key: cord-0010157-kjbktnri
authors: nan
title: Symposium Summaries
date: 2012-09-13
journal: Pediatr Pulmonol
DOI: 10.1002/ppul.22681
sha: 5eb91fe4cea5a06b48b39b8ac338ae42b12cc336
doc_id: 10157
cord_uid: kjbktnri

nan

mutation cystic fibrosis (nmCF) represents one of the more severe forms of the disease and currently lacks disease modifying treatment options. The investigational new drug ataluren is a protein restoration therapy designed to enable the formation of a functioning protein in patients with genetic disorders caused by a nonsense mutation. In patients with CF, ataluren induces ribosomal readthrough of premature stop codons in mRNA, resulting in full-length, functional CFTR.

Pharmacologic proof-of-activity of ataluren has been demonstrated in animal models and Phase 2 clinical trials. Phase 2 studies in 77 patients (ages 6-57 years) with nmCF [2] [3] [4] [5] receiving oral ataluren for periods of 2 through 12 weeks were conducted in 4 separate trials. Significant apical CFTR protein expression was seen in epithelial cells in a 21-patient pediatric study [4] . Ontreatment improvements in total chloride transport in nasal potential difference were followed by reversion to baseline values during the post-treatment period, indicating functional activity of the restored CFTR protein [3] [4] [5] . A positive trend in changes in %-predicted FEV1 was also observed in the 12-week study [5] . Ataluren was generally well tolerated in all Phase 2 clinical trials [2] [3] [4] [5] .

A Phase 3, randomized, double-blind, placebo-controlled clinical trial (Study 009) evaluated the efficacy and safety of ataluren in nmCF was recently completed. Patients were required to be ≥6 years of age, have a nonsense mutation in at least one allele of the CFTR gene, forced expiratory volume in 1 second (FEV1) ≥40% and ≤90% of predicted for age, and a stable regimen of chronic treatment/prophylaxis for CF or for CF-related conditions (if receiving such medications). Patients were stratified by chronic inhaled antibiotic use, age, and FEV1. A total of 238 patients were randomized to receive ataluren 10, 10, 20 mg/kg or placebo at breakfast, lunch, and dinner for 48 weeks.

The primary endpoint was change in %-predicted FEV1; the secondary endpoint was pulmonary exacerbation rate. In the intent to treat (ITT) population, the dif-ference in mean relative change in %-predicted FEV1 from baseline to Week 48 was 3.0% in favor of ataluren (-2.5% for ataluren and -5.5% for placebo, p=0.124). The average treatment effect in %-predicted FEV1 across all post-baseline visits was 2.5% (p=0.0478). Similar positive trends were seen in FEF25-75. The pulmonary exacerbation rate was 23% lower over 48 weeks for ataluren versus placebo (p=0.0992).

The interaction of chronic inhaled antibiotic use with treatment was statistically significant (p=0.0072) for the primary endpoint, indicating that inhaled antibiotic use was a confounder of the overall results. In patients not taking inhaled antibiotics, the difference in mean relative change in %-predicted FEV1 was 6.7% favoring ataluren at Week 48 (nominal p=0.014) and 5.5% on average across all post-baseline visits (nominal p=0.0006). A 43% lower pulmonary exacerbation rate over 48 weeks (nominal p=0.012) favoring ataluren was observed in patients not taking inhaled antibiotics.

The subgroup effect in patients who were receiving antibiotics was driven by patients who were receiving tobramycin, one of several antibiotics used chronically for treatment/prophylaxis of CF. Tobramycin and ataluren both act at the level of the ribosome, and in vitro data indicate that tobramycin antagonizes ataluren at a mechanistic level.

The tertiary endpoints of nasal potential difference and sweat chloride did not demonstrate a treatment difference in this study.

Safety profiles were similar for ataluren and placebo, other than transient cases of mild creatinine elevations, and reversible Grade 3-4 creatinine elevations that were associated with the combination of systemic nephrotoxic antibiotics with ataluren. These Grade 3-4 creatinine elevations were effectively managed in the trial by the prohibition of concomitant therapy of study medication with systemic aminoglycosides and other nephrotoxic antibiotics.

Preliminary data for %-predicted FEV1 from an ongoing open-label extension study (Study 009e) are encouraging, suggesting a persistent treatment effect. Ataluren continues to be generally well tolerated in Study 009e and no new safety concerns have been identified.

Ataluren holds promise as a new drug for treatment of nmCF. Positive trends in lung function and pulmonary exacerbation rate were seen for ataluren over 48 weeks of treatment. Response to ataluren was most apparent in patients not receiving inhaled tobramycin, suggesting a potential antagonist interaction with ataluren. Reductions in both the loss of FEV1 and frequency of pulmonary exacerbations are clinically meaningful endpoints, suggesting that ataluren is potentially disease-modifying.

References: 1. Bobadilla JL, Macek M Jr, Fine JP, Farrell PM. CF: a worldwide analysis of CFTR mutations -correlation with incidence data and application to screening. Hum Mutat. 2002 Jun; 19(6) :575-606.

The cell exploits the emergent properties of proteostasis, a biological language directing the function of the polypeptide sequence encoding the protein fold, to manage the health of the human proteome. The proteostasis system includes a multitude of folding chaperones, trafficking pathways and degradation systems that respond to folding stress responsive signaling pathways of high clinical relevance to manage human healthspan. Proteome/interactome maintenance by the proteostasis network (PN) is very dynamic -promoting/protecting proteome health both inside and outside cells found in diverse tissue and organ environments in response to stress and inherited disease such as cystic fibrosis (CF) (Cold Spring Harb Perspect Biol (2011) 3 pii: a004499). Human biology has taught us that inherited folding chal-lenges such as found in CF alters the kinetics and energetics of key steps in a multistep folding landscape that can be routinely managed by changes in the proteostasis language, a feature of folding biology that offers us numerous opportunities to intercede with therapeutics that modulate the composition and/or function of the PN code to protect the host from folding disease such as found in ∆F508 CFTR variant. By use of proteome-oriented mass spectrometry and bioinformatic tools we are building a multilayered view of healthy biological protein folding and the changes that occur in response to proteostasis targeted therapeutics to alleviate the folding stress that alters the protein fold function found in misfolding disease, potentially pointing toward new therapeutic approaches to CF. G551D mutation demonstrated sweat chloride reductions of almost 60 mmol and significant, dose-related changes in nasal PD. The first randomised, doubleblind, placebo-controlled, multinational phase 3 study was conducted in adults and children over the age of 12 with at least one copy of the G551D mutation. The primary outcome was FEV1, in which significant improvements of over 10 absolute percentage points were maintained out to 48 weeks. These were accompanied by a reduction in the exacerbation rate, significant weight gain, a similar large drop in sweat chloride to that seen in Phase 2 and improvements in the CF-specific quality of life tool, CFQ-R. The drug was shown to be safe, with adverse events equally reported in both arms. A second Phase 3 study in children aged 6-11 years has recently reported similar magnitude of change in FEV1 and significant improvements in weight and BMI. The majority of patients from both these trials rolled over into the open-label phase of the study and long term results are encouraging from both efficacy and safety perspectives. On the basis of these data, ivacaftor is now licensed in the U.S. for children over the age of 6 and adults with G551D and is currently under review in Europe.

In patients with better preserved lung function, such improvements in FEV1 may not be achievable and more sensitive outcome measures may be helpful. A Phase 2 trial studied the effect of ivacaftor in subjects with FEV1>90% using the lung clearance index (LCI) as primary outcome. This is a measure of ventilation inhomogeneity that has been shown in young children and babies to be more sensitive than FEV1 and to accurately reflect underlying structural change. In this study, over a period of 4 weeks, there was a significant treatment effect in LCI (2.2 units), demonstrating benefit even at the early stages of lung disease. A study in children aged 2-5 years is planned.

The tremendous success of this drug has raised questions over how widespread its applicability could be. G551D patients were studied to optimise homogeneity, but in vitro data support the action of this drug on other class 3 mutations, most of which are rare, and a multicentre trial is planned for 2012. Similarly, the drug may increase the open probability of CFTR reaching the cell surface with other classes of mutation; the class IV mutation, R117H, has been shown to be potentiable in vitro and patients with this mutation will also be included in future trials.

Correctors: Class 2 mutations, in which misfolded CFTR is degraded intracellularly and fails to traffic to the cell surface, are present in the majority of CF patients worldwide, making it a very attractive therapeutic target. VX-770 was shown to be ineffective in patients homozygous for this mutation when used as a single agent. VX-809 is the first corrector to have entered clinical trials: in a safety/ dose ranging study, small but significant decreases in sweat chloride were observed after 28 days confirming mechanism of action, but there were no concomitant improvements in nasal potential difference or FEV1. Combination trials with VX-809 and VX-770 are currently underway. Vertex also has another corrector agent, VX-661 which is being studied in a Phase 2 trial both with and without co-administration of VX-770.

Overall, these results provide encouraging evidence that CFTR is a tractable target and have led to a resurgence of enthusiasm for the small molecule approach.

References: 1 We now have an approved and effective CFTR potentiator, i.e., ivacaftor. The first and most obvious goal is to explore the full potential of ivacaftor in the treatment of CF. Obvious indications to assess are: patients with gating mutations other than G551D, patients with G551D younger than 6 years, patients with the R117H-5T complex allele or a class V mutation. Nine CFTR mutations associated with severe gating disturbance and responsive to ivacaftor in vitro have indeed been identified (1) . A trial in subjects carrying these mutations will further establish whether the in vitro efficacy of a CFTR potentiator in Fisher rat thyroid cells translates to in vivo efficacy in patients. Intuitively the efficacy of a CFTR potentiator should be largest in young patients with preserved lung structure. When started early in life, treatment with CFTR modulators has the potential to prevent later disease complications.

The class IV mutation R117H does not only lead to reduced channel conductance but also to impaired channel gating. Since ivacaftor also potentiates wild type CFTR, a clinical trial in subjects with a class V mutation can be considered.

An efficient CFTR modulator will boost the development of new surrogate outcome markers; it will facilitate demonstrating the link between biomarkers and clinical outcome. Patients with at least one G551D mutation and an FEV1 above 90%, had a marked improvement in LCI as well as FEV1. Compared to FEV1, LCI had greater sensitivity since a post hoc analysis led to a lower number needed to include in the trial with LCI as primary outcome parameter rather than FEV1. The same was true for sweat chloride. Further evaluation of improvement in sweat chloride and lung function parameters will solve the question whether sweat chloride can be used as a marker of efficacy: whether it can be used quantitatively, i.e., the response predicts the size of the clinical benefit or qualitatively, i.e., the response predicts benefit but not size of it. Sweat chloride measurements are well standardized, but for LCI this is urgently needed.

To advance the field of biomarkers and clinical outcomes a new mindset is needed. One cannot expect that a biomarker measurement in the first years of life will predict the entire longterm clinical outcome. The solution can be to agree on intermediate markers of beneficial outcome such as improved nutritional status, improved lung disease, lower sweat chloride. Improved lung disease can be defined as delay of chronic Pseudomonas aeruginosa infection, less bronchiectasis, less airway obstruction. There are compelling data from natural history and registries to make these claims. The outcome marker used must then reliably measure this attribute but the outcome measure itself does not need proof of predicting long term outcome. The latter has for a long time been the major factor delaying promotion of new outcome measures to surrogate outcomes. More thought can then go to deciding on the necessary effect size.

As trials with CFTR modulators become more and more mutation specific, for some groups the patient numbers available will be very small. Hence alternative trial designs and drug effect analyses should be developed, such as cross-over trials and responder analysis to assess individual benefit. The intrinsic variability of the primary outcome parameter can be established during the placebo period. When the change in outcome parameter during drug intake exceeds the mean variability plus 1.6 (or 1.9 SD) a responder is identified (2) .

During drug development, great care is taken to guard treatment safety. However, in an orphan disease, only a limited number of patients are exposed to the drug during clinical trials. Ongoing safety assessment post licensing is thus important. A proactive structured plan via CF registries is preferable over solely relying on spontaneous reporting of adverse drug reactions by individual patients and physicians.

Even if a potentiator would be indicated in all patients carrying at least one class III, IV or V mutation, still only about 10% of the CF cohort would benefit from this treatment. It is thus absolutely necessary to find efficient correctors. Given the initial results of combined corrector and potentiator treatment, the correctors under current development will progress in clinical trials. There is however improved understanding of the hurdles involved in F508del CFTR rescue. By inserting "reverting mutations" it was established that efficient correction of F508del CFTR requires improvement of the thermodynamic stability as well as the kinetic profile of the mutant protein. In vitro this requires improvement of NBD1 (nucleotide binding domain) folding as well as restoration of a hydrophobic cluster that can link NBD1 to MSD2 (the second membrane spanning domain) (3, 4) . Finding correctors with dual or additive modes of actions thus seems necessary for highly efficient F508del CFTR rescue. A targeted search could certainly complement the corrector search via high-throughput screening.

We should not overlook the problem associated with drug cost. Equity in drug use is the responsibility of the entire CF community. Very effective drugs should also be made available in countries with a low gross national product. It would also be inappropriate to reserve expensive treatments for rescue of symptomatic patients, since the full drug benefit will be more obvious in patients with preserved lung function. 

Michel Bouvier, Ph.D., F.C.A.H.S.

Mutations in G protein-coupled receptor (GPCR) genes cause various human diseases. In many instances, the mutations do not affect specific functional sites but rather cause conformational changes that are unnecessarily recognized by the endoplasmic reticulum (ER) quality control system as misfolded leading to their intracellular retention and degradation. Examples of such loss of function conformational diseases include nephrogenic diabetes insipidus (NDI) and early-onset severe obesity that result from mutations in the vasopressin receptor type 2 (V2R) and the melanocortin receptor type 4 (MC4R) respectively. For both V2R and MC4R, we found that cell-permeant ligands, which bind to the receptors, increase cell surface expression and rescue their function; an observation leading to the concept of pharmacological chaperones (1) (2) (3) . By binding to the mutated receptors, pharmacological chaperones promote their proper folding, maturation and trafficking. For NDI, a V2R selective pharmacological chaperone was found to significantly improve patients' condition in a small clinical trial establishing its therapeutic value (4) . In order to establish an in-vivo proof of principle that the concept can be extended to early-onset severe obesity, we generated a novel model for MC4Rpormoted obesity. The potential therapeutic value of MC4R pharmacological chaperones is currently being tested in this new model of genetic severe obesity. Although pharmacological chaperones target directly the misfolded receptors, compounds that would interfere with the ER quality control system could also represent potential drug candidates for the treatment of conformational diseases. To identify potential targets, we performed a large scale bioluminescence resonance energy transfer (BRET)-based screen to identify proteins involved in the ER processing of GPCRs. Cornichon homolog-4 (CNIH4) was identified as one of the proteins selectively interacting in the ER with many GPCRs. Through the formation of a complex with COP-II vesicle proteins and other ER-resident proteins CNIH4 was found to play an important role in the ERexport of GPCRs. Whether such proteins may also represent targets for the treatment of conformational diseases remains to be established. 

Beth Hoffman, Ph.D. Vertex Pharmaceuticals, Inc., San Diego, CA, USA Cystic fibrosis (CF) is a chronic, debilitating autosomal recessive disease caused by mutations in the CF transmembrane conductance regulator (CFTR) gene. Many of these mutations, including the most common mutation, F508del, result in defective assembly of the multiple functional domains of CFTR, and severely decreased levels of functional protein at the cell surface. Small molecules, known as CFTR correctors (1, 2) , that permit CFTR to reach the cell surface and function, have been identified based on functional screens, agnostic to mechanism of action. In parallel, considerable research on the biogenesis of CFTR has illuminated details of protein translation, folding, trafficking from the endoplasmic reticulum (ER) to plasma membrane, and the quality control check points along this process with an eye toward developing CFTR correctors.

Numerous neurodegenerative diseases, including Alzheimer's disease, Parkinson's disease, Huntington's disease, Amyotrophic Lateral Sclerosis, exhibit aggregates of one or more proteins characteristic of a specific disorder (3) . Despite the ongoing debate about whether these aggregates are cytotoxic or a protective mechanism, there is general agreement that these aggregates suggest some common etiology. In addition to diseasespecific proteins, aggregates contain other cellular proteins which may reflect a more general cellular stress response. Post-mortem studies have suggested pervasive disturbances of mitochondrial function (4) including energy balance and oxidation. For most of these neurodegenerative diseases, there appear to be both idiopathic (unknown) causes as well as genetic mutations in specific proteins found within these aggregates. Naturally, the genetics represent a sufficiently compelling handle by which to develop potential therapeutic agents, aimed at reducing or inhibiting these proteins.

Despite the genetics of neurodegenerative diseases, the onset of overt disease generally does not appear before middle age. This suggests that the aging process may result in a reduced capacity in protein biogenesis, quality control processes related to protein folding and/ or protein degradation. Mutations in certain genes may combine with this aging process to precipitate the disease process.

Superficially, it might appear that CF and neurodegenerative diseases have very little in common. In CF, there is a loss of CFTR function but no apparent aggregates and relatively unperturbed mitochondrial function. So how can drug discovery efforts in neurodegenerative diseases teach us how to develop better therapeutics for CF?

Research in both CF and neurodegenerative diseases have contributed to expanding interest in regulatory processes of protein translation, folding, degradation and trafficking to different sites within the cell. Approaching protein homeostasis from opposite ends of the process, CF focused on protein biogenesis (increasing CFTR) and neurodegenerative diseases focused on protein degradation (removing toxic misfolded protein), have highlighted the quality control processes of protein folding. Recent advances in the understanding of these quality control processes in neurodegenerative disease will be used to illustrate potential new directions for developing CF therapeutics. In the classical form of α1antitrypsin deficiency (ATD) a point mutation renders a hepatic secretory protein prone to misfolding and polymerization/aggregation. The mutant protein, α1antitrypsin Z (ATZ), accumulates in the endoplasmic reticulum (ER) of liver cells, causing liver fibrosis and carcinogenesis by a gain-of-toxic function mechanism, which we will refer to as its "proteotoxic" effect. We know that there are least two general mechanisms by which ATZ may be degraded after it accumulates in the ER: the proteosomal pathway is responsible for disposal of soluble forms of ATZ and the autophagic pathway is uniquely responsible for disposal of insoluble ATZ and perhaps some soluble ATZ. Our recent studies have also suggested that there is a nonproteosomal, non-autophagic pathway for intracellular disposal of ATZ that is regulated by the insulin signaling pathway. Recently, we investigated the hypothesis that the accumulation of ATZ in hepatocytes and its proteotoxic effects reflect an inability of these disposal pathways to handle the hepatic ATZ load and that pharmacological enhancement of intracellular ATZ disposal prevents or ameliorates liver damage in ATD. Our progress has validated this hypothesis in showing that one drug, carbamazepine (CBZ), which enhances the autophagic degradation of ATZ reduces the hepatic load of ATZ and reverses hepatic fibrosis in a mouse model of ATD. Because CBZ is FDA-approved and has been used safely in humans as an anticonvulsant and mood stabilizer it has rapidly moved into a clinical trial for severe liver dis-ease due to ATD. Using a novel and powerful C. elegans (worm) model of ATD in a high-content automated screening platform we discovered 5 additional compounds that reduce intracellular ATZ accumulation and 4 of these compounds have been found to enhance autophagic disposal of ATZ. Two of these drugs, which are FDA-approved as mood stabilizers, have been shown to reduce ATZ accumulation in mammalian cell line models of ATD. Futher, a pharmacophore model developed by computational analysis of this screen has been validated and can now be used to identify additional drug candidates. Finally, in preliminary studies we found that computational pharmacological strategies for analyzing RNAi screens of the worm model of ATD have identified novel candidate drugs for treatment of ATD. Together, the studies have provided evidence that endogenous proteostasis mechanisms can be utilized for therapeutic interventions for disorders caused by misfolded proteins and autophagy enhancer drugs appear to be particularly promising examples.

Objectives:

1. Learn about mechanisms of proteostasis for misfolded proteins that accumulate in the endoplasmic reticulum.

2. Learn about how proteostatic mechanisms influence the clinical phenotype of misfolded protein disorders.

3. Learn how primitive organisms like C. elegans can be used to model protein misfolding disorders and for development of drug screening platforms.

Brandon A. Wustman, Ph.D. 1 , Pol Boudes 2 and David J. Lockhart 1 1. Amicus Therapeutics, La Jolla, CA, USA; 2. Amicus Therapeutics, Cranbury, NJ, USA A large number of inherited diseases result from mutations that lead to a destabilized, unfolded or misfolded protein. Cystic fibrosis (CF) and Fabry disease are two examples where missense, small insertion or deletion mutations result in translated mutant proteins that retain some function but are largely unable to pass the protein quality control system for export out of the endoplasmic reticulum (ER). Instead, these conformationally challenged proteins are more commonly tagged for disposal by the ER-associated degradation pathway. Phar-macological chaperone therapy is an investigational approach currently in clinical development for the treatment of CF (e.g., Vertex Pharmaceuticals) and Fabry disease (e.g., GlaxoSmithKline Plc/Amicus Therapeutics). Pharmacological chaperones are small molecules that selectively bind and stabilize the native conformation of mutant proteins, allowing them to pass the ER quality control system and to be transported and delivered to their final functional location (e.g., cell surface, lysosomes). Pharmacological chaperone therapy therefore represents a new class of treatment for patients with diseases that result from protein destabilization, unfolding or misfolding.

Fabry disease, an X-linked lysosomal storage disorder, is caused by mutations in GLA, the gene encoding lysosomal α-galactosidase (α-Gal A). The role of α-Gal A within the body is to catalyze the first step in the breakdown of a complex lipid in lysosomes called globotriaosylceramide (GL-3). Reduced or absent levels of α-Gal A activity leads to the accumulation of GL-3 in the affected tissues, including the central nervous system, heart, kidneys, and skin. This accumulation of GL-3 is believed to cause the various symptoms of Fabry disease, including pain, kidney failure, and increased risk of heart disorders and stroke. Currently, regular infusion of recombinant human α-Gal A (e.g., agalsidase beta, agalsidase alpha), termed enzyme replacement therapy, is the standard of care for Fabry disease.

The investigational pharmacological chaperone migalastat HCl (1-deoxygalactonojirimycin HCl) is currently in clinical development for the treatment of Fabry disease. In preclinical studies, migalastat HCl has been shown to be a potent and selective competitive inhibitor that reversibly binds to the active site of α-Gal A (1). Binding of migalastat HCl to α-Gal A increases the thermal stability of the protein (2) and increases the amount of properly processed functional enzyme delivered from the ER to lysosomes, as demonstrated using patientderived lymphoblasts and fibroblasts (1, 3) . The ability of migalastat HCl to reduce GL-3 levels in vivo was investigated using a transgenic (Tg) Fabry mouse model in which the mutant R301Q form of human α-Gal A is expressed on a knockout (KO) background (4) . These Tg/KO mice accumulate GL-3 in multiple disease-relevant tissues. Oral administration of migalastat HCl to Fabry Tg/KO mice resulted in significant and dosedependent increases in α-Gal A activity, with concomitant GL-3 reduction in skin, heart, kidney, brain, and plasma that were comparable to those obtained with intravenous injections of agalsidase beta.

More than 670 mutations (~60% missense) in the GLA gene have been identified for Fabry disease. To help identify which patients might benefit from treatment with migalastat HCl, the in vitro responses of 527 Fabry disease-causing mutations (missense, small deletions and insertions) were assessed using HEK-293 cells that transiently express α-Gal A variants (5) . Approximately 47% of the mutant forms tested in HEK-293 cells demonstrated a significant increase in activity (two-tailed, paired ttest; p-value ≤ 0.05). Taking therapeutically achievable doses into consideration, ~35% of the mutations were amenable to migalastat HCl. Amenable was defined as an increase to ≥ 3% of wild-type and a relative increase of ≥ 1.2-fold when treated with 10 µM migalastat HCl (the C max value in Fabry patient plasma after oral administration of 150 mg of migalastat HCl).

Migalastat HCl was shown to be generally safe and well-tolerated in single-dose (25-2000 mg) and multiple ascending dose (50 and 150 mg) Phase 1 clinical studies, and significantly increased wild-type α-Gal A levels in healthy volunteers. Twenty-six subjects completed Phase 2 studies (FAB-CL-201/NCT00214500, FAB-CL-202/NCT00283959, FAB-CL-203/NCT00283933, FAB-CL-204/NCT00304512) to receive an oral dose of migalastat HCl (25-250 mg BID, QD or QOD) for either 12 or 24 weeks. Twenty-three of these subjects enrolled in a separate, extension study designed to evaluate the long-term safety and efficacy of migalastat HCl (150 mg QOD). Seventeen subjects continue in the ongoing extension study (FAB-CL-205/NCT00526071) and have been receiving migalastat HCl for up to a total of 6 years. During the course of the extension study, migalastat HCl continued to be generally well tolerated, with no drugrelated serious adverse events. The most common adverse events have been headache, arthralgia, diarrhea and fatigue. Preliminary data for renal function released in February 2011 indicated that estimated glomerular filtration rate (eGFR) remained stable up to 4 years for all subjects continuing in the extension study. Among subjects identified as having α-Gal A mutations amenable to migalastat HCl in the HEK-293 assay, the average annual rate of change in eGFR was +1.6 mL/min/1.73m 2 (excluding hyperfiltrators).

Amicus and GSK are conducting two Phase 3 global registration studies of migalastat HCl monotherapy (AT1001-011/NCT00925301, AT1001-012/NCT01218659). Both studies are investigating migalastat HCl (150 mg QOD) in Fabry patients identified as having α-Gal A mutations amenable to migalastat HCl as a monotherapy. The study AT1001-011 is a six-month, placebo-controlled Phase 3 study with an open-label migalastat HCl extension period for up to 18 months. In the initial 140 Fabry subjects screened, 119 (85%) had missense mutations and 97 (82%) had mutations that were amenable to migalastat HCl as a monotherapy. For study AT1001-011 a total of 180 subjects were screened and enrollment was completed with 67 subjects in December 2011. Results from this study are anticipated by the end of 2012. Study AT1001-012 is a randomized, open-label, 18month Phase 3 study investigating the safety and efficacy of migalastat HCI compared to current standard-of-care (e.g., agalsidase beta or agalsidase alfa) for Fabry disease.

A growing number of studies have demonstrated that pharmacological chaperones can be used to increase trafficking and processing for a wide variety of protein classes including enzymes, secreted proteins, transcription factors, G protein-coupled receptors and ion channels. Recently, Vertex Pharmaceuticals announced results from an interim analysis of their Phase 2 clinical trial for VX-809 and KALYDECO™ (ivacaftor). VX-809, a cystic fibrosis transmembrane conductance regulator (CFTR)-targeted pharmacological chaperone, when used in combination with the CFTR potentiator Kalydeco™, was shown to improve lung function in CF subjects who have two copies of the F508del mutation. Half of all CF patients have 2 copies of the F508del mutation and approximately 80% have at least 1 copy of the F508del mutation, a mutation that causes defective protein folding and processing of CFTR in the ER. These studies in CF and Fabry disease demonstrate the initial safety and efficacy of pharmacological chaperones in humans. Pharmacological chaperones, therefore, represent a new class of investigational therapeutics for the potential treatment of diseases caused by mutations that result in destabilized, unfolded or misfolded proteins.

References: 1. Fan JQ, Ishii S, Asano N, Suzuki Y. Accelerated transport and maturation of lysosomal αgalactosidase A in Fabry lymphoblasts by an enzyme inhibitor. Nat Med. 1999;5(1):112-5. The clinical course of CF is affected by interactions among genetic, environmental, and health care-related factors. Genetic explanations for variability in disease progression provide a promising source of information regarding disease biology and pathophysiology, but an understanding of non-genetic factors and the interaction between genetic and environmental characteristics provides more immediate tools for interventions to improve disease outcomes.

Environmental and sociodemographic causes of variability of CF outcomes: Socioeconomic status (SES) as a marker of risk for adverse exposures. The association of socioeconomic factors on CF prognosis is well established (1, 2) and this effect can be noted at an early age (1) . The mechanism is presumably by way of a clustering of detrimental environmental influences, health behaviors and/or difficulties accessing optimal health care; it follows that an understanding of the specific causes of SES-related disparities can provide important clues regarding the effects of adverse exposures for all patients with CF, independent of SES.

Race and ethnicity. The demography of race and the relative contribution of genetic vs cultural determinants is complex. Hispanics, in particular, seem to have an increased prevalence of risk factors whose impact may be seen at an early age (3) . The literature on the relationship between race and outcomes in CF is contradictory, primarily because of the difficulties encountered in controlling for SES when evaluating racial and ethnic disparities (4) .

Environmental tobacco smoke (ETS) and other pollutants. The association of ETS exposure with poor lung function and growth in children with CF was initially reported in 1990 among attendees of a summer camp (5) and has been confirmed in several subsequent publications. ETS exposure is probably a significant contributor to health disparities among low SES patients with CF. Furthermore, variations in a CF-modifier gene (TGF-β1) amplify the negative effects of ETS (6) . Exposure to air pollution (specifically, particulates) has also been shown to be associated with an increase in the likelihood of pulmonary exacerbations and a decrease in FEV1 (7) .

Acquisition of infectious agents. The effect of airway infection on the progression of CF lung disease will be covered by other speakers in this symposium. Acquisition of both Staphylococcus aureus and Pseudomonas aeruginosa (PA) in the CF airway may occur early in life and is associated with more rapid decline in pulmonary function (8) . The likelihood of acquisition of PA is increased by the use of antistaphylococcal prophylaxis; exposure to other CF patients with PA; nebulizer use; hospitalization for viral respiratory infection or treatment with IV antibiotics; and female sex (9) . Respiratory viral infections also appear to be associated with more rapid progression of CF airway disease, at least in the short term (10) .

Stress and mental health considerations. While stress may have a negative impact on disease course by impairing immune function (11) , its role in compromising adherence is probably more important (12) . From the perspective of early childhood lung disease, the impact of parental stress and mental health on disease management is probably the most important consideration (13, 14) .

Dietary and nutritional deficiencies. Nutritional status is highly correlated with pulmonary function in patients with CF, and while the causal relationship is to some extent bidirectional, there is strong evidence that early nutritional inadequacy is deleterious to the lung (15, 16) . The mechanism of this relationship is unclear. 19 Cystic fibrosis (CF) is a life-shortening disease caused by mutations in the cystic fibrosis transmembrane conductance regulator (CFTR) gene (1) . Although bacterial lung infection and the resulting inflammation cause most of the morbidity and mortality, how the loss of CFTR first disrupts airway host defense has remained uncertain. Proposed origins of CF lung disease include reduced mucociliary clearance due to decreased airway surface liquid (ASL) volume or altered mucus, reduced bacterial killing by ASL antimicrobials, defective bacterial elimination by phagocytes, abnormal inflammatory responses, reduced or increased bacterial binding by airway epithelia, and other defects (2) (3) (4) (5) (6) . One or more of these defects could be responsible. Two factors have made it difficult to distinguish between hypotheses and identify initiating insults. First, as with many diseases, clinical manifestations may not reflect original deects, but it is problematic to study CF at its onset in newborn human infants. Second, mice with disrupted CFTR genes do not develop typical CF (7) . To circumvent these obstacles, we generated CFTR -/pigs (CF pigs) (8) , which spontaneously develop hallmark features of CF lung disease (8, 9) . At birth, their lungs lack infection and inflammation, but have a reduced ability to eradicate bacteria (9) . Thus, newborn CF pigs provide an unprecedented opportunity to investigate mechanisms impairing host defense and initiating disease because they allow CF:non-CF comparisons without secondary confounds.

To investigate the abnormalities that impair eradication when a bacterium lands on the pristine surface of a newborn CF airway, we developed a simple assay that interrogated the viability of individual bacteria immobilized on solid grids and placed on the airway surface. We chemically linked biotin to S. aureus, bound streptavidin to gold grids, and combined them to attach S. aureus to grids. We chose S. aureus because we frequently isolate it from porcine CF lungs, and it is the most common organism isolated from young children with CF (9,10). A fluorescent live/dead stain revealed the state of bacteria. Exposing grids to ethanol killed most S. aureus. Importantly, placing grids on the porcine tracheal surface in vivo also killed bacteria.

We found that in newborn wild-type pigs, the thin layer of ASL rapidly killed bacteria in vivo , when removed from the lung, and in primary epithelial cultures. Lack of CFTR reduced bacterial killing. We reasoned that defective killing arose either from reduced amounts of ASL antimicrobial factors or inhibition of their function. We investigated antimicrobials by measuring mRNA, protein, and aggregate activity under optimal conditions and found that non-CF and CF ASL had similar amounts of antimicrobials. We measured Na + and K + concentrations in ASL collected from newborn pigs and found that they did not differ by genotype.

Thus, these factors do not explain defective bacterial killing in CF. We also measured ASL pH and found that it was more acidic in CF, and reducing pH inhibited the antimicrobial activity of ASL. In addition, elevating airway CO 2 reduced ASL pH and inhibited bacterial killing in non-CF pigs, and, conversely, in CF pigs, aerosolizing NaHCO 3 into the trachea increased ASL pH and enhanced killing (as compared to aerosolizing NaCl). These results directly link the initial host defense defect to loss of CFTR, an anion channel that facilitates HCO 3 transport (11) (12) (13) (14) (15) . Without CFTR, airway epithelial HCO 3 secretion is defective, ASL pH falls and inhibits antimicrobial function, and thereby impairs killing of bacteria that enter the newborn lung. These results suggest that correcting ASL pH might prevent the initial infection in patients with CF. That might be accomplished by delivering HCO 3 into airways, altering pH regulation by airway epithelia, enhancing activity of ASL antimicrobials, delivering pH-insensitive antimicrobials, or targeting mutant CFTR. These results also suggest that measuring ASL pH or adapting the bacteria-coated grid method to assay bacterial killing in patients could be useful for assessing potential therapies. Cystic fibrosis (CF) lung disease is characterised by chronic bacterial infections, neutrophil dominated airway inflammation, progressive airway obstruction and bronchiectasis. Patients with CF or other underlying chronic lung diseases (e.g., COPD, asthma) are considered "high-risk" for developing more severe respiratory viral infections. More rapid deterioration of pulmonary function in CF patients has been associated with respiratory viral infections (1) and respiratory virus infection has been shown to result in more severe and prolonged disease in CF infants compared to infection of non-CF infants (2) . Although the frequency of respiratory viral infections involving CF infants is similar to non-CF infants, virus infections were more likely to result in lower respiratory tract infection, impaired pulmonary function, and hospitalization of CF infants (3) . Rhinoviruses (RV) and the paramyxoviruses, respiratory syncytial virus (RSV) and parainfluenza viruses (PIV) are the most commonly isolated respiratory viruses in the human infant population.

The impact of respiratory virus infections in CF infants and the subsequent development of CF lung disease are for the most part understudied. Although many of the clinical studies focused on CF infants are limited in scope and patient numbers, clinical studies focused on the impact of virus infection in patients with other chronic respiratory diseases such as COPD and asthma have rekindled an interest in the potential role of respiratory viruses in initiating and exacerbating CF lung disease.

Experimental studies clearly indicate respiratory virus infection can disrupt critical innate immune mechanisms responsible for the clearance of virus from the lung, e.g., impairment of the cough reflex; reduced efficiency of mucociliary clearance; accumulation of excess or altered airway secretions; and, reduced macrophage-mediated phagocytosis. While such consequences of viral infection are likely to have acute but transient effects on pulmonary function in a non-CF lung these consequences in the CF lung may be more severe and prolonged due to a (1), tobramycin solution for inhalation (2) and inhaled hypertonic saline (3) . In designing an investigator-initiated trial, a systematic approach is extremely valuable. Below, my personal view of the necessary steps is outlined.

1. Identify an important question. This may involve testing a new therapeutic, device or biological, or assessing an established treatment in a new population or for a new indication.

2. Work as a team. Different members of a research team will have complementary areas of expertise. Involve a biostatistician from the beginning. Research coordinators provide important insights into feasibility and burden to families, as well as expertise in critical regulatory issues. For a multicenter trial, you will likely benefit from working with a clinical coordinating center and/or a data coordinating center, such as the CF Foundation Therapeutics Development Network Coordinating Center. This will ensure expertise in clinical trial coordination, including assignment of a project manager, identification and recruitment of sites, assurance of compliance with all regulatory requirements, assignment of a medical monitor, development of case report forms and statistical analysis plan, development of a template IRB application and consent, review of sites' consent forms, tracking of sites, IRB submissions, tracking of enrollment and online and on-site data monitoring.

3. Identify the primary outcome measure. Depending on the age of the participants, the therapy being evaluated and the indication, this could be a clinical outcome such as pulmonary exacerbations or a surrogate endpoint such as lung function.

4. Identify the target population. What are your eligibility criteria inclusion and exclusion criteria)? Here you must balance feasibility, generalizability, safety, and greatest potential for efficacy. 5 . Perform a sample size calculation and assess feasibility. A major downfall of many studies is an inadequate sample size, leading to lack of power to detect a treatment effect. Be sure to account for drop-outs. Poor enrollment is another major downfall of many studies. Is this sample size achievable given the eligibility criteria, budget and timeline?

6. Define the trial design. Will this be a randomized controlled parallel group trial? Would a crossover design make sense? Can it be adequately blinded? Who needs to be blinded? Discuss these issues with your statistician. 7. Identify potential funding sources. Depending on the size and scope of the trial, these may include NIH (R01, R21, R03, STTR/SBIR), the CF Foundation or other foundations. 8 . Evaluate regulatory requirements. Will you need an IND from the FDA? Will you need a DSMB (generally, yes). Will you need a medical monitor? 9. Draft a protocol. This is helpful to ensure that all the details have been addressed. 10 . Draft a grant application. This is quite different from the protocol, as it is aimed at convincing reviewers of the significance, innovation, impact and feasibility of the proposed study. The protocol is generally attached as an appendix.

11. Consider appointing an executive or steering committee. This committee will regularly review study progress, identify site issues that may need addressing, provide scientific input and possibly review ancillary study proposals.

12. Discuss manuscripts and authorship early. This will minimize miscommunication or hard feelings. Consider the national recognition needed for promotion of junior faculty.

13. Meet regularly. The leadership team should meet regularly to ensure timely study start up and implementation, and then timely writing and submitting of manuscripts.

14. Create and adhere to timelines. This will ensure study completion while funding is still available Choosing appropriate outcome measures is a crucial step for the success of investigator-initiated trials. Outcome measures can be divided into clinically meaningful and surrogate endpoints. The Food and Drug Administration defines clinically meaningful endpoints as those that measure directly how a patient feels, functions or survives. In clinical trials, a surrogate endpoint (or marker) is a measure of effect of a certain treatment that may correlate with a real clinical endpoint but does not necessarily have a guaranteed relationship and acts as a substitute for a clinically meaningful endpoint. Most of the endpoints used in CF clinical studies fall into the latter category. Few outcome measures have been validated and are accepted by regulatory authorities; these include lung function measures, namely FEV1 as well as pulmonary exacerbations requiring treatment with antibiotics. FEV1 is often used as a primary outcome measure, but many patients now have FEV1 values within the normal range which limits patients' eligibility to trials. Flows at lower lung volumes such as MEF2575%VC are more sensitive for small airway disease, but also more variable. Measures of ventilation inhomogeneity such as LCI are both sensitive and less variable and are currently explored as outcome measures for clinical trials, but their ability to predict long term lung function decline and mortality still needs to be defined.

Pulmonary exacerbations are associated with significant morbidity and affect patients' functionality and well being. Most phase III studies have used pulmonary exacerbations requiring intravenous antibiotics as an outcome measure as these have been linked to lung function decline and usually require hospitalization. With overall better clinical status of the CF patient population, these more severe pulmonary exacerbations become less frequent events and therefore a large sample size is required to show a significant difference in a clinical trial. The long term impact of milder exacerbations is less clear and needs to be further explored to identify the validity of these exacerbations as outcome measures in clinical studies.

As it is often challenging to receive adequate support for investigator-initiated trials, there is a tendency to compromise in the choice of outcome measures and these studies are therefore at higher risk of being underpowered to provide a definitive answer to the underlying question. Investigator-initiated trials are also often proof of concept studies rather than pivotal phase III studies. Outcome measures for proof of concept trials can vary based on the therapeutic target (CFTR function versus infection versus inflammation); the range of potential outcome measures is therefore rather broad. However, a priori sample size calculations are often challenging as the information for the biological variability of many of these parameters is often rather limited. In addition, there is not necessarily a strong link between surrogate measures providing biological plausibility and established outcome measures for phase III studies. For example, changes in measures of CFTR function will not necessarily correlate tightly with lung function changes as the threshold for response as well as the reversibility of functional impairment will vary between tissues. Assessing the long term variability of outcome measures as well as their changes during clinically significant events such as pulmonary exacerbations can provide important information on their utility in clinical trials.

Introduction: Over the years there have been several fundamental Investigator initiated studies that have been instrumental to the significant progress in the treatment of cystic fibrosis (CF). These pivotal trials which include the study of anti-infectives to treat lung infection (inhaled tobramycin (TOBI) (Ramsey, 1999) ), and the treatment of lung inflammation in CF (4 year ibuprofen (IBU) study (Konstan, 1995) ) have been the foundation for the conduct of today's investigator initiated trials. The CF Foundation continues to support investigator initiated trials with vigor through financial grant support as well as supporting the infrastructure of the CF-TDN Coordinating Center and Site Network. Investigator initiated trials are also funded through federal agencies (NIH, FDA) and pharmaceutical industry sponsors. The opportunities for the seasoned Research Coordinator (RC) are endless in the role of collaborator for these studies whether single site or more complex multicenter projects. The RC has a unique skill set and perspective that allows for operational success of protocols and allows for RC and principal investigator (PI) collaboration.

The translational opportunities provided by IIT allow basic scientists and physicians to pursue questions that are relevant to the treatment and possible cure of CF. It gives a sense of ownership for work in the pursuit of ground breaking advances in the field, providing a bridge from the bench to the bedside. An IIT also provides academic rewards for all members of the team, including publications in peer reviewed journals, further grant support, promotion/tenure and can lead to a national/international reputation within the community in one's area of expertise. For Research Coordinators, IITs also provide an opportunity to provide expertise in the design and implementation of these trials. RCs are able to provide their skill set, knowledge base and perspective to the project. They are able to maintain the balance between answering an important scientific question, providing a logistically manageable protocol while protecting the rights and safety of our patients participating in clinical research.

Physician/Scientist Role in Investigator Initiated Trials: The physicians who develop investigator initiated trials are, in many cases, the best ones to oversee the research as they have the greatest expertise from a clinical perspective. Local institutional review boards and scientific review boards exist to help maintain the integrity of all types of trials. Also standing beside the physicians, who choose to navigate the head to heart conflict, inherent in translational research, must be a strong well educated, well trained, committed and transparent research team consisting of regulatory specialists, research nurses, clinical research coordinators and laboratory personnel who are committed to the success of bench to bedside research (Radtke, SM).

Protocol design: The RC plays an instrumental role in the protocol design because she/he can determine what is actually feasible in study design, for the team, and, most importantly, for the patients participating in the studies. Often times an overly ambitious research plan can lead to low recruitment and/or retention of subjects. This is usually a result of attempting to "do too much" within a single study design, potentially jeopardizing the practical aspects and performance of the study. An experienced RC will easily spot the study impracticalities and work to answer the questions by designing a study schedule that is both practical and able to achieve the objectives of the protocol. Through the use of tools like PORT CF, internal databases and the TDN CCSM, RCs can review patient population availability based on the inclusion/exclusion criteria designed for a specific protocol.

Logistical Planning: RCs have insight into research center staffing, capabilities, space availability, procedure timing and institutional requirements for study implementation. With this expertise, they are the focal resource for the PI in determining what is logistically feasible with regard to study procedures and schedule. In multicenter IITs, the lead RC can work with RCs at all participating sites to identify site specific clinical practices and constraints to be taken into consideration in designing the final protocol.

Data Collection: The experienced RC is a second set of eyes to determine if all data that is needed to meet objectives of the trial is being collected and instrumental in determining what safety data should be collected and at what timepoints. The RC develops organizational tools for visual flow, study specific source documents and case report forms based on the specific aims of the study. RCs are responsible for ensuring the accuracy and completeness of all study data and can assist with database design and data reduction.

Research Integrity: Another important aspect of the RC role is overseeing the integrity of Investigator Initiated Trials. Regulatory and GCP compliance are of the utmost importance in conducting these studies. Ensuring compliance with institutional and oversight agency policy is essential. Building a fully qualified and trained research team for each study is also within the purview of the RC. Use of available tools, particularly in multicenter IITs, serves to standardize the sites' approach to regulatory compliance, as well as to data capture and source documentation. RCs are trained in the regulatory aspects of clinical trials, with additional knowledge gained through experiences with industry initiated clinical trials. The laboratory investigator initiating a translational study may be unfamiliar with the standard documentation of regulatory compliance and will benefit from this expertise of the RC.

Professional Growth: There are many rewards for research coordinators participating in these trials. The synergistic relationship between the PI, RC and the clinical research team establishes an exciting atmosphere for advancement in the treatment of CF while answering important questions to improve care and quality of life for patients and families with CF. IITs offer an opportunity for Research Coordinators to be involved in the authorship process providing their input and expertise while manuscripts are drafted and review and finalization for publication. Most importantly, recognition of the RC role gives one a sense of ownership for the project and allows for professional growth. This experience furthers the commitment of the RC to the success of clinical research, while pursuing knowledge about mechanisms of disease and treatments with the goal of improving outcomes for all patients with CF.

References Investigator-initiated trials are a rewarding experience and require the expertise of an entire cystic fibrosis (CF) research team. Investigator-initiated trials allow an investigator to determine the question to be answered and independently decide the optimal study design to answer that question. Challenges of investigator-initiated trials include obtaining funding, developing a plan to comply with regulatory issues, and maintaining focus to continue working through the many obstacles that will be encountered. Successfully completing a high-quality investigator-initiated trial requires extensive planning and input from all members of the CF research team. The first step is to ask a question.

Like many research projects, this one began with a question during a clinic visit. During fellowship, an adult CF patient with chronic respiratory tract Pseudomonas aeruginosa and methicillin-resistant S. aureus (MRSA) presented with increased symptoms after receiving a course of IV antibiotics directed against P. aeruginosa. Since respiratory infection is a major contributor to morbidity and mortality in CF (1), it was important to know if MRSA could be contributing to this patient's symptoms. A review of the literature suggested that methicillin-sensitive S. aureus (MSSA) may be associated with a milder disease course in CF patients as MSSA is associated with increased survival in a 5-year survivorship model of CF (2) . While MSSA may be associated with better outcomes, there was not a lot of data on the impact, much less treatment, of MRSA in CF at that time. There is concern about MRSA being more virulent than MSSA because of differences in virulence factors and MRSA being associated with lung abscesses in CF patients (3, 4) . Since that day in clinic, the CF community has had increasing concerns about MRSA as the prevalence has continued to increase dramatically and is now detected in 25% of all CF patients (5) . Adding to the concern about MRSA in CF, Dasenbrook and colleagues have published cohort studies using the CF Foundation Patient Registry to show an association between MRSA infection in CF and more rapid lung function decline and worse survival (6, 7) . Furthermore, a recent study by Sanders and colleagues of 8,500 CF exacerbations reported MRSA was a significant risk factor for failure to recover lung function (8) . These studies support the conclusion that MRSA may be a modifiable risk factor for worse outcomes in patients with CF. Therefore, the next step is to determine if treatment of respiratory tract MRSA will improve patient outcomes.

Once the question has been identified, creating a comprehensive research protocol is the most important task. Fortunately, the CF Therapeutics Development Network has a template available to investigators to assist in the development of a thorough protocol. The template serves as an excellent way to organize the project and to begin the task of designing a high quality controlled clinical trial. The protocol template includes sections on the background, study rationale, objectives of the study, study design, primary and secondary endpoints and safety analyses, selection of subjects, a thorough description of the test and control medications, study procedures and guidelines, evaluations by visit, adverse experience documentation, statistical methods and considerations, data safety monitoring plan; data collection, retention and monitoring; administrative, ethical, and regulatory considerations, and provides an example of a schedule of study visits table. For the Persistent MRSA Eradication Protocol (PMEP) trial, the template was an invaluable tool in organizing a comprehensive protocol that evolved after review by the clinical research team, colleagues who conduct clinical trials, the FDA, and peer reviewers after grant application submissions.

An important issue to consider is, investigator-initiated trials of a new drug, or more commonly, off-label use of a drug that is already marketed, may require the investigator to submit an investigational new drug application (IND) to the Food and Drug Administration (FDA). In the application, the FDA will primarily request information that is already contained in the study protocol, reinforcing the importance of a well vetted protocol. Arbit and colleague have published an overview of preparing an IND application to the FDA for investigator-initiated clinical trials (9) . Furthermore, many universities offer a service to help investigators plan an IND application to the FDA. For the PMEP trial, the Johns Hopkins Institute for Clinical and Translational Research Drug and Device Resource Service was consulted and assisted in the IND submission. After a safety review of the application, the FDA concluded that the PMEP trial could proceed with an IND for sterile vancomycin 250 mg reconstituted in 5 mL sterile water and nebulized.

Once a comprehensive, vetted research protocol is assembled, the next step is to obtain funding. Most investigator-initiated studies are phase I or phase II studies, which can be very expensive to conduct. There are many resources to help defray the costs of conducting the trial. CF Foundation Therapeutics (CFFT) offers competitive awards to support clinical research projects directly related to CF treatment and care. The criterion for this award includes any projects that "address diagnostic or therapeutic methods related to CF or the pathophysiology of CF." These are 3-year awards and can either be a single center or multicenter trial. This award is the source of funding for the PMEP trial. Frequently, investigators will receive support from both CFFT and a governmental agency. The FDA Office of Orphan Products Development is another source of funding for investigator-initiated trials. They have an R01 mechanism that supports the clinical development of products for use in rare diseases such as CF. Goss and colleagues are using this mechanism to perform a pharmacokinetic and safety study of IV gallium nitrate in patients with CF (http://clinicaltrials.gov/NCT01093521). For investigators interested in planning a large, multicenter, Phase 3 trial, the National Institutes of Health's U01 cooperative agreement is another available mechanism that was recently used by Rosenfeld and colleagues to conduct the inhaled hypertonic saline in infants and young children randomized trial (10) . In addition to the CF Foundation and government agencies, many pharmaceutical companies have programs for investigator-initiated studies where the company will supply drug and research costs, but the investigators maintain control over the study design and analysis of data.

In conclusion, conducting an investigator-initiated trial is an extremely rewarding experience and requires the expertise of an entire research team. In CF, we are fortunate to have an outstanding clinical research infrastructure to help guide the development of a high quality clinical trial. The PMEP trial is an example of the teamwork involved in conducting an investigator-initiated Phase 2 CF trial and will provide important information about the safety and potential benefits of inhaled vancomycin.

References 

At the time of publication of the Infection Control Guidelines for CF a decade ago (!) [1] , the CF Foundation and co-authors made a commitment to update the guidelines as new information became available. Both CF and infection prevention and control (IP&C) are dynamic disciplines that develop new practices in response to new science and identification of new challenges. In addition, the CF Foundation is committed to developing new care guidelines that incorporate the use of systematic reviews and a priori procedures for crafting recommendations developed by consensus [2, 3] . Hence, a committee consisting of representatives from the interdisciplinary CF care team as well as people with CF and parents of people with CF has been convened to update the IP&C Guidelines. Strategies for updating the guidelines include review of relevant studies in CF, incorporation of relevant national guidelines (e.g., the Centers for Disease Control & Prevention and Healthcare Infection Control Practices Advisory Committee guidelines for Hand Hygiene, Isolation Precautions, and Management of Multidrug Resistant Pathogens, www.cdc.gov/hicpac), and systematic reviews of selected issues if evidence is available.

Update on Clinical Impact of CF Pathogens and Patient-to-Patient Transmission: Studies have continued to show the adverse clinical impact of shared strains acquired through patient-to-patient transmission. For example, Aaron et al demonstrated two dominant clones of Pseudomonas aeruginosa among CF patients in Ottawa; one of these clones, found to be genetically identical to the Liverpool epidemic strain [4] , was associated with an increased rate of death or lung transplantation within a 3-year observation period [5] . Similarly, patients infected with the same Burkholderia dolosa strain had greater decline in lung function and increased risk of dying [6] . Most recently, patients attending the same CF clinic were found to be infected with the same strain of Mycobacterium abscessus subspecies massiliense [7] . Furthermore, Dasenbrook et al. have shown using CFF registry data that chronic infection with methicillinresistant Staphylococcus aureus (MRSA) was associated with greater decline in lung function and increased mortality [8, 9] . Finally, Stone et al. demonstrated that MRSA strains were shared between CF patients and their family members [10] .

Continued Importance of Basic IP&C Principles: Hand hygiene remains the most basic intervention to prevent transmission of potential pathogens. While hand hygiene efforts typically focus on health care providers, it is critical that hand hygiene also be performed by people with CF. Zuckerman et al. have demonstrated that the hands of CF patients can become contaminated with their respiratory tract pathogens during CF clinic despite performing hand hygiene at the start of clinic [11] . Notably, Miroballi et al found that 20% of CF patients or their parents reported that their CF care team had not discussed opportunities to perform hand hygiene with them and only 66% reported they had been told to do so after coughing [12] .

The use of personal protective equipment (masks, gloves and gowns) has been promulgated as means of preventing transmission of potential pathogens. While gloves and gowns are generally reserved for health care providers caring for patients on Contact Precautions, the use of masks by symptomatic patients, so-called "Respiratory Hygiene/Cough Etiquette," has been advocated to prevent transmission of pathogens spread by droplets, such as influenza and SARS, by containing the source of infectious agents [www.cdc.gov/hicpac/pdf/isolation/Isolation2007]. Masks may also play a protective role and prevent acquisition of potential pathogens by non-immune individuals. Notably, in the 2003 IP&C guidelines for CF, consensus for the use of masks by CF patients was an "Unresolved Issue" as available data had not supported their use; transmission of Burkholderia had been halted by implementation of other infection control measures without the use of masks by patients [1] .

The IP&C principle Standard Precautions teaches that if a health care provider is likely to be in contact with infectious body fluids, such as respiratory tract secretions, personal protective equipment should be worn, regardless of the known infection/colonization status of a patient. This is because any patient could potentially harbor an unknown transmissible pathogen. This principle is particularly relevant for CF patients as identification of microorganisms from respiratory tract cultures may lag behind true pathogen status, despite appropriate processing by the clinical microbiology laboratory.

In addition, our understanding of droplet transmission in CF may be incomplete as suggested by recent observations and experimental systems. While the "3 foot rule" has been advocated as the minimal safe distance between patients intended to minimize droplet transmission, infectious droplets with influenza or coronavirus may travel as far as 6 feet [13] . Furthermore, in a forced cough model, infectious nuclei were generated by CF patients infected with P. aeruginosa and some traveled further than 3 feet [14] . Finally, air may remain contaminated with CF pathogens after patients have left the room. Such observations may have implications for mask use by CF patients, particularly in CF clinics.

Special Circumstances: Additional areas for which consensus guidelines will be developed include recommendations for non-healthcare congregate settings in which more than one person with CF may be present such as Great Strides, fund raising events, or pharmaceutical company sponsored events. Finally, the guidelines will also consider the potential role of animals as vectors for pathogens including pet therapy animals, personal pets and farm animals.

Next Steps: The IP&C committee will develop evidence based recommendations when feasible and grade them using the U.S. Preventive Services Task Force grading system [15] . Consensus recommendations will be made for areas for which appropriate published guidelines are available or if systematic reviews were not performed due to limited evidence. Public comment from the CF community will be sought prior to finalizing the guidelines.

References 

John LiPuma, M.D.

As the interdisciplinary committee that has been convened to update the Infection Control Guidelines for CF continues its work, it is clear that the clinical microbiology laboratory remains at the heart of our efforts to provide the data needed to inform our recommendations. Infection control is predicated on our ability to reliably recover and accurately identify infectious agents from human and environmental sources. Microbial genotyping (fingerprinting) remains a critical tool for tracking the acquisition and spread of specific pathogenic strains, and provides a means to monitor the success of infection control policies and practices. However, several unanswered questions challenge our ability to define the optimal use of the microbiology laboratory in infection control efforts.

Because the acquisition of opportunistic pathogens by persons with CF is typically "silent" (i.e., not associated with overt signs or symptoms), detection of infection relies on active microbiologic surveillance. An important question in this regard pertains to the optimal frequency with which surveillance cultures should be obtained. The answer to this question is, however, complicated by our incomplete understanding of the natural history of opportunistic infection in CF. Intuitively, it would seem that a greater frequency of surveillance cultures would afford a greater probability to detect potentially transmissible species, thereby providing more opportunity to intervene with appropriate infection control. However, this granularity of information needs to be balanced against feasibility and carefully considered cost-benefit analyses that are generally unavailable. It is also likely that the optimal frequency of microbiologic surveillance varies with respect to patient age, disease stage and previous culture results.

An important corollary question impacting on infection control policies pertains to our ability to determine when an individual is no longer infected with a specific pathogen. This is especially relevant to current microbial "eradication" treatment strategies wherein aggressive antibiotic use to eliminate initial (or recurrent) infection has become standard practice. Relatively short term follow-up data exist for assessing the success of this strategy for P. aeruginosa infection, and studies of MSRA eradication are currently underway. Nevertheless, little, if any, firm data are available to guide "post-eradication" surveillance practices -and definitions of "infection-free" status (i.e., after previous positive culture) remain elusive. Again, our limited understanding of the natural history of infection with these species in CF is an obstacle to developing firm evidence based guidelines.

Strain genotyping provides the foundation for studies describing the epidemiology of bacterial infection in CF. Several shared -or so-called "epidemic" -strains of P. aeruginosa and B. cepacia complex have been described, and infection with some of these strains has been associated with a greater risk of poor clinical outcome (1). A great number of studies have documented inter-patient transmission of such strains within CF centers, and have served as the basis of implementing infection control measures that have effectively halted further transmission. Since publication of the 2003 Infection Control Guidelines for CF (2) , genotyping studies have enhanced our understanding of the epidemiology and clinical impact of B. cepacia (3) and P. aeruginosa (4, 5) and have also been used to demonstrate that strains of methicillin-resistant Staphylococcus aureus (MRSA) may be shared between CF patients and their family members (6) and that patients attending the same CF clinic may be infected with the same strain of Mycobacterium abscessus (7) . While there is little doubt, based on observations such as these, of the value of genotyping in infection control, there is little evidence-based data upon which recommendations can be made about the optimal frequency of genotyping or the signals that might indicate a specific need for genotyping. Nevertheless, given the success that genotyping has had in identifying spread of specific bacterial strains and guiding effective local infection control measures, it seems appropriate that active genotypic surveillance should be considered an important part of infection control in CF.

Finally, as our appreciation of the complexity of infection in CF continues to expand, particularly within the context of recent CF airways microbiome studies (8), we must be prepared to update infection control guidelines with attention to species not previously included in the repertoire of "typical" CF pathogens. Examples include recommendations regarding the recovery, identification, and genotyping of non-tuberculous mycobacteria (7), recovery of anaerobic species (9) and S. aureus small colony variants (10) , and identification of species within the Streptococcus milleri group (11) . As methods for the rapid and reliable detection of viral respiratory pathogens continue to improve, we will need to consider how best to incorporate these in infection control practices in CF (12) .

Despite these challenges, the considerable improvements made in the recovery, identification, genotyping and detection of microbial pathogens in CF provides ample opportunity to enhance infection control recommendations that will contribute to improved outcomes.

References This presentation will review current information about the role of the inanimate environment in transmission of CF pathogens from a) the natural environment; b) surfaces and water in the home and healthcare facilities; and c) medical equipment and solutions. An understanding of the extent to which organisms in the environment may pose a threat to patients with CF and methods of preventing transmission will reduce the risk of acquiring virulent pathogens and improve outcomes. It is important to demonstrate transmission from a source to a patient in addition to isolation of a specific pathogen from an inanimate source.

CF Pathogens in the Natural Environment: Strict infection control measures to prevent patient-to-patient transmission of Burkholderia spp have reduced but not eliminated infection with certain strains. Acquisition of distinct strains now accounts for the majority of new Burkholderia infections. Some Burkholderia spp are recognized plant pathogens. In 1999-2000, B. cenocepacia isolates recovered from soil samples from New York state agricultural fields that had been planted with onions for several years were indistinguishable from the B. cenocepacia strain that was shared by many patients with CF in the mid-Atlantic region (1). Subsequently, these strains were also found in other cities in the US and Europe and were indistinguishable from isolates from patients with CF (2). Burkholderia spp have also been found in rice, wheat and maize rhizospheres, industrial settings, and human sewage. Of note, the species of Burkholderia found most frequently in the natural environment are found infrequently in CF patients. In contrast, B. multivorans is isolated frequently from CF patients, but only rarely from the environment. P. aeruginosa is widely distributed in the natural environment, mostly in aquatic habitats. Most strains found in aquatic environments are not the same as CF patient strains, with the exception of P. aeruginosa clone C (3).

Healthcare facilities. Most studies of hospital inpatient and outpatient environments do not find a reservoir for shared strains. Although P. aeruginosa was isolated in 19% of surface samples of showers after hospitalized colonized patients had washed, environmental strains were different from clinical strains and no P. aeruginosa was detected before washing (4). In one study of a CF center with a high patient colonization rate with the more virulent Liverpool strain, patient clothes, bed linens, and external surfaces of the patients' personal respiratory therapy equipment were infrequently and transiently colonized (5) . Another study of bacterial surface contamination in 7 different CF clinics found that respiratory tract pathogens were recovered from only 1% of inanimate surfaces in the area of patients known to be colonized with those pathogens, most often from receptionists' counters, bathroom faucets and waiting room chairs; pulse oximeters and stethoscopes were rarely contaminated (6) . In contrast, one study from 2001-2005 recovered P. aeruginosa from 23% of samples from inanimate surfaces, mostly from sinks and drains. Twenty percent of the environmental isolates were genotyped and compared with clinical isolates: 19 of the 21 environmental strains were found to be genetically related to the clinical strains. In this study, it was estimated that each patient who was not colonized with P. aeruginosa had a 5.4% risk of coming into contact with a surface colonized with P. aeruginosa at each clinic visit (7) . Although an early study of Stenotrophomonas maltophilia in patients with CF suggested the potential for acquisition from inanimate surfaces in contact with water, a more recent study demonstrated that isolates associated with persistent colonization of water, traps, and sinks were different from clinical isolates (8) . The variations in these study results may, in part, reflect variations in cleaning and disinfection practices in CF centers. In conclusion, studies to date support the recommendations for surface cleaning and disinfection, but indicate that inanimate environmental contamination is a less frequent source of new acquisition of pathogens.

Contaminated products. Burkholderia spp are the most frequently isolated bacteria in non-sterile and sterile pharmaceutical products that have been recalled. There have been many healthcare-associated outbreaks of Burkholderia infections associated with contaminated skin antiseptics, mouthwash, medications, medical devices, etc., that could serve as a source of Burkholderia spp in CF patients (9) . Although there are no published reports of proven acquisition by CF patients from such products, such sources present a risk for patients with CF. Also, multi-dose vials of medications, especially albuterol, should not be used for >1 patient and unit dose vials are preferred even for the same patient due to outbreaks that have occurred after contamination of medication vials.

Homes. Similarly, studies of the home environment of patients with CF have isolated P. aeruginosa from environmental samples, mostly in bathrooms, but only rarely are the genotypes of patient strains and environmental strains the same. It is likely that the contamination that occurs is from the patient to the environment, which may be of particular concern for multiple patient families. Thus, reasonable surface cleaning and disinfection will likely continue to be recommended, but should not be the primary focus of infection prevention efforts in the home.

CF pathogens recovered from respiratory therapy equipment. The most challenging question about the role of the inanimate environment in transmission is the optimal management of nebulizers and other respiratory therapy equipment, both in the home and in healthcare facilities. Nebulizers are classified as semi-critical items because they come into contact with mucous membranes, but do not penetrate body surfaces. Several studies have recovered CF pathogens from nebulizers and contamination of nebulizers was associated with transmission and therefore, cleaning, disinfecting, rinsing and air drying after each use was recommended in the 2003 guideline for home and in hospitals. Studies of nebulizer care published since the 2003 have found: a) frequent failure to follow recommended practices; b) contamination with CF pathogens when recommended practices are not followed; c) failure of vinegar (acetic acid) to remove S. aureus and E. coli; d) no advantage of hypochlorite based solutions; e) failure to find microbial contamination of the cup of disposable nebulizers after 24 hours; therefore, consideration may be given to disposal of a nebulizer after 24 hours of use when in hospital (10) . Recommendations for care of nebulizers and pulmonary function test equipment will be a priority for the current literature review and guideline update.

Use of cleaning -monitoring protocols. There are now several published reports that demonstrate the benefit for preventing transmission of infectious agents of a) educating personnel who are responsible for cleaning healthcare facility surfaces and b) monitoring the cleaning process using objective measures, (e.g., fluorescent marker systems, ATP bioluminescence assay) (11, 12) . Use of monitoring systems with feedback to the staff members may be beneficial for baseline and post-intervention assessments in ambulatory and inpatient areas where patients with CF receive care.

References 

Jonathan B. Zuckerman, M.D.

Pulmonary/Critical Care, Maine Medical Center, Portland, ME, USA Background: CF patients face not only the reality of developing respiratory tract infection from environmental sources but also the prospect of acquiring new pathogens from other affected individuals. Acquisition of certain bacteria may associate with accelerated pulmonary decline and shortened life expectancy. In some circumstances, changes in respiratory flora may limit options for lung transplantation. Due to the weighty importance of lung function in CF, the issue of infection control is emotionally charged for patients and their families, often provoking fears, protests and a strong desire to negotiate with providers. Clear evidence for inpatient infection control has grown in the past few decades. However, less is known about the risk of cross-contamination in the outpatient setting where residency is relatively brief but where patient traffic is often heavy.

While elucidating the precise mechanisms of bacterial acquisition among CF patients is an area of considerable research interest, there remain practical issues that confront CF programs in order to appropriately balance workflow and financial realities against the goal of maximal patient safety. Longitudinal outpatient care at CF centers is felt to be key to improved outcomes (1). It is therefore critical to address remaining infection control concerns in this setting. For example, in the 2003 evidence-based guidelines for CF infection control, a major unresolved issue was whether or not to advocate the use of masks by CF patients in clinics (2) , as the committee did not feel there were adequate data to recommend this practice for all patients.

Lack of clarity in this area has contributed to variance in practice at CF centers with respect to mask use in clinics. In a survey of 190 U.S. CF care sites, 35% of the programs that responded to a question about outpatient mask use had policies advocating patient use of such protection, despite the absence of a mandate in the practice guidelines (3) . Masks are used to protect healthcare workers from contact with pathogens associated with infected patients, to limit transmission of potential infectious pathogens from coughing patients to others, and to protect patients undergoing sterile procedures from pathogens carried in the respiratory tract of healthcare workers (4) . The effectiveness of surgical masks to protect healthcare workers (5), college students (6), and family contacts (7) from seasonal influenza and other respiratory viruses has been studied, but adherence to mask use is problematic (8) . None of these previous studies may be relevant to preventing CF patient-to-CF patient transmission of bacterial respiratory pathogens in the "real world" outpatient setting, and this served as the basis for the MASC (Management of Air and Surface Contamination in CF clinics) trial.

Contact and droplet transmission: Contact and droplet transmission are recognized routes of cross contamination with CF pathogens. In an initial prevalence survey and before-after trial in outpatient clinics at 7 CF centers we examined patients with positive respiratory cultures for Pseudomonas aeruginosa (PA), Staphylococcus aureus (SA), Stenotrophomonas maltophilia (SM) or Burkholderia species (BS) (9) . Hand carriage and environmental contamination with respiratory pathogens was assessed during clinic visits (Part I) and the effectiveness of hand hygiene performed by CF patients (Part II) was determined using molecular typing of recovered isolates. In Part I, respiratory pathogens were cultured from patients' hands (7%), the exam room air (8%), and less commonly, environmental surfaces (1%). In Part II the hand carriage rate of pathogens was 13.5% and 4 participants without initial detection of pathogens had hand contamination when re-cultured at the end of the clinic visit.

Airborne transmission: The possibility of airborne transmission of CF respiratory pathogens is still somewhat controversial, as nicely described in a recent review (10) . Recovery of pathogens from the air has been demonstrated at different rates depending in part upon experimental conditions. One study of adult CF patients harboring BS showed that air samples (obtained within about 1 meter from the patients) were positive before, during and after chest physiotherapy, respectively (11) . BS was detected in air samples obtained 15 to 45 minutes after the patients left the room, but samples obtained at 60 minutes were negative. Based on this study and others, CF patients have been routinely advised to remain at arm's distance from each other (the "3 foot rule"). Accumulating evidence now challenges the safety of this rule. Recent CDC guidelines state that infectious droplets may project at least 6 feet, farther than previously thought to occur (4) . Data supporting this revision are largely derived from observations of the transmission of viral pathogens and thus, their applicability to CF is uncertain. One recent study in CF patients showed that droplets contaminated with PA sometimes project more than 1 meter from subjects during talking or coughing under experimental conditions (12) . An intriguing study by Wainwright et al. measured the size of potentially infectious droplets generated during forced coughing in CF volunteers. A large proportion of the expectorated infectious droplets were within the respirable range under test conditions (14) . Based largely on epidemiologic observations of viral pathogens, Roy and Milton proposed a schema that modifies the classic categories of transmission to include obligate (only airborne transmission), preferential (can be transmitted by other means such as droplet), and opportunistic transmission (under some circumstances airborne transmission occurs) (13) . It should be emphasized that the relevance of these concepts to infection control in CF remains uncertain.

We felt that a logical next step after our first multicenter outpatient infection control trial was to carefully evaluate patient-derived bioaerosols during typical office visits. Eligible subjects were attending clinic for routine or sick visits and were infected with PA, SA, SM and/or BS. For this study, spirometry was performed in a separate room from the remainder of the encounter (during which time subjects were randomized to wearing or not wearing a surgical mask). Air samples were collected with a single stage impactor (Andersen 1-STG) that drew ambient air onto a blood agar plate. The impactor was positioned >6 ft from the subject and air was collected for 10 min per sample. The following samples were obtained for each subject: during spirometry, 30 min after spirometry, and in the exam room. Plates were processed at a core laboratory using selective media, and bacterial isolates were characterized. The proportion of subjects with the same species isolated from > 1 air sample(s) and historic and/or current respiratory tract specimens was determined. The air exchange rate in each room was measured by the carbon dioxide decay technique. The rate of bacterial air contamination with CF respiratory pathogens was similar to that detected in our previous study. Air contamination was much more frequent during performance of spirometry than during other elements of office encounters and appeared to clear by 30 min after performance of testing. Additional discussion during this session will explore possible associations between rates of air contamination and patient mask use, carriage of specific respiratory pathogens, patient clinical status, cough frequency and exam room ventilation. Pancreatic insufficiency (PI) is a term used to define patients who have lost a significant amount (usually more than 95%) of pancreatic exocrine function and therefore their ability to digest and assimilate nutrients normally. Cystic fibrosis (CF) is by far the most common form of PI in children (1) .

Pancreatic disease starts in utero in CF; significant pancreatic involvement can be found in young infants (2) (3) (4) , and lesions can even be observed in aborted fetuses (5, 6) . Overt PI is present at birth in 50-65% of people with CF, and 20-30% of pancreatic sufficient (PS) patients become insufficient during the first few months and years of life (7) (8) (9) . Pancreatic lesions in CF vary considerably in severity, but in general progress with increasing age (2-4) until complete loss of organ function in ~85% of patients with CF (7, (9) (10) (11) (12) . Whilst sufficient exocrine pancreatic function is present and pancreatic enzyme supplementation is not needed, the pancreas of CF patients with pancreatic sufficiency is never normal. The pancreas is involved in CF irrespective of PS or PI status.

Pancreatic enzyme replacement therapy has been the mainstay for patients with CF and PI, but the current therapy of PI with exogenous pancreatic enzymes is far from perfect. These enzymes do not effectively treat the malabsorption, maldigestion, growth delay, and gastrointestinal symptoms of CF patients with PI (13, 14) . Pancreatic insufficiency has a major impact on growth, nutrition and lung disease (10, 15, 16) , and currently there are no treatments to halt the progression of pancreatic disease in CF.

We found that the CF pigs have pancreatic disease at birth with reduced number of acini, decreased cytoplasmic zymogen granules, and ectatic and plugged ducts surrounded by degenerative exocrine tissue (17) (18) (19) (20) (21) . We identified that neutrophils and macrophages were scattered within dilated acini whereas lymphocyte aggregates were sometimes prominent in the interstitium, but usually found adjacent to the dilated ducts (21) . We studied the subtypes of inflammatory cells in the newborn CF pig pancreas to investigate whether a specific inflammatory pathway was activated. We discovered that there was activation of both the innate and adaptive immune systems in the newborn CF pig pancreata. In contrast, the inflammatory cell populations in the blood and mesenteric lymph nodes were not different between CF and non-CF pigs, suggesting that the pancreatic immune response was a localized rather than a systemic process (21) .

We discovered that as in humans, the pancreatic disease started in utero in CF pigs and progressed over time to complete destruction of the organ. Fetal CF pig pancreata (83-90 day gestation, normal pig gestation is~114 days) had patchy inflammation and acinar atrophy compared to non-CF, with progression in distribution and severity in CF neonates. Large scale transcript profiling revealed that CF fetal and newborn pig pancreata had significantly increased expression of proinflammatory, complement cascade, and profibrotic genes compared to non-CF (22) . Acinar cells exhibited increased apoptosis in CF pig fetal and newborn pancreas. Alpha smooth muscle actin and transforming growth factor beta-1 were increased in both CF fetal and newborn pig pancreata, suggesting activation of profibrotic pathways. Cell proliferation and mucous cell metaplasia were detected in newborn and not fetal CF pigs, indicating that they were not an initiator of pathogenesis, but rather a response (22) .

Two findings in human CF pancreas suggest that inflammation may be contributing to the disease process. First, pancreatitis is a well-known complication of CF in humans with the PS phenotype and it is an important factor leading to PI in these patients (23, 24) . Second, inflammatory cell infiltrates (mixed cell-type) are well-known features of human CF pancreatic pathology, as described in the early autopsy studies (2, 4, 25) .

The findings in humans and in pigs suggest that the CF pancreatic disease starts in utero and advances to insufficiency of the organ via mechanisms that may involve proinflammatory, complement cascade, proapoptotic, and profibrotic pathways. Our studies suggest that the porcine model may be useful for investigating the pancreatic disease mechanisms in CF and that knowledge might enable the development of innovative therapies.

Direct pancreatic function tests such as the secretincholecystokinin test have the highest sensitivity and specificity and remain the gold standard in the assessment of exocrine pancreatic function. However, direct tests are time-consuming, expensive, inconvenient and not well standardized in children. Indirect tests are used routinely in clinical practice because they are non-invasive, simple, less time-consuming and cheaper. The development of new indirect tests has improved diagnostic approach.

Elastase-1 is highly specific for the pancreas and is not degraded during the intestinal passage. Its concentration in feces is five to six times higher than in pancreatic juice and it has been suggested to reflect exocrine pancreatic function well. Fecal elastase levels are significantly correlated to duodenal elastase concentrations as assessed with secretin-cholecystokinin test but probably not under sole secretin stimulation. Fecal elastase test is highly sensitive in severe exocrine pancreatic insufficiency, however, its applicability in milder forms of pancreatic involvement is limited. The use of this test for a long-term observation of declining pancreatic function in individual subjects has been suggested. However, the available data are contradictory.

In preliminary studies low daily variations of fecal elastase levels were documented. However, for higher concentrations discordant results were obtained and for the borderline values outcoming fact may have clinical implications. Moreover, suggested cut-off levels for fecal elastase test are not so obvious. No enzyme degradation over a period of 1 week at room temperature and over 1 year deep freezing was stated. Enzyme replacement therapy has no influence on fecal elastase concentrations in pancreatic insufficient cystic fibrosis patients. Nor was any elastase immunoreactivity found in market preparations. However, exogenous pancreatic enzymes were documented to inhibit endogenous pancreatic secretion in healthy subjects.

Fecal elastase test can be considered to be potentially a reasonable tool in the assessment of exocrine pancreatic function regardless of the cause of pancreatic insufficiency or diminished pancreatic secretion. Specificity of the test in healthy subjects was documented to be very high. However, in many clinical entities, a considerable percentage of false positive results was documented. Fecal elastase concentrations were found to be decreased in celiac villous atrophy and different forms of enteropathy as well as in patients with diabetes mellitus. Similar effect was observed, at least partially due to dilution effect, in acute diarrhea that could be overcome with sample drying. Specific diet (e.g., vegetarian) and enteral/ parenteral feeding may also potentially influence obtained results. Age-specific variations, especially related to the youngest age, should also be considered. In fact, a vast majority of available data on the measurement of fecal elastase is related to the use of monoclonal antibodies based test. Recently, a new polyclonal assay has become available. It seems, however, that direct extrapolation of all data, published for monoclonal test, raises some doubts.

It has been shown that fecal elastase is more sensitive than the other fecal tests. Similarly, it seems to be more practical than all indirect tests available. Therefore, it has become an accepted test for the assessment of exocrine pancreatic function. Having in mind all known limitations, the test might be very useful in clinical practice. Cystic fibrosis (CF), termed after the autopsy findings of "cystic fibrosis of the pancreas" in malnourished infants [1] , is caused by mutations in the gene that encodes for the cystic fibrosis transmembrane conductance regulator (CFTR) protein [2] . The CFTR protein functions on the apical surface of epithelial cells as a cyclic AMP-dependent chloride channel, a bicarbonate channel and as a modulator of other ion channels [3] [4] [5] [6] [7] [8] .

Functionally, the pancreas can be divided into its exocrine and endocrine components. The former is composed of pancreatic acinar and ductular cells, while the latter is primarily composed of pancreatic islet cells. Although CFTR is predominantly expressed in the pancreatic ductal epithelium, both exocrine and endocrine components of the pancreas are affected, to varying degrees, in CF. In health, the human pancreatic ductal epithelium secretes large volumes (1-2L per day) of alkaline fluid made of sodium chloride and bicarbonate, as well other cations (e.g. potassium) [9, 10] . The pancreatic secretion is physiologically intended to flush digestive enzymes secreted by pancreatic acinar cells down the pancreatic-biliary tree and into the duodenum. The alkaline solution also functions to alkalinize acidic chyme emptied from the stomach into the duodenum, and to provide the optimal pH milieu for pancreatic digestive enzyme activity.

In CF disease, exocrine pancreatic damage in its severe form is known as pancreatic insufficiency (PI).

Pancreatic damage begins in utero and continues into infancy or early childhood when complete loss of pancreatic acinar tissue occurs [11, 12] . Only 1-2% of residual pancreatic reserve is required to maintain pancreatic sufficiency (PS) [13] . Although patients with PS CF have sufficient exocrine pancreatic function to maintain normal nutrient digestion without the use of pancreatic enzyme supplements, it is not indicative of normal pancreatic function. However, patients classified as PS demonstrate a very wide range of exocrine pancreatic function, with variations in colipase and lipase secretions of over 250-fold. Pancreatic bicarbonate secretion is impaired in PS CF, even among those with normal pancreatic enzyme output [14] .

The PS status provides insight into the phenotype of affected patients since the exocrine pancreas is the most reliable phenotypic barometer of CFTR function [13, 14] . Patients who carry a mild mutation on at least one allele usually have PS i.e. mutations that confer the pancreatic sufficient phenotype do so in a dominant fashion. PS CF patients are often diagnosed at an older age (notably in late childhood and adulthood), have more subtle disease manifestations, lower sweat chloride concentrations, and better outcomes than PI CF patients [15] [16] [17] . Exceptions occur, in part, due to the complex interactions between CFTR and genetic and/or environmental modifiers [18, 19] .

Unlike PI patients, PS patients are at risk of symptomatic pancreatitis. The presence of a "critical mass" of pancreatic acinar tissue is necessary for symptomatic pancreatitis to occur. However, not all PS CF patients develop pancreatitis, suggesting variability in disease development due to variations in CFTR dysfunction or presence of modifier effects or both. To evaluate the former, a novel classification system for mutation severity, known as the Pancreatic Insufficiency Prevalence (PIP) score, was recently developed and validated [20] . The PIP score for a specific mutation is calculated based on the ratio between PI patients carrying the mutation and all PI and PS patients carrying the same mutation in a homozygous state or heterozygous in a combination with mutations known to have severe consequences, such as F508del or most Class I mutations. The PIP scoring system permitted a more refined classification of the functional severity of CFTR mutations (than can be provided by the conventional 5 class system) and in a far greater number of patients than previous method(s) of classification. Furthermore, this scoring system could be applied to other genotype-phenotype studies. Categorically, mutations can be classified as either mild (≤0.25) or severe (>0.25) on the basis of the PIP score.

In the largest systematic study to date, involving 277 PS CF subjects with and without pancreatitis, the risk of pancreatitis was reported to be associated with severity of CFTR genotype [20] . Patients with PS CF who developed pancreatitis were more likely to carry mild (PIP score ≤0.25) than severe (PIP score >0.25) genotypes. In addition, patients carrying mild genotypes had a significantly higher cumulative proportion and 71% increase in risk of developing pancreatitis at any given time (hazard ratio = 2.4) than those with moderate-severe genotypes. Furthermore, the presence of mild mutations on both alleles conferred a significantly higher cumulative proportion and greater risk of pancreatitis than those with mild/severe and severe/severe allele combinations. In fact, there appeared to be a gradation of risk of developing pancreatitis according to genotype severity. At any given time, patients carrying mild mutations on both alleles (mild/mild) had the greatest risk of pancreatitis (hazard ratio = 4.2) which is followed by patients carrying one mild (mild/severe; hazard ratio = 1.9), when compared with those who carried severe mutations on both alleles. Other phenotypic differences between PS CF patients with and without pancreatitis supported the contention that, as a group, those at risk of pancreatitis carried milder mutations. Patients with pancreatitis were diagnosed with CF at a significantly older age than those without pancreatitis (median (IQR) ages at diagnosis of 14.9 (9.5-27.7) vs. 9.3 (1.5-21.4); P=0.003). Further, patients with pancreatitis had a significantly lower mean (SD) sweat chloride than patients without pancreatitis (74.5 (26.2) vs. 82.8 (25.2) mmol/L; P=0.03).

Monitoring of exocrine pancreatic function on at least an annual basis is recommended due to the risk of pro-gression from PS to PI. Development of steatorrhoea, poor weight gain and/or fat soluble vitamin deficiency should also prompt re-evaluation. In the majority of cases, PS CF patients who carry mild genotypes are expected to remain PS. However, in the presence of symptomatic pancreatitis, there is an increased risk (odds ratio = 5.5) of PS CF patients developing exocrine pancreatic insufficiency [20] . The progression from PS to PI was reported to occur at least a decade after the first documented episode of pancreatitis. 16 

Elizabeth H. Yen, M.D.

Exocrine pancreatic insufficiency (EPI) affects 85% of patients with cystic fibrosis. Untreated EPI during infancy and childhood will quickly manifest as malnutrition and failure to thrive. The most critical of the pancreatic enzymes are the lipases, responsible for the digestion of dietary fats. Salivary amylase and gastric pepsin make sizeable contributions to the digestion of carbohydrates and proteins, respectively, with intestinal brush border enzymes contributing to the final breakdown of these macronutrients. On the other hand, lingual and gastric lipases are minor contributors to fat digestion.

Pancreatic lipase hydrolyzes triglycerides to free fatty acids and glycerol, working at the oil-water interface of an oil droplet. Bile acids increase the surface area at which the lipase enzyme can function through emulsification of dietary fat, and colipase anchors lipase to the bile salts, facilitating efficient action of lipase at the oilwater interface. A failure in any of these mechanisms can lead to fat maldigestion and malabsorption. Therefore, management of EPI is targeted at improving intraluminal fat digestion.

Prior to 2010, pancreatic enzyme products (PEPs) had been available without U.S. Food and Drug Administration (FDA) approval. Problems arising from the variable filling of capsules with active enzymes eventually led to FDA action requiring that all PEPs undergo FDA approval process. Five PEPs are currently approved by the FDA in the U.S. (Creon, Zenpep, Pancreaze, Ultresa, Pertzye): they are all different formulations of the same active ingredient, pancrelipase, which is derived from porcine pancreas and composed of amylases, lipases and proteases. All approved PEPs are enteric coated to prevent inactivation of the enzymes by gastric acid, and they all dissolve at a pH ≥ 5.5. Dosing of PEPs is covered extensively in the product labels of these drugs, but salient points include: (1) dose administration at the start of a meal, (2) not chewing the capsule to prevent enzyme release in the mouth, and (3) for infants and young children, capsule contents should be mixed with an acidic food such as apple sauce and the mouth should be checked for residual beads to prevent oral ulcerations from retained enzymes. Doses exceeding 10,000 units of lipase/kg/day are associated with a higher risk of fibrosing colonopathy and should be avoided.

Troubleshooting persistent fat malabsorption: Persistent fat maldigestion and malabsorption can arise from a variety of causes. Maldigestion can result from noncompliance with PEPs, underdosing with high fat meals, prolonged acidic pH in the small intestine, and small intestinal bacterial overgrowth. Absorption of free fatty acids can be impaired by intestinal inflammation, and by the physical barrier of thick, dehydrated mucus. Additionally, lactase deficiency and Celiac disease can both alter intraluminal conditions that lead to decreased effectiveness of PEPs and malabsorption of nutrients.

Acidic intestinal pH: In CF, lack of CFTR function in pancreatic duct cells and duodenal epithelial cells results in reduced stimulated bicarbonate secretion after a meal (1, 2) . This delay of neutralization of duodenal chyme was recently demonstrated in humans (3) with cystic fibrosis. Interventional studies with gastric acid suppressing medications have demonstrated improvement in fat absorption when combined with PEPs (4). However, clinically relevant long term outcomes, such as improvements in nutritional status, lung function, quality of life, and survival have not been assessed (4) .

The CFTR null mouse model of CF develops SIBO, and shows a robust improvement in weight gain after treatment with metronidazole (5). However, fecal fat absorption is not improved after antibiotic treatment in the CFTR null mouse, indicating alternative mechanisms for weight gain after antibiotic use in this animal model (6) . One human study compared fecal fat absorption in four patients with CF before and after treatment with metronidazole for bacterial overgrowth and showed improvements post-treatment (7) . Oral ciprofloxacin in patients with SIBO also results in improved fecal fat absorption (8) . Larger, randomized control trials in humans are needed to determine the effectiveness of oral antibiotics to treat SIBO on fecal fat absorption and improvements in nutritional status.

Inflammation of the small intestine has been described in patients with cystic fibrosis (9) . Inflammation can damage the absorptive epithelia and alter its absorptive capacity. Both celiac disease and Crohn's disease have been reported in patients with CF, and should be considered in the patient with persistent fat malabsorption and abdominal complaints. Lactase deficiency is a much more common finding in general and can also present with signs of malabsorption. Undigested lactose is fermented by intestinal bacteria, leading to an acidic intestinal milieu that can interfere with PEP activity, leading to fat maldigestion and malabsorption. The best diagnostic modality for lactase deficiency in cystic fibrosis has not been determined, but the gold standard is analysis of brush border disaccharidase activity levels in duodenal biopsies.

Special considerations in liver disease: Patients with CF-associated liver disease (CFLD) have additional nutritional considerations with regard to management of EPI. As a cholestatic disease, the bile acid pool in CFLD is decreased. The decrease in the bile acids results in less effective solubilization of dietary fats, and negatively impacts long chain triglyceride hydrolysis and fat-soluble vitamin absorption. Special attention in these patients is necessary to ensure they do not become deficient in vitamins A, D, E and K, and high dose supplementation may be necessary. Additionally, patients with CFLD have higher caloric demands because of the underlying catabolic state imposed by the liver disease. They may require 1.5 to 2 times resting energy requirement to meet their metabolic needs. PEP dosing should be adjusted to account for the higher fat intake. If supplemental feeds are administered, a formula with a higher percentage of medium chain triglycerides is preferable (10).

Craig Hodges, Ph.D.

The median age of survival for cystic fibrosis has increased from about 2 years in the 1950s to over 37 years today due to newborn screening programs, specialized care and novel therapies. Despite these medical advancements, there is no current solution for the basic defect in CF which is the absence of functional CFTR. To this end, there is much research focused on correcting this underlying defect in CF by either replacing part or all of the gene with exogenous CFTR through gene therapy or correcting the CF patient's endogenous abnormal CFTR with pharmacological compounds [1, 2] . As CFTR correction therapies become possible, there are still many questions that need to be answered. Can spe-cific CF manifestations be ameliorated or even avoided by CFTR correction? When is the optimal timing for correction? Where does CFTR need to be corrected to avoid specific CF manifestations? How much correction is necessary to avoid or improve CF manifestations? How can the utility of new CFTR corrector compounds be assessed as they are identified? To answer these questions, the CF research community needs in vivo models of CFTR correction. In this talk, several CF mouse models will be presented that may serve this function and answer these questions which will help set expectations and goals for future human therapies involving CFTR correction.

Can we achieve CFTR correction and when is the optimal timing for correcting specific CF manifestations? For each clinical CF trait, one must determine if it is reversible upon CFTR correction, if decline can be halted, or slowed, or if the chain of events resulting from CFTR's absence are irreversible. These are important concepts, as currently we only have unaffected to compare with, and reverting a CF patient or mouse to a totally unaffected state is unlikely to be a realistic benchmark. In addition, understanding optimal time points for correcting each CF phenotype will be critical. For example, early postnatal intervention may be effective at preventing pulmonary disease in patients, but is unlikely to have much effect on exocrine pancreatic function, as most of that damage is done by birth. To investigate these questions about CFTR correction, we recently created an inducible model of CFTR correction in the mouse, in which the endogenous CFTR gene is nonfunctional but can be chemically induced to be functional leaving CFTR under its endogenous transcriptional control [3] . This CFTR correction can be completed at specific developmental timepoints throughout the whole mouse. We initially restored CFTR in adult CF mice to prove correction was possible using this model. When we restored CFTR in adult CF mice, we observed significant CFTR correction in all tissues examined displaying robust CFTR expression and/or function. However, since many CF phenotypes manifest very early in life we corrected CFTR in utero halfway through gestation. When we restored CFTR in CF fetuses we observed near normal weight gain, 100% survival to adulthood (compared to 40% without correction) and clear CFTR correction in all tissues tested. The effects of postnatal and juvenile correction are also being investigated.

Where is CFTR correction necessary? The correction of CFTR using our mouse model can also be restricted to specific tissues or cell types and thus we can ask the question of where is CFTR correction necessary to alleviate or avoid specific CF phenotypes. As an example, we used our mouse model to correct CFTR specifically in the intestinal epithelium. While restoration of CFTR in the intestinal epithelium eliminated the intestinal obstruction phenotype, the mice still displayed reduced growth [3] . These results suggest that CFTR correction in the intestinal epithelium is necessary for avoiding CF related intestinal obstruction but other sites of CFTR correction are necessary to correct growth reduction in CF. Since CFTR is widely expressed throughout the body, we can now begin to target where CFTR correction is necessary to improve or avoid manifestations.

How much CFTR correction is necessary? The identification of the minimal requirement of CFTR correction (i.e., percentage of cells with corrected CFTR) is necessary given that proposed treatments for CFTR cor-rection are unlikely to target 100% of cells. One study using CF airway epithelial cell cultures suggested that correction of CFTR in 6-10% of cells may be adequate to restore chloride transport [4] . A more recent in vitro study that utilized differentiated airway epithelial cultures observed that correction of CFTR in 25% of surface epithelial cells restored normal mucus transport rates and airway surface liquid volume, both factors in CF pulmonary disease [5] . The in vivo phenotypic consequence of the amount of CFTR correction in both of these studies is unknown, however both studies suggested that their results be verified if appropriate animal models existed. Until now, a controllable and titratable in vivo correction of CFTR did not exist. Since our inducible CFTR correction model uses a chemical that can be titrated, we are finally assessing how much CFTR correction is necessary to achieve normal CFTR function and what this means for CF manifestations. This aspect of CFTR correction may be very important if future therapeutic CFTR correctors can only achieve a small amount of CFTR correction.

How to assess utility of new CF corrector molecules? While the inducible mouse model of CFTR correction informs us about where, when and possibly how much CFTR correction is necessary, another critical question for the types of therapeutics currently under development is whether murine CFTR is an appropriate target on which to test them. Small molecule approaches to correct folding or increase activity may be quite sensitive to subtle structural differences between the two species, and stop codon suppression sensitive to sequence variation. Thus, we are generating mice carrying a genomic copy of human CFTR, and several of its common mutations, for testing these compounds. The feasibility of this approach has been established by others. Using a transgenic strategy, human CFTR cDNA has been shown to complement the loss of murine CFTR [6] and mice carrying a YAC (yeast artificial chromosome) containing human CFTR display complete complementation of CFTR deficiency [7] demonstrating species-conserved gene regulation. In addition, a BAC (bacterial artificial chromosome) containing the human genomic CFTR locus has been developed and shown to express faithfully in murine cells as well [8] . We are currently using this BAC to introduce human genomic CFTR into mouse embryos to establish a human CFTR expressing mouse line. Once generated, the transgenic line with "wildtype" CFTR will be crossed onto a CFTR-null background and assessed for complementation of murine CFTR deficiency. Ultimately, the mutant versions of this mouse carrying ∆F508, G551D and G542X of human CFTR will be made for the purposes of testing correctors for improving CFTR folding, activation or stop codon suppression, respectively.

By utilizing these novel mouse models of CFTR correction, we hope to determine which CF phenotypes are correctable as well as the appropriate target, opti-mal timing and amount needed for correction. In addition, with CFTR correctors showing significant promise for therapeutic potential, it is critical to characterize the effects of such correctors in in vivo scenarios. A better understanding of CFTR correction will provide much needed insight for the future treatment of CF patients.

References: 1. Kreindler JL. CF: exploiting its genetic basis in the hunt for new therapies. Pharmacol Ther. 2010 

Kalpaj Parekh, M.D.

Background: Lung transplantation is an established treatment modality for a variety of end-stage lung diseases. CF is an autosomal recessive disease leading to multisystem disease with significant morbidity and early death. Severe pulmonary disease is the primary cause of mortality in almost 80% of CF-related early deaths. Lung transplantation is commonly performed for patients with end-stage lung disease from CF with good outcomes and survival benefits. It is the most common indication for lung transplant in the pediatric population and the third most common indication in the adult population. Unfortunately, long-term survival following lung transplantation is still poor compared to that for other solid organ transplants. Within 5 years of receiving a lung transplant, nearly 50% of patients develop bronchiolitis obliterans syndrome (BOS) the most important cause of morbidity and mortality beyond one year following lung transplantation. There is no effective treatment available for OB at present and one of the major reasons is the lack of an animal model that reliably produces human OB pathology.

The ferret is an attractive model for our investigation of pulmonary transplant diseases. In contrast to the mouse lung, the ferret lung is very similar to that of human-with respect to both the cell biology of the airway cells, and the distribution and abundance of submucosal glands throughout the proximal airways. The ferret is also attractive as a model in the sense that both transgenic and knockout animals can be generated using techniques that were pioneered and optimized by the Engelhardt laboratory. We have developed an orthotopic left lung transplant model in the ferret that develops OB.

Orthotopic left lower lobe transplants were successfully performed in 22 outbred domestic ferrets in the absence of immunosuppression (IS) (n=5) and presence of varying IS treatment protocols (n=17). CT scans were performed to evaluate pulmonary parenchyma in the allograft and the IS dosage was modified based on the allograft assessment using the CT scan. At intervals between 3-6 months the allografts were examined histologically for evidence of acute/chronic rejection.

Results: In the absence of IS all allografts (n=5) underwent severe acute rejection with graft loss within 10-14 days. IS protected allografts from acute rejection and early graft loss observed in its absence. Reduction of IS dosage by 50% allowed development of controlled rejection. Allografts developed infiltrates on CT and classic histologic acute rejection and lymphocytic bronchiolitis.

Cycling of IS, to induce repeated episodes of controlled rejection, promoted classic histologic hallmarks of OB including fibrosis-associated occlusion of the bronchiolar airways in all allografts of long-term survivors. The OB lesions in the ferret model recapitulate all the characteristic histological features of the human OB lesions ( Figure 3 ).

We have developed an orthotopic lung transplant model in the ferret with documented longterm functional allograft survival. In the absence of IS, allografts develop acute rejection and lymphocytic bronchiolitis, similar to humans. Long-term survivors develop histologic changes in the allografts that are hallmarks of OB.

Leah Reznikov, Ph.D. 1 Nervous system abnormalities have been reported for decades in cystic fibrosis (CF). These include peripheral neuropathy (1), impaired gastric myoelectrical activity (2), sensorineural hearing loss (3), exaggerated pupillary, saliva and sweat secretion in response to cholinergic stimulation (4), aquagenic wrinkling of the palms (5) and reduced innervation of the sweat gland (6) . Reports that CFTR is expressed in the peripheral (7) and central nervous systems (8) raises the question of whether some nervous system abnormalities might be primary. However, the presence of disease has made it difficult to determine whether alterations in the nervous system are secondary. Evidence supporting the hypothesis that some nervous system abnormalities might be primary includes reports of altered nervous system function in people heterozygous for CFTR mutations, who do not have CF disease.

We utilized a porcine model of CF, which recapitulates many of the attributes of CF disease in humans (9) . We tested the hypothesis that CFTR is expressed in the nervous system, and that loss of CFTR directly affects nervous system function. We studied newborn CF pigs (CFTR-/-) less than 24 hours old to avoid the secondary consequences of disease.

Thus far our data demonstrate that CFTR is present in both the peripheral and central nervous systems, affects axon density, and nerve conduction velocity. Our results suggest that the nervous system is affected by loss of CFTR in the absence of chronic disease. Thus, it is likely that some of the altered nervous system functions in people with CF are primary. Furthermore, our data raise the question of whether altered nervous system function might contribute to the pathogenesis of CF. Animal models, including CF pigs, ferrets and mice will help to address this question.

References 

Michel Bagnat, Ph.D.

Positional cloning of the CFTR gene more than 20 years ago [1] opened a new era for CF research focused on the detailed study of the chloride channel it encodes. The development of specific CFTR activators, inhibitors and correctors has provided a wealth of molecular insights and even drugs that show very promising results in clinical trial [2] [3] [4] [5] . However, many aspects of CFTR biology and their relationship to CF remain unknown. Uncovering roles of CFTR during development and organogenesis will be crucial for understanding how the disease originates and affects the function of multiple organs including the liver and pancreas. To better understand how CFTR functions in organ development and physiology our lab uses the zebrafish model system [6] . We follow an approach that combines forward and reverse genetics to explore some important questions about CFTR biology: 1-What is the role of CFTR in embryonic development?; 2-How does loss of CFTR function lead to exocrine and endocrine pancreas defects?; 3-How is CFTR activity regulated in vivo?

To begin answering these questions we have generated zebrafish CFTR mutants and developed new tools to visualize early, tissue-specific roles of CFTR in the gastroin-testinal tract of live zebrafish. We have also carried out a forward genetic screen to identify new CFTR regulators.

To generate zebrafish mutants for CFTR we used recently developed TAL Effector Nucleases (TALENs) [7] and generated null and hypomorphic cftr alleles. Using these mutants we have uncovered a role for CFTRdependent fluid secretion in early patterning of the zebrafish embryo [8] . We are now using this phenotype to assay human CFTR mutants and drugs using a rescuebased experimental approach.

Pancreatic insufficiency is one of the most consistent and prevalent conditions associated with CF. It is thought to result from the obstruction of the ductal system leading to the destruction of acinar tissue. This hypothesis is very consistent with clinical and electrophysiological data from CF patients. But it does not exclude the possibility that an earlier function of CFTR during pancreas organogenesis may also play a significant role in triggering the disease, particularly in patients where the pancreas is already affected in utero. Using a variety of targeting approaches we have observed that loss of CFTR function in zebrafish leads to defects in the exocrine pancreas early during development, suggesting a role for CFTR in pancreas morphogenesis before the onset of tissue degeneration.

Loss of islet cells, including insulin producing beta cells, is a hallmark of CFRD. While there seems to be a causal relationship between loss of CFTR function and a reduction of ~50% in the number of islet cells [9] , the nature of this connection remains to be determined. Currently, the leading, yet untested, hypothesis is that the destruction of exocrine tissue somehow compromises islet cell survival and/or replenishment. However, it is also possible that CFTR is required for beta cell generation, function or maintenance. We hypothesize that a reduced number of beta cells early during pancreas morphogenesis may increase the physiological burden of the remaining beta cells, in turn contributing to the development of CFRD. Our results show that loss of CFTR function leads to significant reduction in insulin-producing cells at an early stage, suggesting that CFTR plays a role in endocrine pancreas morphogenesis. We are now using an shRNA-based knockdown approach to determine whether CFTR functions cell autonomously or nonautonomously in beta cells.

Previous work from our lab has established the usefulness of zebrafish genetics for uncovering new CFTR regulators [6] . To identify cellular and molecular processes regulating CFTR activity we have taken a forward genetics approach. We have recently completed a 3-generation genetic screen in zebrafish and identified several mutants that hyper-secrete fluid into the lumen of various organs. Our preliminary results suggest these mutants likely affect genes whose products regulate the activity of ion channels, including CFTR. Isolation of these mutations will help us better understand CFTR biology and may provide new molecular targets for CF therapy. Cystic fibrosis (CF), the most common inherited disease in the Caucasian population, is caused by mutations of the CF transmembrane conductance regulator (CFTR). CFTR, a cAMP-regulated anion channel, is confined to the apical plasma membrane (PM) and mediates transepithelial water and electrolyte transport. CFTR belongs to the ATP-binding cassette transporter family and comprises two membrane spanning domains (MSD1, MSD2) with four cytosolic loops (CL1-4) and three cytosolic domains; a regulatory (R) and two nucleotide-binding domains (NBD1, NBD2). The CFTR domain swapped architecture containing multiple domain-domain interfaces is indispensible in coupling the ATP-dependent conformational changes of the NBD1-2 dimer to channel gating [1] .

Newly synthesized CFTR is co-translationally N-glycosylated and undergoes both cotranslational domain folding and posttranslational, coupled multi-domain assembly in the ER, aided by a network of chaperones and co-chaperones [2] [3] [4] . The thermal instability of the NBD1 and its interfaces contribute to the limited folding efficiency of wild-type (wt) CFTR (20-40%) [5] . Upon traversing the Golgi complex, the native core-glycosylated CFTR undergoes complex-glycosylation, a posttranslational modification that indicates the channel biosynthetic processing.

Functional expression defect of CFTR leads to multiorgan pathology, including mucus dehydration, bacterial colonization and recurrent lung infection, the primary cause of morbidity and mortality in CF. Deletion of F508 (∆F508) in the NBD1, the most common CF mutation (~90%), causes global misfolding of the CFTR and degradation by the ER and cytosolic quality control mechanisms, comprising molecular chaperones and multiple ubiquitination machineries (e.g. CHIP, RMA1 and Gp78 E3-ubiquitin ligases) [6] [7] [8] . Ubiquitinated CFTR is retrotranslocated into cytosol for proteasome-mediated degradation, a process known as ER-associated degradation (ERAD). Although most of the ∆F508 CFTR molecules are eliminated by ERAD, a small amount could be detected at the PM by biochemical and functional assays in selected mouse and human tissues [9] [10] [11] . This "residual" ∆F508 CFTR activity can be augmented by exposing the mutant to reduced temperature, chemical chaperones (CC) or correctors by a variety of mechanisms [12] [13] [14] [15] . Although significant efforts have been devoted to the identification of small molecules as therapeutics to increase the ∆F508 CFTR folding efficiency and cell surface density ("correctors") or PM resident channel activity ("potentiators") [15] [16] [17] , we still lack compounds that restore the mutant folding efficiency >15% of its wt counterpart.

While VX-770, the most efficient potentiator that has undergone phase III clinical trial, restores the phosphorylated G551D CFTR and temperature rescued (r)∆F508 CFTR single channel activity via direct binding as a pharmacochaperone (PC) to wt level [16, 18] , "correctors" have modest efficiency and poorly understood mechanism. In principle, correctors may facilitate ∆F508 CFTR conformational maturation by alleviating the kinetic or thermodynamic folding defect and/or domaindomain interactions of the NBD1 by direct binding as a PC or indirectly, as a CC. Alternatively, a corrector may influence the mutant functional expression as proteosta-sis regulators (PRs) by modulating the cellular machinery of the ∆F508 CFTR PM folding, degradation and vesicular trafficking (e.g. HDAC inhibitor and HSF1inducers) [19, 20] .

Several correctors (e.g. corr-4a (C4), VRT-325 (C3) and VX-809) have been isolated and developed by functional high-throughput screenings (HTS) assays and medicinal chemistry [14, 17] . The most promising corrector VX-809 (3-{6-{[1-(2,2-difluoro-1,3-benzodioxol-5-yl)cyclopropanecarbonyl]amino}-3-methylpyridin-2-yl}benzoic acid) restores the ∆F508 CFTR PM expression ≤15% relative to wt CFTR in immortalized and primary respiratory epithelia [17] . Although the limited functional expression of ∆F508 CFTR at the PM was potentiated by the combination of VX-809 and VX-770 in model systems [17] and in CF patients [21] , significant improvement of ∆F508 CFTR folding efficiency is required to achieve wt-like correction. Limited understanding of corrector mechanism and the lack of high resolution CFTR architecture hamper the design of rationally designed, structure-based corrector selection and development.

Based on the recent recognition that ∆F508 mutation not only renders NBD1 energetically unstable [5, 22, 23] but also impairs its interdomain interactions, especially NBD1-MSD2 and coupled domain folding [3] [4] [5] 23 ], we designed a mechanism-based corrector pair selection strategy to isolate synergistically acting corrector molecules. The premise of this approach is that robust correction of the folding defect should eliminate the requirement of potentiator and enhance both ER processing efficiency and the PM stability of the mutant. We hypothesized that a) the modest effect of individual corrector could be attributed to restoration of only one of the ∆F508 CFTR structural defects. Thus second site suppressor mutation stabilizing e.g. the NBD1 may potentiate those correctors that preferentially restore the NBD1-MSD2 interface stability and vice versa. b) Combination of corrector acting on distinct CFTR structural defects similar to suppressor mutations may restore ∆F508 CFTR folding and function synergistically. To achieve our goal, first we selected PCs and CCs that preferentially correct either the NBD1 or the NBD1-MSD2 instability in ∆F508 CFTR. Next, we tested whether combination of correctors targeting complementary structural defects synergistically restores the ∆F508 CFTR ER folding, PM expression and function in non-polarized and respiratory epithelial cells. The results of our ongoing experiments suggest that combination of compounds, including the CFFTI corrector panel (kindly provided by Dr. R Bridges), preferentially targeting distinct structural defects have synergistic and robust rescue of ∆F508 CFTR folding defect in multiple model systems and primary epithelia, outlining a conceptually novel and potentially efficient therapeutic approach in CF. Support by NIH NIDDK, CFFTI, and CFC is acknowledged. Missense mutations of the cystic fibrosis transmembrane conductance regulator, CFTR, which prevent it from functioning as a regulated chloride channel, cause cystic fibrosis. By contrast to nonsense mutations or noncoding mutations that alter transcription or translation, alleles containing missense mutations produce fulllength mutant CFTR protein-albeit protein that fails to function due to a number of pathogenic mechanisms. Rehabilitating these mutant channels to function using a small molecule is a therapeutic approach that has been recently proven effective for the G551D mutation (1).

Chief among these pathogenic mechanisms is disruption of the instructions for folding into the native structure that mediates channel function. The majority of disease-causing missense mutations examined to date, including the predominant ∆F508 mutation, interfere with CFTR folding (2) (Figure) . A smaller number of disease-causing missense mutations form a stable, native structure, but fail to function due to alterations in either the mechanochemical gating cycle (eg. G551D), the ion conduction pathway, or other CFTR mediated activities (3). Thus, although CF is a monogenic disease, pharmacological approaches to treating CF will require distinct small molecules that target pathogenic mechanisms relevant to specific mutations. The feasibility of such an approach will depend on an ability to identify molecules that target these mechanisms and in determining which mechanisms are relevant to each disease-causing muta-tion. In this regard, ivacaftor potentiates not only the G551D mutant channel, but also other disease causing gating mutants (4) .

Mutations, such as ∆F508, that interfere with CFTR folding offer particular challenges due to the complexity of the process of structural maturation and the attendant transient involvement of a multitude of other cellular proteins. The final native structure of CFTR required for function is composed of multiple folded domains in both the membrane and the aqueous cytosol that interact with each other. Mutations could disrupt the hierarchical formation of this complex structure at multiple steps-integration into the membrane (5), co-translational domain folding (6), domain-domain association (7), interactions with other proteins, or some combination of these steps (7, 8) .

∆F508 offers a case in point. F508 resides in the first nucleotide binding domain of CFTR (NBD1) at a surface shown to interact with the fourth intracellular loop (ICL4) in the second transmembrane domain (TMD2). Deletion of this residue causes defective folding of NBD1 and defective interaction of the folded NBD1 with ICL4 (7, 8) . Whereas ∆F508 accounts for approximately 70% of all CF alleles and more than 90% of all patients have at least one ∆F508 allele, small molecules that could correct the folding defect could have therapeutic benefit for the majority of CF patients. Efforts to identify such correctors have been successful but have encountered an efficacy ceiling. This can be rationalized if the extant correctors act on one or the other defective steps but not both. In addition, compounds that act on one pathogenic step would be expected to act in synergy with compounds that act on the other. Methods have been developed to efficiently assess the impact of mutations on the NBD1 folding step and the coupled NBD1 folding-ICL4 association step (8) . These assays facilitate experiments that test if a corrector's mechanism of action is due to promotion of NBD1 folding, NBD1-ICL4 docking or neither. This mechanistic information would suggest compounds that might work in combination to overcome the efficacy ceiling. By contrast to ∆F508, other CF-causing missense mutations may lead to CFTR folding defects by interfering with one or the other of these steps but not both. As such, compounds known to act at the relevant step would be predicted to correct these mutants without an efficacy ceiling. These, and similar analyses, could facilitate efficient integration of knowledge of the specific molecular defects of mutations with the mechanism of actions of correctors, and, thus, evidence based coordination of molecular genetics and molecular pharmacology.

References The major mutation (deletion of phenylalanine at position 508 -F508del-CFTR) accounts for disease in approximately 90% of cystic fibrosis patients. This mutation causes the protein to be misprocessed and retained in the endoplasmic reticulum (ER) where it is subsequently degraded [1] . Low temperature conditions (27°C) enable partial rescue of the functional expression of F508del-CFTR protein on the cell surface [2] . However, upon restoration of F508del-CFTR expressing cells or tissues to normal body temperature (37°C), the mutant exhibits altered channel activity (low probability of being open) and instability with its rapid retrieval from the cell surface [3, 4] . Therefore, effective therapeutic compounds targeting this mutant will need to correct all three defects and 1) restore biosynthesis and 2) channel activity, and 3) prevent unfolding at the cell surface at normal body temperature. There has been considerable progress in identifying "corrector" compounds that exhibit partial efficacy in the functional expression of the major mutant on the surface of affected epithelia. However, the consequence of chemical intervention on each of the specific molecular defects discussed in the preceding paragraph remains unclear and in fact, it has been argued that even the best compounds lack efficacy in preventing the mutant protein from unfolding at the cell surface at 37°C [4] . Part of this uncertainty arises from the use of a diverse range of experimental assays on different cell types. We propose that the development of the next generation of "corrector" compounds would benefit from in vitro methods that can be used to predict the effect of new chemical compounds on preventing unfolding of the full length mutant CFTR protein using the purified protein itself as a discovery tool. This approach has the advantage of selecting compounds that bind directly to the CFTR protein itself to promote a stable conformation. Furthermore, our preliminary studies support the idea that there is a correlation between those small molecules that show superior efficacy in a novel in vitro assay of mutant protein stability and those compounds that show superior efficacy in promoting functional expression of F508del-CFTR in primary cultures of respiratory epithelia.

Our laboratory has developed methods for the purification and functional reconstitution of Wt-CFTR and F508del-CFTR protein [5] [6] [7] [8] [9] [10] . We have shown that the purified and reconstituted mutant protein exhibits regulated chloride channel activity and ATPase activity. In our previous work we employed assays of purified and reconstituted F508del-CFTR activity to determine whether small molecule corrector compounds bound directly to modify activity [5, 6] . We developed a new assay to monitor the propensity for unfolding and aggregation by the purified mutant. First, we showed that this assay reports the differential instability of the mutant protein as it exhibits a much faster rate of aggregation than the Wt protein. Then, using purified mutant protein, we found that the second generation corrector compound: C18 but not VRT-325 (a first generation compound) or its inactive analog stabilized the structure of the whole protein and prevented its aggregation.

Interestingly, the rank order of efficacy in these stability assays corresponds to the relative efficacy of each compound in promoting biosynthetic rescue in primary respiratory epithelial cells and the appearance of the mature, Band C form of the protein. These exciting preliminary data provide the first evidence that there may be a correlation between the ability of small molecules to bind the mutant protein directly to induce a conformational change, thereby protecting it from unfolding and aggregation.

This approach has the advantage of selecting compounds that bind directly to the CFTR protein itself to promote a stable conformation and will aid in determining the binding site for the most promising corrector compounds The knowledge generated through these studies will enable and assist in the rational design of more potent and specific compounds that may one day yield an effective treatment for 90% of CF patients. 

Douglas Cyr, Ph.D. Department of Cell & Developmental Biology, UNC at Chapel Hill, School of Medicine, Chapel Hill, NC, USA CF patients inherit a variety of mutations that cause folding defects in CFTR, which lead to its recognition for premature degradation by Hsp70 dependent E3 ubiquitin ligases (RMA1 and CHIP) on the cytoplasmic face of the ER. The folding defects in ∆F508-CFTR, as well as defects in other rare mutants, are correctable, but the mechanism of action for folding correctors is unknown. ∆F508 occurs in nucleotide binding domain 1 (NBD1) and blocks Clchannel assembly by hindering interactions of NBD1 with intracellular loops exposed by MSD1 and MSD2. Therefore, correction of folding defects in ∆F508-CFTR requires assembly to a conformation to permit escape from multiple ERQC machines and may require repair of more than one folding defect. This talk we describe current knowledge about the mechanism for recognition of misfolded CFTR by ERQC machines and present information on the mechanism for folding corrector action in treatment of CF. Prospects for treatment of ∆F508 homozygotes and compound heterozygotes with different combinations of folding correctors will be discussed. This work is supported by grants from the National Institutes of Health and the Cystic Fibrosis Foundation. Background: Delivery of drugs to the lungs is the mainstay of therapy for many lung diseases including cystic fibrosis (CF). Currently recommended inhalational therapies for CF patients include: antibiotics for eradication and chronic suppression of Pseudomonas aeruginosa, dornase alfa to decrease sputum viscosity, albuterol to prevent or treat bronchoconstriction, and 7% hypertonic saline to rehydrate the airway surface liquid compartment (1) . In addition to these medications, inhaled corticosteroids and long acting beta 2 agonist/inhaled corticosteroid combinations are frequently prescribed to this population. The benefit of topical inhalation therapy lies in the fact that higher in-situ drug concentrations coupled with lower systemic effects and faster onset of action are achieved while lower doses are given.

Aerosol Physiology: Several factors can affect the amount of drug that is deposited in the lungs of CF patients and can be divided into aerosol and patient factors.

Aerosol factors: Therapeutic aerosols generally encompass a range of different particle sizes and are characterized by several parameters including mass median aerodynamic diameter (MMAD), geometric standard deviation (GSD), and respirable fraction (2) . The MMAD is defined as the diameter at which 50% of the particles by mass are larger and 50% are smaller. The GSD is a measure of the spread of the particle size distribution. The respirable fraction is the percentage of aerosol comprised in the 1 µm to 5 µm range. Particles larger than 5 µm mainly deposit in the oropharynx and upper airways, while aerosols between 1 µm to 3 µm will deposit in the peripheral airways. Particles smaller than 1 µm tend to be exhaled. Particles deposit in the airways by different mechanisms. Those particles larger than 3 µm deposit by inertial impaction. This mechanism is also responsible for deposition in upper airways and large intrathoracic airways. Aerosols travelling at a slower speed will suffer less impaction and will enhance intrapulmonary deposition. Particles between 0.5 µm to 3 µm will mainly deposit by sedimentation due to gravity. This mechanism is responsible for deposition in the smaller airways. Finally, particles smaller than 0.5 µm will deposit by diffusion. Breath holding maneuvers facilitates the latter two mechanisms. Drug delivery becomes more complex because some aerosols retain moisture (hygroscopic property) therefore increasing their particle size and potentially altering the site of deposition. Solutions are more easily aerosolized than suspensions.

Patient factors: Several studies demonstrated that low inspiratory flows enhance intrapulmonary deposition and improve peripheral distribution. Large tidal volumes and low respiratory rates enhance intrapulmonary deposition. This provides the physiological basis why dose adjustments are not necessary in younger patients. Also infants and young children are nasal breathers adding a filtering site thus reducing intrapulmonary deposition even further. In addition to this, infants and young children may frequently cry and refuse wearing the mask, resulting in a significant decrease in lung deposition (3). Advanced lung disease will determine a more central and less homogeneous deposition compared to that of CF patients with minimal lung disease. Patients able to follow instructions and able to perform specific maneuvers could achieve a higher intrapulmonary deposition of inhaled drugs. Finally, the best prescribed drug does not work if the patient does not take it. This remains a significant challenge as noted by several studies showing low levels of adherence to prescribed inhalational therapies (4) .

Pitfalls of Using Current Delivery Systems: Several studies have documented how frequently mistakes are made when administering inhaled medications to infants and children. Frequent review of patient and/or caregiver knowledge regarding proper inhalation technique and device care during clinic visits is required due to poor knowledge/skills retention noted when no reinforcement is done.

Nebulizers, pressurized metered-dose inhalers (pMDI) with holding chambers, and dry powder inhalers are frequently used in CF patients. Many new drugs have been approved to be delivered by a specific drug/delivery device combination. Each type of delivery device has specific characteristics, different inhalation technique, different patient instructions and a different maintenance procedure.

Nebulizers: Jet nebulizers require an external gas supply to convert a liquid medication into an aerosol (5) . Continuously operated nebulizers release aerosol during inspiration and expiration. Breath enhanced nebulizers have a higher output during inspiration than during expiration. They require a tight seal between patient and interface to be able to take advantage of their design. Breath actuated nebulizers only release aerosol during inhalation and are not suitable for infants and young children. All these nebulizers use a compressor to operate them. The combination of nebulizer/compressor determines unique aerosol output characteristics. Lower gas flows and higher loading volumes lead to longer treatment times. Also higher gas flows produce smaller size aerosols with a higher respirable fraction (6) .

The amount of medication that remains in the nebulizer after no more aerosol is produced is known as residual or dead volume and can be as high as 1-2 mL for some devices.

While jet nebulizers cool the produced aerosols by 10-12°C, ultrasonic nebulizers do the opposite. This could potentially alter the structure of an inhaled drug (e.g., an inhaled protein such as dornase alfa).

More recently, vibrating mesh nebulizers have been introduced to the market either tied to a specific drug or not. This technology allows a faster and more efficient nebulization compared to jet nebulizers but could lead to overdosing if no dose adjustment is made. Proper care of the mesh is crucial to maintain efficiency of the device.

The concomitant use of nebulizers attached to add-on devices such as PEP devices could potentially decrease the particle size of the aerosol (7).

Cleaning and disinfection following current guidelines is important to avoid bacterial contamination of the devices (8) .

Although no CF-specific medication is available in this type of device, pMDIs are used to deliver corticosteroids, short and long acting bronchodilators, and combination therapy. The use of a valved-holding chamber is recommended to decrease oropharyngeal deposition and overcome actuationinhalation uncoordination (5) . Face masks are used to deliver aerosol in younger children. Efforts should be made to transition the interface to a mouthpiece when developmentally appropriate. Preferred characteristics of holding chambers include: made of antistatic materials, provide feedback regarding the inhalation maneuver and allow the use of the pMDI without removing the canister.

Dry Powder Inhalers: Although no CF specific medications are approved in the U.S., a dry powder inhaler for tobramycin has been approved in Europe. Current dry powder inhalers require the patient to inhale fast to disaggregate the drug and make it ready for inhalation (4). Younger children and subjects with lower lung function have lower inspiratory flows and volumes than older and healthier patients (9) . Evaluating the patient's ability to demonstrate adequate flow is recommended before prescribing these devices. A frequent mistake seen while operating DPIs is that patients either blow into the device or tilt it, therefore losing the dose.

Conclusions: Proper knowledge of aerosol physiology and the devices most commonly prescribed will allow the practitioner to decide on the appropriate modality. Patient education with frequent reinforcement should be incorporated into their clinic visits.

References Topical inhaled therapies are a cornerstone of lung health maintenance in CF. The advantages and disadvantages of various existing aerosol delivery systems were reviewed by Dr. Berlinski in the preceding article. Newer aerosol delivery systems have been developed to improve one or more of the perceived deficiencies of standard aerosol devices. These may include improved device efficiency, airway deposition and/or distribution, convenience, portability, or ease of use. Many new devices and drug formulations reduce the time burden for the patient and family, and electronic devices have the capability of monitoring adherence with therapy. Finally there may be purely commercial reasons for companies to develop new devices. Some of the devices discussed here may be relatively new in the U.S. market but have been available in Europe and other global sites.

Pressurized metered dose inhalers (pMDIs) are not new devices, but many features have been incorporated to improve their function. The conversion from CFC to HFA propellant is almost complete, which in most cases yielded a slower, warmer plume of aerosol. HFA-solution pMDIs (e.g.. beclomethasone, ciclesonide) were designed to yield aerosols with micron-sized particles, thus avoiding upper airway impaction and improving lung deposition and distribution. Dose counters are now incorporated into many pMDIs to indicate when the inhaler is empty, and the first breath-actuated HFA pMDI was recently approved in the U.S., which improves coordination of inhalation and device actuation. Velocitymodifying pMDIs have been developed to reduce the high initial speed of the spray. The Tempo Inhaler combines slow spray velocity with breath actuation to triple the aerosol deposition of a typical pMDI with half the dose-to-dose variability. So far the Tempo has only been combined with a systemic migraine drug, but has the potential to be used for respiratory drugs also. Finally, the slow-mist, propellant-free, multi-dose Respimat inhaler was just introduced in the US for the albuterolipratropium combination, and was used in Phase 3 trials with tiotropium in CF. This device still requires coordination of inhalation and dose-firing, but the mist is very slow and much easier to capture in the inspired airstream, resulting in better lung deposition. Slow-mist inhalers and pMDIs are only capable of delivering microgram drug quantities, so are currently useful mostly for bronchodilators and inhaled corticosteroids, not for CF drugs.

Spacers and valved holding chambers (VHCs) are attachments for inhalers to improve coordination, reduce upper airway dose, or to make inhalers available for use in very young children. These devices are also not new, but several improvements have been made recently. Most of the VHCs now incorporate antistatic materials to reduce drug loss in the chamber, visible flaps to signal that the patient is inhaling properly, low-resistance oneway valves, and improved mask designs to ensure a good seal for small children.

Dry-powder inhalers have been available for decades also, mostly for asthma drugs that have been milled and combined with lactose to improve de-aggregation of particles. Recent advances in dry powder formulations have allowed very high payloads of drug to be delivered via capsule-based DPI devices. Examples include colistimethate (125 mg milled powder per capsule, 1 capsule per dose, recently approved in the EU), mannitol (40 mg per capsule of spray-dried material, 10 capsules per dose, approved in Australia and the EU), and tobramycin inhalation powder (28 mg per capsule of hollow, porous, light particles [Pulmosphere technology], 4 capsules per dose, approved in Canada, EU and South American countries). DPIs are small, portable, breath-actuated, require no power source other than the inspiratory effort, reduce administration time and eliminate the need for refrigeration of medication and complex cleaning rituals compared to nebulizers. Cough is a common side effect of using DPIs that deliver a high drug payload. Children under 6 years of age often cannot use DPIs effectively, but there are new technologies that can actively disperse powder into an aerosol; these devices could be breathactuated or could generate aerosol into a holding chamber for passive inhalation using a mouthpiece or mask.

Many inhaled CF drugs require nebulization, and these devices were reviewed. Vibrating perforated-membrane nebulizers are faster than jet nebulizers, are portable, battery operated and have unique efficiencies of operation. For example, one of the devices from the eFlow platform (Altera) was developed specifically for the nebulization of aztreonam; it has a very low residual volume and can deliver a dose in 2-3 minutes. Another eFlow device (Trio) has been available in the US for a few years, and is similar to the Altera (low dead volume, high output, 2-4 times the delivery efficiency of jet nebulizers). Another eFlow version (Rapid) has been available in the EU and just received approval in the US. The eFlow Rapid is designed to mimic the aerosol characteristics of the PARI LC PLUS jet nebulizer with the exception of a faster delivery time. Bench studies suggest that the delivery of a unit dose of medication would be similar between an LC PLUS and eFlow Rapid, but higher with the more efficient Trio nebulizer. Thus, one must be cautious about substitution of nebulizers for certain drugs that may have toxicity at higher delivered doses. Vibrating mesh devices address convenience, portability, and time burden, but they still require cleaning and disinfection. In the future these electronic devices can be modified to monitor adherence.

Finally, it is well known that if aerosol is inhaled at a very slow rate, deposition in the upper airway decreases, and lower airway deposition and homogeneity of distribution improves. As a bonus, the variability of deposition is reduced (better targeting). Two device technologies help control and slow the inspiratory flow of the patient. The I-neb device uses a vibrating horn and mesh to generate aerosol, monitors the breathing pattern of the patient and aerosolizes only during inspiration (Adaptive Aerosol Delivery, AAD). It can operate in a tidal breathing mode, but its "targeted inhalation mode" (TIM) utilizes a high-resistance mouthpiece and tactile indicators that coach the patient to take slow, deep inhalations. The TIM mode has the benefits of higher deposition efficiency and lower time to administer a dose. A chip in the device also can monitor patient adherence. The I-neb is currently available for the pulmonary hypertension drug iloprost, and also in the UK and a few other countries for colistimethate. The Akita is another device that allows individualized controlled inhalations in combination with either a jet nebulizer or a vibrating mesh. Information about each drug is stored on a "smart card," the device can be programmed to deliver drug centrally or distally in the lung, and adherence can also be monitored.

Recently it was shown that by switching patients using dornase alfa from a jet nebulizer to the Akita mesh device, the peripheral airway function (FEF 75 ) improved, especially in the group that used the device programmed for distal airway delivery. The Akita is currently used in Germany and a few other EU countries. These electronic technologies are relatively expensive, but the improved deposition and homogeneity of distribution (even in partially obstructed airways) may be very important in the development of future expensive therapies, as well as improving the outcomes with older drugs.

New technologies are addressing many of the issues that plagued older aerosol delivery systems. New drug formulations in development can be partnered with new devices to improve convenience, adherence and outcomes. Using newer efficient systems with older drugs may require clinical studies to define proper doses in situations in which there is a dose response for efficacy or toxicity.

References Aerosol formulation properties influence aerosol performance, and hence, therapeutic efficacy. Several factors and formulation issues influence aerosol dispersion performance and pulmonary drug delivery (1) (2) (3) (4) (5) (6) (7) (8) . The solid-state properties of inhalable nanoparticles/microparticles (7, (9) (10) (11) (12) for dry powder inhalers (DPIs) can influence the aerosol dispersion performance (3, 13, 14) . Inhaled antibiotic, antiviral, antifungal, and immunosuppressant aerosols for targeted pulmonary delivery have demonstrated therapeutic advantages in targeting the airways (15) (16) (17) (18) (19) .

For the past several years, the number of FDAapproved DPI products for human use have increased significantly, a trend that is expected to continue for several reasons including increased incidence of several chronic pulmonary diseases, an aging global population, patient preference over other routes, improved patient quality-of-life, and a favorable environmental impact (regarded as "green pharmaceutical products"). Additional advantages of DPIs include fast onset of drug action and minimal side effects due to targeted delivery to the respiratory airways, no drug-food interactions, no hepatic metabolism, and increased physicochemical stability relative to liquid aerosols due to slower degradation chemical kinetics inherent in the solid-state. Currently, DPI inhalers comprise nearly approximately 70% of the approved inhaler products on the market. Indeed, respiratory inhalers are the largest component of the $44.2 billion human respiratory market currently and is expected to continue to increase in the decades ahead.

We have developed and published a robust powder deaggregation aerosol dispersion model to quantitatively predict therapeutic aerosol dispersion performance, allowing for quantitative optimization of aerosol formulations (14, (20) (21) (22) (23) (24) . In aerosol dispersion performance (i.e., inhaled particle delivery to the lung), prediction of fine particle fraction (FPF) generated under certain welldefined aerodynamic airflow parameters enables aerosol formulation optimization. In vitro aerosol performance characterization testing of aerosol formulations under pediatric and adult airflow rate conditions using inertial impaction methods are defined by the United States Pharmacopeia (USP) Aerosol Chapter <601> (25) . The FPF and emitted dose (ED) are calculated, since FPF and emitted dose reflect the mechanisms of aerosolization (fluidization and deaggregation) and also provide insight into predictive deposition modeling in the lungs. The plots of particle distribution among the impactor stages are plotted, which gives a direct and quantitative model of how aerosol particles will distribute across the whole lung region, i.e., predictive in vitro lung deposition modeling into specific lung regions in the context of aerodynamic sizing analysis. 

A. Bruce Montgomery, M.D. Cardeas Pharmaceuticals, Seattle, WA, USA Regulatory guidance available in the United States and Europe give conflicting and often outdated advice in regard to clinical trial design for Phase 3 trials for inhaled antibiotics in CF. For instance, the availability of two approved drugs makes the conduct of placebocontrolled trials longer than 28 days problematic. Comparator drugs require labeled use, making comparator trials difficult to blind as taste, foaming, regimen, device and delivery time differences are present. Currently, there is no consensus on the most appropriate endpoints for evaluation of aerosolized antimicrobials. FEV1 is a surrogate endpoint that it is a predictor of mortality -it is standardized, reproducible, noninvasive, simple, and inexpensive to perform but small statistically significant changes may not be clinically important. FEV1 improvement also has a ceiling effect in patients with mild lung function impairment and spirometry cannot be reliably done in patients under the age of six years. A patient reported outcome is a promising clinical endpoint but there is not an accepted tool that can be used as a primary endpoint for the FDA or the EMA, although the latter recognizes the CFQ-R as a validated secondary endpoint and the FDA grandfathered acceptance of the CFQ-R respiratory domain in the pivotal aztreonam for inhalation study. Exacerbations are an important clinical endpoint that reflects morbidity and is a major driver of cost of care, but exacerbations occur infrequently and a standardized definition has not been reached. Furthermore, an exacerbation endpoint may fail even with an otherwise effective antibiotic. Regulatory authorities will have a difficult time approving any new inhaled antibiotic based on one clinical endpoint alone.

The FDA is now requiring aerosol antibiotics to be a combination drug device approval, changing from the historical practice of a separate device approval. Device approvals have traditionally been by the 510K process that requires substantial equivalence to a predicate device. New devices have been required, if to be included on the drug label, to have clinical trials to prove efficacy and dose. Due to differences in efficiency, particle size, these trials are complex and expensive and seldom, if ever, done. Although many have advocated in vitro methods to have alternative nebulizers approved for inhaled antibiotics, no method has been validated nor accepted by the FDA. In the future, with all new approvals a drug device combination, no one should have an expectation of an in vitro method being accepted.

References The airways of cystic fibrosis (CF) patients are nearly always infected with a large number of microorganisms, and in particular lung infections with P. aeruginosa create severe problems of morbidity and mortality for the patients. Intense antibiotic treatments have prolonged the lives of CF patients, but development of resistance to the drugs limits to varying extent the usefulness of these. Increased knowledge of the bacterial infection routes may provide new therapeutic strategies by disclosing new targets for interference, and specific bacterial adaptive mutations may be useful as genetic markers for chronic states of infections. It is thus the thesis of our research that a detailed molecular and genetic mapping of the evolutionary trajectories of P. aeruginosa during the course of long-term infection in CF airways may pave the way for early diagnosis of chronic infection states and hence for improved treatment regimes.

P. aeruginosa is a versatile environmental organism with a broad metabolic repertoire, which is useful -both in relation to changing environmental conditions and in connection with infections of plants, fungi and animals/humans. In contrast to other fluorescent pseudomonads P. aeruginosa is able to grow at high temperatures (up to 42°C) and has its maximum at 37°C. In addition, its genome encodes a number of virulence factors making it a model example of an opportunistic pathogen. The genome of P. aeruginosa encodes around 6000 genes, of which about 600 have been estimated to exert regulatory functions (1) . It is assumed that this very large regulatory repertoire plays a role during the early stages of infection in CF airways, when switching from an environmental to a host colonization life-style. Prolonged colonization and a switch to de facto chronic infection, however, seems to require genetic adaptation through mutations, which fix the life-style changes of the bacteria. One example of this is associated with the biofilm state of growth of P. aeruginosa, which is connected to the overproduction of alginate (mucoidy). CF patients chronically infected with P. aeruginosa most often display mucoid colonies in their lung expectorates, and in nearly all cases the genetic background for this phenotype includes mutations in the mucA gene encoding an anti-sigma factor, which interferes with one of the main stress regulators AlgU (2) . Thus mutations in mucA can be considered examples of genetic markers, which may be indicative of shifts towards chronic infection.

We have in recent years conducted a comprehensive study of the genomic and phenotypic dynamics of infecting P. aeruginosa lineages in a number of CF patients associated with the Copenhagen CF clinic in Denmark. Since 1973 bacterial isolates from lung expectorates of CF patients have been collected and stored in the clinic, and this strain collection represents bacterial evolution and adaptation covering up to 200,000 bacterial generations of growth in CF airways (3). Our investigations have so far focused on two specific lineages of P. aeruginosa, both of which have been transmitted among several patients connected to the clinic (4). Our aim has been to uncover which genetic changes are the most likely determinants of the apparent dominance of these clones, which have been able to out-compete efficiently a large number of other lineages of P. aeruginosa in the infected patients.

The approach we have taken has been to combine fullgenome sequencing of longitudinal isolates of P. aeruginosa from a number of chronically infected patients with global gene expression analysis and other types of phenotypic characterization. Interestingly, our data show that the two dominant and transmissible lineages in the Copenhagen CF Clinic have followed very different evolutionary trajectories, suggesting that there could be several routes towards persistence of colonization of the CF airways. One of these has involved accumulation of mutations in the genome with a low rate -0.7 x 10 -10 base substitutions per base pair per generation -and most of these are not adaptive but the results of genetic drift. However, a small number of early mutations in global regulatory genes seem to be highly important for the coloization success of this clone (3), and we are currently investigating the combined impacts of these mutations on the overall bacterial phenotype. The common evolutionary profile of the lineage based on analysis of a large number of patients is one of relative phenotypic and genetic homogeneity across the patients, suggesting that the early adaptive regulatory mutations found in all isolates from all infected patients resulted in optimal fitness followed by random neutral mutations with no significant phenotypic consequences.

It is well known that P. aeruginosa changes its phenotype as well as its genotype in the course of chronic infection of CF airways. Our studies suggest that effi-cient adaptation to the human airways may in some cases take place as a consequence of mutations in genes controlling expression of large numbers of genes. It is therefore interesting to search for similar mutations, and for mutations in other global regulator genes, in isolates of P. aeruginosa collected from other patients and from other clinics in order to determine which of these could be considered as genetic markers for chronic infections.

References As lung disease progresses in CF patients, injury to the lung typically exhibits heterogeneity, with the apical portions of the lung becoming more injured than the basilar portions. The etiology for this disease variability is unknown. Understanding the etiology of heterogeneous disease within a CF lung may provide clues to factors that lead to either mild or severe disease in CF patients. The most obvious differences in disease severity occur in CF patients with chronic airway infection, which suggests that bacteria may cause the differential injury in upper and lower lung regions.

Our lab and others have observed that infecting P. aeruginosa strains can evolve into a diverse population of genetic variants within the CF lung. We hypothesized that the genetic variants of P. aeruginosa within the highly injured areas of CF lungs may have more injury potential. To test this hypothesis, we obtained secretions via sterile technique from the major lobar regions of explanted CF lungs. From these secretions, we selected several hundred isolates of P. aeruginosa and confirmed the isolates were derived from a single parent strain by DNA fingerprinting. We then subjected these isolates to a battery of phenotypic tests that categorized isolates into distinct subpopulations. To our surprise, the P. aeruginosa population composition varied tremendously between major anatomic regions of the five CF lungs studied to date. In fact, the most striking observation was that a subpopulation that was the most abundant in one region was low in abundance or absent in other regions.

To identify bacterial functions that contribute to variable disease severity, we performed whole genome sequencing of the most prevalent P. aeruginosa subpopulations infecting anatomic regions of one lung. We found that isolates from mild and severely diseased areas exhibited variants within genes affecting motility, type III secretion, antibiotic resistance and regulators of other pathogenesis functions. P. aeruginosa isolates from the upper regions of the lung were more virulent during acute infection of mouse lungs than isolates from the lower regions. Ongoing experiments are establishing the contribution of genetic variants to the virulence functions in isolates from regions with severe and mild disease. If consistent among several patients, bacterial functions found in severely diseased lung regions may serve as targets for future therapies. Despite the fact that cystic fibrosis (CF) is a genetic disorder, chronic lung infection with the bacterial pathogens Pseudomonas aeruginosa and, less commonly, members of the Burkholderia cepacia complex (BCC), accounts for most of the morbidity and mortality associated with CF (1). BCC infection in CF patients can be transmitted person-to-person, and multiple outbreaks have been described, including one of a highly antibiotic resistant strain of the BCC species Burkholderia dolosa among almost 40 CF patients at a single center in Boston (2) . This outbreak has been associated with accelerated decline of lung function and decreased survival (3).

In collaboration with Roy Kishony's group at Harvard Medical School and Alex McAdam of the Boston Children's Hospital clinical microbiology lab, we recently investigated the genomic diversity of 112 B. dolosa clinical isolates recovered from 14 people with CF from the Boston outbreak over almost two decades (4). The premise of our study was to identify evolutionary patterns that occur recurrently in multiple people infected by the same strain during an epidemic. Applying whole-genome Illumina sequencing and systematic phenotyping of these bacterial isolates, we gained key insights on the disease at three main levels:

1. Epidemic level: A putative person-to-person infection network was uncovered. Accumulation of mutations in the pathogen revealed the specific person-to-person transmission network (who infected whom).

2. Patient level: Organ-to-organ transmission is multiclonal. Isolates recovered consecutively from the blood during bacteremia were not descendants of each other; rather they descended from distinct lung isolates, rejecting the possibility of a clonal transmission from the lung to the blood.

3. Gene level: Adaptive mutations arising independently in different patients identify genes under strong positive selection. These genes illuminated known pathogenic phenotypes, and also implicated novel pathways in pathogenesis.

Notably, we found that a subset of 17 genes having 3 or more point mutations actually received 18 times as many nonsynonymous (protein-altering) mutations as expected by neutral drift and were thus under strong positive selection (dN/dS=18, CI:4.9-152.7). Eleven of the 17 genes encode proteins that belong to functional categories related to pathogenicity. The other six genes have not previously been implicated in the pathogenesis of lung infections.

One of these genes, which had 17 non-synonymous mutations, has homology to FixL of the FixL/FixJ twocomponent system of Rhizobium and Caulobacter. FixL is a sensory histidine kinase that detects oxygen tension and phosphorylates the transcription factor FixJ under low oxygen conditions, leading to differing downstream effects in different species (5) . B. dolosa lacks the genes necessary for nitrogen fixation, suggesting that the fix system is utilized for gene-regulation of a different nature, as has been seen in other species (5) . All of the 24 mutations seen in this pathway were nonsynonymous, and none resulted in a stop codon, suggesting a fine tuning of the pathway rather than its interruption. These results may lead to new therapies of BCC infection in CF patients.

Another pathway under strong positive selection involved O antigen repeats in the lipopolysaccharide (LPS) of the bacterial outer membrane, known for its important role in virulence in related bacterial species. We observed that some of our isolates possessed O antigen, while others did not (i.e., had a rough LPS). Remarkably, a single nucleotide in the glycosyltransferase gene BDAG_02317 correlated exactly with the presence of O antigen repeats. The ancestral genotype encodes a stop codon at this locus, corresponding to the absence of O-antigen repeats. We confirmed this association by showing that complementation with the full-length BDAG_02317 glycosyltransferase gene could restore O antigen presentation in an LPS rough strain. We also found that the last common ancestor of strains from nearly every patient possessed the truncated genotype and that a gain-of-function mutation occurred independently in 9 subjects. These results suggest that a rough LPS may provide an advantage during patient-to-patient transmission during outbreaks among CF patients and illuminate potential new avenues for the development of therapies that might disrupt the transmission of CF bacterial pathogens (4) Progress in understanding exacerbations has been hindered by several factors. First, tractable animal models that replicate chronic bacterial lung infections and disease flares do not yet exist. In their absence, the only way to learn about flare mechanisms is directly from patients and patient-derived samples. Second, primary mechanisms that mediate exacerbations likely produce secondary bacterial and host responses that can confound etiologic understanding. Third, identification of bacterial factors responsible for exacerbations is difficult since bacterial gene and protein expression is highly dependent on environmental conditions, and methods that can be used to model exacerbationassociated changes in bacterial functioning are lacking.

Most CF infections are clonal, meaning that the infecting population is composed of the progeny of a single isolate. Recent findings have illustrated genetic diversity of infecting PA populations within a single patient as well as within differing airway regions of a single patient. Whole genome sequence analysis has shown that infecting PA strains evolve to produce genetically-diverse (but clonally-related) bacterial populations within the host (1). Furthermore, evolved PA subpopulations differ in traits known to affect host inflammatory and injury responses. These include the expression of toxins, pathogen-associated molecular patterns like flagella, and cytotoxic activity. The presence of subpopulations that elicit markedly different host responses could have a profound effect on disease manifestations, and emerging data suggests that flares are associated with marked changes in infecting population composition.

Analogous to the genome, the proteome is the complement of protein levels present in an organism under a given set of conditions. Unlike the genome, the proteome is dynamic and is altered by changes in gene expression, posttranslational modification, protein localization, or other factors to give rise to altered functional levels that enable response to changes in environmental conditions. Our current CF proteomics efforts focus on two unanswered questions:

1) How do exacerbation-associated PA subpopulations differ in protein expression 2) Which PA proteins likely mediate exacerbationassociated host responses in vivo?

These questions have been notoriously difficult to answer due to the difficulty measuring the PA proteome in patient-derived samples such as sputum. To overcome these challenges, we have developed a multi-faceted proteomic approach that includes proteome analysis on dominant PA isolates collected under variable conditions and multiple reaction monitoring (MRM) analysis of whole digested sputum or bacteria enriched from patient sputa.

Isolates that increase in abundance at the onset of flares exhibit marked proteome differences. To understand how composition shifts affect PA population functioning, we compared the proteomes of randomlyselected isolates from subpopulations that were dominant when the patient was "well," dominant at onset of exacerbation ("sick"), and dominant after the patient was on antibiotics ("treatment"). Isolates (12 from each) were independently cultured and then pooled, lysed, and digested with trypsin. Equal aliquots of proteome samples were analyzed in triplicate using reversed-phase capillary liquid chromatography and data-dependent MS/MS using a Velos-FT high performance mass spectrometer. Data were analyzed using MaxQuant (2) with the PA proteome database. A total of 5928 peptides corresponding to 1159 proteins were identified, and 497 proteins showed statisticallyrelevant abundance changes (ANOVA) in the exacerbation or antibiotic treatment isolates relative to those from the well period. These data demonstrate for the first time that P. aeruginosa subclasses dominant during different stages of infection vary greatly at the proteome level. These data also illustrate excellent reproducibility among all technical replicates which is important for such label free quantitation schemes. Also illustrated are the levels of several proteins previously shown to affect aminoglycoside efficacy (3). Excitingly, these proteomics data show that the proteins amgR, amgS, mexY and several others increase in isolates dominant at the onset of exacerbation, suggesting that the PA population in patient airways could exhibit decreased sensitivity to aminoglycosides at exacerbation onset. More importantly, these results illustrate that the proteome of dominant isolates col-lected during different stages of infection can be used to evaluate conditions experienced by the bacteria. This approach can allow new bacterial proteins to be identified that could help better identify the onset of exacerbation conditions, help better understand the mechanisms involved in exacerbation and possibly serve as new therapeutic targets for treatment to further decrease loss of CF patient lung function during exacerbation.

References 

Richard Cohen, M.D.

Because CFRD is often silent and is related to adverse nutritional and pulmonary outcomes, screening is essential. And yet, adherence to published guidelines for CFRD screening at the center level is often cumbersome and suboptimal. This article will review these guidelines, and address barriers to the practical aspects of implementation.

The screening section of the "Clinical Care Guidelines for CF-Related Diabetes" (1) document delineates the following bullet points:

1. "Use of A1C as a screening test for CFRD is not recommended. (ADA-B; USPSTF-D)" This recommendation has basis in a number of studies citing insufficient sensitivity of the A1C as the only screening modality. Sensitivities range from 23% to 100% .Higher sensitivity is achieved at the cost of poorer specificity. And yet, most CF providers still screen with A1C based on the most current ADA screening guidelines (2) with particular reference to Type 2 diabetes, and then exclude the diagnosis of CFRD based on a one-time random test.

2. "Screening for CFRD should be performed using the 2 hour 75 gram OGTT. (ADA-E; Consensus)" In many centers, the practicality of this modality is limited. Alternatives to this algorithm are not infrequent discussion topics.

3. "Annual screening for CFRD by age 10. (ADA-B; USPSTF-B)" 4. "Screening of CF patients during illness (ADA-E; Consensus)". The guidelines suggest an algorithm whereby glucose levels via self-monitoring of blood glucose (SMBG) are monitored for the first 48 hours of an illness, preferably as fasting and two hour postprandial values, either in the hospital setting or at home. Confirmation of SMBG levels by laboratory plasma glucose levels is an imperative, yet implementation of this modality is often very difficult, particularly at home. 5. "Screening for CFRD by measuring mid-and immediate post-feeding plasma glucose levels is recommended for CF patients on continuous enteral feedings, at the time of gastrostomy tube feeding initiation and then monthly at home. Elevated glucose levels detected by SMBG must be confirmed by a certified laboratory. (ADA-E; Consensus)" Doing this properly is an effort to obtain meaningful data for the fine tuning of insulin administration and the subsequent optimization of nutritional repletion parameters.

6. "Women with CF who are planning a pregnancy or confirmed pregnant should be screened for pre-existing CFRD with a 2-h 75-g OGTT if they have not had a normal CFRD screen in the last 6 months. (ADA-E; Consensus)" The guidelines go on to clarify the specifics of this testing during the pregnancy and in the weeks following the pregnancy, consistent with American College of Obstetrics and Gynecology (ACOG) guidelines. 7 . "Pre and post transplant screening." 8. "Screening for chronic complications (ADA-E; Consensus)". This is a reminder that the ADA guidelines necessitate periodic eye exams, regular blood pressure measurements, urine screening for microalbuminuria, and lipid screens in pancreatic sufficient patients with CFRD.

To address implementation of the CFRD guidelines, a year-long quality improvement learning collaborative of cystic fibrosis and endocrine professionals from six CF Foundation (CFF) accredited care centers convened in 2010. This collaborative was sponsored by the CFF and the Pediatric Endocrine Society with the goal of improving quality in screening, diagnosis and care delivery. Participating teams generated CFRD-specific change ideas relevant to these goals and compiled their efforts into a compendium document for dissemination to all accredited care programs: "CFRD Evidence Based Practice and Smart Change Idea Compendium, 2012" (3). The compendium is available on Port CF under "resources."

Yet this still leaves most CF care centers with the cumbersome and labor intensive practicalities of implementing a "process map" for CFRD diagnostic screening, the details of which vary greatly from program to program. Problems lie in the following areas: patient coordinator reviews, location of testing, insurance authorization issues, inconsistent laboratory practices, complacency in following through or acknowledging the need for yearly testing, and patient and family preferences.

To address those barriers, various programs have enacted screening alternatives such as use of liquid or solid "glucola alternatives"(4), creation of CFRD care delivery microsystems within the care center, and generation of RD or RN management protocols. Additional intriguing alternatives include reference to the diurnal variation of pancreatic beta cell function raising the question of implementing the testing later in the day to maximize the yield of positive tests, mixed meal tolerance testing with reference to animal model research, nonfasting screening algorithms with SMBG testing in the clinic visit and administration of glucola on the spot, continuous glucose monitoring (CGM) in lieu of a spot OGTT (5) , and a one-hour plasma glucose value obtained during the OGTT as a more sensitive screening indicator (6) .

The most recent systematic review on CFRD screening was a product of the Department of Public Health, Aberdeen, UK (7) wherein multiple databases were accessed. The conclusions of this systematic review are as follows: screening for CFRD is justified, but the screening for lesser degrees of hyperglycemia is less strong. The authors do surmise that what they refer to as a "pulmonopathy" may be a manifestation of high glucose in pulmonary secretions thus promoting microbial colonization at levels well below the diabetes level.

This has implications for testing and threshold for intervention. The authors suggest obtaining a 1 hour blood glucose level in addition to the 2 hour level (or looking at continuous glucose monitoring), because there may be a temporary high glucose level at 1 hour that normalizes at 2 hours and could be missed by standard testing. Insulin therapy may be clinically useful for these patients, but more research is needed.

References (2), but with only limited data available in the CF population, the 1998 CFF guidelines recommended treatment of CFRD with fasting hyperglycemia and treatment of symptomatic CFRD without fasting hyperglycemia. Fortunately, the CFRD literature was growing: CFRD was associated with a 6-fold greater mortality (3) and with worse clinical status as indicated by poorer pulmonary function, increased frequency of acute pulmonary exacerbations, and poorer nutritional status.

Additionally, small studies suggested insulin therapy helps improve or stabilize pulmonary function and augments nutritional status in CFRD (4, 5) . Ultimately, a randomized control trial found insulin treatment reversed chronic decline in BMI in subjects with CFRD without fasting hyperglycemia (6) and suggested that earlier CFRD intervention may have long-term benefit. The 2010 guidelines for diagnosis and treatment of CFRD reflect these findings (7): CFRD with and without fasting hyperglycemia are no longer differentially treated-both require insulin therapy.

Associations of Dysglycemia and Prediabetes with CF-Relevant Outcomes: The next questions to resolve thus arise: determine the extent to which early glucose abnormalities are associated with CF-relevant outcomes (e.g., nutritional status and pulmonary function) and the extent to which treatment of earlier glucose disturbances benefits CF. Accordingly, the 2010 CFRD guidelines recognized a new category of dysglycemia referred to as Indeterminate Glucose Tolerance (BG1>200mg/dL).

These new guidelines acknowledge few patients have completely normal glucose tolerance. Early glucose changes involve variable postprandial hyperglycemia: 30-, 60-, and 90-minute BG during an OGTT are frequently elevated (8) and are suspected to provide a more sensitive method for detecting glucose intolerance. Moreover, delayed and blunted insulin secretion is found during OGTT even in patients without CFRD (9) . Similarly, delayed and reduced C-peptide responses have been observed during OGTT, with NGT (10,11). Abnormalities are more pronounced with worsening glycemic status (9, 11, 12) . Likewise, intravenous challenge studies have revealed impaired first-phase insulin and C-peptide secretion in response to glucose and other stimulatory agents in CF (11) .

Thus, glucose abnormalities and insulin secretory defects are common but to what extent are they associated with CF-relevant outcomes? Deteriorating clinical status has been observed in patients with CF with normal OGTT but elevated casual BG (13) . Over a 4-year period pulmonary function decline was greater in CF patients with IGT vs NGT and worse insulin secretion at baseline was also associated with greater decline (14) . In the 4-6 years preceding the diagnosis of CFRD or NGT by OGTT, BMI and pulmonary function declined to a greater extent in the CFRD group than the NGT group (15, 16) . Brodsky et al. examined the relationship between BG obtained during a screening OGTT and FEV1% in a pediatric CF population (17) . Not only was elevated BG common in the setting of normal 2-hr BG (BG2), the 1-hr BG (BG1) was negatively associated with FEV1%-predicted after adjustment for age and BMI. Moreover, no relationship between the BG2 and FEV1% was found in this crosssectional study. The findings highlight the potential implications of even subtle glucose abnormalities and early insulin deficiency in CF. In contrast, in a retrospective study of 94 children age 6-9 years, Ode et al. found no difference between nutritional status or pulmonary function in the subsequent five years in subjects with abnormal glucose tolerance (IGT or Indeterminate) vs NGT. The former were at greater risk of developing CFRD (18) .

Treatment of Early Glucose Abnormalities and CF Relevant Outcomes: Studies examining treatment of early glucose abnormalities are now becoming available. Preliminary data from a 12-month trial of daily glargine insulin (@0.2 units/kg) in 14 subjects with either IGT or glucose > 140 mg/dL during continuous glucose monitoring found improvements in FEV1%predicted (19) ; baseline data in the overall cohort of 22 subjects (including 8 subjects with CFRD), mean age 12.5 years, revealed FEV1%-predicted =68.2 ± 6.2 (min/max: 24.0/117.0), which may be a sicker population. In six slightly older subjects with IGT, a trial of daily glargine insulin (@0.3 units/kg/day) was associated with improved BMI and FEV1% predicted (20) . In an 18-month randomized control trial of 34 subjects with IGT (and increased "risk" -low BMI, FEV1%-pre-dicted<80 or decline of >10%), glargine, at a lower dose (0.15 units/kg/day), was not associated with greater improvements in lung function or nutritional status (21) . Similarly, the effects of 12 months of premeal insulin upon pulmonary function and BMI in adults with IGT was not different than the effects of placebo or pre-meal repaglinide, an oral insulin secretagogue (6) .

Future Steps: More detailed information regarding defects in insulin secretion and glucose excursion byeond the OGTT may be needed. Moreover, CF-specific derangements such as inflammation may contribute to insulin secretion and sensitivity defects, but these insulin secretion and sensitivity defects are inadequately depicted by OGTT.

Randomized control trials with adequate power and using insulin doses sufficient to promote anabolism are needed. Moreover, therapies that might promote β-cell function and interrupt progressive defects in insulin secretion (such as incretin-based therapy) are currently underway in subjects with early glucose abnormalities. The ultimate goal of these and other studies is to interrupt the cycle of worsening glucose tolerance, worsening insulin secretion, and worsening nutritional and pulmonary status and the complex interactions that occur among these disturbances. This presentation will develop the concept that mutant CFTR causes systemic redox imbalance and oxidative stress, resulting in islet cell dysfunction and the development of CFRD. Data will also be presented to illustrate how the CFRD environment of redox imbalance to the oxidizing state and hyperglycemia can enhance inflammation and thus accelerate lung decline in CFRD.

Biologic differences between oxidative stress and redox balance: Unregulated overproduction of reactive oxygen species (ROS) has two major effects -oxidative stress and alterations in redox homeostasis (1) . Oxidative stress relates to the presence of high levels of ROS which causes damage to macromolecules such as DNA, lipids, and proteins which then disrupts the function of these molecules. This damage can accumulate and cause cell death. Alterations in redox balance occur with lower levels of ROS, occur rapidly, are not associated with macromolecular damage, and signal normal cellular processes. There are three primary thiol/disulphide couples which control the majority of redox signaling and can function independently of each other. These three couples are glutathione/glutathione disulfide (GSH/GSSG), cysteine/cystine (Cys/CySS), and thioredoxins. Because each respective antioxidant system is in disequilibrium, the concept of oxidative stress being the balance between the sum total of all cellular oxidants versus all cellular antioxidants is an oversimplification.

How redox imbalance may be important in the development of CFRD: The literature on the role of oxidative stress in the development of CFRD is extremely sparse.

However, taking relevant data on the role of oxidant stress in type 2 diabetes and coupling with existing relevant studies in CF (2,3) plus preliminary data from our lab permits the following speculations on the potential role of oxidative stress and redox imbalance in the development of CFRD:

1. Numerous studies have shown that CF patients have a tremendous degree of chronic oxidative stress both in the lung and systemically, the latter exposing the vulnerable islet cells to damage.

2. Evidence suggests that CFTR is expressed in islet cells, including beta cells. Expression of mutant CFTR in beta cells would enhance ER stress, the unfolded protein response, and oxidative stress. Thus, CF beta cells have additional intracellular oxidative stress compared to those expressing wild-type CFTR and therefore are more vulnerable to damage.

3. Mitochondrial dysfunction is linked to type 2 diabetes. Mitochondrial dysfunction in CF was described over 30 years ago. We have found that oxidizing extracellular redox states cause selective ROS generation in the mitochondria and resultant mitochondrial protein damage, which would include oxidation of the major mitochondrial antioxidant thiore-doxin2. Thus, the systemic oxidized environment found in CF patients could exacerbate existing mitochondrial dysfunction and further impair insulin secretion.

4. We measured plasma Cys/CySS redox potential in healthy controls and in CF subjects with normal glucose tolerance (NGT), impaired glucose tolerance (IGT) or prediabetes, and CFRD before and two hours after a standard 75 gm OGTT. Subjects were aged 16 to 40 years. At baseline, healthy controls and CF with NGT had normal redox values for Cys/CySS (range (Eh Cys/CySS [x±SEM] = -86±2.34 mV and -80±4.17 mV respectively) but CF prediabetes and CFRD showed significant redox imbalance to the oxidized state (Eh Cys/CySS = -73±3.62 mV and -71±4.36 mV respectively, p<0.05, note that less negative = more oxidizing). The magnitude of this imbalance was not trivial as with normal aging, there is a 1 mV increase in oxidation for every 5 years of life. Thus this difference translates to these young adults with CF prediabetes and CFRD having the redox potential of an octogenarian. Two hours after the OGTT, CF with NGT, CF with IGT, and CFRD all demonstrated significant worsening of redox imbalance to the oxidizing state resulting in profound levels of systemic oxidation to a similar degree in all three groups that was significantly different from controls (2 hr Eh Cys/CySS: Control = -84±2.59, CF-NGT= -63±4.67, CF prediabetes = -60±4.22, CFRD= -63±2.45, p<0.05). This occurred even in the absence of any detectable hyperglycemia and when there was hyperglycemia, there was no correlation between the degree of hyperglycemia and the degree of redox imbalance. Finally, we have shown in vitro that this degree of redox imbalance decreases insulin secretion and glucose uptake.

Taken together, these data suggest a basic metabolic defect in CF consisting of acute glucose-induced redox imbalance to the oxidizing state which becomes chronic as CFRD develops and which is severe enough to worsen hyperglycemia (through inhibition of insulin secretion and glucose uptake) and ultimately cause beta cell failure.

How hyperglycemia and redox imbalance can worsen CF inflammation: The neutrophil is a key inflammatory effector cell contributing to CF lung disease. Therefore, we have conducted experiments to determine how the CFRD environment of redox imbalance to the oxidizing state and hyperglycemia might enhance neutrophil toxicity and found the following. First, we found that both glutathione and Cys extracellular redox states regulate neutrophil-derived extracellular ROS, with more oxidizing environments resulting in increased ROS production by these normal neutrophils. Additionally, neutrophils maintained in oxidizing conditions with high glucose concentrations had an increase in ROS generation over 8-fold of that seen in neutrophils maintained in oxidizing conditions with normal glucose levels. Interestingly, in the presence of reducing conditions, the increase in ROS generation induced by a high glucose environment was completely ablated, suggesting that the extracellular redox state is a very important regulator of hyperglycemia-induced oxidative stress. Our second series of experiments provide compelling evidence for a previously unrecognized metabolic defect in CF. Through a combinatorial approach of genome wide expression profiling, proteomics, and metabolomics using primary human airway epithelial cells and isolated human blood neutrophils, we found that CF neutrophils and CF airway epithelial cells have an intrinsic metabolic defect in glucose metabolism characterized by a switch to the glycolytic pathway. The implication of this phenomenon for CF neutrophils is a more primed phenotype, as neutrophils mainly depend on glycolysis for energy following activation (4, 5) . The increased glucose levels and redox imbalance to the oxidizing state found in the blood in CFRD are expected to enhance neutrophil priming and are expected to exacerbate responses once neutrophils are activated in the lung. Furthermore, normal adherent human neutrophils switch to glycolysis and generate ROS in the presence of hyperglycemia in a receptor-independent manner. This switch is temporary, lasting only a few hours before glucose is shunted away from the glycolytic pathway and ROS production diminishes. Given the increased expression of several enzymes of the glycolytic pathway in CF neutrophils, this regulatory step may be defective, resulting in prolonged, unabated production of ROS by CF neutrophils when in a hyperglycemic environment. This mechanism may be important not only in CF blood, but also in the CF airway wherein glucose and oxidized byproducts (e.g., advanced glycation endproducts) reach very high levels, thus potentiating neutrophil toxicity.

A final important point to make is that the CF lung has an intrinsic defect in the regulation of a key antioxidant defense mechanism that not only makes it particularly vulnerable to the oxidant stress and redox imbalance of CFRD but also may enhance inflammation. In 2008, Dr. Ziady identified that airway epithelial cells had defective regulation of Nrf2, a transcription factor that drives the expression of a host of genes that protect against oxidant stress and that regulate redox balance (6) . More recently, he and his colleagues (7) have begun to uncover the mechanism of Nrf2 dysfunction and found that cAMP-mediated signaling markedly reduces Nrf2 activity in CF cells. Inhibition of cAMP restored Nrf2 activity AND decreased NF-kB activation. The mechanism of this was via the CREB binding protein, CBP, as Nrf2 and NF-kB competed for CBP and that an inhibitor of cAMP shifted CBP from NF-kB to Nrf2. Since altered cAMP-mediated signaling is a response to loss of CFTR function, this provides a mechanism linking mutant CFTR to decreased Nrf2 and increased NF-kB.

Taken together, these data point to multiple mechanisms whereby pulmonary inflammation could be enhanced in CFRD and also indicate an inability to counter this enhanced inflammation because of defective Nrf2 signaling.

John Engelhardt, Ph.D.

Cystic fibrosis related diabetes (CFRD) is a common complication of CF and affects 20-25% of adolescents and 40-50% over thirty years of age. CFRD is associated with worsening clinical status including reduced pulmonary function and increased mortality compared to CF patients without diabetes. While the pathophysiology of CFRD is multifactorial, delayed insulin secretion appears to be a key hallmark of disease progression. Partial insulin deficiency occurs in part due to islet loss associated with exocrine pancreas disease. However, CF pancreata at CFRD autopsy demonstrate that remaining islets contain roughly half the number of insulin-positive cells found in non-CF controls and this degree of β-cell loss is thought to be insufficient to explain diabetes. Thus, insulin deficiency in CFRD is relative and not absolute. Multiple non-pancreatic organ abnormalities, including those in the lung, liver, and intestine, may modify glucose disturbances and insulin action in CF patients. For example, hepatic insulin resistance and/or enhanced peripheral insulin sensitivity has been observed in CF patients. Furthermore, chronic lung disease in CF is thought to contribute to abnormal glucose regulation and insulin resistance by increasing metabolic demand.

Characterizing early structural and functional endocrine pancreas abnormalities that precede glucose intolerance in CF has been difficult due to the lack of an animal model that spontaneously develops CFRD. CFTR knockout mice demonstrate enhanced sensitivity to low dose streptozotocin-induced hyperglycemia, but are euglycemic in the absence of streptozotocin. The recent development of CFTR knockout models in the pig and ferret provides new avenues for investigating the pathogenesis of CFRD. Both CF pigs and ferrets develop intestinal, hepatic, pancreatic, and lung disease that is pathologically similar to various stages of disease in CF patients; however, the progression of pancreatic disease in the two models is very different. CF pigs are born with significant exocrine pancreatic destruction, inflammation, and fibrosis characteristic of later phase damage in CF patients, whereas CF ferrets have mild pancreatic histopathology at birth. For these reasons, CF ferret and pig models may provide unique insights into various stages of CFRD pathogenesis.

We sought to determine whether alterations in insulin secretion and glucose homeostasis exist in newborn CF ferrets prior to structural loss of the endocrine and exocrine pancreas. Structurally, the newborn CF and non-CF endocrine pancreas was indistinguishable in terms of size and percent insulin and glucagon expressing cells in islets. However, there was an alteration in the distribution of large and small islets between genotypes and enhanced apoptosis in acinar and duct cells of CF animals at birth. Using L-arginine and glucose stimulation tests, our studies demonstrate that fasted newborn CF ferrets have impaired first phase insulin secretion (IFPI). Glucose tolerance tests (GTT) in fasted newborns also demonstrated that newborn CF kits have abnormal glucose tolerance. As expected with IFPI, delayed insulin responses following glucose stimulation to CF kits were accentuated at intermediate time points where hyperglycemia was observed during GTTs. However, the magnitude of insulin overshoot (~6 to 10-fold) at intermediate times points following glucose challenge was far greater for CF kits than that observed in CFRD adults and children with abnormal glucose metabolism. One explanation for the high insulin levels would be systemic insulin resistance often observed in older CF patients, however, no evidence to support systemic insulin resistance in CF newborn kits was found-plasma cortisol levels were normal and insulin/AKT signaling in liver, muscle, and adipose tissue was also not different from non-CF controls.

The abnormal regulation of insulin in fasted glucosechallenged CF animals prompted us to evaluate how insulin secretion was regulated in randomly nursing kits. Given the IFPI observed in newborn CF kits, we hypothesized glucose and insulin levels would be poorly regulated in fed animals. Indeed this was partly the case, plasma insulin, C-peptide, insulin/glucose, and C-peptide/glucose values were highly variable and on average much higher in non-fasted CF kits compared to non-CF controls. Importantly, the molar ratio of plasma C-peptide/insulin was not significantly different between CF and non-CF animals, indicating that the elevated insulin levels in CF kits represents bona fide insulin hypersecretion and that insulin clearance rates were not significantly reduced in CF kits. These tendencies observed in CF kits towards insulin oversecretion might play a role in the hypoglycemia that is observed in humans with CF. Under non-fasted conditions, a relationship between glucose and insulin levels is expected. Indeed, in non-CF non-fasted newborns, glucose and insulin levels were positively correlated such that lower glucoses predicted lower insulin levels. By contrast, in CF non-fasted newborns, insulin did not vary with glucose, such that high insulin levels were observed even when glucose was low. These findings support the notion that insulin secretion by the newborn CF ferret pancreas is not properly regulated in a manner that extends beyond IFPI.

Although mild insulin resistance and early exocrine pancreatic damage may be islet-extrinsic factors than contribute to altered insulin secretion in CF kits, studies in isolated cultured islets suggest islet-intrinsic abnormalities alter glucose-responsiveness of the CF ferret islet. Similar to nursing CF kits, insulin secretion by CF ferret islets is highly variable and less responsive to glucose stimulation in comparison to controls. At low glucose, in vitro cultured CF islets secrete 5.2-fold more of their insulin content relative to non-CF controls. This hypersecretion of insulin by CF islets at low glucose was blocked by activation of K-ATP channels with 100 µM diazoxide. Following stimulation with high glucose, non-CF islets demonstrated a significant induction in percent insulin secretion (5.1-fold), while the percent insulin secretion for CF islets was not significantly different at low and high glucose. The insulin secretory index was 4fold lower for CF islets as compared to non-CF controls. Thus, islet-intrinsic defects in the regulation of insulin secretion may occur in CF. More recent studies in neonatal CF pig islets are beginning to reveal similar abnormalities in glucose-dependent regulation of insulin secretion.

As CF ferrets age, progressive destruction of the exocrine and endocrine pancreas was associated with spontaneous hyperglycemia, glycosuria, and deteriorating glucose intolerance. By 21 days of age, dilated pancreatic acini were lined by cuboidal epithelial cells lacking zymogen granules. The lumen diameter of the acinoductular units was significantly larger in CF pancreata in all age groups compared to non-CF, with a marked increase between postnatal day 9-32 compared to pancreata from younger CF kits. By 21 days, there was loss of lobular parenchyma and replacement by bands of fibrous tissue that separated remnant dilated acinoductular units. No significant differences in inflammation were noted between genotypes of newborn or early neonatal pancreata. However, inflammatory cell infiltration into the interstitium and peripancreatic adipose tissue was significantly higher in CF as compared to non-CF pancreata by 9-32 days and was characterized by the infiltration of neutrophils with fewer lymphocytes and macrophages. Interestingly, progressive destruction and inflammation of the exocrine pancreas in 1-2 month old CF ferrets was associated with the appearance of spontaneous hyperglycemia and glycosuria. Abnormalities in glucose handling with age were evaluated by mixed meal tolerance test (MMTT) in CF and non-CF animals ranging from 45 to 90 days of age. Results from these studies demonstrated a variety of abnormal glucose tolerance phenotypes in CF animals consistent with impaired glucose tolerance and indeterminate glycemia, as defined using human clinical criteria. Despite these observed differences in MMTT profiles of CF animals, the average glucose profile for CF animals was significantly different from non-CF animals at all time points in the glucose curve. Additionally, AUC 0-120min was also significantly (P<0.0001) higher for CF (37,281 +/-4520 mg/dL•min) as compared to non-CF (15,293 +/-1005 mg/dL•min) animals. These older CF ferrets demonstrated a 52% loss of insulin positive cells, which is very similar to the~50% loss in β-cells seen at autopsy in CF patients including those with overt diabetes and glucose intolerance. Relatively little is known about how early phases of pancreatic inflammation and exocrine decline impact glycemic regulation in CF patients. Based on the CF ferret, inflammation association with the onset of exocrine pancreas decline appears closely linked to spontaneous glucose abnormalities and worsening of glucose tolerance.

In summary, studies in CFTR-knockout ferrets suggest that abnormalities in insulin secretion and glucose metabolism may occur very early in CF patients.

Although multiple factors likely contribute to altered insulin regulation in CF, including exocrine pancreas inflammation and changes in systemic insulin resistance and sensitivity, our studies suggest for the first time that islet-intrinsic abnormalities in CF influence glucosemediated regulation of insulin. Future studies of the CF ferret and pig models may help to clarify the mechanism of altered insulin secretion in CF patients.

Clement L. Ren, M.D. Pediatrics/Pulmonology, Univ. of Rochester, Rochester, NY, USA Newborn screening (NBS) for cystic fibrosis (CF) is now available in the entire United States, as well as many other countries across the world. CF NBS algorithms vary amongst different states and countries, but in all cases a positive NBS ultimately leads to diagnostic testing for CF through measurement of sweat chloride [1] . In some cases sweat Clresults are outside the normal range (>29 mmol/L) but below the diagnostic threshold level for CF (<60 mmol/L), leading to an indeterminate result. Infants with indeterminate CF NBS results present a diagnostic challenge to clinicians and a psychological and emotional stress to families. Parad, et al were one of the first groups to describe what they termed "diagnostic dilemmas" [2] . Subsequently, the U.S. CF Foundation convened a panel of experts who proposed the term CFTR-related metabolic syndrome (CRMS) to describe infants who have elevated IRT detected by CF NBS, but initial genetic or sweat Clresults that are nondiagnostic of CF [3] . Guidelines for the definition and follow up of infants with CRMS have been published, but little is known about how the diagnosis of CRMS is being applied and the clinical outcomes of infants with CRMS. The CFF defines CRMS as healthy infants with hypertrypsinogenemia detected through CF NBS and either: (1) intermediate sweat chloride concentrations (between 30-59 mmol/L) on at least 2 occasions and fewer than 2 CFcausing CFTR mutations; or (2) a normal sweat chloride (<30 mmol/L) and 2 CFTR mutations, of which no more than 1 is known to be CF-causing [3] . It is important to distinguish CRMS from CFTR related disorder [1] . The former applies only to asymptomatic infants identified through CF NBS, whereas the latter occurs in older individuals with signs or symptoms of CF and a nondiagnostic sweat test. In 2010, the CFF Patient Registry began including patients with CRMS and CFTR related disorder. In 2009, there were 187 reported new diagnoses of CRMS [4] . CF centers are also asked to report the number of CRMS cases diagnosed at their center as part of the annual renewal process. In 2010, the proportion of positive NBS tests that resulted in a diagnosis of CRMS ranged from 0 to 84% [4] . This wide variation in reported prevalence raises questions about whether the diagnostic criteria for CRMS are being applied correctly.

There are limited data regarding clinical outcomes of infants with CRMS. We recently reported our experience with CRMS over an 8 year period [5] . Compared to CF patients, CRMS patients were more likely to be pancreatic sufficient as assessed by fecal elastase measurement (100% vs. 8%, P<0.01). Their weight for age percentile was normal from birth. A positive oropharyngeal (OP) culture for Pseudomonas aeruginosa (Pa) was found in 25% of CRMS patients compared to 59% in CF patients (P=0.085). One patient with the F508del/R117H/7T genotype was reassigned the diagnosis of CF after he had a positive OP culture for Pa, and his follow up sweat Clat 1 year of life was 73 mmol/L. Other studies have been published that did not focus specifically on CRMS outcomes, but they had infants in their study population that would be considered CRMS who also demonstrated signs of CF disease [6] [7] [8] . Although our study was conducted at a single center and involved a small number of patients, our results and those of others suggest that CRMS infants are at risk for developing signs of CF disease.

In summary, CRMS is an indeterminate outcome arising from CF NBS, and it may be associated with the development of CF disease. Further research is needed to establish accurate diagnosis of CRMS, its prevalence, and its outcome.

References 

Kevin W. Southern, Ph.D., MBChB

Expansion of newborn screening (NBS) for CF across the globe has been dramatic over the last decade. National NBS programmes now exist in many countries including France, Russia, United Kingdom and the United States. There is evidence from a number of sources that NBS impacts positively on patient well being and survival. Newborn screening has led to a "sea change" in how we manage our patients and how our clinics run. It is likely to be the most important intervention for our patients of the past decade and probably the next, with respect to long-term wellbeing (1-6).

All current NBS programmes use the measurement of immuno-reactive trypsinogen (IRT) from a dried blood spot sample as the initial screening test in the first week of life. This is a sensitive test and will only miss a very small proportion of infants with classic CF. A second tier of testing is necessary to improve the specificity of the protocol and a variety of strategies are available, most commonly this involves DNA analysis for common mutations of the cystic fibrosis transmembrane conductance regulator (CFTR) gene. Numerous different screening strategies (over 70) are in operation around the globe. This variance relates mainly to different cut-offs for the IRT assay and different second tier strategies. Whatever protocol is employed, there is a risk of identifying infants with an equivocal diagnosis. These infants have either an intermediate sweat chloride value (between 30-60 mmol L -1 ) with unclear genotype OR the recognition of two CFTR mutations, the molecular consequences of which are unclear and with a normal (or intermediate) sweat chloride value. Protocols with a low IRT cut-off and protocols that employ extended DNA analysis (including gene sequencing) are theoretically more at risk of recognising infants with an equivocal diagnosis. The recognition of infants with an equivocal diagnosis is also dependent on the population screened. There is a good argument that NBS programmes should consider strategies that reduce the recognition of these infants, as a significant number of them will have no longterm medical consequence from this result. Strategies that reduce the recognition of these infants may have an impact on the performance of the programme with respect to the recognition of infants with classic CF, so a balance must be achieved. Regardless of protocol, NBS will always result in the recognition of infants with an equivocal diagnosis. There has been much discussion on the term used to describe these infants. In the U.S., they are designated the term CFTR related metabolic syndrome (CRMS). There is some debate as to how these infants should be managed (7, 8) .

This presentation will consider the question, "Should infants with CRMS and unclear diagnoses receive CF care?" and provide the PRO argument.

Having undertaken a programme of screening and having identified a positive result (including carrier status) there is a moral duty on the health service to report that result to the family, even if the long-term consequences are not clear. The most important message of this presentation is that providing clear and extensive information for the families is key to a good outcome. Invariably, when outcomes are poor it has been as a result of poor communication between healthcare workers and the family or other professionals. This information is not easy to translate and needs to be revisited over and over. Families are often more willing to accept a "grey" situa-tion than healthcare professionals, but they need to have a clear appreciation of the result and of the consequence of that result for the infant and for the family.

The consequence for the infant is that they will likely remain well for the rest of their life. A very small percentage may develop symptoms consistent with a diagnosis of CF, most notably upper airway problems (nasal polyposis and chronic rhinosinuisitis) and in some cases significant lower airway infection and damage. A tiny percentage of these infants may even develop some evidence of pancreatic insufficiency in addition to respiratory disease. There is no doubt that a small proportion of these infants will be at risk of significant disease, however the vast majority will remain well. These infants may be at risk of some "adult" CFTR-related disorders. The most well-characterised is congenital bilateral absence of the vas deferens (CBAVD) in men and this requires careful and sympathetic counseling. Other less well characterised CFTR-related conditions include chronic pancreatitis, rhinosinusitis and isolated bronchiectasis, although there is some debate as to whether a patient that presents with one of these conditions is simply expressing a single organ manifestation of CF (9) .

The consequences for the family relate to the genetic findings and the burden of medicalisation. Parents should be aware that subsequent pregnancies may be at risk of classic CF and again this is a challenging counselling process. Extended family members may opt for genetic investigation as well.

Should these infants be in a CF clinic? There are clear arguments for this, although in many ways this represents an easy option for CF teams to take and it is important that these infants do not simply slot into standard CF care pathways. Infants attending a CF clinic may be more at risk of cross infection, but one would hope in 2012 that clinics will be taking careful measures to avoid such risks. The main argument for attending a CF clinic is that families will receive dedicated care and clear information. They will not become "lost in the system." These are very challenging infants to manage, in some ways much more so than the infant who is homozygous for phe508del. These infants do not need to be seen as frequently as infants with classic CF, in fact following the diagnostic year, they should probably only be seen once a year. But again information for the family is the key to this. The family must be aware that if their child develops a cough that is not responding to antibiotic treatment that they need to contact the CF team and they must have a pathway to do this. There may be a point where we can discharge these children from CF care. At present we do not have sufficient data or experience to define that cutoff but projects around the world (notably CFTR-2 and the ongoing French research project) will help us in this regard.

In summary, we have a profound duty of care to this expanding tribe of children. We should do everything we can to avoid unnecessary recognition of infants with an equivocal diagnosis (CRMS), but once recognised they need dedicated CF care that is sensible and sympathetic to long-term needs.

John Massie, MBBS, FRACP, Ph.D.

The CFTR Metabolic Syndrome (CRMS) describes the situation that follows newborn screening for CF in which infants are left without a clear diagnosis as either having CF or being carriers. This can arise in a variety of ways depending on the newborn screening protocol used. The most common scenario is an infant with an elevated immunoreactive trypsinogen (IRT) and one CFTR gene mutation who has a borderline sweat chloride result, 30-59 mmol/L. Doubt about the diagnosis of CF, or exclusion of carrier status, exists because these infants have no symptoms attributable to CF, have genotype information that is incomplete (or results that are of uncertain clinical significance) and the sweat chloride test is also inconclusive. It should be remembered that newborn screening followed by sweat chloride analysis, can, in most instances, establish the diagnosis of CF, confirm carrier status to those infants identified, and exclude CF to almost all without CF. A small proportion of infants are left without a clear diagnosis, but given the widespread nature of newborn screening for CF in the United States and elsewhere, the number of these infants is not inconsequential. The implications of mislabelling these infants are huge. The consensus document that describes CRMS highlights that some physicians are reluctant to apply a medical label to symptom free infants who may never have symptoms, while others have experience with patients who have had significant symptoms develop. There can be no absolute answer as to who is right, since the development of clinical CF is a complex interaction between the infant's (unknown) CFTR function and environmental factors that cannot necessarily be controlled. In this presentation, however, I argue that it is wrong to treat these asymptomatic infants as if they have CF.

To start it is worth examining the CFF consensus papers on the diagnosis of CF. The key message is that CF is a clinical diagnosis. Without recognised clinical features there is no diagnosis of CF. The difficulty with these statements is that a positive newborn screening test can replace clinical features, but what constitutes a positive newborn screening test is not made clear. As it stands, infants with the CFTR metabolic syndrome, by definition, have no features attributable to CF and therefore should not be labelled as having CF and should not be treated as such. A label is a powerful concept and we should be careful how we apply the label of CF.

This approach may be considered too simplistic, as it is the chance of developing CF that allows us to treat the healthy appearing infant who has two disease causing CFTR mutations, and/or a sweat chloride >60mmol/L. So what is the chance of the infant with the CFTR metabolic syndrome developing CF? We have no way of answering this, but do have a literature full of case reports about patients who may be the exception, and this seems to be driving the case to label all the infants with an uncertain diagnosis after newborn screening with CF. The consensus statements on the diagnosis of CF indicate that the diagnosis of CF is confirmed with evidence of CFTR dysfunction through gene mutation analysis or CFTR function as measured by sweat chloride (other tests of CFTR function such as nasal potential difference or intestinal currents are used as supplementary information in some centres). The difficulty with the genetics is that very few mutations have had functional studies to quantitate CFTR mRNA and predict outcome. CFTR sequencing may reveal sequence variations of uncertain significance, and even if a previous identified sequence has been linked to clinical disease, this is almost exclusively in patients who have presented with some CF-like phenotype, and does not predict that the infant with the same genotype will develop clinical disease. Equally, much hinges on the sweat chloride test, but even on repeating multiple times is unlikely to give the answer in the immediate term. Given the co-efficient of variation of the sweat chloride test is up to 20%, relying too heavily on cut-offs is problematic. These CFTR genetic and functional factors promote doubt as to the likelihood of developing CF, and reinforce the inappropriateness of applying a diagnostic label of CF.

It might be considered that even CRMS is too much of a diagnostic label! The consensus statement on management of these infants at once acknowledges the difficulty achieving consensus on recommendations and at the same time is quite prescriptive in parts of follow-up. I advocate that an approach based on ethical principles is required. To start with, what is needed for the patient and their family is disclosure of the uncertainty of the problem and options open to them. This respects their autonomy to make a decision for follow-up that suits them. This is a partnership of care that takes into account their knowledge of their family, their own circumstances and the way they wish to raise their child. Next, we need to ensure good, evidenced based outcomes for these infants. This is beneficence, and labelling them with CF is unlikely to achieve this. In considering this problem, we should not forget the intent of newborn screening, which is to find those infants with CF who would benefit from early diagnosis. Also remembering how hard it was to prove benefit of newborn screening to patients with a clear diagnosis of CF, there is no evidence of benefit to the CRMS group. In attempting to manage these CRMS infants "in case they have CF" there is little normal data on fundamental aspects of CF care such as throat/cough swabs, and little evidence of key treatments such as prophylactic antibiotics, infant airway clearance and mucolytic therapy. Basic genetic counselling is essential, in particular to offer choices to the parent carrier's family, but on this point it is no different to how carriers discovered by newborn screening are dealt with. Beyond routine health advice such as avoidance of cigarette smoke exposure, breast feeding where possible and annual influenza vaccine from six months of age there is not evidence for CF specific advice in this group. Just as important is making sure that no harm comes to these infants (and their families), this is non-maleficence. There are considerable psychosocial risks to the family or child taking on a sick role when there is no illness. Harm may also arise from medication side-effects, irradiation or general anaesthetic for procedures. There may be considerable harm from more in depth genetic counselling, trying to base future reproductive decisions on uncertain (or wrong) advice. Later there may be adverse insurance or employment outcomes, a consequence of a label misapplied many years before. The principle of justice is perhaps the weakest one upon which to argue, but in a health care setting that is increasingly limited, it is difficult to justify the expense of diagnostic and treatment resources on people with no illness.

In summary I think that a diagnostic label should not be applied to infants with uncertain results after newborn screening for CF. The family should be informed of the uncertainty and a plan of follow-up decided together between the family and physician. Any health advice provided should be evidenced based. Every step should be made to improve newborn screening paradigms to minimise the number of infants who fall into the CRMS group. Better estimates of CFTR function for all the known sequence variations need to be developed. Data should be collected on the outcome of those infants whose families elect active surveillance. Newborn screening for CF provides the opportunity to optimize nutritional status early and to monitor for and potentially treat lung disease prior to the onset of irreversible structural airway damage. Longitudinal studies utilizing bronchoalveolar lavage, infant lung function testing and computed tomography (CT) have established that the hallmark features of CF lung disease, infection, inflammation and structural airway damage, frequently begin in infancy and prior to the onset of symptoms (1) (2) (3) (4) .

Measurement of forced expiratory flows (raised volume rapid thoracic compression technique), lung volumes (body plethysmography) and ventilation inhomogeneity (multiple breath washout) have elucidated the early manifestations of heterogeneous peripheral airway obstruction in CF (5). Diminished forced expiratory flows are associated with infection with P. aeruginosa (6), S. aureus (7) and lower airway inflammation (7, 8) . Although expiratory flows and lung volumes improve postantibiotic therapy for a pulmonary exacerbation, flow limitation and air-trapping/hyperinflation have been reported to persist (9) .

Studies using computed tomography (CT) imaging of the chest have conclusively demonstrated that structural airway damage, including bronchial wall thickening and even bronchiectasis, are common in the first years of life.

Furthermore, structural changes on chest CT in young children with CF are associated with worse pulmonary inflammation and infection (2, 10, 11). Long, et al (12) obtained limited-slice high-resolution CT images of the chest in 34 children with CF (aged 10 weeks to 5.5 years) and 20 healthy disease controls without CF. Their study demonstrated that the airways of the CF cohort had thicker walls with more dilated lumens. Compared with airways from the healthy population, 94% of children with CF were reported to have at least 1 airway that was considered bronchiectatic. Martinez, et al (13) performed limited-slice high-resolution chest CT in younger subjects with CF (8 months to 2.5 years) and found that children with CF had airways with thickened walls and narrower lumens compared with controls. It is unclear whether the observed airway wall thickening is a precursor of the findings of Long, et al in slightly older subjects, or indeed, of bronchiectasis as there are few longitudinal data. However, bronchiectasis can be identified by high-resolution chest CT even in the first few months of life (10) with one prospective population-based cohort study showing that the prevalence of bronchiectasis increased subsequently during the preschool years (11) . Bronchiectasis is already established by 8 years of age in the majority of children with CF. In a recent prospective study (14) in children who had been diagnosed clinically with CF, 51/60 (85%) had evidence of bronchiectasis on volumetric chest CT at a mean age of 7.8 years. Such changes can be missed by spirometry, which can remain normal even when significant and progressive bronchiectasis is present (15) .

The most clearly established benefit of CF newborn screening is improved nutritional outcomes. Nonetheless, even with screening, infants with CF frequently have suboptimal nutritional status (16) . Low levels of insulinlike growth factor 1, first identified in the pig model of CF and subsequently confirmed in infants identified by newborn screening, may partially explain this observation (17) . In addition, lower airway inflammation has been associated with worse nutritional status (18).

Sarath Ranganathan, MBChB, FRACP, Ph.D.

The importance of infancy in CF lung disease: There is convincing evidence that lung disease in CF starts early in life. Evidence is accumulating that lung structural changes (1-4), pulmonary infection and inflammation (3) and diminished lung function (5,6) can be detected early in infancy, even in those who are asymptomatic. Changes that indicate bronchiectasis can be identified in about a fifth of children shortly after diagnosis (3) and its prevalence increases throughout the preschool years (4). The structural changes are progressive in association with classic features of the early pathophysiology (2) and become well-established by school-age (7). These are extremely important data in a condition where 90% of those affected will eventually die as a result of progressive lung disease and where later radiological findings on chest computed tomography (CT) are predictive of poorer prognosis and death in endstage CF lung disease (8) .

As in other chronic respiratory conditions (9), the first few years of life appear to be critical in CF and, unfortunately, this potential life-long trajectory for lung disease cannot be predicted by the presence or absence of symptoms. The finding of the classic pathophysiological features of CF during the first two years of life, a period of extremely rapid lung development, is therefore particularly alarming and invokes the need to change our paradigm for treatment of CF away from limiting, or even reversing, lung disease and, instead, towards its prevention. One such clinical trial is already underway, with prevention of bronchiectasis as its primary endpoint.

Can early treatment improve the outlook for those with CF? Yes! At a population level reported data from several countries indicates that successive cohorts of children with CF have significant improvements in lung function when first measured by spirometry at 6-7 years of age, with little accrued benefit after this age. This indicates that improvements in clinical management of CF lung disease in the preschool years, prior to the first performance of spirometry, has resulted in such detectable improvement. The improved outcomes are likely to be due to multiple factors although we are well aware that there is little evidence to support any specific treatment during the pre-school years. However, our accumulated clinical experience suggests that a combination of aggressive treatment of infections, airway clearance, nutritional treatment and support and attentive clinical supervision have contributed to better outcomes. Including all these facets in management of infants (maximum treatment) is necessary. In addition, newer strategies in CF, such as eradication of infection with Pseudomonas aeruginosa by the use of aggressive antimicrobial therapy, which was also introduced widely into clinical practice without the benefit of randomized control trials but on the basis of results of observational studies, are also indicated. Many would suggest that there would not now be equipoise if randomized control trials of any of these component strategies were to be conducted. Thus, in early CF evidence is required in order for us to withdraw any of the components from the treatment armamentarium in order to preserve the edict of "first, do no harm."

Need for clinical trials in infants with CF: One difficulty faced in early CF management is the reluctance for pharmaceutical companies to conduct randomized-controlled trials of CF therapies in infants once evidence for their efficacy and effectiveness have been proven in phase three trials in older subjects. This reluctance continues despite there now being reasonable study end-points for assessing therapeutic outcomes objectively in pre-school children (10) . Again at a population level, differences in outcomes have been shown to exist at national levels, with worse outcomes evident in Australia where there was more restrictive access to CF therapies versus the United States (11) . It is crucial that all CF therapies are available for use in pre-school children, ideally after appropriate clinical evidence is acquired. The new paradigm means that this is the most important age group to target when designing studies of new interventions and equipoise exists for such studies.

Summary: In conclusion, lung disease starts early in CF, is progressive and is modified by standard CF management paradigms. The presence and progression of early lung disease cannot be predicted by the presence or absence of symptoms. Not only should we treat asymptomatic infants with CF with maximum current therapies, newer therapies, including disease-modifying treatments, will need to be targeted from diagnosis in order to try to prevent lung disease. Future studies need to address how treatment can be individualized and so better targeted, but current evidence indicates that targeting treatment on the basis of symptoms alone is inappropriate.

References 

Alan R. Smyth, M.D., MBBS, MRCP, FRCPCH

Since the landmark randomised controlled trial of newborn screening in cystic fibrosis (CF), performed by Farrell and colleagues in 1997 (1), we have known that early diagnosis of CF in young infants leads to better nutritional outcomes. However, this positive finding was tempered by the observation that, in one of the two CF centres participating in the screening trial, the prevalence of pulmonary infection with Pseudomonas aeruginosa was higher in the screened group (2). This centre did not have in place rigorous measures to prevent cross infection with P. aeruginosa when children attended clinic. Hence early diagnosis of CF may have unintended consequences for these infants. Clinicians have a responsibility to ensure that any intervention in young infants with CF, who are free of symptoms, is evidence based and that the benefits outweigh the risks.

There are sound reasons for wishing to intervene early to prevent lung disease in CF. Airways infection and inflammation develop at an early stage. In a key study, almost 40% of infants with CF had evidence of lower respiratory infection before 3 months of age, as assessed by bronchoalveolar lavage (BAL) (3). When the same infants were followed up for 2 years, new or persistant infection was associated with raised inflammatory mediators and, in those infants whose infection was successfully treated, inflammatory mediators often did not return to baseline levels (3). This infection and inflammation translates into structural damage in the lungs of children diagnosed through newborn screening, with over half having some evidence of bronchiectasis on high resolution CT (HRCT) scan at 5 years (4).

So what "maximum therapies" might we give to asymptomatic infants with CF? The 2009 European best practice guidelines make only four clear statements regarding therapy in infants diagnosed through screening (5). These are: prophylaxis against Staphylococcus aureus (the effects are unclear); immunoprophylaxis against respiratory syncytial virus (no evidence); chest physiotherapy (long term trials needed); vitamin K supplementation (recommended at all ages but dose not validated for infants). The U.S. CF Foundation pulmonary guidelines (6) make only one recommendation that applies to pre-school children and that is not to use antibiotic prophylaxis against S. aureus (at any age).

There is a frequently a delay between the publication of evidence from randomised controlled trials and incorporation into guidelines. So what can we learn from recent trials in young infants with CF which might support "maximal therapy?" As much of the evidence for early infection and inflammation in CF comes from BAL, it seems reasonable to suppose that one should perform a BAL when infants become symptomatic, in order to make an informed choice of antibiotic. The harms must be weighed carefully against the proposed benefits -not least because bronchoscopy in young children requires general anaesthesia. In one of the most important clinical trials to be conducted in screened infants with CF, Wainwright and colleagues (4) randomised infants to BAL directed therapy (initial BAL before 6 months; when hospitalised for respiratory symptoms; before and after P. aeruginosa eradication therapy). This was compared with standard therapy (treatment based on upper respiratory specimens). At 5 years the two groups were compared by looking for P. aeruginosa in BAL and by HRCT score. There was no difference between the two groups for either of the primary outcomes. However bronchoscopy was not without complications: 5% of procedures were followed by a substantial clinical deterioration and 29% by a worsening cough.

Clinicians might be tempted to use therapies in young children that are of proven effectiveness in older patients with CF. The recently published ISIS trial should provide a corrective against this kind of extrapolation (7) . This large, well designed randomised controlled trial enrolled over 300 asymptomatic infants with CF and randomised them to hypertonic saline (7%) or normal saline placebo, nebulised twice daily. The pri-mary outcome was a protocol defined exacerbation of pulmonary symptoms, requiring antibiotics (oral, inhaled or intravenous). There was no difference between the groups in the primary outcome measure (or indeed any of the secondary outcomes). A cough (moderate or severe) was reported in just over one third of patients in each group.

For older drugs, high quality randomised controlled trials in pre-school children are lacking. For instance, 15 randomised controlled trials of dornase alfa have been conducted, which are of sufficient quality to be included in the Cochrane review (8) . However, only one of these included children under 6 years (9). The primary outcome of this study was microbiological, with a reduced prevalence of lower respiratory infection (particularly S. aureus) in the treated group. Only conventional spirometry was performed in this study and so no conclusions can be drawn about lung function in the pre-school children. There is a similar lack of evidence for the use of prophylactic antibiotics to prevent S. aureus, where the Cochrane review includes only 4 trials of 401 patients (10) . Prophylactic antibiotics reduce the number of young children with one or more isolates of S. aureus with a non-significant trend towards more P. aeruginosa with prophylaxis. A definitive answer awaits more clinical data.

This lack of evidence is not restricted to drug therapy. In a Cochrane review of chest physiotherapy vs. no chest physiotherapy, only 6 trials could be included, comprising 66 individuals, none of whom were pre-school children (11) . Studies which have looked at young infants have not been randomised comparisons and so little can be concluded about the potential benefit of chest physiotherapy in young infants. However, these studies do give some idea of likely harms, specifically the occurrence of gastro-oesophageal reflux during percussion with postural drainage (12) . The current UK recommendations are to advise parents of asymptomatic young infants with CF not to perform routine airway clearance but rather to observe their child carefully and start treatment should symptoms develop (13) .

In the coming decade we may see more clinical trials in asymptomatic infants. Currently these trials are limited by the lack of relevant clinical outcome measures, which are validated in this age group and can be performed in multiple centres (14) . This situation may change as outcome measures such as lung clearance index, high resolution CT and BAL are further refined. Until then treatment should be restricted to those therapies where there is good evidence of benefit and where the benefits outweigh the harms. The results of the ISIS trial have been published (1) with an accompanying editorial (2). Briefly, this was a multicenter, randomized, double-blind, placebo-controlled trial of 7% hypertonic saline versus 0.9% isotonic saline inhaled twice daily for a year in 321 patients with CF younger than 6 years of age. This trial was the first to evaluate a non-antibiotic therapy in children with CF in this young age group, and speaks to the ability of this type of study to be done in the future in young children with CF. Assessing potential therapies before the age of 6 years is important because therapies in this age group may be able to prevent lung damage from occurring, rather than treating lung damage that is already irreversible. Hypertonic saline did not have any effects on the primary or secondary outcome measures, which included pulmonary exacerbation rates and measures of respiratory symptoms, respectively. Other potential endpoints may need to be considered in this age group, such as CT scans, lung clearance index or other measures of lung function.

The study also had a subgroup of 74 participants in whom infant lung function testing was performed at the beginning and end of the one-year treatment period. There was a small but significant difference between the two groups in the forced expiratory volume in 0.5 seconds (FEV0.5). However, more than one-third of the participants did not have two accept-able measurements, which could have introduced bias. Further, the clinical significance of this difference in FEV0.5 is not known.

An additional ancillary study at the Toronto site evaluated the change in lung clearance index over the treatment period measured by sulfur hexafluoride multiple breath washout in 20 participants. These results will be presented.

At this point, there is no evidence that initiating treatment with 7% hypertonic saline in CF patients below the age of 6 years would be beneficial in terms of pulmonary exacerbation rates, height, weight, respiratory rate, oxygen saturation, cough or respiratory symptom score. Therefore, this treatment should in general be reserved for those CF patients above the age of 6 years until more evidence is available. However, hopefully different endpoints that look at progression, rather than treatment, of existing disease can be used in future studies to assess whether hypertonic saline have any benefit in these young CF patients. Overview. Activation of "alternative," calcium-activated chloride channels (CaCCs) is considered to be a potential strategy for CF therapy, and potentially for pancreatic and biliary hyposecetion, dry mouth/eye and gastrointestinal hypomotility disorders. Airway epithelial cells express CaCCs, as do secretory epithelial cells and pacemaker cells in the gastrointestinal tract. In CF, the rationale is that activation of CaCCs would restore chloride permeability to CFTR-deficient, CF cells. CaCC-targeted therapy is based on the assumption that stimulating airway epithelial cell chloride permeability by an unregulated mechanism would be beneficial.

Unfortunately, Phase 3 clinical trials of the purinergic (P2Y2) receptor agonist, denufosol, failed to show efficacy, which may be because it produces only transient elevation in cytoplasmic calcium. By high-throughput screening, we have identified small-molecule inhibitors and activators of CaCCs, including those of the TMEM16 class (by target-based screens), the principal CaCCs in airway submucosal gland and salivary gland epithelia, and other, as yet unidentified, CaCCs (by phenotype-based screens). Small-molecule CaCC activators that target CaCCs directly without altering cellular calcium are potentially attractive development candidates because they can produce sustained CaCC activation without generalized calcium elevation, and potentially exploit native epithelial cell regulatory mechanisms involving basolateral potassium channels.

CaCCs. CaCCs are widely expressed in epithelial and non-epithelial cells, where they facilitate epithelial fluid secretion, smooth muscle contraction, neurosensory signaling, and other functions (1). TMEM16A (anoctamin-1, ANO1) was identified in 2008 by the Oh, Galietta and Jan labs as the first bona fide CaCC, as its expression in oocytes and mammalian cells produced outwardly rectifying, calcium-sensitive chloride currents. TMEM16A is expressed in epithelial cells in the airways, salivary gland, pancreas, biliary tract and intestine, as well as in arterial smooth muscle, intestinal pacemaker cells, sensory neurons and some tumors.

Though TMEM16A knockout mice die just after birth because of tracheomalacia, electrophysiological measurements in the neonatal knockout mice suggested the involvement of TMEM16A in chloride secretion in mouse salivary gland and airway epithelia. TMEM16A is also involved in intestinal and vascular smooth muscle contraction, nociception and bile formation. There are multiple TMEM16 family members, with the A and B isoforms functioning as CaCCs. In addition, there are multiple variants of each isoform produced by alternative splicing, which introduces considerable complexity in the analysis of TMEM16 expression, function and pharmacology. Though TMEM16A is likely the major CaCC in salivary and airway submucosal gland epithelia, and in intestinal pacemaker cells (interstitial cells of Cajal), the molecular identity remains unknown of the other, perhaps more than one, epithelial CaCC(s).

CaCC inhibitors. Until recently, available CaCC inhibitors included non-selective anion transport inhibitors with low potency such as DIDS, DPC, NPPB and niflumic acid. In 2008, prior to the molecular identification of TMEM16A as a CaCC, we identified CaCC inhibitors from a phenotype-based high-throughput screening in human intestinal epithelial cells (HT-29), in which YFP-H148Q/L152L was introduced using lentivirus (2). The most potent inhibitor, the arylaminothiophene CaCCinh-A01, had IC50 ~ 1 µM for CaCC inhibition in multiple cell types. CaCCinh-A01 is under evaluation as a candidate therapy for rotaviral diarrhea. Following the identification of TMEM16A, we carried out high-throughput screens to identify inhibitors that target TMEM16A directly, without affecting upstream calcium signaling. We found that some compounds, including tannic acid and related gallotannins (3), function as non-selective CaCC inhibitors that block TMEM16A and other, as yet unidentified, CaCC(s) in multiple cell types. We proposed that CaCC inhibition by gallotannins in red wines and green teas may account, in part, for their health benefits including reduced risk of cardiovascular disease. TMEM16A-selective inhibitors were also identified, including the aminophenylthiazole T16Ainh-A01 (4) and the natural product eugenol (5). T16Ainh-A01 inhibited CaCC chloride current in TMEM16A-transfected cells and in cultures of human salivary gland and IL-4 treated bronchial epithelia, but not in intestine, providing pharmacological data on TMEM16A involvement in CaCC function in various tissues, and for its limited involvement in CaCC activity in unstimulated human airway surface epithelium and intestinal epithelium. T16Ainh-A01 has been widely used in TMEM16-related research. In CF, TMEM16A inhibitors might reduce mucin secretion.

CaCC activators. An activator screen of~110,000 compounds revealed compounds that activated TMEM16A CaCC conductance without increasing cytoplasmic calcium (6) . By patch-clamp, N-aroylaminothiazole "activators" (Eact) strongly increased chloride current at zero calcium, whereas tetrazolylbenzamide "potentiators" (Fact) were not active at 0 calcium but reduced the EC50 for calcium-dependent TMEM16A activation. Of 682 analogs tested, the most potent activator and potentiator produced large and more sustained CaCC chloride currents than purinergic agonists of calcium signaling. Interestingly, analogs of activators were identified that fully inhibited TMEM16A Clconductance, providing evidence for direct TMEM16A binding. Eact increased CaCC conductance in human salivary and airway submucosal gland epithelial cells, and IL-4 treated bronchial cells, and stimulated submucosal gland secretion in human bronchi and smooth muscle contraction in mouse intestine. Phenotypebased screening is in progress to identify activators of non-TMEM16 CaCC(s) in human airway epithelial cells.

References (1). Despite intensive research, established airway colonization in CF patients cannot be eradicated by any known therapy, and intensive antibiotic treatment is only effective at suppressing these persistent infections. Moreover, the frequent use of suppressive antibiotics has made antibiotic resistance a vexing problem in CF patients.

Much attention has been focused on the need for new antibiotics. Our current antibiotic arsenal is inadequate for several reasons. First, resistance due to bacterial genetic mutation is ever increasing for both hospital and community acquired organisms (2, 3) . Second, antibiotics work poorly against biofilms that cause many chronic infections, and no existing treatments target the biofilm growth mode (4). Third, overwhelming infections like severe sepsis have high mortality even when appropriate antibiotics are used (5) . Finally, emerging pathogens with high intrinsic resistance are increasing in frequency worldwide (6) .

In relation to CF, new agents are also needed because conventional antibiotics work poorly in chronic infections, even when the organisms are sensitive when tested ex vivo. These infections resist treatment in large part because the organisms live in biofilms (4) . Biofilms are communities of bacteria associated with surfaces and encased in a polymeric matrix making the bacteria far more resistant to killing than they are in the free-living (planktonic) state (4) . Examples of biofilm infections include CF lung infections, endocarditis, osteomyelitis, wound, sinus, and device infections (7) .

A novel strategy to combat infection is to augment natural defenses, and a key mechanism of host defense is to withhold iron (Fe) from infecting organisms (8) . The importance of Fe availability in acute infections has been established in many systems (3, (8) (9) (10) (11) (12) (13) . Recent work also links Fe availability to chronic infections in that Fe promotes biofilm formation (2, (14) (15) (16) (17) (18) . The central role of Fe in infection suggests that bacteria may be vulnerable to interventions that disrupt Fe metabolism.

We are pursuing a novel approach that uses the metal gallium (Ga) to disrupt bacterial Fe metabolism. Ga has a nearly identical ionic radius as Fe, and many biologic systems are unable to distinguish Ga from Fe (19) . Ga disrupts Fe dependent processes because Ga 3+ cannot be reduced, and redox cycling is critical for Fe's biological functions (19) . Importantly, Ga is already approved by the Food and Drug Administration (FDA) for intravenous (IV) use. Our data shows that Ga kills P. aeruginosa (including antibiotic resistant strains), is active against biofilms, and treats 3 different model P. aeruginosa infections.

In addition, we recently completed a phase 1b clinical trial (R01 FD003704) assessing the safety and pharmacokinetics of a 5 day infusion in two dosing cohorts (100 mg/m 2 /day and 200 mg/m 2 /day) in a total of 20 clinically stable CF adults. Drug safety, sputum pharmacokinetics, and effects on lung function and sputum microbiology will be discussed. This Phase 1b trial is a first clinical step toward testing a novel approach to treat chronic pseudomonal infections of the lung.

Future work could examine Ga maintenance therapy in chronically infected patients in a manner similar to other treatments like inhaled tobramycin. Ga could also be used for treatment of infectious exacerbations, administered alone or in combination with conventional antibiotics. In either approach, Ga treatment may offer a particular advantage in the treatment of multidrug resistant infections that are known to be associated with a poor prognosis (2) . Encouraging results could also set the stage for the use of Ga in other infections, and, perhaps spur the development of other antimicrobial strategies that disrupt micro-nutrient metabolism in bacteria.

References (3, 4) . For chronic infection, aerosolized antibiotics have been demonstrated to improve lung function, reduce the frequency of pulmonary exacerbations, and improve the quality of life (2); there is also evidence for a reduction in the overall cost of health care when used as prescribed (5) .

The beneficial effects of aerosol antibiotic administration are attributed to the production of local concentrations of drug well above those that can be obtained with conventional systemic dosing (6) . These concentrations exceed the minimum inhibitory concentrations (MIC) of organisms considered resistant, allowing for bacterial killing even when the organisms exhibit high MIC levels (7) . Such sputum levels are achievable with low systemic levels, presumably reducing the potential for toxicity (8) .

The current FDA-approved formulations include tobramycin inhalation solution (TOBI), an aminoglycoside antibiotic, and aztreonam inhalation solution (Cayston), a beta-lactam antibiotic. These antibiotics were chosen for a myriad of reasons but one is their activity against Pseudomonas aeruginosa, the most prevalent pathogen in CF airways infection (9) , and known to be associated with worse lung disease (10) . The studies demonstrating their benefit included patients with P. aeruginosa in sputum cultures with the belief that this was the targeted pathogen. However, we know now that CF airways infection is far more complicated and includes many other pathogens such as Burkholderia spp., Stenotrophomonas maltophilia, Achromobacter spp., Staphylococcus aureus (including methicillin resistant, or MRSA), and anaerobes (11) .

There is a critical need for additional options for aerosolized antibiotics for patients with CF. One-third of eligible patients with CF do not use them (9) , either because of drug intolerance, treatment burden, or the perception that it is no longer effective. In addition, adherence is often low (5) presumably because of the increased burden of treatment (12) . Even when adherent to the prescribed regimen of alternate month cycling, there is further decline of lung function and recurrent pulmonary exacerbations (13) . This has convinced some clinicians that a treatment approach of continuous antibiotic usage is preferable, and/or cycling different antibiotics on a monthly basis (14) . These suggest that additional aerosol antimicrobials are needed to provide treatment options for patients with CF and chronic airways infection.

Fluoroquinolones, both oral and intravenously, have been used extensively for treatment of CF airway infections. They have potent activity against CF pathogens and are a logical option for the need of an additional class of aerosol antibiotics. A novel formulation of levofloxacin (MP-376, Aeroquin; Aptalis Pharma, Bridgwater, NJ) has been developed for the treatment of CF airway infections. MP-376 is delivered using a vibrating, perforated-membrane eFlow® nebulizer (PARI Pharma, Munich, Germany) to deliver very high concentrations in the airways while keeping serum exposure low (7, 8) . Early clinical studies of MP-376 in CF patients have demonstrated dose-related increases of levofloxacin in serum and sputum, good safety and tolerability, and evidence of clinical effectiveness compared with placebo (7, 8) .

The results of a Phase 2 randomized, placebo-controlled study assessing the efficacy, safety, and tolerability of three doses of MP-376 in a CF population were recently published (15) . One hundred fifty-one patients were randomized into the study. A notable aspect of this trial (and typical of the ongoing current MP-376 studies) is that the patients were already on a much more aggressive treatment regimen than previous studies; patients had a history of prior use of inhaled antibiotic courses in the prior year, and were also treated with other recommended therapies (2) such as dornase alfa (78%), azithromycin (74%), and hypertonic saline (46%).

As expected for an antibiotic, there was a decrease in the mean sputum density of P. aeruginosa when on inhaled levofloxacin. This occurred despite 62% of P. aeruginosa sputum isolates at baseline being non-susceptible to levofloxacin based on an MIC >2ug/mL, the breakpoint for systemic dosing with levofloxacin. MP-376 also substantially reduced the need for systemic or inhaled anti-pseudomonal antimicrobials and produced improvements in pulmonary function. The drug was found to be safe and generally well-tolerated. The most common complaint was related to the bitter taste of the medication, although no patients discontinued the study drug because of this complaint. There were no treatmentrelated serious adverse events related to MP-376.

Additional studies will soon be completed that will assess whether MP-376 will be a viable addition to the long-term treatment of chronic airway infection in CF. A Phase 3 study comparing MP-376 to placebo in stable patients (age >12 years) with CF (clinicaltrials.gov identifier NCT01180634) has completed its planned enrollment of 330 patients. The treatment duration is 28 days on medication with 28 days of follow up off medication and the primary outcome measure is time until pulmonary exacerbation. A second Phase 3 study is an openlabel, randomized trial to evaluate the safety and efficacy of MP-376 compared to Tobramycin Inhalation Solution (TIS) in stable patients with CF (clinicaltrials.gov identifier NCT01270347). It has also completed its planned enrollment of 267 patients and the study consists of 3 cycles of the standard month on/month off regimen. The primary outcomes include safety (through the entire 168 days of the study) and efficacy (as measured by the predicted FEV1 at Day 28). The results of the placebo-controlled study will be presented.

Provided successful completion of the Phase 3 studies, MP-376 may fulfill the need for an addition to the aerosol antibiotic options suggested earlier. It offers a spectrum of antibiotic activity needed for CF pathogens. In addition it represents an added class of antibiotic potentially useful for patients unable to tolerate the current classes; for those with perceived loss of benefit from the current antibiotics; and as an important partner for future studies testing a treatment strategy of rotating continuous antibiotic usage. It also offers a new antibiotic that could reduce the treatment burden, as it uses an efficient, portable, customized eFlow® nebulizer requiring only 4 to 6 minutes to nebulize a 240-mg dose. Research suggests that a neonatal diagnosis can affect parent child-relationships, but the mechanism remains elusive. This investigation used video recorded observations of parent child interactions during feedings and parent self-reports to examine factors contributing to the quality of parents' relationships with their infants 12 months after receiving normal or abnormal newborn screening results. The sample included 131 mothers and 118 fathers of 131 infants whose newborn screening and follow-up diagnostic results ranged in severity to include cystic fibrosis diagnosis (CF group, n=23), congenital hypothyroidism (CH group, n=35), CF carrier status (CF-C, n=38), or healthy with normal NBS results (H group, n=35). Based on attachment theory and previous research, we hypothesized that factors mediating the quality of parent-child interactions would include parent depression, anxiety, perceptions of child vulnerability, perceptions of child attachment and the level of task-oriented parent behavior.

We operationalized concepts using the following standardized instruments:

-Center for Epidemiologic Studies Depression Scale (1) (CES-D), a 20 item self-report -State Form of State-Trait Anxiety Inventory (2) (STAI), a 20 item self-report -Child Vulnerability Scale (3) (CVS), an 8 item self-report -Attachment Q-Sort (4) (AQS), a 90 item self-report -Task-oriented feeding item, a single observational item that has been associated with less positive and more negative parent behaviors during feeding interactions (5) -Parent-Child Early Relational Assessment (6) (PCERA), a 65 item observational instrument.

Using an exploratory factor analysis, we identified four factors with the best fit for this data set to include Factor 1: Parent Positive Affective Involvement and Verbalization; Factor 2: Parent Negative Affect, Inconsistent and Intrusive Behavior; Factor 3: Infant Positive Affect, Communication, and Social Skills; and Factor 4: Infant Dysregulation and Irritability. We obtained 5-minute video recordings of parents and their infants during typical feeding interactions in their homes.

The analysis employed latent structural equation modeling to construct separate models for mothers' and fathers' data using H group mothers' and fathers' data as a reference. Results showed that mothers and fathers of infants with CF reported significantly higher perceptions of child vulnerability (p< 0.001, p=0.002) than parents in the H group. Mothers who perceived their children as more vulnerable were significantly more likely to also perceive their children as less attached than mothers who viewed their infants as less vulnerable (p=0.001). Mothers and fathers who perceived their infants as less attached were significantly more likely to engage in taskoriented feeding behavior than parents who viewed their infants as more attached (p=0.016, p=0.029). Mothers and fathers who engaged in more task-oriented behavior were also observed to show significantly less overall positive (p< 0.001, p< 0.001) and significantly more overall negative interactions (p< 0.001, p= 0.001) with their infants than parents who used less task-oriented feeding strategies. Fathers who perceived their infants as more vulnerable showed significantly more negative overall interactions with their infants than fathers who perceived their infants as less vulnerable (p< 0.001). When mothers and fathers demonstrated high levels of affective involvement and verbalization, their infants were significantly more likely to exhibit favorable levels of affective expressiveness, communicative skills, and social responsiveness (p< 0.001, p< 0.001). Similarly, when mothers and fathers demonstrated more negative affect and/or inconsistent and intrusive behavior, their infants were significantly more likely to be dysregulated and irritable (p< 0.001, p< 0.001). As predicted there was a significant positive association between the two infant observational factors. Although mothers with lower educational levels were more likely to report depressive symptoms than higher educated mothers, a neonatal diagnosis was not associated with increased anxiety or depression in mothers or fathers relative to the health comparison group 12 months after diagnosis. Fathers of infants identified as CF carriers through newborn screening tended to view their children as significantly more attached than H group fathers (p=0.021). Except for this last result, all findings were in hypothesized directions.

In conclusion, a neonatal diagnosis of CF can lead to parent perceptions of vulnerability that warrant further investigation long-term. Clinical observations or parent reports of infant feeding problems could represent an important sign of more deeply rooted disturbances in the overall parent-child relationship that merit additional clinical evaluation.

Beth Smith, M.D.

Patients with cystic fibrosis (CF) are at increased risk for anxiety and depression (1) . In patients with CF, elevated levels of depression and anxiety have been associated with lower quality of life, respiratory function, and reduced therapeutic regimen adherence (2) (3) (4) (5) . As the survival of patients with CF has improved, there is an increasing adult population and issues around pregnancy have become important. Postpartum depression, the most common complication of childbearing, occurs in 13 percent of women after delivery (i.e., one of every eight). In postpartum depression, anxiety is more common than major depression occurring at other times (6) . Patients with CF who are postpartum therefore represent a unique population at risk for depression and anxiety.

Many new moms experience a transient 7-10 day depression referred to as the postpartum blues, which commonly includes sadness, mood swings, tearfulness, and worry. Symptoms occur during the first few days after childbirth and quickly fade within one to two weeks. As these symptoms do not impair maternal functioning and are time limited they can be treated with reassurance and emotional support. Women should be encouraged to increase their psychosocial supports and obtain help caring for the infant during this time.

Some new moms experience a more severe, long-lasting form of depression known as postpartum depression (PPD). An episode of depression with postpartum onset, as defined by the Diagnostic and Statistical Manual of Mental Disorders (DSM-IV), begins within 4 weeks after delivery. However, many epidemiologic studies define depression occurring within 3 months postpartum as PPD. The pattern of symptoms in women with PPD are indistinguishable from the characteristic major depression that can occur at other times in a woman's life. The symptoms of PPD can include depressed mood, loss of interest or pleasure in activities, insomnia, fatigue, decreased appetite, feelings of worthlessness, poor concentration, and thoughts of death or suicidal ideation. Untreated, PPD can have negative consequences for the infant including an increased risk of insecure attachment and cognitive and behavioral problems (7, 8) . Additionally, for the mother with CF, a postpartum depression can not only interfere with caretaking for the infant but may interfere with the new mom's ability to care for her CF. Addressing maternal depression is essential to ensure the healthy early brain and child development of the newborn (9) and also may promote adherence and ultimately improve maternal health outcomes.

Rarely, an extreme form of postpartum depression known as postpartum psychosis develops after childbirth. Mothers with postpartum psychosis are severely impaired and may have paranoia, hallucinations and delusions. Careful assessment for the presence of suicidal and homicidal thoughts and plans is important. This is a serious condition that requires immediate psychiatric attention.

Screening: The Edinburgh Postnatal Depression Scale (EPDS) is a 10-item questionnaire completed by the mother that can be easily administered in less than 5 minutes (10) . It is in the public domain and is freely downloadable. This screen has been found to be a reliable indicator of the presence of postpartum depression in women in the United States (11) and has been adapted and validated in many languages. Items of the scale correspond to various clinical depression symptoms, such as guilt, sleep disturbance, low energy, anhedonia, and suicidal ideation.

Overall assessment is done by total score, which is determined by adding together the scores for each of the 10 items. If a woman has a total score on the EPDS of 10 or higher or has an affirmative response to questions 10 (a suicide indicator) this constitutes a positive screen and requires a follow-up evaluation and treatment if indicated.

Although formal screening tools are available asking two questions about mood and anhedonia may be as effective as using longer instruments in adults (12) and better fit the clinical practice. For example, "Since your baby was born, have you felt down, depressed or hopeless? Since your baby was born, have you had little interest or pleasure in activities most of the day nearly every day for the past two weeks?" If the woman answers in the affirmative to either question, it is a positive screening result. The level of impairment and distress can be explored with the question, "Has the depression made it hard for you to do your work, take care of things at home, or get along with people?" Treatment: PPD screening improves recognition of the disorder, but improvement in clinical outcomes requires care that ensures adequate treatment and fol-low-up. Responses to a positive postpartum depression screening range from providing reassurance to supportive strategies for postpartum blues and mild depression, to referral for specific interventions for moderate to severe depression (Edinburgh score >13). Education is always important and lets the mother know she is not alone, she is not to blame, and that she will get better. Non-pharmacological interventions are useful in treating PPD, especially in mild illness, as an adjunct to antidepressant medication in more severe depression, and in women reluctant to use medications, for example during breast-feeding. Evidence exists for the use of both cognitive behavioral therapy and interpersonal psychotherapy in the treatment of PPD (13) . Antidepressants are often prescribed to treat PPD; however, few studies have systematically assessed the treatment specifically of PPD. For evidence based guidelines on the pharmacologic treatment of depression in adults, including PPD, see the American Psychiatric Association's practice guidelines at http//www.psych.org/guidelines/mdd2010 (14) . When concerns are significant to warrant referral mothers may be referred to their obstetricians for follow-up and to mental health providers for therapy and/or medication management. If suicidality or psychosis is a concern, or the score on the Edinburg scale is greater than 20, accessing crisis intervention services are necessary. To date, most studies in CF have found that both children and adults have higher rates of depressive and anxious symptoms (Smith et al., 2010; Riekert et al., 2007) than healthy peers and parents have also endorsed elevated psychological symptoms (Quittner et al., 1992; Driscoll et al., 2010) . To date, however, all of these studies have been limited by small sample sizes and single center recruitment. To address the limitations of prior studies, we have been conducting a national study of the prevalence of depression and anxiety in children, adolescents, and adults with CF and parent caregivers. The specific aims of this study were to: 1) use standardized screening tools to evaluate the prevalence of depression and anxiety, 2) to test the relationships between demographic and health variables and psychological symptoms, and 3) to evaluate whether recent medical events, such as IV antibiotic treatments, new diagnosis of CFrelated diabetes, or recent hemoptysis or pneumothorax are predictive of depressive and anxious symptoms.

We screened a total of 1159 patients with CF, ages 12 through adulthood (range = 12-73 years, M age = 25.3). As expected a large percentage of adolescents scored above the clinical cut-off on the CES-D (Center for Epidemiological Studies-Depression) and the HADS-Anxiety Scale. Among adolescents, 22.1% reported elevated depression and 30.2% reported elevated anxiety. This is 2-3 times the prevalence reported in healthy, community samples. In addition, although current lung function and BMI percentile were not significantly related to psychological symptoms, a recent medical crisis, such as hospitalization for IV antibiotics or a new CFRD diagnosis, was associated with scoring in the elevated range.

Based on these data, we would recommend annual screening of depression and anxiety in all patients (ages 8 and older) and parent caregivers. In addition, for those who score in the elevated range, there are several successful, evidence-based treatments for depression and anxiety, including pharmacological and cognitive-behavioral therapies. Anxiety and depression appear to be more common in patients with cystic fibrosis (CF) than in the general population. For those who are living with a chronic illness, anxiety and depression have been shown to have negative impact on health outcome. Recent studies indicated that people with CF experience symptoms of anxiety more frequently than depression (1, 2) . Data from the German arm of The International Depression/anxiety Epidemiological Study (TIDES) reported an elevated anxiety scores found in 20.6% of the patients with CF and 9.6% high level of symptoms of depression. They also found that recent hemoptysis, pneumothorax and recent diagnosis of diabetes were associated with anxiety, whereas impaired lung function and transplant listing status were associated with depression (3). Hegarty and colleagues studied the quality of life in young people with CF including the effect of hospitalization, age and gender and differences in parent/child perceptions. Their data showed that inpatients score significantly lower than outpatients for the Cystic Fibrosis Questionnaire-Revised version (CFQ-R) in the domains regarding emotional state, social, body image and respiratory symptoms (4) . Their findings clearly demonstrated that there are added stresses during hospitalization, which in turn may provide a perfect opportunity for health care providers to encourage patients to express their concerns and seek appropriate referral or support for them.

The adult CF Program at St. Michael's Hospital in Toronto follows more than 400 adults with CF on a regular basis. We have a 15 bed in-patient respiratory unit for adults with CF and other lung diseases. The majority of admissions are patients with CF at any given time. Most of them are admitted due to pulmonary exacerbation for IV antibiotics. However, other medical interventions including desensitization to antibiotics, steroid therapy, insulin regimen, oxygen or Bipap may also be initiated during the same admission. We have observed a fair amount of anxiety and fear and depression from our CF patients during their hospitalizations. According to our Picker Score from 2010, our patients were reporting that their anxiety and fear during their hospitalization were only addressed at 35% of the time. Therefore, in May 2010, as part of the corporate initiative for quality improvement, our unit nurses took on the task to reduce anxiety and fear in our hospitalized patients by utilizing the Best Practice Guideline of Establishing Therapeutic Relationship and Professionalism in Nursing. The frame-work for establishing therapeutic relationship is having knowledge about the disease, the patient, the nurse herself and the system (5). To implement the project, the unit nurses needed to know the source of patients' anxiety and fear, how they like their anxiety and fear addressed. The nurses also needed to have a standardized mechanism to communicate findings to other nurses on patient admission and shift report as well as to the interdisciplinary team on rounds. As for the patients, they need to have opportunities to convey fear and anxiety to their nurses. Both nurses and patients must have confidence in the standardized mechanism to ensure that fear and anxiety are communicated to the relevant team members and from shift to shift for appropriate interventions.

A pre-education survey was conducted to assess nurses' knowledge and comfort level in dealing with patients' anxiety and fear. A one-hour, interactive education session regarding how to establish therapeutic relationships was provided to the unit nurses. More than 80% of unit nurses completed the education session and a post-education survey, which indicated increased knowledge. Fifteen patients completed a survey to identify the source of their anxiety and fear and how their fear and anxiety were addressed. The patients identified ten major sources of their fear and anxiety. Separation from home was considered as the most anxiety provoking, followed by receiving new information about their illness, taking steroids, meeting new health care providers, separation from work, wearing oxygen for the first time, insertion of PICC line, starting on BIPAP, desensitization and finally separation from school.

To engage our patients to express their fear and anxiety, posters were placed in patients' rooms reminding them that the nurses are interested to listen, counsel, address or communicate their issues. Pocket cards containing the list of the ten major sources of anxiety were given to the unit nurses as a reminder to practice active listening for those issues. To facilitate better communication between nurses during shift changes, tape reporting was replaced by Transfer of Accountability (TOA), which requires nurses to give and receive report at the bedside and patients are encouraged to participate if they feel comfortable to do so.

By the first quarter of 2011, our Picker Score improved from 35% to 86% indicating that patients felt their fears and anxieties were being addressed. Based on chart audits, nurses' documentations relating to addressing patients' fear and anxiety had risen from 75% to 95%. Based on unit specific patient satisfaction survey, patients expressed that TOA provides good communication and opportunities to discuss their fear and anxiety during their hospital stay.

This project increased our understanding of the impact of hospitalization on patients. During the past two years, we have proven that implementing the Nursing Best Practice Guideline of Establishing Therapeutic Relationship and Professionalism in Nursing can be an effective way to help ease the added anxiety and fear experienced by our patients during their hospitalization. 

Robert Tarran, Ph.D.

The CF gene product CFTR, is a cAMP regulated Clchannel that is absent or detective in CF-affected epithelia. A key to the understanding of CF airway disease was the finding that CFTR also regulated the epithelial Na + channel (ENaC), leading to raised Na + and fluid absorption in CFTR's absence (1) . This initial discovery that CFTR can regulate ENaC was made over 25 years ago. However, the mechanism of CFTR's influence of ENaC remains obscure. Both CFTR and ENaC reside in the apical membrane of airway epithelia and can be coimmunoprecipitated from native airway epithelia, which suggests a physical interaction, although a specific mechanism remains to be outlined (2) . CFTR may also indirectly regulate ENaC through soluble reporter molecules that are present in the airway surface liquid (3). This latter mode of regulation is variable, and may be lost under some experimental conditions. Pending a mechanistic explanation for the physical regulation of ENaC by CFTR, the possibility of indirect regulation will be discussed below.

Mucus clearance is an important aspect of the mammalian lung's innate defense system. During normal mucus clearance, inhaled pathogens become trapped in the mucus layer and are expelled before they can colonize the airways. The mucus layer is kept away from underlying epithelia by the presence of a ~7 µm periciliary liquid layer (PCL), which acts as a lubricant for the mucus. Together, these layers make up the airway surface liquid (ASL). The rate of mucociliary clearance is strongly influenced by the hydration state of the ASL. ASL volume hydration is largely maintained by Clsecretion through CFTR, with Na + and H 2 O following. However, as ASL moves up the respiratory tract, due to a reduction in surface area as airways converge, excess ASL is absorbed when needed via isotonic Na + -led ASL absorption, which is mediated by ENaC. Our experimental data indicate that ASL volume and ion transport regulation are in dynamic equilibrium, and that either a reduction in CFTR-mediated Clsecretion or an increase in ENaC-mediated Na + absorption will lead to PCL depletion and mucus dehydration. However, the contribution of ENaC dysregulation to chronic CF lung disease is currently controversial. The extracellular loops of α and γ ENaC can be proteolytically cleaved at multiple sites by intracellular furin-type convertases and extracellular serine proteases, leading to activation of the channel and increased Na + absorption. Based upon studies of well differentiated nasal and bronchial cultures grown at the air liquid interface, we conclude that restraint of proteolytic cleavage of ENaC is an essential part of normal airway physiology and vital for limiting Na + absorption rates. Thus, while our data from CF nasal and bronchial epithelial cultures show that ENaC is hyperactive in comparison to normal nasal/bronchial epithelia, bypassing the spontaneous suppression of ENaC cleavage seen in normal airways also leads to ASL volume depletion.

The realization that these ion channels are highly sensitive to soluble reporter molecules that are present in the ASL was crucial to our understanding of how CFTR and ENaC are regulated. These reporter molecules act as feedback sensors to modulate ASL volume. Importantly, these reporter molecules are contained within the ASL and are lost when cultures are mounted in Ussing chambers or following other experimental protocols that require flooding of the mucosal surface. Thus, the study of normal and CF airway epithelia under thin film conditions is essential to understanding the regulation of CFTR and ENaC and studies of airway epithelia under flooded conditions may produce aberrant results. In contrast, thin film studies correlate well with the increased cleavage pattern seen with ENaC in CF airways biochemically (2) .

Adenosine is one such reporter molecule that regulates CFTR activity and removal of adenosine leads to inhibition of CFTR and ASL volume depletion in normal airways. In contrast, adenosine levels are not different in CF airways, but the effector (i.e., CFTR) is absent or defective and so CF airways are unresponsive to changes in adenosine. Using a proteomic screen, we recently identified SPLUNC1 as an ASL volume-sensing molecule that acts as a potent negative regulator of ENaC that serves to prevent excessive volume absorption when ASL volume is at optimum levels (4). Removal or dilution of SPLUNC1 leads to increased ASL absorption and temporary ENaC hyperactivity in normal airway culturesuntil SPLUNC1 levels rise again. For reasons that are currently under investigation, SPLUNC1 is present in CF ASL but is unable to regulate ENaC, suggesting that indirect regulation of ENaC by reporter molecules is defective in CF airways. To better understand defective SPLUNC1 regulation in CF, we identified the ENaC inhibitory domain of SPLUNC1 and further proteomic analysis suggested that this domain may be functionally deficient in CF airways.

Finally, after the PCL has been depleted in CF airways, mucin secretion continues, leading to an increase in the ASL mucin concentration. Due to a salt/water imbalance, this mucus is sufficiently hydrated to increase ASL height but too dehydrated to be functionally cleared (i.e., 10-20% solids in CF vs.~2% solids in non-CF). Thus, we predict that continued mucin secretion in the absence of a functional PCL leads to a greater ASL height in CF compared to normal airways, but with decreased water availability (e.g., dehydration) and reduced ability to be cleared. We propose that this "growing" of dehydrated mucus in CF airways following Clhyposecretion and Na + hyperabsorption leads to mucus plug formation in CF airways. Failure of CF ion transport is likely heterogeneous, reflecting the heterogeneity of CF lung disease, and in some places this may be partially rescued by alternate Clchannels such as the Ca 2+ activated Clchannelat least until an insult such as a viral exacerbation overwhelms the ion transport capacity and tips the balance toward mucus dehydration. We conclude that CFTR may interact with ENaC in multiple ways and that dysregulation of either channel can lead to PCL collapse and mucus dehydration. Thus, any therapies aimed at removing mucus should be directed to normalizing ASL hydration levels and should begin as early as possible after birth.

Funded by the NIH/NHLBI.

Michael Gray, Ph.D.

Despite the growing awareness of the importance of HCO 3 in cystic fibrosis (CF) pathophysiology (1, 2) the mechanism of HCO 3 secretion in CF-affected epithelial tissues is still not fully established, particularly in the airways. HCO 3 is an important component of epithelial secretions and, via its buffering role, controls the pH of the luminal microenvironment. Aberrant HCO 3 secretion is likely to contribute to CF lung pathogenesis as efficient secretion and solubilisation of secreted macromolecules (such as mucus) is a pH-dependent process. Furthermore, abnormal pH/[HCO 3 -] negatively impacts mucus hydration and viscosity, bacterial killing and ciliary beat frequency, that collectively would predispose the lungs to mucus blockage, bacterial infection, and disease (3) (4) (5) .

Previous work in the pancreas, small intestine, reproductive tract and salivary glands have shown that CFTR regulates HCO 3 secretion in these tissues in at least two ways: (i) directly by conducting HCO 3 ions and (ii) indirectly through regulation of members of the SLC26A family of apically-located anion exchangers. Loss of CFTR could therefore lead to aberrant HCO 3 secretion via reduced Cl -/HCO 3 exchange activity, as well as through reduced HCO 3 efflux through CFTR. We have been investigating HCO 3 transport in the Calu-3 cell line, which is used as a model of serous cells of human submucosal glands (SMGs). Our recent results (6) from these cells show that Cl -/HCO 3 exchange makes a major contribution to transcellular HCO 3 transport. Real time measurement of intracellular pH from polarized cultures of Calu-3 cells demonstrate cAMP/PKA-activated Cl --dependent HCO 3 transport across the luminal membrane via CFTR-dependent coupled Cl -/HCO 3 anion exchange. The pharmacological and functional profile of the luminal anion exchanger was consistent with the activity of SLC26A4 (pendrin), which was shown to be expressed by quantitative RT-PCR, Western blot and immunofluorescence. Pendrin-mediated anion exchange activity was confirmed by shRNA pendrin knockdown (KD), which markedly reduced cAMP-activated Cl -/HCO 3 exchange. Furthermore, pendrin activity was dependent on ion transport by CFTR, suggesting that lack of functional CFTR in CF would depress pendrin-mediated HCO 3 secretion. To establish the relative roles of CFTR and pendrin in net HCO 3 secretion, transepithelial liquid secretion rate and liquid pH were measured in wild type, pendrin KD and CFTR KD cells. cAMP/PKA increased the rate and pH of the secreted fluid. Inhibiting CFTR reduced the rate of liquid secretion but not the pH, whereas decreasing pendrin activity lowered pH with little effect on volume. These results establish that CFTR predominately controls the rate of liquid secretion while pendrin regulates the composition of the secreted fluid and identifies for the first time a critical role for this anion exchanger in transcellular HCO 3 secretion in airway serous cells. Physiologically, CFTR-regulated Cl -/HCO 3 exchange in airway SMGs may be important in maintaining the pH of the ASL, which could therefore explain the acidic pH reported in CF airways (7) (8) (9) . In the context of SMGs and CF, lack of pendrin-mediated HCO 3 secretion would therefore predispose the glands to mucus blockage and damage, a characteristic sign of CF airway disease and an important mediator of lung pathophysiology. Therefore, restoring ASL pH through a CFTR-independent, pendrin-mediated, mechanism could represent a potential therapeutic target in the treatment of CF, and current research is directed at this possibility. Furthermore, since pendrin expression has been linked to enhanced mucin gene expression in response to inflam-matory cytokines (10) , suggests that pendrin has multiple roles in airways physiology. Overall, our work suggests that a better understanding of pendrin function and dysfunction in the airways is vital for the development of future therapies for CF airway disease. Several members of the SLC26A family of anion transporters have been found to associate with CFTR and participate in a coordinated regulation of both CFTR function and SLC26A activity (1) (2) (3) (4) . Since the different SLC26A family members show significant diversity in function, substrate, and tissue specificity (5), their coordinated regulation by CFTR may confer the unique anion permeability and transport characteristics of the different organs that are affected by CF. Our lab has recently char-acterized the activity of SLC26A9, found in lung and gastric epithelia, and its coordinated regulation by wt CFTR (2) . In HEK293 cells, SLC26A9 exhibits chloride channel-like properties, is inhibited by CFTR channel blockers, and interacts with CFTR to enhance both constitutive and cAMP/PKA-stimulated chloride secretion. In primary, human bronchial epithelia (HBE) from non-CF donors, constitutive anion secretion bears the pharmacological fingerprint of SLC26A9. Importantly, SLC26A9 activity is absent in HBE from ∆F508 donors even though SLC26A9 mRNA is abundant, suggestive of an interaction between CFTR and SLC26A9 that commences at an early stage of protein maturation. In support of this, SLC26A9 immunoprecipitates (IPs) contain both immature (band B) and mature (band C) wt CFTR.

We have examined whether the loss of SLC26A9 constitutive activity in HBE expressing ∆F508 CFTR reflects a loss of CFTR trafficking or function. HEK293 cells expressing SLC26A9 ± wt CFTR exhibit similar constitutive currents (22 ± 6 pA/pF, n=5, vs. 25 ± 3 pA/pF, n=19). Similarly, co-expression of SLC26A9 with the functionally impaired mutant G551D CFTR results in normal constitutive SLC26A9 currents (28 ± 8 pA/pF, n=13). When SLC26A9 is co-expressed with ∆F508 CFTR, however, the constitutive current is significantly reduced (10 ± 3 pA/pF, n=8, p<0.005). Therefore, reduced SLC26A9 constitutive activity correlates with a loss of mutant CFTR trafficking. Our studies indicate that SLC26A9 does not require co-expression with CFTR to exhibit constitutive activity (2) , which led us to investigate whether co-expression with ∆F508 CFTR altered the expression of SLC26A9 at the plasma membrane.

In HEK293, SLC26A9 protein expression was observed when co-expressed with wt, ∆F508, or G551D CFTR, indicating that loss of CFTR maturation or function does not inhibit SLC26A9 protein synthesis. Indeed, co-expression of SLC26A9 with CFTR increased total protein expression for both, compared to the protein levels of the individually expressed channels. As with wtCFTR, SLC26A9 co-IPs both band B and C G551D CFTR, as well as band B ∆F508 CFTR, confirming that the proteins interact at an early stage of protein maturation. The interaction between SLC26A9 and ∆F508 CFTR does not appear to inhibit maturation of SLC26A9, however, as glycosidase assays reveal that the predominant form of SLC26A9 observed in whole cell lysates from HEK293 expressing SLC26A9 ± CFTR is complex glycosylated. Similarly, biotinylation assays indicate that the levels of SLC26A9 at the plasma membrane are not substantially reduced when co-expressed with ∆F508 CFTR. Consequently, the suppression of SLC26A9 constitutive activity when co-expressed with ∆F508 CFTR does not appear to result from a direct interaction, as the proteins are spatially distinct. Rather, these data suggest that additional proteins comprising the CFTR interactome can also influence SLC26A9 activity.

These results demonstrate that CFTR is able to exert significant influence on the constitutive activity of SLC26A9. Our results have also shown that SLC26A9 exerts a strong influence on the cAMP/PKA-stimulated activity of wtCFTR, as evidenced by a synergistic increase in the forskolin-stimulated current observed during whole cell patch clamp of HEK293 co-expressing SLC26A9 + wtCFTR (2) . However, since both SLC26A9 and CFTR conduct anions, and we lack a spe-cific inhibitor of SLC26A9, determining the mechanism of the observed synergism is challenging. In experiments with SLC26A3 and SLC26A6, Ko, et al. (1) found that the STAS (sulfate transporter and anti-sigma antagonist) domain of these SLC26A family members interacted with the R-domain of wtCFTR, and that co-expression of the STAS domain alone with wtCFTR was sufficient to potentiate the single channel activity of CFTR. We utilized a similar approach, using a STAS domain construct of SLC26A9 (A9 STAS). HEK293 co-expressing A9 STAS + wtCFTR exhibited a forskolin stimulated current that was similar to cells co-expressing SLC26A9 + wtCFTR (-300 ± 39 pA/pF, n = 5, vs. -282 ± 28 pA/pF, n = 18, p=0.4) and significantly greater than cells expressing wtCFTR alone (-197 ± 34 pA/pF, n = 26, p<0.05). Cells expressing A9 STAS + wtCFTR did not exhibit constitutive activity (-2 ± 1 pA/pF, p<0.001), consistent with a lack of SLC26A9's conductive pathway. In excised, inside-out patch clamp of CHO cells expressing wtCFTR ± A9 STAS, the presence of A9 STAS significantly increased the single channel conductance of wtCFTR upon PKA+ATP activation (11.5 ± 1.1 pS, n = 10 vs. 9 ± 1 pS, n = 25, p<0.05). Thus, the observed synergy appears to result, in part, from an increase in the conductance of wtCFTR through an interaction with the STAS domain of SLC26A9.

Summary: Unique members of the SLC26A family of transporters have been found in all tissues affected by CF including the lung, pancreas, and intestine. Our results demonstrate that an interaction with the STAS domain of SLC26A9 (lung) increases the single channel conductance of wtCFTR, while others have shown that an interaction with the STAS domain of SLC26A3 (pancreas, intestine) increases the open probability of wtCFTR (1); in addition, both increase the number of wtCFTR channels at the plasma membrane. How these discreet changes in CFTR activity contribute to the unique anion transport functions of these organs is currently unclear; however, it is clear that even the microscopic properties of CFTR can be modified through interactions specific to the epithelia in which it is expressed. Our results have also demonstrated that co-expression of trafficking deficient ∆F508 CFTR significantly mutes the constitutive activity of SLC26A9, while the functional mutation G551D CFTR or the absence of CFTR does not. Others have shown that specific mutations in CFTR alter the properties of SLC26A3 and A6 (6-7). These observations suggest that SLC26A family members may be advantageous as drug targets for improved function in CF. Finally, knowledge of the interactions between SLC26A9 and ∆F508 CFTR may inform strategies to rescue mis-folded CFTR.

[ Mucus obstruction and airway inflammation are key pathological features of the chronic lung disease that continues to determine most of the morbidity and mortality in patients with cystic fibrosis (CF). Previous studies in human primary airway epithelial cultures and transgenic mouse models demonstrated that depletion of airway surface liquid (ASL) caused by impaired CFTR-mediated Clsecretion and increased Na + absorption mediated by epithelial Na + channels (ENaC) constitutes an important disease mechanism in the CF lung (1-3).

Besides restoration of CFTR function, activation of alternative non-CFTR Clchannels may improve airway surface hydration in the CF lung. Recent evidence suggests that epithelial Clsecretion is increased in Th2-mediated airway inflammation, however, the molecular identity of the underlying Clconductance was not known (4). SLC26A9 is a member of the SLC26 family of anion transporters predominantly expressed in epithelia of the lung and upper GI tract (5) . Recent studies showed that the SLC26A9 protein functions as a Clchannel that contributes to constitutive and cAMP-dependent Clsecretion in human bronchial epithelial (HBE) cells, where it has been reported to interact functionally with CFTR (6). Based on its functional properties in transduced cells and expression pattern in mouse airways, we hypothesized that SLC26A9 may function as an alternative Clchannel that may contribute to ASL homeostasis in health and in Th2-mediated airway inflammation.

To test this hypothesis, we compared transepithelial ion transport in freshly excised bronchial tissues, lung morphology and airway mucus content in wildtype versus SLC26A9-deficient mice under physiological conditions and after intratracheal instillation of IL-13 to model Th2-mediated airway inflammation. Further, in a human study population of 661 children with asthma (i.e. a prototypical Th2-dominated airway disease) and 658 healthy controls, we tested if polymorphisms in the SLC26A9 gene are associated with asthma.

In bioelectric studies in native airway tissues, lack of SLC26A9 had no effect on basal, cAMP-mediated or Ca 2+activated Clconductance (CaCC) under physiological conditions. In airways from wild-type mice, constitutive as well as Ca 2+ -activated Clsecretion were~2-fold increased after intratracheal instillation of IL-13 as expected from previous studies (4). The IL-13 induced constitutive Clconductance was completely abrogated, whereas up-regulation of CaCC was preserved in native airway tissues from SLC26A9-deficient mice. While IL-13 induced goblet cell metaplasia and mucus overproduction to similar levels in both genotypes, lack of SLC26A9-mediated Clsecretion was associated with airway mucus obstruction in IL-13 treated SLC26A9deficient mice that did not occur in wild-type controls.

Testing of effects of genetic variants within the SLC26A9 locus on asthma risk identified several single nucleotide polymorphisms (SNP) in the 3'UTR of SLC26A9 (rs12031234, rs2282429, rs2282430) that were associated with asthma. In silico analyses predicted that the A allele of rs2282430 strengthens binding of hsa-miR-632 to the SLC26A9 3'UTR. To determine functional relevance, we used luciferase reporter assays and studied effects of precursor hsa-miR-632 on the wild-type 3'UTR sequence and the SNP in HEK293 cells. In these studies, transfection of precursor hsa-miR-632 reduced luciferase activity significantly in cells cotransfected with the SNP compared with the wild-type 3'UTR. Collectively, these results suggest that the SNP in the SLC26A9 3'UTR identified in our genetic studies reduces protein expression and indicate that binding of hsa-miR-632 is involved in this process.

Taken together, the results of these studies demonstrate that SLC26A9 constitutes a Clchannel in native airway epithelia that is quiescent under physiological conditions, but constitutively active to promote Clsecretion and hydration of airway surfaces in Th2-mediated airway disease. The observation that genetic deletion of SLC26A9 produced significant airway mucus obstruction in allergic airway disease indicates that SLC26A9-mediated Clsecretion is essential for maintaining ASL homeostasis required for effective MCC in the presence of mucus hypersecretion. In this context, it is noteworthy that up-regulation of transiently active CaCC, was not sufficient to prevent airway mucus obstruction in Th2-mediated airway disease. These results are consistent with observations in patients with CF, in whom increased CaCC-mediated Clsecretion is not sufficient to compensate for deficient CFTR Clchannel function and prevent dehydration-induced mucus plugging and airway disease (3). Therefore, as an alternative Clchannel that supports sustained Clsecretion, SLC26A9 may also serve as a novel therapeutic target to counteract impaired epithelial Clsecretion in CF. 

Cystic fibrosis (CF) is caused by a non-functional chloride and bicarbonate ion channel (CFTR), but the link to the phenomenon of stagnant mucus is not well understood. In contrast to previous ideas, mucin is packed in the mucin granulae of goblet cells in a highly organized way by calcium and low pH mediated interactions of the N-terminal part of the mucin. This will allow an organized release that is triggered by removal of calcium ions, something that bicarbonate is necessary for (1). Mice lacking functional CFTR (Cftr∆508) have no lung phenotype, but show similar ileal problems as humans. The ileal mucosa in CF have a mucus that adhered to the epithelium, was denser, and was less penetrable than that of wild-type mice (2) . The properties of the ileal mucus of CF mice were normalized by secretion into a high concentration sodium bicarbonate buffer (about 100 mM). In addition, bicarbonate added to already formed CF mucus almost completely restored the mucus properties. This knowledge and experimental system used allow novel ways of testing CF therapies and suggest more therapeutic options for a successful CF lung therapy. Cystic fibrosis (CF) is an autosomal recessive inherited disease caused by mutations in the CF transmembrane conductance regulator (CFTR) gene. Despite significant advances in our understanding of CF disease pathogenesis and therapeutic interventions, CF remains a potentially lethal disease. This is primarily due to chronic pulmonary bacterial infections and inflammation (1) . In order to develop and investigate new therapeutic strategies for CF treatment, we need to further our understanding of early CF disease pathogenesis. However, these studies are difficult to perform in infants and young children. These early studies would be more feasible in a CF animal model that develops pulmonary disease similar to humans with CF.

We chose pigs for the development of a CF model, because they are similar to humans in terms of anatomy, physiology, immune system, biochemistry, life span, size, and genetics (2) . At birth, CFTR-/-porcine lungs show no evidence of inflammation in bronchoalveolar lavage fluid, histopathologic analysis, and microarray studies for inflammatory pathways. However, within several months these animals develop lung disease with features similar to human CF lung disease including bacterial infection, inflammation, mucus accumulation, airway remodeling, and airways obstruction. Furthermore, within hours after birth, CFTR-/-piglets have a bacterial host defense defect (3) (4) .

Knowing that CF pigs develop lung disease like humans with CF and that there is a lung host defense defect present at birth, we have focused our studies on the airway surface to better understand the pathogenesis of early CF lung disease. First, we have found that lack of CFTR in the porcine airway leads to more acidic airway surface liquid (ASL) -an observation similar to earlier reports (5) (6) (7) (8) . Furthermore, the more acidic ASL in CF porcine airways impairs bacterial host defense and this defect can be restored with topical bicarbonate treatment. Second, the viscosity of methacholine-stimulated secretions from CF porcine airways is greater than non-CF. These findings suggest that lack of CFTR alters the physiochemical properties of airway secretions and these abnormalities are likely related to lack of CFTR-mediated Cland/or HCO 3 transport. Finally, we are currently investigating the role of CFTR in mucociliary transport in newborn non-CF and CF pigs, prior to the onset of infection and inflammation.

These studies are increasing our understanding of the early events in the pathogenesis of CF lung disease and will hopefully lead to the development of new therapeutic tools for CF. Cystic fibrosis (CF) is caused by mutations in the cystic fibrosis transmembrane conductance regulator (CFTR), an anion transporter of chloride and bicarbonate. Defective or deficient CFTR leads to severe mucoobstructive lung disease and severe morbidity and early mortality due to lung infections. Even though substantial advances in our understanding of CF pathogenesis have been made, we still do not fully understand the pathophysiologic mechanisms that lead to decreased mucociliary clearance (MCC) in CF patients. For example, whether absent CFTR function results in depletion of the airway surface liquid (ASL) in human airways, and whether this explains early deficits in MCC that lead to chronic infection is actively debated (1) . More recent studies in CF pigs have suggested the ASL "low volume" model may not be operative in vivo (2, 3) . These results emphasize the importance of accurately measuring ASL depth in vivo, so that the pathophysiologic relationship between ASL/periciliary layer (PCL) regulation and MCC can be defined. In contrast, innate abnormalities in CF mucus have long been postulated to contribute to the disease (e.g., mucoviscidosis), but whether this is a primary phenomenon, or a result of downstream consequences of the CF defect, also remains unknown (4, 5) . Resolving these issues is imperative, as discerning the nature of the innate CF defect in vivo could resolve a longstanding controversy and result in new therapeutic opportunities that address operative pathophysiology. This knowledge could also lead to new insights regarding the onset and progression of CF lung disease in children, including the small airways.

Two critical barriers underlie the roadblock in our understanding of CF pathogenesis in early lung disease: 1) there have been no tools available for visualizing the important microstructural, functional, and biomechanical features of the respiratory mucosa and mucus in vivo; and 2) it has been difficult to study young CF patients to determine the fundamental defects of this disease, prior to the occurrence of secondary phenomena such as infection and inflammation that confound the study of CF pathogenesis. Recently, our laboratories have developed an innovative technology, termed 1-µm resolution optical coherence tomography (µOCT), which enables real-time cross-sectional microscopy of the functional epithelial surface of living airways and can be readily combined with measures of ion transport in vivo. With µOCT, we have been able to simultaneously and quantitatively monitor ASL and PCL depths, ciliary beating (CBF), and mucociliary transport (MCT) in situ, while also measuring mucus viscosity by native particle tracking techniques without the use of contrast dyes, exogenous microparticles, or other manipulations.

In cell culture studies, we have shown that CF HBE cultures exhibited reduced ASL and PCL depths, CBF, and MCT compared to non-CF cultures, and validated µOCT measurements of each of these parameters using "gold standard" methods. In preparation for the evaluation of MCC by µOCT in vivo, we evaluated intact, full-thickness swine trachea with µOCT and have measured the dynamic response of the functional airway in relation to modulation of ion transport. Results were consistent with findings in human cells and the MCC of native trachea measured by tracking visible particles (6) . µOCT also sensitively detected changes in ASL, CBF, and MCT upon addition of CFTR modulators.

Previous studies have shown that viscosity of sputum and mucus can be monitored by exogenous particle tracking microrheology, a technique in which the mean squared displacements (MSD) of fluorescent microparticles are measured and converted to viscosity by the generalized Stokes-Einstein relation (GSER) (7) . Since similarly sized natural particles or inclusions present in mucus scatter light and are visible using µOCT, and because µOCT has the resolution to determine the sizes of these natural particles (through straightforward deconvolution procedures), we have been able to recapitulate estimates of MSD by tracking endogenous microparticles within mucus. These results demonstrate the potential for µOCT to measure the mechanical properties of unaltered mucus in situ. By evaluating the significant contributors to MCC and CFTR function in the same location without perturbing the airway, µOCT imaging provides a major advantage over current techniques, and forms the potential basis for the first measurements of these parameters in living human subjects.

Presently, we are advancing this technology by building an improved imaging system and a novel pulmonary catheter that will enable µOCT to be used in the airways of living human subjects. In addition, potential difference (PD) electrodes will be integrated within the sheath of the µOCT catheter, enabling colocalized measurements of CFTR ion channel activity. This device will then be employed to investigate the airways of young children with CF, prior to the onset of structural lung disease. We hope to define the earliest events that initiate CF pathogenesis, including the relationship of ASL regulation to mucociliary transport (MCT) and mucus biogenesis while also determining the roles of ion transport towards regulating these pathways. In addition to answering important questions regarding CF pathophysiology, this work will provide a clear path toward the use of µOCT for monitoring functional responses to novel therapeutics.

References Aerosol-based imaging methods provide an opportunity to non-invasively quantify airway epithelial physiology in vivo through the use of radiolabeled probes. Methods for measuring mucociliary clearance have been previously described (1, 2) . These techniques involve the inhalation of a radiolabeled particulate in liquid aerosol. Gamma camera images collected after aerosol delivery are then used to quantify particle clearance over periods from 60 minutes to 24 hours. These studies have provided important mechanistic proof of principle for treatments such as inhaled hypertonic saline (3) and mannitol (4) . Here we explore a novel aerosol-based technique to detect changes in airway liquid absorption. This technique uses multiple radiopharmaceutical probes delivered together in a liquid aerosol to the airway epithelium. We hypothesize that the absorption rate of a hydrophilic small-molecule probe (Indium 111-DTPA) will be influenced by paracellular liquid transport, and will provide a gauge of the total liquid absorption rate from the airway when stimulated by liquid volume addition. The clearance rate of the small-molecule probe will be influenced by mucociliary transport as well as absorption. The mucociliary contribution is estimated by simultaneously measuring the clearance rate of non-absorbable sulfur colloid particles radiolabeled with technetium 99m. The lung distribution patterns and clearance rates of the two probes can be quantified independently based on the different energy levels associated with In111 and Tc99m. We have studied the relationship between DTPA and liquid absorption rate using primary human bronchial epithelial (HBE) cell cultures from CF and non-CF lungs. We have also compared DTPA absorption rates in the lungs of CF patients and healthy controls (5) . Studies to determine DTPA absorption response to a well characterized therapy (7% hypertonic saline) are on-going as are studies of DTPA absorption in pediatric CF subjects and healthy controls.

In vitro observations: Multiple factors are likely to influence the absorption rate of DTPA. No previous evidence for a transcellular pathway for DTPA absorption has been described, so a paracellular path is assumed. Tight junction pore diameter and DTPA concentration in the ASL could influence diffusion across the epithelium. We propose that DTPA absorption can occur directly through paracellular liquid transport if a stimulus for liquid absorption is applied, and that the airway liquid hyper-absorption characteristic of CF will be reflected in increased DTPA absorption rates. To test this hypothesis we measured the DTPA absorption rate in 5 CF and 7 non-CF HBE cell lines (n=6 cultures/line) after the addition of 10 µL of 10 µCi/mL technetium 99m labeled DTPA (Tc-DTPA) in Ringer's. The DTPA absorption rate was significantly increased in the CF HBE cells (54.9±8.6 vs. 32.1±7.3 %/24hrs; p<0.001). Using a previously described optical method for measuring airway surface liquid (ASL) volume (6) we compared liquid and DTPA absorption rates in CF HBEs after the addition of 2.5, 5 and 10 µL of 10 µCi/mL Tc-DTPA. There was a linear relationship between the amount of DTPA absorbed (based on radioactivity) and the ASL volume absorbed for all three volumes tested (R2=0.96, 0.99, and 0.56 respec-tively). Both the DTPA and liquid absorption rates increased linearly as the volume added to the epithelial surface was increased. Previous studies have demonstrated increases in ENaC current in response to ASL volume expansion that would be expected to increase liquid absorption (7) . We also measured DTPA absorption after the establishment of several different transepithelial osmotic gradients. Basolateral mannitol (150 and 300 mM) increased DTPA absorption from the apical surface in a dose-dependent, linear fashion. Apical mannitol (300 mM) decreased but did not prevent the absorption of DTPA (28.6 ± 8.7 vs. 45.9 ± 13.2 %/24 hrs; p=0.005) in CF HBEs, suggesting a diffusive component related to concentration gradient. Notably, the DTPA absorption rate of the CF cells treated with apical mannitol approximates that of non-CF cells. Finally, we sought to determine the relationship between transepithelial electrical resistance (TER) and DTPA absorption rate in CF and non-CF cells. TER was measured using a transepithelial volt-ohmmeter in 30 individual non-CF and 12 CF HBE cultures 24 hours prior to DTPA absorption measurements. While the average TER measured in the CF cells was lower than that of the non-CF cells, there was no relationship between TER and DTPA absorption rate across an extensive range of TER values. DTPA absorption was significantly increased in CF cells when compared to non-CF cells with similar TER levels.

In vivo observations: Our previous studies demonstrated increased rates of DTPA absorption in whole and central lung zones in CF subjects using the multiprobe aerosol method previously described (5) . Current studies are being conducted to assess DTPA absorption response to inhaled hypertonic saline in adult CF subjects and baseline DTPA absorption in pediatric CF and adult control subjects. The figure illustrates characteristic retention curves for the particle and small molecule probes in an adult CF subject and a non-CF control. The gap between the curves (total DTPA clearancemucociliary clearance) represents our measure of absorptive clearance. Summary: We have characterized a novel aerosolbased imaging method to assess liquid absorption across airway epithelium in vitro and in vivo. Studies to date indicate that the absorption of DTPA reflects the basic CF fluid transport abnormality, hyper-absorption, and that this technique will be useful to assess the efficacy of novel therapeutics that seek to correct dysfunctional CFTR and/or modulate airway fluid absorption. 

Anne Stephenson, M.D., Ph.D.

Vitamin A is a fat-soluble vitamin which influences vision, cell differentiation, immune modulation and bone health. Due to fat malabsorption, individuals with cystic fibrosis (CF) are at increased risk for vitamin A deficiency therefore, vitamin supplementation is standard therapy for this patient population. Clinical signs and symptoms of deficiency include xerophthalmia and night blindness, impaired immune function, and dermatologic conditions. More recently, literature suggests that chronic intake of vitamin A in CF may increase the risk of toxicity, particularly following transplantation. Vitamin A toxicity is an established cause of increased intracranial pressure, hepatocellular injury, decreased bone mineral density and fractures.

Vitamin A terminology and units: Confusion exists surrounding the terminology related to vitamin A. Two main forms of vitamin A exist: preformed vitamin A (i.e., retinol), which is the active form of the vitamin, and provitamin A (i.e., carotenoids), which is a precursor of retinol. Beta carotene is the most well-known carotenoid however several others carotenes exists. Typically we ingest preformed vitamin A from animal sources (liver, egg yolks, etc.) whereas provitamin A is found in plant sources (green leafy vegetables, carrots). Oral vitamin A supplements can contain both the active preformed vitamin A (retinol) as well as the provitamin A which is converted to retinol within the body once ingested. Provitamin A is less potent than retinol. In order to understand the physiologic impact of vitamin A supplementation, one needs to calculate the total amount of active vitamin A by converting the provitamin A to its retinol equivalent and adding it to the preformed vitamin A that is ingested. The units associated with vitamin A can be expressed as international units (IU), micrograms (µg) or retinol activity equivalents (RAE).

To convert µg/dL to µmol/L multiply by 0.0349; to convert µmol/L to µg/dL divide by .0349

How do we measure vitamin A stores in the body and what affects retinol levels? Retinol is absorbed through the small intestine, stored in the liver and exported to tissues bound to retinol binding protein (RBP). Individuals with CF are monitored for vitamin A deficiency by measuring serum retinol levels using high performance liquid chromatography. Retinol is the preferred measurement to assess vitamin A status over its derivatives because retinol concentrations are tightly controlled and will not drop until body stores are significantly compromised. In comparison, controversy exists over whether serum retinol levels accurately reflect vitamin A stores in the setting of excess. The ideal measure of body stores of retinol is to assess the liver content however there is no non-invasive way to accurately measure this, short of a liver biopsy. Dietary protein, energy, and zinc are required for synthesis of RBP therefore in severe malnutrition, serum retinol levels may be falsely low due to decreased RBP production. In addition, RBP acts as a negative acute phase reactant therefore vitamin A levels measured during acute inflammatory illnesses may yield misleading results.

The specific formulation of vitamin A supplements affects the risk of toxicity. Provitamin A is highly regulated and selectively converted to retinol. When retinol levels increase, the activation of provitamin A in the gut decreases due to negative inhibition resulting in less available retinol for absorption. Conversely, retinol or active vitamin A, is efficiently absorbed and lacks the negative feedback mechanism increasing the risk of toxicity.

What is the appropriate dose of vitamin A supplementation in CF? In the 2002 CFF Consensus report on nutrition (1), it is recommended that vitamin A levels be measured annually and when serum levels are low, supplementation is initiated. The recommended supplementation varies depending on age. Some concern has been raised recently about elevated retinol levels with chronic supplementation in CF. A cross-sectional study of vitamin A intake and serum retinol levels from 73 pre-adolescent children with CF was compared with healthy age-matched controls (2) . The vitamin A supplementation was higher in the CF cohort as might be expected since vitamin A supplementation is standard therapy in CF. Mean serum retinol levels were also higher in the CF group compared to healthy controls although both groups remained in the normal range. Forty-seven percent of CF subjects, however, had serum retinol values above the 95th percentile reference range. A second study published in 2008 (3) assessed vitamin intake and serum levels in a cohort of children and young adults (mean age 14.5 yrs) with CF compare to healthy controls. In this study 58% of CF subjects had serum retinol levels above the upper limit of normal. Elevated serum retinol levels appear to be even more prevalent post-lung transplantation. Two published studies have shown that retinol levels significantly increase, frequently above the upper limit of normal, following transplantation despite no change in oral supplementation (4, 5) . This finding is not specific to CF as it was also found in those transplanted for reasons other than CF.

Vitamin A and outcomes in CF: Few published studies exist relating serum retinol levels to outcomes in CF. Hakim et al (6) published a study in 2007 showing a negative relationship between serum vitamin A levels and frequency of pulmonary exacerbations. Although levels remained in the normal range, lower levels, were associated with an increased rate of pulmonary exacerbations. Studies in the non-CF population have shown a correlation between bone disease/fracture risk and excess oral vitamin A intake as well as elevated serum retinol levels. One study found that serum retinol levels in the highest quintile (> 2.64 µmol/L) had a higher risk of any fracture (RR=1.64) and hip fracture in particular (RR=2.47) (7) . When serum levels were greater than 3.6 µmol/L, the RR increased to 7.14 for any fracture. Previous studies have shown that 77% of post transplant subjects had levels greater than 2.64 and almost 40% had levels greater than 3.60 µmol/L suggesting a large proportion of post transplant patients are at increased fracture risk from hypervitaminosis A (4).

References (1) for vitamin E (vit E), this family of compounds primarily functions as antioxidants and prevents lipid peroxidation and free radical formation. There are eight naturally occurring forms of vit E (tocopherols and tocotrienols). Fortified foods and vitamin supplements are synthetic compounds, usually alpha-tocopherol. These various forms have different levels of biological activity in humans. Food sources are mainly from vegetable oils/spreads (~20%), and the remaining from cereals, grains, nuts, fruits, vegetables and the fatty portion of meats. Recommended daily allowances (RDA) range from 4 mg/d in infants to 15 mg/d in adults and 19 mg/d for lactation. Gut absorption of vit E is inefficient and dependent on bile acids, pancreatic lipase, micelle formation, uptake in the enterocytes and chylomicron formation. It then travels from the liver with VDLD, VDL and LDL particles. Excretion includes urinary metabolites from alpha and gamma tocopherol with the majority of vit E excreted in the stool (un-absorbed) and minor amounts in bile secreted into stool.

Deficiency is rare except in conditions such as CF and pancreatic insufficiency (PI) where patients experience chronic fat and fat soluble vitamin malabsorption. Deficiency signs are in the peripheral nervous system (sensory neurons, spinocerebellar ataxia, myopathy and retinopathy). Reports of toxicity are uncommon (leukocyte function impairment, coagulopathy), and the tolerable upper limit of daily intake is relatively high (200 to 1000 mg/d based on age) (1). Reports of healthy adults taking chronic supplements as low as 400 mg/d were associated with increased all-cause mortality and have led to more caution (2) .

During the presentation, we will review: -DRI vitamin E values for intake and upper limits for healthy people of all ages (1) -Current CF-specific vitamin E supplemental intake recommendations (3) -Vitamin E content in CF-specific vitamin supplementation products -Selected CF-specific publications -Preliminary findings from our longitudinal nutrition and CF studies Summary: Compared to the healthy population, children and adolescents with CF and PI have a similar vit E dietary intake and massively higher vit E supplement intake. This pattern of intake resulted in mean serum levels of alpha-tocopherol and alphatocopherol adjusted for serum cholesterol that were much higher than the healthy population comparisons (4). This was seen in a series of nutrition and CF studies (5) over more than a decade (1998-2011) with about 300 subjects from multiple CF centers (3 to 13 per study). About 25-60% of the blood values were within the reference range and 5-15% were in the low/deficient range. On typical CF care practice regimens, 30-70% of participants exceeded the expected reference range, depending on the study participant group characteristics.

Patients with CF and PI are in a unique setting. They have lifelong fat and vit E malabsorption, even with the best efforts to balance and optimize food/fat intake and pancreatic enzyme replacement. They have lifelong need for high dose synthetic vit E supplementation that is most often provided in a fixed dose CF-specific vitamin preparation. The evidence suggests that in this setting, a few patients still have insufficient blood levels of vit E (alpha-tocopherol), and this is likely due to poor adherence with the vitamin supplements. Many more patients have levels that are significantly higher than healthy people on a typical diet. The risk of exposure to lifelong high vit E blood levels, even adjusted for the lower cholesterol levels expected in CF and PI, is not known.

References Why do some clinicians still not appreciate the importance of vitamin K supplementation?

Why is there not an understanding of the optimal vitamin K preparation to prescribe?

Why vitamin K supplementation is important: Dietary vitamin K is available as K1 in green leafy vegetables and some vegetable oils, and as gut bacteria synthesised K2. It is not clear how much these contribute to overall vitamin K status. Both can function as the essential co-factor for the posttranslational conversion of specific glutamyl residues to gamma-carboxyglutamyl residues in vitamin K dependent proteins. When undercarboxylated these proteins, e.g., prothrombin and osteocalcin (OC), are functionally defective. This may result in clinically significant bleeding and/or low bone mineral status. In the non-CF population in the absence of malabsorption, because of the efficient liver use of available vitamin K, impaired blood coagulation is rare but the dependent bone proteins are more susceptible to vitamin K deficiency. In the CF population in the absence of supplementation vitamin K deficiency is almost universal and people with CF are therefore likely to have negative coagulation and bone mineral density consequences.

How should we monitor vitamin K levels? The vitamin K assay is not offered outside of selective laboratories, perhaps in part explaining why supplementation has not been universal. Estimation of the prothrombin time has been used as a surrogate measure but correlates very poorly with anything other than severe vitamin K deficiency. A plasma sample that has only 50% of the normal concentration of prothrombin still has a normal prothrombin time. PIVKA-II levels are much more sensitive to vita-min K insufficiency but are not routinely available. Measurement of the fraction of total OC as under-carboxylated OC is a sensitive indicator of the vitamin K status of bone but is generally available only as a research tool.

So what should we do? It is my opinion that all people with CF should receive vitamin K supplements without doctor angst about the difficulties in monitoring blood levels because:

i) The vast majority of patients will have low vitamin K levels if not receiving supplements.

ii) Vitamin K supplementation is really safe. iii) Intuitively and scientifically we can accept that "normal" vitamin K levels are "normal" for a reason and that it is logical in CF care to restore the abnormal to the normal.

And how much should we give? (Question to which I do not have the answer.) It is easier to advise on how much not to give! There is no validated dose. Early USA recommendations were woefully inadequate, 2.5mg weekly for infants and 5mg thereafter (1992) with later recommendations varying between 1mg/day and 10mg/week. More recently supplements of up to 20mg weekly have been recognised as insufficient. It is our policy to prescribe as per the UK CF Trust Guidelines -10mg supplement daily to adults and children above seven years, 5mg daily to children two to 7 years of age who are able to swallow the tablet, and 300µg/kg/day, rounded to the nearest 1mg for those less than two years of age. In practice supplementation is usually delayed until the tablet/capsule preparation is tolerated.

Which preparation? We recommend using phytomenadione, (phylloquinone, vitamin K1). Menadione salts should not be used. Menadiol is a synthetic vitamin K and cannot be compared to K1 because it only works by synthesis of MK-4 in the body by addition of the natural side chain, an inefficient conversion. Menadione also has well known toxicity issues (mutagenicity) especially relevant to children, and has the potential to cause haemolysis and liver damage.

We have neglected vitamin K supplementation in CF care protocols. There is no excuse for continuing to do so now that we are more aware of its important metabolic role. It is safe to supplement but we do need dose validation. Folic acid background: Folic acid is involved in red blood cell, DNA and 1-methyl metabolism, and is a key nutrient required to prevent neural tube defects. Leafy green vegetables and whole grains (fortified) are good sources of folic acid. Digestive tract atrophy, diarrhea, anorexia, weight loss, megaloblastic anemia, fatigue, weakness, shortness of breath, thrombocytopenia, blunted WBC production and response, mood changes and impaired fetal growth as well as neural tube defects are signs and symptoms of deficiency. There is no known folate toxicity. Serum folate status reflects dietary intake whereas RBC levels reflect whole body stores. Recently there has been interest generated in folate supplementation in individuals with CF related to 1 methyl metabolic pathways in which supplementation with 5-methyl tetrahydrofolic acid + vitamin B 12 have been shown to decrease homocysteine levels and alter RBC phospholipid fatty acid profiles and reduce RBC membrane oxidative damage (1). B 6 background: Vitamin B 6 encompasses the pyridoxine family of compounds, involved in protein, lipid, neurotransmitter and RBC synthesis and function. Potatoes, bananas, lentils and spinach are good dietary sources. Skin lesions (perioral, periaural and genital), fissured corners of the mouth, lips, purplish tongue, anemia, impaired leukocyte function, reduced antibody production, mood changes, headaches, sleep disturbances, abnormal brain wave abnormalities, muscle twitching, tingling in the hands and feet, difficulty walking, seizures, and dyslipidemias are findings in deficiency states. Long term intakes of >1000 mg/day may increase risk for peripheral neuropathy. Most commonly status is measured as plasma B 6 or as pyridoxal-5-phosphate (active form); low levels of this active form have been reported in children with CF (2). B 12 background: Vitamin B 12 encompass the cobalamins, which are involved in 1-methyl metabolism DNA synthesis, and in the activation of folate to tetrahydrofolate (active form), fat metabolism, DNA and myelin synthesis, and act as antioxidants to maintain reduced glutathione. Animal protein is a good dietary source; vegetarian diets increase risk of deficiency. There are several perturbations in the digestive and absorptive mechanisms for B 12 in individuals in CF, and the Schilling test is frequently abnormal in patients with CF, which should indicate deficiency. However, B 12 deficiency is not common in the CF population. Atrophy of the digestive tract, megaloblastic anemia, weakness, shortness of breath, thrombocytopenia, blunted WBC development and response, mood changes with hallucinations, paranoid behavior, tingling in hands/feet, sensory loss, unsteady gait, muscle coordination issues are reported in deficiency states. There is no known toxicity. Serum B 12 levels are the most commonly measured lab test to determine status. Methyl malonic acid levels are sensitive index of status and are also commonly used to determine status.

As part of a randomized, double blinded placebo controlled fatty acid + phospholipid trial, we collected dietary intake, anthropometric data, liver enzymes, CBC with differentials, high sensitivity CRP, medication and supplement intake data, vitamin plasma B 6 , serum B 12 and RBC folate data in 110 children 5-18 years of age with CF, pancreatic insufficiency, and FEV 1 > 40% predicted with no known liver/metabolic disease. The status of these vitamins was determined as they are key in the phospholipid and intersecting homocysteine-methionine and 1 methyl metabolism pathways, which were an area of interest for the study.

The subjects were recruited from 10 pediatric CF centers in the Northeast United States. The baseline vitamin B data gathered therein is illustrative of serum status of these vitamins in this populations under current patterns of dietary intake and clinical care. Three day dietary and supplement intake from CF specific and other vitamins were assessed and were expressed in %RDA. Fasting plasma B 6 , serum B 12 and RBC folate levels were measured by means of HPLC, immunoassays respectively. Predictors of serum status including age, gender, dietary and supplement based intake were explored in multiple regression models. Results are presented below.

RBC folate B12 Improvements in CF patient outcomes have occurred over the past several decades due to an improved understanding of the pathogenesis of the disease, leading to the development of a number of new therapies and a better appreciation of how best to use them. CF care has come a long way since alfa dornase became available as the first therapy developed specifically for use in CF. CF clinicians may now prescribe several different medications that help augment airway clearance; a variety of inhaled antibiotics with different modes of delivery; and an assortment of potentially therapeutic immunomodulatory and anti-inflammatory drugs 1 . The excitement surrounding the development of several new CFTR-specific therapies is palpable 2 , and these are all in addition to a number of alternative options that seem to be appearing for treatment of acute pulmonary exacerbations 3 . Up until now, each new CF therapy was typically just added onto the treatment regimen, but we are now going to have to begin to make choices from amongst our new "embarrassment of riches". In the face of this abundance of ther-apeutic options, CF providers will need to learn a vocabulary and new skill set in order to make appropriate decisions in standardizing (as well as personalizing) treatment regimens.

A starting point is to first understand the difference between therapeutic efficacy and effectiveness. Efficacy is a measure of how well a treatment works in an ideal environment, whereas effectiveness measures how well it works when prescribed in practice. Randomized clinical trials (RCTs) that are done in CF are typically explanatory 4 trials that are designed to determine if a treatment has efficacy under ideal experimental conditions. They enroll patients most likely to benefit from the new therapy and attempt to maximize the likelihood that the therapy is used correctly in order to provide evidence of biological activity. When, as has usually been the case in CF, there is no pre-existing proven therapy, RCTs are performed with placebo controls but when an efficacious therapy already exists, the performance of a placebo-controlled trial is ethically problematic. Alternative approaches, which are distinguished primarily by their statistical analyses and power calculations, are superiority trials (which are done to establish that a new therapy is superior to the established one, and are analyzed in the same way as placebo-controlled trials); equivalence trials, which aim to determine that a new intervention is therapeutically similar to another by predefining a range of difference in outcomes that would be considered equivalent; and, most commonly, noninferiority trials, which determine if a new treatment is similar to or better than an existing one, requiring the new one to perform within a limited predefined range of inferiority. The rationale for performing equivalency and noninferiority trials is that the new drug has some other advantage, e.g., fewer side effects, greater ease of administration, or lower cost 5 .

The explosion of interest in comparing effectiveness was initiated by a 2008 report of the Institute of Medicine that called for a program of comparative effectiveness research (CER) that would "assist consumers, clinicians, purchasers and policy makers to make informed decisions that will improve health care at both the individual and population levels" and was fueled by legislation mandating CER priorities and providing a generous funding allocation to support CER 6 . CER does not ignore comparative efficacy, but the interest in "real life" outcomes emphasizes and confers a new status on alternative research methods such as pragmatic trials and analysis of observational (patient registry) data obtained in the course of regular practice. Pragmatic trials are RCTs designed to measure whether an intervention works when used in usual conditions of care; this is in contrast to explanatory trials, which evaluate efficacy. In pragmatic trials, whose focus is generalizability to practice, the intervention is tested in a wide range of subjects and study sites and may be embedded in a context that includes other common adjunctive therapies, so they may not be able to separate out the effect of any individual component. They rely on outcomes of direct relevance to patients such as mortality, severe morbidity, and quality of life 7 .

The potential use of database analyses to do CER is of obvious interest in CF, given the existence of the powerful CFF registry as well as the development of CF registries in Europe and Australia. Patient real-life experience is reflected in registry analyses, but there are important pitfalls 8 . For example, sicker patients are more likely to receive more therapies, making it hard to distinguish an effect of those therapies. This phenomenon, known as confounding by indication (or indication bias) threatens the validity of registry studies unless it can be controlled for in the analysis. A number of techniques have been proposed to accomplish this, including propensity scoring, inverse probability weighting, and the use of instrumental variables 9 , but when prescribing behaviors are influenced by factors that are not collected in the registry, it unlikely that these biases can be completely controlled for 10 .

From a health policy standpoint, an interest in comparing the cost of care is intrinsic to the current interest in comparing effectiveness 6 . It is easy to justify the use of treatments that are more effective clinically when they are also less expensive, or even when they are equal in costs, but real-world relative cost calculations are often quite complex. Cost-effectiveness analysis (CEA) is a form of economic analysis whose practitioners have developed formal analytic approaches to comparing the relative costs and outcomes of two or more courses of action 11 . These calculations involve using a number of assumptions regarding the estimation of costs and outcomes and how these can be extrapolated across different populations. These assumptions and other methodologic issues need to be stated explicitly, and a sensitivity analysis, which examines the impact of variations in these assumptions from a statistical standpoint, should always be reported 12 . The results of a CEA are expressed in terms of an incremental cost-effectiveness ratio (ICER) to evaluate relative benefit. The denominator is a gain in health from an intervention (years of life, premature births averted, etc.) and the numerator is the cost associated with the health gain. Cost-utility analysis (CUA) is a specialized form of CEA that incorporates a quality of life component associated with morbidity to calculate relative indices such as disability-adjusted life years or, most commonly, quality-adjusted life years (QALY) to use in the denominator. QALYs are normally calculated using utility-derived quality of life measures that have been specifically designed for this purpose from surveys of the general population, such as EQ-5D and the SF-6D 13 ; CFQ-R was not designed to be used for this purpose. CUA allows comparison across different health programs and policies by using a common unit of measure, cost/QALY, but the calculation of QALY and use of CUA is controversial. On the one hand, applications for approval of new drugs in the UK are required by the National Institute for Health and Clinical Excellence (NICE) to be accompanied by a CUA that calculates ICER using QALY. In the U.S., the Patient Protection and Affordability and Care Act explicitly prohibits the use of QALY as a tool for establishing what type of health care is cost effective or recommended 14 .

An example of a CUA comparing two different inhaled antibiotics will be presented.

Over the past decade, significant medical advances in cystic fibrosis (CF), have led to the development of new medications and treatments, resulting in significant improvements in life expectancy. However, the daily treatment regimen for CF is now more time-consuming and complex, requiring between two and four hours per day. This presents significant challenges for patients and families, with recent evidence suggesting that rates of adherence are at or below 50% (Eakin et al., Modi & Quittner, 2008) . In addition to the time burden, the costs of these new medications has also risen, resulting in higher copays and out-of-pocket costs for patients (Briesacher et al., 2009) . The purpose of this study was to assess the association between adherence to pulmonary medications and health service utilization and costs for patients identified in a national claims database from 2005 to 2011.

Patients ages 6 years through adulthood with a CF diagnosis and a prescription fill for a pulmonary medication (oral azithromycin, inhaled aztreonam, colistin, dornase alfa, hypertonic saline and tobramycin) were identified using the MarketScan® Commercial Claims and Encounters database. Medication possession ratios (MPRs) were computed for each pulmonary drug during a 12-month study period and then averaged to create a Composite MPR (CMPR) for each patient. Patients were categorized as having low (<0.50), moderate (0.50-0.80) or high (≥0.80) adherence based on their CMPR. All-cause and CF-related health service utilization during the second year after study index date were compared across CMPR categories using negative binomial regression models adjusted for gender, age, region, insurance, comorbidities, index year, and average monthly drug copay in the 6 months prior to the study period. All-cause and CF-related medical service costs were compared between cohorts and modeled with generalized linear model regressions (gamma distribution, log link), controlling for the covariates listed above.

A total of 3,287 CF patients were included in the analyses. After adjusting for patient characteristics, low versus high CMPR was associated with significantly higher rates of all-cause (IRR =1.46, 95% CI=1.20-1.79) and CF-related inpatient/emergency room visits (INPAT/ER) visits (IRR =1.34, 95% CI=1.07-1.69) in the second year after the study index date. Similarly, medium versus high CMPR was associated with significantly higher rates of all-cause (IRR =1.28, 95% CI=1.02-1.59) and CF-related INPAT/ER visits (IRR =1.44, 95% CI=1.12-1.84). All-cause medical service costs were $46,379, $44,824 and $37,067 for patients with low, moderate and high adherence, respectively, during the second year after index date.

In summary, patients with low to medium adherence to pulmonary medications had increased risks of allcause and CF-related inpatient and ER visits, and higher health care costs. Given the challenges of adhering to the CF regimen and the potential costs of poor adherence, interventions to improve adherence behaviors should be developed and evaluated. Overview: This study investigates the effect of socioeconomic status (SES) on outcomes in people with cystic fibrosis (CF) in the UK and Denmark. People with CF from socio-economically disadvantaged backgrounds die younger than those in more advantaged social positions in the UK and the U.S. The key challenge is to understand how and when these inequalities develop, in order to identify promising options for intervention. We demonstrate the importance of social factors for key outcomes in CF, in the context of the universal health systems in the UK and Denmark, and contrast these findings with those from the US and other settings.

Background: Studies across the world have consistently shown that people from socio-economically disadvantaged backgrounds experience worse health than those in more socio-economically advantaged positions [1] . In the UK and internationally, policies have been implemented to try to reduce these inequalities, with limited success [2] . In order to develop more effective interventions we need a better understanding of how these health differences are generated and maintained. CF is a valuable case for understanding how and when health inequalities develop, since unlike other childhood respiratory disorders CF does not discriminate by social class: SES does not affect disease risk but does affect the health outcomes of having CF [3] .

The improvement in survival over successive birth cohorts in CF has been striking, but these improvements have not been shared evenly, both within and between countries [4] . Evidence from the U.S. indicates higher survival rates in the 1980s and 1990s among more advantaged socioeconomic groups, measured by Medicaid status and area-based income, compared with their less advantaged counterparts [5] [6] [7] . For instance the adjusted risk of death was almost four times higher in CF patients with Medicaid cover, a surrogate for low socioeconomic status, compared to those without Medicaid cover [5] . In the UK a cross-sectional study by Britton found a consistent trend from 1959 to 1986 toward higher age at death in CF patients in non-manual, compared with manual occupations [8] . Barr and colleagues recently updated this study, using death registration data in England and Wales, and showed that the socioeconomic divide in premature mortality in CF has persisted with no substantial narrowing for over four decades [9] .

There are significant differences in intermediate CF outcomes such as %FEV1 and weight centile in young people with CF in the U.S. For instance Schechter et al. found a cross sectional difference of 6.7% in %FEV1 by Medicaid status, which increased to 9.2% with adjustment for various confounders [5] . Furthermore O'Connor et al report a stepped social gradient in %FEV1 and weight centile, with an absolute difference of 5.5% between highest and lowest income quintiles in the under 18 age group [7] . A key issue is the role of the healthcare system in mitigating or perpetuating inequalities in CF. Michael Schechter's studies from the U.S. have shown mixed patterns of access to treatment by SES [10, 11] . Young people on Medicaid appear more likely to access care (sick hospital visits, chronic therapies and IV antibiotics). Children living in low-income areas are more likely to receive oral nutritional supplementation, but less likely to receive macrolide therapy, and IV antibiotics. The authors conclude that inequalities in CF outcomes are not explained by differences in healthcare or chronic therapy use. In their study of the effect of socioeconomic status on hospitalisation rates in CF in Ontario Stephenson et al suggest that the provision of universal healthcare in Canada may explain the lack of a socioeconomic differential in hospitalisations [12] .

To investigate these issues in the UK and Denmark, we undertook retrospective cohort studies of population level cohorts.

Methods: For the UK analysis we studied around 9000 people with CF aged under 40 years, with data captured at 58,000 annual reviews between 1996 and 2010. Standardised census-based indices of multiple deprivation (IMD) from the UK constituent countries were used as small area measures of SES. In Denmark the analysis captured 479 patients seen between 1969 and 2010 at the two CF centers in Denmark, with data measured on a monthly basis on around 70,000 occasions. The Danish dataset was linked to population level registers in order to collect individual level socioeconomic data, on patients and their parents (e.g., employment, income, healthcare expenditure). In both countries we explored the effects of socioeconomic status measures on key clinical outcomes (%FEV1, growth and nutritional measures), healthcare use (access to care, use of chronic therapies), and social outcomes (employment and educational outcomes). The analyses employed longitudinal data modelling techniques to examine changes in outcomes over time in groups and individuals whilst allowing for correlation within patients, and potentially informative missing values [13, 14] .

Findings: We have, for the first time, characterised key outcomes over time in people with CF in the UK and Denmark, and explored the impact of social deprivation, as well as other covariates. Our findings, to date, have shown important differences in outcomes by socioeconomic status in the UK, which appear early, and vary across the life-course, perhaps as a consequence of good access to treatment in children from families living in areas with greater social deprivation. The early appearance of inequalities supports the need for interventions that are targeted at the early years and reinforce the importance of screening for cystic fibrosis in newborns. Research to identify interventions to address the early appearance of inequalities needs to be focussed on the early years.

Beth Sufian, J.D.

Children and adults without health insurance coverage or who have inadequate coverage visit their CF care center less frequently and may take less medication than those who have access to comprehensive insurance coverage with low cost share and co-pays.

These individuals with CF have less access to care and treatment than their CF counterparts who either have access to government benefits which provide comprehensive coverage without any cost share or co-pays or whose employer provides health insurance coverage with a low cost share for policyholders.

Individuals with CF who have inadequate or no insurance coverage can benefit from direction from their CF care centers to resources and programs that can help them access assistance with co-pays or obtaining health insurance. Many people with CF and their families are not aware that there are resources that may be able to assist with co-pays for medication or options for free or low cost health insurance coverage. These individuals go without medication or treatment due to financial concerns when they could be benefitting from programs that would allow them access to such medication and treatment.

A study of calls to the CF Legal Information Hotline ("Hotline") and the CF Social Security Project exemplifies the increase in problems related to treatment cost share, co-pays or no access to insurance coverage between 2011 and 2012. The CF Legal Information Hotline was established in October of 1998 and has had over 26,000 calls since that time. The Hotline is funded by a grant from the CF Foundation and is staffed by four attorneys and 3 para-legals. All contacts to the Hotline are free and confidential. The Hotline provides information about government benefits, health insurance benefits, employment and education rights. The Hotline has seen a dramatic increase in calls since the country's financial crisis began in 2008. In addition, in 2012 there has been an increase in callers who indicate they have failed to fill a prescription for medication or failed to attend a CF care center visit because they were not able to afford the co-pay for such medication or treatment. Often callers indicated they had no information on available programs that could assist them with their financial issues.

A survey of clients who have used the CF Social Security Project to help obtain government benefits yielded similar results. The CF Social Security Project allows attorneys to represent people with CF for free in an initial application for Social Security disability benefits or Supplemental Security Income benefits. The Project is a case management project funded by the CF Patient Assistance Foundation. Once the Social Security benefit application is approved the person with CF becomes eligible for Medicaid benefits immediately or Medicare benefits after a 29 month waiting period is met. Those waiting for Medicare can use their monthly cash benefit to pay for COBRA benefits which last until Medicare eligibility begins. The CF Social Security Project has successfully obtained benefits for over 650 children and adults with CF between March 2010 and May 2012. Clients surveyed who had restricted access to medication and treatment or no coverage at all before obtaining government benefits reported that they were able to access medication and treatment without much if any cost share once they became eligible for government benefits.

CF care centers do not always have access to a social worker who can devote time to helping patients access government programs or co-pay assistance programs. Even if the CF center has a dedicated social worker the government benefits application process is complex and time consuming. It is important for CF care centers to provide information to patients at least once a year on the programs available to assist them with issues related to access to insurance benefits, medication and government benefits. In addition, it is important for the CF community to indicate their need for such programs and for industry and CF organizations to fund programs that help patients access assistance for co-pays, insurance coverage or government benefits. 

Proven Efficacy of Azithromycin: To date, five randomized, placebo controlled clinical trials of azithromycin have been performed around the world in different CF patient populations including adults with moderate lung disease, adults and children with moderate to mild lung disease, and children with moderate or mild lung disease both chronically infected or uninfected with Pseuodomonas aeruginosa [1] [2] [3] [4] [5] . These studies have shown significant improvement in lung function ranging from a relative increase in FEV 1 % predicted from 3.6 to 6.2% and/or significant reductions in pulmonary exacerbations, increased weight gain, reduced initiation of new oral or intravenous antibiotics, and/or hospitalizations [1] [2] [3] [4] [5] . Sub-group analysis has demonstrated that a reduction in pulmonary exacerbations occurred in the absence of improvements in lung function [6] .

Safety and Tolerability of Azithromycin: The safety and tolerability of azithromycin have also been studied. Gastrointestinal side effects are a well known effect of macrolide agents. As expected, participants on azithromcyin had increased nausea (33% of azithromycin vs. 16% of placebo participants) and diarrhea (23% of azithromycin vs. 8% of placebo participants), but these toxicities were not associated with discontinuation of drug and were classified by participants as mild to moderate adverse events [3] . A subsequent study performed in patients with milder lung disease did not note this association [5] . In one study, participants on azithromycin experienced more wheezing, that was not associated with an increased rate of discontinuing study drug, hospitalization, use of bronchodilators, or deterioration of lung function [3] . Subsequent studies did not detect this association [5] . Abnormalities in hearing [3] or subjective reports of tinnitus or hearing loss [3, 5] were not increased in azithromycin participants. In contrast, azithromycin was associated with less cough and less productive cough when compared with placebo [5] . No adverse laboratory events defined as changes in liver function, creatinine, or absolute neutrophil counts (ANC) were associated with azithromycin [1, 3, 5] .

Azithromycin has been used since the late nineties in patients with cystic fibrosis who were positive for Pseudomonas aeruginosa. Five prospective, randomized placebo controlled trials have shown beneficial effects on both FEV1 and frequency of exacerbations. However, the consistent improvement in respiratory function after 6 months of therapy has not been demonstrated for extended periods of time. Data reported by Equi show the maximum increase in pulmonary function at 6 months and a decline in efficacy over 1 year, with a tendency towards equivalence between azithromycin and placebo (1). In a trial by Saiman with Pseudomonas aeruginosa uninfected patients, the difference in FEV1 was not statistically significant (2) . The long-term efficacy of azithromicin urgently needs to be investigated. A retrospective study of Tramper-Stranders et al demonstrated an increase in FEV1 in the first year, but a decline in the following three years (3) . Recently, Wilms et al retrospectively studied 74 patients with cystic fibrosis, who were chronically infected with Pseudomonas aeruginosa and who started azithromycin maintenance treatment between 1998 and 2005 (4). Mean duration of follow-up after start of azithromycin was 7.6 years. One year after start FEV1 was + 1.71% compared to baseline, after 18 months + 0.78% and after 24 months FEV1 was back to baseline (-0.19%). During the following years mean annual change in FEV1 was -2.26%. These data suggest that initiation of azithromycin therapy is associated with only temporary improvements in pulmonary function.

A second concern is the rapid development of macrolide resistance in patients with Staphylococcus aureus. In several studies emergence of resistant strains was observed in up to 80% 6-12 months after start of maintenance therapy with azithromycin (3, 5) . It is unclear whether the emergence of these strains is responsible for the only temporary effects in patients with CF. The early positive effects might be counteracted by later detrimental effects of resistant Staphylococcus aureus strains.

Although azithromycin is widely prescribed in cystic fibrosis care, data on the pharmacokinetics of the drug are scarce. Half-life of the drug has been reported to be up to 180 hours with clear storage in polymorphonuclear cells (PMNs). The effects of repetitive dosages on the function of immune cells are largely unknown. Azithromycin has been associated with an increased incidence of nontuberculous mycobacterial infections, particularly of multi-drug resistant species like Mycobacterium abscessus (6) . These possible negative effects on immune cell function might outweigh the initial positive effects of azithromycin.

Storage of azithromycin in PMNs results in high levels of the drug in the inflamed lungs of patients with cystic fibrosis. This raises concerns about interactions with other CF-related drugs. In vitro experiments of Ripoll et al showed inhibitory effects of azithromycin on the hydrolysing activity of dornase alfa (7), while in the study of Equi a reducing effect of azithromycin on the effects of dornase alpha was found.

Nowadays, azithromycin is widely used in patients with CF, many of them already treated for more than 10 years. There is extensive evidence that treatment for more than 6-12 months does more harm than good to these patients. CF lung disease starts early: When a child with cystic fibrosis (CF) is born the lungs show a normal structure. Chronic inflammation and infection occur early in life and contribute to progressive structural lung damage that ultimately leads to respiratory failure. The most important components of structural lung disease in CF are bronchiectasis (BE) -an irreversible widening of airways -and small airways disease. Treatment should start at an early stage with the aim to prevent any irreversible damage from occurring. We, as clinicians, should have zero tolerance with regards to progression of the disease! Therefore, both clinicians and patients need accurate and sensitive measures of the severity of BE and small airways disease so that action can be taken when needed and the effectiveness of interventions can be assessed.

Heterogeneity of CF lung disease: CF lung disease is heterogeneous (1). Unfortunately, we do not have adequate prediction models to identify patients at risk of developing progressive disease. Models using validated bio-markers are needed to personalize treatment and monitoring. Until we have such models we need to use sensitive and accurate tools to determine the current disease status in order to tailor treatment to the needs of the patient. Progression of BE needs to be diagnosed as early as possible since it cannot be reversed and progression needs to be avoided. Currently, chest CT is the gold standard to detect and monitor BE and trapped air. Symptoms, pulmonary function tests (PFT), and chest X-ray (CXR) are not sufficiently sensitive to diagnose and monitor BE. To evaluate the status of CF lung disease using only these insensitive techniques is no better than flipping a coin. "Heads or tails? What's your call?" -is that the way you like to practice medicine? Bronchiectasis cannot be diagnosed by functional tests: Spirometry and other PFTs have never been the gold standard to diagnose BE. PFTs at best tell you that something is wrong with the lungs but not what structural changes are causing these functional changes. Changes in various PFT variables are associated with structural changes. Spirometry served its purpose as a surrogate test to monitor CF lung disease at a time when patients were losing on average 10% of function per year and we had few therapies to stop this progression. Fortunately, prognosis has improved substantially and we have a wide array of potential treatments to pick from. Decline in spirometric indices are, on average, below one percent per year which is well within the noise range of spirometry. Does this mean CF lung disease is stable? Unfortunately not, it was shown that spirometry is just not sensi-tive enough to inform us adequately about the presence or progression of BE (2, 3) . Evaluating progression of CF lung disease only based on PFTs and or CXR is like driving blindfolded. Not a good idea! Chest CT is the best: Unfortunately, CXR is not sensitive to detect, and monitor the most important components of CF lung disease in individual patients. The reasons for this are simple. First, a CXR is a two dimensional representation of a complex 3 dimensional structure making accurate identification of structural abnormalities very difficult. A CXR at best tells you something is wrong with the lung but not what. Second, It has been recognized -since 1945 -that about one third of the lung volume cannot be evaluated using CXR because it is hidden behind the heart and domes of the diaphragm (4). Chest CT has emerged in the early nineties as the gold standard to diagnose BE (5). Chest CT is superior to CXR for detecting and monitoring structural abnormalities in CF patients in early and advanced disease (6) (7) (8) (9) (10) (11) . Chest magnetic resonance imaging is a promising alternative for chest CT but currently inferior to chest CT for detecting early disease and accurately monitoring progression (12) . CT is the only method that can detect and measure progression of structural lung disease in infants. Hence, to diagnose and monitor CF lung disease, chest CT is the current gold standard -why wouldn't one use it?

Image gently: A clear limitation of Chest CT is that it exposes the patient to ionising radiation. Fortunately, current low dose chest CT protocols only expose the patient to an equivalent of radiation that is in the order of 4 months of natural background radiation. When following a bi-annual low dose protocol the risk is very low considering the high morbidity and mortality related to CF lung disease (13) . Further innovations in CT technology and image analysis will allow even further reduction of this dose in the near future.

Therapeutic consequences: Chest CT is the most sensitive and accurate method to determine whether the most important components of end stage CF lung disease -BE and trapped air -are absent, stable or progressive. The clinical relevance of CT detected BE has been well demonstrated (1, 10, (14) (15) (16) (17) (18) . Valid structural information such as that provided by CT is vital if we are to develop truly personalized therapeutic strategies. The treatment burden related to CF therapy is substantial for most patients (19) . The need to intensify treatment and thus increase the burden for the patient requires the support of reliable clinical data. Similarly, if CF lung disease has been stable a reduction in the intensity of treatment can sometimes be considered. Chest CT cannot cure the disease but it is the most sensitive and accurate diagnostic modality to guide personalized treatment.

Summary: Chest CT is the gold standard to diagnose bronchiectasis, a key component of CF lung disease. CT is a sensitive and accurate method to determine the severity and progression of CF related lung disease. Chest CT allows the CF team to personalize the treatment to the needs of the patient. When state of the art low dose chest CT protocols are used radiation risk is considered to be minimal. So take off your blindfolds and start looking at the future of CF lung monitoring! inhaled tobramycin (8) , and following pulmonary exacerbations (5) . CT scans are also associated with future lung disease severity (as measured by future FEV 1 , chest radiography, and pulmonary exacerbations) (9,10) and mortality (11) . Finally, visualization of abnormalities may lead parents and children (or clinicians!) to increase adherence to treatments or accept new treatments that they may otherwise not use (12) .

Despite all of these benefits, routine use of chest CT scans has not been recommended for patients with CF. The primary concern surrounds the impact of ionizing radiation on the future risk of cancer, especially when repeated CT imaging is used in young children. This risk has been estimated to be an additional cancer in 1 in 200 females and 1 in 1,000 males if CT scans were used annually in patients with CF (13), although estimates used in these calculations may overestimate the radiation dose used in modern dosage protocols (14) . On the other hand, it appears that children with CF are at higher risk for malignancies than the general population (15) . More recently, it has been reported that an extra case of leukemia will develop per 10,000 CT scans performed in children in the general population (16) . In a response widely covered in the lay press, the Alliance for Radiation Safety in Pediatric Imaging encouraged parents to ask several questions when their child's physician recommends a CT scan. We can use our answers to two of these questions to assess whether routine chest CT scans should be recommended.

(1) How will having this exam improve my child's care? The answer to this question must be left to the treating physician, as there is no evidence that the use of chest CT scans has affected lung disease progression in children with CF (14) . Additionally, comparisons with PET scans have revealed that CT scans cannot reliably differentiate between localized active infection/inflammation and residual abnormalities (17) . Routine CT scans would logically provide the most information when lung disease is in its early stages, as once lung disease progression has advanced, spirometry, symptoms, and chest radiography can more easily detect changes in disease status (18) . However, most reports of chest CT scan abnormalities in CF have been in older children and adults, meaning there is little standardization of imaging protocols to guide individual radiologists (18) . Finally, to truly maximize the utility of chest CT, quantitative scoring is required. However, scoring protocols are not widespread or standardized, meaning that interpretations will have considerable variability, especially for radiologists not familiar with CF lung disease.

(2) Are there alternatives that will provide the same information? Increasingly, the answer here is yes. Multiple risk factors for lung disease progression have been identified which can guide clinicians in being more or less aggressive in adding new therapies. The presence of mucoid P. aeruginosa alone has been shown to be asso-ciated with the presence of bronchiectasis on chest CT (19) . Chest CT scans are considered the gold standard of structural lung disease, but chest radiography has been shown to be sensitive in detecting concurrent chest CT abnormalities across childhood and adolescence (20) , and as strongly associated with future lung disease progression as chest CT scores (9) . This information can be obtained longitudinally with a fraction of the ionizing radiation. LCI may also be used to detect early functional abnormalities, and LCI has been shown to correlate with chest CT scans and future lung disease severity in young children diagnosed after newborn screening (6) . Finally, recent improvements in MRI technology make it possible to identify structural lung disease with nearly the precision as chest CT; MRI scans have the added benefit of not exposing patients to ionizing radiation and may be used to evaluate functional abnormalities (e.g., regional air trapping, mucociliary clearance) (21) . One of the main advantages of chest CT over chest radiography and spirometry is that small areas of localized structural disease can be identified (12) . However, there is no evidence that directing care against localized structural disease is more effective than treatments directed against modalities that measure averaged disease levels (e.g., LCI, spirometry).

It is apparent from the answers to the above questions that chest CT should not be used routinely in children with CF. There is insufficient evidence to guide clinicians' response to chest CT abnormalities, and there are (or soon will be) sufficient alternatives to chest CT, especially in the young child with mild lung disease who may be especially sensitive to ionizing radiation and for whom appropriate CT scanning protocols are poorly defined. Children with CF who are at high risk for CT abnormalities and/or future lung disease progression can be identified without the use of chest CT. Chest CT should continue to be used as recommended by Borowitz et al: CT scans "may be helpful in infants with symptoms or signs of lung disease who fail to respond to basic interventions" (22) . Filamentous fungi are inhaled routinely with each breath, as "spores" (or conidia), which are produced asexually by moulds that grow in environments that contain organic material. Some organisms -Aspergillus species -and particularly, organisms within the A. fumigatus group, are well adapted to survive in human airways. If spores (conidia) survive in the airway, they can convert back into hyphal, or the filamentous forms of the fungus. This morphotype produces metabolites and products that serve to promote organism tolerance, survival, and polarized growth, such as proteases and other toxins [1] . Recent studies have actually documented that we have different immunologic mechanisms to distinguish between the different forms of the fungus, with conidia being phagocytosed and cleared by both airway epithelia and professional phagocytes without evoking much of an inflammatory response; in contrast, hyphal components evoke production of numerous inflammatory mediators and cellular recruitment through recognition of different cell wall components, such as beta-glucan [2] [3] [4] [5] . Thus, A. fumigatus is one organism that grows as different cell types, and our bodies are well adapted for distinguishing between the two morphologic states.

Although fungi are common in our airways, they are not always found in a hyphal form; healthy airways constantly clear inhaled spores using multiple different mechanisms. In fact, a positive sputum or lavage culture is actually indicative of relatively large fungal burden -and the amount of conidia that are "normally" present do not enable in vitro growth using standard culture methods. The question that arises is one of the "chicken and the egg": when Aspergillus species are present in the airway, is that simply a sign of poor conidial clearance, or is the organism "causing" pathology, and poor clearance? Our thinking on this question has shifted in most recent years, largely as a result of our increased understanding of host immune responses to Aspergillus species, epidemiologic associations, and most importantly, by studies that definitively show that A. fumigatus evokes inflammation -and lung damage in CF animal models.

A small proportion of CF patients develop an entity that was labeled, "allergic bronchopulmonary aspergillosis" (ABPA). This is actually a poorly understood (but accepted) clinical syndrome that is characterized by development of a measurable antibody response (especially a type-II response) to Aspergillus antigens. Notably, these antigens are largely hyphal-excreted, suggesting that the primary defect is in either conidial clearance, or development of systemic immunity/allergic responses (or both). At a minimum, measurement of antibodies directed to the fungus is a sign that the organism is not truly a simple commensal in the airway; inhaled conidia have transitioned into hyphae, and the filamentous form is being recognized by systemic immune responses.

There is more evidence to suggest that Aspergillus species are in fact not simply 'innocent colonizers' in the absence of overt ABPA. Well developed cohort studies actually have shown that recovery of Aspergillus species is both common in the setting of poor pulmonary function, and also an independent risk for exacerbations [6, 7] .

Understanding the inherent biases in these clinical associations, the more compelling arguments that Aspergillus species are implicated in causing lung functional decline are generated by animal models. In one early study, it was shown that CFTR knock-out (-/-) and deltaF508 mutant mice developed profound lung inflammation and biased CD4 T cell responses after inhaled delivery of killed Aspergillus hyphal antigens [8] . Going a step further, we recently demonstrated that epithelial cells (EC) harboring CF mutations demonstrated impaired uptake and killing of conidia, and ECs with CFTR mutation undergo more conidial-induced apoptosis [5] . Germinated (hyphal) forms of the fungus evoke secretion of inflammatory mediators, with CFTR mutation resulting in increased airway levels of macrophage inflammatory protein 2 and KC, and higher lung monocyte chemotactic protein-1. After A. fumigatus inhalation, CFTR-/-mice developed exaggerated lymphocytic inflammation, mucin accumulation, and lung injury. These data definitively demonstrate a critical role for CFTR in mediating EC responses to A. fumigatus, and suggest that the fungus itself elicits aberrant pulmonary inflammation in the setting of CFTR mutation.

Thus, results of these studies suggest a causative role of the fungus in evoking aberrant inflammation and functional decline of lungs of animals with CFTR mutations, and a role of CFTR in mediating clearance of the early inhaled forms. This observation taken further would suggest that treatment of airway "colonization" will have a protective effect, much similar to that of treating P. aeruginosa. However, we have not yet shown this in clinical trials. One small placebo-controlled study evaluating itraconazole did not show a benefit; unfortunately, this drug is not well absorbed in this setting, reaches poor levels in the airways, and the study did not enroll enough people to show a treatment effect [9] . Thus, it may be considered a negative pilot study that suggests the need for more investigation. In vitro, in vivo, and epidemiologic studies suggest that antifungal therapy may benefit people with Aspergillus recovered from sputum, even in the absence of ABPA; a definitive clinical trial is necessary to define the optimal therapeutic strategy. As the CF population ages, and as intensive antibiotic suppressive therapy for bacterial infection becomes more common, the incidence and prevalence of A. fumigatus infection is increasing. In Canadian patients with CF the prevalence of A. fumigatus isolated from sputum rose from 8% in 2001 to 18% in 2009 (1, 2) .

Although many CF patients have Aspergillus that grows in their sputum, it is still unclear whether the organism is simply a commensal organism causing no harm to patients, or whether it may be a significant pathogen in CF patients. A small proportion (<5%) of CF patients who culture Aspergillus from their sputum develop allergic bronchopulmonary aspergillosis (ABPA), a well established complication of CF lung disease (3), however the relevance of Aspergillus in the 95% of patients who culture Aspergillus from their sputum, but who do not develop ABPA, is unknown.

Infection with A. fumigatus in CF patients may be transient, and species genotype can change during the course of the infection (3) (4) (5) , suggesting that Aspergillus may simply be a commensal organism in the airways. However, studies of the serologic response to A. fumigatus in CF patients (without ABPA) suggest elevated IgG and/or IgE antibodies in the majority of patients whose respiratory secretions contain Aspergillus organisms (6) (7) (8) (9) (10) (11) (12) (13) . The presence of a host antibody response to the organism would support the concept that A. fumigatus is not simply colonizing CF airways, but may be acting as a relevant pathogen.

A retrospective cohort study from the pediatric CF clinic in Toronto revealed that CF patients persistently infected with A. fumigatus, who did not have ABPA, had a lower FEV1 compared to uninfected patients, and these patients were also at a higher risk for pulmonary exacerbations requiring hospitalization (RR= 1.94, P= 0.0002) (14) . Similarly, a 3-year prospective cohort study of adult CF patients from Canada revealed that the prevalence of chronic infection with A. fumigatus was twice as high in patients with frequent CF exacerbations compared to those with infrequent exacerbations (15) . However, although Aspergillus colonization of the airways may serve as a marker of a sicker patient, this data does not by itself prove that Aspergillus is causing clinical infection. Furthermore, this data may be confounded by culture frequency; that is patients who have more exacerbations, especially those who are hospitalized, may have more sputum cultures ordered, which would put them at higher likelihood of having a positive culture for A. fumigatus, either through true colonization or via false-positive contamination of sputum cultures.

Diagnosing clinical fungal pulmonary infection in patients with CF is very difficult, since patients almost universally have bronchiectasis, structural lung damage on chest imaging, purulent secretions, and known pathogenic bacteria (such as P. aeruginosa) in their sputum. Determining whether fungus may be playing a pathogenic role in addition to bacteria is almost impossible. Thus the current standard of care amongst CF centers is to forgo antifungal treatment in CF patients who culture A. fumigatus in sputum but who do not have ABPA (16) . However to date, no prospective experimental studies have addressed the question of whether treating Aspergillus in patients with cystic fibrosis, without ABPA, could improve clinical outcomes. Clearly, if treatment of Aspergillus cultured from airway secretions results in improved clinical outcomes compared to a control group that did not receive treatment, then this would indirectly suggest that A. fumigatus is acting as a pathogen in the airways and is causing morbidity. We therefore conducted a double-blind, randomized, placebo-controlled pilot clinical trial to determine if 24 weeks of treatment with the oral antifungal agent itraconazole improved clinical outcomes in CF patients whose sputum was chronically colonized with A. fumigatus. Thirty-five patients with CF whose sputum cultures were chronically positive for A. fumigatus were centrally randomized to receive either oral itraconazole 5 mg/kg/d (N = 18) or placebo (N=17) for a 24 week treatment period (17) .

Over the 24 week treatment period, 4 of 18 (22%) patients randomized to itraconazole experienced a respiratory exacerbation requiring intravenous antibiotics, compared to 5 of 16 (31%) placebo treated patients, P = 0.70. FEV1 declined by 4.62% over 24 weeks in the patients randomized to itraconazole, compared to a 0.32% improvement in the placebo group (between group difference = -4.94%, 95% CI: -15.33 to 5.45, P= 0.34). Quality of life did not differ between the 2 treatment groups throughout the study. Therapeutic itraconazole blood levels were not achieved in 43% of patients randomized to itraconazole.

In summary, our pilot randomized controlled trial did not demonstrate any clinical benefits supporting longterm itraconazole treatment for patients with CF whose sputum was chronically positive for A. fumigatus. Our study was small, and relatively poor absorption of itraconazole was observed with sub-therapeutic plasma levels in some patients. Future larger trials, using newer second generation azoles, with therapeutic monitoring of drug levels, might be required to definitively determine if longterm antifungal therapy is beneficial for these patients.

There are several explanations, other than lack of statistical power and poor drug absorption, which can potentially explain why itraconazole did not appear to be clinically effective. The most obvious explanation may be that A. fumigatus, although often isolated from the sputum of patients with CF, may not play a pathogenic role in CF patients who do not have ABPA. If A. fumigatus is simply a colonizer of the CF airway, rather than an infecting organism causing tissue injury and inflammation, then treatment of this organism would not be expected to provide clinical benefits. To date, studies in CF patients are inconclusive as to whether A. fumigatus is causing infection in these patients, and while it cannot be excluded that individual patients may benefit from therapy, our current study would suggest that treatment of this organism on the basis of sputum isolation alone is not indicated in CF patients. The genetic diversity in the CFTR gene is described in one of three ways: 1) as a clinical case report or case series; 2) as an inventory from labs performing genetic testing; or 3) as a summary report from a particular CF center or geographic region. Projects such as the HapMap and 1000 genomes have begun to catalog diversity across the genome in ostensibly healthy populations. A comparison of the relative frequency of mutations seen in CF patients with the frequency of mutations seen in the general population provides the basis for a case-control study, however there are several problems to consider. First, the mutations seen in CF patients are not uniform throughout the world. It has been well characterized that the incidence of c.1521_1523delCTT (legacy named F508del) in CF patients is greatest in northwestern Europe and declines as you move south and east across the continent. Second, CF is not a uniform disease. There is increasing recognition of atypical cases of CF, CF cases diagnosed in adulthood, and of CFTR related diseases in which there is organ dysfunction related to CFTR mutation that does not meet the diagnostic criteria for CF. Third, variance found in the general (presumably healthy) population represents both neutral CFTR mutations (polymorphisms) as well as disease-causing mutations that may be present in CF carriers. Finally, the large number of CFTR mutations (more than 1900) and the fact that many are very rare means the scale of such a case-control comparison must be enormous. These problems (as well as other inherent biases of a case-control study) make it challenging to sort mutations into CFcausing vs neutral on the basis of population or prevalence data alone. Importantly, this translates to a difficulty using what has been learned about the genetic epidemiology of CF in general to any given genetically unique population.

With the help of CF centers and national registries the Clinical and Functional TRanslation of CFTR (CFTR2) project has assembled clinical data on 39,545 patients from 21 countries (in North America, Europe and Israel) in order to better characterize the disease liability of CFTR mutations that are commonly seen in CF patients. Through this CFTR2 project the 160 mutations seen in nine or more patients worldwide (with an estimated allele frequency that is greater than 0.01%) were analyzed with regards to clinical features of the patients that carry them, functional consequence of the mutation in a cell-culture system, and comparison of the prevalence of the mutations in a CF population vs. a carrier population. Specific details of each mutation can be reviewed on the publicly available CFTR2 website (www.cftr2.org). The mutations assembled for CFTR2 are intentionally biased towards those that cause CF (they were drawn from registries of CF patients).

Keeping that in view, we can examine the effect of a larger panel of well-defined CF mutations (122 disease causing mutations) on screening and diagnosis worldwide. The 122 mutations identified as disease causing (c.350G>A, legacy named R117H for this analysis was not considered disease causing) accounts for 94% (66,297/70,482) of the non-unknown CFTR alleles among the entire CFTR2 population. Among the U.S. population (which represents 21,764 patients) these mutations account for a slightly higher percentage of alleles (95%). The U.S. is not a heterogeneous population, though. States such as California have different ethnic demographics with different incidence of CFTR mutations and therefore potentially more uncharacterized mutations. Likewise, the CF patients diagnosed as an adult more often carry CFTR mutations that are "mild," "atypical" or less penetrant for CF. There were 6 of the 122 CF-causing mutations that were only seen in Europe and not in North America. Many other mutations in the CFTR2 database unique or more common to Europe have unknown or variable clinical significance.

The ability to use aggregate data on CFTR mutation frequency is useful from an epidemiologic perspective, but lacks the accuracy needed to counsel individuals. For the entire CFTR2 database, the extent of genetic testing done on patients is not known. The CFTR2 project is actively seeking data from countries where the population with CF may be smaller, but may be enriched for rare mutations. This information will aid the CF community by expanding the list of CF-causing mutations. However, quantitative assessments of allele frequency worldwide or in a specific population will require great expansion of publicly available databases of genetic variation of the general population, as well as more comprehensive genetic analysis of CF patients. Therefore, specific recommendation on which mutations to test for in diagnosis and screening scenarios are best considered in view of local prevalence and testing objectives.

Peter R. Durie, M.D., FRCP(C) 1 The diagnosis of cystic fibrosis (CF) is usually straightforward in the presence of a typical phenotype and sweat chloride concentration exceeding 60 mmol/L. However, since the discovery of the CFTR gene, a spectrum of conditions associated with CFTR mutations has become recognized, particularly in adolescents and adults presenting with sino-pulmonary diseases, pancreatitis and in men with infertility due to obstructive azoospermia. In fact, different mutations in the CFTR gene give rise to a spectrum of ion channel abnormalities, which together with the influence of non-CF modifier genes and environmental factors result in a heterogeneous spectrum of phenotypes in the various organs affected by CF disease. The term "CFTR-related disorder" is used to define symptomatic individuals who are affected by CFTR dysfunction but do not fulfill the current criteria for CF disease. Furthermore, following worldwide acceptance of newborn screening (NBS) for CF disease, a subset of asymptomatic infants is identified with an equivocal diagnosis. Thus the diagnosis of CF may be difficult to establish or exclude in a subset of individuals at the extremes of age.

A number of studies have attempted to determine a cutoff to distinguish attenuated forms of CF from non-CF. However, no single ion channel measure is likely to provide an absolute cut-off since the clinical manifestations and range of severity of CF disease is not only dependent on the degree of CFTR dysfunction but also on non-CFTR genetic and/or environmental modifying factors. In addition, it is unclear whether ion channel testing on different end-organ epithelia reliably concur with one another (e.g., sweat gland, nasal and intestinal epithelium), and whether a combination of ion channel measurements from different epithelial surfaces improves diagnostic performance.

To address some of these questions, we prospectively enrolled a cohort of 202 subjects with single-organ manifestations suggestive of CF, including those with idiopathic chronic respiratory disease, idiopathic pancreatitis and men with infertility due to obstructive azoospermia. We also prospectively enrolled a training (and validation) sample of 263 non-CF and CF subjects (used for the principle component analysis), including healthy controls, obligate heterozygotes and CF patients with pancreatic sufficiency and pancreatic insufficiency. All subjects underwent extensive genotyping and sweat test and nasal potential difference measurements (NPD) were performed on the same day.

The following limitations in diagnostic tests were identified:

(1) Only a small subset of subjects carried 2 CF-causing mutations.

(2) A large percentage of the patients had equivocal sweat chloride and NPD results.

(3) There was considerable disagreement between sweat chloride and NPD results (normal sweat chloride and abnormal NPD and vice versa).

Due to the limitations of existing diagnostic tests, we evaluated whether combinations of ion channel measurements improved the overall diagnostic performance. We used principle component analysis (PCA) of the training sample to identify the optimal combinations of ion channel measurements that distinguished between non-CF and CF subjects. The best performing PCA method completely separated all non-CF and CF subjects in the training sample and appeared to demonstrate excellent diagnostic performance for both CF and CFTR-related disorders in the validation cohort.

In summary, the diagnosis of CF is not always straightforward and may remain inconclusive despite comprehensive evaluation and application of current diagnostic algorithms. In these circumstances, CFTR sequencing is of limited value and the results of ion channel testing may be equivocal and/or discordant. In such circumstances, integration of ion channel measurements by PCA may assist in the diagnostic assessment of patients with an equivocal diagnosis of CF. The California newborn screening program for CF began in July 2007. California's program conducts 40-mutation panel testing of blood spots found to have immunoreactive trypsinogen levels of 62 ng/mL or higher (top 1.5%). When a single mutation (as opposed to zero or two) is found by the panel (in 5.7% of hypertrypsinogenemic newborns, or 0.088% of all newborns), the blood spot is further tested for the presence of other mutations using DNA sequencing. A second non-panel CFTR mutation -defined broadly at the beginning of the program as a 5T variant (any TG Tract), novel variant, mutation associated with CFTR-related disorders (e.g., pancreatitis, male infertility) or other known CF-causing mutation -has been found in 36% of spots sequenced. Parents of newborns with one mutation are offered telephone genetic counseling. Newborns with two or more CFTR mutations are referred to CF centers for follow up. Follow up includes physical examination and sweat chloride, fecal elastase and other testing. When initial signs and symptoms of CF are not present and the mutation phase is unknown, focused parent testing is offered and the cost of genotyping borne by the screening program. When such newborns are found to be carriers, follow up ceases. When mutations are found to be in trans, then newborns continue to be followed up for at least one year with CF centers, regardless of initial or repeat sweat chloride results. The combination of California's (a) diverse population, (b) screening algorithm with comprehensive genotyping, (c) web-based screening information system to collect diagnoses, test results and clinical information, and (d) collaboratively-developed guidelines for CF center follow up of screen positive newborns has made it possible to examine the natural history of CFTR mutations from early in life in a population-based cohort of hypertrypsinogenemic newborns.

Objective: The objective of this paper is to discuss selected findings from the California Newborn Screening Program for CF that have implications on how CF is diagnosed in the era of newborn screening.

Results: In the first 46 months of the program with a minimum of five months follow-up and after screening over 2 million newborns, 819 were screen positive for CF. Of these, 286 have been diagnosed with CF and 287 with CFTR-Related Metabolic Syndrome (CRMS). The remainder have been found to be carriers, lost to follow up, expired or have other resolutions.

Of the 286 CF newborns, 166 had two mutations detected on the panel and 120 had only one detected by the panel and other mutations detected using DNA sequencing. Close to 100% of the 166 had sweat chloride values of 60 mmol/L or higher, confirming the selection of mutations on the panel as clearly CF-disease causing. Around 66% of the 120 were identified after initial sweat testing or examination, with the remainder (14% of all CF cases) identified subsequently.

Of the 287 asymptomatic CRMS newborns, many have at least one mutation that falls into Classes IV-VI. With age, sweat chloride values have climbed to 60 mmol/L or higher in some genotype groups (e.g., Panel Mutation/TG13-5T).

In the first 24 months of the program and after screening over one million newborns, 34 novel CFTR mutations were found in 38 (7.9%) of the 429 CF screen positive newborns. A minimum of five of these novel mutations are highly likely to be CF-disease causing.

Discussion: In California, the prevalence of CRMS in screen positive newborns is similar to that of CF. Many of the genotypes in the CRMS group are similar to those seen in adults diagnosed with CF. Referral to a CF center at an early age may prevent these individuals from misdiagnoses and less than adequate treatments in the future.

Sweat chloride values have been observed to change with age to levels that are consistent with diminished CFTR function. This suggests that long-term follow up is necessary for at least some genotype groups. Newborn screening programs that rely on one sweat test and do not routinely do comprehensive genetic testing may be missing the diagnosis of CF in these newborns.

With efforts like the Clinical and Functional Translation of CFTR Project (www.CFTR2.org) to establish the clinical significance of CFTR mutations, DNA-based CF newborn screening programs will be able to better select CFTR mutations for their panels and CFTR genotypes for diagnostic referral. However, since new mutations are being found at a high rate by the California program and by others worldwide, especially in non-White groups, there will continue to be a lag in our ability to delineate between mutations that are likely to cause CF, another CFTR-related disorder, or no disease at all. Of the close to 2000 mutations already documented, it is certain that more than the current ACMG-23 mutations are CF-caus-ing. Newborn screening and long-term follow-up will be critical in developing this knowledge base in the future. The diagnosis of cystic fibrosis (CF) in adulthood is nearly always precipitated by the presence of one or more characteristic clinical symptoms. In the absence of the diagnosis, patients are often subjected to extensive diagnostic procedures to rule-out other syndromes, and receive fragmented care focused primarily on reducing symptoms without a clear understanding of their underlying pathophysiology. Establishment of the CF diagnosis can result in a rational multisystem approach to care that is more firmly evidenced based, and can eliminate the need for additional diagnostic procedures and subspecialty consultations. The CF diagnosis also allows access to high quality CF centers, and an ever-expanding formulary of medications and therapies not generally available to non-CF patients. Although adult diagnosed patients have significant morbidity, they demonstrate a sustained response to standard CF center care, and can achieve a near normal lifespan (1) .

Genetic Basis of the Adult Diagnosis: Rarely is the adult diagnosis (AD) a result of a "missed" childhood diagnosis (CD) of classic CF. Rather, AD is typically the result of the presence of one or more class IV-VI CFTR mutations. A "high risk" genotype has been defined as the presence of two mutations from Class I-III, and correlates strongly with a classic clinical phenotype (2). A "low risk" genotype is defined by the presence of one (or more) Class IV-VI mutations, and is typically associated with later diagnosis and pancreatic sufficiency (2, 3) . For subjects in our Colorado database of long-term CF survivors (age >40 yrs) with available mutational analysis (87.3%), only 5.3% of the AD cohort had a "high risk" genotype, compared to 83.3% of the CD cohort (p<0.0001) (1). A similar distribution was found in the CFF Patient Registry database (1).

Prevalence and Trends: Although still a small segment of the CF population, the number of AD patients has increased rapidly since CFTR genotype analysis became widely available. From 1995 From -2005 .3% of the 9766 new cases of CF reported within the CFF Patient Registry were in individuals >18 yrs (4) . This percentage has grown with every year, and is reflected in a steady increase in the mean age of the CF diagnosis, which has increased from 3.2 yrs in 2000 to 3.6 yrs in 2010. This trend has occurred despite a large increase in diagnosis through newborn screening, which has risen from 8.0% to 57.5% of cases over the same 10-year interval (5).

In our experience, the adult CF diagnosis is typically preceded by several years of clinical decline, resulting in an escalation of physician visits, diagnostic testing and subspecialty referrals (6) . The majority (70.6%) of AD patients present with respiratory symptoms (4). Usually cough and sputum production have been present for years, and attributed to asthma or COPD based on clinical features. The detection of recurrent or persistent airway infections is often the event that ultimately triggers testing for CF. In nearly all, radiographic evaluation by HRCT demonstrates bronchiectasis, even in the presence of a normal FEV 1 . The most common airway pathogens at diagnosis are P. aeruginosa and S. aureus, although these are present at a lower frequency than age-matched CD controls (4, 7) . AD patients also appear to have a significantly greater prevalence of nontuberculous mycobacteria (NTM) (4,7), although it is not clear to what extent this finding is due to the much greater age of this cohort. Sinus disease is nearly universal in this population. Other phenotypic features that may lead to diagnosis include male infertility, and bowel symptoms that may include steatorrhea and/or idiopathic pancreatitis (4, 7) . Classic pancreatic insufficiency is not present in the majority of AD patients at diagnosis, but can develop over time. Important factors in establishing CF as the cause of the patient's symptoms include the patient's persistence in seeking a diagnosis, as well as access and willingness to utilize the health care system; these factors may contribute to the observation that the adult diagnosis is made more often in women (1, 7) .

Setting of Care: We have found that the majority of AD patients do not attend CF centers. CF patients > age 40 yrs seen at National Jewish Health from 1992 to 2008 were cross-referenced with the CFF Registry to evaluate for CF center attendance. Analysis of the six most recent years for which each subject was eligible for entry into the CFF Registry demonstrated that on average, 60% of CD patients were seen at a CF center during any given year, compared to 41.7% of AD patients (p<0.001) (1). We also compared the number of deaths captured in the CFF Registry from 1992 to 2005 (n=5,589) to the number of deaths in which CF was listed as the cause of death, or contributing condition in the CDC Multiple Cause-of-Death Record (MCOD) for the corresponding years (n=6,866). The number of deaths for patients aged 1-14 years recorded in the CFF Registry was nearly identical to the MCOD database; however, deaths of infants (age <1 yr) and patients more than 14 years of age were significantly different between the two databases. For patients older than 45 years, which would encompass most of the AD cohort, only 45.9% of the deaths over the 14-year span were captured in the CFF Registry (1).

Response to CF Center Care: Despite differences in the spectrum of infections, organ involvement and CFTR mutations, the response of AD patients to current treatment strategies is very positive. Change in FEV 1 following diagnosis at the Colorado CF Center (n=32) was analyzed by comparing FEV 1 measured at the first CF clinic visit following a new diagnosis to the subject's best FEV 1 for each subsequent 12-month interval. Following diagnosis, FEV 1 improved 7.8+2.2% within 12 months and 11.4+2.3% within 24 months. Over a 4-year period, CF center care resulted in a sustained improvement in FEV 1 over baseline (P=0.0013) (1) . An identical analysis of the CFF Registry database between 1992-2007 also demonstrated a significant improvement in FEV 1 in newly diagnosed adults (n=308) who received care at CF centers nationwide (1). Following diagnosis, FEV 1 increased 9.7+1.9% within 12 months and 10.1+2.9% within 24 months, and was sustained for a 4-year period (p<0.0001). Corresponding to improved lung function, we surveyed AD patients (n=30) at our center, and found strong agreement that CF care was improving their lung symptoms and quality of life, and that the benefits outweighed the additional treatment burden.

Long-term Outcomes and Survival: While CF is generally referred to as "mild" in AD patients, little is known concerning long-term outcomes in this population. FEV 1 values from all clinic visits between 1992 and 2007 (n=35,128) captured in the CFF Registry for patients over 40 years demonstrated that while AD patients have a significantly higher FEV 1 at age 40 years (57.9+3.0%) compared with CD patients (50.0+3.1%; P<0.01), the annual age-related FEV 1 decline after age 40 was not different between the two cohorts (CD, -0.50+0.07% vs. AD, -0.36+0.07%; P=0.19) (1). As expected, median AD survival greatly exceeded CD survival (1) . For patients in the Colorado database who survived past age 40 years, the median CD cohort survival was 51.7 years, versus 76.9 years for the AD cohort (P<0.0001). For patients in the CFF Registry database, the median survival for the CD cohort was 53.5 years, versus 68.2 years for the AD cohort (P<0.0001). However, despite a significantly longer life span, the frequency of death from respiratory failure or transplantation-related complications was 85% for AD, compared to 87% for CD (1) .

Conclusions: The CF diagnosis can significantly improve the lives of adults with progressive bronchiectasis and other manifestations of the disease. AD patients treated at CF care centers demonstrate sustained improvement in FEV 1 , and report a high level of satisfaction with their care. Yet barriers exist that result in a low rate of attendance at CF centers and thus underrepresentation in the CFF Patient Registry. When the clinical consequence of adult diagnosed CF is viewed over the lifetime of the patient, it is apparent that the pulmonary manifestations are not a mild form of the disease, but rather a delayed but equally serious form of the childhood presentation. Thus when diagnosis and treatment are delayed, the morbidity of nonclassic CF can be unnecessarily severe. Although previously regarded as medical curiosities, AD patients represent an important and growing segment of the CF population that deserves increased study (6) .

Supported by the CFF and the Rebecca Runyon Bryan Chair for Cystic Fibrosis. CF patients experience episodic worsening of symptoms commonly referred to as pulmonary exacerbations. Although a standardized definition of a pulmonary exacerbation is lacking, clinical features may include increased cough, increased sputum production, shortness of breath, chest pain, loss of appetite, loss of weight, and lung function decline (1). In school age children, adolescents and adults, pulmonary exacerbations have been shown to adversely impact quality of life (2) and are a major contributor to the cost of care (3). Despite current interventions, they can be associated with irreversible loss of lung function (4) and predict poorer survival (5) . Thus, prevention of exacerbations is of great interest, and time to first exacerbation or exacerbations rate have been common clinical trial endpoints in trials of chronic maintenance therapies enrolling CF patients over 6 years of age (6) (7) (8) . Pulmonary exacerbation is a clinical endpoint defined as affecting how a person feels, functions or survives, that complements surrogate endpoints such as lung function (9) .

In children with CF <6 years of age, there is an emerging role of pulmonary exacerbations as a clinical trial endpoint. A definition specifically for this young age range was developed by an expert panel based on exacerbation definitions for older CF patients (Table) . This definition has been used with slight modifications in three published studies (10) (11) (12) . While no difference between treatment groups in this endpoint was seen in the EPIC or ISIS trials, a 50% relative reduction in exacerbation rate using this definition was seen among Pseudomonas-negative children with CF randomized to azithromycin as compared to placebo (11) .

While pulmonary exacerbation is an attractive endpoint in young children because it is widely accepted as an important clinical endpoint in older CF patients, has a standardized definition in young children, and has been used in many pivotal CF clinical trials, it also has important limitations in these young children. Essentially all of the experience with the use of pulmonary exacerbation as an endpoint in CF clinical trials comes from patients >6 years of age with established lung disease, in whom the goal is to reverse or slow the progression of existing pathology. Pulmonary exacerbations in young children with milder structural airway damage are likely to be quite different. In children with CF <6 years of age, there is a much higher prevalence of viral respiratory infections and lower rates of chronic endobronchial infection, particularly with Pseudomonas aeruginosa. Furthermore, treat-ment of exacerbations is different, with a greater prevalence of oral or inhaled antibiotics and lower prevalence of intravenous antibiotics. The long-term sequelae of pulmonary exacerbations in children <6 in terms of effects on quality of life, lung function and survival are poorly understood. Thus, the impact of pulmonary exacerbations as a clinical trial endpoint in young children with CF will evolve as our understanding of the pathophysiology and outcomes of these exacerbations improves. (10)) Major Criteria > 1 establishes a pulmonary exacerbation 1) Decrease in FEV1 ≥ 10% from best baseline within past 6 months 2) Oxygen saturation < 90% on room air or ≥ 5% decline from previous baseline 3) New lobar infiltrate(s) or atelectasis on chest x-ray 4) Hemoptysis (more than streaks on > 1 occasion in past week)

Minor Criteria > 2 with duration of 3 days establishes a pulmonary exacerbation 1) Increased work of breathing or respiratory rate 2) New or increased adventitial sounds on lung exam 3) Weight loss ≥ 5% or decrease across 1 major percentile for age in past 6 months 4) Increased cough 5) Decreased exercise tolerance or level of activity 6) Increased chest congestion or change in sputum Background: To advance care via clinical and translational research, reliable and valid assessments of lung disease status in toddlers and preschoolers with cystic fibrosis are needed. At present, assessments of infants and of children over age 6 in research protocols often rely on pulmonary function testing, and, to a much lesser extent, have involved computed tomography (CT) scans of the chest. For children from age 2 to 6, PFT assessments are challenging and rarely used in clinical research protocols. CT scans offer an opportunity to assess lung disease status in this age group. Of note, CT scans of the chest have been shown to be more sensitive to intercurrent changes in lung disease than PFTs; however, little data are available in children ages 2 to 6. As part of an NIH-funded, multi-site, randomized clinical trial of behavior and nutrition treatment for toddlers and preschoolers with CF, an ancillary study (funded by the CF Foundation) using CT scans of the chest was completed. The ancillary study afforded the opportunity to obtain CT data at baseline and 18 months later, providing what is to our knowledge, the first longitudinal CT scan data set of patients with CF in this age range. During this workshop, we will present information related to training the study radiologists on the CT scoring system and the protocol followed during the actual, blinded scoring of the scans obtained in this study. Future work will involve assessment of inter-rater reliability across the study radiologists and determination of lung disease status, as well as any change over time, for the sample of toddlers and preschoolers who completed the project.

Training CT Scans: A set of 10 CT scans of young children with varying levels of CF lung disease.

CT Scoring Panels: A set of CF CT scans demonstrating specific lung disease abnormalities obtained from the Training CT Scans. Two panels were developed containing lung disease abnormalities identified by the radiologist and research coordinator to represent varying levels of correlation between the two radiologists' scores.

Consensus Scoring: At the beginning of each scoring session, the research coordinator selected one CT Scoring Panel. The radiologists reviewed and scored each lung abnormality and then came to agreement with the anchor consensus score determined during the training process. This helped to ensure calibration of the radiologists to consensus on known scans immediately before independently evaluating the blinded research study CT images.

Stephanie D. Davis, M.D. 1 Since the late 1960s physiologic measures have been employed to describe the natural history of early CF lung disease (1) . With the advent of newborn screening and improved therapeutics, early airway disease has become more difficult to identify due to it being milder at diagnosis. However, the ability to perform adult type spirometry during infancy (2) and the preschool years (3) has allowed investigators to better understand and identify early lung disease. Recently, infant lung function testing has also been used as an exploratory endpoint in a therapeutic trial (4). These advances over the past decade are now leading to novel approaches to clinical management and research study design.

An ideal endpoint must discriminate disease from health; in early CF this endpoint must be sensitive enough to detect "silent" peripheral airway disease. Flow limitation, diminished forced expiratory volumes and hyperinflation associated with air trapping have been demonstrated in infants and preschoolers with CF (5-8). These findings have been reported in those diagnosed following newborn screening as well as in infants presenting without respiratory symptoms (6, 9) . In fact, nearly 70% of infants tested at baseline have been reported to have an elevated functional residual capacity and 25% of this same cohort had diminished flows in a large US study (5) . Longitudinal follow-up of infants with CF has shown a drop in lung function values over time and diminished values have been reported to occur as early as 6 months (6). As infants with CF are thought to be born with normal airways the "window" for early therapeutics may ideally be in those < 6 months of age.

Lower airway inflammation, infection, clinical symptoms and airway wall thickening have been reported to be associated with lung function measurements in young children with CF. Forced expiratory flows and volumes have been demonstrated to be associated with percent neutrophils and neutrophil elastase as measured in bronchoalveolar lavage fluid in infants with and without respiratory symptoms (10, 11) . Two recent studies reported an association between abnormal lung function indices and infection with Staphylcoccus aureus. In a 10 center study evaluating infant lung function over a 1 year period, hyperinflation as assessed through functional residual capacity was associated with Staphylcoccus aureus (12) . Staphylcoccus aureus was also associated with greater decline in infant lung function over time (11) . The presence of Pseudomonas aeruginosa has also been shown to be associated with diminished lung function measured from infancy through the preschool age group (7,13) and decline over time in infants (11) . Despite these associations causality cannot be inferred from the available data and lower airway infections may simply be a marker of early disease rather than the etiology of these abnormal physiologic findings. The presence of cough has also been associated with diminished flows in both a U.S. and British population (7, 12) . Finally, diminished FEV 0.5 (forced expired volume at 0.5 seconds) values were associated with increased airway wall thickening measured on CT scans of the chest in infants with CF (14) . These validation studies are important when defining infant and preschool lung function testing as an endpoint.

The feasibility of performing lung function testing in a multicenter setting is especially important in rare pediatric lung diseases since therapeutic trials often need several centers to attain adequate sample size. The first multicenter trial using the raised volume technique with multiple testing centers was performed in the U.S. and these results highlighted that expertise with the technique was critical. In sites with extensive experience, > 80% of flow-volume curves were research acceptable; however, sites with less experience reported much lower acceptable percentages (5) . Infant lung function tests did appear safe in this large trial with emesis of chloral hydrate being the most commonly reported adverse event. Finally, the intraclass correlations for the infant lung function indices 1 month apart were quite high, as high as 0.90 for FEV 0.5 (5) . Preschool lung function testing has also been performed in a multicenter setting; however, lower overall acceptability was reported for preschool spirometry compared to infant lung function testing (15) .

The ability of the endpoint to detect changes preand post-therapeutic intervention is critical. In a retrospective study of 10 CF infants, forced expiratory flows and lung volumes significantly improved after an antibiotic course for a pulmonary exacerbation (16) . Recently, the change in FEV 0.5 when adjusted for baseline lung function, height, weight, gender and age was reported to be significantly greater in 22 infants treated with 7% hypertonic saline for 1 year compared to 23 infants treated with isotonic saline for 1 year (4). More data are needed to evaluate the efficacy of using infant and preschool lung function testing as an endpoint in a therapeutic trial, but these initial results are encouraging and indicate that studies using infant lung function as an outcome measure are feasible.

In conclusion, recent advancements in infant and preschool lung function testing highlight the potential applicability of these physiologic measures as validated endpoints for therapeutic trials. With the advent of newborn screening and novel therapeutics such as ivacaftor, defining and validating these tools has become more important than in past decades. International infant and preschool lung function trials are likely forthcoming, thereby highlighting the importance of standard operating procedures and standardized equipment. As we enter this new era of CF care, it is likely that infant and preschool lung function testing coupled with other endpoints such as imaging and measures of ventilation inhomogeneity will better define this early "silent" disease. Lung disease in CF has been demonstrated to start early in infancy, but is not necessarily associated with significant clinical signs and symptoms. Quantifying and tracking early lung disease with sensitive markers has become an area of significant interest both for clinical care and to enable early intervention trials in infants and young children. Early lung disease is focal and charac-terized by mucus plugging and air trapping which may not be captured by global measures of airflow obstruction. Multiple breath washout (MBW) derived measures of lung function such as the lung clearance index (LCI) have been shown to be abnormal in a high percentage of CF patients with normal spirometry and LCI in preschool children predicts future lung function (1, 2) . We have per-formed two interventional studies to date in school age children to assess the ability of MCI to detect a treatment of interventions known to be beneficial for CF patients: hypertonic saline and dornase alfa (3, 4) . Both studies were conducted as double blind randomized controlled trials in a crossover design with intervention periods of 4 weeks in CF patients with an FEV1 of 80% predicted or above and demonstrated significant improvements in LCI that were not detected using FEV1 as an outcome measure. Recently, a third study with ivacaftor in patients carrying at least one allele of a G551D mutation was performed as the first interventional study with LCI as the primary outcome measure in a multicenter setting (5). This study used a similar crossover design and included patients with an FEV1 of at least 90% predicted. Results support the concept that LCI is more sensitive than FEV1 with a treatment effect size that exceeded that of the previous studies with hypertonic saline and dornase alfa.

We also assessed the utility of LCI as part of the Infant Study of Inhaled Saline (ISIS) (6) . Twenty-four of 26 (92%) of the subjects recruited at the Toronto site had paired baseline and follow-up measures at one year. There was a significant reduction in LCI from baseline to follow-up in the hypertonic saline arm; while LCI remained stable in the isotonic saline group. This change was primarily driven by changes in the preschool children. Infants treated with hypertonic saline LCI did not change, where it increased significantly in the isotonic saline group. While the treatment effect was not significant in the overall group (p=0.08), the effect size was higher than in the previously performed short term study in older children (3) . This supports the concept that LCI may be a suitable endpoint for interventional studies in this age group as well.

One of the challenges of moving MBW technology forward as an outcome measure is the lack of standardized equipment. Mass spectrometry based technology using SF 6 as a tracer gas is considered the "gold standard" for measuring LCI, but is only available in a few centers around the world. In an effort to validate a nitrogen based multiple breath washout system, we recently studied 60 healthy children and 60 children with CF on both technologies. We found LCI measured by nitrogen washout to be higher compared with mass spectrometry, but no systematic bias was observed in the differences between these two systems. This difference in LCI between the two MBW systems was more pronounced in CF compared to healthy children and not associated with weight-for-age or height-for-age centiles. However, when compared to FEV1 (percent predicted), the difference between the two systems was greatest for lower values of FEV1. These results suggest that LCI measured by the nitrogen system is more sensitive at detecting lung abnormalities than SF 6 based measurements in children with CF. The data demonstrate the validity of the nitrogen washout system for LCI measurement in school age children and form the basis for subsequent validation studies in preschool children.

While LCI is the main parameter measured by multiple breath washout in CF lung disease to date, it is affected by changes in breathing pattern which may impact on test variability especially in infants and young children. Moment analysis of MBW parameters assesses the shape (skewness) of the washout curve and has been shown to be less affected by variability in breathing frequency or tidal volume. Moment analysis is performed by setting the starting gas concentration at 100% and by plotting end-tidal gas concentrations against the number of lung turnovers (FRCs). Moment 0 is the area under this curve while moments 1 and 2 represent single and double multiplications with the number of lung turnovers. Results are then expressed as moment ratios (M1/M0 and M2/M0) with higher moment ratios being more reflective of ventilation inhomogeneity (VI) in the periphery of the lung. We used two data sets to assess the potential of moment ratio analysis to quantify VI in CF patients. First, we assessed the sensitivity of moment ratios compared to LCI to differentiate health and disease in a cross-sectional validation study of LCI equipment in both healthy and CF children. Subsequently, we reanalysed data from a recently conducted interventional trial (3) to investigate whether moment ratios were able to detect treatment effects of hypertonic saline similar to LCI. In the cross-sectional study comparing health to disease, both LCI the first moment ration were approximately 7 z-scores higher in CF compared with healthy children, whereas the second moment ratio was 12 zscores higher, suggesting greater sensitivity for this parameter compared to LCI. In the interventional crossover trial on hypertonic saline, treatment effects were observed for both moment ratios. These findings suggest that moment ratios may provide a complementary outcome to the LCI in clinical trials that should be further explored in the future.

Chemical chaperones correct the mutant phenotype of the delta F508 CFTR protein

Correction of the F508del-CFTR protein processing defect in vitro by the investigational drug VX-809

What is the role of protein aggregation in neurodegeneration?

Mitochondria and neurodegeneration

Biological and chemical approaches to diseases of proteostasis deficiency

The association between socioeconomic status and gender with median age at death from CF in England and Wales: 1959 to

Exposure of children with CF to environmental tobacco smoke

Interactions between secondhand smoke and genes that affect CF lung disease

Effect of ambient air pollution on pulmonary exacerbations and lung function in CF

Infection, inflammation and lung function decline in infants with CF

Risk factors for initial acquisition of Pseudomonas aeruginosa in children with CF identified by newborn screening

Effects of viral lower respiratory tract infection on lung function in infants with CF. Pediatrics

Stress and asthma

Family correlates of a 10-year pulmonary health trend in CF

Anxiety, depression, and life satisfaction in parents caring for children with CF

Anxiety and depression in CF

Growth and nutritional indexes in early life predict pulmonary function in CF

Recovery of birth weight z score with CF

Gender gap in CF mortality

Is there still a gender gap in CF?

Body image and dieting behavior in CF

Race, ethnicity, and health: can genetics explain disparities?

Infection control recommendations for patients with CF: microbiology, important pathogens and infection control practices to prevent patient-to-patient transmission

Survey of infection control policies for patients with CF in the U.S

guideline for isolation precautions: preventing transmission of infectious agents in healthcare settings

Surgical mask vs N95 respirator for preventing influenza among health care workers: a randomized trial

Mask use, hand hygiene and seasonal influenza-like illness among young adults: a randomized intervention trial

Face mask use and control of respiratory virus transmission in households

Face masks to prevent transmission of influenza virus: a systematic review

Bacterial contamination of CF clinics

Defining routes of airborne transmission of P. aeruginosa in people with CF

Is Burkholderia (Pseudomonas) cepacia disseminated from CF patients during physiotherapy?

A 1-m distance is not safe for children with CF at risk for cross-infection with P. aeruginosa

Airborne transmission of communicable infection-the elusive pathway

References: 1. Durie PR. Pancreatic aspects of CF and other inherited causes of pancreatic dysfunction

CF of the pancreas and its relation to celiac disease: clinical and pathological study

Pathology of CF review of the literature and comparison with 146 autopsied cases

Ultrastructural changes of the pancreas and liver in CF

Pathological confirmation of CF in the fetus following prenatal diagnosis

Decline of exocrine pancreatic function in CF patients with pancreatic sufficiency

Pancreatic function in infants identified as having CF in a neonatal screening program

Pancreatic insufficiency, growth, and nutrition in infants identified by newborn screening as having CF

CF of the pancreas with varying degrees of pancreatic insufficiency

Assessment of pancreatic function in screened infants with CF

Uses and abuses of enzyme therapy in CF

Pancreatic enzyme therapy and clinical outcomes in patients with CF

Improved respiratory prognosis in patients with CF with normal fat absorption

Relative underweight in CF and its prognostic value

Disruption of the CFTR gene produces a model of CF in newborn pigs

Pathology of gastrointestinal organs in a porcine model of CF

CF pigs develop lung disease and exhibit defective bacterial eradication at birth

The DeltaF508 mutation causes CFTR misprocessing and CF-like disease in pigs

An activated immune and inflammatory response targets the pancreas of newborn pigs with CF

Pancreatic damage in fetal and newborn CF pigs involves the activation of inflammatory and remodeling pathways

Recurrent acute pancreatitis in patients with CF with normal pancreatic enzymes

Type of CFTR mutation determines risk of pancreatitis in patients with CF

CF of the pancreas. Morphologic findings in infants with and without diagnostic pancreatic lesions

Identification of the CF gene: cloning and characterization of complementary DNA

Chloride impermeability in CF

Selective activation of CFTR Cl -and HCO 3 -conductances

Correction of the CF defect in vitro by retrovirus-mediated gene transfer

Generation of cAMP-activated chloride currents by expression of CFTR

Purification and functional reconstitution of the CFTR

CFTR as a cAMP-dependent regulator of sodium channels

Physiology of duct cell secretion

Cell physiology of pancreatic ducts

Quantitative evaluation of the development of the exocrine pancreas in CF and control infants

Longitudinal evaluation of serum trypsinogen measurement in pancreatic-insufficient and pancreatic-sufficient patients with CF

Pancreatic function testing in CF

Evidence for a primary defect of pancreatic HCO3-secretion in CF

Independence of apical Cl -/HCO 3 -exchange and anion conductance in duodenal HCO 3 -secretion

Mechanisms of bicarbonate secretion in the pancreatic duct

Intestinal pH and gastrointestinal transit profiles in CF patients measured by wireless motility capsule

Drug therapies for reducing gastric acidity in people with CF

Altered transit and bacterial overgrowth in the CF mouse small intestine

Effect of antibiotic treatment on fat absorption in mice with CF

Intestinal bile acid malabsorption in CF

Oral antibiotic therapy improves fat absorption in CF patients with small intestine bacterial overgrowth

Evidence of intestinal inflammation in patients with CF

Best practice guidance for the diagnosis and management of CF-associated liver disease

Chapter 43: Formulation Challenges of Powders for the Delivery of Small Molecular Weight Molecules as Aerosols

Delivery of Drugs by the Pulmonary Route

Physical characterization of component particles included in dry powder inhalers. I. Strategy review & static characteristics

Physical characterization of component particles included in dry powder inhalers. II. Dynamic characteristics

Physicochemical Characterization & Water Vapor Absorption of Organic Solution Advanced Spray Dried Trehalose Microparticles & Nanoparticles for Targeted Dry Powder Pulmonary Inhalation Delivery

Raman Characterization and chemical imaging of biocolloidal self-assemblies, drug delivery systems and pulmonary inhalation aerosols. A review

Nanomedicine in pulmonary delivery

Advances in Microscopy & Complementary Imaging Techniques to Assess the Fate of Drugs Ex Vivo in Respiratory Drug Delivery

Special Theme Issue-Computational & Visualization Approaches in Respiratory Delivery)

Book Chapter 4: Nanoparticle Lung Delivery & Inhalation Aerosols for Targeted Pulmonary Nanomedicine

Chapter 6: Design & Development of Approved Nanopharmaceutical Products

Lipid Nanoparticulate Drug Delivery & Nanomedicine

Therapeutic liposomal dry powder inhalation aerosols for targeted lung delivery

Surface Analytical Techniques in Solid-State Particle Characterization: Implications for Predicting Performance in Dry Powder Inhalers

Particle interactions in dry powder inhaler unit processes

Pulmonary Inhalation Aerosols for Targeted Antibiotics Drug Delivery

Aerosolized tacrolimus in a lung transplant recipient

Pre-procedural planning

References: 1. Jones DP, Go Y-M

References: 1. Southern KW, Merelle MM, Dankert-Roelse JE, et al. Newborn screening for CF

Is newborn screening for CF a basic human right?

Nutritional benefits of neonatal screening for CF. Wisconsin CF Neonatal Screening Study Group

European best practice guidelines for CF neonatal screening

A survey of newborn screening for CF in Europe

The evolution of airway function in early childhood following clinical diagnosis of CF

Computed tomography reflects lower airway inflammation and tracks changes in early CF

Multicenter evaluation of infant lung function tests as CF clinical trial endpoints

Progression of early structural lung disease in young children with CF assessed using CT

Monitoring of structure and function in early CF lung disease

Lung function from infancy to the preschool years after clinical diagnosis of CF

Infection, inflammation, and lung function decline in infants with CF

Change in pulmonary function following IV antibiotics in infants with CF

Lung disease at diagnosis in infants with CF detected by newborn screening

Bronchiectasis in infants and preschool children diagnosed with CF after newborn screening

Structural airway abnormalities in infants and young children with CF

High-resolution computed tomography imaging of airway disease in infants with CF

Lung Clearance Index and HRCT are complementary markers of lung abnormalities in young children with CF

Nutritional benefits of neonatal screening for CF. Wisconsin CF Neonatal Screening Study Group

Pigs and humans with CF have reduced insulin-like growth factor 1 (IGF1) levels at birth

Evolution of pulmonary inflammation and nutritional status in infants and young children with CF

Acquisition of P. aeruginosa in children with CF

Lower airway inflammation in infants and young children with CF

Effect of bronchoalveolar lavage directed therapy on P. aeruginosa infection and structural lung injury in children with CF

European best practice guidelines for CF neonatal screening

CF pulmonary guidelines: chronic medications for maintenance of lung health

Inhaled hypertonic saline in infants and children younger than 6 years with CF: the ISIS randomized controlled trial

Dornase alfa for cystic fibro

Effect of aerosolized rhDNase (Pulmozyme) on pulmonary colonization in patients with CF

Association of Chartered Physiotherapists in CF. Standards of care and good clinical practice for the physiotherapy management of CF

Guideline on the design and conduct of CF clinical trials: The European CF Society-Clinical Trials Network (ECFS-CTN)

Mechanisms of antibiotic resistance in bacterial biofilms

Reducing mortality in sepsis: new directions

New trends in emerging pathogens

Quorum-sensing signals indicate that CF lungs are infected with bacterial biofilms

Iron and infection: the heart of the matter

Lipocalin 2 mediates an innate immune response to bacterial infection by sequestrating iron

The effect of passage and iron on the virulence of P. aeruginosa

Hepcidin in iron metabolism

Iron in biological systems

Bacterial biofilms: an emerging link to disease pathogenesis

Iron and P. aeruginosa biofilm formation

Iron availability influences aggregation, biofilm, adhesion and invasion of P. aeruginosa and B. cenocepacia

Biofilms, antimicrobial resistance, and airway infection

Loss of microbicidal activity and increased formation of biofilm due to decreased lactoferrin activity in patients with CF

Iron sequestration by human lactoferring stimulates P. aeruginosa surface motility and blocks biofilm formation

Antibacterial drug discovery: is it all downhill from here?

CF pulmonary guidelines: chronic medications for maintenance of lung health

Treatment of early P. aeruginosa infection in patients with CF: the ELITE trial

Comparative efficacy and safety of 4 randomized regimens to treat early P. aeruginosa infection in children with CF

Adherence to tobramycin inhaled solution and health care utilization

Pharmacokinetics and bioavailability of aerosolized tobramycin in CF

Singledose pharmacokinetics of aerosol MP-376 (levofloxacin solution for inhalation) in CF patients: PK-PD implications

Pharmacokinetics and safety of MP-376 (levofloxacin inhalation solution) in CF subjects

CF Foundation. CF Foundation patient registry 2010 annual data report to the center directors

P. aeruginosa and other predictors of mortality and morbidity in young children with CF. Pediatr Pulmonol

The changing microbial epidemiology in CF

High treatment burden in adults with CF: challenges to disease self-management

The long-term use of inhaled tobramycin in patients with CF

Antibiotic therapy against P. aeruginosa in CF: a European consensus

Levofloxacin inhalation solution

References: 1. Radloff LS. The CES-D: A self-report depression scale for research in the general population

Manual for the State-Trait Anxiety Inventory

Attachment Q-Sort (Revision

Task-oriented and bottle feeding adversely affect the quality of mother-infant interactions after abnormal newborn screens

The International Depression and anxiety Epidemiological Study (TIDES): preliminary results from the

Quality of life in patients with CF: association with anxiety and depression

The association between depression, lung function and health-related quality of life among adults with CF

Prevalence and impact of depression in CF

Depressive symptoms in children with CF and parents and its effects on adherence to airway clearance

Symptom features of postpartum depression: are they distinct? Depress Anxiety

Antepartum and postpartum depression: healthy mom, healthy baby

The impact of post-natal depression and associated adversity on early mother-infant interactions and later infant outcome

The CFTR impedes proteolytic stimulation of the epithelial Na + channel

Liquid movement across the surface epithelium of large airways

SPLUNC1 regulates airway surface liquid volume by protecting ENaC from proteolytic cleavage

CF: impaired bicarbonate secretion and mucoviscidosis

A new role for bicarbonate in mucus formation

Influence of external pH on ciliary beat frequency in human bronchi and bronchioles

Wound hypoxia and acidosis limit neutrophil bacterial killing mechanisms

Novel role for pendrin in orchestrating bicarbonate secretion in CFTR-expressing airway serous cells

Abnormal surface liquid pH regulation by cultured CF bronchial epithelium

Hyperacidity of secreted fluid from submucosal glands in early CF

Acid and base secretion in freshly excised nasal tissue from CF patients with ∆F508 mutation

Identification of pendrin as a common mediator for mucus production in bronchial asthma and chronic obstructive pulmonary disease

Airway surface liquid volume regulation determines different airway phenotypes in liddle compared with betaENaC-overexpressing mice

Role of cilia, mucus and airway surface liquid in mucociliary dysfunction: lessons from mouse models

Allergic airway inflammation induces a pro-secretory epithelial ion transport phenotype in mice

Functional characterization of three novel tissue-specific anion exchangers SLC26A7, -A8, and -A9

SLC26A9 is a constitutively active, CFTR-regulated anion conductance in human bronchial epithelia

The porcine lung as a potential model for CF

Disruption of the CFTR gene produces a model of CF in newborn piglets

The ∆F508 mutation causes CFTR misprocessing and CF-like disease in pigs

cAMP stimulates bicarbonate secretion across normal, but not CF airway epithelia

Loss of anion transport without increased sodium absorption characterizes newborn porcine CF airway epithelia

Mucociliary clearance as an outcome measure for CF clinical research

Regional mucociliary clearance in patients with CF

Mucus clearance and lung function in CF with hypertonic saline

The effect of inhaled mannitol on bronchial mucus clearance in CF patients: A pilot study

Absorptive clearance of DTPA as an aerosol-based biomarker in the CF airway

Airway surface liquid volume regulates ENaC by altering the serine protease-protease inhibitor balance: A mechanism for sodium hyperabsorption in CF

Meta-analysis: high-dosage vitamin E supplementation may increase all-cause mortality

References: 1. Flume PA et al. Cystic fibrosis pulmonary guidelines: chronic medications for maintenance of lung health

Therapy for CF-the end of the beginning?

Pulmonary exacerbations in CF

Explanatory and pragmatic attitudes in therapeutical trials

Reporting of noninferiority and equivalence randomized trials: an extension of the CON-SORT statement

Comparative effectiveness research: what happened to incorporating costs of care?

Improving the reporting of pragmatic trials: an extension of the CONSORT statement

Patient registry analyses: seize the data, but caveat lector

Emerging methods in comparative effectiveness and safety: symposium overview and summary

A most stubborn bias: no adjustment method fully resolves confounding by indication in observational studies

The role of cost-effectiveness analysis in health and medicine. Panel on Cost-Effectiveness in Health and Medicine

Guidelines for authors and peer reviewers of economic submissions to the BMJ. The BMJ Economic Evaluation Working Party

Health outcomes in economic evaluation: the QALY and utilities

Legislating against use of cost effectiveness information

S18.2 RELATIONSHIPS BETWEEN ADHERENCE TO PULMONARY MEDICATIONS & HEALTH CARE COSTS: LONGITUDINAL ANALYSES FROM

Ph.D. 1 , Kristin Riekert, Ph.D. 2 and Jie Zhang

References: 1. CSDH. Closing the gap in a generation: health equity through action on the social determinants of health

Health inequalities and CF

International comparison of median age at death from CF

Fitzsimmons SC. The association of socioeconomic status with outcomes in CF patients in the U.S

Non-genetic influences on CF lung disease: the role of sociodemographic characteristics, environmental exposures and health care interventions

Median household income and mortality rate in CF

Effects of social class, sex, and region of residence on age at death from CF

Association between socioeconomic status, sex, and age at death from CF in England and Wales (1959 to 2008): cross sectional study

Association of socioeconomic status with the use of chronic therapies and healthcare utilization in children with CF

Socioeconomic status and the likelihood of antibiotic treatment for signs and symptoms of pulmonary exacerbation in children with CF

Socioeconomic status and risk of hospitalization among individuals with CF in Ontario

Analysis of Longitudinal Data

Understanding the natural progression in %FEV1 decline in patients with CF: a longitudinal study

Long term azithromycin in children with CF: a randomised, placebo-controlled crossover trial

Azithromycin in patients with CF chronically infected with P. aeruginosa: a randomized controlled trial

Long term effects of azithromycin in patients with CF: A double blind, placebo controlled trial

Effect of azithromycin on pulmonary function in patients with CF uninfected with P. aeruginosa: a randomized controlled trial

Marshall BC. Heterogeneity of treatment response to azithromycin in patients with CF

Azithromycin blocks autophagy and may predispose CF patients to mycobacterial infection

Pathophysiology and management of pulmonary infections in CF

Schidlow DV and the CF Foundation Prednisone Trial Group. A multicenter study of alternate-day prednisone therapy in patients with CF

Effect of high dose ibuprofen inpatients with CF

CF pulmonary guidelines: chronic medications for maintenance of lung health

Long term azithromycin in children with CF: a randomised, placebo-controlled crossover trial

Effect of azithromycin on pulmonary function in patients with CF uninfected with P. aeruginosa: a randomized controlled trial

Maintenance azithromycin treatment in pediatric patients with CF: long-term outcomes related to macrolide resistance and pulmonary function

Azithromycin in CF

Long-term, low dose azithromycin treatment reduces the incidence but increases macrolide resistance in S. aureus in Danish CF patients

Azithromycin blocks autophagy and may predispose CF patients to mycobacterial infection

Chest computed tomography scores are predictive of survival in patients with CF awaiting lung transplantation

Detecting early structural lung damage in CF

Progressive damage on high-resolution computed tomography despite stable lung function in CF

Discussion on the stethoscope versus X-rays

CF: scoring system with thin-section CT

CF in children: HRCT findings and distribution of disease

Chest radiology in cystic fibrosis: is scoring useful

Chest computed tomography scans should be considered as a routine investigation in CF

Bronchiectasis in infants and preschool children diagnosed with CF after newborn screening

Effect of bronchoalveolar lavage-directed therapy on P. aeruginosa infection and structural lung injury in children with CF

Lung morphology assessment using magnetic resonance imaging: a robust ultrashort TR/TE 2D steady state free precession sequence in CF routine follow-up

Estimation of cancer mortality associated with repetitive computed tomography scanning scanning in CF

Lung disease at diagnosis in infants with CF detected by newborn screening

Progression of lung disease on computed tomography and pulmonary function tests in children and adults with CF

Computed tomography reflects lower airway inflammation and tracks changes in early CF

Computed tomography correlates with pulmonary exacerbations in children with CF

Bronchiectasis and pulmonary exacerbations in children and young adults with CF

High treatment burden in adults with CF: challenges to disease self-management

Over the past 25 years, computed tomography (CT) disease activity, such as spirometry (1), inflammatory cells on bronchoscopy (5), lung clearance index (LCI) (6), and pulmonary exacerbation frequency (2), and they are integral to diagnosing non-tuberculous mycobacterial (NTM) infections. CT scans have been shown to improve with treatment with DNase

Computed tomography correlates with pulmonary exacerbations in children with CF

Progression of lung disease on computed tomography and pulmonary function tests in children and adults with CF

Bronchiectasis in infants and preschool children diagnosed with CF after newborn screening

Computed tomography reflects lower airway inflammation and tracks changes in early CF

Lung Clearance Index and HRCT are complementary markers of lung abnormalities in young children with CF

Dornase alfa reduces air trapping in children with mild CF lung disease: a quantitative analysis

High resoluin CF patients

Chest computwith CF

Bronchiectasis and pulmonary exacerbations in children and young adults with CF

Chest computed tomography scores are predictive of survival in patients with CF awaiting lung transplantation

Chest computed tomography scans should be considered as a routine investigation in CF

Radiologic imaging in CF: cumulative effective dose and changing trends over 2 decades

Cancer risk among patients with CF and their first-degree relatives

Radiation exposure from CT scans in childhood and subsequent risk of leukaemia and brain tumours: a retrospective cohort study

18F-fluorodeoxyglucose-PET/CT imaging of lungs in patients with CF

High-resolution computed tomography (HRCT) should not be considered as a routine assessment method in CF lung disease

Association between mucoid Pseudomonas infection and bronchiectasis in children with CF

The sensitivity of lung disease surrogates in detecting chest CT abnormalities in children with CF

Computed tomography and magnetic resonance imaging in cystic fibrosis lung disease

CF Foundation evidence-based guidelines for management of infants with CF

The beta-glucan receptor dectin-1 recognizes specific morphologies of Aspergillus fumigatus

Stage-specific innate immune recognition of Aspergillus fumigatus and modulation by echinocandin drugs

CFTR regulates epithelial cell response to Aspergillus and resultant pulmonary inflammation

The effect of chronic infection with Aspergillus fumigatus on lung function and hospitalization in patients with CF

Exacerbation frequency and clini

Aspergillus fumigatus generates an

Treatment of Aspergillus fumigatus in patients with CF: a randomized, placebo-controlled pilot study

Posselt HG. Long-term fungal cultures from sputum of patients with CF

Longitudinal study of Aspergillus fumigatus strains isolated from CF patients

Genetic diversity among isolates of Aspergillus fumigatus in patients with CF

Clinical and serological survey of pulmonary Aspergillosis in patients with CF

Immune responses to Aspergillus in CF

Development of immune responses to Aspergillus at an early age in children with CF

Immune responses to Aspergillus fumigatus and P. aeruginosa antigens in CF & allergic bronchopulmonary aspergillosis

Antibody responses to Aspergillus fumigatus allergens in patients with CF

A 12-year longitudinal study of Aspergillus sensitivity in patients with CF

Sensitization to Aspergillus fumigatus and lung function in children with CF

Serologic IgE immune responses against Aspergillus fumigatus and Candida albicans in patients with CF

The effect of chronic infection with Aspergillus fumigatus on lung function and hospitalization in patients with CF

Exacerbation frequency and clinical outcomes in adult patients with CF

Allergic bronchopulmonary aspergillosis and Aspergillus infection in CF

Treatment of Aspergillus fumigatus in patients with CF: a randomized, placebo-controlled pilot study

CFTR gene mutations in adults with disseminated bronchiectasis

The CFTR gene and ion channel function in patients with idiopathic pancreatitis

Mutations in the CFTR gene and in vivo transepithelial potentials

CF: terminology and diagnostic algorithms

References: 1. Nick JA, et al. Effects of gender and age at diagnosis on disease progression in long-term survivors of CF

CFTR genotype as a predictor of prognosis in CF

Effect of genotype on phenotype and mortality in CF: a retrospective cohort study

Classic respiratory disease but atypical diagnostic testing distinguishes adult presentation of CF

CF Foundation Patient Registry

Manifestations of CF diagnosed in adulthood

Late diagnosis defines a unique population of long-term survivors of CF

Quality of life in young people with CF: effects of hospitalization, age and gender and differences in parent/child perceptions. Child Care Health Dev

Impact of recent pulmonary exacerbations on quality of life in patients with CF

Pulmonary exacerbations are associated with subsequent FEV1 decline in both adults and children with CF

Developing CF lung transplant referral criteria using predictors of 2-year mortality

Efficacy of aerosolized tobramycin in patients with CF

A controlled trial of long-term inhaled hypertonic saline in patients with CF

Comparative efficacy and safety of 4 randomized regimens to treat early P. aeruginosa infection in children with CF

The ATS/ERS Working Group on Infant and Young Children Pulmonary Function Testing Consensus Statement. Raised volume forced expirations in infants: guidelines for current practice

on behalf of the ATS/ERS Working Group on Infant andYoung Children Pulmonary Function Testing. An official ATS/ERS Society statement: pulmonary function testing in preschool children

Inhaled hypertonic saline in infants and children less than six years of age with CF: The ISIS randomized trial

Multicenter evaluation of infant lung function tests as CF clinical trial endpoints

Lung function in infants with CF diagnosed by newborn screening

Lung function from infancy to the preschool years after clinical diagnosis of CF

Spirometry in early childhood in CF patients

Airway function in infants newly diagnosed with CF

Association of lower airway inflammation with physiologic findings in young children with CF

Infection, inflammation, and lung function decline in infants with CF

Association of physiologic measures and clinical characteristics in a multicenter cohort of infants with CF

Lung function from infancy through preschool in a cohort of children with CF

High-resolution computed tomography imaging of airway disease in infants with CF

Lung function distinguishes preschool children with CF from healthy controls in a multicenter setting

References: 1. Gustafsson PM. Peripheral airway involvement in CF and asthma compared by inert gas washout

Lung clearance index at 4 years predicts subsequent lung function in children with CF

Hypertonic saline improves the LCI in paediatric CF patients with normal lung function

The effect of dornase alfa on ventilation inhomogeneity in patients with CF

Sone A, the G551D-CFTR mutation and mild lung disease: comparison of lung clearance index vesus spirometry

Inhaled hypertonic saline in infants and children younger than 6 years with CF: the ISIS randomized controlled trial

In placebo controlled trials, azithromycin has not been associated with treatment emergent methicillin resistant Staphylococcus aureus, small colony variant S. aureus, multi-resistant P. aeruginosa, Burkholderia species, Achromobacter xylosoxidans, Stenotrophomonas maltophilia, and/or non-tuberculous mycobacteria (NTM) [3, 5, 7] . While in vitro data has suggested that azithromycin may reduce effective autophagy (normal cellular process by which damaged organelles and unused proteins are sequestered and intracellular pathogens -including M. abscessus -are killed), the clinical relevance of this observation is unknown and requires additional study [8] . As expected, resistance to macrolide agents did occur; when compared with placebo participants, azithromycin participants had 27% more macrolide-resistant S. aureus and 7% more macrolideresistant H. influenzae [5] . However, numerous other agents are available for treatment of macrolide resistant bacteria in CF patients [9] . No macrolide resistant NTM were detected [3, 5] .Mechanisms of Action: While the mechanism of action of azithromycin in CF is not fully understood, increasing evidence suggests it is functioning as an antiinflammatory agent. In the clinical trials in which inflammatory markers were studies, azithromcyin was associated with reductions in C-reactive protein (CRP) [1] , neutrophil elastase [3] , and most recently with sustained reductions in absolute neutrophil counts, high sensitivity CRP, calprotectin, myeloperoxidase, and serum amyloid A [5, 10] . In this most recent study, reductions in calprotectin and CRP were significantly correlated with improved lung function and weight gain [10] .Thus, azithromycin currently represents the safest anti-inflammatory agent for CF. While associated with improvements in lung function, steroids were associated with decreased growth and cataracts in children with CF [11] . Similarly, while ibuprofen was associated with improvements in lung function, concerns regarding the risk of gastrointestinal bleeding and the complexities of measuring serum levels have substantially curtailed the use of this agent [12] .Evidence-based Recommendations for Use of Azithromycin in CF: Evidence-based guidelines developed by the U.S. CF Foundation have recommended that azithromcyin be used in CF patients 6 years of age and older chronically infected with P. aeruginosa to improve lung function and reduce exacerbations [13] . An updated recommendation that includes the use of azithromycin in CF patients uninfected with P. aeruginosa is pending. Similarly, a Cochrane review conducted by Southern et al. concluded that azithromycin improved respiratory function after 6 months and was associated with a reduction in exacerbations [14] . Thus, evidence-based recommendations consistently recommend the chronic use of azithromycin for patients with CF due to clinical benefits and a favorable safety and tolerability profile.Study Radiologist Training on Scoring System: Two radiologists independently scored 10 Training CT Scans on a lobe specific basis. Each lobe, with the lingula scored as a separate lobe, was evaluated for the extent and severity of bronchiectasis, mucous plugging, peribronchical thickening, parenchymal opacity, ground glass density, cysts or bullae, and air trapping. The scoring system was modified to include a visual analog scale with continuous variables ranging from 0.0 up to 10.0. Following the initial scoring the radiologists discussed all discrepancies and developed common criteria for areas of frequent discrepancies. Inter-rater reliability was assessed on the Training CT Scans using Pearson Correlation Coefficients and these data were used to assess areas in need of more calibration between the study radiologists. Scoring was then repeated by the two study radiologists several months later to again evaluate inter-rater reliability and further the training process of refining radiologist calibration. This completed the training phase of the project.

We wanted to take the information from the training process and use the Training CT Scans as a part of the Research Study Scoring protocol. To meet this goal, two sets of panels using a subset of the Training CT Scans were developed. This development process included a review by one radiologist and the study coordinator of the end results from the two training meetings. Specifically, individual scores of lung abnormalities that represented varying levels of agreement were identified. This resulted in two CT panels that included a set of scores that demonstrated a range of lung disease status and instances of scores that ranged from perfect agreement to large differences in ratings. At the end of this process, the research coordinator summarized individual lung disease abnormalities for all the scans within each panel and asked the two radiologists to consider all the information that came from the training process to then establish a consensus score for each abnormality. The consensus score would be used as the "anchor" rating and a panel would be reviewed at the beginning of each research study scan scoring session. In this way, the culmination of the training process resulted in a way to help the radiologists systematically calibrate with each other on a set of "true anchor" scores as they independently looked at blinded study scans.Protocol for Each Scoring Session: At each consecutive CT scoring session, a CT Scoring Panel was reviewed immediately prior to scoring the blinded, ran-domized research study scans. The radiologists reviewed each abnormality and came to an agreement on the score and communicated the consensus score to the research coordinator. The research coordinator then communicated consensus score back to the radiologists so re-calibration and consensus could be established. The average length of time to review one scoring panel was approximately 50 minutes. Following CT Scoring Panel review, blinded research study scans were independently scored by the two study radiologists.Scoring of Research CT Scans: A total of 11 scoring sessions took place over 27 months. On average, 8 research study scans were reviewed each session. The protocol that used the calibration process was well received by the study radiologists and allowed for a systematic, routine approach to each blinded research CT scan scoring session.Analysis of Research CT Scans: Now that the CT sub-study is complete, analysis has begun on the research data from 42 baseline CT scans and 40 follow-up CT scans. To date, data from one expert coder has been reviewed. Descriptive data from one coder includes the mean presence of bronchiectasis and air trapping, and average lobes with presence of bronchiectasis and air trapping at the baseline and 18 month (follow-up) time points for all participants.Research Results: Mean and SD age of participants at baseline CT scan was 3.67 (1.3) years and at follow-up (approximately 18 months later) was 5.33 (1.28) years. Percentage with presence of bronchiectasis at baseline (N = 42) was 52% and follow-up (N = 40) was 55%. Mean and SD for lobes with bronchiectasis at baseline (N = 40) was 1.25 (1.75) and follow-up (N = 38) was 1.10 (1.43). Percentage with presence of air trapping at baseline (N = 40) was 83% and follow-up (N = 38) was 82%. Mean and SD for lobes with air trapping at baseline (N = 40) was 2.72 (2.19) and follow-up (N = 37) was 2.97 (2.22) .Discussion: Next steps for analyzing the CT data include determining inter-rater reliability on the blinded, randomized research CT scans. Additionally, change scores of the important aspects of lung disease status in young children with CF will be analyzed. Examples of change scores from baseline to follow-up (18 months later) within subjects will be presented at the workshop and compared to existing data from other studies in the same age. Regional differences in lung disease will also be examined and presented.