S. 1, 2 ; Hartman, J.L. 4, 5 ; Ignatova, Z. 3 ; Sorscher, E.J. 1, 2 1. Pediatrics, Emory University, Atlanta, GA, USA; 2. Children's Healthcare of Atlanta, Atlanta, GA, USA; 3. Biochemistry & Molecular Biology, University of Hamburg, Hamburg, Germany; 4. Genetics, UAB, Birmingham, AL, USA; 5. Gregory Fleming James CF Research Center, UAB, Birmingham, AL, USA An emerging question relevant to CFTR pathogenesis involves dynamics of mRNA utilization, translational velocity, and consequent influence on protein folding. Increasing evidence places the ribosome as an integral contributor not only to peptide primary structure, but also as a mediator of final CFTR conformation. Utilizing Fischer rat thyroid (FRT), CF human bronchial epithelia (CFBE), and primary human bronchial epithelia, we have previously shown that siRNA-mediated suppression of ribosomal protein L12 (Rpl12) results in robust enhancement of F508del-CFTR protein stability, maturation efficiency, plasma membrane localization, open channel probability, and transepithelial ion transport. Through use of leading-edge ribosome profiling and "omics" technologies, we found that Rpl12 depletion induces significant slowing of translation initiation and elongation rate, primarily due to increased dwelling occupancy of GC-rich codons within the ribosomal A-site. Rpl12 is located at the base of the 60S P stalk -a structure known to interact with elongation factors -and resides within the GTPase-associated center, which serves as interface for all GTP-bound translation components (e.g. initiation, elongation, and termination factors). We hypothesized that the P stalk may play a role in folding, stability, and/or translational fidelity, and therefore investigated whether Rpl12 silencing might improve other CFTR variants exhibiting defects in processing (G85E, P67L), conductance (A455E), or premature termination (W1282X). Biochemical (Western blot) and functional (Ussing chamber) analyses established that Rpl12 knockdown exerted no effect on P67L maturation, but A455E-, W1282X-, and G85E-CFTR were all partially rescued. Because W1282X correction was of particular interest, we also tested for evidence of translational read-through by assays in FRT cells encoding a C-terminal nano-luciferase cassette cloned downstream (and in-frame) with the complete W1282X CFTR sequence. Our results indicate Rpl12 repression augments nonsense codon read-through of W1282X by several fold, and enhances steady-state levels of W1282X bands B and C. Under Rpl12-depleted conditions, short-circuit current in primary human nasal epithelia with CFTR W1282X/W1282X genotype showed W1282X transepithelial ion transport increased from undetectable values to ~16% of wild-type. Taken together, these results indicate Rpl12 silencing may represent a novel therapeutic strategy for addressing multiple CFTR defects and disease subcategories. For premature termination variants and other difficult-to-treat alleles (e.g. splice mutations), it is unlikely that high-throughput compound library screening will identify an agent that specifically binds CFTR. Findings described here suggest that ribosomal proteins, and Rpl12 in particular, should be evaluated as primary targets for discovery of novel small molecules that induce translational read-through. Supported by CFF (OLIVER17F0; IGNATO17XXO; SORSCH14XXO) and NIH (R01HL136414).

The cystic fibrosis transmembrane conductance regulator (CFTR) is a member of the ATP-binding cassette family of transport proteins, and, in its membrane-bound state, it facilitates the transport of Clions across the apical cell membrane of a variety of cell types (1) . Mutations in CFTR cause cystic fibrosis (CF), a genetic disease in which pathological mucus accumulation leads to chronic lung infections and a life-threatening deterioration of lung function over time. In order to evaluate the effect on CFTR levels after either drug or other treatment methods on a cellular level, a new tool was designed. We developed a polyclonal antibody capable of detecting the extracellular loop 1 (ECL1) of CFTR without the need of membrane permeabilization. Using this approach in combination with imaging flow cytometry enables us to quantify on a single cell level the effect of cyclic compressive stress (CCS) on pseudostratified primary human bronchial epithelial cells (HBE). We are able to quantify and localize CFTR based on a virtual longitudinal cut of ciliated cells and clearly show the increase of plasma membrane CFTR upon CCS treatment or 30°C incubation compared to the nontreated ΔF508 cells.

Methods: To achieve high epitope specificity, we departed from traditional peptide-based antibody design techniques by engineering a CFTR-chimera antigen that deliberately mimics the native conformation of the target, ECL-1 of CFTR. Trapping the ECL-1 sequence between two adjacent alpha-helices limits the random orientations of the loop, thereby increasing the recognition of a structural specific epitope. In this case study, we examined the effect of CCS treatment on ΔF508 HBEs compared to no treatment.

Conclusion: Our preliminary data support the hypothesis that application of our novel antibody in live-cell surface-staining methods can be used to quantify membrane levels of CFTR in human-and mouse-derived primary cells. We envision that this antibody will be of immense value for detecting membrane-bound CFTR in high-throughput drug screening campaigns, particularly when investigator interest is directed towards CFTR read-through modulators' prevention of premature termination of the transcript (2) . Our results might provide a molecular mechanism by which exercise and/or airway clearance could benefit the CF airway epithelium. Chemistry, Bar-Ilan University, Ramat-Gan, Israel The dynamics of wild-type (WT)-CFTR at the atomic-scale as well as how it is affected by CF-causing mutations remains largely elusive even after the publication of several cryo-EM structures of the full length protein, primarily due to two reasons: (1) Experimental structures represent only snapshots of the highly dynamic and allosteric CFTR. Thus, cryo-EM, at least as it stands now is not the tool for studying protein dynamics. (2) No structures of CFTR with CF-causing mutations have been published to date. However, CFTR dynamics is critical to our understanding of the mechanism of action of CF-causing mutations and to our ability to counteract their effects through mutation-specific therapies. On the other hand, the published structures provide excellent starting points for molecular dynamics (MD) simulations which could be utilized to study CFTR's structure, energetics and dynamics under near physiological conditions. Five cryo-EM structures of CFTR were published to date: Nonphosphorylated, inward facing zebrafish (PDB code 5UAR, resolution 3.73Å) and human (5UAK, resolution 3.87Å) CFTR which are nearly identical; phosphorylated, outward facing zebrafish CFTR (5W81, resolution 3.37Å); thermostabilized dephosphorylated and phosphorylated chicken CFTR (resolution 4.3Å and 6.6Å, respectively). Here we used the 5UAK and 5W81 structures as template for generating models of WT human CFTR and of several CF-causing mutants including P67L, G551D, and N1303K. These models were subsequently used as starting points for MD simulations.

Analysis of the resulting trajectories demonstrates that MD simulations can provide insight into the effect of CF-causing mutations on the structure and dynamics of CFTR in agreement with experimental findings. We have previously found a favorable correlation between thermal stability and the fluctuation profiles for several NBD1 constructs. This correlation extends to the full length protein. Thus, WT and G551D-CFTR have similar RMSF profiles indicating similar stabilities. This finding is in agreement with experiment. We also find that P67L leads to enhanced fluctuations in NBD2 relative to the WT protein. Thus, this processing mutation likely reduces the stability of CFTR by destabilizing NBD2. P67L could be rescued by R555K. Indeed, the fluctuations of P67L/R555K-CFTR are even lower than those of WT-CFTR. Based on this correlation we predict N1303K and WT-CFTR to have similar stabilities. Next we demonstrate that for G551D-CFTR the distance between the NBDs in the outward facing conformation is lower than that in WT-CFTR, indicating looser packing of these domains. This is in agreement with the impaired gating caused by the mutation.

A similar yet smaller effect is seen for P67L-CFTR. Indeed this mutation is treatable by ivacaftor. Focusing on G551D, we observe narrowing of the inner pore due to the movement of helix 8 and partial blockage of the extracellular exit point by ECL1. In addition, we observe that an important network of inter-residue interactions in WT-CFTR is lost in the mutant. Finally, we show how MD trajectories can suggest potential binding sites for CFTR modulators, and how such modulators could be docked into these sites. We posit that the hypotheses put forth by the computational analyses could be experimentally validated.

Ubiquitylation regulates various cellular processes, including endoplasmic-reticulum-associated protein degradation (ERAD), membrane protein trafficking, cell signaling, etc. As a substrate for ERAD, CFTR biogenesis may be facilitated by deubiquitinating enzymes (DUBs) that might delay its disposal in the ER. It has been reported that USP19, an ER membrane-anchored DUB, is a target of the unfold protein response and is involved in the turnover of ERAD substrates (1) . In this study, we explored the potential functions of USP19 in modulating CFTR biogenesis vs. degradation. By screening several DUBs including USP5, USP7, USP11, USP13, USP19, and USP28, we found that USP19 specifically recognized the common CFTR mutant, and that its over-expression robustly elevated the levels of immature CFTR and markedly reduced F508del CFTR degradation. Knockdown of USP19 significantly reduced F508del CFTR protein expression in CFBE cells; conversely, whole-cell patch-clamp detected a cAMP-stimulated Clcurrent in CFBE cells expressing F508del CFTR and USP19. In addition, USP19 decreased F508del CFTR ubiquitination, and this deubiquitinase activity required the catalytically activity of this DUB. Interestingly, USP19 interacted with the small ubiquitin-like modifier, SUMO-1, in CFBE cells and promoted the conjugation of F508del CFTR to SUMO-1, which can increase protein solubility and reduce protein aggregation (2) . We have previously reported that SUMO paralogs can modify CFTR protein both in vitro and in vivo (3) . To explore the role of SUMO modification in USP19-mediated stabilization of mutant CFTR, we evaluated the deubiquitinase activity of USP19 using purified NBD1-F508del protein in vitro. We observed that USP19 did not directly promote deubiquitylation of NBD1-F508del; rather, it promoted the deubiquitylation of NBD1-F508del pre-conjugated to SUMO-2. This SUMO paralog forms SUMO poly-chains that are recognized by the SUMO-targeted ubiquitin ligase, RNF4, which promotes mutant CFTR degradation by the proteasome (3) . Moreover, mutation of SUMO-interaction motifs (SIMs) in USP19 prevented its ability to increase F508del CFTR expression and diminished the interaction of USP19 with F508del CFTR in CFBE cells.

These results indicate that USP19-induced F508del CFTR stabilization and functional rescue occurred through a USP19-induced reduction in ubiquitylation of poly-SUMO chain modified F508del CFTR, while also increasing its SUMO-1 conjugation. These results suggest that the modification of mutant CFTR by pathways that utilize different SUMO paralogs play competing roles that determine the fates of wild-type and mutant CFTR. They suggest also that SUMO paralogs themselves can recapitulate the actions of the SUMO-2/3/RNF4 pathway for ubiquitin-proteasome degradation vs. the SUMO-1 pathway, which supports CFTR biogenesis. Modulation of these pathways may provide new therapeutic approaches for the rescue of ERAD-targeted misfolded proteins. (Supported by grants from the NIH (DK 068196) and from CF Foundation Therapeutics. The new challenge for cystic fibrosis (CF) therapy is based on the development of small molecules able to rescue the function of CFTR. Many pharmacological agents have been designed to increase the surface level of mutated CFTR (correctors), as well as its plasma membrane (PM) activity (potentiators). Unfortunately, for the most common CF-causing mutation F508del, their efficacy seems to be time-limited mainly due to a reduced stability of corrected protein (Cholon DM, et al. Sci Transl Med. 2014; 6:246ra96; Veit G, et al. Sci Transl Med. 2014; 6:246ra97; Matthes E, et al. Brit J Pharmacol. 2016; 173:459-70) . Indeed, many factors contribute to PM CFTR stability, including its compartmentalization in PM macromolecular complex composed of phospholipids, sphingolipids, with particular regards for monosialoganglioside 1 (GM1) , and scaffolding proteins such as ezrin and NHERF1 (Monterisi S, et al. J Cell Sci. 2012; 125:1106-17) .

Interestingly, it has been proved that in bronchial epithelial cells the lack of CFTR in the cell PM, such as in the case of the patients carrying the mutation F508del, is associated with a decreased content of GM1 (Itokazu Y, et al. Am J Physiol Cell Physiol. 2014; 306:C819-30) . On the basis of these findings, we investigated the effects of the potentiator ivacaftor (VX-770), the corrector lumacaftor (VX-809) and of GM1 on PM stability and function of CFTR.

Methods and Results: By the use of a radioactive and photoactivatable GM1 derivative, we demonstrated that in bronchial epithelial cells GM1 is closed associated with the wild-type form of CFTR. Then, we analysed CFTR expression upon treatment of CF bronchial epithelial cells individually with VX-809, or VX-770 or combination of both drugs, in presence or not of GM1. In CF cells, the treatment with the corrector VX-809 induces an increase in the mature form of CFTR, an effect that is reversed by combined treatment with the potentiator VX-770. Interestingly, when GM1 is exogenously administered to these cells, the content of the mature form of CFTR remains high even after the combined treatment with corrector and potentiator. Consequently, we analysed the effect of VX-809, VX-770 and GM1 on CFTR scaffolding proteins NHERF1 and ezrin. We found that the expression of both proteins increased in cells treated with GM1 and subjected to double pharmacological treatment with respect to cells treated with corrector and potentiator but without GM1.

Finally, we analysed the effect of treatment with GM1 on CFTR function by performing functional fluorescence assays using live microscopy imaging. We observed that treatment with GM1, in combination with VX-770 and VX-809, increases CFTR function with respect to treatment with only corrector and potentiator.

Conclusion: In conclusion, these results support the role of GM1 in the stabilization and function of CFTR at the PM level, suggesting new therapeutic strategy for the treatment of CF by the use of GM1. The most common CF-associated mutation, F508del causes defects at the level of CFTR mRNA, protein, and channel function. In addition, TGF-β1 blocks the corrector VX-809-mediated functional rescue of F508del-CFTR in primary differentiated human bronchial epithelial (HBE) cells homozygous for F508del. TGF-β1 acts upstream of corrector VX-809 because it represses CFTR mRNA by a mechanism dependent on the TGF-β receptor (TβR)-I and autonomous of epithelial cell de-differentiation. Thus, rescue of the distal F508del-associated defects, at the level of CFTR protein, without correcting the proximal, mRNA repression of F508del-CFTR may limit the in vivo efficacy of corrector/potentiator therapy. While miRNA-145 antagonism was shown to reverse the TGF-β1 mediated rescue of F508del-CFTR, little is known about the mechanisms of TGF-β1 repression of CFTR mRNA. Our previously published data in HBE cells show that the degree of CFTR mRNA inhibition by TGF-β1varied greatly among non-CF controls that included cells from lungs with no disease and COPD. Acquired CFTR dysfunction and elevated TGF-β1 levels were previously reported in COPD. Here, we studied separately HBE cells from CF and COPD lungs procured at the time of transplantation and control lungs with no disease procured from the Center of Organ Recovery and Education (N= 13/group). HBE cells were cultured to full differentiation and TGF-β1 (15 ng/mL) or vehicle control was added to the basolateral medium for 24 hours. While the baseline levels of CFTR mRNA did not differ between the groups, TGF-β1 inhibited CFTR mRNA in cells from COPD lungs similar to CF but had no effect on CFTR in control lungs. These data demonstrate that the disease context modulates TGF-β1effects on CFTR. Work published by several groups including ours, showed that TGF-β1 blocked rescue of F508del-CFTR by corrector VX-809. Recent data demonstrate that the combined use of corrector C18 and CFFT-002 had a superior effect on F508del-CFTR rescue in vitro, compared to a single administration or other small molecules. TGF-β1 blocked the F508del-CF-TR-mediated Cltransport and inhibited ASL volume rescued by correctors C18 and CFFT-002 in CF HBE cells. Together, these data demonstrate that TGF-β1 acts upstream of different small molecule correctors. Next, we examined whether signaling via the canonical TGF-β1 pathway mediates CFTR mRNA repression in CF HBE cells. At baseline, the shRNA mediated depletion of Smad2 (shSmad2) or Smad3 (shSmad3) did not affect CFTR mRNA, compared to the nonsilencing shRNA control. By contrast, shSmad3 accelerated TGF-β1 repression of CFTR mRNA. To study whether the effect was mediated by miRNA-145, we examined the miRNA-145 levels after Smad2 or Smad3 depletion. shSmad3 had a profound inhibitory effect on miRNA-145 at baseline and prevented the TGF-β1-induced upregulation of miRNA-145. These results demonstrate that in CF HBE cells, Smad3 mediates TGF-β1 repression of CFTR mRNA, independent of miRNA-145. In summary, the effects of TGF-β1 on CFTR in HBE cells are modified by the disease background. In CF HBE cells, TGF-β1 inhibits CFTR function rescued by different correctors by a novel mechanism dependent on the canonical mediator, Smad3.

There are over 2,000 mutations reported in the gene encoding cystic fibrosis transmembrane conductance regulator (CFTR), most of these cause CF. Interestingly, only a minority of these mutations are found on the luminal/extracellular face. Our group has demonstrated that a novel chaperone of the ER lumen, endoplasmic reticulum protein of 29kDa (ERp29), has increased expression in response to treatment with 4-phenylbutyrate, promotes the normal biogenesis of wild-type (WT) CFTR and corrects aberrant biogenesis of F508del when overexpressed (Suaud L, et al. J Biol Chem. 2011; 286:21239-53) . While ERp29 must interact with CFTR or F508del on the ER-luminal face of the protein, F508 is located on CFTR's cytoplasmic face within its first nucleotide binding domain; cellular trafficking machinery, including coat complex II (COP II) components that mediate CFTR's exit from the ER and transport to the Golgi, also interact with CFTR's cytoplasmic face. We hypothesize that more rare mutations on CFTR's luminal face may cause abnormal channel biogenesis. ERp29 is suggested to interact with -(F, Y)-X-(F, Y)-or -(F, Y)-(F, Y)-motif on client proteins, and CFTR has exactly one such putative ERp29 binding motif on its luminal face that is potentially accessible to a chaperone. This motif, 1014 Y-I-F 1016 , is located in extracellular loop 5 (ECL5), and there are two rare CF-causing mutations of CFTR within this motif, Y1014C and F1016S. Additionally, there are two described disease-causing CFTR mutations at proline 1013 (P1013H and P1013L) that are adjacent to this motif and are likely to cause a change in structure of ECL5, the region including and surrounding 1014 Y-I-F 1016 motif. To test the hypothesis that these mutations in ECL5 inhibit CFTR biogenesis, we expressed these mutant CFTRs in CFBE41o − CF bronchial epithelial cells. Functional expression of these mutant CFTR, defined as I sc that was inhibited by apical application of 10 µM CFTRinh-172 after treatment of the cells with 10 µM forskolin and 100 µM IBMX and imposition of the basolateral-to-apical chloride gradient in Ussing chambers, was absent. In addition, immunoblots of whole cell lysates of CFBE41o-cells transfected with these mutants demonstrated CFTRs that co-migrated with F508del CFTR and at a lower molecular weight than WT CFTR. These data support the hypothesis that disease-causing CFTR mutations in ECL5 and on CFTR's luminal face inhibit CFTR biogenesis. This work was supported by grants to RCR from the NIH (R01 HL135670) and CFF (RUBENS16G0).

New high-throughput assay formats and innovative screening technologies may allow miniaturized CFTR modulator screens using small quantities of near-native, patient-derived airway epithelial cells. We developed a hollow micropillar array method to screen compounds using epithelial cells cultured on a porous support, with the goal of screening thousands of compounds using a single, commercially available 24-mm diameter Transwell filter containing cultured cells. Test compounds (~1 nL) in an alginate hydrogel were microprinted in hollow cylindrical micropillars (height 150 µm, inner diameter 100 µm) spaced 300 µm apart in a square array configuration (see Figure) . Compounds were delivered by positioning the array near the surface of a cell layer, with ~10 µm distance between the micropillars and cell surface. Compounds were stable in the hydrogel, and efficiently released upon contact with an overlying aqueous solution. Micropillar array geometry, and the viscosity of the hydrogel and overlying solutions, were optimized computationally and experimentally for sustained exposure of cells to test compounds with minimal cross-talk from compounds in neighboring micropillar wells. For proof of concept, a hollow polydimethylsiloxane micropillar array was fabricated to print 100 compounds (10 x 10 micropillar array) in a 3 × 3 square mm area, in which an ~23 µm-high rectangular spacer was fabricated to set a ~10 µm gap between the upper surface of micropillars and the cell layer. CFTR activation by cAMP agonists was measured in CFTR-expressing epithelial cells using a YFP halide indicator. Analysis of iodide quenching curves showed robust detection of CFTR activation with near-zero cross-talk, with a statistical Z'-factor of >0.65. The hollow micropillar array platform developed here should be generalizable to cell monolayers of any type grown on solid or porous supports, and to kinetic or steady-state read-out of an optical signal for enzymes, transporters, receptors or other targets. This approach may be useful for drug screening utilizing small quantity of cells, as in personalized n-of-1 screens using CF patient-derived primary cells. Supported by NIH, CFF and Emily's Entourage.

Treating the "remaining 10%" is a high priority in CF drug development, as is improving the outcome of therapy for the 90% of patients that may soon be treatable with approved and current investigational CFTR modulators. Two of the five most common CF-causing CFTR mutants world-wide are the nonsense mutation W1282X and missense mutation N1303K. We report here that, in both transfected cell models and primary human airway epithelial cell cultures, combination of a corrector and two potentiators increases chloride current from W1282X-CFTR and N1303K-CFTR by up to 10-fold over that of VX-770 alone. High-throughput "synergy screens" done on N1303K and W1282X CFTR, in which >100,000 candidate drugs were tested in combination with VX-770, identified arylsulfonamide-pyrrolopyridine, phenoxy-benzimidazole and flavone co-potentiators. An arylsulfonamide-pyrrolopyridine co-potentiator (ASP-11) added with VX-770 increased N1303K-CFTR current by 7-fold more than VX-770 alone. ASP-11 with VX-770 increased by ~65% current of G551D-CFTR compared to VX-770 alone, was additive with VX-770 on F508del-CFTR, and activated wild-type CFTR in the absence of a cAMP agonist. ASP-11 efficacy with VX-770 was demonstrated in primary CF human airway cell cultures having N1303K, W1282X and G551D CFTR mutations (see Figure) . Structure-activity studies on synthesized ASP-11 analogs produced co-potentiators with EC 50 down to 500 nM. A remarkable finding was the apparent broad selectivity of co-potentiators for different mutant CFTRs. The co-potentiator approach is being tested on less common missense and nonsense CFTR mutations found in NBD1 and NBD2, including L467P, V520F, A559T, R560T, R1158X, R1162X, 3659delC, W1204X, I1234V and Q1313X. These studies support combination potentiator therapy for CF caused by some CFTR mutations that are not treated effectively by single potentiators. Supported by NIH, CFF, CFRI and Emily's Entourage. Approximately 50% of CF patients are heterozygous with a rare mutation on at least one CFTR allele. It is anticipated that approved and "next-generation" CFTR modulators will have therapeutic benefit for most (~90 %) but not all CF patients. We report a personalized CFTR modulator discovery effort involving two siblings that are heterozygous for W1282X and I1234V CFTR mutations. The W1282X mutation, which prematurely truncates CFTR to remove ∼60% of nucleotide binding domain 2 (NBD2), is most commonly found in the Askhenazi Jewish population. The I1234V mutation is a relatively rare CF-causing variant found mainly in subjects of Middle Eastern origin. This mutation, a base pair substitution, c.3700A>G, produces the missense mutant I1234V-CFTR as well as introducing a splice site that produces a 6 amino acid deletion in NBD2 (p.Ile1234_ Arg1239del)). Human nasal epithelial (HNE) cells were obtained from nasal brushings from both subjects and expanded under conditional reprogramming conditions for testing CFTR modulators. Short-circuit current measurements showed no beneficial effect of the approved correctors VX-809 and VX-661. The potentiator VX-770 increased short-circuit current following addition of forksolin by 1.0 ± 0.1 µA/cm 2 . Remarkably, short-circuit current was increased by "co-potentiator" ASP-11 or the flavone apigenin, in the presence of VX-770, to 1.7 ± 0.1 µA/cm 2 and 1.6 ± 0.3 µA/cm 2 , respectively. Comparison of short-circuit current values of similarly cultured non-CF cells (~12-15 µA/cm 2 ) and VX-809-corrected F508del-CFTR cells (~3-5 µA/cm 2 ), suggests that co-potentiator therapy in I1234V/W1282X CF subjects restores chloride conductance to ~11-14% of wild-type CFTR activity, which may have therapeutic benefit. To discover CFTR modulators specific for the I1234V mutation, we have generated transfected FRT cells expressing I1234V-CFTR and Ile1234_Arg1239del-CFTR for high-throughput screening. Supported by NIH, CFF, CFRI.

Short-circuit current in I1234V/W1282X primary human airway epithelial cell cultures in response to 10 µM amiloride, 20 µM forskolin, 5 µM VX-770, 20 µM ASP-11 (left) or apigenin (right), 10 µM CFTR inh -172 and 100 µM ATP.

Introduction: Protein repair therapy is revolutionizing CF management by targeting CFTR defects to improve lung function, enhance nutritional status and increase survival. Ivacaftor (Iva) is a CFTR potentiator that increases CFTR activity at the cell surface. Initial clinical trials established Iva as both efficacious and safe. These trials, along with in vitro studies, led to FDA approval for Iva in CF patients as young as age two with 38 different genotypes. Despite broader medication availability, data beyond 24 weeks in pediatric CF patients is limited, particularly in less common mutations. This retrospective chart review assessed data in pediatric CF patients by mutation class within the year before and after Iva initiation.

Methods: An IRB-approved retrospective chart review of 4 Nemours CF centers identified 27 patients aged 2-18 years with FDA-approved mutations for Iva. Primary outcomes included sweat chloride levels, lung function and nutritional parameters at an average of 10 months, intravenous (IV) antibiotic use and respiratory tract microbiology over 1 year on therapy. Simple, descriptive and comparative statistics were employed for data analysis.

Results: Of 27 patients identified, one was excluded due to selfreported nonadherence. The remaining 26 patients, 11 males (42%), had a mean age of 9.5 years. Fifteen patients had Class III CFTR mutations (58%), 5 patients Class IV mutations (19%) and 6 patients Class V mutations (23%). Overall cohort analysis on therapy showed statistically significant improvement in sweat chloride levels (n=14), with the average starting chloride of 84.3 reduced to 38.4 over approximately 2 months (p =0.00). FEV1 percent predicted (n=21) improved by 11% in 4 months (p=0.05). IV antibiotic use (n=26) declined by 85% over 1 year (p =0.03). BMI increased, however, it was not statistically significant. The proportion of patients growing gram-negative bacteria decreased from 46% to 19% whereas MRSA persisted in 80% of patients.

Data analysis by mutation class is limited by small patient numbers. Patients with Class III mutations (n=15) and Class IV mutations (n=5), normalized repeated sweat chloride values (n=8). FEV1 and BMI improved markedly over 10-12 months of treatment. No patient with a Class III or IV mutation required IV antibiotics in the year following initiation. In contrast, patients with Class V mutations (n=6) did not normalize sweat chloride values (n=3), had marginal improvement in FEV1 over 2 months and decline in BMI over 6 months. There was no change in IV antibiotic use.

Conclusions: CF patients treated with ivacaftor had significant improvement in overall clinical outcomes. However, important variations were evident in each mutation class. Preliminary data suggests CF patients with Class V CFTR mutations may not have a robust response to Iva monotherapy in short term follow-up. This potential variability in genotypic response requires further study to determine which CFTR protein repair therapy is optimal for each patient based on class effect.

Pancreas-on-a-chip was fabricated using photolithography and used to co-culture PDECs on the top chamber and islet cells on the bottom chamber. Insulin secretion from islet cells was monitored during stimulation or inhibition of PDECs to study cell-cell interaction.

Results: Using RNAseq analysis and immunofluorescence microscopy we verified isolated PDECs were primarily epithelial in origin. Isolated islet cells were all positive for dithizone staining. The monitoring of CFTR function of PDECs requires over 1x10 5 cells on filter using short-circuit current assay. However, for pancreas-on-a-chip, we successfully monitored CFTR function with less than 10,000 cells, using an iodide efflux assay. With use of only 10 to 15 islets in the chip, we were able to detect insulin secretion. We observed that the amount of secreted insulin from islet cells decreased by 40% with closing of CFTR channels of PDECs on the top chamber using the CFTR inhibitor.

Conclusion: We have developed pancreas-on-a-chip utilizing cultured patient-derived PDECs and islet cells. Moreover using the chip, CFTR function and insulin secretion were monitored with high sensitivity from a small number of cells. We have successfully co-cultured PDECs and islet cells on the pancreas-on-a-chip and observed an effect of CFTR channel function on the insulin secretion of islet cells. Elucidation of this relationship will help further our understanding of CFRD. cysteine instead of the usual C-X-X-C thioredoxin motif. We have previously demonstrated that ERp29 promotes CFTR biogenesis (Suaud L, et al. J Biol Chem. 2011; 286:21239-53) , and we subsequently demonstrated that it also regulates ENaC biogenesis and functional expression (Grumbach Y, et al. Am J Physiol Cell Physiol. 2014; 307:C701-9) . In this work, we found that overexpression of wt ERp29 increased the abundance of the active form of γ-ENaC, as well as ENaC functional expression whereas ERp29 overexpression of a mutant ERp29 lacking its single Cysteine (C157S ERp29) decreased ENaC functional expression. These observations were not associated with altered expression of β-ENaC at the apical surface, suggesting that ERp29 may modulate ENaC open probability at the apical surface. ERp29 overexpression also promoted the interaction of both ENaC and CFTR with the coat complex II (COP II) ER exit machinery, whereas C157S ERp29 overexpression decreased this interaction. These data suggested a model where ERp29 may promote ENaC cleavage by directing ENaC to the Golgi.

ERp29's C-terminal ER retention motif is KEEL, a KDEL variant that is associated with less robust ER retention. To test whether this motif is critical for ERp29's regulation of ENaC and CFTR, we designed a mutant ERp29 containing a KDEL retention motif (ERp29 KDEL) that should be better returned to the ER from the proximal Golgi by the KDEL receptor, and a mutant that deleted the KEEL motif that would interact less well with the KDEL receptor (ERp29 ΔKEEL). ENaC functional expression was decreased by expression of ERp29 ΔKEEL, while expression of ERp29 KDEL did not elicit any significant effect. As already observed with wt and C157S ERp29 overexpression, β-ENaC expression at the apical surface was not changed by expression of ERp29 KDEL or ΔKEEL. We then further investigated CFTR activity and observed that both ERp29 KDEL and ΔKEEL decreased CFTR functional expression. These data suggest that ERp29's KEEL motif plays a critical role in ENaC and CFTR biogenesis. We therefore tested the hypothesis that this was due to the interaction of this motif with the KDEL receptor, and have preliminarily found that depletion of the KDEL receptor also decreased ENaC processing in the Golgi. Together our findings suggest a key regulatory role for ERp29's association with the KDEL receptor in supporting the biogenesis of CFTR and ENaC, and therefore suggest a role for the KDEL receptor in promoting ENaC biogenesis.

F508del, the most common disease-causing mutation, is present in approximately 85% of CF patients and leads to CFTR misfolding that results in ER retention and early degradation. CFTR folding status is assessed by the endoplasmic reticulum (ER) quality control (ERQC) that comprises a series of checkpoints involving chaperones and trafficking factors. CFTR exit from the ER is mediated by specific sorting motifs that include the 4 retention motifs AFTs (arginine-framed tripeptides, with the consensus sequence RXR) and the diacidic (DAD) exit code that controls the interaction with the COPII machinery. The interactions and regulatory pathways involved in this process regulating these steps remain however largely unknown.

The main goal of this work is to identify traffic factors that regulate CFTR exit from the ER at these specific quality control checkpoints.

We performed co-immunoprecipitation using CFBE cells stably expressing F508del-4RK-CFTR (in which the AFT motifs were abrogated by replacement of one of the arginines by a lysine residue in each motif) or F508del-CFTR (in which AFT motifs are putatively exposed, mediating CFTR retention) and used proteomic profiling and global bioinformatic analysis to identify factors that interact differentially with the two CFTR variants as putative regulators of this specific ERQC checkpoint.

A total of 834 potential CFTR-interacting proteins (CIPs) was identified, among which 198 proteins appear to have higher affinity to F508del-CFTR, as determined by higher abundance in immunoprecipitated proteins. In this subset, we identified an enrichment in proteins involved in RNA processing and complex organization and a decrease in those related to epithelial integrity when compared to the interactome of F508del-CFTR with abrogated AFT motifs.

We selected several putative hits and found 4 proteins involved in the regulation of F508del-CFTR retention and degradation. Among this, we identified kinesin family member C1 (KIFC1) as a stronger interactor with F508del-CFTR versus F508del-4RK-CFTR. We further validated this interaction showing that decreasing KIFC1 levels or activity stabilizes the immature form of F508del-CFTR by reducing its degradation leading to the appearance of the mature form.

The current approach is thus able to identify novel putative therapeutic targets that can be ultimately used to the benefit of CF patients.

Work Cystic fibrosis (CF) usually results from the deletion of an amino acid in the cystic fibrosis transmembrane conductance regulator (CFTR). The 2015 approval of combination therapy with the corrector molecule lumacaftor and potentiator ivacaftor validated modulation of protein biosynthesis as an avenue for pharmacological intervention. These and other drugs now present in combination therapies improve F508del-CFTR folding, increase channel density, and maximize activity at the plasma membrane. Nevertheless, a significant amount of F508del-CFTR is still destroyed by the proteasome through a process known as endoplasmic reticulum associated degradation (ERAD). It has been hypothesized that suppressing this degradation could augment the efficacy of corrector drugs by increasing the foldable pool of F508del-CFTR substrate on which they can act. During ERAD, F508del-CFTR is selected by molecular chaperones, modified with a polyubiquitin chain, and degraded by the proteasome. Prior work indicated that proteasome inhibition fails to rescue F508del-CFTR and instead leads to the accumulation of insoluble, polyubiquitinated protein. More recent studies, however, have shown that disruption of more proximal steps in the ERAD pathway can increase the pool of salvageable CFTR as small-molecule inhibition of ubiquitin activating enzymes synergistically increased F508del-CFTR maturation and activity when combined with chemical correctors (Chung WJ, et al. PLoS ONE. 2016; 11:e0163615) . Here, we sought to precisely target the ubiquitination step in ERAD of mutant CFTR through inhibition of the E3 ubiquitin ligase CHIP, which targets CFTR during both ERAD and the retrieval of F508del-CFTR from the plasma membrane.

To this end, in silico methods revealed two unique sites within CHIP which were targeted as inhibitor pharmacophores. Cell free ubiquitination assays for in vitro CHIP activity identified putative first-in-class CHIP inhibitors. Of these, select hit compounds prevented CHIP ubiquitination of a model substrate with micromolar to millimolar efficacy. One tested compound blunts the ubiquitination of immunopurified F508del-CFTR in vitro. Cellular studies indicate that this molecule stabilizes F508del-CFTR in HEK cells transfected with CFTR bearing plasmid, and prolongs the cellular protein half-life of mutant CFTR. Finally, one putative CHIP inhibitor works synergistically with lumacaftor to prevent proteasomal degradation and enhance F508del-CFTR abundance and maturation. Future efforts are underway to optimize cell free and cellular assay systems, target other druggable sites of CHIP, improve compound efficacy, develop a structure activity relationship for the chemical series identified, and examine compounds in functional assays and in primary lung epithelial cells. These early biochemical studies indicate that suppression of proximal steps in ERAD by chemical inhibition of the CHIP E3 ligase may stabilize CFTR mutant proteins and serve to enhance activity of existing corrector and potentiator compounds to optimize drug therapy for cystic fibrosis.

This work was supported by NIH K08 HL126135 and Cystic Fibrosis Foundation grant BRODSK18G0. on the relationship of the PTC to downstream introns. Therefore, this aspect of the cellular management of regulation of CFTR mRNA harboring PTC mutations is likely lost when CFTR PTCs are encoded in the context of cDNAs that are typically used for modeling nonsense mutant CFTR pathophysiology.

Proteostasis Therapeutics, Inc.'s proprietary CFTR amplifiers confer increased immature CFTR protein levels and stabilized CFTR mRNA through a mechanism that is dependent on translation and independent of the mutation in the CFTR gene. Amplifiers require only the translated sequence of CFTR to confer an increase in CFTR protein and mRNA levels. Amplifiers are complementary to other CFTR modulators, providing additional substrate on which those modulators can exert their mechanistic benefit. This includes read-through agents such as the aminoglycoside G418, the in vitro efficacy of which is enhanced in combinations with amplifier in cell line and primary HBE cell models of PTC mutant CFTR.

Specifically, a combination of amplifier and G418 promotes the functional rescue of CFTR in G542X-derived cell lines and G542X homozygous patient-derived human bronchial epithelial (HBE) cells. Based on this data, we performed an initial high-throughput screen (HTS) in combination with an amplifier in a trafficking assay in a cDNA-based FRT (Fischer rat thyroid) model expressing G542X-CFTR-HRP. Similar to results obtained in an F508del-CFTR HTS, screening in the presence of amplifier enhanced the effectiveness of the assay (5-fold increase in assay window), thus increasing the number of hits. Moreover, screening in the presence of amplifier uncovered modulators that would not have been identified in a standard nonamplifier HTS.

The initial screen resulted in 6 different chemotypes that were progressed to profiling in the Ussing chamber assay in G542X homozygous HBE cells. We report that the low chloride transport activity of these cells could be enhanced by employing a chloride gradient and by adding a corrector in the Ussing chamber assay. All 6 chemotypes were shown to specifically increase endogenous CFTR mRNA in G542X-CFTR HBE cells but not in F508del-CFTR HBE cells.

We have initiated a second-generation PTC read-through HTS to expand upon this approach. Two novel screening strategies that incorporate a more physiologic context for the PTC mutant CFTR are being piloted, and will then be used for a 600K compound screen. Common features include screening on top of amplifier to enrich for specificity and sensitivity of the hits identified, and the inclusion of introns downstream of the PTC mutation to enable the involvement of NMD, allowing for the potential to identify NMD-modulating compounds. Cystic fibrosis (CF) is a fatal multisystem, autosomal recessive disorder caused by mutations in the cystic fibrosis transmembrane conductance regulator (CFTR) gene. CF mutations, of which the most common is F508del-CFTR, affect CFTR expression and function. Recently, long noncoding RNAs (lncRNAs) have gained prominence as pivotal regulators of various biological processes and misregulation of lncRNAs is associated with various diseases. Here we have identified lncRNAs that are aberrantly expressed in CF lung tissues and are likely to regulate CF disease phenotype.

We analyzed lncRNA expression profile in lung tissues, both airway and parenchyma, obtained from CF patients undergoing lung transplant. We have identified 636 differentially expressed lncRNAs in CF airway epithelium and 1974 differentially expressed lncRNAs in CF lung parenchyma compared to matched respective healthy controls (n=4 each group, fold change ≥ 2, p <0.05). 15 lncRNAs exhibit similar expression trend in CF lung airways as well as in CF parenchymal tissues. When all these lncRNAs are compared using a hierarchical cluster algorithm, the dendrogram clearly distinguishes CF tissues from matched controls. The majority of the differentially expressed lncRNAs belong to the class of intergenic lncRNAs. Gene ontology (GO) analyses indicate that the mis-expressed lncRNAs affect CF-relevant biological processes that are associated with signaling pathways; cellular components that include intracellular vesicles and membrane systems; and molecular functions including transcriptional activator activity.

Further validation of selected lncRNAs were performed in multiple CF cell lines. Amongst the top 10 up-regulated lncRNAs in CF airway tissues, uc.363+, uc.8-and ARHGAP19-SLIT1 exhibit a similar expression trend in CF cell lines compared to controls. Interestingly, miRNA-128a and miR-449a regulate the expression of two of the most abundant lncRNAs, uc.8-and ARHGAP19-SLIT1, respectively. Further functional studies will help in the understanding of associated mechanisms and will lead to novel therapeutic targets for CF.

Supported by Collaborative Health Initiative Research Program to RB (I80VP000012).

The identification of proteostasis regulators important for the processing of mutant CFTR, is one possible approach to reducing detrimental health in people with CF. Using a high-throughput functional assay to screen an siRNA library, we identified 37 proteins whose silencing results in a significant rescue of F508del-CFTR activity. The identified targets include proteins associated to F508del-CFTR degradation (like UBA52 and UBXN6) and sumoylation (i.e. UBE2I, UBA2), transcription factors (i.e. MLLT6, CHD4, TRIM24) and also proteins with unknown function like FAU.

Aim: Now our work is focused on: 1. the characterization, prioritization and dissection of the mechanism(s) of action by which downregulation of selected proteins leads to mutant CFTR rescue; 2. the evaluation of global changes induced by target downregulation on cell morphology and functions.

Methods: The effects of silencing/overexpression of specific targets have been evaluated on rescue of mutant CFTR processing and concurrent possible pleiotropic effects by means of biochemical and electrophysiological techniques and by high-content imaging and analysis. The results indicate that, while some of the targets appear particularly promising and specific for CFTR (like FAU), others can be associated with deleterious effects due to dysregulation of specific cell pathways and/or with undesired effects on ion transepithelial transport.

Conclusions: We have identified a panel of proteins that are important for the processing of mutant CFTR, paving the way for using these proteins, and possibly their pathways, as new therapeutic targets to rescue mutant CFTR function.

Cystic fibrosis-causing defects in the chloride channel CFTR can be repaired by developing novel and better correctors (to improve folding) and potentiators (to improve function). Galapagos-AbbVie identified C2 correctors by high-throughput compound screening and Med Chem optimization for cell-surface rescue of F508del-CFTR. These C2 correctors act synergistically with a type I corrector such as C1. C2a corrector was optimized for drug like properties and is in clinical evaluation.

Using a radiolabeling pulse-chase approach in combination with protease susceptibility we track folding and assembly of each CFTR domain with time, and determine when and where correctors work on the newly synthesized protein. To study the effect of C2a on wild-type and F508del CFTR folding, we used unique domain-specific antibodies that allow visualization of folding intermediates in each domain of CFTR. Although the corrector improved exit of F508del CFTR from the ER, the primary folding defect in NBD1 was not corrected. The early conformation of each domain in F508del CFTR (directly after synthesis) did not change upon C2a addition, but after 2 hours the assembly of the TMD1, TMD2 and NBD2 domains had improved, which correlated with the increased amount of Golgi form of CFTR (C band).

To identify which domains of CFTR are influenced by the corrector, we used several approaches. An F508del CFTR construct that lacks NBD2 but does leave the ER (1219x) showed comparable improvement as fulllength F508del CFTR, implying that NBD2 is not required for the action of C2a on CFTR. Because we had excluded both NBDs as target for C2a, we set out to examine whether the corrector acts on the TMDs: we expressed each TMD separately in HEK293T cells and determined degradation kinetics in presence and absence of corrector. Neither TMD1 nor TMD2 showed any change in degradation rate, suggesting that (cell-biological) stability of these individual domains was not influenced by the corrector. Another F508del CFTR construct that lacks NBD2 but does not leave the ER (1202x), however was stabilized by C2a, implying that the corrector does not rescue domain folding itself, but rather plays a role in correcting the domain assembly of CFTR.

To pinpoint the mechanism by which the corrector rescues domain assembly, we are currently studying various designed and CF-patient mutants that are deficient in domain assembly. The preliminary results show that C2a not only efficiently corrects F508del CFTR, but several other CF-causing mutations as well. These results will provide new and important insights into identifying the mode of action of the C2a corrector.

Cystic fibrosis (CF) is the most common, lethal autosomal recessive disorder, and is caused by mutations in the cystic fibrosis transmembrane conductance regulator protein (CFTR), an anion channel that is found in most epithelial cells lining the airway and gut. The most common mutation of CFTR is deletion of phenylalanine at position 508 (CFTR ΔF508), which produces a misfolded protein. Through the ubiquitin proteasome system (UPS), this misfolded protein is ubiquitinated and signaled for degradation via the cytosolic proteasome. This route of degradation is referred to as the endoplasmic reticulum associated degradation (ERAD) system. Previous studies demonstrating experimental restoration of CFTR ΔF508 trafficking to the plasma membrane showed partial function of the chloride channel, raising therapeutic speculations. Some components of how CFTR ΔF508 is degraded are known, however many mechanisms that underly its degradation through ERAD and the UPS are still unknown. Through an siRNA library screen of 22 ER resident ubiquitin E3 ligases, we discovered a ubiquitin E3 ligase, ring finger protein 19B (RNF19B), that aids in the degradation of CFTR ΔF508. We used siRNA-mediated silencing of endogenous RNF19B in the CF epithelial cell line cystic fibrosis bronchial epithelial ΔF508 (CFBE ΔF508) cells to demonstrate that there is a 50% increase in CFTR ΔF508 expression in these cells. We also created an RNF19B plasmid that was co-expressed with CFTR ΔF508 in human embryonic kidney (HEK) 293 cells that demonstrated a 2-fold decrease in CFTR ΔF508 compared to control. This decrease in CFTR ΔF508 expression was also dose-dependent. Cycloheximide-chase (CHX) experiments using HEK 293 cells overexpressing RNF19B and CFTR ΔF508 showed that there was a decrease in CFTR ΔF508 half-life compared to control. Biotinylation experiments using siRNA-mediated silencing of endogenous RNF19B in CFBE ΔF508 cells that were treated with VX-809 demonstrated an increase in mature form of CFTR at the plasma membrane compared to control. A catalytically nonfunctioning RNF19B RING mutant was unable to promote CFTR ΔF508 degradation. Lastly, using siRNA-mediated silencing of endogenous RNF19B in CFBE ΔF508 cells increased forskolin-stimulated short-circuit currents compared to control. Interestingly, there seemed to be cooperatively larger increases in short-circuit currents when the CFBE ΔF508 cells that were silenced with RNF19B were treated for 24 hours with VX-809. Taken together, these data suggest that there is an additional mechanism where the E3 ubiquitin ligase RNF19b is responsible for the ubiquitination and subsequent proteasomal degradation of the misfolded CFTR ΔF508.

Veit, G. 1 ; Bidaud-Meynard, A. 1 ; Avramescu, R.G. 1 ; Schnúr, A. 1 ; Bilodeau, L. 2 ; Frenkiel, S. 3 ; Matouk, E. 4 ; Lukacs, G.L. 1 1. Physiology, McGill University, Montreal, QC, Canada; 2. Institut Universitaire de Cardiologie et de Pneumologie, Québec, QC, Canada; McGill University, Montreal, QC, Canada; 4. Montreal Chest Institute, McGill University, Montreal, QC, Canada Conditional reprogramming (CR) has gained wide application to enhance the amplification and life span of primary epithelial cells, while preserving their characteristics. In CF, this method was used to amplify human bronchial epithelia (HBE) as well as alveolar type II cells, derived from lung samples collected during transplantations, and human nasal epithelia (HNE) obtained from nasal scrapes or brushings. A recent study by Pranke et al. (Sci Rep. 2017; 7:7375) showed that correction of CFTR chloride channel activity in patient-derived CFTR F508del/F508del HNE correlates with the gain in lung function upon lumacaftor/ivacaftor treatment, thus suggesting the predictive value of the HNE culture.

We compared the morphological and electrophysiological characteristics of CR-HBE and CR-HNE. Upon differentiation, both formed pseudostratified epithelia expressing tight junction associated ZO-1, embedded with Muc5AC producing goblet cells and acetylated tubulin expressing ciliated cells, with an overall transepithelial resistance of ~500 Ω*cm 2 . To assess the electrophysiological characteristics of CR-HNE, cells from five individuals with CFTR WT/WT genotype were collected and upon differentiation, their CFTR function was determined by short-circuit current (I sc ) measurement and compared to that of CR-HBE. While the PKA activator forskolin-stimulated CFTR I sc was ~40% lower in CR-HNE than in CR-HBE, likely due to reduced CFTR expression in nasal epithelia, the cAMP sensitivity and fractional potentiator-independent current were similar between CR-HNE and CR-HBE. The functional phenotype of CR-HNE was stable for ~10 passages as indicated by the comparable amiloride-inhibited ENaC and PKA-stimulated, CFTR inh -172-sensitive current densities at passages 3, 6 and 10. Furthermore, CR-HBE and CR-HNE each isolated from 3 patients with CFTR F508del/F508del genotype exhibited similar correction upon treatment with the folding corrector VX-809, resulting in CFTR function corresponding to 24.1 ± 4.7% and 25.1 ± 1.5% of the matching wild-type epithelia currents.

Encouraged by these results, we collected nasal scrapes from ~80 CF patients with various mutations at three clinical sites in Quebec in a first phase. 1) The CR-HNE allow examination of the efficacy and potency of CFTR modulators alone and in combinations. Furthermore, studying cells isolated from multiple patients allows estimation of patient-to-patient variability of these parameters. 2) CR-HNE from CF patients with rare mutations facilitate the identification of mutation-specific therapeutic approaches (precision medicine). 3) By measuring the patient-specific response of CR-HNE to given modulator combinations their clinical response may be predicted. If more CFTR modulators become available, comparative analysis of their efficacy in CR-HNE may guide personalized therapeutic decisions. We will present examples for all three approaches in support of the concept that CR-HNE can serve as a bridge between basic science and clinical CF research.

Intensive research has been oriented to developing drugs that improve CFTR function, with the hope that, by correcting the basic defect, lung disease would be reversed and infection control would improve. Although with the use of lumacaftor/ivacaftor (LUMA/IVA), modest but significant improvement in pulmonary function and a reduction in pulmonary exacerbations occurs, there is no established link with changes in infections. To better understand whether increased CFTR activity would alter the nature of infections in CF patients, we carried out an observational study in a Northern Italian CF Centre. A total of 42 CF patients, 12 years of age or older, who were homozygous for the F508del CFTR mutation, received LUMA/IVA; we studied the 14 patients, who had been taking LUMA/IVA for at least 2 years (3 years in 8 subjects) with particular regard to their airways infections before and after treatment. Pulmonary exacerbations significantly decreased from baseline (from 4.29±0.54 to 2.43±0.40 after 1 year (p<0.05) and to 1.91±0.39 after 2 years (p<0.01)). Improvement after 3 years of treatment was not significant (from baseline 3.75±0.82 to 2.13±0.35). Pulmonary function (FEV 1 percent predicted) improved from baseline but not significantly (from 59.4±5.3 to 62.6.7±5.7 after 1 year and to 65.2±6.4 after 2 years, and from baseline 67.5±4.9 to 77.0±4.5 after 3 years). Body mass index increased from 21.02±0.51 at baseline to 21.83±0.63 after 1 year and to 21.86±0.56 after 2 years, and from 20.64±0.57 baseline to 22.33±0.76 after 3 years. With regard to sputum microbiology the majority of patients included in this study were infected by Staphylococcus aureus (MSSA) at the time of enrollment (64%). Additional species were also present with Achromobacter xyloxidans (36%), Pseudomonas aeruginosa (29%), Aspergillus fumigatus (21%), methicillin resistant (MR) SA (21%), and Haemophilus influenzae (14%) being the most abundant. After the first and second year of treatment, no significant change in microbiological isolation was observed: MSSA (50% after 1 year vs 50% after 2 years), P. aeruginosa (29% after 1 year vs 21% after 2 years), H. influenza Approximately 10% of disease-causing mutations in CFTR result in a nonsense mutation or premature termination codon (PTC), which results in a truncated and usually nonfunctional CFTR protein. During protein synthesis, these stop codons terminate protein translation as they enter the ribosome A site and interact with eukaryotic release factors 1 and 3 (eRF1/3). Sequence context, regulatory components, and small molecules can alter the balance of translation termination and translational readthrough, and therefore have implications for treating nonsense mutations. Understanding how translation termination occurs is therefore important for developing effective nonsense therapies. In this study, we used ribosome profiling to monitor the process of translation termination by obtaining a high-resolution and quantitative profile of actively translating ribosomes across the entire cellular transcriptome (~18,000 and ~16,000 transcripts in human and rat cells, respectively). Ribosome-protected mRNA fragments (footprints) were isolated and sequenced by next-generation sequencing from HEK293 and/or FRT cells prior to and after treatment with an aminoglycoside analog (G418) or a candidate PTC readthrough small molecule (CFF-910), which was identified in a high-throughput screen (Southern Research Institute). Our results revealed that translation termination at native stop codons occurs in two successive and distinct steps. The first occurred as the stop codon entered the ribosome A site, suggesting that the ribosome pauses as eRF1/3 interacts with the stop codon. The second pause occurred as the stop codon translocates from the A site to the P site of the ribosome, and then presumably triggers ribosome disassembly. G418 reduced ribosome occupancy at both pause sites and increased ribosomes present on the 3'UTR consistent with active translational readthrough. In contrast, CFF-910 showed a summative increase of ribosome occupancy only at the first pause site, and 3'UTR footprints lost periodicity consistent with a frame shift-associated readthrough mechanism. These results suggest that G418 and CFF-910 have distinct mechanisms of action on translation termination, with G418 allowing readthrough by near-cognate tRNAs, whereas CFF-910 prolongs ribosome pausing at the stop codon and enables readthrough with out-of-frame translation. Supported by CFF.

Millen, L. 1 ; Musisi, I. 2 ; Patel Thakerar, A. 3 ; Raraigh, K.S. 4 ; Oyem, P. 2 ; Allaire, N.E. 2 ; Benjamin, D. 2 ; Bihler, H. 2 ; Bridges, R.J. 3 ; Cutting, G. 4 ; Mense, M. 2 ; Sivachenko, A. 2 ; Thomas, P.J. 1, 2 1. Physiology Department, UT Southwestern, Dallas, TX, USA; 2. CFFT Lab, Cystic Fibrosis Foundation Theraputics, Lexington, MA, USA; 3. Physiology, RFUMS, Chicago, IL, USA; 4. Genetics, JHMI, Baltimore, MD, USA Hundreds of variants of the CFTR gene have been observed in people with CF. However, experimental evidence for pathogenicity is available for only a fraction of these variants. Moreover, there is a dearth of information regarding the molecular pathology and responsiveness to small molecule therapeutics for the vast majority of the variants. Recent FDA decisions for label extensions based on in vitro functional data highlight the potential value of more data on these poorly understood variants. To fill this gap, we initiated a program to assess 649 putative disease-associated CFTR missense variants, both functionally and biochemically, in a blinded fashion, at three independent sites. The results document those variants: i) likely to be disease-causing due to a lack of function, ii) likely to be benign polymorphisms, iii) that interfere with the folding and maturation of CFTR, iv) that respond to acute treatment with a CFTR potentiator (ivacaftor) and v) that respond to pretreatment with a CFTR corrector (lumacaftor). Our presentation will summarize the progress to date. fibrosis patients carry a PTC on at least one CFTR allele. Translation of a PTC-containing mRNA terminates at the PTC, producing a truncated, usually nonfunctional, polypeptide. PTC suppression therapy utilizes small molecules that suppress translation termination at a PTC to restore synthesis of a full-length polypeptide. PTC suppression, also termed readthrough, is mediated by the base pairing of a near-cognate aminoacyl-tRNA with a PTC and subsequently, its associated amino acid becomes incorporated into the nascent polypeptide at the site of the PTC. However, little is known about the identity of the amino acid(s) inserted at a PTC during readthrough, how the surrounding sequence context influences amino acid incorporation, why certain near-cognate aminoacyl-tRNAs are preferred, or the functionality of full-length CFTR protein generated by readthrough.

Using a readthrough reporter, we recently found that the amino acids inserted at the CFTR W1282X PTC (a UGA codon) in the context of three native CFTR upstream and downstream codons during G418-mediated suppression are leucine, cysteine, and tryptophan. However, we found the amino acids inserted at the CFTR G542X PTC (also a UGA codon) in the context of its native CFTR sequence are arginine, cysteine, and tryptophan. Not only do the amino acids inserted at each CFTR context differ, the proportions of the amino acids inserted also differ. Since the reporters were identical with the exception of the CFTR context, these results demonstrate that the local mRNA sequence context plays a role in near-cognate aminoacyl-tRNA selection during readthrough.

Some full-length CFTR proteins generated by readthrough, termed variant proteins, exhibit reduced maturation and activity, indicating the complexity of nonsense suppression therapy. However, both a CFTR corrector and potentiator enhanced the activity of variant proteins generated by G418-mediated readthrough. These results suggest that PTC suppression therapy in combination with CFTR modulators may be beneficial for the treatment of CF patients who carry a PTC.

We are currently investigating whether it is the mRNA context upstream or downstream of the PTC, or both, that determines which aminoacyl-tRNA is inserted at a PTC. We are also identifying whether the amino acids and their proportions change during readthrough mediated by pharmacological agents other than G418. Additionally, we are examining what amino acids are inserted at other CFTR nonsense mutation contexts in human cells, including those that have a UAG or UAA PTC.

Supported by funding from the CFF and the NIH.

Molecules correcting the trafficking (correctors) and gating defects (potentiators) of the cystic fibrosis causing mutation p.Phe508del are a useful treatment for CF patients bearing p.Phe508del mutation. Until now, 21 exonic variants in cis of p.Phe508del have been identified, albeit at a low frequency. The aim of this study was to evaluate their impact on the efficacy of CFTR-directed corrector/potentiator therapy lumacaftor/ivacaftor.

Mutations were generated by site-directed mutagenesis in both minigenes of the corresponding exon and full-length CFTR. The effect on exon skipping was evaluated in Beas2B cells, the effect on CFTR maturation and function in HEK293 cells.

The analysis by minigene showed that two out of 15 cis variants tested increased exon skipping (c.609C>T and c.2770G>A). Four cis variants were studied functionally in the absence of p.Phe508del, one of which was found to be deleterious for protein maturation (c.1399C>T aka p.Leu467Phe). In the presence of p.Phe508del, p.Leu467Phe prevented the response to lumacaftor/ivacaftor treatment. Preliminary results also indicated a deleterious effect of cis mutations c. 157T>G (p.Leu53Val) Simultaneously, studies were realized in 36 p.Phe508del patients beginning lumacaftor/ivacaftor treatment. 15 patients who displayed an increase in ppFEV 1 of at least 5% were classed as responders (average improvement by 13.5% (2.1); p<0.0001) and 21 patients were classed as nonresponders, displaying a change in ppFEV 1 below 5% (-1.07% (0.8); p=0.82). Analysis of the whole CFTR coding sequence of these patients identified exonic heterozygous variant c.3080T>C(p.Ile1027Thr) in 2 patients with c. 259T>C (p.Phe87Leu) in cis in one of them, and c.609C>T (p.IleI203Ile) in one other patient. These three patients were found as nonresponders to lumacaftor/ivacaftor treatment.

Our results underline the importance of validating treatment efficacy in the context of complex alleles. There are over 2000 mutations associated with the cystic fibrosis transmembrane conductance regulator (CFTR) gene and approximately 280 of the mutations have been confirmed to cause CF. These mutations can be divided into six different defect classes that may be amendable to a precision medicine approach wherein a given mutation class may be addressed using its own therapeutic intervention approach. The deletion of phenylalanine at position 508 (F508del) is the most prevalent mutation among CF patients, with over 85% of patients carrying at least one allele. The F508del mutation impairs the conformational maturation of the protein which attenuates the trafficking of CFTR to the apical surface of bronchial cells of the lungs. Pharmacological chaperones known as correctors can partially rescue this trafficking phenotype, and the early first-to-market correctors, such as lumacaftor (VX-809) and tezacaftor (VX-661), when combined with a potentiator, have been demonstrated to confer only a modest clinical benefit to CF patients homozygous for the F508del mutation.

The more recent clinical stage correctors, such as olacaftor (VX-440) and VX-659, in combination with the early correctors, have demonstrated meaningful improvements in in vitro and early clinical efficacy. Similarly, we have described a novel corrector, PTI-801, that complements both groups of clinical correctors and thus belongs to a new category of correctors. In addition, PTI-801 is distinct from the other correctors in that it protects against the in vitro chronic inhibition of F508del-CFTR stability and activity conferred by ivacaftor . We thus sought to determine whether the corrector mechanisms that have been described by the three reported complementary groups of correctors represent the only nonredundant F508del-CFTR maturation pathways.

To this end, a high-throughput phenotypic screen to discover additional novel CFTR modulators was performed. We describe a novel class of CFTR correctors with distinct characteristics from known clinical stage CFTR correctors. PTI-CCCN is capable of synergizing with lumacaftor, tezacaftor, olacaftor and PTI-CCZ, a corrector similar to PTI-801. PTI-CCCN improves F508del-CFTR chloride transport activity, and nearly doubles this activity when combined with lumacaftor or tezacaftor and ivacaftor. When combined with a more recent corrector such as olacaftor, PTI-CCCN further improves the CFTR activity and band C maturation, and to an even greater extent in the presence of tezacaftor. Intriguingly PTI-CCCN is complementary with PTI-CCZ, indicating they also represent distinct classes of CFTR correctors. Lastly, the combination of PTI-CCCN, PTI-CCZ, the PTI amplifier and potentiator restored CFTR activity in homozygous F508del HBE cells to approximately 140% of normal CFTR.

This new modality of CFTR correction represents a new advance in the development of small molecule modulators to target the dysfunction of F508del-CFTR. These novel correctors expand the diversity of available combinations for known modulators with the ultimate aim to improve clinical efficacy in CF patients.

Cystic fibrosis (CF) is a lethal genetic disease caused by insufficient activity of the chloride channel Cystic Fibrosis Transmembrane conductance Regulator (CFTR). The predominant cause of death in CF is lung failure due to persistent bacterial infection and inflammation of the airways. Interestingly, many of the bacteria common in CF airways secrete the enzyme sphingomyelinase (SMase), which degrades membrane sphingomyelin into phosphocholine and ceramide. Our lab has shown that bacterial SMase applied to the basolateral side inhibits forskolin-activated chloride currents in bronchial epithelial cells from patients (hBEs) (Stauffer BB, et al. Sci Rep. 2017; 7:2931) . This experimental setup precludes a direct interaction between SMase and the hBE chloride channels (CFTR and TMEM16A), since these channels are apical and SMase was applied basolaterally. Furthermore, confocal fluorescence microscopy showed that the sphingomyelin abundance at the apical membrane remained unaltered upon basolateral SMase treatment, indicating that a change in the lipidic environment immediately surrounding these apical chloride channels is unlikely. Still a few mechanisms remain to explain the SMase-mediated inhibition of chloride currents in these cells: 1.) SMase inhibits a basolateral transporter, reducing the chloride influx and thus the efflux; 2.) SMase initiates a signaling cascade leading to an inhibitory modification on apically-located CFTR, TMEM16A, or both; or 3.) SMase initiates a signaling cascade resulting in a reduction in the surface expression of CFTR, TMEM16A, or both. The work herein attempts to determine the basis of SMase-mediated inhibition of chloride currents. To complete this goal, we will utilize hBEs and an immortalized bronchial epithelial cell line in the Ussing chamber to monitor transepithelial current and resistance. Furthermore, we will use various pharmacological activators and inhibitors to isolate CFTR current from TMEM16A current. Lastly, we will use cell-surface biotinylation pull downs to quantify a change in CFTR and/or TMEM16A membrane expression. Deriving the mechanism by which SMase inhibits chloride currents in hBEs is important, as it will inform how bacterial virulence factors may need to be targeted to prevent further inhibition of the already deficient CFTR chloride current to improve treatment for CF patients. (Support: CF Fdn. MCCART17G0.)

Cystic fibrosis (CF) is a chronic lethal genetic disease, which is caused by mutations in the cystic fibrosis transmembrane conductance regulator (CFTR). F508del -deletion of a single phenylalanine residue at position 508is the most common lethal mutation found in CF patients. Fully synthesized F508del-CFTR is rapidly degraded in the endoplasmic reticulum due to F508del-induced misfolding, preventing proper transport of the protein to the epithelial cell membrane. Previously, we have demonstrated that F508del facilitates partial unfolding and consequent aggregation of the first nucleotide-binding domain in human CFTR (hNBD1). This F508delfacilitated aggregation is likely a key contributor to F508del-CF pathogenesis. These findings established a foundation to develop a new highthroughput biophysical screen to identify compounds that directly bind to the hNBD1 domain to prevent the F508del-hNBD1 aggregation in vitro and thereby stabilize F508del-CFTR in vivo. Therefore, we first constructed an isolated hNBD1 that covalently binds to a visible fluorophore at a single labeling site through conjunction reaction between maleimide group of fluorophore and cysteine residue from the protein. Thermal fluorescence self-quenching (tFSQ) of this labeled domain upon unfolding/aggregation provides a robust and accurate high-throughput assay for corrector compounds. We demonstrated several thymidine nucleotide analogs bind ~3-fold tighter than ATP, the physiological ligand of hNBD1. These thymidine analogs offset half of the defect in thermal stability in F508del-hNBD1 and preserve the gating function of F508del-hCFTR channels at 37°C where ATP cannot, showing thymidine analogs merit further exploration as cystic fibrosis drugs.

Meanwhile, the tFSQ assay can be successfully conducted in the presence of moderate detergent environments using the isolated fluorescence-labeled hNBD1, suggesting the tFSQ assay can be performed using whole CFTR as protein target, which gives more advantages in presenting more possible compound-binding pocket compared to the isolated hNBD1 domain. Hence, we engineered a cys-reduced whole hCFTR protein for single-site fluorophore labeling. Similar to isolated hNBD1, the fluorescently-labeled hCFTR provides a strong and reproducible tFSQ signal showing that thymidine also stabilizes the full-length hCFTR more strongly than ATP. This system provides an unprecedented opportunity to characterize thermodynamic and dynamic interactions between the domains in CFTR and is suitable to perform high-throughput screening for compounds directly stabilizing full-length hCFTR in vitro. subject to regulation by phosphorylation dependent (by cAMP and cGMP dependent kinases) and phosphorylation independent (by glutamate) mechanisms in the human sweat duct (Reddy, Quinton, 2003) . We have therefore asked how these different regulatory mechanisms are affected by different CFTR mutations. In addition, we have recently reported (Madireddy R, et.al. Pediatr Pulmonol. 2017; 52(S47) :214 [abstract] ) that CFTR potentiator, VX-770 acutely increased CFTR mediated HCO 3 conductance only in normal but not in CF sweat ducts expressing ΔF508/nonsense mutations (G542X or W1282X). Therefore, the objective of this study was to determine the role of alternative mechanisms of stimulating CFTR-HCO 3 conductance in sweat ducts expressing different mutant forms of CFTR. We studied the role of phosphorylation dependent and independent mechanisms of CFTR mediated HCO 3 permeability in normal CF ducts using electrophysiological techniques on freshly isolated, α-toxin permeabillized and microperfused human sweat ducts (Reddy, Quinton, 2003) . Our general experimental protocol included perfusion of the cytoplasmic bath with 140 mM KGlu while changing the composition of the luminal perfusate from Glu -(gluconate) to HCO 3 -. We have then followed the acute effects of cytosolic application of cAMP, cGMP (10 -5 M each) and glutamate metabolite α-ketoglutarate (1mM) in the presence of 5mM ATP on transepithelial HCO 3 conductance relative to the impermeant anion Glu -. These studies were conducted on sweat ducts from wild-type and the heterozygote CF ducts expressing ΔF508/R117H (n=2 subjects and 8 ducts) and ΔF508/3849 (n=1 subject and 3 ducts) mutant forms of CFTR. Preliminary data indicated that under the experimental conditions cAMP and cGMP together and α-ketoglutarate independently (in the presence of ATP) could stimulate CFTR mediated HCO 3 conductance in wild-type ducts. However, only α-ketoglutarate+ATP could stimulate ΔF508/R117H and ΔF508/5849 heterozygote CF ducts that were incubated overnight with CFTR corrector VX-809 (3mM). These results indicated that the efficacy of different agonists in stimulating aberrant CFTR-mediated HCO 3 - conductance appears to be selectively compromised by different CFTR mutations in stimulating HCO 3 - conductance. In conclusion, these preliminary results suggest that glutamate regulation of CFTR-HCO 3 conductance is relatively unaffected by certain CFTR mutations. These observations further emphasize the need for identifying mutant specific CFTR agonists that can be selectively used for restoring aberrant CFTR mediated HCO 3 conductance affected by different CFTR mutations.

Funded by CFFT and The Nancy Olmsted Trust. Thanks to Mr. K. Taylor for technical assistance.

UBE2L6 IS A UBIQUITIN CONJUGATING ENZYME THAT DIRECTS CFTR ΔF508 DEGRADATION Rajagopalan, C. 1 ; Wei, H. 1 ; Bouhamdan, M. 1 ; Xie, Y. 2 ; Sun, F. 1 1. Physiology, Wayne State University, Detroit, MI, USA; 2. Oncology, Wayne State University, Detroit, MI, USA Cystic Ffibrosis (CF) is one of the most common genetic diseases in the USA and is caused by mutations of the cystic fibrosis transmembrane conductance regulator (CFTR) gene. The most common CF mutation is the deletion of a phenylalanine residue at position 508, resulting in a mutant protein (CFTR ΔF508) that is prematurely degraded by the endoplasmic reticulum-associated degradation (ERAD) system. CFTR ΔF508 escaped from ERAD migrates to the plasma membrane and retains substantial CFTR Clchannel function. Therefore, understanding the molecular mechanism regulating CFTR ΔF508 degradation during its early biogenesis and identifying the ERAD components that may serve as targets for drug development are important for the treatment of CF patients.

Previously, our lab demonstrated that RNF19B (ring finger protein 19B) is a novel E3 ligase in the ubiquitin-mediated degradation of CFTR ΔF508. Knockdown and overexpression of RNF19B have decreased and increased CFTR ΔF508 degradation, respectively. RNF19B directly interacts with CFTR ΔF508. In this study, as a result of screening 34 ubiquitin E2 conjugating enzymes, we found that UBE2L6 is one of several enzymes related to CFTR ΔF508 degradation. Overexpression of UBE2L6 increased CFTR ΔF508 degradation by 2-3-fold. Knockdown of UBE2L6 by siRNA or overexpression of a UBE2L6 dominant-negative mutant substantially decreased CFTR ΔF508 degradation. We further demonstrated that UBE2L6 is physically associated with RNF19B. Moreover, silencing UBE2L6 expression increased VX-809-mediated CFTR ΔF508 maturation. Finally, siRNA-mediated silencing of endogenous UBE2L6 in the CF epithelial cell line CFTR-ΔF increased forskolin-stimulated short-circuit currents compared to control. There was a greater increase in short-circuit currents when UBCE2L6 silenced cells were treated with VX-809. In summary, we unveil a novel E2 enzyme that cooperates with RNF19B to target misfolded CFTR ΔF508 for ERAD. This study presents a novel therapeutic strategy to synergistically restore the function of CFTR ΔF508.

Funding supported by CFF (SUN15XX0) and NHLBI/NIH (HL096800 and HL133162 to FS). epithelial resistance and CFTR function. We used nonhomologous end Introduction: The theratype of individuals with CF who carry at least one very rare CFTR mutation is unknown in most clinical cases. We have previously established planar nasal epithelial cultures (PNECs) from nasal brushings for personalized CFTR function measurements in Ussing chambers and quantified CFTR potentiator responses to ivacaftor in three CF individuals including F508del/1154insTC, F508del/Y563N, and splice variant 3849+10kbC->T/G542X that correlated with a decrease in sweat chloride concentration in a small N-of-1 trial (McGarry ME, et al. Pediatr Pulmonol. 2017; 52:472-9) . The goal of this study was to determine the unknown responses to CFTR-targeted therapies for two CF patients carrying the rare F508del/S1159P (c.3475 T->C; pSer1159Pro) or ultrarare F508del/c.850dupA genotype. The S1159P mutation, a thymine-to-cytosine nucleotide replacement in exon 22 at nucleotide 3475 of the CFTR gene is known to cause a serine-to-proline exchange at amino acid 1159, right before the second nucleotide-binding domain, however the functional impact of the S1159P mutation on CFTR activity is unknown. The c.850dupA mutation is caused by a duplication of an adenine at nucleotide position 850 in exon 6 that presumably leads to termination of CFTR translation at the end of the 4th membrane spanning region in the first transmembrane domain. Currently only one CF patient is known with the F508del/c.850dupA genotype. We determined the level of basal CFTR activity and response to CFTR correctors lumacaftor; tezacaftor) and CFTR potentiator (VX-770, ivacaftor) by using patient-derived cells from nasal brushings. PNECs were generated by conditional reprogramming techniques and used at passage 1. Functional activity of apical F508del/S1159P or F508del/c.850dupA CFTR proteins was assessed by measuring acute responses to forskolin and VX-770 (1 µM) followed by CFTR inhibitor172 (50 µM) without or with overnight CFTR corrector treatment (VX-809, VX-661; 3 µM) with the Ussing assay.

Results: PNECs with F508del/S1159P genotype responded acutely to ivacaftor and cAMP stimulated Clcurrents increased from 5.4±1.0 to 10.6±1.8 µA/cm 2 (1.96-fold) . Mutant F508del/S1159P CFTR function was restored by VX-770 to 29% of normal activity and increased by CFTR corrector treatment to 33% (VX-809) and 40% . PNECs with F508del/c.850dupA genotype responded acutely to VX-770 and cAMP stimulated Clcurrents increased from 1.7±0.3 to 2.1±0.3 µA/cm2 (1.24fold) . Mutant F508del/c.850dup CFTR function was restored to 6% of normal activity without and slightly increased by CFTR corrector treatment to 9% (VX-661) and 10% .

Conclusion: The rare F508del/S1159 genotype is a robust responder to CFTR potentiator treatment whereas F508del/c.850dupA is a weak responder. F508del compound heterozygous PNECs with 30% of normal CFTR activity in Ussing chambers were associated with normalization of clinical biomarkers (sweat chloride, lipase, amylase).

Acknowledgments: Supported by CFF (ILLEK16G0), NIH (30 DK072517), Elizabeth Nash Foundation, and private donations.

Learning Objectives: 1) Recognize the prevalence of CFTR-related disorders without pulmonary manifestations; 2) Understand the basic pathophysiology of these disorders; 3) Implement the knowledge of these disorders into clinical practice.

Case Description: A 48-year-old male with medical history significant for alcohol abuse presented to gastroenterology clinic for evaluation of recurrent acute pancreatitis. He had experienced multiple episodes of epigastric and low back pain associated with elevated lipases over the preceding several months despite an intentional 25-lb weight loss, decreased alcohol consumption and normal lipid levels. Imaging revealed a 9x8 mm cystic mass in the pancreatic head. Pathology results from a FNA showed atypical ductal epithelium concerning for malignancy. He underwent a Whipple with jejunostomy. Biopsy results were consistent with chronic pancreatitis without malignancy. Several months later, the patient decided to undergo genetic testing for chronic pancreatitis with the Ariel Precision Medicine Lab. He was found to have 3 CFTR and 2 CASR variants that increased his risk of pancreatitis. His sweat chloride testing results put him in the intermediate risk category for CF. Given his fertility, lack of pulmonary symptoms and normal spirometry, he was diagnosed with CF-related disorder, explaining the etiology of his recurrent acute pancreatitis.

Discussion: Pancreatic disease is a known complication of CF, and nearly 100% of CF patients will have some type of pancreatic pathology. Exocrine function is the most reliable phenotypic marker with up to 85% of CF patients developing pancreatic insufficiency. Because the presence of acinar tissue is necessary for the development of pancreatitis, the majority of pancreatic-insufficient CF patients do not experience pancreatitis. Within the population of pancreatic-sufficient CF patients, only 10-20% develop pancreatitis. Ledder and coworkers speculate that CFTR and non-CFTR genotypes as well as environmental factors contribute greatly to the development of pancreatitis in these patients (J Gastroenterol Hepatol.. 2014; 29:1954-62) . Determining which genetic mutations can lead to pancreatic disease without true CF has been challenging. LaRusch and coworkers isolated several of these mutations and hypothesized that particular mutations can cause isolated disruptions in bicarbonate secretion through CFTR, leading to complications such as pancreatitis while preserving normal pulmonary function. This patient tested positive for R75Q, one of 9 mutations studied by LaRusch et al and found to be associated with pancreatitis but not true CF. This mechanism is thought to hinge on regulation of WNK1-SPAK activity. The WNK1-SPAK pathway is activated in conditions of low-intracellular chloride, increasing the permeability of the channel for bicarbonate. All 9 variants in the study lead to defective WNK1-SPAK activity, implying that increased intracellular bicarbonate concentrations within the pancreas could lead to higher risk of pancreatitis (PLoS Genetics. 2014; 10(7) :e1004376).

While all genetic variants of CF are not completely understood, emerging evidence shows that various mutations in the CFTR gene can lead to a vast spectrum of phenotypes. A systematic approach to correlating channel function with clinical presentation is vital to appropriately diagnosing and treating patients.

Introduction: CFTR is a cAMP regulated Cland HCO 3 channel that when dysfunctional disrupts microtubule (MT) stability due to reduced function of EPAC1, a MT elongation factor. Loss of MT stability impairs intracellular transport, initiating a signaling cascade believed to contribute significantly to inflammatory responses in CF. These findings suggest that some aspect of cAMP regulation is disrupted in CF cells. Given reports of a CF-like phenotype in patients with a carbonic anhydrase (CA) deficiency, we hypothesized aberrant regulation of CA could be the link between CFTR dysfunction and these cell signaling alterations via soluble adenylyl cyclase (sAC), a bicarbonate sensor and producer of cAMP. Identifying this mechanistic link will assist in the development of new therapies to augment the CFTR modulators and to better control inflammation.

Methods: CA protein expression was determined by Western blot in primary human nasal epithelial (HNE) cells treated with lumacaftor (VX-809). CA mRNA expression was evaluated by qPCR in mouse nasal epithelial (MNE) cells to confirm previous RNAseq data. MT stability and cholesterol trafficking were examined in primary MNE and HNE cells treated with L-phenylalanine, a CA activator, with the addition of sAC inhibitor (KH7) to clarify the mechanism. Cyclic AMP production was determined in HNE cells using a cAMP assay kit from Cayman Chemical.

Results: CA2 expression is decreased at the protein and RNA levels in CF compared to non-CF primary MNE and HNE cells with some patientto-patient variability in primary HNE cells (p<0.05). Similarly, CA2, CA9, and CA12 RNA expression is decreased in CF compared to non-CF HNE cells suggesting a broader disruption of carbonic anhydrase regulation in CF. Given the high level of activity and cytosolic location of CA2, this became the focus for investigations of CA's impact on the disrupted cellular signaling cascade in CF in this study. While activating CA2 with L-phenylalanine restores cholesterol trafficking (n=4) as well as MT structure and elongation (n=3) to wild-type (WT) profiles in primary CF HNE cells, its effect is blocked by inhibition of sAC with KH7 (p<0.05). Activating CA with L-phenylalanine in CF model and primary HNE cells significantly increases cAMP production in IB3s and both WT and CF HNEs (n=3; p<0.05). CFTR-mediated regulation of CA2 expression is demonstrated by reduced CA2 expression in response to inh172 within 24 hours in WT MNE cells (n=3, p<0.05) . Also, treating primary CF F508del homozygous HNE cells with VX-809 leads to an increase in CA2 protein expression but not in CF HNE cells carrying a missense and splice mutation.

Conclusions: CA2 and CA12 expression is reduced in a CFTRdependent manner in multiple CF airway epithelial models including primary mouse and human nasal epithelium. Stimulation of CA activity with L-phenylalanine restores microtubule regulation, intracellular transport, and other CF cellular phenotypes to WT patterns via a sAC-dependent mechanism. These data suggest that reduced CA-mediated bicarbonate production may be a key link between mutant CFTR dysfunction and CF cell regulatory cascades.

Acknowledgments: Supported by the CWRU FRAP, NIH T32, and CFF Third Year Clinical Fellowship Award.

submucosal glands respond to acetylcholine (Choi JY, et al. J Clin Invest. 2007; 117:3118-27) . The beneficial effect of anti-cholinergic therapy for chronic obstructive pulmonary diseases like COPD is well-documented although cholinergic stimulation paradoxically inhibits liquid absorption, increases cilia beat frequency and increases airway surface liquid transport. Here we have investigated the role of bicarbonate in airway mucus bundle transport and the effect of the anti-muscarinic drug ipratropium bromide (Atrovent).

Methods: Mucociliary transport of Alcian blue-stained mucus bundles and bead-collecting strands was evaluated by time-lapse recordings of excised distal trachea and primary bronchi from newborn WT and CFTR-/-(CF) pigs mounted in a heated chamber with aerated Krebs-glucose buffer. Mucus bundle transport was also evaluated after incubation with by the clinically used anti-muscarinic compound ipratropium bromide and stimulation with acetylcholine.

Results: In WT piglet trachea, mucus bundles secreted from the submucosal glands swept with uneven speed over the airway surface. Interestingly, in pigs lacking a functional CFTR channel, the mucus bundles were almost immobile. The stagnant CF mucin bundles were trapped on the tracheal surface attached to the surface goblet cells. Pseudomonas aeruginosa bacteria were moved by the mucus bundles in WT but not CF pigs. As expected the surface liquid transport was increased by acetylcholine and carbachol. In contrast, the mucus bundles were stopped from moving by acetylcholine, an effect inhibited by ipratropium bromide.

Conclusions: CF is characterized by stagnant mucus that permits bacterial overgrowth. We now show that the mucus bundles were essentially stationary in the newborn CF piglets as well as in WT when bicarbonate was removed from the bathing solution. The cholinergic system is important for respiratory physiology as the vagus nerve uses acetylcholine to increase airway smooth muscle tone, submucosal gland secretion and ciliary beat frequency at the same time as it inhibits surface liquid absorption. Increased cholinergic tonus is common in asthma and COPD and the muscarinic inhibitor ipratropium bromide is used for pharmacological treatment of COPD in particular. The use of this treatment is rather counterintuitive except for its inhibitory effect on bronchoconstriction. However, the quick effect of acetylcholine to stop the transport of mucus bundles suggests another and novel mechanism that could explain some of the beneficial effect of anti-cholinergic treatment. Our results further suggest that mucus bundle movement is controlled by attachment-detachment of the mucus bundles to the surface goblet cells. Background: Mucus is critical for the respiratory mucociliary transport defense system, and in large airways, submucosal glands produce most of the mucus. Mucus emerges from submucosal glands to form long strands. However, the function of mucus strands in moving inhaled particles has remained uncertain. We studied newborn pigs to test the hypothesis that movement of large particles requires mucus in the form of strands.

Methods: Mucociliary transport of tantalum disks was assessed by computed tomography in spontaneously breathing newborn pigs with normal airway humidification. We also measured transport of steel or tantalum spheres on pig tracheas ex vivo after submersion in physiologic saline with 5% CO 2 at pH 7.4 and 37°C. To determine the effect of increased mucus production in vivo and ex vivo, we stimulated with the cholinergic agonist methacholine.

Results: In ex vivo experiments, mucus strands moving over the airway surface attached to immobile steel spheres and pulled them along, initiating their movement. Stimulation of mucus secretion with methacholine reduced the delay before mucus moved the spheres. The reducing agents dithiothreitol and tris- (2-carboxyethyl) phosphine (TCEP) disrupted mucus strands. As a result, fewer spheres moved and for those that did, the delay before initiating movement was prolonged. We obtained similar results with CT-scan tracking of microdisks in newborn piglets. Methacholine increased the percentage of microdisks moving and reduced the delay before the microdisks were propelled up airways. Aerosolized TCEP prevented these effects. Once particles were in motion, reducing agents did not alter the speed of these metallic particles ex vivo or in vivo.

Conclusions: These findings indicate that submucosal glands not only secrete abundant mucus, but also produce it in a strand form that is critical to initiate movement of large particles. These findings may also explain why rodents and small mammals lack submucosal glands; large particles do not normally enter their lungs and thus mucus strands may not be needed.

The ability to extract key functional metrics in the native airways microenvironment is important in the investigation of mechanisms behind CF pathogenesis. We previously reported the successful clinical translation of a 1 µm resolution micro-optical coherence tomography (µOCT) imaging technology for the study of functional microanatomy of upper airways in human subjects. Here, we report the latest advancements in this technology and results quantifying morphologic and functional MCC parameters from a clinical study involving cystic fibrosis (CF) and healthy subjects.

Method: An intranasal µOCT probe was used to image multiple regions within the nasal inferior meatus of unsedated subjects. µOCT videos consisting of cross-sectional images that captured airway mucosa and the overlying airway surface liquid layer (ASL) were acquired and stabilized through cross-correlation methods before further image analysis was carried out. The use of a custom-made probe-stabilizing nose cone was also explored to enhance video stability during image acquisition. Key functional microanatomic metrics extracted from stabilized µOCT videos included thicknesses of ASL and periciliary layers (PCL), mucociliary transport rate (MCT), ciliary beat frequency (CBF), epithelial thickness (ET), percent cilia coverage (pCC), and normalized brightness of ASL mucus. In subjects that exhibited an influx of presumably inflammatory cells (bright ~ 10 µm globular structures), the mean count per frame was quantified.

Results: Stark qualitative (e.g. mucus structure) and quantitative differences were observed between CF (n=10) and control (n=10) subjects. Cohort analysis of the key functional microanatomy measures showed a significant decrease in mean PCL thickness, MCT, CBF, pCC, and significant increase in mean glob counts per frame in the CF group as compared to controls. The CF ASL µOCT reflectance intensity was also found to be significantly higher than that of healthy ASL, suggesting elevated mucus viscosity (Shei, et al. Pediatr Pulmonol. 2018; 53(S2):256 [abstract] ). Furthermore, positive linear correlations between PCL, CBF, and pCC versus MCT were observed.

Conclusion: We demonstrated the utility of first-in-kind clinical intranasal µOCT imaging for characterizing multiple defects of the functional microanatomy of CF airways, which include previously observed defects in animal models (e.g. delayed MCT, depleted PCL, and elevated viscosity in CF swine models), as well as previously unrecognized pathologies (e.g. decreased ciliation, inflammatory cell infiltration).

Acknowledgments: Funding from NIH (T32HL105346, 5R01HL116213, R35HL135816, P30DK072482, and UL1TR001417) and CFF (TEARNE16XX0, ROWE14Y0 and ROWE16XX0).

Vladar, E.K. 1 ; Milla, C. 2 ; Axelrod, J.D. 3 1. Pulmonary Sciences and Critical Care Medicine (DOM) , University of Colorado Anschutz Medical Campus, Aurora, CO, USA; 2. Pediatrics, Stanford University School of Medicine, Stanford, CA, USA; 3. Pathology, Stanford University School of Medicine, Stanford, CA, USA Introduction: Cycles of injury and incomplete repair brought on by chronic inflammation and infections elicit structural and functional defects termed remodeling in the cystic fibrosis (CF) airway epithelium. Remodeling precipitates disease phenotypes and progression, but cellular targets and molecular mechanisms are poorly defined. We propose that defects in the basal stem cell population (BCs) cause epithelial remodeling. We show that BCs incur long-lasting damage in CF ("BC injury"), and differentiate into a remodeled epithelium in vitro ("BC dysfunction"). BC injury may stem from global epigenetic alterations eliciting transcriptional changes that lead to reduced self-renewal and differentiation, and/or depletion of specific BC subtypes with self-renewal and/or differentiation potential in the progenitor pool.

Methods: We isolated BCs from healthy and F508del homozygous CF donor nasal brushings. We cultured BCs on Collagen I-coated plastic dishes to assess their ability to proliferate (self-renew) and on Transwell membranes at air-liquid interface to assess their ability to differentiate (multipotency) into luminal airway epithelial cell types. We also tested the barrier and regenerative function of the resulting epithelium.

Results: We find that CF BCs proliferate more slowly and have a diminished capacity to differentiate into ciliated cells upon air-liquid interface culture and this epithelium shows reduced barrier function and wound healing response. We show that BCs isolated from more highly inflamed CF airways display greater dysfunction, however BCs from neonatal CF donors differentiate normally. Importantly, BCs from adults treated with CFTR corrector therapy or BCs differentiated in the presence of these drugs still display dysfunction.

Conclusions: Our results are consistent with a model where BCs are injured by inflammation at the onset of chronic lung disease and drive remodeling and disease progression via maladaptive repair. Unresolved underlying BC dysfunction may also render CF therapies less effective.

Ford, B.D. 1, 2, 3 ; Margaroli, C. 3 ; Forrest, O.A. 3 ; Brown, M. 3 ; Tirouvanziam, R. 3 1. Immunology and Molecular Pathogenesis, Emory University, Atlanta, GA, USA; 2. CF Atlanta, Atlanta, GA, USA; 3. Pediatrics, Emory University, Atlanta, GA, USA Background: The airways of patients with cystic fibrosis (CF) are chronically inflamed, due in part to unrelenting recruitment of polymorphonuclear neutrophils (PMNs) from blood into the airways. Prior studies by our group showed that PMNs recruited to the CF airway lumen acquire a pathological phenotype featuring active elastase-rich granule release, acquisition of immunoregulatory activities, and metabolic licensing (dubbed "GRIM" PMNs). Recently, we showed that GRIM PMNs can be produced in vitro by transmigrating blood PMNs toward CF sputum supernatant (CF SSN) through a small airway epithelium (Forrest OA, et al. J Leukoc Biol. 2018; doi:10 .1002/JLB.5HI1117-454RR). Monocytes can potentially exert inhibitory effects on PMN-driven inflammation, however, their ability to survive upon migration into CF airways and their fate therein is unclear.

Objective: Our goal was to determine whether blood monocytes survive and alter their phenotype and metabolism upon in vitro transmigration toward CF SSN, as a way to mimic their recruitment to CF airways.

Methods: Blood monocytes were isolated from healthy control (HC) and CF subjects and transmigrated to CF SSN or control apical fluid (medium with chemoattractant leukotriene B4 (LTB4), or HC SSN). Additional controls included incubation of HC and CF blood monocytes with CF SSN or medium with LTB4 without transmigration. Cells were analyzed pre-and post-transmigration (and incubation without transmigration) for viability and expression of critical surface markers (e.g., scavenger receptors, activation/exhaustion receptors) using flow cytometry. In addition, cells were analyzed for metabolic capacity along both glycolytic and oxidative phosphorylation (OxPhos) routes using a Seahorse XFp bioenergetics platform.

Results: Monocytes transmigrated to CF SSN did not display any survival disadvantage compared to those recruited to control fluid. In addition, bioenergetics analysis showed enhanced poise for glycolysis-and OxPhos-supported metabolism in monocytes transmigrated to CF SSN compared to control fluid. Finally, phenotypic analysis showed increased expression of the scavenger receptor CD91 and the activation/exhaustion marker programmed death-1 (PD-1), and lower expression of the inhibitory receptor CD172a/SIRPalpha and of the other scavenger receptor CD36 in monocytes transmigrated to CF SSN compared to control fluid. Preliminary comparisons of blood monocytes from CF vs. HC donors suggests subtle differences in their response to transmigration.

Conclusions: Our results suggest the CF airway environment supports the survival of transmigrated monocytes and confers an activated phenotype to them, similar to that seen with transmigrated PMNs. In particular, transmigrated monocytes appear poised for clearance functions with modulation of critical scavenger and inhibitory receptors. However, transmigrated monocytes express high levels of the exhaustion marker PD-1, indicating that they may be ineffective in this environment due to overactivation, thereby leaving GRIM PMNs unchecked.

Persistent inflammation, infections, and lung disease cause hundreds of premature deaths of cystic fibrosis (CF) patients every year. Hence, novel therapeutic approaches aimed at reducing excessive inflammation, promoting clearance of infection, and stimulating lung repair in CF are urgently needed.

In order for inflammation to be beneficial to the host, resolution must occur since uncontrolled inflammation may sustain CF development. Inflammation resolution is an active process modulated by specialized pro-resolving lipid mediators (SPMs). Among SPMs, resolvin (Rv) D1 holds unique anti-inflammatory and proresolution properties that regulate leukocyte infiltration, macrophage efferocytosis, and tissue damage.

Here we show the potential of RvD1 as an emerging strategy to stimulate resolution of chronic inflammation and infection, and to drive lung tissue regeneration in CF.

Gut-corrected Cftr knockout mice were chronically infected using agar-embedded Pseudomonas Aeruginosa (PA) and treated with RvD1 or vehicle. Disease severity, inflammation resolution, and bacterial load were evaluated at 21 days postinfection.

Obtained results showed that RvD1 is host protective from severe disease and bacteria spread as from persistent PA-driven lung disease in CF mice. In particular, RvD1 significantly dampened bacterial burden and LPS amounts in lungs. RvD1 also markedly dampened total leukocytes and PMN infiltration and ameliorated histological scores of lung pathology. In addition, RvD1 treatment increased the percentage of phagocytosis of PA by CF mouse lung macrophages in a dose-dependent manner with a maximum significant enhancement (~ 100%) at 10 nM compared to vehicle-treated cells. Of interest, this enhancement of phagocytosis by RvD1 was associated with a significant reduction in KC (~ 4.7 fold), IL-6 (~ 1.7 fold), and IL-10 (~ 12 fold) released by macrophages during PA engulfment.

Overall, these results indicate that the RvD1 has important functions on innate antimicrobial responses of CF cells in the airway milieu that can contain and limit P. aeruginosa infection. Hence, more studies in appropriate preclinical models and cell systems are required to strengthen our knowledge on the clinical value of RvD1 as an innovative approach to treat P. aeruginosa infections and exert regulatory actions on overshooting, relentless CF inflammation. 1 1. Royal College of Surgeons in Ireland, Dublin, Ireland; 2. School of Immunology, Trinity College Dublin, Dublin, Ireland Introduction: CF lung disease is characterized by persistent, aggressive neutrophilic inflammation. The key functions of neutrophils require energy. ATP is the building block of energy. Glucose can be used to fuel ATP production through two linked metabolic pathways: glycolysis and the TCA cycle. In most cases, the TCA cycle is responsible for aerobic metabolism, while glycolysis is responsible for anaerobic metabolism. However, under certain circumstances, cell metabolism can shift towards a state of aerobic glycolysis. This metabolic reprogramming phenomenon, known as the Warburg effect, has been shown in cancer cells predominantly, but also in inflammatory cells such as macrophages (Palsson-McDermott EM, et al. Cell Metab. 2015; 21:65-80) and results in increased HIF-1α mediated transcription. The proinflammatory cytokine IL-1β is transcribed by HIF-1α as pro-IL-1β, a precursor protein that undergoes cleavage to its active form, typically by inflammasome-driven caspase-1 activity.

Aim: To investigate the presence of a Warburg effect in the CF neutrophil.

Objectives: To demonstrate metabolic reprogramming of the CF neutrophil and its role in the production of IL-1β.

Methods: Ethical approval was granted by Beaumont Hospital ethics committee. Bronchoalveolar lavage fluid (BALF) and peripheral blood samples from patients carrying F508del/F508del (CF, n=12), healthy controls (HC, n=12) and F508del/F508del patients who had received a double-lung transplant (DLT, n=6) were obtained. Highly purified neutrophils were isolated and cytosolic and nuclear fractions prepared. Pyruvate kinase M2 (PKM2), phosphorylated PKM2 and HIF-1α were detected by Western blot with subsequent densitometric analysis. Lactate, pyruvate and the TCA cycle intermediate succinate were measured via colorimetric assay. Pro-IL-1β and cleaved IL-1β were measured by ELISA.

Data: Cytosolic PKM2 was increased in CF compared to HC (P<0.0001), with return to HC levels post-DLT. This pattern was also observed for phosphorylated PKM2 (P<0.0001) and HIF-1α (P=0.003). Nuclear levels of phosphorylated PKM2 dimers and HIF-1α were increased in CF compared to HC (P=0.002, P=0.001). Cytosolic succinate was higher in CF than in HC, with resolution post-DLT (P<0.0001), an effect also observed for lactate (P<0.0001), lactate:pyruvate ratio (P=0.0003) and pro-IL-1β (P=0.0002). Cleaved IL-1β was increased in CF BALF neutrophil cytosols (P=0.001) compared to both HC and DLT but not in peripheral neutrophil cytosols. IL-1β was induced by the TLR4 agonist LPS in a doseand time-dependent manner, with this effect abrogated in a dose-dependent manner by the glycolytic inhibitor 2-deoxyglucose, the succinate dehydrogenase inhibitor DMM and MCC-950, a specific inhibitor of the NLRP3 inflammasome (all P<0.0001). Levels of IL-1β in BALF supernatants were higher in CF than HC, again resolving post-DLT (P=0.001).

Conclusion: A Warburg effect is present in the CF neutrophil, and drives increased production of IL-1β, the processing of which is via NLRP3. Specific inhibition of both the PKM2/succinate/HIF-1α axis and NLRP3 is achievable and represents a therapeutic target. France; 4. Hôpital Trousseau, Paediatric Respiratory Department, Paris, France Introduction: Cystic fibrosis (CF) is the most common lethal genetic disease in the Caucasian population, caused by CFTR (cystic fibrosis transmembrane conductance regulator) gene mutations. CFTR encodes a chloride channel, which is essential for the osmotic balance of airway surface liquid and mucus clearance. The most common mutation is F508del, resulting in the lack of the channel at the apical membrane and leading to an ionic imbalance and a thickened mucus. Hyperviscosity resulting leads to an impaired mucociliary clearance promoting bacterial colonization and establishment of infection/inflammation cycles, which in the long term degrade the pulmonary epithelium. Chronic inflammation is the hallmark of CF lung disease. Indeed, a polynuclear neutrophil invasion and an interleukin-8 (IL-8) hypersecretion were reported in bronchoalveolar lavage fluids, sputum and in primary airway epithelial cells of CF patients. Whereas the origin of inflammation is still discussed, previous works have demonstrated that this chronic inflammation is mainly caused by the alteration of the NF-κB pathway (Tabary O, et al. Am J Pathol. 1998; 153:921-30.) which can be regulated by microRNA (miR). MicroRNAs are a class of noncoding small RNA, which most often bind to the 3'-untranslated region (3'UTR) of target genes mRNAs and thereby repress their translation and/or induce their degradation and are little studied in cystic fibrosis.

Objective: To study the involvement of microRNA in pulmonary inflammation observed in CF patients.

Methods: By an overall analysis of all miR (miRNome), we demonstrated that miR-199a-3p expression was decreased in the bronchial epithelial cells of CF patients. miR-199a-3p is predicted to target the 3'UTR of IKKβ, one of the proteins of the NF-κB pathway.

Results: In this study, we showed on bronchial explants from patients that miR-199a-3p is decreased in CF patients compared to non-CF patients, thus confirming miRNome data. By in vitro studies, we have shown that miR-199a-3p modulates the expression of IKKβ by a direct interaction at its 3'UTR in bronchial epithelial cells from CF patients (CFBE41o-). By miR-199a-3p overexpression experiments, we showed that miR-199a-3p decreased the expression of the IKKβ protein, the activity of NF-κB and the secretion of IL-8 in the CF cell culture supernatants. However, from a fundamental point, the origin of miR-199a-3p expression deregulation remains to be studied. We showed that miR-199a-3p deregulation origin is influenced by pro-inflammatory context and seems due to intracellular calcium concentration modulation in CF cells.

Conclusion: We demonstrated that miR-199a-3p has a negative regulatory role in the NF-κB signaling pathway and that its poorly expression in CF patients contributes to a chronic pulmonary inflammation.

Acknowledgments: Supported by Sorbonne Université -Faculté des sciences, Paris, France; Vaincre la mucoviscidose. Airway acidification contributes to several airway diseases, including cystic fibrosis. In the current study, we examined the effects of acute airway acidification on mucus secretion and transport properties in neonatal male and female piglets. Acute airway acidification induced glycoprotein obstruction of the intrapulmonary airways in both sexes, suggesting a defect in either the secretion or transport properties of mucus. Thus, we examined mucus secretion in response to methacholine ex vivo using lectins. The pattern of mucus secretion was profoundly different in acid-challenged piglets compared to saline-treated controls. Notable differences included a greater expulsion of surface mucus and increased number of mucus strands. Significant dilation of the submucosal gland duct was also observed. We also examined mucus transport properties in submerged tracheas using fluorescent nanospheres. Computer-assisted tracking revealed a decrease in the mean speed of fluorescent particle movement. These findings suggest that acute airway acidification modifies mucus behavior and provide further insight into cystic fibrosis pathogenesis and disease progression.

Supported by R00HL119560-03 (PI, LRR) and 10T2TR001983-01 (Co-I, LRR).

Philippe, R. 1 ; Higgins, G. 2 ; Ringholz, F. 2 ; Ghorbani, S. 1 ; Sassi, A. 1 ; Zare, A. 1 ; Harvey, B. 3 ; Hatton, A. 1 ; Sermet, I. 1 ; McNally, P.G. 3 ; Urbach, V. 1,2,3 1. U1151, INSERM, Paris, France; 2. NCRC, Dublin, Ireland; 3. RCSI, Dublin, Ireland Background: Cystic fibrosis (CF) lung disease is characterised by dysregulated ion transport that promotes chronic bacterial infection and inflammation. In addition, CF patients have an intrinsic unbalance between pro-inflammatory lipid mediators such as LTB4 and specialised proresolution mediators (SPM) that are biosynthesised by lipoxygenases (LO) . In this study, the roles of SPMs, lipoxin (LXA4) and resolvin (RvD1) were investigated on epithelial ion transport and innate defence in CF airways. Furthermore, the impact of CFTR mutation was investigated on LO expression, localisation and activity.

Methods:Studies were performed on primary cultures of bronchial epithelial cells from patients with CF, on human bronchial epithelial cell lines expressing wild-type CFTR or F508del-mutated CFTR and on homozygous F508del-CFTR mice. Airway surface liquid (ASL) and ion transport were investigated using confocal microscopy on human airway epithelial cells and nasal potential difference in mice. LO were studied using RTqPCR, Western blotting and confocal microscopy.

Results: LXA4 and RvD1 increased the ASL height in human CF bronchial epithelial cells. Both SPMs also restored nasal transepithelial potential difference in CF mice by decreasing amiloride-sensitive Na + absorption and stimulating CFTR-independent Clsecretion. In addition, SPMs decreased TNFα-induced IL-8 secretion and enhanced the phagocytic and bacterial killing capacity of human CF alveolar macrophages. Among the LO (15LO, 15LO2, 12LO, 5LO) that are expressed in different cell compartments of CF and non-CF airway epithelial cells, 5LO appeared diffused in the cytoplasm in non-CF airway epithelial cells and at the nuclear envelope in CF cells. In macrophages membrane nuclear localization has been associated with enhancement of LTB4 (pro-inflammatory) biosynthesis at the expense of LXA4 (pro-resolving), while translocation of 5LO in the cytoplasm is associated with an increased LXA4 synthesis.

Conclusion: Since SPMs normalize epithelial ion transport and promote resolution of inflammation in CF airways, abnormal SPM production could contribute to CF airway surface dehydration and persistent inflammation. Translocation of 5LO to the nuclear envelope in CF airway epithelial cells suggests that epithelial 5LO plays a role in the abnormal class switching from pro-inflammatory lipid mediators to SPM in patient airways.

Sheehan, C. 1 ; Bomberger, J.M. 2 ; Madden, D.R. 1 1. Dartmouth College, Hanover, NH, USA; 2. University of Pittsburgh, Pittsburgh, PA, USA Pseudomonas aeruginosa (Pa) is a gram-negative pathogen responsible for a variety of opportunistic infections. In the airway, the bacterium uses a wide array of toxins and virulence factors to evade host immunity and cause disease in patients with respiratory diseases such as cystic fibrosis and chronic obstructive pulmonary disease (COPD). Pa has been shown to reduce mucociliary clearance in airway epithelial cells by altering the trafficking of CFTR, an ion channel that helps maintain chloride levels in epithelial cells (MacEachran DP, et al. Infect Immunol. 2007; 75(8) :3902). Reduction in CFTR-mediated chloride secretion leads to a widening of the mucus layer and an increase in bacterial and fungal infections. Our collaborators determined that the reduction of apical CFTR levels is caused by a specific virulence factor known as the CFTR inhibitory factor (Cif). Cif is an epoxide hydrolase (EH) secreted from Pa that negatively affects ATP-binding cassette (ABC) transporters such as CFTR, transporter associated with antigen presentation 1 (TAP1), and p-glycoprotein, by altering their intracellular trafficking (Bomberger JM, et al. J Biol Chem. 2014; 289:152; Ye S, et al. J Physiol Cell Physiol. 2008; 295:C807-18) . In airway epithelial cells, Cif causes a reduction in apical CFTR levels by inhibiting ubiquitin specific protease 10 (USP10)-mediated deubiquitination of CFTR. Ubiquitinated CFTR signals for degradation which leads to a decrease of CFTR in recycling endosomes and an increase of CFTR in late endosomes fated for the lysosome. Further work demonstrated that Ras GTPase-activating protein-binding protein 1 (G3BP1) regulates USP10 activity and that Cif promotes complex formation between these two proteins which renders USP10 inactive (Bomberger JM, et al. PLoS Pathog. 2011; 7(3) :e1001325). Taken together, these data show that Pa utilizes a novel virulence factor to dysregulate host protein trafficking to promote colonization in the airways. We are currently leveraging our understanding of Cif's enzymatic activity to further explore how epoxides and epoxide hydrolysis are needed for the Cif-mediated CFTR degradation. Through the use of Cif mutants that disrupt catalysis, substrate binding, and substrate specificity, we are gaining a better understanding of how Cif affects the various stages in the CFTR lifecycle. Simultaneously, we are using a biophysical approach to determine how the USP10:G3BP1 complex is formed and regulated. Specifically, we have used biophysical and biochemical assays to elucidate how this complex is formed and regulated in the context of Cif exposure. These data demonstrate that Cif causes more than just CFTR degradation and Cif is affecting other proteins and pathways in epithelial cells. This work opens new areas of exploration on this unique and novel Pa virulence factor.

Introduction: Bitter taste receptors (TAS2Rs) are multifunctional G-protein coupled receptors recently discovered throughout airway smooth muscle (ASM), causing bronchodilation when activated (1, 2) . Approximately 50% of patients with cystic fibrosis (CF) have hyper-reactive airways and many of these patients are refractory to traditional bronchodilator therapy (3) . Long-term bronchodilator therapy with β 2 -agonists in CF is controversial and has shown inconsistent results in treatment of airway hyper-reactivity in patients with CF (3) . Activation of TAS2R agonists increases intracellular calcium leading to ASM relaxation. The cystic fibrosis transmembrane conductance regulator (CFTR) has been detected in sarcoplasmic reticulum in CF and non-CF newborn pigs. In CF piglets, the loss of CFTR function is associated with abnormal calcium signaling and increased airway reactivity (4) . The purpose of this study was to evaluate TAS2R agonists on ASM cells reactivity comparing CF and non-CF models. Our hypothesis was CF models will exhibit enhanced relaxation to TAS2R agonists on ASM cells compared to non-CF cells.

Methods: ASM cell relaxation was assessed using Precision Cut Lung Slices (PCLS) obtained from wild-type (WT) and CFTR -/-(CF) mice (5 males, 2 females both with CF) aged 11.5 +/-1 weeks. After cannulating the trachea, the lungs were filled with 2% agarose, cooled, removed and sectioned using a Compresstome VF-300 tissue slicer. Peripheral lung slices (130 mm thick) were incubated (10% CO 2 ) in DMEM buffer and antibiotics overnight. Slices were mounted onto a modified microscope slide and perfusion chamber with a chamber volume of ~100 mL and a perfusion rate of ~800 mL/min. Initially, lung slices were treated with 1 mM acetylcholine (ACh), followed by 100 mM chloroquine (CQ). Using phase-contrast microscopy, the perfused slices were visualized in real-time, and images were obtained (10x objective) at the end of each round of drug exposure. The images were then analyzed by measuring bronchiole luminal area at baseline, after treatment with ACh, and after treatment with CQ.

Results: A total of 7 lung slices (3 WT and 4 CF) were used in this study. A 2-sample t-test was performed to compare change in bronchiole diameter following CQ administration comparing CF and WT mice. There was a statistically significant difference in percentage of diameter increase after CQ treatment in the bronchioles of CF mice (M=16.9 mm 2 , SD=4.8) compared to control WT mice (M=5.2 mm 2 , SD=5.0), t(5) = -3.11, P=0.026.

Conclusion: Our results suggest that TAS2Rs are functional in CF ASM cells, and that stimulation with the receptor agonist chloroquine leads to an enhanced relaxation response in CF ASM cells compared to non-CF cells.

References : 

The impaired acidification of macrophage phagolysosomes (PL) as a feature of CF is contested (Di A, et al. Nature Cell Biology. 2006; 8:933-44, Haggie PM, Verkman AS. J Biol Chem. 2007; 282(43):31422-8) . If defective acidification is to be developed as a therapeutic target, definitive evidence in human macrophages is required. Measurement of PL acidification with pH-sensitive fluorophores lacks spatial resolution and may not be truly ratiometric (Nunes P. J Vis Exp. 2015;e53402). Surface-enhanced Raman Spectroscopy (SERS)-based nanosensors are a novel alternative to analyse pH in living cells. We have used nanosensors consisting of a gold nanoparticle (NP) functionalised with the pH-sensing molecule para-mercaptobenzoic acid (p-MBA). p-MBA molecules change structure in response to H + ion concentration, which is enhanced by exciting the NP at their plasmon resonance of 785 nm, allowing Raman spectra to be acquired and analysed from individual NPs.

Objective: To precisely measure PL pH in CF and healthy control (HC) human monocyte-derived macrophages (MDMs) using SERS-based nanosensors.

Methods: Peripheral blood mononuclear cells were isolated from whole blood by dextran sedimentation and Percoll ® gradient from 9 HC and 6 CF participants and differentiated into MDMs over 7 days. Functionalised NPs were then incubated with MDMs for 1 hour before live cell imaging on a Renishaw inVia™ Reflex micro Raman Spectrometer, allowing individual NPs to be targeted. pH values were quantified using a Boltzmann curve fitted to calibration data. Light microscopy permitted measurement of NP phagocytosis rates. Transmission electron microscopy (TEM) was used to confirm the presence of NPs in PLs. Fluorescent quantification of MDM phagocytosis using pHrodo™ Green Zymosan BioParticles ® was performed as a comparison. In some experiments, HC cells were treated with CFTR inhibitor . Summary data are expressed as mean±SD. Results were analysed by one-way ANOVA with Tukey's post hoc test, with p<0.05 deemed statistically significant.

Results: TEM confirmed NP localisation to PLs. Raman spectroscopy demonstrated no difference in the pH of HC and CF MDM PLs (5.43±0.12 vs. 5.41±0.22, p=0.9998) . By comparison, measurement by pHrodo™ Green Zymosan Bioparticles ® Conjugates also demonstrated no difference in acidification, but with a much wider SD, reiterating the imprecise nature of non-ratiometric fluorescent microscopy. Phagocytosis rate did not differ between genotypes (HC 90±6% vs. CF 90±3%) underlining that any measurable difference in pH is not attributable to varied NP concentration in MDMs. CFTR inh -172 treatment did not alter pH in HC MDM PLs.

Conclusions: Using novel SERS-based nanosensors, we demonstrate no acidification defect in human CF MDMs. This new technique offers higher accuracy to conventional fluorescent microscopy by targeting the measurement of pH to individual NPs and is thus truly ratiometric. Our data confirm that PL acidification is CFTR-independent in human MDMs and may not be critical in the pathophysiology of CF lung disease.

Armstrong, D.A. 1 ; Chen, Y. 2 ; Nymon, A. 3 ; Hazlett, H.F. 2 ; Dessaint, J.A. 1 ; Salas, L. 2 ; Christensen, B.C. 2 ; Ashare, A. 1, 2 1. Medicine, Lebanon, NH, USA; 2. Geisel School of Medicine at Dartmouth, Hanover, NH, USA; 3. Thayer School of Engineering, Dartmouth College, Hanover, NH, USA Introduction: Lung macrophages are the major cells involved in the pulmonary innate immune response. In the cystic fibrosis (CF) lung, the inability of lung macrophages to successfully regulate the inflammatory response suggests a dysfunctional innate immune system. Various studies have revealed numerous physiological defects associated with CF macrophages including: dysregulation of phagocytic / signaling receptors (Simonin-LeJeune K, et al. PLoS One. 2013; 8(9) :e75667), hyper-responsiveness to microbial stimuli (Bruscia EM, et al. J Immunol. 2011; 186(12) :6990), and impairment in removal of apoptotic cells (Vandivier RW, et al. J Clin Invest. 2002; 109(5) :661). Epigenetics is the study of heritable changes in gene function caused by mechanisms other than changes in the underlying DNA sequence (Wu C, et al. Science. 2001; 293(5532):1103) . Epigenetic mechanisms have emerged as modulators of host defenses that can lead to a more prominent immune response and shape the course of inflammation in the host, both driving the production of specific inflammatory mediators and controlling the magnitude of the host response (Morandini AC, et al. Pathog Dis. 2016; 74 (7)). The most widely studied of the epigenetic modifications is DNA methylation (Marsit CJ, et al. Epigenetics. 2015; 10(8) :708). In this study, we aim to gain insight into innate immune cell dysfunction in CF by investigating alterations in DNA methylation in lung macrophages of CF subjects.

Methods: All analyses were performed using primary lung macrophages from human subjects collected via bronchoalveolar lavage (BAL). Epigenome-wide DNA methylation was examined via Illumina Methyl-ationEPIC (850K) array. Methylation-based sample classification was performed using the Recursively Partitioned Mixture Model (RPMM) and was tested against sample case-control status. Differentially methylated loci were identified by fitting linear models with adjustment of age, sex, estimated cell type proportions and repeat measurement.

Results: We identified 109 differentially methylated CpGs (FDR ≤ 0.1) in CF macrophages, of which 51 are hyper-methylated and 58 are hypo-methylated. Hyper-methylated CpGs included those associated with genes such as TNFSF8 and RUNX3. Gene promoter regions were enriched among CpGs hyper-methylated in CF subjects. Hypo-methylated CpGs included those associated with genes such as S100A14, LSP1, and OSCAR. Hypo-methylated CpGs were typically found within non-promoter CpG islands as well as in putative enhancer regions and DNase hypersensitive regions. Gene/pathway analysis utilizing the Kyoto Encyclopedia of Genes and Genomes tool (KEGG) revealed pathways such as: biosynthesis of unsaturated fatty acids, glycerolipid metabolism, Fc gamma R-mediated phagocytosis and Fc epsilon RI signaling as the top CpG-gene-associated pathways likely affected in CF subjects.

Conclusions: These results support a hypothesis that epigenetic changes, specifically DNA methylation, at a multitude of gene loci in lung macrophages may participate, at least in part, to drive a dysfunctional innate immune response in the CF lung.

Long, M.E. 2,1 ; Gong, K. 2,1 ; Volk, J.S. 2 ; Eddy, W.E. 2 ; Liles, W. 2 ; Manicone, A.M. 2, 1 1. Pulmonary, Critical Care and Sleep Medicine, Univ. of Washington, Seattle, WA, USA; 2. Center for Lung Biology, Univ. of Washington, Seattle, WA, USA Introduction:Pseudomonas aeruginosa pulmonary infections are a driver of progressive lung decline in CF. Recurrent and chronic infections contribute to increased inflammation, subjecting the lung to host-mediated tissue damage. Unfortunately, restoration of CFTR function does not eliminate chronic bacterial colonization in the lung, and initial attempts to eradicate infection after restoring CFTR function have not been successful. Anti-inflammatory strategies to reduce deleterious lung inflammation without compromising host-defense mechanisms may be therapeutically beneficial and will likely be a long-term need in CF. Macrophages are innate immune cells with dichotomous roles in controlling inflammatory responses, and CFTR dysfunction results in increased pro-inflammatory macrophage responses, in part due to over-activation of the MEK1/2-ERK1/2 pathway. My studies discovered that MEK1/2 is a central regulator of macrophage inflammatory and functional responses. MEK1/2 inhibitors are FDA-approved compounds used in certain cancer chemotherapeutic regimens and recent investigations indicate these compounds have potentially beneficial anti-inflammatory/immune modulating properties.

Objective: Determine whether MEK1/2 inhibitors can reduce detrimental inflammation by targeting and modulating immune cell functions without impairing host defense in CF.

Methods: 1) Validate MEK1/2-ERK1/2 pathway activation as a therapeutic target by immunohistochemistry in human CF lung explants. 2) Use murine bone marrow-derived and alveolar macrophages and human blood monocyte-derived macrophage to determine if MEK1/2 inhibitors decrease M1 polarization and increase M2 polarization and efferocytosis of apoptotic cells. 3) Evaluate the therapeutic potential of MEK1/2 inhibitors to reduce inflammation without compromising host defense in experimental murine lung infection models.

Results: Initial analyses of human CF-lung explant tissue indicate high phospho-ERK1/2 activation in mononuclear cells in the lung and in cells localized to sputum clusters in the airway. Treatment of murine and human macrophages with a MEK1/2 inhibitor decreases M1 polarization by LPS and increases M2 polarization with IL-4 and IL-13. MEK1/2 inhibitor treatment also increases macrophage efferocytosis of apoptotic neutrophils. Delivery of a MEK1/2 inhibitor after initiation of P. aeruginosa pneumonia significantly reduces alveolar neutrophilic inflammation without impairing bacterial clearance compared to carrier-treated mice.

Conclusions: Our results indicate that activation of the MEK1/2-ERK1/2 pathway in immune cells in the CF lung is present and that MEK1/2 inhibitors promote reparative properties of macrophages. Our experimental results demonstrate the potential of therapeutic application of MEK1/2 inhibitors to decrease lung inflammation without impairing host defense mechanisms. Future studies will examine dose-responses of the FDA-approved MEK1/2 inhibitor in both gram-negative and gram-positive murine models of acute pneumonia and chronic airways infections.

Acknowledgments: Supported by UW RDP CFFSINGH, CFF LONG18F0, and UW Microbiology Charlie Moore Endowed Fellowship in Cystic Fibrosis.

Saleh, L.A. 1, 2 ; Koloteva, A. 1, 2 ; Boyd, A. 1, 2 ; Baskar, G. 1 (PDE4s), a group of cAMP-specific PDE isoenzymes, are promising targets for CF therapy. PAN-selective inhibition of PDE4s has been shown to stimulate CFTR activity in airway epithelial cells and is well established to exert potent anti-inflammatory effects in sterile models of lung inflammation. However, PAN-selective inhibition of PDE4 enzymes also induces a number of side effects including emesis, nausea, diarrhea and weight loss. These gastrointestinal side effects are strongly contraindicated for a CF therapeutic, given that patients suffer from pancreatic insufficiency and intestinal problems and that gaining and maintaining body weight is already a challenge and is strongly associated with health outcomes in CF. The PDE4 family comprises four genes or subtypes, PDE4A-D. As each plays unique and non-overlapping physiological roles in the body, targeting individual PDE4 subtypes may serve to separate therapeutically beneficial ones from the side effects of the nonselective PDE4 inhibitors available to date. To this end, we assessed the effects of genetic ablation of PDE4A, PDE4B and PDE4D in a mouse model of acute lung infection with P. aeruginosa.

Infection with 10 6 CFUs of P. aeruginosa lab strain PA01 induced ~40% mortality in wild-type mice over a five-day period. Mice deficient in PDE4B exhibited substantially better survival in this model, whereas mice deficient in PDE4D had minor, and mice deficient in PDE4A had no survival benefits. In a second set of experiments, we then assessed lung infection/inflammation in PDE4BKO mice and wild type littermate controls. Preliminary data show that at 16 hours post-infection, the levels of inflammatory cytokines, such as TNFα, as well as the bacterial load are reduced in bronchoalveolar lavage (BAL) fluid and lung tissue of PDE4B knockout mice compared to their wild-type controls. These data suggest that inactivation of PDE4B can serve to alleviate lung inflammation and injury in settings of P. aeruginosa infection without worsening bacterial infection. Future experiments will assess the role of PDE4 subtypes in animal models of chronic P. aeruginosa airway infection that closely mimic the clinical presentation of CF patients. This work was supported by grants from the CFF (SALEH18H0, RICHTE16GO) and the NIH (HL76125, HL141473, HL066299).

Hazlett, H.F. 1 ; Aridgides, D. 1 ; Armstrong, D.A. 1 ; Barnaby, R.L. 2 ; Katja, K. 2 ; Nymon, A. 2 ; Stanton, B.A. 2 ; Ashare, A. 1, 3 1. Medicine, Lebanon, NH, USA; 2. Microbiology and Immunology, Geisel School of Medicine at Dartmouth, Hanover, NH, USA; 3. Medicine, Geisel School of Medicine at Dartmouth, Lebanon, NH, USA Introduction: Macrophages play a critical role in clearing bacterial lung infections. In response to bacterial infections, macrophages migrate into the lungs where they phagocytose and kill bacteria and release inflammatory cytokines and chemokines. However, in cystic fibrosis (CF), macrophages exhibit a reduced ability to kill P. aeruginosa and manifest increased cytokine secretion compared to non-CF macrophages. Several reports have demonstrated that CFTR is expressed in macrophages and the bactericidal activity of macrophages is defective in CF. CFTR correction increases bacterial phagocytosis and killing in CF macrophages. Macrophages also play an important role in maintaining iron homeostasis, which may be clinically relevant in CF since increased iron enhances P. aeruginosa growth. CFTR mutations alter iron sequestration in CF bronchial epithelial cells (CFBEs; Moreau-Marquis S, et al. Am J Phys Lung Cell Mol Phys. 2008; 295:L25-37) . Disruption of nutritional immunity pathways in CFBEs may contribute to P. aeruginosa survival via increased biofilm formation (Hendricks MR, et al. Proc Natl Acad Sci. 2016; 113:1642-7) . The impact of CF on human macrophage nutritional immunity remains unknown. We hypothesized that CF macrophages have altered expression of iron regulatory proteins and this protein expression can be modulated by CFTR correction.

Methods: Peripheral blood monocytes were isolated from healthy (n=6) or CF (n=4) subjects and differentiated into macrophages (MDMs) by treatment with 100 ng/mL M-CSF for 7 days. To correct CFTR function, MDMs were treated with VX-809 (3 µM) and VX-770 (30 nM), followed by treatment with either LPS (10 and 1000 ng/mL) or P. aeruginosa strain PAO1 (MOI=10). Western blotting was performed on total protein. Protein band densitometry was measured using ImageJ. Secreted proteins were quantified using enzyme-linked immunosorbent assay (ELISA). Analyses were performed using two tailed unpaired t-tests and ANOVA with Tukey's test for multiple comparisons.

Results: In response to PAO1 and compared to healthy subjects, CF MDMs expressed decreased heme oxygenase 1 (HO-1) protein levels. An important regulator of heme catabolism and oxidative stress, HO-1 dysfunction is implicated in CF macrophage hyperinflammation (Zhang PX, et al. J Immunol. 2013; 190:5196-206) . Treatment with VX-770/VX-809 increased HO-1 protein expression to similar or greater levels than healthy subject MDMs. CF MDMs expressed elevated ferroportin protein, the major iron exporter, compared to healthy subject MDMs. In CF MDMs, treatment with VX-770/VX-809 reduced ferroportin protein induction by LPS and did not alter hepcidin secretion in response to LPS.

Conclusion: These results suggest that nutritional immunity pathways are altered in CF macrophages in response to bacterial stimuli, and modulation of CFTR function impacts expression of several proteins critical for management of iron in response to bacterial stimuli.

Danahay, H. 1 ; McCarthy, C. 1 ; Fox, R. 2 ; Gosling, M. 1 

The composition of mucus in the CF airway, and in particular the hydration status, significantly affects its clearance and thereby the potential to form plugs, restrict airflow and create a nidus for chronic microbial colonisation. A variety of largely ion channel-based strategies, eg CFTR repair, are being employed to promote mucosal hydration. An alternative approach would be to reduce the excessive production/secretion of the mucin proteins that contribute to the solids component of the mucus gel.

One approach to reduce excessive mucus production in the diseased lung is to reduce the number of mucus producing goblet cells. To identify drug targets that could be regulated to achieve this, we have utilised a 3D culture model of the human airway epithelium, "bronchospheres." Bronchospheres are derived from primary human airway basal cells, and can be cultured to form a well-differentiated mucociliary epithelium without the need for an air-liquid interface. Bronchospheres are cultured in a 384-well assay format that makes them amenable to medium throughput screening. Treating bronchospheres with mediators such as IL-13 induces a mucus hypersecretory phenotype with increased numbers of goblet cells and reduced numbers of ciliated cells. Our hypothesis was that the co-administration of test compounds together with IL-13 would identify compounds capable of preventing goblet cell formation with the opportunity to seed future drug discovery programs.

Bronchospheres were cultured as previously described (Danahay H, et al. Cell Rep. 2015; 10:239-52) . On day 2 after seeding primary human airway basal cells, treatment with IL-13 ± test compounds was initiated. A library of approximately 2,000 pharmacologically active low molecular weight compounds was used, each with a well-annotated mechanism of action. On day 8, media and treatments were topped-up and on day 14 bronchospheres were lysed and RNA isolated. QPCR was then used to assess the expression of cell-specific markers: MUC5AC (goblet cells) and FOXJ1 (ciliated cells). Compounds that induced a ≥2-fold reduction in expression of MUC5AC were classified as hits. This hit list was then refined by checking the expression of FOXJ1. Compounds that had likewise attenuated FOXJ1 expression by ≥2-fold were deprioritised as these likely represented a nonspecific effect on epithelial differentiation. Compounds that either maintained or enhanced FOXJ1 expression in addition to repressing MUC5AC gene expression were prioritised for validation using traditional air-liquid interface cultures.

In total, 92 hit compounds from the bronchosphere screen were tested for effects on goblet and ciliated cell numbers in ALI HBE cultures using quantitative immunohistochemistry. Of these, 38 (41%) significantly attenuated the MUC5AC+ stained area in IL-13 treated HBE and either maintained or increased the FOXJ1+ stained area. Validated hits were then aligned based on their previously reported pharmacological activity to enable common pathways to be identified and to refine our hypotheses through further exemplification of pathway regulators.

Following this screen, we are progressing a lead optimisation program for eventual therapy in respiratory diseases associated with mucus obstruction.

Hall, R.; Cole, P.J. Mucokinetica Ltd, London, United Kingdom Background: There is high medical need for treatments that enhance and sustain mucociliary clearance. In CF phase II clinical studies the osmotically active therapies inhaled hypertonic saline (HS) and inhaled mannitol (Bronchitol; MN) produced 1 hour improvement in whole lung mucociliary clearance; in phase III studies they produced modest benefits vs clinical end points. An ex vivo model of ciliary transport speed (CTS) of secretions was established on pig trachea (Hall R, Cole P. Pediatr Pulmonol. 2017;52(S47):237) to identify potential 2nd generation clinical candidates with improved profile vs ciliary transport. MKA 104 produced slow-onset but considerably longer duration of increased CTS compared to 1st generation treatments.

Objectives: 1) Characterise the effect on CTS time course profile of combining MKA 104 with HS, MN or hyperosmotic sodiun gluconate (HNG), or the P2Y2 agonist UTP. 2) Characterise CTS time course profiles on tracheas from cftr(-/-) piglets.

Methods: Pig tracheas were from a vet college or abattoir. Tracheas were opened to expose the epithelium and maintained in organ culture. Secretions placed at the lower end moved by cilia action to laryngeal end. CTS was calculated by distance moved over time. CTS measurement was repeated several times to set baseline speed for each trachea. The epithelium was covered with either balanced salt solution (BSS), or BSS containing a treatment. At 10 minutes excess fluid was removed and CTS measured again. CTS was measured repeatedly for up to 24 hours to profile efficacy and time course of response to treatments.

Results: 1) Wild-type (WT) pig trachea: HS, MN, HNG and UTP each stimulated CTS immediately following treatment. UTP response peaked at 1 hour and declined back to baseline by 2 hours. Peak CTS response to the osmotic treatments HS, MN and HNG was at 1.5 hours, and CTS declined thereafter. By contrast treatment with MKA 104 alone produced no enhancement of CTS until 40 minutes to 1 hour following treatment, but then increased to achieve greater efficacy and duration than the other treatments. Combination of MKA 104 with each of HS, MN, HNG and UTP resulted in rapid onset enhanced CTS immediately after treatment, this was followed by further enhanced CTS tracking the time course of MKA 104 alone. Combined MKA 104 and HS treatment achieved peak CTS at 5 to 6 hours; CTS was significantly above baseline at 24 hours. 2) cftr(-/-) pig trachea: MKA 104 alone gave a slow-onset response, CTS did not increase until 1 hour after treatment, then increased to peak at 3.5 hours. A second treatment of the same tracheas at 14.5 hours resulted in a second slow onset response that peaked after a further 3.5 hours. Treatment of tracheas with combination of MKA 104 and HS resulted in increased CTS immediately after treatment, and peak response at 3.5 hours. Second treatment at 17 hours again gave an immediate increase in CTS that again peaked after a further 3.5 hours.

Conclusions: 1. Compared to MKA 104 alone, combination treatment with HS, MN, HNG or UTP delivers 1st hour efficacy together with long-duration enhanced CTS. 2. Results suggest that MKA 104 has potential to deliver improved clinical benefits in trials where patients continue use of 1st generation treatments HS or MN.

Sass, L. 1, 2 ; Enos, A. 2 ; Krishna, N. 2, 3 ; Cunnion, K. 1, 2 1. Pediatrics, Children's Hospital of The King's Daughters, Norfolk, VA, USA; 2. Pediatrics, Eastern Virginia Medical School, Norfolk, VA, USA; 3. Microbiology & Molecular Cell Biology, Eastern Virginia Medical School, Norfolk, VA, USA Introduction: Myeloperoxidase (MPO) is a heme-based peroxidase. MPO, which constitutes 30% of the dry weight of neutrophils, catalyzes the generation of hypochlorous acid (i.e. bleach) from hydrogen peroxide and chloride ion to kill pathogenic microorganisms. In cystic fibrosis, neutrophils are recruited to the lungs in response to infection with pathogenic bacteria, like P. aeruginosa, and release MPO. The MPO then catalyzes the generation of hypochlorous acid damaging the delicate lung tissue in an unsuccessful attempt to eradicate pathogenic bacteria.

PIC1 is a short synthetic peptide that strongly inhibits MPO oxidative activity in CF sputum ex vivo (Hair PS, et al. PLoS One. 2017; 12:e0170203) . We subsequently determined that PIC1 can also strongly inhibit the oxidant activity of other heme-based enzymes like hemoglobin and myoglobin (Hair PS, et al. Int J Pept. 2017; 2017:9454583) . The mechanism of antioxidant activity was proven to occur via both single electron transport and hydrogen atom transfer (Gregory Rivera M, et al. PLoS One. 2018; 13:e0193931) .

In order to evaluate the extent to which PIC1 could inhibit MPO oxidative activity in vivo, we developed an inflammatory peritonitis model in rats mediated by intraperitoneal (IP) injection of purified human MPO. PIC1 was then injected after the MPO and the oxidative activity in the peritoneal fluid measured by TMB.

Methods: Young adult rats were injected IP with 0.1 mg of purified MPO. Rats were then injected IP with PIC1 in one of three doses: 4 mg/kg, 20 mg/kg or 80 mg/kg. After 2 hours the rats were euthanized and peritoneal washes with 20 mL of ice cold saline were performed. The peritoneal washes were sedimented for cells and then the supernatant was measured for oxidant activity with a TMB-based assay, which is a standard method for measuring MPO peroxidase activity.

Results: The peroxidase activity of the MPO in the peritoneal fluid was measured 2 hours after IP injection. MPO injection demonstrated an 11-fold increase in peroxidase activity compared with saline control. A dose response was demonstrated for increasing doses of PIC1 correlating with decreased peroxidase activity. The highest dose of PIC1 resulted in a 4-fold decrease in peroxidase activity (P = 0.0004) compared with MPO alone.

Discussion: These findings demonstrate that PIC1 can inhibit MPO oxidative activity in vivo. This is consistent with our prior findings that PIC1 can inhibit MPO oxidative activity in vitro and in CF sputum samples ex vivo. These results provide additional evidence that the inflammatory lung damage in CF mediated by the oxidative activity of MPO can potentially be pharmacologically moderated. Acknowledgment: Supported by Cystic Fibrosis Foundation CUNNIO17I0. is an important part of airway innate defense and is impaired in CF subjects and CF animal models. Previously, we found that MCC velocity (MCCV) was synergistically increased by combining a low dose of carbachol (a Ca 2+ agonist) and forskolin (a cAMP agonist) in ex vivo ferret tracheas (Joo NS, et al. Sci Rep. 2016; 6:36806) . One possible mechanism for synergistic increases in MCCV is synergistic increases in secretion from submucosal glands, produced by low dose carbachol added to forskolin (Choi JY, et al. J Clin Invest. 2007; 117:3118-27 and Joo NS, et al. J Clin Invest. 2010; 120:3161-6) . As one test of this hypothesis, we asked if synergistic increases in MCCV could be produced in rabbit tracheas, which lack airway submucosal glands.

We measured MCCV ex vivo in adult New Zealand rabbit tracheas by following particle movements with time-lapse digital imaging (Jeong JH, et al. Am J Physiol Lung Cell Mol Physiol. 2014; 307:L83-93) . We measured unstimulated basal MCCV with and without pretreatment of 10 µM benzamil to block ENaC driven fluid absorption from the surface epithelia (Joo NS, et al. Sci Rep. 2016; 6:36806) . Tracheas were treated with either 0.3 µM carbachol or 10 µM forskolin, followed by the combined agonists (all added basolaterally). To assess the effect of CFTR inhibition on forskolin-induced MCC, 40 µM BPO-27, a selective CFTR inhibitor (provided by Alan Verkman) was used.

Results: Unlike ferret tracheal MCC where basal MCCV was 1.8 ± 0.3 mm/min at T15, basal MCCV in rabbits was 4.22 ± 1.51 mm/min at T15 (n = 2 rabbits). Forskolin increased T10-30 MCCV by 53% over basal , but carbachol did not stimulate MCCV in rabbit tracheas (n = 3). Combined forskolin + carbachol also failed to synergistically increase MCCV in rabbit tracheas (n = 7). ENaC inhibition by benzamil increased both initial (T0-1) and averaged (T10-30) basal MCCV by 181% and 76%, respectively (n = 2). CFTR inhibition by BPO-27 did not change forskolin-induced MCCV. However, increased numbers of immobile particles remained on the tracheal surface at the end of MCCV measurement periods.

Summary: Six features of MCCV in glandless rabbit tracheas were documented. (1) MCCV was increased by forskolin, but (2) not by carbachol and (3) showed no synergistic increase to the combined effects of carbachol + forskolin. (4) MCCV (measured by movement of fastest particles = majority of particles) was unaffected by CFTR inhibition, but (5) increased numbers of particles were immobile. (6) Inhibition of ENaC increased basal and stimulated MCCV.

Conclusion: These results are consistent with a model in which MCCV is increased by increased amounts of airway surface liquid. Rabbit tracheas lack submucosal glands and also lack carbachol-responsive surface epithelial secretory cells (Joo NS, et al. Sci Rep. 2016; 6:20735) . The high level of basal MCCV in rabbit tracheas is interesting and remains to be explained; a higher level of baseline mucus secretion is a possibility, and at least part of this might be via CFTR.

Introduction: Neutrophilic airway inflammation in cystic fibrosis (CF) is associated with structural lung damage and decreased lung function, even in absence of infection. Necrosis of airway epithelial cells (AEC), a characteristic finding in CF, has been associated with increased neutrophils and elevated interleukin-1alpha (IL-1α). Here, we aimed to measure IL-1α in human paediatric CF airway, assess for associations with inflammatory markers and structural lung changes, and test in vitro whether necrosis of AEC undergoing anoxia drives IL-1α release.

Methods: Bronchoalveolar lavage fluid (BALf) from CF children (n=102, mean age: 3.78; range: 0.16 -7.81 years) with and without respiratory infection was measured for IL-1α, IL-1β, IL-8, neutrophils and neutrophil elastase (NE) activity. Extent of structural lung disease on CT was measured via PRAGMA-CF and associations with IL-1α, IL-1β, IL-8, neutrophils and neutrophil elastase (NE) activity were investigated via multiple linear regression (adjusted for age and sex). CF (AEC CF ) AECs (n=7) were also collected and cultures established. Cells were exposed to anoxia over 48 hours and viability, apoptosis, and necrosis assessed via flow cytometry reported as a percentage of unexposed control (mean ± standard deviation), and IL-1 measured in supernatant. Wilcoxon signed rank test was used to test for significant differences (p<0.05).

Results: IL-1α and IL-1β were detectable in BALf from young children with CF in absence of detectable infection, were increased in the presence of bacterial infection and correlated with IL-8 (r=0.64 and r=0.64 respectively; p<0.0001), neutrophil counts (r=0.71 and r=0.67 respectively; p<0.0001) and NE activity (r=0.26; p<0.01 and r=0.32; p<0.001). When stratified by respiratory infection status, there were associations between IL-1α, IL-1β, IL-8, neutrophil count and NE activity and extent of structural lung disease on CT in children without respiratory infection, however the association between IL-1α and extent of structural lung disease on CT was the strongest (1.20 [0.33, 2 .06], p=0.008). Exposure to anoxia for 48 hours resulted in a significant decrease in cell viability (71.97% ± 34.12; p<0.05), significant increase in cell necrosis (232.8% ± 174.4; p<0.05) but no significant change in apoptosis (124% ± 31.36). IL-1 was measured in supernatant collected from cells exposed to anoxia.

Conclusion: IL-1α is detectable in the CF airway in young children with CF and is associated with structural lung disease on CT, potentially driven by increased necrosis of AEC after exposure to anoxia.

Acknowledgments: Supported by CFA, CFWA, USCF, German Federal Ministry of Education and Research.

Laube, B. 1 ; Carson, K.A. 2 ; Zeitlin, P.L. 3 ; Mogayzel, P.J. 1 1. Pediatrics, Johns Hopkins University, Baltimore, MD, USA; 2. Epidemiology, Johns Hopkins University, Baltimore, MD, USA; 3. Pediatrics, National Jewish Health, Denver, CO, USA Introduction: This study examined the natural history of mucociliary clearance (MCC) in children with cystic fibrosis (CF), the impact of MCC on the evolution of CF lung disease, and the relationship between the natural history of MCC and infection history with Pseudomonas aeruginosa (PA).

Methods: Children ≥6 years of age with CF who had previously undergone an MCC test (MCCv1) at Johns Hopkins underwent the following procedures 3-10 years later: (1) MCCv2 quantified as percent removal of 99m technetium-sulfur colloid from the right lung over 60 minutes (MCC60v2), using gamma camera imaging technology, following radioisotope inhalation, and MCC plus cough clearance (MCC90v2), quantified after 30 programmed coughs between 60 and 90 minutes; (2) lung clearance index (LCI), determined from a multiple breath nitrogen washout test, using the EXHALYZER® D; (3) number of parenchymal abnormalities in a high resolution computed tomography lung scan, scored (CT score) as described by Brody AS, et al. (J Pediatr. 2004; 145:32-8) ; (4) FEV 1 measured by a Microlab spirometer. Age of 1st PA+ culture and the number of PA+ cultures/year between MCCv1 and MCCv2 were documented. Pearson correlation was used to examine the association of: (1) age of 1st PA+ culture and number of PA+ cultures/year between MCCv1 and MCCv2 with percent MCC60v2 and MCC90v2, LCI @2.5%, CT score and percent predicted FEV 1 at the time of MCCv2; and (2) percent MCC60v1 and MCC90v1 with LCI, CT score and FEV 1 at the time of MCCv2. Statistical significance was defined as p<0.05.

Results: Nineteen children with CF (9 males) completed both MCC visits. Mean (±standard deviation) age at time of MCCv1 was 10.3±1.9 years. Median time between visits was 4.6 (range: 3.1-10.0) years. Mean age of 1st PA+ culture was 8.4±4.6 years (n=18). Mean number of PA+ cultures/year between MCCv1 and MCCv2 was 0.98±0.95. Mean percent MCC60 (10.0±5.2 vs. 9.4±6.1) and MCC90 (12.7±5.6 vs. 13.4±7.7) did not significantly differ for the 1st and 2nd visits, respectively. Mean age at 1st PA+ culture was not significantly correlated with any measurements obtained at the time of MCCv2. Number of PA+ cultures/year between MCCv1 and MCCv2 was significantly correlated with MCC60v2 (r=-0.49; p=0.03), MCC90v2 (r=-0.48; p=0.04), FEV 1 (r=-0.46; p=0.046) and LCI (r=0.60; p=0.006). Percent MCC60v1 (r=-0.53; p=0.02) and MCC90v1 (r=-0.59; p=0.008) were significantly correlated with CT score and MCC90v1 (r=-0.52; p=0.02) was correlated with LCI. There was no significant correlation between MCC60v1 and MCC90v1 and FEV 1 .

Pseudomonas aeruginosa infection has a significant impact on the natural history of MCC in children with CF. Children with a greater number of PA+ cultures had slower mucus removal. In addition, percent MCC60 and/or MCC90 may be early biomarkers for the development of parenchymal abnormalities and impaired ventilation in later years in children with CF.

Acknowledgments: Supported by Gilead Sciences and the Johns Hopkins Institute for Clinical and Translational Research, which is funded in part by Grant Number UL1 TR 001079 from the National Center for Advancing Translational Sciences. The major physiological agonist of the CFTR chloride channel is the vasoactive intestinal peptide (VIP), a 28-amino acid neuropeptide that functions as a neuromodulator and neurotransmitter. It is secreted by intrinsic neurons innervating exocrine glands and regulating their secretions. VIP is also a potent vasodilator and bronchodilator. Furthermore it has immunoregulatory activity, and stimulates chloride and water secretion by intestinal and tracheobronchial epithelium. Previous human studies have shown that normal skin, and intestinal and nasal mucosa have rich intrinsic neuronal networks for VIP secretion around exocrine glands. In contrast, CF tissues have little or no VIPergic innervation. These early findings suggest VIP plays a critical role in CF development. Our lab has since confirmed, in vitro and in vivo, the need for chronic VIP exposure to maintain functional CFTR chloride channels at the cell surface of airways and intestinal epithelium as well as normal exocrine tissues morphology.

In this study, we investigated changes in VIP content and innervation in exocrine tissues of 8-to 17-week-old C57Bl/6 CF mice compared to same age wild-types (WT). Paraffin-embedded tissues were sectioned and H&E stained for pathological assessment. VIP was detected by wholemount immunohistochemistry and the amount of VIP was semi-quantified by two blinded investigators using arbitrary scales. Fresh duodenum tissues from 8-week old WT and CF mice (n=5 in each group) were immunostained for VIP and a general neuronal marker (PGP9.5) and imaged with confocal microscopy followed by 3D reconstruction. Lastly, we measured VIP concentration in 8-week-old WT and CF duodenum homogenates with ELISA.

Our results show a strong reduction in VIP immunostaining in young CF mice (8-week old) with minimal signs of inflammation or tissue damage. Such VIP reduction continued to be observed in advanced disease animals (17-week old). Interestingly, 44% to 54% reduction in VIP staining was found in CF sweat glands which are free of inflammation at all ages. In other tissues, we found that the VIP signal was reduced at 8 and 17 weeks respectively, in the lung (44% and 24%), salivary glands (7.34% and 42%) , duodenum (48% and 38%), the endocrine (59.43% and 35.83%) and exocrine pancreas (24.68% and 43.11%). ELISA results confirmed a decrease in VIP concentration in the 8-week-old CF duodenum homogenates (median: 1,616 pg/mL and 888 pg/mL in WT and CF, respectively, n=5). Moreover, we observed a strong reduction in VIP innervation in the duodenum of CF mice at the mucosa and circular muscle layer with fewer fine axons in villi. However, the PGP9.5 signal showed that the general innervation was only slightly disrupted in the mucosal layer.

In conclusion, our data suggest that the low amount of VIP found in CF starts prior to tissue damage and corresponds to a specific reduction in VIPergic innervation that is not a consequence of a general neuronal damage. We propose that this represents an early defect in CF and constitutes an aggravating factor for disease progression that needs to be further investigated at the molecular level.

Supported by Cystic Fibrosis Canada. Mucus plugging of the respiratory apparatus and impaired clearance of mucus from the airways are hallmarks of CF lung disease, leading to the proliferation of pathogenic organisms and concomitant tissue destruction. A range of novel therapeutics targeting mechanisms contributing to abnormal CF mucus are being advanced and are at various stages of preclinical and clinical testing. Here, we evaluated the effect of ARINA-1, a patented inhalational combination product composed of three natural compounds currently in use with favorable safety profiles (ascorbic acid, bicarbonate, glutathione), on key microanatomic features of the mucociliary clearance apparatus as assessed via 1-micron resolution optical coherence tomography. In terminally differentiated primary human bronchial epithelial (HBE) monolayers derived from 4 F508del-homozygous CF donors, ARINA-1 (250 mM, apical) elicited significantly higher absolute (2.8±0.4mm/ min, P<0.0001) and relative (change vs. baseline: Δ2.7±0.4mm/min, P<0.0001) mucociliary transport (MCT) rates vs PBS (0.2±0.1mm/min and Δ0.2±0.1mm/min), peaking at 6 hours. ARINA-1 had no discernable effect on either absolute (not reported) or relative change in airway surface liquid (ASL: Δ2.9±1.2µm) or ciliary beat frequency (CBF: Δ1.2±0.5Hz) relative to PBS (Δ4.8±3.9µm, NS or Δ1.1±0.2Hz, NS). Linear regression analyses established that changes in ASL and CBF with ARINA-1 were significantly associated with changes in MCT, as seen in normal HBE, whereas relationships were weaker for PBS, as in CF (ΔASL: ARINA-1, R 2 =0.40, b=0.22, P<0.001 and PBS, R 2 =0.14, b=0.03, P<0.05; ΔCBF: ARINA-1, R 2 =0.22, b=0.52, P=0.01 and PBS, R 2 =0.10, b=0.10, P=0.08). This distinction is likely related to differences in viscosity and/or adhesion. Consistent with this, we previously reported that ARINA-1 decreases mucus reflectivity (an indicator of viscosity), and particle tracking microrheology analysis of mucus viscosity is in progress. Supplementary studies indicated that the effect of ARINA-1 on MCT was largely driven by the combination of bicarbonate with glutathione (ARINA-1 vs. PBS: Δ1.4±0.3mm/min, P<0.0001; bicarbonate+glutathione vs. PBS: Δ1.3±0.7mm/min, P<0.0001), and that ARINA-1 augmented the effect of lumacaftor (3µM, VX-809) and ivacaftor (10µM, VX-770) (LUMA/IVA) on MCT (ARINA-1+LUMA/IVA vs. PBS: Δ3.0±0.7mm/min, P<0.0001; LUMA/IVA vs. PBS: Δ0.26±0.1mm/min, NS). Assessment of CFTR Cltransport revealed an increase in Isc-dependent current in response to ARINA-1 in CFBE4o-cells expressing wild-type CFTR, largely driven by stimulation of CFTR activity by ascorbic acid, although this response was absent in CFBE41o-F508del cells, implicating CFTR-independent mechanisms for ARINA-1 effect on CF mucus. Overall, results indicate that ARINA-1 increases MCT rate in primary F508del HBE monolayers, suggesting its potential as a novel approach to improving mucus clearance in CF patients. This research was funded by the CFF. We also acknowledge the UAB CF Research Center.

Kramer, E. 1 ; Madala, S. 1 ; Hardie, W. 1 ; Hudock, K.M. 2, 3 ; Davidson, C. 1 ; Ostmann, A. 1 ; Strecker, L.M. 1 ; Clancy, J.P. 1 1. Pulmonary Medicine, Cincinnati Children's Hospital Medical Center, Cincinnati, OH, USA; 2. Pulmonary Biology, Cincinnati Children's Hospital, Cincinnati, OH, USA; 3. Pulmonary & Critical Care Medicine, University of Cincinnati, Cincinnati, OH, USA Introduction: Lung disease in CF is characterized by early airway obstruction, lung remodeling, and inflammation. Studies demonstrate that airway smooth muscle (ASM) dysfunction may be a primary defect contributing to airway hyperreactivity and obstruction in young CF patients. Transforming growth factor-beta (TGFβ) is a pleiotropic cytokine involved in lung remodeling, inflammation, and fibrosis. TGFβ also drives increased ASM tone and proliferation, potentially through release of cytokines such as IL-6. TGFβ is a genetic modifier and biomarker of CF lung disease severity, yet its pulmonary levels in very young CF patients and its mechanism of disease modification in CF are unknown. We hypothesize that CFTR deficiency impacts the pulmonary response to TGFβ, contributing to airway obstruction and ASM dysfunction.

Methods: Cytokine levels in bronchoalveolar lavage fluid (BALF) from CF patients (n = 15) and non-CF diseased control patients (n = 21, tracheotomized patients without significant parenchymal disease) under age 6 years were determined by ELISA or Luminex assay. CF (F508del homozygous) mice and non-CF littermate controls were intratracheally treated with an adenoviral vector containing the TGFβ1 transgene, PBS, or Empty vector. After one week, pulmonary mechanics were determined using the FlexiVent system and lungs were collected for analysis.

Results: Young CF and non-CF patients had similar low bacterial infection rates in BALF cultures (~15%) and similar % neutrophils (p=ns). Despite similar infection and BALF cellularity, young CF patients had elevated TGFβ levels compared to non-CF controls (45.5 vs 26.1 pg/mL respectively, p<0.05). TGFβ levels correlated with pan-elevation of inflammatory markers in non-CF patients, including human neutrophil elastase activity, IL-8, IL-6, and IL-1β. However, in CF patients, only IL-6 levels correlated with TGFβ, indicating unique inflammatory responses triggered in CFTR deficiency. CF mice exposed to subacute, low-dose TGFβ demonstrated increased pulmonary resistance and elastance compared to non-CF mice. A vigorous methacholine response in CF mice was maintained despite the higher baseline resistance. Histology demonstrated TGFβ-induced smooth muscle hypertrophy around smaller conducting airways in CF and non-CF mice.

Conclusions: TGFβ is elevated early in CF lung disease and is associated with increased IL-6 levels, but not other inflammatory mediators. IL-6 is implicated as a driver of ASM dysfunction in CF and other lung diseases and may be an early mediator of TGFβ's detrimental effects. CF mice exposed to TGFβ have greater abnormalities in lung mechanics compared to non-CF mice, indicating a role for CFTR in determining the pulmonary response to TGFβ. Low doses of TGFβ drive ASM thickening in small conducting airways in both CF and non-CF mice. These data implicate TGFβ in exacerbating the detrimental effects of CFTR deficiency on ASM function in early CF lung disease.

Acknowledgments: Supported by the CFF and University of Cincinnati & Cincinnati Children's CCTST CT2 Award (KL2TR001426).

Sendon, C. 1 ; Öz, H.H. 1 ; Di Pietro, C. 1 ; Barone, C. 1 ; Jamali, H. 1 ; Pierce, R. 1 ; Egan, M.E. 1, 2 ; Bruscia, E.M. 1 1. Pediatrics, Yale University, New Haven, CT, USA; 2. Cellular and Molecular Physiology Department, Yale University, New Haven, CT, USA Objectives: Epithelial cell permeability is maintained by cell-cell junction proteins. Caveolin-1 (CAV1), a scaffolding membrane protein, regulates cell junction protein expression. Emerging data show that cell-cell junctions are compromised in cystic fibrosis (CF) human bronchial epithelial cells (HBECs). Our data suggest that CAV1 expression is reduced in CF HBECs. We aim to test whether reduced CAV1 affects CF HBEC permeability and alters cell junction proteins expression.

Methods: We used CRISPR/Cas9-gene edited 16HBE14o-(16HBEge) cell lines carrying CFTR nonsense mutations (W1282X or G542X) in homozygosity (hereafter CF), and the parental 16HBE14o-cell line as isogenic wild-type (WT) control (provided by CFFT). Ussing chamber assay was used for validating absence of CFTR function in 16HBEge cell lines. Electric Cell-substrate Impedance Sensing (ECIS®) instrument was used to measure cell permeability in real time. Cell resistance was monitored either in the absence of inflammatory stimuli, or in response to 10 µg/mL lipopolysaccharide (LPS) (acute and chronic exposure). Cells were grown in ESIS 96-well arrays. Resistance was measured for 10 days to ensure formation of cell junctions, and measured continuously after LPS exposure for 96 hours. CAV1, tight-(Claudin 4, Occludin) and adherens-junction (E-cadherin and B-catenin) protein levels were assessed by Western blot.

Results: Ussing chamber analysis shows that CF W1282X and G542X cells displayed no response to forskolin/IBMX compared to parental WT cells, confirming absence of CFTR function. ECIS analysis revealed that all cell lines gained resistance over time. At 14 days of culture, WT cells showed statistically significant higher resistance compared to CF cells (WT=2027.9±123.3 Ω; CF W1282X=1494.9±35.2 Ω; CF G542X=1604.8±57.7 Ω). Short exposure to LPS (24 hours) minimally affected resistance in WT cells (drop of 43± 11.2 Ω). However, pronounced drop in resistance was seen in CF cells: W1282X =242.6± 89.3 Ω, G542X=126.2± 74.5 Ω. After chronic LPS exposure (96 hours), WT cells regained resistance, which did not differ from untreated WT cells (2154.9± 112 Ω). Both CF cells failed to regain resistance compared to WT treated cells (W1282X=1485.8±22; G542X=1644.4±35, P<0.001). Preliminary data suggest that, at steady-state, CF W1282X cells grown in monolayers and at air-liquid interface have decreased expression of CAV1 and adherens junction proteins (E-cadherin and B-catenin) when compared to WT cells. No differences were observed in levels of tight junction proteins (Claudin 4 and Occludin).

Conclusions: Our preliminary data suggest that loss of CFTR function impairs cell barrier permeability at steady state, and that chronic LPS treatment intensifies this difference between WT and CF cells. We observed that CF W1282X cells displayed decreased expression of CAV-1 and adherens junction proteins compared to WT cells, which may explain the barrier permeability dysfunction. The direct effect of reduced CAV-1 in decreasing the expression of adherens junction proteins in CF cells remains to be evaluated. primary monocytes and HBECs, with and without various small molecule inhibitors. CF monocytes and HBECs were all hyper-responsive to NLRP3 inflammasome activation and inhibiting ENAC, with amiloride or the SPLUNC1-derived S18 peptide, reduced this hyper-responsiveness. We found elevated Na + influx and K + efflux (both are inflammasome activators) post-LPS and ATP in CF monocytes and HBECs. To understand the impact of Na + and K + flux on glycolysis, we measured ATP, glucose, succinate and L-lactate levels, and all elevated in these cells.

Conclusions: Collectively, our findings reveal novel mechanisms underlying the exaggerated inflammatory response, independent of infection, present in CF. We show that CF primary monocytes and macrophages are skewed towards a pro-inflammatory phenotype and increased Na + influx, via ENaC, contributes to metabolic reprogramming and NLRP3 inflammasome activation in CFTR mutated cells. Inhibiton of ENaC with either small molecules or peptides was capable of regulating these perturbations of the innate immune response in vitro. Previous studies suggest that these drugs may accumulate intracellularly, which has implications for both treatment and precision medicine strategies.

Methods: Primary F508del homozygous HBE obtained from the UNC cell culture core were grown at air-liquid interface as previously described. HBE were treated with 1 µM VX-770 (ivacaftor) and 3 µM of either VX-809 (lumacaftor) or VX-661 (tezacaftor) in basolateral media for fourteen days, followed by a 14-day washout period. Electrophysiological responses (by Ussing chambers) and intracellular drug concentrations (by mass spectrometry) were measured at 1, 7, and 14 days in the treatment phase and again after 1, 7, and 14 days of washout (15, 21, and 28 days after treatment start). Drug concentrations were also measured from nasal epithelial biopsies (n=3) obtained from patients on ivacaftor therapy as well as nasospheres and cultured epithelia derived from these samples.

Results: In treated HBE, VX-770, VX-809, and VX-661 accumulated intracellularly within one day, with prolonged exposure leading to increased concentrations of VX-770, but not VX-809 or VX-661 ( Fig 1A) . During washout, VX-809 and VX-661 dropped relatively quickly and were undetectable after 7 days, whereas intracellular VX-770 remained elevated even after 14 days of washout. In electrophysiological studies, treatment-induced increases in peak forskolin current were similar throughout the treatment phase and declined during washout though remained above baseline after 14 days of washout ( Fig 1B) . Preliminary results from nasal epithelial biopsies suggested that intracellular concentrations were well below those measured in HBE but persisted in nasospheres (assessed ~4 days post-collection) though not cultured epithelial cells (assessed ~3 weeks post-collection).

Conclusions: CFTR modulators accumulate intracellularly in HBE, with concentrations of VX-770 increasing with prolonged exposure and persisting during washout. While VX-770 in nasal epithelia from patients on ivacaftor therapy was lower than HBE, drug persisted in nasospheres assessed within a few days of collection. These findings suggest that dosing strategies based on serum pharmacokinetics may not accurately predict CFTR modulator effects in target tissues. Furthermore, the persistence of VX-770 has significant implications for treatment and precision medicine studies of patients on this therapy. There is growing evidence to suggest that NE independently contributes to impaired macrophage phagocytic function. The goal of this study is to evaluate the potential targets of NE protease activity to determine their role in phagocytic function.

Methods: A murine macrophage cell line (RAW 264.7) was cultured and exposed to either 500nM NE or 500nM inactive NE for 2 hours, followed by treatment with an NE inhibitor. The media was collected and concentrated to 1 mL. The protein media samples were enzymatically digested with trypsin via filter assisted sample preparation protocol for LC-MS/MS analysis in technical triplicates. Proteomics datasets were processed in MaxQuant using the Andromeda search algorithm to identify proteins and Perseus to perform relative label free quantification (LFQ). Differentially expressed proteins were analyzed using on-line tools such as UniProt and PeptideCutter to identify functionally significant proteins altered with NE treatment and verification of NE cleavage. After proteomic analysis, a list of target proteins was generated and complementary Western blot analysis was performed.

Results: A total of 841 proteins were identified that were perturbed with NE treatment. Of special interest were proteins involved in phagocytosis, including cell surface receptor proteins for phagocytic and apoptotic cell targets. Also identified were proteins implicated in the cytoskeletal rearrangements necessary for phagocytosis such as phagocytic cup formation, phagocytic synapse formation, and membrane extension during phagosome formation. These are summarized in the Table. All the fragments identified for these peptides were found in extracellular domains.

Conclusions: There are dramatic alterations in the culture media proteome of RAW 264.7 cells when exposed to NE. These perturbations reflect a combination of degraded cell surface proteins implicated in phagocytosis and phagocytic target recognition. We also found unexpected plasma membrane targets that may be required for phagocytic function such as those that orchestrate the necessary cytoskeletal rearrangements. Future directions include evaluating shed/secreted proteins by an analysis of NE effects on biotinylated plasma membrane proteins as well as confirmation of the role of targets by performing siRNA or pharmacologic inhibition of receptors. Nonresolving inflammation and infections are key underlying mechanisms of progressive loss of respiratory function in patients suffering with cystic fibrosis (CF). Resolution of inflammation is an active process in which endogenous chemical mediators, such as resolvin (Rv) D1, promote the termination of the inflammatory response by acting on ALX/FPR2 and GPR32 receptors. Here, we investigated bioactions of RvD1 in chronic lung inflammation and infection by a clinical strain of P. aeruginosa (PA-RP73) in preclinical model of CF. Then, we assessed resolution of inflammation in mice overexpressing hALX/FPR2 (Fpr2/KI) upon PA-RP73 infection. To this end, RP73 agar beads were injected into lungs of C57Bl6/N and Fpr2/KI mice. RvD1 (100 ng/mouse) or vehicle were administered via intragastric gavage. Disease severity, inflammation resolution, and bacterial load were evaluated at 21 days post-infection (DPI) through the analysis of: 1. clinical signs; 2. total and differential bronchoalveolar lavage (BAL) leukocyte counts; 3. lung morphology; 4. PA CFU counting. Total RNA and miRNA were isolated from sorted lung macrophages and used to perform real time PCR. Inoculum of PA-agar beads produced a nonresolving pulmonary inflammation in mice that recapitulates histological and biochemical features of human pathology. Administration of RvD1 significantly reduced airway PA titer (2.99±2.76x10 5 CFU (RvD1) vs. 2.03±0.13x10 6 CFU (vehicle)), neutrophil infiltration (T 50 = from 12 to 10 DPI) and tissue pathology. RvD1 treatment significantly enhanced killing of PA and apoptotic neutrophils by murine lung macrophages (P<0.05). In murine macrophages RvD1 regulated expression of microRNAs miR-21 (P<0.001) and 155 (P=0.029) and downstream genes (toll-like receptors, NF-κB) resulting in reduced inflammatory signaling. Moreover, we observed a greater ability to resolve PA pulmonary inflammation in Fpr2-KI mice compared with wild-type, highlighting the protective role of the receptor in lung inflammation. Indeed, PA infected Fpr2-KI mice showed an enhanced bacterial clearance by 30% and macrophages (P=0.03) in BAL. Collectively, these results unveil beneficial effects and mechanisms of action of RvD1, providing evidence for its anti-inflammatory and proresolution actions in chronic lung inflammation and infection. This is consistent with the observed protective function of RvD1 receptor, confirming its relevant involvement in the resolution process. These findings also point to RvD1 as a prototype of innovative therapeutic strategy for CF based on the exploitation of endogenous resolution mediators. Introduction: Cystic fibrosis is a genetic disease characterized by progressive pulmonary inflammation and infection that leads to declining lung function and ultimately death. Previous studies have shown elevated markers of inflammation, such as cytokine IL-8, in the lungs of individuals with CF that was out of proportion to the concurrent level of infection. It is known that the tissues of CF subjects are characterized by elevated inflammatory markers, but there is debate as to whether this is secondary to a pro-inflammatory infectious environment or due to an inherent feature of cells lacking the CFTR protein. The objective of this study was to use an isogenic cell line to determine the contribution of CFTR to an inflammatory reaction.

Methods: The CRISPR/Cas9 gene editing system was used to create an isogenic cell model of CF from an intestinal epithelial adenocarcinoma cell line, Caco2. Clonal cell lines were propagated and a panel of 8 lines with intact, functional CFTR was compared to a panel of 9 with all alleles of CFTR inactivated by frameshifting indels. These lines all maintained CFTR mRNA production and formation of tight junctions, and those with intact CFTR displayed short circuit currents in response to forskolin, while the CFTR-null lines did not. Western blots verified that protein was present in cells with intact CFTR and was not detected in CFTR-null clones using two separate CFTR antibodies.

Results: Baseline mRNA expression of inflammatory cytokine genes IL6 and CXCL8, the gene encoding IL-8, was not different between the CFTR-null and CFTR-intact cell lines. Following separate stimulations with TNFα and IL1β, all cells responded with increased CXCL8 and IL6 mRNA expression as measured by RT-PCR. Both stimuli caused higher and more varied expression of CXCL8 mRNA in the CFTR-null cells compared to CFTR-intact cells. TNFα at 100ng/mL increased CXCL8 70.88 ± 37.99 fold in wild-type (WT) lines, but 194.31 ± 117.94 fold in CFTR-null cells. IL1β at 0.1ng/mL increased CXCL8 915.81 ± 609.8 fold in WT lines, but 1258.74 ± 1007.08 fold in CFTR-null cells. There was not a significant difference in IL6 production between CFTR intact and null cell lines using either stimulus. Transcriptomes (RNA-seq) from a subset of these lines at baseline and then after TNFα stimulation showed 402 and 264 genes differentially expressed according to CFTR genotype. Gene set (pathway) comparisons were carried out and only one pathway, oxidative phosphorylation, was different in the baseline state, but 9 sets emerged upon stimulation. These data suggest that the absence or loss of CFTR leads to epigenetic differences between CF and non-CF cells, but also that gene and environment interactions are involved in eliciting CF phenotypic differences. Figure: (A,B) Representative Muc5ac and Muc5b airway staining images from WT and MyD88 KO mice, n=7 mice per condition. Representative immunoblot from lung homogenates for Muc5ac (C) and Muc5b (D) with corresponding quantification of band intensity using total protein levels, n=8 mice per condition. (E) RNA expression by qPCR measured by fold change over media control condition, n=12 mice per condition. (F) Immunoblots for Muc5ac from WT and MyD88 KO mice challenged with OVA, with corresponding quantification of band intensity using total protein levels, n=3 mice per condition. Center, Univ of North Carolina, Chapel Hill, NC, USA; 2. Physics and Astronomy, Univ of North Carolina, Chapel Hill, NC, USA CF lung disease arises from a muco-inflammatory state in distal airways that initiates a cycle of infection and airway obstruction. Bronchoalveolar lavage fluid (BALF) is the only current method of studying the biophysics and biochemistry of the distal airway surface liquid (ASL) in early disease. Of particular interest are muco-inflammatory markers including mucins MUC5B and MUC5AC and DNA which are often recovered in the form of unswellable mucus "flakes" that can be found in both health and CF. While the lavage inherently dilutes the ASL milieu, we show that 53% of the mucus-forming mucins in CF are present as insoluble "flakes" or "rafts" that are observable via microscopy. Because a majority of mucins are insoluble, we hypothesize that studying these flakes will yield insight regarding the etiology of CF lung disease. Flakes in BALF present a number of research obstacles mostly arising from the difficulties of extracting the biophysical signals attributable to flakes from the background of lavage fluid. Here, we present a novel methodology that combines particle tracking microrheology (PTMR) and Gaussian Mixture Modeling (GMM) to isolate and study the biophysics and clinical implications of these flakes.

We collected BALF from 29 patients with CF (age 10.5±4.4 years) and 6 healthy adults. We use 1 µm fluorescent polystyrene beads to probe flake rheology via PTMR. Samples were scanned for flakes where bead motion was tracked at 60 fps for 30 seconds. Values of η* were clustered into "watery" and mucus components via GMM. The local complex viscosity (η*, Pa•s) described both the viscous and elastic properties of flakes.

Comparison between η* from bulk (i.e. cone and plate) rheology and PTMR in each CF sample revealed a moderate correlation (r=0.42, p<0.05). Comparing only the mucus component η* from PTMR enhanced the correlation (r=0.60, p<0.01), indicating bulk rheology reflects the thickest components of a heterogeneous sample. Comparing CF to healthy adults, weighted η* (i.e. the sum of mean η* in water or mucus multiplied by the respective GMM mixing proportion, p w or p m ) was increased in CF (log 10 (η*)=-0.96) vs. non-CF (-2.5, p<0.01) . The mean log 10 (η*) of the mucus component was not significantly increased, but the span of η* in the mucus component was ~3 logs larger in CF samples. Furthermore, p m of mucus was increased in CF (p m =0.83) vs. non-CF (p m =0.27) samples (p<0.01). Taken together, these results imply that while flakes appear in both BALF from patients with CF and healthy adults, there are more mucus flakes in BALF from patients with CF and that they can display hyperconcentration relative to non-CF samples.

Inhaled hypertonic saline treatment is used to improve lung health in patients with cystic fibrosis (Elkins MR, Bye PT. Curr Opin Pulm Med. 2006; 12:445-52; Elkins MR, et al. N Engl J Med. 2006; 354:229-40; Donaldson SH, et al. N Engl J Med. 2006; 354:241-50) . The current consensus is that the treatment generates an osmotic gradient that draws water into the airways and increases the airway surface liquid volume, thus improving mucociliary clearance (Tarran R, et al. Mol Cell. 2001; 8:149-58; Goralski JL, et al. Eur Respir J. 2018; in press) . However, there is evidence that hypertonic saline may also stimulate active secretion of airway surface liquid by airway epithelia through the activation of sensory neurons. Since the exact mechanism of action of hypertonic saline is not fully understood, it is difficult to develop procedures to modulate the duration and intensity of the treatment in order to improve hypertonic saline treatment outcome.

The objective of this study is to test the contribution of airway neurons and epithelia on hypertonic saline-stimulated airway surface liquid production in the trachea of cystic fibrosis and wild-type swine using a novel synchrotron-based imaging method. Our results showed that hypertonic saline (7% NaCl) treatment increased airway surface liquid in live wild-type swine. Blocking the nervous system with intramuscular atropine, to block the parasympathetic stimulation of airway gland secretion, and topical lidocaine, to block sensory fibers, reduced the effect of hypertonic saline treatment by ~50% in vivo. Incubating isolated trachea ex vivo preparations from wild-type swine with known inhibitors of ion transport across airway epithelia (i.e., CFTRinh172, bumetanide and niflumic acid in bicarbonate-free saline) also resulted in a 50% reduction in the effect of hypertonic saline treatment. Furthermore, cystic fibrosis swine (CFTR -\-) trachea ex vivo preparations also showed substantially decreased secretory response to hypertonic saline treatment after blockage of neuronal activity with atropine and lidocaine. Our findings indicate that hypertonic saline nebulization treatment-triggered airway surface liquid secretion is mediated in part by the stimulation of airway neurons and the subsequent activation of active epithelia secretion; only ~50% of the secretion is the result of osmotically driven fluid production.

Supported by Cystic Fibrosis Canada and Canadian Institutes of Health Research.

to determine the contribution of ion channels and transporters, as well as signalling pathways in the regulation of ASL pH.

Methods: The ASL of polarised CF and non-CF primary human airway epithelial cells (HAECs) was stained with a mixture of dextran-coupled pHrodo, a pH-sensitive fluorescent dye (0.1 mg/mL) and the dextran-coupled pH-insensitive Alexa Fluor TM 488 (0.1 mg/mL), diluted in a modified Krebs solution, overnight (37°C, 5% CO 2 ) to re-establish thin-film condition. The next day, readings were performed every 5 minutes in a temperature and CO 2 equilibrated plate-reader. Agonists and inhibitors were added either with the dyes if their target was located apically or after 2 hours of baseline readings if their target was on the basolateral membrane. The effects of agonists and inhibitors were also measured on short-circuit current (Isc) in Ussing chambers either in the absence of Cl -, or under a basolateral to apical HCO 3 gradient. Results: A rise in [cAMP]i (using 10 µM forskolin) alkalinised the ASL of non-CF HAECs (t 1/2 = 35 min) which was blocked by the PKA inhibitor H89 and the CFTR inhibitor CFTRInh172. Surprisingly, resting ASL pH was not different between CF and non-CF HAECs and CFTRInh172, in the absence of forskolin, did not affect resting ASL pH of non-CF HAECs. This result was correlated to an absence of response in resting Isc to CFTRInh172 in Cl --free or HCO 3 gradient conditions. In non-CF HAECs, carbonic anhydrase and apical Na + -H + -Exchanger were not involved in resting nor forskolin-induced ASL pH regulation but inhibiting the Na + -Bicarbonate Cotransporter (NBC), using 30 µM S0859, reduced resting and forskolin-induced increases in ASL pH. Interestingly, the inhibition of the basolateral Anion Exchanger 2 (AE2) significantly increased intracellular pH as well as ASL pH. The latter alkalinisation could be potentiated by the addition of FSK.

Conclusion: Our technique enables stable ASL pH measurements to be obtained from primary HAECs under thin film conditions, and gives an insight into the kinetics of the responses to different agonists and inhibitors. An increase in cAMP alkalinised the ASL in non-CF cells which was CFTR dependent. Importantly, we show for the first time the lack of involvement of CFTR in resting ASL pH regulation. Our results also identified NBC and AE2 as potential new therapeutic targets for ASL pH modulation. Introduction: The cystic fibrosis transmembrane regulator (CFTR) is a transmembrane protein, involved in the transport of bicarbonate and chloride ions. In cystic fibrosis (CF) the CFTR is mutated, leading to an absence or reduction in the CFTR function, resulting in abnormal airway surface liquid, recurrent pulmonary infections, and inflammation. The CFTR is assembled and modified in the endoplasmic reticulum (ER), and the accumulation of misfolded or unfolded protein leads to activation of the unfolded protein response (UPR). The most common mutation in CF, ΔF508, results in misfolding of the CFTR and accumulation within the ER. The UPR comprises three signaling pathways involving three ER transmembrane proteins, known respectively as PERK, IRE1α, and ATF6. UPR activation, through its IRE1α arm, has been linked to production of proinflammatory cytokines, including IL-6 and TNF. The aim of this study was to investigate the level and extent of abnormal UPR activation, mainly IRE1α, and proinflammatory cytokine production in patients with CF and also in cell lines harbouring different classes of CFTR mutations.

Methods: Patients' peripheral blood mononuclear cells (PBMCs), primary monocytes, and four human bronchial epithelial cell (HBEC) lines Beas-2b (wild-type), IB3-1 (ΔF508/W1282X), CuFi-1(ΔF508/ΔF508), CuFi-4(ΔF508/G551D), were used to evaluate the activation of the UPR, using quantitative real time PCR (qPCR) and flow-cytometry. Production of IL-6 and TNF was measured by qPCR and ELISA assays. Lipopolysaccharide (LPS), tunicamycin (Tn), thapsigargin (Tg) and 4m8c were used as cellular UPR stimulants or inhibitors to assess UPR activation.

Results: Gene expression revealed a significant increase in IRE1α, XBP1s, and IL-6 levels in CuFi-1 and CuFi-4 HBEC lines, PBMCs, and also in primary monocytes from CF patients, after LPS, Tn, and Tg stimulations. IRE1α protein expression was increased in the three CF cell lines; however, phosphorylation of IRE1α was increased only in the CuFi-1 and CuFi-4 HBEC. IL-6 secretion was significantly increased in all CF cell lines, as well as in PBMCs, and primary monocytes from CF patients. Finally, pre-treatment with the IRE1 inhibitor, 4m8c, abrogated IL-6 production after LPS, Tn, and Tg stimulation in the HBECs, but not in the human monocytes.

Conclusion: Our data suggest a low grade chronic overactivation of the UPR in CF, which predisposes CF cells to overproduction of pro-inflammatory cytokines when intracellular stress is induced. The overproduction of IL-6, in HBECs, is IRE1α dependent, as IL-6 production was reduced when the IRE1 inhibitor was used. We have shown that the UPR is implicated in driving IL-6 overproduction in CF innate immune cells, and this overproduction may contribute to the abnormal inflammatory phenotype seen in patients with CF. 2 1. Cell Biology, Erasmus MC, Rotterdam, NL, Netherlands; 2. Sophia Children's Hospital, Ped. Pulmonol., Erasmus MC, Rotterdam, Netherlands; 3. Center for CF and Airway Disease, Emory University, Atlanta, GA, USA Introduction: Current evidence shows that CF lung disease is initiated in CF infants before the onset of bacterial colonization, suggesting an innate, CFTR-dependent, mechanism as the trigger for early inflammation and tissue remodeling. Previous studies in our laboratory showed that the EGFR/ ADAM17 axis, which controls the shedding of growth factors and cytokines involved in inflammation and tissue remodeling, is more active in primary differentiated CF vs. non-CF airway epithelial cultures (HBEC-ALI) (Stolarczyk M, et al. Am J Physiol. 2018; 314:L555-68) . Since ADAM17 is activated by oxidation of its extracellular proteolytic domain, and substantially reduced by supplementing the medium with glutathione (GSH), this is consistent with a more oxidized extracellular milieu in CF airway epithelia, likely due to deficient GSH transport. We hypothesize that CFTR-related oxidative stress contributes to the progression of early CF lung disease by activating epithelial pro-inflammatory and pro-fibrotic signaling.

Methods: Bronchoalveolar lavage fluid (BALF) and chest CT-scans from children with CF (age 0-5 years, n=23) enrolled in the Rotterdam early CF monitoring program were collected sequentially at 1, 3 and 5 years. Chest CT scans were scored according to PRAGMA-CF method for total disease (%DIS) and bronchiectasis (%BX). We analyzed in cell-free BALF supernatants bioactive lipids responsive to oxidative stress using quantitative HPLC-MS/MS (LACDR, Leiden), neutrophil elastase (NE) activity using a fluorometric assay (Cayman), and myeloperoxidase (MPO) by immune-capture and Amplex red oxidation assay. Cytokines and growth factors involved in inflammation and tissue remodelling were monitored using a Fluidigm-based 96x96 protein array (OlinkTM Immuno-Oncology).

Results: BALF lipid markers associated with oxidative stress, including isoprostanes (8-iso-PGE2) and lysolipids (lysolecithin and lysophosphatidic acid species), correlated positively (Spearman rho > 0.55, P <0.01) with PRAGMA-CF scores (%DIS and %BX), and inflammatory markers (% BALF neutrophils, NE and MPO activities). More than 40 of 92 growth factors and cytokines involved in myeloid cell migration and cell proliferation correlated positively with both lysolipids and isoprostanes. In CF HBEC-ALI culture, shedding of several of these factors was significantly reduced (>2-fold, P<0.001; N=3) by GSH supplementation (AREG, CX3CL1, CCL20, CXCL5, CXCL1, IL8, CX3CL1, TNFRSF12A, VEGFA, CXCL10) or enhanced (>2-fold, P<0.001; N=3) by LPAn, a lysolipid receptor agonist (AREG, IL6, IL8, TRAIL, MCP-1, CCL20, VEFGA, CXCL10).

Conclusions: Bioactive lipid markers in BALF from CF children correlate with early CF lung disease, and may be useful to monitor progression of disease end efficacy of interventions. Together, our data suggest an active role of oxidative stress and lysolipid signalling in the onset and progression of early CF lung disease. Early intervention targeting these pathways may be beneficial to CF patients.

The use of molecular biology approaches has re-emerged as potential therapeutic modalities for cystic fibrosis. Primary cultures of human bronchial epithelial (hBE) cells grown on permeable supports are often used to evaluate the efficacy of these various molecular approaches including the delivery of: nanoparticles, mRNA, DNA, tRNA, viral vectors and small oligonucleotides. In most experiments these molecular reagents are delivered to the apical surface of the hBE cultures in some volume of different compositions and a functional assay of anion secretion performed some time thereafter. The studies reported here were designed to evaluate the volume and time dependence of additions to the apical surface on the secretion of chloride in homozygous F508del-CFTR cells. Equivalent current (Ieq) measurements were made using a MTECC-24 on cells grown for 35 days at an air-liquid interface. Cells were washed four days before the experiments were initiated. PBS (phosphate buffered saline) was added to the apical surface in volumes ranging from 25 to 125 mL. Cells were treated with correctors: C17, C18, C17 plus C18 or DMSO for one to four days. Ieq measurements were made one to four days after the apical volume additions with and without the correctors. The results document volume and time dependent effects on the chloride secretion responses to forskolin and VX-770 stimulation. When measured the next day after the addition of 25 mL of PBS, chloride secretion was stimulated compared to untreated (no volume) controls. However, by day two this stimulatory effect was no longer evident. At 50 mL PBS and above chloride secretion was inhibited on day one and did not return to the level of untreated (no volume) controls for two to three days depending on the volume added. Notably with 75 mL and higher the chloride secretory response was substantially elevated in the corrector treated cells suggesting PBS was acting to up-regulate the effects of the correctors. Indeed, PBS is at least as efficacious as any of the "second mechanism" correctors we have evaluated in the Bridges lab. Future studies will be directed at understanding the mechanism of action of apical PBS on elevating the chloride secretory response. We speculate this mechanism could be used to advantage in the treatment of CF. The results also document the importance of performing appropriate controls when attempting to evaluate molecular reagents delivered to the apical surface. Performed too early after an addition a positive effect of a given reagent could be inhibited by the added volume, performed with a larger volume will require longer duration exposures and small effects could be overshadowed by the volume effect. As always, "Success is in the details." Supported by the CFF. 1 1. Medicine, Univ of Washington, Seattle, WA, USA; 2. Cancer Research, Univ of Chicago, Chicago, IL, USA; 3. St. Vincent's Univ Hospital, Dublin, Ireland Introduction: Evidence is accumulating that abnormal immune responses contribute to CF pathologic airway inflammation, which decreases lung function without eliminating chronic bacterial infections. We sought to investigate how monocyte responses may be impaired in people with CF. In prior work, we used proteomics to study blood monocytes from a cohort of 12 people with CF and G551D CFTR mutations before and 7 days after starting the CFTR potentiator ivacaftor. This study design, which compared cells from the same individuals before and after restoration of CFTR function, identified highly statistically significant changes in the abundance of 21 monocyte proteins despite a small cohort size. The data from this study suggested that monocytes isolated from subjects pre-ivacaftor have heightened responses to the inflammatory cytokine IFNγ compared to cells recovered after initiation of ivacaftor.

Objective: We hypothesized that CFTR dysfunction causes exaggerated IFNγ responses in monocytes.

Methods: We obtained blood monocytes from a second cohort of people (n= 10) with CF and ivacaftor-susceptible CFTR mutations (R117H) before and after starting ivacaftor. Isolated monocytes were plated with and without IFNγ, and then lysed for isolation of RNA or protein. Monocyte IFNγ-induced gene expression before, 2 and 7 days after initiation of ivacaftor was determined by quantitative RT-PCR. Western blots measured relative abundance of STAT1 and phosphorylated STAT1, a key mediator of the intracellular IFNγ signaling cascade. RNAseq was performed to quantify changes in basal gene expression of ex vivo monocytes before and after initiation of ivacaftor. Monocytes were analyzed by flow cytometry to detect changes in surface markers associated with cellular activation state. Cytokine levels were measured in plasma recovered from subjects pre-and post-ivacaftor.

Results: Monocytes isolated 7 days after subjects began ivacaftor therapy demonstrated significantly less induction of gene expression when exposed in vitro to IFNγ. Dampening of monocyte IFNγ-induced gene expression was detected as early as day 2 after initiation of ivacaftor. STAT1 protein levels were unchanged following initiation of ivacaftor, but reductions in IFNγ-induced gene expression were accompanied by marked decreases in IFNγ-induced phosphorylation of STAT1. RNAseq identified approximately 50 monocyte genes for which basal expression was significantly changed after 7 days of ivacaftor, the majority of which were increased in expression post-ivacaftor.

Conclusions: Here we demonstrate that ivacaftor therapy dampened inflammatory responses of peripheral blood monocytes, which could contribute to the decreased symptoms and improved lung function experienced by people with CF receiving ivacaftor. Future studies will focus on determining whether changes in immune cell function are due to ivacaftor increasing CFTR function on monocytes, ivacaftor acting via an off-target effect to modulate monocyte responses, or ivacaftor acting on other cell types to decrease lung and systemic inflammation.

Objectives: Abnormal chloride (Cl -) transport dehydrates airway surface liquid (ASL) in sinonasal epithelium leading to mucus stasis and chronic rhinosinusitis. As an experimental epithelium, rabbit tissue provides an excellent representation of human sinus disease, and the rabbit sinusitis model is established and well suited for therapeutic interventions in vivo. The objective of this study is to evaluate whether ivacaftor reverses the consequences of Pseudomonas aeruginosa-induced acquired CFTR dysfunction.

Methods: Rabbit nasal cavities were evaluated for responsiveness to ivacaftor in vivo (nasal potential difference [NPD] assay). Rabbit nasal epithelial (RNE) cultures were incubated with an ultrafiltrate of P. aeruginosa (PAO1 strain) for 4 hours and tested for acquired CFTR dysfunction. Markers of mucociliary function, including airway surface liquid depth (ASL), periciliary liquid depth (PCL), and ciliary beat frequency (CBF) were measured by micro-optical coherence tomography (µOCT) following PAO1 and/or ivacaftor incubation.

Results: Ivacaftor resulted in a significant 20.3+/-0.1 mV mean NPD polarization that was significantly different than low Clcontrol (12.8+/-1.3; p<0.01). PAO1 exposure induced a state of acquired CFTR deficiency in rabbit nasal epithelium as measured by forskolin-stimulated short-circuit current (I SC ) (control, 37.0 ± 1.1 µA/cm 2 vs PAO1, 26.1 ± 1.1 µA/cm 2 ; p < 0.001). RNE cultures exposed to PAO1 inhibited ASL depth and CBF, whereas co-incubation with ivacaftor restored these markers of mucociliary clearance (p<0.001) as measured by µOCT.

Conclusion: Ivacaftor robustly stimulates CFTR-mediated Clsecretion in RNE, and normalized ASL and CBF in PAO1-induced acquired CFTR dysfunction. Testing CFTR potentiators in P. aeruginosa rabbit sinusitis are planned.

Methods: Primary and immortalized human bronchial epithelial (HBE) cells were obtained from a G551D patient. Human nasal epithelial (HNE) cells were collected from normal and CF subjects. We used these cell systems to study the impact of CFTR activity on mucin properties using biochemical assays (Western blotting, HPLC, IHC), biophysical measurements (macro/microrheology) and scanning electron microscopy (SEM). To isolate effects on individual airway mucins, we pharmacologically inhibited CFTR function in Calu3 cells genetically modified to secrete only MUC5B or MUC5AC, but not both.

Results: As expected, chronic treatment with ivacaftor in G551D primary and immortalized cell cultures resulted in enhanced CFTRmediated short-circuit currents. In parallel, ivacaftor treatment significantly decreased total mucin concentration by ~30%, increased mucus pH by +0.3, and increased ciliary beat amplitude by ~30%, suggesting a physical change in the mucin network surrounding the cilia. Conversely, CFTR inhibition in non-CF HNE cells led to mucin hyperconcentration (1.5-fold change), altered the MUC5B/MUC5AC ratio, and resulted in a notably tighter mucus network by SEM. In G551D models, CFTR rescue decreased MUC5B but not MUC5AC crosslinking as shown by electrophoretic mobility shift post-DTT treatment. In Calu3 cells expressing only MUC5B, CFTR inhibition increased MUC5B concentration and molecular weight, suggesting CFTR function affects mucin multimerization. Similar experiments with MUC5AC-expressing cells are underway. In Calu3 cells individually expressing either MUC5B or MUC5AC, SEM revealed striking differences in their mucin network organization (eg, interwoven vs branching network).

Conclusion: Functional CFTR rescue affected mucin interactions by decreasing mucin entanglement and crosslinking. In contrast, CFTR inhibition increased mucin concentration and molecular weight. MUC5B generally appeared more sensitive than MUC5AC to changes in CFTR function. Since CFTR dysfunction alters the mucin ratio and network organization, both likely affect the overall biophysical properties of airway mucus.

Acknowledgments: Supported by Vertex RIA and CFF. Background: Currently, the mechanisms that underlie the maladaptive formation of bronchiectasis seen in patients with cystic fibrosis are unknown. Prior studies suggest that epithelial repair is dysregulated in individuals with cystic fibrosis, and that aberrant epithelial repair plays a role in the development of bronchiectasis. It has been demonstrated that airway basal cells (BC) have the potential to generate both ciliated and secretory epithelial cells, serving as a proximal airway progenitor cell. Therapies aimed at regulating basal cell proliferation and directing their differentiation toward specific lineages may reverse the disease process and halt the detrimental effects of end-stage lung disease. We hypothesized that airway BC are primed to regenerate injured airway epithelial cells; specifically, epithelial injury triggers changes in the transcriptome within basal cells which allows the regeneration of lung epithelium.

Research Design and Methods:Krt5 CreER(T2) transgenic mouse line in combination with R26R EYFP reporter is used to lineage trace basal cells. Cre recombinase is activated with tamoxifen which results in YFP expression. Experimental mice are treated with naphthalene, which results in depletion of epithelial cells, while control mice receive standard corn oil vehicle. FACS is utilized to isolate YFP+ basal cells 10 days post-naphthalene injury, at which time we can capture basal-cell contribution to epithelial regeneration. Next generation RNA sequencing is performed to characterize differences in gene expression between control BC, and BC after injury.

Results: We used RNA sequencing to compare the transcriptome changes from BC of control and experimental mice after naphthalene injury. Genes commonly associated with basal cells are found to be downregulated after naphthalene injury compared to control (KRT5, KRT14, TP63, NGFR, PDPN). We identified 715 up-regulated genes and 89 down-regulated genes using criteria of log fold-change >2 and adjusted p-value of <0.05. Signaling pathway analysis revealed the activation of pathways including inflammatory signatures such as cytokine-cytokine receptor interactions, natural killer cell mediated cytotoxicity, NF-kappa B signaling, apoptosis, as well as pathways implicated in lung epithelial development such as MAPK signaling, Hedgehog signaling, Notch signaling, and Wnt signaling. Conversely, down-regulated pathways include phosphatidylinositol signaling, VEGF pathways, TGF-beta signaling, and ErbB signaling.

Conclusions: This unbiased transcriptomic study of airway basal cell after injury provides a panel of novel basal cell genes for analysis in a murine airway injury model. Future directions include validation of specific signaling modulators using both quantitative PCR of RNA as well as testing the growth condition of basal cells along with specific inhibitors and agonists in organoid assays. We hope to extend the findings to human model utilizing basal cells isolated from both healthy human controls and patients with cystic fibrosis to determine whether novel regenerative pathways are dysregulated in the setting of human disease.

Kelley, T.; Rosenjack, J. Pediatrics, CWRU, Cleveland, OH, USA Introduction: Finding effective anti-inflammatory therapies is an important clinical goal. Ibuprofen is the standard of care for anti-inflammatory therapies. Recent work has even demonstrated that ibuprofen use improves long-term survival. However, relatively few patients use ibuprofen therapy due to potential adverse effects. Finding other interventions within the CF inflammatory cascade could provide the same therapeutic benefit of ibuprofen without the adverse events that limit use. We have previously identified microtubule alterations in CF cells including reduced acetylation and slower rates of reformation. These changes impair intracellular transport of endosomes leading to pro-inflammatory signaling. Inhibition of histone deacetylase 6 (HDAC6), a cytosolic deacetylase that regulates tubulin acetylation, restores intracellular transport and dampens inflammatory signaling. We also identified that ibuprofen treatment normalizes CF microtubule regulation resulting in more stable microtubules, improved intracellular transport, and related anti-inflammatory signaling. The similar effects of ibuprofen treatment and HDAC6 inhibition at the cellular level lead to the hypothesis that HDAC6 inhibition is an alternative to ibuprofen as an anti-inflammatory therapy.

Methods: To test this hypothesis, we developed an F508del Cftr mouse model (CF) with the expression of Hdac6 knocked-out (CF/Hdac6). We have previously published that the CF/Hdac6 mice exhibit improved growth and weight gain compared to sibling CF mice. Wild-type (WT), CF, Hdac6 -/-, and CF/Hdac6 mice were challenged with 25,000 CFU Pseudomonas aeruginosa (PA) embedded on agarose beads via intratracheal delivery. Changes in weight were monitored over six days post-infection. Bronchoalveolar lavage (BAL) fluid was collected in different groups at 1, 3, or 6 days post infection and analyzed for cell counts, recovered PA CFU, and cytokine production. Both male and female mice between 8-10 weeks of age were tested.

Results: Weight loss in response to infection was identical between CF and CF/Hdac6 mice for days 1 and 2 with both groups losing significantly more weight than WT mice. Beginning on day 3, however, CF/Hdac6 mice begin to rapidly recover and regain weight to WT levels by day 5. CF mice recover much more slowly and have significantly higher weight loss at days 3 through 6 (day 6: WT: -3.0 ± 0.7% body weight; CF: -6.4 ± 1.4%*; CF/Hdac6: -1.9 ± 0.9%; *p < 0.05 compared to WT). These data suggest that CF/Hdac6 mice are able to resolve the infection response faster than CF mice. Analysis of inflammatory cell recruitment demonstrates that CF BAL is dominated by neutrophils at day 3 post-infection compared to WT. CF/Hdac6 mice have values identical to WT mice (day 3: WT: 45 ± 5.0% neutrophils; CF: 73 ± 5%*; CF/Hdac6: 48 ± 10%: *p < 0.05 compared to WT). Cytokine analysis reveals that IL-6 is significantly elevated in CF mice 6 days post-infection compared to WT mice (IL-6 day 6: CF: 16.9 ± 5.2*; CF/Hdac6: 8.8 ± 7.1pg/mL; *p < 0.05).

Conclusions: Data demonstrate that depletion of Hdac6 in CF mice dramatically improves the resolution of the inflammatory response induced by PA challenge and that HDAC6 inhibition represents a potential site of intervention to replicate beneficial effects of ibuprofen.

Acknowledgments: Supported by grants from the CFF and NIH/ NHLBI.

Genovese, M. 1 ; Scudieri, P. 1 ; Musante, I. 1 ; Ferrera, L. 2 ; Bandiera, T. 3 ; Galietta, L.J. 1 1. Telethon Institute of Genetics and Medicine, Pozzuoli, NA, Italy; 2. Istituto Giannina Gaslini, Genova, Italy; 3. Istituto Italiano di Tecnologia, Genova, is a calcium-activated chloride channel with a particular expression in non-ciliated airway epithelial cells. Activation of TMEM16A by calcium agonists results in enhanced chloride and bicarbonate secretion. The physiological role of TMEM16A, which is upregulated under conditions that favor mucus hypersecretion, and its involvement in cystic fibrosis (CF) are still unclear. Our aim is to identify novel pharmacological modulators of TMEM16A function. We are using a high-throughput functional assay to screen a maximally-diverse chemical library (11,300 compounds). For the screening, FRT cells with expression of TMEM16A(abc) isoform and the halide-sensitive yellow fluorescent protein (HS-YFP) are stimulated with a low concentration of UTP to induce partial activation of TMEM16A. TMEM16A potentiators and inhibitors are detected as compounds that accelerate or slow down the rate of HS-YFP quenching, respectively. Controls during the screening included the Ani9 inhibitor (Seo Y, et al. PLoS One. 2016; 11:e0155771) and a TMEM16A potentiator that we identified in a previous study. After screening nearly half of the library, we have selected a set of compounds that enhance or inhibit TMEM16A function. We are giving priority to compounds with stimulatory activity. These hits are evaluated in secondary assays based on: i) null FRT cells, to rule out compounds acting on other channels/ transporters; ii) CFPAC-1 cells (which have endogenous TMEM16A expression), to confirm activity in a second cell type; iii) stimulation of TMEM16A with ionomycin, to bypass purinergic receptors; iv) evaluation of intracellular calcium mobilization with a fluorescent probe, to detect compounds with an indirect mechanism of action. After completion of the library screening, compounds with the best characteristics will be tested in CF human bronchial epithelial cells (short-circuit current experiments) to assess the effect on calcium-dependent chloride secretion. Novel pharmacological modulators of TMEM16A (potentiators and inhibitors) could be useful as tools of research and as possible therapeutic agents to improve mucociliary function in CF and other respiratory diseases. This work is supported by CFF (grant GALIET17G0). The second messenger cAMP exerts a negative constraint on the activation of the innate immune response by limiting production of pro-inflammatory cytokines such as TNFα. Activation of toll-like receptor 4 (TLR4) by bacterial lipopolysaccharides (LPS) induces the expression of the cAMP-phosphodiesterase variant PDE4B2, which hydrolyzes and inactivates cAMP, thus allowing for full activation of the inflammatory response. Counteracting the LPS-induced expression of PDE4B2 by inhibiting its enzymatic activity is a critical mechanism by which inhibitors of type 4 phosphodiesterases (PDE4s) exert anti-inflammatory effects. Probing the potential of PDE4 inhibition for the treatment of cystic fibrosis, we found that treatment with the archetypal PDE4 inhibitor rolipram reduces LPS-induced TNFα production in primary human airway epithelial cells from both non-CF and F508del-CFTR donors by similar levels, suggesting that PDE4 inhibition retains its anti-inflammatory efficacy on a background of altered immune responses in CF. We also observed the established LPS-induced surge in PDE4B2 expression. However, PDE4B2 levels were not only induced by LPS, but also by treatment with PDE4 inhibitor per se; and PDE4 inhibition further amplified the LPS-mediated PDE4B2 induction. This induction is unique to PDE4B2, as other PDE4B splicing variants or PDE4 subtypes PDE4A, PDE4C and PDE4D were not altered by LPS and/ or PDE4 inhibitor treatment. PDE4B2 induction is not unique to rolipram treatment, but was observed upon treatment with a number of structurally distinct compounds, suggesting it is a class effect of nonselective PDE4 inhibitors. Pretreatment with the PKA inhibitor H89 ablates PDE4 inhibitor-induced PDE4B2 expression, suggesting that it results from excessive cAMP/PKA signaling. Indeed, PDE4B2 expression is also induced by treatment with saturating concentrations of the adenylyl cyclase activator forskolin. PDE4B2 protein levels remain elevated for several hours upon washout of PDE4 inhibitor. Thus, PDE4B2 induction may limit the efficacy of PDE4 inhibitors to elevate cAMP and reduce inflammation at clinically relevant, submaximal concentrations. Inactivation of individual PDE4 subtypes using isoform-selective inhibitors did not induce PDE4B2 expression, likely because they trigger smaller, locally restricted increases in cAMP/PKA compared to the nonselective inhibition of total cellular PDE4. Development of subtype-selective PDE4 inhibitors is currently pursued as a way to separate the side effects from the therapeutic benefits of the nonselective PDE4 inhibitors available to date. Our present results indicate that subtype-selective PDE4 inhibitors might have the added benefit of not inducing the pro-inflammatory variant PDE4B2. This work was supported by grants from the CFF and the NIH (HL76125, HL141473, HL066299, DK072517). Introduction: Cystic fibrosis (CF) is a chronic inflammatory condition associated with high circulating plasma levels of the chemokines interleukin (IL)-8 and neutrophil activating peptide-2 (NAP-2). Alpha-1 antitrypsin (AAT) is an acute phase glycoprotein that has been shown to have anti-inflammatory properties by binding IL-8 and NAP-2. Glycosylation is the addition of N-linked oligosaccharides, with different combinations giving rise to different glycoforms. During the inflammatory response to pneumonia, increased levels of IL-6 results in increased sialic acid residues on AAT (sAAT), following upregulation of the enzyme α-2, 6-sialyltransferase (ST6GAL1). sAAT aids resolution by binding IL-8 and NAP-2. Individuals with CF have high circulating levels of IL-6, which could lead to persistently altered glycosylation of AAT. Therefore, our aim was to determine the specific glycoprofile of AAT in CF and the consequence of altered glycosylation.

Methods: Plasma samples were taken from patients with CF every 2 days for 10 days during an exacerbation (n=5, 3 homozygous and 2 heterozygous for the F508del mutation). AAT glycoforms were determined by isoelectric focusing. Plasma levels of IL-6, IL-10, IL-8 and NAP-2 were determined by ELISA. Primary hepatocytes were treated with IL-6, and HepG2 cells with IL-10, for up to 48 hours. The gene and protein expression of ST6GAL1 was determined by qRT-PCR and Western blotting, respectively. The levels of IL-8 and NAP-2 bound to sAAT were determined by ELISA. Neutrophil chemotaxis assays (1x10 5 neutrophils) were performed in response to a combined dose of IL-8 and NAP-2 (0-80 ng/ mL). Statistical significance was obtained using Student's t-test or one-way ANOVA. Ethical approval was obtained from Beaumont Hospital.

Results: Isoelectric focusing patterns demonstrated that the heavily sialylated M0, M1 glycans were present on AAT (sAAT) throughout the exacerbation period (n=5). In hepatocytes following treatment with plasma concentrations of IL-6, ST6GAL1 was shown to be upregulated on the gene and protein level (p<0.05, n=3). Furthermore, the IL-10 concentration required to downregulate ST6GAL1 protein expression in HepG2 cells in vitro was greater than that observed in the plasma of individuals with CF (p<0.05, n=3). There was no significant change in the plasma levels of IL-8 and NAP-2 or the level of IL-8 or NAP-2 bound to sAAT over the 10-day time course (ns, n=5). Although sAAT could inhibit neutrophil chemotaxis to a greater extent than AAT at 10 and 20 ng/mL (p<0.05, n=3), at concentrations of these chemokines found in CF plasma, sAAT could not significantly inhibit neutrophil migration.

Conclusion: In conclusion, this study showed that the glycosylation of AAT is altered in individuals with CF. Despite a significant decrease in plasma CRP levels, sAAT persists throughout the exacerbation period. Perturbed cytokine expression including high levels of IL-6 and low levels of IL-10 in the plasma of individuals with CF, leads to upregulation of ST6GAL1 in hepatocytes. In vivo this persistent expression of sAAT may be an attempt to dampen inflammation, however, sAAT is overburdened by high plasma concentrations of IL-8 and NAP-2 and therefore, fails to significantly inhibit neutrophil chemotaxis.

Salka, K. 1 ; Panigrahi, A. 1 ; Brown, K. 2 ; Perez, G. 1 ; Rose, M. 1 1. Children's National Medical Center, Washington, DC, USA; 2. Solid Biosciences, Cambridge, MA, USA Rationale: Cystic fibrosis (CF) results from mutations in the CFTR gene, which is responsible for transport of chloride and bicarbonate ions. 70% of CF patients have the ΔF508 mutation. Patients with CF experience chronic lung bacterial infection, inflammation, and mucus overproduction. Accumulating data in CF human neonates indicate that lung inflammation is present early in CF before detectable infection. The goal of this study was to determine whether quantitative proteomics would identify a differential status of CF and non-CF human bronchial epithelial (HBE) cells at homeostasis and after a bacterial challenge.

Methods: CF ΔF508 (N=3) and non-CF (N=3) life extended (LE)-HBE cell lines (Fulcher ML, et al. Am J Physiol Lung Cell Mol Physiol. 2009; 296:L82-91) were grown on Transwells and differentiated for 28 days at air-liquid interface (ALI) in media containing heavy lysine 13 C 6 -15 N 2 -Lys and arginine 13 C 6 -Arg to create super SILAC standards (SSS). In a separate experiment, CF (N=3) and non-CF (N=3) LE-HBE cell lines were grown on Transwells and differentiated at ALI for 28 days in nonlabeled media. Apical secretions (AS) were collected 24 and 48 hours after exposure to Pseudomonas aeruginosa (PA), and baseline conditions. AS were mixed in a 1:1 protein ratio with the SSS lysate and prepared for MS/MS analysis by in-gel digestion with trypsin. Statistical tests were performed on Perseus and protein data were analyzed using Ingenuity Pathway Analysis (IPA).

Results: Mass spectrometry analysis identified and quantitated 2057 proteins across samples, of which 78 exhibited differential enrichment or depletion in CF secretions (±1.5 log 2 fold-change; p-value<0.05). Preliminary analysis of the apical secretome of CF and Non-CF LE-HBE cells at baseline conditions shows increased mucin expression (MUC5AC and MUC6) and several matrix metalloproteinases (MMP) in CF cells. After exposure to a bacterial challenge (PA), non-CF cell lines showed a significant differential expression in proteins involved in innate immunity (secretory protein in upper respiratory tracts (SPURT) and complement factor B (CFB)) relative to CF cell lines (±1.5 log 2 fold-change; p-value<0.05). STRING analysis of proteins differentially expressed between CF and non-CF HBEs at baseline conditions showed differences in expression of proteins involved in extracellular matrix organization, leukocyte migration, and immune system processes.

Conclusion: The quantitative proteomic data on AS suggests a change in the inflammatory state, but also a change in the immunological state and extracellular matrix organization, of CF vs. non-CF cells at homeostasis in the absence of infection and after bacterial challenge. Airway remodeling is a key component of CF that leads to increased morbidity and mortality of patients with CF. Understanding changes in extracellular matrix organization and immune system function of CF vs. non-CF cells at baseline may provide insight into what leads to airway remodeling in CF patients. While these findings need to be validated in CF and non-CF primary HBE cells, this data may provide insight into the ability of CF patients to handle infections from a molecular standpoint. Introduction: Chronic infection and inflammation of the respiratory tract, causing severe and irreversible lung tissue damage are a hallmark of cystic fibrosis (CF). Neutrophils are crucial in combating infection and alterations in their function could contribute to abnormal immune responses on multiple levels. There is growing evidence that the CF neutrophil is intrinsically abnormal, however, molecular mechanisms underlying increased primary granule degranulation remain unclear. One protein released from primary granules is the bactericidal permeability-increasing protein (BPI), which has potent bactericidal effects exclusive to gram-negative bacteria. The aim of this study was to investigate the cause of increased primary granule and BPI release and to understand why BPI, present in high quantities in the CF lung, is unable to effectively eradicate Pseudomonas aeruginosa (PA) infection.

Materials and Methods: Neutrophils, plasma and bronchoalveolar fluid (BALF) were isolated from people with CF (PWCF) homozygous or heterozygous for the ΔF508 mutation and healthy controls (HC). Primary granule degranulation was examined using Western blot (WB) analyses and densitometry of immunobands. Active Rac2 levels were determined using a Rac2 activation assay and WB. BPI levels and IgG class BPI autoantibodies (AABs) were detected via ELISA, WB and immunoprecipitation (IP). The bactericidal activity of BPI was determined using a PAO1 bactericidal assay (10 8 cells/mL). Ethical approval was obtained from Beaumont hospital ethics committee.

Results: Elevated levels of BPI were detected in plasma (n=22, p=0.0067) and BALF (n=6, p=<0.01) of PWCF (p=0.01). Ex vivo assays revealed increased release of BPI by CF neutrophils compared to HC cells (n=4, p=0.02). Activation of the low-molecular-mass GTP-binding protein Rac2, involved in the regulation of primary granule trafficking, was significantly increased in CF cells (n= 5, p=0.004), an effect mirrored in HC cells by pharmacological inhibition of CFTR function by p=0.03) . Immunoblotting confirmed that BPI in the CF airways is the fulllength uncleaved protein (n=14) and its inability to kill Pseudomonas was not a result of a reduced pH impeding its action (n=3) or via its interaction with glycosaminoglycans (n=3). By employing a predetermined threshold level for positivity as 3 standard deviations above the mean of HCs, plasma of PWCF (n= 30) proved positive for AABs against BPI (p=<0.0001). By use of protein A IP and WB analyses, BPI in the CF airways was found to be bound to IgG AABs. Upon purification, these AABs inhibited the bactericidal action of BPI (n=3, p=0.035) in a manner comparable to that of a commercial N-terminal directed BPI antibody (n=4, p=0.028).

Conclusion: In neutrophils, CFTR dysfunction results in Rac2 activation and increased degranulation of primary granules. The incidence of AABs directed against BPI is increased in PWCF and these were found to be complexed with BPI in BALF, leading to impaired bacterial killing and persistence of PA in the CF lung. Lung disease is the primary cause of morbidity and mortality in people with cystic fibrosis (CF). Using a CF pig model, we identified at least two primary host defense defects in newborn CF large airways: impaired mucociliary transport and impaired antimicrobial capacity due to a lower airway surface liquid (ASL) pH in CF compared to non-CF. The ASL pH of the large airway is set by the balance of CFTR-mediated HCO 3 transport and ATP12A-mediated proton secretion. However, in distal small airways (diameter <200 µm), which also had an acidic ASL in CF compared to non-CF cell cultures, we found that ATP12A expression and activity is diminished. Therefore, we hypothesize that small airways must use a different mechanism to regulate acid secretion. Microarray data from small airway epithelia revealed that ATP6V0D2, a subunit of the H + -translocating plasma membrane V-type ATPase, could mediate proton secretion in small airways. ATP6V0D2 is expressed on the apical surface of Muc5B + secretory cells, but not in ciliated cells in small airways. Bafilomycin, a V-type ATPase inhibitor, increased ASL pH in small airway epithelia in the presence or absence of HCO 3 -. In addition, bafilomycin decreased ASL viscosity measured by fluorescence recovery after photobleaching assay.

In conclusion, we demonstrated that V-type ATPase contributes to ASL acidification in small airways, which have nominal ATP12A expression and function. Inhibition of V-type ATPase activity could be a novel therapeutic strategy to treat or prevent CF lung disease. 3 channel activity results in airway surface liquid (ASL) acidification. A lower ASL pH causes host defense defects: it reduces the ability of the ASL to kill bacteria and increases ASL viscosity thereby impairing mucociliary clearance. Loss of CFTR activity in the intestine can cause abnormal mucus secretion and intestinal obstruction.

CFTR function is coupled to either ATPase (ATP→ ADP + P i ) or adenylate kinase (ATP + AMP ↔ 2 ADP) activity. It is not known which of the two enzymatic activities is required in vivo. We showed that mutating CFTR amino acid Q1291 disrupts adenylate kinase-but not ATPase-dependent channel activity (Dong Q, et al. J Biol Chem. 2015; 290: 14140-53) . This provided an unprecedented opportunity to address this question.

We first expressed either wild-type CFTR or the Q1291F mutant in well-differentiated primary human airway epithelia from CF donors (that lack endogenous CFTR activity) because they are closest to an in vivo airway epithelium. Both localized to the apical membrane and were expressed in similar amounts. We found that CFTR-dependent Cland HCO 3 short circuit current and transepithelial conductance were significantly lower in epithelia expressing Q1291F CFTR. Airway epithelia from CF donors had a lower ASL pH compared to non-CF. Expression of wild-type CFTR corrected the ASL pH but expression of Q1291F CFTR did not. Likewise, wild-type CFTR but not Q1291F CFTR expression increased the ability of the epithelia to kill bacteria and reduced ASL viscosity. Thus, our results suggest that defective CFTR adenylate kinase activity causes physiologic relevant effects leading to disease.

To further test the hypothesis that CFTR adenylate kinase is required for normal channel function in an epithelium in vivo, we generated a new transgenic mouse model. The intestine of murine CFTR knockout mice (CFTR -/-) mimics human CF intestinal disease. The mice die shortly after weaning. Transgenic intestinal expression of human wild-type CFTR (CFTR -/-; TgWT mice) rescued the lethal intestinal phenotype (Zhou L, et al. Science. 1994; 266:1705-8) . We predicted that if adenylate kinase-dependent gating contributes to normal CFTR function in the intestine, intestinal expression of Q1291F CFTR will either not or only partially rescue CFTR -/mice. We found, in striking contrast to wild-type CFTR, Q1291F CFTR did not improve survival. Necropsies showed intestinal obstruction in CFTR -/and CFTR -/-; TgQ1291F mice at the time of death but not in CFTR -/-; TgWT and CFTR +/+ mice sacrificed at a similar survival time. Expression of wild-type CFTR but not of Q1291F CFTR restored transepithelial Cl − secretion mediated by CFTR in the small intestine of weanling mice and reduced the presence of mucus. Our results indicate that CFTR adenylate kinase activity is essential for CFTR function in vivo. Acknowledgment: Supported by NIH/NIDDK. Genetic modifiers have considerable influence on lung function variation in cystic fibrosis (CF). Transforming growth factor (TGF)-β1 is a known modifier associated with accelerated lung disease in patients homozygous for F508del. Increased TGF-β1 levels are associated with variants of the TGF-β1 gene, present in ~40% of F508del homozygous patients. The relevance of TGF-β1 has been demonstrated by the inhibitory effect on CFTR mRNA and subsequent resistance of F508del-CFTR to corrector-mediated rescue in primary differentiated human bronchial epithelial (HBE) cells. Understanding the regulation of TGF-β signaling in human airways is critical to design strategies to attenuate abnormal TGF-β signaling, allowing the rescue of F508del-CFTR. We aimed to elucidate the mechanisms of TGF-β1 signaling via the proximal pathway at the level of TGF-β receptor (TβR)-I and TβR-II in HBE cells. TGF-β1 initiates a signaling cascade by binding to TβR-II, which phosphorylates and activates TβR-I; in turn, this receptor activates receptor (R)-Smads, initiating an intracellular signaling. In nonstimulated cells, protein phosphatase (PP)1 dephosphorylates TβR-I, protecting it from constitutive activation by TβR-II; however, it remains unknown how TGF-β1 blocks the PP1-mediated inhibition of TβR-I to initiate its signaling. We hypothesized that, after TGF-β1 stimulus, PP1 is inhibited through phosphorylation on its residue T320, thereby allowing the activation of TβR-I. Lemur tyrosine kinase-2 (LMTK2), which is also known to phosphorylate CFTR, organizes a protein network that mediates the inhibitory phosphorylation of the catalytic subunit of PP1 (PP1c) in HeLa cells. PP1 interacts with several proteins that regulate its localization and catalytic activity; however, it is unknown whether LMTK2 also inactivates PP1 in human bronchial epithelial cells.

Experiments performed in human bronchial epithelial (CFBE41o-) cell line demonstrate that TGF-β1 increased the inhibitory phosphorylation of PP1c, in a process mediated by LMTK2. Indeed, siRNA mediated LMTK2 depletion promoted activation of PP1c, leading to a subsequent attenuation of the TGF-β1 signaling via R-Smads. In turn, abundance of the R-Smads protein was increased to counteract these effects. These results were confirmed by a selective PP1c activator, which also inhibited activation of R-Smads despite TGF-β1 treatment. To better understand how PP1c is regulated after TGF-β1 stimulus, we examined mRNA expression of the PP1c-and LMTK2-interacting proteins by mRNA-seq. Changes in the expression level of 27 genes were observed, including GADD34 and p35. Proteomic analysis using liquid chromatography tandem mass spectrometry showed an interaction between PP1c and Inhibitor 2, a protein previously described as a mediator of LMTK2 inhibition of PP1c.

In summary our data demonstrate that PP1c prevents activation of the TGF-β1 pathway in the absence of the ligand while LMTK2 promotes signaling after TGF-β1 stimulus. Our studies may lead to novel therapeutic targets blocking abnormal TGF-β1 signaling, thereby improving the functional rescue of F508del-CFTR by CFTR correctors. Airway submucosal gland serous acini are major sites of CFTR expression in the airways. Submucosal glands secrete a large amount of the fluid and mucus forming the airway surface liquid. Serous acinar cells likely generate most of the fluid secreted by these glands. Agonists that signal through cAMP, such as vasoactive intestinal peptide (VIP), activate CFTR-dependent secretion from intact glands and isolated acinar cells. We studied primary serous acinar cells isolated from submucosal glands dissected from residual human nasal turbinate tissue obtained after sinus surgery. We previously characterized VIP/cAMP-activated, CFTR-dependent Cl --driven fluid secretion from serous cells using DIC imaging of cell volume to track agonist-induced changes in cell solute content reflective of changes in the secretory state of the cells (Lee RJ, Foskett JK. J Clin Invest. 2010; 120(9) :3137-48). This technique was combined with simultaneous quantitative fluorescence microscopy to measure the intracellular concentrations of ions involved in driving fluid secretion ([Cl -] i ) and regulating it ([Ca 2+ ] i ). In the present study, we combined imaging of cell volume and intracellular pH (pH i ) to identify if and how serous cells secrete bicarbonate (HCO 3 -) during VIP/cAMP-stimulated fluid secretion. HCO 3 secretion by serous cells is important for proper airway surface liquid pH as well as polymerization of mucins secreted by more proximal mucus cells in the glands. During VIP stimulation, pH i transiently dropped concomitantly with the fall in Clcontent revealed by cell shrinkage. This likely reflected HCO 3 efflux, as it was almost completely reduced (>90%) in the absence of CO 2 / HCO 3 -. A subsequent alkalinization increased pH i above resting levels; this was predominately inhibited by the Na + /HCO 3 cotransporter (NBC) inhibitor 4,4'-dinitrostilbene-2,2'-disulfonic acid (DNDS). Elimination of driving forces for conductive HCO 3 efflux by ion substation or exposure of serous cells to CFTR inh 172 reduced VIP-induced acidification by >80%. VIP-induced pH i acidification was also markedly reduced (>75%) in serous cells from CF patients. The non-CFTR Clchannel inhibitors niflumic acid and 5-nitro-2-(3-phenylpropylamino)benzoic acid (NPPB) had no effect on VIP-induced acidification. Our data suggest that the majority of HCO 3 efflux during cAMP-evoked serous acinar cell secretion is conductive efflux via CFTR. HCO 3 secretion during VIP stimulation is sustained by NBC activity, whereas we previously showed that cholinergic/Ca 2+ -induced HCO 3 secretion is largely sustained by Na + /H + exchanger isoform 1 (NHE1)-mediated alkalinization. In contrast to some studies of surface epithelial cells and Calu-3 cells, our data do not support a substantial role for apical Cl -/HCO 3 exchange in HCO 3 secretion from primary serous acinar cells. Supported by CFF research grant LEER16G0.

In the past multiple studies have shown a link between hypergammaglobulinemia and more advanced lung disease in patients with cystic fibrosis (CF). Prevalence of hypergammaglobulinemia tends to decrease, due to improving therapies. However, in our experience, the amount of patients who develop hypogammaglobulinemia is increasing. The underlying mechanism, the prevalence and the significance of hypogammaglobulinemia still remain unclear.

Objective: To report the difference in outcome in patients with hypogammaglobulinemia versus patients with normoglobulinemia. To investigate whether there is a difference in lung function, amount of infections and need for hospitalization before and after parenteral substitution with immunoglobulins.

Materials and Methods: In this monocentric retrospective case-controlled study medical records of children with CF born between 1/1/1998 and 31/12/2017 were reviewed. Hypogammaglobulinemia was defined as 2 times or more serum IgG with Z-score < -2SD according to local lab references, not caused by immune modulating therapy. For each case with hypogammaglobulinemia, a matched control was found, based on gender, pancreatic function, birth year (± 4 years) and age at CF-diagnosis (0-2 years vs after 2 years), and genotype (severe versus mild mutations) respectively.

Results: In total 127 children (63 males) were diagnosed and treated in our center during the given period. Median age was 12.4 years and median age at diagnosis of CF was 0.3 years. 78 patients (61%) were homozygote F508del, 39 patients (31%) were heterozygote F508del and 10 patients (8%) had other genotypes. 117 patients (92%) had pancreatic insufficiency. 32 patients (25%) had hypogammaglobulinemia of which 10 were treated with immunoglobulins. Outcome will be compared between cases and matched controls at the age of 5, 10 and 15 years. Changes in lung function and BMI will be compared in the treated vs nontreated patients.

Conclusion: About one-third of the patients with CF have hypogammaglobulinemia at some point during childhood. In this study we compare the evolution of patients with and without hypogammaglobulinemia to explore the significance of hypogammaglobulinemia and how it affects prognosis. Assessment of the effect of treatment with immunoglobulins will assist in the decision to treat patients with CF. Final analysis of the data will be presented at the conference. SLC26A9 is a member of the SLC26 gene family of anion channels and transporters that have diverse functional roles in epithelia. We previously identified, by high-throughput screening, small molecule inhibitors of SLC26A4 (pendrin) and SLC26A3 (DRA) and demonstrated their potential utility to treat CF lung and gastrointestinal disorders, respectively. The SLC26A9 chloride transporter, which is expressed mainly in airway and gastric epithelia, has been proposed as a mutation-agnostic, prosecretory target for CF drug development. In CF subjects, SLC26A9 single nucleotide polymorphisms are associated with susceptibility to meconium ileus, prenatal exocrine pancreatic damage, and CF-related diabetes. Of direct relevance to CF lung disease, treatment responses to ivacaftor in CF subjects with the G551D mutation are associated with a single SLC26A9 polymorphism, and SLC26A9 loss-of-function mutations are found in subjects with diffuse bronchiectasis with or without CF; however, SLC26A9 polymorphisms are not directly associated with CF lung disease severity. These results have been interpreted to suggest that SLC26A9 compensates for CFTR loss-of-function in certain organs, and hence activation of SLC26A9 may provide a mutation-agnostic therapeutic approach for CF. To clarify the role of SLC26A9 in the lung epithelium, we identified small molecule SLC26A9 inhibitors by high-throughput screening and are applying them to study airway and gastrointestinal epithelial physiology. For screening, cDNA encoding human SLC26A9 was cloned into pLION for generation of FRT cells stably expressing SLC26A9 and a halide-sensitive YFP (EYFP-H148Q/I152L/F46L). SLC26A9-mediated transport was assayed from the kinetics of YFP fluorescence quenching in response to extracellular addition of iodide to drive chloride/iodide exchange, in which transport inhibition reduces the rate of fluorescence decrease. 50,000 chemically diverse, synthetic small molecules were screened at 25 mM, utilizing 0.35 mM niflumic acid (a non-selective inhibitor) as positive control. The SLC26A9 screen was robust, with Z'-factor >0.65, and the overall rate for identification of inhibitors was 0.04%, with several distinct chemical classes of inhibitors identified. Inhibitors with different SLC26A9 selectivity were identified, including highly selective compounds that did not inhibit pendrin or DRA. Further optimization of SLC26A9 inhibitors for Objective: Tezacaftor/Ivacaftor (TEZ/IVA) is a new CFTR modulator combination that improves forced expiratory volume in 1 second (FEV 1 ) and forced vital capacity (FVC) in patients aged ≥12 years with CF who are homozygous for F508del-CFTR (F/F; EVOLVE, NCT02347657) and in those heterozygous for F508del-CFTR and a second allele with a CFTR mutation predicted to have residual function (F/RF; EXPAND, NCT02392234). Long-term safety and efficacy are being assessed in EXTEND, an ongoing open-label extension study (NCT02565914). Benefits in lung function are considered to result from 2 physiological effects: (1) a reduction in airway resistance (Raw) due to decreased narrowing and (2) an improvement in breathing capacity (BC) due to opening of collapsed airways. To characterize how TEZ/IVA impacts respiratory function in both F/F and F/RF populations, improvement in FEV 1 mediated by these distinct mechanisms was assessed.

Methods: A retrospective analysis of spirometry data from the placebo and treatment groups of EVOLVE, EXPAND, and EXTEND was performed. Changes in FEV 1 (δFEV 1 ) over time were calculated as a function of changes in Raw (δ[FEV 1 /FVC]•FVC) and BC (δFVC• [FEV 1 /FVC]).

Results: By day 15 of EVOLVE or EXPAND, TEZ/IVA treatment was associated with significant improvements in FEV 1 as compared with placebo in F/F (+108 ± 194 vs -16 ± 176 mL) and F/RF (203 ± 207 vs 0 ± 199 mL) patients aged ≥12 years with CF. These improvements persisted over time. Increases in FEV 1 due to reductions in Raw and improvements in BC were observed, although improvements in BC accounted for the majority of FEV 1 increase (66% in F/F and 61% in F/RF). Raw improvements appeared to plateau between 8 to 12 weeks, but BC improvements continued through the last observation point in EXTEND: 84 weeks in patients with F/F and 56 weeks in those with F/RF.

Conclusions: Our results suggest that TEZ/IVA improves lung function in patients with F/F and F/RF by both reducing Raw and increasing BC. Reduction in Raw effects occurs early and appears to plateau, while increased BC effects progress over time. These findings are potentially consistent with short-term improvement in airway obstruction caused by mucus occlusion and relaxation of smooth airway muscle tone, and long-term improvement from recruitment and beneficial remodeling of lung parenchyma.

Acknowledgments: Sponsored by Vertex Pharmaceuticals Incorporated.

Martino, M.E. 2 ; Ribeiro, C.M. 2, 1 1. Dept. of Medicine, Univ. of North Carolina, Chapel Hill, NC, USA; 2. Marsico Lung Institute/CF Research Center, Univ. of North Carolina, Chapel Hill, NC, USA Background: CF airways exhibit inflammation-triggered mucin overproduction, which leads to mucus obstruction. MUC5AC and MUC5B are the major secreted mucins in the airways of CF patients. Because no therapies are available to treat CF airway mucus overproduction, there is an unmet medical need for new therapies that target mucin synthesis in CF airways. Airway epithelial mucin overproduction activates inositol requiring enzyme 1β (IRE1β), which is only expressed in mucous cells and required for mucin production during allergic inflammation of murine and human airways (1) . Importantly, native inflamed CF airway epithelia exhibit increased levels of IRE1β protein (1), but it is not known whether IRE1β is functionally important for CF airway mucin production.

Objectives: The present study evaluated the role of IRE1β in airway epithelial mucin production utilizing models relevant to inflamed, mucus-obstructed native human CF airways.

Methods: IRE1β mRNA levels, assessed by quantitative RT-PCR, were compared in freshly isolated non-CF and CF human bronchial epithelia (HBE), and correlated with mucin mRNA expression. To reproduce the up-regulation of IRE1β found in native CF airway epithelia, primary non-CF HBE were transduced with a control pQCXIP vector or a pQCXIP vector containing the wild-type IRE1β, grown on permeable supports, and evaluated as well-differentiated cultures. Mucin mRNA levels were studied under basal conditions and after HBE exposure to supernatant from mucopurulent material (SMM) from human CF airways, a translational model for CF airway mucin overproduction (2) . Alternatively, to study the impact of silencing the IRE1β gene on HBE mucin production, lentiviral-delivered CRISPR-Cas9 was used.

Results: IRE1β mRNA levels were up-regulated in CF vs. non-CF HBE and this difference was associated with up-regulation of MUC5AC mRNA expression (n=4). Over-expression of IRE1β up-regulated IRE1β mRNA levels by 9.5 ± 2.2 fold and increased basal MUC5AC and MUC5B mRNA levels by 10.1 ± 5.2 and 4.1 ± 1.6 fold, respectively (mean ± SEM, n=4). Notably, IRE1β over-expression potentiated 72 h SMM-increased MUC5AC and MUC5B gene expression (19.9 ± 4.2 vs. 10.7 ± 4.1 for MUC5AC; 18.0 ± 2.6 vs. 10.5 ± 0.5 for MUC5B; p<0.05, HBE expressing the vector containing IRE1β + SMM vs. HBE expressing the control vector + SMM). The IRE1β CRISPR disrupted the IRE1β reading frame, produced ~75% clones with mutations, and did not affect predicted off-target sites. The IRE1β CRISPR markedly decreased IRE1β protein expression and reduced the basal levels of total (intracellular and secreted) MUC5AC protein by ~ 50% (n=3).

Conclusions: IRE1β and mucin expression is increased in freshly isolated CF HBE. IRE1β over-expression increases basal mucin expression and potentiates SMM-up-regulated mucin production, whereas silencing IRE1β gene expression decreases mucin production. Our study further highlights the functional role of IRE1β in regulation of mucin synthesis and suggests that IRE1β is a novel therapeutic target for CF airway mucus overproduction.

Introduction: Cystic fibrosis (CF) produces chronically progressing lung disease, and monitoring disease status in response to treatment requires sensitive measurement of both lung structure and lung function. Spirometry is currently the gold-standard lung function assessment method, but it typically only provides a single global measurement and it is effort-dependent. Imaging-based techniques such computed tomography are sensitive to small changes in lung structure, but efforts to reduce radiation dosage result in poor temporal and spatial resolution.

We have developed new X-ray imaging-based techniques for detecting, localising and quantifying regional lung disease, allowing us to track disease progression and assess the effectiveness of genetic and pharmaceutical therapies. When high resolution images of the lung are combined with a volumetric particle image velocimetry (PIV) approach, a four-dimensional (3D, over time) dynamic imaging technique known as local airflow analysis (LAA) is formed. This technique was developed at a synchrotron (Stahr CS, et al. Sci Rep. 2016; 6:29438) , but we have translated it to a laboratory X-ray source (Werdiger F, et al. Pediatr Pulmonol. 2017; 52(S47):254 [abstract] ).

Here we present new analysis techniques to assess the distribution of lung disease in β-ENaC mice.

Methods: All imaging was performed on a laboratory-based high-brightness liquid metal jet X-ray source (70 kV, 265W; Excillium, AB, Kista, Sweden) established at the Monash University Laboratory for Dynamic Imaging. A high speed CMOS flat-panel detector (PaxScan, Varian Medical Systems, Palo Alto, CA, USA) was used to capture images at a frame rate of 30 Hz. A rotary stage (Zaber Technologies, Vancouver, Canada) was used to rotate the mice under mechanical ventilation for the CT scan.

β-ENaC-Tg mice (12 mice) that show obstructive disease similar to that in CF were compared to their wild-type (WT) littermates (8 mice) using LAA.

Results: LAA indicated marked heterogeneity of airflow in the β-ENaC-Tg lung, compared to WT littermates-inferred from tissue expansion. Quantitative measures of tissue expansion at the peak in the breathing cycle were used to delineate symptoms of CF lung disease. Lung-disease heterogeneity in β-ENaC-Tg mice was apparent as patchy regions of reduced tissue expansion.

Discussion: Now that LAA analysis techniques have been generated for a small laboratory source, protocols are being adapted for pre-clinical use. The combination of in vivo four-dimensional lung imaging with airflow analysis provides a powerful new lung health assessment tool suited to both the assessment of potential therapies and for locating small changes in lung disease in vivo. In particular, we can now quantify and visualise the subtle regional changes that occur at the onset of lung disease, enabling initial obstructions in lower lung airway regions to be detected for development of monitoring and treatment protocols. To assess the role of miR-199a-5p on TGF-β signaling in CF lung remodeling and inflammation.

Introduction: Morbidity and mortality of cystic fibrosis (CF) patients is driven by chronic infections and hyperinflammatory responses in the lung, leading to structural remodeling, which ultimately causes respiratory failure. CF-affected macrophages (MΦs) are hyperinflammatory, fail to reinstate tissue homeostasis and therefore contribute to disease progression. We previously found that CF MΦs fail to downregulate microRNA-199a-5p (miR-199a-5p) expression after LPS stimulation. This leads to decreased amounts of caveolin 1 (CAV1), which normally terminates pro-inflammatory TLR-4 signaling. A dysregulation of the miR-199a-5p/CAV1 axis has also been described in lung tissues of idiopathic pulmonary fibrosis patients and in animal models of the disease. Dysfunctional miR-199a-5p/CAV1 axis increases TGF-β signaling, which drives tissue remodeling. Using a chronic lung inflammation model, in which Cftr tm1Unc (CFKO) mice develop lung structural remodeling (Bruscia EM, et al. AJP Lung Cell Mol Physiol. 2016; 310:L711-9) , we aim to investigate the role of miR-199a-5p and TGF-β signaling on lung hyperinflammation and remodeling in CF. This is relevant to CF, since TGF-β signaling is also highly increased in late-stage CF lung disease (Harris WT, et al. PLoS One. 2013; :e70196).

Methods: Wild-type (WT) (n=16) and CFKO (n=16) animals were chronically nebulized with LPS (12.5 mg/dose) for 5 weeks, 3 times per week. The first group was sacrificed 24 hours after the last nebulization (T1). Further, a group of mice were sacrificed after 3 (T2) and 6 (T3) weeks of recovery time after the last nebulization. Distinct lung MΦ populations (alveolar MΦ, interstitial MΦ and inflammatory Ly6C + monocytes) were quantified by flow cytometry. CD38 expression was used as pro-inflammatory marker. Whole lung tissue lysates were prepared to measure protein and/or RNA expression of miR-199a-5p, CAV1, TGF-β, Smad-2/Smad-3 phosphorylation, collagen 1 and inflammatory cytokines by Luminex.

Results: We observe increased numbers of alveolar and interstitial MΦs in lung tissues of CF mice at 6 weeks recovery time compared to WT. Both populations also remain with increased expression of the pro-inflammatory marker CD38. Lung tissues of WT mice downregulate miR-199a-5p at 3 weeks recovery time. Lung tissues of CF mice have increased miR-199a-5p levels compared to WT and fail to downregulate miR-199a-5p at recovery time. This correlates with decreased CAV1 protein expression in CF mice. Expression analysis of TGF-β signaling related genes and proteins are ongoing.

Conclusion: We developed a flow cytometry assay, that allows investigation and sorting of MΦ populations during chronic inflammatory lung injury and recovery in lung tissues of CF mice. Our final goal is to assess the impact of distinct MΦ populations and to test whether they have a dysregulated miR-199a-5p/CAV1 pathway, which we hypothesize is inducing TGF-β driven lung remodeling in CF mice.

Introduction: Cystic fibrosis (CF)-related diabetes mellitus (CFRD) is a common comorbidity of CF and a major predictor of worse lung function. Hyperglycemia in CFRD patients contributes to respiratory decline by promoting inflammation, increasing infection risk, and by likely adversely affecting ion channel function, though the mechanisms remain largely unknown. Advanced glycation endproducts (AGEs) are proteins or lipids that become glycated and whose production is accelerated under conditions of hyperglycemia. AGEs bind to the receptor for advanced glycation endproducts (RAGE or AGER) and activation of the membrane-bound form of RAGE (mRAGE) initiates a positive feedback loop that engages pro-inflammatory signaling pathways. Increased levels of RAGE and RAGE ligands in CFRD patients negatively correlate with lung function, but there remains a significant gap in our understanding of how elevated RAGE expression leads to respiratory dysfunction in CFRD. In this study, we sought to understand the effects of hyperglycemia on apical K + (BK) channel current in normal human and cystic fibrosis bronchial epithelial (NHBE and CFBE) cells in vitro and evaluate the role of RAGE signaling in hyperglycemia-induced ion channel dysfunction in CF.

Methods: NHBE and CFBE cells were redifferentiated at the air-liquid interface (ALI). Glucose levels in the media were monitored using OneTouch Verio ® meter. Fully differentiated NHBE or CFBE cells grown on Snapwell filters were mounted in Ussing chambers (EasyMount Chamber) connected to a VCC MC8 voltage clamp unit (Physiologic Instruments). Apical K + currents were measured with ATP stimulation after basolateral permeabilization and amiloride treatment in the presence of a basolateral-to-apical K + gradient. LRRC26, RAGE, MUC5AC, and MUC5B expression were quantified by qPCR using TaqMan Assays and GAPDH as a reference gene.

Results: BK currents were reduced by 50% in NHBE cells under high glucose (12.5 mM) compared to normal glucose (5 mM) conditions (p<0.05; n=6 lungs). BK channel function in nonexcitable cells is directly related to expression of the gamma regulatory subunit LRRC26 and we found that LRRC26 mRNA expression levels negatively correlated with glucose levels. BK currents in CFBE cells trended to decrease in high glucose. Furthermore, RAGE, MUC5AC, and MUC5B mRNA expression levels trended higher in CFBE cells under high glucose. Transepithelial resistance was greatly reduced when CFBE cells were treated with FPS-ZM1, a highly specific inhibitor of RAGE.

Conclusions: BK channels play an important role in CF for airway hydration. Our data indicate that hyperglycemia causes a decrease in apical potassium secretion through BK channels that correlates with elevated levels of RAGE mRNA expression. However, RAGE is also likely to be important for the maintenance of junctional integrity, suggesting approaches to block RAGE signaling broadly could be detrimental in CFRD. As an alternative, soluble forms of RAGE (sRAGE), which function as decoy receptors and intercept pro-inflammatory ligands to prevent mRAGE activation, are being evaluated for their ability to reverse hyperglycemia-induced mucociliary dysfunction in CFBE cells. Airway protective behaviors, such as mucus secretion and airway smooth muscle contraction, require neural circuits. These behaviors are affected in airway diseases, including cystic fibrosis. Yet investigations that examine such behaviors often focus on a single component, thus neglecting the circuit response. In the current study, we compared the transcriptomes of an airway-nervous system circuit (e.g., tracheal epithelia, nodose ganglia, and brainstem) in response to acute airway acidification in neonatal piglets. Intratracheal acid, a potent stimulus of airway sensory nerves, induced sex-dependent airway hyperreactivity and sex-independent airway obstruction. Comparing the tracheal epithelia and nodose ganglia transcriptomes revealed concomitant decreases in transcripts important for neuronal inhibition, in male, but not female, acid-challenged piglets. Increased mRNA for pro-inflammatory markers was observed in the nodose ganglia and brainstem of male acid-challenged piglets, whereas females showed increased transcription of antiviral defense proteins. Comparing the brainstem and epithelia revealed a concurrent decrease in expression of mRNAs important for serotonergic transmission in male, but not in female, acid-challenged piglets. These findings highlight new sex-specific molecules that might be important for cystic fibrosis and other airway diseases. Supported by R00HL119560-03 and 10T2TR001983-01.

Sabz Ali, Z. 1, 2 ; Sweezey, N. 1, 3 ; Palaniyar, N. 1, 2 1. Translational Medicine, The Hospital for Sick Children, Mississauga, ON, Canada; 2. Laboratory Medicine and Pathobiology, University of Toronto, Toronto, ON, Canada; 3. University of Toronto, Toronto, ON, Canada Background: Severe lung disease, the major cause of morbidity and mortality in patients with CF, is associated with severe inflammation and often with chronic Pseudomonas aeruginosa infection. Furthermore, adult females with CF experience worse lung-related outcomes than their male counterparts. However, the mechanism of sex-specific exacerbation of CF lung disease is not fully understood. Studies have reported that neutrophil extracellular traps (NETs) contribute to the extracellular DNA content in the airways of CF patients. Although NETs are effective in trapping pathogens, excessive NET formation (NETosis) can exacerbate CF lung disease. Whether menstrual changes in sex-hormones influence NETosis in CF pathology is not known. Our goal is to investigate the effects of estrogen and progesterone at concentrations relevant to the menstrual cycle on CF-NETosis in the presence of P. aeruginosa.

Methods: Whole blood samples from CF patients were processed using the density gradient Polymorphprep TM for neutrophil isolation. To study NETosis, we incubated the neutrophils with phorbol 12-myristate 13-acetate (PMA) and menstrual relevant concentrations of estrogen and progesterone followed by exposure to P. aeruginosa. NETs were analyzed using a cell impermeable fluorescent DNA binding dye called "Sytox green." The intracellular production of reactive oxygen species (ROS) was studied using a fluorescent dye called "DHR123." NET morphology was confirmed with immunofluorescent confocal microscopy specific for the NET-associated proteins myeloperoxidase (MPO) and citrullinated histone 3.

Results: Neutrophils isolated from CF blood showed typical polymorphonuclear morphology. The nucleus morphology and the absence of cleaved caspase 3 confirmed that the neutrophils were healthy. DHR123based analysis showed robust PMA-mediated ROS production in CF neutrophils. Sytox-green analysis revealed that P. aeruginosa induced NETosis comparable to that of PMA. Immunoconfocal microscopy of neutrophils incubated with P. aeruginosa and PMA showed colocalization of MPO and DNA. The immunoconfocal images confirmed the Sytox green data and also showed the presence of citrullinated histone 3 in subpopulations of neutrophils when exposed to P. aeruginosa and not PMA. Estrogen and progesterone at concentrations relevant to the menstrual cycle modulated NETosis to different degrees.

Conclusion: Exogenous estrogen and progesterone, at concentrations relevant to the menstrual cycle, modulate NETosis in CF neutrophils ex vivo. The approval of lumacaftor/ivacaftor (luma/iva) for CF subjects homozygous for the F508del mutation was based on findings that lumacaftor, a CFTR corrector, improved folding and trafficking of the CFTR protein to the cell membrane, in conjunction with ivacaftor, to augment lung function and nutritional status, and reduce the rate of pulmonary exacerbations by 39% (p<0.001). While reduction of pulmonary exacerbations, the outcome of greatest significance, is presumably linked to improvements in immune function and infection status, precise delineation of immune system effects is unknown. We hypothesized that transcriptome sequencing of whole blood leukocytes would sensitively capture differences in systemic immune responses in CF subjects before and after treatment with luma/iva. Methods: Circulating leukocyte RNA was extracted from peripheral blood samples in CF patients at baseline health prior to treatment and one month following treatment with luma/iva (n=17). Within the patient cohort were two groups: 1) 12 "responders" who stayed on the medication at one month with good tolerance and, 2) 5 "nonresponders" who either discontinued the medication or could only tolerate it at half dose. The isolated total RNA was processed for next-generation sequencing library construction. Pairwise statistical comparisons pre-and post-treatment were performed with the DESeq2 package of R statistical software. Each subject served as their own control. For the statistical comparisons, the cohort was split into responders and nonresponders; pre-and post-treatment samples were compared separately for each group. Weighted gene correlation network analysis (WGCNA) was performed and modules correlated with demographic and clinical parameters.

Results: For all samples, 792 amplicons were statistically significantly different between pre-and post-treatmenr samples at a false discovery rate cutoff of 0.1. The top 10 differentially expressed genes (DEGs) following treatment for all subjects reflected downregulation of leukocyte genes participating in NF-kB, JNK, ERK pathways as well as genes involved in post-translational ubiquitination and regulation of membrane trafficking. When genes were evaluated based on patient tolerance, 903 genes were significantly downregulated and 258 genes upregulated in the responder group. In the nonresponders, 20 genes were significantly upregulated and 2 genes downregulated. Co-expression network analysis of DEGs using WGCNA revealed several modules in the co-expression network were correlated with tolerance of full dose luma/iva. Gene ontology (GO) enrichment analysis of those modules identified key differences in RNA processing, cellular metabolism and mitochondrial function in subjects considered responders versus nonresponders.

Conclusions: Functional, pathway enrichment, and network analyses of DEGs in subjects who are tolerant versus intolerant of luma/iva demonstrate significant differences based on subject tolerance. Specific genes may function as tolerance associated molecular and prognostic indicators in subjects receiving CFTR modulators, identifying targets in the immune modulating actions of these agents. (µOCT) has been shown to provide information on functional microanatomy of the airways. Since viscosity is central to CF pathophysiology, we have extended the µOCT system to incorporate magnetomotive assessment to measure the rheological properties in situ. Here, we investigate the potential of magnetomotive µOCT for particle tracking microrheology (PTM) of CF mucus samples.

Methods: A magnetic system integrated in a µOCT was providing orthogonal fields of sufficient strength to move supermagnetic beads. Within the experiment, the vertical magnetic field was constantly ON. Particle tracks were obtained from µOCT images of mucus from 3 CF human subjects. Images acquired at 40 Hz contained 512 lines and spanned a field of view of 500 µm. Based on demonstrated strategy (Chu KK, et al. Biophys J. 2016; 111:1053-63) , we used a technique similar to PTM, but modified to account for an active force. An automated particle tracking algorithm (Sbalzarini IF, et al. J Struct Biol. 2005; 151:182-95) was used to quantify the motion of the beads (Fig A) . The displacement map (Fig B) was estimated based on intensity-based registration of the sequential images 2.5 seconds apart by means of an open source algorithm (Pennec X, et al. Proc MICCAI. 1999; 597-605) .

Results: The result shows that µOCT imaging can detect a motion of the beads within the sample in the direction of the applied magnetic field. Quantitative analysis of supermagnetic particle velocities within CF mucus demonstrated a mean vertical particle velocity of 14.9 ± 11.2 (std) µm/s. Heterogeneous movement grouped into localized areas of fast and slow motion, likely reflecting properties of the sample itself.

Conclusions: In this work, we demonstrated the capability to spatially modulate magnetic beads in phantoms and CF mucus using a time-varying external magnetic field. Initial results suggest that magnetomotive µOCT technology may be feasible for probing microrheological properties of CF mucus samples in two dimensions, potentially providing an avenue for investigating the local viscoelastic properties of CF mucus in situ.

Acknowledgments: This research was funded by CF Foundation (TEARNE16XX0, ROWE14Y0, ROWE16XX0). We acknowledge the CF Foundation Mucociliary Consortium Core Award and the UAB CF Research Center. Introduction: Over 2,000 mutations have been reported in the cystic fibrosis transmembrane conductance regulator (CFTR) gene. The severity of the disease depends on the variable loss of function from both alleles at the locus. Compared to the most common variant, F508del, phenotypes vary in many lesser-known mutations.

Objective: This retrospective cohort study aims to describe the phenotype of the L206W variant in a cohort of French Canadian adults with cystic fibrosis.

Methods: Between 2000 and 2017, all cystic fibrosis patients who had at least one L206W mutation were included. F508del homozygous controls were randomly chosen from our pool of patients. Demographic data, context of diagnosis, use of antibiotics and pulmonary function tests were gathered from the medical records. Sputum colonisation was assessed for bacteria, mycobacteria and fungi. Extrapulmonary manifestations were also noted such as pancreatic insufficiency, low weight, diabetes, osteoporosis and infertility.

Results Conclusion: Cystic fibrosis adults with a L206W variant tend to follow a mild evolution of the disease, with less exacerbations and fewer extrapulmonary manifestations than typical F508del homozygous patients.

Background: Cystic fibrosis (CF) is a genetic disease with marked heterogeneity in the distribution of CFTR mutations worldwide. Brazilian CF patients had not been extensively investigated, and a recent CFTR genotyping task force offered an opportunity for a better understanding of the distribution of variants throughout the country, by incorporating the data in the Brazilian CF Patient Registry.

Methods: Clinical, functional and microbiological data from the Brazilian CF Patient Registry platform were extracted in October 2017. The Registry is a web based platform, with restricted access and patient identification protected. The genotyping task force was carried out by sequencing CFTR exons and close intronic regions by next generation sequencing (NGS). Genotyping results were incorporated directly in the Registry platform by an integration application that used patient identification codes in an Excel spreadsheet, containing results of NGS. Patients were categorized as Negative, when no variant was identified, Inconclusive, when only one pathogenic variant was identified, and Positive, when a pathogenic variant was found in homozygosis, or at least two pathogenic variants were identified.

Results: Data from 4,654 CF patients were available, and 3,104 (66.7%) of them had a genotyping result. A total of 2,002 patients (64.5%) were positive, 757 (24.4%) were inconclusive, and 345 (11.1%) were negative. The distribution of genotype categories was markedly different across different Brazilian regions, with greater proportions of negative patients in the North (45%) and Northeast (26%) regions. Significant differences were noticed between values of sweat chloride levels, age at diagnosis, FEV1 values, and nutritional endpoints among patients with different genotype categories.

Conclusions: The high proportion of patients with negative results of CFTR genotyping in some Brazilian Regions may indicate that some of the included patients do not actually have CF, and towards that, initiatives to improve the ability to perform reliable sweat chloride testing in these regions are under way. Acknowledgments: CFTR genotyping task force was supported by Vertex Pharmaceuticals.

Silva Filho, L.R. 1 ; Marostica, P. 2 ; Paes, A.T. 3 ; Reis, F.C. 4 ; Athanazio, R.A. 5 ; Damaceno, N. 6 ; GBEFC 7 1. Instituto da Criança HCFMUSP, São Paulo, Brazil; 2. Hospital das Clínicas da UFRGS, Porto Alegre, Brazil; 3. Universidade Federal do Estado de São Paulo, São Paulo, Brazil; Belo Horizonte, Brazil; São Paulo, Brazil; 6. Santa Casa de São Paulo, São Paulo, Brazil; 7. Brazilian CF Study Group, São Paulo, Brazil Background: Several factors are associated with reduced life expectancy in cystic fibrosis (CF). The Brazilian CF Patient Registry includes a significant number of patients and provides an opportunity to investigate some of these factors. The recent improvement of the genetic information in the Registry database allowed a preliminary analysis of the impact of genotype categories on CF patient survival.

Methods: Genetic data incorporated in the Registry was reviewed and standardized according to the most recently available CFTR2 mutation list. New results obtained from a recent CFTR genotyping task force were included in the Brazilian dataset. Genetic, clinical, functional and microbiological data were extracted from the database in October 2017. Figure. Conclusions: While this preliminary analysis of survival did not include corrections for other variables, and may also carry a historical bias of the Registry database, a clear survival effect of genotype category was observed. Acknowledgment: CFTR genotyping task force was supported by Vertex Pharmaceuticals.

Alkindi, H.A. 1,2 ; Alfarsi, A. 1 ; Aljabri, M. 1 1. Child Health, Sultan Qaboos University Hospital, Muscat, Oman; 2. Child Health, Sultan Qaboos University, Alkhod, Muscat, Oman Background: Cystic fibrosis (CF) is an inherited autosomal recessive multi-organs disease with considerable variations in the clinical manifestations among different populations. It is present in the Omani population but there is a lack of published data on the clinical and genetic characteristics of Omani patients.

Objectives: To understand the clinical and genetic characteristics of patients in our CF centre at SQU Hospital in Oman.

Methods: A retrospective cross-sectional study was performed on data collected from January 2010 to December 2016. An analysis was conducted amongst 45 patients who were diagnosed with cystic fibrosis in the CF Centre at SQUH. The diagnosis was based on the clinical features, positive sweat chloride test and/or positive CFTR genetic mutation. The data were collected from the hospital's electronic patient records.

Results: During that period, a total of 45 patients with cystic fibrosis were included in the study. It was composed of 18 males (40%) and 27 females (60%). The age distribution was as follows: 14 (31%) < 5 years, 13 (29%) 5-9 years, 8 (18%) 10-14 years old, 8 (18%) 15-19 years and 2 (4%) patients were over 20 years old. The vast majority of the patients had an early onset of the symptoms. About 64.44% were less than 5 months of age and 82.22% were diagnosed at ≤24 months. The most prevalent symptoms were cough, recurrent chest infection and failure to thrive in 84.44% of the patients. Family history of cystic fibrosis was found in 69% of the patients. The most common mutation was S549R (75.6%) followed by 3120+1G>A. (9.9%) ΔF508 (7.7%) and, 3849+10kbC>T (6.7%). Bacterial growth result at diagnosis was variable. Pseudomonas aeruginosa growth was positive in 14 (31.11%) patients, while Staphylococcus aureus growth was in 9 (20%) patients. However, no bacterial growth was found in 28.69% (13 patients). Pancreatic insufficiency was found in 81% (36 patients). Other CF-related complications were less common, whereas cystic fibrosis-related diabetes was present in 4 (8.89%) patients only. Low bone mineral density was present in 33.3%, with 5 out of 15 patients who underwent the test.

Conclusion: Studying the different phenotypes of cystic fibrosis among the population will help in dealing with the disease. Better management and awareness are needed in order to prevent the complications. In addition, this study has encouraged us to start newborn screening for early detection of cystic fibrosis in order to avoid further complications. Introduction: Ivacaftor is a drug used to treat cystic fibrosis (CF) patients carrying specific gating CFTR mutations. Interpatient variability in the lung response has been shown to be partly explained by rs7512462 in the SLC26A9 gene. We aimed to evaluate in an independent and larger cohort the role of SLC26A9 in the ivacaftor lung response.

We genotyped the French CF Gene Modifier Study cohort (n=4,840) to investigate whether SLC26A9 (Solute Carrier Family 26 Member 9) variants were involved in the lung phenotype heterogeneity. Their influence in the response to ivacaftor was tested in the 30 treated patients who met the inclusion criteria: older than 6 years of age, percent-predicted forced expiratory volume measured in 1 s (FEV 1pp ) in the 3 months before treatment initiation ranging between 40 and 90%. Response to treatment was determined by the change in FEV 1pp from baseline, averaged in 15-75 days, and the first-year post-treatment.

Results: We observed that SLC26A9 variants were not associated with lung function variability in untreated patients and that gain of lung function in patients treated with ivacaftor was similar to clinical trials. We confirmed that rs7512462 was associated with variability in ivacaftor-lung response, with a significant reduction in lung function improvement for patients with the C allele. Other SLC26A9 SNPs also contributed to the ivacaftor-response.

Conclusion: Interindividual variability in lung response to ivacaftor is associated with SLC26A9 variants in French CF patients. Pharmacogenomics and personalized medicine will soon be part of CF patient care.

Acknowledgments Recently published data suggest disruptions in the circadian rhythm as a potential contributing factor in these symptoms. Sleep in mammals is controlled in part by the circadian rhythm which is regulated by discrete genes, including Clock, Bmal1, Period1, Period2, Cryptochrome 1, and Cryptochrome 2. The purpose of this study was to gain a better understanding of disordered sleep experienced in CF by evaluating circadian rhythm gene expression in CF and wild-type (WT) mice.

Methods: We evaluated the circadian rhythm gene expression profiles of CF (F508del/F508del) and WT mice, divided into two subgroups each based on sleep condition. One subgroup of each genotype was permitted to maintain their sleep-wake cycle while the other was deprived of sleep for six hours prior to analysis. Brain, skeletal muscle, jejunum, colon, lung and adipose tissues were collected from each mouse. Quantitative polymerase chain reaction (qPCR) was used to quantify expression of Clock, Bmal1, Period1, Period2, Cryptochrome1 and Cryptochrome2, and expression levels were compared between study groups using independent samples t-tests.

Results: In the rested state, CF mice had altered gene expression in the jejunum (Clock t 7.94 =4.09, p=.008; Bmal1 t 8.91 =3.8, p=.004) and brain (Clock t 6.09 =7.48, p<.001; Cry2 t 9 =7.03, p<.001) when compared to WT mice. In the sleep deprived state, CF mice had altered expression in the brain (Clock t 7.21 =13.42, p<.001; Bmal1 t 8.84 =5.46, p<.001; Per2 t 9.53 =5.47, p<.001), adipose (Cry1 t 10 =-10.2, p<.001; Cry2 t 7.09 =-9.55, p<.001), colon (Cry1 t 10 =-3.42, p=.007) and lung (Per1 t 10 =3.56, p=.005; Per2 t 8.3 =4.95, p=.001) compared to their WT counterparts. In the WT mice, sleep deprivation produced gene expression alteration in the adipose tissue (Per2 t 8.98 =4.26, p=.002; Cry1 (t 5.45 =2.78, p=.005), colon (Cry1 t 10 =3.58, p=.005), brain (Cry2 t 10 =3.54, p=.005), lung (Cry1 t 6.26 =3.76, p=.004; Cry2 t 7.47 =3.81, p=.006), and skeletal muscle (Cry1 t 10 =12.05, p<.001; Cry2 t 9.68 =7.04, p<.001). In the CF mice, sleep deprivation produced dramatically altered transcript levels in the brain (Bmal1 t 6.08 =5.32, p=.002), colon (Cry2 t 9 =-3.45, p=.007) and adipose tissue (Per1 t 6.47 =4.09, p=.005; Cry2 t 9 =-5.17, p=.001). In addition, the CF mice that had been sleep deprived had severely dysregulated expression of all measured genes in the lung apart from Cry1 as compared to the CF mice that had slept normally (Clock t 6.14 =8.86, p<.001; Bmal1 t 7.86 =2.57, p=.003; Per1 t 9 =7.3, p<.001; Per2 t 9 =3.39, p=.008; Cry2 t 9 =4, p=.003).

Conclusions: These data suggest that alterations in circadian rhythm gene expression are detectable in CF mice compared to WT mice, may be a primary effect of the loss of CFTR, and are a potential contributing factor in the disordered sleep experienced by CF patients.

Normal protein folding is maintained by the proteostatic network (PN), which includes molecular chaperones. CF can be viewed as a disease of proteostatic stress, as F508del leads to ineffective protein folding. In vitro work has shown that the PN is inappropriately activated by F508del, leading to protein folding stress and reduced CFTR expression (Roth, et al. PLoS Biol. 2014; 12:e1001998) . We evaluated chaperome activity and CFTR expression in the nasal epithelium of F508del homozygotes, F508del heterozygotes and controls.

Methods: Biopsy of nasal mucosa was performed in 4 F508del homozygotes, 3 F508del heterozygotes, and 6 controls. Libraries were prepared using QuantSeq 3' mRNA-Seq Library Prep Kit FWD for Illumina, then multiplexed and sequenced on the NextSeq 500 platform and aligned to the h38 genome. Estimation of differentially expressed genes (DEG) was performed. Gene Set Enrichment Analysis (GSEA) using a curated list of genes in the chaperome derived from previous work (Brehme, et al. Cell Reports. 2014; 9:1135-50) was performed.

Results: Hierarchical clustering distinguished CF and non-CF subjects, with the exception of a healthy F508del homozygote (49 years old, FEV1>95%) who clustered with controls and was therefore excluded from DEG analysis, which estimated 1687 upregulated genes and 4792 downregulated genes. GSEA using a ranked list of DEGs found significant negative enrichment of the chaperome in F508del homozygotes compared to controls. There were lower CFTR mRNA levels among the CF subjects, which trended toward statistical significance ( Figure) . Two CF subjects with higher CFTR transcript levels (circle in Figure) clustered among controls in a heatmap limited to chaperome genes.

Conclusions: Genes in the PN are negatively enriched in F508del homozygotes compared to controls, potentially due to the stress response from misfolded CFTR, thus extending previous authors' in vitro results to nasal epithelial biospies. Our data suggest an association between reduced chaperome activity and CFTR mRNA expression. Further work needs to be done to validate this finding and investigate whether misfolded CFTR protein could lead to chronic proteostatic stress and reduced CFTR transcription. 1 1. Inst. of Genetic Medicine, JHU, Baltimore, MD, USA; 2. Medicine, JHU, Baltimore, MD, USA; 3. Pediatrics, CWRU, Cleveland, OH, USA Splice variants can have a number of consequences on protein synthesis ranging from the production of no protein, truncated forms of protein or reduced quantities of wild-type (WT) protein. Differentiating among these possibilities is essential for the design and testing of therapeutic strategies for treating splice-site variants. Splice-site variants that permit expression of CFTR protein, whether aberrant or a reduced quantity of normal, may respond to corrector and/or potentiator therapy. To systematically investigate molecular consequences of 46 naturally occurring CF-causing variants (intronic=37 and missense=9), we incorporated 20/26 CFTR introns into 6 different expression mini-genes (EMGs). Missense variants were selected based on splicing predictions using CryptSplice and NNSplice or by their proximity to consensus splice sites. CFTR mRNA and protein were assessed by transient transfection of variant EMG in HEK293 cells and CFTR function was evaluated in CFBE stable cells. Cells expressing WT-EMG served as controls.

Five variants (intronic=3 and missense=2) had no detrimental effect on splicing. The remaining 41 variants exhibited aberrant splicing resulting in either frameshifts (intronic=25 and missense=4) or in-frame deletions (intronic=13 and missense=5). Twenty variants generated more than one splice isoform. Importantly, 9/20 variants (intronic=8 and missense=1) generated reduced amounts of normal spliced transcript. Remarkably, three variants (intronic=2 and missense=1) used noncanonical GC as splice donor site. Our observations using EMGs were corroborated by splicing patterns observed in primary nasal cells from CF individuals (n=5) harboring the same splicing variants.

The effect of missplicing on steady state protein production was evaluated by immunoblotting. Variants (n=41) with splicing defects fell into three groups: Group A -no protein due to a premature termination codon (PTC) introduced due to frameshift (intronic=7), Group B -shortened protein due to an in-frame deletion or PTC due to frameshift at 3' region of CFTR (intronic=19, missense=5), and Group C -full-length protein as a result of residual amounts of normally spliced mRNA isoforms (intronic=8, missense=1). Strikingly, splice variants at 5' region of CFTR (intronic=6) produced shortened protein despite introduction of PTC, potentially due to an alternative start site.

To evaluate the feasibility of using modulators to improve CFTR function in CFBE stable cells, ten variants were selected (missense=2 and intronic=2 from group B, and intronic=6 from group C). Group B variants (e.g. 489+1G>T and I175V) that produce only misfolded protein generated no functional CFTR in response to modulators. Either new modulators or gene correction will be required for group B splice variants. Modulators will not be effective for group A splice variants. However, each variant from group C (e.g. c.3717+40A>G) displayed residual CFTR function and responded to correctors (lumacaftor or tezacaftor) with further enhanced CFTR function upon ivacaftor treatment. This study illustrates that systematic theratyping can optimize assignment of precision therapies to splice-site variants. MD, USA; 3. Genetic Epidemiology Research Branch, National Institute of Mental Health, Rockville, MD, USA; 4. Pediatrics, National Jewish Hospital, Denver, CO, USA; 5. Johns Hopkins Bloomberg School of Public Health, Baltimore, MD, USA Introduction: Bitter taste receptors play an instrumental role in the innate immune response by increasing ciliary beat frequency (1, 2) and mucociliary clearance in airway epithelial cells (3) . Single nucleotide polymorphisms (SNPs) within the TAS2R38 bitter taste receptor gene have recently been linked to chronic rhinosinusitis. The activity of the TAS2R38 protein depends on 3 common SNPs within the TAS2R38 gene, creating a "functional" (PAV) or "nonfunctional" (AVI) haplotype. Patients homozygous for the "functional" haplotype were found to have less severe sinus disease when compared to heterozygotes or patients homozygous for the "nonfunctional" haplotype (4) . Our objective was to determine whether the functional haplotype of the bitter taste receptor TAS2R38 is protective against sinus disease in patients with cystic fibrosis.

Methods: We conducted a retrospective matched case-control study using the prospective Cystic Fibrosis Twin and Sibling Study database (5), where genotyping across the whole genome was previously performed. Four models were conducted matching cases to controls based on monozygotic twins (model 1), dizygotic twins (model 2), nontwin siblings (model 3) , and all patients only controlling for zygosity and family ID (model 4). Secondary outcome analysis was performed using the binary variable of whether patients underwent sinus surgery.

Results: Of the 2,086 genotyped cystic fibrosis patients, 1,406 (67.4%) were heterozygous for the "functional" haplotype (PAV/*). Only 7 patients total were homozygous for the functional haplotype (PAV/PAV). A total of 1,406 patients (748 males and 1,303 Caucasians), were evaluated for sinus disease, for which 530 patients had a positive diagnosis. Of those with sinus disease, 283 underwent surgery for sinus disease. A total of 151 (model 1), 50 (model 2), 1,205 (model 3) , and 1,406 (model 4) matched case controls were used. In all 4 models, there was no significant increase in the odds of protection from sinus disease in the functional haplotype heterozygotes, (OR = 1.3, P= 0.6; OR= 1.3, P = 0.5; OR = 1.0, P= 0.8; OR = 1.1, P = 0.7, respectively). Using sinus surgery as an outcome variable, similar results were found. Patients homozygous for functional haplotype (n=7) were distributed evenly between sinus and no sinus disease.

Conclusions: The presence of "functional" TAS2R38 haplotype does not appear to protect against sinus disease in patients with CF. Prospective studies in this population differentiating severity of sinus diseases are needed.

Background: CF newborn screening (NBS) began in New York State using an IRT-DNA algorithm in 2002. Despite several improvements, a high false positive rate and low positive predictive value (PPV=3.7%) prompted transition to a three-tier IRT-DNA-SEQ algorithm.

Objective: To compare screening metrics and infant outcomes using IRT-DNA and IRT-DNA-SEQ.

Methods: IRT-DNA-SEQ was implemented 12/1/2017. Infants with high IRT (top 5%) were screened for 39 variants using Luminex xTAG CF39v2. Those with one variant or ultra-high IRT (top 0.1%) were comprehensively sequenced using the Illumina MiSeqDx CF Clinical Sequencing Assay (CSA) with supplemental testing to target specific deletions. Detected variants were confirmed using Sanger sequencing. Infants with two CFTR variants were referred to specialty care centers for diagnostic testing; all others were issued reports and considered screen negative. Benign/likely benign variants were not reported.

Results: During the first four months, 38 infants with two variants were referred. With the original IRT-DNA algorithm, 264 infants would have required referral; sequencing identified 161 as carriers and 65 with no clinically significant variants. The average number of referrals per month decreased from 73 to 9. Median age at referral was 14 days (range [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] . CF has been confirmed in nine infants (PPV=25.7%; 9/35); diagnosis is pending for three. Twenty-six infants with two variants and intermediate or low sweat chloride levels are classified as CRMS/CFSPID and will be followed; 25/26 carried at least one variant of varying clinical consequence or uncertain significance. The CRMS/CFSPID to CF ratio increased from 0.9 to 2.9. Median age at first sweat test was 24 days (range 10-72; N=35).

Conclusions: Implementation of IRT-DNA-SEQ resulted in 85.6% reduction in referrals and 7-fold increase in PPV. Referral was delayed by three days compared to IRT-DNA. We project 110 infants will be referred annually, and 650 infants will be reported as negative after comprehensive genetic analysis. Additional testing increases lab workload, turnaround times and cost, but is balanced by reduction in follow-up by NBS staff and care centers, and an overall reduction in healthcare costs due to elimination of most false positive screens. The identification of asymptomatic infants with two CFTR variants, especially variants of uncertain or varying clinical significance, with intermediate/negative sweat results poses new clinical and practical challenges for care teams and families. CF-related diabetes (CFRD) is one of the most important complications of CF, causing worse lung function, malnutrition, and mortality in CF. CFRD develops over time and occurs about 1.6-fold more frequently in females. We previously found variation in CFRD onset to be heritable, i.e., attributable to variants outside of CFTR and identified CFRD modifiers near SLC26A9 and at T2D susceptibility loci (TCF7L2, CDKAL1, CDKN2B-AS1, and IGF2BP2) using data from 3059 individuals. We now report a genome-wide association study (GWAS) using an additional 2681 individuals provided by the International CF Gene Modifier Consortium (total n=5740). Variants at PTMA, SLC26A9 and TCF7L2 associated with CFRD onset with genome-wide significance. Association of variants near PTMA (eg, rs838455; p:3.9e-8, HR:1.75), which encodes for the secreted peptide thymosin alpha 1 (Tα1), is a novel finding. Tα1 is being investigated as a potential single-molecule-based therapy for CF (1, 2) . Variants at SLC26A9 (eg, rs4077468; p:2.75e-8, HR:0.72), which encodes for a chloride/bicarbonate exchanger that interacts with CFTR, were confirmed to associate with CFRD. Variants at the T2D locus TCF7L2 associated with CFRD (e.g. rs7903146; p:4.17e-12, HR:1.51), indicating some degree of shared pathophysiology. Other T2D-associated variants in CDKAL1, IGF2BP2 and CDKN2B-AS1 are also CFRD modifiers, however the CFRD modifier variants at CDKN2B-AS1 are most strongly associated with coronary artery disease, not T2D, in the general population.

Since CFRD occurs more often in females, we investigated sex-specific associations. The CDKN2B-AS1 association with CFRD is sex-specific (e.g., rs1333045; male-only p: 0.10 HR: 0.91; female-only p:3.2e-7, HR: 0.72; interaction p: 7.0e-3). Additionally, through an interaction term GWAS, we identified variants intronic of CNTN3 that have sex-specific associations (e.g., rs79480701; interaction term p: 3.5e-8). These variants associated with CFRD in males and females in opposite directions (male-only p: 8.6e-6, HR: 0.31, female-only p: 4.4e-4, HR: 2.09), therefore may be influencing the earlier onset of CFRD in females, as the common allele is associated with earlier CFRD onset in females.

To further explore overlapping genetic risk factors for CFRD and T2D, a log odds ratio weighted polygenic risk score (PRS) for T2D was constructed from LD-pruned (r-squared<0.1), common (MAF>5%) variants reported in the NHGRI-EBI GWAS Catalog (n=97). The T2D PRS was associated with CFRD (p: 1.3e-14; HR: 2.38), and remained associated after excluding the T2D variants identified above (p: 3.3e-6, HR: 1.80). Similarly constructed PRSs for birth weight, glucose tolerance and proinsulin level associated with CFRD (p<0.05).

These studies identify genetic modifiers of CFRD onset, which may act through CF-specific, T2D-specific, or sex-specific pathways. The CFTR-France database is dedicated to the annotations of rare CFTR variants in the context of their cis-and trans-allelic combinations, associated to patients' phenotypic data. The development of next-generation sequencing (NGS) technologies notably enables the sequencing of the whole CFTR locus and the identification of deep-intronic disease-causing variants in patients. The aim of our work is to collect and annotate all the variants identified by this approach in CFTR-France, in order to provide specific curated analytical data.

Methods: We sequenced the whole CFTR locus of 100 individuals (63 patients and 37 healthy individuals). In order to facilitate the priorization of variants, we developed an in-house tool based on the filtering of ANNOVAR annotated VCF (Variant Call Format) files, in which criteria are editable (e.g., variant's location, gnomAD frequency threshold, quality of reported variants, variant fraction). The variants that fulfilled the filtering criteria were finally included in CFTR-France as a "NGS variants catalogue."

Results: Our specific filtering tool allowed us to select 633 variants, mostly deep intronic variations (n=584). For 70 of the 633 variants, that were already reported in CFTR-France, the genetics and phenotypic data of the corresponding individuals were included in the database. Among the 563 variants that were not in CFTR-France, 215 were neither reported in gnomAD nor in dbSNP databases (210 probably not pathogenic and 5 proven or highly suspected to be disease-causing). Depending on the clinical context, a few novel variants (n=17) were tested by Sanger sequencing, to confirm or refute NGS results, highlighting 2 false positive variants. In CFTR-France, variant's specific web page contains key information about the variant, and links to core and CFTR databases. Specific annotations are also provided for the variants of the "NGS catalogue," including the allele frequency in our population of sequenced individuals, relevant quality criteria, Sanger verification and the resulting variant status (i.e., true or false positive), and the results of in silico predictions or functional studies.

Conclusion: The expanded version of CFTR-France aims i) to offer a valuable tool for the interpretation of CFTR variants identified by highthroughput approaches, thanks to the NGS catalogue, and ii) to provide an open access to our in-house filtering tool for the priorization of the variants contained in any ANNOVAR annotated VCF file. Acknowledgment: Supported by the French association Vaincre la Mucoviscidose.

Macek, M.; Libik, M. Department of Biology and Medical Genetics, Charles University Prague, Prague, Czechia Introduction: We aim to sequence the entire CFTR gene in cases with clinically diagnosed or unconfirmed clinical diagnosis of cystic fibrosis (CF) in countries where there is limited access to reliable sweat testing to confirm the clinical diagnosis of the diseases and/or in whom CF-causing variants (henceforward mutations) were not detected in one or both CFTR alleles.

Methods: Cascade examination of pseudoanonymized cases was conducted by a) Elucigene CF-EU2v1 (elucigene.com) followed by, b) next generation sequencing (NGS) MiSeq (illumina.com) using Devyser CFTR NGS assay covering all coding CFTR exons and all adjacent exon/ intron boundaries, including minimum 20 bp proximal to the 5' end and 10 bp distal to the 3' end of each exon boundary (devyser.com). CFTR variant pathogenicity assignment was according to cftr2.org using the "legacy" nomenclature. The following CF patient cohorts were analyzed, thus far: Armenia (40 cases), Bahrain (40), Belarus (21), Bulgaria (46), Cyprus (37), Georgia (56), Kyrgyzstan (6), Romania (26), Ukraine-(Lviv/Uzhorod; 96), including cohorts (from Croatia, Egypt, Jordan, Kazakhstan) obtained at the time of abstract submission.

Results: The studied populations have both common European alleles (e.g. F508del, N1303K, CFTRdele2,3/21kb/), European alleles at higher frequencies or population specific mutations (e.g. Armenia: 2043delG, 3121-1G>A; Bahrain: 2043delG, 3120+1G>A, 3121-1G>A; Belarus: 2143delT; Cyprus: 4201delTG; Georgia: 1677delTA. Kyrgyzstan: 1898+2 T>C and Ukraine: 2184insA, S466X. In addition, deep intronic mutation was found in a patient of Russian origin -1584+18672 A>G.

Conclusions: Studied populations have distinct distributions of CF-causing mutation that are relevant for confirmation of diagnosis of the disease at this stage of development of CF diagnostics and care. We have identified population specific mutations and raised mutation detection rates to over 95% in most populations studied.

The ability to detect lung disease in children via a noninvasive method would be highly beneficial for cystic fibrosis (CF) clinical care. Development of molecular biomarkers that are prognostic for CF lung disease could allow earlier and hence more effective intervention strategies. Ideally these would be molecules that are stable in body fluids and are mechanistically linked to pulmonary infection and/or inflammation. MicroRNAs are short nucleotide regulatory sequences that bind to target mRNAs to inhibit their expression. The CF lung is known to exhibit an altered miRNA expression profile, including miRNAs with targets such as CFTR, IL-8 and numerous other inflammatory mediators. A gender dichotomy exists in CF, whereby females are at a clinical disadvantage e.g. poorer lung function, earlier colonization by Pseudomonas aeruginosa, greater frequency of exacerbations, and lower median survival age. The aim of this study was to compare the plasma miRNA expression profiles between male and female CF children in order to screen for candidate biomarkers of clinical outcomes that are gender correlated. miRNA expression was quantified in paediatric CF plasma from six males versus six females (Age range: 1-6 years; median age: 3 years; 9 Phe508del heterozygotes) using TaqMan OpenArray Human miRNA Panels. Principal component analysis indicated differences in male versus female miRNA profiles. Two miRNAs were significantly increased in the female samples (hsa-miR-885-5p, fold change: 5.07, adjusted p value: 0.026; and hsa-miR-193a-5p, fold change: 2.6, adjusted p value: 0.031), however only hsa-miR-885-5p was successfully validated by qRT-PCR (p< 0.0001). Gene ontology analysis of validated targets of miR-885-5p identified cell growth, adhesion and migration/motility (including neutrophil chemotaxis), as processes potentially affected, with Ras-related C3 botulinum toxin substrate 1 (RAC1)-mediated signalling featuring significantly. To conclude, there is a significant increase in hsa-miR-885-5p in plasma of CF females versus males under six years of age; this miRNA may therefore have a role in the CF gender gap and may hold future potential as a biomarker. Introduction: Interpatient variability in drug response ranging from lack of efficacy to undesirable adverse drug events due to genetic polymorphisms in drug metabolism, transport, and target receptors is well known. CFTR genotyping is well established but the effects of other gene variants in CF on drug response are less well described. The objective of the study was to determine the distribution of phenotypes and the frequency at which a change in medication prescribing is recommended based on presence of a genetic variant linked to drug response.

Methods: Pharmacogenomics (PGx) testing was offered to all patients attending the adult CF clinic as part of routine care starting January 2018. DNA samples were obtained via buccal swabs. Genotyping was conducted on a selected panel of pharmacogenes at a CLIA accredited clinical laboratory (AlphaGenomix). The patient's medication profile was provided to the lab to generate a customized report on PGx results that were actionable. The Clinical Pharmacogenetics Implementation Consortium (CPIC) and FDA guidelines were used to determine the phenotype and the frequency at which a change in medication prescribing is needed.

Results: PGx testing was performed on 67 adults with the following characteristics: mean age 34 years old; 49% male; mean BMI 23 kg/m 2 , mean percent predicted FEV 1 73%, and 22% were Hispanic. As expected, the highest proportion of patients were extensive (normal) metabolizers for the various CYP enzymes. Notably, the proportion of patients predicted to have altered drug metabolizer status for CYP2D6, CYP2C9, and CYP2C19 was: 18%, 33%, and 43% respectively. These three enzymes are important in the metabolism of opioids (CYP2D6), warfarin (CYP2C9), antidepressants (CYP2C19), and proton pump inhibitors (CYP2C19). In addition, 33% expressed low levels of SLC6A4 (serotonin receptor), which may lead to delayed response to the selective serotonin reuptake inhibitor (SSRI) class of antidepressants; and 25% had gene variants of VKORC1 and CYP2C9, predicting increased sensitivity to warfarin. Based on the medications prescribed, 42% of patients were receiving therapy for which a change in therapy (agent or dose) is recommended. Of the therapies the patients received for which a change in therapy was recommended, the most common drugs with recommendations for change were antidepressants (40%), PPIs (26%), lorazepam (14%) and tacrolimus (7%).

Conclusions: Overall, two in five adult patients with CF may have benefited from a change of therapy according to their genotyping results and the CPIC and FDA guidelines. Most of the changes in therapy would have affected antidepressants and proton pump inhibitors. This study provides evidence in favor of preemptive PGx testing by identifying a significant proportion of patients with actionable PGx variants.

Acknowledgments: Supported in part by funding from AlphaGenomix Laboratories. A TARGETED GENOTYPE  BY SEQUENCING PLATFORM FOR THE  COMPREHENSIVE CHARACTERIZATION OF THE  CFTR The CF Foundation Therapeutics Lab receives CF and normal lung explants from ethnically diverse donors, some of which can harbor rare variants of the gene coding for the CFTR chloride channel protein. Human bronchial epithelial (hBE) cells are isolated from these tissues, banked, and are used in downstream functional assays. Knowledge of the mutational landscape of CFTR will aid in the interpretation of functional data and support basic research studies as well as drug discovery and development in which these cells may get utilized. Therefore, we sought to develop a method that would produce a comprehensive catalog of all variants within exons, introns, untranslated regions, and flanking regions for point mutations, indels, complex alleles and large genomic rearrangements in CFTR.

While PCR based panels are widely employed to assay a limited number of CF associated variants, next generation sequencing (NGS) permits the identification of all known variants as well as private mutations and gross genomic rearrangements without a priori knowledge of the ethnographic origins of donors. Consequently, we developed an unbiased target enrichment strategy of CFTR followed by deep short read sequencing, which is complemented by Oxford Nanopore Minion long read sequencing as a means of phasing alleles and to differentiate carriers from those with two defective alleles. The sequencing data are processed with a bioinformatics pipeline that includes read alignment to genomic reference, variant calling, and alignment and variant QC.

So far, this approach has been applied to 28 hBE lines from CF and normal donors and CRISPR/cas9 gene-edited 16HBE14o-human bronchial epithelial cell lines. We identified known CF-associated variants in CF hBE donor cells and the gene-edited cell lines, but we also identified an unexpected presumably pathogenic R792G variant in a heterozygous non-CF donor. This assay is a significant improvement to existing PCRand NGS-based methods. The complete characterization of CFTR variants will help us in the interpretation of in vitro studies and advance our understanding of how these variants may serve as genetic modifiers of CF.

Santos, S. 1 ; Oliver, K.E. 2 ; Sorscher, E.J. 2 ; Hartman, J.L. 1 1. Genetics, University of Alabama at Birmingham, Birmingham, AL, USA; 2. Pediatrics, Emory University, Atlanta, GA, USA We utilize the yeast homolog of CFTR, YOR1, as a model protein for "phenomic" analysis, or the systematic evaluation of gene interactions. Phenomics reveals potential genetic modifiers and/or therapeutic targets to alleviate CF-causing variants. Quantitative high-throughput cell array phenotyping (Q-HTCP) is a technology we developed to measure gene interactions across the entire collection of yeast gene knockout and knockdown (YKO/KD) strains. Gene interactions are quantified by Q-HTCP-derived cell proliferation parameters that define phenotypic dependency between yor1 mutant alleles and yko/kd mutations, and thus define influence of gene networks potentially relevant to CFTR function. In this context, we perform systematic comparisons of wild-type, parent, single mutant and double mutant backgrounds, and incorporate analysis of yor1-Δ0 (control null allele), and yor1 encoding homologous mutations from CFTR variants of interest. We also show evolutionary conservation of gene interactions using mammalian airway models (primary human bronchial and nasal epithelia). The Q-HTCP workflow for phenomic analysis and genome-wide measurement of gene interaction has recently been refined and optimized further for studies of CF. In particular, we developed and applied an analytical workflow to integrate, compare and contrast gene interaction profiles relevant to CFTR obtained from different screens with the aim to more comprehensively identify functional relationships between all molecular targets that harbor the potential to rescue defective CFTR from numerous mechanistic subcategories. We applied recursive expectation maximization clustering (REMc), developed by our laboratory and used in conjunction with Gene Ontology enrichment analysis, to identify functional gene modules that include components of the proteosome, pathways for protein folding, and targets contributing to translational fidelity. Resulting phenomic data also provides "toxicity" information in that the primary effect of each modifier on cell fitness is assessed in conjunction with its influence on the function of the YOR1/CFTR model. This integrative analysis is intended to provide a systematic, quantitative, and comprehensive summary of genetic influences on CF disease expression resulting from differing alleles in a manner that can be directly correlated with genomewide association studies (GWAS) or emerging properties of complex CFTR alleles. As such, the approach represents a powerful and high-resolution model for developing personalized therapeutic strategies for individuals who possess compound heterozygous variants and/or mutation(s) unresponsive to currently available CFTR modulators. 1 1. SickKids, Toronto, ON, Canada; 2. University Health Network, Toronto, ON, Canada Introduction: Nasal cell cultures generated from nasal brushes of individual patients are being explored as a tool to evaluate individual responses to CFTR-modulating drugs. However, there is insufficient information on whether they represent the response of bronchial epithelial cells.

Method: Nasal and bronchial brushing was performed on CF patients post-lung transplant during surveillance bronchoscopies. The success of generating primary cell cultures was 85% for nasal and 70% for bronchial brushes. CFTR function was measured in Ussing chamber studies (Rte-transepithelial resistance, Ieq-equivalent short circuit current, IeqFSKforskolin induced Ieq). A Pearson correlation was performed between nasal and bronchial outcomes. Further, a Bland-Altman plot was used to evaluate agreement in outcomes between cell types.

Results: Preliminary results on matching nasal and bronchial cell cultures generated from 17 F508del/F508del CF patients from brushes demonstrated a very good correlation of the basic bioelectric parameters, such as baseline Rte (r=0.7, p=0.007), baseline Ieq (r=0.88, p<0.0001) as well as the amiloride-sensitive Ieq (r=0.88, p<0.0001 Birmingham, Birmingham, AL, USA; 3. Cystic Fibrosis Foundation, Bethesda, MD, USA; 4. Massachusetts General Hospital, Boston, MA, USA; 5. Cincinnati Children's Hospital Medical Center, Cincinnati, OH, USA; 6 . University of Colorado School of Medicine, Aurora, CO, USA Background: Human airway epithelial (HAE) cells in differentiated monolayers have formed a key basis for preclinical evaluation of CFTR modulators. This concept has been leveraged through multisite collections of human nasal epithelial (HNE) cells through CF Foundation Therapeutics Development Network sponsored observational studies. In addition, our lab has demonstrated that in vitro improvements in electrophysiologic and mucus transport assays performed on HAE correspond to clinical improvements in sweat chloride and lung function in patients on CFTR modulator therapy.

Objectives: This study aimed to determine whether HNE cells recapitulate key in vitro responses seen previously in human bronchial epithelial cells and if the responses correlate to individual subject's clinical improvement to ivacaftor in GOALe 2 .

Methods: HNE cells were collected from 24 donors participating in the GOALe 2 study and were expanded using conditional reprogramming. After overnight shipment, cells were differentiated at air-liquid interface using Ultroser G supplemented media for 28 days. Inserts were analyzed for forskolin-stimulated CFTR-dependent current using modified Ussing chambers in the presence or absence of ivacaftor (IVA, 10 µM) followed by CFTR Inh -172 (20 µM). In addition, each donor was analyzed for change in mucociliary transport (MCT) rate, airway surface liquid depth (ASL), and ciliary beat frequency (CBF) using micro-optical coherence tomography.

Results: Differentiated monolayers from 24 donors (20 F508del/ G551D, 4 F508del/R117H-5T) have been analyzed. Monolayers demonstrated a mean forskolin+IVA-stimulated I sc of 6.8±4.0 µA/cm 2 (vs 2.3±0.5 µA/cm 2 for vehicle, p<0.005, 48% of wild-type CFTR response). CFTR Inh -172 responses mirrored these differences. Monolayers demonstrated improved MCT (4.3±0.4 mm/min IVA vs 2.9±0.5 mm/min vehicle control, p<0.005) and ASL (13.9±4.9 µm IVA vs 7.3±2.5 µm vehicle control, p<0.05). No significant differences were observed in CBF. When IVA treated I sc was compared to post-ivacaftor sweat chloride in the same subjects, moderate correlation was observed (r = 0.51, p<0.05). There was a trend toward correlation between individual IVA treated I sc and absolute change in percent predicted (pp) FEV 1 (r = 0.74, p=0.06). Improvements in ASL correlated with ppFEV 1 after 6 months of treatment (r=0.51, p <0.05) but neither MCT nor CBF demonstrated a discernible correlation to ppFEV 1 .

Conclusions: Results from the GOALe 2 HNE Study demonstrates that HNE cells recapitulate treatment effects of CFTR modulator treatment. Further, the change in I sc correlated to sweat chloride on an individual level, whereas changes in ASL depth and I sc correlated to improvements in ppFEV 1 . This tool may be useful as an in vitro biomarker to predict treatment response to CFTR-directed therapies.

Acknowledgments: Supported by CFFT, CFF, and NIH. Cystic fibrosis (CF) is a monogenic disease caused by mutations in the gene coding for the chloride channel cystic fibrosis transmembrane conductance regulator (CFTR). Currently, more than 30,000 individuals in the US are living with CF and harbor over 1,700 reported mutations. Counterintuitively, despite being monogenic, the large number of distinct and rare variants of CFTR presents both a serious therapeutic selection challenge and confounds new drug discovery efforts. First-generation CFTR modulators targeting the defective protein are now FDA-approved for 34 genetic variants, which comprise ~59% of the patient population. New highly efficacious triple modulator therapy is currently in clinical development for all patients with at least one CFTR F508del allele. People bearing certain other CF variants will also benefit from the new therapy, but about 5% of all CF cases are not likely to improve from such treatment. These people typically carry two alleles that make no functional protein due to premature termination codon or severe splice mutations as well as some refractory missense mutations, e.g., N1303K.

Drug discovery and development for the CF-causing variants without effective disease-modifying therapy on the horizon, require better models than cell lines heterologously expressing CFTR cDNA constructs. For example, in the context of PTC variants, transcriptional (nonsense mediated mRNA decay), translational (readthrough) and posttranslational modifications (folding/glycosylation) are not fully reflected in cDNA systems and have been shown to affect the amount and/or function of CFTR.

Inducible pluripotent stem cells (iPS cells) offer a promising approach for generating physiologically-relevant systems to aid in CF drug development for all mutations. iPS cells expand nearly limitlessly in an undifferentiated state, are adaptable for gene editing approaches, and can be differentiated into multiple CFTR-expressing cell lineages. This report details how iPS-based protocols are being developed for drug screening. Currently, we are working on an iPS directed differentiation protocol designed to assess CFTR on multiple levels. That is, CFTR responses to potential therapies can be assayed accurately utilizing biochemistry (Western/qRT-PCR), and electrophysiology or organoid swelling for function.

Cell-based therapy approaches may also be built on iPS-derived human bronchial epithelial (hBE)-like cells. Here, we show the single cell transcriptional profile of primary hBE cells compared to differentiated iPS-derived hBE-like cells, data that should further our directed differentiation output to become more physiologically relevant. 1 1. CFFT Lab, Cystic Fibrosis Foundation, Lexington, MA, USA; 2. UT Southwestern Medical Center, Dallas, TX, USA More than 1,700 genetic variants of the CFTR gene have been suggested to cause cystic fibrosis (CF). The ability to assess the effectiveness of both approved drugs and therapeutics in development against rare, diseasecausing CFTR variants is limited by the scarcity of primary tissues and cells for these genotypes. Immortalized cell lines heterologously overexpressing mutant CFTR are commonly used as an alternative. However, these artificial expression systems have fundamentally limited relevance, especially for premature termination codon (PTC) mutations. To address these issues and create specific CF mutations in genomic CFTR, we have developed a CRISPR/Cas9 based gene editing platform that enables the generation of new cell lines with essentially any CFTR variant in its genomic context. For its compatibility with CFTR functional assays, we selected the 16HBE14obronchial epithelial cell line (Dieter Gruenert, UCSF) and optimized a gene editing workflow. We have used this platform to create isogenic, homozygous cell lines for six CF-causing variants (F508del, G551D, N1303K, G542X, W1282X, and Y122X) , and more are currently in development. The resulting cell lines provide a virtually unlimited source of material with specific patient mutations that can be used in a variety of assays.

PTC mutations in the CFTR coding sequence affect about 10% of CF patients. Finding effective therapeutics for CF patients with PTC variants is a serious unmet medical need. Drug-induced readthrough is a promising therapeutic approach that can restore expression of full length CFTR and consequently repress nonsense-mediated mRNA decay (NMD). Highthroughput screening campaigns using immortalized cell lines that heterologously overexpress CFTR PTC variants from cDNA are of reduced relevance because they are not subject to splicing-dependent NMD; so far these models have failed to identify agents for which readthrough activity and rescue of CFTR function translates to patient-derived human bronchial and nasal cells. The 16HBE14o-models of PTC variants are sensitive to NMD, overcoming the limitations of patient-derived primary cells and cDNA based heterologous cell lines. NMD causes reduced basal levels of CFTR mRNA in all three (G542X, W1282X, and Y122X) PTC cell lines. Treatment with a small molecule inhibitor of SMG1, one of the main effectors of the NMD pathway, can normalize mutant CFTR mRNA abundance. Interestingly, the sensitivity to SMG1 dependent NMD and the response to G418 mediated readthrough, as measured by mRNA abundance and CFTR function, varied between the different PTCs. W1282X is the most sensitive to NMD and less amenable to small molecule-mediated readthrough, while Y122X is more amenable to readthrough and only moderately sensitive to NMD. Together, these data suggest that treatment outcomes may vary depending on the specific PTC. These isogenic PTC cell lines will enable the discovery and development of effective NMD modulators and readthrough agents as well as inform therapeutic development strategies for specific PTC variants.

Ali, H. 1 ; Icyuz, M. 2, 4 ; Kelly, S. 2, 3 ; Rab, A. 1, 5 ; Hong, J.S. 1, 5 ; Bedwell, D. 2, 3 ; Hartman, J.L. 2, 4 ; Sorscher, E.J. 1, 5 ; Oliver, K.E. 1, 5 1. Dept of Pediatrics, Emory Univ., Atlanta, GA, USA; 2. Gregory Fleming James CF Research Center, Univ. of Alabama, Birmingham, AL, USA; 3. Dept of Biochemistry & Molecular Genetics, University of Alabama, Birmingham, AL, USA; 4. Dept of Genetics, Univ. of Alabama, Birmingham, AL, USA; 5. Children's Healthcare of Atlanta, Birmingham, AL, USA Introduction: Premature truncation codons (PTCs) account for ~13% of cystic fibrosis (CF) alleles and represent a group of individuals for whom no personalized medicines exist. PTC variants in the CF transmembrane conductance regulator (CFTR), such as W1282X, lead to truncated protein and inadequate mRNA levels attributable to nonsense mediated decay. Patients carrying these mutations typically manifest a severe phenotype and exhibit poor long-term clinical outcome.

Methods and Objectives: We used a novel strategy -genome-wide yeast phenomics -to identify new molecular targets that rescue the CFTR homologue, yeast oligomycin resistance-1 (Yor1). Yor1 is an ATP-binding cassette class C (ABC-C) protein that functions as an oligomycin efflux pump. Yor1-G704X, a mutation exhibiting similar characteristics to W1282X in CFTR, was crossed with ~6,000 single gene mutants from the yeast deletion strain library. Double mutant arrays were analyzed to identify individual gene deletions which enhanced oligomycin resistance. The top 12 molecular targets that robustly rescued Yor1-G704X were evaluated for genetic conservation between yeast and mammalian epithelial models.

Using siRNA, we silenced these homologues in Fischer rat thyroid (FRT) cells stably expressing W1282X-CFTR to assess: (i) plasma membrane localization of HRP-tagged W1282X-CFTR, and (ii) transepithelial ion transport via short-circuit current measurements.

Results: Repression of genes coding for 60S ribosomal proteins (RPL12, RPL8), an RNA processing enzyme (TRIT1), a proteasome constituent (PSMD6), and peroxisomal biogenesis factor 1 (PEX1) led to pronounced rescue of both Yor1-G704X and W1282X-CFTR. In particular, PEX1 knockdown significantly increased W1282X-CFTR trafficking to the cell surface when combined with G418 (an established read-through agent), and induced a 5-fold improvement in chloride transport following combination treatment with G418 and the CFTR corrector, VX-661 (tezacaftor). Moreover, Pex1-depleted FRT cells exhibited higher CFTR-dependent ion transport activity when acutely exposed to the CFTR activator, curcumin, effectively reaching ~ 12% of wild-type levels.

Conclusions: Pex1 is essential for the formation and function of peroxisomes, and its inhibition partially rescues Yor1-G704X and W1282X-CFTR in terms of both processing and function. Similar, strong effects were noted for additional targets not otherwise expected to mediate translational readthrough. Ongoing experiments include evaluation of synergy between PEX1 and other molecular targets, quantitative levels of gene repression, and mechanistic studies (e.g. CFTR readthrough in the setting of specific peroxisome defects). Findings are also being evaluated in primary human CF airway epithelia. This work expands our knowledge regarding molecular pathobiology of nonsense mutations, and in the future, may offer a new therapeutic strategy for CF patients harboring PTC variants. Recently, there has been a rapid expansion of the number of new potential CFTR-directed therapies. These CFTR modulators can be broadly categorized into potentiators, correctors, amplifiers, stabilizers and nonsense suppressors. While in vitro studies may provide support for moving forward with specific modulators, the number of patients needed and the expense of clinical trials limit the ability to test these modulators individually and in combination for FDA approval. Cystic fibrosis animal models allow for in vivo testing of the practicality and efficacy of these CFTR modulators but animal and human CFTR sequence and gating differences are of concern. Without in vivo expression of human CFTR containing CF disease-causing mutations, a true clinical prediction of CFTR-directed therapies effects in patients may not be achieved. We created such a model by combining the many benefits of a mouse model with the ability to express human CFTR containing common CFTR mutations under the control of its adjacent endogenous regulatory elements. The human CFTR-expressing mouse (hCFTR) was created by the insertion of a ~260 kb bacterial artificial chromosome (BAC) containing the human CFTR gene and surrounding sequence. The BAC inserted into a single integration site on mouse chromosome 8. To verify that hCFTR expression could complement the loss of mouse CFTR (mCftr), the hCFTR expressing strain was crossed to a strain carrying a mCftr null mutation. hCFTR positive(+) CF mice displayed CFTR function, normal survival, intestinal histology, and growth while hCFTR(-) CF mice displayed reduced survival due to intestinal obstruction, abnormal intestinal histology and reduced growth. Utilizing the hCFTR expressing mouse strain, we successfully generated five human CFTR mutations (F508del, G542X, W1282X, G551D and 3849+10kb C>T) that allow various CFTR-directed therapies to be tested. We generated these specific CFTR mutations utilizing the genome editing system CRISPR/ Cas9. Optimal single guide RNAs, Cas9 nuclease and oligonucleotides carrying each CFTR mutation were injected into one-cell mouse embryos and implanted in pseudo-pregnant mice. DNA from the resulting pups of these five lines was sequenced using next generation sequencing with a MiSeq instrument. We obtained a high percentage of founder mice with each mutation (F508del(40%), G542X(50%), W1282X(33%), G551D(70%) and 3849+10kb C>T(13%)). Founders with each mutation were bred to mice carrying a mCftr null mutation to create CF mice carrying each hCFTR mutation. CF mice carrying mutant hCFTR have displayed lack of CFTR function and typical CF manifestations (e.g., poor growth and increased incidence of intestinal obstruction). Importantly, hCFTR F508del mice display correction of CFTR function using CFTR modulators (VX-809/VX-770 or VX-661/VX-770) in airway and intestinal cells. We are currently evaluating the ability of other CFTR modulators to restore CFTR function and alleviate CF manifestations in all 5 human mutation carrying strains. These models provide novel tools for in vivo evaluation of CFTR modulators as well as CFTR gene editing strategies for the treatment of CF. Introduction: Nonsense mutations affect 11% of patients with cystic fibrosis (CF) worldwide. Insertion of a premature termination codon (PTC) in the CFTR transcript subjects it to degradation by the surveillance mechanism nonsense mediated decay (NMD). Mouse models of CF are useful to evaluate protein expression and function, but are limited by failure to recapitulate the phenotypic manifestations of human disease. In this study, we generated a novel knock-in G542X rat model using CRIS-PR-Cas9 technology (Horizon Discovery) and characterized its phenotype, NMD sensitivity, and initial treatment response with amikacin, an aminoglycoside shown to induce translational readthrough in vitro and in CF G542X transgenic mice.

Methods: G542X-CFTR function was assessed by in vivo measurements of nasal potential difference (NPD) and ex vivo short circuit current (I sc ) of ileum and trachea. RT-qPCR and Western blot were performed for measuring CFTR transcript and protein levels, respectively. Liquid chromatography-mass spectrometry (LCMS) was used to measure the serum levels of amikacin following SC administration at 170 mg/kg/day for 4 days.

Results: NPD measurements showed no CFTR function in G542X rats compared to wild-type (WT) (Cl --free forskolin = -0.4±1.5 and -13.4±3.5 mV, respectively, P<0.01). Ileal I sc measurements showed minimal forskolin stimulation in G542X rats (29.8±4.0 µA/cm 2 ) compared to WT (188.0±11.7 µA/cm 2 , P< 0.0001). Trachea demonstrated diminished baseline I sc (25.5±9.1 µA/cm 2 ) compared to WT (608.8±33.4 µA/cm 2 , P<0.0001), with minimal CFTR inh -172 sensitive current in G542X (-5.431±5.48) relative to WT (-392.7±76.25, P<0.0001). CFTR mRNA levels of excised lung were ~5 fold less abundant in G542X rats compared to WT (P<0.005), suggesting sensitivity to NMD, as also observed in CF patients with PTCs. As has been shown previously (J Clin Invest. 2007; 117:683-92) , NMD efficiency is variable among CF patients and has been regulated by a negative feedback mechanism (Mol Cell. 2011; 43:950-61) . We have measured 4 conserved NMD factors (UPF1, UPF3b, SMG1 and SMG6) and 2 NMD substrates (RPL3 and SC35). Transcript levels of each were diminished in G542X rats compared to WT (~ 3 fold, P<0.05 for each), suggesting NMD might be globally upregulated in G542X rats in response to the nonsense mutation. Amikacin peak and trough levels at steady state were 177.5±37.5 µg/mL and 0.586 ± 0.154 µg/mL, respectively. While NPD and I sc showed no response to amikacin SC treatment at 170mg/kg/day for 12 days, baseline tracheal currents were significantly higher in amikacin treated rats (206.5±35.23 µA/cm 2 , P<0.05), potentially indicating early evidence of a treatment effect.

Conclusions: We generated a novel rat model expressing the CFTR G542X mutation that is plagued by NMD and is sensitive to respiratory and gastrointestinal abnormalities characteristic of the CF defect. Early evidence suggests global enhancement of NMD, which contribute to low levels of CFTR nonsense transcripts that are expected to limit the efficacy and clinical benefits of readthrough therapy, as seen in CF patients. This model presents a useful tool to evaluate the rescue CFTR expression and function using nonsense mutation directed therapeutic testing in vivo. CF Center, UNC, Chapel Hill, NC, USA; 2. Physiology, Wayne State Univ., Detroit, MI, USA Animal models of cystic fibrosis (CF) have been instrumental in furthering our understanding of CF pathophysiology. Rabbit models of CF have been developed to capitalize on the following features: 1) high (92%) CFTR sequence homology with humans; 2) body weights sizable relative to human newborns, amenable to equivalent imaging/pulmonary function testing; 4) cost-efficiency, compared to pig and ferret; and 5) absence of airway submucosal glands, ideal to study superficial epithelial-specific contributions. CFΔ1 New Zealand White rabbits developed by the Sun laboratory (homozygous for a single nucleotide deletion creating a premature stop codon in CFTR, i.e., a functional CFTR KO) were imported to UNC, bred, and studied with the goal of extending their lifespan and characterizing their phenotype across multiple sites. CF rabbits did not exhibit perinatal meconium ileus, but over time they develop large intestine/ cecum obstruction/stasis causing significant morbidity and mortality. Starting at two weeks of age, all rabbits were given Golytely (osmotic laxative) instead of water, weighed, and their abdomens palpated for masses (indicative of gut stasis) three times/week. Using this approach, 56% of CF kits survived to weaning (7 weeks). We have developed interventional strategies to reduce post-weaning GI disease. If abdominal masses were detected, rabbits were treated twice a day with: 1) 50 mL SQ warmed saline; 2) PO 0.5 mg/kg cisapride (increases motility of lower gut; 3) PO 0.5 mL/kg simethicone (anti-gas agent); 4) PO 0.5 mg/kg metoclopramide (increases motility of small bowel); and 5) PO 0.5 mL/kg lactulose (a laxative). Treatment continued until masses resolved and was successful in rescuing ~70% of CF rabbits exhibiting gut masses. This strategy allowed us to raise 4 CF rabbits to 1 year of age. As previously reported, the bioelectric phenotype of CF rabbits was consistent with absence of functional CFTR in the GI tract and trachea. CF rabbits also exhibited an elevated nasal basal potential difference (PD) compared to wild-type (WT) controls (-36.3±4.0 mV vs. 8.6±3.1mV, respectively. p≤ 0.001, n=5), a greater PD depolarization upon amiloride perfusion (Δ 18.9±1.9 mV vs. 4.4±1.3mV, CF vs. WT respectively), and no response to Clfree buffer. Evidence of olfactory epithelial degeneration was observed in 1-year-old rabbits. At this time, we have not phenotyped sufficient CF rabbits to determine whether the 10% penetrance of muco-inflammatory disease observed in Dr. Sun's lab is also observed at UNC. Using RNAscope in situ hybridization, CFTR expression was detected throughout the WT rabbit distal airway epithelium, similar to humans but not mice, suggesting that CF rabbits may exhibit CFTR-dependent responses to airway challenges. Preliminary data indicate that rabbits infected by respiratory viruses exhibit distal airway muco-cellular obstruction, providing an in vivo model to test the effect of CFTR deletion on viral challenge. Current efforts are geared towards assessing: 1) whether viral infection causes differential incidence/kinetics of muco-inflammatory disease in CF vs. WT rabbits; and 2) whether the nasal cavity of CF rabbits is a useful model to test CFTR modulators.

Funded by the CFF. Background: People with cystic fibrosis exhibit growth defects and brittle bones. That observation has been attributed, in part, to malnutrition and chronic pulmonary inflammation. We tested the hypothesis that disruption of the cystic fibrosis transmembrane conductance regulator (Cftr) gene directly affects bone microarchitecture and integrity by studying bone of newborn Cftr -/pigs.

Methods: We examined femoral cortical and trabecular bones of Cftr -/pigs less than 24 hours after birth using microcomputed tomography (mCT, Skyscan 1076, Bruker). Scans were performed with the following settings: tube voltage, 80 kV; tube current, 0.125 mA; and voxel size, 17 x 17 x 17 mm (x, y, z). Three-dimensional images were reconstructed and analysed using the NRecon GPU version and CTAn (Bruker) software programs, respectively. The cortical bone porosity and structure were defined using a 3.4 mm wide region centered on the middle of the femur. A total of 37 newborn Cftr -/piglets (24 males and 13 females) and 18 newborn Cftr +/+ piglets (8 males and 10 females) was subjected to mCT scan.

Results: Compared to newborn Cftr +/+ pig controls, Cftr -/femoral bone exhibited significantly lower total volume (TV), bone volume (BV) and bone volume density (BV/TV) but only in females. However, the Cftr -/bone mineral density (BMD) in trabecular and cortical tissues was significantly reduced in both sexes, compared to Cftr +/+ piglets. Interestingly, focusing at the porosity of cortical bone in Cftr -/pigs as a determinant of bone fragility associated with high fracture risk, we observed higher closed porosity with a marked increase of closed pore surface in cortical bone of Cftr -/pigs (+18.7% for males and +48% for females). These results suggest a lower bone remodelling, lower interconnectivity within the vascular network, and increased bone fragility in Cftr -/animals. No significant difference was observed in the open cortical porosity, whatever the gender.

Conclusion: Altogether, these data highlight the critical regulatory role of CFTR in bone development and maintenance, and suggest that some bone defects in people with cystic fibrosis are likely primary.

In CF, infection and inflammation often develop in parallel to CF disease complicating the study of CF pathogenesis. In contrast, in the CF gallbladder, disease develops (microgallbladder, epithelial mucinous changes, and lumen obstruction) prior to the onset of inflammatory changes and in the absence of infection. Interestingly, the gallbladder also has one of the highest expressions of CFTR relative to other tissues. Since the gallbladder is rich in CFTR and is defective at birth, we thought it would be an important model to study the role of CFTR in fluid and mucus transport.

Goals: Our goal was to use a gallbladder epithelial organoid model to study the role of CFTR in gallbladder epithelial function.

Methods: CF pig gallbladders were used because their phenotype closely resembles human CF gallbladder disease. Gallbladder tissue was obtained from non-CF and CF piglets, the epithelium was harvested, and cells were suspended in Matrigel supplemented with media. Organoids formed within 48 hours.

Results: Gallbladder epithelial cells from both non-CF and CF pigs developed into organoids. Gallbladder epithelial organoids expressed typical biliary surface proteins, including CFTR on the outer apical membrane of the organoid. Under baseline conditions, without stimulation, CF organoids were smaller and had reduced lumen area relative to non-CF organoids. In response to secretin stimulation, non-CF organoids decreased in both size and lumen area, consistent with apical membrane localization of CFTR on the outside of the organoid. However, this secretin-induced shrinking response was absent in CF organoids. Inhibition of NHE3, a sodium/proton exchanger, caused both non-CF and CF organoids to decrease in size. Finally, adenovirus transduction of CFTR into CF organoids increased the lumen size/total organoid size ratio relative to controls.

Conclusions: Under our culture conditions, porcine gallbladder epithelial organoids display an inside-out orientation. Loss of CFTR function is associated with morphological defects in organoid and lumen size. The size defect was partially reversed with CFTR expression. CFTR appears to play a role in unstimulated fluid absorption in gallbladder epithelia. Finally, gallbladder organoids may represent a novel approach to investigating the pathogenesis of CF.

Acknowledgments: Supported by NIH and Cystic Fibrosis Foundation. Cystic fibrosis (CF) is an autosomal recessive disease caused by mutations in the cystic fibrosis transmembrane conductance regulator gene (CFTR). Not only are multiple organs affected in CF, but the phenotype is difficult to explain solely on the basis of altered anion conductance. A plausible and likely cause of the diverse CF manifestations in multiple tissues is transcriptional dysregulation as a response of CFTR's absence and that genes responding to this dysregulation contribute to disease manifestations. We propose that the transcriptional dysregulation is a consequence of homeostatic mechanisms of the cell that by mobilizing transcriptional machinery to adapt to CFTR's absence, are also influencing other genes in the process, even in tissues where CFTR is not normally expressed, but with whom they share regulatory processes. The CF mouse model allows us to assess this theory without extraneous variation due to genetic heterogeneity. As a model, we used the recently created G542X CF mouse that carries a single nucleotide change in exon 12 of the endogenous mouse Cftr gene and their wild-type (WT) littermates, all male, 7-8 weeks old (C57BL/6J). G542X is one of the five most common CFTR mutations in CF patients. We used RNA from nasal epithelium (3/genotype), whole ileum (4 WT, 5 CF), epididymal fat (5 WT, 4 CF), and PMN (isolated freshly from bone marrow, 3/genotype). RNA-seq libraries were generated using TruSeq stranded Total RNA Ribo-Zero mouse gold kit (Illumina, San Diego, CA), and sequencing carried out using an Illumina HiSeq 2500 Rapid Run flowcell -2x100 bp run. Reads were aligned to the UCSC mouse genome mm10 using Bowtie2 and Tophat. Differential expression determined with Cufflinks/Cuffdiff. Pathway analysis done with OmicsOffice tools for functional classification (TIBCO Spotfire, Palo Alto, CA) and Gene Set Enrichment Analysis (GSEA, Broad Institute). The Table shows number of genes per tissue that have FPKM expression ≥ 1, how many genes are differentially expressed and the relative abundance of Cftr. The higher number of genes different in ileum is likely due to different cell types present in the whole tissue. Ongoing analyses include comparisons with other Cftr genotypes. Skeletal muscle and liver samples are next in the pipeline, as well as nasal and tracheal epithelial cells grown in air-liquid interface (ALI). The last two will allow us not only to compare expression between trachea and nasal epithelium, but also to compare expression between freshly isolated nasal epithelium with epithelium grown in culture: both of which will allow us to better understand the human data, since normally we only have access to human nasal epithelium grown in culture. As we add data from all tissues collected, we will be able to identify transcriptional networks altered in CF and gain better tools to understand potential sites of action. Supported by grants from CFF. FPKM: fragments per kilobase of exon per million fragments mapped 1 1. Inflammation, Amgen, Thousand Oaks, CA, USA; 2. GAU, Amgen, Thousand Oaks, CA, USA; 3. CBSS, Amgen, Thousand Oaks, CA, USA; 4. RIS, Amgen, Thousand Oaks, CA, USA Many potential gene transfer agents are being explored for gene therapy. Cystic fibrosis is a monogenic disease where the chloride channel defect in airway epithelium of the lungs is the primary cause of morbidity and mortality and is amenable to topically applied gene therapy. Previous attempts to correct gene expression in the lungs with lipid-mediated vectors, adenovirus and adeno-associated viruses (AAV) vectors have met with limited success. Herein, we characterize an HIV lentiviral (LV) vector pseudotyped with the ebolavirus (EboZ) envelope protein that expresses the luciferase (FLuc) or green fluorescent protein (eGFP) reporter gene and compare it to AAV2 and VSV-G pseudotyped LV benchmark vectors. Post-nebulization viral vector viability of viral vectors after nebulization by Aeroneb vibrating mesh was 68 ± 6%, 70 ± 11%, and 70 ± 7% of the pre-nebulizer titer for AAV2, VSV-G, and EboZ viral vectors, respectively. eGFP gene transfer to polarized primary air-liquid interface (ALI) cultures of human airway epithelial cells was 22 ± 8% AAV2, 19 ± 12% VSV-G and 34 ± 10% for EboZ viral vectors. Immunofluorescent characterization of ALI culture cells transduced by EboZ pseudotyped lentiviral vector expressing eGFP revealed that EboZ vector transduced goblet, club, and ciliated cell types. Bioluminescent imaging of mice dosed with nebulized or intranasally administered viral vectors expressing FLuc demonstrated that luciferase-mediated photon emission was detectable in the lungs of all treated mice. Immunohistochemical eGFP staining of mouse lungs transduced by eGFP expressing viral vectors revealed that alveolar macrophages were transduced by all the viral vectors and only the AAV2 vector transduced airway epithelial cells. This is likely a consequence of the higher dose of AAV2 (2.5 x 10 12 TU/mouse) delivered compared to the dose for the LVs (8.3 x 10 7 TU/mouse). This work demonstrates for the first time that LV vectors delivered to mice by nebulization are capable of transducing alveolar macrophages in the lower airways. Higher epithelial cell transduction efficiency or higher doses are likely needed to transduce airway epithelial cells.

Supported by Amgen. Nonsense mutations are present in 10% of patients with cystic fibrosis, produce a premature termination codon in CFTR mRNA causing early termination of translation, and lead to lack of CFTR function. There are no currently available animal models which contain a nonsense mutation in the endogenous Cftr locus that can be utilized to test nonsense mutation therapies. In this study, we create a CF mouse model carrying the G542X nonsense mutation in Cftr using CRISPR/Cas9 gene editing. The G542X mouse model has reduced Cftr mRNA levels, demonstrates absence of CFTR function, and displays characteristic manifestations of CF mice such as reduced growth and intestinal obstruction. Importantly, CFTR restoration is observed in G542X intestinal organoids treated with G418, an aminoglycoside with translational readthrough capabilities. The G542X mouse model provides an invaluable resource for the identification of potential therapies of CF nonsense mutations as well as the assessment of in vivo effectiveness of these potential therapies targeting nonsense mutations. Background: Respiratory syncytial virus (RSV) is a leading cause of upper and lower respiratory tract infections among infants, and these infections are highly associated with infant morbidity and mortality worldwide. RSV infection in early life may predispose children with cystic fibrosis (CF) to worse lung disease later in life. However, the reason for this association is not entirely clear. One hypothesis links neutrophil mediated inflammation as a key risk factor for bronchiectasis. The aim of this study was to investigate the effects of RSV infection on CF and non-CF lung epithelial cells.

Methods: F508del-CFTR patients (N=15) and healthy controls (N=13) underwent nasal brushing. Epithelial cells were expanded and seeded on Transwell inserts with air-liquid interface culture (ALI). After 21 days in ALI, cells were infected with RSV. Cells were washed apically with PBS and the basolateral medium was collected at 2, 24, 48 and 72 hours post-RSV infection (hpi). Plaque-forming assays were performed on PBS washings at 2, 24, 48 and 72 hpi and IL-8 was measured in media at 72 hpi. Transepithelial resistance (TER) was measured and cells were either fixed for immunofluorescence or RNA was collected for qRT-PCR at 72 hpi. Lung function data for CF patients were obtained from Toronto CF clinic patient database.

Results: RSV significantly decreased TER at 24, 48 and 72 hpi compared to PBS mock infection in both CF and control cells (all P<0.05). There was no difference in TER between CF and controls post-RSV infection. Immunofluorescence staining suggested disruption of cilia (β-tubulin) and epithelial cell junction markers ZO-1 and E-cadherin at 72 hpi. Plaque assays and qRT-PCR demonstrated higher RSV viral titers in CF cells than in controls at 72 hpi (all P<0.05). IL-8 was found at higher levels in CF as compared to control cells 72 hours post-RSV infection (P=0.023). There was no correlation between RSV titer and IL-8 concentration. However, the levels of IL-8 in the media were negatively correlated to patients' average FEV1 (forced expiratory volume in one second) (FEV1) (P=0.018).

Conclusions: Our data demonstrate that RSV decreases TER in both CF and control epithelial cells and may result from the disruption of epithelial cell junctions. In CF cells, the degree of IL-8 production post-RSV infection is linked to lung function of the CF patients from whom the CF cells were obtained. This suggests the possibility that innate lung epithelial cell immune responses, and specifically IL-8 production, may lead to a more pronounced inflammatory response that could affect lung function in CF patients. CFTR W1282X encodes a premature stop codon after the first third of NBD2 that foreshortens and destabilizes CFTR protein, blunts half-life of the transcript, and omits the carboxy-terminal PDZ anchoring domain, all of which diminish abundance of functional CFTR at the cell surface. Truncated W1282X CFTR exhibits residual activity with a significant CFTR gating defect potentiated by curcumin. G418, an established PTC readthrough agent, has negligible effect on W1282X under conventional in vitro conditions, although prolonged administration at high drug concentrations leads to biochemical evidence of full length CFTR. We also show that G418 unexpectedly stabilizes both wild-type and W1282X mRNA, and increases steady state levels of truncated protein, all of which improve surface localization of wild-type and W1282X CFTR. To further investigate W1282X molecular phenotype, we developed protocols using FRT cell lines encoding wild-type or W1282X CFTR with or without mini-introns surrounding exon 23 (the location of W1282X), and show that W1282X CFTR EMG (expression minigene encoding intronic DNA) produces low levels of protein and mRNA compared to the corresponding cDNA without introns. Augmentation of W1282X CFTR surface protein following treatment with either VX-809 or G418 was found to be much stronger (in proportion) in cells expressing W1282X with mini-introns. Moreover, short circuit current measurements indicate W1282X monolayers chronically exposed to VX-809 or G418 are preferentially enhanced in the presence of intronic DNA. W1282X EMG expression is further increased by the nonsense-mediated decay inhibitor 14 (NMDi14), with stronger activation in the EMG context following co-administration of G418. These findings demonstrate importance of intronic DNA for model systems intended to study W1282X or other CFTR premature truncation variants. Our experiments also emphasize the unanticipated effect of G418 on CFTR mRNA and truncated CFTR protein levels, independent of stop codon suppression, as well as increased full length (wild-type) CFTR following G418 in the FRT model system. We developed a feeder cell-free and serum-free cell culture method (EpiX™), which utilizes small molecules that inhibit TGF-β signaling pathway and modulate cytoskeleton dynamics, to achieve more than one trillion-fold expansion of epithelial stem and progenitor cells from skin, airway, mammary, prostate and other epithelia in standard 2D cell culture. EpiX™ medium enables quick expansion of 50-100 millions nasal epithelial cells from CF patients within 3-4 weeks for ex vivo theratyping to identify drug(s) that could benefit individual patients. The approach may be translated to other disorders where theratyping would be beneficial.

Notable attributes of the EpiX™ technology include: 1. Serum-free and feeder-free formulations allow over one trillion-fold expansion of primary human airway epithelial cells in vitro, while preserving genomic integrity.

2. Airway epithelial cells expanded using EpiX™ technology retain their lineage-committed differentiation capacity, and can seamlessly integrate into standard airway differentiation protocols.

3. EpiX™-expanded airway epithelial cells retain consistent CFTR and ENaC activities even after extensive propagation for at least 8 passages, allowing the generation of millions of cell models for airway epithelial function evaluation.

4. EpiX™ enables the derivation of hundreds of millions of nasal airway epithelial cells in a short time from a small number of primary cells such as nasal brushing samples in 3-4 weeks.

5. Extended proliferation runway empowers genetic engineering (e.g., CRISPR/Cas) to develop high-throughput screening models.

6. Second-generation chemically-defined animal-origin-free formulations satisfy upcoming regulatory guidance. Gene-editing technologies, such as TALENs and CRISPR/Cas9, are changing the way we approach research and medicine, allowing us to introduce any modification into specific genes for disease modelling, or precisely correct existing mutations in patient cells for personalized medicine. In addition, recently refined differentiation protocols of human pluripotent stem cells (hPSCs) into numerous cell lineages highlight hPSCs as an unlimited resource to model all affected tissues in multiorgan diseases. Thus, gene editing of hPSCs would enable development of valuable mutation-customized, fully differentiated, live cultures for CF research and a very powerful platform for tissue-specific CF drug screening. However, gene editing is still very challenging to perform in hPSCs, so we aimed to establish a robust protocol for gene editing hPSCs.

We optimized a stepwise process to minimize time/costs required to correctly edit hPSC lines by: A) Finding an efficient method to transfect designer nucleases, a single-strand oligonucleotide harbouring desired DNA changes, and an additional plasmid to enrich for transfected hPSCs by transient puromycin selection. B) Screening cell populations for nuclease activity by the T7EI assay and C) for desired modifications by a developed sensitive and discriminative screening method, allele-specific PCR (AS-PCR). D) Manually isolating clones from positive populations and rapid clonal AS-PCR screening to identify correctly edited clones. We developed this protocol using a wild-type (WT) human embryonic stem cell (hESC) line, CA1, to produce a homozygous ΔF508 CF-hESC line using CFTR-exon11 targeting TALENs. We also applied this protocol to correct 3 CF iPSC lines (8K, 4D, Em) homozygous for a nonsense mutation, W1282X, using CFTR-exon23 targeting CRISPR/Cas9. All hPSC populations were efficiently nucleofected and showed 16-37% cleavage at the correct sites by the T7EI assay, indicating TALENs or CRISPR activity in hESC and all iPSC lines, respectively. AS-PCR screening identified ΔF508 integration or W1282X correction in hESC and iPSC populations, respectively. 604 hPSC clones were manually isolated and the 342 surviving clones (56%) were rapidly screened by AS-PCR. Only the 33 clones suggested positive (9% of AS-PCR screened) were sequenced and 14 clones (4% of AS-PCR screened) were indeed correctly edited, including 1 homozygous ΔF508 CF-hESC pure clone, 1 heterozygous corrected 4D, 1 homozygous corrected 8K, and 1 homozygous and 1 heterozygous corrected Em iPSC pure clones. Thus, our optimised protocol can be applied to hESCs or iPSCs, using TALENs or CRISPR, to introduce mutations for CF disease modelling (ΔF508) or correct CF-mutations (W1282X) for personalized medicine, respectively, in a 3-6 weeks period. These cell lines are being characterized for pluripotency, chromosomal stability and ability to differentiate into a specific cell lineage.

In summary, we have developed a robust protocol for gene editing hPSCs that should facilitate the production of new "in-house" CF and corrected hPSC lines, expanding the range of tissue-specific models for CF research, and improve CF personalized medicine.

Supported by CF Trust-SRC003, CFIT (SickKids and CF-Canada Foundation), Emily Foundation. Currently, research studies lack the capacity to efficiently model the in vivo environment and functionality of tissues/organs in response to infection, disease, and drugs. Specifically, 2D culture models do not fully recapitulate the complexity and functionality of in vivo human studies, while animal models differ significantly with regards to physiology and metabolics. To address this problem, we have developed a multicell-type, layered 3D airway organoid that can be integrated into Transwell® or microfluidic culture devices at air-to-liquid interface (ALI) for modeling airways disease and infection. This organoid model is more physiologically relevant to the human pulmonary environment, airways disease and therapy evaluation.

Materials and Methods: The multicell-type, layered 3D airway organoids were constructed by first seeding a monolayer microvasculature endothelial cells onto the bottom surface of a porus lung-derived extracellular matrix (ECM) membrane support. Second, lung fibroblasts are seeded onto the upper surface of the membrane within a lung-derived ECM hydrogel with biomechanical properties matching native airway tissue. Finally, airway bronchial epithelial cells were seeded as a monolayer above the hydrogel. Following maintenance in culture for several days, organoids can be exposed to ALI and maintained in standard culture conditions or within a microfluidic system for up to 28 days. Organoid characteristics such as ciliated cell numbers and mucus production can be quantified via antibody staining, while transepithelial electrical resistance (TEER) and short circuit current (Isc) measurements can be performed using an EVOM2 and Ussing chamber. The TEER measurements can be performed daily to evaluate epithelial integrity along with pharmacological blocking and rescue of the CFTR protein. Pseudomonas aeruginosa infection, attachment and biofilm formation can be evaluated using both invasive quantification and noninvasive 3D imaging techniques.

Results and Discussion: At ALI culture, airway organoids are viable and functional for at least 28 days after their fabrication. Microfluidic cultures could be maintained under physiological flow and pressures, with the capability of aerosol delivery to the epithelial surface. Staining of organoids showed an increased presence of mucus, cilia, and other key components of the lung airway often lost during 2D culture. TEER measurements demonstrated the physiological barrier function of the primary epithelial cells with significant increases in TEER values when the cultures were maintained in ALI culture. Ussing chamber analysis demonstrated physiological CFTR responses to inhibition and pharmaceutical rescue. Pseudomonas aeruginosa showed increased levels of attachment and growth on organoids, and enabled identification of more virulent clinical strains. Biofilm formation was also increased in organoids compared to 2D cultures.

Conclusions: Multicell-type, layered 3D airway organoids accurately mimic the biological environment of the human upper airway compared to current 2D cultures, allowing for improved physiology, drug responses and disease modeling. Objective: CFTR modulator therapy has revolutionized care for patients with CF, bringing the underlying protein defect into clinical focus. It is anticipated that highly effective modulators will become available for the majority of CF patients. Nonetheless, for a portion of the CF population harboring rare mutations these drugs remain inaccessible. These subjects may require alternative approaches for drug testing, including individualized, precision medicine.

We developed a CF theratyping pipeline focused on highfidelity predictive models and when possible, robust clinical response characterization. Potential subjects with ≥1 rare/unidentified mutation are referred by CF caregivers. Baseline clinical data are recorded and primary human nasal epithelial (HNE) cells are collected. HNEs are expanded into planar and 3-dimensional models for drug testing, with multiple quality assurance measures (electrophysiologic characteristics, spheroid number/ morphology, internal consistency, immunofluroescence). All results are summarized and provided to patients and clinicians. We have recently enrolled local subjects to undergo trials of the tested agent(s), tracking multiple clinical measures, including sweat chloride, spirometry, BMI, patient-reported outcomes, and advanced MR-imaging.

Results: Over the last year, 60 subjects from 17 centers have been enrolled, including 25 locally. HNE testing has been completed for 45 subjects, complicated by contamination rates of 16% for local and 20% for referred samples. Of those completing testing, 10 (22.2%) demonstrated a statistically significant improvement in CFTR function with ≥1 modulator (Table; compared against wild-type (wt) CFTR Isc = 21 µA/cm 2 , n=13 wtCFTR subjects; *p<0.05 vs Baseline).

Four local subjects underwent HNE testing and clinical characterization on drug, with 3 more in process. Of those, 1 subject demonstrated drug bioactivity in vitro (statistically significant change in CFTR function) and also demonstrated clinical benefit on drug (improvement in >2 clinical outcomes). The remaining 3 subjects demonstrated no drug bioactivity in vitro, and none derived clinical benefit from the same compounds.

Conclusions: This multifaceted pipeline is designed to maximize the ability to detect benefit from CFTR modulation. Use of 2 preclinical model systems provides internal validation, and measurement of multiple clinical outcomes (when available) fully characterizes subjects' drug response. This comprehensive, rigorous focus on both in vitro and in vivo performance allows for continuous feedback, assessment, and refinement of preclinical predictive models. Additional efforts are ongoing to reduce contamination of samples, increase throughput, and validate preclinical models vs in vitro responses.

It is expected that the availability of patient-specific induced pluripotent stem cells (iPSCs) in conjunction with precise genome-engineering approaches will revolutionize human disease modelling and drug screening. Organotypic culture systems based on disease-specific iPSCs exhibit obvious advantages compared to immortalized cell lines and primary cell cultures, however, their implementation in high-throughput (HT) assays is technically challenging and as yet there are very few examples of successful conduct of iPSC-based HT drug screens. Here we demonstrate the development of an organotypic Cl -/Iexchange assay based on cystic fibrosis (CF) disease-specific iPSCs for the application of HT drug screening. CF-iPSCs (homozygous for p.Phe508del) were generated via reprogramming of CD34 pos cells isolated from small volumes of nonmobilized peripheral blood. The resulting CF-iPSCs were analysed regarding their karyotype, pluripotency status and differentiation potential. TALEN-based genome engineering was applied for "footprintless" correction of the p.Phe508del mutation, and targeted introduction of a fluorescence reporter into the cystic fibrosis transmembrane conductance regulator (CFTR) locus facilitated the directed differentiation into intestinal epithelia. Moreover the introduction of a halide-sensitive YFP into the AAVS1 safe harbour locus enabled automated quantitative measurement of CFTR function. Reporter iPSC-derived epithelium revealed a CFTR-channel specific response after forskolin application, which was inhibited after CFTR(inh)-172 treatment. CFTR function in CF iPSC-derived epithelia was partially rescued by treatment with VX-770 and VX-809, and seamless gene correction of the p.Phe508del mutation resulted in full restoration of CFTR function. A functional screen of ~ 40.000 small molecule compounds for modulators of p.Phe508del CFTR on a 384-well scale was conducted. Primary hits are currently evaluated and verified candidate compounds will further be evaluated in secondary assays.

As one of the first functional HT screens utilizing disease-specific iPSC derivatives, our approach will also contribute to clarify whether the general hypothesis holds true that organotypic assays based on patient-specific iPSCs are better predictors of clinical efficacy than conventional cellular assays that typically apply immortalized cell lines.

This work was supported by -E-Rare (INSTINCT) and The German Center for Lung Research (DZL). Introduction: Human induced pluripotent stem cells (hiPSC) can be used to develop organotypic cell culture models to study the pathophysiology of rare and common variants of cystic fibrosis, and the efficacy of experimental therapeutics. In contrast to committed stem/progenitor cells, hiPSC can be clonally modified by gene editing using TALEN and CRISPR/Cas9, and differentiated to multiple epithelial lineages. Moreover, this allows the generation of CFTR corrected lines from CF patients, which provides a valid wild-type control in comparative analysis, and future potential for cell replacement therapy (Merkert S, et al. Stem Cell Res. 2017; 23:95-7) . Developing robust differentiation protocols that allow large scale air-liquid interface (ALI) culture of hiPSC derived airway cells is still challenging.

Approach: We have generated F508del CFTR, corrected, and N1303K CFTR hiPSC, a folding mutant prominent in the Mediterranean population, for which no effective correctors are available. Fluorescent markers monitoring epithelial differentiation: NKX2.1 (NKX2.1eGFP), the airway basal cell marker p63 (p63mVenusnuc NeoR), and CFTR (CFTR-dTomato) were introduced by TALENs in the respective gene loci. In F508del CFTR lines used for HT screening a YFP halide sensitive probe was introduced.

Methods: hiPSC were subjected to a 20-day differentiation and selection protocol, which reproducibly yielded a population of NKX2.1-GFP/p63-Venusnuc neoR expressing cells (~85%). After expansion of p63+ hiPSC in media used to differentiate primary airway basal cells in ALI culture (BEGM, Pneumacult), cells were transferred to standard ECM coated membrane inserts (Costar). At different intervals, cells were analyzed by QPCR, electrophysiology, confocal and electron microscopy.

Results: Depending on the ECM coating, p63+ iPSC on inserts reached variable TEER values within five days before transfer to ALI (500-2500 [BS1] Ohm/cm 2 .). QPCR analysis [BS2] [ORD3] of ALI cultured p63+ hiPSC revealed expression of general airway markers (NKX2.1, SOX2, FOXA2, SOX9, NGFR, CK5), ciliated cells (FOXJ1, CFTR, SCNN1A/ ENACa) and secretory cell markers (CCSP, MUC5AC). Microscopy showed clusters of high cuboid epithelial cells expressing lateral (ECAD) and tight junctions (ZO-1), and airway markers (KRT5, KRT8, KRT14, FOXJ1, P63). Multi-ciliated (TUBIV+) and secretory cells (CCSP, MUC5AC) were observed, confirmed by electron microscopy.

Conclusions: We report efficient generation of hiPSC derived basal airway progenitor cells using a selectable p63 marker, which can be used to generate airway epithelia at air-liquid interface. hiPSC p63+ derived F508del lines are currently tested for CFTR activity and correction, and will be used to validate hits from a current high-throughput screen on iPSC derived CFTR expressing cholanchiocytes.

Bone fragility and low bone mineral density often affect children and young adults with cystic fibrosis (CF) disease and is associated with significant morbidity due to vertebral fractures and decreased lung function (Jacquot J, et al. Osteoporos Int. 2016; 27:1401-12) . We recently reported that bone demineralization is improved by the CFTR potentiator ivacaftor in young patients carrying the G551D CFTR mutation (Sermet-Gaudelus I, et al. J Cyst Fibros. 2016; 15:e67-9) . Due to the presence of CFTR in monocytes, we hypothesized that ivacaftor may impact monocyte differentiation and activation. Monocyte osteoclast precursors fuse and differentiate to form bone-resorbing multinuclear osteoclasts upon stimulation by two essential factors, the M-CSF and the RANKL.

We analyzed RANK and M-CSFR receptors on peripheral blood monocytes by flow cytometry and clinical data in 11 G551D-bearing CF patients (5 females, 6 males, median age 27, range 18-39 years) and evaluated potential changes following 9-and 12-months ivacaftor treatment.

Ivacaftor improves sweat chloride and lung function but does not change the number of blood monocytes. The average percentage of double RANK + /M-CSFR + monocytes (91.6 ± 1.5%) in G551D patients was strongly higher (P=0.00017) compared with 18 healthy individuals (2.7 ± 0.8%, 4 females, 14 males, median age 37, range 20-65 years). Interestingly, this percentage decreased after 9-and 12 months ivacaftor treatment (n=6 and 4 patients, respectively), a reduction which was more pronounced in two patients. Ten out of the 11 patients improved or stabilized their bone mineral density (BMD) by ivacaftor (an improvement > 0.2 of BMD z-score). Flow cytometry data also demonstrate higher expression of both M-CSFR and RANK receptors, reported by unit of monocytic cells in G551D patients compared to 18 healthy donors.

Moreover, we examined ex vivo differentiation and activation of healthy monocytes into osteoclasts for a 21-day period with/or without the addition of inh-172, an inhibitor of CFTR chloride channel activity. Interestingly, multinuclear osteoclasts derived from inh172-treated healthy monocytes were larger, more adherent, and were prone to generate large pits and trenches of dentin resorption. In addition, multinuclear osteoclast-derived inh172-treated healthy monocytes released a very low level of bioactive lipid mediator sphingosine 1-phosphate (S1P), a key mediator in the directed migration of osteoblast/osteoclast precursors attached to the bone surface.

Altogether, these data highlight the critical regulatory role of CFTR in M-CSFR and RANK receptors expression in monocytes, and suggest CF bone disease as a new, cell-type-monocyte dysfunction disease, providing new insights into the pathogenesis of CF bone disease. Introduction: Mice homozygous for the DF508 mutation (the most common disease-associated mutation) exhibit severe intestinal disease and require constant laxative treatment for survival. This pathology mimics the intestinal obstruction (meconium ileus) seen in some CF patients. Genistein is a naturally occurring isoflavone found in soy.

Objective: We aimed to determine whether dietary supplementation with genistein promotes overall growth of the ΔF508 CF mouse, via influences on gluconeogenesis. Clinically, weight gain is associated with improved lung function, thus a means to increase growth and thriving capability is relevant.

Methods: At age 21 days, we maintained male and female ΔF508 mice on three diet regimens for 45 days post weaning; (1) normal diet, (2) normal diet + Colyte and (3) 600 mg genistein/kg diet, 600G. Survival rates and body weight data were determined. At the completion of the diet study, liver and serum were collected and immediately frozen at -80°C until use. Liver tissue was evaluated for total protein expression and histology, and serum used for several ELISA assays.

Results: Body weight at the end of the diet study was significantly greater in males fed genistein, 600G (21.96±0.68 g, n=14) versus those males on Colyte (18.86±0.64 g, n=8). There was no change in overall weight gain in the female groups. Total hepatic GLUT4 protein expression was unchanged in the female groups, yet significantly decreased by 55% with genistein diet in males (compared to Colyte). Serum glucose levels were significantly increased 1.2-fold with genistein diet in males. We are currently exploring whether there are changes in expression of key proteins involved in the gluconeogenesis pathway (for example: glucocorticoid receptor, PEPCK1, 11BHSD-1). Those proteins are responsible for glucose regulation within the liver, and will provide a better illumination of the mechanism of action of 600G to improve weight in males, yet not in females.

Conclusion: We conclude that feeding ΔF508 male mice 600G genistein-diet promoted an increase in serum glucose levels while decreasing hepatic GLUT4 protein expression. There was a sex-dependent effect of 600G genistein diet on weight gain; males fed genistein were 3.3 g heavier, versus those on Colyte (with no change in weight for females). These studies could have implications to provide pharmacological targets to promote growth and ability to thrive in CF male patients. Acknowledgment: Supported by MWU intramural funds (LA). Introduction: Cystic fibrosis patients have a high degree of variability in disease severity, progression and survival. Thus it may not be unexpected that clinical response to CFTR modulator therapy demonstrates individual patient variability. In vitro assays based on experiments performed in patient derived tissues are increasingly being used to identify novel CFTR modulators and predict which modulators may offer benefit for individual patients. Among these patient derived tissues, nasal and bronchial epithelial cells have been identified as they are relatively easy to obtain and may model the CF lung response. However, these cells have a finite lifespan. Induced pluripotent stem cells (iPSCs) are pluripotent stem cells generated from mature somatic cells. These cells contain the genetic information of the individuals from whom the cells are derived and importantly can be passaged indefinitely. Current protocols do exist to transform iPSCs to an epithelial phenotype. However, these protocols are quite complicated and time consuming. A simple method to generate a cell with CFTR function from an iPSC would offer a new model to test CFTR modulators.

Hypothesis: We hypothesized that iPSCs can differentiate into ciliated epithelial cells containing functional CFTR directly via a club cell intermediate and bypassing an endoderm intermediate. We sought to examine the use of air-liquid interface culture (ALI), substrate composition and expansion media as means to differentiate iPSCs.

Methods/Results: iPSCs were initially seeded on inserts coated with collagen type I, which has been previously shown to induce club cell differentiation. However, due to the low attachment rate of iPSCs on collagen type I coated inserts, we explored the options of a combination of a variety of matrices and culturing methods. Various cell densities were also tested to maximize cell adhesion. Our results showed that type I collagen + matrigel significantly increases the iPSC attachment. Furthermore, a seeding density of 600,000 cells/cm 2 and a semidry apical surface allow the cells to maintain a monolayer for later induction process. Pneumacult Ex, a media used for basal cell expansion, was used as our base induction media. However, our preliminary data suggest that Ex media did not induce measurable club cell phenotype over a period of 10 days.

Conclusion: Our initial work suggests that type I collagen + matrigel as a substrate for ALI significantly improves cell adhesion on ALI inserts for later differentiation process. However, Ex media may not be sufficient in inducing club cell differentiation. Furthermore, the use of matrigel, a common matrix for maintaining PSC pluripotency may also hinder this process. Ongoing work will focus on supplementing Ex media with other differentiation factors such as IL-6, which has been shown to promote ciliary cell differentiation, to induce this process. Different matrices such as the laminin series in substitute of matrigel will also be examined. In addition, other intermediary markers will be used to more closely monitor this process. Traditional feeder-free and bovine pituitary extract (BPE)-containing media formulations for the expansion of primary human bronchial epithelial cells (HBECs) typically support the maintenance of mucociliary differentiation potential for a limited number of passages in culture. A novel culture system comprising an inactivated mouse embryonic fibroblast feeder layer in combination with a specialized medium has been reported to improve the expansion of HBECs while still maintaining their differentiation potential, even after extended passaging (Liu X, et al. Am J Pathol. 2012; 180:599-607; Suprynowicz F, et al. PNAS. 2012; 109(49):20035-40) . However this feeder-dependent method is cumbersome and undefined, thus limiting its applications. We have developed an improved medium, PneumaCult-Ex Plus, that is feeder-and BPE-free, and promotes extended passaging of HBECs without the loss of their differentiation potential at later passages. The medium allows for the rapid expansion of HBECs, while maintaining their ability to form a pseudostratified mucociliary epithelium at an air-liquid interface (ALI) using PneumaCult™-ALI for more passages compared to other commercial HBEC expansion media. To measure cell expansion over multiple passages, commercially available HBECs (Passage 1) were thawed and seeded into T-25cm 2 flasks containing either control media (PneumaCult™-Ex or a BPE-containing medium such as BEGMTM) or PneumaCult-Ex Plus at a density of 1 x 10 4 cells/cm 2 . For each culture period, media were fully replenished three times per week and cells were enzymatically dissociated and passaged once cultures reached approximately 80% confluence. ALI differentiation was performed with PneumaCult™-ALI for HBECs expanded in each of the three test media at different passages. The differentiated cells were functionally characterized by immunocytochemistry, qPCR and transepithelial electrical resistance (TEER). The average fold expansion over 8 passages was significantly higher for the cells cultured in PneumaCult-Ex Plus (11.1 ± 2.4 mean ± SD, n=4) compared to PneumaCult™-Ex (4.1 ± 1.9, n=4) or to BEGMTM (3.7 ± 1.6, n=4) . HBECs cultured in either PneumaCult™-Ex or BEGMTM could be differentiated into functional pseudostratified mucociliary epithelium only at early passages (P1 and P2). Conversely, cells expanded in PneumaCult-Ex Plus could be successfully differentiated at each passage for at least 4 passages to generate, by 28 days post-air-lift, a functional pseudostratified mucociliary epithelium containing MUC5AC (7.2 ± 2.4%, n=4) expressing goblet cells and AC-tubulin (30.4 ± 5.4%, n=4) expressing ciliated cells. Barrier function was measured weekly by TEER using EVOM. Fully differentiated cultures in PneumaCult™-ALI showed stable TEER (200 -600 Ω X cm 2 ) for at least 25 weeks (n=4). In summary, we have developed an improved defined BPE-free medium for primary human airway epithelial cell culture that promotes greater expansion of HBECs and maintenance of their mucociliary differentiation potential at later passages. 

Introduction: Next-generation CFTR modulators may significantly improve the health of F508del homozygous cystic fibrosis (CF) patients by correcting and potentiating CFTR function in epithelial tissues. A CFTR modulator triple therapy from Galapagos and AbbVie, including two correctors (C1 and C2a) and a potentiator (P2), have for instance shown promising repair of CFTR function in preclinical epithelial models. This includes models measuring chloride conductance in air-liquid interface cultured bronchial epithelial cells and forskolin-induced swelling in rectal organoids from CF patients. To further determine the efficacy of the CFTR modulator triple therapy and to validate our new in vitro models, we conducted a forskolin-induced swelling assay using nasal and bronchial organoid cultures from CF patients.

Methods: Nasal and bronchial organoids were developed from CF patients with a homozygous F508del mutation (n=2 independent donors) and a wild-type (WT) CFTR control (n=1 donor). CF nasal and bronchial organoid cultures were stimulated with C1 (0.15 µM) and C2a (1 µM) and incubated for 48 hours. Next, organoid swelling was measured after stimulation with forskolin (5 µM) and P2 (2 µM). The efficiency of the modulator therapy was calculated as percentage forskolin-induced swelling of WT control organoids.

Results: C1, C2a and P2 triple therapy increased forskolin-induced swelling in both nasal and bronchial organoids from CF patients. Moreover, we observed comparable responses in nasal and bronchial organoids, demonstrating a higher effect of the modulators on a certain patient. Compared to the triple therapy, C1 and P2 double therapy caused a less prominent increase in forskolin-induced swelling in CF bronchial organoids from 1 patient, while having no effect on the other patient. In addition, double therapy did not improve the swelling of CF nasal organoid cultures from both patients. These results corresponded with the effects of VX-809 and VX-770 in bronchial and nasal organoid cultures.

Conclusion: Altogether, we further confirm the efficacy of the Galapagos and AbbVie CFTR modulator triple therapy in a preclinical setting, demonstrating effectiveness in both nasal and bronchial organoids from F508del/F508del CF patients. Moreover, we demonstrate comparable patient variability in nasal and bronchial cultures. This suggests that forskolin-induced swelling in nasal and bronchial organoid cultures can be further applied as a read-out to determine the efficacy of CFTR modulators in patients with CF.

The cystic fibrosis (CF) rat is a promising animal model for therapy development because it exhibits many of the key features of CFincluding CF lung disease -and is more practical than other CF animal models. Since mid-2017 we have bred 2 lines of CF rat, carrying different CFTR mutations. One carries the Phe508del mutation (class II), predicting a CFTR protein folding impairment and increased CFTR protein degradation; the other a frameshift/stop codon mutation (class IA), predicting an impairment of CFTR protein synthesis, and is essentially a CF knockout (KO) animal. Rats with the 2 mutations are kept under identical clean conditions, and all founder animals were of the same genetic background, making the two models directly comparable. We have observed distinct differences in the severity of CF-related pathologies for these two CFTR mutations.

Methods: The Phe508del model was generated using CRISPR-Cas9 to delete the TTT codon corresponding to 508 in the human sequence. The KO rat was inadvertently created through off-target CRISPR-Cas9 mutations that resulted in one base substitution and two base deletions in the two codons directly upstream of the CRISPR-Cas9 target site; this results in a frameshift and the introduction of a stop codon at position 512 in the CFTR sequence. CF rat production is maintained through the breeding of mutation-matched CF-heterozygous pairs and all animals receive ColonLytely in drinking-water as a precaution against gastrointestinal obstruction.

Results: To date we have produced more than 50 CF rats for each genotype. We have observed increased neonatal mortality for KO rats compared to normal, while Phe508del rats show normal neonatal mortality. Growth curves show lower body weights for KO rats compared to Phe508del rats (to 9 months). Heterozygous male and CF female rat pairs produced offspring for the Phe508del genotype, but not for the KO genotype. Whilst studies are not complete, the nasal potential difference (NPD) measurements show the classic CF phenotype for the KO rats, but Phe508del CF rats show intermediate NPD between the normal and the KO levels. Both lines show increased levels of mucus in the upper and lower intestine, and a tendency of gut obstruction postweaning. Mucus plugging in the tracheal submucosal glands is present in both genotypes, but only the knockout CF rats have developed spontaneous lung disease with airway mucus plugging, as early as 3 months of age. Phe508del CF rats are yet to show evidence of CF lung disease at 6 months.

Conclusions: We report two different models of CF rat; a KO rat that develops spontaneous lung disease with airway mucus plugging at as early as 3 months of age, and a more robust Phe508del rat, which displays milder nonlung CF pathologies and no evidence of CF respiratory disease by 6 months. Analyses are ongoing, but we expect that the milder Phe508del rat model may develop lung disease at later ages, and will be useful for pathology and treatment studies where airway disease is induced, such as through the use of induced viral and bacterial infections. Rationale: MUC5AC and MUC5B are high-molecular-weight glycoconjugates responsible for the viscoelastic properties of mucus, produced in the lungs to prevent pathogen invasion. In cystic fibrosis (CF), ion fluxes and mucus concentrations are abnormal. As a result, the dehydrated mucus layer compresses the periciliary gel layer, collapsing the cilia underneath and impeding lung mucus clearance. It is fairly well accepted that MUC5AC and MUC5B play a critical role in airway mucus transportability and cough clearance; however, their individual role in health and disease has yet to be determined.

Objectives: To characterize the properties of MUC5B and MUC5AC in health and disease with respect to their multimeric organization, viscoelasticity, transportability and clearability from the lungs. Methods: We developed cell lines producing only MUC5ACm or MUC5Bm (m=mono-mucin). We utilized CRISPR-Cas9-mediated deletion to generate A549 cells producing only MUC5AC or MUC5B, as these cells normally produce both mucins at equal ratios.

Results from sequencing, RT-PCR, Western blots, immunohistochemistry, RNA in situ hybridization (ISH) and mass spectrometry confirmed single mucin production (i.e., MUC5ACm and MUC5Bm). Mucus from MUC5ACm and MUC5Bm cells were collected individually by apical cell washing and reconstituted at physiological concentrations (e.g., 2%-5% solids) for experimentation. The biochemical and biophysical properties of mixed and "mono"-mucin mucus were examined by a variety of assays (e.g., light scattering, rheology, mucociliary transport, cohesive strength and cough clearance assays).

Conclusions: Our data suggest that MUC5ACm exhibits higher viscosity and cohesive strength compared to MUC5Bm. Furthermore, MUC5ACm mucus transported less efficiently than MUC5Bm mucus on ciliated human primary airway cultures and therefore MUC5AC appears "stickier" than MUC5Bm. In contrast, cough clearance rate was significantly reduced with MUC5ACm compared to MUC5Bm mucus, suggesting a change in mucin multimer interactions in the presence of MUC5AC. Results from these studies help our understanding of mucus pathogenesis in CF and potentially in other chronic obstructive lung diseases. Center, University Medical Center Utrecht, Utrecht, Netherlands; 2. Pediatric Pulmonology, University Medical Center Utrecht, Utrecht, Netherlands Objective: Mechanisms contributing to patient variability in response to CFTR modulator treatment remain not completely understood. Here, we characterized in more detail transcription, translation and function of R117H-CFTR using patient-specific intestinal organoid cultures and correlated them with response to ivacaftor (VX-770) in vitro. This enables a more careful analysis of relations between CFTR expression and fluid secretion, and may point out that CFTR is the main determinant of patient variability in fluid secretion in in vitro grown organoids.

Methods: Organoids were generated from individuals that were homozygous (n=1) and compound heterozygous (n=13) for R117H-CFTR. We used the novel, automated steady-state lumen area (SLA) and forskolin-induced swelling (FIS) assays to measure (residual) CFTR function and response to treatment with VX-770. In parallel, CFTR protein and mRNA expression levels were determined and all measurements were individually correlated to assess relations between molecular and functional indicators of CFTR.

Results: Paired analysis of residual CFTR function by SLA and FIS in R117H-CFTR organoids demonstrated interpatient variation ranging from low-to-moderate SLA values and FIS responses. SLA values demonstrated a strong correlation with FIS responses at 0.128 µM forskolin stimulation (r = 0.92, p<0.0001), which was also observed upon VX-770 treatment (r = 0.66, p = 0.01). Variability in mRNA and protein expression between R117H-CFTR organoid cultures was detected and demonstrated significant correlation with SLA (r = 0.84 and r = 0.78, respectively) and FIS (r = 0.89 and r = 0.86, respectively) measurements. The study was not powered to analyze correlation with clinical phenotypes, but we observed a trend towards a weak correlation of FIS responses with sweat chloride concentration values (r = -0.47, p = 0.11).

Conclusions: Our results demonstrate that variability in residual R117H-CFTR function and response to VX-770 treatment as measured by CF organoid fluid secretion can be attributed to differences in CFTR mRNA and protein expression levels. The data indicate that differences in CFTR expression may contribute to clinical heterogeneity in R117H-CFTR patients. More than 1,700 genetic variants of the CFTR gene have been suggested to cause Cystic Fibrosis (CF). Independently of their mutation class, a large majority of these variants ultimately result in marginal plasma membrane expression of functionally impaired CFTR. CFTR function is dependent upon proper translocation of CFTR to the membrane. Several assays have been developed to monitor CFTR translocation including mammalian cells heterologously overexpressing CFTR-HRP, a recombinant CFTR cDNA with a horseradish peroxidase (HRP) reporter inserted into its extracellular loop 4 (ECL4). Such high-throughput screening compatible assays have already been successfully used to identify or validate small molecule correctors targeting missense or nonsense variants, e.g., F508del (Phuan PW, et al. Mol Pharmacol. 2014; 86(1) :42-51) or Y122X and W1282X (Liang F, et al. SLAS Technol. 2017; 22(3) :315-24). However, these artificial systems that are based on an overexpression of cDNA, exclude the native CFTR promoter and intronic sequences, and therefore, are poorly suited to study CFTR dysfunctions caused by mutations located in introns and/or affecting splicing, expression regulation, nonsense mediated mRNA decay (NMD), etc. To overcome these limitations, we aimed to create a more biologically relevant model, a cell line with the HRP reporter inserted into the genomic CFTR locus at the ECL4/exon-17 encoding region.

We designed a gRNA targeting ECL4/exon-17 sequence and produced a purified, large (~3600 nucleotides) single-stranded DNA homology directed repair (HDR) template, containing the HRP encoding sequence framed by 5' and 3' CFTR homologous arms. Previously, we have developed an optimized CRISPR/Cas9 gene editing platform to model CFTR variants in the genomic context of 16HBE14o-cells (Dieter Gruenert, UCSF). We adapted this platform for targeted large fragment insertion, which is less efficient and more difficult to screen than small mutations and identified a clonal cell line with an HRP reporter correctly inserted in the ECL4 loop of native CFTR. The expression of the fusion wild-type (wt) CFTR-HRP protein and its ability to properly report CFTR trafficking was confirmed by Western blot detection and cell-surface HRP activity, respectively. Treatments with approved corrector (VX-809) have been described to slightly increase translocation of wt CFTR. As expected, the 16HBEge wt CFTR-HRP cell line elicited an increase in HRP activity with VX-809 treatment, confirming a correlation between CFTR translocation and the HRP assay.

We have also introduced the pathogenic mutations F508del or G542X into the 16HBEge wt CFTR-HRP cell line. These two isogenic cell lines, 16HBEge CFTR-HRP with either the F508del or G542X variant, are currently being characterized and the suitability of these cell lines for highthroughput translocation assays will be evaluated.

Dey, I. 1 ; Meleshkevitch, E. 1 ; Shah, K. 2 ; Bridges, R.J. 1 ; Bradbury, N.A. 1 1. Physiology and Biophysics, Chicago Medical School, North Chicago, IL, USA; 2. NCI, National Institutes of Health, Bethesda, MD, USA LMTK2 is a transmembrane kinase known to bind to and phosphorylate CFTR. We determined if LMTK2 was able to regulate CFTR-dependent chloride and fluid transport using murine enteroids as a functional model. Enteroids were made from wild-type (wt) and LMTK2 -/-mice. Enteroids from LMTK2 -/-mice were able to form in a similar fashion to organoids from wt mice, however the number and complexity of developing crypts was lower in LMTK2 -/-mice compared to wt enteroids. This is likely due to a reduction in ISC, as the stem cell marker Lgr5 was lower in LMTK2 -/enteroids compared to wt. Functional assays were performed using microscopic evaluation of forskolin induced swelling. Following CFTR activation and fluid transport enteroids from wt mice showed a marked rapid swelling that eventually plateaued at 40 times the original lumenal volume, while LMTK2 knockout organoids showed a forskolin-induced increase of only 18-fold the orginal lumenal volume. Enteroids from LMTK2 -/mice were responsive to forskolin but showed a significantly lower rate of swelling and a markedly lower final plateau. Swelling was lost from both wt and LMTK2 -/-enteroids when the samples were incubated with CFTRinh172, indicating that the swelling was indeed due to CFTR. No significant difference in CFTR mRNA levels between wt and LMTK2 -/enteroids was observed, suggesting that the loss of CFTR-mediated chloride and fluid transport was due to altered subcellular trafficking of CFTR rather than a diminution in CFTR protein. These data identify LMTK2 as a viable therapeutic target for abnormal fluid transport in diseases such as cystic fibrosis and secretory diarrhea. Defects in the cystic fibrosis transmembrane conductance regulator (CFTR) of pancreatic ducts leads to abnormal pH regulation in the ductal lumen. The resulting premature activation of the digestive enzymes secreted into the ductal lumen by pancreatic acinar cells leads to tissue necrosis. Our previous studies on pancreatic pathology in CFTR-knockout (KO) ferrets indicate age-dependent changes in acinar tissue, inflammation, fibrosis, adipogenesis, and islet frequency (Am J Pathol. 2018;188(4):876-90). Additionally, islets harvested from CFTR-knockout (KO) newborn ferret pancreas demonstrate reduced islet insulin and lower glucose stimulated insulin secretion likely in part due to enhanced IL-6 secretion from islet-associated ductal cells (Endocrinology. 2017;158(10):3325-38). However, paracrine factors from the pancreatic ducts that influence insulin secretion and their mechanism of action is largely unknown. To this end, we have developed an air-liquid interface polarized primary ductal epithelial cell culture system from newborn wild-type (WT) and CFTR-KO ferrets. The basolateral and apical secretions from the polarized epithelia, upon stimulation with carbachol or IBMX/forskolin, were analyzed by mass spectrometry using "light" and "heavy" stable dimethyl isotope labeling for the quantitation. The fold-change (FC) of proteins enriched in "heavy" labelled CFTR-KO secretions versus "light" labelled WT secretions were determined by calculating the area under the curve of the "heavy" and "light" labeled isotopic peptides (in N=6 animals of each genotype). Initial gene ontology (GO)-term analysis on the CFTR-KO secretions indicated enrichment of immune response and inflammatory processes. Apical secretions had 19 and 28 differentially secreted proteins by CFTR-KO epithelia following carbachol or IBMX/forskolin, respectively. In basolateral secretions, 21 and 7 proteins were differentially secreted by CFTR-KO epithelia following carbachol or IBMX/forskolin, respectively. Paracrine factors, known to influence insulin sensitivity and insulin secretion included insulin growth factor binding protein 7 (IGFBP7; log FC -8.68), thioredoxin (TXN; log FC -2.30), dipeptidyl peptidase 4 (DPP4; log FC -11.35), and thrombospondin 1 (THSB1; log FC 13.25) were among those differentially secreted by the CFTR-KO ductal epithelium compared to WT (p < 0.05). This comparison of CF and non-CF ductal secretomes should prove useful in identifying paracrine factors that impact islet function in CF and hence will aid in understanding the underlying mechanisms leading to aberrant islet insulin secretion in CF. Introduction: CORVO ("Computing ORganoids VOlume") is a tool that processes 3D confocal acquisition to estimate not only the surface area (as many other available tools already do) but also the volume of organoids and their changing with time for an accurate quantification of the time evolution of the organoid structure in response to a treatment. We apply this tool to study the dynamic of human intestinal organoids.

We implemented a C++ computational pipeline for the segmentation and slice-by-slice organoids' image analysis and an R code for the statistical post-processing of the results. CORVO takes two inputs: a video and a TXT file with some parameters. The input video -in AVI format -contains the images of different slices of a set of organoids at different time, captured by a confocal microscope and codified in 5D format. The parameter file -in TXT format -contains the values of some parameters used by the algorithm: 1. the number of slices sampled by the microscope, 2. the minimum area (in pixels) of the organoids to be detected, 3. a real-valued threshold on the intensity values of the organoids to be detected (related to the fluorescence), 4. an integer number related to the "physical" separation (minimum distance) among the organoids, and 5. a flag to display these instructions (0 = no, 1 = yes). The program displays an interactive image, where the boundaries of the organoids are shown in red color. Organoids of interest are selected by the user with a mouseclick. Volume computation follows this step. For each selected organoid, CORVO outputs: 1. the time, 2. the area of the maximum slice, and 3. two estimations of the organoid volume, called the topological and the intensity based volumes of the organoid: the topological (intensity based, resp.) volume is the sum over the slices of the pixels (of the pixel intensity) of the region corresponding to the organoid.

Results: We analyzed wild-type organoids as a model and obtained a statistically significant linear correlation between volume and area of the largest slice, due to the almost regular spheroid shape typical of the wild-type organoid. On the opposite, deviations from linearity may indicate an anisotropic volume increase of the organoid.

Acquisition and analysis of CF organoids is currently in progress. Results will be presented alongside the 3D reconstruction of the dynamics of the organoids swelling.

Conclusion: The statistical analysis of the organoids' volumes computed with CORVO will define the relationship between area and volume computation and will permit evaluation of which variable between the two (area or volume) may be more appropriate for quantification of swelling.

Acknowledgments Introduction: Epithelial tissues have methods to protect the host from environmental pathogens, but, such defenses are dysfunctional in CF patients due to impaired CFTR chloride channel function, epithelial differentiation and barrier function (Cutting GR. Nat Rev Genet. 2015; 16(1):45-56.) . Previous studies have shown that transepithelial electrical resistance (TER) is reduced in CF bronchial epithelial cells over-expressing F508del-CFTR compared to wild-type CFTR. This has been attributed to altered PDZ-motif mediated protein interactions between CFTR and junctional proteins in CF cultures (LeSimple P, et al. J Physiol. 2010; 588:1195-209) . However, such observations have only been investigated in overexpression systems. With the growing acceptance of patient-derived nasal cultures as a surrogate for lower airway epithelium, we sought to determine if the CF-associated defect in TER is recapitulated in this tissue model of disease.

Objectives: To determine possible defects in the paracellular pathway of patient-derived, primary CF respiratory epithelial cultures.

Methods: CFTR-mediated transepithelial chloride currents were measured in Ussing chamber studies of patient-derived airway tissue cultures and using a membrane potential sensitive fluorescent dye. Barrier function was assessed through TER measurements and paracellular permeability, assessed as the rate of fluorescent dextran particles (10 kDa) flux from apical to basolateral compartments.

Results: Bronchial cultures derived from homozygous F508del CF lung explants did not exhibit significant differences in TER measurements compared to non-CF cultures. However, compared to non-CF cultures (n=4 non-CFTR donors), CF cultures (n=7 homozygous F508del donors, n=4 replicates per patient) exhibited a 5-fold increase in paracellular permeability of fluorescent dextran. This paracellular leak phenotype was not uniformly recovered following correction of VX-809. We also derived nasal cultures from 3 CF patients homozygous for W1282X. The small abundance of this mutant protein is expected to lack the C-terminal PDZ motif responsible for protein interactions with tight junction complexes. When grown under air-liquid interface conditions, W1282X nasal cultures displayed defective CFTR channel function and low amounts of protein expression that was insensitive to VX-770/VX-809 treatment, as expected. As in the case of the F508del patient cultures, W1282X cultures did not display defects in TER (200-300Ω cm 2 ) compared to non-CF cultures (300Ω cm 2 ) or paracellular permeability defects when differentiated to the same stages as confirmed through immunofluorescence studies.

Conclusions: Primary bronchial epithelial cells exhibit a paracellular defect in homozygous F508del CF cultures. Absence of increased paracellular permeability in W1282X CF cultures may suggest differential secondary defects to the CF epithelium that are mutation dependent.

Acknowledgments Approximately twelve percent of CFTR mutations are rare mutations that alter pre-mRNA splicing. Splice switching antisense oligonucleotide (SSO) technology has emerged as a novel therapeutic strategy in personalized medicine to modify gene expression by modulating pre-mRNA splicing. One splicing-related mutation in CFTR is the 3849+10kb C>T splice mutation. The 3849+10kb C>T splice mutation creates a de novo 5' splice site resulting in the inclusion of a cryptic exon. This cryptic exon contains a stop codon thereby causing the translation of a truncated CFTR protein ending at amino acid residue 1254. We have designed an SSO to target the cryptic splice site created by 3849+10kb C>T mutation. Treatment with the SSO blocks cryptic splicing and increases the abundance of full-length CFTR mRNA in both patient lymphoblasts homozygous for the splice mutation and primary bronchial epithelial cells from a compound heterozygote patient with the 3849+10kb C>T splice mutation and ΔF508. Importantly, SSO treatment increases cAMP-activated chloride secretion in differentiated primary patient-derived bronchial epithelial cells measured by an equivalent current (Ieq) assay utilizing a TECC-24, demonstrating that the SSOs can restore CFTR function. When analyzed in comparison to current FDA approved CF drugs ivacaftor (VX-770) and ivacaftor/lumacaftor (VX-770 in combination with VX-809), SSO treatment resulted in a greater increase in CFTR function compared to that achieved with either treatment alone. Together, our results demonstrate the ability of SSOs to correct aberrant splicing in the 3849+10kb C>T CFTR splice mutation as measured by an increase in the abundance of full length mRNA as well as the partial restoration of functional protein activity. Supported by the CFF.

Sasaki, S.; Sun, R.; Zhao, C.; Crosby, J.; Monia, B.; Guo, S. Ionis Pharmaceuticals, Carlsbad, CA, USA Introduction: Cystic fibrosis (CF) is an autosomal-recessive genetic disease that affects approximately 70,000 individuals worldwide. CF is caused by mutations in CFTR that result in decreased protein production and function of the chloride channel. Mutations in CFTR most severely affect the lung, causing a build-up of thick mucus and increased susceptibility to infection. As a result, lung disease is the leading cause of morbidity and mortality in CF patients. While many symptomatic treatments exist, quality of life and life expectancy for CF patients are low, and better therapies to improve CFTR expression and function are in need. Thus, we have developed an approach to upregulate CFTR expression using antisense oligonucleotide (ASO) treatment.

ASOs are a therapeutically proven approach to drug discovery for diseases with a genetic basis. They are often used to decrease protein expression via targeted RNase H-mediated degradation of mRNA. More recently, ASOs have been used as steric blocks of splicing factors and regulatory elements to increase mRNA and protein expression (Liang X, et al. Nat. Biotechnol. 2016; 34:875-80; Finkel RS, et al. Lancet. 2017; 388(10063) :3017-26). ASOs can be effectively delivered to various lung cell types by aerosol inhalation (Crosby JR, et al. J Pharmacol Exp Ther. 2007; 321(3) :938-46).

The human CFTR mRNA is a 6 kb transcript with many putative regulatory sites, suggesting that it may be tightly regulated at the mRNA level and could therefore be a good target for ASO-mediated upregulation.

Objective: To identify ASOs that can block inhibitory regulatory elements, such as RNA-binding protein and miRNA sites, in the CFTR transcript and thus increase CFTR protein expression and function. The eventual goal is to develop aerosol-delivered ASOs for therapeutic treatment of cystic fibrosis.

Methods: mRNA expression levels were measured via qPCR. To assay for upregulation of CFTR protein expression, we have developed a CFTR ELISA in conjunction with a cell-based assay for surface CFTR expression to discriminate between membrane and whole-cell CFTR expression. In parallel, we use an ion flux assay with halide-sensitive mutants of YFP to monitor CFTR channel function in a high-throughput manner (Galietta LVJ, et al. Am J Physiol Cell Physiol. 2001; 281:C1734-42) .

Results: We have identified potential structural elements, RNA-binding elements and miRNA sites in the CFTR transcript that may affect its translation. ASOs to block these sites and other previously identified regulatory elements that may inhibit CFTR protein expression have been designed and tested via in-house developed high-throughput assays for CFTR expression and function. ASO inhibition of 5' structural elements increases protein expression and function by 2-to 4-fold, and mechanistic assays are ongoing.

Conclusion: Our preliminary data demonstrates that ASO-mediated upregulation of CFTR presents a promising and novel approach to the development of therapeutics for cystic fibrosis. introduction: CRISPR-cas9 gene editing has been used extensively in the past 5 years in a variety of fields such as generation of mutant cell lines or animal models, gene knock-in. CRISPR is a foreign protein to the host immune system; without shutting down its expression following gene editing, it would cause the host immune system to eliminate the gene edited cells. Currently, gene editing or targeting efficiency in animal models is very low. To overcome these challenges, we tested the helper-dependent adenoviral (HD-Ad) vector for CRISPR-medicated gene targeting to permenantly correct cystic fibrosis transmembrane conductance regulator (CFTR) gene.

Hypothesis: Donor genes that are large in size (6kb-8kb) can be integrated into the human AAVSI locus using the HD-ad system carrying the CRISPR gene.

Both donor DNA and CRISPR-cas9 system were cloned into the HD-ad vector plasmid backbone. Viral vectors were produced in 3-liter cultures and purified by CsCl ultracentrifugation. Transduction of viral vectors were performed in either IB3-1 or A549 cells. The expression of CRISPR was examined via qPCR and Western blot analysis. To quantify the extent of gene integration in the human AAVSI locus, transduced cells were sorted into 96-well plates and allowed to grow till confluency. Then, X-galactosidase staining was performed to quantify the percentage of positive colonies. The LacZ donor with 4kb homology arms yielded the highest integration efficiency (10%), whereas shorter homology arms had lower integration efficiency. Very limited residual CRIPSR expression, which in abundance may cause antigenic response, was found in transduced cells after 9 days, as measured by qPCR, and Western blot analysis. Successful integration of CFTR as donor gene using the HD-Ad integration system was verified using junction PCR analysis and CFTR channel function was rescued as measured by iodide efflux assays.

Conclusion: Efficient donor gene integration can be achieved using the HD-Ad system and residual CRIPSR expression was eliminated in gene-integrated cells. Therefore, HD-Ad is an effective and safe system for CRISPR mediated gene integration. CT, USA; 3. Peds, Yale, New Haven, CT, USA; 4. Genetics, Yale, New Haven, CT, USA Introduction: Advances in noninvasive genetic testing allow for prenatal disease detection using cell-free fetal DNA obtained from maternal blood as early as 7 weeks of gestation (Nature. 2012; 487:320-4) . Despite the possibility of early diagnosis, many neonates and children with CF suffer substantial morbidity and mortality as currently available therapies do not correct the underlying gene defect. There is a growing consensus that treating CF patients early is crucial in preventing or delaying irreversible organ damage. Even at birth, there are signs of multi-organ disease, evidenced by the occurrence of pancreatic insufficiency, reduced birth weight, meconium ileus, tracheomalacia, biliary cirrhosis and absence of the vas deferens. We hypothesize that early intervention through in utero gene editing could correct mutations during the initial stages of pathogenesis, which could allow for normal organ development, disease improvement, and possibly cure. We previously showed that site-specific gene editing to correct the F508del mutation can be achieved efficiently and safely in adult animals via intranasal administration of biodegradable nanoparticles (NPs) loaded with peptide nucleic acids (PNAs) and donor DNA (Nat Commun. 2015; 6:6952) . The aim of this study was to demonstrate that NPs can be safely delivered to developing fetal mice and that NPs loaded with PNA/ DNA can correct the F508del mutation in utero, resulting in sustained postnatal CFTR function.

Methods: We first studied the biodistribution of fluorescent NPs after intra-amniotic (IA) or intravenous (IV) delivery to mouse fetuses at select gestational ages. We next used a GFP mouse model to assess safety and activity of PNA/DNA NPs to edit lung tissue after IA or IV in utero administration. Lastly, we determined if in utero treatment with PNA/DNA NPs targeting the CFTR locus can be used to correct the F508del mutation and lead to sustained postnatal CFTR activity by measuring the nasal potential difference in a mouse model of CF.

Results: PNA/DNA NPs were safely delivered to various fetal mouse tissues at selected gestational ages (E15-E18). IV administration of NPs resulted in widespread NP delivery, including robust NP accumulation in the fetal lung. IA NP delivery resulted in specific accumulation in the fetal lung and gut at gestational ages later than E15, when fetal swallowing and breathing have begun. In a GFP mouse model, both NP delivery routes resulted in gene correction in the lung, with more significant editing achieved after IV NP treatment. Lastly, in utero PNA/DNA NP delivery to F508del fetal mice resulted in significant mutation correction and functional CFTR activity after both IA and IV NP treatment, at a level similar to that of wild-type mice.

Conclusions: PNA/DNA NPs can be safely administered to fetal mice and correct the F508del mutation, resulting in sustained postnatal functional disease improvement. Our findings represent the first systemic in utero correction of a CFTR mutation and establish the feasibility of a fetal gene editing approach that could be used in the treatment of numerous CF disease-causing mutations. Cystic fibrosis (CF) is a common genetic disease caused by mutations in the gene coding for the cystic fibrosis transmembrane conductance regulator (CFTR). Although CF affects multiple organ systems, chronic bacterial infections and inflammation in the lung are the leading cause of morbidity and mortality in people with CF. Gene complementation with a functional CFTR gene repairs this defect, regardless of the disease-causing mutation. In this study, we used a gene delivery system termed piggyBac/adenovirus (Ad), which combines the delivery efficiency of an adenoviral-based vector with the persistent expression of a DNA transposon-based vector. We aerosolized piggyBac/Ad to the airways of pigs and observed widespread pulmonary distribution of vector. We quantified the regional distribution in the airways. We observed transduction of large and small airway epithelial cells of non-CF pigs, with ~30-50% positive surface epithelium. We transduced multiple cell types including ciliated, non-ciliated, basal, and submucosal gland cells. In addition, we phenotypically corrected CF pigs following delivery of piggyBac/Ad expressing CFTR as measured by anion channel activity, airway surface liquid pH, and bacterial killing ability. Combining an integrating DNA transposon with adenoviral vector delivery is an efficient method for achieving CFTR functional correction from a single vector administration.

Moore, P. 1 ; Tarran, R. 2 1. Medicine, University of North Carolina, Carrboro, NC, USA; 2. Cell Biology and Physiology, University of North Carolina, Chapel Hill, NC, USA In the normal lung, chloride secretion via the cystic fibrosis transmembrane conductance regulator (CFTR) and sodium absorption through the epithelial sodium channel (ENaC) are required to maintain airway surface liquid (ASL) volume homeostasis. The importance of this mechanism is highlighted during cystic fibrosis (CF) pathophysiology, where an imbalance in ion transport results in ASL volume dehydration. SPX-101 is a novel peptide-based ENaC antagonist developed in our laboratory that restores ASL height in CF human bronchial epithelial cultures (HBECs). SPX-101 prevents absorption of a test bolus of solution in CF HBECs. However, whether SPX-101 can induce ASL secretion in alreadydehydrated CF cultures remains to be determined. To test this hypothesis, we added SPX-101 as a dry powder in perfluorocarbon to dehydrated CF HBECs, without an additional volume load. Thus, any changes in ASL height would be caused by the induction of secretion, and not due to an inhibition of absorption. The addition of SPX-101 induced a 2-fold increase in ASL height in a dose-dependent manner in CF (ΔF508 homozygote) airway cultures (n=9). This effect was abolished by bumetanide, an inhibitor of the basolateral Na + /K + /2Cltransporter (n=9). Furthermore, this was not an osmotic effect since it was not reprised by the scrambled peptide (control). We hypothesize that SPX-101, via inhibition of ENaC, hypoerpolarizes the apical membrane to induce anion secretion in CF HBECs via a CFTRindependent pathway. Importantly, these data suggest that SPX-101 is mutation agnostic and may be able to rehydrate CF airways in the absence of CFTR correctors or potentiators. However, the nature of the chloride efflux pathway in CF epithelia remains to be identified.

Funded by the CF Foundation and Emily's Entourage. Introduction: Cystic fibrosis (CF) is caused by mutations disrupting the function of the anion channel CFTR. In CF, mucus stagnation and accumulation leads to bacterial colonization, airway inflammation and ultimately tissue destruction and respiratory failure. Sodium hyperabsorption has been suggested to be an important component of CF. ENaC inhibitors are under development to improve mucus clearance in various lung diseases where mucus accumulation is a problem as inhibiting ENaC is thought to rehydrate the airways by means of inhibiting sodium absorption.

Methods: Compound A (Åstrand A, et al. AJP-Lung. 2015 ;308:L22-32) was selected from a library of soluble compounds with good selectivity and potency for inhibiting ENaC (pIC 50 > 9). Excised newborn CFTR -/-(CF) and wild-type (WT) pig airways were incubated with Compound A for 2 hours at room temperature. The transport of Alcian blue-stained mucus bundles was evaluated by mounting the distal trachea and primary bronchi in a heated chamber with aerated Krebs-glucose buffer and timelapses acquired. In addition, WT and CF pig trachea were incubated with Compound A in an Ussing-type chamber and surface pH measured using microelectrodes. Intestinal explants from WT and CF mice were used to evaluate mucus attachment (Gustafsson JK, et al. J Exp Med. 2012; 209:1263-72) and to measure surface pH. Competitive assays with fluorescent Compound A and unlabeled compounds were performed using CHO cells and a CytoFlex cell sorter.

Results: Alcian blue-stained mucus bundles were essentially stagnant in CF pig trachea but were transported in WT. Compound A treatment of CF trachea increased the transport velocity to WT levels. In contrast to WT, CF mouse ileal mucus is attached to the epithelium. Intriguingly, Compound A could detach the attached CF mucus in the ileum where ENaC is not expressed. The classical ENaC inhibitor amiloride at 0.1 mM had no effect. The effect of Compound A was mimicked by a selective NHE3 inhibitor. Compound A increased the pH both in CF trachea and ileum. The binding of fluorescent Compound A was displaced by unlabeled compound in CHO cells expressing NHE.

We have previously shown that bicarbonate transported via CFTR is required for the formation of normal, easily movable mucus in the ileum and that lack of bicarbonate causes attached mucus. As ileum does not express the ENaC protein and the binding at NHE was concentration dependently displaced by unlabeled drug, the Compound A effect could not be due to ENaC inhibition. We now show that Compound A normalizes mucus bundle transport velocity in CF piglet trachea and detaches CF mouse ileal mucus via NHE exchangers, but s0ubtype selectivity remains to be determined. Center, Chapel Hill, NC, USA; 4. Kither Biotech Srl, Torino, Italy Background and Rationale: The underlying cause of cystic fibrosis (CF) is a mutation in the gene encoding the cystic fibrosis transmembrane conductance regulator (CFTR), a cyclic AMP (cAMP)-stimulated chloride channel. The consequent CFTR dysfunction results in obstruction of small airways and airway inflammation and eventually leads to respiratory failure. Recently, a number of CFTR correctors and potentiators, improving membrane expression and gating of the channel respectively, have been developed, but their ability to rescue the basic defect of CF is still unsatisfactory.

Hypothesis and Objectives: We previously showed that phosphoinositide 3-kinase γ (PI3Kγ) acts as a scaffold protein which negatively regulates cAMP by favoring the activation of key cAMP-degrading enzymes, phosphodiesterases 3 and 4 (PDE3 and PDE4). Here, we hypothesize that targeting PI3Kγ scaffold activity enhances cAMP in airway smooth muscle, immune and epithelial cells, leading to concomitant (i) bronchodilation, (ii) anti-inflammatory effects and (iii) CFTR modulation.

Methods: We explored the ability of a cell-permeable peptide targeting the scaffold activity of PI3Kγ (Patent n° PCT/IB2015/059880 -WO/2016/103176) to function as a (i) bronchodilator, (ii) anti-inflammatory agent, and (iii) CFTR modulator. A mouse model of chronic lung inflammation (OVA-sensitized mice) and human primary bronchial epithelial cells (F508del) were used.

Results: We found that, in vivo, the peptide can be efficiently delivered to the lower airways by intratracheal instillation in mice and significantly elevates cAMP in the lungs. Notably, the peptide limits methacholine-induced airway hyperresponsiveness and reduces neutrophilic lung inflammation in OVA mice. In vitro, the peptide potentiates F508del-CFTR currents upon pharmacological correction with VX-809 and, unlike the gold-standard potentiator VX-770, does not interfere with channel stability in the chronic setting. In VX-809-corrected primary cells, the peptide also synergizes with VX-770, by increasing its efficacy by 5-fold and is able to stabilize the CFTR to the plasma membrane.

Conclusions: Overall, the results of this study demonstrate that the peptide targeting PI3Kγ may be exploited therapeutically as a new medicinal product that may offer unique advantages over current existing therapies for CF: i) three independent therapeutic benefits in a single molecule (CFTR rescue, anti-inflammatory and bronchodilator effects); ii) high intrinsic specificity of action (due to its peptidic nature), and iii) limited side effects (thanks to the delivery by inhalation which contributes to maximize its effects on the lungs and, at the same time, reduce possible side effects typical of systemic administration).

Acknowledgments Vasoactive intestinal peptide (VIP) is a 28-amino acid neuropeptide that functions as a neuromodulator and neurotransmitter secreted by intrinsic neurons innervating exocrine glands. VIP is a potent vasodilator and bronchodilator that among other functions regulates exocrine gland secretions by controlling CFTR-dependent chloride secretion, thereby contributing to mucus hydration and local innate defense of the lung. In cystic fibrosis, we found an intriguing marked reduction in VIP around submucosal glands of the lung, pancreas and sweat glands of CF mice, as well as a marked reduction in VIP amount in the small intestine that is related to reduced VIPergic innervation.

Here we report the pre-clinical evaluation of VIP-elastin-like polypeptide (ELP) biopolymer fusion molecules PB1120 and PB1046 that are highly resistant to the peptidases that normally rapidly degrade VIP in vivo and preclude its use as a therapeutic. Also, ELP biopolymer provides slow release from the injection site. In phase 1 and 2 studies, in subjects with hypertension or heart failure, PB1046 was well tolerated and provided exposure of at least a week following each subcutaneous injection. PB1120 has similar potency at the VPAC1 and VPAC2 receptor, whereas PB1046 is relatively more selective for VPAC2. In primary airway epithelial cells from F508del homozygous patients, treatment with PB1120 or PB1046 restored F508del-CFTR protein trafficking and chloride channel activity. Moreover, the molecules enhanced the effect of ivacaftor + lumacaftor.

We have now tested VIP, PB1120 and PB1046 in vivo by injecting 8-to 14-week-old CF mice three times per week for up to five weeks. At 14-weeks old, CF mice receiving PBS (sham) injections presented signs of disease progression with severe tissue damage in the lung and duodenum including extensive remodeling, inflammation, epithelium damage and mucus secretion in the airways as well as goblet cell hyperplasia, villi distortion and severe crypt enlargement in the duodenum. CF mice that received fifteen IP injections of VIP (40 µg/injection) showed much less inflammation and a low level of tissue remodeling in the lung, whereas the intestinal phenotype was not significantly improved. CF mice receiving subcutaneous injections of PB1120 showed some improvement but disease was still progressing in the lung and duodenum. Interestingly, signs of disease were absent or very minimal in mice receiving subcutaneous injections of PB1046 (equimolar to VIP), in both lung and duodenum.

In conclusion, our data indicate that VIP-ELPs are strong potential candidate drugs for the treatment of CF patients with the F508del mutation. In vitro models indicate that combining PB1120 or PB1046 with ivacaftor and lumacaftor produces an additive positive effect on F508del-CFTR function and in vivo studies further demonstrate the benefit of a VIP-ELP treatment at the tissue level.

Introduction: Over 2000 different mutations have been reported in patients with CF and found to occur in all CFTR exons and introns. Of these, 168 are nonsense mutations, 295 are frameshift mutations that are not amenable to current therapies, and therefore new drugs must be developed. Antisense oligonucleotides (AOs) are synthetic RNA analogues that can be designed to anneal to selected splice motifs within pre-mRNAs. AO binding alters the recognition of the splice site by the spliceosome and therefore modulates exon selection. Exon 15 has been selected as an initial target since it has been reported to harbour ~40 mutations and exclusion of this exon will not disrupt the mRNA reading frame. We hypothesize that by skipping exon 15 in patients with amenable mutations such as p.Phe861Leufsx3, studied here, the disease-causing mutation can be bypassed and the induced isoform may retain some residual function, therefore altering the course of disease.

Methods: AO sequences were initially optimised using 2'-O-methyl modified bases on a phosphorothioate backbone (2OMe) and transfected into monolayer primary non-CF (2.6 years M) and p.Phe861Leufsx3/p. Phe508del CF airway epithelial cells (4.1 years M). The ratio of the AO induced, CFTR RT-PCR transcript product missing the target exon, relative to the full length product provides an estimate of AO exon skipping efficiency. The most effective 2OMe AO sequence was identified and re-synthesised as the clinically validated phosphorodiamidate morpholino (PMO) chemistry. Monolayer transfections were repeated. AO mediated modification of protein was shown by Western blot analysis and comparison to size standards. CFTR function before and after PMO application was measured using Ussing chamber studies.

Results: 2OMe AOs were designed, evaluated and further optimised by micro-walking around sequences shown to be capable of modifying splicing. The 2OMe sequence that was most efficient induced an estimated 50% skipping in p.Phe861Leufsx3/p.Phe508del CF airway epithelial cells The PMO induced efficient skipping of exon 15 from non-CF (49%) and p.Phe861Leufsx3/p.Phe508del CF (88%) cells after 7 days in culture. Western blot was used to determine the effect on the induced CFTR protein. Airway epithelial cells from children with CF were also grown at air-liquid interface 28 days to become mucocillary differentiated then assessed for CFTR function using an Ussing chamber.

Conclusion: Exon 15 can be efficiently skipped from the CFTR transcript in both non-CF and CF-derived airway epithelial cells. We propose that exon skipping to remove disease-causing mutations in selected in-frame exons can improve function in amenable CF patients, either alone or in combination with current therapeutics.

Acknowledgments Objectives: Seliciclib (R-roscovitine) is an orally available kinase inhibitor which has undergone extensive phase I and II clinical trials against various cancers and is currently under clinical evaluation against Cushing disease and rheumatoid arthritis (1) . Chronic lung infection with Pseudomonas aeruginosa is linked to progressive pulmonary function decline in cystic fibrosis (CF) patients. Seliciclib displays biological properties suggesting potential benefits in the treatment of CF (2): it partially corrects F508del-CFTR trafficking (3); it stimulates the bactericidal properties of CF alveolar macrophages by lowering their abnormally elevated intraphagolysosomal pH (4); and it displays anti-inflammatory properties due to effects on neutrophils, eosinophils and the Th17/Tregs inflammatory lymphocytes balance (2) . The ROSCO-CF clinical study aims at evaluating this therapeutic potential (5) .

Methods: ROSCO-CF is a Phase IIA, dose-ranging (200, 400, 800 mg seliciclib; 4 days/week, 4 weeks), multicenter, double-blind, placebo controlled clinical study, involving 36 CF patients. It was launched in 2016 to evaluate the safety (primary end point) and effects (secondary end points) of seliciclib in P. aeruginosa infected adult CF patients carrying two CF causing mutations with at least one F508del-CFTR mutation and harboring a FEV-1 ≥ 40%.

Results: The three groups (200, 400 and 800 mg roscovitine, 12 patients each) completed their treatment in March 2017, December 2017 and July 2018, respectively.

Conclusion: We will present the results of this phase 2a clinical trial with the new drug candidate seliciclib (safety, pharmacokinetics, antibacterial and anti-inflammatory effects, spirometry assessment, sweat chloride measurements, pain assessment, quality of life questionnaire answers). To the best of our knowledge, seliciclib (R-roscovitine) could represent a "first-in-class" kinase inhibitory drug for the treatment of CF-associated infection and inflammation. F MRI, using inhaled perfluoropropane (PFP) as an air contrast agent that does not require hyperpolarization, to characterize ventilation abnormalities in subjects with cystic fibrosis. We hypothesized that 19 F parametric ventilation MRI would identify regional ventilation defects and abnormal gas exchange kinetics that differentiate cystic fibrosis (CF) subjects from healthy volunteers.

Methods: This prospective pilot and feasibility study enrolled 13 healthy volunteers and 17 subjects with stable CF between November 2014 and November 2017. The CF subject cohort included 7 individuals with FEV 1 values > 80%. Safety measures, including heart rate, pulse oximetry, and exhaled CO 2 were monitored throughout study procedures. Spirometry was also performed before and after each scan. Dynamic 19 F MRI imaging was acquired during repeated breath holds (~12 seconds) and ventilation kinetic maps were created. Ventilation defects were identified and expressed as the percentages of total thoracic lung volumes, as measured by 1 H imaging. The time constants of PFP gas wash-in and wash-out kinetics were computed on a voxel by voxel basis. Statistical analyses included t-tests with the Satterthwaite correction and nonparametric tests (Wilcoxon Mann Whitney), depending on the normality of the data.

Results: The mean ventilation defect percentage (VDP) was 18.4% ± 7% in healthy controls; 31.3% ± 21.5% in mild CF; and 47.1% ± 17.1% in severe CF. A significant difference was found when comparing the VDP of all CF patients to normal (p=0.009). VDP had a significant negative correlation with FEV1 (-0.56 via Spearman correlation, p=0.011). The median rate constant for gas filling of all lung units (t 1 ) was also increased, though not statistically significant, in CF patients compared with controls (60.7 vs. 48.1) and a significantly higher percentage of voxels had t 1 values above a normal-defined threshold, suggesting delay in alveolar filling. No safety concerns were detected throughout the study.

Conclusion: 19 F MRI revealed distinctly abnormal ventilation images in patients with CF and detected ventilation defects and slowed gas wash-in kinetics in patients with preserved FEV 1 values. PFP inhalation was well tolerated, even in subjects with severely impaired lung function. The ability to characterize regional gas wash-in kinetics adds an additional dimension over traditional hyperpolarized gas techniques that could prove to be useful. Rationale: Sputum culture is the golden standard for detecting Pseudomonas aeruginosa (PA) in cystic fibrosis (CF) patients; however, this is a time-consuming method. Volatile organic compound (VOC) profiles in exhaled breath, as measured by gas-chromatography combined with mass-spectrometry (GC-MS), have already been associated with PA colonization (Robroeks CM, et al. Pediatr Res. 2010; 68:75-80) . Exhaled breath measurements based on pattern recognition of VOCs by electronic noses (eNose) might allow sensitive PA detection at the point-of-care.

Aim: To determine the diagnostic accuracy of exhaled breath analysis by eNose for discrimination between CF patients with and without a PA colonization.

Methods: This was a cross-sectional observational study in CF patients with or without PA colonization. As part of spirometry, exhaled breathprints were collected in duplicate by eNose (SpiroNose) (De Vries R, et al. J Breath Res. 2015 Oct 15; 9(4) :046001). Colonization was defined as 1) PA positive culture at inclusion, or 2) PA present in ≥50% of cultures over the past year (≥4 samples). Data analysis involved signal processing, ambient correction and statistics based on principal component (PC) analysis followed by discriminant analysis. Finally, a receiver operating characteristic (ROC) curve was constructed; area under the curve, including a 95% confidence interval (CI), was determined.

Results: Exhaled breath results of 8 paediatric and 28 adult CF patients colonized (n=20) and noncolonized (n=16) with PA were available. Three principal components (PC1, PC2, and PC3) showed a significant difference (p<0.01) between the two groups, with a cross-validation value of 81% and ROC-AUC of 0.93 (CI 0.84-1.00).

Conclusion: Breath analysis by eNose is able to discriminate between CF patients with and without PA colonization. When independently validated (in progress), these results are supportive of using eNose technology for fast, sensitive and noninvasive detection of PA at point-of-care.

Townsend, S.; Twomey, D.; Cann, A.; Mucha, J. SciBac, Inc., Burlingame, CA, USA Introduction: Chronic Pseudomonas aeruginosa pulmonary tract infections are a primary cause of morbidity and mortality for individuals with cystic fibrosis (CF). The formation of anaerobic biofilms support tolerance to antibiotics like tobramycin which require oxygen and metabolically active P. aeruginosa to be effective (Drenkard E, et al. Microb Infect. 2003; 5:1213-9) . A novel live biotherapeutic with activity against chronic multidrug-resistant infections in CF lungs represents an innovative treatment strategy against the expanding antibiotic resistance public health crisis.

Methods: Patented Directed Recombination by In Vitro Evolution (DRIVE) technology combined metabolically beneficial and therapeutic traits from probiotic strains, such as bacteriocin production and anaerobic growth, into a live biotherapeutic hybrid. Strain characterization was completed using Etest and biological assays for alginate lyase, mucinase, and DNase activity. The spot test and streak test were used to characterize bactericidal activity against 20 multidrug-resistant P. aeruginosa clinical isolates. An acute murine (BALB/c) lung infection model evaluated preliminary safety following intranasal administration.

Results: The in vitro data shows that the DRIVE technology combined multiple therapeutic traits into a single hybrid that metabolizes mucin and produces a bacteriocin that kills multidrug-resistant P. aeruginosa. The hybrid was also observed to penetrate alginate, a component of P. aeruginosa biofilm, and hydrolyze DNA. The hybrid retained resistance to aztreonam and sensitivity to tobramycin. Safety tests in healthy BALB/c mice have shown that intranasal administration of the live biotherapeutic at a concentration of 1 x 10 8 CFU is nontoxic (100% survival) with no adverse effects after 5 days compared to mice treated with 1 x10 7 CFU P. aeruginosa (0% survival).

Conclusions: These preliminary studies suggest DRIVE derived live biotherapeutics facilitate a targeted antibacterial and mucolytic activity that may be administered locally to deliver pluripotent therapeutics directly to the lungs to treat chronic multidrug-resistant P. aeruginosa. The novel live biotherapeutic is being developed for intranasal or nebulized delivery in parallel with the standard 28-day cycle of aztreonam or as a stand-alone treatment. The development of live biotherapeutics for localized treatment for CF-associated P. aeruginosa infections is an innovative strategy to combat chronic multidrug-resistant infections and may result in more positive therapeutic outcomes. Inhaled drugs are typically cleared very rapidly from the airway mucosa. Inhaled amiloride accelerated mucociliary clearance (MCC) in clinical studies, but with a half-life in the airway mucosa of <30 minutes, it was a short acting effect which did not translate into clinical efficacy. To achieve a long duration of action in the clinic with novel inhaled ENaC blockers, it will be necessary to maintain compound levels in the airway lumen at a concentration that will maintain target engagement for a sufficient period of time.

The aim of the present project, was to discover novel inhaled ENaC blocker compounds. A key element of the optimisation phase was to understand lung-specific pharmacokinetics, and how these related to in vivo efficacy. To this end, two key models were used. The first employed intratracheal delivery of test compound into the lungs of rats with sampling of compound levels in the airway, lung tissue and plasma. Compounds were then tested for duration of action and efficacy in the sheep model of MCC.

Over the course of this project, we discovered numerous novel and potent ENaC blockers with a range of pharmacodynamic, pharmacokinetic and physicochemical properties. Of these, ETD001 was selected as a development candidate based upon its overall profile. Notably, ETD001 was not the most potent compound, with an in vitro potency (in nM) of: 40, 31 and 30 on human, sheep and rat ENaC respectively. ETD001 did however show a long residence in the airway lumen of the rat, which was accompanied by a potent and long lasting acceleration of MCC in the sheep following inhaled delivery. Other, more potent examples of novel ENaC blockers, showed significantly greater on-target potency (<5 nM) but failed to show efficacy in the sheep, which could be correlated with a diminished retention of compound in the airway lumen.

Based on the observations that ETD001 showed a long duration of action in vivo and that this was due to sustained compound levels in the airway lumen, we tested the hypothesis that repeat inhaled dosing would further increase the potency of the compound. After a single inhaled dose of ETD001, maximum efficacy in the sheep was observed at 13 µg/kg. In contrast, following twice daily dosing of ETD001 for 3.5 days, the maximal efficacious dose was reduced to 3 µg/kg, a 4-fold increase in potency.

In parallel and consistent with the sheep efficacy data, a 7-day repeat dose study in the rat revealed that lung levels of ETD001 were increased between days 1-7 in a dose-dependent manner. Furthermore, there was no change in blood potassium levels induced by ETD001 at any of these dose levels studied.

In summary, ETD001 is a novel inhaled ENaC blocker with a long duration of action in the airway lumen. Safety and efficacy data support a human tolerated dose that is 30-40 fold over the predicted minimum efficacious human dose. This is significant in view of the potential underdosing of the Vertex ENaC blocker, VX-371, in a recent negative Phase 2 study.

Longitudinal computer tomography studies in children with cystic fibrosis (CF) diagnosed by newborn screening (NBS) as well as cross-sectional magnetic resonance imaging (MRI) studies showed early onset and progression of CF lung disease. However, determination of the effects of NBS on early CF lung disease was hampered due to lacking a contemporary group of clinically diagnosed children.

Objectives: To investigate the effect of CF NBS on lung disease by comparing contemporary newborn-screened to early and late clinically diagnosed preschool children to differentiate between the influence of age at and mode of diagnosis.

Methods: 96 infants and preschoolers with CF (age 0-5years) were followed for 4 years after diagnosis with annual MRI scans (evaluation using dedicated morphofunctional score) and documentation of clinical parameters.

Measurements and Main Results: Over the entire observational span, the global MRI score increased in the entire cohort (P<0.001), clearly indicating the development of lung changes from diagnosis onwards. The number of pulmonary exacerbations (PE) was stable during the observation period, but PE and coughing showed a significant association with MRI scores. Comparing diagnostic patient subgroups, the global MRI score was lower in the NBS group than in the clinically diagnosed groups (P<0.001), whereas disease progression was rather similar in all subgroups (P=0.166).

Conclusions: For the first time, longitudinal MRI data showed the persistence and progression of CF lung disease despite an optimal standard of care. The less severe lung disease in newborn screened children, however, offers a narrow window of opportunity for therapeutic interventions. Taylor Med & Ped, National Jewish Health, Denver, CO, USA; 2. Clin Res Serv, National Jewish Health, Denver, CO, USA; 3. Biostats & Bioinform, National Jewish Health, Denver, CO, USA; 4. Kaiser Permanente, Denver, CO, USA; 5. Med, National Jewish Health, Denver, CO, USA; 6. Pulm, Children's Colorado, Aurora, CO, USA Background: CFTR knockout mice have increased right ventricular torsion and strain in the absence of lung disease (J Cyst Fibros. 2016; 15:34) . Despite substantial improvements in care, patients with CF often develop pulmonary vascular disease (PVD) leading to pulmonary hypertension. The extent to which subclinical PVD contributes to decline in exercise tolerance and quality of life (QOL) in patients with CF is unknown. Identifying those CF patients with PVD prior to onset of right ventricular dysfunction may allow early pharmacologic intervention. Clinical trials have demonstrated that treatment with the phosphodiesterase type 5 inhibitor sildenafil can decrease pulmonary vascular resistance and improve exercise tolerance in non-CF patients.

Objectives: We hypothesized that PVD contributes to exercise intolerance in patients with CF lung disease even in the absence of clinical evidence of pulmonary hypertension, and that sildenafil treatment would be safe and increase exercise capacity, improve cardiac performance during exercise, and enhance QOL.

Methods: In a single-site pilot study, adult subjects with moderate to severe CF lung disease without resting pulmonary hypertension were randomized 3:1 to receive sildenafil 40 mg p.o. t.i.d. or matching placebo for 12 weeks. All patients continued standard of care therapy. At weeks 1 and 12 subjects completed the CFQ-R questionnaire and underwent tests of cardiopulmonary exercise, 6-minute walk distance, sweat chloride, exhaled nitric oxide, pulmonary function, cardiac MRI, and Fitbit™ steps.

Results: 14 patients were randomized; 12 completed all study visits. At screening, subjects' mean age was 33.2 y (range 22-57); 50% (6/12) were female; mean ppFEV1 = 59.2 (range 35-68); mean BMI = 21.7 (range 16.5-27.6); 67% (8/12) were on CFTR modulators. Side effects were mild-moderate and consistent with previous reports. One patient in each arm withdrew for pulmonary exacerbation. There were no study drugrelated serious adverse events. Improvements in exercise tolerance, right heart function (global longitudinal strain on cardiac MRI) and the respiratory component of CFQ-R were observed in those treated with sildenafil (Table) .

Conclusion: In adults with moderate to severe CF lung disease, oral sildenafil administration was safe and improved cardiac performance, exercise tolerance and QOL. A phase III trial of sildenafil is warranted. 1. National Jewish Health, Denver, CO, USA; 2. Vertex Pharmaceuticals Inc, Boston, MA, USA; 3. Mater Hosp, Brisbane, QLD, Australia; 4. Banner Univ Medical Center, Tucson, AZ, USA; 5. Charité Univ Medicine Berlin, Berlin, Germany; 6. St Vincent's Univ Hosp, Dublin, Ireland; 7. Seattle Children's Hosp, Seattle, WA, USA; 8. Univ of Alabama at Birmingham, Birmingham, AL, USA; 9. Children's Hosp of The King's Daughters, Norfolk, VA, USA; 10. St Michael's Hosp, Toronto, ON, Canada; 11. Alfred Hosp, Melbourne, VIC, Australia Introduction: CF patients (pts) with F508del/minimal function (F/ MF) genotypes do not have approved CFTR modulator therapy. Adding a next-generation (NG) CFTR corrector to a corrector/potentiator regimen may substantially improve CFTR function in F/MF pts and enhance benefit over 2-drug regimens in F508del/F508del (F/F) pts.

Objective: Evaluate in vitro efficacy and clinical safety and efficacy of VX-445, an investigational NG CFTR corrector, in triple combination (TC) with tezacaftor (TEZ) and ivacaftor (IVA), or TEZ and VX-561 (deuterated IVA) in CF pts with F/MF or F/F genotypes.

Methods: Effects of TC on CFTR processing/trafficking/function were evaluated in bronchial epithelial cells from F/MF or F/F donors. TC regimens were evaluated in F/MF or F/F pts for up to 4 weeks, the latter after 4 weeks of TEZ/IVA run-in. Primary endpoints were safety and absolute change in percenr predicted (pp) FEV 1 from baseline (BL; in F/F pts, after TEZ/IVA run-in). Sweat chloride (SwCl) and CFQ-R respiratory domain (RD) score were secondary endpoints; reported P values are nominal for these endpoints as no multiplicity adjustment was performed. Data included are through last pt's last visit.

Results: VX-445/TEZ/IVA improved CFTR processing/trafficking and chloride transport more than TEZ/IVA in vitro. Both TC regimens were efficacious (Table) , generally safe, and well tolerated in F/MF and F/F cohorts; most adverse events (AEs) were mild/moderate. Of 95 TC-treated, 4 pts had AEs leading to discontinuation during the treatment period (rash, n=2; chest pain, n=1; increased bilirubin, n=1).

Conclusions: VX-445 TC regimens demonstrated robust, consistent, and clinically meaningful improvements in F/MF pts, in whom previous modulator regimens have not shown efficacy, and in F/F pts beyond those of TEZ/IVA alone. Further development of VX-445 TC in phase 3 is ongoing.

Acknowledgments: Sponsored by Vertex Pharmaceuticals Incorporated. 4 1. NHLI, Imperial College and Royal Brompton Hospital, London and Queen's University of Belfast, Belfast, United Kingdom; 2. National Jewish Health, Denver, CO, USA; 3. University Of Alabama at Birmingham, Birmingham, AL, USA; 4. Celtaxsys, Atlanta, GA, USA Background: The chronic inflammatory component of CF lung disease causes lung injury and lung function decline, leading to morbidity and mortality in people with CF. There are no therapies approved by regulatory authorities for inflammation in CF and previous studies show CFTR modulators do not consistently impact underlying lung inflammation. Acebilustat is a once-daily oral LTA4 hydrolase inhibitor which reduces sputum LTB4, neutrophil DNA and elastase levels, and serum CRP in CF patients (Elborn JS, et al. Clin Transl Sci. 2017; 10:28-34) . In CF patients, we hypothesized acebilustat would decrease lung inflammation, reduce the frequency of CF exacerbations, and diminish lung function decline.

Methods: EMPIRE is a randomized, double-blind, placebo-controlled trial that enrolled CF patients on standard of care to: acebilustat 50 mg or 100 mg, or placebo once daily for 48 weeks. Recruited patients (CFTR mutation agnostic) were selected to enrich for patients with the greatest potential decline in lung function. Key inclusion criteria were age 18-30 years, screening FEV 1 percent predicted (pp) ≥50%, and ≥1 pulmonary exacerbation in the past year. At randomization, patients were stratified by baseline FEV 1 pp (50 to 75% and >75%), number of exacerbations in the 12 months before screening (1 or >1) , and concomitant use of CFTR-modulators (yes/no). Patients were enrolled from 69 sites across North America and Europe. The primary endpoint is the change from baseline in FEV1pp for acebilustat vs placebo, and a key secondary endpoint is the rate of pulmonary exacerbations.

Results: 200 patients were enrolled. At baseline, the mean age was 23.7 years, 51% were female, mean (SD) FEV 1 pp was 70.6 (15.3)%, and 31% were on an approved CFTR modulator. Patients had a mean of 2.1 (1.6) exacerbations in the prior year, with 46.5% having one, 28.5% two, and 25% ≥ three. The last patient, last visit is projected for mid-May 2018, and data lock is projected for mid-Jun 2018. Change from baseline in the FEV1pp and rate of exacerbations will be reported for the study population and also for subgroups based on pre-specified stratification. Safety as assessed by SAEs and the most common TEAEs will be reported.

Conclusions: By addressing persistent, currently untreated underlying lung inflammation, acebilustat has the potential to become a component of foundational therapy for people with CF. The results from this 48-week trial of anti-inflammatory treatment in people with CF regardless of CFTR mutation will be reported.

Acknowledgment: This trial is supported by a grant from CFFT. Background: RPL554 is a first-in-class, inhaled, dual inhibitor of type 3 and 4 isoforms of the phosphodiesterase (PDE) family of enzymes. It has been shown to have bronchodilator and anti-inflammatory effects (1, 2) , both of which are likely to be independent of the mutation status in these patients. Ex vivo data suggests that dual PDE 3 and PDE 4 inhibition may activate the CFTR (3). We investigated the effect of single nebulized doses of RPL554 in CF patients.

Method: A single-centre, randomised, double-blind, placebo-controlled, three-way crossover study was conducted at Papworth Hospital from February to November 2017. Patients received a single dose of RPL554 at 1.5 mg, 6 mg or placebo. The primary objective was to assess the pharmacokinetics (PK) of RPL554. Secondary outcome measures included safety, tolerability and change in FEV 1 . Our exploratory objectives involved investigating the effect on inflammatory biomarkers.

Results: A total of 10 patients (6 males, 4 females, average age = 32.6 years old, most common mutation = F508del) met the inclusion/exclusion criteria and were included in the study. Of these, 9 patients completed all three doses.

The PK profile was consistent with that observed in patients with COPD, although RPL554 C max plasma concentrations were slightly lower in CF patients as compared to historical data with RPL554 in COPD patients. T max (1.2-1.3 hours) and T ½ (7.5 hours at 1.5 mg and 10.1 hours at 6 mg) were similar to previous studies.

The drug also elicited a statistically significant increase in average FEV 1 in patients treated with 1.5 mg (all time points p<0.01) and 6 mg (all p<0.05). This was clinically significant for each time point for both doses. Measurements for exhaled breath condensate (EBC) pH, myeloperoxidase (MPO), IL-8 and TNF α, showed nonsignificant variations from baseline.

RPL554 was well tolerated in this study.

A single dose of RPL554 demonstrated a PK profile consistent with that observed in earlier studies and improved FEV 1 in patients with CF, supporting a larger scale clinical trial evaluation. London, United Kingdom; 2. Vertex Pharmaceuticals Inc, Boston, MA, USA; 3. Indiana Univ. School of Medicine, Indianapolis, IN, USA; 4. Manchester Adult Cystic Fibrosis Centre, Manchester, United Kingdom; 5. Charité Univ Medicine Berlin, Berlin, Germany; 6. St Vincent's Univ Hosp, Dublin, Ireland; 7. Cork Univ Hosp, Cork, Ireland; 8. Schneider Children's Medical Center and Tel Aviv Univ, Tel Aviv, Israel; 9. National Jewish Health, Denver, CO, USA; 10. St Michael's Hosp, Toronto, ON, Canada; 11. Seattle Children's Hosp, Seattle, WA, USA; 12. Boston Children's Hosp, Boston, MA, USA; 13 . Univ of Alabama at Birmingham, Birmingham, AL, USA Introduction: CF patients (pts) with F508del/minimal function (F/MF) genotypes lack approved CFTR modulator therapy. Adding a next-generation (NG) CFTR corrector to a corrector/potentiator regimen may substantially improve CFTR function in F/MF pts and enhance benefit over 2-drug regimens in F508del/F508del (F/F) pts.

Objective: Evaluate in vitro efficacy and clinical safety and efficacy of VX-659, an investigational NG CFTR corrector, in triple combination (TC) with a corrector and potentiator, tezacaftor (TEZ), and ivacaftor (IVA) or TEZ and VX-561 (deuterated IVA).

Methods: Effects of VX-659/TEZ/IVA on CFTR processing/function were evaluated in bronchial epithelial cells from F/MF or F/F donors. VX-659/TEZ/IVA and VX-659/TEZ/VX-561 were evaluated in phase 1/2 trials in F/MF or F/F pts for up to 4 weeks, in F/F pts after 4-week TEZ/ IVA run-in. Primary endpoints were safety and absolute change in percent predicted (pp) FEV 1 from baseline (in F/F pts, after TEZ/IVA run-in). Secondary endpoints were change in sweat chloride (SwCl) and CFQ-R respiratory domain (RD) score; reported P values are nominal for these endpoints as no multiplicity adjustment was performed. Data included are through last pt's last visit.

Results: VX-659/TEZ/IVA improved CFTR processing/trafficking and chloride transport more than TEZ/IVA in vitro. Both TC regimens were efficacious (Table) , generally safe, and well tolerated; most adverse events (AEs) were mild/moderate. One pt on VX-659/TEZ/VX-561 discontinued due to an AE (drug hypersensitivity).

Conclusions: This proof-of-concept study demonstrated robust, consistent, and clinically meaningful improvements with VX-659 TC regimens in F/MF pts, in whom previous modulator regimens have not shown efficacy, and in F/F pts beyond those of TEZ/IVA alone. Phase 3 studies are ongoing to support further development of VX-659 TC regimens. Aberrant pulmonary mucus accumulation and infection are signature traits in CF patients. Mucus plays a vital role in host defense by transporting bacteria out from the lung. MUC5B and MUC5AC are the principal mucin proteins that facilitate mechanical clearance in healthy airways by forming ordered polymers. In contrast, CF airways are characterized by highly concentrated mucus layers. We hypothesize that the origin of this thick immobile mucus in CF airways results from altered structural organization of MUC5B and MUC5AC proteins caused by depletion of water. Hence, understanding the mechanisms of mucin polymer formation in normal versus pathological states is crucial for designing alternative therapies for CF. Previously, Thornton and coworkers reported that the association of adjacent VWD3 domains located in N-terminal region via disulfide bonds drives MUC5B polymer assembly. Despite these studies, absence of accurate atomic-level descriptions of mucin proteins impedes the molecular underpinnings of the polymerization processes of mucins in the lung. Here we generate a potential physiological structural model of the MUC5AC N-terminus using a multilevel computational-experimental methodology. We employ multiple sequence analyses, homology modeling and protein-protein docking techniques for initial model building. Available structural data (e.g., glycosylation sites, cysteine residues known to form disulfide bonds) are applied both for structure validation and as experimental constraints in molecular dynamics simulations for refining the preliminary structural model. This high-resolution structure may serve as a platform to determine the molecular mechanisms of polymerization of MUC5AC protein associated with normal and CF conditions; and reveal sites near oxidated cysteine residues (important for creating intra-and inter-molecular disulfide network in oligomeric states). Molecular description of these sites can expedite the search and design of thiol-based drugs with high efficacy while minimizing off-target interactions and toxicity. Background and Objectives: The most common ΔF508-CFTR mutation results in either the proteasomal-degradation or movement of misfolded CFTR protein into perinuclear aggresome-bodies. Presence of misfolded protein activates reactive oxygen species (ROS) that impairs autophagy, exacerbating proteasomal overload and chronic accumulation of ΔF508-CFTR and other crucial regulatory proteins into aggresome-bodies. Additionally, in CF, the thick airway mucus-buildup, repeated respiratory infections, and chronic lung inflammatory-oxidative stress, together form the physio-chemical barriers that restrict effective drug delivery. Autophagy-inducing FDA-approved drugs such as cysteamine have shown some therapeutic efficacy in CF subjects. In order to overcome these physio-chemical hurdles of drug delivery, a nanoparticle-based drug delivery approach that has shown some promise in CF in our pre-clinical studies was utilized. Briefly, in the present study, we aimed to test a novel nano-formulation with autophagy-inducing antioxidant drug cysteamine (CYS), where we modified the terminal amine groups of cationic dendrimer to form a "cysteamine-like structure" along with an encapsulated drug (cysteamine conjugates) inside the amine surface-decorated dendrimer.

Methods: The CFBE41o-cells were used as the in vitro CF model, where synthesis and characterization of G4-DAB (control) and G4-CYS dendrimer formulations was performed as per our standard protocol and immunoblotting, flow cytometry and autophagy reporter assay were used to quantify the efficacy of the novel dendrimers in rescuing ΔF508-CFTR to the plasma membrane and correcting the resulting autophagy defect, while Pseudomonas aeruginosa-GFP (PaO1) bacteria were used to determine the antibacterial activity of the dendrimer formulation.

Results: We first designed a novel dendrimer formulation (G4-CYS) that significantly increases membrane ΔF508-CFTR expression in CFBE41ocells (p<0.05). Additionally, G4-CYS treatment corrects ΔF508-CF-TR-mediated impaired autophagy as observed by a significant decrease (p<0.05) in Ub-LC3 positive aggresome-bodies. Next, we verified that in nonpermeabilized CFBE41o-cells, G4-CYS significantly (p<0.05) induces ΔF508-CFTR's forward trafficking to the plasma membrane. Furthermore, cysteamine's known antibacterial and antibiofilm properties against Pa were enhanced as our findings demonstrate that both G4-CYS and its control dendrimer, G4-DAB, exhibited significant (p<0.05) bactericidal activity against Pa over the course of an 18-hour bacterial survival. We also found that both G4-CYS and G4-DAB exhibit marked mucolytic activity against porcine mucus (p<0.05). Finally, we demonstrate that G4-CYS not only corrects the autophagy impairment by rescuing ΔF508-CFTR in CFBE41ocells but also corrects the intrinsic phagocytosis defect (p<0.05).

Conclusions: Overall, our data warrants further pre-clinical evaluation of dendrimer-based cysteamine formulation as it holds substantial promise in the treatment of CF due to its superior therapeutic efficacy as compared to direct cysteamine administration.

Correspondence: nvij@vijbiotech.com. Cystic fibrosis (CF) is a hereditary disease caused by mutations in the gene coding for the cystic fibrosis transmembrane conductance regulator (CFTR). Nearly 2,000 genetic variants have been reported, and it is currently unknown how many of these are pathogenic or benign polymorphisms. However, there remains a large number of disease-causing variants, including 74 premature termination codon (PTC) mutations, that cannot be treated with currently available medicines. Therapies for patients with these genotypes remain a critical unmet need.

Substantial efforts in high-throughput screening (HTS) for small molecules that can mediate readthrough of PTCs have not yet been able to identify compounds that promote functional expression of CFTR PTC variants at levels sufficient to expect clinical benefit. In prior HTS campaigns typically a PTC-containing reporter gene was heterologously expressed from cDNA under a CMV promotor in a model cell line. While these HTS assays were able to identify compounds that increased expression of full length functional (reporter) protein in these model systems, hits did not convincingly translate to primary patient cells that express the native form of CFTR PTC alleles.

To facilitate hit validation and translation to more relevant in vitro systems, we have developed new models of native CFTR PTC variants (e.g. G542X, W1282X, and Y122X) and a label-free method to detect full length CFTR as a readthrough product. Employing CRISPR/Cas9 gene-editing, a series of cell lines were generated based on the 16HBE14o-cells*. The models resemble primary human bronchial epithelial (hBE) cells with respect to low CFTR expression, reduced mRNA of PTC variants, and functional expression of CFTR. Inhibition of nonsense-mediated mRNA decay (NMD) by inhibiting SMG1 rescues the mRNA levels of CFTR G542X and W1282X. The readthrough modulator G418 leads to functional expression of CFTR in the G542X model. In parallel, an ELISA assay has been developed with a C-terminal capture antibody, thus capable of detecting readthrough, i.e., small increases in expression of full length CFTR protein.

Further validation of hits from internal HTS campaigns (similar to approaches described above) was carried out with respect to CFTR function and full length CFTR protein. The functional competence of a readthrough product was assessed in electrophysiological assays. In case of efficient readthrough, a reduction of the truncated CFTR protein on Western blots may also be indicative of readthrough. Additionally, CFTR mRNA quantitation was performed to identify hits that can suppress NMD. As proof of concept, we have demonstrated functional CFTR in the 16HBEge CFTR G542X cell line after treatment with a hit compound. Ongoing efforts involve streamlining the hit validation process in native models for efficient identification of promising HTS hits.

* Introduction: A prior Phase 3 study (NCT02279498, the "SOLUTION" study) failed to demonstrate noninferiority of liprotamase, a nonporcine pancreatic enzyme replacement therapy (PERT) containing biotechnology-derived lipase, protease, and amylase without enteric coating, compared with porcine pancrelipase in subjects with cystic fibrosis (CF).

Objective: To determine if higher doses of liprotamase than were used in SOLUTION result in better performance with respect to coefficient of fat absorption (CFA) and coefficient of nitrogen absorption (CNA).

Methods: The RESULT study (NCT03051490) was a Phase 3, randomized, open-label, non-inferiority, active-comparator study conducted in subjects aged ≥7 years with CF-related exocrine pancreatic insufficienvy (EPI) and baseline CFA of ≥80% in the presence of stable porcine PERT. Subjects were randomized 1:1 to receive liprotamase or pancrelipase (Pancreaze®), with the pancrelipase dose initially matched to the pre-study PERT dose in lipase units, and the liprotamase dose at least 25% higher. Based on clinical signs and symptoms of malabsorption, and if sanctioned by a blinded assessor, subjects were allowed dose adjustments during the first 3 weeks up to 10,000 units lipase/kg/day for those under 12 years, and up to 15,000 units lipase/kg/day for those 12 years and older. CFA and CNA were evaluated at Week 4. The primary endpoint was the between group difference in least square (LS) mean change from baseline in CFA, with a noninferiority margin of -15% for the lower bound of the 95% confidence interval (CI).

Results: The study enrolled 138 subjects, with a mean age of 22.1 years, mean baseline CFA of 88.3%. Liprotamase missed the noninferiority criterion for CFA (Week 4, LS mean of 72.3% for liprotamase and 87.9% for pancrelipase, lower CI for treatment difference = -18.2). Liprotamase met the noninferiority criterion for CNA (adjusted means = 95.6% vs 97.4%, lower CI for treatment difference = -2.1). Dose adjustments were more common in the liprotamase arm (58.6% vs 11.7%). Signs of malabsorption were modestly higher in the liprotamase arm.

The most commonly-reported adverse events (AEs) were infective pulmonary exacerbations of CF (4.3%, liprotamase vs 7.4%, pancrelipase). More subjects randomized to liprotamase discontinued from study compared with pancrelipase (11.4% vs 5.7%). Serious AEs were balanced (1.4% for liprotamase vs 2.9% for pancrelipase) and were all deemed related to CF.

Conclusions: Despite a higher dose of liprotamase in the RESULT study, liprotamase was inferior to porcine pancrelipase for CFA, but not for CNA. Safety was similar in both groups. The differences in CFA, dose adjustments and withdrawals suggest that optimal lipid absorption in patients with CF-related EPI treated with liprotamase remains to be achieved. Objective: Early stages of deficient pulmonary physiology are hard to monitor in CF without invasive procedures or radiation. Our method, 3D single breath hold chemical shift imaging (3D-SBCSI), is sensitive to early signs of disease progression using MRI and hyperpolarized Xenon-129 gas, which is inert and nonradioactive.

Method: Using a 1.5T MRI scanner and vest coil tuned to the Xe-129 frequency, 19 healthy, 12 mild CF, and 5 severe CF subjects (N=36) were imaged. Subjects lay supine on the MR table, inhaled a volume (1/3 FVC) of enriched Xe-129 and N 2 , and held their breath for the ~7s acquisition. Xe-129 was polarized to ~30%. Ventilation, 3D-SBCSI, and proton images were post-processed (Tustison N, et al. J Magn Reson Imaging. 2011; 34:831-41 ) and 3 peaks in the Xe-129 spectrum were identified, corresponding to alveolar gas, tissue, and RBC (Mata J. ISMRM. 2010, Stockholm; Mata J. ISMRM. 2009, Honolulu; Qing K, et al. J Magn Reson Imaging. 2014; 39:346-59) . Peaks were analyzed on a voxel-by-voxel basis, and results compared to PFT and blood panel results.

Results: Severe CF subjects had a larger volume of ventilation defects (VD) in their lungs (49.19±3.919%) than mild CF (34.04±14.423%) and healthy subjects (10.46±7.411%) (Fig A) . Severe CF subjects had an average Tissue/RBC peak ratio of 3.26±0.754AU, mild CF 3.04±0.637AU, and healthy 2.63±0.442AU (p<0.05) (Fig B-C) . This correlated well with VD and showed local V/Q mismatches. Tissue/RBC, RBC/Gas, and Tissue/ Gas ratios correlated well with iron concentration and transferrin saturation in blood (|R|>0.70). Two severe CF subjects (S1, S2) had similar FEV1 but different iron levels; with lower iron, Tissue/RBC was higher, Tissue/ Gas was about the same, and RBC/Gas was lower ( Fig D) . The RBC peak shifted more as disease progressed and iron level decreased, indicating a reduced capacity to bind oxygen (Kaushik SS, et al. J Appl Physiol. 2014; 117:577-85; Norquay G, et al. Magn Reson Med. 2017; 77:1399-408) . RBC T2* relaxation was shorter in severe disease which may be linked to acidosis (Schilling AM, et al. Neuroradiology. 2002; 44:968-72) . 3D-SBCSI parameters indicate underlying impairment in gas-exchange and reveal specific physiologic evidence of CF progression on a voxelby-voxel basis.

Conclusion: 3D-SBCSI can detect early physiologic changes in CF that could be used to monitor treatment responses and disease progression before symptoms are clinically apparent. Objective: To determine if a blood transcriptional gene signature could be used to predict risk of exacerbations in the next 12 months.

Methods: CF participants (n=21) were recruited when stable at annual review and followed for 12 months. Peripheral blood RNA was collected into Paxgene RNA tubes. RNA was extracted using the Paxgene Blood RNA Kit (Qiagen) and RNA quality and quantity was assessed using the Bioanalyser (Agilent Technologies). Transcriptional profiles were generated (Illumina HumanRef-8 V4) and analysed using GeneSpring GX14.8.

Results: Participants had mean age 33 years (SD ±13.5y), mean FEV1 65% predicted (SD ±22%).

Unsupervised hierarchical clustering of gene expression profiles revealed 2 distinct clusters. Cluster 2 (n=16) was characterised by significantly lower FEV1 percent predicted (p=0.036), higher residual volume percent predicted (p=0.040), lower BMI (p=0.042), more frequent exacerbations requiring IV antibiotics (p=0.012), and increased presence of Pseudomonas (p=0.011), compared with cluster 1 (n=5). The more severe phenotype in cluster 2 was associated with a reduction in the expression of 119 entities. The most significant downregulation was seen in 4 immune regulatory genes; AIRE, IL-10, IL-17RD, and NLRP8 (p<0.0001).

Conclusions: Transcriptional profile of peripheral blood demonstrating a down-regulation in immunoregulatory genes can predict future risk of exacerbations in the next 12 months. Zwitserloot, A. 6 ; Ratjen, F. 2 1. Royal Brompton Hospital, London, United Kingdom; 2. The Hospital for Sick Children, Toronto, ON, Canada; 3. Marienhospital Wesel, Wesel, Germany; 4. Inselspital, Bern, Switzerland; 5. University Children's Hospital Zurich, Zurich, Switzerland; 6. University Hospital Groningen, Groningen, Netherlands Introduction: The ndd EasyOne Pro ® LAB is an FDA-approved multiple breath nitrogen washout (MBW) device. Previous work to validate the system for use in children by the investigators of this study has led to software changes. This study aims to validate the updated EasyOne Pro ® LAB in a pediatric cohort.

Methods: Healthy children and age-matched subjects with cystic fibrosis (CF) performed plethysmography and MBW on the EasyOne Pro ® LAB device (MBW ndd ; ndd Medical Technologies, Zurich, Switzerland (software v. 2.1.4.5) ) at 2 separate clinically stable visits (range 1 to 4 months apart). A subset of these subjects also performed a MBW test in random order on the Exhalyzer ® D device (MBW EM ; EcoMedics AG, Duernten, Switzerland (Spiroware software v.3.1.6)). MBW tests were reviewed for test quality to meet ATS/ERS acceptability criteria (Eur Respir J. 2013; 41:507-22) ; MBW ndd tests were reviewed using an updated software version (3.2.0.6) to allow for closer inspection of gas traces.

Results: In total, 45 healthy subjects and 41 CF subjects (mean (range) ages 12.3 y (5-17.8)) were tested. 83% of subjects had a successful MBW ndd measurement; main reasons for unsuccessful tests were irregular breathing pattern and leaks. The within-test coefficient of variation in health was within 5% for lung clearance index (LCI) and functional residual capacity (FRC), but slightly higher for LCI in CF subjects (8.8%) . Repeated LCI measurements between visits were reproducible; mean relative difference (95% limits) between visits was -0.9% (-14.8, 13 .0) in health and 2.0% (-23.8, 27.8) in CF. Compared to plethysmography, FRC ndd was consistently lower (mean relative difference (95% CI) was -22.4% (-25.6, -19.2) in health and -27.1% (-30.7, -23.5) in CF) and showed a size dependency with the largest differences observed in children with the smallest FRC.

The average LCI ndd in health was 7.4 (SD 0.6), and showed a size dependency; the smallest children had higher LCI values. The relationship between LCI and size was not observed on the subset of subjects that performed MBW EM (30 health and 27 CF). Subjects had a higher respiratory rate and minute ventilation on MBW ndd compared to MBW EM with a mean difference (95% CI) of 6.2 (5.3, 7. 2) and 2.3 (1.7, 2.8) , respectively. When LCI ndd was recalculated to include dead space to the airway opening matching the ATS/ERS recommendations, the relationship with size was no longer observed, but differences between devices persisted for LCI (Δ(MBW ndd -MBW EM ); (95% CI) was -0.5 (-0.7, -0.3) in health and -1.6 (-2.1, -1.1) in CF.

Conclusions: MBW testing on the EasyOne Pro ® LAB is feasible, repeatable, and reproducible. Both FRC and LCI showed a size dependency which was partially explained by device specific algorithms to calculate outcomes at different geometric points. However, differences between devices were still observed likely reflecting differences in breathing pattern and analysis algorithm. The impact on longitudinal LCI measurements needs to be further explored. Bacterial specific imaging agents have the potential to distinguish infection from other etiologies such as sterile inflammation or cancer. For patients with cystic fibrosis (CF), a bacterial imaging agent could identify the location of the highest bacterial burden and additionally may allow monitoring of antimicrobial treatment efficacy. Here, we report the first-inhuman imaging of carbon-11 trimethoprim, [ 11 C]TMP, a positron emission tomography (PET) radiotracer with attention to the overall biodistribution and uptake in the lungs of acutely infected CF patients.

We recently reported the development and radiosynthesis of PET radiotracers based on the small molecule antibiotic trimethoprim (TMP; Sellmyer MA, et al. Mol Ther. 2017; 25:120-6) . These radiotracers could distinguish live bacteria in rodent models and human dosimetry was extrapolated. [ 11 C]TMP was approved for a human protocol by the University of Pennsylvania IRB and Radioactive Drug Research Committee (RDRC). Inclusion criteria for the study were patients with suspected bacterial infection who were at least 18 years of age. Exclusion criteria for the study were antibiotic therapy with TMP within 48 hours of the study, unstable clinical condition, and pregnant or nursing women. We report on the first 2 patients imaged who participated in the study.

One female (P1) and one male patient (P2) were imaged during CF exacerbation. Both patients showed prompt biodistribution of [ 11 C]TMP throughout the body and rapid removal of the tracer from the blood pool during 60-minute dynamic scanning of the lungs. Renal excretion and hepatic metabolism were evident on whole body imaging at 1 hour. Muscle, lung, brain, and peripheral bone uptake were quite low by 60 minutes after infection. P1, imaged as an outpatient, grew few P. aeruginosa and isolated Aspergillus in her sputum. She was on-treatment with multiple antimicrobials at the time of imaging. Several areas of mucus impaction demonstrated increased uptake above background, showing target to muscle ratios of 1.3-2.0 and some areas of CT abnormality did not correlate with [ 11 C]TMP uptake. P2 was imaged as an inpatient, grew moderate E. coli and many H. influenzae, and was treated with IV antibiotics. He was imaged with [ 11 C]TMP on therapy day 2 and day 9 and showed clinical and radiographic improvement. [ 11 C]TMP uptake decreased in some areas of mucus impaction after treatment including two areas with a 33% decrease in SUVmax.

The ability to monitor an infection in vivo would be a major advance for biomedical imaging. This first-in-human imaging with [ 11 C]TMP in CF patients is an important first step toward an imaging readout of antimicrobial therapy. The reproducibility of the lung clearance index (LCI) was previously described for preschool children (2-5 years of age), suggesting that a physiologically relevant change in LCI is ± 15% (Oude Engberink E, et al. Eur Respir J. 2017; 50:1700433) . It is unknown whether this observed reproducibility is maintained throughout school age. The objective of this study was to compare the reproducibility of the LCI in school age children to that observed in the preschool age range.

Methods: A cohort of healthy children and children with cystic fibrosis (CF), who were previously part of a longitudinal study to measure LCI during preschool age (Stanojevic S, et al. Am J Respir Crit Care Med. 2016; 195:1216-25) , were recruited into this ongoing prospective observational study. LCI was measured by multiple breath nitrogen washout (MBWN 2 ; Exhalyzer ® D, Duernten Switzerland) every three months. Data collected during nonsymptomatic visits, defined as no cough or pulmonary exacerbation, were used to calculate measures of LCI reproducibility between visits. The assumption that the variability is independent of the magnitude of the test was tested using a Bland-Altman plot, and the percent change in LCI was calculated as (LCI 2 -LCI 1 )/LCI 1 .

Results: In this preliminary analysis, 132 LCI measurements were available in 26 healthy children and 27 children with CF. The mean (range) age at first visit was 7.3 (5.0-8.7) years in healthy children and 7.6 (5.2-10.1) years in children with CF. Mean (SD) LCI at first nonsymptomatic visit was 6.8 (0.5) in healthy children and 8.3 (1.5) in children with CF. Similar to what was observed in preschool years, the variability of the LCI increased with the magnitude of LCI. The reproducibility limits were also similar to those observed in preschool children; in healthy children, the mean percent change between visits was -1.3% with 95% limits of agreement of -14 to 12%., whereas in children with CF, the mean percent change was -1.2 (95% limits -18, 16%).

Conclusions: The reproducibility of LCI at school age is similar to previously reported limits observed during preschool age in both healthy school age children and children with CF. These data suggest that ± 15 % change in LCI represents a clinically meaningful change and may be applicable across the entire pediatric age range. Ivacaftor is a CFTR potentiator first approved for use in cystic fibrosis (CF) patients with at least one copy of the G551D CFTR mutation. While ivacaftor is intended for lifelong use, little is known about the long term clinical effects of this therapy in G551D patients.

We conducted a five-year observational extension to the GOAL study involving 35 centers in the US CF Foundation Therapeutics Development Network. CF participants ages 6 years and above with at least one copy of G551D mutation were followed. Evaluations of spirometry, nutritional status, sweat chloride (SC), and CFQ-R were monitored over 5.5 years after ivacaftor initiation (baseline, 1 month, 3 months, 6 months, and annually until 5.5 years).

Results: Of the 151 GOAL participants, 96 enrolled into the 5-year extension (n=52 <18 years of age), with 78 (81%) completing follow-up. Forty-three (45%) were female with mean age of 19.8 years and mean baseline ppFEV 1 of 82%. Absolute change in ppFEV1 was greatest at 6 months (+7.9% [5.8, 10 .1], p<0.0001), while the absolute change at 5.5 years was not statistically different (+0.8% [-2.0, 3.6] ). BMI improved steadily, with maximum absolute change in BMI observed at 5.5 years (+2.5 kg/m 2 [2.0, 3.1] , p<0.0001). CFQR Respiratory domain improved 8.8 ([4.8,12.8] , p<0.0001) at 6 months and 6.7 ([2.5, 10.9] , p=0.002) at 5.5 years. SC had a sustained reduction at 5.5 years of -49.5 mEq/L ([-55.0,-44.1], p<0.0001).

Outcomes differed among the pediatric (mean 11.6 years) and adult (mean 29.5 years) cohorts. The mean baseline ppFEV1 was 67% in adults and 94.7% in children. The absolute change from baseline at 5.5 years was +4.3% ([0.6,8.1] , p=0.0237) in adults and -2.0% by 5.5 years ([-5.9,2 .0], p=0.3228) in children. Baseline BMI was 23.4 kg/m 2 in adults and 17.9 kg/m 2 in children with an absolute increase at 5.5 years of 1.2 ([0.4,2 .0], p=0.003) in adults and 3.6 ( [2.9,4.3] , p<0.0001) in children. The mean change in sweat chloride was -52.4 mEq/L [-60.6,-44.3] in adults and -47.3 mEq/L [-54.9,-39.8] in children (p<0.0001 for both).

At 5.5 years, we observed a correlation between change in SC and absolute change in ppFEV1 in adults (0.48 [0.16, 0.71], p=0.0055), but not in children. Conversely, we observed a positive correlation between change in SC and improvement in weight only in the pediatric group (0.32 [0.02,0.57], p=0.0368).

Conclusions: Clinical improvements were observed over 5.5 years of ivacaftor use in CF patients with a G551D allele. Spirometry remained above baseline in adults, although waned in children, potentially due to a ceiling effect or propensity for decline in children with high lung function. Nutritional status improved in both groups, to a greater extent in children. SC changes significantly correlated with changes in FEV in adults likely reflecting more precise correlations with longer time domains. These results demonstrate long-term effectiveness of ivacaftor monotherapy in this population.

Acknowledgments: On behalf of the GOAL-e2 investigators. Supported by CFFT.

A QUANTITATIVE ASSAY FOR CFTR MODULATORS Guimbellot, J.S. 1 ; Ryan, K.J. 2 ; Anderson, J.D. 1 ; Liu, Z. 1 ; Kersh, L. 1 ; Rowe, S.M. 3 ; Acosta, E.P. 2 1. Pediatrics, Univ. of Alabama at Birmingham, Birmingham, AL, USA; 2. Pharmacology and Toxicology, Univ. of Alabama at Birmingham, Birmingham, AL, USA; 3. Medicine, Univ. of Alabama at Birmingham, Birmingham, AL, USA Introduction: Modulators target the underlying defect in mutant CFTR proteins in a concentration-dependent fashion. Ivacaftor (iva), lumacaftor (luma), and tezacaftor are substrates of cytochrome P450 enzymes. Inhibitors and inducers of P450 affect concentrations and efficacy of many drugs in plasma and target tissues. To understand the effect of metabolism of CFTR modulators on the variability of drug efficacy among the CF population, techniques to quantitate CFTR modulators in plasma and peripheral tissues are essential.

Methods: We developed quantitative LC-MS/MS methods to detect iva, iva metabolites M1-iva and M6-iva, and luma in plasma and wholecell lysates of nasal cells from patients. Subjects donated plasma or nasal epithelial cells (HNE) at random time points for a proof-of-concept study. Iva, M1-iva, M6-iva, and luma were quantified in plasma or HNE after whole cell lysis upon collection (i.e. intracellular drug concentrations). Concentrations of compounds in whole-cell lysate were normalized to volume of HNE.

Results: The assay had a lower limit of detection of 1 ng/mL from 10 µL plasma or whole cell lysate. Intra-and inter-day accuracy and precision were ≥94.7% accuracy with ≤5.5% coefficient of variation (CV) and ≥99.6% accuracy with ≤7.6% CV respectively. Plasma concentrations (PC) of iva varied depending on the drug taken and time of ingestion. In patients taking iva/luma, the PC of iva was quite variable (29.6-256 ng/ mL in samples taken from 1-14 hours post-ingestion), as was that of luma (1910-4950 ng/mL in samples taken from 1-14 hours). HNE intracellular concentration (IC) of iva in two patients who had both taken the drug 14 hours prior were 1.96 and 6.86 ng/mL, while simultaneously obtained PCs were 29.6 and 34.3 ng/mL, resulting in plasma:intracellular (P:I) ratios of 15.1:1 and 5.0:1. M1-iva was detected in plasma and lysate. Despite reliable detection in plasma, luma and M6-iva were not detected in any HNE whole cell lysate. In patients taking iva monotherapy, the PC of iva ranged from 500-2780 ng/mL 2.5-6.5 hours post-ingestion. The HNE IC of iva were 47.9-3370 ng/mL; P:I ratios were 0.21-59.2:1. M1-iva was detected in 2/3 patients taking iva. M6-iva was not measurable in any HNE whole cell lysate, suggesting that this metabolite is produced systemically but is not present in airway epithelia in appreciable amounts.

Conclusions: We developed a method for sensitive and accurate determination of CFTR modulator levels in plasma and tissue. The assay distinguishes parent compounds from metabolites, allowing for accurate determination of metabolism enzyme activity in target tissues of interest and quantitative detection of pharmacologically active parent compounds and metabolites. Assay development for M28-luma, tezacaftor, and tezacaftor metabolites is ongoing. These methodologies will allow for stratifying clinical responses based on therapeutic drug monitoring in plasma and respiratory cells, and ultimately may contribute to the adjustment of treatment strategies to optimize efficacy given the variability we identified in both plasma concentrations and P:I ratios.

Acknowledgment: Supported by CFF. Background: Impairment of the CFTR chloride channel activity observed in cystic fibrosis (CF) patients is the main cause of the deterioration of lung function. In 2008, the chloride channel ANO1 was identified and proposed as a potential therapeutic target in CF. Recently, we have shown that ANO1 activity and expression were reduced in a CF context (Ruffin M, et al. Biochim Biophys Acta. 2013; 1832 :2340 and propose a new therapeutic approach with antisense oligonucleotide blocking a target site (TSB) of miRNA on 3' UTR of ANO1 (Sonneville F, et al. Nat Comm. 2017; 8:710) .

The aim of this work is to analyze the effects of this TSB on primary F508del cells but also on cells with other mutations compared to Vertex drugs.

Methods: We used primary cells from CF patients cultured in air-liquid interface. Cells were obtained from patients with F508del mutation as a control or with other mutations from class I or II. Cells were treated with TSB, lumacaftor/ivacaftor (Orkambi ® ), or tezacaftor during one week and mucociliary clearance and chloride efflux were analyzed after.

Results: Previously, we identified miR-9 as a regulator of ANO1, and we have done experiments using an antisense oligonucleotide designed to specifically link ANO1 3'UTR by masking miR-9 fixation site (TSB ANO1 ; patent PCT/FR2015/051850). We have observed an increase of ANO1 expression, ANO1 chloride activity, migration rate of cells and mucociliary clearance using the TSB ANO1 in CF primary human epithelial bronchial cells cultured in an air-liquid interface but also in CF mice all with F508del mutation. More, we have demonstrated that ANO1 is more efficient than lumacaftor/ivacaftor or tezacaftor.

To increase the proof of concept, we have performed the same experiment on other mutations than F508del. In primary cells with class I or II, we have demonstrated that ANO1 TSB is able to activate total chloride efflux and mucociliary clearance by contrast to Vertex drugs.

Conclusions: Our results demonstrate that our TSB approach is able to correct the main parameters dysregulated in CF for all the mutations even for class I. Based on previous results and on these new results, we propose ANO1 TSB as a candidate drug to treat all CF patients. Objective: To understand whether changes in the L-arginine/nitric oxide (NO) metabolism or airway inflammation can explain the increase in FeNO with ivacaftor therapy.

Methods: This observational trial was approved by the Research Ethics Boards of SickKids and St. Michael's Hospital in Toronto. Pulmonary function (spirometry) and FeNO were measured and sputum samples collected before and for two years (1, 3, 6, 12, 24 months) after initiation of ivacaftor therapy in children and adults with CF and a CFTR gating mutation. Sputum samples were stored at -80°C before quantification; liquid chromatography-mass spectrometry (LC-MS/MS) was used for quantification of L-arginine, L-ornithine and asymmetric dimethylarginine (ADMA), the Griess reagent to detect NO metabolites nitrate and nitrite and an ELISA for cytokine measurements.

Results: Twenty patients were included, 7 pediatric and 13 adults. Mean (IQR) age was 22.2 (12.5, 34.1) years. Baseline FEV 1 was 80% of predicted (76, 91) in the pediatric and 65% (48, 73) in the adult patients. Ivacaftor therapy resulted in a significant increase in pulmonary function and FeNO at the one-month time point and remained increased over the 2-year follow-up. L-arginine (substrate for NO synthases; NOS), and ADMA (NOS inhibitor) levels in sputum showed a transient decrease after 4 weeks but the L-arginine:ADMA ratio, an index of NOS impairment, previously demonstrated to be reduced in CF airways, remained unchanged. Similarly, sputum L-ornithine (product for arginase activity) was decreased initially following ivacaftor initiation; however, the ratio of L-arginine:L-ornithine, used as an index of L-arginine availability for NOS, remained unchanged throughout the observation period. NO metabolites in sputum did not change over time. When including all measurements, FeNO correlated with FVC (r=0.29, p=0.01), FEV 1 (r=0.41, p<0.001) and FEF50 (r=0.76, p<0.001), but changes in FeNO did not correlate with changes in PFTs over time. There was no significant change in any of the 42 different cytokines measured in sputum, however there were significant reverse correlations between changes in FeNO and the changes in some of the cytokines including IL-1α (-0.47, p=0.03) and IFNγ (-0.57, p=0.005), which are both known to induce NOS.

Ivacaftor therapy resulted in a significant and sustained increase in pulmonary function and FeNO in treated CF patients. The increase in FeNO could not be explained by changes in L-arginine bioavailability for NOS but may be related to changes in airway inflammation. Acknowledgment: The study was supported by Vertex Pharmaceuticals. Introduction: MRI in nonsedated pediatric patients may provide a more thorough evaluation of cystic fibrosis liver disease than ultrasound (US) alone due to techniques that evaluate the liver for biliary and diffuse liver disease in a single exam.

Methods: We recruited 17 pediatric subjects with known CF (6:11 M:F, 5-16 years) to undergo MRI and US at two study visits separated by 1-2 weeks. Liver grayscale US and US elastography (USE) were acquired on a Supersonic scanner (Aixplorer, Supersonic Image) with a 2D probe. Ten USE measurements were made in the right hepatic lobe within 5 cm of the skin and averaged. We performed MRI on a 3.0T clinical system (Discovery MR750, GE Healthcare) using a 32-channel cardiac coil. Acquisitions included breath-held and free-breathing chemical shift encoded MRI (CSE-MRI) to measure proton density fat fraction (PDFF), MR Elastography (MRE), biliary imaging (MRCP), and diffusion-weighted imaging. All MRI acquisitions were performed twice per visit to assess same-day repeatability. To assess clinical usefulness, two radiologists independently evaluated MR images, and a third evaluated US images on a 4 point scale: 1: Poor (Non-diagnostic), 2: Fair (Limited diagnostic value), 3: Good (Diagnostic), 4: Excellent (Diagnostic with high degree of confidence). Region of Interest (ROI) measurements were performed on MRE to provide a liver stiffness value in kPa. Image quality scores were evaluated with Wilson binomial confidence intervals. Quantitative measures were compared using Bland-Altman analysis.

Results: 14/17 (82%) subjects completed both visits. All methods produced clinically useful images. Both CSE-MRI methods had good PDFF measurement repeatability, with the novel free-breathing method performing as well as the breath-held method. MRE had good same-day repeatability (LOA= ±0.55 kPa) and intervisit repeatability (LOA= ±1.18 kPa), similar to USE intervisit repeatability (LOA= ±0.99 kPa).

Variability of MRE stiffness values between visits was similar to that of USE with a bias of 0.88 kPa. Normal values of USE liver stiffness have been shown to be below 6.2 kPa (Suh CH et al. Radiology. 2014; 271:895-900) .

Discussion: Comprehensive quantitative liver MRI is feasible in nonsedated pediatric CF patients. MR techniques to measure liver stiffness and fat content, and to assess the biliary system provide information that would otherwise require multimodality imaging.

Bland-Altman plots of stiffness intervisit repeatability for MRE and USE and comparing stiffness values from MRE to USE. Note the significant bias between MRE and USE. Introduction: Functional nuclear imaging biomarkers provide unique insights into cystic fibrosis (CF) lung pathophysiology and are useful for assessing therapeutic efficacy. Our objective is to describe how imaging-based measures of mucociliary clearance (MCC) and small molecule absorption (ABS) reflect differences in CFTR function, depict physiological changes caused by infection, and provide indication of therapeutic response.

Methods: We studied a retrospective CF physiology group (n=22), a prospective CF therapeutic response group (n=12), and healthy controls (n=15). Subjects inhaled radiolabeled particle ( 99m technetium sulfur colloid) and small molecule ( 111 indium-DTPA) probes to measure MCC and ABS. Previous in vitro studies have demonstrated relationships between the ABS measurement and airway surface liquid absoprtion rates. CFTR function was assessed based on genotype, ivacaftor use, and measures of sweat chloride. Pseudomonas aeruginosa (PA) infection history ahead of MCC/ ABS measurements was assessed. Response group subjects performed baseline MCC/ABS and measurements after inhalation of 7% hypertonic saline (HS) and mannitol powder for inhalation.

Results: ABS decreased with increasing CFTR function and was proportional to sweat chloride. MCC function was depressed after PA infection and recovered over 100-200 days. Baseline MCC predicted response to both inhaled HS and mannitol with mannitol providing larger increases in MCC than HS.

Conclusions: Baseline mucus clearance in the CF lung is not universally failed or depressed, at least not over assessment periods of 1-2 hours. PA infection is associated with extended periods of MCC depression. The MCC measurement has a useful role as a biomarker for therapeutic development. The absorption biomarker, ABS, may provide a local measure of ASL absorption and CFTR function in the lung useful for therapeutic development. Background: Pulmonary magnetic resonance imaging using hyperpolarized 129 xenon gas (XeMRI) can quantify ventilation inhomogeneity by measuring the percentage of unventilated lung volume (ventilation defect percent; VDP) in stable pediatric CF patients (1) . The utility of XeMRI to monitor response to therapy in CF has never been assessed. The aim of this study was to assess the utility of XeMRI to capture treatment response in pediatric CF patients undergoing inpatient treatment for a pulmonary exacerbation (PEx).

Methods: CF patients aged 8-18 who were admitted to hospital for inpatient treatment of a PEx were recruited. Exclusion criteria were supplemental oxygen use, forced expiratory volume in one second (FEV 1 ) less than 40% predicted and MRI contraindications. XeMRI (Prisma 3T MRI, Siemens), spirometry, plethysmography, and multiple breath nitrogen washout (Exhalyzer D, EcoMedics) to measure lung clearance index (LCI) were performed within 48 hours of admission and discharge. VDP outcomes were generated from XeMRI scans using k-means (2) and histogram (3) techniques.

Results: 15 participants were included in the study. 10/15 (66.7%) were female, median age was 14 years (IQR 13-16.5) and median baseline (best in 6 months prior to admission) FEV 1 was 85.0% (IQR 61.0 93.0). Xe-MRI was well tolerated on all test occasions with median oxygen saturation nadir of all scans being 91% (IQR 87-94). VDP, LCI and FEV 1 all improved with treatment (see Table) . VDP (using the histogram technique) showed the largest relative improvement of all outcome measures .6], p<0.001). There was no correlation between the magnitude of change of VDP and that of LCI or FEV 1 (p>0.05).

Discussion: XeMRI is well tolerated in pediatric CF patients, even those who are acutely unwell with a PEx. VDP, a measure of ventilation inhomogeneity derived from XeMRI, showed the largest relative improvement following treatment for a PEx. These data support the further investigation of XeMRI as a tool to monitor CF lung disease. 

Mean physiologic and imaging outcome measures before and after treatment. Mean within-subject absolute and relative change in these measures following treatment is also shown. Data are shown as mean value (SD) or mean within-subject change (95%CI).

* denotes a mean within-subject change that is significantly different from 0 (p<0.05)

Gräber, S.Y. 1, 2, 3 ; Dopfer, C. 4, 5 ; Naehrlich, L. 6,7 ; Gyulumyan, L. 4 The combination of the CFTR corrector lumacaftor with the potentiator ivacaftor has recently been approved for the treatment of patients with cystic fibrosis (CF) homozygous for the Phe508del CFTR mutation. The pivotal phase 3 trials examined clinical outcomes, but did not evaluate CFTR function in patients. Hence, we wanted to examine the effect of lumacaftor-ivacaftor on biomarkers of CFTR function in Phe508del homozygous CF patients aged 12 years and older.

This prospective observational study assessed clinical outcomes including FEV1 percent predicted and BMI, and CFTR biomarkers including sweat chloride concentration, nasal potential difference (NPD) and intestinal current measurement (ICM) before and 8-16 weeks after initiation of lumacaftor-ivacaftor therapy.

A total of 53 patients were enrolled in the study and 52 patients had baseline and follow-up measurements. After initiation of lumacaftor-ivacaftor, sweat chloride concentrations were reduced by 18 mmol/L. Further, NPD and ICM showed partial rescue of CFTR function in nasal and rectal epithelia to levels of 10% and 18% of normal, respectively. All patients improved in at least one CFTR biomarker, but no correlations were found between CFTR biomarker responses and clinical outcomes.

We could show that lumacaftor-ivacaftor results in partial rescue of Phe508del CFTR function to levels comparable to the lower range of CFTR activity found in patients with residual function mutations. Functional improvement was detected even in the absence of short-term improvement of FEV1 percent predicted and BMI. Established CF lung infections cannot generally be eradicated, even after the most responsive subjects (those with the CFTR-G551D mutation) receive the most effective CFTR-correcting drug (ivacaftor). Our previous work shows that although ivacaftor treatment initially reduced sputum Pseudomonas aeruginosa burden, counts rebounded in the second year of treatment. Thus, the full clinical benefit of CFTR-correcting drugs might require additional interventions that eliminate lung infection.

We performed a pilot study to attempt infection eradication by combining ivacaftor treatment with intensive antibiotics. We enrolled 12 adult subjects with the CFTR-R117H mutation and chronic lung infections caused by P. aeruginosa (4 subjects), Staphylococcus aureus (6 subjects), or both organisms (2 subjects). Subjects were initially treated with ivacaftor alone for 7 days as our previous work in people with CF and the CFTR-G551D mutation found that a week of ivacaftor significantly reduced sputum P. aeruginosa burden, and such reductions could increase antibiotic activity. P. aeruginosa-infected subjects then received two IV antipseudomonal antibiotics for two weeks, followed by three months of oral ciprofloxacin and inhaled colomycin; S. aureus-infected subjects received oral flucloxacillin for three months (these regimens are called "prolonged antibiotic" treatment, below). We used paired t-tests and Wilcoxon sign-rank tests where appropriate to compare pre-and post-treatment outcome measures.

Treatment with ivacaftor alone for one week increased average percent predicted FEV1 by 7.6% (p = 0.0017), decreased average sweat chloride by 27 mmol/L (p = 0.0001), and reduced average sputum S. aureus and P. aeruginosa CFUs by ~10-fold. During prolonged antibiotic treatment CFUs were further reduced, however one month after completing antibiotic treatment, average P. aeruginosa and S. aureus counts rebounded to pre-antibiotic levels and remained relatively constant (~10-fold lower than pre-ivacaftor levels) over the one-year follow-up period. Studies using a population-level multilocus sequence typing method are in progress to determine whether preexisting P. aeruginosa and S. aureus strains persisted or new infections developed.

Examination of individual subject data showed that one of six S. aureus-infected patients stopped producing sputum after receiving ivacaftor and prolonged antibiotic treatment, raising the possibility of eradication. To explore this further, we are using molecular-and culture-based analysis of throat swabs obtained before and after treatment. Additional work is needed to determine if combining existing or novel anti-infective approaches with highly effective CFTR modulators can eradicate chronic CF lung infections in some subjects.

Supported by an investigator-initiated award from Vertex Pharmaceuticals Incorporated; and by the Cystic Fibrosis Foundation. (EudraCT Trial Number: 2016-001785-29.) 

Iron is an essential element in human cells due to its redox-cycling capacities. Iron levels must be tightly regulated to avoid free iron toxicity via catalysis of reactive oxygen species (ROS) by host cells. ROS activate intracellular redox-sensitive inflammatory pathways and subsequent generation of inflammatory cytokines. Importantly, it has been documented that iron homeostasis is markedly dysregulated in the lungs of CF patients (Ghio A, et al., J Cyst Fibros. 2013; 12:390-8) . To better understand the impact of iron on the inflammatory response of CF airway epithelial cells, DIBI, a novel, highly-specific synthetic iron chelator was used in our in vitro study. We hypothesized, that iron chelation by DIBI could reduce the inflammatory response of CF epithelial cells triggered by lipopolysaccharide (LPS), a component of gram-negative bacteria cell walls. JME/CF-15 cells, a nasal epithelial cell line derived from a F508del homozygous CF patient, were grown on Transwells with differentiation media. To establish the inflammatory response, once polarized, cells were stimulated with LPS under various conditions. Incubating the cells with 200 ng/mL of LPS for 24 hours resulted in significant apical IL-6 secretion of JME/CF-15 cells (measured by ELISA of cell culture supernatant). DIBI administration (100 or 200 µM) significantly reduced LPS-induced IL-6 secretion by 25% or 69%, respectively. With Bradford assays, we verified similar protein content in all experimental groups (control, DIBI, LPS, LPS+DIBI). JME/CF15 cell response to LPS was compared to the response of a standard bronchial epithelial cell line, Calu-3. We found that much higher doses of LPS (5 -10 µg/mL) were necessary to illicit a comparable inflammatory response with Calu-3 cells (measured by increase in IL-6 secretion) compared to JME/ CF15 cells. In conclusion, our results demonstrate that polarized JME/CF15 cells are more sensitive to LPS than non-CF epithelial cells. This inflammatory response can be modulated by iron sequestration using the novel iron chelator, DIBI. These findings highlight the therapeutic potential of iron chelation as a novel approach for attenuating the dysregulated inflammatory response in CF airways.

SENSITIVITY OF HYPERPOLARIZED 129XE MRI TO EARLY REGIONAL VENTILATION HETEROGENEITY OF CF LUNG DISEASE Walkup, L.L. 1 ; Roach, D. 1 ; Thomen, R.P. 2 ; Cleveland, Z.I. 1 ; Clancy, J.P. 3 ; Woods, J.C. 1 1. Center for Pulmonary Imaging Research, Cincinnati Children's Hospital Medical Center, Cincinnati, OH, USA; 2. Bioengineering, Univ. of Missouri, Columbia, MO, USA; 3. Pulmonary Medicine, Cincinnati Children's Hospital Medical Center, Cincinnati, OH, USA Introduction: As CF clinical care has become more personalized, so too must the means of assessing therapeutic response. Sensitive endpoints beyond FEV 1 especially are needed for clinical trials of specific CF genotypes where patient numbers are small and assessing individual patient's response will have greatest impact. This study aimed to demonstrate the sensitivity of hyperpolarized 129 Xe MRI to lung disease in a sizeable group of CF patients and to compare quantitative 129 Xe MRI to FEV 1 and lung clearance index (LCI 2.5 ).

Methods: Hyperpolarized 129 Xe MRI was completed in 48 CF patients (6-45 years old; FEV 1 38%-131%) and 11 control subjects (6-16 years old; FEV 1 89-115%). 129 Xe gas was polarized using a Polarean polarizer for imaging on a Philips 3T MRI scanner. Ventilation images were acquired during a ≤ 16 second breath-hold of ≤ 1L gas. Ventilation deficits were quantified using a signal-intensity threshold of 60% of the mean whole-lung 129 Xe signal and expressed as a percentage (VDP) of the lung volume. 129 Xe VDP was compared to FEV 1 and LCI 2.5 using linear regression.

Results: While control subjects had uniform ventilation, heterogenous ventilation and deficits with a wide spatial and size distribution were observed in nearly all CF patients. For CF patients, 129 Xe VDP was correlated with FEV 1 with an R 2 =0.41 (p<10 -5 ). LCI 2.5 was available for 25 CF patients; mean LCI 2.5 was 10.1 ± 2.3 and correlated with FEV 1 (R 2 0.31, p<10 -4 ) and with 129 Xe VDP (R 2 0.55, p<10 -4 ). In the subgroup of 36 CF patients with mild disease (FEV 1 ≥ 80%; Figure) , mean FEV 1 (103% ± 12.3%) was not different from controls (p=0.78), however, 129 Xe VDP was elevated (16.0% ± 7.0% vs controls 6.3% ± 2.8%; p<10 -7 ).

Conclusion: Hyperpolarized 129 Xe MRI detected ventilation deficits in CF patients with preserved FEV 1 , supporting the sensitivity of 129 Xe MRI to mild CF lung disease. The correlation between LCI and 129 Xe MRI suggests that both assess similar functional components of CF lung obstruction. 129 Xe MRI has spatial sensitivity which may be leveraged for planned procedures like bronchoscopy or to evaluate regional therapeutic response for individual CF patients. Background: Cystic fibrosis (CF) is an autosomal recessive disorder caused by mutations in the cystic fibrosis transmembrane conductance regulator (CFTR) gene. CF patients have defects in ion transport in epithelial cells. Persistent inflammation and infection in CF airway deteriorate lung function profoundly, leading to CF morbidity and mortality. Traditional medical therapies treat only symptoms but not the cause of the disease. Current CFTR-directed drugs, such as ivacaftor, are effective for only a small portion of patients, and the combination of drugs, such as ivacaftor and lumacaftor, has some efficacy for about half of the CF patients. Thus, novel therapeutic strategies for targeting all types of mutations are needed. Our laboratory has been developing CRISPR/Cas9-based gene targeting strategies for permanently correcting CF lung disease. We produced helperdependent adenoviral (HD-Ad) vectors which can deliver gene editing tools for integrating a functional CFTR gene into a genomic safe harbor of pig cells. The objective of this project is to test our CFTR gene targeting strategy in cultured CFTR-deficient pig cells in vitro as well as pig airway in vivo.

Methods: To test CFTR gene targeting strategy, a CFTR-deficient pig cell line was generated by using CRISPR/Cas9 system to fully ablate the translation of CFTR at start codons. At the same time, GFP and mCherry fluorescence cassettes were "knock-in" at target sites through HDR as selective markers to isolate CFTR -/clones. After generating a CFTR-deficient cell line, HD-Ad vector which contained K18 promoter and human CFTR cDNA was transduced into CFTR -/cells. The function of the hCFTR was assessed by forskolin-induced CFTR channel opening through FLIPR and Ussing chamber assays.

Results/Conclusions: Junctional PCR confirmed that the two CFTR alleles from a selected single cell clone were successfully "knock-out" at the target site. qRT-PCR also did not detect CFTR transcript expression from the CFTR -/clone. Both FLIPR and Ussing chamber assays detected increased hCFTR channel-mediated conductance followed by forskolin stimulation in HD-Ad-hCFTR transduced cells. Meanwhile, the inhibition of hCFTR channel conduction was detected by adding CFTR inhibitor (CFTRinh-172). Our study showed that the hCFTR could be functionally expressed in the pig cell line by HD-Ad vector-based delivery. These results showed the potential of HD-Ad vector mediated hCFTR gene transfer strategy to functionally rescue CFTR mutations. Our next step is to integrate the hCFTR transgene specifically into a pig genomic safe harbor, GGTA1 locus, using CRISPR/Cas9 packaged HD-Ad vector. Introduction: There are approximately 2000 mutations that have been identified in the CFTR gene. Clustered regularly interspaced short palindromic repeat (CRISPR)-CRISPR associated nuclease 9 (Cas9), a gene editing tool, is being explored as a possible therapy for CF. A guide RNA (gRNA) directs the Cas9 nuclease to the genomic target site. The Cas9 nuclease then induces a double-strand break (DSB) in the genome which is then repaired by nonhomologous end joining (NHEJ) or homology directed repair (HDR). HDR uses homologous DNA sequences for repair and can be used to introduce exogenous DNA. Helper-dependent adenoviral (HD-Ad) vectors are the preferred choice for the delivery of the CRISPR-Cas9 system as they have a large carrying capacity and a tropism for epithelial cells of the respiratory tract. In addition, HD-Ad vectors lack most viral genes, which may result in a reduced immune response. Our strategy is to use HDR induced by the CRISPR-Cas9 system to integrate a human CFTR minigene at the human CFTR locus, at the same time preventing expression of the mutated CFTR gene.

Objective: To determine which components are required to construct an HD-Ad vector that will correct CFTR channel function after delivery to cultured CF epithelial cells.

Methods: Three donor plasmids containing a lacZ reporter gene flanked by homology arms specific to the region surrounding the insertion site were designed and constructed. One donor plasmid contains a promoterless lacZ reporter gene, the second contains the human cytokeratin 18 (K18) intron upstream of the lacZ reporter gene and the third contains both the K18 5' enhancer/promoter and K18 intron upstream of the lacZ reporter gene. The donor and CRISPR-Cas9/gRNA plasmids were co-transfected in human embryonic kidney (HEK293) cells. One gRNA targets intron 1 and the other targets exon 1 of the human CFTR gene. Beta-galactosidase (β-gal) expression of the integrated lacZ reporter gene was analyzed. Genome targeting efficiency of the two gRNAs was examined using the T7 endonuclease I (T7E1) assay.

Results: Based on β-gal expression, the integrated expression cassette containing the K18 5′ enhancer/promoter and K18 intron produced higher levels of β-gal expression compared to the promoterless lacZ expression cassette and the expression cassette containing only the K18 intron. The genome targeting efficiency of the two gRNAs was determined.

Conclusions: Using the information gained from these in vitro experiments, the HD-Ad vector will be constructed to test the lacZ reporter and CFTR minigene integration at the CFTR locus. Overall, these studies will contribute to the development of a long-lasting treatment effective for all CF patients.

Acknowledgments: This work is supported by Restracomp and CIHR. Background: Pulmonary exacerbations (PEx) are clinically impactful events for individuals with CF. Unfortunately, many individuals with CF fail to regain their baseline lung function despite treatment (Sanders DB, et al. Am J Respir Crit Care Med. 2010; 182:627-32) . Therefore, prevention of PEx remains a clinical priority. Azithromycin is efficacious in reducing PEx risk in CF children and adolescents without chronic P. aeruginosa infection (Saiman L, et al. JAMA. 2010; 303:1707-15 ). However, it does not benefit all patients and there are potential side effects related to its use. We aim to discover blood protein biomarkers that change early following azithromycin treatment and that are capable of identifying individuals who will derive clinical benefit in terms of lung function improvement and PEx risk reduction.

Method: Blood protein biomarkers were examined in subjects from the AZ0004 study (NCT00431964). Individuals with CF aged 6-18 years were included if their FEV 1 was ≥50% predicted and they were P. aeruginosa culture-negative in the prior year. Oral azithromycin was administered three times weekly for 168 days. We received day 0 and 28 plasma and serum samples from 52 subjects in the azithromycin group. One hundred eight plasma proteins were measured using multiple reaction monitoring mass spectrometry (MRM-MS) and 75 serum proteins with the OLINK Proteomics ProSeek® Multiplex Inflammation I panel. We also analyzed 5 proteins (high-sensitivity C-reactive protein, serum amyloid A, myeloperoxidase, G-CSF, and calprotectin) previously measured with ELISA (Ratjen F, et al. Chest. 2012; 142:1259-66) . We examined baseline (ie, day 0) and early change (ie, day 0 to day 28) in biomarker levels to predict clinical response throughout the trial and thus excluded the 12 subjects who received antibiotics between day 0 and day 28. Blood proteins were identified with statistically significant fold-change from day 0 to day 28. Spearman's correlations between baseline and early change in blood biomarker with changes in percent predicted (pp) FEV1 from day 0 to 28 and day 0 to 168 were examined. The Benjamini-Hochberg procedure was used to control for the false discovery rate (FDR).

Results: Serum levels of CCL20 and CX3CL1 changed significantly from day 0 to day 28 of azithromycin treatment (FDR<10%). Baseline serum CD40 levels correlated with early improvements in ppFEV1 (FDR<10%) but not with changes between day 0 to day 168. Early changes in the levels of 15 plasma proteins significantly correlated with changes in ppFEV1 from day 0 to day 28 (FDR<10%) but not with changes between day 0 to day 168.

Conclusions: Baseline and early changes in blood protein levels following 28 days of azithromycin treatment correlate with short-term but not long-term changes in ppFEV1. We are currently evaluating baseline and early change in blood protein levels to predict individuals with reduced PEx risk and the analysis will be ready prior to October 2018. Acknowledgment: Supported by CFFT (QUON15UO). Introduction: L927P CFTR is a rare missense mutation that has no effective CFTR modulator therapy and is not currently included in modulator trials. Though typically considered a conduction (class IV) mutation, characterization is limited. Here we report in vitro CFTR function and response to modulation from human nasal epithelial cells (HNEs) derived from 2 siblings with the L927P/F508del CFTR genotype. These female siblings (ages 16 and 19 years) were receiving supportive CF therapy including dornase alfa, nebulized hypertonic saline, and inhaled antibiotics. Pancreatic status was insufficient based on fecal elastase testing. Sweat chloride concentrations were >100 mmol/L and maximal FEV1 in the year prior to testing was 94% and 77%, respectively. Over the year prior to testing, both siblings had frequent exacerbations and decline in FEV1. Thus, we sought to determine if CFTR modulator therapy could be beneficial for this mutation.

Methods: HNEs were obtained by nasal curettage and expanded in conditional reprogramming conditions (Brewington JJ, et al. J Cyst Fibros. 2018; 17:26-33) . Once confluent, cells were seeded onto semipermeable membranes (for air:liquid interface [ALI] culture) or into matrigel (3-D spheroid culture). Maturation of ALI and 3-D spheroid cultures was 6 weeks and 2 weeks, respectively. Cells were treated with CFTR correctors (VX-809, 3 µM; VX-661, 1 µM) for 72 hours prior to study. Cells at ALI were studied in modified Ussing chambers under voltage clamp conditions, and short circuit current was monitored with a Clsecretory gradient (basolateral to apical). Cells were treated with amiloride (100 µM, apical), forskolin (10 µM, apical and basolateral) + IBMX (100 µM, apical and basolateral), VX-770 (1 µM, apical), and CFTRinh172 (10 µM, apical). Cells in 3-D culture were studied with a fluid transport swelling assay, monitoring spheroid luminal diameter before and 30 minutes after CFTR stimulation (forskolin/IBMX and VX-770). Continuous data were analyzed with paired t-test.

Results: ALI cultures from the younger sister demonstrated 11.4% of wild-type (wt) CFTR function without modulator treatment (2.4 µA/cm 2 versus wtCFTR normative value of 21.0 µA/cm 2 based on 13 wtCFTR+ subjects), consistent with a diagnosis of CF. This improved to 19.2% of wtCFTR function with VX-770 exposure (4.0 µA/cm 2 ; p=0.008 vs control), and 21.4% (4.5 µA/cm 2 ; p=0.04 for VX-809, p=0.0001 for VX-661 versus control). 3-D spheroid cultures from the older sister confirmed these results, though normative wtCFTR values are not yet established. Spheroids shrank (-15 .4%) at baseline, indicating poor CFTR function. There was a trend towards improvement in sphere swelling with exposure to VX-809/VX-770 (+12.9% from baseline; p=0.15) and a robust improvement following VX-661/VX-770 (+47.9% from baseline; p=0.01). Taken together, these data support a trial of VX-661/VX-770 for both sisters.

Conclusion: Based on the HNE results, both siblings were prescribed tezacaftor/ivacaftor. Clinical outcomes including FEV1, exacerbations, CFRQ, and sweat chloride will be reported at 3 and 6 months. This study shows how theratyping can be used to provide supportive data for patient trials of CFTR modulator therapy.

Jubeau, L. 2 ; Juarez-Perez, V. 1,2 1. Stragen, Lyon, France; 2. Alaxia SAS, Lyon, France ALX-009 is a first-in-class antimicrobial product composed of hypothiocyanite (OSCN -) and bovine lactoferrin. OSCNis a highly reactive compound that oxidizes free thiol residues of bacterial proteins with consequent perturbation of the bacterial physiology and induction of bacterial death. Lactoferrin may act by direct interaction with bacterial cell membranes and/or by depriving bacteria of iron due to its iron chelator activity. In previous report, we have demonstrated that ALX-009 has increased in vitro activity against multidrug-resistant (MDR) bacteria infecting the lung of cystic fibrosis (CF) patients.

Post-antibiotic effect (PAE) is the term generally accepted to describe the suppression of bacterial growth that persists after a short exposure of the bacteria to the test product. The PAE is an important pharmacodynamic parameter that may have an impact when defining the clinical dosing regimen. As the PAE of lactoferrin, OSCNand ALX-009 is unknown, the aim of this study was to investigate the PAE and PAE Sub-MIC Effect (PAE-SME) with 0.1, 0.2 and 0.3X MIC of the ALX-009 compounds alone or in combination against five CF bacterial strains.

The PAE and PAE-SME effects were determined by optical density measurements as defined by Dominguez and coworkers (J Antimicrob Chemother. 2001;47:391) with minor modifications. The PAE ranges from 1.15 hours to 5.25 hours for OSCN -; from 0.17 hours to 5.0 hours for bLF and from 1.39 hours to 3.5 hours for ALX-009. These results are comparable to the PAE values published for aztreonam, colistin and tobramycin. PAE-SME tests showed that when 0.3X MIC of bLF or ALX-009 is added to the cultures, the PAE effect is two-to four-fold increased.

These results provide additional evidence on the interests to combine OSCNand bLF and confirm that ALX-009 could be a therapeutic option for the treatment of cystic fibrosis lung infections. ALX-009 is currently being evaluated in phase I clinical studies. Tagalakis The mucosal layer provides protection to bacteria and viruses, but is also a barrier to nanoparticle therapeutics. Nanoparticles allow delivery of therapeutics such as nucleic acids and small molecule drugs. We have shown delivery of siRNA and plasmid DNA (pDNA) by cationic nanoparticles in multiple in vivo models. Our aim was to enhance the efficiency of transfection in the lung by improving the translocation of nanoparticles across the mucosal barrier.

Methods: Fluorescent-labelled siRNA nanoparticles were prepared with a net negative (anionic) or net positive (cationic) surface charge. Both formulations comprised mixtures of cationic targeting peptides with lipids to modulate the surface charge. Size and charge were analysed by NanoZS Zetasizer. Nanoparticles were applied to the surface of a static mucus barrier comprising a layer of CF or non-CF mucus on a semi-permeable transwell membrane over a buffer chamber. Samples were collected at intervals and fluorescence measured. Rate of accumulation of fluorescence (ng/cm 2 ) in the lower chamber was used to calculate diffusion rates.

Results: Data confirmed that anionic formulations diffuse more rapidly in mucus than cationic formulations. PEGylation did not enhance diffusion of either cationic or anionic formulations. Addition of low molecular weight guluronic-rich alginate oligomer (OligoG) with known mucus modulating effects, had a statistically significant improvement in translocation of cationic formulations: mean cumulative concentration of nanoparticles passing the mucosal barrier was >1100 ng/cm 2 , compared to untreated controls (>700 ng/cm 2 ). Low molecular weight mannuronic-rich alginate oligomers (OligoM) showed even greater diffusion of cationic formulations (1800 ng/cm 2 ; p<0.05). Neither alginate oligomers affected diffusion of anionics. Analysis of size and charge showed OligoM reversed the charge of cationic nanoparticles, with size significantly reduced by both OligoM and OligoG. Electron microscopy analysis confirmed that nanoparticles remained intact after diffusion.

Conclusions: Mucus diffusion was not improved by PEGylation of lipid/ peptide nanoparticles in this study. Alginate oligomers enhance diffusion of cationic, but not anionic formulations. OligoM appears to be more effective, which may be due to different surface interactions with the nanoparticle. This study suggests a new approach to enhanced nanoparticle delivery in the lung. Nanoparticle diffusion. 

New CFTR modulators remain needed to normalize CFTR function for the majority of subjects with CF, and especially for those carrying rare uncharacterized CFTR mutations. Up until now, primary human cells such as primary airway cultures have played a dominant role to identify CFTR modulators or responding theratypes. Recently, epithelial fluid secretion assays in intestinal organoids also have been demonstrated as relevant for CF drug development and patient stratification. This model has an advantage over airway cultures in terms of individual accessibility and assay throughput.

We initiated the RAINBOW project in which >1400 FDA-approved drug compounds (Selleckchem) including ivacaftor, lumacaftor and ivicaftor/lumacaftor combined are screened at a single dose (3 µM) for their potential to modify fluid secretion in intestinal organoids. Organoids were grown from more than 150 Dutch CF patients carrying CFTR mutations with a prevalence of <0.5% in the Dutch CF population. As fluid secretion in intestinal organoids is strictly CFTR-dependent, we anticipated that the majority of hit compounds modify CFTR function.

A first screen aimed to select toxic compounds, based on a live-dead screen using propidium iodide staining and visual organoid morphology in two patient-derived organoid cultures. This led to the exclusion of 31 compounds (2.1% of the total library). Next, we performed a primary screen in 85 organoid cultures with various rare genotypes using our high-content 384-well forskolin-induced swelling assay (HTS-FIS). The screening of >60.000 conditions was feasible, and data from >50 screened plates indicated an average Z'-prime factor >0.5 indicating excellent screening performance of the HTS-FIS. Primary hits were then validated using the conventional assay in 96-well plates. Out of 52 confirmed hits that modified epithelial fluid secretion, we selected the 12 most promising hits based on oral route of administration, availability on the Dutch market and likely long-term safety characteristics. The 12 most promising hit compounds will be tested individually and as add-on therapy to ivacaftor/lumacaftor in all organoids from the RAINBOW cohort, of which the results of the first 40 organoid cultures will be presented here. Additionally, preliminary results will be shown that address the mode of action of these hit compounds, focusing on demonstrating CFTR modulation at gene, mRNA and protein level.

Overall, the data supports that fluid secretion assays in intestinal organoids facilitate relative large-scale individual drug screening efforts, potentially leading to the repurposing of CF and non-CF therapies as individual modulators of CFTR. Future experiments are needed to further validate hits and to establish in vivo safety and efficacy. CF-related anemia is one such condition with a multifaceted etiologic basis and associations with poor health.

Methods: Annualized hemoglobin (Hgb) and other laboratory, demographic, and anthropometric data were abstracted from the U.S. CF Foundation Patient Registry (CFFPR) for adult and pediatric registrants before and after therapy with ivacaftor (IVA) or lumacaftor/ivacaftor (LUM/IVA) between January 2010 and December 2016. Univariate and multivariate linear mixed models were used to examine the effect of IVA or LUM/ IVA on Hgb in G551D-CFTR homo-or heterozygotes and F508del-CFTR homozygotes, respectively.

Results: A total of 1,347 registrants (707 males, 640 females) with G551D-CFTR and 12,582 F508del-CFTR homozygotes (6,640 males, 5,942 females) were identified who had never undergone lung transplant (TXP) and had contemporaneous data on Hgb and CFTR modulator use. IVA was associated with average Hgb increases of 0.54 g/dL (95% CI 0.39, 0.69, p <0.0001) and 0.18 g/dL (95% CI 0.01, 0.35, p = 0.037) for males and females, respectively, with G551D-CFTR. LUM/IVA was associated with average Hgb increases of 0.58 g/dL (95% CI 0.48, 0.68, p <0.0001) and 0.26 g/dL (95% CI 0.20, 0.33, p <0.0001) for male and female F508del-CFTR homozygotes, respectively. In multivariate models, IVA positively affected Hgb in males but not females, and LUM/IVA positively affected hemoglobin in both sexes.

Conclusions: IVA and LUM/IVA are independently associated with higher Hgb levels in CF patients.

Comparison of estimated Hgb differences by sex associated with IVA or LUM/IVA in CF patients with one or two G551D-CFTR alleles or two F508del-CFTR alleles, respectively, using CFFPR data collected between 2010 and 2016. Black color used to reflect pre-drug values. Gray color used to reflect post-drug values. Lines denote 95% CI of the estimated mean Hgb value. * p <0.05 for same-treatment, same-sex comparison. Mesenchymal stem cells (MSCs) have been shown to have antiinflammatory, antifibrotic and antimicrobial activity both in vivo and in vitro. We have shown that hMSCs decrease inflammation and infection in a murine model of CF lung infection and inflammation which resulted in the first in CF Phase I Safety Trial. Clinical trials of hMSCs have shown no major adverse events, but modest benefits are sometimes observed due to donor hMSC variability and differences in disease specific clinical targets. The major clinical issue in CF is related to uncontrolled pulmonary inflammation and infection. In these studies we pursued an hMSC optimization strategy, with the goal of enhancing potency and efficacy for specific CF clinical endpoints. Real-time PCR and Luminex technology was utilized to follow hMSC gene expression and ability to produce effectors in response to manipulation to a variety of effectors such as IFNγ, TNFα and IL-1β. Functional endpoints included monitoring the hMSCs for antiinflammatory and antimicrobial activity in vitro using our estabslished potency models. Using this format we have identified two different combinations of stimulants based upon their ability to enhance hMSC antimicrobial (hMSC AM ) and anti-inflammatory (hMSC AI ) potency. The hMSC AM and hMSC AI treated cells augmented antimicrobial activity against Pseudomonas aeruginosa and Staphylococcus aureus with the relative potency sequence of hMSCs AM > hMSC AI >hMSCs not-treated >no hMSCs. The AM and AI combination enhanced hMSCs anti-inflammatory potency as measured by PGE 2 and indomethacin production with the relative potency rating of hMSC AI >hMSC AM >hMSC not-treated >no hMSCs. Taking advantage of the murine model of CF lung infection and inflammation, we investigated the in vivo potency of the hMSC AI and hMSC AM preparations compared to hMSCs not-treated and no-hMSC controls to manage infection and inflammation in the model. Cftr -/mice chronically infected with Pseudomonas aeruginosa and infused with hMSC (regardless of hMSC treatment) had better weight retention and improved clinical score compared to Cftr -/mice not treated with hMSCs. Bronchoalveolar lavage (BAL) evaluation of infection and inflammation status of the hMSC treated Cftr -/mice demonstrated that the hMSC AM and hMSC AI augmented the beneficial therapeutic impact of hMSC not-treated in vivo with a greater decrease in inflammation and infection. Pseudomonas aeruginosa colonization was decreased by 70-80% with the sequential impact of hMSC AM >hMSC AI >hMSC not-treated > no hMSCs (n=3, p<0.05). BAL white cell counts were consistent with the level of infection, with hMSC AI >hMSC AM >hMSC not-treated > no hMSCs (n=3, p<0.05). These studies demonstrate that optimization of hMSCs for inflammation and infection endpoints specific in CF pathophysiology has the potential to enhance therapeutic impact for clinical benefit. Our current efforts are focused on optimzing CF hMSCs and refining to a minimum effector combination. This work was supported by the Cystic Fibrosis Foundation and the David and Virginia Baldwin Fund. Imaging. 2014; 40:1230-7) . In cystic fibrosis (CF), little data exists on LS as no work has been done with limited noninvasive techniques available to investigate mechanical properties of the lung. Our research collaborative has introduced a new powerful imaging technique, Magnetic Resonance Elastography (MRE), as a potential noninvasive means to spatially map LS in CF patients. The aim of our initial study is to investigate the feasibility of MRE to estimate LS in CF patients compared to healthy volunteers.

Methods: MRE was performed on 22 young adult healthy volunteers (12 males, 10 females), and 3 young adult CF patients (1 male, 2 females) in a 1.5T MR scanner (Avanto, Siemens Healthcare, Erlangen, Germany). A SE-EPI MRE sequence was used to obtain 5 axial slices. For MRE acquisition, a passive driver was placed near the apex of the lungs to introduce cyclic mechanical waves at 50Hz. These mechanical waves were encoded in three spatial directions by using 250Hz motion encoding gradient. Each encoding direction required 26 seconds breathhold at residual volume (RV) and total lung capacity (TLC) for all 5 slices. The scan parameters included: FOV:40cm; slice thickness:10mm; TE:11.6ms; TR:400ms; acquisition matrix: 128x64; echo train length: 9. To accurately estimate LS, lung density was estimated using GRE sequence as described previously (Theilmann RJ, et al. J Magn Reson Imaging. 2009; 30:527-34) . The wave data was processed using direct inversion algorithm to obtain spatial LS map and report the mean LS.

Results: The Figure shows the box plot and LS maps obtained at RV and TLC for healthy and CF patients. The mean LS of healthy subjects at RV (0.93±0.22 kPa) and TLC (1.41±0.41 kPa) is lower than that of CF patients' LS at RV (1.06±0.12 kPa) and TLC (1.89±0.42 kPa), respectively. Additionally, LS at TLC is higher than LS at RV. Although it has been observed that most discontinuations occur within the first two weeks of initiation due to respiratory adverse events (Konstan et al. 2017; Wainwright et al. 2015) , there is still limited data with regard to long-term tolerability and rates of discontinuation in the real-world setting. For example, TRAFFIC/TRANSPORT and PROGRESS each reported one serious adverse event related to hypertension (HTN). Here, we report a long-term rate of discontinuation consistent with other observational studies but with a higher rate of HTN leading to treatment discontinuation, including one severe AE in the form of a hypertensive emergency.

Methods: To assess the long-term rate of LUM-IVA treatment discontinuation, we conducted a single-center retrospective cohort study at St. Paul's Hospital (Vancouver, Canada). We reviewed patient charts from April 2016 to May 2018 and noted discontinuations, and the reasons given for discontinuation. For patients who became hypertensive after LUM-IVA initiation, we extracted BP readings for at least a year prior to initiation.

Results: Of 35 patients who were approved for therapy through compassionate use or through a private drug benefit plan, 26 patients started on LUM-IVA. Of the 26 patients who initially started on LUM-IVA, 10 (38%) discontinued therapy: 3 patients (11%) cited worsening respiratory symptoms (increased chest tightness [n=3], shortness of breath [n=2], increased sputum [n=1]); 1 was transplanted; 1 could not renew their insurance coverage; 1 had worsening anxiety; and 4 (15%) developed HTN. Out of the 4 patients that developed HTN, 2 were symptomatic, including one instance of hypertensive emergency characterized by thunderclap headache. Increases in systolic pressure (from baseline) ranged from 25 to 50 mmHg (70 mmHg increase in the case of the hypertensive emergency). Increases in diastolic pressure ranged from 12 to 20 mmHg (19 mmHg increase in the case of hypertensive emergency). For all 4 patients, BP normalized after cessation, although remained elevated above baseline for 3 patients.

Conclusion: Our overall rate of LUM-IVA treatment discontinuation (38%) is consistent with other real-world observational studies involving CF adults (30%-40%). However, we observed a much higher incidence of HTN in our patients (15%). This study underscores the importance of post-marketing surveillance for unexpected adverse effects related to CFTR modulators. Acknowledgment: Supported by FoM Summer Student Research Program, UBC.

Bratcher, P.E.; Zeitlin, P.L. Pediatrics, National Jewish Health, Denver, CO, USA Introduction: Carbocyclic aryl alkenoic acid (CAAA) derivatives encompass a family of compounds also known as butyrates (BA). We and others have published that BA and 4-phenylbutyrate (4PBA) can restore trafficking of specific mutant proteins that cause diseases like cystic fibrosis (CF) (Rubenstein RC, et al. J Clin Invest. 1997; 100:2457-65; Ma W, et al. Elife. 2017; 6; Naia L, et al. J Neurosci. 2017; 37:2776-94) . The hypothesis is that CAAAs offer compelling add-on corrector strategies for F508del CFTR.

Methods: Human nasal epithelial cells (CF and non-CF) were obtained under informed consent by nasal brushings of the inferior turbinate. Brushed cells were expanded using conditional reprogramming culture methods (Reynolds SD, et al. Am J Respir Cell Mol Biol. 2016; 55:323-36) . Cells form well-differentiated monolayers after 4-6 weeks at the air-liquid interface. Functional expression of CFTR was quantified in Ussing chambers, described briefly here. Primary F508del CFTR nasal epithelial cells, NuLi1 (non-CF bronchial), and CuFi1 (F508del bronchial) cell lines were cultured and studied at the air liquid interface. Each monolayer was exposed to vehicle control, 4PBA, 4-phenyl-Δ3-transbutenoic acid (P3TBA), 4-phenyl-but-2-enoic acid (4PB2), VX-809, or VX-661 for 24-48 hours at the indicated concentrations -alone or in combination -at 37°C. The apical and basolateral membrane baths were symmetrical in Ringer's solution, (i.e. no chloride gradient was imposed). Baseline voltage and transepithelial resistance stabilized, and then short circuit conditions were imposed. The sodium current was measured by 100 µM amiloride applied to the apical bath. CFTR was then stimulated by 20 µM forskolin/100 µM IBMX applied to both the apical and basolateral baths; in specific experiments, 1 µM VX-770 was added to the apical bath to maximally stimulate CFTR; next, to confirm that it is CFTR, 10 µM CFTinh-172 was applied apically; and finally 100 µM ATP was applied to the apical bath to activate the calcium-activated chloride secretion.

Results: NuLi1 cells responded in a dose-dependent manner (10nM to 1mM) to pre-incubation with 4PBA with maximal forskolin/IBMX-stimulated Isc (µA/cm 2 (SD)) after 1 mM 4PBA. CuFi1 cells behaved in a similar dose-response manner with maximal forskolin/IBMX-stimulated Isc (0.57 (0.12)) and VX-770-stimulated Isc (1.41 (0.44)) at 1 mM 4PBA. The VX-770-stimulated Isc after combination of 4PBA (1mM) and VX-661 (3 µM) for 24 hours was significantly increased (2.13 (0.59)) compared with 4PBA alone (0.77 (0.38) or VX-661 alone (1.59 (0.27)), N=3, p = 0.0051. Primary F508del nasal epithelial cultures from two donors also demonstrated maximal forskolin/IBMX/VX-770-stimulated Isc as well as CFTR(inh)-172-inhibited Isc when incubated with the combination of 4PBA and VX-661. The CAAAs P3TBA and 4PB2 also increased forskolin/IBMX/VX-770-stimulated Isc and CFTR(inh)-172 inhibited Isc to variable degrees.

Conclusions: 4PBA stimulates additional CFTR-mediated Isc in non-CF bronchial epithelial cells. Importantly, 4PBA and at least 2 other CAAAs correct F508del CFTR in primary F508del nasal epithelial cells and combine with VX-661 in a synergistic relationship.

Day, B.J. Medicine, National Jewish Health, Denver, CO, USA Introduction: Chronic bacterial lung infections are a hallmark of cystic fibrosis (CF) lung disease and are difficult to eradicate. Chronic lung infections result in chronic lung inflammation and oxidative damage which creates a paradox where most anti-inflammatory and antioxidant therapies can suppress an already inadequate lung immune response. Clinical trials using inhaled antioxidants such as N-acetylcysteine (NAC) or glutathione have reported only modest benefits in CF subjects. A potential better approach in CF would be to develop an antioxidant that can also supplement the innate immune system. We have previously reported the beneficial effects of supplementation with the small pseudohalide thiocyanate ( -SCN) in a mouse model of CF lung infection and inflammation. The object of this study was to evaluate the use of -SCN as a counter anion for a manganese porphyrin catalytic antioxidant AEOL10150, which could increase the parent compound's antioxidant spectrum and supplement the lung's innate immunity.

Methods: The antioxidant activities of the parent compound AEOL10150(Cl 5 ) were compared with the anion exchanged AEOL20415(SCN 5 ). Superoxide dismutase activity was measured spectrophotometrically using the cytochrome c assay. Catalase activity was measured using a hydrogen peroxide electrode. Inhibition of lipid peroxidation was assessed spectrophotometrically using a thiobarbituric acid reactive species (TBARS) assay. Peroxidase activity was assessed using an oxygen electrode. The abilities of AEOL10150(Cl 5 ) and AEOL20415(SCN 5 ) to protect human bronchiolar epithelial cells (16HBEs) against hypochlorite (HOCl) and glycine chloramine was assessed spectrophotometrically using a MTT cell viability assay. Antimicrobial activity of compounds was assessed by measuring colony forming units after a 2-hour incubation.

Results: The parent compound AEOL10150(Cl 5 ) is a well characterized broad spectrum catalytic antioxidant with high rates of reaction with superoxide, hydrogen peroxide and lipid peroxides. We found that by anion exchanging out the chloride for thiocyanate the resulting AEOL20415(SCN 5 ) retained comparable antioxidant activities for these reactive oxygen species (ROS). In addition, the AEOL20415(SCN 5 ) gained the ability to scavenge ROS associated with innate immunity, including hypochlorite (HOCl) and glycine chloramine, and could protect human bronchiolar epithelial cells against their toxicity which was not observed with the parent compound AEOL10150(Cl 5 ). We also observed evidence that the manganese porphyrin demonstrated the ability to utilize thiocyanate as a peroxidase substrate to generate hypothiocyanate. AEOL20415(SCN 5 ) demonstrated antimicrobial activity against CF lung pathogens including Pseudomonas aeruginosa and Burkholderia vietnamiensis.

Conclusions: We provide data that one can augment antioxidant activity of positively charged antioxidants by substituting with the pseudohalide thiocyanate as the counter anion. This resulted in enhanced antioxidant spectrum. These novel types of antioxidants may be more beneficial in treating CF lung disease.

Acknowledgments: Supported by Aeolus Pharmaceuticals and a Cystic Fibrosis Foundation Research Grant (DAY18G0). Introduction: Cystic fibrosis (CF) airways exhibit excessive and persistent inflammation partly driven by airway macrophages (AMs). Lenabasum is an oral synthetic cannabinoid receptor type 2 (CB2) agonist that has been shown to reduce the production of key airway pro-inflammatory cytokines known to play a role in CF. Lenabasum acts by activating the resolution of inflammation, an endogenous pathway that restores immunological homeostasis.

In a recently completed double blind placebo-controlled Phase 2 study with 85 adults diagnosed with CF, lenabasum reduced pulmonary exacerbations and decreased inflammatory cells and mediators in sputa while demonstrating a favorable safety profile. The study recruited patients regardless of their CFTR mutation, their lung pathogens or their current standard of care (including approved CFTR-targeting medications).

In this study, we further evaluated lenabasum by testing its effects on AMs from harvested human CF lungs.

Methods: AMs were isolated from lungs excised from CF patients undergoing lung transplantation and studied following 3 days in primary culture. AMs were stimulated with Pseudomonas aeruginosa lipopolysaccharide (LPS, 100 ng/mL) for 6 hours in the absence or presence of lenabasum (1, 3, or 10 µM). Subsequently, the following biomarkers of inflammation were evaluated: 1) the mRNA levels (measured by quantitative RT-PCR) of spliced X box binding protein-1 (XBP-1s), a key endoplasmic reticulum stress transcription factor required for LPS-induced cytokine production in CF AMs, and sphingosine-1 phosphate kinase 1 (SPHK-1), a kinase that phosphorylates sphingosine-1, a sphingolipid metabolite that has been implicated in inflammatory responses and lung fibrosis; 2) secretion of TNF-α and IL-8; and 3) secretion of the lipid mediator prostaglandin 15d-PGJ2 (15d-PGJ2), which has been linked to the resolution of inflammatory responses. Cytokine and lipid mediator secretion were measured in culture media using ELISA.

Results: Lenabasum reduced LPS-stimulated XBP-1s and SPHK-1 mRNA levels, and this response coincided with the reduction of LPSstimulated secretion of TNF-α and IL-8 in CF AMs (Table) . In contrast, lenabasum increased the secretion of the pro-resolution mediator 15d-PGJ2 in CF AMs (Table) .

Conclusions: Our findings provide additional evidence for direct anti-inflammatory and pro-resolution effects of lenabasum on a key cell type that contributes to the pathogenesis of CF inflammatory lung disease. These results offer further support for testing the therapeutic benefit of lenabasum in CF patients. Background: Lumacaftor/ivacaftor (Lum/Iva, Orkambi ® ) is the first CFTR-targeted drug approved for clinical use for patients affected by cystic fibrosis (CF) homoygous for F508del, the most common CFTR mutation. Lumacaftor (VX-809) is a CFTR corrector and ivacaftor (VX-770) is a CFTR potentiator. Previously in the placebo-controlled clinical trials TRAFFIC and TRANSPORT the mean absolute improvement in the percentage of predicted forced expiratory volume in the 1st second (FEV1) ranged from 2.6 to 4.0 percentage points and the rate of events leading to hospitalization or the use of intravenous (IV) antibiotics was lower in the group treated with Lum/Iva (N Engl J Med. 2015; 373:220-31) . In this study we describe the clinical and functional effects on CFTR of this drug in 44 CF patients treated for at least 3 months at the CF Center of Verona in Italy.

Methods: Prescriptions and follow-up were performed according to the rules of Italian regulatory agency for drugs (Agenzia Italiana del Farmaco AIFA) for 47 CF patients. Age of CF patients ranged from 13 -48 years. Analysis of the effects of Lum/Iva focused on their nutritional status expressed as weight and Body Mass Index (BMI), CFTR function measured by Gibson and Cooke sweat test, lung function, frequency of pulmonary exacerbations during the first year of treatment (available only for 16 patients according to the European Consensus Group definition, J Cyst Fibros. 2011;10(S2):S79-81). For 3 patients the treatment was shorter than 3 months and therefore the effects are not available this study. For the others the duration of treatment ranged from 91 days to 4.8 years.

Results: In CF patients treated with Lum/Iva we observed a mean of absolute FEV1 increase of 2.6% ± 8.4 SD, a mean BMI increase of 0.3 ± 1 SD, a mean weight increase of 1.1 kg ± 2.6 SD, a mean reduction of 0.800 ± 0.941 SD of pulmonary exacerbations requiring IV antibiotics or hospitalization, a decrease of sweat chloride values of 19.2 mmol/L ± 12.5 SD vs the mean value of the year before the treatment. In 4 patients the treatment was suspended because of adverse effects, most frequently thoracic oppression. In males and females we obtained very similar results. In severe patients (FEV1 < 50%) we obtained lower improvement of nutritional status in terms of weight and BMI increase than in the other patients (p<0.05) while other effects were not significantly different in these two groups.

Conclusions: CF patients treated with Lum/Iva at the CF Center of Verona have better nutritional status, lung function and less pulmonary exacerbations than before treatment. These effects are consistent with their improved CFTR function. Although conditions of clinical use, in real life, are different than in clinical trials we observe that the effects on relevant clinical outcomes have been achieved at levels expected according to the results obtained in the clinical trials. The variability of response to this drug among homozygous F508del CF patients reported in this study suggests the relevance of personalized medicine and new drug development for CF. Acknowledgment: This study was supported by Lega Italiana Fibrosi Cistica Associazione Veneta Onlus. Introduction: Abnormal chloride (Cl − ) transport has a detrimental impact on mucociliary clearance (MCC) in cystic fibrosis (CF). Ginseng is a medicinal plant noted to have a number of anti-inflammatory and antimicrobial properties. The objectives of the present study are to assess the capability of red ginseng aqueous extract (RGAE) to promote transepithelial Clsecretion in nasal epithelium and inhibit P. aeruginosa biofilm formation.

Methods: Wild-type (WT) and transgenic CFTR -/primary murine nasal septal epithelial (MNSE) cultures were pharmacologically manipulated in Ussing chambers to measure the impact of RGAE on vectorial Clsecretion. Effects on ciliary beat frequency (CBF) were also measured. Biofilm formation of the PAO1 strain of P. aeruginosa incubated with and without RGAE was quantified by crystal violet staining.

Results: RGAE (at 30 µg/mL of ginsenosides) significantly increased Cltransport [measured as change in short-circuit current (ΔI SC =µA/cm 2 )] when compared to vehicle control in both WT and CFTR -/-MNSE (WT vs Control = 49.8 ± 2.6 vs 0.1 ± 0.2, CFTR -/-= 33.5 ± 1.5 vs 0.2 ± 0.3, p < 0.0001). CBF (fold-change/baseline) was significantly increased compared to control in WT MNSE (2.1 ± 0.18 vs 1.2 ± 0.04; p < 0.01). RGAE markedly reduced PAO-1 biofilm formation compared to controls (30 µg/mL ginsenosides = 19.6 ± 3.2% of reduction, p < 0.01). From whole cell patch clamp analysis, using TMEM16A expressing in HEK-293 cells, there was no significant activation of TMEM16A channels by RGAE alone.

However, there was a synergism in the conductance of Clwith combination of UTP and RGAE.

Conclusion: RGAE activates transepithelial Clsecretion, while inhibiting PAO1 biofilm formation. These findings suggest RGAE has therapeutic potential in CF, especially in those infected with P. aeruginosa. RGAE (at 30µg/mL of ginsenosides) significantly increased ΔIsc when compared to vehicle control in both WT and CFTR -/-Murine nasal epithelium. The efficacy of lumacaftor/ivacaftor (ORKAMBI®, Vertex Pharmaceuticals) combination therapy in patients with cystic fibrosis (CF) was evaluated in TRAFFIC and TRANSPORT parallel studies. The mean absolute change in percent predicted (pp) FEV 1 over placebo at 24 weeks in both studies was between 2.6 and 4.0% (p<0.001).

We investigated the real-world efficacy of lumacaftor/ivacaftor in a paediatric CF population, 12 years of age and older, using multiple breath washout (MBW) and spirometry. The data were collected during routinely scheduled outpatients' visits. Here we present pre-and post-lumacaftor/ ivacaftor values, which are an average of two independent measurements recorded at a time of clinical stability, a minimum four weeks apart, before, and then after starting the treatment.

Nitrogen MBW was performed using the Exhalyzer D (Eco Medics AG, Switzerland) and followed the European Respiratory Society (ERS) and American Thoracic Society (ATS) 2013 consensus statement to determine the lung clearance index. Statistical analyses were performed with SPSS Version 24 software (IBM Corporation). The comparison of pre-and post-lumacaftor/ivacaftor lung clearance index (LCI) values and ppFEV 1 values was performed using paired-samples t-test.

Eight patients (three girls), aged between 12.1 and 16.5 years old, had paired data for both lung function assessments. The lumacaftor/ivacaftor treatment resulted in a significant drop in the average LCI value of -1.72 lung volume turnovers, ranging from -3.9 to 0.34, (p=0.021), while the absolute mean change in ppFEV 1 from baseline was 1.5%, ranging from -9.5% to 11.0%, (p=0.502).

The LCI median (range) was 11.25 (7.41-14.33 ) and 8.32 (7.73-13 .45) for pre-and post-lumacaftor/ivacaftor treatment respectively (Fig.1A) . The ppFEV 1 median (range) was 82.5% (49.5%-103.0%) and 86.8% (48.5%-106.5%) for pre-and post-lumacaftor/ivacaftor treatment respectively (Fig.1B) .

Our study demonstrates that the positive effect of lumacaftor/ivacaftor on lung function may be greater than that demonstrated by ppFEV 1 alone. Furthermore, the ability of LCI to detect potentially clinically meaningful improvements suggests it has the potential to act as sensitive outcome measure in clinical trials, which could potentially affect the "sample size" required in clinical trials resulting in significant reduction in the numbers of participants needed to be recruited. Introduction: The lung clearance index (LCI) from the multiple breath washout (MBW) test is a promising surveillance tool for young children with cystic fibrosis (CF). However, longitudinal data are needed to elucidate if MBW can effectively monitor CF lung disease status. Our objective was to determine if LCI can identify structural lung disease progression on chest computed tomography (CT) and correlate with pulmonary infection and inflammation.

Methods: Healthy children (n=75) and children with CF (n=112) aged 3 to 6 years were recruited for MBW testing every 6 months. Children with CF underwent annual bronchoalveolar lavage fluid collection and a chest CT scan. Structural lung disease was assessed using Perth-Rotterdam Annotated Grid Morphometric Analysis (PRAGMA) scores. Differential cell counts were performed, and neutrophil elastase (NE) quantified by ELISA. Comparison of changes in LCI with age between CF and healthy children was performed using 2-way ANOVA. Changes in MBW with CT, infection and inflammation outcomes between visits were assessed using mixed effects models with random subject intercepts. Models were adjusted for age at visit and study centre.

Results: LCI was significantly higher in children with CF compared with healthy children from 4 years of age ( Figure) . Increases in LCI were significantly associated with total lung disease extent (coefficient 0.15; 95% CI: 0.05, 0.26; p=0.003) and air trapping (coefficient 0.05; 95% CI: 0.02, 0.07; p=0.002), but not bronchiectasis (coefficient 0.12; 95% CI: -0.02, 0.27; p=0.09). Increased LCI was also associated with increased neutrophils (coefficient 0.016; 95% CI: 0.01, 0.02; p=0.0001) and the presence of NE (coefficient 0.67; 95% CI: 0.25, 1.09; p=0.002), but not proinflammatory pathogens (coefficient 0.01; 95% CI: -0.42, 0.44; p=0.44). Hospital admission for a pulmonary exacerbation was associated with increased LCI (coefficient 0.47; 95% CI: 0.03, 0.90; p=0.04).

Conclusion: LCI may be a sensitive tool to monitor the progression of structural lung disease, the development of lower respiratory tract inflammation and presence of pulmonary exacerbations as defined by hospitalization in young children with CF. Background: Cystic fibrosis (CF) is a genetic disease resulting from mutations of the cystic fibrosis transmembrane conductance regulator (CFTR) gene. CFTR is responsible for proper movement of chloride and bicarbonate ions across the apical cell surface. Defective CFTR results in an absence of bicarbonate transport, which functions in part to chelate Ca 2+ , a divalent cation found in abundance in mucin granules in normal conditions; this yields a hyperviscous and adhesive mucus on the respiratory tracts of patients with CF. A novel therapeutic, PAAG, is a polycationic derivative of polyglucosamine that is a soluble, nontoxic polysaccharide shown to interact with mucins. Previously, we demonstrated that PAAG improves the viscoelasticity and transport of CF sputum, CF HBE mucus, and purified MUC5B. We have also recently shown that the CF rat displays increased airway mucus viscosity and delayed transport at 6 months of age, allowing for the evaluation of novel mucoactive agents such as PAAG.

Methods: CF rats were aged to 6 months of age then treated with nebulized PAAG (250 µg/mL x 20 mL over 45 minutes whole body) or glycerol vehicle control once daily for 14 days. Upon euthanasia, trachea and lung were excised and opened along the dorsal surface. They were then placed onto a F12 media-soaked gauze and allowed to equilibrate to physiologic conditions (37°C, 5% CO 2 , 100% humidity) for 30 minutes before imaging. Micro-optical coherence tomography (µOCT) was used to evaluate mucus transport en bloc across excised rat airways. MCT rate was determined using time elapsed and distance traveled of native particulates in the mucus over multiple frames.

Results: Mucociliary transport rates were signficantly faster in animals treated with PAAG (0.61 ± 0.27 mm/min, P<0.001) as compared to those treated with PBS (0.17 ± 0.12 mm/min). These data demonstrated PAAG treatment accelerated mucus transport 3.5-fold over control, achieving ~ 70% of MCT rates in normal rats (0.89 ± 0.55 mm/min).

Conclusions: PAAG is currently being evaluated for not only its role in mucus clearance, but also its role in infection (Garcia B, et al. Pediatr Pulmonol. 2018 Introduction: Cell lines such as CFBE41o-as well as primary human nasal (HNE) and bronchial epithelial (HBE) cells are commonly used to study CFTR rescue in airway epithelia. Epithelia are an important site of extra-hepatic drug metabolism, but the inter-individual variability in expression of drug-metabolizing enzymes and transporters (DMET) in the CF population is not known. Since ivacaftor is a substrate of DMET, variation in enzyme expression between individuals may impact drug efficacy. We hypothesized that the RNA expression of proteins involved in transport (ABCB1) and metabolism (CYP3A5, CYP3A4) of ivacaftor, lumacaftor, and tezacaftor would vary in individuals and cell models used in CF research.

Methods: Nasal epithelial brush biopsies were collected from three non-CF and 18 CF individuals, expanded, and cultured at air-liquid interface (ALI) until well-differentiated. CFBE41o-with wild-type (WT) CFTR complementation and Calu3 cell lines were cultured using standard methods. Cells were cultured at ALI until high resistance monolayers were formed. RNA was isolated using the Qiagen DNA/RNA extraction kit. Gene expression of ABCB1, CYP3A5, and CYP3A4 was then analyzed using quantitative real-time RT-PCR and relative expression levels were calculated using ΔΔCt method; all values normalized to GAPDH as the internal control. To standardize comparisons across individuals, the RNA expression of cultures from three distinct non-CF subjects were averaged together to serve as the calibrator. To assess protein expression of the key extrahepatic CYP3A isoform, immunoblots of CYP3A5 were performed.

Results: All CF subjects expressed CYP3A5 in HNE cells, whereas 20% of subjects also expressed CYP3A4 RNA. Fifteen percent of the CF subjects had ≥ 5 fold expression of CYP3A5 relative to control. One CF subject had a ~22 fold higher expression of CYP3A5 relative to the control. Thirty percent of CF subjects expressed ABCB1. Contrary to primary cell cultures, the commonly used cell line CFBE-WT had no expression of either CYP3A5 or CYP3A4, although a relatively high level (~10 fold) of ABCB1 was found. In comparison, expression in Calu3 was 8 fold greater RNA expression of CYP3A5 compared to CFBE-WT. Immunoblots of CYP3A5 in CFBE-WT, HepG2, and Calu3, confirmed elevated expression in Calu3. Immunoblots of primary cultures and of other proteins are ongoing.

Conclusion: The lack of expression of CYP3A5 and CYP3A4 in the cell line CFBE-WT could result in decreased metabolism of drugs (including ivacaftor and tezacaftor, which are substrates of these enzymes), leading to higher intracellular drug concentrations and a more robust effect in functional assays. In individuals with CF, the expression of proteins important in DMET is variable and may contribute to differences in modulator-induced CFTR rescue between in vitro models, as well as differences between models derived from different subjects. Evaluation of other cell models is currently ongoing.

The self-degradative process, termed autophagy, is defective in cystic fibrosis (CF), as a result of abnormal function and production of the cystic fibrosis transmembrane conductance regulator (CFTR) protein.

In CF the accumulation of defective CFTR in the cytosol and a lack of expression of functional CFTR on the cell membrane, leads to perturbation of autophagy, resulting in amplified inflammation. In this study, the anti-inflammatory effect of pre-treating human bronchial epithelial cells (HBECs) and primary monocytes with the proteostasis regulator, cysteamine, and the flavonoid, epigallocatechin gallate (EGCG) was investigated.

Methods: HBEC lines Beas-2b wild-type (WT), IB3-1 (CF -ΔF508/ W1282X), CUFI-1 (CF -ΔF508/ΔF508), CUFI-4 (CF -ΔF508/G551D) and CF patients' monocytes were studied. ELISAs were performed to measure IL-6, TNF and IL-10. Small molecule drugs, cysteamine (250 µM) and EGCG (100 µM), were used to treat both epithelial cell lines and patients' monocytes in vitro. LPS (10 ng/mL) was used to stimulate both HBECs and monocytes. Western blots were used to demonstrate defective autophagy in the cell lines and in CF patient monocytes.

Results: Autophagy proteins were shown to be deficient in CF monocytes compared to healthy controls (HCs), and cysteamine was successfully shown to recover the proteins. The epithelial cell lines and CF patients' monocytes presented with a hyperinflammatory response to LPS, with raised levels of IL-6 and TNF. The levels of IL-6 and TNF were greatly reduced upon addition of cysteamine and EGCG. IL-10 levels were shown to be reduced in CF monocytes compared to HCs. Cysteamine and EGCG reduced the quantity of inflammatory cytokines secreted in HBECs and CF monocytes.

Conclusion: These data show that CF has an exaggerated inflammatory response and a deficient anti-inflammatory response to LPS stimulation in comparison to HC, showing an inherent inflammatory state in CF. The results show that the combination of cysteamine and EGCG is considerably more potent a treatment than when they are used separately. Introduction: Hyper-inflammation drives the irreversible lung tissue damage in CF patients. Macrophages (MΦs) mutated at the CFTR contribute to the hyper-inflammation and defective host defense in CF lung disease. Heme oxygenase-1/carbon monoxide (HO-1/CO), a key signaling pathway involved in resolving inflammation, is dysregulated in CF MΦs and murine CF lungs in response to infection. HO-1 is an inducible enzyme that degrades heme groups, producing anti-inflammatory mediators, such as carbon monoxide (CO), which help to resolve inflammation. CO also has potent bactericidal activity and, in a positive feedback loop, increases endogenous HO-1 expression. The HO-1 pathway is an attractive target for disrupting the hyper-inflammatory response in CF. PEGylated carboxyhemoglobin bovine (Sanguinate®, SG) is a CO-releasing molecule in a Phase II clinical trial for the treatment of sickle cell anemia. The goal of this study was to evaluate whether SG improves resolution of inflammation while strengthening host defense in CF by acting on the HO-1/CO pathway.

Methods: CF-KO and wild-type (WT) bone marrow-derived murine MΦs or peripheral blood-derived MΦs were isolated from healthy donors and CF patients. MΦs were preconditioned for 6 hours with 2 mg/mL SG or vehicle alone prior to challenging with LPS (10 µg/mL) or live Pseudomonas aeruginosa (PA) (PAO1, MOI=1:10). CFU assay was used to assess the bacterial intracellular killing and to determine the bacterial load in the media collected from SG-treated MΦs and then exposed to PA. To test SG's effects in vivo, WT and CF mice were pretreated intravenously with a single dose (320 mg/Kg) of SG or vehicle alone. HO-1 expression in the lung tissues and BAL fluid cell number was assessed at steady state, and in response to LPS (e.g. 3 nebulizations with 12.5 mg PA-LPS every day for 3 days).

Results: SG is a potent dose-dependent inducer of the HO-1 protein in human and murine CF MΦs, increasing CF HO-1 levels to those found in non-CF cells. SG's mechanism of action for inducing HO-1 expression relies on activation of PI3K/AKT signaling. By rescuing the PI3K/AKT pathway and inducing HO-1 to normal levels, SG decreases the hyper-inflammatory response in activated CF MΦs (eg, reduced expression of IL-6, TNF-α and CXCL1), and strengthens the MΦ host defense against PA by improving intracellular and extracellular bacterial killing. Systemic delivery of a single dose of SG is sufficient to induce HO-1 expression in lung tissues. Moreover, SG pretreated mice have a reduced number of neutrophils in the BALF after LPS nebulization with respect to control animals.

Conclusion: Our results suggest that SG treatment stimulates HO-1/ CO pathway, which mediates resolution of the inflammatory response, and promotes host defense against PA. Therefore, SG could help decreasing hyper-inflammation, improving host defense, and, ultimately, minimize CF lung disease. Introduction: Cystic fibrosis (CF) is the most common genetic disease in Caucasians, caused by different mutations in the CFTR gene. Hyperproduction of mucus, gradual decline in forced expiratory volume (FEV1) and persistent hyperinflammation have long been recognized as the age-related pathologies of CF. Fenretinide (FEN), a derivative of vitamin A, has been shown to revert the proinflammatory imbalance of fatty acids and ceramides in CF mice and to inhibit the NF-kappaB pathway in cancer cell lines and macrophages. Lung infection with P. aeruginosa induces the production of both MUC5AC and MUC5B mucins in the lungs. While excessive production of MUC5AC leads to pathological plugging of the airways, the MUC5B gene was shown to be essential for clearing bacterial infections.

Methods: We used the CFTR knockout (KO) mouse on a C57Bl/6 genetic background (CFTR-KO) as a model to study the progression of CF lung disease. We measured airway resistance in response to increasing concentration of methacholine, an experimental parameter in mice that most closely correlates with human FEV1. As a model of chronic infection, mice were infected with P. aeruginosa (PA508) embedded in agarose beads. For histological assessment, lungs were stained with Periodic Acid-Schiff/ Alcian Blue to evaluate mucus production and hematoxylin/eosin (H&E) to evaluate the infiltration of inflammatory cells. We have used SPOC-1 lung goblet cells to analyze mucus production induced by bacterial lipopolysaccharide (LPS). Gene expression was analyzed at the mRNA level by quantitative real-time PCR (qPCR) and at the protein level by ELISA assays.

Results:Our data demonstrate that young CFTR-KO mice do not differ from their wild-type (WT) littermates in the production of mucus at the baseline. Their production of mucus upon bacterial infection is similar to the WT. However, uninfected older CFTR-KO mice show a much higher degree of infiltration of inflammatory cells in the lungs than WT, which can be prevented by FEN treatment. We also show that CF-KO mice display higher methacholine-induced airway resistance than WT and that FEN treatment efficiently corrects this defect. Furthermore, we demonstrate for the first time that FEN treatment efficiently regulates the total production of mucus in the lungs of P. aeruginosa-infected mice. Finally, FEN treatment also prevents the LPS-induced increase in MUC5AC expression, without affecting the level of MUC5B gene expression in the SPOC-1 cell line.

Conclusion: Our data demonstrate age-dependent progression of CF lung disease in CFTR-KO mice. It also demonstrates the impaired lung physiology in CFTR-KO mice compared to their littermate control, which can be improved following FEN treatment. Overall, the protective effects of FEN treatment on CF lung disease and lung physiology as well as FEN-induced selective modulation of LPS-induced production of MUC5AC provide further support for the use of this therapy to prevent the progression of CF lung disease. Rationale: Exercise capacity is an important outcome parameter for individuals with cystic fibrosis (CF) and is a strong predictor of prognosis. Lumacaftor-ivacaftor (Orkambi™) improves FEV 1 percent-predicted (ppFEV 1 ) over 24 weeks and improvements are observed within the first month of treatment. It is unknown if this lung function benefit translates into improvements in exercise endurance, exertional dyspnea, or leg discomfort ratings. Accordingly, the primary purpose of this study was to determine the effects of lumacaftor-ivacaftor on exercise endurance and perceptual responses to exercise following the first month of treatment.

Methods: CF individuals who were F508del homozygous with plans to start lumcaftor-ivacaftor as part of usual care were approached for this study. Visit 1 included pulmonary function tests (PFTs) and a symptom limited incremental cycling test. Visit 2 included PFTs and a symptom limited constant load cycling test (80% peak work rate) with simultaneous assessments of dyspnea and leg discomfort ratings. Participants initiated treatment with lumacaftor-ivacaftor following visit 2. Visit 3 (post) was performed 1 month after treatment with identical procedures as visit 2.

Results: All 3 study visits were completed by 7 subjects. Lumacaftor-ivacaftor resulted in a modest but nonsignificant improvement in ppFEV 1 (pre-treatment: 1.73±0.32 vs. post-treatment: 1.82±0.32 L, p=0.19) and a significant reduction in sweat chloride (pre-treatment:106±14 vs. post-treatment: 86±15 mmol/L, p<0.001). Exercise time improved in all but one subject, but, on average, was not statistically significant (pre-treatment: 9.0±5.5 vs. post-treatment:10.4±6.3 min, p=0.51). Although nonsignificant, relative change in exercise time inversely correlated with changes in sweat chloride (r=-0.82, p=0.13) but not ppFEV 1 (r=-0.36, p=0.44). Exertional dyspnoea at iso-time did not change (pre-treatment: 5.0±1.6 vs. post-treatment: 4.6±1.7, Borg units, p=0.53) but leg discomfort ratings tended to decrease (pre-treatment: 7.0±2.6 vs. post-treatment: 5.4±2.4, Borg units, p=0.10). There was a significant correlation between changes in endurance time and leg discomfort ratings (r=-0.88, p=0.02).

Conclusions: Lumacaftor-ivacaftor did not reduce dyspnea but tended to improve leg discomfort ratings following 1 month of treatment. Decreases in leg discomfort were significantly correlated with improvements in endurance time. Future studies with a larger sample size are needed to verify these findings and assess the long-term effects of treatment on exercise outcomes. MO, USA; 5. National Jewish Health, Denver, CO, USA; 6. Rigshospitalet, Copenhagen, Denmark; 7. University of Utah, Salt Lake City, UT, USA Objectives: To determine the safety, tolerability, pharmacokinetics (PK) and effect of the cystic fibrosis transmembrane conductance regulator (CFTR) corrector PTI-801 in subjects with cystic fibrosis (CF).

Background: Despite significant advances achieved by approved CFTR modulators, substantial medical need remains for CF patients who are not eligible for, or experience suboptimal benefit from, currently approved modulators. PTI-801 is a novel CFTR corrector identified through amplifier enhanced high-throughput screening. In vitro, treatment of human bronchial epithelial (HBE) cells from F508del homozygous donors with PTI-801 showed that the compound enhances the ratio of mature, fully glycosylated CFTR relative to the immature form of the protein.

In vitro, PTI-801 increased CFTR chloride transport activity and when added to lumacaftor+ivacaftor activity was improved by 281%, suggesting a distinct and complementary mechanism of action of PTI-801 relative to lumacaftor+ivacaftor, an already approved corrector and potentiator combination. PTI-801 represents an improvement on current CFTR correctors. PTI-801 is a CFTR corrector currently in clinical trials designed to determine safety, tolerability and effect in CF subjects.

Methods: A randomized, double blind, placebo-controlled Phase 1 clinical study is being conducted in subjects with CF, age ≥18 years, with a forced expiratory volume in 1 second (FEV 1 ) 40-90% of predicted. Study consists of multiple ascending doses of PTI-801 in CF subjects currently receiving lumacaftor+ivacaftor as background therapy. The primary objective is assessment of safety and tolerability with the secondary objective of assessment of the pharmacokinetics (PK) of PTI-801 and change in lung function. Exploratory objectives include changes in sweat chloride values.

Results: Preliminary data from a minimum of 45 subjects from the multiple ascending dose cohorts in patients with CF on background treatment with lumacaftor+ivacaftor, which will include safety, tolerability, PK and lung function evaluation, are expected in advance of the 2018 North American Cystic Fibrosis Conference.

Conclusions: PTI-801 represents a novel third generation CFTR corrector in clinical development. Background: FDL176 is a novel CFTR potentiator that has been shown to increase CFTR-mediated chloride flux in vitro in homozygous F508del-CFTR hBE cells treated with a CFTR corrector. In combination with CFTR corrector FDL169, it has similar in vitro efficacy to lumacaftor/ ivacaftor when hBE cells were treated for ≤1 hour with each potentiator but higher efficacy when treated for ≥24 hours.

Methods: A Phase 1 first-in-human (FIH) study of FDL176 is ongoing evaluating: 1) safety, tolerability and pharmacokinetics (PK) of single ascending doses (SAD) (ranging from 100-1000 mg); 2) effect of food on bioavailability (single dose of 200 mg); 3) bioavailability in healthy females (single dose of 200 mg); 4) 14-day multiple ascending doses (MAD); and 5) bioavailability in CF subjects.

Results: We report here the preliminary data for FDL176 FIH study. Eighty-eight healthy subjects and 7 CF subjects have received at least one oral dose of FDL176 at the time of submission. No serious adverse events (SAEs) were reported. One subject was withdrawn due to an adverse event (AE) of vessel puncture site cellulitis that was considered not related to FDL176; another subject withdrew due to an AE of increased creatinine, mild in severity and considered possibly related to study FDL176. Majority of AEs were mild or moderate in severity and considered to be either unrelated or unlikely related to FDL176. No clinically significant changes in vital signs or 12-lead electrocardiogram data have been reported. C max and AUC 0-24hr increased in a dose-proportional manner with escalating single doses of FDL176 over the dose range studied. FDL176 had a long half-life with individual subject values ranging from 21.25 to 100.99 hours. FDL176 C max and mean AUC 0-tlast were approximately 3-fold larger in fed state than in fasted state. FDL176 mean AUC 0-24hr in females was similar to that observed in fasted males at the same dose. All data including safety, tolerability and PK are expected in advance of the North American Cystic Fibrosis Conference 2018.

Conclusion: FDL176 is well tolerated and demonstrated a favorable safety profile in healthy and CF subjects. Safety and PK profile support further development of FDL176. 

The delivery of therapeutic cargoes to epithelial cells that line the respiratory tract is challenging. Here we employ a new, engineered transduction peptide strategy to carry protein and CRISPR payloads into airway epithelial cells in vitro and in vivo. These synthetic soluble peptides contain both a positively-charged hydrophilic and a hydrophobic outer face. Co-incubation of amphiphilic peptides with recombinant GFP-NLS protein yielded rapid (minutes) and efficient nuclear GFP-NLS delivery in cultured well-differentiated primary human airway epithelial cells, human tracheal explant tissue, and the large and small airway epithelia of mice. Both ciliated and nonciliated airway epithelial cells were transduced. CRISPR endonucleases (S.p. Cas9, A.s. Cpf1) and guide RNA ribonucleoproteins (RNPs) were also readily transferred into difficult-to-transduce human airway epithelia, achieving gene editing at the CFTR and HPRT1 loci. We delivered Cas9 RNPs to the airways of tdTomato (ROSA mT/mG ) mice and achieved editing of LoxP sites with resultant switch to GFP fluorophore expression in epithelial cells of the large and small airways. Amphiphilic peptide transduction of firefly luciferase protein into conducting airway epithelia of mice, followed by bioluminescent imaging revealed that delivery was rapid and restricted to the respiratory tract. This amphiphilic peptide platform provides a new, flexible, chemical-free strategy to deliver protein and CRISPR cargoes to hard-to-transduce airway epithelial cells in vitro and in vivo.

Acknowledgements: NIH P01 HL51670, CFF, Roy J. Carver Trust, IRAP-NRC. (1) . miRNAs are short noncoding sequences that regulate messenger RNA (mRNA) stability and protein synthesis. miR-145 has a direct binding site on human CFTR mRNA. In primary human airway epithelial cells; overexpression of miR-145 inhibits CFTR function and blocks lumacaftor/ivacaftor correction of F508del CFTR. Antagonists to miR-145 block TGF-β signaling and augment F508del correction in polarized human airway epithelia. The experiments described below extend the above findings from human primary cell culture to preclinical animal models.

Hypothesis: miRNA mediates TGF-β suppression of CFTR function in preclinical models.

Methods: C57/BL6 mice and Sprague-Dawley rats were inoculated with 10 8 -10 9 pfu/mL adenoviral vector of bioactive TGF-β (AdV-TGFβ 223/225 ) or adenovirus with sham vector (AdDL70-3). At 7 days post-inoculation, nasal tissue and lungs were harvested for qPCR measure. Harvested rat tracheal explant or rat tracheal epithelial cells (RTEC) were utilized to measure CFTR current in modified Ussing chambers.

Results: In mice, intranasal inoculation of the AdV-TGF-β vector tripled TGF-β transcription (p<0.05) and increased TGF-β signaling 10-fold (p<0.05) in the nares. The increase in TGF-β signaling doubled miR-494 (a rodent miRNA, analogous to miR-145, that directly binds rat CFTR) and reduced CFTR mRNA in lung homogenates by >80% (p<0.05). In rats, adenoviral delivered bioactive TGF-β halved CFTR function in tracheal explants, with reduction of forskolin/IBMX stimulated current by >50% (p<0.05) and diminution in response to CFTR inh -172 sensitive current by >60% (p<0.01). In cultures of primary rat tracheal epithelial cells (RTEC), exogenous TGF-β (5 ng/mL) doubled miR-494 and reduced CFTR function by 50% (p<0.05). Stimulation of primary RTEC with miR-494 (25 nM) alone halved rat CFTR mRNA (p<0.01), successfully modeling in rats our previous findings regarding miR-145 in human primary airway epithelia.

Conclusions: TGF-β potently downregulates CFTR expression and function in vivo in preclinical rodent models. Congruent to our previous in vitro studies in human airway epithelia in which miR-145 mediates TGF-β inhibition of F508del CFTR correction, we identified miR-494 as a CFTR specific miRNA that mediates parallel relationships in the the rat. Our results lay the foundation to manipulate CFTR in vivo and to assess safety and efficacy of miRNA antagonism to augment CFTR correction in rodents. (3DP) is an evolving technology that creates a physical model from a virtual 3D model. Wide range of materials including plastics, metals, biopolymers and biological cells can be used to create patient-specific models, implantable devices, and tissue engineered grafts (Parthasarathy J. Ann Maxillofac Surg. 2014;4:9-18). Over the past two decades, applications of 3DP have centered on dental, CMF plastic reconstructive surgery. Recent advances include orthopedic, cardiac, and neurosurgical specialties (Heller M, et al. Int J Comput Dent. 2017; 19:323-39) . Patient-specific anatomical models (PSAM) created by 3DP are a powerful adjunct to diagnosis, treatment planning, patient education and physician training. We recently used it as a means to assess progression of lung disease in research studies. Models provide an understanding of the spatial relationship of lesions to adjacent anatomical structures, critical in treatment planning. We describe 3Dmodeling and 3DP of PSAM of the CF lung.

Methods: Illustrated is our technique of 3D virtual and 3DP models of the CF lung (Fig) . Inspiratory and expiratory CT images of the thorax from the apices to the lung bases without contrast were acquired from Toshiba/ Aquilon scanner. Pixel resolution was 0.64mm. Data was imported into MIMICS software (Materialise, Belgium) in DICOM format. Within MIMICS segmentation region-growing algorithm was performed on individual 2D slices. The software then converted the segmented DICOMs into a 3D model representing the structural anatomy of the lung and the pathological fibrosis region with the cysts. Complete segmented lung were then exported as. STL files and sent to the 3D printer via GeomagicFreeform software (3DSystems,USA) for production. Surface area of the pathology, and TLV calculated from the 3D virtual models enabled a full evaluation of disease involvement. A 3DP Lab with dedicated workstations for each of the required software and CONNEX3 Objet 350 printer, with capability to print 30µ layers in multiple colors and material types were available. Material assignment critical for proper visualization of the internal CF structures were clear-lung and opaque pink-CF disease. If printed, the model is processed and encapsulated in polyurethane.

Future Directions: We describe 3D modeling and 3DP of the CF lung from high resolution CT imaging data. The added value of 3DP in virtual and physical setting is the ability to evaluate local anatomy and the region of interest in a patient-specific 3D environment for either clinical or research purposes. As we expand our research efforts investigating spatial stiffness of the CF lung using magnetic resonance elastography (MRE; J Magn Reson Imaging. 2014;40:1230-7), this work complements facilitate better understanding of the compliance of the lung tissue and provide better information using MRE data to generate similar virtual and 3DP models we work toward developing the next gen diagnostic tools for CF lung disease.

Zirbes, J.M.; Alvarez, D.; Ryan, S.P.; Milla, C. Pediatrics, Stanford University, Palo Alto, CA, USA FEV1 is routinely used to detect and monitor CF lung disease and treatment guidelines incorporate it for clinical decisions. However, significant CF lung disease can be present in the face of a normal FEV1. The lung clearance index (LCI) obtained by multiple breath washout (MBW) is increasingly being demonstrated of great value as an endpoint for CF clinical trials. If the LCI is incorporated as part of routine CF care, a clear demonstration of its value in addition to existing tools is required.

Objective: This project aims to evaluate the value of the LCI as a clinical decision support tool as well as identify its Minimal Important Clinical Difference (MCID).

Methods: We are prospectively evaluating the LCI within the clinical care setting in a cohort of children followed at a single CF center. Measurements have been incorporated into routine clinic visits.

Results: So far, 50 potential subjects were approached for participation and 41 consented to the study (82%). Population characteristics are presented in the Table. Nine subjects (20%) were not able to complete testing at baseline. In 3 this was due to acute respiratory illness, 3 due to time constraints during the visit and 3 due to inability to complete the required maneuvers. For those subjects that were able to perform MBW, an average of 3.1 maneuvers were attempted (range 1 to 4) and of these an average of 2.31 (range 0-3) were of research quality. Only 3 children (9%) were unable to produce valid maneuvers, due to leaks from inability to keep a tight seal around the mouthpiece. Neither age nor baseline lung function by FEV1 were predictive of chances of success (p>0.2). There was a marginal effect from gender with girls having higher chances of success (p=0.08). Spirometry parameters were within normal and with no evidence for obstruction. In contrast, LCI demonstrates evidence for ventilatory inhomogeneity with 42% of subjects with an abnormal LCI 2.5 (>2 SD from normative data). LCI2.5 distinguished between patients positive or negative for Pseudomonas within the previous year (p= 0.02), which was not seen for FEV1 or FEV1/FVC.

Our initial experience provides evidence for the feasibility of incorporating MBW testing into routine clinical care visits. We expect as subject follow-up continues that subjects who had difficulty with technique will be able to perform valid maneuvers over time. Interim findings confirm the role that LCI can play in the clinical setting to provide evidence for early stage lung disease. Possible associations between longitudinal changes in LCI and quality of life, symptom scores and exacerbations will also be investigated. Rationale: Bronchiectasis and structural airway injury begin early in infants with cystic fibrosis (CF). PRAGMA-CF is an imaging score sensitive to detect early airway injury but requires significant scorer training. Our objective is to develop automated software to score the proportion of visible airways on chest CT scans as a surrogate of airway injury. We evaluated associations between automated airway index (AAI) with PRAGMA-CF %disease scores from chest CT scans of infants from a single center. We evaluated differences between CF infants and disease control and explored correlations with airway inflammation in CF infants.

Methods: Infants with CF had chest CT scans obtained at approximately 1 year of age as part of an international, multicenter observational study. Preliminary analysis was performed on subjects from a single center (Melbourne). Disease control infants from the same center were imaged between 10 and 16 months of age using the same protocol. Bronchoscopy with lavage (BALF) was performed on CF subjects at ~4 months of age. Chest CT scans of CF infants were scored using PRAGMA-CF. Airways and lungs were automatically segmented using Apollo 2.0 lung imaging software (Vida Diagnostics Inc.) Automated software was created using a Random Forest machine learning classifier to predict "airway" or "not airway" on 20x20 pixel regions of axial chest images. AAI was calculated from the volumetric inspiratory chest CT as the proportion of these pixel regions that were scored as "airway" out of the total number of lung regions tested. Sensitivity and specificity for the algorithm to classify "airways" based on Apollo airway imaging was calculated. Student's t-test was used to compare AAI between CF and disease control. Pearson correlation coefficients were calculated between AAI, patient characteristics, PRAGMA scores, and neutrophils from BALF.

Results: CTs from 24 infants (17 CF and 7 controls) collected at a mean age of 13 months were analyzed. Mean PRAGMA-CF %disease was 1.26 in CF. Mean AAI was 7.67 in controls and 8.71 in CF (p =0.36). Sensitivity and specificity of the algorithm to predict "airways" was 93% and 95%. In the CF cohort, AAI was significantly associated with PRAGMA-CF %disease adjusted for atelectasis (r = 0.52; p = 0.03) and with absolute and %neutrophils in BALF at ~4 months (r = 0.51 p = 0.03; r = 0.64 p = 0.005). AAI was not associated with age, weight, or height (r = -0.03, r = 0.05, r = 0.02).

Conclusions: In CF infants, airway index was associated with PRAGMA-CF %disease and prior BALF neutrophils. AAI was not statistically different between CF infants and disease controls in this small cohort.

Acknowledgments: Supported by NHLBI R01HL116211 and CFF DEBOER18A0-Q.

Sharpe, N.A. 1 ; Alroy, I. 1 ; Rot, C. 1 ; Boj, S.F. 2 ; Huertas, P. 3 1. Research and Development, Eloxx Pharmacueticals, Waltham, MA, USA; 2. Hubrecht Organoid Technology (HUB), Utrecht, Netherlands; 3. Clinical Development, Eloxx Pharmacuticals, Waltham, MA, USA Approximately 13% of patients with cystic fibrosis carry at least one nonsense allele in their CFTR and have limited treatment options. ELX-02 is a designer eukaryotic specific ribosome glycoside (ESRG) with improved translational read-through properties and decreased affinity for the prokaryotic and mitochondrial ribosome. In a previous study, when tested with organoids from cystic fibrosis patients carrying homozygous or compound heterozygous nonsense mutations (HUB), ELX-02 showed up to 60% dose-dependent forskolin induced swelling (FIS) in homozygous G542X and W1282 organoids with a slight additive effect noted for VX-770 and up to 40-60% and 40% (respectively) dose-dependent swelling in heterozygous ΔF508/G542X and ΔF508/R1162X organoids with slight to no additive effect noted for VX-770. In these assays the swelling response varied among organoids with the same mutation. To help understand the varied responses, mRNA levels were measured from organoids used in the FIS assay. QPCR was performed on both homozygous and heterozygous organoids from cystic fibrosis patients to define any potential correlation between a swelling response and mRNA levels. ELX-02 (100 µg/ mL) increased total and relative wild-type (WT) CFTR mRNA (2-4X) levels in compound heterozygous ΔF508/G542X and W1282X homozygous organoids. Additionally, the results show a strong correlation between increased mRNA levels and FIS-induced swelling assay in both the compound heterozygous ΔF508/G542X (up to 40% swelling) and W1282X homozygous (up to 60% swelling) organoids. The robust response in the FIS assay correlated with increased total and relative (WT) CFTR mRNA levels. These data suggest mRNA levels vary across individual patients with the same cystic fibrosis nonsense mutations. This variability may be due to differential nonsense-mediated mRNA decay (NMD) in different individual patients. As a consequence, ELX-02 potentiation of functional CFTR translational readthrough in patients with nonsense mutations may be influenced by mRNA levels. The ALBATROSS study aims to assess the safety, efficacy and pharmacokinetics of GLPG2222 in CF patients who are on stable treatment with ivacaftor. GLPG2222 is a novel CFTR corrector, partially rescuing F508del CFTR, in clinical development for the treatment of CF Methods: ALBATROSS is a Phase IIa, multicenter, randomized, double-blind, placebo-controlled study, to evaluate two qd doses (150 mg and 300 mg) of orally administered GLPG2222 for four weeks in ivacaftortreated subjects with cystic fibrosis harboring one F508del CFTR mutation and a second gating (class III) mutation. Male and female subjects aged ≥ 18 years with CF and percent predicted FEV 1 (ppFEV 1 ) ≥ 40% were eligible for inclusion. Subjects were on stable ivacaftor treatment for at least 4 weeks and remained on ivacaftor during the study.

This study included 37 patients (7 on placebo). The majority of patients (76%) harboured G551D as Class III mutation. Overall, GLPG2222 was well tolerated, with adverse events being predominantly mild or moderate, and typical for a CF patient population. The incidence of adverse events was similar in the treatment and placebo groups. There were no serious adverse events reported and no discontinuations due to adverse events.

The targeted exposures of GLPG2222 were achieved in this patient study, with exposures in patients in line with those observed in healthy volunteers.

A statistically significant dose dependent decrease in sweat chloride concentration was observed amounting to a decrease of 6 mmol/L (p<0.05) in the 300 mg cohort, and an absolute change in ppFEV1 of +2.2%. The placebo group had an increase in sweat chloride concentration of 5.6 mmol/L and a decrease in ppFEV1 of -0.8 % (both not statistically significant). Using waterfall plots, the data on changes from baseline for both ppFEV1 and sweat chloride concentration for individual patients will be presented for the two doses evaluated and compared to placebo.

Conclusions: The ALBATROSS results demonstrate that repeated administration of GLPG2222 for four weeks is generally well tolerated. The additional activity observed with treatment with GLPG2222 on top of ivacaftor was in line with what was observed with tezacaftor combined with ivacaftor in a Phase 2 study in this population (Donaldson SH, et al. Am J Respir Crit Care Med. 2018; 197:212-24) . The data support dose selection for future investigational triple combination therapies. Background: Several biomarkers of lung injury are known to be excreted in urine. Urine is easy to collect, freely available and easy to work with making it an ideal body fluid in which to look for biomarkers. Little data exist on baseline levels or natural variability of urinary biomarkers in children or the link between biomarkers and neutrophilic airway inflammation in CF.

Overview: This is a three-part multicentre prospective CF Urinary Biomarker Study (CUBS). Part one involves measurement of urinary and bronchoalveolar lavage (BAL) biomarkers in biobanked matched BAL/ urine samples. Part two involves prospective collection of weekly urine samples at home over 3 months from children with CF and controls. Part three will be a longitudinal study to determine if candidate biomarkers can predict incident bronchiectasis.

Methods: Matched BAL/urine samples (n=40) were collected through the Study of Host Immunity and Early Lung Disease in Children with CF (SHIELD CF) from clinically stable children with CF undergoing annual BAL surveillance as previously described. For the prospective part of the study, urine samples were collected weekly at home from children with CF and healthy controls. Parents were asked to fill out a symptom diary and increase sampling to every two days when children were unwell.

The following urinary biomarkers were assayed in BAL and urine at Mologic labs: A1AT, B2M, C3L1, CC16, cystatin, HNE, Active HNE, HSA, IL-1β, MMP8, MMP9, Active MMP, MPO, NGAL, RBP4, SRAGE, TIMP1, TIMP2, Calprotectin, Creatinine, Fibrinogen, Desmosine, fMLP, CRP, RNase3, IL8, IL6, Siglec8, A1AT, RBP4, C16, B2M, Periostin and PGP. Urinary biomarkers were corrected for creatinine. Neutrophil elastase activity was measured in BAL at NCRC labs.

Results: Forty matched BAL and urine samples from children with CF were assayed. Significant correlations (p<0.05) were found between BAL and urinary levels of HSA, B2M, C3L1, CC16, CYS, MMP8, MMP9, Active MMP, MPO, IL-1β, NGAL, RBP4, SRAGE, TIMP1, TIMP2, Calprotectin, Fibrinogen, CRP, Desmosine, RNase3, Periostin and PGP. Urinary biomarkers were compared between the group of children with and without free neutrophil elastase in BAL. No significant difference in any urinary biomarker was seen between the two groups. In part two, a further 40 children with CF and 22 controls were recruited. Urine collection is almost complete for all patients. Urinary biomarker analysis will follow completion of collection. Preliminary assayed CF samples (3) have shown significant variability in biomarkers over time. Once complete, data on natural variability of urinary biomarkers, difference between CF and control and variation during illness will be presented.

Conclusion: A wide range of urinary biomarkers are produced by children with CF. Levels for many correlate between BAL and urine. Urine biomarkers cannot predict BAL free neutrophil elastase in a cross-sectional sample. Longitudinal analysis of trends is ongoing. Objectives: GLPG2222 is a novel CFTR corrector in clinical development for the treatment of CF. In vitro assays on F508del CFTR show it is highly potent in partially restoring CFTR cell surface expression. Phase I clinical studies have shown that it is generally well tolerated in healthy subjects and CF patients. The FLAMINGO study was designed to assess the safety and efficacy of GLPG2222 in CF patients.

Methods: In this Phase II, multicenter, randomized, double-blind, placebo-controlled study, we evaluated four q.d. doses (50, 100, 200 and 400 mg) of orally administered GLPG2222 for four weeks in adult CF patients homozygous for F508del. Patients with percent predicted FEV 1 (ppFEV 1 ) ≥ 40% were eligible for inclusion if not on concomitant CFTR modulator therapy within four weeks of study start. Main efficacy endpoint was the change from baseline in sweat chloride concentration.

Results: This study included 59 patients (11 on placebo). Overall, GLPG2222 was well tolerated, with adverse events (AE) being predominantly mild or moderate, and typical for this study population. The AE incidence across the treatment and placebo groups was not apparently different. There were no discontinuations due to AE. A dose-dependent decrease in sweat chloride concentration from baseline was observed (maximum decreased for GLPG2222 200 mg group [-18.3mmol /l] at Day 29, [p=0.0001 for pairwise comparison between 200 mg and placebo]). Treatment was not associated with decreases in ppFEV1, suggesting it does not induce bronchoconstriction.

Conclusions: The FLAMINGO results demonstrate that repeated administration of GLPG2222 q.d. for 4 weeks over the dose range tested is generally well tolerated in CF patients homozygous for F508del without evidence of bronchoconstriction. The observed decreases in sweat chloride concentration show treatment associated modulation of CFTR activity. These data validate GLPG2222 as an active CFTR corrector molecule and support the use of GLPG2222 in future investigational combination therapies.

Bryl, B. 1 ; Merrix, S. 1 ; Proud, D. 1 ; Marin, A. 2 ; Byrne, A. 2 ; Duckers, J. 1 

Aim: Hand grip strength (HGS) and fat free mass index (FFMI) are important indicators of skeletal muscle mass and therefore prognosis in patients with respiratory diseases including cystic fibrosis (CF). CTs have become gold standard in assessing lean body mass (LBM) in cachexia in other diseases including cancer. It provides accurate information regarding muscle mass and adipose tissue allowing for singular or longitudinal assessments over time. This information can be used to improve our understanding of LBM and its correlation with lung function and patient outcome. We propose that calculating muscle volume and density on CT imaging would correlate strongly with FFMI and HGS. This could provide important information to centres where HGS and FFMI testing is lacking. We hope to find a positive correlation between LBM obtained by CT, FFMI, HGS and lung function in adult patients with CF.

Methods: Patients of the All Wales Adult Cystic Fibrosis Centre who had a CT thorax between 2013 and 2017 within the local health-board were included. Contrast enhanced images and those with implantable foreign bodies were excluded. This pilot study included 61 patients (37 male, 24 female). The cross-sectional areas and average density (Hounsfield units) of the paraspinal musculature were measured on axial images (mediastinal window setting, width 350 HU and level 40 HU; slice thickness 0.625 mm) at T4 level (rhomboid major, erector spinae, and trapezius muscles) and the T12 level (erector spinae and trapezius muscles). These readings were compared with HGS from a Takei 5401, FFMI calculated using bioeletrical impedance (Tanita BC418ma), BMI and FEV1.

Results: Measurements from T4 showed a significant correlation with FFMI (r=0.59, p<0.001) and HGS (r=0.57, p<0.001). T12 level values also showed a strong correlation with FFMI (r=0.69, p<0.001) and HGS (r=0.70, p<0.001). Only T12 showed a significant correlation with FEV1 (r=0.48, P<0.001) and neither T4 nor T12 showed a significant correlation with BMI (p>0.05).

Conclusion: CT imaging can accurately assess skeletal muscle mass and therefore provide important prognostic information. Further investigation into its use, including reproducibility and increasing the sample size, is currently under way.

Karafilidis, J.; Nygren, K. Santhera Pharmaceuticals, Burlington, MA, USA Introduction: In chronic inflammatory lung conditions, neutrophils are abundantly present in tissue and sputum. Excessive release of neutrophil derived proteolytic enzymes accelerate lung tissue damage leading to progressive lung function decline (1) (2) .

POL6014 is a novel neutrophil elastase (NE) inhibitor, administered as a nebulized drug via the Pari eFlow ® system, in development for cystic fibrosis (CF) by Santhera Pharmaceuticals. Pre-clinical data suggest that POL6014 potently and selectively inhibits NE, reduces neutrophils and inflammatory markers in animal models (3) . Additionally, POL6014 has been shown to inhibit NE in ex-vivo BAL and sputum specimens of CF patients (4) . POL6014 was administered as single ascending (inhalational) doses (SAD), first in healthy volunteers (HV) and then in CF patients.

Objectives: Analyse safety, tolerability and pharmacodynamics of POL6014 from two completed Phase 1 studies.

Methods: Both HV and CF SAD studies were single-center, randomised, double-blind, placebo-controlled, parallel-group studies with 8 subjects per dose randomly allocated to POL6014 or placebo in a 3:1 (active:placebo) fashion. In the HV study (n = 48), the following doses were used: 20, 60, 120, 240, 480, and 960 mg; in the CF study (n = 24), patients received single ascending doses of 80, 160, and 320 mg. Inhalation was performed using eFlow ® nebuliser (Pari Pharma GmbH, Gräfelfing, Germany) around 09.00 am. In CF patients, spontaneous sputum collection was performed pre-dose, 1 to 3 hours and 24 hours after inhalation of POL6014 or placebo.

Results: No deaths and no serious adverse event (SAE) occurred during the HV study or CF study.

Safety-HV Study In the HV study, 27 AEs (24 of which treatment-emergent AEs (TEAE)) were recorded in 13 subjects (27.1%), where all of them had received active compound (POL6014). No AE was reported for subjects having inhaled placebo solution. The most frequently recorded AE was "cough" (5 events).

Safety-CF Study and Pharmacodynamics (PD) In the CF study, a total of six (6) AEs, all of them TEAEs, were recorded in six (6) patients. Five (5) of these patients (83.3%) were on active compound: one (1) patient on 80 mg POL6014 and two (2) patients each on 160 mg and 320 mg POL6014. The most frequently reported AEs were "dizziness" and "headache" with two (2) events each.

Shortly after inhalation of POL6014, a strong reduction of active NE in sputum was observed. Baseline concentrations of active NE were between 5000 and 20,000 ng/mL (0.2 -0.7 µM). Mean (± SD) changes of active NE concentrations between baseline (pre-dose) and 1-3 hours after inhalation were -77.7 (43.5)% for the 80 mg dose group, -98.7 (1.0)% for the 160 mg dose group and -81.4 (40.0)% for the 320 mg dose group (p<0.05 for all three doses).

Conclusions: In these two Phase I SAD studies, orally inhaled POL6014 was safe and well tolerated in both HVs and CF patients. This inhalational agent was safely administered to HVs in single doses ranging between 20 and 480 mg, and to clinically stable CF patients in doses ranging from 80 to 320 mg. Aims: To report on the use of a novel ProteaseTag ® Active Cathepsin G Activity-Based Immunoassay (ABI) for the quantification of active Cat G in sputum sol from patients with CF showing improved specificity over other neutrophilic proteases.

Methods: Spontaneous expectorated sputum was collected randomly from patients with CF (n=13) and processed using 5 parts phosphate buffered saline (PBS), centrifugation at 3000g for 30 minutes at 4°C, followed by aliquoting and storage of the sol supernatant at -80°C. Levels of active Cat G, NE and Pr-3 were quantified using ProteaseTag ® Active Cat G, NE and Pr-3 (ABIs) (ProAxsis Ltd). Quantification of active Cat G was also performed by kinetic assay using the fluorogenic substrate Suc-Ala-Ala-Pro-Phe-AMC (Merck) on a FLUOstar Omega plate reader (BMG Labtech) at λ ex = 380 nm and λ em = 460 nm. Specificity testing of the ProteaseTag ® Active Cat G ABI was performed by measuring Cat G levels in standard curves of active NE and Pr-3 (7.81 ng/mL to 500 ng/mL).

Results: The novel Active ProteaseTag ® Cat G ABI allowed successful quantification of active Cat G, with detection limits ranging from 7.81 ng/ mL to 500 ng/mL. The ABI showed no detection of active NE or Pr-3 at the concentrations tested, thereby indicating specificity for Cat G. The ABI enabled quantification of active Cat G within all of the CF sputum samples (n=13). Furthermore, active Cat G levels showed a positive correlation with both active NE (r=0.92) and active Pr-3 (r=0.68). In addition, whilst active Cat G concentrations quantified by ABI displayed a positive correlation with levels measured by kinetic assay (r=0.91), the kinetic assay required higher standard curve concentrations and therefore was inherently less sensitive.

Conclusions: We have a developed a novel ABI which enables the specific quantification of active Cat G in sputum sol. Levels of active Cat G positively correlate with both active NE and Pr-3 concentrations in sputum samples. This ABI provides improved analysis of active Cat G levels in clinical samples over other currently available substrate-based assays. Rationale: Pharmacological manipulation of the F508del-CFTR by the lumacaftor/ivacaftor combination therapy corrects the mutated protein but modestly improves the clinical status of patients homozygous for F508del. Moreover, a proportion of patients prove nonresponsive to the lumacaftor/ ivacaftor combination treatment. The underlying cause of this variability is unknown and no predictors for therapeutic outcome are available yet.

Objectives: To determine factors involved in the clinical response to lumacaftor/ivacaftor therapy.

Methods: We evaluated 41 children homozygous for F508del over 12 years beginning lumacaftor/ivacaftor treatment at baseline and at 6-months treatment. CFTR activity was assessed at baseline and at 6 months, based on sweat chloride (Cl -) concentration and β-adrenergic peak sweat secretion, nasal potential difference (NPD), rectal biopsy for intestinal short circuit (Isc) current measurements (ICM). Rescue of CFTR activity in patients' primary nasal epithelial (HNE) cells after incubation with lumacaftor was compared to the clinical outcome and the change of in vivo CFTR biomarkers at 6 months therapy. Variation in respiratory improvement was correlated to seric levels of lumacaftor/ivacaftor.

The sweat Clconcentration showed a significant decrease averaging 20 mEq/L (3.9). This improvement was not accompanied by an improvement in the β-adrenergic peak sweat response. There was no correlation between the improvement in sweat Clconcentration and the change in percent predicted (pp) FEV 1 .

The level of functional rescue for F508del-CFTR activity reached up to ~20% of normal in the nasal epithelium and ~15% in rectal biopsies, but there was no significant correlation with the change in FEV 1 at 6 months. Correction of CFTR in primary nasal cells cultures was significantly higher in the patients improving FEV 1 by more than 5% than in those whose FEV 1 variation was less than 5% (26% versus 3.5%, p=0.006) and correlated significantly to FEV 1 absolute change. Drug seric levels increased significantly at 4 hours but this did not correlate with the improvement in ppFEV 1 nor modification of any biomarker of CFTR activity.

Conclusion: This first real life study highlights factors involved in variability of the response to lumacaftor/ivacaftor in F508del homozygous patients. This innovative approach, integrating the patient-specific respiratory cell background, will provide a path to identify subjects with CF who may benefit from CFTR-modulating therapies. The leading cause of morbidity and mortality is chronic lung disease. Most current treatments focus on symptom management and treating secondary disease complications, and few therapies are available that restore CFTR function. Lentiviral vectors are attractive tools for gene therapy because they integrate into the host's genome, which ensures lasting gene correction. CFTR gene addition using lentiviral vectors has shown promising results in CF pigs when the vector is aerosolized in the airways. Challenges for lentiviral gene transfer include optimizing the transduction efficiency and developing methods for high titer vector production.

One strategy to improve transduction and cell targeting is by taking advantage of pseudotyping. Lentivirus vectors can accommodate envelope glycoproteins from other viruses to modify tropism. Currently, the baculovirus envelope protein GP64 has shown the greatest ability to transduce human airway epithelial cells. The objective of this work is to screen and identify additional viral envelopes with similar or improved ability to transduce airway epithelia from the apical surface as GP64, and attain vector titers >10 9 TU/mL. Methods: We produced lentivirus expressing secreted Gaussia luciferase (GLuc) pseudotyped with GP64, Ross River virus (RRV), vesicular stomatitis virus G protein (VSVG), cocal, or two different modified baboon endogenous virus glycoproteins (BaEvTR and BaEvRless). These vectors were then used to transduce well differentiated primary cultures of airway epithelial cells from at least four different human donors at equivalent MOIs from the apical or basolateral surface. Five days after transduction, the apical surface of the cultures was washed with PBS and the relative light units were measured and compared.

Results: RRV, BaEvTR, and BaEvRless envelopes conferred apical transduction similar to GP64 and all produced high titer vectors. The RRV envelope glycoprotein receptor is not well characterized but a single amino acid substitution was shown to broaden its tropism. This makes it an interesting candidate to introduce other modifications that might further increase transduction efficiency in airway cells. Consistent with our results, the receptors for the BaEvs are the neutral amino acid transporters 1 and 2 (ASCT-1/2), which are expressed in airway epithelial cells.

Conclusions: Using a lentiviral pseudotyping screening assay that uses secreted GLuc as a measure of viral transduction, we identified three glycoproteins that produce transduction levels from the apical surface of epithelial cells similar to GP64, and produce high vector titers. Ongoing work is focused on quantifying the number and cell types transduced by each vector pseudotype. Identification of glycoproteins that can apically transduce airway epithelial cells and are compatible with high titer production will help advance lentiviral gene therapy for CF towards clinical applications.

Background: FDL169 is a novel CFTR corrector that has been shown to increase CFTR-mediated chloride flux in-vitro in homozygous F508del-CFTR hBE cells. Furthermore, the combination of FDL169 with the CFTR potentiator FDL176 has similar in vitro efficacy to lumacaftor/ivacaftor when hBE cells were treated for ≤1 hour with each potentiator but higher efficacy when treated for ≥24 hours. Here we report the initial clinical trial results of FDL169.

Methods: This Phase 1b study evaluated safety, pharmacokinetics (PK) and pharmacodynamics of multiple doses of FDL169 for 28 days in subjects with CF homozygous for F508del-CFTR mutation, aged ≥ 18 years old, with sweat chloride value ≥60 mmol/L and forced expiratory volume in 1 sec (FEV 1 ) ≥ 40% of predicted. Twenty-seven subjects were enrolled. In cohort 1, 15 subjects were randomized 2:2:1 to either 400 mg TID, 600 mg TID or placebo; in cohort 2, 12 subjects were randomized 2:1 to 800 mg TID or placebo. Primary objective was safety and tolerability; secondary objective pharmacokinetics (PK); exploratory objectives: changes in sweat chloride and predicted FEV 1 , and respiratory symptoms as measured by the Cystic Fibrosis Questionnaire-Revised (CFQ-R).

Results: One serious adverse event was reported, acute appendicitis unrelated to study drug; no discontinuations due to study drug were reported. Data including safety, tolerability, PK and exploratory endpoints are expected in advance of the North American Cystic Fibrosis Conference 2018.

Conclusion: Safety and PK data from this study will support further clinical development of FDL169 for treatment in CF. Introduction:P. aeruginosa (Pa) lung infection exhibits increased antimicrobial resistance (AMR), which has led to an interest in novel therapies, including application of bacteriophage (phage). In most laboratories, antibiotic sensitivities are tested in planktonic culture; however, in the CF lung Pa employs social strategies like the formation of biofilm to survive in adverse environments. It is known that subinhibitory concentrations of antibiotics can induce bacterial biofilm formation as a survival strategy (Hoffman LR, et al. Nature. 2005; 436:1171-5) . In this experiment, we wished to explore potential interactions between antibiotics at subinhibitory concentrations when combined with low concentrations of a lytic phage cocktail in both planktonic and biofilm modes of growth.

Methods: We chose 6 clinical Pa strains isolated from CF patients at the Royal Brompton Hospital on the basis of (1) sensitivity to the published MIC of both ceftazidime and tobramycin (cef and tob), and (2) sensitivity to a neat phage cocktail (4x10 9 PFU/mL) on standard plaque assay. For assessing planktonic growth, we incubated 96-well plates with 2-fold dilutions of the MIC (to 1/16 MIC) of either cef or tob along with dilute phage preparations (4X10 2-6 PFU/mL) in a checkerboard format for 16 hours. For biofilm, we incubated 96-well plates with Pa alone for 24 hours, and applied 2-fold dilutions of either antibiotic (from 8xMIC to 1/2MIC) along with phage for 16 hours. Wells were stained with crystal violet for biomass quantification.

Results: Each experimental condition was repeated in quadruplicate on 3 separate occasions and appearance categorised descriptively as either no obvious effects, synergistic/additive or inhibitory. Results have been reported with correlation across biological replicates. In total, 4 of the 6 clinical strains showed that addition of low levels of phage enhanced the killing effects of subinhibitory concentrations of tob or cef in planktonic and biofilm. Addition of phage to tob in 1 clinical isolate showed increased biomass suggesting an inhibitory interaction in biofilm. Notably, 1 strain did not show sensitivity to the phage at the concentrations used in this experiment; however, demonstrated synergy with tobramycin in planktonic phase and had moderate effects at biofilm reduction with both antibiotics.

Conclusion: With conventional antimicrobials, choice of agents to target different mechanisms is important to limit the emergence of AMR. In this preliminary study on 6 clinical Pa strains, we discovered substantial variability in response to co-exposure to a bacteriophage cocktail and cef or tob. Whilst in most strains, presence of both antibiotic and phage appeared beneficial, in one there was an inhibitory effect with tobramycin in biofilm. In order to optimise treatment of CF patients colonised with Pa, we think that antimicrobial regimens should be considered on a personalised basis.

Responses to clinically applied phage should be assessed with real time airway culture samples alongside clinically relevant outcomes.

Cohen, I. 1 ; Milla, C. 2 ; Salathe, M. 3, 5 ; Baumlin, N. 3, 5 ; Chung, S. 3 The maintenance of a thin layer of airway surface liquid (ASL) in the ciliated respiratory epithelium as well as proper mucus hydration are critical for appropriate function of the mucociliary apparatus. Cystic fibrosis airway disease is caused by airway dehydration related to the dysfunction of mutant CFTR, resulting in buildup of thick mucus and an inability to clear it from the airways. Simply put, CFTR potentiators improve the function of classes 3 -5 CFTR mutants while CFTR correctors improve trafficking of F508del CFTR, a class 2 mutant. These CFTR modulators are not expected to be active in patients with nonsense mutations (class 1). Brevenal is a small molecule with activity in mucociliary clearance that potentiates calciumactivated chloride channel (CaCC) currents through TMEM16A or ANO-1. In this study, we examined the activity of brevenal in fully differentiated patient-derived cystic fibrosis human bronchial epithelial (CFBE) cells with nonsense mutations in at least one allele (G542X/F508del, G542X/R1162X and R1162X/Unknown). These cells were treated with brevenal, either 500 nM basolaterally or 1 µM apically. ASL depth was measured using meniscus scanning. Basolateral application of 500 nM of brevenal (n=7, from 2 lungs with G542X/F508del and R1162X/unknown) significantly reduced ASL absorption over 24 hours. Brevenal was formulated for apical administration via nebulization similar to our previous in vivo application with cyclodextran. Two-dimensional lateral diffusion of FITC-labeled 70 kD dextran was measured in the ASL on CFBE cells from G542X/ F508del and F508del/R1162X donors by fluorescence recovery after photobleaching (FRAP). One hour after apical administration of 1 µM brevenal, the half-life (t 1/2 ) of fluorescence recovery after photobleaching was reduced compared to control. Ciliary beat frequency (CBF) was measured after FRAP analysis via SAVA software. CBF significantly increased 2 hours after apical administration of 1 µM brevenal. Taken together these findings indicate that by potentiation of TMEM16A activity brevenal corrects the defects of ASL volume and rheology characteristic of mutated CFTR. Alternatively, brevenal may have effects independent of ion transport through apical channels. In either case, the beneficial effects on parameters of mucociliary function seen in vivo and in vitro demonstrate that brevenal may be useful for the treatment of CF airway disease, especially for patients with nonsense mutations currently not amenable to CFTR modulator therapy. Oropharyngeal swabs are much more easily obtained, and the oropharynx may serve as a source of lower airway microbiome constituents through micro-aspiration, especially during infancy. The potential for the oropharyngeal microbiome to serve as a noninvasive, informative biomarker in early stage CF respiratory disease warrants further investigation. To date, studies of the microbiome in CF have largely depended on 16S sequencing, which is less informative than metagenomics shotgun sequencing (MGS). Unlike 16S, MGS is capable of identifying viruses and fungi, as well as species-level identification and functional capabilities of the microbiome constituents. We present metagenomic sequencing data from oropharyngeal swabs from infants with CF.

Oropharyngeal swabs (Copan e-Swabs, white top) were obtained from 10 infants with CF. DNA isolation was performed on swab media using a modified bead beating approach and the DNA Powerlyzer Powersoil kit (Qiagen). Libraries were prepared using the NEB Next Ultra kit, and paired-end sequencing was performed using the Illumina MiSeq Platform. Reads were filtered, assembled into contigs, and annotated using the WEVOTE platform (Metwally A, et al. PLoS One. 2016; 11e0163527) .

Results: From 10 samples, 1,759 unique taxa were identified. Microbial DNA accounted for >50% of reads in 9/10 samples, and >90% of reads in 7/10 samples. The most abundant families and species are listed in the Table. Infants with previous hospitalization trended towards a decrease in Shannon Diversity (p < 0.1) and were enriched with Sanguibacter and Campylobacter species.

Conclusions: Oropharyngeal swabs provide sufficient microbial content for MGS, and findings are consistent with previous literature on oral flora and CF. While there is an association with hospitalization history, further work must be done to elucidate features of the oropharyngeal microbiome associated with disease status, severity, and genotype. We have developed biosynthetic unmodified mRNA that can be aerosolized and efficiently delivered to lung tissue when encapsulated by a lipid nanoparticle, enabling the endogenous production of cystic fibrosis transmembrane conductance regulator (CFTR) for the treatment of patients with CF.

Objective: To demonstrate that lipid nanoparticles efficiently delivered mRNA encoding human CFTR protein (MRT5005) to the lungs in cells, rodents and nonhuman primates resulting in functional protein expression.

Methods: Quantitative PCR data were obtained for exogenous codon-optimized human CFTR mRNA and compared to endogenous levels of host species CFTR mRNA. Human CFTR-specific immunohistochemical (IHC) staining was performed in rat and nonhuman primate studies to determine presence and distribution of hCFTR mRNA-derived hCFTR protein. In vitro chloride-ion channel activity measurements were conducted in Fischer rat thyroid (FRT) and human bronchial epithelial (HBE) cells. Primary HBE cells were cultured and differentiated at an air-liquid interface and treated with MRT5005. Nasal potential difference (NPD) testing in CFTR -/-Sprague Dawley rats was used to estimate CFTR activity after blinded treatment with either MRT5005 or vehicle control. Fluorescent confocal imaging of nonhuman primate lung tissue was performed on hCFTR protein and ZO-1 (tight junction membrane) protein for localization studies.

Results: We achieved supraphysiologic CFTR mRNA levels, as evidenced by significant increases above endogenous CFTR mRNA in normal rats and nonhuman primates. Dose-dependent CFTR expression was observed by IHC in lung tissue, with expression up to 28 days at the higher doses following a single inhaled dose. We have also demonstrated that biosynthetic CFTR mRNA produced fully functional human CFTR protein that was electrochemically active by triggering chloride channel function in both cultured FRT cells and in primary CF epithelial cells. In CFTR -/rats, multiple-dose treatment (QOD x 14 days) of MRT5005 delivered by nasopharyngeal and oropharyngeal administration significantly improved CFTR-dependent chloride transport measured by NPD. Finally, biosynthetic CFTR mRNA lipid nanoparticles were safe in multiple-dose studies in rats and nonhuman primates with no adverse effects seen up to, and including the highest doses tested. These results support the first mRNA therapeutic advanced for the treatment of patients with CF.

Conclusions: A novel lipid nanoparticle efficiently delivered mRNA encoding human CFTR protein to the lungs in rodents and nonhuman primates. The biosynthetic unmodified mRNA encapsulated in a lipid nanoparticle can be aerosolized and delivered to lung tissue, enabling the endogenous production of CFTR for the potential treatment of patients with CF. These data support an ongoing Phase 1/2 study of nebulized MRT5005 in adult patients with CF. Additional information about this study can be found at www.clinicaltrials.gov, NCT03375047. Acknowledgment: Sponsored by Translate Bio, Inc. Ten percent of all cystic fibrosis (CF) patients carry an in-frame premature termination codon (PTC), which leads to the production of truncated, nonfunctional CFTR protein. Compounds have been identified that can suppress termination at PTCs (also called readthrough (RT)) and partially restore the synthesis of full-length protein. However, the amount of full-length CFTR restored by current RT agents is insufficient to correct the CF phenotype. In collaboration with Southern Research (SR), highthroughput screening (HTS) was performed to discover new, efficient RT compounds. Initially, an FDA-approved 1,600-drug library was screened using luciferase-based RT reporters. Using identical RT reporters, eight compounds were identified to induce RT in Fischer rat thyroid (FRT) cells compared to only two in HEK293 cells, suggesting that FRT cells represent a highly sensitive cell line for identifying RT compounds. Due to their responsiveness, FRT cells were chosen for screening a much larger 775,000-compound library. By using a permissive cell line, we anticipated that the RT effects displayed in FRT cell lines might not necessarily be replicated in other cell lines, such as human bronchial epithelial (HBE) cells. However, the molecular structure of the initial RT hits could be modified to increase RT in other cell lines. For example, SRI-37240, a hit from the 775,000 HTS, increased RT in FRT cells up to 10-fold, but it was not active in immortalized 16HBE14o-cells. However, SRI-41315, a derivative of SRI-37240, induced a 15-fold increase in RT in 16HBE14ocells. This indicates that FRTs represent a good cell line for performing an initial HTS since we are less likely to miss any potential scaffolds that are active for RT. We also compared the level of RT induced in 16HBE14oand primary HBE cells by the G418 aminoglycoside control. Current results suggest that similar levels of RT are generated in both cell lines, indicating that RT data correlates well between the cell lines.

PTCs also frequently elicit nonsense-medicated mRNA decay (NMD), a pathway that degrades PTC-containing mRNAs, which further diminishes the level of CFTR protein. Compounds may be identified that not only induce RT, but also inhibit NMD. Theoretically, these dual function compounds have the potential of restoring higher levels of CFTR function than compounds that induce RT alone. We have developed and characterized NanoLuc-based reporters to identify compounds that induce RT and also stabilize CFTR mRNA. The potential of these compounds is currently under investigation. In addition, we are testing SRI-41315 in CFTR-G542X mice to verify its ability to induce RT in vivo. Overall, it is hoped that this approach will identify compounds that can be developed into a therapy for CF patients who harbor a PTC. The goal of this project is to construct a reporter system that measures termination efficiency at premature termination codons (PTCs) versus normal termination codon (NTCs) in order to test whether readthrough (RT) compounds affect PTCs and NTCs differently. Here, we have generated a luciferase-based reporter system which differentiates termination at PTCs and NTCs and are stably expressed in Fischer Rat Thyroid (FRT) cells. The initial characterization of the reporters has been conducted using RT drugs and knockdown (KD) of UPF1, UPF2, UPF3b and PABPC with siRNAs. In addition, eight newly identified RT hits from our recent high-throughput screen (HTS) with Southern Research have been tested with this reporter system.

Generally, all PTCs and NTCs consist of the same three stop codons (UGA, UAG, or UAA). However, PTCs and NTCs remain structurally and mechanistically different. The termination codon closest to the poly(A) tail is defined as the NTC, while the one further away from the poly(A) tail is defined as the PTC. In our reporter system, three individual luciferase genes, Red Firefly (1647 bp), Cypridina (1662 bp) and Gaussia (558 bp) were cloned into the mammalian expression vector, pcDNA3.1zeo+, such that Red Firefly is more distal from the poly(A) tail than Cypridina, while Cypridina is further away from poly(A) tail than Gaussia. A UGAC termination signal is added at the 3' end of Red Firefly and defined as a PTC, while a UGAC added at the 3' end of Cypridina is defined as the NTC. The RT can be measured by Gaussia activity, normalized to Red firefly activity, for both PTC and NTC reporters.

First, we tested the RT effect of the well-known RT agent, G418. Our results showed that G418 increased RT at both PTC and NTC reports in a dose-dependent manner. Interestingly, G418 at a 300 mg/mL dose induced higher RT at the PTC. To examine whether the adjacent proteins around PTCs or NTCs may affect termination efficiency, we are generating KD of these proteins using siRNAs. We are currently investigating the effect of UPF1, UPF2, UPF3B and PABPC KDs alone on RT of a PTC versus NTC, and in combination with G418 treatment. Plus, we are also building additional constructs in which a spacer gene, lac Z, is inserted in order to increase the distance between the PTC or NTC and poly(A) tail to better understand the difference of RT at PTCs and NTCs.

Finally, we tested eight RT hits identified from a previous HTS in this PTC/NTC reporter system. The eight RT hits increased RT at the PTC reporter by 10 to 25-fold, while all eight hits only increased the RT at NTC 3 to 4-fold. These results show that all these RT compounds induce higher RT at the PTC than at the NTC in reporter cells. Overall, we believe these reporters will provide valuable new information about the specificity of RT agents for future drug development. Introduction: Chronic airway inflammation is a major pathogenic factor of CF airway disease. Airway macrophages (AMs) contribute to CF airway inflammation by producing inflammatory mediators. The spliced (active) form of X-box binding protein-1 (XBP-1s) is a potent transcription factor required for cytokine [e.g., interleukin-6 (IL-6), and tumor necrosis factor-α (TNF-α)] production in human CF AMs. Moreover, the inducible nitric oxide synthase (NOS2) is associated with inflammatory responses in macrophages and its expression is increased in AMs from CF mice. Sphingosine-1 phosphate kinase 1 (SPHK-1), a kinase that phosphorylates the sphingolipid metabolite sphingosine-1, is involved in immune responses and associated with increases in inflammatory mediators. Vardenafil (VAR), a phosphodiesterase type 5 inhibitor, decreases lung inflammatory responses in F508del CF mice, but it is not known whether VAR blunts XBP-1s, cytokines, NOS2 and SPHK-1 levels in human CF AMs.

Hypothesis: This study tested whether VAR decreases inflammatory responses associated with up-regulation of XBP-1, NOS2, SPHK-1 and cytokine production in non-CF and CF human AMs.

Methods: AMs were isolated from excised non-CF and CF human lungs and maintained in primary culture for 3 days. To simulate their response to infection, AMs were exposed to supernatant from mucopurulent material (SMM) from human CF airways for 48 hours and treated with VAR (10 µM) for 24 hours. Subsequently, the following inflammatory biomarkers were evaluated: 1) the mRNA levels of XBP-1s, NOS2 and SPHK-1 using quantitative RT-PCR, and 2) IL-6, IL-8 and TNF-α protein secretion using ELISA assays.

Results: CF AMs exhibited higher baseline levels of XBP-1s, NOS2 and SPHK-1 mRNA and cytokine secretion. SMM up-regulated XBP-1s, NOS2, and SPHK-1 mRNA levels and stimulated IL-6, IL-8 and TNF-α secretion in both non-CF and CF AMs. VAR decreased SMM-up-regulated XBP-1s, NOS2 and SPHK-1 mRNA levels, and SMM-increased IL-6, IL-8 and TNF-α secretion in non-CF and CF AMs. These data are summarized in the Table. Conclusions: VAR attenuates IL-6, IL-8 and TNF-α secretion induced by the infectious/inflammatory CF airway milieu in non-CF and CF AMs. These responses are associated with decreased XBP-1s, NOS2, and SPHK-1 expression. Our findings suggest that VAR may be therapeutically beneficial for counteracting the robust airway inflammation of CF patients. Acknowledgment: Funded by the CFF. CFTR is an ion channel expressed in the apical membrane of epithelial cells. Mutations in CFTR cause CF, a disease in which the encoded CFTR suffers from severe folding, functional, and stability defects. The CFTR-associated ligand (CAL) is a regulatory protein that directly interacts with post-endocytic CFTR and targets it for lysosomal degradation (Cheng J, et al. J Biol Chem. 2004; 279:1892-8) . Although there are FDA-approved drugs that improve CFTR folding and function (e.g. tezacaftor and ivacaftor), there are no available therapeutics that directly address the instability of mutant CFTR. Previously, we have shown that selectively perturbing this interaction (via delivery of a competitive peptide inhibitor) leads to increased expression and function of CFTR at the apical membrane (Cushing P, et al. Angew Chem Int Ed Engl. 2010; 49:9907-11) . However, the therapeutic potential of our first-generation CAL inhibitor has been limited by its modest affinity for CAL (Ki = ~40 µM). Using a host of biochemical, structural, and computational approaches, we reengineered our inhibitor and incorporated non-natural amino acids and chemically-modified scaffolding residues. Without loss of specificity, we enhanced the affinity for CAL by 10-fold (Ki = ~4 µM). We also screened >50,000 small molecules from which we identified a subset of compounds that bind to and inhibit CAL. One promising lead inhibits CAL with 1 µM affinity and shows no cellular toxicity at therapeutic concentrations. Complementing existing CF therapies with a CFTR "stabilizer" (i.e. CAL inhibitor) may offer a novel approach for augmenting lung function in a broader spectrum of CF patients.

This work was funded in part by the NIH (R01-DK101451, T32-GM008704, and P30-DK117469) with support from the Novartis Institutes for Biomedical Research.

Singh, A.K. 1 ; Hwang, T. 5 Cystic fibrosis (CF) is a life-threatening, genetic disease caused by mutations in the gene encoding the cystic fibrosis transmembrane conductance regulator (CFTR) protein. This results in abnormal transport of chloride across epithelial cells, leading to dehydration of airway surface liquid and impaired mucociliary clearance. The F508 deletion in CFTR is the most prevalent mutation present on at least one allele in 90% of CF patients. Maximal restoration of F508delCFTR requires two corrector molecules with complementary mechanisms, type 1 (C1) and type 2 (C2), to increase CFTR at the cell surface, in combination with a potentiator (P) to increase channel activity (1) . AbbVie and Galapagos have been developing multiple CFTR modulators, for patients with homozygous and heterozygous F508delCFTR mutations, to explore in triple combination therapies.

For these modulators, we have ongoing efforts to understand the mechanism(s) of action (MoA) that underlie their beneficial effects on CFTR channel function. Total CFTR ion channel function can be derived from the following equation: I = n*Po*i where, I = total current; n = number of channels; Po = open probability; i = single-channel amplitude Our current understanding is that the CFTR modulators on the market or being developed (1) are not likely to affect the conductance (γ) of the CFTR ion channel and hence the single-channel amplitude (i) at a given transmembrane potential. Therefore, the two parameters that work in concordance to yield efficacy as studied with primary hBE cells in TECC and Ussing chamber measurements are that of "n" and "Po." In our pursuit of discovering and developing these C2 CFTR correctors, we have encountered compounds that affect both of these parameters to varying degrees (synergism with C1 correctors and CFTR potentiators). Data for representative examples will be presented.

Acknowledgments: AKS, CMB, AS, SA, YF, TN, TAV, XW and CT are employees of AbbVie. The design, study conduct, and financial support for the research conducted by AbbVie were provided by AbbVie and Galapagos. AbbVie participated in the interpretation of data, review, and approval of the publication. Methods: In vitro internalization of labeled EpL conjugates was evaluated by fluorescence microscopy and by on-cell Western internalization assay. ENaC currents were measured in Ussing chambers using fully differentiated human bronchial epithelial (HBE) cells cultured at the air-liquid interface (ALI). Rats (wild-type [WT] or ovalbumin-sensitized) received oropharyngeal or nebulized doses of conjugates; mRNA and protein expression was analyzed by qPCR and immunohistochemistry.

Results:αvβ6 ligands facilitate receptor-mediated endocytosis of conjugates by epithelial cells in tracking and receptor internalization assays. EpL-RNAi trigger conjugates are internalized by HBE cells in ALI culture. A dose reducing αENaC mRNA expression by ~40% produces similar reductions in ENaC currents. In vivo, the EpL platform produces significant reductions in whole-lung αENaC mRNA expression, increasing potency of the trigger by tenfold and improving uniformity of knockdown. Immunohistochemistry studies demonstrate loss of airway αENaC protein at doses that reduce whole lung αENaC mRNA expression by ~50%, with remaining αENaC protein limited to the distal alveolar epithelium. Reduction of lung αENaC mRNA expression is durable, maintaining >50% knockdown at 3 weeks post-dose and requiring 6-7 weeks for recovery to baseline. After aerosol inhalation of a single 0.16 mg/kg deposited dose of EpL-αENaC RNAi conjugate we observe >50% whole-lung knockdown with no changes in renal αENaC or CFTR mRNA expression.

Conclusions: EpL-RNAi conjugates employing αvβ6 ligands improve functional delivery of αENaC RNAi triggers to the airway epithelium, reducing ENaC currents, producing durable mRNA knockdown and silencing airway ENaC protein expression. ARO-ENaC, Arrowhead's first therapeutic candidate to employ the pulmonary epithelial delivery platform, has been advanced as a novel treatment for cystic fibrosis lung disease. Furthermore, the ability of the EpL-RNAi platform to facilitate functional delivery of RNAi triggers to the airway suggests that additional therapeutic targets in the pulmonary epithelium could be considered, particularly those that are currently inaccessible to traditional small molecule or antibody approaches. In CF, airway dehydration and abnormally viscous and adhesive mucus are thought to cause a profound decrease in mucus transport, precipitating progressive lung disease. Characterizing mucus transport abnormalities in CF patients has been difficult, particularly at the cellular level, due to the limitations of current imaging tools in patients. We have recently developed an imaging technique called micro-optical coherence tomography (µOCT) that obtains cross-sectional images of tissue reflectance from the airways in real time at 1 µm resolution. Here, we investigate whether the intensity of µOCT images of mucus is a reliable indicator of sputum viscosity.

Sputum was obtained from CF patients (spontaneous expectoration, N=9) and normal volunteers (3% HTS induction, N=5). Sputum were imaged by µOCT and the reflectivity compared to dynamic viscosity determined by cone and plate rheometry at near-static shear (shear rate 0.12 Hz). A semi-flexible µOCT probe was used to image mucus in the nasal passages of CF patients (N=10) and an equal number of age-and sex-matched controls. Reflectance intensity of the µOCT images of mucus was quantified by 1) correcting reflectivity for depth-dependent attenuation; 2) computing the intensity histogram of sputum or mucus; 3) calculating mean intensity; and 4) fitting the histogram to a negative exponential function, a measure of the distribution of reflective vs. non-reflective mucus.

Dynamic viscosity was elevated in CF (mean CF: 60.9±113.9 vs mean non-CF: 1.4±2.3; P<0.01) and reflectivity increased as indicated by a rightward shift of the histogram curve (ANOVA P<0.0001). Dynamic viscosity was related to mean mucus intensity in a semi-log fashion (R 2 =0.39, P<0.05). Dynamic viscosity was also related to negative exponential function (R 2 =0.42, P<0.05). These results indicated reflectance intensity and distribution could be used as a µOCT proxy for viscosity. Reflectance intensity of ASL mucus measured from µOCT images of the nasal passages of humans, obtained in vivo, revealed that CF subjects exhibited greater mucus reflectivity than healthy controls, as indicated by rightward shift of the mucus pixel intensity histogram (ANOVA P<0.0001), increased mucus reflectance intensity (131.2% change, P=0.02), and comparison of negative exponentials (mean healthy: -0.10±0.03 vs CF: -0.06±0.03, P=0.001), by subject. In summary, reflectance intensity analysis of airway mucus measured by µOCT is a viable means of assessing native mucus viscosity in vivo in human patients and is highly elevated in the nasal airways of CF subjects.

Acknowledgements: This work was supported by the NIH (T32HL105346, 5R01HL116213, R35HL135816, P30DK072482, and UL1TR001417) and CFF (TEARNE16XX0, ROWE14Y0, and ROWE16XX0). Medicine Center Utrecht, University Medical Center Utrecht, Utrecht, Netherlands; 3. Foundation Hubrecht Organoid Technology, Utrecht, Netherlands; 4. Galapagos NV, Mechelen, Belgium In the last decade impressive therapeutic benefit has been achieved for people suffering from cystic fibrosis (CF) by cystic fibrosis transmembrane conductance regulator (CFTR)-modulator compounds. Currently, three therapies (KALYDECO ® (ivacaftor), ORKAMBI ® (lumacaftor/ivacaftor) and SYMDEKO TM (tezacaftor/ivacftor)) have been approved for people with CF that harbour at least one out of the 33 registered CFTR mutations and are homozygous for F508del-CFTR, respectively. Despite the approval of these therapies and reported significant clinical improvements by ivacaftor, clinical benefits of lumacaftor/ivacaftor remains suboptimal thereby leaving a large population of the CF population without effective treatment options.

To further improve potential benefits for CF patients, Galapagos (GLPG) and AbbVie (ABBV) initiated the development of combinations of a potentiator and two complementary acting correctors. Two distinct sets of correctors (ABBV/GLPG C1, C2a and C2b) with complementary mechanisms and different potentiators (ABBV/GLPG P2 and P4) are in development which synergistically restore CFTR function in primary human bronchial epithelial cells (HBECs). Here, we will present our preliminary data on CFTR function restoration by candidate combination therapies of these compounds in patient-derived intestinal F508del/F508del-CFTR organoids by using the (i) forskolin-induced swelling (FIS) assay and our recently developed (ii) steady-state lumen area (SLA) assay. Various compound combinations were assessed in the FIS and SLA assays, either in the absence (SLA) or presence (FIS) of three different forskolin concentrations, to study maximal CFTR function restoration and potential synergy of compound combinations and compared to SLA measurements of wild-type (WT) or carrier (WT/F508del) organoids or FIS responses of lumacaftor/ ivacaftor-treated organoid cultures.

Additionally, combinations of these CFTR modulators were also evaluated by FIS in organoids harbouring an F508del allele in combination with potentiator nonresponsive alleles of which several examples will be presented. Introduction: Loss of function of the CFTR gene causes cystic fibrosis (CF), a rare disease affecting multiple epithelia including the airways. A mRNA replacement therapy offers many advantages to treat CF patients including: no risk of genomic integration, bypassing the cell's transcriptional regulation and the ability to use the cell's translational machinery to produce complex proteins. Arcturus' LUNAR ® lipid-mediated technology has the potential to deliver functional copies of CFTR mRNA into airway epithelial cells and thereby restore the normal airways physiology in CF patients, an approach that is CFTR mutation agnostic. Here, we report proof of concept for delivery of nebulized LUNAR ® encapsulated mRNA into the murine airway epithelium and the identification of a lead candidate CFTR mRNA.

Methods: Novel CFTR mRNA constructs were designed based on codon and UTR optimization approaches. In vitro screening was performed in transfected CFBE and FRT cells, and their protein lysates were immunoblotted for C-band detection of the mature CFTR protein. Transepithelial conductance (Gt) was assayed in FRT cells transfected with lead candidate CFTR mRNAs, where Gt responses were measured after activation with forskolin followed by VX-770 and inhibition with Inh-172. Screening of LUNAR ® formulations was performed using a GFP mRNA in intratracheally-dosed wild-type (WT) mice. GFP immunohistochemistry and an independent histopathological analysis were used as main endpoints to determine efficacy in vivo.

Results: Optimized CFTR mRNAs transfected in CFBE-cells improved protein expression, over the native sequence, for up to 72hours. Candidate mRNAs were transfected into FRT cells and chloride channel activity was monitored. Although C-band protein expression was similar among mRNA compounds, their functional activity varied from 0.5mS/cm 2 to ~6mS/cm 2 , indicating that codon-optimization can have an impact on protein function.

The efficacy of our nebulized LUNAR ® formulations was confirmed by GFP immunostaining in the murine epithelial airways with consistent staining from trachea to main stem bronchi and terminal bronchioles. No other lung structure was stained for GFP. Treated and control animals did not show any histopathological discrepancy.

Conclusions: We are applying Arcturus' LUNAR ® and mRNA platforms to develop an mRNA replacement therapy for CF. The selective delivery of nebulized formulations into the murine lung epithelial cells validates Arcturus' LUNAR ® technology as a potential lipid-delivery platform for CF. Codon-optimization is a feasible approach to develop improved CFTR sequences with higher protein levels and active chloride channels that might be beneficial for quick mucus clearance in CF patients' lungs. We are currently establishing the efficacy and translatability of our LUNAR ® technology in other animal models with clinical relevance for CF. Introduction: Cystic fibrosis-related diabetes (CFRD) impacts ~50% of adult cystic fibrosis (CF) patients and is one of the most important extrapulmonary complications in CF. Chronic kidney disease (CKD) is a primary pathophysiologic feature of CFRD; is 2-3 times more prevalent in CF patients than in the general population; and is also caused by recurrent use of nephrotoxic drugs such as antibiotics and immunosuppressants following lung transplant surgery (Quon BS, Aitken ML. Paediatr Resp Rev. 2012; 13:206-14) . Therefore, accurate monitoring of kidney function is an important component in the care of CF patients and is likely to increase in importance as the CF patient population continues to age. Unfortunately, current clinical assessments of renal function (i.e., serum creatinine) are known to be insensitive to early-stage CKD. We have previously shown that diffusion tensor imaging -magnetic resonance imaging (DTI-MRI) can sensitively detect early-stage CKD in other patient groups (Lu L. Am J Nephrol. 2011; 34: 220-5) . Therefore, in this initial study, we evaluated the capability of DTI-MRI to detect early-stage CKD in CF patients with clinically normal kidney function in comparison to healthy, non-CF control subjects.

Objective: To determine whether DTI-MRI can detect early-stage CKD in CF patients.

Methods: Adult cystic fibrosis patients (7 female, 4 male; age=16-52) were recruited from the Leroy W. Matthews Cystic Fibrosis Center at Rainbow Babies and Children's Hospital in Cleveland. All CF patients had normal kidney function (eGFR>100 mL/min/1.73m 2 ). Healthy non-CF subjects were recruited as controls (3 female, 2 male; age=26-34). Coronal DTI-MRI data were obtained for each subject's kidneys (b=0, 400 s/mm 2 , 12 directions, 8 imaging slices/subject). A region of interest (ROI) analysis was performed on both the left and right kidneys in order to calculate mean medullary fractional anisotropy (FA) values for each subject. Importantly, medullary FA was shown previously to be sensitive to early-stage CKD (Donnola SB, et al. NMR Biomed. 2018; 31(3) ).

Results: Mean medullary FA values were significantly reduced in CF patients in comparison to healthy non-CF volunteers (0.26±0.03 vs. 0.32±0.05, p=0.05) despite normal kidney function (eGFR>100 ml/ min/1.73m 2 ). These observed differences are indicative of medullary microstructural alterations in the radially-aligned tubules and/or vessels of the renal medulla.

Conclusions: This initial study confirms that DTI-MRI can sensitively assess early-stage CKD in patients with CF despite normal kidney function. These results are consistent with previous DTI-MRI findings in other adult and pediatric patient groups. In addition, these DTI-MRI techniques are available on virtually all modern MRI scanners and require no exogenous contrast agent. Therefore, DTI-MRI may provide a safe, sensitive, and practical tool to detect and monitor early-stage CKD in CF patients in order to assess the impact of CFRD as well as nephrotoxic therapies.

Scott, D.W.; Wu, B.; Stuhlmiller, T.J.; Sesma, J. Spyryx Biosciences, Durham, NC, USA Introduction: Decreased mucocillary clearance due to mucus dehydration is a hallmark of cystic fibrosis (CF). Airway hydration is regulated by the epithelial sodium channel (ENaC) and its natural regulator short palate lung and nasal clone 1 (SPLUNC1). There are reports that SPLUNC1 is decreased in the sputum of CF patients. This suggests that loss of SPLUNC1 could play a major role in mucus dehydration associated with CF. We have developed a novel ENaC regulatory peptide, SPX-101, that can replace SPLUNC1 function. This work investigates the abundance of SPLUNC1 in sputum from healthy (no diagnosed pulmonary disease) and CF donors, the stability of SPLUNC1 and SPX-101 in these sputa and against individual proteases. The functionality of SPX-101 after exposure to diseased sputum and individual proteases was also assessed.

Methods: SPLUNC1 abundance was determined by Western blot analysis. Stability of SPLUNC1 in sputum was determined by Western blot while stability of SPX-101 was assessed by HPLC. The function of SPX-101 after exposure to sputum samples was determined by analysis of ENaC surface density, airway surface liquid (ASL) height, and survival of βENaC-transgenic mice.

Results: SPLUNC1 was significantly reduced, and often absent, in sputum from CF patients as compared to the healthy donors. Recombinant SPLUNC1 was degraded when incubated with CF sputum. The protein was also degraded by multiple individual proteases found in CF sputum such as neutrophil elastase and cathepsin G. In contrast, SPX-101 was stable in CF sputum and when incubated with neutrophil elastase and cathepsin G. Finally, SPX-101 retained the ability to internalize ENaC, increase ASL height, and increase survival of βENaC-transgenic mice after exposure to CF sputum.

Conclusions: SPLUNC1 protein is significantly reduced in the sputum of CF patients as compared to healthy controls. SPX-101, a therapeutically-optimized peptide which mimics SPLUNC1's ENaC regulatory function, is stable in these sputum samples and retains pharmacological activity thereafter. These data support the continued clinical development of SPX-101 for the treatment of cystic fibrosis. Lung function, measured by the percent predicted forced expiratory volume in one second (ppFEV1), has been widely used to evaluate the effectiveness of CF treatments. The correlation between ppFEV1 and patient-reported outcomes is critical for incorporating measures of treatment benefit into the evaluation of new CF treatments. We analyzed the correspondence of ppFEV1 with 8 symptom-based questions from the Cystic Fibrosis Respiratory Symptom Diary-Chronic Respiratory Infection Symptom Score (CFRSD-CRISS). Understanding the relationship of these measures is necessary for planning intervention trials and interpreting outcome measures. We hypothesized that a strong negative correlation would exist between the ppFEV1 and symptom measures from the CFRSD-CRISS items of difficulty breathing, chest tightness, wheeze, cough, and amount of mucus coughed up.

Methods: Data were derived from Standardized Treatment of Pulmonary Exacerbations (STOP), an observational study of CF patients experiencing PEx. CF patients who had CFRSD-CRISS and ppFEV1 measurements on the day of the initial PEx (time 1), 7 days later (time 2), and at the end of intravenous antibiotic treatment (approximately 28 days later; time 3) were included. We examined age-stratified (<18 versus ≥18 years old) characteristics, including the ppFEV1, overall CFRSD-CRISS score, and individual CFRSD-CRISS items. We also calculated age-stratified Spearman correlation coefficients and 95% confidence intervals (95% CIs) between the change in ppFEV1 and the change in CFRSD-CRISS items from time periods 1 and 2 and 1 and 3.

Results: A total of 91 patients were analyzed. As expected, ppFEV1 improved in both age categories by the end of PEx treatment. Symptom scores from the CFRSD-CRISS also improved with treatment for PEx; however, the correlation between ppFEV1 and the specific CFRSD-CRISS measures was low. The correlation of change between ppFEV1 and overall CFRSD-CRISS was -0.42 (95% CI: -0.57, -0.22) from time 2 to 1 and -0.42 (95% CI: -0.58, -0.24) from time 3 to 1. None of the changes in specific CFRSD-CRISS items between time periods had stronger correlations with change in ppFEV1. The subanalysis of patients age ≥18 (N=74) showed similar patterns, with correlations of -0.48 (95% CI: -0.64, -0.28) from time period 2 to 1 and -0.42 (95% CI: -0.58, -0.21) for time period 3 to 1, again with less strong correlations for specific items.

Conclusions: As treatments have become more diverse and effective, it is important to ensure that we accurately measure improvement. The lack of correspondence between direct patient-reported symptoms and ppFEV1 suggests that both measures are needed to evaluate clinical response and that these measures likely capture different aspects of the disease. Further work is needed to determine other phenotypic clinical responses that could be used. Premature termination codons (PTCs) in CFTR result in truncated protein that is nonfunctional and/or unstable, and are the proximate cause of ~11% of CF-causing alleles. High-throughput screening campaign of 750,000 compounds was carried out to identify and develop compounds that induce PTC suppression of mutant CFTR. Since PTC-containing mRNAs are subject to degradation through the nonsense-mediated mRNA decay (NMD), agents that concomitantly inhibit NMD are also of interest for optimal promotion of full-length protein expression and function. Three lead hits, SRI-37240, SRI-38711, and SRI-40003 have demonstrated efficacy in inducing translational readthrough in fluorescence reporters (Du M, et al. Pediatr Pulmonol. 2018; 53(S2) :254[abstract]) and restoring CFTR expression and function as assessed via transepithelial chloride conductance assay (TECC) in Fischer rat thyroid (FRT) cells stably transduced with CFTR-G542X, -W1282X, and -R1162X cDNA. FRT cell monolayers were pretreated with different doses (1-60 µM) of each compound and G418 (250 µg/mL) as a positive control. Scaffolds represented by the original hits SRI-37240 and SRI-38711 restored CFTR activity up to ~30% of wild-type (WT) in different FRT PTC mutations, as previously reported. The new scaffold SRI-40003 reached ~5% of WT activity (0.91 ± 0.01 mS/cm 2 ) in FRT G542X cells, and ~10% in both FRT W1282X (2.13 ± 0.04 mS/cm 2 ) and FRT R1162X (2.16 ± 0.11 mS/cm 2 ), at maximum dose as compared to vehicle (P<0.05). VX-809 (3 µM) addition to SRI-40003 doubled (10-20% WT) the CFTR signal in each of these mutations, indicating an additive effect by restoring function to the resulting missense alleles. Enhanced G542X-CFTR protein expression levels in the presence of SRI-40003 were strongly correlated to functional readouts (R 2 =0.98). Several lead scaffold derivatives are currently being evaluated in FRT G542X cells and demonstrated strong structure-activity relationships, allowing optimization of potency and efficacy. Of particular interest are SRI-41315 (~10% WT, EC50=0.8 µM), SRI-41765 (~17% WT, EC50=1.4 µM) and SRI-41876 (EC50=3.4 µM) derivatives of SRI-37240, SRI-40003 and SRI-38711 respectively. To augment readthrough-mediated CFTR expression, double (SRI agents + CFTR modulators/NMD inhibitors) and triple combination (SRI agents + CFTR modulators + NMD inhibitors) strategies are being explored in primary gene-edited 16HBE14o-(G542X/G542X; W1282X/W1282X) cells with endogenous CFTR context produced by CFFT, immortalized hTERT G542X/G542X cells, and intestinal organoids from nonsense mutation patients. Initial results reflect successful detection of readthrough in primary cell systems, and additional optimization is underway. In conclusion, SRI agents, with and without combination therapy, augment CFTR expression and function in CFTR PTC mutations, and have the potential to accelerate progress in this challenging area of CFTR restoration. This work is funded by the CF Foundation and was assisted by reagents provided by CFFT lab, S. Randell (UNC-Chapel Hill) and R. Bridges (Rosalind-Franklin).

Liu, E.; Lauffer, D.; David, S.; Deshpande, A.; Sui, J. CrowdOut Therapeutics, LLC, Nashua, NH, USA Various CFTR mutations lead to deficient CFTR activities and consequently reduced chloride transport in epithelial cells. As an alternative chloride secretion mechanism, the Ca 2+ -activated Clchannel anoctamin1 (ANO1/TMEM16A) is an important chloride channel protein supplemental to CFTR in the airway epithelial cells. Small molecule TMEM16A modulators could potentially offset the deficit of chloride secretion in CF, producing CFTR-mutation independent benefits to all existing CF treatments and to other respiratory diseases where airway hydration is in demand.

At CrowdOut Therapeutics, we have established a cell-based assay platform with high-throughput screening (HTS) capacity, a collection of about one-million-small-molecule libraries, and a focused target protein TMEM16A. In Jan 2018, we started our HTS campaign to identify small-molecule modulators using epithelial cell lines that transiently express both YFP and TMEM16A. Known activators and inhibitors have shown dose-dependent effects in our assays. Our assays were designed to screen for both TMEM16A boosters, compounds that can boost the expression of TMEM16A in the cell membrane, and TMEM16A activators, compounds that can instantly activate TMEM16A. Cells were either incubated with testing compounds for 24 hours or 15 minutes before being subjected to a YFP Iion flux assay in 384-well format. With a pace of over 30,000 compounds per week, we have collected hundreds of primary screen hits, both TMEM16A boosters and TMEM16A activators. Many of the primary hits were confirmed by our full-dose response validation screen. Confirmed hits will be further validated in patch-clamp experiments and in primary human bronchial epithelial cell assays. Introduction: The glycoproteins MUC5AC and MUC5B are the predominant gel-forming mucins which, with DNA (derived from host cells and colonising microorganisms) and lipids, are thought to be the primary contributors to the rheological properties of the thixotropic matrix which is cystic fibrosis sputum. Physiotherapy and the development of mucolytic and osmotic therapies attempt to tackle this problem in different ways. Cysteamine is a simple aminothiol for which we have already demonstrated antibiotic potentiating properties (Fraser-Pitt D, et al. Antimicrob Agents Chemother. 2016; 60:6200-6; Fraser-Pitt D, et al. Infect Immunol. 2018; 86(6) ) but which also has sputum-thinning and biofilm-disrupting properties (Charrier C, et al. Orphanet J Rare Dis. 2014; 9:189; Devereux G, et al. EBioMedicine. 2015; 2:1507-12) . Here we demonstrate that cysteamine targets cysteines of the von-Willebrand regions in gel-forming mucins in a distinct way to existing thiol mucolytics.

Methods: Rheological analysis of ex vivo CF sputum was performed on a Kinexus Ultra Malvern Panalytical instrument. The ability of cysteamine to bind cysteine was confirmed using LC MS/MS analysis of synthetic peptide. Acidified Alcian Blue was used to visualise mucin in diluted ex vivo sputum samples.

Cysteamine significantly reduces yield stress in freshly expectorated CF sputum and is effective at doses achievable in the lung via oral dosing. The effect on yield stress was greater than the same molar concentration of N-acetylcysteine and physiologically relevant concentrations of DNase I, though DNase I did dose-dependently reduce shear viscosity. Macroscopic and microscopic visualisation of saliva, pig mucin suspensions and diluted ex vivo sputum samples of CF sputum treated with cysteamine show a dose-dependent reduction in polymeric mucin. LC MS/MS analysis of synthetic peptides confirms reversible cysteaminylation activity.

Discussion: Cysteamine can reversibly bind susceptible cysteine thiol side chains and form lysine-like adducts. This can improve the hydrophilicity of mucin, and cysteamine mixed disulphide formation at susceptible von Willebrand cysteines disrupts disulfide bridges between polymers and causes electrostatic repulsion which likely contributes to increased potency over N-acetylcysteine-mediated mucolytic action and was superior to DNase I in reducing yield stress. Cystic fibrosis (CF) is a genetic disorder that results in an abnormally thick mucus production in the respiratory and GI tract. Pseudomonas aeruginosa (PsA) is an opportunistic pathogen in the congested CF airway and results in chronic infection that is difficult to eradicate with conventional antibiotics due to the formation of poorly penetrated biofilms.

We have previously demonstrated that a novel polycationic polymer, poly-acetyl-arginyl-glucosamine (PAAG), improves the viscoelastic properties of CF mucus and abrogates biofilm formation of mucoid bacteria.

In this study, we tested the effect of PAAG using conventional agar bead infection protocol in wild-type (WT) Sprague-Dawley rats. WT rats were nebulized with either PAAG (250 mg/mL) or glycerol control (1.38%) for 45 minutes, 10 days (5 doses) before PsA infection. WT rats were then inoculated with 10 5 CFUs of PsA and continued treatment with PAAG or glycerol for 2 days (2 doses). Following treatment, the lungs were harvested and analyzed for CFU counts and histopathology. Rats treated with PAAG had significantly lower incidence of PsA-infection compared to glycerol control (58% of PAAG-treated vs. 100% of control, P < 0.05). Furthermore, of the infected animals, the PAAG-treated cohort demonstrated fewer CFUs recovered compared to vehicle control (~0.08 vs. 1.2 x 10 6 CFU/mL, P < 0.05).

While agar-coated bead models are the standard for eliciting a semichronic infection in vivo, published studies suggest that PsA delivered in alginate-coated beads better mimics chronic respiratory infection through a more realistic inflammatory response by inhibiting efferocytosis and phagocytosis by macrophages as well as increased production of cytokines; this then fosters chronic infection and biofilms (Hoffmann N, et al. Infect Immunity. 2005; 73:2504-14) . We are now implementing the alginate infections in comparison to agar bead infections using in vitro and in vivo models of CF, including CFBE41o-monolayers and Scnn1b-Tg mice that overexpress ENaC. Thus far we have developed a method to extract pseudomonad alginate and are comparing this to agar beads, which primarily induce chronic infection by hypoxia induction and slow release; nonbacterial alginate beads will serve as an additional control. Outcomes will focus primarily on biofilm production, cytokine analysis and histopathology. These studies will investigate the role of alginate in the persistence of chronic PsA infection and determine whether alginate may play a role in creating a superior model of chronic CF respiratory infection and if novel mucolytics and biofilm disrupters, such as PAAG, are effective in mitigating them.

Introduction: Cysteamine has been used to treat cystinosis for >20 years. Preclinical and Phase IIa studies have shown potentially therapeutically beneficial properties including, antibiotic potentiation, mucus-thinning and biofilm-disruption. We are investigating oral cysteamine (Lynovex®) as an adjunct to conventional treatment of acute exacerbations in cystic fibrosis. The Cysteamine in Acute Respiratory Exacerbations of Cystic Fibrosis (CARE-CF-1) trial was designed to identify the most effective dose regimen and the most suitable patient-reported outcome tool. Additional outcomes were also explored.

Methods: Multicentre, double-blind, placebo-controlled randomised 6-arm study. Adults with CF experiencing acute exacerbations were randomised to Placebo, 450 mg QD, 150 mg TID, 450 mg BID, 300 mg TID and 450 mg TID treatment groups over 14 days. Exacerbations were defined as a Fuch's score 4. All outcomes were at day 14. Primary outcome was change from baseline in sputum gram-negative bacterial load. Secondary outcomes: 1) changes in patient-reported outcome measures from baseline assessment using the CF respiratory symptom diary, chronic respiratory infection symptom score (CFRSD-CRSS), CFQ-R, and Jarad and Sequeiros Symptom Score Questionnaire; 2) blood leukocyte count; 3) change from baseline in sputum neutrophil elastase and IL-8 levels; 4) change in FEV1, weight and BMI; 5) adverse/serious adverse events.

Results: In total 89 participants were recruited, equally distributed among the 6 treatment groups. Mean (SD) age 29.8 years (9.6), 48% female, mean (SD) FEV 1 43.3% (18.3) predicted. 10 participants withdrew from the study, 6 because of AEs/SAEs.

The reduction seen in gram-negative and total burden at 450 mg BID cysteamine at day 14 was not significant. Cysteamine 450 mg BID significantly improved the CFRSD-CRSS score, mean (SD) improvement relative to placebo 10.6 (4.94), p=0.035. 450 mg QD 9.0 (5.24), p=0.090. Analysis of individual domains suggested significant impact on feeling feverish and chest tightness. 450 mg BID significantly reduced white cell count, mean (SD) relative to placebo is 2.43 x 10 9 /l (1.12) p=0.033. Cysteamine at the other dose regimens tested had no significant effect.

The results indicate that cysteamine 450 mg twice daily is a promising adjunct to conventional treatment of CF exacerbations and the most appropriate PROM is the CFRSD-CRSS. These elements will be central to pivotal trials now planned. Center, McGill, Montreal, QC, Canada; 4. Human Genetics, McGill University, Montreal, QC, Canada; 5. Medicine, McGill University, Montreal, QC, Canada; 6. Laurent Pharmaceuticals, Montreal, QC, Canada The cystic fibrosis transmembrane conductance regulator (CFTR) is a tightly regulated anion channel that mediates cAMP-stimulated secretion and is mutated in cystic fibrosis (CF). Severe inflammation and chronic bacterial infection are hallmarks of CF lung disease and cause much of the morbidity and mortality in CF, however the relationship between CFTR function and inflammation remains poorly understood. Several lipid abnormalities have been observed in CF. Very long-chained ceramides (VLCC; C24:0, C26:0), which are considered anti-inflammatory, are reduced, whereas pro-inflammatory long-chained ceramides (LCC; C16:0, C18:0) are increased in CF and may contribute to progression of the disease. Fenretinide (FEN), an exploratory semi-synthetic retinoid, has been shown to reduce inflammation and airway hyper-responsiveness in a CF lung disease model. It also enhances clearance of Pseudomonas aeruginosa from the lungs of CF mice and re-balances levels of LCC, VLCC and other fatty acids in CF patients and mice. Although fenretinide (LAU-7b oral capsules) has been tested as a first-in-class lipid modulator of inflammation in a Phase 1b clinical trial in adult CF patients and is currently cleared for Phase 2 by both FDA and Health Canada, its effects on CFTR membrane distribution and function have not been investigated.

In this study, we used quantitative fluorescence imaging analyses to study FEN treatment effects on the plasma membrane distribution of wild-type CFTR (wt-CFTR) and the most common mutant, F508del-CFTR under control and stressor conditions. We also examined the impact of FEN treatment on CFTR channel function in Ussing chambers. All studies employed a low concentration similar to that measured in the plasma of FEN-treated CF patients. Although FEN did not alter the membrane distribution of wt-CFTR on primary airway epithelial cells under control conditions, it did enhance the partitioning of CFTR into ceramide-rich microdomains during cell stress, suggesting that CFTR recruitment inside these domains may be part of the host defense response that leads to increased pathogen clearance and/or the resolution of inflammation. Short circuit currents (I sc ) mediated by both wt-and F508del-CFTR were increased significantly (30-120%) in a concentration-and time-dependent manner by treatment with FEN. Since we have shown previously that the distribution of CFTR on primary airway epithelial cells depends on membrane lipid composition, the FEN-induced enhancement of functional CFTR expression, at least in part, may be mediated by the rebalancing of fatty acids and ceramide levels in the airway epithelial cells. These results highlight the importance of membrane lipids as determinants of CFTR membrane distribution and function and provide insights into the potential role of CFTR in regulating inflammation resolution and bacterial clearance. Supported by the Cystic Fibrosis Foundation. Background: There are currently several novel CFTR modulators under study, particularly in early-phase clinical trials. Such trials typically are of short duration; for example, two Phase 2 multi-center clinical studies of a triple-combination (TC) CFTR modulator regimen had only one-month treatment periods. Although participation in clinical trials has long been a focus for many CF centers, little has been done to compare health outcomes before participation in brief trials with acute clinical outcomes afterwards.

Objectives: The aim of this retrospective chart review was to characterize respiratory health, antibiotic exposure, and hospitalization before and after brief CFTR modulator study enrollment in adult CF patients in two TC CFTR clinical trials conducted at the Boston Children's Hospital and Brigham & Women's Hospital Cystic Fibrosis Center from 2016 -2018. We evaluated patients' acute clinical outcomes, irrespective of randomization to active drug or placebo, as we remain blinded to each patient's treatment.

Methods: We evaluated FEV 1 percentage change, number of pulmonary exacerbations, use of oral or intravenous antibiotics, and number of hospitalizations during the six-month time period before/after trial participation in a non-random sample enrolled via physician referral. All patients had one copy of the F508del mutation with a minimal function mutation on a second allele. Demographics collected included gender, age, and concurrent microbiology.

Results: Among the four male and six female patients, median age 27.1, there was no significant difference between mean FEV 1 percentage before and after trial participation. Despite no overall group change, seven (70%) patients experienced a decline in FEV 1 up to six months afterwards, including four of the five patients with more severe lung disease (FEV 1 below 60% at study start). FEV 1 declines between one and 24 percentage points (median 11 points) were observed up to six months after participation. Among the patients with more severe lung disease, three received intravenous treatment within a four-month window for CF pulmonary exacerbations (with two hospitalized) and two received oral antibiotics. Notably, the patient experiencing the greatest increase in FEV 1 across trials experienced the greatest FEV 1 decline afterwards but was neither treated with antibiotics nor hospitalized. This patient had received oral antibiotics twice in the six-month period before trial participation.

Conclusions: Brief CFTR clinical trial modulator use may trigger detrimental acute outcomes at trial end in adults with CF, including decline in lung function, initiation of oral or intravenous antibiotics for pulmonary exacerbation, and hospitalization. CF providers might consider prophylactic escalation in therapeutic management of their adult CF patients after brief CFTR clinical trial treatment or other clinical interventions potentially exposing patients to clinically significant fluctuations in CFTR function. This may avoid acute respiratory changes possibly affecting longer term well being. The natural history of CF remains one of progressive lung function decline and premature death. To date it remains an elusive goal to predict future changes in the health of people with CF. Previous attempts at survival analysis have used static pre-defined risk factors to develop models to predict outcomes. Here we introduce a different approach that uses advanced machine learning techniques to analyse longitudinal measures to develop a more useful dynamic survival analysis tool.

Methods: Using UK CF registry we analysed data from 5883 adults with CF from 2009 to 2015. Some 87 variables associated with each patient across all years included demographics (e.g. age, height, weight, BMI), CFTR genotype, treatments, lung function (FEV1), comorbidities, bacterial and fungal infections. The machine learning methods used were a Recurrent Neural Network (RNN) structure using a shared subnetwork. We assessed two competing risks: death from respiratory failure, and death from all other causes. Our method learns complex relationships between trajectories and survival probabilities directly. It is dynamic as it updates its predictions when new data become available. Finally, we compared results to other prediction methods such as Random Survival Forests (RSF), Joint Model (JM), and Cox Proportional Hazards (cs-Cox).

Our method achieved improvements in discriminative performance over other methods in terms of the time-dependent concordance index, which is extended to longitudinal survival. Our method achieved improvements of 5.2% and 14.2% over the best benchmark (8.0% and 15.4% over JM) on average in terms of discriminative performance for death from respiratory failure and death from other causes, respectively. In addition, while much of the current literature has focused on spirometry as the main CF risk factor, our work confirmed the importance of a history of intravenous (IV) and non-IV antibiotic treatments and nutritional status in the risk assessment of people with CF (especially for nonrespiratory-failure deaths).

Conclusions: Our results suggest that the application of our dynamic deep learning approach to survival prediction for people with CF may deliver significant opportunities for individualised and meaningful conversations about health and prognosis between patients and their clinicians.

Comparison of results of c-index between Deep Learning Prediction and other methods for survival prediction at age 30 and 50. Background: Combination therapy with multiple CFTR modulator agents is used to target the underlying defect in the CFTR protein for a variety of mutations. To expand these treatments to individuals with rare mutations, a patient-derived biomarker with low variability and high consistency in outcome measure is needed for clinical use. We report a three-dimensional (3D) organoid model derived from nasal epithelial cells (HNE) for potential use as a biomarker.

Methods: HNE cells were obtained from non-CF and CF patients by nasal brush biopsy. HNE were co-cultured with irradiated 3T3 fibroblasts, followed by culture in matrigel for 10-28 days. After fixation with 4% paraformaldehyde, paraffin-embedded cross-sections were stained using hematoxylin and eosin (H&E); Alcian blue/periodic acid Schiff (ABPAS); and primary antibodies to acetylated tubulin (cilia) and MUC5AC (mucin). CFTR function was tested by swelling assay using a cocktail containing 10 µM forskolin + 100 µM IBMX to stimulate CFTR and 100 µM amiloride to inhibit endogenous ENaC, +/-CFTR modulators. Phase-contrast microscopy was used to measure change in organoid size.

Results: We have optimized methods to reduce variation in culture success and generate highly differentiated organoid morphology from fresh and frozen HNE, non-CF and CF individuals age 1 to 50 years, with a variety of CFTR mutations. H&E staining clearly shows the lumen as well as cilia on the interior in the cross sections. Cilia were also seen to line the lumen in cross-section using anti-acetylated tubulin antibodies. The mucus in the lumen was detected by AB-PAS staining and with immunofluorescent staining using anti-MUC5AC. Growth varied depending on culture conditions; for example, at day 10 spherical organoids cultured in 20% matrigel had an average diameter of 85.5 µm, compared to 63.8 µm in 50% matrigel (p<0.009). Non-CF organoids grown for three weeks swelled with forskolin stimulation. Cross-sectional area of the lumens increased by 4.5 fold in comparison to baseline after 6 hours (p=0.014). CF organoids with G551D/unknown did not change substantially with forskolin stimulation (increase in total size of 1.13 fold) even after 20 hours of stimulation.

Rationale: Cystic fibrosis (CF) lung disease progressively worsens from infancy to adulthood. Changes in early CF airway fluid composition may reveal therapeutic targets to curb early disease progression.

Objective: To identify molecular mediators of lung disease in CF children among airway fluid metabolites.

Methods: Infants at Erasmus MC were diagnosed with CF by newborn screening and enrolled in the I-BALL study to collect bronchoalveolar lavage fluid (BALF) and chest computed tomography (CT) scans at one, three and five years of age. We analyzed 24 patients aged 12-38 months, three of whom were later denoted as CF screen positive, inconclusive diagnosis (CFSPID). CT scans were scored according to PRAGMA-CF, quantifying a composite of total lung damage (PRAGMA-%Dis) and specific subsets (e.g., bronchiectasis). Small molecules in BALF were measured with high-resolution, accurate-mass metabolomics. Myeloperoxidase (MPO) was quantified both by ELISA and activity assays.

Results: Increased PRAGMA-%Dis correlated with BAL neutrophil count. PRAGMA-%Dis correlated with 104 metabolomic features (p<0.05, multiple comparisons-adjusted q<0.25). The most significant annotated feature was methionine sulfoxide, a product of methionine oxidation by MPO-derived oxidants. We confirmed the identity of methionine sulfoxide in BALF by MS/MS and co-elution with an isotopically enriched standard. We used reference calibration to provide nM concentrations of methionine sulfoxide in BALF and confirmed the correlation with PRAGMA-%Dis (ρ=0.582, p=0.0029), extending to bronchiectasis (ρ=0.698, p=1.5x10 -4 ), BALF MPO (ρ=0.803, p=3.9x10 -6 ) and percentage of BAL neutrophils (ρ=0.569, p=0.0046).

Conclusions: BALF methionine sulfoxide correlates with structural lung damage, neutrophil count, and MPO in very young children with CF. Methionine sulfoxide overabundance in early CF is likely connected to bronchiectasis and other lung damage via neutrophilic inflammation and subsequent production of halogenating oxidants by secreted MPO. Further studies are needed to establish whether methionine oxidation directly contributes to early CF lung disease and explore potential therapeutic targets revealed by these findings. London, United Kingdom; 5. University Hospital Southampton, Southampton, United Kingdom Introduction: We currently do not fully understand, and therefore cannot predict, the clinical trajectory of a given individual with CF. The ability to synchronously forecast multiple clinical variables (for example how quickly lung function might decline or the likelihood of developing CF-related diabetes or liver disease) would deliver enormous benefits to individuals with CF, their carers, and CF center teams, providing opportunities to focus screening or interventions, and augment or refine treatments. We therefore chose to explore whether deep learning methods could provide a forecasting framework for individuals with CF.

While deep learning applications to medical problems have gained popularity in recent times, all the existing methods only use covariates at a single time point in making predictions. They thus issue a static survival/ prognostic model. We reasoned however that a patient's medical history (their previous clinical trajectory) would likely influence future clinical outcomes. Predictions could thus be improved by incorporating past information. We therefore chose to integrate historical clinical information into our machine learning architecture using a Recurrent Neural Network (RNN) in the base layer, which dynamically updates its memory state over time.

Methods: We retrospectively analysed longitudinal clinical metadata from the UK Cystic Fibrosis (CF) registry obtained from annual clinical reviews of 10,980 individuals with CF from 2008-2015. A total of 87 variables associated with each patient across all years included demographic information (e.g. age, height, weight, BMI), CFTR genotype, treatments received, lung function (FEV1), CF-related comorbidities, and all bacterial and fungal infections. For this study we considered a joint model for the 2 continuous lung function scores (FEV1 and percent predicted FEV1), 20 binary longitudinal variables of CF-related comorbidities and different infections, and death.

Results: We successfully generated dynamic forecasting of various CF clinical parameters utilising prior longitudinal clinical metadata. We were thus able to simultaneously predict survival probabilities, risks of developing CF-related complications, and the trajectory of lung function decline for each individual with CF.

Conclusions: Our results suggest that the application of our scalable deep learning approach to CF will allow individualised risk stratification for people with CF and provide a framework for delivering personalised care.

Methods: Variants were stably expressed from a single cDNA introduced into the genome of CF bronchial epithelial (CFBE) cells (64 variants), Fischer rat thyroid (FRT) cells (14 variants), or both (3 variants) . CFTR function assessed by short-circuit current (I sc ) in CFBE cells was normalized to wild-type (WT) using CFTR mRNA levels. FRT cell lines with CFTR mRNA levels similar to a cell line expressing WT-CFTR were selected. The I sc responses to ivacaftor (IVA), lumacaftor (LUM), and ivacaftor-lumacaftor combination (IVA-LUM) of 58 variants were quantified and correlated to residual CFTR function of these variants. High responders were identified by response >2SD above the mean of all variants; mid-responders by response >2SD above the mean of remaining variants after high responder removal; average responders were the remainder.

Results: The 67 variants tested in CFBE cells were associated with a range of disease severity. Functional testing was essential to distribute variants according to disease status: 35 were confirmed as CF-causing; 18 were associated with varying clinical consequences; 6 were non-CFcausing; and 8 were not assigned. The level of residual CFTR function was inversely correlated with sweat chloride (R 2 =0.7; p<0.001), an indicator of clinical disease severity. In 58 variants tested with modulators (61 cell lines, including 3 variants in CFBE and FRT cells), the magnitude of drug response correlated with residual CFTR function (IVA: R 2 =0.9; LUM: R 2 =0.6; IVA-LUM: R 2 =0.9; p<0.001 for all). Variants assorted into statistically distinct theratypes based on average (IVA: n=53; LUM: n=52; IVA-LUM: n=52), mid (IVA: n=4; LUM: n=4; IVA-LUM: n=4), or high (IVA: n=4; LUM: n=5; IVA-LUM: n=5) response. Of 44 variants analyzed in CFBE cells, 41 had a significantly higher response to IVA-LUM than either compound alone.

Conclusions: Response to modulators is correlated with amount of residual CFTR function, and nearly all variants received the greatest benefit from IVA-LUM. Since missense variants tested encompass a spectrum of residual CFTR function (<1% to 150%) and a large fraction of those with CF who carry missense variants, we predict that most individuals with missense variants will experience some benefit from combination therapy. Theratyping using methods employed here could be used to identify individuals who may have better or even dramatic clinical responses to CFTR modulators. (PEx) . Critical components of PEx management include optimal antibiotic therapy and maximizing airway secretion clearance. Due to enhanced antibiotic clearance achieving optimal serum concentrations may be difficult. Further, scant human data exist assessing the clinical impact of individualizing beta-lactam therapy and optimizing drug exposure. The purpose of this analysis is to measure the clinical impact of beta-lactam therapeutic drug monitoring (TDM) in patients with PEx.

Methods: Retrospective, single-center, pre-/post-intervention study evaluating the clinical effect of beta-lactam TDM performed on consecutive adult CF patients with PEx admitted to UF Health Shands Hospital from September 2016 to June 2017. The primary outcome was to determine the effect of TDM on time until next exacerbation (TUNE). TUNE was reviewed one-year prior to and one-year after TDM implementation. Secondary outcomes were PEx frequency and rate of decline in forced expiratory volume in one second (FEV1) percent predicted during this 2-year window. During the intervention period, PEx management included betalactam TDM for cefepime (C), piperacillin-tazobactam (PT), or meropenem (M). Peak and trough (T) concentrations were obtained 1 hour after completion of infusion and 30 minutes before the next dose, respectively. Dosing was adjusted to achieve a desired T concentration of at least 4X greater than the minimum inhibitory concentration (Cmin:MIC ≥ 4). Susceptibility testing was performed and interpreted according to CLSI recommendations for Kirby Bauer (KB) disk diffusion. Due to using KB testing, CLSI breakpoints for Pseudomonas were used as the target MIC.

Results: TDM was performed in 32 consecutive adults patients admitted for PEx management. The group was predominantly female (60%), median age of 27 years, weight of 55.2 kg, CrCl of 106.9 mL/min, and single CF mutations (F508del) were observed in 53% (17/32). FEV1 percent predicted on admission was 41.5. P. aeruginosa and S. aureus were the most common isolates.

Implementation of beta-lactam TDM resulted in a shift in TUNE from 203.8 to 233.9 days (p=0.26) observed in the pre-and post-implementation phases, respectively. Patients with a history of ≥ 2 PEx per year experienced the largest shift with an increase in TUNE from 115.1 to 187.6 days (p=0.048). With TDM, the number of PEx per year trended down from 1.78 to 1.34 (p=0.1); in conjunction, the annual rate of decline in FEV1 % predicted shifted from -7.4% to -6.2% (p=0.39). Dosing adjustments were required for all patients resulting in the use of extended infusion (28%) and continuous infusion (72%) strategies to achieve goal concentrations.

Conclusion: Integration of beta-lactam TDM in CF exacerbation management resulted in improvements in TUNE, especially in patients with frequent exacerbations. In conjunction, optimizing drug exposure contributed to fewer exacerbations and slower decline in FEV1 percent predicted. These data highlight the potential impact of beta-lactam TDM and warrant further analysis in larger studies. We are studying a novel antibiotic approach that uses the metal gallium (Ga) to disrupt bacterial iron metabolism. Gallium has a nearly identical ionic radius as iron, and many biologic systems are unable to distinguish gallium from iron. Gallium disrupts iron dependent processes because Ga 3+ cannot be reduced and redox cycling is critical for iron's biological function. Pre-clinical studies show that gallium kills P. aeruginosa (Pa), is active against Pa biofilms in vitro, and treats three different animal models of chronic Pa infections. We tested intravenous (IV) gallium in people with CF and chronic Pa lung infections in a phase 1b safety and pharmacokinetic (PK) study. Gallium treatment appeared safe, had favorable PK, and produced improvements in lung function as measured by FEV 1 and FVC.

Methods: We performed a multicenter placebo-controlled randomized clinical trial at 23 US CF centers. Adults with CF (≥ 18 years), documented chronic Pseudomonas aeruginosa (Pa) infection and FEV 1 ≥ 25% predicted were randomized to receive a five-day continuous intravenous infusion of 200 mg/m 2 /day of IV gallium nitrate or volume-matched placebo (0.9% sodium chloride) and followed through Day 56 for adverse events, inflammatory markers, quantitative sputum Pa density and spirometry. The primary outcome measure was the difference between treatment groups in the proportion of subjects with a 5% or greater relative change in FEV 1 (liters) from baseline to Day 28. Key secondary outcome measures included safety, change in sputum P. aeruginosa density, the emergence of other CF pathogens, and respiratory symptoms using the Cystic Fibrosis Respiratory Symptom Diary -Chronic Respiratory Infection Symptom Score (CFRSD-CRISS).

Results: Enrollment is complete. Of the 128 subjects screened, 120 subjects were randomized and 117 have completed the study as of 4/1/2018. The average age of those randomized was 33 years (SD=10), and 46% of the patients were female. The mean FEV 1 percent predicted was 55% (SD=19). Seven (6%) subjects discontinued or had study drug infusion interrupted or stopped. As of November 2017, (the time of the final DSMB review), there were 21 SAE's noted, of which 16 were pulmonary exacerbations thought unrelated to study drug. No suspected unexpected serious adverse reactions (SUSAR) were noted. The database for the trial will be locked in May 2018.

Conclusions: Gallium nitrate is a novel therapeutic agent to treat chronic Pa infection in CF. The top line results of the trial will be available and presented at NACFC in October 2018.

Acknowledgments: Funding Sources: FDA R01 FD003704; CFFT; NIH/NHLBI UM1HL119073, NIH/NCRR UL1 RR025014. 5 1. Vertex Pharmaceuticals Inc, Boston, MA, USA; 2. Optum, Johnston, RI, USA; 3. National Jewish Health, Denver, CO, USA; 4. University Hospital la Fe Valencia, Valencia, Spain; 5. Imperial College and Royal Brompton Hospital, London, United Kingdom Objective: To examine the impact of tezacaftor/ivacaftor (TEZ/IVA) on disease-related symptoms, functioning, and well-being as measured by Cystic Fibrosis Questionnaire-Revised (CFQ-R) in patients (pts) with cystic fibrosis (CF) homozygous for F508del-CFTR (F508del/F508del). Improvement in the respiratory domain of CFQ-R has been reported previously. Here, we report on the other health domains of CFQ-R.

Methods: EVOLVE (NCT02347657), a phase 3, randomized, doubleblind, placebo-controlled trial, evaluated TEZ/IVA (100 mg QD/150 mg BID) in pts aged ≥12 years with CF and F508del/F508del. CFQ-R consists of 12 domains, including respiratory symptoms (a key secondary endpoint in EVOLVE), and was assessed at baseline (BL) and weeks 4, 8, 12, 16, and 24 . A mixed-effects model for repeated measures (MMRM) was used in the prespecified analysis of the absolute change in score from BL through week 24. In a post hoc analysis, cumulative distribution functions (CDF) were used to compare the distribution of change in scores from BL to week 24 across the 2 treatment groups. No multiplicity adjustment was used in the analyses. CDF differences were assessed using nominal P values obtained from the Anderson-Darling test.

Results: Data from 504 pts with a BL CFQ-R assessment were included in the prespecified analyses. A subset (N=481) with scores at BL and week 24 were included in the post hoc analyses. In the prespecified analysis, an improvement was observed favoring TEZ/IVA over placebo on the following CFQ-R domains: physical functioning, treatment burden, health perceptions, vitality, and social functioning (Table) . In the post hoc CDF analyses, differences favoring TEZ/IVA were observed in a subset (5/12) of domains, including respiratory symptoms, as well as physical functioning, health perceptions, emotional functioning, and treatment burden (all P<0.05). The remaining domains (7/12) 

Conclusion: The present analyses demonstrate TEZ/IVA treatment benefit across a broad range of patient-reported health outcomes beyond respiratory symptoms, including physical functioning. These findings further support the value of TEZ/IVA treatment in CF pts with F508del/ F508del. Acknowledgment: Sponsored by Vertex Pharmaceuticals Incorporated. To examine the impact of tezacaftor/ivacaftor (TEZ/IVA) vs placebo on disease-related symptoms, functioning, and well-being as measured by Cystic Fibrosis Questionnaire-Revised (CFQ-R) in patients (pts) heterozygous for F508del and a second allele with a CFTR mutation predicted to have residual function (F508del/RF). Improvement in the respiratory domain of CFQ-R has been reported previously. Here, we report on the other health domains of CFQ-R.

Methods: TEZ/IVA (100 mg QD/150 mg BID) was evaluated vs placebo in pts aged ≥12 years with CF and F508del/RF in EXPAND (NCT02392234), a phase 3, randomized, double-blind, placebo-controlled, crossover design trial. CFQ-R consists of 12 domains, including respiratory symptoms (a key secondary endpoint in EXPAND), and was assessed at baseline (BL), and weeks 4 and 8. In the prespecified analysis, a linear mixed-effects model was used to estimate the mean absolute change in each domain score from BL to the average of weeks 4 and 8 scores.

In the post hoc analysis, cumulative distribution functions (CDF) were used to compare the distribution of change in scores from BL to week 8 between TEZ/IVA vs placebo. No multiplicity adjustment was used in the analyses. CDF differences were assessed using the Anderson-Darling test (nominal P value).

Results: Data from 244 pts with a BL CFQ-R assessment were included in the prespecified analysis; pts with BL and week 8 scores (N=240) were included in the post hoc analysis. In the prespecified analysis, a TEZ/IVA treatment effect, vs placebo, was observed in health perceptions, vitality, physical functioning, role functioning, social functioning, weight, treatment burden, and emotional functioning (Table) . Post hoc CDF analysis demonstrated consistent findings in a subset of domains (8/12), including respiratory symptoms, as well as in health perceptions, vitality, physical functioning, role functioning, social functioning, treatment burden, and body image (all P<0.05). The remaining domains (4/12) demonstrated no difference.

Conclusions: Treatment with TEZ/IVA demonstrated improvement in patient-reported health outcomes beyond respiratory symptoms in CF pts with F508del/RF. These findings highlight the treatment impact on pts' functioning and further support the overall value of TEZ/IVA treatment benefit.

Acknowledgment: Sponsored by Vertex Pharmaceuticals Incorporated.

McElvaney, O.J.; Gunaratnam, C.; Reeves, E.P.; McElvaney, G.

Introduction:

The key protease implicated in CF airway disease is neutrophil elastase (NE), which is inhibited in vivo by α1-antitrypsin (AAT). Bronchoalveolar lavage fluid (BAL) is the current gold standard method for the assessment of NE activity, and therefore the assessment of synthetic NE inhibitors. However, this method of sampling is invasive and costly. Sputum presents a less-invasive method of sampling the CF airway, but inconsistencies in sputum data between centers have proven problematic.

Aim: This study assessed the "TEmperature-controlled Two-step Rapid Isolation of Sputum" (TETRIS) technique, a standardized protocol developed by this group for the processing of spontaneously expectorated sputum (SS) and induced sputum (IS) for sputum/BAL correlation, and compared TETRIS samples to those processed by conventional methods.

Objectives: To determine 1) whether NE activity measurements in IS from people with CF (PWCF) were comparable to BAL and/or SS samples taken from the same PWCF on the same day if processed appropriately; 2) the stability of NE activity in BAL, SS and IS when frozen over time and whether or not the ability of anti-NE therapies such as AAT to inhibit NE activity ex vivo is reproducible in fresh and stored samples; 3) to measure IL-1β in BAL, SS and IS and assess its stability with or without the addition of protease inhibitors.

Methods: PWCF (n=50) were recruited following ethical approval from Beaumont Hospital ethics committee. NE activity and its inhibition by AAT were measured by FRET. IL-1β was measured by ELISA. Results were correlated with commonly used clinical outcome measures such as FEV1 and the CF-ABLE score (McCarthy C, et al. Chest. 2013; 143:1358-64) , a validated prognostic tool.

Data: BAL NE activity levels were higher in CF than healthy controls (P<0.0001). NE activity in CF BAL correlated with FEV1 (R 2 =0.86, P<0.0001). NE activity in CF BAL also correlated with CF-ABLE score (R 2 =0.89, P<0.0001). NE activity in SS and IS produced by traditional processing both failed to correlate with BAL, while NE activity in TETRIS-processed SS and IS supernatants correlated strongly with BAL (SS: R 2 =0.81, IS: R 2 =0.9, both P<0.0001). The decrease in NE activity per SS sample was significantly lower in the TETRIS group (P<0.0001). Similarly, IS samples processed by traditional methods when compared to the same samples processed by TETRIS displayed a significantly greater loss of NE activity (P<0.0001). When TETRIS processing was applied a correlation with BAL was seen in both IS and SS (SS: R 2 =0.68, P=0.0002; IS: R 2 =0.91, P<0.0001). Sputum processed by traditional methods failed to show a correlation. NE activity (R 2 =0.82, P<0.0001) and IL-1β (R 2 =0.92, P<0.0001) correlated with FEV1. NE activity and IL-1β also correlated with CF-ABLE score (R 2 =0.86, R 2 =0.88, both P<0.0001). The change in IC50 for AAT against NE activity in each sample was greater in samples processed by traditional methods over 7 days and 30 days for both SS (Day 7: P=0.0002, Day 30: P<0.0001) and IS (Day 7: P=0.0003, Day 30: P<0.0001) compared to samples processed by TETRIS.

Conclusions: TETRIS processing is a reliable method of assessing airway inflammation in CF. These data also support the use of IS for disease monitoring and clinical trials.

Yanda, M.; Cebotaru, C.; Guggino, W.; Cebotaru, L. Johns Hopkins University, Baltimore, MD, USA The major hurdle with gene therapy is the development of neutralizing antibodies in response to repeat delivery that could potentially block expression of enough CFTR to be therapeutic. Given that the turnover of airway epithelial cells may make gene transfer with recombinant AAV-based vectors transient, repeat dosing of AAV1 pseudotyped virus will ultimately be required. The goal is to assess whether repeat dosing of AAV1-CFTR vectors administered to primates leads to widespread gene transfer and CFTR expression. To test this, we sprayed into the airways of 2 healthy male and 2 female Rhesus monkeys 5 years of age, 2 doses of 10 13 vg of AAV2-1-Δ27-264-CFTR at 0 and 30 days, respectively, followed by a single dose of 10 13 particles of AAV2-1-GFP at day 60. A similar construct Δ264-CFTR was shown to increase endogenous CFTR in Rhesus macaques via transcomplementation. Monkeys were sacrificed at day 90. This protocol was designed to evaluate the immune response to the AAV1 capsid and the level of expression of the transduced proteins. Both males and females gained weight normally. Likewise, there were no adverse events related to the study indicating that triple dosing with AAV1 vectors is safe. Neutralizing antibody titers were measured in serum and reported as the highest serum dilution that inhibited AAV1 transduction by 50%, when compared to its own AAV1 vector-positive/serum-negative control. All animals used in the study had low neutralizing anti-capsid antibodies (<5 reciprocal dilution, RD) at the beginning of the study. The levels increased in all animals to approximately 380, 30 days after the first dose. Neutralizing antibody titer increased to 1600, 30 days after the second dose of AAV1-Δ27-264-CFTR and to greater than 5000 after the third dosing of AAV1-GFP. Samples were taken for vector genomes in lung regions. At necropsy, approximately 2x10 6 vg/µg DNA was detected for CFTR and 5x10 5 vg/µg DNA for GFP. Both AAV1-CFTR, 2x10 6 vg/µg DNA, and AAV1-GFP, 4x10 6 vg/µg DNA were present in the liver. Widespread immunostaining for CFTR and GFP were detected in trachea, bronchi, and alveoli indicating that transduction of AAV1-CFTR and GFP occurred. Significant expression of both CFTR and GFP were noted throughout the lung. In conclusion, repeat dosing of AAV1 into the lung of Rhesus macaques is safe but leads to increases in neutralizing antibodies against the vector capsid. Detection of GFP protein expression following two doses of AAV1-CFTR suggests that even though increases in neutralizing antibodies are evident, transduction by AAV1 based vectors still occurs throughout the airway. Funded by CFF and NHLBI. Cystic fibrosis (CF) respiratory infections are polymicrobial and difficult to treat. Accumulating evidence suggests interspecies competition and cooperation are key determinants of microbial survival during infection and influence patient outcomes. Our patient studies reveal coinfection with the two most prevalent and problematic pathogens in CF, Pseudomonas aeruginosa and Staphylococcus aureus, correlates with poor lung function and increased frequency of pulmonary exacerbations. Accordingly, cocultivation of P. aeruginosa and S. aureus in vitro dramatically alters each species' tolerance to antimicrobials and production of virulence factors. However, whether these pathogens spatially interact during CF pulmonary infection in a manner sufficient to drive interspecies interactions influencing lung function is unclear. Here we sought to answer this question through fluorescent in situ hybridization (FISH) labeling of P. aeruginosa and S. aureus in ex vivo respiratory samples. Respiratory samples were fixed and optically cleared such that the structure of each respiratory sample and associated microbial community could be maintained and three-dimensional community images were acquired using laser scanning confocal microscopy. Prior to fixing, a section from each sample was homogenized and plated on selective medium to enumerate colony forming units for each species. For each sample determined to be culture positive for both organisms, mixed aggregates of P. aeruginosa and S. aureus could be visualized in close proximity. To understand how these polymicrobial communities form, in vitro assays were designed to visualize and tract single P. aeruginosa and S. aureus cells prior to the development aggregate formation. We observed that P. aeruginosa can sense the presence of secreted products from S. aureus from a distance. Upon sensing S. aureus exoproducts, P. aeruginosa responds by increasing both flagella and type-IV pili (T4P) mediated motility and travels directionally towards S. aureus. Upon arrival at a growing S. aureus colony, P. aeruginosa adopts a unique T4P-mediate motion whereby it orients its cell body perpendicular to the S. aureus colony, moves back-and-forth along its long axis and "burrows" into the S. aureus colony. P. aeruginosa motility is then decreased and depending on the strain and nutritional conditions, P. aeruginosa will either kill S. aureus or form a commensal-like mixed microcolony. Fluorescent reporters for key P. aeruginosa genes important for motility, chemotaxis, and biofilm formation were constructed and imaged by time-lapse microscopy. P. aeruginosa strains deficient in the corresponding genes were also analyzed by microscopy and in traditional macroscopy motility assays in the presence of cell-free S. aureus supernatant. It was determined that these behaviors are controlled at various levels by P. aeruginosa chemotaxis systems (Che and Pil/Chp) and the second messengers cAMP and c-di-GMP. We propose that by understanding how mixed microbial communities form, we may design therapeutics to disrupt formation of communities which have detrimental effects for patients and conversely, bring those together which might possess beneficial interactions. Introduction: Pseudomonas aeruginosa is a gram-negative bacterium linked to morbidity and mortality in patients with cystic fibrosis (CF). P. aeruginosa infects nearly 60% of all CF patients, with up to an 80% prevalence in adults with CF. Once it is established in the airways, P. aeruginosa is nearly impossible to eliminate. In response to the alarming increase in bacterial resistance to antibiotics, recent research efforts have centered upon disrupting quorum sensing (QS) as an alternative therapeutic approach. QS is a chemical communication system employed by P. aeruginosa and another prominent CF pathogen, Burkholderia cenocepacia, to express a wide array of virulence factors necessary for establishing infection. In gram-negative bacteria, QS is often mediated by acyl-homoserine lactone (HSL) signal molecules. P. aeruginosa possesses two complete acyl-HSL QS systems: LasR-I and RhlR-I. In studied strains of P. aeruginosa, LasR regulates RhlR in a "cascade hierarchy" of QS, and mutations in the lasR gene functionally inactivate QS. Mutations in lasR are common in CF patients chronically infected with P. aeruginosa, which has dampened enthusiasm for QS inhibition as a therapeutic strategy. However, we have recently shown that the cascade hierarchy of QS is altered in CF isolates and that RhlR mediates QS in many LasR-null isolates.

Hypothesis: LasR-null, RhlR-active clinical isolates remain virulent. Further, given the rarity with which mutations in rhlR are observed compared to lasR, we hypothesize that inactivation of RhlR may be disadvantageous to P. aeruginosa and therefore presents a therapeutic target.

Methods: We used an in vivo-like three-dimensional (3-D) lung epithelial cell model to assess the virulence of LasR-null clinical isolates that retain RhlR QS activity. We performed co-culture experiments using parent isolates and isogenic rhlR deletion mutants, and grew these strains either in monocultures or co-cultures to determine the effect of RhlR inhibition on intraspecies competition.

Conclusions: RhlR-regulated products are sufficient to induce cytotoxicity in a 3-D lung model in a subset of these LasR-null clinical isolates. Secondly, the co-culture experiments revealed that deleterious mutations in rhlR impart a competitive disadvantage relative to parent isolates, but only in competition. This disadvantage of RhlR mutants occurs through killing of RhlR mutants by the parent strain. Together, these observations support the notion that RhlR may be a more relevant and effective therapeutic target than LasR for treatment of chronic P. aeruginosa infections. Introduction: Upon inhalation, Pseudomonas aeruginosa (PA) adapt to the airway, form biofilms and cause chronic pulmonary infection in cystic fibrosis (CF) patients. Although persistent infection is commonly associated with damaging inflammation and mucus accumulation, we postulated that it is the presence of succinate released from activated airway immune cells that is a major factor in the adaptation of PA to the CF lung.

Methods: Genomic analysis of 17 PA CF isolates was performed. These strains were obtained from sputum of a chronically infected CF patient. Levels of mRNA for mutated genes were analyzed by qRT-PCR and compared to PAO1, a wild-type (WT) laboratory strain. C57Bl/6 mice were infected intranasally either with these mutants or PAO1. Immune cell recruitment to the lung was analyzed by flow cytometry and cytokines by ELISA. HIF1α and IL-1β were analyzed by Western blots. Airway succinate was quantified by using an enzymatic-colorimetric assay. Exogenous succinate was dosed intranasally to infected mice. In vitro studies were conducted to evaluate how succinate repressed the use of other carbon sources, such as glucose. Peripheral blood mononuclear cells (PBMCs) from healthy and CF patients were obtained, infected with WT PA and secreted succinate was quantified.

Results: CF PBMCs infected with PA were found to secrete more succinate than control cells. PA commonly form biofilms to maintain infection in the CF lung. However, biofilm formation is incompatible with the metabolism of succinate (70-80% reduction), known as a catabolite repressor and preferred PA carbon source. The CF strains of PA were less immunogenic mutants that failed to induce succinate secretion by host phagocytes. These mutants failed to activate the inflammasome, an intracellular immune signaling complex that uses succinate to stabilize HIF1α and secrete IL-1β. Phenotypic, genomic and qRT-PCR analyses of these CF isolates confirmed that these strains were either mucoid or small colony variants (SCV). In all of them we found highly conserved gain-of-function mutations not only in pro-biofilm genes, but also in zwf (glucose-6phosphate dehydrogenase, 40-fold), the Entner-Doudoroff and glyoxylate shunt pathways (80-fold), which support exopolysaccharide biosynthesis and minimize the generation of internal oxidants. These antibiotic resistant mutants, optimized for glucose utilization, could not tolerate the oxidative stress generated by succinate. These metabolo-adapted PA strains colonized the airway in a long-term, which was prevented in a 99% with the addition of a single dose of exogenous intranasal succinate.

Conclusions: Succinate generated in the hyperinflammatory CF airway drives PA adaptive mutation, and paradoxically, may provide a useful therapeutic agent.

Acknowledgements: A.P. is supported by NIH 1R35HL135800. SR is supported by a CFF Postdoctoral Fellowship CFF RIQUEL 17F0/ PG008837. Introduction: Stenotrophomonas maltophilia (Sm) is a gram-negative bacillus known to colonize the cystic fibrosis (CF) airway in patients with advanced lung disease. At present, it is still unclear whether presence of this organism is a marker of severe disease or the causative agent of exacerbations. Moreover, although it is well established that Sm can be detected in concert with other common CF pathogens such as Pseudomonas aeruginosa (Pa), the effect of polymicrobial interactions on CF disease progression is also unclear.

Methods: To investigate the contribution of Sm to disease progression, we established a model of acute infection in both wild-type (WT) and CFTR -/-mice, and measured bacterial persistence and virulence as quantified by viable colony counting, weight loss, lung histopathology, differential cell counts, and cytokine analysis. Confocal microscopy of immunofluorescently stained lung sections demonstrated the organization of bacteria in multicellular communities within the lungs. As this bacterial species is frequently co-isolated with Pa, we sought to define this interaction by assessing dual in vitro biofilms via crystal violet staining, viable colony counts and confocal microscopy of fluorescently tagged bacteria. We then established a murine model of co-infection with a mucoid strain of Pa, and the pathological consequences of this were quantified as described for the single-species infection.

Results: Infection with sputum-and nonsputum-derived strains of Sm led to significant weight loss and immune cell infiltration in the lungs of infected mice. Confocal microscopy of lung sections revealed distinct multicellular foci at 24 hours postinfection, consistent with tissue-associated infection. Cytokine analysis showed a robust immune response driven by early inflammatory markers including IL-6, IL-1α, and TNF-α, consistent with an acute exacerbation event. Sm forms biofilms with both mucoid and nonmucoid strains of Pa with no growth defect up to 24 hours, and confocal imaging of these biofilms revealed an evenly distributed structure between the two organisms. During dual infection in a murine model, not only are these organisms able to coexist within the lung, but the presence of Pa confers a significant increase in Sm persistence.

Conclusions: Based on these results, we conclude that Sm initiates significant pathologic changes during an acute infection, and may form biofilm communities within the lungs. In addition, these data show that Sm can establish a cooperative infection in concert with Pa. Current work focuses on understanding the mechanism behind this dual species cooperation, particularly in the context of CFTR -/-mice. Methods: We obtained 24 longitudinally-collected samples from a single patient (PT101) with chronic M. avium complex (MAC) infection over 5 years. DNA was extracted and WGS completed on 24 isolates using Illumina NextSeq. Two independent, high-resolution bioinformatic pipelines at our separate institutions were employed for postsequencing processing and analyses. Genomes were assembled via de novo and reference-based alignments. SNP tables were built from reference-alignments and concatenated core gene alignments and phylogeny inferred with Maximum-likelihood and Bayesian optimality criteria. We identified SNPs affecting protein-coding function and accessory genes unique to our isolates. Additional bioinformatic tools were used to locate antibiotic resistance genes and estimate horizontal gene transfer (HGT).

Results: All 24 isolates were identified as M. avium subsp. hominissuis. Genomes averaged 5.3 Mb with 68.9% G+C content. Phylogenetic inferences using core and reference-based approaches revealed two distinct but closely related MAC clades. Temporally-related isolates were dispersed throughout the two clades, suggesting maintenance of diversity over time. Fifteen core genes were identified with SNPs affecting proteincoding function, including those involved in invasion and persistence in macrophages (mce) and oxidative/antibiotic stress responses. We also identified 102 unique core genes in our isolates that were absent in all other publicly-available complete MAC genomes, many of which likely derived from a large conjugative plasmid. In addition, our preliminary analyses suggest high levels of recombination and potential HGT between MAC and co-infecting pathogens.

Conclusion: Chronic MAC infection in PT101 was characterized by long-term maintenance of population diversity with the accumulation of likely adaptive genomic changes over time. Our dual-institution, two-pronged (de novo and reference-based) approach (1) identified genes that may enhance the development of chronic infection; and (2) highlighted recombination and HGT as major evolutionary modalities for in-host MAC adaptation. Understanding the genetic determinants of chronic NTM infection, which will require implementation of our approach in multiple patients in parallel, will help identify biomarkers of disease, guide clinical decision-making, and eventually inform the design of novel, targeted interventions that can subvert pathogen adaptations. Background: Chronic bacterial airway infections in CF are frequently polymicrobial and may encompass pathogen interactions that affect clinical outcomes. Thus, we aimed to examine for the presence of distinct microbiologic phenotypes and their effect on pulmonary outcomes using a large population-based approach.

Methods: The CF Foundation Patient Registry data (CFFPR) from 2003 through 2015 was divided into two 5-year cohorts and stratified for adolescents (age ≥ 13 years) and adults (≥ 18 years). Culture data on Pseudomonas aeruginosa (PA), Achromobacter xylosoxidans, Burkholderia cepacia complex, Haemophilus influenzae, Staphylococcus aureus, and Stenotrophomonas maltophilia were used. Aside from PA (chronic infection definition), all other organisms were classified as present/absent if at least one culture in the 1-year run-in period was positive. Each 5-year cohort was divided into development and test cohorts (80%:20%). In each cohort, clustering tendency was assessed with visual plots and the Hopkins Statistic (threshold 0.5). The optimal numbers of clusters were determined through direct and statistical methods. Next, two unsupervised clustering algorithms (partitioning around medioids (PAM) and hierarchical) were used to develop the clusters. Clustering analyses were repeated with and without inclusion of patients with no positive cultures (NPOS) at baseline. Internal and external validation methods were applied to constructed clusters. Clinical outcomes of FEV 1 % decline, pulmonary exacerbation (PEx) and hospitalization frequency were assessed between clusters in the follow-up period.

Results: A total of 3,845 adolescents (3,059 /786) and 8,533 adults (6,843/1,690) were analyzed (2005) (2006) (2007) (2008) (2009) (2010) . The adolescents had a mean age of 15 years (SD 1.4); adults had mean age of 28.2 years (SD 9.7) and both groups were composed of 53% males. The development and test groups in the pediatric and adult cohorts were comparable at baseline. In the adolescent group, 8 and 7 distinct clusters were identified with and without NPOS patients respectively with PAM; similarly, 8 clusters were identified using the hierarchical method. In adults, using PAM, 5 and 2 clusters were identified with and without NPOS patients respectively; five clusters were identified with hierarchical methods. In both adults and adolescents, there were significant differences between clusters in subsequent FEV 1 % decline as well as PEx and hospitalization frequency (p<0.001); for example, one cluster in the adult population had a mean FEV 1 of 58.5% compared to another cluster with a mean FEV 1 of 64.6%. Internal and external validation testing of the constructed clusters were robust.

Conclusions: Our analyses of population-based CF registry data were able to identify distinct microbiologic phenotypes in both adolescents and adults that were significantly associated with meaningful pulmonary outcomes. Given the complex polymicrobial milieu of CF airways, consideration of bacterial interactions and the roles they may play on clinical outcomes is of utmost importance. Our approach robustly links microbiologic phenotypes with pulmonary endpoints and shows promise as a prognostic tool to predict clinical outcomes in CF. Rationale: Burkholderia cepacia complex (BCC) represents a classic pathogen group with significant attributable morbidity and mortality in CF. We aimed to determine the clinical outcomes following incident BCC infection by species in a contemporary CF cohort.

Methods: We conducted a retrospective study from January 1997 -February 2018 of pediatric and adult patients with CF followed at The Hospital for Sick Children and St Michael's Hospital (Toronto, Canada). Patients with BCC infection were categorized as prevalent infections if the first positive culture occurred prior to the study period and incident infection if it occurred during the study period. Cox models were constructed to assess the hazard ratio (HR) of death or transplant as a composite outcome. Mixed effects linear and logistic regression models were used to determine the rate of lung function decline (FEV 1 %) and the odds of pulmonary exacerbation (PEx) in the Prevalent and Incident cohorts compared to those without BCC infection. For the Incident cohort with species data, we examined the above outcomes by species (B. cenocepacia including ET-12 vs others). Multivariable models incorporated known confounders using a forward step-wise approach at an a priori p-value of 0.15 using STATA 14.2.

Results: A total of 1313 patients (55% male) were followed over the 21-year period. Of these, 210 (16%) had at least one culture positive for BCC in the study period (131 Prevalent and 79 Incident). In the Incident cohort, the most prevalent species were B. multivorans (40.9%), B. cenocepacia (38.6%) including the ET-12 epidemic strain, and B. gladioli (13.6%). There was no significant difference in the HR of death or transplant in the Prevalent (HR 1.53, 95% CI 0.60 -3.96) and Incident (HR 0.41, 95% CI 0.05 -3.09) groups compared to those with no BCC infection. Both the Prevalent and Incident groups had a greater rate of FEV 1 % decline compared to those with no infection although only significant in the Incident group in an adjusted model (-1.35%/y vs -1.13%/y, p < 0.001). For PEx, both the Prevalent (OR 1.67, 95% CI 1.14 -2.42) and Incident groups (OR 1.66, 95% CI 1.08 -2.53) had significantly increased OR relative to those with no infection in an adjusted model. When species-based analysis was done for the Incident group, although no differences in HR of death or transplant were noted, both the B. cenocepacia and others groups (-1.43%/y vs -1.08%/y in no infection group) had a similar and significantly greater rate of FEV 1 % decline in an adjusted model. Additionally, both species groups had a greater OR of PEx compared to those with no infection though only the non-cenocepacia species group remained significant in an adjusted model (OR 2.04, 95% CI 1.09 -3.81).

Conclusion: Our longitudinal study demonstrated differences in pulmonary outcomes between those with BCC infection compared to those without infection including by species groups. Given the changing epidemiology of BCC, species-based assessment to determine optimal management strategies is important. The source of bacteria causing airway infections in people with CF is often thought to be the environment, although it is unclear which specific environmental exposures pose the greatest risk of infection. Previously, we detected Achromobacter, Burkholderia, nontuberculous mycobacteria (NTM), Pseudomonas aeruginosa and Stenotrophomonas maltophilia in 90% of water and aerator biomass samples collected from homes of individuals without CF. We therefore hypothesized that people with CF are exposed to and may acquire opportunistic bacterial pathogens from their home drinking water.

Objective: To determine if persons with CF acquire infection from bacterial strains found in their home drinking water.

Methods: Drinking water was sampled from the plumbing of homes of 17 children with CF who had recently (within 6 months) become respiratory culture positive for a CF pathogen. All homes were located in Michigan and were served by chlorinated or chloraminated drinking water. Within each home, swabs of biomass from the kitchen faucet aerator, filtered biomass from 1 L of stagnated (> 6 h) premise plumbing water, and 4 L of hot water from the shower or bath were collected. Biomass from an aerator and filtered biomass from 1 L of water were also collected from a second faucet used most frequently by the subject. The genetic relatedness of pathogens detected in drinking water and bacterial strains recovered from subjects was determined using PacBio DNA sequencing of select genomic loci.

Results: DNA sequences from CF pathogens were detected in water from each home. Considerable strain-level diversity was observed in water samples, with P. aeruginosa having the greatest number (N = 122) of distinct strains. Overall, 94% of drinking water samples contained DNA sequences that were closely related (<5 SNPs) to the bacterial strains recovered from subjects residing in those homes. DNA sequences that were identical to those from infecting bacteria were found in the homes of four subjects (subjects infected with Achromobacter, Burkholderia, or P. aeruginosa). Although no household location was found to be the sole source of infecting strains, kitchen aerators and flow through point-of-use filters were the most common sources, identifying these features as potential targets for risk mitigation strategies.

Conclusions: Our findings indicate that home plumbing drinking water can be a source of bacterial strains that cause infection of the respiratory tract of children with CF. The extent to which bacteria in home drinking water contribute to respiratory tract infection of people with CF is not clear and in need of further study. The tools developed in this study allow culture-independent, high-throughput strain-level screening for CF pathogens in environmental samples. This novel approach has far reaching applications for the identification of other sources of clinically relevant pathogens that pose a risk to people with CF.

Acknowledgment: This work was supported by CFF award LIPUMA15G0. (2015). Home: A minimum of 3 water points were sampled from participants' homes; shower, bathroom sink and kitchen sink taps. The participants were divided into 4 groups: 1) CF and active MABSC respiratory infection (<2 years since positive culture; n=10), 2) CF and past MABSC respiratory infection (>2 years since positive culture; n=7) or 3) CF and never infected with NTM (n=8), and 4) homes of people with no lung disease (n=8). All water samples were incubated for 8 weeks. The genomes of presumptive NTM isolates were sequenced.

Results:Hospital: Detection of MABSC in water samples collected at TPCH was variable longitudinally with a range of 0-25% water points positive per time point. At the RCH, MABSC was detected in water samples collected from 2/12 (17%) water points. Home: MABSC was detected in 6/136 (4%) home water points. Genomic analysis of MABSC isolates from hospital and home water showed them to be closely related to respiratory MABSC isolates from people with CF.

Conclusions: The genetic relatedness of the water and respiratory MABSC isolates demonstrates that potable water may be a potential source of MABSC respiratory infection in some people with CF.

The updated cystic fibrosis (CF) infection control guidelines recommend that people with CF wear face masks when in communal hospital areas to prevent the release of infectious aerosols (Saiman L, et al. Infect Hosp and Epidemiol. 2014;35:S1-S67). Recently, we found that short-term wear of face masks (~10 minutes) reduced the release of Pseudomonas aeruginosa aerosols during coughing (Wood M, et al. Am J Respir Crit Care Med. 2018; 197:348-55) . However there is limited evidence to determine if face masks continue to be effective at reducing the release of infectious cough aerosols after longer wear times.

Methods: We recruited 25 adults with CF and chronic P. aeruginosa respiratory infection (Ramsay K, et al. J Cyst Fibros. 2016 ;15(S1):S70) and 10 healthy volunteers to the study. All participants underwent up to five cough tests (randomised order) in our validated cough rig (Knibbs L, et al. Thorax. 2014; 69:740-5) : 1) uncovered cough; 2) coughing with surgical mask immediately applied; 3) coughing with surgical mask worn for 10 minutes; 4) coughing with a surgical mask worn for 30 minutes; 5) coughing with an N95 respirator worn for 10 minutes (optional test). Participants entered the cough rig and performed 2 minutes of tidal breathing before and after a 5 minute cough period. Therefore, an additional 10 minutes was added to the mask wear time (denoted as "total wear"). Quantitative sputum samples and cough aerosols were collected and analysed from participants with CF as per previous methods (Wood et al, 2018). All participants rated their level of comfort after each test.

Results: Surgical masks and N95 respirators were effective at reducing the release of P. aeruginosa aerosols during coughing after 40 minutes total wear and 20 minutes total wear by more than 90%. Participants in both the CF and healthy volunteer groups rated the surgical masks as more comfortable compared to N95 respirators.

Conclusions: Surgical masks are the preferred mask to wear as source control and are effective at reducing the release of infectious P. aeruginosa aerosols during coughing after 40 minutes of total wear. Supported by: Cystic Fibrosis Foundation Therapeutics (USA), The Prince Charles Hospital Foundation (Australia), Advance Queensland (Australia). Species within the M. avium complex are commonly isolated from the environment in soil, water and water supply systems, and are believed to be acquired from these sources. Species within the M. abscessus complex are rarely isolated from environmental sources, and studies have implicated potential for direct or indirect healthcare-associated transmission in CF clinics. There is a need for an evidence-based approach to investigating potential episodes of transmission.

Hypothesis: An investigational tool can be developed that will identify shared healthcare-associated source(s) between two or more patients with highly similar NTM isolates in the setting of CF center care.

Methods: The CF Foundation-funded Colorado CF Research and Development Program was established in 2015 to advance research and treatment of NTM infections in CF. CF care centers are encouraged to submit NTM respiratory isolates for analysis. The NTM Culture, Biorespository, and Coordinating Core provides state-of-the-art support for culture, molecular identification, antimicrobial susceptibility, as well as banking of isolates. NTM isolates are then forwarded to the Molecular Core, which performs genomic "fingerprinting" of NTM isolates via whole genome sequencing applying next generation DNA sequencing. We have identified multiple clusters of NTM isolates which are highly similar between two or more patients, raising the concern for patient-to-patient transmission within CF care centers or acquisition originating from a common source. Using integrated clinical and epidemiological research methods, we have developed an investigational tool that will aid in identification of shared healthcare-associated source(s) between two or more patients with highly similar NTM isolates in the setting of a CF care center. The Healthcare Associated Links in Transmission of NTM (HALT NTM) epidemiological investigation tool facilitates a stepwise process by which centers may perform data abstraction on patients identified with highly similar NTM isolates.

Results: Initial results indicate patient pairs with unrelated strains reveal few or no healthcare-associated points of contact. In contrast, pairs sharing highly related isolates reveal overlapping points of contact. In some cases, a specific instance of overlap has also been identified.

Conclusions: Our findings suggest the newly developed HALT NTM instrument is an epidemiological investigation tool that may systematically identify a shared healthcare-associated source(s) between two or more CF patients with highly similar NTM isolates. This investigation highlights that ongoing evaluation of clusters of highly related NTM isolates among patients within individual CF care centers is warranted. NTM infections may be spread among CF patients within medical care settings. Strict infection control procedures have decreased spread of other bacterial infections among CF patients. However, this finding highlights the need for ongoing investigation into risk factors that may contribute to healthcare-associated transmission of NTM among CF patients.

Funding: CFF.

Background: Although risk factors for failing to respond to intravenous (IV) antibiotic treatment in pulmonary exacerbation (iPEx) have previously been described, much less is understood about risk factors for failure to recover lung function following events treated with oral antibiotics (oPEx). However, these events are common and have a significant impact on lung function in individuals with CF. The aims of this study were thus to identify patient characteristics associated with: 1) failure to recover to ≥90% of baseline lung function following treatment for oPEx, and 2) a shorter time from first oPEx during the study period to the following oPEx.

Methods: This was a retrospective cohort study of patients with CF with oPEx treated at The Hospital for Sick Children and St. Michael's Hospital from 2009 to 2016. We investigated patient characteristics associated with failure to recover ≥90% ("Non-Responder") of baseline lung function (highest forced expiratory volume in 1 second percent predicted (FEV 1 pp) in prior 6 months) following oral antibiotic therapy using a predictive model with generalized estimating equations to account for repeated events. We also compared time to subsequent oPEx between Responders and Non-Responders.

Results: There were a total of 3054 oPEx events in 547 patients during the study period. The median age at oPEx was 25 years and 46% of the entire cohort was female. The majority (83.6%) of patients were pancreatic insufficient and had class I-III mutations (86.2%). Overall, 914/3054 (29.9%) cases of oPEx episodes were considered Non-Responders. Using multivariable analyses, lower BMI (OR 2.00 (95%CI 1.23;3.22), p<0.01) as well as lower baseline lung function (OR 1.02 (95%CI 1.01;1.02), p<0.01) were associated with failure to recover ≥90% of baseline lung function following oPEx. In addition, an increased drop in lung function from baseline to FEV 1 pp at oPEx also increased the odds of being a Non-Responder (OR 1.10 (95%CI 1.08-1.12), p<0.01). The median time to next oPEx was 223.5 days (interquartile range 91-508). Patients who were chronically infected with Pseudomonas aeruginosa (≥2 positive sputum cultures in the year prior) had increased hazards of earlier next oPEx (HR 1.46 (95%CI 1,02;2.90), p=0.04). Other baseline patient characteristics (including age, sex, mutation class, pancreatic status, BMI, lung function measures and other infections) did not contribute significantly to time to next oPEx.

Conclusions: The risk factors for being Non-Responder following oPEx treatment included lower BMI, lower baseline lung function and a greater drop in lung function from baseline. In addition, patients with chronic P. aeruginosa infection had a shorter time to next oPEx. Bacteria in chronic infections often reside in communities composed of micron-sized, highly dense aggregates (~10 1 -10 4 cells). However, how these structured communities form or the role they play in chronic infection is unknown. Although the aggregate mode of growth is associated with increased tolerance to host immune functions and antimicrobial therapies, a primary challenge has been the lack of laboratory systems in which to study aggregates that promote natural aggregate formation in relevant environments. We developed an in vitro model based on chronic infection of the cystic fibrosis (CF) lung, which promotes natural aggregate formation by the bacterium Pseudomonas aeruginosa (PA). When cultured in synthetic CF sputum media (SCFM2), PA readily forms aggregates of similar sizes to those observed in CF lung tissue. In combination with high resolution microscopy and micro-3D-printing, we have established SCFM2 as a platform to observe the assembly of microbial communities, in three-dimensions, in real-time and at the micron-scale. We have exploited this system in two ways; first, to elucidate the life history of PA and the mechanisms that this bacterium utilizes to tolerate antimicrobials, specifically bacteriophage. Our results reveal that this bacterium readily forms aggregates that release migrants to colonize new areas. We show that aggregates allow PA to tolerate therapeutic bacteriophage addition, in part due to exopolysaccharide production; however, seeding of new aggregates by dispersed migrants was inhibited. We propose a model in which aggregates provide a mechanism that allows PA to tolerate an antimicrobial during chronic infection without the need for genetic mutation. Secondly, we assessed interactions between aggregates. As a highly social organism, PA interacts both with itself and other microbes that co-inhabit the same environment. One manner in which PA communicates is via quorum sensing (QS), a phenomenon thought to be critical for emergence of in vivo microbial communities. What remains unanswered is whether such signalling occurs as a localized event (intra-aggregate) or at the community level (inter-aggregate). To answer this question, we used a mixture of micro-3D-printed and naturally formed SCFM2 aggregates to explore the "calling distance" of QS signals. We found that signaling in SCFM2 is primarily intra-aggregate, with only aggregates of > ~5000 cells able to engage in inter-aggregate signaling. In turn, this "calling distance" is impacted by the physical properties of CF sputum. These studies provide benchmark data on the physical requirements required for bacterial tolerance and communication in an environment similar to the CF lung. Using SCFM2 as a model for studying aggregates will allow us to further explore PA chronic infection in the CF lung, from a clinical, evolutionary and ecological perspective, at the finer scale of an individual aggregate. Bacterial-viral co-infections are recognized as a cause of morbidity in chronic lung disease, including cystic fibrosis (CF). Polymicrobial infections occur frequently in the CF airway, but the underlying mechanisms of how viral-bacterial interactions mediate pathogenesis remain poorly understood. In CF patients, clinical observations link the acquisition of chronic Pseudomonas aeruginosa with respiratory virus infections. Recently, we demonstrated respiratory syncytial virus (RSV) infection, and the subsequent innate immune antiviral interferon (IFN) response, promoted biofilm growth of P. aeruginosa. In the current study, our goal was to evaluate innate immune effectors that mediate enhanced bacterial biofilm growth. We performed a high-throughput screen using a collection of over 400 previously identified interferon-stimulated genes (ISGs) inserted into a lentiviral vector. We screened the ISG library with biofilm growth assays using airway surface liquid from ISG-expressing CF airway epithelial cells. Biofilms were visualized using epifluorescence microscopy and quantified using an automated volumetric image analysis program. A Z-score was calculated based on biomass for each ISG screened, resulting in six hits. Hexokinase-2 (HK2) is a screen hit, and was upregulated during RSV infection and IFN-beta treatment. HK2 catalyzes the first step of glycolysis and is known for its role in promoting aerobic glycolysis, or the Warburg effect, where cells increase lactic acid fermentation and lactate secretion. We demonstrated that RSV infection and the IFN response cause robust lactate production by the epithelium. Production of lactate is concurrent with induction of aerobic glycolysis genes in the epithelium, as assessed by dual-species RNA sequencing (dual-seq). Induction of the Warburg effect has been reported for other respiratory viruses, including influenza A virus and adenovirus. Our work is the first to demonstrate a role for Warburg metabolism in RSV infection. Lactate is a preferred carbon source to support P. aeruginosa growth, and dual-seq studies reveal P. aeruginosa lactate utilization genes, lldD, lldE, and lldP, were upregulated in the presence of RSV co-infection. Inhibition of host-cell lactate production during IFN-beta treatment diminished P. aeruginosa biofilm growth. To determine if Warburg metabolism is induced in vivo during RSV infection, we used a mouse model of RSV infection and demonstrated an increase of lactate in the bronchoalveolar lavage fluid. Our results reveal that RSV infection and the IFN response promotes aerobic glycolysis and primes the CF airway environment with additional metabolites to foster P. aeruginosa biofilm growth. By advancing our understanding of host immunological-pathogen crosstalk, we may improve the design of targeted treatments to prevent P. aeruginosa biofilm conversion and decrease morbidity in CF.

Funding by CFF BOMBER14G0, NIH R01HL123771, Vertex Research Innovation Award. Introduction: Decreased mucociliary clearance results in obstruction of the sinus ostia in individuals with cystic fibrosis (CF). These conditions potentiate bacterial colonization associated with the high prevalence of chronic rhinosinusitis (CRS) in the CF population. While chronic bacterial infection of the CF airways is thought to be polymicrobial, a dominance of Staphylococcus aureus is associated with CRS in CF patients. S. aureus is also considered a commensal organism in the airways of ~30% of non-CF individuals, leading us to consider differences in the CF sinus microenvironment that may contribute to S. aureus virulence such as nutrient acquisition and microbiota composition. We hypothesized that mucins, the major component of mucus and enriched in the CF airways, could be a nutrient source for S. aureus contributing to its fitness in the sinuses. Furthermore, we considered bacterial consortia associated with S. aureus colonization of CF sinuses.

Methods: A chemically defined medium (CDM) containing porcine gastric mucin (PGM) or glucose was used to measure growth of S. aureus LAC, MN8, and 2 CF/CRS clinical isolates. To assess bacterial community composition associated with the presence of Staphylococcus, sinus mucus samples collected from CF (n=16) patients and non-CF (n= 23) individuals during endoscopic sinus surgery were analyzed using 16S rRNA gene sequencing.

Results: Under aerobic conditions, S. aureus strains exhibit an overall growth defect on mucin compared to glucose as a source of carbon. However, growth on mucins appears to be more resilient to the depletion of amino acids from CDM, signifying mucin as a potential source of amino acids for S. aureus under respiratory conditions. Growth of S. aureus on mucins under anaerobic conditions typical of the CF sinus microenvironment was not observed. While mucin may not constitute a major nutrient source for S. aureus under nonrespiratory conditions, ordination analysis using 16S rRNA gene sequencing revealed that several anaerobic mucin-degrading members of the Bacteroidetes phylum such as Prevotella and Fusobacterium were enriched in CF samples containing sequences belonging to Staphylococcus. When grown on CDM with mucin in co-culture with an anaerobic mucin-degrading community composed of F. nucleatum, P. melaninogenica, V. parvula, S. gordonii, and S. cristatus, S. aureus LAC exhibited a ten-fold increase in CFUs compared to growth in monoculture.

Conclusions: Neither S. aureus lab strains, nor CF/CRS isolates can utilize porcine gastric mucin efficiently as a nutrient source. Bacterial consortia associated with CF and non-CF patients colonized by Staphylococcus are compositionally distinct, with several mucin-degrading bacteria associated with the sinus microbiome of CF patients. Their presence may be competitive or cooperative for S. aureus utilization of mucins. This ongoing research will better characterize the microenvironment and nutrient availability for S. aureus in chronic bacterial infections of the CF airways. Introduction: Cystic fibrosis (CF) airway is characterized by complex bacterial communities that incite persistent inflammation and airway damage. Based on our rabbit model of sinusitis, blockage of sinus ostia generates a shift in microbiota to a predominance of mucin degrading microbes (MDM) and the shift from acute to chronic sinus inflammation is associated with a robust increase in pathogenic bacteria (e.g. Pseudomonas). MDMs are known to produce mixed acid metabolites (short chain fatty acids (SCFA)), which have the potential to stimulate pathogen growth by offering a carbon source to nonfermenting sinus pathogens (e.g. Pseudomonas, Moraxella, Stenotrophomonas) that cannot utilize sugar. The objective of this study is to evaluate the concentrations of SCFA within the sinonasal mucus and its contribution to the growth of Pseudomonas aeruginosa.

Methods: Rabbit acute rhinosinusitis was induced by blocking the middle meati for two weeks to create an anaerobic environment for MDM. Healthy and sinusitis mucus were collected and co-cultured with PAO1 strain of P. aeruginosa for 72 hours and colony forming units were determined. Targeted quantification of SCFAs in healthy and sinusitis mucus from patients (with and without Pseudomonas infection) was performed via high performance liquid chromatography.

Results: To test whether the sinusitis mucus samples enriched with MDMs and SCFAs could simultaneously stimulate PAO1 growth, the mucus samples were co-cultured with PAO1 in a minimal mucin medium. After 72 hours, a diffusible blue-green pigment (pyocyanin) characteristic of P. aeruginosa growth was observed throughout the co-culture tubes contained with sinusitis mucus (week 2). Colony counts per tube were significantly higher in those tubes contained with the mucus samples from week 2 (8.4x10 9 ±4.8x10 7 ) compared to tubes containing control mucus (1.4x10 9 ±2.0x10 7 ) or no mucus (1.5x10 9 ±2.1x10 7 ) (p<0.0001). To provide evidence of fermentative activity in vivo, three SCFAs (acetate, propionate and butyrate) were quantified using GC-MS within the mucus samples from rabbits on day 0 (Control) and week 2 (Sinusitis). Acetate concentrations were significantly greater in the rabbit mucus samples collected on week 2, relative to day 0 (4.13±0.53 vs 1.94±0.44 mM, p<0.01). Finally, to provide the evidence of fermentative activity in human chronic rhinosinusitis (CRS), we analyzed the presence of SCFAs in human mucus. All SCFAs were significantly higher in CRS compared to controls (acetate, p<0.05; propionate, p<0.0001; butyrate, p<0.01). Of those 9 CRS patients, P. aeruginosa was grown in 5 patients and SCFAs from those patients were also significantly higher than those from controls.

Conclusion: Given that SCFAs are exclusively derived from bacterial fermentation, our evidence suggests a critical role for mucin-degrading bacteria in generating carbon-source nutrients for pathogenic bacteria. MDM may contribute to the development of recalcitrant airway disease by degrading mucins, thus providing nutrients for potential pathogens like P. aeruginosa. GI symptoms, including nutrient malabsorption, growth failure, and GI obstruction, are among the earliest and most severe CF manifestations. Most CF GI microbiome studies have focused on adults, who often receive antibiotics and other confounding therapies. By comparison, children with CF, particularly infants, are likely less impacted by these issues. We hypothesized that infants with CF have GI dysbioses compared with infants without CF that are (1) due to CFTR dysfunction, and are (2) independent of confounders such as diet and antibiotics. As the GI microbiota play important roles in nutrient harvest, vitamin levels, GI health, and growth, all of which relate to long-term health, the CF GI microbiota represents a potential, modifiable target for improving diverse early outcomes.

Methods: We performed shotgun metagenomic sequencing of fecal samples from two studies of infant nutrition: The 28-center Baby Observational and NUtrition Study (BONUS, 231 infants with CF studied during the first year of life), and the companion Healthy Infants Study (25 infants without CF in Buffalo, NY with study visit schedule paralleling BONUS). Samples were collected at 2-3, 4, 6, 9-10, and 12 months. Taxonomic profiles were inferred computationally using Metagenomic Phylogenetic ANalysis (MetaPhlAN).

Results: While differences between the control and CF infant fecal microbiota were small at 3 months of age, the average microbiota of the two groups became progressively distinct. These differences were largely characterized by the relative depletion of Bacteroides and Bifidobacterium spp., and the enrichment of Proteobacteria, in CF microbiota. Development of CF microbiota appeared to lag behind that of the controls, resulting in predicted differences in health-associated microbiota functions. These taxonomic differences could not be explained by differences in diet (including breast versus formula feeding or table food), treatments (including antibiotics), or geographical location.

Conclusions: These findings indicate that infants with CF have progressive fecal dysbioses impacting the abundances of important bacterial taxa that are likely due to CFTR dysfunction rather than influences such as diet or medications. Many taxa depleted in the CF microbiota, such as Bifidobacterium, Roseburia, Faecalibacterium, Eubacterium and Bacteroides spp., are known to be important for GI health, nutrient harvest from undigested food, and somatic growth. Conversely, several of the taxa enriched in CF samples, particularly Proteobacteria, are associated with GI inflammation and are potential pathogens. These results provide a critical baseline upon which the relationships between CF GI microbiomes and clinical outcomes, including somatic growth and intestinal function, can be determined. Introduction: S. aureus is the bacterium cultured most often from respiratory secretions of people with CF. S. aureus adapts genetically during chronic CF respiratory infections, including the formation of slowgrowing, antibiotic-resistant variants known as small-colony variants (SCVs). Because S. aureus SCVs do not grow well on standard culture media, they are not routinely identified by most clinical laboratories, precluding registry-based studies of prevalence and clinical associations. In a single-center study using special laboratory methods, we previously showed that S. aureus SCVs commonly infected children with CF and were associated with worse respiratory disease outcomes (Wolter DJ, et al. Clin Infect Dis. 2013; 57:384-91) . Treatment with sulfonamide antibiotics was a risk factor for SCVs. We sought to determine whether these findings are generalizable throughout the US.

Methods: We performed a 2-year, observational, longitudinal study of 231 CF children aged 6-16 years at enrollment at 5 CF centers across the US. Respiratory specimens collected during routine care visits were cultured by a central laboratory using methods sensitive for S. aureus SCVs. We determined prevalence of SCVs and their independent associations with lung function and exacerbations using generalized estimating equations with identity and logit links, respectively.

Results: S. aureus SCVs were identified in 28% of subjects. Children who ever had SCVs during the study had significantly lower lung function throughout the study than did those without SCVs (enrollment FEV 1 percent predicted 89.5±19.7 versus 95.7±18.8 respectively, p<0.01), but change in FEV 1 percent predicted did not differ after adjusting for baseline FEV 1 . Children with SCVs had significantly increased odds of having exacerbations (OR, 1.7, p<0.004), even when adjusting for age, sex, race, CFTR mutation (F508del homozygous, heterozygous, other) and FEV 1 at enrollment. Subjects with SCVs were more likely in univariate analyses to have been treated with sulfonamide antibiotics prior to SCV detection (OR, 2.2, p<0.04).

Conclusions: More than 25% of CF children in a multicenter US population were infected with S. aureus SCVs. SCV infection is strongly associated with lower lung function and higher exacerbation rates. Because SCVs can be selected by antibiotic treatments, which tend to be provided more frequently to patients with more severe disease, it is not yet clear whether SCV infection causes, or simply indicates, worse lung disease. An interventional study targeting S. aureus SCVs is required to determine whether SCV treatment or prevention improves outcomes. Until then, while we can recommend no changes in current treatment approaches, these results support the general adoption of clinical laboratory methods that identify S. aureus SCVs, and the inclusion of these variants in CF registry data for ongoing surveillance and study.

Acknowledgments: On behalf of the SCVSa study team and subjects. Support by the CFF. . Potential nephrotoxicity and ototoxicity make appropriate monitoring critical. The 2016 CF Foundation Patient Registry (CFFPR) reports a low incidence of hearing loss (1.1% of pediatric patients (≤18 years) and 2.2% overall). In the United States, 13% of the total population ≥ 12 years have hearing loss. A standardized aminoglycoside induced ototoxicity algorithm (AIOA) was implemented in 2017 at Children's Mercy Kansas City (CMKC) to assess CF patients treated with intravenous (IV) and/or inhaled AG.

Methods: The implementation process included a survey of providers, retrospective chart review, observational cohort analysis, and review of literature to develop the AIOA. The algorithm serves as a visual reference for clinicians and provides specific AG monitoring instructions including time of baseline and frequency of subsequent audiograms. It identifies risk factors warranting more frequent screening. The CF center coordinator and pharmacist are responsible for monitoring adherence to the AIOA including 1) identification of new patients for monitoring during preclinic huddles and hospitalizations; 2) review of monthly AG prescriptions; and 3) inpatient AG order review.

Results: Prior to implementation of the AIOA, 12 of 50 patients (24%) treated with IV AG over a two-year period had undergone an audiogram. Of the 70 patients that received at least two courses of inhaled AG in 2016, only 18 (26%) had an audiogram. In the 12 months post-AIOA implementation, 27 of 28 patients (96%) treated with an IV AG had an audiogram. Among these, 15 had an abnormal audiogram (56%); seven (47%) with distortion product otoacoustic emissions (DPOAE) abnormalities and eight (53%) with varying degrees of high frequency hearing loss (HFHL). Four patients were referred to otolaryngology for additional evaluation and therapy (27%) and the use of subsequent AG treatment was avoided in two patients. Over the same 12-month period, 14 of 30 patients (47%) that received inhaled AG for >5 years had an audiogram completed per the AIOA. Among these, seven (50%) had an abnormal audiogram. The majority of patients (6, 86%) had received at least one course of concurrent IV AG treatment. Among the patients with abnormal audiograms, three patients (43%) had DPOAE and four (57%) had varying degrees of HFHL. Two of these patients were referred to otolaryngology. A post-implementation provider survey demonstrated support for the AIOA and use of this information to influence treatment decisions. Barriers to AIOA adherence include lack of provider awareness and difficulty coordinating outpatient audiograms.

Conclusions: Implementation of an AIOA increased the frequency of audiograms among patients treated with IV and inhaled AG. The prevalence of hearing abnormalities at CMKC is much higher than that reported in CFFPR as well as the overall US population. This discrepancy may be secondary to AG usage specific to our center or lack of audiogram testing nationally. The frequent use of AG among CF patients and the high probability of AG induced hearing loss suggest an urgent need to establish an AIOA nationally. Introduction: Antibiotic therapy is essential for the treatment of cystic fibrosis (CF) lung infections. Selection is based on previous culture information, if available, or institution specific antibiograms (ABGM). Most institutional ABGM exclude CF patient cultures so it can be a challenge to assess changes in susceptibility patterns or appropriateness of empiric antibiotic selection. The character of CF airway infection and frequent exposure to antimicrobials increases the risk of developing multidrug resistant infections. Antibiotic selection is important in ensuring adequate antimicrobial coverage, while minimizing adverse effects and the development of resistance. The Children's Mercy Kansas City (CMKC) ABGM excluded CF patients, hindering patient care. A CF-specific ABGM (CF-ABGM) was developed in 2016. This abstract reports initial data and future plans.

Methods: CF patient culture information will be collected from 2016-2026 to develop a CF-ABGM. All CF patient cultures, sputum or throat swab, obtained at CMKC will be included in the CF-ABGM. Patients were identified via a microbiology report and the following data were collected for 2016 and 2017: demographics, microorganism isolates and susceptibility information. Susceptibility information was reported for methicillin-susceptible Staphylococcus aureus (MSSA), methicillinresistant Staphylococcus aureus (MRSA), Pseudomonas aeruginosa (PA), Achromobacter, Stenotrophomonas and Burkholderia species.

Results: Comparing the 2016 and 2017 CF-ABGM to the hospital ABGM, gram-positive and gram-negative microorganisms tested were less susceptible. CF isolates from sputum cultures were less susceptible than throat cultures. Both MSSA and MRSA had significantly lower susceptibility for clindamycin compared to hospital-wide rates (p<0.0001). MSSA and MRSA susceptibility rates for other antimicrobials tested in the CF-ABGM were similar to the hospital ABGM. Over the two-year period, the hospital-wide incidence of MRSA was higher than in the CF population (32-34% vs 27-28%). The most common gram-negative isolate for CF cultures was PA. For every antimicrobial tested, CF cultures had lower susceptibility rates than hospital-wide PA isolates including all aminoglycosides (p<0.0001).

Conclusions: Development of a CF-ABGM demonstrated significantly increased rates of clindamycin resistance for MRSA and MSSA isolates and more resistant PA isolates. This has important clinical implications for empiric antimicrobial selection and monitoring resistance trends over time. Hypothesis: Inhaled administration of QRM-003 will improve patient outcomes and reduce the duration of therapy.

Methods: QRM-003 is a novel inhaled therapy for the treatment of NTM infections, which utilizes clofazimine as the active pharmaceutical ingredient. Clofazimine has repeatedly demonstrated potent antimycobacterial activity against a variety of pathogens, but is limited by poor solubility and bioavailability. Efficacy of QRM-003 was investigated by in vitro macrophage uptake assay, in vitro antibiotic susceptibility testing, and two in vivo mouse models of NTM infection -SCID mouse infection with Mycobacterium abscessus (MABSC), and Beige mouse infection with M. avium (MAC). Animals were treated every other day with a) saline, b) QRM-003 (via aerosolized administration), or c) clofazimine alone (via gavage). Following the final treatment, tissues were recovered to determine bacterial recovery.

Results: QRM-003 was shown to be taken up by macrophages in vitro after 24-hour incubation. Antimicrobial susceptibility testing demonstrated QRM-003 MIC of 1 µg/mL against both MABSC and MAC. QRM-003 demonstrated antimicrobial activity in both animal infection models. In the MABSC infection model, the CFU recovered from lung, spleen and liver were as follows: saline ( Conclusion: QRM-003 has demonstrated potent antimycobacterial activity in vitro, and in vivo administration significantly reduced bacterial recovery in both acute MABSC and MAC infection models. Importantly, this activity was significantly greater than oral administration of the clofazimine alone, despite higher clofazimine dosing via oral administration. Future directions include in vivo infection models to investigate chronic infections and infections in CF-mouse models. Pharmacokinetic evaluations will be performed to quantify the tissue distribution of QRM-003 compared to oral administration of clofazimine. Following that, Qrumpharma will perform inhaled toxicology testing of QRM-003 in two animal species.

Hamdan, M.; Maupin, K. Pediatrics, West Virginia University, Charleston, WV, USA Purpose: Pulmonary exacerbation is a common cause of hospitalization in cystic fibrosis (CF) patients. Obtaining a respiratory culture on admission is important for the determination of appropriate antibiotic treatment. Unless it is clinically indicated, blood cultures may provide minimal diagnostic or clinical yield. However, bloodstream bacterial infections may be more common in specific subsets of CF patients, for example, patients with an intravenous access device (IVD). The purpose of the study is to identify the utility of obtaining blood cultures in pediatric patients with CF admitted for pulmonary exacerbations over a ten-year period.

Methods: We retrospectively analyzed medical records from patients (0-21 years of age) hospitalized for CF pulmonary exacerbations at CAMC Women and Children's Hospital from January 2008 through December 2017. We studied the patients who received a blood culture during their hospital stay. Then we identified whether pathogens were present and whether the patient had an IVD and fever.

Results: Out of the 121 admissions meeting inclusion criteria, 67 (55.4%) included a blood culture. Having an IVD, or requiring oxygen supplementation was associated with receiving a blood culture (p < 0.05). Four of the blood cultures that were positive for growth of bacteria were deemed to be contaminants with results not affecting patients' clinical management. Only 1 of the 67 cultures (1.5%) was positive for a pathogenic organism (methicillin-resistant Staphylococcus aureus) and considered a clinically-significant bloodstream infection. This patient had an IVD and fever at the time of admission.

In our study, we found that obtaining a blood culture from patients admitted with a CF pulmonary exacerbation has minimal clinical value, especially in the absence of IVDs. We plan to use this information to manage CF exacerbation patients in an efficient and cost-effective manner. We recommend further studies to assess the utility of blood cultures in patients admitted with CF exacerbations. Additionally, establishing criteria for obtaining blood culture in patients admitted for CF exacerbation may be useful. Cystic fibrosis (CF) is a devastating life-threatening disease with the majority of mortalities due to respiratory failure. P. aeruginosa (PA) are common gram-negative bacteria leading to chronic infection and inflammation on CF patient lung mucus. It has long been recognized that there is strong adaptive immune response in patients with CF characterized by polyclonal IgG responses as well as T-cell proliferative responses to P. aeruginosa. RNA-seq of normal human bronchial epithelial cells (NHBE) treated with T-cell signature cytokines show unique modules of genes that could be used to classify airway inflammation. To test these modules in CF epithelium we obtained samples from CF patients who underwent bronchial brushing of the 2nd or 3rd generation bronchus on the right for RNAseq. We found evidence of a strong Th1 and Th17 signature in the epithelium which we have also observed in subjects with severe steroid refractory asthma. However, in contrast to severe asthma or COPD we observed extremely high levels of αβ T-cell receptors, γδ T-cell receptors and both B-cell and plasma-cell genes including immunoglobulin genes. We believe this robust mucosal immune response is pathogen specific and this prevents bacteremia. This study is trying to develop PA-specific antibodies with high avidity for treating lung diseases.

Methods: B cells from CF patients receiving clinical bronchoscopies or from CF patients undergoing lung transplant were isolated and purified through FACS Aria III cell sorter. Specifically, cells from bronchoalveolar lavage (BAL) fluid, bronchial brushes and/or mononuclear cells from explant lung tissue were stained and gated on CD19+IgA-IgD-IgM-to obtain memory B cells. 10-100 cells per well were then cultured in 96-well plates on CD40L transfected feeder cells and expanded with IL-2 and IL-21. To characterize PA-specific IgG isotype, series of ELISAs were conducted on supernatants collected from expanded B cells.

Results: We have established a pipeline to clone human B cells from the lungs of CF patients by sorting 10-100 cells from BAL or lung tissue. These cells blast and form colonies and secrete human IgGs. By ELISA, up to 7% make anti-PA-specific antibodies. This study suggests that it is feasible to screen monoclonal antibodies for PA-specific IgG from CF patients and generate antibodies for therapeutic use. We have cloned the variable sequences from several anti-PA hits and are in the process of sequencing these to assess the degree of somatic hyper-mutation as well as cloning these for expression in 293 cells for subsequent assays including FACS. Background: Tobramycin exhibits concentration-dependent activity and peak (Cmax) over minimum inhibitory concentration (MIC) ratio of 8-10 best predicts efficacy. Toxicity has been shown to relate to trough levels and drug exposure. Once-daily (OD) regimen achieves maximal Cmax with similar efficacy and potentially less toxicity compared to multiple daily dosing (MDD) regimens. A 24-hour area-under-concentration (AUC24) of 75-125 mg*h/L further optimizes efficacy and minimizes toxicity. Post-antibiotic effect (PAE) with tobramycin typically lasts 2-6 hours and limited data suggests prolonging time of undetectable level (t<0.5 mg/L) may induce resistance in P. aeruginosa. Although CF guidelines recommend OD tobramycin for treatment of pulmonary exacerbations, there are no recommendations regarding optimal t<0.5 mg/L. Starting March 2013, a monitoring protocol targeting Cmax 20-40 mg/L and t<0.5 mg/L of 2-6 hours was implemented (Protocol group, PG). Prior to this time, dose adjustment was made from a single random level (pre-Protocol group, pPG). Our objectives are to assess the association between tobramycin exposure (AUC24) and toxicity as well as changes in P. aeruginosa resistance.

Methods: Patients under 21 years admitted for treatment of CF exacerbation between Jan 1, 2012 -Dec 31, 2017 were included. Patients with pregnancy, kidney diseases, and ascites were excluded. Demographics, clinical and microbiological data were collected. Nephrotoxicity defined per pediatric RIFLE (pRIFLE) criteria. Patients in pPG received empiric 10 mg/kg every 24 hours and dosage was adjusted with a single level 12-hour postdose. Starting March 2013, patients in PG received empiric 10 mg/ kg every 24 hours and 2-and 10-hour levels were drawn on day 3 for PK determination and dosage adjustment. Primary outcome was differences in tobramycin AUC24 between empiric regimen and after dosage adjustment in PG. Secondary outcomes evaluated differences in toxicity and changes in P. aeruginosa susceptibility patterns between pPG and PG.

Results: 28 patients in PG and 18 patients in pPG were included. Mean age, sex, weight, comorbid conditions and concurrent nephrotoxic agents use were similar in both groups. Of the 28 patients who received tobramycin adjustment in PG, 15 (53.6%) were male, mean age 13.3 (range 4 -19 years) and weight was 42.4 ±13 kg (mean ± SD). Dosage adjustment in PG resulted in significantly higher tobramycin AUC24 (157 vs. 107 mg*hr/L, p<0.007). Compared to patients in pPG, more patients in PG experienced nephrotoxicity (0 vs. 8 (28.6%), p=0.01). No differences in P. aeruginosa susceptibility patterns were observed between pPG and PG.

Conclusion: Current dosage adjustment to meet t<0.5 mg/L of 2-6-hours resulted in tobramycin overexposure and may be associated with higher risk for nephrotoxicity without affecting P. aeruginosa resistance patterns.

mean ± SD; *significant differences observed, p<0.01 , and especially methicillin-resistant SA (MRSA), are a major cause of morbidity in cystic fibrosis (CF). SA, alone or with Pseudomonas aeruginosa, are associated with acute CF exacerbations and occasionally cause bloodstream infections. Using a longitudinal collection of MRSA isolates sampled from sputum, blood, or bronchoalveolar lavage of three CF patients at the Yale Adult CF Center, we performed Illumina and Nanopore whole genome sequencing to query genotypic changes incurred by MRSA over time. Phylogeny was analyzed by mapping individual reads from each isolate against a curated ST typematched genome previously obtained from a non-CF source. Phenotypic characterization of the SA isolates included biofilm assays, transepithelial migration assays, analysis of gene transcription, multiplex ELISAs for cytokines, and flow cytometric analysis of cells from a murine infection model. Our findings revealed three distinct MRSA lineages, with few genomic changes observed relative to a panel of P. aeruginosa CF isolates. Among the genomic changes in one patient's sputum and blood isolates were mutations affecting enzymes (GdpP and DacA) involved in the metabolism of the multifunctional secondary messenger cyclic di-AMP. Isolates harboring these mutations exhibited increased biofilm production, modulation of antimicrobial susceptibility, and changes in the expression of genes involved in central carbon metabolism. These MRSA isolates differed minimally in their immunogenicity or in their ability to traverse lung epithelial cell monolayers. In contrast to P. aeruginosa, which acquire major phenotypic differences over the course of infection in CF, our studies indicate that MRSA can acquire predictable mutations in genes implicated in biofilm production and antibiotic resistance and adapt to the CF airways through epigenetic mechanisms.

Supported by NIH R35 HL135800. Early Pseudomonas aeruginosa (Pa) colonization has been shown to adversely affect long-term pulmonary disease and survival in patients with CF. Early antibiotic therapy has the potential to eradicate initial Pa infection and postpone chronic infection. Inhaled (INH) tobramycin allows for antibiotic delivery to the site of pulmonary infection while avoiding toxicity associated with higher systemic exposure following intravenous (IV) administration. There are limited data evaluating the efficacy and safety of INH tobramycin in patients with CF aged <1 year and particularly in those <6 months. The objective of this study was to evaluate the efficacy and safety of INH tobramycin in early eradication of Pa in infants with CF.

Methods: This IRB-approved retrospective chart review evaluated CF patients with early Pa infection isolated on a respiratory culture prior to 1 year of age. Patients who received antibiotics with anti-Pa activity in the preceding 6 months of index Pa culture or prior to completion of INH tobramycin were excluded. The primary outcome was the frequency of Pa eradication post-treatment. Secondary outcomes included sustained culture negativity at 12 and 18 months and safety assessments.

Results: A total of 32 patients met study inclusion criteria; 13 patients were excluded due to concomitant antibiotic use. Of 18 patients evaluated, 9 patients received INH tobramycin and enteral fluoroquinolone (FQ) and 9 patients received INH tobramycin alone. Mean age at first Pa isolate was 0.34 years and 13 (72%) patients were <6 months of age. The median (IQR) INH tobramycin dose was 80 mg (80,120) and 14 (78%) patients were treated for 28 days and 4 (22%) patients for 14 days. A total of 15 (83%) patients were free of Pa at end of treatment and 10 (56%) patients were eradicated of Pa at 6 months post-treatment. There was no difference in the proportion of patients free of Pa at 6 months (44 vs. 67%, P=0.64) and the time to subsequent Pa infection (log-rank, P=0.27) in patients who received INH tobramycin and FQ versus INH tobramycin alone. There was a nonstatistically significant greater proportion of patients aged ≥6 months with eradication of Pa at 6 months post-treatment compared to those <6 months of age (83 vs. 42%, P=0.15). Of those that did require Pa, median time to next infection was 178 days. Infant pulmonary testing was performed in 16 patients post-treatment. The mean FEV 0.5 was 106±17% in those who eradicated Pa compared to 111±11% in those who did not eradicate Pa. INH tobramycin was relatively well tolerated. Of 3 patients reporting cough associated with therapy, 1 patient required discontinuation of therapy. Of all subsequent Pa isolates, 1 strain was resistant to tobramycin.

Conclusion: Our data suggest that early INH tobramycin therapy is effective in eradicating Pa infection and is well-tolerated in infants. The addition of an enteral FQ to INH tobramycin appears to have similar effectiveness to INH tobramycin alone. This aggressive approach for early eradication and prevention of chronic Pa infection in infants with CF should be considered to prevent rapid progression of lung disease and preserve pulmonary function after infection. During chronic lung infections in people with cystic fibrosis (CF), Pseudomonas aeruginosa transitions from a nonmucoid to a mucoid phenotype characterized by the overproduction of alginate. Additionally, chronic isolates often do not express lipopolysaccharide (LPS) O antigen, a phenotype referred to as LPS-rough. The apparent correlation between alginate expression and loss of O antigen in chronic isolates suggests that alginate and O antigen are inversely regulated. When produced, O antigen is described as long or very long depending on the number of polysaccharide repeat units attached to the LPS molecule. Previous studies reported, and we have confirmed, that PDO300, a mucoid derivative of the nonmucoid laboratory strain PAO1, produces less very long O antigen compared to PAO1, but still expresses long O antigen. We hypothesized that this was due to decreased production of Wzz2, the protein responsible for regulating very long O antigen chain-lengths. To test this hypothesis we analyzed whole cell lysates of PDO300 by immunoblot and showed reduced production of Wzz2 in PDO300 compared to PAO1. This suggests a connection between Wzz2 regulation and alginate production. In order to determine if the down regulation of Wzz2 was clinically relevant we screened a series of nonmucoid and mucoid clinical CF isolates. Overall, the mucoid strains produced less Wzz2 compared to the nonmucoid strains. To further elucidate the relationship between alginate and O antigen production we isolated nonmucoid revertants of a mucoid CF isolate. These revertants had increased Wzz2 compared to the mucoid parent strain. Targeted sequencing of these strains revealed frameshift mutations in algT, the sigma factor that transcribes the alginate operon. Therefore, mutation of algT results in loss of alginate and increased Wzz2 production suggesting an AlgT-dependent pathway for the regulation of Wzz2. These experiments support a model in which a transition between the nonmucoid and mucoid phenotype is accompanied by the down regulation of very long O antigen, through decreased production of Wzz2, prior to the establishment of the LPS-rough chronic phenotype. Objectives: Viral respiratory tract infections are regarded an important cause of respiratory infections early in life. Our aim was to evaluate the presence of respiratory viruses during respiratory symptoms (RS) and asymptomatic episodes in infants with CF compared to healthy controls.

Methods: A total of 20 infants with CF and 45 sex -and age-matched healthy controls were followed for the first 18 months of life. We performed viral MLPA testing (a PCR-based analysis) for 14 different viruses on in total 691 nasopharyngeal samples, of which 508 were taken during prescheduled visits and 183 during RS. Questionnaires were obtained collecting information regarding demographics and environmental exposure. Multilevel logistic regression was used to find differences between groups, and between symptomatic and asymptomatic episodes accounting for repeated measurements and known risk factors for viral infection.

Results: At least one virus was detected in 542 of 691 nasopharyngeal samples (78.4%). Simultaneous detection of 2, 3, 4 and 5 viruses occurred in 24.3%, 7.4%, 1.4% and 0.3% of samples, respectively. We found a significant effect of siblings, daycare attendance, season of sampling and age on viral detection in both groups, while no significant association with breastfeeding was observed. Overall viral detection was significantly lower in children with CF compared to non-CF control children, even after correction for differences in daycare attendance, siblings, age, feeding type, RS and season (57.8% vs 86.4%; OR 0.39 (95% CI: 0.21 -0.73); q=0.006). Adeno-and rhinoviruses were significantly more prevalent in non-CF children, whereas corona-and bocavirus were similarly often detected in both groups. The association between the presence of respiratory viruses and simultaneous occurrence of respiratory symptoms was similar in children with and without CF. Metapneumo-, influenza-, parainfluenza-, rhino-and respiratory synctial virus were significantly associated with RS in both groups, whereas no significant associations were found for adeno-, corona-and bocavirus.

Conclusion: Overall viral detection is lower in infants with CF compared to healthy infants. However, the association between viral presence and simultaneous occurrence of respiratory symptoms is similar in both groups. Our findings might in part be explained by differences in exposure we could not correct for. However, interactions between viruses and microbiota composition may also play a role. Introduction: In patients with cystic fibrosis, several next generation sequencing based studies were established in the last decade to decipher cross-sectionally the structure and composition of the airways microbiota. However, longitudinal studies focusing on the airways microbiota in patients with cystic fibrosis need to be conducted to understand the role of the microbiome in the evolution of the disease. The aim of our study was to analyze the evolution of the microbiota among the airways (nose, throat and lung).

Methods: Using 16S amplicon sequencing targeting the 4th variable regions of the gene, we analyzed 87 nose swabs from 10 patients, 462 throat swabs from 43 patients, and 763 sputum samples from 60 patients for an average time period of 3 years. Samples were treated within the first 24 hours with PMA to ensure no amplification from extracellular DNA from dead cells and contaminant. Sequences were processed with the package DADA2 to ensure a good estimation of the error rate of each runs and trim the reads for good quality and no chimeras.

Results: We observed that each patient possessed a personalized microbiome as the Morisita-Horn distance between patients related microbiome was lower than between nonrelated microbiome for the three airways' compartment. Our results showed that the establishment of a chronic infection by P. aeruginosa was increasing the instability of the throat microbiome while correlating with a stabilization of the microbiome in the sputum indicating that the establishment of the infection in the lower airway correlates with a dysbiosis in the upper area. P. aeruginosa infection in the lower airway was also correlated to an increased decline of the alpha-diversity per year as well as Staphylococcus infection. Anaerobes (Prevotella and Veillonella) were correlated with a stabilization of the decline in the alpha-diversity per year. Finally, we observed that the dominance was correlated to an increase of the stability of the microbiome over time indicating a failure in the elimination process by the lung.

In conclusion, our results showed that the evolution of the microbiome in the lower airway is mostly dependent on the acquisition of a chronic infection by a CF pathogen which is correlated with a break in the dynamic colonization/elimination of the lower airways. On the other side, the maintenance of anaerobes in the lower airways was correlated to a more dynamic microbiome with a stable/increased alpha-diversity over the years.

Acknowledgments: We acknowledge the excellent work of Iris Kühbandner for patient recruitment and sample collection. This study was supported in part by the German Ministry for Education and Research (82DZL00401, 82DZL004A1). A similar model is lacking for patients with cystic fibrosis transmembrane regulator-related metabolic syndrome (CRMS). As an early adopter of state-wide newborn screening for CF, Wisconsin has a long-standing cohort of CRMS patients.

Methods: We performed a retrospective chart review of patients in our CF Patient Registry. We identified 64 unique patients with CRMS. We analyzed only children less than 16 years of age, due to low prevalence beyond that age. We queried the bacterial culture results from throat swab, sputum and bronchoscopy specimens for the eight pathogenic bacteria reported annually by the CFF. To understand the change in flora over the lifecyle of patients, we defined prevalence for each bacterium as at least one positive culture per year in a given patient.

Results: For the bacteria queried ( Figure) , the overall prevalence curves are visually similar to patients with cystic fibrosis, although with lower prevalence. Most patients (53%) grew methicillin-susceptible Staphylococcus aureus (MSSA) in infancy which rose and then declined to 52%, whereas methicillin-resistant Staphylococcus aureus (MRSA) remained low in early childhood before rising later to 10%. For Hemophilus influenzae the prevalence began at 13% in infancy then decreased to 5% and for Pseudomonas aeruginosa the prevalence began at 13% before declining and later rising back to 10%. In total, 91% of patients grew MSSA at least once, 8% grew MRSA at least once, 52% grew Hemophilus at least once, 27% grew Pseudomonas at least once and 11% grew Stenotrophomonas at least once.

Discussion: Patients with CRMS appear to grow a variety of pathogenic bacteria associated with clinical disease in patients with CF. The percentage of patients harboring these bacteria was surprisingly high, which indicates that the CFTR dysfunction in some CRMS patients may not be benign. Interestingly, the prevalence of some of these bacteria over time appears to mimic the long-standing model in CF, albeit with lower overall prevalence. As in CF, more than half our patients grew MSSA beginning in infancy and MRSA prevalence increased in late childhood while Hemophilus decreased at that time. Over one-quarter of patients grew Pseudomonas, although annual prevalence is lower than in CF. Limitations of this study include small sample size. Future directions include analysis of the role these bacteria play in the clinical outcomes in CRMS patients. Understanding the biodiversity and outcomes in these patients will help to define airway clearance therapy and treatment needs, for which little is currently known.

Stenotrophomonas and Achromobacter excluded due to low annual prevalence

Vejzovic, J.; Beagle, A.; Starner, T. University of Iowa Stead Family Children's Hospital, Iowa City, IA, USA Background: Tobramycin is a common intravenous aminoglycoside antibiotic used to treat cystic fibrosis (CF) pulmonary exacerbations due to its activity against Pseudomonas aeruginosa. Aminoglycosides are associated with nephrotoxicity, particularly at high doses or inadequate renal function, requiring therapeutic drug monitoring to confirm concentrations and clearance. In March 2016, a protocol was implemented to ensure that pediatric patients with CF admitted for a pulmonary exacerbation were adequately hydrated while receiving intravenous, extended-interval dosing tobramycin (EIDT). The purpose of this review was to evaluate the efficacy and safety of this protocol.

Methods: We performed a retrospective cohort analysis of pediatric patients admitted to the University of Iowa Hospitals and Clinics with a diagnosis of CF pulmonary exacerbation treated with EIDT. The pre-protocol cohort consisted of patients admitted from January 2014 through February 2016 and the post-protocol from March 2016 through October 2017. The primary outcome was the incidence of acute kidney injury (AKI). Secondary outcomes were the number of dose changes needed to achieve therapeutic tobramycin maximum concentrations (TCmax); clinical improvement based on ppFEV 1 ; incidence of therapeutic and nontherapeutic tobramycin levels; number of days of monitoring per length of stay; and pharmacokinetic (PK) parameters.

Results: There were 131 initial orders that met inclusion and exclusion criteria, 69 pre-protocol and 62 post-protocol. No difference was detected in the occurrence of AKI between the two groups at 17.9% and 24.2% (p= 0.381) respectively; however, in patients with AKI, the number of concurrent nephrotoxic agents was greater in the post-protocol group (6 versus 12). No difference was observed in number of dose changes following the initial dose (27.5% versus 32.3%, p = 0.348). There were more patients in the post-protocol group with subtherapeutic levels (44.9% versus 72.6%), less patients with therapeutic levels (40.6% versus 27.4%), and no patients with supratherapeutic levels (14.5% versus 0%). Volume of distribution (Vd) increased in the post-protocol group (0.44 L/kg ± 0.19 versus 0.51 L/kg ± 0.2). There was no difference in clinical improvement based on change in ppFEV 1 (12.5% versus 12.7%, p = 0.926). The postprotocol group had significantly less days of monitoring per week (3.2 ± 1.4 versus 2.6 ± 2.0, p< 0.01).

Conclusion: Following the implementation of the protocol there was no difference in the occurrence of AKI despite higher number of concurrent nephrotoxic agents suggesting that hydration reduces the risk of AKI. There was no difference in the number of dose changes made to achieve goal TCmax despite differences in number of therapeutic levels. Increased number of subtherapeutic levels could be explained by initial hydration and increased Vd. Clinical improvement was similar between groups which suggests a potential need for re-evaluation of goal TCmax at our institution. Lastly, there was one fewer day of monitoring per week in the post-protocol group which allowed for more pharmacist, nursing, and physician time to be dedicated to patient care. Introduction: Selective culture media and molecular diagnostic testing have led to the identification of additional opportunistic pathogens that may contribute to lung disease severity and progression in cystic fibrosis (CF). One such pathogen is Inquilinus limosus, an oxidase-positive nonlactose fermenter first isolated from the CF airway, and rarely observed elsewhere. Relatively little is known about the virulence and clinical significance of I. limosus infection in CF. The objective of our study was to describe the prevalence, clinical characteristics, and outcomes of I. limosus infection at our CF center and to compare clinical outcomes between I. limosus cases and control subjects chronically infected with Pseudomonas aeruginosa (PA).

Methods: We report a retrospective, case-control study of CF patients with at least one positive I. limosus respiratory culture at the Colorado CF Center between 2006-2016, compared 1:2 with age-matched CF controls with chronic PA infection. Percent predicted FEV1 (ppFEV1) and BMI percentile were modeled versus time from first culture (up to 5 years). The model included a random subject (nested within matched pair) intercept and slope. Annual rate of pulmonary exacerbations requiring IV antibiotics were compared between groups.

Results: Thirteen patients with at least one respiratory culture positive for I. limosus were identified at our CF center; estimated prevalence of 1.8%. The mean age at the time of first positive culture was 12 years; range 4-18 years. The majority of cases (69%) had multiple cultures positive for I. limosus, while 31% had single, transient infections. Susceptibility testing revealed the most active antimicrobials were meropenem, fluoroquinolones, and trimethoprim/sulfamethoxazole, with 95%, 38%, and 24% of isolates susceptible, respectively. ppFEV1 was not different between I. limosus cases versus PA controls at baseline (83.4% vs 86.1%, p=0.71). I. limosus cases had an average ppFEV1 decline of 3.1% per year over 5 years compared to a decline of 2.0% per year for controls (p = 0.45). There was no difference in the pulmonary exacerbation rate between the two groups (both 1.4/year, p=0.90). Finally, BMI percentile showed a trend toward a lower BMI in the I. limosus group compared to the PA controls at baseline (28.4% and 43.4%, p=0.12). I. limosus cases had an increase in BMI percentile of 7.6% over 5 years, compared to a decrease of 5.8% for PA controls (p =0.17)

Conclusions: In this relatively large case series of CF patients with I. limosus infections, initial I. limosus infection was only identified in pediatric patients, and initial infection more commonly led to chronic rather than transient infection. High levels of antibiotic resistance were noted in the I. limosus isolates. Finally, I. limosus infection was associated with clinical outcomes including decline in ppFEV1, yearly rate of pulmonary exacerbations, and BMI that were similar to those observed with chronic PA infection, a pathogen with known increased morbidity and mortality in CF. Methods: In 2017, three CF subjects from YACFP with documented MRSA sputum colonization were hospitalized with MRSA pneumonia. We performed a medical records review to extract demographic, CF-related, microbiologic, clinical complication and therapeutic data. Bacterial analysis of clinical isolates included: MRSA genetic profiling (Illumina and Nanopore long-read sequencing) and growth kinetics (optical density values in relation to time). Isolates were studied in an infection model where 16HBE and THP1 cell lines were infected with MRSA clinical isolates and monitored for cytokine production over 24 hours.

Results: Clinical and microbiologic data are presented in the Table. CRP was increased in all subjects at time of pneumonia diagnosis. All subjects demonstrated sustained decline in FEV1 following MRSA pneumonia. Genetic profiling of MRSA isolates showed unique sequence types, consistent with different growth kinetics for each isolate. Exposure to the MRSA isolates induced variable production of CXCL10, IL-1b, and IL-6 in 16HBE and THP1 cells that did not establish a common pattern of innate response to individual isolates.

Conclusion: Our findings suggest that the MRSA isolates from these subjects were not of clonal origin, and therefore less likely to be have been transmitted during an outbreak. However, a trend towards slower growth in some isolates suggested adaptations that may have caused increased virulence. Here, we present a novel approach to identify the genomic and phenotypic adaptations of MRSA in sputum isolates to determine the likelihood of an individual bacterial clone as the source of an outbreak of severe clinical complications. The primary outcome measure was best percentage predicted FEV 1 (ppFEV 1 ; GLI) for the previous 12 months (under 5 years of age excluded), with a secondary endpoint of requirement for IV antibiotics. Associations between these outcome variables and the infection stratified cohorts were assessed using linear and logistic regression models.

Results: 9,401 patients were included in the study (294 lung transplant recipients excluded). The median age was 19 years (IQR 8, 29) , with 53% males and 49% Phe508del homozygous (Table) . 4,158 patients (44%) isolated Pa and 1,462 (15%) isolated Asp. There were proportionally more females with co-infection (52%) than with Pa alone (46%, p=0.001). CFRD was not reported more frequently in groups with co-infection compared to Pa alone, but was more prevalent in Asp vs clear groups (22% vs 12%, p<0.001). After adjusting for age, sex, Phe508del homozygosity and CFRD diagnosis, patients with Asp vs no infection had a 6.2% lower ppFEV 1 (p<0.001). In patients with Pa infection, there was no additional impact of Asp co-infection on ppFEV 1 . However there was a significantly higher probability of co-infected patients having received IV antibiotics in the preceding 12 months (OR 1.30, p<0.001), which was also the case for the CPa vs CPaAsp groups (OR 1.23, p=0.02).

Conclusions: Co-infection with Pa and Asp is associated with additional use of IV antibiotics despite a lack of evidence that lung function is affected. Causation cannot be ascertained from this cross-sectional data but further longitudinal analyses will explore the impact of co-infection on disease progression. Records from all 2017 banked S. aureus isolates were reviewed to determine if the patient was diagnosed with CF. Isolates were phenotypically characterized by antimicrobial susceptibilities. The first unique isolate from each patient was selected. Pulse-field gel electrophoresis (PFGE) was performed on all isolates and both visual and computer assisted gel analysis was used to group isolates into pulsed field types and subtypes, using a similarity coefficient of 0.8 to define a type. Isolates were also grouped into recognized USA types using to CDC type strains. A small subset of isolates was then selected for whole genome sequencing (WGS).

In 2017 there were 173 CF patients with S. aureus, with 622 banked isolates. Representative isolates from each patient yielded 251 isolates, 97 were MRSA and 154 were MSSA. Antimicrobial resistance profiles are outlined in the Table below. There were 31 and 74 types from MRSA and MSSA, respectively. The largest genotypic cluster of MRSA was USA100 with 34 isolates. USA300 included 4 isolates. The most common subtype included 5 isolates of USA100 that were indistinguishable by PFGE. The largest genotypic cluster of MSSA included 24 isolates and 21 isolates clustered with USA200. There were 2 MRSA isolates and 4 MSSA isolates that did not digest with SmaI and are tetracycline resistant, both traits of livestock-associated S. aureus (ST398). Results from WGS were pending at time of abstract submission.

We found several clusters of genetically-related S. aureus that may represent patient to patient transmission. Additionally, several S. aureus isolates are phenotypically similar to livestock-associated S. aureus. Further evaluation for patient to patient transmission with WGS is ongoing. Background: To prevent the detrimental outcomes associated with chronic infection, antimicrobial treatment is used to eradicate initial P. aeruginosa infection. However, in 10 to 40% of cases, eradication therapy fails and the reasons for this are not entirely understood. We previously showed that staphylococcal protein A (SpA) binds to the exopolysaccharide Psl in P. aeruginosa isolates that fail eradication therapy but not in those successfully cleared, leading to bacterial aggregation within biofilms and tolerance to high concentrations of tobramycin. The goals of this study were to examine the differences in Psl between P. aeruginosa isolates that were successfully eradicated compared to those that persisted, despite antibiotic treatment, in the airways of children with CF.

Methods: We used an initial discovery dataset (the P. aeruginosa isolate collection from the SickKids Eradication Cohort) and confirmed our findings in a larger validation cohort (the P. aeruginosa isolate collection from the Early Pseudomonas Infection Control (EPIC) trial). We compared eradicated to persistent isolates using a definition of eradication based on the first culture obtained after the end of antibiotic treatment. We examined Psl binding using the fluorescently labelled monoclonal antibodies WapR001, WapR016 and Psl0096 visualized by confocal microscopy. We also measured bacterial attachment in the presence of S. aureus filtrates using crystal violet assays.

Results: Compared to eradicated isolates from the SickKids cohort, persistent isolates bound Psl antibody Psl0096 and WapR001 (but not WapR016) significantly more (p<0.01) when grown as biofilms and visualized using fluorescently labelled antibodies. Using crystal violet assay as a measure of bacterial attachment, eradicated isolates from the SickKids cohort had significantly reduced attachment (p<0.01) in the presence of S. aureus filtrates compared to when grown in media alone; this was not observed with the persistent isolates, as SpA bound Psl, permitting pili attachment.

Conclusions: P. aeruginosa isolates causing initial infection in children with CF that fail to be cleared after inhaled tobramycin therapy appear to have higher binding to Psl antibody, Psl0096 and WapR001, compared to successfully eradicated isolates in structurally intact biofilms. These findings using the SickKids cohort will be confirmed by repeating these experiments in a blinded fashion using the EPIC isolate collection. Experiments to determine the contribution of Psl to tobramycin resistance are also ongoing.

Zhang, S.; Shrestha, C.; Bah, A.; Kopp, B.T. Center for Microbial Pathogenesis, Nationwide Children's Hospital, Columbus, OH, USA Introduction: Cystic fibrosis (CF) transmembrane conductance regulator (CFTR) modulators provide CF patients access to life-prolonging therapeutics, but clinical results vary including persistent bacterial infections. Studies by our group and others demonstrated that host immunity plays an integral role in the inability of CF patients to clear bacterial infections. Defective macrophage autophagy, a host defense mechanism to remove misfolded proteins and intracellular pathogens, is an important example of compromised immunity in CF patients. We have shown that human CF macrophages serve as a replicative niche for bacteria by which they avoid host defenses due to dysfunctional autophagy. In CF, dysfunctional autophagy and increased endoplasmic reticulum (ER) stress cooperate to cause heightened inflammation and aggregation of nonfunctional CFTR. Therefore, we sought to determine how three clinically available CFTR modulators that target specific genotypes (ivacaftor, lumacaftor, and tezacaftor) change CF macrophage autophagy and ER stress.

Methods: Human peripheral blood monocyte-derived macrophages (hMDMs) were isolated from patients with CF on/off CFTR modulators and compared to non-CF controls. Macrophage function assays were performed including apoptosis, phagocytosis and bacterial killing, and autophagy/ ER stress pathways interrogated. MDMs were infected with CF clinical isolates.

Results: CFTR modulators increased CFTR expression in hMDMs, which was associated with improved function including phagocytosis, autophagy, killing of bacteria, and decreased ER stress. ER stress pathway interrogation revealed that untreated CF hMDMs had an atypical ER stress response characterized by sustained PERK activation, but low eIF2α activation. In contrast, the IRE1α/XBP(s)-1 ER stress pathway and apoptosis were high in untreated CF hMDMs. CFTR modulators partially restored autophagy through increased beclin-1 phosphorylation and LC3-II conversion and activated the eIF2α pathway while decreasing apoptosis and IRE1a/XBP(s)-1. Results were greatest for ivacaftor-sensitive CFTR mutations compared to F508del (tezacaftor or lumacaftor). RNA-Seq analysis of whole blood from CF patients pre-and post-lumacaftor/ ivacaftor demonstrated activation of eIF2α signaling in clinical responders (+BMI and FEV 1 change).

Discussion: We demonstrated that CFTR modulators differentially improve macrophage function and subsequent bacterial killing based on subject genotype via modification of autophagy/ER stress interactions. Targeted autophagy/ER stress interactions have significant therapeutics implications in CF.

Bonfield, T.L.; Ragavapuram, V.; Sopko, K.; Fletcher, D.

Advances in cystic fibrosis (CF) treatment have improved the quality and survival of patients. However, as CF patients are living longer, nontuberculous Mycobacterium (NTM) infections have increased. NTMs are difficult to treat and contribute to a significant decline in the pulmonary manifestation of CF lung disease. We hypothesize that ineffective CF immunity contributes to deficits in managing NTM infections. Since NTMs are slow growing pathogens, we developed unique culturing conditions and technology to monitor NTM infections in vitro. Our methodology provides the opportunity to monitor Mycobacterium avium, Mycobacterium intracellulare and Mycobacterium avium complex (M. avium + M. intracellular) growth kinetics, survival and metabolic activity. In vitro validation of the NTMs included a focus on macrophages, since they are the primary route of host defense against intracellular bacteria. Bone marrow derived macrophages (BMDMs) were obtained from F508del mice (Cftr -/-) and C57BL/6J controls (Cftr+ /+ ). Cftr -/-BMDMs secreted significantly more IL-6, TNFα, IL-1β and KC when compared to Cftr +/+ BMDMs when treated with either free M. avium or M. intracellulare (see Table) . In an effort to investigate the unique and complex CF immune response to NTMs relative to healthy lungs in vivo we have begun to develop murine models of chronic NTM infection. Agarose beads were embedded with either M. avium or M. intracellulare. We verified the ability of the pathogen-embedded agarose beads to develop colony forming units (CFUs) and induce an inflammatory response from the Cftr -/and Cftr +/+ BMDMs. BMDMs from both Cftr -/and Cftr +/+ mice had a greater response to the pathogen-embedded beads than the free pathogen alone or sterile bead controls (see Table) . Further, BMDMs obtained from Cftr -/mice demonstrated a significantly greater response to the bead preparations and free pathogens compared to the Cftr -/controls demonstrating the unique capacity of CF lung macrophages to have an altered response to NTM infections. We next utilized the validated M. avium and M. intracellulare embedded bead sets in our murine model of CF lung infection and inflammation, with results pending. Our unique technology and in vivo modeling system promises to provide a detailed window into the CF inflammatory response to NTM pathogens for therapeutic testing and mechanistic understanding. This work was supported by The Marcus Foundation and The Cystic Fibrosis Foundation. LasR is the regulator of a quorum-sensing system by which P. aeruginosa coordinates its expression of virulence factors, and is among the most-mutated genes in CF infections; in addition, lasR mutants are more prevalent in later-stage disease, and lasRcolonized patients display accelerated lung function decline (Feltner JB, et al. MBio. 2016 Oct 4; 7(5) ). Loss of lasR is postulated to provide a growth advantage relative to wild-type P. aeruginosa, however little is known about the role of lasR mutation in host cell interactions. Alveolar macrophages (AMΦ) are the principal immune cell in the healthy lung and rest at the interface between the host and respiratory pathogens. They are responsible for pathogen, particulate, and apoptotic cell clearance, in addition to orchestrating the immune response and recruiting other effector cells. Defects in MΦ function have been linked to CF pathogenesis (Barnaby R, et al. AJP Lung Cell Mol Physiol. 2018; 314:L432-8) . We therefore explored the interactions of wild-type and lasR -P. aeruginosa with primary alveolar and immortalized human MΦ.

Methods: We performed a gentamicin-protection phagocytosis assay with P. aeruginosa and THP-1 MΦ. We subsequently investigated the effects of P. aeruginosa-derived soluble factors on cytokine production and metabolism in primary human AMΦ isolated from bronchoalveolar lavage.

Results:lasR -PA14 and CF clinical isolates were less efficiently phagocytosed relative to wild-type. This was not reversed by double mutations in the regulators rhlR or anr, which have previously been shown to augment virulence factor expression in the absence of lasR (Hammond JH, et al. J Bacteriol. 2015; 197:2810-20; Mukherjee S, et al. PLoS Pathog. 2017 ;13 (7)).

Furthermore, we found that soluble products from lasRstrains induce more TNFα and IL-6 than wild-type P. aeruginosa in primary AMΦ. In contrast, wild-type but not lasR mutants inhibited mitochondrial respiration in AMΦ. All strains caused a similar induction in glycolytic rate, classically associated with inflammatory activation (Na YR, et al. J Immunol. 2016; 197:4101-09) .

Conclusions: We illustrate that lasR -P. aeruginosa are less efficiently phagocytosed, trigger more inflammatory cytokine production, and induce similar glycolytic bursts to wild-type strains without impacting mitochondrial function. These effects provide insight into both the high prevalence of lasR mutants in later stages of CF (decreased clearance) as well as their association with increased morbidity (increased inflammation). Given that modulation of quorum-sensing systems has become an attractive prospect in the design of novel CF therapeutics, understanding their role in host-pathogen interactions will be critical to their proper pharmacologic targeting. A key question is whether the intensity of anti-Pa therapy is associated with risk of emergence of drug-resistant respiratory pathogens.

Methods: Cohort study of participants in the Early Pseudomonas Infection Control (EPIC) trial of standardized therapy for newly acquired Pa, with follow-up through the EPIC Observational Study. Exposure was cumulative weeks of inhaled, oral and IV anti-Pa antibiotics, including both EPIC study drug and prescribed antibiotics, during the 18-month trial participation. Outcome was respiratory culture positivity for multidrug-resistant Pa (MDR Pa), methicillin-resistant S. aureus (MRSA), Burkholderia cepacia complex (BCC) or Aspergillus species during five years of followup. Association of antibiotic exposure with risk of pathogen acquisition was assessed using stratified Cox proportional hazards analysis with multiple failure times, adjusted for age, gender, CFTR mutation (F508del homozygous, heterozygous, other) and number of cultures + for pathogens during trial participation.

Results: Of the 249 trial participants, 27 (9%), 147 (59%), 21 (8.4%) and 123 (49%) had at least one + culture during follow-up for MDR Pa, MRSA, BCC and Aspergillus, respectively. There was a significant association between weeks of IV anti-Pa antibiotics and risk of MDR Pa during follow-up (HR 1.13, 95% CI 1.04-1.22, p=0.01). There were no other associations detected between anti-Pa antibiotic exposure and risk of acquisition of any pathogen.

In this cohort of CF patients with new Pa with rigorous prospective ascertainment of antibiotic exposure, there was no evidence of increased risk of acquisition of drug-resistant pathogens with cumulative exposure to inhaled or oral anti-Pa antibiotics. There was an association between number of weeks of IV anti-Pa antibiotics and risk of MDR-Pa, though the effect was small and clinical significance is unclear. These results suggest that standard outpatient management of new Pa with inhaled or oral antibiotics does not promote emergence of drug-resistant pathogens during five years of follow-up. (Pa) is associated with decreasing lung function and increased morbidity and mortality in patients with cystic fibrosis (CF). The persistence of Pa is predicated on its ability to make biofilms which are slimy communities of polymers and microbes that allow Pa to colonize the airway. Pf bacteriophage (Pf) is a filamentous, lysogenic bacteriophage which acts symbiotically when it infects Pa. Pf has been shown to contribute in a concentration dependent manner to adhesion, mucus viscosity, antibiotic tolerance, and inhibition of phagocytosis in Pa biofilms in vitro (Secor PR, et al. Cell Host Microbe. 2015; 18(5) :549-59).

Objective: To measure the amount of Pf in CF sputum and correlate to clinical outcomes.

Methods: Banked sputum samples from adult CF patients at the CF center at Stanford University were analyzed. To assess Pa and Pf load in sputum, DNA was extracted using mechanical homogenization followed by QIAGEN Mini Prep kit. Quantitative PCR was used with probe specific Pa rpIU gene and a probe specifically developed for Pf.

Results: Expectorated sputum was collected from 77 patients. 57 patients had Pa detected with 22 of those also having Pf detected (39%). No Pa or Pf was detected in 23 patients. Patients with Pf in their sputum were significantly older than those with Pa and no Pf (33.4 years vs 25.9 years, p=0.01). The average level of Pf detected was 4.4x 10 9 copies/ mL. Pa levels detected in Pf positive patients were significantly higher than in Pf negative patients (2.2 x 10 8 copies/mL vs 3.2 x 10 8 copies/mL, p=0.01). Pf levels were directly correlated with level of Pa (r 2 = 0.85). Pa isolates from sputum with Pf detected were mucoid in 86% versus 58% in Pf negative sputum (p=0.03). Leeds criteria for chronic Pa infection was met in 100% of patients with Pf in their sputum vs 65% of patients with Pa but no Pf (p=0.006).

Conclusion: Higher Pf load is associated with higher Pa load and Pf is associated with mucoid Pa and chronic infection indicating Pf may support the persistence of Pa. Older age in Pf patients may indicate that Pf is acquired over time. Pf may contribute to worsening pulmonary disease in patients with Pa infection. Pf is a potential target for treating chronic Pa infection in patients with CF. Swords, W.E. 2 1. MUSC, Charleston, SC, USA; 2. UAB, Birmingham, AL, USA; 3. UAB, Birmingham, AL, USA; 4. Dartmouth, Hanover, NH, USA; 5. Synspira Inc., Boston, MA, USA Introduction:Pseudomonas aeruginosa (Pa) forms biofilms in the cystic fibrosis (CF) airway, resulting in chronic infection and antibiotic resistance. Poly (acetyl, arginyl) glucosamine (PAAG) is a synthetic bioactive glycopolymer in clinical development as the agent SNSP113 and alters CF mucin conformation through electrostatic interactions. CF mucins and Pa biofilms share electrophysical properties. We hypothesized PAAG would alter Pa biofilm structure and enhance antibiotic efficacy.

Methods: Mature biofilms formed by Pa strain PA01 or a clinical isolate (Pa529) were cultured and exposed to PAAG (200 µg/mL) and/or the anti-Pa antibiotic tobramycin (10 µg/mL). Respiratory epithelial cells (CFBE41o-) transduced with the F508del CFTR mutation were grown as monolayers and were infected with GFP labeled Pa01. Confocal microscopy images were obtained after biofilms formed and were quantitatively assessed using COMSTAT image analysis software. Bacterial viability of Pa529 biofilms grown on glass confocal microscopy slides was assessed using LIVE/DEAD Baclight stain.

Results: Compared to Pa control and tobramycin alone, both PAAG and combination treatment resulted in significant decrease in maximal thickness ( Conclusion: PAAG has antibiofilm properties, disrupting structural integrity upon exposure to Pa biofilms. This effect enhanced tobramycin efficacy based on confocal imaging assessment and Baclight LIVE/DEAD staining of relatively resistant clinical isolate Pa529. PAAG represents a novel mechanism to combat biofilm associated antibiotic resistance in the CF airway and may improve Pa eradication in chronically colonized patients.

Pa529 biofilms after exposure to LIVE/DEAD stain demonstrate enhanced response using combination therapy. Reference bar is 20 µm. Introduction: Antibiotics are a cornerstone of CF lung disease treatment, yet the microbial determinants of response to antibiotics remain poorly understood. The microbiota in CF respiratory specimens are known to be relatively stable over time, even during antibiotic therapy in later stage disease. Much less is known about how the "metagenome," the predicted functional capacity of the microbiota, including community antibiotic resistance determinants, changes with antibiotic therapy. The commonly used antibiotic inhaled tobramycin is known to reduce sputum densities of Pseudomonas aeruginosa (Pa) and to improve lung disease measures on average in treatment-naïve patients. However, studies have failed to demonstrate a consistent relationship between changes in sputum Pa densities and lung disease outcomes. We hypothesized that metagenomic analysis would identify sputum microbial correlates of clinical response to inhaled tobramycin. Our goal is to understand differential clinical response to this treatment, in the hope of developing more effective antimicrobial therapies.

Methods: We collected sputum samples from 30 individuals with CF before, weekly during and after a standard one-month course of inhaled tobramycin; 157 samples in total. Lung function measurements and symptom scores were collected at study visits from all participants. Using classic culture methods, we determined abundances of traditional CF pathogens Pa and Staphylococcus aureus (Sa) for all samples. We extracted DNA from all samples using a method that focuses on live bacteria and minimizes human DNA. We defined the sputum microbiota and metagenome using shotgun metagenomic sequencing and quantitative PCR.

Results: Both quantitative cultures and quantitative PCR showed a reduction in absolute sputum abundances of both Pa and Sa one week after starting therapy, with bacterial abundances of each taxon recovering nearly to pre-treatment levels by the end of the therapeutic period. By contrast, metagenomic sequencing identified a large shift in microbiota constituency after one week of treatment that did not recover by the end of therapy and was largely characterized by changes in abundances of nondominant, anaerobic taxa. There was notable interpatient heterogeneity in microbial community response in all of the above mentioned microbial metrics. The relationships between these taxonomic changes, clinical measures, and functional metagenomic changes (including antibiotic resistance determinants) are currently being evaluated.

Conclusions: Weekly sampling and metagenomic sequencing indicate that the most substantial shifts in CF sputum microbial communities occur after one week of inhaled tobramycin therapy, with variable community recovery by the end of the treatment period. Standard clinical culture methods provided only a limited view of these changes. Further analysis comparing the sputum microbiota and metagenome to clinical outcomes may identify specific microbial features associated with differential response to tobramycin therapy. Acknowledgment: Supported by Novartis.

Introduction: Mycobacterium abscessus complex (MABSC) lung infections are increasing in prevalence in people with cystic fibrosis (CF). Near identical strains of MABSC circulating in the cystic fibrosis population including dominant global clones have been reported (Bryant JM, et al. Science. 2016; 354:751-7) . This study is investigating the prevalence of shared MABSC strains in the CF population of Queensland, Australia.

Methods: MABSC respiratory isolates stored at the Queensland Mycobacterium Reference Laboratory were included in the study. The collection comprised 179 MABSC isolates from 71 people with CF isolated from 2000 to 2016. All isolates underwent whole genome sequencing (WGS) using the Illumina HiSeq platform.

Results: Of the 71 people with CF (adults = 40; paediatrics = 31) included in the study, 38 participants had a single MABSC isolate sequenced and 33 participants had multiple MABSC isolates sequenced (average number of isolates 4; range 2-11). There was evidence of globally circulating clones in the Queensland CF cohort (especially dominant circulating clone 1 within M. abscessus subspecies abscessus strain) throughout the collection period (2000) (2001) (2002) (2003) (2004) (2005) (2006) (2007) (2008) (2009) (2010) (2011) (2012) (2013) (2014) (2015) (2016) . Further analysis of this dataset will determine if these globally circulating clones are detected in other patient cohorts (e.g. bronchiectasis) and in the healthcare and home environment, specifically potable water.

Conclusions: Global clones of MABSC previously reported (Bryant et al, 2016) were found in the Queensland CF population.

Acknowledgments: Supported by Cystic Fibrosis Foundation Therapeutics (USA), The Prince Charles Hospital Foundation, Advance Queensland, NHMRC Project Grant (APP1102494). Aim: This service evaluation aims to assess the impact of routine screening for RVI in adult CF patients, to determine whether there are potential clinical benefits from this additional testing.

Routine screening for seasonal respiratory viruses in specialist CF clinics occurred between Nov 2017 -Apr 2018. Diagnostic testing was performed using the polymerase chain reaction (PCR)-based Ausdiagnostics 16-well panel able to detect: influenza A, B typing, PIV (1/2/3/4), RSV (A/B), AdV (A-E), RV/EV, hMPV, CoV (229E, OC43, NL63, HKU1), including the BD Veritor point-of-care test able to detect: Flu A, B, RSV as triage.

Results: During Nov 2017-April 2018, 13/37 adult CF patients presenting with an exacerbation had a RVI. 11/37 presented with >1 exacerbation and 4/11 were diagnosed with a RVI during one of these encounters. Of the 13 virus-infected patients, 9 = noninfluenza virus (2 entero/rhinovirus, 5 coronavirus NL63, 1 coronavirus HKU1 and 1 human metapneumovirus), 3 = influenza infections alone (2 influenza A/H3N2, 1 influenza B), and 1 = combined influenza B, coronavirus OC43 and entero/rhinovirus infection. 24/37 patients had no detectable RVI, of whom 19 had been influenza-vaccinated in the preceding six months. Of the 5 unvaccinated patients, 1 = was unconcerned, 3 = vaccine was inaccessible and 1 = unwell post-vaccination in a previous season. Of the 4 patients who had influenza, 3 were unvaccinated due to a lack of concern and 1 was vaccinated earlier but still became infected with late-season influenza. Of patients (n=13) with a RVI, 6 = unemployed, 1 = worked from home and 6 = worked in settings with regular public interactions. Of patients uninfected with RVI (n=24), 17 = worked in occupations with regular public interactions, 5 = unemployed, 1 = college and 1 = worked from home.

Conclusion: Routine RVI screening allowed early influenza and noninfluenza virus detection in 11% and 24% of patients, respectively. RVI screening helped to provide a more definitive diagnosis for patients to better explain the cause of their presenting symptoms. Early RVI detection and management could reduce morbidity/mortality in CF and allow appropriate infection control though the cost-effectiveness of screening remains to be determined. The rise of antibiotic resistance has limited the therapeutic efficacy of conventional antibiotics against CF pathogens. As such, combination therapy and antibiotic cycling have been investigated for pathogen eradication, particularly for Pseudomonas aeruginosa which is a major contributor to the morbidity and mortality of CF patients. Prior work has shown that the concurrent use of antibiotics helps prevent the emergence and spread of resistance. Polymeric nitric oxide (NO) release is an attractive adjunct therapy to antibiotics because it eradicates bacteria through multiple mechanisms, is unlikely to foster resistance, and has previously been reported to act synergistically with silver sulfadiazine (Privett BJ, et al. Mol Pharm. 2010; 7:2289-96) . Nitric oxide-releasing chitosan (COS-EA/NO) both eradicates P. aeruginosa biofilms (Reighard KP, et al. Biofouling. 2015; 31:775-87) and decreases CF sputum viscoelasticity (Reighard KP, et al. ACS Biomater Sci Eng. 2017; 3:1017-26) . The objective of this work was to evaluate how combining NO with antibiotic treatment might improve antibacterial activity. Four different strains of planktonic P. aeruginosa, including three that were multidrug resistant, were exposed to one of six antibiotics (ie, aztreonam, ceftazidime, ciprofloxacin, colistin, meropenem, and tobramycin) with and in the absence of COS-EA/NO in single agent inhibitory, checkerboard, and combination time-kill assays to fully assess antibiotic-NO interactions. All planktonic assays were performed both aerobically and anaerobically to represent extremes of the oxygen gradient found in CF mucus. Biofilms of all four strains were grown aerobically for 2 days in media and subjected to checkerboard assays with COS-EA/NO and a test agent (ie, tobramycin). Finally, biofilms were pretreated with COS-EA/NO at suberadication concentrations before exposure to tobramycin in order to assess the impact of low-dose NO-releasing scaffolds on biofilm sensitivity to antibiotic treatment. Planktonic P. aeruginosa was serially passaged with subinhibitory doses of COS-EA/NO to determine the likelihood of acquired resistance. Most interactions between COS-EA/NO and the antibiotics were found to be additive or indifferent in the planktonic studies, and no antagonism was observed. Synergy was observed in one P. aeruginosa strain between COS-EA/NO and colistin using checkerboard and time-kill assays, and between COS-EA/NO and tobramycin in time-kill assays. Pretreatment of biofilms with COS-EA/NO at 0.2X MBEC significantly increased biofilm sensitivity to tobramycin (~50-90% reduction in MBEC). Continuous exposure to subinhibitory doses of COS/EA-NO did not lead to an increase in MIC after 60 passages. These preliminary data suggest that COS-EA/NO is a viable adjunct therapy for P. aeruginosa eradication in CF patients because it promotes synergistic interactions and is capable of reducing therapeutic doses of antibiotics for biofilm eradication. Additionally, the likelihood of acquired resistance is low, even after continuous exposure to subinhibitory doses of COS/EA-NO. Experiments are currently underway to evaluate cytotoxicity against a primary cell line of human bronchial epithelial cells. Cystic fibrosis (CF) patients have impaired mucociliary clearance and thick mucus, which promote respiratory infections. Pseudomonas aeruginosa and Staphylococcus aureus are two of the most prevalent respiratory pathogens in CF infections. Both organisms are associated with poor lung function and patient outcomes. In addition, P. aeruginosa and S. aureus form biofilms in the airway, which are highly recalcitrant to antibiotic treatment. In this study, we found that P. aeruginosa alters the antibiotic sensitivity profile of S. aureus biofilms. To study polymicrobial interactions, our group has established an in vitro co-culture model on CF-derived bronchial epithelial (CFBE) cells. The CFBE cells are homozygous for ΔF508-CFTR. We used this system along with analogous experiments on plastic dishes to study how interspecies interactions change upon exposure to frontline antibiotics used to treat CF patients. Using this model, we observed that P. aeruginosa supernatant can lead to increased tolerance of S. aureus biofilms to multiple classes of antibiotics, including protein synthesis inhibitors and cell wall-active antibiotics. The small molecule 2-heptyl-4-hydroxyquinoline N-oxide (HQNO) and the siderophores pyoverdine and pyochelin contribute to the ability of P. aeruginosa to protect S. aureus from vancomycin. We propose a model whereby P. aeruginosa forces S. aureus to shift to fermentative growth, leading to a reduced growth rate and decreased susceptibility to cell wall-targeting antibiotics. Additionally, we found that P. aeruginosa supernatant increased the sensitivity of S. aureus biofilms to other classes of compounds, including fluoroquinolones and antiseptics. Treatment of S. aureus with chloroxylenol alone did not decrease biofilm cell viability; however, the combination of chloroxylenol and P. aeruginosa supernatant led to a 4-log reduction in S. aureus biofilm viability compared to exposure to chloroxylenol alone. We found that HQNO, pyoverdine, and pyochelin were responsible for the observed enhanced sensitivity to chloroxylenol. Furthermore, exogenous HQNO can sensitize S. aureus biofilms to chloroxylenol in a dose-dependent manner. Overall, our works shows that interspecies interactions can have dramatic and unexpected impacts on antibiotic efficacy in the context of a polymicrobial disease. (CF) is an autosomal recessive disease associated with significant morbidity and early mortality due to recurrent acute and chronic lung infections (MacKenzie T, et al. Ann Intern Med. 2014; 161:233-41) . The chronic use of multiple antibiotics increases the possibility of multidrug resistance (MDR) and severely limits the options of antibiotic coverage in patients with advanced disease (Sherrard LJ, et al. Lancet. 2014; 384:703-13) . Antibiotic susceptibility determined by culturebased standard techniques may not fully represent antibiotic resistance of all bacteria within the airway. The objective of this study is a) to detect additional antibiotic resistance using molecular methods, and b) to relate the presence of MDR to airway microbiome diversity and pulmonary function.

Methods: This study cohort is from a larger prospective, longitudinal study of CF persons < 21 years of age at CNHS. Spontaneously expectorated sputum samples were obtained at four time points: baseline (B), exacerbation (E), treatment (T), and recovery (R) then homogenized using Sputasol (Fisher Healthcare), vortexed (1 min), centrifuged (12,000 g x10 min) and pellets frozen at -80 ○ C. Bacterial DNA was extracted using QIAsymphony SP (Qiagen) following pretreatment with lysozyme (20 mg/mL) and lysostaphin (200 µg/mL) (Sigma-Aldrich). Extracted DNA was amplified for the V4 region of the 16S rRNA gene, and sequences were processed in mothur (version 1.39.5) (Schloss PD, et al. Appl Environ Microbiol. 2009; 75:7537-41) . The Antibiotic Resistance Genes Microbial DNA qPCR Array (Qiagen) was used to detect antibiotic resistance genes. Clinical culture results and pulmonary function were also noted for each encounter.

Results: Six study participants had 16S sequencing and PCR testing for 19 encounters. When evaluating the presence/absence of MDR bacteria, those samples having resistance detected (n=7) had significantly lower diversity measured by inverse Simpson's index than those without (n=12) (2.193±0 .427 vs 6.023±1.564, p=0.035). Differential abundance of OTUs was measured and samples with MDR bacteria present were more likely to have Streptococcus (p=0.002) and Alcaligenaceae_unclassified (p=0.002). Pulmonary function was decreased when MDR bacteria were present in the corresponding respiratory sample (% predicted FEV 1 , 51±22.9 vs 77±26.7, p=0.054; % predicted FVC, 64.5±22.7 vs 91.6±27.7, p=0.047).

Conclusions: The presence of MDR within the CF airway microbiome was associated with decreased microbial diversity, the presence of Alcaligenes, and decreased pulmonary function. Future studies should incorporate shotgun sequencing with screening against antimicrobial resistance databases to more fully define the mechanisms of resistance.

Low, D. 2 ; Wilson, D. 1 ; Flume, P. 1 1. MUSC, Denver, CO, USA; 2. University of Colorado, Denver, Denver, CO, USA Introduction: Nontuberculous mycobacteria (NTM) are increasingly recognized as a cause of morbidity and mortality in patients with cystic fibrosis (CF). NTM lung disease is defined as compatible symptoms and radiologic findings when NTM are present in respiratory cultures. The decision to start treatment for NTM in CF is confounded by the presence of nonmycobactrial bacterial pathogens. American Thoracic Society (ATS) guidelines recommend optimal treatment of nonmycobacterial pathogens prior to initiating treatment of NTM to facilitate assessment of clinical response to antimycobacterial treatment. In order to understand best practices for the management of NTM lung disease in the CF patient, it is important to understand what is currently driving treatment decisions and monitoring outcomes. For this study we looked at factors associated with initiation of NTM treatment.

Methods: The Cystic Fibrosis Foundation Patient Registry (CFFPR) has collected data regarding NTM since 2010. Data from the CFFPR included microbiologic data (NTM, bacteria), lung function, and pulmonary exacerbations (as reported by centers during a clinic visit or hospitalization). Treatment of NTM was defined as indication of NTM treatment initiation in the annual data set. Descriptive statistics, t-test for continuous variables and chi-square tests for proportions, were used to compare group characteristics.

Results: Between 2010-2014, of the 19,811 individuals aged 10 and older followed in the CFFPR, 15,682 were tested at least once for NTM. 3,230 patients (16.3%) had at least one positive culture for NTM. Of those, 572 (17.7%) had treatment initiated during the study period. Race and gender did not differ significantly between the treated and nontreated groups. Lung function (FEV 1 percent predicted) differed slightly between treated and nontreated groups (69% vs. 67%, p=0.05). The average number of pulmonary exacerbations per patient, defined as any documented exacerbation in year prior to encounter did not differ between groups. Pseudomonas aeruginosa PA [FPA1] infection was lower in the treatment group compared to the nontreated group (80.2% vs. 87.4%, p=<0.001). MRSA infection was also a lower percentage in the treated group vs the nontreated group (47.7% vs. 52.6%, p=0.036).

Discussion: Initiation of treatment for NTM remains a debated topic without clear guidelines. Evaluation of the CFFPR demonstrates that those who have treatment initiated for NTM infection often have lower rate of PA and MRSA infection, which may be indicative that a lack of nonmycobacterial pathogens influences the decision to treat NTM. Surprisingly, there is not a greater rate of exacerbations that might be predicted to influence treatment. We are not finding clear patient-related factors that drive the decision to treat. Further analysis will investigate if there are center-related factors. Trough based therapeutic drug monitoring (TDM) of tobramycin is most useful for the prevention of nephrotoxicity. However, trough levels may fail to capture supratherapeutic peaks or area under the curve (AUC) resulting in higher than necessary drug exposure and potential adverse events. AUC based TDM offers a more accurate estimation of drug exposure. In late 2016, our institutional practice changed from single level to AUC based tobramycin TDM in CF.

Methods: A retrospective chart review of CF admissions for APE between August 2015 and August 2017 was evaluated for tobramycin dose based on TDM strategy. Prior practice targeted a trough ≤ 0.3 mg/L in pediatric patients and a 12-hour level of 1-3 mg/L in adult patients. Current practice utilizes 2-and 8-hour levels with an AUC target of 80-120 mg*hr/L. Patients < 12 years, > 40 years, with a baseline FEV1% < 40, or history of lung transplant were excluded. Characterization of pediatric versus adult patients was determined by admission location. The primary endpoint was average daily dose of tobramycin with AUC versus single level TDM. Secondary endpoints included length of admission, time to next admission, the portion of patients with FEV1% recovery, and change in renal function.

Results: Among 131 patients admitted for APE in the study period, paired data were available for 25 pediatric and 9 adult patients. Baseline characteristics were similar between paired admissions for pediatric patients with a median age of 17, BMI of 21, and baseline FEV1% of 84. Paired adult baseline characteristics were also similar with a median adult age of 26, BMI of 21, and baseline FEV1% of 65. Median tobramycin dose at discharge was 9.5 mg/kg with trough level compared to 7.9 mg/kg with AUC monitoring in paired pediatric data (p = 0.0007). In paired adult data, median tobramycin dose was 8.0 mg/kg with 12-hour level compared to 6.0 mg/kg with AUC monitoring (p = 0.00043). There was no difference in length of admission or portion of patients with FEV1% recovery. Median time to next admission was 192 days (IQR 69-313 days) with trough compared to 80 days (IQR 44-178 days) with AUC monitoring in unpaired pediatric patients (p = 0.0249). Time to next admission was not significant for adult or paired pediatric data. Among unpaired adult patients, serum creatinine increased 0.04 mg/dL with 12-hour monitoring and decreased 0.02 mg/dL with AUC monitoring which was statistically but not clinically significant (p = 0.0005).

Conclusions: Tobramycin doses were significantly lower when utilizing AUC versus single-level TDM in CF patients. Lower tobramycin doses did not appear to be less efficacious in terms of duration of admission and FEV1% recovery. However, tobramycin doses with AUC TDM were correlated to a shortened time to next admission in unpaired pediatric patients which is partially confounded by seasonal trends in CF admissions and poor air quality related to Oregon wildfires in late 2017. Lower tobramycin doses did not result in clinically significant changes in renal function. Further study is needed to validate these findings. Introduction: Biofilm-forming pathogens, the major cause of morbidity in CF, are difficult to treat due to additional protection provided by both the biofilm exopolysaccharide matrix and mucus in the airways. Nitric oxide (NO), an endogenously produced free radical involved in the immune response, holds promise as a CF therapeutic due to broad spectrum antibacterial activity (Carpenter AW, et al. Chem Soc Rev. 2012; 41:3742-52) . Inhalation of NO-releasing biopolymers represents an attractive therapeutic strategy as NO release can be sustained for hours, allowing for near continuous NO delivery. Alginate is a particularly promising NO donor scaffold due to its low toxicity, water solubility, and biocompatibility (Draget KI, et al. Food Hydrocoll. 2011; 25:251-6) . Herein, we report the antibiofilm activity of NO-releasing alginates against CF pathogens.

Methods: Alginate oligosaccharides (~5 kDa, Alg5) were prepared via oxidative degradation of 300 kDa alginates. The carboxylates on Alg5 were modified with alkyl amines, (ie, diethylenetriamine, DETA; spermine, SPER; dipropyltriamine, DPTA; and N-propyl-1,3-propanediamine, PAPA) using carbodiimide chemistry (Ahonen MJR, et al. Biomacromolecules. 2018; 19:1189-97) . Amine-modified alginates were converted to N-diazeniumdiolate NO donors upon reaction with NO at high pressure under basic conditions. Nitric oxide release was monitored in artificial sputum media (ASM, pH 6.5, 37°C) in real time using a chemiluminescence NO analyzer (NOA; Boulder, CO) to simulate release in CF airways. To evaluate the bactericidal activity of the NO-releasing alginates, P. aeruginosa, Burkholderia cepacia, Staphylococcus aureus, and methicillin-resistant S. aureus biofilms were grown in ASM under aerobic or anaerobic conditions for 3 days at 37°C. The biofilms were then treated with control alginates, NO-releasing alginates, or tobramycin. Bacterial viability in the biofilms was assessed using the plate-counting method.

Results: The antibiofilm activity of NO-releasing alginates was evaluated against biofilms of four CF-relevant pathogens. Varying the amine group grafted on alginate imparted tunable NO storage (~0.1-0.3 µmol/mg) and release kinetics (t 1/2 ~0.1-0.4 h). The minimum concentration required to achieve a 5-log reduction in viability after 24 hours was evaluated for all materials (MBEC 24h ). The NO dose delivered was derived from both the MBEC 24h of the alginate samples and the measured NO totals from the NOA. At equivalent concentrations to the MBEC 24h , control alginates did not alter the bacterial viability of the biofilms, whereas the NO-releasing materials were highly antibacterial, implicating NO as the bactericidal agent. Aerobic biofilms required greater NO doses to achieve killing compared to anaerobic biofilms. Alg5-PAPA-DPTA/NO proved to be most effective (NO dose ≤140 µg/mL), regardless of growth conditions, even surpassing tobramycin (>2000 µg/mL).

Conclusions: The antibiofilm action of the NO-releasing alginates was clearly demonstrated against all test pathogens, with enhanced efficacy over current antibiotics. Experiments are underway to evaluate the effect of NO-release kinetics using time-kill assays in ASM.

Acknowledgments: This work was funded by NIH (AI112029) and KnowBIO, LLC. Objectives: To identify fungal taxa and explore the CF lung mycobiome in adolescent and adults with CF utilizing internal transcribed spacer (ITS) sequencing.

Methods: We conducted a prospective cohort study of individuals with CF at the University of Pennsylvania and Children's Hospital of Philadelphia CF Center for sputum analysis with selective fungal culture and ITS sequencing. Exclusion criteria included age younger than 14 years and history of solid organ transplantation. For ITS sequencing, genomic DNA from liquefied sputum was extracted, ITS genes amplified using bar-coded ITS1/ITS2 primers, and ITS amplicons were subjected to Illumina sequencing (MiSeq platform). Sequence data was analyzed using the QIIME2 bioinformatics pipeline. Fungal taxonomic assignments were generated by BROCC. Data for identified taxa were outputted as relative abundances and adjusted by quantitative PCR (qPCR).

Results: Among 77 sputum samples that underwent DNA extraction and amplification, 66 samples from individual CF patients were suitable for analysis after bioinformatic quality control procedures. Clinical characteristics of this subcohort (n=66) included: median age 29 years [23, 41] , 51.5% female, 92.4% pancreatic insufficient, median FEV1 percent predicted 57. 5 [40, 72] , and 69.7% Pseudomonas aeruginosa. Figure 1 depicts the taxonomic assignment and corresponding square-root transformed qPCR-corrected abundance. Candida species (spp), Saccharomyces cerevisiae, Malassezia spp, and Trichosporon spp were most commonly seen. Low level reads of Aspergillus spp and Scedosporium spp were found in CF sputum samples.

Conclusion: Fungal taxa were identified in CF sputum by ITS sequencing; Candida spp, Malassezia spp, and Saccharomyces cerevisiae were the predominant taxa. Trichosporon spp, which has been described as potentially pathogenic in CF, was commonly identified in our cohort by ITS sequencing. Further study of the clinical significance of the presence and recovery of these fungi is needed. Introduction: During their course of infection, bacterial pathogens are exposed to a variety of host factors that may influence their infection process and susceptibility to antimicrobial agents. While interactions between the host microenvironment and the pathogen are known to play a role in the establishment and persistence of an infection, there is limited knowledge on how the host tissue tunes the efficacy of antimicrobial agents. Gaining insights into the host factors that influence the efficacy of antibiotics may help improve our understanding of why there is such a poor correlation between antibiotic efficacy in vitro and in vivo in the cystic fibrosis patient population. This is particularly relevant for the major respiratory pathogen in patients with CF, Pseudomonas aeruginosa.

Study Goals: We previously demonstrated that culturing biofilms of the opportunistic pathogen P. aeruginosa on in vivo-like three-dimensional (3-D) lung epithelial cells (A549 cell line) enhanced the efficacy of aminoglycosides compared to when these same biofilms were cultured on a plastic surface. In the present study, we investigated (i) whether endogenous compounds produced by lung epithelial cells modulate the efficacy of aminoglycosides, (ii) which host factors are responsible, and (iii) what the underlying mode of action is.

Methods: We assessed antibiotic efficacy against P. aeruginosa in the presence of conditioned medium of 3-D lung epithelial cells . Antibiotic efficacy in the presence of 3-D CM or control medium was tested using a biofilm inhibition assay and time-kill curve. Tobramycin uptake was measured using BODIPY-labelled tobramycin in combination with flow cytometry analysis. The intracellular pH was determined using a 2, 7 -bis(2-carboxyethyl)-5(6)-carboxyfluorescein acetoxymethyl (BCECF-AM) assay. A series of chemical inhibitors of metabolic pathways were used to assess which compounds produced by the host were responsible for increased tobramycin efficacy.

Results: The 3-D CM potentiated bactericidal activity of aminoglycosides, including tobramycin. Mechanistic studies indicated that 3-D CM increased the intracellular pH of P. aeruginosa (hereby increasing the ΔpH component of the proton motive force, PMF), resulting in an enhanced tobramycin uptake in a fraction of the bacterial population. Our data suggest that metabolites of the host stimulate bacterial metabolism, hereby increasing the PMF and aminoglycoside uptake. Furthermore, preliminary experiments suggest that this synergistic action between host epithelial cells and antibiotics might not be exerted by cystic fibrosis epithelial cell lines.

Conclusions: We demonstrate that lung epithelial cells contribute to the defence against bacterial pathogens by acting in concert with antibiotics. Introduction: Tertiary lymphoid structures (TLS) are found in the lungs of patients with cystic fibrosis (CF) while they are absent in normal lungs. Like secondary lymphoid organs, TLS harbor segregated B and T cells areas, high endothelial venules, follicular dendritic cells and germinal centers. Persistent bronchopulmonary infection with Staphylococcus aureus (SA), one of the major bacteria infecting CF airways, induces the development of peribronchial TLS in mice in 14 days. Proposed roles of TLS include immune response against infection and autoimmunity. The goals of the present study were to determine the effects of lymphocyte depletion on TLS development (lymphoid neogenesis) and bacterial infection.

Methods: Lymphocyte depletion was obtained by pretreating C57Bl/6 mice with (1) an anti-CD20 monoclonal antibody (mAb) (B-cell depletion) or (2) an anti-CD4 and an anti-CD8 mAbs (CD4/CD8 T-cell depletion) or (3) a combination of anti-CD20, anti-CD4 and anti-CD8 mAbs (combined B-and CD4/CD8 T-cell depletion). Control groups received isotype control mAbs. Lymphocyte depletion in lung tissue was evaluated by flow cytometry. After lymphocyte depletion, mice were infected by intratracheal instillation of agarose beads containing SA (10 5 CFU/mouse). Fourteen days later, lung inflammatory cell infiltration, bacterial load and mouse survival were assessed by immunohistochemistry and cultures of lung homogenates, respectively.

Results: Analysis of mouse lung samples showed that depletion protocols induced a complete depletion of B-cells, CD4/CD8 T-cells or both B-and CD4/CD8 T-cells that persisted for at least 21 days. While TLS were observed in the lungs of persistently SA-infected mice pretreated with control mAbs, these structures were disrupted in the lungs of mice pretreated with anti-CD20 and/or anti-CD4/CD8 mAbs. Lung bacterial load and mouse survival (100% survival) were not different between lymphocyte-depleted groups and controls.

Conclusion: Disruption of SA-induced lymphoid follicles by anti-CD20 and/or anti-CD4/CD8 mAbs before inducing persistent SA-infected mice has no effect on bacterial load and on survival. These results suggest that although B cells and CD4/CD8 T cells are important for lymphoid neogenesis, lymphoid follicles do not play critical roles in infection control. Background: Lung infections with the multidrug-resistant Mycobacterium abscessus complex (MABSC) are difficult to treat in individuals with cystic fibrosis (CF), due to natural resistance towards most antibiotics. MABSC biofilms were previously observed in CF lung sputum (Qvist T, et al. Eur Respir J. 2015; 46:1823-6) where oxygen (O 2 ) consumption caused by polymorphonuclear leukocyte activity creates anaerobic conditions. However, other CF biofilm pathogens have recently been shown to be tolerant to antibiotic treatment due to dormant subpopulations in the anoxic parts of biofilms. We found strong evidence for aggregation and slow metabolism in MABSC isolates from patients with CF. Stopping this shielding process early may be the best shot at achieving bacterial eradication.

Objective: We sought to determine the physiology of biofilm aggregates of MABSC CF isolates and characterize the tolerance mechanisms of recommended antibiotics.

Methods: We carried out susceptibility testing on 33 isolates from 22 patients with CF and MABSC infection and the sequenced reference strain ATCC 19977. Isolates and strain were grown in Müeller Hinton (MH) broth with and without the disaggregating detergent Tween®80 (5%). Aggregated and separated cells were assessed with regard to tolerance pattern to anti-mycobacterial drugs. Time-kill curves generated for oxic and anoxic amikacin treatment in four-fold dilutions from 2 to 512 mg L -1 were determined after 1 and 3 days. In addition, confocal laser scanning microscopy and microrespirometry were used to disentangle biofilm behavior.

Results: Separation of MABCS isolates in 5% MH lead to increased susceptibility to amikacin (Am), tigecycline, kanamycin (Ka), azithromycin (Az), imipenem, cefoxitin and clarithromycin (Cl) (P < 0.05, n = 29-31). Also the frequency of tolerant MABSC isolates revealed significantly decreased tolerance as addition of 5% MH increased both disaggregation and metabolic activity (Am P=0.0059; Ka P=0.0001; Az P= 0.0057 and Cl P=0.0025, chi-square test). We found significantly lower O 2 consumption in MABSC aggregates compared to separated cells indicating slow-growing or dormant bacterial populations (5% MH: P= 0.047, one-way ANOVA). In addition, the bactericidal activity of amikacin in separated MABSC isolates during oxic conditions was enhanced to a maximum bacterial killing of 1-6 log using 2-512 mg L -1 of amikacin.

Conclusion: This study helps explain why drug susceptibility testing is so poorly correlated to treatment outcomes, as planktonic in vitro isolates do not imitate the actual conditions of the CF lung. New therapeutic targets for chronic MABSC infection could therefore be biofilm disruption through disaggregation and increasing O 2 availability during antibiotic therapy. (O 2 ) depletion. Antibiotics can perturb metabolism and respiration in bacteria, which will lead to increased formation of lethal reactive oxygen species (ROS) that contribute to the bactericidal effect. When aerobic respiration is attenuated, tobramycin is less effective (Zemke AC, et al. Antimicrob Agents Chemother. 2015; 59:3329-34) . Thus, re-oxygenation by hyperbaric oxygen treatment (HBOT) may sensitize P. aeruginosa biofilms to tobramycin.

Objective: To investigate the combined effect of HBOT and tobramycin on isolates of P. aeruginosa from intermittently colonized and chronically infected CF patients.

Methods: We tested re-oxygenation by HBOT of an O 2 -depleted biofilm model with the PAO1 strain and 14 mucoid/nonmucoid isolates consecutively isolated from 3 CF patients. Bacteria were embedded in agarose during tobramycin treatment to mimic in vivo conditions. 3-day-old anoxic biofilms were treated with 0-1024 µg/mL tobramycin in two-fold concentrations. The biofilms were further incubated for 90 min ± HBOT (100 % O 2 , 2.8 bar).

Results: PAO1 and mucoid/nonmucoid isolates of P. aeruginosa biofilms receiving HBOT had significantly higher killing rates at 128-1024 µg/mL tobramycin (p<0.001) compared to biofilms receiving tobramycin under anoxic conditions. In fact, these concentrations of tobramycin are actually obtained during inhalation therapy. The minimum biofilm eradication concentration (MBEC) is significantly lowered when the biofilms receive HBOT (P= 0.038, n = 84). Furthermore, the growth rate, which is associated with high susceptibility to antibiotics, was significantly increased by HBOT in 14 of the 15 strains (p<0.0006). The results show that HBOT boosts the efficacy of tobramycin against the biofilm-growing bacteria and may be involved by the formation of ROS that contribute to the bactericidal effect of tobramycin.

Conclusion: This study demonstrates the potential of HBOT as an adjuvant to improve antibiotic activity in chronic P. aeruginosa lung infections in CF patients by enhancing the supply of O 2 for bacterial aerobic respiration. The efficacy of tobramycin is enhanced on mucoid as well as nonmucoid isolates when applying HBOT. Furthermore, the MBEC is significantly decreased when HBOT is applied. This holds potential for reducing the concentration of nebulized tobramycin when combining with HBOT for treating P. aeruginosa lung infections in CF patients. 2 1. Paediatrics, University Hospital Limerick, Limerick, Ireland; 2. Royal College of Surgeons, Dublin, Ireland; 3. Trinity College, Dublin, Ireland; 4. Teagasc Food Research Centre, Cork, Ireland Introduction: Cross-sectional studies in individuals with CF suggest that the lung microbiome is altered early in life, but with significant variability between individuals (1, 2) . In childhood, microbiome diversity appears to be particularly dynamic (2) . Our objective was to design a longitudinal study examining how microbiome diversity in the lower airway changes over time in young children with CF.

Methods: SHIELD CF is a longitudinal study based around the annual bronchoalveolar lavage (BAL) surveillance programme for children aged 1 to 6 years in 3 paediatric CF centres in Ireland (1). BAL is performed under general anaesthesia via laryngeal mask airway and frozen at -80°C for future extraction of DNA. Extracted DNA was normalized and the bacterial 16S rRNA V3 and V4 region was amplified. Illumina overhang adaptor sequences were added and libraries were prepared. The pooled libraries were assessed by Agilent high sensitivity DNA kit and quantified by qPCR using the Kapa Quantification kit for Illumina prior to sequencing using the Illumina MiSeq platform. Raw 16S rRNA gene sequencing reads were quality filtered using PRINSEQ. Denoising, OTU clustering (97% identity), and chimera removal were done using USearch (v7-64bit). OTUs were aligned using PyNAST. Alpha-diversity and beta-diversity were calculated using Qiime (1.8.0). Taxonomy was assigned using a BLAST search against the SILVA SSU 123 database. Further bioinformatics was performed with the phyloseq package in R.3.2.2.

Results: Seven CF subjects had BAL samples taken for five consecutive years beginning at one year of age. Alpha diversity increased over the first 3-4 years but began to drop in the fifth year (Figure) . Additionally, the composition of the lower airway microbiome, even at phylum level, was highly variable between subjects, and also within subjects from year to year.

Conclusion: CF lower airway microbiome diversity increases for the first 4 years of life, before seeming to decline in the fifth year. The lungs of cystic fibrosis (CF) patients are chronically infected with complex, polymicrobial communities of microorganisms that are recalcitrant to antibiotic treatment. Pseudomonas aeruginosa (Pa) and Staphylococcus aureus (Sa) are the most prevalent opportunistic pathogens found in the CF lung, and co-infection with both microbes is common. Co-infection is correlated with worse patient outcomes, including increased frequency of clinical exacerbations. Mucoidy, the overproduction by Pa of the secreted exopolysaccharide alginate, is a known marker of chronic infection, with 87% of CF patients infected with mucoid Pa by age 12. Previous work from our lab has shown that mucoid Pa co-exists with Sa through the transcriptional downregulation of genes essential for the production of antimicrobial exoproducts that Pa normally uses to kill Sa, but the mechanism of this downregulation is not known. In the present study we examine the effects of exogenous alginate on Pa transcriptional regulation and interactions with Sa in order to separate the effects of exogenous alginate from the effects of endogenous alginate synthesis and secretion. We supplied exogenous alginate in co-cultures of Pa and Sa to determine whether exogenous alginate alone can facilitate Sa survival, and found that Sa survival increased significantly in the presence of high concentrations (0.025-2%) of exogenous alginate. To determine whether nonmucoid Pa exposed to exogenous alginate downregulates the same set of genes as mucoid Pa, we used a custom Nanostring codeset containing Pa genes known to be downregulated by mucoid conversion to measure Pa transcriptional changes after a short exposure to exogenous alginate. Interestingly, we found that a different set of genes is downregulated in the presence of exogenous alginate, indicating that protection of Sa occurs by distinct mechanisms in the presence of self-produced versus exogenous alginate. In the presence of exogenous alginate, we find that pqsE, which is essential for full Pa virulence towards Sa in co-culture, is downregulated. The finding that exogenous alginate can attenuate virulence of Pa towards Sa led to the hypothesis that mucoid Pa may be able to influence interactions between nonmucoid Pa and Sa. To test this hypothesis, we co-cultured mucoid Pa, nonmucoid Pa, and Sa on plastic as a biofilm and discovered that there is a delay in Sa killing by nonmucoid Pa in the presence of mucoid Pa. Therefore, mucoid conversion of Pa strains within the CF lung likely exerts indirect effects on microbial composition through changes in nonmucoid Pa gene regulation in response to exogenous alginate, highlighting the importance of intra-species diversity on inter-species interactions. Background: Pseudomonas aeruginosa and Staphylococcus aureus frequently co-infect the lungs of children with cystic fibrosis (CF). We have previously shown that staphylococcal protein A (SpA) binds to P. aeruginosa leading to bacterial aggregation and tobramycin resistance. Our aim in this study was to determine whether SpA could be detected in the sputum of children with CF and measure the resistance to different classes of antibiotics due to SpA mediated aggregation.

Methods: We conducted a prospective observational study of children with CF at the Hospital for Sick Children in June 2017. Sputum was collected and S. aureus bacterial density was measured in colony forming units (CFU)/mL as well as SpA concentrations measured by ELISA. In addition, antimicrobial resistance to levofloxacin, aztreonam, colistin, tobramycin and DNase was measured in P. aeruginosa strains grown as biofilms in the presence of SpA using confocal microscopy.

Results: We report that from 26 patient samples in our prospective study, 17 (65%) were positive for S. aureus and interestingly, SpA can be detected in sputum regardless of the abundance of S. aureus present in the culture. There is no statistically significant correlation (R 2 = 0.3571) between S. aureus CFU count and the concentration of SpA. Notably, 4 out of the 15 isolates had SpA concentrations ranged from 19.64 -254.64 pg/mL, despite being negative for S. aureus in culture. Additionally, the concentration of SpA in filtrates from clinical isolates of S. aureus was not predictive of the extent of P. aeruginosa biofilm inhibition, as measured by crystal violet (CV) assay. Analysis using confocal microscopy confirmed that the tobramycin resistance of wild-type PA01 was independent of the expression levels of SpA in S. aureus. P. aeruginosa biofilms treated with SpA had increased resistance to 1000 µg/mL tobramycin, 40 µg/mL colistin, 200 µg/mL dornase alfa and 4000 µg/mL levofloxacin, but remained susceptible to 500 µg/mL aztreonam. Moreover, persistent isolates from our Sickkids cohort had greater resistance to dornase alfa than eradicated isolates and displayed a more aggregated phenotype.

Conclusions: These findings suggest that the role of S. aureus in chronic P. aeruginosa lung infection may not directly depend on the amount of SpA in CF lungs and additional factors may contribute to inhibition of biofilm growth. Further investigation will be needed to confirm the effectiveness of different classes of antibiotics against aggregating P. aeruginosa in biofilms, as well as contribution of S. aureus to P. aeruginosa antibiotic resistance. Objectives: To describe the utilization of vancomycin and other anti-MRSA antibiotics for the treatment of APE in adult and pediatric patients with CF from the perspective of pharmacists associated with Cystic Fibrosis Canada Accredited Care Centres (CFCACC).

Methods: We conducted an anonymous national quality improvement survey using Survey Monkey®. The survey was sent to 35 CFCACC-associated pharmacists. Survey responses were gathered from December 10, 2017 to January 21, 2018. A combination of open-and closeended questions was used.

Results: Survey was completed by 57% of contacted pharmacists. Vancomycin was selected as the most frequently used antibiotic by 80% and 67% of pharmacists for adult and pediatric patients, respectively. Trough levels (mostly pre-4th dose) are reportedly used to monitor vancomycin, with a target range of 15-20 mg/L. Numerous pharmacists reported not using a loading dose upon vancomycin initiation. Other commonly used anti-MRSA antibiotics included oral linezolid, SMX/TMP, doxycycline, and rifampin.

Conclusion: This study provides the first Canadian data on vancomycin and anti-MRSA utilization for the treatment of APE in patients with CF. Vancomycin was the most commonly used antibiotic and appeared to be the first-line agent. However, there was variability in vancomycin dosing strategies, especially within the adult patient population. More research on optimal vancomycin dosing strategies in this patient population is needed. The data suggest educational opportunities for vancomycin dosing in patients with CF. (Pa) is an opportunistic pathogen that chronically infects cystic fibrosis (CF) lungs by forming multicellular biofilms. A major indication of chronic CF lung disease is intermittent acute pulmonary exacerbations (APE), which lead to intravenous antibiotic treatment against Pa infections. Despite the aggressive and intensive antimicrobial interventions against Pa, the reasons for the poor outcome of treatments during chronic stages of lung infection remains unclear. The emergence of the phenotypically diverse populations of Pa in CF lungs has previously been reported but the impact of this on the patient remains unknown. To unravel the effects of population level diversity on both virulence and host response, we first evolved a diverse in vitro PAO1 population in biofilms over 50 days in synthetic sputum medium (SCFM). Our evolution experiment resulted in the emergence of distinct Pa morphotypes which contribute to population-level diversity. Using an ex vivo pig lung model, we observed differential levels of IL-8 production in alveolar tissues infected with different morphotypes. Furthermore, IL-8 levels were negatively correlated (p=0.0348, r= -0.6109) with total protease production of each evolved isolate. We next studied the levels of protease production and tissue damage by Pa isolates collected from CF patients with varying lung functions. We observed differential levels of protease activity by collected isolates and whole mixed diverse populations during planktonic growth in SCFM compared to co-cultures with bronchial epithelial cells (BECs). Using CF-derived BECs in the air-liquid interface (ALI), we assessed changes in induction of pro-inflammatory responses and integrity of the epithelial layer infected with diverse Pa population compared to the single collected isolates. Our work highlights that phenotypic diversity within Pa populations can influence how this organism can impact the pro-inflammatory responses and damage lung tissue; and that Pa population diversity may contribute to pulmonary exacerbation during chronic CF lung infection. Investigating interactions between phenotypically diverse isolates may help inform new strategies for the treatment and control of biofilms and infection. The use of antibiotics has greatly increased the length and quality of life for cystic fibrosis (CF) patients. However, antibiotic resistance is increasing at an alarming rate with the need for alternative therapeutics. Nitric oxide (NO) is an attractive alternative to conventional antibiotics because it has broad spectrum bactericidal activity and acts in a multimechanistic manner, decreasing the risk of undesirable resistance. Our therapeutic, BIOC51, is a NO-donor modified from a natural biopolymer that releases NO spontaneously in solution. We have shown that BIOC51 is bactericidal against > 15 species of bacteria, including multidrug-resistant Pseudomonas aeruginosa and methicillin-resistant Staphylococcus aureus. The objective of this study was to evaluate the pre-clinical therapeutic potential of BIOC51 against nontuberculosis Mycobacterium (NTM), which affects ~10% of CF patients and is difficult to treat even with today's best antibiotics.

Methods: The antibacterial efficacy of BIOC51 was first evaluated against M. abscessus, M. avium, and M. intracellulare in vitro using standard MIC and MBC assays. For the maximal tolerable dose (MTD) study, 9 mice (3 per group) were administered 0, 150, or 300 mg/kg BIOC51 daily for 3 consecutive days by intratracheal instillation (IT). Mice were observed for signs of adverse effects (i.e., reduced weight, labored breathing, reduced activity, or ruffled hair) at 10 min, 1, 2, 4, and 24 hours after dosing on days 1, 2, and 3. For the in vivo efficacy study, mice were infected IT with 1x10 5 CFU/mouse of M. abscessus. Three mice were sacrificed 1 day post-infection to determine initial bacterial loads. The remaining mice were split into 4 groups of 6 mice each. Mice were treated once daily for 8 consecutive days with either vehicle, BIOC51 (300 mg/kg, IT), amikacin (AMI; 150 mg/kg, s.q.), or clarithromycin (CLA; 250 mg/kg, gavage). Animals were sacrificed 24 hours after administering the last dose. Bacterial loads were determined by plating dilutions of the lung, spleen, and liver homogenates.

Results: BIOC51 kills M. abscessus, M. avium, and M. intracellulare in vitro and significantly reduces in vivo M. abscessus levels (by 2.4 logs). No adverse effects were observed in mice even at the maximum feasible dose (300 mg/kg). No changes were observed in the liver or spleen, which was anticipated given that BIOC51 is a fast-acting and locally-acting therapeutic.

Conclusions: BIOC51, a nitric oxide-releasing biopolymer, is an innovative alternative to antibiotics with strong therapeutic potential for the treatment of NTM lung infections. It has been shown to eradicate NTM in vitro and to reduce NTM levels in vivo with a favorable safety profile. Future work will focus on evaluating BIOC51 efficacy in a rat model of Pseudomonas aeruginosa infection to determine the therapeutic potential of BIOC51 against other CF-relevant pathogens.

Acknowledgment: This work was funded by Vast Therapeutics. (1):101-15). Due to the natural and induced resistance of NTM species to many of the commonly used antibiotic treatment regimens, there is an urgent need for novel therapeutics to combat these infections. This is especially important, as NTM culture positivity is associated with worse clinical outcomes in CF, including an increase in the rate of decline in pulmonary function (Martiniano SL, et al. Pediatr Pulmonol. 2017 ;52(S48):S29-S36). Hydrogen peroxide (H 2 O 2 ) is an important reactive oxygen species that functions as an antimicrobial oxidant during the host immune defense. Additionally, H 2 O 2 can be utilized by the host haloperoxidase system to form hypothiocyanite (HOSCN) and hypochlorite (HOCl), two potent oxidants with roles in immune defense (Chandler JD, Day BJ. Free Radic Res. 2015;49(6):695-710). The role of oxidants in host control of NTM is poorly understood. We measured the susceptibility of genetically diverse clinical NTM isolates belonging to the M. abscessus complex to killing by H 2 O 2 , HOSCN, and HOCl in order to examine the variability amongst strains.

Methods:M. abscessus isolates were exposed to a range of concentrations of either H 2 O 2 or HOCl over a time course of 0 to 4 hours. Additionally, bacteria were exposed to an enzyme system composed of glucose oxidase (GOX) and lactoperoxidase (LPO) London, United Kingdom; 2. Department of Microbiology, Royal Brompton and Harefield NHS Foundation Trust, London, United Kingdom; 3. Imperial College London, London, United Kingdom Background: Isolation of clinically significant filamentous fungi (FF) from patients with Cystic Fibrosis (CF) is increasing. However, the true prevalence is hard to estimate due to different sputum processing protocols across laboratories. Thus, routine respiratory sampling and processing may significantly underestimate the true prevalence of FF in CF patients.

Aims: To compare the prevalence of respiratory tract colonisation from Aspergillus and non-Aspergillus filamentous fungi (NAFF) in a large cohort of adult CF patients followed at the Royal Brompton and Harefield NHS Foundation Trust (>600 patients) prior to and after the adoption of extended fungal cultures (i.e. 28 vs 7 days of incubation) in 2014. Secondarily, use of domiciliary antifungal treatment with azoles was assessed over the same time.

Methods: A retrospective analysis of data collected from the local CF patient registry was performed. Microbiological data were reported annually from sputum cultures in the previous 12 months between 2010 to 2017. Patients were considered colonised if they had ≥1 positive sputum culture over the selected year. Comparison was made between data from two 4-year periods: 2010-13 (T 1 ) vs 2014-17 (T 2 ) before and after extended culture processing was introduced. A case was defined from the number of reported patients each year in the registry, so individual patients recurred in each yearly cohort.

A total of 4776 annual entries were evaluated (T 1 , N=2436; T 2 , N=2340). FF were detected in 25.3% and 27.1% of cases in T 1 and T 2 , respectively (p=ns). Aspergillus was the most common species isolated in both periods (93.8% in T 1 , 80.4% in T 2 ), followed, in order, by Scedosporium, Exophiala and Rasamsonia. Adoption of extended cultures did not affect Aspergillus prevalence, which slightly decreased from 23.7% in T 1 to 21.9% in T 2 (p=ns). In contrast, prevalence of NAFF significantly increased over T 2 up to 2.7 times compared to T 1 (from 2.8% to 7.6%; p<0.001). Most frequently isolated NAFF in both T 1 and T 2 was Scedosporium (2.2% and 3.7%, respectively); however, the highest relative change in prevalence was recorded for Rasamsonia (17.5 times, from 0.04% to 0.7%) and Exophiala (8.3 times, from 0.3% to 2.5%). Annual trends showed an increase in Aspergillus isolates from 23.7% in 2010 up to 25.6% in 2014, with a subsequent decline to 17.5% in 2017 (p=0.006). Similarly, NAFF isolates increased significantly from 2010 (0.7%) peaking in 2015 (8.8%), with a small subsequent fall to 7.8% (p<0.001). Among patients with positive fungal cultures, prescription of azole therapy was significantly higher over T 2 (29.2%, range 26.5% -31.4%) than T 1 (22.7%, range 15.4% -27.9%) (p=0.009).

Conclusions: Use of extended fungal cultures significantly increased NAFF detection. The total number of Aspergillus and total FF isolates did not change significantly. Adoption of extended fungal cultures in T 2 was also associated with increased use of azoles, with a subsequent decline in positive samples for both Aspergillus and NAFF. CF is a systemic disease characterized by a reduction in lung function and a heightened susceptibility to opportunistic bacterial infections, beginning at a young age and developing into chronic, polymicrobial infections by adulthood. Interestingly, bacterial populations colonizing the lungs of CF patients undergo a series of successional changes as patients age; these changes may be a determinant of pulmonary function and overall health. Nontypeable Haemophilus influenzae (NTHi) is a prevalent opportunist infecting younger CF patients. NTHi is a normal commensal of the upper airways, however, in patients with chronic respiratory diseases, such as CF or chronic obstructive pulmonary disease (COPD), this opportunistic pathogen is able to colonize and persist in the upper and lower airways. There remains a significant gap in our understanding of how early stage infections may influence the complex host-pathogen interactions that are at play in the CF lung. Given that bacterial infection early in life is associated with poorer clinical outcome later in life, we hypothesize that these early infections may play a significant role in the pathogenesis of CF, potentially by facilitating infection with pathogens associated with later stage CF. In this study, we sought to model the earliest stages of CF infection in young children to establish a foundation for additional studies modeling the impact of successional changes on the progression of CF and wellness of patients. Given that NTHi infects primarily younger CF patients, we used both young, recently weaned rats and adult rats to elucidate any differences in NTHi infection based upon age. Additionally, this model progressively develops the mucus obstruction phenotype, recapitulating a critical component of CF pathogenesis, unlike the murine model of CF. Adult and juvenile CFTR -/-and wild-type rats were intratracheally infected with a model strain of NTHi, 86-028NP, after which total lung homogenate was collected for bacterial counts, differential cell count, and a cytokine/chemokine Luminex panel. The data show that CFTR -/-and wild-type rats of both age groups are susceptible to infection by NTHi, as there were significant NTHi populations present in the lungs of infected animals up to 72 hours postinfection. This is in striking contrast to mice (including CFTR -/-mice) which are not permissive to NTHi infection. Additional studies addressing the role(s) for specific NTHi factors involved in persistent infections will be presented. The results of these studies provide encouraging evidence that the rat model of CF can be used to examine infection by several early CF pathogens and potentially provide deeper insights into the nature of the pathogenesis of this disease beginning at early life. The isogenic mutant infection data will be presented at the conference, and this work will add to the growing understanding of early CF infections. The results from this study could provide valuable insight into the nature of the CF lung and the host-pathogen interactions that are most important for establishment of infection by common pathobionts. 3 1. Pediatrics, Univ North Carolina Chapel Hil, Chapel Hill, NC, USA; 2. Queens University, Belfast, United Kingdom; 3. University of Washington, Seattle, WA, USA Aim: Epidemiologic studies indicate that chronic but not intermittent MRSA infection in people with CF is associated with worsening clinical outcomes. We hypothesized that phenotypic changes of MRSA occur between early and late infection stages that may be associated with clinical outcomes.

Methods: Assays comparing initial and chronic isolates under aerobic and anaerobic conditions measured δ-hemolysin as a marker for agr function, biofilm formation and susceptibility to different antibiotics (Minimum inhibitory concentration, MIC; E-test™). Patients were classified as "decliner" vs. "stable" based on percentage of clinic visits with prescription of acute antibiotics, change in BMI percentile and FEV 1 percent predicted as available.

Results: Paired isolates were available from 32 subjects with a median age (SD) at time of initial sampling of 8 (9.75) years. Median (SD) time between initial and chronic isolate was 2.2 (1.1) years. Subjects classified as decliners (n=16) tended to be older (12.3 vs. 6.2 years. p=0.08) but did not differ from stable subjects in regards to BMI% (36th vs. 44th percentile; p=0.5) or FEV 1 (68% n=7 vs. 64% n=12; p=0.7) at onset of MRSA infection. Two subjects could not be classified. At the chronic stage, decliners had significantly lower BMI (36th vs. 55th percentile, p=0.04) and FEV 1 (55% n=8 vs. 79% n=12) predicted. Proportion of clinic visits with acute antibiotic prescriptions were 72% vs. 28% (p<0.001) in decliners vs. stable. Typing showed t002 (USA100) in 36%, t008 (USA300) in 37%, with 10 other spa types present in the remaining patients with 75% of isolates PVL positive. Four patients had isolates with different spa types at initial vs. chronic sampling.

There was no difference in hemolysis or biofilm growth for initial vs. chronic isolates; however, biofilm growth was slower under anaerobic compared to aerobic conditions. Similarly, MIC did not differ in initial vs. chronic isolates but differed by anaerobic vs. aerobic condition: the entire isolate collection exhibited lower MICs for vancomycin, and fusidic acid, but higher MICs for TMP-SMX under anaerobic compared to aerobic conditions. Analyses by clinical outcomes showed slower anaerobic biofilm growth, higher rifampin MICs under anaerobic and aerobic conditions, and higher MIC to TMP-SMX under aerobic conditions in isolates of decliners compared to stable subjects.

Conclusion: Duration of infection (i.e. early vs. late isolates) was not associated with phenotypic differences. However, anaerobic conditions as occur in CF lung disease affected growth and antibiotic susceptibilities. Here, 50% of patients had declining status that was associated with specific MRSA phenotypes. 

The goal of this study was to determine eradication rates of new BCC infections with management using a standardized clinical protocol.

Methods: We established a clinical protocol for management of new BCC airway infections in 2010. Children who cultured positive for BCC for the first time were stratified into two groups, low-risk or high-risk, based on standard criteria (e.g. Burkholderia genomovar, lung function, respiratory symptoms, or parental preference). Patients deemed high-risk received a combination of intravenous meropenem and ceftazidime; inhaled tobramycin; and oral sulfamethoxazole/trimethoprim or minocycline over six weeks. Patients considered low-risk were treated as needed based on respiratory symptoms following routine clinical practice and followed with routine cultures. We performed a retrospective study of children who cultured positive for new BCC airway infections in the six years following guideline implementation. We defined eradication as remaining BCC negative on follow-up airway cultures over one year with at least three sputum or bronchoalveolar lavage cultures. We determined the overall rate of eradication and compared eradication between the high-risk and low-risk groups using Fisher's exact test.

Main Results: Thirteen patients (62% male, median age 13 years [range 8-17 y]) had a new BCC airway infection detected. BCC was identified as: B. multivorans (38%), B. cenocepacia (23%), B. gladioli (15%), and other BCC species (23%). Seven patients were stratified as high-risk and six were considered low-risk. There was no statistically significant difference in age, genotype, BMI, or FEV1 between groups. In the high-risk group, 3 (43%) were coinfected with Pseudomonas aeruginosa on the first positive culture for BCC compared to no patients in the low-risk group. BCC was eradicated in 6 of 13 (46%) patients overall. In the high-risk group, BCC was eradicated in 4 of 7 (57%) patients compared to 2 of 6 (33%) in the low-risk group (p=0.6). Of those who did not eradicate, two were infected with B. cenocepacia and one with B. multivorans in the high-risk group, whereas in the low-risk group one had B. multivorans and three had other BCC species.

Conclusions: Over half of patients with new BCC airway infections managed using our standardized clinical protocol remained negative for BCC for one year. Patients deemed low-risk had lower rates of eradication, though the difference was not statistically significant. Further evaluation of approaches to eradicate new BCC airway infections is warranted. Acknowledgment: Supported by the Cystic Fibrosis Foundation HONG16B0. The opportunistic pathogen Pseudomonas aeruginosa forms life-long chronic infections in the lungs of people with cystic fibrosis (CF), leading to a slow decline in lung function and ultimately death. In the lung environment, P. aeruginosa evolve and often acquire mucA mutations, which are associated with mucoid conversion and poor disease prognosis for CF patients. These clinical mucA mutations lead to the production of truncated proteins that are thought to be fully degraded in the cell. Since MucA inhibits the alternative sigma factor AlgU, these mucA mutations lead to the misregulation of the AlgU regulon. Paradoxically, our work shows that a portion of mucA is required for viability in many strains of P. aeruginosa.

Here we examine why mucA is essential. Our results show that mucA is no longer essential in a strain lacking algU and that mucA alleles that abolish the interaction with AlgU are insufficient for viability. These results suggest that the loss of the MucA-AlgU interaction results in bacterial cell death. Because MucA regulates AlgU, we are currently investigating whether aberrant expression of AlgU-regulated genes in the absence of mucA is the cause of the viability defect. To determine if specific sets of genes in the AlgU regulon are responsible for mucA essentiality, we deleted the genes encoding the three major AlgU-regulated transcription factors AlgR, AlgB, and AmrZ. Our work shows that mucA is still essential in strains lacking one of these three transcription factors. While more work is still needed to understand why this interaction is required for bacterial viability in P. aeruginosa, our results suggest that this interaction may serve as a good therapeutic target. Rationale: Episodic respiratory virus infection including human rhinovirus (RV) likely contributes to early progressive CF lung damage. Given the CF airway has been reported to exhibit defective viral responses, it is critical to assess the innate immune responses unique to CF airways and identify potentially new therapeutic strategies. This study profiled the transcriptome of CF airway epithelial cells (AECs) in response to RV infection.

Methods: Primary AECs were cultured from 10 non-CF (age 3.5 ± 1.4 years; 4 males) and 9 CF (age 2.4 ± 2.2 years; 5 males) airways, infected with RV1B (multiplicity of infection: 12.5) and 24 hours later, RNA was extracted and sequenced (RNA-seq) by Illumina Hi-Seq2500. Sequence reads were mapped to human genome transcripts using "HISAT2." Transcript alignments were assembled and quantified by "StringTie." Differential gene-expression analysis was then performed using DESeq2 and upstream analysis by Ingenuity Systems (QIAGEN, Germany).

Results: We identified 1371 and 1329 differentially expressed genes (DEGs; p<0.05) by non-CF and CF AEC respectively in response to RV1B infection. Of these, 337 (25%) were unique to the CF AEC response. Pathway analysis identified common DEGs that were involved in: interferon signalling, antigen processing-cross presentation, RIG-1/MDA-5 signalling and others. Interestingly, primary CF AEC featured 9 unique pathways (p<0.05), with tryptophan catabolism, amyloid formation, DNA damage/ telomere stress induced senescence and toll-like receptor TLR6:TLR2 cascade of interest. Upstream regulators analysis (p<0.05; -2≤ z-score ≥2) suggested downregulation of transcriptional regulators RARA, NCOA1, TAL1 and IKZF3 and upregulation of ANLN, CD3 and JAK1 could be areas of interest in understanding the CF AEC response to RV1B infection.

Conclusions: Overall, primary CF AEC exhibit a similar innate immune response to non-CF AEC in response to RV infection. However, beyond these conserved pathways there were additional unique genes and biological pathways of interest. The results suggest modulating these early innate responses by CF AEC might be a future treatment strategy. 1 1. Microbiology, Vall d'Hebron University Hospital, Barcelona, Spain; 2. Cystic Fibrosis Unit, Vall d'Hebron University Hospital, Barcelona, Spain Background:Burkholderia cepacia complex (BCC) are difficult to manage gram-negative opportunistic pathogens of special relevance for CF patients. They are associated with a bad prognosis. BCC treatment is challenging because of their intrinsic resistance to multiple antibiotics and their remarkable ability for developing acquired resistance. Avibactam is a novel non-β-lactam β-lactamase inhibitor, with activity against Ambler class A (e.g., extended-spectrum β-lactamase [ESBL] and KPC), class C, and some class D enzymes. Recently, the activity of ceftazidime-avibactam against the carbapenemase PenA characteristic of BCC has been documented. Therefore, it could be useful for the treatment of CF patients. The aim of this study was to evaluate the in vitro activity of ceftazidime-avibactam against BCC strains isolated from chronically infected CF patients and to compare it to other antibiotics commonly used for the treatment of respiratory infections caused by these microorganisms.

Methods: The antimicrobial susceptibility of 18 BCC isolates (10 B. contaminans, 3 B. multivorans, 2 B. stabilis, 2 B. cepacia, 1 B. cenocepacia obtained from respiratory samples of 17 CF chronically infected patients) was determined by disc diffusion technique to ceftazidime, meropenem, minocycline and cotrimoxazole, using the specific cut-off points of CLSI (2017) and to ceftazidime-avibactam, piperacillin-tazobactam, cefepime and ciprofloxacin, using the cut-off points of EUCAST for Pseudomonas spp. The MIC to ceftazidime-avibactam was further determined by gradient diffusion technique (E-test).

Results: The resistance rate to conventional antibiotics was high: ciprofloxacin (84%), cefepime (78%), minocycline (83%), cotrimoxazole (78%), meropenem (67%), ceftazidime (56%) and piperacillin-tazobactam (61%) without significant differences between species; 7/18 strains were resistant to all the conventional antibiotics tested. Ceftazidime-avibactam was the most active antibiotic: 78% of the strains were susceptible, with MIC50 and MIC90 values for this antibiotic of 4 and 24 mg/L respectively. Of interest, 5 out of the 7 panresistant strains were susceptible to ceftazidime-avibactam. Six out of the 7 panresistant strains were B. contaminans. The combination of avibactam with ceftazidime restored the susceptibility to ceftazidime in 60% of the strains studied.

Conclusion: Ceftazidime-avibactam is the antibiotic that displayed the best in vitro activity of all those evaluated in this study. Given the high level of resistance of this bacterial genus, these results suggest that ceftazidime-avibactam could be considered a useful therapeutic option for the treatment of BCC. Nevertheless, the possibility of developing secondary resistance warrants extreme caution in its use, particularly in carriers of strains resistant to conventional antibiotics.

Gartner, S. 1 ; Cano, A. 2 ; Barroso, F. 2 ; Rovira, S. 1 ; Spangenber, A. 1 ; Martin-Gomez, M. 2 1. Cystic Fibrosis Unit, Vall d'Hebron Univ Hosp, Barcelona, Spain; 2. Microbiology, Vall d'Hebron Univ Hosp, Barcelona, Spain Background: P. aeruginosa (PsA) chronic airway infection has been linked to worse outcome of CF patients. Eradication may be achieved only before chronic infection establishes, making early detection crucial. Initial PsA isolates usually arise during the first years of life by culturing oropharyngeal (OP) swab samples. RT-PCR may be more sensitive than culture to detect PsA, providing a shorter turn-around time. As compared to conventional ones, flocked swabs in liquid transport medium have been described to enhance the recovery of pathogens from clinical samples. This study was aimed to compare the performance of a real-time PCR versus conventional culture for the detection of PsA in OP samples taken with flocked swabs from CF children aged up to 5 years.

Methods: Retrospective observational pilot study (March-December, 2017) . Follow-up OP samples were collected with the aid of flocked swabs from CF children aged 0-5 years old regularly visited at the Vall d'Hebron Hospital CF Unit. As per standard-of-care protocol, samples were processed for routine culture (vigorous vortexing-centrifugation of the liquid transport media-inoculation of the pellet into standard agar plates-incubation up to 5 days). DNA was obtained from 0.5 mL of sample remnants (EasyMag, BioMérieux), and frozen until assay. A previously validated RT-PCR targeting the gyrBgene (http://www.tdx.cat/handle/10803/290733) was run in bulk by a technician blinded to culture results at the end of the study period in a SmartCycler instrument. Culture results from OP samples taken during the previous 6 months and the 3 months after the study period were also reviewed. Sensitivity (Se), Specificity (Sp), Positive and Negative Predictive Values (PPV, NPV) for the PCR technique, and k index of concordance between PCR and culture results were calculated.

Results: During the study period, 166 OP samples (median 6 samples/ patient, range 2-12) from 29 pediatric patients (median age 4 years) were processed. PsA was detected in culture (C+) of 13 samples from 11 patients (37.9%). The lower detection limit of the PCR was 1.56 log CFU/mL. PCR was positive (PCR+) in 19 OP from 13 patients but only 6 OP from 6 patients were C+/PCR+. In 2 patients, PCR+ preceded C+ results by a mean of 50 days. One patient was PCR+ 1 month after a C+, and in 6 patients no PsA was documented in culture 3 months before or after a PCR+. PCR-/C+ results were obtained in samples with <1 log CFU/ plate. Se, Sp, PPV and NPV of the PCR technique were 46.2, 97.9, 31.6 and 95.2% respectively, with a fair k index of concordance with culture (0.311,SE 0.115).

Conclusion: The studied real-time PCR technique offered a high NPV to rule out the presence of PsA in OP samples collected with flocked swabs from young CF children. Nevertheless, it should be performed along with routine culture, which is still needed to assess the presence of this pathogen in these samples, particularly when the bacterial load is low. Introduction: Bacterial biofilms limit the effectiveness of conventional antibiotics, leading to chronic infections that are difficult to eradicate in patients with CF. NX-AS-401 is a quorum sensing inhibitor that has been found to downregulate genes involved in the formation and maintenance of biofilms (Jakobsen TH, et al. Antimicrob Agents Chemother. 2012; 56:2314-25) . Neem Biotech is testing NX-AS-401 for the ability to disrupt biofilms in CF relevant assays, while having limited impact on lung epithelium.

Objectives: Demonstrate biofilm disruption by NX-AS-401 in CF relevant bacteria. Determine the safety margin for application to the lung epithelium. Demonstrate efficacy in CF patient sputum and in an in vivo lung infection model.

Methods: Minimum biofilm eradication concentration (MBIC) was determined by inoculating compound-treated media with P. aeruginosa or S. aureus and quantifying biofilm formation at 24 h using crystal violet staining. MBEC TM Biofilm assay was conducted as previously described (Innovotech, 2015) , treated with compound and/or tobramycin. Human bronchial epithelial cells (HBECs) were exposed to NX-AS-401 (1 µM -1 mM) for 24 h. Transepithelial electrical resistance (TEER) was measured using an EVOM2 and lactate dehydrogenase (LDH) release was quantified. Uninfected female C57Bl/6J mice were dosed intratracheally with NX-AS-401(0.01 -50 mg/kg). Mice were assessed at 24 and 48 h for clinical signs, weight loss, inflammation and irritation. Sputum from CF patients was added to tubes containing tobramycin, compound alone, tobramycin and compound, or PBS and incubated in aerobic conditions for 24 h. At 0, 6 and 24 h, 100 µl aliquots were removed and total microbial load and P. aeruginosa viable count were determined using blood agar and selective plates. C57Bl/6J mice were inoculated intratracheally with P. aeruginosa colonised agar beads. Treatments of compound and/or tobramycin were dosed every 24 h following established infection for 7 days. Lung CFU burden and histology were compared between treatment groups.

Results: Single dose treatment with NX-AS-401 inhibited the formation of P. aeruginosa and S. aureus biofilms in vitro (IC 50 0.8 µM). Repeat dosing of preformed biofilms with NX-AS-401 in combination with tobramycin resulted in dose-dependent inferred disruption of biofilms and a decreased MBEC. At the active concentrations identified, there was no impact of NX-AS-401 on HBEC integrity, with cytotoxicity EC 50 >500 µM. In vivo intratracheal administration showed no adverse effects at all but the highest 2 doses. Exposure of natural flora in CF patient sputum samples to compound-tobramycin combinations resulted in a reduction of microbial load compared to tobramycin alone. In vivo data showing enhanced clearance of chronic lung infections with combination treatment will be presented. Introduction: In-vitro data suggests ivacaftor may have direct antimicrobial properties against Staphylococcus aureus but not against Pseudomonas aeruginosa. As some people with CF have now received ivacaftor for over 5 years, understanding the long-term impact becomes increasingly important in order to prognosticate and also rationalise adjunctive therapies.

Objective: To investigate the long-term microbiological outcomes associated with the use of ivacaftor.

Methods Table) . .02). Significant reductions in P. aeruginosa remained after adjustment for the number of samples submitted. There were no differences in the use of inhaled antipseudomonals between the ivacaftor and nonivacaftor group (72% vs. 69% respectively, p=0.32) or in the total years of inhaled antibiotics across the study periods (2.8 years respectively vs. 2.7 years, p=0.71).

Conclusions: We found that ivacaftor use is associated with a reduction in sputum positivity for P. aeruginosa and S. aureus. The more pronounced effects we observed against P. aeruginosa suggest the changes are not secondary to a direct bactericidal effect. Over one-third of people receiving ivacaftor remain chronically infected with P. aeruginosa and greater understanding of the outcomes in this population is required.

Crude annual odds-ratios [95% confidence intervals] for a positive respiratory culture.

Langton Hewer, S. 1,2 ; Smyth, A.R. 3 ; Jones, A. 4 ; Williamson, P. 4 Medicine, Bristol Royal Hospital for Children, Bristol, United Kingdom; 2. University of Bristol, Bristol, United Kingdom; 3. University of Nottingham, Nottingham, United Kingdom; 4. Clinical Trials Research Centre, University of Liverpool, Liverpool, United Kingdom Introduction: Infection with the bacterium Pseudomonas aeruginosa (PsA) is common in children and adults with cystic fibrosis (CF). Long term infection is associated with poor outcomes. Newly acquired infection is treated with antibiotics in an attempt to prevent or delay long term infection. However, there is equipoise about the best method to eradicate PsA from the lower respiratory tract. In the UK this has usually involved a combination of oral and/or intravenous antibiotics with nebulised colistin.

Objectives: This study was conceived to test the superiority of 10-14 days of intravenous ceftazidime with tobramycin compared to three months therapy with oral ciprofloxacin.

Design: Multicentre parallel randomised controlled, open label trial. A central web-based randomisation system used computer generated tables to allocate treatments in a 1:1 ratio.

Intervention: Ten to fourteen days of intravenous ceftazidime with tobramycin or 3 months therapy with oral ciprofloxacin. Both treatment arms included three months of nebulised colistin. Follow-up was for up to 2 years.

Main outcome measures: Primary outcome was successful eradication of P. aeruginosa infection at three months post randomisation, and remaining infection free through to 15 months post randomisation. Information on safety and secondary outcomes were collected throughout the trial.

Results: Two hundred eighty-six patients were randomised from 68 UK sites and one site in Italy. The first patient was randomised on 5th October 2010 and the last patient was randomised on 27th January 2017. The final study visit of the final patient took place on 10th April 2018.

The data lock will take place on 31/5/18 and full results will be available in August 2018 and will be presented at NACFC 2018 in full.

Discussion: This is the largest investigator led study of CF in the UK study in which adults and children with CF and early infection with PsA have been allocated two different eradication regimens and will provide evidence as to whether intravenous therapy is superior to oral therapy. This will have implications for patient care and for development of guidelines for PsA eradication. Burkholderia cepacia complex (Bcc) is a group of 22 closely related species of gram-negative bacteria of clinical importance as an opportunistic pathogen causing chronic lung infections in people with cystic fibrosis (CF). Bcc infections are rarely eradicated and the high-level antibiotic resistance means that combatting these chronic infections is particularly challenging. One species in particular, B. cenocepacia, is associated with poor clinical outcomes and an increased rate of mortality. Bcc can survive and persist during chronic infection by mechanisms that are not fully understood.

We previously showed that a series of Irish sequential clinical B. cenocepacia isolates from two chronically infected CF patients increased their ability to attach to CF lung epithelial cells (CFBE41o -) over time of chronic infection (1) . An in-depth proteomic analysis identified 150 proteins that were consistently upregulated over the course of infection. Twenty of these proteins are encoded within the 50-gene low-oxygen-activated (lxa) locus, which is upregulated in response to hypoxic conditions (2) . Their consistent increased abundance over time of infection in both patients' isolates suggests that the lxa locus may play an important role in chronic infection.

There are six universal stress proteins (USPs) encoded on the lxa locus, all of which showed increased abundance in the sequential isolates and have not previously been studied in Bcc. The objective of this project was to determine the role USPs play in the pathogenesis of Bcc during chronic infection.

A single gene deletion mutant (Δusp) of one USP (BCAM0276) in the B. cenocepacia strain K56-2, showed a 90% reduction in attachment to CFBE41ocells when compared to the wild-type K56-2 strain (p<0.005). The Δusp mutant also showed increased sensitivity to peroxide-induced oxidative stress (p<0.0001) and low pH (p<0.05) relative to wild-type Bcc, both of which are relevant to the CF lung and macrophage environments. There was a reduction in both uptake and survival of Δusp in comparison to wild-type, in the U937 human macrophage-like cell line suggesting the USP plays a role in the intra-macrophage survival of Bcc. Ethical approval has been obtained to examine survival and interaction of Δusp in macrophage from CF patients and healthy volunteers.

Overall, USPs, previously associated with adaptation to low oxygen conditions may play a significant role in Bcc pathogenesis and its adaption during chronic infection.

Acknowledgements 

In patients with cystic fibrosis (CF), conventional treatments mostly focused on lessening the manifestations of CF disease (namely bacterial infection and host inflammation) are now supported with new therapeutic approaches to modulate and correct cystic fibrosis transmembrane conductance regulator (CFTR). These new therapies have improved outcomes in individuals with CF positively impacting the clinical status but with an uncertain impact on infection, particularly by Pseudomonas aeruginosa, which is the most prevalent bacterium infecting CF patients. Thus, combination therapies can be beneficial and vaccination can represent a rediscovered opportunity in new health conditions. Although several vaccine formulations against P. aeruginosa have reached clinical testing in the past, none of them has become a licensed product and this bacterium remains a challenging vaccine target. To identify new protein antigens against P. aeruginosa, we explored reverse vaccinology approach integrated with additional bioinformatic tools. A set of 52 antigens, classified as known (31) and unknown (21) functions, were selected on the basis of their localization on the cell surface, absence of similarity with E. coli K12 as well as with human and mouse counterparts. Furthermore, conservation of selected antigens was evaluated by comparative genome analysis in a comprehensive panel of P. aeruginosa genomes available in GenBank from different origin and longitudinally isolated from CF patients. To assess the capability of selected antigens, single or in combinations, to provide protection against P. aeruginosa infection, we established a vaccination protocol in the immunocompetent murine model of acute respiratory infection. This systematic screening identified five combinations, capable of significantly increasing the survival rate among challenged mice. Of note, all combinations included PA5340, a protein of unknown function exclusively present in P. aeruginosa. The maximum proportion of protection in challenged mice was obtained with PA5340 combined with PA3526-MotY. Both proteins were confirmed to be surface exposed by immunofluorescence and able to trigger a specific immune response. Thus, combinations of selected candidates, more than single antigens, are effective to control P. aeruginosa infections in a mouse model of acute pneumonia. Overall this study confirms the capability of reverse vaccinology to provide new impulse to the research of vaccines against P. aeruginosa infection through the rapid identification of novel and promising vaccine candidates. Therefore, we propose these new proteins as potential candidates of a new multicomponent vaccine to combat P. aeruginosa infection in CF patients.

This work was sponsored by Novartis Vaccines and Diagnostics Srl, now acquired by the GSK group of companies.

This Background: Chronic pulmonary infection due to P. aeruginosa (PA) is a negative prognostic factor for cystic fibrosis (CF) patients. Early antibiotic treatment can eliminate the bacteria in more than 70% of cases and thus delay the development of chronic infection. Following eradication, lung re-infection can occur due to PA strains with identical genotype. This may be due to PA re-colonization from the patient's paranasal sinuses. It has been hypothesized that PA undergoes genetic adaptation in the upper respiratory tract of CF patients. However, at the moment, routine microbiological assessment of the upper airways is not part of standard care.

Aims: The aim of the study is to evaluate the microbiological status of CF patients who had undergone eradication therapy in the early stages of infection using genotyping of PA strains.

Methods: From September 2016 to February 2018, 35 patients never colonized by PA or PA-free according to the Leeds' definition (Lee TWR, et al. J Cyst Fibros. 2003; 2:29-34) , were enrolled in a PA early-eradication trial. Lower airways sampling was performed using expectorated sputum or deep throat swab. Upper airways samples were collected by nasal lavage (NL), using the Mainz method (Mainz JG, et al. Thorax. 2009; 64(6) :535-40). Eradication success will be defined as 3 negative, successive PA cultures over 6 months. PA strains were genotyped by BOX-PCR.

Results: PA strains from 35 patients (mean age 13.39 years ± SD 10.97) were analyzed. Four (11.4%) of 35 patients were never colonized, 31 (88.5%) were PA-free (mean time from previous PA isolation 2.14 years ± SD 1.52). In 28 (80%) patients PA was eradicated, in 5 (14.3%) PA treatment failed, 2 (5.7%) patients are still in follow-up. Four (19%) out of 21 NL were positive for PA at the beginning of treatment. Four (11.4%) PA isolates from lower airways showed a mucoid phenotype. A total of 48 PA strains, 24 from throat swab and 17 from sputum, and 7 from NL were further analyzed. No cross infection was detected at first infection among patients. Identical PA genotypes from upper and lower airways were detected in patients in whom simultaneous sampling was performed (1 missed data). In the following 6 months of follow-up all subsequent PA isolates showed an identical genotype. After one year, 6 (17%) patients were recolonized by PA strains and 3 of them (50%) had recolonization by strains of the same genotype while in the 3 others the genotype was different.

Conclusions: The presence of the same strains' genotypes in the nasal lavage and in sputum or swabs suggests that the sinuses play a role in the acquisition and persistence of PA in CF patients. Eradication therapy outcome could be improved by means of strain genotyping, leading to reconsider the percentage of eradication and to a different management of early PA infection. Background: Cystic fibrosis (CF) patients who are chronically infected by Pseudomonas aeruginosa (PA) are periodically treated with antibiotics due to the respiratory exacerbations which they experience. This therapy is associated with a progressive increase in antibiotic resistance, which is regarded with real apprehension by patients and healthcare personnel. The clinical significance of this drug resistance is unclear and so the aim of this study was to evaluate the prevalence of multiresistant (MDR) strains of PA in our CF center and evaluate their clinical impact over time.

Methods: This retrospective study has continued for 4 years in patients who are regularly followed-up in our CF center. We adopted the definition of MDR strains according to the international literature (https://www.cff. org/). Patients infected by MDR PA strains (group A) were compared to those patients infected by non-MDR strains in our center (group B). The clinical impact of infection due to MDR PA infection was analyzed using FEV 1 decline, day hospital stays and antibiotic use.

Results: The incidence of CF patients infected by MDR PA has increased in our center from 12.6 to 17.8%. in the study period. MDR PA strains showed the following rate of resistance to antibiotics: gentamicin 98%, levofloxacin 98%, ciprofloxacin 97%, tobramycin 92%, amikacin 88%, piperacillin-tazobactam 76%, ceftazidime 73%, tixcarcillin-clavulanic 73%, imipenem 69%, meropenem 65%. The mean age of patients in group A is 33.79±11.18 years, that of group B is 29.35±9.5 years (p=0.2). During the 4-year observation period: the annual decline in FEV 1 (% predicted) was -1.3±2.3 per year in group A patient, and -1.8±2.7 per year in group B (p=0.6). In a longitudinal monitoring of pulmonary function of all patients attending our center, the mean annual FEV 1 decrease (±SD) was -1.8±4.9. The average number of days of hospitalization was 11.5±9.9 in MDR-PA-infected patients and 4±6.6 in patients infected with susceptible-PA infection (p=0.01). The average number of days of hospitalization in the Day Hospital (DH) regimen was 38±16.2 in MDR PA-infected patients and 32±14.9 in patients infected with susceptible-PA (p=0.2). The average number of oral antibiotic treatments was 9.1±3.9 in MDR PA-infected patients and 10±4.9 in patients infected with susceptible PA (p=0.5). The average number of intravenous antibiotic treatments was 5.4±2.8 in MDR PA-infected patients and 3.0±3.6 in patients infected with susceptible PA (p=0.03).

Conclusions: The MDR PA prevalence increased in the last four years. Aminoglycosides and fluoroquinolone, the commonly used antipseudomonal agents, had higher resistance rates. No substantial differences in days of DH admission, decline in lung function tests (FEV 1 ) and oral antibiotic therapy was found in patients infected with MDR PA strains. However there is a statistically significant higher difference in days spent in hospital and in use of intravenous antibiotics. Therefore strategies to prevent development and selection of bacteria resistance mechanisms should be evaluated. Complex polymicrobial communities are present in the respiratory tract of cystic fibrosis (CF) patients, but chronic Pseudomonas aeruginosa (Pa) infections account for the majority of morbidity and mortality.

Chronic rhinosinusitis (CRS) is highly prevalent; recent studies suggest the paranasal (PN) sinuses are a site of primary colonization and dissemination of Pa into the lower airways. Despite a robust host immune response and aggressive antimicrobial treatment, Pa adapts to these conditions via mutations that promote chronic lung infections. However, less is known about specific pressures in the PN sinuses or lungs that repeatedly select for these mutations across many patients. Our goal was to investigate evolution of Pa during CF CRS, and identify pathoadaptive traits associated with persistence of Pa in the PN sinuses, sinus microbiota changes, or several measures of clinical disease severity. In a 2-year longitudinal study of 33 CF adults, we collected sinus swabs for bacterial culture and microbiome analysis at quarterly clinic visits and during exacerbations. We genotyped 140 Pa isolates from 18 patients and screened for pathoadaptive phenotypes observed in Pa CF lung isolates, including biofilm matrix overproduction, loss of acute virulence factors, and antibiotic resistance. Most patients were colonized by one clonal lineage of Pa, but isolates displayed diverse pathoadaptive phenotypes, similar to reports of Pa in the CF lung. Whole genome sequencing of 4 patients' longitudinal isolates revealed that while specific mutations varied across all patients, all lineages acquired nonsynonymous mutations in iron acquisition, amino acid metabolism, interbacterial competition, antibiotic resistance, and global regulatory proteins, suggestive of mutational parallelism between patients similar to observations in the lung. Mutation rates were similar across all 4 patients' Pa lineages, but new mutations in phenotypically environmental Pa lineages did not become fixed in the population. In contrast, mutations in Pa lineages phenotypically resembling chronic CF lung isolates became fixed. To assess fitness following periods of putative pathoadaptation, we evaluated isolates collected early and late in the study for traits we hypothesized are important for persistence in PN sinuses: biofilm formation on CF airway epithelial cells (CFBE41o-) and competition against other Pa strains. Competition of established Pa with nonclonal, "invading" Pa isolates from the same patient Methodology: An Institutional Review Board-approved retrospective review of CF patients treated with vancomycin or linezolid between 01/01/2011 and 09/15/2017 was performed. Patients aged 6 to 21 years with a positive MRSA culture within 12 months of admission who received at least 10 days of treatment in the inpatient setting were included. Patients with a documented hypersensitivity to study medication, those unable to perform pulmonary function tests (PFTs) and those with a positive Mycobacterium or Burkholderia cepacia culture in the previous 12 months were excluded. The primary outcome was return to baseline FEV 1 percent predicted (FEV 1 %) within 2 percentage points. Baseline FEV1% was defined as the patient's personal best in the previous 12 months. Repeat encounters were accounted for in analyses.

Results: Forty-three patients (13.3 ± 3.7 years) met inclusion criteria; eleven patients treated with vancomycin in 36 encounters and 32 patients with linezolid in 141 encounters were analyzed. Propensity score weighting was used to balance the populations on covariates, and generalized estimating equations were used to weight multiple admissions for a single patient. Patients had a mean FEV1 of 89% with a mean deficit of 23.8% on admission. The difference in patients who returned to baseline FEV1% was not significant between groups (linezolid 34% vs vancomycin 22.2%, p=0.289). Cox regression demonstrated that after 6 days of treatment, patients receiving linezolid returned to baseline at a greater rate than those receiving vancomycin (HR 3.92; 95% CI 2.07-8.25, p=0.003). Adverse events were infrequent, including four cases of neutropenia and four cases of neuropathy in the linezolid group, and one case of renal injury in the vancomycin group.

Conclusion: Although the proportion of patients who returned to baseline FEV1% was not significantly different between groups, the patients treated with linezolid had an increased rate of return to baseline compared with patients treated with vancomycin after 6 days. Bacterial biofilms cause chronic infections of wounds, medical implants, and in the airways of patients with cystic fibrosis. These biofilms often exhibit antibiotic resistance that makes them medically difficult to treat. The ability to monitor target analytes within the structure of the biofilm is critical for understanding biofilm biology and its response to treatment. Oxygen consumption has been used as a measure of metabolic activity in facultative aerobes, but current methods such as microelectrodes limit monitoring to one dimension and cannot adequately capture dynamics. Biofilms are complex three-dimensional (3-D) structures that require higher resolution approaches to gain a more complete understanding of activity. Traditional measurements require physically moving the microelectrode, which is inherently disruptive to biofilm structure and obscures rapid dynamics within the system, which can potentially alter results. Here, we demonstrate optical oxygen-sensitive nanosensors that were used to measure 3-D oxygen gradients with minimal disruption to the biofilm structure in Pseudomonas aeruginosa biofilms grown from both laboratory strains and CF clinical isolates. We created and characterized optical nanosensors for detection of molecular oxygen which can be incorporated into the biofilm structure during growth. Using these, we obtained confocal microscopy data from which we can determine quantitative 3-D oxygen gradients.

Using this approach, we improved on traditional electrode-based 1-D methods of measuring oxygen profiles by investigating both the spatial and temporal variation in oxygen concentration during biofilm growth and under antibiotic attack. We observed spatial gradation in oxygen concentration during biofilm growth, attributed to nutrient consumption at the edges of the biofilm. We also studied 3-D oxygen gradients during antibiotic attack and found that oxygen was present at greater depths compared to untreated controls, consistent with cell death or a transition to anaerobic respiration. This new approach to biological interrogation provides higher resolution data, with improved temporal resolution, while minimizing the impact of the measurement on the biofilm itself. Increased resolution allows for better in vitro study and screening of antibiotic treatment relevant to treating P. aeruginosa biofilm infections.

Funding for this research provided by the Children's Hospital Colorado Research Institute, the Colorado School of Mines Technology Transfer Office, and startup funds from Colorado School of Mines. Persistent polymicrobial infections in individuals with cystic fibrosis (CF) are a significant cause of morbidity and mortality. Staphylococcus aureus and Pseudomonas aeruginosa are the most frequently isolated bacterial pathogens from CF children and adults, respectively. Notably, P. aeruginosa acquisition correlates with seasonal respiratory virus infections, and CF patients experience severe exacerbations during respiratory viral co-infection. We observed P. aeruginosa biofilm growth on CF airway epithelial cells (AECs) is enhanced during respiratory virus co-infection and hypothesized virus infection also alters microbial community dynamics in the CF airways. Using co-culture systems, we observed that cultured individually with CF AECs, P. aeruginosa and S. aureus each exhibit enhanced biofilm growth on RSV-infected cells. However, when S. aureus and P. aeruginosa are co-cultured, RSV co-infection leads to a dramatic reduction in S. aureus. We observe this trend with laboratory strains and clinical isolates collected from CF patients with chronic rhinosinusitis. Additionally, we observed P. aeruginosa exhibits enhanced production of the antimicrobial pyocyanin during RSV co-infection. Exogenous addition of pyocyanin at concentrations observed during P. aeruginosa-virus co-infection decreases S. aureus biofilm populations on CF AECs. When co-cultured with a P. aeruginosa pyocyanin mutant, S. aureus populations did not decrease during virus co-infection. We are currently investigating if this outcome is mediated by specific P. aeruginosa and host interferon-mediated antimicrobial mechanisms activated during respiratory virus infection. We observe RSV infection increases expression of the host interferon stimulated gene, IDO1, resulting in elevated levels of the secondary metabolite kynurenine. IDO1 controls tryptophan metabolism during infection via the kynurenine pathway and has antimicrobial effects towards a range of pathogens, and P. aeruginosa utilizes kynurenine to produce the quorumsensing signal PQS, which regulates pyocyanin production. Dual hostpathogen RNA sequencing to evaluate transcriptomic changes in CF AECs, P. aeruginosa and S. aureus during virus co-infection confirmed that in CF AECs, expression of IDO1 and SLC6A14, encoding the kynurenine transporter hATB 0+ , is increased during co-infection with all three pathogens. The P. aeruginosa genes pqsABCD and phzM responsible for production of the PQS signal and pyocyanin, respectively, also show increased expression, supporting our hypothesis that these pathways mediate host-pathogen crosstalk and antimicrobial activity during polymicrobial infection. This work will further our understanding of how virus co-infection alters interactions between bacterial pathogens and the host and offer insight on the role of viral-bacterial interactions in shaping host-associated polymicrobial communities.

This study was funded by CFF BOMBER14G0, NIH R01HL123771 and R61HL137077 (JMB) and CFF KIEDRO17F0 and CFF KIEDRO18F5 Postdoc to Faculty Transition Award (MRK).

Makhecha, S. 1 ; Bentley, S. 1 ; Balfour-Lynn, I. 2 1. Pharmacy, Royal Brompton Hospital, London, United Kingdom; 2. Paediatric, Respiratory, Royal Brompton Hospital, London, United Kingdom Introduction: Aminoglycoside-induced ototoxicity has been known to occur with long term therapy, by causing apoptosis of the inner and outer hair cells in the cochlea, generating free radicals, which can lead to irreversible loss of hearing and balance. The hearing loss caused by aminoglycosides affects high frequency tones. N-acetylcysteine (NAC), a thiol-containing antioxidant has been used to reduce ototoxicity caused by aminoglycosides in adults (1) . We therefore, introduced NAC orally to all children receiving amikacin for M. abcessus. Though NAC has low toxicity, it has been associated with nausea, vomiting, rhinorrhoea, pruritis and tachycardia (1) . The aim of this study was to assess the safety and tolerability of NAC in a small group of CF children on amikacin for M. abcessus.

Methods: A retrospective case note review was conducted over one year (April 2017 (April -2018 . Children were identified from pharmacy records and clinical data collected from case notes and discharge letters. Data on adverse events and discontinuation of treatment were collected.

Results: We identified 11 children who received NAC as prophylaxis for aminoglycoside ototoxicity, mean age 12 years (range 9-16 years). Doses were used in accordance with local guidelines with 5/11 children receiving 600 mg twice daily and 6/11 receiving 300 mg twice daily. No children discontinued therapy due to side effects, however, 2/11 did not like the formulation and were therefore changed to an alternative oral formulation. Additionally, no children were identified as having aminoglycoside-induced hearing loss. The median duration of therapy was 14.5 days (range 5-23 days).

Discussion: NAC was tolerated well in all children, however, a further study needs to be conducted to evaluate the most appropriate dose and formulation in children and efficacy on prevention of aminoglycoside otoxicity. We are now considering using it for all aminoglycoside courses. There is limited data supporting this with many citing a concern of resistance development. Utilization of multiple antibiotics can increase a patient's risk of developing acute kidney injury (AKI) or result in other adverse effects. In this study, we evaluated the efficacy of single versus double intravenous antibiotic coverage in pediatric patients with a CF exacerbation in the setting of chronic PsA infection.

Methods: This is a retrospective review of pediatric patients with CF admitted to Children's of Alabama for a PsA pulmonary exacerbation. Patients included were aged 0 to 21 years with a positive sputum culture of PsA between January 2015 to July 2017. Patients were cohorted based on treatment with one or two anti-PsA antibiotics. Patients were excluded if they were unable to perform spirometry. The primary endpoint is change in FEV 1 from admission to discharge. Secondary endpoints are change in weight from admission to discharge, incidence of neutropenia and AKI, and 30-day readmission rates. Statistical analysis was performed using Graphprism 2.0. IRB approval was obtained through the University of Alabama, Birmingham.

Results: There were 424 inpatient stays evaluated in this study with 69 (16%) in the single antibiotic group (group 1) and 355 (84%) in the double antibiotic group (group 2). The average length of stay was 9.71 days in group 1 and 11.35 days in group 2. The average change in FEV 1 was 9.511% in group 1 and 12.52% in group 2 (p=0.159). The mean weight change was 0.78 kg in group 1 and -0.27 kg in group 2 (p=0.75). The 30-day readmission rate was 11% for group 1 and 17% for group 2 (p=0.15). The AKI incidence for group 1 was 0.47% and group 2 was 8.96% (p=0.0426). The most commonly used antibiotic in group 1 was piperacillin/tazobactam (n=21). The two most commonly used antibiotics in group 2 were tobramycin (n=288) and piperacillin/tazobactam (n=113).

Conclusions: Based on this retrospective study, there was no statistical difference between lung function, nutritional status or time to readmission in single versus double antibiotic coverage in treating PsA infections in CF pediatric patients. There was a significant difference in the number of patients that developed AKI in the double coverage group. Our data suggests that this was more common with utilization of aminoglycosides in conjunction with another antibiotic class. This data supports the observation that treatment with single anti-PsA antibiotics may have similar efficacy while reducing the incidence of AKI in this population. Further research is needed to evaluate other longitudinal factors such as rate of lung function decline or development of antimicrobial resistance impacted by single antibiotic treatment of PsA in CF exacerbations. Introduction: Airway infections in cystic fibrosis (CF) with Pseudomonas aeruginosa (PA) are common, frequent, persistent, and once chronic, are rarely eradicated. As highly adaptive opportunists, PA in the CF population and individual, present with various phenotypes, virulence factors, and antibiotic tolerances. New or complementary therapeutic strategies are needed which can be effective across the diverse PA population and within the varied CF host environments. Our objective was to compare physiological, phenotypic, and structural characteristics of PA from a variety of sources for uniformity or heterogeneity of features which may impact the results of in vitro investigations of host-pathogen interactions and the potential for adjuvant therapeutic successes.

Methods: A collection of 18 PA was obtained from various CF and non-CF clinical and nonclinical sources. Each clinical isolate was examined for enzymatic activities, motility, mucoid character, production of fluorescent and visible pigments, and structural features, employing automated ID system, liquid and agar culture, fluorescence and electron microscopy, and a reporter eukaryotic cell culture model. The panel was screened for correlations between characteristics and carbohydrate binding, and effects on epithelial cells.

Results: As a PA collection, extensive heterogeneity was observed for this organism. The physiological flexibility of the species was evident for 12 of the 48 activity assays, with the pattern suggestive of mucoid isolates adaptively losing activities, with the exception of lipase which was more frequent among mucoid isolates. Phenotypic, structural and functional heterogeneity was observed in the presence of pigments, flagella, pili, motility, and mucoidy. Agar colony phenotypes and growth patterns were not always predictive of the presence or absence of flagella and/or pili as viewed by TEM. Binding studies indicating positive result, revealed broad ranges between strains of numbers of binding organisms (i.e. ~ 600 -3500 for specific fluorescent glycopolymer bound cells/microscopic field). Cell culture challenges with a panel of PA conditioned media yielded significant variation in dose-dependent effects on mammalian cell survival (range 22-79%, compared to 60% for virulence factor pyocyanin control), and on activation or shut down of stress and/or inflammation pathways. The sources, pigments, or structural features did not present consistent obvious correlations between attribute and effect.

Conclusion:P. aeruginosa from various sources can express widely varying traits, which are not necessarily predicted by their source, growth condition or phenotype. To reach the appropriate conclusions in vitro, hostpathogen investigations would benefit from verification, and quantitation where possible, of the actual features presented by the P. aeruginosa, as used in the study. 4 1. Paediatrics, University Hospital Limerick, Limerick, Ireland; 2. Teagasc Food Research Centre, Cork, Ireland; 3. Royal College of Surgeons, Dublin, Ireland; 4. Trinity College, Dublin, Ireland Introduction: Cross-sectional studies in individuals with CF suggest that the lung microbiome is altered early in life, but with significant variability between individuals (1, 2) . Sequential sputum sampling in adults has shown greater reductions in diversity in those with advancing lung disease compared to clinically stable individuals, with significant antibiotic effects seen. Our objective was to explore the effect of current and prior antibiotic exposure on the lower airway microbiome of young children with CF.

Methods: SHIELD CF is a longitudinal study based on an annual bronchoalveolar lavage (BAL) surveillance programme for children aged 1 to 6 years in 3 paediatric CF centres in Ireland (1). BAL is performed under GA via laryngeal mask and frozen at -80°C for future analysis. Extracted DNA was normalized and bacterial 16S rRNA V3 and V4 region amplified. Illumina overhang adaptor sequences were added and libraries prepared. Pooled libraries were assessed by Agilent high sensitivity DNA kit and quantified by qPCR using the Kapa Quantification kit for Illumina prior to sequencing using the Illumina MiSeq platform. Raw 16S rRNA gene sequencing reads were quality filtered using PRINSEQ. Denoising, OTU clustering (97% identity), and chimera removal were done using USearch (v7-64bit). OTUs were aligned using PyNAST. Alpha-diversity and beta-diversity were calculated using Qiime (1.8.0). Taxonomy was assigned using a BLAST search against the SILVA SSU 123 database. Further bioinformatics was performed with the phyloseq package in R.3.2.2. Antibiotic use on the day of the study and antibiotics taken in the preceding year were recorded.

Results: Data were available on 241 BALs from 106 children with CF with mean (SD) age at the time of BAL acquisition of 3.6 (1.8) years. At bronchoscopy 33 (13.7%) were on prophylactic flucloxacillin and 21 (8.7%) were on long term azithromycin, with neither affecting alpha or beta diversity. Flucloxacillin did not affect the relative abundance of Staphylococcus, or Pseudomonas at the genus level, but azithromycin resulted in lower relative abundance of Pseudomonas. Regarding prior antibiotic exposure; the number of courses of oral antibiotics taken in the previous year ranged from 0 to 12 (median 4 courses), resulting in a mean duration on oral antibiotics of 42 days (range 0 to 168). The number of admissions for IV antibiotics ranged from 0 to 6, with days on IV antibiotics ranging from 0 to 43. There was no correlation between total cumulative days on any (intravenous or oral) antibiotics and Shannon diversity (R-value 0.03, p-value 0.77).

Conclusion: Current or prior exposure to antibiotics in clinically stable preschool children with CF did not affect the alpha or beta diversity of the lower airway microbiome, however azithromycin exposure reduced relative abundance of Pseudomonas Recently, it has been demonstrated that patients may be chronically infected with multiple lineages of Pa epidemic clones. It is unclear how common Pa strain sharing and multiple lineage infection are in new-onset Pa infections occurring earlier in CF, when infections are aggressively treated with antibiotic eradication therapy (AET) and epidemic strains are infrequently encountered. We performed whole genome sequencing (WGS) of Pa isolated from sputum of children prior to initiation of inhaled AET, to determine the frequency of mixed infection and strain sharing, and its impact on AET failure.

Methods: Using Illumina technology, we sequenced genomes of 342 Pa isolates collected from 65 children with 75 episodes of new-onset Pa infection (episodes at least 1 year apart, with AET failure occurring in 27%) between 2012 and 2016. Up to 10 isolates, picked from multiple morphotypes in relative proportion to abundance in sputum, were sequenced per episode. An initial phylogenetic tree was constructed using the software Assembly and Alignment Free (AAF). Closely related strains were further investigated by mapping the Pa genomes to a closely related reference, using 2 pipelines: 1) Bacteria and Archaea Genome Analyser (BAGA) if a finished genome was available as reference; and 2) an in-house pipeline accommodating draft assemblies as reference. Maximum likelihood phylogenetic trees were generated from Single Nucleotide Polymorphisms (SNPs), and strain sharing, due to direct or indirect transmission, was inferred based on detection of appropriate topological signal in these trees (strains from different patients exhibiting paraphyletic relationships). Univariate logistic regressions were used to assess associations between mixed infection, strain sharing and AET failure. All statistical analyses were done using SAS 9.04.01 (SAS Institute, Cary NC, USA).

Results: A large number of patients shared Pa strains with other patients (N=25/65, 40%). The majority of patients with multiple episodes experienced strain replacement, rather than recurrence of the same strain (N=5/8, 62.5%). Mixed infection (two or more strains present in sputum sample concurrently) occurred in 12/75 episodes (16%). Having a mixed infection was significantly associated with sharing of Pa strains (unadjusted OR 10.7, 95% CI 2.2; 53.7, p <0.01) but was not associated with AET failure. Furthermore, strain sharing was not associated with AET failure.

Conclusions: A substantial proportion of patients with new-onset Pa infection were infected with a strain shared with other patients. Mixed lineage Pa infections were relatively frequently observed in new-onset episodes and were associated with strain sharing between patients; however, mixed infections and strain sharing were not associated with AET failure in this cohort. Nitric oxide (NO) production by host cells may play a role in innate microbe defense mechanisms. In preclinical models, antimicrobial activity of exogenous high-dose NO has been reported against a broad spectrum of bacterial species, including Mycobacterium smegmatis. In a recent prospective pilot study of inhaled NO in NTM-CF patients, it was shown improvement in lung function and reduction in sputum MAB and Pseudomonas aeruginosa bacterial load. In this study, we investigate whether high-dose NO exhibits antibacterial activity against multidrug-resistant MAB clinical strains in vitro.

Methods: A continuous horizontal-flow NO delivery system (Ghaffari A, et al. Nitric Oxide. 2005;14:21-9) was used to treat planktonic and immobilized (agar) bacterial cultures to specific doses of NO. Serially collected clinical strains including an outbreak strain were inoculated at 10 6 CFU/mL (2 mL) in saline, Middlebrook 7H9, or an artificial sputum media (Sriramulu DD, et al. J Med Microbiol. 2005 ;54:667-76) and treated with 250 ppm NO or medical air (control) for up to 10 hours. Time-kill assay was performed to assess NO activity against MAB. Samples were cultured on 7H11 agar every 2 hours followed by CFU analyses.

Results: We observed no growth of MAB B1 strain cultured in saline after 6.0-hour and 2.0-hour treatments at 250 ppm and 400 ppm NO, respectively. In artificial sputum, time-kill curves for MAB B1, B5, and B8 clinical strains showed 4.2-Log 10 , 4.0-Log 10 , and 2.8-Log 10 reductions, respectively, following 10-hour treatment with 250 ppm NO compared to controls. All controls were treated with medical air only inside the same NO delivery system simultaneously and remained fully viable. In addition, we controlled for the minute effect of NO on media pH by growing MAB in artificial sputum media (pH 5.0). Viability curves did not show a significant difference in MAB B1, B5, and B8 growth patterns between pH 7 and pH 5.

Conclusion: These findings demonstrate the antibacterial activity of exogenous NO against multidrug-resistant MAB clinical isolates in vitro. These data suggest further evaluation of inhaled high-dose NO therapy as a potential anti-mycobacterial agent in NTM lung infections is warranted. Acknowledgement: We thank the AIT Therapeutics team for providing the in vitro NO delivery system used in this study.

Background: Cystic fibrosis (CF) is an inherited multi-organ disease with a reduced survival. One important risk factor for reduced survival is chronic pulmonary infection with Pseudomonas aeruginosa (PSA). From first detection to chronic infection with PSA exists a window of opportunity to eradicate primary or new colonization of PSA in the respiratory system. Therefore the German Society of Pediatric Pneumology published in July 2013 a guideline related to PSA eradication in patients with CF. The guideline is not legally binding. In 2017 we asked how well accepted the guideline is in daily clinical practice.

Method: A survey with 15 questions regarding PSA screening and eradication regime was emailed to all CF centers (CFC) in Germany. For the evaluation the questions were categorized as diagnostic, initial therapy, control of success and rescue treatment. The survey could be answered online or be printed, filled out and send back via fax.

Results: The survey was answered by 29 out of 90 CFC. They were responsible for 2090 CF patients of whom 1154 are children. This covers slightly more than one-third of all German CF patients and 45% of all CF patients <18 years. 11 CFC treat <50 patients and 18 treat >50 patients. 4 CFC just deal with children, 3 only with adults and 22 with both. For diagnosis 5 (17%) clinics deviate from the guideline: they started therapy after the second PSA positive test, instead of right after the first positive test. The initial therapy in 9 (31%) CFC differed from guideline. 7 (24%) of them used a combination of oral ciprofloxacin and inhaled tobramycin, while the guideline proposed either a) the combination of oral ciprofloxacin with inhaled colistin or b) monotherapy with inhaled tobramycin. Treatment a) was favored by 12 (41%), while treatment b) was favored by 6 (21%); 2 (7%) CFC used both therapies on a regular basis. The guideline was again ignored by 15 (52%) CFC regarding the way treatment success was monitored. The time span of PSA testing was too short in 14 (48%) CFC, 5 (17%) CFC did not do a PSA-antibody check-up 3 months after the end of the eradication treatment, 7 (24%) did not use antibody testing regularly. Five (17%) CFC had not stated any rescue therapy. All in all, 21 (72%) CFC deviate from guideline in at least one category of PSA eradication recommendations. For 63% of all CF patients PSA eradication regimen was at least in one part different from the guideline. The percentage of larger CFC (>50 CF patients) which followed the guideline was low but still higher (33% vs 18%) compared to smaller CFC.

Discussion: Deviations from guideline recommendations are frequent in all aspects of PSA eradication. It may contribute to the low acceptance rate, that the guideline is a recommendation based on just a few studies. One reason for the low acceptance rate by the smaller CFC could be a lack of frequent personal experience with PSA eradication. The low acceptance of this guideline, which is based mainly on personal experience and not on randomized control trials (RCTs), stresses the need for RCTs in this clinically important area to improve the evidence of recommendations and thereby their acceptance. At the end improving the quality of guidelines will improve the quality of clinical care for CF patients. The M. abscessus complex is composed of three species, M. abscessus, M. bolletii, and M. massiliense, all of which are difficult to treat due to antibiotic resistance. Several studies have found that point mutations occur in phoR during chronic infection in CF patients. The genes phoR with phoP compose a putative two-component system whose function in M. abscessus complex is unknown, although homologs in Mycobacterium tuberculosis are required for virulence and survival in macrophages. We hypothesized that phoRP is critical for M. abscessus complex infection and that point mutations that occur during infection make the bacteria more fit in the host by modulating pho activity.

Methods: We compared the ability of clinical M. abscessus complex isolates to invade and survive within THP-1-derived macrophages and used whole-genome sequencing to identify mutations between the isolates. We constructed a phoRP deletion mutant in M. abscessus ATCC 19977 and compared its ability to survive within THP-1-derived macrophages. We also measured the intracellular survival of the phoRP deletion mutants within CF-like macrophages generated by treating THP-1-derived macrophages with a CFTR inhibitor (CFTR(inh)172).

Results: We identified one pair of M. massiliense isolates from the same patient, 8 months apart, that differed by only 14 point mutations, one of which was in phoR. The later isolate of this pair, NTM0633, which had a phoR point mutation, was significantly better able to survive within THP-1-dervived macrophages than the earlier isolate, NTM0295. The M. abscessus ATCC 19977 phoRP deletion mutant was less able to survive within THP-1-dervived macrophages compared to parental strains after 2 days, demonstrating the importance of this pathway for the virulence of M. abscessus. The decreased ability of the phoRP deletion mutant to survive within macrophages was also observed in CF-like macrophages.

Conclusions:M. abscessus complex phoRP is a two-component system that is important for bacterial survival within macrophages (both CF and non-CF). These findings suggest that modulation of the phoRP pathway allows M. abscessus complex to adapt to the host and that the phoRP pathway could be a novel therapeutic target. Acknowledgment: Supported by Department of Anesthesiology, Critical Care and Pain Medicine, Boston Children's Hospital. Patients with the genetic disease cystic fibrosis (CF) are often chronically infected by the opportunistic pathogen Pseudomonas aeruginosa, leading to significant morbidity. P. aeruginosa virulence is largely regulated through a cell-cell signaling mechanism termed quorum sensing (QS). Typical QS consists of a LuxI-type signal synthase producing a diffusible acyl-homoserine lactone (AHL) signal, which binds a cognate LuxR-type receptor-regulator that activates transcription of target genes. In laboratory strains and conditions, P. aeruginosa employs two AHL synthase/receptor pairs arranged in a hierarchy, with the LasI/R system controlling the RhlI/R system and many downstream virulence factors. Cryptically, P. aeruginosa isolates with inactivating mutations in LasR are frequently isolated from chronic CF infections and are associated with attenuated virulence phenotypes. Our recent work has shown that many of these chronic isolates still employ AHL QS using only the RhlI/R system. Methods: We examined the scope and dynamics of Las-independent Rhl QS in P. aeruginosa using a global transcriptomics approach complemented with phenotypic and biochemical analysis. We assembled a cohort of chronic CF isolates with wild-type lasR, variant lasR, and lasR-null alleles to investigate the regulatory shift from Las to Rhl QS in isolates that span the spectrum of QS genotypes characteristic of CF infection. We transcriptionally profiled each isolate at low-and high-cell density with either added AHL signal or a signal-degrading enzyme to compare QS-ON and QS-OFF states. We introduced fluorescence-based transcriptional gene-fusion reporter plasmids into these isolates to profile the induction kinetics of QS genes dynamically throughout growth.

Results and Conclusions: We analyzed disparate QS regulatory regimes in P. aeruginosa, with the hypothesis that there are specific patterns of adaptation that allow the chronic CF infection lifestyle. Our global analysis of a cohort of chronic CF isolates found radical rewiring of QS. Both timing and magnitude of canonical QS-controlled genes was changed in CF isolates compared to the laboratory wild-type strain PAO1. Furthermore, the hierarchy of Las and Rhl QS appears restructured, with lasR-null isolates still able to activate traditionally Las-regulated genes. Our results implicate Rhl instead of Las as a possible therapeutic target for the disruption of QS-controlled virulence in chronic (not acute) infections. Our work lays a foundation for development of alternative therapeutic strategies directed at P. aeruginosa from chronic CF infections, as well as an improved fundamental understanding of patho-adaptation in QS. Methods: Prospective data on patients undergoing treatment for NTM was collected as a part of each individual's IND with an IRB-approved protocol. These data included forced expiratory volume in 1 second percent predicted (FEV1), body mass index percentile (BMI), and frequency of admissions per year. Bacterial and acid fast bacilli (AFB) cultures were obtained quarterly and safety data including blood work and vital signs monthly. Based on guidelines, our treatment strategy was to provide 3-drug initiation including intravenous (IV) antibiotics for 8-12 weeks. Specific antibiotics were chosen based on susceptibilities. IV amikacin and oral azithromycin were used in all cases. Oral regimens followed for 10-12 months with clofazimine, azithromycin and inhaled amikacin.

Results: In 2017, our clinic cared for 150 pediatric patients (age 2 weeks-20 years). AFB was positive in 17 patients (incidence of 11%) and 9 were Mycobacterium avium complex (MAC) and 8 were M. abscessus. Treatment was not needed for MAC positive cultures as they were only grown <2 times. M. abscessus grew <2 times in 5 patients and 3 were already on treatment from 2016. Of the five patients undergoing treatment for M. abscessus in 2017, treatment with IV tigecycline was used in 3 and IV imipenem in 2. Patients were switched to inhaled amikacin and oral clofazimine around day 68 (range 56-77 days). All 5 have completed 1 year of total treatment (range 365-433 days). Only 1 patient has cleared M. abscessus for a year. Indication for initiation of treatment in all was a decline in FEV1 from 1 year prior (range: -11 to -3%). Four had further decline in FEV1 over the treatment course. Decline in BMI from the year prior was noted in 4 patients and 3 of these patients had further decline during treatment. The frequency of admissions was reduced from an average of 3/year in the year prior to treatment to 1.8/year. There were no adverse events or safety issues related to use of clofazimine.

Conclusions: Our data suggest that clofazimine use is safe and tolerable for patients with M. abscessus. However, the efficacy is unknown as our patients did not clear NTM. There was a benefit in reduction in admissions during the year of treatment. Further recommendations for treatment initiation, duration and prolonged clofazimine use within the pediatric population is needed. Results: The LCI in patients infected with Aspergillus fumigatus was significantly (p <0.05) higher than in patients with normal respiratory flora. There were also statistically significant differences in LCI measured in patients with chronic Pseudomonas aeruginosa infection and those with first Pseudomonas aeruginosa infection (p <0.05). Futhermore, significant statistical differences (p<0,05) were found between two groups of patients (FEV1> 70% predicted and FEV1 <70% predicted) with chronic Pseudomonas aeruginosa infection. Analyses have shown significant statistical differences (p<0,05) between patients with normal respiratory flora and intermittent as well as chronic Pseudomonas aeruginosa infection.

Conclusions: LCI results are associated with microbiological status in CF. Children with Aspergillus fumigatus as well as chronic Pseudomonas aeruginosa airway infection usually had higher LCI. It suggests that LCI increases more in children with airway infection. Early eradication of pathological flora positively affects the maintenance of lower LCI.

Background: There is emerging evidence for the role of posaconazole in the management of Aspergillus-related CF lung disease. The tolerability and efficacy of posaconazole in paediatric CF patients is not well established.

Methods: We report a prospective study over a fifty-one month period from February 2014 to May 2018 evaluating the safety, tolerability and efficacy of posaconazole in children with CF.

Results: Fifteen children (6 males, median age 11.6 years, range 6-17 years) received a total of twenty-one courses of posaconazole for the duration of four to twelve weeks. Of these patient episodes, five received posaconazole following an incomplete response to corticosteroids for emerging or active ABPA. Six patient episodes required a combination of corticosteroids and posaconazole for difficult-to-treat ABPA. A subgroup of patients (n=10) with persistent isolates of Aspergillus fumigatus in the sputum, in the absence of serological markers of ABPA, received posaconazole monotherapy for pulmonary exacerbations not responding to conventional broad-spectrum antibiotic treatment. Children under 12 years and less than 40 kg received a dose of posaconazole liquid of 18-24 mg/kg/day in two divided doses. Older children were prescribed posaconazole as per manufacturer's recommended dose. Posaconazole was well tolerated. Rash and gastrointestinal upset occurred in one child who went on to tolerate a subsequent course omitting the loading dose. A second child failed to tolerate posaconazole on two separate occasions experiencing flushing and influenza-like symptoms. Both these children were over 16 years of age. Full blood count, electrolytes and liver function were monitored monthly and remained within normal limits for all patients. Posaconazole blood levels were available for sixteen patient episodes, of which 81% achieved a therapeutic level of >1.25 mg/L. There was a modest but significant improvement in FEV1 (percent predicted) demonstrated for the cohort as a whole with a mean change of 7% (-14 to 26%) following posaconazole therapy.

Conclusion: Posaconazole is well tolerated in children as young as six years old with CF, therapeutic blood levels are readily achieved in those who adhere to therapy, and improvements in lung function are observed. We identified that triclosan, an FDA-approved antibacterial, works synergistically with tobramycin to increase killing of Pa cells within mature biofilms. The combination eradicated greater than 99% of the strain PAO1 in vitro even though neither treatment was effective on its own (Maiden M, et al. Antimicrob Agents Chemother. 2018 May 25;62(6)). Tobramycin/ triclosan was effective against 6/7 CF mucoid Pa clinical isolates tested, including a tobramycin-resistant strain that overexpresses the MexXY efflux pump.

To understand the mechanism of action of tobramycin/triclosan, targeted experiments ruled out a number of potential mechanisms based on the known activities of triclosan and tobramycin. To elucidate the mechanism(s) of tobramycin/triclosan synergy, we evolved mutant Pa that are resistant to tobramycin/triclosan by serially passaging biofilms. To accomplish this, biofilms were grown on MBEC™ pegs for 24 hours and then treated with tobramycin/triclosan for 24 hours. Viable cells were recovered and allowed to form new biofilms for another cycle of treatment at ever-increasing concentrations of tobramycin/triclosan. By gradually increasing the selection pressure, we evolved 191 single colony isolates that are ~200x more resistant tobramycin/triclosan.

To determine the molecular mechanism of resistance, Illumina HiSeq was used to determine whole genome sequences of six resistant isolates. A single nucleotide polymorphism in the fusA1gene was identified in 6/6 mutants. One of these mutations, L40Q, has recently been independently identified and verified in several clinical Pa isolates as providing resistance to tobramycin (Bolard A, et al. Antimicrob Agents Chemother. 2018 Jan 25;62 (2)). The fusA1 gene encodes for elongation factor g, which is essential for protein translation, although it is not the direct target of tobramycin. Current work is focused on elucidating how this mutation renders Pa resistant to tobramycin, making the combination ineffective, and defining the mechanisms of synergy. This understanding will be important for the clinical development of tobramycin/triclosan to treat CF patients. Methods: We performed WGS on 521 NTM isolates from 282 US patients, including those from the recently expanded multisite CF PATIENCE and PREDICT clinical trials. Sequence reads for each isolate were mapped to the designated reference genome and single nucleotide polymorphisms (SNPs) were identified, then used to estimate genomic relatedness among isolates. The genomic relationships of NTM isolates within and between US CF patients were investigated to assess: i) clonality and possible transmission of isolates between US CF patients; ii) the presence of multiple strains of a species within a patient; iii) the presence of multiple species per patient.

Results: We found that 58% of patients with M. abscessus subsp. abscessus infections had isolates that belong to two previously reported dominant clones (DC) and 56% of patients with M. abscessus subsp. massiliense had isolates in two other DCs. Furthermore, we identified clusters, composed of isolates from two or more patients within empirically defined SNP thresholds, in M. abscessus group (n= 21 clusters; 70/192 patients = 36%), M. chimaera (n=3 clusters; 7/17 patients = 41%) and M. intracellulare (n=1 cluster; 2/33 patients = 6%). No M. avium clones or clusters were observed within US CF patients. Of 101 patients with 2 or more isolates, 14 patients (~14%) cultured both M. abscessus group and M. avium complex (MAC) isolates. In patients with multiple M. abscessus isolates, we typically observed clonality; however, six patients (6/72 = 8%) cultured both subspecies abscessus and massiliense. We observed multiple strains of MAC species isolated in 32% (9/28) in patients over time. The overall mean SNPs within MAC species (µ avium =5867 SNPs; µ chimaera = 16900 SNPs; µ intracellulare = 25799 SNPs) was more than observed in M. abscessus (µ abscessus = 4900 SNPs; µ bolletii = 12934 SNPs; µ massiliense = 25642 SNPs), which suggests different microbial demographics among the NTM infecting the US CF patient population.

Conclusions: We observed both clonality and clusters of M. abscessus, M. chimaera and M. intracellulare among CF patients. WGS identified M. abscessus infections are more likely to be clonal over time compared to MAC infections, in which 32% of patients have mixed genotypes. The hypothesis of independent environmental acquisition of NTM is most consistent with the genetic variation observed in MAC. This contrast between clonality of M. abscessus and heterogeneity of MAC suggests that they have different ecological niches and/or modes of transmission. Our results also emphasize the need for WGS in investigations of NTM infection and treatment in the CF population. Introduction: Pulmonary infections are associated with airway inflammation in children with CF. These interactions drive the progression of structural lung damage and lung function decline. P. aeruginosa and S. aureus are the predominant organisms that have been associated with the pathogenesis of CF lung disease, the significance of other infections is less certain.

Objective: To evaluate the longitudinal association of respiratory infections with the development of structural lung disease in young children with CF.

Methods: Infants diagnosed with CF after newborn screening, participating in the AREST CF cohort study were included. Participants prospectively underwent chest CT scans and bronchoalveolar lavage (BAL) at 3-6 months,1 year and annually up to 6 years of age. The prevalence of lower airways pathogens and the association with CF-CT scores and neutrophilic inflammation (neutrophil elastase activity and IL-8 levels) was investigated longitudinally, accounting for repeated measurements, pancreatic insufficiency, age, gender, homozygosity for ΔF508 and co-infections.

Results: A total of 262 preschool children underwent 925 BALs and chest CTs. P. aeruginosa, Aspergillus species, H. influenzae and S. aureus, S. pneumoniae and M. catarrhalis were the most commonly cultured organisms. P. aeruginosa, Aspergillus species, H. influenzae and S. aureus were significantly associated with markers of inflammation as well worse CF-CT scores (Table) . Each of these pathogens had different associations with the different CF-CT subscores. P. aeruginosa was most significantly associated with bronchiectasis (p<0.001) and Aspergillus species was most significantly associated with trapped air (p<0.001).

Conclusion:P. aeruginosa, Aspergillus, S. aureus and H. influenzae are associated with neutrophilic inflammation and development of structural lung disease. P. aeruginosa is the pathogen most associated with the development of bronchiectasis whereases Aspergillus is the pathogen most associated with the development of trapped air. Further research is required to understand the cause and effect of these associations and to determine the need for eradication protocols also for Aspergillus, H. influenzae and S. aureus.

BE -bronchiectasis, BWT -bronchial wall thickening, TA -trapped air, MP -mucus plugging, Diff -mean difference, CI -confidence interval. Introduction: Children with cystic fibrosis (CF) who become infected with Pseudomonas aeruginosa (PA) are treated with antibiotics, but even with active and prolonged therapy not all children eradicate the infection. Failure to eradicate PA infection (hereafter referred to as "persistence") leads to chronic PA infection and poor clinical outcomes. We hypothesized that differences in the accessory genomes of PA strains influenced their persistence in CF. We sought to develop a machine learning model based on the presence or absence of specific PA accessory genomic elements (AGEs) to predict which children would develop persistent infections and could therefore be targeted for more aggressive treatments.

Methods: Whole-genome sequencing was performed on all PA isolates collected during the Early Pseudomonas Infection Control Clinical Trial (EPIC CT). Baseline PA isolates (collected prior to initiation of anti-PA antibiotic therapy) from 186 newly infected patients in the EPIC CT were categorized as causing a persistent vs. eradicated infection based on whole-genome sequence results of follow-up cultures collected as part of the trial. The programs Spine, AGEnt, and ClustAGE were used to catalog the AGEs present in each isolate. A random forest machine learning approach was utilized to create a prediction model of eradication based on the presence or absence of AGEs among the baseline isolates.

Results: Of the 203 baseline PA isolates from 186 patients, 146 were eradicated and 57 were persistent. A random forest machine learning model was generated using AGEs as features and persistence vs. eradication as outputs. This model was based on 13 discriminatory AGEs and had an internal accuracy of 77%, a sensitivity of 87%, a specificity of 52%, a PPV of 85%, and an NPV of 61%. Of the 13 AGEs, 4 were highly associated with persistence (p < 0.05) and 4 were highly associated with eradication (p < 0.05). As the initial data had relatively few persistent isolates, we employed ROSE, a data balancing method that draws artificial samples from the feature space neighborhood around the minority class. Balancing of the data resulted in generation of an improved model based on 81 discriminatory AGEs with an internal accuracy of 92%, a specificity of 92%, a sensitivity of 93%, a PPV of 92%, and an NPV of 92%.

Conclusions: A machine learning model based on PA AGEs had a high level of internal accuracy and has the potential to identify strains that are more likely to cause persistent PA infections in individuals with CF. We next plan to validate this model using an independent cohort of newly infected CF patients.

Acknowledgements: Supported by CFF HAUSER1510, SINGH15UO and NIH R21 HL129930, RO1 AI118257. Thank you to the TDN participating sites and patients participating in the EPIC trial. Purpose: Management of paranasal sinuses complications of cystic fibrosis (CF) is a challenge for the otolaryngologist. Paranasal sinuses system represents a P. aeruginosa reservoir affecting lower airways system. Early diagnosis and follow-up of ENT complications of CF, like mucoceles and polyposis, are important in order to treat them. A clinical, symptoms, radiological score for treatment, stratification and evaluation of CF patients has been developed.

Method and Materials: Between May 2015 and December 2017, 40 patients attending our ENT CF clinic were evaluated for paranasal sinuses complications with maxillo-facial CT (18 males, 22 females; mean age16.80 years, range 6-30 years). Radiological results were described qualitatively and quantitatively using Lund-Mackay score. CT report with the radiological score, endoscopic Meltzer's Score and SNAQ-11 questionnaire for clinical ENT symptom were used for the assessment of a CF sinus score (CFSS), with a maximum score of 112 points. ENT treatment was chosen depending on CFSS results.

Results: CT guided Lund-Mackay score: mean result 18.25 (range 6-24). By CFSS findings, 15 patients were addressed to endoscopic sinus surgery (mean result 65.5; range 32-75); the remaining patients were addressed to follow-up (mean result 25.8; range 14-42). Most relevant radiological findings were pluri-sinonasal disease, medialization of maxillary sinuses walls, mucocele, and bone resorption. At endoscopic evaluation polyposis was the most frequent finding (60%); finally, nasal obstruction and rhinorrhea were registered as the most referred symptoms (45% and 25% each).

The need for an objective quantification of radiological, clinical, symptoms assessment of patients with ENT complications of CF has prompted the development of a clinical score, currently in its second year of testing, CFSS. In the future it could become a mean for ENT stratification treatment and for comparison among ENT specialists. (2005) (2006) (2007) (2008) (2009) (2010) (2011) (2012) (2013) (2014) (2015) (2016) set out to identify rapid growing mycobacteria (RGM) isolated from sputum of cystic fibrosis (CF) patients attending the CF Centre of Milan. The study aims to determine the prevalence of the different subspecies of Mycobacterium abscessus complex (MABSC) in chronically colonized patients, the possible persistence of the same strain throughout the colonization period and to evaluate the correlation between mutations responsible for macrolidesand amikacin-resistance and the related minimum inhibitory concentration (MIC).

Methods: Out of the 892 patients followed at our centre, 150 were positive for RGM at least once during the study period (99 intermittently and 51 chronically colonized). One strain a year per patient was identified both by MALDI-TOF MS and then by rpoB gene sequencing.

In 314 MABSC strains isolated from the 51 chronic patients, the erm gene was sequenced to evaluate inducible resistance to clarithromycin. Furthermore, 23S and 16S region were sequenced to investigate constitutive resistence to macrolides and aminoglycosides, respectively.

Broth microdilution method according to CLSI (2011-M24) was carried out to evaluate antibiotic susceptibility and inducible resistance to clarithromycin.

Moreover one strain per year from 34 chronic patients was studied by whole genome sequencing (WGS).

Results: MALDI-TOF MS identified RGM strains at species level, in agreement with rpoB gene sequencing results: 51 patients were chronically colonized by MABSC. Out of the 99 intermittent patients, 47 were colonized by MABSC, 40 by M. chelonae and 7 by others RGM.

According to rpoB gene sequencing, within the 51 chronic patients M.a. abscessus accounted for 70%, M.a. bolletii for 17% and M.a. massiliense for 13%.

Microdilution tests detected amikacin susceptibility in 312 MABSC strains; mutation of the 16S rRNA gene was highlighted in the remaining 2 strains. 23S rRNA analysis did not show constitutive resistance to macrolides in any strains. Resistance to clarithromycin at day 14 was found in 234 strains (74.5%). Out of the 80 susceptible strains, 41 (M.a. massiliense) presented truncated erm gene, 16 (M.a. abscessus) presented a mutation in T28->C and 23 (M.a. abscessus) also presented truncated gene erm, therefore the gene was inactive.

WGS data showed 28 patients were chronically colonized by the same strain, whereas in 6 patients a subspecies switch was highlighted.

Conclusions: MALDI-TOF MS represents a useful tool for routine clinical diagnostic procedures, as it allows a quick, simple, reliable identification at species level. Molecular techniques remain mandatory to define subspecies.

Genotype analysis shows that chronic infections are more frequently supported by M.a. abscessus.

The majority of MABSC isolates exhibit resistance to clarithromycin. Clarithromycin susceptibility of all M.a. massiliense strains is confirmed by the presence of the truncated erm gene. Data indicate the importance of molecular analyses. as different responses to pharmacologic treatment may be explained by inducible resistance to clarithromycin. 

Objectives: Chronic lung infection is the major cause of morbidity and mortality in cystic fibrosis patients, so an improved ability to detect specific infecting organisms would allow more directed treatment and potentially improve CF patient outcomes. Routine culture-based methods, employed in the clinical detection and identification of CF-associated infections, frequently fail to detect fastidious, anaerobic and slow-growing bacteria, viruses and fungi. Our aim was to develop the first molecular assay that provides comprehensive detection of infectious organisms (bacteria, viruses and fungi) found in the complex polymicrobial community of CF lungs. The proposed assay would provide accurate, specific and rapid delivery of results for both infection detection during exacerbations and for routine surveillance of CF airways.

Methods: Genus-and species-specific primer sets were designed to amplify each of the targets, employing novel, highly multiplexing RT-PCR technology. Amplified products were subsequently hybridised to a biochip array, which facilitated spatial discrimination of individual target organisms, allowing easy interpretation of the infection profile of the test sample. A cohort of CF sputum samples has been collected across the UK, Ireland and US, each with full microbiome analysis, performed through NGS. A subset has also been forwarded for routine culture. This cohort provides comparative data against which to assess performance of the array.

Results: We have developed a highly multiplexed molecular diagnostic array that can detect up to 135 species and one antibiotic resistance marker associated with CF infection. The assay can detect 117 bacterial targets, including aerobic, anaerobic, gram-positive and gram-negative species, 8 respiratory viruses, 10 fungal species and the MecA gene. All assay components, including the RT-PCR reagents are stable at room temperature (18 -25°C), with only one RT-PCR reaction and one biochip array necessary for all infection targets. Assay time from nucleic acid to data interpretation is under 3½ hours.

We have developed a multiplex cystic fibrosis infection detection array, capable of rapid detection of viruses, fungi, and both aerobic and anaerobic bacteria. This has the potential to significantly improve routine infection screening of CF patients and accelerate identification of infections associated with exacerbations, with the prospect of improving patient care pathways and outcomes. In addition, the comprehensive nature of the array provides opportunities for non-CF patients to avail of improved lung infection detection, such as those with other chronic lung conditions, including COPD. On a broader theme, more accurate and timely identification of specific infecting organisms can lead to improved antibiotic stewardship, slowing the advance of widespread antibiotic resistance. Acknowledgments: Work supported by Innovate UK and Randox Laboratories Ltd.

cystic fibrosis (CF). Sodium nitrite damages essential iron-sulfur clusters in bacterial proteins, which are repaired with available iron supplies. Ga(III) is an iron mimic that, when available, replaces the iron in these clusters causing bacterial growth arrest and death.

Objective: We tested the hypothesis that Ga(III) and nitrite would have synergistic antimicrobial activity against P. aeruginosa.Methods and Results: First, we determined if nitrite and Ga(III) were synergistic against a range of CF clinical isolates grown aerobically on M9 agar plates using a checkerboard design. The combination has been found to be synergistic aerobically in 11/16 P. aeruginosa clinical isolates, as well as the lab strains PAO1 and PA14 (synergy defined by an FIC ≤0.5). Anaerobically, synergy was tested for PAO1 and PA14 using checkerboard plates with 1% KNO 3 as a terminal electron acceptor. Anaerobically, the Ga(III) MIC for PA01 was 100-200 mM and the MIC for PAO14 was 800-1600 mM. The anaerobic combination MIC for 1.25 mM nitrite was 25 mM for both PAO1 and PA14. In liquid culture, the Ga(III)-nitrite interaction was iron dependent, as increasing the amount of iron in the media blocked the effect. Because bacteria grow in biofilms or aggregates in chronic lung infections, we next tested if Ga(III) and nitrite were synergistic in two biofilm models. Abiotic biofilms were grown on microtitre plates for 24 hours in the presence of Ga(III) and nitrite in a checkerboard design. In the biofilm model, Ga(III) MIC for PAO1 was 313 mM and nitrite MIC was 14 mM. The combination MIC for PAO1 was 3 mM nitrite and 78 mM Ga(III), giving a synergistic FIC of 0.46. Finally, we used a biotic biofilm assay where PAO1 biofilms were grown on the apical surface of human airway epithelial cells (CFBE41o-). Against mature biofilms, nitrite in the assay was bacteriostatic. Ga(III) was bactericidal at concentrations greater than 2.5 mM, with killing to below the limit of detection by 10mM. Addition of nitrite caused additional antibiofilm activity (p<0.01 by one-way ANOVA).

Conclusion: Ga(III) and nitrite have synergistic antimicrobial activity that is iron dependent. Combining a compound that attacks Fe-S clusters with an iron mimetic is a novel conceptual approach targeting core bacterial metabolism.

Acknowledgments: ACZ: CFF ZEMKE16Q0, NHLBI K23HL131930 Background: Left untreated, new infection with P. aeruginosa (Pa) becomes chronic, which is associated with worsening lung function, inadequate nutrition, more pulmonary exacerbations and higher mortality. Infection with Pa has been shown to be of great concern to parents and patients with cystic fibrosis (CF).

Aim: The aim of the study was to evaluate the clinical and microbiological outcome of Pa eradication treatment in patients diagnosed through neonatal screening.

Methods: This is a retrospective study of patients diagnosed through neonatal screening between 1984 and 2017. All patients underwent microbiological surveillance every 3 months with oropharyngeal swabs or sputum. From 1993 until today all patients diagnosed with Pa underwent eradication treatment. We retrospectively applied the definition of sustained/ nonsustained eradicators by calculating the time to appearance of mucoid phenotypes and the FEV 1 trend in these two categories (Mayer-Hamblett N, et al. Clin Infect Dis. 2015; 61:707-15 ). Patients' microbiological status, regarding Pa colonization, was evaluated using the Leeds definition.

Results: The first Pa infection was observed in 118 (77%) of 153 patients diagnosed through neonatal screening. Age at first infection (mean±SD) was 3.6±4.9 years in males, 4.5±5.4 years in females (p=0.3). Mean age (±SD) at first acquisition was 3±3.7 years in 66 patients with 2 alleles with minimal dysfunction and 5±6 years in 19 patients with at least 1 one allele with residual function (p=0.07). Mucoid phenotype was detected in 8 (7%) out of 116 patients at first infection. No substantial pattern of antibiotic resistance to tobramycin, ciprofloxacin or colistin was detected. From 1993 to 2017, 109 patients were treated with early eradication treatment. Thirteen (27%) out of 48 nonsustained eradicators and 6 (9.8%) out of 61 patients with sustained eradication developed over time Pa chronic infection (OR=3.46). The time to mucoid state (mean±SD) was 9.7±4.1 years in 12 patients with sustained eradication and 5±4.7 in 21 nonsustained eradicators (p = 0.02). In the same time period the FEV 1 decline per year was -2±3.7% of predicted in sustained and -2.2±4.4 in nonsustained eradicators (p=0.9). At the present, 73% of all our patients diagnosed through neonatal screening who underwent early eradication treatment are Pa-free, 11% are Pa-intermittent and 16% have chronic infection (mean 14±8.3, median 12.2, range 1.3-34.2 years).

Conclusions: CF patients acquire Pa infection in their earliest infancy, but the mean acquisition age is different in those with serious mutations compared to those with at least one allele with residual function. In the absence of a gold standard which defines eradication, the use of "sustained eradicator" allows us to identify a category of patients with reduced risk of developing chronic infection and longer time to developing mucoid status. In our cohort of patients diagnosed through neonatal screening, our careful microbiological surveillance and early eradication treatment have resulted in reduced number of patients with chronic Pa infection. 

Objective: Aspergillus fumigatus (asp) is the most common detected filamentous fungus in CF airways. But still little is known about the exact aspergillosis entities in CF. Therefore, a study to evaluate patients with CF regarding their different Aspergillus entities, such as asp bronchitis, -infection, -colonization, -sensitization and ABPA, was established in 2016.

Methods: Patient with CF producing sputum and with the ability to perform pulmonary function testing were included into a prospective, longitudinal study (DRKS00012862). Total serum IgE, asp-specific IgE, asp IgG, sputum galactomannan (GM), sputum qPCR, sputum culture, asp-specific T cells response (ARTE), and clinical data were measured on the basis of an established protocol used in all patients to define Aspergillus categories.

Results: 200 patients have been included into the study from 12/2016-05/2018. 46.5% were male with a median age of 28 years (range 6-69) and a median BMI of 19.6 kg/m 2 (range 11.7-34.2); 46.5% were homozygous for F508del. The median FEV 1 pp was 49 (range 14-133). IgE was 56 kU/L (range 0-4359), mean asp-specific IgE was 0.22 kU/L (range 0-91), mean asp IgG was 65.5 mg/L (range 2.7-200), sputum GM was elevated in 48%, qPCR was positive in 39%, asp-specific T-cell response was positive in 28% (TH1+TH2).

Conclusion:Aspergillus categorizations using recommended definitions could not be confirmed in all patients. Using asp-specific T cell responses expands the Aspergillus categories and enables a distinction to be made between allergic and infectious disease. 

Introduction: Infections with nontuberculous mycobacteria (NTM) present an increasing problem in patients with predisposing diseases such as cystic fibrosis (CF) or immunodeficiency. In pulmonary infections caused by rapid growing mycobacteria, Mycobacterium abscessus is the most frequent pathogen. Diagnosis of M. abscessus infections depends largely on symptoms of disease and confirmation by mycobacterial culture of sputum samples or bronchoalveolar lavages. Microbiological cultivation however, is time-consuming and not always feasible or reliable. A suitable assay should therefore be able to differentiate between infection with tuberculous mycobacteria and NTM, and to distinguish between ongoing infection and colonization.

Aims: We set out to identify immunodominant antigens in M. abscessus by ELISA and immunoblot methods. Selected antigens should not cross react with tuberculous mycobacteria, and preferably only be recognized during infection, but not upon bacterial colonization.

Materials and Methods: Heat-killed bacterial suspensions of the three different M. abscessus subspecies were sonicated, and alkaline-stable lipids were extracted with chloroform-methanol. The composition of lipids was analyzed by thin layer chromatography (TLC). Antigenic lipids were identified by blotting TLC-separated fractions to PVDF membranes and probed with sera from M. abscessus infected patients and experimentally infected animals. The immunodominant fractions were eluted and used to establish an ELISA-protocol for the detection of antibodies in blood and sputum.

Results: Sera from infected CF patients and experimentally infected animals showed a strong reaction against the low molecular fraction in crude bacterial lysates. A more detailed analysis revealed that both lipoarabinomannan (LAM) and lipids from the cell wall were recognized. Since LAM is an essential component of all mycobacterial cell walls, and therefore not appropriate to specifically indicate NTM infection, we further analyzed the response against specific components of the lipid fraction, which are not found in tuberculous mycobacteria. Patients with culture-confirmed M. abscessus disease showed a strong serum antibody response against the glycopeptidolipid (GPL) fraction of TLC-separated alkali-stable lipids. Purified PGL were used to establish an ELISA protocol to detect antibodies in sera and sputum of infected patients.

Summary: We established ELISA-and Western blot-based methods to identify immunodominant antigens in NTM infections. GPL from the cell wall were among the most dominant antigens in crude lysates and are not shared with tuberculous mycobacteria. Most patients, but none of the controls showed a strong and specific antibody response against this fraction within purified lipids in all M. abcessus subspecies. These preliminary results indicate that glycopeptidolipids of M. abscessus might provide a specific target for antibody-mediated diagnostic assays. 

Introduction: During its persistence in cystic fibrosis (CF) airways, P. aeruginosa often develops a multidrug resistance (MDR) phenotype by the accumulation of pathoadaptive mutations. A better understanding of the changes that P. aeruginosa undergoes during the adaptive process resulting in the acquisition of an MDR phenotype is essential for the development of new therapies.

Aims: The primary objective of this study was to analyse the adaptive process of a collection of 40 P. aeruginosa isolates belonging to the same clonal lineage collected from the sputum of a single CF patient over an 8-year period, and to integrate phenotypic, genomic and transcriptomic approaches to unveil yet uncharacterized resistance mechanisms.

Results: While most early isolates were susceptible to almost all antibiotics tested, resistant phenotypes remarkably increased over time in the persistent population. Hypermutable strains emerged over time and showed a significant correlation with the increase of resistance. The mucoid phenotype was mainly present in early and intermediate isolates and did not correlate with resistance; strong biofilm producers were present mostly in the intermediate/late stage of infection, and a significant correlation was found between biofilm production and MDR phenotype. Genomic analyses of the population indicated that the acquisition of antibiotic resistance was evident in specific phylogenetic clusters; however, the emergence of an MDR phenotype over time could not be comprehensively explained by mutations found in known resistance genes or by other pathoadaptive mutations within the genomes, thus supporting the hypothesis that the MDR phenotypes may arise from a combination of yet uncharacterized genomic determinants and/or a differential modulation of the transcriptional profiles. Seven closely related clones with different resistance profiles, were then grown in artificial sputum medium (ASM) and in microaerophilic conditions to mimic the CF environment. For all selected clones, the resistance to antibiotics increased significantly in ASM and biofilm, often resulting in a switch from a susceptible to a resistant phenotype. To characterize the expression profiles of these clones, we performed global gene expression analyses through RNA-seq. Late MDR clones showed a two-fold increased number of differentially expressed genes compared to the growth conditions (planktonic vs. sessile) in respect to the susceptible early one and, when analysed in relation to the isolation time, late clones presented expression profiles more similar to each other than when they were compared to the early one.

Conclusions: Comparative sequence analysis of the genomes in the longitudinal isolates identified putative pathoadaptive mutations within the genomes, that could not fully explain the increase of antibiotic resistance. Transcriptomic analyses revealed the presence of several differentially expressed genes in relation to growth conditions and time of isolation.

We are currently further studying these clones to define a comprehensive correlation between the genotypic mutations in the population and the transcriptomic profiles of the isolates to their corresponding susceptible or resistant phenotype. 

Pseudomonas aeruginosa is the most common pathogen colonizing the lungs of cystic fibrosis (CF) patients. Besides P. aeruginosa, a variety of other microorganisms are present in the CF lung, often as part of a polymicrobial biofilm. The impairment in microbial clearance causes a continuous stimulation of the immune system which is driven at least in part by P. aeruginosa, and eventually leads to irreversible lung damage. The role of other microbiome members in inflammation remains mostly unexplored. Over the last decades, the focus for treatment of CF has been on remedies to ameliorate airway clearance and to eradicate bronchial infections with antibiotics. However, there is a growing interest in addressing the inflammatory response as a therapeutic target for CF patients. This research therefore aims to investigate whether the CF lung microbiome contains anti-inflammatory bacteria that could be exploited to dampen the epithelial immune response induced by P. aeruginosa.

The interleukin-8 (IL-8) production by organotypic three-dimensional (3-D) lung epithelial cells (A549 cell line) was evaluated after infection with P. aeruginosa by an ELISA. Subsequently, we analyzed the IL-8 induction by P. aeruginosa PAO1 co-cultured with other bacteria commonly found in the CF lung: Staphylococcus aureus, Streptococcus anginosus, Rothia mucilaginosa, Achromobacter xylosoxidans or Gemella haemolysans.

While single culture of P. aeruginosa induced a high IL-8 response in 3-D lung epithelial cells after 4 hours of infection, exposure to single cultures of other bacteria, including S. aureus, S. anginosus, A. xylosoxidans, G. haemolysans and R. mucilaginosa, only induced low IL-8 responses. Interestingly, the P. aeruginosa-induced IL-8 response was strongly reduced when co-cultured with R. mucilaginosa. Co-cultures of P. aeruginosa with the other tested bacteria did not show this effect. R. mucilaginosa also exerted anti-inflammatory properties when using other pro-inflammatory stimuli (P. aeruginosa lipopolysaccharide [LPS], S. aureus, or H 2 O 2 ) or various P. aeruginosa CF isolates. The anti-inflammatory effect of R. mucilaginosa was also observed using CF cell lines (CFBE, IB3). In addition, this bacterium was able to significantly reduce the P. aeruginosa-induced MCP-1, IL-6 and GM-CSF response in 3-D lung epithelial cells. With regard to the mode of action, qPCR array results have shown a potential effect of R. mucilaginosa on the NF-κB-pathway. An inhibitory effect on NF-κB-pathway activation was confirmed using Western blotting and a 3-D model of a NF-κB-reporter lung epithelial cell line. Based on these results, we conclude that R. mucilaginosa is able to significantly reduce the inflammatory response in 3-D lung epithelial cells via inhibition of the NF-κB-pathway. These data suggest the presence of potentially beneficial bacteria in the CF lung microbiome, which could be exploited as a novel way to treat the excessive inflammation in CF patients. Introduction: Early identification of pulmonary exacerbations is vital for the management of cystic fibrosis to break the cycle of chronic infection and inflammation. Noninvasive airway sampling in preschool children can be inaccurate and methods such as cough swabs are ineffective to detect organisms such as tuberculosis mycobacterium. Emerging CF pathogens such as nontuberculous mycobacteria (NTM) are essential to detect in order to plan patient management and optimise infection control measures.

Objectives: To assess the feasibility, effectiveness and reproducibility/ validity of using a disposable mask aerosol sampling system as a method of detecting pathogens in the CF lung and to explore the effect of chest physiotherapy on overall microbial yield.

Methods: This is a prospective cross-sectional study including CF children who are able to tolerate a face mask. A suitably sized standard oxygen face mask fitted with a gelatin-based sampling mix which enabled the detection of Mycobacterium abscessus and Pseudomonas aeruginosa was worn by the child during tidal breathing for a period of 15 minutes. Routine microbiology sampling using cough swab, sputum and/ or BAL was completed alongside facemask sampling. The child completed their routine airway clearance with the physiotherapist (including either PEP, acapella, autogenic drainage or nippy clearway). Following this a separate face mask was worn with the neutral sampling mix for a further 15 minutes of tidal breathing. Further microbiology sampling was then completed. These masks were labelled (pre-and post-physiotherapy) and along with the microbiology samples were sent to the microbiology laboratory and stored in the fridge. The samples were then transferred to the laboratory, dissolved in water and placed in the centrifuge. Subsequently the results were analysed in the microbiology laboratory using PCR for analysis specifically for M. abscessus, P. aeruginosa and looking at the yield of expelled microbes.

Data Collection: 11 CF children (6 male) with a mean age 12 years, age range 1-16 years, from the Leicester cohort were included in this pilot study. 8/11 were known to be NTM naïve, 2/11 had previously isolated NTM and 1 patient actively has NTM. 18% (n=2) had a positive yield for M. abscessus, which was not detected on standard clinically indicated airway sampling. One patient with confirmed M. abscessus on BAL, was negative on a paired facemask sampling system. 1/11 patients was known to be actively isolating P. aeruginosa but was negative on the paired face mask sampling. Although 45% (n=5) had a trend towards an increase in microbial yield (prGen16s) following airway clearance, no strong correlation was noted. All patients tolerated the face mask sampling well.

Conclusion: Mask aerosol sampling systems may provide an effective alternative to cough swab and sputum samples to detect M. abscessus. Airway clearance methods did not demonstrate an effect on the overall bacterial burden detected by the face mask collection system. Larger studies are required to confirm whether such methods can be implemented into routine clinical practice. Introduction: Extended-interval dosing of tobramycin in cystic fibrosis (CF) patients has been established as equally efficacious and potentially less nephrotoxic than the traditional dosing strategy. The optimal dosing strategy has not been well-defined and it is not known if one can predict, with sufficient precision, future dosing without repeating levels. The optimal pharmacodynamic parameter that predicts aminoglycoside efficacy has traditionally been correlated to peak concentrations. Recently there has been evidence supporting the use of 24-hour area under the concentration-time curve (AUC 0-24 ) as a pharmacodynamic marker of both efficacy and safety. We present here a large single-center experience of dosing tobramycin based on 2-and 8-hour levels with subsequent extrapolation of peak (C max ) and trough (C min ) concentrations as compared to the AUC 0-24 , and a determination of within-patient variability with these strategies.

Methods: Following an IRB-approved protocol, all CF patients who received at least one administration of tobramycin between 2015 and 2017 were analyzed. First, we calculated C max , C min , volume of distribution (V d ) and AUC 0-24 for each dosing and compared resultant in-range levels and need for dosing adjustments based on these metrics. Then, we assessed the consistency between measurements by calculating the coefficient of variance (CV) for each patient. Using these metrics, predictors of out-ofrange initial tobramycin levels (20 £ C max < 30) as well as out-of-range AUC 0-24 (70 £ AUC 0-24 < 120) were identified. Additionally, we identified factors predicting reproducible drug levels.

Results: A total of 89 CF patients and 314 tobramycin doses were available for analysis. The mean age at dosing was 33 years (9-62), mean dosing weight 61.1 kg (24.8-88) and V d was 0.4 L/kg (0.17-1.17) with a mean administered dose of 10 mg/kg (5.6-15.8) . The mean C max was 24.2 µg/mL (9.0-54.4) with a mean AUC 0-24 of 97 mg*hr/L (33-247). The average increase dose change was 113 mg (n=51) and 119 mg decrease (n=49). Goal AUC 0-24 and C max were achieved in 79% and 68% of doses, respectively. Dosing adjustments were pursued for 139 patients when extrapolated C max was used as the goal parameter, and would have only been required in 50 instances had AUC 0-24 been the target. Elevated or reduced Vd, and elevated creatinine clearance were associated with increased CV.

Conclusions: The results of this study indicate that our dosing protocol can effectively and reliably produce an appropriately extrapolated C max as well as AUC 0-24 , and dosing based on a targeted AUC 0-24 may lead to fewer dose adjustments without a significant difference in dosing variability. Further, V d and kidney function affect clearance, however, if these metrics are stable, repeat dosing may be considered without repeating drug levels. However, larger, controlled studies are needed to understand the pharmacodynamic correlation of AUC 0-24 , consistency between dosing and clinical outcomes. Stenotrophomonas maltophilia is an important colonizer in the lungs of people with cystic fibrosis (CF) and is positively correlated with poor lung function, CF-related diabetes, shorter time to pulmonary exacerbation, and shorter time to transplant or death. We were interested in understanding the transcriptional response of Stenotrophomonas to CF sputum and identifying genes required to grow on its components. For the transcriptional response, we used RNA-Seq to analyze the transcriptomes of S. maltophilia K279a and two CF clinical S. maltophilia isolates in response to the artificial CF sputum formulation SCFM2. To identify genes required to metabolize sputum components, we have focused on mucin utilization. There were a large number of conserved transcriptional responses to SCFM2 in the three Stenotrophomonas strains, including those predicted to be implicated in low-oxygen growth, virulence, and metabolism. Also interesting were transcripts whose expression changes were CF-isolate specific, including those involved in cell motility, surface-adhered biofilms, and stress responses. These CF-specific changes correlate with in vitro phenotypes observed by us and others, including loss of swimming motility, decreased attached biofilms, and greatly increased oxidative resistance among CF isolates. For mucin degradation, we have initially screened 6000 transposon mutants for inability to use mucin as a sole carbon source. Three mutants have been identified in carbohydrate metabolism genes that suggest potentially important utilization pathways for the carbohydrate component of mucin.

We are currently screening more and also repeating this screen with mucin as both the sole carbon, nitrogen, and methionine source. These findings have increased our understanding of Stenotrophomonas response to the CF lung environment and suggested pathways that might be critical for their growth and survival. 

Introduction: Nontuberculous mycobacteria (NTM) cause pulmonary infections in cystic fibrosis (CF) patients and are difficult to treat due to innate resistance to antibiotics. Effective drugs against NTM include the aminoglycoside, amikacin (AMK), and the macrolides: azithromycin (AZM) and clarithromycin (CLR). Drug resistance (DR) mutations have been observed previously in NTM, but their prevalence has not been widely reported for clinical CF-NTM isolates in the United States (US). Understanding DR genotypes and their relationships to antimicrobial susceptibility test (AST) results could improve antibiotic choices and treatment outcomes.

Methods: All CF isolates in the NTM Genomic Database at National Jewish Health (n=330) from 183 US patients including six clinically relevant NTM species and subspecies were screened for DR mutations in the 16S ribosomal RNA (rRNA) and 23S rRNA genes corresponding to AMK and macrolide resistance, respectively. An additional mutation in the erm(41) gene controlling inducible macrolide resistance was studied only in M. abscessus with full-length erm(41) genes (n=162). ASTs for AMK, AZM and CLR were performed for a subset of isolates (n=82), and minimum inhibitory concentrations (MIC) were compared with DR genotypes. A subset of M. abscessus (n=29) isolates was also tested for inducible CLR resistance in 14-day incubation assays.

Results: We evaluated DR mutations in 330 CF-NTM isolates including six species and subspecies: M. abscessus ssp. abscessus (47%), M. abscessus ssp. massiliense (12%), M. abscessus ssp. bolletii (2%), M. avium (24%), M. intracellulare (10%), M. chimaera (4%), and M. chelonae (1%). Of 59 patients with two or more longitudinal samples, all within-patient isolates shared the same genotypes. In the entire patient cohort, DR mutations at position 1408 in the 16S rRNA, corresponding to AMK resistance, were found in 5.5% of patient isolates and four of six species/ subspecies. DR mutations at positions 2058 and/or 2059 in the 23S rRNA, which confer macrolide resistance, were observed in 4.4% of patients and four of six species/subspecies. Of NTM isolates with ASTs, the 16S rRNA mutation corresponded to AMK MICs >64 µg/mL, and 23S rRNA mutations resulted in AZM MICs >256 µg/mL and CLR MICs >32 µg/mL. For M. abscessus with full-length erm(41) genes, 16% of patients isolates had T-to-C mutations at position 28 which was confirmed by the lack of inducible macrolide resistance in extended incubation CLR assays.

Conclusions: DR mutations in the 16S and 23S rRNAs are present in low frequencies (~5%) in CF-NTM populations, and correspond to elevated MICs in AST results. As many as 16% of M. abscessus isolates with full length erm(41) genes also have the T28C mutation and are susceptible to CLR suggesting additional treatment options for a subset of CF patients. This study is ongoing and will be updated prior to the 2018 NACFC. Introduction: Studies of the cystic fibrosis (CF) lung microbiome have shown that as patients age microbial diversity decreases due to the dominance of bacterial pathogens, but the drivers of this microbial succession are poorly understood. Antimicrobial activity of inflammatory cells is likely to be a major selective pressure on the lung microbiota, but a mechanism behind how these cells shape the chemical environment of the lung and select for the pathogens residing there has yet to be elucidated.

Methods: We used a multi-omics approach with 16S rRNA gene sequencing, metabolomics and peptidomics on patient sputum samples (n=101) to understand the relationship between disease severity, the lung metabolome and the lung microbiome.

Results: The CF metabolome existed in two chemical states. One, from patients with more severe disease that contained lower microbial diversity but higher metabolite diversity, and the other, from patients with better lung function with higher microbial diversity and lower metabolite diversity. Sputum from patients in the severe disease group had a higher abundance of peptides and amino acids and were dominated by pathogens such as Pseudomonas aeruginosa. Signatures of these peptides indicated they were sourced from proteolysis by neutrophil elastase and cathepsin G on host proteins, and accordingly, the severe disease group had higher proteolytic activity from these enzymes. The relative abundance of P. aeruginosa was associated with high amino acids in sputum and their metabolism. This pathogen uses these compounds as its principlal carbon source creating a connection between host inflammation, proteolysis and pathogen persistence.

Conclusions: In the CF lung, neutrophil recruitment in response to pathogen infection results in rampant proteolysis producing free amino acids and peptides. P. aeruginosa comes to dominate this niche space in the CF lung due to its resistance to neutrophilic attack and subsequent metabolism of its preferred carbon source that becomes readily available. The objective of this project is to identify and characterize specific factors which contribute to the fitness and virulence of P. aeruginosa in low oxygen environments of the CF lung.

Background: Many studies indicate that microbes in the CF lung encounter a low oxygen environment. In Pseudomonas aeruginosa, the LasR transcription factor controls a complex quorum sensing network that coordinates expression of a broad set of genes, many of which are secreted proteins and small molecules that contribute to virulence. LasR loss-of-function (LOF) strains regularly emerge in the lungs of patients with CF (Hoffman LR, et al. J Cyst Fibros. 2009;8:66-70) and we have found that LasR LOF strains have increased expression of genes involved in growth in low oxygen (Hammond JH, et al. J Bacteriol. 2015; 197:2810-20) .

Methods: We applied computational biology approaches to identify factors involved in microoxic respiration. Using laboratory strains and CF clinical isolates, we performed genetic and biochemical approaches to study the roles of these factors in microoxic respiration and catabolism of substrates enriched in CF sputum.

Results: We have shown that clinical and laboratory strains with LasR LOF display higher activity of an oxygen-sensitive transcription factor Anr compared to their otherwise isogenic LasR+ counterparts. Our data show that increased Anr activity is associated with increased burden in an acute murine lung infection model. Our data also show that LasR LOF strains can outcompete their isogenic LasR+ counterparts in a microoxic competition assay. Anr controls the expression of a high affinity cytochrome oxidase (cco2) that is important for microoxic respiration (Comolli JC, Donohue TJ. Mol Microbiol. 2004; 51:1193-203) . The expression of cco2 is correlated with expression of PA1673 across over 1000 samples, and both genes are highly expressed in LasR LOF strains. We have shown that PA1673 is also critical for microoxic fitness. Biochemical characterization confirmed that PA1673 is a hemerythrin, a class of di-iron proteins, that can bind oxygen reversibly with a K d of 1 µM. Genetic analyses suggest a model in which PA1673 works in conjunction with the high affinity cytochrome oxidases to help cells manage intracellular oxygen. Our data indicate that PA1673 sequesters oxygen away from enzymes involved in the catabolism of carbon sources that require oxygen as a co-substrate. PA1673 significantly inhibits the catabolism of phenylalanine, tyrosine, and choline, which are substrates previously shown to impact signaling and virulence factor regulation.

Conclusions: Taken together, our data suggest the P. aeruginosa cells have strategies for managing oxygen in low oxygen environments and that these pathways are particularly important in LasR LOF strains that arise from LasR+ cells during chronic CF lung infections. This knowledge may be useful in the development of future treatment strategies for patients with CF.

Acknowledgments: Supported by NIH T32-HL134598 and NIH R01 AI091702 to D.A.H., COBRE Pilot Program as part of P20-GM113132 to E.P. and D.A.H., as well as the CF Research Development Program of the CFF. 

Introduction:Burkholderia contaminans represents a highly problematic pathogen for patients with cystic fibrosis (CF) due to multiple reasons including: high antibiotic resistance, transmissibility among patients, and highly unpredictable clinical outcome ranging from asymptomatic infections to a devastating and fatal pneumonia similar to a "cepacia syndrome." The aim of this work was to evaluate the clinical status of patients colonized with B. contaminans and to analyze whether it is associated with the genetic characteristics of the isolates recovered along their chronic infections.

Methods: This study was carried out with 17 patients colonized with B. contaminans attended at the Sor María Ludovica Children's Hospital, La Plata, Argentina, in the period 2005-2017. Among them, 14 patients were chronically infected, while 3 presented transient infections. For each patient, the first isolate which was believed to have initiated the infection and the last isolate available of the chronic infection were analyzed by multilocus sequence typing, (MLST). For all the isolates the antibiotic sensitivities against the main antibiotics recommended by the Clinical and Laboratory Standards Institute (CLSI) Guidelines were determined. Patients' clinical status was assessed through nutritional parameters, such as Z-score BMI and BMI percentile together with respiratory outcomes, including persistent cough, shortness of breath and pulmonary exacerbation frequency.

Results: The MLST analysis revealed a low genetic diversity among isolates: only ST 102 and ST 872 were found. Although ST 872 was the prevalent lineage from 2005 to 2015 it was completely substituted by ST 102 in the last 3 years. Among the 10 patients colonized with ST 872, two different scenarios were found. Half of them evolved slowly, and showed a stable clinical outcome for many years, while the other half evolved rapidly to an adverse outcome. Remarkably, for the 4 patients colonized with ST 102, the nutritional and respiratory outcomes worsened in less than 3 years of the chronic infection.

As regards the antibiotic sensitivity it was observed that along the chronic infection 80% of the isolates belonging to the ST 872 lineage developed resistance to at least three of the antibiotics tested, while the isolates belonging to the ST 102 lineage are still sensitive to some of the antibiotics used in clinical treatments.

Conclusions: Through this investigation we could find for the first time, a direct association between the B. contaminans lineage that is colonizing the respiratory tract of cystic fibrosis patients, with their nutritional and respiratory outcome. Our results also showed that for patients colonized with ST 102 lineage, different treatments for early eradication of B. contaminans might be implemented to eradicate the infection and improve the clinical outcomes of the patients. Cross-sectional studies of NTM prevalence in the CF population have noted the rare occurrence of patients co-infected with more than one species or subspecies of NTM. We hypothesize this problem is likely underrecognized. The PREDICT Trial (NCT02073409) provides a unique opportunity to investigate this through review of extensive NTM culture history and data prospectively collected from subjects followed for up to 5 years.

Methods: Subjects with CF and at least one positive NTM culture in the past 2 years were eligible. All available historic NTM culture data were compiled, and additional cultures were obtained at every visit. NTM identification was performed by standard methodology, including rpoB molecular species identification, and in some cases whole genome sequencing (WGS). Subjects underwent a diagnostic algorithm based on CFF/ECFC NTM Consensus Guidelines. Primary endpoint was the diagnosis of NTM disease.

Results: Enrollment started in December 2013. Results from 44 adults and 7 children (age <18 y) were analyzed. An average of 27.4 cultures (SD 19.0, range 2-76 cultures) were available for each subject over an average duration of 8.2 years (SD 4.0, range 0-18.6 y). Both M. avium complex (MAC) and M. abscessus complex (MABSC) were present in 29.4% of subjects. Among patients with MAC (n= 43), 24% had more than one subspecies within the complex. For patients with MABSC (n=25), 12% had more than one subspecies. Thirty patients had more than one isolate analyzed by WGS. Among subjects with MAC 42.9% (6/14) demonstrated the presence of >1 genetically distinct strain within the same subspecies. No patient with MABSC (0/16) was found with >1 genetically distinct MABSC strain. Overall, 51% of all subjects, (54.5% of adults) enrolled in the PREDICT trial demonstrated the presence of NTM co-infection with 2 or more NTM at the complex, species, subspecies, or strain level. NTM co-infection was not associated with a lower FEV 1 at time of enrollment or greater number of cultures over the study period. Subjects in the PREDICT Trial who met ATS criteria for NTM disease had a greater number of distinct strains than those with clinically indolent infection (2.0±0.2 vs. 1.4±0.2, p=0.02).

Conclusions: Although only a subpopulation of CF patients appear at risk for acquiring NTM, patients with a positive NTM culture will commonly acquire NTM co-infection over time. Recovery of greater numbers of NTM strains is associated with progression to NTM disease.

The current analysis utilizing a range of historic culture techniques underestimates the presence of genetically diverse strains within species for older culture data, but with increasing WGS analysis, the true rate of NTM co-infection can be determined. This study is ongoing and additional data will be added prior to the 2018 NACFC. Acknowledgment: Supported by CFF. Monitoring patient response to antibiotic therapy, particularly the achievement of a subclinical threshold of NTMs is also currently difficult, relying on an integration of patient culture results and symptoms.

Objective: To evaluate CF sputum headspace for molecules that classify patients as NTM-positive or NTM-negative.

Methodology: Sputum from 19 patients (12 paired (longitudinal)) from National Jewish Health, Denver, (stored at -80°C prior to analysis, sample volume > 0.5 mL). Headspace concentrated onto solid phase microextraction fiber. Analysis by comprehensive 2D gas chromatography coupled to a time-of-flight mass spectrometer. Collectively, sputum headspace contained > 400 volatile features. After removal of artifacts and features that were not significantly different between NTM-positive and NTM-negative samples, hierarchical clustering analysis (HCA) was performed based on the chromatographic area of the remaining 26 molecules.

Results: HCA shows classification of patient samples predominantly into two groups -NTM-positive and NTM-negative. Classifications correlated to culture-status of the sample for the majority of patients (two patients classified as NTM-positive, that were classified by culture as NTM-negative). The 26 putative biomarkers were combined into a "sputum volatile organic compound (VOC) score" (SVS) ranging from 0 (no NTM) to 4 (NTM-positive). Paired analysis: three patients that cleared infection (based on sputum culture) had an SVS that went from 4 to 0; 2 patients who did not clear infection over time but were culture negative at the time of sputum sampling had an SVS between 0 and 1; 1 patient transitioned from culture negative to culture positive generated an SVS that went from 0 to 4.

Conclusion: This pilot study presents the possibility that volatile molecules arising from patient sputum could be used to more rapidly establish the culture status of patients suspected of NTM disease as well as track culture status through treatment. Acknowledgment: Supported by CFF Pilot award to J. Nick (PI; NICK2017).

Volatile organic compound (VOC) in paired samples from 6 CF patients before and after treatment (tx) for NTM. The panels show patients who cleared NTM after tx (left) vs. those who have not yet cleared (right). Introduction: CF respiratory disease is a major determinant of reduced quality and length of life and is characterized by chronic, recalcitrant infections. Historically, these infections have been analyzed by classic clinical culture and, more recently, 16S amplicon sequencing. This newer molecular method has demonstrated the CF sputum microbiota to be polymicrobial, often comprising taxa not detected using traditional culture methods. These more recent studies have thus far failed, however, to find consistent predictors of either therapeutic success or clinical status. Metagenomic sequencing of total CF sputum DNA has the potential to characterize both the taxonomic and functional components of sputum microbiota, potentially providing further insight into the persistence of CF respiratory infections. Metagenomic sequencing of CF sputum is hampered, however, by overwhelming quantities of host DNA, leaving a small proportion of microbial reads available for metagenomic analysis. Furthermore, a sizable portion of bacterial DNA in sputum originates from dead bacterial cells, masking the true viable microbial contribution. In this study, we compared the abilities of four methods to deplete human DNA from CF sputum samples. Our goal is to develop a method for metagenomic sequencing of CF sputum that enables functional characterization of these complex microbial communities.

Methods: DNA was extracted from 8 CF sputum samples using four methods for depleting human and extracellular bacterial DNA in addition to standard DNA extraction. The extent of human DNA depletion was determined using both quantitative PCR and metagenomic sequencing. The effect of each depletion method on the calculated taxonomic and antibiotic resistance gene profiles was defined using metagenomic sequencing. We measured the efficiency of bacterial extracellular DNA reduction using mock communities constructed with bacterial isolates from CF sputum samples.

Results: As defined by both quantitative PCR and metagenomic sequencing, hypotonic lysis of eukaryotic cells followed by nuclease digestion of extracellular DNA resulted in the greatest decrease in human DNA. Metagenomic sequencing also showed an increase in microbial metagenomic sequencing depth, increased detection of low-abundance taxa and improved detection and characterization of antibiotic resistance genes using this refined method. Metagenomic sequencing of mock communities demonstrated that this method provided the most accurate representation of the viable constituency of microbial communities and that certain taxa that exude large quantities of extracellular DNA may have been previously overrepresented.

Conclusions: Metagenomic analysis indicated that depletion of both human and extracellular bacterial DNA before metagenomic sequencing provided the most in-depth picture of the viable taxonomic composition and functional capacities of the CF sputum microbiota. We hope this revised DNA extraction method will lead to a more accurate depiction of how these microbial communities contribute to respiratory disease in CF. was shown to be common in children and adults with CF, and to include isolates that overexpress capsular polysaccharides. We hypothesized that these pneumococcal variants would have altered biofilm phenotypes, as well as persistence within the lung.

Methods: Pneumococcal biofilm formation on a glass surface was assessed using an in vitro continuous-flow biofilm culture system (Stovall Convertible flow-cell). Biofilms were stained with LIVE/DEAD BacLight (Molecular Probes) and visualized by confocal laser scanning microscopy. Quantitative measures of biofilm structure were generated from three-dimensional vertical Z-stack images, and used for nonparametric statistical comparison of biofilm properties. Mice (age-matched C57/BL6 or gut-corrected CFTR -/-mutants) were infected via the intranasal route, euthanized at varying time points as indicated, and bacterial load was assessed by plate count and/or quantitative PCR. Inflammatory responses in lung tissue were assessed by histopathology.

Results: Hypercapsulated pneumococcal isolates formed biofilms with significantly greater density and biomass than pneumococcal controls. Biofilm formation was also significantly affected by colony phase-variation; phase locked variants differed significantly in formation and density of pneumococcal biofilms. Bacterial colonization and persistence was significantly elevated in CFTR -/-as compared to normal C57/BL6 mice, with a modest increase in bacterial load for hypercapsulated strains. Pneumococcal infection also resulted in significant lung inflammation.

Conclusions: Long-term infection in the CF airway may select for pneumococci that overexpress surface polysaccharides. These variants have an increased propensity for biofilm formation, which may facilitate bacterial resistance to clearance. The significance of S. pneumoniae as a pathogen in patients with CF merits additional work; while these infections may be problematic, there are also indications that pneumococci can compete with other bacteria and potentially impede colonization by Pseudomonas aeruginosa. Introduction: S. aureus (Sa) is the bacterium most frequently cultured from CF respiratory samples, and Sa infection has been associated with lung disease severity. Thus, the identification of Sa can impact treatment decisions and patient outcomes. Typically, the first step of Sa detection in clinical samples is growth on mannitol salt agar (MSA), a selective medium for cultivating and identifying staphylococci. Fermentation of mannitol (MAN) by Sa creates lactic acid, changing the color of a pH indicator in MSA which allows for easy discrimination between Staphylococcus species. MAN fermentation is believed to be conserved in Sa. Therefore, this test is often used by laboratories as a screen to identify this species. However, the existence of nonfermenting (MAN-neg) Sa has been described in non-CF infections. The role of MAN-neg Sa in CF has not been investigated. We sought to determine if MAN-neg Sa isolates, 1) are prevalent in CF respiratory samples, 2) can predominate individual sputum samples, and 3) have lost the ability to ferment MAN through common mechanisms.

Methods: To determine prevalence, respiratory samples collected during a 2-year longitudinal study of 231 CF children at 5 CF centers were cultured on MSA, and representative isolates were tested for coagulase activity and then screened for MAN fermentation. To determine the frequency of MAN-neg Sa in individual sputum samples, 100 colonies growing on MSA from each of 5 sputum samples collected from different patients were tested for coagulase activity and MAN fermentation. In both cases, Sa identity was confirmed for agglutination-positive, MAN-neg isolates by species-specific (nuc) PCR. Whole genome sequencing (WGS) was performed on isogenic pairs of MAN-positive and MAN-neg isolates cultured from 5 different patients to identify candidate mechanisms for loss of MAN fermentation.

Results: In the multicenter study, 89 (5.9%) of 1,508 Sa isolates were MAN-neg. Among the 200 Sa culture-positive patients in this study, MAN-neg Sa was detected in at least 1 respiratory sample from 42 (21%) patients. In sputum samples from 2 of 5 different patients tested, all Sa colonies (n=100 each) were MAN-neg. A sample from a third patient contained a mixture of MAN-positive (78%) and MAN-neg (22%) Sa. All colonies cultured from sputum samples from the last 2 patients were MAN-positive. WGS analysis of MAN-positive and MAN-neg pairs identified mutations in mtlR, a transcriptional regulator of the mannitol operon, in 3 clonally-unrelated MAN-neg isolates relative to their MAN-positive counterparts.

Conclusions: The ability to ferment mannitol in MSA is a hallmark diagnostic test for the identification of Sa in the clinical and research laboratories. However, results from this study demonstrate that MAN-neg Sa isolates are prevalent among CF patients and can dominate infecting Sa populations. Therefore, laboratories that process CF clinical samples should not solely rely on MAN fermentation for the identification of Sa. Whether MAN-neg isolates represent a clinically-important adaptive change is not yet known. 

Pseudomonas aeruginosa is an opportunistic pathogen known for its flexible metabolism, allowing it to colonize a variety of environmental niches and cause a wide range of human infections. In a study of the volatile metabolomes of 24 P. aeruginosa clinical isolates from various human infection sites, we observed that less than 20% of volatiles were conserved among all of the isolates, and individual isolates have a diverse metabolic repertoire (Bean H, et al. J Breath Res. 2016; 10:047102) . We hypothesized that during chronic CF lung infections, the metabolic diversity within and between P. aeruginosa isolates would be reduced as the infection adapts to the host environment and loses virulence factors. To test this hypothesis we compared the P. aeruginosa volatile metabolomes of chronic cystic fibrosis lung infection isolates from early and late infection stages, and also compared these data to the volatilomes of acute infection isolates.

Methods: P. aeruginosa isolates from acute infections and early and late chronic infections were cultured aerobically in rich media for 24 hours. The headspace of cell-free culture supernatants was sampled by solid phase microextraction (SPME) and analyzed by comprehensive two-dimensional gas chromatography time-of-flight mass spectrometry (GC×GC-TOFMS). Univariate and multivariate statistical analyses and alpha and beta diversity metrics were used to compare the volatile profiles of early vs. late chronic infections and early vs. other acute infections.

Results and Conclusions: In aggregate, we observe the same metabolic diversity in P. aeruginosa isolates from acute infections, early stage chronic infections, and late-stage chronic infections if we compare the presence and absence of chemical species. However, when the abundance of the chemical species is considered, we observe a significant change in the diversity of the volatile metabolome during the course of chronic infections, with an overall reduction in the metabolic output. As a result, the volatile metabolome can be used to classify an unknown P. aeruginosa isolate as an early versus late infection isolate. Introduction: Cystic fibrosis (CF) is a genetic, multi-organ disease caused by absent or dysfunctional CFTR mediated anion secretion, which in the lung, leads to mucus obstruction, chronic infection/inflammation, and increased acidity in the airway surface liquid (ASL). Burkholderia cepacia complex (Bcc) bacteria are extremely pathogenic to CF patients, and acquisition of Bcc is associated with a significant increase in mortality. Short Palate Lung and Nasal Epithelial Clone 1 (SPLUNC1) is secreted by airway epithelia and serves as a multifunctional innate defense protein. We tested the hypotheses that SPLUNC1 is required to reduce Bcc growth in the airways and that its antimicrobial activity is impaired in CF airways.

Methods: SPLUNC1 was knocked down by shRNA in primary human bronchial epithelial cells (HBECs) and HBECs were incubated with Bcc clinical isolate J2315. Additionally, SPLUNC1 knockout (KO) mice and wild-type (WT) littermates were infected with J2315. We have elucidated the crystal structure of SPLUNC1 and screened novel mutants for pH-sensitivity of antimicrobial activity. J2315 was incubated with SPLUNC1 mutants at pH range of 6 -7.5. Non-CF and CF HBECs were then pretreated with 4 µM SPLUNC1 mutants for 1 hour, and then infected with J2315 at MOI of 30 for 2 hours. Washings of the ASL were collected for bacterial counts. Additionally, sputum from non-CF and CF patients were collected, treated with SPLUNC1 and SPLUNC1 mutants, and infected with J2315.

Results: SPLUNC1 KO mice had increased bacterial burden in the lungs compared to WT littermates. Knockdown of SPLUNC1 in HBECs resulted in increased bacterial burden compared to WT non-transduced HBECs. SPLUNC1's antimicrobial activity against J2315 was impaired under acidic conditions, but was restored at ≥pH 7.0. However, some SPLUNC1 mutants reduced J2315 bacterial growth at acidic pH. While SPLUNC1's antimicrobial activity in CF ASL and sputum was impaired, SPLUNC1 mutants reduced J2315 bacterial burden in CF ASL and CF sputum to similar levels as seen in non-CF ASL and non-CF sputum.

Conclusions: Our data suggest that (i) SPLUNC1 is needed in the airways to reduce Bcc growth, and (ii) its antimicrobial activity is pH-dependent. SPLUNC1 mutants, however, retain pH-independent antimicrobial activity. While the impact of CF on SPLUNC1-Bcc anti-microbial interactions is currently under investigation, our data suggest that understanding this interaction may lead to novel therapies to treat bacterial infections in CF lungs. Introduction: Azithromycin (AZM) is a macrolide antibiotic that inhibits bacterial protein synthesis. Studies have shown that AZM use in CF reduces pulmonary exacerbations (PEx), lung function decline and improves nutritional status. This benefit is particularly observed in those chronically infected with Pseudomonas aeruginosa-an organism with intrinsic AZM resistance. Alternate mechanisms including host immune modulation, antivirulence and antiviral effects have been proposed to explain AZM benefits. However, the CF airways are infected with a diverse community of organisms and it may be through AZM effects on the CF microbiome that AZM provides benefit. Herein, we sought to determine the effects of AZM on microbial communities and determine if baseline microbiota associates with clinical response.

We examined our longitudinal sputum biobank to identify patients naïve to AZM, with sputum samples ≤18 months pre-and post-AZM initiation. Samples were excluded if collected within 14 days of PEx or change in therapy. Demographics and dynamic variables of disease were recorded. Total genomic DNA from sputum was extracted and the V3-V4 region of the 16S rRNA gene was sequenced. Patients were classified as Responders (R) if their rate of FEV 1 decline improved in the 18 months following AZM initiation relative to the 18 months prior and Non-responders (NR) if no improvement was observed.

Results: We assessed 74 sputum samples from 29 CF patients (18 female; 11 male) consisting of 44 pre-AZM samples (median -154 days (IQR -372 to 0) from AZM) and 30 post-AZM samples (median +365 days (IQR +245 to +415)). Patients had a median age of 24.7 years (IQR 20.9-30.5), FEV 1 54.5% predicted (IQR 35-85) and 72% were chronically infected with P. aeruginosa. Median annual FEV 1 decline did not differ following AZM initiation in the cohort; -1.18%/year (IQR -5.22-+2.01) pre-AZM and -1.9%/year (IQR -5.5-+6.8) post-AZM, P=0.82. AZM did not induce gross changes in the CF microbiome. No difference in Shannon diversity index (SDI) was noted between pre-and post-AZM samples (1.44 vs 1.33, P=0.65). Of the cohort, 41% were classified as R. Demographics at baseline of R vs NR were similar with respect to age, gender, FEV 1 , co-morbidities and cultured pathogens (P>0.05). Baseline SDI prior to AZM did not differ in R compared to NR (1.1 (IQR 0.43-2.25) vs 1.35 (IQR 0.47-2.04), P=0.85). Furthermore, the relative abundance of the top 10 microbiome members did not differ based on R status. However, R trended towards a higher relative abundance of Pseudomonas 70.6% (IQR 0.1-95.7%) compared to NR 38.5% (IQR 0.5-84.4%) in post-AZM samples (P=0.67) and lower abundance of Stenotrophomonas (P=0.03). Microbial community structure clustered based on baseline FEV 1 % (P=0.02) and by individual patient (P=0.001), but had no association with AZM or rate of FEV 1 decline.

Conclusions: AZM does not appear to induce significant changes in the CF microbiome and response to AZM is independent of baseline community structure. Modification of microbial community structure may not be a mechanism by which AZM exerts its clinical benefits. Cystic fibrosis patients experience disproportionately high rates of chronic rhinosinusitis (CRS) because of epithelial CFTR dysfunction, with upwards of 95% of individuals displaying radiography characteristic of the disease. CRS is associated with microbial colonization of the sinonasal mucosa, and it has recently been appreciated that metastasis of dominant organisms (i.e. Staphylococcus aureus, Pseudomonas aeruginosa) in patients with this disorder may significantly contribute to development of chronic lower airway infection. Chronic colonization with P. aeruginosa in CF patients correlates with trends in seasonal respiratory virus infection (respiratory syncytial virus (RSV), influenza A, rhinovirus (RV)) (5,6). We have recently demonstrated that respiratory virus infection (RSV, RV) promotes increased P. aeruginosa biofilm growth through bronchial epithelial dysregulation of iron sequestration. In addition, we observed an increase in both iron and P. aeruginosa biofilm growth on rhinovirus-infected primary CF sinonasal epithelial cells. To test the hypothesis that sinus virus infection and iron levels affect the microbiota present in the CF paranasal sinuses, we utilized longitudinal 16s rRNA sequencing, 52-plex cytokine panels, and iron quantification in a prospective cohort of 33 CF CRS patients undergoing routine nasal endoscopy with evaluation and debridement. Sinus samples were also subjected to q-RT-PCR respiratory virus panels, where virus infections were detected in 18% of patient visits and the most common viruses detected were human rhinovirus and coronavirus. Using a longitudinal mixed effects model, viral infection was associated with increased sinus iron concentrations. Patients presented with low microbial diversity (alpha diversity), with Staphylococcus sp. and Pseudomonas sp. representing the most abundant taxa and virus detection was associated with a further reduction in microbial diversity measures. In a univariate analysis, pseudomonads and Stenotrophomonas were significantly enriched, whereas streptococci were significantly reduced on virus-positive visits. Together, these data suggest a dynamic, exploitative interplay between the host epithelium and its microbial constituents modulated by nutritional immunity during virus coinfections.

This study was funded by CFF ZEMKE16Q0 (Shwachman) and NHLBI K23HL131930, GILEAD ISR Award, NIH NCATS UL1 TR0000005, University of Pittsburgh CTSI Pilot Program. 

Objective: Novel immunological classification of aspergillosis in adult cystic fibrosis was proposed by Baxter and cowerkers (1). This work proposed phenotypic classification as: Class 1; No disease, Class 2; Allergic bronchopulmonary aspergillosis (ABPA), Class 3; Aspergillus sensitised, and Class 4; Aspergillus bronchitis.

The aim of this study was to reclassify the original patient cohort from Manchester Adult Cystic Fibrosis Centre (MACFC) after 10 years to investigate their clinical outcomes and determine if clinical phenotype had changed over time.

Methods: All patients (n=129) from the original patient cohort were followed up until May 2018.

Data collected included clinical outcome, FEV 1 percent predicted, BMI, age, gender, co-pathogens, antibiotic treatment days, azole naïve/therapy, and previously identified latent class analysis of Aspergillus disease. Data were tested for normality and between-group comparisons were calculated with one-way Anova or Kruskal-Wallis. Survival was assessed with Kaplan Meier and then re-analysed with Cox Regression to adjust for other prognostic factors.

Patients eligible for aspergillosis phenotypic reclassification had sputum Aspergillus PCR, sputum Aspergillus galactomannan, and serum immunology (Aspergillus specific IgG, Aspergillus specific IgE, total IgE) tested.

Results: There were 79 patients who survived without transplant to the current day and were eligible for phenotypic reclassification.

To the current day there was no statistically significant difference in survival outcomes between Class 1 (n=49), Class 2 (n=23), Class 3 (n=19), and Class 4 (n=38) (p=0.521).

The sole predictor of survival was baseline FEV 1 percent at consent to the initial study (p<0.001). There was no statistical difference in rate of decline of FEV 1 (p=0.59), rate of decline of BMI (p=0.82), or IV therapy days required (p=0.85) across the classes. The highest proportion of patients treated with azoles were in the ABPA group (n=14, 60.9%) with a median treatment duration of 26 months. A small number of patients with Aspergillus bronchitis were treated for clinical disease (n=8, 21%) with a median duration of 5.5 months treatment.

When reclassifying patient's aspergillosis phenotype, it is clear that sputum Aspergillus galactomannan is markedly raised (median=5.37, mean=5.4) in comparison to Baxter's work (median=0.44, mean=1.42). Initial data suggests that some patients do change class of disease over time, however full results are awaited for latent class analysis re-phenotyping.

Conclusions: There appears to be no difference in mortality, clinical decline, or antibiotic treatment burden across the Aspergillus phenotypic classes. Baseline FEV 1 remains the strongest predictor of survival.

Reference Objectives:Aspergillus is a ubiquitous organism and CF lungs are vulnerable to infection. We performed environmental fungal air sampling prior to and during building work carried out adjacent to Manchester Adult Cystic Fibrosis Centre (MACFC) to evaluate and monitor our air quality. We then undertook further environmental sampling to ascertain if high efficiency particulate air (HEPA) filtration was efficient in lowering environmental fungal load.

Methods: An SAS Microbial Air Sampler sampled 1 m 3 of air in assigned locations throughout our ward on a weekly basis for 14 months and outdoor samples were taken for comparison. Each plate was cultured for 4 days at 30°C.

For the naturally ventilated outpatient department (OPD), baseline air sampling was performed at the beginning of the day in 4 rooms that had their windows closed overnight and 2 rooms that had their windows open overnight. In 2 of the closed window rooms HEPA filters were then switched on, and samples were taken in each of the 6 rooms at 1 hour and at the end of the day, along with an outdoor sample at each time point as a control.

Results: The predominant organism cultured was Aspergillus fumigatus, followed by Penicillium spp. The ward sampling revealed: Site 1, outdoor air: maximum yield 59 colony forming units (CFU) A. fumigatus (range 0-59 CFU, median 9), and 8 CFU Penicillium (0-8, median 0). Site 2, ward corridor: 29 CFU A. fumigatus (0-29 CFU, median 2.5), 15 CFU Penicillium (0-15, median 0.5). Site 3, patient room: 12 CFU A. fumigatus (0-12 CFU, median 2), 9 CFU Penicillium (0-9, median 0). Site 4, positive pressure anteroom: 2 CFU A. fumigatus (0-2 CFU, median 0), 1 CFU Penicillium (0-1 CFU, median 0). Site 5, patient room: 58 CFU A. fumigatus (0-58 CFU, median 4.5), 5 CFU Penicillium (0-5 CFU, median 2.5). There was a clear rise in Aspergillus burden demonstrated during the summer months in all areas except positive pressure anteroom (>10 air changes/ hour), which persistently yielded negligible fungal growth.

OPD sampling has identified an array of fungal species, including A. fumigatus, A. niger, and Penicillium with no clear pattern of decreased fungal burden in the rooms using HEPA filtration.

The rooms with windows opened tended to reflect the outdoor control sample as may be expected, however, rooms with windows closed and no HEPA filtration had variable and often high fungal yield. Further work is continuing to increase test days and data.

Of note there were patient complaints of rooms overheating due to HEPA machines, and of unacceptable noise emitted.

Conclusions: High levels of A. fumigatus were persistently isolated from ward sites 1, 2, 3 and 5, at baseline and during building works, with peak counts found in the summer months. Fungal ingress onto the ward was demonstrated in all sites except in the positive pressure anteroom. The clinical efficacy of HEPA filtration in reducing environmental fungi appears to be limited.

Data provides further evidence of need for effective ventilation to reduce fungal ingress in CF centres. Correlation with patient outcome data is required. Introduction: Chronic Pseudomonas aeruginosa (PA) lung infections in cystic fibrosis (CF) patients present an infection paradox; antibiotics often fail to fully eradicate antibiotic susceptible bacteria, which successfully adapt and persist in the lungs of the patients. Our objective is to understand why antibiotic treatment fails in CF patients infected with PA. It has been suggested that bacterial persistence is associated with the presence of "persister" subpopulations. Persister bacteria are susceptible cells that survive antibiotic treatment and can resume growth when antibiotics are no longer present, resulting in antibiotic tolerance. We hypothesize that treatment failure may be worsened due to the presence of high persister mutants with significantly increased persister subpopulations in presence of antibiotics.

Methods: A collection of 495 PA isolates from 40 CF patients was screened by high-throughput methods involving ciprofloxacin (CIP) treatment. This screening was used to quantify different levels of persistence (high (HiP) and low (LoP)) among the clinical isolates. Determinations of fitness of HiP variants relative to wild-type phenotype was carried out in a biofilm system simulating the antibiotic dosing used to treat CF patients. In addition, whole genome sequencing (WGS) data was used to search for frequencies of specific impact mutations within the HiP population.

Results: Approximately 25% of the clinical isolates showed HiP phenotypes, and these isolates are occurring frequently in the bacterial populations derived from 65% of the patients included in the study. The distribution of CIP resistant isolates (cipR) varied greatly with a significant enrichment of cipR among HiP variants (60% cipR among HiP, 27% cipR among LoP). Interestingly, isolates with HiP phenotype show significantly increased fitness compared with LoP isolates in biofilm investigations, where CIP is added to the biofilm populations following concentration changes simulating the pharmaco-kinetic and -dynamic conditions when treating CF patients. Mutation analysis of WGS revealed increased impact (missense and nonsense) mutations in 13 genes.

Conclusions: The present study demonstrates that PA persisters are clinically relevant as a possible explanation for antibiotic eradication failure. High percentage occurrence of HiP isolates contributes to the maintenance of PA in the lungs of CF patients, possibly due to their increased fitness in antibiotic containing environments. Analysis of HiP mutants led us to conclude that several HiP genes are involved in persistence by which rpoN was the most frequent mutation associated with HiP. Introduction: Antibiotic efficacy achieved in vitro infrequently correlates with clinical outcomes. For example, if a P. aeruginosa isolate demonstrates susceptibility to a given antibiotic in the clinical lab, that compound can be ineffective when administered in vivo. Conversely, if a pathogen is found to be resistant to an antimicrobial in vitro, treatment with that same compound can result in the resolution of patient symptoms. This paradox raises the question: do conventional antibiotics have a therapeutic effect beyond the targeted pathogen?

Objective: Here we tested our hypothesis that aztreonam and other conventionally used antimicrobials (trimethoprim, tobramycin, meropenem and piperacillin) can exploit interspecies relationships among CF bacterial community members. Specifically, under conditions where P. aeruginosa relies on co-colonizing organisms (anaerobes) for growth nutrients (Flynn JM, et al. PLoS Pathog. 2016; 12:e1005846) , P. aeruginosa may be indirectly targeted by inhibiting growth of its metabolic partners.

To test this hypothesis, we collected sputum from twenty adult subjects and used enrichment culturing to isolate both P. aeruginosa and anaerobic bacteria from each sample. Predictably, both subpopulations showed variable susceptibility to each antibiotic under oxygen-limited conditions. Paired subpopulations from each subject were then co-cultured in a defined mucin medium, where P. aeruginosa was dependent on anaerobic mucin degradation to acquire nutrients for growth. Under these conditions, the antibiotic resistance profiles of the anaerobic community were frequently predictive of overall bacterial community growth (determined by quantitative PCR). In most paired communities, P. aeruginosa was inhibited at significantly lower concentrations of multiple antibiotics, relative to minimum inhibitory concentrations determined by P. aeruginosa monoculture under nutrient-replete conditions (i.e. non-cooperative growth).

Conclusions: These results underscore the importance of both growth environment (e.g. oxic versus hypoxic, nutrient bioavailability) and bacterial community interactions (e.g. cooperative versus monoculture) on the efficacy of antibiotic compounds. Specifically, our data support the "weakest link" model in microbial ecology, where the antibiotic tolerance of a co-operative polymicrobial community drops to that of the least resistant member (Adamowicz EM, et al. bioRχiv. 2018:243949) . These data also suggest that targeting interspecies interactions may have implications for CF bacterial community growth, particularly in instances where a given pathogen exhibits resistance when cultured in isolation. Acknowledgment: Supported by Gilead Sciences. Introduction:Mycobacterium abscessus is a rapidly growing nontuberculous mycobacteria (NTM) pathogen with an increasing prevalence, high virulence, and antibiotic resistance that is implicated in lung function decline. M. abscessus growth studies are largely confined in vitro to the use of simple bacterial broth as the growth media, or to complex in vivo systems. However, bacterial growth is highly dependent on growth conditions. The CF airway is enriched in neutrophils and neutrophil corpses. Dead and dying neutrophils modify the growth of CF pathogens, including Pseudomonas aeruginosa, in ways that can undermine therapy, such as increased antibiotic susceptibility by biofilm-like aggregate formation. We sought to understand if M. abscessus growth is modified in the presence of dead and dying neutrophils. Recognizing the mode of growth and antibiotic response of M. abscessus in CF airway conditions may provide insight into development and testing of novel antimicrobials.

Methods:M. abscessus single cell suspensions were prepared in PBS by sonication. Human neutrophils were isolated from peripheral blood and frozen in RPMI and human plasma to induce cell death and lysis. M. abscessus was inoculated into either media alone or in the presence of neutrophil lysates. DNase or antibiotics were added at the time of inoculation. Bacterial growth was determined by serial dilution and plating. Limited DNase treatment was used in some microscopy experiments.

Results:M. abscessus growth was enhanced in the presence of neutrophil lysates. The presence of catalase in neutrophil lysates may account for some of the growth activity, as catalase alone enhanced growth of CF isolates. The majority of M. abscessus biomass occurred on or within the tangle of dead neutrophils, while fewer surface-attached M. abscessus aggregates were observed. Neutrophil lysates induced both smooth and rough morphotypes of M. abscessus to form aggregates. Addition of DNase at the time of M. abscessus inoculation reduced M. abscessus aggregate formation. A hallmark of biofilm growth is an increase in antibiotic resistance. However, M. abscessus grown in the presence of neutrophil lysates demonstrated similar susceptibility to amikacin and azithromycin compared to growth in media.

Conclusions: Neutrophil lysates provide a growth advantage to M. abscessus and alter the mode of growth from planktonic or surface-attached microcolonies to dense, suspended aggregates. The data suggest that DNA acts as a scaffold for aggregate formation, and catalase promotes enhanced growth. M. abscessus growth in neutrophil lysates and growth in media were similarly susceptible to antibiotics. Elements of the CF airway environment may provide a growth advantage to M. abscessus. Recognizing changes in mode of growth in conditions modeling the CF airway may aid in developing therapeutic strategies and novel therapeutic targets.

Acknowledgments: Supported by CF Foundation grants MALCOL16I0 and NICK17G0. In efforts to better understand microbial triggers of CF pulmonary exacerbations, multiple studies have compared microbiota of respiratory samples collected during periods of baseline health to those collected at the time of pulmonary exacerbation. This approach has most often utilized a limited number of baseline samples under the assumption that CF airway microbiota are relatively stable during periods of clinical stability. However, the day-to-day dynamics of these communities are largely unknown. Our objectives were to 1) determine the stability of CF airway microbiota during periods of baseline health; 2) determine the stability of these communities between baseline periods; and 3) identify factors associated with variation in baseline community stability.

Methods: Daily sputum samples and self-reported surveys of symptoms and medication use were collected from a single adult with CF over 731 days. Baseline samples were selected based on daily survey results and physician medical records. Baseline samples were defined as those at least 21 days after completion of episodic antibiotics and more than 14 days prior to start of treatment with episodic antibiotics for pulmonary exacerbation. Baseline samples underwent DNA sequencing of the V4 region of the bacterial 16S rRNA gene. Measures of microbial ecology were calculated, and compared within and between baseline periods.

Results: Four periods of baseline clinical stability were identified (N=227 samples; range 15-99 samples per baseline period). With nMDS ordination analysis, the centroids of each baseline period differed from the centroids of the other baseline periods (AMOVA, p<0.005 for each comparison). Day-to-day Bray-Curtis (BC) similarity between samples was high within each baseline cluster (median 0.831), and was similar across baseline periods; however, outlier samples were observed (range of day-to-day BC similarity 0.233-0.928). The average pairwise BC similarity of samples obtained at weekly or greater intervals was lower than the average day-to-day BC similarity (repeated measures ANOVA, p<0.0031). Within each baseline period, changes in maintenance antibiotic regimen were associated with shifts in bacterial community composition (AMOVA, p<0.02).

Conclusions: Based on analyses of daily sputum samples from an adult with CF over a 2-year period, day-to-day stability of airway microbiota is high during baseline clinical state; however, the presence of outlier samples demonstrates the limitations of using a single sample as representative of the baseline microbiota. Microbial community structures changed over time between baseline periods, suggesting the need to use the proximal baseline period when analyzing changes that occur between baseline and exacerbation. Finally, changes in airway microbiota were associated with changes in maintenance antibiotic regimens, emphasizing the need to account for maintenance antibiotics in considering baseline community dynamics. We anticipate these findings will inform the design of future studies investigating changes in airway microbiota between baseline and exacerbation.

Acknowledgments: Support by awards from CFF (LIPUMA13I0, LIPUMA15P0, CAVERL17A0), NIH (1R56HL126754-01A1, 1R01HL136647-01). Introduction: Antibiotic (ABX) selection for CF pulmonary exacerbation (PEx) management is frequently guided by respiratory cultures and antimicrobial susceptibility testing (AST) to identify pathogen-specific resistance patterns. Although evidence-based AST guidelines exist for acute systemic infections, its clinical utility in CF PEx management is unclear. Small, single-center analyses in CF have failed to illustrate any AST clinical benefit. This study aimed to describe the frequency of AST in pediatric PEx, to determine if AST ordered at PEx diagnosis is associated with an alteration to initially-prescribed ABX regimens, and to test relationships with time to next PEx requiring ABX.

Methods: This retrospective cohort study utilized data from the Pediatric Health Information System database from Jan 2010 through Dec 2016. The clinical transaction classification code "Sensitivity/Susceptibility Testing" was used to identify AST. Children and adolescents (1-18 years) were studied. Inclusion criteria included daily oral and/or IV ABX use each hospitalization day, respiratory culture within 48 hours of admission, and for PEx with AST testing, AST ordered on or after the day of respiratory culture. An ABX switch was defined as any addition/subtraction in oral, inhaled, and/or IV ABX occurring a minimum of 5 days after respiratory culture to ensure enough time was available for the AST to result. Associations between AST and ABX switching were modelled by multivariate logistic regression models and fixed-effects to adjust for hospital-level variation. Kaplan-Meier and Cox proportional hazards survival analyses were used to study effects on time to next IV-treated PEx.

Results: Over the 7-year study period, 3341 PEx were available for analysis, of which 1905 (57%) had AST performed during hospitalization. PEx patients that had AST ordered were older (11.5 vs 8.3 y) and more likely to use azithromycin (37.1% vs 28.1%) and insulin (12.8% vs 6.1%) during hospitalization. Sex (52.4% vs 50.9% female), length of stay (11.0 vs 10.0 days), and use of oral steroid treatment (24.7% vs 24.3%) were not different between groups with and without AST ordered. Ordering AST at admission was associated with increased ABX switching (32.9% switching vs 20.2% no switching, p<0.001) in both unadjusted and adjusted models. ABX switching was also associated with length of stay (per day increase, OR 1.18; CI 1.15-1.22, p<0.001), and number of admission ABX prescribed (per 1 increase, OR 1.58; CI 1.40-1.79, p<0.001). We intend to perform survival analysis using Cox-proportional hazards regression to determine if the use of AST is associated with a longer time to next PEx requiring IV ABX.

Conclusions: AST is expensive, burdensome, and not readily available at all centers, and thus studies evaluating its clinical impact are essential. CF patients with AST ordered at hospital admission for PEx are significantly more likely to have ABX treatments switched after five days despite a lack of objective evidence that this practice is beneficial. Additional modeling (including survival analyses for time to next PEx) will test for correlation of AST and clinical outcomes. Rationale and Objective: Pseudomonas aeruginosa (Pa) pulmonary infections are the primary cause of mortality in cystic fibrosis (CF) patients. Today, there is no way to predict early Pa colonization. We hypothesized that the airways microbiota may provide clues to decipher this issue. In a microbiome-based pilot study, we found that Porphyromonas was negatively correlated with Pa. In a previous study (Bernarde C, et al., PLoS One. 2015; (4)), we demonstrated a sustainable increase of Porphyromonas after initiation of ivacaftor, which was positively correlated with the FEV1. Taken together these results suggest that Porphyromonas could be a favorable prognostic biomarker in CF, especially in the context of Pa pulmonary infection. To go further, we tested this hypothesis in a larger cohort study.

Methods: We analyzed the airways microbiota of a biocollection settled via a prospective multicenter cohort study of 96 CF patients (median age: 13.8 y) followed for 3 years. Only patients Pa-free for at least one year were included. For each patient, sputum was collected every three months up to the first Pa positivity in culture. Bacteriological cultures were performed according to the French SOPs; clinical data were collected. For 188 sputum samples (47 CF patients including 22 who became Pa positive at the end of the follow-up), microbial composition, Porphyromonas relative abundance (RA), and diversity were analyzed by 16S rRNA gene amplicon paired-end sequencing on the MiSeq Illumina platform and thanks to a validated pipeline (https://github.com/dridk/mucobiome). Porphyromonas qPCR schemes covering the main Porphyromonas species were designed in order to analyze the absolute abundance (AA).

Results: The CF airways core microbiota was composed of 12 genera, including Streptococcus, Neisseria, Haemophilus, Staphylococcus, Granulicatella, Prevotella, Gemella, Fusobacterium, Porphyromonas, Leptotrichia, Oribacterium, and Capnocytophaga in descending order. We found that Porphyromonas RA was more abundant with a lower IQR in PA negative samples than in Pa positive samples. This result was confirmed by Porphyromonas qPCR (P=0.039, t-test). We observed that 75% of patients initially highly colonized by Porphyromonas (≥10 3 genome equivalent/mL) remained uncolonized by Pa; Conversely, only 37% of patients initially weakly colonized (<10 3 genome equivalent/mL), remained uncolonized by PA (P=0.0088, chi-squared test). Finally, in patients harboring Pa in culture at the end of the study, we noticed a decrease of Porphyromonas AA throughout the follow-up; conversely, for patients remaining Pa negative, Porphyromonas AA remained stable or slightly increased.

Conclusion: Given this first insight, we suggest that Porphyromonas molecular detection may be interesting in the follow-up toolbox of CF to enlarge the window of opportunity in the management of Pa early infection. Further studies on replication cohorts, including young children cohort, are needed to validate Porphyromonas as a predictive biomarker of Pa infection. Acknowledgment: Supported by the French CF association, Vaincre la Mucoviscidose. Methods: Patients aged ≥6 years with FEV1 25-90% predicted and ≥2 PA-positive respiratory cultures were enrolled from CFF-accredited centers in the US. Respiratory samples were collected annually and processed at a central reference laboratory. The primary outcome was the annual change in the proportion of subjects whose least susceptible PA isolate had ≥4-fold increase in aztreonam minimum inhibitory concentration (MIC) and was above the parenteral aztreonam susceptibility breakpoint (MIC >8 µg/ mL). Other outcomes included overall changes in PA susceptibility to aztreonam, i.e., the proportion of subjects whose highest MIC was susceptible (≤8 µg/mL) vs resistant (>8 µg/mL), susceptibility to other anti-PA antibiotics over time, and the association between AZLI exposure (≥1 course in the 12 months prior to respiratory cultures) and changes in these susceptibility patterns. Isolates from subjects with PA-positive respiratory cultures were compared for annual MIC changes from baseline through Year 5 for all microbiologic outcomes.

Results: From August -November 2011, 510 subjects were enrolled from 31 CF centers. Their mean age at enrollment was 26 years of age (range 6 to 71); 377 (74%) were ≥18 years and mean FEV1 percent predicted was 60% (range 17-104%). In the year prior to enrollment, 276 (54%) subjects received >1 AZLI course and 130 (26%) received ≥5 AZLI courses. Evaluable subjects whose PA isolates met the primary outcome included: Year 1: 13% (41/312); Year 2: 22% (58/266); Year 3: 19% (49/258); Year 4: 16% (36/232); Year 5: 15% (30/201). No association with AZLI exposure (yes/no or number of courses) was seen between subjects who did/did not meet the primary outcome. PA susceptibility to aztreonam (MIC ≤8 µg/mL) was: Baseline: 62%; Year 1: 63%; Year 2: 54%; Year 3: 51%; Year 4: 57%; Year 5: 51%. Among subjects with AZLI use in the 12 months prior to respiratory cultures, the aztreonam median MIC (MIC50) for all PA isolates was 8 µg/mL in Years 1 to 4, and 16 µg/ mL in Year 5. Among subjects without AZLI use, the aztreonam MIC50 varied from 2 to 4 µg/mL in Years 1 to 5. There were no 4-fold or greater increases in MIC50/MIC90 for all PA isolates from baseline through Year 5 for aztreonam or other anti-PA antibiotics.

Conclusions: AZLI use was not associated with increased antimicrobial resistance (as defined by the predetermined primary outcome measure) in CF patients during 5 years of observation. Interpretation of antibiotic exposure and its association with resistance was limited by incomplete capture of exposure to antibiotics in the CFFPR. Acknowledgment: Supported by Gilead Sciences. Objectives: A 5-year prospective observational study was conducted to assess the association of changes in Pseudomonas aeruginosa (PA) susceptibility to aztreonam with selected clinical outcomes in a subset of patients in the CF Foundation Patient Registry (CFFPR) following approval of Cayston® (aztreonam for inhalation solution [AZLI]) for treatment of chronic PA infections.

Methods: Patients aged ≥6 years with FEV1 25-90% predicted and ≥2 PA-positive respiratory cultures were enrolled from CFF-accredited centers in the US. The primary outcome measure was the annual change in the proportion of subjects whose least-susceptible PA isolate had ≥4-fold increase in aztreonam minimum inhibitory concentration (MIC) and was above the parenteral aztreonam susceptibility breakpoint (MIC >8 µg/mL) compared to the previous year. A secondary outcome was decreased susceptibility comparing these parameters from baseline to Year 5. Secondary clinical outcomes were collected from the CFFPR and included decline in FEV1 percent predicted measured from baseline to Year 5, pulmonary exacerbations defined as intravenous antibiotic treatment in the home and/ or hospital, and hospitalizations (all-cause).

Results: 510 subjects were enrolled from August to November 2011 from 31 CF centers of whom 377 (74%) were adults ≥18 years of age. The mean age at enrollment was 26 years of age (range 6 to 71) and the mean FEV1 percent predicted was 60% (range 17-104%). The primary outcome was met in 13%, 22%, 19%, 16%, and 15% of subjects in Years 1-5, respectively. Most AZLI use (≥1 course in the 12 months prior to respiratory cultures) occurred in subjects with moderate or severe lung disease (defined as FEV1 percent predicted <75%), but AZLI exposure (yes/no or number of courses) was not associated with meeting the primary outcome. From

Year 4 to Year 5 the FEV1 percent predicted dropped by 0.8% and 1.4% for those who met (n=30) and did not meet (n=171) the primary outcome, respectively. The annual rates of FEV1 decline from baseline to Year 5 for those who met (n=56) and did not meet (n=153) the secondary susceptibility outcome were 1.86% and 1.84%, respectively. The mean annual number of pulmonary exacerbations per subject differed among those who met/did not meet the secondary decreased susceptibility outcome: 2.0 vs. 1.2, respectively, p<0.001. The mean annual number of hospitalizations per subject who did/did not meet this susceptibility outcome was 1.6 vs. 1.1, respectively, p=0.003.

Conclusions: Over a 5 year observation period, subjects meeting the primary or secondary susceptibility outcomes did not exhibit a greater decline in lung function. Those meeting the secondary susceptibility outcome did have higher rates of exacerbations and hospitalizations both of which are associated with use of antibiotics. In this observational study, these trends are noted, but causality between meeting the secondary susceptibility outcome and clinical outcomes cannot be inferred. Introduction: Cystic fibrosis (CF) is a progressive, genetic lung disease that is characterized by persistent infections and mucus buildup that results in worsening lung function. Due to the risk for MRSA, vancomycin is a common treatment in patients with CF. A common, but difficult to predict reaction that can occur with vancomycin treatment is known as red man syndrome (RMS). The primary objective of this study was to compare the prevalence of RMS in pediatric CF patients versus non-CF patients. The secondary objectives were to evaluate the use of pre-treatment diphenhydramine and risk factors for RMS, specifically examining vancomycin infusion rate (mg/kg/minute), vancomycin infusion duration (1-h vs. greater than 2-h), and vancomycin dose (mg/kg).

Methods: In this IRB-approved retrospective study, data were collected for pediatric patients with and without CF, aged 30 days to 18 years, during the period of January 1, 2015 to July 31, 2017 who received vancomycin. CF patients who received a course of IV vancomycin during the study period were age-and gender-matched with a pediatric patient without CF that also received vancomycin therapy during the study period.

Results: Seventy-nine CF patients were included in the study and matched with 79 non-CF patients. In the CF group 70.9% of patients had a history of RMS or a new onset of RMS during the study period vs. 20.3% in patients without CF (P < 0.001). Of those patients that experienced RMS 73% in the CF group and 27% in the non-CF group received pre-treatment diphenhydramine (P < 0.001). Patients with CF received higher median vancomycin doses (CF: 18.3 (15.5-19.6) mg/kg vs Non-CF: 15 (14.7-16.6) mg/kg) and slower median vancomycin infusion rates (CF: 0.16 (0.14-0.25) mg/kg/min vs Non-CF: 0.25 (0.22-0.27) mg/kg/min) when compared to patients without CF (P < 0.0001, P = 0.0001, respectively). Comparing patients with CF that experienced RMS and those that did not, there was no difference in vancomycin dose or vancomycin infusion rate but CF patients that experienced RMS were more likely to be older than those that did not (CF with RMS: 12 (7-16) years vs. CF without RMS: 7 (1-12) years, P = 0.003). During the study period 14% of CF patients and 11% of non-CF patients experienced new onset of RMS (P = 0.632). Analysis of CF patients without RMS compared to those with new onset of RMS during the study period revealed no difference in vancomycin dose, but patients that experienced RMS had faster median vancomycin infusion rates than CF patients that did not experience RMS (CF new RMS: 0.32 (0.26-0.32) mg/kg/min vs. CF without RMS: 0.16 (0.15-0.3) mg/kg/min, P = 0.034).

Conclusions: The prevalence of RMS was approximately 3.5 times higher in the pediatric CF population compared to the general pediatric population, with no difference in the rate of new onset RMS during the study period. CF patients that experienced new onset RMS received faster vancomycin infusion rates than those that did not experience RMS. CF patients may benefit from slower vancomycin infusion rates to prevent the initial occurrence of RMS. Objective: Nontuberculous mycobacteria (NTM) are isolated increasingly from sputum of cystic fibrosis (CF) patients. Effective treatment depends on accurate species identification and determination of macrolide (MA) and aminoglycoside (AG) resistance. Current approaches use time-consuming Sanger sequencing and culture-based antimicrobial susceptibility testing. An accurate test to identify NTM species and resistance markers in one day is a significant unmet need. The Hain NTM-DR line probe assay (LPA) achieves this with a rapid hybridization method. We tested and validated the LPA against a panel of whole genome sequences from 284 NTM isolates.

Methods: Genomic DNAs from the Colorado CF Research and Development Program (RDP) collection and non-CF control DNAs were PCR amplified, hybridized, developed, and interpreted according to the manufacturer protocol (Hain Lifescience, Nehren, Germany). Results were compared to whole genome phylogenomic trees to identify species. The LPA antimicrobial resistance results were compared to corresponding sequences of three extracted NTM genomic loci: erm(41), 16S rRNA (rrs), and 23S rRNA (rrl).

Results: Species identification and antimicrobial resistance are shown in the Figure. Conclusion: The NTM-DR line probe assay offers an accurate and faster alternative to currently available approaches for identifying the most frequently encountered NTM species in the M. abscessusand M. avium complexes. This assay promises to be an asset for health care providers who require an understanding of antimicrobial susceptibility to implement the most effective treatments for CF patients. Introduction: Anaerobic bacteria constitute an important part of the CF microbiome, but little is known about specific interactions of these bacteria with the lung epithelia. We investigated the pathogenic potential of two closely related gram-negative anaerobes, Prevotella melaninogenica and Prevotella scopos, and the Prevotella hemolysin PhyA, by assessing their ability to cause lung epithelial cell toxicity in addition to erythrocyte lysis. We further studied the post-transcriptional regulation of PhyA by its neighboring gene phyZ that is predicted to encode an acyl-transferase of unknown function.

Methods: We assayed cytotoxicity in the CF epithelial cell line (IB3) and its corrected counterpart (C38) following infection with P. melaninogenica or P. scopos. We sequenced the genome of P. scopos 361B and identified putative hemolysin genes shared between the two species. We screened putative hemolysins using an epithelium detachment assay of each gene subcloned into E. coli. We used mass spectrometry to analyze the amino acid modification in the PhyA protein expressed in the presence and absence of PhyZ, and assessed its effect on hemolysis and cytotoxicity.

Results:P. melaninogenica and P. scopos killed both the uncorrected IB3 and corrected C38 cell lines, with P. scopos displaying a much more robust phenotype. Some of the putative hemolysins showed cell detachment capacity with the strain encoding PhyA having the most robust cell lifting ability. Expression of PhyA in the presence of PhyZ abrogates the hemolytic capacity of PhyA. Mass spectrometry analysis showed that when expressed with PhyZ, PhyA becomes deacetylated at multiple lysine residues.

Conclusions: Our data suggest that P. melaninogenica and P. scopos, common commensal organisms, may have pathogenic potential in CF. The PhyA hemolysin may be a virulence gene that could mediate host tissue destruction. The mechanism of PhyA toxicity is not known. The activity of this protein may be post-translationally regulated by the product of the phyZ gene. We have previously demonstrated that oral-derived anaerobic bacteria are abundant in bronchoalveolar lavage fluid (BALF) samples derived from asymptomatic CF infants. Without the cross feeding of metabolites liberated by mucin-degrading oral anaerobes, the growth of Pseudomonas aeruginosa on a mucin nutrient source is limited. However, the environmental factors that enable the growth of this pathogen in infant airways in vivo have yet to be defined. In lieu of these observations, we hypothesize that anaerobic bacteria serve as contributors to the onset of chronic infection by Pseudomonas aeruginosa in young CF patients.

Methods: We utilized banked CF and non-CF pediatric BALF samples from the University of Minnesota and the Cystic Fibrosis Foundation Therapeutics Biobank Repository. A three-fold approach was designed to gauge the environmental conditions that exist in infant airways. We first measured mucin integrity in BALF via slot blotting and ELISA in order to generate a ratio of intact to total mucins. To identify the mucin-degrading anaerobes and survey the bacterial community implicated in this process, we performed qPCR and 16S rDNA sequencing. Thirdly, we quantified the abundance of mucin-degradation byproducts, acetate and propionate, via gas chromatography and mass spectrometry. Within our analysis, we tested for correlations that exist among these three parameters.

Results: Preliminary experiments revealed significant variation in glycoprotein degradation among BALF samples, a difference that may be attributed to variation in microbiome composition. In accordance with these results, 16S sequencing and qPCR demonstrated that these samples harbored unique bacterial communities, some of which contained high relative abundances of CF-associated anaerobes. Metabolites acetate and propionate were also found to be present at high relative abundances in pediatric CF samples. Most importantly, we have observed a striking negative correlation between the ratio of intact to total mucins and the abundance of both anaerobic bacteria and mucin-degradation byproducts.

Conclusions: These results suggest a mechanistic role of anaerobic communities in the progression of CF airway disease.

Acknowledgments: Supported by the Cystic Fibrosis Foundations LAGUNA 16Y5, LAGUNA 17A0 and the Esselen Charitable Cystic Fibrosis Fund. While established Pseudomonas aeruginosa lung infections in people with CF are generally thought to be clonal, meaning that only one strain infects at a time, there are situations in which multiple P. aeruginosa strains co-exist in CF lungs. For example, multiple P. aeruginosa strains can sometimes be detected early in infection, before a single strain dominates. In addition, epidemic strains can displace established P. aeruginosa, likely after some period of co-existence. It is also possible that CF lungs commonly harbor nondominant P. aeruginosa strains at a low abundance, or multiple strains of other pathogens like Staphylococcus aureus. A key roadblock to understanding is that identification methods typically examine small numbers of isolates from sputum, so co-existing strains could easily be missed. Improved methods could enable early detection of superinfection by epidemic lineages, and improve understanding of strain competition during the initiation and chronic phases of infection. We have developed two methods that combine Multi-Locus Sequence Typing (MLST) and next-generation sequencing to detect co-existing P. aeruginosa or S. aureus strains that have different sequence types in CF sputum. In the first method, sputum is spread onto selective agar to allow for growth of the targeted bacterial species. After incubation, hundreds of colonies are scraped from the plate and the DNA from the population is isolated and pooled. The seven standard MLST loci used for strain typing (Curran B, et al. J Clin Microbiol. 2004; 42:2:5644-9; Enright M, et al. J Clin Microbiol. 2000; 38:3:1008-15 ) are amplified and sequenced using Illumina. Population-level allelic variation at all seven loci is then analyzed to determine if multiple sequence types are present. If strain-level complexity is manageable, the relative proportion of the sequence types can be measured. Control experiments show that strains can be identified if present at approximately 1% relative abundance. A second version of the method amplifies MLST loci from DNA extracted directly from sputum. This approach could be particularly useful for identifying strains that are present in lungs, but exhibit growth defects in culture, or to measure strain relative abundance in vivo without artifacts induced by growth differences. Additional information on the methods, and proof of principle analysis of CF sputum samples will be presented. These tractable methods to simultaneously interrogate hundreds of isolates in individual CF sputum samples could lead to new insights into strain dynamics during CF infection pathogenesis. Background: Nontuberculous mycobacteria (NTM) are cystic fibrosis (CF) pathogens that are clinically challenging to diagnose and eradicate with treatment. The current method for pathogen detection is sputum culture that has significant limitations including slow growth and low sensitivity due to necessary decontamination procedures. Culture-independent detection of NTM with 16S rRNA sequencing has shown poor sensitivity and correlation to culture. Therefore, we have designed primers in the rpoB gene to specifically amplify the NTM "microbiome" that includes a range of NTM species and subspecies. We hypothesize that an NTM-specific molecular assay will have advantages over sputum culture for detecting bacterial burden and predicting clinical significance of NTM in the CF airway.

Methods: The rpoB amplicon sequencing method was evaluated in sputa from adult CF patients (n=19) with differing NTM culture status and disease histories, as well as positive (n=6) and negative (n=2) controls. DNA was extracted from raw sputum and controls, and rpoB amplicons were sequenced on the Illumina MiSeq. Sequence reads were annotated to the species or subspecies level using a custom rpoB sequence database derived from clinical isolates in our NTM Genomes Database. The molar abundance of PCR-amplified sequencing library (nmol/L) and normalized read counts were associated with culture status and disease phenotype.

Results: Positive controls included genomic DNA with 2-3 clinically relevant species mixed in equal nanogram proportions. We were able to annotate >95% of sequences amplified from controls as known NTM species in our database, and we identified expected proportions of NTM species in each mixture. The cross-sectional CF study cohort included both culture positive (n=7) and negative (n=12) sputum samples and patients with the following disease phenotypes: i). never positive for NTM by sputum culture (n=8); ii). history of NTM disease, on treatment or previously treated (n=4); or iii). currently positive for active NTM disease (n=7). Culture negative samples had significantly lower nmol/L of PCR-amplified libraries (p = 0.026) and normalized NTM read counts (p < 0.0001) than culture positive samples. There were also significantly lower NTM read counts in patients that were never positive for NTM compared to those with active NTM lung disease (p = 0.002). In patients with previous or current NTM disease diagnoses, we observed a reduction in NTM read counts in patients on antibiotic treatment compared to those not on treatment. Interestingly, we observed read counts above background levels for two or more NTM species or subspecies in 9/19 or 47% of patient samples.

Conclusions: Using an NTM-specific target, our culture-independent molecular assay reveals the NTM microbiome in raw sputum with high sensitivity and specificity compared to sputum culture. It also offers the potential to discriminate clinical phenotypes. The frequent and unexpected observation of mixed species infections suggests a complex NTM ecology not previously described in CF airways and not detected by standard culture techniques. One of the most widespread and destructive opportunistic pathogens in the cystic fibrosis airways is Pseudomonas aeruginosa, however, P. aeruginosa does not colonize the airways alone. Microbes that are common in the oral cavity such as Rothia mucilaginosa are also present in cystic fibrosis sputum, and have different metabolic capacities than P. aeruginosa. We examined the metabolic interactions between P. aeruginosa and R. mucilaginosa using stable isotope assisted metabolomics. Glucose-derived 13 C was incorporated into glycolysis metabolites, lactate and acetate, and some amino acids by R. mucilaginosa. P. aeruginosa utilized labeled substrates derived from R. mucilaginosa to generate different primary metabolites in hypoxic versus aerobic conditions. Furthermore, fluorescence lifetime imaging of NADH was used to track changes in cellular metabolism and indicated a shift in bacterial central metabolism in different oxygen levels and during cross-feeding interactions. These results indicate P. aeruginosa may get a metabolic boost from its microbial neighbor, which could contribute to its colonization of the cystic fibrosis airways. This work was supported as a pilot project from the UC Davis West Coast Metabolomics Center funded by NIH DK097154, and T.G. is supported through the BEST IGERT program funded by the National Science Foundation DGE-1144901. The majority of cystic fibrosis (CF) patients die as a result of respiratory failure caused by chronic bacterial colonization and inflammation in the airways. Although Pseudomonas aeruginosa is the most common gram-negative CF pathogen, other gram-negative microbes, such as Stenotrophomonas maltophilia, are becoming increasingly dominant in the CF airways. Specifically, S. maltophilia is estimated to infect as many as 30% of CF patients and is associated with a three-fold increase risk of mortality or lung transplant in CF patients. In order to investigate S. maltophilia's adaptation to CF lungs we (1) sought to identify a CF-specific accessory genome amongst a collection of S. maltophilia strains, (2) identified which genes are expressed by S. maltophilia in the lung, and (3) characterized growth, gene expression, and metabolite production for S. maltophilia growing in vitro at a range of pH. Our multiscale 'omics approach indicates that acidic conditions are stressful for S. maltophilia and this bacterium copes through increased expression of DNA repair genes. Taken together, our findings suggest that S. maltophilia may not have evolved to be a CF lung pathogen, and that its negative effects on the human host may be a byproduct of its attempt to cope with the unfavorable environment of the CF lungs. Introduction:Haemophilus influenzae (HI) is one of the most frequently identified organisms in the lungs of CF patients. Indeed, it is commonly referred to as an early airways colonizer. Whereas HI has been extensively studied in pediatric populations, little is known about it in adults with CF.

Methods: We performed a retrospective cohort study of patients attending the Calgary Adult CF Clinic between 2002-2016 to identify patients who had HI recovered from sputum. Patient demographics and clinical data were obtained for each visit in which HI was recovered. HI collected in real-time underwent susceptibility testing and were then stored at -80°C. First and last isolates as well as yearly isolates for each patient were subsequently recovered and assessed for relatedness using pulsedfield gel electrophoresis (PFGE). Our primary outcome of interest was to determine the risk of pulmonary exacerbation (PEx) at incident HI strain infection. Secondary outcomes sought included defining the epidemiology, natural history, and resultant clinical course following HI infection in adults with CF. Persistent infections were defined as recovery of the same pulsotype of HI sequentially in yearly isolates (≥2).

Results: Over the study period the sputum from 71/239 (30%) patients cultured HI, with 52/71 (73%) culturing HI for the first time. Antimicrobial resistance rates were 0% for ciprofloxacin, 22.5% for TMP/SMX and 16.2% for ampicillin. 137 isolates from 66 of the 71 patients (93.0%) underwent PFGE. Of these 137 isolates, 102 represented distinct incident infections with specific HI pulsotypes. This revealed that many patients experienced serial infections with individual strains of HI ultimately replaced by new pulsotypes; 37 patients experienced -1 pulsotype, 23 -2 pulsotypes, 5 -3 pulsotypes and 1 -4 pulsotypes during the 14 years of observation. Patients were no more likely to experience a PEx at incident HI infection compared to visits preceding or following (14/52 versus 17/104, RR = 1.65 [0.88-3.07], p=0.14). Of those infected with HI, 28 (39%) cultured more than one pulsotype and 13 (18%) had persistent infection. No demographic, treatment, or co-infection factors associated with risk of progression to persistent infection. Of the persistently infected patients 8 (62%) ultimately cleared the infection by the end of the study period suggesting persistence does not denote permanence. Five clusters of clonal HI were identified, comprising isolates from 12 (A), 3 (B), 4 (C), 3 (D), 2 (J) individuals. Only those isolates in Cluster C were collected at similar time points (range from 2012-215).

Conclusions: HI was common amongst adults with CF although the bulk of infections were indeed transient. No association with incident infection and PEx risk was observed. While clonality was observed amongst isolates, clusters often comprised isolates collected at markedly different time intervals which may suggest that the clonality observed may be due to prevalent strains in the environment independently acquired by several patients as opposed to patient-patient transmission. As median projected survival in CF increases, a growing challenge is the strategic management of infections. Colistin, a polymyxin antibiotic, is a powerful treatment option in CF acute pulmonary exacerbations (APE) secondary to Pseudomonas aeruginosa. However, little is known about how it is metabolized and how to best optimize dosing for therapeutic efficacy and reduced toxicity. FDA dosing recommendations are not based on modeling studies, which increases the risk of treatment failures and toxicity. Newer, highly accurate analytical techniques are available in research settings which allow accurate pharmacokinetic (PK) and pharmacodynamic (PD) studies of the drug, to better understand its metabolism in CF. In this study, we utilized a novel mass spectrometry assay to determine serum colistin levels and establish PK parameters in subjects undergoing treatment for APE.

Methods: Steady state PK and PD properties of intravenous colistin peak, midpoint and trough concentrations in CF adults admitted for APE treatment were assessed by liquid chromatography-tandem mass spectrometry. Measurement of pharmacologically active colistin A and B from serum was performed (University of Colorado School of Pharmacy). Subjects were enrolled as part of an ongoing single-center prospective clinical trial evaluating colistin versus tobramycin in APE. Steady state PK, including peak and trough concentrations (Cmax and Cmin), area under the curve (AUC 0-24 ), half life (T ½), volume of distribution (VD), and clearance (CL), were calculated for all time points. Pharmacodynamics were evaluated via AUC to mean inhibitory concentration (MIC) ratio (AUC 0-24 :MIC), with optimal efficacy for treatment of P. aeruginosa pulmonary infection at ratios >60 (Dudhani RV, et al. Antimicrob Agents Chemother. 2010; 54:1117-24) .

Results: In CF adults with APE (n=22) undergoing either twice or three times daily dosing with colistin, the mean total daily dose was 2.9 mg/kg/ day. The T½, determined based on clearance and volume of distribution, was on average 7 hours. Clearance (CL) was widely variable, with a mean CL of 6.8 L/hr (±SD 4 L/hr). Mean MIC values to colistin were low, with a mean ±SD of 0.34 ±0.36 µg/mL. The mean AUC 0-24 :MIC ratio for study subjects was 191 (SD ±163), and 81% of subjects reached PD targets with a ratio >60.

Conclusion: When given during APE, colistin clearance varies markedly between adults, reflecting a wide range of renal function and raising concern for risk of accumulation and toxicity. When given at 8-hour intervals, colistin often accumulates, since mean T1/2 was 7 hours. Given that MICs are quite low, desirable PK/PD targets are being achieved with total daily dosing of 2.9 mg/kg/day. While the use of colistin has been demonstrated to be safe in CF, prior studies have used far greater doses (~ 4 fold). The current study demonstrates that lower total daily dosing can achieve effective bacterial killing with less renal exposure, when MICs are <0.5 µg/mL; though assessment of best dosing intervals is still being evaluated in the current clinical trial.

Acknowledgment: This study was funded by CF Foundation grant SAAVED15A0. Microbial antibiotic resistance is spreading, and the risk that we fall into a new pre-antibiotic era is real. Some human pathogens are resistant to all known antibiotics. In this context, attractive approaches pertaining to the interference in bacterial chemical communication (known as quorum sensing (QS)), have the potential to control pathogens without killing commensal bacteria. Numerous bacterial pathogens produce and utilize acyl homoserine lactones (AHLs) as chemical signal molecules to coordinate, in a cell density dependent manner, pathogen-critical bacterial behaviors including virulence and biofilm formation. Consequently, enzymatic quenchers termed lactonases, which hydrolyze AHLs, inhibit pathogenicity.

The main cause of morbidity and mortality in patients with cystic fibrosis (CF) is chronic lung infection. The most prevalent pathogen is Pseudomonas aeruginosa, but other important pathogens have been isolated including Burkholderia cepacia. Both pathogens utilize quorum sensing for virulence, biofilm formation, and colonization. Lung colonization is a long process where P. aeruginosa forms biofilm making it difficult to fight using current therapies. New strategies to inhibit biofilm formation and fight pathogens including multiresistant ones are much needed.

We investigated the ability of enzymatic quenchers to inhibit both virulence and biofilm formation on CF clinical isolates of P. aeruginosa. We used two enzymes, which have been engineered to be highly stable and highly active against AHLs. Because P. aeruginosa is known to utilize two parallel QS circuits, one using 3-oxo C12 AHL, the other one C4 AHL, we used two enzymes with different substrate specificity. For instance, one enzyme disrupts only the 3-oxo C12 AHL based circuit (Ssopox-W263I) and a second enzyme disrupts both circuits (GcL) used by P. aeruginosa for quorum sensing.

We have tested the effects of these enzymes on 39 isolates of P. aeruginosa, and demonstrated that enzyme treatment reduced biofilm formation in 23% of the strains by up to 88%. Addition of the lactonase reduced elastase in 53% of the clinical isolates by up to 63%. We saw a reduction in protease up to 69% with addition of lactonase in the tested strains. These results are consistent with recently published data that showed reduction in virulence factors and biofilm formation of P. aeruginosa isolated from diabetic foot ulcers using Ssopox-W263I (Guendouze A, et al. Front Microbiol. 2017; 8:227) .

The present study demonstrates the ability of these engineered lactonases to disrupt the complex communication system of these bacteria to reduce virulence and biofilm formation. With future investigations on mixed species communities and combination therapy with antibiotics, we expect these studies to provide the key data to assess the potential of signal disruption to treat and prevent infections relevant to CF patients, transition to animal model studies, and delineate the importance of signaling in chronic lung infections. Introduction: Pseudomonas aeruginosa populations chronically infecting the CF lung exhibit profound population heterogeneity. To improve quality of life, lung function, and reduce exacerbation occurrence, patients with chronic P. aeruginosa infection are prescribed inhaled antibiotics such as aztreonam (AZLI). These typically reduce sputum P. aeruginosa burden by 1-2 log. However, clinical improvements do not correlate with Pseudomonas killing. We postulated that inhaled antibiotics disproportionally kill virulent P. aeruginosa, and as such populations collected under AZLI selective pressure would have lower pathogenicity.

To study P. aeruginosa pathogenicity we utilized a Drosophila melanogaster chronic fly infection model. To determine if isolates of the same genetic background have differential virulence, 4 isolates collected from a single patient at a single point in time were independently assessed. Strain-specific PCR assays and pulse field gel electrophoresis (PFGE) were used to confirm strain genetic background. Fly survival proportions were measured daily over a period of 14 days. To determine if AZLI induced changes in P. aeruginosa population pathogenicity, sputum samples were collected from CF patients at two time points: day 0 (prior to AZLI) and after 28 days of AZLI 75 mg inhaled TID. From each sputum sample, approximately 80 randomly selected individual isolates were independently collected and stored, and subsequently pooled together to create an artificial community. Proportions of surviving flies were measured daily over a period of 14 days and colony forming units were counted in order to analyze morphotype dynamics.

Results: Four separate and distinct isolates with identical pulsotypes, with a positive PCR assay confirming the Prairie Epidemic Strain status (two protease positive, two negative) exhibited differential fly killing in our preliminary study. When combinations of two individual isolates were assessed, attenuation of virulence was observed. A total of 640 P. aeruginosa isolates collected from sputum samples 28 days apart from five patients were included (median age 32 years and FEV1 53% predicted, 80% F508del homozygous). Patient/time point specific P. aeruginosa artificial communities resulted in differential fly killing in 3/5 patients. However, in no instance was the artificial population of P. aeruginosa created from isolates collected under AZLI selective pressure less virulent. Within the model, there was no evidence of changes in the distribution of morphotypes over time, either in the flies themselves or the substrate filter.

Conclusion:We observed that genotypically identical isolates of P. aeruginosa have different virulence potential in vivo using the D. melanogaster fly model. Combining individual isolates can, however, modulate this effect. Furthermore, pooled populations collected under different therapeutic influences can have differential virulence. However, we found no evidence that AZLI induced a community of lower virulence potential that might explain its therapeutic benefit.

Acknowledgements: This work was partially supported by an NSERC discovery grant to DGS Wan, Y.G. 2 1. Medicine, Dartmouth Medical School, Norwich, VT, USA; 2. Microbiology and Immunology, Dartmouth Medical School, Hanover, NH, USA CFBE (ΔF508) were cultured to near confluency and infected with respiratory syncytial virus (RSV; RFP-labeled) for 24-48 hours, and subsquently overlaid with Pseudomonas aeruginosa (PA) for 2 hours prior to harvest. Uninfected ΔF508 cells were similarly infected with RSV with PA co-cultured for 2 hours. Wild-type controls were used as well for both conditions. RNA isolated from the PA isolated from RSV-infected ΔF508 cells vs. uninfected ΔF508 cells was isolated and subjected to Nanostring™ analysis for a panel of PA virulence factors. The results revealed a significant enhancement of expression of multiple virulence factors expression by PA, induced by the RSV-infection of these cells, compared with induction by co-culture with uninfected cells. Particularly induced in the presence of RSV were phenazines, and associated pathways, as well as elastase and iron acquisition genes. These data suggest that respiratory virus infection in cystic fibrosis may directly enhance the virulence of colonizing PA, and may help to explain the frequently observed association between respiratory virus expression and acute exacerbation of cystic fibrosis.

Athern, K.; Young, D.C.; Bhakta, Z. University of Utah, Salt Lake City, UT, USA Introduction: Due to recent advancements in the management of cystic fibrosis (CF), the number of patients entering adulthood continues to increase and in many countries the percentage of adult patients exceeds that of pediatric patients. Parallel to this shift, the number of pregnancies occurring in women with cystic fibrosis has risen steadily, with 235 pregnancies reported in 2015 versus 124 pregnancies in 1990 (Cystic Fibrosis Foundation. Patient Registry Annual Data Report. 2015). As pregnancy becomes more common in cystic fibrosis, many questions have arisen around the impact of pregnancy and childbirth on clinical outcomes and disease management. One area that remains uncertain is the impact on patient specific pharmacokinetic parameters. Both CF patients and pregnant patients experience changes in medication metabolism, distribution, and excretion. However, the effect of pregnancy on medication clearance in cystic fibrosis patients is unknown.

Objectives: The primary objective of this study was to determine how pregnancy impacted antibiotic clearance in pregnant CF patients hospitalized for an acute pulmonary exacerbation (APE).

Methods: This study was a retrospective chart review of pregnant CF patients at the University of Utah hospital from January 1, 2004 to January 31, 2018. Patients were included within the analysis if they were hospitalized for an APE during pregnancy and had at least one APE hospitalization outside of pregnancy (either prior to or following birth). Data collected included, but was not limited to, antibiotic plasma concentrations on day 2 and day 7 of hospitalization, goal plasma concentration, calculated antibiotic clearance rates, and calculated elimination half-lives. Safety and efficacy data points included serum creatinine, blood urea nitrogen, time to next exacerbation, and forced expiratory volume in one second at baseline, on day 7, and on date of discharge.

Results: A total of 49 patients were screened for inclusion within the study and 4 patients met inclusion criteria. Two patients had 3 hospitalizations available for analysis (prior to, during, and following pregnancy) and two patients had 2 hospitalizations available (during and either prior to or following pregnancy). When compared to clearance rates prior to pregnancy, patient #1 experienced a 22% decrease in clearance (310mL/min to 243mL/min) while patient #2 experienced a 47% decrease (355mL/min to 189mL/min). Both patients maintained a reduced initial clearance rate during the post-pregnancy APE and only one dose adjustment occurred. Patients #3 and #4 did not display a change in clearance rates during pregnancy. Day 7 clearance rates were similar across available hospitalizations for 3 patients and 1 patient did not have day 7 levels available during pregnancy.

Conclusions: This is the first study to evaluate antibiotic clearance rates in pregnant CF patients. While the number of included patients was small, the variability in results supports the need for close monitoring of antibiotic therapy in this population. The impact of this study is unclear as no significant changes occurred with the antibiotic regimens. More research is needed to ensure optimal antibiotic management in this growing population.

Jorth, P. 1 ; Spero, M.A. 2 ; Newman, D.K. 2 1. Academic Pathology, Los Angeles, CA, USA; 2. Biology and Biological Engineering, Caltech, Pasadena, CA, USA Mucoid Pseudomonas aeruginosa evolve during chronic cystic fibrosis (CF) infections and make robust biofilms that protect against antibiotics and the immune response. Previous measurements suggested that alginate gene expression is induced in response to low oxygen, however, these measurements were made in bulk liquid cultures. Because P. aeruginosa is thought to grow as biofilms in vivo and biofilms are known to exhibit large oxygen gradients, we wondered whether alginate production within biofilms might be localized to particular regions. We used confocal microscopy to determine where alginate transcripts are expressed within biofilm populations grown using the agar block biofilm assay (ABBA), which permits rapid study of biofilms; the ABBA assay establishes oxygen gradients over the entire aggregate population and within individual aggregates. The hybridization chain reaction (HCR) was used to detect alginate mRNA transcripts within aggregate biofilms. Using an inducible alginate gene expression system, we showed that HCR microscopic analysis is quantitative. Measuring alginate gene expression using ABBA, we found that alginate gene expression is not detected in single cells nor in small aggregates that begin to form after 8 hours of growth. However, after approximately 12 hours, alginate gene expression was observed in intermediate-sized aggregates growing approximately 100-400 µm below the agar surface where oxygen would be expected to begin to be limiting. Surprisingly, in the mature fully-formed 18-hour aggregates, biofilm-grown mucoid P. aeruginosa only expressed alginate genes growing near the agar surface, where oxygen concentrations are highest. Moreover, within these mature aggregates, alginate gene expression was highest in cells growing near the aggregate surface. Together, these patterns of alginate gene expression suggest that biofilm P. aeruginosa express alginate genes in microaerobic zones rather than in purely anaerobic conditions, as had been suggested by previous studies in liquid cultures. This could be especially important in CF lung mucus where oxygen gradients are steep and suggests that alginate may be most protective for bacteria growing near the air-mucus interface. Going forward, we will be examining where alginate genes are expressed in vivo in model infection systems and in CF patient specimens. 1 1. Indiana Univ., Indianapolis, IN, USA; 2. Univ. of Nebraska, Omaha, NE, USA; 3. Erasmus Univ., Rotterdam, Netherlands; 4. Telethon Kids Inst., Perth, WA, Australia; 5. Washington Univ., St. Louis, MO, USA; 6. Univ. of Melbourne, Melbourne, VIC, Australia Background: Structural lung disease in CF begins early. Viral infections are linked with short-and long-term morbidity in children and adults with CF. Respiratory viral infections in CF, especially in early life, may alter disease trajectory.

Objective: Evaluate associations between respiratory symptoms and viral infections in infants with CF during the 1st year of life and chest CT abnormalities at 1 year of age.

Methods: From 2013-17, 78 infants diagnosed with CF at 4 centers in the US and Australia were enrolled and followed through ~1 year of life. Parents completed weekly surveys assessing respiratory symptoms and collected nasopharyngeal (NP) swabs when symptoms were present. Infants underwent volumetric chest CT scans at ~1 year of age. CT scans were analyzed using the Perth-Rotterdam Grid Morphometric Analysis for CF (PRAGMA-CF) system by LungAnalysis (Erasmus Medical Center) . Encounters were classified based on respiratory symptoms present: Abnormal Breath Sounds (ABS), Coughing, Nasal Congestion, or No Symptoms. NP swabs were analyzed for the presence of viruses by the GenMark Respiratory Viral Panel (Special Projects Lab-Washington University). Generalized linear models were used to test associations between rates of symptom reporting (encounters with symptom(s) / all encounters), number of hospitalizations, number of antibiotic prescriptions, or detection of specific viral groups (ever vs never) and PRAGMA-CF results.

Results: Infants underwent chest CT at a median (range) age of 1.0 (0.8-2.0) years. The median (range) of lung involvement was 0 (0-2.6)% for bronchiectasis (BE), 0 (0-1.3)% for mucus plugging (MP), 1.0 (0-3.4)% for airway wall thickening (AWT), and 1.1 (0-5.9)% for Total Disease (BE+AWT+MP) in 65 infants with technically acceptable inspiratory scans; and 3.5 (0-40.9)% for air trapping (AT) in 58 with acceptable expiratory scans. There were no statistically significant associations with PRAGMA-CF results and rate of symptom reporting (for any of the symptoms), number of hospitalizations, or courses of antibiotics. Adjusting for age, study site, and the rate of other symptoms, the reporting rate of ABS was associated with AWT (p=0.04) (N=62). The median (range) rate of ABS was 3.0 (0-51.4)%. Amongst 56 infants with adequate NP testing, influenza was detected at least once in 5%, human metapneumovirus (HMPV) in 5%, respiratory syncytial virus (RSV) in 14%, parainfluenza in 25%, coronavirus in 25%, and human rhinovirus in 82%. In multivariate models adjusting for age, site, and other viruses, infants ever positive for RSV had higher Total Disease (p=0.02) (N=54); while any HMPV infection was associated with increased AT (p=0.01) (N=48).

Conclusions: Rate of abnormal breath sounds and confirmed infection with RSV or HMPV during infancy were associated with increased early structural lung disease at 1 year of age.

Acknowledgments: Supported by NIH (HL116211), NHMRC (GNT1043768), CFF (SANDERS16Y5). Background: Inhaled hypertonic saline (HS) has been shown to be an effective mucociliary clearance agent in persons with CF >6 years of age. The ISIS (Infant Study of Inhaled Saline) study, a randomized controlled trial of HS in infants and young children, demonstrated no effect of HS on pulmonary exacerbations, the primary endpoint, but did demonstrate a significant effect on infant lung function in a substudy and on the lung clearance index (LCI, a measure of ventilation inhomogeneity) in a singlecenter ancillary study. We aimed to assess the treatment effect of HS on the LCI in preschool children with CF in a large, multicenter clinical trial. Our primary hypothesis is that, compared to isotonic saline (IS), HS will improve the LCI during the 48-week treatment period.

Methods: The SHIP (Saline Hypertonic in Preschoolers) Study is a randomized, double-blind, controlled trial of inhaled 7% hypertonic saline (HS, active agent) vs. 0.9% isotonic saline (IS, control agent) for 48 weeks in preschool patients with CF, in order to evaluate a treatment effect of HS on LCI. As secondary analyses, we are also evaluating longitudinal associations of LCI with spirometric indices and clinical endpoints. Participants have a confirmed diagnosis of CF, ages ≥36 to ≤72 months at screening, have not used HS in the prior month and were able to provide acceptable LCI results at screening and enrollment. IRB approval was obtained at each site as well as parental/guardian informed consent. Participants were pre-treated with albuterol and randomized to inhale 7% HS or 0.9% IS twice daily for 48 weeks. The primary endpoint is the change in LCI via nitrogen multiple breath washout from enrollment to Week 48. Secondary endpoints include FEV 0.75 measured by preschool spirometry, protocol-defined pulmonary exacerbations, modified CFQ-R, the CF Respiratory Sign Diary (an observer-reported outcome) and rates of treatment-emergent CF respiratory pathogens. All sites were trained and certified in multiple breath washout and preschool spirometry, and all tests are over-read for acceptability.

Interim Result: 188 participants were screened and 150 randomized at 25 North American sites between 4/21/2015 and 08/04/2017. At enrollment. mean (SD) age was 54.2 (11.0) months; 54.0% were female; mean (SD) weight and height percentiles were 52. 2 (25.8) and 48.2 (26.0), respectively; mean (SD) LCI 2.5 and LCI 5.0 were 9.3 (2.0) and 6.0 (0.8), respectively; and mean (SD) FEV 0.75 was 0.81 (0.30) L. To date, overall multiple breath washout and spirometry measurement acceptance rates are 88% and 82%, respectively. Last patient visit will occur in July, 2018 and results will be available in September, 2018 Introduction: Pulmonary exacerbations (PEx) are common in people with cystic fibrosis (CF), yet there are limited data to define optimal treatment decisions, thus intravenous (IV) antibiotic therapy to treat PEx varies widely in the US. STOP2 (Standardized Treatment Of PEx 2) is an ongoing multicenter, randomized, controlled open-label trial that evaluates early response to PEx IV antibiotic treatment, where early robust responders (ERR) are eligible for a shorter treatment duration, and non-ERR (NERR) are candidates for a longer IV antibiotic course. We report STOP2 metrics at 60% enrollment.

Methods: CF adults ≥18 years of age treated for a PEx with IV antibiotics at 58 sites (up from 33 at the same time last year) are being enrolled to evaluate safety and efficacy of differing durations of IV antibiotic treatment. Treatment can be at home or inpatient, airway clearance and continuation of chronic medications are encouraged, and antibiotic dosing and selection guidelines are provided. Participants' forced expiratory volume in one second (FEV 1 ) and symptoms [using the Chronic Respiratory Infection Symptom Score (CRISS)] are measured. Those exceeding the thresholds of an 8% predicted FEV 1 increase and 11 point CRISS improvement on Day 7-10 are classified as ERR and randomized 1:1 to 10 vs 14 days total IV antibiotic treatment; the remaining are classified as NERR and are randomized 1:1 to 14 vs 21 days. The primary endpoint is change in FEV 1 from treatment start to 14 days after scheduled end of IV antibiotic course. The study plans to evaluate 310 ERR participants to test the hypothesis that 10 days is not inferior to 14 days of IV antibiotics, and 570 NERR to test the hypothesis that 21 days is superior to 14 days of IV antibiotics.

Results: An open April 2018 report to the CF Foundation Data Monitoring Committee to assess safety and protocol adherence had 626 patients screened with 580 (93%) randomized (vs an expected 15% loss before randomization). Enrollment was approximately 60 participants ahead of schedule. Of those randomized, 49% were female, 61% were 18-30 years old, 58% had FEV 1 <50% predicted at treatment start, 58% had ≥2 PEx in the year prior, 24% were treated entirely outpatient, and 11% received oral steroids prior to randomization. 416 (72%) were NERR while 164 were ERR; a 2.5:1 ratio which is closer to the projected 2:1 ratio than was observed a year prior at the first interim (3.4:1) . 18 (11%) ERR participants deviated from randomized IV duration, and 72 (17%) NERR deviated. At the time of the interim there were 38 withdrawals (7%), vs an expected withdrawal rate of 15%.

Conclusions: Persons with CF experiencing a PEx are rapidly enrolling in STOP2. Fewer participants are meeting ERR criteria than expected, but withdrawals before and after randomization are less than anticipated. STOP2 has the potential to transform PEx care standards by identifying optimal treatment durations -a necessary first step to investigate other facets of PEx treatment.

Acknowledgments: This work was supported by CFF. The authors acknowledge TDN sites, CF patients and families. (Kerem E, et al. N Engl J Med. 1992; 326:1187-91) . Since the amount of time a CF patient spends on the lung transplant wait list is dependent on allocation score and can be up to 2 years, understanding characteristics involved in a patient's probability of a lung transplant is important. This study seeks to develop models to identify characteristics that affect the probability of, and time to lung transplant. The models are translated into nomograms to provide physicians with a graphical representation to explain to patients how their characteristics are related to their lung transplant prognosis.

Methods: A subset of the Cystic Fibrosis Foundation Patient Registry (CFFPR) was analyzed from 2003 and 2014, including pediatric and adult CF patients aged 6-40 years (n=28,797). The CFFPR is composed of health care information from US CF patients who are treated at CF Foundation-accredited care centers and have consented to participate. Logistic and Cox regression models were used to predict the probability of a lung transplant and the time to lung transplant, respectively. Predictors considered were: weight, height, sex, most recently documented age (age at review), FEV 1 (percent predicted), Pseudomonas aeruginosa infection, forced vital capacity (FVC) (percent predicted), reason for diagnosis, F508del genotype, baseline insurance status, methicillin-resistant Staphylococcus aureus (MRSA) infection, Burkholderia cepacia infection, CF-related diabetes (CFRD), pancreatic enzyme usage, ibuprofen usage, corticosteroid therapy, race, and number of overnight pulmonary hospitalizations in most recently documented year. Significance of predictors (p ≤ 0.05), Akaike information criterion (AIC), and concordance index were used to select the final models.

Results: The final Cox regression model identified 5 predictors as significant: height, number of overnight pulmonary hospitalizations, age at review, CFRD, and baseline insurance status. The following additional predictors were significant in the final logistic regression model: weight, F508del genotype, reason for diagnosis, pancreatic enzyme usage, FEV1 percent predicted, and race. The concordance indices were 0.86 and 0.84, respectively. Nomograms based on the final logistic and Cox regression models were developed to explain lung transplant prognosis.

Conclusion: With nomograms demonstrating the probability of and time to lung transplant, physicians can communicate to CF patients how their specific set of characteristics typically affect lung transplant prognoses. This may allow physicians and patients to work together to decide changes in treatment to prevent or delay the need for a lung transplant, as well as agree on a threshold for listing the patient for a lung transplant.

Hinds, D. 1 ; Hatch, J.E. 2 ; Ren, C.L. 2 ; Sanders, D.B. 2 1. Pediatrics, Indiana University School of Medicine, Indianapolis, IN, USA; 2. Pediatric Pulmonology, Allergy, and Sleep Medicine, Indiana University School of Medicine, Indianapolis, IN, USA Introduction: The majority of pulmonary exacerbations (PEx) in children with CF are treated with oral antibiotics (ABx), although to date, descriptions of oral ABx treatments only include those prescribed at clinic visits. We hypothesized that the majority of oral ABx are prescribed over the phone, and treatment outcomes are worse for patients seen in clinic compared to those prescribed ABx over the phone.

Methods: We performed a retrospective chart review of oral ABx prescriptions for children with CF followed at the Riley Hospital for Children CF Center from July through Dec 2016. Patients whose initial ABx during this time frame were prescribed over the phone were grouped for comparison against those first treated in clinic. We documented treatment location, the duration and description of symptoms, and treatment with IV ABx in the following three months. We recorded the best FEV 1 in the 6 months before and 3 months after treatment, and, for patients treated in clinic, at the time of treatment. This project was reviewed and approved by our local IRB.

Results: 156 (of approximately 330 total patients at our center) children with CF received 242 courses of oral ABx in Jul-Dec 2016; 56% of all prescriptions were provided over the phone. ABx choice did not differ between phone and clinic, with all p-values >0.05 (Table) . For the first prescription per patient in the study window, ABx were prescribed equally in clinic and over the phone. Most demographics, clinical characteristics, and respiratory cultures results were not significantly different between the two groups. Children provided phone ABx were more likely to have private insurance (60% vs 38% of clinic ABx, p=0.01), to report symptom duration of < 7 days (81% vs 47%, p=0.013), and report cough (94% vs 80%, p= 0.02). Documentation of symptom duration was missing for 14% and 56% of phone and clinic ABx, respectively. There were no differences between phone and clinic ABx in extension of oral ABx (17% vs 6%, p=0.08), or treatment with IV ABx in the subsequent 3 months (20% vs 29%, p=0.27). Among 42 and 39 children treated in phone and clinic with FEV 1 at both time points, 76% and 87%, respectively, recovered to 6-month baseline FEV 1 (p=0.26) in the subsequent 3 months.

Conclusion: At our CF center, over half of oral ABx are provided over the phone. Short-term outcomes were not different for phone or clinic ABx, but ABx prescribed by phone should be accounted for in analyses of pulmonary exacerbations. Further analysis and research is required to determine the clinical significance of the differences in insurance status and symptom duration we observed in the phone ABx group. Methods: In this IRB-approved prospective study, 6 cystic fibrosis (CF) patients (mean 11.8±3.8 years) admitted for an APE were recruited. Two patients were recruited from each FEV1 (percent predicted) category (mild ≥ 70, moderate 50 -70, severe ≤ 50). Signed informed assents/ consents were obtained. Subjects completed a chest MRI, using UTE and T1 mapping (modified look-locker inversion recovery), spirometry and CFRSD-CRISS questionnaire within 24 hours of admission (time 0), at 1 week (time 1) and at 2 weeks (time 2). Two cases did not complete last MRI (time 2 MRI). MRI scoring (Eichinger, et al. Eur J Radiol. 2012; 81(6) :1321-9) was performed by two experienced pediatric radiologists blinded to the patients' names and dates of studies. T1 mapping was used as a surrogate for perfusion images for functional scoring. A region of interest analysis on T1 mapping images was used to calculate the mean normalized T1 (nT1) values for all lobes. The percent lung volume without T1 value (T1 null) on the system generated analysis map was calculated. The primary outcome was to assess the differences in mean nT1, percent volume of T1 null, and MRI scores during treatment. Correlation between the mean nT1, T1 null volume and MRI scoring with spirometry and CFRSD-CRISS questionnaire was evaluated. Statistical analysis was performed by t-test, linear mixed effects model, Pearson, Spearman's and Kendall's correlation tests. P<0.05 was considered significant.

Results: All but one case showed decrease in whole lung percent volume of T1 null from time 0 to 1 (p 0.02), however, no changes were noted between time 1 and 2 (p 0.8). Similarly, the linear mixed effects model suggests that the whole lung mean nT1 value is increasing by 90.52 in the next visit compared to the previous visit on average, between time 0 and 1 and time 1 and 2 (p 0.002). There was a significant decrease in questionnaire scoring (p 0.01) and an increase in the FEV1 (p 0.04) between time 0,1 and 2. No significant difference in FEF 25-75 was detected during treatment. There was a strong correlation between T1 null percent volume and FEV1 (p 0.004) and FEF 25-75 (p 0.03) for all MRIs. No correlation between mean nT1 value, MRI scoring, T1 null percent volume, FEV1 and FEF 25-75 with the questionnaire was found (all p>0.05). MRI scoring correlated with spirometry results only on time 0 (p 0.01 and p 0.04). There was no significant change in MRI scoring over time.

Conclusion: MR T1 mapping sequence can be used as an imaging biomarker in the evaluation of CF pulmonary exacerbations. It is a safe imaging method, without side effects of radiation or contrast agent, to detect CF pulmonary regional disease and as an objective measure for clinical trials. Larger scale studies are needed. London, United Kingdom; 2. Great Ormond Street Hospital for Children NHS Foundation Trust, London, United Kingdom; 3. Hospital for Sick Children, Toronto, ON, Canada; 4. Royal Brompton Hospital, London, United Kingdom; 5. NHLI, Imperial College, London, United Kingdom; 6. Kings College Hospital, Univ. Hospital Lewisham and Royal London Hospital, London, United Kingdom Introduction: Preservation of lung health is essential for maintaining quality of life and extending survival in patients with cystic fibrosis (CF). In longitudinal studies of clinically diagnosed CF children lung clearance index (LCI) becomes abnormal earlier than spirometry, and pre-school children with elevated (abnormal) LCI have worse lung function at school age (Am J Respir Crit Care Med. 2011; 183:752-8) .

Hypothesis: Pre-school LCI is a sensitive early marker of abnormal lung function during adolescence. We aimed to relate pre-school LCI to spirometry and LCI in adolescence, and determine whether LCI tracked through childhood.

Methods: The clinically diagnosed London Cystic Fibrosis Collaboration (LCFC) cohort was recruited from specialist paediatric CF clinics in London between 1999 London between -2002 . The cohort has been followed up at pre-school, school age, and now adolescence. Adolescents between 12-17 years with at least one prior test within the LCFC performed spirometry and multiple breath washout (respiratory mass spectrometer using SF 6 as a tracer gas). The primary outcome was forced expiratory volume in one second (FEV 1 ) z-score at adolescence. An abnormal FEV 1 was defined as <-1.96 z scores. Mixed effects linear regression was used to determine LCI tracking through childhood. Structural equation models were used to conduct a mediation analysis to determine the direct effects of pre-school LCI on adolescent LCI, independent of school age LCI.

Results: At adolescence, 43 subjects with CF were tested. The mean (SD) zFEV 1 was -1.61 (1.5); 30% (13/43) of subjects had abnormal FEV 1 . There was a weak but statistically significant correlation between pre-school LCI and adolescent zFEV 1 (r= -0.41, p=0.01); children with higher LCI at pre-school were more likely to have a lower adolescent zFEV 1 . There was no significant correlation between pre-school and adolescent zFEV 1 (r=0.28, p=0.08). Pre-school LCI had a sensitivity of 91% and specificity of 46% for abnormal adolescent FEV 1 , whereas pre-school spirometry had a sensitivity of 42% and specificity of 90%.

There was significant tracking of LCI throughout childhood (slope coefficient (95% CI): 0.17 (0.12; 0.22); p <0.001; intraclass correlation = 0.67). Moreover, pre-school LCI, independent of school age LCI, was associated with adolescent LCI.

Conclusions: Pre-school LCI tracks through to adolescence, and is associated with adolescent zFEV 1 . These findings further emphasize the window of opportunity to intervene to prevent the development of lung disease at an early age. The mechanism by which such pathways can alter disease trajectories, and how best to prevent this, requires further investigation, but elevated pre-school LCI is an ominous warning sign of impaired lung function in adolescents.

Acknowledgments: Funding: UK Cystic Fibrosis Trust and National Institute for Health Research (NIHR).

Muston, H. 1 ; Slaven, J.E. 2 ; Clem, C. 1 ; Tiller, C.J. 1 ; Davis, S.D. 1 ; Ren, C.L. 1 1. Pediatric Pulmonology, Indiana Univ. School of Medicine, Indianapolis, IN, USA; 2. Biostatistics, Indiana Univ. School of Medicine, Indianapolis, IN, USA Rationale: Abnormal lung function in cystic fibrosis (CF) is present early in life, and nutritional status in early childhood affects lung function in later life. There are limited data on which infant pulmonary function test (iPFT) measurements are most sensitive to detecting early lung disease.

Objective: To describe the results of clinically-indicated iPFTs in infants with CF at 12-24 months of age who were diagnosed by newborn screen (NBS).

Methods: We performed a retrospective chart review of CF NBS infants who had iPFTs obtained at 12-24 months of age at Riley Hospital for Children from 2012-2017. Forced expiratory flows (FEF) were obtained via the raised volume rapid thoracoabdominal compression technique and functional residual capacity (FRC) was measured via plethysmography using the Jaeger BabyBody Device (Carefusion). Demographics and clinical data were obtained from the electronic medical record. Chi-square analysis was performed comparing the number of infants with significantly elevated FRC to the number of infants with significantly diminished FEF. Infants were considered to have significant hyperinflation or obstruction if the z-score was ≥ 2 for FRC or ≤ -2 for FEF.

Results: We identified 23 infants for this study. Characteristics of this cohort included 47.8% males, 95.7% were Caucasian and 87% of subjects had Class I-III CF genetic mutations. Respiratory culture results included a high prevalence of Staphylococcus aureus (70%) and Pseudomonas aeruginosa (48%), consistent with previously described studies. Mean birth weight z-score for the cohort was -0.57. Over the first year of life, 91% of subjects received at least one course of oral antibiotics, with a median of 2 antibiotic courses. The mean age at iPFT testing was 74.2 weeks. The iPFT results are summarized in the Table. The mean FVC z-score was -0.56 (95% CI: -1.09 to -0.04). All other FEF measures were within the normal range. In contrast, the mean FRC z-score was 2.80 (95% CI: 1.67 to 3.94). An abnormal FRC (z-score ≥ 2) was seen in 52% of infants, whereas an abnormal FEV 0.5 (z-score ≤ -2) was abnormal in only 13% of the group (p=0.011) and an abnormal FVC (z-score ≤ -2) was abnormal in only 4% of the group (p<0.001).

Conclusion: In this cohort of US infants diagnosed with CF through NBS, FEF measures were for the most part normal, but hyperinflation was common by 1 year of age. Our results suggest that FRC may be a more sensitive measure of early lung disease in this population compared to FEF measures. procoagluant factor FVIII activity, platelet hyperreactivity, and antiphospholipid antibodies. Additional risk factors such as surgery, immobility, and frequent placement of peripherally inserted central catheters also exist in this population. Current treatment guidelines for VTE recommend long-term anticoagulant therapy with a direct oral anticoagulant over the vitamin K antagonist warfarin. Patients with CF diagnosed with VTE pose unique therapeutic challenges stemming from malabsorption of vitamin K, and often have underlying liver dysfunction to further complicate blood coagulation and metabolism of anticoagulants. In addition, many medications prescribed to patients with CF have pharmacokinetic properties that alter the metabolism of anticoagulant drugs. Despite these challenges, there are currently no recommendations for VTE management specific to patients with CF.

Methods: The primary objective was to survey Cystic Fibrosis Foundation (CFF)-accredited care centers in the U.S. to describe current VTE treatment practices. Secondary objectives included characterizing variance in treatment patterns and challenges in patients with CF diagnosed with VTE. An online survey was distributed via e-mail by the CFF. Over a 6-week period, respondents completed the survey and results were captured via REDCap and summarized using descriptive statistics.

Results: There was a total 71 respondents from 126 CFF-accredited care centers in the U.S (response rate 56.3%). An average center treats 0-5 VTE episodes per year, with some treating upwards of 20 episodes. Centers reported using the following anticoagulants: unfractionated heparin (UFH) (18.3%), low-molecular-weight heparin (LMWH) (71.8%), fondaparinux (5.6%), warfarin (35.2%), dabigatran (4.2%), rivaroxaban (33.8%), apixaban (36.6%), edoxaban (1.4%), betrixaban (1.4%), and argatroban (1.4%). Centers had an average confidence level (scale of 0-100 with 100 being extremely confident) of 51 in managing anticoagulant therapy in CF patients. When asked to what extent a VTE management guideline specific to CF patients would be helpful (scale of 0-100 with 100 being extremely helpful), the average response was 74. Frequently identified VTE treatment challenges for centers were navigating drug-drug interactions and managing therapeutic complications. For common VTE scenarios, centers responded that pediatric CF patients are most often treated with LMWH and warfarin, while adult CF patients are treated with a variety of anticoagulants including LMWH/UFH and warfarin, apixaban, and rivaroxaban. Treatment duration was often for 3 to 6 months, although some patients were treated indefinitely.

Discussion: Results indicate drug interactions between CF and anticoagulant therapies are often encountered and that choice of anticoagulant differs between pediatric and adult patients. Overall, centers have low confidence in managing VTE in patients with CF and most respondents indicated that a CF-specific VTE treatment guideline would be helpful.

Drevinek, P. 1 Introduction: Add-on therapy of chronic Pseudomonas aeruginosa infection with macrolide azithromycin became part of standard CF care due to its anti-inflammatory properties and the ability to interfere with some of the microbial virulence factors. Clinically significant improvement in lung function, weight gain and reduction in number of pulmonary exacerbations (PEx) have been demonstrated for the first 6 to 12 months of the treatment, but benefits from continuation of the therapy beyond 12 months were not evident. Keeping P. aeruginosa infected patients on a long-term azithromycin maintenance therapy is still a common practice, despite lack of information about its optimal duration or positive impact on patients' health status in the long run. To evaluate usefulness of a prolonged azithromycin treatment, we analyzed data from 25 CF patients who stopped their therapy with azithromycin after two or more years of administration.

Methods: Following parameters were collected from patient medical records and the national Czech CF registry: date of birth, start and stop date of azithromycin therapy, best annual FEV 1 percent predicted (pp) value, number of PEx per calendar year and yearly Staphylococcus aureus infection status including its susceptibility to macrolides.

Results: Mean duration of azithromycin use was 6.8 years (range 2.1 to 14.7 y), mean age at initiation of therapy was 12.2 (2.1 to 29.6), mean age at the cease of therapy was 19.0 (10.7 to 31.8). FEV 1 pp at the last year on therapy and at the following year differed by 1.875 (CI -1.09, 4.84; p=0.2035) , a mean change in number of PEx was 0.5 (CI -0.28, 1.28; p=0.1965) . Of note, the change of best annual FEV 1 pp measured two years vs. one year before the end of treatment was 8.15; p=0.079) . Resistance of S. aureus to erythromycin was reported in 12 patients while on therapy. In 5 of them, the bacterial isolates became susceptible to macrolides within the first year post-azithromycin.

Conclusions: We did not observe any beneficial effect of a years-long azithromycin treatment on clinical outcome measures such as FEV1 or PEx. The decrease in lung function during the last two years on therapy was on average higher than between the last year on therapy and the first year after the cease of azithromycin. Norling, T. 5 1. CaRACS, Berlin, Germany; 2. Xellia doo, Zagreb, Croatia; 3. SHL Group AB, Nacka Strand, Sweden; 4. Medspray BV, Enschede, Netherlands; 5. Pharmaero ApS, Copenhagen, Denmark A new inhalation product (ColistAir ® ) providing a liquid formulation of colistimethate sodium (CMS) for inhalation together with a novel aqueous droplet inhaler (ADI) device was developed for the management of Pseudomonas aeruginosa lung infection in cystic fibrosis (CF) patients. The ADI is a compact, portable device comprising a reusable spring-operated powerpack requiring no electricity and disposable mouthpieces intended for single day use. CMS is supplied as a ready-to-use stable liquid formulation in prefilled cartridges, which are inserted into the mouthpiece by the user. Each cartridge contains the daily treatment, i.e. 2 doses and is discarded daily together with the used mouthpiece. A new mouthpiece and cartridge are used each day, hence no cleaning of the device is required, offering greater convenience to patients compared to other devices.

An open-label, crossover study was conducted in healthy volunteers (HV) to assess the safety, tolerability, pharmacokinetics (PK) and lung deposition of a single CMS dose administered as ColistAir ® in comparison to 2 MIU CMS nebulized via eFlow ® and to the dry powder CMS inhalation product Colobreathe™ (1.66 MIU CMS) . Several dose levels of ColistAir ® were initially tested in the present study to assess the number of actuations required to achieve plasma exposures similar to the 2 comparators. The selected ColistAir ® dose (10 actuations, corresponding to 1.4 MIU CMS) and the 2 comparator products were then administered to 18 HV in a crossover design. Safety, tolerability and PK where evaluated in all 3 treatment arms, whereas lung deposition was assessed by scintigraphy in the ColistAir ® and nebulized CMS (eFlow ® ) arms only, where the test and reference product were radiolabeled with technetium-99m ( 99m Tc).

No serious or severe adverse events (AE) were reported. All treatment regimens were well tolerated. The most common treatment related AE was mild cough, predominantly in the Colobreathe™ arm. Mean lung deposition was greater for ColistAir ® St. Michael's Hospital, Toronto, ON, Canada; 2. Div. of Pulmonology, Lausanne University Hospital, Lausanne, Switzerland; 3. Div. of Allergy and Clinical Immunology, St. Michael's Hospital, Toronto, ON, Canada Introduction: Cystic fibrosis (CF) patients are prone to allergic bronchopulmonary aspergillosis (ABPA), a complex immunologic response to Aspergillus antigens. Omalizumab, a recombinant monoclonal anti-IgE antibody, has been used off-label for difficult-to-control ABPA. The aim of this study was to evaluate treatment efficacy and safety of omalizumab in adult CF patients with ABPA, a population for which data are extremely limited in the literature.

Methods: All adult CF patients who received omalizumab in the Toronto adult CF center between 2005 and 2017 were eligible for this study. Patients fulfilling at least the minimal diagnostic criteria for ABPA (Stevens DA, et al. Clin Infect Dis. 2003; 37:S225) were included in the analysis. We compared changes in FEV 1 (max FEV 1 percent (%) predicted, slope, area under the curve -AUC), use of systemic corticosteroids (SCS), use of IV antibiotics and days of hospitalization of the period before and up to 1 year of omalizumab treatment. Comparisons were performed with the Wilcoxon signed-rank test. A linear mixed effects model was used for FEV 1 % predicted slope and the Trapezoidal rule for FEV 1 % predicted AUC.

Results: During the study period, 11 of 27 (40%) adult CF patients receiving omalizumab fulfilled ABPA criteria (female 55%, median age 29 years). The median IgE was 889 IU/mL, 55% had positive skin test and 27% had specific IgE for Aspergillus. Colonization with Aspergillus was found in 82%. Diabetes was reported in 82%, nasal polyps in 46% and drug allergies in 64% of patients. At omalizumab initiation, 10 patients were on SCS and 3 on systemic antifungals. Omalizumab was initiated due to one or a combination of the following: poor control despite first line treatment (n=8), secondary effects to SCS or antifungal agents (n=8).

Within 1 year of follow-up, 2 patients died and 2 were transplanted. No omalizumab-related adverse effects were observed. The max FEV 1 % predicted was not significantly different (median change 3.9, p=0.12) for the period before and while on omalizumab treatment but the rate of FEV 1 decline (median change of slope -0.021, p=0.019) and the variability of FEV 1 (median change of AUC 1099.5, p=0.027) improved significantly on omalizumab. The cumulative dose of SCS was not significantly different (p=0.37) but 4 (36%) patients decreased the SCS dose by more than 50% compared to baseline. Days on IV antibiotics (p=0.51) and hospital days (p=0.097) did not differ significantly before and while on omalizumab therapy.

Conclusion: In adult CF patients with ABPA not responding to first line treatment or presenting with significant secondary effects due to SCS and/or antifungal agents, omalizumab should be considered. In this cohort, omalizumab was well tolerated and decreased the rate of FEV 1 decline and the variability of FEV 1 . A subset of patients decreased the dose of SCS by more than 50% of baseline. Larger studies are needed to identify patient characteristics that may predict response to omalizumab.

De Groof, J.; Vermeulen, F.; Proesmans, M.; Boon, M.; De Boeck, K. CF Reference Centre, University Hospitals Leuven, Leuven, Belgium Introduction: Routine use of chest CT in the follow-up of patients with cystic fibrosis (CF) remains controversial, the most important disadvantage being the associated radiation exposure early in life. Together with the increasing life expectancy, long-term effects of radiation are not to be ignored. The aim of this study is to assess the cumulative effective dose (CED) of all CT and conventional radiography (CR) examinations in patients with CF in order to improve future dose estimations.

Methods: The effective dose (ED) of each CT and CR for paediatric patients with CF, followed for at least one year between 2000 and 2017 in the Leuven CF centre, were calculated retrospectively, using console data for each CT and age-specific conversion factors. Published ED estimates were used to approximate CR EDs. From 2005 on, routine chest CT once every 2 years, starting from the age of 3 years, was included in the follow-up protocol of the Leuven CF centre.

Results: A total of 221 patients were included for a total of 2083 patient years, 1175 CT examinations (with 960 chest CTs) and 6207 CRs (with 5182 chest x-rays). The median total CED for each patient was 11.70 mSv with a p25 and p75 of 4.74 mSv and 22.47 mSv respectively. Median annual CED was 1.23 mSv with a p25 and p75 of 0.72 mSv and 1.91 mSv respectively. On average chest CT accounted for 67% and chest x-ray for 19% of the total CED. There was no significant difference in median annual CED between the 2000-2005 and 2012-2017 periods (0.82 mSv vs 0.83 mSv respectively). In the same periods, the number of annual CTs per patient increased from 0.45 to 0.64 (p 0.03), annual chest CTs per patient increased from 0.35 to 0.54 (p <0.01). Evolution of chest CT ED according to time period and age at time of examination was calculated (Table) .

Discussion and Conclusion: Chest CT EDs in our study population were comparable to reference ranges found in literature. Despite the use of routine chest CT from 2005 on and thus an increase in the annual number of chest CTs, median annual CED did not significantly change, reflecting dose reductions attained with the use of CF-specific low dose chest CT protocols, especially in patients 0-15 years old. Future technological advances and development of new ultra-low dose CT protocols will most likely reduce ED even further. MUSC, Charleston, SC, USA; 2. Microbiology and Molecular Genetics, Univ. of Pittsburgh, Pittsburgh, PA, USA; 3. Pulmonary, Allergy and Critical Care Medicine, Univ. of Pittsburgh, Pittsburgh, PA, USA; Univ. of Pittsburgh, Pittsburgh, PA, USA Over 95% of individuals with CF have radiographic sinus disease, but 25-75% of patients report symptoms. Predictors of more symptomatic sinus disease are poorly understood. Our objective was to determine which clinical variables predict more severe chronic rhinosinusitis (CRS).

Methods: In an observational study, we analyzed a longitudinal cohort of 33 adults with CF and symptomatic sinus disease who were referred to CF otolaryngology clinic from 2015-17. Study visits coincided with routine or symptom-driven clinical visits. A rigid nasal endoscopy was performed as clinically indicated. Objective and subjective disease severity were measured using the Lund-Kennedy scoring (mLK) and the Sino-Nasal Outcome Test (SNOT22). Samples were obtained for culture.

Results: Participants were followed for a mean of 14.7 months. The cohort had a mean age of 31.5 years and a mean FEV 1 of 62.9% predicted (IQR 49-89); 13/33 were homozygous for F508del. CF-related diabetes (CFRD) was diagnosed in 42% and 48% had allergic rhinitis. Endoscopic sinus cultures most commonly revealed P. aeruginosa (64% of subjects) and S. aureus (64%). The mean SNOT-22 score was 38.7 . At baseline CFRD and sinus exacerbation visit were both associated with higher SNOT-22 scores (p<0.05, Kruskal-Wallis test). Age, BMI, FEV 1 , F508del homozygosity and sex were not associated with worse sinus symptoms at baseline. Using mixed effects models, CFRD was associated with a 3.8 point increase in SNOT22 scores per year (fitted model shown in Fig) . Nasal steroid use was not associated with different baseline SNOT22, but it was associated with 3.2 point decrease in severity per year. Endoscopic disease (mLK scales) was higher with CFRD (1.87 points, p<0.005), and at exacerbation visits (1.29 points, p=0.006). Nasal steroid use was associated with a 2.24 point decrease in mLK over all visits (p<0.001), while allergic rhinitis was associated with a 1.8 point increase in mLK per year (p<0.001).

Conclusions: CFRD, allergic rhinitis, and nasal steroid use may modulate CRS disease severity in adults.

Acknowledgments Objective: To characterize clinical presentations of CF lung disease in infancy.

Methods: The study population were infants enrolled in FIRST (Feeding Infants Right… from the STart), an ongoing multicenter prospective observational study to identify optimal feeding. The present study utilized data from 159 children who reached age 12 months (mo) by 12/31/17. After excluding 15 children due to early withdrawal, low birth weight, and missing data, 144 children enrolled at 1.7±1.0 mo were included in the final analyses. Respiratory outcomes, i.e., positive Staphylococcus aureus cultures (Sa), positive Pseudomonas aeruginosa cultures (Pa), hospitalizations, antibiotics treatments (Ab), chronic antibiotics use (Chronic Ab), acute upper respiratory infections (AUI), and pulmonary exacerbations (pEx), were obtained at routine CF center visits using questionnaires filled out by physicians or study coordinators. We applied a k-means clustering analysis to jointly examine frequency of pulmonary outcomes aforementioned at ages 0-6 mo and 6-12 mo.

Results: Of 144 children, 25 had meconium ileus (MI), 114 had no MI but pancreatic insufficiency, and 20 were pancreatic sufficient. At 0-6 mo, the prevalence of Pa was 9.7%, Chronic Ab 4.2%, Sa 52.1%, hospitalization 28.5%, Ab 46.5%, AUI 66.7%, and pEx 25.7%. At 6-12 mo, the percentages were 11.5%, 9.7%, 60.4%, 23.6%, 63.2%, 83.3%, and 34.7%. The clustering analysis identified 3 clusters, which appear similar at both 0-6 and 6-12 mo (Figure) . On average, Cluster 1 had little lung disease; Cluster 2 had elevated Sa, but few other lung disease outcomes; Cluster 3 had several markedly increased disease outcomes, including hospitalization, Ab, AUI, and pEx. The majority of infants in cluster 1 (64%) and 2 (73%) at 0-6 mo remained in the same clusters as at 6-12 mo, but half of Cluster 3 (52%) switched to cluster 1 or 2 at 6-12 mo.

Conclusion: Our preliminary analyses reveal distinct lung disease presentations in CF infants.

Acknowledgments: Supported by R01DK109692, R01HL113548. Introduction: Acute pulmonary exacerbations (PEx) are currently treated with oral or intravenous antibiotics, yet inhaled antibiotics may represent a better choice since they achieve much higher concentrations at the site of action with less systemic exposure.

Objectives: The objectives of this study are to investigate the clinical and microbiological effects of treating PEx with aztreonam lysine for inhalation (AZLI) plus one intravenous antibiotic.

Methods: An open-label randomised cross-over study was designed to include 32 exacerbations from 16 subjects at a large adult CF centre in the UK. Key inclusion criteria include sputum growth of P. aeruginosa within the last 6 months and no previous use of AZLI. Over the course of consecutive hospitalisations patients received two different treatments (Treatment A: AZLI plus intravenous colistimethate. Treatment B: Two intravenous antibiotics as chosen by admitting CF physician). Primary outcome is average difference in change of FEV1 at 14 days between each arm. Secondary outcomes include time to next exacerbation, quality of life measures and sputum microbiology (CFU counts and 16S rRNA sequencing).

Results: At the time of interim analysis 9 study subjects had completed both arms of the study. At day 7 there was no difference between the average change in FEV1 between the AZLI arm and the IV arm (+9% vs. +11%, p=0.41). At day 14 the AZLI arm had greater improvement in FEV1 (+16% vs. 11%, p=0.008). The AZLI had a numerically longer time to next exacerbation but this was not statistically significant (119 vs. 57.5 days, p=0.6).

Conclusion: Study completion results expected September 2018 and are to be presented at NACFC 2018.

Rotolo, S.M. 1 ; Collins, J. 1 ; Dowell, M. 2 1. Pharmacy, Univ of Chicago Medicine, Chicago, IL, USA; 2. Pediatrics, Univ. of Chicago Medicine, Chicago, IL, USA Introduction: Azithromycin's high oral bioavailability, mild side effect profile, and relative lack of drug interactions in addition to its anti-inflammatory and immunomodulatory effects make it a beneficial prophylactic agent for patients with a variety of disease states. Despite these advantages, the risk of life-threatening arrhythmia due to QTc prolongation may limit its use. This is well-documented with treatment doses of azithromycin, but studies regarding EKG monitoring with the use of prophylactic azithromycin are lacking. Additionally, with increased focus on mental health screening in the CF population, more patients may be connected to care for their depression and anxiety, potentially leading to initiation of antidepressants or anxiolytics associated with QTc prolongation.

The purpose of our study was to determine the frequency of QTc monitoring in patients taking prophylactic doses of azithromycin concomitantly with an antidepressant, and whether or not patients that were monitored experienced QTc prolongation in the inpatient setting. The primary objective was to describe the incidence of QTc monitoring. The secondary objectives were to identify patient-specific factors associated with QTc monitoring and QTc prolongation.

Methods: This study was a single-center, retrospective chart review conducted at University of Chicago Medicine. Eligible patients included those who received both azithromycin and an antidepressant during an inpatient stay, identified via a computer-generated report, regardless of indication for use. We collected patient age, sex, weight, azithromycin dosing regimen and indication for use, antidepressant agent, whether or not monitoring was done, and QTc interval in patients who did receive monitoring.

Results: 73 patients met inclusion criteria. This included patients with COPD, CF, asthma, lung transplant, HIV/AIDS, and other indications. The incidence of QTc monitoring in patients on concomitant prophylactic azithromycin and antidepressant therapy in the inpatient setting was 84.9%. Patients with a diagnosis of COPD were significantly more likely to be monitored than those with other disease states (100%, p=0.014), and those with CF were least likely to have been monitored (56%, p=0.035). Of patients that were monitored, 25.8% had an episode of QTc prolongation. These patients were younger in age (median 35.5 vs 59 years, p=0.0395) than those who had normal QTc. Of the 5 patients with CF that received monitoring, 2 had prolongation.

Conclusions: Our results demonstrate that there is a risk of QTc prolongation in patients on concomitant prophylactic azithromycin and antidepressants. This is inconsistent with previous studies that reported no concerning EKG changes in CF patients on concomitant prophylactic azithromycin and antidepressants (Pediatr Pulmonol. 2015; 50(S41):352; Pediatr Pulmonol. 2017; 52(S47) :327). However, other authors have found borderline elevated QTc intervals in adolescent males on prophylactic azithromycin only (J Cyst Fibros. 2016; 15:192-5) . The variation in findings reinforces the need for larger, multicenter studies to define the true incidence of QTc prolongation and clear guidelines on appropriate monitoring for this patient population. The CF Foundation (CFF) guidelines for respiratory care includes recommendations regarding the use of inhaled corticosteroids (ICS). The 2013 recommendations are derived from results of numerous randomized controlled trials looking at ICS use in CF patients. These recommendations were reported as a "D," meaning the magnitude of net benefit was zero or negative in regards to giving ICS to patients without asthma or ABPA in ages greater than 6 to improve lung function, quality of life or reduce pulmonary exacerbations (Mogayzal M, et al. Amer J Respir Crit Care Med. 2013; 187:680-9) . Because of the lack of benefit and potential concern for harm with the use of ICS, namely due to risk of longitudinal growth impairment, the CFF follows ICS use in each of the certified CF centers and presents this data in their yearly report. In an effort to work towards lowering our center's use of ICS in CF patients, we looked closely at ICS use over time, differences by age group, ICS type and prescribing habits of different CF providers.

Methods: The 2016 CF Registry report for the Baylor College of Medicine CF center was reviewed, including ICS use in patients aged 6-17 who did not have a diagnosis of asthma or ABPA. In order to better approach quality improvement efforts to decrease ICS use, a query of our electronic medical record (EMR) was formulated to look at current ICS use versus the 2016 data. After receiving the data set, 4 pulmonary fellows sorted through patient charts individually to determine which patients had a diagnosis on their problem list of asthma or ABPA. We looked for justification of ICS use in their most recent clinic encounter (e.g history of wheezing). The data were then broken down by CF physician and provided to them individually for feedback and improvement purposes.

Results: Initial review of the queried data from our EMR shows a decrease in ICS use without a diagnosis of asthma or ABPA from 24.6% as noted in the 2016 CF Registry report to 13% in April 2018. Marked variability was noted amongst our CF providers as to their prescribing habits. Their individual patient populations vary from 4 to 65 total patients with a percentage of ICS use in these patients (without asthma/ABPA) ranging from 0 to 50%. Patients with documented asthma or ABPA were slightly more likely to be on an ICS/LABA combination inhaler (38%) instead of ICS alone when compared to patients who had documented justification for use of ICS that was not asthma or ABPA (27%). Those with a documented justification (not asthma or ABPA) were most likely to be on high dose ICS/ LABA (27%) vs other groups. Of those patients without documented justification for ICS use, 77% were on ICS alone (without a LABA).

Conclusions: Our CF providers have made great strides in improving both their documentation of asthma/ABPA diagnosis as well as their prudent use of ICS in our center. We found there is variability amongst providers as to how they utilize these medications and their documentation regarding that. We can continue to improve our use of these medications while understanding the individual practice style of our physician cohort is not homogeneous. Background: CF is a disease with equal prevalence across sexes, but prior studies have shown that women have worse outcomes, such as increased mortality. It is well known that pulmonary exacerbations (PEx) in CF patients can lead to decreased lung function, lower quality of life and shortened survival, but there is a paucity of data evaluating the existence of sex differences in frequency, treatment, and outcomes in PEx. The objective of our study was to evaluate sex-based differences in frequency of PEx, antibiotic duration, treatment location and treatment response at our institution.

Methods: This retrospective cohort study was conducted in CF patients ≥ 18 years of age who were seen in the Johns Hopkins Adult CF Center from June 2016 to June 2017. Patients with a history of lung-transplant were excluded. Baseline demographic variables were obtained and differences were determined using chi-squared testing for categorical variables and t-tests for continuous variables. The primary outcome of interest was the frequency of PEx between women and men. A PEx was defined as a worsening of respiratory symptoms coupled with treatment with intravenous (IV) antibiotic therapy. Secondary outcomes measured were antibiotic duration, treatment location and treatment response for those patients with pre-and post-PEx spirometry data. For comparisons between the groups, unpaired t-tests were used to compare continuous variables and Fischer's exact tests were used for categorical variables.

Results: There were 146 women and 158 men seen at the Johns Hopkins Adult CF center during the study period (mean age 33.9 years; 49% female, mean BMI 23.8). A total of 240 PEx occurred in 147 patients, with 138 PEx in 82 women and 102 PEx in 65 men. Of the 82 women, 51 had 1 PEx, 15 had 2 PEx, and 16 had ≥ 3 PEx in the study time period. Of the 65 men, 43 had 1 PEx, 12 had 2 PEx, and 10 had ≥ 3 PEx. Overall, 56% of women seen in the study period had at least 1 PEx compared to 41% of men (p=0.009). Mean duration of IV antibiotic treatment per exacerbation was 15.1 (SD 6.5) days for women compared to 14.5 (SD 7.8) days for men (p=0.46). There was a trend toward more women (24%) being treated solely at home with IV antibiotics compared to men (17%)(p=0.15). There were both pre-and post-PEx spirometry data for 127 of the PEx. Treatment response did not differ significantly between women and men, with 71.3% women having a 10% increase in FEV 1 (liters) compared to 70.2% of men (p=0.90).

Conclusions: In our single-center cohort study, women had a statistically significantly higher percentage of PEx. There was a trend towards women spending less time in the hospital for treatment of PEx, which can be associated with worse outcomes. It is unclear why women with CF have shorter survival than men, but our findings suggest increased frequency of PEx and different treatment patterns may contribute to worse outcomes in women.

Chadwick, H.K. 1, 2 ; White, H. 1, 2, 3 ; Shaw, N. 2 ; Etherington, C. 2 ; Clifton, I. 2 ; Pollard, K. 2 ; Whitaker, P. 2 ; Peckham, D. 1, 2 1. Respiratory Medicine, Leeds Institute of Biomedical and Clinical Sciences, Univ. of Leeds, Leeds, United Kingdom; 2. Adult Cystic Fibrosis Unit, Leeds Teaching Hospitals Trust, Leeds, United Kingdom; 3. Nutrition & Dietetic Group, Leeds Beckett University, Leeds, United Kingdom Introduction:P. aeruginosa (PA) is the most common pathogen and is associated with worse lung function, nutrition and increased mortality. Aztreonam lysine inhalation solution (AZLI, Cayston®, Gilead) is licenced for use in patients with CF aged six years and older, with chronic PA (CPA). The aim of this audit was to assess the usage and clinical outcome of AZLI in a large adult CF unit.

Methods: Data were extracted between 2011 and 2017. A total of 168 patients received at least one dose of AZLI. Indications included CPA, PA eradication and combination therapy in acute exacerbations. Patients who received at least 50% of doses (91/168 patients) were included with 15 subjects being excluded due to pregnancy, lung transplant and recent treatment. Data were extracted for age, weight, BMI, lung function, CRP, plasma viscosity (PV), white cell count (WCC), and days of IVs.

Results: A total of 76 patients met inclusion criteria; PA (n=62) and Burkholderia cepacia complex (BCC) ± PA (n=14). Of these 76 patients, 3 died within 1 year of prescription of AZLI, 4 died within 2 years, and 6 died within 3 years. Prior to the initiation of AZLI, the majority of patients had failed to respond to alternative inhaled antipseudomonal treatment; promixin monotherapy (PA n=17; BCC/PA n=2), tobramycin monotherapy (PA n=28; BCC/PA n=8), or a combination of the 2 treatments (PA n= 14; BCC/PA n=2). AZLI was prescribed on an alternate month basis with another agent in 49 PA (20 promixin, 29 tobramycin) and 8 BCC patients (3 promixin, 5 tobramycin) . There were no differences in clinical characteristics at baseline between PA and BCC patients except for weight where BCC patients were heavier. AZLI therapy was initiated in patients with low FEV 1 . In both groups there was a significant decline in FEV 1 prior to initiating therapy (p<0.001) which subsequently stabilised (NS). There was a significant increase in CRP and PV in patients with PA over the study period (Table) .

Conclusion: AZLI therapy stabilises lung function decline in patients with PA even with moderate and severe disease. Data suggest that AZLI was being prescribed late in the disease course and should be considered sooner. Further work to assess adherence to therapy is needed.

Carr, S.B. 1, 3 ; Leadbetter, J. 2 ; Caine, N. 4 ; Nyangoma, S. 3 ; Carryer, B. 2 ; Miller, J.E. 2 ; Richardson, J. 2 ; Bilton, D. 3 1. Paediatric Respiratory, Royal Brompton Hospital, London, United Kingdom; 2. Pharmaxis Ltd, Sydney, NSW, Australia; 3. Imperial College, London, United Kingdom; 4. CF Trust, London, United Kingdom Introduction: Inhaled mannitol (Bronchitol ® ) is an inhaled hyperosmotic medicinal product which in CF changes the viscoelastic properties of mucus, increases hydration of the periciliary fluid layer and contributes to increased mucus through mucociliary activity and cough provocation. Inhaled mannitol received a licence for use in adults from the European Medicines Agency (EMA) in 2012 with the recommendation to perform a post-authorisation study to assess long-term safety.

Methods: Using data from the UK CF Registry, a safety study was designed to monitor the use of mannitol in adults with CF. Reports were submitted to the EMA (6 monthly for first 3 years then yearly) over the 5-year period July 2012 -June 2017. All patients recorded as using mannitol in the UK were followed up and matched to similar unexposed patients using annual review data from the year prior to initiating mannitol. A propensity score model was used for matching using age, BMI, percent predicted FEV 1 , medication use, hemoptysis history and presence of chronic infections. The main outcomes were hemoptysis and bronchospasm; secondary outcomes included sepsis, abscess, cough fracture and changes in microbial infections. Outcomes were compared between exposed and matched unexposed groups using logistic regression models, adjusted for baseline factors. Reasons for stopping mannitol were also recorded.

Results: 446 adults without transplant were started on mannitol during 5-year study period with 131 subsequently discontinuing treatment (median exposure 15.7 months). The full unexposed population was 5484 subjects from which 947 were selected as matches using the propensity score model. The exposed and matched unexposed groups were similar in age, infection status, and rate of decline in FEV 1 (median 1.7% vs 1.8% per year). The exposed group had lower BMI (mean 21.75 vs 22.58; p<0.001) and baseline FEV1 (mean 59.2% vs 63.6%; p<0.001) and more days of IV antibiotic use in the year prior to exposure (mean 31 vs 24; p<0.001) and the history of hemoptysis was higher (8.84% vs 5.6%; p=0.02). Despite these baseline differences there was no significant difference in the main outcome measures of hemoptysis (11.7% vs 8.9%; p=0.397) and bronchospasm (defined by introduction of inhaled corticosteroids) 12.9% vs 13.0%, p=0.397). No events of cough fracture, pulmonary abscess or septicaemia were reported. Incidence rates (per 100 patient years) of new infections: PA 4.3 vs 4.6, SA 6.3 vs 6.8, Aspergillus 6.9 vs 4.8 (all not statistically significant), NTM 7.0 vs 7.0. The median annual rate of decline in FEV 1 slowed in both groups: 1.5% vs 1.6%.

Conclusion: The long-term safety data generated from this 5-year realworld study complements the extensive safety data from a broad clinical trial program with no new safety signals identified. This study confirms that there was no increase in hemoptysis or use of inhaled corticosteroids in the exposed population and no differences in other safety outcomes, including the rate of acquisition of new infections. This is despite the fact that the exposed population appeared sicker at baseline with higher IV antibiotic use and lower FEV 1 .

Messore, B. 1 ; Clivati, E. 1 ; Biglia, C. 1 ; Bonizzoni, G. 1 ; Demichelis, S. 2 ; Albera, C. 1 ; Ricciardolo, F. 3 ; Bena, C. 1 ; Bellocchia, M. 1 1. AOU S.Luigi Gonzaga, Orbassano, Italy; 2. Adult Cystic Fibrosis, AOU S.Luigi Gonzaga, Orbassano, Italy; 3. Department of Clinical and Biological Sciences, AOU S.Luigi Gonzaga, Orbassano, Italy Background: A.fumigatus is a ubiquitous, spore-forming fungus associated with multiple pulmonary disorders including allergic bronchopulmonary aspergillosis (ABPA), aspergilloma, invasive aspergillosis, allergic asthma and hypersensitivity pneumonitis.

In patients with cystic fibrosis (CF), growth of A. fumigatus hyphae within the bronchial lumen triggers an immunoglobulin E (IgE)-mediated hypersensitivity response that results in airway inflammation, bronchospasm, and bronchiectasis.

The management of ABPA in CF includes systemic and inhaled corticosteroids and itraconazole, considered the mainstay of therapy for ABPA. Adjunctive measures, such as the use of bronchodilators, and environmental manipulation, are frequently employed. More recently, small-scale studies have shown that anti-IgE therapy -omalizumab (OM) -may be an effective treatment for ABPA. The role of OM therapy in severe asthma is also well recognized.

Little is known about treating chronic allergic severe asthma in CF patients having previously experienced acute ABPA (ASA/ABPA).

Aim: to evaluate the effect of OM in CF patients with ASA/ABPA in terms of pulmonary exacerbation rate (PEr), airway patency, exercise tolerance and life quality.

Methods: We assessed the value of lung function, 6-minute walk test (6MWT), respiratory exacerbations, hospitalizations, days on oral and iv antibiotics, days on corticosteroids and quality of life six and twelve months before and after OM, in 7 CF patients with ASA/ABPA treated with OM (3 M, 4 F), mean age 31 years (range 24-41); mean FEV1 percent predicted (pp) pre-OM 57.4 (range 33-76), median OM therapy 1050 mg/month (range 600-1200 mg/month). OM dose was determined according to serum IgE levels. All patients gave consent to this analysis.

Results: We analyzed 7 CF patients (2 with ΔF508 homozygous mutation, 2 with ΔF508 heterozygous mutation), 6 with P. aeruginosa chronic respiratory infection, all sensitized to A. fumigatus, 6 with prick test positive to 2 or more perennial aeroallergens. We found a significant reduction in PEr (1.71 ± 0.28 vs 0.57± 0.20, p=0.002 at 6 months; 1.17 ± 0.65 vs 1.00 ± 0.45 p=0.004 at 1 year); and days on oral antibiotic therapy (24.36 ± 3.91 vs 7.43 ± 2.68, p=0.001 at 6 months; 49.83 ± 8.42 vs 14.00 ± 6.26, p=0.004 at 1 year). We observed an improvement in distance at 6 MWT pp (104.80 ± 5.36 vs 108.75 ± 4.9, p=0.012 at 1 year). There were no differences in hospitalization and IV antibiotics, corticosteroid use, lung function test and quality of life at both 6 months and 1 year.

Conclusions: In adult CF patients with ASA/ABPA treatment with omalizumab significantly decreases pulmonary exacerbation rate, oral antibiotics use and improves exercise tolerance, suggesting that anti-IgE therapy might have a role in improving clinical outcomes in this subset of CF patients.

Introduction: The survival of children with cystic fibrosis (CF) varies throughout European countries. One factor influencing a patient's outcome is the colonization of Pseudomonas aeruginosa (PSA). This study attempted to determine whether the varying survival rates occur due to differences in Eastern and Western European PSA treatment and its accessibility.

Methods: Using a survey focusing on PSA treatments, we gathered information from 36 European countries. The results were collected with the aid of doctors, CF centers and CF organizations. The survey contained 25 questions that were sectioned into 3 blocks: general information (eg guidelines, accessibility), PSA detection (eg diagnostic methods) and treatment regimen (eg start of therapy, medication). The survey was sent via email which allowed participants to answer either through an online link or by printing the PDF-version and sending it back via email or fax. The link to the online-survey also appeared in the European Cystic Fibrosis Society (ECFS) newsletter of September 2017. We received 45 online responses, 7 PDF and 5 faxes. In reference to Eastern Europe, this includes the following countries: former Soviet Union, Bosnia and Herzegovina, Bulgaria, Croatia, Czech Republic, Hungary, Poland, Republic of Macedonia, Romania, Serbia, Slovakia and Slovenia.

Results: The survey showed that in 33% of Eastern European countries, not every patient had access to this therapy whereas there was full accessibility in the West. In addition, only 2 of the Eastern European countries stated that the patients were required to pay for their own treatments. In all other cases, the costs were covered by a local or national health service. With the exception of 2 clinics in Western Europe, treatment was given according to guidelines. National guidelines existed in most of the countries, but local guidelines were more common in the West. In Eastern Europe, national guidelines and ECFS guidelines were both equally used. In the Western countries, the majority of clinics followed local guidelines (43.3%) or ECFS guidelines (33.3%), while national guidelines were used by only 13.3% of the clinics. In the diagnostic block, we found few differences between the East and the West. A noticeable disparity could be seen in the involvement of specialized laboratories in detecting PSA. While 96.7% of the Western clinics worked with specialized laboratories, only 74.1% of the clinics in Eastern Europe collaborated with them. The therapy block showed only minor differences. Although different medications were used, we noticed a preference for the following: ciprofloxacin, 40 mg/kg, every 12 hours, oral over 3-5 weeks and/or tobramycin, 300 mg, every 12 hours, inhaled over 4 weeks. The majority of Eastern and Western clinics used throat swab and sputum for the check-up of the eradication success. A few Western clinics also used antibody tests.

Conclusion: Even though the therapy was paid for in most countries, there were differences in its accessibility. Beyond that we found minor disparities. The Western as well as the Eastern countries used local, national or international guidelines. Diagnostic means may correlate to the country's financial resources. The use of therapeutics seemed similar in both Eastern and Western Europe.

Ratkiewicz, M.M. 1 ; Royse, C. 2 1. Pulmonology, Phoenix Children's Hospital, Phoenix, AZ, USA; 2. Univ. of Arizona, Phoenix, AZ, USA Introduction: Most patients with CF will meet diagnostic criteria and the diagnosis is straightforward, however in a subset diagnosis is not clear. Sweat chloride is the gold standard for confirming a diagnosis of CF (Farrell PM, et al. J Pediatr. 2017; 181S:S4-15) . Patients with intermediate sweat chlorides are a diagnostic dilemma. Asymptomatic infants with abnormal newborn screen (NBS) and older children and adults with concerning symptoms all require additional evaluation.

Methods: A retrospective chart review was performed on all patients with intermediate sweat chlorides at Phoenix Children's Hospital (PCH) between August 2012 -August 2014. Intermediate results were defined according to CF Foundation guidelines at that time (30-59 mmol/L for ≤6 months, 40-59 mmol/L for >6 months). Clinical and demographic data were gathered and patients classified according to diagnosis of CF, CFTR-related metabolic syndrome (CRMS), CF-excluded, or remained indeterminate.

Results: There were 250 intermediate results in 98 patients. Most had repeat testing with a mean of 2.4 and range of 1 to 6. Additional clinical information was available for 73. Ten were not seen by pulmonary and did not have repeat testing. Of the 73 evaluated at PCH, 21 (29.3%) were diagnosed with CF, 41 (56.2%) were excluded from a CF diagnosis, 5 (6.8%) with CRMS and 6 (8.2%) remain indeterminate. Of those diagnosed with CF, 14 trended to have diagnostic sweats, while 7 remained intermediate, and were diagnosed by genetic testing or clinical features. CFTR sequencing is shown in the Table for those diagnosed with CF. There were 21 who had testing due to an abnormal NBS. 8 were diagnosed with CF. Infants diagnosed with CF had a higher initial mean sweat chloride (43.6 mmol/L, SD 8.2) compared to those in whom a diagnosis of CF was excluded (35.6 mmol/L, SD 6, p=0.02). The mean age of diagnosis of CF was 10 months with a range of 1 to 38 months. Isolation of methicillin-sensitive Staphylococcus aureus occurred more frequently in those with CF (61.9% vs 30.4%, p =0.04). Isolation of only normal oropharyngeal flora was more common in those who were not diagnosed with CF (48% vs.4.8%, p=0.002).

Conclusions: Some patients are not being appropriately evaluated, highlighting an educational need. Our study also illustrates the complexity of a CF diagnosis in a subset of patients who do not meet standard diagnostic criteria. It demonstrates the limitations of sweat chloride testing and full sequence genetic testing, and the importance of a thorough evaluation by a pulmonologist at an accredited CF center. The mainstay therapy for acute bronchopulmonary aspergillosis (ABPA) in pediatric patients with cystic fibrosis (CF) is oral systemic corticosteroids. Adverse effects including hypertension, hyperglycemia, weight gain, osteoporosis, and cataracts are documented with the use of oral glucocorticoids. Case reports and small studies show high dose pulse intravenous (IV) methylprednisolone has been effective in treating ABPA, but limited data exist describing what adverse effects, if any, are associated with its use. The primary aim of this study is to determine the adverse effects associated with high dose pulse IV methylprednisolone for the treatment of ABPA in pediatric CF patients and the incidence at which they occur. Secondly, this study will evaluate the changes in pulmonary function tests and IgE serum concentrations after administration of high dose pulse IV methylprednisolone. Risk factors associated with the development of adverse effects will also be investigated.

Methods: A retrospective chart review was performed on pediatric patients with CF between August 2011 and August 2017. Inclusion criteria were as follows: age 18 years or less, diagnosis of ABPA, admission to acute care unit at Texas Children's Hospital, and pulse dose IV methylprednisolone administration during hospitalization. Lung transplant recipients were excluded from the study and patients concurrently treated with omalizumab were excluded from the IgE analysis. Descriptive analysis was performed on the data gathered.

Results: A total of 19 patients with 51 encounters met inclusion criteria for the primary objective. Patients were treated with an average methylprednisolone dose of 20 mg/kg and received an average of two consecutive doses per admission. Length of stay for each hospital encounter was 12.6±7.7 days. Majority of the patients (80%) were noted to have hyperglycemia during admission with the highest average glucose of 398 mg/dL, but returned to baseline in 97.5% of patients by the time of discharge. Only three encounters were notable for hypertension with an average high blood pressure of 137/74 mmHg. Few patients received opthalmascopic (n=5) and bone density exams (n=8) pre-and post-methylprednisolone therapy, therefore data analysis was not completed on these adverse effects. Methylprednisolone administration resulted in a reduction in IgE serum concentrations (866.2 vs 511.5 IU/mL, p=0.003) but no change in FEV 1 (60.5 vs 63.9 % predicted, p=0.44) . Patients with known CF-related diabetes were more likely to develop hyperglycemia after methylprednisolone administration than those without (96.2% vs. 68.2%, p=0.015).

Conclusion: Hyperglycemia was the most common adverse effect experienced by CF patients with ABPA receiving high dose pulse IV methylprednisolone and was more common in patients with a history of CF-related diabetes. Pulse dosing of methylprednisolone was also found to significantly decrease IgE serum concentrations.

Garinis, A.C. 1, 2 ; D. 2, 1 ; Hunter, L. 3 ; Feeney, P. 2,1 ; Steyger, P. 1,2 ; Powers, M.R. 1 ; Allada, G. 1 1. Oregon Health & Science Univ., Portland, OR, USA; 2. National Center for Rehabilitative Auditory Research, VA Portland Health Care System, Portland, OR, USA; 3. Cincinnati Children's Hospital, Cincinnati, OH, USA Introduction: Ototoxicity occurs when a chemical damages inner ear sensory cells or the auditory nerve. Patients with cystic fibrosis (CF) are at risk for acquiring ototoxic hearing loss due to the preferential use of intravenous (IV) aminoglycoside (AG) antibiotics for the treatment of acute pulmonary exacerbations involving Pseudomonas aeruginosa (a common colonizer of CF). An objective measure to detect early ototoxic changes and risk of further hearing damage would be highly valuable for CF patients. Transient evoked otoacoustic emissions (TEOAEs) are an objective measure of pre-neural cochlear outer hair cell (OHC) function and can be used to assess the strength of the medial olivocochlear (MOC) reflex, an indicator of cochlear neuronal integrity. Animal models indicate that both cochlear OHCs and MOC efferent neurons are an initial target of AG-related auditory damage. This study explores the clinical utility of both TEOAEs and the MOC reflex as early indicators of ototoxicity, that when combined with other patient factors (e.g., age, gender, and AG dosing) could be used to identify patients at risk for ototoxicity.

Methods: CF patients, ages 15 + years, were recruited from the Oregon Health & Science Univ. Hearing sensitivity was measured for frequencies between 0.25-16 kHz. Clinical 226-Hz tympanometry, a test of middle ear function, was conducted to rule out middle-ear pathology (e.g., otitis media) in each patient. TEOAEs were measured to determine OHC function in response to low and moderate level clicks. Patients with present TEOAEs were then tested for their MOC reflex by presenting contralateral broadband noise to one ear as TEOAEs were recorded in the opposite ear.

Results: CF patients with normal hearing (≤25 dB hearing level) from 0.25-8.0 kHz were categorized into two groups: those receiving active IV-AG treatments and those not actively on treatment. TEOAE and MOC reflex data from this study were compared to published age-matched normative data in non-CF ears with normal hearing. Preliminary data shows that patients receiving IV-AG treatments had TEOAE responses that were lower in amplitude compared to normative data and to CF patients not receiving treatment. TEOAE response growth as the stimulus level was increased from low to moderate levels showed greater variability in the treatment group. MOC reflex inhibition was also greatly reduced in CF patients on IV-AG treatment compared to both control group comparisons.

Conclusions: IV-AG dosing may have a significantly negative effect on auditory function at the cochlear and brainstem levels in CF patients with normal hearing. Further exploration of these objective hearing tests may show promise for identifying ototoxicity before functional hearing loss occurs. Preserving hearing is critical to improve quality of life, maintain social connections and academic or career success in CF patients.

Acknowledgments: Funded by NIH-NIDCD Award, 1R21DC016128-01A1. The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH or the VA.

Hunter, L. 3, 4 ; Garinis, A.C. 1, 2 ; Dong, M. 4, 3 ; Macdonald, K. 4 ; Burg, G. 3 ; Blankenship, C. 3 ; Feeney, P. 2,1 ; Clancy, J.P. 3, 4 ; Vinks, A. 3, 4 1. Oregon Health & Science University, Portland, OR, USA; 2. National Center for Rehabilitative Auditory Research, VA Portland Health Care System, Portland, OR, USA; 3. Cincinnati Children's Hospital, Cincinnati, OH, USA; 4. College of Medicine, University of Cincinnati, Cincinnati, OH, USA Introduction: Ototoxicity (sensory hearing loss, SNHL) is a well-established side effect of life-saving intravenous aminoglycoside (AG) treatments in cystic fibrosis (CF) patients. Previous studies have demonstrated wide variation in the incidence of AG-induced SNHL, likely due to genetic variants, differences in pharmacokinetics, drug regimen, and kidney function. The goal of this study was to examine the effect of cumulative IV-AG exposure on hearing function in pediatric patients with CF, and to estimate exposure using pharmacokinetic (PK) modeling of tobramycin and amikacin.

Methods: Participants with CF (n=50) received IV tobramycin or amikacin in combination with vancomycin. Age-matched participants without CF (n=50) or a history of AG treatment were also assessed. Hearing thresholds for all participants were tested using behavioral audiometry (0.25-16 kHz). Hearing status was classified as normal hearing (< 20 dB) or SNHL (one threshold ≥ 25 dB, or two or more thresholds ≥ 20 dB). Total number of lifetime IV-AG doses and demographic information was collected from the electronic medical record. Participants were placed in four quartiles based on their cumulative IV-AG exposure. Amikacin and tobramycin individual cumulative exposures (expressed as area under the curve, AUC) were estimated using a PK model-based Bayesian approach with MWPharm. The PK modeling allowed estimation of pharmacokinetic parameters and systemic AGs exposure (AUC) on the basis of medication history, taking into account the varying status of the patient with respect to body weight and kidney function over time.

Results: CF participants demonstrated mild ranging to severe SNHL across all exposure quartiles of IV-AGs, with poorer hearing thresholds compared to age-matched controls for both ears. There was a significant association between the number of cumulative IV-AG doses and the prevalence of SNHL. Patients exposed to amikacin in addition to tobramycin had more severe SNHL. PK modeling demonstrated that patients with poorer renal clearance had poorer hearing outcomes for tobramycin. Amikacin was associated with poorer hearing outcomes, but the sample size was small, and all exposed patients had SNHL, primarily in the 9-16 kHz range.

Conclusions: The total number of lifetime IV-AG doses for tobramycin was associated with a higher prevalence of SNHL, as significant hearing loss occurred in all quartiles of exposure. All patients treated with amikacin had significant SNHL. There was a high prevalence of ototoxicity in this pediatric cohort, consistent with similar studies in adult CF patients. Prospective, longitudinal investigations of exposure via PK modeling are needed to better understand the relation between AGs and ototoxicity in CF patients. Acknowledgment: Funded by the NIH-NCATS Award 5UL1TR001425-04 and CCHMC Place Outcomes Research Award. The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH or the VA. Wales, Cardiff, United Kingdom; 2. School of Medicine, Cardiff University, Cardiff, United Kingdom; 3. Biosciences, Cardiff University, Cardiff, United Kingdom; 4. Microbiology, University Hospital for Wales, Cardiff, United Kingdom Introduction: Children with CF are often nonproductive of sputum even during exacerbation. Effective pathogen detection is challenging, but critical, since infection and lung disease begin early in life. The role of sputum-induction as an infection-diagnostic has not previously been systematically addressed in young children with CF.

Methods: Children with CF aged between 6 months and 18 years were recruited at the Children's Hospital for Wales (Cardiff, UK). We matched samples from cough swab, sputum-induction, and single-lobe, two-lobe, and six-lobe bronchoalveolar lavage (BAL) for within-patient comparisons of pathogen yield, in order to evaluate the relative contribution of each approach. Primary outcomes were comparative pathogen yield from sputum induction and cough swab in stage 1 of the study, and from sputum-induction, and single-lobe, two-lobe, and six-lobe BAL in stage 2.

Results: 124 patients were recruited to the trial and underwent 200 paired cough swab and sputum-induction procedures in stage 1. 167 (84%) procedures were successful. Of the 167 paired samples, 63 (38%) sputum-induction samples were pathogen-positive compared with 24 (14%) cough swabs (p<0.0001). 86 pathogens were isolated in total. 79 (92%) pathogens were isolated from sputum-induction compared with 27 (31%) from cough swab (p<0.0001). In subgroup analysis, similar benefits were seen from sputum-induction in children <6 or ≥6 years of age, and in both asymptomatic and symptomatic patients. For stage 2, 35 patients underwent a total of 41 paired sputum-induction and BAL procedures. 39 pathogens were isolated from 28 (68%) of the 41 paired samples. Sputum-induction identified 27 (69%) of the 39 pathogens, compared with 22 (56%; p=0.092) on single-lobe, 28 (72%; p=1.0) on two-lobe, and 33 (85%; p=0.21) on six-lobe BAL. In a sensitivity analysis, using sputum-induction and 6-lobe BAL as a combined gold standard, sensitivity was 0.63, 0.59, and 0.85, for sputum-induction, 2-lobe BAL, and 6 lobe BAL respectively.

Conclusion: Sputum-induction is superior to cough swab for pathogen detection, and is a credible approach to sampling the lower airway in symptomatic children with CF. Performing sputum-induction before BAL will correctly describe the lower airway pathogen environment in almost two-thirds of patients, and could substantially reduce the number of bronchoscopy procedures required. Both sputum induction and six-lobe BAL contribute independent, incremental gains in pathogen detection compared with the current gold-standard two-lobe BAL. We propose that sputum-induction and six-lobe BAL combined are used as standard of care for comprehensive lower airway pathogen detection in children with cystic fibrosis. Rationale: Patients with CF experience inflammation due to bacterial colonization and neutrophil recruitment that increase during an acute pulmonary exacerbation (APE). Because 20% of CF patients admitted for APE may not return to their previous spirometry baseline, CF patients and their providers are eager to restore baseline FEV 1 percent predicted (FEV 1 pp) with APE treatment. There is insufficient evidence to recommend systemic steroids during APE treatment; however, our center has utilized a 5-7 day "rescue" steroid burst started at least a week into admission when treatment fails to demonstrate expected improvement. We hypothesized that rescue steroids do not result in significantly improved FEV 1 pp compared to matched controls.

Objective: After obtaining IRB approval, encounters of CF patients hospitalized from 6-1-2013 through 8-31-2017 were assessed. Inclusion criteria: 6-20 years of age with a diagnosis of CF that can perform spirometry, and admitted for at least 12 days duration. Exclusion criteria: steroids ordered for ENT procedure, diagnosis of CF-related diabetes, FEV 1 pp < 40% at baseline, or have a history of ABPA. Demographic data including age, gender, length of stay, antimicrobial therapy agents, time until next antimicrobial treatment and best baseline spirometry were collected. Corticosteroid treated encounters were age, gender and date matched with patient encounters who did not receive steroids. Data were assessed for FEV 1 pp at admission, midpoint, discharge, and at follow-up visit. Data were analyzed by t-test and Fisher's exact test.

Results: 182 of 817 CF patients during the study period received systemic steroids for various indications. Applying our criteria, there were 65 rescue steroid and 48 matched control encounters. Our interim analysis is in the Table. Conclusion: CF patients treated with rescue steroids had a lower improvement in FEV 1 pp at the midpoint of admission compared to control but a similar rate of return to baseline FEV 1 pp at discharge and followup. Our data suggest time until next antibiotic was longer in the treatment group, however, length of stay was longer and change in antimicrobial therapy was more frequent, both of which could contribute to this outcome. These data suggest a prospective trial should be considered to understand rescue corticosteroids in CF APE management.

Johnson, R.; Rice, J.; Wagner, B.; Juarez-Colunga, E.

Cystic fibrosis (CF) clinical trials often utilize time to first pulmonary exacerbation (PEx) or total number of PEx as endpoints. These outcomes fail to capture patterns or timing of exacerbations and can incompletely capture how covariates influence the risk of further exacerbations. Analysis of gap times between exacerbations provides a better framework to understand risks of subsequent PEx in children with cystic fibrosis. Gap time models can be adapted to account for the progressive nature of chronic diseases or infections in biomedical studies by including previous number of exacerbations in the model. We propose a statistical model to explain elevated risks after exacerbations and when and how these risks return to some baseline level. We demonstrate these methods on data from individuals enrolled in the Early Pseudomonas Infection Control (EPIC) study and provide a summary of multiple adaptations of these models. We found that a change-point provided a better fit to the data and that risk of a subsequent exacerbation dropped after approximately 1.8 years. Furthermore, we determined that increased number of copies of the ΔF508 mutation and female gender were associated with higher rates for time to first exacerbation and greater risks of having a future exacerbation, and higher numbers of previous exacerbations were also significantly associated with greater risks of having an exacerbation. Introduction: Oral antibiotics are frequently prescribed for pulmonary exacerbations in children with cystic fibrosis (CF). Antibiotics may be prescribed during clinic visits or via telephone triage following family-initiated contact for increased respiratory symptoms. However, antibiotic prescribing patterns and patient outcomes are not well understood. The objective of this study was to characterize the frequency of antibiotic prescriptions in clinic and by telephone triage in our pediatric CF center. We also sought to characterize symptoms leading to oral antibiotic prescriptions and to determine the number of hospitalizations within the 3 months following an oral antibiotic course.

Methods: A retrospective chart review was performed in all patients followed at our CF center 6-17 years of age who received a course of oral antibiotics for a pulmonary exacerbation between July and December 2016. The percent of our center's overall population in this age range who were prescribed oral antibiotics over the 6-month period was calculated. The location of antibiotic prescription was determined, symptoms were captured and frequency of hospitalization within the 3 months following an oral antibiotic course was compared between the two groups.

Results: Of 236 eligible CF patients, 139 (59%) received a total of 215 antibiotic courses over the 6-month period (range: 1-5 antibiotic courses per patient). Antibiotics were more frequently prescribed by telephone triage compared to in-clinic (122 vs. 93 [56.7% vs. 43 .3%], p=0.05). Amoxicillin-clavulanic acid was the most commonly prescribed antibiotic (103 [47.9%]) followed by trimethoprim-sulfamethoxazole (44 [20.5%]) and fluoroquinolones (37 [17.2%]). Antibiotics were prescribed for a median of 14 days (range 7-28 days). Cough was the most common symptom leading to antibiotic prescriptions and was present in 201/215 encounters (93.5%). Of those encounters without a reported increase in cough, 12/14 were seen in clinic and decreased lung function led to the antibiotic prescription. Twenty-nine patients were hospitalized a total of 36 times (range: 1-3 hospitalizations per patient) within three months of being prescribed an oral antibiotic for a pulmonary exacerbation. There was no difference in hospitalization rate between patients prescribed antibiotics in clinic versus over the phone (18/93 [19.4%] vs. 19/122 [15.6%], p=0.5).

Conclusions: Oral antibiotics are frequently prescribed in pediatric patients as treatment for a pulmonary exacerbation, and initial management is often provided via telephone triage. Cough is the most commonly reported symptom. Importantly, those without a noted increase in cough or other symptoms were often prescribed antibiotics due to decreased lung function which may be missed in those evaluated only by telephone. Hospitalizations occurred in a minority of the patients in the three months after oral antibiotic treatment. Acknowledgment: Supported by the Cystic Fibrosis Foundation (#HOPPE16A0). London, United Kingdom; 3. Cystic Fibrosis Foundation, Bethesda, MD, USA; 4. Royal Bromptom Hospital, London, United Kingdom; 5. Surgery, Johns Hopkins School of Medicine, Baltimore, MD, USA Rationale: Therapeutic advances in cystic fibrosis (CF) have increased median predicted survival to 47 years in the United States (US) and United Kingdom (UK), yet many individuals with CF develop progressive disease. Lung transplantation (LT) remains an effective treatment for end-stage lung disease, but survival is limited after LT to over 7 years (US) and over 8 years (UK). Transition programs from pediatric to adult care for CF have improved outcomes, yet little is known about the transition from CF to LT care.

Methods: This is a parallel cross-sectional survey of CF providers in accredited CF centers in the US and specialist CF centers in the UK. The survey explores CF center experience with LT, transition practices for patients needing LT, and communication between CF and LT teams.

Results: 92/122 (75.4%) US CF centers and 26/28 (93%) UK specialist adult centers and several pediatric centers answered the survey. 206 surveys (US) and 114 surveys (UK) were completed. Half of respondents in the US and UK have over 10 years of experience in CF care. 36.4% in the US and 39.7% in the UK have a LT program associated with their center. In both countries, only 69.4% (US) and 67.5% (UK) of CF teams meet formally to discuss patients who may require LT, and in addition, 40.4% (US) and 46.8% (UK) hold these meetings on an "as needed" basis. Most CF teams provide education and assess patient understanding of LT, but only 46.1% (US) and 54.4% (UK) have a visit focused on LT, and only 20.9% (US) and 12.3% (UK) have a member of the LT team at that visit. Most centers in the US (77.3%) and in the UK (79.8%) reported a dedicated LT team meeting, but only 20.3% (US) and 21.1% (UK) have a CF caregiver present at this meeting. Respondents in both the US and UK reported concern that patients were discussed too late (21.8% vs. 21.9%), referred too late (19.3% vs. 18.4%), became too sick to refer (29.7% vs. 14.0%), and were impacted negatively due to timing (36.6% vs. 28.7%).

Conclusion: There is no standardized process for coordinating care from CF to LT in both the US and UK; responses from CF caregivers demonstrate a wide variation in practice. Further work is needed to develop best practice guidance in this highly complex transition process to help improve outcomes for individuals with CF who undergo LT. The CF Foundation Respiratory Education Committee recommends CF patients take their nebulized respiratory treatments (tx) in a certain order. The order of respiratory tx is believed to optimize the efficacy of each medication and promote maximum effort of airway clearance from the patient. The current pragmatic recommended order of the respiratory tx is: bronchodilator, dornase alfa, 7% hypertonic saline, inhaled antibiotics, and ending with inhaled steroids. This recommendation is a suggested order, as there are no data to support.

After meeting with respiratory therapists from other CF centers, we discovered that some centers differ in how the respiratory regimen ends.

While some centers recommend their patients to end their respiratory tx order with inhaled steroids, others recommend ending with inhaled antibiotics. Does changing medication order affect the patients' FEV1? Are there other factors that may be involved?

Method: -The CF care center identified patients using both inhaled steroids and antibiotics.

-A baseline FEV1 was recorded for these patients for their current respiratory order of tx.

-The patients that participated were then asked to switch the order of inhaled steroids and inhaled antibiotic in their daily respiratory tx routine.

-A subsequent switch FEV1 was recorded at their next quarterly visit.

-Patients involved in the test were invited to provide feedback on how the medication order changed their respiratory tx regimen.

Results: In total, 26 patients were identified as using both inhaled steroids and inhaled antibiotics and were selected for the A/B variation study. 73% of selected patients completed the the A/B variation study and returned for the post-study FEV1 testing. Differences in FEV1 scores before and after switching the order of tx were observed in 74% of study patients. In 42% of study patients a higher FEV1 score was correlated with finishing tx with inhaled antibiotics compared to 32% finishing tx with inhaled steroids. The small sample size prevents the result from being statistically significant (p value is 0.2493). It would take a sample size of >125 patients to show statistical significance for the performance differences observed.

In post-study feedback 32% of patients reported that finishing with inhaled antibiotics (which are prescribed cyclically) in their tx order resulted in better adherence of their respiratory tx regimen for taking inhaled steroids for times when antibiotics were not prescribed.

Conclusion: At the conclusion of this study there are several essential takeaways:

-At this point, there is no statistically significant difference between finishing one's tx regimen with inhaled steroids or inhaled antibiotics.

-Following the suggested feedback from patients in the study, our CF care center will recommend a pragmatic order of inhaled steroids before inhaled antibiotics to promote better adherence for inhaled steroids. This is done mainly to promote the consistency of the adherence and messaging to patients rather than the effectiveness of the prescribed medications.

-To confirm the results observed, a larger sample size will be required. Expanding the patient sample population through the inclusion of test results from other care clinics may be necessary to reach the required levels for statistical significance.

Leemans, G. 1 ; De Hondt, A. 1 ; Ides, K. 2 ; Van Holsbeke, C. 1 ; Belmans, D. 1 ; Becker, B.C. 3 ; Van Hoorenbeeck, K. 4 1. Fluidda, Inc., Kontich, Belgium; 2. Universiteit Antwerpen, Antwerpen, Belgium; 3. Clinical Research, St. Paul, MN, USA; 4. Pediatrics, Universitair Ziekenhuis, Antwerpen, Belgium Introduction: High-frequency chest wall oscillation (HFCWO) therapy for airway clearance (AWC) is common in patients with cystic fibrosis (CF). A limitation is requirement for AC power. A mobile HFCWO device (mHFCWO) that generates chest wall pressure and oscillating airflow comparable to standard HFCWO (stdHFCWO) devices is now available. Our study evaluated effectiveness of the mHFCWO (The Monarch ® System) compared to stdHFCWO (The Vest ® System).

Methods: The study was a randomized open-label crossover. Subjects received HFCWO with both devices. Inclusion criteria were age ≥ 15 years, stable CF medications (prior 4 weeks) and daily sputum production. Patients were excluded for FEV 1 percent predicted < 30% or > 90%, history of pneumothorax (past 6 months), hemoptysis requiring embolization (past 12 months), IV antibiotics (past 4 weeks), exacerbation of allergic bronchopulmonary aspergillosis or anticipated hospitalization in next 3 weeks. Patients were randomized to mHFCWO or stdHFCWO day 1. Each subject received alternate therapy after a washout period (2 -7 days) . Subjects performed 1 morning treatment each day. 4 subjects performed an afternoon treatment. Treatments were 30 minutes, multiple frequencies, with intensity settings 6 to 10 on each device. Subjects collected sputum during 30-minute therapy and 1-hour posttherapy. We compared mean wet weight of sputum. To further evaluate mHFCWO therapy, CT scans were done before therapy and after sputum collection. Functional respiratory imaging (FRI), described by De Backer and coworkers (Radiology. 2010; 257:854-62 ) was used to evaluate airway geometry using CT scans. Brody scores were performed on CT scans by a radiologist blinded to the study.

Results: 9 stable patients with CF were enrolled. 1 subject exited for inability to complete study visits. 8 subjects completed the study. Mean wet sputum weight was similar between mHFCWO and stdHFCWO therapy (6536 ± 8554 vs. 5801 ± 5824 mg, p=NS). FRI showed statistically significant differences in airway geometry and patency for mHFCWO after therapy (iVaw: 49.442 ± 50.117 vs 44.516 ± 49.637, p<0.001), suggesting mucus shifting. Comparison of pre-and post-therapy Brody scores showed statistically significant improvement after therapy with mHFCWO (57.71 ± 16.55 vs 55. 2 ± 16.98, p=0.001) .

Conclusions: In our study, mHFCWO was comparable to stdHFCWO for sputum production. Post-therapy improvement in AWC is shown by FRI results and improved Brody scores. 1, 2 1. Medicine, Nursing and Health Sciences, Monash Univ., Melbourne, VIC, Australia; 2. Respiratory Medicine, The Alfred Hospital, Melbourne, VIC, Australia; 3. Physiotherapy, Australian Catholic University, Ballarat, VIC, Australia Background: Regular maximal exercise testing is recommended as best practice in adults with cystic fibrosis (CF). While CPET is the gold standard measure of exercise capacity it is not always available to all patients with CF around the world. Annual exercise capacity measurement is recommended as best practice to monitor physical fitness and to prescribe an optimal exercise program. The previously used 3-minute step test has floor and ceiling effects in adults with CF.

Objectives: To develop a simple, maximal, incremental, externally paced exercise capacity test suitable for adults with CF across the entire range of lung function, age, height and BMI; to evaluate the feasibility and safety of the test.

Step Test Exercise Protocol (A-STEP) was developed using an interval timer App, a 20 cm step, an oximeter (SpO 2 , heart rate (HR)) and a recording sheet. To test feasibility and safety during stable state all of these measures together with shortness of breath and leg fatigue using a 0-10 numerical rating scale were recorded at baseline (pre-test) and at the end of each level of the test until the subject reached the maximal level by achieving scores of 8, 9 or 10 out of 10 for SOB or LEGS or both. The test started at 18 steps per minute and increased each minute to a maximum of 15 levels (46 steps/minute). Blood pressure (BP) was recorded before and at the end of the exercise test. Patients with an FEV 1 below 20% were excluded together with those with co-morbidities that precluded maximal exercise testing such as musculoskeletal, cardiac and other organ problems and pregnancy. The Alfred Wellness Score (Awescore) was completed.

Results: Baseline demographic data was reported as mean (standard deviation (SD)) range for 40 adults (20 male) with stable CF aged 31.7 FEV 1 57.2 (20.4) 27-98% predicted (9 with mild lung disease (3 male (M)), 23 with moderate (14 M), and 8 with severe (3M)); Height 167.0 (8.9) 146-185.5 cm; BMI 22.1 (2.7) 17.7-30.0; AweScore 73.6 (11.1) 52-96%. Pre/Peak A-STEP (for SpO 2 , HR, SOB/ LEGS post for BP) (mean (SD) range) were: SpO 2 : 97.9 (1.79); 92.1 (5.0) 78-99%; HR: 79.6 (13.35); 164. systolic BP: 115.7 (14.7); LEGS: 0.3/8.4 . Twenty-three (58%) reported maximal ratings of 9 or 10/10 for SOB, LEGS or both with all but one reaching at least 8/10. One extremely fit patient could not reach peak SOB or LEGS because of in-coordination at the highest level. The mean number of levels completed for those with mild lung disease was 11.67, moderate 10.17, and severe 7.38. Recovery: 36 participants were at baseline SpO 2 in 5 mins or less. One participant achieved baseline HR in this time. There were no adverse events.

Conclusion: We developed a new maximal exercise field test for adults with stable CF. The A-STEP was feasible and safe across the entire range of lung function, age, height and BMI. Future research aims to validate the A-STEP against the gold standard cardiopulmonary exercise test. Intrapulmonary percussive ventilation has been availabe in Europe and North America for decades and only recently introduced to Australia in the form of the Metaneb. There is little evidence for use of the Metaneb in cystic fibrosis (CF). As part of the clinical introduction of the Metaneb, patient feedback was sought to provide some evidence for its use.

Aim: To determine whether use of the Metaneb in a large adult CF service is feasible and safe; to evaluate patient reported outcomes relating to airway clearance therapy (ACT) using the Metaneb.

Methods: Following review of radiology and discussion with each patient's treating physician, patients not achieving expected improvement in lung function and sputum clearance with their usual ACT techniques were selectively treated with the Metaneb. This provided nebulised mucolytic therapy with hypertonic (HS) or isotonic saline (IS) combined with constant positive airway pressure (CPEP) and constant high frequency oscillations (CHFO). The rationale for using the Metaneb was to mobilise secretions through lung recruitment, collateral ventilation and creation of shear forces. Patients were treated with repeated 2 to 3 minute cycles of each modality with rests for forced expirations and expectoration as required. Dosage (number of cycles and number of treatments) was individualised. Patient reported outcomes are presented as values on a number of visual analogue scales from -5 (most negative outcome) to 0 (no different) to +5 (most positive outcome) comparing the Metaneb with their usual ACT.

Results: Thirty two patients (12 male Conclusions: Patients found the Metaneb to be more effective in terms of chest clearance compared to their usual ACT. The time required for ACT was similar to usual ACT. They all reported a preference for the Metaneb to clear persistent mucus plugging. The Metaneb was feasibile for use with CF inpatients and outpatients and there were no adverse events. . We used data from the Patient and Family Experience of Care survey to implement a process improvement in response to patient-reported experiences. Because our timeliness of care scores are consistently in the lower quartile, we designed an intervention to reduce wait times in clinic.

Objective: To improve timely care and productivity of clinic visits by reducing clinic wait times.

Methods: The CFF sends Patient and Family Experience of Care (PFEC) surveys to families after clinic visits and our team reviews monthly CFLN reports. Using PFEC data, we implemented a QI intervention called "Take Five" to improve clinic wait times, using Plan-Do-Study-Act (PDSA) cycles, SMART Aims (Specific, Measurable, Attainable, Relevant, Time Bound), and 90-day goals. Take Five is a verbal cue system. Staff exiting a patient's room address the next provider by saying "take 5" and the oncoming person then has five minutes to get report, put on personal protective equipment, and enter the patient room. We measured visits for 67 patients before the initiative and 70 after. Wait episodes were defined as any period following check-in at the front desk that families were not face-to-face with staff. Episodes were categorized as short (0-5 minutes), moderate (6-9), long (10-14), and excessive (≥15). To measure the process, we time-stamped multiple clinic dates. All data were analyzed in Excel and we used graphs and run charts to track results.

Results: Our response rate of PFEC survey returns is 22%, which is comparable with other centers in the CFLN. One measure of timely care is how promptly patients are brought to their exam room, and our score for this category is consistently low at the 26th percentile. Because patients at our center see between three and eight different disciplines at each visit, breaks between providers can collectively add up to long overall waits and increase time spent in clinic. We found that families averaged six waiting episodes per visit, and prior to Take Five spent an average of 36 total minutes waiting. After the intervention this decreased to 22 minutes. Wait time as a percent of total visit time dropped from 35% to 23%, and the average wait per episode decreased 45%, from 6.5 to 3.6 minutes. The amount of moderate, long, and excessive wait episodes went from 37% to 18%. Average total time in clinic decreased by 10 minutes.

Conclusions: The Patient and Family Experience of Care survey is a helpful tool to identify areas of concern which can be used to drive QI efforts. Feedback from families indicated dissatisfaction with timeliness of care and we successfully designed a targeted intervention around reducing waiting times between providers in clinic. Our goal is to use this process to turn over rooms quickly and decrease time patients spend between check-in and the start of their visit. We have also used these results to implement a communication protocol for excessive waits. 501w NEBULIZER CLEANING: BACK TO BASICS Presta, C. 1 ; Johnson, Z. 2 ; Watts, K. 3 ; Conway, J. 4 1. Pediatric Pulmonology, Advocate Children's Hospital Cystic Fibrosis Center, Park Ridge, IL, USA; 2. Resident, Advocate Children's Hospital, Park Ridge, IL, USA; 3. Director, Advocate Children's Hospital Cystic Fibrosis Center, Park Ridge, IL, USA; 4. Quality Improvement, Advocate Children's Hospital Cystic Fibrosis Center, Park Ridge, IL, USA Introduction: During routine respiratory assessments, discrepancies between the Cystic Fibrosis Foundation (CFF)'s standard of care guidelines and actual practice regarding nebulizer cleaning were uncovered. The aim of this project is to improve infection control of nebulizer equipment through education to patients and parents.

Methods: A pre-and post-assessment questionnaire was created to evaluate the level of knowledge related to nebulizer cleaning practice. The CFF nebulizer cleaning video was installed on all computers in patient exam rooms. Using PortCF, a list was created of all patients. At the initial visit, families complete the pre-assessment, watch the video, and receive hands-on education. The questionnaire and cleaning methods are reviewed and all questions are answered during the assessment. A follow-up visit is completed at 3 months, with the questionnaire repeated to assess knowledge improvement and reviewed. Plan-Do-Study-Act (PDSA) methodology was used to test redesign ideas.

Results: 2 PDSA cycles identified definitive knowledge deficit, need for ongoing education, and necessity for nebulizer supplies. Of the 34 patients enrolled (2 patients were excluded) in the CF center a pre-assessment has been performed on 29 patients (90.7%). A lack of knowledge was identified with a mean score of 78.2%, with the target of 90% or higher. Moreover, those falling below the mean, there was not distinct majority in a specific age group. Currently, 17 patients have completed the post-assessment questionnaire with a mean score of 94.4%, which is above the target of 90% or higher. Despite a small "n" size, improvement in educational knowledge is apparent. Also, at the follow-up visits, positive feedback was noted from families about the education provided.

Discussion: Utilization of online educational resources and hands-on education during clinic visits improved knowledge base and infection control practices for our families. A new policy is being developed to include yearly evaluations, continued re-education, and quarterly supply distribution.

Pre-Score Prior to Nebulizer Education We initiated this quality improvement (QI) project to improve the FEV1 in our patients with the lowest lung function (FEV1<80% predicted).

The project started in November, 2017 by creating a list of patients with an FEV1 <80% predicted.

Global Aim: To improve FEV1 in our CF patients 6-21 years old. Specific Aim: To increase the mean FEV1 percent predicted by 5% in patients with FEV1 below 80% predicted at the start of the project within 6 months.

Methods: Plan: Identify patients with FEV1 <80% predicted. Do: Collect data on patients with FEV1 <80% and develop individualized treatment plan for each patient. Develop fishbone to identify potential contributing factors to suboptimal lung function. Create flow diagrams to outline clinic processes to improve identification of patients and to help address barriers to improved lung function.

Study: Study barriers to adherence, education and clinic flow that may result in FEV1 <80% and assess these barriers by both standardizing care across providers and disciplines and individualizing care for each patient.

Act: Implement the individualized treatment plan.

Results: Forty-one pediatric CF patients were identified (44% male). Mean age was 16.5 years (range 7.9-21.6 years), 51% of patients were homozygous F508del. Mean FEV1 percent predicted at baseline was 59.2 (range 29-79%). Key findings from the flowchart of our center's process:

Patients: Need for better patient education on treatment technique, cleaning of equipment and appropriate settings. Providers: A more standardized approach between all providers on patient management and on patient education was needed. To address these points, each discipline created a flow chart to be followed in clinics to standardize our approach and education to the patients and families. Key findings from the center's fishbone: Need for continuing education for patients/families, need to give out CF treatment plan with all medications listed, all changes that were discussed during the clinic visit and the plan for ongoing treatment after every clinic visit. Our results showed improvement in mean FEV1 percent predicted by 6.7% (SD ± 18) in 6.5 months' duration which exceeded our target goal of 5%.

Conclusion: Through this project which is ongoing, it was clear that all team members, patients and families should partner to influence FEV1 percent predicted improvement. The flow charts that were created have been instrumental in organizing and focusing our efforts to reach our goal. The team members and patients/families feel ownership of this project. Since the flow charts helped our team members and patients' families in the lower FEV1 group, they are being applied to all patients regardless of their FEV1 percent predicted to help improve the overall FEV1 for all our patients. (Rajan S, et al. Semin Respir Infect. 2002; 17:47-56) . At Ann & Robert H. Lurie Children's Hospital of Chicago (Lurie Children's), the 2014 CF Foundation's Patient Registry data revealed the median FEV1 percent predicted (ppFEV 1 ) for patients 6 to 17 years of age was below the national average since 2011. When comparing overall ppFEV 1 over the past 10 years, our ppFEV 1 rose little (from 85.8% to 89.9%), while the country rose from 85.8% to 92.5%. As a result, we initiated a quality improvement project called REACT: the Re-Education of Airway Clearance Techniques. The REACT program was originally designed and implemented by Dr. Zanni in 2006 and he reported an increase in median ppFEV 1 of 9% in 6-17 years old after implementing REACT (Zanni RL, et al. BMJ Qual Saf. 2014; 23:i50-5) .

Aim: We aim to improve ppFEV 1 of patients over 6 years of age at Lurie Children's Pediatric Cystic Fibrosis Program.

Methods: Our center designed a REACT program similar to Dr. Zanni's with his permission. An introduction letter was sent, explaining REACT and inviting patients to voluntarily participate. A "call to action" letter was sent prior to an annual respiratory review, requesting patients to bring their respiratory care equipment to participate in the REACT program. Participants were seen by the respiratory therapist, who reviewed knowledge about airway clearance techniques (ACTs), assessed and corrected ACTs, assessed function and use of patient equipment, reviewed care of equipment, discussed treatment plans and options, instructed new techniques when applicable, reviewed the importance of adherence to treatment and assessed any barriers to treatment. Patients who did not have correct technique and/or were not adherent to treatment plan were asked to return to clinic sooner for further education and follow-up. The primary outcome measure is ppFEV 1 , secondary outcome measures are BMI percentile and FEF 25-75 percent predicted (ppFEF 25-75 ). Outcome measures for ppFEV 1 , and BMI percentile were tracked for individuals in the year prior and year after initial REACT session.

Results: From July 2016 to May 2018, 81 (57%) out of 143 eligible patients participated in REACT during their annual respiratory review. Eleven patients were excluded for insufficient data. Patients were equally distributed by gender and ranged from 6 to 22 years, with 46% 6-12 years, 39% 13-17 years and 16% over age 18. Mean ppFEV 1 and ppFEF 25-75 remained constant pre-and post-REACT, 95% and 82% respectively. A small increase in BMI occurred, going from 61% pre-REACT to 64% post-REACT.

Conclusions:The REACT program at Lurie Children's has not yet resulted in a robust improvement in ppFEV 1 . We will continue the REACT program due to success at identifying and correcting gaps in knowledge and skills in ACTs, and addressing adherence issues. We plan to improve our implementation and increase participation with the goal of improving the lung function at our CF center.

Phillips, A.L. 1 ; Aikens, C. 1 ; Hoover, W. 2, 3 ; Thomas, L.K. 2 1. Physical and Occupational Therapy, Children's of Alabama, Birmingham, AL, USA; 2. Pulmonary, Children's of Alabama, Birmingham, AL, USA; 3. Pediatric Pulmonary Medicine, Univ. of Alabama at Birmingham, Birmingham, AL, USA Introduction: Newborn screening advances have allowed for significantly earlier CF diagnosis. This is complemented by focused interdisciplinary care in outpatient clinics. Our team has clinically observed several patients with evidence of delayed development (DD), without standardized developmental screens performed in clinic. With the addition of physical therapy (PT) to the outpatient clinic team, our aim was to improve the number of children with CF between ages of 0-5 screened with a standardized developmental assessment tool from 0-100% within 18 months.

Methods: The implementation of developmental screening into clinic workflow was tested through plan-do-study-act cycles. Two PTs were incorporated into 3 of 4 half-day clinics per week. Children ages 0-5 were screened in outpatient CF clinic or outpatient PT clinic by a licensed PT using standardized tools. The Alberta Infant Motor Scale (AIMS) was used for infants ages 0-12 months and the Ages and Stages Questionnaire (ASQ) was used for children ages 1-5 years. The ASQ was completed by the family while in clinic and scored by the physical therapist. If these screens were positive, or other concerns were found, the Bruininks-Oseretsky Test of Motor Proficiency 2nd edition (BOT2) and the Peabody Developmental Motor Scale (PDMS) were used to further assess those children seen in the outpatient PT clinic. Appropriate referrals to community services were made as needed.

Results: Over the course of 16 months, 61 children of a total 66 possible were screened (92.4%). Thus the implementation of DD screens by a PT into clinic flow was a success. Of those screened, 17 (27.8%) were identified as having DD. Delay in gross motor skills/coordination were found in 13 children with 4 of those 13 also having speech delays. Delays in speech and/or personal social skills consistent with spectrum or ADHD type behaviors were found in 3 children, and 1 had exclusively a speech delay. Gross motor delay made up 21.3% of the total population screened, speech delay made up 13.1% of the population, and combined delays made up 11.4% of the population.

Discussion: Developmental screening can be accomplished effectively in an outpatient CF clinic setting, though it is not without its challenges. Not all children were screened within the 0-5 age range. Additionally, the ASQ is a family driven questionnaire and at times, parents verbalize anxiety about having one more thing to be concerned about (ie, delayed development) and may not accurately answer questions. Also, while the ASQ benefits a clinic setting, there is an overall lack of standardized tools that are time efficient to be performed in most outpatient clinical settings. Finally, future aims should address prevention of DD in a population not previously identified as being at-risk, as well as family education on appropriate development and enhancement of such development. Other factors that may influence or exacerbate delay, ie, socioeconomic status, access to healthcare, adherence to additional treatment plans/recommendations, and severity of disease should be included in future interventions. Acknowledgment: Support by CFF Award CCPT032-16. (CF) is characterized by a progressive decline in lung function resulting in impaired aerobic capacity that is linked to mortality. Habitual physical activity (HPA) is recommended for all individuals with CF due to its positive effect on pulmonary function, exercise tolerance, and quality of life. Research has identified barriers to adherence to routine medical care in CF; however, less is known about barriers specific to HPA. The purpose of this project was to assess the utility of a questionnaire to identify self-perceived barriers to HPA in CF.

Methods: Self-perceived barriers to HPA were assessed in adults with CF as part of routine care using a modified version of the Barriers to Being Active Quiz (BBAQ) from the U.S. Centers for Disease Control (CDC). The BBAQ calculates a total score ranging from 0 to 63 with a higher score indicating more barriers to HPA. Domain scores are also calculated for 7 common categories of barriers identified by the CDC. A score of 5 or greater in any domain category identified important barriers to HPA to overcome. Data were retrieved from retrospective chart reviews and the relationships between the BBAQ score and clinical measures of fitness were evaluated (SPSS v 25. statistical software).

Results: Seventy-one subjects with a mean age of 33.6±10.4 years (FEV 1 : 63.1±25.9 percent of predicted, BMI: 23.5±4.0) completed the BBAQ with a mean score of 15.0±9.8 (range: 0 to 39). Lack of time, energy, and willpower were identified as important barriers in 23%, 39%, and 39% of our sample respectively, and social influence was identified in 18%. In contrast, fear of injury, lack of skill, and lack of resources were rarely identified (<3%). Lack of time was more prevalent in mild lung disease whereas social influence and lack of skill were more prevalent in severe lung disease. Total BBAQ score was related to grip strength (r= -0.51, p=0.000), and perceived dyspnea (r=0.30, p=0.02) and leg fatigue(r=0.43, p=0.002) during a 3-minute step test but not related to disease severity.

Conclusion: The modified BBAQ was feasible and effective in identifying self-perceived barriers to HPA in adults with CF and led to meaningful discussions to promote HPA. The observed relationships between the BBAQ score and clinical measures of physical fitness suggests that self-perceived barriers to HPA may contribute to deconditioning. Individualized approaches are warranted; however, interventional strategies targeting lack of energy and willpower along with time management and social influence may increase HPA in adults with CF. Though lack of skill was rarely reported, it was only observed in individuals with severe lung disease. Research is indicated to determine the relationship between self-perceived barriers to HPA and actual HPA to develop appropriate behavioral interventions to increase adherence to current guidelines in adults with CF.

Acknowledgment: This project was partially supported by the CF Foundation's Award for a Physical Therapist Grant.

Bowen, M. 1 ; Duong, Q. 1 ; Russell, C. 1 ; Lamberti, J. 2 ; Connors, G. 1 ; Brown, W. 1 1. CF Center at Inova, Falls Church, VA, USA; 2. Respiratory Care Services, Inova Fairfax, Falls Church, VA, USA Objective: Published guidelines for outpatient pulmonary rehabilitation for cystic fibrosis patients are weak. Our objective was to develop and implement a Cystic Fibrosis (CF) Specific Outpatient Pulmonary Rehabilitation Program.

Background: Outside of lung transplant, there are no uniform pulmonary rehabilitation (PR) practices or programs in place geared towards CF patients. In many settings, CF patients participate in a PR program with patients who are typically much older and more likely to have other lung diseases. In many cases, contact isolation practices are not incorporated into these programs due to logistical issues or lack of awareness of the infection control guidelines. Reimbursement for PR is available depending on individual health insurance plans, some may have annual or lifetime limits for coverage.

Methods: We sought to create a unique outpatient CF PR program tailored to the specific needs of this patient population. Challenges included: availability of space, integration of infection control practices, insurance reimbursement, as well as staff availability. In collaboration with our PR respiratory therapists (RT) and exercise physiologist, we designed a CF comprehensive PR program that includes 16 sessions over an 8-week timeframe coached 1:1 by a CF RT. Patients undergo a structured educational component and exercise training. Education includes: infection control, airway clearance techniques, sleep hygiene, disease management, benefits of exercise, utilization of community resources, anatomy/physiology of the lung and coping skills for anxiety and depression. Treadmill and weight machines are utilized for aerobic and strength training purposes. Sessions are customized based on individual need. Oxygen and NIOV (noninvasive open ventilation) devices (typically for patients with FEV1<40%) are provided, as needed.

Sufficient Data: Assessment and outcome tools are measured and recorded for each patient at the beginning and completion of the program. Characteristics captured include age, gender, genotype, weight, body mass index, and spirometry. Assessments include pre-and post-program six-minute walk test (6MWT) parameters: total distance, maximum Borg dyspnea score, maximal heart rate, and nadir O 2 saturation. Oxygen and NIOV use are recorded, if applicable. Physical assessment tests are also performed at the beginning/end of the program to include: hand grip strength, 8 feet up and go, and sit to stand. Lastly, questionnaires aimed at assessing quality of life are completed before/after the program and include CFQ-R Respiratory, PHQ and GAD-7.

Conclusions: CF-specific PR can serve as an important part of the CF treatment plan, anywhere along the disease spectrum. The creation of a CF-specific program is safe, feasible and usually covered by health insurance. Goals of CF PR are unique and include augmentation of stamina and management of symptoms during physical activity, but also increase knowledge and comfort with airway clearance and exercise training in CF disease management. CF-specific PR guidelines are needed to provide structure and uniformity so that these types of programs can be available at CF care centers around the country. 1, 2 1. Medicine, Loyola Univ. Chicago, Maywood, IL, USA; 2. Medicine, Loyola Univ. Medical Center, Maywood, IL, USA Introduction: In patients with cystic fibrosis (CF) referral to lung transplant (LT) is recommended once the patient has reached an FEV 1 < 30% predicted (J Heart Lung Transplant. 2015; 34:1-15) . Need for advanced respiratory support (ARS) which includes mechanical ventilation and/or extracorporeal membrane oxygenation prior to LT is often necessary in order to bridge to transplant. The aim of this investigation is to assess the factors associated with ARS prior to LT in CF candidates.

Methods: This is a retrospective cohort of all LT referrals to Loyola University Medical Center from 2010-2017. Included are all candidates with a pulmonary diagnosis of CF referred for first time lung transplant.

Results: During the Investigation period there were 78 individuals with CF referred for consideration of LT. Of those referred, N=37 (47%) underwent LT at LUMC, 10 patients were declined, 9 were too early and actively followed, 10 did not wish to pursue lung transplant or were lost to follow-up, 6 died while awaiting LT, 4 are actively listed, and 2 were out of network for insurance. Of those who went on to active listing for LT, 16 required ARS prior to LT or death, and 28 did not require ARS prior to transplant. Those requiring ARS had lower FEV 1 percent predicted at referral compared to those who did not require ARS (median 22.5% vs. 27% [IQR 23-30], p=0.04) (Figure) . Survival from referral to transplant or death was lower in the ARS group compared to the no ARS group (median 14.9 [IQR 6.1-37.0] months versus 57.3 months, p=0.008), and lung allocation score (LAS) at listing was higher versus 37.8 ], p=0.009) despite not needing ARS at listing. CF patients who required inpatient transplant evaluations due to the need for expedited listing had an increased odds ratio of needing ARS ], p=0.001). There was no difference in post-transplant survival amongst those that required ARS versus those that did not.

Conclusions: CF lung transplant recipients who required ARS as a bridge to LT had a significantly shorter time from LT referral to transplant or death, had lower FEV 1 percent predicted at referral and higher LAS at listing. Those who required inpatient evaluation for transplant were over 4 times as likely to require ARS prior to LT. ARS prior to LT did not impact post-transplant survival. 

Ward, N. 1, 2 ; Ward, B. 3 ; Stiller, K. 4 ; Kenyon, A. 1 ; Holland, A.E. 2, 5 1. Cystic Fibrosis Service, Royal Adelaide Hospital, Adelaide, SA, Australia; 2. Physiotherapy, La Trobe Univ., Melbourne, VIC, Australia; 3. School of Physical Sciences, Univ. of Adelaide, Adelaide, SA, Australia; 4. Allied Health, Central Adelaide Local Health Network, Adelaide, SA, Australia; 5. Physiotherapy, Alfred Health, Melbourne, VIC, Australia Introduction: Adherence to airway clearance techniques (ACTs) in cystic fibrosis (CF) is often reported to be lower than for other therapies. Assessing adherence to ACTs has relied upon participant self-reports due to the lack of objective measures. Self-reports are less accurate than electronic measures for other therapies such as medications. We developed the positive expiratory pressure therapy recorder (PEPtrac) to objectively measure adherence (session attempts and within-session quality) with positive-expiratory pressure (PEP) ACTs.

Aim: To evaluate the PEPtrac in a benchtop study (PEPtrac-0) and its feasibility in a clinical population (PEPtrac-1).

Methods: PEPtrac-0: Five PEPtrac devices underwent independent calibration evaluation. The PEPtracs were then used by an investigator with the following ACT devices: PariPEP S™, combined PariPEP S™ and LC Sprint™ nebuliser, Acapella DH™, PariOPEP™ and Aerobika™. All sessions were video-recorded. An investigator blinded to the specified treatment regimen but familiar with each device's technique independently scored each session's videos to determine: time commenced, number of breaths and sets performed, and session duration. These data were compared to the PEPtrac data to assess percentage agreement for the number of breaths and sets and mean differences in commencement time and session duration.

PEPtrac-1: adults with CF were recruited to use either the PariPEP S™, Acapella DH™ or Aerobika™ with a PEPtrac, 1-2 times per day for one week. Participants also completed a daily treatment diary and a questionnaire on the PEPtrac design, including ease of use (VAS: 0 = extremely easy, 10 = extremely difficult). Data are presented as means (SD Background: Standardized exercise testing should be part of the regular assessment of patients with cystic fibrosis (CF). Benefits of physical activity in this patients include improved pulmonary function and functional capacity. An important test, that simulates normal childhood activities, is the 6-minute walk test (6MWT).

Objective: The goal of our study was to compare maximal distance, expressed as percentage predicted, covered during 6MWT in two differents groups: Sport and No Sport.

Methods: From January 2017 to April 2018, 58 stable CF patients were included. 6MWT was conducted as ATS recommendation. Children were instructed to walk as far as possible around two cones, separated by 30 meters, to complete laps at the best pace possible during six minutes. The percentage predicted of distance covered was used as the measure of performance. All subjects were asked if they were playing some sport: any activity involving physical exertion and skill in which an individual or team competes against another. Medical verbal information and Clinical Records were used to confirm that data. Group A was formed by sport practicants and Group B, by children involved in physical activities mainly at school (less than 2 hours per week). We compared 6MWD, FEV1, age, height and BMI between group A and B. Distance and percentage according to Geiger reference equation were measured (Geiger R. J Pediatr. 2007; 150:395-9) . Values were expressed as mean and standard deviation. T-test for unpaired mean was used to detect difference in distance predicted percentage between groups. Significance level was set at α< 0.05.

Results: Anthropometric characteristics and 6MWT data of the different groups are shown in the Table. The mean difference of 6MWD between both group was 31.2 meters, 95% CI [-1.208, 63.608 ]. This difference was not statistically significant (p=0.05). The difference in distance predicted percentage was 4.8%, 95% CI [1.188, 8.412 ] p=0.01. Pearson's correlations between FEV1 and distance predicted percentage in group A was r=0.28 and in group B r=0.40.

Conclusions: In our study, those patients who regularly practice sports activity walked more distance and obtained better percentage predicted according to reference equation versus patients who perform physical exercise at school but no sport. This difference was statistically significant. We found that FEV1 was the only independent variable significantly different between groups. We concluded that 6MWT may be useful to evaluate exercise capacity in our hospital when a CPT in laboratory is impossible or not recommended. 

Introduction and Objectives: Cystic fibrosis (CF) is the most common fatal genetic disease affecting multiple body systems. In Canada, the median survival of people with CF has reached 53.3 years. As people with CF are living longer, there are more attention directed towards rehabilitation related issues, such as muscular issues arising from the disease. The objectives of this study were to compare peripheral skeletal muscle structure and function in adults with CF to healthy controls; and to investigate their associations with clinical variables in adults with CF.

Methods: A systematic review was performed using PRISMA guidelines. Electronic databases were searched including MEDLINE, EMBASE, CINAHL, AMED, and CENTRAL. The eligibility criteria were human studies with measurements in peripheral skeletal muscle structure and/ or function in adults with CF. Articles published in language other than English, and studies with data from both children and adults combined were excluded. Descriptive statistics [median (min-max)] were computed. Quality assessment was performed using the Modified Downs and Black checklist. A meta-analysis was also performed. Only studies with healthy controls were included. Standardized mean differences (SMD -healthy controls minus adults with CF), and their variance were calculated for each study. A random-effects model was used to calculate the effect sizes and confidence intervals (CI); and I 2 was calculated to assess the heterogeneity of the studies.

Results: Out of 1,360 articles retrieved, 19 were included in the systematic review; nine of which were used in the meta-analysis. For adults with CF, the median sample size was 18 (7-64); median age was 28.0 years (22.9-35.2 years); BMI=20.6 kg/m 2 (17.7-22.1 kg/m 2 ); FEV 1 =56.1% (20.9-72.0%) predicted. The quality of the studies was 15 (11-20) out of 27. The majority of the studies (15 out of 19) focused on quadriceps muscle strength and/or leg muscle size. The meta-analysis showed that compared to healthy controls, adults with CF have smaller leg muscle: 3.3 (95% CI: [0.5, 6.1], p<.021, I 2 =96.7%); lower quadriceps muscle strength: 3.1 (95% CI: [2.1, 4.0] , p<.0001, I 2 =86.7%); and reduced handgrip strength: 3.9 (95% CI: [2.1, 5.8] , p<.0001, I 2 =89.4%). There was a significant but small correlation between handgrip strength and FEV 1 (r=0.24, 95% CI: [0.020, 0.45], p<.035, I 2 =0%). No significant correlations were found between quadriceps muscle strength and FEV 1 and FVC.

Conclusion: Adults with CF have smaller leg muscle, weaker quadriceps muscle and handgrip strength than healthy controls, although this should be interpreted with caution because of the high heterogeneity among studies. There was a mild association between lung function and handgrip strength but not with quadriceps muscle strength. Future studies should include larger sample sizes and with older adults with CF. Factors that contribute to muscle weakness in adults with CF and its clinical implications need to be investigated further. The Cystic Fibrosis Foundation (CFF) recommends aerobic exercise for cystic fibrosis (CF) patients of all ages as an adjunctive therapy for airway clearance and for its additional benefits to overall health, including higher quality of life scores, lower anxiety and depression scores, and better adherence. Exercise also helps to maintain lung function, maintain good posture, build muscle strength, and increase endurance. The 2015 CFF Patient Registry Report showed 45.2% of adults identify exercise as one of their methods of airway clearance. This data is consistent with patient data from our center (44.6%); however, exercise is more prevalent as the primary airway clearance method.

Objective: The aim of this project is to understand current exercise regimens and identify the barriers to completing these routines in adult CF patients at a large CF care center.

Methods: A survey was distributed to a cohort of adult CF subjects during a clinic appointment. The survey consisted of seven questions that asked about the participant's current exercise regimen, type and intensity of exercise, and perceived barriers to exercise. Overall patient demographics and pulmonary function were also recorded.

Results: The survey was completed by 43 CF subjects (ages 21-64). Lack of time was the top barrier to exercise (23%), followed by lack of energy (22%) and lack of motivation (20%). More than half (56%) of the population surveyed is working or going to school more than 20 hours per week. Nineteen different types of exercise were reported, with walking (69%), lifting weights (44%), and yoga (39%) being the top three. More than half (58%) of the patients reported exercising two or more times a week at moderate to high intensity, and this same group also stated that exercise is very or extremely important to them. This perceived importance of exercise is consistent across a range of FEV 1 , percent predicted (28-114%). Time is a significant barrier for patients working 20 or more hours per week (p=0.024). Depression/anxiety and a lack of motivation is significantly higher in patients working less than 20 hours per week compared to other groups (p=0.011, p=0.022 respectively). The barrier of "not feeling well" had little variation between all groups.

Conclusion: There are several barriers to exercise in an adult CF cohort, including time, energy, and motivational constraints. Despite barriers, patients are able to incorporate exercise regimens into their schedules, even with a wide range of pulmonary function. A personalized approach to a patient's barriers to exercise by multidisciplinary CF care team members may expand current exercise participation rates and attitudes towards exercise. A larger study is needed to determine how exercise-based interventions would impact overall respiratory and mental health.

Suggett, J. 1 ; Coppolo, D.P. 2 ; Meyer, A. 1 1. Science and Technology, Trudell Medical International, London, ON, Canada; 2. Monaghan Medical, Syracuse, NY, USA Introduction: Oscillating positive expiratory pressure (OPEP) devices can be used to manage a variety of conditions, such as CF, COPD and bronchiectasis through a general mechanism of opening / vibrating airways and loosening mucus, however, the specific mechanism by which this is achieved differs between devices. This investigation assessed the positive pressure oscillation waveforms of various devices and linked the critical performance attributes of pressure pulse amplitude and frequency in order to compare potential effectiveness.

Methods: A simulated OPEP exhalation maneuver was generated based on previous research (Meyer, et al. Am J Respir Crit Care Med. 2014; 189:A3036) in which a flowmeter recorded the waveforms of healthy volunteers and an average profile was scaled so the peak expiratory flow rate was 30 L/min, being more patient representative. This waveform was then used to operate, via a breathing simulator, a range of different OPEP devices: Aerobika* (TMI, Canada); vPEP (D R Burton, US); VibraPEP (Curaplex, US); Acapella Choice (Smiths Medical, US); Flutter (Axcan, US), (n=3 devices, 3 replicates of each). The pressure / time waveforms were recorded for each device, set at their highest resistance to enable direct comparison. Analysis of each device waveform was performed in order to determine the total pressure pulse impact around optimum oscillation frequency (TPPI F ). This is defined and calculated as the sum of all discernable (>1.0 cm H 2 O) pressure pulse amplitudes, in a single exhalation waveform, that are in the frequency range of 10-15 Hz.

Results: Each device waveform had its own unique pressure pulse pattern. The Table summarizes the TPPI F values for each device.

Discussion and Conclusions: The use of the TPPI F value to assess the therapeutic effectiveness of air flow oscillations is supported by the twin assumptions of a) effectiveness is dependent, in part, on the ability of the device to generate and maintain a pressure amplitude / spike throughout the maneuver (Van Fleet H, et al. Respir Care. 2017; 62(4) :451) and b) effectiveness is optimized at a frequency of approximately 13 Hz (Silva CEA, et al. Respir Care. 2009; 54(11) :1480-7). The TPPI F values showed the Aerobika* device to be the most effective, with a 54% higher value than the second ranking device and more than four times the value of the lowest performing device. The combination of pressure and frequency in assessing device performance also supports the value of a pressure manometer attachment to OPEP devices. It is recognized that the implications of these reported laboratory differences should be assessed in a clinical setting also, however they do provide useful relevant insights when selecting a device for clinical practice.

Bass, R. 1, 2 1. Physiotherapy, Newcastle Upon Tyne Hospitals, Newcastle Upon Tyne, United Kingdom; 2. Cystic Fibrosis Trust, London, United Kingdom Objective: To describe the experience of the first Physiotherapy Fellowship Programme, its benefits and opportunities.

Methods: The Cystic Fibrosis Trust worked alongside the Association of Chartered Physiotherapists in Cystic Fibrosis (ACPCF) to develop a programme designed to enhance expertise and leadership within physiotherapy. The fellowship was created to inspire therapists to work towards obtaining the desired level of experience required to lead a cystic fibrosis (CF) physiotherapy service. The clinical fellowship allows for dedicated learning and development time whilst providing appropriate clinical backfill cover. It has supported attendance at international conferences, key national courses and meetings. It has also permitted intensive development opportunities at other centres, working alongside a variety of experienced clinicians. The programme provided the chance to review current practice within the home centre and, the identification and implementation of local service improvement ideas and research proposals.

Results: The learning opportunities available addressed a number of key areas: airway clearance, infection control measures and exercise prescription. Throughout the fellowship, the information and knowledge gained has been shared with the local physiotherapy team, developing skills and service further. The ongoing research exploring the use of online exercise videos has the potential to benefit the wider CF community and physiotherapy profession. The experiential learning has greatly increased confidence, allowing a positive contribution to service discussions and developments. Challenges faced during the fellowship were primarily due to it being the first of its nature within physiotherapy.

Conclusion: This fellowship programme has been beneficial to the fellow, the home CF centre and the national CF community. Further fellowship opportunities will be enhanced due to the experiences gained in this inaugural year.

Hemphill, M.T.; Cross, C.; Helton, R.; Jones, K.; Allen, K.S.

Introduction: Approximately one-half of the variation in cystic fibrosis (CF) disease status is secondary to nongenetic variables. One variable that has been a focus of attention is the role of excercise as it relates to quality of life in CF patients (Collaco JM, et al. BMC Pulm Med. 2014; 14:159) . Thus far, studies aimed at determining an association between exercise and quality of life have produced mixed results (Schmidt AM, et al. Physiother Theor Pract. 2011; 27:548-56; Rovedder PM, et al. Respir Med. 2014; 108:1134-40; Hebestreit H, et al. BMC Pulm Med. 2014; 14:9) . Although results have been variable, there is a positive association between exercise and quality of life in the general population. Furthermore, physical activity has been shown to be safe in CF. Therefore, it seems reasonable to assess regular exercise and health-related quality of life in CF. This study aims to evaluate personal fitness devices (PFD) in adults with CF.

Materials and Methods: This is a prospective single-group, unblinded study, in which adult patients wear a PFD for 12-14 months. They are given information about the benefits of physical activity and set a personal fitness goal at an intial visit. Cystic Fibrosis Questionnaire-Revised (CFQ-R) and 6-minute walk test (6MWT) are then performed to obtain control data. Patients are evaluated at 3-month intervals, with CFQ-R and 6MWT data collected at 6 and 12 months. The primary outcome evaluated is detection of a minimal clinically important difference improvement in CFQ-R over 12 months. Secondary outcomes are change in CF exacerbations, forced expiratory volume in one second (FEV1), admission rates, BMI, and 6MWT. The t-test was used to determine statistical significance.

Results: Thus far, 12 patients have completed 6-month follow-up. Average initial CFQ-R is 72.2. Most patients selected a certain number of steps as their goal, with average being 7100 steps. At follow-up, the average patient-reported number of steps was 6250. No statistically significant difference has been detected in either overall CFQ-R (p 0.65) or any subcategory. In regards to secondary outcomes, only change in BMI reached statistical significance (23.8 to 24.4, p 0.006). Weight also increased, but did not reach statistical significance (146.0 to 149.1, p=0.94). Although not statistically significant (p 0.25), there was a definite increase in 6MWT distance of 244 meters.

Discussion: At this point in the study, no difference has been observed in the primary outcome of CFQ-R score. This may be due to several factors. First, a follow-up of only 6 months may not be sufficient to detect a difference. As patients exercise for up to 12 months, a more clear difference may be observed. Next, it appears that patients with higher FEV1 were enrolled initially (average FEV1 62.3% for first 12 versus 54.4% for final 8 patients). As follow-up data for those with more advanced disease are included, it would be reasonable to expect a clearer signal of improvement. Furthermore, there was an increase in 6MWT distance (though not statistically significant), which may become more apparent as 12-month data are collected. Based on patient feedback, we believe that the study examines a sustainable exercise regimen not previously studied.

Welsner, M. 1 ; Sutharsan, S. 1 ; Stehling, F. 2 ; Koerner-Rettberg, C. 3 ; Benzrath, S. 3 ; Gruber, W. 2 1. Dept. of Pulmonary Medicine, Adult Cystic Fibrosis Center, Ruhrlandklinik, West German Lung Center, University Hospital Essen, Essen, Germany; 2. Pediatric Pulmonology and Sleep Medicine, Cystic Fibrosis Center, Children's Hospital, Essen, Germany; 3. Dept. of Paediatric Pulmonology, University Children's Hospital of Ruhr University Bochum, Bochum, Germany Introduction: In the interdisciplinary treatment of patients with cystic fibrosis, physical activity and exercise has proven to be particularly effective and is therefore an important part of current therapy recommendations. The project CFmobil will investigate the effects of a partially supervised exercise program on cardiopulmonary performance, pulmonary function parameters, health-related quality of life, actigraphy-measured habitual physical activity and sleep parameters over 12 months.

Methods: Participants receive an individualized and supervised exercise program based on daily physical activity, personal interests and physical performance. Patient contacts are scheduled every 2 weeks within the first three months, then at intervals of four weeks. Based on the stress test at the beginning, after 6 and 12 months, the intensity of the training will be adjusted during the intervention period. The participants wear a wrist activity monitor (wActiSleep-BT, Actigraph, Pensacola, USA) for 28 days each 4 weeks prior to the start of the intervention period and after 3 respectively 12 months. During these periods, daily physical activity and sleep parameters are recorded in the home environment.

Results: A total of 136 adult CF patients were included in the study. 39 patients (23 male, mean age 27 years ± 7.8, FEV1 percent pred. 53.3 ± 21.6, BMI 19.9 ± 2.7) reached time point T2 after three months of intervention. At T2 there was a nonsignificant increase in actigraphy-measured steps taken per day (8501 ± 3088 vs. 8757 ± 3837) and a shift in physical activity intensity (MET/min/day): light (<3 MET/min/day 569. 9 ± 433.3 vs. 443.7 ± 348.4, p<0 .05), moderate (3-6 MET/min/day 99.9 ± 84.8 vs. 81 ± 62.1, p<0.05), vigorous (>6 MET/min/day 10 ± 11.7 vs. 12.6 ± 13.2, p>0.05) and very vigorous (>9 MET/min/day 3.5 ± 7.1 vs. 5.5 ± 11.8, p>0.05) . There was only a small non-significant influence of the increased and intensified daily activity on the actigraphically measured sleep parameters within the first three months of intervention: Sleep Efficiency (%) 90.5 ± 2.9 vs. 90.9 ± 3.4, Total Time in Bed (min) 493 ± 70 vs. 509 ± 105, TST (min) 449 ± 67 vs. 467 ± 104, WASO (min) 43 ± 13 vs. 42 ± 14, # of Awakenings 10.9 ± 4.2 vs. 10.6 ± 5.1, Awakening (min) 4.2 ± 1.4 vs. 4.4 ± 1.6.

Conclusion: Within the first three months of a partially supervised exercise program, there was a tendency for adult CF patients to increase daily activity with an increased number of steps taken per day and more intense exercise phases. Despite the increased daily activity, the sleep habits of the participants do not appear to change significantly. It remains to be seen if the participants will experience a further increase in daily activity within the CFmobil project and whether there will be an impact on sleep habits after prolonged participation. 

Background: Exercise improves aerobic capacity and may stabilize lung function in adults with cystic fibrosis (CF). However, many adults with CF do not exercise regularly. Common barriers include lack of motivation, support, and feedback. In-person peer support groups are precluded due to CF infection control. Wearable fitness trackers integrated into social media may increase exercise activity and tolerance by providing patients their own data and virtual support communities.

Objective: To determine if a personalized exercise prescription provided by a physical therapist, plus use of a wearable fitness tracker integrated with a social media platform is associated with increased exercise tolerance in adults with CF compared to an exercise prescription alone.

Methods: This was a pilot randomized clinical trial (RCT) at a large CF center. We randomized 40 adults with CF, stratified by age and FEV1, to either receive a fitness tracker and exercise prescription (N=19) or prescription alone (N=21). Recruitment occurred May 2016 -April 2017. Subjects were followed for 12 months to account for seasonal changes which might affect exercise frequency and to assess for sustained effects.

Outcomes: The primary outcome was the mean level completed on a submaximal graded exercise test (GXT) 12 months post-randomization. Secondary outcomes included change in GXT and FEV1 over 1 year, mean FEV1 at 1 year, and mean scores on the CF Questionnaire-Revised (CFQR), General Anxiety Disorder (GAD-7), and Patient Health Questionnaire (PHQ-9).

Results: Stratified randomization resulted in 2 groups with similar demographic characteristics. Combined, the cohort mean age was 35 (SD=14 years) and 55% male. Baseline GXT was 6.1 (SD=1.8) for the intervention group and 5.7 (SD=1.6) for the control group. At 1 year of followup, there was no significant difference in the mean GXT (5.9 (SD=1.4) for the intervention group vs 6.3 (SD=1.7) for the control group), nor was there a significant difference in the change in GXT in either group. There was no significant difference in mean FEV1 (67% pred. (SD=27) for the intervention group vs 66% (SD=26) for the control group) at 1 year. Rates of decline in FEV1 over the year did not differ statistically between groups. Mean scores on the CFQR, GAD-7, and PHQ-9 showed no significant differences between the 2 groups at baseline or at 1 year. Daily tracker activity data indicated increasing frequency of days with 0 daily steps over time.

Discussion: In this pilot RCT, provision of a fitness tracker with an exercise prescription did not result in significant improvements in exercise tolerance, pulmonary function, or patient reported outcomes compared to a prescription alone. Short-term studies suggest that trackers increase exercise frequency. However, our longer-term study is consistent with other longer-term studies assessing the effects of trackers on peripheral artery disease and weight loss. Inconsistent use of the tracker over time may account for some of these findings. The current data do not support routine use of fitness trackers in adults with CF to increase exercise tolerance. Acknowledgment: Support by the Steven J Vertuccio Fund.

Reimpell, P. 1 ; Fuchs, C. 2 ; Junge, S. 3 ; Framke, T. 3 1. PARI GmbH, Starnberg, Germany; 2. PARI Medical Holding GmbH, Starnberg, Germany; 3. Pediatric Pneumology, Medical School Hannover, Hannover, Germany Introduction: High therapy adherence is important for effective treatment of CF. However, the time burden of the many necessary therapies makes this difficult. The study's primary objective was to determine average patient adherence with respect to nebulization therapy in daily life without any intervention to enhance adherence. The baseline value may be used as a reference for future adherence-increasing interventions. Secondary objectives included effects of number of comorbidities, treatment burden and P. aeruginosa-status on average adherence. Adherence effects on lung function parameters and BMI scores were also investigated.

Method: The observational study was carried out with 36 CF patients (6 to 17 years old, 39% female) at the pediatric CF center of the Hannover medical school in Germany and followed a prospective, single arm, single blinded observational study design. Data collection was supported by telemetric methods. The subjects were a pediatric patient cohort suffering from cystic fibrosis who regularly self-administered nebulization therapy. Telemetric monitoring was facilitated through a Bluetooth-enabled eFlow rapid nebulizer (PARI Pharma GmbH, Germany), which recorded the period during which the device was nebulizing medication. Data were wirelessly transmitted from the nebulizer via a paired mobile phone to a central server for data analysis. The total study duration was set out to be 6 months, of which only the first 3 months were telemetrically monitored. Patients had a consultation with their physician at three points: (1) study start, (2) between the two phases and (3) at the study end. During the consultation, the lung function parameters FEV 1 , FVC, MEF25, BMI-Z scores and other clinical parameters were measured. There was no special intervention to increase therapy adherence. The investigating clinician did not know a patient's adherence during any point of the study due to the single blinding to inhibit bias.

Results: In the study, patient adherence averaged at 60%.The study found that adherence declined as the study progressed. Between the first and third month, adherence decreased by 9% on average. It was also shown that patients with less arduous treatment plans were more adherent than those with more daily prescribed nebulizations. In accordance, patients who were prescribed antibiotics for nebulization were also found to be less adherent. Patients with high adherence rates (>75%) had a better FEV 1 development with an increase of 3.9% compared to an average decline of FEV 1 of the total patient group of 1% over the monitored period of 6 months.

Conclusions: Therapy adherence is an important parameter in clinical trials and is also associated with better patient health outcomes. Using a new device and awareness of ongoing adherence tracking, probably motivated some patients to be more adherent in the beginning of the study resulting in the decrease of adherence over time. Consistent adherence is especially important for patients who need to nebulize antibiotics. New technology, like apps with gamification may offer opportunities to help these patients engage with their disease and follow their therapy more diligently. Background: Oscillating positive expiratory pressure (OPEP) devices are indicated for the removal of excess secretions and reduce gas trapping in patients with hypersecretory pulmonary diseases such as cystic fibrosis (CF). OPEP works when the patient exhales actively against a fixed resistor, generating mean intrapulmonary pressures of 10-20 cmH 2 O with rapid fluctuations of at least 1 cmH 2 O. The recommended duration of expiration, not including breath-hold, is 3-4 times the length of inspiration representing an inspiration: expiration (I:E) ratio of 1: 3 or 1: 4. This prolonged, steady exhalation splints open collapsed small airways. The cycle should be repeated 10-20 times with 2-3 additional "huff" coughs to clear any loosened secretions.

In this study, we sought to evaluate paediatric CF patient performance of OPEP therapy to determine compliance with target therapeutic pressures.

Methods: A convenience sample of twenty-one paediatric patients was recruited for the study. Each participant had a well-documented history of using an OPEP device twice daily, and had received standardised training and instructions from the same specialist physiotherapist. Performance was evaluated using a bespoke designed flow and pressure sensor placed in-line between the participant's mouth and device. Participants were asked to perform ten expirations as per their normal routine. Each expiration was recorded separately for post hoc analysis.

None of the participants achieved the target therapeutic pressure ranges during expiration. The mean pressures generated across all participants were 16.19 ± 6.78 cmH 2 O, while mean flow = 31.28 ± 8.85 L/minute. The mean expiration length was 2.53 ± 1.36 seconds. The mean I:E ratio across the 21 participants ranged from 1: 0.36 to 1: 2.42, with an overall average of 1: 1.24 (n=209 expirations).

Interpretation: Despite standardised instruction, the results demonstrate that there is considerable variation between participants and overall poor technique during OPEP therapy. It was noted by the authors that several parents were actively encouraging their child to complete shorter and more forceful expirations. This well-intentioned, but ultimately misinformed emphasis appears to rely on the premise that "more is better" as would be the norm during lung function testing. The outcomes of this study suggest the airway clearance effects of OPEP could be compromised due to poor technique as there appears to be a conflation between FEV 1 testing and OPEP therapy.

Mean pressure and flow generated for all recorded breaths. Therapeutic target range shown in shaded box. Introduction: High-frequency chest wall oscillation (HFCWO) vests are the current standard of care to help manage certain respiratory conditions. Although they have been shown to result in increased mobilization of mucus, the mechanism of action is not fully understood. One postulated mechanism is increased cephalad airflow bias, but it remains unclear if evidence for this effect is supported by standard clinical spirometry measurements. We performed this study to investigate short-term application of motor and compressor-based HFCWO vests on spirometry parameters TV, PEF, FVC, FEV1 and FEF25-75%. Another aim of the study was to determine if there were any significant differences between the two device types.

Methods and Analysis: We conducted a prospective, three-arm study in healthy subjects. Each subject was assessed according to ATS guidelines using standard spirometry techniques and equipment. In each arm, a motorbased vest (AffloVest®) and one of the three compressor-based vests (The Vest®, SmartVest®, inCourage®, each n=10, total n=30) was evaluated. Consecutive subjects in each arm were fitted with the two different types of HFCWO vests in alternating order. Lung function tests were performed at baseline and while using each vest. The primary objective of the study was to compare each spirometry measurement at baseline vs during use of the HFCWO vests.

Results: Enrollment included 32 subjects and 2 subjects withdrew. There were no statistically significant differences in TV or PEF between baseline and during HFCWO use for any groups. FVC was significantly decreased vs baseline in the aggregate compressor-group (mean = 4.12 L, range [2.29-6 .73] L vs 4.29 L, [2.48-6 .57] L, p=0.019) as well as FEV1 (3.30 L, [1.92-5.83 ] L vs 3.51 L, [2.05-5.54 ] L, p<0.005) and FEF25-75% (3.19 L/s, [1.19-6 .22] L/s, vs 3.71 L/s, [1.77-6 .43] L/s, p<0.005). Only FEF25-75% was significantly decreased vs baseline in the AffloVest group (3.54 L/s, [1.63-6 .37] L/s vs 3.71 L/s, [1.77-6.43 ], p = 0.031). FEV1 and FEF25-75% were statistically significantly lower in the compressor-based group compared to the AffloVest group.

Discussion: We show for the first time that compressor-based HFCWO vests significantly decreased FVC, FEV1 and FEF25-75% during use. Only FEF25-75% was significantly decreased in the AffloVest group during use, and the decrease was significantly less than in the compressor-based group. It has been speculated that increased cephalad airflow bias is the mechanism of action of HFCWO vests. However, no increase in any spirometry parameters were found in this study and the link between alterations in airflow and mobilization of mucus has never been established in clinical studies using HFCWO vests. Furthermore, airflow measurements taken at the subject's mouth have not been correlated to any effects in the flexible airways, where cephalad airflow bias results in mucus mobilization. This study demonstrates that the concept of HFCWO induced cephalad airflow bias during use is not supported by standard pulmonary spirometry measurements.

Acknowledgements: Supported by International Biophysics Corporation, manufacturer of the AffloVest. During the newborn screening process, 18 children have been diagnosed with "Cystic Fibrosis Screen Positive, Inconclusive Diagnosis" (CFSPID) also referred to as "CFTR-related Metabolic Syndrome" (CRMS).

Objectives: To describe the use of airway clearance in the children with CFSPID at BC Children's Hospital.

Subjects: Eighteen children were diagnosed with CFSPID, at BCCH, since the inception of the newborn screening program in 2009 (4 females and 14 males). The age ranges at the time of chart evaluation was 6 months to 15 years. Sixteen of the subjects were diagnosed with CFSPID through newborn screening, one was diagnosed because of a sibling diagnosis and one was diagnosed by genetic testing due to the presence of nasal polyps. All are pancreatic sufficient.

Methods: A chart review was completed in February 2018 of the 18 children diagnosed with CFSPID. Five children that later converted to a CF diagnosis from an initial CFSPID diagnosis were included. Data were collected on their airway clearance teaching and techniques, as well as hospitalizations, Pseudomonas aeruginosa (Psa) acquisition and sweat chloride testing.

Results: A physiotherapist was present at the initial teaching sessions for each CFSPID diagnosis. At this time, education was given that focused on exercise and activity. An airway clearance technique (ACT) was taught to families and encouraged to use when the child was unwell with respiratory symptoms. At the time of the chart review, 11/18 children continued with the same physiotherapy program that was initiated at diagnosis. In this group, two children used the active cycle of breathing as their ACT and 9 children used modified postural drainage with percussion and vibration. None of these children were hospitalized or acquired Psa and all had sweat chloride levels less than 60 mmol/L. Seven of the 18 children were recommended to commence an ACT on a daily basis. Daily ACT was commenced due to chronic respiratory symptoms (2 children) or a subsequent positive sweat chloride test (5 children) . In this group, two of the children used positive expiratory pressure (PEP) mask as their ACT and the remaining 5 children used modified postural drainage with percussion and vibration. Two children were hospitalized for a pulmonary exacerbation and received IV antibiotics. One child acquired Psa. All 18 children continue to be followed by the physiotherapists at the CF clinic. Education as well as exercise and activity remain a priority for all. ACTs are reviewed at each clinic visit.

Conclusions: The CFSPID population is diverse and demonstrates varied clinical presentations. Due to the heterogeneity of this population, a personalized approach to their care is required. It is important that education, exercise and activity, are discussed with all patients and their families. Airway clearance techniques should be considered and a physiotherapist should maintain a role in their care. More data are required to determine the efficacy of airway clearance in the CFSPID population.

Richards, K. 1 ; Bernardo-Stagg, K. 1 ; Hill, S. 1 ; Miller, H. 1 ; Barker, D. 1 ; Asfour, F. 2 1. PFT Lab, Primary Children's Hospital, Salt Lake City, UT, USA; 2. Pediatric Pulmonology, Univ. of Utah, Salt Lake City, UT, USA Background: The standard of care at our Pediatric CF outpatient clinic was to provide verbal instruction and patient demonstration with each new respiratory therapy added to a patient's regimen and review therapies at their annual respiratory assessment (ARA). Families were instructed to call with questions. Feedback received from families of newly diagnosed infants (age ≤ 12 months) revealed feeling overwhelmed by the volume and complexity of the respiratory treatment regimen, which includes metered dose inhaler (MDI) with spacer, chest physiotherapy (CPT) and nebulized inhaled medications. We hypothesized that follow-up with families of newly diagnosed CF infants post-education (by phone and in clinic) would help them successfully and confidently perform these therapies.

Methods: In January 2016, we initiated a follow-up process of calling the caregiver approximately 10 days following education (MDI/spacer and CPT, nebulizer teaching) using a standardized set of questions to assess proper performance and level of comfort with each therapy. At patient's next clinic visit, the caregiver demonstrates the technique and answers the questions again. We assign a 1 (correct) or 0 (incorrect) value to the standardized questions and rate comfort level on a 1 (uncomfortable) -4 (very comfortable) scale. Information is compared to data gathered at patient's ARA. Gathered data will be used to assess if our phone and in-clinic follow-up improved the proportion of caregivers correctly performing respiratory therapies after initial education and sustained at ARA. We also compared retrospective-age matched controls by evaluating ARAs in the 2 years prior to this intervention.

Results: Thirty-two newly diagnosed infants participated in this project, 27 received a phone call and all 32 received a clinic follow-up. Proper technique was assessed for each therapy and at least 2/3 of caregivers (61%-94%) performed correctly with a trend towards improvement with in-clinic follow-up compared to phone call. Of those who completed ARA, correct technique was sustained. Correct technique was greater in this group when compared to the matched controls (See Table) . In general, the proportion of caregivers who rated comfort level 4 (very comfortable) was greater in clinic follow-up (58%-90%) as compared to phone (48%-92%).

Discussion: Families of infants with CF have a complex daily care regimen. Effectiveness of respiratory treatments is hinged on proper adherence and utilization. Time spent in follow-up improved patient care by allowing for early identification of improper technique with re-education when indicated. This follow-up is a vital component of CF care and warrants ongoing utilization and evaluation. Introduction: Airway clearance therapy (ACT) is a standard daily treatment for CF. Adherence to ACT in children with CF has been estimated at 50%, but studies have relied on self-reports. We compared self-reported ACT adherence to actual usage data from home ACT equipment, a high-frequency chest wall oscillation (HFCWO) vest, and identified factors associated with overestimation of adherence in self-reports.

Methods: Pediatric patients who perform ACT with HFCWO vest (N=161) were eligible to participate. Objective adherence data were obtained from the HFCWO device, which records cumulative utilization time. Two readings were collected: a baseline within 2 days of enrollment, and a follow-up 5 weeks later. The difference between readings, divided by the number of days between, provided a mean daily use. Objective adherence was represented as a ratio (%) of mean-to-prescribed daily use. Self-reported adherence data were collected with a caregiver survey at study enrollment using questions from the validated Treatment Adherence Questionnaire-CF. Adherence rates were coded as low (<35% of prescribed), moderate (36-79% of prescribed), and high (≥80% of prescribed), following previously used cut-offs. An overestimation (Yes/No) was present when self-reported adherence was at least one category higher than objective adherence (e.g., high vs moderate, moderate vs low).

Results: Of those who enrolled (n=120), 116 (97%) completed the study. After accounting for missing data, the final analysis included 110 participants. The majority of caregivers (85%) were female. Mean caregiver age was 41.14 (SD 8.01) years, mean patient age was 11.19 (SD 4.72) years. Forty-one percent of households had income <$50,000 and 35% had public health insurance. Prescribed daily ACT was 60 minutes or longer for 84% of patients. In 9% of cases, ACT took place in multiple households (e.g., divorced parents or care split between parents and grandparents). Mean time between data readings was 37 days. Mean adherence rate by actual usage data was 61% of total prescribed time. Only one-third of patients (35%, n=38) were highly adherent, and 28% (n=31) were low-adherent. In contrast, nearly two-third of caregivers (65%, n=72) self-reported high adherence and only 8% (n=9) self-reported low adherence (p<0.001). Nearly half of caregivers (46%) overestimated their adherence. In multiple logistic regression adjusting for demographic and clinical characteristics, adherence overestimation was associated with multiple ACT locations (OR 5.30, p<0 .05) and prescribed ACT duration ≥ 60 min (OR 3.85, p<0.05). Among patients with prescribed ACT ≥ 60 min, after adjusting for demographic and clinical covariates, income <$50,000 increased 3-fold the odds of overestimating adherence (OR 3.04, p<0.05).

Conclusions: Caregiver self-reports overestimate actual adherence to ACT. The overestimation increases with treatment in more than one household and longer duration of prescribed ACT. Among patients with prescribed ACT of 60 min or longer, overestimation increases with low income. Our data underscore the necessity of objective measures of adherence, particularly with low-income children and those receiving treatments in multiple locations. , however other articles state that exercise over several months has negligible effects on lung function (Kriemler S, et al. Lung. 2016; 194(4) :659-64). While many factors play a role in a person's lung function, including but not limited to; female gender, frequent or productive cough, low BMI (<66th percentile), ≥1 pulmonary exacerbation (Cogen J, et al. Pediatr Pulmonol. 2015; 50(8) :763-70), patterns have been noted in the clinic while several patients have been undergoing supervised exercise therapy.

Methods: The following cases were a sample of convenience and completed as a retrospective study. The records of a 64-year-old female with cystic fibrosis (patient x) and a 17-year-old female with cystic fibrosis (patient y) were reviewed. Neither patient is on CFTR modulator therapy, since they are each heterozygous for F508del and class I mutations. Both patients have had multiple sessions of high-intensity exercise training, which include lower extremity strengthening and treadmill training. Target heart rates were determined utilizing the modified shuttle walk test and Karvonen's formula. FEV1 readings were obtained in the cystic fibrosis clinic during regularly scheduled visits as determined by the provider. Frequency of exercise was determined through chart review/patient report or directly from physical therapy notes.

Results: Over the span of 4 years, patient x had multiple exacerbations and medical issues not related to cystic fibrosis. However, every time she presented with an increase in lung function, she had reported exercising on a regular basis or had completed a physical therapy routine (minimum of an increase of 12%). With patient y, a steady decline in FEV1 has been noted since March 2015 (72%) to January 2018 (33%). Multiple medical interventions have taken place since then, including hospitalizations, oral and IV antibiotics. Patient y did achieve a high of 43% FEV1 on 1/25/2018 but dropped again to 34% by 3/9/2018. She started exercise therapy on 2/26 for 1 hour 1x a week, then 2x a week in the middle of March 2018. She also reported completing more exercise independently at home during this time. Since March 9th, the FEV1 has not been lower than 38% (with a high of 42% on 3/26).

Discussion: While these cases are quite different, with different medical histories and interventions, one constant that is interesting is a significant increase in FEV1 during periods of high intensity exercise (target heart rate of 60-85%) or self-reported exercise. The improvements in FEV1 cannot attributed exclusively to exercise, as these were uncontrolled studies that included other environmental factors and medical interventions. The observations here warrant further study into the true effect of exercise on FEV1, in a controlled setting with a larger sample size, possibly multicenter and statistically similar groups. Introduction: Exercise, physical activity and airway clearance have been shown to decrease the decline of lung function over time in patients with cystic fibrosis (CF). Physical therapy (PT) is uniquely designed to address these aspects of CF care. Quality improvement (QI) data was collected to determine the utilization and impact of increasing access to physical therapy intervention in these patients.

Method: PT services were initiated in both the pediatric and adult CF clinics at Mass. General Hospital for a total of 8 hours per week in 2016. This was increased to 16 hours starting in May of 2017 with the implementation of the CF PT RFA from the CF Foundation (CFF). Patients were seen at their regular quarterly visits and scheduled for comprehensive annual PT visits during this time. PT interventions consisted of consultation on airway clearance, exercise, physical activity, musculoskeletal issues and pelvic floor issues. Retrospective QI data was analyzed to determine utilization of PT services and assess effect. Data collection and analysis are currently ongoing over the three-year grant duration.

Results: PT CF clinic visits increased in both the pediatric and adult CF clinics from 2015 to 2017. Adult clinic saw an increase in PT utilization in clinic from 0% of patients in 2015, to 62% in 2016 and 82% in 2017. PT visits in the pediatric clinic increased from 0% of patients in 2015, to 45% in 2016 and 93% in 2017.

Adult patients that had been seen by PT in CF clinic received IV antibiotic days for an average of 23.5 days in 2016 to 21.6 days in 2017 whereas the patients who did not see PT in clinic received IV antibiotics for a longer duration of 31.33 days in 2016 to 30.5 days in 2017. This analysis has not yet been performed in the pediatric CF population.

Discussion: Overall utilization of physical therapy increased dramatically from 2015 to 2017 in both pediatric and adult clinics as expected. Adult patients who saw a physical therapist averaged 7.83 less IV antibiotic days in 2016 and 8.9 less days in 2017 which represents a significant savings of cost to the healthcare system and a potential decrease in antibiotic-related side effects. Future and ongoing analysis of the effect of PT on pediatric antibiotic days, adult and pediatric quality of life measures (AweScore), physical activity scores (Habitual Activity Estimation Score), and measures of aerobic capacity are needed to correlate the decrease in IV antibiotic days in patients with PT intervention. Additional analysis of subsequent years is also required to further validate these observed data trends. . Application of these criteria in pediatrics can be challenging. This pediatric CF center had no formal process for determining or communicating PFT reliability to the data coordinator. All PFT values of pediatric CF patients were being entered in the CFFPR, potentially affecting CFFPR data accuracy, patient summary reports, and possibly causing undue anxiety in patients/caregivers. The aim of this project was to increase documentation of PFT reliability at this CF center from 8% to 95% by September 30, 2018. Additionally, PFT reliability processes across CF centers were assessed.

Methods: Baseline assessment of PFT reliability documentation at this center was completed with a manual review of PFT reports in the electronic medical record from 4th quarter 2017. The process was mapped with change ideas generated. A series of plan-do-study-act cycles tested a template for documenting PFT reliability. The ATS/ERS based template, embedded in the PFT report, included 5 questions: 1) Is maneuver free of artifact? 2) Is extrapolated volume within normal limits? 3) Is exhalation time within normal limits for age? 4) Are the criteria for a reproducible maneuver met? and 5) Is the test reliable? Primary caregiver's (PC's) knowledge of their child's FEV 1 percent and associated level of worry (None, A little, Some, A lot) were assessed with an electronic survey disseminated via social media for this center's PCs. Benchmark PFT reliability data from other CF centers were collected with an electronic survey sent to the RRT listserv.

Results: Baseline chart review of PFTs (n=208) at this center revealed that only 17 (8%) PFT reports documented reliability. Since implementation, the template has been used on 37 (77%) of 48 PFTs, and 5 of 37 (13.5%) have been identified as unreliable. The data collection is ongoing. In PC surveys (n=21), 19 (90%) reported knowledge of their child's FEV 1 % and 12 (57%) reported that this number affects their worry "a lot." Of 35 pediatric CF centers polled, 26 (74%) responded that they have a process to determine PFT reliability. Of 24 centers reporting process specifics, 6 (25%) use subjective assessment by PFT tech; 15 (63%) use other methods, including ATS/ERS criteria, loop assessment/exhalation time, or MD review. The 24 centers also responded that 19 (79%) document PFT reliability directly on the report, and 15 (63%) stated they do not enter unreliable PFT results in the CFFPR.

Conclusion: Project results show that documentation of PFT reliability can be improved with a QI process. PC surveys indicate that PFT values can cause worry and confirm the need for PC understanding of PFT reliability. The data collection is ongoing to assess adherence to template use, variation in reliable PFT values, change in FEV 1 percent mean, variation in reliability criteria questions, and ultimately clinical usefulness of the template. Survey of CF centers reveals that not all centers have a formal process for determining PFT reliability. The project has a potential to improve the quality of CFFPR data.

Engelun, M.; Clabaugh, J.; Dunitz, J.M. Minnesota CF Center, University of Minnesota, Minneapolis, MN, USA Introduction: Physical therapy (PT) can provide interventions specific to individuals with cystic fibrosis (CF) to augment airway clearance and improve overall quality of life. One intervention is physical activity, which has been shown to improve exercise capacity and slow the decline in lung function in individuals with CF (Dwyer TJ, et al. Curr Opin Pulm Med. 2011; 17:455-60) . The positive effects of physical activity, in addition to postural re-education, pain control and incontinence management has been reported. However, it has not been identified how many individuals with CF could benefit from skilled PT in each area. Barriers to individual participation are present, even though PT has become an important part of the CF interdisciplinary team.

Purpose: 1) To demonstrate the prevalence of adults with CF who would benefit from PT treatment for postural impairments, aerobic deconditioning, pain impairing airway clearance, and incontinence management. 2) To identify barriers that prevented PT treatment during the first 18 months of PT presence in CF clinic.

Methods: Data were retrospectively collected by two therapists from PT screens and evaluations completed in CF clinic. Each individual was briefly screened for PT needs by the same therapists from May 2017 to April 2018, which was within the first 18 months of PT presence in CF clinic. If PT was indicated, an evaluation and treatment was attempted. Categories of needs included: postural re-education, aerobic conditioning, pain control, and incontinence management. If patients refused PT services, a barrier was identified. Barriers were organized into five categories: 1) participated (PT evaluation completed), 2) none (no skilled PT needs), 3) declined (reason not stated, seeking PT elsewhere, or did not want to participate), 4) time constraint (unable to stay for appointment) and 5) financial concern (individual did not know their PT coverage/copay/out of pocket expenses).

Results: Data were collected on 180 individuals, 94 male, median age 34 years (age range 19-66). Analysis showed 141 (78.3%) of 180 individuals demonstrated a need for PT. Specifically, 67.8% had postural needs, 53.9% had aerobic needs, 22.2% had pain and 30.6% reported incontinence. In total, 85.1% of 141 individuals with needs reported a barrier to participating, including: 61.7% declined participation, 13.5% reported a time constraint and 9.9% indicated a financial concern. Likewise, 14.9% participated in evaluation and treatment.

Conclusion: This analysis demonstrates adult individuals with CF have a need for PT, most significantly in the areas of postural re-education and aerobic conditioning. A fair number of individuals also reported incontinence impairments; however, we suspect this impairment is still underreported. Notably, these data were collected within the first 18 months of new PT presence in our CF clinic. This analysis has helped identify specific barriers to individual participation with PT, and continued PT presence will likely begin to minimize these barriers. The ongoing success of providing PT in the clinic will require the continued support of the CF interdisciplinary team promoting the importance of PT involvement in the care of individuals with CF. ). The goal of this study was to determine if adherence predicted lung function decline in the same cohort 3 years later.

Methods: We followed a single-center cohort of pediatric CF patients who perform ACT with HFCWO vest. Objective adherence data from the home HFCWO device were collected over a 5-week period in 2014. Cumulative utilization time divided by the number of days in the study period provided a mean daily use value. The ratio (%) of mean-to-prescribed daily use objectively quantified adherence. Patient sociodemographic and clinical data were abstracted from the CF Foundation Patient Registry. For each patient, peak FEV 1 percent predicted (highest FEV 1 percent for the year, representative of optimal lung health) and mean FEV 1 percent predicted (average FEV 1 percent for the year) were calculated for 2014 and 2017, respectively. Bivariate and multivariate analyses were performed with simple and multiple linear regression.

Results: Complete data over the 3-year study period were available for 88 study participants. Mean patient age in 2014 was 13 (SD 4) years, and mean ACT adherence rate was 60% (SD 31). In 2014, FEV 1 percent predicted values were 100% (SD 16) for peak and 92% (SD 17) for mean. Over the 3-year study period, peak FEV 1 percent declined 6% (SD 12) and mean FEV 1 percent declined 4% (SD 12). Adherence to HFCWO ACT was not associated with absolute FEV 1 percent values in 2014 or 2017, but was correlated with less decline in peak FEV 1 percent (b= -0.9, p<0.05). After adjusting for maternal and paternal education, household income, and patient age, the relationship between ACT adherence and decline in peak FEV 1 % remained significant (b= -0.8, p<0.05). When ACT adherence was coded as low (<35% of prescribed), moderate (36-79% of prescribed), and high (≥80% of prescribed), high adherence was associated with 8% less decline in peak FEV 1 % (b= -7.9, p<0.05) 3 years later.

Conclusion: Adherence to ACT with HFCWO does not correlate with cross-sectional FEV 1 percent values at baseline or 3 years later, but it limits peak FEV 1 % decline. Compared to low-adherent patients, highly adherent patients have 8% less decline in peak FEV 1 % over a 3-year period. High adherence to ACT with HFCWO may have a protective effect for lung health in pediatric CF patients.

Arcinas, R.R. 1 ; Sathe, M.N. 2 ; Varghese, P.A. 3 ; Sharma, P. 4 1. Physical Medicine and Rehab, Children's Health Dallas, McKinney, TX, USA; 2. Pediatric Gastroentorology, UT Southwestern, Dallas, TX, USA; 3. Pulmonology, Children's Health, Dallas, TX, USA; 4 . Pediatric Pulmonology, UT Southwestern, Dallas, TX, USA Introduction: As the average age of life expectancy of people with cystic fibrosis (CF) increases, clinicians have begun to take note of decreased bone mineral density (BMD) in patients. The CF Foundation (CFF) guidelines recommend obtaining at least one DEXA scan between the ages of 8 and 18 years old to ensure proper bone growth and development. Our CF Center at Children's Health in Dallas sought to implement the DEXA scan guidelines in 2013 which consisted of a plan to foster a multidisciplinary approach. This included partnering with Physical Medicine and Rehabilitation so that a physical therapist (PT) can be involved to consult and implement exercise prescriptions for those patients with abnormal DEXA scan results. Our CF clinic obtained a 3-year grant support from the CFF for PT services which started in December 2016.

Objective: A review of implementation of outpatient (OP) PT services since December 2016 reveals the following challenges in the first year which need to be addressed: Poor education for the need of physical therapy with Texas insurance providers, tests and measures needed to attain approval for therapy services, timeliness of approval for services, compliance of families to follow through with recommendation of services. From Dec 2016 -Aug 2017: Out of the 55 patients initially evaluated, 4% of those evaluated were denied services due to poor education of insurance providers, 35% received OP services, 27% were followed at their quarterly CF clinic visits, 23% refused PT services when recommended, and 11% were unable to receive PT services due to poor procedures in scheduling them in a timely manner.

The PT collaborated with insurance companies to identify what is needed for approval of services; PT created algorithm for recommendation of OP services using CPT evaluation guidelines; created system wide procedure for initiation of services; continued education to patients regarding importance of OP PT if recommended.

Conclusion: The increase in education has helped stakeholders understand the need for OP PT services. This has improved compliance with recommendation of receiving OP PT services. Since administering the following assessments, there have been no insurance denials for services: the Bruninks-Oseretsky Test of Motor Proficiency, 2nd Edition (BOT-2) Strength and Agility Composite to test strength; 3-Minute Step Test for Aerobic Capacity; Postural Assessments and Range of Motion measurements for ribcage expansion and musculoskeletal impairments that may inhibit proper weight bearing contributing to a decreased BMD. Procedures for insurance verification, authorization and initiation of treatment in a timely manner is an ongoing process.

Bessette, H.C.; Wilson, A.C.; MacDonald, K.D. Pediatrics, Oregon Health & Science University, Portland, OR, USA Rationale: CF patients receiving inpatient care are essentially confined to their rooms due to infection control guidelines. At our institution, inpatient stays for acute pulmonary exacerbations are 10-14 days for 9-17 year olds. Hospital stays are associated with depressed and anxious mood in children and adolescents. We hypothesized: 1) an in-room exergaming program would lead to improved activity levels measured by actigraphy compared to control 2) standard measures of mood would be higher in the exergaming group compared to control.

Methods: An IRB-approved protocol was employed to recruit CF patients being admitted for inpatient care to be randomized to either daily participation for 2 hours with an in-room XBOX Kinect and continuous actigraphy or actigraphy alone. Mood surveys (PANAS) were given on day of enrollment, day 8, and day of discharge. Members of the research team checked-in with participants every few days or to provide technical support. Comparison of mean values was by t-test with α of 0.2 and β 0.8, significance of 0.05.

Results: 31 persons have enrolled with 1 withdrawal and 5 subjects' data withdrawn for lost materials, surveys, or early discharge. The mean self-reported activity entries per day was 1.7 in the active group entries versus 1.2 entries in control (p=0.004). Compliance between written activity entries and recorded actigraphy active intervals for the active group was high (85%). Recorded actigraphy active intervals for the active group and control group were not significantly different. The mood survey reports both positive affect and negative affect; mean score for negative affect at discharge was 6.5 for active group versus 8.3 in control (p=0.29). On average, the negative affect score decreased by 4.2 for active group and by 2 for the control group from enrollment to discharge (p=0.26). The change in mood during hospital stay for active group versus control group was not significantly different.

While this represents our preliminary findings, the trial is still actively enrolling.

Alaa, A. 3 ; Daniels, T.W. 6 ; Floto, R.A. 1, 2 ; van der Schaar, M. 4, 5 1. Medicine, University of Cambridge, Cambridge, United Kingdom; 2. Cambridge Centre for Lung Infection, Royal Papworth Hospital, Cambridge, United Kingdom; 3. UCLA, Los Angeles, CA, USA; 4. University of Oxford, Oxford, United Kingdom; 5. Alan Turing Institute, London, United Kingdom; 6. University Hospital Southampton, Southampton, United Kingdom Introduction: The ability to accurately predict the prognosis of individuals with cystic fibrosis (CF) would allow targeted interventions with appropriate medical treatments and optimally timed referral for lung transplantation. Existing algorithms and clinical decision aids lack sufficient precision to meaningfully influence healthcare policy or clinical management. We therefore wondered whether machine learning-based approaches might improve prognostic accuracy and potentially guide clinical decision making.

Methods: We took retrospective longitudinal data from the UK Cystic Fibrosis Registry; a database sponsored and hosted by the UK Cystic Fibrosis Trust that covers 99% of the national cohort of CF individuals. The Registry comprises clinical metadata for each individual including demographic variables, CFTR genotype, a list of disease-related variables obtained at annual clinical review including information on microbial infections, co-morbidities and complications, lung function, weight, intravenous antibiotics usage, medications, and transplantation. We used AutoPrognosis, a state-of-the-art machine learning method for prognostic scoring (1) to automatically construct a prognostic model for predicting 3-year mortality based on the variables at baseline.

Machine learning experiments were conducted using data from adult patients in 2012, the most recent cohort for which 3-year mortality data was available. 115 variables were extracted for each patient and fed into AutoPrognosis to derived an agnostic, data-driven approach for discovering risk factors.

We focused initially on death or lung transplantation within 3 years of baseline data collection as the key outcome. We compare the performance of AutoPrognosis with all existing prognostic models in terms of various diagnostic accuracy metrics that capture the models' sensitivity, specificity and predictive values. In particular, we evaluate the models' discriminative power (AUC-ROC), the informedness (Youden's J statistic), and its clinical usefulness (measured by area under precision-recall curve (AUC-PR), average precision and the F1 score).

Results: AutoPrognosis outperformed all existing models with respect to all diagnostic metrics under consideration, displaying satisfactory discriminative power (AUC-ROC of 0.89 (95% CI: 0.88-0.90) and a J statistic of 0.67 (95% CI: 0.65-0.69). More importantly, AutoPrognosis displayed an even more significant gain when it came to clinical usefulness (with AUC-PR of 0.58 (95% CI: 0.54-0.62), an average precision of 0.59 (95% CI: 0.55-0.63) and an F1 score of 0.60 (95% CI: 0.57-0.63).

Our results indicate that competitive machine learning approaches significantly improve prognostic forecasting and will support optimized referral for lung transplantation. BMI <17 kg/m 2 is a contraindication to LTx at most LTx centers. Low BMI is associated with death without LTx in CF patients with advanced lung disease (Ramos KJ, et al. Chest. 2017; 151:1320-8) . We hypothesized that underweight cystic fibrosis (CF) patients would have post-transplant survival outcomes that are acceptable and comparable to non-CF diagnoses.

Methods: Using OPTN data as of December 13, 2016 in the UNOS/ STAR file, we identified patients with CF, idiopathic fibrosis (IPF), and chronic obstructive pulmonary disease (COPD; without notation of alpha-1 antitrypsin disease) listed for LTx between May 2005 and Dec 2015. Patients were stratified by BMI (kg/m 2 ) ≥ or < 17 at the time of LTx. Analyses were restricted to COPD and IPF patients with bilateral LTx, because patients with CF only undergo bilateral LTx and survival with bilateral LTx is superior to single LTx. We calculated Kaplan-Meier estimates of median post-LTx survival by diagnosis and within the BMI strata for patients with CF. We used Cox proportional hazards (PH) modeling to estimate the hazard ratio (HR) of post-LTx mortality (censored at lost to follow-up; includes time following a repeat LTx if applicable) for patients with CF and IPF, with COPD as the reference group. Cox PH models were adjusted for mean annual LTx center volume and calendar year of LTx.

Results: Among 2,183 patients with CF, 6,451 patients with IPF, and 4,855 patients with COPD, the median BMI (IQR) at the time of LTx was 19 (18-21), 28 (25-30), and 24 (21-28), respectively. Median post-transplant survival for patients with CF and BMI <17 (n=351) was 7.0 years (95% CI 4.5, 7.9); patients with IPF (all BMIs) 6.5 years (6.1, 7.1) and patients with COPD (all BMIs) 6.5 years (6.2, 7.0). Compared to COPD, the unadjusted and adjusted HR (95% CI) for post-transplant mortality for CF patients with BMI <17 were both 1.06 (0.89, 1.25) . For IPF patients the unadjusted and adjusted HR relative to COPD were both 1.04 (0.96, 1.12). Patients with CF and BMI <17 had increased post-transplant mortality compared to non-underweight (BMI ≥17kg/m 2 ) CF patients (adjusted HR 1.23, 95% CI 1.03, 1.47, p=0.02). Patients with COPD and BMI <17 had no difference in post-LTx mortality compared to non-underweight COPD patients (HR 1.20, 95% CI 0.80, 1.79). Less than 1% of patients with IPF had BMI <17.

Conclusions: Patients with CF and BMI <17 have comparable post-LTx survival outcomes to patients with COPD or IPF who undergo bilateral LTx. For patients with CF, BMI <17 should not be an absolute contraindication to LTx.

Dellon, E.P. 1 ; Basile, M. 2 ; Hobler, M.R. 3 ; Georgiopoulos, A. 4 ; Goggin, J.L. 5 ; Chen, E. 6 ; Goss, C.H. 3 ; Hempstead, S.E. 7 ; Faro, A. 7 ; Kavalieratos, D. 8 1. University of North Carolina, Chapel Hill, NC, USA; 2. Northwell Health, Manhasset, NY, USA; 3. Univ of Washington, Seattle, WA, USA; 4. Massachusetts General Hospital, Boston, MA, USA; 5. Univ of California San Diego, La Jolla, CA, USA; 6. Rush Univ Medical Center, Chicago, IL, USA; 7. Cystic Fibrosis Foundation, Bethesda, MD, USA; 8. Univ of Pittsburgh, Pittsburgh, PA, USA Background: Individuals with CF who are considering or have undergone lung transplantation may benefit from many aspects of palliative care (PC) along the transplant continuum. Currently, no practice standards exist, and incorporation of PC into care around transplant likely varies from center to center. We aimed to better understand PC in lung transplant by surveying key stakeholders in CF care.

Methods: A mixed methods approach with closed and open ended survey questions was utilized. Patients, caregivers, and CF care team members ("providers") answered questions about PC that included specific questions about lung transplant. Thematic analysis was used to characterize responses to open ended questions that highlighted the specific beliefs and concerns of stakeholders.

Results: Of 70 patient participants, 24 (34%) recalled discussing lung transplant with their CF care team and 11 (15%) were transplant recipients; 26 of 100 caregivers (26%) recalled transplant discussions with the CF team, and 5 (5%) had a loved one who had undergone transplant. Of 350 CF care team member participants, 37 (11%) also provide transplant care. While 90% of providers rated PC as very valuable in CF, only 81% of CF and transplant providers versus 93% of non-transplant providers rated it as such (P=0.004). When asked what makes PC different in CF from other conditions, 22% of providers selected the option of transplant for advanced disease. Transplant was noted to be a barrier to PC by 30% of transplant providers versus only 16% of non-transplant providers (P=0.03). In written responses, many providers described PC as important for those who are not pursuing or not eligible for transplant to help with care planning; some felt it is also helpful for those who are waiting for transplant. Those who emphasized the importance of PC throughout the transplant process noted that changing goals of care, decision making, and symptom management are common concerns along the continuum. Some described feeling that the goals of transplant and PC are mutually exclusive, equating them with life extension and end of life, respectively. Caregivers described the importance of PC throughout the transplant process given medical and emotional vulnerability and emphasized the need for individualized care.

Conclusions: There are varying opinions about the role and importance of PC for individuals with CF who are considering or who have had a lung transplant. Further studies are needed to better understand how PC fits into lung transplant care for individuals with CF. Clarification of what PC entails and dispelling myths about the goals of transplant and PC being in conflict can be addressed through PC education for patients, caregivers, and providers. Background: We previously showed that bone mass was reduced and bone quality severely impaired in rib specimens from cystic fibrosis (CF) patients collected at the time of lung transplant (LTx). To prevent further deterioration of bone due to the immunosuppressive therapy, osteoporosis medications are initiated soon after LTx in most patients. However, very little is known about long-term bone mineral density (BMD) changes in CF patients after LTx.

Objectives: The study objectives were to: 1) examine femoral neck (FN) and lumbar spine (LS) BMD changes in CF patients up to 10 years post-LTx and 2) assess predictors of FN and LS BMD changes at 1 and 5 years post-LTx.

Methods: This is a retrospective cohort study of CF patients who underwent LTx between January 1, 2000 and December 31, 2015 and had FN and LS BMD measured during the waiting period and at least once post-LTx. Post-transplant timepoints ranged from 6 months to 10 years. Annualized percentage BMD changes were calculated to take into account the varying time intervals between BMD measurements. Potential predictors of post-LTx BMD percentage changes at 1 and 5 years post-LTx were assessed using multivariate linear regression adjusted for potential confounders.

Results: The study cohort consisted of 131 CF patients (46% males) with a median age at transplant of 28 years (IQR 24-35). Prior to transplant, 24% of patients had a normal bone status while 50% had osteopenia and 27% osteoporosis. Median Z-scores for LS and FN BMD were -1.4 (IQR -2.1 to -0.7) and -1.1 (IQR -1.8 to -0.3) respectively. While 20% of patients were on bisphosphonate therapy in the three months preceding LTx (average dose: 7.8±2.7 mg/d), the proportion and doses increased substantially in the post-LTx period from 84% of patients at 6 months (15±47 mg/d) to 97% at 10 years (19±53 mg/d). LS and FN Z-scores kept declining during the first year post-LTx (median LS and FN BMD Z-scores: -1.8 [IQR -2.3 to -0.7] and -1.5 [IQR -2.3 to -0.8]) but increased thereafter to reach -0.5 for both sites at 10 years. The few patients who did not receive bisphosphonates experienced a greater reduction in BMD at 1 year post-LTx (LS and FN BMD Z-scores: -2.3 [IQR -2.8 to -1.6] and -2.1 [IQR -3.0 to -0.7]). The annual percentage changes in BMD since pre-LTx differed significantly between skeletal sites and ranged from -2.0 and -4.8% at 6 months post-LTx (p=0.0005) to 0.0 and -0.5% at 5 years post-LTx (p=0.003) for LS and FN respectively. Multivariate analysis revealed that weight changes at 5 years were positively associated with annual percentage changes in both LS and FN BMD at 5 years whereas this relationship was not present at 1 year. No significant association was detected between post-LTx BMD changes and pre-LTx BMD, diabetes status and immunosuppressant and corticosteroid doses.

Conclusion: CF recipients experience on average a decline in BMD over the first year post-LTx that is greater at the FN than LS. However, this effect is transient as recuperation and even improvement are observed over the long term. Introduction: Information on the impact of lung transplant (LT) on patient-centered outcomes for patients with cystic fibrosis (CF) is limited; even less well described are the determinants of this effect. We studied the impact of LT on physical disability and health-related quality of life (HRQL) in patients undergoing LT for cystic fibrosis. We also studied the effects that changes in lung function and frailty had on disability and HRQL.

Methods: In a single-center prospective cohort study from 2010-2017, assessments of frailty, disability, and HRQL were performed before and at 3 and 6 months after LT. HRQL was assessed by the Medical Outcomes Study Short Form Physical Component Summary scale (SF12-PCS), SF12 Mental Component Score (SF12-MCS), the respiratory-specific Airway Questionnaire 20-Revised (AQ20R), and the Euroqol 5D (EQ5D). Physical disability was assessed using the Lung Transplant Value Life Activities Scale (LT-VLA). Physical frailty was assessed using the Short Physical Performance Battery (SPPB) (see Table footer for ranges and minimally clinically important differences for HRQL, disability, and frailty measures). Changes in HRQL and disability before and 6 months after transplant were compared by paired t-tests. We tested the association of changes in SPPB and FEV1 (scaled to 200 mL) on HRQL and disability by unadjusted linear mixed effects models as well as models adjusting for SPPB, FEV1, and gender.

Results: Twenty-three patients with mean age 31 (±7.87) (52% female) underwent LT. Subjects reported improvements in HRQL and less disability ranging from 1.5-to 6-fold the MCID in each metric used (Table, part 1 ). In general, improvements in frailty and FEV1 were independently associated with improved HRQL and less disability (Table, part 2). In adjusted models, for example, each 1-point improvement in SPPB or 200mL improvement in FEV1 was associated with a reduction in disability by -0.15 points (95%CI -0.20 to -0.09; MCID=0.3) and -0.07 points (95%CI -0.09 to -0.05), respectively.

Conclusion: Lung transplant affords patients with CF large and meaningful improvements in HRQL and lessened disability. Notably, improvements in physical frailty may be as important a determinant of improved disability and HRQL as allograft function. Methods: We performed a retrospective cohort study utilizing the lung transplantation database at Duke Hospital and United Network for Organ Sharing (UNOS). The study included those over age 18 who underwent lung transplantation at Duke Hospital from May 1, 2005 to April 30, 2015 for indication of CF. Analyses excluded those who had multi-organ or repeat lung transplantation. Weight gain was assessed by change in body mass index (BMI) from initial pre-transplant evaluation to time of transplant surgery. Cox proportional hazards models were used to examine relationships with mortality and chronic lung allograft dysfunction (CLAD).

Results: The cohort included a total of 202 lung transplant recipients. The median BMI at time of initial evaluation was 19.7 kg/m 2 (SD = 2.5) and 19.5 kg/m 2 (SD = 2.4) at time of transplant surgery. A subset of participants (43%) had surgical feeding tubes prior to transplant surgery. Examination of BMI changes revealed that 33% gained, 28% maintained, and 39% lost weight prior to surgery. Analyses of weight gain and mortality demonstrated a trend towards protective association, albeit nonsignificant (HR = 0.82 [0.63, 1.06], P = .123). Lower baseline BMI was the strongest predictor of weight gain (B = -0.28, P = .001), although the use of a feeding tube (B = -0.29, P = .029) and 6-minute walk distance (6MWD) were also related (B = 0.26, P = .047). Greater weight gain trended towards longer CLAD-free survival (HR = 0.77 [0.59, 1.01], P = .058) (Figure) . Albumin was significantly associated with mortality after controlling for weight change (HR = 0.55 [0.31, 1.00], P = .049).

Conclusion: Weight gain before lung transplant surgery may be associated with improved survival, but is difficult to achieve for many patients with advanced CF. Acknowledgments: Supported by the NIH 5T32HL007538-32. Objective: To describe how pain is experienced and managed in lung transplant recipients with CF.

Methods: A cross-sectional online survey was developed with 4 domains: demographics, pain characteristics (types, diagnoses, impact on mood, and impact on function), pain reporting, and management strategies. Individuals with CF ages 14 and older status-post-lung transplantation completed online surveys which were distributed nationally via the CF Foundation Community Voice listserv, social media sites, and CF center patient listservs.

Data: The survey was completed by 40 individuals; mean age was 35.8 years (range=19-64), 75% female, 45% with private insurance and 82.5% were more than one year post-transplant. The majority of respondents endorsed experiencing pain related to their CF disease (83%), 59% reported having their pain actively managed by a healthcare provider, and 31% reported having an official chronic pain diagnosis. The most common types of pain reported were nonsinus related headaches (70%), lower back pain (65%), acid reflux (65%), sinus pain (65%), abdominal pain, (60%) and joint pain (58%). Participants reported experiencing an average of six different types of pain (range 0-12). A total of 37% reported that pain often or always affected daily life, 24% reported pain affected mood a lot or a great deal, and 35% reported experiencing pain daily. Clinically, respondents reported discussing pain with their CF (56%) and primary care (30%) teams more than their transplant team (17%), while in the social setting pain was mostly discussed with significant others (35%). Acetaminophen was the most commonly used over-the counter medication (92%). Twenty-six (70%) reported receiving prescription pain medications and 30% reported receiving an opioid; 73% of those receiving prescriptions reported receiving more than one type of prescription pain medication. Alcohol (8%) was the most commonly reported nonprescribed substance used for pain relief and 5% reported using nonprescribed oxycodone Hot packs (50%) and exercise (36%) were the most common nonpharmacologic therapies used for pain. Procedural pain relief was used by some, including steroid injections (10%) and nerve blocks (10%). Conclusions: Pain attributable to CF is common in those who have undergone lung transplantation, with a prevalence similar to that experienced by the general CF population. Those with pain desire management help from their CF care providers, and many receive multiple prescriptions for pain medications including opioids. There is a need to further explore pain and the use of both pharmacologic and nonpharmacologic therapies in this population to improve health outcomes and quality of life. Method: Adult lung transplant recipients at a large medical center between January 1, 2010 and April 30, 2018 were retrospectively evaluated for DGE using 4-hour solid gastric emptying scintigraphy (GES) before and after lung transplantation. Not all patients who received lung transplantation completed this test. DGE was defined as greater than 70% of contents remaining in the stomach at 90 minutes or greater than 10% of contents remaining at 240 minutes.

Results: During this time frame, 307 patients received a lung transplant with 147 of these patients completing GES. Of the 39 transplant recipients with CF, 7 completed GES prior to transplant. All 7 patients had normal gastric emptying. Overall, 17 patients completed GES after transplant with DGE occurring in 10 (59%) of these patients. Patients were tested a mean of 121 days after transplant. A second GES was completed a mean of 415 days after transplant in 5 patients. Significant improvement in gastric emptying occurred 80% of the time. More lung transplant recipients with CF exhibited delayed gastric emptying than in other lung transplant recipients.

Conclusions: Delayed gastric emptying is very common after lung transplantation in patients with CF, affecting a greater percentage than in non-CF lung transplant recipients. There is a significant improvement in gastric emptying between 12 and 18 months after transplantation. Future analysis will be required to assess the effect of delayed gastric emptying in CF on post-transplant outcomes. The International Depression Epidemiological Study (TIDES) study showed that rates of anxiety and depression are higher among patients with CF and their caregivers than the general population. A previous analysis showed that CF patients with a positive depression screen had increased 5-year mortality. We sought to investigate intermediary events and changes that may be driving this association.

Objective: To determine whether a positive screen for depression in patients with CF is associated with longitudinal decline in FEV 1 % predicted and the rate of pulmonary exacerbations (PEx).

The study sample was composed of US participants in TIDES who completed the Hospital Anxiety and Depression Scale (HADS) and/or Center for Epidemiologic Studies Depression Scale (CES-D) during a stable visit between 2006 and 2010, were 12 years of age and older, and are followed in the CF Foundation Patient Registry. Patients with a positive screen for depression on either scale were compared to those with a negative screen on both using t-tests, repeated measures ANOVA, and Poisson Regression.

Results: The analysis included 1,037 patients with CF who met inclusion criteria. Of patients with a positive depression screen, 37.4% had no PEx, 29.4% had one PEx, and 33.2% had two or more PEx in the year of screening (baseline year), compared with 49.0%, 26.0%, and 25.0%, respectively among those with a negative screen (p=0.003). Of this cohort, 745 (70%) had yearly FEV 1 and PEx data and did not receive a transplant over 5 years of follow-up. Poisson regression modeling found a significant interaction between depression screen and baseline PEx (p=0.008). The difference in mean PEx over 5 years between those with positive vs negative screen with 0 PEx during the baseline year was 1.0 PEx; in those with 1 baseline PEx the difference was 2.5 PEx; and for those with 2+ baseline PEx it was 0.4 PEx. In the baseline year, the mean FEV 1 % predicted for patients with a positive depression screen was 65.0% (SD 23.8) vs 68.6% (SD 22.9) for those with a negative screen (p = 0.03). At one year, patients with a positive screen had increased variability in FEV 1 % predicted, with a mean relative change of 8.5 (SD 8.9) compared with 7.1 (SD 6.9) in those who screened negative (p=0.01). Repeated measures ANOVA found no significant association between depression and lung function change over 5 years for the group overall (p = 0.47); no differences were found when stratified by age group, although a nonsignificant trend towards greater decline was seen among depressed patients with more advanced lung disease at baseline.

Conclusions: CF patients with a positive screen for depression are more likely to experience more PEx on follow-up. The failure to find a significant effect in change in FEV 1 % predicted may be due to survivor bias (ie, those who died with accelerated rate of decline were excluded from analysis). Our findings provide some insight into the previously reported association of depression with greater 5-year mortality in CF. Methods: Using the CF Foundation Patient Registry data (CFFPR) from 2010-2016, we compared high (>75% of encounters) vs. low/no (<25% of encounters) AZM use in those reporting use of inhaled tobramycin or aztreonam (>50% of clinic encounters for ≥ 1 year). Rate of FEV 1 pp decline and PEx risk were estimated using linear mixed effects and Poisson regression respectively for comparison by AZM use overall and within each inhaled antibiotic cohort. Propensity score (PS) matching was applied to identify smaller groups closely balanced in 14 baseline parameters with potential to affect AZM use (FEV 1 pp, age, mutation class, PEx in prior year, comorbidities, medications, etc.).

Dataset: 10,002 people in the CFFPR were identified as using inhaled tobramycin in conjunction with high (n=5972, 59.7%) or low/no use of AZM (n=4029, 40.3%). High AZM users were older (mean age 21.1 vs. 18.1 years) and had lower FEV 1 pp vs low/no AZM (76.1% vs. 83.6%). PS matching generated equally balanced groups consisting of 2,440 people per group. 3,109 patients met the definition of chronic inhaled aztreonam with either high AZM (n=2105) or low/no AZM (n=1004) usage. High AZM users were older (26.7 vs 24.8 years) and had lower FEV 1 pp (69.1% vs. 75.5%). PS matching generated 583 people/group.

Results: Overall, high AZM use did not associate with a lower rate of FEV 1 pp decline or lesser risk of acute PEx. Among the tobramycin cohort, high AZM users had a slightly greater rate of decline in FEV 1 pp than low/ no AZM users in both the overall (p=0.006) and the PS matched cohorts (-1.63%/year vs. -1.52%/year, p=0.04). Among the aztreonam cohort, high AZM users had a similar but slightly lower rate of FEV 1 pp decline than low AZM users in both the overall (p=0.31) and PS matched cohorts (-0.88%/year vs. -0.95%/year, p=0.16). More notably, high AZM use did not significantly reduce the risk of PEx in either population: adjusted relative risk (aRR) of PEx with inhaled tobramycin and high AZM use (vs. low/ no use) was 1.07 (p=0.04) in the total cohort and 0.992 (p=0.86) in the PS matched cohort. With inhaled aztreonam, aRR with high AZM use was 1.00 (p=0.96) in the total cohort and 0.95 (p=0.44) in the PS matched cohort.

Conclusions: Our analyses of contemporary CF registry data suggest that ongoing chronic AZM use may not attenuate the rate of FEV 1 pp decline or acute PEx risk in those using inhaled antipseudomonal antibiotics. The duration of prior AZM use in this population is unknown but likely of relevance. An association between AZM use and greater rate of decline in FEV 1 pp was observed in those using inhaled tobramycin but not inhaled aztreonam, which is consistent with a growing body of evidence indicating that AZM can adversely interact with tobramycin.

McKone, E.F. 1, 2 ; Kirwan, L. 3 ; Zolin, A. 4 ; Jackson, A. 3 1. St. Vincent's University Hospital, Dublin, Ireland; 2. University College Dublin, Dublin, Ireland; 3. Cystic Fibrosis Registry of Ireland, Dublin, Ireland; 4. University of Milan, Milan, Italy Introduction: Dornase alfa is a mucolytic treatment for CF patients that leads to improved lung function and reduced exacerbation frequency. The aim of the study was to examine the effect of dornase alfa on the rate of decline in lung function in CF patients that have been enrolled in the European Cystic Fibrosis Society Patient Registry (ECFSPR).

Methods: Data on all CF patients aged 6 years and older enrolled in the ECFSPR from 2008 to 2014 were eligible for inclusion in this analysis. There were 10,742 patients that started dornase alfa treatment during the study period. Treatment was defined as inhaled dornase alfa for >3 months in any calendar year. To examine the effect of dornase alfa on the rate of decline in lung function, we included patients with at least 1 year's data available before and 1 year after commencement of treatment. A sensitivity analysis included patients with at least 2 year's data before and after commencement of treatment. The primary outcome measure was longitudinal changes in forced expiratory volume in one second as percent predicted (FEV 1 %pred). FEV 1 %pred was calculated using Global Lung Function Initiative equations.

Results: The study population (n=6,065) was 52% male with mean age 18.9±0.16 years and mean FEV 1 %pred of 76±0.3%pred at time of starting dornase alfa. The effect of dornase alfa on the rate of decline in lung function depended on age group (p=0.008). The largest effect was observed in the <12 years group, with an absolute difference in FEV 1 %pred decline of 0.37% per annum between the treatment periods (before -1.39±0.15%, after -1.02±0.10%, p=0.002). In the 12 to <18 years group an absolute difference of 0.31% per annum was observed (before -1.46±0.20%, after -1.15±0.13%, p=0.041). There was no significant difference in the ≥18 years group (before -0.83±0.13%, after -0.69±%, p=0.163). The sensitivity analysis (n=2,024) produced similar results.

Conclusions: Dornase alfa is an effective therapy that reduces lung function decline over time. These effects are most pronounced in children and adolescents with CF.

Acknowledgments: Presented by the authors on behalf of the European Cystic Fibrosis Patient Registry and supported by a funded grant from Roche. Introduction: Forced expiratory volume in 1 second (FEV 1 ) is routinely used to guide clinical decisions for patients with cystic fibrosis (CF). Although its diagnostic and evaluative use is extensive, the lack of reproducibility standards (within-subject biological variability of serial measurements) has long been described as a major limitation.

Objective: To define the normal range of FEV 1 reproducibility in healthy individuals in order to understand what a clinically meaningful change is for CF patients.

Methods: Spirometry measurements from 6 longitudinal studies in healthy individuals were combined. Z-scores for FEV 1 were calculated using the Global Lung Function Initiative (GLI) reference equations. A conditional change score (Cole TJ. Arch Dis Child. 1995; 73:8-16) based on the correlation between successive FEV 1 measurements in healthy individuals, and correction for regression to the mean was derived. The change score was validation in a CF population using FEV 1 measures from clinically stable CF patient visits from the Toronto CF Database (TCF).

Results: In total, 43,937 FEV 1 measurements from 8535 healthy individuals ranging in age from 6 to 80 years old were used to define the reproducibility of FEV 1 . The between-subject variability of FEV 1 was much greater than the within-subject variability between successive measurements, confirming that comparison of individual results to a reference range cannot be used to accurately track individual disease progress. The correlation between repeated measurements was independent of age, but decreased with time (r = 0.874 -0.034*time(years)), such that the correlation for measurements 3 months apart is 0.87. Adjusting for the initial lung function a change score can be estimated (zFEV 1 (current) -r*zFEV 1 (previous))/ (√(1-r 2 ). A change score within ±1.96 is considered within the normal limits of reproducibility, whereas values outside this range reflect either a meaningful deterioration or improvement. For example, a drop from -1.5 Z-scores (~82% predicted) to -2.3 Z-scores (~71% predicted) in 3 months is calculated to be outside the limits of normal variability, and therefore a meaningful deterioration. Contrarily, an increase in FEV 1 from -1.5 Z-scores (~82% predicted) to -0.5 Z-scores (~93% predicted) in 3 months is within the expected range of normal variability, and would not be considered meaningful. When applied to 48,711 stable FEV 1 measurements from 1098 CF patients (ranging in age from 5-74 years old), 82.4% of stable visits were within the range of variability in health, whereas 15.2% were identified as having a meaningful drop and 3.7% were increasing significantly. Notably, only 46.1% of stable visits were within a 10% relative change in FEV 1 percent predicted. Further analysis will investigate application of the change score for pulmonary exacerbations and monitoring treatment effects.

Conclusion: These results suggest it may be possible to more accurately identify the normal range of FEV 1 reproducibility in CF than the currently used limits of a 10% change to improve tracking of lung function changes over time.

Acknowledgment: This work was funded by a Cystic Fibrosis Research Innovation Award from Vertex Pharmaceuticals. Introduction: The median age of survival (MSA) for Canadians living with cystic fibrosis (CF) has been a useful estimate to inform people born with CF today of their expected survival. However, it is of limited use to those patients currently living with CF, many of whom have reached or even surpassed the median survival age of a given CF population. Further, several countries have reported lower median survival age in females with CF. It is unclear how the survival age varies dependent on reaching a given age and whether the survival sex gap persists in individuals who have lived past a certain age. The objectives of our study were to (1) calculate conditional survival estimates for Canadians living with CF over time, (2) examine whether there is a sex-based conditional survival gap at varying ages, and (3) to compare the factors that influence risk of death before and after 30 years of age.

Methods: We conducted a retrospective cohort analysis using data from the Canadian CF Registry. We compared 3 time periods: 1992-1996, 2002-2006 and 2012-2016 . Within each period, median age of survival was calculated using a rolling 5-year window. Conditional survival estimates by time period and sex were calculated using the Kaplan-Meier method.

Multivariable models were created using Cox proportional hazards models to determine which predictors were associated with an increased risk of death before or after age 30. The first included only static covariates (sex, race, age of diagnosis) while the second also included time-varying predictors (BMI, FEV1, microbiology, exacerbations, CF-related diabetes). Subjects under age 6 and post-transplant clinical measurements were excluded in the 2nd model.

Results: Between 2012-2016, 4,881 patients (53.9% male) were included. Median survival age (MSA) increased from 52.6 (95% CI 50. 4-56.8) years to 61.0, 64.3, and 68.9 years when conditioning on living to age 35, 40 and 45 years, respectively. Although females consistently had a lower MSA for each conditional analysis compared to males, the differences were not statistically significant. The MSA increased 21 years between the 3 time periods; however, the conditional survival MSA given that a patient lives to 35 years of age has remained relatively stable over time.

Females (HR 1.59 (95% CI: 1.08-2.38), p=0.02) and non-Caucasian race (HR 2.70 (95% CI: 1.3-5.56), p<0.01), were associated with an increased risk of death in the under 30 years of age cohort (n=2972) and age of diagnosis < 2 years (HR 1.6 (95% CI: 0.98-2.62, p=0.06) was borderline-significant. However, these factors were no longer statistically significant in the risk of death over age 30 group (n=1614). The model including time-varying covariates is in the process of being analyzed.

Conclusions: MSA is higher for those who reach age 35 and older compared to the overall MSA that is reported in Canada. Although the MSA did not differ between males and females, females were at a higher risk of death before the age of 30 years compared to males. Analyses of timevarying models are ongoing and will be completed for NACFC.

Cosgriff, R. 1 ; Carr, S.B. 2,3 ; Lee, A. 1 ; Yip, M. 1 ; Gunn, E. 1 ; Charman, S. 1 1. Impact, Cystic Fibrosis Trust, London, United Kingdom; 2. Royal Brompton Hospital, London, United Kingdom; 3. Imperial College London, London, United Kingdom Objective: As genotype specific cystic fibrosis (CF) therapies are increasingly available, complete, accurate genotype data is vital. This work has progressed the utility of genotype data in the UK CF Registry by improving the data entry and cleaning, and harmonizing with other CF databases.

Background: The UK CF Registry is a database of the 99% of people with CF who consent to their data being collected. Genetic data remain associated with each patient longitudinally. In 2016, 9544 (98.4%) people with CF on the Registry had a value recorded for both alleles.

Methods:Software development. In 2016 the UK CF Registry moved to a new software platform, which includes a "lookup table" of the nucleotide, protein, and legacy name of every mutation in the CFTR2 (cftr2.org) database. It also contains alternative names for the commonest genotypes (e.g. DF508 instead of F508del). A Registry user can "start typing" part of any variation of a genotype name, triggering the list to auto-refine remaining options. By removing the need for users to scroll through a long list of genotype names, less time is spent on data entry, and fewer common genotypes are entered as "other" free text values.

CFTR2 has been agreed to as the Gold Standard by the CF Registry Global Harmonization group. Use of standardised nomenclature and coding will enable easier comparison of genetic data across different countries. The CFTR2 team updates participants when new genotypes have been coded and categorised. This process is built into the UK CF Registry development release cycles. In summer 2018 we increased the list of pre-defined genotypes from 220 to 382.

Data cleaning. Users can enter genotypes other than those present in the CFTR2 genotype look up table into an "other" field. Due to the historic nature of most demographic data in the Registry, there are still mutations incorrectly categorised as "other." For this reason, mutation data are cleaned after data lock each year.

Genotype cleaning searches the "other" field for identifiable information attributed to known mutations. For example, a mutation in the "other" field named "F508Del" would be flagged for correction. Truly unknown mutations are thereby identified by a process of elimination. Data cleaning protocols standardise responses to facilitate accurate mining of legacy names. Standardisation is an essential part of data mining of human entered data, where syntax variations can result in missed cases. The expansion of the software to include all genotypes will ensure the "other" field functions solely as a repository for truly unknown mutations.

Results: In the last five years, the completeness of genetic data has increased from 96% to 98%. In cleaning 2017 genotype data, 92 (6%) "Other" entries were converted to "unknown," 18 (1%) to F508del, and 74 (5%) into predefined mutations.

Conclusions: The UK CF Registry is increasingly relied upon to provide mutation information to the CF community, for clinical trials feasibility, health technology appraisal, and the commissioning of CF services.

Quality improvement of genetic data is an ongoing process, supported by advances in research, education of those involved in entering and analysing UK CF Registry data, and development of the software.

Munck, A. 1 ; Bourmaud, A. 2 ; Pracros, J. 3 ; Bellon, G. 4 ; Picq, P. 2 ; Farrell, P. 5 1. CF Center, Hôpital Robert Debré, AP-HP, Univ. Paris7, Paris, France; 2. INSERM, CIC-EC1426, UMR1123, Hôpital Robert Debré, AP-HP, Univ. Paris7, Paris, France; 3. Radiology, Hospices Civils de Lyon, Univ. of Lyon, Lyon, France; 4. CF Center, Hospices Civils de Lyon, Univ. of Lyon, Lyon, France; 5. Dept.of Pediatrics, Univ. 

Objective: To characterize the longitudinal phenotypic expression of a positive newborn-screened cohort with an equivocal CF (eCF) diagnosis, reassign labeling if applicable and better define its prognosis.

Methods: A multicenter cohort with an eCF diagnosis (n=63) was locally matched with infants diagnosed definitely with CF. Both cohorts were prospectively compared on baseline characteristics and cumulative data. At a planned endpoint of 6-7 years, a standardized evaluation assessed the pulmonary phenotype (primary criteria: CT scan Brody score plus chest X ray (Wisconsin (WCXR) and Brasfield (BCXR)) scores as rated by centralized reading), FEV1, pathogens and anthropometry, fecal elastase1, treatments (T) and sweat chloride (SC) in the eCF cohort. Subsequently, within the eCF cohort, two substudies based on updated CF-causing mutations and final SC compared those reassigned to a CF diagnosis (final CF diagnosis (FCF)) or not (final eCF diagnosis (FeCF)) and those carrying ≥1 R117H or not. To identify independent factors associated with Brody scores we used general linear regression models (GLM). Variables were introduced in the model based on clinical judgment and adjusted for sex.

Results: Compared to infants with CF, the eCF cohort had lower immunoreactive trypsinogen* and SC values*, delayed initial visit at CF clinic* with less symptoms*, no pancreatic insufficiency and better nutritional status*. With similar duration of follow-up, they had fewer visits*, hospitalizations*, less P. aeruginosa* and MSSA (p=0.004), CF co-morbidities* and T burden* compared to CF. At the endpoint the eCF cohort had a milder pulmonary phenotype based on Brody (0.0 [0.0; 2.0] vs 13 [2.0; 31.0])*, WCXR* and BCXR* scores, FEV1 (p=0.005), less symptoms and a better nutritional status. At the endpoint, 28/63 (44%) of the eCF cohort fulfilled diagnostic criteria for CF. Compared with the FeCF cohort (n=35), the FCF (n=28) had a delayed initial visit (p=0.01) with a lower weight z-score (p=0.02) and fewer had an R117H mutation (p=0.002); we found no differences in cumulative, endpoint data but more frequently MSSA (p=0.04) with fewer cough symptoms (p=0.01). Compared to the no R117H cohort (n=36), the R117TH (all 7T/9T) cohort (n=27) had an earlier initial visit (p=0.02), a lower SC value (p=0.04). There were no differences in cumulative, endpoint data, but a lower proportion of final SC ≥60 mmol/L (p=0.05). The GLM on the global cohort (n=126) found a positive association between Brody scores and IRT*, SC values*, genotype (p=0.0006), WCXR*, FEV1 (p=0.0003), P. aeruginosa*.

Conclusion: The matched eCF and CF cohorts demonstrated significant differences in outcomes. By a mean age of 7.5 years, 44% of the eCF cohort had been reassigned to a CF diagnosis. Our data demonstrate the value of following children with eCF in CF clinics and the need of further studies with longer term monitoring and larger cohorts. *p<0.0001 Acknowledgements: Funded by VLM, SFM, AFDPHE.

Terlizzi, V. 1 ; Mergni, G. 1 ; Buzzetti, R. 2 ; Centrone, C. 3 ; Zavataro, L. 1 ; Braggion, C. 1 1. Pediatric Dept., Cystic Fibrosis Centre, Meyer Children's Hospital, Florence, Italy, Florence, Italy; 2. Epidemiologist, Bergamo, Italy, Bergamo, Italy; 3. Diagnostic Genetics Unit, Careggi University Hospital, Florence, Italy Background: Increased implementation of cystic fibrosis (CF) newborn screening (NBS) has led to identification of infants with a positive NBS test but inconclusive diagnostic testing, classified as "CF transmembrane conductance regulator-related metabolic syndrome" (CRMS) in the North American nomenclature and "CF screen positive, inconclusive diagnosis" (CFSPID) in Europe (Ren CL, et al. J Pediatr. 2017 ;181S:S45-S51.e1). We retrospectively evaluated the prevalence and clinical outcome of CFSPID patients by two NBS programs in the period 2011-2016 at CF Centre of Florence, Italy.

Methods: We retrospectively evaluated CFSPID and CF patients diagnosed by CF NBS in the years 2011-2016 and followed until 31.12.2017. In this period, a pilot project was aimed to assess the diagnostic impact of DNA analysis on the NBS algorithm (immunoreactive trypsinogen (IRT)meconium lactase -IRT2). IRT was considered elevated for values above the 99th centile laboratory cut-off. The CFSPID definition was according to Munck A, et al. (J Cyst Fibros. 2015; 14:706-13) . All CFSPID patients had a repeat sweat chloride (SC) test every 6 months and an extended CFTR gene analysis (detection rate 98%). During follow-up we reclassified children as: CF diagnosis, healthy carrier or healthy in presence respectively of 2 consecutive pathological SCs, or 2 consecutive normal SCs for age and 1 or 0 CF-causing mutation. We kept the CFSPID definition when SC was persistently in borderline range or in the presence of 2 CFTR mutations, at least 1 of which had varying clinical consequence (https://www.cftr2.org/).

Results: Of 179,684 babies screened from January 2011 to December 2016, 1520 (0.8%) screened IRT-positive at day 3. Infants called to perform SC test were 359 by NBS algorithm and 181 by DNA analysis. We identified 32 CF diagnosis and 50 CFSPID (CF:CFSPID ratio 0.64:1). One of 179,602 cases was false negative by NBS and diagnosed after an episode of dehydration with hypochloremic metabolic alkalosis at age 10 months. 20/50 (40%) CFSPID cases were diagnosed only by the IRT-DNA algorithm, 13/50 (26%) only by IRT-meconium lactase-IRT2, while both protocols identified the remaining 17 cases (34%). 37/50 CFSPID patients had a conclusive diagnosis on 31.12.2017: 5 (10%) CF, 17 (34%) healthy and 15 (30%) healthy carrier; 13/50 (26%) cases were asymptomatic with persistent borderline SC and followed as CFSPID (CF:CFSPID ratio 2.8:1). CFTR genetic analysis impacted on sensitivity and positive predictive value. Evaluating CFSPID babies with at least 2 consecutive borderline SCs at diagnosis, as required to classify an infant as CRMS (Borowitz D, et al. J Pediatr. 2009 ;155:S106-16), we had 23/50 cases (which all babies diagnosed as CF, 21.7%) with a ratio CF:CFSPID of 1.4:1.

Conclusions: At the end of the six-year period of our experience we had a CF:CFSPID ratio of 2.8 CF for every 1 case of CFSPID and 10% of CFSPID infants progressed to a CF diagnosis. It seems more reasonable to have at least 2 chloride values in the borderline range before defining a baby as CFSPID. Introduction: Following the introduction of newborn screening for CF in 2007, we implemented a pathway to inform, diagnose and educate parents whose child had a positive screening result. This pathway was developed to incorporate the UK national standards for delivering results within 5 working days. However, there is no universally agreed approach to how the information is delivered to families when the CF centre receives the result indicating a sweat test is required. We have followed the same pathway whereby the CF Nurse Specialist (CNS) delivers the presumed positive result face-to-face alongside the family's Health Visitor in the family home. Once diagnosis is confirmed the next day, we invite the family to attend a 2-day education programme facilitated by the CNS incorporating all members of the multidisciplinary team. It is known that initial diagnosis and education can influence parental adjustment and commitment to the CF team and care process. Although anecdotally we have had positive feedback from families, we have not formally reviewed the process. We wished to evaluate parental experience.

Methods: A qualitative online questionnaire was developed with parental input, consisting of 83 questions. Parents had the option of leaving free text/ comments on their experiences. It was anticipated it would take 20 minutes to complete. It was sent to 101 families who had gone through the entire NBS process at the Royal Brompton between 2007 and 2016. They had to have a positive CF or CFSPID result.

Results: Response rate was 47%. Important findings were that 95% of families agreed that the initial screen positive result should not be given over the telephone and highlighted the importance of face-to-face communication. Ideally the CNS should visit 2-3 hours after the initial phone call from the Health Visitor as this wait is recounted as one of the most stressful parts of the process. The recommendation to have both parents present was supported by 100% of respondents. Parents felt it was vital that the initial screening result was given by an experienced CNS who was an expert in the condition. This allowed them to be prepared for meeting the Consultant and having the sweat test the following day. There was no consensus as to whether genetic results (if it was diagnostic for CF) should be disclosed at the initial home visit. The 2-day education programme was highly valued with 77% preferring 2 days vs 1 day or split days. It was found that parents still wanted more information and practical guidance on how to administer medications started at education.

We have modified the process by ensuring families are visited by the CNS 2-3 hours following the initial phone call wherever possible. The Pharmacy team now spend more time showing parents how to reconstitute medications and the use of oral syringes. The CNS team are working on more written information to provide following education. Results confirmed that the process puts the families first and is highly valued. We will continue to give results face-to-face within the family home and offer the initial education via our thorough 2-day education programme.

Belkin, R.A.; Matusov, Y.; Resuello, D.; Sager, J.S.; Wright, R.S.; Li, J. Santa Barbara Cottage Hospital, Santa Barbara, CA, USA Introduction: CF is underdiagnosed in adults. Adult-diagnosed CF and "CF not resolved" (CFNR) more often presents with a milder presentation and is diagnostically more difficult. There is potential benefit from screening/diagnosis (institution of CF-specific therapies, genetic counseling for family members, and sense of relief for patients without prior diagnosis; Clin Chest Med. 2016;37:47-57). Which adults should be screened for CF remains unclear. We sought to identify clinical phenotypes associated with adult diagnosis of CF/CFNR to help establish which group of patients should be targeted for CF screening in adults.

Methods: This was a single-center, retrospective case control study conducted at Santa Barbara Cottage Hospital CF Program, analyzing the records of 78 adult patients from 2005-2018 who were referred for CF workup (sweat chloride and/or genetic testing) based on clinical suspicion for CF. Clinical characteristics were collected, including signs/symptoms at diagnosis, physical exam findings, antecedent history, imaging, microbiologic characteristics of sputum cultures, and presence of comorbidities. Subsequently, those patients who met diagnostic criteria (J Pediatr. 2017;181S:S4-15) for CF/CFNR were compared with those who did not (controls), both as a combined CF/CFNR group and then separately as CF and CFNR groups (via Fisher's exact test for binary variables; two-sample t-test and Kruskal-Wallis rank sum for continuous variables). Finally, CF and CFNR groups underwent pairwise comparison for significance of differentiating factors.

Results: Of 78 patients evaluated, 16 met the diagnostic criteria for CF, 32 for CFNR, 30 were non-CF/controls. Patients were predominantly women (n=55) with a median age of 60 (range 44-71) years in all groups. 12 patients had sweat chloride ≥ 60 mmol/L; 27 had sweat chloride 30-59 mmol/L. The most common identified mutation among CF/CFNR patients was F508del (n=15), although 17 patients did not have an identified mutation. The presence of pancreatic insufficiency was the most prominent clinical feature that differentiated CF patients from controls (p=0.04; CF vs CFNR p=0.004, CF vs control p<0.001), but not CFNR patients from controls. Pancreatic enzyme supplementation and fat-soluble vitamin deficiency was associated with the CF group compared to controls, but not for the CFNR group vs controls (p=0.003, p=0.02, respectively). Family history of CF and Mycobacterium avium complex (MAC) lung disease were associated with CF/CFNR patients vs controls (p=0.01, p=0.07, respectively). Lung function and other microbiology data were not significantly different between the groups. Pancreatitis and infertility were more prevalent in the CF/CFNR group. The most common diagnosis in the control group was idiopathic bronchiectasis.

Conclusions: Pancreatic insufficiency, CF family history, presence of MAC, pancreatitis and infertility could help differentiate adult-diagnosed CF patients from those who do not have CF among adults with a clinical suspicion for CF. This group should be considered for screening for CF. Multivariate analysis and studying a larger population are the next steps. Education on screening for CF in adults is of paramount importance.

McGarry, M.E.; Ly, N.P. Pediatric Pulmonary, University of California, San Francisco, San Francisco, CA, USA Introduction: Use of chronic medications to prevent lung disease in cystic fibrosis (CF) has led to great advancements in morbidity and mortality. Despite standardized recommendations of medication therapy, there is great variation in prescriptions of these medications. In other diseases, medications are underutilized in minorities. Minorities with CF have more severe lung function, which may be related to medication prescriptions.

Objective: We sought to ascertain chronic medication prescriptions in minorities with CF compared to non-Hispanic white patients.

Methods: This is a cross-sectional analysis of 8,409 non-Hispanic white and 1,066 Hispanic subjects 6-18 years old in the CF Foundation Patient Registry in 2015. We compared the use of chronic medications by subject race/ethnicity using logistic regression adjusted for the following covariates: adjusted for age, pancreatic insufficiency, CFTR mutation class, and insurance status. Chronic medications were: alfa dornase, azithromycin, and hypertonic saline. We compared whether subjects were on all 3 recommended medications (alfa dornase, azithromycin, hypertonic saline). Race/ ethnicity was self-identified non-Hispanic white or Hispanic. We then did sub-analyses by FEV 1 percent predicted.

Results: Hispanic subjects were 72% more likely to be prescribed dornase alfa than non-Hispanic white subjects (aOR 1.72, 95% CI 1.48-2.00, p<0.001). Hispanic subjects were 87% more likely to be prescribed hypertonic saline than non-Hispanic white subjects (aOR 1.87, 95% CI 1.42-2.45, p<0.001). Hispanic subjects were 87% more likely to be prescribed azithromycin than non-Hispanic white subjects (aOR 1.89, 95% CI 1.35-2.62). Hispanic subjects were 171% more likely to be prescribed all 3 medications than non-Hispanic white subjects (aOR 2.71, 95% CI 1.71-4.29, p<0.001).

In subjects with FEV 1 percent predicted less than 80%, there was no difference between ethnicities in prescriptions of any of the medications. In subjects with FEV 1 percent predicted 80% or greater, Hispanic subjects were more likely to be prescribed hypertonic saline (aOR 1.77, 95% CI 1.31-2.39), dornase alfa (aOR 1.67, 95% CI 1.43-1.95, p<0.001), azithromycin (aOR 1.73, 95% CI 1.19-2.53, p=0.004), and to be on all 3 medications (aOR 2.77, 95% CI 1.59-4.58, p<0.001).

Conclusion: In CF, unlike in other diseases, Hispanics are more likely to be prescribed recommended chronic medication than non-Hispanic whites, but only in those with normal lung function. Medication prescriptions do not explain why Hispanics have more severe pulmonary function. Besides prescriptions, there may be differential adherence or correct use of these medications by ethnicity due to health literacy or language barriers. Efficacy of these medications may be different in Hispanic patients as very few Hispanics were included in the clinical trials of the efficacy of these medications. (J Cyst Fibros. 2017; 16:702-8) of a prognostic score developed from the French CF Registry that demonstrated the ability to accurately predict adults with CF at high risk for transplant or death in three years. This score incorporated the following variables which are common to many registries: percent predicted FEV 1 , BMI, use of long-term oxygen therapy or BiPAP, the number of IV antibiotic courses in a year, whether a patient was hospitalized in the year, use of oral corticosteroids, and infection with B. cepacia complex. Our goal for this study was to determine how well the score would predict patient outcomes in an independent population, using the Canadian registry.

Methods: Patients aged 18 years or older who were followed in the Canadian CF Registry in 2011 and whose vital status was known in 2014 were selected for this study. Patients were excluded if they were transplanted prior to 2011 or were lost to follow-up in the study period. Variables were defined in the same way as used in the original study, however the number of IV antibiotic courses for Canada combined hospitalizations and home IV courses. The prognostic score was calculated using the scoring tool from the Nkam article and modeled using logistic regression. Predictive ability was assessed using a c-statistic and a one-sided p-value of 0.05.

Results: Of the 2,074 adults followed in 2011, 1,123 (54.1%) were male, 128 (6.2%) had FEV1 percent predicted under 30%, 23 (1.1%) had a BMI less than 16 kg/m 2 , 113 (5.4%) indicated home oxygen use, 332 (16.0%) were infected with B. cepacia complex, 12 (0.6%) used BiPAP, 97 (4.7%) used oral corticosteroids, 561 (27.0%) were hospitalized, and 144 (6.9%) had more than 2 courses of IV antibiotics throughout the year. Within 3 years, 224 patients (10.8%) received a lung transplant (n=116) or died without a transplant (n=108). The median prognostic score was 1 (range 0-8.5), with 147 (7.1%) patients identified as high risk (score≥4). For every one unit increase in the score, the odds of being transplanted or dying was 3.2 (95% CI: 2.75-3.62, p<0.0001). The score showed great discrimination power of detecting patient outcome with a c-statistic of 0.899.

The risk score developed on the French data performed very well when applied to the Canadian CF data, despite some differences in data collection and variable definitions. This clinical scoring tool may be useful for identifying those patients at high risk of death or transplant within three years. Introduction: Cystic fibrosis (CF) patients are prescribed medications known to be associated with ototoxicity, including aminoglycosides and macrolides. Ototoxicity is reported in the CF Foundation Patient Registry to occur in 3% of CF patients. Although there has been focus on hearing-related events, there is little attention to vestibular function, which may also be adversely affected by these medications. We report the updated results of our screening program using the well-validated screening instruments, Tinnitus Functional Index (TFI) and Vertigo Symptom Scale (VSS).

Methods: The TFI and VSS are adapted into a web-administered format for ease of implementation in the adult CF clinic. All patients are invited to complete the surveys in the clinic setting. The instruments have well-established threshold scores and threshold changes in a score. TFI total scores (potential range 0-100): <25 (no need for intervention), 25-50 (possible need), and >50 (recommend intervention). VSS total scores (potential range 0-60): ≥6.5 (clinically elevated dizziness) and ≥ 12 (severe dizziness), both requiring further assessment and intervention. All patients who score in the upper ranges are referred to Otolaryngology (ENT) and Audiology for further evaluation. ENT would then evaluate them rendering a diagnosis via physical exam, vestibular testing and/or audiometry. We also conducted a chart review collecting data associated with ototoxicity (e.g. aminoglycoside use).

Results: Seventy-five patients have completed the surveys as part of routine clinical care. The overall mean TFI score is 8±15.6 and overall mean VSS total score is 10±11.6 (clinically elevated dizziness). The proportion of patients in each scoring category is shown in the Figure. Thirty-four patients (45%) had a VSS score that warranted further evaluation (13% clinically elevated dizziness; 32% clinically severe dizziness). Four patients have completed evaluation with three (75%) found to have clinical vertigo (2 with benign paroxysmal positional vertigo; 1 with bilateral vestibular hypofunction). The fourth patient was found to have orthostatic dysautonomia and/or headaches causing her dizziness.

Conclusion: Screening for ototoxicity has found a high proportion of patients reporting clinically relevant vestibular dysfunction. Although the numbers of subjects with complete evaluation are small, they are demonstrating diagnoses that could be related to CFTR function (BPPV) or drug toxicity (BVH). Assessment of other correlative factors is ongoing.

Rizvi, L. 1 ; Griffin, K. 1 ; Sykes, J. 1 ; Abhyankar, P. 1 ; Wassermann, J. 2 ; Tullis, E. 1 1. Respirology, St. Michael's Hospital, Toronto, ON, Canada; 2. Anesthesia, St. Michael's Hospital, Toronto, ON, Canada Introduction: As people with cystic fibrosis (CF) are surviving longer, the likelihood of them encountering surgery and anesthesia is also increasing. Due to the complex nature of CF disease, an understanding of type and impact of anesthesia and surgery in CF patients is required to optimize the pre-, peri-and postoperative status of patients. There are only a few studies on this topic to date, therefore a further investigation is needed on the care of adult CF patients who undergo surgery.

Objective: To describe the types of surgery typically seen in adults with CF and the demographic and clinical profile of patients requiring surgery.

Method: A retrospective chart review of the surgeries on adults with CF at St. Michael's Hospital (SMH) between the years of 1999 and 2016 was performed. This study was approved by SMH's research ethics board. We excluded patients that had transplant surgery, endoscopy, lithotripsy, obstetrical procedures such as caesarian section, or underwent surgery at another hospital. If two procedures were performed on the same day with the same goal of care, they were counted as one surgery (e.g., septoplasty and functional endoscopic sinus surgery). However, if a surgeon performed two unique tasks with differing purposes (e.g., a hernia repair as well as an ileostomy); they were counted as separate surgical events.

We analyzed 173 surgical records involving 119 CF patients. Median age at the time of surgery was 29.6 years (range 14.9-75.2 years), median BMI was 22.13 (range 15.06-38.01), 60.5% were male; 79.8% of all patients were pancreatic insufficient and 44.5% had CF-related diabetes. Surgery for chronic sinusitis and/or nasal polyps was the most the common type of surgery (52.6% of operations) followed by abdominal surgery (appendectomy, cholecystectomy, etc.) (30.1% of operations). The majority of operations required general anesthesia (79.8%). Fifty-two of the 173 operations (30.1%) were day surgery, 70 operations (40.5%) required an overnight stay for monitoring and only 47 (27.2%) of the operations were performed on inpatients. In 23 of 173 surgeries (13.3%), the pre-operative forced expiratory volume in 1 second (FEV 1 ) was <40% predicted. The median FEV 1 in the 138 patients requiring general anesthesia was 66.7% predicted (range 26-123.5%) and 14 (10%) had FEV 1 of <40% predicted. Median length of ICU stay for the patients who required postoperative admission was 4 days (range 1-26 days) and median days to hospital discharge from surgery for all patients was 1 day (range 0-37 days) and for emergent cases (N=24) was 9 days (range 0-28 days). The majority (77.3%) of patients that had surgery are still alive but there were 4 deaths (3.4%) within 30 days of surgery.

Conclusion: Surgery and anesthesia administration can be carried out safely in CF patients, even in patients with low FEV 1 . Most patients have day surgery or an overnight admission. It is difficult to distinguish the impact of the anesthesia from that of surgery on patient outcomes. This experience has helped guide best practice guidelines for our adult CF centre regarding anesthesia management for CF patients undergoing nontransplant surgery. Methods: Longitudinal data in a single pediatric CF center were obtained from the CF Foundation Patient Registry. Socio-environmental measures included TSE, mother's and father's education, annual household income, and health insurance type. Clinical data included genotype, microbiology, and spirometry. Spirometric measures at each age from 6 to 15 years old were compared to normative (Wang, FEV 1 %: normal=100) predictive equations. For each participant, the highest and the lowest FEV 1 % at each age were obtained. Data analyses were performed separately for peak FEV 1 % (highest recorded FEV 1 % at each age, representing optimal lung health) and nadir FEV 1 % (lowest recorded FEV 1 % at each age, representing severity of exacerbations). Multivariable analyses used growth curve models and fixed effect models of lung function.

Results: The sample (N=142) was 46% female and 56% F508del homozygous. TSE, present in 28% of children, increased with low paternal education (OR 6.6, p=0.001), low maternal education (OR 3.7, p=0.004), and low income (OR 3.3, p=0.005). At age 6, smoke-exposed children had nadir FEV 1 % 9% lower (p=0.003) than unexposed children, and the difference reached 19% (p=0.017) by age 15; peak FEV 1 % followed a similar pattern (Figure) . The annual rate of lung function decline was triple in exposed vs unexposed children (2.17 vs 0.83 for peak, 2.17 vs 0.67 for nadir). In fixed effect models, the difference in longitudinal trajectories by TSE reached significance by age 14 (b= -7.17, p=0.041). In nested growth curve models, TSE and income had independent effects on lung function. Nadir FEV 1 % decreased 6% with TSE (p=0.042) and increased 1% with each $10,000 income (p=0.035); peak FEV 1 % decreased 4% for homozygous F508del genotype (p=0.037) and increased 1% with each $10,000 income (p=0.039). In smoke-exposed children, each $10,000 income increased nadir FEV 1 % by 2% (p=0.050).

Conclusions: TSE and income are independent predictors of pulmonary decline in pediatric CF. Smoke cessation should be an essential component of CF care, emphasized at CF diagnosis. Low-income patients with smoke exposure should be prioritized recipients of interventions to address pulmonary disparities. Introduction: Canadian cystic fibrosis (CF) survival has dramatically increased over the past two decades. Forecasting the CF population's age distribution and organ dysfunction would inform the planning of resources for these complex patients. In a recent study, the flow method was used to forecast a 50% increase in the Western European CF population by 2025, with the adult population experiencing the largest increase (Burgel PR, et al. Eue Respir J. 2015; 46(1) :133-41). However, it is unclear if such forecasts apply to other CF populations for which the survival rates may differ. Further, the accuracy of such predictions has not been evaluated, particularly for long-term predictions.

Objectives: The objectives were to: (1) calculate the short-and long-term forecasted characteristics of the Canadian CF population in terms of: (i) adult vs. pediatric patients, (ii) CF-related diabetes (CFRD) and (iii) transplanted patients; (2) assess the accuracy of short-and long-term flow method forecasting using longitudinal Canadian CF Registry data.

Methods: This population-based cohort study used longitudinal Canadian CF Registry (CCFR) data from 1996-2015. Short-and long-term projections were defined as 5-and 15-years in the future, respectively. We applied the flow methodology to calculate annual entering (new diagnoses, re-tracked patients), transitioning (moving from pediatric to adult care at age 18) and exiting (death, lost-to-follow-up) flow rates and averaged them over a 5-year period. The average flow rate was then applied to the base population to determine the 5-and 15-year projected numbers of patients in Canada within each subgroup. We compared the estimates with the actual number of patients in the registry.

Results: The short-term predicted number of total CF patients using the flow method was minimally overestimated by 1.3% when compared to actual registry data. The number of adults was overestimated by 4.1% whereas the model underestimated the pediatric patients by 1.1%. In contrast, long-term estimates underestimated the number of adults by 29.6% and pediatric patients by 7%. Short-term predictions comparing actual with predicted numbers of CFRD patients were similar; however long-term CFRD projections for both adult and pediatric patients were substantially underestimated by 56.5% and 74.1% respectively. Analyses are ongoing to compare the projected number of transplanted patients.

Conclusion: The flow methodology accurately forecasted short-term 5-year predictions of pediatric and adult patients, while the long-term 15-year predictions were significantly underestimated. The flow methodology assumes a constant rate of change over time based on historic data and, given significant changes in survival and CFRD rates in the considered period, long-term predictions were less accurate. A sophisticated modelling technique that can account for changing flow rates over time may allow for more accurate predictions.

In 1993, the Central Connecticut Cystic Fibrosis Center, in conjunction with the clinical laboratory at UConn Health, began to provide diagnostic services for cystic fibrosis at birth including measurement of immunoreactive trypsinogen, mutation analysis and sweat testing. All testing is performed at one location to help solidify diagnosis of CF within a few days after blood spot receipt. In 2009, when CF newborn screening was mandated in CT, our program began screening 7/10 newborns with the remaining newborns being screened at Yale. Over the ensuing 2 decades, the screening process was refined and in 2013 sweat tests were performed on approximately 49% of positively screened newborns in our program. In late 2014, we began genetic counseling for cystic fibrosis via telemedicine with genetic counselors at the University of Florida (SIP grant funding) at the time of sweat testing. Initially, we did not have many families availing themselves of genetic counseling; however, after changing to an "opt out" rather than an "opt in" method of scheduling as well as a letter to pediatricians reminding them of the standard of care for CF diagnosis during the newborn period, acceptance of the service improved. Though the initial outcomes of the project were measured in a family satisfaction survey, we perceived an increase in our completed sweat tests and wished to investigate further.

Methods: We reviewed the annual newborn screening/sweat test records of infants born between January 1, 2013 (prior to genetic counseling) and December 31, 2017 (after 2 years of added genetic counseling).

Results: The 2013 data revealed that 62 of 146 (49%) positively screened infants completed a recommended sweat test. In contrast, 2017 data revealed that 116 of 148 (80%) positively screened infants completed a recommended sweat test (P<0.0001 by chi-square). Furthermore, in 2017, approximately 71.5% of these infants had the sweat test performed by 42 days of age. All CF positive infants were identified with two mutations on genetics prior to sweat testing. Before providing this "free service" of genetic counseling via telemedicine <1% of our positively screened infants received genetic counseling for cystic fibrosis whereas after 2 years of this program >99% receive genetic counseling.

Conclusions: Providing genetic counseling for cystic fibrosis via telemedicine at the time of sweat testing for patients with positive CF newborn screen, not only provides great family and patient satisfaction, but also increases the likelihood that families receive genetic counseling for CF. Furthermore, our experience suggests enhancement of screening programs with free genetic counseling via telemedicine could decrease the time to CF diagnosis by incentivizing families to complete the sweat test.

additional 208 identified as CF carriers (11.4%). This group of CF carriers includes those whose mutations were found to be in cis or were retroactively determined to be non-CF causing (e.g. 1525-42G>A and -741T>G). California's CF newborn screening algorithm begins with measuring immunoreactive trypsinogen (IRT) on all newborn blood spot specimens. Screen-positive specimens based on a high IRT cutoff receive additional testing using a panel of 40 known deleterious CFTR mutations designed for California's diverse population. Specimens with two panel mutations identified (0.4% of high IRT) were immediately reported as screen-positive for CF. Specimens with one panel mutation identified (5.3% of high IRT) were further tested by focused DNA sequencing, which can detect previously known as well as novel mutations/variants. Specimens with one panel mutation and no further mutations/variants identified through focused sequencing were reported as CF carriers and those with additional mutations identified were reported as screen-positive for CF. All screen-positive cases were referred for diagnostic testing and evaluation at one of 15 CF specialty care centers in the state, which also ordered molecular testing on parent samples in some cases. Mothers of newborns identified as CF carriers were sent a letter notifying them of their child's carrier status and they were offered free telephone genetic counseling. Approximately 9.1% of families with newborns identified as CF carriers received genetic counseling through this service. Screen-negative cases include specimens with low IRT, or specimens that had an elevated IRT and no panel mutations identified. Among the 42 false negative CF cases, 21 (50%) were missed due to low IRT, 15 (36%) were missed due to rare mutations not included on the CFTR mutation panel, and 6 (14%) were missed due to no additional mutations identified by DNA sequencing. The ratio of CF screen-positive cases to confirmed cases was 2.9:1. Overall CF prevalence was 1 in 7,911 births. California's CF newborn screening algorithm has proven quite robust in its application to the State's diverse, multi-ethnic population and may serve as a useful model for newborn screening programs more broadly.

Introduction: Our adult CF clinic is the largest in Western Canada and is involved in both investigator-initiated and industry-sponsored clinical trials. While barriers and facilitators to research participation have been studied extensively in other disease areas (eg, cancer), there remains a limited understanding of the factors that influence participation in research for individuals with CF. The purpose of this study is to gain a better understanding of the types of research that individuals with CF are most interested in and to identify barriers and facilitators to participation.

Methods: A survey consisting of 68 questions covering a wide variety of factors potentially influencing research participation was developed by our adult CF research team and was reviewed and approved by the CF clinical team and our Patient Advisory Council (PAC). The format of the anonymous survey includes both multiple choice and Likert-style questions, whereby patients rank the degree to which they agree with the statements on a scale of 1 to 10.

Results: The survey has been completed by 110 adult CF patients thus far. Many individuals with CF (64%) do not actively seek information regarding research studies but if they are to receive information most would like to receive it via e-mail or directly from the CF physician or nurse. Individuals with CF are most interested in clinical trials that target the root cause for CF, infection, and inflammation. They are also most interested in studies that involve pulmonary exacerbations or exercise. Identified barriers to research participation include being too busy and living too far away from the CF clinic. Facilitators to research participation include ongoing access to the study drug following trial completion, compensation for study participation, and study visits being conducted in the patient's hometown. Very few individuals (<5%) are willing to attend study visits more than twice per week. Individuals with CF are most comfortable receiving experimental treatments in the form of a pill or via nebulizer and least comfortable receiving treatments ntravenously or subcutaneously. Most individuals with CF are comfortable providing sputum, saliva, blood, and urine samples and genetic information but are least comfortable providing stool samples.

The important patient-derived information generated by this study will allow us to optimize the selection and design of clinical trials in CF to enhance the likelihood of successful subject recruitment. The findings from this study will also help inform the research priorities for the newly formed Canadian CF Clinical Trial Network. We are currently in the process of expanding the survey to our affiliated pediatric clinic to determine if the research interests and barriers/ facilitators are similar across the age spectrum.

Acknowledgment: This research is supported by Cystic Fibrosis Canada. Clinical research is a valuable addition to a cystic fibrosis (CF) center. After assessing the number of patients enrolled in clinical trials, it was concluded that the CF center could improve the number of patients enrolled in clinical research. The CF center further determined through patient interviews that many potential patients (especially in pediatrics) had little knowledge of current trials or knew how to find appropriate clinical trials in which they qualify.

Purpose: The purpose of this project was to incorporate a culture of research into our CF center by increasing baseline knowledge of clinical trials.

Methods: Patients were initially surveyed to determine knowledge gaps. An anonymous survey was given to both the adult and pediatric clinics over a 2-week period (4-6 clinics). Initial results showed that the clinic needed to provide education to the patients about research and introduce the Clinical Research Coordinator (CRC) to all patients in order to create a positive relationship. The initial survey included 6 adult patients and 14 pediatric patients. Of the 20 patients surveyed, 16 were aware research was being conducted, but only 10 felt that they were routinely updated. Out of 20 patients, 17 were interested in participation in clinical trials and 9 were interested in referrals. Eight patients said that there were barriers to their participation in clinical research. The stated barriers written on the survey were: lack of information, cost, number of visits, availability, too many needles, change in disease, medication discontinuation, other commitments, side effects, time, money, and age of child. Most patients preferred to be contacted in person, some were also willing to use US mail, e-mail, phone/text.

Intervention: The CRC began attending all clinics, meeting with patients for a few minutes each visit and was available for their questions/ concerns about participation in clinical trials. At the current time, the CRC has been attending adult and pediatric CF clinics for 6 months. The CRC has been sharing basic information related to clinical trial participation and use of CF Foundation resources during clinic.

Results: Post-survey results indicated the following: the 13 surveyed were aware research was being conducted in our clinic and all felt that they were routinely updated. Out of 13 surveys returned, 8 were interested in participation in clinical trials locally, with 5 of those interested in referrals. Four surveys answered yes that there were barriers to their participation. The stated barriers written on the survey were "If I would have to stop any current medications," "time commitment," "lung function too low," and "distance."

Conclusion: Patients and families in both the pediatric and adult centers seem to be receptive to meeting with the Clinical Research Coordinator. These meetings have helped create a trusting relationship between the CRC and the patients. A culture of research will take time to develop and will require maintenance of this relationship over time. The CF center feels that this relationship is key to creating patient and family engagement and to increase both clinical trial enrollment and a culture of research. Griffin, K.; Rizvi, L.; Tullis, E. Respirology/CF, St. Michael's Hospital, Toronto, ON, Canada Introduction: Understanding the barriers to research participation by cystic fibrosis (CF) patients is important to meet the recruitment targets of the increasing number of clinical trials for CF patients.

Objectives: This study aims to summarize the reasons for refusal of patients who were potentially eligible to participate at the time of recruitment for a specific clinical trial, and who did not agree to participate.

Methods: As part of the research recruitment practices at the Toronto Adult CF Centre, the Toronto CF Patient Registry is used to identify potential participants for clinical trials. The Toronto CF Patient Registry allows for identification of patients with specific CFTR gene mutations, filtered by the forced expiratory volume in one second (FEV 1 ), age, and microbiology as required by a protocol. Additional pre-screening using the electronic medical record and patient chart is then completed for study-specific inclusion/exclusion criteria to identify eligible patients to consider for the trial. Patients may be approached at their clinic visit or via e-mail or telephone. If patients are not interested in participating in the trial, the reason for refusal is recorded.

In this study, we describe responses from patients who were pre-screened by research staff and invited to participate in at least one randomized, double-blind, placebo-controlled research study between February 2015 and December 2017.

Results: A total of 6 clinical trials were conducted at the Toronto Adult CF Centre during the study period. Of these, 3 were phase 2 and 3 were phase 3 clinical trials involving CFTR modulators, or anti-inflammatory compounds. In total, 82 patients were deemed to be eligible based on the pre-screening process and were invited to participate in one or more clinical drug trial. Twenty-six (32%) patients approached agreed to participate and scheduled a screening visit, 21 (26%) did not respond or follow-up with research staff about their intention to participate in the trial and 35 (43%) did not agree to participate.

Ten (29%) of the 35 patients who declined participation reported that they were too busy because of work, 6 (17%) reported being generally too busy, (e.g. unable to make additional visits to clinic), 2 (6%) were not able to participate as they were away for school, 3 (9%) planned to start a family in the near future and so were not agreeable to interventions to prevent pregnancy, 5 (14%) were not interested in participating in clinical trials in general, 4 (11%) became ineligible after they were initially approached and thus were not screened and 5 (14%) did not provide a specific reason for declining.

Discussion: A third of patients who are eligible and approached are initially interested in participating in clinical trials. In our adult CF centre, 25% of patients do not reply but it is not clear if this is due to lack of interest or failure to connect. The decision to participate or not in clinical research appears to be related to the time commitment required of clinical trials interfering with work and personal schedules and prior commitments. Future studies should explore reasons for not being interested in participating in clinical trials as well as lack of response from patients on offers to participate in trials from their CF centre. Background: A critical step in the eQUIP-CR process is performing an assessment of the clinical research program by surveying key personnel within and related to the research process. This is essential to identifying operational improvement needs and evaluating the culture of research in our program.

Objectives: 1) Distribute the eQUIP-CR Program Assessment to determine if we are meeting the needs of our clinical and administrative partners in our continued efforts to recruit subjects and conduct research. 2) Compile survey results and form a cross-sectional work group to address potential areas of change.

Methods: Our previous assessment was conducted in 2012 with initiation of eQUIP-CR. In mid-2017 we determined that re-assessment was needed. In contrast to our prior assessment, we expanded the distribution to include not only members of the CF Team but also pulmonary MDs, nurses, front desk staff/medical assistants (MA) and administrative staff. We also included individuals from outside our clinical section, such as the translational research unit, IRB, Corporate Compliance and Investigational Pharmacy. Surveys were completed anonymously; however, individuals were asked to mark their name on a return log so that respondents could be tallied.

Results: Of the 29 surveys distributed, 22 were returned. The highest response rate was represented by clinic staff including 6/6 RNs and 2/2 front desk/MA staff. Overall MD response was 4/7. Administrative staff response was 1/2. CF Team response was 100%. Survey results were compiled, and the following topics were prioritized: 1) Flow of research subjects during study visits in clinic; 2) Communication with staff about studies/research activities; and 3) Annual updates at staff meetings. An invitation to participate in an eQUIP work group was extended to clinical staff and one individual from each of three key areas was identified as a delegate to represent their peers, including PFT lab, nursing, and front desk/MA. The work group was led by the research team and met to discuss strategies to implement process improvements for the identified priorities. Individuals were eager to contribute and share ideas. The entire group benefited from different viewpoints and various approaches to problem solving. The initial improvement recommendation was to create a research visit intake sheet, providing specific instructions on activities needed for a research visit in clinic. We have begun to implement this tool, with ongoing revision based on feedback from the work group.

Next steps: We plan to create a monthly CF research update on current studies, local research activities and highlights from the CFF and TDN to be distributed within our section and to our institutional research colleagues. A "State of the CF Research Program" will be presented annually to staff, faculty and section leadership.

Conclusions: Previous eQUIP-CR efforts have focused on improvements within the CF Team. This project expanded the scope to include the clinical team. This has helped to foster the culture of research and identify gaps in quality in our research program. These efforts will improve the conduct of research at our center and enhance communication and partnership between the research and clinical teams. Background: The amount of time and effort required to participate in CF clinical trials varies drastically depending on the intensity, complexity, and feasibility of a study. Such variability can make it difficult to delegate and manage various responsibilities. The CFFT TDN Clinical Research Metrics Report provides CF centers an overview of enrollment and start-up metrics, which allows sites to track annual progress. However, in order to benefit from this report, understanding how to interpret these data is critical. A lack of understanding prompted our research team to reassess how our site tracks TDN Metrics, develop visual tools to aid with comprehension, and set goals for future improvement.

Objective: Create customizable spreadsheets to develop visual tools that help track, interpret, and react to changes in TDN Metrics.

Methods: In January, 2018, an electronic database was created to visualize patient enrollment dates in graph form. Using the TDN weighted (complexity) score, assessment reports were generated based on dates of enrollment, and total weight score was calculated on a quarterly basis since quarter-four, 2011. Utilizing the TDN Metrics Report as a guide, our site set the minimum enrollment goal equal to 10% of our CF patient population, and adjusting for weighted score. To reach this goal, an average weighted score of 10 needed to be maintained quarterly. Graphs were evaluated and communicated to the study coordinators to determine the number of subjects that need to be enrolled per quarter in order to reach the enrollment goal.

Results: During the first quarter of 2018, our site earned a total weighted enrollment score of 13.6, 3.6 points above the quarterly goal. A similar increase in weighted enrollment score has been observed during quarter two, reaching a score of 11.5 by mid-May. Both quarters achieved higher scores when compared to all quarters since quarter two, 2014.

Conclusion: While the number of weighted units greatly increased after implementation of this tool, additional data must be collected to assess sustainability. Instead of using TDN ranking as a goal, these customized visualizations were used at our site to help achieve site-specific enrollment goals and determine strategies for internal growth. Similar visualization tools have also been developed for study start-up metrics to help identify areas of achievement and opportunities for improvement. Viewing progress and establishing clear goals has not only increased weighted enrollment, but has also improved job satisfaction at our site. Objective: We launched the Seattle Children's Grams to Grow Initiative to increase W/L %ile of infants at our center using quality improvement (QI) methodology.

Methods: A multidisciplinary QI team, supported in part by the CF Learning Network (CFLN), included 2 parent partners who co-produced the interventions. We organized barriers in a fishbone diagram. In a cross-sectional retrospective chart review of patients <24 months, we created a Pareto chart to prioritize system and patient level issues. Interventions focused on chronic care processes (1) Clinic follow-up: reliable appointment scheduling by scheduler, medical assistant (MA), and family;

(2) Data support through Registry and CFLN: consistent and current infant nutrition data to track metrics; and (3) Standardized approach: therapeutic nutrition algorithm to target expected grams per day of weight gain. The Model for Improvement framework was used to guide Plan-Do-Study-Act (PDSA) cycles to test interventions iteratively. Run charts displayed process metrics monthly. Mann-Whitney was used to compare annual W/L %ile pre-and post-interventions.

Results: The first PDSA ramp (12/2017-3/2018) targeted reliable scheduling of clinic follow-up. Thirteen infants <24 months were followed at our center in this 90-day cycle. Following changes to scheduler job aide and MA script, proportion of patients with ≥1 visit scheduled at a time increased from 71% to 92%. The proportion of clinic visits occurring at guideline-recommended frequency (e.g. monthly for patients <6 months old) increased from 60% to 82%. In Registry PDSA cycles to improve outcome tracking, we reduced mean time for new diagnosis infants to consent to Registry from 112 days, SD 135d (6/2014-9/2017, N=21 infants) to 28 days, SD 16d (10/2017-4/2018, N=6 infants). Reliable documentation of nutritional supplements increased from 33% (2/6 visits) to 63% (5/8 visits) . Nutrition algorithm to standardize provider practice reached team consensus and was launched 5/2018. Algorithm adherence metrics are pending. To date (12/2017-3/2018), proportion of monthly clinic visits in which infants gained expected grams per day increased from 60% (12/2017) to 73% (3/2018) . Annual median W/L %ile of infants significantly increased from 4/2016 -3/2017 to 4/2017 -3/2018 , p=0.006).

Conclusion: Development of interventions to target system failures and patient needs demonstrates promising improvements to early nutritional outcomes in CF infants at our center.

Acknowledgments Introduction: Many infants with cystic fibrosis (CF) struggle with optimizing nutrition, which is critical for appropriate growth, lung health, immunologic and neurologic development. Although the newborn screen for CF has helped improve clinical outcomes, growth in children with CF is delayed when compared to healthy children (JAMA Pediatr. 2017; 171:546-54) . Weight-for-length (WFL) should be measured monthly in the first year of life and then every 3 months if meeting nutritional goals. If not, then measurements may need to be more frequent (J Pediatr Gastroenterol Nutr. 2002; 35:246-59) . In 2016, our center's average WFL percentile, 59.5%, was below the national average of 65.6%, motivating us to examine this population through quality improvement (QI).

Methods: We categorized our CF patients from 0-2 years of age based on our registry data and their CDC WFL percentiles: failure (<10th percentile), at-risk (10th to <25th percentile), concerning (25th to <50th percentile), and optimal (>50th percentile). We targeted the at-risk and failure groups and created a fishbone diagram to identify potential barriers to achieving optimal growth. We designed a survey of questions that would target these barriers, such as access to food, formula, medications, and feeding tolerance. We created 3 plan-do-study-act (PDSA) cycles, over 4 weeks each.

Results: We initially identified a total of 38 infants; 2 were at-risk and 5 were in failure based on stated WFL criteria. For our 1st cycle, we called these 7 families weekly and intervened for any obstacles that arose in between clinic visits. At the end of our first cycle, 4 infants previously <10th percentile graduated to the at-risk category. For our 2nd cycle, we contacted 10 infants via phone and MyChart (electronic medical record (EMR) messages) weekly. For our 3rd intervention, we encouraged MyChart use for families to allow for more flexibility and time for response. Overall, our number of infants in failure decreased from 5 to 1 after 3 PDSA cycles. Our infants at-risk peaked after our 1st cycle from 2 infants to 9 but decreased to 2 infants by the end of our 3rd cycle. Our infants >25th percentile WFL increased from 12 to 16 (Figure) .

Conclusion: Optimizing infant nutrition is essential for future overall health and is an important obstacle for CF providers and families. Through QI, we identified barriers and risk factors that further stratify this population, such as pancreatic sufficiency or socioeconomic status. By identification of barriers, more frequent intervention, and use of EMR, we have improved overall infant WFL percentile at our CF center.

Seid, M. 1 ; Thurmond, S. 1 ; Moore, E. 2 ; Britto, M. 1 ; Amin, R. 1 ; Zeribi, K. 3 ; Thomas, L.K. 4 ; Griffin, N. 1 1. CCHMC, Cincinnati, OH, USA; 2. Family Partner, Hershey, PA, USA; 3. Shift Results, Seattle, WA, USA; 4. Children's of Alabama, Birmingham, AL, USA Objective: To share learning and results from the CF Learning Network (CFLN) pilot phase. Sponsored by the CF Foundation, the CFLN involves people with CF, families, clinicians, and researchers collaborating to coproduce a healthcare system in which people with CF have healthrelated quality of life no different from the general population, achieve their goals, and lead full, productive lives.

Methods: Co-designed by a multi-stakeholder group, the CFLN makes it easier for everyone to work together to improve health and healthcare, thereby leveraging existing capabilities, infrastructure, intrinsic motivation, and collective intelligence of the whole CF community to improve clinical outcomes (FEV1) as well as quality of life. Key change concepts are 1) create a common vision with shared knowledge and know-how, effective teamwork, and timely, accurate data; and 2) ensure reliable chronic care (personalized, evidence-based, timely care; effective self-management; strong care partnerships; joyful work environment). The CFLN focused on timely (monthly) data entry into PortCF, integrating patients and family partners (PFPs) into quality improvement (QI) teams, collaborative pre-visit planning, and population management. The pilot phase (2016 -2018) tested this design with 29 CF programs and a team of community innovators (people not affiliated with a current CFLN program).

Results: As of April 2018, teams sustained high participation in network activities (median = 86% on a bundled "all or none" metric). Spring 2018 Community Conference attendees, surveyed about the CFLN Community, feel that they have a role in the community (88% agree or strongly agree), the community can solve problems (89%), they have influence over the community (75%) and feel hopeful about the future of the community (89%). Process measures have improved: the percent of clinical encounters entered into the registry within 30 days increased from the baseline median of 65% to 85% and average days between encounter date and entry dropped from 42 to 6. Most (88%) community innovators reported leading or co-leading improvement initiatives and 100% of PFPs felt valued or highly valued by their CF QI Team. Thirteen of the 29 CF programs were able to increase to at least 50% of clinic visits collaboratively planned with patients and families. We did not expect clinical outcomes to show statistical improvement within the first two years. Although there are changes in lung function and BMI%, it is too soon to tell if these represent statistically significant improvement.

Discussion: The CFLN has resulted in a highly motivated, collaborative, inclusive community improving chronic care processes that have been shown, in the chronic care literature, to be linked to clinical and quality of life outcomes. Post-pilot plans include further network maturation, strategies for spread and scale, and continued focus on implementing changes in all aspects of the chronic care model. As well, the CFLN will build infrastructure to pursue "real-world research" to accelerate implementation, improvement, and personalized care to improve outcomes. From an early age, patients with cystic fibrosis often experience stepwise declines in lung function following acute worsening of chronic bacterial infections. To date, no protocols have been published by the Cystic Fibrosis Foundation regarding clinic-based management of pulmonary exacerbations in children. We hypothesize that with a standardized protocol to treat decline in lung function, patients will better maintain pulmonary function over time. Our project aims to develop, implement, and assess the efficacy of a treatment algorithm for management of pulmonary exacerbations presenting to the Pediatric Cystic Fibrosis Clinic at Childrens' Health Dallas.

In patients able to perform PFTs, an increase in pulmonary symptoms is assessed in the context of decline from baseline FEV1; percent decline in FEV1 of <5%, 5-10%, or >10% corresponds to treatment modifications of proportional magnitude. Treatment modifications range from increased airway clearance to oral antibiotics or inpatient admission. In patient populations unable to perform PFTs, symptoms are the primary indicator of the presence and severity of an exacerbation, and depending on severity of respiratory symptoms and the presence of other systemic symptoms (weight loss, fever, hemoptysis), therapy is escalated accordingly. To assess the algorithm's capability to parallel provider judgement regarding exacerbation management, patients demonstrating increased pulmonary symptoms were evaluated by providers while a concurrent algorithm-based assessment was performed. Immediately following the patient encounter, providers were informed of the algorithm recommendations and asked to evaluate the appropriateness of algorithm-recommended treatment modifications, laboratory work-up, and follow-up time via survey.

Algorithm-based treatment and work-up recommendations were found to coincide with physician judgement in 90% of encounters to which the algorithm was applied. Algorithm-recommended follow-up times were in accord with provider judgement 70% of the time. While data collection is still in progress, we have demonstrated that use of an algorithm is appropriate for the identification and treatment of pulmonary exacerbations. Longitudinal data regarding the recovery of lung function in patients treated with consideration of algorithm recommendations will serve to determine if algorithm-based management better improves long-term preservation of lung function when compared to past practice at the University of Texas Southwestern/Children's Health Pediatric CF clinic. If successful, an evidence-based algorithmic approach to exacerbation management will serve to standardize care among providers, making follow-ups and sick visits more efficient and amenable to team-based care. The patient experience survey specifically asks questions relating to being taken directly to an exam room on arrival and maintaining a 6-foot distance from other CF patients. The data fell below the 50th percentile compared to other CF programs. The registration process did not follow the CF Foundation guidelines of a private area and maintaining a 6-foot distance from others with CF. The aim of changing the process was to improve the patient experience through changing the registration process resulting in improvement to the 75th percentile survey scores by end of 2017.

Methods: Through review of data and understanding patients' and families' frustration, a team was assembled to review changing the registration and arrival process. The team comprised registration, nursing, quality improvement, guest services manager, and management. Quality improvement methodologies of process mapping were used to examine the process from arrival, registration, and to exam room and Plan-Do-Study-Act (PDSA) was used to test redesign ideas.

Results: The team went through 3 PDSA cycles and identified several areas of improvement. First, the area where patient's check-in was changed from 2nd floor registration desk to reception desk on the main floor. This allowed for more space to accommodate the recommended 6 feet between CF patients. Second, patient registration was changed from 2nd floor registration desk to the exam room on arrival, which expedited arrival to exam room, forgoing the waiting area, and allowing the CMAs to register and check-in patient concurrently. Lastly, patients were escorted to exam rooms on arrival and taken almost immediately to an exam room after notification to CF team and eliminated 2nd floor waiting room. Before the improvement work, 47% (n= 17) of families reported being brought to exam room on arrival on the experience of care survey. At the onset of the improvement up to March 2018, this result significantly improved to 88% (Fischer's exact test p = 0.023), exceeding the initial goal of 80%. Maintaining 6 feet or more from another with CF improved from 65% to 88%, but was not statistically significant (p=0.223). Phone interviews with patients and families revealed positive feedback as well as positive verbatim comments from the patient experience of care survey. The CF team expressed improved clinic functioning.

Discussion: The change in the process was successful as objectives were met and the CF team will continue to find additional areas to improve the process as it strives to reach the top 10% in the country. recommends that cystic fibrosis (CF) patients' growth and nutritional status be monitored routinely at quarterly visits or sooner if necessary as normal/optimal growth is associated with improved pulmonary function and survival. The 2001 Consensus Nutrition Report states that CF patients should have normal growth. The guidelines recommend the use of BMI% for clinical evaluation as it reflects both genetic as well as health factors. We reviewed our data in the 2014 registry, our first year for our new CF program, and found that our mean BMI% for patients 2-20 was 21%. We instituted a process improvement (PI) project to help us achieve a mean of 50% BMI.

Methods: We devised a PI project with the aim of improving the nutritional state of our patients whose BMIs <50% such that the mean BMI% of our patients aged 2-20 was at least 50th percentile. We determined a number of key drivers to include: following CFF growth and nutrition recommendations for CF patients, following CFF dosing recommendations for pancreatic enzyme replacement therapy, and aggressively managing nutrition in patients as soon as identified by newborn screening as having CF. We used a number of change strategies such as developing a standardized nutrition algorithm for the evaluation and treatment of CF patients 2-20 years, providing parents with growth charts at every visit with identification of nutrition goals, verifying CFF-recommended PERT dosing and patient administration/adherence, increasing the frequency of visits for patients with BMI% <26% (or at risk as defined by rate of weight gain <50% expected for age [≤ 5-8 g/day] or downward crossing of BMI% in last 3 months), prescribing nutritional supplementation (oral and enteral) and appetite enhancers, combining behavioral and nutritional interventions, evaluating for nonpancreatic causes of malabsorption, and evaluating for CF-related diabetes if patient consistently failed to thrive. Furthermore, as part of the PI project a fishbone diagram (Ishikawa diagram) was constructed to look at the barriers faced to achieve the aim. And finally an algorithm was developed to be followed with each patient at each clinic visit based upon a CF Nutrition Risk Protocol from Pulmonary and Gastroenterology services at the Women's and Children's Hospital of North Adelaide, SA, Australia.

The aim of our PI project of increasing the mean BMI% to 50% in our patients age 2-20 years was achieved within two years of initiating the PI project. Not only did the mean improve to 53% but the median BMI% for 2-20 years went from 42 to 56% and the weight for length and weight percentile for patients <24 months improved from 63 to 77% and 21 to 37% respectively.

Conclusions: Baylor College of Medicine opened a new CF center with the Children's Hospital of San Antonio in October 2013. We faced numerous barriers when we set out to provide quality care to patients, many of whom had not been followed consistently in a local CF center. By standardizing care, following the 2001 Consensus recommendations, and most importantly by having a dietitian see the high risk patients frequently and consistently we were able to reach our goal in two years. Psychosocial issues (denial of disease, homelessness, autism), and 3. Silent droppers. Many of the silent droppers often had FEV1 percentile that was well over 100% but was drifting downward. Our team decided to focus on the silent droppers, making sure they were not missed and that there was a standard approach.

Objective: Improve treatment and follow-up of all patients (children and adults) who have decreases in FEV1 percentile of 5% or more from baseline Methods: We developed an algorithm to guide management for patients with FEV1 percentile 5-10% below baseline and >10% below baseline. We also agreed to arrange office follow-up in 2-6 weeks for everyone with declines in FEV1 percentile of 5% of more. We tracked our adherence to the protocol and follow-up visits in Excel with run charts to document our progress and reviewed these monthly with the entire team. After months of discussion and tracking, we changed our report from relative change to absolute change in FEV1 percentile and our team's definition of baseline FEV1 percentile changed as well.

Results: Over a 15-month period, there were 73 episodes where a patient had a decrease in FEV1 percentile 5-10% below baseline, and 74 episodes where a patient had a decrease in FEV1 percentile >10% below baseline. Initially, adherence to the protocol for patients with declines in both groups was between 60-80%. After 9 months of tracking the protocol, and changing the definitions of FEV1 percentile decline and baseline FEV1, adherence to the protocol has been close to 100% for 5 months. Follow-up visit scheduling has gone from 80% to 90% over the 15 months. Median pediatric FEV1 percentile went from 76.2 in 2015 to 94 in 2017. Median adult FEV1 percentile went from 56.3 in 2016 to 60.6 in 2017.

We have successfully standardized our interventions for declines in FEV1 percentile. Compared with before the intervention, all care providers are starting to think more about baseline FEV1 and document it in their clinic notes. We are also seeing patients back more often after seeing a smaller drop in FEV1 to document recovery to baseline. the waiting room to reduce the risk of cross-contamination of pathogens between patients. Our large pediatric CF clinic noted excessive waiting room time and long clinic visits. These problems created an elevated infection exposure risk for patients and low satisfaction among patients, families, and staff. Our primary aim for this project was to minimize or eliminate the amount of time spent in the waiting room for patients with CF. Our secondary aims included reducing total time of clinic visits, reducing the time spent in room without a care provider, and improving patient and provider satisfaction with clinic flow.

Methods: We created a multidisciplinary team including physicians, nurses, nurse assistants, hospital epidemiology staff, ancillary care providers (psychology, social work), and quality improvement coaches. We utilized Lean methodology to identify root causes of our clinic flow problems, develop potential interventions, redesign our clinic processes, reduce waste, and monitor the impact of our changes. Using a week-long period of experimentation, data gathering, and rapid plan-do-study-act cycles, we tested multiple interventions to achieve our desired outcomes. Successful interventions included: direct rooming for patients with CF to eliminate time spent in the waiting room, pre-visit planning questionnaires given at the beginning of clinic visits, a clinic workroom flow board, moving ancillary care team members to a more central location, scheduling future clinic appointments during current visits, and a flag system to indicate when rooms were ready to be cleaned. Data were collected with direct observation on a sample of patients, provider satisfaction surveys, and Quality Data Management (QDM) patient surveys.

Results: At 90 days after implementation of our changes, the median time spent in the waiting room was reduced from 16 minutes to 2 minutes. Time spent in exam room for patients was reduced from a baseline of 111 minutes to 93 minutes. Time to room cleaning after the patients left the room was reduced by 52%, and the downtime without a patient in a room was reduced by 21 minutes. The time a patient spent in the room without a care team provider did not change. Providers reporting being "very satisfied" with clinic infection control increased from 20% to 50%. QDM patient survey responses reporting being brought directly to an exam room increased from 40% to 100%. Providers indicated a positive response as well, with double the number of providers indicating they were now "somewhat satisfied" or "very satisfied" with clinic flow.

Discussion: Using quality improvement methodology and tools, we were able to implement changes to our clinic flow that resulted in reduced infection exposure risk, decreased clinic appointment times, and improved patient and staff satisfaction. Inclusion of representatives from all aspects of patient care, identification of inefficiency in our processes, and improving communication were keys to our success.

Andrews, A.; Meier, E.; Oermann, C.M. Pulmonology, Children's Mercy Kansas City, Kansas City, MO, USA Introduction: Cystic fibrosis hospitalizations can be very difficult for patients and their families. Patients requiring frequent hospitalization are often disengaged from inpatient care due to physical and emotional health issues. Cooperation with and active participation in prescribed therapies (airway clearance, nutrition, exercise) varies among patients but may have a significant effect on the outcome of hospitalization. Weight gain, return to previous baseline pulmonary function, and duration of a hospital stay may be influenced by patient engagement. Recurrent struggles with adherence to therapy can lead to frustration among inpatient care team members. CF Care Team (CFCT) members recognized this challenge and developed an Inpatient Expectation Agreement (IEA) to improve patient engagement with care, staff-patient interactions, and improve outcomes of hospitalization.

Methods: CFCT members were surveyed to assess thoughts on behavioral contracts, patient engagement strategies, and expectations of hospitalized CF patients. Results indicated that 66% of staff agreed that an IEA was needed. An IEA was developed and reviewed with the CFCT. Staff feedback was incorporated and a draft document was provided to the CF Parent Family Advisory Council (CFPFAC) for review. Following discussion, it was decided that children school-age and above would receive the IEA. Additionally, the CFPFAC suggested the use of an unsigned "agreement" rather than "contract" requiring signature. Hospital administration (Patient Advocates and Risk Management) approved the final IEA. A decision was made that the ambulatory CFCT would present the IEA and discuss with the patient/family at the time of hospital admission. An algorithm for IEA use was developed. Education was provided through CFCT meetings, email reminders, and posters in clinic. A tracking tool was developed to ensure the IEA was being provided to the patient/family.

Results: A post-implementation survey was developed and sent to the CFCT. Outcomes included staff satisfaction and the percent of hospitalized patients receiving the IEA. Survey results indicate that, despite education, 33% of staff were not aware of the IEA and 41% of staff forgot to present to patients. Eight patients have received the IEA. Three CF providers have presented the IEA at admission. Following the survey, additional education was provided via email and CFCT meetings. Overall, 77% of staff indicated that the IEA is helpful.

Conclusions: The development of an IEA has led to improved communication among CFCT staff (ambulatory and inpatient) and patients/ families. Improved engagement of patients with inpatient care should drive outcome improvement. Future directions for this project include the assessment of patient responses, changes in behavior and impact on duration of hospitalization. 

Background: Children with CF are at risk of developing CF-related diabetes (CFRD). CFRD is associated with poor nutritional status, decreased lung function, more frequent pulmonary exacerbations, and decreased survival. It is estimated that 20% of adolescents with CF have CFRD (Moran A, et al. Diabetes Care. 2009; 32:1626-31) . International guidelines recommend annual screening with oral glucose tolerance test (OGTT) for children 10 years of age and older. Alberta Children's Hospital CF clinic cares for 140 children and adolescents aged 0 to 17 years, of which 62 are 10 years or older in 2018. Review of 2017 data showed that 5% of eligible patients had OGTT completed. We aimed to improve the frequency of screening for CFRD.

Objectives: To increase the percentage of eligible patients screened for CFRD with OGTT from 5% in 2017 to a goal of 100% by December 31, 2018.

Methods: A multidisciplinary quality improvement (QI) team was formed, and The Dartmouth Institute Microsystem Academy QI methods were used. The process began with identification of patients who met criteria for CFRD screening, and ended with patients completing OGTT. This was achieved by multiple plan-do-study-act (PDSA) cycles, including: addition of OGTT to the patient care roadmap, identification of patients, discussion in pre-clinic meetings, and addition of a visual prompt on the patient chart. The second step was to increase staff and patient knowledge of CFRD screening. Further PDSA cycles included: formal educational session for clinic staff, discussion of CFRD screening with patients and families by the clinic physician, provision of an educational information sheet (http://www.cysticfibrosis.ca/cystic-fibrosis-related-diabetes-cfrd), and improvement in written patient instructions for booking OGTT. Data were collected during each week's clinic to determine the percentage of eligible patients who successfully received education and information on OGTT booking. Data were also collected on number of patients who completed OGTT.

Results: Data from January to December 2017 revealed that 5% (3/58) of eligible patients had completed OGTT for CFRD screening. Baseline data collected for 4 clinics in January 2018 indicated 0% (0/16) of patients had received information regarding CFRD screening. In the first 3 months of the QI initiative, of eligible patients identified at each clinic a median of 90% received CFRD education and OGTT booking information. OGTT has been completed in 34% (12/35) of patients who received information in clinic during the first 3 months. In total, 19% (12/62) of all eligible patients completed OGTT in the first quarter of 2018.

The percentage of all eligible patients completing OGTT for CFRD screening increased from 5% over the 12 months of 2017 to 19% in the first 3 months of the QI initiative. It is anticipated that the percentage will continue to increase in 2018, when the remainder of eligible patients attend clinic, receive CFRD information, and pending OGTTs are completed.

Robinson, K.J.; Prior, C.; Lahiri, T. Pediatrics, University of Vermont Children's Hospital, Burlington, VT, USA Background: Individuals with cystic fibrosis (CF) require more energy to breathe normally, fight lung infections and compensate for fat malabsorption. Maintaining adequate nutrition has been associated with higher pulmonary function. For these reasons, patients with CF require more calories than people without CF. Energy needs of people with CF are estimated to be 1½ to 2 times the needs of those without CF. For some patients and families the financial burden associated with the cost of these additional calorie needs is more than their household budget can allow. A quality improvement project was performed on the inpatient pediatric service at the University of Vermont Children's Hospital to increase screening for food insecurity (FI), a social determinant of health, using the Hunger Vital Sign. The Hunger Vital Sign is a validated two-question screening tool that assesses the risk for FI. The Hunger Vital Sign is embedded in the CMS Accountable Health Communities Health-Related Social Needs Screening Tool, which we built into our version of Epic. Because of this process, we discovered that up to 24% of inpatient CF patients had positive FI screens compared to the known rate of 14% of household FI in Vermont.

Methods: Our SMART Aim was to screen 90% of patients for FI using the Hunger Vital Sign at quarterly visits to our outpatient pediatric CF clinic in 6 months. Our multidisciplinary team developed a key driver diagram to identify essential attributes to reach our goal and outline related interventions. We engaged our CF patient and family advisory board to seek approval to begin screening for FI and educated families with an article in our newsletter. We developed a process flow map of the screening process and related interventions. We partnered with various organizations to develop interventions for families who screened positive for FI. FI screening was expanded to the adult CF clinic once the screening process and related interventions were refined in the pediatric clinic.

Results: To date, 86 patients (56 pediatric patients and 30 adults) have been screened. There were positive screening results for FI in 14% of pediatric patients and 27% of adults. In parallel with this project, our institution developed a food pantry which will provide food to patients and families with positive FI screening results.

Conclusions: FI was more common among children admitted to the University of Vermont Children's Hospital compared to known rates among all Vermont households. This suggests a need to screen patients and families with CF for FI in clinic and provide reliable interventions based on the negative potential impact of FI. Screening for FI in pediatric and adult CF clinics can be performed using the Hunger Vital Sign with support from patients and families. Innovative strategies to intervene for patients who have a positive FI screen should be explored, including a food pantry. Successful implementation of a FI screening process suggests that CF centers should consider screening for other social determinants of health due to the frequency of visits patients with CF have with their CF center and the integration of social work in the CF clinic. Objective: Because of different nutritional needs, psychosocial maturity levels, and use of aggressive interventions we sought to develop a standardized algorithm to implement across our center based on age groups and patient/family acceptance of interventions. We are a new center with largely a young CF population and found wide variability of nutrition interventions being used among providers at different times in the disease process and in different settings. We wanted to be conscious of the different maturity levels of our patients and patient/family acceptance of interventions, and tailor interventions accordingly. We developed our algorithm to provide family-and patient-centered care based on the different needs of diverse age groups while adhering to known successful interventions appropriate per age. For the greatest impact on long-term nutritional and pulmonary outcomes, our algorithm is more aggressive with our younger subset. The goal of this initial project was for >50% of our patients 0-2 years old to have a weight for length (W/L) >50th percentile (%ile) on the WHO growth chart.

Methods: Our algorithm is based on the following age groups: 0-5 years, 5-10 years, and >10 years. It uses color zones based on W/L %ile for patients 0-2 years and body mass index %iles for older patients (green zone: >50th %ile, yellow zone: 25-49th %ile, orange zone: 10-24th %ile, red zone: <10th %ile) and age growth velocity recommendations for patients <10 years. Through several plan-do-study-act cycles, we created a 7-page document in PowerPoint with check boxes and progressive interventions to allow different providers to easily see interventions already and to be implemented. Input was received from physicians across different specialties as well as the CF team's ancillary staff. Weight, weight %ile, height, height %ile, and W/L %ile were collected monthly from patients 0-2 years old who designate LCH as their home CF center. If a patient was seen twice in 1 month the later data was used.

Data: The data demonstrate a positive trend of nonrandom variation starting in December 2017. The next steps on the improvement process will focus on bringing more stability to the process to assure goal is met consistently.

Conclusions: Our initial data shows that algorithm use in our 0-2 year old population has resulted in improved standardization of interventions, more consistent use by different providers in different care settings, and increased percentage of patients with W/L >50th %ile. This tool enables us to improve the nutritional outcomes of all patients in a consistent and efficient manner as our center grows. Additionally, our patients and families are beginning to see greater emphasis on growth and nutrition and their importance to overall CF care and outcomes.

Acknowledgments: Thank you to the LCH CF team and LCH QI team. Background: Sweat chloride testing is required to diagnose cystic fibrosis (CF), however, obtaining sufficient amounts of sweat for the measurement of chloride is a challenge that plagues many CF centers. Benchmarks have been set to attain a quantity not sufficient (QNS) of <5% for patients >3 months (LeGrys VA, et al. J Pediatr. 2007; 151:85-9) and <10% for patients ≤3 months or younger (LeGrys VA, et al. J Pediatr. 2010; 157:1035-7) . At our institution, use of the Gibson Cook methodology in prior years resulted in a QNS rate of <3%, however, with a recent change to the Macroduct Sweat Collection System the rate of QNS results increased to 15%. In order to reduce this gap in outcomes, the laboratory and CF center personnel collaborated through quality improvement processes.

Objectives: Reduce sweat chloride testing QNS rates to guideline based values.

Changes implemented included: re-education of personnel, limiting the number of staff performing sweat collection, retraining staff regardless of experience, holding weekly educational huddles where best practice for sweat collection and weekly QNS including causes and solutions are discussed, and increasing the frequency of direct observation of the collection process by leadership. To ensure that the sweat collection process was optimal, the following process changes were implemented: blotting the skin between electrodes with gauze before stimulation to avoid short circuiting current flow, use of photos to emphasize correct placement of pilogel disc on the inner volar surface of the arm, and wrapping the Macroduct collector with a disposable stretch bandage during the 30 minutes collection. Additionally, collaboration between the lab and CF center identified several risk factors for sweat QNS based on CLSI recommendations to use as a guide to identify inpatients with sweat orders who were at increased risk for sweat QNS. Ordering providers were encouraged to obtain a pulmonary consult prior to rescheduling the patient for sweat collection. Results from sweat tests were tracked prior to implementation for 8 months and 8 months post-implementation, with data reviewed by both lab and CF center personnel. Comparisons between proportions was done using Fisher's exact test.

Results: Pre-implementation, a total of 269 patients had sweat tests (43% ≤ 3 months) with a QNS rate of 15% for all ages (21% ≤ 3 months; 11% > 3 months). Post-implementation, a total of 196 patients had sweat tests (44% ≤ 3 months) with a QNS rate of 5% for all ages (10% ≤ 3 months; 1% > 3 months). QNS rates decreased from 15% pre-implementation to 5% post-implementation for all ages (p<0.003). Similar reduction occurred in infants ≤ 3 months, from 21% to 10% (p=0.0354) and patients >3 months, from 11% to 1% (p= 0.0006).

Conclusions: Collaborative work by the CF center and lab led to implementation of several change ideas that demonstrated an effective reduction in sweat QNS rates. These change ideas can be similarly utilized in other institutions where higher than acceptable QNS rates are observed. 1 1. Medicine, Pulmonary, Loyola University Medical Center, Maywood, IL, USA; 2. Respiratory Care, Loyola University Medical Center, Maywood, IL, USA; 3. Social Work, Loyola University Medical Center, Maywood, IL, USA; 4. Nutritional Services, Loyola University Medical Center, Maywood, IL, USA Background: The adult CF program at Loyola University Health System manages about 100 patients, 50% of whom are post-transplant. Therefore, we see a complicated population of patients, many of whom have end-stage lung disease. Our lung function averages are below national averages, utilization of chronic pulmonary therapies has been lower than expected without a clear reason, and less than half of our patients thought that airway clearance was working. Feedback from patients suggested that sometimes they did therapies because they have been told to, as opposed to doing them because they thought they helped. As part of the FUN LLC 2 quality improvement program, we decided to implement a modified version of the ReACT airway clearance education program, first described by Zani (BMJ Qual Saf. 2014; 23:i50-5) . In order to have time in clinic to have 30-minute education sessions, we decided we would need to improve clinic flow first.

Methods: A timeline monitoring tool was implemented to improve clinic flow. This was then adjusted to become an in-clinic communication tool. 29 patients then completed 30-minute clinic educational sessions on airway clearance techniques, treatment order, nebulizer care. Patient assessment of time spent doing airway clearance (minutes per week), comfort with airway clearance (scale of 1 to 10) and nebulizer care (scale of 1 to 10) were assessed before and after educational sessions. Later motivational interviewing was incorporated to address adherence problems.

Results: Although clinic visit time increased with the addition of the educational sessions, downtime did not increase and staff satisfaction improved. Self-reported time spent doing airway clearance improved by 18.9% (from 270 minutes per week to 313 minutes), comfort with airway clearance improved by 14.6% (from 8.2 to 9.4 on a scale of 1 to 10) and comfort with nebulizer care improved by 12% (from 8.6 to 9.4 on a scale of 1 to 10). There was a trend towards improvement in FEV1 in those patients completing the program (FEV1 increased from 64% to 68% predicted after the sessions).

Conclusions: The FUN LLC 2 program provided the tolls necessary to organize a multidisciplinary group to work together, collect data, propose solutions and implement changes. Our changes improved clinic workflow and staff satisfaction and allowed time for 30-minute educational sessions. These education sessions improved patients' assessment of comfort with airway clearance and nebulizer care. Adherence to airway clearance is a complicated and multifaceted problem, of which a patient's knowledge deficit is only one small part.

Varghese, P.A. 1 ; Sharma, P. 2 ; Sathe, M.N. 3 ; Scott, A. 1 1. Cystic Fibrosis, Children's Medical Center of Dallas, Carrollton, TX, USA; 2. Pulmonology/Cystic Fibrosis, UT Southwestern, Dallas, TX, USA; 3. GI/Cystic Fibrosis, UT Southwestern, Dallas, TX, USA Problem: Co-production is a novel concept consisting of patients and families working with the CF team to develop information and expertise. It allows families and patients to actively engage in the care and processes within the center. Productive clinic visits involve pre-visit planning on the part of the CF care team as well as patients and families. Successful co-production will increase patient satisfaction and improve health outcomes.

Assessment: Prior to participation in the CF Learning Network, our pre-visit planning process included weekly patient review to prepare for the following week. Electronic patient review data sheets were filled out by each discipline on Mondays and Tuesdays, then printed for use in Wednesday meetings. Patient review sheets were utilized to ensure annual labs, chest x-ray, oral glucose tolerance tests, and DEXA scans were completed. Review sheets were placed in the patient's chart on the day of clinic for provider to utilize. While this process worked efficiently and effectively for the team, we recognized a lack of input and collaboration from patients and their families. Therefore, we aimed to increase the involvement of families in pre-visit planning from 0% to 25% by September 30, 2017. This goal was met by December 2017. Our new goal for 2018 is to have 75% pre-visit participation by December 2018. To more successfully meet the annual goal of 75% participation, monthly review of participation rates will be monitored. Our monthly pre-visit participation goal has been set at 50%.

Interventions: We proposed to manually send out a pre-visit questionnaire via MyChart message 10 days in advance to get family input for their clinic visit. First steps included educating and enrolling families onto MyChart as well as educating on co-production and pre-visit planning. Next, a concise and direct Pre-Visit Questionnaire was co-produced with our family partner and Patient Advisory Council. Our plan-do-study-act cycles in 2017 showed that MyChart yielded low response rates and therefore distribution was expanded to include mail and phone surveys.

Outcomes: At the end of 2017, 176 out of 538 questionnaires were completed. We had a 32% response rate per month in 2017. Phone surveys yielded more responses. However, due to staffing changes, phone surveys and manually sending out the questionnaires were discontinued. As of April 2018, the questionnaires are being linked to scheduled appointments and send automatically via MyChart. 100 questionnaires have been sent out and zero has responded. We conclude this is due to families not being accustomed to the questionnaire link with appointments. Education regarding this change will be disseminated via our clinic quarterly newsletter as well as the Parent Advisory newsletter.

Conclusion: Feedback from families who have participated in pre-visit planning has been overwhelmingly positive. Time study completed in 2017 demonstrates patients who participated in patient review had an average reduction of 35-40 minutes in their total clinic time. We will explore the idea of a dedicated voicemail for pre-visit planning and continue to find the most suitable method of pre-visit planning with our families. 579m Introduction: Online patient portals have been increasingly used to support communication between patients/families and health teams. These portals have been well received by parents of children with chronic conditions. Research into patient portals in pediatric care has explored their ease of use for families. However, little research has been conducted specifically in families with CF. The aims of this study were to 1) assess at baseline and 10 months after implementation the attitudes of families about how patient and carer-facing applications (apps) impacts communication; 2) to explore the experiences of patients/families with CF communicating with hospital staff using the My Health Memory (MHM) app. Its initial functions included syncing appointments with the native phone calendar, with links to information about the appointment and telehealth app when requested, rescheduling appointments, messages to allied health and nursing, and storage of communication in the patients' electronic health record held at the hospital and the family's phone.

Methods: 28 parents of 33 children with CF attending a busy tertiary children's hospital completed an online survey (2 parents who were approached declined to participate citing being "too busy"). The 24-question survey administered 10 months post-implementation asked parents about their experiences of the MHM app, and demographics information about them and their child/ren. Responses to open-ended questions were thematically analysed and descriptive statistics applied.

Results: Most participants (96.4%, N=27) reported using the app at least once a month. Overall, 96.4% (N=27) participants reported the app as being helpful or very helpful. Features participants highlighted as helpful included the convenience of messaging communication (66.7%, N=14), appointments tracking (23.8%, N=5), being a central point of information and contact (9.5%, N=2), phone notifications (4.8%, N=1), and access to discharge summaries (4.8%, N=1). Suggestions for improvement centred around the technical challenges experienced by 47.6% (N=10) of the 21 respondents to this question. Future-looking suggestions included access to test results and medical records (23.8%, N=5), and the ability to send images (9.5%, N=2) . Seventeen participants reported that the app had changed their interactions with their child's care team. Examples include the app making it "easier to communicate with treating staff" (70.6%, N=12) leading parents to "feel more connected and very well supported" (17.6%, N=3), and greater confidence in speaking with hospital staff as they felt they were not intruding on the other person's work.

Conclusions: A patient portal enabling parents of children with CF to manage their child/ren's appointments and communicate with hospital staff was well received. Despite nearly half the participants reporting technical challenges, most parents stated that the MHM positively changed their interactions with hospital staff. This lends further support to the use of patient portals to improve care experiences and empower patients/families in a busy tertiary CF clinic.

Paranjape, S.M.; Peeler, D.; Mogayzel, P.J. Pediatric Pulmonology, JHU/APL, Baltimore, MD, USA Background: Children with cystic fibrosis (CF) often report poor sleep, increased daytime sleepiness and fatigue. Prior studies have not consistently demonstrated obstructive sleep apnea (OSA) or sleep disordered breathing (SDB) in CF. Polysomnography (PSG) has mainly been performed to assess hypoxemia or the need for noninvasive ventilation in severe, end-stage lung disease. Treatment of OSA has been shown to improve nutritional status in CF (MacDonald KD, et al. Respir Care. 2009; 54(12):1727-31) . We previously demonstrated altered respiratory patterns during sleep, particularly increased inspiratory flow limitation (IFL) and respiratory rate, in children with CF compared to healthy controls that are independent of age, BMI, and lung function (Paranjape S, et al. Pediatrics. 2015; 136(5) :920-6).

Purpose: The purpose of this study was to assess sleep in children with CF in the outpatient setting using a validated screening instrument.

Methods: We used a 6-item OSA screening tool (Kadmon G, et al. Int J Pediatr Otorhinolaryngol. 2013; 77(9):1461-4) in all of the outpatient clinics (excluding sleep clinics) of the Johns Hopkins Pediatric Pulmonology Division. A cumulative score > 2.72 indicates high risk for OSA. We surveyed 419 children and families between December 2017 and May 2018.

Results: One hundred sixty-three respondents with CF (39%) ranged in age from 2 months to 21 years. High risk cumulative scores > 2.72 were noted in 7 CF respondents (cumulative score range 2.72-3.75). These 7 respondents ranged in age from 18 months to 14 years. Only two had undergone prior PSG and none had prior ENT evaluation; one was pancreatic sufficient. Among all children with high risk cumulative scores (32/419, 8%), median (range) BMI z-score was lower in children with ]) compared to children without CF (1.59 [-1.54-2.75]) .

Conclusions: We have demonstrated that a simple, validated screening tool can be deployed in a busy clinic to identify children with potential sleep disordered breathing. OSA and sleep disordered breathing are prevalent but underdiagnosed conditions that could potentially lead to worse clinical outcomes in CF and other pediatric chronic lung diseases, particularly with respect to growth and nutritional status. In children with CF, regardless of lung disease severity, there is a need for effective and efficient clinical screening to identify sleep disorders and prioritize sleep assessment.

Acknowledgment: Department of Pediatrics, Johns Hopkins University School of Medicine.

Weiland, J. 1 ; Hente, E. 1 ; Mullen, L. 1 ; Chamberlin, J. 2 ; Siracusa, C. 1 ; McPhail, G.L. 1 1. Pulmonary, Cincinnati Children's Hospital, Cincinnati, OH, USA; 2. Anderson Center, Cincinnati Children's Hospital, Cincinnati, OH, USA The Cystic Fibrosis (CF) Center at Cincinnati Children's Hospital values co-production and family involvement in quality improvement (QI). CF Center leadership asked that a patient/caregiver steering committee be re-established with a focus on QI. As a result, the CF Improvement Partners (CFIP) was created to build a collaborative team of caregivers, patients, and care team members for co-production of QI initiatives.

Team development was facilitated by an institutional "Parent Program Specialist" who sent introductory letters and provided follow-up phone calls to families with a prior history of engagement in CF initiatives. Of the 20 families contacted, 8 caregivers and 1 patient expressed interest in joining the group. The inaugural meeting in January 2018 reviewed the purpose of the team, reviewed core concepts of the family-centered care philosophy and introduced QI concepts (Smart Aims, Key Driver Diagrams and PDSA cycles). A brief review of our current QI initiatives was presented. Family partners were surveyed to determine which initiatives would receive top priority for the upcoming academic year. Our family partners had the strongest interest in mental health screening for caregivers, which we discussed in further detail in a subsequent team meeting. During this meeting, the CFIP team designed a process map to define the first steps for implementing screening for caregiver depression and anxiety. An introductory letter to caregivers describing the screening initiative was also reviewed by the family partners, resulting in significant editing of content, format and our family-engagement approach.

At the inaugural meeting, CFIP team members completed a baseline survey to determine comfort working together and knowledge of QI science. The initial survey was completed by 10 CFIP members: 4 family members, 1 patient, and 5 care team members. On a 5-point Likert scale, 20% of respondents felt their ideas were valued by others some of the time, 60% most of the time and 20% all of the time. In terms of confidence related to doing QI work, 20% felt somewhat confident, 60% very confident and 20% extremely confident. When asked about confidence in leading QI initiatives, 40% were somewhat confident, 40% were very confident and 20% were extremely confident. Free-text feedback in regard to ways that the team could perform successfully included: small breakout teams, collaboration, regular meetings, open and effective communication, success in projects we take on, planning ahead, constant brainstorming and attention toward real problems.

Preliminary observations of CFIP have shown high engagement in co-production, with a significant opportunity for all parties to learn from each other. This year our family partners chose to focus on caregiver mental health screening. Moving forward, we will continue to teach QI methodology, with comfort and knowledge being re-assessed. We will survey the team about the value of working as partners in improving care and outcomes, in addition to measuring process improvements and outcomes related to co-produced initiatives. Weiland, J.; Mullen, L.; Siracusa, C.; Amin, R.; McPhail, G.L. Pulmonary, Cincinnati Children's Hospital, Cincinnati, OH, USA Background: Five years ago at Cincinnati Children's Hospital Cystic Fibrosis Center, we did not have a systematic approach to recognizing and managing declines in lung function. Median FEV1 percent-predicted (FEV1PP) in 6 -17 year olds had dropped from 96% in 2012 to 91% in 2013. We were determined to standardize and ensure the highest quality care. Over the past five years, we have utilized quality improvement (QI) methodology to develop and implement initiatives to improve our management of FEV1 decline and, ultimately, lung function outcomes.

Methods: The first step in our journey was to standardize care. A multidisciplinary cystic fibrosis (CF) team developed an FEV1 Decline Algorithm. Once established, we created a Key Driver Diagram (KDD) to guide implementation and tracking of improvements. Utilizing QI methodology, our focus was on early recognition of decline in lung function. Patients with a 10% or greater decline in lung function from their best in the past year were put into the "red zone" category. Our SMART Aim was to decrease the percent of patients with lung function in the red zone from 16.1% to 15% over the first year. Our journey to improve outcomes has included a bundle of initiatives. Using the KDD as a guide, we began with weekly emails to the care team with a list of patients who had fallen into the red zone. The team reviews patients against the algorithm to be sure that all steps have been taken to return them to their baseline best. A checklist is used in team meetings to ensure that each key aspect of care is discussed. Along with our Infectious Disease, Microbiology and Pharmacy colleagues, we created an Antibiotic Guideline which provides a first and second line choice of antibiotics, along with dosing recommendations, for each of the common CF bacteria. We implemented a process to involve primary CF physicians into the inpatient care of patients, utilizing an admission goal sheet and an inpatient care guideline.

Results: In 2015, after implementation of the bundle there was a decrease in the percent of patients in the red zone from 16.1% to 11% with an increase in median FEV1PP in 6 -17 year olds to 100%. In 2016, however, we saw our trend begin to reverse. The percent of patients in the lung function red zone increased to 24%. Our median FEV1PP for 6 -17 year olds declined to 97%. With the algorithm as a guide, we identified a gap, determining that patients were not consistently being asked to follow-up in a timely manner. With focus on this, we tracked how often providers asked patients new to the red zone to return for follow-up within 28 days of the decline and treatment. We have seen this improve from 68% to 90%. Actual return in 28 days increased from 50% to 88%. Red zone for lung function, as of April 2018 was at 10%. We await 2017 registry results for median lung function.

Discussion: Implementation of a bundle of initiatives has led to improved outcomes for lung function in our CF center. Tracking results of each initiative on a run chart and following QI methodology is key to achieving and sustaining improvements.

Henter, E.; Harmelink, J. Pediatric Pulmonology, Univeristy of Wisconsin, Madison, WI, USA The Pediatric Cystic Fibrosis (CF) QI Workgroup is composed of clinic staff in the University of Wisconsin Pediatric CF Center and a parent representative. Through meetings with the parent representative and Quality Data Management (QDM) survey results, concerns were expressed about the room cleaning process in the Pediatric CF Center. QDM surveys are sent quarterly to families by the Cystic Fibrosis Foundation to provide feedback to CF centers for quality improvement purposes. With the information gathered through the QDM surveys and our parent representative, a discussion took place with clinic management regarding current processes in place for training staff on room cleaning. From these discussions, it was determined that no standardized process was currently in place. Two members of the Pediatric CF QI Workgroup worked with clinic management to measure and help create a standardized training process on room turnover.

The purpose of this project is to determine how well clinic rooms are being cleaned between patients and to standardize the process in which clinic staff is trained regarding room cleaning.

In collaboration with Environmental Services, a product was identified that would allow measurement of surface cleaning. DAZO ® dabbers, filled with florescent marking gel, were dotted onto high touch surface areas. This dot was not visible to the naked eye. After a surface was cleaned, a black light could be used to assess if dots were removed, and therefore how well the surface was cleaned. A control cleaning was completed by auditors to verify effectiveness of this measurement method.

Results were calculated by how many fluorescent dots remained from the total number of dots applied. Three rounds of initial audits were completed with a passing rate of 20%. Staff were educated by Environmental Services through a formal, established program on room cleaning. After staff went through the education process, another round of auditing was completed six months later with a passing rate of 80%, a 60% improvement over initial audits.

When a standardized program was implemented to educate staff on proper techniques for room cleaning, the quality and efficiency of cleaning significantly improved. Moving forward, we plan to work with clinic management and clinical nurse specialist on implementing a standardized education protocol for room cleaning. In addition, we will continue quarterly audits to maintain staff room cleaning competency.

*DAZO® is a registered trademark of Kleancheck Systems LLC and is licensed exclusively to Ecolab for unrestricted use in healthcare facilities. The CF Foundation guidelines for patient follow-up in cystic fibrosis (CF) clinic provides a framework for appropriate treatment and chronic care for CF patients. The guidelines recommend at least 4 clinical care visits, 4 respiratory cultures, and 2 pulmonary function tests per year. The Pediatric CF Program at Vanderbilt has averaged anywhere from 60-83% compliance with these guidelines. After reviewing our outcomes from the Center Specific Report from 2016, we chose to address this outcome as the focus of our participation in the FUN LLC3 program.

Methods: Through the FUN LLC3 Program, we evaluated our current system through the development of the 5Ps (Purpose, Patients, Professionals, Processes, Patterns). We created a flowchart of our current process for appointment scheduling and follow-up in clinic, as well as brainstormed methods to improve our current outcomes. Brainstorming included the use of affinity mapping, which allowed for the less vocal team members to participate more freely. We developed plan-do-study-act (PDSA) cycles which have been adjusted and are ongoing to improve our outcomes.

Identification of patients who were not currently meeting guidelines revealed that the patients fell into 2 main groups: 1. those for whom social concerns have been raised, and 2. those for whom follow-up appointments were scheduled at 13-16 week intervals rather than 12 weeks. We developed several PDSA cycles to aid with the second group of patients since ongoing social work interventions were occurring for the former group. PDSA cycle #1: educate the front desk and scheduling staff regarding the follow-up requirements for CF. PDSA cycle #2: ensure providers prescribe 12-week follow-up rather than the less clear 3-month follow-up. PDSA cycle #3: develop master clinic list for documentation of prescribed follow-up during CF clinic. PDSA cycle #4: reconcile master list with actual scheduled follow-up and reschedule if necessary.

Results: This quality improvement (QI) work is ongoing and will be a continuous QI effort for our program indefinitely. Providers are prescribing 12-week, instead of 3-month follow-up regularly. The master clinic list completion has increased from <50% to 100% with incorporation of the list into the QI meeting agenda. P chart of correct patient follow-up scheduling reveals a mean of 56%, however we anticipate a shift in the mean in the next few weeks with 4 points above the mean at the time of this abstract. The number of patients scheduled outside of the prescribed time-frame declined from 5 to 2 per clinic session.

Discussion: This QI project is ongoing with many barriers and limitations to appropriate follow-up scheduling for CF patients. Future efforts include incorporation of educational materials developed by the family advisory board regarding the importance of regular follow-up in CF clinic. Taketani, E.A.; Vanderbilt, N.; Gravelle, J.; Rushing, S.; Sokolow, A.; Driskill, C.; O'Connor, M.; Thomas, C.S.; Nevel, R.; Brown, R.F. Vanderbilt University Medical Center, Nashville, TN, USA Introduction: Adherence with chronic medical therapies in a disease such as CF declines with increased complexity of the daily maintenance treatment. Many of our patients are started on inhaled corticosteroids (ICS) at a young age due to symptoms and/or family history of asthma. However, these therapies are never reassessed to determine if they are needed as patients age. Based on 2015 Center Specific Report data, our rates for ICS use were one of the highest in the US.

We developed a quality improvment (QI) project aimed at reducing the use of ICS in patients without a diagnosis of asthma. Through our QI team, we developed multiple plan-do-study-act (PDSA) cycles to address this rate. PDSA #1: Verify the authenticity of the data and correct mistakes in the data collection/reporting. PDSA #2: Develop algorithm for discontinuation of ICS. PDSA #3: Implement algorithm which included tracking of patients who had to restart ICS. Data for algorithm success were collected and graphed in a P Chart taking into account autocorrelation of these data.

Results: We started in 2015 with a rate of 52.9% of our patients on ICS without a diagnosis of asthma. In-depth analysis of the registry entry process revealed that the data entry person was including intranasal steroids as ICS. Correction of this mistake resulted in a decline in 2016 registry data to 31.3% on ICS prescribed therapy however no asthma diagnosis documented; yet despite this decline, our center remained above the national average for 2016 (27.4%). An algorithm was developed and implemented in late 2016 and continued through 2018 for a trial of discontinuation of ICS therapy. This algorithm included questions about filling of ICS, need for systemic steroids in previous 2 years and PFTs at time of consideration to discontinue ICS. By 12/31/2017, the rate of ICU use had declined to 8%, and the current rate for May 2018 is 5% of patients with CF and no diagnosis of asthma who are treated with ICS. Due to the autocorrelation of the data, a rate below the lower control limit (4.78%) would be needed to show a significant decline which would be anticipated over the next few months. Two patients had to restart ICS during implementation of the algorithm and were subsequently diagnosed with asthma. No hospitalizations or systemic steroids were required for either of those patients.

Conclusions: The use of this algorithm has significantly diminished the use of ICS in patients without a diagnosis of asthma in our CF program at Vanderbilt without significant adverse events from discontinuation of the ICS. Future steps include the development of a similar algorithm for other nonessential CF medications such as proton pump inhibitors. should be used to assess activity. We use the Habitual Activity Estimation Scale (HAES), a validated questionnaire, to measure self-reported physical activity levels at home. Additionally, patients complete the Cystic Fibrosis Questionnaire-Revised (CFQ-R) to evaluate quality of life. We aimed to assess the value of physical activity self-report in relation to perceived quality of life to inform future areas of improvement in our clinical program.

Objective: To determine if increased physical activity levels (HAES) correlate with CFQ-R physical, health perception, and respiratory domains for patients enrolled in the SCH CF Exercise Program.

Methods: Patients in the exercise program are ages 5 years and older, complete both HAES and CFQ-R, and meet with the athletic trainer who evaluates metrics of endurance, core strength, balance, coordination and agility then sets goals to target areas of improvement. One data set (HAES, CFQ-R) per patient was reviewed from their most recent exercise visit. Through HAES, patients self-report activity in four levels of intensity (inactive, somewhat inactive, somewhat active and active) on a typical weekday and weekend day. We defined "activity" as somewhat active plus active time, and compared reported activity on weekday and weekend by Wilcoxon test. We used Spearman rank correlation to assess the association between CFQ-R specified domains and hours per day of activity, using a weighted average of weekday and weekend.

Results: From 10/2015 -4/2018, 62 patients were eligible. Patients had a mean age of 12.2 years (SD 4.2), mean FEV1 percent predicted of 97.5% (SD 15.4%), mean BMI percentile of 51.4 (SD 26.3%) and were 55.7% female. Participants reported higher median hours of activity on weekday versus weekend (8.2 hours [IQR 6.4, 9.8] vs 7.0 hours [IQR 5.6, 8.9] , p < 0.001). HAES-reported hours of activity moderately correlated with the CFQ-R physical (ρ = 0.430, p = 0.001) and health perception domains (ρ = 0.403, p = 0.002), but not respiratory domain.

Discussion: Patients reporting more hours of activity significantly correlated to improved quality of life scores in both physical and health perception domains. We will target individuals with less time spent being active to make exercise recommendations with the goal of improving quality of life, and tailor exercise recommendations to increase weekend day activity.

Introduction: A recent James Lind Alliance Priority Setting Partnership identified the top 10 questions for clinical research in CF (Thorax 2018; 73:388-90) . The first of these was: "What are effective ways of simplifying the treatment burden of people with CF?" To gain a deeper understanding of this, we conducted further research with the lay and professional communities. Our objective is to develop an outline for a clinical trial that aims to simplify treatment burden.

Methods: This work is led by a steering group, representative of the CF community (both lay and professional). A questionnaire was produced using Surveymonkey TM , open for 4 weeks in March-April 2018. We promoted the survey via Twitter TM (@questionCF), professional networks, the UK CF Trust and NIHR. Data were analysed by descriptive statistics.

Results: We received 941 survey responses: 189 (20%) from people with CF (PwCF); 394 (48%) from family; and 300 (32%) from health professionals. PwCF reported a median of 10 (0-52) current treatments, taking 2 hours (0-13h) per day to complete; 30 (10%) were on IV antibiotics. We had responses from 21 countries, with 390 (87%) from the UK and 31 (7%) from USA. There was broad consensus between lay and professional respondents over the most important and most burdensome treatments (see Table) . We asked health care professionals: "With the advent of CFTR modulators it may be possible to stop or reduce some existing treatments for those patients taking these drugs. Would you support a stopping trial if this was to be carried out?" and 129 (78%) said they would.

Conclusions: There is broad consensus between lay and professional communities over the most burdensome treatments. There is support among professionals for a trial of stopping existing treatments for those on CFTR modulators. We will use this methodology to explore other priority research questions and the approach may be useful in other conditions. Acknowledgments: Supported by the CF Trust and Univ. Nottingham. The CF Learning Network (CFLN) is a collaborative network created to improve health, care, and time for people with cystic fibrosis (CF). An integral part of improving care in the CFLN is to co-produce with patients and families impacted by the disease. Co-production is the collaborative communication between lived experience of CF and those with CF knowledge expertise. Centers meet and discuss patients coming to a clinic visit and formulate a plan of care without any involvement of patients/ families. Pre-visit planning (PVP) serves to take the experience of people with CF to the care teams prior to the upcoming clinic visit. In CFLN, our team began the implementation of the PVP process in November 2016. The aim of the study was to empower adults to share their concerns and needs in a collaborative manner prior to the clinic visit. The electronic patient portal (EPP) system became active in January 2016 and provides documentation in the medical records. A memo encouraging clinical communication to be conducted via the EPP was sent electronically via mail and reinforced at each clinic visit by team. The PVP message was developed in collaboration with the Partner Advisory Board and delivered through EPP. The PVP message included the specific team members the patient wished to communicate with, topic to discuss or specific question to ask a team member. The PVP responses are reviewed in the weekly pre-clinic meeting and a personalized care plan is developed. After a test phase, data collection began in August 2017. Data points included date PVP was sent, date patient responded, did the patient attend the clinic visit, type of clinic visit, time patient arrived in the clinic and time patient was discharged from the clinic, and lastly, which interdisciplinary members participated in the clinic visit. The number of patients scheduled for clinic visits from August 2017 to April 2018 were 935. The total PVP sent out was 681. 223 patients responded to the PVP (32.3%). The data were stratified to examine patients who responded to the PVP and attended the clinic visit (155 patients) or responded to the PVP and did not attend clinic (68 patients). 336 patients attended clinic and did not respond to the PVP vs 122 patients who did not respond to the PVP and did not attend clinic. The average time to response to the PVP was 11.1 days. Team members (MD, RN, RD, SW, RT, Pharm.) participated in the clinic visit approximately 82% of the time except for the dietitian due to staff changes (52%). Adults who responded to the PVP spent an average of 158 minutes versus patients who did not respond to the PVP who had an average time of 146 minutes. CF adults desire to be more engaged and their concerns/needs to be actively discussed at clinic visits. Subjectively, the team felt empowered to collaborate with the patients who responded to the PVP including co-creating care plan. Use of the EPP serves as a good platform for PVP. Potential barriers include lack of access, smart phone capability, and health literacy. Our experience has shown that adults with CF selectively choose whether to respond given their perception of clinical importance. McPhail, G.L.; Mullen, L.; Weiland, J.; Wimmers, H.; Selker, N.; Chapman, M.; Siracusa, C. Pulmonary, Cincinnati Children's Hospital, University of Cincinnati, Cincinnati, OH, USA Background: Incorporating families' priorities into pre-visit clinic planning (PVP) facilitates partnership and co-production. Since 2001, the Cincinnati Children's Hospital Medical Center (CCHMC) CF team has worked to incorporate the voice of the family in PVP. The initial approach was a paper survey, developed by families and clinicians, mailed to families with a self-addressed stamped envelope in which to return a completed survey. The response rate was low and the project was discontinued. The second approach was to send a survey for families to access via the family portal of the electronic medical record (MyChart). The response rate was low at 5%, possibly because "pre-visit survey" could not be included in the subject line of the message sent to families or because the volume of MyChart messages received by families was high. HIPAA and hospital regulations were barriers to using email and online formats to improve PVP response rates. In 2016, the CCHMC CF team joined the CF Learning Network. We learned new ideas to facilitate co-producing PVP with families.

Objective: To better include families' priorities into PVP. Our aim was to increase the percentage of families involved in PVP from 0% to 80% from March to September 2018 with the utilization of the MyChart app and unencrypted email.

Methods: We designed a Day of Visit (DOV) form with our family partners. We did n-of-1 testing in several CF clinics to get family feedback to revise the DOV form. The DOV form was given to the families upon arrival to clinic. The form listed patient outcomes, risk stratification for FEV1, BMI and CF-related diabetes status, tests that would be discussed or completed at the visit, and team members that were expected to see the family that day. The DOV form had a section for families with the question "What matters to you?" The goal was to incorporate families' priorities into the clinic plan at the point of care. With our family partners and hospital administration we then designed an electronic communication agreement that would allow us to send the DOV form via unencrypted email in advance of clinic visits in the future.

Results: 100% of our patients and families are now receiving the DOV form when they arrive in clinic. The family can write down "What matters to you?" for that visit, and these priorities are addressed at the point of care by the CF team. With the implementation of the MyChart app and unencrypted email, we hope to have input from the patient and families far in advance of our own weekly CF team clinic PVP meetings.

Discussion: Incorporating families' priorities into PVP is important for family activation and engagement. Barriers exist to prevent families from participating readily in PVP. Our families found value in our PVP form that shared patient outcomes, the plan for the clinic visit that day, and elicited their priorities for the visit that day. Using unencrypted electronic communication, we hope to incorporate the families' priorities into our team PVP far in advance of clinic. This will allow us to design clinic visits to focus on families' areas of interests and needs and promote family activation and engagement. Hogan, C.; Sam, S.; Kleinhenz, M. Department of Medicine, University of California, San Francisco Medical Center, San Francisco, CA, USA Introduction: Data from the UCSF 2016 Center Specific Report indicated 45% of our center's adult cystic fibrosis (CF) patients achieved the standard of four visits with the care team, one sputum culture, and 2 pulmonary function tests (4V+1S+2P). This was below the national average of 53.6% and lower than the center's 2015 performance, 54%. Review of our clinic identified opportunities to improve processes of care: address available appointments, tailor communications with patients to their preferences, and control lag time between providers during multidisciplinary care team visits. Reports from other CF care centers inspired the idea for a "clinic quarterback (QB)" to coordinate these tasks.

Objective: Increase the proportion of patients with 4V+1S+2P to at least 54% by the end of 2018.

Methods: The patient roster was reviewed to determine which patients required 4V+1S+2P and which patients had fewer than four visits in the preceding year. The available appointments were inventoried and when compared with the number of appointments needed to serve our patient roster, found to be inadequate. After a feasibility study and training program, the center added an additional adult CF clinic session with a full care team and trained CF providers. A part-time research assistant was hired as the "clinic QB" to manage patient relations and clinic flow. The clinic QB tracked scheduling and communicated with patients using a variety of methods: telephone calls, text messages, and Tele-Health sessions. The clinic QB enabled a reconfigured appointment template, a personalized visit itinerary for each patient vetted in the pre-clinic huddle, quick, smooth transitions between care team members, and a process to wrap-up the visit including scheduling future appointments.

Results: In 2016, the center had 284 available appointments when 363 appointments were needed for guideline care. The second clinic session opened in winter 2017 to close this gap. Year-end data from 2017 showed clinic expansion did not produce the desired results: only 50% of patients achieved 4V+1S+2P. 61 patients were surveyed about communication preferences: 44 (72%) preferred text messaging. All six patients offered Tele-Health visits declined in favor of in-person visits. Staggering appointment slots, organizing visit itinerary for each patient, and directing the sequence of visit activities yielded an average appointment of 83 minutes. With the clinic QB cuing provider transitions, the aggregate lag time between care team members averaged 10 minutes, 12% of the visit. During the observation period, each exam room was used once; data do not reflect 30 minutes required for infection prevention and control practices if fewer rooms were available. Assuming patients attend all scheduled appointments, year-to-date data suggests 65 patients (68%) will complete 2 visits with the care team by the end of quarter 2 and are on track to achieve guideline care.

Conclusion: Tailored communication strategies and personalized agendas for each clinic visit are ways to partner with patients in achieving guideline care. Introducing a "clinic QB" created new ways for the care team to manage patient relationships, optimize clinic flow and improve attendance in the adult CF center. Jue, V.; Hogan, C.; Kleinhenz, M. UCSF Medical Center, San Francisco, CA, USA Background: Cystic fibrosis (CF) patients use proton pump inhibitors (PPIs) for gastroesophageal reflux disease (GERD) and/or to create a basic gastric environment to optimize absorption of pancreatic enzymes. PPIs or histamine 2 receptor antagonists (H2RAs) are recommended agents to create a basic gastric environment. PPIs are prescribed for 51% of CF patients (2016 CF Foundation Patient Registry) . Evidence suggests that PPI use >1 year can lead to an increased risk of bone fractures, pneumonia, C difficile infections, and malabsorption of vitamins (D Johnson, et al. Clin Gastroentero Hepatol. 2013; 11:458-64) . H2RAs are an alternative with fewer potential complications. Chronic PPI use may contribute to potential effects on microbiology and bone health. Guidelines from the American College of Gastroenterology recommend an 8-week course to treat GERD. Literature encourages clinicians to attempt to de-escalate or discontinue use of PPI. Primary objective of this quality improvement study is to educate our center's patients on the risks associated with chronic PPI use, examine the rationale for initial PPI prescription, and de-escalate or discontinue PPI therapy when appropriate.

Method: The initiative began in November 2017 at University of California San Francisco (UCSF) Adult CF Center (109 patients). The pharmacist identified patients prescribed PPIs. During quarterly visits the dietitian collected baseline gastrointestinal information from each patient using the iCan GI Symptom Tracker (2017 AbbVie, Inc). Patients were asked about the PPI indication, duration of use, medications failed, lifestyle/ diet changes, and side effects. Patients were educated on the potential side effects of chronic PPI therapy and asked one of the following: willingness to discontinue, decrease dose, or switch to a H2RA. Those patients that did decrease dose or switch to a H2RA received a follow-up telephone call from the pharmacist at 1 week, 1 month, 3 months, and 1 year. Repeat completion of the iCan GI Symptom Tracker occurs at subsequent quarterly visits. Lung transplant, chronic steroid use, and incarcerated patients were excluded. Data are collected on a secured Excel spreadsheet.

Results: To date, 34 patients received medication education on potential chronic side effects of PPIs. Out of the 34 patients, 17 provided baseline GI symptom data. Symptoms are controlled for 15 patients who changed to a H2RA. Symptoms resolved for 3 patients who self-discontinued PPIs after reading about the chronic side effects. There were 9 patients who declined, reporting symptoms returned on the H2RA. Five patients were actively sick. One patient needed both a PPI and H2RA for symptom control. One patient is currently still tapering off of a PPI. For the patients that agreed to try to de-escalate or discontinue PPIs, 1 patient had the 3-month follow-up phone call and all others had at least a 1-month call.

Conclusion: Adults with CF adhere to complex medical regimens. Embedding a pharmacist within the care team provides the opportunity to assess the indications for PPI therapy within the context of adult health recommendations. Decisions about PPI in adults with CF requires patient education and follow-up of health status when a change from long-term PPI is indicated. Introduction: Quarterly cystic fibrosis (CF) patient visits with the CF team ensures early detection of lung function changes and improves outcomes. Our program started a quality improvement (QI) initiative in 2014 to improve patient participation to clinic quarterly. Over the past 4 years, we continued this initiative, and ran multiple plan, do, study, act (PDSA) cycles in order to improve our process. Our team participated in Learning and Leadership Collaborative (LLC) 2015-2016, Virtual Improvement Program (VIP) 2016-2017 and in July of 2017, our adult CF program applied and was invited to participate in the Cystic Fibrosis Learning Network (CFLN). As part of CFLN we aimed to improve patient team communication pre-clinic to better address patient needs and we integrated a new process (pre-visit planning or PVP) in quarterly visits QI.

Methods: Port CF reports are utilized routinely to generate a list of patients due for a 90-day visit. Patients due for 90-day visits are called by the CF administrative assistant (AA) to schedule an appointment. During fall 2017, a PVP survey was designed by the care team with the assistance of our patient partner. Patients who are scheduled to be seen in clinic are called by the AA to confirm their appointments 1-7 days in advance, during these phone calls the AA also administers the PVP surveys. At the beginning of this process, patients were surveyed on their preferred PVP contact method and over the past 2 months, we have begun contacting them via their preferred method of communication. A list of confirmations and all PVP responses are shared with the team via email; data are also stored in a secure server. Each care team member sends an email 1-2 days before clinic with a list of patients they would like to see, the reasons, and desired lab tests. Itineraries are created for the patients according to their PVP answers and their needs.

Results: Our adult center serves 100 patients with CF seen regularly every 4 months. During the past 4 years, the patient percentage seen in clinic 4 or more times a year has increased from 38. 5% in 2013, to 50% in 2014, 66.7% in 2015 and 67% in 2016 according to registry data. We started administering the PVP questionnaires in December 2017 and so far 68 patients have been surveyed, with a 61% response rate. Since changing to patient's preferred method of contact, the patient response has increased by 21%. At the beginning of the year, our program started participating in CF HealthCheck; so far we have enrolled 9 patients. For the next few months, our team will concentrate on combining surveys and questionnaires into one pre-visit message through MyChart. Our organization is requesting all providers change all patient communications to MyChart. This will be another challenge and PDSA for our team. Patients found the pre-visit planning survey helpful (87%) and staff satisfaction with the survey was 100%.

Conclusion: A standardized approach to pre-clinic visit preparation ensures better communication between the team and the patients and is received well both by patients and staff. A standardized process of patient contact pre-clinic and quarterly increases clinic participation.

Gamel, B. 1 ; Noyes, S. 2 ; Seid, M. 1 1. Pulmonary Medicine, Cincinnati Children's Hospital, Cincinnati, OH, USA; 2. James M. Anderson Center for Health Systems Excellence, Cincinnati Children's Hospital, Cincinnati, OH, USA Objective: To describe the role of Patient and Family Partner (PFP) in the CF Learning Network (CFLN), their level of engagement (LOE) in coproducing quality improvement (QI) initiatives with their CF program teams, and factors that might contribute to PFP satisfaction with the coproduction relationship.

Background: The CFLN -29 pediatric and adult care programs and a team of community innovators -is a community of patients and families, clinicians, and researchers who collaborate to improve outcomes for people with CF. Every program involved in the CFLN is expected to have at least one PFP integrated into the program QI team. In addition to being involved in QI at the program level, PFPs attend virtual monthly network webinars and semi-annual community conferences in person. They are supported by a PFP Program, which serves a professional development function, including training, mentoring, and shared learning from PFPs who work well with their program teams. We hypothesized that PFPs with a higher LOE (those who are leading or co-leading QI work) would be more satisfied in their working relationship with their QI team.

Methods: PFPs are surveyed monthly. PFPs are asked to indicate on a scale of 1-7 how engaged they are in the QI work. A "1" indicates they are just learning about the work, while a "7" indicates they are leading or co-leading a QI initiative. PFPs are also asked "how satisfied are you in the working relationship between you and your QI care team?" (highly satisfied, satisfied, neutral, unsatisfied and highly unsatisfied). We interviewed several high-functioning PFP-program teams to understand what factors matter in engagement and satisfaction.

Results: This session will share updated results. In April, 2018 14 of the 57 PFPs responded to the monthly survey. Of these, 12 were actively participating, co-leading, or leading a QI initiative at their program and all 14 were satisfied or highly satisfied with the working relationship with their program team. LOE rating was not related to satisfaction with the working relationship. High-functioning teams told us that the most important factor in a successful, satisfied relationship was the match in expectation of LOE between the QI team and the PFP. PFPs with a lower LOE who are working on QI teams who expect their PFPs to have a lower LOE may be more satisfied with the relationship. This may be the same for PFPs with a higher LOE who are working on a QI team who expects to have a PFP with a high LOE. Lower satisfaction in the relationship may come from asymmetry of QI team's expectation of the PFPs engagement.

Conclusions: PFPs are increasingly involved in QI with CF program teams. Contrary to our hypothesis, we found no relationship between level of engagement and satisfaction. Preliminary evidence suggests that matching involvement to desired level of involvement is important. Other dimensions might also be important. Our next step is to develop a tool that teams and PFPs can use to negotiate these dimensions.

Acknowledgements: Supported by Award SEID15A0 from the CFFT. Introduction: Infection transmission is a notable risk, both perceived and real, in CF clinics. However, infection control practices remain variable and generally unsubstantiated by quantifiable data. We seek to identify practices which minimize risk to the patient population, maximize overall clinic efficiency, and improve patient experience. To accomplish this, we are quantifying key interactions during clinic and simultaneously surveying bacterial populations to create an incisive model of transmission risk.

Methods: Three data types inform our model.

Interactions. BLE beacons provide real-time estimates of proximity among study participants and between participants and objects of interest. These beacons emit a Bluetooth signal at regular intervals and can be tuned to detect interactions between 3-50 feet. Cell-phone applications developed for this study are used to capture these interactions.

2) Participant Surveys. Both patients and clinic staff are enrolling in the study. Participants complete a series of population-specific questionnaires surveying their health history, perceived level of interaction with others, and perceptions of transmission risk. Results are compared to events/risks identified in the final quantified model.

3) Temporal Microbiomes. Swabs (~500/day) of surfaces (i.e. exam tables, stethoscopes) are taken multiple times throughout each clinic day. Participants also provide swabs of nasal passages, throats, and hands. Microbial content of swabs is assessed by DNA sequencing, and analyses are conducted to determine whether transmission of bacteria is correlated with interaction or survey data.

In order to assess viability of BLE beacons as a marker of personal interactions, 6 beacons were placed in clinic for one day. Interactions of one pulmonologist were captured during this time via cell-phone app. Interactions were also recorded independently by a study team member.

Results: Beacons tuned with site-specific parameters were placed in two exam rooms, a staff teaming space, a sanitizing station at patient check-in, the vitals collection area, and the staff's gowning/gloving station. During a 2-hour period of observation, 49 interactions were recorded. The pulmonologist spent 1.2 hours in the teaming space, 12.2±2.6 minutes interacting directly with each patient, and 1.7 minutes at the gowning/gloving station. Beacon-generated results concurred with those recorded by the study team.

Conclusions: We have shown that BLE beacons are a viable means of quantifying real time interactions within the outpatient clinic. Upon study completion in June 2018, ~50 patients and ~25 staff members will have contributed to the interaction model, surveys, and microbiome sampling. Broad implementation of these methods could aid in the development of both generalizable and site-specific best practices for infection control. The University of Minnesota Adult CF Program is a large adult CF clinic with over 400 patients and outpatient clinic that is conducted five days per week. Historically the CF multidisciplinary team conducted pre-visit planning (PVP) during weekly team meetings. The team reviewed the upcoming clinic schedule and discussed how patients' needs would be addressed. The team also decided which ancillary specialists would see which patients during their upcoming clinic visits. In an effort to increase co-production of care, a need for improved PVP was identified. Due to the size of the adult CF patient population, a sustainable and efficient method for conducting PVP with patient input was needed. Utilizing an electronic method of communication incorporated into the electronic medical record (EMR) was desired in order to improve efficiency and enhance team member and patient satisfaction.

Objective: To improve co-production through collaborative pre-visit planning utilizing the EMR.

Methods: Patients were educated about and enrolled in the EMR patient portal, MyChart. Care team members gave verbal and written information about MyChart to patients during their clinic visits. A CF PVP Questionnaire was developed. In the questionnaire, patients are asked about their main priority or concern for the upcoming visit and what they most want to discuss with the team. Patients are also given the opportunity to select which team members they would like to see. The questionnaire includes examples of common topics each team member can address. The PVP tool has been piloted in a small sample of patients. Patient and team member satisfaction levels were measured utilizing surveys.

Results: As a result of this project, the number of patient questionnaires was reduced from six to three and many redundant questions were eliminated. To date, 25% of patients who have received the survey have responded. A 1-5 Likert scale was used to assess patient satisfaction. The statement: "Pre-visit planning improved my communication with my care team" received an average score of 4.1. The statement: "I liked having the option of selecting the CF members I wanted to see today" received an average score of 4.2. The care team's satisfaction with the system received an average score of 3.9.

Conclusion: The EMR can be utilized to enhance co-production of care and include patients in pre-visit planning. The current response rate precedes any patient education. We anticipate that response rate will improve once we have educated our patients about the goals and benefits of the system. Introduction: Individuals with CF are estimated to require 1.5-2 times the nutritional needs of those without CF. A diet high in fat and calories is recommended, with fat contributing about 40% of caloric intake. BMI can be impacted by food and beverage intake. CF patients tend to consume an excessive amount of saturated fat, while having a low intake of polyunsaturated fat. Shifting the focus from high saturated, trans-fat and calorie-dense foods and beverages, to nutritionally-dense foods and beverages, we aimed to provide distinctive nutrition education to make smarter food choices to assist them in reaching the CF Foundation's BMI goals (≥ 22 kg/m 2 for women and ≥ 23 kg/m 2 for men). 41% of our adult CF population was below BMI goal.

Objective: Reduce the proportion of adult CF patients with a below goal BMI while making more sustainable and healthier food choices.

Patients completed a 24-hour dietary recall with all items consumed at meals and snacks, including quantity. A team member then reviewed their results with a STOP-LIGHT System which listed categories of foods based on both calorie and nutrient density: Red (STOP!), Yellow (Caution), and Green (GO!). Red Light foods contain a large amount of calories that contain little or no nutrition. Foods listed in the Green Light category provide high protein and high calories in the form of medium-lean meats, healthy fats, and whole grains. Foods listed in the Yellow Light category fall somewhere in between Red and Green. The patients were shown where a smarter food choice could be made (eg, lemon water instead of soda) and what types of food correlated to the specific colors (eg, candy is "Red"). Patients then set 1-2 dietary goals for making smarter food/ beverage choices. At their next clinic visit, another 24-hour dietary recall was completed. A team member again reviewed the recall with the patient to determine if/where the patient can make further improvements.

Results: 51 patients were enrolled, with 46 educational assessments completed. 22 of the 46 patients were seen at follow-up. 19 successfully achieved a nutrition-related goal they set for themselves. The Figure shows prior to initiating the STOP-LIGHT System, 32 of 78 (41%) patients were below goal BMI whereas after implementing the STOP-LIGHT System, this number was reduced to 22 (28%) below goal.

Conclusions: The STOP-LIGHT System provided an easy and effective tool that complemented visual learning to auditory reinforcement in nutritional education. Adult CF patients and the CF team as a whole participated in improving nutritionally appropriate food choices.

Acknowledgment: This work was conducted as a part of LLC-2 Project.

Szczesniak, R.D. 1 ; Grossoehme, D.H. 2 ; Brokamp, C. 1 ; Su, W. 1 ; McPhail, G.L. 2 ; Seid, M. 3, 2 ; Pestian, J. 4 ; Clancy, J.P. 2 1. Biostatistics and Epidemiology, Cincinnati Children's Hospital, Cincinnati, OH, USA; 2. Pulmonary Medicine, Cincinnati Children's Hospital, Cincinnati, OH, USA; 3. Anderson Center, Cincinnati Children's Hospital, Cincinnati, OH, USA; 4. Biomedical Informatics, Cincinnati Children's Hospital, Cincinnati, OH, USA Introduction: Lung function monitoring is imperative for CF clinical management. Population-level FEV1 decline averages 1-3% predicted/ year but patient-specific bouts of rapid decline are far less subtle and often unpredictable at the patient level.

Objective: To develop a tool (graphical user interface, GUI) for clinicians and patients to use at point of care to forecast and prevent rapid decline.

Methods: We performed a mixed methods study, first using US CF Foundation Patient Registry data on 27,296 patients to establish a predictive model of rapid FEV1 decline. A clinician GUI prototype was developed using RShiny. Clinician focus groups totaling 16 CF care providers were conducted to understand provider perspectives on rapid decline and optimize the GUI for identifying personalized points of rapid decline throughout the clinical course of each patient.

Clinician point-of-care identification of patients experiencing rapid decline tended to be retrospective and framed using linear patterns. Narrative and visual data from clinicians indicated that discussions with patients and caregivers avoid the term rapid decline, but clinicians used the term in team management meetings. GUI as presented (see Figure) was considered useful for pre-visit planning, personalized treatment choices and early testing with new therapies. There was consensus that a patient-specific GUI should differ from the clinician GUI.

Conclusions: This study yielded a revised prototype platform with the potential to serve as a viable point-of-care tool. Future work includes implementation of this clinical prototype, evaluated prospectively under real-world settings, with the aim of improving the pre-visit planning process and monitoring of treatment for CF point of care. This work will include patient-centered design, in order to produce a viable point-of-care GUI for patients.

Online Prediction Tool for Rapid Decline. This patient (186) is an F508del heterozygous male with first available FEV 1 at age 6. 1 years, born 1999-2005 , and baseline FEV 1 of 84% predicted. His FEV 1 loss became more severe with age, and he was at high risk of rapid decline age 12. The gray shaded area shows data used for training the prediction model; the red shaded area covers the two-year forecast period. He had relatively few exacerbations and attended clinical visits routinely. He was not diagnosed with CF-related diabetes, but had used Medicaid insurance and experienced infections with Pseudomonas and MRSA. Wukovich, R. 1 ; Roesch, E. 1 1. Rainbow Babies & Children's Hospital, Cleveland, OH, USA; 2. University Hospitals Cleveland Medical Center, Cleveland, OH, USA Background: Optimal nutrition is associated with better pulmonary and survival outcomes for individuals with cystic fibrosis (CF). However many barriers can make optimizing nutrition challenging, including psychosocial issues/irregular visits, gastrointestinal problems and CF-related diabetes (CFRD). It is often challenging to identify all of the specific barriers during a busy outpatient clinic visit. Our center sought to systematically identify individual patient barriers in our high-risk patients, towards the ultimate goal of improving the nutritional status in our clinic.

Methods: CF pediatric patients with a BMI <30th percentile were targeted for intervention. We developed a structured questionnaire to assess patient/caregiver satisfaction with weight, confidence in weight management, nutritional goals and perceived barriers to optimizing nutrition. Additional questions aimed to identify specific concerns such as medications, supplies and transportation. Duration of interview, next appointment status, concerns/barriers, BMI percentile at time of call, satisfaction with current weight, confidence with weight management and patient goals were documented. If concerns were identified, the interviewer notified the appropriate CF team member.

Results: Our cohort included 23 patients. To date, we have interviewed 11 families either by phone or in clinic (48% success rate). The average duration of each encounter was 15 minutes (range 7-25 minutes). Although 60% of patients/caregivers were confident with managing nutrition, the majority were not satisfied with their/their child's weight. Among patients and families contacted, 30% did not have a follow-up appointment scheduled. To further evaluate adherence to 4 visits per year with subspecialists, we found the 6 patients with CFRD had an average of only 0.67 visits with endocrine in the last year (range 0-2 visits/year). The 5 patients with g-tubes had an average of 1.8 visits with GI in the last year (range 0-3 visits/year).

When asked about specific barriers, responses included access to nutritional supplements, transportation issues, medication costs, blood sugar levels, trouble with tube feeds, appetite stimulants and appropriate food options. All but one patient/family had goals, examples included weight gain, treatment with new modulators and staying healthy.

Conclusions: The development of a structured questionnaire improved our understanding of patient/caregiver perceptions of nutrition and allowed the personalized identification of specific nutritional concerns. Our findings to date indicate that patient/caregiver calls are helpful in addressing issues between clinic visits and with scheduling follow-up appointments. Increased patient contact may contribute to improved weight gain, better treatment adherence and clinic attendance, particularly in multidisciplinary clinics. We plan to assess change in BMI every 6 months to see if our ongoing interventions are improving nutritional outcomes and continue to improve our clinic processes.

Data collection is ongoing and will be updated prior to NACFC. Acknowledgment: Supported by the CF Foundation Fun LLC3. The goal of the CF Foundation's (CFF) Patient and Family Experience of Care survey (PFEC) is to provide a forum for people accessing CF clinics to give feedback that can be used to improve care experience. PFEC is in its 4th year of data collection with 10,947 competed surveys (May 2018) and 88% of the participants were satisfied with the survey process with 23% stated it was "easy," "quick," and/or "thorough." However 17% stated it was "too long" especially the phone survey taking about 20 minutes to complete. Since 56% participate by phone, the goal was to shorten PFEC to reduce the time burden.

Objective: To report the process of revising the PFEC.

The national PFEC steering committee (PFEC consultant, parent and research advisor, CFF leadership, and survey host leadership) met twice monthly for 7 months to develop a revised PFEC. The committee decided the main focus areas for the revised PFEC were infection prevention and control (IPC), mental health, and partnership. The first 2 were guidelines of care and the last was relative to a new initiative, Partnership for Sustaining Daily Care. The committee reviewed PFEC's psychometric statistics such as percent missing, ceiling effect, and variation among CF clinics. The committee also asked for input from the CFF Community Voice in which 96 people participated in the Experience of Care Feedback Survey (October 2017). They were asked whether to keep or remove PFEC questions, whether new questions should be included, and their suggestions of new survey questions.

Results: PFEC has 5 IPC guideline-based questions, which remained on the revised PFEC with 1 new question about clinic environment cleanliness as suggested by the Community Voice participants. To address mental health, PFEC has 1 question about whether the care team asked about mental health, which remained on the revised PFEC. One question was added about whether mental health treatments or counseling has helped. Two validated surveys; CollaboRATE and IntegRATE were added to the revised PFEC. CollaboRATE has 3 questions measuring shared decision making in clinical encounters, which covered suggestions from the Community Voice and partnership concepts. IntegRATE has 4 questions about information sharing, consistent advice, mutual respect, and role clarity, which were similar suggestions from the Community Voice and comments received from the open-ended PFEC questions. An additional question was included about whether the person felt like a valued partner. Several PFEC questions were removed: demographic-type questions since they were too personal (Community Voice and survey process feedback), most of the self-care questions since they showed similar results across questions and little program variation, and most of the CF team questions. The revised PFEC has 30 questions, which has 20 less questions than the PFEC used since 2015. Numerous new questions were vetted by the committee, however most could not be included.

Conclusion: The revised PFEC will continue to provide an opportunity for people to give feedback to the CF clinic and will take less time to complete. 18 and older) ., A recent clinic survey revealed that 89% of our patients (n=36) do not take their inhaled antibiotics as prescribed.

Objective: We wish to identify reasons for low adherence to prescribed chronic inhaled antibiotics in the UT Health East Adult CF Program and to develop an intervention to increase adherence. We aim to improve the use of chronic inhaled antibiotics as prescribed from approximately 11% to 75%.

Methods: Patients received a pre-clinic inhaled antibiotic questionnaire asking the following questions: 1. What inhaled antibiotics are prescribed to you? 2. Do you take them as prescribed? 3. Do you know why you take them? 4. If you do not take them as prescribed, what is the reason (time, insurance, can't remember, too tired, equipment)? Patients then received education on the order of therapy in both a picture and descriptive form handout. Patients also received education on what Pseudomonas is, how it progresses, and why inhaled antibiotics are prescribed. The patient was given a choice of three time management methods: wall/written calendar, phone calendar, or Mango Health app. The CF team tested several medication adherence apps (CF-Buzz, Mango Health, MyMeds, Dosecast). Mango Health App was chosen due to ease of use and the suggestion that electronic applications may help with medication adherence.

Results: Pre-clinic survey revealed 100% (36/36) of the patients knew which inhaled antibiotics they were prescribed. However, only 11% (4/36) of these patients took the inhaled antibiotics as prescribed. Approximately 30% of the patients did not understand why they took inhaled antibiotics. According to 60% of the patients, the primary reason inhaled antibiotics were not taken as prescribed was due to time. When given the choice of a time management method: 42% of patients chose the wall/written calendar, 15% of patients chose the phone calendar, 29% of patients chose the Mango Health App, and 14% of patients declined all methods. Post-survey revealed 100% of the patients know why they are prescribed inhaled antibiotic. A time management tool is currently being used in 66% of patients. The number of patients taking inhaled antibiotics as prescribed increased from 11% to 60%.

Conclusion: Our adult patients were not as educated about their disease process nor understood the reason for taking inhaled antibiotics. Younger patients were more receptive to the use of electronic applications (Mango Health or phone calendar) for adherence to therapy. Older patients were less proficient with electronics and preferred a written calendar. Combination of education and an adherence tool were very beneficial in improving understanding of the disease process and treatment.

Glazer, L.; Hochstadt, R.; Sadeghi, H.; Sewell, W. Pediatric Pulmonary, Columbia Univ. Medical Center, New York, NY, USA Introduction: Cystic fibrosis (CF) clinic visits are lengthy and involve many providers. Past clinic time studies at our center showed mean time of 119 minutes and median of 109.5 minutes with average downtime of 23 minutes. Priorities may vary for team and family. Prior to this project, only a phone call confirming the appointment was made before clinic visit. CF visits may begin before the clinic appointment. The aims of this project were to involve families and patients in their visit before the appointment, to improve efficiency, to prioritize needs of the patients, to increase family input into care (coproduction) and to improve patient satisfaction.

Methods: A quality improvement (QI) team was formed followed by recruitment of patient/family partners. Four plan-do-study-act (PDSA) cycles were completed. The first cycle, in October 2017, involved creating a survey, which was emailed to families one week prior to their scheduled appointment. The survey inquired three (3) specific categories and write-in option for respiratory symptoms, weight and GI symptoms, insurance, and other. The second cycle was from November through December 2017 that included a Monday phone call on the week of the appointment for those who have not responded to the email, Spanish language addition to the survey and placing patient identifier. The third cycle was from January through February 2018 and included incorporation of the responses in the weekly CF team meeting and inclusion of our center's second patient care site. The fourth cycle from March through April 2018 consisted of adjusting survey questions based on feedback from patient/family partners by adding a question on equipment concerns.

Results: The response rate during the first and second PDSA cycles were 20% and 23% respectively. When phone call was added to email, the total response rate increased to 39% and 53% in the third and fourth PDSA cycles. Table below shows that number of responders increased with more survey categories.

Conclusion: This initiative instituted coproduction of care with families. Families feel heard. It also improved organization and teamwork. Challenges include time commitment for team and families, language barriers for some parents, exploring the optimum mode of communication and expansion to second care site with different support staff. Future plans will include bimonthly 15-minute QI team tele-meeting including parent/ family partners and assessment of patient/family satisfaction.

de Monestrol, I. 1, 4 ; Ericson, P. 2 ; Hjelte, L. 1, 4 ; Moström, J. 5 ; Hager, A. 3 ; Lindblad, S. 4 1. Karolinska University Hospital, Stockholm, Sweden; 2. Sahlgrenska University Hospital, Gothenburg, Sweden; 3. Genia, Upstream Dream, Stockholm, Sweden; 4. Karolinska Institute, Stockholm, Sweden; 5. National Cystic Fibrosis Association, Uppsala, Sweden Objectives: As new life-changing, yet expensive, therapies for cystic fibrosis (CF) become available, Sweden and its responsible authorities strive to create a system that enables stuctured follow-up and evaluation of new treatments for optimal care and best use of resources. In 2016, key CF stakeholders formed the Sweden CF Coalition, a national collaborative learning network that enables persons with CF, families, clinicians, researchers and others to work together toward common goals. The Coalition aims to reduce the burden of illness for patients by radically improving the ability of patients, families and professionals to co-produce improved clinical practice and better care at home.

Methods: In June 2017, the Coalition formed a Coordination Group to focus on developing a national system for orderly introduction and follow-up for new therapies for optimal health, with lumacaftor/ivacaftor as a first example. The Coordination Group developed a Driver Diagram to further outline the mission and set core accountable work streams and evaluation measures. The work streams aim to develop: clinical guidelines; clinical decision support; a health information exchange; a patient support system; a national learning network and international collaboration. The group is composed of representatives from all four CF centers, the National Cystic Fibrosis Association, the CF Working Group of the Society of Medicine, the National Quality Registry Development Group, the Patient Support System Development Group, the Chair of the CF Coalition and Coalition advisors. The Group meets monthly to report back from the various work streams and to ensure progress is made towards the mission.

Results: The result of the work streams, after one year, is the co-creation of the first system for introduction and follow-up of new CF therapies. The first version of the system includes:

• Consensus on the customer journey for the first year on lumacaftor/ ivacaftor;

• A new module for lumacaftor/ivacaftor follow-up meetings in the National Quality Registry, representing a first step in developing an encounter-based quality registry;

• The first health information exchange developed, with measures for follow-up and evaluation of the therapy, including both clinic-and patient-reported data;

• A special module for patient treatment follow-up and adherence support developed within the currently used mobile patient support system, Genia. The module is integrated with clinical workflows and health IT systems and is developed with industry collaboration;

• A pilot to include shared-care sites with the patient support system and Registry for seamless care and follow-up for rural patients; and

• A pilot at one CF center with a dedicated nurse and physiotherapist for introduction and follow-up of lumacaftor/ivacaftor. Conclusion: The system is ready to be tested and implemented in the summer of 2018. The advantages of a Coordination Group, consisting of key stakeholders, include the ability to gather and share lessons in real time and make needed iterations in rapid cycles, based on quality improvement methodology.

Videna, S. 1 ; Metz, K. 1 ; Batchelor, D. 1 ; Drescher, C. 1 ; Taskar, V. 1 ; Tummons, D. 1 ; Forseen, C. 1 ; Curry, D. 1 ; Self, S. 2 1. Augusta University, Augusta, GA, USA; 2. Pediatric Pulmonary, UAB, Birmingham, AL, USA Introduction: Using the Dartmouth Microsystems approach during LLC-2, our theme was to improve nutritional outcomes in our adult CF patients. 41% of our patients were below their goal BMI which is defined by the CF Foundation as ≥23 kg/m 2 in males and ≥22 kg/m 2 in females. We embarked on a series of interventions towards our global aim which was to reduce the proportion of patients below goal BMI by 10%.

Objective: Our specific aims were obtaining accurate height and weight for 100% of our patients, standardizing these measurements in the clinic, assessing nutritional knowledge in our patient population, and assessing their attitudes towards attaining the goal BMI.

Methods: We surveyed patient barriers towards attaining their goals and performed a cost/utilization analysis of programs available for enzymes, supplements, and vitamins. We also revised a comprehensive nutritional algorithm that any adult CF team member could use during the clinic visit. Our change idea was the STOP-LIGHT System which was targeted, visual nutrition education to provide repeated exposure to information regarding appropriate nutritionally dense foods. The foods were categorized as: Red (STOP!), Yellow (Caution), and Green (Go!). Foods listed in the Green Light category contain high protein and high calories in the form of medium-lean meats, healthy fats, and whole grains. Foods listed in the Yellow Light category fall somewhere in between the Red and Green Light categories.

Results: Of the 41% patients who were below goal BMI at our center, 47% were female and 53% were males. After standardizing height measurement in clinic, the number who were below goal BMI fell to 36%. 34 patients who were below BMI were surveyed, 23 wanted to gain weight while 11 wanted to maintain current weight. Among the 23 who wanted to gain weight, 12 listed no barriers, 6 mentioned cost and the remaining, nausea, lack of appetite, exercise, transportation. The Table shows the distribution of their insurance status and availability of programs for enzymes, vitamins and supplements. STOP-LIGHT data are presented separately.

Conclusions: 1. Understanding the limitations of our measurements (height/weight) was the first step towards improving nutritional outcomes. 2. Knowing the specific attitudes of patients who were below goal BMI helped plan an individual approach. 3. Optimization of utilization of assistance programs helped distinguish mental and financial barriers in the below goal BMI group. 4. STOP-LIGHT System provided objective, nonjudgmental education to bridge the barriers between knowledge and practice.

Acknowledgment: This work was conducted as a part of LLC-2. We undertook a quality improvement (QI) initiative to study lung function decline in our adolescents and implement interventions based on QI principles to slow FEV1 decline and improve median lung function.

Objectives: 1) To describe variable rates of lung function decline and risk factors for rapid decline in our CF adolescents. 2) To reduce rate of lung function decline.

Methods: We created a retrospective cohort of CF patients (n=69), ages 8 to 18 years as of 1/1/2013 who had at least one FEV1 percent predicted (pp) value measured each year from 2013-2016 and no history of lung transplant. Individual baseline FEV1pp was calculated for each year as the average of the best value of each quarter, according to the method used for the CF Foundation registry report. Individual annual rate of decline in baseline FEV1pp from 2013-2016 was also calculated. Risk factors for rapid decline were examined using registry data. Four QI interventions were implemented. First, lung zones were created based on approximate quartiles for rate of decline, ranging from A (no decline) to D (very rapid decline >3.5%/year). Lung zone and rate of decline were added to our clinic preview sheet. Second, a table with this decline data was added to the respiratory therapist's clinic note for each patient. Third, a rapid decliner checklist was created to be completed during clinic preview for patients in the C or D zones. Fourth, a lung health algorithm for providers was generated to encourage uniformity in care and focus attention on rapid decliners.

Results: Median baseline FEV1pp for our cohort decreased from 91.5 in 2013 to 81.3 in 2016. The mean annual rate of decline for our cohort was -2.1% per year (SD 3.3) . There was significant variability in rates of decline in the cohort: 25% (n=17) had no decline in FEV1pp, while the bottom quartile had rates of decline of -3.7 to -13.3% per year. CF-related diabetes patients diagnosed by 2013 had significantly faster decline in FEV1pp (-5.1 vs. -1.5%/year, p=0.02) , as did patients with Medicaid insurance compared to private insurance (-3.5 vs. -1.2%/year, p=0.009) . Patients diagnosed with prediabetes by 2016 had a tendency towards faster decline in FEV1pp (-2.1 vs. -0.8%/year, p=0.08) . Of the original cohort, 49 patients received care at our program in 2017 and were exposed to the QI interventions. In this group, median FEV1pp increased from 81.3 to 83.3 in 2017. The mean rate of decline was only 0.7% in 2017 and 45% (n=22) improved their baseline FEV1pp with a median improvement of +3.6. The mean annual rate of decline improved from -2.1%/year for 2013-2016 to -1.7%/year for 2013-2017.

Conclusions: Our results demonstrate improvements in lung function and slowing of decline with QI interventions. These improvements may be attributable to a more standardized and proactive approach to decreases in lung function and increased clinician attention to patients with rapid decline, especially those with high baseline lung function. The risk factor analysis led us to implement additional QI projects around improving CF-related diabetes control. Introduction: CF patients age 18 to 29 are at risk for decreased or halted clinic attendance and participation in their healthcare due to a variety of age related risk factors such as moving away to college and transitioning to healthcare independence. According to the 2016 CF Registry report, the national median FEV1 in this age group is 74.2%, compared with 72.5% at our center. While some patients have regular contact with their care center during this time, others are at risk for decreased contact and a decline in overall health status including FEV1. The specific aim of this project is to improve the FEV1 in this age group by 1% within the next 12 months.

Methods: CF patients at the center 18 to 29 years old were identified using portCF. We created a survey containing questions addressing sickness, equipment issues, perception of respiratory function, ability to meet nutrition goals, gastrointestinal issues, insurance concerns or medication refills, clinical trials, goals for the next year, transportation to next appointment, and preference for communication method in between appointments. Our first Plan-Do-Study-Act (PDSA) cycle involved calling every patient in this age range within our center and documenting parameters pertaining to this call in centralized spreadsheet. Up to two call attempts per patient per quarter were made. Successful (or unsuccessful) contact, duration of call, next appointment status, major concern(s), and whether or not the patient was hospitalized were documented in the central spreadsheet by each QI team member conducting the call. If concerns were reported by the patient, the caller contacted the appropriate team member at that time.

Results: During the first quarter of 2018, our QI team called 63 patients and was able to make contact on 28 of the 74 calls made (37.8% success rate). The average length of each successful phone call was 8 minutes (range 2 to 30 minutes). Of the patients who were successfully contacted, 57.1% were found to not have their next appointment scheduled, and 7 patients scheduled appointments at the time of the call. First quarter major concerns reported by the patients included GI (n=3), Medications (n=2), Respiratory (n=3), Insurance (n=4), Mental Health (n=4), and Other (n=11).

Conclusions: Our findings indicate that making quarterly phone calls may be helpful in addressing patient concerns, which could contribute to increased attendance at clinic appointments, medication adherence, and study participation. Calls will continue to be made and documented each quarter by the QI team and will be available at the conference. Acknowledgment: Supported by the Cystic Fibrosis Foundation Fun LLC3. Rationale: Pulmonary exacerbations (PEx) are a substantial burden for CF patients and many do not recover lost lung function. Despite this, best practices are lacking for the length, type and location (hospital or home) of IV treatments. Recent analysis showed IV treatments in the hospital were more successful than those in the home, including in return to 90% of baseline FEV 1 (Ann Am Thorac Soc. 2018; 15:225-33) . Our previous quality improvement (QI) initiatives focused on hospital care and communication. As our Home IV Antibiotic Therapy (HIAT) use was increasing, we began a QI project to similarly improve our HIAT. Specific aims include improving discharge planning, standardizing follow-up practices, and safe care administration.

Methods: We evaluated our HIAT process, including initiation, planning, and discontinuation. As HIAT occurs outside of the healthcare system many care gaps exist, such as lack of regular documentation and problems occurring away from aid. Multiple types of problems existed, with medical complications and central line issues being most common and labs errors identified as most avoidable. When to use HIAT was also not standardized or well defined, leading to confusion for the family and care team.

To address these concerns we designed a set process and leveraged the electronic health record to standardize care/communication. The new process includes setting individualized criteria for HIAT use, including a specific FEV 1 goal, minimum number of hospital days, and stable social situation/compliance. Follow-up is now scheduled at 7 days for all, with lab tests generally coordinated. Charting templates, including a Home IV Road Map, were devised to allow consistent and accurate communication and monitoring.

Results: After implementation our tools and process were quickly adopted, with 100% compliance by 6 months. As for safety, we now better monitor potential issues and have had no more errors with laboratory tests. As for efficacy, our average rate of return to ≥90% of baseline FEV1% by first follow-up visit was 87% before intervention and increased to 96% after intervention. All patients have successfully met this goal for >1 year, nearing special cause variation (G-and P-chart, limited by sample size). Interestingly, despite starting the project due to increasing HIAT, we had a regular decrease with frequency essentially halving every year. Reasons for this are multiple, including increased awareness of advantages/disadvantages of hospital versus home, patient transitions, and precise selection of appropriate candidates. Most start the course in the hospital and transition to home. Overall, we feel our care quality and communication has improved with this process and tools.

Conclusions: Exacerbation treatment is an essential part of CF care, with much uncertainty on best practices. Using set criteria, we have standardized our HIAT care plans and have had successful return to baseline FEV 1 , similar to our inpatient care. We have had a subsequent decrease in overall HIAT use, with multiple causes and temporally related to our project. Further work is needed to better define PEx treatment needs. Next steps include increased integration with home care companies and our adult program.

Acknowledgment: Supported by a QI Grant from the CFF. amikacin, tobramycin and gentamicin. The ototoxicity of aminoglycosides is well-established and individuals with CF are at an increased risk for developing hearing loss (HL) due to the toxicity of these medications. Along with concerns for auditory dysfunction, the use of aminoglycosides is also vestibulotoxic. A recent study by Handelsman and coworkers, (Pediatr Pulmonol. 2017; 52:1157-62) , investigated the vestibulotoxicity of aminoglycosides in patients with CF, finding an incidence of 79% among their cohort diagnosed with some degree of vestibular dysfunction. In previous years, patients with CF were not routinely referred for a comprehensive audiologic and vestibular evaluation. A quality improvement (QI) initiative to obtain baseline hearing and vestibular evaluations and establish an audiologic protocol was developed and incorporated into the routine care for patients with CF seen at the Univ. of Chicago Medicine (UCM). By obtaining baseline and follow-up evaluations, UCM established a monitoring program to assess the auditory and vestibular systems, increased early identification of patients with HL, and identified ways to preserve hearing, while also educating our patients on the importance of hearing protection due to the synergistic effects of noise exposure during treatment with ototoxic medications.

Methods: Patients were recruited for testing in conjunction with their annual CF clinic visit and data are collected on an on-going basis. The initial phase focused on obtaining hearing evaluations (pediatrics) and the second phase is focused on obtaining hearing and vestibular evaluations (adults). A baseline evaluation was completed unless patients had a known HL or were unable to complete testing due to age or current medical status. Individuals found to have a HL were referred for follow-up diagnostic testing, treatment, and management. Inpatient protocol allows for patients to receive a full diagnostic evaluation prior to and after ototoxic IV antibiotic treatment.

Results: Patients seen for a baseline evaluation with normal hearing were referred for an annual evaluation. Patients diagnosed with HL were treated and seen for routine follow-up in Audiology. Since initiating the project, an increase to 84% of pediatric patients were tested over a 12-month period and of those tested, three were identified with bilateral HL. Over the next 12 months, our goal is to increase from 12% to 50% of adults seen for testing.

Conclusions: Through improved awareness, availability, and convenience of testing, the number of patients referred to Audiology increased from previous years when no clinical protocol was in place. This QI initiative resulted in identifying HL and vestibular dysfunction which may have otherwise been missed. Routine auditory and vestibular evaluations as part of an ototoxicity monitoring program should be implemented and become standard practice for patients with CF. The need also exists to establish ototoxicity monitoring guidelines and recommendations through the CF Foundation. Introduction: Compliance with airway clearance therapy (ACT) and optimizing airway clearance are significant challenges in cystic fibrosis (CF) pulmonary management. These challenges need effective management to ensure airway clearance is optimized and lung health is maintained. The Duke Pediatric Pulmonary Function Laboratory administers an annual and a follow-up needs assessment questionnaire for CF patients. The follow-up questionnaire is conducted every 3 months to ensure knowledge is maintained in three important areas: pulmonary medications, ACT, and infection control. Knowledge in these areas is important to optimize lung health and facilitate the transition from the pediatric pulmonary function lab to the adult pulmonary function lab. Ultimately, the questionnaire is a systematic approach to maintain and reinforce knowledge in pulmonary medications, ACT, and infection control. The aim of this quality improvement (QI) project was to assess the impact of this systematic approach on optimal airway clearance devices and settings. The ultimate goal is to remedy suboptimal airway clearance and to use this opportunity to further educate CF patients and caregivers.

The QI methodology framework used is the Plan-Do-Study Act (PDSA). The questionnaire follows CF Foundation guidelines. The ACT data on pre-implementation and post-implementation respiratory therapist (RT) review was evaluated. The implementation strategy for the RT-driven CF needs assessment QI initiative is to gather data on the patient from both the annual and follow-up needs assessment review. Each patient and/or caregiver was given an annual questionnaire to fill out. The decision of who will complete the questionnaire was dependent upon the cognitive level of the patient and ability to understand the questions. The follow-up questionnaire is filled out by the RT and reviewed with the patient. Any deficits in knowledge or ACT methods are immediately addressed with an explanation and hand-out to reinforce knowledge sharing and proper technique, and if needed, a phone call is placed to ensure deficits in knowledge were met. The design of the questionnaire is fill in the blank, multiple choice, and yes/no.

Results: A total of 52 CF patients were evaluated. The questionnaire identified 19 patients with vests that did not fit, 34 patients that had suboptimal vest settings, and 4 patients that did not have a vest or other ACT device. Following our intervention 0 patients had vests that did not fit (100% reduction), 9 patients had suboptimal vest settings (74% reduction) and 4 patients did not have a vest or other ACT device (0% reduction) due to the patients' primary ACT being chest physiotherapy.

Conclusions: Thus, our intervention resulted in a significant decline in the number of patients without optimal vest size and settings and no ACT device. The outcomes of the QI initiative proved the effectiveness of an RT-driven needs assessment questionnaire in evaluating appropriate vest size and vest settings. Ultimately, further studies are needed to evaluate the effectiveness of this RT-driven CF needs assessment questionnaire on CF pulmonary exacerbations. Introduction: CF patients are at risk for rapid pulmonary progression after periods of exacerbation, which can cause irreversible damage to the airways. Given the link between lung function and survival, it is imperative to screen and intervene early in patients with large pulmonary function declines. Our center has engaged in a quality improvement (QI) initiative to target lung function declines within our adult population.

Objective: To improve lung function, as measured by FEV 1 , in adult CF patients with rapid pulmonary decline at the UAB Adult CF Center.

Methods: A multidisciplinary rapid pulmonary progression algorithm has been implemented for patients classified with rapid decline status. We defined baseline FEV 1 as the average of the two best FEV 1 measures from the past 12-18 months and entry criteria as the following: ≥ 3 exacerbations OR FEV 1 of < 40% AND decline of > 5% in FEV 1 from baseline over the past year OR < 3 exacerbations and a decline of > 10% in FEV 1 from baseline over the past year after pulmonary exacerbation treatment. The algorithm focused on improving FEV 1 through intensive clinic visits with Pharm.D., RT, MD, and NP with briefer clinic follow-up intervals of 4-6 weeks. Pulmonary interventions included: identification of adherence issues through pharmacy assessments with MMAS tool, counseling on optimization of respiratory therapies, use of additional respiratory clearance strategies, and referrals to other specialists. Patients received a written pulmonary action plan at the end of each visit summarizing pulmonary recommendations and interventions, medication changes, and specific collaborative patient and team goals for follow-up.

Results: At present 7 patients (3 female, 4 males) have completed 3 or more visits in the algorithm. 81% of patients that completed a second visit and 71% of patients that completed a third visit returned to clinic in the recommended 4-6 week follow-up interval. The average increase in FEV 1 following initiation of algorithm, between initial visit and visit 2 was 2.05% with an average net improvement in FEV 1 between initial visit and visit 2 of 4.57%. Average FEV 1 for females at initial visit was 60.8% (n=9) and average FEV 1 for males at visit 1 was 42.3% (n=9). By visit 2, average FEV 1 for females had improved to 66% (n=8) and average FEV 1 for males to 46.6% (n=8). At visit 3, average FEV 1 for females was 65.6 % (n=3) and average FEV 1 for males improved to 53.75% (n=4). Of the 7 patients that completed 3 visits on the algorithm, 3 exited from the rapid pulmonary progression protocol based on sustained improvement in FEV 1 at prior baseline over subsequent visits.

Conclusions: Implementation of a multidisciplinary rapid pulmonary progression algorithm, use of a written FEV 1 action plan, and a unified team message on improving pulmonary status led to improvements in pulmonary function and stabilization of decline in the rapidly progressing subset of patients in the adult CF center. Future work will continue use of this pathway with a focus on sustaining FEV 1 and further optimizing pulmonary status in subsequent visits as well as implementation of automated identification of patients through CF patient registry tools.

Stewart, J. 1,2 ; Hamilton, J. 1,2 ; Meihls, S. 1 ; Hamaker, M. 2 ; Asfour, F. 2 1. Pediatrics, Intermountain Cystic Fibrosis Center, Salt Lake City, UT, USA; 2. Pediatrics, University of Utah, Salt Lake City, UT, USA Introduction: Clinical care guidelines for cystic fibrosis-related diabetes (CFRD) include screening annually with a 75g 2-hour oral glucose tolerance test (OGTT) starting at 10 years of age (Moran A, et al. Diabetes Care. 2010; 33:2697-708) . Center-specific registry data showed that in 2012 OGTT rates hit a peak of 81.1%. In 2013, our rates started to decline; rates were 68.1% in 2014, 65.6% in 2015, and 61.7% in 2016. This decline is likely due to a decrease in clinic appointments and changes in providers at our center. This was the impetus for this quality improvement project.

Objectives: Increased screening for CFRD by increasing OGTT rates among patients over 10 years old without a diabetes diagnosis.

Methods: Interventions were multifaceted and included an inpatient and outpatient focus. Outpatient interventions included: increased number of providers, increased availability of clinic appointments, and increased communication with patients and families regarding the importance of OGTT testing and CFRD screening. We added a monthly CFRD clinic with a dedicated endocrinologist which improved awareness of the relationship between CFRD and pulmonary outcomes. We obtained OGTT during annual clinic visits in the outpatient setting and inpatient at the end of hospitalization, when patients are at baseline state of health. Additional inpatient interventions included: education with nursing and phlebotomy regarding OGTT administration and CFRD, and a dedicated CF nurse practitioner who rounds on inpatients Mondays, Wednesdays, and Fridays with a pulmonologist. This has improved communication with inpatient resident teams and endocrinology.

Results: According to registry data we calculated for 2017 the OGTT rate for patients over 10 years old without a diabetes diagnosis was 80%. This is significantly improved from previous annual rates, which were: 68.1% in 2014, 65.6% in 2015, and 61.7% in 2016. Conclusions: Multifaceted interventions at our pediatric center resulted in an increased rate of screening for CFRD that is in accordance with the current guidelines. In an effort to continually improve patient care, further inventions will be made in order to hold these gains and continue to increase rates of CFRD screening. This is an ongoing QI project. Future implications will include interventions related to positive CFRD screening to improve overall CF patient care and outcomes. inpatient (49%) and median FEV1 predicted (80.7%) were lower compared to the national 82.2% and 92.9% respectively.

Objective: We devised a patient survey to evaluate opinions regarding LOS in order to identify specific improvement opportunities. IRB exemption was obtained. The survey was given from April 2017-April 2018.

Methods: Paper surveys were administered to families at time of visit. Response data were entered into REDCap. Survey consists of two demographic items, average number of hospitalizations over the past 3 years, age and thirteen Likert-type items assessing perceptions on hospital LOS. Age was recorded as a continuous variable, and average number of hospitalizations as categorical with 5 response categories. Likert-type items used a 5-point scale, and responses were coded as 1="Strongly Disagree"; 2="Agree"; 3="Neither agree nor disagree"; 4="Agree"; 5="Strongly agree." Frequency of responses was used to order Likert-type items from most agreement to least agreement. Additionally, mean (95% CL) is reported for each Likert-type item. Frequency (%) is reported for the average number of hospitalizations. SAS v9.3 was utilized for analysis.

Results: Nineteen out of 35 families completed the survey. Most agreed that patients should remain hospitalized for a majority of the 14-day IV antibiotics course (mean Likert scale value (MLSV) 3.58 (95% CL for mean (2.97, 4.19) ). Most agreed that longer LOS leads to closer monitoring, prompt therapy adjustments (MLSV 4.05 (95% CL (3.68, 4.43) ) and better outcomes (MLSV 3.68 (95% CL for mean (3.10, 4.26) ). Most common negative associations with increased LOS were perceived increased exposure to infection (MLSV 3.63 (95% CL (3.12, 4.14) ) and decrease in school performance (MLSV 3.74 (95% CL (3.23, 4.24) ). Most disagreed that hospitalizations decrease both the stress of managing CFE (MLSV 2.63 (95% CL (1.98, 3.28) ) and timely delivery of treatments (MLSV 2.68 (95% CL (2.15, 3.22) ).

Conclusions: Our objective was to identify specific intervention opportunities to improve perception of increased LOS. Although the majority of patients agreed that longer stays result in improved outcomes, the most common concerns were exposure to infection and decreased school performance. A majority of participants disagreed that longer LOS leads to timelier delivery of therapy and decreased stress. Next steps in quality improvement (QI) will consist of continual work with the hospital staff to ensure timely delivery of treatment, patient education about hospital infection control measures and minimization of interruptions in school work. As per 2016 CFF Center specific report, our center showed increased duration of treatment completed in the hospital (76.2%) and improved FEV1% (95.7%) reflecting early stages of our QI initiatives. Objective: To determine the prevalence of probiotic supplementation, understand clinical practice of CF healthcare providers, and identify associations of probiotic use to gut microbiome (GM) and fecal calprotectin (a marker of intestinal inflammation) in young children with CF.

Method: We utilized data from FIRST (Feeding Infants Right… from the STart), a multicenter prospective observational study that includes a cohort of 183 infants born from 1/2012 to 12/2017 and enrolled at age 1.8±1.0 mo from six CF centers in the US. Of 142 infants who reached age 24 mo by 6/30/18, 5 withdrew at 3 mo and 9 had low birth weight. The remaining 128 infants had data on probiotic use collected prospectively at CF center visits (monthly before age 6 mo, bi-monthly at age 6-12 mo and quarterly thereafter). The GM ancillary study was initiated in 2015; and 54, 68 and 82 FIRST subjects had GM and calprotectin data at age 6, 12 and 24 mo, respectively, as of 6/30/18. A survey on healthcare providers' recommendations on probiotic supplements was sent via the US CF Foundation to 153 CF center directors in 6/2018; as of 7/25/18, 103 CF healthcare providers completed the survey.

Results: Overall, 52 (41%) of 128 FIRST subjects received probiotic supplements in the first 2 yr of life. The prevalence of probiotic use varied greatly among CF centers (15%, 19%, 44%, 64% and 73%, p<0.001), increased with birth year (30% and 48% among those born in 2012-13 and 2014-16, respectively, p=0.03) , and was higher in infants with meconium ileus than those with pancreatic insufficiency or pancreatic sufficiency (68%, 39% and 17%, respectively, p=0.004). Among probiotic users, 28 (54%) started in infancy and used >12 mo (long-term users), 11 (21%) used on and off in the first 2 yr of life, and 13 (25%) began in the second yr for a duration of 1-9 mo. About 80% of CF healthcare providers recommended probiotics to their pediatric patients (age 0-18 yr) in the past yr. More specifically, 12% of providers reported recommending probiotics at least half of the time to infants with CF, this percentage increased to 25% for patients 13-24 mo of age and 43% for patients 2-10 yr of age.

Cross-sectional analyses showed that GM diversity increased from age 6 to 24 mo. Relative abundance of the top two phylla reversed: Acidobacteria decreased (33% at 6 mo, 20% at 12-24 mo) and Firmicutes increased (30% at 6 mo, 45% at 12-24 mo). Bifidobacterium (the most abundant genus) decreased from 34% at 6 mo to 17% at 12-24 mo. The prevalence of abnormal fecal calprotectin (>120 ug/g) was higher at 6 mo (44%) than 12 mo (18%) or 24 mo (26%), p=0.04. GM diversity and fecal calprotectin did not differ significantly between long-term users and nonusers of probiotics at age 6, 12 or 24 mo.

Conclusion: Our preliminary analyses showed that gut microbiome diversity increased and intestinal inflammation decreased from 6 to 24 mo of age but no significant differences were noted by probiotic use. More analyses adjusting for antibiotic therapy are underway. Acknowledgment: Funded by R01DK109692.

Connolly, L.; Bourke, B.; CFLD Research Group; Rowland, M.

Introduction: The pathophysiology of cystic fibrosis-related diabetes (CFRD) is not completely understood, however, decreased insulin secretion is considered the primary defect. There is conflicting evidence on the role of insulin resistance (IR) in CFRD. IR is a common feature of non-CF liver disease. The aim of this study was to investigate whether liver disease (LD) contributes to IR in CF.

Methods: Participants over the age of 8 enrolled in a long-term prospective study of cystic fibrosis liver disease (CFLD) were invited to participate in this study. CFLD was classified according to the North American Cystic Fibrosis Foundation guidelines as 1. CFLD, 2. Nonspecific CFLD, and 3. No liver disease (NoLD). CFRD was classified according to the American Diabetes Association guidelines as normal (fasting<5.6 mmol/L, 2h<7.8 mmol/L) or impaired (fasting≥5.6-6.9mmol/L, 2hr≥7.8-11.0mmol/L). Fasting insulin, glucose, and HOMA2 were compared between those with CFLD, and 2 control groups i) those with NoLD and impaired glucose tolerance (IGT) and ii) those with NoLD and normal glucose tolerance (NGT). Exclusion criteria included insulin therapy, nonspecific CFLD, type 1 diabetes mellitus, corticosteroid use, CFTR corrector therapies, FEV1<40% or a recent pulmonary exacerbation. As puberty influences IR, we analysed postpubertal participants separately and stratified this group based on gender. Fasting glucose (mmol/L) and fasting insulin (pmol/L) were measured using a hexokinase glucose assay method and a chemiluminescent microparticle immunoassay. Insulin resistance was determined using the HOMA2 Calculator © (University of Oxford 2013). Data were presented as mean±SD and compared with ANOVA. P-value <0.05 was considered significant.

Results: Seventy-one participants were included of whom 26 had CFLD (mean age 16.5±4. 8 (M:19, 73 .1%)),17 had no evidence of LD but were classified as IGT (mean age 15.7±4.3 (M:13, 76.5%)) and 28 had no evidence of LD with NGT (mean age 15.9±4.9 (M:19, 67.9%)). There was no evidence of IR in any of the 3 groups (CFLD 1.1±0.6, NoLD/IGT 0.86±0.36 and NoLD/NGT 0.90±0.36, p=ns). Glucose and insulin measures were similar in all 3 groups. Forty participants were postpubertal, of whom 15 had CFLD (mean age 20. 1± 2.8 (M:10, 66 .7%)), 11 had NoLD with IGT (mean age 18. 2± 2.9 (M:9, 81 .8%)) and 14 had NoLD with NGT (mean age 20.2±2.9 (M:9, 64.3%)). HOMA2 values were again similar between the groups (CFLD 0.96±0.37, NoLD/IGT 0.74±0.13, NoLD/NGT 0.96±0.36, p=ns) . Those with CFLD and those with NoLD/IGT had significantly higher glucose measures compared to those with NoLD/NGT (CFLD 5.52±0.55, NoLD/IGT 5.63±0.23, NoLD/NGT 5.2±0.38, p=0.03) . Insulin values were similar in all groups. A gender stratified analysis of postpubertal participants demonstrated females with CFLD had significantly higher insulin values compared to males (F:63. 68±19.26, M:42.67±14.29, p=0 .03) and were more insulin resistant compared to males with CFLD (F: 1.23±0.39, M:0.82±0.28, p=0.03) . There was no difference in insulin or HOMA2 values between males and females with NoLD/IGT or with NoLD/ NGT. Glucose values were similar between males and females across all groups.

Conclusion: Females with CFLD, known to have a more severe phenotype, are more insulin resistant than males with CFLD. Verkman, A.S. 2 1. Pediatrics, University of California, San Francisco, San Francisco, CA, USA; 2. Medicine and Physiology, University of California, San Francisco, San Francisco, CA, USA Gastrointestinal (GI) problems are common in CF including constipation, meconium ileus, meconium plug syndrome and distal intestinal obstruction syndrome. Lack of CFTR-mediated chloride secretion in the GI tract is thought to be the main cause for these problems. The prevalence of CF-associated constipation is very high and it is often refractory to dietary modifications and currently available laxatives. Here, we report a novel target for drug therapy of GI problems in CF -SLC26A3 (down-regulated in adenoma, DRA), a chloride/anion exchanger expressed in the luminal membrane of intestinal epithelial cells where it facilitates electroneutral NaCl absorption. SLC26A3 loss of function in humans and mice causes severe chloride-losing diarrhea. SLC26A3 inhibition is a unique and CFTR-independent approach to restore luminal chloride concentrations in CF GI tract. The objective of this study was to identify SLC26A3 inhibitors by high-throughput screening as novel tools for studying intestinal fluid absorption mechanisms and test their efficacy in a mouse model as first-in-class drug candidates for anti-absorptive therapy of constipation. A screen of 50,000 synthetic small molecules was done in FRT cells co-expressing murine SLC26A3 and a yellow fluorescent protein halide sensor, followed by structure-activity studies on analogs of active compounds. The 4,8-dimethylcoumarin DRA inh -A250 fully and reversibly inhibited SLC26A3-mediated chloride exchange with bicarbonate, iodide and thiocyanate with IC 50 down to 150 nM (see Figure) . DRA inh -A250 at 10 µM was selective for SLC26A3, as it did not inhibit the homologous anion exchangers SCL26A4 (pendrin), SLC26A6 (PAT-1) or SLC26A9, nor did it affect other chloride channels and intestinal ion transporters. In mice, intraluminal DRA inh -A250 fully blocked fluid absorption in closed distal colonic loops, and oral DRA inh -A250 prevented loperamide-induced reductions in stool weight, number of pellets and stool water content in wild-type and CF (ΔF508 homozygous) mice. These studies support a major role of SLC26A3 in colonic fluid absorption and suggest the therapeutic utility of SLC26A3 inhibition in constipation and other hyposecretory GI disorders associated with CF. Supported by NIH and CFF.

Concentration-dependent inhibition of SLC26A3-mediated chloride/bicarbonate exchange by DRA inh -A250, whose chemical structure is shown.

Woodley, F.W. 4 ; Moore-Clingenpeel, M. 6 ; Nemastil, C.J. 5 ; Machado, R.S. 1 ; Hayes, Jr., D. 5 ; Kopp, B.T. 5 ; Kaul, A. 3 ; Di Lorenzo, C. 4 ; Mousa, H. 2 ; Jadcherla, S. 7 1. Pediatric Gastroenterology, Universidade Federal de Sao Paulo, Sao Paulo, Brazil; 2. Gastroenterology, Rady Children's Hospital, San Diego, CA, USA; 3. Gastroenterology, Cincinnati Children's Hospital Medical Center, Cincinnati, OH, USA; 4. Gastroenterology, Nationwide Children's Hospital, Columbus, OH, USA; 5. Pulmonary Medicine, Nationwide Children's Hospital, Columbus, OH, USA; 6. Biostatistics Core, Nationwide Children's Hospital, Columbus, OH, USA; 7. Neonatology, Nationwide Children's Hospital, Columbus, OH, USA Background: Children with CF have increased gastroesophageal reflux (GER). Chemical clearance of acid GER is significantly prolonged in these children when compared to symptomatic age-matched children without CF. Because prolonged acid exposure is known to decrease baseline impedance, we posited that children with CF would have lower esophageal basal impedance values, when compared to symptomatic children without CF.

Objective: Compare proximal (Z1) and distal (Z6) baseline impedance values (PBI and DBI, respectively) in children with and without CF.

Methods: Mean (±SD) baseline impedance values from combined pH-multichannel intraluminal impedance (pH-MII) tracings for 16 children with CF (median age 8.2 years, range 3.1-17.7 years) and 16 children without CF (median age 8.3 years, range 3.0-17.2 years) were calculated and compared within groups using paired t-tests and between groups using two-sample t-tests. Chi-square tests were used to evaluate group differences in abnormality; significance was a two-sided p<0.05. Abnormal PBI and DBI was < 2230Ω. Study was retrospective. All patients were off anti-reflux medications.

Results: Nine (56%) children with CF and 8 (50%) children without CF had abnormal DBI (OR=1.3, 95% CI=0.72, 1.3) (p=0.7232). Six (38%) children with CF and 1 child (6%) without CF had an abnormal PBI (OR=9, 95% CI=0.08, 9.0) (p=0.0829). Mean DBI was significantly lower than mean PBI in both cohorts ([CF, p=0.015 Method: Bio-electrical impedance analysis (BIA) measurements were performed in CF patients with TF (2F, 7M) and without TF (matched controls, Co). Percent fat, body fat mass index (BFMI) and fat free mass index (FFMI) were calculated. Patients were matched for age, gender, pancreatic function. Results are given as median (quartiles).

Results: At the start of TF patients were aged 12.8 years (8.7; 15.6) , had a height z-score of -2.7 (-2.8; -1.7) and a BMI z-score of -2.1 (-2.9; -1.4) . The median time they were on TF before the BIA measurement was 1.5 years (0.84; 2.7) . There was a significant improvement in BMI z-score (p= 0.038) and L z score (p=0.008).

The age at BIA measurement was 14.4 years (12.5; 16.5) . TF patients were significantly shorter than Co (p=0.001) but the BMI z-score as well as pulmonary function (FEV1%, FVC%) were not significantly different between TF and Co.

The body composition, however, was significantly different with a higher percent fat (p=0.024) (TF: 24.7 (19.6; 33.1) and Co: 12.9 (10.8; 22.1) ), a higher BFMI (p= 0.019) (TF: 4 (3.3; 5.7); Co: 2.5 (1.95; 3.85) ), and a lower FFMI (p= 0.04) (TF: 11.8 (10.9; 13.4); Co 15.1 (13.2; 17.25) in TF patients. There was no relation of body composition with age at start of TF, duration of TF or pulmonary function. The BMI z-score at start of TF was positively correlated with percent fat (p=0.032) and BFMI (p=0.01).

Conclusion: Patients receiving TF have less FFM. Although tube feeding is often the only remaining way to restore nutritional status in CF patients, it leads to an increase in body fat. Different TF types as well as interventions combining activity to TF should be evaluated in the future. Background: Linear growth is a major determinant of vital capacity and FEV 1 . Adolescent and adult patients with cystic fibrosis (CF) with CFTR-p.G551D have sustained increases in weight following ivacaftor treatment. A study of CF pigs and infants with CF showed CFTR could be directly involved in growth hormone signaling. Thus, CFTR modulators could enhance linear growth in children with CF. We hypothesized that children with CF would have increased height Z-scores after treatment with CFTR modulators.

Methods: We studied a retrospective cohort of 105 patients with CF aged 2-18 years at the University of Iowa Pediatric CF Center. Longitudinal anthropometric data were obtained from the medical record. To assess linear growth potential, we recorded maternal and paternal height. To examine the effect of CFTR modulators, we performed a nested case-control study. Patients treated with ivacaftor or ivacaftor-lumacaftor were matched to untreated subjects with CF by age and sex. Trends in height, weight, and BMI Z-scores were analyzed using linear mixed-effects models with time and sex as fixed effects and individual subjects as random effects.

Results: Preschool-aged children with CF had normal weight and BMI Z-scores, but below average height Z-scores (Z=-0.22 ± 0.09, p=0.019 at age 2). Diminished height Z-scores persisted throughout childhood and adolescence (Z=-0.30 ± 0.12, p=0.014 at age 10). Parental height Z-scores were +0.31 for mothers and +0.29 for fathers, indicating that the children with CF did not achieve their linear growth potential. Adolescent females had disproportionate decreases in BMI and weight Z-scores with time. To examine the effect of CFTR modulators, we compared subjects receiving ivacaftor or ivacaftor-lumacaftor to age and sex-matched controls. Patients who received ivacaftor (N=8) had increasing height Z-scores compared to controls (N=16) before starting ivacaftor (+0.12 units/year for ivacaftor group vs. -0.14 units/year for control group, P < 0.001). This advantage was maintained after starting ivacaftor (+0.18 units/year for ivacaftor vs. +0.06 units/year for control, P =0.002). The ivacaftor-lumacaftor group (N=15) had similar height and weight compared to the matched control group (N=15) before therapy. After starting therapy, treated patients had greater increases in height Z-scores compared to matched controls (+0.025 units/ year for ivacaftor-lumacaftor vs. -0.065 units/year for control, P = 0.018) Conclusions: Children with CF have decreased height Z-scores from early childhood through adolescence and fail to reach their biological growth potential. Our findings in early childhood are similar to those reported in the BONUS study. Patients treated with ivacaftor and ivacaftor-lumacaftor had increased linear growth compared to age and sex-matched controls. However, patients prescribed ivacaftor had better growth than controls prior to starting therapy. Improvements in height could impact vital capacity and FEV 1 . 2 1. Department of Food and Nutrition, Rush University Medical Center, Chicago, IL, USA; 2. Rush University Medical Center, Chicago, IL, USA Introduction: The association of body mass index (BMI) and pulmonary function is well described in the cystic fibrosis (CF) literature. Current recommendations from the CF Foundation suggest BMI goals of ≥22 kg/m 2 for females and ≥23 kg/m 2 for males. Obtaining and preparing food may be important life skills to help maintain weight and BMI in the optimal range. Our goal was to determine whether there is an association between these life skills and BMI.

Methods: A "Self Care Skills Worksheet" was adapted from existing tools (Sawicki and Wood [Transition Readiness Assessment, 2009] , Wood et al [TRAQ, 2014] , and Reiss, Gibson, and the Institute for Child Health Policy at UF, 2005) . The worksheet examined multiple questions regarding self-care skills. This analysis will focus on one question of interest: "Do you grocery shop and cook for yourself for most of your meals/food?" Patients indicated their perceived ability to complete this skill on a Likert scale: I always do this (4), I am learning to do this (3), I don't know how to do this, but I want to learn (2), I don't know how to do this (1), I don't need to do this (0). For the purpose of this data analysis answers were categorized into two groups: "yes" which included those who selected 3 or 4 and "no" for those who selected 0, 1, or 2. The worksheet was administered June-August 2015 to patients ages 18-25 in the pediatric and adult centers. A Mann Whitney U was used to compare median [interquartile range] BMI between patients who answered yes (3 or 4) versus those who answered no (0-2). A p-value of 0.05 was utilized to determine significance.

Results: A total of 15 patients ages 18-25 completed the worksheet. The sample size was majority male (n=10) with a median age (years) of 23 [21] [22] [23] [24] , BMI (kg/m 2 ) of 19. 9 [17.0-22.2] , and forced expiratory volume in 1 second (%) of 60 . Ten patients reported always or learning to grocery shop and cook independently. A Mann-Whitney U test revealed a significant difference in BMI between those who answered yes (21.4 [19.8-23.1] ) versus those who answered no (16.9 [16.7-18 .0], p =0.001) to the key self-care skill regarding grocery shopping and cooking.

Conclusion: A significant difference in BMI was seen for those who cook and grocery shop for themselves versus those who do not perform these tasks on their own. This result highlights the importance of educating patients on basic life skills including grocery shopping and food preparation. Future educational interventions should focus on ensuring that patients feel comfortable performing these self-care skills which in turn could result in increased BMI values. Further analysis with larger sample sizes is necessary to determine the true association between self-care skills and BMI. Introduction: Cystic fibrosis is a systemic genetic disease characterized by chronic inflammation and micronutrient deficiencies. CF Foundation guidelines recommend annual assessment of liposoluble vitamin levels and provide guidance regarding supplementation. Iron deficiency has been described in multiple CF studies. However, few care centers assess iron status routinely and there is much controversy related to supplementation. Iron deficiency is difficult to assess in CF since iron, transferrin, transferrin saturation (TSAT), and ferritin levels are affected by inflammation, and different deficiency cutoffs have been proposed for use in chronic inflammatory diseases. Moreover, while iron deficiency is associated with worsening lung disease, it is not well understood whether deficiency contributes to disease or is part of an adaptive protective mechanism to prevent pathogens from utilizing iron to harm the host. At our center, we perform a routine iron panel in clinic in stable patients annually for women and every 3 years for men. We conducted a single center retrospective analysis to identify prevalence of iron deficiency in our adult patient population.

Methods: Medical charts were reviewed for demographic information, FEV1, vitamin levels, CBC, use of acid blockers and CFTR modulators, colonization of Pseudomonas aeruginosa, and presence of CF-associated comorbidities. Prevalence rates were calculated using a variety of parameters to classify iron deficiency, including TSAT <20% and ferritin levels <30 ng/mL, <41 ng/mL, and <100 ng/mL. For statistical analyses, iron deficiency was determined if either TSAT <16% or ferritin <12 ng/mL. Patients whose iron panels were drawn during hospitalization or exacerbation were excluded from analysis. Data were analyzed using SPSS to identify risk factors significantly associated with iron deficiency and to determine the correlations between biochemical markers of iron status.

Results: Of the 105 patients at our center, 67 were included in the analysis. Using a definition of either TSAT <16% or ferritin < 12 ng/mL, the prevalence of iron deficiency was 41.8% (n=67). By gender, 30.0% of men (n=30) had iron deficiency whereas 51.4% (n=37) of women did. When using more liberal parameters of TSAT <20% or ferritin <41 ng/ mL, deficiency was 67.2% (n=67). Serum iron and TSAT correlated very strongly whereas transferrin correlated poorly with other biomarkers. Iron deficiency was not associated with pancreatic insufficiency or use of proton pump inhibitors or H2 blockers.

Conclusion: These results suggest a high prevalence of iron deficiency in CF patients even in a stable population, and that malabsorption does not seem to be a key player in iron deficiency. Moreover, iron deficiency rates were much higher in CF compared with the general population according to CDC rates (9-12% in women and 2% in men) regardless of which ferritin and TSAT levels were used to define iron deficiency. More studies are needed to understand how to diagnose and manage this important nutrient in CF. With more information, perhaps assessment of iron status will become routine practice in more CF centers to enhance the quality of care provided for patients.

Phong, R.Y. 1 ; Taylor, S.L. 2 ; Kirk, K. 3 ; Robinson, B.A. 3 ; Jhawar, S. 1 ; Nandalike, K. 1 1. Pediatric Pulmonology, Univ. of California Davis Medical Center, Sacramento, CA, USA; 2. Department of Public Health Sciences, Univ. of California Davis, Sacramento, CA, USA; 3. Div. Pulmonary & Critical Care Medicine, Univ. of California Davis Medical Center, Sacramento, CA, USA Background: There is ongoing interest in quantifying malnutrition in children with cystic fibrosis (CF). New standardized diagnostic indicators for malnutrition using weight for length/body mass index (WFL/BMI) z-scores and mid-upper arm circumference (MUAC) z-scores are being used by dietitians in clinical practice (Becker P, et al. Nutr Clin Practice. 2015; 30:147-61) . However, the applications of these diagnostic indicators in the pediatric CF population have not been well described.

Objective: To compare MUAC z-scores to WFL/BMI z-scores in diagnosing and classifying malnutrition in children with CF, and to explore the relationship between malnutrition classification and pulmonary function testing (PFT) values.

Method: A retrospective observational study of pediatric CF clinic visits from November 2017 through April 2018 was carried out, and patients aged 2 months through 18.5 years old who received MUAC, weight, height and PFTs were included. Previously published malnutrition classifications using MUAC z-scores and WFL/BMI z-scores were utilized: -1 to -1.9 z score is equivalent to mild malnutrition; -2 to -2.9 z-score is moderate malnutrition; and -3 or greater z-score is severe malnutrition (Abdel-Rahman, et al. Nutr Clin Practice. 2017; 32:68-76; Becker P, et al. 2015) . PFT values included absolute value of FEV1, FVC, ratio of FEV1/ FVC and FEF25-75 (L/sec) as obtained from PFT during clinic visits.

Results: We assessed 42 patients (14 females). We found MUAC and WFL/BMI z-scores were significantly, positively correlated (Pearson's correlation r = 0.87, t40 = 11.1, p < 0.001), however there was poor agreement in categorizing malnutrition as evidenced by Cohen's kappa of 0.24 (95% CI: 0.046, 0.44). The WFL/BMI z-scores identified 85% of patients as adequately nourished, however MUAC z-scores suggested that 50% of our patients had some degree of malnourishment (Table) . None of the PFT values were significantly related to either MUAC or WFL/BMI z-scores, with correlations less than 0.3. This could be due to small sample size and 9/42 patients unable to perform PFT due to age limitation.

Conclusions: MUAC and WFL/BMI z-scores are useful to assess malnutrition in children with CF, however MUAC z-scores may be a more sensitive measure to diagnose and quantify malnutrition. Though we found no statistically significant connection between PFTs and malnutrition, the relationship between MUAC z-scores and PFTs needs further assessment in a larger and more diverse sample size. (BM) for most infants diagnosed with CF. However, recent studies have shown that BM intake rapidly decreases over the first 6 months of life. We initiated a quality improvement project to optimize the provision of BM in babies with CF.

Methods: Since 2015, our CF team has partnered with hospital-based International Board-Certified Lactation Consultants (IBCLC) to provide breastfeeding support to mothers who had a child newly diagnosed with CF. IBCLCs have been incorporated into the initial diagnosis visit and all breastfeeding mothers offered IBCLC support; additional follow-up has occurred as needed. Medical record review was used to determine duration of BM intake prior to the intervention and between 2015-18, the years post-intervention. We included infants with classic CF who were diagnosed within the first month of life and were being provided some BM at the time of diagnosis. Length of BM provision was defined as the time in months until the last clinic visit where any BM was being provided. We assessed anthropometric data and surveyed the mothers of the infants who had met with the IBCLC.

Results: The pre-intervention group consisted of 14 infants who were being provided some BM at diagnosis. The post-intervention group consisted of 17 infants whose mothers were offered IBCLC support (one mother with 3 previous children declined to meet with the IBCLC).

Mothers in the pre-intervention group provided BM to their child for an average of 6.0 months (SD 5.5) while those in the post-intervention group provided BM to their child for 6.9 months (SD 4.5) (p=0.63).

Among mothers in the post-intervention group, 16/17 (94%) continued to provide BM past the 2nd visit (first visit after diagnosis) while only 8/14 (57%) in the pre-intervention group continued to do so after the first visit (p=0.03).

There was no change in weight z-score (WAZ) or weight-for-length z-score (WFLZ) at ages 6 or 12 months when compared with WAZ and WFLZ at diagnosis (all comparisons p value > 0.05). The WFLZ at 6 months was a mean of -0.02 (SD 1.07) and 0.14 (SD 1.02) at 12 months.

Mothers of infants with CF were also surveyed about their breastfeeding experience and 83% indicated they planned to breastfeed a minimum of 9 months or longer; but only half met this goal. Half of the respondents agreed that the stress of learning her baby had CF affected her BM supply and CF cares made her feel overwhelmed. Mothers commented that the IBCLC provided suggestions to optimize milk intake and regain self confidence in breastfeeding under stressful circumstances.

Discussion: Our small study indicates two important trends. Firstly, mothers in our IBCLC-intervention were less likely to quit breastfeeding after the diagnosis visit and secondly, there was a small increase in the time that BM was provided, although this was not statistically significant. Growth of these children was not affected adversely. The mothers themselves suggested that the IBCLC provided valuable support at a stressful period in their lives. We hypothesize that by providing support to mothers of infants with CF, the IBCLC can prolong the duration that infants are provided with BM. Weight caught up to healthy children by eight months, but height did not. We hypothesized that growth impairment during preschool ages in CF is associated with a reduced adult height.

Methods: The CF Foundation Patient Registry was used in a retrospective case-control manner to identify adults with CF aged 18-19 years between 2011 and 2015 with any height measurements (standardized to CDC height for age percentiles) and retrospectively ascertained their height for age percentiles between the ages of 2 and 4 years.

Results: Among 3, year olds with CF, only 26% were at or above the median height for their age and 25% (n=915) had a height percentile below the 10th percentile in adulthood. Between the ages of 2 and 4 years, 29% (1034/3566) had at least one height percentile measure below the 10th percentile, and 54% of those short in early childhood went on to be below the 10th percentile in early adulthood (an increased odds of 7.7 [95% CI=6. 5 -9.1] compared to children with CF between ages 2-4 without a height below the 10th percentile). Among the children consistently below the 10th percentile between 2 and 4 years, 67% were below the 10th percentile as adults (an odds ratio of 9.0 [95% CI=7. 3-11.0] ). Among the children with CF between the 10th and 25th percentiles at ages 2-4 years, 58% of them were below the 25th percentile as adults (with 24% below the 10th). Only 13% of children between the 10th and 25th percentile for height at age 2-4 years were greater than the 50th percentile as adults. The maximum height measurements between ages 2-4 were highly correlated with maximum height at age [18] [19] [95% CI= 0.62, 0.66] (Kramer MS, et al. JAMA. 2000; 284:843-9) . They are at greater risk for feeding intolerance after birth, which translates to poor growth over the first years of life (Santos IS, et al. BMC Pediatrics. 2009; 9:71) . The impact of preterm birth on growth in infants with cystic fibrosis (CF) has not been analyzed.

Objective: The goal of this study is to compare growth in the first two years of age for children with CF born late preterm (34-37 weeks gestation) and those born full-term (>37 weeks). We hypothesize that children with CF born late preterm will be more likely to have growth failure at two years of age compared to term infants.

Methods: We used data from the US Cystic Fibrosis Foundation Patient Registry. We included children diagnosed with CF by newborn screen (NBS) between 2010 and 2013. Information collected included clinical characteristics, demographics, nutritional indices at two years of age, and various health variables (Table) . The primary outcome was weight-forlength (WFL) percentile at two years.

Results: A total of 2492 infants born with CF were included. 165 (7%) were born late preterm. The majority of all infants were male, Caucasian, and had commercial insurance. Most of the infants had a severe mutation (class 1-3) and were on pancreatic enzymes by age two. A larger proportion of late preterm infants were supplementing with calorie dense feeds (72% vs 66%, p=0.2) and had feeding tubes (28% vs 15%, p<0.05).

A univariate analysis showed more late preterm infants were hospitalized between birth and age two compared to term infants (OR 1.43, 95% CI 1.01-2.03). The majority of all infants were greater than the 50th percentile WFL at age two (55% and 57%). Late preterm infants had similar odds of having a WFL above the 50th percentile at 2 years compared to late term infants (OR 0.92, 95% CI 0.65-1.30).

Conclusion: There are differences in feeding practices among late preterm and term infants with CF but no difference in WFL at age two. This suggests late preterm infants catch up to their term peers with CF, possibly in part due to more agressive feeding regimens. Future research should determine when these infants catch up to their peers, if this is sustained during childhood, and how this influences lung function. Introduction: Failure to thrive is common in infants diagnosed with cystic fibrosis (CF). Newborn screening (NBS) for CF is thought to be beneficial in improving nutritional status and growth of patients. In Norway, CF was included in the national newborn screening program in 2012.

Objective: The aim of this study was to evaluate growth in young children with CF, 0-2 years of age, after the advent of NBS. The growth in NBS-children was compared to the Norwegian growth charts from 2009 and to the growth of patients diagnosed by clinical symptoms.

Methods: Only children born at full-term were included. They were divided into three groups; children diagnosed in the NBS program, children diagnosed with meconium ileus (MI) (2006) (2007) (2008) (2009) (2010) (2011) (2012) (2013) (2014) (2015) (2016) , and children diagnosed traditionally based on clinical symptoms < 2 years of age (2006) (2007) (2008) (2009) (2010) (2011) . We collected data from the Norwegian growth chart for each child; weight, length and BMI are reported as standard deviation scores (SDS) from the 50th percentile for age and gender at birth, 1 year and 2 years of age.

All statistical analyses were performed using IBM SPSS software. T-test, ANOVA-test and linear regression were used. P-values ≤ 0.05 were considered statistically significant.

Results: A total of 23 young children diagnosed by NBS were included; 20 of these had reached 1 year of age, and 15 had reached 2 years of age at inclusion. Ten children with MI (2006 MI ( -2016 , and 18 children with other clinical symptoms diagnosed before NBS were included at 2 years of age.

Children with MI had reduced growth compared to children diagnosed by NBS and young children presenting with clinical symptoms before 2 years of age (Table) . SDS-BMI was not significantly different between the three groups. We also performed regression analyses in the group of children without MI, controlling for pancreatic insufficiency (PI) and screening. Screening, but not PI was significantly (p=0.027) associated with increased SDS-length at 2 years of age. SDS-BMI and SDS-weight were not associated with either PI or screening.

Children diagnosed by NBS (n=23) had significantly lower weight and length at birth compared with the 50th percentiles for Norwegian growthcharts (SDS-weight -0.81 (p=0.001), SDS-length -0.89 (p=0.025)). At 1 year of age (n=20), mean values for weight, length and BMI did not differ significantly from the 50th percentiles, but at 2 years of age (n=15), weight and BMI, but not length, were significantly lower than the 50th percentiles (SDS-weight -0.29 (p=0.045), SDS-length 0.21 (ns), SDS-BMI -0.57 (p=0.051)).

Conclusions: NBS seems to be beneficial in improving nutritional status and growth in young children < 2 years of age in CF. Children diagnosed in the NBS-program did not attain normal growth parameters for same age and gender. Our group of young children diagnosed by NBS is small, and the data should be further explored in the future. CRSA, Sorbonne Université, INSERM, Paris, France; 3. Institut Pierre Louis D'epidemiologie et de Sante Publique, Sorbonne Université, INSERM, Paris, France; 4. Pediatric hepatology unit, Hôpital Necker Enfants Malades, Paris, France Introduction: Cystic fibrosis-related liver disease (CFLD) is a common symptom in patients with cystic fibrosis (CF). However, its prevalence, risk factors, and evolution are unclear. We analyzed a large database of patients with CF to investigate the incidence of CFLD and related risk factors.

Patients and Methods: We retrospectively analyzed 3,328 CF patients with pancreatic insufficiency born after 1985 and recruited into the French CF Modifier Gene Study since 2004. We determined liver status, age at CFLD and severe CFLD onset, sex, CFTR genotype, history of meconium ileus, treatment with ursodeoxycholic acid (UDCA), and respiratory and nutritional status.

Results: The incidence of CFLD increased by approximately 1% every year, reaching 32.2% at age 25. The incidence of severe CFLD increased only after the age of 5, reaching 10% by age 30. Risk factors for CFLD and severe CFLD were male sex, CFTR F508del homozygosity, and history of meconium ileus. Increasingly precocious initiation of UDCA treatment did not change incidence of severe CFLD in patients. Finally, patients with severe CFLD had worse lung function and nutritional status than other CF patients.

Conclusion: CFLD not only occurs during childhood, but also later in the lifetime of patients with CF. Male sex, CFTR F508del homozygosity, and history of meconium ileus are independent risk factors for CFLD development. Earlier use of UDCA over the last 20 years has not changed the incidence of severe CFLD, questioning the use of UDCA in young children given its possible adverse effects. 1 1. Pulmonology, Nemours Children's Specialty Care, Jacksonville, FL, USA; 2. Gastroenterology & Nutrition, Nemours Children's Specialty Care, Jacksonville, FL, USA Background: Good nutritional status is associated with better outcomes in cystic fibrosis (CF). The combination of pancreatic insufficiency, frequent pulmonary exacerbations, and malnutrition make it challenging to obtain optimal nutritional status by oral intake alone. In some individuals, enteral tube feeds can help achieve improved nutritional status. As our CF center's median BMI is less than the national average, we decided to examine our process for recommending gastrostomy tubes (G-tube).

Objectives: To improve our understanding of parents' perceptions about G-tube placement and possible barriers as an option to improve nutritional status. The secondary objective was to analyze pre-and post-G-tube placement anthropometric data to determine if the intervention was effective in improving nutritional status.

Methods: We reviewed our current practice of approaching patients at risk or in nutritional failure. To obtain information on parents' perception of our process of approaching the need for G-tube placement we created a survey using open-ended and multiple-choice questions. These were administered electronically during clinic visits. Parents of 12 patients with G-tubes were surveyed. Pre-and post-G-tube nutritional data were collected by retrospective chart review.

Results: Sixty-five percent of patients were female. Fifty-eight percent of patients had two copies of ΔF508 gene. Fifty-eight percent of patients were considered to have good adherence with using the G-tube per clinic recommendations. The average age of placement was 6.8 years (6 mo to 15y). The average age of first discussion was 5.6 years. Average length of time from discussion to placement was 7.4 months (1 mo to 1.85 y). One-third of parents felt somewhat bad at the time of the first G-tube discussion. Twenty-five percent of patients felt neutral, 16.7% felt relieved and 16.7% felt somewhat relieved. Benefits reported to having the G-tube were related to weight gain. Stressors reported include maintenance, time consuming, and patient dislike. Themes that emerged included adding to nightly routine, avoiding certain sports and some clothing changes.

A paired samples t-test was used to compare the mean BMI z-score before and after G-tube placement. A significant improvement in BMI z-score was seen 6 months (p = 0.024, n = 12) and 1 year (p = 0.029, n = 10) after G-tube placement. A paired samples t-test was calculated to compare the mean pre-G-tube placement. There was no significant difference found in BMI z-score 3 years after G-tube placement (p = 0.061, n = 8).

Conclusions: The qualitative survey results indicate parents experienced a range of reactions to the initial G-tube discussion. We can conclude that at the end of the process, parents were satisfied with the process and felt adequately informed. Use of the G-tube in the first year resulted in improved nutritional status. The lack of significant results at 3 years may be due to less adherence over time. Further investigation is needed to determine appropriate patient candidates, factors contributing to nonadherence with recommendations and reasons for delay from first discussion to actual placement. Harris, A.; Evans, M. Nutrition, Royal Children's Hospital, Melbourne, VIC, Australia Background: Children with cystic fibrosis (CF) who are at risk of poor nutritional status can consequently be at risk of low lung function. The nutritional management of CF requires strict monitoring of dietary intake to ensure adequate growth, CF-specific estimated nutritional requirements are met and appropriate enzyme distribution. Our current practice requires patients to return a 3-day food diary prior to their annual review. We aimed to investigate compliance with nutritional screening recommendations, nutritional status and identify those patients "at nutritional risk" <10 years of age in our clinic.

Method: Retrospective data were collected for all patients aged between 2 to 10 years during 2015 and 2016. Growth data were collected from routine clinic visits and converted to z-scores. Three-day food diaries were analysed for energy, protein and fat intake and results were compared to Australasian Nutrition Guidelines for CF. Patients were excluded if they had comorbidities that are known to affect growth. Patients were deemed "at risk" if BMI was <25th percentile or weight loss occurred in the preceding 2 months. Statistical analysis was performed using SPSS.

Results: Data were collected on 127 patients (73 male, 54 female). In total 121/127 were classified pancreatic insufficient (PI) and 6/127 were pancreatic sufficient (PS). Of these, 12/127 patients were deemed at nutritional risk. Food diaries were returned and analysed for 33/127 patients, 2/12 from at risk patients. All PS patients failed to return food diaries. Mean age 80.2 ± 34.9 months (range: 24.0-134.7), mean weight z-score 0.11 ± 0.8 (range: -2.2-3.0), height z-score -0.02 ± 0.9 (range: -2.4-3.1) and BMI z-score 0.20 ± 0.7 (range: -1.7-2.7). For patients that returned food diaries, mean energy intake was 127.7% (range: 95.9-185.3) of estimated energy requirement (EER), 128.5% (range: 94.5-180.1) of their estimated protein requirement (EPR) and 120.3% (range: 68. 9-192.5) of estimated fat requirement (EFR). No significant relationship was observed between BMI z-score and energy intake (r= -.17, p= .925), protein intake (r= .094, p= .603), or fat intake (r= -.051, p= .777). In males only, age was significantly correlated to energy (r= .620, p= .006), protein (r= .614, p= .007) and fat (r= .485, p= .041). No significant relationship was observed for age and BMI z-score (r= -.62, p= .245) in the total population. Weight z-score and BMI was significantly correlated in PS patients (r= .915, p= .011). No significant relationship was observed for BMI z-score and dietary intake for the at-risk group.

Conclusion: Overall, patients were noncompliant with dietary monitoring requirements with few food diaries returned. However nutritional status was adequate in our cohort with a small number of at-risk patients identified. There were insufficient data to confidently be able to determine significant relationships between anthropometric outcomes and dietary intake. As a consequence we will review our nutritional screening protocols to determine the most effective and reliable method of collecting accurate dietary data in this population. This will assist us in providing meaningful dietary interventions to our at risk population. Introduction: Somatic growth in children with cystic fibrosis (CF) is strongly associated with pulmonary and clinical outcomes. Unfortunately, achieving caloric intake sufficient for growth is often difficult. Attempts to improve dietary intake in CF, including behavioral interventions and liquid nutritional supplements, variably promote weight gain. Development of alternative, palatable, nutrient-dense sources of supplemental calories may benefit children with CF as it could plausibly improve their nutritional status and, secondarily, pulmonary outcomes. Ready-to-Use Supplemental Food (RUSF), composed of peanut butter paste, flour, milk powder, sugar and oil, is an inexpensive, palatable, energy-dense, and effective treatment of undernutrition in low and middle-income countries. We sought to develop a RUSF as a source of supplemental nutrition for children with cystic fibrosis and determine its palatability and tolerability.

Methods: Using a linear programming tool, we developed six custom CF-RUSFs with sufficient caloric density (495kcal/100g), protein, and fat for children with CF. A taste trial was performed to identify the two most palatable CF-RUSFs as measured by dichotomous likability and rank score. We then determined acceptability of the CF-RUSF through a two-week crossover acceptability study. Two randomly assigned cohorts were provided CF-RUSF for 1 week, after which they received the crossover lead candidate for 1 week. Families were instructed to provide CF-RUSF equivalent to approximately 20% of recommended caloric intake, but were allowed to present the CF-RUSF in the manner the child preferred (plain, on bread/toast, with fruit, etc). Daily intake, gastrointestinal symptoms, overall acceptability by 5-point hedonic scale, and CF-RUSF preference were measured.

Results: For the taste trial, 9 children with CF, age 3-10 years with mean BMI percentile of 39% (three with gastrostomies) were enrolled. The two lead candidate CF-RUSFs were liked by at least 50% of the children with an average rank out of 6 of 4.1 (RUSF-1: standard unflavored) and 3.8 (RUSF-2: cinnamon flavored). The acceptability study enrolled 10 children (age 3 to 12 years, four with gastrostomies) with a mean BMI of 49% (SD and mean weight-for-age percentile 34%) into the acceptability study. Adverse symptoms including mild diarrhea and mild abdominal pain were reported by 1 of 10 participants, no child developed constipation or vomiting. Participants consumed CF-RUSF on an average of 60.4% (SD 31%) of study days. Mean caloric intake from RUSF was 170.36 kcal (SD 172.9) . Average acceptability on the hedonic scale was 4 (IQR 1.8) for both RUSFs, although 63% of participants indicated a preference for the standard unflavored RUSF. There was no significant difference in RUSF consumption between either RUSF or week 1 and 2 of the study.

Conclusion: CF-RUSF is acceptable and well tolerated by all children with cystic fibrosis. Further studies will seek to determine acceptability as compared to supplements and effect on weight.

to -0.2 kg weight change pre-intervention, a more than 5-fold increase. The trend toward increased weight gain was sustained at 6-month follow-up, although the weight gain (0.8 kg) was somewhat attenuated 3 months after the intervention had ended. None of the weight changes reached statistical significance, likely due to the small sample size.

Conclusion: Delivering HCHP snack-food boxes to patients' homes was associated with a weight gain at 3-and 6-month follow-up. Providing HCHP snack boxes to CF patients at increased nutritional risk may help achieve optimal nutrition. The intervention needs to be tested in a larger sample.

Acknowledgment: This project is funded by a QI grant from the CF Foundation.

Nutritional characteristics of participants pre-and postintervention (N=6), Mean (SD) In clinical trials, ivacaftor (IVA) and lumacaftor/ ivacaftor (LUM/IVA) have been shown to significantly improve lung function (i.e. FEV1), decrease the frequency of CF pulmonary exacerbations and improve body mass index (BMI) in people with CF who carry specific CFTR mutation(s). Upon starting CFTR modulators, the Minnesota (MN) CF Center observed weight gain in excess of what is reported in the original clinical trials. Although optimal nutrition is key to longevity, excessive weight gain may negatively impact lung function. Therefore, we sought to determine if CFTR modulator therapy was associated with excessive weight gain.

Objective: To compare weight gain over time in those patients with CF started on IVA or LUM/IVA with those who were not treated with a CFTR modulator. Our secondary objectives were to evaluate the correlation between CFTR modulator use and pulmonary outcomes (lung function and pulmonary exacerbations).

Methods: Using an internal database, a retrospective, observational study was conducted at the MN CF Center to compare weight gain over time in people with CF who started CFTR modulator therapy versus those who qualified but did not start the drug. Demographics, anthropometric measurements (i.e. height, weight, BMI) and lung function data were collected from the initiation of therapy to present. Linear mixed effects models were used for the longitudinal weight change, with status on IVA or LUM/IVA modeled as a time-varying variable. Interaction between age and IVA or LUM/IVA status was included to estimate the modification effect of CFTR modulator therapy on weight change. An age 2 term was also included to model nonlinear weight changes. Analyses were stratified by age: <18 and 18. Statistical models were adjusted for sex, hospitalizations, home IV courses, CF related diabetes, FEV1, pancreatic status, and Pseudomonas aeruginosa status.

Results: Of 100 patients who qualified for IVA, 24 were ever on IVA (10 patients <18). Of 352 patients who qualified for LUM/IVA, 182 were ever on LUM/IVA (110 patients <18). Weight change over time was observed to be nonlinear up to age 18, particularly in patients on therapy. After adjusting for covariates, there was no statistically significant difference in weight change over time in patients who were on IVA compared to those who were not. In contrast, there was a significant association between weight change and LUM/IVA status among patients < 18. Accelerated weight gain was observed after initiation of therapy with an estimated effect of 2.827 pounds per year more than those were not on therapy (95% CI:

1.641-4.013, p-value=3.0e-6). However, a negative interaction between age 2 and LUM/IVA was estimated: -0.118 (95% CI: -0.165 --0.0704, p-value=1.0e-6), suggesting that the weight gain difference post-therapy slows over time. Further analysis will be complete at the time of abstract presentation.

Conclusion: Pediatric patients with CF showed significantly accelerated weight gain after initiation of therapy with LUM/IVA. Patients on IVA alone did not replicate this finding, likely limited by sample size.

Patterson, R. 1 ; Perrin, F.M. 2 ; Joshi, D. 3 ; Hurt, K. 4 1. Brighton and Sussex Medical School, Brighton, United Kingdom; 2. Respiratory Medicine, King's College Hospital, London, United Kingdom; 3. Institute of Liver Studies, King's College Hospital, London, United Kingdom; 4. Brighton and Sussex University Hospitals, Brighton, United Kingdom Background: Cystic fibrosis-associated liver disease (CFLD) is the most common cause of extrapulmonary mortality in CF, responsible for 2.5% of deaths. Liver disease can be broadly categorised into steatosis or cirrhosis, with or without associated portal hypertension (PHT).

Diagnosis of CFLD is challenging, as the clinical hallmarks of disease often present late. It is therefore difficult to make a diagnosis prior to the onset of significant disease. Abnormalities in laboratory tests are also poor predictors of CFLD and liver biopsy is generally avoided, due to the focal nature of CFLD and the risk of life-threatening complications.

As part of the diagnostic process for CFLD, transient elastography (TE) is increasingly being used. This form of TE is a validated part of diagnosis and monitoring in other liver diseases, such as viral hepatitis and nonalcoholic steatohepatitis. The aim of this study was to evaluate the role of TE in assessing disease severity in patients with CFLD.

Methods: The Debray criteria (Debray D, et al. J Cyst Fibros. 2011; 10(suppl. 2) :S29-S36) were used to identify patients with CFLD. Patients were included if they met two or more of the criteria. Patients were then classified into subtypes using clinical findings, laboratory investigations, ultrasound findings and oesophagogastroduodenoscopy results. TE results were also collected (IQR, success rate, type of probe used).

The Mann-Whitney U test was used to compare the liver stiffness measurements (LSMs) for the cirrhotic and noncirrhotic populations with CFLD. A receiver-operator curve analysis with 95% confidence interval was also calculated to assess the diagnostic accuracy for the detection of CFLD. Once more of the cohort has received a Fibroscan it may be possible to extend this analysis to a comparison between all identified subtypes of CFLD.

Results: There were 225 patients registered at the single centre, n=66 (29%) were defined as having CFLD (n=32 (48%) were male). Of these, n=6 (9%) had received an orthostatic liver transplant. Cirrhosis was identified in n=15 (23%) patients, of whom n=7 also had PHT. Heterogenous changes were observed in n=9 (14%) of patients, whilst fatty changes were found in n=26 (54%) patients.

To date, 12 (18%) patients have undergone TE assessment. Of these, half have cirrhotic changes, whilst the other half do not. Mean LSM among the cirrhotic patients was 30.0 ± 26.0 kPa. Mean LSM among noncirrhotic patients was 5.6 ± 1.4 kPa. LSM of > 9.0 kPa predicted cirrhosis with 100% specificity and sensitivity.

The preliminary results have demonstrated that TE is a useful tool for identifying cirrhosis in CFLD. Completion of data collection should allow for the analysis of the use of TE in different subtypes of CFLD. Background: Our group has previously shown that children with CF consume significantly more energy-dense, nutrient-poor (junk) foods than controls. There is a concern that this dietary pattern may result in suboptimal intake of essential vitamins and minerals (micronutrients) needed for growth and development. In addition, there are no comprehensive and contemporary dietary studies in the paediatric CF population that address this issue.

Objectives: To evaluate micronutrient intake in children with CF against recommended dietary intakes (RDIs) and compared with age-and sex-matched controls.

Methods: Children with CF and matched controls between 2-18 years old were recruited from Sydney Children's Hospital, Australia. All children or their carers completed a validated food frequency questionnaire (The Australian Child and Adolescent survey) which measures habitual intake for the previous 6 months. Dietary data were reported on for 13 micronutrients. Absolute intake as well as energy-adjusted intake (amount of micronutrient per 1000 kcals) was calculated for each.

Results: A total of 82 children with CF and 82 controls were recruited. Children with CF consumed significantly more total energy than controls [3142 (2531-3822) kcal/day vs 2216 (1660-2941) kcal/day; p < 0.001]. Absolute micronutrient intake in children with CF was significantly higher in all apart from vitamin C and folate. However after adjusting for energy, controls consumed significantly more of all micronutrients except vitamin A, sodium, calcium and phosphorous. In addition, children with CF of preschool age consumed more riboflavin, folate, potassium, magnesium, calcium and phosphorous per 1000 kcals than children with CF of school age. Children with CF failed to meet RDIs for certain key micronutrients: folate (26.8%), iron (15.9%), calcium (9.8%), magnesium (8.5%) and potassium (6.1% of AI).

Conclusion: Despite children with CF consuming more total energy and absolute micronutrient quantities than controls (apart from vitamin C and folate) a concerning proportion still fell short of requirements for several key micronutrients. Furthermore the micronutrient density of the CF diet is less than that of controls and appears to decrease once the child is of school age. This study has reinforced the importance of diet quality alongside quantity. Introduction: Cystic fibrosis (CF) is associated with impaired nutrient absorption in the setting of increased energy expenditure, and lower body mass index is associated with worse lung function. Parenteral nutrition (PN) may be a useful tool in optimizing nutritional status. While consensus recommendations exist to support the safe and effective use of parenteral nutrition in the general population, there is a paucity of high quality evidence to identify best practices in patients with cystic fibrosis. The objective of this study is to characterize the patterns, complications, and benefits of parenteral use in children and adolescents admitted to a single large CF center.

Methods: A retrospective review of patients between the ages of 0 to 19 years with a diagnosis of CF admitted between January 1, 2014 and April 1, 2017 was conducted utilizing the electronic medical record. Patients who received PN at any point during an admission were enrolled. Those with insufficient data related to the objectives of this study were excluded. Data collected for each episode of PN included patient demographics, enteral feed characteristics, estimated energy needs, indications for initiation and discontinuance of PN, complications of PN, and serial anthropometrics.

Results: Twenty-five episodes of inpatient PN use were identified in 18 patients, with a mean age at admission of 14 years, (range of 1.5 to to 19.4 years). The average BMI percentile for age was 16.6 (range: 0 to 62.6). An indwelling enteral tube was present in 60% of the episodes; an indwelling central line was present in 48%. Bronchopneumonia and/or malnutrition was documented in 68% of admission diagnoses with CF liver disease and/or CF-related diabetes in 40%. The most common indication for PN was as supplementation to oral intake (56%) followed by poor gastrointestinal tolerance (44%). A nasogastric tube was offered and declined in 28% of episodes. The median number of days to PN initiation was 6 (range: 0 to 103 days). The median duration was 9 days (range: 1 to 77 days). PN delivered an average of 57% of estimated energy needs and 87% of estimated protein needs. Complications attributed to PN were documented in 72% of the episodes and included line infection, electrolyte abnormalities, elevated triglycerides, elevated BUN, and glycemic control issues. In 1 episode, PN temporarily discontinued due to a complication attributed to PN (line infection). Eighty percent of episodes with PN ended when adequate enteral intake resumed. Pre-and post-PN weights were available for 13 of the episodes, with an average weight gain of 1.7 kg. Weight gain was not seen in PN durations less than 4 days.

Conclusion: The patterns of inpatient use of parenteral nutrition in children and adolescents with cystic fibrosis in this large CF center are highly individualized and markedly variable. Complications attributed to PN were common but rarely led to its premature or temporary discontinuance. The potential benefits of PN use in acute illness in CF warrants further investigation, with particular attention to whether or not significant weight gain is achieved and/or maintained. This study highlights an opportunity for the development of a focused guideline for the use of PN in this population.

Kent, D.S. 1 ; Remer, T. 2 ; Blumenthal, C. 3 ; Egert, S. 4 ; Gaskin, K.J. 1 1. Gastroenterology, James Fairfax Inst. of Paed Nutr and Univ. Sydney, Children's Hospital at Westmead, Westmead, NSW, Australia; 2. Univ. of Bonn, DONALD Study, Dortmund, NRW, Germany; 3. Nutr. and Diet, Westmead Hospital, Westmead, NSW, Australia; 4. Univ. of Bonn, Bonn, NRW, Germany Background: Infants identified by neonatal screening programs as having cystic fibrosis (CF) need further evaluation to determine their pancreatic phenotype as those with fat malabsorption i.e. pancreatic insufficient (PI) require oral pancreatic enzyme replacement therapy (PERT), whereas those with sufficient endogenous pancreatic enzyme secretion with normal fat absorption are pancreatic sufficient (PS) and PERT is unnecessary. The gold standard method for the indirect assessment of pancreatic function status in infants is a 72-hour faecal fat balance test. However, most centres globally refuse to do faecal fat balance tests for a variety of obvious reasons and are looking for possible alternatives. The 13 C mixed triglyceride (MTG) breath test has been proposed for the noninvasive assessment of fat digestion and absorption. Using mass spectroscopy (MS) to analyse samples this has proved a very sensitive but complex and expensive method. Nondispersive isotope selective infrared spectrometry (NDIRS) developed during the 1990s has been used as a potential substitute for the more expensive MS technique. However, using NDIRS technique in a group of 54 healthy infants of less than 5 months of age, we have found NDIRS technique to be too insensitive. Therefore we wanted to investigate whether with age increase in the first two years of life this test would improve and gain sensitivity.

Aim: The aim of this study was the longitudinal assessment of these babies to validate the potential of the 13 C MTG breath test with NDIRS technique as suitable assessment of fat absorption in the first two years of life.

Methods and Patients: A group of 53 babies were longitudinally assessed with the 13 C MTG breath test and NDIRS technique during babies first two years of life with at least three consecutive breath tests. Babies 4 1. National Heart and Lung Foundation, Imperial College London, London, United Kingdom; 2. Paediatric Respiratory Medicine, Royal Brompton Hospital, London, United Kingdom; 3. London School of Hygiene and Tropical Medicine, London, United Kingdom; 4. Division of Respiratory Medicine, Hospital for Sick Children, Toronto, ON, Canada Introduction: Early nutrition and somatic growth in cystic fibrosis has been shown to be an important predictor of lung function (forced expiratory volume in 1s, FEV1), from age 5 years. The aim of this study is to describe the relationship between trajectories of early growth, and a patient's first lung function measurement.

Methods: Data from two cohorts drawn from the UK and Canadian CF registries are used. Patients born from the year 2000 with at least one weight and one FEV1 measurement were included. Early growth was measured by weight for age z-scores (WFA). Longitudinal weight measurements from age 1 to 5.5 years were summarised using a mixed effects linear regression with random intercepts (WFA at age 1) and slopes (rate of change in WFA over time). Patient specific intercepts and slopes were then used to predict first FEV1 percent predicted (FEV1%). Other predictors included: sex; year of birth; socio-economic status; genotype class; method of diagnosis and age at first Pseudomonas (Pa) infection.

Results: Individual growth trajectories were estimated for 1964 and 699 patients in the UK and Canadian registries respectively. In both cohorts, WFA at age 1 and the rate of change in WFA over time between the ages of 1 and 5.5 were significant predictors of first FEV1% (see Figure for UK results). The greater the child's WFA at 1 year, the greater his or her first FEV1%. Furthermore, steeper WFA decline was associated with worse first FEV%. Further factors which were independently (and negatively) associated with first FEV1% included: low socio-economic status; younger age at first Pa infection and genotype class ("high" risk rather than "low" or missing). The final prediction models were similar between the UK and Canada, with the exception that age at first Pa was not significantly associated with first FEV1%, and a difference between males and females was found.

Conclusion: Patient weight at age 1, as well as the rate of change in weight in the first five years are significant predictors of first FEV1% in two contemporary, largely newborn screened populations. These findings further emphasize the importance of monitoring growth and nutrition from diagnosis and interventions aimed at maintaining healthy weight. Objectives: The impact of lumacaftor-ivacaftor (LUM-IVA) on body composition and nutrition support usage is not widely studied, particularly in those with severe lung disease. We aimed to evaluate changes in body composition in adults with CF with FEV 1 < 40% predicted over 12 months of LUM-IVA treatment. We also aimed to compare weight and nutrition support usage in the year prior to commencing LUM-IVA, with that during the first year of treatment.

Methods: Data were analysed for 24 CF adults (13 male, baseline age (mean±SD): 33.5±8.3 years, FEV 1 %predicted: 34.7±7.4%, BMI: 20.1±2.6kg/m 2 ), who received LUM-IVA for 12 months under a managed access program. Body composition (tetrapolar multifrequency BIA, SECA, Germany) was measured at 0 (baseline), 1, 6 and 12 months. Absolute changes (Δ) in BMI, weight, fat-free mass (FFM) and fat mass (FM); and %Δ in weight from baseline were determined. Usage of oral nutritional supplements (ONS) and enteral nutrition (EN) during the year prior to commencing LUM-IVA was compared with that during the first year of treatment. Analyses were performed using linear mixed effects regression modelling and Pearson's correlation.

Results: No change in mean weight was seen over the year prior to commencing LUM-IVA treatment (58.4±11.6 kg vs 58.4±12.1 kg, p=0.97). No weight change was seen at one month (p=0.91). Weight increased significantly in the first 6 months (mean±SE %Δweight 0-6 mth: 3.9±1.7%, p=0.03) and was maintained by 12 months (%Δweight 0-12 mth 4.2±1.7, p=0.02). BMI increased significantly by 6 months (0.87±0.25 kg/m 2 , p=0.005), plateauing by 12 months (ΔBMI-6-12 mth: 0.03±0.25, p=1.0). Fat mass increased significantly by 6 months (ΔFM-0-6 mth: 2.4±0.5 kg, p=0.001), with no further increase by 12 months (ΔFM6-12 mth: 0.3±0.5 kg, p=1.0). No changes were seen in mean FFM (ΔFFM0-12 mth: 0.85±1.0 kg, p=0.41), indicating overall preservation of FFM. Lower baseline BMI was associated with higher %weight gain (r= -0.58, p=0.003) and FFM (r= -0.56, p=0.005) at 12 months. Four patients (17%) gained >5% FFM whilst two lost >5% FFM. The % of patients with BMI<18.5 kg/m 2 decreased from 33% at baseline to 13% at 12 months (p=0.003, McNemar's test). Fourteen patients (58%) received ONS and 8% enteral nutrition in the previous year; no change at 1 year. The median number of scripts dispensed for ONS or EN users was unchanged (8 per year) .

Conclusions: BMI, weight and fat mass gains seen after six months on LUM-IVA plateau by one year. In this cohort, underweight patients were more likely to improve BMI and body composition. No change in nutrition support utilization was seen in the first year on LUM-IVA. Mechanisms underlying body composition changes, and reasons for plateau in gains, require further investigation, including improved appetite, salt/hydration or exercise, or via amelioration of catabolism. BACKGROUND: Due primarily to malabsorption, infants with CF have increased nutritional needs that may not be met by prolonged, exclusive breastfeeding. The Adequate Intake (AI) of linoleic acid (LA) and alpha-linolenic acid (ALA) is 4.4 and 0.5 g/day, respectively, based on an average intake of 780ml/day of breast milk in exclusively breastfed healthy infants, ages 2-6 mo. Factoring in 15% fat malabsorption, infants with CF who are exclusively breastfed would require 5.1 g/day of LA and 0.6 g/day of ALA to meet their needs.

OBJECTIVE: To determine the association between maternal diet and breast milk (BM) composition and corresponding intake of LA and ALA in breastfed infants with CF.

METHODS: We utilized data from FIRST (Feeding Infants Right… from the STart), an ongoing multicenter prospective observational study that includes a cohort of 183 infants born from 1/2012 to 12/2017 and enrolled at age 1.8±1.0 mo from six CF centers in the US. Of 183, 79 infants enrolled in FIRST after the BM ancillary study was initiated in 2015; 48 were breastfed and 29 mothers (60%) agreed to participate in the BM study. This analysis includes data from 70 food frequency questionnaires (reflecting maternal dietary intake) and 77 BM samples collected from 20 days to 6.9 mo postpartum from 23 mothers. For exclusively breastfed (exB) infants (n=7), LA and ALA intakes (g/day) were estimated based on LA and ALA concentration in their mothers' BM samples (16 samples) multiplied by 780 ml. For partially breastfed (pB) infants (n=15), LA and ALA intakes were estimated based on BM (proportion of BM to total number of feedings multiplied by 780 ml) plus formula intake recorded in three-day food records (35 records).

RESULTS: Maternal LA intake was 14.9±6.8 g/day with 47% below the AI (13 g/day) and maternal ALA intake was 1.6±0.8 g/day with 36% below the AI (1.3 g/day). These findings are similar to dietary intakes reported from NHANES (National Health and Nutrition Examination Survey). Caloric density of BM was as expected, 19.5±3.6 kcal/oz (range: 12.3-31.6), but LA and ALA contents were low [LA: 4.0±1.7 g, (61%<AI); ALA: 0.3±0.2 g (87%<AI) per 780 ml]. LA in maternal diet and BM samples (expressed as % of total fat) were significantly correlated (r=0.42, p=0.004), but the correlation for ALA was not significant. LA intake in exB infants was lower (3.9±1.9 g/day) compared to that in pB infants (5.5±2.0 g/day), p=0.007. ALA intake in exB infants was also lower than that in pB infants (0.35±0.21 versus 0.44±0.18 g/day, respectively) but the difference was not significant, p=0.10. The prevalence of low LA intake was almost 2-fold high in exB infants (63%<AI) than that in pB infants (34%<AI), p=0.06. A similar trend was observed in ALA intake (exB: 88%<AI, pB: 66%<AI), p=0.18. CONCLUSIONS: Maternal LA and ALA intake were similar to women in NHANES but the majority of breast milk samples were low in LA and ALA. Supplemental formula intake reduced the prevalence of low dietary LA by about half and ALA by a quarter, and represents an effective strategy for improving essential fatty acid intake in breastfed infants with CF. Lung function decline is correlated strongly with nutritional status. CFTR modulator therapies are designed to correct the malfunctioning protein made by the CFTR gene. Currently, three CFTR modulators have been approved by the FDA: Kalydeco (ivacaftor), Orkambi (ivacaftor/lumacaftor), and Symdeko (tezacaftor/ivacaftor). Age and CFTR mutations limit patient eligibility for treatment with CFTR modulators. Studies have shown an improvement in nutritional status following treatment with ivacaftor or ivacaftor/lumacaftor.

Objective: The purpose of this review was to examine nutritional outcomes in CF patients receiving a CFTR modulator in a postclinical trial setting and compare outcomes between patients who received ivacaftor and ivacaftor/lumacaftor at Children's Healthcare of Atlanta at Scottish Rite.

Methods: An IRB-approved, single-center, retrospective analysis was performed for patients treated with CFTR modulators from January 2012 to April 2018. Height, weight, body mass index (BMI) percentile, vitamin D levels, duration of therapy, and pancreatic enzyme dose were evaluated both pre-and post-modulator therapy.

Results: Fourteen of 22 (63.6%) eligible patients are currently receiving ivacaftor, and 29 of 54 (53.7%) eligible patients are receiving ivacaftor/ lumacaftor at Children's Healthcare of Atlanta at Scottish Rite. The mean age of our patients was 9.45 years, and they received CFTR modulator therapy for a mean duration of 21 months. Significant increases in height and weight were observed following treatment. Additionally, a significantly lower dose of pancreatic enzymes was utilized (see Table) . No differences were observed in BMI percentile or Vitamin D level following treatment. When comparing modulators, a significant increase in Vitamin D level was noted in patients treated with ivacaftor in comparison to ivacaftor/lumacaftor (median 9.4 unit vs. -0.1, p=0.019). A trend towards greater decreases in pancreatic enzyme dosing was also noted in the patients on ivacaftor (379 unit vs 189; p=0.11).

Conclusion: The results of this review lend support to prior evidence that CFTR modulator therapies may improve nutritional outcomes. These results also suggest there may be an improvement in pancreatic function related to the use of CFTR modulators as indicated by a decrease in pancreatic enzyme dose post-modulator therapy. Due to the small sample size and retrospective nature of this review, a larger, multicenter study on CFTR modulators and the impact on nutrition outcomes and pancreatic function is necessary to further support these findings. Introduction: Cystic fibrosis (CF) is caused by mutations in the cystic fibrosis transmembrane conductance regulator (CFTR) gene leading to loss of functional anion channels. Low weight-for-age percentiles and body mass index of CF patients strongly correlate with lung function and survival. CF patients also exhibit elevated energy expenditure and preferential reliance on carbohydrates. Despite high-calorie high-fat diets coupled with pancreatic enzyme replacement therapy, CF patients struggle to maintain a healthy BMI. Studies also showed that CF adolescents with good lung function and nutritional status still manifest elevated energy expenditure and low exercise performance indicating intrinsic nature of energy inefficiency in CF. We and others demonstrated that CF mouse models display similar energy imbalance, including reduced energy intake and storage, and high energy expenditure, all contributing to reduced growth that is independent from apparent malnutrition, exocrine pancreatic insufficiency, and progressive lung disease. Our further investigations of metabolic alterations in CF mice revealed abnormally low fasting glucose levels and high ketone body levels possibly indicating difficulty of maintaining normoglycemia (e.g. inefficient gluconeogenesis) and high lipid oxidation. Akin to high-fat high energy diets prescribed to CF patients, CF mice had poor growth and fat accretion outcomes when fed high-fat diet.

Objective: In this work, we investigated whole-body metabolic responses of CF mice to carbohydrate restriction by feeding ketogenic diet short term. This strategy would impose a metabolic shift towards lipid β-oxidation, ketogenesis, and gluconeogenesis to maintain energy balance.

Methods: CF (R117H) and wild-type (WT) mice (N=16/group) were fed ketogenic (89.5% fat, KG) or standard chow (9% fat, CHOW) for 1 week. Changes in body weight, blood glucose and ketones were measured before and after respective diet feedings. Indirect calorimetry data (IDC) plan, we wanted to assess if providing a single high-dose vitamin D 3 will better achieve these recommendations. A similar study had been proven efficacious in improving vitamin D levels in children with CF in Australia; however, this study did not adequately assess safety or feasibility.

Methods: This study is a Phase II clinical drug trial with the purpose of assessing the safety, efficacy, and feasibility of an oral, one-time, high-dose (250,000 International Units (IU)) vitamin D 3 to pediatric patients with CF aged 3 years to 18 years followed in the Johns Hopkins All Children's Pediatric CF Center. Patients were identified as potential subjects based on annual bloodwork and any patient with 25OHD <30 ng/mL was approached for inclusion. Exclusion criteria included any history of kidney disease or stones, previous history of hypercalcemia or hypercalciuria, abnormal creatinine levels 2 times the upper limit of normal (ULN), CF liver disease or cirrhosis, elevated liver transaminases 5 times ULN, parathyroid disorder or pregnancy. Subjects were provided a single dose during CF clinic. Safety was assessed with a serum calcium (Ca) level and a serum phosphorus (Phos) level measured at 1 week and 3 months following the dosage. If Ca or Phos were elevated, a parathyroid hormone level (PTH) and spot urine Ca to creatinine ratio was obtained. To assess feasibility, participants answered a 10-item questionnaire on any gastrointestinal symptoms experienced and overall experience with the ease of the regimen.

Results: Thirty patients have consented for this study with 3 screen failures. Mean age was 15.8 years and 75% were female. Mean 25OHD level at the start of the study was 21.08 ng/mL. One patient experienced a serum Ca level higher than 9.9 mg/dL at 1-week postdose, indicating hypercalcemia, but upon further testing had normalized. Three patients had slightly elevated serum Phos levels at 1 week and after additional safety testing had normalized Phos levels. All had normal fasting PTH levels. One patient had elevated Phos at 3 months with elevated PTH but PTH normalized in a week and Phos normalized. The feasibility questionnaire revealed that all patients found the oral, one-time, high-dose vitamin D 3 easy to take, would take it again, and preferred it over a daily vitamin D 3 supplement.

Conclusions: Oral, high-dose Vitamin D 3 can be safely administered to pediatric patients with CF and is quite feasible to perform in clinic. This should be considered as a treatment strategy for patients with vitamin D insufficiency. We are continuing to collect efficacy data on the remaining patients to aid with designing a large multicenter trial for this treatment alternative. In the general population, poor diet quality contributes to increased risk of cardiometabolic diseases; however, this relationship has not been studied in cystic fibrosis (CF) populations. We sought to describe diet quality and examine its relationship with body composition and fat distribution in adults with CF.

Methods: This prospective, cross-sectional study included 24 adult CF subjects and 25 healthy controls. Analysis of 3-day food records was completed using Nutrition Data System for Research software, and 2015 Healthy Eating Index (HEI) diet quality scores were calculated. Diet variables were adjusted per 1000 kcal. Whole and regional body composition was assessed by dual energy X-ray absorptiometry. Multiple linear regression analyses between dietary and body composition variables were controlled for age, sex, and disease status (CF vs control).

Results: The mean age and sex distribution of CF subjects (28.7 y, 50% male) were similar to controls (mean age 27.1 y, 36% male, all p>0.05). CF subjects reported higher total calorie intake vs controls (2,935 ± 778 vs 2,017 ± 602 kcal/day, p<0.001), although the percentage of daily caloric intake from dietary total fat, carbohydrate, or protein intake did not differ between groups (p>0.05). CF subjects consumed more trans-fatty acids (1.17 ± 0.65 vs 0.73 ± 0.61 g/1000 kcal/day, p< 0.01), added sugars (28.81 ± 14.16 vs 13.43 ± 7.77 g/1000 kcal/day, p< 0.001), and refined grains (82 ± 14% vs 66 ± 29%, p=0.03), and less total dietary fiber (7.4 ± 2.29 vs 12.9 ± 6.32 g/1000 kcal/day, p< 0.001) and whole grains (18 ± 14% vs 34 ± 29%, p=0.03) than healthy controls. CF subjects also had a higher glycemic load (286.6 ± 82.9 vs 156.4 ± 55.0, p< 0.001) and lower HEI scores (47.3 ± 8.7 vs 63.1 ± 14.1, p< 0.001) than controls. Among all subjects, added sugar was positively associated with visceral adipose tissue (VAT) (β=0.29, p=0.001) and percent body fat (β= 3.1, p=0.04). Total dietary fiber was inversely associated with VAT (β= -0.31, p=0.008). Refined grains were positively associated with total body fat (β= 6.42, p=0.04). Whole grains and HEI score were inversely associated with total body fat (β= -6.44, p= 0.04 and β= -0.13, p= 0.03, respectively). Dietary intake was not significantly associated with fasting glucose or FEV1% predicted (all p>0.05).

Conclusions: Despite no differences in habitual macronutrient intake, adults with CF reported lower diet quality compared to healthy controls, as indicated by higher intakes of added sugar, trans-fatty acids, and refined grains, and lower intakes of total fiber, whole grains, and lower HEI scores. Furthermore, these novel data implicate poor diet poor quality as a significant predictor of VAT and total body fat in a CF population. The implications of these findings will require rigorous longitudinal studies. Our study emphasizes the importance of diet quality in CF subjects, as opposed to simply focusing on macronutrient quantities. These data may be used to identify nutritional targets for improvement of health outcomes in CF. Introduction: Mostly on cross-sectional data 1-hour high glucose and low insulin values during oral glucose tolerance test (OGTT; G60 and I60 respectively) have been associated with lower pulmonary function and weight.

Objectives: To describe prospectively over a 4-year period the clinical decline (weight and pulmonary function) associated with hyperglycemia and hypoinsulinemia at 1 hour of the OGTT in 2 large Canadian (Montreal) and French (Lyon) cohorts of adult patients with CF.

Methods: Inclusion criteria were patients with forced expiratory volume in one second (FEV1) ≥30% and OGTT G60 and I60 values available at inclusion. Linear mixed models were used to assess the association between annual changes of FEV1 or BMI and G60 at baseline (<11 mmol/l versus ≥11 mmol/L) and I60 (<24 versus ≥24 mU/L 3 ) at baseline and over a 4-year follow-up period. Models were adjusted for age, BMI and Pseudomonas colonization at baseline, gender, CFTR mutation, the year of study entry (2003) (2004) (2005) (2006) (2007) (2008) (2009) (2010) (2011) (2012) (2013) (2014) (2015) (2016) and the group of patients (Canada and France).

Results: The cohort included 119 French patients (mean ± standard deviation (SD): 24.9 ± 6.8 years) and 210 Canadian patients (24.5 ± 6 years) with a mean baseline FEV1 of 63.9 ± 20.8% and 71.4 ± 18.9% and mean baseline BMI of 20.3 ± 2.2 kg/m 2 and 21.4 ± 2.8 kg/m 2 , respectively. The average follow-up time ± SD was 3.5 years ± 3.0. FEV1 at baseline was significantly higher for Canadian patients than French ones (+ 5.3%, 95% confidence interval (CI): 1.1 to 9.4; p = 0.013). The annual rate of decline in the FEV1 was not significantly different between groups and was estimated at -1.0% (95% CI: -1.4 to -0.5). G60 ≥ 11 mmol/L at baseline was associated with a lower mean inclusion FEV1 of −3.5% (95% CI: −1.4 to -0.5, p = 0.082) whereas no significant difference was shown in decline of FEV1 over time even when considering Canadian and French patients. I60 at baseline had no influence on inclusion FEV1 and on FEV1 decline during follow-up.

BMI at baseline was greater for Canadians than for French patients (+1.9 kg/m 2 95% CI: 1.1 to 2.6) but the change over time (+0.2 kg/m 2 , 95% CI: 0.1 to 0.3) was similar in both cohorts. No association between baseline G60 and I60 and BMI at inclusion and over time was shown.

Conclusion: Canadian CF patients had a better pulmonary and nutritional status than French patients at their entry in the cohort, but changes over time were comparable between the groups. Higher G60 and lower I60 values at baseline were not associated with different clinical degradation over time. These data do not support a direct negative role of high G60 and low I60 values on clinical decline of CF adult patients. 2009;32:1626-31) . CFRD is associated with pulmonary function decline and increased mortality (Miller RJ, et al. Can Respir J. 2008; 15:291-4) . In 2010, the CFRD Clinical Care Guidelines were updated, recommending all patient with CF begin CFRD screening at 10 years old. This study uses PortCF registry data of patients with CF to examine the impact of earlier screening practices at centers from 2009-2015.

Methods: The cohort was defined as patients with CF and without CFRD that turned 10 years old from 2009 to 2015. The cohort was identified to examine the impact of the 2010 guidelines recommending a 2 hour OGTT screening test for CFRD of patients entering the screening age. CF clinics with less than 50 patient-years of data were excluded. Screening rates for each clinic were calculated from patients at risk of CFRD; specifically number of patient-years with a CFRD screen divided by the total number of patient-years. Kaplan-Meier curves were used to show the distribution of age at CFRD diagnosis and the differences in curves were tested using the log rank test, with significance level set at 0.05.

Results: Of 189 centers that met inclusion criteria, the average rate of annual screening was 35%. Of 4010 patients screened, 512 (13%) were diagnosed. The screening rate of the patients' clinic was significantly associated with time to diagnosis (p-value=0.0001). Clinics with higher screening rates tended to diagnose CFRD more frequently and earlier ( Figure) . Centers with the lowest (0%-20%), middle (40%-60%), and highest (80%-100%) screening rates, diagnosed 7% (95% CI: 4-10%), 15% (13-18%), and 25% (15-34%) of patients by the age of 15, respectively.

Conclusions: Centers that screen for CFRD more tend to diagnose CFRD at a younger age. An average annual screening rate among this cohort of 35% suggests there are significant opportunities for improvement in early diagnosis of CFRD. Background: CF-related diabetes (CFRD) is the most frequent comorbidity in CF. According to current guidelines insulin is the only accepted medical treatment in addition to dietary advice. Insulin treatment is an additional burden to CF patients who have a very stressful treatment and therefore oral antidiabetic drugs (OAD) are also used at least partially in many countries to treat CFRD. Recently OAD were shown to be as safe and effective as insulin for treatment of early diagnosed CFRD for at least 24 months (Lancet Diabetes Endocrinol. 2018 Feb;6(2):114-21). We wanted to know how newly diagnosed (by annual OGT screening) CFRD is treated in real life.

Methods: Data from the German CF registry from 2004-2016 were used. Newly diagnosed CFRD was considered when CFRD was documented for the first time in the annual report and at least in 2 years before no CFRD was documented. Treatment options were insulin, OAD, dietary advice and no treatment at all. We followed treatment till end of documentation to document changes in treatment (eg OAD to insulin).

Results: Between 2004 and 2016 data from 6105 CF patients were available and 483 (52.6% female) had been newly diagnosed with CFRD. Age (mean±SD) at diagnosis was 23.9±10.0 years. Initial treatment (multiple answers were possible): Insulin 229 (47.4%), OAD 56 (11.6%), dietary advice 41 (8.5%) and no treatment 198 (41.0%). One year after CFRD was diagnosed treatment changed from OAD to insulin in 7 patients, from insulin to OAD in 3 patients, from dietary advice only to insulin in 1 and also 1 to OAD and from no treatment of CFRD to insulin in 43, to OAD in 5 and to dietary advice in 10 patients. Four or more years after initial CFRD diagnosis 25 (44.6%) of patients initially treated with OAD had changed to insulin, 6 (2.6%) from insulin to OAD, 14 (7.0%) from no treatment to OAD, 93 (47.0%) to insulin and 29 (14.7%) to dietary advice. In at least 62 (31.9%) of all patients with initially no treatment there was still no documented treatment of CFRD at end of documentation.

Discussion: These real life data demonstrated that current treatment of CFRD is far away from what is recommended by guidelines. Not only OAD was used as alternative treatment to insulin but more surprising also only dietary advice or even no CFRD related treatment was documented. CF registry data reports (eg 2015 Patient Registry-Annual Data Report from the CF Foundation) showed cross-sectional data including all CFRD patients with no specification of the duration of CFRD. Nevertheless even in this report only 77.6% of all CFRD patients were on chronic insulin treatment and 8% had no treatment at all. There is still room to improve CFRD treatment and it might be that the treatment burden of insulin is in part a reason for this low acceptance together with minimal clinical signs in the early stage of CFRD. It might be that OAD which were shown to be as safe and effective as insulin in the treatment of CFRD with a lower treatment burden compared to insulin are an alternative medical treatment at least for those who used no medical treatment for CFRD at all. Introduction: Cystic fibrosis-related diabetes (CFRD) is more commonly seen in patients with advanced lung disease. We hypothesized that the incidence of diabetes in pediatric patients may have changed over time as the pulmonary health of CF patients has improved.

Methods: This cohort study used prospectively collected Canadian CF Registry (CCFR) data from 2000 to 2016 inclusive. Patients between 10 and 18 years, who were annually screened by oral glucose tolerance test, were included in this analysis with patients receiving lung transplantation censored at transplant. Risk factors for CFRD were investigated using a nested case-control design. Each patient with CFRD was matched to four randomly selected CF patients by pancreatic status, age and year at CFRD diagnosis. Baseline for time dependent risk factors, such as FEV1 percent predicted and sputum microbiology, was the clinic visit closest to CFRD diagnosis. Univarable conditional logistic regression analyses were done to assess the association of each individual variable with CFRD. Multivariable conditional logistic regression was performed to evaluate the independent association of these variables. Incidence rates of CFRD for three consecutive time periods (2000-2004, 2005-2009 and 2010-2016) were calculated and compared using Poisson logistic regression analysis.

Results: There were 3139 patients meeting inclusion criteria during the study period. Patients with CFRD (n=514) were diagnosed at a median of 14.9 years (IQR 12.6 to 16.7 years). In the univariable analysis, compared to patients without CFRD, patients in the CFRD group were more likely to be female (55.5% vs. 40.4%, p<0.001), homozygous ΔF508 (64.4% vs. 58.7%, p=0.04) and have poorer nutritional status (%BMI<12th percentile 21.9% vs. 16.8%, p=0.009). In addition, the CFRD group had worse lung function (%predicted FEV1 70.6% ±23.0 vs. 81.1% ±19.5, p=<0.001) and had higher rate of sputum culture positive for Pseudomonas aeruginosa at baseline (44.9% vs. 38.2%, p=0.03). All risk factors, except nutritional status, were significantly associated with CFRD in the multivariable model. The overall incidence rate of CFRD was 2.36 per 100 patient years. There was a significant reduction in the incidence of CFRD over the consecutive study periods (p=<0.001) declining from 4.56 to 2.25 per 100 patient-years.

Conclusion: The incidence of CFRD in a pediatric population is decreasing over time which is in contrast to the rising rates of type 1 and 2 diabetes in the general population (Mayer- Davis EJ, et al. New Engl J Med. 2017; 376:1419-29) . Whether this change in incidence is causally related to improved pulmonary health or associated with a lower degree of pulmonary and systemic inflammation requires further study.

Acknowledgments: This abstract is funded by Irwin Family Foundation. Objective: Patients with cystic fibrosis (CF) have an estimated fracture rate of 15-25% yet bone mineral density (BMD) is not always abnormal in such patients. Multiple studies have evaluated TBS (trabecular bone score) in other high fracture disease groups as a bone quality measure and evaluated factors that can affect TBS. However, no study has evaluated TBS in CF patients. This study evaluated TBS in CF patients and correlated with factors potentially influencing fracture risk.

Methods: A retrospective observational study where all patients with CF, 18 years or older who underwent DXA (dual X-ray absorptiometry) scan between 2009 and 2017 were included, TBS was applied to DXA lumbar spine images. Higher TBS score indicates better microarchitecture. Clinical and biochemical data were collected at the time of the scan through chart review and survey. Data are presented as mean ± SD.

Results: One hundred thirty-six CF patients were included. Mean age was 28.1 ± 9.9 years, BMI was 21.8 ± 3.4 kg/m 2 , and 25-hydroxyvitamin D was 28.7 ±11.1 ng/mL. There were 58.5% males and 41.5% females, F508del mutation was present in 93% of patients. Cystic fibrosis-related diabetes was present in 36.6% (mean HbA1c 7.2% ± 1.98). Mean FEV1 was 64.4 ± 25.5% predicted. Mean TBS was 1.41± 0.10, lumbar spine BMD (L-spine BMD) was 1.09 ± 0.15 g/cm 2 . Lumbar spine Z-score was ≤-1 and <-2 in 45.7% and 28.7% of patients, respectively.

Multiple regression analysis and Pearson's correlation calculations showed that TBS significantly negatively correlated with age (decrease of 0.002 per year, r = -0.21, P 0.04), and positively correlated with FEV1 (increase by 0.001 per 1% increase in predicted FEV1, r = 0.4, P 0.007). There was a significant positive correlation between TBS and lumbar spine BMD (r = 0.62, p 0.00) (Figure) . Negative correlation between mean HbA1c, presence of F508del and TBS was observed, however it did not reach statistical significance (r = -0.05 for both).

Sixty percent of patients completed a survey finding previous fracture in 7 patients (8.14%). Previous exposure to systemic steroids was present in 75% of patients who completed the survey (duration of exposure in 27% was longer than 3 months).

Our study shows positive correlation between TBS and FEV1 as well as L-spine BMD. However, observed mean TBS score in cystic fibrosis patients was higher than mean score reported in other disease groups such as type 1 diabetes. While this could be explained by younger age groups in cystic fibrosis, further studies are needed to evaluate TBS cutoff point for fracture prediction in cystic fibrosis and utility of TBS use in CF patients. Background: Cystic fibrosis (CF) is a life-limiting, multisystem disease affecting pulmonary, hepatic, intestinal, endocrine and exocrine pancreatic function. The mechanisms linking metabolic disease, including cystic fibrosis-related diabetes (CFRD), to declining lung function and mortality remain unclear. Hepatic steatosis is a marker of insulin resistance in adults with type 2 diabetes and can progress to steatohepatitis and cirrhosis. We sought to determine the presence of hepatic steatosis in CF patients and diabetes and non-diabetes related risk factors associated with the accumulation of liver fat.

Methods: We recruited 30 pancreatic insufficient CF subjects aged 10-40 years with either normal glucose tolerance (NGT) or CFRD. Subjects with known CF liver disease, pregnancy, baseline FEV1<30% or an untreated pulmonary exacerbation were excluded. Clinical characteristics (BMI, FEV1, CFTR modulator use) and fasting blood chemistries (glucose, lipids, hepatic function, HbA1C) were assessed. Proton density fat fraction (PDFF) was measured on a Siemens Skyra 3T MRI, with steatosis defined as PDFF >5%. Continuous variables were analyzed with a Wilcoxon rank sum test and discrete variables were analyzed using chi square test or Fisher's exact test.

Results: Twenty-one subjects completed the study (12 CFRD and 9 NGT). Nine subjects were actively receiving the CFTR modulator lumacaftor/ivacaftor and all had at least one copy of ΔF508. Mean PDFF was 5.1% and 7 subjects (33%) met criteria for hepatic steatosis. Hepatic fat was significantly lower in patients on modulator (2.1%) than in patients on no modulator (7.3%) (Figure, p=0.001) , and an associated trend of lower triglycerides, lower postprandial glucose and higher HDL were noted with modulator use. Hepatic fat in the NGT group (6.6%) was higher than in the CFRD group (3.9%, p=0.04), which likely reflects higher modulator use in the CFRD group (7/12) versus the NGT group (2/9). When patients on CFTR modulators were excluded, there was no difference in PDFF between CFRD (7.5%) and NGT subjects (7.2%, p=1.00).

Conclusions: We found no association between hepatic steatosis and CFRD. Use of the CFTR modulator lumacaftor/ivacaftor was associated with significantly lower hepatic fat. This suggests that CFTR modulators have unexpected effects on substrate metabolism. Whether these changes are advantageous remains unknown. Prospective studies of patients initiating CFTR modulator therapy are needed to fully understand the effects of these promising new therapies.

Pancreatic insufficient (PI) and sufficient (PS) CF subjects underwent dual-energy X-ray absorptiometry (DXA) for body composition, peripheral quantitative computerized tomography (pQCT) for muscle area and fat area at 66% tibial length (MA and FA, respectively) and oral glucose tolerance testing (OGTT) for glucose tolerance status. Sex-and age-adjusted LBM index, FM index, MA and FA Z scores (LBMI-Z, FMI-Z, MA-Z and FA-Z, respectively) were generated using contemporary reference data. Glucose tolerance was defined as normal (NGT) 1-hour OGTT glucose (PG1) < 150 mg/dL and 2-hour OGTT glucose (PG2) < 140; early glucose intolerance (EGI), PG1 ≥150 and PG2 < 140; impaired (IGT), PG2 ≥ 140 and < 200; CFRD, PG2 ≥ 200). Multivariate regression was used to assess the associations between LBMI-Z or MA-Z and fasting glucose, PG1, PG2, FMI-Z, FA-Z and BMI-Z, and interactions between OGTT glucose and FMI-Z or FA-Z were tested.

Results: DXA and OGTT were performed in 40 subjects (50% male, median age 18.8 years (range: 9.6-39.2), 90% PI, 11-NGT; 16-EGI; 10-IGT; 3-CFRD, median PG1 179 (83-399), median PG2 127 (36-336), median BMI-Z 0.3 (-1.2-2.1) and median FEV1%-predicted 96 (59-134)). pQCT was performed in a smaller proportion (n=26). As expected, LBMI-Z was positively associated with BMI-Z (partial β-coefficient 0.8, CI: 0.4-1.1, p<0.001), adjusted for FEV1%-predicted and age. After adjustment for age, LBMI-Z was positively associated with FMI-Z (partial β-coefficient 1.8, CI: 0.2-3.3, p=0.025) but this relationship was blunted with increasing PG1 (partial β-coefficient -0.007, CI: -0.01-0.00006, p=0.05). In contrast, PG2 did not alter the association between LBMI-Z and FMI-Z as prominently (β-coefficient -0.007, CI: -0.01-0.0009, p=0.08).

MA-Z was positively associated with FA-Z (partial β-coefficient 1.4, CI: 0.2-2.6, p=0.02) and this relationship was also blunted with increasing PG1 (partial β-coefficient -0.007, CI: -0.01-0.0004, p=0.04). There was no interaction between PG2 and FA-Z (p=0.3).

Conclusions: Individuals with CF, glucose intolerance and a normal median BMI, exhibit LBM deficits that are masked by an increase in FM. Whether insulin therapy can prevent or reverse these LBM deficits requires further study. Diabetes is a frequent comorbidity of cystic fibrosis (CF), with ~80% of individuals carrying severe CFTR mutations developing diabetes by middle age. Although CF-related diabetes (CFRD) occurs primarily due to impaired insulin secretion, impaired secretion of pancreatic polypeptide (PP) occurs early prior to the onset of CFRD. PP is an understudied islet hormone and its role in glucose homeostasis remains largely enigmatic. Several studies suggest that PP is an intra-islet regulator of insulin secretion, and that its absence is associated with hepatic insulin resistance-which is known to occur in CF. We hypothesized reduced PP secretion may contribute to abnormal glucose levels in CF. Herein, we found that PP secretion was reduced during oral glucose tolerance testing (OGTT) and hypoglycemic glucose clamps in CF ferrets. Similarly, reduced PP secretion following OGTT was observed in young CF human subjects. Despite reduced PP secretion, CF ferrets and human pancreas retain immunoreactive PP cells. To evaluate the impact of PP on glucose metabolism, we ectopically expressed PP in wild-type (WT) ferret muscle using recombinant adeno-associated virus (AAV). Mixed meal tolerance tests (MMTTs) at three weeks after AAV-PP infection demonstrated lower glucose and lower insulin, suggesting that PP enhances insulin sensitivity. Thus, we hypothesized that low PP levels in CF may decrease insulin sensitivity. Indeed, CF ferrets exhibited altered hepatic expression of genes related to insulin resistance/sensitivity and glucagon/fasting glucose metabolism regulation. PP knockout mice lack a glycemic phenotype. However, the mouse PP amino acid sequence differs significantly from human, whereas the ferret PP amino acid sequence is nearly identical to human. Therefore, we created a PP-KO ferret. PP-KO ferrets displayed fasting hyperglycemia and hyperinsulinemia, despite normal body weight. Following OGTT, PP-KO ferrets exhibited glucose intolerance. While loss of PP did not significantly affect insulin secretion, it markedly decreased glucagon secretion. Moreover, acute administration of PP to PP-KO ferrets before OGTT improved glucose tolerance and increased glucagon secretion, without alteration of insulin levels. We investigated the impact of PP supplementation on postprandial hyperglycemia during normal chow feeding in CF ferrets. Blood glucose was monitored immediately before and one hour after feeding. Intramuscular PP was administered for three days. CF ferrets showed normalization of postprandial glucose levels following PP supplementation, while there was no effect seen in controls. Glucose clamps of PP-KO and CF ferrets are ongoing to better understand the role of PP in regulating insulin action. Taken together, these results suggest that PP increases insulin sensitivity and that low PP levels may be involved in the pathogenesis of CFRD. Further studies on these models may provide new insights into mechanisms of PP action on glucose metabolism and the role of PP in the pathogenesis of CFRD. Objective: Recent retrospective reviews in children with cystic fibrosis (CF) have shown that severity of lung disease, poor nutritional status, and low vitamin D levels were associated with lower bone mineral density (BMD). There remains a gap in our understanding of bone disease in adult patients with CF, and the role of CF-related diabetes (CFRD) in bone health has not been well-studied. We hypothesized that lumbar spine BMD is associated with lung function and CFRD in an adult population with CF.

Methods: We performed a retrospective review of CF patients who had BMD measurements between 2004 and 2017, including 101 adults and 26 children. Lumbar spine BMD and Z-scores from dual energy x-ray absorptiometry (DEXA) were obtained. Clinical and demographic variables were recorded. Laboratory data, pulmonary function testing, and diagnoses of CFRD and pancreatic insufficiency (PI) were included when present within the year of each DEXA. Pearson correlations, bivariate regression and multivariable regression were used to determine the relationships between predictor variables and BMD Z-score.

Results: The mean age of CF patients was 31±11 years old; 56% were F508del homozygote, 56% were female, 3.4% had CFRD and 22% had PI at the time of DEXA (n = 127). The mean lumbar spine Z-score was -0.8±1.35; 4% of patients had a low BMD for age, defined as lumbar spine Z-score ≤ -2. Lumbar spine Z-score correlated with FEV 1 % predicted (r = 0.17, p = 0.05) and age (r = 0.19, p = 0.01). In bivariate analysis, age (β 0.005, p = 0.03), gender (β 0.84, p = 0.03) and CFRD (β 0.06, p < 0.05) were the strongest predictors of lumbar spine Z-score. After adjusting for age and gender, FEV 1 % predicted was the only remaining significant predictor of lumbar spine Z-score (β 0.011, p = 0.03) (Table) .

Conclusion: Prior to adjustment for age and gender, CFRD was associated with lumbar spine Z-score. After adjusting for age and gender, FEV 1 % predicted remained as a significant predictor of lumbar spine Z-score, consistent with previous studies. Loss of significance for CFRD in the multivariate model could be due to the low proportion of patients with CFRD, or to differing effects of hyperglycemia and diabetes medications. A 10% increase in FEV 1 % associated with a +0.1 higher Z-score. Thus, preserving lung health may reduce the risk of developing osteoporosis. We plan to further investigate the role of CFRD in bone health, predictors of low Z-scores, and rate of fractures in this population.

RANK, RANKL and OPG, although not much is known about the contribution of neutrophils to the RANK/RANKL/OPG expression or bone turnover. Therefore, we sought to investigate the role of neutrophils and the RANK/RANKL/OPG pathway in patients with CF to understand their impact on bone loss.

Material and Methods: We studied fresh plasma and isolated neutrophils from patients with CF and compared the levels of RANK, RANKL and OPG by ELISA and flow cytometry to bone mineral density. In addition, we used a CFTR -/-mouse model of lung infection to understand the factors affecting bone loss. Whole blood was collected for plasma analysis of bone formation (P1NP), bone resorption (CTX-1) and RANKL/OPG ratio. MicroCT and immunohistochemistry or immunofluorescence for RANKL and neutrophil elastase was done on bone specimens from mice.

Results: We found a higher plasma RANKL/OPG ratio in patients with CF relative to healthy controls. We demonstrated that this increase in the RANKL/OPG ratio in patients with CF correlated with bone mineral density. Further, CFTR -/-mice showed lower P1NP levels, higher CTX1 levels and a higher RANKL/OPG ratio relative to controls. The RANKL/ OPG ratio was significantly increased in neutrophils from infected versus uninfected mice. Bone analysis showed high expression of neutrophil elastase and RANKL along with smaller cortical perimeter in tibia of CFTR-/-mice.

Conclusions: Our results demonstrate that in plasma and neutrophils from patients with CF there is an increase in the RANKL/OPG ratio correlating with low bone mineral density. This is supported by mouse studies showing high levels of neutrophil elastase in bone as well as reduced cortical perimeter. These data lay the foundation for considering therapies that may target the RANKL/OPG pathway as a treatment for CFBD in patients with CF. Rationale: Cystic fibrosis-related diabetes (CFRD) is a prevalent comorbidity that negatively impacts clinical outcomes including lung function, weight, and ultimately survival. Its impact on the health-related quality of life of individuals with cystic fibrosis (CF) has not been explored previously. The objective of this study is to investigate whether the quality of life of adults with CFRD and impaired glucose tolerance (IGT) differs from individuals with normal glucose tolerance.

Methods: Individuals with CF who attended our adult CF clinic were included if they completed the Cystic Fibrosis Questionnaire-Revised for adolescents and adults over 14 years old (CFQ-R 14+) during at least one stable visit. Stable visits were defined as no oral or IV antibiotic requirements within 4 weeks of their clinic date. Sociodemographic, clinical characteristics, and CFQ-R 14+ domain scores were compared between individuals with CFRD on insulin (n=28), CFRD not on insulin (n=26), impaired glucose tolerance (n=29), and normal glucose tolerance (n=63). Clinical variables were compared using ANOVA for continuous variables and the chi-square test for categorical variables. CFQ-R domain scores were compared across groups using the Kruskal-Wallis test. Multivariate linear regression was performed to examine the relationship between CFRD/IGT status (vs. normal glucose tolerance) and each of the twelve CFQ-R 14+ domains following adjustment for age, sex, body mass index (BMI) and FEV 1 percent predicted (ppFEV 1 ).

Results: A total of 146 adult subjects completed the CFQ-R 14+ during a stable visit. Of the twelve domains in the questionnaire, the Treatment Burden domain was significantly worse (p<0.001) for individuals with CFRD on insulin compared to the other categories [CFRD on insulin mean score 46.8 (SD 21.9); CFRD not on insulin 65.4 (25.2); IGT 65. 5 (22.7) ; and normal blood glucose tolerance 69.0 (22. 4) ]. The Treatment Burden domain included the difficulty and time spent on treatment each day. None of the other CFQ-R domains differed significantly between the four categories. Following multivariable adjustment, the Treatment Burden domain score remained significantly lower for the CFRD on insulin group compared to the normal glucose tolerance group (adjusted β-coefficient -13.90; SE = 4.86; p<0.005) indicating greater treatment burden. The Treatment Burden domain score was not lower for CFRD not on insulin or IGT groups compared to the normal blood glucose tolerance group. CFRD on insulin also had lower Role Functioning and Digestion domain scores compared to normal glucose tolerance but these associations were not statistically significant.

Conclusions: CF adults diagnosed with CFRD who are on insulin report a significantly lower CFQ-R 14+ Treatment Burden domain score. Future studies should focus on finding ways to ease the burden of blood glucose monitoring and to identify alternatives to insulin to decrease treatment burden and improve the quality of life of individuals with CF. Objective: CF-related diabetes (CFRD) is one of the most common complications in patients with CF. CFRD development is associated with worsening pulmonary function and general mortality. The current gold standard for the diagnosis of CFRD is the 2 hour 75 g oral glucose tolerance test (OGTT), with annual testing recommended by the American Diabetes Association. Given the cumbersome nature of this test, simpler alternatives have been explored. The use of HbA1c in CFRD screening has been highly controversial with recent studies generating conflicting results. Interestingly, superior specificity was noted in studies examining more recent HbA1c data. Thus, we explore whether improvements in HbA1c test performance may influence the reliability of this marker in CFRD screening.

Methods: HbA1c proficiency testing data published by the College of American Pathologists (CAP) was reviewed to assess changes in test imprecision from 2002-2017. A retrospective analysis was then conducted where all OGTT results from 2012-2018 from both adult and pediatric CF patients were extracted from the laboratory information system at DynaLIFE Medical Labs in Edmonton, Canada. OGTT results with an HbA1c performed within ± 100 days were included for analysis. Two HbA1c tests were used during this time frame: BioRad Variant II Turbo 2.0 high performance liquid chromatography (2012-December 2016) and Roche c513 immunoassay (December 2016 (December -2018 . Correlation and ROC curve analyses were performed to assess the reliability of HbA1c as a CFRD screening tool.

Results: HbA1c test imprecision declined steadily over time (from 7% in 2002 to 3% in 2017), reflecting improvements in commercially available assays. This mirrors the improvement in specificity of HbA1c in CFRD screening over the same time period in previously published studies. The imprecision of the HbA1c tests used in our local laboratory was 1.9% for the newer Roche assay versus 2.4% for the older BioRad assay.

In the adult population, ROC curve analysis demonstrated that HbA1c >5.5% can identify CFRD with a sensitivity of 95% and specificity of 45% (AUC=0.833; p<0.0001; n=108), supporting the notion that this test may be effective in CFRD screening. Subanalysis of individual HbA1c assays showed the newer Roche assay had a sensitivity of 100% and specificity of 86% (AUC=0.924; p<0.0001; n=25) while the older BioRad assay had a sensitivity of 95% and specificity of 50% (AUC=0.833; p<0.0001; n=83).

Similar results were observed in the pediatric population, where HbA1c >5.5% identified CFRD with 100% sensitivity and 48% specificity (AUC=0.846; p=0.0002; n=55). Subanalysis of individual HbA1c assays was not possible due to the low sample size.

Conclusions: This study highlights the steady improvement in HbA1c testing over time, which may explain the controversy surrounding the use of HbA1c in CFRD screening. More refined HbA1c assays with lower imprecision may enable a simpler and more convenient approach to CFRD screening in both adults and children with CF. Efforts are underway to prospectively determine the optimal HbA1c cutoff using current assays and whether combining HbA1c with other markers of glycemic control can further improve screening performance. 1 1. Pediatr. Gastroenterol., UT Southwestern, Dallas, TX, USA; 2. Physical Rehab., Children's Health, Dallas, TX, USA; 3. Pediatr. Endocrinol., UT Southwestern, Dallas, TX, USA; 4. Clin. Nutrition., Children's Health, Dallas, TX, USA; 5. Pediatr. Pulmonol., UT Southwestern, Dallas, TX, USA; 6. Pulmonol, Children's Health, Dallas, TX, USA Introduction: Cystic fibrosis (CF)-related bone disease (CFBD) is commonly seen in the adult CF population. Adult studies have shown that respiratory illness and inflammation, over time, negatively affects bone mineral density (BMD). There is limited and conflicting data about overall outcomes on BMD in pediatric CF (Endocrinol Metab Clin North Am. 2016; 45(2) :433). With extended life expectancy and increased emphasis on disease prevention for CF patients, pediatric CFBD is a topic that demands further investigation.

Methods: In an effort to improve screening and management of CFBD at our CF center, we initiated a quality improvement project in April 2016 to improve screening in patients 8 years or older. Our team formulated a DEXA scan algorithm for mild, moderate, and severe bone loss; advertised DEXA scan recommendations in our newsletter; and ordered scans for eligible patients at weekly review meetings to increase surveillance. We then retrospectively reviewed 145 DEXA results and charts after IRB approval and gathered demographics, symptoms, co-morbidities, as well as nutritional status, genetic testing, anthropometric measures, medication exposure, exercise, hospitalization frequency and relevant serum studies.

Results: Out of 145 DEXA scans reviewed, the majority of our patients had normal BMD (n=79, 54%). The remaining were categorized as mild (22%), moderate (16%) or severe (8%) based on Z-score for age, gender and body size. From 145 patients, 59.3% were male, with the largest percentage of males seen in the severe category (Table) . Mean FEV1 percent predicted was lower as BMD decreased. Most patients had at least one ΔF508 mutation (79%), with a higher percentage seen in patients with abnormal DEXAs. Of patients with normal DEXAs, 49% were on proton pump inhibitors (PPIs) whereas PPI use in patients with abnormal DEXAs ranged from 63.6-69.6%. Mean vitamin D level was similar amongst all 4 categories, 28.4 (27-30 ng/mL).

Conclusions: A baseline DEXA scan is an important screening tool for initiating a thorough bone health assessment. We have increased our use of DEXA scans in detecting CFBD and have intervened with appropriate vitamin and mineral supplementation as well as physical therapy. By studying these patients, we can also better understand the effect of potential risk factors, such as pancreatic insufficiency, malnutrition, respiratory status, medication use, and decreased physical activity. More focus should be placed upon prevention and management of pediatric CFBD including improved screening, appropriate nutritional and exercise interventions, and strategies early in life to prevent complications and to promote better health in adulthood. Background: Individuals with CF have a 35-50 percent chance of developing CFRD by adulthood. Nearly 10 percent of children with CF will have CFRD by age 10 years. The prevalence of CFRD increases with age and a delay in diagnosis leads to a negative impact on lung function, weight and survival. Additionally, recent studies suggest an association between INDET and poor pulmonary function, as well as high risk of developing CFRD. The 2-hour OGTT is the accepted standard for diagnosis of CFRD. However, performing annual OGTTs is labor intensive and inconvenient, leading to poor adherence and delayed diagnosis.

Methods: Hemoglobin A1C values and 2-hour OGTT results of fifteen CF patients (8-21 years) were compared in a retrospective review. The accepted ADA definitions for normal, IGT, and diabetes mellitus for hemoglobin A1C and OGTT were used. The accepted definition of INDET added in 2009 by the North American CFRD Consensus Committee was used. Pearson correlation coefficients were calculated for A1C at each OGTT time point of 0 hour, 1 hour and 2 hours separately.

Results: No significant correlation between hemoglobin A1C and 0 minute glucose values were found. In addition, we found no significant correlation between hemoglobin A1C and 120 minute glucose values. Therefore, the utility of a hemoglobin A1C as a screening tool for NGT, IGT or CFRD is limited. However, a significant correlation (0.7052; p=0.005) was noted between hemoglobin A1C and 60 minute glucose values, suggesting a role for hemoglobin A1C to identify patients with INDET.

Conclusion/Implications: Findings suggest a hemoglobin A1C value may be used as an effective predictor of INDET. In comparison to a time consuming and labor intensive OGTT, the use of a point-of-care screening hemoglobin A1C may allow for earlier detection and intervention of glycemic disturbance, resulting in better outcomes in the CF population. It has been established that the endocrine pancreas is affected in the CF patient. There is a correlation between glucose abnormalities, morbidity and mortality in CF patients (Ode KL, Moran A. Lancet Diabetes Endocrinol. 2013; 1:52-8) . Glucose abnormalities include CF-related diabetes (CFRD) and impaired glucose tolerance (IGT). CFRD is one complication in CF, occurring in more than 40% of adults and 25% of adolescents, which is preceded by episodes of IGT (Ntimbane T, et al. Clin Biochem Rev. 2009; 30:153-77; Schwarzenberg SJ, et al. Diabetes Care. 2007; 30:1056-61) . It is very well documented that reduced insulin secretion and insulin resistance (IR) lead to CFRD (Nathan BM, et al. Curr Opin Endocrinol Diabetes Obes. 2010; 17:335-41; Preumont V, et al. Diabetes Care. 2007; 30:1187-92) . We have developed a CF rabbit that is more representative of human CF than any other CF animal model. No literature regarding CF-related diabetes in rabbits is available. For this purpose, we have developed 3 different mutations. For the first one we added one nucleotide (+1, caused frameshift at amino acid 479) or deleted one nucleotide (-1, caused frameshift at amino acid 478) which caused a frameshift in the CFTR at exon 1. Another mutation resulting in a deletion of 3 amino acids (from 477 to 479) is the -9 deletion or ΔPSE-477-479. All these mutations affect the NBD1 of the CFTR channel. The mechanisms underlying CFRD are poorly understood. We have observed that CF rabbits have a tendency of low plasma insulin, which is an early sign of diabetes.

To study the impact of CFTR deficiency in insulin sensitivity and glucose tolerance, wild-type (WT) and CFTR knockout rabbits were used to test glucose tolerance tests (GTT) and plasma insulin. Our results regarding the genotype +1/+1 showed that 80% of the CF rabbits showed impaired glucose tolerance. As for insulin, the CF rabbits had an insulin resistance which will explain the phenotype of impaired glucose. For the other CFTR mutation -1/+1 genotype, all the CF rabbits studied showed an abnormal glucose tolerance with 75% showing a severe impaired glucose. In addition, the insulin measurements showed less insulin production compared to WT after the injection of glucose. The lower insulin in the CF rabbits is compatible with an insulin insufficiency which will explain the delay in the glucose clearance. As for the -9/-9 genotype, 62% of the CF rabbits with -9/-9 genotype showed an impaired glucose clearance, with 38% showing the typical CFRD phenotype. The insulin measurements showed a normal insulin production compared to the WT which means, there is an insulin resistance in rabbits with abnormal glucose clearance. Taken all together, our CF rabbits with a frameshift mutation such as +1/+1 and -1/+1, are more susceptible to have CFRD than -9/-9 deletion of 3 amino acids. The data suggests that no expression of CFTR protein has a more dramatic effect on the uptake of the glucose than the expressed ΔPSE-477-479 CFTR. Our CF rabbit is the perfect animal model to study the many problems associated with CF patients with CFRD including chronic pancreatic inflammation, dysfunction of the immune system, oxidative stress, and impaired insulin production and secretion.

Funding supported by CFF (SUN15XX0) and NHLBI/NIH (HL096800 and HL133162 To FS).

Methods: This was a random order cross-over trial comparing two configurations of the BP, insulin-only and bihormonal (insulin + glucagon) with usual care (UC). Three subjects with CFRD treated with multiple daily injections of insulin or insulin pump therapy were recruited for this study. Subjects participated in three arms (BP in the insulin-only vs. bihormonal configuration vs. UC) with each arm one week in duration. The insulin-only BP was set at a target of 120 mg/dL and the bihormonal BP at a target of 110mg/dL, which were selected as the default targets for these configurations based on studies in T1D. Subjects had no restrictions on their daily activity and continued their diet, exercise and work throughout all arms. During the BP arms, subjects discontinued their UC insulin and received insulin only from the BP. They kept a diary of insulin dosing during the UC arm, and when available insulin pump data were downloaded and collected. They answered brief emailed surveys daily regarding symptomatic hypoglycemia, psychosocial impact/quality of life, and any other adverse events.

Results: The mean glucose was nominally lower in both the bihormonal BP (140±16 mg/dL) and insulin-only BP arms (147±10 mg/dL) and relative to the usual care with insulin pump or multiple daily injections (165±43 mg/ dL). Hypoglycemia was low in all three arms (0.35%, 0.25%, and 0.37% <54 mg/dL, respectively). No adverse events were reported. These test subjects found the BP easy to use, and surveys suggested improved satisfaction and reduced burden of diabetes treatment.

Conclusions: This pilot study represents the first report of any artificial pancreas system in the management of CFRD. We demonstrate the feasibility of blood glucose control with the BP in this patient population. Larger-scale studies are needed to further investigate the utility and benefits of the BP in CFRD. Background: Cystic fibrosis-related diabetes (CFRD) is the most common comorbidity of adult cystic fibrosis and has significant impact on lung function, weight maintenance and survival.

Annual screening for CFRD using the 75 g oral glucose tolerance test (OGTT) is the recommended screening tool used to diagnose CFRD but the wide availability of continuous glucose monitoring (CGM) technology has shown that clinically relevant glucose excursions can occur in patients with normal OGTT results. Furthermore, the introduction of oral DPP-4 inhibitors as oral insulin secretagogue drugs that are effective at managing postprandial glucose excursions with minimal risk of hypoglycaemia allows for earlier treatment than has previously been the case with insulin. This is relevant as even modest glucose excursions in daily life have been shown to impact on lung function and weight, even in patients without a formal diagnosis of CFRD.

Prior to 2017 we conducted OGTT as the primary method of screening for CFRD and offered CGM to those with abnormal results to determine if treatment was appropriate. Based on our experience of poor correlation between OGTT results and observed glucose excursion we changed policy in 2017 to offer all patients CGM. Here we present our early experience of its use as a screening and assessment tool for dysglycaemia in CF patients.

Methods: All non-CFRD patients were offered CGM (using the Medtronic iPro2 device) as part of the annual review. CGM was worn for 5 days alongside compilation of a structured food diary.

CGM traces were reviewed in a multidisciplinary meeting comprising cystic fibrosis and diabetes specialist health professionals, with a treatment plan agreed for individuals with significant glucose excursions. In the event of a gliptin being advised as the treatment, a repeat CGM was offered to review the impact of the treatment.

Results: In 2016, 92 patients underwent OGTT; 15 were abnormal and went on to have CGM. Of these 8/15 had significant postprandial glucose excursion (recurrent postmeal glucose > 8 mmol/l, 144 mg/dl) of which 4 were managed by dietary modification, 1 using a DPP-4 inhibitor and 3 with prandial insulin.

In 2017, 99 patients were offered CGM as primary screening. 24 declined and were offered OGTT. Of the 75 who completed CGM, 43 were identified as having significant glucose excursions. Of these, 10 were managed with dietary intervention, 21 with DPP-4 inhibitor and 12 with insulin.

Conclusions: Introduction of routine CGM screening of all patients attending an adult cystic fibrosis service has enabled identification of a much higher prevalence of abnormal glucose excursions (57.3% of those screened, 43.4% of clinic population) than was apparent from OGTT followed by CGM the previous year (8.6%), enabling early active and varied intervention. Prospective studies will be needed to determine the impact of this approach and early intervention on preservation of lung function and outcomes of CF. 

Introduction: A host of epidemiologic data suggests that women with cystic fibrosis (CF) have worse outcomes than men. In addition, women with CF demonstrate earlier colonization with respiratory pathogens and an increased rate of exacerbations near ovulation. The etiology of this sex based disparity is unclear, but sex hormones may contribute to these differences. High estrogen levels are associated with increased inflammation in CF animal models of infection. We, therefore, sought to prospectively evaluate markers of inflammation and lung health in men and women with CF. Women off hormone contraception were followed through ovulatory cycles and subsequently started on oral contraceptive pills (OCP) to understand the impact of OCP on lung function, respiratory symptoms, sputum and plasma inflammatory markers as well as sputum cytology.

Materials and Methods: We followed women with CF who were not on hormone contraceptives for the duration of an ovulatory cycle including study visits that corresponded to menses, ovulation and luteal phase. In addition, subjects were subsequently placed on a standard hormone combination OCP, Loestrin (ethinyl estradiol/norethindrone). Males were recruited at a 1:1 ratio in an age and FEV1% matched fashion as controls. Data collected included lung function, symptom questionnaires, sweat tests, plasma for hormone levels and inflammatory markers, sputum for inflammatory markers, bacterial density, and cytology.

Results: Males (n = 22) and females (n = 22) with CF were enrolled in this single-site study. A subset of females consented to participate in the OCP arm of the study (n = 12). Male and female populations were similar with a mean age of 29.8 ± 7.2 in females and 34.7 ± 11.2 in males, p = 0.117 and mean FEV1% of 57.9 ± 23.2 in the females and 59.9 ± 23.0 in the males, p= 0.774. Levels of estradiol, progesterone and luteinizing hormone proved to be in expected ranges, marking points of menses, ovulation and luteal phase and effective contraception in the women. No adverse events were reported with use of OCP. Analysis of lung function showed no significant changes, however, CFQ-R respiratory scores were on average 9.2 points lower in women at ovulation relative to menses and improved 6.6 points once placed on OCP. In addition, interim analysis of inflammatory markers revealed increased levels of sputum white blood cells, percentage of neutrophils, IL-8 and IL-1β in females at ovulation relative to menses which decreased on OCP. These markers were also higher in females at ovulation than in male controls. Further analysis of additional serum and sputum inflammatory markers and correlations between hormone levels and markers will be determined.

Conclusions: Our results demonstrate that females with equal FEV1% to their male counterparts have higher markers of inflammation and that addition of OCP suppressed inflammatory markers. These studies support a potential benefit to OCP therapy in women with CF and serves as the foundation for further mechanistic understanding of how sex hormones are involved in the inflammatory responses in patients with CF.

Acknowledgements Introduction: Individuals with CF are surviving into adulthood as a result of advancements in medical treatments. Consequently, the CF care model has increasingly evolved to proactively address issues related to adulthood (ie, transition to adult care, family planning, and transplant). Educational attainment is the foundation upon which many future outcomes are later met. Yet, systematic assessment and delivery of tailored interventions to promote school success are lacking for the student population with CF.

Methods: The primary aims of this longitudinal research study are to: 1) describe the educational needs in pediatric CF, 2) describe the frequency and predictors of school absences, and 3) examine the quality of life of this school-age sample. Eligible participants include school-age (K-12) patients diagnosed with CF, English speaking, and enrolled in a public/private school. Enrolled participants and one primary caregiver complete baseline measures to assess perceptions of educational issues, executive functioning, child emotional-behavioral functioning, resilience, and quality of life. Subsequently, participants complete the Brief School Needs Inventory (BSNI) with the CF school liaison specialist (SLS). The BSNI identifies and stratifies educational needs into risk levels to inform individualized interventions. Medical and demographic data are extracted through a medical chart review and school records (GPA, attendance) are collected from schools.

Results: Mean age of the enrolled sample (n=54) is 10.35 years, and approximately 60% are female. Mean highest FEV1% in last two years = 108 (17.1). Of the 70% of enrolled participants who have completed the BSNI to date, 26% fall into the high educational risk category, 37% in the moderate, and 37% in the low-risk category. This abstract will present baseline data from the aforementioned measures (to be completed by summer 2018) and examine predictive relationships with school absences. See the Table for concerns endorsed by parents on the BSNI. Notably, no parents have declined school intervention services once enrolled.

Conclusions: Preliminary results show: 1) a variety of educational support needs requiring individualized assessment and treatment and 2) high acceptability and need from parents for educational support to be systematically included in the CF care model. Results reinforce a significant gap in the care model that requires attention and action from the CF community.

Acknowledgment: This study and the CF SLS position are part of the CF WELL: Cystic Fibrosis Wellness, Education and Learning Laboratory program funded by the Boomer Esiason Foundation. Introduction: There has been heightened awareness in the CF community about the importance of behavioral health's impact on overall health. As the chronic care model continues to expand to more proactively and effectively respond to emotional and behavioral concerns, equipping CF care providers with increased understanding and communication skills around behavioral health can begin to increase comfort and the likelihood that meaningful, more effective interactions occur around these topics.

Methods: A significant element of our newly-established Behavioral Health program at Cincinnati Children's CF Center (Behaviors, Emotions & Thoughts To Enhance Resilience in CF; BETTER-CF) includes the development of an education series delivered to all care team providers to introduce behavioral health topics and provide opportunities for interactive learning. Monthly topics are presented by an interdisciplinary panel of CF care team members (led by a psychosocial team member) at a full center meeting, and involves didactic, role-play, and case-based learning. A needs survey was administered in February 2018 asking CF care members to report current comfort and frequency of addressing behavioral health topics prior to the start of the educational series.

Results: Nineteen care team members responded including physicians, fellows, nurses, social workers, chaplains, pharmacists, dietitians, and respiratory therapists. The majority reported being somewhat comfortable bringing up behavioral health topics (47%) or fielding behavioral health questions/concerns (32%). However, the majority reported providing behavioral health resources to families only some of the time (42%). Comfort with specific topics is detailed in the Table. A post-survey will be completed at the end of this year's educational series.

Conclusions: In order to proactively address behavioral health concerns, team members and patients/families need to feel comfortable raising and responding to issues. Education and practice with active and supportive strategies has the opportunity to promote trusting, therapeutic relationships and allow for more timely provision of needed resources. Standard clinical training does not adequately prepare many providers to address these topics sensitively and effectively. Our results demonstrate a need to provide more systematic training about PTSD, anxiety and depression, parenting, and suicidal ideation.

extracted from patients' electronic medical records from 3 months prior and 3 months post-study engagement.

Results: 144 patients completed initial screening, with 40% and 36.5% endorsing elevated anxiety and depression, respectively. To date, 23 patients (16%; 13 women, 10 men) opted to participate in ACT, with 74% selecting telehealth. Retention was exceptional, with 96% completing all 6 sessions. ACT was associated with decreased anxiety on the BAI (t=2.99, p<.05), a trend on the GAD-7 (t=1.92, p=.07), depressive symptoms on the BDI-II (t=1.80, p<.05) and rigid attachment to one's thoughts on the CFQ13 (t=4.19, p<.01). ACT was also associated with a trend toward increased FEV 1 /FVC ratio from 3 months prior to 3 months post-treatment (t=-1.77, p=.09). There were no significant outcome differences between telehealth and in-person participants.

Conclusions: Telehealth-delivered ACT is a feasible and potentially effective treatment for symptoms of anxiety and depression and could lead to improved lung function among people with CF. Near-perfect retention suggests ACT is well-received via Webcam. Given that ACT with CF showed associated reductions in anxiety and depressive symptoms, detachment from maladaptive beliefs and coping, and a trend toward improved lung function, a larger RCT of ACT with CF is warranted to identify mediators of lung function improvement, such as treatment adherence.

Background: While medical advances have improved CF life expectancy, managing and completing the daily treatment regimen is often a struggle. Individuals with CF, families, and clinicians report difficulty having effective discussions about these challenges. Prior CF Foundation research identified effective patient-provider communication as critical to success with sustaining daily care.

Objective: To systematically elicit feedback from CF clinicians, people with CF, and families about communication styles, needs, and preferences; and to identify opportunities to impact adherence through improved communication.

Methods: Qualitative research was conducted at 5 geographically and demographically diverse US accredited CF care centers (2 adult; 3 pediatric). Methods included: 1) observation of outpatient clinical interactions between individuals with CF, family members and clinicians; 2) 1:1 semi-structured interviews of individuals with CF, family members, and clinicians; 3) in-person clinician focus groups; 4) virtual focus groups of family members and adults with CF. Participants were asked about their communication experiences and needs in CF care. De-identified transcripts of handwritten notes and audio-recordings were coded and analyzed using content analysis and constant comparative methods to identify emergent themes. Using a refined coding list, coders established reliability using consensus gaining strategies.

Results: A total of 80 participants completed study activities: 56 clinicians, 14 adults with CF, and 10 family members. Four themes emerged: Eliciting Psychosocial Concerns, Childhood Development and Transitions, Negotiating Shared-Decision Making, and Suggested Communication Resources. Participants recounted psychological, family, and social factors, anxiety around clinical testing, and economic/health insurance challenges as adversely affecting adherence. Participants identified childhood as a pivotal opportunity to begin discussion about daily care and described the complexity of preparing adolescents to assume responsibility for self-care. Adults with CF, clinicians and parents described significant communication challenges in co-creating treatment plans; they stressed the value of effective communication resources to develop trust and motivate behavior change.

Conclusion: Themes identified a need for enhanced shared-decision making and relationship-centered communication in which individuals with CF fully partner with clinicians to discuss treatment plans and daily care. To improve adherence, clinicians, parents, and adults with CF desire resources to facilitate discussion of psychosocial needs, collaborative development of treatment plans, and early rapport-building with children and young adults. The findings highlighted the interpersonal, organizational and system-level barriers contributing to clinical communication challenges, and the need to create CF-specific resources to address these obstacles. interventions have a large evidence base for depression/anxiety prevention and treatment. However, traditional CBT protocols require adaptation to address the emotional challenges of coping with CF, stressors related to disease management, and barriers to access to care.

Objective: To develop a CF-specific CBT preventive intervention for adults integrated into routine CF care, with input from adults with CF and care team members. The program consists of 8 45-minute modules flexibly delivered by CF care providers in clinic, on the inpatient unit, or by telephone.

Methods: In-depth feedback was collected via audio-recorded telephone interviews with 16 adults with CF from 3 CF centers (MGH, U. Miami, U. Buffalo). Purposive sampling was used to ensure diverse representation by gender, age, ethnicity, and disease severity. Focus groups of CF care team members were conducted at MGH and Buffalo. Semistructured interviews elicited discussion on intervention structure and content; tailoring to meet CF-specific mental health needs; and integrating the intervention into CF care.

Results: Patients ranged from 21-53 years (M=35); 50% were female; 3 were Hispanic, 1 was post-transplant. Of the 10 providing this data, patient-reported FEV1 ranged from 25-113% pred (M=72). 8 reported a diagnosis of depression or anxiety and 4 more reported having had depression/anxiety symptoms; 13 had participated in mental health treatment (therapy or medication). Interviews revealed core themes related to the experience of coping with CF. Top stressors included worry about disease progression, treatment burden, and financial stress. Patients emphasized the value of a CF-specific mental health intervention and provided positive feedback about the proposed program structure. Patients felt that delivery by a CF team member was advantageous (eg, knowledge of CF; familiarity to patient; easier access to care). Most felt that telephone delivery was a positive feature; several indicated the importance of having at least one face-to-face meeting. Interviews confirmed the relevance and acceptability of CF-specific content. Feedback informed design of the intervention manual and patient workbook, and a training program for CF care providers. Direct quotes from CF adults were included in the workbook to highlight patient contributions, increasing authenticity and resonance of material.

Conclusions: Results suggest the acceptability of a CF-specific CBT-based approach to promote positive well-being and prevent depression and anxiety for adults with CF. Patients and providers identified themes informing intervention development, highlighting the value of including the CF community in program development. The CF-CBT program is currently being piloted across CF centers.

Acknowledgments: Vertex Circle of Care Charitable Grant, CFF Therapeutics.

Richards, C. 1 ; Pinsky, H. 1 ; Friedman, D. 2 ; Yonker, L.M. 2 ; Georgiopoulos, A. 2 1. Massachusetts General Hospital, Boston, MA, USA; 2. Pediatrics, Massachusetts General Hospital, Boston, MA, USA Introduction: Substance use disorders are common in both general and medically-ill populations, with devastating impact on health and self-care (Facing Addiction in America: The Surgeon General's Report on Alcohol, Drugs, and Health; HHS 2016 Nov). In particular, alcohol and opiate misuse represent major public health problems. However, there has been minimal study to date of substance use patterns in adults with CF.

Objective: This retrospective, single-center study aimed to 1) assess the prevalence of documented alcohol and opiate misuse in adults with CF, and 2) to compare psychiatric and medical status in CF adults with and without documented alcohol/opiate misuse.

Methods: Following an IRB-approved protocol, the medical records of all patients 18 years and older seen between 2005-2017 at the Massachusetts General Hospital Adult and Pediatric CF clinics were queried using relevant keyword searches of electronic medical records (EMR). Mean values and confidence intervals for each clinical variable were calculated. Groups with and without documented alcohol/opiate misuse were compared using one-sided Student's t-tests and linear regression where appropriate.

Results: 229 medical records were reviewed. 15 adults (6.6% of clinic) had a documented history of opiate misuse and an additional 13 (5.7% of clinic) had a documented alcohol misuse history, together representing 12% of our total clinical population. Males comprised 71% (n=20) of this cohort. Depression and anxiety were the most common psychiatric comorbidities: n=14 (50%, p=0.68) and n=16 (57%, p=0.18) respectively. Attention deficit-hyperactivity disorder, bipolar disorder, and obsessive compulsive disorder were represented at lower prevalence. Patients with alcohol and/or opiate misuse had significantly more inpatient admissions per year (mean 2.06; p<0.001) and a higher no-show rate to clinic (17%, p<0.001). They also had higher mean depression screening scores than the total adult CF population, PHQ-9: 10.4 vs 5.1 (p<0.001), as well as higher mean anxiety scores, GAD-7: 10 vs 4.4 (p<0.001). There were no significant differences in lung function or BMI between the two groups.

Conclusions: This is among the first studies to examine the prevalence and correlates of alcohol/opiate misuse in adults with CF. CF adults with comorbid misuse of these substances had significantly more admissions and a higher rate of missing clinic appointments. These patients were no more likely to carry a diagnosis of depression or anxiety; however, their self-reported depression and anxiety symptoms were significantly more severe. These findings indicate that alcohol/opiate misuse is common and impacts both routine and inpatient care. Our review may underestimate prevalence, as we do not systematically screen for problematic substance use, and have included only those cases with clinically significant misuse clearly and verifiably documented in the EMR. Further research into the prevalence and effects of substance misuse in CF is indicated. Objectives: Adhering to the complex CF treatment regimen is challenging. Surprisingly, there are no standardized tools to identify adherence barriers. Using the newly created Barriers Assessment Screener (BAS; 21 items) we evaluated adherence barriers identified by people with CF treated in pediatric and adult CF centers to understand their perceptions of: (1) frequency (yes/no occurred in past 6 months) and (2) adherence interference (1=never to 5=always gets in the way of doing treatments) of common barriers.

Methods: 12 pediatric (PEDS) and 9 adult (ADULT) US CF centers participating in the Success with Therapies Research Consortium Barrier's Study administered the BAS and a demographic survey during a clinic visit. Medical history and prescribed medications were abstracted from the medical record. Eligibility criteria were: age 13+ years, diagnosed with CF, English speaking, prescribed a chronic pulmonary medication, no other family member enrolled, no developmental delays precluding completion of surveys, and no hospitalization expected in the next 28 days.

Results: 407 people enrolled (ADULT=200; PEDS=207). Mean age was ADULT=32.8 (SD=10.9) and PEDS=17.1 (SD=3.1), 50% were male, 85% were White non-Hispanic, and 30% had public or no health insurance. Participants from adult centers endorsed more total barriers as occurring in the past 6 months compared to pediatric centers (ADULT M=8.7 (SD=4.3) vs. PEDS M=6.9 (SD=4.6), p<.001) and adults endorsed 11/21 items statistically significantly more frequently (ps<.05). However, the 4 most commonly endorsed barriers were the same in both groups: too tired (ADULT=67%; PEDS=48%), rather stay/go to sleep (ADULT=64%; PEDS=61%), too lazy (ADULT=61%; PEDS=56%), and too busy (ADULT=55%; PEDS=53%). The least commonly endorsed barriers for pediatric centers were financial problems (11%) and a lack of social support (13%); for adult centers it was not having a schedule/routine (21%) and financial problems (22%). If a barrier was present, both adult and pediatric center participants rated the level interfering with their adherence similarly except for one item: not having a daily routine/schedule for which adult participants reported this as more problematic than pediatric participants (ADULT M=2.90 (SD=1.09) vs. PEDS M=2.26 (SD=1.09) p<.01). The mean adherence interference rating for barriers was usually in the "rarely" to "sometimes" range (i.e., mean scores were in the 2-3 range) with the exception that 75% of adult center and 69% of pediatric center participants rated not having social support as "never" interfering with adherence.

Conclusions: Both adolescents and adults with CF face many barriers that may negatively impact adherence. While participants treated in adult centers endorsed significantly more barriers that those in pediatric centers, once endorsed, the perceived degree of interference with adherence was similar. The next step in validating the BAS is to evaluate whether it is associated with treatment adherence. If so, the BAS may be a clinically useful tool for helping people with CF sustain daily care. Background: "Primary" palliative care (PC) skills for CF care teams are not clearly defined, but in general for serious illness include managing basic physical and emotional symptoms, and having basic discussions about prognosis, goals, suffering, and code status. Complex symptom management and addressing conflicts around goals and decisions are considered "specialty" PC skills for which PC consultation may be appropriate. We aimed to understand primary PC skills of CF care teams from the perspectives of team members, individuals with CF, and family caregivers.

Methods: CF care team members ("providers"), adults with CF ("patients"), and family caregivers ("caregivers") rated the ability of CF care teams to provide various aspects of PC using a 5-point scale from "poor" to "excellent." Median ratings were compared between and among groups using Mann-Whitney and Kruskal-Wallis tests.

Results: A total of 520 participants, including 70 patients, 100 caregivers, and 350 providers, completed surveys. CF care teams consistently rated their PC skills higher than patients or caregivers rated providers' skills. Providers rated their teams "very good" at pain and depression assessments, discussing lung transplant, and discussing prognosis, and "good" at discussing advance care planning (ACP), code status, end of life, and hospice. Patients and caregivers agreed that teams provide "very good" pain assessment, but rated teams "good" at assessing depression (P<0.001) and discussing prognosis (P=0.006), and "poor" at discussing lung transplant (P<0.001), ACP (P<0.001), code status (P<0.001), end of life (P<0.001), and hospice (P<0.001). Providers, patients, and caregivers affiliated with adult CF care teams rated teams more highly then providers, patients, and caregivers affiliated with pediatric teams at discussing lung transplant (P<0.001), end of life (P=0.006), ACP (P<0.001), code status (P=0.012), and hospice (P=0.016). A majority of patients (69%) and caregivers (60%) felt CF care teams should definitely receive more PC training.

Conclusions: Discrepancies exist among patient/caregiver and provider perceptions of PC skills in CF, and skills of adult and pediatric teams may differ. While patients, caregivers, and providers agree that CF care teams are skilled in some "primary" PC skills like pain and depression assessment and discussing prognosis, patients and caregivers feel providers' skills are lacking in discussing lung transplant, ACP, code status, end of life, and hospice. Education for all groups could promote awareness of PC, and CF care teams may benefit from specific PC training to enhance "primary" PC skills as well as understanding when and how to utilize specialty PC services. The role of the CF social worker at Cincinnati Children's has expanded to include transition coordination, mental health screening and referrals, and administration of the Psychosocial Assessment Tool-CF (PAT-CF), in addition to the daily expected responsibilities of a hospital social worker. Each of these roles demands the most up-to-date data entry and analysis to effectively use the information during clinical practice and previsit planning meetings. Manual entry, individual management of data, and data reporting places additional high demand on social workers and is not feasible given the multiple responsibilities for a full caseload. Automated data management and reporting is ideal to track psychosocial outcomes that social workers monitor clinically. These topics include mental health screening and intervention, adherence tracking, transition/educational objectives, and other general psychosocial services (eg, insurance, financial aid).

Program Summary: To address these complex needs, our CF center created an automated Psychosocial Report. It is believed that if each team member is made aware of the mental health, psychosocial risk factors and transition progress of each patient during previsit planning meetings, we can examine patterns, understand needs, and provide better outcomes. The Psychosocial Report data is populated from the electronic medical record and external databases with most current data. The form includes patient demographics (including contact information), insurance carrier, and supplemental security income status. The report includes the last 3 mental health screen scores (GAD-7 and PHQ-9) and dates completed, with a color-coded, risk-stratified legend to define the scoring parameters. The PAT-CF score is also color-coded and stratified by risk, as well as transition/education assessments and parameters. Further information pulled into the report includes mental health diagnoses, date of the next psychology appointment, and date seen by the school liaison specialist. Finally, the most recent social work note and (if applicable), psychology instructions are automatically pulled into the report.

Results: The latest and most complete version of the Psychosocial Report has been live since April 2018 and is being used in previsit planning and weekly rounds to discuss patient progress, examine mental health screening trends, and explore psychosocial barriers to quality of life and care. Stratification is used to flag patients who are at high risk and discussions can occur to put resources in place to collaboratively address identified needs. The development of this report has elevated the importance of mental health and psychosocial information for all care team members so that this information can be easily pulled and discussed in real time in previsit planning meetings. Social workers are now able to significantly decrease time spent in preparation for previsit planning meetings and accurately report psychosocial factors impacting patient care, permitting more time for purposeful clinical interaction and resource identification. The nutritional demands of children with cystic fibrosis (CF) are complex due to their difficulty absorbing nutrients and gaining weight. Body mass index (BMI) percentile is a critical factor in CF health outcomes, serving as a predictor of lung function and mortality across the lifespan. Guidelines recommend calorie intake between 110-200% of the daily energy needs for youth without CF, BMI at the 50th percentile or higher, 40% of calories from fat, 3 meals and 2-3 snacks per day, and use of enzymes for pancreatic insufficient patients. Almost half of youth with CF fall short of these nutritional health goals. Research using objective measures showed daily enzyme adherence among adolescents ranged from 42.5% to 77.1%. While nutritional adherence is multifactorial, one barrier yet to be examined among adolescents with CF is attempts to conceal CF diagnosis, symptoms, and treatments in order to appear "normal." Attitudes toward disease disclosure develop within the family system, thus examining disease disclosure within the context of adolescent relationships and family attitudes about privacy and information sharing may shed light on the relation between disclosure and nutritional adherence. We hypothesized that better adolescent nutritional adherence behaviors would be predicted by families with more favorable attitudes towards sharing private information, better relationship quality with caregivers and friends, and more comfort in disease disclosure.

Methods: Adolescents between 12-17 years with a confirmed diagnosis of CF, pancreatic insufficiency, access to the internet, and proficiency in reading English were eligible to participate. Exclusion criteria included developmental delay or intellectual disability, history of a lung transplant, or diagnosis of diabetes. Adolescents and a caregiver completed measures of disease disclosure, caregiver privacy attitudes, and relationship quality. Teens used MEMSCaps for 1 month to track enzyme adherence and logged their nutritional intake for 2 weeks during this same period using Myfitnesspal™.

Results: Better enzyme adherence was predicted by greater comfort in disease disclosure activities and lower adolescent/maternal discord. Higher dietary intake and eating frequency were predicted by greater caregiver privacy with nonfamily members and lower teen closeness with friends.

Conclusions: Results indicate that adolescents who feel more comfortable in disclosure behaviors take enzymes reliably in different environments. However, teens with closer friendships eat less frequently and consume fewer calories, which suggests that these adolescents are spending more time with friends, leaving less time for self-management. Teens who reported lower levels of discord with their mother demonstrated better enzyme adherence, perhaps suggesting a developmentally appropriate shift of parent/adolescent independence without associated conflict. Higher caregiver privacy outside of the family predicted higher caloric intake and eating frequency, which could reflect developmentally appropriate ratio of monitoring and autonomy in these families. Findings from this sample suggest that daily structure and positive, appropriately supportive relationships are factors that can be targeted by clinicians and families. Introduction: Advances in the care and treatment of cystic fibrosis (CF) have led to improved mortality rates; therefore, considerably more patients with CF are living into adulthood. Among other treatments, medical nutrition therapy is required and research has found that social, behavioral, and physical factors influence the ability of CF patients to follow dietary recommendations. A socioeconomic factor that has not been adequately investigated with regards to dietary compliance of CF patients is food insecurity.

Objective: The aim of this case study was to estimate the prevalence of food insecurity among CF patients residing in Idaho. The correlation between food insecurity and health outcomes (lung functioning and body mass index) was also explored.

Methods: Conducted between August 2013 and March 2015, all data were collected using the Research Electronic Data Capture (REDCap) system. Participants included adult patients and parents of children with CF. All participants (n= 87), patients of the St. Luke's Cystic Fibrosis Center of Idaho, were volunteers. During their regular quarterly visit, patients or parents of children with CF completed a computer-based survey; body fat measurements were also collected. In patients over the age of six, a forced expiratory volume in one second (FEV 1 ) using a MicroLoop spirometer was assessed using predicted equations from the National Health and Nutrition Examination III (NHANES III). The Centers for Disease Control and Prevention (CDC) BMI calculator was employed for this study. For validity, the survey questions drew from the Mountain West CF Consortium Questionnaire (MWCFCQ) tool, which was designed to help identify and test educational interventions of nutrition management for parents of children with CF. The Institutional Review Board at St. Luke Health System evaluated and approved all aspects of the study.

Conclusion: Food insecurity rates among CF patients were found to be significantly higher compared to the community-at-large; however, no specific correlation between food insecurity and BMI or lung function emerged. The degree of food insecurity in both study groups, adults with CF and parents of children with CF (43% and 33%), was notably and statistically significantly above state levels (15%) (Idaho Food Bank. http://idahofoodbank.org/news-blog/food-insecurity-in-idaho. 2015). Furthermore, the prevalence of food insecurity was higher than reported by an earlier study conducted in other states in the Mountain West region of the United States (26.3%) (McDonald CM, et al. Infant Child Adolesc Nutr. 2009 ;I(6):325-31.) This case study highlights the need for continued research around food access issues and this patient population, as well as the value of CF advocacy organizations promoting efforts to address food insecurity issues.

Background: The Cystic Fibrosis Reproductive & Sexual Health Collaborative (CFReSHC) is a partnership between patients, providers, and researchers that aims to improve the sexual and reproductive health (SRH) of women with CF. After addressing priority areas for CF SRH research, we sought to solidify the structure of CFReSHC in alignment with its goals and mission. We aim to evaluate the engagement process in building this sustainable research partnership.

Methods: From May 2017 to July 2018, CFReSHC finalized its organizational structure and conducted multiple monthly meetings. We employed video conferencing software and shared electronic documents as a platform. Both private and public social media groups and a monthly newsletter were created to facilitate continuous engagement and communication. To assess the engagement process of CFReSHC, we anonymously surveyed CFReSHC leaders, patients, and researchers (ongoing).

Results: CFReSHC consists of a Governance Board (GB) of 4 women with CF, 3 clinician-researchers, and 1 community engagement expert; a Research Advisory Panel (RAP) of all GB members and 6 additional CF/SRH researchers; and a Patient Task Force (PTF) of 63 women with CF. The GB has three subcommittees: Communication, Meeting Planning, and Governance. Monthly PTF and RAP meetings mirror each other in focus and have included the following topics (based on previously identified SRH priorities): pregnancy and parenthood, female hormones and CF, body image, and urinary incontinence. CFReSHC participants have collaborated on 3 grant applications, 2 letters of intent, and an online patient mini-conference around SRH in CF.

The majority of the GB (71%) understood the expectations for their participation in CFReSHC. Over half (56%) felt their skills/expertise was valued. Two GB members reported difficulty understanding research discussions. Over 40% of the GB did not feel email communication was helpful for their engagement with CFReSHC. Based on this feedback, CFReSHC instituted mechanisms for providing feedback and procedures for explaining and discussing research during meetings. Communication was transitioned to a cloud-based platform. Survey results assessing the PTF's and RAP's attitudes toward the engagement process are pending.

Conclusion: This project demonstrates an online engagement process for building a sustainable national patient-provider-researcher partnership that respects participant values and goals. Specifically, CFReSHC has continued to advance inquiries into female SRH in line with relevant patient-centered outcomes in the provision of comprehensive CF care. To assess individual Research Voice member perspectives and experiences related to clinical research, CF Foundation staff conducted telephone interviews with Research Voice members.

Methods: CF Foundation staff interviewed a convenience sample of Research Voice members from those indicating prior experience in clinical research. Staff used a structured interview process to ensure consistency. The interview guide included the following questions:

-Have you participated in a clinical trial or research study? If so, did you participate in an investigational trial or an observational trial?

-Tell us about your experiences with the clinical trial process.

-What, if any, barriers or challenges did you encounter? -What advice would you offer to future clinical trial teams? -Based on your understanding of CF research, what opportunities do you see for improving the way research is conducted and prioritized?

Due to the nature of the interview process, Research Voice members self-reported past participation in observational or interventional studies; responses were not verified. One response was excluded because the interviewee did not actually have any prior experience in clinical research. Individual response coding was performed, and initial keywords were identified. Keywords were then sorted, classified into themes and validated.

Results: Between December 4, 2017 -January 26, 2018, five CF Foundation staff members interviewed eighteen people with CF, nine parents of people with CF, and one spouse (n=28). The theme analysis identified several common challenges associated with clinical research participation, listed below, with percent of respondents answering affirmatively (answers not exclusive). These include:

Communication and Coordination (57%): Specific issues noted included lack of knowledge of trial opportunities, poor coordination within research sites, and lack of clear documentation.

Accessibility (50%): Issues identified included challenges associated with distance to research site, travel requirements, and time burden associated with site visits.

Trial Design (43%): Issues noted included exclusion criteria, time burden associated with design, and invasiveness of study design.

It is interesting to note that side effects were cited by only 4 interviewees. Conclusion: Although this qualitative study is limited in scope, these results demonstrate that communication and coordination, accessibility, and elements of study design are potential barriers to participation in clinical research and should be focus areas for improving the clinical trial experience for CF community members. Including the patient perspective in clinical trial design to identify the biggest barriers to participation may help improve trial enrollment and retention. Background: Parents of children with cystic fibrosis are at increased risk for anxiety and depression. As such, current guidelines recommend the screening of parental mental health. After successfully implementing patient-specific screenings, we focused on implementing parental screening. Despite having a written explanation of the rationale, only 6% of the parents within our clinic who were offered mental health screenings completed the assessment during their clinic visit, however 80% agreed to take information home to read on their own.

Methods: Due to the perceived lack of parental engagement in mental health screenings after the initial intervention, an emphasis upon targeted and more comprehensive parent mental health screenings was prioritized. Educational intervention included a psychoeducational training during a CF Family Education Day, which included information on the association between parental depression and health outcomes. "At-risk" parents are identified by the CF team during weekly interdisciplinary meetings, and the psychologist subsequently meets with identified parents in-person. Following a positive screening, treatment recommendations and referrals are made.

Results: Of the parents identified as "at-risk" and directly approached for screening, there was 100% participation in completing the screenings in clinic. This time-intensive approach, however, led to only 11 total parents being screened in this fashion. Of these, 5 had elevated scores on the PHQ-8 and/or GAD-7 which lends support to the assumption that the CF treatment team is effective in identifying "at-risk" parents. Parent screenings were sometimes performed in conjunction with family-based psychological interventions to assist in patient care, and as such, were eligible for reimbursement via Health and Behavior (H&B) Codes.

Conclusion: In the previous 2 years, psychoeducation on the importance of parent screenings provided in both a large group setting and in individualized settings did increase the rate of screenings completed in clinic. The screening completion rate was higher when each parent that was individually approached by the psychologist agreed to complete the packet in session. Given that each parent screening required more intensive clinical time, fewer parents were able to be met with in clinic. Therefore the increase in the "quality" of screenings did adversely impact the "quantity" of individual parents approached for screening. These findings suggest the best approach for parent screenings should be a combination of our 2 trialed methods: a social worker to continue the discussion and provide parent info during annual review of patient, and a psychologist to continue to meet 1:1 with identified "at-risk" parents as this proved to be an effective strategy for parents needing more time.

Channon, K.; Massey-Chase, R.; Heise, R.; Linkson, L.; Elston, C. Kings College Hospital, London, United Kingdom Introduction: Adherence to treatment in cystic fibrosis (CF) is complex and multifactorial. At our centre we proactively monitor adherence, completing I-neb downloads at every patient contact. If a downward trend in adherence is identified or patients report concerns then further support is offered. One approach to improve adherence is through our CF@home service. Strategies to optimise adherence include identifying barriers, setting goals, providing education on timing of nebulised drugs and practical problem solving, such as setting reminders. The aim of this study is to assess the impact of this home-based intervention on adherence to nebulised therapy.

Method: A retrospective case note review of home visits between July -December 2017 identified 22 visits (17 patients) where the primary reason for the visit was to support adherence. Data collected included; adherence at first visit (measured by I-neb download), intervention delivered, adherence at follow-up contact and exacerbation frequency measured by IV days per annum. Change in adherence was defined as ±5% of adherence at initial visit.

Results: 17 patients were seen; 7 male, age range 18 -42 years, FEV 1 predicted 20 -93% (median 71%). IV days: 0 -69 days (median 14 days). Median adherence at first visit was 29% (range 0-100%). Time between contacts ranged from 14 -154 days (median 40 days). 10/17 (59%) patients demonstrated improved adherence of 13 -88% (median 30%) at their follow-up contact. Adherence decreased in 6/17 (35%) patients and 1 patient had no change.

Those patients where adherence improved demonstrated practical barriers to adherence such as struggling to fit nebulisers around daily routine, forgetting, misconceptions about timings between nebulisers and a lack of understanding of the role of the drugs. The interventions delivered on these visits were a combination of problem solving, goal setting and education.

The patient whose adherence was static had baseline adherence of 0% and was unable to find her I-neb having just moved house. Follow-up contact identified that she was not ready to make a behaviour change due to external factors.

Of the 6 patients where adherence decreased; 2 are individuals who have complex psychosocial barriers to adherence; 3 have shown a long-term fluctuating pattern of adherence; 1 did not have a clinically significant decline as their adherence fell from 100% to 92%. All patients whose adherence did not improve were offered frequent contact with the team, comprising both clinic appointments and home visits.

Conclusion: Identifying the barriers to adherence requires an individualised approach. Michie and coworkers (Implement Sci. 2011; 6; 42) acknowledge that capability, opportunity and motivation are required to change adherence. Determining practical barriers and solutions in the home environment is an effective way of addressing "opportunity to change" and led to improved adherence in the short term in a significant number of patients. Where motivation is the main issue, an extended intervention to address willingness/readiness to change is necessary. Further evaluation is ongoing to assess the long-term impact of this focused adherence support at home. Methods: A multidisciplinary group at the University of North Carolina Adult CF Center completed a quality improvement project focusing on the implementation of outpatient ACP meetings as routine care for patients with CF. The target population was patients age ≥40 years or those with an FEV 1 of ≤40%, although patients who fell outside these parameters but expressed interest were eligible to participate. Patients were informed about the initiative via a quarterly newsletter and directly by the center's licensed clinical social workers (LCSW). An educational handbook was provided prior to the scheduled meeting. Patients were scheduled for a one-hour, billable ACP meeting with their CF provider, LCSW, dietitian, nurse coordinator, and designated support people. Effort was made to minimize interruptions by scheduling the meeting as the first or last visit of the day. Patients were invited to complete a survey using REDCap before and after their ACP meeting to assess their attitude toward end of life (EOL) care and ACP. Frequency of responses was tallied for each question.

Results: Of the 39 patients approached for ACP meetings, 17 patients have completed the meeting, 6 are scheduled, 9 have yet to be scheduled, and 7 have declined. Of those that have completed the meeting, 12 have an FEV1 ≤40%, 2 are ≥40 years of age, and 2 have both an FEV1 ≤40% and age ≥40 years. Seven pre-surveys and four post-surveys have been completed so far. Over 85% of patients felt that patients with cystic fibrosis have unique needs when it comes to ACP. All patients either agreed or strongly agreed it was important to discuss their prognosis and their wishes about EOL while they are healthy. However, 28.6% of patient did not feel comfortable talking about EOL care. When asked if they wished their provider would discuss advanced disease and EOL care, 71.4% of patients responded favorably. Prior to the ACP meeting 85.7% of patients had not discussed EOL care with their loved ones though 100% of patients stated they would like that opportunity. Only 3 patients reported documenting their wishes using an advanced directive, but 7 patients entered formal advanced directives into their chart after their meeting. The majority (71.5%) of patients would like to know more about lung transplantation.

Conclusions: In a multidisciplinary care model at a large academic CF center, we successfully implemented sustainable ACP for CF patients and continue to evaluate the effect of the program. While there is not yet enough post-data for quantitative analysis, qualitative feedback from patients has been positive. It is important for healthcare providers to meet the needs of this special patient population by ensuring that ACP is in place for patients prior to crisis situations. Objectives: CF management places significant burden on youth and their caregivers. Both are at risk for mental health concerns, such as anxiety/ depression, and families often face challenges in effectively communicating about CF-related issues. A diverse range of family psychosocial needs makes it difficult for care teams to identify the most pressing issues for patients; assessments that identify patient needs and priorities would be beneficial. This study developed and validated the CF Psychosocial Needs Interview (CF-PNI), a semi-structured interview that evaluates the psychosocial needs of pediatric CF patients and their families.

Methods: Families from a pediatric clinic participated (N=23; 14 male; 18 White; patient M age=9.6, SD=5.35). Families completed the CF-PNI with a psychology practicum student during a routine clinic visit. In the interview, families described the most significant challenges they encountered when managing CF, as well as the frequency and intensity of psychosocial stressors related to 3 domains of daily life (Treatment Adherence, Mental Health, Communication about CF). Family perceived level of need for psychosocial support in each domain was scored on a 5-point Likert scale; higher scores indicated greater need for support. Caregivers also completed scales evaluating treatment adherence (CF Treatment Adherence Rating Scale; CF-TARS), perceptions of patient quality of life (Cystic Fibrosis Questionnaire-Revised; CFQ-R), patient behavior (Strengths and Difficulties Questionnaire; SDQ), and family functioning (Family Assessment Device-General Functioning; FAD-GF). Patient lung function (% predicted FEV 1 [FEV 1pp ]) and body mass index (BMI) were also collected. Convergent validity was established with bivariate correlations between interview scores, parent rating scales, and health outcomes. Conclusions: The CF-PNI appears to be a feasible semi-structured interview for evaluating the psychosocial needs of patients with CF and their families. Despite the lack of a significant association between Treatment Adherence need scores and the subjective CF-TARS measure, needs scores from the interview demonstrated evidence for convergent validity across all three domains. Future work should further evaluate the psychometric properties of the CF-PNI, such as replication of current findings in a larger sample and evaluation of the stability of CF-PNI need scores over time.

Acknowledgment: Funded by CFF Student Traineeship Award MOLITO17H0. For adolescents and young adults, perceptions of treatment burden and health (ie, disease perceptions) may buffer the impact of anxiety/depression on CF health outcomes and serve as useful intervention targets. We hypothesized that treatment burden and health perceptions would moderate associations between mental health symptoms and poor health outcomes.

Methods: A subset of participants (13-25 years) from 12 pediatric and 9 adult CF care centers participating in the Success with Therapies Research Consortium's cross-sectional, observational Barrier's Study. Participants completed measures of depression (Patient Health Questionnaire (PHQ-8)), anxiety (Generalized Anxiety Disorder scale (GAD-7)), and health-related quality of life (Cystic Fibrosis Questionnaire-Revised (CFQ-R); subscales include Treatment Burden (CFQR-TB) and Health Perceptions (CFQR-HP)). Lung function (percent predicted FEV 1 (FEV 1 pp)) and body mass index (BMI) were extracted from health records. Separate moderation analyses using the PROCESS macro evaluated main and interactive effects for each independent predictor (PHQ-8, GAD-7) and each moderating variable (CFQR-TB, CFQR-HP) on FEV 1 pp and BMI, controlling for age and sex.

Data: 255 adolescents and young adults participated (M age=17.97; 87.5% White; 49.4% male; M FEV 1 pp=82.75 (SD=21.01); M BMI=21.66 (SD=3.39)); 84.2% of PHQ-8 and 83.3% of GAD-7 scores were in the minimal/mild range. Models predicting FEV 1 pp explained 12.4%-16.3% of the variance in FEV 1 pp (p-values<.001). CFQR-TB and CFQR-HP scores were significantly associated with FEV 1 pp (β=.31-.40, ps=.007-.037); PHQ-8 and GAD-7 scores were not significantly associated with FEV 1 pp. Models predicting BMI explained 5.0%-8.4% of the variance in BMI (ps=.003-.019); the model including PHQ-8 and CFQR-TB was nonsignificant. No hypothesized predictors were associated with BMI. CFQR-TB and CFQR-HP scores did not moderate associations between the PHQ-8, GAD-7, and FEV 1 pp or BMI.

Conclusions: Treatment burden and health perceptions emerged as significant correlates of lung function, but not BMI. Moreover, symptoms of depression and anxiety were not associated with CF health outcomes in this sample. Our hypothesis was not supported, suggesting that patients' perceptions of their health and burden may be better indicators of lung function than measures of psychological functioning. Even though at least 16% of our sample was moderately or severely anxious/depressed, measures of anxiety/depression were not a proxy for health status. Future research is needed to better understand the mechanism by which treatment burden and heath perceptions may be associated with lung function beyond the influence of psychological functioning. The CF Inpatient Child Life (CL) team recognized that patients with cystic fibrosis (CF) have many obstacles that prevent them from maintaining and engaging in interactions with peers, family, and community while hospitalized. Patients are restricted from socialization with peers or participating in developmentally appropriate group activities due to standard infection prevention and control practices. Socialization is a key component of mental and overall health. Peer interaction aids patients in developing positive relationships, improves their adherence with medical care, and enhances their ability to cope with a chronic illness. The CL team collaborated to create telepresence opportunities for patients with CF during hospitalizations.

Methods: Telepresence was facilitated through the hospital telemedicine "robot" and video conferencing using iPad technology. The robot represents the patient as a single entity, allowing them to virtually ambulate and engage with peers in group settings. Video conferencing allows patients to interact with those present at the activity, as well as other patients in isolation who are utilizing the technology. Patients are able to engage in a variety of social opportunities including teen art activities, playgroups, virtual therapy dog visits, connect with professional athletes, and other special events. Prior to the implementation of these virtual opportunities, patients in isolation were unable to participate in these group experiences.

Results: The telepresence experience has increased opportunities for socialization and normalization of peer interactions among CF children in isolation during hospital admission. Patients verbalized enjoyment of participation in these activities and operation of the telepresence technology. We gathered verbal feedback from 13 patients, ages 4 to 19 years old, who used the robot 16 times and the iPads 15 times. Patient-reported benefits included the ability to connect with peers, meet patients with a similar diagnosis, engage in age appropriate group activities, socialize with staff members, and focus on therapeutic interactions in a stressful environment. Limitations included hindered communication due to excessive background noise and patients not being able to physically attend events.

Conclusions and Future Plans: The use of telepresence among hospitalized CF patients led to an ability to connect and engage in therapeutic group experiences previously hindered by isolation. The Inpatient CL team will continue utilizing telepresence technology while expanding these services to additional inpatient populations. We will enhance this technology by placing a large computer monitor, camera, and microphone in the unit playroom to allow patients to "attend" playgroups and other programs. Expansion will also include collaboration between the Inpatient CL team, hospital based school teachers, Project Rise (Reaching for Independent Successful Employment), and Social Work. These telepresence interactions will utilize iPad technology for school counseling, career building assistance, and creation of a peer-to-peer mentoring program. Our vision is to facilitate the creation of a virtual community for hospitalized CF patients who require isolation.

Robnett, J. 2 ; Haberman, D. 1 ; Burbidge, K. 1 ; Zobell, J.T. 1 ; Collingridge, D. 1 1. Intermountain Healthcare, Salt Lake City, UT, USA; 2. Pediatrics, University of Utah, Salt Lake City, UT, USA Introduction: Individuals with CF experience symptoms of anxiety and depression at a rate that is 2-3 times higher than those in the general population (Quittner, et al. Thorax. 2014; 69:1090-97) . In 2013 the CF Foundation recommended annual mental health screenings for all adolescent and adult patients.

In data from the CDC, Utah had the 5th highest number of reported suicides in the United States in 2016. This is an epidemic within the general population and we suspect higher rates of anxiety and depression exist in the CF population in Utah.

Objective: To explore the prevalence of anxiety and depression in the adolescents we serve at Intermountain Pediatric CF Center and to examine the relationship between mental health symptoms and physical health outcomes in these patients.

Methods: We began to complete mental health screening with adolescents age 12 in November 2016 using the PHQ9 and GAD7. In 2017, 106 of 110 eligible patients completed these screenings. We completed a retrospective chart review to obtain demographic information (age, sex, private or government insurance), physical health data (number of hospitalizations, BMI, FEV1) and dornase alfa refill data for each of these patients. Analyses were completed for all variables.

Results: Rates of depression in our adolescents were more than twice the reported rates in the TIDES study (24% vs 10%). Additionally, 8% of adolescents responded positively to question 9 of the PHQ9 (suicidal ideation). Prevalence of anxiety was 29%.

A t-test revealed that there was a sex difference on both mental health measures, with females reporting higher rates of both depression and anxiety. This difference was statistically significant for the GAD7 (p=0.046), but due to low statistical power caused by increased variability, was not statistically significant for PHQ9 (p=0.13). However, the difference on the PHQ9 is still likely clinically significant.

Results of a generalized estimating equations (GEE) analysis found that increased scores on the PHQ9 were negatively correlated with monthly dornase alfa refill adherence (p=0.05). However, scores on the GAD7 were positively correlated with refill adherence (p=0.04).

Conclusions: Adolescents with cystic fibrosis at our center endorsed higher rates of depression than reported in the TIDES study (24%). Additionally, 8% of these patients indicated suicidal ideation. These rates are alarming, as Utah has one of the highest rates of suicide in the country. Additionally, rates of anxiety are similarly high, with 29% of patients endorsing symptoms at the mild level or above. Findings indicate that females endorse higher rates of both anxiety and depression.

Dornase alfa adherence was found to be positively correlated with symptoms of depression and negatively correlated with symptoms of anxiety. This finding needs further study to determine clinical implications.

This data supports the importance of conducting mental health screenings. While this process adds a significant burden to clinic flow and requires appropriate staffing, we feel strongly that the emotional distress our patients report is critical to address as part of their overall treatment.

Bruschwein, H.; Soper, M.; Albon, D. Pulmonary, University of Virginia, Charlottesville, VA, USA Introduction: Patients with cystic fibrosis (CF) have higher rates of depression and anxiety than the average population, with 19% of adults with CF endorsing elevated rates of depression and 32% endorsing elevated rates of anxiety (Quittner AL, et al. Thorax. 2014; 69:1090-7) . Therefore, mental health treatment for these patients should not be overlooked. In 2015, the International Committee on Mental Health in Cystic Fibrosis recommended annual screening of patients with CF with the PHQ-9 and GAD-7 (Quittner AL, et al. Thorax. 2016; 71:26-34) . In response, many CF clinics implemented screening programs as part of outpatient clinic visits.

Although there has been increased awareness of the mental health needs of patients with CF, this has primarily been done through screening and interventions in the outpatient setting. Little research has been done to examine the prevalence of anxiety in the medical inpatient setting, especially in patients with CF who have been hospitalized for a CF exacerbation. This quality improvement pilot study aimed to identify prevalence of anxiety in patients with CF during hospitalizations for CF exacerbations and at post-hospitalization clinic visits.

Methods: Patients with CF who were hospitalized for a CF exacerbation were screened with the GAD-7 anxiety screener during hospitalization. Patients were screened again during their post-hospitalization follow-up clinic visit.

Results: The screeners were completed by 19 patients. During hospitalization, 53% of patients endorsed elevations in anxiety (32% endorsed moderate symptoms of anxiety, 21% endorsed mild symptoms, and 47% endorsed minimal symptoms). On the post-hospitalization anxiety screener, 32% endorsed elevations in anxiety (6% endorsed moderate symptoms of anxiety, 26% endorsed mild symptoms, and 68% endorsed minimal symptoms).

Conclusion: Awareness of the mental health needs of patients with CF is increasing and many CF clinics conduct mental health screening during outpatient clinic visits. However, anxiety during hospitalizations may be overlooked and undertreated. The prevalence of anxiety during hospitalizations for CF exacerbations appears to be significantly higher than the baseline rate of 32% found by Quittner and coworkers (2014). While anxiety appears to decrease after discharge and long-term treatment may not be needed, there is a need for short-term interventions to treat anxiety during hospitalizations. Future research should focus on development and implementation of anxiety interventions during hospitalization. Additional screening and further data collection are in process and will be included in the final presentation.

Acknowledgements: Mental Health Coordinator Grant supported by the Cystic Fibrosis Foundation. Objective: Several research studies indicate that individuals with cystic fibrosis (CF) and their caregivers have a high prevalence of depression and anxiety. Integrating mental health supports and services in CF clinical care is a proactive approach and has the ability to assist patients and caregivers with improving their overall quality of life; however, the process of doing so presents many challenges (e.g. availability of qualified providers and long wait times). The goal of this project is to highlight the process that has been developed to efficiently and effectively triage and treat the mental and behavioral health needs of individuals with CF and their caregivers.

Methods/Process: With the support of the Mental Health Coordinator grant, the CF social worker was hired as the Mental Health Coordinator and a psychologist with CF experience was hired as the mental health provider and consult. The providers worked together to develop a triage process for meeting the mental and behavioral health needs of patients and their caregivers. The social worker provides psychosocial education at every clinic visit and administers annual mental health screenings to patients and caregivers. Individuals with elevated screens are referred to the psychologist in order to receive immediate interventions. Individuals are able to receive 3 therapy sessions from the psychologist free of charge. If individuals need additional support beyond the initial sessions, the social worker assists with locating providers in close proximity to his/her home or the individual has the option to continue billable services with the CF psychologist. In order to support the individual's needs in his/her community, the social worker has begun to develop a mental health network based on the counties of residence for the entire CF clinic's patient panel. A provider survey has been developed and contact information for providers is continually being gathered. The survey also gathers information on the provider's interest in obtaining additional information about CF.

Mental health screens July 2016 -June 2017: 7 patients (aged 12 and older) and 31 caregivers. The psychologist provided services to 10 patients and 1 caregiver and 2 of the patients received telemedicine services. The psychologist continued therapy services with 4 patients beyond the three sessions. No community referrals were given. To date, 42 community providers were sent a survey. Findings: The model served to efficiently identify CF patients and caregivers in need of mental health support and quickly provided access to treatment, decreased wait-time by at least 10 weeks, saved caregivers time in regards to completing a provider search, intake and scheduling.

Limitations: Transition to insurance coverage, confusion about which mental health issues are covered by the grant and the hesitancy of patients to transition to a community provider.

Recommendations: Continue current model. Gather data on satisfaction and sustainability. Build a database of community providers and provide information, consultation and training as needed. Share the model with other CF centers. however, it has also presented some unique challenges that warrant further inquiry and discussion.

Program Summary: For the PS component of care, patients met with either a licensed psychologist or a post-doctoral fellow under supervision. Consistent with CFF recommendations, patients (or their parents if under 13 years of age) were administered the PHQ-9 and GAD-7 to assess depression and anxiety symptoms. The PS provider also routinely assessed general emotional well-being, treatment regimen adherence, diet and sleep patterns, academic and social functioning, and status of and reliance on support system. Parents of younger patients were also asked to complete a short version of the Pediatric Symptom Checklist (PSC). During the visit, these screening measures were reviewed and discussed. When patients or their families identified problem areas in session, brief cognitive-behavioral interventions, psychoeducation, and outpatient PS referrals were provided. If patients screened negative, reassessment was recommended twice a year. If patients screened positive, reassessment was recommended at the following visit.

Results/Discussion: Over the first year, approximately 280 PS visits occurred as a component of patients' care during CF clinic follow-up visits. These visits occurred during 55 separate clinic sessions, with the PS team averaging 5 visits per clinic. A significant barrier to treatment, discovered soon after implementation, was the impact of poor health insurance coverage for mental health care. On multiple occasions, families ultimately opted out of future PS visits because of the additional provider copay required by insurance. Very few patients declined to see a PS provider based solely on fact they were not interested in PS support, which is optimistically indicative of movement away from the stigma associated with mental health issues preventing needed care. A second barrier was the extreme paucity of qualified PS providers across Oklahoma. Often, a family would be open to assistance to address an identified PS issue, but the challenge was finding a qualified provider covered under the patient's insurance plan. In summation, the primary barriers to care during the first year of implementation were additional cost burdens on families that already oftentimes had significant medical expenses and a limited referral base of qualified PS providers. To address these two issues moving into the second year of implementation, the PS team plans to use psychologists-in-training more strategically to provide supervised psychology services when families' insurance coverage for licensed PS provider visits is poor/cost prohibitive. The PS team is also working on the development of a seminar-based training program to better equip and build the referral base of local providers that can confidently and competently provide outpatient mental health services to Oklahoma's CF population. to infection control guidelines. To better support our patient population, an online support group was created to provide a space for CF patients to gain support and address feelings of isolation.

Methods: An email was sent out to the UNC CF population in February 2017 to survey interest in an online support group. Responses were received from 40 out of 289 patients, with 87.5% of respondents endorsing interest in participating in an online support group. Staff then worked with the UNC legal team to create a participant agreement. The first support group was offered January through March 2018 through the Bluejeans platform and included four CF patients and one LCSW facilitator. Participants were selected from those who voiced interest and who had already established a therapeutic relationship. A second eight-week support group is currently underway. The format for the group includes one introductory/ rapport building session, six focused topic sessions selected by the group participants, and one termination session. Potential topics offered include: Managing Daily Life, Managing My Mood, CFRD, Transplant, Affording Care, and Relationships. Participants are asked to complete a pre-and post-survey to gauge the effectiveness of the group. The survey includes four Likert scale questions measuring the feeling of connection to the CF community, difficulty discussing aspects of care, feelings of isolation having CF, and comfort with technology.

Results: To date, a total of nine patients have participated (eight females and one male). Three participants from the first group (75%) completed preand post-surveys. Post-survey data are not yet available from the second group. In this limited sample, we observed improvements in feelings of isolation due to CF and comfort with technology. Free-text responses provided by these patients include the response, "I don't remember ever really meeting any other people with CF prior to this group, so it was nice to meet others and learn that things that affected me also affected them (more or less) so I did not feel as 'crazy' for feeling the way I did and isolated." Other positive aspects of the group noted were "the automatic connection to others," and "the topics and size of the group." Of note, members of the group expressed interest in extending their participation, and a peer-led group option is being arranged.

Discussion: The implementation of CF mental health guidelines at UNC has allowed opportunities for further rapport building and conversations around problems coping with chronic illness. Although we are early in the process, development of this resource and early responses from participants suggests that it can help patients develop a support system that includes other CF patients, while maintaining contact precautions. Ideas for the future include CF population-specific groups, such as a men's group or a young adult group.

Crowley, E.; Brown, C.D. Indiana University, Indianapolis, IN, USA Objective: In a prior retrospective analysis of patients with CF transferring to adult care, social complexity was both a strong predictor of hospitalizations and had a strong negative association with lung function after transferring to adult care (Crowley EM, et al. Pediatr Pulmonol. 2018; 53:735-40) . In a prospective cohort of adolescent subjects, we hypothesized that social complexity (measured by the Bob's Level of Social Support, BLSS), transition readiness (TRAQ), and treatment complexity (TCS) are associated with health care use and lung function (percent predicted (pp) FEV 1 ).

The study design is a cross-sectional, observational study of adolescent patients with CF. Patients with a diagnosis of CF between the ages of 14 to 19 receiving care in the pediatric CF center at Indiana University were approached for enrollment. Outcome variables included annual hospitalization rate, annual outpatient visit rates (measured as the number of hospitalizations or outpatient visits over the 2 years prior to enrollment) and ppFEV 1 at enrollment. Dependent variables of interest included ppFEV 1 , age, TCS, BLSS and TRAQ. To understand the relationship between the dependent variables and the outcome variable, multivariate linear regression was used.

Results: Fifty patients were enrolled with an average age of 16.5 ± 2 years. Sixty-eight percent were female. Forty percent had Medicaid insurance. The mean ppFEV1 was 86.2 ± 22.7% using GLI equations. The median assessment scores were: TRAQ 64 (IQR 56-73), TCS 14 (IQR 12-17), and BLSS 7 (IQR 6-8). The average annual hospitalizations were 1.0 ± 1.4 and average annual outpatient visits were 4.6 ± 1.6. After adjusting for age, TCS, BLSS, and TRAQ, the only significant predictor of outpatient visit rate was ppFEV 1 (β = -0.04, p < 0.0001). However, age, BLSS, and ppFEV 1 were significant predictors of hospitalization rate (Table) . When adjusting for age, BLSS, and TRAQ, the only significant predictor of ppFEV 1 was TCS (β = -3.4, p < 0.0001).

Conclusions: When adjusting for other factors, our results suggest that transition readiness knowledge is not significantly associated with health care use outcomes. In addition, these findings suggest that treatment complexity is strongly related to lung function, which would be expected because of the nature of the measure. These findings also validate those previously found--that social complexity is a strong predictor of hospitalizations and it is regardless of self-care knowledge. Patients with high social complexity are a high-risk group where efforts to develop interventions throughout childhood and adolescence may mitigate potentially preventable health care use and improve future outcomes. As the CF population ages and treatment options evolve, it is increasingly important for patients to discuss and document advance care decisions (Dellon EP, et al. J Cyst Fibros. 2016; 15:96-101) . Over the past 5 years, our adult CF center has been involved in quality improvement initiatives to improve palliative care, which includes ACP and EOLC. The objective of this study was to evaluate whether any changes have occurred in our patients' perceptions of ACP and EOLC over this time period.

Methods: Surveys that included questions about ACP and EOLC were administered between April-June 2013 and again between March-May 2018 to patients in a single adult CF center. Survey completion was voluntary and responses were blinded. Questions were adapted prior surveys (Sawicki GS, et al. J Palliat Med. 2008; 11:1135-41) to assess patients' attitudes and decisions regarding ACP and EOLC. Seven questions related to ACP and EOLC were selected for evaluation. Chi-square analysis was performed to determine differences between 2013 and 2018. A p-value of 0.05 was utilized to determine significance.

Results: In 2013, a total of 37 surveys were returned, representing 86% of surveys distributed. In 2018, a total of 28 surveys have been returned, representing 52% of available patients. In both 2013 and 2018, most patients reported thinking about their future "occasionally/often" (59% vs 61%), not having talked with anyone about EOLC (57% vs 56%), and not having completed health care proxy forms (75% vs 82%). A statistically significant increase was found from 2013 to 2018 in those who thought "very much or somewhat" (43% vs 68%, p<0.05) compared to "not very much" about whom they would like to make or communicate decisions on their behalf if they were unable to do so. While most patients continued to report being unsure about their wishes for future medical treatment if they become very ill (47% vs 46%, p=0.46), there was a trend toward more patients reporting to have specific wishes in 2018 (53% vs 73%, p=0.22). Nonstatistically significant increases also were seen from 2013 to 2018 in the percentage of individuals who were very comfortable with talking with their family about the medical care they want if too ill to make their own decisions (57% vs 70%, p=0.29), and talking with their CF team about the medical care they want if too ill to make their own decisions (54% vs 71%, p=0.15). when the patient is assessed utilizing the PRAP assessment and continues as outlined by the PRAP algorithm. By working on this process, we expect to identify opportunities to develop and support psychosocial coping skills for patients younger than 12 years old, a population not currently evaluated under the CFF mental health screening guidelines.

Method: 106 PRAP assessments of patients ages 3 to 11 years old were completed by a CCLS during an outpatient clinic visit or prior to an invasive procedure from 2016 to early 2018.

PRAP takes into account a number of aspects that can affect coping. These include communication, special needs, anxiety and coping, temperament, caregiver stress, past healthcare encounters, invasiveness of encounter, and developmental impact. Each category is assessed on a zero to three point scale. For the purposes of this abstract, we specifically looked at the patients' anxiety and coping.

Results: In a retrospective review of the data, 39 assessments identified patients who did not require support at the time of the clinic visit, 46 assessments identified patients who were "able to cope with support," 19 assessments identified patients who were "visibly distressed but were able to calm with support," and 2 patients who were "unable to utilize coping strategies or were not easily calmed during their clinic visit or after an invasive procedure." (Psychosocial Risk Assessment in Pediatrics (PRAP). Cincinnati Children's Hospital Medical Center. 2012.) Patients whose PRAP showed increased need for support received interventions to meet their needs. Interventions included increased child life support, connecting patients to the mental health coordinator before the age of 12, or consultation with a pediatric psychologist.

Discussion: Evaluating younger patients utilizing PRAP aids in identifying those patients who need help to develop coping skills. Use of this assessment also provides opportunities to connect patients to mental health services earlier if needed. Introduction: Given the improvements in the diagnosis and treatment of cystic fibrosis (CF) and the increasing number of CF living into adulthood (Elborn SJ, et al. Eur Respir J. 2016; 47:420-8) , a need to focus on the self-management and health care transition from paediatric to adult health care has become paramount. Although in most Italian systems there has been ongoing centralised CF care with combined paediatric and adult services, projects to create independent adult CF centres are increasing. Thus, it is important to develop and implement the knowledge about readiness to transition and patients' functional and psychological status. In addition, we know that anxiety and depression symptoms have been associated with worse health outcomes (Goldbeck L, et al. Cochrane Database Syst Rev. 2014; 6:CD003148) . To date, the relationship between self-reported health care transition readiness tools and psychological symptoms of Italian CF patients has never been systematically explored. This study investigates the relationship between transition readiness and patients' functional and psychological status in adult CF patients which are followed in a pediatric center.

Methods: We recruited 52 adult CF patients (F/M=23/29, mean age 28.6±8.3 SD y, FEV 1 74±23% predicted, BMI 21.7±2.6). All patients completed: a)Transition Readiness Assessment Questionnaire (TRAQ); b) Readiness to transfer questionnaire-CF (RTQ-CF); c) Patient Health Questionnaire-9 (PHQ-9); d) Generalized Anxiety Disorder-7 scale (GAD-7); e) Cystic Fibrosis Questionnaire-Revised Teen/Adult (CFQ-R + 14).

Results: Age and pulmonary function were strongly related to readiness (TRAQ and RTQ-CF). Looking at TRAQ domains we found a relationship between age and "overall score" (R = 0.32, p = 0.02), "appointment keeping" (R = 0.41, p = 0.002) and "talking with providers" (R = 0.28, p=0.04). RTQ-CF overall score was related to FEV1 both expressed as absolute and percent of predicted (R = 0.33, p=0.01 and R = 0.28, p=0.04). Inverse relationships were found between: 1) anxiety and BMI, FEV1, FVC (R = -0.27, p=0.04; R = -0.29, p=0.03; R = -0.32, p=0.02 respectively); 2) depression and BMI (R = -0.37, p=0.006). In TRAQ results, CF females had higher scores of "managing daily activities" compared to CF males (p = 0.005) and there was a relation between "overall score" of TRAQ and respiratory domain of CFQ-R (R = 0.40, p=0.02). In CF males we observed a positive relationship between anxiety/depression and BMI (R = -0.40, p=0.05; R = 0.51, p=0.01).

Conclusions: Transition readiness scores are related to age and disease severity CF in adults. Depression, anxiety symptoms and quality of life domains were related to patients' functional status. Considering the need to develop and implement effective transition programme in Italy, our data suggest that it will be important to use combined tools of clinical and psychological variables to guide educational interventions by providers to support transition. (CF) and their caregivers were at risk for elevated levels of depression and anxiety. The concerning rates of depression and anxiety in patients and caregivers contributed to the International Committee on Mental Health in CF developing a consensus statement for screening and treating depression and anxiety. In this statement, clinic-based screening for depression and anxiety for patients and caregivers was outlined using the Patient Health Questionnaire (PHQ) and the Generalized Anxiety Disorder (GAD-7) questionnaire. The CF Center at Ann & Robert H. Lurie Children's Hospital of Chicago (Lurie Children's) started screening in patients in 2014 and incorporated caregiver screening in 2016. The objective of this study was to assess the change in rates of depression and anxiety in patients and their caregivers since the 2016 screening and treatment referral protocol was initiated.

Methods: Lurie Children's utilized an interdisciplinary team of a coordinator, pediatric social worker, and adult clinical psychologist to implement depression and anxiety screening for patients and caregivers in our pediatric program. The team collected annual screens, using the PHQ and GAD-7, follow-up screens where indicated as well as psychoeducation on mental health and self-care, and provided referrals to mental health treatment based on patients' and caregivers' severity of symptoms.

Results: In 2016, the rates of elevated depression were 8% for the patients and 29% for the caregivers, and the rates of elevated anxiety were 32% for the patients and 28% for the caregivers. Twenty-three percent of the patients and 15% of the caregivers declined the depression and anxiety screen. In 2017, the patient rate of elevated anxiety improved in 2017, with only 13% of reporting elevated scores, whereas the rate of elevated depression remained the same at 8%. Caregiver rate of elevated depression and anxiety improved over the year: 14% had elevated anxiety and 12% had elevated depression scores. Twenty-four percent of patients and 41% of caregivers declined the screen. The most recent data for 2018 will also be presented.

Conclusion: Over the two years where both patients and caregivers were screened, rates of depression and anxiety declined, except in regards to depression rates in patients, which remained stable. The overall improved rates of depression and anxiety may reflect the positive impact of psychoeducation, support, follow-up, and treatment referrals, which calls for continuation of screening and referrals. A high percentage of patients and caregivers declined the mental health screen, which may have influenced the rates of elevated depression and anxiety. The reasons behind patients and caregivers declining the screens need to be examined in future research to address this barrier to implementation.

Foil, K. 1 ; Christon, L.M. 1 ; Szentpetery, S.E. 2 ; Flume, P. 1 1. Medicine, Medical University of South Carolina, Charleston, SC, USA; 2. Pediatrics, Medical University of South Carolina, Charleston, SC, USA Introduction: Annual screening for at least 1 primary caregiver (CG) of children with CF (ages 0-17) is recommended (Quittner AL, et al. Thorax. 2016; 71:26-34) due to a 2-3 fold increased risk for depression and anxiety (Quittner AL, et al. Thorax. 2014; 69:1090-7) . Care centers are tasked with development of care and referral pathways, screening implementation, care coordination and follow-up. The first year of CG screening in a southeastern pediatric CF center is reviewed.

Methods: A 3-year CF Foundation Mental Health (MH) grant allowed inclusion of a clinical psychologist in clinic to lead MH initiatives, and a genetic counselor assisted. Delivery of a resource packet and self-screens (PHQ-8 and GAD-7) began in 4/2017. Institutional Legal/Compliance concerns about inclusion in the electronic medical record (EMR) were resolved with CPT 96161: screening a CG for patient benefit; and screening began in 11/2017. CGs voluntarily completed an electronic PHQ-8 and GAD-7 at their child's visit. Supportive feedback and self-care tips were given to all. CGs with elevated scores in the moderate-severe range were encouraged to seek diagnostic assessment and counseling or pharmacologic treatment. CGs needing MH care at our institution had a separate medical record number created for referral and subsequent care. Others were provided a statewide network of counselors or guided to a primary care physician (PCP).

Results: Between 4/2017 and 11/2017 we had 75 eligible patients, corresponding to 72 CGs (3 sib pairs). MH resources and self-screens were provided to CGs of 67 (89%) patients. From 11/2017 to 4/2018 there were 2 new diagnoses, and CG score capture was completed for ≥1 CG of 55 of 77 (71.4%) patients. A total of 54 CGs completed the screening; this includes 1 CG for each of 47 patients, 3 CGs of siblings, and both parents of 2 patients by request. CGs were mothers (72%), fathers (15%), custodial extended family (6%), grandparent (5%), and foster parent (2%). The number of CGs screened per clinic ranged 0-5 (mean 2.25).

For CGs that scored in the moderate-severe range in either scale, 5 (36%) accepted referrals to institutional behavioral medicine. Six (43%) agreed to see their local MH service and/or PCP. Three (21%) declined referral due to perceived lack of need (1), no insurance (1) or bad timing (1). CGs not screened include 7 who were eligible but were missed (acute care needs, time conflicts, one needs screens in Spanish), 7 who were not yet due for screening, and 7 with no clinic visit in this period.

Conclusions: These data demonstrate the implementation and feasibility of CG MH screening in clinic. CGs were receptive to screening and referrals. Rates of depressive and anxious symptoms in our CG cohort are consistent with published rates. Recommendations for navigating institutional barriers to implementation are offered. Future efforts will target screening of remaining CGs within the year, tracking follow-up adherence and interventional outcomes, and identifying areas for quality improvement. 

Tluczek, A. 1 ; Grob, R. 2 ; Greene, L. 3 ; Van Gorp, S. 5 ; Raymond, K. 4 ; Collins, L. 4 ; George, C. 4 1. School of Nursing, Madison, WI, USA; 2. Center for Patient Partnerships, Madison, WI, USA; 3. Happy Heart Families, Seattle, WA, USA; 4. CF Foundation, Bethesda, MD, USA; 5. See What I Mean Consulting, Cedar Falls, IA, USA Objective: This quality improvement project was designed to increase our understanding of parent, patient, and provider perspectives about working together.

Methods: Staff of the Partnerships for Sustaining Daily Care initiative conducted audio-recorded interviews with CF providers from various disciplines (N = 20), parents of children with CF ages 2 to 22 years (N = 7) and one patient. Interviewers introduced partnership as the focus of discussion. We used grounded dimensional analysis of transcribed de-identified interview data to inductively identify and describe social processes associated with clinical partnership as reported by interviewees.

Results: Provider interviews revealed two central dimensions of partnership: mutuality and reciprocity. For example, a provider defined partnership as, "when both [patient and provider] have a voice and we have a shared interest and we are trying to develop a plan so that we both have buy in." Providers' self-reflections suggested that their actions are motivated by respect, compassion, empathy, and caring about the patients and families they serve. They also value the long-term nature of these relationships. Although providers reportedly adopted a partnership framework, their exemplar stories across disciplines included strategies that were sometimes discrepant with their philosophical descriptions of partnership and were more aligned with an "adherence" or "compliance" framework that involved persuading patients to follow-through with "what she [patient] should be doing." Parents consistently described partnership as parents and providers "working together" in the best interests of the child/patient in which each member contributes parallel and complementary expertise. Parents described how their unique vantage points of time with and proximity to their children allows for observations of subtle changes or idiosyncrasies that confer an understanding of their child's psychosocial, developmental, and health needs. Parents used these parental understandings to advocate for their children's evolving health care needs while maintaining a "normal" life. While providers tended to focus on the child's physical needs, parents offered more holistic perspectives about the child and family life. In combination, they become the recipe for a successful and mutually gratifying partnership.

Conclusions: The most notable difference between provider and parent interviews was that parents acknowledged the expertise that both parent and provider contribute to partnerships, whereas providers tended to focus on strategies to facilitate child/patient adherence to the prescribed treatment plan. Provider did not acknowledge parental expertise. In fact, one parent recounted a provider's pleasant surprise at the favorable outcome resulting from a nutrition plan that honored parental preferences regarding breastfeeding. Parents also talk about partnership upstream, e.g. at time of treatment plan development; providers discussed downstream partnership strategies used to enhance adherence to the treatment plan.

Introduction: Recent improvements in treatment regimens have extended the life expectancy of cystic fibrosis (CF) patients. However, the early mortality, constant hospitalization, and time commitment to managing CF directly contribute to poor mental health in these patients. Mental health issues such as increased risk of anxiety and depression have become a forefront topic in the CF community, including how to identify and treat these conditions. Depression and anxiety can have adverse effects on CF patients in that they are less likely to be compliant with treatment, have worse lung function, and have increased number of hospital visits. The goal of this study was to determine whether CF mouse models exhibit the phenotypes of anxiety and depression and to explore potential mechanisms. We have shown that slower microtubule reformation rates as well as decreased acetylated microtubule levels are seen in CF tissues and hypothesize that these microtubule alterations contribute to behavior changes.

Methods: Knocking out histone deacetylase 6 (Hdac6 -/-) in CF epithelial cells restores the level of acetylation levels and reformation rates toward normal levels, as well as improves growth and reduces inflammation in CF mouse models. Therefore, we used wild-type (WT), F508del (CF), Hdac6 -/-, and CF mice bred with Hdac6 -/mice to knockout Hdac6 in a CF mouse model (CF/Hdac6) to test our hypothesis. FABP-driven Cftr corrected mice (FD) were used as controls since these mice express Cftr in the brain. The Elevated Plus Maze (EPM) and an Open Field Box (OF) were used to measure anxiety and a Tail Suspension (TS) assay was used to measure depression using 4-, 8-and 12- week-old male and female mice with approximately 10 mice in each group.

Results: A higher level of depression was seen in the 8-week-old CF mice as measured by immobility time compared to WT mice. The depression phenotype did not occur in the CF/Hdac6 mice (WT: 79.7 sec ± 8.6, CF: 105.1 sec ± 7.4, CF/Hdac6: 78.8 sec ± 6.5). Both the 4-and 8-week-old CF mice appeared more anxious as seen not only by traveling less distance in the EPM but as well as in the OF box when compared to WT mice (EPM 4 week: WT: 641.2 cm ± 56.6, CF: 367.7 cm ± 36.4; 8 week: WT: 878.1 cm ± 84.7, CF: 658.5 cm ± 82.1) (OF 4 week: WT: 1220.5 cm ± 117.3, CF: 652.8 cm ± 77.9; 8 week: WT: 1662.9 cm ± 121.9, CF: 1140.1 cm ± 107.7). All differences are significant as determined by t-test. Depletion of Hdac6 had no impact on anxiety measures, whereas, FD mice were consistent with WT values showing Cftr-dependence for the phenotype.

Conclusions: CF mice display behavior consistent with heightened anxiety at both 4 and 8 weeks of age, and depressive behavior at 8 weeks of age suggesting a possible post-pubertal link to depression. Hdac6 depletion reverses the depression phenotype, but not anxiety related behavior, suggesting the two phenotypes are regulated by separate mechanisms. Hdac6 depletion fixing post-pubertal depression is consistent with our previous work showing that Hdac6 knockout restores post-pubertal growth in CF mice and points to Hdac6 regulating a neurological developmental process in CF. Acknowledgment: Supported by a grant from the CFF. The age of rapid development of CFTR modulator therapy has brought new hope and promise for improved quality of life and longevity for individuals with cystic fibrosis. The approved CFTR modulator therapies have differing efficacies and are indicated for select populations of patients. The long-term trajectory of lung function stability and potential increased life span for any of these individuals can be speculated but is still largely unknown. Thanks to persistent advancement of maintence therapies and implementation of best practice, even prior to the advent of CFTR modulator therapy, the age of survival of individuals with CF has been on the rise. Despite improvements in pulmonary health for individuals with CF, it remains a chronic illness with a heavy disease burden. Increased rates of anxiety and depression are found in individuals with cystic fibrosis (Quittner AL, et al. Thorax. 2014; 69:1090-7) . Some CFTR modulator therapies have known medication interactions resulting in decreased efficacy of SSRIs. The complex overlay of promising new therapies, high rates of depression, and many individuals still left without CFTR modulator therapy brings to question what impact CFTR modulator therapy has had on mental health over the last 5 years. The Cystic Fibrosis Center Western New York has been uniformly screening individuals with cystic fibrosis ages 12 and up for the last 5 years. We aim to analyze our existing data to identify trends of mental health change that have been found in relation to the availability and prescription of CFTR modulator therapy. These factors combined likely have a significant impact on mental health and outlook on life for individuals with CF and may be helpful for the CF care team to provide optimal mental health monitoring and treatment.

Methods: PHQ9 and GAD7 data available from 2013-2017 will be examined for changes in depression and anxiety incidence and severity in relation to CFTR modulator therapy availability, prescription and efficacy as measured by health care outcomes (FEV1, BMI, exacerbation rate and mortality).

Results: Results being analyzed at present. Discussion: Pending full result review. (Quittner, Thorax, 2015) . In 2013, the CF Center of Western New York began routine depression screening. We previously evaluated health outcomes related to incidence of depression and hypothesized that rate of decline of health outcomes since our screening initiation would be worse in depressed individuals. We evaluated the effect of depression on patient health outcomes including FEV 1 percent predicted (pp), BMI, 25-hydroxyvitamin D levels and hospitalizations.

Methods: Individuals with CF, ≥12 years old, were assessed annually from 2013-2017 for symptoms of depression with PHQ-9. Patients who screened positive for depression received further assessment and intervention based on severity. Health outcome measures (FEV 1 pp, BMI, 25-OH D) at the time of annual screening and incidence of hospitalization in the year prior was recorded.

Results: There was a trend toward a difference in FEV 1 pp in adults who screened positive for depression in 2017 vs. those who did not (66.5 vs 75.2%, p=0.09). When adults were examined for FEV1 difference if they ever screened positive for depression vs. not, the difference was more pronounced and statistically significant (65.8 vs 78.4%, p=0.01). A greater proportion of hospitalizations for those who screened positive for depression vs. those who did not was again seen reaching statistically significance (chi-square = 6.16, p = .01). Longitudinally, from 2013 to 2017, combined data were analyzed and longitudinal multilevel modeling is still underway. Predicted FEV 1 percent decline was significantly greater in individuals who screened positive for depression vs those who did not with a decrement of -6.39% last year with an increased gap in FEV1 decline over time seen in individuals who screened positive for depression vs. those who did not. Again, these data are still being examined with further analytic scrutiny.

Conclusion: Depression screening in CF is relevant to important health outcomes. Significant decreased mean FEV1 pp was seen in adults who were depressed vs. those who were not over a 5-year time period. These findings support the recommendation to continue to screen and treat for depression in CF as education and treatment may potentially impede lung function decline and improve health outcomes. Gilmore, D. 1 ; Hooper, C. 1 ; Nemastil, C.J. 1 Background: Marijuana (THC) has potential adverse psychosocial and physical effects in patients with CF. Despite reports of 20% or more usage among the CF population, the effects of THC on adherence, healthcare utilization, mental health disease, and traditional measures of CF health are poorly understood.

Objective: The primary objective of this study was to compare key data between CF patients who self-reported THC use and control CF patients without self-reported THC use.

Methods: Patients with CF between the ages of 16 and 44 years who self-reported use of THC between 2012 and 2016 were identified from electronic medical records. A control group of age-matched patients was also identified from the remaining CF population. A retrospective chart review was performed to collect data on demographics, healthcare utilization, presence of co-morbid conditions, lung function and nutritional parameters.

Results: Among a total population of 239 patients in the targeted age group, 22 individuals reported THC use (prevalence 9.2%). There was no statistical difference in age, presence of F508del mutation, FEV1, or BMI between groups. Compared to controls, patients who self-reported THC had a significantly higher percentage of no-shows to CF clinic appointments (42% vs. 33%, p=0.02). THC reporters were also more likely to have a clinical diagnosis of depression and/or anxiety (73% vs. 41%, p=0.03) and more likely to use tobacco products (36% vs. 4%, p=0.01). While THC-reporters did have more frequent emergency department utilization than controls (median 1 visit/year vs. 0 visits, p=0.05), there was no statistical difference between groups in terms of frequency of hospitalization for pulmonary exacerbations.

Conclusions: This study demonstrates that patients who self-reported use of THC were more likely to have no-shows to CF clinic, a marker of suboptimal adherence. Furthermore, THC self-reporters were also more likely to suffer from mental health disease and use tobacco. These factors may have profound long-term sequelae. More research is needed to better understand the relationship between THC use and its adverse effects on adherence, mental health disease, healthcare utilization, and pulmonary and nutritional health. The CF Mental Health Team at Children's of Alabama (COA), consisting of two licensed social workers and one licensed professional counselor/mental health coordinator (LPC/MHC), collaborate to implement the mental health guidelines as recommended by the CF Foundation. The team shares the responsibility of screening patients and families and the LPC/MHC enters data into both a database kept by the CF team and into the CF Registry (CFR). Due to a recent change in program manager (PM), a question arose about when anxiety and/or depression were marked as a complication in the CFR.

Methods: LPC/MHC reached out via listserv to the CFR team to ask what criteria should be considered when marking anxiety and/or depression as a complication in the CFR. The registry team responded that these complications should be indicated when a clinical diagnosis of anxiety and/ or depression has been made. LPC/MHC next created a report in the CFR of all patients (n=50) who had the complication of anxiety and/or depression marked during 2017. Each encounter (clinic visit/hospitalization) was reviewed in the electronic medical record (EMR) to ascertain if there was an indication of anxiety and/or depression in the physician note. Mental health screening (PHQ-9 and GAD-7) scores as well as if the patient was on any psychotropic medication were cross-referenced.

Results: Of the 50 patients that had anxiety and/or depression marked as a complication submitted to the 2017 CFR, 17 (34%) did not have any documentation in EMR to support the diagnosis, were not on any psychotropic medications throughout the calendar year, and/or had no elevated scores on mental health screenings. Six percent of patients were under the age of 12 and had not ever completed a screening. Significant discrepancies between inpatient and outpatient providers' documentation of anxiety and/ or depression were noted. In addition, when mental health screenings for 2017 were reviewed, there were 13 patients with moderate-severe depression and/or anxiety scores that were not indicated in the CFR as having these complications.

Discussion: Our results indicate inconsistencies in our documentation of the field "complication of anxiety and/or depression" in the CFR. Some of the complications were marked in error while others had simply not been removed when the complication was resolved. This brief study has sparked corrective actions designed by the CF Mental Health Team, CF Center Director, and new PM as to how we can be more consistent and accurate in what information is reported. As a result, physicians were notified about their patients that did not meet criteria for diagnosis and these complications are in the process of being removed from the registry. Also, physicians will be asked to include an indication of anxiety and/or depression in their note in the EMR. LPC/MHC will have quarterly meetings with PM to review discrepancies between EMR, mental health screening scores, and registry as we continue to work toward greater accuracy and consistency.

Schleich, E.; Green, D.M. Pediatric Pulmonary, Johns Hopkins All Children's Hospital, St. Petersburg, FL, USA Background: People with cystic fibrosis (CF) are at risk of developing depression and anxiety at a rate of two to three times that of the community. Screening guidelines for depression and anxiety in CF were published by the CF Foundation in 2015. Since July 2016, the Johns Hopkins All Children's Hospital (JHACH) CF clinic has employed a part-time Mental Health Coordinator (MHC) to implement these guidelines and provide screening to patients over the age of 12. In this study, we wanted to assess the effect visits with the MHC has on patient scores of depression and anxiety.

Methods: A retrospective chart review of pediatric CF patients followed at JHACH was provided. Patients were included if they were >age 12 and had been provided at least two screenings. The Patient Health Questionnaire-Adolescent (PHQ-A) and the Generalized Anxiety Disorder 7 (GAD-7) were provided as the screening tools. PHQ-A has been validated as a measure of depression and GAD-7 has been validated as a marker of anxiety within the CF population. Patients self-administered the screenings after brief education about depression and anxiety, symptoms, and reason for screening. Depression and anxiety were defined as a score of 5 or greater on the PHQ-A or the GAD-7. Linear regression was applied to determine any relationship of the number of visits with the MHC in change in scores. The regression was also adjusted for age and gender.

Results: 58 patients were included in the study and mean age was 15.4 years and 38% were male. PHQ-A identified 14 patients in the mildly depressed range, 3 in the moderately depressed range and 1 in the severely depressed range. The average change in PHQ-A for these patients was -1.2 (range: -12 to 19) with 7 falling below the range for depression. Based on the GAD-7, 11 patients scored in the mildly anxious range, 8 in the moderately anxious range and 3 in the severely anxious range. The average change in GAD-7 for these patients was -2.1 (range: -12 to 8) with 7 falling below the range for anxiety. For the entire group, 23 had a decline in PHQ-A score, 18 stayed the same and 17 had an increase in scores. On GAD-7, 23 had a decline, 16 stayed the same and 19 had an increase in scores. Depression and anxiety were both observed in 12 patients. The average number of visits with the MHC was 6.2 but for those with depression it was 9.8 and for those with anxiety was 9.9. Linear regression showed that there was an association in PHQ-A score decline with number of clinic visits with the MHC (p=0.047) but not for . For those with depression, there was a trend towards improvement with the number of visits with MHC (p=0.097) however there was no difference for those with anxiety (p=0.211). Multivariate analysis including sex and age did not affect the results.

Conclusions: Our data suggest that visits with an MHC can improve scores on the PHQ-A. Reasons for this improvement include increased trust with the MHC, more forthcoming patients, screening without parents present, and improved knowledge of depression and anxiety with increased exposure to the MHC. As depression and anxiety are identified risk factors for poor outcomes in the CF population, continued access to MHC could be of extreme benefit to a CF center. The CF Medication Belief Questionnaire (CF-MBQ) is a validated tool that measures health beliefs related to dornase alfa, hypertonic saline, inhaled antibiotics, and azithromycin (Eakin MN, et al. J Cyst Fibros. 2017; 16:637-44) . As more effective, targeted therapies are developed, persons with CF may question the need to continue other chronic medications. We hypothesized that use of a CFTR modulator would be associated with significantly lower scores on CF-MBQ Self-Efficacy, Importance, and Motivation scales and significantly higher scores on the Decisional Balance "reasons to miss" medications subscale.

Methods: Results are from a subanalysis of the BARRIERS study, a multicenter, cross-sectional, observational study of persons with CF ages 13 years and older, conducted at 13 pediatric and 9 adult CF care centers in the Success with Therapies Research Consortium. Data collection included demographics, medications, comorbidities, and patient-reported measures (e.g., CF-MBQ). CF-MBQ subscales were compared between subjects prescribed a CFTR modulator to those that were not, using Student's t-tests with unequal variances.

Results: Of 408 patients, 189 (46.2%) were prescribed a CFTR modulator with 141 (74.6%) prescribed lumacaftor/ivacaftor. Among the 408 patients, 344 (71.7%) completed the CF-MBQ. The mean age was 24.8±11.2 years, 205 (50.3%) were female, and characteristics were generally similar between groups. However, FEV1 percent predicted was higher among persons prescribed a CFTR modulator (78.7±21.5) versus not (70.9±25.1, p=0.001). Contrary to prediction, no significant group differences were found between CF-MBQ subscale scores (see Table) .

Conclusions: Persons with CF do not perceive other chronic medications (i.e., dornase alfa, hypertonic saline, inhaled antibiotics, and azithromycin), as less important when prescribed a CFTR modulator. This may be attributed to perceived additive nature of combined therapy or continued value from symptom relief with previous therapies. Future analyses will include comparison by specific CFTR modulator and CFTR modulator eligibility (based on genotype) and prescribed use. MI is a collaborative, nonjudgmental, patient-directed and goal-oriented approach that is especially effective with patients who are ambivalent or resistant to changing their behaviors. Instead of relying on traditional information delivery models, MI focuses on eliciting the patient to discuss his or her own motivations for change and ability to make such changes. MI has been successfully used to improve patient treatment adherence and positive health behaviors with a wide range of illness populations, including cystic fibrosis (CF). MI is increasingly familiar to healthcare providers, but many find it difficult to obtain training and sustain practice of MI skills due to time limitations and lack of ongoing support.

Methods: Over the course of one month, four brief MI training sessions (~15-20 minutes) were delivered as part of weekly pulmonology division meetings with healthcare providers across multiple disciplines, including, physicians, fellows, nurses, social workers, and respiratory therapists. A clinical psychology post-doctoral fellow with specialized training in MI conducted the sessions and offered ongoing support to providers with implementing the skills in clinic. The trainings focused on the "spirit of MI," evoking patient change talk, core MI skills (open-ended questions, affirmations, reflections, and summary statements), and effective information exchange techniques (ask-provide-ask and elicit-provide-elicit). A brief one-page handout summarizing MI skills covered in the sessions was provided at the end of the training series. Pre-and post-surveys were collected to assess provider beliefs and behaviors pertaining to patient communication and satisfaction with the brief trainings.

Results: Overall, CF healthcare providers reacted very positively to the brief MI training sessions, feeling it was very useful. All participants indicated the training made them consider ways to improve communication with patients. The majority of participants indicated they liked having multiple, brief training sessions over the course of one month instead of having a single longer training. Approximately 40% of participants expressed interest in participating in additional follow-up training sessions to continue practicing MI skills. Two subgroups (physician fellows/ residents and registered dietitians) specifically invited the speaker for additional training sessions. Strategies participants indicated they intended to use with patients were: active listening, eliciting more patient input, doublesided reflections, and open-ended questions.

Conclusions: Offering brief MI training sessions during existing pulmonology division meetings is a convenient way to teach MI skills and encourage their continued use. Healthcare providers benefitted from the MI training sessions and indicated they would use MI strategies to improve communication and facilitate patient behavior change. Future plans include offering advanced training modules and additional opportunities for practicing skills with live coaching and feedback.

were performed between PHQ-9/GAD-7 scores and CFQ-R's 13 domains scores.

Results: Preliminary data analysis has shown that a high percentage of patients reported scores above the cut-off level of 5 both for depression symptoms (Dep vs No Dep: N=35; 43%) and anxiety symptoms (Anx vs No Anx: N=36; 43%). For Depression, of the 13 quality of life domains, the correlation test shows that the Vitality, Emotional, and Weight domains are the most anti-correlated ones (cor =-0.65 p= 1.24e-09; cor = -0.57 p= 2.1e-07; cor=-0.42 p=3.4e-05). In the case of Anxiety, the most anti-correlated are the Emotional, Vitality, and Digestion domains (cor=-0.56 p= 3.8e-07; cor=-0.51 p=4.0e-06; cor=-0.32 p= 0.001). When we considered the two subsets: Anx/Dep vs No Anx/Dep, the patients with clinical Anxiety show an anti-correlation with the Emotional domain (cor=-0.47 p=0.002), while the patients with clinical Depression show a significant anti-correlation with Treatment Burden, Emotional, and Health quality of life domains (cor= -0.40 p= 0.008; cor= -0.47 p= 0.002; cor= -0.36 p= 0.015).

Conclusion: Depression and anxiety scores allied with specific Health-Related Quality of Life domains, but the Dep vs No Dep analysis shows that depression is mostly associated with several Health-Related Quality of Life domains compared to Anxiety patients.

Snell, C. 1, 2 ; Fuchs, C. 3, 4 ; Ashley, A. 5 1. Psychiatry, Boston Children's Hospital, Boston, MA, USA; 2. Harvard Medical School, Boston, MA, USA; 3. Boston Medical Center, Boston, MA, USA; 4. Boston University School of Medicine, Boston, MA, USA; 5. Pulmonary, Boston Children's Hospital, Boston, MA, USA Background: In 2016, with the support of the CF Foundation MHC grant, a robust mental health program was fully implemented within our large pediatric and adult CF center. The center assembled a mental health care team including a pediatric psychologist and an adult psychologist who as part of their roles conducted trainings for CF social workers focused on brief, evidence-based mental health interventions for use in a clinic setting. The CF social workers were the team members primarily responsible for mental health screening, clinic-based mental health evaluation, and providing psychoeducation and brief mental health interventions when possible in clinic. This is an analysis of the trainings provided and their utility as demonstrated by use within the outpatient CF clinic in the three months after trainings concluded.

Methods: From July 2016 through August 2017, the adult and pediatric psychologists conducted 17 trainings on brief interventions for use in a medical clinic setting. Trainings were 30 minutes in length and occurred at the beginning of each biweekly CF mental health team meeting. The trainings focused on CF-specific issues including pain management, coping with medical trauma and procedural anxiety, body image, mindfulness, and complementary medicine. The trainings also focused on using cognitive-behavioral therapy (CBT), acceptance and commitment therapy (ACT) and other evidence-based treatment modalities with the CF population. The team's social workers were surveyed about what topics they were interested in, and surveyed following each presentation to give feedback and inform future training topics. To assess the practical application of these trainings, we will perform retrospective chart review from September 2017-November 2017 on all patients with CF >12 years old who had a mental health assessment by a social worker in the ambulatory clinic. Social workers were asked at the time of each screening to chart which interventions they used with each patient in clinic.

Results: Surveys administered to the social work staff (N= 6) after the presentations indicate that the presentations were perceived to be relevant, informative, and likely to be applied to their clinical work. The most highly rated topics in terms of informativeness, relevance, and applicability to clinical work were medical trauma (mean rating out of 5= 5) and ADHD (mean= 5). The lowest rated topics were body image (mean= 4.4), procedural anxiety (mean= 4.6), and values clarification (mean= 4.6).

Conclusion: Preliminary results suggest that the series of psychologist-led skills trainings offered as part of the MHC grant were well-received by the CF social workers. Additional planned analyses will assess the extent to which the trainings were applied in practice and improved the provision of mental health care for our adolescents and adults with CF.

Dell, M.L. 1 ; Nemastil, C.J. 2 1. Psychiatry, The Ohio State University, Dublin, OH, USA; 2. Nationwide Children's Hospital, Columbus, OH, USA Background: Dexmedetomidine is an α-adrenergic agonist approved by the FDA and used for sedation of CF patients on mechanical ventilation or undergoing surgical procedures. It has sedative, analgesic, and anxiolytic properties useful in severe anxiety, agitation, aggression, and delirium. However, little is reported regarding nonprocedural use of dexmedetomidine in CF patients in intensive care settings. (Gee SW et al. J Pediatr Pharmacol Ther. 2015; 20:329-34; Nelson S, et al. BioMed Res Int. 2015; 2015:635737; Popat K, et al. Adv Anesthesia. 2006; 21:177-92.) Objective: The aim of this study was to review procedural and nonprocedural dexmedetomidine use in CF patients with comorbid psychiatric diagnoses.

Methods: After IRB approval, CF patients seen by psychiatry between Jan 2013 and June 2018 were identified by retrospective chart review. CF patients included in this study received dexmedetomidine and had more or more psychiatric diagnoses. Factors considered included sex, age, psychiatric diagnosis, substance use, family psychiatric history, psychotropic medications, transplant status, and treatment adherence.

Results: A total of 39 subjects, 16 males and 23 females, met inclusion criteria. The mean age of dexmedetomidine first administration was 21.4 ± 8.8 years. Sixteen of the 39 subjects received multiple administrations during the 5½-year study period. Twenty-five of the 39 subjects received the medication for procedures only. Psychiatric diagnoses included depression (16), anxiety (18), panic disorder (11), obsessive compulsive disorder (5) , psychosis (4), cannabis abuse (4), alcohol abuse (3), ADHD (2), bipolar disorder (1), intellectual disability (1) . Two of these 25 had received lung transplants. Fourteen of the 39 subjects received dexmedetomidine for procedures and during ICU stays. These included 6 females and 8 males, mean age 25.6 ± 9.4 years. All 14 had delirium and/or were not able to come off ventilator due to severe agitation/anxiety. Psychiatric diagnoses included anxiety (14), depression (14), panic disorder (10), cannabis use (4), alcohol use disorder (3), benzodiazepine abuse (3), intellectual disability (1), ADHD (1). These patients received an average of 7.3 psychotropic medications (range 2 -14), and average 5.1 analgesic medications (range 1-10). Seven of 14 were post-lung transplant and 8 of 14 died during the hospitalization they received dexmedetomidine for nonprocedural reasons. Family histories of individuals requiring nonprocedural dexmedetomidine had higher rates of mood, anxiety, and substance use disorders than for the procedural only group. Nonadherence to treatments was common in both procedural and nonprocedural dexmedetomidine groups.

Conclusion: While commonly used for procedural sedation, dexmedetomidine is also a helpful agent for agitated delirium and when extreme anxiety contributes to difficulty weaning of ventilators or other higher levels of respiratory support. Identifying psychiatric and medical factors associated with benefit and optimally timed initiation of nonprocedural dexmedetomidine may lead to decreased time on ventilators and in ICU levels of care.

Quittner, A.L. 1 ; Eckmann, T. 2 ; Riekert, K.A. 2 1. Miami Children's Research Institute, Nicklaus Children's Hospital, Miami, FL, USA; 2. Medicine, Johns Hopkins, Baltimore, MD, USA Objectives: Adherence to the CF regimen across adolescents and adults is quite poor and one key to improving adherence may be to identify an individual's barriers to each component of treatment. Items were developed with all major stakeholders, including patients, parents, healthcare intervention, and referral to treatment for EAU in CF clinics is warranted to potentially prevent health consequences of EAU. Acknowledgment: Supported by K23AA022126. Introduction: Rates of depression in cystic fibrosis (CF) are elevated, but the longitudinal clinical course and rates of recovery and recurrence have not been demonstrated. The purpose of this study was to describe rates and patterns of depression in individuals with CF over time. We present findings of a 5-year prospective, longitudinal study.

Methods: Our CF center designed and implemented a standardized process for ongoing depression screening in individuals with CF age ≥ 12, providing interventions tailored to the severity of depression and for those who screen positive for suicide risk. We detailed the rates and course of depression and treatment approaches throughout the 5-year period (2013 to 2017) .

Results: During the five-year period, there were 33 adolescents in the pediatric program and 117 adults in the adult program. The average one-year incidence rate of clinically significant depressive symptoms was 20% (18% in 2013, 19% in 2014, 18% in 2015, 26% in 2016, and 20% in 2017) . The five-year prevalence of depression (2013-2017) was 29% and highest among individuals aged 18-25 (19% of individuals aged 12-17; 39% of those 18-25; 32% of those 26-49; and 18% of those 50 and older) and among females (33%) compared to males (26%). The 5-year prevalence of suicidal ideation was 12%. Difficulties with energy, sleep, and feeling sad, depressed, or hopeless were the most common symptoms reported by individuals with CF and depression. With stepped care treatment 38% of individuals with depression had a clinical recovery (PHQ-9 score< 5) at 3 months, 14% at 6 months, 11% at 9 months and 3% by one year (10% are still actively receiving stepped care treatment). We found the course of depression was chronic in the remaining 24% of our population who experienced persistent (symptoms for ≥ two years in adults and ≥ 1 year adolescents), unremitting depression. The likelihood of depression recurrence was only 8% over the 5-year study time.

Conclusions: In patients with CF the annual incidence and 5-year prevalence rates of depression are higher than in the general population. However, with stepped care treatment we achieved high rates of recovery and lower rates of recurrence than reported in the general population. These findings provide support that screening and stepped-care treatment strategies tailored to the severity of symptoms with the goal of recovery are feasible in a CF center and reduce depressive symptoms and the risk of recurrence in individuals with CF.

Introduction: Pain in individuals with CF negatively affects health outcomes including quality of life and mortality, yet little is known about CF care provider attitudes and beliefs towards pain management in this population.

Objective: To compare current CF care provider approaches and beliefs surrounding pain management to patient reported experiences with and attitudes towards pain management.

Methods: An online provider survey was developed to assess CF provider perceptions and approaches to pain in CF care centers. It was distributed nationally via CF Foundation (CFF) provided listservs across disciplines. A similar patient survey was developed for individuals with CF to assess how pain is currently managed and their preferences regarding pain management. It was distributed nationally via the CFF Community Voice listserv, social media sites, and CF center patient listservs.

Data: The survey was completed by 242 providers [physicians (15%), advanced practice providers (23%), nurses (15%), social workers (14%), other disciplines (33%)], 63% female; 396 completed the patient survey, mean age: 33.1 years (range: 14-79), 76% female. When asked to estimate the percentage of their patient population who experience pain, 66% of providers responded that less than half of their patients experience pain, and 30% responded that pain should not be considered a symptom of CF. Comparatively, 82% of patients responded that they experience pain related to their CF disease, 34% reported experiencing pain daily, and 33% responded that pain affects their daily life. The majority of providers and patients, 64% and 66% respectively, believe that pain should be managed by the CF care team, with 75% of patients reporting discussing pain with their CF provider. 34% of patient respondents reported having an active pain management plan which was managed by their CF care team in 65% of cases. Of those with a pain management plan, 72% reported being satisfied to very satisfied. The majority of providers reported being somewhat to strongly comfortable in assessing pain (79%), making pain diagnoses (70%), and in prescribing pain interventions (64%). 54% of providers are concerned that opioid pain management will make patients ineligible for transplant and 92% of providers somewhat to strongly agree that clinical guidelines should be developed for treatment of CF-related pain.

Conclusions: Providers vastly underestimated the percent of patients who experience CF-related pain and 30% did not believe pain is a symptom of CF, however, the majority of both CF care providers and patients agreed that pain should be managed by CF care teams. Despite reports of providers being comfortable providing pain management, our findings suggest that additional education and CF-specific pain management guidelines are needed.

Polineni, D. 1 ; Durkin, K.M. 2 ; Ruvalcaba, E. 3 ; Bord, E. 4 ; Duncan, C.L. 2 1. Univ. of Kansas Medical Center, Kansas City, KS, USA; 2. Department of Psychology, West Virginia Univ., Morgantown, WV, USA; 3. Johns Hopkins Univ., Baltimore, MD, USA; 4. Boston Children's Hospital, Boston, MA, USA Introduction/Aim: Patient adherence to cystic fibrosis (CF) treatment regimens is similar to that in other chronic diseases, with estimated adherence rates ranging from 30-70%. Particularly with recent FDA approval of CFTR modulators, enhancing adherence to therapies is expected to improve both lifespan and quality of life for individuals with CF. We report on the development and validation of a new measure (CF-Care Behavior Survey; CF-CBS) to assess patient nonadherence to the most common components of CF care.

The CF-CBS is a self-administered survey composed of two sections. The first section includes 4 major questions, pertaining to estimated adherence to CF therapies, barriers to adherence, and motivation for improving adherence patterns (importance of/confidence in doing so). These same questions are asked separately for each of 12 different treatment components (e.g., airway clearance, inhaled antibiotics, dietary supplements). The second section involves questions regarding the patient's approach to volitional or purposeful changes to his or her overall CF care regimen. Content validation of the CF-CBS is being conducted as one-key component of a broader multisite, pilot and feasibility trial to test telecoaching as an adherence promotion intervention. The validation sample will consist of about 40 patients (ages 14-25) and approximately 20-25 parents of adolescent patients. Patients and parents individually complete the CF-CBS at study enrollment, followed by an individualized, cognitive debriefing interview utilizing a standard interview guide with prompts. All cognitive debriefing interviews are audio-recorded and transcribed verbatim by a paid third-party service. Cognitive debriefing interview transcripts are divided equally among three coders, with 30% of interviews independently coded by two coders to assess reliability. Transcripts will be coded for consistency in participant interpretation of survey items and need for clarification in item content. Discrepancies will be discussed and resolved with the study principal investigators. The CF-CBS measure will be revised based on aggregated results and feedback.

Results: To date, 27 patients and 11 caregivers have completed cognitive debriefing interviews and coding is underway. Recruitment and data collection are expected to be completed by July 2018. Patient and parent feedback will be summarized and the revised measure will be displayed as part of the presentation.

Conclusion: The CF Care Behavior Survey (CF-CBS) is anticipated to be a unique and useful tool for measuring adherence to CF care, including volitional and inadvertent nonadherence. Because adherence often varies as a function of type of treatment, this new measure has significant potential for measuring variation in adherence across the different CF regimen components. Further, it is expected to be valuable to use during targeted interventions (e.g., tele-coaching) and in measuring outcomes for interventions promoting adherence more broadly.

Acknowledgment: Grant Support: Cystic Fibrosis Foundation Therapeutics/Success with Therapies Research Consortium.

Polineni, D. 1 ; Durkin, K.M. 2 ; Ruvalcaba, E. 3 ; Chen, G.J. 1 ; Muther, E.F. 4 ; Prickett, M. 5 ; Saavedra, M.T. 6 ; Duncan, C.L. 2 1. Pulmonary and Critical Care Medicine, University of Kansas, Kansas City, KS, USA; 2. Psychology Department, West Virginia University, Morgantown, WV, USA; 3. Johns Hopkins University, Baltimore, MD, USA; 4. Children's Hospital of Colorado, Aurora, CO, USA; 5. Northwestern University, Chicago, IL, USA; 6. National Jewish Health, Denver, CO, USA Introduction: The flexibility of tele-coaching affords the opportunity to take a practical and patient-centered approach to identify individualized adherence concerns and address them with tailored, efficacious interventions. The primary objective of this presentation is to report upon key stakeholder (i.e., patient, parent, provider) feedback regarding the initial feasibility and logistics of our tele-coaching intervention to enhance adherence in adolescents and young adults with CF.

Methods: This multicenter project involves a qualitative assessment to obtain stakeholder input, via focus groups, regarding feasibility, structure, and logistics of the individualized tele-coaching intervention. For this project, 40 participants with CF (CF Patient Cohort; age 14-25 years old) and 15-25 multidisciplinary CF care providers (Provider Cohort) will take part in focus groups to inform the development of our tele-coaching intervention. All focus groups are conducted via web-based tele-conferences, audio-recorded, and transcribed verbatim through a paid service. Transcripts are coded by independent coders using qualitative data analysis software (e.g., ATLAS.ti) to identify and analyze domain themes and categories across transcripts. Reliability of coding will be calculated via Kappa statistic.

Results: Thus far, rates of participation by eligible families have been high (65%). To date, a total of 16 patients and 20 providers have participated in focus group sessions, and retention rates from enrollment to focus group participation are 78% in teen patients and 53% in adult patients. This presentation will include a review of findings from our full sample on acceptability of study methods, reaction to intervention plan, and feedback on structure and logistics for intervention delivery. Preliminary findings highlight receptiveness of patients and providers to video-call based coaching, with enthusiasm for enhancing continuity of care between clinic visits and improving provider access for long-distance care/communication. Access to high quality internet connection has been listed as a concern.

Conclusion: Findings are expected to provide guidance in tailoring our tele-coaching intervention to stakeholder preferences and recommendations, in an effort to enhance its feasibility and acceptability. In future directions, our intervention will be validated via a second series of focus groups including participants of the Patient Cohort, as well as a group of 20-25 parents (i.e., Parent Cohort) who live with participants age 14-18 in the Patient Cohort. Our research design will inform study procedures for a subsequent project executing a randomized controlled trial to test the efficacy of this stakeholder informed tele-coaching intervention. Acknowledgment: Grant Support by the Cystic Fibrosis Foundation Therapeutics/Success with Therapies Research Consortium.

Walsh, K.; Hunt, S.; Ren, C.L. Riley Hospital for Children, Indianapolis, IN, USA Background: Previous studies have shown that there is a high prevalence of anxiety and depression in caregivers of children with chronic disease, and that this significantly impacts their children's mental health and long-term outcomes. Research has also shown the importance of breaking bad news to caregivers, and how this alone can impact caregiver's perceptions of their child's diagnosis, quality of life and life expectancy. There are limited data on the mental health of parents of infants diagnosed with CF through newborn screening (NBS). The goals of this study are to determine the prevalence of mental health disorders (MHDs) in parents at our CF center and identify factors associated with the presence and severity of these mental health diagnoses.

Methods: We screened parents of CF patients ≤2 years for anxiety and depression from February 2017 to May 2018, using the GAD-7 and PHQ-9 surveys respectively. We collected demographic information and conducted chart reviews to identify clinical features associated with the presence of parental MHDs.

Results: A total of 35 parents were approached (24 mothers, 11 fathers) and 29 (83%) agreed to participate. Screening is ongoing and updated results will be presented at the conference. The PHQ-9 identified 9 parents (31%) with mild (n=5), moderate (2) or moderately severe (2) levels of depression, while the GAD-7 identified 11 parents (38%) with mild (n=6), moderate (4) or severe (1) levels of anxiety. 44% of parents of children with a history of meconium ileus (MI), 38% of female parents, and 21% parents of children with a history of ≥2 hospitalizations screened positive for depression. Additionally, 43% of female parents and 21% of parents of children with a history of ≥2 hospitalizations screened positive for anxiety. There was no association found between marital status, number of children with CF, socioeconomic status, and insurance status with either depression or anxiety.

Conclusions: The proportion of parents of CF NBS infants who screen positive for depression or anxiety is similar to that reported in other parents of children with chronic disease. This study aimed to identify clinical factors that may contribute to parental susceptibility to higher rates of depression and anxiety. Female parents, parents of infants with a history of MI and parents of infants with a history of multiple hospitalizations are more likely to have features of depression and/or anxiety, and CF caregivers should be more vigilant in screening for MHD in families with these characteristics. Educating both care providers as well as parents on the risk factors associated with mental health issues will be a vital part of improving their access extremely impaired (33% each). Of those with moderate anxiety, 60% felt somewhat impaired, 32% felt very impaired, and 4% felt extreme impairment. Finally, of patients with severe anxiety, 6% reported no impairment, 33% reported somewhat, 39% reported feeling very impaired, and 22% felt extremely impaired.

Data analysis is ongoing to evaluate relationship to objective physical parameters including BMI, FEV1, and changes from known FEV1 baseline.

Conclusions: Symptom severity based on GAD-7 and PHQ-9 scores are not reliable predictors of self-reported functional impairment. Further studies are needed to investigate the utility of general screening tools like GAD-7 and PHQ-9 for the CF population.

Saavedra, M.T. 1 ; Poch, K. 1 ; Lindwall, J. 3 ; Muther, E.F. 3 ; Eckmann, T. 2 ; Psoter, K. 2 ; Quittner, A.L. 4 ; Riekert, K.A. 2 1. National Jewish Health, Denver, CO, USA; 2. Johns Hopkins, Baltimore, MD, USA; 3. Children's Hospital Colorado, Aurora, CO, USA; 4. Nicklaus Children's Research Institute, Miami, FL, USA Introduction: High treatment complexity in cystic fibrosis (CF) presents a challenge for disease self-management and treatment adherence. Adherence rates decline when comorbid conditions such as anxiety and depression are present. In the current study, we evaluated whether higher levels of anxiety and depression modified perceived treatment burden when compared to actual treatment complexity.

Methods: Data was obtained from the recently completed multicenter Barriers study, a cross-sectional observational study performed at 22 pediatric and adult sites participating in the Cystic Fibrosis Foundation Success with Therapies Research Consortium. Adolescents and adults completed the CFQ-R, PHQ-8 and GAD-7. Treatment Burden scale is a domain of the CFQ-R. The Treatment Complexity Scores (TCS) were derived from chart review. For all analyses, variables were treated as continuous. Models were conducted separately for anxiety and depression. Unadjusted and multivariable linear regression evaluated the main effects of treatment complexity, anxiety/depression and the interaction between them on Treatment Burden, controlling for age, gender, income, insurance status, lung function, number of IV courses in the prior year, and presence of CF-related diabetes.

Results: Prevalence of anxiety in this sample was distributed as follows--minimal (63%), mild (22%), moderate (9%), and severe (2%). Depression prevalence was similar-minimal (61%), mild (25%), moderate (10%), moderately severe (2%) and severe (2%). GAD7 scores were not associated with treatment complexity score or CFQ-R Treatment Burden in unadjusted and adjusted models. PHQ8 scores were associated with CFQ-R Treatment Burden (rho=-0.12, p=0.03) in the unadjusted model; however, this association did not persist following inclusion of covariates (p=0.17). There was no significant interaction between psychological symptoms and TCS on CFQ-R Treatment Burden.

Conclusion: A large percentage of subjects with CF manifest symptoms of depression and anxiety. Approximately 33% of this sample was anxious above the clinical cut-off, and 39.4% were depressed. Depression, but not anxiety, was related to greater Treatment Burden. While controlling for multiple demographic and medical variables in this sample removed the association between treatment complexity and psychological symptoms, the association is an important consideration. Further studies are needed in CF adolescents and adults in order to identify at which point depression starts to impact treatment adherence and how the CF multidisciplinary care team model can better identify vulnerable subjects prior to the point at which behavioral health becomes a barrier to CF treatment adherence.

Acknowledgments: This study was funded by CF Foundation grants BARRIERS16PE0, SAWICK14PE1 & RIEKERT15PE0.

Cassidy, J.; Nay, L.; Nelson, B.; Carlin, K.; McNamara, S. Seattle Children's Hospital, Seattle, WA, USA Background: Children with cystic fibrosis (CF) are exposed to painful and anxiety-producing procedures during routine outpatient care, such as quarterly deep oropharyngeal (OP) cultures. Child life specialists (CLS) are trained to recognize developmental concerns in the context of health care and to intervene using developmentally appropriate play, rehearsing coping skills and offering psychological preparation to ameliorate anxiety and distress. With the addition of a CLS at one CF clinic per week, we designed and implemented a pilot study to evaluate the effect of child life intervention on behavioral distress and self-reported pain for children with CF undergoing clinical OP culture. We hypothesized that repeated exposure to CLS intervention would decrease the child's observed behavioral distress and self-reported pain and decrease parental anxiety relative to the procedure over time.

Methods: Children with CF ages 4-9 years who attended CF clinic approximately quarterly with OP cultures as standard care were enrolled. Patients were assigned to the intervention group (individualized CLS sessions at each study visit) or to the control group (no CLS) based on day of clinic attendance and CLS presence. At each study visit, a trained observer rated the child's behavioral distress during the OP swab procedure using the Procedure Behavior Checklist (PBCL). Immediately following the OP swab, the FACES scale was administered to measure the child's experience of pain. The State-Trait Anxiety Inventory (STAI) was used to measure parental anxiety. Descriptive statistics, plots, and linear mixed-effects models were used to examine if the CLS intervention resulted in significant decreases in child distress, pain, or parental anxiety over time.

Results: Thirty-four patients were consented with 28 (N=17 intervention group, N=11 control group) completing 3-4 study visits over a 15-month period. Participants were 54% female; mean age at enrollment was 6.03 years (SD 1.64). No statistical significance was found between the intervention and control groups for behavioral distress scores (PBCL), self-reported pain following procedure scores (FACES), or parental anxiety scores (STAI).

Discussion: Our results are biased to the null because majority of participants (93%) had previous CLS intervention, with 79% having CLS interaction in the year prior to enrollment. The longitudinal nature of the study resulted in variability in the intervention group; 10 of 17 participants had an intervening CF clinic visit without CLS present. A mixed intervention group was created for analysis to differentiate these participants from the pure intervention group, resulting in small group sample sizes for analysis. Length of time between study visits ranged from 42-287 days. Other potential contributing factors include multiple nurses performing the OP swabs, turnover in PBCL scorers, and other procedures/confounding events occurring at a clinic visit. Subjectively, parents endorse CLS services in clinic, although these opinions were not measured. Larger studies with CLS naïve patients and more control over potentially confounding factors are needed to assess the effect of CLS intervention in outpatient clinics. Introduction: Cystic fibrosis (CF) management has been largely based on an integrated, collaborative team approach with airway clearance, antibiotic regimens for respiratory infections, inflammatory modulation and nutritional management (Zemanick ET, et al. J Cyst Fibros. 2010; 9:1-16) . Cystic fibrosis transmembrane conductance regulator (CFTR) modulators such as ivacaftor (Kalydeco), ivacaftor/lumacaftor (Orkambi), and more recently tezacaftor/ivacaftor (Symdeko) have changed the way in which CF management has been approached, with more targeted treatments at the protein level based on CF mutations. The decision to start one of these medications is multifactorial and involves many stakeholders. Caregivers and patients may consider side effects, including possible unknown long term effects, expectation of clinical benefit, prior experience with similar medicines, social media or other influences (Singer E, et al. Health Expect. 2011; 17:4-14) . Patient choice on medications may also be influenced by how the information is presented by the CF provider (Fraenkel L, et al. Med Decis Making. 2016; 37:230-8) . In light of the changing landscape of CF medical management, this study aims to better understand the decision making process including the influence of social media on individuals and caregivers of those with CF who are eligible for CFTR modulators, as well as to discern if the way providers present the information is consistent.

Methods: A cross-sectional survey of pediatric and adult patients with CF at two CF centers in Buffalo, NY and Boston, MA eligible for ivacaftor/lumacaftor or tezacaftor/ivacaftor is being conducted. Eligibility is defined as individuals 6 years of age or older with two copies of the F508del mutation, or for tezacaftor/ivacaftor additional eligible patients with one copy of F508del and another specified mutation, who were clinically eligible for either of the medications. A 24-question survey will be administered electronically during clinic visits. Most survey responses are "tick box" or rank-ordered questions, with open-ended responses to assess what information and factors led the individual/caretaker to a decision regarding the use of CFTR modulators. A separate survey for medical providers will inquire about the information they give to patients regarding starting CFTR modulators. Multiple response frequencies and the mean rank order of survey questions will be analyzed. Text responses will be analyzed thematically.

Results: Surveys will be collected and analyzed by October 1, 2018. The collection of approximately 50 surveys is planned. IRB submission is ongoing.

Conclusions: An understanding of the decision making process surrounding beginning CFTR modulators may help providers understand how to present information to patients regarding relevant therapies in an era of reliance on the internet and social media. Comparison of provider presentation of the information to patient/caregiver perception will help inform providers on ways to educate patients with CF regarding CFTR modulators. The Stanford Integrated Psychosocial Assessment for Transplant (SIPAT) is a tool used by some transplant programs to identify candidates who may display behaviors which may be harmful post-transplant. The SIPAT has been validated in organ transplant evaluation, and a score of <20 suggests an acceptable candidate, a score between 20-40 suggests a minimally acceptable candidate, and a SIPAT score >40 suggests a poor candidate. Use of SIPAT in transplant candidates with cystic fibrosis (CF) has not been previously evaluated.

Methods: Transplant recipients at Loyola University Medical Center from 2012-2017 who had undergone psychosocial assessment with SIPAT.

Results: There were 21 individuals with CF included in the cohort. Mean age 33.2 years, with 52% being female and 91% Caucasian. Body mass index at time of SIPAT was 20.4 kg/m 3 . Median SIPAT score was 7 (IQR 6-9). Substance abuse post-transplant was found in 3 patients (14%), and support failure was found in 3 patients (14%). Optimal adherence was noted in 13 (65%) post-transplant, with occasional nonadherence in 3 (15%) and harmful nonadherence in 4 (20%). Elevated SIPAT score pretransplant was not associated with poor outcomes following transplant, including prolonged hospitalization, readmissions, or acute rejection in the first year following transplant. Those who developed harmful nonadherence and substance abuse following lung transplant trended to have high SIPAT scores pre-transplant.

Conclusions: Use of SIPAT in pre-transplant candidates with CF may identify patients at risk for development of harmful nonadherence and substance abuse following lung transplant.

Snell, C. 1,2 ; Bond, J. 3 ; Helfand, L. 3 1. Psychiatry, Boston Children's Hospital, Boston, MA, USA; 2. Harvard Medical School, Boston, MA, USA; 3. Pulmonary, Boston Children's Hospital, Boston, MA, USA Background: Studies of adherence to daily treatments and of emotional wellbeing among youth with CF have consistently found that adolescence is a particularly challenging developmental period. As a result of these findings, there has been an emphasis within the CF community on developing mental health screenings and interventions for adolescents with CF. However, there are fewer widespread resources and programs available for younger children with CF and their parents.

Objective: Using brief, anonymous, clinic-based parent surveys, we explored common adherence challenges and barriers to adherence among children with CF ages 3 to 8 years. We also developed an interview assessment tool and parent handouts on common adherence challenges, as part of a clinic-based program for children who have recently turned three years old.

Methods: CF center social workers administered brief, five-question surveys to parents during their child's CF clinic visit. After the first three surveys given, parents were asked to provide feedback on the questions' clarity. For all adherence and resource-related questions, parents were able to indicate more than one response, and were encouraged to indicate the most challenging area. No identifying information was included in the surveys and responses were anonymous.

Results: Twenty-six parents of children with CF ages 3-8 years completed the survey. Parents reported that the most challenging parts of their child's medical routine were airway clearance (n= 14), getting enough calories (n= 13), and medical procedures (n=11). They reported that what makes it challenging for their child to do CF care is the child not liking the way the treatment feels (n= 10), the child having a high activity level/ being distractible (n= 9), and the child not wanting to appear different (n= 9). The most commonly endorsed barriers to adherence were finding time to do treatments (n= 15) and forgetting to do treatments (n= 6). In terms of resources, parents reported they were most interested in receiving written materials (n= 12), or attending a parent workshop (n= 10).

Conclusion: Parents of young children with CF report that three areas are the most challenging parts of their children's medical routines: airway clearance, getting enough calories, and medical procedures (e.g. blood draws). Parents were interested in learning more about how to manage these challenges through written materials and workshops. As future directions, we are developing an interview assessment tool for use with parents, and developing parent education handouts on each of the three challenging areas. We plan to implement the following as a standard of care for children who have recently turned 3 years old: assessment of adherence challenges, discussion of relevant handouts, and making psychology referrals if needed. (2):263-83). To improve education, a children's story was written about a child with CF experiencing hospitalization who learns about the cause of CF and the importance of self-care behaviors such as treatment adherence. The purpose of this pilot study was to evaluate the feasibility and usefulness of an educational book tailored to the developmental level of preschool and young school age children (ages 4-10 years) with CF. Endpoints were measured by children's knowledge gained (pre-/post-intervention test) and parent questionnaire to determine feasibility, usefulness, and perceived satisfaction. A pre-/post-test randomized controlled trial design was utilized with 40 children with CF between ages 4-10 years and their parent (child-parent dyad), admitted to the hospital or seen in the outpatient CF clinic. The test group received the CF book (Breathing Better: Peter's Trip to the Hospital) and the control group received a generic children's book (Going to the Hospital). The evidence-based knowledge pre-/post-test is an 11 multiple choice questionnaire developed by a nurse clinician, a Child Life Specialist, and a nurse researcher. Procedures consisted of study staff completing the pretest with the child, demonstrating reading of the book, and requesting parents to read the assigned book at least once per day for 3 days. Follow-up included completion of the post-test questionnaire with the child and parent. Analysis of descriptive statistics for demographics and parent questionnaire, chi square for group comparison, and t-test for pre-/post-test comparison were completed. A total of 40 children with CF (53% male; mean age 7.05 years; median 7 years) along with their parent dyad completed the study with 20 in each group. Demographic comparison between the groups were similar. Preliminary comparisons of pre-/post-test mean scores between the groups were non-significant. The level of difficulty per item for post-test intervention group were moderate (6 out of the 11 items were scored correct by 70% or less of the respondents). However, parents with the CF book reported their child enjoyed the book (100%), had improved understanding of the disease (90%), learned something he/she did not previously know about CF (child 90%; parent 45%), increased discussion about CF between child and parent (90%), and would recommend the book to other parents of children with CF (100%). Findings from this study have led to revision of the pre-/post-test questions. Additional enrollment will be requested to re-evaluate knowledge assessment. Although knowledge assessment was insignificant, the usefulness and feasibility of the book were positive. With diagnosis and coping education at the core of Nursing and Child Life roles, this CF-specific children's book met the study's secondary aim of being useful to child-parent dyads enrolled in the study.

Barr, E. 1 ; Hadjiliadis, D. 2 ; Mueller, R. 3 ; Valverde, K. 1 ; Dorgan, D. 2 ; Spielman, K. 3 1. Genetic Counseling, Arcadia Univ., Philadelphia, PA, USA; 2. Pulmonology, Univ. of Pennsylvania, Philadelphia, PA, USA; 3. Univ. of Pennsylvania, Philadelphia, PA, USA Introduction: Cystic fibrosis (CF) clinics have guidelines for disease management, but recommendations for timing and content of conversations about sexual and reproductive health (SRH) vary. As life expectancy for CF has increased, SRH has become increasingly relevant. Men with CF are almost always infertile due to congenital bilateral absence of the vas deferens (CBAVD), though assisted reproductive technology (ART) can enable biological fatherhood.

Purpose: This study aimed to evaluate knowledge, experiences and attitudes towards CBAVD, ART, and sexual health to provide insight for a larger scale study that may inform the development of guidelines to meet the SRH needs of CF males.

Methods: A piloted survey was electronically sent to all males with CF (ages 18+) cared for by Penn Medicine Adult CF Program. The survey consisted of 33 close-ended questions and 4 supplementary open-ended responses. Close-ended responses were analyzed with descriptive statistics and chi-square tests; open-ended responses were coded thematically by the first author.

Results: Fifty-three participants responded, a response rate of 35.3%. Participants had a median age of 37, ranging 18-61. All subjects reported knowing CF impacts male fertility. Median age (± SD) subjects recalled first learning this was 16 (± 6.4 years). Twenty-nine (54.7%) learned from a healthcare provider (HCP) and 9 (16.9%) from parents. Average age of discussion for participants who learned from a HCP or parents was 19.4 and 14.8 years, respectively. Other modes included friends, written, and online materials. Eleven participants (20.8%) wished they learned sooner, majority (10) of whom learned after age 16 (p<0.001). Upon learning, 29 participants (54.7%) wanted to know if there were still options to have children, and 20 (37.7%) wanted to know if there were options to fix it. One subject expressed "it would be nice to have this as standard procedure to find out about fertility at clinic one or two times over teen and young adult years." Twenty-two (41.5%) have visited a fertility specialist, and 17 (31.7%) have had semen analysis.

Conclusion: The majority of adult men with CF are aware of the disease's impact on male fertility, but knowledge regarding ART options and uptake of services remains limited. Average age of participants who learned from parents was 5 years earlier than those who learned from a HCP. Emotional responses did not differ significantly based on who facilitated the discussion. Further, these data suggest a preference for learning about reproductive impact of CF and ART options by age 16, which should be further explored and confirmed. These findings provide insight to the range of emotions and questions that men experience when learning about CF-related male infertility. Upon learning this, many have questions regarding their ability to have children. As one man conveyed in an openended response, boys with CF should learn in "early adolescence, so that he has plenty of opportunities to ask pertinent questions as they arise or become relevant in his life. Background: Parents often struggle with how to motivate their children to complete treatments and to become more independent with treatments as they get older. Treatment adherence is extremely important for overall functioning and health, but can also be a significant source of conflict between parents and children with CF. Frequently, parents tire of "nagging" their children to do their treatments and children feel annoyed and may rebel against treatments. A promising approach to improving treatment adherence and parent-child conflict is the adapted use of motivational interviewing (MI). MI is a style of communication designed to facilitate behavior change. Health providers commonly use MI to improve treatment adherence in adolescents with chronic illnesses; however, this approach has not yet been studied as a skill for parents to use. We examined whether basic MI skills could be taught to parents of children with CF. MI skills were incorporated into education on developmentally appropriate parenting strategies for children and adolescents with CF. Methods: We piloted several different methods for educating parents on MI skills. We first offered a one-time in-person and web-streamed session on MI during a parent education night (N=6). A separate 3-session telehealth group was piloted with a small number of parents who expressed interest in developing MI skills (N=4). To participate in training sessions, the parent had to have a child with CF (aged 8-18 years) and have expressed concerns with treatment adherence. Additionally, we developed a one page introductory handout on MI to distribute to parents during clinic visits.

Results: There were many parents interested in learning MI skills; however, there were several barriers to parents' participation, largely related to scheduling. The original plan to run a 3-session telehealth group was revised due to these problems. Instead, we piloted offering 3 brief individual MI training sessions to parents either in-person or via phone. Session 1 focused on the "spirit of MI," session 2 focused on core MI skills (open-ended questions, affirmations, reflections, and summary statements), and session 3 addressed ways to apply MI skills to improve treatment adherence. To date, three parents have been involved in individual MI training. All parents have responded very positively to learning MI skills and have been able to demonstrate skills during practice exercises. Additional data collection is ongoing.

Conclusions: Brief training in MI skills for parents of children with CF could improve parenting strategies used to facilitate treatment adherence. Although we were unable to conduct these trainings via group-based telehealth due to scheduling problems, adapting training sessions to be done individually in-person and via phone increased feasibility. Additionally, Support Evaluation List, ISEL), health assessment (Memorial Symptom Assessment Scale, MSAS), treatment activity questionnaires (Tool for Adherence Behaviour Screening, TABS and other surveys), health-related quality of life measure (CFQ-R), and disease severity measures (forced expiratory volume in one second [FEV 1 ] and pulmonary exacerbations). Using data from three waves, separate linear regression analyses were conducted to assess predictors and outcomes of social support, controlling for weight percentile, diagnosis of diabetes, pancreatic sufficiency status, and colonization with Burkholderia cepacia, Staphylococcus aureus, and Pseudomonas aeruginosa.

Results: Females and those who were employed reported greater social support. In turn, greater social support was associated with fewer mental and physical health symptoms, digestive symptoms, and eating disturbances over time. Social support was also associated with elevated emotional, social, and role functioning as well as vitality and improved body image. Moreover, those with more support experienced less treatment burden and better overall perceptions of their health. Age, income, education, marital status, FEV 1, and exacerbations were not associated with improved social support. Social support was not related to physical functioning, problems gaining weight, respiratory symptoms, or treatment activity.

Conclusion: Social support has been shown to have a complex interaction with health in other chronic illnesses, and the same is true for adults with CF. This study fills gaps in the literature by examining predictors and outcomes of social support in this emerging adult population. The present study suggests particularly that unemployed adult males with CF report lower social support, perhaps identifying a subgroup for whom targeted interventions may be appropriate. Background: Peer support can be an important component of chronic illness management and social support, which can be difficult among people with cystic fibrosis (CF) and their families due to factors such as the treatment burden, infection protection and control guidelines, and day-to-day life demands. Support from peers who have experienced or are experiencing similar challenges can be especially meaningful during times of transition, such as adjusting to a new CF diagnosis, entering adulthood, making career decisions, or considering transplant. Members of the CF community have expressed a sense of isolation and desire for connection with peers. Virtual one-to-one and group connections can be a source of support to address unmet needs for peer connection among people with CF and their family members. We report on the use and satisfaction with this type of connection.

Methods: Two virtual programs were implemented to facilitate connections among people with CF and their family members. CF Peer Connect is a topic-based peer mentoring program enabling one-to-one connections through video, phone, email, or text conversations. Participants request mentorship on a particular topic and are matched with a screened and trained peer mentor for short-term mentoring in order to share experiences and learn from a peer who understands life with CF.

Virtual events (BreatheCon, CF FamilyCon and CF MiniCons) are online events that bring together groups of people with CF and their families in real time. These virtual events feature keynote panels, small-group video breakout sessions, mindfulness activities, and group chats for conversation on issues that are unique to people living with CF and their families. People with CF and their families set the agenda, and all small group video breakout sessions are facilitated by trained community members. Topics for virtual events and peer mentoring address common life transitions, including going to college, transition to adulthood, relationships, starting a family, going on disability, adjusting to a later CF diagnosis, and lung transplantation.

Results: Since 2016 CF Peer Connect has trained 156 peer mentors (ages 22-72) and matched 205 individuals (ages 16-66) with a peer mentor. Among people who sought mentorship, 71% reported their needs were mostly or completely met, and 67% stated they would recommend or highly recommended the program. Many participants reported this was the first time they were able to talk with another person who has CF. Six virtual events have been hosted since 2016 with a total of 1,014 participants (754 unique), ranging in age from 18-78. Grouped post-event surveys show that 88% of participants were likely to recommend the program to a friend.

Conclusion: Virtual connections in one-to-one and group settings can be an effective way to facilitate meaningful peer connections around life and life transitions for people with CF and their families. Both programs provide safe peer support with high levels of satisfaction in a population that faces challenges meeting with others who understand the experience and disease burden. Introduction: Michigan Medicine has a large Adult CF Center with over 300 patients. Nurses have historically filled the role as a center coordinator, which was true for our center as well. With a growing center and high work volume, our center explored the cost of adding another coordinator at the center. The cost of adding an additional RN was not feasible. After much discussion, our center decided to offer the position to our clinical social worker, who would work part time in clinic providing direct patient care and part-time as the center coordinator. The addition of social work to the coordinator role adds a multidisciplinary approach to the role.

Methods: When roles were first established, center and RN coordinators met with the director to establish and define the roles. Although both coordinators work closely together, it was important to establish parameters around each role. The workload was divided between both coordinators based on education, expertise, and skills. The RN assumed work that required having a medical education and social work took on process improvements and logistical management at the center. We meet formally once per week to have a check-in on projects we are working on and to reassess the needs of the center. Informally, we are in almost daily communication so we are cognizant of what is happening at our center. The program was piloted in 2015 and continues to evolve as we strive to meet the needs of the patients and the center's goals.

Results: The decision to create a dual shared role has allowed both our center coordinator and RN coordinator to utilize their skill sets to customize the role. Working as a team allows us to look at situations with different views to provide more diversified perspective. Splitting the role provided the RN more time engaging in clinical activities with patients. The social worker receives direct feedback during patients' annual assessments which allows coordinators to utilize this feedback to improve processes throughout the clinic. Projects we have worked on since we became center coordinator and RN coordinator include: development of a Patient Advisory Board, weekly inpatient rounds, and weekly team huddle with all members of the outpatient care team. Additionally, we implemented a transplant task force to evaluate and provide recommendations on our current transplant referral process. We have worked closely on organizing center data, such as genotypes and sweat tests, where previously there had not been time allotted for projects like this. Standardization has allowed us time to focus on projects to improve our patients' satisfaction, and process improvement. The processes we are implementing at our center will provide future coordinators structure with the role. The one challenge we face is that our offices are not physically in the same location, but this can be explored in the future.

Conclusion: Having two coordinators, one with a social worker skill set and one with nursing clinical background, has made our team stronger. Utilizing strengths of different disciplines benefits the center and allows for both coordinators to be more productive within their designated role. Our center feels this model works well and anticipates utilizing this dyad for the foreseeable future. during this time frame to obtain caregiver perspectives about their child's MH and barriers in addressing MH concerns (n=28). MH screening data were analyzed from 2015 to May 2018 to determine the impact of receiving MH intervention embedded in CF clinic on MH screening results. Analyses will determine if patients with mild, moderate and/or severely elevated MH symptoms experienced improvement in their symptoms based on the frequency of visits with MH provider in CF clinic.

Results: We found 47% of adolescents in our CF center are experiencing elevated concern for depression and/or anxiety. Those with elevated MH symptoms are significantly more likely to be female (p=0.05), have public insurance (p=0.008) and have a lower household income (p<0.001). Those with elevated MH symptoms had lower lung function than those without elevated MH screening (p<0.05). Results from caregiver survey showed 85% of caregivers believed their child's emotional health affects their ability to manage CF. Yet, 45% did not follow through with community MH referrals. Specific barriers will be reported. There was an overall trend of improvement in MH screening results following initial screening and subsequent visits with CF MH provider. Data will be presented to describe the exact relationship between number of visits by CF MH provider and MH screening scores over time.

Conclusions: Results of this study provide evidence of the value of embedding MH providers in improving health outcomes of pediatric patients with CF. Patients and caregivers experience barriers to accessing community MH services. Therefore, the relationship between the number of MH visits in CF clinic and improvements in MH screening could serve as an indicator of feasibility and efficacy of providing MH interventions within a busy CF clinic.

Mehdi, N.; Butler, D.J. Pulmonology, Univ. of Oklahoma CF Center, Moore, OK, USA Introduction: Our CF center is the only accredited center in Oklahoma. We have shared multidisciplinary team for adult and pediatric centers. Our team with less than part time social worker makes knowledge and implementation of health care transition limited. Our center needed to establish a standard process of assisting our patients with changing from a pediatric to adult model of care. CF R.I.S.E. is approved by AAP. CF R.I.S.E. program material, sponsored by Gilead, helps patients learn to balance the challenges of adulthood and CF. This program provides educational materials for managing their CF and health care independently. The materials detail understanding CF disease, treatments, CF rights, and responisibilities for CF care at home and daily life. This program provides access to credible, CF Foundation-supported guidelines for care.

Methods: Members of the transition team include CF physicians, CF coordinator, dietitian, social worker, mental health coordinator. The program is championed by the pediatrics CF director and dietitian. CF staff members and target population of CF patients 12 years and older completed paper Pre-Surveys that evaluated our current practice of transitioning patients from pediatric to adult programs in June 2017. Pre-Test Surveys were recorded and kept in clinic. Target patients (2-3 each week) were identified during their clinic visit to complete the Pre-Test Survey. These patients were enrolled into the CF R.I.S.E. portal to complete General CF Knowledge Assessment and CF Responsibility Checklist modules on paper or via tablet online. Scores were recorded. Education materials were given upon completion to review. Education materials include General CF Knowledge, ACT, Depression & Anxiety, Managing CFRD, Yearly CF Visit Tests, Exercise and Sexuality. Post-Survey will be given after 6 months of reviewing education topics. Scores will be recorded and compared to Pre-Test Survey and module scores. We predict improvement in CF knowledge measured by improvement in test scores. We predict a more organized method for beginning transition from pediatric to adult CF care.

Results: Staff members (12) completed the Pre-Test with 95% staff members strongly agreeing our center needs a process to clearly address transition from pediatric to adult care. Patients (n=21) completed enrollment and knowledge assessments. Patients who appeared to be most compliant during CF clinic visit interview typically had lower CF knowledge assessment scores. Our average CF Health Knowledge score was 80% out of 100%. Those with higher knowledge assessment scores report less compliance during CF clinic visit interview. We had difficulty with patients registering for the CF R.I.S.E. portal at home or via mobile app. This is despite stating significant interest in the program. Patients completing paper surveys and assessment have been successful. Clinic tablets allow patients to register for the CF R.I.S.E program during their clinic visit. Once registered, patients complete modules and additional assesments during subsequent clinic visits.

Conclusion: Online enrollment with clinic tablets will increase participation. We anticipate full time social worker and mental health coordinator to champion this program August 2018. This will allow additional support to improve the progress of the program. Background: CF is a rare, complex, and life-limiting chronic disease. CF care is provided on a specialized medical surgical unit at SJH, and practices on this unit can differ widely from others at the same hospital. Care is coordinated with the outpatient clinic at National Jewish Health. NJH is the largest adult CF care and research center in the US, with around 500 patients. All NJH adult CF patients needing inpatient care are admitted to SJH Medicine North, a 24-bed unit. The average length of CF admission is 13 days. The average daily CF census is 17 patients and projected to grow, requiring a move to a larger, 30-bed unit in 2018. The interdisciplinary team includes NJH pulmonologists, hospitalists, endocrinology, ENT, pharmacy, nursing, RT, PT, dietitian, behavioral health, and case management. Education for nursing and ancillary staff includes a CF Education Day for all CF-care RNs, CNAs, RTs, and PTs; bimonthly CF education seminars; and yearly CF-specific e-Learning.

Method: The poster provides an overview of the multidisciplinary CF care on the SJH Medicine North unit, including nursing care, isolation and infection prevention, and psychosocial challenges. It describes a "day in the life" of a CF patient, including treatment burden and patient-centered care. Examples of personalized care include accommodating patient preferences and meeting the unique psychosocial, nutritional, and exercise needs of each patient.

Outcomes: Success of nurse-driven, CF-specific care initiatives on the unit is demonstrated by low hospital-acquired infection rates and improved medication administration. No CLABSI (central line associated blood stream infection) occurred on Medicine North in 2017, with 4,930 total central line days. Only one hospital-acquired C. difficile infection occurred in 2017, despite heavy use of antibiotics. Reports of IV antibiotics left clamped declined by more than half from 2016 to 2017 after implementing a new process.

Conclusion: CF is a lifelong illness, requiring more frequent hospital admissions as the disease progresses. SJH staff develop a family-like relationship with CF patients and families. Including them in decisions about their care and accommodating preferences helps build rapport and improve treatment adherence. By understanding the unique needs of the CF population, even SJH staff who have limited interaction with CF patients can contribute to improving their care. Close collaboration with NJH outpatient staff ensures a smooth transition when hospitalization is needed. Background: At University of New Mexico Hospital, we noted that many times there was a delay from the time patients were admitted from clinic to the hospital to the administration of their first IV antibiotic dose. This delay was due to lack of venous access. Issues identified were that the increased from 22% in mid-Nov 2017, to 70-100% by late Jan 2018. Since then, while still improved from baseline, nursing presence during morning rounds has been more variable, (50-100%).

Conclusion: Shifting the start time of morning rounds and use of a checklist can improve participation and engagement of bedside nurses in nurse-led, multidisciplinary morning rounds on a CF inpatient unit. Background: Clinic visits with shared priorities are one means to build strong care partnerships between patients with cystic fibrosis (CF), their families, and the CF care team. As part of the CF Learning Network (CFLN), we focused on collaborating with parents of infants at our CF center to pre-plan upcoming clinic visits. By pre-emptively assessing and sharing family concerns and issues with the team, we endeavored to improve co-production of clinic visits with families of infants, who are among our most vulnerable patients with whom we are building supportive relationships.

Objective: To use iterative quality improvement (QI) methods to implement a nurse (RN)-driven process to increase participation of families of infants in pre-clinic visit planning (PVP).

Methods: A multidisciplinary QI team, including 2 parent partners used Plan-Do-Study-Act (PDSA) cycles to develop an RN-driven PVP process. We focused on families of patients <2 years of age. We reviewed PVP processes at other CFLN sites, and developed 3 questions identifying (1) family concerns, (2) priority team members to see, and (3) medication/ equipment access issues. The questions and process were finalized with input from CF RNs, medical assistants, and QI team parent partners. The RN would reach families by phone a week prior to the scheduled appointment. Parental responses are documented in the electronic medical record, and forwarded to the clinic coordinator. The clinic coordinator emailed the family-identified issues to the CF team 1 day prior to clinic. Contacted families received a survey during clinic visits to measure satisfaction. An RN survey measured process satisfaction, barriers, and process time after each attempted week of contact.

Results: From 2/21/18-5/8/18, RNs completed 6 of 12 (50%) attempted PVP calls. Three calls were not attempted due to high RN workload. All 6 PVP responses were shared with the CF care team the day prior to clinic. On the in-clinic survey, all 6 parents reported the call to be somewhat to very helpful. Five of 6 reported that they saw the requested team members. One parent did not answer. Five of 6 reported that they agreed or strongly agreed that the concerns they pre-identified were addressed during the clinic visit. Five RN surveys were completed. RNs reported an average of 18 minutes on the process, including the call and documentation, and that the time spent on patient care/coordination was the same or somewhat more in 4 of 5 surveys. RNs felt 4 of 5 parents were receptive to the call. The time spent on PVP was somewhat to completely worth it for 4 of 5 calls. The biggest challenge identified by RNs was time spent. RNs anecdotally felt calls improved the relationship with newly diagnosed families.

Conclusions: Early PDSA cycles over a 10-week period demonstrate high (100%) parent satisfaction with RN-driven PVP calls to pre-identify and address concerns in clinic visits. RN satisfaction was limited by time spent. Future PDSA cycles will test protecting RN time to address RN satisfaction.

Acknowledgment: Supported by Cystic Fibrosis Foundation SEID15A0. We sought to identify gaps and explore areas of opportunities that will lead to successful and sustainable quality improvement for the Mayo Clinic Cystic Fibrosis Center. This study took place following anecdotal reports from cystic fibrosis (CF) patients who had expressed concerns about the respiratory care they received during their inpatient visits. The CF care team decided to explore these concerns further. The specific objectives of this exploratory study were: a) to understand the patient experience including their respiratory therapy regimen; b) to identify areas of improvement during an inpatient CF hospitalization; and c) to understand and identify potential barriers associated with multidisciplinary team efforts in CF patient care.

Methods: We used a microsystem approach focusing on coordination of the multidisciplinary system of care for CF patients. This involved a purposeful sampling of major stakeholders who are involved in receiving or providing care for a CF diagnosis. Data were gathered using qualitative research methods via interviews with patients and the CF care team. This included 23 participants (physician=3; registered nurse=3; respiratory therapist=9, child life specialist=1, adult patients=5, parents of pediatric patients=2). Patient interviews focused on the overall inpatient care experience, therapy regimen, challenging factors to their care, and interactions with care team members/clinical staff. All interviews were audio recorded and transcribed verbatim. Care team interviews focused on admission processes, perception around roles, and respiratory care responsibilities. Perspectives on the challenges of CF care and possible practice improvement opportunities were explored. Data were analyzed using thematic analysis.

Results: Patient data: Functional aspects of care concerns included: lack of clear, consistent communication among the care team; coordination of services; and inconsistencies of patients' treatment plans. The major concern related to interpersonal aspect of patient care was physicians listening to patient concerns and thoughts. Care team data: undefined care team responsibilities specific to patient care; communication regarding different treatment plans; delay in respiratory care needs; inconsistencies in the use of airway clearance modalities; patient nonadherence to medication and treatment schedules; lack of parental involvement; and the inability to meet patient psychosocial needs during hospital stay. Care team practice improvement recommendations: improving care coordination through clear communication among teams, setting guidelines around care team roles and responsibilities, developing specific CF patient guidelines that encompass individual needs of patients and clearly delineate care team roles in regards to respiratory care, developing regulations that mandate family engagement, and preventing care delays through respiratory therapy alerts mechanism during CF patient admission.

Conclusion: The findings of the study will assist in developing an institutional CF guideline to improve care consistency and patient outcomes.

Dye, A.E. 3rd Floor Respiratory, Children's Healthcare of Atlanta, Atlanta, GA, USA Background: Children with cystic fibrosis (CF) must learn at a young age to manage CF by abiding to the prescribed treatment plans including medication self-administration, especially pancreatic enzymes. Pancreatic enzymes, essential with meals and snacks, are needed to absorb nutrients and ensure weight gain. Weight loss and decrease in pulmonary function test (PFT) scores often result in hospital admissions. Once hospitalized, CF patients must rely on licensed professionals administering their daily medications. In April 2015, a 35-bed pediatric hospital unit implemented a limited scope policy allowing CF patients 10 years of age or older or patients urinary incontinence, screening for intimate partner violence, and taking a sexual history. Nearly two-thirds (64%) felt identification of local women's health specialists familiar with CF would facilitate SRH care. Approximately half desired screening tools, national guidelines, and management algorithms for CF SRH concerns. Thirty-five percent desired provider SRH educational materials to view at the point-of-care and 33% desired online case-based learning. Few (<10%) were interested in role playing or simulation-based SRH skills learning.

Conclusion: Most interprofessional CF care team members lack the necessary comfort and skills to address SRH with AYA women with CF. Topics and formats for additional provider training identified in this work should be used to develop targeted interventions to improve this aspect of comprehensive CF care.

Acknowledgments: Support by the CF Foundation ( These guidelines allow only one CF patient in the event room at the same time for health safety concerns. Yet CF patients and families are continually requesting more education and updates from the CF Care Center. Thus, there was a growing need for a more effective means to incorporate larger numbers of CF patients at group education events.

In partnership with the local CFF chapter, our center used a combination of rotating small group discussions and video teleconferencing (VTC) to expand education opportunities for CF patients while continuing to comply with the CFF infection control guidelines.

Method: 1. A planning committee was formed including representatives from both the adult and pediatric center teams as well as members of the local CF Foundation.

-The planning committee booked an event room, invited vendors, and lined up speakers.

2. Invitations were sent out to our patients and families via clinic handouts, flyers, email, social media, and mail.

-In the invitation, attendees were welcomed to join the event in person or via VTC. Patients were invited to attend virtually to meet CFF infection control guidelines.

-Gift bags and raffle prizes were available to both on-site and virtual attendees to promote a larger turnout.

3. Topics for the event were planned from previous CF Family Education Day evaluations, clinic visits, and social media feedback from the CF community.

-The event agenda included an invited key note speaker, CF Compass representative, and multiple round tables.

4. The round tables were divided by topics applicable to all patients' ages and needs. Each round table had a moderator/educator between the two CF center teams. Additional laptop availability for the "hot topic" round tables was provided, which allowed patients to call in via VTC.

-The 8 different round table topics were offered across three 30-minute sessions. Participants could rotate to different tables per their need on topic education.

Results: Prior to the implementation of the current CF infection control guidelines, we had over 100 participants. After the guidelines, we had a total turnout of 60 attendees for both on-site and virtual attendance. Over 10 patients participated virtually. Overall about 20% of the attendees chose to participate virtually. Patients and families could be active participants and obtain education information through round table selections. The center team members interacted and participated with patients and families that were physically present as well as the ones connected virtually for the event.

Conclusion: We received positive reviews on our post-CF Education Day evaluations from both the virtual and physically-present attendees at the event. We will continue this teaching format for our future CF Family Education Day, as well as our upcoming CF Staff Education Day. Introduction: Sleep issues are not often discussed in the CF office visit. Sleep loss and sleep disorders can increase risk of anxiety/depression, produce lack of energy and result in poorer adherence. The nationwide implementation of Mental Health screening in CF provides opportunity to report sleep issues.

Objective: Measure how well equipped the CF centers are to handle sleep issues, what support is needed.

Methods: We developed a survey of attitudes and concerns at CF centers regarding sleep knowledge, ability to identify/treat sleep disorders and access sleep consultation. We partnered with the CF Foundation contacting centers via email to participate. Survey results uploaded to REDCAP. Categorical variables analyzed by chi-square test, continuous variables by ANOVA and ordinal variables (Likert scale) via Kruskal-Wallis test.

Results: Sixty-eight centers completed our survey. Respondents split between pediatrics (25), adult (22) and all age centers (21). Most centers were small/medium sized (41 with 1-150 patients, 18 with 150-300 patients and 11 with >300 patients). Center asking about sleep: All responding agreed (52.9%) or strongly agreed (22.1%) with routinely asking. All centers reported sleep issues important in care of patients regardless of age group, size of center (38.2 % agreed, 52.9% strongly agreed). Comments: patients underreport sleep issues, behavioral therapy access is limited. Testing: All responded agreed (33.8%), strongly agreed (54.4%) they are comfortable ordering sleep tests. Request: referral guidelines for age/symptoms/FEV1 to prompt testing. Sleep referral common practice overall (32.4% agreed or 44.1% strongly agreed). Convenient referral process: agreement (39.7%) or strong agreement (40.9%). Some reported limited to very limited access to testing. Sleep equipment: Utility of testing for NIPPV/Bilevel/O2 strongly agree for pediatric centers (60%) vs. adult (33.3%) or all age centers (33%). Many center directors reported frustration over approved indications for NIPPV. Sleep medications: Significant difference in comfort prescribing sleep medication (p 0.0016): Adult centers reported comfort prescribing for sleep issues (54.4%) agreed or strongly agreed (4.5%) while centers caring for children disagreed or strongly disagreed (pediatrics 32%, 20%); all age centers (33.3%, 28.6%). Safer drugs for treatment of insomnia was concern raised by some. Additional Resources: Most centers reported value in educational resources for sleep diagnosis/ treatment for patients and providers (48.5% agreed, 32.4% strongly agreed). Requests: general information on sleep/good sleep habits, insomnia and treatment, sleep/wake rhythm disorders.

Conclusions: Improving CF outcomes requires a comprehensive approach. Untreated sleep disorders contribute to poorer adherence to therapy. While CF centers report comfort in evaluation of sleep problems, this survey highlights areas to address. Sleep education resources for patients and providers, criteria for testing, improved qualifications for therapies needed. Improved early NIPPV for progressive respiratory failure is needed but will require proof of medical benefit.

Acknowledgments: Thanks to CFF. Introduction: Cystic fibrosis (CF) education is ongoing in CF centers. In the past, the University of Mississippi has held an annual CF Patient and Family Education Day in Jackson, MS. This event was held on Saturdays lasting eight hours. Attendance by family representatives has historically been poor due to distance, infection control restrictions and childcare. despite expressed interest. In addition, patients with frequent exacerbations, who would strongly benefit from additional education and support, agreed to participate but did not attend. Ways to address barriers to patient participation should be studied, especially in patients with frequent exacerbations.

individualized genetics consult. Subjects repeated the knowledge and perceptions survey immediately following genetic consultation as well as again 4-8 weeks later to assess for retention of knowledge.

Results: A preliminary analysis was performed following enrollment of the first 9 subjects. Participants were 78% male with a mean age of 31.2 ± 7.2 years (mean±SD). Most subjects self-identified as either in a committed relationship or engaged/married (67%). In addition, 44% expressed they wanted children in the future, 44% expressed they did not want children, and 11% already had children. The majority of subjects, 78%, noted they had never talked to a healthcare provider about fertility options.

The average pre-test knowledge score was 12.78 ± 6 out of 21 total questions. Initially, 33% of subjects incorrectly answered at least 1 of 2 questions regarding assisted reproductive technologies. Genetic consultation lasted 29.4 ± 5.8 minutes. Following the consult, the average immediate post-test knowledge score increased to 15.2 ± 4.4 (p<0.05) with all but one subject improving their score.

Conclusion: Consistent with previous data, few individuals with CF reported having previously structured discussions regarding CF fertility or fertility options. There was wide variability in this small, preliminary cohort as to family building preferences. Even a short genetic consultation was associated with a significant improvement with CF-specific genetic knowledge. Further studies must be performed to assess whether retention of knowledge is sustained and whether genetic knowledge is associated with reproductive perceptions. The CF Foundation (CFF) provides clinical guidelines to inform and improve CF care delivery. Guidelines may be evidencebased (developed from a systematic review of the best available evidence) or consensus-based (developed primarily from the opinions and experience of experts). Prior research has suggested that many CF patients are not receiving guideline-recommended care. The CFF Guideline Steering Committee (GSC) thus conducted a survey designed to examine how CF centers learn about these guidelines, challenges to their implementation, and if implementation differs by type of center (adult, pediatric, or affiliate).

Methods: The CFF GSC (composed of members of the multidisciplinary care team, CFF staff, two parents of children with CF, and one adult with CF) developed the survey. The survey consisted of 22 multiple choice and free text questions. It was sent to CF center and program directors and coordinators via SurveyMonkey; 699 people received the survey. The survey was distributed on February 22, 2018. A reminder was sent on March 8 and the survey closed on March 16.

Results: Individuals at 58 adult programs, 73 pediatric programs, and 17 affiliate programs completed the survey (response rate=21%). Respondents noted hearing about new guidelines from CFF emails (96%) and listservs (44%). Among pediatric program respondents, 27% reported always reviewing the guidelines as a care team compared to 9% and 6% of adult and affiliate program respondents, respectively (p< 0.001). Guidelines are most commonly disseminated to team members in meetings (n=70) and via email (n=71); affiliate program respondents were less likely to use email (29% vs 50% and 51% of adult and pediatric program respondents, respectively). On a question where respondents could select multiple responses, 45% of participants noted using all the guidelines, 33% noted they may not use some guidelines because the strength of evidence is not enough to change current practice, 14% said they do not agree with some of the recommendations, and 11% said they do not agree with the entirety of some guidelines. Barriers to guideline implementation include lack of time during clinic visits, lack of clinic space, insufficient staff, funding limitations, and a lack of patient buy-in. Of note, 15% of respondents were unaware of the executive summaries, 31% were unaware of the clinician-targeted educational handouts, and 32% were unaware of the patient-targeted educational handouts.

Conclusion: The majority of CF center and program directors and coordinators are aware of CF guidelines, but are less aware of supplemental materials designed to facilitate their implementation. Structural barriers to implementation of guidelines were identified. One dissemination strategy, reviewing guidelines as a team, was rarely done, especially by adult and affiliate centers. Further research is needed to understand which strategies result in better implementation of guidelines.

Cathcart, F. 1 ; Holder, E. 1 ; Medhurst, N. 2 ; Madge, S. 1 1. Adult Cystic Fibrosis, Royal Brompton Hospital, London, United Kingdom; 2. Cystic Fibrosis Trust, London, United Kingdom Background: Over the last decade there have been significant advances in CF however, as CF remains a life limiting disease, high quality, end-oflife care must remain a priority. Recent literature suggests a paucity in both CF-specific end-of-life documentation and education for clinicians. In order to mitigate this we developed a two-stage strategy. The first stage was to develop a CF specific Advance Care Plan with an accompanying guideline for healthcare professionals, now available on the UK CF Trust website. The next stage was to provide a CF-specific, end-of-life education platform for the multidisciplinary team.

Objective: To develop a CF-specific end-of-life care course for healthcare professionals.

Method: A proposal was developed and funding sought to film and design an online course. A modular programme was developed and an international, multi-professional faculty were invited to contribute.

Results: The course content is divided into four modules: end-of-life care, supportive and palliative care, lung transplantation and non-invasive ventilation, symptom control and advance care planning. Each module consists of short presentations and animations. Contributors include clinicians from both CF and palliative care, patients and family members. The online course can be accessed via a link through the UK CF Trust website and is free to all. Participants can work through the modules in their own time with a certificate of completion at the end. Anonymised data are collected on the participant's profession, country and self-assessment confidence score.

Conclusion: This online, end-of-life course allows flexible, easily accessible free training to all members of the CF MDT. Continued audit of the anonymised data will not only reveal the number and speciality of participants but the usefulness of the course in promoting confidence in end-of-life care. Hemoptysis is the coughing up of blood. Many factors (diabetes, S. Aureus) can contribute to hemoptysis. It is thought that chronic infection and inflammation in the airway is the cause of the bleeding with the majority of the blood coming from the bronchial artery (1) . CF patients (pt) report the occurrence of hemoptysis. Pts struggle with describing the amount of hemoptysis and fail to notify the clinic. Our center created a structured educational tool to help CF pts better identify the amount of blood they cough up. The aim of the project is to educate the pts on when to notify the staff and better describe the amount of hemoptysis they experienced. The study investigated the effect of the hemoptysis educational tool in CF adult pts. We educated pts in clinic about hemoptysis, amounts that should be reported and when they should inform the clinic. Participants included CF patients who had appointments in April, 2018 at CF Center, Keck Hospital of USC. Pts were given a pre-test, a hemoptysis handout (2) , visual aids and a post-test. The pre-test examined their hemoptysis knowledge. The handout was an informational sheet detailing a description of hemoptysis, symptoms, treatments and visual representation of when to call the clinic if hemoptysis occurs. Study staff reviewed the handout with visual aids that represented different common household items to describe the amounts of hemoptysis. Pts completed the post-test to establish what they learned. Data were collected: pts' demographics, PFT, pre/post-test questionnaire. The pre-test assessed if pt coughed up blood, how often and when, does the pt know when to call the CF clinic if and when they have hemoptysis, and if they know what hemoptysis is. The post-test included the evaluation of the educational tools. 36 pts were educated on the hemoptysis project. Average age (years) was 32.7±8.3. 53% Male; 47% Female. Average FEV1 (liters) measured for PFT was 2.47±1.03. Average of FEV1 was 69%. Based on the pre-test, majority of pts (78%) reported coughing up blood. More than half of pts (53%) were aware of what hemoptysis is. 61% of pts did know when to call the CF clinic when experiencing hemoptysis and 39% did not know when to call. After being educated, 89% pts were aware of what hemoptysis is. 81% pts now knew when to call the clinic when experiencing hemoptysis. Post-test surveyed pts on the helpfulness of the educational tools. 100% of pts found the handout, visual aids or both tools were helpful. Hemoptysis is a medical condition that happens to CF pts. It is important to assess how well pts are informed about hemoptysis. Our findings suggest that pts have coughed up blood and need to understand the importance of this medical condition. There is significant evidence of pts encountering hemoptysis. Findings determine that the educational tools made an impact on pts' knowledge. Overall, the educational study was evaluated on the basis of how well patients became informed and aware about hemoptysis in order to call the clinic and accurately describe the hemoptysis episode. Over the past several decades, life expectancy for patients living with cystic fibrosis (CF) in the United States has increased to an average of 43 years, with many patients exceeding that life expectancy by significant margins (Cystic Fibrosis Foundation Patient Registry. 2017 Annual Data Report. Bethesda, MD). Due to this increased longevity, patients who were previously limited to the pediatric world are now living long enough to transition to adult providers. The transition from pediatric to adult care is a growing topic of interest, and requires significant adjustment for both patients and providers. As a part of this transition, patients with cystic fibrosis are admitted to Internal Medicine wards for exacerbation therapy, which can result in inpatient hospital stays of several weeks.

We have developed a pair of surveys for patients and resident physicians at an academic CF care center to identify the points of primary conflict during these lengthy admissions. Preliminary data from our mixed method cross-sectional study has elucidated several barriers to optimizing quality patient care. From the perspective of resident physicians, challenges include: unfamiliarity in managing CF exacerbations, lack of autonomy, and difficulty managing patient expectations. From the perspective of patients, we see frustrations with unclear admission and discharge goals, blood sugar management, and frequent medication changes. Based upon these results, we are tailoring interventions in the form of informational primers for both patients and residents that we hope will mediate possible areas of miscommunication during hospital admissions. We hope that our research and interventions can improve patient care at our institution and be utilized by other CF care centers to improve the quality of care for this unique patient population.

McEwen, C.; Rotolo, S.M. University of Chicago Medicine, Chicago, IL, USA Background: Cystic fibrosis-related diabetes (CFRD) is one of the most common comorbidities associated with Cystic Fibrosis (CF) and affects approximately 20% of adolescents and 40-50% of adults. CFRD shares features of type 1 diabetes mellitus (T1DM) and type 2 diabetes mellitus (T2DM) but is a unique disease. Over time, thick mucus around the pancreas results in scarring and ultimately leads to decreased insulin production. Additionally, many individuals with CF have a component of insulin resistance. An oral glucose tolerance test (OGTT) is the gold standard for diagnosis of CFRD and should be conducted annually starting at age 10 years. Based on low rates of adherence with annual screening at our CF center, we designed a two phase quality improvement effort. Phase 1 consisted of updating the order capabilities for OGTT in the electronic medical record (EMR) to streamline provider ordering and improve clarity of instructions to laboratory staff. Phase 2 consisted of a patient education effort, utilizing materials designed by a multidisciplinary group with endocrinology representation.

Methods: A retrospective analysis of OGTT adherence rates during pre-(n=50) and post-phase 1 implementation (n=52) of CF patients over age 10 years from a single CF center was conducted. Individuals with a previous diagnosis of T1DM or CFRD were excluded. The primary outcome was to determine adherence rates with OGTT guidelines pre-and postimplementation of the new OGTT order in the EMR.

Results: There was no significant difference in age (p= 0.58), percent predicted FEV1 (p=0.88), and body mass index (BMI) (p=0.32) between pre-implementation and phase 1 implementation. Overall OGTT adherence rate was 28% in both pre-implementation and phase 1 (p=1.0). Adherence amongst adult patients was 22% pre-implementation and 29% after phase 1 (p=0.60). Adherence amongst pediatric patients was 43% pre-implementation and 27% after phase 1 (p=0.45).

Conclusion: There was no significant increase in overall adherence to annual OGTT recommendations from pre-implementation to phase 1. Amongst adults alone, while there was no statistically significant improvement in adherence between the time points, clinically the rates of adherence rose from 22% to 29%. This trend may be the result of EMR changes or increased provider awareness during the immediate postimplementation period. Another limitation to the current data is incomplete data collection from phase 1 (scheduled to end July 2018). Despite these limitations, the trend towards increased adherence in adults is encouraging and our findings further demonstrate the need for phase 2 of targeted education. Finally, compliance rates of annual OGTT testing seen in this cohort are comparable to the national average amongst CF centers. This nationwide overall low adherence may leave room to explore alternative measures of glucose metabolism that have a decreased burden to patients.

Porco, K.; Landon, C. Ventura Medical Center, Ventura, CA, USA Rationale: Attain Health is a patient-centered care program uniting integrative health and physical performance coaching with community engagement and peer accountability to optimize the health of people living with cystic fibrosis. Patients understand the benefits that exercise, sound nutrition, time management, mindfulness, and productive treatments have on their health and life expectancies. Patients are overwhelmed by the demands of cystic fibrosis, feel out of control and difficult to evoke change. These constant demands combined with the fear of cross-contamination leave individuals isolated and without peer support.

Methods: Each client met with a Performance Coach and Integrative Health Coach in weekly virtual meetings, weekly group coaching sessions and educational webinars from experts in the CF community. A platform of Integrative Health and Physical Performance Coaching offered clients the opportunity to set personal health goals, develop implementation plans, establish habits, and overcome inevitable setbacks to achieve their health goals. This was achieved through a virtual community, providing accountability and the community to support each other to integrate positive change. Each client was offered a ninety-minute consult with the Integrative Health Coach (IHC). The IHC discussed the parameters of the program, vision work, create an outline of the client's goals for the program, and the group coaching and counseling. A 90-minute consult followed with the performance coach conducting a movement assessment screen and establishing fitness goals. The performance coach created a strength, conditioning, and movement program for the client. The clients have unlimited access to the IHC to maintain motivation, employ mindfulness techniques, and work toward their ultimate health goals. The client had the opportunity to engage in video conferences with the performance coach to teach proper exercise technique, and address any difficulties that the client is having with the exercise program.

Results: The three-month pilot program served fifteen clients, ages 17-54, from the United States, Canada, and Denmark. Mean increase in FEV1 was 3.3% over the three-month study period with a range of -2% to +15%. Of clinical significance is one patient with FEV1 of 21% increased to 33% and was taken off the transplant list. BMI improved in 10/15 patients. Two patients had goals of weight loss (BMI 29 and 31) and were successful.

Conclusion: The pioneering cystic fibrosis community, first to arrive in the genomics to proteomics revolution, also requires humanomics to overcome the imposed isolation of bacterial cross-contamination. Telehealth and integrated healthcare offer the promise of relief. Existing expenditure analyses are out-of-date with current treatment regimens and mostly limited to people with private insurance. We assessed recent trends in healthcare expenditures for people with CF on Medicaid, a safety-net insurance program for Americans with low incomes.

Methods: A retrospective analysis of a convenience sample of Medicaid claims data was conducted for individuals aged 0-to-64 years who were continuously enrolled for at least 1 calendar year during 2010-2016. Mean annual expenditures during a calendar year were calculated for individuals who met a claims-based CF case definition, with two or more outpatient claims with a diagnosis code for CF or one inpatient claim associated with CF. Average annual growth rates were calculated through linear regression of the natural logarithm of annual expenditures.

Results: Average Medicaid spending adjusted for inflation nearly doubled from roughly $49,000 per patient in 2010 (n=1170; 72% pediatric) to approximately $83,000 per patient in 2016 (n=1123; 66% pediatric). Inflation-adjusted spending on outpatient and inpatient care increased by 0.1% and 1.4% per year, respectively, whereas pharmaceutical spending increased by 15.1% per year. By 2016, 65.6% of mean annual expenditures were accounted for by pharmacy spending, with much of the growth accounted for by spending on specialty drugs. The annual growth rate in pharmaceutical spending rose by 33.1% during 2014-2016, the years during which lumacaftor/ivacaftor was introduced. In 2016, 2% of patients had filled prescriptions for ivacaftor and 10.6% had prescriptions reimbursed for lumacaftor-ivacaftor; ivacaftor and lumacaftor/ivacaftor accounted for 8% and 21% of all pharmaceutical spending, respectively.

Conclusions: These are the first publicly available estimates of healthcare utilization and expenditures for publicly insured US patients with CF. Per-patient expenditures for Medicaid-insured patients with CF increased considerably during 2010-2016; specialty drugs were largely responsible for this increase. The new, genotype-targeted CFTR modulators contributed to the increase in Medicaid expenditures, although the frequency with which prescriptions for ivacaftor and lumacaftor-ivacaftor were reimbursed remained low in comparison with CF patient registry data as well as data on privately insured patients with CF. It is not known how many of the patients in the Medicaid sample had dual coverage with private insurance and whether they had prescriptions for CFTR modulators reimbursed by those private plans. It is important for further research using other data sources, i.e., patient registry data, to ascertain whether there is indeed lower use of CFTR modulators by Medicaid-insured patients and, if that is the case, to identify potential barriers to access for that population. Finally, economic evaluations of CF therapies should use up-to-date estimates of healthcare costs that reflect current clinical practice. Methods: This study was conducted using insurance claims data from Inovalon's Medical Outcomes Research for Effectiveness and Economics Registry (MORE2 Registry®). Data were evaluated for each year from 2009 to 2015. Individuals were included in the study if they had at least two claims with CF listed as a diagnosis more than 30 days apart with at least one claim for a CF-specific medication or if they had previously been matched to the CFF Patient Registry. Individuals also had to be continuously enrolled for at least one calendar year. Utilization was categorized by inpatient, emergency department, outpatient, professional services, medications, and other. Utilization was stratified by age, sex, race, payer type, class of medications, transplant status, and diagnosis of CF-related diabetes or depression.

Results: A total of 6,306 people with CF were analyzed in the study with approximately 1,900-2,800 individuals analyzed in each calendar year. Of the 1,923 individuals studied in 2015, the mean age was 19 years, 52% were male, and 71% had Managed Medicaid. Of note, the percent of individuals with Managed Medicaid was much higher in 2015 than prior years, suggesting an increase in Inovalon's Medicaid managed care member population, not the CF population itself. In 2015, 32% of studied individuals had at least one inpatient hospitalization, and among them there was an average of two hospitalizations with an average total length of stay of 18 days. In addition, 36% had at least one emergency department visit. Among those with at least one visit, there was an average of two visits to the emergency department. Almost all individuals had at least one outpatient visit (93%) with an average of nine visits. Within the population, 4% of individuals had prescription claims for ivacaftor, 15% for lumacaftor/ivacaftor, 73% for pancreatic enzyme replacement therapy, 73% for mucolytics, 41% for inhaled antibiotics and 71% for other antibiotics. The hospitalization rates of Medicaid and privately-insured individuals were similar, but the average length of stay for hospitalized Medicaid patients was 1.5 times that of the commercial patients. Individuals with Medicaid had a higher prevalence of emergency department visits and equal numbers of outpatient visits. All comorbidities examined were associated with greater utilization, but the most substantial difference was seen with depression. Those with depression were 2.3 times more likely to have had at least one hospitalization and, among those hospitalized, had more than double the length of stay compared to those who are not depressed.

Conclusion: Claims data provided a comprehensive assessment of the utilization of health services and prescription medications by people with CF. These data suggest overall high utilization with subsets of the population demonstrating higher utilization. Further research should investigate drivers of the variability within the CF population and explore the impact on health outcomes and cost of care. Daniels, M.L. 1 ; Wong, E.T. 2 1. Pulmonary/Critical Care, University of North Carolina, Chapel Hill, NC, USA; 2. Pharmacy, University of North Carolina Healthcare, Chapel Hill, NC, USA Background: Disease-specific therapies have changed the course of CF by improving symptoms and extending life expectancy. CFTR sequencing availability has increased the number of adult diagnoses, often with residual function mutations. Drug coverage is particularly challenging for adult CF patients with Medicare through disability or age ≥65 years. While Medicare supplemental plans (Part D) cover CF-specific therapies, cost sharing for Tier 5 drugs range from 25-33% of total drug cost. Medicare patients are ineligible for manufacturer patient assistance programs (PAP) due to the Federal anti-kickback statute.

According to the 2016 Cystic Fibrosis Foundation (CFF) Patient Registry Annual Data Report, 2783 (18.4%) of the 15,154 adult patients have Medicare/Indian Health Services. As CF management improves, the number of Medicare patients will continue to rise and the cost of specialty medications will adversely affect this growing population. We sought to describe the prevalence of UNC Adult CF Center patients with Medicare who may be unable to afford disease specific therapies (dornase alfa, inhaled antibiotics (abx), and CFTR modulators) due to high cost copays and PAP ineligibility.

Methods: Retrospective chart review of current UNC Adult CF Center patients in Spring 2018 screened for Medicare as primary medical coverage. Prescription coverage was classified as Part D only; Part D/Medicaid; Part D/commercial; commercial only; Tricare; or none. Demographic data included age, gender, BMI, FEV1 percent predicted, and CFTR mutations. Descriptive data included number of specialty medications.

Results: Of the 306 UNC Adult CF Center patients, 56 (18.3%) have Medicare (9 due to age) with median age of 56.5 years, median FEV1 of 49%, and median BMI of 21.8. Nearly 70% (39/56) are either on a modulator (26/56) or qualify (13/56) for a modulator. Drug coverage for these patients: Medicare Part D only (n=15); Part D/Medicaid (n=26); Part D/Commercial (n=2); Commercial Only (n=8); Tricare (n=4); none (n=1).

Of the 4.6% with Medicare Part D only for drug coverage, 7 use dornase alfa, 4 use one inhaled abx, and 3 use two inhaled abx. Six of the nine patients eligible for CFTR modulators are receiving treatment. Half of the Medicare D only patients use at least 2 specialty medications. Of the 6 patients who will age into Medicare within 5 years, 4 are CFTR modulator eligible, 4 use dornase alfa, and 2 use inhaled abx.

A cost-analysis example of a Part D patient using alternating two inhaled abx, ivacaftor, dornase alfa, and pancreatic enzymes has an estimated annual patient cost of $47,932.52.

Conclusion: Nearly 5% of patients at the UNC Adult CF Center have Medicare Part D with more expected to acquire Part D due to age or disability over time. Part D patients incur high copay costs due to specialty drug cost and cost sharing. It is critical for pharmaceutical companies and the CFF to work together within the law of the anti-kickback statute, such as through an independent charity PAP or a "coalition model" PAP, to ensure that all Part D patients can afford these necessary drugs. Introduction: As the cystic fibrosis (CF) population increases and treatments escalate in complexity, costs related to CF care are expected to rise and could put tremendous strain on health care systems. CF health care costs have been explored in several countries. A recent scoping review identified 28 studies that reported overall CF health care costs; interestingly cost drivers varied across studies, with hospitalizations accounting for the majority of cost in some and pharmaceuticals dominating in others (Hollin H, et al. Appl Health Econ Health Policy. 2016; 14:151) . However, there are no Canadian studies of national CF health care costs or hospitalization costs. The aim of this study was to examine the hospitalization costs of CF patients in Canada.

Methods: Record level data were obtained from the Canadian Institute for Health Information databases. CF patients were defined based on at least one hospital admission with an ICD-10 code of E84. Costs were estimated using an aggregate costing strategy. The primary objective was to determine the annual hospital costs of CF patients in the 2014 fiscal year using a public payer perspective. Secondary objectives included (1) examining trends in annual hospital costs from 2010-2014 and (2) comparing provincial differences in hospital costs per patient and per hospitalization in 2014. Cost estimates were determined per capita CF patient and converted to 2014 Canadian dollars. Calculations assumed a fixed cost for a treatment or service for a given year. A sensitivity analysis was performed in which transplant patients were excluded.

Results: The number of CF patients requiring hospitalization and hospitalization events increased from 2010 to 2014 (913 vs 964 and 1564 vs 1741, respectively). Of those hospitalized in 2014, 50% were female and median age was 21 years (IQR, 14-30). In 2014, CF patients had a median of 2 hospitalizations (IQR, 1-4) and a median 10 hospital days (IQR, (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) . CF-related complications accounted for 70% of admitting diagnoses in 2014. The total cost of all CF hospitalizations increased from 2010 to 2014 ($29 vs $34 million). The mean annual hospital cost per patient was $35,684 compared to $31,745 in 2010. The mean cost per hospitalization remained similar ($20,955 vs $19,782) . Sensitivity analysis with transplant patients excluded revealed that the total annual hospital costs and the mean cost per patient decreased by $1.1 million and $742, respectively. There was a significant difference in mean annual hospital cost per patient among provinces in 2014, with Saskatchewan reporting the highest costs. Regression analyses to identify factors associated with higher mean cost per hospitalization are pending.

Conclusion: Over a five-year period, there was an overall increase in the total cost of all CF hospitalizations and the mean annual hospital cost per patient, while the mean cost per hospitalization remained similar. In 2014, hospital cost differences were identified among provinces. The increasing financial burden of inpatient CF care appears to be related to increased hospitalizations, however further investigations are under way to explain these findings.

Dufeck, E. 1 ; Ehlert, K. 1 ; Tran, L. 2 ; Espinosa, R. 1 1. University of MN CF Center, Minneapolis, MN, USA; 2. Fairview Pharmacy Services, Minneapolis, MN, USA Background: The University of Minnesota CF Center applied for and was granted financial assistance from the CF Foundation to implement outpatient pharmacy services. As part of this grant a certified pharmacy technician was integrated into the CF multidisciplinary team in March 2017. The technician is shared amongst the pediatric and adult centers. Duties of the technician include: prescription benefit investigation, completing prior authorizations (PA), completing insurance appeals within 24 hours of receiving PA denial, routing prescriptions to correct pharmacies, obtaining co-pay assistance from pharmaceutical manufacturers, assisting patients in securing funding through HealthWell and other grants, face-to-face patient education on medication coverage, and serving as a resource to the multidisciplinary team and patients. The technician also attends weekly outpatient rounds and is involved in CF center quality improvement initiatives.

Objective: To evaluate the impact of the pharmacy technician on PA management. Additionally to characterize PAs needed for medications prescribed in a large pediatric and adult CF center.

Methods: A retrospective review of all PAs completed from March 1, 2017 to March 30, 2018 was completed. The data were collected from the electronic medical record (EMR). The pharmacy technician processed test claims through the health system outpatient pharmacy dispensing software to determine if a PA was needed.

Results: In the one-year time frame 439 test claims were submitted to insurance. Of these, 415 (94%) required a PA. Among the PAs that were required, 73% were for specialty medications and 27% were for nonspecialty medications. Of the PAs submitted 353 (85%) were initially approved. There were 62 (15%) PA denials. In 38 (61%) of the denials, an alternative medication was prescribed. The remaining 24 denials went on to be appealed. Out of these appeals 14 (58%) were approved and 10 (42%) were denied. PA was not needed for 24 (6%) of all the test claims processed, including 15 specialty medications and 9 nonspecialty medications.

Conclusion: Integration of the pharmacy technician has had a positive impact on the multidisciplinary team and patients. The technician has a high rate of successful PA and appeal completions thus mitigating the burden of this process from nurse coordinators. This has allowed for nurse coordinators to increase time spent in direct patient care. Approximately one-fourth of PAs were for nonspecialty medications, which are often not FDA-approved for CF. Interestingly, 6% of test claims revealed no PA was necessary. The ability of the technician to run real-time test claims through the insurance has conserved resources by preventing unnecessary claims. Introduction: Three inhaled antipseudomonal antibiotic formulations are currently approved for the management of chronic Pseudomonas aeruginosa infections in cystic fibrosis (CF) patients in the USA. This retrospective study using US MarketScan databases assessed the use of antipseudomonal antibiotics (oral/IV) and hospitalizations due to pulmonary exacerbations in CF patients, who were treated with tobramycin inhalation powder (TIP), tobramycin inhalation solution (TIS) or aztreonam for inhalation solution (AZLI).

Methods: Patient data between 01 May 2012 and 30 June 2016 were retrieved for analysis. CF patients aged ≥6 years at the index date with their first prescription/claim of index treatment (TIP, TIS or AZLI) during the identification period (01 May 2013 to 30 June 2015) and who had at least one further claim within 1-year after the index date were analyzed. Use of antipseudomonal antibiotics (oral/IV) and hospitalizations due to pulmonary exacerbations in CF patients in 1-year post-index were primary outcomes of interest. In addition, time-to-first use of antipseudomonal antibiotics (oral/IV) and time-to-first hospitalization for CF pulmonary exacerbation in the post-index period were estimated.

Results: Among 755 eligible patients, 191, 283, and 281 were in TIP, TIS and AZLI cohorts, respectively. Compared with TIP cohort (35.1%), higher proportion of patients in the AZLI cohort (47.7%, p<0.01) received IV antipseudomonal antibiotics pre-index. More patients in AZLI cohort were hospitalized in the pre-index period due to exacerbation (43.8%, p<0.01) compared to TIP (30.4%) cohort. Proportion of patients in the TIP cohort (73.8%) who received antipseudomonal antibiotics (oral/IV) in 1-year after index date was comparable with TIS cohort (73.1%), and was numerically lower than that of the AZLI cohort (80.4%). Use of IV antipseudomonal antibiotics post-index was significantly lower in the TIP cohort (36.7%) than in the AZLI cohort (47.3%, p=0.02). The median timeto-first use of any antipseudomonal antibiotic was significantly longer in the TIP cohort (133 days) than in AZLI cohort (107 days). In all three cohorts (TIP vs TIS vs AZLI), the proportion of patients hospitalized (40.3% vs 36.0% vs 44.8%), average number of hospitalizations (0.8 vs 0.7 vs 0.9), and duration of stay (17.3 vs 19.8 vs 21.6 days) for CF pulmonary exacerbation in 1-year after the index date were comparable. A trend towards delayed hospitalization due to pulmonary exacerbations was observed in patients in the TIP cohort (median time-to-hospitalization, 849 days) compared with AZLI cohort (538 days).

Conclusion: Patients in the TIP cohort have reduced antipseudomonal antibiotic use and reduced risk of hospitalization due to exacerbations compared with patients in ALZI cohort, while TIP and TIS cohorts were comparable.

FDA approval. The patients were grouped into three urgency groups for obtaining tez/iva. Group one included patients who did not tolerate lum/ iva and had frequent exacerbations or recent lung function decline. Group two included stable patients who were not on lum/iva. Group three included patients who were stable on lum/iva but preferred to switch to tez/iva. A request was placed in Epic for an electronic prescription to be created. A newsletter was released to patients regarding FDA approval of tez/iva and expected medication interactions, side effects, safety and dosing based on comorbidities as well as need for lab test monitoring. We created a smart phrase in Epic for patient education to be included in after-visit summary. Charts were screened prior to each visit for patient eligibility to tez/iva and in-clinic education was initiated. Patient's prescription status was collected in a secure server.

Results: Our adult center serves 100 patients with CF that are seen quarterly. Tezacaftor/ivacaftor-eligible patients represent 60% of our population. Group one included 22 patients and currently 21/22 have received tez/iva. Pre-authorization is pending for one patient. Group 2 included 12 patients and 8/12 patients are taking tez/iva. Two patients are in the process of obtaining prior authorization. One patient decided to wait as she wishes to conceive. One patient was never prescribed due to no insurance coverage. In the third group, 8/26 patients have started tez/iva and 11 patients have not been prescribed tez/iva due to noncompliance with clinic visits. The 7 remaining patients have not started tez/iva: one patient would prefer to wait as she wishes to conceive, one refused the medication, and four have not been approached yet by the team or have decided to wait to start over the summer. In the 3 months after tez/iva approval, 66% of eligible patients are either taking tez/iva or are in the process of initiating.

Conclusion: A standardized approach to prescribing the new CFTR modulators ensures timely access to these medications for eligible patients. The main barrier to patient access to new drugs is patient noncompliance with clinic visits especially when routine lab tests are required to monitor for drug toxicity. (CFF, 2017) . Experiences from CFF Compass suggest barriers to accessing proper care and resources may exist throughout the transplant process. We sought to investigate these experiences to understand and identify the most common transplant access issues encountered in CF.

Methods: CFF Compass is a free, personalized service to help with the insurance, financial, legal, and other issues people with CF are facing. To understand the transplant process and barriers to care, we analyzed case notes of 382 transplant-related calls to CFF Compass from February 2016 to April 2018. These calls were reviewed and sorted into topics.

Results: Although each transplant experience is unique, several reoccurring needs emerged. These needs fell into the following categories: insurance coverage (106), general financial support (82), housing/relocation (74), medication/device access (47), general information requests (47), social security/disability (45), transportation (42), food (25), legal help (14), utilities (11), and locating a transplant center (10). It was not uncommon for CFF Compass callers to identify one or more of these needs. For example, of the 74 individuals with housing or relocation concerns, 34% identified transportation needs and 26% required assistance with food resources.

Of the 382 CFF Compass transplant cases, 224 calls originated from people other than the person with CF, such as parents, care center staff, spouses, and friends. The top needs varied by caller; for example, 36% of patients called about insurance coverage, and 31% of parents called for help with housing/relocation. Friends were most likely to call with a general inquiry (50%), such as details about the transplant process or how to donate organs themselves.

We analyzed the characteristics of transplant patients by reviewing gender, state of residency, and stage of the transplant process. Inquiries for males were 39%, 43% for females, and 18% of cases did not specify the gender of the patient. Compass received most calls from Florida (36), California (28), New York (28), and Texas (27). Two-thirds of the calls occurred pre-transplant while one-third occurred post-transplant. The top three pre-transplant needs related to insurance (67), financial (64), and housing (54) while post-transplant needs were insurance (33), medication (29), and social security/disability (18).

Conclusion: The path of the transplant journey is complex and access needs exist throughout the pre-transplant and post-transplant process. Not surprisingly, access issues for CFF Compass callers were mostly financial and logistical. Without these needs met, people with CF cannot access proper care which could potentially affect long-term outcomes and quality of life. There is significant need for education and navigational support around insurance, as well as logistical support for lodging and transportation. A coordinated approach to addressing access barriers is critical for people with CF who are considering or have had a lung transplant. There is increasing recognition of the important role that pharmacists play in outpatient cystic fibrosis (CF) care. Recently, Yale's Adult CF Program received funding from the CF Foundation to introduce a pharmacist into our outpatient clinic, which serves 84 adult patients (age 20 to 81 years). To assess patient's perceptions and needs for outpatient pharmacy support, a survey was created and distributed in the clinic. Based upon survey results, we are designing patient education and intervention strategies.

Methods: A survey was developed to assess medication knowledge, medication access issues, and perceptions of the role of a pharmacist within CF clinic. Patients were given the survey during their CF outpatient clinic visit.

Results: To date, 58 of 84 (70%) patients have completed the survey. None have declined. For medication knowledge, 91% of patients reported that they had knowledge of what a majority of their medications are being used for. Conversely, 53% reported conversations with their provider about medication side effects but 47% reported no such discussion. For medication access, 24 patients reported that they had missed taking medications due to pharmacy related problems. When such problems occurred, 55% were resolved within 1 week. However, 14% of patients reported issues could take up to a month to resolve. The final question addressed patient perceptions of how a pharmacist would be most useful to them as part of the CF care team (see Figure) .

Conclusion: To date, the majority of patients have completed this survey, which identified patient knowledge, medication access issues, and patient perceptions about pharmacist within the clinic. While the majority of patients are familiar with medication indications, many do not recall discussion(s) about side effects. Importantly, 41% of patients reported missing doses of their medications due to a pharmacy related issue. In addition, patients identified assistance with refills and insurance as a significant area that they anticipate clinic pharmacy will assist with. Given these results, we are implementing patient-specific education and assessing the pharmacist role to assist with pharmacy issues to improve medication adherence.

Conclusions: 1) Provider-mediated enrollment is an effective way to encourage CF caregivers to register with an HROs platform; 2) Over 6 months, Folia demonstrated high rates of continued engagement of CF caregivers. We also observed significant network effects among caregivers, resulting in dozens of registrations at locations outside of ME.

Further Work: As of May 2018, we have initiated pilots at Dell MC in Austin and University of Vermont MC in Burlington to test Folia with adult patients and to discover how to incorporate Folia data into clinical decision-making.

5 ; Ting, A. 13

Cystic Fibrosis Newborn Screening Consortium 1 1. Wadsworth Center, NYSDOH

Children's Hosp. of Buffalo

Kings College Hospital

United Kingdom Lung function is a significant predictor of lumbar spine z-score after adjusting for age, gender, and other covariates

Children's Hosp. of Philadelphia

USA to have an AD (17% v. 13%, p=0.593) and have a MPOA (39% v. 23%, p=0.160) but this did not reach statistical significance. Conclusions: Patients with CF who received a LT were less likely to have had an EOL conversation with a pulmonologist and to receive PC than CF patients who died from end stage CF without a LT. Further areas of research should explore how and when to discuss EOL and the timing of PC consultation for patients who are LT candidates. 721 PSYCHOLOGICAL WELL-BEING OF CHILDREN WITH CF AND PARENTS Puckey

1,4 1. Paediatrics, Royal Brompton and Harefield NHS Foundation Trust

United Kingdom providers, and payers. The major purpose of this study was to examine the frequency of these barriers and across adolescents and adults. Methods: Adolescents and adults from 12 pediatric (PEDS) and 9 adult (ADULT) US CF centers, participating in the Success with Therapies Research Consortium (STRC), completed a demographic survey and the CF-ABQ (CF Adherence Barriers Questionnaire) modules on an Ipad. Patients "clicked" on a pictorial icon representing the treatments they are currently prescribed and then rated the barriers for that specific treatment on: difficulty (1= "not at all hard

Most frequent barriers to enzymes across ages were: 1) Taking enzymes after eating (PEDS M=2.69, ADULT M= 2.68 out of 5); 2) Forgetting to bring enzymes with you (PEDS M = 2.54, ADULT M =2.37 out of 5); 3) Not taking enzymes in front of others

Acknowledgments: This work is supported by the NIH, CF Foundation, Trucode Gene Repair, Inc., and the Galpert Family Fund. Acknowledgments: This work was supported by the Italian Cystic Fibrosis Foundation (Fondazione per la Ricerca sulla Fibrosi Cistica) and by the Italian Ministry of Health (Ricerca Corrente). Acknowledgments: Funding from the NIH (R01-AI091699, Acknowledgments: Supported by CFRI, CFF and CFFT. Acknowledgments: Supported by the French Association "Vaincre La Mucoviscidose" and "SATT Lutech." Acknowledgment: Supported by CFF. Acknowledgements: NIH P01 HL51670, CFF, Roy J. Carver Trust, Acknowledgements: We are grateful to Janet Allen and Rebecca Acknowledgments: Support by CFF Therapeutics (US), TPCH Foun-Acknowledgment: Supported by the CF Foundation. Acknowledgment: Supported by Novartis. Acknowledgments: The study was supported NIH (R00HL111217, R01HL131012 and R44HL123299) and CFF (CLEVEL16A0). Acknowledgments: Supported through a CF Foundation Clinical Acknowledgment: Supported by Gilead Sciences UK & Ireland Ltd.Within and between group effects (PA v BCC) from 1 year pre-AZ-LI to 1 year post-AZLI Acknowledgment: Supported by the Cystic Fibrosis Foundation Second Year Clinical Fellowship.Introduction: Anxiety and depression are common among individuals with CF, associated with decreased lung function (J Cyst Fibros. 2008;7:S107), health-related quality of life (J Cyst Fibros. 2008;7:581-4), and increased healthcare costs (Pediatr Pulmonol. 2014;49:1177-81). However, research is lacking regarding effective and feasible interventions for these concerns among CF populations. This study concludes our pilot study that suggested Acceptance and Commitment Therapy (ACT; Behav Ther: 2004;35:639-65) is a feasible treatment, including via telehealth; reducing anxiety, depression, and cognitive fusion among CF patients (Pediatr Pulmonol. 2017;52(S47):476). Telehealth is ideal for CF patients, given exposure precautions, exacerbations, and distance from clinic. ACT differs from traditional cognitive behavioral therapy by emphasizing acceptance of painful inner experiences, thereby reducing the need for behaviors such as medication nonadherence that function to avoid anxiety and depression associated with thinking about having CF. We hypothesize that our ACT with CF protocol (Bennett DS, et al. Drexel Univ College of Medicine. 2016) will also improve lung function among people with CF.Methods: Adults with CF and elevated anxiety and/or depressive symptoms (screened via GAD-7 and PHQ-9) participated in 6 weekly manualized ACT sessions, delivered by a licensed psychologist (CVO'H) and graduate student trainees. Patients selected treatment in person or via telehealth (using HIPAA-compliant WebEx). Participants completed psychometric measures of depression (PHQ-9, BDI-II), anxiety (GAD-7, BAI) and cognitive fusion (CFQ13) at baseline, after 3 and 6 weeks of ACT, and 3 months post-treatment. Lung function (FEV 1 /FVC) was Acknowledgment: Support by CFF. . We have previously produced a tool to explore the level of understanding of a child about CF ("Helping Others Learn about CF," Forest 2006 M1359/WBR/May2006). We report a patient-and family-driven QI initiative aiming to create a user friendly tool to assist sharing information about CF by an adult (usually parent) with a child.

Methods: A literature review about information sharing by adults with children across any health condition was undertaken. This included research on the development of cognitive processing and what may affect this (Bluebond-Langer, et al. 2001 "Psychosocial Aspects of CF"' Arnold Publishers), taking into account social, spiritual, cultural and any other factors which may influence this. We sought comments from service users, adults and older CF paediatric patients and peers including the UK Psychosocial Professionals group (UKPP-CF).Results: The resultant leaflet covers advice about talking to a child about CF, accounting for the family's culture of sharing information, and also provides further information about accessing other resources. Areas include: Why? -including encouraging honesty between patient and carers, promotes adherence to treatment. Who? -ensuring consistency amongst carers; When? -a process rather than a single event: What? -areas of the body that are affected, how did they get CF, what needs to be done every day to keep well, genes, fertility. How? -answering the child's questions as they arise in an age-appropriate way, opportunities of hospitalisation or outpatient clinic visits, encouraging learning about how the healthy body works and how a CF body can change this; honesty and what to do if the child asks a very difficult question. Other areas include pacing discussions about CF, checking understanding, and advice that the parent looks after their own emotional well-being. This has been distributed to parents; feedback from over 100 parents and colleagues has been positive.Discussion: This quality improvement initiative aimed to collate research, service user and colleague views to produce a leaflet to offer guidance to parents about informing children about CF. The leaflet has received positive feedback and is now used as an integral tool by our CF team. We hope that this will be useful in other centres to aid information sharing with children with CF. Furthermore the concepts could be taken and utilised for information givers to children about other diseases/conditions. (Quittner AL, et al. Thorax. 2014;69:1090-7). CF Mental Health Guidelines recommend screening for depression and anxiety in individuals with CF using the Patient Health Questionnaire-9 (PHQ-9) and the Generalized Anxiety Disorder-7 scale (GAD-7) (Quittner AL, et al. Thorax. 2016;71:26-34). Psychological symptoms have been associated with worse Health-Related Quality of Life HRQoL (Goldbeck L, et al. Cochrane Database Syst Rev. 2014;6:CD003148). It is well known that it is important to evaluate the effects of the disease and its treatments on daily functioning in several domains (e.g., respiratory symptoms, physical functioning, and vitality). To date, the relationship between psychological symptoms assessed considering the CF Mental Health Guidelines and Health-Related Quality of Life domains of Italian CF patients has never been systematically explored. This study investigates the relation between depression/anxiety symptoms and HRQoL domains in CF patients.

Methods: We recruited 82 consecutive CF patients (F/M=47/35, mean age (SD) [range]=26(9) [14-54]). We excluded patients with other chronic diseases. All patients completed: a) PHQ-9 for depression symptoms; b) GAD-7 for anxiety symptoms; c) Cystic Fibrosis Questionnaire-Revised Teen/Adult (CFQ-R + 14). Pearson's correlation and score range analyses Acknowledgments: Support by CFF grants BARRIERS16PE0, SAWICK14PE1 and RIEKERT15PE0.

Lowery, E.M. 1 ; Afshar, M. 2, 1 ; West, N.E. 5 ; Kovacs, E.J. 3 ; Smith, B. 4 ; Joyce, C. 2 1. Medicine, Loyola Univ Chicago, Maywood, IL, USA; 2. Public Health, Loyola Univ Chicago, Maywood, IL, USA; 3. Surgery, Univ of Colorado Denver Anschutz Medical Campus, Aurora, CO, USA; 4. Psychiatry, Univ at Buffalo School of Medicine, Buffalo, NY, USA; 5. Medicine, Johns Hopkins Univ, Baltimore, MD, USA Introduction: Excessive alcohol use (EAU), a harmful pattern of drinking that includes both binge drinking and heavy use, occurs in 25% (binge) and 7% (heavy use) of the US population, respectively (NSDUH; SAMHSA 2015). Little is known about alcohol use in individuals with cystic fibrosis (CF). The objective of this investigation is to examine alcohol consumption patterns in individuals with CF using a self-report survey.Methods: Individuals with CF, 18 years of age or older, were recruited for participation in the study through social media and internet-based platforms. The survey included 36 questions that collected information about demographics, alcohol use, CF-related comorbidities, and self-reported health outcomes.Results: 1135 individuals initiated a survey with 952 meeting inclusion criteria and completing the survey. Of those included in the analysis cohort, N=729 (77%) currently consume alcohol, N=171 (18%) previously consumed alcohol, and N=52 (5%) never consumed alcohol. Amongst the people with CF who currently consume alcohol, N=174 (24%) met the requirements for heavy alcohol consumption and N=354 (49%) for binge-pattern consumption, with a combined N=371 (53%) meeting criteria for EAU. The figure shows reported typical alcohol use patterns in individuals with CF. Almost 30% of current drinkers experienced symptoms of harmful alcohol use including black-outs and guilt after drinking. Amongst the respondents who met criteria for EAU, 7% required oxygen, 6% had a lung transplant, 10% had CF liver disease and 32% had CF-related diabetes. Those with EAU reported more hospitalizations than those without EAU [244 (62%) vs 182 (54%), p=0.034]. Characteristics associated with EAU after multivariable adjustment included younger age, unmarried status, male gender and age ≤15 years at time of first alcoholic drink, which increased the odds of developing EAU by almost 6-fold [5.43 (2.97-9.94), p<0.01].Conclusion: EAU is occurring at a much higher proportion in individuals with CF, and a substantial proportion of CF individuals with EAU have multiple co-morbidities, where abstinence is recommended. Screening, brief Acknowledgment: Supported by CFFT grant SMITH14Q10. 

Mississippi is a state that has one of the highest BMIs in the general US population. Center-specific data revealed that CF patients at the University of Mississippi Medical Center did not follow the national trend nor did they meet the CF Foundation's BMI percentile (%ile) recommendation of >50%ile. In 2015, registry data for patients ages 2-19 revealed an average BMI of 38.3%. As BMI >50%ile is correlated with better lung function and increased life span, our goal was to increase BMI %ile. The target population for the intervention was the 2-10 y old age group as they might be most amenable to change and set a strong foundation as they enter adolescence and adulthood.Methods: A multidisciplinary team was assembled including CF assistant center director, center coordinator, social worker, nurses, respiratory therapists, dietitian, GI and Endocrine physicians and parents of a patient with CF. The team met weekly with a quality coach and conducted an assessment of our patients' BMIs by reviewing center-specific data for 2014 and 2015. A fishbone diagram of factors contributing to low BMI was developed. Global and specific aims for improvement were written. The team developed a list of change ideas. Clinical, functional, costs and satisfaction outcomes using our Clinical Value Compass were examined and ideas for change were rapidly evaluated using a Plan-Do-Study-Act (PDSA) model.Interventions: Using the PDSA model a number of change ideas were tested including:1. Increasing the number of patients seen by our dietitian at each clinic visit. Prior to starting our LLC2 cycle, we had a part-time dietitian who saw 20-60% of our CF patients each month. Soon after starting our project, a full-time dietitian was hired and "patients seen" increased to 93-100% of patients/month.2. Increasing the number of patients using nutritional supplements. Prior to the FUN LLC2, only 2% of patients in our target age group were receiving financial assistance for nutritional supplements. After submitting prescriptions to their insurance company and utilizing Healthwell program as a payer of last resort, the number using supplements increased to 75% in our target population.3. Optimizing pancreatic enzyme dosage in the target population. Average dose increased from 8604 lU lipase/kg/day to 9168 lU lipase/kg/day. 4 . Utilizing a nutrition tracker to keep track of all information related to our patients. 5 . Creating a nutrition algorithm to determine which patients needed closer follow-up and further interventions.6. Creating and distributing nutritional quizzes to assess patients' and families' knowledge.7. Hosting multiple local Education days around the state with greater attendance, family input and satisfaction.Results: Prior to the start of the LLC2 cycle 18 patients age 2-10 years were above the 50%ile for BMI and 18 patients below the 50%ile. At the end of the cycle in October 2017, 21 patients were above the 50%ile and 14 were below the 50%ile. Of the 14 patients below the 50%ile 7 showed increased BMIs by at least 5%.Conclusion: This project resulted in increased focus on BMI amongst all disciplines and improved BMIs in our target population. Acknowledgment: This project was supported by CFF though the Fundamentals Learning Leadership Collaborative 2.

Wang, J. 1 ; Benitez, D. 2 ; Avila, K.J. 1 ; Beringer, P. 1 ; Rao, A.P. 3 1. School of Pharmacy, USC, Los Angeles, CA, USA; 2. Keck Medical Center of USC, Los Angeles, CA, USA; 3. Pulmonary Critical Care & Sleep Medicine, Keck School of Medicine, Los Angeles, CA, USA Introduction: The Cystic Fibrosis Learning Network (CFLN) is a collaborative network created to improve health and care through accelerated quality improvement (QI) initiatives. In 2016, a pediatric center developed a systematic approach to acute declines in lung function utilizing an algorithm (Schechter MS, et al. Pediatr Pulmonol. 2016; 51(S45) :402[abstr], Pediatr Pulmonol. 2017;52(S47):416[abstr]). As part of the CFLN, our CF care center selected a QI initiative on population management of lung function decline. An algorithm was created to utilize a similar systematic approach unique to our adult population. Discussion and development of an algorithm started in April 2017 with implementation in November 2017. The purpose of this initiative was to measure improvement in pulmonary health outcomes after implementation of the algorithm.Methods: Patients were stratified into the algorithm based on FEV1 decline at the clinic visit. Relative declines in FEV1 are calculated during clinic based on their best FEV1 in the past year. Each arm of the algorithm is broken down into the following acute FEV1 declines: <5%, 5-10%, >10%. Process measures utilized during the QI process include time to next follow-up clinic visit within 6 weeks, usage of antibiotics (intravenous (IV) or oral), and distribution of interventions. Study outcomes include percent of patients returning to baseline and mean FEV1 over the study period.Results: In patients with acute FEV1 declines ≤-5%, 28/41 (68.3%) were brought back within 6 weeks at the start of the QI process in April 2017. From May until October 2017, patients were brought back a median of 52.4% of the time. After fully implementing the algorithm, patients were brought back a median of 69.4% between November 2017 to April 2018. Antibiotic usage increased after implementing the algorithm (November 2016-March 2017: 29% vs November 2017-February 2018: 44%). The most successful intervention for returning patients back to baseline was medication optimization or education (59.3%). Admission for IV antibiotics was successful in bringing patients back to baseline 50% of the time (n=18). Average FEV1 improved throughout the QI process, starting at 66% in April 2017 to a peak of 70.1% in November 2017 and has remained stable.Conclusions: Currently there is no clear definition for a pulmonary exacerbation; providers look at a variety of subjective and objective measures. The current algorithm utilizes relative decline in FEV1 and further stratifies patients into a treatment arm developed through QI processes in collaboration with the care team. Antibiotic usage has increased, however the rate of returning patients back to baseline in this study is not higher than previously reported data. The most successful intervention was medication optimization and education, when bringing patients back within 6 weeks. Since implementation of the algorithm, there has been an increase in mean FEV1 at our center. This demonstrates the utility of developing a systematic approach to improve outcomes in CF care and that a close follow-up may identify declines in lung function before progression to a pulmonary exacerbation.Conclusions: When children with CF are distinguished from among other symptomatic children referred to our clinic, baseline esophageal impedance tends to be lower for the CF cohort at both positions, but only significantly so in the proximal esophagus when the children with CF have GERD. Also, children with CF are apparently more likely to have abnormal PBI, compared to other symptomatic children without CF. Larger, multicenter studies are needed to further examine mucosal integrity in the proximal esophagus of children with CF and GERD.

Pals, E. 1 ; Verkade, H. 1 ; Gulmans, V. 2 ; Koning, B.d. 3 ; Koot, B. 4 ; Meij, T.d. 5 ; Hendriks, D. 6 ; Gierenz, N. 7 ; Vreugdenhil, A. 8 ; Houwen, R. 9 ; Bodewes, F. 1 1. University Medical Center Groningen, Groningen, Netherlands; 2. Dutch Cystic Fibrosis Foundation, Baarn, Netherlands; 3. Erasmus University Medical Center, Rotterdam, Netherlands; 4. Academic Medical Center, Amsterdam, Netherlands; 5. VU University Medical Center, Amsterdam, Netherlands; 6. Haga Hospital, The Hague, Netherlands; 7. Radboud University Medical Center, Nijmegen, Netherlands; 8. Maastricht University Medical Center, Maastricht, Netherlands; 9. University Medical Center Utrecht, Utrecht, Netherlands Introduction: Between 2 and 10% of patients with cystic fibrosis develop CF-related cirrhosis (CFC). The lifetime impact of CFC on mortality has so far not been quantified. The availability of a national registry on CF care and relevant outcomes offers a unique opportunity to assess the mortality risk of CF patients with or without CFC. We aimed to determine the role of CFC on CF patient survival based on national data of the Netherlands.Methods: We identified all CFC patients in the Netherlands who had been alive at 1-1-2009 , based on the review of medical records. CFC was defined either by ultrasound demonstration of the combination of inhomogeneous liver parenchyma and splenomegaly, or by the histological diagnosis. A control group was obtained from the national CF registry, consisting of 980 CF patients without CFC. We compared the groups with regard to survival and age at death, in the period 1-1-2009 to 31-12-2014. In case of mortality in either of the two groups, the primary cause of death was retrieved from the medical records.Results: We identified 103 CF patients with CFC and 980 CF patients without CFC at the start of the study period. In the subsequent 6 years, compared to the control group, the mortality rate was significantly higher in CFC patients (22/103, 21%) than in non-CFC controls (63/970, 6%; P<0.01). The median age of death was 10 years lower in CFC than in non-CFC patients (27 vs. 37 years, respectively; P=0.01). A significantly higher proportion of CFC patients died before the age of 25 years, compared to the control group (45% vs. 27%; P<0.01). At the start of the study period, CFC and non-CFC patients did not significantly differ in pulmonary function tests (forced expiratory volume at 1 s; % of predicted forced vital capacity), nor in body weight or body mass index, indicating similar clinical conditions at that moment. In the diseased CFC patients, the cause of death had been attributed to pulmonary disease in 68% of the cases, and to chronic or acute-on-chronic liver failure in 18%, several times subsequent to an elective surgical procedure.Conclusion: Our national data compellingly demonstrate that CFC is a critical risk factor for early mortality in CF patients and suggest that CFC has a negative effect on life expectancy of 10 years. The mortality seems either due to a strongly accelerated decline in pulmonary function or to liver failure, for example after an elective surgical procedure. A more detailed identification of CFC patients who have the highest risk of mortality seems now indicated for the consideration of preemptive liver transplantation.

Beattie, S.; Abdel-Rahman, S.; Snell, Y.A. The Children's Mercy Hospitals and Clinics, Kansas City, MO, USA Background: It is well documented that optimal growth and nutrition is related to positive outcomes in persons with cystic fibrosis (CF). Specifically, CF guidelines state patients with body mass index (BMI) ≥50th percentile support better outcomes in relation to their FEV1 status (Stallings V, et al. J Am Diet Assoc. 2008; 108:832-9) . Children with CF are vulnerable to malnutrition for a variety of reasons including: increased daily nutritional needs, malabsorption/maldigestion and recurrent/chronic respiratory infections. A recent article reported concerns about BMI percentile being a poor predictor of nutritional status in children with CF (Konstan M, et al. J Cyst Fibros. 2017; 16:158-60) . Pappas and coworkers found that mid-upper arm circumference (MUAC) z-score classified more CF patients as malnourished compared with BMI z-score based on recommended malnutrition classifications (Pediatr Pulmonol. 2016; 51(S45):435[abstract] ). We conducted this study to examine the relationship between BMI z-score (zBMI), MUAC z-score (zMUAC), dietitian-derived malnutrition classifications and FEV1 in our CF patients.Methods: Age, weight, height, BMI, MUAC and FEV1 were collected for patients aged 4-19 years at CF standard of care clinic visits and/or during hospitalizations during 2015-2018. Based on the Academy of Nutrition and Dietetics (AND) and American Society for Parenteral and Enteral Nutrition (ASPEN) criteria for malnutrition classifications, we looked at concordance using the z-score threshold for malnutrition classifications among zBMI, zMUAC, weight z-score (zWt), height z-score (zHt) and percent FEV1. Standard descriptive statistics were used to analyze data.Results: To date, 322 observations (49% male, median age 12.2 years) were obtained in 209 unique CF patients. Though there is a significant relationship between nutritional status and FEV1 across the CF population (Table) , linear regression demonstrates that zBMI accounts for only 31% of the variability in FEV1 (rsq=0.310, p<0.01), compared with 25% for zMUAC (rsq=0.252, p<0.01), 31% for zWt (rsq=0.357, p<0.01) and 9% for zHt (rsq=0.090, p<0.01).Conclusions: Anthropometric measures alone are not robust predictors of pulmonary lung function as measured by FEV1 due to the broadness of the data given other variables. No single anthropometric measure shows a strong correlation with lung function. A cut-off point for using BMI ≥50th percentile fails to identify FEV1 status.

Van Biervliet, S. 1 ; Van Braeckel, E. 2 ; De Baets, F. 3 ; Van daele, S. 3 ; Van Meerhaeghe, S. 4 ; Declercq, D. Gent University Hospital, Gent, Belgium; 2. Pneumology, Gent University Hospital, Gent, Belgium; 3. Pediatric Pneumology, Gent University Hospital, Gent, Belgium; 4. CF Centre, Gent University Hospital, Gent, Belgium Background: Body mass index is a clinical outcome parameter for the evaluation of the nutritional status in CF. However, recent studies indicate that lean body mass (LBM) might be a stronger indicator. In the nutritional management of CF, tube feeding (TF) is often used to improve the nutritional status. The influence of TF on body composition was not yet evaluated.

Tarn, V.E. 1,2 ; Self, S. 3 ; Anderson, V. 3 ; Oates, G.R. 1 ; Gutierrez, H. 1 1. Pediatrics, UAB, Birmingham, AL, USA; 2. Pediatric Pulmonary Center, Birmingham, AL, USA; 3. Pediatric Pulmonary Division, Children's of AL/UAB CF Center, Birmingham, AL, USA Introduction: Despite intense nutritional interventions, including quarterly visits with dietitians, optimized pancreatic replacement therapy, oral supplements, G-tubes, and behavioral therapy, some patients with CF struggle with poor growth. As part of a pilot project for pediatric CF patients at significant nutritional risk with BMI < 25 percentile, we tested the effectiveness of delivering high calorie, high protein (HCHP) snack boxes to the patient's home in an effort to improve nutritional outcomes.Methods: At enrollment, the patient/family received a digital scale and a 3-month weight gain goal. Instructions on how to record weekly weight, and information sheets on caloric intake and recipe ideas were provided. Patients/families selected preferred HCHP snacks from a list of available items. Snack boxes with a 2-week supply of HCHP snacks were ordered by the CF dietitian and shipped from an online distributor to the patient's home every 2 weeks over a 3-month period. Follow-up telephone calls at 2-week intervals were made to record weekly weight from the home scale and to note changes in snack box preferences. Weight, BMI%, and BMI z-scores were recorded from clinic visits at four time points: 3 months pre-intervention, intervention (baseline), and at 3 and 6 months postintervention.Results: A total of 9 patients (6 male, 3 female), ages 12-20 years, were enrolled. During the 3-month intervention period, each patient received 7 snack-box shipments, each costing approximately $50, including delivery fee. Three patients were excluded from data analysis (1 death, 1 delivery issue to college campus, and 1 lost to follow-up). The Table presents the nutritional characteristics of participants (N=6, ages 12-17) before and after the intervention. Although the BMI% and BMI z-scores of participants declined due to an increase in linear growth, patient weight increased. The 3-month weight change at the end of the intervention was 1.1 kg, compared were divided into 4 consecutive age groups: first 13 C breath test on babies less than 5 months of age, second breath test between 6-9 months of age, third breath test between 9-12 months and fourth breath test between 9-24 months and with a different test meal (of yoghurt).Results: A percentage of between 39% (9-24 months old) to 56% (9-12 months old) of individual results were found lower compared to the minimum reference point in all age groups. Intra-individual variation of the 13 CO 2 response, in which no rate-limiting variation in pancreatic exocrine function was expected, was also assessed and showed that out of the 53 longitudinally assessed babies with three to four consecutive MTG breath tests, only 6 babies (3%) had consistent results above the minimum 13 C cut-off point during the three to four breath test assessments with NDIRS technique.Conclusion: The 13 C MTG breath test with NDIRS technique is too insensitive to determine pancreatic function status in babies within the first two years of life and therefore the faecal fat balance test remains the gold standard test in assessing CF patient's pancreatic function in the first two years of life. Introduction: Risk factors for Clostridium difficile infection (CDI), a common hospital-acquired infection, include recent antimicrobial exposure, hospitalization and advanced age. Patients with a chronic disease such as cystic fibrosis (CF) require multiple courses of antimicrobials, and potentially lung transplantation, meaning they are at high risk of acquiring C. difficile or developing CDI. Recently, there have been cases of severe CDI in CF patients reported in Australia. The objectives of this project were to determine the prevalence of CDI or C. difficile colonisation in CF patients in Western Australia, and to study the molecular epidemiology of C. difficile strains from these patients.

Methods: From late June to December 2017, 24 CF patients who came for treatment or consultation at the Department of Respiratory Medicine at Sir Charles Gairdner Hospital (SCGH) were recruited with informed consent. Faecal samples were collected on this initial visit, and after subsequent visits, and screened for C. difficile and toxins using EIA. All samples were also cultured for C. difficile and any isolates ribotyped and toxin gene profiled. Before the study started, 177 environmental samples were collected from the CF outpatient clinic and inpatient ward with MWE Polywipe sponges which were also cultured for C. difficile.Results: The prevalence of C. difficile colonisation in CF patients on the initial visit was 46% (11/24), increasing to 54% through the period of study as two patients became positive with C. difficile after antimicrobial treatment. One patient was persistently positive throughout the study. None of these three patients was symptomatic at the time of sample collection although one required CDI therapy within 6 weeks. The prevalence of C. difficile in the environment was 3% (6/177); 1/90 samples from the outpatient clinic and 5/87 samples from the ward. PCR ribotyping revealed 8 distinct ribotypes in patients, with the nontoxigenic strain RT 039 the most common (8 isolates) . RT 014/020 (3), RT, 046 (2), RT 103 (1), RT 012 (1), and RT QX 360 (1) were all toxigenic. The nontoxigenic RT 010 comprised 5 of 6 isolates found in the environment, however, this strain was not isolated from patients. Routine surveillance data for C. difficile in non-CF patients revealed that RT 014/020 was the most common ribotype found in WA during the study period. All toxigenic ribotypes found in CF patients were found in non-CF patients.Discussion: About half the C. difficile strains found in CF patients were toxigenic but no toxin was detected in the faeces of any culture-positive patient, suggesting possible protective effects of the CF intestinal milieu or microbiome. These asymptomatic carriers are however a potential source of C. difficile, and infection prevention and control procedures should be applied appropriately. Further studies should focus on the effect of the CF intestinal microbiome, CFTR mutations and CFTR modulator therapy on the epidemiology of CDI and C. difficile colonisation in CF patients. 1 1. Ghent University Hospital, Ghent, Belgium; 2. European Cystic Fibrosis Nutrition Group, Ghent, Belgium; 3. Universitair Ziekenhuis Brussel, Brussels, Belgium Introduction: Nutritional therapy is one of the cornerstones in CF therapy due to its strong relation with longevity. Although tube feeding (TF) can be useful to improve the nutritional status, its sustained beneficial effect on long term is debatable.

Objective: To evaluate the long-term effect of TF on BMI, height (H); weight and pulmonary function (FEV1%).Methods: All paediatric CF patients followed at 2 Belgian centers, starting TF between 2000 -2014 for a minimal duration of 1 year were included. BMI, weight and height measurements were translated into z-scores based on national reference data. In SPSS ® (version 24) a linear model for repeated measures was used. Results are given as medians and quartiles between brackets.Results: The total number of patients on tube feeding is 25 (11.1% of patients) (5 infants-toddlers (≤5 years), 7 school-aged children (≤12 years), 13 adolescents (13-18 years)). The median time on TF is 5 years (2.5; 6.1). At start the median age was 12.4 years (7-14.6 ). The results are presented in the Table. Initially, BMI improved significantly (p=0.002), but this evolution stabilized two years after the initiation of TF. Patients starting TF during puberty tend to return to their original BMI z-score. H z-score also improved significantly (p=0.007) beginning 2 years after the initiation of TF. However, effect was statistically and clinically significant in infants and school-aged children but less pronounced clinically significant in adolescents. Patients tend to have an improvement in FEV1% in the first years after the start of TF, however, on the long-term this effect is not sustained.Conclusion: The initiation of TF in paediatric CF patients leads to a significant improvement of BMI, height and lung function in the first year. However, when TF is started during puberty the effect on BMI is temporary without improved growth. Adolescents are also significantly more stunted than other age groups at start of TF. This age group is probably a more challenging group to convince or this might reflect older practices. Treatment adherence might also be an issue.were collected to assess substrate utilization (lipid vs. glucose) and changes in energy expenditure.Results: KG diet feeding resulted in marked weight loss in CF mice, while WT mice maintained body weight. Both CF and WT mice maintained normal glycemia and ketone body levels at fasting/refeeding when fed CHOW, however, when fed the KG diet, CF mice became hypoglycemic (WT KG 151.7±7.3 vs. CF KG 64.6±4.5 mg/dL, p<0.0001) with trending high ketone bodies (WT KG 3.3±0.2 vs. CF KG 4.3±0.5 mM, p=0.07) . Respiratory quotient showed CF mice switched completely to lipid utilization, while normally elevated energy expenditure markedly decreased in CF mice after 1 week of KG diet feeding.Conclusions: Overall, removal of dietary carbohydrates uncovered an abnormal pattern of energy utilization. CF mice poorly adapted to lipid and ketone body utilization to maintain energy homeostasis and body weight. Also, CF mice did not maintain blood glucose, likely via inefficient gluconeogenesis. This work highlights the importance of dietary carbohydrates in CF to maintain energy balance and blood glucose levels. Background: There is strong evidence that a higher weight-for-length or BMI are associated with better pulmonary function in pediatric patients with cystic fibrosis (CF). Better nutritional status in early childhood is associated with improved outcomes and survival at 18 years (Zemel BS, et al. J Pediatr. 2000; 137(3) :374-80). For this reason, the CF Foundation nutritional goal for all children with CF is a weight-for-length or BMI ≥ 50th percentile.Objective: To design and implement a nutritional care algorithm using a multidisciplinary quality improvement approach aimed at improving weight-for-length/BMI status in children with CF ages 1-12 years over 3 consecutive visits.Methods: We developed an evidence-based guideline (EBG) with inclusion criteria for CF patients with weight-for-length or BMI <50th percentile, or weight <10th percentile, without a G-tube, who were 1-12 years old. The EBG took a step-wise approach to evaluation and incorporated components of Tier 2 and Tier 3 of the CFF EBG Preschool Nutrition Algorithm. The EBG included in-clinic assessments and questionnaires by nursing and nutrition as well as development of a multidisciplinary plan to address defined symptoms or behaviors of concern labeled "red flags." The red flags encompassed the areas of dietary concerns, GI symptoms, enzyme administration and medication concerns, behavioral concerns, and weight gain velocity. Parameters were defined for expedited referral to GI with a goal of appointments within one month. The EBG was implemented in February 2017 and patients were identified for inclusion at our pediatric CF Center multidisciplinary meeting as part of our quality improvement review.Results: In the first year of implementation, there were 49 patients identified as "at nutritional risk." We tracked patient progress over a period of 3 visits. Overall, the median weight-for length/BMI of the patients identified increased from the 32.54th percentile at visit 1 to the 36.25th percentile at visit 3.Conclusion: About half of the identified patients increased their weight-for-length/BMIs from visit 1 to 2 and visit 2 to 3. Five patients were referred to GI as a result of their nutritional status. The nutrition EBG and our multidisciplinary approach increased the weight-for length/BMI percentile in patients at nutritional risk. Our median starting weight-for length/BMI at the 32.54th percentile indicated the need for a revision to our inclusion criteria. We have amended the EBG to include CF patients with weight-for-length or BMI < 35th percentile. We will continue to utilize our assessment tools and track the weight-for length/BMI of these patients over consecutive visits.

Smieja, M. 1 ; Fuentes, D. 2 ; Chang, C. 2 ; Das, S. 3 ; Rennig, A. 3 ; Sipos, T. 3 ; Schindler, T. 4 1. Children's Minnesota, Minneapolis, MN, USA; 2. Chiesi USA, Inc., Cary, NC, USA; 3. Digestive Care Inc., Bethlehem, PA, USA; 4. University Hospitals Cleveland Medical Center, Cleveland, OH, USA Background: Patients with exocrine pancreatic insufficiency (EPI) require pancreatic enzyme replacement therapy (PERT) to meet their nutritional needs. Some patients with EPI may also need a gastrostomy tube (G-tube) to maintain optimal nutritional status, and administering PERT via a G-tube is often necessary and/or preferred for these patients. Pancrelipase (Pertzye) 4,000 United States Pharmacopeia (USP) lipase units capsules are a PERT approved for administration via G-tube. These pancrelipase (Pertzye) capsules feature spherically shaped, enteric-coated microspheres with a distinct diameter range that is smaller than those in the product's larger available capsule sizes. The aim of this study was to examine G-tube administration of the pancrelipase (Pertzye) capsule contents delivered via various delivery configurations and vehicles.Methods: This in vitro study assessed the feasibility, push force, and enteric-coated microsphere stability and integrity of pancrelipase (Pertzye) 4,000 USP lipase units capsule contents delivered via G-tube. Capsules were opened and their contents sprinkled onto one of 3 brands of applesauce (pH <4.0; Mott's ® , Musselman's ® , and Gerber ® Stage 2 brand). Mixtures were stirred to achieve uniform microsphere suspension, transferred into a syringe (size 10, 12, or 35 mL), and delivered into a G-tube (size 14 French; Kimberly Clark MIC-KEY ® or Covidien Nutriport ® brand) using steady pressure over 10 to 12 seconds. Feasibility and push force were assessed using developed methods, while microsphere stability/integrity was assessed using standard methods.Results: G-tube administration of pancrelipase (Pertzye) 4,000 USP lipase units capsule contents was feasible, as assessed by the successful delivery of microspheres through multiple compatible syringe/G-tube configurations. The 35-mL syringe had the greatest compatibility with the different G-tube brands examined. With a single exception, the contents of 2 pancrelipase (Pertzye) capsules suspended in 10 mL of any brand of applesauce were delivered through each compatible syringe/G-tube configuration tested without obstruction. For the 12-mL syringe/MIC-KEY G-tube configuration, some clogging was noted using Gerber Stage 2 applesauce; reduction of the capsule contents by half (ie, suspension of 1 capsule in 10 mL of applesauce) resulted in successful delivery in this configuration. Addition of suspended, enteric-coated microspheres to applesauce did not meaningfully affect the push force required for delivery via G-tube. Microsphere stability and integrity, as assessed by mean lipase activity and dissolution lipase activity, were maintained at ≥95% and ≥89% of initial levels, respectively, exceeding prespecified acceptance tolerances.Conclusions: Delivery of pancrelipase (Pertzye) 4,000 USP lipase units capsule contents suspended in applesauce via G-tube is feasible and maintains microsphere stability and integrity. These findings support the FDA's approval of the updated pancrelipase (Pertzye) product label to include the G-tube administration route.

Carr, V. 1 ; Hodge, D. 2 ; Green, D.M. 2 1. Clinical Nutrition, Kate Farms, Carpinteria, CA, USA; 2. Johns Hopkins All Childrens, St. Petersburg, FL, USA Background: Children with cystic fibrosis (CF) are at risk of developing a vitamin D insufficiency, defined as a 25-hydroxy-vitamin D (25OHD) serum level <30 ng/mL. Vitamin D insufficiency has been linked to increased inflammatory markers, poor bone health, and increased pulmonary exacerbations in those with CF. Screening and treatment guidelines were published by the CF Foundation in 2012. Since that time, our clinic has continued to observe poor vitamin D levels despite additional supplementation per these guidelines. As part of a quality improvement Methods: A retrospective chart review of patients followed at the Children's Hospital of Philadelphia CF Center who were prescribed insulin prior to nocturnal gastrostomy tube feeds was conducted. Individuals with CF liver disease (CFLD), Type 1 diabetes, or treated with daytime insulin were excluded. Weight, height, BMI, one-hour and two-hour glucose from oral glucose tolerance test (OGTT), and FEV 1 %-predicted were recorded from CF Center visits 6-12 months pre-and post-insulin initiation; subjects were receiving nocturnal tube feedings throughout this study period. Weight-Z, Height-Z and BMI-Z were calculated using CDC 2000 reference data.Data are presented as median [min; max]. Wilcoxon signed-rank tests were used to compare maximum BMI-Z, Weight-Z, Height-Z, and FEV1 percent predicted pre-and post-insulin initiation.Results: Of the 60 patients on nocturnal tube feeds, 9 (7M/2F) patients met inclusion criteria. None had CF-related diabetes (CFRD) by OGTT. Age at time of insulin initiation was 10 years [7; 13.6 ]. One-hour OGTT glucose was 165 mg/dL [126; 333] and two-hour OGTT glucose was 107 mg/dL [82; 195] . BMI-Z post-insulin (1.07 [-0.098; 1.29 ]) tended to be higher than pre-insulin (0.76 [-0.82; 1.39]), p = 0.09. Weight-Z post-insulin (-0.073 [-1.14; 0.74]) tended to be higher than pre-insulin (0.28 [-1.05; 0.72]), p = 0.07. Height-Z post-insulin (-1.63 [-2.15 ; 0.91]) was not different than pre-insulin (-1.46 [-2.17 ; 0.95]), p = 0.11. FEV1 % predicted post-insulin (99 [74; 120]) was not different than pre-insulin (102 [83; 118], p = 0.91.Discussion: Six to twelve months of treatment with insulin prior to overnight enteral feeds in nondiabetic patients with CF tended to be associated with improved nutritional status. While this study did not directly assess nocturnal hyperglycemia during tube feeds, these data are consistent with the hypothesis that insulin therapy may address such undetected glucose elevations during overnight enteral feeds. Prospective studies should focus on the contribution of overnight enteral feeds to perturbed glucose homeostasis as well as the role of empiric insulin initiation in optimizing nutritional status and preserving β-cell function in individuals receiving overnight enteral feeds.

Tommerdahl, K. 1 ; Brinton, J. 2 ; Vigers, T.B. 1 ; Green, M. 1 ; Nadeau, K. 1 ; Zeitler, P. 1 ; Chan, C.L. 1 1. Endocrinology, Children's Hospital Colorado, Aurora, CO, USA; 2. Biostatistics, Colorado School of Public Health, Aurora, CO, USA Background: In youth with cystic fibrosis (CF), current guidelines define abnormal glycemia (AG) during an oral glucose tolerance test (OGTT) as follows: fasting plasma glucose ≥100 mg/dL, 1 hour glucose (1hG) ≥200 mg/dL, and/or 2 hour glucose (2hG) ≥140 mg/dL. In type 2 diabetes, the oral disposition index (oDI) is a reliable measure of insulin secretion relative to sensitivity and is a strong predictor of diabetes progression. Additionally, oDI has been described to be abnormal in the early development of cystic fibrosis-related diabetes (CFRD). Morphologic characteristics of the OGTT glucose curve are reported to identify risk of prediabetes and type 2 diabetes in obese youth. Whether such OGTT characteristics better predict oDI abnormalities in CF when compared to traditional OGTT criteria has not been studied. Our aim was to compare 3 alternate methods of classifying OGTT glucose results in CF youth: 1. shape of the curve (biphasic vs monophasic), 2. time to glucose peak (<30 minutes vs ≥30 minutes), and 3. 1hG <155 mg/dL vs ≥155 mg/dL, to traditional OGTT criteria defining AG to determine which best identifies lower oDI.Methods: CF youth 10-18 years underwent a 2-hour 75 g OGTT with glucose and insulin concentrations obtained every 30 minutes. Participants taking insulin were excluded. In addition to traditional OGTT criteria, participants were classified by the 3 alternate methods as AG (monophasic shape of curve, glucose peak at >30 minutes, 1hG ≥155 mg/dL) or normal glycemia (NG) (biphasic shape of curve, glucose peak at <30 minutes, 1hG <155 mg/dL). oDI ((Δ 0-30 minutes insulin/ Δ 0-30 minutes glucose)* 1/fasting insulin) was calculated. Mean oDI and clinical outcomes including BMI, %FEV1, and %FVC were also calculated and compared to predefined morphologic groups with ANOVA. Associations between oDI and clinical outcomes were evaluated with linear regression.Results: N=52 CF youth completed an OGTT: NG, N=11; AG, N=41; mean±SD age 13±4 years; 37% male; BMI z-score 0.0±0.8; FEV1% 88±16.3; FVC% 97±14.8. Peak glucose and insulin concentrations differed significantly between the NG and AG groups defined by traditional criteria (p<0.05); however, the oDI comparison for insulin was not significantly different. In contrast, early vs. late time to peak glucose and 1hG < vs. ≥155 mg/dL showed a significant oDI difference (0.17±0.09 vs. 0.09±0.05, p=0.0302; 0.16±0.09 vs. 0.08±0.05, p=0.0184) while monophasic vs biphasic did not (0.17±0.09 vs. 0.09±0.06, p=0.2047). Neither traditional OGTT nor alternate morphologic groups predicted BMI or pulmonary function outcomes and oDI was not associated with BMI, %FEV, or %FVC.Conclusions: Alternate OGTT measures including time to peak glucose and 1hG better identify abnormalities in insulin secretion as identified by oDI than traditional OGTT criteria. Further studies, including prospective follow-up, are required to determine whether these alternate methods identify those at greater risk for future clinical decline in CF.

Sheikh, S.; Kelly, A.; Rubenstein, R.C. The Children's Hospital of Philadelphia, Philadelphia, PA, USA Objectives: While nutritional recommendations in cystic fibrosis (CF) focus on body mass index (BMI), this measure is a surrogate for body composition. Lean body mass (LBM) correlates more strongly with pulmonary function than does BMI, and the relationship between BMI and LBM is perturbed in CF. Moreover, LBM deficits are one potential mechanism through which cystic fibrosis-related diabetes (CFRD) confers increased pulmonary morbidity. We hypothesized that worsening glucose tolerance is associated with LBM deficits relative to fat mass (FM) and BMI.Cystic fibrosis-related diabetes (CFRD) is present in over 40% of adult CF patients, and is characterized by decreased insulin secretion. CFRD shares many characteristics of both types of classical diabetes, though it may more closely resemble type 2 diabetes, because of the decreased insulin response, lack of auto-pancreatic antibodies and similar genetic predispositions. As in type 2 diabetes, it is hypothesized that ER stress and impaired insulin biogenesis could contribute to CFRD. Our group has demonstrated that the ER chaperone ERp29 (endoplasmic reticulum protein of 29 kDa) is critical for the biogenesis of CFTR (Suaud L, et al. J Biol Chem. 2011; 286:21239) and ENaC (Grumbach Y, et al. AJP Cell Physiol. 2014; 307:C701) .Rationale: The biogenesis of insulin and ENaC share many key features. 1. Both proinsulin and ENaC can be processed by furin-like convertases in the Golgi into a more active from (insulin and cleaved, higher P o ENaC, respectively). 2. Both proteins are likely transported from the ER to the Golgi by COP II machinery. 3. Both proinsulin and ENaC can bypass this processing in the Golgi to be secreted (as proinsulin) or arrive at the apical membrane (as uncleaved, low P o ENaC) in a less active form. In addition, elevated secreted proinsulin/insulin ratios are seen in both CFRD and Type 2 diabetes, suggesting impaired insulin processing (Sheikh S, et al. Diabetes. 2017; 66:134) .Hypothesis: ERp29 is a critical factor in promoting the efficient conversion of proinsulin to insulin.Preliminary Results: Proinsulin co-immunoprecipitates with ERp29: INS-1 rat insulinoma cells and Min-6 mouse insulinoma cells were each lysed under nondenaturing conditions. Lysates were subjected to immunoprecipitation with rat anti-C-peptide and rabbit ERp29 antibodies and precipitated proteins were resolved by SDS-PAGE. ERp29 co-precipitated with anti-C-peptide, which may suggest co-precipitation with proinsulin, and both insulin and proinsulin co-precipitated with anti-ERp29. ERp29 regulates insulin secretion: INS-1 cells were transfected with plasmids encoding wild-type or mutant ERp29s (C157S, ΔKEEL, KDEL). Cells transfected with the mutant ERp29s had decreased insulin secretion in glucose-stimulated insulin secretion experiments. Similarly, siRNA-mediated depletion of ERp29 increased proinsulin secretion in Min-6 cells. These data, taken together with the immunoprecipitation data, suggest that functional ERp29 interacts with and plays an important role in insulin biogenesis and secretion. ERp29 and proinsulin form a complex with the KDEL receptor: Lysates of Min-6 cells prepared under nondenaturing conditions were subjected to immunoprecipitation with either anti-C-peptide, anti-ERp29 or anti-KDEL Receptor. Precipitated proteins were resolved by SDS-PAGE. Both ERp29 and proinsulin were found to co-precipitate with the KDEL receptor.Conclusion: Together, these data support the hypothesis that ERp29 plays a role in the conversion of proinsulin to insulin, as well as begin to suggest a role for the KDEL receptor in this process.Acknowledgments: Supported by grants from NIH (R01 HL135670) and the Commonwealth of Pennsylvania. Cystic fibrosis (CF) is an autosomal recessive disorder, that is often associated with malnutrition, secondary to intestinal malabsorption, inadequate intake, and increased energy expenditure. With an early CF diagnosis, more focus on nutritional failure and increased use of nutritional supplements, the prevalence of malnutrition among patients with CF has significantly declined. While the prevalence of overweight and obesity in the general population are substantial, nutritional status in the CF population is varied and recent studies indicate that the prevalence of obesity is also increasing in children and adults with CF. The objective of this study is to determine the prevalence and factors associated with overweight and obesity in adults with CF. We performed a cross-sectional analysis, which included 484 adult patients with CF who were seen at the University of Minnesota CF Center between January 1, 2015 to January 31, 2017.

Cross-sectional analyses were performed among 484 patients with CF age >18 years old. BMI was dichotomized as ≥25 vs.<25 kg/m 2 . Bivariate association between predictors and the dichotomous BMI was evaluated using chi-square test or two-sample t-test depending on the nature of the variable. Multivariate analyses were conducted using logistic regression models. Variables that were significant at the 0.10 level in bivariate analysis were included in multivariate analysis. Backward model selection method was used to select the best set of predictors.Mean age of this cohort was 35.2 (SD=11.6), 95% were white and 45% were female. 5.2% were underweight (BMI < 18.5), 62.6% had normal weight (BMI ≥ 18.5-24.9 kg/m 2 ), 25.6% were overweight (BMI ≥ 25-29.9 kg/m 2 ) and 6.6% patients were obese (BMI ≥30 kg/m 2 ). Bivariate analyses of associations with BMI ≥25 kg/m 2 (32% of population), show that gender (25% in female vs. 38% in male, p=0.0036), age at CF diagnosis (mean age of 11.6 among BMI ≥25 vs. 4.8 among BMI <25, p<0.0001), CF transmembrane conductance regulator (CFTR) genotype classification (62% in mild vs. 25% in severe vs. 30% in unclassified, p<0.0001), and pancreatic insufficiency status (26% in pancreatic insufficient patients vs. 63% in pancreatic sufficient patients, p<0.0001) were significantly associated with BMI ≥ 25 kg/m 2 . CF-related diabetes was trending towards significance (28% vs. 36%, p=0.067). Multivariate regression analyses show that the best set of predictors of BMI ≥ 25 kg/m 2 were gender (female vs. male OR=0.46, p=0.0013), CFTR genotype classification (mild vs. severe OR=3.44, p=0.0003), and age at CF diagnosis (OR=1.03, p=0.0064). CFTR modulator (lumacaftor/ivacaftor) was not significantly associated with BMI ≥ 25 kg/m 2 .In our patient population, prevalence of overweight and obesity is significant. Male gender, older age at CF diagnosis, mild CFTR genotype and pancreatic sufficient status were associated with overweight/obesity. Background: Aging cystic fibrosis (CF) patients are at high risk of developing glucose intolerance, which can lead to CF-related diabetes (CFRD). Decrease in insulin secretion over time is the main hypothesis to explain this increasing prevalence but mechanisms, especially in older patients are still not well elucidated.

Objective: Assess evolution of glucose tolerance and insulin secretion/ sensitivity in aging CF patients.Methods: This is a prospective observational analysis in the older adult CF patients (n=46) from the Montreal Cystic Fibrosis Cohort (MCFC) who were followed-up for at least 4 years. Baseline and follow-up 2-hour oral glucose tolerance test (OGTT; with glucose and insulin measurements every 30 minutes for 2 hours), pulmonary function test (FEV1) and anthropometric measurements were performed the same day. Inclusion visit in the MCFC and last visit were included in this analysis. Insulin sensitivity was measured by the Stumvoll Index.Results: After a mean follow-up of 9.9 ± 2.6 years, mean age at follow-up was 43.5 ± 8.1 years old. An increase of body weight (+2.6 ± 6.5 kg, p = 0.01) and a decrease in pulmonary function (FEV1; 73.4 ± 21.2% to 64.5 ± 22.4%, p ≤ 0.001) were observed. Overall, insulin secretion is maintained at follow-up but all OGTT glucose values increased (for all values, p ≤0.028). At follow-up, 28.3% of patients had a normal glucose tolerance while 71.7% had abnormal glucose tolerance (AGT). AGT patients decreased their insulin sensitivity over time (p = 0.029) while it remained the same in patients with normal glucose tolerance (p = 0.917).Conclusion: In older CF patients the progression of impaired glucose tolerance is occurring with stable insulin secretion but reduced insulin sensitivity. Introduction: Inflammation has been implicated in the pathogenesis of both diabetes and its chronic complications in adults. Little is known about inflammation levels in three common types of diabetes in youth: Type 1 diabetes (T1D), Type 2 diabetes (T2D) and cystic fibrosis-related diabetes (CFRD).

Objective: To investigate circulating serum biomarkers associated with inflammation in T1D, T2D and CFRD in adolescents, by exploring biomarkers by diabetes type and complication status and to assess their determinants.Methods: The study group (mean age 15±3 years; 49% female) consisted of 134 T1D, 32 T2D, 32 CFRD, 37 controls without diabetes and 11 with CF but normal glucose tolerance. Vascular inflammation was assessed by sE-selectin (ELISA) and systemic inflammation by hsCRP (turbidimetry), as well as WCC and ESR in those undergoing diabetes complications assessment. Nephropathy was defined as elevated albuminuria (≥20µg/min). Cardiac autonomic neuropathy was assessed by measures of heart rate variability (HRV) on a 10-min continuous electrocardiogram, peripheral nerve function by vibration and thermal threshold testing and retinopathy by 7-field stereoscopic fundal photography. Descriptive statistics, ANOVA and regression analyses were performed, with significance at p<0.05.Results: T1D participants had higher HbA1c than T2D and CFRD (8.5 [7.7-9.5 ] vs 6.6 [5.6-9.2] vs 5.9 [5.7-6.5 ]%, p<0.001) and longer diabetes duration (8 [5-11] vs 2 [0-3] vs 2 [1] [2] [3] [4] years, p<0.001). Comparing T1D vs T2D or CFRD, inflammatory markers were lower in T1D (p<0.05). sE-selectin and hsCRP levels in CFRD were higher though not significantly than in CF subjects without diabetes (p=0.2; p=0.15). T1D and CFRD compared to T2D had less albuminuria (1 and 0 vs 19%) and less peripheral nerve abnormalities (5 and 0 vs 19%). HRV abnormalities were less common in T1D in contrast to T2D and CFRD (28 vs 54 vs 50%). In multivariable analysis, hsCRP and ESR were associated with complications after adjusting for HbA1c, BMI and diastolic blood pressure (hsCRP: odds ratio (OR)=2.2, p=0.02; ESR: OR=2.5, p<0.04). In addition, BMI was independently associated with hsCRP (correlation coefficient (CC)=0.3, p=0.001) and ESR (CC=0.2, p<0.03).Conclusion: Inflammatory markers, including sE-selectin, hsCRP, WCC and ESR are elevated in adolescents with diabetes, being higher and comparable in T2D and CFRD than in T1D. The association between inflammation and diabetes complications is consistent with inflammation driving vascular pathology in diabetes. Background: Cystic fibrosis-related diabetes (CFRD) is the most common nonpulmonary manifestation of CF and is associated with decline in pulmonary function, compromised nutritional status, and increased mortality. The bionic pancreas (BP) is a closed-loop system of glycemic management that employs continuous glucose monitoring (CGM) coupled with mathematical algorithms to automatically administer insulin with or without glucagon. The BP has been shown to both simultaneously lower mean CGM glucose and time spent in hypoglycemic ranges in patients with type 1 diabetes while improving quality of life measures.

Hjelm, M.; Nemastil, C.J.; Salvator, A.; Hayes, Jr., D.; Tumin, D. Pulmonary, Nationwide Children's Hospital, Columbus, OH, USA Introduction: Cystic fibrosis (CF) patients have an increased prevalence of depression and anxiety compared to the general population, which contributes to poor treatment adherence and worsened clinical outcomes. Little is known regarding clinical factors contributing to the development of poor mental health in the CF population. Another co-morbidity in this population is CF-related diabetes (CFRD), which affects 35% of patients by 30 years of age. Research assessing the impact of newly diagnosed CFRD is limited, but qualitative studies have revealed patients with feelings of increased disruption, anxiety, and confusion related to their CFRD treatment regimen. No studies have been conducted evaluating anxiety or depression relative to a diagnosis of CFRD. We hypothesized that CFRD increases risk of anxiety and depression in CF patients and is associated with worse clinical outcomes.Purpose: Our purpose was to evaluate the association between new CFRD diagnosis and the mental health of patients with CF, particularly a diagnosis of anxiety and depression.Methods: A single-center, retrospective cohort study was completed including CF patients seen at our institution from 2009-2017. Diagnoses of CFRD, depression, and anxiety were identified using ICD-9 and ICD-10 codes and verified by manual chart review. Group comparisons according to new CFRD diagnosis were assessed using two-sample t-tests or Wilcoxon rank sum tests for continuous variables, and chi-square or Fisher's exact tests for categorical variables. Multivariable logistic regression was used to predict anxiety and depression according to CFRD diagnosis.Results: The analysis included 238 patients with a mean age of 31 ± 12 years. Of the patients included in the analysis 83 (35%) had a new diagnosis of CFRD, 71 (30%) were diagnosed with depression, and 86 (36%) were diagnosed with anxiety. On multivariable analysis adjusting for sex, CFTR mutation, and age, patients with CFRD were more likely to have depression than CF patients without CFRD (OR = 2.1; 95% CI: 1.1, 3.7; p=0.02). The adjusted association between CFRD diagnosis and anxiety did not reach statistical significance (OR = 1.6; 95% CI: 0.9, 2.9; p=0.11).Conclusions: Our results suggest an association between CFRD diagnosis and worse mental health among patients followed at our institution. The association of CFRD diagnosis with increased likelihood of depression may be stronger than its association with increased likelihood of anxiety. Mental health sequelae of CFRD diagnosis may complicate CF patients' adherence and self-management.Acknowledgements: Supported by CF Foundation Clinical Fellowship Grant. Background: People with CF experience myriad physical, emotional, and socioeconomic burdens, all of which affect quality of life (QoL) for patients and their families. Although specialty palliative care (PC) reduces suffering and improves QoL for individuals with serious illness, no evidence exists for its impact in CF. As such, we conducted the first randomized clinical pilot trial of a specialty PC intervention in CF.

Methods: We used patient-reported outcome measures to identify unmet palliative needs among patients at the University of Pittsburgh Adult CF Center. Using these data, we developed InSPIRe:CF (Integrating Supportive care to Prepare, Improve QoL, and Reduce suffering:CF), a protocolized, patient-centered PC intervention based on the Chronic Care Model. InSPIRe:CF embeds a PC clinician within the CF center, working alongside usual CF care providers. InSPIRe:CF is delivered concomitantly with usual CF care. Patients receive at least four in-person visits (and follow-up calls as needed) with a PC clinician, addressing four main foci: symptom management, emotional support, advance care planning and goals of care, and coping and resilience. This single-site, two-arm, unmasked, parallel randomized pilot clinical trial evaluated the feasibility, acceptability, and perceived effectiveness of InSPIRe:CF. We measured feasibility via enrollment and outcome assessment rates. We conducted individual semi-structured interviews to evaluate acceptability, perceived effectiveness, and barriers and facilitators to trial participation. Two investigators independently analyzed qualitative data using thematic analysis.Results: We randomized 50 adults with CF to InSPIRe:CF plus usual care, or usual care alone (approach-to-randomize rate, 79%). Fifty-six percent of our sample was male, with a median age of 32 years (range: 18-67), and median FEV 1 of 41% predicted (range: 20-82% predicted) at enrollment. Of 50 participants randomized, two died and one was lost to follow-up. Overall, 67% of intervention participants reported the intervention was not burdensome and 100% agreed or strongly agreed that they were satisfied with the PC clinician's care. Sixty-seven percent of participants agreed or strongly agreed that the intervention improved their physical symptoms, 62% improved their QoL, and 100% felt that all patients with CF should receive specialty PC. Preliminary themes identified from analysis of summative interviews include: 1) appreciation of the patient-centeredness of a PC approach; 2) endorsement of the importance of goals of care elicitation and advance care planning discussions; and 3) a desire to expose individuals with CF to specialty PC at earlier stages of their disease.Conclusions: Embedding specialty PC within usual CF management is feasible, acceptable, and perceived to be effective among individuals living with CF. Given these promising findings, further randomized clinical trials are warranted to establish the effectiveness and optimal delivery models of specialty PC in CF.Acknowledgments: Supported by CFF PILEWS14QI0, NHLBI K0133466. The aim of this study was to evaluate the implementation of these guidelines in clinical practice.

Methods: CF patients (12-18 years., n=17) and caregivers of CF patients (0-18 years., n=79) completed the GAD-7 and PHQ-9 questionnaires to assess symptoms of anxiety and depression. A comparison of screening and formal clinical assessment outcomes was undertaken. Referral pathways of treatment were investigated.Results: Participants reported more and different psychosocial problems (e.g. stress, relationship problems) at clinical assessment that required psychosocial treatment. Treatment was provided for 48 participants (10 CF adolescents, 25 mothers and 13 fathers) who mainly scored nonelevated on the PHQ-9 and/or GAD-7. Some participants refused treatment, indicating barriers to receiving psychosocial treatment.Conclusion: This is the first clinic evaluation of the ICMH recommendations statements. Deviation from the guidelines occurred in this study as all participants underwent clinical assessment allowing an evaluation of the screening process. Many CF patients and caregivers, who screened in the nonelevated range on depression and anxiety also required treatment, highlighting the complexity and demand for mental health provision. Future studies should identify barriers towards psychosocial treatment. (6):493-500). However, optimal timing of end of life (EOL) discussion in relation to lung transplant (LT) has not been established for CF patients. In this study we aim to describe EOL care practices for CF patients who died of end-stage CF without receiving a LT compared to patients who received a LT.

Methods: We conducted a retrospective chart review of all CF patients who died without receiving a LT or received their first LT from August 1, 2012 -December 31, 2017. We identified patients using the CF Foundation registry, then conducted a manual chart review. Included patients were ≥18 years of age at the time of death or LT. The primary location for their CF care was the Univ. of Texas Southwestern. Statistical analyses utilized chi-square contingency table analysis comparing patients who received a LT to those that died without receiving a LT on each categorical variable, and a Student's t-test for independent groups to compare the two groups on each numerical value. The primary outcomes were 1) whether an EOL discussion occurred, 2) the time of EOL discussion prior to death or LT, 3) if the patient was evaluated by PC, and 4) whether they had an advanced directive (AD) or medical power of attorney (MPOA).Results: We identified 23 patients who died without LT and 40 patients who received a LT for end-stage CF. Median age of those who died was 27, and for those who received a transplant was 28. Patients who died were 39% female, and 53% were female in the LT group. Among patients who received a LT, 6/40 (15%) had EOL conversations with a pulmonologist compared with 17/23 (74%) who died (p<0.0001). Patients who died were more likely to have PC involved (57% v. 5%, p=0.001). Timing of EOL conversations was similar for patients who died (9/17, 53% occurred >6 months prior to death) compared to those who had received a LT (2/6, 33% >6 months prior to LT) (p=0.640). Patients who died were more likely Prior research has shown that pediatric CF patients with low or moderate adherence to medications are likely to have increased rates of hospitalization and higher healthcare costs. In adults, the role of medication adherence in maintaining stable lung function is less clear.Objective: To examine the relationships between medication adherence, barriers to adherence, and lung function as measured by ΔFEV 1 in adults with CF.Methods: This is a retrospective cohort study of patients presenting for follow-up from January 2018 to April 2018. Patients were included if they were age ≥ 18, using at least one key CF lung therapy (bronchodilator, hypertonic saline, dornase alfa, airway clearance, inhaled antibiotics), had a PFT, and completed a survey assessing adherence and barriers to adherence at the same visit. Post-lung transplant patients were excluded. Lung function was measured by ΔFEV 1 , defined as the percent change in FEV 1 on the day of the visit vs their personal best in the previous 12 months. Stable lung function was defined as ΔFEV 1 > -5%, and an acute decline in lung function as ΔFEV 1 ≤ -5%. Adherence rates were reported in two groups, with ≤ 50% adherence being "low" and > 50% adherence being "moderate-high." Barriers to adherence were assessed using survey responses.Results: A total of 88 out of 130 patients met inclusion criteria for this study. The average age was 32 ± 9 years. The majority of patients were white (92%), male (55%), and had at least a high school education (82%). Almost one-third (26%) identified as Hispanic/Latino. The odds of presenting with an acute decline in lung function with low adherence vs. moderate-high adherence was not significantly different (OR, 1.33; 95% CI, 0.81, 2.14). The most frequently reported barriers to adherence related to the lifestyle burden of treatments -patients indicating they don't have time to perform a therapy (46% of reported barriers) or they forget (26%). Significantly more patients endorsed forgetting as a barrier in the stable group (19.3%) vs. the acute decline group (11.2%), χ 2 (1, N = 88) = 5.038, P = .02.Conclusions: In this adult CF population, the estimated odds of presenting with an acute decline in lung function were 33% higher for patients reporting low adherence vs. those reporting moderate-high adherence. While not significant, this trend mirrors previous research that found worse adherence to pulmonary medications was associated with higher acute healthcare use in CF patients age ≥ 6 years (Quittner AL, et al. Chest. 2014; 146 (1):142-51). The lifestyle burden of CF treatments was identified as the major barrier to adherence. This is in agreement with previous studies in pediatric CF populations that identified "time management" and "forgetting" as top barriers to adherence (Modi AC, et al. J Pediatr Psychol. 2006; 31(8) :846-58). The significant increase in the number of stable patients endorsing forgetting as a barrier to adherence highlights an opportunity for the CF care team to engage patients in discussions regarding the role of each treatment in maintaining overall lung health, and to design patient-specific interventions to overcome barriers related to treatment burden.

Passamano, I.Y. Cystic Fibrosis Relief Fund, Inc., Houston, TX, USA Introduction: Hurricane Harvey is the second most destructive hurricane in U.S. history. It covered 1260-square-miles of Harris County Texas with over 1.5 ft of water, including 1/3 of Houston. Over 100,000 people evacuated their homes. Among those affected were ≈200 households with a person with cystic fibrosis. However, none of the existing relief agencies specifically addressed the unique needs of people with cystic fibrosis affected by the hurricane.Study Question: Is there a need for disaster relief targeted to individuals with cystic fibrosis affected by disasters?Method: The Cystic Fibrosis Relief Fund, Inc. (CFRF) was established by Isabella Passamano days after Hurricane Harvey as a nonprofit corporation to raise funds and deliver assistance to individuals with cystic fibrosis affected by the disaster. The CFRF coordinated efforts with local pediatric and adult CF centers in Houston. CF center social workers referred to CFRF individuals and families needing assistance or confirmed that an individual or family was in need. This coordination allowed CFRF to identify individuals most in need, identify their specific need, reduce cost and delays, deliver assistance without complex paperwork or unnecessary vetting procedures. Requests for relief are received in writing by email and presented to the CFRF Board to determine if:1. The request is by or for an individual with CF; 2. The need is caused by the hurricane; 3. The individual lacks resources (i.e., lacks available income or resources to meet their immediate need of food, healthcare, clothing or shelter); and 4. The request is needed to meet an immediate need of food, healthcare, clothing, or shelter.CFRF raised funds from donors and delivered assistance to 42 individuals or families. CFRF directly paid vendors for large goods (e.g., mattresses, refrigeration for medicine) and services (e.g., rent, utilities). Direct payment assured that the funds were used for the intended purpose. Funds for food and fuel were delivered via debit cards for easy distribution and use.Discussion: Requests were for items needed to re-establish a household such as clean bedding; mattresses; food; rent; relocation costs; kitchen items; clothing; and fuel. 50% of funds dispersed were for mattresses and bedding; 20% for food; 14% for housing and relocation expenses; 9% for household and kitchen items (including refrigeration for medications); 4% for clothing; and 3% for fuel, gasoline, and utilities. Requests are reviewed and confirmed within 2 days, and relief is delivered in 5 days. 81% of recipients received food assistance and at least one other form of aid. 100% of the individuals requesting relief received assistance.CFRF expanded its scope to include individuals with CF affected by Hurricane Irma in Florida and by California wildfires in late 2017.The CFRF is now expanding its scope to form a National CF Relief organization to deliver assistance to CF families affected by food insecurity and by future disasters.Conclusion: Individuals with cystic fibrosis need targeted disaster relief because they have unique needs not met by FEMA or other relief agencies. Coordination with CF center in the affected area is essential to a quick and proper assessment of requests. CFRF effectively and rapidly delivered relief to individuals with CF affected by three disasters. Background: In accordance with new CF mental health guidelines, the UNC Adult CF team performs universal screening for anxiety and depression and offers individual therapy (in-person or via teletherapy) for those with positive screens. Through this process, we identified feelings of isolation and loneliness as being an important theme. Since the early 1990s, CF patients have had limited exposure to other patients with CF due Conclusion: There are many factors that influence patient readiness to discuss ACP and EOLC, and completion of advance directives. With improving life expectancy and care options for adults with CF, planning for the future remains crucial. In this small single-center study, there is suggestion of an increase in thinking, talking, and planning for the future amongst our adult outpatient population over the past 5 years. Larger, multicenter longitudinal studies could better elucidate patterns. The multidisciplinary CF team should continue efforts to educate and encourage our patients to plan for the future. (Kazmerski TM, et al. J Cyst Fibros. 2018; 17:57-63) . The goal of this study is to explore contraceptive use and pregnancy history among reproductive-aged women with cystic fibrosis (CF), and to ultimately assess the association of contraceptives and pregnancy with health status. We also evaluate contraceptive use and proportion of pregnancies coinciding with CFTR modulator use. Current recommendations suggest that women on CFTR modulators use contraception other than combined hormonal contraceptive pills, if they wish to avoid pregnancy (Ren CL, et al. Ann Am Thorac Soc. 2018; 15:271-80) .

Methods: Women with CF aged 18-50 were recruited from three adult CF centers in the United States (University of Washington, National Jewish Health and University of Texas Southwestern Medical Center) and matched with their health information from the Cystic Fibrosis Foundation Patient Registry (CFFPR). They consented to CFFPR access and completed an online questionnaire, which asked about prior contraceptive use, pregnancy history and CFTR modulator start and stop dates.Results: One hundred and fifty women with CF between the ages of 18-50 who were identified in the CFFPR completed the survey. Ninety-five percent (n=143) reported ever using contraception and 62% (n=93) were currently using contraception. Among 34% (n=51) of women who reported having ever been pregnant, 49% (n=25) had at least one unplanned pregnancy. Overall contraceptive use and type by age group (18-24 years vs. 25-50 years) between 2012 and 2016 were similar. Among CFTR modulator users, 70.2% reported currently using contraception, compared to 58.9% of women not on CFTR modulators (p=0.162). Two women of 48 CFTR modulator users reported unplanned pregnancies after modulator use initiation. One of these women was using combined hormonal contraceptives, the other was not using anything.Conclusion: Reproductive-aged women with CF reported similar rates of contraceptive use and unintended pregnancy to that of reproductive-aged women in the general United States (US) population. Unplanned pregnancy occurred among 4.2% of women using CFTR modulators, suggesting further research that evaluates modulator use and safety in reproductive-aged women is warranted. Linkage of these findings to CFFPR data is forthcoming and will be available at the NACFC presentation.

Siracusa, C. 1 ; Filigno, S.S. 2 1. Pediatr. Pulmon. Medicine, Cincinnati Children's Hospital, Cincinnati, OH, USA; 2. Behav. Med. and Clin. Psychol., Cincinnati Children's Hospital, Cincinnati, OH, USA Background: For the past decade, the Cystic Fibrosis (CF) Center at Cincinnati Children's Hospital has been addressing the mental health needs of our patients through a model that includes social work, a psychologist embedded in the CF clinic, and an inpatient behavioral medicine consultation service. Our center was supported by the CF Foundation (CFF) mental health coordination grant and has been implementing the CFF-recommended mental health guidelines since 2016. In order to better address the mental, emotional and behavioral needs of our patients and families, we began planning a multifaceted, multidisciplinary program to address observed gaps in our care model at the end of 2017. By centralizing, optimizing, and expanding our past and current efforts, we developed a new program: BETTER-CF (Behaviors, Emotions & Thoughts To Enhance Resilience in CF). The three cores of this program include innovations in data management, education, and clinical care.Initiatives: 1. Data Management: We worked with our internal application specialists to create accurate, real-time, psychosocial databases and reports that include mental health screens, standardized psychosocial assessments, psychiatric diagnoses, and key aspects from psychologist and social work notes. These reports are generated with information pulled from the EMR, and are being used in previsit planning, team communication, and clinical care (Moore S, et al. Pediatr Pulmonol. 2018 ;53(S2):409).2. Education: We implemented a behavioral health education series for CF care team members. This case-based series has been incorporated into our center meetings, combines didactics and interactive learning, and is presented by various members of the care team (Filigno SS, et al. Pediatr Pulmonol. 2018 ;53(S2):406). A needs assessment guided the development of this programming.3. Clinical Care: We began incorporating aforementioned data reports into our weekly previsit planning meetings, in addition to our weekly center discussion of current inpatient admissions. We also are in early stages of developing a novel inpatient care model with a focus on improved mental health support, continuity from outpatient care, and self-management; led by a psychologist in a newly-created position. Redesigning this milieu will require education and training, and buy-in from all members of the inpatient team as well as patients and families. Finally, we have worked with caregivers to develop our process for caregiver mental health screening and resource allocation in response to elevated scores.Next Steps: We will continue to test and expand the described program, as well as begin to incorporate new initiatives (e.g. incorporating adherence data into pre-visit planning). We ultimately foresee this program to be incorporated in all aspects of CF care, including outpatient visits, inpatient stays, and integrating into self-management at home. We anticipate a cultural change at our CF center that encourages every interaction with patients and families to be positive and therapeutic. Once the program is refined and optimized, we plan to spread our robust resources to other programs in our division and share our learnings with other CF centers. Aim: We aim to improve the identification of the need for increased psychosocial support for patients 3 to 11 years old. The process begins Introduction: It is recognised that among CF patients psychological and behavioural problems are common and varied (CF Trust, 2011) . The TIDES study (Quittner A, et al. Thorax. 2014 ;69:1090-7) reported a high incidence of anxiety and depression among CF patients >11 years (y) and their parents/carers (parents). Both recommend patients 12+ y and parents of patients aged 0+ y should be screened annually for psychological problems. This study aims to take up these recommendations and also to look at psychological well-being of patients >3 y.

Method: Over a 12-month period psychological screening was completed during CF annual review (AR). The PHQ9 (Spitzer R, et al. JAMA. 1999; 282:1737-44) and GAD7 (Spitzer R, et al. Arch Intern Med. 2006; 166:1092-7) were used to screen parents for symptoms of depression and anxiety. In addition the SDQ (http://www.sdqinfo.com) was used: parents completed the SDQ for child aged 4-10 y; patients >10 y also completed a self-report SDQ. Information and the questionnaires were sent prior to AR and scored on the day.Results: Of 264 families sampled age was 0-17 y, mean 8 y and 125 (47%) were female. Of at least one parent and child, 189 (72%) completed all age appropriate measures, 28 (11%) completed some measures and 47 (18%) declined. PHQ9 and GAD7 data were provided by 70% of mothers and 47% of fathers; 190 with children <12 y, 74 >11 y. SDQ was completed by 72% of parents with children >3 y and 70% of children aged >10 y. Parents who completed questionnaires reported 9% anxiety and 7% depression in the moderate-severe range. Parent-rated concern on SDQ was 7-16% across the subscales, and in the self-reported 2-10%. Mothers reported significantly higher levels of anxiety (p=.002) and depression (p=.001) than fathers. Parents of patients >11 y were significantly more anxious (p=.04) and depressed (p=.01) than those of <12 y. There was no significant difference between the two age groups in SDQ emotional distress scores but parents of children >11 reported greater impact of challenges on their child (p=.02).Discussion: This study screened psychological well-being in 264 CF patients 0-17 y and their parents of a CF centre. Subject recruitment was high. Anxiety and depression reported by parents were lower than many studies including TIDES. Challenges for emotional well-being in patients both as reported by their parents and the patients (>11 y) indicated a lower incidence of psychological distress than other CF studies and is similar to the general population. Patients >10 y reported a lower incidence of problems than their parents did about them. Parents of patients <12 y reported themselves as significantly less anxious and depressed than those of >11 y. Annual monitoring will be continued to offer interventions to promote adherence to treatment and health-related quality of life. Further research to explore the reasons for the difference in the mood of parents of patients <12 y and those of older paediatric patients is indicated. Conclusions: The ABQ modules provide an individualized "snapshot" of the most common or difficult barriers to CF treatments. It takes only a few minutes to complete the ABQ-CF modules, which have 4 to 9 questions. This tool can be used in clinic to initiate a patient-centered conversation about what is "getting in the way" and lends itself to a brief, behavioral intervention to solve the barrier. Additional reliability and validity analyses will be performed with the ABQ-CF, medication possession ratios, and the CFQ-R Treatment Burden scale.Acknowledgments: Supported by: CF Foundation grants BARRI-ERS16PE0, SAWICK14PE1 and RIEKERT15PE0.

Quittner, A.L. 1 ; Eckmann, T. 2 ; Riekert, K.A. 2 1. Miami Children's Research Institute, Nicklaus Children's Hospital, Miami, FL, USA; 2. Medicine, Johns Hopkins, Baltimore, MD, USA Objectives: The CF regimen is complex and burdensome, with low rates of adherence across treatments. Discrepancies between the patients' and clinics' understanding of the treatment plan may partially explain poor adherence. We compared adolescents' and adults' reports of their treatments with the prescribed treatment plan (PTP) completed by the research coordinator.Methods: Adolescents and adults from 12 pediatric (PEDS) and 9 adult (ADULT) CF centers, participating in the Success with Therapies Research Consortium (STRC), completed demographics and the ABQ-CF (Adherence Barriers Questionnaire) on which they endorsed their treatments. The PTP was completed by the research coordinator via chart review. Eligibility criteria were: age 13+ years, diagnosed with CF, English speaking, on a chronic pulmonary medication, and no hospitalization expected in next 28 days.Results: 407 people were enrolled (PEDS=207; ADULT=200). Mean age PEDS=17.1 (SD=3.1) and ADULT=32.8 (SD=10.9), 50% were male, 85% were White non-Hispanic, and 30% had public/no health insurance. Mean FEV 1 percent predicted was PEDS = 84.4 (SD=20.2), ADULT = 64.2 (SD = 22.9). Overall, agreement was fairly good, with discrepancies (percent difference between patient endorsement vs PTP) ranging from a low of .50% (enzymes, ivacaftor) to a high of 12% (TOBI Podhaler) across ages. Agreement using Kappa coefficients (.70 criteria) indicated lowest convergence for vitamins and dornase alfa, with borderline-low kappas for TOBI Podhaler across centers. Next, we examined the magnitude of the discrepancies by clinic type and found significantly greater discordance for hypertonic saline for PEDS (8.2%) vs ADULT (2.5%) and greater discordance for colistin (ADULT = 7.0%; PEDS= 1.9%) and TOBI Podhaler (ADULT = 12%; PEDS = 4.4%) in ADULT vs PEDS centers (p's <.05). Across rates of discordance, patients reported they were not on a medication the clinic had prescribed.Conclusions: Overall, there was substantial agreement between patients and clinics on their prescribed regimen; however, there were several instances of participants not identifying a medication prescribed in the chart. Agreement was highest for enzymes and ivacaftor and lowest for vitamins, dornase alfa and TOBI Podhaler. Disagreements between adolescents and adults were found for hypertonic saline, with adolescents endorsing this less often than adults; fewer adults reported prescriptions for colistin and the Podhaler than adolescents. These results suggest that providing a written treatment plan to patients at each clinic visit, such as the PTP, would be useful for increasing patients' understanding of their regimen. Treatments under-reported by patients may be those with a greater number of barriers. Adherence was also lowest for hypertonic saline in a national study of adherence across ages (Quittner AL, et al. Chest. 2014; 146:142-51) . A key limitation: we did not evaluate patients' understanding of the frequency or dose of these treatments, which could reveal additional discrepancies. Data from the newly developed CF-Adherence Barriers Questionnaire (CF-ABQ) may provide more insight into reasons for poor adherence. and treatment of their own mental health, while also improving the care and treatment of our CF patients.Acknowledgments: Funding by CFF grants CC182-16 and CMHC182-15. Methods: Adult patients with CF were recruited from outpatient pulmonary clinics and inpatient units at the University of Alabama Hospital and Children's Hospital of Alabama. Total of 123 patients completed a baseline survey and 97 of them (79% retention) completed the same survey 6 months later. The surveys included the Pittsburgh Sleep Quality Index to assess sleep quality and the Brief Symptom Inventory to measure depression symptoms. Correlations examined concurrent associations between sleep quality and depressive symptoms. Then, linear regressions predicted depressive symptoms at follow-up from sleep quality at baseline, controlling for FEV1.

Results: The 123 participants had a mean age of 31.9 years (SD=11.6 years; age range: 19-67 years), 43% were male, 93% were White and 7% were African American. At both baseline and follow-up, better sleep quality was associated with fewer depressive symptoms (r=.49 and .35, p<.001). In the linear regression, poor sleep quality at baseline significantly predicted increased depression symptoms at follow-up (beta=.35, p=.001).Discussion: Better understanding of risk factors for depression in patients with CF may lead to interventions to prevent or predict the development of depression in this population. Since sleep quality may predict depression, screening for sleep quality may be helpful in identifying patients at risk for developing depression.Acknowledgments: Supported by Center for Palliative and Supportive Care, University of Alabama at Birmingham. We have observed that a subgroup of adults with cystic fibrosis (CF) reports widespread chronic pain and fatigue that do not resolve with conventional CF specific treatment. Their symptoms appear similar to those that meet criteria for fibromyalgia. Fibromyalgia (FM) is a centralized pain state in which the central nervous system is the origin or amplification of pain (Clauw D. JAMA. 2014; 311:1547) . Thus, people with fibromyalgia will feel more pain than would normally be expected from nociceptive causes. We used a standard FM measure to assess the chronic pain experienced by adults with cystic fibrosis and compared the results to those reported for patients with fibromyalgia.

Methods: We used an existing fibromyalgia survey that combines a Symptom Severity Index and Michigan Body Map. This survey reports out a total score on the FM symptom scale (FS) and subscores of chronic pain widespreadness (WPI) and severity (SS). A FS score ≥12 meets criteria for fibromyalgia. We added two qualifying questions to the standard FM survey to identify pain specifically related to abdominal and sinus disease in CF. The FM survey was offered to 24 consecutive patients during their annual outpatient physical therapy evaluation and to 12 patients during hospitalization for treatment of a pulmonary exacerbation. One patient did not fully complete the survey and could not be scored.Results: The FS scores for the 35 patients who completed the study were normally distributed with a mean (SD) of 10.7 (6.7). Fourteen patients (40%) met FS criteria (≥12) for fibromyalgia. Positive responses to the additional CF-specific questions regarding abdominal and sinus pain were widely distributed among the patients and not concentrated in the highest FS scoring group. The mean WPI score was 5.2 and mean SS was 5.6. The mean FS score for our subjects (10.7) was considerably higher than the mean reported for a comparison population of 16,273 patients presenting for elective surgery (5.4; See Figure) . Our patients whose scores met the FS criteria reported pain in more nonjoint areas than those who did not meet FS criteria, which is consistent with comparison of FM and non-FM populations (Lu L, et al. Arthritis Rheumatol. 2017 ;69(suppl 10);[abstract]).Conclusion: The FM survey results from our adult CF patients reveal a high level of widespread pain with 40% scoring in the range seen in patients with fibromyalgia. This frequency is much higher than the 2-8% reported for control populations. We suggest that this FM survey tool can be used to evaluate the widespread chronic pain that occurs in adults with CF and to measure the effects of treatment directed to alleviate the chronic pain. Introduction: For patients with cystic fibrosis (CF), indicators of physical health and disease severity are closely monitored by a team of health professionals at periodic clinic visits. Body mass index (BMI) and forced expiratory volume (FEV 1 ) are key metrics of health that guide therapeutic intervention. Because these health outcomes are strongly determined by patients' adherence to treatment regimen and adherence greatly varies across patients, it is important to examine factors that contribute to treatment adherence and health outcomes. One such factor is Health Locus of Control (HLC), which indicates the degree to which individuals believe their health is controlled by their own behavior vs external influences. HLC has three key dimensions: Internal (belief that one's actions affect one's health), Powerful Others (belief that powerful others like medical professionals affect one's health), and Chance (belief that one's health is influenced by chance; Wallston KA, et al. Health Educ Monogr.1978; 6:160-70) . Unique relationships have been identified between these dimensions of HLC and different health behaviors (Abbott J, et al. Thorax. 1996; 51:1233-8) . For instance, CF patients with higher Powerful Others HLC have been shown to have greater adherence to physiotherapy, pancreatic enzyme, and vitamin therapies. These particular behaviors have the potential to help maintain FEV 1 and BMI in CF patients (Mcllwaine MP, et al. Curr Opin Pulm Med. 2014; 20:613-7; Sinaasappel M, et al. J Cyst Fibros. 2002; 1:51-75) . This study examined whether key dimensions of HLC predict health outcomes in patients with CF.

Methods: We examined HLC and health outcomes in 123 adult CF patients (53% male; ages 19-67; mean age=31.9, SD=11.6; 93% White, 7% African American) recruited from inpatient and outpatient facilities in the Southeastern US. At baseline, each participant completed the Multidimensional HLC Scale (Wallston et al, 1978) , an 18-item measure that distinguishes three dimensions of HLC (Internal, Powerful Others, and Chance). BMI and FEV 1 scores were obtained from the CF Foundation registry for baseline and 6-month follow-up. Correlations and linear regressions tested whether the three HLC dimensions at baseline predicted BMI and FEV 1 at follow-up.Results: At baseline, average BMI was 23.8 (SD=5.4) and mean FEV 1 was 61.1% (SD=23.0%). Baseline Powerful Others HLC was positively correlated with follow-up BMI (r=.200, p=.034) and served as a unique predictor of higher follow-up BMI in the regression model (Beta=.213, p=.036) . No dimension of HLC predicted FEV 1 scores at follow-up.Discussion: In previous work, Powerful Others HLC predicted dietary maintenance in CF patients (Abbott et al, 1996) , which may explain why this dimension predicted higher BMI at follow-up in the present study. Interestingly, Internal HLC predicted better diet in community samples (Cheng C, et al. Health Psychol Rev. 2016; 10:460-77 Background: Transition into adulthood and self-care is a critical time for young adults with cystic fibrosis (CF). This stage of adolescent development, in addition to the high treatment burden of CF and the possibility of switching care providers, has been shown to induce fear and anxiety in young adults (Tierney S, et al. Int J Nurs Stud. 2013; 50(6) :738-46). While transition may be emotional, little is known regarding the presence of clinical symptoms of depression and anxiety leading up to and shortly after transfer to adult care. To better understand the effects of transition and its potential impact on mental health we investigated whether young adults with CF experience increased symptoms of depression or anxiety surrounding healthcare transition.Method: Patients attending the Emory+Children's pediatric clinic who were within two years of their scheduled transfer to the adult CF clinic (Pre-Transfer cohort) were matched with a similar age group of patients who had already transferred to the adult clinic (Post-Transfer cohort). Anxiety and depression screening was performed at least annually on all patients utilizing the Generalized Anxiety Disorder 7-item scale (GAD7) and the Patient Health Questionnaire 9-item scale (PHQ9) respectively. The results of annual screenings collected in the calendar year prior and after transition were used. Of 23 subjects in the Pre-Transfer cohort, 12 had screenings both before and after adult care transfer. Continuous categories such as age, PHQ9 score, GAD7 Score and time to/from transfer were compared with unpaired, 2-tailed t-test. Categorical data such as sex was compared with chi-squared test. Significance was determined at p ≤ 0.05.Results: There were no differences between Pre-(n=23) and Post-Transfer (n=10) cohorts in terms of age (19.7±0.60 vs. 19.17±1.03), sex (40% female vs. 52% female), time to/from transfer, GAD7 (4.38±5.58 vs. 2.7±2.21) or PHQ9 (5.23±5.93 vs. 3.4±4.09) scores. It is notable that GAD7 and PHQ9 scores were relatively low in both groups with the average scores in all categories below the meaningful clinical cut off (< 5), with the exception of Pre-Transfer cohort PHQ9 scores (5.23±5.93).When comparing GAD7 and PHQ9 scores in the Pre-Transfer group before and after their own transfer of care, there was a nonstatistical trend towards decreased scores following transfer (PHQ9 -5.23±5.93 vs. 3.22±4.12 before and after transfer, GAD7 -4.38±5.58 vs. 3.72±4.51). Again, individual subject's before and after transfer screening scores were generally low. Only 25% of subjects had an elevated PHQ9 score prior to transfer and 33% of subjects had an elevated GAD-7 score prior to transfer.Conclusion: Given the extent of other contributing factors it is difficult to determine what role healthcare transition plays in the context of mental health and whether PHQ9 and/or GAD7 scores tend to normalize once young adults with CF are well established in the adult care environment. Small cohort sizes limit the power of the analysis in this study and suggest an opportunity for continued exploration. Introduction: Depression in individuals with cystic fibrosis (CF) has been associated with adverse health outcomes, including a steeper rate of decline in lung function (Frederick C, et al. Pediatr Pulmonol. 2017; 52:500) and increased mortality risk (Schechter MS, et al. Pediatr Pulmonol 2017; 52:404-5) . Further, a recently published study found major depression as a significant risk factor for early death in the general population and this association persists over long periods of time (Gilman SE, et al. CMAJ. 2017;189:E1304-10). The goal of this retrospective review was to describe rates and patterns of depression in individuals with CF that died since our center designed and implemented a standardized process for ongoing depression screening with interventions tailored to the severity of depression.

Methods: This was a retrospective, observational chart review of all patients under the care of The Cystic Fibrosis Program of Western New York, who died since 2013 that evaluated: (1) whether these patients had higher rates of depression and (2) the course and treatment of depression. All patients with depression data were included.Results: Nine (70%) of the 13 patients who died met the inclusion criteria, having at least one mental health screening assessment available for review. One patient was not screened secondary to dementia and three did not screen secondary to illness severity at the time of screen initiation. Three of nine patients who were screened died post-lung transplant (LT). The average age at death was 42 years in the LT group screened. No patients who underwent LT screened positive for depression. Four patients in the screened non-LT group (80% non-LT group, 44.4% of all deceased patients screened) met criteria for clinically significant depressive symptoms (2 mild, 1 moderate, 1 severe). Two patients endorsed suicidal ideation. Of the individuals with depression 50% had recurrent depression (N=2) and 50% had chronic depression (N=2). The median/average age of death of individuals without LT was 27 years and 27.25 years respectively.Discussion: Our limited data support an association between depression with rates that are higher than the rates of individuals living with CF. The average and median age of death in individuals who died and screened positive for depression is less than the median age of survival in all individuals with CF aligning with previous research showing early mortality in individuals with CF that died without receiving a lung transplant. It is unclear whether depression predisposed patients to an earlier mortality risk or if worsening health increased the rates of depression. This is likely a multifactorial process however with this small compelling set of data and significant mortality implications, more longitudinal research is needed to optimize screening, evaluation and treatment of mental health in individuals with CF. The association between pediatric chronic disease and caregiver (CG) anxiety or depression has been well documented. Annual screening for ≥1 primary CG of children with CF is recommended given the 2-3-fold increased risk for depression and anxiety in this population (Quittner AL, et al. Thorax. 2014; 69:1090-7) . Risk factors within CF-CG cohorts and correlation of CG mental health with patient health status are less known. Driscoll and coworkers (Pediatr Pulmonol. 2009; 44:784-92; J Pediatr Psychol. 2010; 35:814-22) found increased depression among female CF-CGs as patient lung function decreased, elevated CF-CG depression with increased family stress and lack of employment, and no association between patient zBMI and CG depression. The first year of CG screening in a southeastern pediatric CF center is reviewed with attention to associations between CG anxiety and depressive symptoms and CF health outcomes and CG employment.

Methods: CGs voluntarily completed an electronic PHQ-8 and GAD-7 at their child's visit. Supportive feedback and self-care tips were given to all. CGs with elevated scores in the moderate-severe range were encouraged to seek diagnostic assessment and treatment. Pearson's correlations were used to examine relations of log-transformed CG screen results with child health outcomes of FEV1, BMI z-score (zBMI) and number of hospitalizations in the past year. Household employment status was reviewed when available.Results: A total of 54 CGs were screened. CGs were mothers (72%), fathers (15%), custodial extended family (6%), grandparent (5%), and foster parent (2%). Mean CG age was 37.5 years (9.1). Mean CG scores were 4.8 (PHQ-8) and 4.2 (GAD-7); 20 (37%) CGs had elevated PHQ scores (score ≥5) while 19 (35%) had elevated GAD-7 scores (score ≥5). CG elevations for both anxiety and depression symptoms were strongly associated (rho=0.83, p<0.001) with 16 CGs ≥5 on both screens. Patient FEV1, number of hospitalizations, and CG employment outside the home were not significantly correlated with CG depression/anxiety. Child zBMI was significantly positively correlated with increased caregiver depression/anxiety symptoms (rho= 0.26 (log-PHQ-8), rho= 0.28 (log-GAD-7), p<0.05).Conclusions: These data provide insight into associations of CG mental health with patient health outcomes in our care center. Rates of CG depressive and anxious symptoms are consistent with published rates. Both CG depression and anxiety increased with higher patient zBMI in our center, which differs from previous published data. Additionally we found no association of FEV1, hospitalizations or employment with CG symptoms. Overall, few studies have examined health outcomes in relation to CG depression and anxiety. Further investigations should be conducted with larger sample sizes, as bi-directional impacts of CG mental health and patient health could be important in CF care. Maintaining child nutritional status and goal weights is a primary focus of pediatric CF care, and perhaps parents who have been persistently working on these goals experience increased stress. This study highlights the need for continued CG screening with future attention to factors that increase CG risk and associations with pediatric CF outcomes. Methods: We administered surveys to 3 stakeholder groups: a) individuals with CF, b) caregivers, and c) CF clinicians/care team members via the CF Foundation Community Voice web platform and clinician listservs. CF individuals and caregivers were asked: 1) if they thought CF clinicians should receive more training in PC; 2) whether availability of lung transplant for CF affects PC delivery; and 3) further thoughts. CF clinicians were asked: 1) how they would describe PC in the context of CF care, 2) how the availability of lung transplant affects how PC fits into CF care; and 3) further thoughts. All open-ended responses were read and analyzed by two qualitative researchers.

Results: Seventy individuals with CF, 100 CF caregivers, and 350 CF clinicians responded to our survey. Within each group there was variation regarding what PC is, who should administer it, and when. Among all groups, several participants equated PC with end-of-life care. As stated by a CF participant, PC is "not something really talked about at my stage of CF." Others saw CF-specific PC as "an integral part of the journey for our CF care." Additionally, caregivers highlighted their own PC needs: "The emotional needs of all involved need to be addressed more fully." Most participants believed that PC was appropriate and necessary for lung transplant, stating that "Palliative care could help patients better understand transplant, and also help the care team get patients to transplant, and help them be more successful after the transplant." However, several participants perceived PC and transplant as mutually exclusive specifically linking PC to opioid use: "The biggest barrier [to PC for lung transplant] is the use of narcotics to alleviate pain." When specifically asked about who should deliver PC services, clinicians' responses also varied with some stating that they "prefer to leave that discussion to a specialist" and others stating "Basic palliative care is the responsibility of every CF provider at every point in the disease."Conclusion: Among CF stakeholders, there are varied beliefs and practices about PC as it relates to people with CF. As a result, many people with CF may not be receiving PC that is appropriate to their needs across the spectrum of disease. Education and outreach are needed in order to increase access to quality PC for individuals with CF. Fibrosis provides care to approximately 180 adults. CF Foundation-accredited care centers have coordinated efforts to provide on-site preventative screening, evidenced-based, focused therapies, medication management and community referrals as indicated. We have implemented annual screening since 2014 with the PHQ-9 and GAD-7 as clinically indicated. Collaborative team discussions have focused on the utility and appropriateness of general screening tools such as the GAD-7 and PHQ-9 for the CF population. Given that the DSM-5 diagnostic criteria for both depressive and anxiety disorder requires significant functional impairment, we sought to assess the relationship between self-reported functional impairment and scores from GAD-7 and PHQ-9 screening.

Methods: This is a retrospective chart review examining adult cystic fibrosis patients seen by the CF Collaborative Care team between January 1, 2017 to December 31, 2017. Screening results were included in this study only if both total score and self-reported level of impairment were recorded. Patients were tested at routine, nonsick visits. Primary outcomes were total scores on PHQ-9 and GAD-7 as well as self-reported levels of functional impairment.Results: Out of 276 total visits, 207 (75%) contained complete screening records that included both total scores and self-rated functional impairment for both GAD-7 and PHQ-9. The severity breakdown of PHQ-9 scores was 56% (115 patients (pts)) minimal, 24% (49 pts) mild, 12% (24 pts) moderate, 8% (16 pts) moderately severe and 1% (3 pts) severe. The severity breakdown of GAD-7 scores was 61% (127 pts) minimal, 18% (37 pts) mild, 12% (25 pts) moderate, and 9% (18 pts) severe. For both anxiety and depression, the majority of patients with minimal symptoms reported no functional impairment (75% and 73%, respectively). For those with mild symptoms, 76% with depressive symptoms and 81% of those with anxiety reported being "somewhat" impaired. For those with moderate depressive symptoms, 13% reported no impairment, while 42% reported "somewhat," 38% reported "very," and the remaining 8% reported extreme impairment. Patients with moderately severe depressive symptoms reported 6% "not at all," 31% somewhat, 38% very, and 23% extremely impaired. Patients with severe depressive symptoms equally reported being somewhat, very, or individual sessions allowed for tailoring each session to meet the needs of the individual family. Future plans include revisiting the possibility of telehealth group for parents to allow parents to learn from each other's experiences and foster increased social support. Acknowledgments: Research supported by a grant from the Cystic Fibrosis Foundation. Background: Early development of the InformedChoices decision aid (DA) for CF advance care planning involved exploration of CF adults' information needs regarding invasive mechanical ventilation (IMV) and lung transplant (LT). We learned that CF adults felt it was necessary, for their own decisions, to hear personal stories by other CF people about their own experiences. In direct response, we collected CF adult and caregiver narratives about IMV and LT to include in the DA. We present results from analysis of the narratives.Methods: We recruited participants from CF clinics at Northwell Health, Univ. Pennsylvania and CFF Community Voice. Interviews were conducted in person or by web-conference. Open-ended questions focused on: 1) process (leading up to, during and after IMV or LT); 2) information seeking; and 3) feelings about experiences. Interviews were audio recorded and transcribed for thematic analysis.Result: Fifteen participants were recruited (10 CF adults and 5 caregivers); Participants were able to give personal accounts of experiences undergoing IMV and/or LT. Some described IMV in detail, with one person stating that "It felt like a railroad track going down my throat"; others described the moments prior to transplant surgery, and post-op recovery: "I had to relearn how to walk. I remember getting very weak because I couldn't pass my swallow test, so they wouldn't let me eat." Others described their depression post-transplant: "I don't know if I could have been prepared for the depression I felt, especially with the complications. It felt so debilitating because the thing I went in to have surgery for was fixed but now I had a new problem." Caregivers described the loss of their loved one from chronic rejection or the decision to decline IMV: "It wasn't about not wanting a breathing tube, it was the futility of the intervention." Participants also discussed topics clinicians never brought up, including ECMO or needing additional organ transplant. IMV due to respiratory failure was rarely discussed with 4/6 participants who experienced this having "no idea it could happen." Several LT recipients brought up their organ donors with 2 of them developing close relationships with their donor's family. Importantly, most LT recipients stated that they would not change their decision to undergo transplant despite complications: "It's still been a wonderful five years -look at the things that I couldn't do before. It's given me a whole new life, I can breathe!" Conclusion: Given these results, decision making for CF adults about late-stage care can be enhanced by hearing stories from others who have had similar experiences, especially because these experiences often cover topics not disucssed in clinic. By including narratives in our DA, we hope to offer CF adults the opportunity to learn from the experiences of others. Health, Univ North Carolina Chapel Hil, Chapel Hill, NC, USA Introduction: Routine visits to a primary care physician (PCP) allow for symptom management, anticipatory guidance, and wellness checks. Yet according to the CDC, in 2015, 23.3% of all adults ages 18-44 did not see a PCP for even a single visit. As the average age of cystic fibrosis patients rises and more patients enter higher risk categories for diseases like breast and colon cancer, anticipatory guidance is even more necessary. Because CF adults are already tasked with the burden of having frequent subspecialty visits, many do not maintain a relationship with a PCP but instead identify their pulmonary subspecialist as their primary care physician. However, physicians trained as pulmonary subspecialists may fail to implement important routine health screenings.Methods: A retrospective chart review was completed for a convenience sample of 92 adult CF patients, including all patients over the age of 40 followed in an adult CF clinic. Topics of interest were identified from the United States Preventative Services Task Force (USPSTF) Recommendations for Primary Care Practice. Charts were abstracted for grade A and B recommendations from the USPSTF to identify a pattern of missed screening opportunities.Results: In 59% of cases, the patient had no PCP identified, or the primary pulmonologist was listed as the PCP. Out of those with a separate PCP identified (38 patients), only 61% (23 patients) demonstrated evidence of specialist communication with the PCP. Out of 37 female patients who met age criteria (age >40 years) for screening mammography, only 8 (22%) were up to date on screening. 51 women were within the recommended age range for cervical cancer screening, yet only 12% (6 patients) had documentation of meeting USPSTF screening recommendations. Of the 92 patients reviewed, 50 met CF Foundation (CFF) guidelines criteria for screening colonoscopy. Of these, 24 patients (48%) had received a screening colonoscopy at some point in the past, with 19 patients (38%) having received testing in accordance with either CFF guidelines or the USPSTF. Due in large part to recent concerted quality improvement efforts by the CF team, only 6% of the included patients had not been screened for depression. Only 3% of patients had not been screened for diabetes.Conclusions: Collaboration between specialists and PCPs provides the most comprehensive care for patients. Most physicians would agree that maintaining up to date knowledge in both general internal medicine and subspecialty care is challenging; further, physicians who identify as a subspecialist may not see it in their realm of practice to perform primary care screening activities. In a sampling of CF adults seen at a major academic institution, many of the screening activities that a generalist routinely performs, such as breast cancer evaluations and sexually transmitted infection counseling, were missing. This suggests that these patients likely felt they are receiving comprehensive health care by their pulmonologist, even when missing several crucial screening examinations. Notably, screening rates were much higher when the complication or disease in question was specifically associated with CF, such as diabetes, depression screening, and colonoscopy.

Flewelling, K.D. 1 ; Sellers, D. 3 ; Sawicki, G.S. 2 ; Robinson, W. 1 ; Dill, E. 1 1. Psychology, University of Colorado Denver, Aurora, CO, USA; 2. Harvard Medical School, Boston, MA, USA; 3. Cornell University, Ithaca, NY, USA Introduction: A growing population of adults with cystic fibrosis (CF) has emerged in recent decades. Although social support has been linked to a variety of mental and physical health outcomes in those with and without chronic illness, this construct has rarely been studied in the adult CF population. The current study examines both predictors and outcomes associated with social support in adults with CF. Based on previous research, it was hypothesized that those who are older, male, who have less income and education, who are not married or employed, and who have lower FEV 1 scores and more pulmonary exacerbations would have less social support. Further, it was hypothesized that individuals who report lower social support would have poorer mental and physical health as well as treatment activity and disease-specific health-related quality of life.Methods: Participants in the study included 250 adults with CF who took part in an eleven-wave longitudinal study known as the Project on Adult Care in Cystic Fibrosis (PAC-CF). Participants were administered a battery of measures including a social support evaluation (Interpersonal

Fries, L.M. 1 ; Dunitz, J.M. 2 ; Brady, C. 9 ; Yeley, J. 6 ; Jansma, B. 7 ; Allada, G. 8 ; George, C. 3 ; Willey Courand, D. 10 ; Tomczyk, C. 11 ; Brown, R.F. 5 ; Benitez, D. 4 1. Medicine, State Univ of NY at Buffalo, Buffalo, NY, USA; 2. Univ of MN, Minneapolis, MN, USA; 3. CFF, Bethesda, MD, USA; 4. Keck Med Ctr of USC, Los Angeles, CA, USA; 5. Vanderbilt Univ Med Ctr, Nashville, TN, USA; 6. Indiana Univ Health, Indianapolis, IN, USA; 7. Children's Hosp of Philadelphia, Philadelphia, PA, USA; 8. OR Health and Sciences Univ, Portland, OR, USA; 9. Children's Hospitals of Minnesota, Portland, MN, USA; 10. UTHSCSA, San Antonia, TX, USA; 11. Univ (2):377-81) was sent to CF program directors (PDs) and APPs and compared to 2009 survey data to determine workforce gaps, APP scope of practice and training.Results: Responses were received from 124 PDs and 107 APPs. Of the PDs, 93 employed APPs in their programs. Reasons for APPs included job appropriate for APP (61%), service expansion (66%) and physician shortage (51%). Of the 56% of PDs recruiting physicians, 80% reported difficulty recruiting. The three qualities which APPs reported most important in a job are practicing at the top scope of practice, collaboration with physicians and schedule/work hours. Sixty-three percent of APPs serve leadership roles including QI leadership, CF clinic coordinator, registry coordinator, newborn screening, among others. Sixty-four percent of APPs practice at their top scope of practice. Reasons preventing APPs from practicing at the top scope of practice include lack of other staffing, poor understanding of APP role, PD responsibility delegation and hospital regulations. Only 10% of PDs report that they do not plan to use APPs in the future. Varying state, hospital, insurance and program regulations are barriers to APP billing and independent practice. Based on both surveys, training for APPs in CF is mostly on the job which 70% of the APPs felt was adequate, but 97% of APPs felt a more formal training program would be beneficial. PDs and APPs prefer the following training formats: online modules (40% PDs/48% APPs), year-long mentorship with senior APP (39% PDs/42% APPs), 3-month (40% PDs/39% APPs) or 1-month in-person fellowship (15% PDs/30% APPs).Discussion: APPs are an important part of CF clinical care teams. The care provided by APPs fills a gap in physician coverage and is uniquely tailored to a chronic care model due to the nature of APP training. The taskforce performed a literature search to evaluate current fellowship training opportunities for APPs new to practice or a particular specialty. Discussion is underway to develop a more formal training opportunity for CF APPs.Acknowledgment: Supported in part by the Vanderbilt Institute for Clinical and Translational Research grant (UL1 TR000445 from NCATS/ NIH).

Raraigh, K.S. 1 ; Langfelder-Schwind, E. 2 1. Institute of Genetic Medicine, Johns Hopkins University, Columbia, MD, USA; 2. The Cystic Fibrosis Center, Mount Sinai Beth Israel, New York, NY, USA Introduction: Genetic counseling in the cystic fibrosis (CF) clinic began in 1971 and has been repeatedly shown to provide value to patients, families, and CF clinicians, but inclusion of genetic counseling services as part of a multidisciplinary CF clinic has not been widely adopted. We sought to investigate the attitudes and barriers to provision of genetic counseling in CF clinics.Methods: Center directors and clinic coordinators at US CF Foundation-accredited CF centers were surveyed using a web-based tool distributed via email listserv.Results: The survey was completed by 125 center directors and clinic coordinators. Eighty-nine percent of respondents indicated that genetic counselors (GCs) provide a unique service to CF patients and their families and 84% indicated that GCs provide a valuable service in the CF center. Of clinics involving a GC as part of the CF team, 85% "strongly recommend" or "recommend" this relationship for all CF centers. Overall, 79% of respondents indicated that GCs at their institution had a "high level of understanding of CF"; however, when intragroup differences were examined, it was noted that 100% of those whose CF teams included a GC agreed with this statement, compared to 42% who indicated no routine GC involvement with their CF team. Involvement of a GC in CF newborn screening (NBS) follow-up was also examined, as nearly all states' processes include DNA analysis. However, nearly 40% of pediatric respondents indicated that their CF center does not regularly utilize or refer to GCs as part of NBS follow-up. Of those who do, 78% indicated that GCs improve efficiency of their NBS follow-up procedures, but differences were noted between groups: 94% of clinic coordinators indicated that GCs improved NBS follow-up efficiency vs. 65% of center directors. While only 39% of respondents indicated that their CF center staff has the time to provide genetic counseling regarding CF NBS, further analysis showed differing opinions between the groups: 29% of clinic coordinators indicated that their staff has time to provide genetic counseling vs. 50% of center directors. Barriers to the involvement of GCs and provision of genetic counseling services were identified through specific survey questions and from 65 open-ended responses and included: cost/benefit; demonstration of expertise and knowledge of CF; accessibility (due to scheduling or geographical constraints); and perception of need for a GC's services.Conclusions: Preliminary data suggests that center directors and center coordinators that have an engaged GC believe that genetic counselors provide a unique and valuable service to CF patients and their families, have a high level of understanding of CF, and are accessible to their patients. Those who do not work closely with GCs are more likely to report a negative view of genetic counselors' value, knowledge, and accessibility. Understanding the relationship between the CF centers' perceptions of access and referral patterns may play a key role in improving the connections between GCs and CF centers, and further exploration of the barriers to including GCs as part of the CF team may provide insights into the variability in practices among CF centers.

Ariefdjohan, M. 1 ; Towler, E. 2 ; Sagel, S. 3 ; Muther, E.F. 1, 3 1. Psychiatry, Univ. of CO School of Medicine, Aurora, CO, USA; 2. Breathing Institute, Children's Hospital Colorado, Aurora, CO, USA; 3. Pediatrics, Univ. of CO School of Medicine, Aurora, CO, USA Background: Adolescents with cystic fibrosis (CF) report high levels of depression and anxiety, with prevalence as high as 50% as compared to up to 17.5% among healthy counterparts. Unaddressed mental health (MH) concerns have been associated with poorer health outcomes. While the CF Foundation has invested in integrating MH into CF Centers, barriers still exist for those with CF to receive effective interventions to treat MH. Referrals to community MH providers also remain challenging. This study aims to 1) confirm the relationship between MH concerns and health outcomes in our CF center, 2) examine caregiver perceptions of adolescent MH concerns and barriers to community referrals, and 3) explore the possibility of whether frequency of contact with CF MH provider is correlated with changes in MH symptoms, as well as whether this relationship influences overall health outcomes.Methods: MH screening data (PHQ-9, GAD 7) from February 2015 to August 2016 (n=145) were analyzed to evaluate differences in lung function and demographic characteristics among patients with elevated scores (≥5) as compared to those without concern for depression/anxiety. An online survey was administered to caregivers of patients with elevated screening PICC consult team is for inpatient line placement only, they are unavailable after 1600, and no other options are available except for a peripheral line. This delay in care was an issue for both patients and providers. For many patients peripheral lines are not an option due to poor venous access. Currently patients wait in clinic for admission from 4-10 hours with no IV access.Objective: To improve the time from admission to the administration of first IV antibiotic dose.Methods: A multidisciplinary conference was held with the PICC consult team, CF provider and CF nurse to discuss methods to improve time from PICC insertion to initiation of IV antibiotics.Results: It was agreed that the PICC consult team would place the lines in clinic for CF patients being admitted for a CF exacerbation and IV antibiotics. Once admission is identified, the PICC team is notified and an order is placed for the PICC line. The PICC team inserts the line in the patient's clinic room. This process was initiated January 1, 2018 and since initiation, 100% of PICC lines were placed in the clinic while the patient waits for an inpatient bed. Due to implementing our new process the time from admission to IV antibiotics has decreased from an average of 8 hours and 50 minutes to 1 hour and 47 minutes. As well our patients verbalize an increased satisfaction with the new process.Conclusion: Since initiation of this process January 1, 2018 patient care has become more streamlined and we have greater patient satisfaction. Inpatient staff have verbalized that PICC insertion prior to admission has allowed for staff to utilize their time more efficiently. Based on the success of this process the adult center plans to look at additional aspects of care to identify areas that can be streamlined to improve patient care. The next step in this project addresses the need to start the IV antibiotics in clinic while waiting on an inpatient bed. (2):377-81) was sent to CF adult patients and families to evaluate access to APPs, understanding of differences between APPs and physicians, and comfort with the care provided by APPs.

Results: Responses were received from 121 people (62 adult patients, 58 parents/guardians and 1 spouse). Sixty-seven percent reported APPs providing care in their current CF program and 19% did not know if an APP was present. Of those who have access to APPs, 88% have received care by an APP. Clinical care provided by an APP was most commonly in the settings of routine (71%) and sick (60.5%) CF care although hospitalizations (33%) and hospital follow-ups (34%) were reported. In addition, 15% reported research study visits provided by APPs. Seventy-four percent of patients and families are comfortable receiving care from an APP although 81% are not comfortable with only an APP providing their care. Comments reported preference for a physician physically seeing the patient at least occasionally. The majority felt their APP would contact a collaborating physician to discuss complicated patients or concerns outside of an APP scope of practice with only 17% reporting less confidence that this communication would occur. Insurance coverage was reported to cover APP care by 54% of respondents, however 46% reported that they did not know if insurance covered this care.Discussion: The overall experience with APPs in CF programs as reported by patients and families is a positive one although many report they would still want a physician to physically see them or their family member at least on occasion. Many patients and families reported not knowing if APPs are involved in their care, so further education of the CF community is indicated.Acknowledgments Background: Due to a consistent incidence of intravenous piggyback infusion (IVPB) errors cystic fibrosis (CF) patients are placed at a higher risk of complications compared to other patients. Nurses caring for CF patients on the Advanced Pulmonary Care Unit (APU) give between 10 to 20 IVPBs, primarily antibiotics, in a twelve-hour shift. IVPB medications left clamped had been an ongoing problem on the APU with 35 reported cases of IVPB left clamped from January 1, 2016 through December 31, 2016. IVPB antibiotics left clamped can cause anaphylaxis in desensitized patients, an increase in antibiotic resistance, increased length of stay with an approximate cost of $1040.00 per day, and poor satisfaction in our patients, caregivers and physician partners.Objective: The purpose of this process change is to eliminate IVPB infusions from remaining clamped during IVPB setup and administration. In addition this change would increase patient, caregiver and physician satisfaction and prevent negative impacts due to IVPB left clamped.Method: A multidisciplinary team worked to improve the process of administering IVPB medications left clamped, by developing a "no clamp method" (NCM) when using IVPB tubing to administer medications. To ensure compliance of process change, registered nurses (RN) were individually educated and signed a contract regarding RN's agreement to use NCM. Post-implementation was done by tracking IVPBs left clamped, occasional audits of RNs using the new process and monthly reminders to RNs.Conclusion: The APU has had less occurrences since beginning NCM. In 2016 there were 35 IVPBs left clamped reported compared to 16 cases in 2017. In November of 2017 there was a spike of 4 cases. Audits and reeducation was done with a decrease back to a baseline of 2 or less cases a month. This process change has helped our team establish a comprehensive way to reduce IVPBs left clamped. This has resulted in safer practice for patients, and an increase in satisfaction for all parties.

Bray, L.; Ladores, S.L.; Burgess, B.E.; Mrug, S. Univ. of Alabama at Birmingham, Hoover, AL, USA Background: With the life expectancy of individuals with cystic fibrosis (CF) doubling over the last two decades, it is critical that healthcare providers address both quantity and quality of life. Health-related quality of life (HRQoL) encompasses physical, social, psychological, and functional aspects of daily living and how CF impacts each one. Women with CF consistently report having an overall poorer quality of life across international studies, but the cause of the gender difference is unknown. The purpose of this study was to explore gender differences in HRQoL and then build upon those results to better understand gender-specific facilitators and barriers to HRQoL.Methods: Gender-specific facilitators and barriers to HRQoL were explored using a sequential explanatory mixed methods design. The sample included 123 adults with CF who were recruited within a tertiary care center in southeast United States. Data collected included the CF Questionnaire-Revised (CFQ-R), which is a disease-specific HRQoL questionnaire, a demographics form, and CF Foundation Patient Registry data. The CFQ-R has 12 separately scored domains of HRQoL (physical functioning, body image, social functioning, health perceptions, respiratory symptoms, treatment burden, emotional functioning, role functioning, vitality, eating problems, digestive symptoms, and weight). Scores range from 0 to 100, with higher scores representing better HRQoL. The CFQ-R scores and descriptive statistics were analyzed using SPSS Statistical Software v. 23. A subsample of 15 men and 15 women completed a 30-45 minute, semi-structured interview to build upon the survey results and described gender-specific facilitators and barriers to HRQoL. The interviews were transcribed verbatim and analyzed using Braun and Clarke's method of thematic analysis and NVivo 11 software.Results: Fifty-seven men and 66 women with CF, aged 19-67 were included in the quantitative analyses. Though significance was not reached (p< 0.05), women reported a better HRQoL compared to men in the areas of body image, weight, and digestive symptoms. In the remaining nine areas, women reported a poorer HRQoL than men, with statistically significant differences in the areas of physical functioning (p = 0.02), social functioning (p = 0.03), emotional functioning (p = 0.02), and weight (p = 0.02). Five main themes emerged from the qualitative data that described facilitators and barriers to HRQoL: 1) Biological and Physiological Factors; 2) External Factors; 3) Functional Status; 4) Perceptions of Preferences, Values, and Mental Health; and 5) Perceived Symptom Status. Facilitators and barriers specific to men were associated with assuming male roles, utilizing counseling resources, and being open with others about having CF whereas, women reported not wanting to be labeled by others, being content with thinness, and social media.Conclusions: Factors that contribute to HRQoL are complex and diverse. These results show where the gender differences in HRQoL occur and provide insight into potential facilitators and barriers that could be used to improve the HRQoL of adults with CF. Acknowledgment: Supported by UAB Center for Palliative and Supportive Care.

Paca, K.; Devoogd, R.; Brennan, L.; Towler, E. Children's Hospital of Colorado, Denver, CO, USA Background: Pulmonary exacerbations (PEx) are a frequent cause of hospital admissions for patients with cystic fibrosis (CF) and intravenous (IV) antibiotics are part of the standard of care for PEx. Patient /Family satisfaction surveys at our pediatric institution identified complaints due to a delay in IV access and first antibiotic administration upon admission in our CF patients. Because this delay could negatively impact quality of care and outcomes, we examined our process of obtaining IV access and timing of first antibiotic dose.Objective: To improve timeliness of peripheral IV (PIV) or port access and administration of the first IV antibiotic dose within two (2) hours from time of admission.Methods: A literature review was done to determine a standard time frame for first antibiotic administration in CF, but little is published. A process map was developed to delineate the steps involved in admitting a patient for a PEx. A chart audit over a one-year period was performed; data collected included patient origin, presence of an admission letter from a provider, arrival time to hospital unit, timing of antibiotic order and verification, time and type of IV access, type of RN obtaining access, and time the first antibiotic was given. Nurses were surveyed regarding barriers to initiating IV antibiotic therapy. Potential barriers included lack of supplies, patient/family obstacles, no orders from medical team, lack of time or ability to start IV, difficulty obtaining IV access itself, antibiotic unavailable from pharmacy, and prioritizing admissions for CF patients experiencing PEx. Two plan-do-study-act (PDSA) tests of change included (1) making a "PIV Kit" including necessary supplies for starting IVs and (2) changing the admission order set to include an order to obtain IV access Results: In the baseline audit, time from admission to IV access (via PIV or port) ranged from 0.38 to 19.42 hours (hrs) with a median of 2.53 hrs. The time from admission to first antibiotic dose ranged from 0.9 hrs to 22.80 hrs with a median of 3.80 hrs. During the PIV kit trial, the time from admission to PIV placement ranged from 0.52 hrs to 5.3 hrs, with a median of 2.2 hrs; and the time from admission to first antibiotic dose ranged from 1.23 hrs to 6.17 hrs, with a median of 3.45 hrs. After the admission order set modification, the time from admission to IV access ranged from 0.23 hrs to 4.03 hrs, with a median of 2.33 hrs. Time from admission to first antibiotic dose ranged from 1.87 hrs to 6.32 hrs, with a median of 3.28 hrs. These changes were not significant.Conclusions: After two PDSA cycles, we observed a small but not significant improvement in overall median time to first antibiotic administration on admission. There is significant variability in timing, and we are not meeting our goal of 2 hours or less in many patients. The PIV kit seemed to have the largest impact on timing to IV access. Continued data collection is underway to determine further areas for improvement. Results from the nurse survey also identified opportunities for education surrounding the importance of timely IV access and antibiotics in CF exacerbations and impact on patient care and satisfaction. Background: Most patients with cystic fibrosis (CF) who require inpatient admission to Cincinnati Children's Hospital are hospitalized on our Complex Pulmonary Unit (A7C1). Several years ago, this unit implemented nurse-led rounds in order to improve nurse-physician communication, patient care, and patient-family experience. However, we found that bedside nurses often could not be present at morning rounds due to the timing of many other patient care needs, which conflicted with the unit morning rounding time. We also found that participation in morning rounds was not always a priority for bedside nurses, since they did not appreciate the value of their participation.

Objective: To increase the participation of bedside nurses during morning multidisciplinary rounds.Methods: Members of our A7C1 Operational Excellence (OpEx) team met in November 2017 to review and improve our NRC Health (National Research Corporation) data regarding nurse-physician communication. There was consensus among team members that participation of bedside nurses during morning rounds was a key driver toward improving nurse-physician communication and patient-family experience. Using the "5 Whys," we determined that the 8:40 AM start time for morning rounds conflicted with the timing of many patient assessments, treatments and medications, and precluded consistent involvement of the bedside nurse. We noticed that bedside nurses were often more able to participate on Tuesdays, when our unit morning rounds started at 9 AM, due to hospital-wide Grand Rounds. We hypothesized that shifting our rounding start time to 9 AM would enable more bedside nurses to participate in morning rounds. Initially, we shifted our rounding time from 8:40 AM to 9:00 AM. However, our resident team felt that they did not have enough time after rounds to complete patient care activities before their didactic lectures. Consequently, we changed our rounding start time to 8:50 AM. Through the "5 Whys," our A7C1 OpEx team also determined that bedside nurse engagement was a key driver toward improving participation in morning rounds. To improve bedside nurse engagement in morning rounds, another intervention that we trialed was the use of a checklist which gave the bedside nurse ownership of specific information to be shared with the team during morning rounds.Results: Qualitative feedback from A7C1 nurses was excellent. Bedside nurses were able to complete most patient care tasks before the start of morning rounds, and better able to participate in nurse-led morning rounds. Furthermore, bedside nurses felt that their contributions during rounds were appreciated and valued. Quantitative measurement of bedside nurse presence on morning rounds during implementation of these interventions with compliant caregivers to have limited stores of pancreatic enzymes in the patient room safe. Easy access for self-administration of enzymes was allowed rather than the traditional method of nurse administration. Utilizing the primary nurse concept with these CF admissions allowed patient/nurse bonding and familiarity with food preferences, motivational factors, and compliance issues, suggestions/recommendations. Methods: One month post-implementation, a satisfaction questionnaire was administered at the end of hospitalization for participating CF patients. The study measured patient level of satisfaction with immediate access to bedside pancreatic enzymes versus traditional nurse administration. The study also measured primary nurse preference between the traditional methodology and self-administration of bedside enzymes. The results were analyzed using the 5-point Likert satisfaction scale with 5 being very satisfied, 1 being not satisfied.Results: A total of 14 primary nurses were surveyed one month post-implementation, with 12 preferring bedside enzymes and 2 preferring traditional nurse administration. Of the 11 CF patient survey participants, 9 experienced a decrease in wait time from the traditional 10-20 minutes to 5-10 minutes. Two patients had less than 5-minute wait times with self-administration. Eight of 11 patients using the new practice, scored a "5," very satisfied, or a "3," satisfied.Conclusion: Hospitalized pediatric CF patients had higher levels of satisfaction with self-administration of pancreatic enzymes. Immediate access to enzymes stored in the room's safe increases their autonomy and parallels a self-care practice usually performed at home. Additionally, a decrease in wait times optimizes nutritional growth and healing while hospitalized. 4 1. Pulmonary Medicine, Yale, Madison, CT, USA; 2. Yale Cancer Center, New Haven, CT, USA; 3. Yale New Haven Hospital, New Haven, CT, USA; 4. Yale University School of Medicine, New Haven, CT, USA Background: Cystic fibrosis (CF) represents a relatively small patient population that is medically complex and requires a multidisciplinary team approach to provide effective care. As the CF population ages, the number of hospital admissions will increase, due to a larger patient pool and an increase in disease complications. As such, inpatient nursing staff must have a basic understanding of the disease and its medical management to provide satisfactory care. This was identified as a programmatic need when adult patient admissions were centralized to one floor at Yale New Haven Hospital. A survey was administered to inpatient nurses intended to identify gaps in knowledge, and to inform future education interventions to ultimately improve patient care and patient experience.Methods: A review of Pubmed and Cochrane databases was conducted using keywords: inpatient, guidelines, management, nursing assessment tool, and cystic fibrosis, to identify patient-centered perceptions of inpatient care. Potential gaps in nursing (RN) knowledge were identified based on these results. To assess RN knowledge regarding CF, a questionnaire in use by our CF center's pediatric program was adapted for use in the adult inpatient unit at Yale New Haven Hospital. The 20-question survey consisted of questions across 5 disciplines: Medical Provider, Respiratory Therapy, Physical Therapy, Mental Health and Nutrition. The questionnaire was administered to 30 floor nurses on the floor in August 2017.Results: 13 out of 30 nurses successfully completed the assessment. Scores ranged from 40 -75 with a mean score of 75 out of 100. No one scored 100 on the pre-test. Gaps in knowledge were identified in all 5 disciplines. The greatest knowledge gaps were in respiratory and physical therapy knowledge, both with 62% incorrect answers. Nutrition and mental health had 8% incorrect answers, patient care 13%, general knowledge 20%, and medications 26% incorrect answers to the survey. The questions answered incorrectly most frequently were related to nebulizer cleaning and airway clearance (77% and 69% incorrect answers respectively).Conclusion: This quality improvement project sought to assess inpatient nurse understanding of CF to improve patient experience during hospitalization. The assessment identified areas of knowledge gaps that are being addressed by CF team-based education. Additional interventions include inpatient nurses joining the adult CF team for outpatient clinic visits. The educational program will be reassessed using a post-education assessment to continue to identify areas of improvement. 4 ; Miller, E. 5, 6 ; Sawicki, G.S. 1 1. Boston Children's Hospital, Boston, MA, USA; 2. Institute for Healthcare Improvement, Boston, MA, USA; 3. Boston University School of Public Health, Boston, MA, USA; 4. Beth Israel Deaconess Medical Center, Boston, MA, USA; 5. Children's Hospital of Pittsburgh of UPMC, Pittsburgh, PA, USA; 6. Center for Women's Health Research and Innovation, Pittsburgh, PA, USA Background: As health outcomes continue to improve, sexual and reproductive health (SRH) has become an important disease-specific concern for adolescent and young adult (AYA) women with cystic fibrosis (CF). However, SRH is not traditionally addressed in the CF care model. This study aimed to identify the educational needs and preferences of interprofessional CF care team members regarding SRH in AYA women with CF.Methods: Interprofessional CF care team members were recruited to participate in an anonymous online survey via provider-specific listservs and an international conference mobile app. Questions explored provider knowledge and skills in providing general and CF-specific SRH information, factors for optimizing SRH care, and preferred SRH topics and formats for additional training. Results were summarized using descriptive statistics.Results: A total of 523 providers completed the survey. Among participants, 39% were physicians/NPs/PAs, 19% nurses, 20% social workers, and 22% were from other disciplines. Forty-nine percent cared for pediatric patients, 28% for adult patients, and 24% for both. Seventy percent felt comfortable asking family members or partners to leave the room during SRH discussions with AYA women and 66% knew the policies around patient SRH confidentiality. Half (51%) felt they knew how to respond to the SRH concerns of AYA women with CF; however, only 33% reported they knew which questions to ask and 32% felt confident taking a sexual history. Only 29% were comfortable with their current CF-specific SRH knowledge. Regarding CF-specific SRH concerns, 44% were comfortable providing CF pre-pregnancy counseling and 41% providing contraceptive counseling, yet only 19% were comfortable managing urinary incontinence and 11% addressing sexual functioning concerns.Participants identified the following key SRH focus areas for further provider training: pregnancy/parenthood planning, sexual functioning, The University of Mississippi Medical Center provides the only accredited CF center in the state of Mississippi. Therefore patients may drive up to 3 hours to the center. A more convenient way to provide face-to-face education to our patients and families was needed. The goal of this project was to provide education to a larger patient population by increasing attendance at Cystic Fibrosis Patient and Family Education Days.Methods: A spreadsheet was developed including the addresses of the entire pediatric and adult CF population. The areas of highest patient populations in the state were then assessed. These areas were narrowed down to 5 locations to host CF Education Days. The locations were no more than a 1 to 2 hour drive for the target population in the area. Dates were chosen throughout the 2017 calendar year and local restaurants booked to host the events. Patients and parents were called to assess if a dinner during the day or at night would be most convenient for them. Invitations to the Education Days were distributed by mail to each of the patients. Based on family/ patient preferences, 4 events were held at night while 1 was held during the day each during a 2-hour timespan. A meal was served to the attendees. A center update and education regarding improving nutrition in CF patients was presented by the social worker, dietitians, respiratory therapist, nurses and nurse practitioners. Evaluations were collected from the attendees at the end of each program to assess the program and location.Results: In 2015 the University of Mississippi CF center had 7 attendees representing 4 families at Education Day. In 2017, we visited 5 locations around the state and had 40 attendees representing a total of 25 patient families. Some of the comments from the evaluations were as follows:"Having flexible locations that reaches out all over the state is very helpful!" "For a new learner of CF health issues this has been very informative, thank you all for your time!" "We really appreciate ya'll coming to us. Thank you!" "So glad we were able to attend. By it being close to our hometown we can come every time ya'll have one."There was an overwhelming positive response to the 5 locations across the state by all of the attendees. Conclusion: By providing 5 different locations for CF Education Days attendance was increased from 7 attendees in 2015 to 40 attendees in 2017. This method of hosting Education Days has proven successful in providing education to the CF families/patients. For 2018, the CF team is once again traveling across the state to 5 different locations focusing education efforts on pulmonary issues.

Cotton, C. Pulmonary Medicine, UNMH Adult CF Center, Albuquerque, NM, USA Background: Cystic fibrosis (CF) is an inherited autosomal recessive disorder caused by mutations of the CFTR gene. The CF mutation 3849+10kbC>T is rare, often manifesting with more mild disease, often a single organ affected, diagnosed later in life, typically associated with normal/intermediate sweat chloride levels. In the world population the prevalence of CF patients with this mutation is approximately 0.2% (1), in Europe 0.82% (2) and in the United States 1.6% (3). We describe a high prevalence of this mutation within certain patient populations in the CF center at University of New Mexico.Problem: In 2015-2017, 7 adult patients diagnosed with CF at our center were noted to have at least one copy of 3849+10kbC>T mutation. All of these patients self-reported as Native Indian, receiving care at pueblo clinics. With late diagnosis, patients often presented with severe lung disease and little improvement despite initiating aggressive therapies. This late diagnosis is due in part to atypical presentation and if CF is suspected, negative or indeterminate sweat test result in a lack of further work-up.Method: To determine the best methods in educating the outlying providers/population, discussion with pueblo providers and three of the newly diagnosed patients helped direct the methods used. These methods were in the form of PowerPoint presentations to pueblo staff, an educational article directed to the public in a pueblo newsletter (which is also distributed to other pueblos), teleconferences and discussion with a pueblo disability council.

Since the initiation of the educational programs/information there has been one patient diagnosed as a direct result of these educational methods as noted by their primary provider. After educational discussions pueblo providers have identified multiple patients that need further screening. This is an ongoing process as more pueblos request information.Conclusions: Prior to these educational programs, our data supported that patients with this mutation tended to go undiagnosed until later in life. This is due to atypical presentation, negative or indeterminate sweat chloride which often halted any further evaluation for CF. Since initiation of the education one patient has been diagnosed. Positive feedback from all the participating pueblo clinic staff reinforces the indication of previous lack of information about CF in this particular patient population. Our hope is with continued education, open communication with pueblo clinic staff and CF clinic availability other patients will be diagnosed much earlier and initiate earlier interventions.References : Introduction: Cystic fibrosis (CF) is a genetic disease with a high treatment burden that impacts all areas of a patient's life. In addition, patients with CF have limited contact with each other due to infection prevention and control guidelines. As a result, patients may feel isolated in caring for their CF. In order to improve education and support for patients, virtual peer support groups were developed and conducted via an online video platform. The goal of the online groups was to educate and empower patients to be active participants in their care while increasing social support through connection with other CF patients.Methods: Cohorts of patients were identified and invited to participate in the groups (group 1-stable patients, group 2-young adult patients who use exercise as ACT, group 3-patients with frequent exacerbations). Each group was to include 10 patients and to run for 6 weekly, standardized sessions. The HIPAA-compliant BlueJeans video conference platform was used and patients had the option to use mobile phones, tablets, or computers to attend the group. Patients received a schedule of the sessions, which were moderated by CF care team members. The telemedicine dept provided support for the program. Session topics included personality styles, co-production, lifestyle change, nutrition, time management, mental health, home IVs, hospitalizations, and CF therapies. The curricula were modified to meet the needs of each group. At the conclusion of the group, patients were surveyed to assess effectiveness of the group and areas for improvement.Results: Group 1 had 10 participants. Post-completion survey indicated 100% of the participants strongly agreed they would recommend the group to others, 100% felt the group met expectations (20% Agree, 80% Strongly Agree), and 100% felt comfortable speaking and sharing during sessions (20% Agree, 80% Strongly Agree). Group 2 did not occur due to lack of interest in the identified cohort. In Group 3, two of 10 confirmed participants participated. They expressed strong satisfaction with the group but acknowledged that increased attendance may have enhanced the experience. The formal post-survey has not yet been distributed, results will be included in the final presentation. Challenges of the virtual groups included patient availability/scheduling, technological difficulties, patient lack of access to technology, and lack of patient attendance despite confirmed interest in the group.Conclusion: An online group can be an effective way to improve education and enhance support for patients with CF while providing a meaningful experience for both patients and care team members. However, there are barriers to implementation, including lack of patient attendance 4 ; Stecenko, A. 4 ; Bellcross, C. 2 ; Wei, J. 5 1. Cystic Fibrosis, Children's Healthcare of Atlanta, Atlanta, GA, USA; 2. Genetic Counseling, Emory University School of Medicine, Atlanta, GA, USA; 3. Genetics, Northshore University Health System, Evanston, IL, USA; 4. Pulmonology, Emory University, Atlanta, GA, USA; 5. Emory University, Atlanta, GA, USA Background: Adolescents with cystic fibrosis (CF) often lack the necessary knowledge about CF-specific inheritance and fertility issues to make informed decisions about family planning. There are few evidencebased educational tools for addressing these knowledge gaps. Our aim was to develop an education self-directed module (SDM) on genetics, inheritance, and reproductive issues, and compare knowledge improvement to in-person genetic counseling (GC) on the same topics.Methods: CF patients aged 15-23 years were randomly assigned to complete the SDM or an in-person GC session. Knowledge was measured using a questionnaire several months before and immediately after the intervention. We used paired t-tests to assess improvement in overall score and in each category of the questionnaire (Inheritance, Carriers, and Reproduction). Relative improvements between the GC and SDM groups were compared using independent samples t-tests.Results: A total of 38 participants completed the study, with 19 participants in each intervention group (GC and SDM). From pre-intervention to post-intervention, mean knowledge scores increased from 49% correct to 66% correct (p=.001) in the SDM group and increased from 46% correct to 74% correct (p<.001) in the GC group. The GC group improved significantly in all three questionnaire categories (p≤0.001), while the SDM group improved significantly only in Inheritance and Reproduction (p≤0.011). Overall improvement was not significantly different between the two groups (p=0.070), though this may reflect a lack of statistical power. The GC group did improve significantly more in the Carriers category than the SDM group (p=0.031).Conclusions: The significant improvement from pre-to post-intervention provides evidence for both genetic counseling (GC) and the self-directed module (SDM) as interventions for targeting the significant knowledge gaps in adolescent CF populations. Our results suggest that the SDM may be less effective in educating patients than a GC session, but a larger study is needed to confirm this. Given that many CF clinics are unable to recruit genetic counselors, the SDM provides a reasonable alternative for the purpose of patient education. London, London, United Kingdom Background: Standardization of the nitrogen multiple breath washout (MBW) test for clinical trials includes the training and qualification of MBW operators. In response to feedback and the changing needs of the CF community, a 2-day instructor-led training session for the Exhalyzer ® D (EcoMedics AG, Duernten, Switzerland) MBW device was condensed to a single day hands-on training with supplemental eLearning modules.

Objective: To evaluate the addition of eLearning modules within the restructured training session.

ELearning modules were developed using the Articulate RISE web-based platform. Participants completed a pre-training module prior to the hands-on session. Pre-training topics included physiological principles of MBW, Exhalyzer ® D equipment and a pre-training quiz. After the 1-day hands-on training session, participants completed the posttraining module which consisted of more advanced topics including testing preschool age children and MBW test quality control. Participants were asked to complete a knowledge test and feedback survey at the end of the training.Results: To date, 28 participants from 17 sites have completed the revised MBW training program, of which 61% (17/28) were naïve to MBW testing. Most of the participants indicated that they were confident in conducting a MBW test with an "agreed" or "strongly agreed" rating on a 5-point Likert scale (26/28; 93%). This compared favorably to feedback from the previous 2-day training program, where the reported confidence in conducting a MBW test was 83% (20/24). Furthermore, all responded that the eLearning resources enhanced the overall learning experience (28/28; 100%). Although some participants reported barriers including initial technical difficulties in accessing the eLearning modules (9/28; 32%), which have since been addressed, there was overwhelmingly positive feedback regarding the design and ease of use of the eLearning modules (28/28; 100%).The average score for the pre-training quiz was 80%, and participants indicated that the pre-training modules prepared them for the hands-on session (26/28; 93%). After completing the post-training module, operators felt confident in recognizing a successful MBW test (23/28; 82%). The post-training knowledge test scores also demonstrated proficiency in understanding of MBW test quality (average score of 82%). To date, 100% of trainees have successfully completed the qualification process for testing children 6 years and older; an improvement from the completion rate observed in the previous 2-day training program (91/102; 89%) (Isaac SM, et al. Pediatr Pulmonol. 2016 ;51(S45): 228).Conclusion: The eLearning MBW training modules are a valuable resource to complement the revised hands-on training as it offers trainees the flexibility and ability to learn at their own pace. Despite the abbreviation of the instructor-led training time, the combined online eLearning and hands-on approach exceeded the prior training qualification success and has the potential to enhance both the effectiveness and efficiency of the MBW training program. Background: Due to medical advances over the last few decades, most patients with cystic fibrosis (CF) are living well into their adult years. Many contemplate parenthood, though CF can have significant effects on fertility and health. A large number of patients report never having any consultations about fertility issues with their healthcare providers, despite noting their desire to have these discussions. Previous studies have shown that many individuals with CF have poor understanding of CF inheritance and genetics, though no studies have been conducted exploring this patient population's perceptions about different family building options. In this study, our aim was to assess whether a genetic counseling intervention would be associated with a change in knowledge and/or perceptions about genetics, fertility, assisted reproductive technologies and family building options among adult patients with CF.

Methods: Adults 18 years and older presenting to the Emory+Childrens' Adult Cystic Fibrosis clinic were approached for study inclusion (IRB 00102445). Participants completed a pre-intervention survey to measure their knowledge of CF genetics as well as their perceptions and understanding of assisted reproductive technologies and other family building options. Following the initial questionnaire, the subjects partook in an was an employee of Novartis Ireland Ltd. during the conduct of the study.

Gajo, M.; Kim, J.; Keens, T.G. Children's Hospital Los Angeles, Los Angeles, CA, USA Background: The care for cystic fibrosis (CF) patients is complex and often involves multiple chronic medications and long treatment times. It is well-established that a pharmacist is valuable in the ambulatory care setting. Incorporating a CF pharmacist in a multidisciplinary team presents an opportunity to guide prescribers in optimizing a medication regimen and to assist patients in improving their medication knowledge. Pharmacists may be valuable especially in the pediatric setting and have been a part of the Children's Hospital Los Angeles (CHLA) CF multidisciplinary team since 2013. Children may be at greater risk of medication errors and poor medication adherence due to unique challenges such as off-label use of medications, frequent dose changes, lack of pediatric-friendly formulations, and variable pharmacokinetic and pharmacodynamic profiles. Therefore, the objective of the study was to evaluate the role of the pediatric pharmacist in the multidisciplinary CF clinic setting.Methods: A retrospective chart review was completed from July 1, 2013 to March 31, 2018. Inclusion criteria were patients with a diagnosis of CF taking one or more medications. A few patients who were not seen by a pharmacist were excluded from the study, and they were also excluded if they had no medications. Each patient's medication regimen was reviewed by the pharmacist to assess the appropriateness of therapy. The interventions made in the clinic were documented and categorized.Results: There were 2,398 total patient encounters for 248 patients during the study period. There was a total of 1,148 pharmacist interventions documented. Pharmacists made an average of 2.6 interventions per clinic day. Of these interventions, 67% were made specifically for patients and/ or caregivers, and 33% were made for the prescribers. The most common categories of interventions included correction of medication dose or frequency (22%), optimization of medication scheduling (16%), correction of inhaled medication administration sequence (8%), and dosage recommendations for the prescribers (8%). Of the total number of interventions, 98% were documented as accepted and appropriate changes were made based on the pharmacist's recommendations. Examples of infrequent but notable interventions included patient referrals to subspecialists for followup, recommendations to order labs for monitoring, and identification of undocumented allergies.Conclusion: We conclude that pharmacists made significant and impactful interventions during the study period. We speculate that the pharmacist plays a valuable role in a multidisciplinary CF clinic setting by optimizing the patient's medication regimen while improving their medication knowledge. Introduction: Food insecurity is a barrier to care that has emerged through interactions between Cystic Fibrosis (CF) Foundation Compass, care centers and CF families. Food quantity and quality are vital for people with CF to manage their disease, making investigating this social determinant of health (SDOH) important. To better understand the scope, barriers, impacts, and interventions related to food insecurity in the CF population, we conducted a needs assessment.

Methods: Between Dec. 2017 and May 2018 Compass conducted a needs assessment to assess food insecurity in CF. This included reviews of published research and reports on SDOH and food insecurity, existing CF-specific literature and research, existing programs and resources, food-related Compass calls, and 3 informal surveys. The first survey (12/2017) collected information via discipline listservs on use of Supplemental Nutrition Assistance Program (SNAP), Special Supplemental Nutrition Program for Women, Infants, and Children (WIC) and public housing. The second survey (12/2017) gathered emails from CF social workers to understand food insecurity screening procedures and interventions. The third survey, given at the Nutrition and Social Worker Consortium (4/2018), collected data on the perceived scope of food insecurity in the CF population.Results: The first survey yielded 45 respondents, the second 20, and the third 94. Results from surveys 1 and 2 demonstrate food insecurity is problematic in CF. In the third survey, 54% of respondents indicated 20-39% of their population experience food insecurity while 17% indicated 40-59% of their population are food insecure. In relation to WIC, 37% feel the benefits are sufficient while 12% indicate SNAP benefits are sufficient for people with CF. Clinicians estimated 85% of families are experiencing intermittent or constant food insecurity. One respondent commented, "families just get used to living without." Less than half of CF clinicians (44%) actively screen for food insecurity.Compass has fielded 110 cases related to food insecurity since Feb. 2016. Of these, 67% of callers identified one or more financial needs in addition to food such as housing, transportation, utilities and medical bills. These findings were similar to the third survey results; respondents noted families primarily struggled with transportation, housing and utilities.Respondents from the surveys indicated unique characteristics of food insecurity in the CF population such as the need for increased calories, the cost of medical care which limits resources to purchase food and the cost of high calorie/quality food.Conclusion: Little data and research exist on food insecurity in the CF population; more research is imperative to fully assess the impact on people with CF and families. High medical expenses limit available resources to purchase food needed for the high calorie diets, increasing the likelihood of food insecurity. Clinicians noted federal programs are insufficient to meet the nutritional needs of people with CF; the means to address and intervene are largely dependent upon community resources. Additionally, families prefer not to apply for federal programs due to a perceived stigma. CF clinicians expressed overwhelming interest in this topic to learn more about food insecurity screening and to identify potential interventions.

Patel, K.; Martin, M.; Kirchner, K. Children's Healthcare of Atlanta at Scottish Rite, Atlanta, GA, USA Background: Cystic fibrosis (CF) patients have increased energy needs due to pancreatic insufficiency (PI), infections, increased work of breathing, and higher baseline metabolic rate. Nutrition supplements are recommended when oral intake is inadequate to meet caloric needs. Furthermore, CF patients have increased needs for fat soluble vitamin supplementation due to malabsorption associated with PI. The out-of-pocket cost for nutrition supplements and vitamins can be burdensome. In 2016, the HealthWell Foundation created a fund aimed at providing financial assistance for patients with CF to help cover the cost of FDA-approved medications, vitamins and nutritional supplements. To qualify for the grant, patients must have health insurance and meet grant income guidelines. In 2016, the combined grant amount for treatments, vitamins and nutrition supplements was $15,000. Currently the grant provides $15,000 for treatments and $2,000 for vitamins and nutrition supplements.Objective: The purpose of this review was to examine nutritional outcomes in CF patients that utilized the HealthWell Foundation grant for nutrition assistance at Children's Healthcare of Atlanta at Scottish Rite.Methods: Patients approved for the grant between January 2016 and March 2018 who used funds for nutrition assistance were identified through the HealthWell Foundation provider portal. Using linear regression models, associations between dollar amounts paid, grant duration, and change in vitamin D, weight, and BMI percentiles were analyzed. Models were adjusted for baseline (pre-grant) measurements.Results: Over the three years, 26 of 46 (57%) approved patients used the grant for nutrition assistance. The mean duration of grant was 18 months. The mean amounts used for nutrition assistance were $738.21 in 2016, $533.63 in 2017, and $79.50 in 2018. Grants for 2017 and 2018 were still open for use at the time of this review. Grant duration was positively associated with positive weight change (kg) (β=0.33, SE=0.13, p=0.0206). There were no significant associations between grant amount or grant duration and change in vitamin D levels or BMI percentiles.Discussion: Positive weight trends were noted in patients using the HealthWell Foundation grant. The grant was underutilized with 43% of approved patients not using any funds for nutrition assistance and an average of only 4.9% (2016) of the awarded amount was used for nutrition assistance. As a result of this review, we are revising our EPIC nutrition note to better identify and enroll eligible patients. We are more proactively using the provider portal to track grant use and expiration. This review emphasized that improved education of clinic staff, pharmacies, and CF families is important to ensure this grant is used to its full potential, as it may positively impact nutritional outcomes if the maximum benefit is applied. Background: Folia is a cloud-based platform available on web and mobile devices that enables the capture of home-reported outcomes (HROs) by family caregivers and patients (families). Datapoints collected include symptom presence, severity, characteristics; treatment adherence; biometrics; problematic behavior frequency; and wellness indicators. Folia utilizes multiple-choice questions to simplify data capture. Data are reported to the provider in a PDF document at or before an appointment.

Objectives: 1) Demonstrate the effectiveness of in-clinic distribution to enroll cystic fibrosis (CF) caregivers in an HROs program; 2) Establish the ability of Folia to engage caregivers in capturing HROs.Methods: The Folia team engaged with the Maine Medical Center pediatric CF Clinic ("the Clinic") to pilot in-clinic distribution over 6 months from Oct 11, 2017 through Apr 11, 2018 (new-patient enrollment continues). During appointments, clinicians offered a one-page informational flyer with an invite code for access to Folia. Families were invited to sign up online. In addition to in-clinic enrollment, the Clinic mailed a letter to all families. The Folia team also collaborated with the Southern ME CF Family Advisory Board to hold an in-person event for interested families.To measure the success of initial in-clinic enrollment, we collected a daily count of new Folia users who registered using the invite code. We measured Folia engagement via basic usage statistics, including average number of log-ins per week, ratio of monthly active users vs registered users, and total Folia questions answered. We administered a qualitative survey at 8 weeks. At participant request, we added referral codes for CF caregivers outside of ME.Data: Total clinic population of 80 families; 78 visited clinic. Approximately 45 (58%) received in-clinic information; all received mailing. Registrations: 37 with ME invite code; 44 with referral codes. Engagement: 1.9 log-ins/week/registered-user; 28 active users (35%); 24,560 answers as of May 15, 2018 (877/active-user) . analysis. The types of assistance provided were categorized for reporting purposes.Results: The data indicated that, out of 32 people with CF and their families affected by a disaster, 12 needed assistance accessing or replacing damaged medications and devices, 9 needed help finding a shelter or temporary housing until the storm passed, 3 needed a generator as an alternate/back-up source of electricity, 3 needed help paying for home repairs caused by storm damage, 2 requested guidance on effects of smoke inhalation in people with CF, and 1 family requested tips on disaster and emergency preparedness.Conclusion: Having an emergency preparedness plan during a natural disaster or emergency is important for everyone, but especially critical for people living with a chronic illness. People with CF must have access to medications, treatments, water and shelter, to sustain their daily care. Proactive measures taken by people with CF may help with preparedness and ease some stress and burden so that families can focus on their safety and well-being.

Jue, V. 1 ; Sieng, D. 2 ; Kleinhenz, M. 2 1. Pharmacy, UCSF Medical Center, San Francisco, CA, USA; 2. UCSF Medical Center, San Francisco, CA, USA Background: Quality assurance data collected at University of California San Francisco (UCSF) Medical Center in 2013 found the top medication issues for patients transitioning out of the hospital include insurance, access, prescription issues, side effects, and directions. CF adults wrestle with issues not only with hospitalization but with new medications, reformulation of existing therapies, changes in insurance plans and alignment of services with specialty pharmacies. To assess how a pharmacy technician (TECH) could service the needs of patients in our Adult CF Program (n=109), we tracked and quantitated the range and extent of a pharmacist's activities delegated to a TECH.Method: Since the addition of a TECH in Dec. 2017, the care team catalogued and tracked pharmacy related activities performed by the TECH. The TECH was tasked with processing high-risk medication prior authorizations (PAs) and treatment authorization requests (TARs). High-risk medications include but are not limited to: inhaled antibiotics, inhaled mucolytics, and CFTR modulator therapy. Other responsibilities include routing high-risk medications to correct specialty pharmacies and facilitating sign up for grants to lower copays. Data are collected on a secured Excel spreadsheet.Results: From Dec. 2016 -Nov. 2017, the pharmacist obtained 87 PA/ TAR approvals and 17 denials. Denials were managed by: letter of appeal (LOA), change to formulary product, discount cash price, or through CF grants/patient assistance programs. Time to resolution of denials ranged 1-104 days. Overall, 106 medication insurance issues were tackled which included: 17 inhaled antibiotics, 13 inhaled dornase alfa, 8 pancreatic enzymes, 3 inhaled hypertonic saline, 10 special CF vitamins, 5 azithromycin, and 50 nonspecialty medications. From Dec. 2017 -Apr. 2018, the TECH managed 50 PA/TAR approvals and 16 denials. Time to resolution of denials ranged 1-180 days. 129 medication insurance issues were resolved: 25 inhaled antibiotics, 24 inhaled dornase alfa, 7 pancreatic enzymes, 3 inhaled hypertonic saline, 10 special CF vitamins, 3 azithromycin, and 57 nonspecialty medications. Additionally, with release of tezacaftor/ivacaftor in Mar. 2018, 61 eligible patients were assessed for therapy. To date, the TECH completed 31 tezacaftor/ivacaftor PAs with 3 denials requiring submission of LOA by the pharmacist. Notably, while the pharmacist encouraged patients to sign up for Healthwell grants, the technician has enrolled 5 patients.Conclusion: Embedding TECHs in the care team supports optimum patient care in managing medication PA/TARs required for patient access to high-risk medications. The TECH is the ideal pharmacy agent to track high-risk medication access and use by managing and updating medication outcomes. Participation in a high functioning care team increased job satisfaction.

Fairservice, L.R.; Bomersback, T.; Bjornson, C.; Bendiak, G. CF Clinic, Alberta Children's Hospital, Calgary, AB, Canada Background: Ivacaftor is one of a group of CFTR modulator medications used to treat cystic fibrosis patients. It was initially approved for patients with at least one copy of the G551D mutation in the CFTR gene. In clinical trials, sweat chloride was used as a surrogate marker to document the effect of ivacaftor on CFTR function. Due to local government funding criteria, our clinic has longitudinally measured sweat chloride in a cohort of pediatric pediatrics treated with ivacaftor. There is little data in the published literature regarding the long-term impact of ivacaftor on sweat chloride measurements, or their variation with time.Objective: The aim of this quality assurance review was to describe the variability in repeated sweat chloride measurements in a pediatric cohort treated with ivacaftor.Methods: Sweat chloride results were reviewed for all patients in our clinic who had received treatment with ivacaftor between 2013 -2018. Descriptive statistics were used to review the variation in measurements for each patient. Given the nature of this work as quality improvement, formal review was waived by the institutional ethics review board.Results: Eight patients were identified who had been treated with ivacaftor. Median age at initiation of treatment with ivacaftor was 6.5 years (range: 3 -14) . Four patients (50%) were male. All patients had one copy of the G551D gene mutation. Median sweat chloride pre-treatment was 109.5 mmol/L (88 -122). Median sweat chloride at the first measurement after initiation of treatment was 47 mmol/L (38 -88).Five patients had repeated annual measurements of sweat chloride after initiation of therapy to ensure ongoing eligibility for government funding. Median number of serial measurements was 5 (4 -8) . (See Table. )Conclusions: Within-patient variability in annual sweat chloride measurements was greater than anticipated in patients being treated long-term with ivacaftor. While adherence to therapy is a primary consideration, we query whether other factors, such as variability in timing of medication administration with respect to sweat chloride measurement, or variation in appropriate fat and enzyme co-administration with the medication, could account for these differences. Based on the findings of this quality assurance review, our team is taking steps to advocate for alteration of the current government funding criteria mandating annual sweat chloride testing, to include other measures of efficacy and adherence. Background: Cystic fibrosis (CF) patients are often diagnosed by newborn screen shortly after birth. At the first visit at the CF center pancreatic enzymes are commonly prescribed. This can be an overwhelming time for parents with the diagnosis and a new baby. A pharmacist presence in the outpatient CF clinic can have a significant impact on patient care. The objective of this study was to determine the impact of a pharmacist in a CF newborn screen clinic on the length of time from pancreatic enzyme prescribing to patients receiving medication.

Methods: A retrospective study was conducted and included CF patients diagnosed by newborn screen and seen in the CF newborn screen clinic from January 1, 2016 to February 1, 2018. Patients were excluded if they were diagnosed in the hospital and sent home on pancreatic enzymes from the hospital. The patient's medical record was reviewed for date of first visit, date of prescription. Pharmacy records and contact with the pharmacy were used to determine when the patient received the medication and the copay amount if applicable.Results: The study included 15 patients in the no-pharmacist group and 9 patients in the pharmacist group. Sex and age at first visit were similar between the two groups, no-pharmacist group 10 (66.7%) male versus 6 (66.7%) male in the pharmacist group (p=1) and the median (IQR) age at first visit was 15 (12-22) days in the no-pharmacist group and 15 (11. 5-25.5) in the pharmacist group (p=0.9). The median (IQR) number of days from first visit to obtaining pancreatic enzymes in the no-pharmacist group was 3 (1-28) days with a range of 1-125 days. The median (IQR) number of days from first visit to obtaining pancreatic enzymes in the pharmacist group was 1 (0-1) day which was a statistically significant difference from the no-pharmacist group (p=0.003). Patients in the no-pharmacist group had copays range from $0-$106 for the first fill. In the pharmacist group all patients had $0 copay.Conclusions: This is the first study looking at the potential pharmacist impact in a CF newborn screen clinic. The addition of the pharmacist to the CF newborn screen clinic team significantly decreased the time from prescribing to patients obtaining pancreatic enzyme supplementation and reduced patient out of pocket cost. Background: The CF population continues to age, and lengthy hospitalization for IV antibiotics is not feasible for many working adult patients. The University of Kansas Health System (TUKHS) is a Midwestern academic medical center in Kansas City (KC), KS. The adult CF center at TUKHS cares for more than 230 adults with CF from five area states, with 49% of patients residing outside of the KC greater metro area. Previous quality improvement (QI) work from our center suggests increased residential distance to our center may impact care. Since CF patients are at higher risk for lung function loss during pulmonary exacerbation, we established a post-hospital follow-up clinic for patients discharged home on IV antibiotics, who at time of discharge had not yet recovered their baseline lung function. We aimed to test the feasibility of this clinic with such a large patient catchment area. Here, we report on our experience with a follow-up clinic developed as a QI initiative.

Methods: Patients hospitalized at TUKHS for CF exacerbation who were discharged home on IV antibiotics with an FEV1 that was not yet at baseline (>5% actual FEV1 decline from previous best in prior 6 months) were offered a follow-up clinic appointment. The appointment was scheduled at day 14 of IV antibiotic treatment to assess clinical response. If FEV1 was not yet at baseline, IV antibiotic extension for at least 1 week was recommended. Central lines were removed in the follow-up clinic if antibiotic course was complete. Distance traveled, mean antibiotic duration, attendance in clinic, and patient satisfaction were assessed.Results: Forty-four patients between Aug. 2017 and April 2018 qualified for outpatient IV antibiotic follow-up. Three patients (7%) met criteria but were not offered a follow-up appointment and were excluded from analysis. Twenty-eight patients (64%) maintained their follow-up appointment and traveled an average round trip distance of 117.4 miles (range 1-774 miles). Ten patients (23%) with average distance to travel of 150.9 miles declined the follow-up appointment. Three patients (7%) with average distance to travel of 125.7 miles did not show or canceled their appointment. Mean antibiotic duration of those who attended the follow-up clinic was longer than those who did not attend clinic at 17.1 vs. 16.4 days, respectively. Patients who declined or did not show for follow-up appointment had lower FEV1 than patients that maintained the appointment at 47% vs 55.6%, respectively (p=0.3). Of patients who completed the survey, when asked if the follow-up clinic was beneficial, the average score was a 4.5 on a Likert scale, with 4.0 being "agree" and 5.0 being "strongly agree."Conclusion: Many adult patients prefer to complete antibiotics at home instead of the hospital. The distance to CF center may impact patient care and follow-up in these patients. Despite a large catchment area, our results suggest pulmonary exacerbation follow-up clinic was valued by patients and has a positive effect on post-hospital outcomes. Future work will include evaluation of the impact post-hospital follow-up clinic has on probability of recovery to baseline FEV1 in a larger sample size.