Modulator Treatments for Cystic Fibrosis: Effectiveness and Value Final Evidence Report and Meeting Summary June 7, 2018 Prepared for: ©Institute for Clinical and Economic Review, 2018 AUTHORS: ICER Staff/Consultants University of Minnesota Modeling Group Ethan M Balk, MD, MPH Karen M. Kuntz, ScD Associate Professor of Health Services, Policy and Practice Professor Brown University School of Public Health University of Minnesota School of Public Health Thomas A. Trikalinos, MD, PhD Kael Wherry, MS Associate Professor of Health Services, Policy and Practice Graduate Student Brown University School of Public Health University of Minnesota School of Public Health Kristin Mickle, MPH Ian Williamson, MBA Research Lead, Evidence Synthesis Graduate Student Institute for Clinical and Economic Review University of Minnesota School of Public Health Geri Cramer, BSN, MBA Research Lead, Evidence Synthesis Institute for Clinical and Economic Review Rick Chapman, PhD, MS Director of Health Economics Institute for Clinical and Economic Review Sonya Khan, MPH Program Director, Midwest CEPAC Institute for Clinical and Economic Review Daniel A. Ollendorf, PhD Chief Scientific Officer Institute for Clinical and Economic Review Steven D. Pearson, MD, MSc President Institute for Clinical and Economic Review DATE OF PUBLICATION: May 3, 2018 Ethan Balk and Tom Trikalinos served as the lead authors for the report. Geri Cramer and Kristin Mickle led the systematic review and authorship of the comparative clinical effectiveness section. Rick Chapman was responsible for oversight of the cost-effectiveness analyses. Geri Cramer and Rick Chapman developed the budget impact model. Sonya Khan authored the section on coverage policies and clinical guidelines. Dan Ollendorf and Steven Pearson provided methodologic guidance on the clinical and economic evaluations. The role of the University of Minnesota (UMN) School of Public Health Modeling Group is limited to the development of the cost-effectiveness model, and the resulting ICER reports do not necessarily represent the views of the UMN. We would also like to thank Ariel Jurmain, Erin Lawler, Molly Morgan, David Whitrap, and Leslie Xiong for their contributions to this report. ©Institute for Clinical and Economic Review, 2018 Page i Final Evidence Report – Cystic Fibrosis About ICER The Institute for Clinical and Economic Review (ICER) is an independent non-profit research organization that evaluates medical evidence and convenes public deliberative bodies to help stakeholders interpret and apply evidence to improve patient outcomes and control costs. Through all its work, ICER seeks to help create a future in which collaborative efforts to move evidence into action provide the foundation for a more effective, efficient, and just health care system. More information about ICER is available at http://www.icer-review.org. The funding for this report comes from government grants and non-profit foundations, with the largest single funder being the Laura and John Arnold Foundation. No funding for this work comes from health insurers, pharmacy benefit managers, or life science companies. ICER receives approximately 20% of its overall revenue from these health industry organizations to run a separate Policy Summit program, with funding approximately equally split between insurers/PBMs and life science companies. For a complete list of funders and for more information on ICER's support, please visit http://www.icer-review.org/about/support/. About Midwest CEPAC The Midwest Comparative Effectiveness Public Advisory Council (Midwest CEPAC) – a core program of ICER – provides a public venue in which the evidence on the effectiveness and value of health care services can be discussed with the input of all stakeholders. Midwest CEPAC seeks to help patients, clinicians, insurers, and policymakers interpret and use evidence to improve the quality and value of health care. The Midwest CEPAC is an independent committee of medical evidence experts from across the Midwest, with a mix of practicing clinicians, methodologists, and leaders in patient engagement and advocacy. All Council members meet strict conflict of interest guidelines and are convened to discuss the evidence summarized in ICER reports and vote on the comparative clinical effectiveness and value of medical interventions. More information about Midwest CEPAC is available at https://icer-review.org/programs/midwest-cepac/. The findings contained within this report are current as of the date of publication. Readers should be aware that new evidence may emerge following the publication of this report that could potentially influence the results. ICER may revisit its analyses in a formal update to this report in the future. ©Institute for Clinical and Economic Review, 2018 Page ii Final Evidence Report – Cystic Fibrosis In the development of this report, ICER’s researchers consulted with several clinical experts, patients, manufacturers and other stakeholders. The following clinical experts provided input that helped guide the ICER team as we shaped our scope and report. None of these individuals is responsible for the final contents of this report or should be assumed to support any part of this report, which is solely the work of the ICER team and its affiliated researchers. For a complete list of stakeholders from whom we requested input, please visit: https://icer- review.org/material/cf-stakeholder-list/ Expert Reviewers Manu Jain, MD Professor, Department of Medicine (Pulmonary and Critical Care); Department of Pediatrics Northwestern University Dr. Jain is a member of the Vertex Pharmaceuticals Advisory Board, and a Site PI for Vertex Phase 2 and 3 studies. He has received more than $5,000 in honoraria or consultancies during the previous year. Brian P. O’Sullivan, MD Professor of Pediatric Pulmonology Geisel School of Medicine, Dartmouth College No relevant conflicts of interest to disclose, defined as more than $10,000 in healthcare company stock or more than $5,000 in honoraria or consultancies during the previous year from health care manufacturers or insurers Cystic Fibrosis Foundation • Contributions: CFF has received charitable contributions and/or fees for service >$5,000 from health care companies, including Vertex Pharmaceuticals. • Equity Interests: CFF has the option to acquire equity interests >$10,000 from a pharmaceutical company unrelated to this report. • Intellectual Property: CFF has entered into therapeutic development award agreements that may result in intellectual property and royalty rights from various pharmaceutical companies. • Research Support: CFF provides financial support to the Therapeutics Development Network (TDN) which delivers high-quality clinical trials to CF patients in the search for better therapies and a cure. CFF provides financial support to the Data Safety Monitoring Board whose primary responsibility is to protect the safety and welfare of people with CF who participate in TDN- approved studies. • Other Relationships: CFF facilitated, but did not participate in, the development of the CFF Pulmonary Guidelines: Use of CFTR Modulator Therapy in Patients with CF. • For more information on CFF’s interactions, see www.cff.org/industry. ©Institute for Clinical and Economic Review, 2018 Page iii Final Evidence Report – Cystic Fibrosis Return to Table of Contents Table of Contents Executive Summary ............................................................................................................................ ES1 1. Introduction ....................................................................................................................................... 1 1.1 Background .................................................................................................................................. 1 1.2 Scope of the Assessment ............................................................................................................. 6 1.3 Definitions .................................................................................................................................. 10 1.4 Insights Gained from Discussions with Patients and Patient Groups ........................................ 12 1.5. Potential Cost-Saving Measures in Cystic Fibrosis .................................................................... 13 2. Summary of Coverage Policies and Clinical Guidelines ................................................................... 14 2.1 Coverage Policies ....................................................................................................................... 14 2.2 Clinical Guidelines ...................................................................................................................... 15 3. Comparative Clinical Effectiveness .................................................................................................. 19 3.1 Overview .................................................................................................................................... 19 3.2 Methods ..................................................................................................................................... 19 3.3 Results ........................................................................................................................................ 23 3.4 Summary and Comment ............................................................................................................ 51 4. Long-Term Cost Effectiveness .......................................................................................................... 54 4.1 Long-Term Cost Effectiveness .................................................................................................... 54 4.4 Summary and Comment ............................................................................................................ 79 5. Other Benefits and Contextual Considerations ............................................................................... 81 5.1 Other Benefits ............................................................................................................................ 82 5.2 Contextual Considerations ......................................................................................................... 82 6. Value-Based Price Benchmarks ........................................................................................................ 83 7. Potential Budget Impact .................................................................................................................. 84 7.1 Overview .................................................................................................................................... 84 7.2 Methods ..................................................................................................................................... 84 7.3 Results ........................................................................................................................................ 86 8. Summary of the Votes and Considerations for Policy ..................................................................... 89 8.1 About the Midwest CEPAC Process ........................................................................................... 89 8.2 Voting Results ............................................................................................................................ 91 ©Institute for Clinical and Economic Review, 2018 Page iv Final Evidence Report – Cystic Fibrosis Return to Table of Contents 8.3 Roundtable Discussion and Key Policy Implications .................................................................. 96 References ......................................................................................................................................... 101 Appendix A. Search Strategies and Results........................................................................................ 109 Appendix B. Previous Systematic Reviews and Technology Assessments ........................................ 113 Appendix C. Ongoing Studies ............................................................................................................. 117 Appendix D. Comparative Clinical Effectiveness Supplemental Information .................................... 133 Genetic Specific Data on CFTR Modulators ................................................................................... 134 Forest Plots from Meta-Analysis .................................................................................................... 138 Appendix E. Comparative Value Supplemental Information ............................................................. 145 Appendix F. Evidence Tables .............................................................................................................. 152 Appendix G. Summaries of Public Comments Delivered at Public Meeting...................................... 211 Appendix H. Conflict of Interest Disclosure ....................................................................................... 214 ©Institute for Clinical and Economic Review, 2018 Page v Final Evidence Report – Cystic Fibrosis Return to Table of Contents List of Acronyms Used in this Report AHRQ Agency for Healthcare Research and Quality AE Adverse event BMI Body mass index BSC Best supportive care CADTH Canadian Agency for Drugs and Technologies in Health CF Cystic fibrosis CFF Cystic Fibrosis Foundation CFFPR Cystic Fibrosis Foundation Patient Registry CFQ-R Cystic Fibrosis Questionnaire-Revised CFRD Cystic fibrosis-related diabetes CFTR Cystic fibrosis transmembrane conductance regulator gene CI Confidence interval GI Gastrointestinal HRQOL Health related quality of life IV Intravenous LCI Lung clearance index MCID Minimum clinically important difference NIH National Institute of Health NICE National Institute for Health and Care Excellence (UK agency) PEx Pulmonary exacerbation PERT Pancreatic enzyme replacement therapy ppFEV1 Percent predicted forced expiratory volume in 1 second SAE Serious adverse event USPSTF United States Preventive Services Task Force VC Vital capacity WAC Wholesale acquisition cost WTP Willingness to pay ©Institute for Clinical and Economic Review, 2018 Page vi Final Evidence Report – Cystic Fibrosis Return to Table of Contents Executive Summary Background Cystic fibrosis (CF) is a progressive genetic disease that affects many organ systems, though most of its morbidity and mortality is associated with its impact on the respiratory system. According to the Cystic Fibrosis Foundation Annual Report, there were 30,000 individuals living with CF in the US in 2016.1 Given that the eligible patient populations for treatment with the drugs under review in this assessment were under 10,000 for each drug, we performed this review using ICER’s framework for treatments of ultra-rare disorders (https://icer-review.org/wp-content/uploads/2017/11/ICER- Adaptations-of-Value-Framework-for-Rare-Diseases.pdf). The pathogenesis of CF is linked to the cystic fibrosis transmembrane conductance regulator (CFTR) gene. In epithelial cells, the CFTR gene is transcribed and translated to produce the CFTR protein, which is in turn, transported to the apical membrane, the part of the membrane that faces inwards towards the open lumina of an organ, such as the airways within the lung. There the protein acts as a chloride ion gate and contributes to the regulation of salt transport in and out of the cell. Mutations to the CFTR gene lead to thickened secretions in the lung, gastrointestinal tract, pancreas, and other organs. Due in part to the thickened lung secretions, people with CF commonly have frequent acute pulmonary infections requiring antibiotic treatment and hospitalization. Ultimately, most people with CF suffer progressive damage to their airways, leading to bronchiectasis and ultimately to respiratory failure, which is responsible for the majority of CF- related deaths. A little over 300 different mutations are known to cause CF.2 Patients with CF carry pathogenic mutations in both copies of the CFTR gene. The most common pathogenic mutation is the F508del mutation. About 86% of all patients have at least one copy of the mutation; these patients are approximately evenly split between homozygous (two copies of the mutation) and heterozygous (one copy of F508del and another mutation).3,4 Another relatively common mutation is G551D, which is found in approximately 5% of CF patients.3 In patients with at least one copy of G551D some of the protein folds correctly, but when it reaches the apical membrane it does not open appropriately to let chloride ions flow normally. The impact of CF and the complexity of its management are associated with multiple physical and social challenges as well as economic insecurity, which can severely affect the quality of life of CF patients, their caretakers, and the rest of their families. ©Institute for Clinical and Economic Review, 2018 Page ES1 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Management The core treatment regimen for CF has historically aimed to control symptoms. It includes aggressive airway hygiene with chest physiotherapy, airway clearance devices, bronchodilators, inhaled and systemic antibiotics, inhaled hypertonic saline, and aerosolized recombinant human DNase to reduce sputum thickness. Also helpful is management of the diet, with pancreatic enzyme replacement therapy and insulin if necessary. The treatment burden for CF patients is high, with patients reporting that they spend upwards of two hours a day completing treatment activities.5 Lung transplantation remains the last-line intervention for CF patients with end-stage disease. Patients who undergo successful lung transplantation no longer suffer from CF in their lungs but continue to have symptoms related to CF in other organ systems. While improvements in supportive care have improved the prognosis for CF patients, these treatments are directed only at symptom management. Recently introduced agents known as CFTR modulators directly address the pathophysiology of the disease and are the focus of this review. CFTR modulator drugs Modulator drugs increase CFTR-mediated ion transport. Two types of modulator drugs have been developed, with complementary modes of action. The effectiveness of modulators depends on the CF-causing mutation. CFTR potentiators, such as Kalydeco® (ivacaftor monotherapy), increase the likelihood that the CFTR channel will transport ions through the cell membrane, i.e., they increase the channel’s “open probability”. Kalydeco has been approved for patients with various “gating” (e.g. G551D) and other mutations that result in residual CFTR protein function in the cell membrane (e.g., R117H). CFTR correctors, such as lumacaftor and tezacaftor, increase the amount of normal or mutated CFTR protein that gets transported to the apical (luminal) membrane, thereby increasing the amount of CFTR protein on the cell surface. Orkambi® (lumacaftor/ivacaftor) and Symdeko™ (tezacaftor/ivacaftor) are considered in patients homozygous for the F508del mutation. While Symdeko has also been studied in patients who are heterozygous for the F508del allele with a residual function mutation, it was approved by the FDA in February 2018 not only for these populations but for other mutations potentially responsive to Symdeko based on laboratory assessments.6 For the purposes of this report we use trade names to facilitate ease of interpretation of the data, with the exception of unapproved doses of lumacaftor with ivacaftor. ©Institute for Clinical and Economic Review, 2018 Page ES2 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Insights Gained from Discussions with Patients and Patient Groups We held semi-structured discussions via teleconference with parents of children with CF as well as with adult patients with CF and identified several cross-cutting themes. The first theme identified from these discussions pertained to aspects of the CF experience that have a strong impact on quality of life from the patient’s and family’s perspective. First, daily care is demanding. Aggressive airway hygiene, a mainstay of standard CF management, is a time- consuming process. Additionally, patients routinely take many pills and inhalation treatments as part of standard care and are concerned by the prospect of even more interventions (e.g., more pills for the modulator treatments, or additional medications to manage emerging complications of CF, such as CF-related diabetes). The high daily demands of standard care take a toll on patients and caregivers. Second, CF patients often endure frequent and severe complications from their disease. Hospitalizations typically last for many days or weeks leading to substantial time lost from school, work, and leisure for both patients and caregivers. Hospitalizations and specialized care can be associated with additional logistical hindrances and expenses if it is necessary to travel to a facility with experience in CF management. Third, even so-called minor complications of CF are pervasive and reduce quality of life. For example, chronic sinusitis can be accompanied by the inability to smell or taste foods, which reduces appetite and contributes to malnutrition. All of the above can greatly limit the ability of CF patients to participate in the social, athletic, work, and other functions that their peers engage in. Another theme in our discussions with patients and caregivers reflected the challenges of adhering to CF management. The daily management of CF is demanding; skipping airway hygiene for a day creates time for other activities and may not have an immediately perceptible negative impact on clinical function. Thus, children or young adults who move on to the next stage of their lives (e.g., leaving home to go to college) may be tempted to lapse in terms of adherence. A third theme was related to financial insecurity due to management of the disease. While all patients we spoke with had insurance coverage, their co-payments varied for CF-related treatment. Uncertainty about future insurance coverage for treatments was also commonly raised. Additional expenses are associated with hospitalizations including travel, accommodation, arranging for care of other children, and other concerns. Further, parents with inflexible work schedules risk losing their jobs after exhausting their sick time. Comparative Clinical Effectiveness We evaluated evidence of the efficacy, safety, and effectiveness of CFTR modulators in comparison with other CFTR modulators or placebo in our target population of individuals with cystic fibrosis. We included any age group with a genetic mutation for which a CFTR modulator had been or was expected to be approved. Comparative trials of CFTR modulators (vs. other intervention or ©Institute for Clinical and Economic Review, 2018 Page ES3 Final Evidence Report – Cystic Fibrosis Return to Table of Contents placebo) were typically powered to detect differences in the change from baseline in percent predicted forced expiratory volume in 1 second (ppFEV1), a measure of respiratory function. While we abstracted both change from baseline and differences between treatment groups, we note that there is no universally agreed-upon definition of a clinically-important difference given the substantial heterogeneity in respiratory function inherent in CF.7 We also captured data on the following additional outcomes: mortality, pulmonary exacerbation, weight and body mass index (BMI), and quality of life. We also sought patient-reported outcome data and incorporated it in the review if available. We sought evidence on harms from any study design. We evaluated treatment in three distinct populations: 1. Kalydeco for patients with gating and residual function mutations. This included individuals with G551D and non-G551D gating mutations and those with R117H residual function mutations. 2. Orkambi and Symdeko for individuals homozygous for the F508del mutation. 3. Symdeko and Kalydeco for individuals heterozygous for the F508del mutation with a second mutation amenable to Symdeko. We first describe the evidence regarding clinical benefits for each population. Next, we describe the evidence on harms for the CFTR modulators collectively. 1. Kalydeco for patients with gating and residual function mutations Clinical Benefits Key Findings: Children, adolescents, and adults with G551D and non-G551D gating mutations experienced statistically significant and clinically meaningful gains in ppFEV 1 and reductions in the rate of pulmonary exacerbations with Kalydeco compared to placebo in 24-week studies. Longer- term follow-up suggests lung function improvements, including reduced rates of pulmonary exacerbations, are durable through three years. Limited evidence also suggests one-year reductions in rates of death, organ transplantation, and hospitalizations. Statistically significant gains in body weight and respiratory symptom-related quality of life with Kalydeco were reported for G551D and non-G551D gating mutation populations aged 12 and older compared to placebo. Statistically significant improvements in lung function or weight were not observed in adult patients with R117H residual function mutations. In a small sample of children aged 6 to 11 years with R117H residual function mutations, those on Kalydeco experienced statistically significant worsening of lung function and trended towards decreased respiratory symptom-related quality of life scores compared to placebo. ©Institute for Clinical and Economic Review, 2018 Page ES4 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Four randomized controlled trials (RCTs) – STRIVE, ENVISION, KONNECTION, and KONDUCT – evaluated the safety and efficacy of Kalydeco in individuals with at least one G551D, non-G551D gating, or R117H mutation.8-11 All four studies required a baseline ppFEV1 ≥40%. All four trials randomized participants to receive either 150 mg of Kalydeco or placebo twice daily for 24 weeks. A fifth comparative study compared over 1,600 people (implicitly with any relevant mutation) taking Kalydeco with over 8000 matched controls not taking Kalydeco; the conference abstract reported one-year follow-up data.12 We also evaluated three noncomparative studies: KIWI,13 a Phase III single-arm study that included children aged 2-5 with a G551D gating mutation; GOAL,14 a longitudinal cohort study of individuals aged 6 years and older with at least on G551D mutation; and PERSIST,15 which followed eligible STRIVE and ENVISION participants for an additional 96 weeks on Kalydeco. Study findings are summarized in Table ES 1 below. For people 6 years and older with gating mutations (G551D and non-G551D), studies have mostly found improvements in the primary pulmonary, weight, and quality of life outcomes with Kalydeco compared to placebo over 24 to 48 weeks. Studies have reported significant improvements in ppFEV1 compared to placebo of 10.4 percentage points (95% CI 8.6 to 12.3, by meta-analysis) over 24 to 48 weeks, significant reductions in risk of pulmonary exacerbations (34% vs. 56%, hazard ratio 0.455, P=0.001), increases in weight (2.8 kg or 0.7 kg/m2), and clinically significant improvements in the respiratory domain of the CFQ-R quality of life instrument of about 5 to 10 points, although the difference with placebo was nonsignificant in the study of 6 to 11 year olds with the G551D mutation. Long-term follow-up (96 weeks) of these people on continued Kalydeco treatment found maintenance of their improvements in ppFEV1 (10.7 percentage points, 95% CI 7.3 to 14.1). Other long-term follow-up studies found continued lowered risk of pulmonary exacerbations compared to matched controls on best supportive care (RR 0.64, 95% CI 0.58 to 0.70) and lowered annual risk of death (RR 0.41, 95% CI 0.20 to 0.84). Based on a single study of people with the R117H gating mutation, Kalydeco improved respiratory function and quality of life in people aged 18 years and older; however, among the small subset of study participants 6 to 11 years old, Kalydeco was not more effective than placebo. For those 18 and older, ppFEV1 improved by 5 percentage points and the respiratory domain of CFQ-R improved by 12.6 points. For the 17 children aged 6 to 11 years, ppFEV1 worsened on Kalydeco, going down 6.3 percentage points compared to placebo; the respiratory domain of CFQ-R was also reduced, but not significantly so. In both age groups there were no differences in risk of pulmonary exacerbation (hazard ratio 0.93) or change in BMI. ©Institute for Clinical and Economic Review, 2018 Page ES5 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Table ES1. Summary of Kalydeco (150 mg 2x/day) on Clinical Efficacy Outcomes for G551D-, non-G551D Gating Mutations, and R117H- CFTR Mutations Age Studies ppFEV1 Pulmonary Weight (Diff) CFQ-R Other (RR) Duration (N) (Absolute Diff), Exacerbation Respiratory Percentage Points Domain (Diff) G551D Mutation Randomized Controlled Trials ≥6 yr STRIVE HR 0.455 (0.29, 0.73)† Weight (kg): 10.4 (8.6, 12.3)* 9.7 (6.5 to 13.0)* 48 wk (N=213) ENVISION nd† 2.8 (1.8, 3.8)* Non-G551D Mutation Randomized Controlled Trial ≥6 yr BMI (kg/m2): KONNECTION 10.7 (7.3, 14.1) nd 9.6 (4.5, 14.7) 8 wk (N=39) 0.7 (0.3, 1.0) R117H Mutation Randomized Controlled Trial ≥6 yr BMI (kg/m2): KONDUCT HR 0.93 (nd) 24 wk (N=69) 0.3 (−1.6, 2.1) 6-11 yr (N=17)‡ −6.3 (−12.0, −0.7)§ −6.1 (−15.7, 3.4)§ ≥18 yr (N=50)‡ 5.0 (1.2, 8.8) 12.6 (5.0, 20.3) Any Indicated Mutations (Implied) Nonrandomized Comparative Study ≥6 yr # US cohort nd RR 0.64 (0.58, 0.70) nd nd Death: 0.41 (0.20, .84) 1 yr (N=1256 **) Organ Txp: 0.15 (0.04, 0.59) Hospitalization: 0.64 (0.58, 0.70) Results in bold font are statistically significant. Abbreviations: BMI: body mass index, CFQ-R: Cystic Fibrosis Questionnaire-Revised, Diff: difference between Kalydeco and placebo, HR: hazard ratio, nd: no data (not reported), ppFEV1: predicted percent forced expiratory volume in one second, RR: risk ratio, Txp: transplantation, wk: weeks, yr: year. * Pooled (meta-analyzed). † Pulmonary exacerbations reported only in STRIVE study. ‡ Inconsistent results for different age groups. Only two participants were between 12 and 17 years and were excluded from subgroup analyses. § Favoring placebo. # Implied. ** On Kalydeco, matched with 6000 controls . ©Institute for Clinical and Economic Review, 2018 Page ES6 Final Evidence Report – Cystic Fibrosis Return to Table of Contents 2. Orkambi and Symdeko for patients with homozygous F508del mutation Clinical Benefits Key Findings: Orkambi and Symdeko both provided small but statistically significant improvements in absolute ppFEV1 compared to placebo after 24 weeks of treatment; however, the magnitude of effect varies by age, dose, and baseline lung function. In longer-term follow-up (96 weeks), those on Orkambi had slower decline in ppFEV1 than matched controls. Neither Orkambi nor Symdeko provided statistically significant short-term improvement in BMI or BMI- for-age z score compared with placebo. Both Orkambi and Symdeko provide improved respiratory-related quality of life compared with placebo. Orkambi and Symdeko reduced pulmonary exacerbation events over 24 weeks, including those requiring intravenous antibiotics and hospitalizations, compared with placebo. Indirect comparisons yielded no material differences between Orkambi and Symdeko in key clinical outcomes. Six key studies including four randomized controlled trials, one single arm trial and one long-term, open-label extension study were identified (see Table ES 2).16-20 Two randomized trials of Orkambi (TRAFFIC and TRANSPORT) were analyzed together, with a subsequent open-label extension study.16,19 Three of the trials (and the open-label extension study) evaluated Orkambi in people 12 years or older (mean age 25 years) or children aged 6 to 11 years old. The single arm study also evaluated Orkambi in children aged 6 to 11 years old. The single randomized trial of Symdeko included mostly adults (mean age 26 years). All primary studies evaluated 24 weeks of therapy; the open-label extension followed people for an additional 96 weeks of therapy. TRAFFIC, TRANSPORT and EVOLVE, included people with ppFEV1 between 40% and 90% (mean 60%); the other trial of Orkambi, Ratjen et al., included younger children who had lung function closer to normal (ppFEV1 >70%; mean 90%). The trials evaluated various doses of lumacaftor (all used the same dose of Kalydeco, 250 mg twice daily). TRAFFIC/TRANSPORT evaluated both lumacaftor 600 and 800 mg total daily; the FDA approved dosage for adults is 800/500 mg daily (Orkambi). As study reporting allows, we focus on data for the FDA approved dose. The Orkambi trial of children 6 to 11 years old used the FDA approved dosage of 400/500 mg daily for this age range. The Symdeko trial also used the FDA approved dosage for adults (100/300 mg daily). Study findings are described by therapeutic comparison below and summarized in Table ES 2 on page ES9. ©Institute for Clinical and Economic Review, 2018 Page ES7 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Orkambi People taking Orkambi had modest, but statistically significant, improvements in lung function over six months compared to placebo. Both adults and adolescents 12 and older and children 6 to 11 years had net increases in ppFEV1 of 2.8 (95% CI 1.8 to 3.8) and 2.4 (95% CI 0.4 to 4.4) percentage points, respectively, compared to placebo. The effect of Orkambi on weight was inconsistent across trials. TRAFFIC found no significant difference in weight change compared with placebo, but the identically designed TRANSPORT study found significant weight gain on the drug; pooled analysis found a small, but statistically significant weight increase of 0.24 kg/m2 (95% CI, 0.11 to 0.37) compared to placebo. The open-label extension study found continued weight gain of about 0.75 to 1 kg/m2 over 96 weeks. The randomized trial of children 6 to 11 years old found no differences in weight measures. The respiratory domain of the quality of life measure CFQ-R was statistically significantly different in adolescents and adults between Orkambi and placebo (2.2 points; 95% CI 0.0 to 4.5), although this did reach the recognized clinically important difference of 4.0.21 A similar, though statistically nonsignificant effect was found in the trial of children (2.5 points; 95% CI -0.4 to 5.4). TRAFFIC/TRANSPORT reported a significant reduction in risk of pulmonary exacerbations among those taking Orkambi (rate ratio 0.61, 95% CI 0.49 to 0.76). Similarly decreased rates of pulmonary exacerbations were found in the 96-week extension study (0.65 events/year, 95% CI 0.56 to 0.75). The pediatric trial did not report on pulmonary exacerbations. Symdeko The randomized trial of Symdeko in adolescents and adults reported modest but significant improvements in ppFEV1 compared to placebo after 24 weeks (4.0%, 95% CI 3.1 to 4.8). Symdeko resulted in a clinically and statistically significant improvement in the respiratory domain of CFQ-R (5.1 units; 95% CI 3.2 to 7.0) compared to placebo and significantly lower rate of pulmonary exacerbations (rate ratio 0.65; 95% CI 0.48 to 0.88). However, BMI and BMI z-score were not significantly different between drug and placebo (0.06 BMI units [95% CI -0.08 to 0.20]; -0.04 z score units [95% CI -0.15 to 0.07]). Orkambi vs. Symdeko No study has compared the two CFTR modulators approved for this population. However, by indirect comparison (network meta-analysis) of the two studies of adolescents and adults, we found no statistically significant differences in effects on ppFEV1, pulmonary exacerbations, BMI z-score, or quality of life as assessed using the respiratory domain of the CFQ-R. Detailed results are available in the full report (see Section 3). ©Institute for Clinical and Economic Review, 2018 Page ES8 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Table ES2. Summary of Orkambi and Symdeko on Clinical Efficacy Outcomes for Homozygous F508del CFTR Mutations Age Studies ppFEV1 Pulmonary Weight (Diff) CFQ-R Duration (N) (Absolute Diff), Exacerbation, Respiratory Percentage Points Rate Ratio Domain (Diff) Orkambi* vs. Placebo Randomized Controlled Trials 6-11 yr Ratjen et al. 2.4 (0.4, 4.4) nd BMI: −0.1 kg/m2 (−0.1, 0.3) 2.5 (−0.4, 5.4) 24 wk (N=204) BMI z-score: 0.0 (−0.2, 0.2) ≥12 yr TRAFFIC 2.8 (1.8, 3.8) 0.61 (0.49, 0.76) BMI: 0.24 kg/m2 (0.11, 0.37) 2.2 (0.0, 4.5) 24 wk (N=1108) TRANSPORT BMI z-score: nd Extension Study (vs. Matched Controls) ≥12 yr TRAFFIC 42% slower rate 96 wk (N=2043)† TRANSPORT of decline† Symdeko (100/500 mg) vs. Placebo Randomized Controlled Trial Mean 26 yr EVOLVE 4.0 (3.1, 4.8) 0.53 (0.34, 0.82) BMI: 0.06 kg/m2 (−0.08, 0.20) 5.1 (3.2, 7.0) 24 wk (N=504) BMI z-score: 0.04 (−0.15, 0.07) Symdeko vs. Orkambi Network Meta-Analysis Indirect comparison EVOLVE vs. Tr/Tr 1.2 (−0.1, 2.5) 0.87 (0.53, 1.42) 2.9 (0.0, 5.8) EVOLVE vs. Ratjen BMI z-score: −0.04 (−0.29, 0.21) Results in bold font are statistically significant. Abbreviations: BMI: body mass index, CFQ-R: Cystic Fibrosis Questionnaire-Revised, Diff: difference between Kalydeco and placebo, nd: no data (not reported), ppFEV1: predicted percent forced expiratory volume in one second, Tr/Tr: TRAFFIC/TRANSPORT, wk: weeks, yr: year. * Data are presented for the now-approved dosages of lumacaftor (400 mg/day for children 6-11 years old and 800 mg/day for older patients). † Open label extension study of TRAFFIC/TRANSPORT (n=455) compared with 1588 matched controls. ©Institute for Clinical and Economic Review, 2018 Page ES9 Final Evidence Report – Cystic Fibrosis Return to Table of Contents 3. Symdeko and Kalydeco for patients with heterozygous F508del mutation and a second mutation amenable to Symdeko Clinical Benefits Key Findings: Based on a single short-term (8 week) cross-over trial, Symdeko and Kalydeco both improved absolute and relative ppFEV1 compared with placebo. Symdeko provides a statistically significant benefit over Kalydeco. Clinically-important and statistically significant improvements in respiratory symptom-related quality of life were observed for both Symdeko and Kalydeco compared with placebo. At 8 weeks, BMI and pulmonary exacerbations were not significantly different between the two drugs and compared with placebo, however; the follow-up duration was likely too short to adequately evaluate these outcomes. A single trial, EXPAND, evaluated both Symdeko (100/300 mg daily) and Kalydeco (300 mg daily) monotherapy (compared to placebo) in patients heterozygous for the F508del mutation with a second mutation amenable to Symdeko. EXPAND was a cross-over trial in which participants took drug for only 8 weeks (n=234). Participants were 12 years or older with ppFEV1 between 40% and 90%, and stable lung disease.22 Findings are summarized in Table ES3 on the following page. Compared to placebo, both interventions provided statistically significant improvement in absolute ppFEV1: 6.8 percentage points for Symdeko (95% CI 5.7 to 7.8) and 4.7 percentage points for Kalydeco (95% CI 3.7 to 5.8). While the clinical significance of these improvements is unknown, these are larger in absolute terms than those seen in the homozygous population. Symdeko also resulted in statistically superior improvement compared to Kalydeco (difference 2.1 percentage points; 95% CI 1.2 to 2.9). Symdeko and Kalydeco both yielded clinically and statistically significant improvements in quality of life using the CFQ-R respiratory domain score as compared to placebo (Symdeko 11.1 points, 95% CI 8.7 to 13.6; Kalydeco 9.7 points, 95% CI, 7.2 to 12.2), with no significant difference seen in comparisons between the two drugs. While taking either CFTR modulator, patients had fewer episodes of pulmonary exacerbation (11 and 9 events, respectively) than while taking placebo (20 events), but the differences were not statistically significant. In addition to the randomized trial data reported in Table ES3, EXPAND reported subgroup differences in effects of Symdeko on ppFEV1 based on age. Those less than 18 years old showed a 12.0 percentage point improvement in absolute ppFEV1 (95% CI, 9.3 to 14.8), whereas those 18 years and older saw a 6.0 percentage point increase (4.9 to 7.0); however, data should be interpreted with caution given only 11 patients under the age of 18 received Symdeko. ©Institute for Clinical and Economic Review, 2018 Page ES10 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Table ES3. Summary of Symdeko and Kalydeco on Clinical Efficacy Outcomes for Heterozygous F508del CFTR Mutation Age Study ppFEV1 Pulmonary Weight (Diff) CFQ-R N (Absolute Diff), Exacerbation, BMI, kg/m2 Respiratory Duration Percentage Points Rate Ratio Domain (Diff) ≥12 yr EXPAND Symdeko (100/300 mg) vs. Placebo (Randomized Controlled Trial) N=234 6.8 (5.7, 7.8) 0.54 (0.26, 1.13) 0.34 vs. 0.18 (nd*) 11.1 (8.7, 13.6) 8 wk (cross-over) Kalydeco (300 mg) vs. Placebo (Randomized Controlled Trial) 4.7 (3.7, 5.8) 0.46 (0.21, 1.01) 0.47 vs. 0.18 (nd*) 9.7 (7.2, 12.2) Symdeko (100/300 mg) vs. Kalydeco (300 mg) (Randomized Controlled Trial) 2.1 (1.2, 2.9) 1.18 (0.49, 2.87) 0.34 vs. 0.47 (nd*) 1.4 (−1.0, 3.9) Results in bold font are statistically significant. Abbreviations: BMI: body mass index, CFQ-R: Cystic Fibrosis Questionnaire-Revised, Diff: difference between Kalydeco and placebo, nd: no data (not reported), ppFEV1: predicted percent forced expiratory volume in one second, wk: weeks, yr: year. * Insufficient data to allow calculation of confidence interval; implied nonsignificant. ©Institute for Clinical and Economic Review, 2018 Page ES11 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Harms For all three CFTR modulators, harms were nonserious and generally uncommon. Serious adverse events, as defined by the studies, commonly occurred at the same or lower rates among those taking the CFTR modulators than those taking placebo, including adverse events ascribed to the drugs. No deaths during CFTR modulator trials were related to the drugs. However, reasons for CFTR modulator discontinuation included elevated liver enzymes, creatinine kinase levels, hemoptysis, bronchospasm, dyspnea, pulmonary exacerbation, and rash. Across studies, summary (i.e., meta-analyzed) rates of discontinuation due to adverse events were: • Orkambi 6.3% (95% CI 3.7, 9.6) • Symdeko 2.5% (95% CI 0.1, 8.3) • Kalydeco 1.2% (95% CI 0.3, 2.5) • Placebo 2.1% (95% CI 1.1, 3.4). Chest tightness (“abnormal respiration”) is a concern that we heard from patients and clinicians, however, the adverse event was only sparsely reported in the literature. In TRAFFIC and TRANSPORT, abnormal respiration was more common with Orkambi (800/500 mg) than placebo (8.7% vs. 5.9%); in the open-label extension study, reported rates of abnormal respiration were between 10-17% over 96 weeks.16,19 Of note, those with baseline ppFEV1 <70% reported more chest tightness than those with baseline ppFEV1 ≥70% (11-20% vs. 6-8%).23 A real-world cohort study reported that nearly 20% of patients reported chest tightness.24 Abnormal respiration was not reported to be a concern for Symdeko and clinical data showed no to low reporting of this side effect.18,22 Symdeko also has fewer drug interactions than Orkambi.6,25 Controversies and Uncertainties CFTR modulator data is unfolding, with the evidence base for some regimens limited to a few published studies. Outcomes of interest, particularly related to weight changes and pulmonary exacerbations, are not consistently reported across studies. Thus, conclusions on individual outcomes are based mostly on one or two trials. Evidence of the comparative effects of CFTR modulators (versus placebo) beyond six months is sparse and largely inconclusive; however, with non-comparative data out to three years, Kalydeco effectiveness has been widely accepted in the clinical community for certain mutations. For the homozygous F508del mutation population, there are no trials that directly compare the two treatment options, Symdeko and Orkambi. For the heterozygous F508del mutation population, there is only a very short-term (8 week) crossover trial comparing treatment options to each other or to placebo. A key uncertainty relates to the relationship between improvements in lung function (as measured by ppFEV1) and reductions in the rate of pulmonary exacerbations. While some level of benefit in ©Institute for Clinical and Economic Review, 2018 Page ES12 Final Evidence Report – Cystic Fibrosis Return to Table of Contents lung function was seen in all studies, exacerbations were not measured consistently and benefit was not uniformly seen. While there are structural explanations in some cases (e.g., the 8-week crossover EXPAND study may have been too short to capture differences in exacerbations), the degree to which reductions in exacerbation rates are contingent on or independent from effects on lung function remains uncertain. In addition, data on the durability and nature of CFTR modulator effects on lung function are only just emerging. Specifically, there is evidence indicating that these agents provide improvements in lung function over the short term (albeit to varying degrees depending on agent and population), but information on slowing of the rate of lung function decline over the longer term is not yet mature and still developing. Research on CFTR modulators is hampered by a number of factors inherent to the population of people with CF. CF genetics are highly complex and variable, and the disease affects relatively small populations when considered by type of mutation. In addition, the recent FDA approval of Symdeko was not limited to the population studied in the EXPAND trial, which required at least one F508del mutation. Therefore, we cannot state with any certainty how generalizable the results from EXPAND are to patients with other mutations, for whom outcomes data are currently unavailable. Additionally, where two drugs for the same population are available, there are little head-to-head data. For example, in the homozygous F508del population, we do not have randomized studies looking at Symdeko versus Orkambi. Other patient characteristics are also likely to impact the effectiveness of the drugs. Limited evidence suggests that, in contrast with adults, children with the R117H mutation do not receive a benefit with Kalydeco, while adolescents heterozygous for the F508del mutation may have a greater benefit with Symdeko than adults. Additionally, variation within and across studies in the care delivered as part of CF symptom management increases the difficulty in interpreting the findings regarding added benefits of CFTR modulators. Even within studies, there was wide variation in the concomitant therapies being used by study participants. It is unknown whether there are any interactions between the effect of the CFTR modulators and any of the concomitant therapies. It is possible that the modulators have little incremental benefit when used with some standard of care therapies or, alternatively, that some of the concomitant therapies may enhance their effects. It is also likely that this variability makes even general indirect comparisons between active therapies that we conducted somewhat problematic to interpret. Nearly 85% of people with CF in the United States receive care at accredited CF centers, which provide multidisciplinary clinical care. This high-quality, specialized approach to care has improved survival for people with CF. Many of the CF trials discussed in this report were conducted in such accredited CF centers, thus improvements in health outcomes seen among these patients (those ©Institute for Clinical and Economic Review, 2018 Page ES13 Final Evidence Report – Cystic Fibrosis Return to Table of Contents assumed to be receiving best supportive care) likely reflect added benefits of CFTR modulators. We identified uncertainties, however, regarding whether beneficial gains in survival are distributed unequally due to differences in access to US CF care centers. For example, Canadian CF patients have been living longer since the mid-1990s and currently live, on average, 10 years longer than American CF patients despite higher usage of mucolytics.26,27 When comparing the US and Canada, the difference between Canadian and US survival disappeared when US patients receiving Medicare and Medicaid were excluded from survival data, suggesting CF patients receiving care through public health insurance are missing out on 10 years of life.26,28 It is unclear whether patients are receiving different care depending on their insurance type or whether American CF patients with public insurance are more likely to have important socioeconomic disadvantages that affect their CF management. While long-term studies are underway to evaluate the impact of CFTR modulators on long-term survival, ensuring access to the highest quality CF care in the interim may improve the survival of all CF patients. Percent predicted FEV1 was the primary outcome for most studies. However, it important to note that ppFEV1 is a surrogate measure of disease severity that attempts to measure lung function relative to what is predicted in healthy persons of the same age and sex. Additionally, it remains unclear what magnitude of change in ppFEV1 is clinically relevant. Evaluation of adverse events among people with CF is challenging because the most frequently reported events may be due to the underlying disease, as evidence by the higher rates of adverse events among those taking placebo than CFTR modulators. Finally, cystic fibrosis is a multisystem disease, yet many aspects of the disease have not been systematically researched. Our evaluation of the impact of CFTR modulators is highly dependent on those outcomes measured in the trial data, namely pulmonary function, weight, respiratory symptom-related quality of life and the number, type and annualized rate of pulmonary exacerbations. Summary and Comment Kalydeco for patients with cystic fibrosis caused by gating and residual function mutations: • Kalydeco provides improvements in ppFEV1 (5.0 to 10.7 percentage points in different populations), weight, and respiratory-symptom-related quality of life (9.6 to 12.6 points) for children, adolescents, and adults (over 24 weeks). Longer-term follow-up (up to three years) shows lung function, weight, and quality of life gains are durable across all gating mutations. • However, limited data suggest 6 to 11 year olds with the R117H mutation may not have improved respiratory function and quality of life with Kalydeco treatment. • Pulmonary exacerbations were less frequent (HR=0.46), shorter, and required fewer hospitalizations and intravenous antibiotics for patients taking Kalydeco. ©Institute for Clinical and Economic Review, 2018 Page ES14 Final Evidence Report – Cystic Fibrosis Return to Table of Contents • Fewer patients (across populations) discontinued Kalydeco due to adverse events (1.2%) than with placebo (2.1%). Across all subpopulations, rates of discontinuation due to adverse events and severe adverse events were similar for Kalydeco and placebo. Given the relatively consistent evidence arising from controlled trials of lung function improvement, with clinically significant improvements and associated reductions in pulmonary exacerbations, and with no evidence of significant harms, we have high certainty Kalydeco provides a substantial (moderate-large) net health benefit relative to best supportive care. We therefore assign a rating of “superior” (A) to the comparative clinical effectiveness of Kalydeco in this population. Orkambi for patients with cystic fibrosis caused by two copies of the F508del mutation: • Orkambi improved ppFEV1; however, changes in absolute ppFEV1 were relatively modest (2.4 to 2.8 percentage points). • At 24 weeks, BMI increases with Orkambi among those aged 12 years and older (0.61 kg/m2), which was maintained over the subsequent 96 weeks; but no significant difference was found in a study of younger children. • Treatment improved respiratory symptom-related quality of life in patients age 12 and older (2.2 points); a similar improvement was found in a smaller study of children 6-11 years old, but the effect was not statistically significant. • The rate of pulmonary exacerbation was lower for patients aged 12 and older taking Orkambi (rate ratio = 0.61); data were not reported in the study of younger children. • Chest tightness (abnormal respiration) was reported as a side effect for those taking Orkambi ranging from 8% in the Phase III trials to 20% in a real-world post-approval study. • Rates of discontinuation due to adverse events were higher for Orkambi (4.6%) than for placebo (1.6%) within a trial in this population. Similar results were seen among all studies across populations (6.3% vs. 2.1%, respectively). In two large Phase III trials and an accompanying 96-week open-label extension study, Orkambi provided improvements in ppFEV1 as well as a reduced rate of decline in lung function; however, lung function improvements were modest, and patients also reported drug-drug interactions as well as abnormal respiration and other side effects leading to discontinuation. Thus, for patients homozygous for the F508del mutation, we have high certainty Orkambi provides a small net health benefit relative to placebo (i.e. best supportive care), and therefore assess the evidence to be “incremental” (B). Symdeko for patients with cystic fibrosis caused by two copies of the F508del mutation: ©Institute for Clinical and Economic Review, 2018 Page ES15 Final Evidence Report – Cystic Fibrosis Return to Table of Contents • Treatment with Symdeko improved absolute ppFEV1 (4.0 percentage points) and respiratory-related quality of life (5.1 points) compared to placebo over 24 weeks. No significant differences in weight were reported. • Treatment reduced the rate of pulmonary exacerbation over 24 weeks (rate ratio = 0.53). • In this population, rates of discontinuation due to adverse events were similar for Symdeko (2.8%) and placebo (3.1%). Similar results were seen among all studies across populations (2.5% vs. 2.1%, respectively). A single, parallel-arm, Phase III trial showed a moderate improvement in ppFEV1 with Symdeko, and reductions in the rate of pulmonary exacerbation; however, the trial was relatively short in duration. Discontinuation due to adverse events was lower than seen in the trial of Orkambi. While a single, short-duration trial only provides moderate certainty, for patients homozygous for the F508del mutation, we judge the net health benefit of Symdeko to be “incremental or better” (“B+”), indicating moderate certainty of a small or substantial net health benefit and high certainty of at least a small benefit. Symdeko for patients with cystic fibrosis caused by one copy of the F508del mutation and a second mutation amenable to Symdeko: • Treatment with Symdeko resulted in improvement in absolute ppFEV1 (6.8 percentage points) and respiratory sympton-related quality of life (11.1 points). • The treatment effect on pulmonary exacerbations and BMI was exploratory only, due to small patient numbers and short trial duration (8 weeks). While a single trial showed evidence of improvement in lung function for Symdeko compared with placebo, the study was of short duration (eight weeks) and used a crossover design. Longer-term studies to confirm effects on pulmonary exacerbation and weight gain are necessary. As above, the current trial evidence provides only moderate certainty, but the level of benefit demonstrated suggests that Symdeko provides a small or substantial net health benefit, with high certainty of at least a small net health benefit relative to placebo (i.e., best supportive care). Therefore, we assess the evidence to be “incremental or better” (“B+”) in patients heterozygous for the F508del mutation with an approved residual function mutation. ©Institute for Clinical and Economic Review, 2018 Page ES16 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Long-Term Cost Effectiveness We conducted a cost-effectiveness analysis using a de novo microsimulation model comparing CFTR modulator treatments plus best supportive care to best supportive care alone for CF patients. We modeled the same three populations described in the project scope (i.e., those with gating mutations, homozygous for F508del, and heterozygous for F508del with a residual function mutation potentially responsive to treatment). The CFTR modulators of interest for these three populations were: 1. Gating mutations: Kalydeco (with patients initiating treatment at two years old). 2. Homozygous for the F508del mutation: Orkambi or Symdeko (with patients initiating treatment at six years old). 3. Heterozygous for the F508del mutation and a residual function mutation that are potentially responsive to treatment: Symdeko or Kalydeco (for patients initiating treatment at 12 years old). CF is a condition which falls under ICER’s ultra-rare disease framework. Therefore, we considered whether to adopt dual base-case analyses based on health system and societal perspectives. However, while the impact of this disease can be substantial on patient and caregiver productivity, and informal caregiver time, the impact of treatment with the CFTR modulators on societal costs is not expected to be substantial in proportion to the health system costs, because the drugs do not greatly reduce the daily burdens associated with usual CF supportive care. We therefore present the results from a societal perspective as a scenario analysis rather than as part of a dual base case. Outcomes were estimated over a lifetime time horizon using one-year time increments from treatment initiation until death. The primary health outcome was quality-adjusted life years (QALYs) but we also report life expectancy and the lifetime number of acute pulmonary exacerbations. Costs and health outcomes were discounted at 3% per year. A comprehensive list of model assumptions, along with the rationale for each, is available in Section 4 of the report. The primary model variable was percent predicted forced expiratory volume in one second (ppFEV1), modeled as a continuous variable. For each population, a cohort of CF patients begins the model at the age of drug initiation. Each simulated patient is assigned a ppFEV1 value drawn from a distribution and then experiences annual age-specific declines in lung function. In addition to ppFEV1, the model tracked the values of other variables for each simulated person: weight-for-age z-score, number of acute pulmonary exacerbations per year (defined as exacerbations requiring intravenous antibiotics), pancreatic sufficiency, lung transplantation, and diagnosis of CF-related diabetes or B. cepacia infection. During any given year, a simulated person may experience a change in their ppFEV1, experience one or more pulmonary exacerbations, be diagnosed with diabetes mellitus or B. cepacia infection, or undergo lung transplantation. The annual risk of death is influenced by all of these variables. EQ-5D utility values derived from a sample of cystic fibrosis ©Institute for Clinical and Economic Review, 2018 Page ES17 Final Evidence Report – Cystic Fibrosis Return to Table of Contents patients were assigned based on lung function or receipt of lung transplantation; disutilities were assigned for acute pulmonary exacerbations. For the treatment arms, we allowed the initial ppFEV1 and weight-for-age z-score values to change based on trial results or assumptions in the absence of data. We also allowed the risk of acute pulmonary exacerbation to decrease with treatment, independent of the improvement in ppFEV1. All costs were adjusted to 2017 US dollars using the personal consumption expenditure (PCE) price index. Annual net drug acquisition costs for each was derived from the Federal Supply Schedule (FSS) to determine discounted (net) prices of Kalydeco and Orkambi (Table 4.5).29 As Symdeko was only recently approved by the FDA, information on its net pricing was not yet available. We therefore applied the FSS discount rate for Orkambi (3.2%) to the wholesale acquisition cost (WAC) of Symdeko to arrive at an estimated net price. We assumed that annual CF-related healthcare costs over an individual’s lifetime consisted of three components (not including the cost of the CFTR modulator drugs): disease management, acute pulmonary exacerbations requiring IV antibiotics, and transplant-related costs. Both disease management and pulmonary exacerbation components incorporated a gradient cost structure that was derived from Lieu et al. to reflect increasing costs with increasing disease severity categories (mild, moderate, and severe ppFEV1 categories).30 An age-related adjustment (<18 or 18+) was included in the exacerbation component. To derive current best supportive care costs, we used two average annual cost estimates based on an unpublished analysis of 2016 commercial payer and Medicaid claims data ($130,879 and $83,173 in 2016 US dollars) (S. Grosse, personal communication, April 12, 2018). Transplant-related costs include the one-time cost of receiving a lung transplant followed by an annual cost associated with post-transplantation care. Base-Case Results Overall, all three CFTR modulator therapies provided substantial health benefits (range of 5.0-6.1 gain in discounted QALYs; 3.5-4.3 gain in discounted life years) at a substantial increase in direct medical costs (range of $4.1-$6.3 million in discounted costs) (Table ES4). The incremental cost-effectiveness ratios for Kalydeco for individuals with a gating mutation were approximately $1.5 million and $960,000 per life year and QALY gained respectively (Table ESY). For individuals who are homozygous for the F508del mutation the incremental cost-effectiveness ratios for Orkambi and Symdeko versus best supportive care were approximately $891,000 per QALY and $974,000 per QALY, respectively, and approximately $1.3 million and $1.4 million per life year gained, respectively. For individuals who are heterozygous for the F508del mutation with a residual function mutation, the incremental cost-effectiveness ratios for Kalydeco and Symdeko in this population were approximately $940,000 QALY and $841,000 per QALY, respectively, and approximately $1.3 million and $1.2 million per life year gained, respectively. ©Institute for Clinical and Economic Review, 2018 Page ES18 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Table ES4. Results for the Base Case for CFTR Modulators Plus Best Supportive Care (BSC) Compared to BSC Alone, By Study Population (Discounted at 3% per Year) Population and Average CFTR Drug Cost Total Cost Total Life Years Total QALYs Treatment Number of PEx CF Individuals with A Gating Mutation BSC $0 $2,227,765 32.75 22.16 15.92 Kalydeco Plus BSC $7,443,121 $8,666,308 18.86 26.52 22.65 CF Individuals Homozygous for F508del Mutation BSC $0 $2,108,199 26.02 20.77 14.74 Orkambi Plus BSC $5,847,893 $6,983,336 11.45 24.57 20.21 Symdeko Plus BSC $6,290,005 $7,478,684 13.36 24.70 20.25 CF Individuals Heterozygous for F508del Mutation with Residual Function Mutation BSC $0 $2,081,180 25.51 18.98 12.92 Kalydeco Plus BSC $6,447,156 $7,557,596 10.85 23.07 18.74 Symdeko Plus BSC $5,934,935 $7,091,919 12.68 23.25 18.88 CFTR: Cystic fibrosis transmembrane conductance regulator; PEx: pulmonary exacerbations; QALYS: quality adjusted life years; BSC: best supportive care Table ES5. Incremental Cost-Effectiveness Ratios Compared to Best Supportive Care (BSC) for the Base Case Treatment vs. BSC Cost Per LY Gained Cost Per QALY Gained Cost Per PEx Averted CF Individuals with a Gating Mutation Kalydeco Plus BSC $1,476,543 $956,762 $463,571 CF Individuals Homozygous for F508del Mutation Orkambi Plus BSC $1,280,892 $890,739 $334,495 Symdeko Plus BSC $1,367,400 $974,348 $424,212 CF Individuals Heterozygous for F508del Mutation and Residual Function Mutation Kalydeco Plus BSC $1,340,171 $941,110 $373,541 Symdeko Plus BSC $1,174,508 $840,568 $390,600 BSC: best supportive care; LY: life year; QALY: quality adjusted life years; PEx: pulmonary exacerbation Sensitivity and Scenario Analysis Results To demonstrate effects of uncertainty on both costs and health outcomes, we varied input parameters using available measures of parameter uncertainty (i.e. standard errors) or reasonable ranges to evaluate changes in cost per addition QALY for CFTR modulators plus best supportive care versus best supportive care alone. All analyses were most sensitive to assumptions about lung function-specific utilities, the independent effect of the drugs on the reduction of acute pulmonary exacerbations, and the discount rate; while changes some of these changes resulted in large variation in cost-effectiveness estimates, in no case did the results approach commonly cited thresholds. ©Institute for Clinical and Economic Review, 2018 Page ES19 Final Evidence Report – Cystic Fibrosis Return to Table of Contents We also evaluated the uncertainty in the model parameters simultaneously by conducting a probabilistic sensitivity analysis. For all CFTR modulators in all CF populations evaluated, the number of iterations in which the CFTR modulators were cost-effective at a WTP threshold of $500,000 per QALY or less was 0%. For example, the 95% credible interval for the incremental cost- effectiveness ratios for Kalydeco compared with best supportive care was $669,500 to $1,591,500 per QALY for CF individuals with gating mutations. In a scenario analysis we incorporated the costs associated with lost productivity in individuals with CF. For individuals with a gating mutation we projected that the difference in lifetime (discounted) indirect costs was $31,600. Including productivity losses in the analysis resulted in incremental cost-effectiveness ratios for Kalydeco very similar to those seen in the base case ($952,100 per QALY societal vs. $956,800 per QALY base case). Estimates for the incremental cost-effectiveness ratios for the CFTR modulators for the other two populations also tracked very closely with base case estimates. We did not include impacts on patient educational levels or caregiver costs in this analysis, given the lack of evidence that this varies by lung function or is impacted by CFTR modulators. The addition of direct non-health care costs that are not affected by CFTR modulator treatments would likely result in an increase in total societal costs, due to our modeled increase in life expectancy with modulator therapy. In the base case we assumed that CFTR modifiers would result in 50% of the annual decline in ppFEV1 that would be seen for best supportive care, after a 2-year period without any decline. In another scenario analysis we varied that assumption from 0% (i.e., no declines in ppFEV1 over the individual’s lifetime) to 100% (i.e., the same annual declines as those on best supportive care after the first two years on drug). As an example, for CF individuals with a gating mutation, the incremental cost-effectiveness ratio for Kalydeco was $620,400 per QALY when we assumed that there was no long-term decline in ppFEV1 (i.e., the drug increased ppFEV1 at the start of therapy and individuals’ lung function remained constant for the remainder of their lifetime). Similar declines in ICERs were found with other drugs and populations, but again did not approach commonly-accepted thresholds. Two other scenarios were explored. In one scenario we explored the impact of assuming that ppFEV1 would not fully recover after a pulmonary exacerbation. Assuming a 5% absolute decline in ppFEV1 for each pulmonary exacerbation experienced reduces the cost-effectiveness ratios by approximately 25%. We also examined the impact of allowing an independent increase in utility above that due to lung function improvement. Assuming a 5% increase in utility with CFTR modulator drugs reduced base-case cost-effectiveness ratios by approximately 15%. ©Institute for Clinical and Economic Review, 2018 Page ES20 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Threshold Analyses The annual price for each drug at which the drug for CF individuals with relevant mutations would be cost-effective at thresholds of $50,000, $100,000, $150,000, $200,000, $300,000, and $500,000 per QALY is shown in Table ES6. Table ES6. Threshold Analysis Results Price to Price to Price to Price to Price to Price to Estimated Estimated Achieve Achieve Achieve Achieve Achieve Achieve Annual Annual $100,000 $150,000 $50,000 $200,000 $300,000 $500,000 WAC Net Price per per per QALY per QALY per QALY per QALY QALY QALY CF Individuals with A Gating Mutation Kalydeco $311,719 $309,842 $55,145 $69,142 $83,146 $97,142 $125,149 $181,149 CF Individuals Homozygous for F508del Mutation Orkambi $272,886 $264,090 $55,562 $67,820 $80,063 $92,321 $116,822 $165,824 Symdeko $292,258 $282,850 $53,210 $65,467 $77,718 $89,976 $114,484 $163,501 CF Individuals Heterozygous for F508del Mutation and Residual Function Mutation Kalydeco $311,719 $309,842 $60,295 $74,175 $88,054 $101,934 $129,693 $185,211 Symdeko $292,258 $282,850 $57,921 $71,969 $86,016 $100,071 $128,166 $184,356 WAC: wholesale acquisition cost; QALY: quality adjusted life year gained Since Kalydeco and Symdeko are each used for treatment in two different populations, we also calculated population-weighted threshold prices using estimated numbers of patients in each population (3,000 CF individuals with gating mutations, 8,464 CF individuals homozygous for F508del mutation, and 6,195 CF individuals heterozygous for F508del mutation and residual function mutation). For Kalydeco, the blended annual price across the two relevant populations ranged from approximately $58,600 at the $50,000 per QALY threshold to approximately $183,900 at the $500,000 per QALY threshold. For Symdeko, the blended annual price across the two relevant populations ranged from approximately $55,200 at the $50,000 per QALY threshold to approximately $172,300 at the $500,000 per QALY threshold. Summary and Comment We developed an individual-level microsimulation model to project the lifetime benefits and costs of CFTR modulator therapies for three different CF populations. The drugs increased lung function, increased weight-for-age z-scores, and decreased the number of acute pulmonary exacerbations and lung transplantations over the lifetime of individuals. The drugs did not impact non-lung aspects of the disease, nor did they decrease the need for CF-related supportive care. Overall, all ©Institute for Clinical and Economic Review, 2018 Page ES21 Final Evidence Report – Cystic Fibrosis Return to Table of Contents drugs (plus best supportive care) evaluated were very effective compared with best supportive care alone in all populations studied, with quality-adjusted life year gains ranging from 5.47 to 6.73 (discounted). With (discounted) CFTR drug-related costs ranging from $4.9 million to $7.4 million, the incremental cost-effectiveness ratios of drugs plus best supportive care compared with best supportive care alone were approximately $0.9 million per QALY for all drugs in all populations considered. Our results were robust to variations to parameter estimates, adopting a modified societal perspective, or using life years gained as the health outcome, except for unit drug costs. Other Benefits and Contextual Considerations Our reviews seek to provide information on other benefits offered by the intervention to the individual patient, caregivers, the delivery system, other patients, or the public that would not have been considered as part of the evidence on comparative clinical effectiveness. These elements are listed in the table below. As CFTR modulators were evaluated under ICER’s framework for a serious ultra-rare condition (https://icer-review.org/wp-content/uploads/2017/11/ICER-Adaptations-of- Value-Framework-for-Rare-Diseases.pdf) additional elements appear in the table that are assessed for such conditions. ©Institute for Clinical and Economic Review, 2018 Page ES22 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Other Benefits Table ES7. Potential Other Benefits and Contextual Considerations (Not Specific to Any Disease or Therapy) Potential Other Benefits This intervention offers reduced complexity that will CFTR modulator treatment is often additive to significantly improve patient outcomes. current treatment regimens, and may therefore increase complexity of daily, routine CF care. However, reductions in the rate and/or intensity of pulmonary exacerbations may reduce patient and caregiver burden over time. This intervention will reduce important health disparities No impact identified across racial, ethnic, gender, socio-economic, or regional categories. This intervention will significantly reduce caregiver or As described above, CFTR modulators are not broader family burden. likely to reduce the daily burden of managing CF, but may reduce patient/caregiver burden with regard to managing exacerbations. This intervention offers a novel mechanism of action or CFTR modulators are the first and only approach that will allow successful treatment of many treatments that target the underlying defect in patients for whom other available treatments have failed. the CFTR protein caused by specific mutations in the CFTR gene. This intervention will have a significant impact on improving In patients with FEV1<40%, CFTR modulators the patient’s ability to return to work or school and/or their may increase the patient’s ability to work and overall productivity. improve overall productivity. This intervention will have a significant positive impact No impact identified outside the family, including on schools and/or communities. This intervention will have a significant impact on the entire No impact identified “infrastructure” of care, including effects on screening for affected patients, on the sensitization of clinicians, and on the dissemination of understanding about the condition, that may revolutionize how patients are cared for in many ways that extend beyond the treatment itself. Other important benefits or disadvantages that should have No impact identified an important role in judgments of the value of this intervention. ©Institute for Clinical and Economic Review, 2018 Page ES23 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Contextual Considerations Potential Other Contextual Considerations This intervention is intended for the care of individuals with Cystic fibrosis significantly impacts both length a condition of particularly high severity in terms of impact and quality of life. on length of life and/or quality of life. This intervention is intended for the care of individuals with Patients with cystic fibrosis have a high lifetime a condition that represents a particularly high lifetime burden of illness. burden of illness. This intervention is the first to offer any improvement for While CFTR modulators are the first to target patients with this condition. disease pathology, advancements in supportive care have also greatly improved prognosis for CF patients. Compared to best supportive treatment, there is significant Serious side effects of CFTR modulators appear uncertainty about the long-term risk of serious side effects to be minimal compared to the effects of the of this intervention. underlying disease; however, long-term data are not yet available. Compared to best supportive treatment, there is significant The long-term effects of CFTR modulators on the uncertainty about the magnitude or durability of the long- rate of disease progression are starting to term benefits of this intervention. develop but remain sparse. The magnitude and durability of CTFR modulator benefit has not been reliably quantified at this time. There are additional contextual considerations that should No impact identified have an important role in judgments of the value of this intervention. Potential Budget Impact We used results from the same model employed for the cost-effectiveness analyses to estimate the total potential budget impact of Symdeko in cystic fibrosis, specifically for those heterozygous or homozygous for the F508del mutation. Potential budget impact was defined as the total differential cost of using Symdeko plus best supportive care, rather than relevant existing therapy for the treated population, calculated as differential health care costs (including drug costs) minus any offsets in these costs from averted health care events. We estimated the eligible prevalent population in the United States, derived from the 2016 Cystic Fibrosis Foundation Patient Registry Annual Data Report,1 at 8,464 cystic fibrosis patients over the age of 6 with two copies of the F508del mutation, and 6,195 cystic fibrosis patients over the age of 12 with one copy of the F508del mutation. Table ES8 shows the per-patient budget impact calculations for Symdeko in those homozygous for the F508del mutation relative to current care assuming Orkambi plus best supportive care in 50% and only best supportive care in 50%, based on prescribing rates for Orkambi.1 The average ©Institute for Clinical and Economic Review, 2018 Page ES24 Final Evidence Report – Cystic Fibrosis Return to Table of Contents potential budgetary impact when using the WAC ($292,258) was an additional per-patient cost of approximately $117,300 and approximately $109,100 using the discounted WAC ($282,850). At the three cost-effectiveness threshold prices (at $50,000, $100,000 and $150,000 per QALY), there would be estimated cost savings, because while there would be increased costs from using Symdeko in addition to best supportive care, these additional costs would be more than offset by the replacement of Orkambi at net price by Symdeko at the much lower assumed threshold prices. Table ES8. Per-Patient Budget Impact Calculations Over a Five-year Time Horizon for Individuals Homozygous for F508del Mutation Average Annual Per Patient Budget Impact WAC Discounted $150,000/ $100,000/ $50,000/ WAC QALY QALY QALY Symdeko+BSC $300,749 $292,545 $113,699 $98,765 $92,331 Orkambi+BSC (50%) & $183,418 BSC (50%) Difference $117,331 $109,128 ($69,719)* ($84,653)* ($91,078)* WAC: wholesale acquisition cost; QALY: quality adjusted life year; BSC: best supportive care *Indicates cost-saving Table ES8 shows the per-patient budget impact calculations for Symdeko in those with one F508del mutation and a residual function mutation, compared to current care assuming Kalydeco plus best supportive care in 50% and best supportive care in 50%. The average potential budgetary impact when using the WAC ($292,258) was an additional per-patient cost of approximately $92,800 and approximately $84,600 using the discounted WAC ($282,850). At the three cost-effectiveness threshold prices (at $50,000, $100,000 and $150,000 per QALY), there would be estimated cost savings, again because the increased costs from using Symdeko in addition to best supportive care would be more than offset by the replacement of Kalydeco at net price by Symdeko at the much lower assumed threshold prices. Table ES9. Per-Patient Budget Impact Calculations Over a Five-year Time Horizon for Individuals with F508del Mutation and Residual Function Mutation Average Annual Per Patient Budget Impact WAC Discounted $150,000/ $100,000/ $50,000/ WAC QALY QALY QALY Symdeko+BSC $301,966 $293,776 $122,441 $110,212 $97,983 Kalydeco +BSC $209,185 (50%) & BSC (50%) Difference $92,781 $84,591 ($86,744)* ($98,973)* ($111,202)* WAC: wholesale acquisition cost, QALY: quality-adjusted life year, BSC: best supportive care *Indicates cost-saving ©Institute for Clinical and Economic Review, 2018 Page ES25 Final Evidence Report – Cystic Fibrosis Return to Table of Contents For the combined populations of interest, the annual potential budgetary impact of treating the entire eligible population with Symdeko at the net price over five years is 95% of the $915 million threshold, but exceeded the threshold by 2% using WAC. While the total number of patients eligible for treatment with Symdeko is relatively low (n = 14,659), the increased cost per patient from using Symdeko over the current treatment mix leads to a total estimate approaching the budget impact threshold. Note that this number may actually be understated, because the approved FDA label for Symdeko allows treatment beyond those having at least one copy of the F508del mutation, so long as the mutation is responsive to Symdeko (through in vitro or clinical data).6 Table ES10. Estimated Total Potential Budget Impact of Symdeko for Treatment of Eligible Populations Using Net Prices Over a Five-year Time Horizon Eligible N Treated per Annual BI per Total BI Percent of Population Year Patient (millions) Threshold Homozygous F508del Symdeko 8,464 1,693 $109,128 $552,527,040 60% Heterozygous F508del with Residual Function Mutation Symdeko 6,195 1,239 $84,591 $312,510,796 34% Total Eligible US CF Population* Symdeko 14,659 2,932 $172,274 $865,037,837 95% BI: budget impact * Annual BI per patient for total US CF population weighted by percentage contribution. Value-Based Price Benchmarks Our value-based benchmark prices for Kalydeco, Orkambi, and Symdeko are presented in Table ES11. As Kalydeco and Symdeko are each used for treatment in two different populations, we calculated blended threshold prices weighted by estimated numbers of patients in each population. For each drug, the discounts required to meet both threshold prices (>70%) are much greater than the currently assumed discount from WAC. Table ES11. Value-Based Benchmark Prices for Kalydeco, Orkambi, and Symdeko Annual Price Annual Price Annual Net Discount from Annual to Achieve to Achieve Price (with WAC to Reach WAC $100,000 $150,000 Mark-Up) Threshold Prices per QALY per QALY Kalydeco $311,719 $309,842 $72,533 $86,453 72% to 77% Orkambi $272,886 $264,090 $67,820 $80,063 71% to 75 % Symdeko $292,258 $282,850 $68,215 $81,225 72% to 77% QALY: quality-adjusted life year ©Institute for Clinical and Economic Review, 2018 Page ES26 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Midwest CEPAC Votes The Midwest CEPAC Panel deliberated on key questions raised by ICER’s report during the public meeting on May 17, 2018. The results of these votes are presented below, and additional information on the deliberation surrounding the votes can be found in the full report. Comparative Clinical Effectiveness 1) For individuals with approved gating, non-gating, and residual function mutations (including but not limited to G551D and R117H), is the evidence adequate to demonstrate that the net health benefit of treatment with Kalydeco (ivacaftor) with best supportive care is greater than that of best supportive care alone? Yes: 12 votes No: 0 votes 2) For individuals who are homozygous for the F508del mutation, is the evidence adequate to demonstrate that the net health benefit of treatment with Orkambi (lumacaftor/ivacaftor) with best supportive care is greater than that of best supportive care alone? Yes: 11 votes No: 1 votes 3) For individuals who are homozygous for the F508del mutation, is the evidence adequate to demonstrate that the net health benefit of treatment with Symdeko (tezacaftor/ivacaftor) with best supportive care is greater than that of best supportive care alone? Yes: 12 votes No: 0 votes 4) For individuals who are homozygous for the F508del mutation, is the evidence adequate to distinguish the net health benefit between treatment with Symdeko with best supportive care and Orkambi with best supportive care? Yes: 1 votes No: 11 votes ©Institute for Clinical and Economic Review, 2018 Page ES27 Final Evidence Report – Cystic Fibrosis Return to Table of Contents 5) For individuals who are candidates for Symdeko combination therapy because they carry one F508del mutation and residual function mutation that is potentially responsive to Symdeko, is the evidence adequate to demonstrate that the net health benefit of treatment with Symdeko with best supportive care is greater than that of best supportive care alone? Yes: 11 votes No: 1 votes Other Benefits and Contextual Considerations When compared to best supportive care, does Kalydeko, Orkambi, or Symdeko offer one or more of the following “other benefits”? (yes, no, uncertain) Potential Other Benefits # of votes This intervention offers reduced complexity that will significantly improve patient outcomes. 4 / 12 This intervention will reduce important health disparities across racial, ethnic, gender, socio- 0 / 12 economic, or regional categories. This intervention will significantly reduce caregiver or broader family burden. 8 / 12 This intervention offers a novel mechanism of action or approach that will allow successful 10 / 12 treatment of many patients for whom other available treatments have failed. This intervention will have a significant impact on improving the patient’s ability to return to work 7 / 12 or school and/or their overall productivity. This intervention will have a significant positive impact outside the family, including on schools 3 / 12 and/or communities. This intervention will have a significant impact on the entire “infrastructure” of care, including 2 / 12 effects on screening for affected patients, on the sensitization of clinicians, and on the dissemination of understanding about the condition, that may revolutionize how patients are cared for in many ways that extend beyond the treatment itself. Other important benefits or disadvantages that should have an important role in judgments of the 7 / 12 value of this intervention. ©Institute for Clinical and Economic Review, 2018 Page ES28 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Are any of the following contextual considerations important in assessing Kalydeco’s, Orkambi’s, or Symdeko’s long-term value for money in patients? (yes, no, uncertain) Potential Other Contextual Considerations # of votes This intervention is intended for the care of individuals with a condition of particularly high 12 / 12 severity in terms of impact on length of life and/or quality of life. This intervention is intended for the care of individuals with a condition that represents a 12 / 12 particularly high lifetime burden of illness. This intervention is the first to offer any improvement for patients with this condition. 5 / 12 Compared to best supportive treatment, there is significant uncertainty about the long-term risk of 3 / 12 serious side effects of this intervention. Compared to best supportive treatment, there is significant uncertainty about the magnitude or 10 / 12 durability of the long-term benefits of this intervention. There are additional contextual considerations that should have an important role in judgments of 7 / 12 the value of this intervention. Long-Term Value for Money 1) For individuals with approved gating, non-gating, and residual function mutations (including but not limited to G551D and R117H), given the available evidence on comparative clinical effectiveness and incremental cost effectiveness, and considering other benefits and contextual considerations, what is the long-term value for money of Kalydeco with best supportive care compared with best supportive care alone? Low: 10 votes Intermediate: 2 votes High: 0 votes 2) For individuals who are homozygous for the F508del mutation, given the available evidence on comparative clinical effectiveness and incremental cost effectiveness, and considering other benefits and contextual considerations, what is the long-term value for money of Orkambi with best supportive care compared with best supportive care alone? Low: 11 votes Intermediate: 1 votes High: 0 votes 3) For individuals who are homozygous for the F508del mutation, given the available evidence on comparative clinical effectiveness and incremental cost effectiveness, and considering other benefits and contextual considerations, what is the long-term value for money of Symdeko with best supportive care compared with best supportive care alone? Low: 11 votes Intermediate: 1 votes High: 0 votes ©Institute for Clinical and Economic Review, 2018 Page ES29 Final Evidence Report – Cystic Fibrosis Return to Table of Contents 4) For individuals who are candidates for Symdeko combination therapy because they carry one F508del mutation and residual function mutation that is potentially responsive to Symdeko, given the available evidence on comparative clinical effectiveness and incremental cost effectiveness, and considering other benefits and contextual considerations, what is the long-term value for money of Symdeko with best supportive care compared with supportive care alone? Low: 11 votes Intermediate: 1 votes High: 0 votes Key Policy Implications Following its deliberation on the evidence, the Midwest CEPAC Panel engaged in a moderated discussion with a policy roundtable about how best to apply the evidence on modulator treatments for cystic fibrosis to policy and practice. The policy roundtable members included one patient advocate, one caregiver, two clinical experts, and two payers. The discussion reflected multiple perspectives and opinions, and therefore, none of the statements below should be taken as a consensus view held by all participants. Key Recommendations on Pricing and Access • The prices for CFTR modulators are too high, harming patients and families today while threatening the health care system’s ability to maintain access for all patients to important future clinical advances. Benefiting from monopoly pricing power, the company bears a significant social responsibility to change its pricing approach by committing to the following two actions: • Abandon vague claims that prices are justified by the need to invest in future research and instead join the growing number of biotech innovators who provide a transparent, explicit justification for their prices based on the ability of treatments to improve the length and quality of patients’ lives; • Accept that the process for determining a reasonable price for new drugs requires innovators, especially those with monopoly pricing power at their disposal, to exercise restraint and be open to an independent process to evaluate fair pricing that includes the full engagement of the innovator, patients, patient advocacy groups, clinical experts, insurers, and other stakeholders. • Public and private payers should continue to affirm their commitment to provide access to important clinical advances for CF and should remove superfluous requirements for coverage approval and continuation. ©Institute for Clinical and Economic Review, 2018 Page ES30 Final Evidence Report – Cystic Fibrosis Return to Table of Contents • Since insurance coverage denial for CF drugs is off the table, payers should be willing to develop and adopt new approaches to moderate the impact of monopolistic pricing power. • Patient organizations that have a leading role in funding, organizing, promoting, and otherwise fostering innovative research on new treatments should demand commitments from manufacturers for sustainable pricing of the products patients helped bring to the market. • Professional societies should fully exercise their responsibility by bearing witness to the impact on their patients of failed pricing and insurance policies and by demanding to be part of the public process that should guide pricing to balance the needs for affordability and for investments in future innovation. Recommendations to Improve Future Research • Future studies should measure and report a broad set of outcomes to better assess the health and economic impact of CF interventions to patients, their caregivers, and their health system. • Manufacturer-sponsored research should enroll patients who are often encountered in clinical practice, but who are routinely excluded from clinical trials. • Leverage all available resources to maximize the evidence base. Because CF is relatively rare, effort should be made to maximize use of all existing data, including routinely collected information. ©Institute for Clinical and Economic Review, 2018 Page ES31 Final Evidence Report – Cystic Fibrosis Return to Table of Contents 1. Introduction 1.1 Background Cystic fibrosis (CF) is the most common life-shortening genetic disease in Caucasian populations. Its birth prevalence varies by ethnic descent. In the US approximately 1 in 3,000 Whites are born with CF, but it is less common among in Latinos (1 in 4,000-10,000) and African Americans (1 in 10,000- 20,000). According to the Cystic Fibrosis Foundation Annual Report, the overall prevalence of CF in the US in 2016 was 30,000.1 Although rare, CF represents a substantial economic burden. In 2013, CF-related hospital costs alone were estimated to exceed $1.1 billion.31 Pathogenesis Over 1800 cystic fibrosis transmembrane conductance regulator (CFTR) gene mutations have been described to be associated with CF, but the functional significance of only a subset is known. Based on the Clinical and Functional Translation of CFTR repository, a little over 300 mutations have been characterized in detail.2 CF-causing mutations result in absent, not functioning, or abnormally functioning CFTR protein. Patients with CF carry pathogenic mutations in both copies of the CFTR gene. People with pathogenic mutations in only one copy of the CFTR gene do not manifest CF but are carriers of the disease. The most common pathogenic mutation is the F508del mutation. This mutation (a loss of phenylalanine at the 508th position) causes the protein to misfold and become marked for degradation. About 86% of all CF patients have at least one copy of the mutation; of these patients, approximately 41% are heterozygous and 46% are homozygous.3,4 Another common mutation is G551D, which is found in approximately 5% of CF patients.3 In patients with at least one copy of G551D some of the protein folds correctly, but when it reaches the apical membrane it does not open appropriately to let chloride ions flow normally. The following is an oft-used classification scheme for mutations that are known to cause CF. A classification system for the most common pathogenic mutations of the CFTR gene describes five classes: • Class I (transcription-stopping or "X-group") mutations result in no CFTR protein being produced. • Class II mutations ("folding mutations") result in protein formation (folding) and trafficking defects that hinder the transport of the CFTR to the apical membrane of cells. This group includes the most common CF-causing mutation, F508del. • Class III mutations ("gating mutations") result in a non-functioning CTFR protein on the apical membrane of cells. An example is the aforementioned G551D mutation, which is responsible for approximately 5% of CF cases. ©Institute for Clinical and Economic Review, 2018 Page 1 Final Evidence Report – Cystic Fibrosis Return to Table of Contents • Class IV and V mutations are associated with residual function (reduced functionality) of CFTR. CF is a progressive disease that affects many organ systems, though most of its morbidity and mortality are associated with its impact on the respiratory system. In epithelial cells, the CFTR gene is transcribed and translated to produce the CFTR protein, which is in turn, transported to the apical membrane, the part of the membrane that faces inwards towards the lumen of an organ. There it acts as a chloride ion gate and contributes to the regulation of salt transport in and out of the cell. Mutations to the CFTR gene can affect the amount of CFTR protein that is produced and transferred to the apical membrane or the CFTR protein's ability to regulate chloride and sodium ion flow.32 Failure to express normally-functioning CFTR protein in the apical (luminal) membrane of epithelial cells leads to thickened secretions in the lung, gastrointestinal tract, pancreas, and other organs. These thickened secretions are an integral part of the cascade that cause the primary manifestations of CF. In the lungs, the thickened secretions lead to decreased mucociliary clearance and chronic bronchial infection, which result in lung destruction over time. Daily aggressive pulmonary hygiene (i.e., nebulized medications and chest physiotherapy) are necessary to maintain health. Recurrent pulmonary exacerbations occur despite best care and require antibiotic treatment, increased pulmonary hygiene, and often hospitalization. Infections are associated with bacteria expected in bronchiectasis of other causes and tend to occur early in CF. The bronchi of many CF patients are eventually colonized with Pseudomonas aeruginosa, Burkholderia cepacia complex and other pathogens which are commonly resistant to most antibiotics. Chronic and repeated lung infections contribute to progressive damage in the airways, leading to bronchiectasis and ultimately to respiratory failure, which is responsible for the majority of CF-related deaths. CF affects all epithelia, and thus also affects other organ systems. Dysfunction in the epithelia of the intestine, pancreas, and liver can cause intestinal malabsorption, pancreatic insufficiency and CF-related diabetes, as well as biliary cirrhosis. Most men with CF are infertile because the vas deferens is not fully developed, but women with CF are subfertile, in part due to changes in cervical mucus, but are usually able to become pregnant and give birth. The disease and its management are therefore associated with multiple physical and psychosocial problems and economic insecurity, which can severely affect the quality of life of CF patients, their caretakers, and the rest of their families. Diagnosis All 50 US states and the District of Columbia now provide newborn screening for CF. Most states use some combination of blood testing for pancreatic injury and CFTR gene mutation analysis for screening. Patients who carry CF-causing mutations in each copy of the CFTR gene manifest CF. The diagnosis of CF is made by measuring the concentration of chloride ions in sweat following an ©Institute for Clinical and Economic Review, 2018 Page 2 Final Evidence Report – Cystic Fibrosis Return to Table of Contents established protocol. CF diagnosis is definitive in patients with sweat chloride concentrations above 60 mEq/L (as measured with established protocols in certified labs) and who have a clinical picture consistent with CF. Most CF patients have been diagnosed in childhood, although some patients with milder presentations have been diagnosed as adults. In the US in 2016, the median age at diagnosis for all patients was four months of age; 62% of new CF diagnoses were detected through newborn screening.1 Early diagnosis before symptom onset allows early treatment and, thus, is associated with better lung and nutritional outcomes later in life.33 Clinical Presentation While lung function is normal at birth, lung infections tend to occur early in life. Repeated and chronic infections can lead to bronchiectasis at a young age. Acute pulmonary infections requiring antibiotic treatment (pulmonary exacerbations) occur and can rapidly deteriorate pulmonary function. Pulmonary exacerbations are associated with increased lung damage, earlier mortality, higher healthcare costs, and lower quality of life.34,35 End-stage lung disease results in respiratory failure and death. CF patients with Class I, II, and III mutations tend to have somewhat lower lung function compared to those with Class IV and V mutations.1 The gastrointestinal (GI) system is also commonly affected in CF patients. Malabsorption of fat due to insufficient pancreatic enzymes, known as pancreatic insufficiency, affects an estimated 85% of CF patients and makes reaching a normal weight difficult for CF patients.36 Pancreatic damage that leads to an insufficiency of pancreatic enzymes often occurs within a few months after birth. 36 Similarly to lung function, pancreatic sufficiency and weight are influenced by genotype; F508del homozygous individuals are typically the most underweight, and F508del heterozygotes with G551D and R117H mutations showing slightly better nutrition.37 Over 80% of Cystic Fibrosis Foundation Patient Registry (CFFPR) patients are prescribed pancreatic enzyme replacement therapy (PERT) as part of their CF regimen to aid in fat metabolism and weight gain.1 Children born today show significant improvements in reaching and maintaining sufficient weight compared to CF patients born in 1987.1 As children mature into adulthood, clinical guidelines aim for adults 20 years and older to have a body mass index (BMI) at or above 22 for women and 23 for men.1 Lung function and weight are also closely related for CF patients, as shown in Figure 1.2. ©Institute for Clinical and Economic Review, 2018 Page 3 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Figure 1.2. FEV1 Percent Predicted Versus BMI Percentile for Children Six to 19 Years in 20161 Management The core treatment regimen for CF has historically aimed to control symptoms. It includes aggressive airway hygiene with chest physiotherapy, airway clearance devices, bronchodilators, inhaled and systemic antibiotics as needed or chronically, inhaled hypertonic saline, and aerosolized recombinant human DNase to reduce sputum thickness by breaking down free inflammatory cell DNA, as well as nutritional support through pancreatic enzyme replacement therapy, insulin, and diet. The treatment burden for CF patients is high, with patients reporting that they spend upwards of two hours a day completing treatment activities.5 Organ transplantation remains the last-line intervention for CF patients with end-stage disease. Advances in the early diagnosis and management of CF have led to longer survival than in earlier eras. In the 2016 annual report of the US Cystic Fibrosis Foundation Patient Registry, 53% of CF patients in the US were adults. The median predicted survival of CF patients born in 2016 is estimated to be 47.7 years.1 According to an NIH fact sheet “In 1962, the predicted median survival for CF patients was about 10 years, with few surviving into their teenage years.”38 Today, nearly 75% of those registered in the CFFPR over 18 years old were considered to have normal lung function or mild lung impairment; in 1987, this proportion was only about one-third.1 Likewise, lung function was severely impaired in about one-third of patients in 1987; today that number is 4%.1 While improvements in supportive care have improved the prognosis for CF patients, these treatments are directed only at symptom management. Recently introduced agents that modulate the pathophysiology of the disease, namely, Kalydeco®, Orkambi® and Symdeko™ represent a new class of treatments, and are the focus of this review. ©Institute for Clinical and Economic Review, 2018 Page 4 Final Evidence Report – Cystic Fibrosis Return to Table of Contents CFTR modulator drugs Modulator drugs increase CFTR-mediated ion transport. Two types of modulator drugs have been developed, with complementary modes of action. The effectiveness of modulators depends on the CF-causing mutation. For example, patients who are homozygous for class I mutations cannot respond to modulator-based treatments because there is no CFTR protein to be modulated. A full list of mutations for which each drug is approved is available in Appendix D. CFTR potentiators, such as ivacaftor (Kalydeco), increase the likelihood that the CFTR channel will transport ions through the cell membrane, i.e., they increase the channel’s “open probability”. Kalydeco has been approved for patients with various “gating” (e.g. G551D, a Class III mutation) and other mutations that result in residual CFTR protein function in the cell membrane (e.g., R117H). CFTR correctors, such as lumacaftor and tezacaftor, increase the amount of normal or mutated CFTR protein that gets transported to the apical (luminal) membrane, thereby increasing the amount of CFTR protein on the cell surface. Combinations of CFTR correctors and potentiators are considered in patients with “folding” (e.g., F508del, a Class II mutation) and/or residual function mutations. Orkambi (lumacaftor/ivacaftor) and Symdeko (tezacaftor/ivacaftor) are considered in patients homozygous for the F508del mutation. Symdeko is also considered in patients who are heterozygous for the F508del allele and carry a residual function mutation. For the purposes of this report we use trade names to facilitate ease of interpretation of the data, with the exception of unapproved doses of lumacaftor with ivacaftor. The use of these agents has generated great interest on the part of clinicians, patients, and their families. These drugs are the first of their kind to address the underlying genetic deficiencies leading to CF. Added to best supportive care, these drugs have been shown to improve respiratory function and weight, and they may slow the rate of decline of respiratory function over time. While generally safe, there may be some tolerability issues in some populations. Uncertainties around the use of modulators exist because most data are relatively short-term (or at best up to only about 3 years) and on surrogate endpoints, and evidence about longer-term benefit and increased survival does not yet exist. In addition, currently marketed CFTR modulators are very expensive, and alignment of their cost to patient benefit is not well understood, especially considering that these regimens will be incremental costs on top of current treatments comprising best supportive care. All stakeholders will therefore benefit from a comprehensive review of the clinical evidence and potential economic impact of adding CFTR modulator treatments to best supportive care. ©Institute for Clinical and Economic Review, 2018 Page 5 Final Evidence Report – Cystic Fibrosis Return to Table of Contents 1.2 Scope of the Assessment The scope for this assessment is described on the following pages using the PICOTS (Population, Intervention, Comparators, Outcomes, Timing, and Settings) framework. Evidence was collected from available randomized controlled trials and observational studies. Our evidence review included input from patients and patient advocacy organizations, data from regulatory documents, information submitted by manufacturers, and other grey literature when the evidence meets ICER standards (for more information, see https://icer- review.org/methodology/icers-methods/icer-value-assessment-framework/grey-literature-policy/). Analytic Framework The analytic framework for this assessment is depicted in Figure 1.1. Figure 1.1 Analytic Framework The diagram begins with the population of interest on the left. Actions, such as treatment, are depicted with solid arrows which link the population to outcomes. For example, a treatment may be associated with specific health outcomes. Outcomes are listed in the shaded boxes: those within the rounded boxes are intermediate outcomes (e.g., changes in lung clearance index), and those within the squared-off boxes are key measures of benefit (e.g., health-related quality of life). The key measures of benefit are linked to intermediate outcomes via a dashed line, as the relationship ©Institute for Clinical and Economic Review, 2018 Page 6 Final Evidence Report – Cystic Fibrosis Return to Table of Contents between these two types of outcomes may not always be validated. Curved arrows lead to the adverse events of treatment which are listed within the blue ellipse.39 Populations We reviewed evidence in three distinct populations: 1) The first population included individuals with CF and mutations consistent with the FDA- approved indications for Kalydeco. In this population, we reviewed evidence on Kalydeco. We included studies of individuals with mutations that have either gating or other (residual) functional implications (e.g., R117H). 2) The second population included individuals with CF who are homozygous for the F508del mutation. In this population we reviewed evidence on both Orkambi and Symdeko. 3) The third population included individuals with CF who are heterozygous for the F508del mutation and a residual function mutation that is potentially responsive to Symdeko. In this population we reviewed evidence on Symdeko and Kalydeco. Within these populations, subgroups of interest were defined according to presence of advanced nonreversible lung disease (e.g., patients who have predicted FEV1 below 40%, between 40-90%, or above 90%) and age (groups as defined in each study). Predicted FEV1 is a measure of lung function defined as the forced expiratory volume during the first second of expiration, adjusted for age, height, sex, and race.40,41 Other subgroups of interest were people with advanced non-pulmonary disease, such as recurrent pancreatitis, diabetes, liver transplantation, poor growth, and infertility. We included studies of individuals of any age, regardless of their past medical history, comorbidities, or the severity of their CF; however, we sought to exclude studies conducted in individuals after lung transplantation (for whom CFTR modulation therapy would not affect lung function). We imposed no other restrictions regarding population eligibility. Interventions and Comparators We examined the following comparisons in the following three appropriate populations: 1. For individuals who are candidates for Kalydeco, we compared adding Kalydeco to best supportive care versus best supportive care alone and placebo. 2. For individuals who are homozygous for the F508del mutation, we compared adding Orkambi or Symdeko to best supportive care versus best supportive care alone. We also compared Orkambi to Symdeko. 3. For individuals who are candidates for Symdeko because they carry one F508del mutation and residual function mutation that is potentially responsive to Symdeko, we compared adding Symdeko to best supportive care versus adding Kalydeco to best supportive care versus best supportive care alone. ©Institute for Clinical and Economic Review, 2018 Page 7 Final Evidence Report – Cystic Fibrosis Return to Table of Contents We excluded studies of lumacaftor and tezacaftor monotherapy, based on stakeholder feedback, neither is intended to be used as monotherapy. We excluded studies of Kalydeco, Orkambi, or Symdeko conducted in populations for whom the drugs are not approved or are not anticipating approval based on their genetic mutations. We also excluded studies of composite treatment strategies that, for example, start with Kalydeco and shift to a combination regimen after a period of time – if they were conducted in populations in which at least one of the regimens is not approved. Settings All settings were considered. Studies conducted in any country were considered. Outcomes Outcomes of interest included patient-centered outcomes, other clinical outcomes, important physiologic measurements, adverse events, and costs. Clinical outcomes pertain to measures of health status or events. Examples of clinical outcomes of interest include: • Mortality • Pulmonary exacerbations (acute and severe worsening of pulmonary symptoms) • Hospitalizations • Lung transplantation • Acute pancreatitis • Fertility Physiologic measurements are surrogate or intermediate measures for symptom severity, disease progression, or patient-centered outcomes. Examples of physiologic measurements of interest include: • FEV1 (predicted), including rate of FEV1 decline • Lung clearance index (LCI) • Weight, BMI, and growth (surrogate measures of nutrition status) • Fasting glucose and related measures of glucose control or diabetes Patient-centered outcomes include many outcomes that are also classified as clinical or cost outcomes listed separately below, but also include specific outcomes that directly relate to the lived experiences of patients and their families. Examples of patient-centered outcomes of interest include: ©Institute for Clinical and Economic Review, 2018 Page 8 Final Evidence Report – Cystic Fibrosis Return to Table of Contents • Disease-specific quality of life (specifically, as measured with the Cystic Fibrosis Questionnaire-Revised [CFQ-R] respiratory domain or other measures where available.42 • Mental health and affect, including depression, worry, and anxiety (as measured with validated instruments) • Functional status, including work, social/family, emotional, physical, etc. (as measured with validated instruments) • Time lost from school or work • Ability to participate in athletic activities and social functions • Financial insecurity • Caregiver burden Adverse events pertain to complications, harms, or other such events caused by or attributed to the intervention, not the disease process. Examples of adverse events of interest include: • Liver dysfunction • Upper respiratory infections • Gastrointestinal complaints (e.g., nausea, diarrhea, abdominal pain) • Headache • Rash • Chest discomfort • Dyspnea • Cataracts • Adverse events leading to treatment discontinuation Other outcomes were considered and reviewed depending on relevance to patients and availability of data. Evidence on drug-drug interactions from eligible studies was also included. We excluded measures of cellular (as opposed to organ) function and other blood, serum, or urine laboratory measures (other than glucose), such as sweat chloride, fecal elastase, sputum inflammatory measures, and nasal potential difference. While these outcomes may help to demonstrate whether the modulators address the basic defects in CF, they are not directly pertinent to clinical outcomes. We also excluded novel or “candidate” measures, such as metrics based on high resolution computerized tomography. Timing Randomized controlled and non-randomized comparative studies of all follow-up durations were eligible. Observational studies had to report outcomes at least one month following treatment. ©Institute for Clinical and Economic Review, 2018 Page 9 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Single-dose studies of any type were excluded. Our focus was on studies in which patients are prescribed a course of treatment. Potential Major Advance for a Serious Ultra-Rare Condition ICER is assessing CFTR modulator treatments under an adaptation of the ICER value framework focused on treatments for serious, ultra-rare conditions because we believe the assessment meets the following proposed criteria: • An eligible population for the treatment indication(s) included in the scope of the ICER review is estimated at fewer than approximately 10,000 individuals • There are no ongoing or planned clinical trials of the treatment for a patient population greater than approximately 10,000 individuals The US candidate population for treatment with modulators may be as small as 1,200 individuals (for Kalydeco) and is anticipated to involve 10,000 individuals or less in each genetically-specified population. 1.3 Definitions Disease and Pathophysiology Cystic Fibrosis (CF): We relied on each study’s definition of CF. However, the diagnostic criteria are standard. The diagnosis of CF is definitive in patients who have sweat chloride concentrations above 60 mEq/L (as measured with established protocols in certified labs) and who have a clinical picture consistent with CF. See Section 2, for a summary of current diagnosis guidelines. Heterozygous (for a genetic variation): The state of carrying the genetic variation only in one chromosome. Homozygous (for a genetic variation): The state of carrying the genetic variation in both chromosomes in a chromosome pair. Mutations: Heritable changes in the DNA, here, of the CFTR gene. More than 1,700 different CFTR mutations at different loci (places) of the CFTR gene have been identified, with varying effects on the quantity and function of the CFTR protein.3 A subset of these mutations are known to be pathogenic (see below). Pathogenic mutations: Mutations that substantially affect the quantity of functional CFTR protein on the cell membrane, causing CF. Based on the Clinical and Functional Translation of CFTR repository, a little over 300 mutations are known to cause CF.2 A patient manifests CF and its complications if they have pathogenic mutations in both copies of the CFTR gene. ©Institute for Clinical and Economic Review, 2018 Page 10 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Outcomes Absolute change: the numeric difference between the endpoint value (however defined) and the baseline (starting) value. Forced expiratory volume in one second (FEV1): the volume of air a person can exhale during a forced breath after a full inhalation, measured in the first second of the breath.43 FEV1 is reported in liters and measures the capacity of a person’s lungs. Lower FEV1 values indicate increasing lung impairment or damage. FEV1 is measured via spirometry. Percent predicted forced expiratory volume in one second (FEV1): measured FEV1 as a percentage of the predicted FEV1 value for a healthy individual of the same age, sex, and height.41 A clinically relevant change in absolute percent predicted FEV1 has been considered to be three to five points or greater.7 CF-related diabetes: We accepted each study’s definition of CF-related diabetes. While we may refer to CF-related diabetes as “diabetes” in this report, CF-related diabetes does not have the same pathophysiology as type I or II diabetes mellitus in people without CF. During a period of stable baseline health CF-related diabetes is diagnosed with standard diabetes criteria. However, modified criteria are used to diagnose CF-related diabetes during acute illness or continuous feedings.44 Cystic Fibrosis Questionnaire-Revised (CFQ-R): A validated survey which measures health-related quality of life (HRQOL) in CF patients.45 The CFQ-R measures quality of life and physical disease symptoms using the following scales: physical functioning, emotional functioning, social functioning, body image, eating problems, treatment burden, respiratory symptoms, and digestive symptoms, among other domains specific to older patients. Scores range from 0-100 with an increasing score indicating better quality of life. In general, a four-point change is considered clinically meaningful (the minimum clinically important difference, or MCID).21 This report primarily focuses on the CFQ-R respiratory domain score since it was reported in the pivotal trials of the CFTR modulators. Lung Clearance Index (LCI): A novel surrogate outcome that assesses the uneven distribution of lung ventilation, an indicator of obstructive lung disease. It represents the number of lung volume turnovers required for the lungs to clear a tracer gas to reach 2.5% of starting tracer gas concentration.46 Technical issues limit the feasibility of its use to adults and older children. Reductions from baseline indicate an improvement. Pulmonary exacerbations (PEx): New or change in antibiotic therapy (IV, inhaled, or oral) for any four or more of the signs/symptoms: change in sputum; new or increased hemoptysis; increased cough; increased dyspnea; malaise, fatigue, or lethargy; temperature above 38 degrees Celsius; anorexia or weight loss; sinus pain or tenderness; change in sinus discharge; change in physical ©Institute for Clinical and Economic Review, 2018 Page 11 Final Evidence Report – Cystic Fibrosis Return to Table of Contents examination of the chest; decrease in pulmonary function by 10%; and radiographic changes indicative of pulmonary infection).16 The CFTR modulators’ manufacturer informed us that the same definition was used in all clinical trials, but different sub-definitions were reported in studies (e.g., PEx requiring hospitalization or requiring antibiotics). Pulmonary abnormality or chest tightness: An adverse effect that has been associated with modulator therapy (primarily Orkambi) often leading to discontinuation. Weight for age z-score: A score that corresponds to the weight percentile of a child considering the distribution of weights of healthy children of the same age. For example, a weight for age z-score of -1.3 corresponds to the 10-th percentile of age specific weight values. An increase in the z-score from -1.3 to -1.2 corresponds to climbing from the 10th to the 12th weight percentile among children of the same age. An increase in the z-score from -0.3 to -0.2 would correspond to climbing 4 percentiles (from the 38th to the 42nd percentile). 1.4 Insights Gained from Discussions with Patients and Patient Groups We held semi-structured discussions via teleconference with parents of children with CF as well as with adult patients with CF, and identified cross-cutting themes, as described in further detail below. The first theme pertained to aspects of the CF experience that have a strong impact on quality of life from the patient’s and family’s perspective. First, daily care is demanding. Aggressive airway hygiene, a mainstay of standard CF management, is a time-consuming process. Additionally, patients routinely take many pills and inhalation treatments as part of standard care and are concerned by the prospect of even more interventions (e.g., more pills for the modulator treatments, or additional medications to manage emerging complications of CF, such as CF-related diabetes). The high daily demands of standard care take a toll on patients and caregivers. Second, CF patients often endure frequent and severe complications from their disease. Hospitalizations (e.g., secondary to pulmonary exacerbations), typically last for many days or weeks leading to substantial time lost from school, work, and leisure for both patients and caregivers. Hospitalizations and specialized care can be associated with additional logistical hindrances and expenses if it is necessary to travel to a facility with experience in CF management. Third, even minor complications of CF are pervasive and cannot be discounted in terms of reduced quality of life. For example, chronic sinusitis can be accompanied by the inability to smell or taste foods, which reduces appetite and contributes to malnutrition. All of the above can greatly limit the ability of CF patients to participate in the social, athletic, work, and other functions that their peers engage in. Another theme referred to the challenges of adhering to CF management. The daily management of CF is demanding, and a main goal of treatment is to delay the progression of the disease; ©Institute for Clinical and Economic Review, 2018 Page 12 Final Evidence Report – Cystic Fibrosis Return to Table of Contents skipping airway hygiene on a day both releases precious time for other activities and may not have an immediately perceptible negative impact on clinical function. Thus, children or young adults who move on to the next stage of their lives (e.g., leaving home to go to college) may be tempted to lapse in terms of adherence. A third theme was related to financial insecurity induced by the management of the disease. While all patients with whom we spoke have insurance coverage, their co-payments vary for CF-related treatment. Uncertainty about future insurance coverage of all treatments was also commonly raised. Additional expenses are associated with hospitalizations including travel, accommodation, arranging for care of other children, and other concerns. Further, parents with inflexible work schedules risk losing their jobs after exhausting their sick time. 1.5. Potential Cost-Saving Measures in Cystic Fibrosis As described in its Final Value Assessment Framework for 2017-2019, ICER will now include in its reports information on wasteful or lower-value services in the same clinical area that could be reduced or eliminated to create headroom in health care budgets for higher-value innovative services (for more information, see https://icer-review.org/final-vaf-2017-2019/). ICER encourages all stakeholders to suggest services (including treatments and mechanisms of care) currently used for people with CF that could be reduced, eliminated, or made more efficient. Some patients and caregivers we spoke with expressed concern about the very large cost associated with some CF treatments, including CFTR modulators, for what may be a modest gain in quality of life. In responses to the draft scoping document, stakeholders focused on potential ways in which CFTR modulators could offset costs by reducing pulmonary exacerbations and prolonging the decline in lung function leading to lung transplant. These potential changes in healthcare resources were captured in ICER’s economic models of the modulators themselves. We did not receive any suggestions on low-value services, but we heard from patient groups that randomized withdrawal studies are currently being planned to help inform possible changes to the current CF care regimen. ©Institute for Clinical and Economic Review, 2018 Page 13 Final Evidence Report – Cystic Fibrosis Return to Table of Contents 2. Summary of Coverage Policies and Clinical Guidelines 2.1 Coverage Policies To understand the insurance landscape for modulators treatments for cystic fibrosis, we reviewed publicly available 2017 coverage policies and formularies for Midwestern state Medicaid programs (Missouri), Centers for Medicare and Medicaid Services (CMS) policies, and major commercial plans in individual marketplaces across Missouri and other Midwestern states, including Anthem Blue Cross Blue Shield, Aetna, Blue Cross Blue Shield Kansas City, and Cigna Missouri. We surveyed each plan’s coverage policies for the three modulator treatments: Kalydeco, Orkambi, and Symdeko. No coverage policies were found for Symdeko as it was recently approved in February 2018. All the plans surveyed provided prior authorization criteria for the coverage of Orkambi or Kalydeco. Specifically, for Orkambi, all plans required a documented diagnosis of CF, as well as a CF mutation test documenting that the patient is homozygous for the F508del mutation.47-50 Plans varied on age requirements, some, like Cigna, allowing in patients six years or older, while other plans, like Anthem, required patients to be 12 years or older.49,50 For Kalydeco, all plans also required patients be over the age of two and have a definitive documented diagnosis of CF, as well as a CF mutation test documenting that the patient has one mutation that is responsive to Kalydeco based on its label (i.e. any of the following mutations: G551D, G1244E, G1349D, G178R, G551S, S1251N, S1255P, S549N, S549R, or R117H).47,50-52 Some plans also specifically call out that Kalydeco is not approved for any CF patients with a homozygous F508del mutation without the concurrent treatment with lumacaftor. We received anecdotal reports from patient advocacy groups and clinical experts that some patients have experienced difficulty accessing modulator treatments prescribed to them. Several of the examples of coverage denials appeared to be errors in the administration of the policy – for example, denial of coverage despite the patient having a covered mutation. One example of a coverage policy that went beyond current FDA labeling was Florida Medicaid’s prior authorization criteria for Orkambi, requiring that patients age 6-18 must have undergone a baseline ophthalmic examination to monitor for lens opacities and cataracts.53 ©Institute for Clinical and Economic Review, 2018 Page 14 Final Evidence Report – Cystic Fibrosis Return to Table of Contents 2.2 Clinical Guidelines There are a number of guidelines on the treatment and management of cystic fibrosis. These guidelines focus on different aspects of disease management, including diagnosis, care delivery, nutritional considerations, respiratory care guidelines, infection prevention, and management of other comorbid conditions like CF-related diabetes, liver disease and bone disease. Below, we have summarized guidelines from the Cystic Fibrosis Foundation, the National Institute for Health and Care Excellence, and the European Cystic Fibrosis Society. Cystic Fibrosis Foundation (CFF) Diagnosis54 The CFF guidelines recommend that diagnosis of CF begin with the clinical presentation of CF, followed by a sweat chloride test. Guidelines suggest that a sweat chloride test result greater than or equal to 60 mmol/L results in a CF diagnosis. A result less than or equal to 29 mmol/L suggests that CF is unlikely. For test results between 30 and 59 mmol/L, CFF recommends genetic testing to determine if any CFTR mutations are present. This is then followed by a clinical evaluation at a CFF- accredited care center for physiologic testing to make a more definitive diagnosis. Nutritional and GI Care Guidelines55 In the care and management of patients with CF, CFF recommends a focus on the patient’s nutritional status as a key component of clinical care for all patients, outlining guidelines for the caloric intake for patients, monitoring of growth and weight status of patients, and dosing of pancreatic enzyme replacement therapy (PERT). CFF recommends that for patients older than two years of age, energy intake should be 110-200% above those of healthy patients with similar age, sex, and size in order to see weight gain. It also recommends that the maintenance of normal weight, for both children and adults, was associate with better FEV1, as well as survival. CFF recommends that children and adolescents maintain a BMI at or above the 50th percentile in order to see benefit in FEV1 measurements. Finally, CFF recommends that PERT dosing should be 500- 2500 units lipase per kg body weight per meal in order to help bolster absorption of dietary fat and prevent macro- and micronutrient deficiencies. Respiratory Care Guidelines56 CFF has a series of guidelines relating to respiratory care for patients with CFF. These include chronic medications to maintain lung health, pulmonary exacerbations clinical care, CF airway clearance therapies, and pneumothorax and hemoptysis care guidelines. CFF lists a series of chronic medications that can be used in the management of respiratory care of CF patients. CFF recommends the use of some inhaled antibiotics, such as tobramycin and ©Institute for Clinical and Economic Review, 2018 Page 15 Final Evidence Report – Cystic Fibrosis Return to Table of Contents aztreonam, for all patients. It recommends mucolytics such as dornase alfa in patients at all stages of the disease, and hypertonic saline in all patients. CFF also suggests that anti-inflammatories, such as ibuprofen and azithromycin, may be beneficial for some patients. Finally, CFF recommends the use of Kalydeco in patients with at least one copy of the G551D mutation. CFF acknowledges that the guidelines were published prior to the label expansion for Kalydeco and the approval of Kalydeco and lumacaftor for patients with the homozygous F508del mutation. Pulmonary Exacerbations57 For the treatment of acute pulmonary exacerbations, which the guidelines describe as an increase in respiratory symptoms accompanied by an acute decrease in lung function, CFF lists a series of treatment recommendations, as well as a series of treatments it does not recommend. CFF recommends the continuation of chronic medications for maintenance of lung health during exacerbations. It recommends that airway clearance therapy techniques be increased during exacerbations. CFF recommends daily dosing of aminoglycosides rather than dosing three times a day during exacerbations. CFF states there is insufficient evidence to recommend the following treatments: delivery of IV antibiotics in a non-hospital setting, the continuation of inhaled antibiotics in patients being treated with the same antibiotics via IV, and the routine use of corticosteroids in the treatment of exacerbations, among others. Airway Clearance Therapy (ACT)58 CFF recommends the use of airway clearance for clearance of sputum, augmentation of cough, maintenance of lung function and improved quality of life in patients with CF. They do not recommend one form of ACT over another form, and rather suggest that each individual patient may have unique factors that would make one form of ACT more beneficial than another for that individual. CFF recommended aerobic exercise as well due to its overall health benefits. Infection Prevention and Control59 In order to better prevent the spread of infection in patients with CF, these guidelines recommend a series of precautions and policies, particularly for use in health care settings. These precautions include hand hygiene, contact precautions, mask use by CF patients, minimizing wait times in outpatient waiting rooms/common areas, and placement of patients with CF in single-patient rooms in inpatient settings. National Institute for Health and Care Excellence (NICE)60 Diagnosis NICE guidelines, which are written primarily for the United Kingdom, recommend diagnosis using a sweat test or a cystic fibrosis gene test in people with a series of qualifications, including family ©Institute for Clinical and Economic Review, 2018 Page 16 Final Evidence Report – Cystic Fibrosis Return to Table of Contents history, recurrent and chronic pulmonary disease, persistent chest X-ray changes among others. For individuals with a positive sweat test result, a clinical assessment that suggests CF, or a gene test that suggests one or more CF mutations, NICE recommends referral to specialist CF centers. Provision of Care to CF Patients: NICE outlines extensive guidelines around appropriate and comprehensive care to patients with CF and their families. NICE recommends the provision of adequate information and support to newly diagnosed individuals and their families, particularly information around local support and advocacy services, how to manage the risks of cross-infection, and transition to adult care. Care delivery itself should be provided by a multidisciplinary team made up of clinicians, dietitians, pharmacists, psychologists and physiotherapists, as well as social workers that are based at specialist cystic fibrosis centers. NICE recommends that these centers should plan patient care, minimizing the risk of cross-infection and maintain patient registries that track condition, treatments, and outcomes. Other recommendations include considering the use of telemedicine and home visits to minimize risk of infections. Annual and Routine Reviews NICE recommends that patients with CF undergo a comprehensive annual review that includes assessments of pulmonary function, nutritional and intestinal absorption, liver disease, CF-related diabetes, psychological status, and the patient’s exercise program. NICE states that these reviews should occur regularly for patients with CF and should occur more frequently in newly diagnosed or very young patients. Airway Clearance Techniques NICE recommends offering individualized airway clearance technique plans to patients based on their ability to clear mucus from their lungs, their (and their family or caregiver’s) preference, as well as any other factors that may impact adherence to the plan. NICE specifically recommends against offering high-frequency chest wall oscillation as a technique for patients with CF except in exceptional circumstances, as evidence does not demonstrate that it is a more effective technique than others. Mucoactive Agents NICE recommends the use of mucoactive agents for patients with CF with clinical evidence of lung disease. The first choice should be dornase alfa. If the patient does not respond, clinicians should consider the use of dornase alfa with hypertonic saline, or hypertonic saline alone. For those patients who cannot use dornase alfa, clinicians should consider mannitol dry powder for inhalation, particularly for children. NICE does not recommend Orkambi for the treatment of patients who are homozygous for the F508del mutation. ©Institute for Clinical and Economic Review, 2018 Page 17 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Infection and Nutrition NICE has extensive guidelines on the management of a series of bacterial infections through the use of oral, inhaled or intravenous antibiotics, depending on the strain. In addition, NICE outlines guidelines for the management of patient’s nutritional needs through caloric intake, nutritional needs and pancreatic enzyme replacement therapy, where appropriate. ©Institute for Clinical and Economic Review, 2018 Page 18 Final Evidence Report – Cystic Fibrosis Return to Table of Contents 3. Comparative Clinical Effectiveness 3.1 Overview To inform our review of the comparative clinical effectiveness of CFTR modulators in patients with cystic fibrosis, we extracted evidence from available clinical studies, whether in published or unpublished form (e.g., conference abstracts or presentations, FDA review documents). We focused on evidence of the efficacy, safety, and effectiveness of CFTR modulators in comparison with other CFTR modulators or placebo in our target population of individuals with cystic fibrosis of any age with a genetic mutation for which a CFTR modulator has been approved (see Appendix D). Our review focused on assessing the intermediate and long-term outcomes and harms assessed in available studies. We sought evidence on the following outcomes primarily: pulmonary exacerbation, percent predicted FEV1, weight/BMI, and quality of life measures. When reviewing clinical evidence in ultra-rare populations, ICER acknowledges the challenges of study design, recruitment, and availability of data on long-term outcomes. As such, when possible we aim to add to our findings specific context regarding areas of challenges in study design. 3.2 Methods Data Sources and Searches Procedures for the systematic literature review assessing the evidence on CFTR modulators followed established best research methods.61,62 We conducted the review in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines.63 The PRISMA guidelines include a list of 27 checklist items, which are described further in Appendix Table A1. We conducted the literature searches in PubMed and EMBASE. No limitations were placed on the searches regarding publication date, language, age, country, study design, or publication type (e.g., peer-reviewed or conference proceeding). All search strategies were generated utilizing the Population and Interventions described above. The search strategies included a combination of indexing terms (MeSH terms in MEDLINE, searched through PubMed, and EMTREE terms in EMBASE), as well as free-text terms, and are presented in Appendix Tables A2- A3. The date of the most recent search is December 19, 2017. To supplement the database searches, we performed a manual check of the reference lists of included trials and reviews and invited any interested stakeholder to share references germane to the scope of this project. Further details of the search algorithms, methods for study selection, quality assessment, and data extraction are available in Appendix Tables A2-3, Figure A2, and F1. ©Institute for Clinical and Economic Review, 2018 Page 19 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Study Selection We included all relevant randomized clinical trials and nonrandomized comparative studies of any size and duration. We also included single-arm (i.e., non-comparative) studies with at least 100 participants and at least one month of follow-up. We excluded studies evaluating Kalydeco and Orkambi combination therapy in populations outside their respective FDA-approved indications, as well as studies of composite treatment strategies that started with Kalydeco and later shifted to a combination regimen. In vitro and non-human studies were excluded, as were single-dose and pharmacokinetic studies. We excluded conference proceedings and abstracts reporting data also available in full-text peer-reviewed publications. We supplemented our review of published studies with data from known conference proceedings (within the last five years), regulatory documents, information submitted by manufacturers, ClinicalTrials.gov, and other grey literature when the evidence meets ICER standards and is not duplicative (for more information, see http://icer-review.org/methodology/icers-methods/icer- value-assessment-framework/grey-literature-policy/). Data Extraction and Quality Assessment Main trial data was extracted directly into SRDR™ (https://srdr.ahrq.gov). All eligible citations were extracted into Microsoft Word tables. Elements included a description of patient populations, sample size, duration of follow-up, funding source, study design features (e.g., open-label or cross- over periods), interventions (drug, dosage, frequency, schedules), outcome assessments (e.g., timing, definitions, and methods of assessment), results, and quality assessment for each study. Data were extracted from the full articles by a single reviewer and validated by a second reviewer. Assessment of Level of Certainty in Evidence We used the ICER Evidence Rating Matrix (see Figure 3.1) to evaluate the evidence for a variety of outcomes. The evidence rating reflects a joint judgment of two critical components: • The magnitude of the difference between a therapeutic agent and its comparator in “net health benefit” – the balance between clinical benefits and risks and/or adverse effects AND • The level of certainty in the best point estimate of net health benefit.64 ©Institute for Clinical and Economic Review, 2018 Page 20 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Figure 3.1. ICER Evidence Rating Matrix Assessment of Bias As part of our quality assessment, we evaluated the evidence base for the presence of potential publication bias. Given the emerging nature of the evidence base for newer treatments, we performed an assessment of publication bias for CFTR modulators using the clinicaltrials.gov database of trials. We scanned the site to identify studies completed more than two years ago that would have met our inclusion criteria and for which no findings have been published. Any such studies may indicate whether there is bias in the published literature. For this review, we did not find evidence of any study completed more than two years ago that that has not subsequently been published. We did learn of one study in patients with one copy of the F508del mutation and ©Institute for Clinical and Economic Review, 2018 Page 21 Final Evidence Report – Cystic Fibrosis Return to Table of Contents another mutation that results in no residual CFTR function, but this study was stopped early for futility.6 Quality of Individual Studies We rated all identified randomized control trials to be good quality using criteria from the US Preventive Services Task Force (USPSTF).65 See Appendix F Table for full trial ratings. Trials of good quality had study arms that were comparable at baseline, authors employed valid instruments to evaluate outcomes, and differential attrition was not observed. Fair-quality studies reported slight imbalances in baseline characteristics, showed some differences in follow-up between trial arms, and used less reliable measurement instruments to assess outcomes. We did not assign a quality rating to non-comparative studies or references that were obtained from grey literature sources (e.g., conference proceedings). Meta-Analysis We conducted meta-analysis for each outcome of interest, including harms, for which there were data from at least two studies that were sufficiently similar in population, intervention (e.g., dose), and other characteristics. From comparative studies, we meta-analyzed data on clinical, physiologic, and patient-centered outcomes. In part based on which outcomes had enough data to meta-analyze from sufficiently similar studies, we conducted meta-analyses of percent predicted FEV1, weight (in kg, BMI or as a BMI normalized to age and sex [z score]), CFQ-R respiratory domain, and pulmonary exacerbations. For harms outcomes, we combined data from single-arm studies and individual arms of comparative studies. We conducted meta-analyses of the proportion of participants receiving each drug (and placebo) who experienced severe adverse events (Grade 3 or 4) as well as drug discontinuation due to adverse events. Pulmonary abnormalities (chest tightness) were too infrequently reported to allow meaningful meta-analysis. Where data were reported for the same study participants at multiple time points (e.g., in both the RCT and the extension study), we included data from the longest duration of follow-up (i.e., the extension study) in the meta- analysis. When feasible, we also conducted meta-regression with study duration as a covariate; for these analyses we used all available data. All meta-analyses were conducted with random effects model restricted maximum likelihood analyses. Harms were analyzed as arcsine transformed data.66 Estimates of indirect comparisons were obtained as linear combinations of the direct estimates, following Bucher et al.67 ©Institute for Clinical and Economic Review, 2018 Page 22 Final Evidence Report – Cystic Fibrosis Return to Table of Contents 3.3 Results We evaluated treatment in three distinct populations: I. Kalydeco in gating and residual function mutation populations. This included individuals with G551D and non-G551D gating mutations and those with R117H residual function mutations. II. Orkambi and Symdeko in individuals homozygous for the F508del mutation. III. Symdeko and Kalydeco in individuals heterozygous for the F508del mutation with a second mutation amenable to Symdeko. Study Selection Our literature search yielded 1,897 potentially relevant references (Figure A1) of which 49 met eligibility criteria. The primary reasons for study exclusion included regimens for CFTR modulators outside the scope of the review (i.e. studies in other CF genetic populations or assessing other CF therapy regimens), non-clinical outcomes (e.g., in vitro studies), lack of outcomes of interest, and non-comparative study designs with either follow-up less than one month or study size less than 100 participants. Abstracts presented before 2012 were also excluded. Kalydeco: We included 35 articles on Kalydeco treatment in gating and residual function mutations; 19 articles were peer-reviewed publications and 16 were abstracts without associated peer- reviewed publications. Seven Phase III clinical trials were included, four of which were randomized clinical trials and three of which were single-arm studies; these were reported in ten included publications and seven conference abstracts. All randomized controlled trials were considered good quality. Seventeen references (10 publications, seven conference abstracts) reported randomized controlled trials data. An additional ten non-randomized controlled studies were reported in four publications and six conference abstracts, and four single-arm studies were reported by four publications and three abstracts. Three of the single-arm citations reported results from the GOAL study. One additional publication reporting on GOAL and a randomized control trial was included. Orkambi: We included ten articles on Orkambi treatment in individuals who are homozygous for the F508del mutation (seven peer-reviewed publications and two abstracts). Of the ten citations, four were randomized controlled trials and six were single-arm studies. All randomized controlled trials were considered good quality. Symdeko: We included three articles on Symdeko treatment, all of which were peer-reviewed randomized controlled trials (one Phase II, two Phase III). All randomized controlled trials were considered good quality, although parallel arm design is more impactful than short-term, crossover design. ©Institute for Clinical and Economic Review, 2018 Page 23 Final Evidence Report – Cystic Fibrosis Return to Table of Contents We report the results for the CFTR modulators by population of interest in the sections that follow, given the genetic specificity of the disease. We were unable to locate evidence in the following subgroups of interest: people with recurrent pancreatitis, diabetes, or liver transplantation. Some outcomes (e.g., pregnancy) were reported for CFTR modulators in general, without sufficient details to outline results by genetic subpopulation or drug regimen. Clinical Benefits Clinical Benefits of Kalydeco in Gating and Residual Function Mutation Populations Key Findings: Children, adolescents, and adults with G551D and non-G551D gating mutations experienced statistically significant and clinically meaningful gains in ppFEV 1 and reductions in the rate of pulmonary exacerbations with Kalydeco compared to placebo in 24-week studies. Longer- term follow-up suggests lung function improvements, including reduced rates of pulmonary exacerbations, are durable through three years. Limited evidence also suggests one-year reductions in rates of death, organ transplantation, and hospitalizations. Statistically significant gains in body weight and respiratory symptom-related quality of life with Kalydeco were reported for G551D and non-G551D gating mutation populations aged 12 and older compared to placebo. Statistically significant improvements in lung function or weight were not observed in adult patients with R117H residual function mutations. In a small sample of children aged 6 to 11 years with R117H residual function mutations, those on Kalydeco experienced statistically significant decreases in lung function and trended towards decreased respiratory symptom-related quality of life scores compared to placebo. Harms associated with Kalydeco are discussed separately, below. Four key randomized controlled trials – STRIVE, ENVISION, KONNECTION, and KONDUCT – evaluated the safety and efficacy of Kalydeco in individuals with at least one G551D, non-G551D gating, and R117H mutations (Table 3.1).8-11 All four studies required a baseline ppFEV1 ≥ 40%; upper limits were 90% for ages 12 and up and 105% for ages 6-11. All four trials randomized participants to receive either 150 mg of Kalydeco or placebo twice daily for 24 weeks. STRIVE, ENVISION, and KONDUCT were parallel group studies that assessed the mean absolute change from baseline in ppFEV1 through 24 weeks of treatment as the primary outcome, with additional data collection through 48 weeks in STRIVE and ENVISION. KONNECTION was a two-part, cross-over trial that randomly assigned participants to receive either Kalydeco twice daily for eight weeks followed by eight weeks of matched placebo or eight weeks of matched placebo followed by eight weeks of Kalydeco. The short-term duration of this study is an important limitation. Primary and secondary outcomes were the same as STRIVE, ENVISION, and KONDUCT except these were reported at eight weeks. KIWI, a Phase III single-arm study that included children aged 2-5 with a G551D gating mutation, assessed change from baseline in weight and BMI z-scores (difference in standardized deviations ©Institute for Clinical and Economic Review, 2018 Page 24 Final Evidence Report – Cystic Fibrosis Return to Table of Contents from normal population, for age and sex) as secondary efficacy endpoints (Table 3.1).13 Lung function measures were not included in this study because children under five years cannot perform spirometry reproducibly. Children were required to weigh at least 8 kg and to have at least one gating mutation at screening to qualify for enrollment. Long-term safety of Kalydeco was assessed in two open-label studies: PERSIST and GOAL. PERSIST followed eligible STRIVE and ENVISION participants for an additional 96 weeks, during which all participants received 150 mg of Kalydeco twice daily (Table 3.1).15 GOAL was a longitudinal cohort study of individuals aged six years and older with at least on G551D mutation and without prior history of Kalydeco use; participants received 150mg of Kalydeco twice daily.14 Key outcomes of GOAL included spirometry (ppFEV1), weight, CFQ-R scores, and hospitalizations. Additional details for the studies described above are summarized in Appendix F. ©Institute for Clinical and Economic Review, 2018 Page 25 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Table 3.1. Key Trials of Kalydeco Efficacy Conducted in G551D, non-G551D Gating Mutations, and R117H Residual Function Mutation Populations Study Quality and Study Design PERSIST15 KIWI13 KONNECTION*10 STRIVE8 ENVISION9 KONDUCT11 Single-arm, open-label Single-arm, open- RCT, Phase III cross-over RCT, Phase III RCT, Phase III RCT, Phase III extension label trial design Good Good Good Good Good Good Follow-up Duration 48 weeks 48 weeks 96 weeks 24 weeks 8 weeks 24 weeks Mutations G551D G551D G551D G551D non-G551D gating R117H Included Ages Included 12+ 6-11 6+ 2-5 6+ 6+ Treatment Groups Kalydeco Kalydeco Kalydeco Kalydeco Kalydeco Kalydeco Placebo Placebo Placebo Placebo No. of participants 161 52 144 34 39 69 % Female 52% 52% 53% 18% 44% 57% Age, mean (range) 25.5 (12-53) 8.9 (6-12) NR† NR (2-5) 22.8 (6-57) 31 (NR) ppFEV1, mean 63.6% 84.2% NR† N/A 78.4% 72.9% Weight, mean 61.5 kg 30.9 kg NR† NR NR NR† Weight z-score‡ NR NR NR† −0.2 0.084 NR† BMI z-score‡ NR 0.08 NR† NR 0.359 NR RCT: randomized controlled trial; BMI: body mass index; ppFEV1: percent predicted forced expiratory volume in one second *All participants received both Kalydeco and placebo; randomization determined one of two treatment orders: eight weeks of Kalydeco followed by eight weeks of placebo OR eight weeks of placebo followed by eight weeks of Kalydeco. A four- to eight-week washout period bridged the two treatment periods. †Data reported by treatment arm but not for overall trial population ‡Z-score = 0 indicates average weight for age and sex ©Institute for Clinical and Economic Review, 2018 Page 26 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Percent Predicted Forced Expiratory Volume (ppFEV1) Treatment differences (between-group differences comparing Kalydeco and placebo groups) in mean absolute and relative ppFEV1 changes are shown in Table 3.2. Table 3.2. Summary of Kalydeco Clinical Efficacy Outcomes for G551D-, non-G551D Gating Mutations, and R117H-CFTR Mutations Versus Placebo FEV1, Mean CFQ-R Respiratory Weight, Mean BMI, Mean Absolute Change Domain, Mean Absolute Change Absolute Change Population from Baseline, Absolute Change from Baseline, Kg from Baseline, Percentage Points from Baseline, (95% CI) Kg/m2 (95% CI) (95% CI) Points (95% CI) G551D †9 ‡ Ages 6-11 10.0 2.8§ 5.1 NR (n=52) (4.5 to 15.5) (1.3 to 4.2) (−1.6 to 11.8) 8.6 Ages 12+†8 10.5 2.8 NR (NR) (n=161) (8.5 to 12.5) (1.3 to 4.1) p<0.001 Non-G551D gating mutations Ages 6+#10 10.7 0.70 9.6 NR (n=39) (7.3 to 14.1) (0.34 to 0.99) (4.5 to 14.7) R117H ¤ 11 ‡ Ages 6+ 2.1 0.26‡ 8.4‡ NR (n=69) (−1.13 to 5.35) (−1.57 to 2.10) (2.17 to 14.6) Ages 6-11 −6.3 −0.18‡ −6.1‡ NR (n=17) (−12.0 to −0.7) (−2.38 to 2.01) (−15.7 to 3.4) Ages 18+ 5.0 0.31‡ 12.6‡ NR (n=50) (1.2 to 8.8) (−1.90 to 2.51) (5.0 to 20.3) N/A: not applicable for trial; NR: not reported *Ages 2-5 (KIWI), a single-arm study where all participants received Kalydeco (50 or 75 mg, based on weight) †Ages 6-11 (ENVISION) and ages 12+ (STRIVE) show treatment difference (Kalydeco vs. placebo) at 48 weeks ‡Adjusted, least squares mean and mixed-effects model for repeated measures § Adjusted, least squares mean and linear mixed model #Cross-over study design (8 weeks) followed by a 16-week open label extension (KONNECTION); treatment difference (Kalydeco vs. placebo) at 8 weeks ¤Ages 6+ (KONDUCT), treatment difference (Kalydeco vs. placebo) at 24 weeks. Treatment differences by age group shown in italics; age 12-17 subgroup (n=2) was too small for subgroup analysis All randomized controlled trials reported mean absolute change from baseline ppFEV1 (Table 3.2). Differences between Kalydeco and placebo groups’ mean absolute change from baseline after 48 weeks of treatment showed significant gains on Kalydeco in ppFEV1 for G551D individuals aged 6-11 (treatment difference: 10.0 percentage points; 95% CI 4.5 to 15.5; baseline ppFEV1 84%) 9 and 12 and older (treatment difference: 10.5 percentage points; 95% CI 8.5 to 12.5; baseline ppFEV1: ©Institute for Clinical and Economic Review, 2018 Page 27 Final Evidence Report – Cystic Fibrosis Return to Table of Contents 64%).8 Lung function outcomes at 24 and 48 weeks were comparable. Meta-analysis of the two RCTs comparing Kalydeco to placebo in patients with G551D mutations yielded a difference in ppFEV1 of 10.4 percentage points (95% CI 8.6 to 12.3), favoring Kalydeco (Appendix D, Figure D6).8,9 Results from the GOAL observational study show similar ppFEV1 gains for non-G551D gating mutations before and after Kalydeco treatment initiation (treatment difference: 10.7 percentage points; 95% CI 7.3 to 14.1).14 Lung function effects depended on age for R117H individuals in the KONDUCT study. Analysis of all participants showed a non-significant 2.1 percentage point difference (95% CI −1.13 to 5.35 percentage points) in ppFEV1 between Kalydeco and placebo groups.11 When stratified by age, however, children aged 6-11 on Kalydeco had significant declines in absolute ppFEV1 (difference: −6.3 percentage points, 95% CI −11.96 to −0.71 percentage points, p=0.03) compared to those on placebo, though the trial authors note the overall group’s lung function was stable except for one child who experienced a pulmonary exacerbation.11 In contrast, those aged 18 and older experienced significant gains in ppFEV1 (difference: 5.0%; 95% CI 1.15 to 8.78) compared to those on placebo. Only two participants in the study were aged 12-17, which precluded statistical analysis. Two publications explored long-term ppFEV1 outcomes: one Phase III single-arm open-label extension (PERSIST) and one non-randomized comparative study. PERSIST enrolled G551D individuals who completed STRIVE or ENVISION and assessed long-term safety and efficacy over an additional 96 weeks of Kalydeco use.15 Absolute change from baseline ppFEV1 was evaluated as a secondary outcome. Gains were similar for patients originally randomized to Kalydeco and placebo in both studies and averaged 9-10 percentage points over 96 weeks. This magnitude of effect is similar to what was observed in STRIVE over 24 weeks. Additional post-PERSIST analyses matched G551D individuals aged six and older who received Kalydeco during STRIVE, ENVISION, and/or PERSIST with up to five age-, sex-, weight-, and ppFEV1- comparable F508del homozygous individuals using the Cystic Fibrosis Foundation Patient Registry (CFFPR).68 Treatment differences showed G551D participants on Kalydeco during a Phase III trial gained a mean absolute 10.7 percentage points (p<0.001) compared to F508del receiving only standard care. The annualized rate of ppFEV1 decline showed those on Kalydeco experienced a modest but statistically significant difference in the rate of lung function decline (0.8 percentage points; 95% CI 0.06 to 1.55%) over three years compared to those receiving only standard care (Appendix F).68 ©Institute for Clinical and Economic Review, 2018 Page 28 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Weight and BMI Outcomes related to nutrition were reported using a variety of measures, ultimately limiting direct comparisons of nutritional outcomes (Table 3.2). STRIVE and ENVISION both reported mean absolute changes from baseline weight, while KONNECTION and KONDUCT reported mean absolute changes in BMI. ENVISION and KONNECTION also reported absolute changes in BMI-for-age z- scores. Overall, participants with G551D mutations in STRIVE and ENVISION receiving Kalydeco experienced a statistically significant mean 2.8 kg weight gain from baseline compared to those on placebo after 48 weeks (STRIVE 95% CI 1.3 to 4.1; ENVISION 95% CI 1.3 to 4.2).8,9 These effects represent about a 10% weight gain in children aged 6-11 years and about a 5% weight gain in adults. Meta-analysis of the two trials yielded the same estimate, with a tighter confidence interval: 2.8 kg (95% CI 1.8 to 3.8) (Appendix D, Figure D7). Age-stratified analysis (≤20 and >20 years old) showed a similar trend of weight gain for those on Kalydeco compared to placebo (Appendix F).69 Those under 20 years of age benefitted to a greater magnitude compared to those aged 20 and older (4.9 kg, 95% CI: NR vs. 2.9 kg, 95% CI 1.35 to 4.47 kg). Individual-level response analysis in this study suggested weight gain and increased lung function were not correlated, though both outcomes improved with Kalydeco treatment. The 34 children ages 2-5 years receiving Kalydeco in the single arm, open-label KIWI study showed a statistically significant mean increase in weight z-score (0.2, SD 0.3; p<0.0001) and BMI z-score (0.4, SD 0.4; p<0.0001). Non-G551D gating mutation individuals on Kalydeco experienced a statistically-significant 0.7 kg/m2 (95% CI 0.34 to 0.99 kg/m2) BMI increase after eight weeks of treatment compared to placebo.10 R117H individuals again had mixed results in subgroup analyses by age, and Kalydeco treatment effects were non-significant in all groups analyzed.11 Based on the data reported in the article, there was no statistically significant difference in weight change among younger and older participants, though most R117H participants (87%) were pancreatic sufficient and at a normal body mass at baseline. Quality of Life using Cystic Fibrosis Questionnaire– Revised (CFQ-R) All four randomized controlled trials collected CFQ-R respiratory domain scores, as shown in Table 3.2. Three of four trials reported significant, clinically meaningful increases from baseline CFQ-R respiratory domain scores for Kalydeco groups compared to placebo. Participants aged 12 and older reported significant improvements in quality of life regarding respiratory symptoms. STRIVE, KONNECTION, and the subset of KONDUCT participants who were aged 18 and older reported a mean absolute increase of 8.6 (95% CI NR, p<0.001), 9.6 (95% CI 4.5 to ©Institute for Clinical and Economic Review, 2018 Page 29 Final Evidence Report – Cystic Fibrosis Return to Table of Contents 14.7), and 12.6 (95% CI 5.02 to 20.25) points on the CFQ-R Respiratory domain compared to placebo, respectively.8,10,11 The KONNECTION study included children as young as six years, but the study average age was 22.8 years; therefore, we assume most participants were aged 12 or older. Meta-analysis of these three trials yielded a summary estimate of the difference between Kalydeco and placebo of 9.7 units (95% CI 6.5 to 13.0) (Appendix D, Figure D8). Participants aged 6-11 years (G551D and R177H), however, showed conflicting results in CFQ-R respiratory domain score improvement. G551D participants reported a non-significant 5.1 (95% CI −1.6 to 11.8) point improvement compared to placebo 9, while R117H participants reported a −6.1 (95% CI −15.68 to 3.41) point change 11; R117H findings may have been impacted by the small sample size, however (n=17). These studies were not meta-analyzed. One additional analysis of STRIVE CFQ-R outcomes reported scores for all domains included in the questionnaire (Appendix F).70 Treatment differences in health perceptions (7.6 points, p<0.001), physical functioning (4.4 points, p=0.006), respiratory symptoms (8.6 points, p<0.001), social functioning (4.3, p=0.003), vitality (5.5 points, p=0.002), and weight (5.3 points, p=0.053) domains exceeded the MCID threshold of four points. Treatment differences in the other domains also favored Kalydeco over placebo, though effects were not clinically meaningful. For the respiratory domain, 57% of those taking Kalydeco reported improvement in CFQ-R scores versus 25% on placebo (p<0.05). Likewise, 29% of Kalydeco recipients versus 54% of those on placebo reported a CFQ-R respiratory domain score decrease (p<0.05). Pulmonary Exacerbations Pulmonary exacerbations reported in randomized clinical trials are shown in Table 3.3. Pulmonary exacerbations were generally reported as either an outcome or adverse event, and in some cases as both, complicating in-depth understanding and analysis. Our meta-analysis and summary results for pulmonary exacerbations use the "outcome" data, not the adverse event data. ©Institute for Clinical and Economic Review, 2018 Page 30 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Table 3.3. Pulmonary Exacerbations in G551D Gating and R117H Residual Function Populations, by Reported Outcome Definition STRIVE8 KONDUCT11 Follow-up Duration 48 weeks 24 weeks Placebo Kalydeco Placebo Kalydeco (n=78) (n=83) (n=35) (n=34) Modified Fuch’s Criteria No. PEx 99 (1.3/subject) 47 (0.6/subject) 17 (0.5/subject) 13 (0.4/subject) No. Subjects with PEx 44 28 13 11 Hazard ratio (p value) 0.455 (0.001) 0.93 (NR) Required IV Antibiotics No. PEx (% of all PEx) 47 (47) 28 (60) 7 (41) 2 (15) No. Subjects with PEx NR NR 6 2 Required Hospitalization No. PEx (% of all PEx) 31 (31) 21 (45) 8 (47) 2 (15) No. Subjects with PEx NR NR 6 2 PEx: pulmonary exacerbations; NR: not reported In addition, pre-specified definitions of pulmonary exacerbation were not always available in published studies, appendices, or protocols. During conversations with the manufacturer, however, we heard all published clinical trials used the same protocol definition of a pulmonary exacerbation (modified Fuch’s criteria). We noted two discrepancies in pulmonary exacerbations reported as adverse events and outcomes. ENVISION reported four exacerbations in the Kalydeco group and three in the placebo group as outcomes; however, eight exacerbations are reported for each group when categorized as adverse events.9 Second, the KONDUCT study reported 13 and 11 exacerbations in the Kalydeco and placebo groups, respectively, and report three additional exacerbations (one in placebo, two in the Kalydeco group) as adverse events.11 STRIVE was the only randomized comparative study showing a treatment effect on the incidence of pulmonary exacerbations (Table 3.3). STRIVE participants receiving Kalydeco experienced approximately half as many pulmonary exacerbations compared to the placebo group over 48 weeks (55% risk reduction, p<0.001).8 ENVISION reported exacerbations among 4 of 26 (15%) Kalydeco and 3 of 26 (12%) placebo recipients over 48 weeks.71 The frequency of pulmonary exacerbations was similar (33-46%) during the additional 96 weeks of Kalydeco treatment during.15 Exacerbations during KONNECTION were reported by cross-over period with short-term intervention (8 weeks): 9 of 38 (24%) and 11 of 39 (28%) of participants experienced a pulmonary exacerbation during the eight-week Kalydeco and placebo periods, respectively.10 ©Institute for Clinical and Economic Review, 2018 Page 31 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Among the different definitions of pulmonary exacerbation explicitly or implicitly used by studies, we were most interested in pulmonary exacerbations requiring IV antibiotics and hospitalization because these are often associated with additional financial costs and reduced quality of life. STRIVE and KONDUCT were the only two studies to explicitly report these outcomes. The rate of exacerbations requiring IV antibiotics through 48 weeks was 0.71 for Kalydeco and 0.40 for placebo recipients. Thus, our calculations provide a rate ratio of 0.56 (NS).8 As shown in Table 3.3, there was no consistent trend in the Kalydeco and placebo groups in the rate of exacerbations requiring hospitalization or IV antibiotics. Meta-analysis of pulmonary exacerbations per modified Fuch’s criteria in STRIVE and KONDUCT yielded a summary odds ratio of 0.51 (95% CI 0.26 to 1.00) and a summary relative risk of 0.65 (95% CI 0.48 to 0.89) (Appendix D, Figures D9 and D10). KONDUCT did not report a p-value or confidence interval for the hazard ratio, implying statistical nonsignificance. However, assuming a nonsignificant p-value of either 0.10 or 0.50 yielded almost identical summary hazard ratios of about 0.67 (95% CI 0.33 to 1.35) (Appendix D, Figure D11). The two studies, though, had very different estimates of hazard ratios and the meta-analysis is statistically heterogeneous. A post hoc analysis of STRIVE participants assessed post-exacerbation lung function recovery.72 Lung function recovery, defined as returning to ≥100% of pre-exacerbation ppFEV1, was assessed two-to eight-weeks (“short-term recovery”) after antibiotic treatment for an exacerbation and again using the end-of-study ppFEV1 measurement (“long-term recovery”). Short-term (53.7% vs. 57.1%), and long-term recovery rates (46.6% vs. 47.7%) were similar for the placebo and Kalydeco groups. However, other related outcomes favored Kalydeco over placebo: 57% lower rate of pulmonary exacerbations (RR: 0.43; 95% CI 0.29 to 0.68); statistically significantly shorter pulmonary exacerbations (mean normalized days per patient: 13.5 [SD 27.3] vs. 36.7 [SD 49.5], respectively; p<0.001); fewer patients treated with IV antibiotics for an exacerbation (patients treated with IV antibiotics: 18.1% vs. 34.6%, respectively; p=0.02); and shorter antibiotic treatments (mean normalized days per patient of IV antibiotic therapy: 6.7 [SD 19.4] vs. 11.0 [SD 20.3], respectively; p=0.02) compared to placebo. A large, non-randomized, comparative, long-term study also reported significantly lower risks of pulmonary exacerbations associated with Kalydeco (N=1667 on Kalydeco).12,73 The study implicitly included all people with available data receiving Kalydeco, regardless of mutation. The annual risk of an exacerbation was assessed by matching individuals on Kalydeco to similar patients on best supportive care (US 6200, UK 2069).12 Over a one year period 6 to 12 year-old US children taking Kalydeco experienced a significantly lower annual risk of pulmonary exacerbation compared to those on best supportive care (RR: 0.34, 95% CI 0.22 to 0.52).73 Analysis of all ages (not reported in the abstract) showed those on Kalydeco also experienced a statistically significant decrease in the annual risk of pulmonary exacerbation (RR: 0.64, 95% CI 0.58 to 0.70) in the US cohort of 1256 participants on Kalydeco; similar results were seen in a UK cohort of 411 patients.12 The annual risk of other clinical outcomes in the US cohort were also lower for patients on Kalydeco compared to ©Institute for Clinical and Economic Review, 2018 Page 32 Final Evidence Report – Cystic Fibrosis Return to Table of Contents placebo, including death (RR: 0.41, 95% CI 0.20 to 0.84), organ transplant (RR: 0.15, 95% CI 0.04 to 0.59), and hospitalization (RR: 0.64, 95% CI 0.58 to 0.70), with similar but nonsignificant results for death and organ transplantation.12 Clinical Benefits of Orkambi and Symdeko in Individuals Homozygous for the F508del Mutation Key Findings: Orkambi and Symdeko both provided small but statistically significant improvements in absolute ppFEV1 compared to placebo after 24 weeks of treatment; however, the magnitude of effect varies by age, dose, and baseline lung function. In longer-term follow-up (96 weeks), those on Orkambi had slower decline in ppFEV1 than matched controls. Neither Orkambi nor Symdeko provided statistically significant short-term improvement in BMI or BMI- for-age z score compared with placebo. Both Orkambi and Symdeko provide improved respiratory-related quality of life compared with placebo. Orkambi and Symdeko reduced pulmonary exacerbation events over 24 weeks, including those requiring intravenous antibiotics and hospitalizations, compared with placebo. Indirect comparisons yielded no material differences between Orkambi and Symdeko in key clinical outcomes. Harms associated with Orkambi and Symdeko are discussed separately, below. Two treatment regimens were reviewed for individuals homozygous for the F508del mutation: Orkambi and Symdeko. Across these two treatments, we identified six key trials including four Phase III randomized controlled trials, one single arm trial and one long-term, open-label extension trial. Five of the six trials were of Orkambi. Two placebo-controlled, parallel-arm Phase III RCTs of Orkambi, TRAFFIC and TRANSPORT, enrolled patients ages 12 and older with two copies of the F508del mutation.16 Inclusion criteria included a screening FEV1 between 40-90% predicted and stable disease.16 Two doses of lumacaftor were tested against placebo (lumacaftor 600 mg daily or 400 mg twice a day, both with ivacaftor 250 mg twice a day).16 Study design was identical in both trials, so data were pooled by the author and are presented here. A subgroup analysis by baseline ppFEV1 is also reviewed in this section where data are available. A single placebo-controlled, parallel-arm Phase III randomized controlled trial evaluated 200 mg of lumacaftor twice daily in combination with 250 mg ivacaftor twice daily in children ages 6-11 years with two copies of the F508del mutation. Inclusion criteria specified a minimum weight of 15 kg, ppFEV1 > 70% and lung clearance index (LCI2.5) of 7.5 or more lung volume turnovers at screening.17 Exclusion criteria were similar to TRAFFIC/TRANSPORT. One randomized, placebo-controlled, parallel-arm trial of Symdeko, EVOLVE, enrolled 510 cystic fibrosis patients ages 12 and older who were homozygous for the F508del mutation for 24 weeks of follow-up.18 Inclusion and exclusion criteria were similar to TRAFFIC/TRANSPORT. ©Institute for Clinical and Economic Review, 2018 Page 33 Final Evidence Report – Cystic Fibrosis Return to Table of Contents The long-term safety of Orkambi was assessed in two open-label continuation studies. PROGRESS followed eligible TRAFFIC and TRANSPORT participants for an additional 96 weeks, during which all participants received either 600 mg of lumacaftor daily (combined with 250 mg of ivacaftor twice daily) or 400 mg of lumacaftor twice daily (combined with 250 mg of ivacaftor twice daily).19 Milla et al. reported on 58 children ages 6-11 years old receiving 200 mg of lumacaftor twice daily in combination with 250 mg ivacaftor twice daily during follow-up of 24 weeks.20 The primary endpoint of both open-label studies was based on treatment-emergent adverse events and other physiologic measures. Across all studies, outcomes of interest included ppFEV1 (as both absolute and relative changes), weight or BMI (or BMI Z score), CFQ-R respiratory domain, and number or rate of pulmonary exacerbations. See Table 3.4 for a comparison of baseline patient characteristics and outcome measures across key trials and Table 3.5 for a summary of results across trials. For simplicity, results present outcomes by the differing doses of lumacaftor only, as the dose of Kalydeco did not differ. Table 3.4. Included Trials in the Homozygous F508del Population TRAFFIC/TRAN PROGRESS19 SPORT*16 Ratjen et al.17 Milla et al.20 EVOLVE18 Study Design and Single-arm, RCT, Phase III, RCT, Phase III, Single-arm RCT, Phase III, Study Quality open-label Ages 12+ Good study Good extension Good Follow-up Duration 24 weeks 24 weeks 96 weeks 24 weeks 24 weeks Orkambi* Orkambi Symdeko Treatment Groups Orkambi* Orkambi Placebo Placebo Placebo No. of Participants 1108 204 1029 58 504 % Female 49% 59% 48% 53% 49% Age, mean (range) 25.1 (12-64) 8.8 (6-11) 25.0 (SD~10) 9.1 (6-11) 26.3 (SD~10) ppFEV1, mean 60.6% 89.8% 60.3% 91.4% 60.0% 2 2 2 BMI, mean 21.2 kg/m 16.4 kg/m 21.2 kg/m 16.89 kg/m2 21.04 kg/m2 *An additional arm, 600 mg daily lumacaftor with ivacaftor was studied; Pooled analysis ppFEV1: percent predicted forced expiratory volume in 1 second; BMI: body mass index ©Institute for Clinical and Economic Review, 2018 Page 34 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Table 3.5. Summary of Clinical Efficacy Outcomes from Randomized Controlled Trials for Patients Homozygous for F508del Orkambi Symdeko TRAFFIC and TRANSPORT*16 Ratjen et al.17 EVOLVE18 Lumacaftor Orkambi Orkambi Symdeko 600 mg qd (400 mg q Placebo (200 mg q Placebo (100 mg Placebo w/ivacaftor 12 hrs) 12) daily) FEV1, Absolute Change†, 1.1 -1.3 3.0 2.5 -0.32 3.4 -0.6 Percentage (-0.4 to (-2.8 to (p<0.001) (p<0.001) (p=0.40) (2.7 to 4.0) (-1.3 to 0.0) Points (p-value 2.6) 0.2) or 95% CI) FEV1, Relative Change†, % (p- 5.4 4.6 -0.17 6.3 -0.5 NR NR value or 95% (p<0.001) (p<0.001) (p<0.001) (5.1 to 7.4) (-1.7 to 0.6) CI) Lung Clearance Index (LCI), -1.0 0.1 Absolute NR NR NR (-1.3 to (-0.2 to NR NR Change (95% -0.8) 0.3) CI) BMI, Absolute 0.4 0.3 0.18 0.12 Change†, kg/m2 0.41 0.37 0.13 (0.3 to (0.1 to (0.08 to (0.03 to (P-Value or (p<0.001) (p<0.001) (p<0.007) 0.5) 0.4) 0.28) 0.22) 95% CI) BMI-For-Age Z Score, 0.1 0.1 -0.06 -0.02 Absolute NR NR NR (0.0 to (0.0 to (-0.14 to (-0.10 to Change, (95% 0.2) 0.1)† 0.02) 0.06) CI) CFQ-R, Respiratory Domain 5.5 3.0 4.9 4.1 1.9 5.0 -0.1 Absolute (3.4 to (1.0 to (p<0.001) (p<0.001) (p=0.02) (3.5 to 6.5) (-1.6 to 1.4) Change†, 7.6) 5.0) Points (P-Value Or 95% CI) Pulmonary 251± 173± (0.80 152± (0.70 78‡ (0.64 122‡ (0.99 Exacerbation, (1.14 per NR NR per 48 wk) per 48 wk) per yr) per yr) No. (Rates) 48 wk) All data change from baseline to follow-up; q=every, qd=daily CI: confidence interval *Pooled results †least-square means ‡Number of events (annualized estimated event rate) † Nonsignificant (P-value >0.05). ©Institute for Clinical and Economic Review, 2018 Page 35 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Percent Predicted Forced Expiratory Volume (ppFEV1) and Lung Clearance Index (LCI) Orkambi The key Orkambi randomized controlled trials reported absolute and relative changes in ppFEV1 between baseline and 24 weeks.16,17 For individuals ages 12 and older enrolled in TRAFFIC and TRANSPORT, least-squares mean absolute change in ppFEV1 was 3.0 percentage points 600 mg/day lumacaftor/ivacaftor arm, 2.5 percentage points in the Orkambi arm, and -0.32 percentage points in the placebo arm between baseline and 24 weeks (Table 3.5).16 The differences compared to placebo were 3.3 (95% CI, 2.3 to 4.3) percentage points for 600 mg daily lumacaftor/ivacaftor arm and 2.8 (95% CI, 1.8 to 3.8) percentage points for Orkambi.16 Konstan et al. performed a post hoc analysis by matching participants from TRAFFIC/TRANSPORT taking Orkambi with controls from the US CFFPR (homozygous F508del) to assess changes to the annual rate of ppFEV1 decline.74 Based on 455 patients taking Orkambi and 1,588 matched controls, the authors found Orkambi produced a 42% slower rate of decline in ppFEV1 (1.33 vs. 2.29 percentage points per year; p-value < 0.001).74 Although changes in ppFEV1 in the randomized trials were positive and significant, a post-approval study at a single hospital (n=116, mean age=24.7 years (range 12-59), 62% female, baseline ppFEV1=67.4) found no benefit of Orkambi after an average of four months use in a real-world cohort of children and adults (n=116; mean change in ppFEV1 0.11%; 95% CI, -39% to 20%).24 The ppFEV1 was reported as a secondary endpoint in the two trials in the 6-11 year old population, as lung function is often preserved in younger children.17 Milla et al. reported no statistically significant difference in absolute change in ppFEV1 from baseline to 24 weeks in an open-label Phase III trial.20 A randomized placebo-controlled trial of 206 children found participants taking 200 mg of lumacaftor twice a day in combination with 250 mg of ivacaftor twice a day (Orkambi) experienced a statistically significant absolute change in ppFEV1 of 2.4 percentage points (95% CI 0.4 to 4.4) compared with placebo; however, this was primarily driven by decreases in ppFEV1 in the placebo group between baseline and 24 weeks.17 The within-group change in the Orkambi arm did not show a statistically significant improvement.17 Relative changes in ppFEV1 were not reported in either trial. In an effort to capture the respiratory benefit of Orkambi, lung clearance index (LCI2.5) was used as the primary efficacy endpoint in the trial. LCI is a novel surrogate outcome that measures the number of lung volume turnovers required for the lungs to clear a tracer gas to reach 2.5% of starting tracer gas concentration.17 Reductions from baseline indicate an improvement. In both trials of Orkambi in the 6-11 year old population, Orkambi provided a statistically significant improvement from baseline with a change of -0.88 (95% CI, -1.40 to -0.37) and -1.0 (95% CI, -1.3 to - ©Institute for Clinical and Economic Review, 2018 Page 36 Final Evidence Report – Cystic Fibrosis Return to Table of Contents 0.8).17,20 In the RCT, the difference between Orkambi and placebo was also statistically significant (difference of -1.1, 95% CI -1.4 to -0.8).17 Subgroup analysis In TRAFFIC and TRANSPORT, similar changes in absolute ppFEV1 over 24 weeks compared with placebo were found for Orkambi (400 mg twice daily) for patients with baseline ppFEV1 < 40% (3.3%, 95% CI 0.2 to 6.4, n=29) and patients with baseline ppFEV1 ≥ 40% (2.8%, 95% CI 1.7 to 3.8, n=336), as well as for patients with baseline ppFEV1 < 70% (3.3%, 95% CI 2.1 to 4.4, n=245) and patients with baseline ppFEV1 ≥ 70% (1.9%, 95% CI −0.2 to 4.0, n=114).23 A 24-week, open-label Phase IIIb study of individuals with advanced lung disease (ppFEV1<40%) reported a statistically significant decline in ppFEV1 (-1.7%; 95% CI, -3.2 to -0.1) for the first 15 days followed by a return to baseline at week four, remaining stable until study completion.75 Symdeko In the homozygous population, one RCT (EVOLVE) reported absolute and relative changes in ppFEV1 for Symdeko.18 The primary efficacy endpoint, absolute change from baseline in percentage of predicted FEV1 through 24 weeks, showed a statistically significant improvement in absolute ppFEV1 of 3.4 percentage points (95% CI, 2.7 to 4.0).18 Compared with placebo, Symdeko provided 4.0 percentage point improvement (95% CI 3.1 to 4.8).18 Relative change from baseline in percentage of predicted FEV1 through week 24 showed a statistically significant improvement both within the active drug arm (6.3%, 95% CI, 5.1 to 7.4) and between Symdeko and placebo (6.8%, 95% CI, 5.3 to 8.3).18 Orkambi versus Symdeko No study has directly compared Orkambi and Symdeko. As shown in Table 3.6, the absolute change in ppFEV1 was significantly greater with both drugs than with placebo. By indirect comparison (network meta-analysis), the difference in absolute change in ppFEV1 between the two drugs is nonsignificant: 1.2 percentage points (95% CI -0.1 to 2.5, p=0.073). ©Institute for Clinical and Economic Review, 2018 Page 37 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Table 3.6. Absolute Change in ppFEV1 in Patients Homozygous for the F508del Mutation Orkambi vs. Placebo* Symdeko vs. Placebo† Symdeko vs. Orkambiⱡ FEV1, Absolute Change, 2.8 4.0 1.2 Percentage Points (95% (1.8 to 3.8) (3.1 to 4.8) (-0.1 to 2.5) CI) *Two studies included (TRAFFIC and TRANSPORT); data for lumacaftor 400 mg twice daily with ivacaftor 250 mg twice daily only †One study included (EVOLVE; n=504) ⱡ TRAFFIC, TRANSPORT and EVOLVE (n=1612); the comparison of Symdeko and Orkambi is an indirect comparison between the two placebo-controlled trials Weight and BMI Orkambi BMI was reported as absolute change from baseline in all Orkambi trials (Table 3.5). In trials with younger patients, BMI-for-age z-score was also reported. Results in BMI varied across trials. In the TRAFFIC trial (n=549), neither active treatment dose arm showed a difference in BMI compared to placebo.16 However, in TRANSPORT, an identically designed trial of 559 participants, least-squares mean absolute change in BMI was significantly higher in the two active comparator arms compared to placebo.16 It is not clear why the effect of Orkambi on weight differed in the two trials except to note that the increases in BMI were only about 1-2% from participants’ baseline BMIs. In a pooled analysis, lumacaftor 600 mg daily with ivacaftor showed a statistically significant increase of 0.28 kg/m2 (95% CI, 0.15 to 0.41 kg/m2) compared to placebo and Orkambi showed a statistically significant increase of 0.24 kg/m2 (95% CI, 0.11 to 0.37) versus placebo.16 After 96-weeks on Orkambi, individuals in PROGRESS (open-label extension of TRAFFIC and TRANSPORT) had an absolute change in BMI of 0.76 to 0.96 kg/m2 (95% CI, 0.56 to 0.97 kg/m2 and 95% CI, 0.81 to 1.11 kg/m2 depending on original assignment arm).19 Both BMI-for-age z-score and weight-for-age z- score in participants under the age of 20 in TRAFFIC/TRANSPORT showed improvement with Orkambi versus matched controls (see Appendix Figure D1).16 Results of absolute change in BMI in children 6-11 years old also varied between studies. In the open-label, single-arm, Phase III study, children saw an absolute change in BMI of 0.64 kg/m2 (95% CI, 0.46 to 0.83 kg/m2) at 24 weeks (a 3.8% increase from baseline).20 However, in the randomized controlled trial, there was no difference in absolute BMI between Orkambi and placebo.17 BMI-for- age z-scores also showed a significant increase from baseline to 24-weeks in the single-arm study (0.15 kg/m2; 95% CI, 0.08 to 0.22 kg/m2) yet showed no difference compared to placebo in the RCT.17,20 Weight-for-age z-scores changed from a baseline mean of -0.03 (1.03) to 0.13 (95% CI, 0.07 to 0.19) at 24 weeks (least-squares mean using mixed-effects model for repeated measures).20 ©Institute for Clinical and Economic Review, 2018 Page 38 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Subgroup analysis In TRAFFIC and TRANSPORT, similar changes in BMI over 24 weeks compared with placebo were found for Orkambi (400 mg twice daily) for patients with baseline ppFEV1 < 40% (0.3, 95% CI −0.2 to 0.8, n=29) and patients with baseline ppFEV1 ≥ 40% (0.2, 95% CI 0.1 to 0.4, n=336), as well as for patients with baseline ppFEV1 < 70% (0.2, 95% CI 0.0 to 0.3, n=245) and patients with baseline ppFEV1 ≥ 70% (0.3, 95% CI 0.1 to 0.6, n=114).23 Symdeko Absolute change in BMI from baseline to 24 weeks in the EVOLVE trial showed within-person improvement of 0.18 kg/m2 (95% CI, 0.08 to 0.28) in the Symdeko arm and 0.12 kg/m2 (95% CI, 0.03 to 0.22) in the placebo arm (<1% increases from baseline).18 The difference in absolute change in BMI between treatment and placebo was non-significant.18 BMI-for-age z-score change from baseline to 24 weeks was non-significant for both arms (see Table 3.5).18 Long-term data on the effect of Symdeko on BMI or BMI-for-age z-score is not available yet. Orkambi versus Symdeko No study has directly compared Orkambi and Symdeko. As shown in Table 3.7, the absolute change in BMI Z score was similar for both drugs versus placebo; thus, by indirect comparison (network meta-analysis), the difference in Z score between the two drugs is nonsignificant: -0.04 z score units (95% CI -0.29 to 0.07) Table 3.7. Meta-analysis of Change in BMI-for-age Z score in Patients Homozygous for the F508del Mutation Orkambi vs. Placebo* Symdeko vs. Placebo† Symdeko vs. Orkambiⱡ BMI-for-age Z 0.0 -0.04 -0.04 score, (95% CI) (-0.2 to 0.2) (-0.15 to 0.07) (-0.29 to 0.21) *One study included (Ratjen et al.; n=204). †One study included (EVOLVE; n=504) ⱡRatjen et al. and EVOLVE (n=708) Quality of Life using Cystic Fibrosis Questionnaire– Revised (CFQ-R) Orkambi Adolescents and adults receiving Orkambi in TRAFFIC and TRANSPORT reported improved respiratory symptoms on the CFQ-R after 24 weeks as compared to individuals randomized to placebo (2.2 points; 95% CI 0.0 to 4.5, see Table 3.8; individual arm results in Table 3.5).16 While statistically significant, this value did not meet the MCID of 4.21 These benefits lasted through 72 ©Institute for Clinical and Economic Review, 2018 Page 39 Final Evidence Report – Cystic Fibrosis Return to Table of Contents weeks for all participants who enrolled in the open-label extension study, PROGRESS.19 At 96 weeks, patients continued to report improved symptoms, however, the benefits did not statistically differ from baseline in most patients.19 Respiratory symptom quality of life was mixed in children ages 6-11 years. Milla et al. reported a statistically and clinically significant improvement in CFQ-R between baseline and 24 weeks in an open-label trial (5.4 points; 95% CI, 1.4 to 9.4).20 These findings were similar in the randomized controlled trial where children randomized to Orkambi reported an absolute change from baseline to 24 weeks of 5.5 points (95% CI, 3.4 to 7.6), however, children randomized to placebo also reported fewer respiratory symptoms (3.0 points; 95% CI, 1.0 to 5.0).17 Orkambi was not found to confer a statistically significant benefit when compared to placebo.17 Other domains of the CFQ-R were not reported in the key studies. Subgroup analysis In TRAFFIC and TRANSPORT, estimates of relative effects of Orkambi compared with placebo on CFQ-R over 24 weeks varied based on baseline ppFEV1 category, but because of high variability in the score across the study, differences across subgroups were not statistically significant.23 Symdeko Individuals enrolled in the Symdeko arm of the EVOLVE study showed a clinically and statistically significant improvement in respiratory symptoms from baseline to 24 weeks (5.0 points; 95% CI, 3.5 to 6.5) while individuals randomized to placebo showed a slight but nonsignificant decline.18 Compared with placebo, Symdeko improved respiratory domain quality of life (difference of 5.1 points; 95% CI, 3.2 to 7.0).18 Other domains of the CFQ-R were not reported in the key studies. Orkambi versus Symdeko No study has directly compared Orkambi and Symdeko. As shown in Table 3.8, both drugs resulted in statistically significant improvements in respiratory symptom-related quality of life, but the effect was larger with Symdeko. By indirect comparison (network meta-analysis), Symdeko was just nonsignificantly more effective to improv CFQ-R respiratory domain score than Orkambi: difference 2.9 units (95% CI -0.0 to 5.8, p=0.054). ©Institute for Clinical and Economic Review, 2018 Page 40 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Table 3.8. Meta-analysis of Quality of Life in Patients Homozygous for the F508del Mutation (CFQ- R) Respiratory Domain Score Orkambi vs. Placebo* Symdeko vs. Placebo† Symdeko vs. Orkambiⱡ CFQ-R, absolute change, 2.2 5.1 2.9 score (95% CI) (0.0 to 4.5) (3.2 to 7.0) (-0.0 to 5.8) *Two studies included (TRAFFIC and TRANSPORT); lumacaftor 400 mg twice daily with ivacaftor 250 mg twice daily only †One study included (EVOLVE; n=504) ⱡ TRAFFIC, TRANSPORT and EVOLVE (n=1612) Pulmonary Exacerbations Table 3.9. Reported Annualized Pulmonary Exacerbation Rates Per Patient Year in Patients Homozygous for the F508del Mutation TRAFFIC/TRANSPORT PROGRESS EVOLVE Follow-up 24 weeks 96 weeks 24 weeks Duration Placebo Orkambi Orkambi* Placebo Symdeko No. Subjects 371 369 369 256 248 Modified Fuch’s Criteria No. PEx 251 152 NR NR NR No. PEx per Pt 1.14 0.70 0.65 0.99 0.64 Yr (95% CI) (0.97 to 1.34) (0.57 to 0.84) (0.56 to 0.75) NR NR Required IV Antibiotics No. PEx per Pt 0.58 0.25 0.32 Either IV Either IV Yr (0.47 to 0.72) (0.19 to 0.33) (0.26 to 0.38) antibiotics or antibiotics or hospitalizations hospitalizations (or both) (or both) 0.54 events/yr 0.29 events/yr Required Hospitalization No. PEx per Pt 0.45 0.17 0.24 Either IV Either IV Yr (0.36 to 0.57) (0.12 to 0.25) (0.19 to 0.29) antibiotics or antibiotics or hospitalizations hospitalizations (or both) (or both) 0.54 events/yr 0.29 events/yr PEx: pulmonary exacerbations *Lumacaftor 400 mg twice daily with ivacaftor 250 mg twice daily, ±total 120 weeks data (96 weeks after 24 in TRAFFIC/TRANSPORT) ©Institute for Clinical and Economic Review, 2018 Page 41 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Orkambi Patients receiving Orkambi in TRAFFIC and TRANSPORT reported fewer pulmonary exacerbation events (modified Fuch’s criteria) from baseline to 24 weeks than patients randomized to placebo (Table 3.9).16 The rate ratio between active drug and placebo was 0.65 (95% CI, 0.55 to 0.77) with the greatest reduction in the Orkambi arm (0.61, 95% CI 0.49 to 0.76).16 Orkambi provided statistically significant reductions in pulmonary exacerbations requiring antibiotics (56% fewer than placebo) and hospitalizations (61% fewer than placebo).16 Pulmonary exacerbations reported during TRAFFIC/TRANSPORT are also shown in Table 3.9. After 96 weeks, those who continued on Orkambi maintained a stable reduction (Table 3.9).19 The number of events requiring hospitalization per patient-year increased slightly after an additional 96 weeks. Similarly, the number of events requiring intravenous antibiotics per patient-year also increased slightly from 0.25 (95% CI, 0.19 to 0.33) at the end of the randomized clinical trial to 0.32 (95% CI, 0.26 to 0.38) at the end of the open-label extension study (Table 3.9). Pulmonary exacerbation events were not reported as an outcome in studies of children 6-11 years old. Symdeko Pulmonary exacerbations reported during EVOLVE are shown in Table 3.9. Patients in the EVOLVE trial randomized to Symdeko showed a statistically significantly lower rate of pulmonary exacerbation compared to those randomized to placebo (RR 0.65; 95% CI 0.48 to 0.88).18 The rate of pulmonary exacerbations requiring antibiotics or hospitalization was also significantly lower in the Symdeko arm compared to the placebo arm (RR 0.53; 95% CI, 0.34 to 0.82).18 Orkambi versus Symdeko As shown in Table 3.10, both drugs significantly reduce the rate of pulmonary exacerbations to a similar extent. Indirect comparison (network meta-analysis) between Symdeko and Orkambi found no statistically significant difference in pulmonary exacerbations between the two drugs, with an estimated rate ratio of 0.87 (95% CI 0.53 to 1.42). Table 3.10. Meta-analysis of Pulmonary Exacerbations in Patients Homozygous for the F508del Mutation Symdeko vs. Orkambi vs. Placebo* Symdeko vs. Placebo† Orkambiⱡ Pulmonary Exacerbations, Rate 0.61 0.53 0.87 Ratio, Score (95% CI)Ⱡ (0.49 to 0.76) (0.34 to 0.82) (0.53 to 1.42) *Two studies included (TRAFFIC and TRANSPORT), 400 mg dose only. †One study included (EVOLVE; n=504) ⱡPulmonary exacerbations defined as infective or requiring intravenous antibiotics or hospitalization ©Institute for Clinical and Economic Review, 2018 Page 42 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Clinical Benefits of Symdeko and Kalydeco in Individuals Heterozygous for the F508del Mutation Key Findings: Based on a single short-term (8 week) cross-over trials, Symdeko and Kalydeco both improve absolute and relative ppFEV1 compared with placebo. Symdeko provides a statistically significant benefit over Kalydeco. Respiratory symptom-related quality of life was statistically significantly improved by both Symdeko and Kalydeco compared with placebo. At 8 weeks, BMI and pulmonary exacerbations are not significantly different between the two drugs and compared with placebo, however, the follow-up duration was likely too short to adequately evaluate these outcomes. Harms associated with Symdeko and Kalydeco are discussed separately, below. There is one key trial of Symdeko and Kalydeco in patients heterozygous for the F508del mutation with a second mutation that is responsive to Kalydeco (see Appendix D for list of secondary genes and gene specific efficacy outcomes). The EXPAND trial is a Phase III, randomized, double-blind, placebo-controlled, three intervention crossover trial in which each patient received two of the three interventions for eight-week periods separated by an eight-week washout period.22 The three interventions included Symdeko (tezacaftor 100 mg daily with ivacaftor 150 mg twice daily), Kalydeco (ivacaftor 150 mg twice daily) or placebo. Individuals were included if they were aged 12 or older, had a percentage of predicted FEV1 at screening between 40-90%, a diagnosis of cystic fibrosis and stable lung disease. Exclusion criteria included laboratory values in the abnormal range, acute respiratory infections or changes in pulmonary disease 28 days prior to first drug, had a history of transplant or recently used other CFTR modulators. Individuals were randomized to one of six intervention sequences.22 The quality of the study was good, although it provided short-term (eight week) data relative to the parallel-arm RCTs in patients homozygous for the F508del mutation (i.e., 24 weeks in EVOLVE and TRAFFIC/TRANSPORT). The primary efficacy endpoint was absolute change in ppFEV1 from baseline to an average of the four-week and eight-week measurements in the first intervention and was compared to the same timepoints in the second assigned intervention. Key secondary endpoints included CFQ-R respiratory domain score and relative change in ppFEV1. Exploratory endpoints included the rate of pulmonary exacerbations and BMI.22 Of the 246 patients who received treatment, 95% (n=234) completed both intervention periods and provided efficacy data. The average age at screening across all subjects was 34.8 (SD 14.2) years, 55% of subjects were female, average ppFEV1 was 62.3% (SD 14.5), average BMI was 24.2 (SD 5.1) kg/m2, and average baseline CFQ-R score was 68.1 (SD 17.7).22 While all patients had one F508del mutation, the second mutation varied. Table 1 of the EXPAND manuscript describes the cohort as being 60% class V noncanonical splice and 40% class II to IV residual function mutations in the second allele at baseline.22 ©Institute for Clinical and Economic Review, 2018 Page 43 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Table 3.11. Summary of Results in the EXPAND Trial in Patients Heterozygous for a F508del Mutation22 Symdeko (N=161) vs. Kalydeco (N=156) Symdeko (N=161) vs. Placebo (N=161) vs. Placebo (N=161) Kalydeco (N=156) ppFEV1, Absolute Changeⱡ, Percentage 6.8 (5.7 to 7.8) 4.7 (3.7 to 5.8) 2.1 (1.2 to 2.9) Points (95% CI) FEV1, Relative Change, % 11.4 (9.6 to 13.2) 8.1 (6.3 to 9.9) 3.3 (1.8 to 4.8) (95% CI) BMI, Absolute Changeⱡ, 0.34 Symdeko 0.47 Kalydeco 0.34 Symdeko kg/m2 (Variance Data 0.18 placebo 0.18 placebo 0.47 Kalydeco Not Reported) CFQ-R, Absolute 11.1 (8.7 to 13.6) 9.7 (7.2 to 12.2) 1.4 (-1.0 to 3.9) Changeⱡ, Points (95% CI) Pulmonary Exacerbation, Rate Ratio vs. Placebo 0.54 (0.26 to 1.13) 0.46 (0.21 to 1.01) 1.18 (0.49 to 2.87) (95% CI) CI: confidence interval Percent Predicted Forced Expiratory Volume (ppFEV1) Change in ppFEV1 was measured as an average of the results at four weeks and eight weeks compared to baseline.22 Compared to placebo, both interventions provided statistically significant improvement in absolute ppFEV1: 6.8 percentage points for Symdeko (95% CI 5.7 to 7.8) and 4.7 percentage points for Kalydeco (95% CI 3.7 to 5.8)(Table 3.11).22 The difference between Symdeko and Kalydeco was also statistically significant but clinically modest, favoring Symdeko (2.1 percentage points; 95% CI 1.2 to 2.9).22 These changes compared to baseline ppFEV1 of 62%. Subgroup Analysis The EXPAND trial analyzed the difference in absolute change in ppFEV1 by age, baseline ppFEV1, class of residual function mutation, sex, use of concomitant medications and colonization of pseudomonas aeruginosa. Most of the subgroups showed similar relatively consistent treatment effects for Symdeko versus placebo; however, age < 18 vs. ≥ 18 years seemed to modify the effect. Those less than 18 years old showed a 12.0 percentage point improvement in absolute ppFEV1 (95% CI, 9.3 to 14.8) whereas those 18 years and older saw a 6.0 percentage point increase (4.9 to 7.0).22 The confidence intervals were wider in the under 18 subgroup due to small numbers (< 15% of each arm).22 Similar results were seen in the same subgroups with Kalydeco compared with placebo. ©Institute for Clinical and Economic Review, 2018 Page 44 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Body Mass Index BMI was a non-powered exploratory endpoint in the EXPAND trial given the short time frame on each intervention sequence. BMI increased 0.34 kg/m2 for Symdeko (1.4% increase from baseline), 0.47 kg/m2 for ivacaftor (1.9%), and 0.18 kg/m2 for placebo (0.7%) (Table 3.11).22 No data were reported to allow an estimate of statistical significance. Quality of Life using Cystic Fibrosis Questionnaire– Revised (CFQ-R) Symdeko provided significantly better quality of life using the CFQ-R respiratory domain score compared to placebo (11.1 points; 95% CI 8.7 to 13.6) (Table 3.11).22 Kalydeco also provided significantly better respiratory symptom-related quality of life compared to placebo (9.7 points; 95% CI, 7.2 to 12.2).22 No significant benefit was found between Symdeko and Kalydeco on CFQ- R.22 The proportion of patients that received a clinically significant improvement in CFQ-R was 65% in the Symdeko group, 58% in the Kalydeco group and 33% in the placebo group.22 Pulmonary Exacerbations The placebo group in the EXPAND trial reported the greatest number of pulmonary exacerbations overall (n=20 events; estimated event rate per year of 0.63) (Table 3.12). The Symdeko group reported 11 events (0.34 estimated event rate per year) and the Kalydeco group reported 9 events (0.29 estimated event rate per year) (Table 3.12). The rate ratio versus placebo was not statistically significant for either drug. Estimated indirect analysis of Symdeko compared to Kalydeco showed no significant differences between the drugs; however, this is not unexpected since pulmonary exacerbation was an exploratory endpoint and the study was of a limited duration (8 weeks). Data on the number of events or event rates of pulmonary exacerbations requiring IV antibiotics or hospitalization were not reported. ©Institute for Clinical and Economic Review, 2018 Page 45 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Table 3.12. Reported Annualized Pulmonary Exacerbation Rates in Patients Heterozygous for the F508del Mutation EXPAND*22 Follow-Up Duration 8 weeks Placebo Kalydeco Symdeko Modified Fuch’s Criteria No. Subjects 161 156 161 No. PEx’s 20 9 11 Estimated Event Rate per 0.63 0.29 0.34 Year Rate Ratio vs. Placebo - 0.46 0.54 95% CI - (0.21 to 1.01) (0.26 to 1.13) PExs: pulmonary exacerbation; CI: confidence interval *Pulmonary exacerbations requiring IV or hospitalization not reported Harms Frequencies of adverse events for all three CFTR modulators are reported in Table 3.13. Serious adverse events occurred less frequently in all modulators compared to placebo. Reasons for CFTR modulator discontinuation included elevated liver enzymes, creatinine kinase levels,76 hemoptysis, bronchospasm, dyspnea, pulmonary exacerbation and rash.16 No deaths during CFTR modulator trials were related to the drugs. ©Institute for Clinical and Economic Review, 2018 Page 46 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Table 3.13. Percent of Patients Reporting Adverse and Serious Adverse Events from RCTs Kalydeco Orkambi Symdeko 8 11 STRIVE KONDUCT TRAFFIC/TRANSPORT±16 EVOLVE18 EXPAND22 G551D R117H Homozygous F508del Homozygous F508del Heterozygous F508del 48 weeks 24 weeks 24 weeks 24 weeks 8 weeks Active Placebo Active Placebo Active Placebo Active Placebo Active Placebo N 83 71 34 35 369 370 251 258 162 162 Any Adverse Event 82 (99%) 78 (100%) 32 (94%) 35 (100%) 351 (95.1%) 355 (95.9%) 227 (90.4%) 245 (95.0%) 117 (72%) 126 (78%) (AE) Any AE Grade ≥3 NR NR NR NR NR NR 22 (8.8%) 29 (11.2%) 4 (2%) 9 (6%) Any Serious Adverse 20 (24%) 33 (42%) 4 (12%) 6 (17%) 64 (17.3%) 106 (28.6%) 31 (12.4%) 47 (18.2%) 8 (5%) 14 (9%) Event (SAE) Any AE Leading to 1 (1%) 4 (5%) 0 0 17 (4.6%) 6 (1.6%) 7 (2.8%) 8 (3.1%) 0 1 (1%) Discontinuation AE Resulting in Death 0 0 0 0 0 0 0 0 0 0 Most Common Adverse Events ALT Increased 3 (3.6%) 3 (3.9%) NR NR 10 (2.8%) 16 (4.4%) 8 (3.2%) 13 (5.0%) 1 (0.6%) 1 (0.6%) AST Increased 3 (3.6%) 3 (3.9%) NR NR 11 (2.9%) 11 (3.0%) NR NR 1 (0.6%) 0 Infective PEx of CF 34 (41%) 50 64.1%) 13 (38%) 14 (40%) 132 (35.8%) 182 (49.2%) 75 (29.9%) 96 (37.2%) 21 (13%) 31 (19%) Cough 27 (32.5%) 33 (42.3%) 10 (29%) 9(26%) 104 (28.2%) 148 (40%) 66 (26.3%) 84 (32.6%) 23 (14%) 16 (10%) Increased Sputum NR NR 5 (15%) 4 (11%) 54 (14.6%) 70 (18.9%) 36 (14.3%) 42 (16.3%) 14 (9%) 11 (7%) Dyspnea NR NR NR NR 48 (13%) 29 (7.8%) 16 (6.4%) 18 (7.0%) 9 (6%) 11 (7%) Abnormal 11ⱡ Respiration/Chest NR NR NR NR 32 (8.7%) 22 (5.9%) 11ⱡ (4.4%) 2 (1.2%) 0 (4.3%) Tightness Hemoptysis 9 (10.8%) 17 (21.8%) 0* 6* (23%) 50 (13.6%) 50 (13.5%) 26 (10.4%) 35 (13.6%) 12 (7%) 14 (9%) Diarrhea 11 (13.3%) 10 (12.8%) 5 (15%) 4 (11%) 45 (12.2%) 31 (8.4%) 17 (6.8%) 23 (8.9%) 13 (8%) 10 (6%) Nausea 13 (15.7%) 9 (11.5%) NR NR 46 (12.5%) 28 (7.6%) 23 (9.2%) 18 (7.0%) 9 (6%) 10 (6%) Fatigue NR NR NR NR NR NR 16 (6.4%) 31 (12.0%) 12 (7%) 16 (10%) NR: not reported ± TRAFFIC/TRANSPORT, 400 mg only; ALT/AST: alanine aminotransferase/aspartate aminotransferase ⱡ Chest discomfort=0%, *Participants>18 years (24 ivacaftor; 26 placebo) ©Institute for Clinical and Economic Review, 2018 Page 47 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Common side effects of CFTR modulators include rash, dizziness, headache, and upper respiratory tract infection,76 and nasopharyngitis.6 Additional side effects are reported in Table 3.13. FDA labels for all three modulators include monitoring for elevated liver enzymes (alanine and aspartate transaminase) and cataracts, as these have been reported with CFTR modulator use.76,6,25 Concomitant use of CFTR modulators with CYP3A inhibitors is not recommended due to drug interactions. Through stakeholder input, ICER was told that chest discomfort (often reported as chest tightness or abnormal respiration), was one of the primary reasons for Orkambi discontinuation. In TRAFFIC and TRANSPORT, abnormal respiration was reported in 8.7% of individuals receiving 400 mg lumacaftor twice daily compared to 5.9% of individuals receiving placebo.16 The long-term follow- up study, PROGRESS, reported rates of abnormal respiration between 10-17%.19 Individuals in the placebo arm in TRAFFIC/TRANSPORT reported higher rates of chest tightness than those originally randomized to active drug out to 96-weeks.19 Additionally, individuals with baseline ppFEV1 < 70% predicted reported more chest tightness than those with baseline ppFEV1 ≥ 70% (11-20% vs. 6-8%, respectively in the 400 mg lumacaftor twice daily arm).19 A real world cohort study at the Johns Hopkins Cystic Fibrosis Center after Orkambi approval (n=116) showed that nearly 20% of patients reported chest tightness.24 For Symdeko, chest discomfort was reported as zero in the F508del homozygous population and 1.2% in the heterozygous population.18,22 Orkambi is reported to have significant drug interactions that are not seen with Symdeko.6,25 Meta-Analyses of Harms Across Interventions Eleven publications provided data on rates of discontinuation due to adverse events. 4,8-10,15,16,18- 20,22,77 The studies evaluated ivacaftor 300 mg/day (five studies), Orkambi 800/500 mg/day (five studies), Symdeko 100/300 mg/day (three studies), and placebo (eight studies). Studies or study arms of nonstandard doses were omitted from analysis. With one exception, described below, across studies, duration of intervention did not correlate with drug discontinuation rates by metaregression. Summary rates of discontinuation due to adverse events were: Kalydeco 1.2% (95% CI 0.3, 2.5), Orkambi 6.3% (95% CI 3.7, 9.6), Symdeko 2.5% (95% CI 0.1, 8.3), and placebo 2.1% (95% CI 1.1, 3.4) (Appendix D, Figures D12-15). The three Symdeko studies were heterogeneous, with a small study having a higher discontinuation rate (2/17, 11.8%) than the other two studies (0 and 2.8%) resulting in a wide confidence interval.4 A crude comparison across interventions suggests that discontinuation due to adverse events is significantly more likely to occur with Orkambi than Kalydeco, Symdeko, or placebo, which all had similar rates of drug discontinuation due to adverse events. For Orkambi, no correlation with treatment duration was evident (by meta- regression) from four to 72 weeks (P=0.37); however, inclusion of the study arm of people on drug ©Institute for Clinical and Economic Review, 2018 Page 48 Final Evidence Report – Cystic Fibrosis Return to Table of Contents for 96 weeks (with a discontinuation rate of 7.4%) yielded a significant correlation of 0.4% per month (95% CI 0.1, 0.7; P=0.018). Two publications provided data on grade 3 or 4 severe adverse events.18,22 The studies evaluated ivacaftor 300 mg/day (1 study), Symdeko 100/300 mg/day (2 studies), and placebo (2 studies). In both studies, the drugs were taken for 24 weeks. Summary rates of grade 3 or 4 severe adverse events were: Kalydeco 5.1% (95% CI 2.6, 9.9), Symdeko 5.3% (95% CI 0.8, 13.3), and placebo 8.4% (95% CI 3.6, 14.9) (Appendix D, Figures D16- 17).18 However, for both Symdeko and placebo, the reported rates of grade 3 or 4 severe adverse events were considerably lower in EXPAND than in EVOLVE; this resulted in statistical heterogeneity between the two studies. Nevertheless, within and across studies, all interventions had similar rates of grade 3 or 4 severe adverse events. Controversies and Uncertainties Many factors limit or complicate our ability to interpret the clinical benefits of CFTR modulators. Perhaps the largest limitation is the complexity of CF genetics, which directly impact disease severity and progression. Each population reviewed–gating and residual function mutations (Class III), heterozygous F508del, and homozygous F508del (Class II)–has unique genetic and disease variability marked by a general deterioration in lung and pancreatic function. As such, interpreting clinical trial outcomes from relatively small samples in short periods of time (one year or less), may provide a limited picture of clinical benefit. In addition, the FDA approval of Symdeko was not limited to the population studied in the EXPAND trial which required at least one F508del mutation. Therefore, we cannot state with any certainty, how generalizable the results from EXPAND are to patients with different genetic makeup. Additionally, the myriad therapies employed in best-practice CF symptom management may increase the uncertainties of the benefits of CFTR modulators. Standard-of-care treatments include dornase alfa and hypertonic saline; azithromycin, tobramycin, and aztreonam are also used in those with Pseudomonas aeruginosa infections. Data from the CFFPR indicate 88% of registry patients use dornase alfa and 70% use hypertonic saline; of those who are Pseudomonas aeruginosa- positive, two-thirds or more use inhaled tobramycin and azithromycin (69% and 66%, respectively), 43% use inhaled aztreonam, and most participants in CFTR modulator trials were concurrently taking some or all these standard-of-care treatments during study treatment. As expected, these interventions positively impact pulmonary status in many or most patients. Both dornase alfa and tobramycin have been shown to improve FEV1 in children with CF (3-6% and 8-20%, respectively).8,71 In contrast, hypertonic saline use, which was shown to decrease the risk of pulmonary exacerbations by 66% compared to placebo8, was not permitted during Kalydeco Phase III trials, a restriction which may limit the applicability of the study to typical care. The open-label extension study allowed the use of hypertonic saline; however, no data was available for our review. These interactions should be systematically evaluated in future studies. ©Institute for Clinical and Economic Review, 2018 Page 49 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Interpreting lung function using FEV1 comes with numerous uncertainties. FEV1 is a surrogate measure of disease severity that attempts to measure lung function relative to what is predicted in healthy persons of the same age and sex. Despite being well-defined in literature and widely used in clinical trials and clinical practice, it remains unclear what magnitude of change in FEV1 is clinically relevant. While there is precedent for FDA approval based on 2-3% absolute change in ppFEV1, it is unclear how this translates to improved survival and/or quality of life. Similarly, the lung clearance index is a new surrogate outcome that has had limited long-term use. While validation studies are ongoing, there have also been debates about which tracer gas is most optimal and adequate training and diffusion of the procedure. There are also few direct correlation studies between lung function surrogates such as ppFEV1 and lung clearance index in people with CF and hard clinical endpoints such as lung transplant or death.12 Stakeholders identified uncertainties around CFTR modulator treatment decisions considering their personal experiences. One parent, for example, shared that their child experienced beneficial weight gains while on Orkambi but simultaneously experienced lung function deterioration. Not only does this patient’s experience provide an example of often difficult decision making needed regarding tradeoffs on the apparent effects of the drugs (here weight gain vs. lung function), but it also highlights that not all patients will respond to CFTR modulator treatment the same or as predicted based on the study evidence. Nearly 85% of people with CF in the United States receive care at an accredited CF center, which provide multidisciplinary clinical care. This high-quality, specialized approach to care has improved survival for people with CF. Many of the CF trials discussed in this report were conducted in such accredited CF centers, and these trials demonstrated improvements in health outcomes among those receiving best supportive care are likely with the addition of appropriate CFTR modulators. We identified uncertainties, however, regarding whether beneficial gains in survival are distributed unequally due to differences in access to US CF care centers. For example, Canadian CF patients have been living longer since the mid-1990s and currently live, on average, 10 years longer than American CF patients.26,27 When comparing the US and Canada, the difference between Canadian and US survival disappeared when US patients receiving Medicare and Medicaid were excluded from survival data, suggesting CF patients receiving care through US public health insurance have a survival disadvantage.26,28 It is unclear whether patients are receiving different care depending on their insurance type or whether American CF patients with public insurance are more likely to have important socioeconomic disadvantages that affect their CF management. While long-term studies are underway to evaluate the impact of CFTR modulators on long-term survival, ensuring access to the highest quality CF care in the interim may improve the survival of all CF patients. Evaluating “adverse events” in studies of people with CF is challenging because the most frequently reported events in studies are likely not side effects due to the drugs, but instead are adverse outcomes due to the underlying disease that occur while patients are taking the drugs. The “adverse events” reported across all trials included outcomes expected with CF, like cough or ©Institute for Clinical and Economic Review, 2018 Page 50 Final Evidence Report – Cystic Fibrosis Return to Table of Contents increased sputum production. For example, pulmonary exacerbation, a very common event for people with CF, was reported as both a clinical outcome and an adverse event, sometimes in the same study. Furthermore, across studies, specific adverse events commonly occurred more frequently among those taking placebo than those taking CFTR modulators; this was even found for adverse events that were ascribed to the drugs. For example, in STRIVE, serious adverse events were about twice as common with placebo than with ivacaftor;8 in EXPAND, more patients taking placebo had adverse events considered by the investigator to be related or possibly related to the trial regimen with placebo than with Symdeko.22 Finally, cystic fibrosis is a multisystem disease, yet many aspects of the disease have not been systematically researched. Thus, our rating of the impact of CFTR modulators is highly dependent on those outcomes measured in the trial data, namely pulmonary function, weight, respiratory symptom-related quality of life and the number, type and annualized rate of pulmonary exacerbations. 3.4 Summary and Comment Table 3.14. ICER Evidence Rating for Use of Kalydeco for Cystic Fibrosis Caused by the G551D, non- G551D Gating, and R117H Residual Function Mutations. Population/Genetic Group ICER Evidence Rating G551D, Other Gating, Non-G551D Gating Mutations, And R117H Residual Function Mutations Kalydeco A Kalydeco for patients with cystic fibrosis caused by gating and residual function mutations: • Kalydeco provides improvements in ppFEV1 (5.0 to 10.7 percentage points in different populations), weight, and respiratory-symptom-related quality of life (9.6 to 12.6 points) for children, adolescents, and adults (over 24 weeks). Longer-term follow-up (up to three years) shows lung function, weight, and quality of life gains are durable across all gating mutations. • However, limited data suggest 6-11 year olds with the R117H mutation may not have improved respiratory function and quality of life with Kalydeco treatment. • Pulmonary exacerbations were less frequent (HR=0.46), shorter, and required fewer hospitalizations and intravenous antibiotics for patients taking Kalydeco. • Fewer patients (across populations) discontinued Kalydeco due to adverse events (1.2%) than with placebo (2.1%). Across all subpopulations, rates of discontinuation due to adverse events and severe adverse events were similar for Kalydeco and placebo. ©Institute for Clinical and Economic Review, 2018 Page 51 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Given the relatively consistent evidence arising from controlled trials of lung function improvement, with clinically significant improvements and associated reductions in pulmonary exacerbations, and with no evidence of significant harms, we have high certainty Kalydeco provides a substantial (moderate-large) net health benefit relative to best supportive care. We therefore assign a rating of “superior” (A) to the comparative clinical effectiveness of Kalydeco in this population. Homozygous F508del mutations Table 3.15. Evidence Rating for the Use of Orkambi for Cystic Fibrosis Caused by Two Copies of the F508del Mutation Population/Genetic Group ICER Evidence Rating Homozygous F508del Mutation Orkambi B Symdeko B+ Orkambi for patients with cystic fibrosis caused by two copies of the F508del mutation: • Orkambi improved ppFEV1; however, changes in absolute ppFEV1 may not be considered clinically important (2.4 to 2.8 percentage points). • At 24 weeks, BMI increases with Orkambi among those aged 12 years and older (0.61 kg/m2), which was maintained over the subsequent 96 weeks; but no significant difference was found in a study of younger children. • Treatment improved respiratory symptom-related quality of life in patients age 12 and older (2.2 points); a similar improvement was found in a smaller study of children 6-11 years old, but the effect was not statistically significant. • The rate of pulmonary exacerbation was lower for patients aged 12 and older taking Orkambi (rate ratio = 0.61); data were not reported in the study of younger children. • Chest tightness (abnormal respiration) was reported as a side effect for those taking Orkambi ranging from 8% in the Phase III trials to 20% in a real-world post-approval study. • Rates of discontinuation due to adverse events were higher for Orkambi (4.6%) than for placebo (1.6%) within a trial in this population. Similar results were seen among all studies across populations (6.3% vs. 2.1%, respectively). In two large Phase III trials and an accompanying 96-week open-label extension study, Orkambi provided a consistent magnitude of approximately 3 percentage point improvement in ppFEV1 as well as a reduced rate of decline in lung function over time, however, patients also reported drug- drug interactions and side effects leading to discontinuation. Thus, for patients homozygous for the F508del mutation, we have high certainty Orkambi provides a small net health benefit relative to placebo (i.e. best supportive care), and therefore assess the evidence to be “incremental” (B). ©Institute for Clinical and Economic Review, 2018 Page 52 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Symdeko for patients with cystic fibrosis caused by two copies of the F508del mutation: • Treatment with Symdeko improved absolute ppFEV1 (4.0 percentage points) and respiratory-related quality of life (5.1 points) compared to placebo over 24 weeks. No significant differences in weight were reported. • Treatment reduced the rate of pulmonary exacerbation over 24 weeks (rate ratio = 0.53). • In this population, rates of discontinuation due to adverse events were similar for Symdeko (2.8%) and placebo (3.1%). Similar results were seen among all studies across populations (2.5% vs. 2.1%, respectively). A single, parallel-arm, Phase III trial showed a moderate improvement in ppFEV1 however, the trial was relative short in duration. For patients homozygous for the F508del mutation, we have moderate certainty that Symdeko provides a small or substantial net health benefit, with high certainty of at least a small net health benefit relative to placebo (i.e., best supportive care). Therefore, we assess the evidence to be “incremental or better” (“B+”). Heterozygous F508del with a residual function mutation Table 3.16. Evidence Rating for The Use of Symdeko For Cystic Fibrosis Caused by a Single Copy of The F508del Mutation with An Approved Residual Function Mutation Population/Genetic Group ICER Evidence Rating Heterozygous F508del with Residual Function Mutation Symdeko B+ Symdeko for patients with cystic fibrosis caused by one copy of the F508del mutation and a second mutation amenable to Symdeko: • Treatment with Symdeko resulted in clinically relevant improvement in absolute ppFEV1 (6.8 percentage points) and respiratory symptom-related quality of life (11.1 points). • The treatment effect on pulmonary exacerbations and BMI was exploratory only due to small numbers and short duration A single 8-week cross-over trial provided evidence of the improvement in lung function compared with placebo. Long-term studies to confirm these data are required. For patients heterozygous for the F508del mutation with an approved residual function mutation, we have moderate certainty that Symdeko provides a small or substantial net health benefit, with high certainty of at least a small net health benefit relative to placebo (i.e., best supportive care). Therefore, we assess the evidence to be “incremental or better” (“B+”). ©Institute for Clinical and Economic Review, 2018 Page 53 Final Evidence Report – Cystic Fibrosis Return to Table of Contents 4. Long-Term Cost Effectiveness 4.1 Long-Term Cost Effectiveness Overview The objective of this analysis was to estimate the cost-effectiveness of CFTR modulator treatments plus best supportive care for CF patients. We modeled three different populations based on mutation status, and three different CFTR modulators or combinations of modulators that have indications in one or more CF populations. We evaluated Kalydeco for individuals with gating mutations, and Orkambi and Symdeko for individuals who are homozygous for the F508del mutation. For patients who are heterozygous for the F508del mutation with a residual function mutation, we evaluated Symdeko and Kalydeco as possible CFTR modulator treatments. The model structure for this assessment is described below. CF is a condition which falls under ICER’s ultra-rare disease framework. Therefore, we considered dual base-case analyses that reflect both health system and societal perspectives. While the impact of this disease on patient and caregiver productivity, informal caregiver time, education, and disability costs can be substantial, the impact of treatment with the CFTR modulators on societal costs is not expected to be as substantial, because the drugs do not greatly reduce the daily burdens associated with usual CF supportive care. We therefore present the results from a modified societal perspective as a scenario analysis rather than as part of the base case. Outcomes were estimated over a lifetime time horizon using one-year time increments from treatment initiation until death. The primary health outcome was quality-adjusted life years (QALYs) but we also report life expectancy and the lifetime number of acute pulmonary exacerbations. QALYs are a measure that combines both length of life and quality of life into a single measure, and are the recommended metric for use in cost-effectiveness analyses.78 The impact inventory is provided in Appendix Table E1. Costs and health outcomes were discounted at 3% per year. The model was developed in TreeAge software version 2017 (Williamstown, MA). A preliminary version of the results in this section of the report were presented publicly on April 26, 2018 and included some data inputs based on 2016 costs; this version includes results generated from re-running the models with those data inputs updated to 2017 values. ©Institute for Clinical and Economic Review, 2018 Page 54 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Cost-Effectiveness Model: Methods Model Structure We developed a de novo discrete-time microsimulation model. The primary model variable was percent predicted forced expiratory volume in one second (ppFEV1), modeled as a continuous variable. This model type was chosen to account for the continuous nature of ppFEV1 and to capture the primary effect of the CFTR modulator drugs (i.e., increase in ppFEV1 or slowing the decline of ppFEV1 over the longer term). For each population, a cohort of CF patients begins the model at the age of drug initiation. We assigned a gender distribution based on the current prevalent CF population.1 Each simulated patient is assigned a ppFEV1 value drawn from a distribution and then experiences annual age-specific declines in lung function. The means and standard deviations (SD) of the initial ppFEV1 distributions were set so that when the cohort reached the average ages reported in the relevant clinical trials, the means and ranges of the ppFEV1 matched those observed in the relevant trials. For example, for individuals with a G551D mutation we set the starting distribution so that the population was similar to the ppFEV1 mean and range (84.2%; 44.0%-133.8%) of the ENVISION trial at age nine (mean age) and the mean and range (63.3%; 31.6%-98.2%) of the STRIVE trial at age 26 (mean age).8,9 In addition to ppFEV1, the model tracked the values of other variables for each simulated person: weight-for-age z-score, number of acute pulmonary exacerbations per year (defined as exacerbations requiring intravenous antibiotics), pancreatic sufficiency, lung transplantation, and diagnosis of CF-related diabetes or B. cepacia infection. During any given year, a simulated person may experience a change in their ppFEV1, experience one or more pulmonary exacerbations, be diagnosed with diabetes mellitus or B. cepacia infection, or undergo lung transplantation if their ppFEV1 falls to 30% or below. The annual risk of death is influenced by all of these variables. Figure 4.1 shows a diagram of the model, with the risk of pulmonary exacerbation and lung transplantation dependent on the ppFEV1 value. Persons are simulated for their lifetime, accumulating life years, QALYs (i.e., life years weighted by a quality-of-life value) and costs each year. For the treatment arms, we allowed the initial ppFEV1 and weight-for-age z-score values to change based on trial results or by assumption if no trial evidence existed. We also allowed the risk of acute pulmonary exacerbation to decrease with treatment, independent of the improvement in ppFEV1. ©Institute for Clinical and Economic Review, 2018 Page 55 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Figure 4.1 Model Framework Target Population We considered three distinct populations for this analysis. The first population includes individuals with CF and gating mutations, such as the G551D mutation, consistent with the FDA-approved indications for Kalydeco. The age of treatment initiation is two years old, consistent with FDA labeling. The initial distribution of ppFEV1 in this population was assumed to be normal with a mean (SD) of 96.37 (12.02). The second population includes individuals with CF who are homozygous for the F508del mutation. This population is eligible for treatment with Orkambi or Symdeko, and we assumed that the age of treatment initiation was six years old for both treatments given that recommended age for Symdeko will likely be lowered with additional trials, as was the case for Orkambi. The initial distribution of ppFEV1 in this population was assumed to be normal with a mean (SD) of 88.09 (13.39). The third population includes individuals with CF who are heterozygous for the F508del mutation and a residual function mutation that is potentially responsive to Symdeko. This population is eligible for treatment with Symdeko combination or Kalydeco, and the age of treatment initiation is 12 years old. The initial distribution of ppFEV1 in this population was assumed to be normal with a mean (SD) of 81.93 (15.41). For all populations, we truncated the ppFEV1 distributions at a mimimum of 44 and maximum of 134. We did not evaluate treating individuals with CF and the R117H mutation (although evidence is summarized in Section 3) because this is a small population with very limited trial evidence and a substantially different prognosis compared with individuals with gating mutations. ©Institute for Clinical and Economic Review, 2018 Page 56 Final Evidence Report – Cystic Fibrosis Return to Table of Contents We found that individuals with gating mutations or who are homozygous for the F508del mutation are similar in terms of their expected ppFEV1 trajectories and in terms of other variables (e.g., pancreatic sufficiency). In general, individuals heterozygous for the F508del mutation with a residual function mutation have a better prognosis, and have a higher percentage with pancreatic sufficiency.22,79 We assumed that best supportive care consists of the following pulmonary and pancreatic therapies (percent utilization): dornase alfa (87.5%), inhaled tobramycin (69.4%), inhaled aztreonam (43.2%), azithromycin (65.5%), hypertonic saline (70.7%), oxygen (10.4%), non-invasive ventilation (2.8%), pancreatic enzyme replacement therapy (86.5%) and supplemental feeding (tube or oral, 56.4%).1 Individuals with or developing CF-related diabetes were assumed to require oral hyperglycemic agents (3.9%), intermittent insulin (5.9%) and chronic insulin (76.3%), and to require diabetes- specific follow-up care (e.g., HbA1c measurements). We assumed that best supportive care applied to all individuals, whether on CFTR modulators or not, but that the intensity of therapy varied by lung function category. Acute pulmonary exacerbations were defined as those that involve treatment with IV antibiotics either in the hospital or with home treatment. We estimated disease management costs for all CF individuals, including annual clinic visits and all other costs except those for acute pulmonary exacerbations and lung transplantation; the disease management costs varied by level of ppFEV1. Acute pulmonary exacerbations and lung transplantation were costed separately. The rationale for this approach was that the disease management costs for a given level of ppFEV1 will be the same for patients in both arms (modulator therapy vs. no modulator therapy). Disease management costs will vary as individuals who live longer will have higher management costs, although individuals on modulator therapy will also have better lung function, resulting in reductions in these costs. Treatment Strategies For each population, we compared the eligible CFTR modulator treatment(s) plus best supportive care best supportive care alone. We did not compare CFTR modulator treatments directly with each other. Key Model Characteristics and Assumptions We made several assumptions for this analysis (Table 4.1). ©Institute for Clinical and Economic Review, 2018 Page 57 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Table 4.1. Key Model Assumptions Assumption Rationale ppFEV1 does not increase over time. We made this assumption because average lung function generally declines with age. Best supportive care is the same in all treatment We only assume that CFTR modulator therapy will arms, given the same ppFEV1 category. have an impact on costs associated with acute pulmonary exacerbations, lung transplantation, and ppFEV1-specific disease management. All other costs of supportive care not associated with lung function (e.g., pancreatic insufficiency, CF-related diabetes)will not be affected by CFTR modulator therapies, which has been supported by limited data. The weight-for-age z-score is constant over the There is limited evidence for how weight-for-age z- lifetime of a patient. score changes over time and this assumption has been used in other CF economic evaluations. The risk of B. cepacia infection over time does not The occurrence of B. cepacia infection was depend on lung function severity. incorporated only because it impacts CF-specific mortality risk and was modeled only as a function of age. The drug effects are modeled as an increase in These are the well-documented effects of CFTR ppFEV1, an increase in weight-for-age z-score, and a modulator drugs. decrease in the annual number of acute pulmonary exacerbations relative to best supportive care alone. CFTR drugs decrease the annual number of acute Modeling the impact of ppFEV1 changes and an pulmonary exacerbations through the increase in independent effect of drug treatment on acute ppFEV1 (the risk of exacerbations depends on lung pulmonary exacerbation rates allowed us to calibrate function). There is also an independent effect of to the reductions in exacerbations observed in clinical drugs on acute pulmonary exacerbation, trials. independent of the lung function effect. Treatment discontinuation rates are the same as Because we used trial effectiveness estimates, we those reported in the trials. There is no further drug assumed the same percentage of patients are taking discontinuation after the end of the trial time the drug in the model as in the trials, irrespective of horizon. available data on real-world discontinuation. Model Inputs Clinical Inputs We modeled the ppFEV1 trajectories through age-specific annual declines.35,80 To match the mean ppFEV1 values observed in the drug trials, we allowed the decline for ages under nine to be slightly higher than reported in the literature for CF individuals with a gating mutation or who are homozygous for the F508del mutation. The annual risk of having acute pulmonary exacerbation was modeled as a function of ppFEV1, age, and the number of acute pulmonary exacerbations the previous year.81-83 The annual risk of lung transplant was 0% for individuals with ppFEV1 >30% as ©Institute for Clinical and Economic Review, 2018 Page 58 Final Evidence Report – Cystic Fibrosis Return to Table of Contents per guidelines.84 The annual risk of diabetes was modeled as a function of age and sex.85 We assumed that 5% of CF individuals with a gating mutation or who are homozygous for the F508del mutation had pancreatic sufficiency at diagnosis and that this proportion was stable over lifetime.86 For CF individuals heterozygous for the F508del mutation with a residual function mutation, we estimated that 84% had pancreatic sufficiency at diagnosis based on the EXPAND trial population.22 Similarly, we assumed that weight-for-age z-score is constant for each person throughout life (in the absence of modulator therapy), which was set to -0.23.69 The risk of B. cepacia infection over time was derived from age-specific prevalence values from the CFF Registry and does not depend on lung function severity.1 Base-case values are listed Table 4.2. Table 4.2. Key Model Inputs Baseline Value Source Annual Decline in ppFEV1 Age 6-8 years -1.12 (-2.00 for gating or F508del homozygous mutation*) Age 9-12 years -2.39 Konstan, Age 13-17 years -2.34 2007;Konstan, Age 18-24 years -1.92 201235,80 Age ≥25 years -1.45 Annual Rate of Acute Pulmonary Exacerbation by Age and ppFEV1 Age <18 8.5938*exp(-0.035*ppFEV1) Goss, 2007; Age ≥18 3.7885*exp(-0.026*ppFEV1) Whiting, 201481,82 Hazard Ratio for Increase in Rate of Pulmonary Exacerbation (Relative to 0 Exacerbations the Prior Year) 1 Exacerbation the Prior Year 1.6 VanDevanter, 2 Exacerbations the Prior Year 2.4 201683 3+ Exacerbations the Prior Year 4.0 Number of Pulmonary Exacerbations Per Year: 1, 2, 3+ (Conditional On 1+) Age < 5 0.76 / 0.19 / 0.05 Age 5-10 0.68 / 0.20 / 0.12 Age 11-17 0.54 / 0.22 / 0.24 Goss, 200781 Age 18-29 0.48 / 0.23 / 0.29 Age ≥30 0.53 / 0.27 / 0.20 Annual Risk of Lung Transplantation ppFEV1 >30 0 Thabut, 201387 ppFEV1 ≤30 0.647 Annual Risk of CF-Related Diabetes (Male, Female) Age 0-9 0.008, 0.016 Age 10-19 0.039, 0.060 Age 20-29 0.049, 0.071 Adler, 200885 Age 30-39 0.065, 0.072 Age 40+ 0.051, 0.029 *Assumed higher declines for youngest age group for individuals with a gating mutation or who are homozygous for the F508del mutation to fit trial-specific means for each population. ©Institute for Clinical and Economic Review, 2018 Page 59 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Clinical Probabilities/Response to Treatment To model the treatments’ effects, we assumed that there is an immediate increase in ppFEV1 and improvement in weight-for-age z-score, as observed in the trials or by assumption if no trial evidence existed (Table 4.3). We assumed no ppFEV1 decline on drug for the first two years and then a decline that is 50% of the best supportive care rate thereafter.68,74 We assumed that the increase in weight-for-age z-score would persist for a patient’s lifetime.68 The drug trials reported reductions in acute pulmonary exacerbation rates (e.g., rate ratios). When available we used the rate ratios for acute pulmonary exacerbations that required IV antibiotics. We assumed that part of the decline in number of acute pulmonary exacerbations was due to the increase in ppFEV1. However, we also allowed for an independent effect of the drugs on reducing the acute pulmonary exacerbation rates. For example, the rate ratio for Kalydeco + best supportive care versus best supportive care alone was 0.56.8 The model-generated rate ratio for a population similar to STRIVE was 0.83 when we assumed that the decline in acute pulmonary exacerbations with drug was only due to the increase in ppFEV1. We assumed that Kalydeco also had an independent effect on the reduction in acute pulmonary exacerbations by reducing the chance that an individual will experience an exacerbation and reducing the number of multiple acute pulmonary exacerbations among those patients experiencing at least one exacerbation. We varied these assumptions until the model-generated rate ratio was 0.56. The independent effect from Kalydeco for CF individuals with gating mutations was to reduce the risk of exacerbation and the number of multiple exacerbations (given at least one) by 22%. This approach assumes that the reduction in exacerbation rate was a combination of a lower percentage of patients experiencing an exacerbation in a year and fewer exacerbations among those who do experience at least one. ©Institute for Clinical and Economic Review, 2018 Page 60 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Table 4.3. Treatment Effectiveness Inputs Change in Acute Increase in ppFEV1 Weight-For Age Pulmonary Source (Mean, 95% CI) Z-Score (Mean, Exacerbation RR 95% CI)* CF Individuals with a Gating Mutation 10.0 (4.5-15.5) 0.56 0.35 (0.20-0.51) Davies, 2013;Ramsey, 2011;Borowitz, Kalydeco 2016;McKone, 20148,9,15,69 CF Individuals Who are Homozygous for the F508del Mutation Orkambi 2.8 (1.8-3.8) 0.44 Same as above Wainwright, † 4.0 (3.1-4.8) 0.54 Same as above 2015;Konstan, 2017;Taylor-Cousar, Symdeko 2017; NICE, 201616,18,19,88,89 CF Individuals Who are Heterozygous for the F508del Mutation with a Residual Function Mutation Symdeko 6.8 (5.7-7.8) 0.54 (0.26-1.13)‡ Same as above Rowe, 201722 Kalydeco 4.7 (3.7-5.8) 0.46 (0.21-1.01)‡ Same as above *Change in weight-for-age z-score reporting is variable and not consistent. We assumed that all drugs would achieve the same effect on weight-for-age z-score as observed in Borowitz et al.69 †Rate ratio (RR) is for exacerbations with either IV antibiotics or hospitalization (or both). We assume that all hospitalizations would involve IV antibiotics. ‡RR reported for pulmonary exacerbations defined by modified Fuch’s criteria (not necessarily requiring IV antibiotics). Mortality Each year simulated individuals face a risk of dying. We modeled this probability as a combination of their age-specific mortality rate based on the US life tables90 and a CF-specific rate. CF-specific mortality rates were a function of sex, ppFEV1, weight-for-age z-scores, number of acute pulmonary exacerbations, diagnosis of CF-related diabetes, pancreatic sufficiency, and B. cepacia infection.91 The Liou analysis also found that S. aureus infection was an independent predictor of mortality; however, the impact of infection was to decrease the mortality rate. Because we found no explanation as to why infection with S. aureus would be associated with better survival, and because of the recent rise in methicillin resistant S. aureus1, we opted to not include this characteristic in the mortality rate function. The following equation was used to model the annual mortality rate for age a (ℎ 𝑎 ) for non-transplanted patients91: ©Institute for Clinical and Economic Review, 2018 Page 61 Final Evidence Report – Cystic Fibrosis Return to Table of Contents ℎ 𝑎 = 𝑏 𝑎 𝑒(𝐾) 𝐾 = 0.15(𝑆𝐸𝑋 − 0.47) − 0.042(𝑝𝑝𝐹𝐸𝑉1 − 67.7) − 0.0280(𝑊𝐹𝐴 + 0.85) + 0.350(#𝑃𝐸 − 1.1) + 0.440(𝐷𝐼𝐴𝐵 − 0.061) − 0.140(𝑃𝑆 − 0.053) + 1.410(𝐵𝐴𝐼 − 0.032) − 0.280(#𝑃𝐸 − 1.1)(𝐵𝐴𝐼 − 0.032) The patient-specific parameters that affect mortality among non-transplanted patients were SEX (0 male, 1 female), ppFEV1 (%), WFA (weight-for-age z score), #PE (number of acute pulmonary exacerbations in the current year), DIAB (0 no diagnosis of diabetes, 1 yes), PS (0 no pancreatic sufficiency, 1 yes), BAI (0 no B. cepacia infection, 1 yes). The age-specific baseline hazard () was a product of the age-specific rates from the US life tables90 and an adjustment factor that was needed to match the life expectancy targets of a CF cohort. Survival after lung transplant was a function of time since transplant and was better than prior to transplant.87 Utilities We used the linear interpolation of EQ-5D utilities by ppFEV1 conducted by Schechter et al. (Table 4.4).92 These utilities were used to weight each year of life to accumulate QALYs over an individual’s lifetime. The extrapolation was based on EQ-5D values estimated for ppFEV1 groups (0.86 for >70%, 0.81 for 40%-69%, and 0.64 for <40%) among cystic fibrosis patients provided to Tappenden et al. for a NICE economic evaluation.93 Because we modeled ppFEV1 as a continuous variable, we used a linear function to assign utilities based on ppFEV1 (utility = 0.593047 + ppFEV1*0.003476). We used similar assumptions as Tappenden et al. and applied a short-term utility decrement of 0.17 during the year in which an acute pulmonary exacerbation occurred.93 We used the same utility used by Schechter et al.92 for the first year after lung transplantation (0.32) based on quality of life study of lung transplantation in patients with cystic fibrosis.94 Subsequent years after transplantation were set to a utility equivalent to a ppFEV1 of 70%-79%: 0.838. Table 4.4. Utility Values by Level of ppFEV1 (Derived from Schechter et al.) ppFEV1 (%) Utility >90 0.920 80-89 0.873 70-79 0.838 60-69 0.801 50-59 0.765 40-49 0.729 30-39 0.692 20-29 0.653 <20 0.625 ppFEV1: Percent predicated forced expiratory volume in 1 second ©Institute for Clinical and Economic Review, 2018 Page 62 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Adverse Events Serious and severe adverse events were generally comparable across treatment groups and often higher in the placebo arms. Therefore, we did not explicitly model adverse events in terms of added costs or disutilities but assumed that patients who experienced a bothersome adverse event would discontinue the drug. As the discontinuation rates typically reported in the trials were greater than the reported discontinuation rates due to adverse events, we assumed that the reported discontinuation rates included discontinuation due to adverse events. Economic Inputs Drug Acquisition Costs Annual net drug acquisition costs for each medication were used in the model. We could not calculate net prices for all drugs using our standard source (SSR Health, LLC), as this source did not include consistent publicly-disclosed net sales figures for the specialty drugs in this review. We therefore used data from the Federal Supply Schedule (FSS) to determine discounted (net) prices of Kalydeco and Orkambi (Table 4.5).29 The FSS is a price schedule set forth by the U.S. General Services Administration (GSA) that is used in negotiation with manufacturers of drugs, medical equipment, and supplies and service contracts for the VA and other federal organizations. As Symdeko was only recently approved by the FDA, information on its net pricing was not yet available. We therefore applied the FSS discount rate for Orkambi (3.2%) to the wholesale acquisition cost (WAC) of Symdeko to arrive at an estimated net price. Table 4.5. Drug Cost Inputs WAC per Net price per Annual Drug Intervention Administration Unit Unit/Dose*95 Unit† Cost Kalydeco Oral twice daily 150mg tablet $426.72 $424.15 $309,841.58 Orkambi Age 6-11 years Oral, 2 tablets 100mg/125mg $186.78 $180.76 $264,085.53 twice daily Age 12+ years Oral, 2 tablets 200mg/125mg $186.78 $180.76 $264,085.53 twice daily Symdeko Oral 100mg/150mg $400.08 $387.20 $282,656.00 (once/twice) daily *WAC as of January 12, 2018 †FSS prices as of January 2, 2018 Some prior cost-effectiveness analyses in CF have attempted to account for possible price changes over time, by assuming that the drug prices will decrease upon loss of patent exclusivity.82,96,97 For ©Institute for Clinical and Economic Review, 2018 Page 63 Final Evidence Report – Cystic Fibrosis Return to Table of Contents example, Dilokthornsakul et al. assumed that the prices of Kalydeco and Orkambi would drop to 10% of WAC after patent expiration.96,97 We chose not to make such an assumption in our current analysis, because attempts to model price changes over time would add an additional layer of uncertainty and speculation to our analysis, and while there have been calls to include price changes in cost-effectiveness analysis, the current convention is not to include estimates of changes in drug price throughout the life cycle.98,99 Estimating such changes may be especially difficult in the US market, where drug prices are mostly unregulated, and changes in prices occur relatively frequently. The timing of entry of other competitors (branded or generic) is difficult to predict, due to the possibility of patent litigation and “pay for delay” agreements. Generic drugs are generally expected to have discounted pricing relative to branded competitors, but the size of that future discount is difficult to estimate, particularly for rare diseases with limited to no competition. This was recently evidenced by the introduction of a new generic version of trientine hydrochloride (Syprine®), which entered with a 14% discount off a brand price that had increased by a factor of 30 in recent years.100 Finally, even products with historically stable pricing may be sold to or acquired by another manufacturer, who could decide to change pricing in dramatic and unpredictable fashion. Administration and Monitoring Costs We assumed that there were no additional costs associated with the administration and monitoring of the CFTR modulator drugs above best supportive care. Health Care Utilization Costs We assumed that annual CF-related healthcare costs over an individual’s lifetime consisted of three components (not including the cost of the CFTR modulator drugs): disease management, acute pulmonary exacerbations requiring IV antibiotics, and transplant-related costs. We used an approach similar to that taken by Dilokthornsakul et al. in their cost-effectiveness analyses.96,97 Both disease management and pulmonary exacerbation components incorporated a gradient cost structure that was derived from Lieu et al. to reflect increasing costs with increasing disease severity categories (≤40% ppFEV1, severe; between 40% and 70% ppFEV1, moderate; ≥70% pp FEV1, mild).30 An age-related adjustment (<18 or 18+) was included in the exacerbation component. The 2016 CFF Patient Registry data were used to calculate the adjustment, reflecting a higher proportion of total treatment duration spent in the hospital versus home IV treatment for children with a pulmonary exacerbation than for adults.1 This resulted in a lower cost per exacerbation for adults. Average cost estimates based on 1996 data30 do not include all currently available CF treatment and therefore are not likely reflective of current best supportive care costs. Several other studies have found higher average annual medical costs even after adjusting for inflation101;102. To derive current best supportive care costs, we used two average annual cost estimates provided by Scott ©Institute for Clinical and Economic Review, 2018 Page 64 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Grosse from the CDC based on his analysis of 2016 commercial payer and Medicaid claims data ($130,879 and $83,173 in 2016 US dollars) (S. Grosse, personal communication, April 12, 2018). We applied a 5% reduction to account for transplant-related costs, excluded CFTR-related costs, and updated to 2017 US dollars using the personal consumption expenditure (PCE) price index. We then calculated a weighted average based on health insurance information reported in the 2016 CFFPR showing a 60%/40% insurance mix (private/other).1 This resulted in an average annual cost estimate of $77,143, which was used to calibrate the best supportive care cost estimates. Transplant-related costs include the one-time cost of receiving a lung transplant followed by an annual cost associated with post-transplantation care. Estimates for the cost of a transplant and initial year following a transplant were derive from a 2017 Milliman Research Report.103 Annual costs were reduced for all subsequent years following the first year post-transplant based on estimates from a study of inpatient and outpatient billing services of lung transplantation patients at the University of Washington.104 The CF-related disease management and exacerbation costs were assumed to be zero for individuals in post-transplant years. Cost estimates are shown in Table 4.6 and are reported in 2017 US dollars. Table 4.6. Direct Costs by Disease Severity ppFEV1 70% ppFEV1 40%-69% ppFEV1 <40% Disease Management $25,367 $33,462 $57,210 PEx* (age <18) $52,988 $83,956 $124,386 PEx* (age 18+) $48,015 $76,322 $109,372 Lung Transplant $905,191 Post-Transplant (Year 1) $273,665 Post-Transplant (Year 2+) $103,913 *PEx = acute pulmonary exacerbation requiring IV antibiotics Productivity Costs For the societal perspective, we used data provided by CFF regarding employment status as a function of age and lung function. The data provided showed that employment rates for patients with ppFEV1 40% were similar to the general population. However, employment rates were lower for patients with ppFEV1 <40%. We estimated a 50% increase in the loss of productivity for patients with ppFEV1 <40% and assumed an average weekly wage of $857 (Bureau of Labor Statistics) plus a fringe rate. Thus, we assumed that changes in lung function increase the chance that a person is employed. We also added productivity losses to the cost of acute pulmonary exacerbations. Because there is no evidence on the impact the CFTR modulator therapies have on employment and education status, we were only able to model these effects through ppFEV1. ©Institute for Clinical and Economic Review, 2018 Page 65 Final Evidence Report – Cystic Fibrosis Return to Table of Contents A large impact on caregiver costs from CFTR modulator treatment would require that caregiver burden be associated with lung function (e.g., the primary characteristic which modulator treatments change) or have direct evidence that the CFTR modulators reduce caregiver burden. However, Neri et al. found no relationship between caregiver burden, as measured by the General Strain Index, and patient factors such as ppFEV1 or occurrence of acute pulmonary exacerbations.105 Angelis et al. did find that direct non-health care costs were of the same magnitude as direct health care costs (in the United Kingdom) but did not report societal costs by lung function category.106 Therefore, we did not include impacts on caregiver costs in this analysis, given the lack of evidence that it varies by lung function or is impacted by CFTR modulators. The addition of direct non-health care costs that are not affected by CFTR modulator treatments would result in an increase in total societal costs due to the substantial increase in life expectancy with modulator therapy. Sensitivity Analyses We ran one-way sensitivity analyses to identify the key drivers of model outcomes, using available measures of parameter uncertainty (i.e., standard errors) or reasonable ranges for each input, as described in the model inputs section above. Probabilistic sensitivity analyses were also performed by jointly varying all model parameters over 1,000 simulations, then calculating 95% credible interval estimates for each model outcome based on the results and reporting the percent of the simulations where the drug was cost-effectiveness for a given willingness to pay (WTP) threshold (varying from $50,000 per QALY to $500,000 per QALY). We use normal distributions for parameters in the mortality model and drug effect parameters, beta distributions for utilities and probabilities, and truncated normal distributions for costs. Additionally, we performed a threshold analysis by systematically altering the price of CFTR modulators to estimate the maximum prices that would correspond to a set of given WTP thresholds. Scenario Analyses We performed four scenario analyses. In the first we present our results that used a societal perspective. In the second we varied our assumption about long-term effectiveness of the CFTR modulator drugs. In our base case we assume that, after two years, individuals on CFTR modulator therapies would experience 50% of the annual ppFEV1 decline that those receiving best supportive care alone would experience. In scenario analyses we assume that the annual decline in lung function with the CFTR modulator drugs varied between 0% long-term decline (i.e., no long-term lung function decline experienced with drug) to 100% (i.e., long-term decline with drug is the same as best supportive care after two years). This range was supported by the simulated standard error of the long-term percent decline (99% credible interval 1%-99%). In a third scenario analysis we incorporated an additional decrease in ppFEV1 that is not recovered when individuals experience a pulmonary exacerbation. This effect is supported by a study,107 although the magnitude of this effect is unclear, and it is uncertain the degree to which this effect is already captured in the other benefits of CFTR drugs (e.g., decrease in long-term decline in lung function). In a scenario analysis, ©Institute for Clinical and Economic Review, 2018 Page 66 Final Evidence Report – Cystic Fibrosis Return to Table of Contents we varied the additional absolute decline in ppFEV1 due to a pulmonary exacerbation between 0% (i.e., no additional decline in ppFEV1 due to pulmonary exacerbation) to 5% (i.e., a 5% absolute decline in ppFEV1 for each pulmonary exacerbation experienced). In a final scenario analysis, we explored the assumption that CFTR modulator therapies have a quality-of-life benefit in addition to respiratory improvements. An analysis of STRIVE CFQ-R findings reported scores for domains other than the respiratory domain and found clinically significant improvements in certain domains (e.g., physical functioning, health perception, vitality, weight). 70 Although a CFQ-R score does not directly translate into a utility, we varied an independent utility effect (i.e., using a multiplier to the lung-function-informed utility) due to CFTR therapy from 1 (no independent effect) to 1.05 (a 5% increase in utility with drug), above that due to lung function improvement. Model Validation We used several approaches to validate the model. First, we provided preliminary methods and results to manufacturers, patient groups, and clinical experts. Based on feedback from these groups, we refined data inputs used in the model. Second, we varied model input parameters to evaluate face validity of changes in results. Simulated individuals were compared to observed statistics of CF patients: median age of survival, percent in lung function categories (≤40% ppFEV1, severe; between 40% and 70% ppFEV1, moderate; ≥70% pp FEV1, mild) by age, and median ppFEV1 by age.1 We also performed model verification for model calculations using internal reviewers. Finally, we compared results to other cost-effectiveness models in this therapy area. Cost-Effectiveness Model: Results Base Case Results The base case results are shown in Tables 4.7 and 4.8. All CFTR modulators are compared to best supportive care. We did not compare the drugs with each other for CF populations with two CFTR modulator alternatives because of the lack of substantive differences between them in the meta- analysis results and in the modeling results. For individuals with a gating mutation, the total discounted lifetime costs for Kalydeco plus best supportive care and best supportive care only were approximately $8,666,300 and $2,227,800, respectively. The total discounted QALYs (and life years) for Kalydeco plus best supportive care and best supportive care alone were 22.65 (26.52) and 15.92 (22.16), respectively. The incremental cost-effectiveness ratios for Kalydeco in this population were approximately $956,800 per QALY gained and $1,476,500 per life year gained. For individuals who are homozygous for the F508del mutation the total discounted lifetime costs for Orkambi, Symdeko and best supportive care were approximately $6,983,300, $7,478,700 and $2,108,200, respectively. The total discounted QALYs (and life years) for Orkambi, Symdeko and best supportive care were 20.21 (24.57), 20.25 (24.70) and 14.74 (20.77), respectively. The ©Institute for Clinical and Economic Review, 2018 Page 67 Final Evidence Report – Cystic Fibrosis Return to Table of Contents incremental cost-effectiveness ratios for Orkambi and Symdeko versus best supportive care in this population were approximately $890,700 per QALY and $974,300 per QALY, respectively, and approximately $1,280,900 and $1,367,400 per life year gained, respectively. For individuals who are heterozygous for the F508del mutation with a residual function mutation, the total discounted lifetime costs for Kalydeco, Symdeko and best supportive care were approximately $7,557,600, $7,091,900 and $2,081,200, respectively. The total discounted QALYs (and life years) for Kalydeco, Symdeko and best supportive care were 18.74 (23.07), 18.88 (23.25) and 12.92 (18.98), respectively. The incremental cost-effectiveness ratios for Kalydeco and Symdeko in this population were approximately $941,100 per QALY and $840,600 per QALY, respectively, and approximately $1,340,200 and $1,174,500 per life year gained, respectively. Table 4.7. Results for the Base Case for CFTR Modulators Plus Best Supportive Care (BSC) Compared to BSC Alone, By Study Population (Discounted at 3% per Year) Average Population and Treatment CFTR Drug Cost Total Cost Total Life Years Total QALYs Number of PEx CF Individuals with A Gating Mutation BSC $0 $2,227,765 32.75 22.16 15.92 Kalydeco Plus BSC $7,443,121 $8,666,308 18.86 26.52 22.65 CF Individuals Homozygous for F508del Mutation BSC $0 $2,108,199 26.02 20.77 14.74 Orkambi Plus BSC $5,847,893 $6,983,336 11.45 24.57 20.21 Symdeko Plus BSC $6,290,005 $7,478,684 13.36 24.70 20.25 CF Individuals Heterozygous for F508del Mutation with Residual Function Mutation BSC $0 $2,081,180 25.51 18.98 12.92 Kalydeco Plus BSC $6,447,156 $7,557,596 10.85 23.07 18.74 Symdeko Plus BSC $5,934,935 $7,091,919 12.68 23.25 18.88 CFTR: cystic fibrosis-related diabetes; PEx: pulmonary exacerbations; QALYS: quality adjusted life years; BSC: best supportive care Table 4.8. Incremental Cost-Effectiveness Ratios Compared to Best Supportive Care (BSC) for the Base Case Treatment vs. BSC Cost Per LY Gained Cost Per QALY Gained Cost Per PEx Averted CF Individuals with a Gating Mutation Kalydeco Plus BSC $1,476,543 $956,762 $463,571 CF Individuals Homozygous for F508del Mutation Orkambi Plus BSC $1,280,892 $890,739 $334,495 Symdeko Plus BSC $1,367,400 $974,348 $424,212 CF Individuals Heterozygous for F508del Mutation and Residual Function Mutation Kalydeco Plus BSC $1,340,171 $941,110 $373,541 Symdeko Plus BSC $1,174,508 $840,568 $390,600 SBSC: best supportive care; LY: life year; QALY: quality adjusted life years; PEx: pulmonary exacerbation ©Institute for Clinical and Economic Review, 2018 Page 68 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Sensitivity Analysis Results To demonstrate effects of uncertainty on both costs and health outcomes, we varied input parameters using available measures of parameter uncertainty (i.e. standard errors) or reasonable ranges to evaluate changes in cost per additional QALY for CFTR modulators plus best supportive care versus best supportive care alone. Because utilities depending on the ppFEV1 value were a linear equation, we varied the slope of the line (base case, 0.003476). Drug cost variation is described more completely as part of threshold analyses (see below). The impacts of variations in input values on cost-per-QALY estimates are shown for Symdeko in CF individuals homozygous for F508del mutation in Figure 4.2, and in individuals heterozygous for F508del mutation and residual function mutation in Figure 4.3. The analyses were most sensitive to assumptions about the independent effect of drugs on the reduction of acute pulmonary exacerbations, the discount rate, and lung function-specific utilities; while changes in the former resulted in large variation in cost-effectiveness estimates, these did not approach commonly cited thresholds. Also, while not shown in the Figure, we recognize that the difference in resource intensity and costs by level of lung function might have changed over time (our source for this differential was published in 1996), and so varied the differential in background costs across ppFEV1 categories by multiplying costs by a factor of 0.5-1.5 times the base case value (with the higher value resulting in larger absolute cost differences across the three categories), and again found that the cost per QALY estimates did not approach commonly used thresholds. Results were similar for the other drugs in each population, with results shown in Figures E1-E3 in Appendix E. ©Institute for Clinical and Economic Review, 2018 Page 69 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Figure 4.2. Tornado Diagram for One-Way Sensitivity Analyses of Cost per QALY Gained for Symdeko Plus Best Supportive Care Versus Best Supportive Care Alone in CF Individuals Homozygous for F508del Mutation Independent PEx Reduction [0.5-1.0] Discount Rate [1%-5%] Slope of Utility Function [0.002-0.005] Avg. Annual BSC Costs [$49,979-$149,937] Avg. PEx Cost [$41,253-$123,759] Absolute ppFEV1 Gain [3.1%-4.8%] Parameter Input High Parameter Input Low Avg. Transplant Costs [$589,428-$1,768,284] Avg. Annual DMCosts [$19,340-$58,020] $800,000 $1,050,000 $1,300,000 PEx: acute pulmonary exacerbation; BSC: best supportive care; DM: disease management; Probability of transplant among individuals with ppFEV1<30%. ©Institute for Clinical and Economic Review, 2018 Page 70 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Figure 4.3. Tornado Diagram for One-Way Sensitivity Analyses of Cost per QALY Gained for Symdeko Plus Best Supportive Care Versus Best Supportive Care Alone in CF Individuals Heterozygous for F508del Mutation and Residual Function Mutation Independent PEx Reduction [0.5-1.0] Slope of Utility Function [0.002-0.005] Discount Rate [1%-5%] Avg. PEx Cost [$41,253-$123,759] Avg. Annual BSC Costs [$54,274-$162,823] Absolute ppFEV1 Gain [5.7%-7.8%] Parameter Input High Parameter Input Low Avg. Transplant Costs [$589,428-$1,768,284] Avg. Annual DM Costs [$19,340-$58,020] $700,000 $950,000 $1,200,000 PEx: acute pulmonary exacerbation; BSC: best supportive care, DM = disease management, Probability of transplant among individuals with ppFEV1<30%. We also evaluated the uncertainty in the model parameters simultaneously by conducting a probabilistic sensitivity analysis (Table 4.9). For all CFTR modulators in all CF populations evaluated, the number of iterations in which the CFTR modulators were cost-effective at a WTP threshold of $500,000 per QALY or less was approximately 0%. For example, the 95% credible interval for the incremental cost-effectiveness ratios for Kalydeco compared with best supportive care was $669,500 to $1,591,500 per QALY for CF individuals with gating mutations. Scatterplots showing the cost and effectiveness results from the probabilistic sensitivity analyses can be found in Figures E4-E6 in Appendix E. ©Institute for Clinical and Economic Review, 2018 Page 71 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Table 4.9. Probabilistic Sensitivity Analysis Results: CFTR Modulators Versus Best Supportive Care Cost- Cost- Cost- Cost- Cost- Cost- Effective Effective Effective Effective Effective Effective CF population and CFTR Modulator at at at at at at $50,000 $100,000 $150,000 $200,000 $300,000 $500,000 per QALY per QALY per QALY per QALY per QALY per QALY CF Individuals with a Gating Mutation Kalydeco plus BSC 0% 0% 0% 0% 0% 0% CF Individuals Homozygous for F508del Mutation Orkambi plus BSC 0% 0% 0% 0% 0% 0% Symdeko plus BSC 0% 0% 0% 0% 0% 0% CF Individuals Heterozygous for F508del Mutation and Residual Function Mutation Kalydeco plus BSC 0% 0% 0% 0% 0% 0% Symdeko plus BSC 0% 0% 0% 0% 0% 0.2% CFTR: cystic fibrosis transmembrane conductance regulator gene; BSC: best supportive care; Scenario Analyses Results Modified Societal Perspective We incorporated the costs associated with lost productivity in individuals with CF (Table 4.10). For individuals with a gating mutation we projected that the difference in lifetime (discounted) indirect costs was $31,600. Including productivity losses in the analysis resulted in incremental cost- effectiveness ratios for Kalydeco very similar to those seen in the base case ($952,100 per QALY societal vs. $956,800 per QALY base case). Estimates for the incremental cost-effectiveness ratios for the CFTR modulators for the other two populations also tracked very closely with base case estimates (Table 4.10). ©Institute for Clinical and Economic Review, 2018 Page 72 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Table 4.10. Incremental Cost-Effectiveness Ratios Compared to Best Supportive Care (BSC) for the Societal Perspective Treatment vs. BSC Incremental Costs (Direct) Incremental Costs (Indirect) Cost Per QALY Gained CF Individuals with a Gating Mutation Kalydeco plus BSC $6,438,543 -$31,635 $952,061 CF Individuals Homozygous for F508del Mutation Orkambi plus BSC $4,875,137 -$30,639 $885,140 Symdeko plus BSC $5,370,485 -$30,891 $968,744 CF Individuals Heterozygous for F508del Mutation and Residual Function Mutation Kalydeco plus BSC $5,476,416 -$26,054 $936,633 Symdeko plus BSC $5,010,739 -$27,306 $835,987 BSC: best supportive care; QALY: quality adjusted life year Long-Term Effectiveness Assumptions In the base case we assumed that CFTR modifiers would result in 50% of the annual declines in ppFEV1 as for best supportive care, after the first two years without any decline. In this scenario analysis we varied that assumption from 0% (i.e., no declines in ppFEV1 over an individual’s lifetime) to 100% (i.e., the same annual declines as those on best supportive care after the first two years on drug) (Table 4.11). For CF individuals with a gating mutation, the incremental cost-effectiveness ratio for Kalydeco was $620,400 per QALY when we assumed that there was no long-term decline in ppFEV1 (i.e., the drug increased ppFEV1 at the start of therapy and individuals’ lung function remained constant for the remainder of their lifetime). Similar declines in ICERs were found with other drugs and populations (Table 4.11). ©Institute for Clinical and Economic Review, 2018 Page 73 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Table 4.11. Incremental Cost-Effectiveness Ratios ($ per QALY) Compared to Best Supportive Care for the Long-Term Effectiveness Assumption Treatment vs. BSC 0% Decline 25% Decline 75% Decline 100% Decline CF Individuals with a Gating Mutation Kalydeco plus BSC $620,428 $751,624 $1,271,535 $1,772,535 CF Individuals Homozygous for F508del Mutation Orkambi plus BSC $566,976 $698,108 $1,191,460 $1,647,556 Symdeko plus BSC $615,966 $761,672 $1,314,815 $1,886,539 CF Individuals Heterozygous for F508del Mutation and Residual Function Mutation Kalydeco plus BSC $651,429 $774,607 $1,152,209 $1,443,267 Symdeko plus BSC $580,459 $688,044 $1,038,188 $1,289,044 BSC: best supportive care ppFEV1 Recovery After Pulmonary Exacerbation Assumptions In the base case we assumed that CF individuals’ ppFEV1 would fully recover to baseline following pulmonary exacerbations, allowing only for the natural decline in lung function and the impact of the CFTR drugs on that natural decline. In this scenario analysis we varied that assumption from 0% (i.e., no additional decline in ppFEV1 due to pulmonary exacerbation) to 5% (i.e., a 5% absolute decline in ppFEV1 for each pulmonary exacerbation experienced) (Table 4.12). For CF individuals with a gating mutation, the incremental cost-effectiveness ratio for Kalydeco was $737,900 per QALY when we assumed that there was a 5% absolute decline in ppFEV1 for each pulmonary exacerbation experienced. Similar declines in ICERs were found with other drugs and populations (Table 4.12). ©Institute for Clinical and Economic Review, 2018 Page 74 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Table 4.12. Incremental Cost-Effectiveness Ratios ($ per QALY) Compared to Best Supportive Care for the Lung Function Recovery After Pulmonary Exacerbation Assumption Treatment vs. BSC 1% Decline 3% Decline 5% Decline CF Individuals with a Gating Mutation Kalydeco plus BSC $826,217 $749,865 $737,931 CF Individuals Homozygous for F508del Mutation Orkambi plus BSC $732,581 $608,234 $569,114 Symdeko plus BSC $827,295 $706,465 $678,570 CF Individuals Heterozygous for F508del Mutation and Residual Function Mutation Kalydeco plus BSC $772,962 $641,731 $606,196 Symdeko plus BSC $700,135 $595,378 $570,023 BSC: best supportive care Independent Utility Effect In the base case we assumed that CF individuals’ utility was based only on lung function (i.e., ppFEV1, pulmonary exacerbations, lung transplantations). In this scenario analysis we varied an independent utility effect (i.e., using a multiplier to the lung-function-informed utility) due to CFTR therapy from 1 (no independent effect) to 1.05 (a 5% increase in utility with drug), above that due to lung function improvement (Table 4.13). For CF individuals with a gating mutation, the incremental cost-effectiveness ratio for Kalydeco was $836,500 per QALY when we assumed that there was a 5% increase in utility due to drug that in independent of lung function improvement. . Similar declines in ICERs were found with other drugs and populations (Table 4.13). ©Institute for Clinical and Economic Review, 2018 Page 75 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Table 4.13. Incremental Cost-Effectiveness Ratios ($ per QALY) Compared to Best Supportive Care for the Non-Respiratory Utility Assumption Treatment vs. BSC 1% Increase 2% Increase 4% Increase 5% Increase CF Individuals with a Gating Mutation Kalydeco plus BSC $927,566 $901,055 $855,659 $836,511 CF Individuals Homozygous for F508del Mutation Orkambi plus BSC $859,468 $830,519 $778,983 $756,152 Symdeko plus BSC $940,146 $908,528 $852,381 $827,580 CF Individuals Heterozygous for F508del Mutation and Residual Function Mutation Kalydeco plus BSC $911,513 $883,952 $833,545 $810,438 Symdeko plus BSC $814,291 $789,831 $745,070 $724,539 BSC: best supportive care Threshold Analysis Results Unit and annual prices necessary to reach cost-effectiveness thresholds of $50,000, $100,000, $150,000, $200,000, $300,000 and $500,000 per QALY are listed in Tables 4.14 and 4.15 respectively, for each CF population and CFTR modulator. Threshold prices were higher for the CF population heterozygous for F508del mutation and residual function mutation, and slightly higher for Orkambi compared with Symdeko for CF individuals homozygous for F508del mutation on an annual cost basis. A discount of approximately 37%-44% would be necessary to reach a cost- effectiveness threshold of $500,000/QALY. Larger discounts would be needed to achieve cost- effectiveness thresholds of $300,000 or less per QALY. ©Institute for Clinical and Economic Review, 2018 Page 76 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Table 4.14. Threshold Analysis Results Presented as Price per Unit Unit Unit Price Unit Price Unit Price Unit Price Unit Price Price to to WAC per Net Price to Achieve to Achieve to Achieve to Achieve Achieve Achieve Unit per Unit $100,000 $150,000 $200,000 $300,000 $500,000 $50,000 per QALY per QALY per QALY per QALY per per QALY QALY CF Individuals with A Gating Mutation Kalydeco $426.72 $424.15 $75.49 $94.65 $113.82 $132.98 $171.32 $247.98 CF Individuals Homozygous for F508del Mutation Orkambi $186.78 $180.76 $38.03 $46.42 $54.80 $63.19 $79.96 $113.50 Symdeko $400.08 $387.20 $72.84 $89.62 $106.39 $123.17 $156.72 $223.82 CF Individuals Heterozygous for F508del Mutation and Residual Function Mutation Kalydeco $426.72 $424.15 $82.54 $101.54 $120.54 $139.54 $177.54 $253.54 Symdeko $400.08 $387.20 $79.29 $98.52 $117.75 $136.99 $175.45 $252.37 WAC: wholesale acquisition cost; QALY: quality adjusted life year gained Table 4.15. Threshold Analysis Results Presented as Annual Prices Price to Price to Price to Price to Price to Price to Achieve Achieve Achieve Annual Annual Achieve Achieve Achieve $50,000 $100,000 $150,000 WAC Net Price $200,000 $300,000 $500,000 per per per per QALY per QALY per QALY QALY QALY QALY CF Individuals with A Gating Mutation Kalydeco $311,719 $309,842 $55,145 $69,142 $83,146 $97,142 $125,149 $181,149 CF Individuals Homozygous for F508del Mutation Orkambi $272,886 $264,090 $55,562 $67,820 $80,063 $92,321 $116,822 $165,824 Symdeko $292,258 $282,850 $53,210 $65,467 $77,718 $89,976 $114,484 $163,501 CF Individuals Heterozygous for F508del Mutation and Residual Function Mutation Kalydeco $311,719 $309,842 $60,295 $74,175 $88,054 $101,934 $129,693 $185,211 Symdeko $292,258 $282,850 $57,921 $71,969 $86,016 $100,071 $128,166 $184,356 WAC: wholesale acquisition cost; QALY: quality adjusted life year gained Note that Kalydeco and Symdeko are each used for treatment in two different populations. Therefore, we also calculated population-weighted threshold prices using estimated numbers of patients in each population. (We assumed approximately 3,000 CF individuals with gating mutations, 8,464 CF individuals homozygous for F508del mutation, and 6,195 CF individuals heterozygous for F508del mutation and residual function mutation.) The blended unit price for Kalydeco across both relevant populations varied from $80.24 at $50,000 per QALY, $99.29 at $100,000 per QALY, $118.35 at $150,000 per QALY and $251.73 at $500,000 per QALY. The blended annual prices across the two relevant populations at the $50,000, $100,000 and $150,000 per QALY threshold prices were approximately $58,600, $72,500 and $86,500, respectively, and at the ©Institute for Clinical and Economic Review, 2018 Page 77 Final Evidence Report – Cystic Fibrosis Return to Table of Contents $500,000 per QALY threshold price was approximately $183,900. Blended unit prices for Symdeko across both of its relevant populations were $75.57 at $50,000 per QALY, $93.38 at $100,000 per QALY, $111.19 at $150,000 per QALY, and $235.89 at $500,000 per QALY. The blended annual prices across the two relevant populations at the $50,000, $100,000 and $150,000 per QALY threshold prices were approximately $55,200, $68,200 and $81,200, respectively, and at the $500,000 per QALY threshold price was approximately $172,300. Prior Published Evidence on Costs and Cost-Effectiveness Model validation was also conducted in terms of comparisons to other model findings. We searched the literature to identify models that were similar to our analysis, with comparable populations, settings, perspective, and treatments. We identified two prior published, US-based cost-effectiveness analyses of CFTR modulator drugs, both from the same group. Dilokthornsakul and colleagues have modeled the long-term costs and outcomes of Kalydeco treatment of CF patients with the G551D mutation (2016)97 and, more recently, Orkambi treatment of CF patients with homozygous F508del mutation (2017).96 They developed a Markov model with a lifetime horizon and US payer perspective, comparing each treatment to usual care. Our model in the current analysis was informed by these prior models, and therefore shares some similarities, including time horizon, perspective, and the base-case assumption of 50% decline in efficacy two years after treatment initiation. The prior models included health states for three categories defined by lung function (mild: ppFEV1 ≥ 70%, moderate: 40% ≤ ppFEV1 < 70%, and severe: ppFEV1 < 40%), while the ICER analysis models ppFEV1 as a continuous value. Although base case outcomes in the 2016 analysis97 were undiscounted, results were also presented using a discount rate of 3%. Discounted incremental QALYs were 5.21, incremental lifetime costs approximately $3,527,000, and the base-case incremental cost–effectiveness ratio was approximately $680,000 per QALY (2013 US$ converted to 2017 using the personal consumption expenditure [PCE] price index). Our current model estimated incremental QALYs of 6.73, incremental costs of $6,438,543, and an incremental cost-effectiveness ratio of approximately $956,800 per QALY. Starting age for treatment in the earlier Kalydeco model was 25 years old, while we modeled treatment initiation at two years old. Kalydeco WAC was $426.72 per tablet, which was only slightly higher than the net price used in our analysis ($424.15), but Dilokthornsakul et al. assumed that the drug price would drop to 10% of that amount after patent expiration in 2027. This assumption, along with the later age of treatment initiation, may have led to the lower lifetime costs observed in the analysis by Dilokthornsakul and colleagues. The same model was later adapted by Dilokthornsakul and colleagues to examine the lifetime costs and outcomes of Orkambi combination treatment of CF patients with homozygous F508del mutation.96 Starting age for treatment with Orkambi was 25 years old, while the ICER analysis ©Institute for Clinical and Economic Review, 2018 Page 78 Final Evidence Report – Cystic Fibrosis Return to Table of Contents modeled treatment initiation at six years old. The WAC for Orkambi was $117.88 per tablet, which was lower than the net price used in our analysis ($180.76). Dilokthornsakul et al. again assumed that the drug price would drop to 10% of WAC after patent expiration. Their analysis estimated a gain of 2.42 QALYs with an incremental lifetime cost of approximately $2,698,000, or approximately $1,115,000 per QALY (all discounted; costs converted to 2017 dollars). Our current model for Orkambi estimated incremental QALYs of 5.47, incremental lifetime costs of $4,875,137, and an incremental cost-effectiveness ratio of $890,739 per QALY. Again, the later age of treatment initiation and the assumption of a lower future price may have led to the lower lifetime costs calculated in this analysis than those from our current model. Prior to these analyses, Whiting and colleagues had modeled the cost-effectiveness of Kalydeco treatment of CF patients aged six years or older (with median age = 20 years) with G551D mutation in the United Kingdom.82 They modified a deterministic simulation model developed by Vertex Pharmaceuticals, adding in lung transplantations. This analysis was conducted from the UK National Health Service perspective, with a lifetime horizon and 3.5% discount rate for costs and outcomes. For long-term effects of Kalydeco treatment on ppFEV1 decline, they modeled three different scenarios: conservative, with same rate of decline as for standard care; intermediate, with 66% rate of decline; and optimistic, with stable ppFEV1 over lifetime. The cost of Kalydeco used in the model was £182,000 (approximately $306,000 in 2017 US$), with the assumption that it would decline to £20,000 in 14 years, due to loss of patent exclusivity. They used UK-based utility values and costs for usual care, making these results less comparable to our US-based analysis. This model led to estimated QALY gains of 1.27 (in the conservative scenario) to 5.26 (in the optimistic scenario), the latter being closest to our current model estimate of 6.73 incremental QALYs. The incremental cost-effectiveness ratio was estimated to vary between £335,000 and £1,274,000 per QALY (approximately $563,000 to $2,141,000 in 2017 US$). 4.4 Summary and Comment We developed an individual-level microsimulation model to project the lifetime benefits and costs of CFTR modulator therapies for three different CF cohorts. The drugs increased lung function, increased weight-for-age z-scores, and decreased the number of acute pulmonary exacerbations and lung transplantations over the lifetime of individuals. The drugs did not impact non-lung aspects of the disease, nor did they decrease the need for CF-related supportive care. Overall, all drugs (plus best supportive care) evaluated were very effective compared with best supportive care alone in all populations studied, with quality-adjusted life year gains ranging from 5.47 to 6.73 (discounted). With (discounted) CFTR drug-related costs ranging from $4.9 million to $7.4 million, the incremental cost-effectiveness ratios of drugs plus best supportive care compared with best supportive care alone were approximately $0.9 million per QALY for all drugs in all populations considered. Our results were robust to variations to parameter estimates, adopting a societal perspective, or using life years gained as the health outcome, except for unit drug costs. ©Institute for Clinical and Economic Review, 2018 Page 79 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Limitations There are several limitations to our analysis that deserve mention. We used ppFEV1 as the primary marker of lung function to characterize the progression of CF over time. Trials generally did not include patients with either very low or very high lung function, which may impact the generalizability of our results. Furthermore, based on available evidence, only the effect of the CFTR modulators on lung function, weight and acute pulmonary exacerbations are included in the model. As any surrogate marker of disease, it is not a perfect marker for progression. We did not have direct measures of CFTR modulator benefit on EQ-5D utilities above that associated with ppFEV1. We conducted a scenario analysis to examine the potential impact of this and found that a 5% increase in non-respiratory-related utility would increase the ICER by approximately 13% for all drugs and populations. In addition, limited evidence exists about the drugs’ impact on individual’s ability to work or attend school, or the degree to which caregiver burden is reduced by CFTR modulator drugs. Such information would better inform our analysis from a societal perspective. More importantly, we only had short-term measures of drug effect and had to make assumptions about their effect over the lifetime of the patient. In addition, we used trial-based estimates of discontinuation of these therapies to be consistent with the efficacy estimates; real-world patterns of discontinuation may differ from these. Conclusions We found that CFTR modulator therapies plus best supportive care substantially improve patient health outcomes compared to best supportive care. Because of the high cost of these drugs, however, the cost of CFTR modulator therapies exceed commonly used cost-effectiveness thresholds. For ultra-rare diseases, decision-makers often give special considerations that lead to coverage and funding decisions at higher willingness-to-pay thresholds. We evaluated thresholds up to $500,000 per QALY and still found that drug prices would need to be reduced by about 40% to be considered cost effective at this threshold. ©Institute for Clinical and Economic Review, 2018 Page 80 Final Evidence Report – Cystic Fibrosis Return to Table of Contents 5. Other Benefits and Contextual Considerations Our reviews seek to provide information on other benefits offered by the intervention to the individual patient, caregivers, the delivery system, other patients, or the public that would not have been considered as part of the evidence on comparative clinical effectiveness. These general elements are listed in the table below, and the subsequent text provides detail about the elements that are applicable to the comparison of adding versus not adding CFTR modulators to standard care for CF patients. Table 5.1. Potential Other Benefits or Contextual Considerations (Not Specific to Any Disease or Therapy) Potential Other Benefits This intervention offers reduced complexity that will significantly improve patient outcomes. This intervention will reduce important health disparities across racial, ethnic, gender, socio-economic, or regional categories. This intervention will significantly reduce caregiver or broader family burden. This intervention offers a novel mechanism of action or approach that will allow successful treatment of many patients for whom other available treatments have failed. This intervention will have a significant impact on improving the patient’s ability to return to work or school and/or their overall productivity. This intervention will have a significant positive impact outside the family, including on schools and/or communities. This intervention will have a significant impact on the entire “infrastructure” of care, including effects on screening for affected patients, on the sensitization of clinicians, and on the dissemination of understanding about the condition, that may revolutionize how patients are cared for in many ways that extend beyond the treatment itself. Other important benefits or disadvantages that should have an important role in judgments of the value of this intervention. Potential Other Contextual Considerations This intervention is intended for the care of individuals with a condition of particularly high severity in terms of impact on length of life and/or quality of life. This intervention is intended for the care of individuals with a condition that represents a particularly high lifetime burden of illness. This intervention is the first to offer any improvement for patients with this condition. Compared to best supportive treatment, there is significant uncertainty about the long-term risk of serious side effects of this intervention. Compared to best supportive treatment, there is significant uncertainty about the magnitude or durability of the long-term benefits of this intervention. There are additional contextual considerations that should have an important role in judgments of the value of this intervention. ©Institute for Clinical and Economic Review, 2018 Page 81 Final Evidence Report – Cystic Fibrosis Return to Table of Contents 5.1 Other Benefits CF represents a major and lifelong burden to patients and their caregivers. As described in Section 1.5, important aspects of the lived experience of CF patients and their informal caregivers are not captured by quality of life instruments or by the typically used outcomes in trials and registries. It is possible that there are improvements in the quality of life with CFTR modulator treatment that may not be fully reflected in our model estimate. However, we also heard from individual patients and their caregivers that use of CFTR modulators is typically additive to their daily burden of disease management, thereby increasing (rather than reducing) the complexity of managing the disease. The time costs associated with CF and its complications are very large and extend over a lifetime. While the time costs of patients are, theoretically, accounted for when estimating QALYs, the time costs of their informal caregivers are very difficult to estimate. 5.2 Contextual Considerations The major contextual consideration pertains to the fact that the evidence is sparse, especially for the long-term effects of CFTR modulators on the rate of progression of the disease. Our modeling analyses suggest that reductions in the rate of CF progression with these medications may improve both unadjusted and quality-adjusted life expectancy relative to supportive care alone. The magnitude and sustainability of such effects have yet to be reliably quantified. Currently, the CFTR modulators are the only available intervention that targets the basic pathophysiology of the disease. Novel treatments, e.g., a triple combination of VX-445 and VX-659 (novel CFTR corrector) with tezacaftor and ivacaftor, and treatment advances that are likely to be realized in the next decade may be associated with better outcomes and may eventually substantially change the typical course of the disease. With the uptake of systematic newborn screening in the last several years, an increasing number of CF patients are diagnosed early, before the onset of symptoms or the establishment of irreversible lung, pancreatic, liver, and other complications. Early and aggressive management of CF, with or without CFTR modulator therapy, is expected to change the course of the disease in these patients. While CFTR modulator therapies may play a role in improving health, overall improvements in the management of care of the disease have substantially improved the prognosis for the CF population, possibly to the detriment of new therapies trying to prove a significant clinical response. However, even with these gains in longevity and quality of life over the last few decades, the United States still lags other comparable countries in terms of health benefits in the CF population. ©Institute for Clinical and Economic Review, 2018 Page 82 Final Evidence Report – Cystic Fibrosis Return to Table of Contents 6. Value-Based Price Benchmarks Our value-based benchmark prices for Kalydeco, Orkambi, and Symdeko are presented in Table 6.1. As Kalydeco and Symdeko are each used for treatment in two different populations, we calculated blended threshold prices weighted by estimated numbers of patients in each population. For each drug, the discounts required to meet both threshold prices (>70%) are much greater than the currently assumed discount from WAC. Table 6.1. Value-Based Benchmark Prices for Kalydeco, Orkambi, and Symdeko Annual Price Annual Price Annual Net Discount from Annual to Achieve to Achieve Price (with WAC to Reach WAC $100,000 $150,000 Mark-Up) Threshold Prices per QALY per QALY Kalydeco $311,719 $309,842 $72,533 $86,453 72% to 77% Orkambi $272,886 $264,090 $67,820 $80,063 71% to 75 % Symdeko $292,258 $282,850 $68,215 $81,225 72% to 77% QALY: quality-adjusted life year ©Institute for Clinical and Economic Review, 2018 Page 83 Final Evidence Report – Cystic Fibrosis Return to Table of Contents 7. Potential Budget Impact 7.1 Overview We used results from the same model employed for the cost-effectiveness analyses to estimate the total potential budgetary impact of Symdeko in cystic fibrosis, specifically for those heterozygous or homozygous for the F508del mutation. We used the WAC for Symdeko, an estimate of discounted WAC, and the cost-effectiveness threshold prices at $50,000, $100,000, and $150,000 per QALY in our estimates of budget impact. We did not include the other therapies modeled above in this potential budget impact analysis, given their established presence on the market. 7.2 Methods Potential budget impact was defined as the total differential cost of using Symdeko plus best supportive care, rather than relevant existing therapy for the treated population, calculated as differential health care costs (including drug costs) minus any offsets in these costs from averted health care events. All costs were undiscounted and estimated over a five-year time horizon, given the potential for cost offsets to accrue over time and to allow a more realistic impact on the number of patients treated with the new therapy. The potential budget impact analysis included the candidate populations eligible for treatment: those patients with cystic fibrosis who may be eligible for Symdeko. To estimate the size of the potential candidate populations for treatment, we used inputs from the Cystic Fibrosis Foundation Patient Registry Annual Data Report (2016), which includes prevalence and treatment estimates from the Cystic Fibrosis Foundation Patient Registry.1 In this analysis, we assumed that all CF patients homozygous for the F508del mutation over the age of six would be eligible for Symdeko. We also assumed that all patients over the age of 12 and heterozygous for an F508del mutation with an allowed residual function mutation were eligible for Symdeko. Note that while the approved FDA label for Symdeko allows treatment beyond those having one F508del mutation with a second mutation amenable to Symdeko, we did not include such patients because of the lack of published data on the number of individuals with less frequently occurring mutations, making it infeasible to calculate a reliable number of additional patients likely to be treated. To calculate the number in the first population, we used the estimate of F508del mutation prevalence (24,901) multiplied by the percent who are homozygous (41%) as described by the CFFPR Annual Data Report (2016).1 We then estimated the proportion of patients over the age of six years in the overall cystic fibrosis population (82.9%). Applying these proportions to the prevalent population, our budget impact model assumes 8,464 cystic fibrosis patients with two copies of the F508del mutation in the United States will be eligible for Symdeko. We assumed that 20% of these patients (1,693) would initiate Symdeko in each of the five years. ©Institute for Clinical and Economic Review, 2018 Page 84 Final Evidence Report – Cystic Fibrosis Return to Table of Contents To calculate the population with heterozygous F508del mutation, we used the same estimate of F508del mutation prevalence (24,901) multiplied by the percent who are heterozygous (45.8%) as described by the Cystic Fibrosis Foundation Patient Registry Annual Data Report (2016).1 We then multiplied by the proportion of patients over the age of 12 (66.9%) and subtracted the number of G551D and R117H patients (2,145) as defined in the 2016 CFF Patient Registry Annual Data Report (because these two mutations are not included on the Symdeko label).1 Administration, 2018, 113} In total, our budget impact model assumes 6,195 cystic fibrosis patients with one copy of the F508del mutation will be eligible for Symdeko in the United States. This number may be understated because the approved FDA label for Symdeko allows treatment beyond those having one F508del mutation, so long as the mutation is responsive to Symdeko (through in vitro or clinical data).6 We assumed that 20% of the patients (1,239) would initiate Symdeko in each of the five years. ICER’s methods for estimating potential budget impact are described in detail here and have recently been updated. The intent of our revised approach to budgetary impact is to document the percentage of patients that could be treated at selected prices without crossing a budget impact threshold that is aligned with overall growth in the US economy. Briefly, we evaluate a new drug that would take market share from one or more drugs and calculate the blended budget impact associated with displacing use of existing therapies with the new intervention. For this analysis, in the population homozygous for the F508del mutation, we assumed that Symdeko (plus best supportive care) would replace Orkambi in 50% of eligible patients and would be added to best supportive care in 50% of the eligible patients being treated. According to the CFFPR Annual Data Report (2016), prescribing rates for Orkambi are 52.5% across all eligible patients.1 For the population heterozygous for an F508del mutation with an allowed residual function mutation, we assumed that Symdeko (plus best supportive care) would replace Kalydeco in 50% of eligible patients and would be added to best supportive care in 50% of the eligible patients being treated. In the absence of data on treatment mix in this specific population, we based our assumption on the prescribing rate of Kalydeco in the R117H mutation population as a surrogate (approximately 50% of eligible patients).1 Using this approach to estimate potential budget impact, we then compared our estimates to an updated budget impact threshold that represents a potential trigger for policy mechanisms to improve affordability, such as changes to pricing, payment, or patient eligibility. As described in ICER’s methods presentation (http://icer-review.org/wp-content/uploads/2018/03/ICER-value- assessment-framework-update-FINAL-062217.pdf), this threshold is based on an underlying assumption that health care costs should not grow much faster than growth in the overall national economy. From this foundational assumption, our potential budget impact threshold is derived using an estimate of growth in US gross domestic product (GDP) +1%, the average number of new drug approvals by the FDA over the most recent two-year period, and the contribution of spending ©Institute for Clinical and Economic Review, 2018 Page 85 Final Evidence Report – Cystic Fibrosis Return to Table of Contents on retail and facility-based drugs to total health care spending. Calculations are performed as shown in Table 7.1. For 2017-18, therefore, the five-year annualized potential budget impact threshold that should trigger policy actions to manage access and affordability is calculated to total approximately $915 million per year for new drugs. Table 7.1. Calculation of Potential Budget Impact Threshold Item Parameter Estimate Source 1 Growth in US GDP, 2017 (est.) +1% 3.20% World Bank, 2016 2 Total health care spending, 2016 ($) $2.71 trillion CMS NHE, 2014 CMS National Health Contribution of drug spending to total health care 3 17.7% Expenditures (NHE), 2016; spending (%) Altarum Institute, 2014 Contribution of drug spending to total health care 4 $479 billion Calculation spending ($) (Row 2 x Row 3) Annual threshold for net health care cost growth for ALL 5 $15.3 billion Calculation new drugs (Row 1 x Row 4) Average annual number of new molecular entity 6 33.5 FDA, 2017 approvals, 2015-2016 Annual threshold for average cost growth per individual $457.5 7 new molecular entity Calculation million (Row 5 ÷ Row 6) Annual threshold for estimated potential budget impact $915 million 8 for each individual new molecular entity (doubling of Calculation Row 7) 7.3 Results Table 7.2 illustrates the per-patient budget impact calculations for Symdeko in those homozygous for the F508del mutation, compared to current care assuming Orkambi plus best supportive care in 50% and only best supportive care in 50%. Potential budget impact is presented based on WAC ($292,258 per year), discounted WAC ($282,850 per year), and the prices to reach $150,000, $100,000, and $50,000 per QALY in this population ($75,166, $63,315, and $51,463 per year, respectively). ©Institute for Clinical and Economic Review, 2018 Page 86 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Table 7.2. Per-Patient Budget Impact Calculations Over a Five-year Time Horizon for Individuals Homozygous for F508del Mutation Average Annual Per Patient Budget Impact WAC Discounted $150,000/ $100,000/ $50,000/ WAC QALY QALY QALY Symdeko+BSC $300,749 $292,545 $113,699 $98,765 $92,331 Orkambi+BSC (50%) & $183,418 BSC (50%) Difference $117,331 $109,128 ($69,719)* ($84,653)* ($91,078)* WAC: wholesale acquisition cost, QALY: quality adjusted life year, BSC: best supportive care *Indicates cost-saving The average potential budgetary impact when using the WAC ($292,258) was an additional per- patient cost of approximately $117,300 and approximately $109,100 using the discounted WAC ($282,850). At the three cost-effectiveness threshold prices (at $50,000, $100,000 and $150,000 per QALY), there would be estimated cost savings, ranging from approximately $69,700 per patient using the annual price ($75,166) to achieve $150,000 per QALY to approximately $91,000 using the annual price ($51,463) to achieve a $50,000 per QALY cost-effectiveness threshold. Note that we estimate overall savings because while there would be increased costs from using Symdeko in addition to best supportive care, these additional costs would be more than offset by the replacement of Orkambi at net price by Symdeko at the much lower assumed threshold prices. Table 7.3 illustrates the per-patient budget impact calculations for those with one F508del mutation and a residual function mutation, compared to current care assuming Kalydeco plus best supportive care in 50% and best supportive care in 50% of patients. We present the potential budget impact results based on WAC ($292,258 per year), discounted WAC ($282,850 per year), and the prices to reach $150,000, $100,000, and $50,000 per QALY for Symdeko in this population ($85,960, $71,922, and $57,883 per year, respectively). Table 7.3. Per-Patient Budget Impact Calculations Over a Five-year Time Horizon for Individuals with F508del Mutation and Residual Function Mutation Average Annual Per Patient Budget Impact WAC Discounted $150,000/ $100,000/ $50,000/ WAC QALY QALY QALY Symdeko+BSC $301,966 $293,776 $122,441 $110,212 $97,983 Kalydeco +BSC $209,185 (50%) & BSC (50%) Difference $92,781 $84,591 ($86,744)* ($98,973)* ($111,202)* WAC: wholesale acquisition cost, QALY: quality-adjusted life year, BSC: best supportive care *Indicates cost-saving ©Institute for Clinical and Economic Review, 2018 Page 87 Final Evidence Report – Cystic Fibrosis Return to Table of Contents The average potential budgetary impact when using the WAC ($292,258) was an additional per- patient cost of approximately $92,800 and approximately $84,600 using the discounted WAC ($282,850). Importantly, at the three cost-effectiveness threshold prices (at $50,000, $100,000 and $150,000 per QALY), there would be estimated cost savings, ranging from approximately $86,700 per patient using the annual price ($85,960) to achieve $150,000 per QALY to approximately $111,200 using the annual price ($57,883) to achieve a $50,000 per QALY cost-effectiveness threshold. Again, it should be noted that these overall savings would result from the mix of increased costs from using Symdeko in addition to best supportive care as well as the potential savings from replacement of Kalydeco at net price by Symdeko at the much lower assumed cost- effectiveness threshold prices. For the combined populations of interest, the annual potential budgetary impact of treating the entire eligible population with Symdeko over five years did not exceed the $915 million ICER budget impact threshold at discounted WAC and the three cost-effectiveness threshold prices for $50,000, $100,000, and $150,000 per QALY, but barely exceeded the threshold (by 2%) at WAC. The annual potential budgetary impacts of treating the entire eligible populations using net prices (discounted WAC) are compared to the $915 million threshold in Table 7.4. The potential annual budget impact we estimated for Symdeko in the combined populations is 95% of the $915 million annual budget impact threshold at the net price. While the total number of patients eligible for treatment with Symdeko is relatively low (n = 14,659), the increased cost per patient from using Symdeko over current treatment mix leads to a total estimate approaching the budget impact threshold. Table 7.4. Estimated Total Potential Budget Impact of Symdeko for Treatment of Eligible Populations Using Net Prices Over a Five-year Time Horizon Eligible N Treated per Annual BI per Total BI Percent of Population Year Patient (millions) Threshold Homozygous F508del Symdeko 8,464 1,693 $109,128 $552,527,040 60% Heterozygous F508del with Residual Function Mutation Symdeko 6,195 1,239 $84,591 $312,510,796 34% Total Eligible US CF Population* Symdeko 14,659 2,932 $172,274 $865,037,837 95% BI: budget impact * Annual BI per patient for total eligible US CF population weighted by percentage contribution. ©Institute for Clinical and Economic Review, 2018 Page 88 Final Evidence Report – Cystic Fibrosis Return to Table of Contents 8. Summary of the Votes and Considerations for Policy 8.1 About the Midwest CEPAC Process During Midwest CEPAC public meetings, the Midwest CEPAC Panel deliberates and votes on key questions related to the systematic review of the clinical evidence, an economic analysis of the applications of treatments under examination, and the supplementary information presented. Panel members are not pre-selected based on the topic being addressed and are intentionally selected to represent a range of expertise and diverse perspectives. Acknowledging that any judgment of evidence is strengthened by real-life clinical and patient perspectives, subject matter experts are recruited for each meeting topic and provide input to Midwest CEPAC Panel members before the meeting to help clarify their understanding of the different interventions being analyzed in the evidence review. The same clinical experts serve as a resource to the Midwest CEPAC Panel during their deliberation, and help to shape recommendations on ways the evidence can apply to policy and practice. After the Midwest CEPAC Panel votes, a policy roundtable discussion is held with the Midwest CEPAC Panel, clinical experts, patient advocates, payers, and when feasible, manufacturers. The goal of this discussion is to bring stakeholders together to apply the evidence to guide patient education, clinical practice, and coverage and public policies. Participants on policy roundtables are selected for their expertise on the specific meeting topic, are different for each meeting, and do not vote on any questions. At the May 17, 2018 meeting, the Midwest CEPAC Panel discussed issues regarding the application of the available evidence to help patients, clinicians, and payers address important questions related to the use of CFTR modulator treatments for cystic fibrosis. Following the evidence presentation and public comments (public comments from the meeting can be accessed here, starting at minute 1:50:20), the Midwest CEPAC Panel voted on key questions concerning the comparative clinical effectiveness, comparative value, and other benefits and contextual considerations related to CFTR modulators. These questions are developed by the ICER research team for each assessment to ensure that the questions are framed to address the issues that are most important in applying the evidence to support clinical practice, medical policy decisions, and patient decision-making. The voting results are presented below, along with specific considerations mentioned by Midwest CEPAC Panel members during the voting process. ©Institute for Clinical and Economic Review, 2018 Page 89 Final Evidence Report – Cystic Fibrosis Return to Table of Contents In its deliberations and votes related to value, the Midwest CEPAC Panel considered the individual patient benefits, and incremental costs to achieve such benefits, from a given intervention over the long term. There are four elements to consider when deliberating on long-term value for money (see Figure X below): 1. Comparative clinical effectiveness is a judgment of the overall difference in clinical outcomes between two interventions (or between an intervention and placebo), tempered by the level of certainty possible given the strengths and weaknesses of the body of evidence. Midwest CEPAC uses the ICER Evidence Rating Matrix as its conceptual framework for considering comparative clinical effectiveness. 2. Estimated incremental cost-effectiveness is the average incremental cost per patient of one intervention compared to another to achieve a desired “health gain,” such as an additional stroke prevented, case of cancer diagnosed, or gain of a year of life. Alternative interventions are compared in terms of cost per unit of effectiveness, and the resulting comparison is presented as a cost-effectiveness ratio. Relative certainty in the cost and outcome estimates continues to be a consideration. As a measure of cost-effectiveness, the Midwest CEPAC voting panel follows common academic and health technology assessment standards by using cost per quality-adjusted life year (QALY), with formal voting on “long- term value for money” when the base case incremental cost-effectiveness ratio is between $50,000 per QALY and $175,000 per QALY. 3. Other benefits refer to any significant benefits or disadvantages offered by the intervention to the individual patient, caregivers, the delivery system, other patients, or the public that would not have been considered as part of the evidence on comparative clinical effectiveness. Examples of other benefits include better access to treatment centers, mechanisms of treatment delivery that require fewer visits to the clinician’s office, treatments that reduce disparities across various patient groups, and new potential mechanisms of action for treating clinical conditions that have demonstrated low rates of response to currently available therapies. Other disadvantages could include increased burden of treatment on patients or their caregivers. For each intervention evaluated, it will be open to discussion whether other benefits or disadvantages such as these are important enough to factor into the overall judgment of long-term value for money. There is no quantitative measure for other benefits or disadvantages. 4. Contextual considerations include ethical, legal, or other issues (but not cost) that influence the relative priority of illnesses and interventions. Examples of contextual considerations include whether there are currently any existing treatments for the condition, whether the ©Institute for Clinical and Economic Review, 2018 Page 90 Final Evidence Report – Cystic Fibrosis Return to Table of Contents condition severely affects quality of life or not, and whether there is significant uncertainty about the magnitude of benefit or risk of an intervention over the long term. There is no quantitative measure for contextual considerations. Figure 8.1 Conceptual Structure of Long-term Value for Money 8.2 Voting Results Comparative Clinical Effectiveness 1) For individuals with approved gating, non-gating, and residual function mutations (including but not limited to G551D and R117H), is the evidence adequate to demonstrate that the net health benefit of treatment with Kalydeco (ivacaftor) with best supportive care is greater than that of best supportive care alone? Yes: 12 votes No: 0 votes Comments: After some discussion of the relatively small sample sizes available for the Kalydeco trials, the panel voted unanimously in the affirmative, based primarily on the sizable improvements in lung function observed. ©Institute for Clinical and Economic Review, 2018 Page 91 Final Evidence Report – Cystic Fibrosis Return to Table of Contents 2) For individuals who are homozygous for the F508del mutation, is the evidence adequate to demonstrate that the net health benefit of treatment with Orkambi (lumacaftor/ivacaftor) with best supportive care is greater than that of best supportive care alone? Yes: 11 votes No: 1 votes Comments: Prior to voting on this question, patient and clinical experts described some of the perceived differences between Orkambi and Kalydeco – for example, while Orkambi produces clinical benefit, the size of the benefit is modest in comparison to Kalydeco, which had already been on the market. However, Orkambi is available to a much broader set of patients, and for a different type of mutation with a different prognosis. Some panel members voiced hesitation about Orkambi’s net health benefits due to adherence concerns due to the chest tightness sensations some patients experience. 3) For individuals who are homozygous for the F508del mutation, is the evidence adequate to demonstrate that the net health benefit of treatment with Symdeko (tezacaftor/ivacaftor) with best supportive care is greater than that of best supportive care alone? Yes: 12 votes No: 0 votes Comments: The panel voted unanimously in the affirmative. Several panelists noted that Symdeko’s clinical effectiveness was similar to that of Orkambi, but with a lower discontinuation rate due to adverse events. 4) For individuals who are homozygous for the F508del mutation, is the evidence adequate to distinguish the net health benefit between treatment with Symdeko with best supportive care and Orkambi with best supportive care? Yes: 1 votes No: 11 votes Comments: While clinical experts mentioned that side effects and current stability on treatment might determine whether to choose Symdeko or Orkambi, the panelists’ vote was driven by a lack of head-to-head comparisons of the two agents and the results of the indirect comparison, which were not statistically significant. ©Institute for Clinical and Economic Review, 2018 Page 92 Final Evidence Report – Cystic Fibrosis Return to Table of Contents 5) For individuals who are candidates for Symdeko combination therapy because they carry one F508del mutation and residual function mutation that is potentially responsive to Symdeko, is the evidence adequate to demonstrate that the net health benefit of treatment with Symdeko with best supportive care is greater than that of best supportive care alone? Yes: 11 votes No: 1 votes Comments: Most of the panelists voted in the affirmative, noting a moderately-sized level clinical benefit. One panelist felt that the study design (8-week crossover trial) and small sample size precluded definitive conclusions at this time. Other Benefits and Contextual Considerations When compared to best supportive care, does Kalydeko, Orkambi, or Symdeko offer one or more of the following “other benefits”? (yes, no, uncertain) Potential Other Benefits # of votes This intervention offers reduced complexity that will significantly improve patient outcomes. 4 / 12 This intervention will reduce important health disparities across racial, ethnic, gender, socio- 0 / 12 economic, or regional categories. This intervention will significantly reduce caregiver or broader family burden. 8 / 12 This intervention offers a novel mechanism of action or approach that will allow successful 10 / 12 treatment of many patients for whom other available treatments have failed. This intervention will have a significant impact on improving the patient’s ability to return to work 7 / 12 or school and/or their overall productivity. This intervention will have a significant positive impact outside the family, including on schools 3 / 12 and/or communities. This intervention will have a significant impact on the entire “infrastructure” of care, including 2 / 12 effects on screening for affected patients, on the sensitization of clinicians, and on the dissemination of understanding about the condition, that may revolutionize how patients are cared for in many ways that extend beyond the treatment itself. Other important benefits or disadvantages that should have an important role in judgments of the 7 / 12 value of this intervention. Comments: Panel members voted that modulator treatments for CF provide several other benefits that are not necessarily captured in the clinical data. There was some discussion over whether modulator treatments will reduce or increase complexity of treatment. One one hand, it is an additional pill to take and does not replace any existing treatments. However, a clinical expert noted that some patients who have been successful on Orkambi have been able to stop inhaled ©Institute for Clinical and Economic Review, 2018 Page 93 Final Evidence Report – Cystic Fibrosis Return to Table of Contents antibiotics, which simplified their drug regimen. Four panelists voted that modulator treatments offer reduced complexity that will improve patient outcomes. Ten panelists felt that modulator treatments provide a new mechanism of action, although a panelist who did not vote for this other benefit mentioned that it is unknown whether this new mechanism will actually allow for the successful treatment of patients who have been failed by other therapies. Another benefit that several panelists noted was potential reductions in caregiver or broader family burden, based on a lower rate of pulmonary exacerbations and associated reductions in hospital days. These same benefits were also felt to have the potential effect of improving a patient’s ability to return to work or school, and to have a positive impact outside the family. Seven panelists voted that other benefits not listed above were also important. For example, one panelist said that modulator-related benefits like weight gain were inconsistently measured in clinical trials. Another panelist noted the potential for improved mental health of patients and caregivers the reduction in pulmonary exacerbations. Are any of the following contextual considerations important in assessing Kalydeco’s, Orkambi’s, or Symdeko’s long-term value for money in patients? (yes, no, uncertain) Potential Other Contextual Considerations # of votes This intervention is intended for the care of individuals with a condition of particularly high 12 / 12 severity in terms of impact on length of life and/or quality of life. This intervention is intended for the care of individuals with a condition that represents a 12 / 12 particularly high lifetime burden of illness. This intervention is the first to offer any improvement for patients with this condition. 5 / 12 Compared to best supportive treatment, there is significant uncertainty about the long-term risk of 3 / 12 serious side effects of this intervention. Compared to best supportive treatment, there is significant uncertainty about the magnitude or 10 / 12 durability of the long-term benefits of this intervention. There are additional contextual considerations that should have an important role in judgments of 7 / 12 the value of this intervention. Comments: The panel unanimously voted that CF patients have a condition of particularly high severity and a high lifetime burden of illness. There was some discussion about whether this intervention is the first to offer improvement for patients with this condition. Modulator treatments are the first type of treatment to modify the disease mechanism itself, rather than addressing symptoms only. On the other hand, a clinical expert noted there are many treatment options available to patients with CF, and best supportive care for CF patients has improved greatly over the past few decades. Five panelists voted that this intervention is the first to offer ©Institute for Clinical and Economic Review, 2018 Page 94 Final Evidence Report – Cystic Fibrosis Return to Table of Contents improvement, and several noted that they voted this way due to the modulators’ novel mechanism of action. Ten panelists voted that there is uncertainty about the long-term benefits of this intervention. Several panelists noted that, while there is early evidence that modulator treatments may slow the rate of decline of lung function over the long run, the magnitude of effect is currently very uncertain. Long-Term Value for Money Comments: When considering the long-term value for money of modulator treatments, the CEPAC panel discussed the treatment development process. The Cystic Fibrosis Foundation (CFF) played an integral role in funding the early research that led to CFTR modulator development. One panelist asked a representative from CFF if they attempted to exert control or influence over the price. The response was that the Foundation had concerns about sustainable pricing, but did not “have a seat at the table” when pricing decisions were made. Following this discussion, the Midwest CEPAC panel voted on the long-term value for money of CFTR modulator treatments. In all cases, a majority of the panel voted that the long-term value of these therapies was low. 6) For individuals with approved gating, non-gating, and residual function mutations (including but not limited to G551D and R117H), given the available evidence on comparative clinical effectiveness and incremental cost effectiveness, and considering other benefits and contextual considerations, what is the long-term value for money of Kalydeco with best supportive care compared with best supportive care alone? Low: 10 votes Intermediate: 2 votes High: 0 votes 7) For individuals who are homozygous for the F508del mutation, given the available evidence on comparative clinical effectiveness and incremental cost effectiveness, and considering other benefits and contextual considerations, what is the long-term value for money of Orkambi with best supportive care compared with best supportive care alone? Low: 11 votes Intermediate: 1 votes High: 0 votes ©Institute for Clinical and Economic Review, 2018 Page 95 Final Evidence Report – Cystic Fibrosis Return to Table of Contents 8) For individuals who are homozygous for the F508del mutation, given the available evidence on comparative clinical effectiveness and incremental cost effectiveness, and considering other benefits and contextual considerations, what is the long-term value for money of Symdeko with best supportive care compared with best supportive care alone? Low: 11 votes Intermediate: 1 votes High: 0 votes 9) For individuals who are candidates for Symdeko combination therapy because they carry one F508del mutation and residual function mutation that is potentially responsive to Symdeko, given the available evidence on comparative clinical effectiveness and incremental cost effectiveness, and considering other benefits and contextual considerations, what is the long-term value for money of Symdeko with best supportive care compared with supportive care alone? Low: 11 votes Intermediate: 1 votes High: 0 votes 8.3 Roundtable Discussion and Key Policy Implications Following its deliberation on the evidence, the Midwest CEPAC Panel engaged in a moderated discussion with a policy roundtable about how best to apply the evidence on modulator treatments for cystic fibrosis to policy and practice. The policy roundtable members included one patient advocate, one caregiver, two clinical experts, and two payers. The discussion reflected multiple perspectives and opinions, and therefore, none of the statements below should be taken as a consensus view held by all participants. The names of the Policy Roundtable participants are shown below, and conflict of interest disclosures for all meeting participants can be found in Appendix H. Table 8.1 Policy Roundtable Members Name Title and Affiliation Mary Dwight Senior VP of Policy and Advocacy, Cystic Fibrosis Foundation Jeremy Olimb Pastor, and father of children with cystic fibrosis David Orenstein, MD, Antonio J and Janet Palumbo Professor of Cystic MA Fibrosis, Children’s Hospital of Pittsburgh Manu Jain, MD, MS Professor of Medicine and Pediatrics, Director of Adult CF; Feinberg School of Medicine, Northwestern University Erik Schindler, Manager, Clinical Pharmacy; UnitedHealthcare PharmD, BCPS Pharmacy Jane Horvath, MHSA Senior Policy Fellow, National Academy for State Health Policy ©Institute for Clinical and Economic Review, 2018 Page 96 Final Evidence Report – Cystic Fibrosis Return to Table of Contents The roundtable discussion was facilitated by Dr. Steven Pearson, MD, MSc, President of ICER. The main themes and recommendations from the discussion are organized by type and summarized below. Much of the policy roundtable discussion centered on the difficulty in balancing the very real benefits realized by patients and families from the advent of CFTR modulators, including the hope that future innovation will bring even more effective treatments to a broader segment of the CF community, with the concern that the high costs for these drugs contribute not only to the financial toxicity experienced by CF patients but to the rising difficulties in maintaining access to affordable care for everyone. At the root of this discussion were the high prices for CFTR modulators, prices that the ICER analysis and subsequent votes of the Midwest CEPAC found to be far too high to align in reasonable fashion with the benefits and cost-offsets of these drugs. The ensuing discussion allowed a deeper exploration of the perspectives of patients and their families, of CF advocacy groups, and of public and private insurers wrestling with these challenges. Key policy themes and recommendations for future action are highlighted below. Key Recommendations on Pricing and Access 1. The prices for CFTR modulators are too high, harming patients and families today while threatening the health care system’s ability to maintain access for all patients to important future clinical advances. Benefiting from monopoly pricing power, the company bears a significant social responsibility to change its pricing approach by committing to the following two actions: • Abandon vague claims that prices are justified by the need to invest in future research and instead join the growing number of biotech innovators who provide a transparent, explicit justification for their prices based on the ability of treatments to improve the length and quality of patients’ lives; • Accept that the process for determining a reasonable price for new drugs requires innovators, especially those with monopoly pricing power at their disposal, to exercise restraint and be open to an independent process to evaluate fair pricing that includes the full engagement of the innovator, patients, patient advocacy groups, clinical experts, insurers, and other stakeholders. The first CFTR modulator was approved six years ago, so changes to the treatment pathway and adjustments to clinical practice are therefore reasonably mature. The manufacturer should no longer make vague justifications for the high prices of the CFTR modulators based on general statements about research and development costs or prospects for future innovation and continued investment in new treatments for CF. These arguments increasingly ring hollow in the absence of any quantification or further details to contextualize them. The manufacturer bears further responsibility to change their approach to justifying their pricing during a phase when they have enjoyed sustained company ©Institute for Clinical and Economic Review, 2018 Page 97 Final Evidence Report – Cystic Fibrosis Return to Table of Contents growth, rising profits and stock values, and have funded substantial stock repurchase programs. In addition, any benefit of the doubt given by payers when Kalydeco was first approved for a small subset of the CF population has vanished with the introduction of newer treatments for much larger groups of patients. The manufacturer should therefore be fully transparent about the calculus made for pricing of the CFTR modulators, and be willing to engage in processes intended to produce independent judgments of what fair pricing and sustainable access look like for CF innovations. 2. Public and private payers should continue to affirm their commitment to provide access to important clinical advances for CF and should remove superfluous requirements for coverage approval and continuation. Payers need to strive to provide broad access to treatments that improve patients’ lives while also seeking to control costs so that health care can be affordable for all. In the case of the CF drugs it is important for payers to seek to control costs without using access restrictions as a key feature of negotiation. Patients and their families need to know that insurers will help them receive these new drugs. Testimony provided at the policy roundtable highlighted that some payers impose unnecessary, and at times illogical, requirements for documentation prior to approval of insurance coverage. Examples include requirements for periodic genetic testing or other re-affirmation that a patient continues to have CF, a disease with immutable mutations and no current cure, or requiring that patients must be failed by supportive care medications before trying CFTR modulators, which are intended as add-on treatment to best supportive care rather than as a replacement. Such requirements pose an unnecessary burden, have no benefit for the patient, and do not engender trust in the payer by the treating clinician. 3. Since insurance coverage denial for CF drugs is off the table, payers should be willing to develop and adopt new approaches to moderate the impact of monopolistic pricing power. Historically, both public and private payers have had limited bargaining power in situations where a single manufacturer exercises its pricing power in a rare and underserved disease. Depending on the degree of engagement by the manufacturer in efforts to determine a value-based price, it may be necessary for payers to develop new approaches that can give them the ability to reign in excessive pricing. One prominent example is the recent program implemented by the New York Medicaid program. Under its new budget law, the program is empowered to highlight medications that contribute to growth in pharmaceutical spending above a designated budget cap. Selected medications can be relegated to a public process in which New York’s Drug Utilization Review Board (DURB) uses a variety of data points, including the results of independent drug value assessments, to determine a target price for supplemental rebate negotiation by Medicaid. If negotiations fail to reach at least 75% of the targeted rebate, the state has the right to require disclosure of research and development information from the manufacturer, along with other potential penalties. ©Institute for Clinical and Economic Review, 2018 Page 98 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Orkambi was the subject of the New York Medicaid meeting on April 26, 2018 and the DURB unanimously selected as a target price the price from the ICER report needed to meet a cost-effectiveness threshold of $150,000 per QALY. Further negotiations are ongoing, but the New York example already demonstrates how payers should explore additional tools that may give them additional leverage in moderating the health-system impact of excessively priced medications. 4. Patient organizations that have a leading role in funding, organizing, promoting, and otherwise fostering innovative research on new treatments should demand commitments from manufacturers for sustainable pricing of the products patients helped bring to the market. In the policy roundtable discussion, the CF Foundation described its central role in fostering the development of CFTR modulators as well as convincing manufacturers of the benefits of investing in CF innovation. It is likely that without the CF Foundation’s efforts, the drug developers would have prioritized other diseases and the innovations that resulted in CFTR modulator therapies may not have been realized. However, the CF Foundation has not had a “seat at the table” to discuss the pricing of these innovations. While other patient organizations should hold the CF Foundation up as an example of how to fund research and development in an underserved area, they must also couple this with a requirement of the developers to commit to sustainable pricing practices. 5. Professional societies should fully exercise their responsibility by bearing witness to the impact on their patients of failed pricing and insurance policies and by demanding to be part of the public process that should guide pricing to balance the needs for affordability and for investments in future innovation. Clinical experts on the roundtable agreed that the physician community could do more to advocate for fair and aligned pricing for the products they prescribe to their patients. There is considerable excitement in the clinical community about the potential for both short-term and sustained clinical benefit with the CFTR modulators, but physicians also have a front- row seat to the inequities and access challenges posed by the pricing of these drugs. The oncology community is an important model for physician activism, having highlighted the financial toxicity associated with new cancer regimens. The CF clinical community should consider a similar effort, given the financial challenges posed by CFTR modulators and other supportive-care treatments for CF. Recommendations to Improve Future Research 1. Future studies should measure and report a broad set of outcomes to better assess the health and economic impact of CF interventions to patients, their caregivers, and their health system. ©Institute for Clinical and Economic Review, 2018 Page 99 Final Evidence Report – Cystic Fibrosis Return to Table of Contents The evidence review showed a paucity of evidence on patient-centered outcomes pertaining to extrapulmonary manifestations of the disease, including but not limited to: mental health and affect, quality of life beyond the respiratory domain, impact on the endocrine, gastrointestinal, and functional effects of CF; impact on caregivers, including quality of life, affect, and time costs; and information on costs, including out of pocket costs, informal caregiver time, and transportation costs. Specifically, a CF core outcomes set (COS) should be developed and applied. A CF-specific COS is under development (see http://www.comet-initiative.org/studies/details/882 ; http://www.comet- initiative.org/studies/details/120) that is considering many of these measures. 2. Manufacturer-sponsored research should enroll patients who are often encountered in clinical practice, but who are routinely excluded from clinical trials. Clinical trials have often excluded patients who have a very high or very low ppFEV1, the very young, or those who have CF-related diabetes. Conducting studies in a broader set of populations can help assess hypothesized benefits of CFTR modulators across the spectrum of the disease, and allow payers to more readily accept the FDA’s increasingly broad decisions regarding indicated populations for treatment. Evidence suggests that CFTR modulators slow the rate of progression of the disease; thus, it is reasonable (and some might argue more ethical) to conduct studies in younger patients or, more generally, in patients who do not yet have extensive irreversible organ damage. Finally, targeting patients with extra-pulmonary manifestations such as CF related diabetes can measure the effects of outside of the respiratory system. 3. Leverage all available resources to maximize the evidence base. Because CF is relatively rare, effort should be made to maximize use of all existing data, including routinely collected information. There was considerable discussion of the CF Foundation’s registry at the policy roundtable, which collects a considerable amount of data on approximately 95% of the CF patients in the U.S. While research on some of the evidence gaps discussed above is ongoing and publications are forthcoming, the CF Foundation should consider broadening researcher access to the underlying data, using innovative designs (e.g., nested clinical trials, N-of-1 studies), and other measures to speed the generation of publicly-available information on these important topics. **** This is the first ICER review of modulator treatments for cystic fibrosis. ©Institute for Clinical and Economic Review, 2018 Page 100 Final Evidence Report – Cystic Fibrosis Return to Table of Contents References 1. Cystic Fibrosis Foundation. Cystic Fibrosis Foundation Patient Registry 2016 Annual Data Report. cff.org2017. 2. Clinical and Functional Translation of CFTR. CFTR2.org. 2018; https://www.cftr2.org/. 3. Foundation CF. What is Cystic Fibrosis. 2017; https://www.cff.org/What-is-CF/About-Cystic- Fibrosis/. Accessed October 24, 2017. 4. Donaldson SH, Pilewski JM, Griese M, et al. Tezacaftor/Ivacaftor in Subjects with Cystic Fibrosis and F508del/F508del-CFTR or F508del/G551D-CFTR. American journal of respiratory and critical care medicine. 2017. 5. Sawicki GS, Sellers DE, Robinson WM. High treatment burden in adults with cystic fibrosis: challenges to disease self-management. Journal of cystic fibrosis : official journal of the European Cystic Fibrosis Society. 2009;8(2):91-96. 6. Food and Drug Administration. U.S. Prescribing Information - Symdeko. In: HHS, ed. www.fda.gov2018. 7. National Institute for Health and Care Excellence (NICE). Lumacaftor–ivacaftor for treating cystic fibrosis homozygous for the F508del mutation. nice.org June 2016 2016. 8. Ramsey BW, Davies J, McElvaney NG, et al. A CFTR potentiator in patients with cystic fibrosis and the G551D mutation. The New England journal of medicine. 2011;365(18):1663-1672. 9. Davies JC, Wainwright CE, Canny GJ, et al. Efficacy and safety of ivacaftor in patients aged 6 to 11 years with cystic fibrosis with a G551D mutation. American journal of respiratory and critical care medicine. 2013;187(11):1219-1225. 10. De Boeck K, Munck A, Walker S, et al. Efficacy and safety of ivacaftor in patients with cystic fibrosis and a non-G551D gating mutation. Journal of cystic fibrosis : official journal of the European Cystic Fibrosis Society. 2014;13(6):674-680. 11. Moss RB, Flume PA, Elborn JS, et al. Efficacy and safety of ivacaftor in patients with cystic fibrosis who have an Arg117His-CFTR mutation: a double-blind, randomised controlled trial. The Lancet Respiratory medicine. 2015;3(7):524-533. 12. Bai Y, Higgins M, Volkova N, et al. Real-world outcomes in patients (PTS) with cystic fibrosis (CF) treated with ivacaftor (IVA): Analysis of 2014 US and UK CF registries. Journal of Cystic Fibrosis. 2016;15:S41. 13. Davies JC, Cunningham S, Harris WT, et al. Safety, pharmacokinetics, and pharmacodynamics of ivacaftor in patients aged 2-5 years with cystic fibrosis and a CFTR gating mutation (KIWI): an open-label, single-arm study. The Lancet Respiratory medicine. 2016;4(2):107-115. 14. Rowe SM, Heltshe SL, Gonska T, et al. Clinical mechanism of the cystic fibrosis transmembrane conductance regulator potentiator ivacaftor in G551D-mediated cystic fibrosis. American journal of respiratory and critical care medicine. 2014;190(2):175-184. 15. McKone EF, Borowitz D, Drevinek P, et al. Long-term safety and efficacy of ivacaftor in patients with cystic fibrosis who have the Gly551Asp-CFTR mutation: a phase 3, open-label extension study (PERSIST). The Lancet Respiratory medicine. 2014;2(11):902-910. 16. Wainwright CE, Elborn JS, Ramsey BW, et al. Lumacaftor-Ivacaftor in Patients with Cystic Fibrosis Homozygous for Phe508del CFTR. The New England journal of medicine. 2015;373(3):220-231. ©Institute for Clinical and Economic Review, 2018 Page 101 Final Evidence Report – Cystic Fibrosis Return to Table of Contents 17. Ratjen F, Hug C, Marigowda G, et al. Efficacy and safety of lumacaftor and ivacaftor in patients aged 6-11 years with cystic fibrosis homozygous for F508del-CFTR: a randomised, placebo- controlled phase 3 trial. The Lancet Respiratory medicine. 2017;5(7):557-567. 18. Taylor-Cousar JL, Munck A, McKone E, et al. Tezacaftor-Ivacaftor in Patients with Cystic Fibrosis Homozygous for Phe508del. The New England journal of medicine. 2017. 19. Konstan MW, McKone EF, Moss RB, et al. Assessment of safety and efficacy of long-term treatment with combination lumacaftor and ivacaftor therapy in patients with cystic fibrosis homozygous for the F508del-CFTR mutation (PROGRESS): a phase 3, extension study. The Lancet Respiratory medicine. 2017;5(2):107-118. 20. Milla CE, Ratjen F, Marigowda G, Liu F, Waltz D, Rosenfeld M. Lumacaftor/Ivacaftor in Patients Aged 6-11 Years with Cystic Fibrosis and Homozygous for F508del-CFTR. American journal of respiratory and critical care medicine. 2017;195(7):912-920. 21. Quittner AL, Modi AC, Wainwright C, Otto K, Kirihara J, Montgomery AB. Determination of the minimal clinically important difference scores for the Cystic Fibrosis Questionnaire-Revised respiratory symptom scale in two populations of patients with cystic fibrosis and chronic Pseudomonas aeruginosa airway infection. Chest. 2009;135(6):1610-1618. 22. Rowe SM, Daines C, Ringshausen FC, et al. Tezacaftor–Ivacaftor in Residual-Function Heterozygotes with Cystic Fibrosis. New England Journal of Medicine. 2017;377(21):2024-2035. 23. Elborn JS, Ramsey BW, Boyle MP, et al. Efficacy and safety of lumacaftor/ivacaftor combination therapy in patients with cystic fibrosis homozygous for Phe508del CFTR by pulmonary function subgroup: a pooled analysis. The Lancet Respiratory medicine. 2016;4(8):617-626. 24. Jennings MT, Dezube R, Paranjape S, et al. An Observational Study of Outcomes and Tolerances in Patients with Cystic Fibrosis Initiated on Lumacaftor/Ivacaftor. Annals of the American Thoracic Society. 2017. 25. Food and Drug Administration. Label- Orkambi. In: FDA; 2016. 26. Flume PA, VanDevanter DR. The Cystic Fibrosis Survival Gap: Why Do Canadians Fare Better Than Americans? Ann Intern Med. 2017;166(8):599-600. 27. Stephenson AL, Sykes J, Stanojevic S, et al. Survival comparison of patients with cystic fibrosis in Canada and the United States: a population-based cohort study. Annals of internal medicine. 2017;166(8):537-546. 28. Donaldson SH, Solomon GM, Zeitlin PL, et al. Pharmacokinetics and safety of cavosonstat (N91115) in healthy and cystic fibrosis adults homozygous for F508DEL-CFTR. Journal of cystic fibrosis : official journal of the European Cystic Fibrosis Society. 2017;16(3):371-379. 29. Pharmaceutical Pricing. U.S. department of Veterans Affairs; 2018. Accessed January 16th, 2018. 30. Lieu TA, Ray GT, Farmer G, Shay GF. The cost of medical care for patients with cystic fibrosis in a health maintenance organization. Pediatrics. 1999;103(6):e72. 31. Agrawal A, Mehta D, Sikachi RR, Du D, J. W. Nationwide trends of hospitalizations for cystic fibrosis in the United States from 2003 to 2013. Intractable Rare Dis Res. 2017;6(3):191-198. 32. O'Sullivan BP, Freedman SD. Cystic fibrosis. The Lancet. 2009;373(9678):1891-1904. 33. Cystic Fibrosis Foundation, Borowitz D, Parad RB, et al. Cystic Fibrosis Foundation practice guidelines for the management of infants with cystic fibrosis transmembrane conductance regulator-related metabolic syndrome during the first two years of life and beyond. Journal of Paediatrics. 2009;155(6):S106-116. 34. Britto MT, Kotagal UR, Hornung RW, Atherton HD, Tsevat J, Wilmott RW. Impact of recent pulmonary exacerbations on quality of life in patients with cystic fibrosis. Chest. 2002;121(1):64- 72. ©Institute for Clinical and Economic Review, 2018 Page 102 Final Evidence Report – Cystic Fibrosis Return to Table of Contents 35. Konstan MW, Morgan WJ, Butler SM, et al. Risk factors for rate of decline in forced expiratory volume in one second in children and adolescents with cystic fibrosis. The Journal of pediatrics. 2007;151(2):134-139, 139.e131. 36. Cystic fibrosis: Assessment and management of pancreatic sufficiency. UpToDate; 2017. https://www.uptodate.com/contents/cystic-fibrosis-assessment-and-management-of- pancreatic-insufficiency. Accessed March 6, 2018. 37. Comer DM, Ennis M, McDowell C, et al. Clinical phenotype of cystic fibrosis patients with the G551D mutation. QJM : monthly journal of the Association of Physicians. 2009;102(11):793-798. 38. National Institutes of Health. Cystic Fibrosis Fact Sheet. 2013; https://report.nih.gov/nihfactsheets/ViewFactSheet.aspx?csid=36. 39. Woolf S, Schünemann HJ, Eccles MP, Grimshaw JM, Shekelle P. Developing clinical practice guidelines: types of evidence and outcomes; values and economics, synthesis, grading, and presentation and deriving recommendations. Implementation Science : IS. 2012;7:61-61. 40. Wang X, Dockery DW, Wypij D, Fay ME, Ferris BG, Jr. Pulmonary function between 6 and 18 years of age. Pediatric pulmonology. 1993;15(2):75-88. 41. Hankinson JL, Odencrantz JR, Fedan KB. Spirometric reference values from a sample of the general U.S. population. American journal of respiratory and critical care medicine. 1999;159(1):179-187. 42. University of Miami. Cystic Fibrosis Questionnaire- Revised. 2017; http://www.psy.miami.edu/cfq_QLab/index.html. Accessed October 30, 2017. 43. Lung Institute. What is FEV1? Here's what you need to know. 2016; https://lunginstitute.com/blog/what-is-fev1-heres-what-you-need-to-know/. Accessed March 6, 2018. 44. Moran A, Brunzell C, Cohen RC, et al. Clinical care guidelines for cystic fibrosis-related diabetes: a position statement of the American Diabetes Association and a clinical practice guideline of the Cystic Fibrosis Foundation, endorsed by the Pediatric Endocrine Society. Diabetes Care. 2010;33(12):2697-2708. 45. Quittner AL, Buu A, Messer MA, Modi AC, Watrous M. Development and validation of The Cystic Fibrosis Questionnaire in the United States: a health-related quality-of-life measure for cystic fibrosis. Chest. 2005;128(4):2347-2354. 46. Robinson PD, Latzin P, Verbanck S, et al. Consensus statement for inert gas washout measurement using multiple- and single- breath tests. The European respiratory journal. 2013;41(3):507-522. 47. Aetna. Specialty Pharmacy Clinical Policy Bulletins Aetna Non-Medicare Prescription Drug Plan: Cystic Fibrosis. 2015; http://www.aetna.com/products/rxnonmedicare/data/2015/MISC/cystic_fibrosis.html. Accessed February 2018. 48. City BCBSK. Orkambi (lumacaftor/ivacaftor) Medical Policy. 2017; http://medicalpolicy.bluekc.com/MedPolicyLibrary/Prescription%20Drugs/Standard%20Prescrip tion%20Drugs%20Pharmacy%20Benefit/07-17%20Orkambi%20(lumacaftor-ivacaftor).pdf. Accessed February 2018. 49. Anthem. Orkambi (ivacaftor/lumacaftor). 2015; https://www11.anthem.com/provider/noapplication/f0/s0/t0/pw_e235679.pdf?na=pharminfo. Accessed February 2018. 50. Cigna. 2017 Cigna Prior Authorization Criteria. 2017; https://www.cigna.com/iwov- resources/medicare-2017/docs/prior-authorization-chs.pdf. Accessed February 2018. 51. City BCBSK. Kalydeco (ivacaftor) Medical Policy. 2017; http://medicalpolicy.bluekc.com/MedPolicyLibrary/Prescription%20Drugs/Standard%20Prescrip ©Institute for Clinical and Economic Review, 2018 Page 103 Final Evidence Report – Cystic Fibrosis Return to Table of Contents tion%20Drugs%20Pharmacy%20Benefit/07- 17%20Kalydeco%20(ivacaftor)%200817%20update.pdf. Accessed February 2018. 52. Anthem. Kalydeco (ivacaftor). 2015; https://www11.anthem.com/provider/noapplication/f0/s0/t0/pw_e181473.pdf?na=pharminfo. Accessed February 2018. 53. Agency for Health Care Administration SoF. Orkambi: (lumacaftor; ivacaftor). 2016. 54. Farrell PM, White TB, Ren CL, et al. Diagnosis of Cystic Fibrosis: Consensus Guidelines from the Cystic Fibrosis Foundation. The Journal of pediatrics. 2017;181(Supplement):S4-S15.e11. 55. Stallings VA, Stark LJ, Robinson KA, Feranchak AP, Quinton H. Evidence-Based Practice Recommendations for Nutrition-Related Management of Children and Adults with Cystic Fibrosis and Pancreatic Insufficiency: Results of a Systematic Review. Journal of the American Dietetic Association.108(5):832-839. 56. Mogayzel PJ, Jr., Naureckas ET, Robinson KA, et al. Cystic fibrosis pulmonary guidelines. Chronic medications for maintenance of lung health. American journal of respiratory and critical care medicine. 2013;187(7):680-689. 57. Flume PA, Mogayzel PJ, Jr., Robinson KA, et al. Cystic fibrosis pulmonary guidelines: treatment of pulmonary exacerbations. American journal of respiratory and critical care medicine. 2009;180(9):802-808. 58. Flume PA, Robinson KA, O'Sullivan BP, et al. Cystic Fibrosis Pulmonary Guidelines: Airway Clearance Therapies. Respiratory Care. 2009;54(4):522-537. 59. Saiman L, Siegel JD, LiPuma JJ, et al. Infection Prevention and Control Guideline for Cystic Fibrosis: 2013 Update. Infection Control and Hospital Epidemiology. 2014;35(S1):S1-S67. 60. National Institute for Health and Care Excellence. Cystic Fibrosis: diagnosis and management (NICE Guideline). NICE; October 25, 2017 2017. 61. Cook DJ, Mulrow CD, Haynes RB. Systematic reviews: synthesis of best evidence for clinical decisions. Ann Intern Med. 1997;126(5):376-380. 62. Higgins J, Green S,. Cochrane Handbook for Systematic Reviews of Interventions Version 5.1.0. The Cochrane Collaboration; 2011. 63. Moher D, Liberati A, Tetzlaff J, DG A. Preferred reporting items for systematic reviews and meta- analyses: the PRISMA statement. Annals of Internal Medicine. 2010;8(5):336-341. 64. Ollendorf DA, Pearson SD. An integrated evidence rating to frame comparative effectiveness assessments for decision makers. Medical care. 2010;48(6 Suppl):S145-152. 65. Agency for Healthcare Research and Quality. U.S. Preventive Services Task Force Procedure Manual. 2008. 66. Rucker G, Schwarzer G, Carpenter J, Olkin I. Why add anything to nothing? The arcsine difference as a measure of treatment effect in meta-analysis with zero cells. Statistics in medicine. 2009;28(5):721-738. 67. Bucher HC, Guyatt GH, Griffith LE, Walter SD. The results of direct and indirect treatment comparisons in meta-analysis of randomized controlled trials. Journal of clinical epidemiology. 1997;50(6):683-691. 68. Sawicki GS, McKone EF, Pasta DJ, et al. Sustained Benefit from ivacaftor demonstrated by combining clinical trial and cystic fibrosis patient registry data. American journal of respiratory and critical care medicine. 2015;192(7):836-842. 69. Borowitz D, Lubarsky B, Wilschanski M, et al. Nutritional Status Improved in Cystic Fibrosis Patients with the G551D Mutation After Treatment with Ivacaftor. Digestive diseases and sciences. 2016;61(1):198-207. ©Institute for Clinical and Economic Review, 2018 Page 104 Final Evidence Report – Cystic Fibrosis Return to Table of Contents 70. Quittner A, Suthoff E, Rendas-Baum R, et al. Effect of ivacaftor treatment in patients with cystic fibrosis and the G551D-CFTR mutation: patient-reported outcomes in the STRIVE randomized, controlled trial. Health and quality of life outcomes. 2015;13:93. 71. Davies J, Sheridan H, Bell N, et al. Assessment of clinical response to ivacaftor with lung clearance index in cystic fibrosis patients with a G551D-CFTR mutation and preserved spirometry: a randomised controlled trial. The Lancet Respiratory medicine. 2013;1(8):630-638. 72. Flume PA, Wainwright CE, Elizabeth Tullis D, et al. Recovery of lung function following a pulmonary exacerbation in patients with cystic fibrosis and the G551D-CFTR mutation treated with ivacaftor. Journal of cystic fibrosis : official journal of the European Cystic Fibrosis Society. 2017. 73. Bai Y, Higgins M, Volkova N, et al. Real-world outcomes in young (6-to 12-year-old) patients (pts) with cystic fibrosis (CF) treated with ivacaftor (IVA): Analysis of 2014 US and UK CF registries data. Journal of Cystic Fibrosis. 2016;15:S57-S58. 74. Konstan M, McKone E, Moss R, et al. Evidence of reduction in annual rate of fev1 decline and sustained benefits with lumacaftor and ivacaftor (LUM/IVA) in patients (PTS) with CF homozygous for f508del-cftr. Pediatric pulmonology. 2016;51:260. 75. Taylor-Cousar JL, Jain M, Barto TL, et al. Lumacaftor/ivacaftor in patients with cystic fibrosis and advanced lung disease homozygous for F508del-CFTR. Journal of cystic fibrosis : official journal of the European Cystic Fibrosis Society. 2017. 76. Food and Drug Administration. Ivacaftor Prescribing Information. In:2017. 77. Boyle MP, Bell SC, Konstan MW, et al. A CFTR corrector (lumacaftor) and a CFTR potentiator (ivacaftor) for treatment of patients with cystic fibrosis who have a phe508del CFTR mutation: a phase 2 randomised controlled trial. The Lancet Respiratory medicine. 2014;2(7):527-538. 78. Sanders GD, Neumann PJ, Basu A, et al. Recommendations for conduct, methodological practices, and reporting of cost-effectiveness analyses: Second panel on cost-effectiveness in health and medicine. Jama. 2016;316(10):1093-1103. 79. Sawicki GS, Konstan MW, McKone EF, et al. Rate of Lung Function Decline in Patients with Cystic Fibrosis (CF) Having a Residual Function Gene Mutation. American journal of respiratory and critical care medicine. 2017;195:A4847. 80. Konstan M, Wagener J, VanDevanter D. Risk factors for rate of decline in FEV1 in adults with cystic fibrosis. Journal of Cystic Fibrosis. 2012;11(5):405-411. 81. Goss CH, Burns JL. Exacerbations in cystic fibrosis: epidemiology and pathogenesis (part 1). Thorax. 2007;62(4):8. 82. Whiting P, Al M, Burgers L, et al. Ivacaftor for the treatment of patients with cystic fibrosis and the G551D mutation: a systematic review and cost-effectiveness analysis. Health technology assessment (Winchester, England). 2014;18(18):1-106. 83. VanDevanter DR, Kahle JS, O'Sullivan AK, Sikirica S, Hodgkins PS. Cystic fibrosis in young children: A review of disease manifestation, progression, and response to early treatment. Journal of cystic fibrosis : official journal of the European Cystic Fibrosis Society. 2016;15(2):147-157. 84. Lynch Jr, Sayah D, Belperio J, Weigt S. Lung transplantation for cystic fibrosis: results, indications, complications, and controversies. Seminars in Respiratory and Critical Care Medicine. 2015;36(2):299-320. 85. Adler A, Shine B, Chamnan P, Haworth C, Bilton D. Genetic determinants and epidemiology of cystic fibrosis-related diabetes: results from a British cohort of children and adults. Diabetes care. 2008;31(9):1789-1794. 86. Ahmed N, Corey M, Forstner G, et al. Molecular consequences of cystic fibrosis transmembrane regulator (CFTR) gene mutations in the exocrine pancreas. Gut. 2003;52(8):1159-1164. ©Institute for Clinical and Economic Review, 2018 Page 105 Final Evidence Report – Cystic Fibrosis Return to Table of Contents 87. Thabut G, Christie J, Mal H, et al. Survival benefit of lung transplant for cystic fibrosis since lung allocation score implementation. American journal of respiratory and critical care medicine. 2013;187(12):1335-1340. 88. The National Institute for Health and Care Excellence. NICE Committee Papers. 2016. 89. The National Institute for Health and Care Excellence. Lumacaftor–ivacaftor for treating cystic fibrosis homozygous for the F508del mutation. 2016. 90. Arias E, Heron M, Xu J. United States life tables, 2014. National vital statistics reports. 2014;66(4). 91. Liou T, Adler F, FitzSimmons S, Cahill B, Hibbs J, Marshall B. Predictive 5-Year Survivorship Model of Cystic Fibrosis. American Journal of Epidemiology. 2001;153(4):345-352. 92. Schechter MS, Trueman D, Farquharson R, Higuchi K, Daines CL. Inhaled aztreonam versus inhaled tobramycin in cystic fibrosis. An economic evlaution. Annals of the American Thoracic Society. 2015;12(7):1030-1038. 93. Tappenden P, Harnan S, Uttley L, et al. The cost effectiveness of dry powder antibiotics for treatment of Pseudomonas aeruginosa in patients with cystic fibrosis. PharmacoEconomics. 2014;32:159-172. 94. Busschbach JJ, Horikx PE, van den Bosch JM, Brutel de la Riviere A, de Charro FT. Measuring the quality of life before and after bilateral lung transplantation in patients with cystic fibrosis. Chest. 1994;105(3):911-917. 95. Red Book Online® Search. Truven Health Analytics; 2018. http://www.micromedexsolutions.com.ezp- prod1.hul.harvard.edu/micromedex2/librarian/CS/E7F89E/ND_PR/evidencexpert/ND_P/evidenc expert/DUPLICATIONSHIELDSYNC/A4E796/ND_PG/evidencexpert/ND_B/evidencexpert/ND_App Product/evidencexpert/ND_T/evidencexpert/PFActionId/redbook.FindRedBook?navitem=topRe dBook&isToolPage=true. Accessed January 15th, 2018. 96. Dilokthornsakul P, Patidar M, Campbell JD. Forecasting the Long-Term Clinical and Economic Outcomes of Lumacaftor/Ivacaftor in Cystic Fibrosis Patients with Homozygous phe508del Mutation. Value in Health. 2017. 97. Dilokthornsakul P, Hansen RN, Campbell JD. Forecasting US ivacaftor outcomes and cost in cystic fibrosis patients with the G551D mutation. The European respiratory journal. 2016;47(6):1697- 1705. 98. Moreno SG, Ray JA. The value of innovation under value-based pricing. Journal of Market Access & Health Policy. 2016;4:10.3402/jmahp.v3404.30754. 99. Garrison LP, Jr., Mansley EC, Abbott TA, 3rd, Bresnahan BW, Hay JW, Smeeding J. Good research practices for measuring drug costs in cost-effectiveness analyses: a societal perspective: the ISPOR Drug Cost Task Force report--Part II. Value in health : the journal of the International Society for Pharmacoeconomics and Outcomes Research. 2010;13(1):8-13. 100. Dagoni R. Outrage as Teva Charges $18,275 for Generic Drug in US. Jerusalem Post2018. 101. Ouyang L, Grosse SD, Amendah DD, Schechter M. Healthcare expenditures for privately insured people with cystic fibrosis. Pediatric pulmonology. 2009;44(10):989-996. 102. O’Sullivan AK, Sullivan J, Higuchi K, Montgomery AB. Health care utilization & costs for cystic fibrosis patients with pulmonary infections. Managed Care. 2011;20(2):37-44. 103. Bentley TS, Phillips SJ. 2017 U.S. organ and tissue transplant cost estimates and discussion. Milliman;2017. 104. Ramsey SD, Patrick DL, Albert RK, Larson EB, Wood DE, Raghu G. The cost-effectiveness of lung transplantation. A pilot study. University of Washington Medical Center Lung Transplant Study Group. Chest. 1995;108(6):1594-1601. ©Institute for Clinical and Economic Review, 2018 Page 106 Final Evidence Report – Cystic Fibrosis Return to Table of Contents 105. Neri L, Lucidi V, Catastini P, Colombo C. Caregiver burden and vocational participation among parents of adolescents with CF. Pediatric pulmonology. 2015. 106. Angelis A, Kanavos P, Lopez-Bastida J, Nicod E, Serrano-Aquilar P. Social and economic costs and health-related quality of life in non-institutionalised patients with cystic fibrosis in the United Kingdom. BMC Healt5h Serv Res. 2015. 107. Sanders DB, Bittner RC, Rosenfeld M, Hoffman LR, Redding GJ, Goss CH. Failure to recover to baseline pulmonary function after cystic fibrosis pulmonary exacerbation. American journal of respiratory and critical care medicine. 2010;182(5):627-632. 108. Patel S, Sinha IP, Dwan K, Echevarria C, Schechter M, Southern KW. Potentiators (specific therapies for class III and IV mutations) for cystic fibrosis. Cochrane Database Syst Rev. 2015(3):Cd009841. 109. Konstan MW, Ramsey BW, Elborn J, et al. Safety and efficacy of treatment with lumacaftor in combination with ivacaftor in patients with CF homozygous for F508DELCFTR. Pediatric pulmonology. 2015;50:269-270. 110. McColley SA, Konstan MW, Ramsey BW, et al. Association between changes in percent predicted fev1 and incidence of pulmonary exacerbations, including those requiring hospitalization and/or iv antibiotics, in patients with CF treated with lumacaftor in combination with ivacaftor. Pediatric pulmonology. 2015;50:282. 111. Accurso FJ, Rowe SM, Clancy JP, et al. Effect of VX-770 in persons with cystic fibrosis and the G551D-CFTR mutation. The New England journal of medicine. 2010;363(21):1991-2003. 112. Guigui S, Wang J, Cohen RI. The use of ivacaftor in CFTR mutations resulting in residual functioning protein. Respiratory medicine case reports. 2016;19:193-195. 113. Konstan MW, Plant BJ, Elborn JS, et al. Efficacy response in CF patients treated with ivacaftor: post-hoc analysis. Pediatric pulmonology. 2015;50(5):447-455. 114. Heltshe SL, Mayer-Hamblett N, Burns JL, et al. Pseudomonas aeruginosa in cystic fibrosis patients with G551D-CFTR treated with ivacaftor. Clinical infectious diseases : an official publication of the Infectious Diseases Society of America. 2015;60(5):703-712. 115. Barry PJ, Plant BJ, Simmonds NJ, et al. Ivacaftor decreases mortality in G551D patients with severe lung disease. Pediatric pulmonology. 2015;50:275-276. 116. Volkova N, Bai Y, Higgins M, et al. Disease progression in patients (pts) with cystic fibrosis (CF) treated with ivacaftor (IVA): Analysis of real-world data from the UK CF Registry. Journal of Cystic Fibrosis. 2016;15:S41. 117. Elborn S, Wainwright C, Sermet-Gaudelus I, et al. Pulmonary effects of the investigational CFTR potentiator, ivacaftor, in two phase 3 trials in subjects with CF who have the G551D-CFTR mutation. American journal of respiratory and critical care medicine. 2012;185. 118. Flume P, Wainwright C, Tullis E, Rodriguez S, Davies J, Wagener J. Pulmonary exacerbations in CF patients with the G551D-CFTR mutation treated with ivacaftor. Journal of Cystic Fibrosis. 2013;12:S63. 119. Bai Y, Higgins M, Volkova N, et al. Ivacaftor long-term safety study: Analysis of 2013 us CF foundation patient registry data. Pediatric pulmonology. 2015;50:284. 120. Mainz J, Narayanan S, Suthoff ED, et al. Patient-reported outcomes among patients (pts) with cystic fibrosis and the G551D-CFTR mutation treated with ivacaftor (IVA) compared with those homozygous for the F508del-CFTR mutation. Journal of Cystic Fibrosis. 2016;15:S115. 121. Accurso FJ, Ratjen F, Altes T, et al. Effect of withdrawal of ivacaftor therapy on CFTR channel activity and lung function in patients with cystic fibrosis. Journal of Cystic Fibrosis. 2013;12:S62. 122. Davies JC, Sheridan H, Lee PS, Song T, Stone A, Ratjen F. Effect of ivacaftor on lung function in subjects with CF who have the G551D-CFTR mutation and mild lung disease: A comparison of lung clearance index (LCI) vs. spirometry. Journal of Cystic Fibrosis. 2012;11:S15. ©Institute for Clinical and Economic Review, 2018 Page 107 Final Evidence Report – Cystic Fibrosis Return to Table of Contents 123. Elborn JS, Rodriguez S, Lubarsky B, Gilmartin G, Bell S. Effect of ivacaftor in patients with cystic fibrosis and the G551D-CFTR mutation who have baseline FEV1 >90% of predicted. Pediatric pulmonology. 2013;48:298. 124. Plant BJ, Konstan M, Aherns R, et al. Lung function, weight, and sweat chloride responses in patients with cystic fibrosis and the G551D-CFTR mutation treated with ivacaftor: A secondary analysis. Journal of Cystic Fibrosis. 2013;12:S62. 125. Suthoff E, Rendas-Baum R, Vera-Llonch M, Bayliss M, Sermet-Gaudelus I, Quittner AL. Patient- reported treatment effects of ivacaftor beyond respiratory symptoms in patients with cystic fibrosis (CF). Pediatric pulmonology. 2014;49:250. 126. Hathorne H, Brand KM, Britton LJ, et al. The investigation of quality of life and adherence in patients with the G551D mutation receiving ivacaftor therapy. Pediatric pulmonology. 2015;50:427. 127. Wainwright C, Bell S, Morton J, et al. The effect of ivacaftor in individuals with CF and severe lung disease. Pediatric pulmonology. 2014;49:376-377. 128. Barry PJ, Plant BJ, Nair A, et al. Effects of ivacaftor in patients with cystic fibrosis who carry the G551D mutation and have severe lung disease. Chest. 2014;146(1):152-158. 129. Edgeworth D, Keating D, Ellis M, et al. Improvement in exercise duration, lung function and well- being in G551D-cystic fibrosis patients: a double-blind, placebo-controlled, randomized, cross- over study with ivacaftor treatment. Clinical science (London, England : 1979). 2017;131(15):2037-2045. 130. Stalvey MS, Pace J, Niknian M, et al. Growth in Prepubertal Children With Cystic Fibrosis Treated With Ivacaftor. Pediatrics. 2017;139(2). 131. Fink A, Sawicki GS, Morgan WJ, Schechter MS, Rosenfeld M, Marshall BC. Treatment response to ivacaftor in clinical practice: Analysis of the us CF foundation patient registry. Pediatric pulmonology. 2015;50:361. 132. Heltshe SL, Godfrey EM, Josephy T, Aitken ML, Taylor-Cousar JL. Pregnancy among cystic fibrosis women in the era of CFTR modulators. Journal of cystic fibrosis : official journal of the European Cystic Fibrosis Society. 2017. ©Institute for Clinical and Economic Review, 2018 Page 108 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Appendix A. Search Strategies and Results Table A1. PRISMA 2009 Checklist # Checklist Item TITLE Title 1 Identify the report as a systematic review, meta-analysis, or both. ABSTRACT Structured 2 Provide a structured summary including, as applicable: background; objectives; data sources; Summary study eligibility criteria, participants, and interventions; study appraisal and synthesis methods; results; limitations; conclusions and implications of key findings; systematic review registration number. INTRODUCTION Rationale 3 Describe the rationale for the review in the context of what is already known. Objectives 4 Provide an explicit statement of questions being addressed with reference to participants, interventions, comparisons, outcomes, and study design (PICOS). METHODS Protocol and 5 Indicate if a review protocol exists, if and where it can be accessed (e.g., Web address), and, if Registration available, provide registration information including registration number. Eligibility 6 Specify study characteristics (e.g., PICOS, length of follow-up) and report characteristics (e.g., Criteria years considered, language, publication status) used as criteria for eligibility, giving rationale. Information 7 Describe all information sources (e.g., databases with dates of coverage, contact with study Sources authors to identify additional studies) in the search and date last searched. Search 8 Present full electronic search strategy for at least one database, including any limits used, such that it could be repeated. Study Selection 9 State the process for selecting studies (i.e., screening, eligibility, included in systematic review, and, if applicable, included in the meta-analysis). Data Collection 10 Describe method of data extraction from reports (e.g., piloted forms, independently, in Process duplicate) and any processes for obtaining and confirming data from investigators. Data Items 11 List and define all variables for which data were sought (e.g., PICOS, funding sources) and any assumptions and simplifications made. Risk of Bias in 12 Describe methods used for assessing risk of bias of individual studies (including specification of Individual whether this was done at the study or outcome level), and how this information is to be used studies in any data synthesis. Summary 13 State the principal summary measures (e.g., risk ratio, difference in means). Measures Synthesis of 14 Describe the methods of handling data and combining results of studies, if done, including Results measures of consistency (e.g., I2) for each meta-analysis. Risk of Bias 15 Specify any assessment of risk of bias that may affect the cumulative evidence (e.g., Across studies publication bias, selective reporting within studies). Additional 16 Describe methods of additional analyses (e.g., sensitivity or subgroup analyses, meta- Analyses regression), if done, indicating which were pre-specified. RESULTS Study Selection 17 Give numbers of studies screened, assessed for eligibility, and included in the review, with reasons for exclusions at each stage, ideally with a flow diagram. ©Institute for Clinical and Economic Review, 2018 Page 109 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Study 18 For each study, present characteristics for which data were extracted (e.g., study size, PICOS, Characteristics follow-up period) and provide the citations. Risk of Bias 19 Present data on risk of bias of each study and, if available, any outcome level assessment (see within Studies item 12). Results of 20 For all outcomes considered (benefits or harms), present, for each study: (a) simple summary Individual data for each intervention group (b) effect estimates and confidence intervals, ideally with a Studies forest plot. Synthesis of 21 Present results of each meta-analysis done, including confidence intervals and measures of Results consistency. Risk of Bias 22 Present results of any assessment of risk of bias across studies (see Item 15). Across Studies Additional 23 Give results of additional analyses, if done (e.g., sensitivity or subgroup analyses, meta- Analysis regression [see Item 16]). DISCUSSION Summary of 24 Summarize the main findings including the strength of evidence for each main outcome; Evidence consider their relevance to key groups (e.g., healthcare providers, users, and policy makers). Limitations 25 Discuss limitations at study and outcome level (e.g., risk of bias), and at review-level (e.g., incomplete retrieval of identified research, reporting bias). Conclusions 26 Provide a general interpretation of the results in the context of other evidence, and implications for future research. FUNDING Funding 27 Describe sources of funding for the systematic review and other support (e.g., supply of data); role of funders for the systematic review. From: Moher D, Liberati A, Tetzlaff J, Altman DG. The PRISMA Group (2009). Preferred Reporting Items for Systematic Reviews and Meta-Analyses: The PRISMA Statement. PLoS Med 6(6): e1000097. doi:10.1371/journal.pmed1000097 ©Institute for Clinical and Economic Review, 2018 Page 110 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Table A2. Search Strategies of Medline 1996 to Present with Daily Update and Cochrane Central Register of Controlled trials using PubMed® #1 Search cystic fibrosis[MeSH Terms] #2 Search cystic fibrosis transmembrane conductance regulator[MeSH Terms] #3 #1 or #2 #4 Search cystic fibrosis transmembrane conductance regulator (CFTR) potentiator #5 Search cystic fibrosis transmembrane conductance regulator (CFTR) corrector #6 Search cystic fibrosis transmembrane conductance regulator (CFTR) modulator #7 Search CFTR potentiator #8 Search CFTR corrector #9 Search CFTR modulator #10 Search ivacaftor #11 Search lumacaftor #12 Search tezacaftor #13 Search VX-770 #14 Search VX-809 #15 Search VX-661 #16 Search Kalydeco #17 Search Orkambi® #18 #4 or #5 or #6 or #7 or #8 or #9 or #10 or #11 or #12 or #13 or #14 or #15 or #16 or #17 #19 #3 and #18 Table A3. Embase Search Strategy #1 ‘cystic fibrosis transmembrane conductance regulator (CFTR) potentiator’ #2 ‘cystic fibrosis transmembrane conductance regulator (CFTR) corrector’ #3 ‘cystic fibrosis transmembrane conductance regulator (CFTR) modulator’ #4 ‘CFTR potentiator’ #5 ‘CFTR corrector’ #6 ‘CFTR modulator’ #7 ‘ivacaftor’:de OR ‘ivacaftor’:ab,ti #8 ‘lumacaftor’:de OR ‘lumacaftor’:ab,ti #9 ‘tezacaftor’:de OR ‘tezacaftor’:ab,ti #10 ‘ivacaftor plus lumacaftor’:de OR ‘ivacaftor plus lumacaftor’:ab,ti #11 ‘ivacaftor plus tezacaftor’:de OR ‘ivacaftor plus tezacaftor’:ab,ti #12 #1 or #2 or #3 or #4 or #5 or #6 or #7 or #8 or #9 or #10 or #11 ©Institute for Clinical and Economic Review, 2018 Page 111 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Figure A1. PRISMA Flow Chart Showing Results of Literature Search for CFTR Modulators 1897 potentially relevant references screened 1545 citations excluded Cell or Animal: 659 Population: 146 Intervention: 174 Comparator: 0 Outcomes: 0 Study Design: 566 352 references for full text review 303 citations excluded (conference abstract duplicated peer-reviewed publication, abstracts older than five years, no outcome of interest, sample size <100 for observational studies) 49 TOTAL 1) 10 RCTs a. 19 publications b. 9 conference abstracts & presentations 2) Non-randomized comparative studies a. 4 publications b. 6 conference abstracts & presentations 3) Single-arm studies a. 6 publications b. 5 conference abstracts & presentations ©Institute for Clinical and Economic Review, 2018 Page 112 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Appendix B. Previous Systematic Reviews and Technology Assessments We identified two completed technology assessments on ivacaftor and two assessments on Orkambi, one from the National Institute for Health and Care Excellence (NICE) in the UK and three from the Canadian Agency for Drugs and Technologies in Health (CADTH). These reviews are summarized below. Of note, NICE expects to publish a proposing an appraisal document on Symdeko treatment for treating cystic fibrosis in people with the F508del mutation. Technology Assessments NICE Technology Assessment Report: Lumacaftor-ivacaftor for treaing cystic fibrosis homozygous for the F508del mutation [TA398] (July, 27, 2016) https://www.nice.org.uk/guidance/ta398/chapter/1-Recommendationsksjhdf;alskjnef;awnefaw The National Institute for Health and Care Excellence (NICE) performed a review of Orkambi in 2016. NICE did not recommend treatment for CF patients 12 years or older who are homozygous for the F508del mutation. The decision was based on the clinical evidence and cost-effective analysis. For clinical effectiveness, NICE examined the TRAFFIC, TRANSPORT, and PROGRESS clinical trials. Despite the general good quality of these trials, the results might not be generalizable to patients with mild or severe CF due to the inclusion criteria. Furthermore, the clinical evidence was insufficient to determine the long-term effect of Orkambi, since the treatment period in the main trials was 24 weeks. NICE noted that reporting the average of week 16 and week 24 results, rather than week 24 data alone, was more favorable to Orkambi. Concerning safety, NICE concluded that Orkambi was generally well tolerated. NICE assessed cost effectiveness of Orkambi based on the manufacturer’s microsimulation model. NICE concluded the manufacturer’s model might overestimate the benefits of Orkambi treatment and substantially underestimate the costs. NICE is currently developing guidance on Symdeko combination therapy for treating cystic fibrosis with the F508del mutation. The review is now on the scoping stage, and the publication date is to be announced. ©Institute for Clinical and Economic Review, 2018 Page 113 Final Evidence Report – Cystic Fibrosis Return to Table of Contents CADTH: Ivacaftor CADTH Canadian Drug Expert Committee Final Recommendation (November, 19, 2015) https://www.cadth.ca/sites/default/files/cdr/complete/SR0430_complete_Kalydeco_R117H_Nov- 23-15_e.pdf Common Drug Review – Clinical Review Report (March, 13, 2015) https://www.cadth.ca/sites/default/files/cdr/clinical/SR0430_KalydecoR117H_CL_Report.pdf Orkambi CADTH Canadian Drug Expert Committee Final Recommendation (October, 28, 2016) https://www.cadth.ca/sites/default/files/cdr/complete/SR0471_complete_Orkambi-Oct-28-16.pdf This review from the Canadian Agency for Drugs and Technologies in Health (CADTH) focused on assessing ivacaftor in the treatment for cystic fibrosis in patients 18 years and older with the CFTR R117H mutation. CADTH recommended ivacaftor for treating cystic fibrosis in adult patients with the CFTR R117H mutation if the following criteria and condition are met: first, patients have confirmed diagnosis of CF with chronic sinopulmonary disease; second, discontinuation criteria should be developed for non-responders in consultation with physicians; third, there is a substantial reduction in price. CADTH assessed the clinical effectiveness of ivacaftor in the R117H residual function mutation population, which showed ivacaftor was associated with modest, clinically relevant changes in ppFEV1 and respiratory symptoms compared to placebo. No significant treatment effect was observed in the time to pulmonary exacerbations. Ivacaftor was associated with few serious adverse events or withdrawals due to adverse events in trials. Considering the limited sample size (n=69, KONDUCT) and short duration of the studies, CADTH concluded additional data are needed to determine the long-term safety of ivacaftor. After assessing the manufacturer’s economic model and conducting a Common Drug Review Reanalysis (CDR), CADTH concluded that for ivacaftor to be cost-effective, a price reduction of at least 98% would be necessary. Following ivacaftor, CADTH reviewed Orkambi. CADTH recommended that Orkambi not be reimbursed for the treatment of CF in patients aged 12 years and older who are homozygous for the F508del mutation. The clinical evidence suggested that the magnitude of ppFEV1, BMI, and pulmonary exacerbations improvement with Orkambi compared to placebo was of uncertain clinical significance. ©Institute for Clinical and Economic Review, 2018 Page 114 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Previous Systematic Reviews We identified one systematic review on ivacaftor.108 This review included four randomized control trials: one Phase II dose-ranging study (n=19), one adult (n=167) and one pediatric (n=52) Phase III trial with G551D mutation populations and one trial with homozygous F508del participants (n=140). The trial evaluating ivacaftor among the F508del population was also included. No clinical differences were reported for CFQ-R, lung function, pulmonary exacerbations, or weight outcomes. Adults taking ivacaftor reported significantly higher CFQ-R respiratory domain scores through 48 weeks compared to those taking placebo. Children on ivacaftor did not report similar improvements compared to placebo. Children and adults receiving ivacaftor both reported significant improvements in relative change from baseline in FEV1 at 24 weeks, and adults reported similarly significant improvement in FEV1 through 48 weeks. Pooled data showed significant improvements in absolute change from baseline in ppFEV1 at both 24 and 48 weeks for ivacaftor groups compared to placebo. Both studies reported improvement in weight and decreased rates of pulmonary exacerbations among ivacaftor groups. Pooled data from both Phase III studies showed increased rates of coughing and episodes of decreased pulmonary function among placebo recipients. Adults taking ivacaftor reported dizziness more frequently than placebo recipients. Neither trial reported a difference in study drug interruptions or discontinuations between placebo and ivacaftor groups. Overall, the authors concluded the Phase III trials in G551D populations showed sufficient efficacy and safety compared to placebo through 48 weeks of treatment, supporting the use of ivacaftor in children and adults at least six years old. We identified one systematic review and guidline document from the Cystic Fibrosis Foundation for the use of ivacaftor and Orkambi.33 The guideline was designed to advise the use of these medications for clinicians, CF patients, and their families. A multidisciplinary committee was assembled to develop clinical questions using the Patient-Intervention-Comparison-Outcome format. A systematic review on ivacaftor and Orkambi was conducted to find relevant publications. The published peer-reviewed literature was from database inception through April 2016 in Ovid, EMBASE, PubMed, Cochrane Library, and Google Scholar. RCTs reflecting PICO criteria were included in the meta-analysis. The evidence was assessed using the Grading of Recommendations, Assessment, Development, and Evaluation (GRADE) approach, and recommendations were based on the results. In summary, for adults and children age six and older with CF due to gating mutations other than G551D or R117H, the guideline panel made a conditional recommendation for treatment with ©Institute for Clinical and Economic Review, 2018 Page 115 Final Evidence Report – Cystic Fibrosis Return to Table of Contents ivacaftor. For those with two copies of F508del, the guideline panel made a strong recommendation for treatment with Orkambi for adults and children age 12 and older with an ppFEV1 <90%; and made a conditional for treatment with Orkambi for (1) adults and children age 12 or older with ppFEV1 >90% and (2) children age six to 11. ©Institute for Clinical and Economic Review, 2018 Page 116 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Appendix C. Ongoing Studies Estimated Title, Trial Sponsor, Study Design Treatment Arms Patient Population Key Outcomes Completion ClinicalTrials.gov Identifier Date Ivacaftor A Phase 3, 2 Part, Open-Label Phase III 1. Experimental: Part A- Ivacaftor Inclusion Criteria Primary Outcome Measures June 2020 Study to Evaluate the Safety, Group 1: Participants 12 to < 24 • Confirmed diagnosis of CF by • Part A: AEs, serum chemistry and Pharmacokinetics, and Open label months sweat chloride value or CF hematology abnormal lab values, standard Pharmacodynamics of Ivacaftor Group 2: Participants 6 to < 12 mutation criteria. 12 lead ECGs, vital signs, and in Subjects With Cystic Fibrosis Non- months • Must have 1 of the following 9 ophthalmologic examinations [ Time Group 3: Participants 3 to < 6 CFTR mutations on at least 1 Frame: Day 1 - Day 70 ] Who Are Less Than 24 Months randomized months) allele: G551D, G178R, S549N, • Part B: Same as above [ Time Frame: Day of Age and Have a CFTR Gating S549R, G551S, G1244E, S1251N, 1 - Week 24 ] Mutation Single group Group 4: Participants 0 to < 3 S1255P, or G1349D. • Part A: Peak concentrations (C3-6h) of months assignement • No clinically significant ivacaftor, M1 ivacaftor, and M6 ivacaftor [ Vertex Pharmaceuticals abnormalities in hematology, Time Frame: after 4 days of IVA treatment ] 2. Experimental: Part B – Ivacaftor serum chemistry, and vital signs • Part A: Ctrough of IVA, M1 IVA, and M6 Incorporated Estimated Group 5: Participants 12 to < 24 IVA [ Time Frame: after 4 days of IVA enrollment: 35 months Exclusion Criteria NCT02725567 treatment ] Group 6: Participants 6 to < 12 • History of abnormal liver months function or abnormal liver Secondary Outcome Measures Group 7: Participants 0 to < 6 function at screening months • History of solid organ or • Part B: Peak concentrations (C3-6h) of hematological transplantation IVA, M1 IVA, and M6 IVA [ Time Frame: • Hemoglobin (Hgb) <9.5 g/dL at through Week 24 ] screening • Part B: Ctrough of IVA, M1 IVA, and M6 • Chronic kidney disease of Stage IVA [ Time Frame: through Week 24 ] 3 or above • Part B: Absolute change from baseline in • Presence of a non-congenital or sweat chloride [ Time Frame: through progressive lens opacity or Week 24 ] cataract at Screening A Phase 3, 2-Arm, Open-label Phase III 1. Experimental: Ivacaftor will be Inclusion Criteria Primary Outcome Measures June 7, 2021 Study to Evaluate the Safety administered every 12 hours from Ivacaftor Arm: Subjects From • Safety assessments based on the number and Pharmaco-dynamics of 2-Arm Day 1 through the morning dose of Study 124 (above) Part B: of subjects with AEs and SAEs [Time Frame: Long-term Ivacaftor Treatment the Week 104 Visit. • Must have completed the last Baseline - safety follow-up (up to 24 weeks in Subjects With Cystic Fibrosis Open label study visit of Study 124 Part B. after last dose) ] Who Are Less Than 24 Months 3. No Intervention: Observational Ivacaftor Arm: Subjects Not From Arm Study 124 Part B: Secondary Outcome Measures of Age at Treatment Initiation Non- Absolute change in sweat chloride [ Time randomized Frame: Baseline - Week 104 ] ©Institute for Clinical and Economic Review, 2018 Page 117 Final Evidence Report – Cystic Fibrosis Return to Table of Contents and Have a CFTR Gating • Confirmed diagnosis of CF, or 2 Mutation Parallel CF-causing mutations. assignement • One of the following CFTR Vertex Pharmaceuticals mutations on at least 1 allele: Incorporated Estimated G551D, G178R, S549N, S549R, enrollment: 75 G551S, G1244E, S1251N, S1255P, NCT03277196 or G1349D. Exclusion Criteria Ivacaftor Arm: Subjects Not From Study 124 Part B: • History of any illness or condition that might pose an additional risk in administering ivacaftor to the subject An acute upper or lower respiratory infection, or pulmonary exacerbation, or changes in therapy for pulmonary disease within 4 weeks of Day 1 Phase 3b, Randomized, Double- Phase III 1. Experimental: Inclusion Criteria: Primary Outcome Measures June 30, 2018 blind, Placebo-controlled, TEZ 100 mg/IVA 150 mg fixed-dose • Prior discontinuation of • Incidence of respiratory adverse events Parallel Group Study to Assess 2-Arm combination tablet in the morning; lumacaftor/ivacaftor, with at least (AEs) [ Time Frame: At Day 56 ] the Safety, Efficacy, and IVA 150 mg tablet in the evening. 1 respiratory sign or symptom Number and proportion of subjects with Tolerability of Symdeko Randomized considered related to therapy. respiratory AEs will be reported (TEZ/IVA) in an Orkambi- 2. Interventions: • Resolution or stabilization of experienced Population Who Double-blind Drug: TEZ/IVA; IVA qualifying event(s) >28 days prior Secondary Outcome Measures Are Homozygous for the to Screening. • Absolute change in ppFEV1 [ Time Frame: Placebo matched to TEZ/IVA fixed- F508del CFTR Mutation Parallel • Discontinuation of from baseline to the average of the Day 28 dose combination tablet in the assignement lumacaftor/ivacaftor therapy must and Day 56 measurements ] morning; placebo matched to IVA Vertex Pharmaceuticals have occurred <8 weeks from the • Relative change in ppFEV1 tablet in the evening. Incorporated Estimated first dose of lumacaftor/ivacaftor. • Absolute change in CFQ-R score Interventions: Placebo enrollment: 90 • Homozygous for F508del • Tolerability, defined as the number and 2. NCT03150719 mutation in the CFTR gene proportion of study participants who discontinue treatment [ Time Frame: • FEV1 ≥25% and ≤90% of through Day 56 ] predicted normal for age, sex, and • Number and proportion of subjects who height. discontinued TEZ/IVA will be reported. • Exclusion Criteria: ©Institute for Clinical and Economic Review, 2018 Page 118 Final Evidence Report – Cystic Fibrosis Return to Table of Contents • Recent rapid or progressive Safety assessments based on the number deterioration in respiratory status. of subjects with adverse events (AEs) and • Receiving continuous oxygen at serious adverse events (SAEs) >2L/min or on face-mask ventilation. • An acute upper or lower respiratory infection, pulmonary exacerbation, or change in therapy for pulmonary disease within 28 days before Day 1. • Documentation of colonization with organisms associated with a more rapid decline in pulmonary status. • History of lung transplantation since most recent initiation of lumacaftor/ivacaftor. • Participation in an investigational drug study or use of a CFTR modulator within 28 days or 5 terminal half-lives of the investigational drug or modulator (whichever is longer). A Phase 1/2 Study of VX-445 in Phase II 1. Experimental Part A: VX-445 in Inclusion Criteria: Primary Outcome Measures April 6, 2018 Healthy Subjects and Subjects Healthy Subjects (HS) • Parts A, B, and C: • Absolute change in sweat chloride With Cystic Fibrosis 2-Arm Part A includes single dose •Female subjects must be of non- concentrations [Parts C, D, E, and F only] [ escalation. childbearing potential. Time Frame: from baseline through Day 29 Vertex Pharmaceuticals Randomized •Between the ages of 18 and 55 ] 2. Experimental: Part B: VX-445 in years, inclusive. Incorporated • Relative change in ppFEV1 [Parts D, E, HS •BMI of 18.0 to 32.0 kg/m2, Parallel inclusive, and a total body weight and F only] Part B includes multiple-dose NCT03227471 assignement >50 kg • Absolute change in CFQ-R score [Parts D, escalation. • Parts D, E, and F: E, and F only] Estimated •Body weight ≥35 kg. • Maximum observed concentration 3. Experimental: Part C: VX-445 in •Parts D and F: Heterozygous for enrollment: 224 (Cmax) of VX-445,TEZ and metabolites (M1- Triple Combination (TC) with F508del and an MF mutation TEZ/IVA in HS TEZ and M2-TEZ), IVA and metabolites (M1- •Part E: Homozygous for F508del IVA and M6-IVA) and VX-561 [ Time Frame: Multiple-dose escalation of VX-445 •FEV1 value ≥40% and ≤90% of in TC with TEZ/IVA predicted mean for age, sex, and from Day 1 through Day 43 ] height. ©Institute for Clinical and Economic Review, 2018 Page 119 Final Evidence Report – Cystic Fibrosis Return to Table of Contents 4. Experimental: Part D1: F/MF • Area under the concentration versus time genotypes TC Exclusion Criteria: curve during a dosing interval (AUCtau) of 100 mg VX-445 qd in TC with TEZ • Parts A, B, and C: VX-445, TEZ and metabolites (M1-TEZ and and IVA for 4 weeks. • History of febrile illness within M2-TEZ), IVA and metabolites (M1-IVA and 14 days before the first study drug M6-IVA) and VX-561 5. Experimental: Part D2: F/MF dose. • Observed pre-dose concentration genotypes TC-High • Glucose-6-phosphate (Ctrough) of VX-445, TEZ and metabolites Subjects will receive VX-445 in TC dehydrogenase (G6PD) deficiency. (M1-TEZ and M2-TEZ), IVA and metabolites with TEZ and IVA for 4 weeks. • Parts D, E, and F: (M1-IVA and M6-IVA) and VX-561 • History of clinically significant F/MF genotypes TC-Mid cirrhosis with or without portal Secondary Outcome Measures VX-445 in TC with TEZ and IVA for 4 weeks. hypertension. • Absolute change in sweat chloride • Glucose-6-phosphate concentrations [Parts C, D and E only] Experimental: Part D2: F/MF dehydrogenase (G6PD) deficiency. • Relative change in ppFEV1 [Parts D and E genotypes TC-Low • Lung infection with organisms only] VX-445 in TC with TEZ and IVA for 4 associated with a more rapid • Absolute change in CFQ-R score [Parts D weeks. decline in pulmonary status. and E only] • Maximum observed concentration • History of solid organ or Experimental: Part E: (Cmax) of VX-445,TEZ and metabolites (M1- hematological transplantation. F/F genotype - TC TEZ and M2-TEZ), and IVA and metabolites VX-445 in TC with TEZ and IVA for 4 (M1-IVA and M6-IVA) [ Time Frame: from weeks Day 1 through Day 43 ] Active Comparator: TEZ/IVA • Area under the concentration versus time TEZ and IVA for 4 weeks. curve during a dosing interval (AUCtau) of VX-445, TEZ and metabolites (M1-TEZ and Experimental: Part F: F/MF M2-TEZ), and IVA and metabolites (M1-IVA genotypes - TC and M6-IVA) Observed pre-dose concentration (Ctrough) VX-445 in TC with TEZ and VX-561 for 4 weeks. of VX-445, TEZ and metabolites (M1-TEZ and M2-TEZ), and IVA and metabolites (M1-IVA and M6-IVA) A Phase 2, Randomized, Phase II 1. Experimental: Part 1: F/MF Inclusion Criteria Primary Outcome Measures March 20, Double-blind, Controlled Study genotype -TC Low • Body weight ≥35 kg. • Safety and tolerability as assessed by 2018 to Evaluate the Safety and 2-Arm 80 mg of VX-659 qd in TC with TEZ • Subjects must have an number of subjects with adverse events Efficacy of VX-659 Combination and IVA for 4 weeks eligibleCFTR genotype. (AEs) and serious adverse events (SAEs) Therapy in Subjects Aged 18 Randomized F/MF genotype - TC Mid • Part 1 and Part 3: Heterozygous [From baseline through safety follow-up 240 mg of VX-659 qd in TC with TEZ (20 Weeks)] Years and Older With Cystic for F508del and an MF mutation • Absolute change in percent predicted Fibrosis Parallel and IVA for 4 weeks. (F/MF) forced expiratory volume in 1 second assignement F/MF genotype - TC High • Part 2: Homozygous for F508del (ppFEV1) [Baseline through Day 29 ] 400 mg VX-659 qd in TC with TEZ (F/F) and IVA for 4 weeks. ©Institute for Clinical and Economic Review, 2018 Page 120 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Vertex Pharmaceuticals Estimated Comparator: F/MF genotype - • FEV1 value ≥40% and ≤90% of Secondary Outcome Measures Incorporated enrollment: 105 placebo for 4 weeks. predicted mean for age, sex, and • Absolute change in sweat chloride height concentrations [From baseline through Day NCT03224351 2. Experimental: Part 2: F/F 29 ] genotype – TC Exclusion Criteria • Relative change in ppFEV1 400 mg of VX-659 qd in TC with TEZ Ivacaftor Arm: Subjects Not From • Absolute change in CFQ-R and IVA for 4 weeks Study 124 Part B: • Maximum observed concentration of VX- Comparator: F/F genotype - TEZ/IVA • History of clinically significant 659, TEZ, M1-TEZ, IVA, M1-IVA, and VX-561 cirrhosis with or without portal [Day 1 through Day 29 ] 3. Experimental: Part 3: F/MF hypertension. • Area under the concentration vs time genotype - TC • Glucose-6-phosphate curve during a dosing interval of VX-659, 400 mg of VX-659 qd in TC with TEZ TEZ, M1-TEZ, IVA, M1-IVA, and VX-561 dehydrogenase (G6PD) deficiency and VX-561 for 4 weeks • Observed pre-dose concentration of • Lung infection with organisms Comparator: F/MF genotype - drugs above associated with a more rapid Placebo decline in pulmonary status. History of solid organ or hematological transplantation. A Phase 3, Open Label Study to Phase III 1. Experimental Part A: Cohort 1 Inclusion Criteria Primary Outcome Measures September Evaluate the Pharmacokinetics, VX-661 50 mg qd + IVA 75 mg q12h • Subjects who weigh ≥15 kg • Part A: Maximum observed concentration 2018 Safety, and Tolerability of VX- 2-Arm Interventions: without shoes at the screening of VX-661 and ivacaftor [Day 1 and Day 14 ] 661 in Combination With Drug: VX-661 • All genotypes as specified by the • Part A: Area under the concentration Ivacaftor in Subjects 6 Through Open label Drug: Ivacaftor study protocol are eligible in Part versus time curve during a dosing interval 11 Years of Age With Cystic A. of VX-661 and ivacaftor Fibrosis, Homozygous or Non- Cohort 2 • The following genotypes are • Part B: Safety and tolerability of VX-661 in Heterozygous for the F508del randomized VX-661 50 mg qd + IVA 150 mg eligible in Part B: combination with ivacaftor as determined CFTR Mutation q12h o homozygous for the F508del by adverse events and serious adverse Interventions: CFTR mutation Parallel events [ Time Frame: from baseline Drug: VX-661 o heterozygous for the F508del Vertex Pharmaceuticals assignement through 29 Weeks] Drug: Ivacaftor CFTR mutation and with a second Incorporated • allele with a CFTR mutation Estimated • Secondary Outcome Measures 2. Experimental: Part B: VX-661 + predicted to have residual NCT02953314 enrollment: 72 • Part A: Cmax of selected metabolites for IVA function. VX-661 + IVA 75 mg q 12h or IVA o heterozygous for the F508del VX-661 and Ivacaftor [ Time Frame: Day 1 150 mg q 12h CFTR mutation and with a second and Day 14 ] Interventions: CFTR allele with a gating defect • Part A: AUCτ of selected metabolites for Drug: VX-661 that is clinically demonstrated to VX-661 and Ivacaftor Drug: Ivacaftor be ivacaftor responsive • Part A: Safety and tolerability of VX-661 • A sweat chloride value ≥60 in combination with ivacaftor as mmol/L or chronic sinopulmonary determined by adverse events (AEs) and and/or gastrointestinal disease serious adverse events (SAEs) [From consistent with a diagnosis of CF baseline through Day 31 ] ©Institute for Clinical and Economic Review, 2018 Page 121 Final Evidence Report – Cystic Fibrosis Return to Table of Contents • Subjects who are homozygous • Part B: Cmax of VX-661, M1-661, M2-661, for the F508del-CFTR mutation ivacaftor, M1-ivacaftor, and M6-ivacaftor [ must have a sweat chloride value Time Frame: Day 1 through Week 16 ] ≥60 mmol/L. • Part B: AUCτ of VX-661, M1-661, M2-661, • Subjects with ppFEV1 of ≥40% ivacaftor, M1-ivacaftor, and M6-ivacaftor • Subjects who are willing to • Part B: Absolute change in ppFEV1 remain on their stable CF • Part B: Relative change in ppFEV1 medication regimen through Day • Part B: Absolute change in weight 14 (Part A) or through Week 24 • Part B: Absolute change in weight for age (Part B) or, if applicable, through z-score the Safety Follow up Visit. • Part B: Absolute change in height • Female subjects of childbearing • Part B: Absolute change in height for age potential must have a negative z-score serum pregnancy test at the • Part B: Absolute change in body mass Screening Visit and a negative index (BMI) urine pregnancy test at the Day 1 • Part B: Absolute change in BMI for age z- Visit before receiving the first dose score of study drug. • Part B: Absolute change in sweat chloride • Subjects of childbearing Part B: Absolute change in CFQ-R score potential who are sexually active must meet the contraception requirements Exclusion Criteria • An acute upper or lower respiratory infection, pulmonary exacerbation, or changes in therapy for pulmonary disease within 28 days before Day 1 • A standard 12 lead ECG demonstrating QTc >450 msec at the Screening Visit. • Ongoing or prior participation in an investigational drug study or use of commercially available CFTR modulator (except physician- prescribed ivacaftor for approved indications) within 30 days of screening. Pregnant and nursing females ©Institute for Clinical and Economic Review, 2018 Page 122 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Intestinal Current Oberservational 1. Baseline measurements will be Inclusion Criteria Primary Outcome Measures June 2019 Measurements (ICM) to performed within a 4-week interval • Phe508del homozygous subjects • ICM Absolute change from baseline of the Evaluate the Activation of Case-only prior to the start of oral treatment aged 12 years and older with cumulative chloride secretory ion current Mutant CFTR in Subjects With with lumacaftor + ivacaftor. cystic fibrosis response to forskolin/IBMX and carbachol Cystic Fibrosis Aged 12 Years Estimated According to the phase 3 study • FEV1 ≥ 40% of predicted normal in rectal tissue as a CFTR biomarker and Older, Homozygous for the enrollment: 125 results by week 4 the gain of FEV1 for age, gender and height [Measurement at the baseline visit within a p.Phe508del-CFTR Mutation, levels off, drug levels are in steady (Knudson standards) or FEV1 > 4-week interval prior to the start of oral Treated With Lumacaftor in state and all reversible initial 35% of predicted normal for age, treatment with lumacaftor and ivacaftor; Combination With Ivacaftor reductions of lung function are gender and height at baseline, second measurement at a day 10 - 14 resolved. Thus the second stable lung function during the weeks after the initiation of oral treatment assessment will be performed preceding three months and no with lumacaftor and ivacaftor] during the initial steady state at a acute upper or lower respiratory Hannover Medical School day 10 - 14 weeks after the infection or pulmonary Secondary Outcome Measures initiation of oral treatment with exacerbation during the preceding NCT02807415 lumacaftor + ivacaftor. four weeks • Spirometry Absolute change from baseline in percent predicted FEV1 2. Study participants will be Exclusion Criteria [Measurement at the baseline visit within a 4-week interval prior to the start of oral requested to record the • An acute upper or lower administration of Orkambi® by date treatment with lumacaftor and ivacaftor; respiratory infection or pulmonary and time for 7 days before the second measurement at a day 10 - 14 exacerbation at baseline scheduled visit to perform weeks after the initiation of oral treatment • Advanced liver disease as functional CFTR assays. with lumacaftor and ivacaftor] documented by sonography • NPD Absolute change from baseline of • Abnormal liver function at the Sermet score of nasal transepithelial The local patient databases at the baseline, defined as ≥ 3 upper limit potential difference measurements (NPD) three sites will be searched for all of normal in minimum 3 of the as a CFTR biomarker [Measurement at the subjects who fulfil the inclusion following: serum aspartate baseline visit within a 4-week interval prior criteria. After all subjects have been removed from the list who fulfill transaminase, serum alanine to the start of oral treatment with one or more exclusion criteria, the transaminase, gamma-glutamyl lumacaftor and ivacaftor; second eligible subjects will be randomly transpeptidase, or total bilirubin measurement at a day 10 - 14 weeks after assigned to rank numbers. Subjects • Abnormal blood creatine the initiation of oral treatment with will then be contacted in the phosphokinase at baseline lumacaftor and ivacaftor] sequence as they appear in the rank • Creatinine clearance < 60 Sweat chloride testing Absolute change number list. mL/min from baseline of the chloride concentration in Gibson-Cooke pilocarpine iontophoresis • Co-medication with strong sweat test as a CFTR biomarker CYP3A inhibitors and inducers [Measurement at the baseline visit within a • Non-congenital lens opacities 4-week interval prior to the start of oral • Haemorrhoids (bleeding risk treatment with lumacaftor and ivacaftor; when taking rectal suction second measurement at a day 10 - 14 biopsies for ICM) ©Institute for Clinical and Economic Review, 2018 Page 123 Final Evidence Report – Cystic Fibrosis Return to Table of Contents • History of nasal surgery that weeks after the initiation of oral treatment removed the respiratory with lumacaftor and ivacaftor] epithelium • Topical treatment of nostrils in the 3 days prior to baseline Disturbing nasal aspects of secretions, erythema, crustae, ulcera, edema at baseline A Study of the Effect of Single center 1. The participants will have been Inclusion Criteria Primary Outcome Measures June 2018 Combination Lumacaftor and previously screened to make sure • Age 18 years old or greater • Change in glycated hemoglobin Ivacaftor on Markers of Open label they are candidates for the study. • Patients diagnosed with cystic (hemoglobin A1C) [Time Frame: 1 year] Hyperglycemia in Persons With These patients will be contacted fibrosis (CF), genotype • A blood test will be used to determine Cystic Fibrosis Single-group prior to their first visit to discuss homozygous Phe508del the hemoglobin A1c change while on the assignement enrollment in the study. • Subject has been started on medication. Massachusetts General lumacaftor/ivacaftor for clinical • Change in units of insulin used over a Hospital Estimated 2. At the study visit the participant reasons, with no contraindication period of 6 months to 1 year. will come to the CRC or DRC for a for starting the drug [Time Frame: 1 year] enrollment: 50 NCT02858843 research visit. The following will • Contraindications for taking drug occur at this study visit: informed include abnormal liver enzyme Secondary Outcome Measures consent; brief medical history; tests, renal dysfunction, pregnancy • Change in glycemia contingent on genetic weight and height; vital signs and or nursing mothers risk score [Time Frame: 1 year] blood pressure; blood draw for DNA • The investigators will examine how extraction, A1c and an extra Exclusion Criteria change in glycemia is dependent on research tube for storage. This will • Does not have a HgbA1c within 1 genotype at variants associated with type 2 be scheduled at a time that is year prior to starting medication. diabetes and insulin secretion using genetic convenient to the patient. • Has not been on the risk scores. combination therapy for at least 2 • Pulmonary function test (PFT) forced months expiratory volume at one second (FEV1) measurements [Time Frame: 1 year ] The investigators will compare how PFT measurement of FEV1 are related to changes in glycemia Effects of Orkambi on Obeservational 1. Experimental: Part 1: F/MF Inclusion Criteria Primary Outcome Measures December Exertional Dyspnea, Exercise genotype -TC Low • Confirmed diagnosis of CF and • Change in iso-time dyspnea rating from 2019 Performance, and Ventilatory Case-only 80 mg of VX-659 qd in TC with TEZ homozygous for the F508del baseline (visit 2) to visit 3 and 4 during Responses in Adults With Cystic and IVA for 4 weeks mutation in the cystic fibrosis constant-load exercise tests. [Parameters Fibrosis Estimated F/MF genotype - TC Mid transmembrane conductance will be measured during 3 visits. Visit 2 will enrollment: 16 240 mg of VX-659 qd in TC with TEZ regulator (CFTR) gene occur before the participants go on University of British Columbia and IVA for 4 weeks. • Aged 19 years or older Orkambi. Visit 3 and 4 will occur at 1 month F/MF genotype - TC High • Forced Expiratory Volume in 1 and 3 months after initiating full dose of second (FEV1.0) < 90% predicted ©Institute for Clinical and Economic Review, 2018 Page 124 Final Evidence Report – Cystic Fibrosis Return to Table of Contents NCT02821130 400 mg VX-659 qd in TC with TEZ • Body mass index greater than 16 drug, respectively. All visits will be and IVA for 4 weeks. or less than 30 kg/m2 completed within 4 months.] Comparator: F/MF genotype - • Currently non-smoking or a past • placebo for 4 weeks. smoking history of less than 20 Secondary Outcome Measures pack-years • Cardio-respiratory responses 2. Experimental: Part 2: F/F • Chronic activity-related dyspnea genotype – TC Exclusion Criteria • Quality of life measured using the St. 400 mg of VX-659 qd in TC with TEZ • Chronic airway infection with George's Respiratory Questionnaire. and IVA for 4 weeks Mycobacterium abscessus, Physical activity measured using the Comparator: F/F genotype - TEZ/IVA Burkholderia cepacia complex, or International Physical Activity other organisms with infection Questionnaire (long version) and Recent 3. Experimental: Part 3: F/MF control implications based on the Physical Activity Questionnaire. genotype - TC treating physicians 400 mg of VX-659 qd in TC with TEZ • Use of supplemental oxygen or and VX-561 for 4 weeks desaturation less than 85% with Comparator: F/MF genotype - exercise Placebo Diagnosis of pneumothorax in the past 4 weeks Personalized Therapy of Cystic 2-Arm 1. Cystic fibrosis, treated Inclusion Criteria Primary Outcome Measures October 2020 Fibrosis: Set-up of Response Cystic fibrosis patients treated • Cystic fibrosis patients treated • Correlation between biological markers Markers Non- either by Ivacaftor or by the by CFTR modulators (Ivacaftor or and clinical and physiological outcome randomized association Ivacaftor-Lumacaftor the association Ivacaftor- [Time Frame: 6 months] Hôpital Necker-Enfants Procedure: Nasal swab; rectal Lumacaftor) • Malades Parallel biopsy. • Cystic fibrosis patients non Secondary Outcome Measures assignement treated by CFTR modulators Correlation between biological markers and NCT02965326 2. Cystic fibrosis, non treated • Patients in whom cystic fibrosis clinical and physiological outcome [Time Estimated Cystic fibrosis patients, non treated diagnosis has been suspected, but Frame: 12 months] by a CFTR modulator excluded by physiological and enrollment: 75 Procedure: Nasal swab; rectal genetic investigations biopsy. Exclusion Criteria 3. Non-Cystic fibrosis • Pregnant or lactating women Patients in whom cystic fibrosis • Contraindication to nasal swab diagnosis has been suspected, but • Contraindication to rectal biopsy excluded by physiological and genetic investigations Procedure: Nasal swab; rectal biopsy. A Randomized, Double-blind, Randomized Experimental Sequence 1: Ivacaftor Inclusion Criteria Primary Outcome Measures September 25, Placebo-controlled, Crossover → Placebo • Confirmed diagnosis of CF and at 2018 Study to Evaluate the Efficacy Double-blind least one of the following: ©Institute for Clinical and Economic Review, 2018 Page 125 Final Evidence Report – Cystic Fibrosis Return to Table of Contents of Ivacaftor in Subjects With Subjects will be randomized to increased sweat chloride level, • Correlation between biological markers Cystic Fibrosis Who Are 6 Years Placebo- receive Ivacaftor, 150 mg every 12 identification of 2 CF causing and clinical and physiological outcome of Age and Older and Have controlled hours (q12h) for 8 weeks in mutations, or demonstration of [Time Frame: 6 months] Either a 3849 + 10KB C→T or Treatment Period 1 followed by abnormal nasal epithelial ion • D1152H-CFTR Mutation Single group Placebo matching Ivacaftor for 8 transport. Secondary Outcome Measures assignement weeks in Treatment Period 2. A • A 3849 + 10KB C→T or D1152H Correlation between biological markers and Vertex Pharmaceuticals washout period of 8 weeks will be mutation on at least 1 CFTR allele. clinical and physiological outcome [Time Incorporated Crossover study maintained between the 2 periods. • FEV1 ≥40% of predicted and Frame: 12 months] ≤105% of predicted at screening. Experimental: Sequence 2: Placebo NCT03068312 Estimated → Ivacaftor Exclusion Criteria enrollment: 50 Subjects will be randomized to • A G551D, G1244E, G1349D, receive Placebo matching to G178R, G551S, S1251N, S1255P, Ivacaftor for 8 weeks in Treatment S549N, S549R, or R117H mutation. Period 1 followed by Ivacaftor 150 • For subjects <18 years of age at mg q12h for 8 weeks in Treatment the Screening, evidence of Period 2. A washout period of 8 cataract/lens opacity determined weeks will be maintained between to be clinically significant by the the 2 periods. ophthalmologist. Use of any moderate or strong inducers or inhibitors of cytochrome P450 (CYP) 3A, including consumption of certain herbal medications and certain fruit and fruit juices, within 14 days before Day 1. A Phase 3, Open-label, Rollover Phase III 1. Experimental Part A: VX- Inclusion Criteria Part A: Primary Outcome Measures September Study to Evaluate the Safety 661/ivacaftor • Completed study drug Part A: Safety and tolerability of long-term 2019 and Efficacy of Long Term Open label VX-661 100 mg/ ivacaftor 150 mg Treatment Period in a parent treatment of VX-661 in combination with Treatment With VX-661 in fixed dose combination (FDC) tablet study (NCT02070744, ivacaftor based on adverse events (AEs), Combination With Ivacaftor in Non- daily (qd) in the morning and NCT02347657, NCT02516410, ophthalmologic exams, clinical laboratory Subjects Aged 12 Years and randomized ivacaftor 150 mg tablet qd in the NCT02392234, NCT02412111) or values, standard digital electrocardiograms Older With Cystic Fibrosis, evening study drug treatment and the (ECGs), vital signs, and pulse oximetry Homozygous or Heterozygous Parallel Safety Follow up Visit for subjects [Time Frame: from baseline through Study for the F508del-CFTR Mutation assignment 2. No Intervention Part: A from NCT02508207. Completion (up to 3 years)] Observational Cohort • Previously received at least 4 • Long-term Follow-up weeks of study drug before Vertex Pharmaceuticals Estimated Secondary Outcome Measures discontinuing in Part A of Study Incorporated enrollment: • Parts A and B: Absolute change from 3. Experimental Part B: VX- NCT02565914 to participate in 1116 baseline in ppFEV1 [Time Frame: from 661/ivacaftor another qualified Vertex study. NCT02565914 baseline through Week 96] ©Institute for Clinical and Economic Review, 2018 Page 126 Final Evidence Report – Cystic Fibrosis Return to Table of Contents VX-661 100 mg/ ivacaftor 150 mg • Completed the last required visit • Part A: Relative change from baseline in fixed dose combination (FDC) tablet of another qualified Vertex study ppFEV1 daily (qd) in the morning and before or during the Returning • Parts A and B: Number of pulmonary ivacaftor 150 mg tablet qd in the Visit in Part A Study NCT02565914. exacerbations evening • <18 years of age (age on the • Parts A and B: Absolute change from date of informed consent/assent baseline in body mass index (BMI) in the parent study) • Parts A and B: Absolute change from • Completed study drug baseline in BMI z-score for subjects aged Treatment Period in a parent <20 years study or study drug treatment and • Part A: Absolute change from baseline in the Safety Follow up Visit for CFQ-R score subjects from NCT02508207, but • Part A: Absolute change from baseline in do not elect to enroll in the body weight NCT02565914 Treatment Cohort; • Part A: Absolute change from baseline in or body weight z-score for subjects aged <20 • Received at least 4 weeks of years study drug treatment and • Part A: Absolute change from baseline in completed visits up to the last height z-score for subjects aged <20 years scheduled visit of the Treatment • Part A: Time-to-first pulmonary Period of a parent study (and the exacerbation Safety Follow up Visit for subjects • Part A: Pharmacokinetic (PK) parameters: from NCT02508207), but do not trough concentrations of VX-661 , a VX-661 meet eligibility criteria for metabolite (M1-661), ivacaftor, ivacaftor enrollment into the Treatment metabolite (M1-ivacaftor) Cohort • Part A: Observational Cohort: Safety, as • Part B: determined by related serious adverse • Did not withdraw consent from events (SAEs) [Time Frame: from baseline the parent study or Part A of Study through study Completion (up to 3 years)] NCT02565914. Part B: Safety and tolerability assessments • Completed study drug treatment including number of subjects with adverse during the Treatment Period in events (AEs) and serious adverse events [ Part A of NCT02565914, Studies Time Frame: from baseline through safety NCT02730208 or NCT03150719, or follow-up visit ] other eligible Vertex studies. • Previously received at least 4 weeks of study drug before discontinuing Study NCT02565914 to participate in another qualified Vertex study, which is defined as a Vertex study of investigational CFTR modulators that allows ©Institute for Clinical and Economic Review, 2018 Page 127 Final Evidence Report – Cystic Fibrosis Return to Table of Contents participation of subjects in Study NCT02565914. • Exclusion Criteria: • History of drug intolerance in the parent study that would pose an additional risk to the subject. • Participation in an investigational drug trial (including studies investigating VX-661/ivacaftor or lumacaftor/ivacaftor) other than the parent studies of NCT02565914 or other eligible Vertex studies investigating VX- 661 in combination with ivacaftor, or use of a commercially available CFTR modulator. A Phase 2, Randomized, Phase II 1. Experimental: Inclusion Criteria Primary Outcome Measures September Placebo-Controlled, Double- VX-661/ivacaftor • Homozygous for the F508del • Change in CT imaging score from baseline 2018 blind Study to Evaluate the Randomized Fixed-dose combination tablet of CFTR mutation at Week 72 [Time Frame: from baseline at Effect of VX-661 in VX-661 100-mg/ivacaftor 150-mg • Confirmed diagnosis of CF Week 72] Combination With Ivacaftor on Double-blind and an evening dose of ivacaftor • Percent predicted forced • Chest Imaging Endpoints in 150-mg to be taken approximately expiratory volume (ppFEV1) ≥70% Secondary Outcome Measures Subjects Aged 12 Years and Placebo- 12 hours after the morning dose of predicted normal for age, sex, • Safety and tolerability assessments Older With Cystic Fibrosis, controlled and height during screening. including number of subjects with adverse Homozygous for the F508del 2. Experimental: • Stable CF disease as judged by events (AEs) and serious adverse events Placebo the investigator (SAEs) CFTR Mutation Estimated visually-matched tablets to be taken [Time Frame: Through week 72] enrollment: 40 on the same schedule as the active Exclusion Criteria Vertex Pharmaceuticals treatment. • An acute upper or lower Incorporated respiratory infection, pulmonary exacerbation, or changes in NCT02730208 therapy (including antibiotics) for pulmonary disease within 28 days before Day 1 (first dose of study drug) • Pregnant or nursing females. • Sexually active subjects of reproductive potential who are not willing to follow the contraception requirements. ©Institute for Clinical and Economic Review, 2018 Page 128 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Any contraindication to undergoing chest imaging, as per the site's institutional guidelines A Phase 3, Rollover Study to Phase III 1. Experimental: Inclusion Criteria Primary Outcome Measures July 26, 2019 Evaluate the Safety of Long- Subjects <6 years of age and <14 kg • Completed 24 weeks of LUM/IVA Safety and tolerability assessments based term Treatment With Non- at enrollment: LUM 100 mg/IVA 125 treatment and the Safety Follow- on the number of subjects with adverse Lumacaftor/Ivacaftor randomized mg q12h. Subjects <6 years of age up Visit in Study VX15-809-115 events (AEs) and serious adverse events Combination Therapy in and ≥14 kg at enrollment: LUM 150 Part B (Study 115B, NCT02797132) (SAEs) [Time Frame: From baseline through Subjects Aged 2 Years and Open label mg/IVA 188 mg q12h. Subjects ≥6 • Completed 24 weeks of LUM/IVA safety follow-up (up to 98 weeks).] Older With Cystic Fibrosis, years of age at enrollment, treatment and the Safety Follow-• Homozygous for the F508del- Parallel regardless of weight: LUM 200 up Visit in Study 115B, but do not Secondary Outcome Measures mg/IVA 250 mg q12h. want to enroll in the Treatment CFTR Mutation assignment • Absolute change from baseline in sweat Cohort. chloride [From baseline through 96 weeks] • Received at least 4 weeks of • Absolute change from baseline in body Vertex Pharmaceuticals Actual LUM/IVA treatment and mass index Incorporated enrollment: 50 completed visits up to Week 24 • Absolute change in BMI-for-age Z-score and the Safety Follow-up Visit, if • Absolute change from baseline in weight NCT03125395 required, of Study 115B but are • Absolute change in weight-for-age Z- not taking LUM/IVA at the end of score the Study 115B Treatment Period • Absolute change from baseline in stature because of a drug interruption and (height) either did not receive Vertex • Absolute change from baseline in stature- approval to enroll in the for-age Z-score Treatment Cohort or do not want • Time-to-first pulmonary exacerbation to enroll in the Treatment Cohort. • Number of pulmonary exacerbations • Permanently discontinued • Number of Cystic Fibrosis (CF)-related LUM/IVA in Study 115B after hospitalizations receiving at least 4 weeks of • Absolute change from baseline in fecal treatment and remained in the elastase-1 (FE-1) levels study from the time of treatment • Absolute change from baseline in serum discontinuation through the Week levels of immunoreactive trypsinogen (IRT) 24 Visit and Safety Follow-up Visit, • Change from baseline in sputum if required. microbiology cultures • Absolute change from baseline in lung Exclusion Criteria clearance index (LCI)2.5 • Prematurely discontinued Absolute change from baseline in LCI5.0 LUM/IVA treatment in Study 115B. • Subjects with a history of allergy or hypersensitivity to LUM/IVA. • Liver function test (LFT) abnormality meeting criteria for ©Institute for Clinical and Economic Review, 2018 Page 129 Final Evidence Report – Cystic Fibrosis Return to Table of Contents LUM/IVA treatment interruption at the completion of Study 115B, for which no convincing alternative etiology is identified. • QTc value at the completion of Study 115B that would pose an additional risk to the subject in the opinion of investigator, and which should be discussed with the Vertex medical monitor Participation in an investigational drug trial (including studies investigating LUM and/or IVA) other than Study 115B. A Phase 3, Rollover Study to Phase III 1. Experimental: • Inclusion Criteria Primary Outcome Measures August 2018 Evaluate the Safety and Efficacy Experimental: Treatment Cohort: • Subjects entering the Treatment • Treatment Cohort: Assess safety and of Long-term Treatment With Non- lumacaftor/ivacaftor (6 through 11) Cohort must meet both of the tolerability of long term treatment of Lumacaftor in Combination randomized Lumacaftor (LUM) 200 mg every 12 following criteria: lumacaftor in combination with ivacaftor, With Ivacaftor in Subjects Aged hours (q12h)/ivacaftor (IVA) 250 mg • Completed study visits up to based on adverse events and changes in q12h (subjects aged 6 through 11 Week 24 of Study 109 or Week 26 6 Years and Older With Cystic Open label clinical laboratory values, vital signs, and of Study 011B and did not Fibrosis, Homozygous for the years) spirometry [Time Frame: up to 4 weeks permanently discontinue F508del-CFTR Mutation Parallel treatment after last dose [last dose = Week 96]] assignment 2. Experimental: • Subjects entering the • Treatment Cohort: Observational Cohort must meet 1 Vertex Pharmaceuticals Secondary Outcome Measures lumacaftor/ivacaftor (12 and older) of the following criteria: Incorporated Estimated • Treatment Cohort: Absolute change in enrollment: 240 LUM 400 mg q12h/IVA 250 mg • Completed 24 weeks of study Lung Clearance Index 2.5 (LCI2.5) (subjects q12h (subjects aged 12 years and drug treatment in Study 109 or NCT02544451 from Study 109 and the Study 011B LCI older) completed 24 weeks of study drug treatment and the Week 26 Safety Substudy only) [From baseline to Week 96] Follow up in Study 011B. • Absolute change in sweat chloride 3. No intervention: Received at least 4 weeks of study • Absolute change in body mass index Observational cohort • Absolute change in CFQ-R score Long-term follow-up drug and completed visits up to • Observational Cohort: Safety, as Week 24 of Study 109 or Week 26 determined by serious adverse events of Study 011B. [Time Frame: 2 years] • Treatment Cohort: Absolute change in Lung Clearance Index 5.0 (LCI5.0) • Absolute change in ppFEV1 • Relative change in ppFEV1 • Absolute change in body mass index (BMI)-for-age-z-score ©Institute for Clinical and Economic Review, 2018 Page 130 Final Evidence Report – Cystic Fibrosis Return to Table of Contents • Absolute change in weight • Absolute change in weight-for-age-z- score • Absolute change in height • Absolute change in height-for-age-z-score • Absolute change in Treatment Satisfaction Questionnaire for Medication (TSQM) domains • Time-to-first pulmonary exacerbation (subjects from Study 109 only) • Event of having at least 1 pulmonary exacerbation • Number of pulmonary exacerbations • Rate of change in LCI2.5 (subjects from Study 109 and the Study 011B LCI Substudy only) • Rate of change in LCI5.0 Rate of change in ppFEV1 Observational Study of Observational 1. Experimental: Inclusion Criteria Primary Outcome Measures December Outcomes in Cystic Fibrosis Ivacaftor • Male or female with confirmed • Number of pulmonary exacerbations and 2020 Patients With Selected Gating Cohort study Observational model: cohort diagnosis of CF16 duration of treatment for pulmonary Mutations on a CFTR Allele • At least 1 allele with 1 of the exacerbations [Time Frame: 48 Months] (The VOCAL Study) Estimated following CFTR mutations: G178R, • Percentage of patients with cultures enrollment: 90 S549N, S549R, G551S, G1244E, positive for Pseudomonas aeruginosa Vertex Pharmaceuticals S1251N, S1255P, G1349D • Percentage of patients with cultures Incorporated • Six years of age or older positive for bacteria other than Pseudomonas aeruginosa and for fungi Exclusion Criteria NCT02445053 • Absolute change in percent predicted • Previously exposed to ivacaftor, FEV1 except currently treated patients • Absolute change in weight, weight-for- who started ivacaftor treatment age Z score, body mass index (BMI), and within 6 months of enrollment BMI-for-age Z-score • Currently enrolled in a ivacaftor • Incidence and prevalence of interventional study or other comorbidities during ivacaftor treatment interventional therapeutic clinical compared to the period before ivacaftor study directed at CFTR modulation treatment • History of organ transplantation • Incidence and cause of deaths • Incidence and reason for organ transplantations ©Institute for Clinical and Economic Review, 2018 Page 131 Final Evidence Report – Cystic Fibrosis Return to Table of Contents A Study in US Cystic Fibrosis Observational 1. Cohort 1: Intervention Inclusion Criteria Primary Outcome Measures December Patients With the R117H-CFTR The Interventional cohort will not • Non Interventional Cohort • FEV1 and forced vital capacity [FVC]) [ 2019 Mutation to Confirm the Long- Cohort study be utilized. • Male or female with confirmed Time Frame: 36 Months ] term Safety and Effectiveness diagnosis of CF • Pulmonary exacerbations, use of IV of Kalydeco, Including Patients Estimated 2. Cohort 2: Non Intervention • Must have at least 1 allele of the antibiotics <18 Years of Age, Combining enrollment: 150 A Non-Interventional Cohort R117H-CFTR mutation • Height and weight measurements. BMI, Data Captured in the Cystic comprising pediatric (<18 years of • Enrolled in the US CFF Patient BMI-for-age z-score, and weight-for-age z- Fibrosis Foundation Registry age) and adult R117H-CFTR patients Registry score treated with commercially-available • With a record of ivacaftor From an Interventional Cohort • Death or transplantation ivacaftor. treatment initiation from 01 and a Non-Interventional • Hospitalizations January 2015 through 31 Cohort • Symptomatic sinus disease, Pulmonary 3. Cohort 3 - Historical December 2016 A Historical Cohort comprising data • complications, CF-related diabetes and Vertex Pharmaceuticals distal intestinal obstruction syndrome, from an earlier time period for • Historical Cohort Incorporated Hepatobiliary complications, Pancreatitis pediatric (<18 years of age) and • Patients with CF in the CFF adult patients with the R117H-CFTR Patient Registry as of 01 January • Information for the above shown CF- NCT02722057 mutation who have never been 2009 related complications as recorded in the exposed to ivacaftor and matched • Must have at least 1 allele of the registry will be evaluated on age, gender, and lung function R117H-CFTR mutation to patients in the Non- • Patients with no evidence of any Select pulmonary microorganisms Interventional Cohort. prior ivacaftor exposure Source: www.ClinicalTrials.gov (NOTE: studies listed on site include both clinical trials and observational studies) ©Institute for Clinical and Economic Review, 2018 Page 132 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Appendix D. Comparative Clinical Effectiveness Supplemental Information Figure D1. Effect of 400 mg Lumacaftor Twice Daily with Ivacaftor Compared to Matched Controls on Weight-for-Age and BMI Z-score ©Institute for Clinical and Economic Review, 2018 Page 133 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Genetic Specific Data on CFTR Modulators Ivacaftor The effect of ivacaftor differs by mutation.76 Below are the in vitro response thresholds and stratified efficacy data from clinical trials, adapted from the FDA label (prescribing information).76 Figure D2. Net Change Over Baseline (% of untreated normal) in CFTR-Mediated Chloride Transport Following Addition of Ivacaftor from FDA Label76 *Clinical data exist for these mutations ©Institute for Clinical and Economic Review, 2018 Page 134 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Figure D3. Efficacy Outcomes of Ivacaftor by Genetic Mutation from FDA Label 76 ©Institute for Clinical and Economic Review, 2018 Page 135 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Symdeko6 The effect of Symdeko differs by mutation.76 Below are the in vitro response thresholds and stratified efficacy data from clinical trials, adapted from the FDA label (prescribing information).76 Figure D4. Net Change Over Baseline (% of Untreated Normal) in CFTR-Mediated Chloride Transport Following Addition of Symdeko from FDA Label6 *Clinical data exist for these mutations; #F508del represents data from one allele ©Institute for Clinical and Economic Review, 2018 Page 136 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Figure D5. Efficacy Outcomes of Symdeko by Genetic Mutation from FDA Label6 ©Institute for Clinical and Economic Review, 2018 Page 137 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Forest Plots from Meta-Analysis Figure D6. A Meta-Analysis of ppFEV1 for Ivacaftor Versus Placebo in Patients with Gating and Residual Function Mutations (Difference in Change in Absolute Percentage Points Between Study Arms) C.I: confidence interval, IVA: ivacaftor, Phet: chi-square P value for heterogeneity ©Institute for Clinical and Economic Review, 2018 Page 138 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Figure D7. Meta-Analysis of Weight for Ivacaftor Versus Placebo in Patients with Gating and Residual Function Mutations (Difference in Change in Weight, in kg, Between Study Arms) C.I.: confidence interval, IVA: ivacaftor, Phet: chi-square P value for heterogeneity Figure D8. Meta-Analysis of CFQ-R Respiratory Domain for Ivacaftor Versus Placebo in Patients with Gating and Residual Function Mutations (Difference in Change in Scores Between Study Arms) C.I.: confidence interval, CFQ-R: Cystic Fibrosis Questionnaire-Revised, IVA: ivacaftor, Phet: chi-square P value for heterogeneity. ©Institute for Clinical and Economic Review, 2018 Page 139 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Figure D9. Meta-Analysis of Odds Ratio of Pulmonary Exacerbations for Ivacaftor Versus Placebo in Patients with Gating and Residual Function Mutations Abbreviations: C.I.: confidence interval, IVA: ivacaftor, OR: odds ratio, P het = chi-square P value for heterogeneity. Figure D10. Meta-Analysis of Risk Ratio of Pulmonary Exacerbations for Ivacaftor Versus Placebo in Patients with Gating and Residual Function Mutations C.I.: confidence interval, IVA: ivacaftor, Phet = chi-square P value for heterogeneity, RR: risk ratio. ©Institute for Clinical and Economic Review, 2018 Page 140 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Figure D11. Meta-Analysis of Hazard Ratio Pulmonary Exacerbations for Ivacaftor Versus Placebo in Patients with Gating and Residual Function Mutations C.I.: confidence interval, HR: hazard ratio, IVA: ivacaftor, Phet = chi-square P value for heterogeneity. Figure D12. Meta-Analysis of Proportion of Patients Who Discontinued Ivacaftor Due to Adverse Events Studies in grey provide shorter-term results than subsequent studies and are not included in the meta-analysis C.I.: confidence interval, IVA: ivacaftor, Phet: chi-square P value for heterogeneity. ©Institute for Clinical and Economic Review, 2018 Page 141 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Figure D13. Meta-Analysis of Proportion of Patients Who Discontinued Lumacaftor/Ivacaftor Due to Adverse Events Studies in grey provide shorter-term results than subsequent studies and are not included in the meta-analysis C.I.: confidence interval, LUM/IVA: lumacaftor/ivacaftor (with daily dosage in mg per drug), Phet : chi-square P value for heterogeneity Figure D14. Meta-Analysis of Proportion of Patients Who Discontinued Tezacaftor/Ivacaftor Due to Adverse Events C.I.: confidence interval, Phet: chi-square P value for heterogeneity, TEZ/IVA: tezacaftor/ivacaftor (with daily dosage in mg per drug) ©Institute for Clinical and Economic Review, 2018 Page 142 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Figure D15. Meta- Analysis of Proportion of Patients Who Discontinued Placebo Due to Adverse Events C.I.: confidence interval, Phet: chi-square P value for heterogeneity Figure D16. Meta-Analysis of Proportion of Patients with Grade 3 or 4 Adverse Events on Tezacaftor/Ivacaftor C.I.: confidence interval, Phet: chi-square P value for heterogeneity, TEZ/IVA: tezacaftor/ivacaftor (with daily dosage in mg per drug) ©Institute for Clinical and Economic Review, 2018 Page 143 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Figure D17. Meta-Analysis of Proportion of Patients with Grade 3 or 4 Adverse Events on Placebo Abbreviations: C.I.: confidence interval, Phet: chi-square P value for heterogeneity ©Institute for Clinical and Economic Review, 2018 Page 144 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Appendix E. Comparative Value Supplemental Information Table E1. Impact Inventory Included in This Analysis from… Sector Type of Impact Perspective? Notes on Sources Health Care Sector Societal Formal Health Care Sector Health Outcomes Longevity effects   Health-related quality of life effects   Adverse events   Modeled through discontinuation rate. Medical Costs Paid by third-party payers   Paid by patients out-of-pocket   Future related medical costs   Future unrelated medical costs   Informal Health Care Sector Health-Related Costs Patient time costs NA  Unpaid caregiver-time costs NA  Transportation costs NA  Non-Health Care Sectors Productivity Labor market earnings lost NA  Cost of unpaid lost productivity due to illness NA  Cost of uncompensated household production NA  Consumption Future consumption unrelated to health NA  Social services Cost of social services as part of intervention NA  Legal/Criminal Number of crimes related to intervention NA  justice Cost of crimes related to intervention NA  Education Impact of intervention on educational achievement of NA  population Housing Cost of home improvements, remediation NA  Environment Production of toxic waste pollution by intervention NA  Other Other impacts (if relevant) NA  ©Institute for Clinical and Economic Review, 2018 Page 145 Final Evidence Report – Cystic Fibrosis Return to Table of Contents One-Way Sensitivity Analyses Figure E1. Tornado Diagram for One-Way Sensitivity Analyses of Cost per QALY Gained for Ivacaftor Plus Best Supportive Care Versus Best Supportive Care Alone in CF Individuals with Gating Mutations Discount Rate [1%-5%] Independent PEx Reduction [0.5-1.0] Absolute ppFEV1 Gain [4.5%-15.5%] Slope of Utility Function [0.002-0.005] Avg. Annual BSC Costs [$49,622-$148,866] Avg. PEx Cost [$41,253-$123,759] Transplant Costs [$589,428-$1,768,248] Parameter Input High Parameter Input Low Avg. Annual DM Costs [$19,340-$58,020] $750,000 $1,000,000 $1,250,000 PEx: acute pulmonary exacerbation; BSC: best supportive care; DM: disease management Probability of transplant among individuals with ppFEV1<30% ©Institute for Clinical and Economic Review, 2018 Page 146 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Figure E2. Tornado Diagram for One-Way Sensitivity Analyses of Cost per QALY Gained for Orkambi Plus Best Supportive Care Versus Best Supportive Care Alone in CF Individuals Homozygous for F508del Mutation Independent PEx Reduction [0.5-1.0] Discount Rate [1%-5%] Slope of Utility Function [0.002-0.005] Avg. Annual BSC Costs [$49,979-$149,937] Avg. PEx Cost [$41,253-$123,759] Avg. Transplant Costs [$589,428-$1,768,284] Absolute ppFEV1 Gain [1.8%-3.8%] Parameter Input High Probability of Transplant [0.0-0.87] Parameter Input Low Avg. Annual DM Costs [$19,340-$58,020] $700,000 $950,000 $1,200,000 PEx: acute pulmonary exacerbation; BSC: best supportive care; DM: disease management Probability of transplant among individuals with ppFEV1<30% ©Institute for Clinical and Economic Review, 2018 Page 147 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Figure E3. Tornado Diagram for One-Way Sensitivity Analyses of Cost per QALY Gained for Kalydeco Plus Best Supportive Care Versus Best Supportive Care Alone in CF Individuals Heterozygous for F508del Mutation and Residual Function Mutation Independent PEx Reduction [0.5-1.0] Slope of Utility Function [0.002-0.005] Discount Rate [1%-5%] Avg. PEx Cost [$41,253-$123,759] Avg. Annual BSC Costs [$54,274-$162,823] Absolute ppFEV1 Gain [3.7%-5.8%] Parameter Input High Avg. Transplant Costs [$589,428-$1,768,284] Parameter Input Low Avg. Annual DM Costs [$19,340-$58,020] $800,000 $1,050,000 $1,300,000 $1,550,000 PEx: acute pulmonary exacerbation; BSC: best supportive care; DM: disease management Probability of transplant among individuals with ppFEV1<30% ©Institute for Clinical and Economic Review, 2018 Page 148 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Probabilistic Sensitivity Analyses Figure E4. Scatterplot of Cost and Effectiveness for Ivacaftor Plus Best Supportive Care and Best Supportive Care Alone in CF Individuals with Gating Mutations (1,000 Iterations) $10.0 $9.0 $8.0 Total Direct Medical Costs (Millions) $7.0 $6.0 $5.0 $4.0 $3.0 Best Supportive Care $2.0 $1.0 $0.0 10 12 14 16 18 20 22 24 26 28 30 Effectiveness (QALYs) ©Institute for Clinical and Economic Review, 2018 Page 149 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Figure E5. Scatterplot of Cost and Effectiveness for Lumacaftor/Ivacaftor Plus Best Supportive Care, Tezacaftor/Ivacaftor plus Best Supportive Care, and Best Supportive Care Alone in CF Individuals Homozygous for F508del Mutation (1,000 Iterations) $9.0 $8.0 $7.0 $6.0 Total Direct Medical Costs (Millions) $5.0 Best Supportive Care $4.0 Lumacaftor-Ivacaftor plus Best Supportive Care $3.0 $2.0 $1.0 $0.0 10 12 14 16 18 20 22 24 26 28 Effectiveness (QALYs) ©Institute for Clinical and Economic Review, 2018 Page 150 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Figure E6. Scatterplot of Cost and Effectiveness for Kalydeco Plus Best Supportive Care, Tezacaftor/ivacaftor plus Best Supportive Care, and Best Supportive Care Alone in CF Individuals Heterozygous for F508del Mutation and Residual Function Mutation (1,000 Iterations) $9.0 $8.0 $7.0 $6.0 Total Direct Medical Costs (Millions) $5.0 $4.0 $3.0 Best Supportive Care Tezacaftor-Ivacaftor $2.0 Ivafactor Monotherapy $1.0 $0.0 10 12 14 16 18 20 22 24 26 28 30 Effectiveness (QALYs) ©Institute for Clinical and Economic Review, 2018 Page 151 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Appendix F. Evidence Tables Study Design and Author & Year of Duration of Publication, Interventions (n) & Inclusion and Exclusion Follow-up, Patient Characteristics Outcomes Harms (Trial), Dosing Schedule Criteria (Sites & geographical Quality Rating location) Symdeko Taylor‑Cousar 18 Phase 3, randomized, N=504 Inclusion Age ppFEV1 Any AE, n (%) double-blind, • 12 years of age or older Mean, years (SD) Mean absolute change from (1) 227 (90.4) NEJM multicenter, placebo- (1) TEZ/IVA: 100 mg of • Confirmed diagnosis of (1) 26.9 (11.2) baseline, percentage points (2) 245 (95.0) tezacaftor once daily CF (95% CI) controlled, parallel- (2) 25.7 (9.5) 2017 • Two Phe508del alleles (1) 3.4 (2.7 to 4.0) group trial and 150 mg of ivacaftor Grade 3/4 AE, n (%) • Percentage of the (2) -0.6 (-1.3 to 0.0) twice daily (n=248) Female, n (%) Difference=4.0 (3.1 to 4.8) (1) 22 (8.8) EVOLVE - Homozygous predicted FEV1 between Trial conducted in 91 (1) 121 (48.8) (2) 29 (11.2) F508d 40% and 90% at sites in the United (2) Placebo (n=256) (2) 125 (48.8) ppFEV1 screening States, Canada, and Stable disease Mean relative change from SAE, n (%) Good Europe from January Percent predicted FEV1 baseline, % (95% CI) (1) 31 (12.4) 30, 2015, to January (1) 6.3 (5.1 to 7.4) Exclusion (ppFEV1) (2) 47 (18.2) 20, 2017. (2) −0.5 (−0.7 to 0.6) Mean, percentage Difference =6.8 (5.3 to 8.3) points (SD) Discontinuation d/t AE, n Duration of follow- (1) 59.6 (14.7) (%) Pulmonary exacerbation up: 24 weeks (2) 60.4 (15.7) (1) 7 (2.8) (PEx), no. of events (annualized estimated (2) 8 (3.1) BMI event rate) Mean, kg (SD) (1) 78 (0.64) Infective PEx of CF, n (%) (1) 20.96 (2.95) (2) 122 (0.99) (1) 75 (29.9) (2) 21.12 (2.88) (2) 96 (37.2) BMI Mean absolute change from *CFQ-R respiratory baseline, kg/m2 (95% CI) Cough, n (%) domain (1) 0.18 (0.08 to 0.28) (1) 66 (26.3) Mean, score (SD) (2) 0.12 (0.03 to 0.22) (2) 84 (32.6) (1) 70.1 (16.8) Difference=0.06 (−0.08 to (2) 69.9 (16.6) 0.19) Headache, n (%) (1) 44 (17.5) ©Institute for Clinical and Economic Review, 2018 Page 152 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Study Design and Author & Year of Duration of Publication, Interventions (n) & Inclusion and Exclusion Follow-up, Patient Characteristics Outcomes Harms (Trial), Dosing Schedule Criteria (Sites & geographical Quality Rating location) * Scores on (CFQ-R) CFQ-R Respiratory domain (2) 37 (14.3) range from 0-100, Mean absolute change from higher scores indicating baseline, points (95% CI) (1) 5.0 (3.5 to 6.5) a higher patient- (2) −0.1 (−1.6 to 1.4) reported QoL with Difference=5.1 (3.2 to 7.0) regard to respiratory status. Rowe 22 Phase 3, randomized, N=248 Inclusion Age ppFEV1 Any AE, n (%) double-blind, • 12 years of age or older Mean, years (SD) Mean absolute change from (1) 126 (78) NEJM multicenter, placebo- (1) Placebo (n=162) • Confirmed diagnosis of (1) 32.6 (13.9) baseline (2) 114 (73) CF Within-group, L (SD) controlled, two- (2) 36.3 (15.2) (3) 117 (72) 2017 (2) IVA: Kalydeco, 150 • One Phe508del allele (1) −0.02 (0.21) period, three- (3) 35.6 (13.5) mg every 12 hours and one allele with a (2) 0.17 (0.23) intervention (n=157) (3) 0.23 (0.25) Grade 3/4 AE, n (%) EXPAND - residual-function crossover trial mutation Sex (1) 9 (6) Heterozygous F508d Female, n (%) Between-group, least- (2) 8 (5) (3) TEZ/IVA; tezacaftor • Percentage of the Trial conducted at 86 100 mg once daily with predicted FEV1 between (1) 46 (58) squared mean differences, L (3) 4 (2) Good ivacaftor 150 mg every 40% and 90% at (95% CI) sites from March 27, (2) 40 (49) 12 hours (n=162) screening Iva v. Plac: 4.7 (3.7 to 5.8) 2015, to Feb 16, (3) 48 (58) SAE, n (%) • Stable disease Tez/Iva v. Plac: 6.8 (5.7 to 2017. 7.8) (1) 14 (9) Incomplete block Type of Residual Tez/Iva v. Iva: 2.1 (1.2 to 2.9) (2) 10 (6) design Exclusion Duration of follow- Function Mutation, (3) 8 (5) up: 24 weeks • Any comorbidity or lab n (%) ppFEV1 Randomized 1:1:1:1:1:1 abnormality that may Class V Mean relative change from Discontin d/t AE, n (%) to 6 blocks each baseline, % confound study results (1) 48 (60) (1) 1 (<1) containing two Within-group, % (SD) or increase potential (2) 48 (59) (2) 2 (<1) interventions of 8 (1) -0.16 (9.45) harm to participant (3) 50 (60) (3) 0 weeks with an 8-week (2) 8.40 (10.76) • PE or change in (3) 11.17 (12.39) washout period treatment within 14 Class II-IV Infective PEx of CF, n (%) between. Participants days first dose (1) 32 (40) Between-group, least- (1) 31 (19) were randomized to • Prolonged QT/QTc (2) 33 (41) squared mean differences, % (2) 20 (13) receive two of three (95% CI) interval ©Institute for Clinical and Economic Review, 2018 Page 153 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Study Design and Author & Year of Duration of Publication, Interventions (n) & Inclusion and Exclusion Follow-up, Patient Characteristics Outcomes Harms (Trial), Dosing Schedule Criteria (Sites & geographical Quality Rating location) interventions studied • Solid organ transplant (3) 33 (40) Iva v. Plac: 8.1 (6.3 to 9.9) (3) 21 (13) for 8 weeks each with • Used inhibitors or Tez/Iva v. Plac: 11.4 (9.6 to an 8-week washout inducers of CYP3A4 ppFEV1 13.2) Cough, n (%) Tez/Iva v. Iva: 3.3 (1.8 to 4.8) period between. • Participation in another Mean, percentage (1) 30 (19) trial in last 3 months points (SD) CFQ-R (2) 17 (11) • Pregnancy or breast- (1) 62.1 (14.0) Mean change from baseline, (3) 23 (14) feeding (2) 62.8 (14.6) points • History or evidence of (3) 61.8 (14.9) Within-group: NR Headache, n (%) cataracts or lens (1) 13 (8) BMI Between-group, least- (2) 11 (7) opacity squares mean difference, • Use of restricted Mean, kg (±SD) (3) 19 (12) points (95% CI): medications or foods in (1) 24.6 (5.0) Iva vs. Plac: 9.7 (7.2 to 12.2) specified window (2) 24.5 (5.5) Tez/Iva vs. Plac: 11.1 (8.7 to Hemoptysis, n (%) before first dose (3) 23.6 (4.6) 13.6) (1) 14 (9) • Unwilling to take Tez/Iva vs. Iva: 1.4 (−1.0 to (2) 17 (11) CFQ-R Respiratory 3.9) (3) 12 (7) contraceptives during study if of reproductive domain PExs potential Mean, mean (±SD) Increase in creatinine, n Number of events • Colonization with (1) 67.8 (17.5) (1) 20 (%) organisms associated (2) 70.0 (17.7) (2) 9 (1) 5 (3) with more rapid (3) 66.5 (17.9) (3) 11 (2) 8 (5) decline in pulmonary (3) 6 (4) Pancreatic Estimated event rate/year status (1) 0.63 insufficiency, n (%) (2) 0.29 Yes (3) 0.34 (1) 11 (14) (2) 11 (14) Rate ratio v. placebo (95% (3) 11 (13) CI) • (2) (0.21 to 1.01) No (3) (0.26 to 1.1.3) (1) 56 (70) ©Institute for Clinical and Economic Review, 2018 Page 154 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Study Design and Author & Year of Duration of Publication, Interventions (n) & Inclusion and Exclusion Follow-up, Patient Characteristics Outcomes Harms (Trial), Dosing Schedule Criteria (Sites & geographical Quality Rating location) (2) 61 (75) (3) 60 (72) Missing (1) 13 (16) (2) 9 (11) (3) 12 (14) Donaldson 4 Phase 2, randomized, N=41 Inclusion Pooled Homozygous ppFEV1 AE in all homozygous placebo-controlled, • Confirmed diagnosis of F508del Mean (least-squares) F508del Am J Resp Crit Care multicenter, dose- Multiple doses in trial. CF (1) N=17 absolute change from Med Only reporting relevant • Homozygosity for the baseline, percentage points escalation study (2) N=24 Any AE, n (%) dose (95% CI) Phe-508del CFTR (1) 92 (86.8) (1) 3.75 (NR) 2017 37 centers in US, mutation Age (2) 30 (90.9) (1) TEZ/IVA: 100 mg qd (2) -0.14 (NR) Canada, Germany • Age of 18 years or Mean, years (±SD) Difference=3.89 (0.94 to tezacaftor and 150 mg Phase 2 and UK. Enrollment: ivacaftor q 12 hours older (1) 31.0 (9.3) 6.83) Any Serious AE, n (%) Feb 2012 to March (n=17) • ppFEV1 at the time of (2) 30.2 (7.8) (1) 8 (7.5) Good screening that was 40- ppFEV1 2014 (2) 5 (15.2) (2) Placebo (n=24) 90% of the predicted Mean (least-squares) Sex normal values relative change from Duration of follow- Female, n (%) Serious PEx, n (%) • Body weight of at least baseline, percent (95% CI) up: 56 days for (1) 11 (64.7) (1) NR (NR) (1) 7 (6.6) 40 kg and BMI of at safety; 28-days (2) 8 (33.3) (2) NR (NR) (2) 5 (15.2) least 18.5 kg/m2 efficacy Difference=7.04 (1.77 to • ppFEV1 12.31) Discontinuation due to Exclusion Only reporting on • Any comorbidity or lab Mean, percentage AE, CFQ-R Respiratory domain homozygous abnormality that may points (SD) n (%) Mean absolute change from F508del, TEZ/IVA confound study results (1) 58.7 (16) baseline, points (p-value) (1) 2 (11.8) 100/150mg or increase potential (2) 57.8 (15.3) (1) 3.79 (p=0.1679) (2) 0 (0) combination and harm to participant (2) NR (NR) placebo • PE or change in BMI Difference=6.81 (p=0.2451) Cough, n (%) treatment within 14 Mean, kg (SD) (1) 17 (16.0) days first dose (1) 23.0 (3.7) (2) 6 (18.2) (2) 21.7 (2.4) ©Institute for Clinical and Economic Review, 2018 Page 155 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Study Design and Author & Year of Duration of Publication, Interventions (n) & Inclusion and Exclusion Follow-up, Patient Characteristics Outcomes Harms (Trial), Dosing Schedule Criteria (Sites & geographical Quality Rating location) • Pregnancy or breast- feeding • Unwilling to take contraceptives during study if of reproductive potential • History of solid organ transplant • Participation in another trial in last 3 months • History of alcohol, medication, or illicit drug use within 1 year before screening • Orkambi Wainwright 16 Two phase 3, double- N=1108 Inclusion Age Pooled Analysis, least- Any AE, n (%) blind, placebo- • Confirmed diagnosis of Mean, years squares means (1) 356 (96.5) NEJM controlled, (1) LUM/IVA: 600 mg of CF (1) 24.5 (2) 351 (95.1) ppFEV1 randomized trial lumacaftor once daily in • Homozygosity for the (2) 25.3 (3) 355 (95.9) 2015 Mean absolute change from combination with 250 Phe-508del CFTR (3) 25.4 baseline Duration of follow- mg of ivacaftor every mutation Within-group, percentage Discontinuation d/t AE, n TRAFFIC and up: 24 weeks 12 hours (n=368) • Age of 12 years or Sex points (p-value) (%) TRANSPORT - older Female, n (%) (1) 3.0 (p < 0.001) (1) 14 (3.8) Homozygous F508d 187 centers in North (2) LUM/IVA: 400 mg of • Percentage of (1) 182 (49.5) (2) 2.5 (p < 0.001) (2) 17 (4.6) America, Australia, lumacaftor every 12 predicted FEV1 at the (2) 182 (49.3) (3) -0.32 (p =0.40) (3) 6 (1.6) Good and Europe hours in combination time of screening that (3) 181 (48.8) with 250 mg of was 40- 90% of the Between-group difference, ≥ One SAE, n (%) Enrollment between ivacaftor every 12 predicted normal percentage points (95% CI) ppFEV1 (1) 84 (22.8) April 2013 and April hours (n=369) values (1) 3.3 (2.3 to 4.3) Mean, percentage (2) 64 (17.3) 2014 • Stable cystic fibrosis (2) 2.8 (1.8 to 3.8) points (3) 106 (28.6) disease (3) NA ©Institute for Clinical and Economic Review, 2018 Page 156 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Study Design and Author & Year of Duration of Publication, Interventions (n) & Inclusion and Exclusion Follow-up, Patient Characteristics Outcomes Harms (Trial), Dosing Schedule Criteria (Sites & geographical Quality Rating location) All data reported are (3) Placebo: (1) 60.8 pooled groups of two Lumacaftor-matched Exclusion (2) 60.5 ppFEV1 Infective PEx of CF, n (%) studies – TRAFFIC placebo every 12 hours • Any comorbidity that (3) 60.4 Mean relative change from (1) 145 (39.3) and TRANSPORT in combination with increases risk in the baseline (2) 132 (35.8) Within-group, % (p-value) ivacaftor-matched study (cirrhosis, BMI (3) 182 (49.2) (1) 5.4 (p < 0.001) placebo every 12 hours Torsades de Pointes) Mean, kg/m2 (2) 4.6 (p < 0.001) (n=371) • Abnormal lab values (1) 21.0 Cough, n (%) (3) -0.17 (p =0.80) • Respiratory event (2) 21.5 (1) 121 (32.8) within 4 weeks of first (3) 21.0 (2) 104 (28.2) Between-group difference, day on drug (3) 148 (40.0) % (95% CI) • Colonization with (1) 5.6 (3.8 to 7.3) certain bacteria (2) 4.8 (3.0 to 6.6) Headache, n (%) • Prolonged QT interval (3) NA (1) 58 (15.7) • History of transplant (2) 58 (15.7) • Using strong inhibitors, BMI (3) 58 (15.7) moderate inducers, or Mean absolute change from strong inducers of baseline, kg/m2 (p-value) Within group CYP3A within 14 days (1) 0.41 (p<0.001) of first day on drug (2) 0.37 (p<0.001) • History of cataract or (3) 0.13 (p=0.007) lens opacity or evidence of cataract or CFQ-R Respiratory domain lens opacity Mean absolute change from determined to be baseline, points (p-value) clinically significant (1) 4.9 (p<0.001) (2) 4.1 (p<0.001) (3) 1.9 (p=0.02) PEx No. of events; Rate Ratio (95%CI) (1) 173; 0.70 (0.56 to 0.87) ©Institute for Clinical and Economic Review, 2018 Page 157 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Study Design and Author & Year of Duration of Publication, Interventions (n) & Inclusion and Exclusion Follow-up, Patient Characteristics Outcomes Harms (Trial), Dosing Schedule Criteria (Sites & geographical Quality Rating location) (2) 152; 0.61 (0.49 to 0.76) (3) 251; NA Elborn 23 See Wainwright See Wainwright See Wainwright Data reported are Pooled Analysis Pooled Analysis stratified – see Study < 40% vs. ≥40% ppFEV1 < 40% vs. ≥40% ppFEV1 Lancet Resp Med Prespecified design and follow-up Lumacaftor 400mg q 12 hrs/ Both doses (600mg & subgroup analyses of Ivacaftor 250 mg q 12hrs 400mg) 2016 pooled efficacy and Age safety data by lung Median, years (range) ppFEV1 Any AE, n (%) function. (1) 23.0 (12–64) Mean absolute change from (1) 350 (96) TRAFFIC and (2) 27.0 (13–44) baseline vs. placebo, (2) 52 (98) TRANSPORT Subgroup For Demographics (3) 23.0 (12–57) percentage points (95% CI) (3) 649 (96) analysis data: (4) 26.0 (12–57) (1) reference (1) Placebo n=371 (5) 18.5 (12–53) (2) 3.3. (0.2 to 6.4) Infective PEx of CF, n (%) (<40%ppFEV1=2 (3) 2.8 (1.7 to 3.8) (1) 182 (50) 8) Sex (2) 27 (51) Female, n (%) ppFEV1 (3) 248 (37) LUM 400 mg q12 Iva (1) 181 (49%) Mean (least-squares) 250 mg q12, n=731 (2) 31 (58%) relative change from Cough, n (%) (2) Baseline ppFEV1 (3) 331 (49%) baseline vs placebo, % (95% (1) 147 (40) <40% n=53 (4) 269 (51%) CI) (2) 21 (40) (3) Baseline ppFEV1 (5) 93 (46%) (1) reference (3) 203 (30) ≥40% n=687 • (2) 9.1 (0.7 to 17.4) (4) Screening ppFEV1 (3) 4.5 (2.7 to 6.3) Headache, n (%) ppFEV1 <70% Mean, percentage (1) 57 (16) n=527 points (range) BMI (2) 10 (19) (5) Screening (1) 60.4 (33.9–99.8) Least-squares mean vs. (3) 103 (15) ppFEV1 (2) 37.2 (31.1-39.9) placebo, kg/m2 (95% CI) • ≥70% n=204 (3) 62.5 (40.0-96.5) (1) reference • (4) 54.0 (31.1-69.8) (2) 0.3 (-0.2 to 0.8) ©Institute for Clinical and Economic Review, 2018 Page 158 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Study Design and Author & Year of Duration of Publication, Interventions (n) & Inclusion and Exclusion Follow-up, Patient Characteristics Outcomes Harms (Trial), Dosing Schedule Criteria (Sites & geographical Quality Rating location) (5) 77.9 (70.0–96.5) (3) 0.2 (0.1 to 0.4) For Results at 24 • weeks: BMI CFQ-R Respiratory domain Mean, kg/m2 (SD) Least-squares mean vs. (1) Placebo (1) 21.0 (2.9) placebo, points (95% CI) (2) LUM 400 mg (2) 20.9 (3.4) (1) reference q12 Iva 250 mg (3) 21.3 (3.0) (2) −4.2 (−12.0 to 3.7) q12, (4) 21.2 (2.9) (3) 2.9 (0.5 to 5.3) • FEV1<40% (5) 21.4 (3.3) (3) LUM 400 mg q PEx 12 Iva 250 mg q Event rate ratio (95%CI) 12, FEV1≥40% (1) reference (2) 0.59 (0.33 to 1.05) (3) 0.61 (0.48 to 0.77) PEx No. events requiring IV antibiotics, rate ratio (95%CI) (1) Reference (2) 0.56 (0.27 to 1.17) (3) 0.42 (0.30 to 0.58) PEx No. events requiring hospitalization, rate ratio (95%CI) (1) reference (2) 0.67 (0.27 to 1.65) (3) 0.36 (0.23 to 0.54) Konstan 19 Phase 3, multicenter, N=1030 Inclusion Age Pooled Analysis, least- Death, n (%) parallel group, open- • Confirmed diagnosis of Mean, years (SD) squares means (1) 2 (0.5) Lancet Resp Med label trial. (1) LUM/IVA: continued CF (1) 25.1 (9.3) (2) 1 (0.5) ppFEV1 400 mg of lumacaftor • Homozygosity for the (2) 24.9 (10.1) 2017 every 12 hours in F508del-CFTR mutation • Age of 12 years or older ©Institute for Clinical and Economic Review, 2018 Page 159 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Study Design and Author & Year of Duration of Publication, Interventions (n) & Inclusion and Exclusion Follow-up, Patient Characteristics Outcomes Harms (Trial), Dosing Schedule Criteria (Sites & geographical Quality Rating location) PROGRESS - Patients who combination with 250 Sex Mean absolute change from Discontinuations for two Homozygous F508d completed TRAFFIC mg of ivacaftor every Exclusion Female, n (%) baseline, percentage points groups, n (%) or TRANSPORT 12 hours (n=340) • Any comorbidity or lab (1) 164 (48) (95% CI) – Wang-Hankinson 170 (33) abnormality that may 72 weeks participated in the (2) 86 (49) (1) 0.5 (−0.4 to 1.5) study in 191 sites in (2) LUM/IVA: Placebo confound study results Discontinuation d/t AE, n (2) 1.5 (0.2 to 2.9) 15 countries transitioned to 400 mg or increase potential ppFEV1 (%) lumacaftor every 12 harm to participant Mean, percentage 38 (7) 96 weeks Duration of follow- hours in combination • History of drug points (SD) (1) 0.5 (−0.7 to 1.6) up: 96 weeks; with ivacaftor 250 mg intolerance in the prior (1) 60.4 (14.2) (2) 0.8 (−0.8 to 2.3) Infective PEx of CF, % however, main every 12 hours (n=176) study (2) 60.2 (13.8) 65 efficacy outcomes • Pregnancy or breast- ppFEV1 reported at 72 weeks At 72 weeks (primary feeding BMI Mean absolute change from Cough, % efficacy), those on • History of poor Mean, kg/m2 (SD) baseline, percentage points 44 LUM/IVA in compliance with study (1) 21.4 (2.9) (95% CI) – GLI Traffic/Transport had drug or procedures (2) 20.9 (2.8) 72 weeks Increased sputum, % received 96 weeks of (1) 0.9 (0.0 to 1.9) • Participation in an 22 active drug. (2) 1.9 (0.6 to 3.2) investigational drug Pseudomonas positive, trial no. Hemoptysis, % 96 weeks (1) 261 20 (1) 1.1 (0.0 to 2.2) (2) 126 (2) 1.1 (−0.5 to 2.6) ppFEV1 Mean relative change from baseline, % (95% CI) At 72 weeks (1) 1.4 (−0.3 to 3.2) (2) 2.6 (0.2 to 5.0) At 96 weeks (1) 1.2 (−0.8 to 3.3) (2) 1.1 (−1.7 to 3.9) ©Institute for Clinical and Economic Review, 2018 Page 160 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Study Design and Author & Year of Duration of Publication, Interventions (n) & Inclusion and Exclusion Follow-up, Patient Characteristics Outcomes Harms (Trial), Dosing Schedule Criteria (Sites & geographical Quality Rating location) BMI Mean absolute change from baseline, kg/m2 At 72 weeks (1) 0.69 (0.56 to 0.81) (2) 0.62 (0.45 to 0.79) At 96 weeks (1) 0.96 (0.81 to 1.11) (2) 0.76 (0.56 to 0.97) CFQ-R Respiratory domain Mean absolute change from baseline, points (95% CI) At 72 weeks (1) 5.7 (3.7 to 7.5) (2) 3.3 (0.7 to 5.9) At 96 weeks (1) 3.5 (1.3 to 5.8) (2) 0.5 (−2.7 to 3.6) PEx, No. of events per patient- year (95%CI) (1) 0.65 (0.56 to 0.75) (2) 0.69 (0.56 to 0.85) PEx, No. of events requiring hospital admission per patient-year (95%CI) (1) 0.24 (0.19 to 0.29) (2) 0.30 (0.22 to 0.40) ©Institute for Clinical and Economic Review, 2018 Page 161 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Study Design and Author & Year of Duration of Publication, Interventions (n) & Inclusion and Exclusion Follow-up, Patient Characteristics Outcomes Harms (Trial), Dosing Schedule Criteria (Sites & geographical Quality Rating location) PEx, No. of events requiring intravenous antibiotics per patient-year (95%CI) (1) 0.32 (0.26 to 0.38) (2) 0.37 (0.29 to 0.49) Konstan 109 See Konstan 2017 N=176 See Konstan 2017 See Konstan 2017 ppFEV1 Most commonly reported Mean (least-squares) AEs: Pediatric Pulmonology Interim analysis of (1) LUM/IVA: 400 mg of relative change from PROGRESS at 24 lumacaftor every 12 baseline, percent (SE); p- Infective PEx of CF (48%) 2015 value weeks hours in combination 24 weeks of PROGRESS* Cough (39%) with 250 mg of (1) 2.6 (0.47); p<0.0001 Abstract ivacaftor every 12 (2) 3.5 (0.64); p<0.0001 Headache (17%) hours (n=340) BMI Dyspnea (17%) Mean (least-squares) (2) LUM/IVA: Placebo absolute change from Abnormal respiration baseline, kg/m2 (SE); p- (14%) transitioned to 400 mg lumacaftor every 12 value 24 weeks of PROGRESS* hours in combination (1) 0.56 (0.06); p<0.0001 with ivacaftor 250 mg (2) 0.37 (0.08); p<0.0001 every 12 hours (n=176) CFQ-R Respiratory domain Mean absolute change from baseline, points (SE); p- value 24 weeks of PROGRESS* (1) 6.3 (0.85); p<0.0001 (2) 5.1 (1.17); p<0.0001 PEx Event rate per year (95%CI) (1) 0.6 (0.5 to 0.8) ©Institute for Clinical and Economic Review, 2018 Page 162 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Study Design and Author & Year of Duration of Publication, Interventions (n) & Inclusion and Exclusion Follow-up, Patient Characteristics Outcomes Harms (Trial), Dosing Schedule Criteria (Sites & geographical Quality Rating location) (2) 0.6 (0.5 to 0.8) *Interim analysis McColley 110 See Wainwright Stratified analysis by: See Wainwright 2015 See Wainwright 2015 Rate Ratio (95% CI), drug vs. NA 2015 placebo Pediatric Pulmonology • ≤0% or PEx Post hoc analysis • >0% absolute 2015 ≤0% absolute improvement: TRAFFIC and improvement in 0.74 (0.55 to 0.99) TRANSPORT ppFEV1 Abstract evaluating the >0% absolute improvement: association between AND 0.53 (0.40 to 0.69) changes in percent predicted FEV1 and <5% relative improvement: • ≥5 or 0.62 (0.47 to 0.80) PE rates • <5% relative improvement in ≥5% relative improvement: ppFEV1 from baseline 0.60 (0.44 to 0.82) to Day 15 PEx requiring • hospitalization ≤0% absolute improvement: 0.40 (0.23 to 0.69) >0% absolute improvement: 0.38 (0.24 to 0.59) <5% relative improvement: 0.31 (0.19 to 0.51) ≥5% relative improvement: 0.50 (0.31 to 0.82) PEx requiring antibiotics ≤0% absolute improvement: ©Institute for Clinical and Economic Review, 2018 Page 163 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Study Design and Author & Year of Duration of Publication, Interventions (n) & Inclusion and Exclusion Follow-up, Patient Characteristics Outcomes Harms (Trial), Dosing Schedule Criteria (Sites & geographical Quality Rating location) 0.49 (0.33 to 0.74) >0% absolute improvement: 0.40 (0.28 to 0.58) <5% relative improvement: 0.37 (0.25 to 0.54) ≥5% relative improvement: 0.54 (0.37 to 0.80) Taylor-Cousar75 Open-label N=46 Inclusion Mean age, years Primary endpoint: safety Any AE, n (%): 43 (93) prospective study of • Confirmed diagnosis of (range) and tolerability Journal of Cystic LUM/IVA in patients LUM/IVA 400 mg q 12 CF 32.1 (17 to 56) AE leading to treatment Secondary outcomes: discontinuation: 8 (17) Fibrosis homozygous for hours with IVA 250 mg • Homozygosity for the Sex: Male, n (%) F508del with q 12 hours (n=28) F508del-CFTR mutation 30 (65) Mean absolute change in Serious AE: 18 (39) 2017 ppFEV1<40% • Age of 12 years or older ppFEV1 (least-squares) from ½ dose necessary for • ppFEV1<40%, adjusted ppFEV1 baseline (95% CI): AE leading to death: 1 (2) Six centers in United 39% of patients at start for age, gender and Mean, percentage Day 15: -1.7pp (-3.2 to -0.1) States of study (n=18) height points (range) Week 24: -0.4pp (-1.9 to AE with incidence >10%: 29.1 (18.3 to 42.0) 1.1) Infective PE: 27 (59) Duration of follow- Exclusion Respiration abnormal: 26 up: 24 weeks • Current use of invasive BMI (57) mechanical ventilation Mean, kg/m2 (range) Mean absolute change in Cough 21 (46) 21.4 (15.7 to 28.5) CFQ-R respirator domain Dyspnea 20 (43) • Any comorbidity that score (LS) from bassline may confound study Documentation of (95% CI): results or increase being on lung Week 24: 2.5 (-1.0 to 5.9) potential harm to transplant list at participant screening, n BMI change from baseline, • Abnormal liver or renal Yes: 2 mean (SD): function No: 25 Week 24: 0.29 kg/m2 (0.17) Unavailable: 19 Also measured: Annualized all-cause hospitalization event rate in ©Institute for Clinical and Economic Review, 2018 Page 164 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Study Design and Author & Year of Duration of Publication, Interventions (n) & Inclusion and Exclusion Follow-up, Patient Characteristics Outcomes Harms (Trial), Dosing Schedule Criteria (Sites & geographical Quality Rating location) the 24 weeks prior to study compared with the 24 weeks on LUM/IVA 1.15 events/year compared with 2.78 events/year prior to study start IV antibiotic duration (days) in the 24 weeks prior to study compared with the 24 weeks on study drug. Found LUM/IVA led to decreased normalized total duration (11.38 days) vs. prior 24 weeks (19.89 days). Mean difference of -8.52 (3.67), p=0.0369 Jennings 24 Retrospective N=116 Exclusion: Homozygous F508del ppFEV1 Reported Side Effects, n observational study, • Previous exposure to 100% Mean change from baseline, (%) Annals ATS pre/post treatment (1) Pre-LUM/IVA LUM/IVA percentage points (range) 46 (39.7) Sex 0.11 (−39 to 20) with LUM/IVA • Participation in a 2017 (2) Post-LUM/IVA M:F Discontinuation clinical trial 54:62 20 (17.2) One center: Johns Hopkins Age Chest Mean, years (range) tightness/discomfort 23 Duration of follow- 24.7 (12-59) (19.8) up: 11 months ppFEV1 Dyspnea Subgroup by age and Mean, percentage 12 (10.3) points (range) FEV1 67.4 (20-115) Increased cough/congestion 10 (8.6) ©Institute for Clinical and Economic Review, 2018 Page 165 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Study Design and Author & Year of Duration of Publication, Interventions (n) & Inclusion and Exclusion Follow-up, Patient Characteristics Outcomes Harms (Trial), Dosing Schedule Criteria (Sites & geographical Quality Rating location) CF-related diabetes (CFRD), No. (%) Diarrhea 26 (22.4) 5 (4.3) Pseudomonas positive Nausea No. (%) 3 (2.6) 71 (61.2) Decreased appetite MRSA positive 2 (1.7) No. (%) 35 (30.2) Rash 2 (1.7) B. cepacia complex positive, No. (%) Discontinuation by 8 (6.9) subgroup, adjusted odds ratio (95% CI): Proton-pump inhibitor use, No. (%) Age: 1.00 (0.95 to 1.06) 51 (44) Female: 3.12 (1.04 t0 9.34) Anti-depressant use, Baseline ppFEV1 <40%: No. (%) 2.35 (0.74 to 7.50) 21 (18.1) Azole use, No. (%) 6 (5.2) Ratjen 17 Phase III, N=206 Inclusion: Mean age, years (SD) LCI Any AE, n (%) randomized, double- • Age 6-11 (1) 8.7 (1.6) Mean (least-squares) (1) 98 (95) Lancet Resp Med blind, placebo- (1) LUM/IVA: • Confirmed diagnosis of (2) 8.9 (1.6) absolute change from (2) 98 (97) Lumacaftor 200 mg and baseline, score (95% CI)* controlled, cystic fibrosis 2017 Sex 24 weeks Any SAE, n (%) multinational trial ivacaftor 250 mg q 12 • Weight at least 15 kg Female, n (%) (1) −1.0 (−1.3 to −0.8) (1) 13 (13) (n=104) • ppFEV1≥70% and lung (1) 63 (61) (2) 0.1 (−0.2 to 0.3) (2) 11 (11) Homozygous F508del Nine countries: USA, clearance index (LCI) ≥ (2) 58 (57) Difference: −1∙1 (−1.4 to Australia, Belgium, (2) Placebo (n=102) −0.8) p<0∙0001 Study discontinuation, n 7.5 Canada, Denmark, ppFEV1 (%) • homozygous F508del ©Institute for Clinical and Economic Review, 2018 Page 166 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Study Design and Author & Year of Duration of Publication, Interventions (n) & Inclusion and Exclusion Follow-up, Patient Characteristics Outcomes Harms (Trial), Dosing Schedule Criteria (Sites & geographical Quality Rating location) France, Germany, Mean, percentage BMI (1) 1 (1)* respiration Sweden, and the UK • Exclusion: points (SD) Mean (least-squares) abnormal • Any comorbidity or lab (1) 88.8 (13.7) absolute change from (2) 0 (0) Duration of follow- (2) 90.7 (10.8) baseline, kg/m2 (95% CI) abnormality that may up: 24 weeks 24 weeks Elevated liver enzymes of confound study results Weight (1) 0.4 (0.3 to 0.5) clinical significance, n (%): or increase potential Mean, kg (SD) (2) 0.3 (0.1 to 0.4) (1) 13 (13) Enrollment: July 23, harm to participant (1) 29.4 (6.5) Difference: 0∙1 (−0.1 to 0.3) (2) 8 (8) 2015 to Sept 20, 2016 • Acute respiratory tract (2) 30.2 (6.8) p=0.2522 infection, PE, or Cough, n (%) LCI ppFEV1 (1) 46 (45) changes in therapy for Mean (SD) Mean (least-squares) (2) 47 (47) pulmonary disease (1) 10.3 (2.4) absolute change from within 28 days of (2) 10.3 (2.2) baseline, percentage points Infective PEx of CF, n (%) treatment initiation (95% CI) (1) 20 (19) • History of solid organ 24 weeks (2) 18 (18) transplant (1) 1.1 (−0.4 to 2.6) (2) −1.3 (−2.8 to 0.2) Oropharyngeal pain, n Difference: 2.4 (0.4 to 4.4) (%) p=0.0182 (1) 15 (15) (2) 10 (10) CFQ-R • Mean (least-squares) Pyrexia, n (%) absolute change from (1) 15 (15) baseline, points (95% CI) (2) 20 (20) 24 weeks (1) 5.5 (3.4 to 7.6) (2) 3.0 (1.0 to 5.0) Acute change in ppFEV1 Difference: 2.5 (−0.1 to 5.1) immediately after study p=0.0628 drug administration @ day 1, mean absolute *Decreases in LCI reflect change (SD) improvements in lung < 2 hours post-dose function while increases in (1) -5.5 (8.2) LCI indicate lung function (2) -0.1 (5.1) ©Institute for Clinical and Economic Review, 2018 Page 167 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Study Design and Author & Year of Duration of Publication, Interventions (n) & Inclusion and Exclusion Follow-up, Patient Characteristics Outcomes Harms (Trial), Dosing Schedule Criteria (Sites & geographical Quality Rating location) decline 4-6 hours post-dose (1) -7.7 (7.3) (2) -1.4 (7.1) 24 hours post-dose (1) -4.1 (10.1) (2) -1.7 (6.8) Milla 20 Open-label, phase III N=58 (54 completed 24 Inclusion: Mean age, years (SD) ppFEV1 All adverse events n (%): weeks) • Age 6-11 at screening 9.1 (1.53) Mean (least-squares) 55 (94.8) Am J Respir Crit Care Duration of follow- • Confirmed diagnosis of absolute change from Med up: 24 weeks active Lumacaftor 200 mg q Sex Female, n (%) baseline, percentage points Serious adverse event n cystic fibrosis 31 (53.4) (95% CI) (%): med with 2 week 12 hours with 250 mg • ppFEV1≥40% 24 weeks 4 (6.9) 2017 washout of ivacaftor q 12 hours • Homozygous F508del ppFEV1 2.5 (-0.2 to 5.2) • Stable disease Mean, percentage Interruption of treatment Homozygous F508del points (SD) BMI due to an adverse event, • Exclusion: 91.4 (13.7) Mean (least-squares) n (%): 6 (10.3) • Any comorbidity or lab absolute change from abnormality that may Weight baseline, kg/m2 (95% CI) Discontinuation due to an confound study results Mean, kg (SD) 24 weeks adverse event, n (%): or increase potential 31.5 (6.1) 0.64 (0.46 to 0.83) 2 (3.4) harm to participant Weight-for-age z-score BMI-for-age z-score Elevated liver enzymes of Mean (SD) Mean (least-squares) clinical significance, n (%): -0.03 (1.03) absolute change from 11 (19.3) baseline (95% CI) BMI-for-age z-score 24 weeks Serious events, n (%): Mean (SD) 0.15 (0.08 to 0.22) Infective PEx: 2 (3.4) 0.01 (0.90) Ileus: 1 (1.7) Weight-for-age Z score Elevated liver Mean (least-squares) transaminase levels: 1 absolute change from (1.7) baseline (95% CI) 24 weeks Respiratory events n (%): 0.13 (0.07 to 0.19) Dyspnea: 1 (1.7) ©Institute for Clinical and Economic Review, 2018 Page 168 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Study Design and Author & Year of Duration of Publication, Interventions (n) & Inclusion and Exclusion Follow-up, Patient Characteristics Outcomes Harms (Trial), Dosing Schedule Criteria (Sites & geographical Quality Rating location) Respiration abnormal: 1 CFQ-R (1.7) Mean (least-squares) Wheezing: 2 (3.4) absolute change from baseline, points (95% CI) Common adverse events, 24 weeks n (%): 5.4 (1.4 to 9.4) Cough: 29 (50) Nasal congestion: 12 LCI (exploratory endpoint; (20.7) n=30) Infective PEx: 12 (20.7) Mean (least-squares) Headache: 12 (20.7) absolute change from baseline, score (95% CI)* Cataract, n (%): 1 (1.7) 24 weeks -0.88 (-1.40 to -0.37) *Decreases in LCI reflect improvements in lung function while increases in LCI indicate lung function decline Boyle 77 Double-blind, N=35 Inclusion: Only LUM/IVA group ppFEV1 Any AE, n (%) placebo-controlled, • Age 18+ baseline provided - Mean (least-squares) (1) 10 (91) Lancet Respiratory phase 2 trial with 3 Three cohorts: only • Confirmed diagnosis of placebo pooled (mixed absolute change from (2) 20 (74) reporting on cohort 3, baseline, percentage points • cohorts cystic fibrosis hetero and 2014 (95%CI) • SAE, n subjects (%) days 28-56 (combo) • ppFEV1≥40% homozygous) (1) 6.1 (2.0 to 10.2) (1) 1 (9); 2 events (1 PE) 24 centers in • At least one F508del (2) −1.6 (−4.2 to 1.1) (2) 4 (15); 6 events (4 PE) Homozygous F508del Australia, Belgium, (1) LUM/IVA: 400 mg Age (we only report on two Difference: 7.7 (2.7 to 12.6) Germany, New lumacaftor q 12 hours copies) Mean, years (SD) PEx of CF, n (%) Zealand or US with 250 mg ivacaftor • q Exclusion: (1) 25.5 (6.7) ppFEV1 (1) 2 (18) 12 hours (n=11) (2) 30.8 (12.4) Mean (least-squares) (2) 7 (26) • Any comorbidity or lab Enrollment: Oct 2010 relative change from abnormality that may baseline, percentage points to May 2012 Sex Discontinuation d/t AE, n confound study results (95%CI) ©Institute for Clinical and Economic Review, 2018 Page 169 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Study Design and Author & Year of Duration of Publication, Interventions (n) & Inclusion and Exclusion Follow-up, Patient Characteristics Outcomes Harms (Trial), Dosing Schedule Criteria (Sites & geographical Quality Rating location) Duration of follow- (2) Placebo (n=24; or increase potential Female, n (%) (1) 8.2 (1.8 to 14.7) 1/15 up: 28 days pooled across cohort 2 harm to participant (1) 5 (45) (2) −2.1 (−6.3 to 2.2) and 3) • PE or change in (2) 9 (33) Cough, n (%) treatment within 14 (1) 3 (27) days first dose BMI (2) 6 (22) • Prolonged QT/QTc Mean, kg/m2 (SD) interval (1) 21.7 (2.9) Headache, n (%) • Solid organ transplant (2) 22.6 (2.7) (1) 2 (18) • Used inhibitors or (2) 5 (19) inducers of CYP3A4 Weight • In another trial in last 3 Mean, kg (SD) (1) 60.7 (10.3) months (2) 66.0 (10.6) Ivacaftor Ramsey 8 Phase 3, randomized, N=161 Inclusion Age ppFEV1 Any AE, n (%) double-blind, • 12 years of age or older Mean, years (range) Mean absolute change from (1) 82 (99) NEJM placebo-controlled (1) IVA: 150 mg of • Confirmed CF diagnosis (1) 26.2 (12-53) baseline, percentage points (2) 78 (100) ivacaftor twice daily • G551D mutation on at (2) 24.7 (12-53) (95% CI) international trial 2011 least one CFTR allele (1) 10.1 SAE, n (%) (n=83) • FEV1 between 40-90% Sex (2) −0.4 (1) 20 (24) Duration of follow- Female, n (%) Difference=10.5 (8.5 to 12.5) (2) 33 (42) STRIVE – G551D (2) Matched Placebo of predicted value for up: 48 weeks (1) 44 (53) persons of their age, sex, (n=78) (2) 40 (51) PEx Interruption d/t AE, n (%) and height Good No. of events (rate per (1) 11 (13) Exclusion ppFEV1 subject) (2) 5 (6) • History of illness or Mean, percentage (1) 47 (0.59) condition that may points (2) 99 (1.38) Discontinuation d/t AE, n confound results or (1) 63.5 (%) pose safety risk (2) 63.7 PEx (1) 1 (1) • Acute respiratory No. of subjects (2) 4 (5) Weight (1) 28 infection, PE, or Mean, kg (2) 44 PEx, n (%) changes in therapy for (1) 61.7 RR (95% CI): 0.43 (0.27 to (1) 11 (13) pulmonary disease (2) 61.2 0.68) (2) 26 (33) ©Institute for Clinical and Economic Review, 2018 Page 170 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Study Design and Author & Year of Duration of Publication, Interventions (n) & Inclusion and Exclusion Follow-up, Patient Characteristics Outcomes Harms (Trial), Dosing Schedule Criteria (Sites & geographical Quality Rating location) within 4 weeks of enrollment BMI Weight Hemoptysis, n (%) • Abnormal liver and Mean, kg/m2 Mean change from baseline, (1) 1 (1) renal function (1) 21.7 kg (95% CI) (2) 4 (5) • History of solid organ (2) 21.9 (1) 3.1 or hematological (2) 0.4 transplant *CFQ-R Respiratory Difference=2.7 (1.3 to 4.1) • Pregnancy, breast- domain feeding, or planning (1) NR CFQ-R Respiratory domain pregnancy (2) NR Absolute change from • On-going participation baseline, points in another clinical trial * Scores on (CFQ-R) (1) 5.9 • Using inhaled range from 0-100, (2) −2.7 hypertonic saline higher scores indicating Difference=8.6 treatment a • Concomitant use of higher patient- CPY3A4 inhibitors or reported QoL with inducers regard to respiratory status. Davies 9 Phase 3, randomized, N=52 Inclusion Age ppFEV1 Any AE, n (%) double-blind, • 6-11 years of age Mean, years (range) Mean adjusted* change (1) 26 (100) Am J Respir Care Med placebo-controlled (1) IVA: 150 mg of • Confirmed CF diagnosis (1) 8.9 (6-12) from baseline, percentage (2)25 (96.2) trial ivacaftor twice daily • G551D mutation on at (2) 8.9 (6-12) points (95% CI) 2013 (n=26) least one CFTR allele (1) 10.7 SAE, n (%) Duration of follow- • FEV1 of 40-105% of the Sex (2) 0.7 (1) 5 (19) ENVISION – G551D up: (2) Matched Placebo predicted value for Female, n (%) Difference= 10.0 (4.5 to (2) 6 (23) 48 weeks (n=26) persons of their age, (1) 17 (65) 15.5) Good sex, and height (2) 10 (38) Interruption d/t AE, n (%) • Body weight ≥15kg Weight (1) 1 (4) ppFEV1 Mean adjusted* change (2) 3 (12) Exclusion Mean, percentage from baseline, kg (95% CI) • History of illness or points (range) (1) 5.9 Discontinuation d/t AE, n condition that may (1) 84.7 (52.4-133.8) (2) 3.1 (%) confound results or (2) 83.7 (44.0-116.3) Difference=2.8 (1.3 to 4.2) (1) 0 pose safety risk (2) 1 (4) Weight CFQ-R Respiratory domain ©Institute for Clinical and Economic Review, 2018 Page 171 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Study Design and Author & Year of Duration of Publication, Interventions (n) & Inclusion and Exclusion Follow-up, Patient Characteristics Outcomes Harms (Trial), Dosing Schedule Criteria (Sites & geographical Quality Rating location) • Acute respiratory Mean, kg (range) Mean adjusted* change PEx of CF, n (%) infection, PE, or (1) 31.8 (18.8-62.6) from baseline, (95% CI) (1) 8 (31) changes in therapy for (2) 30.0 (17.8-46.3) (1) 6.1 (2) 8 (31) pulmonary disease (2) 1.0 within 4 weeks of BMI Difference=5.1 (−1.6 to 11.8) Cough, n (%) enrollment Mean, kg/m2 (range) (1) 13 (50) • Abnormal liver and (1) 17.1 (14.2-26.0) PExs† (2) 19 (73) renal function (2) 16.8 (13.8-22.1) No. reported • History of solid organ (1) 4 Headache, n (%) or hematological CFQ-R Respiratory (2) 3 (1) 7 (27) transplant domain (2) 4 (15) • On-going participation Mean, points * Least squares mean and in another clinical trial (1) 78 mixed-effects model for • Using inhaled (2) 80 repeated measures. hypertonic saline Adjusted for all available. treatment • Concomitant use of † Protocol-defined CPY3A4 inhibitors or exacerbations. Additional inducers exacerbations were reported as AEs, but difference in definitions were not available. McKone 15 Phase 3, open-label N=192 Inclusion Age ppFEV1 Any AE, n (%) extension • G551D mutation on at Mean, years (SD) Mean absolute change from STRIVE and ENVISION Lancet Respir Med (1) IVA: 150 mg of least one CFTR allele (1) baseline, percentage points placebo groups: Duration of follow- ivacaftor twice daily • Had completed either a.) 27.7 (9.8) (SD) Week 1-48: 82 (92%) 2014 up: a.) STRIVE IVA (n=77) STRIVE or ENVISION b.) 26.0 (9.6) (1) Week 48-96: 81 (92%) 96 weeks b.) STRIVE placebo study c.) 9.8 (1.9) a.) 9.4 (10.8) PERSIST – G551D (n=67) • Negative urine d.) 9.8 (1.8) b.) 9.5 (11.2) STRIVE and ENVISION c.) ENVISION IVA pregnancy test for c.) 10.3 (12.4) ivacaftor groups: Good (n=26) women of child- Sex d.) 10.5 (11.5) Week 48-96: 100 (97%) d.) ENVISION placebo bearing potential had Female, n (%) Week 96-144: 95 (92%) (n=22) • Participants of child- (1) BMI bearing potential and a.) 41 (53) Mean absolute change from SAE, n (%) Note: Groups a) and c) who are sexually active b.) 35 (52) baseline, kg/m2 (SD) All SAEs: 82 (43%) on IVA for 48 weeks must meet c.) 17 (65) (1) Week 1-48: 38 (20%) ©Institute for Clinical and Economic Review, 2018 Page 172 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Study Design and Author & Year of Duration of Publication, Interventions (n) & Inclusion and Exclusion Follow-up, Patient Characteristics Outcomes Harms (Trial), Dosing Schedule Criteria (Sites & geographical Quality Rating location) prior to PERSIST start, contraceptive d.) 9 (41) a.) 1.2 (2.2) Week 48-96: 44 (23%) then followed for requirements b.) 1.0 (1.6) additional 96 weeks on ppFEV1 c.) 0.30 (0.6) STRIVE and ENVISION ivacaftor (144 weeks Exclusion Mean, percentage d.) 0.37 (0.5) placebo groups: total); Groups b) and d) • History of illness or points (SD) Week 1-48: 15 (17%) on ivacaftor for 96 condition that may (1) Weight Week 48-96: 19 (21%) weeks of PERSIST after confound results or a.) 71.9 (18.5) Mean absolute change from 48 weeks of placebo in pose safety risk b.) 62.2 (18.7) baseline, kg (SD) STRIVE and ENVISION prior trial (96 weeks • History of study c.) 94.9 (14.5) (1) ivacaftor groups: total). treatment intolerance d.) 83.6 (17.4) a.) 4.1 (7.1) Week 48-96: 23 (22%) • Pregnancy, breast- b.) 3.0 (4.7) Week 96-144: 25 (24%) All patients in PERSIST feeding, or planning BMI c.) 14.8 (5.7) received ivacaftor pregnancy Mean, kg/m2 (SD) d.) 10.1 (4.1) Deaths, n (%) • Concomitant use of (1) (1) 2 CPY3A4 inhibitors or a.) 23.0 (4.0) CFQ-R Respiratory domain inducers b.) 21.9 (3.5) Mean absolute change from Discontinuation d/t AE, n c.) 18.6 (2.9) baseline, points (SD) (%) d.) 16.8 (2.2) (1) (1) 3 (2) a.) 6.8 (19.6) Weigh b.) 9.8 (16.2) PEx, no. of events (%) Mean, kg (SD) c.) 10.6 (18.9) (1) (1) d.) 10.8 (12.8) STRIVE and ENVISION a.) 66.0 (14.9) placebo groups: b.) 61.4 (13.1) Week 1-48: 30 (34%) c.) 37.9 (11.7) Week 48-96: 35 (39%) d.) 32.4 (8.9) STRIVE and ENVISION ivacaftor groups: Week 48-96: 46 (45%) Week 96-144: 46 (45%) Cough, n (%) (1) STRIVE and ENVISION placebo groups: Week 1-48: 27 (30%) ©Institute for Clinical and Economic Review, 2018 Page 173 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Study Design and Author & Year of Duration of Publication, Interventions (n) & Inclusion and Exclusion Follow-up, Patient Characteristics Outcomes Harms (Trial), Dosing Schedule Criteria (Sites & geographical Quality Rating location) Week 48-96: 16 (18%) STRIVE and ENVISION ivacaftor groups: Week 48-96: 32 (31%) Week 96-144: 27 (26%) Headache, n (%) (1) STRIVE and ENVISION placebo groups: Week 1-48: 11 (12%) Week 48-96: 7 (8%) STRIVE and ENVISION ivacaftor groups: Week 48-96: 14 (14%) Week 96-144: 17 (17%) De Boeck 10 Two-part, double N=39 Inclusion Age ppFEV1 Any AE, n (%) blind, randomized, • Confirmed diagnosis of Mean, years Mean absolute change* Ivacaftor: 28 (73.7) J Cyst Fibros controlled, crossover (1) IVA-Placebo: 150 mg CF (1) 23.8 from baseline, percentage Placebo: 31 (83.8) study of ivacaftor every 12 • A non-G51D gating (2) 21.7 points (95% CI) 2014 hours for 8 weeks mutation on at least (1) 7.5 SAE, n (%) Trial conducted in 12 followed by placebo one allele Sex (2) −3.2 Ivacaftor: 4 (10.5) KONNECTION – non- sites in the United q12 hours for 8 weeks • Age of 6 years or older Female, n (%) Difference=10.7 (7.3 to 14.1) Placebo: 7 (18.9) G551D gating States, France, and (n=20) (1) 7 (35.0) mutations Belgium. Exclusion (2) 10 (52.6) BMI Infective PEx of CF, n (%) (2) Placebo-IVA: Mean absolute change from (1) 9 (23.7) • History of illness or Fair Duration of follow- Placebo q12 hours for 8 ppFEV1 baseline, kg/m2 (95% CI) (2) 11 (29.7) condition that may up: 8 weeks weeks followed by Mean, percentage (1) 0.7 confound results or ivacaftor 150 mg q12 points (2) 0.02 Cough, n (%) pose safety risk hours for 8 weeks (1) 77.7 Difference=0.7 (0.34 to 0.99) (1) 6 (15.8) • Acute respiratory (n=19) (2) 79.1 (2) 7 (18.9) infection, PE, or CFQ-R respiratory domain changes in therapy for Both treatment groups BMI-for-age z-score Mean absolute change from Headache, n (%) pulmonary disease observed a 4-8 week Mean, score baseline, points (95% CI) (1) 5 (25) ©Institute for Clinical and Economic Review, 2018 Page 174 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Study Design and Author & Year of Duration of Publication, Interventions (n) & Inclusion and Exclusion Follow-up, Patient Characteristics Outcomes Harms (Trial), Dosing Schedule Criteria (Sites & geographical Quality Rating location) washout between within 4 weeks of (1) 0.50 (1) 8.9 (2) 7 (39) placebo and ivacaftor enrollment (2) 0.23 (2) −0.7 • History of solid organ Difference= 9.62 (4.5 to Discontinuation d/t AE, n or hematological 14.7) (%) transplant (1) 0 • On-going participation *Mixed-effects model for (2) 0 in another clinical trial repeated measures. within 30 days of screening • Using inhaled hypertonic saline treatment • Concomitant use of CPY3A4 inhibitors or inducers • Evidence of cataracts or lens opacity at screening Moss 11 Phase 3, multicenter, N=69 Inclusion Age ppFEV1 Protocol-defined PEx of placebo controlled, • 6 years of age or older Mean, years (SD) Mean absolute change from CF, n patients (%) NEJM double blind, parallel (1) IVA: 150 mg of • Confirmed diagnosis of (1) 29.2 (16.6) baseline, percentage points (1) 11 (32.3) group trial ivacaftor every 12 CF (2) 32.7 (17.4) (SD) (2) 13 (37) 2015 hours for 24 weeks • Arg117His-CFTR (1) 2.6 (1.2) Duration of follow- (n=34) mutation Sex (2) 0.5 (1.1) Protocol-defined PEx of KONDUCT – R117H up: 24 weeks • ppFEV1 of at least 40 Female, n (%) Difference=2.1 (95% CI:-1.13 CF, n events (event rate) (2) Placebo (n=35) (1) 19 (56.0) to 5.35) (1) 13 (0.249) Good (2) 20 (57.0) (2) 17 (0.295) Exclusion ppFEV1 • Gating mutation (1 or ppFEV1 Mean relative change from more) Mean, percentage baseline % (SD) SAE, n patients (%) • History of illness or points (SD) (1) 4.8 (1.9) (1) 4 (12) condition that may (1) 75.7 (19.3) (2) -0.2 (1.8) (2) 6 (17.5) confound results or (2) 70.2 (18.9) Difference= 5.0 (95% CI:- pose safety risk 0.24 to 10.31) ©Institute for Clinical and Economic Review, 2018 Page 175 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Study Design and Author & Year of Duration of Publication, Interventions (n) & Inclusion and Exclusion Follow-up, Patient Characteristics Outcomes Harms (Trial), Dosing Schedule Criteria (Sites & geographical Quality Rating location) • Acute respiratory BMI Needing admission to infection, PE, or Mean, kg (SD) BMI hospital, n patients changes in therapy for (1) 24.5 (6.3) Mean absolute change from (events) pulmonary disease (2) 23.1 (6.0) baseline, kg/m2 (SD) (1) 2 (2) within 4 weeks of (1) 0.49 (0.67) (2) 6 (7) enrollment CFQ-R Respiratory (2) 0.23 (0.65) • Abnormal liver function domain Difference=0.26 (95% CI:- at screening Mean, points (SD) 1.57 to 2.10) Needing intravenous • History of solid organ (1) 75.3 (20.1) antibiotic therapy, n or hematological (2) 66.4 (24.4) CFQ-R respiratory domain patients (events) transplant Mean absolute change from (1) 2 (2) • History or alcohol, baseline, points (SD) (2) 6 (8) medication, or illicit (1) 7.6 (2.2) drug abuse within 1 (2) -0.8 (2.2) year of study initiation Difference=8.4 (95% CI:2.17 • On-going participation to 14.61) in another clinical trial within 30 days of screening • Any “non-CF-related” illness within 2 weeks of study initiation • Concomitant use of CPY3A4 inhibitors or inducers Davies 13 Two-part, open-label, N=34 (Part B, only) Inclusion Part B reported (only) Part A results not reported Harms Part A not single-arm, phase 3 • Children aged 2–5 reported Lancet Respiratory study Part A: 4-day ivacaftor years Age Part B results: q 12 hours for • Weight 8 kg or more N (%) Harms Part B: 2016 15 hospitals in the pharmacokinetic and • Confirmed diagnosis of Age 2: 9 (26%) Mean weight-for age z- scores, mean (SD) – across Patients with any AE, n USA, UK, and Canada CF Age 3: 11 (32%) safety (two doses) - 50 (%) KIWI – gating • CFTR gating mutation Ages 4 and 5: 14 (41%) both doses mg if they weighed <14 on at least one allele Difference between 24 (1) 10 (100) mutations kg (n=4), and 75 mg if weeks and baseline: 0.2 (2) 23 (96) (Gly551Asp, Gly178Arg, Sex Ser549Asn, Ser549Arg, (0.3), p<0.001 ©Institute for Clinical and Economic Review, 2018 Page 176 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Study Design and Author & Year of Duration of Publication, Interventions (n) & Inclusion and Exclusion Follow-up, Patient Characteristics Outcomes Harms (Trial), Dosing Schedule Criteria (Sites & geographical Quality Rating location) Part B enrolled June they weighed ≥14 kg Gly551Ser, Gly970Arg, Female, n (%) SAE, no. events (no. pts, 28, 2013 to Sept 26, (n=5) Gly1244Glu, 6 (18) Mean BMI-for-age z-scores, %) 2013 Ser1251Asn, mean (SD) – across both (1) 4 (3, 30) Part B: 24-week safety Ser1255Pro, or doses Weight-for-age z-score (2) 3 (3, 13) Gly1349Asp) Difference between 24 (1) 50 mg (n=10) Mean, score (SD) weeks and baseline – 0.4 (2) 75 mg (n=24) -0.2 (0.8) SAE: Infective PEx of CF, n Exclusion (0.4), p<0.001 (%) • History of illness or Height-for-age z-score, condition that may Mean height-for-age z- (1) 1 (10) Mean, score (SD) scores, mean (SD) – across confound results or (2) 1 (4) -0.3 (0.8) both doses pose safety risk • Acute respiratory Difference between 24 AE: Infective PEx of CF, n infection, PE, or Mutations, n (%) weeks and baseline: -0.1 (0.3), p=0.84 (%) changes in therapy for G551D homozygous: (1) 1 (10) pulmonary disease 1(3) IRT, ng/mL (marker of (2) 4 (17) within 4 weeks of enrollment pancreatic stress), mean G551D heterozygous (SD) • Abnormal liver function with F508del: 26 (76) Cough, n (%) baseline to week 24 – (1) 4 (40) at screening 20.7 (24) p=0.002 • History of solid organ (2) 15 (63) or hematological G551D heterozygous not F508del: 5 (15) FEV1 not reported since transplant Vomiting, n (%) spirometry is not a reliable • Use of moderate or measure in very young (1) 3 (30) strong inducers or Ser549Asn children (2) 7 (29) inhibitors of CPY3A4 heterozygous: • Participation in a 2 (6) Hepatic enzyme clinical study of elevation, n (%) investigational or (1) 3 (30) marketed drug within (2) 2 (8) 30 days of screening ©Institute for Clinical and Economic Review, 2018 Page 177 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Study Design and Author & Year of Duration of Publication, Interventions (n) & Inclusion and Exclusion Follow-up, Patient Characteristics Outcomes Harms (Trial), Dosing Schedule Criteria (Sites & geographical Quality Rating location) Rowe 14 Longitudinal cohort, N=153 Inclusion: Age ppFEV1 Not reported single arm, • Male or female ≥ 6 Mean, years (SD) Absolute change from Am J Respir Care Med observational study (1) IVA: 150 mg of years of age at Visit 1 21 (11.3) baseline, percentage points (95% CI) ivacaftor twice daily • Must have a clinical 2014 1 mo: 6.7 (5.2 to 8.3) Duration of follow- diagnosis of cystic Age categories, n (%) 3 mo: 5.4 (4.0 to 6.7) up: 6 months fibrosis and the Ages 6-11:38 (25) 6 mo: 6.7 (4.9 to 8.5) GOAL following CFTR Ages 12-17: 33 (22) mutations: Ages 18-29: 52 (34) 6 mo, by age group (SD) • Included mutations: Ages 30+: 30 (20) Ages 6-11: 4.3 (11.1) Ages 12-17: 8.1 (8.2) G551D on at least 1 Ages 18+: 7.4 (10.7) allele with any known Sex or unknown mutations Female, n (%) Weight allowed on second 70 (46) Mean absolute change from allele; R117H on at baseline, kg (95%CI) least 1 allele with any ppFEV1 1 mo: 1.2 (0.9 to 1.4) known or unknown Mean, percentage 3 mo: 1.7 (1.3 to 2.1) points (SD) 6 mo: 2.5 (1.9 to 3.1) mutation on the second allele except 82.4 (25.9) 6 mo, by age group (SD) G551D; a non-G551D Ages 6-11: 3.7 (2.9) gating mutation on one By age Ages 12-17: 3.3 (3.3) allele: (G178R, S549N, Ages 6-11: 104.3 (16.2) Ages 18+: 1.5 (3.5) S549R, G551S, G970R, Ages 12-17: 91.2 (18.3) Ages 18+: 69.1 (23.3) BMI G1244E, S1251N, Mean absolute change from S1255P, G1349D) with baseline, kg/m2 (95% CI) any known or unknown Weight 1 mo: 0.4 (0.3 to 0.5) mutation on the Mean, kg (SD) 3 mo: 0.6 (0.4 to 0.7) second allele except Pooled not reported 6 mo: 0.8 (0.6 to 1.0) G551D or R117H By age 6 mo, by age group (SD) Ages 6-11: 30.6 (7.7) Ages 6-11: 1.1 (1.2) Exclusion Ages 12-17: 0.9 (1.0) NR Ages 12-17: 56.1 (15.7) Ages 18+: 0.5 (1.3) ©Institute for Clinical and Economic Review, 2018 Page 178 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Study Design and Author & Year of Duration of Publication, Interventions (n) & Inclusion and Exclusion Follow-up, Patient Characteristics Outcomes Harms (Trial), Dosing Schedule Criteria (Sites & geographical Quality Rating location) Ages 18+: 66.5 (13.7) CFQ-R Respiratory domain BMI Mean absolute change from baseline, (95% CI) Mean, kg/m2 (SD) 1 mo: 9.7 (7.1 to 12.4) 21.3 (4.5) 3 mo: 10.9 (8.1 to 13.7) 6 mo: 7.4 (4.1 to 10.7) By age Ages 6-11: 17.2 (2.4) 6 mo, by age group (SD) Ages 12-17: 21.0 (4.1) Ages 6-11: -0.7 (16.7) Ages 18+: 23.3 (4.1) Ages 12-17: 7.6 (14.6) Ages 18+: 11.7 (20.7) CFQ-R Respiratory domain Mean, points (SD) Pooled not reported By age Ages 6-11: 83.6 (12.2) Ages 12-17: (76.2) (15.6) Ages 18+: 62.4 (20.5) Flume 72 Post-hoc analysis of N=See STRIVE See STRIVE See STRIVE PEx See STRIVE participants who No. subjects (%) J Cyst Fibros experienced PExs (1) IVA: 150 mg of Characteristics of (1) 28 (33.7) from STRIVE ivacaftor twice daily (2) 44 (56.4) participants who had 2017 randomized clinical (n=83) trial (Ramsey, 2011) ≥1 protocol-defined No. of PExs (event rate) (2) Matched placebo PEx during study (1) 47 (0.589) STRIVE This study analyzed (n=78) (baseline data prior to (2) 99 (1.382) Good only those who PEx) No. of days per pt with reported a PEx event, mean (SD) during STRIVE (1) 13.54 (27.27) (2) 36.67 (49.54) ©Institute for Clinical and Economic Review, 2018 Page 179 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Study Design and Author & Year of Duration of Publication, Interventions (n) & Inclusion and Exclusion Follow-up, Patient Characteristics Outcomes Harms (Trial), Dosing Schedule Criteria (Sites & geographical Quality Rating location) Duration of follow- (1) n=2 up: (2) n=44 No. of pts treated with IV 48 weeks (STRIVE) antibiotics for PEx, n (%) (1) 15 (18.1) (2) 27 (34.6) Age Mean, years (SD) No. of events treated with (1) 26.9 (7.81) IV antibiotics, n (event rate) (2) 24.4 (9.29) (1) 28 (0.397) (2) 47 (0.711) Age, n (%) (1) No. subjects hospitalized for PEx (%) • <18: 4 (14.3) (1) 11 (13.3) • ≥18: 24 (85.7) (2) 23 (29.5) (2) No. of PExs treated by • <18: 11 (25.0) hospitalization (event rate) (1) 21 (0.311) • ≥18:33 (75.0) (2) 21 (0.489) Weight No. of subjects reporting Mean, kg (SD) increased cough during a (1) 63.01 (13.95) PEx (%) (2) 59.33 (14.7) (1) 46/47 (97.9) (2) 95/99 (96.0) BMI No. of subjects reporting Mean, kg/m2 (SD) PEx with full long-term (1) 21.94 (3.42) functional recovery* (%) (2) 21.68 (3.92) (1) 13/28 (46.4) (2) 21/44 (47.7) BMI-for-age z-score * Full long-term Mean, score (SD) recovery=return to ≥100% of (1) −0.95 (0.94) ©Institute for Clinical and Economic Review, 2018 Page 180 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Study Design and Author & Year of Duration of Publication, Interventions (n) & Inclusion and Exclusion Follow-up, Patient Characteristics Outcomes Harms (Trial), Dosing Schedule Criteria (Sites & geographical Quality Rating location) (2) −0.54 (0.95) ppFEV1 measurement most closely preceding PEx. ppFEV1 prior to first PEx Mean, percentage points (SD) (1) 68.36 (20.67) (2) 61.64 (16.75) Accurso 111 Multicenter phase 2 Part 1 Inclusion Sex ppFEV1 All AEs, no. reported (%) double-blind, N=20 • 18 years of age or older Females, n (%) Mean relative change from Part 1: 7 (88) NEJM placebo-controlled, • Diagnosed with CF Part 1: 11 (55) baseline, percentage points Part 2: 6 (86) two-part dose- (1) IVA: ivacaftor every • G551D mutation on at Part 2: 9 (47) (95% CI) 2010 ranging study (N=39). 12 hours in 25, 75 or least one CFTR allele Part 1 Mild AEs, no. reported 150mg dosage for 14 • ppFEV1≥40 Age 25mg: 4.9 (-2.6 to 12.5) (%) Phase 2 Part 1: Participants days, then 25, 75, or Median, years (range) 75mg: 10.0 (4.5 to 15.6) Part 1: 5 (63) randomly assigned to 150mg dosage for 14 Exclusion Part 1: 30 (19-51) 150mg: 10.5 (3.3 to 17.7) Part 2: 5 (71) receive 25, 75, or days post-washout • History of illness or Part 2: 21(18-42) Placebo: 0.7 (-8.8 to 10.2) 150mg of ivacaftor, period (n=4 per group) condition that may Moderate AEs, no. or placebo, every 12 confound results or BMI Difference: reported (%) hours for two 14-day (2) Placebo (n=4) pose safety risk Median, kg/m2 (range) 25mg vs placebo: p=0.45 Part 1: 0 periods separated by • Acute respiratory Part 1: 23 (17-29) 75mg vs. placebo: p=0.09 Part 2: 1 (14) a washout period. Part 2 infection, PE, or Part 2: 22 (20-25) 150mg vs placebo: p=0.10 N=19 changes in therapy for Severe AEs, no. reported Part 2: New pulmonary disease ppFEV1 ppFEV1 (%) participants within 4 weeks of Median, percentage Median relative change Part 1: 2 (25) randomly assigned to (1) IVA: 150 (n=8) or points (range) from baseline, percentage Part 2: 0 250mg (n=7) of enrollment receive either 150 or • Abnormal liver or renal Part 1: 56 (42-109) points (range) 250mg of ivacaftor, ivacaftor every 12 Part 2: 69 (40-122) Part 2 Discontinuation in Part 2: hours for 28 function at screening or placebo, every 12 • History of solid organ 150mg: 8.7 (2.1 to 31.3) 0 hours for 28 consecutive days CFQ-R Respiratory 250mg: 4.4 (0 to 18.3) or hematological consecutive days. transplant domain Placebo: 7.3 (5.2 to 8.2) (2) Placebo (n=4) Median, score (range) • Pregnancy or breast- Duration of follow- feeding Part 1: NA Difference up: Part 2: 72.2 (16.7-88.9) 150mg vs. placebo: p=0.56 ©Institute for Clinical and Economic Review, 2018 Page 181 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Study Design and Author & Year of Duration of Publication, Interventions (n) & Inclusion and Exclusion Follow-up, Patient Characteristics Outcomes Harms (Trial), Dosing Schedule Criteria (Sites & geographical Quality Rating location) 28 days • Ongoing participation 250mg vs. placebo: p=0.78 in another therapeutic clinical trial, or prior CFQ-R Respiratory domain participation in an Median change from investigational study baseline, points (range) without appropriate Part 2 at 28 days washout 150mg: 8.3 (0 to 16.7) 250mg: 11.1 (-5.6 to 33.3) Placebo: 0 Difference 150mg vs. placebo: p=0.46 250mg vs. placebo: p=0.47 Guigui 112 Non-randomized N=11 Inclusion: ppFEV1 ppFEV1 NR comparative study of • Ivacaftor provided by Mean, percentage Mean, percentage points Respir Med Case Rep ivacaftor (1) Ivacaftor (n=7) insurance company (at points (SD) effectiveness in time of study, ivacaftor Year 1 (1) 50 2016 individuals with (2) Regular care (n=4) was not approved to (1) NR (2) NR residual function treat those with (2) 61 (15) mutations at a single residual function BMI CF center mutations). Year 3 Mean, kg (SD) (1) 60 (NR) (1) 19.5 (2) Duration of follow- (2) 54 (14) (2) 22 (3) up: 3 years (one month BMI CFQ-R Respiratory after initiating Mean, kg/m2 (SD) domain ivacaftor treatment Year 3 Mean, score (SD) and every three (1) 22.3 (3) (1) 50 (5) months after) (2) 21 (3) (2) 48 (6) CFQ-R No. of PEs per year Mean, points (SD) (SD) Year 3 (1) 4.4 (2) (1) 95 (5) (2) 4.6 (2) (2) 50 (4) ©Institute for Clinical and Economic Review, 2018 Page 182 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Study Design and Author & Year of Duration of Publication, Interventions (n) & Inclusion and Exclusion Follow-up, Patient Characteristics Outcomes Harms (Trial), Dosing Schedule Criteria (Sites & geographical Quality Rating location) No. of PEs per year (SD) Year 3 (1) 2 (2) (2) 5.5 (3) Sawicki 68 Non-randomized N=1,075 Inclusion: ppFEV1 ppFEV1 NR comparative study; G551D Mean, percentage Annualized rate of decline, Am J Respir Crit Care G551D individuals 6+ (1) Ivacaftor (n=189), • Participation in STRIVE, points (SD) percent (SE) Med years of age who G551D only ENVISION, and/or Year 3 (1) 65.7 (19.5) received ivacaftor PERSIST (1) -0.91 (0.34) (2) 67.5 (20.4) 2015 during a phase 3 (2) Regular care • Have at least 3 FEV1 (2) –1.72 (0.16) study (STRIVE, (n=886), F508del measures over ≥6 Difference = 0.80 (95% CI: BMI-for-age z-score ENVISION, and/or homozygous only months after 30 days 0.06 to 1.55)* Mean, score (SD) PERSIST) were on ivacaftor (1) -0.16 (0.90) matched to up to 5 ppFEV1 (2) -0.12 (0.92) F508del homozygous F508del homozygous Treatment difference individuals using the • 2010 baseline during a Weight-for-age z-score Year 3 Cystic Fibrosis clinically stable 10.70 (p<0.001) Mean, score (SD) Foundation Patient encounter and (1) -0.21 (0.96) Registry (CFFPR). matching by propensity BMI (2) -0.17 (0.92) score to a G551D Mean BMI-for-age z-score Individuals were individual participating (SE)* matched by in one of the Phase 3 Year 3 propensity score studies (1) 0.087 (0.08) which included sex, (2) -0.23 (0.04) baseline age, year of CF diagnosis, sweat BMI-for-age z score, chloride value, CF- estimated rate of change* related diabetes, (1) -0.016 weight-for-age z (2) -0.024 score, BMI, use of p=0.72 inhaled medications and ppFEV1 (among Weight others) Mean weight-for-age z- score (SE) Year 3 (1) 0.08 (0.08) ©Institute for Clinical and Economic Review, 2018 Page 183 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Study Design and Author & Year of Duration of Publication, Interventions (n) & Inclusion and Exclusion Follow-up, Patient Characteristics Outcomes Harms (Trial), Dosing Schedule Criteria (Sites & geographical Quality Rating location) Duration of follow- (2) -0.22 (0.04) up: p<0.001 up to 3 years Weight-for-age z score, estimated rate of change* (1) NR (2) NR p=0.29 *Estimation and significance of rate change differences done by mixed model. Borowitz 69 Pooled and stratified See STRIVE and See STRIVE and Age Weight Not reported data from STRIVE and ENVISION ENVISION Mean, years (SD) Mean (least-squares) Dig Dis Sci ENVISION Ages ≤20 change from baseline, kg* randomized clinical (1) 12 (4.2) Ages ≤20 2016 trials (2) 12 (4.3) (1) 4.9 (2) 2.2 Ages >20 Difference=2.7 (95% CI:1.14 (1) 31 (8.4) to 4.29) (2) 29 (8.0) Ages >20 ppFEV1 (1) NR Mean, percentage (2) NR points (SD) Weight Ages ≤20 Mean weight-for-age z- (1) 77.5 (17.64) score, change from (2) 77.9 (19.01) baseline* Ages ≤20 Ages >20 (1) 0.29 (1) 60.3 (15.03) (2) -0.06 (2) 59.1 (15.57) Difference=0.35 (95%CI: 0.202 to 0.508) BMI Mean, kg (SD) Ages >20 Ages ≤20 ©Institute for Clinical and Economic Review, 2018 Page 184 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Study Design and Author & Year of Duration of Publication, Interventions (n) & Inclusion and Exclusion Follow-up, Patient Characteristics Outcomes Harms (Trial), Dosing Schedule Criteria (Sites & geographical Quality Rating location) (1) 18.5 (2.92) (1) NR (2) 18.2 (2.38) (2) NR Ages >20 BMI (1) 22.6 (3.73) Mean change from baseline, (2) 23.1 (3.42) kg/m2* Ages ≤20 BMI-for-age z-score (1) NR Mean, score (SD) (2) NR Ages ≤20 (1) -0.179 (0.9533) Ages >20 (2) -0.220 (0.8516) (1) 0.9 (2) -0.1 Ages >20 Difference=1.0 (95% CI: 0.44 (1) NR to 1.49) (2) NR BMI Mean weight at Mean BMI-for-age z score baseline, kg (SD) change from baseline* Ages ≤20 Ages ≤20 (1) 43.3 (16.18) (1) 0.26 (2) 41.8 (15.12) (2) -0.13 Difference=0.39 (95% CI: Ages >20 1.35 to 0.573) (1) 64.9 (13.87) (2) 65.4 (13.26) Ages >20 (1) 2.7 (2)-0.2 Difference=2.9 (95%CI: 1.35 to 4.47) *At 48 weeks. Konstan 113 Post-hoc analysis of See STRIVE and See STRIVE and Tertiles, by absolute ppFEV1 PEx, mean no. of days STRIVE and ENVISION ENVISION ENVISION change in ppFEV1, Mean absolute change from experienced (SD) looking at ivacaftor percentage points: Lower ivacaftor: ©Institute for Clinical and Economic Review, 2018 Page 185 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Study Design and Author & Year of Duration of Publication, Interventions (n) & Inclusion and Exclusion Follow-up, Patient Characteristics Outcomes Harms (Trial), Dosing Schedule Criteria (Sites & geographical Quality Rating location) Pediatr Pulmonol efficacy on an Ivacaftor (n=109) baseline, percentage points 15.61 (30.57) individual-response Lower tertile: (95% CI)* Lower placebo: 2015 level. FEV ≤5.56 (n=37) Lower Tertile 29.79 (50.63) Middle tertile: Ivacaftor: 1.58 Difference=14.18 Subgroups were FEV >5.56 and ≤13.5 Placebo: -6.39 defined by tertiles (n=36) Difference=7.97† (6.48 to Middle ivacaftor: (thirds) of FEV1 Upper tertile: 9.47) 14.59 (26.45) response. Patients FEV>13.59 (n=36) Middle placebo: were assigned to a Lower ivacaftor vs. pooled 33.64 (49.67) tertile within Placebo (n=100) placebo difference=2.29† Difference=19.05 treatment groups Lower: (0.40 to 4.19) based on the FEV ≤−2.65 (n=34) Upper ivacaftor: absolute change Middle: Middle Tertile 5.83 (15.94) from baseline in FEV >−2.65 and Ivacaftor: 9.37 Upper placebo: ppFEV1 through 48 ≤1.74 Placebo: -0.29 28.02 (40.24) weeks of treatment. (n=33) Difference=9.66† (8.77 to Difference=22.19 Upper: 10.55) (p=0.0019) FEV<1.74 (n=33) Upper Tertile Age Ivacaftor: 21.19 Mean, years (SD) Placebo: 5.59 Ivacaftor Difference=15.60† (13.00 to Lower: 23.1 (13.7) 18.19) Middle: 24.9 (10.6) Upper: 18.3 (8.3) Weight Mean change from baseline, Placebo kg (95% CI)* Lower: 22.1 (11.2) Lower tertile difference= Middle: 23.4 (11.4) 0.62 (2.10 to 5.13)† Upper: 18.0 (8.7) Middle tertile difference= 1.89 (-0.18 to 3.97) ppFEV1 Upper tertile difference= Mean, percentage 2.65 (0.39 to 4.91)† points (SD) Ivacaftor CFQ-R Lower: 72.1 (23.0) ©Institute for Clinical and Economic Review, 2018 Page 186 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Study Design and Author & Year of Duration of Publication, Interventions (n) & Inclusion and Exclusion Follow-up, Patient Characteristics Outcomes Harms (Trial), Dosing Schedule Criteria (Sites & geographical Quality Rating location) Middle: 64.5 (18.2) Mean absolute change from Upper: 68.9 (11.7) baseline, points (95% Ci) Lower tertile difference: Placebo 4.42 (-1.04 to 9.89) Lower: 73.1 (19.7) Middle tertile difference: Middle: 66.7 (18.7) 11.3 (6.85 to 15.74)† Upper: 64.6 (18.8) Upper tertile difference: 6.26 (1.06 to 11.47)† Weight Mean, kg (SD) *Through 48 weeks of Ivacaftor treatment Lower: 56.5 (22.5) †Significant difference vs. Middle: 58.3 (15.1) placebo Upper: 48.8 (15.8) Placebo Lower: 53.1 (21.4) Middle: 57.0 (15.9) Upper: 50.7 (17.8) Quittner 70 Analysis of STRIVE See STRIVE See STRIVE See STRIVE CFQ-R treatment difference Not reported CFQ-R data broken (ivacaftor vs. placebo) Health Qual Life down by individual Body Image* Outcomes survey scales: Body 2.7 (p=0.086) Image, Digestive Digestive Symptoms 2015 Symptoms, Eating 0.5 (p=0.732) Problems, Emotional Eating Problems* Functioning, Health 3.3 (p=0.002) Perceptions, Physical Emotional Functioning* Functioning, 2.1 (p=0.096) Respiratory Health Perceptions* Symptoms, Role 7.6 (p<0.001) Functioning, Social Physical Functioning* Functioning, 4.4 (p=0.006) Treatment Burden, Respiratory Symptoms* Vitality, and Weight. 8.6 (p<0.001) ©Institute for Clinical and Economic Review, 2018 Page 187 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Study Design and Author & Year of Duration of Publication, Interventions (n) & Inclusion and Exclusion Follow-up, Patient Characteristics Outcomes Harms (Trial), Dosing Schedule Criteria (Sites & geographical Quality Rating location) Role Functioning Participants ages 14+ -0.6 (p=0.651) completed the Social Functioning* Teen/Adult version; 4.3 (p=0.003) those under 14 at Treatment Burden baseline completed 3.3 (p=0.042) the Child version. Vitality Parents of 12 and 13 5.5 (p=0.002)* year-olds completed Weight the Parent/Caregiver 5.3 (p=0.053) CFQ-R. *Placebo reported decrease Minimal clinically in CFQ-R score between important difference baseline and 48 weeks. (MCID) defined as 4 points for CFQ-R scores. Heltshe 114 Combination data See GOAL See GOAL PA infection duration PA culture positivity, odds Not reported from GOAL and in year prior to ratio* Clin Infect Dis Cystic Fibrosis treatment with 0.65 (35% reduction) Foundation Patient ivacaftor, n/N (%) 2015 Registry analyzing Persistent* PA prevalence after Pseudomonas 59/145 (40%) ivacaftor initiation by aeruginosa (PA) Intermittent baseline category, n/N incidence, 30/148 (20%) infection free (%)* prevalence, and Infection-free Persistent association with 59/148 (40%) 5/48 (10%) clinical outcomes Intermittent during treatment *Note: participants 21/30 (70%) with ivacaftor. with persistent infection tended to be GOAL data (6 mos. of older, had lower FEV1, Frequency of PA isolation ivacaftor) and higher after ivacaftor initiation, supplemented with hospitalization rates at n/N (%)* CFFPR data from year baseline. More frequent before and year after 7/143 (5%) ©Institute for Clinical and Economic Review, 2018 Page 188 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Study Design and Author & Year of Duration of Publication, Interventions (n) & Inclusion and Exclusion Follow-up, Patient Characteristics Outcomes Harms (Trial), Dosing Schedule Criteria (Sites & geographical Quality Rating location) ivacaftor treatment Less frequent initiation for 36/134 (27%) comparison. No change 91/143 (68%) Duration of follow- up: 2 years Reduction in PA frequency (Median follow-up in was not significantly the CFFPR=12.5 associated with mos.) improvements in FEV1, BMI, hospitalization, or exacerbation rate. *On ivacaftor vs. before ivacaftor. Bai 73 Non-randomized N=1,324 NR NR US data only See Outcomes comparative long- J Cyst Fibros term post-approval (1) IVA (n=215) Deaths, n/N (%) observational safety (1) 0/215 (0) 2016 study using data (2) Standard of care (2) 2/1109 (0.2) from UK and US CF (n=1,109) Abstract patient registries. Organ transplants, n (%) (1) 0 (0) Comparators not (2) 1 (0.1) receiving ivacaftor were matched to Hospitalizations, n (%) ivacaftor recipients (1) 25 (11.6) based on age, sex, (2) 338 (30.5) and genotype RR (95% CI)=0.38 (0.26 to severity. 0.56) Duration of follow- PEx, n (%) up: (1) 20 (9.3) 1 year (2014) (2) 307 (27.7) RR (95% CI)=0.34 (0.22 to 0.52) ©Institute for Clinical and Economic Review, 2018 Page 189 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Study Design and Author & Year of Duration of Publication, Interventions (n) & Inclusion and Exclusion Follow-up, Patient Characteristics Outcomes Harms (Trial), Dosing Schedule Criteria (Sites & geographical Quality Rating location) Cystic fibrosis related diabetes (CFRD), n (%) (1) 16 (7.5) (2) 131 (11.9) RR (95% CI)=0.63 (0.38 to 1.03) Hepatobiliary complications, n (%) (1) 3 (1.4) (2) 62 (5.6) RR (95% CI) =0.25 (0.08 to 0.79) Pulmonary complications, n (%) (1) 61 (28.4) (2) 392 (35.4) RR (95% CI)=0.80 (0.64 to 1.01) Bai 12 Non-randomized N=7,456 NR NR US data only See Outcomes comparative long- J Cyst Fibros term post-approval (1) IVA (n=1,256) Deaths, n/N (%) observational safety (1) 8/1256 (0.6) 2016 study using data (2) Standard of care (2) 97/6200 (1.6) from UK and US CF (6,200) RR (95% CI)=0.41 (0.20 to Abstract patient registries. 0.84) Only US data is reported Organ transplants, n (%) (1) 2 (0.2) Comparators not (2) 68 (1.1) receiving ivacaftor RR (95% CI)=0.15 (0.04 to were matched to 0.59) ivacaftor recipients based on age, sex, Hospitalizations, n (%) ©Institute for Clinical and Economic Review, 2018 Page 190 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Study Design and Author & Year of Duration of Publication, Interventions (n) & Inclusion and Exclusion Follow-up, Patient Characteristics Outcomes Harms (Trial), Dosing Schedule Criteria (Sites & geographical Quality Rating location) and genotype (1) 346 (27.6) severity. (2) 2671 (43.1) RR (95% CI)=0.64 (0.58 to Duration of follow- 0.70) up: 1 year (2014) PE, n (%) (1) 349 (27.8) (2) 2684 (43.3) RR (95% CI)=0.64 (0.58 to 0.70) Barry 115 Non-randomized N=56 NR NR Deaths, n/N See Outcomes comparative (1) 5/21 J Cyst Fibros prospective cohort (1) Ivacaftor (n=21) Ivacaftor group (2) 12/21 study measuring received drug in prior 2015 effects of ivacaftor (2) Standard of care multi-center cohort Lung transplant, n/N on death and (n=35) study and had baseline (1) 1/21 Abstract transplantation FEV1 <40 and (2) 8/21 among CF patients continued treatment with FEV1 <40. during prospective Mulivariate model, all cohort study. subjects: Duration of follow- Ivacaftor therapy associated up: with improved survival Median = 1126 days (HR=0.24, p=0.047) Male sex associated with improved survival (HR=0.13, p=0.012) Volkova 116 Non-randomized N=1,642 NR ppFEV1 ppFEV1 PEx comparative long- Mean, percentage Mean, percentage points Annual risk, % J Cyst Fibros term post-approval (1) Ivacaftor (n=277) points (SD) (SD) 2013 observational safety 2013 (1) 49.5 (1) 70.6 (24.8) 2016 study using a United (2) Standard of care (1) 75.8 (25.7) (2) 56.8 (2) 71.4 (23.6) Kingdom CF registry. (n=1365) (2) 70.6 (24.3) Abstract 2014 PEx 2014 (1) 34.3 Annual risk, % ©Institute for Clinical and Economic Review, 2018 Page 191 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Study Design and Author & Year of Duration of Publication, Interventions (n) & Inclusion and Exclusion Follow-up, Patient Characteristics Outcomes Harms (Trial), Dosing Schedule Criteria (Sites & geographical Quality Rating location) 2012 registry data (1) 51.6 (1) 77.8 (25.6) (2) 57.0 served as baseline. (2) 44.3 (2) 70.8 (24.2) Patients with a record of ivacaftor in Annual risk of Annual risk of 2013 and 2014 were hospitalization for PEx, hospitalization for PEx, % matched about 1:5 to % 2013 comparator patients (1) 48.0 (1) 38.3 without a history of (2) 43.4 (2) 44.3 ivacaftor use with comparable age, sex, CFRD, % 2014 and genotype (1) 17.3 (1) 24.6 severity. (2) 23.2 (2) 45.6 Distal intestinal Annual risk of Cystic obstruction syndrome, fibrosis-related diabetes, % % (1) 6.5 2013 (2) 7.4 (1) 18.8 (2) 25.6 2014 (1) 20.6 (2) 28.4 Annual risk of distal intestinal obstruction syndrome (DIOS), % 2013 (1) 5.1 (2) 7.5 2014 (1) 4.7 (2) 8.1 Elborn 117 Subgroup analysis of N=213 See STRIVE and Age ppFEV1 See STRIVE and ENVISION STRIVE and ENVISION ENVISION N ivacaftor/n placebo Mean absolute change from ©Institute for Clinical and Economic Review, 2018 Page 192 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Study Design and Author & Year of Duration of Publication, Interventions (n) & Inclusion and Exclusion Follow-up, Patient Characteristics Outcomes Harms (Trial), Dosing Schedule Criteria (Sites & geographical Quality Rating location) Am J Resp Crit Care ivacaftor treatment (1) Ivacaftor (See STRIVE baseline, percentage points Med effect on mean STRIVE and ENVISION) <18: 19/17 (p-value) absolute change 18+: 64/61 STRIVE 2012 from baseline (2) Placebo (See STRIVE <18: 11.9 (p=0.0003) ppFEV1 at 24 weeks and ENVISION) ENVISION 18+: 9.9 (p<0.0001) Abstract by baseline age and <18: 26/26 FEV1. 18+: 0 ENVISION <18: 12.5 (p<0.0001) Duration of follow- Low FEV1 18+: NA up: N ivacaftor/n placebo 24 weeks STRIVE (ppFEV1<70%) Low FEV1 (1) 49 STRIVE: 10.7 (p<0.0001) (2) 45 ENVISION: NA ENVISION Mid FEV1 (ppFEV1<70%) STRIVE: 10.6 (p<0.0001) (1) 4 ENVISION: 9.3 (p=0.1322) (2) 8 High FEV1 Mid FEV1 STRIVE: NA N ivacaftor/n placebo ENVISION: 6.9 (p=0.1920) STRIVE (ppFEV1≥70) (1) 34 (2) 33 ENVISION (ppFEV1 70- 90%) (1) 12 (2) 6 High FEV1 N ivacaftor/n placebo STRIVE (Not defined) (1) 4 (2) 5 ©Institute for Clinical and Economic Review, 2018 Page 193 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Study Design and Author & Year of Duration of Publication, Interventions (n) & Inclusion and Exclusion Follow-up, Patient Characteristics Outcomes Harms (Trial), Dosing Schedule Criteria (Sites & geographical Quality Rating location) ENVISION (ppFEV1>90%) (1) 10 (2) 11 Flume 118 Analysis of PEx N= 213 See STRIVE and See STRIVE and Incidence of protocol- Not reported incidence and ENVISION ENVISION defined signs and symptoms J Cyst Fibros incidence of (1) IVA: ivacaftor group of a PEx, no. times reported protocol-defined PEx from STRIVE (n=83) (% of total events) 2013 signs and symptoms Increased cough reported in STRIVE. (2) Placebo (n=78) (1) 99 (26.7) Abstract (2) 145 (23.3) Change in sputum (1) 73 (19.7) (2) 110 (17.7) Malaise, fatigue, lethargy (1) 45 (12.1) (2) 76 (12.2) Dyspnea (1) 33 (8.9) (2) 64 (10.3) Bai 119 5-year observational N=5,931 Not reported Patients treated with No. of deaths, annual risk See Outcomes post-authorization ivacaftor were (%) Pediatr Pulmonol safety study (1) IVA (n=999) matched 1:5 with (1) 5 (0.5) patients in the CFFPR (2) 66 (1.3) 2015 (2) Comparator who never received Unadjusted relative risks* Analyzed results of ivacaftor on age, (95% CI) = 0.37 (0.15 to 0.93) the US CF Foundation (n=4,932) gender, and CFTR Abstract Patient Registry genotype. No. of organ (CFFPR) data in 2013 transplantation, annual risk (%) (1) 2 (0.2) (2) 53 (1.1) ©Institute for Clinical and Economic Review, 2018 Page 194 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Study Design and Author & Year of Duration of Publication, Interventions (n) & Inclusion and Exclusion Follow-up, Patient Characteristics Outcomes Harms (Trial), Dosing Schedule Criteria (Sites & geographical Quality Rating location) Average duration of Unadjusted relative risks ivacaftor exposure (95% CI) = 0.19 (0.05 to 0.76) was 1.4 years No. of hospitalization, Duration of follow- annual risk (%) up: 5 years (1) 247 (24.7) (2) 2055 (41.7) Unadjusted relative risks (95% CI) = 0.59 (0.53 to 0.66) No. of PEx, annual risk (%) (1) 256 (25.6) (2) 2037 (41.3) Unadjusted relative risks (95% CI) = 0.62 (0.56 to 0.69) *Unadjusted relative risks for ivacaftor vs comparator cohort as well as their 95% CIs based on normal approximation were calculated by the authors. Mainz 120 Compared CFQ-R N=209 Inclusion Sex CFQ-R Respiratory domain scores of G551D • 12 years of age or older Female, n (%) Mean (least-squares) score, J Cyst Fibros patients on IVA (≥ 3 (1) IVA* (n=72) • G551D-CFTR mutation (1) 43 (60.3) points* • Caregivers of pts aged (1) 75.4 months) to (2) 73 (35.2) 2016 (2) Caregiver, standard 6-11 completed a one- (2) 62.5 homozygous F508del of care (n=137) time survey comprising on standard of care Mean no. of CFQ-R Digestive Symptoms Abstract the CFQ-R, EQ-5D-5L, and in a real-world WPAI comorbidities, n domain setting (prior to *The mean duration of (1) 1.5 Mean (least-squares) score* patients on ivacaftor (1) 85.4 LUM/IVA (2) 2.0 was 22 months. (2) 78.0 availability). p<0.01 ©Institute for Clinical and Economic Review, 2018 Page 195 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Study Design and Author & Year of Duration of Publication, Interventions (n) & Inclusion and Exclusion Follow-up, Patient Characteristics Outcomes Harms (Trial), Dosing Schedule Criteria (Sites & geographical Quality Rating location) Clinical data was CFQ-R Eating domain collected from Mean (least-squares) score* patient medical (1) 91.1 (2) 84.2 records. Duration of follow- CFQ-R Health Perceptions up: domain survey administered Mean (least-squares) score* once (1) 67.6 (2) 58.6 CFQ-R Physical Functioning domain Mean (least-squares) score* (1) 74.6 (2) 66.6 CFQ-R Treatment Burden domain Mean (least-squares) score (1) 65.3 (2) 54.8 CFQ-R Vitality domain Mean (least-squares) score* (1) 63.5 (2) 55.9 CFQ-R Weight domain Mean (least-squares) score* (1) 80.7 (2) 64.2 EQ-5D-5L index score* (1) 0.90 ©Institute for Clinical and Economic Review, 2018 Page 196 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Study Design and Author & Year of Duration of Publication, Interventions (n) & Inclusion and Exclusion Follow-up, Patient Characteristics Outcomes Harms (Trial), Dosing Schedule Criteria (Sites & geographical Quality Rating location) (2) 0.81 VAS score (p-value)* (1) 75.7 (2) 70.0 School productivity loss (%) (1) 24.6 (2) 33.6 Activity impairment (%) (1) 21.6 (2) 28.3 *Statistically significant difference between ivacaftor and standard of care Accurso 121 3 randomized, N= Not reported Not reported ppFEV1 Not reported blinded, phase 2 Mean change from baseline, J Cyst Fibros studies in G551D (1) Study 101: Ivacaftor percentage points (SE); p- patients had cross- treatment lasted 14 value 2013 over designs. (1) 5.2 (2.0); NR days (n=4) (2) 7.1 (2.7); p=0.0104 (3) 8.8 (2.7); p=0.0313 Abstract (2) Study 106: Ivacaftor treatment lasted 28 days (n=18) (3) Study 107: Ivacaftor treatment lasted 28 days (n=8) ©Institute for Clinical and Economic Review, 2018 Page 197 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Study Design and Author & Year of Duration of Publication, Interventions (n) & Inclusion and Exclusion Follow-up, Patient Characteristics Outcomes Harms (Trial), Dosing Schedule Criteria (Sites & geographical Quality Rating location) Davies 122 Phase 2, randomized, N=7 (interim analysis) Inclusion Age ppFEV1 Any AE, n/N double-blind, • 6 years of age or older Mean, years (SD) Treatment difference for During placebo: 5/7 J of Cyst Fibros placebo-controlled, Participants were • Confirmed diagnosis of 14.0 (8.6) the mean change from During ivacaftor: 6/7 crossover, randomized to one of CF, with GG551D-CFTR baseline, percentage points 2012 multicenter study. mutation LCI (p-value) two treatment orders: SAE, n/N • FEV1 of at least 90% 7.2 (p=0.1264) Mean (SD) 1/7 Duration of follow- LCI of at least 7.4 Abstract (1) 150mg of ivacaftor 9.2 (1.9) up: LCI 12 weeks (2 four- every 12 hours for 4 Mean change from baseline week treatment weeks, washout for 4 ppFEV1 treatment difference (p- periods with four- weeks, and 150mg Mean, percentage value) week washout placebo every 12 hours points (SD) −2.22 (p=0.0097) between) 98.5 (6.4) for 4 weeks OR (2) 150mg of placebo every 12 hours for 4 weeks, washout for 4 weeks, and 150mg ivacaftor every 12 hours for 4 weeks ©Institute for Clinical and Economic Review, 2018 Page 198 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Study Design and Author & Year of Duration of Publication, Interventions (n) & Inclusion and Exclusion Follow-up, Patient Characteristics Outcomes Harms (Trial), Dosing Schedule Criteria (Sites & geographical Quality Rating location) Elborn 123 Post-hoc analyses on N=31 Inclusion ppFEV1 48 Week Data: Not reported STRIVE, ENVISION • FEV1 of at least 90% at Mean, percentage Pediatr Pulmonol (and Study 106 which (1) STRIVE IVA (n=4) baseline in STRIVE, points (SD) ppFEV1 ENVISION Absolute change from not reported here) (1) 95.6 (2.7) 2013 (2) STRIVE Placebo baseline, percentage points randomized, placebo (2) 93.8 (3.0) (SD) controlled, double- (n=5) (3) 99.3 (12.4) (1) 9.1 (3.0) Abstract (4) 101.7 (6.5) blind, multicenter (2) −7.7 (13.7) studies. (3) ENVISION IVA (3) 1.5 (13.5) (n=10) (4) −4.4 (8.3) Weight Duration of follow- Mean, kg (SD) up: (4) ENVISION Placebo Weight (1) 59.2 (20.1) 48 weeks (n=12) Absolute change from (2) 58.8 (2.2) baseline, kg (SD) (3) 37.4 (12.5) (1) 8.2 (7.6) (4) 29.8 (7.3) (2) −1.6 (2.7) (3) 7.0 (3.7) (4) 3.0 (2.3) Plant 124 Secondary analyses N=209 See STRIVE, ENVISION See STRIVE, ENVISION ppFEV1 Not reported of STRIVE and Treatment difference in J Cyst Fibros ENVISION, including mean change from baseline, analysis of ppFEV1 (1) IVA: 48 weeks of percentage points (p-value) and body weight by ivacaftor (n=109) STRIVE 2013 FEV1 response (<5% <5% FEV1 improvement: 4.2 and ≥5% (2) Placebo: 48 weeks (p<0.0001) Abstract improvement). of placebo (n=100) ≥5%: 6.2 (p=0.0023) Duration of follow- ENVISION up: <5%: 1.6 (p=0.5093) 48 weeks (see STRIVE ≥5%: 9.8 (p=0.0522) and ENVISION) Weight ©Institute for Clinical and Economic Review, 2018 Page 199 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Study Design and Author & Year of Duration of Publication, Interventions (n) & Inclusion and Exclusion Follow-up, Patient Characteristics Outcomes Harms (Trial), Dosing Schedule Criteria (Sites & geographical Quality Rating location) Treatment difference in absolute change from baseline, kg (p-value) STRIVE <5%: 3.3 (p<0.0001) ≥5%: 1.7 (p=0.3313) ENVISION >5%: 2.0 (p=0.0582) ≥5%: 3.4 (p=0.0094) Suthoff 125 Analysis of patient- (1) IVA: 150 mg of See STRIVE See STRIVE CFQ-R Respiratory domain Not reported reported quality of ivacaftor twice daily Percent of subjects Pediatr Pulmonol life outcomes, via reporting* (2) Matched placebo Improvement (p-value) CFQ-R, from STRIVE. 2014 (1) 57 (2) 25 Duration of follow- STRIVE up: Decline 48 weeks (1) 29 Abstract (2) 54 CFQ-R Social Functioning domain Percent of subjects reporting* Improvement (p-value) (1) 49 (2) 29 Decline (1) 30 (2) 50 CFQ-R Vitality domain ©Institute for Clinical and Economic Review, 2018 Page 200 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Study Design and Author & Year of Duration of Publication, Interventions (n) & Inclusion and Exclusion Follow-up, Patient Characteristics Outcomes Harms (Trial), Dosing Schedule Criteria (Sites & geographical Quality Rating location) Percent of subjects reporting* Improvement (p-value) (1) 49 (2) 23 Decline (1) 36 (2) 50 CFQ-R Treatment Burden domain Percent of subjects reporting* Improvement (p-value) (1) 44 (2) 22 Decline (1) 26 (2) 41 CFQ-R Health Perceptions domain Percent of subjects reporting* Improvement (p-value) (1) 44 (2) 17 Decline (1) 28 (2) 45 ©Institute for Clinical and Economic Review, 2018 Page 201 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Study Design and Author & Year of Duration of Publication, Interventions (n) & Inclusion and Exclusion Follow-up, Patient Characteristics Outcomes Harms (Trial), Dosing Schedule Criteria (Sites & geographical Quality Rating location) CFQ-R Physical Functioning domain Percent of subjects reporting* Improvement (p-value) (1) 35 (2) 12 Decline (1) 13 (2) 40 CFQ-R Eating Problems domain Percent of subjects reporting* Improvement (p-value) (1) 25 (2) 10 Decline (1) 12 (2) 27 CFQ-R Weight Problems Percent of subjects reporting* Improvement (p-value) (1) 19 (2) 13 Decline (1) 9 (2) 28 ©Institute for Clinical and Economic Review, 2018 Page 202 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Study Design and Author & Year of Duration of Publication, Interventions (n) & Inclusion and Exclusion Follow-up, Patient Characteristics Outcomes Harms (Trial), Dosing Schedule Criteria (Sites & geographical Quality Rating location) *p<0.05 for difference between treatment groups in the percent improved and declined Hathorne 126 Quality of life N=151 See GOAL See GOAL Statistical significance of Not reported analysis using GOAL • improvement in CFQ-R Pediatr Pulmonol study data. Ivacaftor (single arm) domains after 6 mo of treatment by sex (p-value)* 2015 Treatment Burden domain Data was measured (1) females (p=0.0002) before and 6 months (2) males (p=0.0034) GOAL after initiation of ivacaftor. Health Perceptions domain Abstract (1) females (p=0.0292) (2) males (p=0.0121) Physical Functioning domain (1) females (p=0.0429) (2) males (p=0.0110) Role Functioning domain (1) females (p=0.0001) (2) males (p=0.0061) * Authors do not define whether changes in quality of life (CFQ-R scores) meet a minimum clinically important difference. Unclear whether statistical significance of improvement meets threshold for clinical importance. ©Institute for Clinical and Economic Review, 2018 Page 203 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Study Design and Author & Year of Duration of Publication, Interventions (n) & Inclusion and Exclusion Follow-up, Patient Characteristics Outcomes Harms (Trial), Dosing Schedule Criteria (Sites & geographical Quality Rating location) Wainwright 127 12 months data from N=331 Inclusion Age Median hospital admission Not Reported the Australian CF • 15-54 years of age Mean, years (SD) count (IQR) Pediatr Pulmonol Data Registry (1) IVA: n=17 • Confirmed diagnosis of (1) 29 (7.3) (1) 0.6 (0.0 to 1.8) CF (2) 27 (8) (2) 2.4 (0.6 to 3.5) (ACFDR). 2014 (2) Matched placebo: • Pancreatic insufficient Difference: p=0.007 n=314 patients with G551D ppFEV1 Duration of follow- Length of stay in hospital, Abstract mutation Mean, percentage up: days (IQR) Patients were assessed • FEV1 < 70% points (SD) 24 weeks (1) 2.9 (0.0 to 27.5) every 2-3 months post- (1) 38.3 (12.4) (2) 23.5 (8.2 to 45.2) (2) 45.4 (14.5) Difference: p=0.015 treatment. (n=17) BMI Data were collected Mean, kg (SD) retrospectively from (1) 20.4 (2.6) patient records and the (2) 20.5 (2.8) physician declaration form required every 3 months for supply/resupply of ivacaftor. Barry 128 Retrospective case- N=56 Inclusion Age ppFEV1 No adverse events control study of • Confirmed diagnosis of Mean, years (range) Mean, percentage points reported in the treatment Chest patients receiving (1) IVA: cases had at CF (1) 22 (20-31) (SD) group. ivacaftor on the least 3 months • At least one G551D (2) 23 (21-27) (1) 30.7 (9.9) 2014 compassionate use allele (2) NR treatment with 2 previously listed control program in the UK • ppFEV1 < 40% ppFEV1 and Ireland. ivacaftor by the time of ppFEV1 subjects underwent lung • Minimum of 3 months Mean, percentage transplantation. data collection (n=21) Median absolute change treatment with ivacaftor points (SD) Duration of follow- • from baseline, percentage up: (2) Matched control (1) 26.5 (7.2) points (IQR) Exclusion (2) 30.3 (7.5) 1-1.75 years (1 year subjects: each case was (1) 3.8 (0.2 to 7.7) • Patients with FEV1 before ivacaftor matched up to 2 (2) 0.6 (-2.1 to 2.8) <40% were excluded Weight treatment and 90- control subjects (n=35) from phase 3 clinical Median, kg (IQR) 270 days on Weight trials (1) 49.8 (44.4-60.7) ivacaftor) Median, kg (IQR) ©Institute for Clinical and Economic Review, 2018 Page 204 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Study Design and Author & Year of Duration of Publication, Interventions (n) & Inclusion and Exclusion Follow-up, Patient Characteristics Outcomes Harms (Trial), Dosing Schedule Criteria (Sites & geographical Quality Rating location) Median time on (2) 54 (49.0-62.4) (1) 51.6 (48.6 to 66.8) ivacaftor: 237 days (2) NR BMI Mean, kg/m2 Weight (1) 19.1 (2.9) Median change from (2) 20.2 (5.2) baseline, kg (IQR) (1) 2.3 (-0.4 to 4.2) Sex (2) 0.6 (-0.5 to 3.2) Female, % BMI (1) 52 NR, kg/m2 (2) 49 (1) 20.2 (2) NR BMI Median change from baseline, kg/m2 (IQR) (1) 0.84 (NR) (2) 0.2 (NR) Davies 71 Phase 2, multicenter, N=20 Inclusion By arm (treatment Results are pooled for all Any AE, n (%) placebo-controlled, • Confirmed diagnosis of order 1 or 2) subjects during ivacaftor Ivacaftor: 13 (72%) Lancet Respir Med double-blind 2x2 Demographics: CF and placebo weeks. Placebo: 15 (79%) crossover study. (1) Placebo → IVA: 28 • At least one G551D- Age 2013 CFTR allele Mean, years (SD) ppFEV1 days of placebo twice SAE, n Duration of follow- • ppFEV1 > 90% (1) 19.8 (13.35) Mean, percentage points up: daily, 28-day washout (2) 13.4 (7.12) (95% CI) Ivacaftor: 3 • Age of 6 years or older 28 days period, and 28 days of Ivacaftor: 104.97 Placebo: 1 • Weight ≥ 15 kg 150 mg ivacaftor twice ppFEV1 Placebo: 94.85 • LCI > 7.4 daily (n=10) Mean, percentage Difference= 8.67 (2.36 to 14.97) points (SD) (2) IVA → Placebo: 28 (1) 92.6 (7.43) CFQ-R Respiratory domain days of 150 mg (2) 101.8 (11.59) Mean, points (95% CI) ivacaftor twice daily, BMI Ivacaftor: 83.33 28-day washout period, Mean, kg (SD) Placebo: 79.97 (1) 22.7 (6.96) ©Institute for Clinical and Economic Review, 2018 Page 205 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Study Design and Author & Year of Duration of Publication, Interventions (n) & Inclusion and Exclusion Follow-up, Patient Characteristics Outcomes Harms (Trial), Dosing Schedule Criteria (Sites & geographical Quality Rating location) 28 days of placebo (2) 19.4 (3.71) Difference= 3.99 (−5.32 to twice daily (n=10) 1.33) Sex Results, at 28 days Female, n (%) LCI (95% CI) (1) IVA (n=18) (1) 4 (40) Ivacaftor: 8.13 (2) 6 (60) Placebo: 9.40 (2) Placebo (n=17) Difference= −2.16 (−2.88 to 1.44) CFQ-R Respiratory domain Mean, score (SD) (1) 71.7 (13.4) (2) 75.6 (18.2) LCI Mean (SD) (1) 8.88 (1.46) (2) 9.17 (1.66) Edgeworth 129 Single-center, N=20 Inclusion All participants Results are pooled for all All participants double-blind, • Aged between 16 and subjects during ivacaftor Clin Sci (London) placebo-controlled, (1) IVA: ivacaftor 150 75 years Age and placebo weeks. No. hospitalizations for randomized, mg twice daily for 28 • Confirmed diagnosis of Mean, years (range) PEs 2017 crossover study. CF 32 (18-65)* ppFEV1 days (n=10) 5 • At least one G551D- Mean absolute change from Duration of follow- CFTR allele ppFEV1 baseline, percentage points up: (2) Matched Placebo: (95% CI) Abdominal discomfort, n 150 mg of placebo • ppFEV1 ≥ 25% Mean, percentage 84 days; 28 days of (1) 14.1 (9.4 to 18.8) 3 twice daily for 28 days points (range) treatment; 28 days of Exclusion (2) 0.4 (-4.3 to 5.1) washout; 28 days of (n=10) 54 (23-110) Difference = 13.7 (7.0 to • Known adverse Elevated creatinine other treatment 20.3) reaction to ivacaftor BMI kinase, n • Deemed unlikely to Mean, kg/m2 (SD) 1 BMI physically complete a 25.8 (18-36.4) Mean absolute change from CPET study baseline, kg/m2 (95% CI) Sex (1) 1.9 (1.1 to 2.7) Female, n (%) (2) 0.7 (-0.2 to 1.5) ©Institute for Clinical and Economic Review, 2018 Page 206 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Study Design and Author & Year of Duration of Publication, Interventions (n) & Inclusion and Exclusion Follow-up, Patient Characteristics Outcomes Harms (Trial), Dosing Schedule Criteria (Sites & geographical Quality Rating location) 8 (40) Difference = 1.2 (0.1 to 2.3) • CFQ-R Respiratory domain Mean absolute change from baseline (95% CI) (1) 16.1 (−29.9–62.0) (2) −6.1 (−41.0 to 28.8) Difference: 22.2 (−26.3 to 70.6) Stalvey 130 Post-hoc analysis on N=83 See GOAL and ENVISION Weight-for-age z-score Weight Not reported GOAL and ENVISION Mean, score (p-value) Mean weight-for-age z- Pediatr Pulmonol GOAL: (1) 0 score at endpoint (p-value) Duration of follow- (2) 0.08 (1) 0.27 (p<0.0001 vs. 2017 up: (1) IVA: n=35 baseline) (3) -0.16 GOAL: 6 mo (2) 0.44 (p<0.001 vs. ENVISION: 48 weeks placebo) GOAL and ENVISION ENVISION: Age (3) -0.36 (p<0.001 vs. Mean, years (SD) (2) IVA: n=25 (1) 8.7 (1.6) baseline) (2) 8.5 (1.8) (3) Placebo: n=23 (3) 8.8 (1.8) ppFEV1 Mean, percentage points (SD) (1) 106.4 (14.6) (2) 87.3 (14.6) (3) 83.8 (20.8) BMI Mean, kg/m2 (SD) (1) 17.1 (2.4) ©Institute for Clinical and Economic Review, 2018 Page 207 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Study Design and Author & Year of Duration of Publication, Interventions (n) & Inclusion and Exclusion Follow-up, Patient Characteristics Outcomes Harms (Trial), Dosing Schedule Criteria (Sites & geographical Quality Rating location) (2) 17.2 (2.7) (3) 16.8 (1.8) Sex Female, n (%) (1) 16 (45.7) (2) 14 (56) (3) 9 (39.1) Fink 131 Retrospective N=403 NR Mean age at treatment ppFEV1 Not reported observational cohort start, years (median) Mean change from baseline, Pediatr Pulmonol study using US Cystic Ivacaftor (single arm) 21.4 (18.5) percentage points (SD) Fibrosis Foundation 5.4 (9.1) 2015 Patient Registry Females, % comparing 49 Mean difference in no. PEx’s Abstract nutritional and reported (SD) pulmonary outcomes -2.1 (1.1) in the 12 months preceding and 12 Weight months on ivacaftor. Mean change in from baseline, kg (SD) 4.3 (4.7) Percent without change in weight or lung function 13 Percent with change in weight and lung function 42 Percent with change in only weight 37 ©Institute for Clinical and Economic Review, 2018 Page 208 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Study Design and Author & Year of Duration of Publication, Interventions (n) & Inclusion and Exclusion Follow-up, Patient Characteristics Outcomes Harms (Trial), Dosing Schedule Criteria (Sites & geographical Quality Rating location) Percent with change in only lung function 8 Percent with FEV1 response and baseline FEV1 of: ≥80: 43 40-79: 60 <40: 48 Percent with weight response and baseline FEV1 of: ≥80: 84 <80: 72 Multiple Regimens Heltshe 132 Retrospective, Pre-and post-phase III Women with cystic Genotype, N (%) The number of women with NA observational, trials of ivacaftor (2009- fibrosis between the ages Homozygous F508del: CF in the childbearing years J Cyst Fibros epidemiologic 2013) and of 15-44 (childbearing 31,989 (46.7) increased annually from analysis using the US lumacaftor/ivacaftor 5,335 in 2005 to 7,164 in years) Heterozygous F508del: 2017 CF Foundation (2013-2014) 2014 Patient Registry 22,533 (32.9) Manuscript between 2005-2014 G551D: 2,860 (4.2) Slight downward trend in R117H: 1,182 (1.7) pregnancy rates (2% Other: 9,884 (14.4) reduction per year) Pregnancy rate per 100 consistent with national woman-years (all trends. Pregnancy rates were lower years): 25.5 during years of clinical trials (compared to pre-trial) but rebounded post-approval for ivacaftor (no data on lumacaftor/ivacaftor). Number of live births grew from 2005-2009 (70.1%) to ©Institute for Clinical and Economic Review, 2018 Page 209 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Study Design and Author & Year of Duration of Publication, Interventions (n) & Inclusion and Exclusion Follow-up, Patient Characteristics Outcomes Harms (Trial), Dosing Schedule Criteria (Sites & geographical Quality Rating location) 2013-2014 (73.4%) in registry population. Percent live births were higher in the CF population than the overall US population (64.6%) ©Institute for Clinical and Economic Review, 2018 Page 210 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Appendix G. Summaries of Public Comments Delivered at Public Meeting This section includes summaries of the public comments prepared for the Midwest CEPAC Public Meeting on May 17, 2018 in St. Louis, Missouri. These summaries were prepared by those who delivered the public comments at the meeting and are presented in order of delivery. A video recording of all comments can be found here (https://youtu.be/fw-FHQWNvfE), beginning at minute 1:50:20. Conflict of interest disclosures are included at the end of the statement. Two speakers did not submit a written public comment summary. 1. Michael Boyle, MD, Senior Vice President of Theraputics Development, Cystic Fibrosis Foundation At time of publication, ICER has not received a summary of public comment from Dr. Boyle. Dr. Boyle is an employee of Cystic Fibrosis Foundation (CFF), which provides research and clinical trial support to health care companies, including Vertex Pharmaceuticals, that results in the Foundation’s receipt of payments, equity interests, and/or fees for service >$5,000 from Vertex Pharmaceuticals and other healthcare companies. Dr. Boyle is also an uncompensated Adjunct Professor of Medicine at Johns Hopkins University. 2. Siri Vaeth, MSW, Associate Director, Cystic Fibrosis Research, Inc. (CFRI) I speak on behalf of my daughter with cystic fibrosis, CFRI, the Cystic Fibrosis Engagement Network, and the thousands of people whose lives were quantified in this shocking cost benefit analysis. We share collective outrage with the methodology and outcome of this report. Emily’s Entourage, Rock CF Foundation and the Bonnell Foundation join CFRI in rejecting your assertion that ICER engaged with us as stakeholders. The report radically downplays the life-threatening, horrifying complications of CF. Individuals with CF are suffering. Previously there was no way to stop declining lung function, regardless of adherence to medical regimen. For many who have benefitted from CFTR-modulating drugs, it is the first time they have genuine hope that they will survive. Your report only benefits those seeking to avoid paying for life-saving drugs - not those who are living with, and dying from, this disease. As a rare disease, CF has few therapeutic options. This report has alarming implications, providing payers with justification to refuse coverage of these ©Institute for Clinical and Economic Review, 2018 Page 211 Final Evidence Report – Cystic Fibrosis Return to Table of Contents therapies, making them inaccessible to patients and discouraging investment in and development of new drugs for the CF and rare disease communities. CF is not a manageable disease. An exacerbation is not a minor event. Lung transplants are only life extending and fraught with their own extreme risks. You did not conduct due diligence to understand our experience. You crunched numbers to determine that life-saving therapies which reduce pain and suffering are not cost effective. You bear responsibility for the impact of this deeply flawed report. CFRI provides a broad range of educational, psychosocial, and advocacy programs that receive grant funding from several pharmaceutical companies, including Vertex. 3. Chad Riedy, National Advocacy Co-Chair, Cystic Fibrosis Foundation Being diagnosed with cystic fibrosis in 1984 at the age of three years old, I was not expected live to see my twelfth birthday and for a good portion of my life we did not have therapies like hypertonic saline, inhaled antibiotics, or the vest. I have seen how they have changed the way cf is treated and the difference they have made, but they are the not the answer. They simply help us manage our disease but do not treat the root cause of our disease. CFTR modulators are much closer to the answer and, while not perfect, they are a vital piece. Since being on Symdeko for the past five months, I can carry my kids up the stairs to bed. I can help a neighbor move his couch. I can bike with my family and take a walk to the farmer’s market without fear. While taking it has not lessened the amount of time I still have to spend on treatments and other cf related responsibilities, it has provided a better quality of life right now and real hope for the days to come. Hope that with decreased exacerbations and lung function deterioration I will be able to grow old and gray with my wife, Julie, see Liam and Tate grow up and have many more days enjoying life. You can’t quantify these impacts in data and you can’t quantify time with family. Symdecko has been life-changing for me and my family. CF Fighters deserve to have every medicine and every therapy in their playbook and CFTR modulators are the best play yet. We are fighters. We are hopeful. Mr. Riedy is the National Advocacy Co-Chair at CFF, a volunteer position. CFF paid for Mr. Riedy’s travel expenses to this meeting. ©Institute for Clinical and Economic Review, 2018 Page 212 Final Evidence Report – Cystic Fibrosis Return to Table of Contents 4. Mike Price, Parent of a Child with Cystic Fibrosis At time of publication, ICER has not received a summary of public comment from Mr. Price. No relevant conflicts of interest to report. 5. Juliana Keeping, Parent of a Child with Cystic Fibrosis; Communications Director, Patients for Affordable Drugs My name is Juliana Keeping, and my 5-year-old son, Elijah, has cystic fibrosis. I’m the communications director for Patients For Affordable Drugs, a nonprofit patient advocacy organization that formed in response to soaring prescription drug costs in the U.S. I support ICER’s conclusion that CF drugs are overpriced. Drugs don’t work if people can’t afford them. To demonstrate that point, I’ve brought with me the story of a 40-year-old CF patient named Lora Moser from Austin, Texas. Lora Moser family and friends raised $750,000 for the development of Orkambi, Kalydeco and Symdeko. She has a prescription for Orkambi she can’t fill as the co-pay is $4,400 per month.. Not only is Lora not receiving her medicines, people with CF all over the world are not getting our charity-funded drugs, because these drugs are priced out of reach. The federal government and the Cystic Fibrosis Foundation took on all the risk in creating these medicines, but Vertex priced the drugs as if it alone took on the risk. CF drugs have made multi-millionaires out of Vertex executives like Jeffrey Leiden. According to Axios, his compensation in 2017 alone was $78.5 million. Vertex has also bought back a half billion in its own stock. Instead of buying its shares to enrich investors and paying executives like Leiden tens of millions, Vertex should use its corporate tax breaks to lower the prices of its medicines. The problem is the price. Orkambi, Kalydeco and Symdeko cost far too much. Ms. Keeping owns shares of stock in Vertex Pharmaceuticals. Patients for Affordable Drugs is funded in part by the Laura and John Arnold Foundation, which also provides funding to ICER. ©Institute for Clinical and Economic Review, 2018 Page 213 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Appendix H. Conflict of Interest Disclosure Table H1. ICER Staff and Consultant COI Disclosures Name Organization Disclosures Ethan Balk, MD, MPH Brown University None Rick Chapman, PhD, MS ICER None Geri Cramer, BSN, MBA ICER None Ariel Jurmain, BA ICER None Sonya Khan, MPH ICER None Karen Kuntz, ScD University of Minnesota None Kristin Mickle, MPH ICER None Daniel Ollendorf, PhD ICER None Steven Pearson, MD, MSc ICER None Thomas Trikalinos, MD, PhD Brown University None Kael Wherry, MS University of Minnesota None Ian Williamson, MBA University of Minnesota None Table H2. Midwest CEPAC Panel Member COI Disclosures Name Organization Disclosures Eric Armbrecht, PhD St. Louis University * Ralph Brindis, MD, MPH, MACC, FSCAI, FAHA UCSF * Don Casey, MD, MPH, MBA Medecision, IPO4Health * Rena Conti, PhD University of Chicago * Stacie Dusetzina, PhD Vanderbilt University * Elbert Huang, MD, MPH University of Chicago * Jill Johnson, PharmD University of Arkansas * Timothy McBride, PhD Washington University in St. Louis * Scodd Micek, PharmD St. Louis College of Pharmacy * Reem Mustafa, MD, MPH, MhD University of Kansas * Rachel Sachs, JD, MPH Washington University in St. Louis * Stuart Winston, DO St. Joseph Mercy Health System * * No relevant conflicts of interest to disclose, defined as more than $10,000 in healthcare company stock or more than $5,000 in honoraria or consultancies during the previous year from health care manufacturers or insurers. ©Institute for Clinical and Economic Review, 2018 Page 214 Final Evidence Report – Cystic Fibrosis Return to Table of Contents Name Title COI Declaration Mary Dwight Senior Vice President of CFF provides research and clinical trial support to health Policy and Advocacy care companies, including Vertex Pharmaceuticals. CFF Cystic Fibrosis Foundation has received charitable contributions and/or fees for service >$5,000 from Vertex Pharmaceuticals and other health care companies. Jane Horvath, Senior Policy Fellow Employee of the National Academy for State Health MHSA National Academy for State Policy. Health Policy Manu Jain, MD, MS Professor of Medicine and Member of the Vertex Pharmaceuticals Advisory Board, Pediatrics, and Director of and Site PI for Vertex Phase 2 and 3 studies. Has received Adult CF more than $5,000 in honoraria or consultancies during Feinberg School of the previous year. Medicine, Northwestern University Jeremy Olimb Pastor and father of No conflicts of interest to report. children with cystic fibrosis David Orenstein, Antonio J and Janet No conflicts of interest to report. MD, MA Palumbo Professor of Cystic Fibrosis Children’s Hospital of Pittsburgh Erik Schindler, Manager, Clinical Pharmacy Employee of UnitedHealthCare. PharmD, BCPS UnitedHealthcare Pharmacy ©Institute for Clinical and Economic Review, 2018 Page 215 Final Evidence Report – Cystic Fibrosis Return to Table of Contents