key: cord-0010179-nco1zza8
authors: nan
title: Symposium Summaries
date: 2011-09-28
journal: Pediatr Pulmonol
DOI: 10.1002/ppul.21582
sha: 517529f6d4397d6328ffcb6a65c13a8b571e6b9f
doc_id: 10179
cord_uid: nco1zza8

nan

Mitchell L. Drumm, Ph.D., Tracey L. Bonfield, Ph.D. and Craig A. Hodges, Ph.D. Depts. of Pediatrics & Genetics, Case Western Reserve Univ., Cleveland, OH, USA CF is unquestionably a systemic disorder, affecting every organ system of the body to some degree. What is not clear is how loss of CFTR gives rise to the myriad of physiologic effects characteristic of CF. For the past several decades, most of the focus has been on epithelial cells and epithelia, as they are clearly affected by the loss of CFTR and provide an explanation for elevated sweat electrolytes, airway mucous rheology and some gastrointestinal manifestations. However, CFTR mRNA is expressed in most cell types of the body and channel activity has been reported in many as well (lymphocytes, muscle, neurons, for example), but its role in most non-epithelial cells is not clear. Even less clear is the effect of not having CFTR in these non-epithelial cell types.

We have begun to address this concept in an animal model, the mouse, in which the murine Cftr gene has been modified with loxP sites to allow conditional inactivation of Cftr (1) . This allele has been used to examine the effect of Cftr absence from lymphocytes and showed that mice lacking Cftr in CD3+ lymphocytes have an exuberant Th-2-biased response to Aspergillus, resembling the response of CF patients and indicating that this is a function of the CF immune system rather than the CF epithelium (2) . More recently, we have inactivated Cftr in myeloid cells and challenged these mice with an inoculum of P. aeruginosa and compared the effects to mice that had received bone marrow transplants. All mice mounted an inflammatory response, but whereas control mice resolved the response within a few days, CF mice, non-CF mice with myeloid Cftr inactivation and non-CF mice receiving CF bone marrow all experienced prolonged inflammatory indices (~2 weeks). These results also point to dysregulated inflammatory responses by the CF immune system.

Conditional inactivation of Cftr provides insight into disease pathogenesis, but another important concept is whether correcting for the loss of Cftr can alter CF pathophysiology. To begin to address this concept, a second conditional allele was generated, but in this case a nonfunctional form of Cftr was generated with loxP sites but constructed so that it could be conditionally activated. Both alleles were used to examine the effects of manipulating Cftr function in gastrointestinal epithelium. Inactivating Cftr in gut epithelium resulted in intestinal obstruction, but did not impair growth. Restricting Cftr function to the gut epithelium completely prevented obstruction, but these mice were indistinguishable from CF mice in terms of growth characteristics, length and weight (3) .

As the organ systems affected by CF do not function in isolation, it is important to know how loss of CFTR in one cell type, tissue or organ affects the function of other cells, tissues or organs, and these novel mouse models provide insight into these interactions. 

David K. Meyerholz, D.V.M., Ph.D. Pathology, Univ. of Iowa, Iowa City, IA, USA Cystic fibrosis (CF) is caused by mutations in the gene encoding the CF transmembrane conductance regulator (CFTR). CF is characterized by clinical disease in many organs including the intestines (meconium ileus), pancreas, liver/gallbladder, skin, nasosinus, vas deferens and lung. However, the understanding of how CFTR loss causes disease phenotypes in these diverse organs is limited. Because lung disease is the principal cause of mor-tality in CF, it is an important area of investigational study. CF mouse models were developed nearly twenty years ago and from these, much has been learned regarding CFTR and CF disease. Recently, the advent of pig and ferret models has provided new perspectives for understanding CF disease and pathogenesis. In pigs, both CFTR-Null and CFTR-∆F508 models were born with meconium ileus, pancreatic destruction, parameters of focal biliary cirrhosis, and segmental atresia of the vas deferens; however, at birth, the lung lacked inflammation or airway obstruction typical of advanced CF lung disease (1) (2) (3) . To determine if CF pigs would develop lung disease, we surgically corrected the meconium ileus of neonatal pigs and observed their postnatal growth for clinical evidence of lung disease. In the weeks and months after surgery, serial CT scans, bronchoalveolar lavages, cultures, and pathologic examinations revealed that CF pigs developed lung disease characterized by a heterogenous, intermittent to progressive process causing mucopurulent airway obstruction with air trapping, atelectasis and infection by a variety of bacterial pathogens (3) (4) . To better understand the mechanisms responsible for disease, we examined electrolyte transport in newborn CF airways; we found a loss of chloride and bicarbonate transport with no increase in sodium absorption (5) . We challenged newborn CF pigs with S. aureus (clinical isolate from a CF pig) and found that within hours of inoculation, CF pigs had impaired eradication of bacteria compared to their wild-type littermates. These data suggest that lack of CFTR caused a host defense defect in eradicating bacteria before the onset of airway inflammation (4) . We also morphometrically examined trachea and large airways to see if loss of CFTR caused microanatomic changes. CF trachea and primary bronchi had reduced caliber and circularity with hypoplastic submucosal glands, accentuated smooth muscle bundles and discontinuous cartilage along the anterior surface. We found no change in MUC5AC positive goblet cells, cellular height or ciliary height of surface epithelium. To see if the tracheal phenotype translated to humans, we retrospectively examined pediatric CT scans and published trachea data. We found that infants with CF had detectably reduced tracheal circularity and caliber as predicted by the CF pig model (6) . We studied smooth muscle as a possible contributor to this phenotype because of its altered morphology at birth and because altered pathophysiology of airway smooth muscle might explain some of the observed anatomic changes including reduced caliber, reduced circulatory and discontinuous cartilage. Ongoing studies suggest a potential role of CFTR in airway smooth muscle function. We are also studying neonatal and fetal CF pig lung using CT imaging, morphometric analysis and gene expression to assess genotypic differences in airway and parenchymal structure and development. In summary, CF pigs spontaneously develop lung disease like that in humans. This model has helped us address questions regarding CF pathophysiology, and it is helping us identify novel features of the CF phenotype in humans, expanding our clinical understanding of the disease. 

John F. Engelhardt, Ph.D. Dept. of Anatomy & Cell Biology, Univ. of Iowa, Iowa City, IA, USA Animal modeling of cystic fibrosis has been challenging, with species-and strain-specific differences in organ biology and CFTR function influencing the emergence of disease pathology. We have developed a CFTR-deficient ferret model (1) that develops many of the pathologies observed in humans with CF, including defective airway chloride transport and submucosal gland fluid secretion, variably penetrant meconium ileus, pancreatic, liver and vas deferens disease, and most importantly lung infections during the neonatal period (2) . Newborn CFTR-knockout ferrets fail to thrive and succumb to lung infection if untreated shortly after birth. This phe-notype can be partially overcome by antibiotic therapy, supplementation with elemental diet, and drug therapy to raise gastrointestinal pH and normalize liver function tests. Using these rearing methods, CF ferrets can be reared to adulthood. In contrast to neonates, adult CF ferrets contract a more slowly progressive lung disease characterized by excessive mucus production, plugging of the airways, and lethal bacterial infections.

To determine whether early lung infections during the first weeks of life observed in CFTR-knockout ferrets were secondary to impaired nutritional status, we performed controlled bacterial challenge experiments in KO It is embarrassing that after nearly three decades of defining electrolyte disorders in cystic fibrosis (CF), we do not yet have a unifying cause-and-effect mechanism to explain why a basic defect in anion transport causes angry mucus in CF. Understanding what make mucus hostile enough to kill should be critical to design strategies for effective therapy.

To date the most widely held concept for explaining the abnormally "thick" CF mucus holds that Na + dependent fluid absorption via ENaC is exuberant and excessive in CF, causing "dehydrated" mucus. Since thick mucus, or mucoviscidosis, underlies the pathology of all anatomically distressed tissues in CF, the same basis or phe-nomenon that causes abnormal mucus should occur in each affected organ. Yet, four of the major organs that are affected by thickened mucus in CF cannot "dehydrate" mucus via hyperactive ENaC because they lack this cation channel for absorbing Na + and fluid. Consequently, another mechanism common to mucus formation and dependent on CFTR should be present in those organs. An inspection of the organs affected in CF suggests that depressed HCO 3 transport is a common defect in these systems.

This fact raises the question of whether altering HCO 3 transport alters mucus. Briefly, we and others have recently observed that disruption of HCO 3 secre-and WT newborns within 6-12 hrs after birth. At this time, total weight and body mass index were similar between the two genotypes. Controlled injection of antibiotic resistant PAO1 into the trachea of newborn KO and WT animals demonstrated two orders of magnitude difference in bacterial killing over a 6 hr period. Interestingly, newborn (0-3 day old) WT and KO animals demonstrated no significant differences in mucociliary clearance, suggesting that defects in bacterial eradication were not due to impaired clearance during the newborn interval. However, mucociliary clearance rapidly became impaired within CF animals by 1-2 weeks of life. With advances in medical treatments for CF lung disease, the incidence of CF-related diabetes (CFRD) is climbing-now~50% of CF patients develop CFRD by the age of 30 years (3, 4) . Although CFRD is only infrequently seen in infants,~20% of adolescents are affected (5) . CFRD is now the main secondary complication in CF, and responsible for a 6-fold increase in morbidity and mortality (3) . The pleiotropic nature of CFRD, which does not conform to either type 1 or type 2 diabetes, has made dissection of its etiology and impact on CF lung disease difficult to understand. Declining clinical status and lung function often occurs years before CFRD is diagnosed, and it is uncertain if the pre-diabetic state influences the onset of more severe lung disease or if more severe lung disease influences emergence of CFRD.

We have begun to utilize the CFTR-knockout ferret model to address the pathophysiology of CFRD. Although histopathology is mild in the CF ferret pancreas at birth, pancreatic disease in CFTR-knockout ferrets is rapidly progressive and leads to destruction of both the exocrine and endocrine pancreas during the first month of life. Such pathology is accompanied by hyper-glycaemia, glycosuria, and fibrosis of the pancreas. To evaluate whether defects in the endocrine pancreas occur prior to the emergence of pancreatic fibrosis and destruction of the exocrine pancreas, we evaluated endocrine pancreas structure and function in newborn CFTRknockout ferrets. Newborn CFTR-knockout ferrets demonstrated no significant changes in the percentage of parenchymal insulin and glucagon positivity in the pancreas. Despite these findings, subtle abnormalities in glycemic regulation were observed. For example, newborn CF ferrets demonstrate abnormal glucose tolerance tests despite the minimal histopathology in the pancreas at birth. CF kits also demonstrate impaired acute insulin secretion following L-arginine or glucose challenge and an exaggerated second phase insulin response to glucose. These findings have begun to implicate CFTR functions in the endocrine pancreas that may contribute early in life to the metabolic and lung pathology in cystic fibrosis. tion concurrent with mucus secretion causes several alterations that appear consistent with mucus abnormalities presumed to occur in CF, namely: 1) mucus discharge is impeded; 2) mucus swelling is reduced; 3) mucus diffusivity is decreased (viscosity is increased); 4) mucus adhesiveness appears increased; and 5) stimulated mucocilliary transport is diminished.

In this presentation, we will try to: 1) compare some of the more outstanding disturbances in mucus secretion and clearance in humans and animal models; 2) defend the notion that poor HCO 3 transport is common to CF affected tissues; and 3) explain at least to a limited extent in terms of electrostatics why HCO 3 is critical for normal mucin release and formation.

We will consider the possible importance of these observations to adapt strategies for therapy with HCO 3 -. We will discuss these findings and their potential relevance to approaches by which mucocilliary clearance might be improved in airways of human animals -no joke.

Acknowledgments: The authors' work is supported by MCCC-Cystic Fibrosis Foundation; NIH-RO1 HL084042; CFRI, and Nancy Olmsted Endowment. Special thanks to Mr. S. Clemens for assistance with the title.

Vin Tangpricha, M.D., Ph.D.

Program, Emory Laney Graduate School, Atlanta, GA, USA Vitamin D is an important secosteroid hormone for the maintenance of normal calcium homeostasis for optimal skeletal mineralization. Vitamin D can be produced in the skin or obtained from the diet from a limited number of foods. After entering the circulation, vitamin D is converted to 25-hydroxyvitamin D (25(OH)D) by enzymes in the liver and then to 1,25-dihydroxyvitamin D (1,25(OH)2D), the more active form of vitamin D, by enzymes in the kidney (1) . Other tissues in the body have the ability to convert 25(OH)D to 1,25(OH)2D, which may act as an autocrine and/or paracrine hormone to regulate other extra-skeletal processes (2) .

Patients with cystic fibrosis (CF) have been found to have sub-optimal vitamin D status by several recent published reports (3, 4) . Potential causes of vitamin D deficiency in CF include fat malabsorption due to pancreatic insufficiency (5) , inadequate sunlight exposure (6) , suboptimal supplementation and/or intake of vitamin D containing foods, diminished vitamin D stores due to decreased adiposity, and decreased circulating half-life of 25(OH)D due to alterations in vitamin D binding protein (4) .

Randomized controlled trials have demonstrated efficacy of vitamin D therapy in the prevention of osteoporotic fractures in non-CF patients when adequate levels of serum 25(OH)D are obtained and when adherence to vitamin D is high (1) . Recent vitamin D guidelines from the Institute of Medicine have focused on optimizing bone health for the general population (7) . More recent guidelines developed in 2011 co-sponsored by the Endocrine Society, Canadian Society of Endocrinology and Metabolism and the National Osteoporosis Foundation call for screening of vitamin D deficiency in patients with CF and recommend much higher levels of serum 25(OH)D and higher doses of vitamin D in those at risk for vitamin D deficiency, including patients with CF (1) .

Although direct randomized controlled trial evidence in the CF population is currently lacking, vitamin D deficiency is a likely major contributor to CF bone disease. More focus should be placed on prevention of vitamin D deficiency in the CF population given that it is well established that vitamin D deficiency is nearly universal in CF patients. New guidelines to be released by the CF Foundation will address screening, prevention and treatment of vitamin D deficiency in the CF population. Vitamin D deficiency is a concern for the general population, but particularly for individuals with cystic fibrosis, the majority of whom experience maldigestion with resultant malabsorption. Knowing how to treat and prevent vitamin D deficiency is of utmost importance for optimal nutritional care. For this reason, the Cystic Fibrosis Foundation formed an evidence based guidelines committee, to develop practice guidelines regarding the assessment and management of vitamin D levels in persons with cystic fibrosis (1) . Using questions developed by this committee, a Cochrane systematic review was performed. Figure 1 details a summary of the search and review process.

For the assessment of vitamin D status and treatment goals, the CF Foundation recommends that all individuals with CF have serum 25-hydroxyvitamin D measured annually, preferably at the end of winter, with a goal to maintain serum 25-hydroxyvitamin D at ≥ 30 ng/mL, but not to exceed 100 ng/mL. For individuals with CF with a serum 25-hydroxyvitamin D level < 30 ng/mL, there should be assessment for adherence to the prescribed regimen. The CF Foundation recommends against the use of serum 1,25-dihydroxyvitamin D as the measurement to assess vitamin D status and against the routine measurement of parathyroid hormone, osteocalcin, alkaline phosphatase, or other indirect markers to assess vitamin D status in all individuals with CF. The CF Foundation is not able to make a recommendation for or against having all individuals with CF fast prior to the measurement of serum 25-hydroxyvitamin D. All individuals with CF should have serum 25-hydroxyvitamin D levels rechecked after 3 months any time the dose of vitamin D3 has been changed. All of the above recommendations for the assessment of vitamin D status and treatment goals are based on consensus recommendations and have a low certainty of net benefit using the USPSTF grading system.

Regarding the use of vitamin D compounds and treatment strategy, the CF Foundation recommends all individuals with CF be treated with vitamin D3 (cholecalciferol) to achieve and maintain serum 25-hydroxyvitamin D levels ≥ 30 ng/mL. This recommendation is made with moderate certainty and moderate benefit using the USP-STF grading system; all other recommendations in this section have a low certainty and unknown benefit. The CF Foundation is not able to make a recommendation for or against the use of an oil versus a powder-based formulation of vitamin D3 in all individuals with CF. The CF Foundation recommends all individuals with CF who are prescribed vitamin D3 (in addition to their CF-specific vitamins) take once-daily vitamin D3 therapy or its weekly dose equivalent to maintain serum 25-hydroxyvitamin D levels ≥ 30 ng/mL, and those individuals with refractory vitamin D deficiency be treated with calcitriol, doxercalciferol or paricalcitol only in consultation with a specialist with expertise in vitamin D therapy.

For the treatment of infants with CF, birth to 12 months, the CF Foundation recommends an initial dose of 400-500 IU vitamin D3 per day. This dose is provided in 1 ml of CF specific multivitamins. For infants with a serum 25-hydroxyvitamin D level < 10 ng/mL, it is recommended that assessment for rickets and urgent management and treatment in consultation with a specialist with expertise in vitamin D therapy be completed. For infants with a serum 25-hydroxyvitamin D level ≥ 20 but < 30 ng/mL and with confirmed adherence to the prescribed regimen, the dose of vitamin D3 may be increased to 800-1000 IU per day. For infants with a serum 25-hydroxyvitamin D level < 20 ng/mL or with a persistent serum 25-hydroxyvitamin D level ≥ 20 but < 30 ng/mL and with confirmed adherence, the dose of vitamin D3 may be increased to a maximum of 2000 IU per day. The CF Foundation recommends all individuals with CF, birth-12 months, who are unable to achieve a serum 25-hydroxyvitamin D level ≥ 30 ng/mL after treatment with 2000 IU vitamin D3 per day and with confirmed adherence, be managed in consultation with a specialist with expertise in vitamin D therapy.

For treatment of children 1-10 years of age, the CF Foundation recommends treatment with an initial dose of 800-1000 IU vitamin D3 per day. For those with a serum 25-hydroxyvitamin D level ≥ 20 but < 30 ng/mL and with confirmed adherence to the prescribed regimen, the dose of vitamin D3 may be increased to 1600-3000 IU per day. If serum 25-hydroxyvitamin D level < 20 ng/mL or there is a persistent serum 25-hydroxyvitamin D level ≥ 20 but < 30 ng/mL and confirmed adherence, the dose of vitamin D3 may be increased to a maximum of 4000 IU per day. For those who are unable to achieve a serum 25-hydroxyvitamin D level ≥ 30 ng/mL after treatment with 4000 IU vitamin D3 per day and with confirmed adherence, management should be done in consultation with a specialist with expertise in vitamin D therapy.

For the treatment of individuals with CF who are 10 years of age and older, the CF Foundation recommends an initial dose of 800-2000 IU vitamin D3 per day. For those with a serum 25-hydroxyvitamin D level ≥ 20 but < 30 ng/mL and with confirmed adherence to the pre-scribed regimen, the dose of vitamin D3 may be increased to 1600-6000 IU per day. For those with a serum 25-hydroxyvitamin D level < 20 ng/mL or with a persistent serum 25-hydroxyvitamin D level ≥ 20 but < 30 ng/mL and with confirmed adherence, the dose of vitamin D3 may be increased to a maximum of 10,000 IU per day. The CF Foundation recommends all individuals with CF, age >10 years, who are unable to achieve a serum 25-hydroxyvitamin D level ≥ 30 ng/mL after treatment with 10,000 IU vitamin D3 per day and with confirmed adherence, be managed in consultation with a specialist with expertise in vitamin D therapy. All of the above recommendations for the assessment of vitamin D status and treatment goals from birth through adulthood are based on consensus recommendations and have a low certainty of net benefit using the USPSTF grading system.

Regarding the use of UV lamps, the CF Foundation is not able to recommend for or against the use of ultraviolet lamps in the management of vitamin D deficiency in all individuals with CF. This recommendation has a low certainty and an unknown benefit, using the USPSTF grading system. These consensus and evidence based recommendations for assessment and management of vitamin D will provide guidelines and standardized care in the treatment of individuals with CF. This evidence based review provides the framework for continued future research regarding best clinical practice in vitamin D management for all individuals with CF.

References: 1. Tangpricha 

Robert Aris, M.D.

Probably. The complexity of this question should not be over-estimated because vitamin D has been shown to be an important immunoregulatory molecule vis a vis transplant rejection in addition to being involved in calcium, bone biology, and fracture prevention. Furthermore, extra-skeletal sequelae of vitamin D deficiency include insulin resistance, hypertension, and malignancy (1) , all of which are major comorbidities after lung and solid organ transplantation. The following remarks will focus on the role of vitamin D and both host response to pathogen and transplant immunology. While beyond the scope of this short summary, a myriad of papers largely from the bench and pre-clinical studies, suggest that vitamin D, mainly in the form of 1,25 OH 2 D, augments CD4+, CD25 high regulatory T cells, which are able to immunomodulate allogeneic T cells and thus protect the lung graft from rejection (2) . In addition, vitamin D regulates antigen presentation and contributes to the production of tolerogenic dendritic cells (3) . Recently, two clinical studies in renal transplant recipients showed that vitamin D reduced HLA DR and costimulatory molecule expression in peripheral blood leukocytes, and expanded regulatory T cell populations, results which might be anticipated to have a salubrious effect on transplant rejection (4, 5) . Taken together, these results suggest that vitamin D metabolites might protect the lung allograft by reducing overall alloreactivity (rejection). In addition through antimicrobial peptide (cathelicidin) networks and toll like receptors, vitamin D plays a crucial role in pathogen defense against bacterial and mycobacterial infections (6) . In this way, vitamin D "primes" the innate immune response through macrophage and epithelial cell activation (7) . Thus, in a post-transplant environment, vitamin D may have broad reaching positive effects on the health of the patient and the graft.

But what happens to serum 25OHD levels after lung transplantation? First, virtually every disease that results in end-stage lung disease and the need for lung transplantation, including CF, is associated with serious vitamin D deficiency (8) . The same is true for other solid organ transplantation. Thus low 25OHD levels are the norm during advanced illness. The reported prevalence of vitamin D insufficiency after lung transplant ranges from 51% to 97% and of severe deficiency from 26% to 33% (8, 9) . The etiology of this problem is due the chronic disease, inadequate vitamin D intake and low sun exposure due to immobility. In addition, because of an increased risk of skin cancer (7.4% incidence by year 3 and up to 80% after 20 yrs, largely linked to azathioprine use) in organ recipients, many patients dramatically limit their sun exposure (10) . Fortunately newer DNA synthesis inhibitors like mycophenolate (cellcept®) may cause less risk.

Prospective studies in lung, heart, and liver transplant recipients indicate that serum 25OHD levels tend to gradually increase from the low levels seen before transplantation to normal levels (9, 11) . While some of this change is probably related to the use of vitamin D supplements, it is certainly not all related because, as it pertains to CF individuals, they are already supplemented carefully before lung transplant. Even more importantly a number of transplant medications alter the calciumvitamin D-PTH axis. Thus, PTH, a usually reliable measure of the health of the calcium-vitamin D axis, is a poor biomarker of vitamin D deficiency after transplant because calcineurin immunosuppressants (cyclosporine and tacrolimus) reduce renal function and raise PTH levels over time. In addition calcineurin immunosuppressants raise 1,25OH 2 D levels and lower Vitamin D binding protein production, the latter of which would be expected to lower serum 25OHD levels over time (12) . Furthermore, corticosteroids increase 25OHD catabolism, an effect which may explain, in part, why 25OHD levels go up over time after transplant since corticosteroids are slowed weaned down to a nadir near the 6 month anniversary (13) . Lastly macrolides and azoles, used commonly to slow the advent of chronic rejection or treat fungal infections, respectively, inhibit 25OHD catabolism via the CYP3A4 enzyme and thus would contribute to higher serum 25OHD levels. Therefore vitamin D metabolism after transplant is very complicated. One 2-yr RCT of the efficacy of pamidronate in treating postlung transplant osteoporosis in CF patients treated with routine immunosuppressants, but not macrolides, puts things into perspective. This study demonstrated that serum 25OHD, 1,25OH 2 D and PTH levels went up 50% on average despite no changes in vitamin D supplementation in the patients. Since half the patients were receiving placebo, the effects were not due to bisphosphonate treatment itself (14) . Interestingly other fat soluble vitamin (A and E) levels go up in patients after lung transplantation, despite no difference in supplementation, attesting to the likelihood that drug-drug or drug-vitamin interactions may be taking place at the level of CYP3A4, the complexities of which have not been fully unraveled (15) . Sometimes these levels go so high that cessation of supplements is a consideration.

What are the consequences of low vitamin D status after transplantation? Low vitamin D status (25OHD or 1,25OH 2 D) has been linked to low femoral neck Z-scores (liver transplant), indirectly to vertebral fractures (heart transplant) and death (heart transplant) (6) , while few studies linking adverse outcomes in lung transplant have been undertaken. Of course, vitamin D deficiency after organ transplant must be taken in the context of the fact that both steroids and calcineurin inhibitors are direct causes of osteoporosis and fracture, as is diabetes, a common complication after transplantation. Therefore it has been historically difficult to link a particular causal factor (eg, vitamin D insufficiency) directly to an adverse outcome because of the myriad of potentially influential and overlapping variables.

What are the best interventions for low vitamin D after lung or liver transplantation? Vitamin D supplementation in various forms usually has a salubrious effect on bone endpoints, measured most commonly by bone densitometry as fracture endpoint trials are usually underpowered after transplant. Most vitamin D intervention studies have shown a slowing in bone density decline after transplant. Bisphosphonates, in this regard, are more potent drugs and are considerably better at improving bone density than vitamin D supplements alone. There is not a one-size-fits-all supplementation protocol, so supplement doses should be guided by serum 25OHD levels, similar to what is done before transplant. And realizing for all the fat soluble vitamins, some patient may require large, small or no supplements. It should also be noted that as GFR declines in lung transplant recipients over time (as a consequence of immunosuppressants), a transition from ergo-or chole-calciferol to calcitriol (Rocal-trol®) may be necessary because the kidney will no longer be able to 1-hydroxylate the precursor molecule to the most biologically active metabolite, 1,25OH 2 D.

References Classical function of the active form 1,25-dihydroxyvitamin D is maintaining calcium homeostasis by stimulating calcium transport from intestine and kidney into the blood, and by having direct negative feedback effect on parathormone release. However, several recent observations have initiated a re-evaluation of the physiologic and pharmacologic actions of vitamin D. The nuclear vitamin D receptor has been found to be expressed in most tissues in the body, and CYP27B1, the enzyme required to convert circulating 25(OH)D to 1,25-dihydroxyvitamin D, has been discovered in several cell types, including immune cells and epithelial cells (1, 2) . This has been followed by subsequent description of "non-classical functions of vitamin D," which comprise anti-proliferative, differentiation-regulating, anti-diabetic, and immunomodulatory properties of vitamin D, and are so far firmly supported by numerous in vitro and in vivo animal experiments (3) .

The in vitro findings of immunomodulatory properties of vitamin D consistently suggest that vitamin D stimulates the innate immune system and enhances antimicrobial activity at early stages of infection (4, 5) , but provides negative feedback mechanism at later stages of infection (6) . The inhibitory effects of vitamin D on adaptive immune response include induction of the potent antiinflammatory cytokine IL-10. This complex immunomodulatory role of vitamin D has been reviewed in detail (1) . So far there are no large human studies on the proposed role of vitamin D in the immune system during infection.

The vast majority of children and adults with CF have serum levels of 25-hydroxyvitamin D (25(OH)D) below 30 ng/mL (75 nmol/L), the upper cut-off value for vitamin D insufficiency, despite following the currently recommended vitamin D supplementation regimens (7, 8) .

We hypothesized that chronic vitamin D insufficiency may contribute to the ongoing CF lung inflammation, which is associated with high IgG levels, low IL-10 levels, and poor lung function and studied the relation between vitamin D and serum IgG, and, between vitamin D and lung function (9) . The study was done in a large, well-characterized Scandinavian CF population. Serum 25(OH)D and total IgG were measured, spirometry car-ried out and vitamin D intake data gathered using a seven-day dietary food record. Multiple linear regression analyses were performed for IgG and FEV1 as dependent variables, and serum 25(OH)D, daily food and supplemented vitamin D sources of intake as independent variables. The model was controlled for age, gender, genotype, CF-related diabetes, infection/colonization status, long-term oral corticosteroid treatment, long-term treatment with macrolides, pancreatic insufficient vs. sufficient phenotype and Body Mass Index z-score. Serum total IgG levels were negatively associated with serum 25(OH)D (R 2 =0.386; beta=-5.3; p<0.001), supplemented vitamin D intake per kg bodyweight (R 2 =0.388; beta=-2.1; p<0.001), and total vitamin D intake per kg bodyweight (R 2 =0.415; beta=-1.5; p=0.002). Serum 25(OH)D was positively associated with FEV1 (R 2 =0.318; beta=0.16; p=0.028). The study supports the proposed role of vitamin D in the immune system during infection and substantiates prospective studies.

The Third National Health and Nutrition Examination Survey (14 000 noninstitutionalized adults) revealed a significant association between 25(OH)D levels and pulmonary function (10) . The effect that 25(OH)D levels in patients with chronic lung disease may have on pulmonary function has not been extensively studied. However, vitamin D has been linked in some studies with atopy-and asthma-associated phenotypes in children with established disease. Recently it has been shown that children with inadequate vitamin D levels are at increased risk of developing atopy, and subsequently bronchial hyperresponsiveness and asthma (11) . Low vitamin D levels at age 6 was a predictor of atopy and asthma at 14.

Vitamin D may treat or prevent allergy to Aspergillus in patients with CF, according to the results of a recent in vitro study (12) . Using peripheral CD4+ T cells isolated from patients with ABPA, the investigators found that the costimulatory molecule OX40 ligand (OX40L) was critical in driving Th2 allergic responses to A. fumigatus, whereas CD4+ T cells from patients without ABPA did not mount enhanced Th2 responses and contained more regulatory T cells expressing transforming growth factor beta (TGF-β). In patients with ABPA, increased Th2 reactivity was correlated with lower mean serum levels of vitamin D. Adding 1,25 OH-vitamin D3 to cells from patients with ABPA markedly decreased expression of OX40L, increased expression of TGF-β needed for development of allergen tolerance, increased Treg TGFβ expression, and decreased Th2 responses. Also, 1,25(OH)2 D3 has been shown to induce increases in antimicrobial proteins and secretion of antimicrobial activity against pathogens including Pseudomonas aeruginosa (13) . 1,25(OH)2 D3 thus directly regulates antimicrobial peptide gene expression, revealing the potential of its analogues in treatment of opportunistic infections.

Vitamin D supplementation has been associated with a decreased incidence of type 1 diabetes (14) and plasma 25(OH)D has been described to be associated with type 2 diabetes (15) . The exact molecular pathways are not completely understood but vitamin D has been suggested to act via its immunomodulatory effects in type 1 diabetes (16) (17) and to increase insulin secretion and peripheral insulin sensitivity in type 2 diabetes (18) (19) . We assessed the relationship between vitamin D and CFrelated diabetes, glucose tolerance and HbA1c in 898 Scandinavian CF patients. Vitamin D insufficiency degree (OR = 1.36; p = 0.032) and serum 25(OH)D < 30 nmol/L (OR = 1.79; p = 0.042) were significant risk factors for CF-related diabetes. Accordingly, log e -HbA1c value was positively associated with serum 25(OH)D < 30 nmol/l and < 50 nmol/l, as well as with vitamin D insufficiency degree (adj.R 2 = 20.5% and p < 0.05 in all). In subgroup analyses, serum 25(OH)D < 30 nmol/l determined the log e -HbA1c value in paediatric patients (adj.R 2 = 20.2%; p = 0.017), but not in adults. Improving vitamin D status may have some anti-diabetic effect in CF, especially in children. Prospective studies on the proposed role of vitamin D insufficiency in the pathophysiology of diabetes mellitus are needed Evidence-based medicine (EBM) has been championed as a "new paradigm" for medical education and practice (1) . EBM focuses on medical knowledge derived from controlled clinical trials, asserting that such knowledge represents the "best evidence" that can be used to support a clinical decision. The value of clinical research for clinical medicine is limited, however, by the relative lack of research, the relevance of the research to clinical practice and by the difficulty of applying populationbased knowledge to the treatment of individuals. Other forms of medical knowledge, including those derived from pathophysiologic understanding and clinical experience, might be helpful, but have been devalued by EBM (2) . Though acknowledging the need to integrate various kinds of medical and non-medical knowledge into clinical decision making, proponents of EBM have said very little about how such integration should actually take place. The process by which clinicians are to weigh and balance what may be conflicting information coming from not only published clinical research but also from personal experience, pathophysiologic understanding of disease, the preferences of individual patients or other "non-evidentiary" sources has not been fully elucidated.

Five topic areas are relevant to any clinical decision (Table) (3). Recognizing that each form of medical knowledge (empirical, experiential and pathophysiologic) has significant strengths and limitations when it comes to utilizing that knowledge in the care of individual patients, hierarchies of knowledge advanced by EBM cannot be helpful in deciding on individual treatment decisions. Rather, the clinician must weigh all relevant information in attempting to come to the best decision for an individual patient. The process of clinical decision making must begin with the patient at hand, with the clinician then considering all empirical, experiential and pathophysiologic facts and reasons that bear upon the case before reaching a conclusion regarding the most medically appropriate course of action. Patient preferences and system features need to be factored in at this point in order to finalize a treatment choice or recommendation.

The care of persons with cystic fibrosis has greatly benefited from high-quality clinical research performed over the last several decades. CF care, however, requires individualization. Many treatments in CF (e.g., pancreatic enzyme replacement) are not supported by high-quality clinical research yet are clearly defensible as best practice. New therapeutic interventions (e.g., chronic azithromycin) can be expected to work for some with CF, but not all. Clinicians may find it appropriate to expand the use of approved therapies (e.g., TOBI) to patients who would not have been eligible for study participation or for indications not subjected to clinical investigation. Finally, not all clinical research demonstrating effectiveness of an intervention will be compelling enough to change practice. Features beyond the study design must be considered. When interventions not well supported by clinical research results are advocated by experts or individual practitioners, the five topics outlined can serve as framework, making the reasoning explicit and open to rebuttal.

The optimal practice of clinical medicine, though requiring the knowledge of the results of clinical research, demands that clinicians attempt to deliver the best care to a particular individual. This cannot be accomplished by an appeal to clinical research results alone.

The 

Patrick A. Flume, M.D.

A clinical practice guideline is a systematically developed statement designed to assist practitioners and patients in making decisions about appropriate health care for specific circumstances (1) . The Cystic Fibrosis Foundation has invested considerable resources into the development of guidelines for the diagnosis and management of cystic fibrosis (CF) and its complications. As these guidelines are only recently published, it is premature for us to ask whether these guidelines have had a meaningful effect in the lives of patients with CF, but in principle, if clinicians provide interventions which are known to improve clinical outcomes, then patients should benefit (2) .

The development of guidelines alone is not sufficient for there to be benefit. The critical "next step" is the implementation of those guidelines. A lesson learned from guidelines in other disease conditions is that clinicians often don't utilize them. Studies have demonstrated that patients receive only about 50% of recommended evidence-based therapies, and mostly because of a lack of using recommended therapies (3) . The key factors of successful guidelines are: they should cover an area of practice which varies greatly and clinicians are aware of this; their evidence is fairly secure; the indication for their use is common in the practice of the targeted clinicians; clinicians are aware of the gaps in their own knowledge about the clinical condition covered by the guidelines; and the benefits of implementation are large (2) . For patients this means that the change will improve outcomes and for clinicians, the care should be quicker or easier.

We can use the guidelines on the use of chronic medications to maintain lung health (4) as an example of guideline development and implementation. These guidelines were meant to address an area of great importance to the CF community. The CFF Patient Registry (CFFPR) had demonstrated to us that there was wide variation in the use of such medications across the center network (5) . The development of the guidelines followed strict methodologies to ensure sound recommendations based on the evidence. A committee constructed the relevant questions, a systematic review of the literature was performed, and there was an a priori plan of how to assess the evidence.

The CF community was asked to offer commentary prior to submission for publication and the manuscript reporting the guidelines were subject to peer review. The CFF then invested resources into the education of the CF community regarding the guidelines including webinars and printed materials.

Since that time, there has been additional evidence that supports the recommendations provided in the guidelines. Analyses of the CFFPR have shown a decreased hospitalization rate in those patients on chronic macrolides over 2 years (6), a reduced mortality in those patients using inhaled tobramycin and/or dornase alfa (7) , and a slower rate of decline of FEV1 in patients using inhaled antibiotics (8) . The Canadian Study of High-Dose Ibuprofen in CF was under-powered and yet still found a reduced rate of decline of FVC but not FEV 1 (9) , and an analysis of ibuprofen usage in the CFFPR found a 23% reduction in rate of decline of lung function in treated compared to those not treated (10) . Novel analyses of adherence to medications have shown a lower rate of hospitalization (11) and a reduced occurrence of exacerbations (12) .

Just as the development of the guidelines was based upon a systematic approach, the implementation of guidelines is facilitated by a systematic approach. Success is often based upon assuring that all patients receive the therapy. This can be accomplished using a quality improvement approach that engages the clinical team, patients and families, educating all to be sure there is agreement, and then executing an action plan to make these guidelines part of standardized care. A successful implementation of the chronic medications guidelines has been demonstrated at one center where prescription of medications increased from ~60% to ~90% of eligible patients (13) .

Such reports are encouraging, but there remains one additional challenge acknowledged in the guidelines. We should not presume that all patients should be treated with all recommended therapies, so how should we introduce therapies to our patients? In addition, there has been an increasing trend by clinicians to introduce therapies earlier under the premise that this will better preserve lung function in the long run. The evidence supporting the medications included in the guidelines primarily demonstrated improvement in lung function (i.e., FEV 1 ), but we recognize that changes in FEV 1 are a relatively late event of the ongoing pathogenesis of CF lung disease. So if we choose to intervene earlier, how do we rationally institute a practice change? What are the appropriate diagnostic tests and how will we measure success?

There is one diagnostic test that has already established a consistent place in the evaluation and treatment of CF patients -periodic culture of the oropharynx. The practice of antibiotic treatment of early Pseudomonas infection has become fairly standard among CF centers. This is done with the belief that early infection can be eradicated and the time to establishment of chronic infection can be delayed. Once chronic infection has been established, the use of chronic aerosolized antibiotics should be used as recommended in the guidelines. Such an approach is well-accepted for the treatment of Pseudomonas, and there is increased interest in a similar approach to other bacterial pathogens.

Whereas we have a clear diagnostic test to guide the use of inhaled antibiotics, what about the other medications recommended in the guidelines? We recognize that the rate of disease progression is variable among patients; there are patients with milder manifestations of disease such that many have evaded diagnosis until later in life. Some have suggested that computed tomography (CT) of the chest or bronchoscopy might provide information that could guide therapy, but these come with concerns of increased cost, potential toxicity, and an increased treatment burden and they have not found a place in standard practice. Epidemiologic studies have shown the relationship between early nutritional status (e.g., weight and height for age percentiles) and eventual pulmonary health (14) . Such observations may provide the needed information about when to introduce therapies (even long before symptoms), but they do not demonstrate which therapy to introduce first. We can assess tolerance of the medication, but assessment of benefit will be more challenging to prove. In addition, we must consider the possibility of unintended consequence that could occur with therapies.

In conclusion, it is likely that the clinical practice guidelines will provide benefit to patients, but only if they are followed. That is, the clinicians must prescribe the medications and the patients must adhere to a treatment regimen. The key to success of implementation of the recommendations contained within the guidelines is to develop a systematic approach as in a quality improvement project until they become standardized therapy. For those patients who are not yet on therapies, a systematic approach should be used to assess, diagnose and treat with these therapies. It is highly unlikely that formal clinical trials will be performed that will answer the question as to which therapy first and when, and so the CF community must standardize their approach and share their results. Only in these comparative effectiveness studies will we be able to know the optimal approach with greater certainty Although cystic fibrosis (CF) has an increasingly good prognosis, early death is still too common. This review focuses in particular on what we can learn from severe asthma; most asthmatic children respond to low doses of inhaled corticosteroids, and detailed protocols have been described to rationalise management (1). In summary, the umbrella term used at the start of the process is "problematic, severe asthma" (2) and the steps are to determine:

1. Is the diagnosis correct ("not asthma at all")? 2. Are there significant co-morbidities ("asthma plus")?

3. Is it "difficult" asthma (improves if the basics are got right)?

4. Is it true severe, therapy resistant asthma, which remains a problem even when the basics are got right, and therefore may be treated with "beyond the guidelines" therapy?

The hypothesis is that a similar approach may be useful when pulmonary CF is apparently very severe. There is no evidence to guide us in what we term "challenging CF"; what follows is personal practice based on our experience with asthma.

Definition of "challenging CF." These are arbitrary (as in severe paediatric asthma), and the following definitions are suggested. Children may become challenging very quickly (e.g., failure to improve after two or more courses of intravenous antibiotics in quick succession) or gradually, with a series of almost subclinical deteriorations.

-Any child with lung function > 2 Z-scores below CF charts (3) -Any child who receives > 3 courses of intravenous antibiotics annually (whether elective or unplanned) -Any child requiring home oxygen -Any child in "nutritional failure" (e.g., body mass index (BMI) > 2 Z scores below the mean; drop in weight or body mass index (BMI) centiles by 10% over a year) -Any child with a severe CF pulmonary complication (e.g., massive hemoptysis, pneumothorax, therapy resistant ABPA, oral steroid dependent) -Any child whose self-or parent-reported symptoms are significantly overestimated or underestimated -Any child in whom there is refusal or extreme reluctance to give prescribed treatment by the caregivers Challenging CF: Is it really CF? In most cases the diagnosis will have been firmly established. It is at least worth considering whether another diagnosis with a specific therapy may have been missed.

Challenging CF: Is it CF plus? Of the comorbidities which may complicate asthma, obesity and food allergy are not likely relevant in CF. The relationship between upper and lower airways is complex (4), and severe rhinosinusitis should be sought and treated. Gastroesophageal reflux (GER) and dysfunctional breathing patterns (e.g., vocal cord dysfunction) may both complicate CF. GER may be difficult to diagnose, lipid laden macrophages have a good negative but a poor positive predictive value, measurement of pepsin in lavage may be helpful (5) .

It is CF, comorbidities have been excluded as far as possible, now what? The next step is a detailed multidisciplinary assessment, which includes a home visit. By analogy with asthma (6), the following issues are addressed:

-Medication adherence: relevant in more than half the asthmatics. Prescription records are obtained, whether the medications are accessible and in date, and whether an inappropriately young child is being left to take medications unsupervised is determined (7) . Consider measuring blood levels of appropriate medications (prednisolone, theophylline, itraconazole, voriconazole) and using nebulisers or airway clearance devices with a microchip which records adherence.

-Chest physical therapy practice: if equipment is prescribed, is it available or obviously unused at the back of a cupboard? -The home environment: exposure to tobacco smoke is checked by cotinine measurements, allergen sensitization and exposure is checked, especially the presence of mould or fungi.

-Psychosocial issues: most referrals to psychology in the context of asthma were made only after discussions in the home, where parents are more likely to talk about sensitive issues.

-Education: knowledge of CF and the treatments is re-checked.

After gathering these data, a multi-disciplinary team meeting decides whether further more detailed investigation is needed, or whether addressing the basic management steps is the way forward. Clearly if standard therapies such as rhDNase, hypertonic saline and macrolides are not optimally deployed, then this should happen before invasive investigations are undertaken.

The next step: detailed investigation. Nutrition and pulmonary status are so tightly bound up that both should be evaluated; the detail of nutritional assessment will depend on BMI. Some of the tests may already have been done in the immediate past, and may therefore not need repeating.

-Assessment of lung structure and function: inspiratory and expiratory CT scan, full lung function including bronchodilator reversibility, and lung clearance index.

-Fibre optic bronchoscopy, broncho-alveolar lavage and endobronchial biopsy to exclude infection with atypical organisms, airway malacia and other unexpected structural disease, and determine the pattern of airway inflammation. Consider 16s rRNA studies and anaerobic culture, although results may be difficult to interpret. Induced sputum may be a complementary investigation.

-Exclusion of GER by pH monitoring, give consideration to an impedance study, and refer to gastro-enterology if any doubt.

-Exclude unexpected nocturnal sleep disordered breathing with polysomnography.

-Exclude insulin deficiency: random blood sugar monitoring, oral glucose tolerance test, continuous monitoring (CGMS).

-If there is concomitant nutritional failure, a full assessment by the dietitian including a 3 day fecal fat, celiac screen, stool culture for Giardia, and urinary electrolytes. A referral to gastroenterology is made if there is still diagnostic doubt.

What next? Unlike in asthma, there is insufficient evidence to be dogmatic about the outcome of this sort of process. Clinical scenarios we have encountered include:

-"Distal and dry"' bad lung disease -distal airway disease with air trapping on inspiratory and expiratory CT scans, little sputum and minimal airway secretions on bronchoscopy. GER has to be excluded, and therapeutic choices include pulsed methyl prednisolone (which perhaps is underused because of fears of adverse effects), intravenous immunoglobulin (8) and other immunosuppressive agents.

-"Pan-airway and productive" -bad proximal bronchiectasis and marked purulent airway secretions. We use rotating nebulised antibiotics (TOBI™ and colistin month on month off of each; there may be a role for AZLI™ here); planned three-monthly courses of intravenous antibiotics; and even long term home intravenous colistin.

-"Inflammation predominant" -Very occasional responses to methotrexate and cylosporin have been described on an anecdotal basis (9) (10) (11) (12) . Prior to an attempted trial of these agents, a most careful evaluation of more conventional approaches is mandated. Immunosuppression brings with it the risk of invasive infections such as aspergillosis (13) .

-Psychological issues which are of course part of non-adherence, but there is much more to it than this. In particular in the teenage years, a lot of work may be needed. Issues include breakdown of family relationships and the use of non-adherence as a weapon; "over-calling" the child's symptoms; dysfunctional breathing patterns.

Challenging CF: Conclusions. We propose the use of a protocol based on that used in really severe asthma. In the context of asthma, we have learned that much "nightmare asthma" is readily amenable to basic management if implemented effectively. We hypothesise that the use of a similar protocol will identify patients with CF who need closer attention to conventional management, and a subgroup who will require innovative therapies.

References Nontuberculous mycobacteria (NTM) are ubiquitous in the environment existing as biofilms, aerosols, or dusts. NTM are extraordinarily hardy, and can withstand standard disinfectants, chlorine, glutaraldehyde, mercury, and formaldehyde. They have intrinsic resistance to broad classes of frontline antibiotics used for the treatment of bacteria. NTM survive under a wide range of conditions, and have been isolated from distilled water, hospital water heated to 55°C, ice machines, public drinking water and even showerheads. Therefore, it is not surprising that these organisms are often present in sputum from patients with cystic fibrosis (CF)(1).

In the setting of CF lung disease, the significance of NTM in respiratory secretions is a major diagnostic and therapeutic challenge. NTM have traditionally been categorized by the Runyon classification. In this system, organisms from group III (slow-growing), such as M. avium complex (MAC), and group IV (rapid-growing), such as M. abscessus, appear most commonly as CF pathogens. The incidence of NTM disease appears to be increasing in CF, and also in other lung diseases. Increased recovery of NTM in CF airway samples is due in part to specialized processing techniques that have reduced overgrowth by bacterial co-infections, as well as increasing physician awareness. The possibility also exists that exposure to NTM is becoming more common in our modern environment. However, acquisition of NTM is clearly an age-related complication of CF (2), and as median survival increases, so will the prevalence of NTM infection.

Who to treat. When NTM is first recovered from a CF respiratory sample, thorough consideration is needed to determine if treatment is required. Not uncommonly, NTM is recovered once, and is not seen in subsequent cultures (2) . In other patients, NTM is recovered on multiple occasions, but with no apparent clinical consequences (3). However, in some patients, NTM is clearly associated with accelerated progression of lung disease. As the usual signs, symptoms, and radiographic features of NTM disease broadly overlap with CF lung disease, it is frequently difficult to determine when NTM is an active pathogen. Although Olivier et al. suggested an algorithm to diagnosis NTM disease in the setting of CF based on their nested cohort study, there have been no prospective trials to validate this algorithm (3) . Important considerations include the species of NTM, quantity of organism, number of sep-arate positive cultures, clinical and radiographic evidence of disease progression, and exclusion of other causes of accelerated disease progression. Generally speaking, rapid growers such as M. abscessus are more virulent than slower growers such as M. avium (4) . Nevertheless, slow growing NTMs are capable of causing disease in CF patients.

Finally, it appears to us that NTM in children is more likely to have significant and sometimes devastating consequence, whereas an indolent clinical course is more often seen in older patients or those with an adult diagnosis.

The following diagnostic approach is one followed at the Colorado CF Center, and generally follows ATS consensus guidelines (5) . A minimum of two positive cultures and radiographic progression must be present. However, the most important consideration is that clinical decline is unexpectedly rapid in the setting of aggressive treatment of known pathogens and comorbidities. Often NTM are first looked for (and found) in association with a clinical downturn. An essential part of determining the significance of a new NTM infection is to first aggressively treat other infecting pathogens such as P. aeruginosa or S. aureus infections with IV antibiotics. If the patient does not return to baseline with treatment, we consider a second course of antibacterial antibiotics. A thorough review of airway clearance and other basic elements of CF care should be performed, and therapies maximized. Finally, all potential comorbidities must be addressed, in particular new bacterial infections (i.e. B. cepacia), CF-related diabetes, poorly controlled sinusitis, chronic aspiration and ABPA. Through exclusion of these well-recognized causes of accelerated decline, the contribution of NTM can be established. The case for treatment should be presented to the patient and family, and it is important that both the caregiver and patient be convinced of the need for treatment, as therapy is often difficult and a variety of complications and side effects are likely. As part of the initiation of treatment, the patient should be made aware that there is no proven treatment strategy, the rate of success is not known, and even if treatment is successful the patient is at risk for reactivation or re-infection. However, in many cases treatment is clearly beneficial.

How to treat. In the absence of CF specific recommendations, an initial approach to treatment is based on guidelines developed for the general population (5) . For M. abscessus, our standard approach is three-drug thera-py with parenteral amikacin and imipenem (or cefoxitin), combined with an oral macrolide for 3 months. Following this acute intensive treatment phase, therapy can often be changed to an oral macrolide and a fluoroquinolone, in combination with inhaled amikacin. For MAC, our standard approach is 3-drug therapy with a macrolide, ethambutol, and rifampin. However, in the presence of cavitary disease, amikacin can be used for the first 3 months of treatment, and continued up to 6 months if sputum cultures remain positive. Clearly, antibiotic selection will need to be individualized based on pre-existing allergies and toxicities. To a lesser extent, the antibiotic susceptibility of the organism may be a consideration as well; in particular the presence of resistance to clarithromycin in patients who received longterm therapy with azithromycin. Initially, the CF team needs to work closely with the patient to find a drug combination that is tolerated, and in some cases this approach is most effectively accomplished in the inpatient setting. Once a stable drug combination is identified, a critical consideration is monitoring of drug pharmacokinetics, as CF-specific alterations in oral absorption and volume of distribution can result in unexpectedly low drug levels (6) . In general, administration of oral agents with food and enzymes appears to enhance absorption, which may directly contradict recommendations for the general population (i.e. take on an empty stomach). Monitoring for drug-related toxicities is mandatory, such toxicities include development of ototoxicity associated with aminoglycosides, and visual changes associated with ethambutol. A CBC (with platelets) and liver function tests are checked at each visit to monitor for bone marrow suppression and hepatitis, which can be induced by a variety of the antimicrobials. Patients are typically seen in clinic every month during treatment, for sputum culture and laboratory analysis in addition to routine spirometry and physical exam. Standard CF care should be continually emphasized, especially aggressive airway clearance. Throughout the course of treatment, the potential for pulmonary exacerbation due to P. aeruginosa, S. aureus or other typical CF infections must be considered, even if the NTM treatment may have partial overlapping antimicrobial activity against these organisms. When to stop treatment. As with other aspects of NTM treatment, the duration of therapy is not known. In non-CF lung disease, success is defined by clearing the sputum of organisms. The ATS recommendation is that treatment continue for 12 months following clearing of the sputum (5). In our experience, this endpoint is often difficult to achieve. As other biofilm-forming infections such as P. aeruginosa are not routinely cleared from the CF airway, it is expected that NTM infections are sometimes impossible to eradicate as well. However, as with P. aeruginosa, we have seen sustained clinical response to treatment following antimicrobial therapy, despite continued presence of NTM in sputum isolates. Our approach is to initially attempt to clear the sputum, with a prolonged (6-12 month) treatment course. If treatment is not well tolerated despite attempts to introduce alternative drug combinations, a shorter course of therapy can be offered, with clinical improvement as an endpoint. In all patients with a history of NTM, we monitor all subsequent cultures for NTM, expecting that additional courses of NTM treatment will be needed episodically.

References and MsbA (3) and the mammalian drug transporter ABCB1 (4). This type of ABC transporter has 2 x 6 transmembrane alpha helices forming two transmembrane domains (TMDs) and these each have four long intracellular helical extensions of transmembrane helices which form cytoplasmic loops (ICLs). Four of these loops bind to pockets in the two cytoplasmic nucleotide binding domains (NBDs), with one loop from each transmembrane domain crossing over to contact the NBD on the opposite side of the transporter (1, 2) . The structures of Sav1866, MsbA and ABCB1 show overall folds that are broadly similar, however ABCB1 (and some lowerresolution structures for other MsbA orthologs) show a wide separation of the NBDs and an inward-facing (i.e., cytoplasm-facing) orientation of the two TMDs (3, 4) . In contrast, Sav1866 and some of the MsbA orthologs display closely associated NBDs and a more outward-facing configuration of the TMDs (1-3). These structural studies imply that major conformational changes can occur in the relative orientation of the four domains. Similar, but smaller conformational changes have also been observed in a separate group of bacterial ABC proteins which carry out the import of substances into the cell (5) (6) (7) (8) (9) (10) (11) . Large conformational changes were previously proposed to be part of the substrate translocation mechanism for ABC transporters, and the formation of a sandwichdimer of NBDs in the presence of nucleotides will be addressed in more detail in my presentation (12) .

Structural data derived from 2D crystals of CFTR have been published (13) . These crystals were obtained in the presence of the non-hydrolysable ATP analogue AMP-PNP (13) , and with two separate crystal forms present. The data generated by these studies was estimated to have a spatial resoilution of ~20Å, typical for negatively-stained specimens where the detail is limited by the granular nature of the uranyl acetate stain and by dehydration in the high vacuum of the microscope (14, 15) . Despite these drawbacks, the structural data showed that CFTR within the 2D crystals was in a monomeric state, and that it was structurally homologous to ABCB1 (16), as expected. Conditions have now been found that yield better 2D arrays of CFTR and higher resolution data has been extracted from these crystals in the unstained and frozen-hydrated state (Rosenberg, M.F., et. al., submitted). These conditions require the absence of nucleotide, hence the protein should be quiescent. Under these conditions, the protein was packed in two crystalline layers stacked on top of each other. This new unit cell packing and the structures of the two CFTR molecules in the unit cell will be described in the symposium.

Single particle electron microscopy has been employed to study CFTR structure in the non-crystalline state. An early study of CFTR embedded in a heavy-atom stain reported dimeric complexes (17) . A later study also reported dimeric complexes, although there was an unexpected void in the centre of the complex (18) . Later, structural data were obtained for purified CFTR single particles in the unstained and frozen-hydrated state (19) and these data also showed dimeric CFTR complexes consistent with two side-by side molecules but with no large cavity in the centre of the dimer. Labelling of the CFTR particles using 1.8nm-diameter gold particles was used to establish domain-domain interactions (20) , and provided the first direct 3D data on the location of the Rregion. I will describe the locations of the CFTR domains and discuss the correlation between the low-resolution domain labelling studies and the recent higher resolution 2D-crystal studies.

References As a member of the ABC-transporter superfamily of proteins, CFTR is composed of multiple modular protein domains. The CFTR functional complex is composed of a dimer of transmembrane domains (TMDs), a dimer of nucleotide-binding domains (NBDs), and a regulatory (R-) domain. ATP binding between the NBDs putatively induces NBD dimerization. This dimerization, in turn, is coupled to channel gating via NBD interactions with the intracellular loops of the TMDs. Disruption of local domain structure and alterations in the appropriate domain-domain assembly are associated with multiple disease-causing CFTR mutations (1-3).

Structures of mammalian and bacterial ABC transporter proteins, including the NBDs from CFTR, provide templates for understanding the biosynthetic and functional defects associated with these CF-causing mutations (3) (4) (5) (6) . Despite topological differences in the TMDs, the available ABC-transporter structures demonstrate conservation across two functional interfaces. Multiple disease-associated mutations can be mapped to both of these conserved domain-domain interaction interfaces.

The TMD-NBD interfaces are formed by intracellular loops from the TMDs and one surface of each NBD. These contacts putatively couple movements in the NBDs to conformational changes in the TMDs and regulate channel gating. Previous studies have shown that disruption of this interface leads to defects in channel biosynthesis and function (7, 8) . Recent work has also shown that alterations to this interface can be made to promote the proper trafficking of mutant CFTR, suppressing the effects of disease-causing mutations, including ∆F508 (9).

The second interface, the NBD-NBD interface, is critical for ATP binding and hydrolysis. The ATP-binding and hydrolysis sequences are distributed across both NBDs as "half-sites." ATP binding leads to NBD dimerization and complementation of the ATP-associated sequences across the NBD dimer (10) . While alterations in the NBD-NBD interface and canonical ATP-associated sequences often result in changes in channel function, biosynthetic defects have also been associated with several of these sites (8, 11) .

Using multiple approaches to evaluate these domaindomain interactions, we are refining our understanding of the physical basis of these interactions. Computational, biochemical and cell biological approaches have been employed to probe structural and functional constraints of the TMD-NBD and NBD-NBD interface. Both interfaces show limited plasticity; however, specific alterations within both interfaces have positive impacts on domain-domain interactions.

Multiple human ABC-transporters show altered biosynthesis and trafficking as a result of mutations within the TMD-NBD interface. The sensitivity of these systems to substitution at the TMD-NBD interface demonstrates potential commonality in the biosynthetic pathways of this protein family. Additionally, profiling disease-causing mutations within the intracellular loops of human ABCtransporters and tolerated versus non-tolerated substitutions provides insight into the physico-chemical requirements for maintenance of the TMD-NBD interaction interface.

The ability to positively influence protein trafficking and function suggests that small molecule targeting to these interfaces may serve in the development of corrector and potentiator compounds.

Acknowledgements: This work was funded by the NIH and CFF.

References CFTR misfolding and subsequent degradation constitute a primary defect in most patients with cystic fibrosis (1) . Therefore, an important goal in developing novel CF therapies is to better understand how CFTR acquires and maintains its folded state in the cellular environment. In particular, attention has focused on NBD1 where the ∆F508 mutation confers a temperature-sensitive phenotype, decreases thermal stability, interferes with interactions between NBD1 and the membrane spanning domains, and changes NBD2 (2) (3) (4) (5) . WT NBD1 exhibits a melting temperature near 37°C which is reduced by loss of Phe508, resulting in an unstable form of the mutant protein. Moreover, refolding of NBD1 in vitro is very inefficient, suggesting that proper NBD1 structure is dependent upon the specific folding pathway. Because it takes 5-7 minutes to synthesize full length CFTR protein in cells, whereas secondary protein structure forms on a millisecond time scale, many aspects of folding, i.e., membrane insertion (6) and initial domain folding (7) occur cotranslationally as nascent CFTR emerges from the ribosome. This initial domain folding is followed by slow domain assembly (8) and release from a complex set of chaperone machinery (9) , as confirmed by the timing of proteolytic changes in the various CFTR domains and its relatively long residence time (30-120 min) in the ER.

Recent studies using limited proteolysis and fluorescence resonance energy transfer (FRET) have begun to tease out early folding steps that give rise to NBD1 struc-ture. A primary defect in ∆F508 CFTR arises in NBD1 and can be identified by characteristic proteolytic patterns for the recombinant, in vitro translated, and cell expressed NBD1 domain, as well as full-length CFTR. This basic defect can be rescued by a secondary suppressor mutation, I539T, which like Phe508 is located in the α-subdomain region of NBD1 (residues 495-564). A fascinating feature of this suppressor mutation is that mice already carry a Thr residue at this position, which likely explains increased NBD1 WT and ∆F508 thermal stability compared to human NBD1, and perhaps their milder lung disease phenotype. Consistent with the above-mentioned results, I539T fully rescues α-subdomain stability and conformation (10) . Examination of ribosome-bound nascent chains by FRET has further revealed that NBD1 folding begins cotranslationally with rapid collapse of a short N-terminal subdomain (residues 389-~500) that forms a minimal ATP-binding site (11) . This region contains the three-stranded ABC β-subdomain, the unstructured regulatory insertion (RI), as well as regions previously ascribed to the F1-type ATPase core. The CF-causing mutation L441P interferes with N-terminal subdomain folding and disrupts CFTR trafficking, whereas removal of the regulatory insertion, which improves trafficking of the full-length protein (12) , also improves N-terminus folding efficiency. Thus efficient folding of the N-terminal subdomain appears to be a key event that may provide a template upon which the subsequent α-subdomain and the α/β core assemble. Impor-tantly, kinetic analysis of N-terminal subdomain folding, as well as analysis of regions C-terminal to this subdomain, indicates that ∆F508 specifically impacts subsequent folding of the α-subdomain and/or the α/β-sheet core.

References CFTR has two fascinating distinctions. First, it is the only member of the ABC transporter superfamily that is known to function as an ion channel. Other ABC transporters are pumps that use the free energy from ATP hydrolysis to drive substrate transport uphill against concentration gradients (1) . Anion transport through the CFTR channel is a passive process driven solely by prevailing electrochemical gradients. Second, CFTR is the only known ion channel that couples an enzymatic activity to channel gating. Channel opening is conformationally linked to ATP binding and the associated dimerization of the two nucleotide binding domains (NBDs). Channel closing is facilitated by ATP hydrolysis and the subsequent release of the hydrolysis products from one of the two ATP binding sites (2) (3) (4) (5) . The rate of channel closing in the absence of ATP hydrolysis is otherwise very low due presumably to the tightness of the NBD dimer (5) . Thus, the ATPase activity of CFTR has been adapted to energize a cycle of channel gating rather than chloride transport against an electrochemical gradient.

This unique enzymatic activity of the CFTR channel begs the question-how does its gating compare to that of conventional ligand-gated channels? The gating of more typical ligand-gated channels (e.g., nicotinic acetylcholine channels) obeys allosteric principles described decades ago (6-9). The first principle of an allosteric mechanism for a typical ligand-gated channel is that ligand binding shifts the equilibrium between closed and open states but is not absolutely required for channel opening. Unliganded openings are possible (albeit with a lower probability) as are constitutive mutations that increase the frequency of unliganded openings. Such mutations often locate to the symmetry axis between the ligand binding site(s) and the channel gates (7) (8) (9) . CFTR differs from conventional ligand-gated channels because it usually consumes its ligand (ATP) during the gating cycle. But, it is becoming clear that CFTR channel gating does share some important features with the gating of conventional ligand-gated channels that reversibly bind their ligands. Several groups have observed that wild type CFTR channels can open (albeit rarely) in the absence of ATP (10, 11) . In addition, a number of constitutive mutations that enhance unliganded CFTR activity have been reported (11, 12) . Our group produced several such mutations within the long cytosolic loops that connect the NBDs to the transmembrane domains (TMDs) (11) . These constitutive mutations locate to the putative symmetry axis of the CFTR channel between the NBDs and the TMDs when mapped onto the available crystal structures of other ABC transporters. The constitutive mutations in the cytosolic loops also enhance ligand (ATP) sensitivity, a cardinal feature of a classic allosteric gating process (reciprocity principle). These findings support a "hybrid" gating mechanism for CFTR channels that blends the allosterism of a ligandgated channel with its unique enzymatic activity (13) .

The constitutive mutations that we and others have produced can be used as tools to explore fundamental aspects of CFTR channel regulation. Two specific examples that we discuss here are: (i) defining the mechanisms by which R domain phosphorylation enhances CFTR channel activity; and (ii) promoting unliganded activity as a means to circumvent the defective gating of CF mutant channels. Regarding the first example, the main physiologic cue for CFTR channel activation in vivo is phosphorylation of its R domain by cyclic nucleotide dependent kinases (14) . The underlying mechanisms are poorly understood although there is evidence that NBD dimerization is regulated by R domain phosphorylation (15) . By introducing constitutive mutations into channel constructs that cannot dimerize their NBDs (e.g., an NBD2 deletion construct), we determined that R domain phosphorylation strongly promotes CFTR channel activity in the absence of either ATP binding or NBD dimerization (11) . These and other data support a model in which the R domain interacts with the TMDs and/or the connecting cytosolic loops to modulate channel opening. Regarding the second example, the most prevalent CF-causing mutations (G551D and F508del) strongly disrupt ATP-dependent channel gating. Importantly, the very low activities of these CF mutant channels can be markedly improved by introducing constitutive mutations in the cytosolic loops (11) . Defining how these loop mutations promote unliganded channel opening may inform new strategies for augmenting the activities of CF mutant channels. (4) , and with the support of the US Cystic Fibrosis Foundation, states throughout the United States began implementing CF newborn screening programs. By December 2009, babies in all 50 states and the District of Columbia were being screened for CF. Prior to implementation of newborn screening, babies would typically present with failure to thrive, steatorrhea, respiratory symptoms, or other complications of untreated CF. While some infants with CF diagnosed by newborn screening still present with symptoms, and 15-20% of infants are still presenting with meconium ileus, US CF clinics are seeing a different type of infant with CF than in previous decades. New guidelines provide treatment and monitoring recommendations for the care of infants, with a focus on infants identified early through newborn screening (5,6); however, the US CF community is still gaining an understanding of the new face of the infant with CF.

Every year, there are approximately 1,000 new CF diagnoses in the United States reported to the US CFF Patient Registry (7) . The proportion of new diagnoses made by newborn screening grew from 5-10% in 1996-2001 to almost 45% in 2009 (420/960), demonstrating an annual increase in individuals diagnosed by newborn screening. The median age of diagnosis in the 1990s and early 2000s was approximately six months of age in the United States (8) . Current US CFF data show a decrease in the median age at diagnosis to less than two months of age in recent years. Almost 62% of all children under the age of 2 followed at US CF centers in 2009 were diagnosed by newborn screening, and an additional 18% were identified by meconium ileus (7) . Over 80% of children who are followed at US CF care centers under the age of two were identified with an early diagnosis.

Many reports have described the differences in the early course of CF between children diagnosed conventionally after the development of symptoms and those identified by newborn screening, primarily focusing on the benefits of early diagnosis on long term outcomes (9) (10) (11) or on the severe complications of a late diagnosis (12, 13 ). An analysis of the US 2000-2002 CFF Patient Registry demonstrated that infants diagnosed with symptoms had a higher preva-lence of complications throughout childhood compared to those diagnosed by newborn screening. The prevalence of stunting and wasting in the year of diagnosis, defined as height or weight for age less than the 3rd percentile, respectively was reported at 3 times the rate in symptomatically diagnosed infants (stunting 213/810, 26%; wasting 265/815, 33%), compared to newborn screened infants (stunting 21/239, 9%; wasting 26/240, 11%). Pseudomonas aeruginosa was identified in the year of diagnosis at almost twice the rate in symptomatic infants (228/774, 29%) compared to newborn screened infants (32/211, 15%, ref 14) .

Fewer complications in early childhood are reported to the US CF Registry since nationwide implementation of CF newborn screening. In infants without meconium ileus, the annual prevalence of stunting in children diagnosed in the first two years of life declined steadily from 6-10% in the era before widespread implementation of newborn screening (1996-2003) to 2-4% in 2007-2009 (7) . Wasting in the first two years of life declined steadily in the years between 1996 and 2009, from a high of 15% in 1996 to approximately 3% in 2009. The cumulative incidence of P. aeruginosa infection in the first two years of life decreased from 30% to 23% between 2002 and 2009. While other complications including hypoelectrolytemic dehydration, bleeding and hemorrhage related to vitamin K deficiency are not closely tracked by the US CFF, we believe that these complications are being seen less frequently because of earlier diagnosis and initiation of treatment.

Infants identified by newborn screening are presenting earlier and appear healthier than their counterparts in the era before newborn screening was widely implemented. Early bacterial infections and growth outcomes have improved dramatically since the introduction of screening. Many of these differences are due to the impact of earlier diagnosis and the initiation of aggressive care concurrent with the introduction of screening. However, the many factors that contribute to poor growth have not been defined and persist despite pancreatic enzyme supplementation. Although published guidelines for the clinical management of infants with cystic fibrosis exist (5), there is a scarcity of evidence to dictate care. New therapeutic trials to treat early P. aeruginosa in infants, and new initiatives to improve growth and nutrition are positively impacting the course of early CF in concert with newborn screening, yet we are in need of more studies and interventional trials in infants with CF to optimize outcomes in newborn screened infants. Newborn screening provides the opportunity for earlier surveillance and initiation of treatment prior to irreversible damage occurring, further changing the face of the newborn with CF.

References Although survival has improved for successive cohorts since the early 1960s data from British and US data registries confirm that whilst the median age of survival has improved, the mortality rate for adult sufferers has remained virtually unchanged for 4 decades. The failure to impact on adult mortality appears at odds with data that indicate a majority of individuals with CF have apparently normal lung function in mid-childhood. However, US data indicate that, consistent with mortality rates, the best modern treatment has failed to significantly alter the trajectory of decline in lung function in adults. One plausible explanation for this conundrum is that commonly used measurements of lung function, such as forced expiratory volume in one second (FEV1), underestimate the presence and progression of lung disease. Evidence supporting this assertion comes the Netherlands. De Jong et al. compared changes in FEV1 with assessments of structurual lung disease obtained using computed tomography (CT) using a validated scoring system (1). This group observed that despite stable lung function, CT scores worsened over time. de Jong and colleagues also observed structural changes in infants and young children with CF using CT. Therefore, currently available data suggest that structural changes detected by CT could be useful for detecting early manifestations of CF lung disease and for describing disease heterogeneity. Based on these observations, the Australian Respiratory Early Surveillance Team for Cystic Fibrosis (AREST CF) collaboration combined CT together with bronchoalveolar lavage (BAL) and measurements of infant and preschool lung function in an early surveillance program for children diagnosed with CF following newborn screening (NBS). This innovative surveil-lance program has provided the first comprehensive snapshot of the earliest pathophysiological changes in the lung associated with CF and the significant factors that contribute to structural lung disease, notably bronchiectasis.

Our first cross-sectional analyses have established that structural lung disease including bronchiectasis occurs soon after birth, is common in the first years of life and is associated with inflammation and infection (2, 3) . Limited (3-slice) CT and BAL were performed in children 2mo-6 yr with CF diagnosed following NBS. Scans were analyzed for the presence and extent of abnormalities. The prevalence of bronchiectasis was 22% and increased with age (P = 0.001). Factors associated with bronchiectasis included absolute neutrophil count (P = 0.03), neutrophil elastase concentration (P = 0.001), and P. aeruginosa infection (P = 0.03). Bronchial dilatation was detected in nearly 20% of children between 2 and 5 months of age and overall nearly 80% of children at this age have evidence of pulmonary disease manifest by infection, inflammation or radiological changes (2) .

Early lung function data using the raised volume rapid thoraco-abdominal compression (RVRTC) technique have also demonstrated that the mean FEV(0.5) z score did not differ between infants with CF and healthy control subjects less than 6 months of age (-0.06 and 0.02, respectively; P = 0.87). However, the mean FEV(0.5) z score was lower by 1.15 in infants with CF who were older than 6 months of age compared with healthy infants (P < 0.001) (4). FVC and FEF(75) followed a similar pattern. These data suggest that lung function, measured by forced expiration, is normal in infants with CF at the time of diagnosis by newborn screening but is diminished in older infants suggesting the optimal timing of therapeutic interventions aimed at preserving lung function may be within the first 6 months of life.

We are now in a position to report the first longitudinal data from the AREST CF populations.

We examined 301 paired 3-slice CT scans obtained 1 year apart in children whose first scan was at the age 1-3 years. Bronchiectasis was never present in 74 pairs (25%), and was detected at either the initial or subsequent scan in 227 pairs (75%). Bronchiectasis was detected at the initial scan in 133 scan pairs (44%) and persisted at the subsequent scan in 98 pairs (74%). Airtrapping present at the initial scan persisted in over 80% of subsequent scans. The extent of bronchiectasis increased over 1 year in 63% of scans, and air trapping in 47%. Radiological progression of bronchiectasis and air trapping was associated with severe CFTR genotype, pulmonary infection and worsening neutrophilic inflammation.

We have recently reported data from 37 infants who had at least two repeat lung function measurements (5) . Mean (SD) z-scores for FVC were -0.8 (1.0), -0.9 (1.1) and -1.7 (1.2) when measured at the first visit, one-year or two-year visits respectively. Mean (SD) z-scores for FEV0.5 were -1.4 (1.2), -2.4 (1.1) and -4.3 (1.6) respectively. In those infants in whom free neutrophil elastase was detected FVC z-scores were 0.81 lower (p=0.003), and FEV0.5 z-scores 0.96 lower (p=0.001) respectively. Significantly greater decline in FEV0.5 z-scores occurred in those infected with S. aureus (p=0.018) or P. aeruginosa (p=0.021). Pseudomonas aeruginosa infection in children less than 5 years detected by bronchoalveolar lavage can be eradicated in more than 85% of cases using standard antibiotic regimes indicating the importance of early detection of endobronchial infection (6) .

We argue that intervening at diagnosis to prevent irreversible lung damage in childhood will reduce adult mortality and that outcome measures need to be defined that reflect the onset, progression and heterogeneity of lung disease to enable clinical trials with potential diseasemodifying agents (7) . (1). The California NBS algorithm is a three-step process based on the determination of IRT as a first step. Specimens with an elevated IRT are sent to Stanford University Molecular Genetics Laboratory for CFTR mutation analysis (Asuragen Signature® CF 2.0 ASR) with a panel of 40 mutations determined to occur with high frequency in California (2, 3, 4) . A unique feature of the California NBS algorithm is that specimens with only one mutation identified undergo further testing by temporal temperature gradient gel electrophoresis (TTGE) and focused sequencing (Ambry Genetics Aliso Viejo, CA from July 2007 to July 2009; Stanford University Molecular Genetics Laboratory from 2009 and onward). This methodology is not constrained to the detection of known mutations, as in many commercially available panels. Thus, it can detect previously known as well as novel CFTR mutations and variants (4). Only infants whose specimens have at least two mutations or variations identified are considered positive to screening and referred to accredited CF Centers (CFCs) in California for confirmatory testing through sweat chloride determination. From the time of implementation 4 years ago, over 2 million babies have been screened and the incidence of screen positives has been approximately 1 in 4,200.

The application of more sophisticated genetic based assays to newborn screening algorithms results in an increased identification of infants with mutated CFTR alleles that result in less severely affected organ function. Although more than 1,800 variants in the CFTR gene have been identified to date, for many of these variants their pathogenic potential is plainly unknown because clinical data is sparse and/or they have not been functionally characterized ex vivo (5). In addition, among the California screen-positive newborns, novel variants have been identified in 4% of those with only one mutation identified by the California-40 panel.

Specific mutations and variants associated with normal sweat electrolyte levels and mild phenotypic manifestations have been described (6, 7) . Distinguishing sweat test-negative CF patients from just carriers of CFTR mutations is important for the early recognition of complications in those with the disease and for the provision of genetic counseling to carriers. Infants who are positive to the screen are referred to CF Foundation-accredited CF care centers for sweat testing and diagnostic follow up. Those with elevated sweat chloride test values are conclusively diagnosed with CF and followed up using current CF infant care guidelines. Those with lower values are clinically evaluated using follow up guidelines developed by consensus of the California CF Consortium of CF centers in partnership with the California Department of Public Health Genetic Diseases Branch (8) . This includes guidelines for preparation of infants for sweat testing, testing for fecal elastase levels, close monitoring of growth parameters and respiratory microbiology (oropharyngeal, nasopharyngeal or bronchoalveolar lavage depending on the clinical circumstances). Further, guidelines for repeat sweat chloride testing and result interpretation are in place.

Given that newborns identified by NBS are usually asymptomatic, we have embarked on a longitudinal project to monitor these infants for the development of any disease manifestations. Not surprisingly, only infants with severe mutations had sweat chloride values in the diagnostic range. The remainder had values at presentation below the indeterminate cut off (<30 mmol/L). Of importance, a significant proportion of these infants (25%) have developed clinical problems that have led to a diagnosis of CF based on the presence of mutations plus clinical manifestations, and this despite their low sweat chloride values (9) . Although CFTR genotype and sweat chloride are correlated, neither measure is a perfect predictor of the development of clinically apparent CF disease. Ruling out a diagnosis of CF should not be based solely on sweat chloride values. Our experience so far emphasizes the importance of longitudinal follow up at CF Centers regardless of initial sweat chloride results. The information being generated regarding the clinical and prognostic implications of a genetic diagnosis of CF based on the detection of alleles with variants and novel mutations will be of great importance to further inform the natural history of the disease in this genetically diverse group. The diagnostic criteria for cystic fibrosis (CF) necessitates that evidence of CFTR dysfunction must be present for a diagnosis to be made (1) . That criterion can be satisfied by identification of two known CF-causing mutations. The list of mutations included in the consensus statement on diagnosis was drawn from the 23 mutations identified by the American College of Medical Genetics (ACMG) as having direct or empirical evidence of sufficient loss of CFTR function to cause disease (when paired with another CF-causing mutation) (2). However many more mutations in the CFTR gene than these 23 with well characterized disease liability have been identified (3) . The Clinical and Functional Translation of CFTR (CFTR2) project was funded by the U.S. CF Foundation to address this gap in knowledge between well characterized (usually the most commonly seen) mutations, and the less commonly seen mutations with unknown disease liability. CFTR2 has compiled patient information on nearly 40,000 CF patients from North America and Europe to identify all mutations that are seen in CF patients, and then to prioritize those mutations for functional analysis.

One hundred sixty mutations are seen in nine or more patients in CFTR2, which represents an allele frequency greater than 0.01%, and, in total, account for 97% of the identified alleles. Nonsense mutations (n=35), mutations of the canonical nucleotides of the splice donor/acceptor sites (n=15), and insertion/deletion mutations that cause a frameshift (n=30) account for 50% of these 160 mutations. These classes of mutations would be expected to introduce a premature termination codon and lead to nonsense mediated RNA decay and no protein product. The clinical characteristics of patients who carry these types of mutations are all consistent with CF, with sweat chloride concentrations clearly in the diagnostic range and with high rates of pancreatic insufficiency. The remaining 80 mutations cannot be declared deleterious on the basis of their mutation type alone, and therefore require further functional assessment. Missense changes represent a particular diagnostic challenge as they can result in a protein that is produced in inadequate quantities by the cell, one that does not conduct chloride, or a protein that is fully functional despite the amino acid substitution. The CFTR2 project in association with Vertex Pharmaceuticals coordinated the production of cell lines stably expressing the missense mutations (n=65) among the top 160 for analysis of RNA and protein levels, protein maturation, and chloride conductance. Mutations predicted to alter splicing that have not been previously studied are being analyzed as well. As a result of these efforts, the number of mutations that can be rigorously characterized as CF-causing is expanded from 23 to 130 (details on www.cftr2.org). The remaining 30 mutations have varying clinical consequences as they sometimes are associated with CF, but not fully penetrant as they also may result in a single organ system CFTR-related disorder or no phenotype (4).

However, this still represents only a fraction of the total number of mutations that have been described. The CFTR2 project will continue to characterize mutations on the basis of the clinical manifestations in patients that carry them and the functional consequence of that nucleotide and/or protein change; but will never be able to totally annotate the genetic variability in CFTR. Therefore, the challenge of CFTR mutations of uncertain clinical significance and of lower penetrance must be considered in CF diagnosis and screening when genetic information is used in testing. The inclusion of a mutation of uncertain clinical consequence or incomplete penetrance in a newborn screening panel creates a situation (not unlike a borderline biochemical test) that leaves the family and primary care provider with unpredictable prognosis, and requires a detailed follow-up plan from the testing organization and possibly CF care center. The CFTR2 project will be able to better inform CF diagnosis, but will not eliminate all diagnostic uncertainties. It will remain up to the screening organization to balance the appropriate sensitivity and specificity for their screen with the resources available for testing and follow-up.

References 

Helge Hebestreit, M.D.

Since the early 1980s, reports of patients with cystic fibrosis (CF) participating in competitive sports including marathon running have shown that strenuous exercise is possible despite the multi-organ disease affecting lungs, muscles, sweat glands, etc. Nowadays, most patients with CF engage in some endurance activities, team and individual sports, and/or strength training (1). Recent observational studies have shown that a high level of physical activity is associated with a slower decline in pulmonary function, especially in girls (2) . Furthermore, physical activity correlates positively with exercise capacity in CF (3) which, on the other hand, is related to survival (4). Randomized controlled trials assessing the effects of physical training in CF have shown that regular exercise may improve physical fitness (5-10) and pulmonary function (6) (7) (8) 10, 11) in CF. Furthermore, several studies have shown that physical conditioning can improve quality of life in patients with CF (7, 9, 10) .

Although there is some evidence suggesting that a supervised exercise intervention might be more effective than a partially unsupervised intervention over short time periods, the latter approach has definitively shown longterm benefits (10,11) which may persist even after the intervention has ended (10) . Although regularly repeated activities will improve performance especially in tasks similar to the trained activities, with respect to pulmonary function training mode does not seem to matter in the long run. In other words, the effects of strength training and aerobic training on pulmonary function seem to be comparable (12) .

Given the available evidence in favor of an active lifestyle in patients with CF, regular physical activity and exercise is considered important for a patient's health by most CF health professionals (13, 14) . Exercise is well accepted among patients as well (15) 

Heather Chambliss, Ph.D.

Supporting behavior change can be a challenge for both health professionals and patients, yet it is the key for the long-term maintenance of physical activity and fitness. Skillbuilding strategies derived from behavioral theories can be incorporated across a variety of programs and settings.

One way to structure the physical activity counseling session is to consider an A's Approach (1):

Assess -explore current physical activity habits as well as factors that may support or hinder activity;

Advise -provide clear, structured advice considering patient needs, current behavior, self-efficacy, and physical activity guidelines;

Assist -provide behavioral assistance through behavioral skill building;

Arrange -develop a plan for monitoring short-term and long-term progress and provide feedback to support goals.

Health professionals can assist patients in meeting physical activity goals by supporting behavioral skill building. Behavioral skills which can be effective in mobilizing motivation and progress include self-monitoring, goal setting, problem solving, social support, stimulus control, and relapse prevention among others (2-4).

Self-monitoring is a skill which is consistently shown to be an important predictor of successful behavior change across health behaviors. The skill of self-monitoring involves keeping a record of current habits and future goals and can be accomplished using written logs, calendars, or technology.

Goal setting involves establishing concrete behavioral goals that meet "SMARTS" criteria. Both patients and professionals agree that goals are important; however, setting the right goal is the key.

When professionals assess factors that hinder physical activity, barriers will emerge. Problem solving, or following a stepwise approach to address barriers using the "IDEA" method, assists patients in overcoming barriers both in the current situation and in the future.

An important role of the health professional is to serve as a source of social support, providing knowledge and encouragement. Individuals should be encouraged to build a support team by identifying specific types of help needed and matching support to meet those needs.

Another behavioral skill is stimulus control, or identifying and managing triggers that prompt behaviors. Managing the environmental and psychosocial prompts that lead to healthy or unhealthy choices is a skill that helps people take control of their daily habits.

Finally, patients should be taught skills to help prevent relapse, or falling back into old behaviors. Relapse prevention teaches people how to identify high risk situations and plan ahead so that they can manage them more effectively. In addition, patients should be counseled on how to prepare for potential lapses so that they can get back on track when they do not meet their goals.

It can be helpful to consider behavioral skill building from a "toolbox" approach, so that each skill represents one tool that can be used as needed. Practitioners can help patients prepare for a physically active lifestyle by giving them the tools they need for long-term change. This type of approach can help increase motivation by establishing realistic goals, increasing confidence, and having a tailored plan of action. Thus, approaching behavior change using a skill building approach allows flexibility to tailor skills to individual needs, readiness for change, and fitness goals.

References 

Nancy Alarie, B.Sc. P.T.

There is no longer any doubt that physical activity and exercise contribute positively to the health and wellbeing of CF patients. There are several methods available to measure exercise tolerance (cardiopulmonary exercise testing, 6MWT, shuttle tests) but what about the assessment of a patient's level of activity? While the assessment of exercise tolerance is valuable and necessary it does not provide the clinician with information on the patient's day to day activities and may not be easily accessible to all clinicians. Without valid and reliable assessment tools the efficacy of our interventions becomes impossible to determine.

The World Health Organization defines physical activity as "any bodily movement produced by skeletal muscles that requires energy expenditure. The term "physical activity" should not be mistaken with "exercise." Exercise is a subcategory of physical activity that is planned, structured, repetitive, and purposeful in the sense that the improvement or maintenance of one or more components of physical fitness is the objective. Physical activity includes exercise as well as other activities which involve bodily movement and are done as part of playing, working, active transportation, house chores and recreational activities."

There are several methods available to monitor/measure habitual physical activity including: (a) doubly labelled water that measures total energy expenditure, (b) motion sensors (pedometers, accelerometers) that measure activity counts, (c) recall questionnaires and activity diaries, and more recently (d) portable physical activity monitor (SenseWear Pro 3 Armband, SWA, Body Media, Pittsburgh, USA). Each of these methods assesses a different aspect of habitual physical activity and each has definite benefits and drawbacks. The following is intended to provide an insight into the strengths and weaknesses of available tools.

The "Gold Standard" in the literature is described as direct observation. Although ideal and applicable in the research setting its use as a measure of daily patterns of activity is limited. Doubly labelled water (where the hydrogen and oxygen in water have been replaced by sta-ble isotopes) measures daily energy expenditures by calculating the elimination rates of the isotopes over time (by sampling of saliva, urine or blood). However this technique is costly and does not provide information regarding the pattern or intensity of activity.

Motion sensors detect body movements and provide an estimate of physical activity. Pedometers are small electronic devices used to estimate the mileage walked or total number of steps taken over a period of time. They are relatively inexpensive, re-useable, objective and nonreactive. On the downside they only report total counts over the observational period and cannot assess the intensity or pattern of the activities performed.

Accelerometers are more sophisticated electronic devices that measure accelerations produced by body movement. Accelerometers provide an objective, nonreactive, and re-usable tool for assessing physical activity and the newer generation of devices will be able to automatically detect patterns of activity (i.e., no need to manipulate the device). However, their ability to assess cycling, locomotion on a slope, or other activities with minimal torso movement is limited.

Subjective measures of habitual activity such as questionnaires and diaries are typically cheap, easy to administer, and often quick to perform. Only those that have been validated against more stringent measures of physical activity should be used.

The Habitual Activity Estimation Scale (HAES) has been used in several chronic illnesses. This scale has several benefits over other measures as it was designed to be used in the clinical setting, can be completed in less than 15 minutes (resulting in 100% completion rates), data can be quickly analyzed and clinicians can provide immediate feedback to patients and caregivers. Patients older than 11 years of age can complete it without assistance. The HAES has been shown to be a reliable and valid instrument for the assessment of physical activity in pediatric and adult CF populations.

The International Physical Activity Questionnaire (IPAQ) is a self report measure of physical activity in adults. "The purpose of the IPAQ is to provide a set of well-developed instruments that can be used internationally to obtain comparable estimates of physical activity. There are two versions of the questionnaire. The short version is suitable for use in national and regional surveillance systems and the long version provides more detailed information often required in research work or for evaluation purposes." Preliminary research in CF shows that it tends to underestimate physical activity in patients with lower energy expenditure activities and overestimate physical activity in patients with higher energy expenditure activities.

The Seven-Day Physical Activity Recall (7D-PAR) is an interviewer administered recall of a person's physical activity at various intensity levels over the previous 7 days. This measure is used frequently in adult studies and has demonstrated acceptable testretest reliability and validity when compared to electronic monitoring.

Preliminary studies in children have indicated adequate validity in relation to direct observation. Thus far only one reported abstract investigated the use of 7D-PAR in a CF population and found that the ICC between accelerometry and 7D-PAR was not strong enough to assess individual patients.

Portable physical activity monitors, specifically the SenseWear Pro 3 have been designed to improve the measurement of physical activity. This monitor integrates a bi-axial accelerometer, galvanic skin resistance, heat flux, skin and near body temperature to estimate the energy expenditure and step count. The armband is worn on the upper right arm. This system was evaluated in a population of adult CF patients and the diagnosis of CF did not seem to impact the system's ability to accurately estimate energy expenditure. However it did tend to underestimate energy expenditure with increasing exercise intensity (consistently reported in other studied populations).

Although the ideal method for measuring physical activity remains elusive, regular assessment by the clinician is key to the management of our CF patients.

Use of physical activity questionnaires has many practical values but should be limited to those that have been assessed for validity and reliability in the CF population. Pedometers and accelerometers decrease the subjectivity of questionnaires but have limited ability to assess cycling movements and upper limb movements. Doubly labelled water and direct observation appear to be the most objective measures but their applicability in the clinical setting is non-existent. Portable activity monitors do show promise but are not without limitations.

References It is clear that regular, intentional, physical activity (PA) should be part of a healthy lifestyle, including for patients with cystic fibrosis. There are both disease specific, e.g., improved airway clearance, improved lung function, better control of CF-related diabetes, etc., as well as more general, e.g., cardiovascular and musculoskeletal health, benefits of PA. It is equally clear that there is no "one-size-fits-all" target for PA. The following is intended to provide a general guide for PA recommendations followed by suggestions for children.

PA recommendations of any kind should be based on knowledge of the clinical status, including pulmonary function, and of the exercise response, including whether desaturation occurs and at what level of exercise.

PA targets can be attained using a combination of both formal, e.g., exercise prescription and leisure-time activity recommendations, and informal, but no less intentional, habitual or lifestyle PA recommendations. In any case the recommendation should be based on the patient's response to three critical questions: "What kinds of activities do you like to do?," "Who is going to be your partner?" and "What sorts of things keep you from being physically active?" Success in meeting PA targets will be improved if participants can select from a variety of favorite activities and participate in these activities on an "as desired" basis. Success is further enhanced by making a commitment with an activity partner. For children and young adults, promoting activity with best friends significantly increases time spent in moderate-to-vigorous activity while adults benefit from the accountability associated with a committed partner. Finally, PA is more easily integrated into a lifestyle with the identification of, and removal or minimization of barriers -excuses, in many cases.

Making PA a regular part of one's daily routine is difficult for most people, particularly if they are currently sedentary. It is therefore best to start the exercise recommendation with advice and encouragement as to how to become a regular exerciser. For some the first step is to find ways to incorporate activity into one's daily routine. These lifestyle changes might include changing one's commute or shopping habits to include walking or riding a bicycle for all or part of the distance or simply parking the car at the far side of the parking lot and walking, taking the stairs instead of the elevator, walking during work breaks, etc. An initial goal should be to accumulate 30 minutes of this kind of activity 5 days a week. The use of a pedometer can enhance motivation for lifestyle PA. Recognizing that less than 5,000 steps per day represents a sedentary lifestyle the initial goal should be to increase the daily step count to >10,000 per day over a period of time.

As the individual approaches the goal of 30 minutes, 5 days a week or 10,000 steps the recommendation should include advice on increasing exercise intensity to moderate to vigorous levels. Moderate exercise intensity should allow the person to maintain a conversation without becoming breathless while vigorous intensity would involve a sense of breathlessness. Once a person has become accustomed to an increase in habitual, lifestyle PA it becomes easier to move to leisure-time PA or more formal exercise recommendations.

The compendium of physical activities (1-3) provides a listing of hundreds of activities and their energy cost. Energy cost is listed in terms of METS (Metabolic Equivalents, with 1 MET = 3 mL O 2 /kg/min) allowing one to recommend a variety of specific PA at moderate (3-6 METs) or vigorous (>6 METs) activity intensities. For instance, encouraging activities such as general carpentry or sweeping floors or bicycling with light effort would fulfill a recommendation for moderate activity while encouraging jogging at 6 mph, shoveling snow or higher intensity bicycling would fulfill a recommendation for vigorous activity. In this way a variety of activities can be combined to attain the goal of a minimum of 30 minutes of moderate-vigorous activity 5 days a week.

Some patients will require a more formal exercise prescription to be carried out in an inpatient or outpatient rehabilitation setting; some may find that specific exercise prescriptions make it easier to attain the above-described goals outside of a rehabilitation setting. This more specific prescription approach also allows for safe recommendations for strength training activities. The prescription for aerobic exercise will use heart rate as the marker for exercise intensity and a target of >60% of Heart Rate Reserve calculated from a peak/maximal exercise test (HRR = (Peak HR -Rest HR)% intensity + Rest HR).

The most effective way to involve children in regular activities that will meet the recommendation is to engage family and friends in an active lifestyle, including leisure activities and sports. Children can be effectively disengaged from sedentary activities by providing rewards for reducing those activities, allowing them to select their own alternatives. Similarly, connecting children with active friends is effective in increasing the amount of time spent in moderate to vigorous activity on a daily basis.

References Staphylococcus aureus is one of the most significant causes of human bacterial infection, contributing to morbidity and mortality in individuals of all ages (1). The virulence properties of this microorganism facilitate infection of a wide array of tissues and organs, including the human lung. Historically, S. aureus pneumonia has been a scourge of the critical care environment, not often documented as a cause of community acquired disease. However, an alarming trend has been noted for over a decade, as an increase in the incidence of severe S. aureus pneumonia in otherwise healthy adults and children has been reported. Recent studies suggest that approximately 50,000 cases of S. aureus pneumonia occur in the U.S. alone each year, with more than half caused by methicillin-resistant S. aureus (MRSA) (2) . Increasing antimicrobial resistance among S. aureus strains and the appearance of particularly virulent isolates of this pathogen within the community have rendered conventional antimicrobial therapies obsolete, resulting in increased mortality and cost of care (3). Concomitant with these changes, individuals with cystic fibrosis have experienced an increasing burden of drug-resistant S. aureus disease (4). Novel therapeutic approaches to both prevent and treat invasive pulmonary infection with S. aureus, and especially MRSA, are necessary. As S. aureus is a key colonizing organism and pathogen in both the normal and CF host, understanding the principles that govern the interaction of the host with the pathogen in both patient populations is likely to provide keen insights that will enhance our ability to combat S. aureus lung disease.

Studies of disease pathogenesis. As a first step in understanding the pathogenesis of S. aureus pneumonia, animal models have been used to define the pathogenencoded virulence factors that facilitate lung infection. Initial work in the field led to the observation that staphylococcal Protein A, a cell-wall associated protein, contributes to lung disease by virtue of its ability to interact with the tumor-necrosis factor-alpha receptor, TNFR1, and stimulate proinflammatory responses (5) . Disruption of the interaction of Protein A with TNFR1, either by genetic deletion of Protein A in engineered S. aureus, or in TNFR1 knockout mice, mitigates disease (5). Subsequently, studies in an adult, immunocompetent mouse model of S. aureus pneumonia have revealed a profound defect in virulence when mice were infected with strains engineered to lack expression of α-hemolysin (Hla), a pore-forming exotoxin that is produced by almost all strains of S. aureus (6) . Further investigation of αhemolysin in S. aureus pneumonia revealed the central role of this single toxin in the modulation of virulence, correlating α-hemolysin expression with mortality in animal studies and direct injury to the pulmonary alveolar epithelium (7) . As a toxin that forms injurious pores in the eukaryotic cell membrane, the mechanism of αhemolysin-induced injury to the lung appears conceptually simple -the toxin can induce irreversible damage to the plasma membrane, culminating in cell death and organ dysfunction. The recent identification of A Disintegrin and Metalloprotease 10 (ADAM10) as the cellular receptor for α-hemolysin is anticipated to yield further insights into the mechanism of toxin action, and may also permit a clearer understanding of host susceptibility to disease (8) . ADAM10 plays an essential role in the maintenance and turnover of the epithelium (9) . The use of this protein as a toxin receptor raises the interesting possibility that the toxin may not only bind to the receptor to gain access to the host cell, but may take advantage of the native function of ADAM10 to specifically injure the epithelium of the lung.

In addition to studies on S. aureus virulence factors such as Protein A and α-hemolysin, the nature of the host immunologic response to the pathogen plays a critical role in establishing disease severity. Depletion of alveolar macrophages leads to enhanced mortality from experimental MRSA pneumonia, however does not impact on the bacterial load in the lung tissue. In contrast, depletion of host dendritic cells was associated with an increased bacterial load in the lung, but no alterations in mortality (10). The differential requirements for specific immune cell populations in governing distinct facets of disease pathogenesis further underscore the complexity of the host-pathogen interaction in S. aureus pneumonia.

Implications for prevention and therapy. Both active immunization with a non-toxigenic variant of αhemolysin and passive immunization with rabbit anti-αhemolysin immune sera protected animals from lethal S. aureus pneumonia (7). Targeting the toxin in this fashion may therefore be a highly effective strategy to limit S. aureus lung disease. In addition, it is reasonable to propose that strategies that interfere with the ability of the toxin to bind to its host receptor, ADAM10, may also be of clinical value. Host-targeted approaches to the prevention or amelioration of disease are also essential to consider, including strategies to manipulate the host cytokine and chemokine response to infection and minimize intrapulmonary inflammation (10) . At present, the field remains limited by a lack of knowledge of the hostpathogen interaction in human disease. While experimental model systems have provided keen insight into the complexities of this interaction, this knowledge will need to be extended to the human population through well-designed clinical studies aimed at identifying factors that contribute to host susceptibility as well as evaluate the contribution of staphylococcal virulence factors to disease. Moreover, it will be essential to understand the host-pathogen interaction in the setting of underlying CF as the nature of this interaction may be fundamentally different, demanding distinct approaches to the prevention and treatment of S. aureus lung disease.

Lisa Saiman, M.D., M.P.H. Pediatrics, Columbia University, New York, NY, USA Introduction. The molecular epidemiology of bacterial pathogens can be integrated with social networking to increase our understanding of infectious disease transmission in community-based settings (1). This strategy has been used as a public health measure to study transmission of Staphylococcus aureus (2, 3) in the community, including closed communities such as prisons (4) or among intravenous drug users (5) . Integrating molecular epidemiology with social networking strategies could potentially be used to better define transmission of S. aureus in CF and has implications for prevention.

Molecular typing strategies. Several molecular typing strategies may be used to distinguish S. aureus strains. These include pulse-field gel electrophoresis (PFGE), multi-locus sequence typing (MLST), S. aureus protein A typing (spa typing), and potentially microarray technology. Strategies used for community settings must be able to successfully perform, store, and compare huge amounts of sequencing data required to define unique and shared S. aureus strains.

Social networking strategies. Community-based and population settings are constantly changing making a total understanding of all factors involved in disease transmission impossible. Nonetheless, social network methodologies use group dynamics including social contacts, environments, and risk behaviors to provide insights into disease transmission. Three social network levels of how individuals are connected to others can be studied: the dyad, the personal network, and the community. For the dyad and personal network, direct relationships are evaluated while for the community network, both direct and indirect relationships are evaluated. There are obviously limitations to social networking studies. Such studies cannot readily observe risk behaviors, particularly those that may be deliberately hidden; links between people can be difficult to assess over long time intervals or large distances; and non-responders can have significant, but difficult to predict, influences on data.

Integrating molecular epidemiology and social network strategies. Investigators emphasize that bacterial molecular relatedness does not describe transmission dynamics without concurrent interview data that provides supporting epidemiology. Investigators further emphasize that even partial interview data can reveal links that may be brief, anonymous, or through a central individual.

Social networking techniques potentially can elucidate infectious disease transmission dynamics including the settings where transmission could have occurred which is central to developing preventive strategies. For instance, integration of data from DNA fingerprinting with social network analysis allowed identification of infectious TB cases and the sites of transmission (6) . A similar strategy was used to explore transmission of gonorrhea (7) . Another salient example of this integration occurred in a study of S. aureus transmission among inhalational drug users (2) .

Relevance to CF. In CF, acquisition of both methicillin-susceptible S. aureus (MSSA) and methicillinresistant S. aureus (MRSA) is most likely different than acquisition of other potential pathogens. S. aureus, including MRSA and MSSA, can be colonizing flora of the skin (e.g., axilla), anterior nares, throat, and anogenital tract. Unlike other CF pathogens, person-to-person transmission of this microorganism can occur among family members and close contacts with and without CF (8) . Several molecular epidemiology studies of MRSA in CF have demonstrated acquisition of so-called community acquired strains (8) (9) (10) .

Promoting an increased understanding of the role of social networking in CF could increase the effectiveness of prevention efforts and sustain eradication efforts of MRSA. Integration of molecular typing and social networking is feasible in CF due to the interdisciplinary infrastructure that exists in CF care centers throughout the world.

Future Directions. Integration of molecular typing and social networking strategies within existing CF research on MRSA could facilitate our understanding of transmission and prevention. Future studies could investigate the virulence determinants that are associated with colonization and transmission of MRSA The emergency of small colony variants (SCV) of MRSA after MRSA treatment may further increase challenges in CF treatment.

There are many antibiotics options for MRSA treatment. Vancomycin, a glycopeptide, has been the traditional antimicrobial agent of choice for MRSA treatment. Many have argued that traditional dosing, 15 mg/kg every twelve hours, is inadequate, especially for severe infections such as bacteremia and pneumonia. Observational studies that found high vancomycin MICs (≥ 2 mcg/mL) are independently associated with poorer outcomes, specifically lower treatment response and higher infection-related mortality. Many recommend that serious MRSA infections require dosages that exceed traditional recommendations. Recent guidelines by the ATS/IDSA Committees recommend targeting higher vancomycin troughs of 15-20 mcg/mL in patients with health-care related MRSA pneumonia, rather than the lower targets (of 10 or less) recommended in the past.

The recent emergence of glycopeptide-intermediate and vancomycin-resistant S. aureus (GISA and VRSA) strains is very concerning. GISA strains have been reported to emerge in patients with CF, typically after prolonged exposure to vancomycin. These strains may be missed by traditional microbiologic testing. There are very limited data for newer glycopeptide, telavancin, which is approved only for the treatment of skin and skin structure infections (SSTIs).

Several older antimicrobial agents have been used increasingly for the treatment of MRSA. However, data on efficacy is very limited. TMP-SMX is active in vitro against many MRSA strains. The largest published trial on the use of TMP-SMX for S. aureus found that TMP-SMX demonstrated a lower clinical cure rate for S. aureus infection compared to vancomycin (85% vs. 98%). Many experts consider it a viable treatment option for SSTIs, but data on pulmonary infections are very limited.

Clindamycin, a lincosamide, has been used successfully to treat MRSA infections, however, antimicrobial resistance is greater than 10% in some areas. Some MRSA isolates develop resistance when exposed to lincosamides and macrolides, such as erythromycin. This inducible resistance can be detected via the D-test. Most experts argue that confirming a negative D-test is critical when considering clindamycin treatment of a serious S. aureus infection. The ability of clindamycin to inhibit pvl expression in vitro is a theoretic advantage of using clindamycin for treatment, although the clinical benefit of this inhibition unclear.

The tetracyclines doxycycline and minocycline are active against MRSA, although their use in younger children is contraindicated. These antibiotics have successfully been used to treat MRSA infections in small case series and clinical trials, although there are extremely limited data for patients with CF. Tigecyline, a new and related drug lacks an oral formulation and there are no compelling data that this newer drug provides advantages over older tetracyclines. A newer generation cephalosporin, ceftaroline, is the only commercially available beta-lactam with activity against MRSA. It is approved for the treatment of communityacquired pneumonia, but there are no data on CF treatment.

Linezolid, an oxazildanone antimicrobial, comes in both oral and intravenous formulation, which has made it useful in patients with CF. There are retrospective data suggesting that in treatment of ventilator associated pneumonia it may have greater efficacy and lower mortality compared with vancomycin. However, its pharmacokinetics in adults and children with CF is controver-sial; linezolid dosing may need to be increased to 3 times per day in these patients. Additionally, linezolid-resistant MRSA have been reported in patients with CF. Furthermore, bone marrow and neurologic toxicity with prolonged linezolid treatment can be problematic.

Daptomycin, a lipopeptide with bactericidal activity against S. aureus, should not be used in the treatment of pneumonia as it is inactivated by pulmonary surfactant. There are data rifampin may provide additional benefit to standard therapy in the treatment of S. aureus, but data are inconsistent and rifampin is associated with many drug interactions. Rifampin therapy is limited by the appearance of rifampin resistance during therapy, especially when used as monotherapy.

Data on optimal MRSA decolonization and eradication are not clear and continue to emerge. Eradication of colonization of other household members may be important for eliminating colonization in the MRSA colonized patient. In CF patients, S. aureus biofilms in the upper airways (e.g., sinuses) and lower airways present challenges for eradication efforts. In non-CF patients, there are data suggesting that mupirocin in combination with chlorhexidine may eliminate short-term colonization and reduce skin infection with MRSA, but these regimens do little to prevent airway colonization. A recent investigation in CF patients using aggressive regimens for MRSA eradication demonstrated that regimens that typically combined 2 oral antibiotics plus nebulized vancomycin plus topical skin and nasal disinfection may be effective at eradicating MRSA colonization, with eradication rates of approximately 80%. Some patients required multiple courses of therapy to achieve eradication. The clinical benefit of eradication is incompletely understood.

References These studies have been limited however by small study populations, lack of control groups, single-center retrospective design, variable follow-up, and failure to distinguish incident vs persistent MRSA infection (4) (5) (6) (7) . Doe reported the largest experience to date with a retrospective review of 37 patients (7) . Although many different eradication regimens were used, most of the patients in their cohort were treated with a combination of two oral antibiotics (rifampicin, fusidic acid or trimethoprim) and nebulized vancomycin. They reported eradication of MRSA in 81% of the participants at 6 months. There was no distinction made between those with incident MRSA and those with persistent MRSA infection in the study analysis, although they report that approximately 38% of those included had had multiple positive MRSA cultures. Garske focused on treatment of CF adults with persistent MRSA (4). The study enrolled seven CF adults with persistent MRSA (average FEV1 36% predicted, all with chronic PA, 6/7 with previous IV MRSA therapy); 5 of the 7 patients (71%) were culture-negative six months after completing a six month treatment regimen of oral fusidic acid and rifampin (4) . Most other studies have focused on patients with incident MRSA infection rather than persistent infection. Macfarlane reported on 17 CF patients with incident MRSA treated with oral and IV antibiotics. Utilizing a protocol which started with oral rifampicin and fusidic acid, they achieved a 94% eradication rate at 12 months (5). A study by Solis of 15 CF children treated with five days of oral and nebulized vancomycin was associated with a 55% eradication rate (6) .

The key questions to be answered in future CF MRSA research can be divided into three main areas: 3) The effect of MRSA molecular characteristics and patient characteristics on treatment response and patient outcomes.

a. Are there MRSA molecular epidemiology characteristics which aid in predicting the persistence and severity of MRSA infection in CF?

b. Are there specific patient characteristics which aid in predicting the persistence and severity of MRSA infection in CF?

c. What role does the environment play in determining the frequency and severity of MRSA infections in CF?

Clinical Trial Designs for MRSA in CF. There are currently studies in progress or about to be commenced which address each of the key areas described above. These include: 1) STAR-CF too Trial. (Staph Aureus Resistance in CF, Treat or Observe?). This is a multi-center trial (UNC and TDN coordinating center) with 6 months follow-up designed to assess if an early eradication protocol is effective for eradication of MRSA (and will provide opportunity to obtain data regarding spontaneous disappearance versus persistence of MRSA). Patients are assigned to either an observational arm or an interventional arm treated with two oral antibiotics, topical mupirocin, and environmental decontamination. Eighty patients with new respiratory tract MRSA infection will be enrolled and randomized 1:1. Primary outcome measure is proportion of subjects in each arm with MRSA negative respiratory cultures at day 28. Secondary outcome measures include proportion of subjects treated with oral, inhaled and IV antibiotics over the 6 month study and number of days of use, and proportion of subjects with a protocol defined pulmonary exacerbation.

2) PMEP Trial. (Persistent Methicillin-resistant Staph Aureus Eradication Protocol) -This is a 6 month 2-center trial (Case and Johns Hopkins) in individuals ≥ age 12 designed to assess the effect of an aggressive 28-day inhaled and oral antibiotic combination protocol for CF individuals with persistent MRSA infection. Participants must have two positive MRSA respiratory cultures in the last 2 years at least 6 months apart and a positive MRSA respiratory culture at screening. Study design is a double-blind, comparatorcontrolled, parallel-group study with 1:1 assignment to either vancomycin for inhalation (250mg BID) or taste matched placebo. In addition, both groups will receive oral rifampin, a second oral antibiotic (TMP-SMX or doxycycline), mupirocin intranasal cream and chlorhexidine body washes. Forty patients with persistent respiratory tract MRSA infection will be enrolled. Primary outcome measure is percentage of patients MRSA free by induced sputum respiratory tract culture one month after completion of 4-week eradication protocol in intervention arm vs control arm. Secondary outcome measures include proportion of patients MRSA free at 3 and 6 months after completion of treatment protocol, change in FEV 1 from baseline on days 28 and 118, time to first exacerbation, and change in MRSA CFUs.

3) Characterization of MRSA infection/colonization in CF. While not a clinical trial, this is the largest ongoing study that seeks to define the molecular epidemiology characteristics of MRSA infections in CF and if molecular sub-types of MRSA differ in effect on exacerbations, lung function and nutrition. This is a multicenter study (coordinating center UNC) aiming to enroll 450 patients ages 1-18. Outcomes of interest: Characterize clinical impact of SCCmec MRSA sub-types on exacerbations and decline of FEV 1 and nutrition using prospective and longitudinal study designs; Evaluate the best antibiotics to be used based on MRSA susceptibility obtained from different U.S. CF centers; Measure colo-nization rate in nose and skin in patients with known MRSA and in their main caregiver in a sub-group of CF patients with recent acquisition.

References: 1. 

Michael R. Knowles, M.D.

Cystic fibrosis (CF) is a recessive "monogenic" disorder caused by mutations in CFTR. There is a wide range of lung disease severity in CF, even for patients who are homozygous F508del. Studies of twins and sibs have determined that non-CFTR genetic factors account for >50% of variability in lung disease (1) . Identification of non-CFTR gene modifiers is particularly relevant in CF, as almost all patients are now being diagnosed through neonatal screening programs. To intensively study gene modifiers, a North American CF Gene Modifier Consortium was formed by investigators at UNC/CWRU, Johns Hopkins, and Toronto, Canada. The Consortium developed a quantitative and standardized lung disease phenotype, based on multiple measures of FEV 1 over 3 years, adjusted for survival (2) . A GWAS (Illumina 610 Quad) was performed in 3,467 CF patients, and two loci with genome-wide significance were discovered (3). One locus was identified at chr 20q13.2 by linkage analysis of the Twin and Sibs (Hopkins) cohort (LOD score of 5.03), and SNPs under the linkage peak replicated in UNC/CWRU and Canadian association population patients. A robust association at chr 11p13 was discovered (p=3.3 x 10 -8 ) in F508del homozygotes in UNC/CWRU and Canadian patients, and replicated in the family-based (Hopkins) population (joint analysis of three studies, p=1.5x10 -9 ). The significant SNPs at chr 11p13 are located in an intergenic region, 3' to EHF (an epithelial-specific transcription factor) and APIP (an inhibitor of apoptosis) (4) (5) (6) (7) . EHF is expressed in airway epithelia, and may serve as a regulator of differentiation under conditions of stress, and/or play a role in inflammation (8, 9) . APIP inhibits apoptosis, and cis expression QTL (eQTL) patterns are reported for lymphocytes. The direction of the APIP eQTLs and "risk" SNPs in the GWAS suggest increased expression of APIP, and inhibition of apoptosis, is associated with worse CF lung disease. This is consistent with the concepts that delayed neutrophil clearance caused by reduced apoptosis can cause inflammation in the CF airways, and inhibition of apoptosis can contribute to mucus cell metaplasia (10, 11) . The intergenic region at chr 11p13 has predicted regulatory features, including open chromatin domains. Ongoing work will test the hypothesis that genomic variation in this region alters expression of nearby genes, which has biologic relevance to CF airways disease. The study of genes that modify severity of "monogenetic" diseases, such as CF, is progressing rapidly, and mechanistic insights are beginning to emerge. The extension of the Consortium to trans-Atlantic (France) sites will provide additional power to identify new loci for studies to link genetic variation to mechanisms of lung disease severity.

References Meconium ileus (MI) presents in 15% of CF patients, and is a highly heritable (H >0.88) and early complication. With the resources of the North American CF Gene Modifier Consortium (1), we sought to identify genes that contribute to MI susceptibility. Using the traditional genome-wide association study (GWAS) approach, two highly significant genome regions were identified with five SNPs with p values < 5 x 10 -8 . The regions correspond to the SLC6A14 and SLC26A9 genes, and their SNP associations to MI have been replicated in additional North American and French CF patient sets. However, they explain only 5% of the MI phenotypic variability. To identify additional modifier genes that contribute to MI, we developed the hypothesis-driven GWAS (GWAS-HD) to reprioritize genotyping results considering current molecular understanding of CF. A major aspect of CF pathophysiology is impaired fluid and electrolyte flux at the epithelial interfaces of a number of CF-affected organs including the airway and intestine. The single cell epithelial layer is unique in that it forms a highly selective and tight barrier between body organ and surface or ductal interfaces. Epithelial function is achieved by cell polarization whereby many determinants and reg-ulators of fluid, solute and ion transport, including the CFTR ion channel reside specifically at the apical membrane. With these considerations, we proposed our "apical hypothesis" and used data from the Gene Ontology Consortium (2) to derive two lists of genes to test for relationship to MI susceptibility, one with gene products that localize to the apical membrane (155 genes), the second, with gene products that localize to the nuclear envelope (224 genes). The nuclear gene group showed no association to MI as anticipated, however, the "apical gene" group showed significant association, p-value = 0.0002. Together, the apical genes account for 17% of phenotypic variability in MI. Efforts to understand the biology of these contributing apical genes, and their relation to MI susceptibility and other CF phenotypes will be discussed.

References Diabetes develops frequently in people with cystic fibrosis over time, occurring in 25-50% of adolescents and adults with CF. People who fall into this subset of CF patients tend to have worse lung function, poorer nutritional status, and increased mortality. Treatment of diabetes (a.k.a. CF-related diabetes or CFRD) has been shown to correlate with improvement in lung function and body mass index.

CFRD develops as the capacity of the pancreatic beta cells to secrete insulin declines over time. In general, insulin resistance is not thought to play a substantial role, although CF patients can exhibit transient and significant drops in insulin sensitivity when ill or having a pulmonary exacerbation. At first, hyperglycemia may be transient, occurring only when ill or taking systemic glucocorticoid medications. Over time, insulin secretion diminishes, and daily exogenous insulin may become necessary. There is some evidence that insulin may benefit CF patients prior to meeting diagnostic criteria for CFRD.

The mechanisms underlying the development of diabetes in some CF patients are not well known. Diabetes occurs almost exclusively in CF patients with severe exocrine pancreatic insufficiency, suggesting that pancreatic autodigestion of islet cells may play a role. In support of this hypothesis, CF patients have demonstrated decreased glucagon and pancreatic polypeptide (made in the alpha and delta cells of the islets)(1). However, other lines of evidence suggest other factors are involved. First, the rate of CFRD varies widely even among CF patients with the same severe CFTR mutations (e.g., homozygous F508del). Second, epidemiologic data suggest that other factors (CF liver disease, transplant medications, ABPA) play a role, at least in a subset of people with CFRD (2) . Third, studies of pancreatic tissue from diabetic and non-diabetic CF patients failed to detect correlation in CFRD status with islet cell mass or number. In general, pancreatic islets are preserved as the exocrine tissue becomes fibrotic. Interestingly, these studies found that a specific marker of CFRD was the presence of islet amyloid, a marker for beta cell dysfunction also specific to type 2 diabetes (T2D)(3). Finally, knock-out mouse studies have suggested there may be an intrinsic defect in CFTR-deficient pancreatic beta cells (4) .

The identification of islet amyloid in CFRD islets suggested similarity to type 2 diabetes (T2D). T2D is caused at least in part by genes (40-60% heritability) and bears some phenotypic and genetic similarity to CFRD. Patients who develop T2D have decreased sensitivity to insulin in the target tissues (e.g., skeletal muscle, liver, and fat cells). Initially, beta cell proliferation increases the insulin secretory capacity. Over time, however, beta cell function diminishes, resulting in hyperglycemia, and eventually, treatment with insulin. At diagnosis, patients with T2D may have lost 80% of insulin secretory capacity.

In the CF twin and sibling study, having a strong family history of T2D was correlated with a 3-fold increased risk of CFRD (5), suggesting that there may be genetic risk factors shared between T2D and CFRD. The degree to which variation in genes beyond CFTR (a.k.a. modifier genes, or genetic modifiers) affects CFRD, was tested in the CF Twin and Sibling study. Two phenotypes were considered: presence (vs. absence) of CFRD, and onset of CFRD by Cox regression (referred to as CFRDage). Comparison of concordance rates for CFRD and CFRDage between identical twins, non-identical twins and siblings with CF, showed that a high proportion (40-100%) of the variation in CFRD risk was due to genetic modifiers (6) .

Common variants in ~40 genetic loci have been associated with risk of T2D and with other glucose-related metabolic traits in people without CF. Initially using data from the CF Twin and Sibling study (TSS; Johns Hopkins University) and the Gene Modifiers Study (GMS; UNC Chapel Hill/Case Western), we found that a susceptibility gene for T2D, TCF7L2, also confers risk for CFRD (5) . Those studies have since been replicated in a set of CF patients in the Canadian Consortium for CF genetic studies (CGS; University of Toronto), strengthening the overall evidence for this gene's involvement in CFRD (meta-analysis using the CFRDage trait, n = 4483 patients, rs7903146 SNP, hazard ratio = 1.34 per allele, P = 4.4×10 -7 ).

In addition, variants (single nucleotide polymorphisms or SNPs) in 6 additional risk loci for T2D were also tested for association with CFRD onset (CFRDage). These loci were selected based on effect size in T2D and estimated power for detection using ~4483 CF patients. Study-wide significant association with CFRD was found for SNPs in CDKAL1 (rs7754840, P=1.7×10 -3 ; rs7756992, P=2.1×10 -5 ) and CDKN2A/B (rs1412829, P=1.2×10 -5 ). As was the case with the TCF7L2 SNP, the high-risk allele for T2D was the same as the high-risk allele for CFRD. The 4 remaining loci did not show statistically significant evidence for association with CFRD, (study-wide P<0.1 in IGF2BP2; P>0.1 in KCNJ11, HHEX, SLC30A8). Relatively weak associations with CFRD could have been missed due to limitations in study power. Susceptibility genes affecting both T2D and CFRD suggest that some pathways of disease pathogenesis are shared.

Using the families in the Twin and Sibling Study, we are also conducting a genome-wide linkage study for onset of CFRD. Linkage studies can be sensitive to both common and rare variants which affect the risk of disease. A genome-wide linkage map of 20,066 SNPs was typed for the TSS and met all quality-control measures. Nonparametric linkage (variance components) methods were used to detect linkage to the onset of CFRD using data from 122 phenotyped sibling pairs with CF and severe exocrine pancreatic insufficiency. The quantitative trait used for linkage was the Martingale residual for CFRD onset after Cox regression (CFRDresidual). This CFRD phenotype has identical heritability (~1.0) and correlation with pertinent covariates (e.g. sex, lung function, or T2D SNPs) as CFRDage. Empiric LOD scores were calculated by simulation. Genome-wide significant evidence for linkage was found at chromosome 2q35-2q36.1 (LOD = 3.8, approximately equivalent to studywide P=0.035). Genome-wide suggestive evidence for linkage was found at chromosome 12p12 (LOD = 2.24). These studies identify a region of chromosome 2 which may harbor novel genetic modifiers of CFRD.

We are also conducting a genome-wide association study for CFRD onset using data from the International CF Gene Modifier Consortium. This type of study will be sensitive to common variants (>5% allele frequency in the population) of moderate effect size, similar to those already identified based on overlap with T2D susceptibility loci. We hypothesize that, in addition to loci shared with T2D which act on common diabetes pathways, there may be loci unique to CFRD which act on CFTR-specific or CFRD-specific pathways. Modifier loci in either category will present opportunities for future studies of diabetes pathogenesis and development of new therapies, both in CFRD and in T2D. Also, identification of a genetically at-risk subset of CF patients may have clinically relevant implications for CFRD screening strategies.

These studies show that the risk of CFRD is heavily influenced by genes other than CFTR. Three genetic modifiers of CFRD have been identified which are also susceptibility genes for type 2 diabetes, indicating that some pathways are shared between these two forms of diabetes. In general, these variants have associated with decreased insulin production and beta cell function in studies of non-CF patients, suggesting that beta cell function may be important in CFRD pathogenesis. Studies of CF patients, who are sensitized to develop diabetes at a young age, can inform on the pathophysiology of other forms of diabetes such as type 2 diabetes. Novel genetic modifiers of CFRD, such as may be found at chromosome 2q35-2q36. Advances in genotyping technology have made it technically and financially feasible to carry out highdensity genotyping of polymorphisms simultaneously. Such advances provide the possibility to detect much more quickly than ever before genomic variants that influence, or cause, a phenotype. While these technolo-gies make it possible to identify disease-causing locations in the genome, narrowing the list of possible genes involved from approximately 25,000 to only a handful, the identity of the genes involved and how they impart their effects may not be immediately apparent. Functional assessment is critical and thus cellular and animal models are needed to confirm any suspected gene's effects.

For these types of studies, the mouse is ideally suited as it is possible to manipulate any suspected gene and place it in a CF context. In fact, there are hundreds of lines of mice with mutations in specific genes and hundreds of stem cells for the creation of additional lines, if desired. Thus, we have begun surveying putative CFmodifying genes in mice by placing mutant forms of the modifiers on CF backgrounds.

One of the prominent features of all CF animal models (mouse, ferret and pig) is the propensity for intestinal obstruction in the distal ileum. Thus, modifiers of obstruction would be ideal for testing this concept. Blackman and colleagues used a family-based approach (linkage analysis) and found that several regions of the genome, including one on chromosome 8, significantly associate with meconium ileus in CF patients (1) . Further analysis of the chromosome 8 region suggested that the gene involved is MSRA (Henderson, unpublished) , encoding methionine sulfoxide reductase A, an enzyme that catalyzes the reduction of methionine sulfoxide to methionine. We introduced a null allele of MsrA into a CF line (Cftrtm1Unc) and compared the incidence of intestinal obstruction in CF mice with the three different MsrA genotypes. At 40 days, only 17% of CF mice (n=30) with wildtype MsrA survived, while 42% of those herterozygous for the MsrA mutant allele (n=33) and 61% homozygous for the mutant allele (n=46) survived. These results confirm the modifying effects of MSRA, but an enzyme that alters the state of methionine as a modifier opens a new avenue in our thinking about the pathophysiology of gastrointestinal obstruction in CF.

We have more recently reported a high density genotyping survey identifying two regions of the genome, one on chromosome 11 and one on 20, that associate significantly with CF lung disease (2) . Neither of these regions neatly identifies a single gene and therefore we are in the process of examining several of the genes in those regions. To date, the most proximate genes to the chromosome 20 region are MC3R (melanocortin 3 receptor) and CBLN4 (cerebellin-like 4). We have obtained mice with mutant alleles of each of these genes and are currently crossing them with various CF lines to generate CF mice with different levels of modifier gene function. These mice are then put through a panel of assays to determine if altering putative modifier gene function has discernible effects on any CF-related phenotype.

References The fatal lung disease cystic fibrosis (CF) is a loss of protein function disorder caused by misfolding and premature degradation of the cystic fibrosis transmembrane conductance regulator (CFTR). CFTR is a Clchannel that controls hydration of epithelial cell surfaces in airways and glands (1). Most CF patients inherit the CFTR∆F508 mutant allele whose protein product exhibits subtle folding defects that lead almost all nascent forms to be degraded (2) . Patients that exhibit partial CFTR function have mild CF symptoms, so restoration of CFTR∆F508 activity to modest levels is a therapeutic goal.

F508 is located in NBD1 and is not essential for Clconductance (1), but its deletion leads pools of nascent CFTR∆F508 to accumulate in a foldable, but kinetically trapped conformation (3) . CFTR∆F508 misfolding appears to involve subtle defects in NBD1 folding that cause co-translational misassembly of an intermediate that is degraded rapidly and may not accumulate (4) .

Folding of the small pool of nascent CFTR∆F508 that is spared initial degradation is arrested because of defective contact formation between NBD1 and regions that include intracellular loops exposed by MSD2 and misfolding of NBD2. Consequently ~ 99% of CFTR∆F508 is degraded prematurely by the ubiquitin-proteasome system. CFTR folding is assisted by several different molecular chaperones. The ER-associated Hsp40 Hdj-2 (DNAJA1) helps attract cytosolic Hsc70 to the ER membrane surface to facilitate co-translational folding and assembly of NBD1 (5). The ER luminal chaperone calnexin appears to act after Hdj-2 to facilitate association of regions within CFTRs membrane-spanning and cytosolic domains (6) . Terminal steps in folding of full-length CFTR are facilitated by Hsp90 and its associated co-factors (7) .

Paradoxically, the selection of nascent forms of CFTR and CFTR∆F508 for proteasomal degradation is also facilitated by molecular chaperones. The cytosolic E3 ubiquitin ligase CHIP interacts with Hsc70 and/or Hsp70 to form a quality control machine that utilizes the polypeptide binding activity of Hsc/Hsp70 to target misfolded CFTR for proteasomal degradation (8) . In addition, the ER-associated E3 RMA1/RNF5 acts in association with Derlin-1 and the E2 Ubc6e to ubiquitinate CFTR (4). How selection of CFTR for degradation by the RMA1 E3 machinery and CHIP E3 complex are synergized is not entirely clear. However, the RMA1 E3 complex may act coincident with translation to recognize folding defects in CFTR that involve the misfolding and defective assembly of NBD1 into a complex with the Rdomain (4, 6) . In contrast, the CHIP E3 may act posttranslationally to recognize misfolded regions of CFTR that include NBD2 (4).

Deletion of F508 renders CFTR highly sensitive to changes in RMA1 activity, which suggests that the RMA1 E3 plays a critical role in selection of CFTR∆F508 for premature degradation (4) . Yet, it is unclear how the RMA1 complex distinguishes between on-and off-pathway forms of nascent CFTR. RMA1 interacts with the transmembrane quality control factor Derlin-1 and siRNA knockdown of Derlin-1 enhances cell surface expression of CFTR (4). Derlin-1 can form complexes with MSD1 of CFTR and Derlin-1 overexpression promotes ER retention and proteasomal degradation of CFTR (4). So, it is possible that Derlin-1 is part of a complex that acts as a membrane chaperone to scan the assembly status of CFTR's membrane regions and targets misassembled forms to RMA1 for ubiquitination. This seems logical, except the region in CFTR that is ubiquitinated by RMA1 is cytosolic. Therefore, quality control factors in addition to Derlin-1 may assist RMA1 in the selection of misfolded CFTR for ubiquitination. Clues to the identity of such a factor come from the analysis of the Hsp40/DnaJ family of Hsc70 co-chaperones. Hsp40 proteins utilize a conserved J-domain to regulate Hsc70 ATPase activity and specify client proteins of Hsc70. One mechanism for specification of Hsc70 function is for a specialized Hsp40 to attract Hsc70 to function in a discrete cellular location. Studies in yeast identify a unique Hsp40 sub-type that is integrated into the ER membrane and exposes its J-domain to the cytosol (9). This sub-family is represented in yeast by HLJ1 (high copy lethal DnaJ 1) and deletion analysis suggests that HLJ1 functions with cytosolic Hsc70 in ER quality control (10). RMA1 is not expressed in yeast, so we pondered whether or not an HLJ1-like protein might act with RMA1 in human cells to assist in the selection of CFTR for degradation. Indeed, HLJ1-like Hsp40s are conserved in the human genome and this talk will describe the characterization of human DNAJB12 (JB12). JB12 is a Type II Hsp40 that contains a J-domain and G/F-like region and is similar to HLJ1 in that it is localized to the ER membrane and exposes its J-domain to the cytosol. Yet, JB12 differs from HLJ1 in that it contains N-and C-terminal extensions that were added during the course of evolution. Data presented will demonstrate that JB12 acts with Hsc70 and RMA1 to facilitate proteasomal degradation of CFTR and CFTR∆F508. JB12 forms complexes that contain RMA1 and Derlin-1 and increasing JB12 levels dramatically increases association of Hsc70 with ER localized forms of CFTR and RMA1. Depletion of endogenous JB12 results in a significant 3-fold increase in the folding efficiency of CFTR and permits a pool of CFTR∆F508 to escape the ER. JB12 appears to direct Hsc70 to function with the E3 RMA1 in degradation of nascent CFTR∆F508 and CFTR. Since fluctuations in its activity impact on the fate of nascent CFTR, JB12 is capable of exerting control over the folding efficiency of nascent CFTR and CFTR∆F508 and JB12. Inactivation of JB12 increases the potency of small molecule folding correctors and JB12 appears to be a therapeutic target for treatment of CF 

Nadia A. Ameen, MBBS Pediatrics, Yale University School of Medicine, New Haven, CT, USA Cystic fibrosis transmembrane conductance regulator (CFTR) chloride channels are in the apical domain of both crypt and villus enterocytes where they play a critical role in anion and fluid secretion. Genetic mutations that lead to absence of functional CFTR on the apical plasma membrane of enterocytes result in intestinal obstruction and cystic fibrosis. On the other hand, enterotoxins elaborated by Vibrio cholerae and Escherichia coli utilize CFTR as their apical exit pathway to elicit massive fluid secretion and secretory diarrhea following cAMP and cGMP signaling (1, 2) . In native epithelium and cultured polarized intestinal cells, both the number of CFTR channels and anion transport on the cell surface are regulated by protein kinase A, protein kinase Gphosphorylation and endocytic recycling and exocytosis of CFTR-containing vesicles by membrane trafficking (3,4) (Fig. 1) . But the mechanisms that control the intracellular trafficking itinerary of CFTR in intestinal cells are poorly understood. Polarized intestinal epithelial cells possess an apical brush border (BB) with tightly packed microvilli or membrane protrusions on their apical surface. Microvilli increase the apical surface area of enterocytes and are supported by bundled arrays of core actin filaments (~20) and cross-linking proteins that provide the structural support for each microvillus. The microvillus core rootlets descend into the apical cytoplasm of the enterocyte at the terminal web (TW), a region rich in cross-linking filaments composed of myosin and spectrin (5) (Fig.2) . Members of the unconventional myosin family of actin binding proteins are abundant in intestinal epithelial cells where they play major roles as cargo transporters and as mediators of membrane-cytoskeleton adhesion. The subcellular localization, known functions and diseases associated with intestinal myosins are listed in Table 1 . The distribution of CFTR along the microvillar membranes, in clathrincoated pits and vesicles, apical recycling and exocytic vesicles in the native enterocytes, suggest a role for myosins in regulating intracellular traffic (6) . Myosin 6 is the only myosin family member that has been shown to associate with CFTR in the intestine. This minus end directed motor is unique among actin-based motors because it moves away from the plasma membrane. In cultured cells, myosin 6 associates with clathrin coated pits and coated vesicles and interacts with clathrin, the endocytic adaptors AP-2 and the C terminus of disabled 2 (Dab2) to regulate clathrin-mediated endocytosis (7) . In the intestine, myosin 6 is present in both crypt and villus enterocytes and localizes to the lower third of microvilli, the terminal web and intermicrovillar domain, the sites of clathrin-coated pit formation and initiation of apical endocytosis. Absence of myosin 6 leads to decreased negative membrane tension in the enterocyte BB microvillus and intermicrovillar membrane in myosin 6 mutant mice (Myo6 sv/sv) (Fig. 2, 3) , markedly delayed apical endocytosis of CFTR, redistribution of the endocytic adaptor Dab2 in the terminal web region, and exaggerated fluid secretion following treatment with E. coli enterotoxin due to accumulation of CFTR on the surface of crypt and villus enterocytes in the small intestine (8, 9) . Myosin 6 was suggested as the candidate motor that would move CFTR down the microvillar membrane and sequester channels into clathrin coated pits for endocytosis, but its distribution in the enterocyte brush border does not support such a role. How CFTR migrates down the microvillus to the intermicrovillar membrane of enterocytes in preparation for endocytosis remains unknown. Myosin 6 distribution in the enterocyte BB contrasts to rat kidney proximal tubules, where it localizes along the entire length of microvilli, and plays a role in redistribution of NHE3 Na + /H + exchanger along the microvilli during acute hypertension (10) .

Myosin 1a (Myo1a) is the major plus end actin motor and counterpart to myosin 6 in the enterocyte BB. In contrast to myosin 6 that is present in both crypt and villus enterocytes, Myo1a is essentially confined to the BB of mature villus enterocytes in small intestine where it bridges the actin core to the microvillar membrane (Fig.3) . Myosin 1a is in lipid rafts, binds to and plays a critical role in tethering and retention of the BB hydrolase sucrase isomaltase (SI) in the microvillar membrane (11) . Lack of this plus end motor in Myo1a KO mice leads to bnormalities in microvillar membrane tension, blebbing and abnormal membrane protrusion in BB of villus enterocytes in the small intestine (Fig.3) . Preliminary studies of CFTR localization and anion transport in Myo1a KO mice indicate that CFTR accumulates in the terminal web region, is absent from the BB and redistributes to the basolateral domains of villus enterocytes in the small intestine. Analysis of CFTR anion transport (Ussing chambers) in intestinal tissues from Myo1a KO mice reveal a marked (>50%) reduction in CFTR current (Isc) compared to WT controls but only in intestinal segments that possess villi. Ongoing studies will seek to determine the specific role that Myo1a plays in regulating CFTR localization and function in the enterocyte BB.

Lianwu Fu, Ph.D., Andras Rab, Zsuzsa Bebok, M.D. and James F. Collawn, Ph.D.

Cell surface proteins often contain short (4-6 amino acid) cytoplasmic tail signals that allow for their efficient entry into clathrin-coated pits and clearance from the cell surface (reviewed in (1)). The two most common types of internalization signals consist of a tyrosine-based signal (NPXY or YXXΦ) and a di-leucine-based signal (D/EXXXLL/I) where X is any amino acid and Φ is a large hydrophobic residue. For the cystic fibrosis transmembrane conductance regulator (CFTR), two such signals have been identified in the C-terminal tail, YDSI and DSIQKLL (2, 3) . There is evidence that the YDSI and DSIQKLL signals are recognized by the adaptor protein-2 (AP-2) complex at the cell surface (3) (4) (5) . Interestingly, both tyrosine-based and di-leucine-based signals have been shown to act as lysosomal targeting signals (6) . Furthermore, earlier studies indicate that ubiquitination can serve as an internalization signal as well as a signal for lysosomal targeting (6, 7), making it difficult to differentiate between internalization and lysosomal targeting signals.

The AP-2 complex is composed of alpha-2, beta-2, mu-2 and sigma-2 subunits. AP-2 provides two functions: (1) it recognizes and binds to the internalization signal; and (2) it binds to and promotes clathrin assembly. The beta-2 subunit binds to clathrin, the mu-2 binds to the internalization signal of the cell surface protein, and the alpha-2 subunit recruits accessory proteins that are important for coated-pit formation and scission (reviewed in (6)). For CFTR, there is evidence that the tyrosine-based signal interacts with AP-2. Weixel and Bradbury demonstrated that a photoactivatable peptide containing the YDSI sequence could be UV cross-linked to the 50 kDa mu-2 subunit of the AP-2 complex. Furthermore, they demonstrated that overexpression of a dominant negative mu-2 subunit inhibited CFTR internalization in HeLa cells (4) . More recent studies in HEK cells by Collaco and Ameen demonstrated that the alpha-2 subunit was important for CFTR internalization and importantly, GST pull-down assays indicated that the alpha-2 subunit is a binding partner for CFTR (5) . Although it is unclear if both subunits of the AP-2 complex interact directly with the cytoplasmic tail of CFTR, both studies support the idea that the AP-2 directly interacts with CFTR and this interaction is important for the clearance of CFTR from the cell surface.

In our studies, we investigated the relationship between CFTR endocytosis and CFTR protein half-life. In that regard, we examined two different types of mutations in CFTR, one in which we ablated the tyrosinebased signal in CFTR, and another which included 2 nat-urally occurring CFTR mutations, R31L and R31C. In the first, we mutated the YDSI sequence to ADSA and found that CFTR internalization was reduced by more than 75%, yet this change had no effect on CFTR protein half-life (8) . In the second, we noticed that both R31 mutants appeared to create potential internalization signals, e.g., YRQR31 -> YRQC31 or YRQL31. Analysis of these CFTR mutants in endocytosis assays revealed that both mutants were internalized at double the rate of the wild type protein. More importantly, the mutations had no effect on the protein half-life (9) . And finally, studies by Bradbury demonstrated that another naturally occurring mutation, N287Y, generated a tyrosine-based signal in CFTR's cytoplasmic loop 2. The appearance of this second internalization signal almost doubled the rate of CFTR internalization and again, this mutation had no effect on CFTR half-life (10) . These studies indicated that CFTR mutations that introduced a tyrosine-based signal dramatically affected internalization, but had no effect on CFTR half-life, indicating that the two processes are not necessarily coupled.

Based on the above studies, it appeared that AP-2 interactions were not critical for CFTR down-regulation. Furthermore, it is clear that CFTR endocytosis is affected by a number of proteins besides AP-2. Details regarding the role of these accessory proteins in CFTR endocytosis are now beginning to be appreciated. Swiatecka-Urban, Stanton and colleagues demonstrated that when CFTR was immunoprecipitated from polarized Calu-3 cells, a number of proteins were brought down in a complex that includes myosin VI, Dab2, and clathrin (11) . They also showed that myosin VI, a minus-end directed actin motor, was important for CFTR internalization by demonstrating that a dominant-negative form of myosin VI interfered with CFTR internalization in HEK293 cells. Further support for a myosin VI role in endocytosis came from studies by Ameen and Apodaca in intestinal cells from myosin VI-deficient mice (12) . Coimmunoprecipitation studies performed in polarized intestinal CaCo-2 cells illustrated that endogenous CFTR brought down a complex that included Dab2, alpha-AP-2, myosin VI, and clathrin (5). Furthermore, Collaco and Ameen demonstrated that in the intestine, CFTR interacts indirectly with Dab2 through the AP-2 complex (alpha-2 subunit) (5).

Dab2 (Disabled 2) is particularly interesting because like AP-2 it can act as an adaptor protein. Maurer and Cooper demonstrated that Dab2 functions as an adaptor independently from AP-2 in LDL receptor internalization (13) . Furthermore, they showed that Dab2 is important in megalin internalization in visceral endoderm cells (14) . Both of these receptors have an NPXY type tyrosine-based signal. Interestingly, apical CFTR is increased in the intestines of Dab2 knockout mice, suggesting that it is important in CFTR internalization and/or down-regulation (5) . In this talk, we will present a model to explain the roles of AP-2 and Dab2 in CFTR endocytosis and lysosomal targeting in polarized human airway epithelial cells. We will also discuss how these two adaptors differ with regard to their effects on CFTR down-regulation.

William E. Balch, Ph.D.

The cell exploits the environment sensitive (1) and emergent properties of proteostasis (2-4) to generate and maintain proteome balance (3) in diverse cell, tissue and organismal systems during development and in response to aging and disease such as cystic fibrosis (CF). The proteostasis network (PN) is a system of signaling pathways (UPR*, HSR*, ARE* and IR*), and chaperone (Hsp40/70/90, chaperonins, and sHSPs*) and degradative components (UPS*, ALES*) that respond to environmental stress through genetic and epigenetic mechanisms to manage folding and membrane trafficking (5) . Physical, pathological and inherited challenges to the energetics of the biological fold (its "landscape") can compromise proteome balance (2). Our goal is to understand how the inherited misfolding prone protein ᭝F508 CFTR is managed by proteostasis.

By use of systems level proteomic, genomic and chemical tools we are building a dynamic, multi-layered view of the healthy CFTR biological protein fold and the changes that occur in response to energetically compromised folding stress such as is observed in CF and ᭝F508. We provide evidence that chemical biology man-agement of the PN can alter the composition of the local proteostasis program to restore function. The discovery of tools that redirect the ᭝F508 fold to function highlights the potential value of the emergent properties of the proteostasis network to therapeutically rebalance the epithelial cell proteome to potentially benefit CF patient health span. *UPR (unfolded protein response); HSR (heat shock response); ARE (anti-oxidant response elements); IR (inflammatory response); UPS (ubiquitin-proteasome system); sHSPs (small heat shock proteins); ALES (autophagy-lysosomal-endosomal system). In people with cystic fibrosis pulmonary exacerbations are associated with a significant reduction in quality of life, more rapid decline in FEV 1 and reduced survival (1). There is a good evidence base to help with treatment decisions for chronic treatment, mostly aimed at improving FEV 1 and reducing pulmonary exacerbations (2) . However, the evidence base for many aspects of the treatment of pulmonary exacerbations is less extensive. The important issues of 1) when to treat; 2) which antibiotics and other therapies to use; 3) how long to treat; and 4) what to do at the end of a pulmonary exacerbation, are largely based on custom and practice. Most published studies on exacerbations are observational with few studies exploring and comparing alternate options.

When to start treatment? Defining pulmonary exacerbations has been a subject of much effort in the past 10 years (3). A number of constructs have been developed to help with a definition that will be useful particularly in clinical trials but there has been less focus on a pragmatic definition for use in clinical decision making. A tool developed in the Akron CF Centre has shown some utility in focusing physicians on the symptoms of pulmonary exacerbation and in one study has resulted in more consistent initiation of antibiotic therapy (4).

Significant variation in decision making to initiate antibiotics was seen in a recent vignette-based study (5) . Systemic symptoms and measures of pulmonary function were more strongly related to a decision to treat with IV antibiotics while symptoms of cough and sputum and crackles on exacerbation were less important. It may also be that people with CF perceive an exacerbation differently to clinicians.

Defining an exacerbation for clinical practice should focus on operationally describing the symptoms that in an individual should be considered an indication to start IV antibiotics. These may be heterogeneous and need to be individualised for each patient. It is likely that a low threshold for initiation of antibiotic therapy is best practice, as better performing centres use more antibiotics (5) .

Oral, inhaled or intravenous? The decision to commence antibiotics for an exacerbation is determined by severity of the symptoms and clinical signs and physio-logical measurements associated with the exacerbation and by the organism(s) identified. Most exacerbations in people with CF not chronically infected with Pseudomonas aeruginosa are associated with Staphylococcus aureus or Haemophilus influenzae, though on some occasions no bacteria are isolated by conventional methods (6) . These organisms can be effectively treated with an appropriate dose of oral antibiotics directed against these organisms. This is usually higher than in non-CF situations. Treatment should be started empirically and then adjusted accordingly when the organism has been identified and anti-microbial susceptibility testing undertaken. Methicillin resistant Staphylococcus aureus (MRSA) associated exacerbations can be treated with oral antibiotics but when severe may require intravenous treatment.

For patients with P. aeruginosa infection and a mild exacerbation, when there is no change in lung function and only a modest increase in cough and sputum, treatment with an oral agent such as ciprofloxacin is a rational approach. For exacerbations with severe symptoms or a significant reduction in FEV 1 , intravenous antibiotics are indicated. There are no data to support the use of inhaled antibiotics for pulmonary exacerbations, though a number of centres use inhaled tobramycin instead of an intravenous aminoglycoside in patients who have impaired renal function. Treatment of exacerbations associated with other Gram negative infections or nontuberculous mycobacterial infection is challenging and may require the use of oral, intravenous and inhaled therapies (7) .

Choice of intravenous antibiotic. Most CF guidelines recommend combining an antipseudomonal extended-action penicillin, cephalosporin, monobactam or carbapenem and aminoglycoside for the treatment of a pulmonary exacerbation (6) . The rationale for this combination is largely to prevent the development of resistance. Again, doses are usually higher than those normally used (6) . There is no good evidence that single agent therapy is better or worse than dual agent therapy but the resistance concerns remain with single agent therapy. The role of anti-microbial sensitivity testing to direct antibiotic choice is an area of some controversy. The accuracy of sensitivity testing has been questioned and in one study there was no impact on effectiveness of treatment whether antibiotics were used to which the P. aeruginosa was sensitive compared to resistant (8) . Synergy testing has been explored in multi-resistant P. aeruginosa and Burkholderia cepacia complex organisms but again appears to offer no advantage over an empirical decision (9) . However, recent work from our CF centre has suggested that failure to recover from a pulmonary exacerbation is more likely if the treatment is with antibiotics to which the patient's P. aeruginosa is not sensitive. Pragmatically, antibiotic sensitivity testing may be considered a guide to antibiotic choice but other considerations such as allergy to antibiotics, previous good response and patient preference should also be considered.

Length of time for IV antibiotics. A few studies have addressed this particular issue. Most of the improvement in FEV 1 and other metabolic and inflammatory markers improves after 7 days of IV antibiotics (10). This was confirmed in a large study of FEV 1 recovery which demonstrated almost all of the recovery was achieved within 14 days (11) . This is difficult to study as convention has been to treat for 14 days. However, this study suggests that it is reasonable in practice for IV antibiotic courses to be for 10-14 days depending on response. If there is full resolution of symptoms, inflammatory markers (white cell count and CRP) and full recovery of FEV 1 , then a 10-day course is a reasonable choice. For exacerbations which are not resolving there is very little data to help from studies for decision making. At some centres the IV course is continued for 3 or 4 weeks while an alternative approach is to change the IV combination for a further 2 weeks and monitor progress. Management of exacerbations should be complemented by increased airway clearance, consideration of nutritional loss, active management of blood glucose, and correction of hypoxia and hypercapnia, if required. There is no evidence to support the routine additional use of oral corticosteroids.

What to do at the end of a pulmonary exacerbation? All patients should have their chronic therapy reviewed at the end of an exacerbation. Consideration of factors that may have contributed to the episode should include adherence to regular therapies, undertreatment and new complications such as diabetes. The importance of prevention of exacerbations should be discussed and strategies implemented to reduce further episodes and if they do develop, early initiation of therapy.

Microbial diversity and exacerbations. A number of recent studies have demonstrated that there is a large number of different bacterial genera in the sputum from people with CF (12) . These include a range of anaerobic bacteria, Streptococci species, as well as viruses and fungi. The relevance of these observations to clinical care is currently unclear. Molecular information from nonculture based methods of identification may become increasingly available and may prove to be a useful tool in determining the most appropriate antibiotic therapy for exacerbations. Several current ongoing studies may help to clarify this.

Conclusions. Treatment of pulmonary exacerbations remains a major part of cystic fibrosis care, yet is an area in which we do not have sufficient studies to allow us to optimise care. It is clear that exacerbations are associated with poorer outcomes and studies focused on how to maximise the treatment of exacerbations will be important. Beta-lactam antibiotics exhibit time-dependent bactericidal activity (1). More importantly, beta-lactam antibiotics exert maximum bactericidal activity when their concentrations are 3-4 times the minimum inhibitory concentration (MIC) (1). In addition, beta-lactam concentrations must remain above the MIC for >70% of the dosing interval in order to achieve maximal bactericidal activity (2) . Vogelman et al. reported that the efficacy of ticarcillin against P. aeruginosa reached its maximal benefit as the time above MIC approached 100% (3). This is of particular concern because patients with cystic fibrosis (CF) exhibit increased renal clearance of beta-lactam antibiotics, such as ticarcillin, by as much as 30% compared to healthy volunteers (4). Our preliminary data demonstrated beta-lactam clearance in CF patients increases by as much as 23 to 56% during a 14-day hospital stay (5) . Intermittent administration of ticarcillin/clavulanate in patients with CF resulted in T>MIC against P. aeruginosa at a susceptible MIC of 64 µg/ml of merely 17% (in press). As a result, patients with CF require higher doses or more frequent administration of beta-lactam antibiotics (4).

In contrast, continuous infusion (CI) of beta-lactam antibiotics maintains serum levels above MIC at concentrations that maximize bactericidal activity (1). In patients with CF, Rappaz et al. reported that the ceftazidime concentrations delivered via CI exceeded an MIC of 4 µg/ml against P. aeruginosa 100% of the time (6). Hann et al. concluded that CI of cefepime in patients with CF effectively achieves and maintains concentrations above MIC and reduces the total daily dose by 20% (7). Kuti et al., demonstrated that meropenem delivered via CI was able to maintain serum concentrations >4 µg/ml for 100% of the dosing interval (8) . While there is sufficient evidence to conclude that CI is more effective at increasing T>MIC for any infecting organism, there is limited data regarding the efficacy of continuous versus intermittent delivery of beta-lactam antibiotics. In the largest published trial to date, Hubert et al. compared ceftazidime delivered via CI versus thrice daily infusion (9) . The authors concluded that the regimens were equivalent. However, the mean change in FEV1% predicted was significantly better in patients with resistant bacteria when ceftazidime was delivered via CI vs. intermittent infusion (p<0.05) (9) . Also, the mean difference in time to next intravenous antibiotic course was 0.4 months longer with continuous versus short infusion ceftazidime (p=0.04). Beta-lactam antibiotics delivered via CI are well tolerated. No difference in laboratory values (i.e. serum creatinine or liver function tests) was seen between ceftazidime delivered via CI versus intermittent infusion (6, 9, 10) . In addition, cefepime and meropenem administered via CI are well tolerated (8, 11) .

Aminoglycosides, such as tobramycin, exhibit concentration-dependent bactericidal activity (12) . As a result, the bacterial killing effect is correlated with peak serum concentrations. Maximal bactericidal activity has been demonstrated when the aminoglycoside peak serum concentration/minimum inhibitory concentration (MIC) ratio is greater than 10 (13). Once-daily dosing can achieve higher peak serum concentrations and peak/MIC ratios and minimize the development of adaptive resistance (1, 14) . Once daily dosing of aminoglycosides may also reduce the risk of nephrotoxicity (15, 16) . This has led the CFF to recommend once-daily dosing of aminoglycosides to be the preferred method of administration in all CF patients with acute pulmonary disease (17) . However, limitations exist with the use of once-daily dosing of aminoglycosides in CF. The optimal peak serum concentration and peak/MIC ratios have not been established for acute exacerbations of CF (17) . The potential for nephrotoxicity, ototoxicity, and vestibular toxicity necessitate appropriate therapeutic drug level monitoring (15, 18) . Finally, data regarding efficacy of once versus twice/three times-daily dosing of aminoglycosides is limited. The largest study published to date regarding the optimal aminoglycoside dosing strategy in CF patients (pediatric and adult) was the TOPIC trial (16) . The authors concluded that oncedaily administration of tobramycin was equivalent to thrice-daily dosing and resulted in less nephrotoxicity in children. 

Pulmonary exacerbations are a common complication of cystic fibrosis (CF) lung disease, which lead to significant morbidity and accelerate lung function decline (1,2). Despite aggressive therapy, many patients do not return to their pre-exacerbation lung function after treatment (3) (4) (5) . Therefore, developing the best possible approach to therapy is of vital importance. Unfortunately, systematic review of exacerbation therapy has not provided answers to several important questions regarding the optimal approach to therapy (6) . Treatment with antibiotics is the mainstay of exacerbation therapy. Oral antibiotics can be effective in treating the majority of exacerbations (7) . However, patients with more severe exacerbations or recurrent symptoms require intravenous (IV) antibiotic therapy.

There is controversy regarding the optimal method for delivery of IV antibiotics. Traditionally, admission to the hospital has been the preferred approach for the administration of IV antibiotics. However, advances in the percutaneuos placement of central venous catheters allow for the delivery of IV antibiotics in an outpatient setting. Systematic review suggests that the use of a longer central line may provide benefit over shorter IV catheters for antibiotic therapy (8) . Peripheral insertion of central catheters has gained widespread use throughout the world for the delivery of antibiotics to patients with CF. However, there are reports of venous thrombosis associated with the use of these catheters (9) (10) (11) (12) . Although totally implantable vascular access devices are frequently used in patients with CF there are no randomized studies about the value of these catheters (13) . However, their widespread use suggests that they are safe and effective (14, 15) . Factors that influence the decision to place an implantable catheter include: frequency of exacerbations, difficulty of placement of a peripherally inserted central catheter and patient preferences. Ultimately, the decision about the type of venous access used should be made on an individual patient basis.

The ability to provide stable access for the delivery of IV antibiotics has allowed patients to receive antibiotic therapy in an outpatient setting. However, there is concern among many care providers that therapy delivered at home will not result in comparable outcomes to treatment in the hospital. It is important to remember that the delivery of antibiotics is only one component of the overall therapy of pulmonary exacerbations. Only one study has compared home and hospital treatment of pulmonary exacerbations in a prospective fashion (6, 16) . Wolter et al (17) found similar outcomes between those patients that completed most of their therapy at home compared to those that remained in the hospital for their entire treatment course. However, several retrospective studies have found that patients treated in a hospitalized setting do in fact have better outcomes (18, 19) . More recently, Collaco et al (3) examined the outcomes of 1,278 exacerbations experienced by 479 subjects in the US twin and sibling study. These authors found that the outcome of patients treated entirely in the hospital or entirely at home or in some combination of both locales had similar outcomes with respect to lung function and time to the next exacerbation. Studies have suggested that there can be benefits to different aspects of quality of life measures with either hospitalization or home therapy (17, 20) . Therefore, consideration of the differing burdens of hospitalization or home IV administration must be taken into account when making treatment decisions. The disruption of home and family life must be weighed against the certainty of increased airway clearance, appropriate rest and adherence to medication delivery schedules. For home IV therapy to be successful, the patient must have the time, resources and family support to administer the antibiotics and provide intensive airway clearance and inhalational therapy.

Many considerations must be taken into account when making the decision on how to treat a patient experiencing a pulmonary exacerbation. Unfortunately, there are few studies to help guide these decisions. The type of venous access and the locale of therapy have the potential to greatly influence the outcome of therapy. Therefore, careful consideration should be given when making these decisions.

References Studies of management of patients with CF are limited by the relatively small patient population and heterogeneity of the patient population. Although there is evidence from Phase 3 trials and consensus statements to guide chronic therapies in CF (1-4), the management of pulmonary exacerbations is not as well studied. Consensus statement for management of pulmonary exacerbations was hampered by lack of data, poor quality or conflicting data (5) . CF physicians also face the challenge of managing individual patients where specific clinical issues complicate therapy; such as pregnancy, drug allergies or where unusual or multiple bacteria are present. In these situations, where the literature is not very helpful, one usually contacts CF colleagues to ask for advice. This session aims to review an approach to these problems.

With improving survival and increasing numbers of adults with CF, care centres are managing more and more pregnancies. Large registry (6) and single centre studies (7, 8) have demonstrated that pregnancy itself does not have an impact on survival. However, clinicians need to be able to manage pulmonary exacerbations in pregnant women with CF. Pregnancy is an exclusion criteria for drug trials and animal teratogenicity studies are difficult to extrapolate, so product monographs are usually not helpful. The usual statement that use in pregnancy is not recommended unless the potential benefits to the mother outweigh the potential risks to the fetus may be the case when treating a woman with a CF pulmonary exacerbation.

There are some drugs that should be avoided in pregnancy including aminoglycosides, tetracyclines and chloramphenicol. Aminoglycosides cross the placenta and there are case reports of fetal ototoxicity including deafness. Clinical stability is an important goal during pregnancy and inhaled tobramycin reduces the number of pulmonary exacerbations. Median levels of tobramycin one hour after receiving 300 mg of tobramycin solution for inhalation are 0.94 µg/ml (range of 0.18 -03.62) (9) . Serum levels after lower doses of inhaled tobramycin can be measured. Tetracyclines may be used to treat B cepacia complex. Tetracyclines cross the placenta and cause yellow or grey staining of deciduous teeth in children born to mothers who received tetracycline in the 2 nd and 3 rd trimester. No increased incidence of malformations was seen in 2 cohort studies of 274 infants whose mothers used tetracycline in the 1 st trimester (10, 11) . Tetracyclines deposit in the skeleton where they inhibit bone elongation and calcium uptake. This appears to be a temporary inhibition of bone growth. Trimethoprim-sulfamethoxazole may be used to treat Stenotrophomonas and B cepacia complex. Sulfonamides should not be used after 32 weeks of pregnancy as they displace bilirubin from binding to albumin and can cause kernicterus in the newborn. Trimethoprim should not be used in the first trimester as it is a folate antagonist and may result in increased risk of neural tube defects. Cartilage toxicity has been demonstrated in animal studies of quinolones. However, a controlled study prospectively enrolling and following 200 women exposed to fluoroquinolones during pregnancy did not show increased risk of major congenital malformations in the group exposured in the first trimester and no clinically significant musculoskeletal dysfunctions in the children (12) . A series of observational cohort studies suggest that azithromycin does not cause an increased risk of congenital abnormalities when used in pregnancy (10, 11, 13) . There is no data available on the use of aztreonam or meropenem in pregnancy. Cephalosporins and penicillins are generally considered safe in pregnancy. Another common challenge is in the CF patient with unusual or multiple bacteria present in sputum cultures. CF clinicians are just becoming familiar with treating Stenotrophomonas and Achromobacter, and now we have to decide if we should treat other gram negative bacteria such as Ralstonia, Pandoraea and Inquilinus (14) (15) (16) . These bacteria are closely related to Burkholderia species but it is not clear if they are pathogens although there are case reports of significant illness in CF patients infected with these bacteria (17) . In a similar way, we see nontuberculous mycobacteria (NTM) in combination with the usual CF pathogens and have to decide if they are bystanders or causing disease. One general approach is to treat the other bacterial pathogens and see if there is a clinical response. If no clinical response occurs, then determining persistence of the NTM with serial cultures and "burden" of infection with smears and using a combination of clinical symptoms and radiographic information (HRCT scan) may help with this decision. A trial of therapy, with frequent reevaluation, may be the best way of dealing with this uncertainty.

When approaching the challenge of making decisions and managing patients in the absence of evidence based guidelines, it is helpful to collect all the information you have and utilize resources. In our centre, the CF physicians get together on a regular basis to discuss the patients with challenging medical problems and put together a plan for care. Pulling together a case presentation with a literature review to present to your colleagues followed by a thoughtful discussion and involvement of other experts such as pharmacy is extremely helpful. Of course, email communication with CF experts around the world identified through internet and a literature search is not only helpful but can be reassuring to know that you are not alone in handling these challenges.

References 

Cara I. Kimberg, Ph.D.

It is generally accepted that cystic fibrosis (CF) does not impact cognitive function; however, this is substantiated primarily through clinical observations and there are few empirical studies to support this claim. It is important that the cognitive skills of individuals with CF are better understood on a population level, as even subtle deficits may impact disease management. Based on the similarities between CF and other disease groups, there is reason to believe that individuals with CF may be at-risk for attentional deficits or executive dysfunction (i.e., impairments in planning, organization, shifting, initiating, and emotional control). For example, individuals with sleep apnea, a disorder characterized by pauses in breathing or shallow breaths, are found to demonstrate executive dysfunction and behaviors consistent with attention deficit hyperactivity disorder (ADHD)(1). Of note, when sleep apnea is treated, these neurobehavioral deficits remit. Similarly, patients with chronic obstructive pulmonary disease (COPD) performed worse on tests of attention, executive function (EF) and processing speed compared to controls. The degree of impairment was predicted by low levels of forced expiratory volume (FEV) and forced vital capacity (FVC)(2,3). Using the negative cognitive effects of chronic or intermittent hypoxia demonstrated by patients with sleep apnea and COPD as a model, it is plausible that, due to compromised pulmonary functioning, individuals with CF may demonstrate a similar pattern of deficits. This may be especially true for pediatric patients, as oxygen deprivation could have more adverse consequences on the developing brain.

To date, limited research has been conducted on the neurocognitive function of individuals with CF. A study by Zinman and colleagues (4) explored the differential impact of oxygen versus room air at night on cognitive function and found no group differences in processing speed, memory or achievement. Koscik and colleagues (5) determined that children with vitamin E deficiency at traditional diagnosis had poorer cognitive performance on a task of general intellectual ability several years later compared to children diagnosed through newborn screening who did not demonstrate this nutritional deficit. Lastly, Dobbin and colleagues (6) evaluated the impact of pulmonary exacerbations on objective neurobehavioral performance. Pulmonary exacerbations were found to negatively impact neurocognitive performance, as well as a driving simulation task. Importantly, treatment of the exacerbation led to improved cognitive function.

The value of understanding the cognitive skills, particularly in the domains of attention and EF, of individuals with CF is highlighted through intervention research. For example, in the context of an adherence intervention, it is possible that responders and nonresponders differ based on EF skills. An intervention that utilizes text messages to remind patients to take their medication may be successful for adolescents with executive dysfunction, as the text message directly targets their weakness (i.e., poor organization and planning). In contrast, this intervention may have no effect on adolescents with good EF, as the text message does not address the underlying reason for poor adherence (e.g., depressive symptoms). Thus, the success of an intervention may be influenced by the cognitive profile of the participants. It is important for researchers to be mindful of cognitive function as a potentially confounding factor when designing and implementing interventions. Clinically, it is essential for multidisciplinary CF team members to be aware of individual cognitive impairments. According to the Diagnostic and Statistical Manual of Mental Disorders (DSM-IV-TR, ref 7), 3%-7% of school-age children are diagnosed with ADHD; therefore, it would be expected that a subset of children with CF also meet criteria for this diagnosis. In fact, a retrospective study by Georgiopoulos and Hua (8) determined that 9.6% of children and adolescents with CF, who were seen by the Massachusetts General Hospital CF Program, were referred to a psychiatrist and subsequently received a diagnosis of ADHD. The behavioral profile of ADHD, which includes symptoms of impulsivity, inattention, hyperactivity and executive dysfunction, has the potential to significantly complicate disease management if not properly treated. For example, an impulsive child may eat without taking enzymes or an adult with poor working memory skills may forget the order in which to complete his/her treatments. These instances of poor disease management are consistent with the Georgiopoulos and Hua study (8) , as many of the patients ultimately diagnosed with ADHD presented with poor treatment adherence.

To evaluate attention and EF, standardized measures including performance-based tests and questionnaires can be administered. Such measures are given as part of a clinical or research neuropsychological assessment. Neuropsychological evaluations focus on the link between brain structures/pathways and cognitive skills. Test measures for the assessment are selected based on the referral or research question and evaluate general intellectual function, academic achievement, language, visual-spatial and visual-motor abilities, attention, learning and memory, EF and behavioral/emotional functioning. A performance profile, based on the determined pattern of strengths and weaknesses across the multiple domains, is used to guide recommendations. In sum, there are identified factors that may place individuals with CF at risk for attentional and EF impairment. Further research, using neuropsychological assessment, is needed to delineate the cognitive profile of these patients, as cognitive function may impact disease management and research findings.

Anna M. Georgiopoulos, M.D.

Prevalence of ADHD in pediatric medical illness and CF. Attention Deficit-Hyperactivity Disorder (ADHD) occurs in 10% of children (1) and 4-5% of adults (2) in the general population. Chronic medical illness increases risk for psychiatric illness (3) . A study of chronically ill children revealed ADHD in 9.2% vs. 4.3% in controls (4) . While ADHD is strongly familial (5), medical stressors such as intrauterine growth retardation (6) , perinatal hypoxia (7), and malnutrition (8) could encourage phenotypic expression in genetically predisposed patients. The risk of ADHD is higher in males with congenital adrenal hyperplasia (18.2%) and familial male precocious puberty (44.4%) (9) . ADHD rates of 38%-50%+ occur in fetal alcohol syndromes (10), congenital heart disease (11), chromosome 18 abnormalities (12), and velocardiofacial syndrome (13) . In a cohort of 188 pediatric CF outpatients, ADHD was diagnosed by a child psychiatrist in 9.6%; 59% had a family history of ADHD (14) . These data may underestimate the prevalence of ADHD in CF, as they exclude those declining evaluation or diagnosed elsewhere. In another study of CF children, 21% met criteria for a disruptive behavior disorder (15) . By comparison, the rate of ADHD in diabetic children (2.1%) was lower than in complicated epilepsy (12.0%) (16) .

Prevalence of ADHD in adults with CF and pediatric patient caregivers. The Adult ADHD Scale Self Report Symptom Checklist v1.1 (ASRS) was administered to adult CF outpatients and caregivers of pediatric CF patients as part of an ongoing study of psychopathology in CF. Effects on treatment adherence and medical outcomes. ADHD is associated with functional impairment including in self care (2, 17) and substance abuse (18) . Longitudinal data links impulsivity with early mortality (19) . Task avoidance, inattention and difficulty following directions may interfere with health regimens. Oppositionality, forgetting, and trouble managing the time required for care routines are commonly reported in CF families (20, 21) . White et al. (15) identified a trend between externalizing disorders and lower CF adherence.

In our study of pediatric CF patients with ADHD (14), 61% (11 of 18) presented with CF non-adherence. The 3 who refused ADHD treatment and the 2 who had medication trials resulting in minimal or no improve-ment had no change in adherence. One had much improved ADHD symptoms after treatment, but minimal improvement in adherence. The remaining 5 demonstrated improvement in adherence after successful ADHD treatment.

In our adult CF pilot sample described above, total score on the CF Questionnaire-Revised (CFQ-R) was higher (p=0.035) in patients meeting criteria for probable ADHD (47.7±9.1) vs. those without ADHD (37.0±13.3). Two subscales, Physical (65.6±16.0 vs. 45.0±26.7; p=0.04) and Vitality (58.3±19.1 vs. 44.8±20.9; p=0.09) accounted for this difference. FEV1 and BMI did not differ significantly in the ADHD vs. non-ADHD groups. There was significantly less self-reported adherence to exercise, nasal irrigation and diabetes care (p<0.05) in CF patients with any psychopathology (depression, anxiety and/or ADHD) than with no psychopathology. However, there were no significant differences in self-reported adherence between CF adults with ADHD vs. no ADHD. Several limitations complicate this finding, including sample size; the high prevalence of depression/anxiety in the non-ADHD group; sample bias toward CF-ADHD patients who do attend clinic and are more likely to adhere to care; and lack of insight/decreased accuracy in self-reported adherence in patients with ADHD.

Treatment of ADHD in CF. In our pediatric cohort (14) , ADHD medication treatment was attempted in 13 cases, achieving an improvement rating of much or very much improved in 8 cases. In 3 cases, the Best Regimen consisted of stimulant monotherapy (mixed amphetamine salts; methylphenidate); 2 consisted of non-stimulant monotherapy (atomoxetine; nortriptyline); 2 used a combination of 2 non-stimulants (atomoxetine or nortriptyline + guanfacine); and 1 used a stimulant plus a non-stimulant (buproprion XL+ Orosmethylphenidate). The decision about whether and how to treat ADHD in CF requires careful consideration of psychiatric and medical status. In our study, stimulants were more effective but more often associated with weight loss than non-stimulants. The 2 patients treated with nortriptyline demonstrated increases in BMI percentile (from 22% to 63% and from 34% to 60%) along with improvement in ADHD. Tricyclic antidepressants are considered third-line ADHD agents, after stimulants and FDA-approved nonstimulants. However, the risk/benefit ratio may differ in CF, where low weight and comorbid anxiety and depression are common. If stimulants are used, boosting caloric intake, adding cyproheptadine or other appetite stimulators and treating constipation may help maintain weight. Conclusion. ADHD is common and treatable in patients with CF. Oppositional behaviors, distractibility, forgetfulness or extreme motor activity interfering with CF care should trigger consideration of ADHD. Stimulants, nonstimulants and combination therapies are viable treatment options. Further research on the prevalence and treatment of ADHD in CF and its impact on medical adherence and outcomes is warranted.

Bruce J. Masek, Ph.D.

Nonadherence with medical treatment regimens is prevalent in an estimated one out of two children with chronic medical illness. The model of adherence risks described by Ickovics and Meisler (1) provides a useful framework for the development of interventions to improve adherence to medical therapy in children with chronic illness. In this model five factors are considered, which include: 1) child's developmental stage; 2) disease symptoms; 3) complexity of treatment regimen; 4) child-parent relationship; and 5) clinical setting. Behaviorally-based interventions that address one or more of these factors are generally effective for increasing treatment regimen adherence for childhood illnesses (2) . The CF patient with ADHD may present a special challenge to health care providers in terms of nonadherence due to behavioral noncompliance, which is so often a feature of the disorder. CBT approaches focused on skills-based interventions to treat ADHD symptoms in adults have recently gained empirical support (3) . The aim of this talk will be to describe CBT interventions to promote medical adherence in children and adult CF patients with ADHD that can be delivered by a member of the multi-disciplinary health care team.

References 

Gregory P. Downey, M.D. 1. Medicine, National Jewish Health, Univ. of Colorado, Denver, CO, USA; 2. Immunology, Univ. of Colorado Denver, Denver, CO, USA Cystic fibrosis (CF) is an autosomal recessive disorder caused by mutations in the cystic fibrosis transmembrane conductance regulator (CFTR), a cAMP regulated Clchannel (1, 2) . While CF is a multi-organ disease, the clinical course of affected individuals is dominated by recurrent respiratory tract infections leading eventually to respiratory failure (3, 4) . Although the life expectancy of affected individuals has increased remarkably in recent years, the vast majority of CF patients will die from pulmonary disease consequent to progressive inflammatory damage (4-6). One of the characteristic features of CF is persistent and neutrophil-predominant inflammation involving the lower respiratory tract, initially centered in the airways. In fact, pulmonary inflammation is observed very early in life, even in the absence of detectable infection, suggesting the possibility of a primary abnormality of regulation of inflammation (7) . Paradoxically, despite the presence of increased numbers of neutrophils, bacteria survive and often thrive in CF lungs, suggesting a defect in antimicrobial functions of phagocytes in the CF airway (7) (8) (9) (10) (11) (12) (13) (14) . Whether this reflects an intrinsic neutrophil defect or is a consequence of the abnormal milieu of the CF lung remains unre-solved (15) . We will review selected aspects of abnormal neutrophil function that have been reported and consider genetic versus inflammatory neutrophil reprogramming in CF.

One possibility is that there is an intrinsic defect in phagocyte function. Indeed, there are reports of defects in CF neutrophils, including increased oxidant production (16) , decreased shedding of L-selectin (17), differential expression of TLRs (18) , and alterations in Clflux and intracellular pH regulation (19) that may hamper antimicrobial functions and predispose to inflammatory tissue injury. However, because expression of CFTR mRNA and protein is normally very low in professional phagocytes such as neutrophils and macrophages (20) , it is possible that these reported defects are secondary to extrinsic factors, such as abnormalities in the airway surface liquid (ASL) or to the systemic effects of pulmonary inflammation consequent to infection. However, some of these abnormalities in neutrophil function have been detected in asymptomatic CF carriers who have no apparent indication of systemic or pulmonary inflammation, compatible with the notion of an intrinsic phagocyte defect.

As noted above, another possibility is that phagocyte dysfunction is secondary to defective epithelial function consequent to CFTR mutations leading to dehydration and increased viscosity of the airway mucous (21) , to increased concentrations of proteases such as elastase in ASL leading to cleavage and inactivation of membrane receptors on the phagocytes or degradation of anti-microbial factors in the ASL (22, 23) , or to abnormal function of epithelial derived antimicrobial factors such as βdefensins in the ASL (24). An additional possibility is that there is failure to clear effete neutrophils from the CF lung, either as a consequence of delayed apoptosis of the neutrophils or defective clearance of the apoptotic neutrophils by the airway epithelial cells (25) .

From the above discussion it is apparent that defective neutrophil function in CF is likely multifactorial with contributions from primary defects in neutrophil function, from the abnormal composition of the ASL, and from abnormal clearance of apoptotic neutrophils leading to a prolonged and augmented inflammatory response in the CF lung. Loss of CFTR function has a profound impact on host defense in the respiratory tract, resulting in chronic infection and inflammation. The sequence of events leading from loss of CFTR anion channel function to lung disease pathology is complex, likely multifactorial, and a subject of intense investigation. Thus, no single hypothesis or pathologic scheme may suffice to explain the progressive lung disease associated with CF. Here are briefly summarized some facets of the antimicrobial defenses of the airways and how they may be altered in CF and contribute to lung disease.

Secreted host defense molecules of the airways. The surface and submucosal gland epithelia elaborate a multilayered, constitutive and inducible array of antimicrobials (1). This antimicrobial shield includes a number of peptide, protein, and lipid components. Some examples include the alpha-and beta-defensin peptides, the human cathelicidin peptide LL-37, lysozyme, lactoferrin, SLPI, surfactant proteins A and D, elafin, lipocalin-2, CCL20, S100 family members calgranulin A and B, and PLUNC. In addition to these polypeptides, secreted lipids including cholesteryl esters contribute directly to broad spectrum antimicrobial activity (2, 3) . A lactoperoxidase based oxidative system also contributes to host defense through the production of the potent antimicrobial hypothiocyanite (4) (5) (6) . The combined effects of these molecules are to inactivate inhaled or aspirated microbes without stimulating inflammation. Their mechanisms of action are diverse and include nutrient sequestration, opsonization, cell signaling, surfactant/dispersant (7), bacteriocidal, and bacteriostatic activities. Synergy and additivity contribute to the success of this redundant system. In concert, this polyfunctional system is remarkably effective, yet in CF this arm of innate mucosal immunity breaks down.

Loss of CFTR function impairs airway antimicrobial defenses by several mechanisms. CFTR is an anion channel. In addition to playing an important role in regulating the airway surface liquid (ASL) volume and composition through Cltransport, CFTR also conducts other anions including HCO 3 -, SCN -, and glutathione. Therefore impaired CFTR channel activity can have several effects on the ASL environment. Impaired bicarbonate secretion may alter ASL pH (8) and characteristics of secreted mucins (9, 10) . The combined effect of subtle changes in ASL composition and volume create an unfavorable environment for antimicrobials. Furthermore, changes in mucin composition and viscosity (11) may further hamper their effects by altering mucin charge properties and by reducing mucociliary clearance. Secretion of antimicrobials by surface and submucosal gland epithelia is both constitutive and responsive to neurohumoral stimuli. CFTR-dependent secretion is defective in CF and therefore the net abundance of antimicrobials may be reduced (12, 13) . Early evidence in a porcine model indicates that tracheal submucosal gland tissue mass is reduced in CF, a finding that if confirmed in humans, may provide an additional anatomical contribution to the host defense defect early in life. Loss of CFTR may alter the transcriptional program of cells, resulting in changes in the abundant gene products that contribute to host defense. For example, Erzurum and coworkers noted a reduction in the functional expression of interferon inducible gene products in primary CF airway epithelia, perhaps contributing to a reduction in antiviral defenses (14, 15) . Such a change in antiviral immunity could also contribute to inflammation and secondary bacterial colonization (16) .

The impact of disease on antimicrobials. In addition to the many direct links between loss of CFTR function and antimicrobial defenses, as the disease progresses and inflammation and infection persist additional secondary effects of disease further compromise airway innate immunity. A number of proteases including neutrophil elastase, cathepsins S and G, and bacterial products, can directly cleave host defense molecules and further reduce or destroy their activities. Chronic infection and inflammation also trigger extensive remodeling of the epithelium, changing the proportions of cell types present, causing goblet cell metaplasia, and further changing the secretory profile of the cells.

Lessons from new animal models. While mouse models contributed greatly to disease understanding, they fail to develop lung disease similar to humans with CF. Recently, new models of CF were developed using gene targeting in pigs (17) and ferrets (18, 19) . These new models reproduce several features of CF disease (18, 20, 21) . At birth, the airways of CFTR targeted pigs are free of inflammation but manifest a host defense defect without the secondary consequences of infection (17, 21) . Within the first months of life CF pigs spontaneously develop lung disease manifested by infection with bacteria, airway remodeling and mucus hypersecretion. Moreover, after intrapulmonary bacterial challenge with S. aureus, CF pigs fail to eradicate bacteria as effectively as wild-type pigs (17) . These results suggest that impaired bacterial elimination is the pathogenic event initiating a cascade of inflammation and pathology. CF pigs also exhibit altered tracheal cartilage and luminal narrowing at birth suggesting that the prenatal development of the lung may be perturbed (22) . Such anatomic abnormalities could further predispose to airway infection by influencing airflow and particle deposition patterns. Adult CF ferrets also develop a lung disease phenotype with similarities to human CF, including bacterial colonization (personal communication, John Engelhardt). Chronic lung diseases such as cystic fibrosis (CF), chronic bronchitis (CB) and asthma are characterized by a hypersecretion of mucus, which often exacerbates morbidity and hastens mortality. A characteristic of these diseases is that they are often accompanied by mucus stasis and enhanced susceptibility to infection and inflamma-tion. The molecular basis for the optimum mucus properties that ensure effective flow remain unexplained.

We have previously shown that there are at least 135 different proteins readily identifiable in respiratory mucus. Our observations revealed that mucus is comprised of an array of biomolecules ranging in molecular weight from around 6kD to 100MD. These may be split into two distinct groups, the first and major group being globular type proteins of MW between 6kD to 200kD with a diverse array of proposed functions. The second group is the mucins, which are large space-filling glycoconjugates typically of MW 200kD to 100 MD, most of this mass being carbohydrate in origin. Our studies have yielded evidence that at least 30 of these proteins are involved in distinct, very high molecular weight, multiprotein complexes centered around the gel forming mucins MUC5B and MUC5AC. These complexes constitute a discrete secretory entity we call the "mucin interactome." Based on these studies, we hypothesize that the key rheological functions of mucus do not emanate from the biophysical properties of the mucins themselves but from the dynamic interactions between mucins and globular proteins to form the functioning mucus gel.

To determine the composition of the mucin interactome we have undertaken a combination of biochemical methods such as S1000 gel chromatography, CsCl density gradient centrifugation, co-sedimentation and coimmunoprecipitation experiments combined with LC-MS/MS mass spectrometry. Proteomics analysis of these steps revealed that the gel forming mucins MUC5B and MUC5AC and membrane tethered mucins MUC1, 4 and 16 are predominant but associated with at least 30 other proteins. The number of the mucin interacting proteins identified was three times lower after using chaotropic agents and/or detergents, indicating that these complexes are affected by such agents. Our rheological experiments indicated an extreme effect of Triton and different dilutions of GuHCl on the reduction of viscosity and abolishing the network viscoelasticity. Our data revealed a number of protein candidates that resist dissociation from the mucin-interactome of HTBE secretions after such treatments such as LPLUNC1, glutathione S-trans, DMBT1, defensin alpha, surfactant-associated protein B, trefoil factor, complement C3, PLUNC, polymeric-IG receptor, gelsolin, WAP four-disulfide protein2. While the range of specific protein functionalities can be viewed as surprisingly large, e.g. antimicrobial, anti-protease, anti-oxidant and wound healing, we argue that they truly fall into just one general category, i.e., innate host defense. In other words, the mucus gel is part and parcel of innate immunity.

These binding partners of mucins may have a function in modulating mucin/mucus conformation and interactions. On the other hand, just because protein components are attached to the mucins does not necessarily mean that they will modify the rheology in a significant way. In many instances, we might intuit that the mucins carry protective proteins that do not interfere directly in the mechanical properties of the mucus. Therefore, we surmise that the mucin interactome in the airways can be functionally separated into two main roles: 1. Mucin/mucus structure and organization: Interacting proteins may play a role in the mucin structure and organization for effective mucociliary clearance and increased stability of the barrier function. 2. Increased fracture rates in adults with CF are wellrecognized(1). Vertebral and rib fractures are particularly concerning in CF as these fractures can compromise pulmonary and nutritional status and impact quality of life. While reported fracture rates in children are not consistently increased(2-3), increased fracture risk in adults may find its roots in childhood. Thus, preserving bone health and identifying individuals with increased fracture risk are important elements in the care of children and adults with CF.

Bone mineral density (BMD), measured by dual energy X-ray absorptiometry (DXA), is an established method of osteoporosis assessment in adults(4) although interpretation of DXA is less well-established in children. Moreover, in CF, the relationship between DXA measures and fracture risk are not established although suspected. In considering DXA, a review of the International Society for Clinical Densitometry Position Statement is worthwhile (5) . In children, osteoporosis requires the presence of both low BMD-Z<-2 (low bone mineral content/density for chronologic age) and clinically significant fracture (vertebral, lower extremity long bone, or 2 upper extremity long bone fractures), as in premenopausal women and men<age 50y. For the child/adolescent with diseases that threaten skeletal and nutritional health, spine and whole body DXA is recommended and measures of lean body and fat mass should be considered. Assessing bone mineral content (BMC) relative to muscle mass (partly influenced by height) may improve our understanding of bone health (6) . Because bone size highly influences DXA, BMC/BMD should be adjusted for height or height age in children with short stature or delayed puberty. How one implements this recommendation, however, remains a challenge.

Both normal and compromised BMC have been reported in the pediatric and adult CF populations (7) (8) (9) (10) (11) . Bone mineral accrual during puberty appears compromised in CF. Short stature and delayed puberty, including delayed and decreased peak height velocity, are common in CF (12) (13) , and DXA-BMD may over-estimate bone mineral deficits. CFF guidelines recommend obtaining DXA in adults and in children age>8y if <90% ideal body weight, FEV1<50%predicted, prolonged/frequent systemic glucocorticoid use, delayed puberty, or history of fracture (14) . Studies examining how often to obtain DXA are lacking, but obtaining repeat studies 6-12 months after initiation of a bone related therapy may be appropriate.

Vertebral compression fractures are frequently asymptomatic. In other cases, acute pain can develop and compromise respiratory status. Compression fractures increase the risk of additional compression fractures. Progressive kyphosis can limit vital capacity and compromise abdominal size, leading to decreased appetite. Thus, lateral spine films may be useful for identifying asymptomatic vertebral compression fractures in individuals with decreased BMD-Z.

Because DXA provides an "areal" measure of BMD, other tools directed at measuring bone qualities may be useful in CF. Peripheral quantitative computed tomography provides a three-dimensional measure of bone density and provides data on not only cortical and trabecular bone but bone geometry and strength. Reference data for children are not yet available (15) .

Numerous factors threaten bone health in CF. Laboratory evaluation should be directed at identifying potential mediators upon which to intervene: 25OHD, iPTH, phosphorus, zinc, PIVKA, urine calcium/phosphorus/creatinine (hypercalciuria and tubular phosphate reabsorption). To screen for exogenously induced adrenal insufficiency measure AM ACTH and cortisol. In males consider measuring testosterone; females with irregular menses/amenorrhea: iLH, FSH, prolactin. Urine ntelopeptides may help guide therapy choice.

Two categories of therapeutic agents are available for the treatment of osteoporosis: antiresorptive and anabol-ic. Bisphosphonates have primarily been used in CF (16) (17) , but newer agents will also be reviewed.

Bisphosphonates are antiresorptive agents that improve bone mineral density through their affinity for bone tissue where they concentrate at sites of resorption. Bisphosphonates bind directly to bone mineral, are engulfed by osteoclasts, impair osteoclast resorptive function, and ultimately suppress bone turnover markers. Because each agent varies by affinity for bone binding and degree of osteoclast inhibition, no two agents share the same properties. Bisphosphonates are available in oral and intravenous formulations and, depending upon the agent, are prescribed daily, weekly, monthly, quarterly or yearly. Bisphosphonates are not FDA approved for use in children, but a growing body of literature supports their use for pediatric metabolic bone disorders.

Potential side effects may limit their use in CF. In individuals with gastro-esophageal reflux, esophageal erosions have been described with use of oral bisphosphonates. To decrease this risk, oral bisphosphonates are taken with a full glass of water and the patient upright for at least 30 min. Additionally, oral bisphosphonates should be taken on an empty stomach to increase absorption. Hypocalcemia is more commonly described in patients using intravenous formulations of bisphosphonates and may occur up to several days after the completion of the infusion. Prior to using intravenous bisphosphonates, 25OHD and calcium status should be assessed and deficiencies corrected through supplementation. After the first intravenous infusion, patients may experience flu-like symptoms, bone, joint, and muscle pain; non-steroidal anti-inflammatory agents may provide relief. Symptoms tend to decrease with subsequent infusions.

The incidence of osteonecrosis of the jaw, defined as exposed bone in the maxillofacial area with delayed healing for more than eight weeks despite appropriate care (18) , is increased in the setting of bisphosphonate therapy (estimated incidence 0.7/100,000 patient treatment years (18) . Active malignancy, concurrent treatment with glucocorticoids, extensive and invasive dental work done before treatment, and severe dental and gingival disease increase the risk.

In 2005, a number of case reports describing unusual, low-energy, subtrochanteric femoral fractures in patients taking the oral bisphosphonate alendronate for 5-10 years (19) raised the concern that long-term bisphosphonate therapy may compromise bone integrity. Currently, conclusive evidence directly linking long-term bisphosphonate use to these fractures is lacking and insufficient to warrant discontinuing these drugs in osteoporotic patients (20) . However, individuals appear to derive diminished benefit from bisphosphonate use for more than 5 years (21) , and ongoing dialogue discussing risks and benefits of continued therapy is important.

Teriparatide is the only anabolic agent FDAapproved for treatment of postmenopausal women, men with primary or hypogonadal osteoporosis, and osteoporosis associated with sustained systemic glucocorticoid therapy, at high risk for fracture. While chronically elevated levels of PTH lead to osteoclast activation and reduced bone mineral density, intermittent PTH dosing enhances bone modeling via induction of bone lining cells to differentiate into mature osteoblasts and a direct interaction with osteoblasts that leads to decreased cell death. It is administered as a daily subcutaneous injection and is approved for a 2 year treatment period. Potential side effects include transient hypercalcemia, hypercalciuria, and hyperuricemia. In animal studies, concern for osteosarcoma development in individuals with open epiphyses has been raised, and this drug currently has a FDA black box warning against use in children.

Denosumab was FDA approved in 2010 for treatment of postmenopausal women with osteoporosis at high risk for fracture and patients who have failed or are intolerant to other osteoporosis therapy. It is a human monoclonal antibody to RANK ligand and inhibits osteoclast activation. It is administered via subcutaneous injection every 6 months for a period of 3 years. Prior to initiation of denosumab, vitamin D and calcium repletion are recommended. Side effects include hypocalcemia, skin irritation/infection at injection site, and pancreatitis. Increases in BMD during treatment are lost within 12 months of therapy discontinuation.

Compromised bone health in CF is common, but the best strategies for prevention, screening, and management remain to be defined.

Kristina L. Penniston, Ph.D., R.D., C.D.

As the survival of patients with cystic fibrosis (CF) improves, the incidence of delayed complications, such as bone disease, increases. Although first described decades ago, bone disease in patients with CF has become more prevalent as patients age. Estimates of osteopenia and osteoporosis in young adults with CF range from 20-40%. Fracture incidence and rates of kyphosis are higher than in the non-CF population and tend to occur at a younger age. The pathophysiology of bone disease in CF is multifactorial, and nutritional status alone may not correctly predict fracture risk. Low bone mineral density is correlated with severity of lung disease, yet it is not uncommon in adult CF patients with normal lung function. Peak bone mass is often not achieved in patients with CF, and this can be attributed to multiple factors, including nutrition. Moreover, studies show higher rates of bone loss in adults with CF as compared with age-matched controls. These observations combine to place individuals with CF at high risk for low bone mineral density.

Bone health is important as it is an important contributor to mobility and therefore a necessary component of a healthy and active life. Bone health is an effector of quality of life, with bone disease associated with multiple decrements, including those related to self-esteem, activity levels, and depression. As a growing number of adult patients with CF are considered for lung transplantation, a life-saving treatment for many, severe bone disease is an exclusion factor.

Proficiency in identifying and managing CF-related bone disease and its sequelae is required of today's CF health care team member, including non-dietitians. In addition to vitamin D deficiency, major diet-or nutrientrelated effectors of CF-related bone disease include: (i) malnutrition, general or protein-calorie; (ii) deficiency of micronutrients important for bone, e.g., vitamin K, magnesium, potassium; (iii) negative calcium balance from malabsorption, corticosteroids or both; (iv) caffeine and alcohol; and (v) dysbiosis secondary to malabsorption, chronic antibiotic use, or both. In addition, inflammation secondary to recurrent infection and sarcopenia secondary to malnutrition, both of which are mediated by diet, are linked with bone disease.

Early prevention strategies and interventions to correct risk factors are important, and all members of the CF health care team should be involved. Targeted food-and supplement-based therapies, in combination with pharmacologic therapy if needed, can successfully alter the natural course of CF-related bone disease.

Anne K. Swisher, Ph.D., P.T., C.C.S.

Numerous studies have shown that people with cystic fibrosis (CF) have low bone density compared to their peers. This is true both in childhood bone accrual and adult bone maintenance. This low bone density is associated with many factors, including chronic inflammation, low body weight, low levels of physical activity and CF-related diabetes (CFRD). Lack of healthy bone is concerning, as risk of fragility fractures increase in the CF population. Fractures can lead to pain and impaired posture, particularly fractures of the vertebral spine. This impaired posture, in turn, can compromise breathing. Fractures and their sequellae can lead to limitations in ability to participate in important life functions and declines in quality of life for these patients.

Exercise and physical activity are important ways to improve and maintain bone health in people with CF at any age. Exercise is thought to work through mechanical forces of adequate ground reaction force, combined with appro-priate rest intervals, and of increasing stimulus over time in order to cause a positive balance in bone deposition and/or resist resorption. Recent studies regarding chronic inflammation in CF also suggest that the anti-inflammatory effects of regular physical activity and exercise may help fight bone loss through other mechanisms than mechanical forces. Recently, it has been shown that both CFTR and vitamin D receptors are present in skeletal muscle; therefore, muscle contraction may affect, or be affected by these proteins.

Bone-building exercise is typically considered to involve high-velocity impact, which often involves activities such as jumping. In addition, resistance training imparts a strain to bone through tendinous attachment which stimulates a site-specific response. Postural training, particularly of the thoracic spine and ribcage, are especially important in CF, as this is a site of risk for fragility fracture. Aerobic conditioning, typically using weight-bearing activity, can also stimulate bone retention. Recently, peak aerobic capacity was shown to be the strongest predictor of bone mineral density in patients with CF. Thus, improving overall exercise capacity may additionally benefit bone density.

However, the art of exercise prescription for bone health comes into play when determining the most appropriate exercise for a specific patient in his or her specific circumstances. This presentation will discuss principles of bone health exercise for patients of different ages (infancy, pre-pubertal, adolescence, adulthood, and aging with CF) as well as adapting for severity of lung disease and other situations such as the pregnant patient with CF, or the patient who already has osteoporosis. In the end, the exercise must be the "best" exercise that the patient is willing and likely to perform.

Melenie A. Meyers, R.N.

Cystic fibrosis (CF) affects 30,000 children and adults in the United States (70,000 worldwide). More than 45% are adults with a median age in the mid 30s due to a progressive increase in life expectancy over the past three decades.

CF bone disease, i.e., osteopenia, osteoporosis and increased risk of bone fractures, is a common problem occurring in 50-75% of adult CF patients that can readily be assessed by measuring bone density with dual energy x-ray absorptiometry (DEXA).

Based on CF Foundation Registry data, osteoporosis occurs in 2.6% of adults and 0.2% of children with CF, and osteopenia occurs in 3.3% of adults and 0.2% of children with CF. These estimates are likely low since screening DEXA scans are not done routinely in all patients. Thus, while CF bone disease is clearly a problem in adults, it is also important in pediatric CF patients.

The importance of CF bone disease should not be underestimated since it is thought to increase mortality by 5-10%. The clinical consequences of CF bone disease include pain due to cough-induced rib fractures, which can lead to impaired respiratory function from reduced ventilation and cough, reluctance to perform airway clearance and reduced exercise. Others include kyphosis, delayed bone healing, digestive problems, transplant viability, and psychosocial issues.

Education is a cornerstone of preventing CF-related bone disease but requires addressing the myriad risk fac-tors affecting bone health. Education is especially important in pediatric programs since bone health in later life is established at an early age. Puberty is critical for bone development and bone mass peaks by about age 30 and serves as a "bone reservoir" for the remainder of life. Improved comprehension of fracture prevention, monitoring practices, testing algorithms, and features of CF predisposition to low bone mineral density are all important to education programs.

Learning to fold into our practice, monitoring, screening and education of the multitude of CF complications is a challenge. As health care providers, we have the responsibility of educating ourselves, staff, patients and families about CF and its many complications.

Education, prevention, early recognition and treatment are the most effective strategies for sustaining bone health to help maintain the quality of life of many individuals with CF.

Utilizing "best practices" methods, screening and educational tools that deliver a consistent message that can be shared across the many CF Centers will help to improve CF care and outcomes. 

Nico Derichs, M.D.

Defective CFTR-mediated chloride transport is a prominent feature in cystic fibrosis (CF). The CFTR basic defect in native human CF tissues manifests itself as absent or reduced function of the CFTR chloride channel. Modulation of ion transport is an attractive strategy for treatment of CF. The rationale is that modifying the ion transport properties of CF epithelial cells towards normal would be of clinical benefit. Having good pre-clinical bio-assays to prioritize ion channelmodulating CF drugs is important in streamlining the CF drug pipeline in selecting the very best drug candidates to move forward in clinical trials. At present the primary endpoint in pre-clinical evaluation of candidate drugs is electrophysiological responses in well-differentiated primary cultures of human bronchial epithelial cells from CF subjects. There are caveats in the use of this model, however, as e.g. it requires cell growth in culture for at least 3 weeks, where phenotype changes can occur and the model might not be representative of the in vivo condition. Moreover, prediction of clinical drug effect by in vitro data has been shown to be difficult for some CFTR modulating compounds in clinical trials. This leads to the development and optimization of alternative pre-clinical surrogate assays of epithelial cell function in the airways and intestine, two important and accessible CF tissues.

Intestinal current measurement (ICM) is a relatively simple ex vivo test which involves native human rectal biopsies. ICM has been used for genotype-phenotype studies (1-3) and for difficult CF diagnosis (4-7). In recent years, it was further developed for preclinical drug development in CF. Similar studies can be performed with ex vivo native bronchial tissue from human lung explants (8) . There are at present no data from clinical trials on changes in ICM induced by disease modifying drugs, however substantial experience has been gained in preclinical human ex vivo corrector studies (9) . ICM may be useful in phase IIa and b clinical trials in adult CF patients, CF children and infants. Preliminary data on intra-individual variability confirm the promising role of ICM as outcome parameter in CF therapeutic studies (Derichs N, unpublished) , although more information on repeatability is needed. Another limitation of the ICM is the very low number of centers with expertise. In addition, the short viability of the rectal biopsies impairs transport to a central laboratory for analysis. ICM has at present most application in preclinical drug testing of potentiators and correctors, since the translation of these CFTR modulators from in vitro efficacy to in vivo clini-cal trials can be optimized by ex vivo incubation studies on native human tissues. Under appropriate incubation conditions intestinal biopsies will remain viable and functional for at least 24h, making the testing of slowacting CFTR correctors feasible (8) . In preclinical ICM corrector studies on F508del homozygous CF patients, defective CFTR-mediated chloride secretion could be corrected to levels of 25% of healthy control after 24 hour ex vivo incubation with corrector-29 (10). However, as is true for NPD, the minimal clinically relevant improvement in ICM is still unknown.

In summary, recent experiences support the utility of ex vivo assays like ICM in native rectal tissue for CF drug development. Preclinical testing of candidate CF drugs can help to prioritize and optimize compounds for translation into clinical trials.

References CF is caused by mutations in the gene encoding the CF transmembrane conductance regulator (CFTR), an epithelial anion channel involved in chloride and bicarbonate transport. Absent or defective CFTR results in abnormal ion transport in the airways, sinuses, sweat duct, gastrointestinal tract, hepatobiliary tract, and vas deferens. In the lung, mucus stasis ensues, resulting in chronic infection and end organ dysfunction, and is a common cause of morbidity and early mortality. Although there are over 1500 disease-causing mutations in CFTR, they can be categorized into several major classes based on the molecular pathogenesis induced by each mutation (1) . CFTR mutations can also be generally grouped by severity of the anion transport abnormality conferred by the mutation. While numerous factors alter the genotype-phenotype correlation in CF, including environmental factors and genetic modifiers to CFTR activity, the severity of the ion transport defect is also correlated with long-term outcome (2) . As such, measurement of CFTR activity in CF subjects can confer diagnostic and prognostic information, suggesting the importance of accurately measuring CFTR activity. As restoring CFTR function through small molecule pharmaceuticals is one therapeutic approach that has shown considerable promise for the treatment of CF (3), accurate and reproducible measurements of CFTR activity is also essential to evaluating the efficacy of these CFTR modulators in development.

The traditional measurement of CFTR activity involves measurement of sweat chloride concentration following stimulation of sweating with the cholinergic agonist pilocarpine, and has long been used as a diagnostic marker in CF. Standardization of the method using the Macroduct collection system has facilitated use of sweat chloride analysis in the context of clinical trials, and also allows analysis in a central laboratory blinded to treatment assignment (4). Use of sweat chloride analysis for the evaluation of the efficacy of the CFTR potentiator VX-770 and the CFTR F508del processing corrector VX-809 has been highly informative. As predicted by genotype-phenotype correlations in CF, changes in sweat chloride following modulation of CFTR have been highly sensitive to small changes in CFTR activity, and have allowed relative efficacy to be determined with relatively simple study designs and small numbers of study subjects (3) . Emerging methods of sweat testing, including sweat rate analysis by evaporimetry or quantitation of the sweat output from individual glands, can also be used to quantify the function of sweat glands (as opposed to the sweat duct), and may provide additive diagnostic information regarding the impact of CFTR modulation on glandular epithelium. Whether this method confers additional information in the context of CFTR modulation remains to be determined.

Unlike sweat chloride analysis, nasal potential difference (NPD) measurements allow the isolation of sodium and chloride transport, providing a means to independently analyze CFTR activity, ENaC function, and ion transport induced by alternative chloride channels. The technique is performed in vivo, and can also be used to evaluate proof-of-concept studies involving local (e.g., nasal) administration of biologic agents, including pharmaceuticals and gene therapy vectors. NPD has been sensitive to the detection of improved chloride transport induced by various CFTR modulators, including premature termination codon suppressors, CFTR potentiators, and gene therapy agents, and also has been sensitive to agents that inhibit ENaC activity (4) . Although the measurement requires specialized equipment and training to conduct appropriately, recent advances in methodology have improved the quality of tracings, provided a means for central interpretation and analysis, and greatly facilitated ease of use, providing a means to use NPD in clinical trials involving a large number of operators and participants (5) . The technique can also be adapted to measure CFTR activity in the lung, however potential difference measurements do require a greater number of subjects to demonstrate a detectable difference in CFTR activity compared to sweat chloride analysis, particularly when evaluating an agent with marginal activity. Similar to NPD, intestinal current measurements (ICM) provide a means to isolate sodium and chloride transport. An added utility of the measure is that it provides a means to test the activity of ion transport modulators ex vivo on full-thickness human tissues. Recent standardization of ICM measurement technique may provide a means to use the method in multicenter trials.

As data regarding the modulation of CFTR activity involving a number of different treatment strategies continue to emerge from long-term trials that also measure clinical outcome, use of CFTR activity measures may become useful as a surrogate outcome measure in CF.

Scott D. Sagel, M.D.

Cystic fibrosis (CF) lung disease is characterized by a self-perpetuating cycle of airway obstruction, chronic bacterial infection, and vigorous inflammation that results in bronchiectasis, progressive obstructive lung disease, and marked shortening of life expectancy. Since airway inflammation plays a central role in the progression of CF lung disease, inflammatory biomarkers that can be used to monitor disease activity and evaluate response to therapy would be extremely valuable. Many inflammatory cells, mediators, and enzymes are involved in the complex pathophysiology of CF lung disease, so that there are many possible biomarkers to study and there is a high degree of redundancy (1, 2) . Informative biomarkers may be identified through discovery-based proteomic or metabolomic approaches or through targeted measurement of purported candidates (cytokines, proteases, oxidants, matrix components, tissue breakdown products).

The ideal biomarker would be clinically and biologically relevant, reproducible, sensitive and specific to treatment effects, and feasible (3) . At present, no single inflammatory biomarker clearly meets all of these criteria. Individual biomarkers will be appropriate in trials of targeted anti-inflammatory therapeutic agents such as a specific anti-cytokine or anti-protease drugs. However, for clinical trials of anti-inflammatory treatments that target multiple pathways of the inflammatory response or for therapeutic agents aimed at correcting cystic fibrosis transmembrane conductance regulator (CFTR) protein function or restoring airway surface liquid, a combination of biomarkers will likely prove more valuable. This symposium presentation will review the current state of knowledge of biomarkers of inflammation relevant to CF lung disease, and the tools to measure inflammation. This presentation will critically examine the existing data supporting the validity of pulmonary biomarkers in CF, with an eye towards their application as surrogate endpoints or outcome measures in CF clinical trials.

Due to the localized nature of enhanced inflammation in CF, respiratory secretions from the lower airway are the optimal source of inflammatory cells and soluble mediators, and can be obtained directly or indirectly from the airways. The most common direct method is by performing bronchoscopy with bronchoalveolar lavage (BAL) or bronchial washing. Much of what has been learned about infection and inflammation in the CF lung has been through the use of BAL and much of the experience evaluating the "biologic efficacy" of therapeutics in CF has been through the use of BAL (reviewed in reference (2)). BAL offers the advantages of being able to be performed in any age group, and for its ability to sample the same area of the lung for studies that require repeat sampling (as is commonly done with intervention studies). Disadvantages include the need for an experienced team to perform the procedure, the need for sedation to perform the procedure, the inability to conveniently perform repeat sampling more than once or twice throughout an intervention study, and the high costs associated with performing the procedure. As a result of these drawbacks, investigators have relied on spontaneously expectorated sputum for many years, but this requires patients to have enough lung involvement and resultant respiratory secretions that allow regular sputum production. Thus, it is not very useful for examining inflammation in early lung disease, and in young children regardless of disease severity, since young children have difficulty expectorating sputum. To overcome the obstacle of obtaining sputum from subjects with milder lung disease, inducing a sputum specimen with hypertonic saline has been used to obtain lower airway secretions, and much effort has been expended in recent years in an attempt to validate the methodology of obtaining, processing, and assaying induced sputum (reviewed in detail in reference (2)). However, sputum induction is difficult to perform in patients age 8 years and younger, and multicenter studies have largely limited inclusion to those 10 years of age or greater. Additionally, biomarkers of inflammation in both induced and expectorated sputum show considerable variability, especially in comparison to BAL.

Other ways of indirectly assessing markers of inflammation in the lower airways have been considered, including exhaled gases and breath condensate, but they too have their limitations. Exhaled NO (eNO) is reduced in CF (4); therefore, consideration could be given to measuring eNO in exhaled breath in a study of an anti-inflammatory drug that might impact the pathway leading to the upregulation of NO. While investigators have measured some inflammatory mediators in exhaled breath condensate, the usefulness of these measures as outcomes remains to be defined as well.

A systemic marker of lung inflammation would be ideal. Blood and urine can be obtained from subjects of any age and disease severity, and may reflect the status of inflammation throughout the lung, rather than one segment as is assessed by BAL, or heterogeneous segments as is assessed with sputum. However, a systemic marker may not be sensitive enough to detect a meaningful change in lung disease, given that the inflammatory response to infection in the CF lung is largely confined to the lung.

In CF, we need more information about how pulmonary biomarkers relate to lung disease development, severity, and progression, and how they change in response to treatments. In addition, there is a relative paucity of data of how pulmonary biomarkers may be affected by concurrent therapies, or how they relate to important clinical outcomes in CF, including rate of decline in FEV1, pulmonary exacerbation frequency, and mortality. In the future, pulmonary biomarkers will likely be useful in predicting disease progression, indicating the onset and resolution of a pulmonary exacerbation, and assessing response to current therapies or candidate therapeutics.

Supported by the CF Foundation (SAGEL07A0) and the National Institutes of Health (U54 HL096458-07).

Christiane De Boeck, M.D., Ph.D.

In CF lung disease, the course of FEV1 relates to long-term outcome. However, using FEV1 as the primary outcome parameter in phase 3 trials has become difficult, since the mean year to year change in FEV1 is low. Apart from adolescence, during which it is higher, the year to year change is only in the order of 1% (1). Even though the mean change is low, the loss of lung function is not homogeneously distributed between patients: 15-20% of children experience a year to year change of 10% or more and 30-40% have a year to year change > 5%. One could focus on subjects at higher risk of loss of lung function but then the study would lose in terms of generalizability: the findings may not be applicable to subjects with a more stable disease course. When a large treatment effect is anticipated, the use of FEV1 as primary outcome remains a good option e.g., such as proven in the evaluation of VX-770.

The use of pulmonary exacerbations as outcome in CF clinical trials is also not obvious. We do not have a universal definition for this item. Pulmonary exacerbations requiring treatment with intravenous antibiotics are a major driver of lung disease progression, but they are infrequent in young and/or rather healthy subjects with CF (2). Less stringent definitions of pulmonary exacerbations have been used or are under evaluation but it is less certain how these relate to long-term loss of lung function.

The need for new surrogate outcome parameters is thus obvious: the multiple breath inert gas washout test (MBW) and chest CT are prime candidates, especially in children or subjects with mild lung disease.

CF lung disease starts in the small airways and FEV1 is insensitive to changes in the small airways. MBW is non-invasive and provides information about this "silent" lung zone where disease starts (3, 4) . The result of MBW, a test of gas mixing efficiency, is usually reported as the lung clearance index (LCI) i.e., the number of "lung turnovers" (FRCs) required to clear the lungs of an inert gas to 1/40th of the starting concentration. LCI is more sensitive than FEV1 (4). LCI has been shown to be responsive to interventions such as treatment with intravenous antibiotics (5), hypertonic saline (6) and rhD-NAse (7) . Recent data link early LCI abnormalities to longer term lung disease course: an abnormal LCI at preschool age predicted an abnormal FEV1 at school age (8) . Several issues remain to be resolved: which MBW apparatus is best, which inert gas is most appropriate, are results obtained by different techniques comparable, what is the test repeatability, what constitutes a significant change. Preliminary data are promising: given the increased sensitivity of LCI and depending on the research question, power calculations come up with numbers around 50 children in each study arm (6) . This compares favorably to using FEV1 as outcome measure, which will require hundreds of children in each treatment arm (9) .

The second outcome measure with the potential to solve the current problem of loss of sensitivity is scoring of chest CT. Several studies have proven this parameter's sensitivity. At a mean age of 4 months, 84% of infants with CF detected after newborn screening are asymptomatic but 81% have an abnormal CT (10) . At the age of 5 years, 7 of 11 subjects already have bronchiectasis (11) . Bronchiectasis scoring on chest CT correlates well with measurement of LCI (3, 12) . Chest CT has the disadvantage of a radiation burden but it has the advantage of objectively quantitating structural CF lung disease (13) . Again, using this parameter as an outcome will offer the advantage of feasibility in terms of study size. Depending on the hypothesis put forward, the sample size calculations for use of chest CT scoring in longterm studies result in about 50 subjects per group (13) .

CFTR dysfunction impedes chloride transport to the airway lumen. Together with an increased sodium reabsorption, caused by the loss of CFTR inhibition on ENaC function, this leads to dehydration of the airway surface lining (ASL) fluid. It follows that the mucociliary clearance is greatly reduced, because the cilia are unable to function properly. This theory is challenged. Measurement of the ASL depth and the ASL and mucociliary layer viscosity is possible with phase contrast X-ray imaging (14) or confocal fluorescence photobleaching (15) . Also measuring airway liquid absorption and mucociliary clearance by nuclear medicine techniques are interesting biomarkers close to the fundamental defect of CF lung disease (16) . These measurements of mucociliary composition and clearance are however still in the exploratory phase and not yet appropriate for Phase 3 clinical trials.

Last but not least, we are eventually entering the era of disease modifying drugs. These treatments with CFTR potentiators and correctors and nonsense mutation readthrough probably have the highest potential in young children with relatively healthy lungs. Paradoxically showing improvement in this population is extremely difficult. Several gating mutations reported are very rare and cannot be studied individually. Therefore we should dare to think out of the box. Providing proof of efficacy in a Phase 3 trial in a small population is impossible. When studying CFTR correctors and potentiators in (pre)school children with a very slow disease progression or in a very limited population, it is appropriate to use a biomarker closely linked to the causal pathway of the disease as primary outcome. CFTR bio-assays such as sweat chloride and nasal potential difference measurement or intestinal current measurements are good candidates (17) . An alternative is to only collect safety data especially if efficacy has already been proven in another age category. Proof of clinical benefit can then follow during pharmacovigilance via Phase 4 trials. The results of this 24 week, 352 patient double-blind placebo-controlled protocol in CF patients of any genotype ≥5 years old with FEV1 ≥75% predicted were positive, in that the pre-specified primary FEV1 endpoint was met, and a 24 week open-label extension seemed to confirm this lung function effect, suggesting, in the authors' heady words, "denufosol has efficacy and safety profiles suitable for early intervention in cystic fibrosis lung disease." The developers of denufosol, Inspire Pharmaceutials, and the CF community looked forward to confirmation in the FDA-required second Phase 3 trial (TIGER-2) followed swiftly by submission of a New Drug Application to the FDA for approval. Instead, less than 3 weeks after publication of the TIGER-1 paper, Inspire announced that the 466 patient, 48 week placebo-controlled TIGER-2 trial did not even come close to meeting any of its endpoints (2). Inspire stock tanked; the company lost $400m in one day (3) . It promptly shelved the program, restructured and laid off employees; by mid-May its remaining ophthamologic franchise was acquired by Merck (4). Inspire expired. What happened?

Denufosol is a third-generation nucleotide P2Y2 agonist derived from UTP with increased stability on the respiratory epithelial mucosal surface (t 1/2~3 hr). Much elegant in vitro work has established that denufosol increases rates of PIP2 hydrolysis, increasing IP3 and release of Ca 2+ intracellularly; this increases chloride secretion through calcium-activated chloride channels (CACC), stimulates ciliary beat frequency, increases mucin secretion from goblet cells and surfactant from type II alveolar cells, and inhibits sodium absorption via the epithelial sodium channel, resulting in improved airway hydration and increased mucociliary transport (5, 6) . Because it is hydrolyzed by endogenous nucleotidases at the epithelial surface and rapidly metabolized after IV infusion there is virtually no systemic exposure and thus little chance for systemic toxicity, as verified in preclinical animal toxicity studies and later in the clinical trials. Denuofosol was formu-lated to be delivered by conventional jet nebulization for clinical trials.

Granting retrospective advantage, it does appear hints of impending disaster can be found in the development program results, but were ignored or misinterpreted by both a commercial enterprise and diverse stakeholder community eager for success for both unique and common reasons. By focusing on these at best ambiguous if not frankly negative harbingers, it is possible that we may learn not to repeat the mistakes of the past.

1. Mechanistic Rationale. The results of TIGER-2 suggest that the preclinical assessment of denufosol's effect on CF respiratory epithelium, clear in so many excellent in vitro experiments, may be flawed. The key finding by researchers at UNC, after the clinical program was well under way, that rate of change-dependent shear stress releases endogenous luminal ATP, thereby activating the CACC to secrete chloride, opens the possibility that this channel may be more physiologically active in CF than previously anticipated, and thus less clinically augmented by pharmacologic stimulation (7) . Might exercise, airway clearance regimes or even activities of daily living already be doing the CFTR-bypass work it was supposed denufosol might be needed to accomplish? Now that the molecular identity of at least one CACC has been discovered as TMEM16A it may be possible to evaluate this with more precision (8) . The clinical effect of inhaled osmotic agents such as hypertonic saline and mannitol points to the viability of strategies to improve airway surface hydration. But the fate of other investigational drugs activating CACC and/or inhibiting ENaC may be affected by further research in this area.

2. Phase 1. The determination of optimal dose and interval are crucial in clinical drug development. The first trials of denufosol showed single-dose tolerability at up to 200 mg in non-CF adult nonsmokers. In adult and pediatric CF patients Phase 1 highest tested dose was 60 mg as a single dose or in b.i.d. dosing for 5 days, and since this was well-tolerated this dose was carried forward (9) . Although there was somewhat more wheezing at 60 mg vs lower doses, could higher doses have been more effective with acceptable levels of intolerability? If acute increases in airway hydration were feared as potential cause of adverse events, strategies to deal with this via dose titration could have been tried. The only pharmacodynamic measure was change in expectorated sputum weight, but this is a fickle outcome that did not show clear dose-response, was only seen after single and not repeat doses, and seems an inappropriate biomarker given the decision to treat only patients with FEV1≥75% predicted in Phase 3, as many patients with mild disease do not expectorate.

3. Phase 2. The Phase 2 program for denufosol incorporated several crucial junctures in regime and target population. In the first protocol 89 patients ≥8 years old with FEV1≥75% were enrolled in a 28 day controlled trial in randomized parallel groups receiving 20, 40, 60 mg or placebo t.i.d. (10) . Patients taking inhaled antibiotics, hypertonic saline, or macrolides were excluded. Tolerability was good. With respect to efficacy, there was no dose-response in FEV1 change from baseline in actively treated patients. While the denufosol-treated pooled group change was significantly better than placebo, the effect in protocol completers was less than intentto-treat, and most importantly, the placebo group (n=21) experienced an unusually sharp decline in FEV1 (adjusted mean decline of 140 mL or ~2.5% over 28 days) and other lung function measures, changes ascribed to regression to the mean. In addition there were no differences in CFQ-R, HRCT scores, sputum weight, or exacerbations. Was the calculation that nominal doses ≥20 mg achieve initial airway surface concentrations exceeding maximally effective doses as calculated from nasal potential difference measurements, and thus predictive of no doseresponse over the trial dose range, correct (moreover, even if true, is initial concentration enough to count on)? Taken together, the results now seem suspect in indicating any efficacy, given the aberrant placebo decline in lung function, the questionable approach to pharmacodynamic dose-response, and the lack of concurrent changes in multiple clinically relevant outcome measures. Notably, the second Phase 2 trial enrolling 72 patients with FEV1 modified to 60-90% showed no efficacy and a smaller treatment effect in those with ≥75% FEV1 due to a more typical flat placebo response (11). 4. Phase 3. Based on Phase 2, TIGER-1 extended the approach of selecting patients with FEV1 ≥75% predicted, necessarily reducing magnitude of potential lung function treatment effect. TIGER-1 barely met its primary endpoint of a statistically significant treatment effect on FEV1 (L) at 24 weeks (45 mL, p=0.047). But ominous signs were many: the magnitude was small; the effect of growth on lung volume was not incorporated (nonsignificant trends were seen for FEV1 % predicted); no secondary endpoints (other spirometric measures, exacerbations, CFQ-R) were attained. The kinetics of response were puzzling for an active drug: both active and placebo FEV1 were down at 28 days and then rose slowly over time. On the other hand, post-hoc analyses of patients on minimal background therapy (0-2 drugs), of adolescents, and of FEV1 loss in patients with exacerbation appeared to suggest efficacy by showing trends to greater denufosol treatment effect than seen in the overall study population (1, 12, 13) . Based on TIGER-1 top line data, TIGER-2 was redesigned after launch to optimize chances for success by extending the placebo-controlled treatment period to 48 weeks, increasing sample size to boost power, and capping entry FEV1 at 110%. Despite all this, TIGER-2 results showed superimposable results for denufosol and placebo in all outcomes.

This sobering story shows that drugs likely to succeed for CF should undergo continued basic research on mechanism given our substantial ignorance of CF pathogenesis and pathophysiology; that we should confirm optimal dose (preferably by dose-dependent relevant clinical pharmacodynamics), interval and delivery, as well as safety, in Phase 1; that we should obtain robust exploratory efficacy endpoints on several congruent biomarker and/or clinical outcomes, with expected placebo responses as well as safety, in Phase 2; and that we should achieve multiple and clinically meaningful as well as statistically significant efficacy outcomes and safety in Phase 3. Short of this, the stakes are too high and the price too dear.

In the past five years a significant number of new therapies have been developed for the treatment of cystic fibrosis. A number of new inhaled antibiotics are available or are in Phase III clinical trials that will significantly help in controlling chronic infection and so hopefully maintaining lung function (1). How these therapies will be used together remains to be determined. Antiinflammatory therapies, such as those directed against human neutrophil elastase show some promise but are still in early phase clinical trials (2). More novel approaches aimed at regulating downstream inflammatory pathways in immune or epithelial cells and other aspects of innate immunity are currently being investigated but are also only in early phase studies (3) .

Advances in therapies for airways clearance also show considerable promise with inhaled mannitol (Bronchitol) in the final stages of clinical development. These studies indicate that Bronchitol significantly improves FEV1 which is sustained over 12 months and also reduces pulmonary exacerbations (4). Bronchitol is likely to be a useful adjunct to current therapies as it is equally effective in those on Pulmozyme as those not.

The most exciting progress in pipeline therapies is in the development of therapies which affect the basic defect of cystic fibrosis. These drugs include PTC124, VX-770 for G551D patients and VX-809 in combination with VX-770 for F508del mutations, and the Gene Therapy programme using a plasmid liposome construct.

PTC124 is designed to overcome stop mutations by causing readthrough in the translational step and so resulting in the manufacture of full length protein. Proof of concept studies have been undertaken and two of these have shown significant changes in nasal potential difference and some improvement in FEV1 (5, 6) . These results are encouraging and the pivotal phase 3 study has been completed and the results should be available in the very near future.

VX-770 is a small molecule which potentiates CFTR which is expressed on the cell surface (Class III mutations). This means it is likely to be useful across all Class III mutations but it has been particularly studied in patients with the most common, archetypal class III mutation, G551D. In a proof of concept study, VX-770 significantly reduced sweat chloride concentration and nasal potential difference with improvements in FEV1 (7). A pivotal Phase 3 study has been completed and demonstrates significant and sustained improvements in FEV1 by 10%. This is associated with significant improvements in CFQ-R respiratory domain, a significant reduction in pulmonary exacerbations, a significant weight gain and a reduction in sweat chloride concentrations. This demonstrates for the first time that a treatment that corrects the functional, basic defect in vitro translates through to significant improvement in relevant clinical measurements of lung function, symptoms and systemic measurements of CFTR function. VX-770 has also been studied in F508del homozygous patients and shows no significant clinical efficacy though there is a small improvement in sweat chloride concentrations. The most recent study examining the effect of the combination of VX-770 and VX-809 in F508del homozygous patients suggests both drugs have no major safety issues when used together and has demonstrated a modest, potentially meaningful reduction in sweat chloride concentrations (8) .

Gene therapy has the potential to overcome the functional defect in CF in all classes of mutations. The UK Gene Therapy consortium has recently completed a single dose, safety study combining an innovative plasmid that expresses CFTR for around 4 weeks and the liposome GL67. This will enter a multidose one year study in 2012. This will determine if this approach has therapeutic potential.

Current pipeline therapies for CF show considerable promise. The improvements in patients with at least one G551D mutation following VX-770 show that CFTR is a tractable target and improving CFTR function results in significant improvements in clinical status. Further results from the current, ambitious programmes of development of therapies which are aimed at treatment of the functional defect in CF are eagerly awaited. Mutations in the cystic fibrosis transmemebrane conductance regulator (CFTR) gene are the cause of cystic fibrosis (CF). The deletion of phenylalanine at position 508 (F508del) is the most common CF mutation, and at least one copy is found in approximately 90% of CF patients. As such, rescue of the F508del CFTR mutation using small molecule approaches remains an important goal stimulating drug discovery and development. Success in this endeavor could have a considerable impact on CF outcomes. The F508del mutation primarily induces misfolding of CFTR, resulting in premature degradation of the protein prior to reaching the cell surface. The mutation also confers abnormalities in channel gating, thereby severely reducing ion transport activity even when the mutant protein resides at the cell surface. Knowledge of the molecular biology of the F508del defect suggests approaches to "correct" CFTR misprocessing may need to be combined with agents that "potentiate" CFTR gating to confer a meaningful clinical improvement (1) .

In collaboration with the Cystic Fibrosis Foundation, Vertex Pharmaceuticals used high-throughput screening to evaluate hundreds of thousands of compounds in a cell-based screening platform designed to identify correctors of F508del processing and potentiators of CFTR activity (2) . VX-770, an investigational agent and robust potentiator of CFTR function, increased short-circuit current in G551D/F508del primary human bronchial epithelial cells (HBE) to approximately 50% of activity observed in non-CF HBE (3) . VX-770 also demonstrated in vitro activity in a subset of HBE cells derived from F508del CFTR homozygous individuals, suggesting that VX-770 can potentiate a small pool of F508del at the cell surface resulting in approximately 15% of non-CF HBE CFTR activity, prior to correction of F508del processing. VX-770 has completed Phase 2 testing and progressed through Phase 3 testing for the treatment of CF in individuals with the G551D gating mutation (4). Prolonged (48 week) treatment of CF subjects with at least one copy of G551D CFTR exhibited marked improvement in pulmonary function, a reduced chance of experiencing a CF pulmonary exacerbation, and a ~45 mEq improvement in sweat chloride, in addition to demonstrating an acceptable safety profile. In contrast, treatment of F508del homozygous CF patients with VX-770 was safe, but resulted in a small change in sweat chloride (~3 mEq) and did not alter clinical outcome; VX-770 treatment improved sweat chloride by > 10 mEq in only a small subset (~15%) of F508del homozygous individuals. These results suggest that while demonstrating detectable bioactivity in some individuals, treatment with VX-770 alone is not sufficient as a single therapy in F508del homozygous subjects, and indicate the importance of correcting F508del processing in combination with a CFTR potentiator as also suggested by in vitro studies.

Using cell-based strategies, Vertex Pharmaceuticals and other laboratories have also discovered agents that correct F508del CFTR processing, resulting in rescued F508del protein at the cell surface and partially restored anion transport (2, (5) (6) (7) . The lead corrector candidate in the Vertex program is the investigational agent VX-809, and in F508del homozygous HBE cells, VX-809 restored CFTR activity to~16% of non-CF HBE in vitro and was accompanied by improved band B to C formation, indicating the increased presence of CFTR residing at the cell-surface. In CF patients homozygous for F508del, sweat chloride improved 8 mEq over placebo (7 mEq overall) at the highest dose tested (200 mg daily) and demonstrated a reasonable safety profile. Unfortunately, 4 weeks of VX-809 monotherapy was not associated with a detectable clinical improvement in spirometry, suggesting bioactivity of VX-809 is not sufficient to improve clinical outcomes when used alone. When VX-770 is added to VX-809 in vitro, anion transport activity doubles in F508del HBE cells, suggesting an additive effect when a potentiator is combined with a CFTR corrector. Interim results of the first CF clinical trial assessing the benefit of a corrector and potentiator in combination were recently reported. In this trial, two weeks of 200 mg VX-809 treatment was followed by the addition 150 mg or 250 mg of VX-770 twice daily for 7 days. Results demonstrated VX-770 further decreased sweat chloride compared to VX-809 alone (by 9 mEq above VX-809 treatment alone at the VX-770 250 mg dose group and 13 mEq (total change) following the combination of VX-809 and VX-770). These findings demonstrate the proof of concept that significant improvements in sweat chloride are possible upon combined treatment of F508del homozygous individuals with a CFTR potentiator and a CFTR corrector. Based on genotype-phenotype correlations in the disease, 13 mEq may be sufficient to confer improved clinical outcomes (8) . Further testing of combination strategies involving CFTR correctors and potentiators are warranted in CF, and may confer clinical improvement upon long-term testing in the most common mutation found in CF patients.

Arnold L. Smith, M.D. and Bonnie W. Ramsey, M.D.

In the late 1970s, intravenous (IV) aminoglycosides, and particularly tobramycin, became an important part of the regimen for the management of CF pulmonary exacerbations. Aminoglycosides had much broader therapeutic indices than the polymyxins, while retaining excellent antipseudomonal activities. While studying the pharmacokinetics of IV tobramycin, we observed that tobramycin clearance in persons with CF was increased compared to healthy volunteers (1). Although aminoglycosides were known to be cleared from the systemic circulation in the urine, the glomerular filtration rate in persons with CF dramatically underpredicted their rate of aminoglycoside elimination (1). We speculated that the increased non-renal systemic clearance could be due to expectoration of the antibiotic in sputum. In order to improve aminoglycoside quantitation over the existing bioassay in complex matrices, such as sputum, we had developed radio-enzymatic assays (2) and were surprised to learn that levels of tobramycin in CF sputum during IV treatment were almost 10 times higher when assessed by RAE than by bioassay (3). Our observations suggested that clearance of (bound and inhibited) tobramycin by sputum expectoration could account for the accelerated tobramycin clearance in persons with CF (4). In addition, we were able to determine that sputum inhibition of tobramycin activity was saturable (suggestive of adsorptive bind-ing) and could be overwhelmed by addition of significantly higher levels of tobramycin. Our observation that CF sputum had a saturable inhibitory effect on tobramycin suggested that topical administration of high quantities of tobramycin by aerosol could overcome sputum inhibition, produce an active unbound fraction and result in improved tobramycin efficacy. Initially, our intention was to explore this treatment modality as a therapy to treat patients experiencing a pulmonary exacerbation.

While these studies were being conducted, we also completed a clinical trial of the synthetic penicillin azlocillin for the treatment of CF pulmonary exacerbations (5, 6) . During that study, we recognized that a subset of our patients could be characterized as exacerbation "repeaters," meaning that they were always at an increased risk for an exacerbation, having many such episodes in a year. At the same time, other CF clinicians, and in particular physician-scientists in Denmark, had reported improved patient outcomes by use of scheduled antibiotic treatments in their patients (7) . These observations led us to consider the merit of using scheduled "high dose" inhaled tobramycin as a prophylactic therapy to reduce risk of exacerbation in high risk patients.

When we began these investigations, aerosol generation was an uncharacterized route for chronic delivery of drugs, and particularly of drugs for which the potential for systemic toxicity was substantial. In order to better characterize any potential safety associated with prolonged administration, we completed an open-label, single arm 3-month study of inhaled tobramycin in 22 subjects (8) . This was primarily a safety study to assess the effects of prolonged aerosol antibiotic administration on renal and 8th cranial nerve function; we meticulously tracked high-frequency audiometry, vestibular function, glomerular filtration rates and proximal renal tubular function in these subjects (8) . There were no significant safety concerns with the treatment, and we observed that mean pulmonary function (assessed as the forced expiratory volume in 1 second; FEV1) increased during the first 4 weeks of treatment but then appeared to level off and even decrease during the following 2 months of treatment. In addition, nearly three quarters of patients had at least one P. aeruginosa isolate with an MIC above the tobramycin parenteral breakpoint at the end of 3 months of treatment (8) . When respiratory secretions were assessed a month after cessation of treatment, the prevalence of tobramycin-resistant isolates had decreased dramatically (as had FEV1). This "reversion" of resistance phenotype in CF P. aeruginosa clinical isolates after removal of antibiotic pressure had been previously observed in studies of aminoglycosides, fluoroquinolones and beta-lactams (9) (10) (11) , and was interpreted to mean that antibiotic resistance in P. aeruginosa could be "transient," and reversible.

In order to confirm the improvement in FEV1 observed in the 3-month inhaled tobramycin safety study, we subsequently conducted a randomized, placebo-controlled, taste-masked, cross-over study in which one group of patients received aerosolized high-dose tobramycin for 1 month followed by 2 months of placebo and the other group received 1 month of placebo followed by 2 months of inhaled tobramycin (12) . This study confirmed a statistically significant improvement in FEV1 associated with tobramycin inhalation that (again) was not observed to increase beyond 4 weeks of treatment, and again showed no evidence of renal damage or neurotoxicity. In addition, although the protocol called for administration of inhaled tobramycin tid, review of adherence records and patient reporting indicated that less than 50% of midday doses had been administered during the study, suggesting that a) the observed benefits of inhaled tobramycin were likely to be retained with a bid dosing schedule and b) a tid dosing schedule was probably unrealistic given the delivery system being employed. By today's standards, it is understandable why the delivery system in use was problematic with respect to adherence: in order to receive a dose, a patient/parent had to dissolve 1.2 g of tobramycin sulfate in 30 mL saline, transfer 15 mL to an Ultraneb 100/99 ultrasonic nebulizer, add another 15 mL of saline, turn on the aerosol generator and let it run for 2 minutes to assure linear output, and then take 200 breaths (12) . After use, the apparatus had to be rinsed and dried.

At this point, the potential benefit to the CF population of availability of a safe, effective, optimized formulation and delivery system for inhaled tobramycin was evident, but the cost of such an undertaking was daunting. The CF Foundation and Seattle Children's Hospital approached a small Seattle biotech company, PathoGenesis Corporation, to take on the task in return for rights to commercialization. We had collected evidence that measurable benefit of continuous treatment with inhaled tobramycin was maximized within 1 month, and that emergence of isolates with reduced tobramycin susceptibility was low 4 weeks after inhaled tobramycin administration had been discontinued, and thus the month-on/month-off schedule of inhaled tobramycin was born. In Phase 3 studies, inhaled tobramycin (now 300 mg delivered bid by PARI LC+ jet nebulizer) again showed an FEV1 benefit and relatively infrequent emergence of tobramycin-resistant isolates (13) ; the systematic study of >500 patients permitted the identification of an increased risk of tinnitus and hoarseness associated with this treatment (13) .

In the intervening years since the validation of inhaled tobramycin as an effective chronic therapy, aerosol delivery systems have become faster, more efficient and more convenient. In addition, the expansion of molecular techniques and refinement in the study of CF microbiology has revealed that our assumptions regarding "transient" resistance and microbial ecology were overly simplistic.

Finally, tobramycin is no longer the only well-characterized inhaled antibiotic for management of persons with CF whose airway secretions yields P. aeruginosa, and even more inhaled antibiotics are in clinical development. Although other inhaled antibiotics have been and will be studied independently in placebo-controlled trials, optimal management of our patients will likely be achieved by coordinated use of these therapies. Thus, the next phase of this journey is to consider how best to study their use together. Other: And…. Doctor: Adherence remains poor. The drugs appear safe. Resistance as defined for drugs given intravenously probably has no relevance to inhaled antibiotics.

Other: Right! Let us review the efficacy of daily inhaled antibiotic therapy. Hodson showed that daily nebulised antibiotics improved respiratory function and reduced the incidence of hospital admissions (1) . Subsequent studies show a slower decline in respiratory function, improved clinical scores and weight, decreased P. aeruginosa density, exotoxin A, or elastase (2) . These studies are weakened by small patient numbers and the use of varied antibiotics at varied doses administered by varied delivery systems, but meta-analyses confirm positive treatment effects (3) . No see-saw clinical responses were observed.

The first evidence based study of inhaled antibiotic treatment used a 4-week alternating "on/off" regimen for Tobramycin Solution for Inhalation (TSI) (4) . Because of the vogue for evidence based medicine -"The use of cyclic, suppressive inhaled antibiotic therapy with 28 days of therapy followed by 28 days off therapy has become the standard of care for CF patients > 6 years of age with chronic P. aeruginosa infection"(5) -the on/off TSI regimen has become the comparator for all trials of new medications, despite the fact that in that(4) and every subsequent study, patients show a roller coaster response (up in the "on" months, down in the "off" months), challenging the claim that improvement in lung function was maintained during each "off" period (4) .

I suggest there is no rationale for allowing patients to deteriorate every other month and that the reasons promulgated for alternate month therapy are flawed. A challenge to a continuous daily regimen should show at least lack of inferiority. This has not been addressed or proven. Instead we chart the "month off" fall in respiratory function and quality of life (QoL) or, by a variety of interventions, nibble at the on/off approach so that patients may take therapy only once daily but continuously, or take an alternative inhaled antibiotic in the "off" month. The result? None of us knows what constitutes optimal inhaled antibiotic treatment. I suggest that at least in part we have accepted the on/off regimen as a baseline for treatment not so much because of fears of drug toxicity and bacterial resistance, but because halving the annual cost makes the prescription more palatable to those who pay.

Arguments for on/off treatment plead the case for better adherence, less toxicity risk, less probability of developing increased bacterial antibiotic resistance, but where is the evidence? Patients renege on significant parts of treatment regimens, in particular chest physiotherapy and aerosolised antibiotic protocols because these are time consuming and without immediate gratification. We have shown a median adherence of 36%, measured by nebuliser download (6) . Rather than argue that an on/off regimen will increase adherence, (hypothetical, not evidence-based), we should argue that with such a protocol and known adherence patterns patients are likely to be in receipt of their prescribed aerosolised antibiotic for less than 2 out of every 8 weeks. Positive results seen during clinical trials probably reflect the stimulus effect of taking part in the study and bias possibly introduced by the patients agreeing to enrol being those most motivated to take their treatment. Even so every study of on/off treatment shows a fall towards baseline values in the "off" month. How much more pronounced will this be in the reality non-clinical trial situation of poor adherence? In practice patients probably achieve the equivalent of an on/off regimen when prescribed continuous treatment! What is the likelihood of drug toxicity? There are no data for long term continuous use of inhaled TSI or aztreonam (AZLI) but safety data available are reassuring. A review of studies from 1965-95(7) and a 1999 Consensus Conference (8) show no evidence of renal or ototoxicity. Extra vigilance is needed for patients with renal insufficiency (9) and with the more efficient mesh based nebulisers, although the risk is low (10) . Good safety data are published for tobramycin inhaled powder (TIP) (11) and AZLI (5) .

Is bacterial resistance with continuous daily inhaled therapy likely to be a practical issue? There are no longterm data for the newer antibiotics but despite the widespread use of twice daily nebulised colistin in Europe, resistance has not been a problem. Reassuringly, studies of on/off therapy with TSI, TIP, and AZLI have not shown significant increases in resistance to the inhaled preparation, increased resistance to other antibiotics, or increased isolation of naturally resistant bacteria, e.g. B. cepacia complex, Stenotrophomonas, Achromobacter. Moreover, "susceptible thresholds for parenteral administration might not be relevant to inhaled therapy as the drug concentration achieved at the site of infection with inhaled antibiotics, including TSI, can be significantly higher than the systemic concentration" (12) , and the clinical response appears to be independent of the MIC (13) . Indeed, Moss advises that the argument about increased bacterial resistance is less important than addressing the progressive fall in lung function that accompanies P. aeruginosa infection (13) .

In conclusion we must ask why we wish to risk compromising patients' immediate well-being and longevity by only treating them half the time if we accept that: 1) inhaled antibiotic therapy results in a significant mean improvement in lung function and QoL, a mean decrease in sputum P. aeruginosa density that significantly correlates with increases in FEV1 (13) , and reduced hospital admissions;

2) patients are likely to suffer an annual loss of lung function of 1-4% and that TSI may impact on the natural course of the disease by reducing this rate of fall (14) ;

3) small airways disease starts early, is only detectable by standard spirometry when substantial obstruction has occurred, is exacerbated by P. aeruginosa acquisition, and improved by TSI therapy(15); 4) a delay in starting inhaled antibiotic treatment is associated with an impaired response suggesting an irreversible component to lung function decline(13); 5) patients will have recurrent respiratory infective exacerbations resulting in long term decline in FEV1, especially in children so that "preventing exacerbations may ultimately be more important than the approach taken to treat the exacerbation"(16); 6) only treating exacerbations will not arrest the progressive decline in lung function(17); and 7) most patients will not be more than 50% adherent to inhaled treatment regimens.

The argument for an alternating four week on/off treatment regimen has to be a solecism.

Donald VanDevanter, Ph.D.

We have entered a new age with respect to inhaled antibiotic therapies for the management of patients with CF and chronic Pseudomonas aeruginosa lung infections. Members of two different antipseudomonal antibiotic classes (aminoglycosides and monobactams) have now been specifically formulated for inhalation and studied extensively in large controlled clinical trials (1) (2) (3) (4) . The commercial availability of these agents and the likely regulatory approval of other inhaled antibiotics in the near future creates an opportunity for use of these agents "in concert" in the management of individual patients. Although there may be an expectation that use of multiple inhaled antibiotics can produce better outcomes than monotherapy, there is and will be little objective evidence of improved outcomes for combining these therapies derived from registration studies. Rather, it will likely be up to the CF community to determine how to best combine these therapies and whether resulting outcomes are "better."

There are a number of outcomes that have been used to study antibiotic therapies, all of which could potentially support studies of inhaled antibiotic combinations. These include clinical endpoints such as effect on pulmonary exacerbation, quality of life, and weight gain, as well as the surrogate endpoint of pulmonary function assessed by spirometry (5) and "markers" of benefit such as change in high-resolution computerized tomography (HRCT) (6) . Other marker endpoints that would lend themselves to studying chronic antibiotics include lung clearance index (LCI) by multiple-breath washout (7) , and markers of inflammation in respiratory secretions (8) . Emergence of local and systemic toxicities and changes in infection microbiology have been employed to study safety issues (1).

There will be challenges in designing trials to demonstrate an additional benefit of combining inhaled antibiotics, including powering of a superiority endpoint using an active comparator (inhaled monotherapy), maintenance of blinding when each aerosol treatment has a different (and identifiable) formulation and delivery device, differences in patient response that may result from infection history, airway microbial community, and previous antibiotic exposures, and choice(s) of treatment rotation schedule. While these are difficult challenges, they seem tractable and it is likely that a study comparing inhaled antibiotic monotherapy (or monotherapies) to combination therapies could be designed.

More important than design questions is the discrepancy between the likely duration of even the most ambitious randomized, controlled, prospective trial (a few months to a year?) and the length of time that a patient with a chronic lung infection might benefit from chronic suppressive antibiotic therapy (perhaps decades). Funda-mental clinical questions are likely to remain unanswered upon successful completion of a 6 month or even 1 year study comparing inhaled monotherapy to combination therapy. How would clinicians extrapolate risks and benefits identified in such a study to extended use of inhaled antibiotics? How likely is it that the study would have recognized and addressed the underlying motivation for combining inhaled antibiotic therapies in the first place? Is this really about increased amplitude of benefit, or rather increased duration of benefit?

It is difficult to find researchers or caregivers who do not intuit that combination of inhaled antibiotic classes should produce better outcomes than monotherapy over the lifespan of a patient with CF. Why is this the case and what assumptions underlie this belief? One assumption (A) appears to be that there is a finite capacity for any single antibiotic to produce benefit when administered chronically to a patient with a chronic lung infection: at some point, a patient will become refractory to response with that antibiotic. This seems a rational and reasonable assumption. A companion assumption (B) is that this capacity for benefit can be manipulated by variation of antibiotic dose, schedule and regimen. This is the underlying tenet of chronic intermittent inhaled tobramycin: continuous use of inhaled tobramycin for 3 months showed decreasing pulmonary function benefit after 1 month (9), while chronic intermittent inhaled tobramycin use showed continued efficacy after 6 months (1) to two years (10) . This construct has never been objectively tested in a single study and these historical studies differ in important ways (e.g., dose, dose schedule, delivery device, sample size). However, this assumption also seems reasonable.

The promise of an additional benefit derived from a second inhaled antibiotic can be found in both assumptions (A) and (B), above. If assumption (A) is true, then access to 2 antibiotics increases the amount of time that a patient can experience the benefit of chronic inhaled antibiotic suppression compared to the situation in which that patient has access to only monotherapy. This relative benefit is unlikely to be demonstrated in a 1-year prospective study, since we already know that the benefits of inhaled antibiotic monotherapy can extend beyond 1 year (10, 11) . If assumption (B) is true, then there may be a way to "rotate" the 2 inhaled antibiotics such that the effective duration over which each antibiotic provides benefit exceeds the total benefit that would be achieved if each antibiotic was used successively as a monotherapy. This is all very well and good and may even be true. However, this "benefit" of inhaled antibiotic rotation (extending the effective life of each chronic inhaled antibiotic) would presumably play out over the course of years or even decades of a patient's life, and is unlikely to be captured by the endpoints we are considering in relatively short (6-12 month) controlled studies.

If the true promise of inhaled antibiotic rotation lies in delaying that time at which a patient becomes refractory to all available inhaled antibiotic therapies, then we have yet to consider a clinical trial endpoint likely to capture that benefit. Shorter term endpoints may justify or encourage combinational antibiotic use by clinicians, but they will not allow us to answer this greater question: have we increased the effective time span that we can treat patients? Not that we are unable to measure refrac-toriness…it is actually quite simple. When an individual becomes refractory to a chronic therapy that he or she had previously responded to, then treatment cessation should not result in immediate clinical deterioration. Our challenge is that we already know that most patients remain responsive to chronic intermittent inhaled antibiotic monotherapy for more than a year, as evidenced by consistent, reproducible declines in FEV 1 and QoL observed during their off-drug periods (10, 11) . Thus, choosing a "time to refractoriness" endpoint for a study of a year or less duration would be misguided, at best.

In conclusion, although prospective randomized studies comparing the superiority of combining inhaled antibiotics to monotherapy are feasible, they are unlikely to address underlying motivations (and greatest potential benefits) of combining therapies: extension of the effective time during which infections can be suppressed. Rather, it will be the job of providers to consider whether patients under their care continue to benefit from inhaled antibiotics (rotated or not) after months and years of use, to consider careful and deliberate N of 1 "withdrawal studies" to assess refractoriness, and to record these data in a manner that will allow us to characterize outcomes using retrospective, observational analyses. Today, the standard dosing interval for FDA-approved inhaled antibiotics is 28 days, with an implied 28 day rest period. This regimen was originally conceived based upon interpretation of results from clinical studies with inhaled tobramycin conducted decades ago, but never objectively tested against other possible regimens. The availability of multiple classes of inhaled antibiotics allows the clinician to consider a number of approaches for combining different inhaled antibiotics into a single management plan. How will we go about identifying those approaches that are the most likely to provide improved outcomes for the population?

The strategy of intermittent (e.g., month on-month off) antibiotic dosing has been contested in clinical practice. This has been done primarily because the clinician and patient have decided that there are unacceptable outcomes associated with that regimen, such as loss of lung function or occurrence of pulmonary exacerbations. Indeed, in the original trials of inhaled tobramycin, 40% of patients on tobramycin required treatment for an exacerbation (1) . Some may argue that the reduction in lung function observed in the "off" months demonstrates worsening infection and inflammation and more aggressive intervention may prevent further loss of lung function. In these cases, the patient would have been treated by either adding a second drug, usually an off-label intravenous antibiotic (e.g., colistin), or continuous use of the inhaled tobramycin (2) .

Today there are two inhaled antibiotic formulations FDA-approved for treatment of persons with cystic fibrosis (CF) and Pseudomonas, and several other agents in latestage clinical trials. If all goes according to plan, clinicians will have the "luxury" of being able to prescribe one or more of several classes of anti-pseudomonals in the near future. Although FDA approval will have required demonstration of efficacy of an inhaled antibiotic compared to a placebo, and there will be some head-to-head data available as well, what clinicians will really need is some indication as to whether management of our patients will benefit from using these drugs in some sort of combination, as opposed to as single agents. One can propose a number of approaches for combining different inhaled antibiotics into a single management plan (Table) . Presumably, some approaches are more justifiable from a mechanistic standpoint than others. Similarly, some are likely to be more feasible from a treatment burden, adherence, and/or cost perspective. How will we go about identifying those approaches that are the most likely to provide improved outcomes for the population?

The two most basic questions involve a comparison of intermittent (e.g., month on-month off) against continuous (no time off) therapy (Table-Strategy 1 vs. Strategy 3) or a comparison of a combination (or rotation) of antibiotics compared to a single drug (Table-Strategy 2 vs. Strategy 3). The addition of more antibiotics increases the complexity even further as we can then consider intermittent therapy using a rotation of antibiotics.

The design of treatment strategies for a trial is not difficult; several are shown in the Table, but the challenge lies in the efficacy endpoints that would help us know which treatment strategy is the most effective.

Whereas previous studies of aerosol antibiotics have demonstrated improvement in FEV 1 and quality of life, and a reduction in exacerbations, these were short-term trials powered based upon comparison against placebo. How would those endpoints work in a comparison against another inhaled antibiotic, one of which has been in use for a long period of time?

As noted earlier, the suggestion that a different treatment strategy should be considered is based upon the observation that the current FDA-approved strategy of intermittent therapy is not optimal. So it would seem that important clinical parameters that prompt a change in treatment approach to either continuous therapy or rotation of antibiotics (or both) include loss of lung function (i.e., over time), impaired quality of life, and frequent exacerbations. These, too, would not be difficult endpoints to measure in a clinical trial, but the feasibility of performing such trials is challenging because the studies would likely require longer durations (i.e., it is unlikely that we would see a meaningful difference in the short term) and power analyses suggest that large numbers of patients would be required.

There are additional challenges to the feasibility of performing such trials. The current antibiotic choices use different drug-device combinations and it may be impossible to completely blind patients and investigators to the treatment. A double-blind, double-dummy approach to circumvent this problem may be unrealistic. Thus, an open-label design may be the best approach, even with the potential for bias. What is the impact of previous antibiotic exposure? This could be accounted for by randomization of patients based on prior usage, but the addition of randomization factors results in a greater number of patients required for the study.

Finally, how would such a study be funded? Industry is required to demonstrate the efficacy and safety of their drug being studied; they are not required to demonstrate the optimal treatment strategy, but typically follow established paradigms under FDA guidance. Comparative effectiveness trials using the CFF Patient Registry seems the most likely avenue but we cannot expect third party payors to be interested in paying for the added medications in the absence of existing evidence. It seems a partnership between the CFF, industry, the FDA, and some third party payors would be necessary to perform such a trial.

References This presentation addresses the following objectives: 1. Establish goals of introducing end of life care discussions in the plan of care for CF patients and families. As a result of recent data suggesting efficacy for treatments attacking basic defects in mutant CFTR function, optimism for finding a cure for cystic fibrosis is at its highest level since the identification of the CFTR gene (and its accompanying promise of gene therapy). However, patients and families still live with the reality of a disease that continues to impose the need for choice from among a wide range of palliative and aggressive treatments as the illness progresses and premature death approaches. Careful integration of end of life discussions, with goals appropriate to the current condition and disease trajectory of individual patients, can help patients and families navigate these choices deliberately, with confidence, and with hope toward the best possible outcomes.

2. Identification of challenges associated with discussing end of life care with CF patients and their families. Traditionally, care teams have adopted strategies that have been successful in helping patients and families to accept aggressive and intensive monitoring and intervention as necessary parts of their lives. Enthusiastic and encouraging emphasis on the benefits to be achieved, and reinforcement of behavior on the part of families that reflects the same enthusiastic optimism is an important part of this approach. Though this approach has been important in translating the introduction of new therapies into better FEV1s, higher BMIs and longer lives, it can also send the unwanted message that fears regarding premature death and progressive illness themselves endanger the best outcomes, and are not appropriate topics to bring up with the team. Conversely, inviting such discussions can be seen by the care team and family as "taking away hope," depriving patients and those caring for them of powerful tools that can address directly such fears and, in many cases, allay them.

3. Identification of "triggers" signaling the need for, and opportunities to introduce, end of life care discussions. Impending death, or refusal to undergo lung transplantation with progressive severe lung dysfunction, certainly signals the need for end of life discussions. However, patients signal their need for such discussions, and their opportunity to benefit from them, earlier in their course. Teams can learn to recognize these signals and how best to respond to them.

Though end of life discussions surrounding impending death are less frequent in children than they are in adults, they still occur. Of equal importance are concerns of children and their families regarding how the threat of shortened life span may affect quality of life, adherence to treatment and normal development in patients who are far from respiratory failure or decisions regarding transplant. The multidisciplinary team has become the cornerstone in the delivery of cystic fibrosis (CF) care and is well understood by team members, patients, and caregivers. Multidisciplinary team members contribute to the high quality of care that has come to be expected in Cystic Fibrosis Foundation Accredited Care Centers and are essential in the care of the CF patient at end of life. Unfortunately, not all members of the multidisciplinary team fully understand their roles and responsibilities in end of life or palliative care of CF patients. To better assess team members' understanding of roles and educational exposure in caring for CF patients at end of life a survey was developed and distributed via the established list serv's within the CF community. The survey was distributed to 244 nurses, 421 dietitians, 423 social workers and 192 respiratory therapists. Responses were received from 45 (18%) nurses, 102 (24%) dietitians, 57 (13%) social workers, and 51 (27%) respiratory therapists. The results of the survey indicate that although the majority of respondents have 15 or more years of experience within their discipline, most have been working with CF patients for 5 years or less. The majority of respondents or 44% reported they "somewhat understand" their role in caring for CF patients at the end of life, with social workers feeling most confident in their role and dietitians feeling least confident. All disciplines expressed the need for additional education related to advanced care planning and care of the CF patient at end of life. This information is presented in an initial effort to assist multidisciplinary team members to recognize and understand their respective roles in end of life care of patients with CF and will examine the roles of the nurse, social worker, respiratory therapist, and dietitian.

The role of the CF nurse may differ slightly depending upon whether or not care is provided in primarily an inpatient or an outpatient setting. Regardless of the setting the CF nurse works, similar principles should be used in providing end of life care. Four guidelines that can be used by the CF nurse in the care of the patient at the end of life and in advanced care planning are 1) providing the patient and family with support at the end of life or in advanced care planning; 2) supporting informed discussions about care goals and treatment options; 3) assisting the patient by facilitating the implementation of the patient's wishes regarding end of life care; and 4) knowing and understanding current legal issues surrounding end of life care and advanced care planning (1). The CF nurse may also need to assure that the patient and family are comfortable with decisions that have been made, and if not, may need to refer them to appropriate sources such as the palliative care team, social work, or psychology as appropriate. As is true in virtually all aspects of nursing, often the nurse spends the most time with the patient and family. This places the nurse in a position to identify needs of the patient and family and to address these needs or communicate them to the appropriate team members. It is through these roles, along with effective communication skills that the CF nurse is able to provide effective care at end of life.

Survey results indicate that social workers are most familiar and comfortable with their discipline's role in caring for patients during end of life. This is not unexpected as social workers are commonly consulted at these difficult times. The role of the CF social worker in the care of the CF patient at the end of life and in advanced care planning is consistent with the role in caring for any patient at the end of life, or in advanced care planning. However, the unique, often long-term relationship of the CF social worker with the CF patient may provide a strong foundation of support to the patient and family in dealing with sensitive end of life issues. According to the National Association of Social Workers Standards for social work practice in palliative care and end of life, there are several standards that social workers can use to help them work more effectively in end of life settings. Some of those standards include 1) Ethics and Values; 2) Knowledge; 3) Assessment; 4) Intervention/Treatment Planning; 5) Attitude/Self-Awareness; 6) Empowerment and Advocacy; 7) Interdisciplinary Teamwork; and 8) Cultural Competence (2) . Through following and utilizing these standards in everyday practice social workers can be very effective in helping the CF patient and their family at end of life and at the time of advanced care planning.

Respiratory therapists are essential members of the CF multidisciplinary team across the continuum of care, including care delivered towards the end of life. Although approximately 80% of respiratory therapists report that they "somewhat" or "fully" understand their role in end of life care of CF patients, 50% feel minimally or not at all prepared to deliver that care. The role of the respiratory therapist extends across the outpatient and inpatient settings and includes participation in family conferences regarding goals of care, management of invasive and noninvasive ventilation support, and education regarding appropriate use and delivery of airway clearance therapy during end of life. Respiratory therapists may be asked to deliver hands-on chest physiotherapy toward end of life due to increased effectiveness or additional contact provided by the respiratory therapist (3). The respiratory therapist often spends a significant amount of time at the bedside for delivery of nebulized medications and airway clearance therapy and often develops close relationships with patients and families. The value of these close, often long-term relationships should not be underrated and may be a significant asset during discussions about end of life and treatment preferences.

Dietitians are integral members of every CF multidisciplinary team. In the survey, dietitians reported the least understanding of their role in end of life care. This lower level of understanding is not surprising as 99% of dietitians also responded that they have received less than 10 hours end of life education in CF. Nutrition is an essential component of CF disease management and remains a necessary area of focus during end of life as well. Patients with CF have long been taught to maintain high calorie, high protein diets due to pancreatic malabsorption and many older patients also have co-morbidities of CF-related diabetes (CFRD). Maintenance of BMI has been a goal of most CF patients (or at least of their dietitians) across their life span and nutritional issues at end of life may be complicated by this lifelong effort.

The multidisciplinary team is essential to the care of CF patients across the lifespan, including toward the end of life. The established team provides an ideal environment for delivering education to ensure that each clinician fully understands their role and is adequately prepared to care for the CF patient and their family toward end of life.

Walter M. Robinson, M.D., M.P.H.

A distinctive chronic care model, with an emphasis on prevention, routine clinical monitoring, early intervention, diagnostic standardization, and multi-disciplinary care, was introduced into US CF care in the 1960s and 1970s. Improved survival attributed to this model of care led to its widespread adoption in cities across North America. Accreditation of care centers by the US CFF promoted the model as the standard of care for cystic fibrosis. As the CF population aged, the model was adopted by centers caring for adults, with some important modifications. Monitoring of local outcome data was made routine with the widespread adoption of the CFF Clinical Registry, and was later expanded by the CFF using well-known quality improvement methods. Details of the CF care model are now routinely cited as examples of efficient and effective care that could be applied to other clinical conditions.

A specific model for the care of the dying patient in North America began almost simultaneously. Drawing on a shift in American popular concern about the "medicalization" of dying, and utilizing clinical models developed in the UK and elsewhere, a system of care for those with terminal illness slowly gained strength in the U.S. and Canada. Initially based outside hospitals and using the principles of hospice, this shift in clinical care for the dying gained strength in the 1990s as both clinicians and patients began to recognize the need for better symptom control in life-limiting illness. The increasing use of supportive technologies, including those formerly limited to intensive care units, and the development of new therapies with both high risk and great potential for benefit meant that a greater proportion of deaths in the US occurred after an aggressive but ultimately futile intervention. The palliative care model, distinct in both location and principles from the hospice model, began to emerge in clinical centers across the continent. This model developed unevenly in pediatrics, but grew rapidly in adult medicine. The palliative care model emphasized control of symptoms (even if the cause of the symptoms could not be eliminated), management of both treatment and disease side effects, shared decision-making with patients and families, recognition of the social and spiritual aspects of life-limiting illness (which many perceived to be lost in an increasingly bureaucratized and depersonalized death), and above all, effective communication with patients and families about prognosis and treatment options.

Integration of these two care models-the CF chronic disease model and the palliative care model-can be difficult for a host of reasons, yet there are also factors distinctive to the care of children and adults with CF that can promote better palliative care.

Barriers to Effective Palliative Care in Cystic Fibrosis.

1. Lack of CF-specific symptom research. Studies performed over the last decade have amply demonstrated the breadth and depth of the symptom burden in CF, but few studies have examined specific treatments for those with CF.

2. Lack of systematic study of CF care in the last three months of life. There is a scarcity of up-to-date, multicenter, prospective observational studies of the clinical and social characteristics of death due to CF, such as medications used and their efficacy, patterns of care, moral and ethical distress of caregivers, patients and families, and a host of other factors. Such studies are widely performed in oncology, and the methods used in oncology could easily be adapted to gain clinical insights into the relative success and failure of specific approaches to CF care in the last three months of life.

3. An "either/or" attitude to curative and palliative care among CF clinicians. The CF clinical care model emphasizes close adherence to a difficult regimen; we rightly emphasize that adherence can extend life, often dramatically so. In our zeal to promote adherence, we can often become so optimistic about the remarkable advances made in CF care that we may find it difficult to discuss the negative aspects of living with CF. We may minimize side effects, ignore (or more often, fail to ask about) symptoms we have no easy answer for, and we may be reluctant to discuss declining quality of life or eventual death. We are not unique among clinicians in this regard, but the dramatic advances made in CF, as well as the increasingly available possibility of rescue by lung transplant, may make us particularly reluctant to utilize effective palliative care approaches which could increase quality of life. This difficulty can be amplified by our long-standing and close relationship with many patients and families, as we may avoid discussion of any distressing aspects of life with CF with those to whom we have become so close. 4 . Shift in age at death. Changing demographics of death in CF have meant that skill in caring for the dying child or adolescent with CF has been lost; the death of a child or adolescent is now fortunately rare enough that our pediatric skills may be rusty. Inpatient clinical staff on adult wards are now caring for the majority of deaths from CF, and we need to shift our focus to this clinical setting.

Factors Promoting the Integration of Palliative Care into CF care.

A multi-disciplinary team approach to care. A team approach is a fact of life in the CF care model, and some members of the team may be more skilled at symptom assessment or communication about difficult topics.Yet all members of the team should become familiar with effective palliative care, and all members should learn to promote symptom assessment and treatment as well as open discussion of prognosis and treatment options. Under no circumstances should the burden of communication about prognosis or treatment options be left only to one team member.

2. A pattern of routine and preventive care. The routine nature of the CF care model gives ample time and opportunity to develop the rapport essential for shared decision-making. This is a luxury many of our colleagues in other clinical settings do not have, and we must make the most of it. Early discussion of advanced treatment options is a realistic goal in the CF model of care, although studies suggest we do not take advantage of the opportunity.

3. Widespread availability of expert palliative care teams. Palliative care services are increasingly available in clinical centers across North America, so that expertise in symptom management is far more accessible than in years past. However, anecdotal evidence and single center studies suggest that palliative care teams are invited onto the CF team only rarely and often at too late a stage to be helpful. This is another effect of the "either/or" thinking regarding palliative and curative therapy in CF, but we can and must overcome this outdated approach. Lung transplantation for most patients as a palliative rather than curative treatment, and comparing the expected survival benefit from lung transplantation with improvement in quality of life remains a topic of considerable discussion in the cystic fibrosis and transplant communities.

When CF patients and their families are referred for Palliative Care, the primary goals are relief of symptoms and improvement in quality of life, whether or not such therapies may or may not shorten the duration of survival. Therefore, it is critically important to reconcile the concerns of patients, family, and caregivers during the terminal stages of cystic fibrosis lung disease with the requirements of that patient also listed for lung transplantation. Opiate analgesics may be exceedingly helpful in relieving pain associated with osteoporotic compression fractures, as well as providing relief of dyspnea in patients with any end-stage lung disease. However, osteoporosis will not improve following transplant, and it is important to keep in mind that regular preoperative use of these medications induces tolerance and leads to higher opiate doses needed to manage postoperative pain.

These higher doses may contribute to postoperative delirium and prolonged need for mechanical ventilation, delayed gastric emptying associated with nausea, emesis and aspiration pneumonia. While opiate pain medications are also effective in relieving dyspnea and anxiety associated with the end of life, inadvertent postoperative cessation of these medications may also be associated with delirium and withdrawal.

The use of noninvasive positive pressure ventilation (NIPPV, BiPAP) and/or endotracheal intubation and mechanical ventilation increases the Lung Allocation Score (LAS) and moves patients up the lung transplant waiting list. NIPPV is indicated for patients with chronic hypercarbic respiratory failure, and is particularly valuable as a bridge to transplantation in those cystic fibrosis patients already listed. Its use may provide valuable respiratory muscle rest, allowing patients to maintain ambulation and fitness critically important to successful transplant outcomes. Endotracheal intubation and positive pressure ventilation may keep the already listed patient alive until transplant, but the presence of an artificial airway that needs to be secured substantially limits activity and compromises overall conditioning. The presence of an endotracheal tube severely limits the patient's ability to communicate with family members and staff, critically important in assessing patient comfort needs at the end of life. The use of either of these modes of mechanical ventilatory support in the end-stage cystic fibrosis patient not listed for lung transplantation is of questionable value. Cystic fibrosis (CF) is an autosomal recessive disease, caused by mutations such as [∆F508] in the CFTR gene, which prevent trafficking of the CFTR protein to the apical surface of lung epithelial cells. One important consequence is inability to activate the cAMP-activated chloride channel activity, which is intrinsic to CFTR. However, another consequence is the genesis of a profoundly hyper-proinflammatory phenotype in the airway, principally manifested by high levels of IL-8 and certain other potent cytokines and chemokines. A frequent cause for debate is whether the pathophysiology of the CF airway is due to persistent infection, or due to intrinsic inflammation, or due to a mixture of both? Furthermore, because of operational defects in the innate immune system in the lung, it is possible that inflammation in the CF airway is potentiated by CFTR-dependent defects in immune cells, including neutrophils, macrophages and lymphocytes.

Is it Persistent Infection? One possibility for CF lung pathophysiology is that for an as yet unknown reason, the epithelial cells in the CF lung are hypersensitive to bacteria and other infectious agents. Consistently, cultured CF lung epithelial cells respond to bacteria and lipopolysaccharide (LPS) by expressing IL-8 at a greater rate than do lung epithelial cells from non-CF, control patients (1, 2) . It has also been shown that Pseudomonas produces a toxin that suppresses whatever function is left in [∆F508]CFTR (3) . Taken together, these pieces of information can be interpreted as consistent with enhanced sensitivity on the part of CF epithelial cells to infectious agents, possibly involving intrinsic activation of NFκB signaling.

Is it Intrinsic Inflammation? With respect to potential intrinsic connections between mutations in CFTR and production of IL-8 protein, the IL-8 mRNA in CF lung epithelial cells has been recently shown to be greatly stabilized relative to IL-8 mRNA in control cells (4) . In this case, the mechanism involves low levels of the ARE-binding protein tristetraprolin (TTP), which regulates mRNA stability by interacting directly with ARE domains in the 3' untranslated region (3'UTR) of the IL-8 mRNA. In addition, MAPKinases have direct control over IL-8 mRNA as well (5) . We have also investigated the relationship between the presence of mutant CFTR and disease-specific microRNAs (6, 7) . The microRNA miR155 was most elevated in CF cells, where it serves to activate signaling from PI3K to AKT. In turn, activated AKT drives downstream activation of MAPkinases, thereby stabilizing the IL-8 mRNA.

Is it Other Immune Cells in the CF lung? The proinflammatory phenotype of the CF lung has long been associated with objective defects in those neutrophils, macrophages and lymphocytes present in the airway. However, until recently, these defects have been ascribed to damage to otherwise normal immune cells, inflicted by the cytokine storm of epithelial cell dysfunction, infection, and obstruction. Macrophages, either resident in the CF lung, or studied in samples of peripheral circulation, have been found to express functional CFTR, which regulates phagosome acidification in macrophages (8) . More recently, it has been shown that disease-causing mutations in CFTR can determine the functional responses of alveolar macrophages (9) .

Likewise, neutrophils express CFTR, and apparently use it to translocate chloride into the phagolysosome, where it combines with H 2 O 2 to produce bacteriocidal HOCl (10, 11) . CF neutrophils are found to be relatively inefficient at bacterial killing, and this defect may therefore also contribute to failure of the CF airway to clear the airway of bacteria. We have recently observed in CF neutrophils from blood that IL-8 mRNA is massively elevated, and that genes associated with regulation of ER stress are particularly prominent (12) .

Finally, in adaptive immunity experiments, lack of Cftr in CD3+ lymphocytes leads to an exaggerated IgE response when transferred into Cftr+ wildtype mice (13) . The data clearly identify a Cftr deficiency in lymphocytes rather than the epithelial cells as the source of the dysfunctional signal.

The available data therefore seem to suggest "both," with an emphasis on CFTR-dependent proinflammato-ry signaling in lung epithelial cells. However, the available data also yield insight into why the CF airway cannot resolve infection. Disease-causing CFTR mutations affect the function of all the cells making up the lung's defensive armamentarium, leading to a perfect storm of functional incompetence. This insight may have relevance to the problem of drug development for CF, because candidate therapeutics will have to rescue CFTR function in more than just the CF epithelial cell. some studies indicating that airway inflammation in CF only follows respiratory infection (14) (15) (16) (17) . Similarly, in pigs with mutated CFTR genes no inflammation is present at birth; but very early in life, CF piglets develop signs of CF lung disease including infection, airway inflammation and mucus accumulation suggesting a defect in host-defence mechanisms (18) .

The controversy of whether lung inflammation in CF can occur in the absence of infection, while important for understanding the basic pathophysiology of CF, is less relevant for the treatment of CF patients. Airway infection and inflammation are virtually universal in CF patients (regardless of the order in which they develop) and reducing inflammation is likely to improve CF lung disease. We must now identify the molecular pathways that can be targeted to safely and effectively reduce lungdamaging inflammation in CF. Background. Pseudomonas aeruginosa is an opportunistic pathogen and a leading cause of nosocomial infections, especially in a setting of epithelial cell injury and in immunocompromised patients, including cystic fibrosis (CF) patients (1) . Mutations in CFTR reduce or eliminate chloride secretion in the lung and thereby reduce the volume of airway surface liquid, which in turn suppresses mucociliary clearance of P. aeruginosa and other pathogens trapped in the mucus (2, 3) . Accordingly, individuals with CF are susceptible to chronic infections, primarily by the opportunistic pathogen P. aeruginosa. Over time the CF lung evolves into a highly inflamed and purulent environment that is the proximate cause of morbidity and mortality in these patients. By late adolescence, 80% of CF patients are chronically infected with P. aeruginosa, the dominant pathogen in CF airways (4) (5) (6) . As is the case with CF, patients with COPD and bronchiectasis are particularly susceptible to bacterial infection, including colonization by P. aeruginosa, which is particularly difficult to eradicate due to the high basal levels of antibiotic resistance and the ability of P. aeruginosa to form antimicrobial resistant biofilms (3, (7) (8) (9) (10) (11) .

In the CF lung, the virulence factors produced by this microbe and the chronic immune response mounted by the host in a fruitless attempt to clear these infections combine to cause extensive damage to the lung tissue, and frequently, respiratory failure and death of the patient (4-6). Patients with chronic P. aeruginosa colonization often acquire respiratory virus infections, which may trigger acute exacerbations. Synergism between bacteria and viruses in inducing infection in the airway is welldocumented (12, 13) . CF patients show reduced ability to clear P. aeruginosa acquired during respiratory viral infections compared to healthy controls, often requiring anti-pseudomonal treatment (14) . Furthermore, 85% of new pseudomonal colonization of CF patients followed a viral upper respiratory tract infection within 3 weeks (15) . Most studies to date characterize an infection where the virus predisposes patients for a secondary bacterial infection, but scientists and clinicians now recognize that the synergism is not unidirectional (16, 17) . In fact, intrapulmonary influenza titers are also elevated in mice with chronic P. aeruginosa infection compared to P. aeruginosa free mice (18) . And although the prevalence of viral disease is similar in CF and healthy controls, CF patients with viral lung infections have prolonged and exacerbated symptoms and longer hospitalizations (12, 19) . The abundance of literature documenting respiratory viral and bacterial, often P. aeruginosa in CF, dual infections have not established a mechanism of how these polymicrobial infections arise or how they are maintained during the course of the ensuing respiratory disease. Our long-term goal is to define a mechanism for the prolonged and exacerbated viral infections observed in the clinic in CF patients chronically colonized by P. aeruginosa.

Results. We previously reported that a secreted toxin from P. aeruginosa, Cif (PA2934), reduces CFTR-mediated chloride secretion by human airway epithelial cells by enhancing the amount of ubiquitinated CFTR and its degradation via the lysosomal pathway (20) (21) (22) . The aim of the current study was to determine if Cif regulates other ABC transporters, specifically the transporter associated with antigen processing (TAP) to alter viral antigen presentation. Patients with TAP deficiency, like CF patients, have recurrent bacterial infections, nasal polyps, and chronic purulent rhinitis in the first 6 years of life, and chronic bacterial infections, bacterial pneumonia, and bronchiectasis later in life. Thus, we hypothesized that Cif facilitates the degradation of the TAP complex, resulting in reduced viral antigen presentation via the major histocompatibility complex (MHC) class I molecules and therefore, diminishes the immune response to viral pathogens. P. aeruginosa toxin Cif reduced TAP1, but not TAP2, protein abundance in polarized human airway epithelial cells (CFBE41o-cells) in a time-dependent manner. The Cif-mediated reduction in TAP1 was more dramatic in CF airway cells (CFBE41o-), as compared to WT-CFTR complemented CFBE41o-cells. Cif inhibited ubiquitin specific protease 10 (USP10), a deubiquitinating enzyme, resulting in increased poly-ubiquitination and proteasomal degradation of TAP1. The inhibition of USP10 activity by Cif was more profound in CF cells (CFBE41o-), as compared to WT-CFTR complemented CFBE41o-cells. USP10 deubiquitinating activity is regulated by protein-protein interactions with G3BP1 (Ras-GAP SH3 domain binding protein), where interaction of the N-terminus of USP10 with G3BP1 inactivates USP10 DUB activity (23) . Cif does not alter the protein abundance of USP10, but does enhance the interaction between G3BP1 and USP10 to inhibit USP10 DUB activity. In CF airway cells, G3BP1 protein abundance is increased, leading to a reduced basal level of USP10 activity and thus, greater susceptibility to the Cif toxin. Functionally, the Cif-mediated decrease in TAP1 expression reduced peptide uptake into the endoplasmic reticulum and decreased MHC class I molecule abundance at the plasma membrane, effects that were enhanced in CF cells compared to WT-CFTR complemented CF cells. Furthermore, Cif decreased MHC class I presentation of viral antigen and eliminated CD8 T cell recognition of influenza Ainfected cells.

Summary. We propose that P. aeruginosa, by secreting the Cif toxin, inhibits the ability of CD8 T cells to recognize and eliminate viral infections, severely compromising the immune defenses in the lungs of CF patients. This is the first bacterial toxin that inhibits the viral immune response of the host, potentially illuminating a mechanism for the clinical observation that viral-bacterial co-infections dramatically increase the morbidity and mortality of CF patients chronically colonized by P. aeruginosa. This study also identifies a new therapeutic target for preventing co-infections in the CF airway. Defective cystic fibrosis transmembrane conductance regulator (CFTR) function is responsible for cystic fibro-sis (CF) lung disease, the most life-threatening complication of CF. Lung function declines and fails as a con-sequence of airway inflammation, chronic bacterial infections and progressive tissue destruction. Given the critical role of lung inflammation in the pathogenesis of CF lung disease, a better understanding of the relationship between infections and inflammation is critical for the development of novel therapies targeting inflammation. An important question within this relationship has been whether inflammation or infection comes first.

A limited and remarkably consistent collection of bacteria colonizes and infects the CF lung in an age dependent sequence (1), suggesting an intrinsic defect that provides a niche for specific infections. Pulmonary infections start early in children with CF, with Staphylococcus aureus common in young infants (1) . Other organisms, such as Haemophilus influenzae and Pseudomonas aeruginosa, are believed to become more prevalent in older children and adults. Up to 80% of the patients are eventually infected with P. aeruginosa (2) . P. aeruginosa can be found as free-swimming, planktonic bacteria or as micro-colonies encased in an exopolymeric matrix, also known as biofilms. It has been suggested that chronic infections of P. aeruginosa in CF represent a prototypical biofilm infection, with the formation of large cellular aggregates of the bacteria and distinct production of molecular patterns (3) .

The airway epithelium is the first line of defense against foreign particles and pathogens; it not only forms a physical barrier, but also plays an important role in the detection of pathogens and the initiation of inflammatory responses. This is accomplished via an intricate network of signaling pathways, both within and between cells. Activation of this network occurs via pattern-recognition receptors (PRRs), so-called because they recognize molecular patterns from bacteria, viruses and certain selfderived molecules. The two best-characterized families of PRRs are the Toll-Like Receptors (TLRs) and the NODlike Receptors (NLRs). A convergence point of host defense mechanisms against bacterial pathogens activated by many PRRs is the p38α Mitogen-Activated Protein Kinase (MAPK) (4). Interestingly, nematodes deficient for pmk-1, a p38 MAPK ortholog, are susceptible to bacterial killing by P. aeruginosa (5) . This highlights the ancient role of p38α MAPK in regulating host defense against this pathogen and its usefulness as a marker of engagement of host innate immune responses to P. aeruginosa.

In human airway epithelial cells (AECs), activation of p38α MAPK has been coupled to TLR5 in response to P. aeruginosa (6) . Interestingly TLR5 has been reported to be an anti-inflammatory target and modifier gene in CF (7, 8) . However, it is unclear whether P. aeruginosa grown as a biofilm activates AECs in the same way as planktonic P. aeruginosa. Comparing the activation of p38α MAPK in response to planktonic or biofilm derived P. aeruginosa materials, we discovered that planktonic P. aeruginosa activates p38α MAPK in a TLR-dependent fashion more strongly than biofilm P. aeruginosa, which acted in a TLR-independent fashion. This highlights the complex relationship between P. aeruginosa and the airway epithelium in CF; airway epithelial cells distinguish different bacterial forms of the same clinical isolate using distinct pathways for their detection.

Moreover, in CF a "hyper-inflammatory" phenotype has been described, resulting in increased synthesis of pro-inflammatory cytokines (9) (10) (11) (12) . Furthermore, we recently described that CFTRdelF508 AECs are hypersensitive to bacterial stimuli, resulting in increased MAPK activation and IL-6/IL8 synthesis (13, 14) . We also found that CFTRdelF508 AECs are more sensitive to reactive oxygen species (14) . A pro-oxidative shift is associated with enhanced intracellular signaling (15) , which could explain increased MAPK activation and pro-inflammatory cytokine secretion. Interestingly, this hyper-responsiveness holds true in response to both planktonic and biofilm derived material from P. aeruginosa.

These results suggest that p38α MAPK reaches a functional threshold of activation required to engage host defense mechanisms earlier in CF. This lower threshold of activation means that AEC from CF patients may engage pro-inflammatory responses more frequently than their non-CF counterpart, resulting over time in increased tissue damage.

Meconium ileus (MI) is a severe intestinal obstruction which presents at birth. It is the earliest clinical manifestation of cystic fibrosis (CF) and occurs in ~15% of CF patients. It occurs with equal frequency in males and females, and the proportion of observed variation in MI that can be attributed to inherited genetic factors, i.e. the heritability of MI, has been estimated to exceed 88% (1). For this reason, there have been a number of MI gene mapping studies; in human these include Zielenski et al. The International Gene Modifier Consortium consists of data ascertained as part of the University of North Carolina/Case Western Reserve extremes of lung phenotype study, the Johns Hopkins University CF Twin and Sibling study, the Canadian consortium for gene modifiers population-based study, and the French CF population-based study. Using the North American data collections we conducted a genome-wide association study (GWAS) of MI in CF, comprised of 3,763 CF patients with severe CFTR genotypes, genotyped at 543,927 single nucleotide polymorphisms (SNPs) across the genome, to identify variants associated with MI. Five genome-wide significant SNPs (p<5x10 -8 ) were identified in two genomic regions that include the genes SLC26A9 (min p =9.88x10 -9 ) on chromosome 1 and SLC6A14 (min p=1.28x10 -12 ) on chromosome X; neither of which had been previously identified in other MI modifier studies. The association with these genes was replicated in an independent collection of North American data from the consortium, and in the French CF cohort (p=0.001 and p=0.0001 for SLC6A14 and SLC26A9, respectively). None of the associated SNPs are in the coding regions of these genes, but rather are located just upstream of their respective transcription start sites, suggesting that the variants may affect gene expression.

The associated SNPs, however, account for less than 5% of the phenotypic variance in MI, suggesting other genetic factors are yet to be identified, and that there is potential for substantial genetic heterogeneity. To try to identify additional genes contributing to the MI variance, we developed the Hypothesis-Driven GWAS, which provides a way to incorporate hypothesized mechanisms for MI based on known biology into the assessment of genetic association. The Hypothesis-driven GWAS is based on the observation that CF disease is caused by impaired fluid and electrolyte flux at the epithelial interfaces of all CF-affected organs including the airway, intestine, pancreas, liver and vas deferens. In all these organs, the single cell epithelial layer is unique in that it forms a highly selective and tight barrier between body organ and ductal interfaces. Epithelial "function" is achieved by cell polarization whereby many determinants and regulators of fluid, solute and ion transport, including the CFTR ion channel reside specifically at the apical membrane, with contributing features from basal and lateral surfaces. We hypothesized that with loss of CFTR, other transporters that result in variation in any residual or adapted epithelial functions contributed by apical membrane constituents could modify CF phenotypes, such as MI.

We tested the apical hypothesis, by assigning SNPs of genes that encode proteins that localize to the apical plasma membrane to a high priority group and comparing their association evidence to those genes that are not localized to the apical plasma membrane. Hypothesisdriven GWAS was implemented using two statistical procedures. First, using the prioritized apical membrane list, the stratified false discovery rate control (SFDR) (4) was applied to the data to re-evaluate the initial association evidence for any given SNP at the genome-wide level. Then a permutation-based test was used to determine the statistical significance of the apical hypothesis as a whole, testing all high priority SNPs jointly to assess whether a preponderance of apical constituents contribute to MI susceptibility.

A list of 157 genes was annotated as localized to the apical plasma membrane using GO Consortium data (5). This list included SLC26A9 but not SLC6A14 despite it being localized to the brush border membrane. After applying the GWAS-HD, we observed that the group of constituents from the apical plasma membrane was associated with MI (p=0.0002), and this association was replicated in the independent French cohort (p=0.022). The SFDR indicated that SNPs in two additional genes, SLC9A3 and ATP2B2, also reached genome-wide significance after re-prioritization with the apical hypothesis. The gene-based analysis in the French cohort confirmed the SLC9A3 association. Using multivariate regression modeling we were able to estimate that 36 different genes from the apical plasma membrane gene list, including SLC9A3 and SLC26A9, are associat-ed with MI. These genes along with SLC6A14 can account for~17% of the phenotypic variance.

To conclude, the multiple associated genes suggest the genetic basis of MI is quite heterogeneous, which might explain the lack of consensus across MI modifier studies to date. Thus far, we have been able to account for 17% of the phenotypic variability, However, the predicted heritability of MI exceeds 88%, suggesting that there are many more modifiers to be found. The intestines are affected early in CF, often with clinical manifestations at birth. Additionally, intestinal dysfunction continues throughout life in CF. CF intestinal issues range from the very specific life threatening problem of intestinal obstruction, to those that primarily affect quality of life, such as bloating and flatulence. The impact of CF on the intestines contributes to poor nutrition and failure to thrive, which are strongly associated with worse outcomes. While loss of exocrine pancreatic function is the major contributor to impaired digestion in CF, even with optimal pancreatic enzyme replacement therapy, nutritional problems are still common which points to impaired intestinal function.

Loss of CFTR function results in decreased volume of fluid with abnormal electrolyte composition on the epithelial surfaces and in the lumens of affected organs. As in the airways, pathogenesis in the CF intestine appears to follow the same sequelae as a consequence of loss of CFTR activity: mucus accumulation/obstruction, microbial infection, and inflammation.

There have been relatively few investigations of the human CF small intestine, mostly due to its inaccessibility. Much of our current understanding of CF pathogenesis in the intestine is a result of studies using genetically modified mouse models of CF. Use of transgenic mice has contributed greatly to our understanding of normal electrolyte transport mechanisms in the intestine including the roles of CFTR (1). The major phenotype of CF mice is the lethal obstruction of the distal small intestine. There are over a dozen strains of CF mice, including total knockout of the Cftr gene as well as those with specific mutations in CFTR known to cause disease in humans (2) . Comparison of these different CF strains reveals 10-15% of normal CFTR function is sufficient to avoid intestinal obstruction which has important implications for therapies in development such as CFTR small molecule correctors. The most severely affected mouse strain is a conventional knockout which totally lacks Cftr mRNA and protein expression (3) . Half of such CF mice die before weaning and most of the survivors die shortly after weaning as they begin to consume solid food. Post-weaning intestinal obstruction can be prevented by feeding the mice a liquid diet (4) or giving them an osmotic laxative solution to drink (5) .

The initiating event in CF pathogenesis in the intestine is due to the abnormal fluid produced when CFTR is dysfunctional; secreted epithelial fluid in CF is lower in volume and lacking normal bicarbonate ion levels. The importance of these secretions in CF is shown by the fact that improving the hydration state of the gut using liquid food (4), orally administered osmotic laxative (5), or crossing CF mice with transgenic mice that have disruptions in electrolyte absorption pathways (which contribute to reduced fluid volumes in CF), can prevent intestinal obstruction (6, 7) .

Changes in the CF mouse small intestine largely confirm those reported in human CF, namely: accumulation of mucus, bacterial overgrowth, a mild immune response, altered eicosanoid metabolism, slow intestinal transit, fat malabsorption, and impaired mucosal barrier function. CF mice, also similar to human CF, are of low body weight indicating maldigestion and malnutrition.

When examined histologically, the CF intestine exhibits excessive mucus which physically blocks the intestinal crypt lumen and accumulates along the intestinal surfaces. Blockage of the crypt lumen has been proposed to prevent access of Paneth cell derived antimicrobials to the intestinal lumen, and CF mice are more susceptible to infection with pathogenic bacteria (8) . The mucus that accumulates in the CF mouse intestinal lumen becomes heavily colonized with colonic type bacteria (9) , and bacterial overgrowth happens within 4 days of birth (10) . There is also a loss of bacterial diversity in the CF mouse small intestine (9, 10) . Both the change in composition of bacterial species and their overgrowth are forms of microbial dysbiosis. Microbial dysbiosis has wide-ranging effects not only on intestinal function, but also on the development and behavior of the systemic immune system (11) . The implications of gut microbial dysbiosis in CF are not well understood.

Accompanying microbial dysbiosis of the CF mouse small intestine, there is an activation of innate immune defenses, which includes large infiltrations of neutrophils and mast cells (12) . Microarray analysis showed that there are many changes in gene expression in the CF mouse small intestine, a lot of which are components of the innate immune system (12) . Among the altered genes are several involved in eicosanoid metabolism. There are increases in some phospholipase A2s, which cleave membrane phospholipids to produce arachidonic acid, the precursor for a variety of eicosanoids. Also, there is decreased expression of the major prostaglandin degrading enzymes, prostaglandin dehydrogenase (Hpgd) and prostaglandin reductase (Ptgr1). Coincident with these changes in gene expression there are elevated levels of PGE 2 and PGF 2 a (13). An important functional significance of elevated PGE 2 is that this eicosanoid has a dominant relaxant effect on intestinal circular smooth muscle, thus providing a mechanistic explanation for decreased small intestinal motility that occurs in CF (14) .

Most strains of CF mice have impaired weight gain even though these mice are all pancreatic sufficient. This indicates there are intestinal issues contributing to CF malnutrition. A major nutritional issue in human CF is poor fat digestion/assimilation. Similar to human CF, CF mice have impaired lipolysis as well as impaired post-lipolytic uptake of fatty acids, and the former was improved by inhibition of gastric acid (15) . The mechanisms responsible for impaired fat assimilation are only partly known and a complete understanding of fat maldigestion in CF remains an important goal.

CF mice have been used to test therapeutic approaches to improve intestinal function. Direct eradication of bacterial overgrowth using oral broad spectrum antibiotics reduces inflammation, decreases mucus accumulation, improves barrier function, and improves body weight gain (9, 16) . Interestingly, antibiotics do not improve intestinal motility in CF mice and, in fact, impair motility in wild type mice which is accompanied by changes in eicosanoid metabolism similar to that of control CF mice (14) . These observations illustrate the complex interplay between the microbiota and gut function, and serve as a cautionary tale about potential unintended consequences of manipulating the gut microbiota.

Treatment of CF mice with laxative at weaning sufficiently improves luminal hydration such that there is a significant decrease in mucus accumulation, eradication of bacterial overgrowth, less inflammation, improved motility, improved barrier function, and a small increase in weight gain (10, 17) . Laxatives are commonly used in the CF clinic to treat GI symptoms, but the mouse data suggest that chronic laxative use may be a good prophylaxis for CF patients. This idea needs further investigation.

We will continue to learn important lessons about intestinal dysfunction in CF using mouse models, as well as from newly developed CF pigs and ferrets. One goal is to discover how to improve intestinal function in CF which is expected to further contribute to greater longevity and quality of life. Also, as the pathogenesis of CF intestinal disease has many similarities with CF airway disease, novel therapeutic approaches that benefit the intestine may lead to analogous airway therapies. Because of the central role the normal gut microbiota has for systemic immune system development and function, there is also a real potential that therapies that improve intestinal function in CF may have broader benefits such as reducing the hyperresponsive airway immune responses characteristic of this disease. increased mucus release in the presence of HCO 3 may be due simply to a HCO 3 dependent fluid lavage effect. To determine whether decreased mucus release in the absence of HCO 3 is due to decreased HCO 3 --dependent fluid secretion, we compared fluid secretion rates from closed intestinal sacs with and without HCO 3 in the bath solutions. Increases and decreases in weight were interpreted as fluid secretion and absorption, respectively. 5-HT or PGE2 markedly increased fluid secretion, which was completely blocked by bumetanide (a Na + -K + -2Clcontransporter inhibitor). Removing HCO 3 from the bath had no detectable effect on basal or stimulated fluid secretion. Likewise, DIDS did not reduce fluid secretion. These findings indicate that the HCO 3 -dependent increase in stimulated mucus release is unlikely due to HCO 3 dependent fluid secretion and lavage effect (7).

How does HCO 3 help mucus release? To test our hypothesis that HCO 3 enhances mucus release by neutralizing H + (changing pH) and/or sequestering Ca 2+ , porcine stomach mucin was dissolved with Ca(OH) 2 and divided into equal aliquots. Equal moles of NaCl and NaHCO 3 were added to each aliquot, respectively. HCO 3 decreased both the pH and [Ca 2+ ] dramatically. Subsequently, each mucus solution was filtered through an Immobilon-P membrane. The weight of mucus retained on the filter membrane from the mucus solution treated with NaCl was significantly higher than that treated with NaHCO 3 , indicating that HCO 3 reduced the size of mucus aggregates in solution. Furthermore, to better define the role of pH and [Ca 2+ ] in HCO 3 --induced mucus disaggregation, we replaced the Ca 2+ with NaOH equimolar for OHin the initial mucus solution, and then treated aliquots of the resulting NaOH mucus solution with equimolar NaCl or NaHCO 3 . The pH of the mucus solution treated with NaHCO 3 decreased significantly, but the amount of mucus retained on the filter membrane was the same for mucus solutions with NaCl or with NaHCO 3 . In this experiment, the initial pH of the aliquots of NaOH mucus solutions were similar to the mucus solutions with Ca(OH) 2 , but differed in the concentration of [Ca 2+ ]. These results indicate that the disaggregating effect of HCO 3 is probably due to sequestering the [Ca 2+ ] from the mucus (10) .

To further prove that HCO 3 disaggregates mucus by decreasing [Ca 2+ ] in mucus, we collaborated with Prof. Wei-Chun Chin and Dr. Eric Y. T. Chen at the University of California, Merced. Optical emission spectrometry was used to quantify the relative amount of bound Ca 2+ displaced from mucus after adding HCO 3 -. Porcine stomach mucus with Ca 2+ was added with HCO 3 and assayed at increasing intervals of time. The mucus sample was filtered through an Isopore membrane and redissolved in 1% HNO3 solution. As HCO 3 incubation time increased, the amount of mucus-bound Ca 2+ decreased, showing that HCO 3 readily sequesters bound Ca 2+ from mucus.

To show that HCO 3 can directly disperse aggregated mucus gels, Ca 2+ was added to porcine gastric mucus and equilibrated to form mucus gels. HCO 3 dispersed aggregated mucus gel particles, and significantly decreased the mucus gel in size. Addition of EGTA (Ca 2+ chelator) had a similar effect. Again, these data indicate that HCO 3 likely disperses Ca 2+ cross-linked mucus gels by sequestering free as well as Ca 2+ bound in mucus networks (10) .

Conclusions. Our results suggest that: 1) Normal intestinal mucus release requires CFTR-dependent HCO 3 secretion; and 2) HCO 3 --dependent mucus release is probably not due to a lavage effect of HCO 3 --dependent fluid secretion, but due to mucus gel dispersion upon removing Ca 2+ from the mucus matrix. These findings provide insights for the pathogenesis and perhaps new directions for treatments in CF. Aside from Na + absorption or Clsecretion, correcting the defective management of HCO 3 in CF may be necessary for optimal therapies in CF.

Daniel Gelfond, M.D.

Sciences/Children's Hospital of Buffalo, Buffalo, NY, USA Introduction. CFTR is found throughout the entire gastrointestinal (GI) tract. Its dysfunction contributes to the intestinal problems in patients with CF as well as pancreatic and hepatobiliary disease. Clinical symptoms related to intestinal manifestations are not exclusive to the CF population, however patients with CF often have higher prevalence of intestinal complaints.

In CF patients the incidence of GERD is estimated to be 6-8 times higher when compared to the general population. Children with CF tend to have a higher prevalence of GERD when compared to the adult CF population (1). The principal mechanism causing GERD is transient (inappropriate) relaxation of the lower esophageal sphincter. Esophageal pH recordings in children with CF have a higher esophageal exposure to gastric acid along with increased symptoms of regurgitation and heartburn when compared to non-CF children (2) . Long-term complications of GERD such as esophagitis and esophageal stricture is more common in CF children and adults (3). There is a well recognized relationship between GERD and chronic obstructive pulmonary disease (4). Despite supporting evidence for aggressive GERD therapy to improve pulmonary status, there is poor correlation between severity of GERD and FEV1 or FVC in adult patients (5) . Use of proton pump inhibitors has become the first line of therapy in reducing acid production.

Meconium ileus (MI) is an acute condition of newborn infants due to dehydrated thickened meconium that causes mechanical obstruction in the distal ileum along with failure to pass meconium within the first 24-48 hours after birth. Approximately 20% of infants with CF present with meconium ileus at birth. MI is very suggestive of CF, yet up to half of infants presenting with MI do not carry a diagnosis of CF (6) . Non-CF diseases associated with abnormal gut motility, such as Hirschsprung disease and chronic intestinal pseudo-obstruction, have been associated with MI-like disease, suggesting that decreased peristalsis may allow for increased resorption of water, thus favoring the development of MI. Mouse models of CF have shown that severe neonatal bowel obstruction can occur despite normal pancreatic function. These models also highlight that defective HCO 3 excretion leads to an acidic and dehydrated luminal environment. Lack of HCO 3 likely leads to compacted, dehydrated mucus that contributes to MI. Abnormal intestinal motility may also contribute to the development of MI. Furthermore, there is growing evidence for multiple loci modifier gene involvement in the pathogenesis of MI. Medical management with enema irrigation is successful in up to fifty percent of infants. Some infants with MI require urgent surgical interventions either secondary to failed medical management or complications of enema therapy (7) . Overall outcomes of medical and surgical intervention of infants with MI are encouraging. Understanding of the relationship between MI and genetic predisposition as well as environmental impact on infants with CF clearly deserves further investigations.

Distal intestinal obstruction syndrome (DIOS) occurs in patients of all ages including adults with CF and is predominantly localized to the ileocaecum. Accumulation of viscid fecal material with strong adhesion to villi and crypts of the mucosa is a hallmark of this syndrome. Clinically it is differentiated from constipation by an acute onset of symptoms suggestive of partial or complete intestinal obstruction. DIOS can also present as intermittent symptoms of abdominal pain and distention likely attributed to a partial breakdown and re-accumulation of mucofeculent material over a strongly adherent mucoid mass. Radiographic images suggestive of right lower quadrant fecal loading along with clinical symptoms and physical findings of a palpable mass is sufficient to establish a diagnosis pending exclusion of other surgical and medical causes of right lower quadrant abdominal pain with clinical signs of obstruction (8) . DIOS is seen more frequently with pancreatic insufficient patients as well as those with a prior history of MI or prior episodes of DIOS (8) . In European studies, prevalence of DIOS was found to be up to seven times greater in adult patients compared to pediatric patients with CF (9) . Other mechanisms such as intestinal dysmotility, mucosal inflammation, tissue hypertrophy, fat malabsorption and defective chloride and water secretion into gut lumen are likely to contribute to obstructive process and collectively contribute to DIOS (8) . Pharmacologic management is centered on luminal rehydration with osmotic stool softeners to mobilize the mucus mass and relieve obstruction. Enemas and nasogastric decompression can be used in cases of severe obstruction. Surgical intervention remains the last resort once medical management fails or there are signs of bowel ischemia (8) . Emphasis is also placed on prevention of recurrent episodes of DIOS by adherence with pancreatic enzyme replacement therapy, hydration and routine stool softeners.

Constipation is a frequent complaint in patients with CF and likely shares some of the underlying mechanism with DIOS such as dysmotility and decreased water secretion secondary to CFTR defect. Similar to DIOS, fat malabsorption and prior history of MI were noted to be independent risk factors in development of constipation in the pediatric CF patients (10) . Although overall incidence of constipation in patients with CF did not differ from the general population (32% vs. 34-37% respectively), it was noted to be 1.5 times as prevalent in patients with pancreatic insufficiency when compared to the pancreatic sufficient patients (11) . Despite these observations, there is no correlation between the dose of pancreatic enzymes and self-reported constipation in patients with pancreatic insufficiency (11) . In contrast to the general population where the likely cause of constipation in a pediatric age group is deficiency in dietary fiber and inadequate fluid intake, these factors were not associated with constipation in children with CF (10). Therapeutic options in the CF patient with constipation are not different from the medical management of constipation in general population. Osmotic stool softeners are usually first line of treatment along with infrequent supplementation with enemas or stimulant laxatives.

In CF patients, thick mucus secretions along with intestinal dysmotility can predispose patients to intestinal stasis facilitating development of small bowel bacterial overgrowth (SBBO). Bacterial overgrowth within the segments of the small intestine produce toxic byproducts and metabolites often leading to enterocyte damage, malabsorption and malnutrition. Our current understanding of the pathophysiology of SBBO in CF has been facilitated by studies using the CF mouse model (12) . SBBO can present with symptoms of abdominal pain, distention and diarrhea. There are significant limitations in studying SBBO in patients. Diagnostic modalities such as imaging, culture of duodenal aspirates and breath hydrogen testing all have limitations. Empiric therapy with enteric antibiotics and probiotics is frequently initiated to suppress and potentially eradicate bacterial overgrowth in the small bowel. species, suggesting that the lungs of CF pigs have a host defense defect against a wide spectrum of bacteria. Second, the number of bacteria in the CF pig lung is variable over time, similar to humans with CF, suggesting that not all host defenses are impaired. Third, the CF pig has provided additional insight regarding the temporal and causal relations between inflammation and infection in the CF lung. Newborn CF pig lungs show no inflammation but are less often sterile than controls. Fourth, after introduction of bacteria into CF pig lungs, pigs with CF fail to eradicate bacteria as effectively as wild-type pigs. Finally, ion transport studies in CFTR-/-newborn nasal and tracheal/bronchial epithelia show markedly reduced Cland HCO 3 transport (7). Yet, lack of CFTR does not increase transepithelial Na + or liquid absorption or reduce periciliary liquid depth. These results suggest that impaired bacterial eradication is the pathogenic event that initiates a cascade of inflammation and pathology in CF lungs. Furthermore, these studies suggest that reduced anion permeability, and not increased Na + transport, initiates CF airway disease.

Pedro A. Piedra, M.D.

Respiratory illnesses are a major cause of morbidity and mortality worldwide and respiratory viruses are a major contributor to the burden of respiratory illnesses. In industrialized countries, influenza outbreaks are associated with peaks in cardio-respiratory related hospitalization and deaths among all age groups; respiratory syncytial virus (RSV) is the major cause of bronchiolitis and hospitalization in infants, and rhinovirus is associated with acute exacerbations of chronic respiratory disease such as asthma, chronic obstructive airway disease and cystic fibrosis. Other respiratory viruses such as adenovirus, human metapneumovirus, parainfluenzaviruses, coronaviruses, enteroviruses, human bucavirus, and human respiratory polyomaviruses contribute to the overall societal impact both in disability-adjusted life year (DALY) and financial cost.

Common features related to most of the human respiratory viral pathogens are 1) infection rates are highest in infants and young children; 2) recurrent infections occur throughout life and are milder except in elderly adults and those with co-morbid conditions; 3) illness burden and disease severity is greatest at the extremes of lifethe young and the elderly; 4) virus-specific serum neu-tralizing antibody (maternally derived, passively administered, or infection-induced) protects against severe lower respiratory tract illness; and 5) co-morbid conditions such as chronic lung disease, heart disease, and immunodeficiency increase the likelihood of severe respiratory disease.

Cystic fibrosis, an inherited chronic lung disease, is unique among the co-morbid conditions with regard to virus-related respiratory tract morbidity. Persons with cystic fibrosis maintain an increased risk for experiencing severe viral respiratory illness throughout life because of their progressive pulmonary inflammation and disease. Infants and young children with CF, like their healthy counterparts, have comparable high rates of infection with respiratory viruses. They differ, however, in their severity and duration of illness. They are significantly more likely compared to healthy infants and young children to develop 1) a lower respiratory tract illness; 2) have a slower illness resolution; 3) an increased risk of hospitalization; 4) a persistent deterioration in lung function; and 5) an increased risk for bacterial colonization with Pseudomonas aeruginosa, Staphylococcus aureus and Haemophilus influenzae (1-3). Increasing age normally reduces the risk for severe respiratory disease related to viral infections. Unfortunately, older children with CF experience frequent bouts of respiratory exacerbations. These events are often related to an acute respiratory viral infection. Respiratory viruses, in particular, rhinovirus, influenza virus, and parainfluenza virus are frequently detected during acute exacerbations in older children resulting in clinical deterioration, intravenous antibiotic treatment, and lung inflammation (4) (5) (6) (7) . Improvement in viral diagnostics with use of new molecular diagnostic platforms such as the real-time polymerase chain reaction (rt PCR) can improve our understanding of the role of respiratory viral infections on progressive lung disease experienced in children with CF.

Currently we have limited capabilities in preventing respiratory viruses. The approved products are vaccines against influenza infection and a humanized monoclonal antibody, palivizumab, administered prophylactically for the prevention of RSV-related hospitalization. Palivizumab is recommended for a highly select group of infants and young children with medical conditions that increase the risk of severe RSV disease. One such group is children under two years of age with chronic lung disease. CF children may derive benefit from monthly administration of palivizumab during the RSV season, although few studies are available on the benefit of palivizumab in children with CF.

Influenza vaccine is the only approved vaccine for the prevention of a respiratory viral infection. In the United States, the current influenza vaccine recommendation by the Advisory Committee on Immunization Practices (ACIP) of the Centers for Disease Control and Prevention (CDC) is annual influenza vaccination for all persons 6 months of age and older (8) . There are currently two major types of influenza vaccine that are FDA approved; the live attenuated influenza vaccine (LAIV) and the inactivated influenza vaccine (TIV). LAIV is approved for persons 2 to <50 years old who are in good health and TIV is approved for all persons 6 months of age and older. It is important to note that there are a number of vaccine manufacturers that produce TIV and most are not approved for children less than 4 years of age. Both LAIV and TIV are safe and efficacious with LAIV having superior efficacy compared to TIV in children. The efficacy of LAIV and TIV vary from year to year and in general ranges from 60 to 90%, in part, dependent on how well the vaccine antigens antigenically match the circulating viruses and on the immunocompetence of the host.

There are at least three major vaccination strategies to reduce influenza infection. Vaccination of the individual has a direct benefit by providing immunologic protection. Influenza vaccination coverage in the CF population is very high in the U.S. and Europe. The CF Foundation, utilizing their patient registry from 2006-2007, reported influenza vaccination coverage of approximately 90% in children and adults with CF. Underutilization occurred among Hispanics and persons requiring oxygen therapy. A second strategy is maternal immunization. Vaccination of the pregnant women has both a direct benefit to the pregnant women and an indirect benefit to the newborn infant. Infants born to mothers who are vaccinated during pregnancy have a significantly lower risk of influenza infection during their first six months of life. Recall that the influenza vaccine is not approved for use in infants less than 6 months of age. Maternal vaccination against influenza, unfortunately, has been a neglected practice accounting for the lowest vaccination coverage of all target populations. Another strategy is vaccination of select groups who spread the virus efficiently to others such as school-age children, persons who care for or live with persons at high risk, and healthcare workers. This strategy has a direct benefit to the vaccinated person and an indirect benefit to their contacts. Vaccination of school age children has been associated with indirect protection against influenza related complications in unvaccinated household members and the community while vaccination of healthcare workers has been associated with decreased morbidity and mortality in their patients. Surprisingly healthcare workers have suboptimal coverage. Mandatory influenza vaccination programs, although controversial, appears to be the most effective way to significantly improve influenza vaccination coverage among healthcare workers.

idiopathic bronchiectasis associated with NTM have noted a relatively homogenous phenotype of tall, asthenic, post-menopausal women with CFTR abnormalities noted in approximately 30-50% (9,10). The reasons for this female predominance are not clear but the potential role of female hormones in airway surface defenses has been investigated. Progesterone receptors have been noted at the base of the airway cilia (11) . Progesterone has an inhibitory effect on cilia beat frequency that is blocked by estrogen (12) . The relative change in balance between progesterone and estrogen that occurs after menopause or perhaps with an asthenic morphotype might alter airway clearance to allow retention of aerosolized environmental organisms like the NTM. Indeed, patients with primary ciliary dyskinesia also appear to have a marked susceptibility to NTM infections with a similar prevalence that appears to increase with age as in CF (13) . Investigators have also suggested a potential role of female hormones in macrophage function and intracellular killing of mycobacteria (14) .

While Mycobacterium avium complex has been the predominant species isolated in CF patients, M. abscessus seems to be associated with the most morbidity and mortality. Moreover, there appears to be a biphasic interaction with the host whereby it can be intermittently cultured from the respiratory tract of some patients over a prolonged period with no apparent effect on the clinical course followed by a rapid clinical decline associated with an increase in mycobacterial burden (15) . It is unclear whether factors accounting for this change in disease severity are host related, represent changes in virulence factors of the organism, or both. Recent studies in France have suggested a change in colony morphology from smooth to rough may correlate with this change in virulence (16) . We have not observed a consistent association with clinical course in examination of colony morphology of serial isolates. However, ongoing studies in our lab have focused on identification of possible virulence factors in serial isolates that may be associated with these periods of rapid decline.

A key factor separating tuberculosis from disease caused by NTM has been the apparent mode of acquisition with TB transmission occurring from a human host reservoir and NTM acquisition presumed to be exclusively from an environmental reservoir. Prior epidemiologic studies that have examined the potential for humanto-human transmission of NTM have not found evidence to support this (2). However, a recent outbreak of lethal disease associated with a clonal strain of M. massiliense among CF patients in an adult lung transplant center raises significant concerns for the potential of human-to human transmission among a particularly vulnerable population (17) .

In summary, CF represents a particularly susceptible population to NTM infections. It is likely this susceptibility relates to altered airway clearance favoring retention of inhaled organisms from contaminated environmental aerosols on the airway surface. The reasons for the age and gender associations are not entirely clear. While these organisms affect a minority of CF patients, disease associated with the M. abscessus group can be particularly devastating. The recently described outbreak of M. massiliense among CF patients awaiting transplant might warrant similar infection control considerations for some species as are recommended for other virulent organisms such as Burkholderia cepacia. Identification of specific inducible virulence factors that might allow a more refined approach to infection control and better targeting of aggressive treatment regimens is of key importance.

References: 1. Lowy FD. Staphylococcus aureus infections

Methicillin-resistant-Staphylococcus aureus hospitalizations, United States

Community-associated MRSAresistance and virulence converge

Challenging and emerging pathogens in cystic fibrosis

Staphylococcus aureus protein A induces airway epithelial inflammatory responses by activating TNFR1

Surface proteins and exotoxins are required for the pathogenesis of Staphylococcus aureus pneumonia

Vaccine protection against Staphylococcus aureus pneumonia

Role of a disintegrin and metalloprotease 10 in Staphylococcus aureus alpha-hemolysin-mediated cellular injury

The "a disintegrin and metalloprotease" (ADAM) family of sheddases: physiological and cellular functions

Participation of CD11c(+) leukocytes in methicillin-resistant Staphylococcus aureus clearance from the lung

Signal-dependent membrane protein trafficking in the endocytic pathway

Efficient endocytosis of the CFTR requires a tyrosine-based signal

Multiple endocytic signals in the C-terminal tail of the CFTR

Mu 2 binding directs the CFTR to the clathrin-mediated endocytic pathway

Alpha-AP-2 directs myosin VI-dependent endocytosis of CFTR chloride channels in the intestine

Signals for sorting of transmembrane proteins to endosomes and lysosomes

Misfolding diverts CFTR from recycling to degradation: quality control at early endosomes

Mutations in the amino terminus of the CFTR enhance endocytosis

A mutation in the CFTR generates a novel internalization sequence and enhances endocytic rates

Myosin VI regulates endocytosis of the CFTR

Defective CFTR apical endocytosis and enterocyte brush border in myosin VIdeficient mice

The adaptor protein Dab2 sorts LDL receptors into coated pits independently of AP-2 and ARH

References: 1. Powers ET, Balch WE. Protein folding: Protection from the outside Nature

Cell Biology. The proteome in balance

Emergent properties of proteostasis in managing cystic fibrosis

Reduced histone deacetylase 7 activity restores function to misfolded CFTR in cystic fibrosis

Enhanced beta-lactam elimination during treatment of an acute exacerbation in cystic fibrosis (CF)

Continuous infusion of ceftazidime with a portable pump is as effective as thrice-a-day bolus in cystic fibrosis children

Pilot study of continuous infusion cefepime in adult patients with cystic fibrosis

Pharmacokinetic properties and stability of continuous-infusion meropenem in adults with cystic fibrosis

Continuous versus intermittent infusions of ceftazidime for treating exacerbation of cystic fibrosis

Continuous vs thrice-daily ceftazidime for elective intravenous antipseudomonal therapy in cystic fibrosis

Is there a rationale for the continuous infusion of cefepime? A multidisciplinary approach

The aminoglycosides

Association of aminoglycoside plasma levels with therapeutic outcome in gram-negative pneumonia

Adaptive resistance to tobramycin in Pseudomonas aeruginosa lung infection in cystic fibrosis

Aminoglycoside use in cystic fibrosis: therapeutic strategies and toxicity

Once versus three-times daily regimens of tobramycin treatment for pulmonary exacerbations of cystic fibrosis-the TOPIC study: a randomised controlled trial

Cystic fibrosis pulmonary guidelines: treatment of pulmonary exacerbations

References: 1. Beebe DW, Gozal D. Obstructive sleep apnea and the prefrontal cortex: toward a comprehensive model linking nocturnal upper airway obstruction to daytime cognitive and behavioral deficits

Cognition and chronic hypoxia in pulmonary diseases

The effect of chronic or intermittent hypoxia on cognition in childhood: A review of the evidence

Nocturnal home oxygen in the treatment of hypoxemia cystic fibrosis patients

Cognitive function of children with cystic fibrosis: Deleterious effect of early malnutrition

The effect of infective exacerbations on sleep and neurobehavioral function in cystic fibrosis

American Psychiatric Association: Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition, Text Revision

The diagnosis and treatment of attention deficit-hyperactivity disorder in children and adolescents with cystic fibrosis: A retrospective study

The prevalence and correlates of adult ADHD in the U.S.: results from the National Comorbidity Survey Replication

Psychiatric disorder in a sample of the general population with and without chronic medical conditions

Chronic illness, disability and mental and social well-being: Findings of the Ontario Child Health Study

Genetics of adult attention-deficit/hyperactivity disorder

Behavioural symptoms of attention deficit/hyperactivity disorder in preterm and term children born small and appropriate for gestational age: a longitudinal study

Birth insult interacts with stress at adulthood to alter dopaminergic function in animal models: possible implications for schizophrenia and other disorders

A follow-up study of the influence of early malnutrition on development: behavior at home and at school

Psychiatric characterization of children with genetic causes of hyperandrogenism

Distinguishing between attentiondeficit hyperactivity and fetal alcohol spectrum disorders in children: clinical guidelines

Cardiovascular monitoring of chil-Cardiovascular Nursing

Psychiatric syndromes in individuals with chromosome 18 abnormalities

Affective disorders and other psychiatric diagnoses in children and adolescents with 22q11

Deletion Syndrome

The diagnosis and treatment of attention deficit-hyperactivity disorder in children and adolescents with CF: a retrospective study

Adherence and psychopathology in children and adolescents with CF

A population survey of mental health problems in children with epilepsy

Functional impairments in adults with self-reports of diagnosed ADHD: A controlled study of 1001 adults in the community

Psychoactive substance use disorders in adults with attention deficit hyperactivity disorder: effects of ADHD and psychiatric comorbidity

Psychosocial and behavioral predictors of longevity. The aging and death of the "Termites"

Barriers to treatment adherence for children with CF and asthma: what gets in the way?

Identifying barriers to treatment adherence and related attitudinal patterns in adolescents with CF

Cystic fibrosis

Cystic fibrosis

Cystic fibrosis: pathogenesis and future treatment strategies

Determinants of mortality from cystic fibrosis in Canada, 1970-1989

Update in cystic fibrosis 2009

Early pulmonary inflammation in infants with cystic fibrosis

Bronchoalveolar lavage findings in cystic fibrosis patients with stable, clinically mild lung disease suggest ongoing infection and inflammation

The relationship between infection and inflammation in the early stages of lung disease from cystic fibrosis

Nasal and bronchoalveolar lavage fluid cytokines in early cystic fibrosis

Lower airway inflammation in infants and young children with cystic fibrosis

Cystic fibrosis lung inflammation: early, sustained, and severe

Early pulmonary infection, inflammation, and clinical outcomes in infants with cystic fibrosis

Disturbed myeloperoxidase-dependent activity of neutrophils in cystic fibrosis homozygotes and heterozygotes, and its correction by amiloride

Neutrophil adhesion molecule surface expression and responsiveness in cystic fibrosis

Circulating and airway neutrophils in cystic fibrosis display different TLR expression and responsiveness to interleukin-10

Altered intracellular pH regulation in neutrophils from patients with cystic fibrosis

Expression of the CFTR gene in cells of non-epithelial origin

Human beta-defensin-1 is a salt-sensitive antibiotic in lung that is inactivated in cystic fibrosis

Dysfunctional CFTR inhibits phagocytosis of apoptotic cells with proinflammatory consequences

Antimicrobial lipids: novel innate defense molecules are elevated in sinus secretions of patients with chronic rhinosinusitis

A novel host defense system of airways is defective in cystic fibrosis

The lactoperoxidase system links anion transport to host defense in cystic fibrosis

Thiocyanate transport in resting and IL-4-stimulated human bronchial epithelial cells: role of pendrin and anion channels

PLUNC is a novel airway surfactant protein with anti-biofilm activity

Hyperacidity of secreted fluid from submucosal glands in early cystic fibrosis

Role of epithelial HCO3 transport in mucin secretion: lessons from cystic fibrosis

Cystic fibrosis: impaired bicarbonate secretion and mucoviscidosis

Hyperviscous airway periciliary and mucous liquid layers in cystic fibrosis measured by confocal fluorescence photobleaching

Hyposecretion of fluid from tracheal submucosal glands of CFTRdeficient pigs

Substance P stimulates human airway submucosal gland secretion mainly via a CFTR-dependent process

Cystic fibro-Interferon Cytokine Res

Impaired innate host defense causes susceptibility to respiratory virus infections in cystic fibrosis

Innate inflammatory responses of pediatric cystic fibrosis airway epithelial cells: effects of nonviral and viral stimulation

Cystic fibrosis pigs develop lung disease and exhibit defective bacterial eradication at birth

Disease phenotype of a ferret CFTR-knockout model of cystic fibrosis

Adeno-associated virustargeted disruption of the CFTR gene in cloned ferrets

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

FDA Drug Safety Communication: Ongoing safety review of oral bisphosphonates and atypical subtrochanteric femur fractures

References: 1. Matel JL, Milla CE. Nutrition in cystic fibrosis

Symposium 6: Young people, artificial nutrition and transitional care. The nutritional challenges of the young adult with cystic fibrosis: Transition

European cystic fibrosis bone mineralisation guidelines

Bone disease in cystic fibrosis: What's new?

A roadmap to the brittle bones of cystic fibrosis

Skeletal health of children and adolescents with inflammatory bowel dis

Osteoporosis and inflammation

How can we prevent multisystem complications of cystic fibrosis? Semin Respir Crit Care Med

Intestinal inflammation is a frequent feature of cystic fibrosis and is reduced by probiotic administration

References: 1. Swisher AK. Not just a lung disease: Peripheral muscle abnormalities in cystic fibrosis and the role of exercise to address them

European cystic fibrosis bone mineralization guidelines

The anti-inflammatory effect of exercise

Bone mineral density in cystic fibrosis: benefit of exercise capacity

Bone health, daily physical activity and exercise tolerance in cystic fibrosis patients

Weight-bearing exercise and bone mineral accrual in children and adolescents: A review of controlled trials

Exercise during childhood and adolescence: A prophylaxis against cystic fibrosis-related low bone mineral density?

atypical (e.g., nonclassic) manifestations. References: 1. Rowe SM, Miller S, Sorscher EJ. Cystic fibrosis

Mutations in the cystic fibrosis transmembrane regulator gene and in vivo transepithelial potentials

Effect of VX-770 in persons with cystic fibrosis and the G551D-CFTR mutation

Detection of cystic fibrosis transmembrane conductance regulator activity in early-phase clinical trials

An international randomized multicenter comparison of nasal potential difference techniques

Sputum biomarkers of inflammation in cystic fibrosis lung disease

Year-to-year changes in lung function in individuals with cystic fibrosis

Exacerbations in cystic fibrosis. 1: Epidemiology and pathogenesis

Multiple-breath inert gas washout and spirometry versus structural lung disease in cystic fibrosis

Using index of ventilation to assess response to treatment for acute pulmonary exacerbation in children with cystic fibrosis

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

The effect of dornase alfa on ventilation inhomogeneity in patients with cystic fibrosis

Inhaled corticosteroids and lower lung function decline in young children with CF

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

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

Randomised controlled trials in cystic fibrosis: what, when and how?

Phase contrast X-ray imaging for the non-invasive detection of airway surfaces and lumen characteristics in mouse models of airway disease

Hyperviscous airway periciliary and mucous liquid layers in CF measured by confocal fluorescence photobleaching

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

New clinical diagnostic procedures for cystic fibrosis in Europe

Aerosol antibiotics: considerations in pharmacological and clinical evaluation

Lung deposition of inhaled alpha(1)-proteinase inhibitor in cystic fibrosis and alpha(1)-antitrypsin deficiency

Highlights of a workshop to discuss targeting inflammation in cystic fibrosis

Inhaled dry powder mannitol in cystic fibrosis: an efficacy and safety study

Effectiveness of PTC124 treatment of cystic fibrosis caused by nonsense mutations: a prospective phase II trial

References: 1. Rowe SM, Miller S, Sorscher EJ. Cystic fibrosis

Rescue of DeltaF508-CFTR trafficking and gating in human cystic fibrosis airway primary cultures by small molecules

Rescue of CF airway epithelial cell function in vitro by a CFTR potentiator, VX-770

Effect of VX-770 in persons with cystic fibrosis and the G551D-CFTR mutation

Nanomolar affinity small molecule correctors of defective Delta F508-CFTR chloride channel gating

Structural analog of sildenafil identified as a novel corrector of the F508del-CFTR trafficking defect

Mutations in the cystic fibrosis transmembrane regulator gene and in vivo transepithelial potentials

References: 1. Hodson ME, et al. Aerosol carbenicillin and gentamicin treatment of Pseudomonas aeruginosa infection in patients with CF

Evidence for using nebulised antibiotics

Systematic review of nebulised antibiotics in cystic fibrosis

Intermittent administration of inhaled tobramycin in patients with CF

An 18 month study of the safety and efficacy of repeated courses of inhaled aztreonam lysine

Accurate assessment of adherence

Inhalation of antibiotics in CF

Use of aerosolised antibiotics in patients with cystic fibrosis

Vestibular toxicity due to inhaled tobramycin in a patient with renal insufficiency

Serum tobramcyin levels following delivery of tobramycin via Eflow advanced nebuliser

Tobramycin inhalation powder for P. aeruginosa infection in cystic fibrosis

Effect of chronic intermittment administration of inhaled tobramycin on respiratory microbial flora

Effect of TOBI® on long

Cystic fibrosis lung disease starts in the small airways

Pulmonary exacerbations are associated with subsequent FEV1 decline

National Association of Social Workers. Standards for social work practice in palliative and end of life care

Pulmonary complications of cystic fibrosis

Effect of diseaserelated pain on the health-related quality of life of children and adolescents with CF

Self-reported physical and psychological symptom burden in adults with CF

Chronic pain in cystic fibrosis

The immediate effect of musculoskeletal physiotherapy techniques and massage on pain and ease of breathing in adults with CF

Pain in children and adults with cystic fibrosis: A comparative study

Frequency and self-management of pain, dyspnea, and cough in CF

Testing a model of pain appraisal and coping in children with chronic abdominal pain

Acute rib fracture pain in CF

Prevalence of pain in adults with CF

Pain, coping, and disability in adolescents and young adults with cystic fibrosis: A web-based study

End-oflife care in cystic fibrosis

Effects of lung transplantation on inpatient end of life care in CF

Palliative and end-of-life care in CF: what we know and what we need to know

Cystic fibrosis: End stage care in Canada

Caregiver perspectives on discussions about the use of intensive treatments in CF

Family caregiver perspectives on symptoms and treatments for patients dying from complications of CF

Exaggerated activation of nuclear factor-kappaB and altered iKappaB-beta processing in cystic fibrosis bronchial epithelial cells

A P. aeruginosa toxin that hijacks the host ubiquitin proteolytic system

CFTR regulates phagosome acidification in macrophages and alters bactericidal activity

Diseasecausing mutations in the CFTR determine the functional responses of alveolar macrophages

CFTR expression in human neutrophils and the phagolysosomal chlorination defect in CF

CFTR-mediated halide transport in phagosomes of human neutrophils

Genome wide transcriptome profiling of CF blood neutrophils reveals inflammation, gender and CFTR-dependent expression differences

References: 1. Chmiel JF, Konstan MW. Inflammation and anti-inflammatory therapies for cystic fibrosis

Cystic fibrosis and innate immunity: How chloride channel mutations provoke lung disease

Toll-like receptors as therapeutic targets in CF

Innate immunity mediated by TLR5 as a novel antiinflammatory target for CF lung disease

A common dominant TLR5 stop codon polymorphism abolishes flagellin signaling and is associated with susceptibility to legionnaires' disease

TLR5 as an anti-inflammatory target and modifier gene in CF

Early pulmonary inflammation in infants with cystic fibrosis

Role of IKK and ERK pathways in intrinsic inflammation of CF airways

CFTR inhibition mimics the cystic fibrosis inflammatory profile

Mechanisms of the noxious inflammatory cycle in CF

MAPK signaling pathways regulate IL-8 mRNA stability and IL-8 protein expression in CF lung epithelial cells lines

Calcium-dependent regulation of NF-(kappa)B activation in CF airway epithelial cells

Lower airway inflammation in infants and young children with CF

Lower respiratory infection and inflammation in infants with newly diagnosed cystic fibrosis

Innate immune response in CF airway epithelia: hyperinflammatory?

Lower airway inflammation in infants with CF detected by newborn screening

Post-operative infections in cystic fibrosis and non-cystic fibrosis patients after lung transplantation

Metabolic and Molecular Basis of Interited Disease

Pathogen-host interactions in P. aeruginosa pneumonia

Cystic Fibrosis Foundation Patient Registry

Pathophysiology and management of pulmonary infections in CF

How human neutrophils kill and degrade microbes: an integrated view

Antimicrobial mechanisms of phagocytes and bacterial evasion strategies

Inside the neutrophil phagosome: oxidants, myeloperoxidase, and bacterial killing

Myeloperoxidase: friend and foe

CFTR expression in human neutrophils and the phagolysosomal chlorination defect in CF

CFTR-mediated halide transport in phagosomes of human neutrophils

The role of chloride anion and CFTR in killing of P. aeruginosa by normal and CF neutrophils

Cystic fibrosis since 1938

New concepts of the pathogenesis of CF lung disease

Cystic fibrosis pathogenesis and the role of biofilms in persistent infection

Lung infections associated with CF

Infection and inflammation in CF: a short review

Mechanisms of biofilm resistance to antimicrobial agents

The epidemiology, pathogenesis and treatment of P. aeruginosa infections

The DeltaF508-CFTR mutation results in increased biofilm formation by P. aeruginosa by increasing iron availability

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

Effects of reduced mucus oxygen concentration in airway Pseudomonas infections of CF patients

RSV mediates P. aeruginosa binding to cystic fibrosis and normal epithelial cells

How do viral infections predispose patients to bacterial infections?

High P. aeruginosa acquisition rate in CF

Effects of upper respiratory tract infections in patients with CF

Synergistic role of staphylococcal proteases in the induction of influenza virus pathogenicity

Role of Staphylococcus protease in the development of influenza pneumonia

P. aeruginosa inhibits endocytic recycling of CFTR in polarized human airway epithelial cells

A P. aeruginosa toxin that hijacks the host ubiquitin proteolytic system

Long-distance delivery of bacterial virulence factors by P. aeruginosa outer membrane vesicles

Ras-GAP SH3 domain binding protein (G3BP) is a modulator of USP10, a novel human ubiquitin specific protease. Oncogene

The changing microbial epidemiology in cystic fibrosis

Quorumsensing signals indicate that CF lungs are infected with bacterial biofilms

UCSD-Nature Molecule Pages

A conserved p38 MAP kinase pathway in Caenorhabditis elegans innate immunity

The p38 Immun

TLR5 as an anti-inflammatory target and modifier gene in CF

Innate immunity mediated by TLR5 as a novel antiinflammatory target for CF lung disease

Dysregulated cytokine production in human cystic fibrosis bronchial epithelial cells

Cytokine secretion by CF airway epithelial cells

Altered respiratory epithelial cell cytokine production in CF

P. aeruginosa drives CXCL8 synthesis via redundant toll-like receptors and NADPH oxidase in CFTR∅F508 airway epithelial cells

Oxidative stress and gene regulation

Ph.D. 1,2 and Gene Modifiers for the International Consortium of CF

Dalla Lana School of Public Health

Relative contribution of genetic and nongenetic modifiers to intestinal obstruction in CF

Detection of a cystic fibrosis modifier locus for meconium ileus on human chromosome 19q13

Modifier gene study of meconium ileus in CF: statistical considerations and gene mapping results

Were genomewide linkage studies a waste of time? Exploiting candidate regions within genome-wide association studies

Gene ontology: tool for the unification of biology

S19.3 ROLE OF BICARBONATE IN INTESTINAL MUCUS SECRETION Ning Yang 1 and Paul M

Cystic fibrosis: a disease of vulnerability to airway surface dehydration

Evidence for a primary defect of pancreatic HCO3-secretion in cystic fibrosis

A functional CFTR protein is required for mouse intestinal cAMP-, cGMP-and Ca(2+)-dependent HCO3-secretion

Dual role of CFTR in cAMP-stimulated HCO3-secretion across murine duodenum

Molecular mechanism of product storage and release in mucin secretion. II. The role of extracellular Ca++

Molecular mechanism of mucin secretion: I. The role of intragranular charge shielding

Normal mouse intestinal mucus release requires cystic fibrosis transmembrane regulator-dependent bicarbonate secretion

Altered ion transport by tracheal glands in cystic fibrosis

Hyposecretion of fluid from tracheal submucosal glands of CFTRdeficient pigs

References: 1. Malfroot A, Dab I. New insights on gastro-oesophageal reflux in CF by longitudinal follow up

Gastroesophageal reflux in patients with CF

Complications of gastroesophageal reflux in patients with cystic fibrosis

Chronic obstructive pulmonary disease and its comorbidities

Characteristics of gastroesophageal reflux in adults with CF

Clinical and genetic characteristics of meconium ileus in newborns with and without CF

Issues in the management of simple and complex meconium ileus

Guidelines for the diagnosis and management of distal intestinal obstruction syndrome in cystic fibrosis patients

The agerelated incidence of meconium ileus equivalent in a CF population: the impact of high-energy intake

Pancreatic enzyme therapy and clinical outcomes in patients with CF

Altered transit and bacterial overgrowth in the CF mouse small intestine

Selected physiologic compatibilities and incompatibilities between human and porcine organ systems

The porcine lung as a potential model for CF

Cystic fibrosis pigs develop lung disease and exhibit defective bacterial eradication at birth

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

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

Role of respiratory syncytial virus in early hospitalizations for respiratory distress of young infants with CF

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

Association of respiratory viral infections with pulmonary deterioration in patients with cystic fibrosis

The impact of respiratory viral infections in patients with CF

The role of respiratory viruses in cystic fibrosis

Human metapneumovirus and respiratory syncytial virus infections in older children with CF

Nontuberculous mycobacterial pulmonary disease in CF

Nontuberculous mycobacteria. I: Multicenter prevalence study in cystic fibrosis

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

The epidemiology of nontuberculous mycobacterial diseases in the U.S. Results from a national survey

Centers for Disease Control and Prevention

Isolation prevalence of pulmonary non-tuberculous mycobacteria in Ontario

Pulmonary nontuberculous mycobacterial disease prevalence and clinical features: An emerging public health disease

Pulmonary nontuberculous mycobacterial disease: Prospective study of a distinct preexisting syndrome

Prospective analysis of cystic fibrosis transmembrane regulator mutations in adults with bronchiectasis or pulmonary nontuberculous mycobacterial infection

Sex hormone-dependent activation of the progesterone receptor in airway epithelial cells inhibits cilia beat frequency

Primary ciliary dyskinesia: Diagnostic and phenotypic features

Mycobacterium abscessus infection in CF. Colonization or infection?

Overexpression of proinflammatory tlr-2-signalling lipoproteins in hypervirulent mycobacterial variants

Mycobacterium abscessus subspecies massiliense in a respiratory outbreak in a lung transplant and CF center

Lung transplantation is indicated for patients with chronic end-stage lung disease who are failing maximal medical therapy, or for whom no effective medical therapy exists. The primary goal of lung transplantation is to provide a survival benefit, and several studies demon-strate that the survival of cystic fibrosis patients is improved following lung transplantation.International Society of Heart and Lung Transplantation (ISHLT) guidelines for referral of a cystic fibrosis patient to a lung transplantation program include:

Stuart E. Turvey, M.D., Ph.D. and Christoph J. Blohmke Dept. of Pediatrics, Univ. of British Columbia, Child & Family Research Institute, Vancouver, BC, Canada Lung disease, the major cause of death in cystic fibrosis (CF) is caused by chronic infection and inflammation. Current therapies for CF address airway infection with antibiotics, and airway obstruction using clearance techniques combined with mucolytics. Safe and clinically-acceptable therapies to target airway inflammation are likely to augment current treatments and improve the clinical outcome in CF (1) . Clinical trials with oral corticosteroids, high-dose ibuprofen and azithromycin have demonstrated that anti-inflammatory therapy is beneficial for patients with CF, improving important clinical outcomes such as lung function and body weight. Clinical experience with these relatively non-specific antiinflammatory medications provides "proof of concept" evidence that targeting inflammation can be beneficial in CF. Our challenge is now to identify new, specific antiinflammatory targets.Modulating the function of the innate immune system is a particularly attractive treatment approach, since activation of the innate immune system is central to the inflammatory response occurring in the CF lung (2). Innate immunity relies upon a series of germline-encoded receptors, including Toll-like receptors (TLRs), to sense infectious organisms and to trigger an acute inflammatory response. Work by our group and others suggests that modulation of TLR function, and particularly TLR5 inhibition, has therapeutic potential to improve the inflammatory manifestations of CF (3, 4) . Emerging evidence indicates that the TLR5-flagellin interaction plays a central role in driving the inflammatory response triggered by P. aeruginosa and inhibition of TLR5 normalizes the inflammatory response generated by CF airway epithelial cells following exposure to P. aeruginosa (4) .A powerful translational research approach for identifying and validating new therapeutic targets is through the identification of genetic variants that modify CF clinical phenotypes. The TLR5 gene harbours a polymorphism-TLR5 c.1174C>T (rs5744168)-that encodes a premature stop codon. Since allele T has been shown by our group and others (5) to be associated with significantly impaired flagellin responsiveness, we predicted that CF patients carrying the T allele would have improved clinical outcomes due to the endogenous antiinflammatory effects of this TLR5 variant. In a large representative CF cohort, adults with CF carrying the TLR5 premature stop codon (CT or TT genotype) had improved nutritional status, as measured by zBMI, compared with CF patients homozygous for the common fully-functional allele (CC genotype) (6) . Nevertheless, our results must be interpreted with some caution given that the association between the TLR5 c.1174C>T SNP and zBMI in adult CF patients was modest (P=0.044) and the modifying impact of TLR5 will only be firmly established through replication studies in other CF cohorts.A major challenge confronting our CF research community and our efforts to successfully control pulmonary inflammation is to understand the molecular mechanism(s) mediating inflammatory immune responses in the CF airway. Despite an increasing appreciation of how CF pathogens interact with the airway epithelium, it remains unclear whether inflammation in CF is inherently triggered by loss-of-function CFTR mutations alone or due to the failure of the innate immune system to control infection of the airways. Many studies support the concept that excess inflammation is a fundamental component of CF. For example, even in the absence of culturable infections increased pro-inflammatory molecules are found in bronchoalveolar lavage (BAL) of CF patients (7) and immortalized fetal CF cells showed enhanced activation of pro-inflammatory pathways (8) . While the lack of CFTR channel function (9) and endoplasmic reticulum (ER) stress-induced inflammatory responses (10) have been suggested as potential causes, the mechanisms underlying the exaggerated inflammatory phenotype in CF are not clear. In some studies p38 MAPK and ERK signalling is dysregulated in CF with potentially detrimental consequences including increased activation of pro-inflammatory transcription factors (NF-κB, AP-1) (11, 12) . Other reports suggest that Ca 2+ signalling or an imbalance in the oxidative stress response in CF potentially mediates the enhanced activation of pro-inflammatory signalling (10, 13) .Nevertheless, the evidence supporting a hyperinflammatory phenotype in CF is far from unanimous with CF and early microbe-host interactions. The most devastating anomaly of CF occurs in the lung characterized by chronic bacterial infection, abnormal airway inflammation, extensive neutrophil infiltration and small airway obstruction (2) (3) . CF lung infection has a unique pathogen profile which is distinct from other lung infections. Pseudomonas aeruginosa, Staphylococcus aureus, Haemophilus influenzae, Stenotrophomonas maltophilia, Achromobacter xylosoxidans, Burkholderia cepacia are the most prevalent, among which P. aeruginosa predominates (4) (5) (6) . Strikingly, all the CF organisms except S. aureus are opportunistic pathogens, which do not cause infections in normal hosts but in neutropenic patients (6) . It is not fully understood why CF patients are particularly susceptible to these organisms and how the organisms manage to escape the host defense at the early infection stages when there is little antibiotic selection and no biofilm formation. It is the early microbehost interactions and the early subversion of the host defense that leads to pathogen colonization and subsequently persistent infection in CF lungs.Polymorphonuclear neutrophils and anti-bacterial defense. Neutrophils constitute 50-60% of the circulating leukocyte pool in humans and are the major cell type to combat extracellular bacterial infections (7) . Neutrophils ingest bacteria by a process known as phagocytosis (8) . After the bacteria are contained in the membrane-bounded phagosomes, neutrophils produce toxic oxidants including hydrogen peroxide (H 2 O 2 ) and hypochlorous acid (HOCl) to achieve an efficient killing (9-10). HOCl, the active component in common laundry bleach, is synthesized by myeloperoxidase (MPO) predominantly presented in neutrophils. This MPO-catalyzed chemical reaction, H 2 O 2 + Cl -+ H + Ç HOCl + H 2 O, requires Clas a substrate. Thus, the availability of this anion to neutrophil phagosomes is a rate-limiting factor, which affects the production of HOCl. Such a defect in phagocytic innate immunity may preferentially allow certain bacterial strains to evade the compromised host defense.Neutrophils express CFTR and mobilize this chloride channel to their phagosomes. Human peripheral blood neutrophils express CFTR as noted by immunofluorescent staining. When neutrophils phagocytose bacteria, CFTR is recruited to the phagosomes (11) . PLB-985 cells are premyelocytic cell line. We have established a permanent cell line expressing EGFP-wt-CFTR. After induction with 1.25% DMSO, the cells differentiate into mature neutrophils which internalize opsonized 3-micron Latex beads. CFTR is rapidly mobilized to the nascent phagosomal membrane and persists through phagosomal maturation. Therefore, CFTR recruitment to phagosomal membrane is an early event, which reflects the importance of the CFTR channel in neutrophil functions.

To directly measure chloride levels in neutrophil phagosomes, we conjugated chloride probe to zymosan particles. After phagocytosis by normal or CF neutrophils, fluorescence of the probe was quantitatively measured and the dynamics of the chloride level in phagosomes in response to extracellular chloride level changes were recorded. The data demonstrated that CF neutrophils are deficient in transporting chloride to the organelle (12) . Furthermore, CF neutrophil microbicidal function is impaired, especially in a low chloride environment (13) .Summary. Mounting evidence indicates that CFTR plays an important role in phagocytic host defense, which may contribute to CF lung pathogenesis.The mucus layer coating the gastrointestinal tract is one of the first lines of innate host defense. However, in cystic fibrosis (CF), a hallmark of the disease is the accumulation of abnormally thick and sticky mucus in the lung, intestine, and various other exocrine organs. Although the accumulation of thick mucus is thought likely to play a central role in the development of CF, how mutations in the CFTR gene lead to mucus accumulation has not been determined.Why HCO 3 -? Several hypotheses are proposed to explain the pathogenesis of abnormally thick mucus in CF. Probably the most prevalent is excessive Na + and fluid absorption that dehydrates mucus due to unregulated epithelial Na + channel (ENaC) activity (1). However, it is very difficult to extend this rationale from the airways to other CF affected organs, where ENaC is not expressed, for example, the small intestine. Impaired HCO 3 secretion in major CF-affected organs such as pancreas and small intestine has been documented for many years (2-4), but the possible involvement of HCO 3 with mucus formation has received little attention. The major constituents of normal mucus are mucin glycoproteins, which are large, heavily glycosylated proteins with a defining feature of tandem repeating sequences of amino acids. Intracellularly, the Ca 2+ and H + cations shield the high density of fixed anionic sites on the oligosaccharides that surround the protein cores of the mucin monomers (5, 6) . We hypothesized that HCO 3 as an important biological buffer may play a crucial role in unpackaging and expanding mucins from condensed granules to the final mucus substance by neutralizing H + and /or sequestering Ca 2+ .Can HCO 3 help mucus release in the small intestine? To validate our hypothesis, we designed experiments to determine the impact of HCO 3 on mucus release. We made a simple temperature-controlled mucus perfusion system. Mouse intestinal segments were perfused luminally with glucose-free Ringers solution. The perfusate was collected and assayed by WGA-lectin binding or by PAS staining of the filtrand retained on Immobilon-P film (7) . After testing several agonists, we found that serotonin (5-HT) and prostaglandin E2 (PEG2) are two potent mucus secreting stimulants. Then, stimulated mucus release was tested in the presence and absence of HCO 3 or different inhibitors. After removing HCO 3 from the bath solution, released mucus in perfusate was significantly decreased compared to the HCO 3 -Ringers solution. Further, the Na + -HCO 3 contransporter (NBC) inhibitor, DIDS, also markedly reduce 5-HT and PGE2-induced mucus release. These results suggest that HCO 3 is crucial for normal mucus release. We observed that PGE2 and 5-HT did not induce significant mucus release either in CFTR knock-out mice or in the presence of CFTR inhibitor, glyH-101, indicating that normal mouse intestinal mucus release requires CFTR-dependent HCO 3 secretion (7). Is the HCO 3 --dependent mucus release due to HCO 3 --dependent fluid secretion? In addition to HCO 3 -, fluid secretion also appears to be crucial for mucus transport (7-9), which raises a possibility that

David A. Stoltz, M.D., Ph.D. Internal Medicine, Univ. of Iowa Hospitals & Clinics, Iowa City, IA, USA Despite significant advances in our understanding of CF disease pathogenesis and therapeutic interventions, CF remains a potentially lethal disease. Lung disease causes most of the morbidity and mortality in cystic fibrosis (CF) (1) . Understanding the pathogenesis of the lung disease has been hindered, however, by the lack of an animal model with characteristic features of CF. We chose pigs for the development of a new CF model, because they are more similar to humans than are mice in terms of anatomy, physiology, immune system, biochemistry, life span, size, and genetics (2, 3) . We now know that, within months of birth, CF pigs (both CFTR-/-and CFTR∆F508/∆F508) spontaneously develop features typical of human CF lung disease, including airway inflammation, infection, remodeling, mucus accumulation, and airway obstruction (4) (5) (6) . Findings from the CF pig lung have provided insight into CF disease pathogenesis. First, CF pig lungs contain multiple bacterial Recent work suggests that the airways of cystic fibrosis (CF) patients harbor a vast array of bacteria not previously implicated in CF infections. These findings have fundamentally altered concepts of CF lung disease. For example, culture-independent analysis of bronchoalveolar lavage (BAL) samples indicated that 25% of children harbored organisms not typically associated with CF. Studies using newer methods are even more striking. Infecting populations consisting of between 206-1329 bacterial taxa per patient were found in upper airway samples analyzed using DNA microarrays. If correct, these findings would require that we fundamentally change established ideas about CF infections.

An important outstanding question, however, is how to interpret these results given the potential that sputum, throat and BAL samples can be contaminated with bacteria from the upper airway. Contamination is a concern for a number of reasons. First, upper airways contain abundant flora, and throat, sputum, and bronchoscopy samples can be mixed with upper airway secretions. Second, culture-independent methods employ 20-35 cycles of PCR amplification, thus even trace amounts of contaminating DNA can produce a robust signal. Third, like the lung, the nasal and pharyngeal epithelium produce mucus that could encase resident bacteria. This complicates efforts to wash contaminating bacteria away, or differentiate them from lung organisms by microscopy. Finally, clinical labs typically disregard organisms known to be oropharyngeal flora to avoid confounding results. Research applications of culture-independent methods bypass these criteria. These factors raise the possibility that non-culture based detection methods could overestimate lung microbial diversity due to upper airway contamination.The goals of this study were to answer three key questions. First, what bacterial species are present in the lower airways of CF patients with established chronic infections? Identifying the resident bacteria is critical to understanding the mechanisms producing disease. Second, how does oropharyngeal contamination affect culture-independent measurements performed on upper airway specimens? Addressing this question is vital to interpreting culture-independent measurements of infecting population composition at early disease stages. Third, are there consistent differences in the makeup of bacterial populations in distinct anatomic regions of the lung? Finding such differences might explain the regional variation in disease severity often seen in CF lungs. To answer these questions we devised a sterile dissection technique to isolate secretions from the lobar segments of CF lungs removed at the time of lung transplantation. We measured the composition of infecting populations in these samples using 16S rDNA amplicon pyrosequencing and compared results to paired throat and sputum samples collected hours earlier.

Kenneth N. Olivier, M.D., M.P.H.

Recent epidemiologic studies suggest that cystic fibrosis patients are particularly vulnerable to nontuberculous mycobacterial infections. Prior to 1990, nontuberculous mycobacteria (NTM) had been described in only a handful of CF patients (1). A large cross sectional study that sampled approximately 10% of the U.S. CF population in the mid 1990s with 3 sputum specimens prospectively collected over a year found an overall prevalence of NTM in 13% (2) . This prevalence appeared to exponentially increase with age ranging from 10% in children around age 10 to almost 50% in CF patients over the age of 40. A single site survey of CF patients over the age of 40 found very notable differences in those patients diagnosed early in life (median age of diagnosis around age 2) and later in life (median diagnosis approximately age 27) with the latter group being mostly female and 75% having NTM in their lower airway (3) . Similarly in the general population, CDC surveys of state mycobacteriology lab isolates reported in the 1980s and mid '90s suggested prevalence of 2-8 per 100,000 (4, 5) . However, more recent studies have noted an increasing prevalence over time with prevalence and relative female predominance increasing markedly by decade over age 50 (6) (7) (8) .Single site studies of patients presenting with apparent