Extended-Release Opioid Agonists and Antagonist Medications for Addiction Treatment (MAT) in Patients with Opioid Use Disorder: Effectiveness and Value Final Evidence Report December 3, 2018 Prepared for ©Institute for Clinical and Economic Review, 2018 ICER Staff and Consultants Reiner Banken, MD, MSc Milon Waththuhewa, PharmD, MSC Senior Fellow Program Manager Institute for Clinical and Economic Review Institute for Clinical and Economic Review Ifeoma Otuonye, MPH Foluso Agboola, MBBS, MPH Research Lead, Evidence Synthesis Director, Evidence Synthesis Institute for Clinical and Economic Review Institute for Clinical and Economic Review Katherine Fazioli, BS Kristin Mickle, MPH Senior Research Assistant Research Lead, Evidence Synthesis Institute for Clinical and Economic Review Institute for Clinical and Economic Review Varun Kumar, MBBS, MPH, MSc Dan Ollendorf, PhD Health Economist Director, Value Measurement & Global Health Initiatives Institute for Clinical and Economic Review Center for the Evaluation of Value and Risk in Health Institute for Clinical Research and Health Policy Studies Rick Chapman, PhD, MS David Rind, MD, MSc Director of Health Economics Chief Medical Officer Institute for Clinical and Economic Review Institute for Clinical and Economic Review Sumeyye Samur, PhD, MS Steve D. Pearson, MD, MSc Health Economist President Institute for Clinical and Economic Review Institute for Clinical and Economic Review Laura Cianciolo, BA Program Associate Institute for Clinical and Economic Review DATE OF PUBLICATION: December 3, 2018 Reiner Banken served as the lead author for the report. Ifeoma Otuonye led the systematic review and authorship of the comparative clinical effectiveness section in collaboration with Katherine Fazioli, Kristin Mickle, and Foluso Agboola. Varun Kumar developed the cost-effectiveness analyses and inputs for the budget impact model. Sumeyye Samur conducted the budget impact analysis and Rick Chapman was responsible for oversight of the cost-effectiveness analyses and the budget impact model. Laura Cianciolo authored the section on coverage policies and clinical guidelines. Dan Ollendorf, David Rind and Steve Pearson provided methodologic guidance on the clinical and economic evaluations. We would also like to thank Alexandra Ellis for her support on the coding and model development, Milon Waththuhewa, Serina Herron-Smith, and Matt Seidner for their contributions to this report. ©Institute for Clinical and Economic Review, 2018 Page ii Final Evidence Report - MAT in Patients with OUD Return to Table of Contents 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. The treatments included in this review are not manufactured by any members of this program. For a complete list of funders and for more information on ICER's support, please visit http://www.icer-review.org/about/support/. About New England CEPAC The New England Comparative Effectiveness Public Advisory Council (CEPAC), a core program of the Institute for Clinical and Economic Review (ICER), is a nationally-recognized community forum. The New England CEPAC convenes three times each year at public meetings to review objective evidence reports and develop recommendations for how stakeholders can apply evidence to improve the quality and value of health care. More information about the New England CEPAC is available here: https://icer-review.org/programs/new-england-cepac/. The New England CEPAC directly engages clinicians, patients, and payers during public meetings to discuss implications of the evidence for clinical decision-making and coverage policies. Application of evidence takes shape through new medical policies, benefit designs, and patient and clinician tools to improve clinical care and patient outcomes. All members of the New England CEPAC must meet ICER’s criteria for Conflict of Interest. 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 iii Final Evidence Report - MAT in Patients with OUD Return to Table of Contents 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/mat-stakeholder-list/. Expert Reviewers Sarah Wakeman, MD Medical Director, Substance Use Disorder Initiative Massachusetts General Hospital, Harvard Medical School 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. Ruth Potee, MD Medical Director Franklin County House of Corrections & Franklin Recovery Center, Greenfield, MA 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. Sean Murphy, PhD, MA Associate Professor of Research Director, CHERISH Consultation Service Weill Cornell Medicine, Department of Healthcare Policy & Research 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. Michael Miller Communications & Chapter Director Young People in Recovery 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 iv Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Table of Contents Executive Summary ................................................................................................................................ 1 Comparative Clinical Effectiveness .................................................................................................... 5 1. Introduction ....................................................................................................................................... 1 1.1 Background .................................................................................................................................. 1 1.2 Scope of the Assessment ............................................................................................................. 8 1.3 Definitions .................................................................................................................................. 11 1.4 Insights Gained from Discussions with Patients and Patient Groups ........................................ 13 1.5. Potential Cost-Saving Measures in Extended-Release Opioid Agonists and Antagonist MAT in Patients with OUD............................................................................................................................ 14 2. Summary of Coverage Policies and Clinical Guidelines ................................................................... 15 2.1 Coverage Policies ....................................................................................................................... 15 2.2 Clinical Guidelines ...................................................................................................................... 17 3. Comparative Clinical Effectiveness .................................................................................................. 20 3.1 Overview .................................................................................................................................... 20 3.2 Methods ..................................................................................................................................... 21 3.3 Results ........................................................................................................................................ 23 3.4 Summary and Comment ............................................................................................................ 48 4. Long-Term Cost Effectiveness .......................................................................................................... 51 4.1 Overview .................................................................................................................................... 51 4.2 Methods ..................................................................................................................................... 52 4.3 Results ........................................................................................................................................ 67 4.4 Summary and Comment ............................................................................................................ 73 5. Potential Other Benefits and Contextual Considerations ................................................................ 76 5.1 Potential Other Benefits ............................................................................................................ 76 5.2 Contextual Considerations ......................................................................................................... 77 6. Value-Based Price Benchmarks ........................................................................................................ 79 7. Potential Budget Impact .................................................................................................................. 80 7.1 Overview .................................................................................................................................... 80 7.2 Methods ..................................................................................................................................... 80 ©Institute for Clinical and Economic Review, 2018 Page v Final Evidence Report - MAT in Patients with OUD Return to Table of Contents 7.3 Results ........................................................................................................................................ 81 7.4 Access and Affordability............................................................................................................. 82 8. Summary of the Votes and Considerations for Policy ..................................................................... 83 8.1 About the New England CEPAC Process .................................................................................... 83 8.2 Voting Results ............................................................................................................................ 85 8.3 Roundtable Discussion and Key Policy Implications .................................................................. 89 References ........................................................................................................................................... 94 Appendix A. Search Strategies and Results........................................................................................ 104 Appendix B. Previous Systematic Reviews and Technology Assessments........................................ 109 Appendix C. Ongoing Studies ............................................................................................................. 111 Appendix D. Comparative Clinical Effectiveness Supplemental Information .................................... 115 Appendix E. Comparative Value Supplemental Information ............................................................. 149 Appendix F. 2014 APA Clinical Guideline ........................................................................................... 164 Appendix G. Public Comments........................................................................................................... 165 Appendix H. Conflict of Interest Disclosures ..................................................................................... 170 ©Institute for Clinical and Economic Review, 2018 Page vi Final Evidence Report - MAT in Patients with OUD Return to Table of Contents List of Acronyms Used in this Report AATOD American Association for the Treatment of Opioid Dependence AE Adverse event AMCP Academy of Managed Care Pharmacy AHRQ Agency for Healthcare Research and Quality ASAM American Society of Addiction Medicine CADTH Canadian Agency for Drugs and Technologies in Health CDF Cumulative distribution function COWS Clinical Opiate Withdrawal Scale DSM Diagnostic and Statistical Manual of Mental Disorders of the American Psychiatric Association ED Emergency department FDA United States Food and Drug Administration HCV Hepatitis C virus HHS United States Department of Health and Human Services HIV Human immunodeficiency virus LCD Local Coverage Determination MAT Medication for addiction treatment and/or medication-assisted treatment NCD National Coverage Determination NIDA National Institute on Drug Abuse OLE Open-label extension OUD Opioid use disorder PCP Primary care physician PICOTS Population, Intervention, Comparators, Outcomes, Timing, and Settings PRISMA Preferred Reporting Items for Systematic Reviews and Meta-Analyses RCT Randomized clinical trial REMS Risk Evaluation and Mitigation Strategy SAE Serious adverse event SAMHSA US Substance Abuse and Mental Health Services Administration SF-36 36-item Short Form Questionnaire SOWS Subjective Opiate Withdrawal Scale TIP Treatment Improvement Protocol UHC UnitedHealthcare UMP Utilization Management Policy US United States USPSTF United States Preventive Services Task Force VAS Visual analogue scale ©Institute for Clinical and Economic Review, 2018 Page vii Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Executive Summary Background Opioid use disorder (OUD) is an increasingly common public health concern that is central to the public health crisis in the US known as the opioid epidemic. In 2016, it was estimated that 2.1 million people suffered from an OUD in the US and 116 Americans died daily from opioid-related drug overdoses.1 and overall life expectancy in the US decreased in 2015 due to the opioid epidemic.2,3 On October 27, 2017, the Acting Secretary of Health and Human Services declared a nationwide public health emergency regarding the opioid crisis4. The Council of Economic Advisers estimates the overall economic cost of the opioid crisis to society to be $504 billion, or 2.8% of US gross domestic product.5 The diagnosis of OUD is based on criteria related to the following dimensions: impaired control, social impairment, risky use, increased tolerance, and withdrawal.6 The diagnostic criteria for moderate to severe OUD roughly correspond to the concept of opioid addiction.7 OUD is to be considered a chronic, treatable illness that requires long-term treatment and is marked by periods of “remission” (reduction in or elimination of signs and symptoms) and relapse. Considering the chronic nature and behavioral impacts of OUD, the primary aim of treatment is recovery rather than cure. A person in recovery refers to an individual who abstains from further use, reduces their substance use to a safer level, or takes steps to mitigate the potential physical and emotional harm resulting from continued use.8 Though some individuals enter and sustain recovery on their own, recovery is mostly achieved via access to evidence-based clinical treatment and recovery support services.9 Misuse of opioids occurs in many different patient subpopulations comprising patients who have followed many different paths to the disorder. In all age groups, medical use of prescription opioids can lead to OUD, but younger adults are more likely to abuse heroin and synthetic opioids, while older individuals are more likely to move from therapeutically-appropriate use of opioids for acute or chronic pain to misuse of those same opioids.10 Overall, OUD patients do present with important psychiatric comorbidities, especially depression, post-traumatic stress disorder, and personality disorders.11 The use of medication for the treatment of OUD is called MAT, for “medication for addiction treatment” (MAT), also known as “medication-assisted treatment”, and is considered as one of the essential elements for countering the opioid epidemic.12. The FDA has approved three medications (in various forms) for the treatment of OUD: methadone, buprenorphine, and naltrexone.13 All three drugs are to be used in combination with counseling and psychosocial support14, described as a “multipronged approach that can include counseling, vocational training, psychosocial therapy, ©Institute for Clinical and Economic Review, 2018 Page ES1 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents family support, and building connections to community resources,”14 also including safe/supportive housing as an essential dimension for many patients. Table ES1 provides an overview of the three FDA-approved drugs for the treatment of OUD. Table ES1. Comparison of Medications for OUD15,16-18 Methadone Buprenorphine Naltrexone Mechanism of Action at mu- Agonist Partial agonist Antagonist Opioid Receptor Medically supervised Maintenance, following Phase of Medically supervised withdrawal, withdrawal, medically supervised Treatment maintenance maintenance withdrawal Sublingual buccal, subdermal Route of Oral, intramuscular Oral implant, subcutaneous extended- Administration extended-release release Usual sublingual/buccal Limited effectiveness of oral Effective Dosage by Usually 60mg–120mg stabilizing dose between 12 mg– naltrexone due to limited Mouth daily 16 mg daily treatment retention Regulation through Not regulated through Controlled Schedule II Schedule III Controlled Substances Act Substances Act Only available in Prescribed by physicians, nurse opioid treatment practitioners and physician Availability programs with Available by prescription assistants with a SAMHSA SAMHSA certification prescribing waiver and DEA registration In a recently published draft guidance document, the FDA recommends using a decrease in opioid use as a primary efficacy endpoint for demonstrating the effectiveness of drugs for OUD. The FDA further states that “sponsors and other stakeholders often mistakenly believe that using a change in drug use patterns as the endpoint always requires complete abstinence.” Long-term studies should demonstrate that observed reductions in drug use predict clinical benefit, even if opioid use has not completely stopped.19 By accepting and recommending a primary endpoint of a clinically relevant decrease in the use of opioids, rather than abstinence, the FDA endorses certain dimensions of “harm reduction strategies” that aim to minimize death, disease, and injury from continuing drug use, with a focus on improving daily social function and productivity. Despite the essential role of MAT in treating OUD and in preventing harm, including death, an important gap persists between the need for and the availability of MAT. More than 30 million people live in US counties without a single prescriber for addiction treatment20 and currently only about 20% of patients with OUD are receiving treatment21 Expanding access to OUD medications is considered an important public health strategy for countering the opioid epidemic.12 ©Institute for Clinical and Economic Review, 2018 Page ES2 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Populations in prisons and jails present a unique challenge for MAT, as regular use of heroin or other opioids is common prior to incarceration. The lack of access to MAT by incarcerated populations drives diversion for self-medication to control withdrawal and cravings. This diversion reinforces negative beliefs about opioid agonist therapy in correctional settings.20 During imprisonment, tolerance of opioids is diminished and the risk for death from overdose is greatly increased upon release. Extended-release formulations have generated clinical interest because of their potential to improve retention in treatment and circumvent some of the access challenges seen with current forms of MAT. These formulations are currently available only for buprenorphine and naltrexone. Table ES2 provides an overview of extended-release medications for OUD that are currently available or under consideration by the FDA. Table ES2. Extended-Release Formulations for OUD Medications Substance Name and Company FDA Approval FDA Recommended Dosing After at least seven days of treatment with a transmucosal buprenorphine‐containing product delivering the equivalent of 8 to 24 mg of Sublocade™, Indivior buprenorphine daily, Sublocade abdominal (Subcutaneous Nov 30, 2017 subcutaneous injections are initiated with 300 mg injection) monthly for the first two months followed by a maintenance dose of 100 mg monthly. Maintenance dose can be increased up to 300mg monthly. Buprenorphine For patients on maintenance treatment with a transmucosal buprenorphine‐containing product Probuphine®, Titan delivering the equivalence of buprenorphine 8 mg Pharmaceuticals, Inc. May 26, 2016 or less per day. Four Probuphine implants (Subdermal implant) inserted subdermally in the upper arm for six months of treatment, after a new insertion in the other arm, transitioned back to a transmucosal. PDUFA date CAM2038, Braeburn expected for (Subcutaneous N/A December 26, injection) 2018 After an opioid-free duration of a minimum of Vivitrol®, Alkermes December 10, seven to 10 days. Administered 380 mg Naltrexone (Intramuscular 2010 for OUD intramuscularly every four weeks or once a injection) month. After completed withdrawal, Vivitrol is administered by a healthcare provider as an intramuscular (IM) gluteal injection, alternating buttocks for each subsequent injection. As with all medications ©Institute for Clinical and Economic Review, 2018 Page ES3 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents for OUD, treatment with Vivitrol should be accompanied by psychosocial support.22 As naltrexone is not regulated by the Controlled Substances Act, Vivitrol can be prescribed without any specific requirements. Treatment with Sublocade replaces a daily dose of buprenorphine with a transmucosal product with extended-release formulation of buprenorphine. For treatment to be initiated, patients need to be on a stable transmucosal dose of 8 to 24 mg buprenorphine for at least seven days. Sublocade is administered through abdominal subcutaneous injection and forms a solid mass upon contact with body fluids. If administered intravenously, it can cause life-threatening pulmonary emboli, as mentioned in a black box warning in the FDA label. Sublocade can only be prescribed by physicians, nurse practitioners, and physician assistants holding a SAMHSA waiver. Treatment with Probuphine implants involves surgical subdermal insertion on the inside of the upper arm of a set of four rods, each 2.5 mm in diameter and 26 mm in length, each rod containing the equivalent of 80 mg of buprenorphine, at steady-state releasing the equivalent of an 8 mg daily transmucosal buprenorphine dose.12 The implants must be removed after six months and a second set of rods can be placed in the other arm. After this second insertion, patients must transition back to a transmucosal buprenorphine-containing product.23 The CAM2038 buprenorphine injection is currently under regulatory review with an expected approval date in December 2018. In clinical studies, this subcutaneous injection has been administered weekly or monthly with multiple dose strengths, in any subcutaneous tissue. The manufacturer proposes to include treatment initiation in the indication for CAM2038.24 If the FDA retains this proposal in the label to be approved, this would eliminate the need for prior treatment with transmucosal buprenorphine. Insights Gained from Discussions with Patients and Patient Groups As part of our review, we spoke with organizations working with individuals and families affected by OUD. There was a consensus that MAT is often difficult to access, in part because of the stigma attached to OUD. This stigma is rooted in a widespread belief that drug addiction is a moral failing rather than a medical condition that is best addressed through treatment. OUD needs to be considered a chronic disease that can affect widely varying populations in terms of age, background, and other factors. The expression, “treatment is not one-size-fits-all,” was used by several organizations to stress the importance of patients having access to different treatment options on their road to recovery; some patients enter recovery without the assistance of MAT, while others require MAT for long periods of time or even their entire lives. Equal access to all types of medications is considered important. For example, we received comments that Vivitrol® (Naltrexone, Alkermes) is currently more easily available than other medications. Buprenorphine extended-release medications are considered important new ©Institute for Clinical and Economic Review, 2018 Page ES4 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents treatment options that could improve recovery and should be widely available for consideration by patients and physicians. It was also mentioned that peer support is particularly important for young people entering the recovery process, as they usually lack a strong existing social network compared to older adults. Several organizations stressed that better daily functioning and well-being, and eventually recovery, are the most important outcomes of treatment. For some this may involve complete abstinence from non-medical opioid use, for others a reduced and controlled level of use. It was mentioned that this corresponds specifically to the discussions at the public meeting on Patient-Focused Drug Development for Opioid Use Disorder convened in April 2018 by the FDA.25 Comparative Clinical Effectiveness Our review focused on the efficacy, safety, and effectiveness of extended-release medication versus each other and versus transmucosal formulations of buprenorphine/naloxone. We do recognize that methadone being used since the 1960s has a very strong evidence base. However, as a schedule II substance regulated through the Controlled Substances Act, it cannot be legally dispensed for MAT through community pharmacies or physician offices, but only as part of highly structured treatment programs. Due to this very different context of use, it has not been chosen in the present assessment as a comparator for the extended-release medications for OUD, but we have identified and summarized previous systematic reviews that are detailed in the report. In this review of the comparative clinical effectiveness of newer, extended-release treatments for MAT (two buprenorphine injections, one buprenorphine implant, and a naltrexone injection), we systematically identified and synthesized the existing evidence from clinical studies. We evaluated studies of patients 16 years or older with OUD. For the comparison of the interventions of interest versus each other and versus transmucosal formulations of buprenorphine/naloxone, we extracted any relevant data, whether in published or unpublished form (e.g., conference abstracts or presentations, FDA review documents). Due to important differences in study characteristics and outcomes assessed, we did not compare the interventions of interest through direct or indirect quantitative assessments. We sought evidence on different outcomes as detailed in Section 1.2 of the report. A detailed description of the methods is available in Section 3.2. Our literature search identified a total of 557 potentially relevant references. Among the 23 references included for the present analysis, 18 references report findings from 11 comparative trials. Five of these trials were identified as key trials evaluating the four drugs of interest and analysed for the comparability of evidence (Table ES3). In four of the key trials, three of the interventions of interest: CAM2038 (Braeburn), Probuphine® (Buprenorphine/Naloxone, Titan Pharmaceuticals Inc.) and Vivitrol were compared to buprenorphine/naloxone, while the remaining one key trial for Sublocade™ (Buprenorphine/Naloxone, Indivior) was placebo-controlled with no ©Institute for Clinical and Economic Review, 2018 Page ES5 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents active comparator. We identified no head-to-head trials of the interventions of interest. The trials and their quality ratings are described in detail in the full report. ©Institute for Clinical and Economic Review, 2018 Page ES6 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Table ES3. Comparability of Evidence: Key Trials Across the Interventions of Interest Treatment Detoxification/ Time of Trial Study Design Duration Outcomes Induction Period Randomization (Weeks) Detoxification: none Phase III RCT • Urine samples used to assess Lofwall At start of CAM2038 Non- 24 abstinence 2018 Induction: one day of 4 mg bup/1 induction inferiority mg nal • Outcome measured over 24 weeks Detoxification: none • Combination of urine samples and Induction: run-in induction phase Trial 13- 24 self-report used to assess abstinence Sublocade Phase III RCT with SL bup/nal film followed by After induction 0001 • Outcome measured over 24 weeks open-label phase with 8 to 24 mg doses of bup/nal for four to 11 days Detoxification: none Phase III • Urine samples and self-report used to Rosenthal Induction: stable dose of 8 mg/day Probuphine Non- 24 After induction assess abstinence 2016 or less of sublingual inferiority • Outcome assessed over 24 weeks buprenorphine received for at least 24 weeks Detoxification: yes, protocols and Before • Abstinence not reported X-BOT Phase IV length of time varied by site induction • Time to relapse event reported Vivitrol Phase III RCT • Urine samples used to assess Tanum After Non- 12 Detoxification: yes abstinence 2017 detoxification inferiority • Outcome measured over 12 weeks Bup/nal: buprenorphine/naloxone, mg: milligram, SL: sublingual ©Institute for Clinical and Economic Review, 2018 Page ES7 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Clinical Benefits Mortality We sought evidence on the effect of the interventions of interest on reducing mortality. However, we found no relevant data on this outcome. All-Cause Discontinuation Discontinuation rates appeared similar with CAM2038, Probuphine, and Vivitrol compared with sublingual buprenorphine/naloxone. However, tests of statistical significance were not reported. Of note, significantly more patients discontinued before induction with Vivitrol compared to buprenorphine/naloxone. Results from the placebo-controlled trials of Sublocade and Probuphine showed substantially greater attrition in the placebo group than in the active treatment arms. The most common reasons for discontinuation were lack of efficacy, adverse events, withdrawing consent, being unable to complete induction, loss to follow-up, and withdrawal symptoms. Abstinence and Relapse Outcomes Abstinence from opioid use was variably defined in available trials. For most interventions, the number of opioid-negative urines did not statistically differ in comparison to sublingual buprenorphine/naloxone. Results from the Probuphine trials showed statistically significantly greater abstinence than buprenorphine/naloxone on various measurements. Participants on Sublocade treatment were also more likely to be abstinent, but in comparison to placebo. Relapse to opioid use was a measure specific to trials of Vivitrol; a statistically significantly higher rate of relapse was seen with Vivitrol versus buprenorphine/naloxone in the intent-to-treat group. Diminishing Illicit Use of Opioids Vivitrol was the only intervention with data on diminishing illicit use of opioids which was assessed in one key trial. That trial found that Vivitrol decreased use of heroin and other illicit opioids when compared to buprenorphine/naloxone over the duration of the trial. Opioid craving – Visual Analog Scale Opioid craving scores on CAM2038 and Probuphine were not statistically significantly different from those on buprenorphine/naloxone. Sublocade decreased opioid craving compared with placebo. One trial found numerically lower opioid craving scores with Vivitrol than buprenorphine/naloxone, but statistical significance was not reported. Opioid Withdrawal No significant differences were shown for CAM2038 and Probuphine each in comparison to buprenorphine/naloxone in the Clinical Opiate Withdrawal Scale (COWS) and Subjective Opiate ©Institute for Clinical and Economic Review, 2018 Page ES8 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Withdrawal Scale (SOWS). Only the higher dose arm of Sublocade showed any significant difference from placebo. There were no COWS or SOWS data for Vivitrol. Health-Related Quality of Life Evidence on health-related quality of life and patient outcomes were reported only in trials of Vivitrol. Results showed an overall increase in quality of life in patients receiving Vivitrol compared with placebo. Patient satisfaction with treatment occurred more with Vivitrol than with buprenorphine/naloxone. Healthcare Utilization Limited data were reported on healthcare utilization, and only for Vivitrol. Evidence from available trials found no differences in healthcare utilization between Vivitrol and treatment as usual. Results from one observational study showed reduced inpatient admissions with Vivitrol. Other outcomes No data on incidence of infectious diseases, functional outcomes, employment-related outcomes, diversion, and accidental pediatric exposure were reported in the trials that met our inclusion criteria. Harms Serious adverse events were generally uncommon and similar in trials of CAM2038, Probuphine, and Vivitrol in comparison to buprenorphine/naloxone and in the Sublocade trial vs placebo. Low numbers of participants discontinued due to adverse events in the trials of CAM2038, Probuphine, and Vivitrol when compared to buprenorphine/naloxone. The most common adverse events reported in the trials were injection/implant site pain, gastrointestinal issues, headaches, and insomnia (see Table 3.5 in report). Controversies and Uncertainties As mentioned previously, differences in trial designs, population selection, comparators, and outcome measures precluded formal comparisons between the different extended-release formulations. All four formulations also differ in their labeled or potential treatment indications; for example, only CAM2038 has the possibility of starting OUD treatment directly after diagnosis. Sublocade and Probuphine must be preceded by daily transmucosal use of buprenorphine and Vivitrol by a period of medically supervised opioid withdrawal. Probuphine implants should be used for patients on maintenance treatment with a transmucosal buprenorphine‐containing product delivering a low to moderate dose, the equivalent of buprenorphine 8 mg or less per day. The effective required buprenorphine dosage for most patients is between 12 and 16 mg daily, therefore only patients who can tolerate such doses may be suitable for Probuphine implants. ©Institute for Clinical and Economic Review, 2018 Page ES9 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Various outcome measures were used in the trials of the interventions of interest. Outcome measures are based on different calculations of negative urine samples (Appendix Table D5), using the term relapse to designate a certain percentage of positive urine samples. However, the term relapse refers to a person with OUD who is being treated and is in remission experiences a loss of control. A relapse is different from a return to opioid use that is limited in scope and time and that does not involve the return of the signs or symptoms of OUD. It is not certain to which degree different rates of negative urine samples constitute a meaningful measure of success, even for the short duration of the trials. The lack of any clear guidance on how to obtain the opioid-free state needed for starting Vivitrol makes comparisons between the evidence for the extended-release agonist formulations and the extended-release antagonist formulation very difficult. Head-to-head trials of agonist formulations should be possible but have not yet been conducted. In the real world, OUD patients often present with important psychiatric comorbidities, such as depression, post-traumatic stress disorder, and personality disorders.11 Patients with psychiatric comorbidities are largely excluded from the trials (refer to Appendix Table D2), thus limiting their generalizability. This is not limited to the evidence on extended-release formulations but present in the evidence base for all MATs.26 As noted by SAMHSA in the 2018 TIP, no evidence clearly predicts which patients are best treated with Vivitrol versus methadone or buprenorphine formulations.12 The treatment sequences for different subpopulations with OUD cannot be based solely on the available evidence, but rather must be informed by clinical knowledge and the local context. The evidence on the use of the extended-release formulations is subject to the same general limitations as for the other medications for OUD. It is not yet known if or when to best taper these medications,12 and evidence is lacking on the added value of the different types of counseling and psychosocial support required by the FDA label the most recent clinical practice guideline.12 The available research focuses on short-term outcomes and does not provide any evidence of observed reductions or patient control of drug use that are of clinical and social benefit, even if opioid use has not completely stopped.19,27 In addition, questions around the impact of extended- release formulations on critically important outcomes, such as overdose and other OUD-associated mortality, health-related quality of life, work productivity, educational attainment, and incarceration have largely gone unanswered by the evidence currently available. Summary and Comment Using the ICER Evidence Matrix shown in Section 3.4 of the report, we assigned evidence ratings independently for each of the interventions of interest compared to transmucosal buprenorphine/naloxone for study participants with OUD being considered for MAT. ©Institute for Clinical and Economic Review, 2018 Page ES10 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Table ES4. Evidence Ratings (Versus Transmucosal Buprenorphine/Naloxone) Comparisons Evidence Rating CAM2038 C+ Sublocade I Probuphine P/I Vivitrol C CAM2038 Evidence for CAM2038 is comprised of one 24-week Phase III trial in comparison to buprenorphine/naloxone. Results showed CAM2038 to be non-inferior to buprenorphine, but not significantly different in abstinence, opioid craving, and opioid withdrawal. For participants with OUD being considered for MAT, we have moderate certainty that CAM2038 provides a small, or substantial net health benefit given the increased convenience and provider interaction associated with subcutaneous injections, but high certainty that it is at least comparable as it is a buprenorphine-containing treatment. Therefore, we consider the evidence on CAM2038 to be comparable or better (C+). Sublocade Evidence for Sublocade is limited to one 24-week Phase III trial compared to placebo. In the absence of a direct comparison of Sublocade to buprenorphine/naloxone, we consider the evidence on Sublocade compared to buprenorphine/naloxone to be insufficient (I). Probuphine Evidence for Probuphine compared to buprenorphine/naloxone comprises two 24-week Phase III trials. For participants with OUD being considered for MAT, we have moderate certainty of a comparable or small net health benefit for the trial populations. However, we have concerns that the study population may not be reflective of the more general population being considered for MAT. Therefore, we consider the evidence on Probuphine in comparison to buprenorphine/naloxone to be promising but inconclusive (P/I). Vivitrol Evidence for Vivitrol compared to buprenorphine/naloxone consists of data derived from two trials: one 24-week Phase IV trial, and one shorter 12-week Phase III trial. Results showed that Vivitrol is non-inferior to buprenorphine/naloxone on a variety of abstinence outcomes. Vivitrol has the most mature evidence base of any of the treatments of focus for this review. Differences observed between Vivitrol and buprenorphine/naloxone are due at least in part to differences in treatment ©Institute for Clinical and Economic Review, 2018 Page ES11 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents intent and goals. Therefore, we considered the evidence on Vivitrol in comparison to buprenorphine/naloxone to have high certainty of a comparable net health benefit (C). Long-Term Cost Effectiveness We conducted an economic evaluation to estimate the cost-effectiveness of the interventions of interest among adult patients considered for OUD treatment, from a U.S. health sector perspective. Costs and outcomes in the model were discounted at 3% annually, and the model had four-week cycle lengths and was run over a five-year time horizon. Our model compared buprenorphine extended-release subcutaneous injections (CAM2038 [investigational]), extended-release injectable naltrexone (Vivitrol), and buprenorphine subdermal implant (Probuphine), to a transmucosal buprenorphine/naloxone, specifically generic sublingual (SL) buprenorphine/naloxone in the base case analysis; another extended-release subcutaneous injection (Sublocade) was compared to SL buprenorphine/naloxone in a scenario analysis. We developed a decision tree with pre-MAT initiation rules for each intervention of interest based on pre-treatment induction/detoxification protocols in the key clinical trials and FDA labels, in keeping with an intention-to-treat perspective. MAT-specific initial state probabilities in the subsequent Markov model were assigned based on pre-MAT initiation outcomes (Figure 4.1A) for each MAT. The Markov model comprised five health states, namely, “MAT with Illicit Use of Opioids”, “MAT with NO Illicit Use of Opioids”, “OFF MAT with Illicit Use of Opioids”, “OFF MAT with NO Illicit Use of Opioids” and “Death” (Figure 4.1B). Patient flow through the model was unidirectional, in that once in a health state, patients could not move to a previously occupied health state. For patients treated with CAM2038, or Sublocade, and their respective comparators, those successful in the pre- MAT initiation started the model in the “MAT with Illicit Use of Opioids” health state, while those who were unsuccessful entered the model in the “OFF MAT with Illicit Use of Opioids” health state. With increasing abstinence over time, those successful transitioned to the “MAT with NO Illicit Use of Opioids” health state, where they could remain or transition over to permanent abstinence from illicit use (“OFF MAT with NO Illicit Use of Opioids”) or relapse (“OFF MAT with Illicit Use of Opioids). Patients treated with Vivitrol, or Probuphine, and their respective comparators upon success with pre-treatment protocols entered the model in the “MAT with NO Illicit Use of Opioids” health state and could transition to a state of permanent abstinence from illicit use or relapse. Patients could die from all-cause mortality from any health state, or from opioid-related overdose while illicitly using opioids and not on MATs. The modeled cohort focused on adult patients diagnosed with OUD seeking MATs and had a mean age of 36 years; 30% of the cohort were female and there was a 50:50 split of illicit use of prescription opioids and injection drugs. Some of the key modeling assumptions included: • Return to pre-treatment choice of illicit use of opioids (prescription or injection), upon relapse following MAT ©Institute for Clinical and Economic Review, 2018 Page ES12 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents • 10% of all patients who remained abstinent from illicit use while on MATs for one year or more could transition to permanent abstinence from illicit use • A single cost and utility were chosen for illicit use (by prescription or injection, and concurrent use of MATs); illicit use was not modeled according to different levels of use • Constant mortality from opioid overdose irrespective of duration of illicit use • Incidence of HIV and HCV infections only among persons who inject drugs (PWID) and not among those who illicitly used prescription opioids • Serious adverse events of illicit use that were not related to overdose were not included A detailed description of model choices, assumptions, and respective rationale for each can be found in Section 4, Table 4.3 of this report. Treatment efficacy estimates, namely, abstinence from illicit use of opioids for CAM2038 and its comparator, and relapse to illicit use of opioids for Vivitrol and Probuphine and their respective comparators were derived from the respective key clinical trials. Treatment discontinuation was also derived from trials. Wherever Kaplan-Meier (KM) curves available, they were digitized and fit with parametric curves that fit the digitized data, and were extrapolated beyond the trial duration, for the modeled time horizon. Mortality in the model was a function of background mortality increased by overdose-related mortality among those who illicitly used opioids without MATs, as well as mortality from HCV or HIV infection among PWID. Health state utilities were sourced from a US-specific cross-sectional survey study where available, with calculations made to estimate utilities when on MATs and illicitly using opioids using data from a UK-specific study. We applied disutility multipliers in PWID with HIV or HCV to reflect these comorbidities and treatment associated with them. Due to a variation in reporting of adverse events (AE) among trials and their being reported by severity, and AE-related costs assumed to not impact overall costs substantially, we did not include AEs in the model. MAT drug costs in the model were sourced from the Federal Supply Schedule database for all except Vivitrol, which was provided to us as a net price by the drug’s manufacturer. For the SL buprenorphine/naloxone comparator, we used the undiscounted WAC in accordance with ICER’s Reference Case. We included administration and monitoring costs for all interventions of interest as appropriate. Non-drug health care costs were sourced from a claims analysis and awarded to health states in the model as appropriate. We also included costs for specific HIV and HCV-related treatment as well as non-drug treatments costs for these co-morbidities. For the modified societal perspective, we included the costs of lost productivity, and the costs of criminal justice and incarceration that were applied to the percentage of patients assumed employed or involved in crime-related activities in this hypothetical patient cohort. Details regarding inputs, sensitivity analyses, and scenario analyses are available in Section 4 of the report. We ran scenario analyses for a Sublocade vs. SL buprenorphine/naloxone comparison that ©Institute for Clinical and Economic Review, 2018 Page ES13 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents included separate threshold analyses for drug price and efficacy for Sublocade to achieve cost- effectiveness threshold of up to $150,000 per QALY, and one where we assumed treatment efficacy and adherence the same as those seen in the CAM2038, with Sublocade’s FSS price and also a favorable assumption reflecting 100% success with pre-Sublocade treatment induction. Base-Case Results Treatment with any of the interventions of interest resulted in higher costs relative to SL buprenorphine/naloxone, while QALY gains were seen only for CAM2038 and Probuphine. We do not report the incremental costs of CAM2038 relative to its comparator since CAM2038 does not yet have a known price, and any incremental cost is only based on non-drug costs. All interventions and their respective comparators showed similar life-year outcomes (4.62 years). Table ES5. Incremental Cost Effectiveness Results of the Interventions of Interest Versus Respective SL Generic Buprenorphine/Naloxone Comparators Incremental Cost-Effectiveness Intervention Incremental Costs Incremental QALYs Ratio (Cost per QALY gained) CAM2038 -- 0.06 -- More costly, less effective Vivitrol $10,300 (0.03) (dominated) Probuphine $2,700 0.01 $265,000 QALY: quality-adjusted life year All incremental results are versus each interventions respective SL buprenorphine/naloxone comparator, over a five-year time horizon. Sensitivity Analyses One-way sensitivity analyses showed that results were generally most sensitive to intervention discontinuation rate (relapse to illicit use of opioids), the incidence of HCV, and intervention costs. Results of our probabilistic analyses showed that none of the 1,000 simulations resulted in incremental cost-effectiveness ratios that were at or below the $150,000 per QALY threshold for Vivitrol, while only 12.4% of simulations resulted in incremental cost-effectiveness ratios that were at or below the same threshold for Probuphine. We do not report probabilistic results for CAM2038 since we have no price for this treatment and cannot calculate an incremental cost- effectiveness ratio. ©Institute for Clinical and Economic Review, 2018 Page ES14 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents In the Sublocade versus generic SL buprenorphine/naloxone scenario, a threshold analysis for price showed per cycle Sublocade costs similar to that of CAM2038’s, assuming Sublocade’s efficacy to be the same as CAM2038’s. The second threshold analysis examining the efficacy required for Sublocade to reach a $150,000 per QALY cost-effectiveness threshold showed that even if patients treated with Sublocade achieved complete abstinence from illicit use of opioids with 100% treatment adherence, at its current price the cost-effectiveness of Sublocade would still exceed $215,000 per QALY. A third scenario comparing Sublocade to generic SL buprenorphine/naloxone that assumed Sublocade’s efficacy to be the same as CAM2038’s, using Sublocade over a five-year time horizon resulted in an incremental cost-effectiveness ratio of approximately $577,000 per QALY gained. Modeling interventions from a societal perspective resulted in higher total costs, with incremental results directionally similar to the base case findings. All other scenario analyses produced results similar to the base-case analyses, except for when Vivitrol was analyzed using a “per protocol” approach, which resulted in its cost-effectiveness being approximately $1 million per QALY. Threshold Analyses We could not calculate threshold prices for Vivitrol based on the base case estimates since it was less effective relative to its comparator. Threshold analyses for Cam2038 and Probuphine were calculated at the $50,000, $100,000, and $150,000 per QALY willingness-to-pay (WTP) thresholds (Table ES6). Table ES6. Threshold Analysis Results WAC per Net Price Unit Price to Achieve Unit Price to Achieve Unit Price to Achieve Unit per Unit $50,000 per QALY $100,000 per QALY $150,000 per QALY CAM2038* - - $219† $313† $406† Probuphine $4,950‡ $3,640‡ $1,165‡ $1,741‡ $2,318‡ *No list or net prices for CAM2038 were available as of the date of this report. † Price per four-week dose. ‡ Price per implant lasting six months. Summary and Comment Our analyses indicate that all the interventions of interest generate similar life years, with only marginal differences in QALYs relative to their respective comparators. Only CAM2038 and Probuphine produce incremental QALYs relative to the respective comparators while Vivitrol does not. Recognizing that Vivitrol is used in a population with specific treatment goals and intent that are different from those associated with the other MATs, we analyzed its cost-effectiveness in a “per protocol” scenario which resulted in an incremental cost-effectiveness ratio of approximately $1 million per QALY, well above commonly accepted cost-effectiveness thresholds of $100,000 to $150,000 per QALY. The findings remained robust in most sensitivity and scenario analyses. While ©Institute for Clinical and Economic Review, 2018 Page ES15 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents evidence was inadequate to compare Sublocade to SL buprenorphine/naloxone, even under extremely favorable assumptions its cost effectiveness exceeded commonly accepted thresholds. We were limited by the lack of data for effective modeling around patients cycling though multiple MATs and intervention re-use. We lacked treatment-specific quality-of-life estimates and robust data on diversion and switching to other opioids, and differences in trial designs prevented the use of a comparator with normalizable efficacy estimates. The findings of our analysis suggest that the interventions of interest result in only marginal changes in QALYs relative to generic SL buprenorphine/naloxone, but universally higher costs, with resulting ratios when calculable, well above commonly-cited thresholds of $50,000 to $150,000 per QALY gained. Potential Other Benefits and Contextual Considerations Our reviews seek to provide information on potential 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. Potential Other Benefits Table ES7. Potential Other Benefits Other Benefits Description This intervention offers reduced complexity Extended-release formulations are important additional that will significantly improve patient treatment options that could improve long term recovery by outcomes. lowering the constraints of daily adherence to transmucosal buprenorphine formulations. This intervention will reduce important health In correctional settings extended-release formulations offer disparities across racial, ethnic, gender, socio- the potential of decreasing diversion and may diminish economic, or regional categories. negative beliefs about opioid agonist therapy and improve general access to MAT for inmates. Regulator could consider not to subject extended-release formulations to waivers in the future, thus increasing overall and regional access to MAT This intervention will significantly reduce N/A caregiver or broader family burden. This intervention offers a novel mechanism of N/A action or approach that will allow successful treatment of many patients for whom other available treatments have failed. ©Institute for Clinical and Economic Review, 2018 Page ES16 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Other Benefits Description This intervention will have a significant impact N/A on improving return to work and/or overall productivity. Other important benefits or disadvantages Administration by a health professional can contribute to that should have an important role in prevent accidental poisoning in children that currently occurs judgments of the value of this intervention. with transmucosal products. Contextual Considerations Table ES8. Potential Contextual Considerations Contextual Consideration Description This intervention is intended for the care of OUD is considered a public health emergency with an individuals with a condition of particularly high epidemic of deaths that decrease the overall life severity in terms of impact on length of life and/or expectancy in the US. Providing access to extended- quality of life. release medications, can contribute to diminish the consequences of the opioid epidemic. This intervention is intended for the care of OUD is a chronic disease that that carries a stigma individuals with a condition that represents a affecting self-esteem, social relations, and work. particularly high lifetime burden of illness. Extended-release formulations could improve long-term care. This intervention is the first to offer any N/A improvement for patients with this condition. Compared to transmucosal formulations of N/A buprenorphine/naloxone, there is significant uncertainty about the long-term risk of serious side effects of this intervention. Compared to transmucosal formulations of There is significant uncertainty about the magnitude or buprenorphine/naloxone, there is significant durability of the long-term benefits of extended-release uncertainty about the magnitude or durability of formulations, given the 6-month duration of nearly all the long-term benefits of this intervention. trials of these agents. Probuphine implants cannot be used for longer than 12 months according to the FDA label. For the other formulations, their duration of appropriate use is unknown and will only be better defined through clinical experience and long-term observational study. There are additional contextual considerations that For antagonist therapy with Vivitrol, its action cannot be should have an important role in judgments of the reversed, so it becomes impossible to use opioids for value of this intervention. emergency pain management. Regional analgesia or non- opioid analgesics need to be used ©Institute for Clinical and Economic Review, 2018 Page ES17 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Value-Based Benchmark Prices We calculated value-based prices for CAM2038 and Probuphine (Table ES9). Since Vivitrol was less effective relative to its comparator in the base case, and since we did not have adequate data to model Sublocade versus SL buprenorphine/naloxone in the base case analysis, we did not estimate their value-based prices. Table ES9. Value-Based Benchmark Prices for Cam2038 and Probuphine Annual Price to Achieve Price to Achieve Discount from WAC Required to WAC $100,000 per QALY $150,000 per QALY Achieve Threshold Prices CAM2038* -- $4,082† $5,301† -- Probuphine $4,950 ‡ $1,741 ‡ $2,318 ‡ 53% to 65% QALY: quality-adjusted life year *No list or net prices for CAM2038 were available as of the date of this report. † Annual price. ‡ Price per implant lasting six months. Probuphine implant cannot be used more than twice in the treatment for OUD for each patient. Potential Budget Impact We used results from the cost-effectiveness model to estimate the potential total budgetary impact of CAM2038 in the patients aged 18 years and above with OUD. Given the presence of other MATs in the US marketplace for over a year, we excluded them from the budget impact analysis. In the absence of a list or net price or any published price estimate for CAM2038, we calculated its budget impact using only the prices to reach WTP thresholds between $50,000 to $150,000 per QALY. We applied the 2015 OUD prevalence estimate to the projected 2018 to 2022 US population to calculate the candidate population for treatment with CAM2038. This resulted in approximately 312,000 treatment eligible patients each year. The per-patient budget impact using CAM2038’s prices to reach $150,000, $100,000, and $50,000 per QALY gained WTP thresholds for CAM2038 ($5,301, $4,082, and $2,863 per year, respectively) compared to generic buprenorphine/naloxone are presented below in Table ES10. Table ES10. Per-Patient Budget Impact Calculations Over a Five-Year Time Horizon Average Annual Per Patient Budget Impact $150,000/QALY $100,000/QALY $50,000/QALY CAM2038 $35,420 $33,883 $32,346 Generic $31,653 Buprenorphine/Naloxone Difference $3,768 $2,231 $694 QALY: quality-adjusted life year ©Institute for Clinical and Economic Review, 2018 Page ES18 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents As shown in Figure ES11, only 24% and 40% of the entire population could be treated each year at the prices that would reach the $150,000 to $100,000 per QALY thresholds respectively, before the total budget exceeded the ICER annual budget impact threshold. The entire eligible population could be treated each year at the price that would reach $50,000 per QALY. Figure ES11. Potential Budget Impact Scenarios at Different Prices of CAM2038 to Treat Adults with OUD $6,000 $150,000 per QALY $5,000 $100,000 per QALY $4,000 Annual Price $3,000 $2,000 $1,000 $0 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Percentage of Patients Treated without Crossing BI Threshold Each Year In the absence of a list or net price for CAM2038 as of the date of publication of this report, its potential budget impact is uncertain. However, if it is priced similarly to Sublocade, the expectation would be that the budget impact of CAM2038 would be offset by lower use of Sublocade. ICER is not issuing an access and affordability alert at this time. However, as use of both agents may increase over time, health systems likely to be covering large numbers of patients with OUD may wish to pay close attention to the actual use and costs of extended release injectable buprenorphine whether administered as Sublocade or CAM2038. New England Comparative Effectiveness Public Advisory Council Votes The New England Comparative Effectiveness Public Advisory Council (NE CEPAC) deliberated on key questions raised by ICER’s report at a public meeting on November 8th, 2018 in Newton, Massachusetts. The results of these votes are presented below, and additional information on the deliberation surrounding the votes can be found in the full report. ©Institute for Clinical and Economic Review, 2018 Page ES19 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Patient population for all questions: Patients 16 years or older with opioid use disorder, who are being considered for MAT. 1) Is the evidence adequate to demonstrate that the net health benefit of the buprenorphine subcutaneous extended-release injection Sublocade (Indivior) is superior to that provided by transmucosal formulations of buprenorphine/naloxone? Yes: 0 votes No: 13 votes 2) Is the evidence adequate to demonstrate that the net health benefit of the buprenorphine subcutaneous extended-release injection CAM2038 (Braeburn) is superior to that provided by transmucosal formulations of buprenorphine/naloxone? Yes: 4 votes No: 9 votes 3) Is the evidence adequate to demonstrate that the net health benefit of the buprenorphine subcutaneous extended-release injection Probuphine (Titan Pharmaceuticals Inc.) is superior to that provided by transmucosal formulations of buprenorphine/naloxone? Yes: 1 vote No: 12 votes 4) Is the evidence adequate to demonstrate that the net health benefit of the Naltrexone subcutaneous extended-release injection Vivitrol (Alkermes) is superior to that provided by transmucosal formulations of buprenorphine/naloxone? Yes: 2 votes No: 11 votes 5) Is the evidence adequate to distinguish the net health benefit among the following interventions: (1) the two buprenorphine subcutaneous extended release injections (Sublocade and CAM2038); (2) the buprenorphine implant (Probuphine); (3) naltrexone intramuscular extended-release injection (Vivitrol)? Yes: 1 vote No: 12 votes ©Institute for Clinical and Economic Review, 2018 Page ES20 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents 6) Does treating patients with one of the extended-release interventions (CAM2038, Sublocade, Probuphine, or Vivitrol) offer one or more of the following potential “other benefits” vs transmucosal formulations of buprenorphine/naloxone? CAM2038 and Sublocade offer reduced complexity that will significantly improve patient 10/13 outcomes. Probuphine offers reduced complexity that will significantly improve patient outcomes. 4/13 Vivitrol offers reduced complexity that will significantly improve patient outcomes. 8/13 These interventions will reduce important health disparities across racial, ethnic, gender, 7/13 socioeconomic, or regional categories. These interventions will significantly reduce caregiver or broader family burden. 6/13 CAM2038 and Sublocade offer a novel mechanism of action or approach that will allow 7/13 successful treatment of many patients for whom other available treatments have failed. Probuphine offers a novel mechanism of action or approach that will allow successful 3/13 treatment of many patients for whom other available treatments have failed. Vivitrol offers a novel mechanism of action or approach that will allow successful treatment 8/13 of many patients for whom other available treatments have failed. These interventions will have a significant impact on improving patients’ ability to return to 5/13 work and/or their overall productivity. There are other important benefits or disadvantages that should have an important role in 7/13 judgments of the value of these interventions. ©Institute for Clinical and Economic Review, 2018 Page ES21 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents 7) Are any of the following contextual considerations important in assessing the long-term value for money of the extended-release interventions (CAM2038, Sublocade, Probuphine, or Vivitrol)? These interventions are intended for the care of individuals with a condition of particularly 11/12 high severity in terms of impact on length of life and/or quality of life. These interventions are intended for the care of individuals with a condition that represents 12/12 a particularly high lifetime burden of illness. These interventions are the first to offer any improvement for patients with this condition. 0/12 There is significant uncertainty about the long-term risk of serious side effects of CAM2038. 2/12 There is significant uncertainty about the long-term risk of serious side effects of Sublocade. 3/12 There is significant uncertainty about the long-term risk of serious side effects of 3/12 Probuphine. There is significant uncertainty about the long-term risk of serious side effects of Vivitrol. 3/12 There is significant uncertainty about the magnitude or durability of the long-term benefits 9/12 of CAM2038. There is significant uncertainty about the magnitude or durability of the long-term benefits 9/12 of Sublocade. There is significant uncertainty about the magnitude or durability of the long-term benefits 2/12 of Probuphine. There is significant uncertainty about the magnitude or durability of the long-term benefits 9/12 of Vivitrol. There are additional contextual considerations that should have an important role in 6/12 judgments of the value of this intervention. Note: Only 12 Council members voted on this question because one member had to leave early due to a medical emergency. Long-Term Value for Money As described in ICER’s value assessment framework, questions on “long-term value for money” are subject to a value vote only when incremental cost-effectiveness ratios for the interventions of interest are between $50,000 and $175,000 per QALY in the primary “base case” analysis. As shown in the analysis, the estimates for Probuphine, Sublocade, and Vivitrol exceed the higher end of the range. Consequently, all three interventions were deemed “low value” without formal voting by the public Council. CAM2038 was not yet approved at the time of the meeting, and no price was available, so an incremental cost-effectiveness ratio could not be calculated; as a consequence, a value vote was not taken. Key Policy Implications Following its deliberation on the evidence, the NE CEPAC engaged in a moderated discussion with a policy roundtable about how best to apply the evidence on extended-release MATs for patients with Opioid Use Disorder to policy and practice. The policy roundtable members included two clinical experts, two payer representatives, representatives from Indivior, Alkermes and Braeburn, ©Institute for Clinical and Economic Review, 2018 Page ES22 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents and an expert on the treatment of substance use disorders in criminal justice settings (one patient advocate was invited to participate in the roundtable, but did not attend the meeting). 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 top-line policy implications are presented below, and additional information can be found in Section 8.3. All Stakeholders • Decrease stigma by aligning efforts around education that enhances awareness that Opioid Use Disorder (OUD) is a chronic disease requiring long-term treatment. Manufacturers • Bring the price of extended-release medications into alignment with their clinical value. Payers • For treatments for OUD whose prices are aligned with clinical value, payers should create coverage criteria that present no barriers to access. In particular, prior authorization criteria for Sublocade and similar extended-release treatments should be flexible enough to support evidence-based individualized treatment decisions. Payers and Policy Makers • Avoid legislative action favoring one form of MAT • Coordinate MAT for individuals leaving the correctional system and ensure continuity of care • Take action to address the many long-term policy goals that are yet to be achieved in order to improve treatment for OUD. Regulators • Consider eliminating restrictions on prescribing extended-release formulations of buprenorphine Researchers • The research community should work with clinicians and manufacturers to identify clinical characteristics that would better predict which OUD patients are likely to benefit most from the available MAT approaches ©Institute for Clinical and Economic Review, 2018 Page ES23 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents 1. Introduction 1.1 Background Opioid use disorder (OUD) is an increasingly common public health concern that is central to the public health crisis in the US known as the opioid epidemic. In 2016, it was estimated that 2.1 million people suffered from an OUD in the US and 116 Americans died daily from opioid-related drug overdoses.1 Overall life expectancy in the US began to decrease in 2015 due to the opioid epidemic,2 and this trend continued through 2016, the first such decrease since the 1960s.3 On October 27, 2017, the Acting Secretary of Health and Human Services declared a nationwide public health emergency regarding the opioid crisis4. The Council of Economic Advisers estimates the overall economic cost of the opioid crisis to society to be $504 billion, or 2.8% of US gross domestic product.5 Treatment of OUD that is grounded in the use of medication, collectively known as medication for addiction treatment (MAT), has received increasing attention in recent years as one of the essential elements for countering the opioid epidemic. (Note: this term is used interchangeably with “medication-assisted treatment;” we refer to both in this report through the MAT acronym.) The 2010 Affordable Care Act increased access to substance abuse treatment at the federal level, both for commercial plans and Medicaid. In 2015, the Department of Health and Human Services created a specific grant program for MAT that extended to 11 states and expanded to others in subsequent years. The 21st Century Cures Act of 2016 allocated $1 billion over two years to enhance states’ response to the epidemic28 and recent state legislation has been enacted to increase access to MAT, either by expanding OUD treatment programs or enhancing health insurance coverage.29 Initiatives in New England states are often considered models for other programs: Vermont’s “Hub and Spoke” is referred to as a success for integrating treatment facilities and programs into its health care system,29-32 and Rhode Island is a leading example for providing access to MAT in correctional facilities29,33. In 2014, ICER conducted an assessment on clinical, delivery system, and policy options for the management of patients with opioid dependence.34 The report found that “long-term maintenance treatment approaches using methadone or Suboxone® (Indivior) to reduce the craving for opioids have been found to be more effective than short-term managed withdrawal methods that seek to discontinue all opioid use and detoxify patients” and concluded that coordinated efforts are needed to improve access to opioid dependence treatment. Since that initial review, newer, extended- release medications for addiction treatment have been approved or are currently undergoing regulatory review. The present report does not seek to revisit the policy challenges and options highlighted in 2014, but to specifically assess the effectiveness and value of these newer medication options in patients with OUD. ©Institute for Clinical and Economic Review, 2018 Page 1 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Opioid Use Disorder Opioids are substances that act on specific receptors in the brain and produce various effects such as pain relief, euphoria, respiratory depression, and constipation.35 They are either extracted from opium, obtained from the pods of poppy varieties, or produced semi-synthetically or synthetically. Opioids reduce pain by affecting the mu receptor in the brain and spinal cord.36 The mu receptor in the brain is also central to the feelings of reward or pleasure, leading to abuse.37 The analgesic effects are mediated mainly through the spinal mu receptor’s release of substance P,36,38 the central neurotransmitter for pain. The rewarding effect involves the dopaminergic system, which is implicated in all addictive behaviors including those of alcohol and nicotine.39 The concepts and terminology around illicit drug use are constantly evolving. In the 1980s, a committee of experts convened by the American Psychiatric Association defined by consensus a set of diagnostic criteria for compulsive, uncontrolled, drug-seeking behavior. However, some members of the committee felt that using the term “addiction” to define this behavior constitutes a moral judgment that stigmatizes patients. The term “dependence” was chosen instead, even though this term was not directly related to the physical dependence that leads to withdrawal symptoms by abrupt cessation, rapid dose reduction, or administration of an opioid antagonist. The term was used in versions III and IV of the American Psychiatric Association’s Diagnostic and Statistical Manual of Mental Disorders of the American Psychiatric Association (DSM) as a term for the compulsive, uncontrolled, drug-seeking behavior that is known to others as addiction. An intermediate state between drug use and dependence called “abuse” was also created in DSM-III.40 In 2013, DSM-5 replaced the categories of substance abuse and dependence with a single classification of OUD, based on criteria related to the following dimensions: impaired control, social impairment, risky use, increased tolerance, and withdrawal.6 The language of OUD is now generally accepted and has led to a general change in “terminology that will not reinforce prejudice, negative attitudes, or discrimination.”12 OUD is generally considered to be a chronic, treatable illness that requires long-term treatment and is marked by periods of “remission” (reduction in or elimination of signs and symptoms) and relapse. Considering the chronic nature and behavioral impacts of OUD, the primary aim of treatment is recovery rather than cure. Recovery is defined as a process of change through which individuals improve their wellness and health, live self-directed lives, and strive to reach their full potential. A person in recovery refers to an individual who abstains from further use, reduces their substance use to a safer level, or takes steps to mitigate the potential physical and emotional harm resulting from continued use.8 Though some individuals enter and sustain recovery on their own, recovery is mostly achieved via access to evidence-based clinical treatment and recovery support services.9 Misuse of opioids occurs in many different patient subpopulations comprising patients who have followed many different paths to the disorder. For example, recreational users obtain and use ©Institute for Clinical and Economic Review, 2018 Page 2 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents opioids to get high, but as they do not use the drugs in a compulsive and uncontrolled manner, they are not considered to have OUD.41 This does not mean that they do not need support and treatment, however, as recreational use is considered one of the precursors to OUD. In all age groups, medical use of prescription opioids can lead to OUD, but younger adults are more likely to abuse heroin and synthetic opioids, while older individuals are more likely to move from therapeutically-appropriate use of opioids for acute or chronic pain to misuse of those same opioids.10 Overall, OUD patients do present with important psychiatric comorbidities, especially depression, post-traumatic stress disorder, and personality disorders.11 Medications for Addiction Treatment The 2014 assessment by ICER stressed the importance of medication-based long-term maintenance treatments, and their superiority over medication-free “detoxification” protocols.34 The central role of medications in the treatment of OUD has been confirmed by more recent assessments.42 The Treatment Improvement Protocol (TIP) for Medications for OUD published in 2018 by the Substance Abuse and Mental Health Services Administration (SAMHSA) states as one if its key messages that “the science demonstrating the effectiveness of medication for OUD is strong.”12. The FDA has approved three medications (in various forms) for the treatment of OUD: methadone, buprenorphine, and naltrexone.13 All three drugs are to be used in combination with counseling and psychosocial support14, described as a “multipronged approach that can include counseling, vocational training, psychosocial therapy, family support, and building connections to community resources,”14 also including safe/supportive housing as an essential dimension for many patients. The FDA also questions the term “MAT,” “Because OUD is a chronic illness, we should consider treating it much like we would any other chronic condition. We do not think of the medications used to treat diabetes or hypertension as ‘medication assisted treatment.’ We simply call it ‘treatment.’ OUD should be viewed similarly.”14 Table 1.1 provides an overview of the three FDA- approved drugs for the treatment of OUD. ©Institute for Clinical and Economic Review, 2018 Page 3 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Table 1.1. Comparison of Medications for OUD15-18 Methadone Buprenorphine Naltrexone Mechanism of Action at mu- Agonist Partial agonist Antagonist Opioid Receptor Medically supervised Maintenance, following Phase of Medically supervised withdrawal, withdrawal, medically supervised Treatment maintenance maintenance withdrawal Sublingual buccal, subdermal Route of Oral, intramuscular Oral implant, subcutaneous extended- Administration extended-release release Usual sublingual/buccal Limited effectiveness of oral Effective Dosage by Usually 60mg–120mg stabilizing dose between 12 mg– naltrexone due to limited Mouth daily 16 mg daily treatment retention Regulation through Not regulated through Controlled Schedule II Schedule III Controlled Substances Act Substances Act Only available in Prescribed by physicians, nurse opioid treatment practitioners and physician Availability programs with Available by prescription assistants with a SAMHSA SAMHSA certification prescribing waiver and DEA registration Methadone is a complete synthetic mu opioid receptor agonist that does not produce a euphoric effect as opioids do.12 However, access to methadone treatment is very limited in the US, as it cannot be legally dispensed through community pharmacies or physician offices, but only as part of highly structured treatment programs 43,44 Buprenorphine is a semi-synthetic mu opioid partial agonist, meaning that it binds and activates the receptor, but the activation is partial with a ceiling effect on its different actions, including the “high” that is achieved. Buprenorphine has historically been administered sublingually or in the form of buccal tablets to improve bioavailability. Prescription of buprenorphine in settings outside of methadone treatment programs requires a waiver that can be obtained by physicians, nurse practitioners and physician assistants, and the number of active OUD patients that any one practitioner can have is capped at 30 patients in the first year, 100 patients thereafter, and 275 by special designation.12 The combination of buprenorphine with the opioid antagonist naloxone, marketed as Suboxone, is frequently employed in order to avoid intravenous abuse. Naltrexone is a semi-synthetic mu receptor antagonist, meaning that it binds to the receptor but does not produce a response. Through its high affinity for the receptor it blocks the activation of the receptor by other opioids and displaces other opioids if they are already bound to the receptor. For the treatment of OUD, the patient must first undergo opioid withdrawal therapy for seven days, which involves abstaining also from buprenorphine and methadone, but taking only symptomatic ©Institute for Clinical and Economic Review, 2018 Page 4 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents medication. Attaining the period of opioid abstinence represents a challenge for many patients with OUD, and therefore MAT with oral naltrexone is not recommended due to low retention of patients, except under very specific circumstances.12 As stated by Nora Volkow, director of the National Institute on Drug Abuse, “these medications reduce withdrawal symptoms, improve mood, and help restore physiological balance—allowing the patient’s brain to heal while he or she works towards recovery.”45 This applies specifically to treatments with agonists. Naltrexone, at least with oral administration, has not been shown to normalize dopaminergic and stress responsive pathways.16 While this essential role of MAT has been established, it is not yet known definitively if or when to taper these medications. Some patients with OUD may achieve recovery without MAT, many need the medications for years, and others require lifelong treatment.12 In a recently published draft guidance document, the FDA recommends using a decrease in opioid use as a primary efficacy endpoint for demonstrating the effectiveness of drugs for OUD. The FDA further states that “sponsors and other stakeholders often mistakenly believe that using a change in drug use patterns as the endpoint always requires complete abstinence.” Long-term studies should demonstrate that observed reductions in drug use predict clinical benefit, even if opioid use has not completely stopped.19 By accepting and recommending a primary endpoint of a clinically relevant decrease in the use of opioids, rather than abstinence, the FDA endorses certain dimensions of “harm reduction strategies” that aim to minimize death, disease, and injury from continuing drug use, with a focus on improving daily social function and productivity. Despite the essential role of MAT in treating OUD and in preventing harm, including death, an important gap persists between the need for and the availability of MAT. More than 30 million people live in US counties without a single prescriber for addiction treatment20 and currently only about 20% of patients with OUD are receiving treatment21 Expanding access to OUD medications is considered an important public health strategy for countering the opioid epidemic.12 Populations in prisons and jails present a unique challenge for MAT, as regular use of heroin or other opioids is common prior to incarceration. For example, nearly 50% of people on arrival at the Middlesex Sheriff’s office in Massachusetts require medically supervised withdrawal from opioids46. However, currently few correctional settings in the US offer MAT for inmates.20 The lack of access drives diversion for self-medication to control withdrawal and cravings. This diversion reinforces negative beliefs about opioid agonist therapy in correctional settings.20 During imprisonment, tolerance of opioids is diminished and the risk for death from overdose is greatly increased upon release. ©Institute for Clinical and Economic Review, 2018 Page 5 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Extended-Release Medications Extended-release formulations have generated clinical interest because of their potential to improve retention in treatment and circumvent some of the access challenges seen with current forms of MAT. These formulations are currently available only for buprenorphine and naltrexone. Table 1.2 provides an overview of extended-release medications for OUD that are currently available or under consideration by the FDA. Table 1.2. Extended-Release Formulations for OUD Medications Substance Name and Company FDA Approval FDA Recommended Dosing After at least seven days of treatment with a transmucosal buprenorphine‐containing product delivering the equivalent of 8 to 24 mg of Sublocade™, Indivior buprenorphine daily, Sublocade abdominal (Subcutaneous Nov 30, 2017 subcutaneous injections are initiated with 300 mg injection) monthly for the first two months followed by a maintenance dose of 100 mg monthly. Maintenance dose can be increased up to 300mg monthly. Buprenorphine For patients on maintenance treatment with a transmucosal buprenorphine‐containing product Probuphine®, Titan delivering the equivalence of buprenorphine 8 mg Pharmaceuticals, Inc. May 26, 2016 or less per day. Four Probuphine implants (Subdermal implant) inserted subdermally in the upper arm for six months of treatment, after a new insertion in the other arm, transitioned back to a transmucosal. PDUFA date CAM2038, Braeburn expected for (Subcutaneous N/A December 26, injection) 2018 After an opioid-free duration of a minimum of Vivitrol®, Alkermes December 10, seven to 10 days. Administered 380 mg Naltrexone (Intramuscular 2010 for OUD intramuscularly every four weeks or once a injection) month. Of the four extended-release formulations, only Vivitrol was available at the time of the ICER report in 201434. As discussed in the previous section, naltrexone is an opioid receptor antagonist and it is essential that the patient undergoes opioid withdrawal before Vivitrol can be started. An intravenous or subcutaneous naloxone challenge is recommended to ensure complete withdrawal before starting Vivitrol.12 Patients transitioning from buprenorphine or methadone to Vivitrol can experience withdrawal symptoms for as long as two weeks after having stopped the agonist treatment. After completed withdrawal, Vivitrol is administered by a healthcare provider as an ©Institute for Clinical and Economic Review, 2018 Page 6 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents intramuscular (IM) gluteal injection, alternating buttocks for each subsequent injection. The injection should be made with one of the customized needles provided with the product. As with all medications for OUD, treatment with Vivitrol should be accompanied by psychosocial support.22 As naltrexone is not regulated by the Controlled Substances Act, Vivitrol can be prescribed without any particular requirements. Treatment with Sublocade replaces a daily dose of buprenorphine with a transmucosal product with extended-release formulation of buprenorphine. For treatment to be initiated, patients need to be on a stable transmucosal dose of 8 to 24 mg buprenorphine for at least seven days. Sublocade is administered through abdominal subcutaneous injection using the syringe and safety needle included with the product. Sublocade forms a solid mass upon contact with body fluids and if administered intravenously, can cause life-threatening pulmonary emboli, as mentioned in a black box warning in the FDA label. To minimize the risk that would arise from intravenous self- administration, the Risk Evaluation and Mitigation Strategy (REMS) program does not allow the drug to be dispensed directly to the patient. Sublocade must be administered by a healthcare provider.47 Buprenorphine is a Schedule III substance regulated by the Controlled Substances Act. It can only be prescribed by physicians, nurse practitioners, and physician assistants holding a SAMHSA waiver. Treatment with Probuphine implants involves surgical subdermal insertion on the inside of the upper arm of a set of four rods, each 2.5 mm in diameter and 26 mm in length, each rod containing the equivalent of 80 mg of buprenorphine. The implants must be removed after six months and a second set of rods can be placed in the other arm. After this second insertion, patients must transition back to a transmucosal buprenorphine-containing product.23 Peak buprenorphine plasma concentrations occur 12 hours after implant insertion, then slowly decrease, and after about four weeks reach steady-state concentrations comparable to daily transmucosal buprenorphine doses of 8 mg or less.12 The CAM2038 buprenorphine injection is currently under regulatory review with an expected approval date in December 2018. In clinical studies, this subcutaneous injection has been administered weekly or monthly with multiple dose strengths, in any subcutaneous tissue. The manufacturer proposes to include treatment initiation in the indication for CAM2038.24 If the FDA retains this proposal in the label to be approved, this would eliminate the need for prior treatment with transmucosal buprenorphine. During an FDA advisory committee meeting in November 201748, the committee members expressed some concerns about the trial design and the resulting clinical data on effectiveness and safety.49 In January 2018, the FDA requested additional clinical information from the manufacturer, although the nature of the additional data requested is currently unknown. ©Institute for Clinical and Economic Review, 2018 Page 7 Final Evidence Report - MAT in Patients with OUD 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 individuals and advocacy organizations, data from regulatory documents, information submitted by manufacturers, and other grey literature when the evidence met 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. ©Institute for Clinical and Economic Review, 2018 Page 8 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents 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 clinical or health outcomes. Outcomes are listed in the shaded boxes: those within the rounded boxes are intermediate outcomes (e.g., short-term abstinence from non-medical opioid use), 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 between these two types of outcomes may not always be validated. Curved arrows lead to the potential harms of an action (typically treatment), which are listed within the blue ellipsis.50 ©Institute for Clinical and Economic Review, 2018 Page 9 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Populations The key population of interest for the review included patients aged 16 years and above with OUD in various treatment settings. Given different patient incentives for seeking treatment and differing mechanisms of action for the treatments themselves, we focused on a range of patients with OUD who are being considered for MAT. For the subpopulations we focused on adolescents and young adults (up to 25 years), people who inject drugs, and pregnant women. Interventions The list of interventions was developed with input from patient organizations, clinicians, manufacturers, and payers on which drugs to include. The full list of interventions is as follows: • Buprenorphine subcutaneous extended-release injection (Sublocade) • Buprenorphine implant (Probuphine) • Buprenorphine subcutaneous extended-release injection (CAM2038) • Naltrexone intramuscular extended-release injection (Vivitrol) Comparators Comparisons were primarily made to other common medications used for OUD (e.g., buprenorphine/naloxone in sublingual and buccal formulation), as well as to placebo. As described further in Section 3, indirect comparisons of the interventions of interest to each other were not feasible due to differences in study populations, timing of randomization, and outcome measures between key trials. Outcomes The outcomes of interest are described in the list below. • Short-term and long-term abstinence from ongoing use of opioids • Diminishing illicit use of opioids • Opioid withdrawal syndrome • Health system utilization (number of emergency department (ED) visits, number of primary care physician (PCP) visits, days of inpatient hospitalizations) • Infectious (HIV, hepatitis), injection reactions, and other complications through continued use of injectable opioids • Functional outcomes (cognitive, occupational, social/behavioral)51 • Craving/desire for opioids • Accidental pediatric exposure • Mortality (overdose deaths, suicide) • Health-related quality of life ©Institute for Clinical and Economic Review, 2018 Page 10 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents • Employment-related outcomes • Adherence/treatment discontinuation (number of times treated in detox/rehab, duration of abstinence) • Diversion • Other patient-reported outcomes • Other adverse events Timing Evidence on intervention effectiveness and harms was derived from studies of any follow-up duration. Settings The settings of interest included outpatient (including office-based), inpatient, and correctional facility settings in the US. 1.3 Definitions Agonist An agonist is a chemical that binds to a receptor and activates the receptor to produce a biological response. A partial agonist, such as buprenorphine, binds and activates the receptor, but the activation is partial, even at maximal receptor occupancy. Antagonist An antagonist binds to a receptor but does not produce a response, and in the case of naltrexone also blocks the activation of the receptor by other opioids. Harm Reduction Harm reduction for OUD includes policies, programs and practices that aim to minimize death, disease, and injury from continuing drug use, without the explicit goal of reducing or stopping use. Syringe exchange programs are an example of a harm reduction strategy to reduce HIV and HCV infections in people who inject drugs. Although there is often misunderstanding and unnecessary controversy surrounding harm reduction, its goals are congruent with the goals of treatment and recovery. MAT with agonists can be viewed as an example of a harm reduction strategy, when diminishing opioid use, rather than complete abstinence, is accepted as a valid outcome. ©Institute for Clinical and Economic Review, 2018 Page 11 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Maintenance Treatment Providing medications to achieve and sustain clinical remission of signs and symptoms of OUD and support the individual process of recovery without a specific endpoint (as with the typical standard of care in medical and psychiatric treatment of other chronic illnesses).12 MAT Medication for addiction treatment (MAT) is helping individuals sustain recovery using medications approved by the FDA in combination with individualized psychosocial supports. It is also often called medication-assisted treatment. Different international organizations are using the term Psychosocially Assisted Pharmacological Treatment of Opioid Dependence to refer to the combination of specific pharmacological and psychosocial measures used to reduce both illicit use of opioids and harms related to opioid use, and improve quality of life.17 OUD Opioid use disorder (OUD) is defined by DSM-5 by the presence of a certain number of the following signs and symptoms: impaired control, social impairment, risky use, increased tolerance, and withdrawal. OUD replaces what DSM-IV termed “opioid abuse” and “opioid dependence.” The diagnostic criteria for moderate to severe OUD roughly correspond to what is considered addiction, which is defined as a “primary, chronic, neurobiological disease, with genetic, psychosocial, and environmental factors influencing its development and manifestations, characterized by behaviors that include one or more of the following: impaired control over drug use, compulsive use, continued use despite harm, and craving.”7. Recovery Recovery is a process of change through which individuals improve their health and wellness, live self-directed lives, and strive to reach their full potential. Four major dimensions support a life in recovery: health, home, purpose, and community. Though some individuals enter and sustain recovery on their own, recovery is mostly achieved via access to evidence-based clinical treatment and recovery support services.9 A person in recovery refers to an individual who abstains from further use, reduces their substance use to a safer level, or takes steps to mitigate the potential physical and emotional harm resulting from continued use.8 A person can be considered in recovery while on MAT. Relapse A process in which a person with OUD who is being treated and is in remission experiences a loss of control. A relapse is different from a return to opioid use that is limited in scope and time and that does not involve the return of the signs or symptoms of OUD. The different ©Institute for Clinical and Economic Review, 2018 Page 12 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents operational definitions of relapse in clinical trials of medications for OUD are based on different levels of return to opioid use as measured by toxicology tests and questionnaires. Remission Remission refers to the disappearance of signs and symptoms of the disease. DSM-5 defines remission as present in people who were diagnosed with OUD but no longer meet OUD criteria, except for craving. Remission is an essential element of recovery.12 Withdrawal Opioid withdrawal is defined by DSM-5 by the presence of at least three of the following signs or symptoms: dysphoric moods; nausea or vomiting; muscle aches; lacrimation or rhinorrhea; pupillary dilation, piloerection, or sweating; diarrhea; yawning; fever, insomnia. A withdrawal syndrome is a sign of physical dependence and can be produced by abrupt cessation, rapid dose reduction, and/or administration of an antagonist, and is a sign of prior physical dependence. Addiction medicine professionals use the term withdrawal management instead of detoxification27, which was defined under the Narcotic Addict Treatment Act of 1974 as a treatment to achieve an opioid-free state. 1.4 Insights Gained from Discussions with Patients and Patient Groups As part of our review, we spoke with organizations working with individuals and families affected by OUD. There was a consensus that MAT is often difficult to access, in part because of the stigma attached to OUD. This stigma is rooted in a widespread belief that drug addiction is a moral failing rather than a medical condition that is best addressed through treatment. OUD needs to be considered a chronic disease that can affect widely varying populations in terms of age, background, and other factors. The expression, “treatment is not one-size-fits-all,” was used by several organizations to stress the importance of patients having access to different treatment options on their road to recovery; some patients enter recovery without the assistance of MAT, while others require MAT for long periods of time or even their entire lives. Equal access to all types of medications is considered important. For example, we received comments that Vivitrol is currently more easily available than other medications. Buprenorphine extended-release medications are considered important new treatment options that could improve recovery and should be widely available for consideration by patients and physicians. It was also mentioned that peer support is particularly important for young people entering the recovery process, as they usually lack a strong existing social network compared to older adults. ©Institute for Clinical and Economic Review, 2018 Page 13 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Several organizations stressed that better daily functioning and well-being, and eventually recovery, are the most important outcomes of treatment. For some this may involve complete abstinence from non-medical opioid use, for others a reduced and controlled level of use. It was mentioned that this corresponds specifically to the discussions at the public meeting on Patient-Focused Drug Development for Opioid Use Disorder convened in April 2018 by the FDA.25. 1.5. Potential Cost-Saving Measures in Extended-Release Opioid Agonists and Antagonist MAT in Patients with OUD As described in its Final Value Assessment Framework for 2017-2019, ICER now includes 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 MAT that could be reduced, eliminated, or made more efficient. ICER encouraged all stakeholders to suggest services (including treatments and mechanisms of care) currently used for people with OUD that could be reduced, eliminated, or made more efficient. We have not received any suggestions for potential cost-saving measures but continue to seek such input. ©Institute for Clinical and Economic Review, 2018 Page 14 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents 2. Summary of Coverage Policies and Clinical Guidelines 2.1 Coverage Policies To understand the insurance landscape for MAT, we reviewed publicly available representative coverage policies for Sublocade, Probuphine, and Vivitrol. We reviewed policies from the Centers for Medicare and Medicaid Services (CMS), MassHealth, and from regional and national commercial insurers (Aetna, Anthem, Blue Cross Blue Shield of Massachusetts [BCBS of MA], Cigna, Harvard Pilgrim, and United Healthcare [UHC]). At the time the Evidence Report was published, we were unable to survey policies pertaining to CAM2038, as the FDA had yet to issue a decision on the medication. Limited information is available regarding Medicare coverage of treatment for OUD. National Coverage Determinations (NCD) describe policies regarding physician-provided, hospital outpatient, and freestanding clinic services for drug abuse treatment. Coverage is subject to general limitations applicable to these settings of care.52,53 We were unable to locate Local Coverage Determinations (LCD) for any treatment for OUD. Details of the utilization management policies (UMPs) for Sublocade, Probuphine, and Vivitrol are broadly summarized below. We were unable to locate specific UMPs for Harvard Pilgrim or Cigna, but we located a general Cigna policy for the treatment of substance abuse, which defines established standards of effective care.54 Sublocade We located UMPs for Aetna, Anthem, and UHC. All payers require a diagnosis of moderate-to- severe opioid dependence. Additionally, the three payers require patients to have initiated treatment first with a transmucosal or sublingual buprenorphine-containing product before beginning treatment with Sublocade. Anthem and UHC specify that patients undergoing treatment with Sublocade may not receive supplemental oral, sublingual, or transmucosal buprenorphine. Aetna and Anthem state that psychosocial counseling must accompany treatment with Sublocade. UHC specifies a further requirement that the initial authorization for Sublocade may not exceed six months.55-57 Sublocade is not listed on the 2018 formularies for Aetna’s 3-Tier Value Plan, Anthem’s National 3- Tier Drug Plan, Harvard Pilgrim’s Value 3-Tier Plan, or UHC’s Traditional Three-Tier Plan. However, all plans offer alternative branded and generic forms of buprenorphine and buprenorphine/naloxone for the treatment of OUD. Cigna covers Sublocade under a patient’s ©Institute for Clinical and Economic Review, 2018 Page 15 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents medical benefit. BCBS of MA was the only payer to include Sublocade on its 3-Tier Plan formulary. Sublocade is placed on the mid-range tier and does not require prior authorization.58 Probuphine We located UMPs for Aetna, Anthem, BCBS of MA, and UHC. All payers require a diagnosis of moderate-to-severe opioid dependence. Further, all payers state that the patient must have achieved prolonged stability on transmucosal buprenorphine before initiating treatment with Probuphine. Anthem, BCBA of MA, and UHC also specify that patients must be currently maintained on an appropriate dose of sublingual or transmucosal buprenorphine. Aetna, Anthem, and BCBS of MA specify that psychosocial counseling must accompany treatment with Probuphine. UHC specifies two further requirements: one, the patient may not receive supplemental oral, sublingual, or transmucosal buprenorphine, and two, the patient cannot have had an opioid positive urine drug screen in the past 90 days prior to the insertion of Probuphine.55,57,59,60 Probuphine is not listed on the 2018 formularies for Aetna’s 3-Tier Value Plan, BCBA of MA’s 3-Tier Plan, Harvard Pilgrim’s Value 3-Tier Plan, or UHC’s Traditional Three-Tier Plan.58,61-64 Cigna covers Sublocade under a patient’s medical benefit. Anthem covers Probuphine on its three-tier plan on the highest formulary tier and requires prior authorization.65 Vivitrol Since Vivitrol has fewer prescribing restrictions and criteria than both Sublocade and Probuphine, we were unable to locate UMPs from commercial and regional payers indicated for the treatment of OUD. Vivitrol is not listed on the 2018 formulary UHC’s Traditional Three-Tier Plan.62,64 Aetna, Anthem, BCBS of MA, and Harvard Pilgrim cover Vivitrol on the highest formulary tier.58,61,64,65 Anthem is the only payer that requires prior authorization. Medicaid Policies To begin treatment with Sublocade, MassHealth requires a diagnosis of opioid dependency, documentation that the patient is stabilized on buprenorphine for at least seven days, and evidence that the patient needs the extended-release injection formulation. Prior authorization is required. Similarly, Probuphine may be prescribed if the patient is diagnosed with opioid dependence, is currently stabilized on buprenorphine or equivalent for at least six months and requires the implant formulation due to an adverse reaction, contraindication, or diversion with other therapeutic alternatives. Prior authorization is required. Vivitrol may be prescribed without prior authorization and is not subject to a utilization management policy.66 ©Institute for Clinical and Economic Review, 2018 Page 16 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents The Department of Vermont Health Access (DVHA) covers Sublocade and Probuphine as non- preferred brands for the treatment of OUD. All covered forms of buprenorphine require prior authorization, and patients undergoing treatment with Sublocade and Probuphine must receive psychosocial counseling and therapy. Patients may be prescribed Vivitrol if they pass the naloxone challenge and partake in a comprehensive treatment plan that includes psychosocial counseling and therapy.67 2.2 Clinical Guidelines We reviewed guidelines on MAT issued by major US clinical societies, working groups, and health technology assessment organizations. A majority of these guidelines do not include recommendations concerning Probuphine and Sublocade because they were only recently approved. The third long-acting buprenorphine formulation, CAM2038, is currently under review by the FDA and, as such, was not listed in any clinical practice guidelines. The 2010 guideline from the American Psychiatric Association, which was summarized in ICER’s 2014 report on opioid dependence, has not been updated since and may be found in Appendix F. American Association for the Treatment of Opioid Dependence (AATOD) AATOD Guidelines for Using Naltrexone (Vivitrol) in OTPs (2012)68 The 2012 AATOD guidelines state that Vivitrol may be an effective treatment for patients who struggle with a daily dosing routine. Before a patient begins treatment with Vivitrol, they must be opioid-free for at least seven to 10 days. Additionally, the AATOD recommends that a patient passes the naloxone challenge before initiating treatment with Vivitrol. The guidelines also state that Vivitrol treatment should be combined with drug rehabilitation programs, psychological counseling, and/or Narcotics Anonymous meetings. Academy of Managed Care Pharmacy (AMCP) Findings and Considerations for the Evidence-Based Use of Medications Used in the Treatment of Substance Abuse Disorder (2016)69 In their 2016 recommendations, the AMCP recommends that all patients with substance abuse disorder, including opioid addiction, should be offered medication. However, the decision to begin medication for opioid addiction should be an individualized decision between the doctor and patient. In conjunction with pharmacologic treatment, the AMCP recommendations also emphasize that psychosocial treatment should be utilized as an important aspect of recovery. ©Institute for Clinical and Economic Review, 2018 Page 17 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents American Society of Addiction Medicine (ASAM) The ASAM National Practice Guideline for the Use of Medications in the Treatment of Addiction Involving Opioid Use (2015)70 Buprenorphine (Sublingual) In their 2015 guidelines, the ASAM recommends that sublingual buprenorphine should be initiated in opioid-dependent patients once they begin to experience mild-to-moderate opioid withdrawal. Psychosocial treatment should be employed in conjunction with buprenorphine in order to maximize the benefit of treatment. In order to prevent diversion, the ASAM recommends that patients undergoing treatment should have frequent doctor visits (especially early in treatment when weekly visits are recommended), urine drug tests, and visits for pill counts. In addition, the ASAM states that patients should be tested for buprenorphine, additional substances, and prescription medications. Lastly, the ASAM states that if a patient switches from buprenorphine to naltrexone, there should be a seven to 14 day waiting period before treatment with naltrexone is initiated. Naltrexone (Vivitrol) The ASAM guidelines state that Vivitrol is also an effective treatment for OUD, and may be especially effective in patients with contraindications to buprenorphine or for patients for whom treatment with buprenorphine failed. The ASAM notes that Vivitrol may also be effective for patients confined to prison or inpatient habilitation, patients with co-occurring opioid disorders, and patients living in locations where methadone or buprenorphine treatment is unavailable. The ASAM strongly recommends psychosocial treatment in conjunction with Vivitrol. Lastly, the ASAM notes that treatment with Vivitrol does not have a specified timeframe and that the duration of treatment is based on clinical determinations and a patient’s situation. National Institute on Drug Abuse (NIDA) Principles of Drug Addiction Treatment: A Research-Based Guide (2018)71 The 2018 NIDA guidelines offer recommendations for both drug addiction broadly and addiction to opioids. Overall, the NIDA emphasizes that the treatment program must take into consideration not only the individual’s psychological, social, professional, and legal situation, but also their age, gender, and ethnicity. The NIDA recommends that the treatment plan be continually reviewed and modified if indicated. Behavioral therapy and psychosocial counseling are also essential to treating drug addiction, and may include individual, family, or group therapy. ©Institute for Clinical and Economic Review, 2018 Page 18 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents The NIDA states that pharmacologic therapy is also an important aspect of drug addiction treatment, and that methadone, buprenorphine (sublingual, injectable, and an implantable form), and naltrexone may be effective for individuals with opioid addiction. Individuals undergoing pharmacologic therapy for addiction must be continually monitored to ensure the efficacy of the treatment. Substance Abuse and Mental Health Services Administration (SAMHSA) Medications for Opioid Use Disorder for Healthcare and Addiction Professionals, Policymakers, Patients, and Families: Treatment Improvement Protocol (TIP) 63 (2018)12 In the 2018 TIP 63, SAMHSA states that opioid use disorder should be approached as a chronic illness, and to be effective, treatment for the disorder may need to be ongoing and continuous. Patients seeking treatment should be fully informed about opioid use disorder and the various pharmacologic treatment options available, including methadone, injectable naltrexone, and extended-release buprenorphine formulations. SAMHSA emphasizes that treatment should be individualized to the patient’s needs, considering patient preference, occupation, use of other substances, and past treatment. SAMHSA stresses that the length of treatment with injectable naltrexone and buprenorphine may vary, and some patients may require lifelong treatment. Patients who begin treatment should be monitored closely within the first few days and weeks of induction to encourage retention and to assess the effectiveness of treatment. To improve the efficacy of pharmacologic treatment, SAMHSA recommends that patients have access to counseling, recovery, and addiction services. Additionally, pharmacologic treatments should be integrated with outpatient and home treatment, and some patients may require different levels and sites of care that range from inpatient treatment to outpatient counseling. SAMHSA and the US Department of Health and Human Services (HHS) Federal Guidelines for Opioid Treatment Programs (2015)44 The 2015 SAMHSA/HHS guidelines refer specifically to treatment with methadone and all products containing buprenorphine (the guidelines also note that extended-release injectable naltrexone may be appropriate for some patients, but is not subject to these SAMHSA/HHS regulations). The guidelines state that the pharmacologic selection should be determined by a clinician and should take into consideration the patient’s medical history, psychological state, complicating conditions, age, gender, and past and present substance use. Dosages should be adjusted when clinically indicated, and the SAMHSA/HHS guidelines note that dosage caps and ceilings should be eliminated. The guidelines state that medication-assisted treatment may be continued indefinitely and recommends against fixed treatment lengths. Patients undergoing pharmacologic treatment should also undergo psychosocial treatment. ©Institute for Clinical and Economic Review, 2018 Page 19 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents 3. Comparative Clinical Effectiveness 3.1 Overview In this review of the comparative clinical effectiveness of newer, extended-release treatments for MAT (two buprenorphine injections, one buprenorphine implant, and a naltrexone injection), we systematically identified and synthesized the existing evidence from clinical studies. Full PICOTS criteria are described in Section 1.2. In brief, we evaluated studies of patients 16 years or older with OUD. Our review focused on the efficacy, safety, and effectiveness of Sublocade, Probuphine, CAM2038, and Vivitrol versus each other and versus transmucosal formulations of buprenorphine/naloxone. We extracted any relevant data, whether in published or unpublished form (e.g., conference abstracts or presentations, FDA review documents). Due to important differences in study characteristics and outcomes assessed, we did not compare the interventions of interest through direct or indirect quantitative assessments. Essential to our review was the evidence on the clinical benefits common to trials on MAT and reported tolerability/harms. We sought evidence on all outcomes listed below: • Relapse and abstinence outcomes based on urinalysis or self-report • All-cause discontinuation/study retention • Craving/desire for opioids • Opioid withdrawal syndrome (Clinical Opiate Withdrawal Scale [COWS], Subjective Opiate Withdrawal Scale [SOWS]) • Diminishing illicit use of opioids • Health-related quality of life (EuroQol-5 dimensions questionnaire, 36-item short form health survey [SF-36]) • Health system utilization (number of emergency department [ED] visits, number of primary care physician [PCP] visits, days of inpatient hospitalizations) • Functional outcomes (cognitive, occupational, social/behavioral) • Mortality (overdose deaths, suicide) • Employment-related outcomes • Other patient-reported outcomes • Infections (HIV, hepatitis), injection reactions, and other complications through continued use of injectable opioids • Accidental pediatric exposure • Diversion • Adverse Event (AE): serious adverse events (SAEs), discontinuation due to AEs, any AE reported by ≥ 5% of a trial arm ©Institute for Clinical and Economic Review, 2018 Page 20 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents 3.2 Methods Data Sources and Searches Procedures for the systematic literature review assessing the evidence on MAT for OUD followed established best research methods.72 The review was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines.73 The PRISMA guidelines include a list of 27 checklist items, which are described further in Appendix Table A1. We searched MEDLINE, EMBASE, Cochrane Database of Systematic Reviews, and Cochrane Central Register of Controlled Trials for relevant studies. Each search was limited to English language studies of human subjects and excluded articles indexed as guidelines, letters, editorials, narrative reviews, case reports, or news items. We included abstracts from conference proceedings identified from the systematic literature search and submitted by manufacturers. All search strategies were generated utilizing the Population, Intervention, Comparator, and Study Design elements described in Section 1. The search strategies included a combination of indexing terms (MeSH terms in MEDLINE and EMTREE terms in EMBASE), as well as free-text terms, and are presented in Appendix Tables A2-A4. The date of the most recent search was September 25, 2018. To supplement the database searches, we performed a manual check of the reference lists of included trials and reviews and invited key stakeholders to share references germane to the scope of this project. We further supplemented our review of published studies with data from conference proceedings, regulatory documents, information submitted by manufacturers, and other grey literature when the evidence meets ICER standards (for more information, see http://icer-review.org/methodology/icers-methods/icer-value-assessment-framework/grey- literature-policy/). Study Selection Two reviewers independently screened the abstracts and full-texts of studies using DistillerSR (Evidence Partners, Ottawa, Canada), with any incongruencies resolved through consensus. We included relevant published and unpublished randomized clinical trials (RCTs) of any sample size as well as non-randomized comparative studies where available. To support the comparative evidence and to gain insights into the duration of treatment benefits and harms, we included non- comparative single arm observational studies with a minimum of 50 participants, and open-label extensions (OLEs) of RCTs of any size and duration. Studies assessing transmucosal buprenorphine formulations not containing naloxone and any routes of administration outside of scope (e.g., naltrexone implant) were excluded. We excluded conference abstracts reporting data that were also available in a full-text peer-reviewed publication. A detailed protocol of the methods was registered on Prospero (CRD42018103836). ©Institute for Clinical and Economic Review, 2018 Page 21 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Data Extraction and Quality Assessment Data were extracted into Microsoft Excel by one researcher and independently verified by another researcher. Data elements included a description of patient populations, sample size, duration of follow-up, funding source, study design features (e.g., open-label), interventions (drug, dosage, frequency, schedules), outcome assessments (e.g., timing, definitions, and methods of assessment), results, and quality assessment for each study. Quality assessment was based on US Preventive Services Task Force (USPSTF)74 criteria that included presence of comparable groups, non- differential loss to follow-up, use of blinding, clear definition of interventions and outcomes, and appropriate handling of missing data. For more information on data extraction and quality assessment, refer to Appendix D. Assessment of Level of Certainty in Evidence We used the ICER Evidence Rating Matrix to evaluate the level of certainty in the available evidence of a net health benefit among each of the interventions of focus (see Appendix D).75 Assessment of Bias We assessed the presence of publication bias by utilizing the ClinicalTrials.gov database of trials. Search terms included “buprenorphine injection,” “Sublocade,” “CAM2038,” “buprenorphine implant,” “Probuphine,” “naltrexone,” “Vivitrol,” and prior terms for these therapeutics. Publication bias was evident if any registered trials meeting our inclusion criteria remained unpublished after more than two years since their completion. We did not identify any completed, but unpublished trials of our interventions. Therefore, we did not find any evidence of publication bias. We did identify three registered and ongoing trials of the injectable naltrexone of focus. These trials are described in the Ongoing Trials sections in Appendix C and not included in any qualitative analyses. Data Synthesis and Statistical Analyses Data on relevant outcomes are summarized in evidence tables in Appendix D and are synthesized qualitatively in the text of the report. Relevant data include those listed in the data extraction section. Where possible, data on key outcomes of interest were evaluated on an intent-to-treat (ITT) basis. Due to differences between the studies in terms of the study design, patient characteristics, (including dosing and frequency), and outcomes (including definitions and methods of assessments), we were unable to directly or indirectly compare the interventions of interest by quantitative assessments. Hence, we focused on narratively describing the comparisons made within the clinical trials of each intervention. ©Institute for Clinical and Economic Review, 2018 Page 22 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents 3.3 Results Study Selection Our literature search identified a total of 557 potentially relevant references (see Appendix Figure A1). We included 23 references, of which 18 focused on comparative clinical trials, three were on OLEs, and two on observational studies. These references consisted of 15 publications, five conference abstracts, and three web-based references. Primary reasons for study exclusion included use of interventions outside of our scope, different study population (e.g., recreational opioid users), small sample size (sample size <50 for observational studies), and conference abstracts with duplicate data to the full-text publications. The 18 references of comparative trials correspond to 11 trials, of which five of the trials were identified as key trials evaluating the four drugs of interest (Table 3.1). In four of the key trials, three of the interventions of interest (CAM2038, Probuphine and Vivitrol) were compared to buprenorphine/naloxone, while the remaining one key trial (Sublocade) was placebo-controlled with no active comparator. We identified no head-to-head trials of the interventions of interest. Below, we describe these trials and efficacy results, followed by a discussion of the tolerability and harms. Trial Characteristics CAM2038 (Buprenorphine Subcutaneous Extended-Release Injection) Data to inform our assessment of CAM2038 in patients with OUD were obtained from one published Phase III RCT (Lofwall 2018).76 This 24-week multicentered non-inferiority trial compared CAM2038 on a weekly and monthly basis to sublingual buprenorphine/naloxone in patients with moderate-to-severe OUD aged 18 to 65 years. There was no detoxification period. Eligible participants were already in opioid withdrawal and able to tolerate an initial sublingual dose of 4 mg of buprenorphine/1 mg of naloxone administered at the start of the study. Participants were randomized on the first day to receive weekly or monthly subcutaneous injections of CAM2038 and daily sublingual placebo tablets, or subcutaneous placebo injections and sublingual buprenorphine/naloxone tablets. Doses were given on a flexible schedule of 16, 24, or 32 mg of weekly CAM2038 for the first 11 weeks (phase 1), and 64, 96, 128, or 160 mg of monthly CAM2038 from weeks 12 to 24 (phase 2).76 Reinductions were not allowed if participants missed a visit. The primary outcome was the mean percentage of urine samples testing negative for illicit opioids over 24 weeks.76 Additionally, responder rate, which is defined as percentage of patients with no evidence of illicit opioid use (as measured by negative urine test results and self-report of drug use) in phase one (for at least two of the three assessments at weeks nine to 11), and in phase two (for five of the six assessments from week 12 to 24) was a co-primary endpoint. Secondary outcomes ©Institute for Clinical and Economic Review, 2018 Page 23 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents included mean percentage of opioid-negative samples from weeks four through week 24, and study retention. Sublocade (Buprenorphine Subcutaneous Extended-Release Injection) Data used to inform our assessment of Sublocade was mainly taken from a six-month Phase III randomized placebo-controlled trial.77 Participants and investigators where blinded to treatment in the trial. Eligible participants ages 18 to 65 first underwent an open-label run-in induction phase with sublingual buprenorphine/naloxone film followed by an open-label run-in dose adjustment period for four to 11 days to achieve a buprenorphine /naloxone dose of eight to 24 mg.77 Approximately three-fourths completed the run-in phase and were randomized to either the 300 mg dose of Sublocade injection, 100 mg dose of Sublocade injection, or placebo. After randomization, the dose of sublingual buprenorphine/ naloxone was tapered and then discontinued. Those randomized to the 100 mg dose group received an initial monthly dose of 300mg Sublocade for two months before receiving a monthly 100 mg dose for four months, while the 300 mg group received a monthly dose of 300 mg for the six months. In addition, participants received individual counseling at least once a week.77 The primary outcome was the percentage of urine samples combined with self-reports negative for illicit opioid use from weeks five to 24. Secondary outcomes included percentage of participants with treatment success defined as having ≥80% of urine samples and self-report negative for illicit opioids from weeks five to 24, percentage of urine samples negative for opioids from weeks five to 24, percentage of self-reports negative for illicit opioid use from weeks five to 24, opioid craving, opioid withdrawal, participants who completed the last visit (completers), participants who were abstinent, and clinician-reported ratings.77 In addition, we identified one OLE, an extension of the Sublocade trial.78 The study population included a combination of participants who were rolled over after completing the RCT and those newly enrolled into the OLE. All participants underwent an induction period of sublingual buprenorphine/naloxone film for up to two weeks to achieve a dose ranging from eight mg to 24 mg/day. All participants received manual-guided individual counseling sessions varying in frequency.78 Probuphine (Buprenorphine Implant) Data used to inform our assessment of Probuphine was obtained from one Phase III RCT (Rosenthal 2016).79 Rosenthal 2016 was a placebo-controlled, multicentered, six-month trial conducted in the US on participants 18 to 65 years of age with OUD. In this non-inferiority trial, participants were required to be clinically stable and must have received a stable dose of sublingual buprenorphine (8 mg/day or less) for at least 24 weeks with no opioid withdrawal or illicit opioid-positive urine samples for at least 90 days before the study began. Patients were then randomized to receive ©Institute for Clinical and Economic Review, 2018 Page 24 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents either four daily sublingual buprenorphine tablets with four placebo subdermal implants or four buprenorphine implants with daily sublingual placebo tablets.79 Both the participants and investigators in the trials were blinded to treatments.79 All participants received manual-guided counseling sessions during each visit and 2 mg/day of supplemental sublingual buprenorphine as needed during the study duration.79 The primary outcome in the Rosenthal 2016 trial was the difference in proportion of responders, defined as participants with no illicit opioid use based on monthly urine drug tests and self- reporting in four of the six months screening.79 Secondary outcomes assessed in Rosenthal 2016 included treatment retention, percentage of illicit opioid use per month, cumulative percentage of negative illicit opioid urine results during six months, opioid craving, and opioid withdrawal. In addition, we identified two six-month placebo-controlled randomized trials and one six-month OLE all conducted in the US on participants of same age range as Rosenthal 2016.80-82 In Rosenthal 2013 and Ling 2010, participants were required to complete an open-label induction phase with buprenorphine/naloxone (12 to 16 mg/day sublingual tablets for three consecutive days) within 10 days of screening before randomization.80,81 Participants who experienced severe withdrawal symptoms (>12 on the Clinical Opiate Withdrawal Scale) or severe cravings for opioids (>20 on the 0-100 Visual Analog Scale) during the induction phase were excluded.80,81 Participants in both trials were randomized to receive either 320 mg Probuphine implants (four 80 mg subdermal implants) or four placebo implants, with an additional open-label non-inferiority comparison to buprenorphine/naloxone administered in a third arm of participants in Rosenthal 2013. Participants and investigators were blinded to Probuphine and the placebo implant in both trials.80,81 Participants also received manual-guided counseling sessions biweekly to weekly and 2 mg/day of supplemental sublingual buprenorphine as needed during the study duration.80,81 The OLE found was a continuation of Rosenthal 2013. Participants who completed the Rosenthal 2013 trial underwent a brief induction phase with sublingual buprenorphine/naloxone (12-16 mg/day) then received four buprenorphine implants in the opposite arm.82 The primary outcome in Ling 2010 was the percentage of the total urine samples that were negative for illicit opioids during first 16 weeks.80 In Rosenthal 2013, the primary outcome was the percentage of total urine samples that were negative for illicit opioids during the 24 weeks trial period. In addition, the combination of percentage of urine samples with negative self-report of illicit use was identified as a coprimary endpoint in Rosenthal 2013.81 Secondary outcomes assessed in these trials included retention, percentage of illicit opioid use per month, opioid craving, and patient-reported and clinician- reported withdrawal scale. The OLE assessed the percent of opioid-negative urine samples, study retention, and reductions in opioid use. Vivitrol (Naltrexone Intramuscular Extended-Release Injection) Data used to inform our assessment of Vivitrol were taken from two key trials: one Phase IV trial (Lee 2018, X-BOT) and one Phase III trial (Tanum 2017).83,84 X-BOT was a 24-week multicentered, ©Institute for Clinical and Economic Review, 2018 Page 25 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents open-label, randomized controlled trial that compared Vivitrol to buprenorphine/naloxone. Participants were 18 years or older, diagnosed with OUD, and had used non-prescribed opioids within 30 days of the trial.83 Detoxification protocols and length of time varied by site: two sites used clonidine or comfort medications (no opioids); four sites used three to five days of methadone tapers, while the remaining two sites used three to 14 days of buprenorphine tapers.83 Timing of randomization also varied (during detoxification or after completion of detoxification), but this was designed a priori to assess the difficulty of completing detoxification. Participants were randomized to receive either 380 mg/month of Vivitrol or 8 to 24 mg/day of sublingual buprenorphine/naloxone film and stratified by treatment site and opioid use severity. Before induction with Vivitrol, participants had to complete detoxification, have opioid-negative urine, and a negative naloxone challenge (minimal or no withdrawal symptoms after administration of ≥0.4 mg naloxone).83 Missed Vivitrol injections required participants to be reinducted with a repeat naloxone challenge.83 The primary outcome in X-BOT trial was the time to relapse event. Relapse was defined as the use of non-study opioids beyond 20 days after randomization.83 Other secondary outcomes were the proportion of participants who were successfully inducted on an initial dose, frequency of non- study opioid use, and opioid craving.83 The second key trial of Vivitrol (Tanum 2017) was a 12-week, multicenter, non-inferiority, open- label, randomized controlled trial that was conducted in outpatient settings in Norway comparing Vivitrol to sublingual buprenorphine/naloxone in participants 18 to 60 years of age diagnosed with OUD.84 Participants with other drug or alcohol use disorders were excluded from the trial. Following screening, participants were referred to a detoxification unit.84 After detoxification, participants were randomly assigned to 4 to 24 mg/day of oral buprenorphine/naloxone or 380 mg/month of Vivitrol.84 The primary outcome in Tanum 2017 measured study retention, proportion of total number of urine drug tests without illicit opioids, and number of days of heroin and other illicit opioids.84 Secondary outcomes included number of days of injecting intravenous drugs, and heroin craving.84 In addition to the two key trials described above, we identified nine other trials of Vivitrol (four RCTs and two OLEs).85-90 All were government funded in the US and Russia, and ranged in duration from eight to 78 weeks. Comparisons to Vivitrol included placebo or usual treatment, which varied in definition across the trials. The detoxification period also ranged from a week to a month with up to a week of Vivitrol induction. Participants were ages 18-60 years and met either DSM-IV for substance dependence and abuse or DSM-5 criteria for OUD with the majority of the trials assessing treatment in those who were or had been incarcerated. In contrast to these trials, the study population in the key trials was not restricted to incarcerated individuals. Appendix Tables D1-D3 contain data on the study design and baseline characteristics for all studies included. ©Institute for Clinical and Economic Review, 2018 Page 26 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Table 3.1. Trial Characteristics of Key Trials Trial Treatment Arms Patient Characteristics Intervention Period Primary Outcomes CAM2038 N=428; Mean age: 38.4; Male: 61.4%; Heroin as • Mean percentage of urine samples with CAM2038* primary opioid: 70.8%; Prescription drugs as primary test results negative for illicit opioids Lofwall 2018 24 weeks SL bup/nal* opioid: 29.2%; Intravenous drug use: 52.3%; Mean during trial period years since OUD diagnosis: 4.5 • Responder rate Sublocade N=504; Mean age: 45.2% between 30 and 45; Male: • Percentage of urine samples negative Sublocade 300mg/100mg Trial 13-0001 66.7%; Heroin as primary opioid: NR; Prescription Sublocade 300mg/300mg 24 weeks for illicit opioid combined with self- (Unpublished) drugs as primary opioid: NR; Intravenous drug use: Placebo reports negative for illicit opioid use NR; Mean years since OUD diagnosis: NR Probuphine N=177; Mean age: 39.0; Male: 59.1%; Heroin as Probuphine 320 mg primary opioid: 21.0%; Prescription drugs as primary Rosenthal 2016 SL bup/nal ≤8 mg 24 weeks • Proportion of responders opioid: 74.4%; Intravenous drug use: NR; Mean years since OUD diagnosis: 6.2 Vivitrol • Percentage of urine samples that were N=159; Mean age: 36.1; Male: 72.3%; Heroin as negative for illicit opioids Vivitrol 380 mg primary opioid: NR; Prescription drugs as primary Tanum 2017 12 weeks • Study retention SL bup/nal 4-24 mg opioid: NR; Intravenous drug use: 85.5%; Mean years • Number of days of heroin and other of heavy opioid use: 9.3 illicit opioids N=570; Mean age: 34.0; Male: 70.4%; Heroin as Vivitrol 380 mg primary opioid: 81.0%; Prescription drugs as primary Lee 2018 X:BOT 24 weeks • Time to a relapse event SL bup/nal 8-24 mg opioid: 15.8%; Intravenous drug use: 63.2%; Mean years of opioid use: 12.5 *Flexible dosing. Bup/nal: buprenorphine/naloxone, d: day, mg: milligram, OUD: opioid use disorder, SL: sublingual ©Institute for Clinical and Economic Review, 2018 Page 27 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Quality of Individual Studies We rated the one trial of the buprenorphine implant, one trial of CAM2038, one trial of Sublocade, and three trials of Vivitrol to be of good quality. These trials had comparable arms at baseline, did not have differential attrition, were patient and physician/investigator blinded, had clear definitions of intervention and outcomes, and used an intent-to-treat analysis or a modified version. Most of the good quality rated trials did not impute missing data in their primary outcomes, but a few used various imputation techniques across some of the outcomes reported. Three trials of Vivitrol and two trials of Probuphine were rated fair, as they had incomparable groups at baseline, differential attrition during follow-up, or were missing up to two of the USPSTF criteria. No trials were rated poor. Further details on the ratings of all included trials are in Appendix Table D3. Comparability of Evidence Across Interventions of Interest As noted above, we identified five key trials for this review. Although these trials were similar in eligibility criteria, patient characteristics, and study duration, we were unable to compare the interventions of interest to each other through quantitative indirect assessment primarily due to variations in study characteristics. For example, five of the studies (all three Probuphine trials, the CAM2038 trial, and Sublocade trial) randomized participants following induction, while randomization was conducted before induction in the two Vivitrol trials. In addition, as seen in Table 3.1 above, there are variations in the outcomes assessed in these trials that are further complicated by the use of non-standard clinical measures. For example, time to relapse was an outcome assessed in only one key trial of Vivitrol, while number of days of heroin and other illicit opioid use was only assessed in the other Vivitrol trial. Furthermore, although the majority of the studies were 24 weeks in length, the timing of outcome assessment differed across trials. The percentage of urine samples negative for opioid use was assessed in the Sublocade trial (combined with negative self-reporting) and CAM2038 trial with durations ranging from five to 24 weeks and four to 24 weeks respectively. These differences are summarized in Table 3.2. below. ©Institute for Clinical and Economic Review, 2018 Page 28 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Table 3.2. Comparability of Evidence: Key Trials Across Interventions of interest Treatment Time of Trial Study Design Duration Detoxification/Induction Period Outcomes Randomization (Weeks) Detoxification: none Phase III RCT • Urine samples used to assess abstinence CAM2038 Lofwall 2018 24 At start of induction Non-inferiority Induction: one day of 4 mg bup/1 mg • Outcome measured over 24 weeks nal Detoxification: none • Combination of urine samples and self- Induction: run-in induction phase 24 report used to assess abstinence Sublocade Trial 13-0001 Phase III RCT with SL bup/nal film followed by After induction • Outcome measured over 24 weeks open-label phase with 8 to 24 mg doses of buprenorphine/naloxone for four to 11 days Detoxification: none • Urine samples and self-report used to Phase III Probuphine Rosenthal 2016 24 Induction: stable dose of 8 mg/day or After induction assess abstinence Non-inferiority less of sublingual buprenorphine • Outcome assessed over 24 weeks received for at least 24 weeks Detoxification: yes, protocols and • Abstinence not reported X-BOT Phase IV Before induction length of time varied by site • Time to relapse event reported Vivitrol Phase III RCT • Urine samples used to assess abstinence Tanum 2017 12 Detoxification: yes After detoxification Non-inferiority • Outcome measured over 12 weeks Bup/nal: buprenorphine/naloxone, d: day, mg: milligram, OUD: opioid use disorder, SL: sublingual ©Institute for Clinical and Economic Review, 2018 Page 29 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Clinical Benefits Mortality We sought evidence on the effect of the interventions of interest on reducing mortality. However, we found no relevant data on this outcome. All-Cause Discontinuation Discontinuation rates appeared similar with CAM2038, Probuphine, and Vivitrol compared with sublingual buprenorphine/naloxone. However, tests of statistical significance were not reported. Of note, significantly more patients discontinued before induction with Vivitrol compared to buprenorphine/naloxone. Results from the placebo-controlled trials of Sublocade and Probuphine showed substantially greater attrition in the placebo group than in the active treatment arms. The most common reasons for discontinuation were lack of efficacy, adverse events, withdrawing consent, being unable to complete induction, loss to follow-up, and withdrawal symptoms. CAM2038 At 28 weeks, a similar proportion of people discontinued in the CAM2038 arm compared to the sublingual buprenorphine/naloxone arm (41% vs 43%; p-value: not reported).76 More than 80% of the total participants who discontinued withdrew consent or were lost to follow-up. Other reasons cited for discontinuation included clinical (physician’s) decision and adverse events. Sublocade During the initial open label two-week run-in period at the start of the Sublocade trial, about a quarter of the total study population (24.2%) did not complete the trial and were not randomized to receive treatment.77 Following the run-in period and after randomization, the number of participants who discontinued were similar for both the 300-mg arm and the 100-mg arm of Sublocade (36% vs. 38%, respectively; p-value: not reported), but substantially lower than for the placebo arm (66%), although no statistical significance was reported.77 Participants mostly withdrew consent or were lost to follow-up (>20% in both active arms). Other reasons included lack of efficacy, adverse events, protocol violations, withdrawal symptoms, non-compliance with study drug, withdrawal by physician’s decision, or site closure by sponsor.77 Additionally, in the OLE, 50% of participants who were newly enrolled dropped out of the trial.78 Probuphine Of the three trials of Probuphine, the Rosenthal 2016 reported very low discontinuation rates, with a total of 7% in the Probuphine arm and 6% in the buprenorphine/naloxone arm with most participants withdrawing or lost to follow-up.79 Importantly, however, those participants who entered the Rosenthal 2016 trial were already stable on buprenorphine/naloxone. In contrast, ©Institute for Clinical and Economic Review, 2018 Page 30 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents higher rates of discontinuation were generally observed in Rosenthal 2013 and Ling 2010. In Rosenthal 2013, discontinuation in the Probuphine arm and sublingual buprenorphine/naloxone arm was 36% versus 37%, respectively (p-value: not reported) with a greater proportion of patients discontinuing in the placebo implant arm (80%).81 Nearly half of the participants in this trial discontinued for reasons unspecified. A similar finding was observed in Ling 2010 (refer to Appendix Table D6). Major reasons cited for discontinuation in these trials included loss to follow- up, consent withdrawal, lack of treatment efficacy, non-compliance/non-adherence, issues from adverse events, or incarceration. Vivitrol In the X-BOT trial, a higher percentage of participants discontinued in the Vivitrol arm than in the buprenorphine/naloxone arm (28% vs. 22%); however statistical significance was not reported. Of note, significantly more participants discontinued before induction in the Vivitrol group compared to the buprenorphine/naloxone group (28% vs. 6%, p-value<0.0001).83 In Tanum 2017 trial, 30% of patients discontinued from the Vivitrol arm compared to 38% in the buprenorphine/naloxone arm, (statistical significance was not reported). However, the study reported similar retention time between the two arms (mean days: 69.3 vs. 63.7; p-value: not significant). In longer-term OLEs of Vivitrol trials lasting 48 to 52 weeks, results showed up to 52% attrition.89,90 Over 80% of the participants who discontinued withdraw consent or were lost to follow-up in both the X-BOT and Tanum 2017 trials. Other reasons cited for discontinuation in these trials include induction, detoxification failure, adverse effects, and incarceration. Abstinence and Relapse Outcomes Abstinence from opioid use was variably defined in available trials. For most interventions, the number of opioid-negative urines did not statistically differ in comparison to sublingual buprenorphine/naloxone. Results from the Probuphine trials showed statistically significantly greater abstinence than buprenorphine/naloxone on various measurements. Participants on Sublocade treatment were also more likely to be abstinent, but in comparison to placebo. Relapse to opioid use was a measure specific to trials of Vivitrol; a statistically significantly higher rate of relapse was seen with Vivitrol versus buprenorphine/naloxone in the intent-to-treat group. CAM2038 In the Lofwall 2018 trial, abstinence was primarily measured by the proportion of opioid-negative urine samples over 24 weeks. The proportion of urine samples that were opioid negative was 35.1% with CAM2038 and 28.4% with buprenorphine/naloxone. This difference of 6.7% (95% CI - 0.1% to 13.6%) excluded the investigator-chosen non-inferiority margin of -11%.76 As a secondary outcome, the mean percentage of opioid-negative samples along with self-report of abstinence ©Institute for Clinical and Economic Review, 2018 Page 31 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents from weeks four to 24 was assessed, and it was found to be significantly higher in the CAM2038 group compared to the buprenorphine/naloxone group (Mean: 35.1% vs. 26.7%, p=0.004). Additionally, during weeks 13 to 24, the mean percentage of opioid-negative samples was significantly higher for participants solely receiving CAM2038 monthly than those receiving buprenorphine/naloxone with a difference of 8.5% (95% CI 1.2% to 15.7%, p=0.02). The proportion of responders to treatment was also reported, defined as having no illicit opioid use assessed by urine tests and self-report both negative during phase one (at least two of three assessments at weeks nine to 11, and week 12) and phase two (at least five of six assessments from weeks 12 to 24, and the last month of treatment). The proportion of treatment responders with CAM2038 and sublingual buprenorphine/naloxone was 17.4% and 14.4%, respectively (difference = 3.0, 95% CI - 4.0 to 9.9).76 Sublocade In the Sublocade trial, abstinence was measured by the percentage of opioid-negative urine samples combined with self-reports negative for illicit opioid use during weeks five to 24. Results of the primary analysis showed that the proportion of participants with 90% or more negative samples was similar between the 300-mg arm and 100-mg arm (24% and 21%), but significantly higher than the placebo arm (2%, p<0.0001 for both comparisons to placebo).77 In a secondary analysis, the numbers of weeks abstinent during weeks five to 24 was the same for the two active arms and also significantly higher than placebo (8.5 for both vs. 1.0, p<0.0001).77 Additionally, results at the 24th week of the treatment period showed a statistically significant higher proportion of abstinent participants in the 300-mg arm than in the 100-mg arm of Sublocade or placebo (44% vs. 37%, vs. 2%, p<0.0001).77 Probuphine In the Rosenthal 2016 trial, abstinence was measured by the combination of urine drug tests and self-report both negative to illicit use of opioids. The primary analysis of the proportion of responders was defined as participants with at least four to six months with no illicit use of opioids based on urine samples and self-report. A higher proportion of participants were abstinent with Probuphine than buprenorphine/naloxone in all participants who received treatment (96.4 vs. 87.6; p=0.03).79 Furthermore, in a separate sensitivity analysis in the intent to treat population, a higher proportion of participants were abstinent with Probuphine than buprenorphine/naloxone over the six-month period (80.5% vs. 66.7%, p= 0.04).79 At six months, a greater proportion of participants were abstinent with Probuphine than buprenorphine/naloxone (85.7% vs. 71.9%; p=0.03).79 The two similar Probuphine trials (Ling 2010 and Rosenthal 2013) differed in the measurement of abstinence. Ling 2010 measured abstinence as the percentage of a total of 48 urine samples negative for illicit opioid use during the first 16 weeks, and showed a greater percentage of reported opioid-negative samples with the buprenorphine implant than placebo (40.4% vs. 28.3% ©Institute for Clinical and Economic Review, 2018 Page 32 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents p=0.04).80 Rosenthal 2013 defined abstinence as the percent of opioid-negative urines during the entire study duration. Rosenthal 2013 found a greater proportion of reported opioid-negative urine samples in the buprenorphine implant arm than with the placebo implant (31.2% vs. 13.4%, p<0.0001); the proportion with open label sublingual buprenorphine/naloxone was 33.5%. Vivitrol We found data on abstinence and relapse outcomes in three Vivitrol trials. The abstinence-based outcome presented in Tanum 2017 was defined as the proportion of the total number of urine drug tests with no opioid use. The proportion of the total number of opioid-negative urine drug tests with Vivitrol and buprenorphine/naloxone were 0.9 and 0.8, respectively (difference = 0.1, 95% CI - 0.04 to 0.2).84 In addition, Krupitsky 2011 showed a statistically significant median proportion of weeks with confirmed abstinence to be greater for Vivitrol compared to placebo (90.0% vs. 35.0%, respectively) during weeks five to 24. Confirmed abstinence was defined as a negative urine drug test and negative self-report for opioid use.86 Also, a significantly higher proportion of participants with confirmed abstinence occurred with Vivitrol than placebo (35.7% vs. 22.6%, Relative Risk (RR) =1.58, 95% CI 1.06 to 2.36, p=0.0224).86 During weeks three to 24, the X-BOT trial found a higher proportion of participants in the intent-to-treat group relapsed after 20 days on Vivitrol compared to those on sublingual buprenorphine/naloxone film (65% vs. 57%, p=0.036). In the per-protocol group, more than half the participants relapsed in both groups, and the proportions were not statistically significantly different from each other. 83 Due to its antagonistic action, patients on Vivitrol would not experience any effect from taking opioids, but the relapse rates in the per protocol group may be indicative of continued craving while on Vivitrol. Diminishing Illicit Use of Opioids Vivitrol was the only intervention with data on diminishing illicit use of opioids which was assessed in one key trial. That trial found that Vivitrol decreased use of heroin and other illicit opioids when compared to buprenorphine/naloxone over the duration of the trial. CAM2038 No data on diminishing opioid use were reported in the CAM2038 trial. Sublocade No data on diminishing opioid use were reported in the Sublocade trial. Probuphine No data on diminishing opioid use were reported in the Probuphine trial. ©Institute for Clinical and Economic Review, 2018 Page 33 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Vivitrol Tanum 2017 was the only key trial reporting on diminishing illicit use of opioids. On average during the 12-week trial, patients treated with Vivitrol had fewer days of heroin use and illicit drug use than patients treated with buprenorphine/naloxone (heroin use: -3.2 days, 95% CI -4.9 to -1.5); illicit drug use: -2.7, 95% CI -4.6 to -0.9). At 12 weeks, results also showed that participants receiving Vivitrol had fewer mean days of heroin use than those receiving buprenorphine/naloxone (1.1 vs. 4.1, respectively; mean difference = -3.6 days, 95% CI -6.0 to -1.2), although differences were not statistically significant for other illicit opioid use.84 Its associated 48-week OLE showed a continued decrease in heroin use days and illicit drug use days.89 Evidence on intravenous drug use was mixed; the eight-week trial (Lee 2015) showed a higher percent of participants with heroin use after release from prison with Vivitrol than treatment as usual in the first month.91 In contrast, at 24 weeks in Lee 2016, the proportion of criminal justice offenders using any intravenous drugs was numerically higher in the treatment as usual group versus Vivitrol, although this was not statistically significant.92 Opioid craving – Visual Analog Scale Opioid craving scores on CAM2038 and Probuphine were not statistically significantly different from those on buprenorphine/naloxone. Sublocade decreased opioid craving compared with placebo. One trial found numerically lower opioid craving scores with Vivitrol than buprenorphine/naloxone, but statistical significance was not reported. Opioid craving is generally defined as a desire to use opioids. It is commonly measured with self- reported questionnaires using the Visual Analogue Scale (VAS). A total of eight studies in our included study set assessed for opioid craving using the VAS, and they are summarized in Table 3.3. below. ©Institute for Clinical and Economic Review, 2018 Page 34 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Table 3.3. Opioid Craving – VAS Scores* in Key Trials Study Duration of Treatment Arm Mean VAS Over Duration Mean VAS Change from N p-Value Study Design Follow-Up Dosage of Follow-Up Baseline CAM2038 CAM2038 213 17.3 (SD: 25.5)† NR Lofwall 201876, RCT 24 weeks NR Bup/nal, SL 215 17.3 (SD: 25.5)† NR Sublocade Sublocade 300mg/100mg 192 ¤ NR 2.1 (SE: 1.63) vs. placebo: p=0.0003 Trial 13-0001, RCT 24 weeks Sublocade 300mg/300mg 193¤ NR -0.9 (SE:1.63) vs. placebo: p<0.0001 Placebo 96¤ NR 11.5 (SE: 2.48) -- Probuphine -2.3 (SD: 11.15)‡; Probuphine 84 NR -2.7 (SD: 12.58)† Rosenthal 201679, RCT 24 weeks NS for both -2.8 (SD: 19.57)‡; Bup/nal, SL 89 NR -1.9 (SD: 18.97)† Vivitrol Vivitrol 56 0.83 (95% CI: -0.81 to 2.43)§ NR Tanum 201784, RCT 12 weeks NR Bup/nal, SL 49 2.69 (95% CI: 1.77 to 3.60)§ NR Bup: buprenorphine, mg/d: milligrams per day, nal: naloxone, RCT: randomized controlled trial, SL: sublingual, VAS: visual analog scale *Opioid craving measured on 100 mm scale, where 0=no craving and 100=strongest craving, unless otherwise noted. †Mean VAS need-to-use score, where 0=no need and 100=strongest need. ‡Mean VAS desire-to-use score, where 0=no desire and 100=strongest desire. §Craving for heroin, rated on a scale of 0=no craving to 10=very strong. #Data reported are percentage of patients reporting opioid craving increases, decreases, or no changes compared to baseline, where 0=no craving and 10=strongest craving. ¤Number of participants analyzed for outcome. ©Institute for Clinical and Economic Review, 2018 Page 35 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents CAM2038 Lofwall 2018 showed no differences in opioid cravings between CAM2038 injections and sublingual buprenorphine/naloxone over 24 weeks. Groups were not compared statistically; however, data presented in a graph indicates that CAM2038 and buprenorphine/naloxone resulted in identical Mean VAS scores for opioid craving after 24 weeks of treatment (17.3, SD: 25.5).76 Likewise, both groups showed a similar pattern in VAS scores over the trial duration. Sublocade Compared to placebo, the two Sublocade arms each showed a decrease in opioid cravings (300-mg arm: mean difference = -12.4, 95%CI -17.5 to -7.3, p<0.0001; 100-mg arm: mean difference = -9.4, 95% CI -14.6 to -4.3). Refer to Table 3.3 above.77 Probuphine In the Rosenthal 2016 trial, there were no significant differences between Probuphine and buprenorphine/naloxone in the mean change from baseline of desire-to-use and need-to-use scores by six months.79 In the two trials (Ling 2010 and Rosenthal 2013) comparing Probuphine to placebo implants, a statistically significant benefit favoring Probuphine over placebo was seen when comparing mean VAS scores over 24 weeks (see Appendix Table D7).80,81 Ling 2010 reported a mean VAS score of 9.9 (95% CI: 7.8 to 12.0) versus 15.8 (95% CI: 12.7 to 18.9) in the buprenorphine and placebo groups, respectively. Similarly, Rosenthal 2013 reported a mean VAS score of 10.2 versus 21.8 in Probuphine and placebo implant groups, respectively. Rosenthal 2013 also compared the same active implant to open-label buprenorphine/naloxone and found no statistically significant difference between the two treatments. Vivitrol One key study of Vivitrol (Tanum 2017) comparing Vivitrol to buprenorphine/naloxone assessed opioid craving specific to heroin using a 0-10 VAS scale. The study found cravings to be numerically less intense among participants on Vivitrol compared to the buprenorphine/naloxone arm; however, the statistical significance of this difference was not reported (Table 3.3).84 In addition, two placebo-controlled trials of Vivitrol also reported on opioid cravings. The NEW HOPE trial compared Vivitrol to placebo and reported the percentages of participants reporting improved, worsened, or stable opioid cravings among a small subset of participants.88 These results were not statistically compared. Approximately 10% more participants in the Vivitrol group versus placebo reported experiencing decreased opioid cravings (43.8% vs. 33.3%, respectively) through 24 weeks of treatment while similar proportions of the Vivitrol and placebo groups reported increased opioid craving (18.8% vs. 20.0%, respectively) 88 ©Institute for Clinical and Economic Review, 2018 Page 36 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Krupitsky 2011 reported a statistically significant treatment difference in opioid craving by VAS scale favoring Vivitrol over placebo treatment after 24 weeks of treatment (Vivitrol mean VAS change from baseline: -10.1; placebo: 0.7; difference = -10.7, p<0.0001).86 Additional data averaging results from weeks eight to 24 showed similar results (Vivitrol: -9.4, placebo: 0.8).93 Opioid Withdrawal No significant differences were shown for CAM2038 and Probuphine each in comparison to buprenorphine/naloxone in the Clinical Opiate Withdrawal Scale (COWS) and Subjective Opiate Withdrawal Scale (SOWS). Only the higher dose arm of Sublocade showed any significant difference from placebo. There were no COWS or SOWS data for Vivitrol. COWS is administered by clinicians and measures 11 common opioid withdrawal symptoms. Higher scores indicate worse clinician-reported withdrawal symptoms. Four common thresholds are used to indicate mild (5-12), moderate (13-24), moderately severe (25-36), and severe (scores >36) opioid withdrawal symptoms. SOWS is a self-reported measure for grading 16 common opioid withdrawal symptoms. Each symptom is graded on a scale of 0 (not at all) to 4 (extremely) for a total score of 64, where higher scores indicate more severe patient-reported withdrawal symptoms. Both COWS and SOWS evaluate common opioid withdrawal symptoms including pulse rate, sweating, restlessness, pupil size, bone and joint aches, gastrointestinal upset, tremors, yawning, goosebumps, and anxiety or irritability. Three key trials included in our systematic review measured COWS over 24 weeks. Two studies reported both COWS and SOWS data; no studies reported only SOWS. Reporting of opioid withdrawal symptoms was more homogenous than reporting of opioid craving. Neither the COWS or SOWS scores, however, have an established and/or validated definition of a minimum clinically important difference, so it is unclear whether the changes reported in the trials summarized below represent clinically meaningful improvements for patients with OUD. CAM2038 Lofwall 2018 showed no differences between CAM2038 injections and sublingual buprenorphine/naloxone over 24 weeks.76 Groups were not compared statistically, however, graphical data published shows nearly identical weekly (weeks one to 12) or monthly (weeks 12-24) COWS scores over 24 weeks. COWS data estimated from this graph reported mean COWS scores of 3.3 (SD 3.5) and 2.7 (SD 4.0) for the CAM2038 and sublingual buprenorphine/naloxone groups, respectively. ©Institute for Clinical and Economic Review, 2018 Page 37 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Sublocade The Sublocade trial showed a significant difference in COWS and SOWS in the 300-mg arm when compared to placebo (COWS -1.1 vs. -0.1, mean difference: -1.0, 95% CI -1.72 to -0.23; SOWS -2.0 vs. 0.7, mean difference: -2.6, 95% CI -4.32 to -0.90), but no significant difference in the 100-mg arm.77 Probuphine Rosenthal 2016 showed no significant differences in COWS and SOWS from baseline between Probuphine and buprenorphine/naloxone.79 In addition, two studies – Ling 2010 and Rosenthal 2013 – evaluated buprenorphine implants versus placebo.80,81 Both studies showed buprenorphine implants were statistically superior compared to placebo implants over 24 weeks of treatment in COWS (2.49 vs. 4.52, p<0.001) and SOWS (5.30 vs. 8.42, p<0.0001). As noted above, Rosenthal 2013 also included an open-label buprenorphine/naloxone group.81 Exploratory analyses showed buprenorphine implants was associated with worse withdrawal symptoms compared to open-label buprenorphine/naloxone (COWS 2.49 vs. 1.71, , p=0.0005; SOWS 5.30 vs. 2.83, , p=0.0006). Vivitrol There was no evidence on COWS or SOWS in the key trials of Vivitrol. Health-Related Quality of Life and Other Outcomes Evidence on health-related quality of life and patient specific outcomes were reported only in trials of Vivitrol. Results showed an overall increase in quality of life in patients receiving Vivitrol compared with placebo. Patient satisfaction with treatment occurred more with Vivitrol than with buprenorphine/naloxone. CAM2038 No data on health-related quality of life, incidence of infectious diseases, functional outcomes, employment-related outcomes, diversion, accidental pediatric exposure, or other patient outcomes were reported in the CAM2038 trial. Sublocade No data on health-related quality of life, incidence of infectious diseases, functional outcomes, employment outcomes, diversion, accidental pediatric exposure, or other patient outcomes were reported in the Sublocade trial on clinicaltrials.gov. ©Institute for Clinical and Economic Review, 2018 Page 38 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Probuphine No data on health-related quality of life, incidence of infectious diseases, functional outcomes, employment-related outcomes, diversion, or other patient outcomes were reported in the Probuphine trial. Vivitrol Health-related quality of life was measured only in the Vivitrol placebo-controlled trial. This was assessed using the 36-item short form health survey (SF-36) and the VAS self-rating assessment of patients’ general health EuroQol-5 dimensions questionnaire.86 In the 24-week trial, the difference in mean change from baseline in patients’ VAS assessments showed a statistically significant increase in general health with Vivitrol when compared to placebo (14.1 vs. 2.7, difference of 11.4, 95% CI 5.0-17.8, p-value = 0.0005).86 The mean score for the mental component of SF-36 was also higher in the Vivitrol arm than the placebo arm (50.37 vs. 45.28; difference of 5.09 95%CI 2.09 to 8.09; p-value not reported).86 In addition, patient satisfaction was assessed in one of the key trials of Vivitrol (Tanum 2017). Tanum 2017 measured satisfaction with treatment by the visual analog scale (VAS) with 0 indicating very low and 10 indicating very high. By 12 weeks, participants’ satisfaction with treatment was found to be higher among participants receiving Vivitrol than those receiving buprenorphine/naloxone (estimated values: 8.61 vs 3.66; p-value not reported).84 No data on incidence of infectious diseases, functional outcomes, employment-related outcomes, accidental pediatric exposure, or diversion were reported in the Vivitrol trials that met our inclusion criteria. Health Care Utilization Limited data were reported on health care utilization, and only for Vivitrol. Evidence from available trials found no differences in health care utilization between Vivitrol and treatment as usual. Results from one observational study showed reduced inpatient admissions with Vivitrol. Health care utilization was reported in a post-hoc analysis of Lee 2016.92,94 During the 24-week treatment phase of the study, the percentage of participants with health care utilization, defined as any ED visits or hospital admissions, did not significantly differ between patients randomized to Vivitrol and treatment as usual (31.5% vs. 35.0%, respectively).94 When stratified by health care utilization type, there also were no significant differences in the percentages of participants randomized to Vivitrol and treatment as usual in terms of drug detox hospitalizations (2.7% vs. 2.1%), psychiatric hospitalizations (1.4% vs. 3.5%), or ED visits (25.3% vs. 28.0%).94 However, significantly fewer patients randomized to Vivitrol had medical or surgical hospitalizations ©Institute for Clinical and Economic Review, 2018 Page 39 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents compared to patients randomized to treatment as usual (6.9% vs. 14.0%; incidence rate ratio [IRR]=0.37 95% CI: 0.16, 0.88; p=0.02).94 We also identified one observational study assessing health care utilization among cohorts of patients with OUD receiving treatment with Vivitrol (n=1,041), non-specific buprenorphine (n=20,566), and nonpharmacological therapy (n=6,883).95 Comparing the 12-month period before treatment (baseline) to the 12-month period after starting treatment (follow-up), the mean number of inpatient admissions was reduced by 46.6%, 20.8%, and 15.1% in the Vivitrol, buprenorphine, and nonpharmacological therapy cohorts, respectively (p<0.05 baseline vs. follow-up for all); the study did not report the statistical significance of the differences among the cohorts. The mean days in inpatient care was significantly reduced in the Vivitrol (-56.8%) and buprenorphine cohorts (-8.8%) (p<0.05 baseline vs. follow-up for both) but not in the nonpharmacological therapy cohort (- 0.6%). In addition, the mean number of ED visits was reduced by 26.1% in the Vivitrol cohort, 13.3% in the buprenorphine cohort, and 15.5% in the nonpharmacological therapy cohort (p<0.05 baseline vs. follow-up for all). Harms Serious adverse events were generally low and similar in trials of CAM2038, Probuphine, and Vivitrol in comparison to buprenorphine/naloxone and in the Sublocade trial vs placebo. Discontinuation due to adverse events was not reported in most trials. Results from one Vivitrol trial showed that similarly low numbers of participants discontinued when compared to buprenorphine/naloxone. The most common adverse events reported in the trials were injection/implant site pain, gastrointestinal issues, headaches, and insomnia. In the CAM2038 trial (Lofwall 2018), the incidence of serious adverse events (SAEs) occurred in 2.3% of participants receiving CAM2038 arm versus 6% of those receiving sublingual buprenorphine/naloxone at 24 weeks (see Table 3.4 below). Seven participants discontinued due to AEs in the CAM2038 arm, while three discontinued in the buprenorphine/naloxone arm.76 Although no participant died from overdose, there was one unrelated death in the buprenorphine/naloxone arm of the trial. The most commonly occurring AEs in the CAM2038 arm (see Table 3.4 below) were injection-site pain, injection-site pruritus and erythema, headache, constipation, and nausea. Additional AEs of urinary tract infection and insomnia occurred in the buprenorphine/naloxone arm.76 During the 24-week Sublocade trial, results showed SAEs occurring in <5% of participants in the 300 mg Sublocade arm versus 2% of participants in the 100 mg arm and 5% in the placebo arm. One death occurred in the 300-mg Sublocade arm unrelated to overdose. Ten participants discontinued due to AEs in the 300-mg arm, seven discontinued due to AEs in the 100-mg arm, and two discontinued in the placebo arm. The most common AE in the Sublocade arms were similar to ©Institute for Clinical and Economic Review, 2018 Page 40 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents CAM2038, and included gastrointestinal disorders, injection site pruritus and pain, upper respiratory tract infections, nasopharyngitis, headache, and insomnia. In the 24-week Probuphine key trial,79 similar proportions of participants receiving Probuphine and placebo had an SAE (2.3% vs. 3.4%). Similar results were seen in the other trials with the exception of Ling 2010 which had a higher rate of SAEs occurring in the placebo arm.80,81 There was one accidental pediatric exposure in the buprenorphine/naloxone arm that led to the notification of child protective services.79 Deaths were unreported in Rosenthal 2016. However, one death occurred in the Rosenthal 2013 trial and was overdose-related in the buprenorphine/naloxone arm.81 In the Rosenthal 2016 trial, only one participant discontinued due to an AE in the Probuphine arm.79 Whereas, no participant discontinued due to AEs in the Rosenthal 2013 trial and four discontinued due to AEs in the Ling 2010 trial both in the Probuphine arms (see Appendix Table D9).80,81 The most common AEs were related to the implant site (erythema, itching, pain), headaches, gastrointestinal problems, and nasopharyngitis. In the X-BOT trial, at 24 weeks the incidence of AEs was 14% and 11% in the Vivitrol and sublingual buprenorphine/naloxone arms, respectively.83 Six participants discontinued due to the occurrence of an AE in the Vivitrol arm, while eight discontinued in the buprenorphine/naloxone arm. Out of the five total fatalities that occurred, two were in the Vivitrol arm with the other three in the buprenorphine/naloxone arm. Overdose was the cause of death in two of the three fatalities in the Vivitrol arm and three of the four in the buprenorphine/naloxone arm.83 No deaths occurred in the Tanum 2017 trial. At 12 weeks, a similar proportion of participants that had experienced a SAE in the Vivitrol arm was 8.5%, whereas 4.2% experienced a SAE in the buprenorphine/naloxone arm.84 Four participants discontinued due to AE in the Vivitrol arm, while six discontinued in the buprenorphine/naloxone arm in this trial.84 The most common AEs (also reported in four other trials of Vivitrol versus usual treatment or placebo) were insomnia, psychiatric disorders (anxiety and depression symptoms), and injection site problems. The label for Vivitrol warns of an increased risk of opioid overdose fatalities, as participants on Vivitrol have reduced tolerance to opioids given that complete withdrawal is a prerequisite for Vivitrol therapy. The label reports that cases of opioid overdose deaths have been reported in patients who used opioids at the end of a dosing interval, after missing a dose, and after discontinuation.96 We identified a review of case narratives of overdose fatalities among patients who received Vivitrol that were reported to the FDA Adverse Event Reporting System. Results show that most of the overdose deaths occurred within 28 to 56 days after the last reported Vivitrol dose.97 However, it is unclear whether this time interval corresponds to a biological rebound risk of overdose. ©Institute for Clinical and Economic Review, 2018 Page 41 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Table 3.4. % of SAE, Discontinuation due to AE (%), and Deaths SAE, n (%) Discontinued Due to AE, n (%) Death, n (%) CAM2038 (Lofwall 2018) 76 CAM2038 5 (2.3) 7 (3.3) 1 (0.5) Bup/Nal 13 (6.0) 3 (1.4) 0 Trial 13-000177 Sublocade (300mg/300mg) 7 (3.5) 10 (5.0) 1 (0.5) Sublocade (300mg/100mg) 4 (2.0) 7 (3.4) 0 Placebo 5 (5.0) 2 (2.0) 0 Buprenorphine Implant79 Probuphine 2 (2.3) 1 (1.1) NR Bup/Nal 3 (3.4) 0 NR X-BOT83 Vivitrol 29 (14) 6 (2.1) 3 (1.1) Bup/Nal 29 (11) 8 (2.8) 4 (1.4) Tanum 201784 Vivitrol 8.5 4 (5.6) 0 Bup/Nal 4.2 6 (8.3) 0 ©Institute for Clinical and Economic Review, 2018 Page 42 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Table 3.5. ≥5% AEs Headache Nasopharyngitis Injection/Implant Injection/Implant Site Insomnia Depression (%) GI Upset (%) (%) (%) Site Problems (%) Pain (%) (%) CAM2038 (Lofwall 2018)76 6.1 (pruritus); 7.5 (constipation); 7.0 CAM2038 7.5 8.9 5.6 5.6 (erythema) (nausea) 6.8 (pruritus); 7.4 (constipation); 7.9 Bup/Nal 7.9 7.9 2.8 5.6 (erythema) (Nausea) Trial 13-000177 Sublocade 8.0 (constipation), 8.0 8.5 5.0 9.5 (pruritus) 6.0 8.5 (300mg/300mg) (nausea), 5.5 (vomiting) Sublocade 9.4 (constipation); 8.9 9.4 5.4 6.4 (pruritus) 4.9 6.4 (300mg/100mg) (nausea); 9.4 (vomiting) 0 (constipation), 5 Placebo 6.0 1.0 4.0 (pruritus) 3.0 11 (nausea), 4 (vomiting) Buprenorphine Implants79 Probuphine 6.9 8.0 13.8 6.9 8.0 Bup/Nal 3.4 4.5 7.9 2.2 1.1 X-BOT83 Vivitrol 16.3 (any) 12.0 Bup/Nal NA 20.6 Tanum 201784 16.9 (anxiety or Vivitrol 5.6 11.3 depression) 8.3 (anxiety or Bup/Nal NA 4.2 depression) ©Institute for Clinical and Economic Review, 2018 Page 43 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Comparator Evidence Methadone and Buprenorphine/Naloxone We did not conduct a systematic search for methadone or buprenorphine/naloxone; rather, we identified and summarized previous systematic reviews published by Cochrane and CADTH, as listed below: • Methadone maintenance therapy versus no opioid replacement therapy for opioid dependence98 • Buprenorphine/naloxone versus methadone for the treatment of opioid dependence99 We were unable to identify a systematic review for buprenorphine/naloxone in comparison to no active treatment. Therefore, we summarized the results from the sole placebo-controlled trial of buprenorphine/naloxone.100 Fudala et al. was a multicenter, placebo-controlled trial including 326 patients with OUD (DSM-IV). Patients were randomly allocated to daily treatment with sublingual tablets of buprenorphine/naloxone, buprenorphine alone, or placebo for four weeks. The trial was terminated early because buprenorphine/naloxone and buprenorphine alone showed much greater efficacy than placebo. After four weeks, 268 patients from the placebo-controlled phase and 193 newly-enrolled patients entered a 48- and 52-week open-label safety study of buprenorphine/naloxone, respectively. We also highlighted findings from a systematic review of RCTs of buprenorphine-containing formulations to supplement the results from Fudala et al.101 This systematic review evaluated the efficacy of the buprenorphine-containing formulations at low (2-6 mg), medium (7-15 mg), and high doses (≥16 mg) compared to placebo; of note, the authors of this systematic review considered 1 mg of buprenorphine as placebo in addition to true placebo. We summarized results for the medium and high dose levels as the doses of buprenorphine in Fudala et al. and the key trials in our review are within the medium and high dose range. Mortality Four studies included in the 2009 Cochrane systematic review of RCTs comparing methadone maintenance treatment to therapies not involving opioid agonists (i.e., placebo, detoxification, nonpharmacological therapy, wait-list controls) for the treatment of OUD reported the number of deaths among patients. Three out of 287 (1%) patients on methadone maintenance treatment and eight out of 289 (2.8%) patients receiving no opioid agonist therapy had died at one to 36 months of follow-up.98 Five of the eight deaths associated with no opioid agonist therapy resulted from fatal overdoses, and the causes for the other three deaths were not reported. At three years of follow- up, only one of the three deaths associated with methadone maintenance treatment resulted from an alleged overdose.102 Results from the meta-analysis found that methadone maintenance treatment was not statistically significantly more effective in preventing deaths compared to no ©Institute for Clinical and Economic Review, 2018 Page 44 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents opioid agonist therapy, although the risk ratio points to a reduced risk of mortality for patients on methadone maintenance treatment (risk ratio [RR]: 0.48; 95% CI: 0.10 to 2.39).98 We did not find any information from systematic reviews on buprenorphine/naloxone compared to no active treatment or methadone in the ability to prevent deaths. All-Cause Discontinuation/Study Retention The 2009 Cochrane systematic review found methadone maintenance treatment to be more effective in retaining patients in treatment compared to therapy with no opioid agonists. To analyze treatment retention, the authors of this review stratified the included studies by those conducted before and after 2000 since they speculated that differences in results occurred over time; however, it is unclear why the authors chose the year 2000 as a cut point and why they did not conduct similar subgroup analyses for the other outcomes (mortality and abstinence). Nevertheless, treatment retention was found to be significantly higher with methadone maintenance treatment compared to no opioid agonist therapy in the older studies at six to 32 weeks (68.1% vs. 25.1% RR: 3.05; 95% CI: 1.75 to 5.35, three studies, 505 patients) and even higher in the newer studies at four to 24 weeks (73.4% vs. 16.4%; RR: 4.44; 95% CI: 3.26 to 6.04, four studies, 750 patients).98 In Fudala et al, 10.1% of patients receiving buprenorphine/naloxone, 3.8% receiving buprenorphine, and 11.0% receiving placebo discontinued from the four-week placebo-controlled trial; reasons for discontinuation were not reported.100 Of the 472 patients assessed for safety in the 48- and 52- week open-label phase of Fudala et al., 385 (81.6%) patients received at least eight weeks and 261 (55.3%) patients received at least six months of treatment with buprenorphine/naloxone.100 The 2014 Cochrane systematic review and meta-analysis of RCTs of buprenorphine-containing formulations showed treatment with buprenorphine-containing formulations was associated with significantly higher treatment retention compared to placebo at medium doses at two to 48 weeks (65.3% vs 37.6% RR: 1.74; 95% CI: 1.06 to 2.87, four studies, 887 patients) and high doses at four to 48 weeks (65.5% vs. 39.7% RR: 1.82; 95% CI: 1.15 to 2.90, five studies, 1,001 patients).101 Five studies summarized in the 2016 CADTH review comparing buprenorphine/naloxone and methadone reported the percent of patients completing treatment with methadone and buprenorphine/naloxone. Three of these studies measured retention at six months and found methadone retained between 46.4% and 74.1% of patients and buprenorphine/naloxone retained between 30.3% and 50.0% of patients103-105; two of these studies reported statistical significance favoring methadone (p<0.01).103,105 Opioid Abstinence The 2009 Cochrane systematic review found methadone maintenance treatment to be effective in reducing illicit use of opioids compared to no opioid agonist therapy. A meta-analysis of six trials ©Institute for Clinical and Economic Review, 2018 Page 45 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents including 1,129 patients showed 45.7% of patients on methadone maintenance therapy versus 66.5% of patients on no opioid agonist therapy had positive hair or urine analysis at four to 16 weeks of follow-up (RR: 0.66; 95% CI: 0.56 to 0.78).98 In Fudala et al., buprenorphine alone and buprenorphine/naloxone both reduced use of opioids as measured by urine tests. At four weeks, the percent of negative urine samples was 17.8% in the buprenorphine/naloxone arm and 20.7% in the buprenorphine alone arm compared to 5.8% in the placebo arm (p<0.001 for both vs. placebo).100 The overall rate of opioid-negative urine samples increased during the open-label study to 67% at week 52.100 The 2014 Cochrane meta-analysis showed maintenance treatment with buprenorphine significantly reduced the use of opioids as measured by the percent of positive urine tests compared to placebo at high doses at four to 48 weeks (standardized mean difference [d]: -1.17; 95% CI: -1.85 to -0.49; three studies; 729 patients) but not at medium doses at two to 16 weeks (d: -0.08 95% CI: -0.78 to 0.62; two studies; 463 patients).101 Seven of the 10 studies included in the 2016 CADTH review measured the use of opioids with urine tests. These studies presented mixed evidence on the efficacy of buprenorphine/naloxone and methadone in reducing illicit use of opioids. Point estimates in five of these studies showed patients treated with buprenorphine/naloxone were numerically less likely to test positive for opioids compared to patients receiving methadone, with two studies reporting statistical significance. A longitudinal one-year study including 3,812 outpatients with OUD reported 47% of patients receiving buprenorphine/naloxone compared to 70% of patients receiving methadone had opioid- and cocaine-positive urine samples at one year of follow-up (p<0.001).106. Also, a 12-week RCT reported that the percentage of urine tests positive for opioids was 0.2% for patients treated with buprenorphine/naloxone versus 1.5% for patients treated with methadone (p=0.03).107 Point estimates in the remaining two of the seven studies suggested patients receiving methadone were numerically less likely to test positive for opioids compared to patients treated with buprenorphine/naloxone, with one study reporting statistical significance; this open-label extension of a RCT of methadone versus buprenorphine/naloxone followed 795 patients for an average of 4.5 years and found 42.8% versus 31.7% of patients randomized to buprenorphine/naloxone and methadone, respectively, had positive urine samples at follow up (p<0.01).103,108 Controversies and Uncertainties As outlined in the section on comparability of evidence, differences in trial designs, population selection, comparators, and outcome measures precluded formal comparisons between the different extended-release formulations. All four formulations also differ in their labeled or potential treatment indications; for example, the manufacturer proposes that the indications for CAM2038 should include OUD treatment directly after diagnosis.24 Sublocade and Probuphine must be preceded by daily transmucosal use of buprenorphine and Vivitrol by a period of medically supervised opioid withdrawal. Probuphine implants should be used for patients on maintenance ©Institute for Clinical and Economic Review, 2018 Page 46 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents treatment with a transmucosal buprenorphine‐containing product delivering a low to moderate dose, the equivalent of buprenorphine 8 mg or less per day. The effective required buprenorphine dosage for most patients is between 12 and 16 mg daily, therefore only patients who can tolerate such doses may be suitable for Probuphine implants. Various outcome measures were used in the trials of the interventions of interest. Outcome measures are based on different calculations of negative urine samples (Appendix Table D5) and then defining relapse based on some percentage of positive urine samples. However, the clinical term “relapse” refers to a person with OUD who in remission and then experiences a loss of control. A relapse is different from a return to opioid use that is limited in scope and time and that does not involve the return of the signs or symptoms of OUD. It is not certain to which degree different rates of negative urine samples constitute a meaningful measure of success, even for the short duration of the trials. The lack of any clear guidance on how to obtain the opioid-free state needed for starting Vivitrol makes comparisons between the evidence for the extended-release agonist formulations and the extended-release antagonist formulation very difficult. Head-to-head trials of agonist formulations should be possible, but have not yet been conducted. In the real world, OUD patients often present with important psychiatric comorbidities, such as depression, post-traumatic stress disorder, and personality disorders.11 Patients with psychiatric comorbidities are largely excluded from the trials (refer to Appendix Table D2), thus limiting their generalizability. This is not limited to the evidence on extended-release formulations, but present in the evidence base for all MATs.26 As noted by SAMHSA in the 2018 TIP, no evidence clearly predicts which patients are best treated with Vivitrol versus methadone or buprenorphine formulations.12 The treatment sequences for different subpopulations with OUD cannot be based solely on the available evidence, but rather must be informed by clinical knowledge and the local context. The evidence on the use of the extended-release formulations is subject to the same general limitations as for the other medications for OUD. It is not yet known if or when to best taper these medications, 12 and evidence is lacking on the added value of the different types of counseling and psychosocial support required by the FDA label and the most recent clinical practice guideline.12 The available research focuses on short-term outcomes and does not provide any evidence regarding observed reductions or patient control of drug use that are of clinical and social benefit, even if opioid use has not completely stopped.19,27 In addition, questions around the impact of extended-release formulations on critically important outcomes, such as overdose and other OUD- associated mortality, health-related quality of life, work productivity, educational attainment, and incarceration have largely gone unanswered by the evidence currently available. ©Institute for Clinical and Economic Review, 2018 Page 47 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents 3.4 Summary and Comment Using the ICER Evidence Matrix (Figure 3.1), we assigned evidence ratings independently for each of the interventions of interest compared to transmucosal buprenorphine/naloxone for study participants with OUD being considered for MAT. We recognize that comparisons of Sublocade, Probuphine, and Vivitrol each versus placebo, for which we have relevant data, have shown incremental benefits. However, the most policy relevant comparisons are those involving the interventions of interest with the active treatment of transmucosal buprenorphine/naloxone. Figure 3.1. ICER Evidence Rating Matrix ©Institute for Clinical and Economic Review, 2018 Page 48 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Table 3.6. Evidence Ratings (Versus Transmucosal Buprenorphine/Naloxone) Comparisons Evidence Rating CAM2038 C+ Sublocade I Probuphine P/I Vivitrol C CAM2038 Evidence for CAM2038 is comprised of one 24-week Phase III trial in comparison to buprenorphine/naloxone. Data was limited on clinical outcomes due to the limed number of trials available for synthesis. Results found CAM2038 to be non-inferior to buprenorphine, but not significantly different in abstinence, opioid craving, and opioid withdrawal. Similarly, while discontinuation rates were high, they did not differ between the active arms, and safety profiles were also comparable. For participants with OUD being considered for MAT, we have moderate certainty that CAM2038 provides a small, or substantial net health benefit given the increased convenience and provider interaction associated with subcutaneous injections, but high certainty that it is at least comparable as it is a buprenorphine-containing treatment. Therefore, we consider the evidence on CAM2038 to be comparable or better (C+). Sublocade Evidence for Sublocade is limited to one 24-week Phase III trial compared to placebo. Presently, there are no head-to-head trials comparing Sublocade to buprenorphine/naloxone. Therefore, we consider the evidence on Sublocade compared to buprenorphine/naloxone to be insufficient (I). Probuphine Evidence for Probuphine compared to buprenorphine/naloxone comprises two 24-week Phase III trials, although only one was considered key. Due to the inclusion criteria and trial design, the populations in the trials may be different from the general population being considered for MAT. The key trial included only participants who were clinically stable and receiving buprenorphine tablets for at least 24 weeks before the trial. Additionally, the other trial excluded participants with severe opioid withdrawal symptoms and cravings, which may have inflated the reported benefits of Probuphine on abstinence outcomes in this trial. No significant differences were found for opioid craving and opioid withdrawal. Similar rates of discontinuation occurred between both active arms, along with similar proportions of serious adverse events. For participants with OUD being considered for MAT, we have moderate certainty of a comparable or small net health benefit for the trial populations. However, we have concerns that the study population may not be reflective of the more general population being considered for MAT. Therefore, we consider the evidence on Probuphine in comparison to buprenorphine/naloxone to be promising but inconclusive (P/I). ©Institute for Clinical and Economic Review, 2018 Page 49 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Vivitrol Evidence for Vivitrol compared to buprenorphine/naloxone consists of data derived from two trials: one 24-week Phase IV trial, and one shorter 12-week Phase III trial. Results found that Vivitrol was non-inferior to buprenorphine/naloxone on a variety of abstinence outcomes. However, a higher rate of relapse was seen with Vivitrol compared to buprenorphine/naloxone in the intent-to-treat group. Results showed a significant reduction only in heroin use, but not for the use of other illicit opioids. A higher rate of discontinuation was found during induction with Vivitrol than buprenorphine/naloxone, which speaks to the difficulties encountered in attempts to successfully withdraw from all opioid use. In terms of safety, serious adverse events were similar between both active arms. However, while not a phenomenon observed during the clinical trials, the label for Vivitrol warns against the increased risk of opioid overdose deaths based on spontaneous post- marketing adverse event reporting. Vivitrol has the most mature evidence base of any of the interventions of interest for this review. Differences observed between Vivitrol and buprenorphine/naloxone are due at least in part to differences in treatment intent and goals. Therefore, we considered the evidence on Vivitrol in comparison to buprenorphine/naloxone to have high certainty of a comparable net health benefit (C). ©Institute for Clinical and Economic Review, 2018 Page 50 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents 4. Long-Term Cost Effectiveness 4.1 Overview The primary aim of this analysis was to estimate the lifetime cost-effectiveness of certain drugs used for MAT among a cohort of patients who were considered for OUD treatment, from a US health care sector perspective. A decision-analytic approach was employed. The model compared buprenorphine extended-release subcutaneous injections (CAM2038 [investigational] and Sublocade), extended-release injectable naltrexone (Vivitrol), and buprenorphine subdermal implant (Probuphine), to a transmucosal buprenorphine/naloxone, specifically generic sublingual (SL) buprenorphine/naloxone. We decided against a “no treatment” comparator given the target population of interest, as well as the availability of a common active comparator. We also decided to move our comparison of Sublocade vs. generic SL buprenorphine/naloxone from the base case to a scenario analysis, as further review and public comments we received led us to conclude that the inputs to the NMA were insufficient to support such an indirect treatment comparison. Key model outcomes, namely, quality-adjusted survival and health care costs were summed over a five-year time horizon for each treatment option. We deviated from the ICER Reference Case life-time horizon because of relatively high rates of treatment discontinuation and restart in the MAT environment. While previous models have employed even shorter time horizons,109,110 we used a five-year horizon to help capture potential downstream effects of MATs. Costs and outcomes were discounted at 3% per year. Incremental outcomes and costs were calculated comparing each intervention to SL buprenorphine/naloxone. The model was developed in hēRo3℠ with some components of the model, such as survival distributions, developed in RStudio (version 1.1.442). hēRo3 is a Web-based, health economic modeling platform that supports the development of both Markov cohort and partitioned survival models (Policy Analysis Inc., Brookline, MA). Calculations in hēRo3 are performed in the programming language, R, using an open-source health-economics modeling package, called “heRomod” (https://github.com/PolicyAnalysisInc/heRomod), that runs on a cloud-based platform. heRomod is a modified version of the open-source, health-economics modeling package, HEEMOD (http://cran.r-project.org/package=heemod). An extensive set of unit tests is available to validate calculations of the modeling package. Further details on hēRo3 activities and functions are available in Appendix E. ©Institute for Clinical and Economic Review, 2018 Page 51 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents 4.2 Methods Model Structure We developed a de novo decision analytic model for this evaluation, informed by key clinical trials and prior relevant economic models. The base case analysis took a health care sector perspective and thus focused on direct medical care costs only. Costs and outcomes were discounted at 3% per year. The analytic framework for this assessment is depicted in Figure 4.1. Figure 4.1. Model Schematic 4.1A. Decision Tree BUP/NAL: buprenorphine/naloxone Patients in comparator arms of individual treatments enter the model in health states in the same manner as their respective interventions. Sublocade analysis included only as a scenario ©Institute for Clinical and Economic Review, 2018 Page 52 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents 4.1B. Markov Model The model focused on an intention-to-treat cohort of OUD patients attempting to initiate treatment with MAT at model entry. Model cycle length was set at one month (four weeks)”MAT , reflecting prior economic models evaluating MATs.109,110 We acknowledge that treatment duration with Vivitrol and Probuphine tend to be shorter compared to those with other MATs that are meant to be maintenance therapies. We hence modeled shorter time-horizons in scenario analyses, while keeping the base-case time horizon at five years. Initial treatment pathways differed for each intervention based on trial design and FDA label, and patients were assigned initial state probabilities accordingly (Figure 4.1A). For CAM2038, patients who tolerate an initial dose of 4 mg buprenorphine/naloxone initiate therapy on day one of the model with a maximum dose of 8 mg generic SL buprenorphine/naloxone, followed by trial-based doses of CAM2038 injections.111 Sublocade patients undergo a “run-in” phase during which they are stabilized on 8-24 mg per day of transmucosal buprenorphine product for 11 days, aligning with the mean run-in phase in the Sublocade trial.112 For Vivitrol, patients undergo a detoxification period (completely opioid-free) for seven days prior to initiating treatment with Vivitrol, aligning with the FDA label and one of the three detoxification regimens in the key trial.113,114 For Probuphine, patients were required to be “clinically stable” for at least three months on ≤8 mg per day of a buprenorphine-containing product prior to Probuphine implant insertion, as seen in its FDA label and key trial.115,116 Details on initial state probabilities based on success/failure of the pre- MAT treatment rules are presented below in Table 4.1. ©Institute for Clinical and Economic Review, 2018 Page 53 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Table 4.1. Intention to Treat Analysis: Intervention-Specific Initial Health State Probabilities, Based on Run-In/Detoxification/Stabilization Protocols from Key Clinical Trials MAT with NO OFF MAT with OFF MAT with MAT with Illicit Illicit Use of Illicit Use of NO Illicit Use Use of Opioids Opioids Opioids of Opioids CAM2038*111 0.997* 0 0.003* 0 Generic SL 0.997* 0 0.003* 0 Buprenorphine/Naloxone*111,117 Vivitrol†114 0 0.721 0.279 0 Generic SL 0 0.941 0.059 0 Buprenorphine/Naloxone†114 Probuphineǂ116 0 0.891ǂ 0.109ǂ 0 Generic Buprenorphine/Naloxone ǂ116 0 0.891ǂ 0.109 ǂ 0 *4 mg buprenorphine/naloxone testing for tolerance to buprenorphine product. Percentage tolerance calculated as per pre-randomization data across both treatment arms. † Seven-day opioid detoxification period. ǂ Three-month period of being ‘clinically stable’ on ≤8 mg per day of transmucosal buprenorphine-containing product. Dose chosen for the model during “clinically stable” phase was 8 mg per day and ‘clinically stable’ was assumed as abstinent from illicit use of opioids. Percentage “clinically stable” calculated as per pre-randomization data across both treatment arms. Note that patients do not need to be in complete opioid withdrawal when initiating treatment with CAM2038. In the CAM2038 arm, patients who tolerated the 4 mg buprenorphine/naloxone test dose started the model in the “MAT with Illicit Use of Opioids” health state while those who did not entered the model in the “OFF MAT with Illicit Use of Opioids” health state. The proportion of patients who successfully completed a “run-in” phase with Sublocade entered the model in the “MAT with Illicit Use of Opioids” health state, while those who fail this “run-in” phase entered the model in the “OFF MAT with Illicit Use of Opioids” health state. Patients transitioned to “MAT with NO Illicit Use of Opioids” from “MAT with Illicit Use of Opioids” health state as they abstained from illicit opioid use, defined as a negative urine sample for opioids plus self-reporting of no illicit use of opioids, or to “OFF MAT with Illicit Use of Opioids” health state due to MAT discontinuation. For CAM2038 and its comparator, the proportion of discontinuation of MATs from the “MAT with NO Illicit Use of Opioids” and “MAT with Illicit Use of Opioids” to the “OFF MAT” health states were assumed the same, and were based on the “illicit use of opioid” status at the time of discontinuation among patients in the Sublocade trial as this was the only data source from which we could parse out discontinuation based on illicit use status.118 Patients successful at detoxification prior to initiating Vivitrol treatment entered the model in the “MAT with NO Illicit Use of Opioids” health state, while those who did not complete the detoxification period entered the model in the “OFF MAT with Illicit Use of Opioids” health state. Similarly, for Probuphine, patients “clinically stable” at three months on ≤8 mg per day on a buprenorphine-containing product ©Institute for Clinical and Economic Review, 2018 Page 54 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents entered the model in the “MAT with NO Illicit Use of Opioids” health state, while those not “clinically stable” over the same period entered the model in the “OFF MAT with Illicit Use of Opioids” health state. Relapse in the Vivitrol and Probuphine trials is defined by a positive opioid urine sample. Upon relapse, patients enter the “OFF MAT with Illicit use of Opioids” health state. For Probuphine, upon relapse to illicit use, patients are modeled such that they enter the “MAT with Illicit Use of Opioids” health state if relapse occurs within six months of implant insertion, and once implant is removed after the six-month period, all relapsed patients transition to the “OFF MAT with Illicit Use of Opioids” health state. The implant can be removed only at 24 weeks following implant insertion, thus effectively rendering 0% discontinuation for the first six months when using the implant. We found no evidence on immediate removal of implant upon relapse to illicit use of opioids. As Vivitrol is an opioid antagonist and blocks other opioids from binding to opioid receptors, a relapse (failure of abstinence from illicit opioid use) is considered equivalent to MAT discontinuation, with patients transitioning from “MAT with NO Illicit Use of Opioids” to “OFF MAT with Illicit Use of Opioids.” Since patients on Vivitrol have not been taking an opioid or opioid agonist for a period of time (because they are taking an opioid antagonist), there is an increased sensitivity (decreased tolerance) to opioids, resulting in an increased risk of mortality from opioid overdose among patients relapsing to illicit use. However, we found no robust data on this increased risk for mortality, so did not attempt to model this. For those treated with Probuphine, although “clinically stable” could mean illicitly abusing at least low doses of opioids while on ≤8 mg per day of buprenorphine-containing product, our model considered “clinically stable” as abstinent from illicit opioid use. Patients not “clinically stable” were defined as those who did not meet inclusion criteria pre-randomization in the trial.116 The label also indicates use of Probuphine for no longer than six months in one arm, after which a new implant can be administered subdermally in the contralateral arm, but only if the new implant is administered immediately after the previous implant has been removed.115 However, treatment efficacy using a second set of implants has not been studied. Thus, for patients abstinent from illicit opioid use on Probuphine at the end of the six-month implant period, the next intervention in the treatment pathway was assumed to be generic SL buprenorphine/naloxone. Beyond the six-month duration of the implant, patients followed the same pathway as those who have been treated with generic SL buprenorphine/naloxone for nine months, depending on which health state they were in at the time of implant removal. The comparator treatment versus each intervention is SL buprenorphine/naloxone and it follows the same rules associated with initial state probabilities as the interventions when entering the Markov model. The comparator was attributed trial-specific efficacy and discontinuation. Comparator price was based on generic formulations of SL buprenorphine containing product. In all treatment arms, patients could discontinue MAT when not illicitly using opioids and could move to one of two health states: (1) “OFF MATs with NO illicit use of opioids,” occurring in an ©Institute for Clinical and Economic Review, 2018 Page 55 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents assumed 10% of all patients who remained in the “MAT with NO illicit use of opioids” health state for at least 12 months; or (2) “OFF MAT with illicit use of opioids,” among all other patients. Once in the “OFF MAT with NO illicit use of opioids” or “OFF MAT with illicit use of opioids” health state, we assumed that patients could not re-enter either the “MAT with illicit opioid use” or “MAT with NO illicit use of opioids” health states in the model (Figure 4.1B). Patients remained in the model until death. All patients could transition to death from all causes from any of the alive health states. In addition, patients could die from opioid overdose in health states where they illicitly use opioids. Target Population The populations of focus for this review generally included adults diagnosed with opioid use disorder seeking MAT. Base case population characteristics in the trials were reasonably similar in age and gender ratio. Trial populations varied mostly by type of OUD (prescription or injectables). We therefore used a weighted average from all key trials considered for this analysis to derive the percentage of patients with illicit prescription opioid use and injection drug use (Table 4.2).111,112,114,116,117 Table 4.2. Base Case Model Cohort Characteristics Percent Illicit Use of Prescription Opioids Mean Age Percent Female (vs. Injectable Drug Use) Baseline Characteristics 36 years 30% 50.7%* *Weighted average across interventions and comparators in the trials included for this analysis. Treatment Strategies The list of interventions was developed with input from patient organizations, clinicians, manufacturers, and payers on which MATs to include. All listed interventions were compared to generic SL buprenorphine/naloxone. Although SL buprenorphine/naloxone is the common comparator across all interventions, its efficacy and rate of treatment discontinuation varies by intervention due to the varied opioid use status of populations entering the key clinical trials considered for this analysis. The MAT interventions evaluated were: • Buprenorphine subcutaneous ER injection o CAM2038 (Investigational), Braeburn Pharmaceuticals o Sublocade, Indivior Pharmaceuticals (scenario analysis) • Injectable ER naltrexone – Vivitrol, Alkermes Pharmaceuticals • Buprenorphine ER subdermal implant – Probuphine, Titan Pharmaceuticals Key Model Choices and Assumptions Our model was informed by the key choices and assumptions listed in Table 4.3. ©Institute for Clinical and Economic Review, 2018 Page 56 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Table 4.3. Key Model Choices and Assumptions Model Assumption Rationale Patients continue receiving ancillary counseling Treatment with MAT is associated with ancillary services while on MAT, irrespective of whether they counseling services, based on stakeholder input. maintain abstinence or whether they relapse. Patients on MAT, upon relapse to illicit use of We found no robust published evidence on the illicit opioids, are assumed to return to the same opioid use of specific opioids by category in patients who have use (prescription or injectable) used pre-MAT. relapsed on MAT. Long-term discontinuation/relapse for all There exists no robust data on long-term interventions was assumed the same as seen in the discontinuation/relapse for all accessed interventions. trials, if using point estimates, or were extrapolated using the same parametric curve functions used to fit trial-specific data. We assumed that 10% of all patients who remained We found no published evidence indicating the in the “MAT with no illicit use of opioids” health percentage of MAT recipients remaining off opioids state for at least 12 months transitioned to an “OFF when they stop MAT. Given the frequency of relapse in MAT with NO illicit use of opioids” health state. this population, we assumed a relatively low rate of permanent abstinence and tested this rate in sensitivity analyses. The model assumed only a single cost and utility We found no published evidence of categories of associated with each health state and does not reduction in illicit use of opioids use while on or after categorize levels of reduction of illicit use of opioids. MAT. Opioid overdose-related mortality was assumed to There exists no robust published evidence on opioid occur only during periods of illicit use of opioids and overdose-related mortality by MAT type and was assumed the same whether on concurrent MAT concurrent illicit opioid use. or otherwise. Mortality from opioid overdose was held constant We found no robust published evidence on time- over time. dependent mortality from opioid-overdose among OUD patients. Incidence of Human Immunodeficiency Virus (HIV) A significant proportion of HIV and HCV cases among and Hepatitis C Virus (HCV) infections was only those who illicitly use opioids occur in PWID. We found attributed to people who inject drugs (PWID). no published evidence on HIV and HCV incidence among non-PWID illicit opioid users. The model assumed a constant disutility associated We found no robust evidence on time- and disease- with HIV infection and treatment with anti- status-dependent change in disutility among those retroviral therapy (ART). infected and diagnosed with HIV and treated with ART. ©Institute for Clinical and Economic Review, 2018 Page 57 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Model Assumption Rationale Among PWID diagnosed with HCV, the model Patients with spontaneous HCV infection clearance or assumed a constant disutility only for those for those without cirrhosis successfully treated with direct- whom there was no spontaneous clearance of HCV acting antiviral therapy are assumed to have no HCV- specific disutilities affecting their quality of life. infection and who fail treatment. Additional HCV-specific health care costs, as well as The proportion of individuals meeting these conditions HCV-specific mortality, were attributed only to would be expected to be quite small given the high cure those patients diagnosed with HCV who are without rates associated with current treatments. HCV infection clearance and not cured with treatment. HIV drug (anti-retroviral therapy) costs were We found evidence that not all HIV-diagnosed attributed to all PWID diagnosed with HIV infection, individuals enroll in supportive community-care based while 75% of these individuals were attributed costs programs.119 for HIV-specific community care-based programs. Serious adverse event (SAE)-related costs or The trials vary in reporting of SAEs, with most reporting disutilities were not included in the model. only percentage of SAEs and not specific non-relapse- related SAEs. Individual adverse events when reported were not reported by category of severity. We assume that background health care costs (sourced from a claims analysis) include costs associated with treating SAEs. Model Inputs Clinical Inputs Transition Probabilities Treatment Efficacy Transition probabilities related to treatment efficacy were derived from relevant trial data (Table 4.4).111,114,116,117 All trial efficacy estimates were converted to per-cycle transition probabilities and held constant throughout the modeled time horizon. Table 4.4. MAT Treatment Efficacy Abstinence from Illicit Use of Opioids at 24 Weeks* Intervention Comparator (SL Buprenorphine/Naloxone) CAM2038 117 34.2% 27.4% Vivitrol114 48%† 44%† Probuphine 116 80.5%† 67.4%† *Abstinence estimates over the 24-week trial duration, were converted to per cycle transition probabilities. † KM curves used for modeling relapse over time are based on per protocol observations and not intention-to-treat (ITT) observations since ITT approach has been taken into consideration in the decision tree prior to Markov model entry. ©Institute for Clinical and Economic Review, 2018 Page 58 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Treatment Discontinuation Treatment discontinuation to “OFF MAT with illicit use of opioids” could occur from both “MAT with no illicit use of opioids” and “MAT with illicit use of opioids” for CAM2038 and generic SL buprenorphine products. The proportion discontinuing from each of these states was derived from data in the Sublocade trial (calculated from academic-in-confidence data) and applied to the overall discontinuation rate reported for CAM2038 and generic SL buprenorphine products in the trials, since as stated earlier, this trial was the only source from which we could parse out MAT discontinuation based on illicit use status. Treatment discontinuation was estimated from the trial- reported Kaplan-Meier (KM) curves for discontinuation for CAM2038, Vivitrol, Probuphine, and their respective SL buprenorphine/naloxone comparators.111,114,116,117 Discontinuation rates for each MAT at the end of trial period is presented in Table 4.5. Note that this comparison was only made for discontinuation; as noted in Section 3, differences in study populations, outcome measures, and run-in protocols prevented a formal and comprehensive NMA. We fit parametric survival curves to KM data utilizing the approach described by Hoyle and Henley.120 First, we extracted data points from digitized copies of the trial curves, then used the extracted values, the number of remaining patients at each time interval, and maximum likelihood functions to estimate curve fits to the underlying individual patient data. The fitted model curves included the distributional forms of exponential, Weibull, exponential, log-normal, log-logistic, and gamma. The base-case parametric function was selected based on best model fit using Akaike information criterion (AIC) values and visual comparison (see Appendix Table E2). Beyond trial duration, discontinuation was extrapolated using the best-fitting curve function seen within the trial period. Table 4.5. MAT Treatment Discontinuation Discontinuation Comparator (SL Intervention Buprenorphine/Naloxone) CAM2038117 31%* 27.4%* Vivitrol†114 52% 54% Probuphine †116 0% ‡ 32.6% *KM curves used for modeling discontinuation over time. † KM curves used modeling treatment discontinuation are based on per protocol observations and not intention-to- treat (ITT) observations since ITT approach has been taken into consideration in the decision tree prior to Markov model entry. ‡ 0% discontinuation because Probuphine will be implanted for the duration of six months irrespective of abstinence or relapse to illicit use of opioids. Patients who remain abstinent at the time of Probuphine implant removal were assumed to have the same discontinuation rate as those treated with SL buprenorphine product. ©Institute for Clinical and Economic Review, 2018 Page 59 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Comorbidities Associated with OUD Key OUD-related comorbidities with significant public health impact include HCV and HIV infections among PWID. A cohort study and a meta-analysis based on four US-specific surveys on PWID reported annual incidence of HIV and HCV among PWID as 0.055% (95% Confidence Interval: 0.042% to 0.080%) and 26.7%, respectively.121,122 These rates were converted to per-cycle probabilities in the model. While endocarditis is also a potential adverse effect among PWID, it was not included in the model due to relatively low incidence and associated mortality, with available data being non-recent.123-125 Mortality Opioid overdose-related mortality was estimated from observational data, while all-cause gender- and age-specific mortality was sourced from the Human Mortality database’s US-specific tables.126,127 Increased risk of mortality associated with HIV and HCV was attributed to PWID (Table 4.6).128,129 Among PWID diagnosed with HCV, the increased mortality risk from HCV was applied only to those for whom there was no spontaneous clearance of HCV infection along with treatment failure.130,131 Table 4.6. Mortality Inputs Parameter Value Opioid-Related Overdose Death126 13.3 per 100,000 people* HR of Death from HIV128 3.15 (95% CI: 2.59 to 3.82)† MRR of Death from HCV129 2.37 (95% CI: 1.28 to 4.38)† All-Cause Mortality 127 US Life Tables All values were converted to per cycle transition probabilities. HR: hazard ratio, MRR: mortality rate ratio, CI: confidence interval *Among all illicit users of opioids. †Compared to PWID without infection. Utilities Health state utilities were derived from a study that used an online US cross-sectional survey.132 The study comprised hypothetical descriptive vignettes for OUD and associated MAT-related health states that were developed based on inputs from literature, clinical expert opinion, and people diagnosed with OUD. Quality of life assessments were undertaken using the standard gamble technique. For each health state, two sets of vignettes were developed, one including physical/emotional descriptors, and another “expanded” version adding societal to the physical/emotional descriptors (i.e., employment, criminal justice, and family relationship-specific aspects). The study excluded comorbidity-associated vignettes because its primary focus was assessing quality of life associated with OUD alone. Health state utilities when on MAT with concurrent use of illicit opioids were calculated by applying the ratio of utilities when illicitly using ©Institute for Clinical and Economic Review, 2018 Page 60 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents opioids with and without MAT (from a UK study133) to the base utility when illicitly using opioids when OFF MATs (from the cross-sectional survey) (Table 4.7). Health state utilities in the “OFF MAT with NO Illicit use of opioids” health state was sourced from a nationally representative survey study conducted in the US.134 Table 4.7. Health State Utilities Parameter Value (Range – 95% CI) MAT with NO Illicit Use of Opioids132 0.766 (0.738 – 0.795) Relapse – OFF MAT with Illicit Use of Opioids 0.694 (0.660 – 0.727) (Prescription)132 Relapse – OFF MAT with Illicit Use of Opioids (IDU)132 0.574 (0.538 – 0.611) MAT with Illicit Use of Opioids (Prescription) 133 0.700* (0.660† - 0.727ǂ)§ MAT with Illicit Use of Opioids (IDU)133 0.618* (0.538† – 0.727ǂ)§ OFF MAT with NO Illicit Use of Opioids134 0.852ǂ(0.736 – 0.901)§ HIV Disutility Multiplier135 6.9% (1% - 19.5%)§ HCV Disutility Multiplier 136 7% (1% - 16%)§ IDU: injection drug use, CI: confidence interval *Based on utilities reported by Connock et al., 2007, the ratio of utilities in health states with illicit use of prescription opioids while ON and OFF MATs is approximately 1.01, while the same with illicit use of injectable opioids is approximately 1.07. These ratios were applied to the “relapse” OFF MAT illicit use of opioids health state utilities to derive utilities for prescription and injection-related health states of “MAT with Use of Illicit Opioids.” † Same lower bound as when “OFF MAT with Illicit Use of Opioids;” same upper bound as when OFF MAT with illicit use of prescription opioids. ǂ Calculated as age-range specific population-weighted mean starting in the 30-39 years age range. § Calculated ranges are not 95% Cis. For PWID diagnosed with HIV, we applied a 6.9% disutility to their baseline health state utilities. This estimate was derived from an economic evaluation that assessed the cost-effectiveness of HIV prevention programs among PWID in the U.S.135 The model sourced baseline quality of life estimates for PWID and HIV stage- and treatment-specific multipliers from published literature. We applied multipliers specific to anti-retroviral therapy (ART) and symptomatic HIV to arrive at a 6.9% reduction from baseline utility among PWID diagnosed with HIV and treated with ART. Detailed calculations are available in Appendix Table E3. For PWID diagnosed with HCV, we applied a 7% disutility to their baseline health state utilities. This disutility was derived from estimates used in a U.S.-specific cost-effectiveness model assessing anti- HCV treatments in patients diagnosed with HCV.136 The applied disutility was held constant over time and attributed only to PWID for whom there was no spontaneous clearance of HCV infection or failure of anti-viral treatment. ©Institute for Clinical and Economic Review, 2018 Page 61 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Adverse Events The trials vary in reporting of SAEs, with most reporting only percentages of SAEs and not specific non-relapse-related SAEs. Individual adverse events, when reported, were not reported by category of severity. For these reasons, and because separate costing of SAEs was not expected to affect model results in a material fashion, we did not attempt to estimate SAE costs for any treatment of interest. We did, however, use background health care costs from a claims analysis by Shah et al. that included costs associated with treating SAEs. We found no evidence on disutility associated with serious adverse events in this population, so no impact on utility from SAEs was assumed. Economic Inputs The model included all treatment costs associated with each individual regimen, including drug acquisition costs, drug administration costs, and supportive care costs (e.g. clinician visits, counseling, and monitoring). Drug Acquisition Costs We found no estimates on net price from SSR Health for the currently approved interventions. In the absence of SSR net price data, we used net price as reported in the Federal Supply Schedule (FSS) for all interventions except Vivitrol.137 For Vivitrol, we used the net price provided to us by the manufacturer, which was derived from IQVIA estimates.138 There is no listed price available for CAM2038, as the drug is currently under review, so we calculated only threshold prices (i.e., prices that would achieve certain cost-effectiveness thresholds) for this MAT. For generic SL buprenorphine/naloxone, we used the average of generic Wholesale Acquisition Costs (WAC). Based on the regimen dosage specified earlier in the clinical evidence review section (Table 1.2), the model utilized the lowest-cost combination of vials, tablets, or implants for each regimen. All MAT drug costs are listed in Table 4.8. ©Institute for Clinical and Economic Review, 2018 Page 62 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Table 4.8. Drug Cost Inputs Net price per Net Price Discount Annual Net Intervention WAC per Dose*139 Dose137 from WAC Price CAM2038 24/96 mg -- -- -- -- Sublocade 300 mg‡ $1,580 $1,206.83† 24% $15,688.79 Vivitrol 380 mg $1,309 $759.25§ 42% $9,870.25 Probuphine 296.8 mg $4,950 $3640.32† 26% $3,640.32# Generic SL ER Buprenorphine/Naloxone $8.32 -- -- $3,037.46 16 mg Generic ER Oral $4.39 -- -- $1,603.02 Buprenorphine 8 mg¤ WAC: wholesale acquisition cost, FSS: Federal Supply Schedule *WAC as of October 17, 2018. † FSS price as published on October 1, 2018 ‡ Included only in a scenario analysis § Manufacturer-provided net price # One-time cost; does not include MAT cost following implant removal ¤ For clinical stabilization period for Probuphine Administration and Monitoring Costs We included costs of administering CAM2038 and Vivitrol, once per cycle (Table 4.9).140 We also included administration costs associated with insertion and removal of Probuphine, as Probuphine- associated costs were not available in the background cost publication (Table 4.9).140,141 Table 4.9. Administration Costs (National Average Non-Facility Price) Parameter Value Probuphine Implant Insertion (CPT® Code: 11981)140 $145.90 Probuphine Implant Removal (CPT® Code: 11982)140 $163.08 SC/IM Injection Administration (CPT® Code: 96372)140 $20.88 Health Care Utilization Costs Non-MAT (non-drug) background health care costs have been sourced from a retrospective cohort study using claims data from the Truven Health Analytics MarketScan® database.141 This analysis reports baseline and follow-up costs specific to treatment with Vivitrol, methadone, buprenorphine, and non-pharmacological therapy in patients diagnosed with OUD. Patients were followed up for one year and costs included those associated with inpatient admissions, emergency department (ED) visits, outpatient visits, and pharmacy costs. We calculated the population-weighted average costs of inpatient, ED, and outpatient visits among the Vivitrol and buprenorphine treated populations at baseline and follow-up, and attributed these costs to the “OFF MAT with Illicit Use of Opioids,” “ON MAT with Illicit Use of Opioids,” and the “MAT with NO Illicit Use of Opioids.” We ©Institute for Clinical and Economic Review, 2018 Page 63 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents also excluded pharmacy costs, to avoid double-counting of costs of MAT. When illicitly using opioids while on MATs and when using MAT while not illicitly using opioids, we added MAT drug- specific (except CAM2038) and associated administration costs. While there was a decrease in inpatient and ED costs between follow-up and baseline, there was an increase in outpatient costs, which we believe is attributed to patients making more frequent physician office visits owing to treatment with MATs. We assigned health care costs to the “OFF MAT with NO illicit use of opioids” health state based on health care costs for the general population without OUD. These costs were sourced from the Health Care Cost Institute’s 2016 report, which described costs based on claims analyses in the population under 65 years old with employer-sponsored insurance.142 Components of cost included inpatient, professional, outpatient and prescription drugs. All costs shown in Table 4.10 are per-cycle costs. Table 4.10. Health Care Costs per Cycle ON or OFF MAT with MAT with No Illicit OFF MAT with NO Illicit Illicit Use of Opioids141 Use of Opioids141 Use of Opioids142 Inpatient Admissions $385.08 $332.94 -- Emergency Department Visits $81.01 $70.97 -- Outpatient Visits $480.78 $727.98 -- All Health Care Costs -- -- $427.84 All costs calculated and presented as per-cycle costs, using annual costs reported in source publications. For PWID diagnosed with HIV or HCV, we attributed drug and other non-drug costs associated with these comorbidities.119,143 For individuals with PWID and HIV, we attributed only 75% of costs associated with HIV-related community care programs, to reflect the proportion of those diagnosed with HIV who participate in such programs.119 Spontaneous clearance of HCV infection has been reported in 24.4% (95% Confidence Interval: 19.5% to 29.1%) of HCV-diagnosed PWID, based on a meta-analysis of 28 reports, of which seven were US-specific.131 Among those for whom no spontaneous HCV infection clearance occurs, treatment with glecaprevir/pibrentasvir (Mavyret™, AbbVie, Inc.), a pan-genotypic eight-week treatment for HCV, was initiated. Ongoing HCV-related health care costs were attributed only to those who failed initial treatment with glecaprevir/pibrentasvir. Costs of glecaprevir/pibrentasvir were estimated from the FSS.137 Estimates for treatment success with glecaprevir/pibrentasvir were sourced from treatment efficacy trial data presented in the drugs’ prescribing label.130 Appropriate HIV- and HCV-related costs were also attributed to those PWID in the “OFF MAT with NO illicit use of opioids” health state. All HIV and HCV-related costs (per cycle) are presented in Table 4.11. All costs were inflated to 2018 levels using the health care component of the personal consumption expenditure index,144 in accordance with the ICER Reference Case. ©Institute for Clinical and Economic Review, 2018 Page 64 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Table 4.11. HIV and HCV Treatment Costs per Cycle HIV119 HCV130,137,143 Drug Costs $1,865.04 $19,389.08* Other Treatment Costs $396.22† $849.22‡ HIV: human immunodeficiency virus, HCV: hepatitis C virus HCV drug cost for patients achieving a cure is assumed to be that of glecaprevir 100 mg/pibrentasvir 40 mg (Mavyret) for eight weeks, assuming all new cases of diagnosed HCV patients have no liver cirrhosis. *FSS net price per four weeks. †Assuming only 75% of diagnosed individuals attend HIV-specific community care programs. ‡Calculated as additional cost for HCV care relative to cost of non-HCV, non-MAT health care costs in PWID diagnosed with HCV and treatment with Mavyret failed. Societal Costs We also included costs associated with lost productivity, criminal justice, and incarceration in a scenario analysis that took a modified societal perspective. For lost productivity, based on trial population baseline characteristics, we estimated that 34% of the population diagnosed with OUD were employed.111,114,116-118 Birnbaum et al. reported productivity costs which included lost wages, excess disability, medically-related absenteeism, lost wages from incarceration, and presenteeism associated with opioid abuse, dependence, and misuse in the US.145 Combining these estimates with SAMHSA data146, we calculated the productivity loss costs per person (Table 4.11). We then applied these costs to approximately 34% of the modeled cohort while in health states that included illicit use of opioids. The costs of criminal justice and incarceration were sourced from a retrospective cohort study that included data from the California Outcomes Monitoring System, Automated Criminal History System, Offender Based Information System, and National Death Index.147 Patients included in the study were those diagnosed with OUD with uniquely identifiable criminal justice records. Criminal justice and incarceration costs comprised costs of policing, court, corrections, and medical expenses, cash losses, property theft, and consequences related to criminal victimization. Approximately 43% of the entire sample was involved in criminal justice and incarceration-related events; we hence applied these costs to the same percentage within our cohort. This study reported daily costs of criminal justice and incarceration when on opioid agonist therapy and “post- treatment,” which in our model referred to costs when on MAT (with and without illicit use of opioids) and off MAT (only with illicit use of opioids), respectively (Table 4.12). Details of our calculation are available in Appendix Tables E4 and E5. ©Institute for Clinical and Economic Review, 2018 Page 65 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Table 4.12. Societal Costs per Cycle Value Productivity Loss145,146 $1,334.26* Criminal Justice and Incarceration147 When ON MAT (With and Without Illicit Use of Opioids) $1,089.23† When OFF MAT (Only with Illicit Use of Opioids) $5,446.13† *Applied to only 34.42% of cohort. †Applied to only 43.24% of cohort. Sensitivity Analyses We conducted one-way sensitivity analyses to identify the impact of parameter uncertainty and key drivers of model outcomes. Inputs for one-way sensitivity analyses are presented in Appendix Tables E6 to E12. Probabilistic sensitivity analyses were also performed by jointly varying all model parameters over 1,000 simulations. Details of distributions used for the probabilistic analyses can be found in Appendix Tables E13 to E19. Additionally, we performed a threshold analysis across incremental cost-effectiveness thresholds ranging from $50,000 to $150,000 per QALY gained. Scenario Analyses For the final report, we moved the Sublocade vs. SL buprenorphine/naloxone comparison from the base case analysis as presented in our draft report, to scenario analyses. First, we conducted a threshold analysis calculating for Sublocade’s per dose price to reach cost-effectiveness thresholds between $50,000 and $150,000 per QALY. Second, we conducted another threshold analysis calculating for efficacy (abstinence from illicit use of opioids), to reach cost-effectiveness thresholds of up to $150,000 per QALY. Finally, we conducted a third scenario analyses that assumed Sublocade to have the same efficacy and discontinuation rate as CAM2038 relative to its comparator, as seen in the CAM2038 key trial. We used Sublocade’s FSS price (in the second and third scenario analyses) and induction regimen but favored Sublocade assuming 100% successful induction (“run in”) with Sublocade and 99.7% successful induction with its comparator SL buprenorphine/naloxone, again as seen in the CAM2038 key trial. Please refer to the model methods sub-section for Sublocade’s “run in” protocol. For all other interventions, multiple scenario analyses were conducted to evaluate the impact of key model choices and assumptions on the robustness of the results and conclusions. • We included a modified societal perspective that included the costs associated with productivity loss and criminal justice and incarceration among patients who illicitly use opioids. • We modeled shorter time-horizons of one and two years. As stated in the model structure sub-section, we acknowledge that treatment with the buprenorphine products (except Probuphine) is meant to be long-term, while treatment with Vivitrol or Probuphine is often ©Institute for Clinical and Economic Review, 2018 Page 66 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents intended to be for a shorter time horizon, up to one year, based on feedback received from stakeholders. • We varied population characteristics such that the entire cohort entering the model were either illicit users of prescription opioids or were PWID. • We conducted an analysis with a modified model structure that excluded the “OFF MAT with NO Illicit Use of Opioids” health state, to model a scenario in which patients cannot permanently abstain from illicitly using opioids. • We included an analysis that followed a “per protocol” approach and not an “intention-to- treat” approach wherein all patients entered the model in the “MAT with Illicit Use of Opioids” for CAM2038, Sublocade, and their respective comparators, or in the “MAT with NO Illicit Use of Opioids” for Vivitrol, Probuphine, and their respective comparators. • We modeled a second implant for Probuphine following the first implant to see its effect on longer-term outcomes. Model Validation Model validation followed standard practices in the field. 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. We then tested all mathematical functions in the model to ensure they were consistent with the report (and supplemental Appendix materials). Independent modelers also tested the mathematical functions in the model as well as the therapy- specific inputs and corresponding outputs. We also conducted sensitivity analyses with null input values to ensure the model produced findings consistent with expectations. Finally, we compared the ICER model to previously published models. We searched the literature to identify models that were similar to our own, with comparable populations, settings, perspective, and treatments. 4.3 Results Base-Case Results In the comparison of CAM2038 versus generic SL buprenorphine/naloxone, CAM2038 produced marginally higher QALYs (3.26 vs. 3.20) and similar life years (4.62), as seen in Table 4.13. Note that in all results below life years are different only at the third or subsequent decimal places. CAM2038’s higher rate of abstinence was offset by its higher rate of discontinuation relative to its comparator (as seen in the trial data), which led to a marginally higher QALY gain. Since there exists no list or net price for CAM2038, we could not calculate its incremental cost-effectiveness ratio. ©Institute for Clinical and Economic Review, 2018 Page 67 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Table 4.13. Base Case Results for CAM2038 versus Generic SL Buprenorphine/Naloxone Treatment MAT Drug Costs Other Costs Total Cost Life Years QALYs CAM2038* -- $66,100 -- 4.62 3.26 Generic SL Buprenorphine/Naloxone $5,400 $64,700 $70,100 4.62 3.20 QALY: quality-adjusted life year *No List or net price available yet for CAM2038. In the comparison of Vivitrol versus generic SL buprenorphine/naloxone, Vivitrol produced marginally fewer QALYs (3.25 vs. 3.28) and similar life years (4.62), as seen in Table 4.15. Drug costs with Vivitrol are higher relative to its comparator, while nondrug costs between the two treatments are similar. Thus, with higher costs and lower effectiveness, Vivitrol is dominated by generic SL buprenorphine/naloxone (Table 4.14). Table 4.14. Base-Case Results for Vivitrol versus Generic SL Buprenorphine/Naloxone Incremental MAT Drug Other Treatment Total Costs Life Years QALYs Cost per QALY Costs Costs Gained More costly, Vivitrol $15,900 $65,500 $81,500 4.62 3.25 less effective Generic SL $5,900 $65,200 $71,200 4.62 3.28 -- Buprenorphine/Naloxone QALY: quality-adjusted life year In the comparison of Probuphine versus generic SL buprenorphine/naloxone, Probuphine produced QALYs that were essentially identical (3.38 vs. 3.37) and similar life years (4.62), as seen in Table 4.15. Drug costs are higher with Probuphine, by approximately $2,300, while non-drug costs are higher by approximately $400, resulting in higher total cost with Probuphine. Probuphine’s cost- effectiveness ratio relative to its comparator is approximately $265,000 per QALY gained; as previously noted, however, only one implant was assumed for base case analyses, consistent with the clinical trial design, even though the FDA label allows for a second implant. Table 4.15. Base-Case Results for Probuphine versus Generic SL Buprenorphine/Naloxone Incremental MAT Drug Other Cost per Treatment Total Costs Life Years QALYs Costs Costs QALY Gained Probuphine $11,000 $66,900 $77,900 4.62 3.38 $265,000 Generic SL $8,600 $66,500 $75,100 4.62 3.37 -- Buprenorphine/Naloxone QALY: quality-adjusted life year ©Institute for Clinical and Economic Review, 2018 Page 68 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Sensitivity Analysis Results One-way sensitivity analyses showed that results were most sensitive to intervention discontinuation rate (relapse to illicit use of opioids), the incidence of HCV, and intervention costs for Vivitrol and Probuphine. Since CAM2038 currently has no price, we do not present tornado diagrams specific to CAM2038’s incremental cost-effectiveness ratios. Intervention-specific tornado diagrams are presented in Appendix Figures E1 and E2. Results of the probabilistic analyses showed that none of the interventions reached cost- effectiveness thresholds of $50,000 per QALY gained over 1,000 simulations. Zero percent of simulations reached the $100,000 or $150,000 per QALY gained threshold for Vivitrol, while 1% and 12.8% of all simulations for Probuphine achieved the $100,000 and $150,000 per QALY gained threshold (Table 4.16). Table 4.16. Probabilistic Sensitivity Analyses Percentage of 1,000 Simulations at or Below Willingness-To-Pay Thresholds $50,000 per QALY $100,000 per QALY $150,000 per QALY Vivitrol 0% 0% 0% Probuphine 0% 1% 12.8% QALY: quality-adjusted life year We also present the percentage of simulations where MAT interventions are more costly and more effective relative to their respective comparators (Table 4.17). Again, since CAM2038 currently has no published price, we do not present probabilistic results on its incremental cost-effectiveness. Hexbins and cost-effectiveness acceptability curves for each intervention are presented in Appendix Tables E3 to E6. Table 4.17. Cost-Effectiveness Plane Quadrant Vivitrol Probuphine Northeast (More Costly and More Effective) 1.2% 76.8% Northwest (More Costly and Less Effective) 98.8% 23.2% Southwest (Less Costly and Less Effective) 0% 0% Southeast (Less Costly and More Effective) 0% 0% Scenario Analyses Results Sublocade versus SL Buprenorphine/Naloxone As explained in the methods section, in the first scenario we conducted a threshold analysis calculating the price of Sublocade that would reach cost-effectiveness thresholds between $50,000 and $150,000 per QALY. This resulted in Sublocade monthly (four week) prices nearly the same as those of CAM2038 in its threshold analysis (Table 4.20). The second scenario was a threshold ©Institute for Clinical and Economic Review, 2018 Page 69 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents analysis to calculate the efficacy (abstinence from illicit use of opioids) required to reach cost- effectiveness thresholds of up to $150,000 per QALY. This analysis showed that even if Sublocade use resulted in 100% adherence (0% discontinuation) and abstinence, its incremental cost- effectiveness ratio relative to SL buprenorphine/naloxone would still be well above the $150,000 per QALY threshold, at approximately $215,000 per QALY. In the third scenario analysis we assumed Sublocade and its comparator would have the same efficacy (abstinence from illicit use of opioids) and treatment adherence as CAM2038 and its comparator. For MAT costs, we use Sublocade’s and its comparators’ dose costs according to approved regimens. All other aspects of this scenario are similar to the base case model. As with CAM2038, the proportion of patients who successfully completed a “run-in” phase with Sublocade entered the model in the “MAT with Illicit Use of Opioids” health state, while those who fail this “run-in” phase entered the model in the “OFF MAT with Illicit Use of Opioids” health state (Figure 4.1A). This analysis was favorable to Sublocade since it assumed 100% “run-in” success for the Sublocade arm, and 99.7% “run-in” success for its SL buprenorphine/naloxone comparator as seen in the CAM2038 trial. Under these assumptions, Sublocade’s cost-effectiveness relative to generic SL buprenorphine/naloxone was estimated at approximately $577,000 per QALY over five years. Modified Societal Perspective In the modified societal perspective, total QALYs did not differ as we did not attribute additional disutilities that may be associated with productivity loss or criminal justice and incarceration; however, total costs in each treatment arm were greater. Intervention-specific costs associated with lost productivity and criminal justice and incarceration are presented in Appendix Tables E20 to E22. Including societal costs increased the total costs across all interventions and their comparators, but did not change the base case findings, specifically for Vivitrol which was more costly and less effective than its comparator. For Probuphine, however, including societal costs led to Probuphine being the dominant strategy, as slightly less costly and slightly more effective than generic SL buprenorphine/naloxone (Tables 4.18 and 4.19). However, the generalizability of the findings for Probuphine are limited, as the population in the trial (i.e., clinically stable on SL buprenorphine products for six months) is quite different from the eligible population in actual practice (i.e., patients diagnosed with OUD seeking MAT). Table 4.18. Vivitrol versus Generic SL Buprenorphine/Naloxone Treatment Total Costs QALYs Incremental Cost per QALY gained Vivitrol $200,000 3.25 More costly, less effective Generic SL $178,000 3.28 -- Buprenorphine/Naloxone QALY: quality-adjusted life year ©Institute for Clinical and Economic Review, 2018 Page 70 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Table 4.19. Probuphine versus Generic SL Buprenorphine/Naloxone Treatment Total Costs QALYs Incremental Cost per QALY gained Probuphine $155,000 3.38 Less costly, more effective Generic SL $156,000 3.37 -- Buprenorphine/Naloxone QALY: quality-adjusted life year Other Scenario Analyses Shorter time horizons resulted in results directionally similar to those observed in the base case analyses, with CAM2038 producing higher QALYs relative to its comparator, Vivitrol being dominated by its comparators, and Probuphine producing incremental cost-effectiveness ratios well above willingness-to-pay (WTP) thresholds of $50,000 to $150,000 per QALY gained. Detailed results of this scenario analysis can be found in Appendix Tables E23 and E24. Conducting analyses in a cohort comprising only PWID diagnosed and seeking MAT for OUD (i.e., no persons illicitly using prescription opioids) resulted in interventions and comparators with fewer QALYs and higher costs for all MATs, compared to those in the base case analyses. Detailed results of this scenario analysis can be found in Appendix Tables E25 to E27. Varying the model structure to exclude the “permanent abstinence from illicit use of opioids” health state resulted in marginally lower health outcomes (QALYs) and higher costs for all MATs. This is because all patients discontinuing MAT move to the “OFF MAT with Illicit Use of Opioids” health state, which involves lower utilities and higher costs than the “OFF MAT with no illicit use” health state. Detailed results of this scenario analysis can be found in Appendix Tables E28 to E30. We employed a “per protocol” approach, allowing for all patients in the CAM2038 and its comparator arms to enter the model in the “MAT with Illicit Use of Opioids” health state, and the Vivitrol, Probuphine, and relevant generic SL buprenorphine/naloxone comparator arms to enter the model in the “MAT with NO Illicit Use of Opioids” health state. In this scenario, we found that results were similar to the base case analysis, except in the case of Vivitrol, which resulted in an incremental cost effectiveness ratio of slightly more than $1 million per QALY gained relative to its comparator. Detailed results of this scenario analyses can be found in Appendix Tables E31 to E33. Aligning with the prescribing label for Probuphine, we modeled a scenario where a second implant is inserted immediately after removal of the first implant. Since there is no trial-related efficacy on the extended use of Probuphine, we assumed that the efficacy during the first six months extended through the subsequent six-month period. Relative to the base case, extended use of Probuphine for an additional six months resulted in increased total costs of approximately $3,200 and QALYs of 0.015 over five years, resulting in a lower incremental cost-effectiveness ratio of approximately ©Institute for Clinical and Economic Review, 2018 Page 71 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents $236,000 per QALY gained. Detailed results of this scenario analysis can be found in Appendix Table E34. Threshold Analyses Results Prices required to achieve willingness-to-pay thresholds of $50,000, $100,000, and $150,000 per QALY for CAM2038 and Probuphine are presented below in Table 4.20. Vivitrol showed inferior effectiveness but at higher costs relative to its respective comparator (generic SL buprenorphine/naloxone) in the base-case analyses. Therefore, we did not calculate threshold prices for Vivitrol but recommend that its price per unit be no more expensive than generic SL buprenorphine/naloxone. Table 4.20. Threshold Analysis Results WAC per Net Price Unit Price to Achieve Unit Price to Achieve Unit Price to Achieve Unit per Unit $50,000 per QALY $100,000 per QALY $150,000 per QALY CAM2038* -- -- $219 † $313† $406† Probuphine $4,950 ‡ $3,640 ‡ $1,165 ‡ $1,741‡ $2,318‡ QALY: quality-adjusted life year *No list or net prices for CAM2038 were available as of the date of this report. † Price per four-week dose. ‡ Price per implant lasting six months. Model Validation Model validation followed standard practices in the field. We tested all mathematical functions in the model to ensure they were consistent with the report (and supplemental Appendix materials). We also conducted sensitivity analyses with null input values to ensure the model was producing findings consistent with expectations. Further, independent modelers tested the mathematical functions in the model as well as the specific inputs and corresponding outputs. Prior Economic Models We searched the literature to identify models that were similar to our analysis, with comparable populations, settings, perspective, and treatments. Carter, Dammerman, and Frost recently examined the cost-effectiveness of subdermal implantable buprenorphine (Probuphine) versus sublingual buprenorphine for OUD treatment, using a US societal perspective that included both direct medical costs and non-medical costs such as lost productivity and criminal justice costs.110 Their analysis used a shorter time horizon (12 months) and allowed for a second implant after the first six months. Their model did not include a health state for off treatment without relapse, but was otherwise similar to those from Jackson et al. and Schackman et al.109,148 They estimated that Probuphine treatment would lead to a slight increase in QALYs gained (0.031) and higher drug costs but lower overall costs (by approximately -$4,400), largely due to decreases in ED/hospitalization ©Institute for Clinical and Economic Review, 2018 Page 72 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents and criminal justice costs. A key difference between this model and the current ICER analysis is that all patients were assumed to start “On treatment, not relapsed,” whereas the ICER analysis assumed that 10.9% of patients started in the “Off MAT with Illicit Use of Opioids” health state, to represent those patients who were not “clinically stable” for at least three months on ≤8 mg per day of a buprenorphine-containing product prior to Probuphine implant insertion. Jackson et al. analyzed the cost-effectiveness of injectable extended-release naltrexone (Vivitrol) compared to methadone maintenance and buprenorphine maintenance treatments for OUD.148 They estimated the incremental cost per opioid-free day over a six-month time horizon, using a state health program perspective. They found that Vivitrol would cost approximately $72 per opioid-free day (2015 US$) compared to methadone maintenance treatment, while buprenorphine maintenance was dominated (i.e., more costly but less effective) by methadone maintenance. The analysis by Jackson et al. did not include quality of life estimates or calculate cost per QALY, precluding direct comparison with our model. In addition, the cohort in their model was assumed to be on treatment and did not seem to account for patients who did not complete the detoxification period required for Vivitrol treatment initiation. Schackman et al. examined cost-effectiveness over a two-year horizon of long-term buprenorphine/naloxone treatment compared to no treatment for OUD, from a health care system perspective.109 Their base case reported an incremental cost per QALY of $35,100 for buprenorphine/naloxone compared to no treatment, and only a slight change to $35,200 when using a five-year time horizon. They assumed an annual cost for buprenorphine/naloxone of approximately $4,700 (compared to approximately $3,000 per year in our analysis); no generic forms of the treatment were available at the time of their analysis. Estimated cost per QALY in their analysis would decrease to $23,000 if the price of buprenorphine/naloxone were reduced by 50% (2010 US$). Unlike our model, their model assumed a cohort of “clinically stable” OUD patients who had already completed six months of outpatient buprenorphine/naloxone treatment, and who entered the model as “In treatment off drugs.” As the authors point out, the inclusion of costs and outcomes for patients in the first six months of treatment (including those who do not become “clinically stable”) would likely lead to higher cost-effectiveness ratios. Other analyses have examined comparators outside the scope of the present analysis, such as diacetylmorphine versus methadone treatment (Nosyk 2012)149, or in different countries, such as the UK (Connock 2007)133. 4.4 Summary and Comment Results of our assessment of the long-term cost-effectiveness of MATs for OUD suggest that all therapies generated very similar life years, with only marginal differences in QALYs relative to their respective comparators. Vivitrol was a dominated strategy (being more costly and less effective) relative to generic SL buprenorphine/naloxone. Our analysis indicates that only CAM2038 and Probuphine produce incremental QALYs relative to generic SL buprenorphine/naloxone, and then only marginally. Due to a lack of data, we couldn’t calculate the cost-effectiveness of Sublocade in ©Institute for Clinical and Economic Review, 2018 Page 73 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents the base case analysis, but even under unreasonably favorable assumptions in a scenario analysis, Sublocade still wasn’t cost-effective. We recognize that the population pursuing Vivitrol for MAT may have different treatment intent and goals, given the need for complete opioid withdrawal. We tested this in a per-protocol analysis that assumed successful withdrawal at model entry. While in this scenario Vivitrol produced greater QALYs than comparator treatment, this gain came at a cost of over $1 million per QALY gained. We did not calculate an incremental cost-effectiveness ratio for CAM2038, given the lack of an available price. However, threshold analyses suggest that this agent should be priced between $219 and $406 per four-week dose to fall within commonly-cited ranges for cost-effectiveness. Probuphine’s incremental cost-effectiveness ratio was well above the $150,000 per QALY WTP threshold relative to its comparator, a conclusion that did not change across multiple scenario and sensitivity analyses from the healthcare sector perspective. Using a modified societal perspective showed directionally similar results compared to the base case analyses for all interventions except Probuphine, which under a modified societal perspective became the dominant strategy (as slightly less costly and slightly more effective) relative to its comparator. As described above, however, findings for Probuphine are reflective of the population in which this MAT was tested (i.e., clinically stable on SL buprenorphine products for six months), so the generalizability of these results to the broader MAT population is very limited. Key model drivers included treatment discontinuation rates, intervention costs, and the incidence of HCV infection among PWID. Probabilistic sensitivity analyses confirmed the robustness of our base case findings. Shorter time horizons showed directionally similar results compared with those seen in the base case for all interventions. Changing population characteristics to include only PWIDs resulted in higher costs and lower QALYs for all interventions relative to the base case analyses. Aligning with comments that OUD be considered a chronic disorder, our scenario that excluded permanent abstinence from illicit use of opioids resulted in poorer health outcomes and higher costs. Limitations Our model has several limitations. While we acknowledge that OUD is a relapsing condition with patients cycling through the same or different therapies multiple times, we did not model re-use of MATs once patients relapsed, or relapse to illicit use among those considered permanently abstained from illicit use of opioids, as inadequate data exist for these estimates. Additionally, among illicit users of opioids, treatment efficacy and discontinuation may depend on type of illicit use (prescription or injection) which we do not consider as we did not find estimates on these specific to individual MATs. We modeled the pre-Markov decision tree based on trial-reported estimates, which may differ in a real-world setting. Also, quality of life among illicit users may differ ©Institute for Clinical and Economic Review, 2018 Page 74 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents based on levels of illicit use, which our model does not consider due to lack of data on these levels. Our estimates of utility came from a study that utilized a direct method of elicitation, wherein participants (with or without OUD) were described hypothetical OUD-related health states and asked to rate them between 0 and 1. However, the health state vignettes were specific to buprenorphine-containing compounds and methadone and not to Vivitrol. We model health care costs based on those for a commercially-insured population, which may not be representative of the real-world OUD demographic. Additionally, we used a weighted average of health care costs across populations using different MATs to arrive at single, state-specific health care costs, as we wanted to model costs for a “typical” patient eligible for MAT. While our objective was to identify MATs with best value relative to current treatment practices, we could not compare all MATs to a comparator with a single efficacy estimate, and instead and had to rely on trial-specific comparator estimates due to differences in population characteristics and trial design. While we acknowledge that adherence to MATs differ due factors such as the route and frequency of administration, we have no robust and/or long-term data to account for this in our model. Finally, our model does not capture diversion and/or switching to other opioids while on specific MATs, due to a lack of robust data on these estimates. We acknowledge the importance of these issues. Conclusions In conclusion, the findings of our analysis suggest that CAM2038, Vivitrol and Probuphine result in only marginal changes in QALYs relative to generic SL buprenorphine/naloxone, but universally higher costs. The incremental cost-effectiveness of these therapies versus generic SL buprenorphine/naloxone therefore falls outside commonly-cited thresholds of $50,000 to $150,000 per QALY gained. Even with assumptions extremely favorable to Sublocade, its incremental cost- effectiveness versus generic SL buprenorphine/naloxone also falls outside these commonly-cited thresholds. ©Institute for Clinical and Economic Review, 2018 Page 75 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents 5. Potential Other Benefits and Contextual Considerations Our reviews seek to provide information on potential other benefits offered by the intervention to the individual patient, caregivers, delivery system, other patients, or 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 extended-release opioid agonists and antagonist MAT to transmucosal formulations of buprenorphine/naloxone. Table 5.1. Potential Other Benefits and Contextual Considerations 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 return to work and/or overall productivity. 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 buprenorphine/naloxone, there is significant uncertainty about the long-term risk of serious side effects of this intervention. Compared to buprenorphine/naloxone, 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. 5.1 Potential Other Benefits As stressed by several organizations representing patients with OUD, “treatment is not one-size- fits-all” and patients need to have access to different treatment options on their road to recovery. Extended-release formulations are important additional treatment options that could improve long ©Institute for Clinical and Economic Review, 2018 Page 76 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents term recovery by lowering the constraints of daily adherence to transmucosal buprenorphine formulations. Extended-release buprenorphine formulations are currently subjected to the limits of number of patients that health care providers can treat annually. These limits intend to control diversion of buprenorphine products that are taken without direct medical observation. As extended-release buprenorphine products are administered by health care professionals, the risk of diversion by the patient is extremely low compared to transmucosal buprenorphine products. Regulators could consider not requiring waivers for extended-release formulations thus increasing overall access to MAT. Additionally, administration by a health professional should also prevent accidental poisonings in children that currently occur with transmucosal products. For OUD patients who are subjected to a program with external monitoring with important consequences of adherence, such as healthcare professionals, pilots, probationers or parolees, the use of extended-release formulations may also significantly improve rates of retention.12 In correctional settings with their high prevalence of OUD, extended-release formulations offer the potential of decreasing diversion. It must be noted, however, that the risk of diversion of transmucosal buprenorphine products is, at least in part, related to the fact that inmates with OUD are entering withdrawal and buprenorphine is diverted for controlling withdrawal.20 Offering buprenorphine through extended-release formulations may diminish negative beliefs about opioid agonist therapy and improve general access to MAT for inmates. 5.2 Contextual Considerations OUD is considered a public health emergency4 with an epidemic of deaths that decrease the overall life expectancy in the US2,3 and impacts all parts of society: families, the health system, social services, the judiciary system, and the economy. For the affected person, OUD is a chronic disease that is often compared to other chronic diseases, such as diabetes, but that carries a stigma affecting self-esteem, social relations, and work.12 Providing access to extended-release medications, can contribute to diminish the consequences of the opioid epidemic. Compared to transmucosal formulations of buprenorphine/naloxone or to methadone, there is significant uncertainty about the magnitude or durability of the long-term benefits of extended- release formulations, given the 6-month duration of nearly all trials of these agents. In addition, Probuphine implants cannot be used for longer than 12 months according to the FDA label. For the other formulations, their duration of appropriate use is unknown and will only be better defined through clinical experience and long-term observational study. ©Institute for Clinical and Economic Review, 2018 Page 77 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents For antagonist therapy with Vivitrol, its action cannot be reversed, so it becomes impossible to use opioids for emergency pain management. Regional analgesia or non-opioid analgesics need to be used.22 ©Institute for Clinical and Economic Review, 2018 Page 78 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents 6. Value-Based Price Benchmarks ICER’s value-based price benchmarks are meant to showcase drug prices that are required to align with value, defined as a willingness-to-pay (WTP) price range of between $100,000 to $150,000 per QALY. In cases where prices fall outside the upper bound and sometimes within this range, we present value-based prices. We calculated value-based prices for CAM2038 and Probuphine (Table 6.1). Since Vivitrol was less effective relative to its comparator in the base case, and since we did not have adequate data to model Sublocade versus SL buprenorphine/naloxone in the base case analysis, we did not estimate their value-based prices. Table 6.1. Value-Based Benchmark Prices for CAM2038 and Probuphine Annual Price to Achieve Price to Achieve Discount from WAC Required to WAC $100,000 per QALY $150,000 per QALY Achieve Threshold Prices CAM2038* -- $4,082 † $5,301 † -- Probuphine $4,950 ‡ $1,741 ‡ $2,318 ‡ 53% to 65% QALY: quality-adjusted life year *No list or net prices for CAM2038 were available as of the date of this report. †Annual price. ‡Price per implant lasting six months. Probuphine implant cannot be used more than twice in the treatment for OUD for each patient ©Institute for Clinical and Economic Review, 2018 Page 79 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents 7. Potential Budget Impact 7.1 Overview We used the cost-effectiveness model to estimate the potential total budgetary impact of CAM2038 in the patients aged 18 years and above with OUD. We calculated budget impact using the prices to achieve willingness-to-pay (WTP) thresholds between $50,000 to $150,000 per QALY gained in our estimates of budget impact. Since CAM2038 hasn’t been approved for use yet, no WAC or net price exists for the drug and we hence could not calculate budget impact at these prices. We did not include Probuphine, Sublocade, or Vivitrol in our calculations given their presence in the US marketplace for one year or longer. 7.2 Methods We used results from the same model employed for the cost-effectiveness analyses to estimate total potential budget impact. Potential budget impact was defined as the total differential cost of using each new therapy 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 five-year time horizons. The five-year timeframe was of primary interest, 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: patients aged 18 years and above with OUD. To estimate the size of the potential candidate populations for treatment, we used the reported prevalence for the year 2015 as those diagnosed with OUD, and applied it to the 2015 adult population to derive a point estimate of prevalence.150 We then applied the estimated prevalence to the projected 2018 to 2022 adult population in the US to derive the average number of OUD patients each year over the five-year period. This resulted in a population size of approximately 1.5 million patients over five years, or approximately 312,000 patients each year. While not all patients diagnosed with OUD seek treatment with MAT and only providers with adequate addition treatment training can prescribe certain MATs, we do not have data on the former or data on the later-specific to CAM2038. We hence assumed that all patients diagnosed with OUD were eligible for treatment with CAM2038. ICER’s methods for estimating potential budget impact are described in detail elsewhere151 and have been recently updated. The intent of our revised approach to budgetary impact is to document the percentage of patients who could be treated at selected prices without crossing a budget impact threshold that is aligned with overall growth in the US economy. For 2018-19, the five-year annualized potential budget impact threshold that should trigger policy actions to manage access and affordability is calculated to total approximately $991 million per year for new drugs. To ©Institute for Clinical and Economic Review, 2018 Page 80 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents estimate potential budget impact, 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. In this analysis, we assumed that all patients diagnosed with OUD would be treated with CAM2038 in place of generic buprenorphine/naloxone. 7.3 Results Table 7.2 illustrates the per-patient budget impact calculations in more detail, based on the prices to reach $150,000, $100,000, and $50,000 per QALY gained WTP thresholds for CAM2038 ($5,301, $4,082, and $2,863 per year, respectively) compared to generic buprenorphine/naloxone. Table 7.2. Per-Patient Budget Impact Calculations Over a Five-Year Time Horizon Average Annual Per Patient Budget Impact $150,000/QALY $100,000/QALY $50,000/QALY CAM2038 $35,420 $33,883 $32,346 Generic $31,653 Buprenorphine/Naloxone Difference $3,768 $2,231 $694 QALY: quality-adjusted life year The average potential budgetary impact at the three cost-effectiveness threshold prices for the drug ranged from approximately $3,768 per patient using the annual price to achieve $150,000 per QALY to approximately $694 using the annual price to achieve a $50,000 per QALY cost- effectiveness threshold. The annual potential budgetary impact of treating the entire population over five years did not exceed the $991 million ICER budget impact threshold at the price ($2,863) to achieve $50,000 WTP, approaching approximately 87% of the threshold. However, as shown in Figure 7.1, only 24% and 40% of the entire population could be treated each year at the prices that would reach the $150,000 to $100,000 per QALY thresholds respectively, before the total budget exceeded the ICER annual budget impact threshold. ©Institute for Clinical and Economic Review, 2018 Page 81 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Figure 7.1. Potential Budget Impact Scenarios at Different Prices of CAM2038 to Treat Adults with OUD $6,000 $150,000 per QALY $5,000 $100,000 per QALY $4,000 Annual Price $3,000 $2,000 $1,000 $0 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Percentage of Patients Treated without Crossing BI Threshold Each Year 7.4 Access and Affordability In the absence of a list or net price for CAM2038 as of the date of publication of this report, its potential budget impact is uncertain. However, if it is priced similarly to Sublocade, the expectation would be that the budget impact of CAM2038 would be offset by lower use of Sublocade. ICER is not issuing an access and affordability alert at this time. However, as use of both agents may increase over time, health systems likely to be covering large numbers of patients with OUD may wish to pay close attention to the actual use and costs of extended release injectable buprenorphine whether administered as Sublocade or CAM2038. ©Institute for Clinical and Economic Review, 2018 Page 82 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents 8. Summary of the Votes and Considerations for Policy 8.1 About the New England CEPAC Process During New England CEPAC public meetings, the New England CEPAC 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 New England CEPAC 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 New England CEPAC during their deliberation, and help to shape recommendations on ways the evidence can apply to policy and practice. After the New England CEPAC votes, a policy roundtable discussion is held with the panelists, 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 November 8th meeting, the New England CEPAC discussed issues regarding the application of the available evidence to help patients, clinicians, and payers address important questions related to the use of Sublocade, CAM2038, Probuphine and Vivitrol and MAT. Following the evidence presentation and public comments (public comments from the meeting can be accessed at https://www.youtube.com/watch?v=vJoVYaApQU8&feature=youtu.be starting at minute 1:14:00. the New England CEPAC voted on key questions concerning the comparative clinical effectiveness, comparative value, and potential other benefits and contextual considerations related to Sublocade, CAM2038, Probuphine and Vivitrol. 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 New England CEPAC members during the voting process. ©Institute for Clinical and Economic Review, 2018 Page 83 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents In its deliberations and votes related to value, the New England CEPAC considers 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 8.1 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. New England 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 New England 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. Potential 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 potential 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 potential 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 potential 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 84 Final Evidence Report - MAT in Patients with OUD 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 Patient population for all questions: Patients 16 years or older with opioid use disorder, who are being considered for MAT. 1) Is the evidence adequate to demonstrate that the net health benefit of the buprenorphine subcutaneous extended-release injection Sublocade (Indivior) is superior to that provided by transmucosal formulations of buprenorphine/naloxone? Yes: 0 votes No: 13 votes Comments: The Council unanimously judged that there was inadequate evidence to distinguish the net health benefit between treatment with Sublocade and transmucosal formulations of buprenorphine/naloxone as there were no available head-to-head trial data and indirect comparisons were not possible. ©Institute for Clinical and Economic Review, 2018 Page 85 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents 2) Is the evidence adequate to demonstrate that the net health benefit of the buprenorphine subcutaneous extended-release injection CAM2038 (Braeburn) is superior to that provided by transmucosal formulations of buprenorphine/naloxone? Yes: 4 votes No: 9 votes Comments: A majority of the Council determined that the evidence was inadequate to demonstrate that the net health benefit of CAM2038 is superior to transmucosal formulations of buprenorphine/naloxone. One Council Member who voted in the affirmative highlighted the more favorable outcomes for CAM2038 on urine screening outcomes, stating that although the outcome is not meaningful to patients, it was a reasonable surrogate. 3) Is the evidence adequate to demonstrate that the net health benefit of the buprenorphine subcutaneous extended-release injection Probuphine (Titan Pharmaceuticals Inc.) is superior to that provided by transmucosal formulations of buprenorphine/naloxone? Yes: 1 vote No: 12 votes Comments: A majority of the Council judged the evidence to be inadequate to demonstrate that the net health benefit of Probuphine is superior to transmucosal formulations of buprenorphine/naloxone. Several Council Members noted that the study population (patients who were well-managed on an 8 mg per day of buprenorphine/naloxone) represented patients with less-severe OUD, limiting the generalizability of the study. Two Council Members also noted that Probuphine is not considered to be a long-term therapy, as patients will need to transition back to transmucosal buprenorphine/naloxone after a maximum of 24 weeks on Probuphine. 4) Is the evidence adequate to demonstrate that the net health benefit of the Naltrexone subcutaneous extended-release injection Vivitrol (Alkermes) is superior to that provided by transmucosal formulations of buprenorphine/naloxone? Yes: 2 votes No: 11 votes Comments: A majority of the Council judged that the evidence was inadequate to demonstrate that the net health benefit of Vivitrol is superior to transmucosal formulations of buprenorphine/naloxone. Council Members voting in the negative noted that the patients who are most likely to use Vivitrol will be seeking to completely taper off opioids (e.g., for professional requirements) and represent a subset of the broader population with OUD. Council Members also noted higher relapse rates in the Vivitrol trial versus transmucosal buprenorphine. One Council Member voting in the affirmative noted that Vivitrol may work better than transmucosal buprenorphine for the subset of patients able to successfully complete the mandatory detoxification period. The two Council Members ©Institute for Clinical and Economic Review, 2018 Page 86 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents voting in the affirmative noted that the long history of the drug increased their certainty regarding its benefit and safety profile, and Vivitrol’s potential protective effect against lethal overdoses. 5) Is the evidence adequate to distinguish the net health benefit among the following interventions: (1) the two buprenorphine subcutaneous extended release injections (Sublocade and CAM2038); (2) the buprenorphine implant (Probuphine); (3) naltrexone intramuscular extended-release injection (Vivitrol)? Yes: 1 vote No: 12 votes Comments: A majority of the Council determined that the evidence was inadequate to distinguish the net health benefit among Sublocade, CAM2038, Probuphine and Vivitrol. The Council underlined the lack of head-to-head trials, substantial heterogeneity in studies that prevented indirect comparisons, and the lack of data beyond 24 weeks of treatment. 6) Does treating patients with one of the extended-release interventions (CAM2038, Sublocade, Probuphine, or Vivitrol) offer one or more of the following potential “other benefits” vs transmucosal formulations of buprenorphine/naloxone? CAM2038 and Sublocade offer reduced complexity that will significantly improve patient 10/13 outcomes. Probuphine offers reduced complexity that will significantly improve patient outcomes. 4/13 Vivitrol offers reduced complexity that will significantly improve patient outcomes. 8/13 These interventions will reduce important health disparities across racial, ethnic, gender, 7/13 socioeconomic, or regional categories. These interventions will significantly reduce caregiver or broader family burden. 6/13 CAM2038 and Sublocade offer a novel mechanism of action or approach that will allow 7/13 successful treatment of many patients for whom other available treatments have failed. Probuphine offers a novel mechanism of action or approach that will allow successful 3/13 treatment of many patients for whom other available treatments have failed. Vivitrol offers a novel mechanism of action or approach that will allow successful treatment 8/13 of many patients for whom other available treatments have failed. These interventions will have a significant impact on improving patients’ ability to return to 5/13 work and/or their overall productivity. There are other important benefits or disadvantages that should have an important role in 7/13 judgments of the value of these interventions. Comments: The Council identified several other important benefits such as the reduction of stigma (i.e., less frequent dosing schedules require fewer doctors’ visits that could lead to improved privacy for patients) and the potential of the injectable forms to reduce diversion. The Council also underlined the notion that access to a wider range of treatment options could lead to improvements in population-level health because although evidence demonstrates similar benefits for each drug, patient subpopulations may respond more favorably to one drug over another due to differences in indications, routes of ©Institute for Clinical and Economic Review, 2018 Page 87 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents administration, and mechanisms of action. While there is currently inadequate evidence to identify which patients will respond better to each therapy, the clinical community may gain important insights as these drugs are more frequently used. One Council Member noted that injectable forms of buprenorphine may be superior to transmucosal buprenorphine in ED settings, as they will protect against overdoses soon after patients are discharged. One Council Member noted that extended-release drugs administered on a monthly basis in clinical settings provide an opportunity for increased patient-provider interaction over therapies that are dispensed at a pharmacy, while another member noted the potential for extended-release forms to reduce health disparities in rural areas by requiring fewer visits to the provider. Conversely, another Council member expressed concerns that the less- frequent dosing schedule for Sublocade may lead to decreased patient adherence as a result of infrequent contact with their provider. 7) Are any of the following contextual considerations important in assessing the long-term value for money of the extended-release interventions (CAM2038, Sublocade, Probuphine, or Vivitrol)? These interventions are intended for the care of individuals with a condition of particularly 11/12 high severity in terms of impact on length of life and/or quality of life. These interventions are intended for the care of individuals with a condition that represents 12/12 a particularly high lifetime burden of illness. These interventions are the first to offer any improvement for patients with this condition. 0/12 There is significant uncertainty about the long-term risk of serious side effects of CAM2038. 2/12 There is significant uncertainty about the long-term risk of serious side effects of Sublocade. 3/12 There is significant uncertainty about the long-term risk of serious side effects of 3/12 Probuphine. There is significant uncertainty about the long-term risk of serious side effects of Vivitrol. 3/12 There is significant uncertainty about the magnitude or durability of the long-term benefits 9/12 of CAM2038. There is significant uncertainty about the magnitude or durability of the long-term benefits 9/12 of Sublocade. There is significant uncertainty about the magnitude or durability of the long-term benefits 2/12 of Probuphine. There is significant uncertainty about the magnitude or durability of the long-term benefits 9/12 of Vivitrol. There are additional contextual considerations that should have an important role in 6/12 judgments of the value of this intervention. Note: Only 12 Council Members voted on this question because one member had to leave early due to a medical emergency. Comments: As Probuphine is administered for six months only, the Council did not consider there to be uncertainty related to its long-term benefits or side effects. One Council Member mentioned that adoption of implantable treatments may encourage development ©Institute for Clinical and Economic Review, 2018 Page 88 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents of new treatment paradigms. From a broader perspective, one Council Member noted that the price of generic buprenorphine will continue to drop, which will make other treatment options less attractive by comparison. Finally, one Council Member noted that RCTs may underestimate benefits in the complex and heterogeneous population, as a subpopulation may benefit more from a drug than another. Long-Term Value for Money As described in ICER’s value assessment framework, questions on “long-term value for money” are subject to a value vote only when incremental cost-effectiveness ratios for the interventions of interest are between $50,000 and $175,000 per QALY in the primary “base case” analysis. As shown in the analysis, the estimates for Probuphine, Sublocade, and Vivitrol exceed the higher end of the range. Consequently, all three interventions were deemed “low value” without formal voting by the public Council. CAM2038 was not yet approved at the time of the meeting, and no price was available, so an incremental cost-effectiveness ratio could not be calculated; as a consequence, a value vote was not taken. 8.3 Roundtable Discussion and Key Policy Implications Following its deliberation on the evidence, the New England CEPAC Panel engaged in a moderated discussion with a policy roundtable about how best to apply the evidence on extended-release medications for addiction treatment in patients with opioid use disorder to policy and practice. The policy roundtable members included one policy expert, two clinical experts, two payers, and three manufacturer representatives (one patient advocate was invited to participate in the roundtable, but did not attend the meeting). 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 G. Table 8.1. Policy Roundtable Members Name Title and Affiliation Barbara Henry, RPh Lead Pharmacy Specialist, Harvard Pilgrim Health Care Kimberly Lenz, PharmD Clinical Pharmacy Manager, MassHealth Richard Malamut, MD Chief Medical Officer, Braeburn Lewis Nelson, MD Professor and Chair, Department of Emergency Medicine; Chief, Division of Medical Toxicology, Rutgers New Jersey Medical School Amy K. O’Sullivan, PhD Head of Health Economics and Outcomes Research, Alkermes Maria Schiff, MPH Senior Officer, Substance Use Prevention and Treatment Initiative, The Pew Charitable Trusts Ann Wheeler, PharmD, BCPP National Director of Managed Care Medical Affairs; Head of Behavioral Health Medical Affairs, Indivior Joe Wright, MD Medical Director, Boston Health Care for the Homeless Program; Clinician, CareZone ©Institute for Clinical and Economic Review, 2018 Page 89 Final Evidence Report - MAT in Patients with OUD 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 audience and summarized below. All Stakeholders Decrease stigma by aligning efforts around education that enhances awareness that Opioid Use Disorder (OUD) is a chronic disease requiring long-term treatment. Only a small fraction of patients with OUD have access to evidence-based treatment. Some estimate that overall only about 3 or 4% of patients in the US currently have access to MAT, and there are important geographic disparities. There is stigma associated with a perception of OUD as a moral failure, which presents an obstacle both for making MAT widely available as well as for patient acceptance of MAT. Manufacturers Bring the price of extended-release medications into alignment with their clinical value. The opioid epidemic represents a public health crisis, and we heard repeatedly that there should be access to all available therapies. But the current pricing of the extended-release formulations represents a missed opportunity to address this crisis. There is an important shared responsibility between manufacturers and insurers to match responsible pricing with insurance programs that help every patient obtain access to the type of treatment that will work best for them. Manufacturers bear sole responsibility for prices that greatly exceed the added clinical benefits of new treatments and therefore make this kind of policy outcome difficult. Payers For treatments for OUD whose prices are aligned with clinical value, payers should create coverage criteria that present no barriers to access. In particular, prior authorization criteria for Sublocade and similar extended-release treatments should be flexible enough to support evidence-based individualized treatment decisions. Based on available evidence and clinical expert testimony at the CEPAC meeting, the following prior authorization considerations should guide the development of payer policies for coverage of Sublocade and similar extended-release OUD treatments: Patient eligibility 1. Concomitant behavioral health counseling. Although counseling has been shown to be important in supporting beneficial outcomes of OUD treatment, prior auth criteria requiring documentation of ongoing counseling may present a barrier to treatment. Instead, payers may wish to ask the prescriber to attest that they have a relationship through which they can refer the ©Institute for Clinical and Economic Review, 2018 Page 90 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents patient to counseling or care management. If this is not available, the payer should consider reaching out to the provider/patient to help arrange these services. 2. Medical necessity for Sublocade versus buprenorphine. In an environment with a significant price difference between transmucosal buprenorphine therapy and extended-release formulations, payers may consider requiring that patients attempt treatment with transmucosal therapy before transitioning to extended-release formulations. This would likely not be necessary if the price for extended-release formulations were to align with value. For patients for whom Sublocade treatment is being considered, stabilization on buprenorphine at a dose ≤ 24 mg per day for at least seven days is a reasonable prior auth criterion to ensure that patients can be adequately managed with Sublocade alone. The medical necessity for Sublocade over less expensive buprenorphine can also be supported by specific home and life situations. For example, patients with unstable home and life situations, where storage of the daily transmucosal doses is challenging, are likely to have greater chance to benefit from OUD treatment through an extended-release agent. Similarly, patients who desire treatment but have demonstrated that they cannot easily maintain the commitment to daily treatment are viewed as good candidates for Sublocade. 3. Coverage of multiple extended-release agents. The likely advent of CAM2038 as another option for extended-release injectable treatment will raise questions about whether clinicians and patients require access to both Sublocade and CAM2038. If payers are able to achieve value-based prices for both of these products, both should be covered. If this is not the case, clinical experts did not argue that there was an obvious clinical reason necessitating that both Sublocade and CAM2038 be available. Accordingly, payers may consider granting formulary placement to only one of these agents if doing so allows for superior affordability to be achieved. Mechanisms for appeal to allow coverage for additional treatment options should always be available and not overly burdensome. Payers and Policy Makers Avoid legislative action favoring one form of MAT The media has reported on efforts by Alkermes, the manufacturer of Vivitrol, to give this drug preferred status in some treatment contexts through state legislation. In the face of an epidemic, it is particularly important that all stakeholders align to reduce barriers to care and avoid actions that could be interpreted as creating stigma against any treatment options that are well supported by evidence. Effective MAT requires access to all evidence-based treatment options, as stated in the 2018 joint policy statement by the American Correctional Association and American Society of Addiction on OUD treatment in the justice system.152 ©Institute for Clinical and Economic Review, 2018 Page 91 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Coordinate MAT for individuals leaving the correctional system and ensure continuity of care Individuals leaving the correctional system often experience a significant gap in OUD treatment and access to medical care. This treatment gap leads to very large increase in overdose death rates after release. Systems should be created to provide MAT to those individuals who need it, to connect them with needed care, and to provide the necessary insurance coverage for this care. Take action to address the many long-term policy goals that are yet to be achieved in order to improve treatment for OUD. In 2014 ICER published a report on the management of patients with opioid dependence 34 and identified several actions needed from payers and policymakers153. Policy makers should recognize that most of these long-term policy goals are still to be achieved. We encourage them and others to read the policy recommendations and other material available at https://icer-review.org/wp- content/uploads/2016/01/FINAL-Payer-Action-Guide.pdf. Regulators Consider eliminating restrictions on prescribing extended-release formulations of buprenorphine Access to MAT remains a major challenge for effectively managing OUD in the US. Training and reporting requirements needed for obtaining a waiver for prescribing buprenorphine, as well as limitations on the number of patients providers can treat, constitute important impediments to access to MAT. A clinical expert at the meeting pointed out that no such limitations exist for providers who choose to prescribe Oxycontin or other opioids. Given the low risk of diversion of extended-release buprenorphine formulations, waiver requirements and numeric limitations do not seem justified for these medications. Researchers The research community should work with clinicians and manufacturers to identify clinical characteristics that would better predict which OUD patients are likely to benefit most from the available MAT approaches All stakeholders agree that for MAT treatment is not “one-size-fits-all” and that different patients and patient groups need different approaches for treatment. However, evidence is still lacking on which patients are most likely to benefit from the available options for therapy. All pivotal trials should include an active comparator arm, and treatments should be compared head-to-head in post-registration observational analyses as well. Pivotal trials should use the broadest possible eligibility criteria and, importantly, define patient clinical characteristics and OUD use in consistent terms that will facilitate evaluations of comparative effectiveness. Outcome measures for all pivotal trials should include quality of life and other patient-centered outcomes. ©Institute for Clinical and Economic Review, 2018 Page 92 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents **** This is the second ICER review of MAT for OUD. The first ICER review can be found here: https://icer-review.org/topic/opioid-dependence/ ©Institute for Clinical and Economic Review, 2018 Page 93 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents References 1. HHS. What is the U.S. Opioid Epidemic? 2018; https://www.hhs.gov/opioids/about-the- epidemic/index.html. Accessed Aug 8, 2018. 2. Dowell D, Arias E, Kochanek K, et al. Contribution of Opioid-Involved Poisoning to the Change in Life Expectancy in the United States, 2000-2015. Jama. 2017;318(11):1065-1067. 3. Wakeman SE, Barnett ML. Primary Care and the Opioid-Overdose Crisis - Buprenorphine Myths and Realities. N Engl J Med. 2018;379(1):1-4. 4. HHS Press Office. HHS Acting Secretary Declares Public Health Emergency to Address National Opioid Crisis. 2017; https://www.hhs.gov/about/news/2017/10/26/hhs-acting-secretary- declares-public-health-emergency-address-national-opioid-crisis.html. 5. Council of Economic Advisers. The Underestimated Cost of the Opioid Crisis. 2017; https://www.whitehouse.gov/sites/whitehouse.gov/files/images/The%20Underestimated%20C ost%20of%20the%20Opioid%20Crisis.pdf. Accessed April 23, 2018. 6. SAMHSA. EXHIBIT 2.13. DSM-5 Criteria for OUD in TIP 63: Medications for Opioid Use Disorder. 2018; https://store.samhsa.gov/shin/content//SMA18-5063FULLDOC/SMA18- 5063FULLDOC.pdf#page=57. Accessed Aug 13, 2018. 7. American Academy of Pain Medicine, American Pain Society, American Society of Addiction Medicine. Definitions Related to the Use of Opioids for the Treatment of Pain: Consensus Statement of the American Academy of Pain Medicine, the American Pain Society, and the American Society of Addiction Medicine. 2001; http://www.asam.org/docs/default- source/public-policy-statements/1opioid-definitions-consensus-2-011.pdf? Accessed 2018-08- 03, 2018. 8. White House Office of National Drug Control Policy. Changing the Language of Addiction. In: Washington, DC:: White House Office of National Drug Control Policy; 2017: https://obamawhitehouse.archives.gov/blog/2017/01/13/changing-language-addiction. Accessed 2018-08-09. 9. SAMHSA. Recovery and Recovery Support. 2018; https://www.samhsa.gov/recovery. Accessed 2018-08-03. 10. McBain R, Rose AJ, LaRochelle MR. The U.S. opioid epidemic: One disease, diverging tales. Prev Med. 2018;112:176-178. 11. Ross J, Teesson M, Darke S, et al. The characteristics of heroin users entering treatment: findings from the Australian treatment outcome study (ATOS). Drug and alcohol review. 2005;24(5):411- 418. 12. SAMHSA. TIP 63: Medications for Opioid Use Disorder. Treatment Improvement Protocol (TIP) series 2018; https://store.samhsa.gov/shin/content//SMA18-5063FULLDOC/SMA18- 5063FULLDOC.pdf. Accessed April 18, 2018. 13. FDA. Information about Medication-Assisted Treatment (MAT). 2018; https://www.fda.gov/Drugs/DrugSafety/InformationbyDrugClass/ucm600092.htm. Accessed Aug 12, 2018. 14. FDA. CDER Conversation: Treatment for Opioid Use Disorder. 2018; https://www.fda.gov/Drugs/NewsEvents/ucm611659.htm. Accessed Aug 10, 2018. 15. SAMHSA. EXHIBIT 1.2. Comparison of Medications for OUD in in TIP 63: Medications for Opioid Use Disorder. 2018; https://store.samhsa.gov/shin/content//SMA18-5063FULLDOC/SMA18- 5063FULLDOC.pdf#page=28. Accessed Aug 13, 2018. ©Institute for Clinical and Economic Review, 2018 Page 94 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents 16. Bart G. Maintenance medication for opiate addiction: the foundation of recovery. J Addict Dis. 2012;31(3):207-225. 17. World Health Organization, International Narcotics Control Board, United Nations Office on Drugs and Crime. Guidelines for the psychosocially assisted pharmacological treatment of opioid dependence. 2009; xviii, 111 p. Available at: http://www.who.int/substance_abuse/publications/opioid_dependence_guidelines.pdf. 18. ASAM. The ASAM National Practice Guideline for the Use of Medications in the Treatment of Addiction Involving Opioid Use. 2015; https://www.asam.org/docs/default-source/practice- support/guidelines-and-consensus-docs/asam-national-practice-guideline-supplement.pdf. Accessed Aug 10, 2018. 19. FDA. Opioid Use Disorder: Endpoints for Demonstrating Effectiveness of Drugs for Medication- Assisted Treatment - Draft Guidance for Industry. In:2018: https://www.fda.gov/downloads/Drugs/GuidanceComplianceRegulatoryInformation/Guidances /UCM615743.pdf. 20. Wakeman SE, Rich JD. Barriers to Medications for Addiction Treatment: How Stigma Kills. Subst Use Misuse. 2018;53(2):330-333. 21. Saloner B, Karthikeyan S. Changes in Substance Abuse Treatment Use Among Individuals With Opioid Use Disorders in the United States, 2004-2013. Jama. 2015;314(14):1515-1517. 22. FDA. Vivitrol label. 2015; https://www.accessdata.fda.gov/drugsatfda_docs/label/2015/021897s029lbl.pdf. Accessed Aug 10, 2018. 23. FDA. Probuphine label. 2018; https://www.accessdata.fda.gov/drugsatfda_docs/label/2018/204442s006lbl.pdf. Accessed Aug 13, 2018. 24. Braeburn Pharmaceuticals. Slides for the November 1, 2017 Joint Meeting of the Psychopharmacologic Drugs Advisory Committee and the Drug Safety and Risk Management Advisory Committee. 2017; https://www.fda.gov/downloads/AdvisoryCommittees/CommitteesMeetingMaterials/Drugs/Ps ychopharmacologicDrugsAdvisoryCommittee/UCM586734.pdf#page=4. Accessed Aug 13, 2018. 25. FDA. Public Meeting on Patient-Focused Drug Development for Opioid Use Disorder-April 17, 2018. 2018; https://www.fda.gov/ForIndustry/UserFees/PrescriptionDrugUserFee/ucm591290.htm. Accessed April 23, 2018. 26. Dennis BB, Roshanov PS, Naji L, et al. Opioid substitution and antagonist therapy trials exclude the common addiction patient: a systematic review and analysis of eligibility criteria. Trials. 2015;16:475. 27. FDA. FDA Briefing Information for the March 27, 2018 Meeting of the Psychopharmacologic Drugs Advisory Committee. In:2018: https://www.fda.gov/downloads/AdvisoryCommittees/CommitteesMeetingMaterials/Drugs/Ps ychopharmacologicDrugsAdvisoryCommittee/UCM602417.pdf#page=13. 28. Barlas S. U.S. and States Ramp Up Response to Opioid Crisis: Regulatory, Legislative, and Legal Tools Brought to Bear. P T. 2017;42(9):569-592. 29. Takkunen A, Hanson K. States fight the opioid epidemic by expanding access to treatment. National Conference of State Legislatures Blog 2018; http://www.ncsl.org/blog/2018/08/08/states-fight-the-opioid-epidemic-by-expanding-access- to-treatment.aspx. Accessed August 17, 2018. 30. Clemans-Cope L, Wishner JB, Allen EH, Lallemand N, Epstein M, Spillman BC. Experiences of three states implementing the Medicaid health home model to address opioid use disorder— ©Institute for Clinical and Economic Review, 2018 Page 95 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Case studies in Maryland, Rhode Island, and Vermont. Journal of substance abuse treatment. 2017;83:27-35. 31. Mohlman MK, Tanzman B, Finison K, Pinette M, Jones C. Impact of Medication-Assisted Treatment for Opioid Addiction on Medicaid Expenditures and Health Services Utilization Rates in Vermont. Journal of substance abuse treatment. 2016;67:9-14. 32. Simpatico TA. Vermont responds to its opioid crisis. Prev Med. 2015;80:10-11. 33. Green TC, Clarke J, Brinkley-Rubinstein L, et al. Postincarceration Fatal Overdoses After Implementing Medications for Addiction Treatment in a Statewide Correctional System. JAMA psychiatry. 2018;75(4):405-407. 34. Institute for Clinical and Economic Review. Management of Patients with Opioid Dependence: A Review of Clinical, Delivery System, and Policy Options. In: Institute for Clinical and Economic Review, ed. Boston2014: https://icer-review.org/wp-content/uploads/2016/01/CEPAC-Opioid- Dependence-Final-Report-For-Posting-July-21.pdf. Accessed 2018-04-11. 35. Health NIo. Opioids, opioid antagonists. https://livertox.nih.gov/Opioids.htm 36. Jensen TS. Opioids in the brain: supraspinal mechanisms in pain control. Acta Anaesthesiol Scand. 1997;41(1 Pt 2):123-132. 37. Fields HL, Margolis EB. Understanding opioid reward. Trends Neurosci. 2015;38(4):217-225. 38. Holden JE, Jeong Y, Forrest JM. The endogenous opioid system and clinical pain management. AACN Clin Issues. 2005;16(3):291-301. 39. Hou H, Wang C, Jia S, Hu S, Tian M. Brain dopaminergic system changes in drug addiction: a review of positron emission tomography findings. Neurosci Bull. 2014;30(5):765-776. 40. O'Brien C. Addiction and dependence in DSM-V. Addiction (Abingdon, England). 2011;106(5):866-867. 41. Savage SR, Joranson DE, Covington EC, Schnoll SH, Heit HA, Gilson AM. Definitions related to the medical use of opioids: Evolution towards universal agreement. Journal of Pain and Symptom Management. 2003;26(1):655–667. https://www.jpsmjournal.com/article/S0885- 3924(03)00219-7/pdf. Accessed 2018-08-06. 42. California Health Benefits Review Program. Analysis of California Assembly Bill 2384 Medication- Assisted Treatment. In:2018-05-15: http://analyses.chbrp.com/document/view.php?id=1343. Accessed 2008-04-18. 43. Institute for Clinical and Economic Review. Management of Patients with Opioid Dependence: A Review of Clinical, Delivery System, and Policy Options - Appendices. In: Boston2014: https://icer-review.org/wp-content/uploads/2016/02/Appendices-Final-Opioid-Dependence- Report-CEPAC-10-6.pdf. Accessed 2018-07-11. 44. SAMHSA. Federal Guidelines for Opioid Treatment Programs. HHS Publication No. (SMA) PEP15- FEDGUIDEOTP 2015; https://store.samhsa.gov/shin/content/PEP15-FEDGUIDEOTP/PEP15- FEDGUIDEOTP.pdf. 45. Volkow ND. Medications for opioid use disorder: bridging the gap in care. Lancet (London, England). 2018;391(10118):285-287. 46. Middlesex Sheriff’s Office. The MATADOR Program - Utilizing Incarceration to Tackle Addiction and Save Lives: Implementing Medication Assisted Treatment Programs in Jails. In: Middlesex Sheriff’s Office, Middlesex County, MA; 2018: https://www.sheriffs.org/sites/default/files/MATADOR%20White%20Paper_April2018.pdf. Accessed 2018-08-10. 47. FDA. Sublocade label. 2018; https://www.accessdata.fda.gov/drugsatfda_docs/label/2018/209819s001lbl.pdf. Accessed Aug 14, 2018. ©Institute for Clinical and Economic Review, 2018 Page 96 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents 48. FDA. 2017 Meeting Materials, Drug Safety and Risk Management Advisory Committee. 2017; https://www.fda.gov/AdvisoryCommittees/CommitteesMeetingMaterials/Drugs/DrugSafetyand RiskManagementAdvisoryCommittee/ucm536632.htm. Accessed August 14, 2018. 49. FDA. Minutes for the November 1, 2017 Joint Meeting of the Psychopharmacologic Drugs Advisory Committee and the Drug Safety and Risk Management Advisory Committee. 2017; https://www.fda.gov/downloads/AdvisoryCommittees/CommitteesMeetingMaterials/Drugs/Ps ychopharmacologicDrugsAdvisoryCommittee/UCM592675.pdf. Accessed Aug 14, 2018. 50. Woolf S. An organized analytic framework for practice guideline development: using the analytic logic as a guide for reviewing evidence, developing recommendations, and explaining the rationale. Clinical Practice Guideline Development: Methodology Perspectives. AHCPR Pub 1994; 105-113. Available at: https://ntrl.ntis.gov/NTRL/ under number PB95195848. 51. Maglione MA, Raaen L, Chen C, et al. Effects of medication assisted treatment (MAT) for opioid use disorder on functional outcomes: A systematic review. Journal of substance abuse treatment. 2018;89:28-51. 52. CMS.gov. National Coverage Determination (NCD) for Treatment of Alcoholism and Drug Abuse in a Freestanding Clinic (130.5). In: Services CfMM, ed. cms.gov2018. 53. CMS.gov. National Coverage Determination (NCD) for Treatment of Drug Abuse (Chemical Dependency) (130.6). In: Services CfMM, ed. cms.gov2018. 54. Cigna. Cigna Standards and Guidelines/Medical Necessity Criteria for Treatment of Mental Health and Substance Use Disorders. In: Cigna, ed. Revised Edition ed. cigna.com2018. 55. Aetna. Buprenorphine Implant (Probuphine) and Extended-Release Injectable (Sublocade). 2018; http://www.aetna.com/cpb/medical/data/900_999/0910.html. Accessed 08/15/18. 56. Anthem. Sublocade (buprenorphine extended-release). In:2018. 57. UnitedHealthcare. Medical Benefit Drug Policy: BUPRENORPHINE (PROBUPHINE & SUBLOCADE). In. uhcprovider.com2018. 58. BlueCrossBlueShieldMA. Medication Look Up. 2018; https://www.bluecrossma.com/wps/portal/members/using-my-plan/manage-my- plan/pharmacy-coverage/medication-look-up/. Accessed 08/15/18, 2018. 59. Anthem. Probuphine (buprenorphine implant). 2018. 60. BlueCrossBlueShieldMA. Pharmacy Medical Policy: Medical Utilization Management (MED UM) & Pharmacy Prior Authorization Policy. 2018. 61. Aetna. 2018 Aetna Pharmacy Drug Guide: Aetna Value Plus Plan. 2018. 62. Cigna. Cigna Value 3-Tier Prescription Drug List. 2018. 63. Care HPH. 3-Tier Value Formulary Drug Tier Look-Up (2018). 2018; https://www.harvardpilgrim.org/hphconnect/formulary/drugLookup.htm?&formularyId=VALUE _3_TIER&year=2018. Accessed 08/15/18, 2018. 64. UnitedHealthcare. Your 2018 Prescription Drug List: Traditional Three-Tier. 2018. 65. Anthem. National Drug List - Three (3) Tier Drug Plan. 2018. 66. MassHealth. Table 36: Drug and Alcohol Cessation Agents. 2018; https://masshealthdruglist.ehs.state.ma.us/MHDL/pubtheradetail.do?id=36. Accessed 08/15/18, 2018. 67. DVHA. Department of Vermont Health Access Managed Care Entity Vermont Substance Use Disorder Medication Assisted Treatment Guidelines. 2018. 68. AATOD. AATOD Guidelines for Using Naltrexone (Vivitrol) in OTPs. 2012. 69. AMCP. AMCP Addiction Treatment Advisory Group: Findings and Considerations for the Evidence-Based Use of Medications Used in the Treatment of Substance Use Disorder. 2016. 70. ASAM. The ASAM National Practice Guideline for the Use of Medications in the Treatment of Addiction Involving Opioid Use. 2015. ©Institute for Clinical and Economic Review, 2018 Page 97 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents 71. National Institute on Drug Abuse. Principles of Drug Addiction Treatment: A Research-Based Guide (Third Edition). In:2018: https://d14rmgtrwzf5a.cloudfront.net/sites/default/files/675- principles-of-drug-addiction-treatment-a-research-based-guide-third-edition.pdf. 72. Sigmon SC, Dunn KE, Saulsgiver K, et al. A randomized, double-blind evaluation of buprenorphine taper duration in primary prescription opioid abusers. JAMA psychiatry. 2013;70(12):1347-1354. 73. Moher D, Liberati A, Tetzlaff J, Altman DG. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. International journal of surgery (London, England). 2010;8(5):336-341. 74. U.S. Preventive Services Task Force Procedure Manual. Agency for Healthcare Research and Quality; July 2008. Publication No. 08-05118-EF. 75. Review IfCaE. ICER Evidence Rating Matrix: A User’s Guide. 2017. 76. Lofwall MR, Walsh SL, Nunes EV, et al. Weekly and Monthly Subcutaneous Buprenorphine Depot Formulations vs Daily Sublingual Buprenorphine With Naloxone for Treatment of Opioid Use Disorder: A Randomized Clinical Trial. JAMA internal medicine. 2018;178(6):764-773. 77. ClinicalTrials.gov. Trial 13-0001 Sublocade vs Placebo. https://clinicaltrials.gov/ct2/show/results/NCT02357901?term=NCT02357901&rank=1&sect=Xf e897d0156#outcome9. 78. ClinicalTrials.gov. Trial 13-0003 Safety and Tolerability Study of Depot Buprenorphine in Treatment Seeking Subjects With Opioid Use Disorder. 2018; https://clinicaltrials.gov/ct2/show/results/NCT02510014?term=NCT02510014&rank=1&sect=X4 301256. 79. Rosenthal RN, Lofwall MR, Kim S, Chen M, Beebe KL, Vocci FJ. Effect of Buprenorphine Implants on Illicit Opioid Use Among Abstinent Adults With Opioid Dependence Treated With Sublingual Buprenorphine: A Randomized Clinical Trial. Jama. 2016;316(3):282-290. 80. Ling W, Casadonte P, Bigelow G, et al. Buprenorphine implants for treatment of opioid dependence: A randomized controlled trial. JAMA: Journal of the American Medical Association. 2010;304(14):1576-1583. 81. Rosenthal RN, Ling W, Casadonte P, et al. Buprenorphine implants for treatment of opioid dependence: randomized comparison to placebo and sublingual buprenorphine/naloxone. Addiction. 2013;108(12):2141-2149. 82. Beebe KL, Chavoustie S, Ling W, Sigmon S, Leiderman D, Bailey G. Buprenorphine implants for the treatment of opioid dependence: Six and 12 month outcomes. Neuropsychopharmacology. 2012;38:S266-S267. 83. Lee JD, Nunes EV, Jr., Novo P, et al. Comparative effectiveness of extended-release naltrexone versus buprenorphine-naloxone for opioid relapse prevention (X:BOT): a multicentre, open- label, randomised controlled trial. Lancet (London, England). 2018;391(10118):309-318. 84. Tanum L, Solli KK, Latif Z-e-H, et al. Effectiveness of injectable extended-release naltrexone vs daily buprenorphine-naloxone for opioid dependence: A randomized clinical noninferiority trial. JAMA psychiatry. 2017;74(12):1197-1205. 85. Lee JD, Friedmann PD, Kinlock TW, et al. Extended-Release Naltrexone to Prevent Opioid Relapse in Criminal Justice Offenders. N Engl J Med. 2016;374(13):1232-1242. 86. Krupitsky E, Nunes EV, Ling W, Illeperuma A, Gastfriend DR, Silverman BL. Injectable extended- release naltrexone for opioid dependence: a double-blind, placebo-controlled, multicentre randomised trial. Lancet (London, England). 2011;377(9776):1506-1513. 87. Lee JD, Friedmann PD, Boney TY, et al. Extended-release naltrexone to prevent relapse among opioid dependent, criminal justice system involved adults: rationale and design of a randomized controlled effectiveness trial. Contemp Clin Trials. 2015;41:110-117. ©Institute for Clinical and Economic Review, 2018 Page 98 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents 88. ClinicalTrials.gov. NEW HOPE. 2017; https://clinicaltrials.gov/ct2/show/results/NCT01246401?sect=X5643012. 89. Solli KK, Latif ZE, Opheim A, et al. Effectiveness, safety and feasibility of extended-release naltrexone for opioid dependence: a 9-month follow-up to a 3-month randomized trial. Addiction (Abingdon, England). 2018. 90. Krupitsky E, Nunes EV, Ling W, Gastfriend DR, Memisoglu A, Silverman BL. Injectable extended- release naltrexone (XR-NTX) for opioid dependence: Long-term safety and effectiveness. Addiction (Abingdon, England). 2013;108(9):1628-1637. 91. Lee JD, McDonald R, Grossman E, et al. Opioid treatment at release from jail using extended- release naltrexone: A pilot proof-of-concept randomized effectiveness trial. Addiction (Abingdon, England). 2015;110(6):1008-1014. 92. Lee JD, Friedmann PD, Kinlock TW, et al. Extended-Release Naltrexone to Prevent Opioid Relapse in Criminal Justice Offenders. New England Journal of Medicine. 2016;374(13):1232- 1242. 93. Krupitsky E. Naltrexone for opiate dependence: Oral, implantable, and injectable. European Neuropsychopharmacology. 2011;21:S104-S105. 94. Soares WE, Iii W, Donna R, et al. Healthcare utilization in adults with opioid dependence receiving extended release naltrexone compared to treatment as usual. Journal of substance abuse treatment. 2018;85:66-69. 95. Shah A, Duncan M, Atreja N, Tai KS, Gore M. Healthcare utilization and costs associated with treatment for opioid dependence. J Med Econ. 2018;21(4):406-415. 96. Vivitrol pacakage insert. 97. Saucier R, Wolfe D, Dasgupta N. Review of Case Narratives from Fatal Overdoses Associated with Injectable Naltrexone for Opioid Dependence. Drug Safety. 2018:1-8. 98. Mattick RP, Breen C, Kimber J, Davoli M. Methadone maintenance therapy versus no opioid replacement therapy for opioid dependence. Cochrane Database of Systematic Reviews. 2009(3). 99. CADTH. Buprenorphine/naloxone versus methadone for the treatment of opioid dependence: A review of comparative clinical effectiveness, cost-effectiveness and guidelines. 2016. 100. Fudala PJ, Bridge TP, Herbert S, et al. Office-based treatment of opiate addiction with a sublingual-tablet formulation of buprenorphine and naloxone. N Engl J Med. 2003;349(10):949- 958. 101. Mattick RP, Breen C, Kimber J, Davoli M. Buprenorphine maintenance versus placebo or methadone maintenance for opioid dependence. Cochrane Database of Systematic Reviews. 2014(2). 102. Newman RG, Whitehill WB. Double-blind comparison of methadone and placebo maintenance treatments of narcotic addicts in Hong Kong. Lancet (London, England). 1979;2(8141):485-488. 103. Saxon AJ, Hser YI, Woody G, Ling W. Medication-assisted treatment for opioid addiction: methadone and buprenorphine. Journal of food and drug analysis. 2013;21(4):S69-s72. 104. Neumann AM, Blondell RD, Jaanimagi U, et al. A preliminary study comparing methadone and buprenorphine in patients with chronic pain and coexistent opioid addiction. J Addict Dis. 2013;32(1):68-78. 105. Proctor SL, Copeland AL, Kopak AM, Herschman PL, Polukhina N. A naturalistic comparison of the effectiveness of methadone and two sublingual formulations of buprenorphine on maintenance treatment outcomes: findings from a retrospective multisite study. Experimental and clinical psychopharmacology. 2014;22(5):424-433. ©Institute for Clinical and Economic Review, 2018 Page 99 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents 106. Curcio F, Franco T, Topa M, Baldassarre C. Buprenorphine/naloxone versus methadone in opioid dependence: a longitudinal survey. European review for medical and pharmacological sciences. 2011;15(8):871-874. 107. Otiashvili D, Piralishvili G, Sikharulidze Z, Kamkamidze G, Poole S, Woody GE. Methadone and buprenorphine-naloxone are effective in reducing illicit buprenorphine and other opioid use, and reducing HIV risk behavior--outcomes of a randomized trial. Drug and alcohol dependence. 2013;133(2):376-382. 108. Hser YI, Evans E, Huang D, et al. Long-term outcomes after randomization to buprenorphine/naloxone versus methadone in a multi-site trial. Addiction. 2016;111(4):695-705. 109. Schackman BR, Leff JA, Polsky D, Moore BA, Fiellin DA. Cost-effectiveness of long-term outpatient buprenorphine-naloxone treatment for opioid dependence in primary care. Journal of general internal medicine. 2012;27(6):669-676. 110. Carter JA, Dammerman R, Frost M. Cost-effectiveness of subdermal implantable buprenorphine versus sublingual buprenorphine to treat opioid use disorder. J Med Econ. 2017;20(8):893-901. 111. Alderks CE. Trends in the Use of Methadone and Buprenorphine at Substance Abuse Treatment Facilities: 2003 to 2011. In: The CBHSQ Report. Rockville (MD): Substance Abuse and Mental Health Services Administration (US); 2013:1-7. 112. Treatment Seeking Participants With Opioid Use Disorders Assessing Tolerability of Depot Injections of Buprenorphine. National Institute of Health; 2018. https://clinicaltrials.gov/ct2/show/study/NCT02357901. Accessed July 1, 2018. 113. U.S. Food and Drug Administration. VIVITROL® (naltrexone for extended-release injectable suspension) Intramuscular. In:2015. 114. Lee JD, Nunes EV, Jr., Novo P, et al. Comparative effectiveness of extended-release naltrexone versus buprenorphine-naloxone for opioid relapse prevention (X:BOT): a multicentre, open- label, randomised controlled trial. The Lancet. 2018;391(10118):309-318. 115. U.S. Food and Drug Administration. PROBUPHINE (buprenorphine) implant for subdermal administration CIII. In:2002. 116. Rosenthal RN, Lofwall MR, Kim S, et al. Effect of buprenorphine implants on illicit opioid use among abstinent adults with opioid dependence treated with sublingual buprenorphine: A randomized clinical trial. Jama. 2016;316(3):282-290. 117. Lofwall MR, Walsh SL, Nunes EV, et al. Weekly and monthly subcutaneous buprenorphine depot formulations vs daily sublingual buprenorphine with naloxone for treatment of opioid use disorder: A randomized clinical trial. JAMA internal medicine. 2018;178(6):764-773. 118. Indivior. Data on File. 2018. 119. Bernard CL, Owens DK, Goldhaber-Fiebert JD, Brandeau ML. Estimation of the cost-effectiveness of HIV prevention portfolios for people who inject drugs in the United States: A model-based analysis. PLoS medicine. 2017;14(5):e1002312. 120. Hoyle MW, Henley W. Improved curve fits to summary survival data: application to economic evaluation of health technologies. BMC medical research methodology. 2011;11:139. 121. Lansky A, Finlayson T, Johnson C, et al. Estimating the number of persons who inject drugs in the united states by meta-analysis to calculate national rates of HIV and hepatitis C virus infections. PloS one. 2014;9(5):e97596. 122. Page K, Morris MD, Hahn JA, Maher L, Prins M. Injection drug use and hepatitis C virus infection in young adult injectors: using evidence to inform comprehensive prevention. Clinical infectious diseases : an official publication of the Infectious Diseases Society of America. 2013;57 Suppl 2:S32-38. 123. Berlin JA, Abrutyn E, Strom BL, et al. Incidence of infective endocarditis in the Delaware Valley, 1988-1990. The American journal of cardiology. 1995;76(12):933-936. ©Institute for Clinical and Economic Review, 2018 Page 100 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents 124. Spijkerman IJ, van Ameijden EJ, Mientjes GH, Coutinho RA, van den Hoek A. Human immunodeficiency virus infection and other risk factors for skin abscesses and endocarditis among injection drug users. J Clin Epidemiol. 1996;49(10):1149-1154. 125. Sande MA, Lee BL, Millills J, Chambers HF. Endocarditis in intravenous drug users. 2nd ed. New York: Raven Press; 1992. 126. Opioid Overdose Death Rates and All Drug Overdose Death Rates per 100,000 Population (Age- Adjusted). 2016. https://www.kff.org/other/state-indicator/opioid-overdose-death- rates/?currentTimeframe=0&selectedDistributions=opioid-overdose-death-rate-age- adjusted&selectedRows=%7B%22wrapups%22:%7B%22united- states%22:%7B%7D%7D%7D&sortModel=%7B%22colId%22:%22Location%22,%22sort%22:%22 asc%22%7D. Accessed 07/11/2018. 127. Human Mortality Database. 2016. https://usa.mortality.org/. Accessed 07/15/2018. 128. Lappalainen L, Hayashi K, Dong H, Milloy MJ, Kerr T, Wood E. Ongoing Impact of HIV Infection on Mortality among Persons who Inject Drugs Despite Free Antiretroviral Therapy. Addiction (Abingdon, England). 2015;110(1):111-119. 129. El-Kamary SS, Jhaveri R, Shardell MD. All-Cause, Liver-Related, and Non–Liver-Related Mortality Among HCV-Infected Individuals in the General US Population. Clinical Infectious Diseases: An Official Publication of the Infectious Diseases Society of America. 2011;53(2):150-157. 130. U.S. Food and Drug Administration. MAVYRETTM (glecaprevir and pibrentasvir) tablets, for oral use. In:2017. 131. Smith DJ, Jordan AE, Frank M, Hagan H. Spontaneous viral clearance of hepatitis C virus (HCV) infection among people who inject drugs (PWID) and HIV-positive men who have sex with men (HIV+ MSM): a systematic review and meta-analysis. BMC Infectious Diseases. 2016;16(1):471. 132. Wittenberg E, Bray JW, Aden B, Gebremariam A, Nosyk B, Schackman BR. Measuring benefits of opioid misuse treatment for economic evaluation: health-related quality of life of opioid- dependent individuals and their spouses as assessed by a sample of the US population. Addiction (Abingdon, England). 2015;111(4):675-684. 133. Connock M, Juarez-Garcia A, Jowett S, et al. Methadone and buprenorphine for the management of opioid dependence: a systematic review and economic evaluation. Health technology assessment (Winchester, England). 2007;11(9):1-171, iii-iv. 134. Sullivan PW, Ghushchyan V. Preference-Based EQ-5D index scores for chronic conditions in the United States. Med Decis Making. 2006;26(4):410-420. 135. Bernard CL, Brandeau ML, Humphreys K, et al. Cost-Effectiveness of HIV Preexposure Prophylaxis for People Who Inject Drugs in the United States. Annals of internal medicine. 2016:10.7326/M7315-2634. 136. Chhatwal J, Kanwal F, Roberts MS, Dunn MA. Cost-effectiveness and budget impact of hepatitis C virus treatment with sofosbuvir and ledipasvir in the United States. Annals of internal medicine. 2015;162(6):397-406. 137. Pharmaceutical Prices. U.S. Department of Veterans Affairs; 2018. Accessed July 15, 2018. 138. Alkermes. Data on File. 2018. 139. Redbook. 2018. Accessed July 11, 2018. 140. Physician Fee Schedule Search. 2018. https://www.cms.gov/apps/physician-fee- schedule/search/search-criteria.aspx. Accessed August 14, 2018. 141. Shah A, Duncan M, Atreja N, Tai KS, Gore M. Healthcare utilization and costs associated with treatment for opioid dependence. Journal of Medical Economics. 2018;21(4):406-415. 142. Health Care Cost Institute. 2016 Health Care Cost and Utilization Report. Washington, DC: Health Care Cost Institute;2018. ©Institute for Clinical and Economic Review, 2018 Page 101 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents 143. Johnson RL, Blumen HE, Ferro C. The burden of hepatitis C virus disease in commercial and managed Medicaid populations. Milliman;2015. 144. National Economic Accounts. United States Department of Labor; 2018. https://www.bea.gov/national/#gdp. Accessed August 10, 2018. 145. Birnbaum HG, White AG, Schiller M, Waldman T, Cleveland JM, Roland CL. Societal costs of prescription opioid abuse, dependence, and misuse in the United States. Pain medicine (Malden, Mass). 2011;12(4):657-667. 146. Substance Abuse & Mental Health Data Archive. National Survey on Drug Use and Health (NSDUH-2007). Substance Abuse & Mental Health Data Archive;2007. 147. Krebs E, Urada D, Evans E, Huang D, Hser YI, Nosyk B. The costs of crime during and after publicly funded treatment for opioid use disorders: a population-level study for the state of California. Addiction (Abingdon, England). 2017;112(5):838-851. 148. Jackson H, Mandell K, Johnson K, Chatterjee D, Vanness DJ. Cost-Effectiveness of Injectable Extended-Release Naltrexone Compared With Methadone Maintenance and Buprenorphine Maintenance Treatment for Opioid Dependence. Substance abuse. 2015;36(2):226-231. 149. Nosyk B, Guh DP, Bansback NJ, et al. Cost-effectiveness of diacetylmorphine versus methadone for chronic opioid dependence refractory to treatment. CMAJ : Canadian Medical Association journal = journal de l'Association medicale canadienne. 2012;184(6):E317-328. 150. Davenport S, Weaver A, Caverly M. Opioid prescription drug patterns in diagosed and non- diagnosed opioid use disorder populations. Milliman;2018. 151. Pearson SD. The ICER Value Framework: Integrating Cost Effectiveness and Affordability in the Assessment of Health Care Value. Value in health : the journal of the International Society for Pharmacoeconomics and Outcomes Research. 2018;21(3):258-265. 152. ASAM, ACA. Joint public correctional policy on the treatment of opioid use disorders for justice involved individuals. 2018; http://www.aca.org/ACA_Prod_IMIS/DOCS/ACA- ASAM%20Press%20Release%20and%20Joint%20Policy%20Statement%203.20.18.pdf. Accessed Nov 13, 2018. 153. Institute for Clinical and Economic Review. An Action Guide for Management of Opioid Dependence: Next Steps for Payers and Policymakers. In: Review IfCaE, ed.2014: https://icer- review.org/wp-content/uploads/2016/01/FINAL-Payer-Action-Guide.pdf. 154. Lofwall MR, Bailey GL, Frost MP, et al. An efficacy and safety comparision of CAM2038 vs SL BPN/NX for the treatment of patients with OUD who inject opioids. 2018. 155. Lofwall MR, Frost M, Bailey GL, et al. Phase 3 trial comparing depot buprenorphine (CAM2038) to sublingual buprenorphine/naloxone for OUD treatment: Subanalysis among persons with heroin as the primary opioid of abuse. 2018. 156. Haight B, Andorn A, Laffont C, et al. RBP-6000 buprenorphine monthly depot demonstrates sustained clinical efficacy and safety in phase III opioid use disorder trials. Neuropsychopharmacology. 2017;43:S463-S464. 157. 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. 158. Hourly Earnings (MEI). Organisation for Economic Co-operation and Development; 2018. https://stats.oecd.org/Index.aspx?DataSetCode=EAR_MEI#. Accessed August 20, 2018. 159. Databases, Tables & Calculators by Subject. United States Department of Labor; 2018. https://www.bls.gov/data/#prices. Accessed August 20, 2018. ©Institute for Clinical and Economic Review, 2018 Page 102 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents APPENDICES ©Institute for Clinical and Economic Review, 2018 Page 103 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Appendix A. Search Strategies and Results Table A1. PRISMA 2009 Checklist # Checklist item Pages TITLE Title 1 Identify the report as a systematic review, meta-analysis, or both. p. 40 ABSTRACT Provide a structured summary including, as applicable: background; objectives; data sources; study Structured 2 eligibility criteria, participants, and interventions; study appraisal and synthesis methods; results; p. 1 - 11 Summary 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. P. 21 Provide an explicit statement of questions being addressed with reference to participants, interventions, Objectives 4 p. 9 - 11 comparisons, outcomes, and study design (PICOS). METHODS Protocol and Indicate if a review protocol exists, if and where it can be accessed (e.g., Web address), and, if available, 5 p. 23 Registration provide registration information including registration number. Eligibility Specify study characteristics (e.g., PICOS, length of follow-up) and report characteristics (e.g., years 6 p. 9 – 11, p. 22-23 Criteria considered, language, publication status) used as criteria for eligibility, giving rationale. Information Describe all information sources (e.g., databases with dates of coverage, contact with study authors to 7 p. 22 - 23 Sources identify additional studies) in the search and date last searched. Present full electronic search strategy for at least one database, including any limits used, such that it Search 8 Appendix Table A2 could be repeated. Study State the process for selecting studies (i.e., screening, eligibility, included in systematic review, and, if 9 p. 25 - 31 Selection applicable, included in the meta-analysis). Data Describe method of data extraction from reports (e.g., piloted forms, independently, in duplicate) and Collection 10 p. 23 - 24 any processes for obtaining and confirming data from investigators. Process List and define all variables for which data were sought (e.g., PICOS, funding sources) and any Data Items 11 p. 10 - 14 assumptions and simplifications made. Risk of Bias in Describe methods used for assessing risk of bias of individual studies (including specification of whether Individual 12 this was done at the study or outcome level), and how this information is to be used in any data p. 23 Studies synthesis. ©Institute for Clinical and Economic Review, 2018 Page 104 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Summary 13 State the principal summary measures (e.g., risk ratio, difference in means). p. 21 Measures Synthesis of Describe the methods of handling data and combining results of studies, if done, including measures of 14 NA Results consistency (e.g., I2) for each meta-analysis. Risk of Bias Specify any assessment of risk of bias that may affect the cumulative evidence (e.g., publication bias, 15 p. 23 Across Studies selective reporting within studies). Additional Describe methods of additional analyses (e.g., sensitivity or subgroup analyses, meta-regression), if done, 16 NA Analyses indicating which were pre-specified. RESULTS Study Give numbers of studies screened, assessed for eligibility, and included in the review, with reasons for 17 Appendix Fig. A1 Selection exclusions at each stage, ideally with a flow diagram. Study For each study, present characteristics for which data were extracted (e.g., study size, PICOS, follow-up 18 p. 25 - 29 Characteristics 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 item 12). p. 33 - 45 Within Studies Results of For all outcomes considered (benefits or harms), present, for each study: (a) simple summary data for Individual 20 p. 33 - 45 each intervention group (b) effect estimates and confidence intervals, ideally with a forest plot. Studies Synthesis of 21 Present results of each meta-analysis done, including confidence intervals and measures of consistency. NA Results Risk of Bias 22 Present results of any assessment of risk of bias across studies (see Item 15). p. 23 Across Studies Additional Give results of additional analyses, if done (e.g., sensitivity or subgroup analyses, meta-regression [see 23 NA Analysis Item 16]). DISCUSSION Summary of Summarize the main findings including the strength of evidence for each main outcome; consider their 24 p. 33 - 52 evidence relevance to key groups (e.g., healthcare providers, users, and policy makers). Discuss limitations at study and outcome level (e.g., risk of bias), and at review-level (e.g., incomplete Limitations 25 p. E9 – ES10 retrieval of identified research, reporting bias). Provide a general interpretation of the results in the context of other evidence, and implications for Conclusions 26 p. 52 - 55 future research. FUNDING Describe sources of funding for the systematic review and other support (e.g., supply of data); role of Funding 27 p. iv - v 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 105 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Table A2. Search Strategy of Medline 1996 to Present with Daily Update, PsycINFO, Cochrane Central Register of Controlled Trials via Ovid, September 25, 2018. # Search Terms 1 opioid related disorder*.mp. (narcotic* or opiate* or opioid* or heroin) and (misuse or abus* or addict* or habit* or 2 dependenc* or withdraw).ti,ab. 3 1 or 2 4 (buprenorphine or Sublocade).mp. 5 (buprenorphine implant or Probuphine).mp. 6 (buprenorphine or CAM2038).mp. 7 (naltrexone or Vivitrol).mp. 8 (extended release or slow release or controlled release or sustained release).mp. 9 (4 or 6 or 7) and 8 10 (medication assisted treatment or (medication adj3 addiction treatment) or MAT).mp. 11 5 or 9 or 10 12 3 and 11 clinical trial.pt. or clinical trial, phase I.pt. or clinical trial, phase ii.pt. or clinical trial, phase iii.pt. or clinical trial, phase iv.pt. or controlled clinical trial.pt. or multicenter study.pt. or randomized controlled trial.pt. or double-blind method/ or clinical trials as topic/ or clinical trials, phase i as topic/ or clinical trials, phase ii as topic/ or clinical trials, phase iii as topic/ or clinical trials, phase 13 iv as topic/ or controlled clinical trials as topic/ or randomized controlled trials as topic/ or early termination of clinical trials as topic/ or multicenter studies as topic/ or ((randomi?ed adj7 trial*) or (controlled adj3 trial*) or (clinical adj2 trial*) or ((single or doubl* or tripl* or treb*) and (blind* or mask*))).ti,ab,kw. or (4 arm or four arm).ti,ab,kw. cohort studies/ or longitudinal studies/ or follow-up studies/ or prospective studies/ or retrospective studies/ or cohort.ti,ab. or longitudinal.ti,ab. or prospective.ti,ab. or 14 retrospective.ti,ab. or case-control studies/ or control groups/ or matched-pair analysis/ or retrospective studies/ or ((case* adj5 control*) or (case adj3 comparison*) or control group*).ti,ab,kw. 15 13 or 14 16 12 and 15 (abstract or addresses or autobiography or bibliography or biography or clinical trial, phase I or comment or congresses or consensus development conference or duplicate publication or 17 editorial or guideline or in vitro or interview or lecture or legal cases or legislation or letter or news or newspaper article or patient education handout or periodical index or personal narratives or portraits or practice guideline or review or video audio media).pt. 18 16 not 17 19 (animals not (humans and animals)).sh. 20 18 not 19 21 limit 20 to english language 22 Remove duplicates from 21 ©Institute for Clinical and Economic Review, 2018 Page 106 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Table A3. Embase Search Strategy, September 25, 2018. # Search Terms #1 ‘opiate addiction’/exp or ‘opiate addiction’ (narcotic* or opiate* or opioid* or heroin) and (misuse or abus* or addict* or habit* or #2 dependenc* or withdraw) #3 ‘drug abuse’ and ‘substance abuse’ #4 #1 or #2 or #3 #5 (buprenorphine or Sublocade) #6 (‘buprenorphine implant’ or Probuphine) #7 (buprenorphine or CAM2038) #8 (naltrexone or Vivitrol) #9 (extended or slow or controlled or sustained) and release #10 (#5 or #7 or #8) and #9 #11 (medication assisted treatment or medication NEAR/3 addiction treatment or MAT) #12 #6 or #10 or #11 #13 #4 and #12 #14 ‘clinical’:ab,ti AND ‘trial’:ab,ti OR ‘clinical trial’/exp OR random* OR ‘drug therapy’:lnk 'clinical article'/exp OR 'controlled study'/exp OR 'major clinical study'/exp OR 'prospective #15 study'/exp OR 'cohort analysis'/exp OR 'cohort':ab,ti OR 'compared':ab,ti OR 'groups':ab,ti OR 'case control':ab,ti OR 'multivariate':ab,ti #16 #14 or #15 #17 #13 and #16 #17 AND (‘chapter’/it OR ‘conference review’/it OR ‘editorial’/it OR ‘letter’/it OR ‘note’/it OR #18 ‘review’/it OR ‘short survey’/it) #19 #17 not #18 #20 ‘animal’/exp or ‘nonhuman’/exp or ‘animal experiment’/exp #21 ‘human’/exp #22 #20 and #21 #23 #20 not #22 #24 #19 not #23 #25 #24 and [english]/lim #26 #24 and [medline]/lim #27 #25 not #26 ©Institute for Clinical and Economic Review, 2018 Page 107 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Figure A1. PRISMA Flow Chart Showing Results of Literature Search 630 references identified 8 references identified through literature search through other sources 626 references after duplicate removal 626 references screened 450 citations excluded 153 citations excluded 176 references assessed 14 Population for eligibility in full text 19 Intervention 2 Comparator 26 Outcomes 89 Study/publication type 3 Not retrievable 23 total references 11 RCTs (2 secondary publications) 3 OLEs 2 Observational studies 15 publications, 3 clinicaltrials.gov results page, 5 conference abstracts ©Institute for Clinical and Economic Review, 2018 Page 108 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Appendix B. Previous Systematic Reviews and Technology Assessments We identified two systematic reviews and one technology assessment on the treatment of OUD using naltrexone: (1) the induction and adherence rates of naltrexone (XR-NTX, Vivitrol) in patients with OUD, (2) the effectiveness of sustained-release naltrexone and its adverse events, and (3) a NICE health technology appraisal on naltrexone as a treatment option for the management of OUD. These reviews and technology assessment are summarized below. Jarvis, B., Et al. (2018). “Extended-Release Injectable Naltrexone for Opioid Use Disorder: A Systematic Review” Cochrane Database of Systematic Reviews This systematic review evaluated the success of introducing naltrexone, patient’s adherence to treatment and its overall efficiency as a treatment option for patients with OUD. Thirty-four studies met the inclusion criteria as peer-reviewed studies with patients who were considered for treatment for opioid use, met the criteria for opioid abuse, or have OUD but were not required to be hospitalized. The pooled results from all studies indicated that the efficiency of naltrexone was lowest when used in patients who did not yet undergo detoxification (62.6% to 85% success rate, respectively). Investigational studies found higher rates of adherence to the treatment (47%), whereas medical records indicated that only 10.5% adhered to treatment outside of a trial setting. The study concluded that extended-release naltrexone is not clinically significant because the need for patient detoxification significantly lowers the pool of patients eligible to complete the treatment successfully. By six months, only 47% of participants were still adhering to the treatment. Lobmaier, P. Et al. (2008). “Sustained-Release Naltrexone for Opioid Dependence (Review)” Cochrane Database of Systematic Reviews The purpose of this systematic review was to evaluate the effectiveness of sustained-release naltrexone injection and its adverse effects separately in participants with OUD, participants with alcohol use disorder, and healthy participants. RCTs were included in the review and for the evaluations of the efficacy and safety of naltrexone injection. Researchers concluded that not enough reports exist to evaluate the effectiveness of naltrexone injection. One included trial found that the naltrexone injection’s effectiveness was dependent on dose. The high-dose treatment group in the study took a longer amount of time before they dropped out of treatment as compared to the low-dose or placebo group. When evaluating the amount and severity of adverse events (AEs), participants with OUD reported feeling fatigued and having administration-site specific conditions. Six out of ten participants with OUD in one trial, Waal 2003, reported dysphoria but the trial had no control group. In a separate trial (Waal 2006) patients with opioid dependence reported irritability, headache, and nausea, but this decreased as the study continued. Researchers ©Institute for Clinical and Economic Review, 2018 Page 109 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents concluded that there is not enough evidence to evaluate the effectiveness of sustained-release naltrexone as a treatment for OUD. NICE Health Technology Appraisal NICE issued a technology appraisal for naltrexone as a treatment for the management of OUD. In the assessment of naltrexone’s clinical effectiveness, researchers found 17 studies on the clinical effectiveness of naltrexone: one systematic review, 13 randomized-controlled trials (RCTs) and 3 non-randomized comparative studies. Two RCTs were conducted in a prison setting with follow-ups ranging from 20 days to a full year. All pooled studies mainly focused on reporting retention rates, relapse of opioid use, and re-incarceration in the case of the prison RCTs. Pooled analysis of the relapse rates showed a statistically significant reduction in risk of opioid use with naltrexone as compared to placebo. NICE researchers assessed that the pooled data confirmed naltrexone use showed a significant reduction in relapse. However, there was no difference in retention to treatment with naltrexone as opposed to other treatments, nor was there a significant reduction in mortality of patients being treated with naltrexone. ©Institute for Clinical and Economic Review, 2018 Page 110 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Appendix C. Ongoing Studies Appendix Table C. Ongoing Studies of Partial Opioid Agonists and Full Opioid Antagonist Title, Trial Sponsor, Estimated ClinicalTrials.gov Study Design Treatment Arms Patient Population Key Outcomes Completion Identifier Date Vivitrol A Strategy to Improve Phase III, Experimental: Regimen 1 Inclusion Criteria: • Percent of patients August 1, 2019 Succes of Treatment Randomized 1. A documented history of treatment successfully transitioned off • Rapid Monday to Friday Dicontinuation in trial, parallel with buprenorphine or buprenorphine oral naltrexone-induction Buprenorphine assignment buprenorphine/naloxone for at least • Percent of patients abstinent procedure Reponders 6 months with sustained abstinence from any opioids at 25-week • Intervention: Drug: Enrollement : 60 from illicit opioids for at least 3 trial endpoint Vivitrol New York State (currently months. Psychiatric Institute recruiting) 2. Aged 18 to 60 years Experimental: Regimen 2 3. In otherwise good health NCT03232346 • 5-week buprenorphine 4. Seeking buprenorphine taper from maintenance discontinuation and willing to accept dose of 8, 6, or 4mg randomization to either taper from buprenorphine or injection • Intervention: Drug: naltrexone Buprenorphine Exclusion Criteria: 1. Lifetime history of DSM-5 diagnosis of schizophrenia, schizoaffective disorder, or bipolar disorder 2. Individuals who meet DSM-5 criteria for any substance use disorders - severe, other than opioid and nicotine use disorder. 3. A recent history of binge-use of alcohol or sedative-hypnotics ©Institute for Clinical and Economic Review, 2018 Page 111 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Title, Trial Sponsor, Estimated ClinicalTrials.gov Study Design Treatment Arms Patient Population Key Outcomes Completion Identifier Date Injectable Phase III, Experimental: Vivitrol (XR- Inclusion Criteria: • Compare outcomes of the February, 2019 Pharmacotherapy for randomized, NTX) 1. Be at least 18 years of age or older, three intervention groups, Opioid Use Disorders parallel • 50 participants will be 2. Meet criteria for DSM-5 opioid use measured by a combination of (IPOD) assignment randomized to the long- disorders self-reports and urine drug acting naltrexone 3. Be detained for at least 48 hours, screens for opioids at 6- University of California, Enrollement : months post-intervention. condition (XR-NTX) which 4. Have an expected release date Berkeley 151 will include monthly within one year, injections of study drug. 1. Plan to reside in area after release. NCT02110264 Experimental: XR-NTX+PN Exclusion Criteria: • 50 participants will be 1. Have a medical or psychiatric randomized to receive condition that would make long-acting naltrexone participation unsafe in the judgment (XR-NTX) and will be of the medical staff or the PI, assigned to a patient 2. Have current or chronic pain or have navigator (PN). plans to undergo pain Active Comparator: ETAU treatment/therapy, 3. Have known sensitivity to naltrexone • 50 participants will be or naloxone, randomized to the drug- 4. Have participated in an education/treatment-as- investigational drug study within the usual group. past 30 days prior to screening, ©Institute for Clinical and Economic Review, 2018 Page 112 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Title, Trial Sponsor, Estimated ClinicalTrials.gov Study Design Treatment Arms Patient Population Key Outcomes Completion Identifier Date • Intervention: Behavioral: 5. Have a current pattern of alcohol, ETAU benzodiazepine, or other depressant or sedative hypnotic use, as determined by the study physician which would preclude safe participation in the study. Long-Acting Naltrexone Phase III, Experimental: Vivitrol at place Inclusion Criteria: • treatment adherence August 2020 for Pre-release randomized, of residence 1. Adult male or female inmate at MTC, • XR-NTX+ MMTx vs. XR-NTX- Prisoners paralell OTx following release from • One injection of long- BPRU, JPRU, BCCC, or MCIW and be assignment prison acting naltrexone (XR- eligible for release within 30 days Friends Research • Any illicit opioid used NTX) in prison, followed 2. History of opiate disorder • re-arrest [ Time Frame: 12- Institute, Inc. Enrollment : 240 by 6 monthly injections 3. Suitability for XR-NTX treatment months following release post-release at the 4. Currently opioid-free by history, with from prison ] NCT02867124 participants's place of negative urine for all opioids and no • re-incarceration [ Time Frame: residence utilizing mobile signs of opiate withdrawal 12-months following release 5. Willingness to enroll in XR-NTX from prison ] medical treatment • criminal activity [ Time Frame: Interventions: Drug: XR-NTX treatment 1,2,3,4,5,6,7 and 12-months Other: place of residence Exclusion Criteria: following release from prison Active Comparator: Vivitrol at 1. Liver function test levels greater ] opioid treatment program than three times normal • Injection drug use and HIV 2. Active medical illness that may make sexual risk factor [Time participation hazardous Frame: 6 and 12-months ©Institute for Clinical and Economic Review, 2018 Page 113 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Title, Trial Sponsor, Estimated ClinicalTrials.gov Study Design Treatment Arms Patient Population Key Outcomes Completion Identifier Date • One injection of long- 3. Untreated psychiatric disorder that following release from prison acting naltrexone (XR- may make participation ] NTX) in prison, followed 4. History of allergic reaction to XR-NTX by 6 monthly injections 5. Creatinine above normal limits post-release at a 6. Suicidal ideation (within the past 6- community opioid months) treatment program. 7. Body Mass Index (BMI) > 40 Interventions: Drug: XR-NTX 8. Unadjudicated charges that may result in transfer to another facility Other: opioid treatment and/or additional prison time. program Long Acting Naltrexone Phase IV, non- Experimental: Extended Inclusion: • Perceived change in recovery August 2023 for Opioid Addiction: randomized, release naltrexone 1. 18-65 years old through week 52 Focus on Sustained parallel • Perceived change in quality of • 380 mg every 4 weeks 2. Meets criteria for DSM-5 opioid use Abstinence and assignment, life through week 52 No Intervention: Treatment disorder Recovery (NaltRec) open-label 3. Completing a stay in a controlled As Usual (TAU) environment with restricted access Lars Tanum, MD Enrollment: 300 • Daily sublingual to substances of abuse with a buprenorphine in flexible minimum duration of seven days NCT03647774 dose Exclusion: 1. Severe psychiatric disorder 2. Alcoholism defined by the criteria in DSM-5 Source: www.ClinicalTrials.gov (NOTE: studies listed on site include both clinical trials and observational studies) ©Institute for Clinical and Economic Review, 2018 Page 114 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Appendix D. Comparative Clinical Effectiveness Supplemental Information We performed screening at both the abstract and full-text level. Two reviewers independently screened the abstracts and full-texts of studies identified through electronic searches according to the inclusion and exclusion criteria described earlier using DistillerSR (Evidence Partners, Ottawa, Canada), with any incongruencies resolved through consensus. We did not exclude any study at abstract-level screening due to insufficient information. For example, an abstract that did not report an outcome of interest would be accepted for further review in full text. We retrieved the citations that were accepted during abstract-level screening for full text appraisal. Each full-text was independently reviewed by two reviewers and conflicts resolved by a third reviewer. Reasons for exclusion were categorized according to the PICOTS elements during full-text review. We used criteria published by the US Preventive Services Task Force (USPSTF) to assess the quality of RCTs and comparative cohort studies, using the categories “good,” “fair,” or “poor” (see Appendix Table D3).74Guidance for quality ratings using these criteria is presented below, as is a description of any modifications we made to these ratings specific to the purposes of this review. Good: Meets all criteria: Comparable groups are assembled initially and maintained throughout the study; reliable and valid measurement instruments are used and applied equally to the groups; interventions are spelled out clearly; all important outcomes are considered; and appropriate attention is paid to confounders in analysis. In addition, intention to treat analysis is used for RCTs. Fair: Studies were graded "fair" if any or all of the following problems occur, without the fatal flaws noted in the "poor" category below: Generally comparable groups are assembled initially but some question remains whether some (although not major) differences occurred with follow-up; measurement instruments are acceptable (although not the best) and generally applied equally; some but not all important outcomes are considered; and some but not all potential confounders are addressed. Intention to treat analysis is done for RCTs. Poor: Studies were graded "poor" if any of the following fatal flaws exists: Groups assembled initially are not close to being comparable or maintained throughout the study; unreliable or invalid measurement instruments are used or not applied equally among groups (including not masking outcome assessment); and key confounders are given little or no attention. For RCTs, intention to treat analysis is lacking. Note that case series are not considered under this rating system – because of the lack of comparator, these are generally considered to be of poor quality. ©Institute for Clinical and Economic Review, 2018 Page 115 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents ICER Evidence Rating We used the ICER Evidence Rating Matrix (see Figure D1) to evaluate the evidence for a variety of outcomes. The evidence rating reflects a joint judgment of two critical components: a) 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 b) The level of certainty in the best point estimate of net health benefit.75 Figure D1. ICER Evidence Rating Matrix ©Institute for Clinical and Economic Review, 2018 Page 116 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Table D1. Key Baseline Characteristics of Included Studies Prescription Mean (SD) Years Heroin as Mean (SD) Mean (SD) Male, White, Employed, Recent Opioid Drugs as IV Drug of Opioid Study Arm N Primary Age at First Age % % % Use, % Primary Use, % Use/OUD Opioid, % Opioid Use Opioid, % Diagnosis CAM 2038 4.3 (7.8) since CAM2038 213 38.7 (11.2) 56.8 74.6 35.7 NR 71.4 28.6 53.5 NR Lofwall diagnosis 201876 4.7 (6.0) since SL bup/nal 215 38.0 (10.9) 66.0 76.3 33.5 NR 70.2 29.8 51.2 NR diagnosis 31.6 3.06 (5.29) since CAM2038 114 37.3 (11.6) 54.4 78.9 29.8 tested positive 85.1 NR 100 NR CPDD diagnosis for fentanyl Injection 21.8 Poster154 2.76 (3.54) since SL bup/nal 110 37.2 (11.5) 66.4 82.7 25.5 tested positive 86.4 NR 100 NR diagnosis for fentanyl 35.5 4.03 (8.74) since CAM2038 152 38.9 (11.2) 56.6 66.4 27.6 tested positive 100 0 63.8 NR CPDD diagnosis for fentanyl Heroin 25.8 Poster155 3.19 (4.67) since SL bup/nal 151 38.9 (11.4) 69.5 69.5 27.2 tested positive 100 0 62.9 NR diagnosis for fentanyl Sublocade 21.7% (18-29), Sublocade 43.3% (30-44), 203 67.0 69.0 NR NR NR NR NR NR NR 300mg/100mg 31.5% (45-59), 3.4% (60+) Trial 13- 000177 22.4% (18-29), Sublocade 47.3% (30-44), 201 67.2 71.6 NR NR NR NR NR NR NR 300mg/300mg 26.4% (45-59), 4.0% (60+) ©Institute for Clinical and Economic Review, 2018 Page 117 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Prescription Mean (SD) Years Heroin as Mean (SD) Mean (SD) Male, White, Employed, Recent Opioid Drugs as IV Drug of Opioid Study Arm N Primary Age at First Age % % % Use, % Primary Use, % Use/OUD Opioid, % Opioid Use Opioid, % Diagnosis 23.0% (18-29), 45.0% (30-44), Placebo 100 65.0 78.0 NR NR NR NR NR NR NR 30.0% (45-59), 2.0% (60+) Trial 13- De Novo 412 38.4 (12.10) 63.8 71.6 NR NR NR NR NR NR NR 000378 (Study following Roll-Over 257 41.6 (11.07) 65.8 64.9 NR NR NR NR NR NR NR 13-0001) Probuphine 6.2 (5.93) since Probuphine 87 38 (11.2) 59.8 94.3 78.1 NR 17.2 75.9 NR NR Rosenthal diagnosis 201679 6.2 (6.95) since SL bup/nal 89 39 (10.8) 58.4 95.5 72.0 NR 24.7 73.0 NR NR diagnosis Probuphine 114 36.4 (11.0) 63.2 83.3 NR NR 66.7 33.3 NR NR NR Rosenthal Placebo 54 35.2 (10.3) 57.4 83.3 NR NR 51.9 48.1 NR NR NR 201381 implants SL bup/nal 119 35.3 (10.9) 60.5 81.5 NR NR 63.0 36.1 NR NR NR Probuphine 108 35.8 (11.0) 66.7 75.9 NR NR 63.9 36.1 NR NR NR Ling 201080 Placebo 55 39.3 (11.7) 72.7 72.7 NR NR 61.8 38.2 NR NR NR implants Vivitrol 21.2 (6.5) 12.8 (9.0) opioid Vivitrol 283 34 (9.5) 69.0 73.0 17.0 NR 81.0 15.0 63.0 opioid use use ©Institute for Clinical and Economic Review, 2018 Page 118 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Prescription Mean (SD) Years Heroin as Mean (SD) Mean (SD) Male, White, Employed, Recent Opioid Drugs as IV Drug of Opioid Study Arm N Primary Age at First Age % % % Use, % Primary Use, % Use/OUD Opioid, % Opioid Use Opioid, % Diagnosis Lee 21.4 (7.6) 12.2 (9.0) opioid 201883 (X- SL bup/nal 287 33.7 (9.8) 72.0 75.0 20.0 NR 81.0 16.0 64.0 opioid use use BOT) prior 30 days: 8.9 (7.8) heavy 7.6 days using opioid use; Vivitrol 80 36.4 (8.8) 76.3 90.0 NR heroin; NR NR 90.0 NR 6.9 (5.8) heroin 8.2 days using use other opioids Tanum prior 30 days: 201784 12.0 days 9.6 (10.5) heavy using heroin; opioid use; SL bup/nal 79 35.7 (8.5) 68.4 88.6 NR NR NR 81.0 NR 14.5 days 6.7 (5.2) heroin using other use opioids 22.5 6.7 36.2 Continuing (95%CI (95%CI 54 (95% CI: 33.9, 81.5 NR NR NR NR NR Solli Vivitrol 21.1, 24.0) 5.5, 7.8) 38.4) 201889 heroin use heroin use 63.2 (Tanum 21.4 6.7 35.1 2017 OLE) Inducted on (95%CI (95%CI 63 (95% CI 32.9, 71.4 NR NR NR NR NR Vivitrol 19.5, 23.4) 5.2, 8.1) 37.2) heroin use heroin use prior 30 days: 42.1 Vivitrol 153 44.4 (9.2) 84.3 20.4 17.0 30.9 any NR NR during NR NR Lee opioids lifetime 201692 prior 30 days: 40.0 Treatment as 155 43.2 (9.4) 85.2 19.4 18.7 38.1 any NR NR during NR NR usual opioids lifetime ©Institute for Clinical and Economic Review, 2018 Page 119 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Prescription Mean (SD) Years Heroin as Mean (SD) Mean (SD) Male, White, Employed, Recent Opioid Drugs as IV Drug of Opioid Study Arm N Primary Age at First Age % % % Use, % Primary Use, % Use/OUD Opioid, % Opioid Use Opioid, % Diagnosis 7-days pre- arrest: 13.0 44.0 40 Vivitrol 16 100 NR 31.0 prescription NR NR during NR NR [range 26-52] drug; 94.0 lifetime Lee heroin 201591 7-days pre- arrest: 18.0 24.0 Treatment as 47 17 100 NR 12.0 prescription NR NR during NR NR usual [range 39-58] drug; 100.0 lifetime heroin prior 30 days: 9.1 (4.5) since Vivitrol 126 29.4 (4.8) 90.0 98.0 NR 88.0 heroin; NR NR NR NR dependence Krupitsky 12.0 201186 methadone; 10.0 (3.9) since Placebo 124 29.7 (3.6) 86.0 100 NR 13.0 other NR NR NR NR dependence opioids prior 30 days: Continuing 9.0 (4.2) since Krupitsky 67 29.5 (5.0) 92.5 100 NR 89.5 heroin; NR NR NR NR Vivitrol dependence 201390 8.8 (Krupitsky methadone; 2011 OLE) Inducted on 9.8 other 9.4 (4.0) since 47 29.4 (3.8) 85.1 100 NR NR NR NR NR Vivitrol opioids dependence 20.1 (11.2) heroin use; Vivitrol 66 46.6 (8.3) 83.3 NR NR NR NR NR NR NR 2.8 (7.2) other NEW opioid use HOPE88 18.4 (10.2) Placebo 27 43.9 (7.8) 77.8 NR NR NR NR NR NR NR heroin use; ©Institute for Clinical and Economic Review, 2018 Page 120 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Prescription Mean (SD) Years Heroin as Mean (SD) Mean (SD) Male, White, Employed, Recent Opioid Drugs as IV Drug of Opioid Study Arm N Primary Age at First Age % % % Use, % Primary Use, % Use/OUD Opioid, % Opioid Use Opioid, % Diagnosis 3.2 (5.4) other opioid use Observational Naltrexone 1041 29.9 (10.93) 69.0 NR NR NR NR NR NR NR NR Shah Nonpharmaco- 201895 6883 33.2 (13.43) 38.9 NR NR NR NR NR NR NR NR logical therapy 95% CI: 95% confidence interval; bup/nal: buprenorphine/naloxone; mg: milligram; N: number of participants; NR: not reported; OLE: open-label extension; OUD: opioid use disorder; SD: standard deviation; SL: sublingual ©Institute for Clinical and Economic Review, 2018 Page 121 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Table D2. Study Designs of Included Studies Study/ Detox Intervention Total Location Inclusion NCT/ Centers Funding Period Induction Period Period Follow-Up Exclusion Criteria of Sites Criteria Phase (Days) (Weeks) (Weeks) CAM 2038 Pharmacotherapy for OUD Lofwall 201876 DSM-5 criteria Industry 1 day of 4 mg within 60 days; NCT02651584 Multicenter US 0 24 28 for moderate (Braeburn) bup/1 mg nal AIDS; chronic pain requiring Phase III or severe OUD opioid therapy Sublocade Condition requiring chronic Open-label run-in opioid treatment; induction phase substance use disorder Trial 13-000177 with SL bup/nal for DSM-5 criteria (DSM-5) with regard to any NCT02357901 Industry 3 days followed by for moderate Multicenter US 0 24 24 substances other than Phase III (Indivior) a 4- to 11-day SL or severe OUD opioids, cocaine, cannabis, Unpublished bup/nal open-label for 3 months tobacco, or alcohol; run-in dose- received MAT for OUD in adjustment period prior 90 days ©Institute for Clinical and Economic Review, 2018 Page 122 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Study/ Detox Intervention Total Location Inclusion NCT/ Centers Funding Period Induction Period Period Follow-Up Exclusion Criteria of Sites Criteria Phase (Days) (Weeks) (Weeks) De novo subjects: Condition requiring chronic De novo opioid treatment; subjects: substance use disorder Open-label run-in Trial 13-000378 DSM-5 criteria (DSM-5) with regard to any induction phase NCT02510014 De Novo: for moderate substances other than with SL bup/nal for Phase III 49 weeks or severe OUD opioids, cocaine, cannabis, Industry 3 days followed by Unpublished Multicenter US 0 NR for 3 months tobacco, or alcohol; (Indivior) a 4- to 11-day SL (Study Roll-Over: received MAT for OUD in bup/nal open-label following 13- 25 weeks Roll-over prior 90 days run-in dose- 0001) subjects: adjustment period Completed Roll over subjects: trial 13-0001 Major protocol deviations or adverse events in trial 13-0001 Probuphine DSM-IV diagnosis of opioid Stable dose of 8 Chronic pain requiring Rosenthal dependence mg/day or less of opioids; AIDS; primary 2016 Industry and no Multicenter US 0 SL buprenorphine 24 24 diagnosis of substance NCT02180659 (Braeburn) evidence of for at least 24 dependence other than Phase III opioid weeks opioids or nicotine withdrawal or illicit opioid- positive urine Rosenthal Government Open-label DSM-IV AIDS; current dependence 201381 (NIDA) and induction phase diagnosis of on psychoactive substances Multicenter US 0 24 24 NCT01114308 Industry with 12-16 mg/day current opioid other than opioids or Phase III (Titan) of SL bup/nal for dependence nicotine (DSM-IV); received ©Institute for Clinical and Economic Review, 2018 Page 123 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Study/ Detox Intervention Total Location Inclusion NCT/ Centers Funding Period Induction Period Period Follow-Up Exclusion Criteria of Sites Criteria Phase (Days) (Weeks) (Weeks) three days within MAT for OUD in prior 90 10 days of days; current diagnosis of screening chronic pain that required opioid treatment AIDS; current dependence Open-label on psychoactive substances induction phase DSM-IV other than opioids or Ling 201080 with 12-16 mg/day Industry diagnosis of nicotine (DSM-IV); received NCT 00447564 Multicenter US 0 of SL bup/nal for 24 24 (Titan) current opioid MAT for OUD in prior 90 Phase III three days within dependence days; current diagnosis of 10 days of chronic pain that required screening opioid treatment Vivitrol Vivitrol arm: DSM-5 criteria ≥3 days since last for OUD; used Protocols opioid use and Serious medical, psychiatric Lee 201883 opioids other and length pass naloxone disorder, or other (X-BOT) Government than Multicenter US of stay challenge (≥0.4 24 36 substance use disorder; NCT02032433 (NIDA) prescribed varied by mg) chronic pain requiring Phase IV within 30 days site opioids prior to SL bup/nal arm: consent Varied Government Vivitrol arm: (Research ≥3 days since last DSM-IV Other drug or alcohol Tanum 201784 Council of opioid use and ≥7 days criteria for dependence; serious NCT01717963 Multicenter Norway Norway and pass naloxone 12 48 opioid somatic or psychiatric Phase III the Western challenge (2-4 mg) dependence illness Norway Regional SL bup/nal arm: ©Institute for Clinical and Economic Review, 2018 Page 124 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Study/ Detox Intervention Total Location Inclusion NCT/ Centers Funding Period Induction Period Period Follow-Up Exclusion Criteria of Sites Criteria Phase (Days) (Weeks) (Weeks) Health Three to four-day Authority) dose titration to reach target dose Other drug or alcohol DSM-IV dependence requiring a criteria for high level of care; an opioid untreated psychiatric NA Lee 201692 Vivitrol arm: Pass dependence; disorder or medical Government (participants NCT00781898 Multicenter US naloxone challenge 24 78 had been condition; a current (NIDA) had to be Phase II/III (>0.8 mg) incarcerated; diagnosis of chronic pain opioid-free) opioid-free requiring opioids; drug status at overdose in the previous 3 randomization years requiring inpatient hospitalization Opioid- dependent adults Academic incarcerated in (NYU) and NA (partici- NYC DOC Chronic pain requiring Lee 201591 Vivitrol arm: Pass Industry pants had to meeting DSM- opioids; serious, NCT01180647 Multicenter US naloxone challenge 8 8 (Alkermes be opioid- IV criteria for uncontrolled medical or Phase III (0.8 mg) provided free) opioid psychiatric illnesses study drug) dependence prior to arrest; opioid-free at randomization Krupitsky DSM-IV AIDS-indicator disease; ≤30 days 201186 Industry criteria for psychosis, bipolar disorder, Multicenter Russia (pre-study 7 days 24 76 NCT00678418 (Alkermes) opioid major depressive disorder detox) Phase III dependence with suicidal ideation; ©Institute for Clinical and Economic Review, 2018 Page 125 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Study/ Detox Intervention Total Location Inclusion NCT/ Centers Funding Period Induction Period Period Follow-Up Exclusion Criteria of Sites Criteria Phase (Days) (Weeks) (Weeks) present dependence on substances other than opioids or heroin If recent opioid use Academic or DSM-IV for Prescription of opioid pain (Yale), anticipated opioid medications or expressing a NEW HOPE88 Government withdrawal, If recent opioid use dependence; need for them; already NCT01246401 (NIH), and five-day or anticipated confirmed HIV enrolled in an opioid Multicenter US 24 48 Phase I/II Industry buprenor- withdrawal, three- infection; substitution therapy Unpublished (Alkermes phine five days released from program; in opioid provided withdrawal prison within withdrawal (3-5 days since study drug) protocol ±30 days last opioid ingestion) was employed Observational MAT for opioid dependence Diagnosis of in one month prior opioid Shah 2018 95 Industry 24 excluded from the Multicenter US NR NR 12 months dependence Observational (Alkermes) months buprenorphine and (ICD-9 CM) for nonpharmacological 6 months therapy cohorts Bup/nal: buprenorphine/naloxone; DSM: Diagnostic and Statistical Manual of Mental Disorders; ICD-9 CM: International Classification of Diseases, Ninth Revision, Clinical Modification MAT: medication for addiction treatment; mg; milligram; NR: not reported; NYC DOC: New York City Department of Corrections; OUD: opioid use disorder; SL: sublingual ©Institute for Clinical and Economic Review, 2018 Page 126 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Table D3. Quality Ratings of Included Trials Clear Definition of Comparable Non-Differential Patient/Physician Interventions Clear Definition of Primary Handling of Study USPSTF Rating Groups Follow-Up Blinding (Including Outcomes Missing Urine Tests Initiation) CAM 2038 Lofwall 2018 76 Yes Yes Yes Yes Yes Considered positive Good Sublocade Trial 13-0001 77 Yes Yes Yes Yes Yes Considered positive Good Probuphine Data were randomly imputed with 20% Rosenthal 201679 Yes Yes Yes Yes Yes Good relative penalty against Probuphine Rosenthal 201381 Yes No Yes Yes Yes Considered positive Fair Ling 2010 80 Yes No Yes Yes Yes Considered positive Fair Vivitrol Lee 2018 (X-BOT) 83 Yes Yes No Yes Yes Considered positive Good Tanum 2017 84 Yes Yes No Yes Yes Considered positive Good Lee 201692 Yes Yes No No Yes Considered positive Fair Lee 201591 No Yes No Yes Yes Considered positive Fair Krupitsky 2011 86 Yes No Yes Yes Yes Considered positive Fair NEW HOPE 88 Yes Yes Yes Yes Yes Considered positive Good USPSTF: United States Preventive Services Task Force ©Institute for Clinical and Economic Review, 2018 Page 127 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Table D4. Key Abstinence and Relapse Outcomes in Included Studies I Relapse and Opioid Use Opioid-Negative Urine Samples Study Arm N Week B/W Arm Description Data B/W Arm Comparison Description Data Comparison CAM2038 CAM2038 213 35.1 (2.5) tx diff (95% CI) 6.7 (-0.1, 13.6) 1-24 NR SL bup/nal 215 28.4 (2.5) p(NI)<0.001 p(S) is NS CAM2038 213 mean (SE) % of negative 35.8 (2.6) tx diff (95% CI) 1-12 NR urine samples 5.9 (-1.3, 13.1) Lofwall 201876 SL bup/nal 215 29.9 (2.6) p is NS CAM2038 213 33.9 (2.6) tx diff (95% CI) 13-24 NR 8.5 (1.2, 15.7) SL bup/nal 215 25.4 (2.6) p=0.02 CAM2038 213 mean (SE) % of negative 35.1 (2.5)* 4-24 NR p=0.004 SL bup/nal 215 urine samples, CDF 26.7 (2.5)* CPDD Injection CAM2038 114 mean (SE) % of negative 30.9 (3.3)* 4-24 NR p<0.001 Poster154 SL bup/nal 110 urine samples, CDF 15.4 (2.7)* CPDD Heroin CAM2038 152 mean % of negative 29.9* 4-24 NR p<0.001 Poster155 SL bup/nal 151 urine samples, CDF 12.7* Sublocade Sublocade vs. placebo: 300mg/100 194 41* p<0.0001 mg # of participants with Trial 13-000177 Sublocade 5-24 NR ≥90% negative urine vs. placebo: 300mg/300 196 samples, CDF 48* p<0.0001 mg Placebo 99 2*  ©Institute for Clinical and Economic Review, 2018 Page 128 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Relapse and Opioid Use Opioid-Negative Urine Samples Study Arm N Week B/W Arm Description Data B/W Arm Comparison Description Data Comparison Roll-Over 257 25 % of participants with 28.8* Trial 13-0003 78 NR ≥90% negative urine NR De Novo 412 49 15.0* samples, CDF Probuphine vs. placebo: Probuphine 114 31.2 p<0.0001 Placebo 1-24 NR 54 13.4  implant SL bup/nal 119 33.5  vs. placebo: p<0.0001 Probuphine 114 39.6 vs. SL bup/nal: 1-16 NR p is NS mean % of negative Rosenthal 2013 81 Placebo 54 urine samples, CDF 17.9  implant SL bup/nal 119 37.8  vs. placebo: p<0.0001 Probuphine 114 28.9 vs. SL bup/nal 17-24 NR p is NS Placebo 54 7.2  implant SL bup/nal 119 29.6  36.6 (95% Probuphine 108 CI: 30.5, mean % of negative 42.6) Ling 201080 1-24 NR p=0.01 urine samples 22.4 (95% Placebo 55 CI: 15.3, implant 29.5) ©Institute for Clinical and Economic Review, 2018 Page 129 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Relapse and Opioid Use Opioid-Negative Urine Samples Study Arm N Week B/W Arm Description Data B/W Arm Comparison Description Data Comparison 40.4 (95% Probuphine 108 CI: 34.2, 46.7) 1-16 NR p=0.04 28.3 (95% Placebo 55 CI: 20.3, implant 36.3) Probuphine 108 NR Placebo 17-24 NR p<0.001 55 NR implant Vivitrol (1): n (%) (1): 185 participants (65) Vivitrol 283 4 (0-19) who (2): 8.4 (1): OR (95%CI) 1.44 relapsed† (3.0, 23.4) Lee 201883 (1.02, 2.01), p=0.036; median (IQR) weekly- 3-24 (2): median p<0.0001 (X-BOT) (1): 163 (2): HR (95%CI): 1.36 negative urine samples (IQR) (57) (1.10, 1.68), p=0.004 SL bup/nal 287 relapse-free 10 (3-20) (2): 14.4 survival (5.1, 23.4) weeks (1): 0.8 mean (SD) (1.5) days using (2): 1.2 tx diff (95%CI) Vivitrol 63 (1): heroin, (2.2) (1): -3.0 (-4.9, -1.2), (2): other (3): 2.96 p=0.001 Tanum 201784 4 illicit NR (1.32, 4.58) (2): -2.9 (-4.8, -0.9), opioids, and (1): 3.7 p=0.004 (3): IV drugs (7.4) (3): NR SL bup/nal 65 during (2): 4.2 preceding 4 (7.9) ©Institute for Clinical and Economic Review, 2018 Page 130 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Relapse and Opioid Use Opioid-Negative Urine Samples Study Arm N Week B/W Arm Description Data B/W Arm Comparison Description Data Comparison weeks (95% (3): 3.97 CI) (1.90, 5.22) (1): 0.8 (1.9) (2): 1.8 Vivitrol 59 (4.7) tx diff (95%CI) (3): 3.75 (1): -3.3 (-5.1, -1.5), (2.13, 5.34) p<0.001 8 NR (1): 4.4 (2): -2.6 (-4.6, -0.7), (9.1) p=0.007 (2): 4.0 (3): NR SL bup/nal 55 (8.5) (3): 4.08 (2.38, 5.72) (1): 1.1 (2.3) (2): 2.0 Vivitrol 57 (5.0) tx diff (95%CI) (3): 4.51 (1): -3.6 (-6, -1.2), (2.46, 6.61) p=0.003 12 NR (1): 4.1 (2): -2.4 (-4.9, 0.1), (8.4) p is NS (2): 4.4 (3): NR SL bup/nal 50 (8.7) (3): 4.56 (2.33, 6.68) Vivitrol 63 mean (SD) group 0.9 (0.3) tx diff (95%CI) 1-12 NR proportion of total # of 0.1 (-0.04, 0.2), SL bup/nal 65 negative samples 0.8 (0.4) p<0.001 ©Institute for Clinical and Economic Review, 2018 Page 131 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Relapse and Opioid Use Opioid-Negative Urine Samples Study Arm N Week B/W Arm Description Data B/W Arm Comparison Description Data Comparison (1): 0.1 mean (0.0, 0.2) (95%CI) days Continuing (2): 0.2 28 using (1): tx diff (95%CI) Vivitrol (-0.1, 0.4) heroin, (1): 0.3 (-0.5, 1.0), (3): 2.4 (2): other p=NS Solli 201889 (Tanum (-0.7, 5.5) 36 illicit (2): 0.7 (-0.1, 1.6), NR 2017 OLE) (1): 0.8 opioids, and p=0.088 (-0.3, 1.9) (3): IV drugs (3): 1.2 (-2.2, 4.6), Inducted (2): 0.6 30 during p is NS on Vivitrol (0.0, 1.2) preceding 4 (3): 6.0 weeks (2.2., 9.9) (1): median (1): 10.5 Vivitrol 153 weeks to (2): 43.1 74.1 relapse‡ (3): 5.9 (1): HR (95% CI) 0.49 (2): % of (0.36, 0.68), p<0.001 participants (2): OR (95% CI) 0.43 OR (95% CI) Lee 201692 24 who (0.28, 0.65), p<0.001 % of negative samples 2.30 (1.48, 3.54), (1): 5.0 Treatment relapsed‡ (3): OR (95% CI) 0.67 p<0.001 155 (2): 63.9 55.7 as usual (3) % of (0.25, 1.82), (3): 8.6 participants p is NS reporting IV drug use (1): % of (1): 38 OR (95%CI) Vivitrol 16 59 OR (95%CI) participants (2): 25 0.08 (0.01, 0.48), 4 3.5 (1.4, 8.5), Treatment who (1): 88 p<0.004 Lee 201591 17 % of negative samples 29 p<0.009 as usual relapsed§ (2): 6 (2): NR (2): % of (1): 50 Vivitrol 16 8 OR (95%CI) 59 OR (95%CI) participants (2): NR ©Institute for Clinical and Economic Review, 2018 Page 132 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Relapse and Opioid Use Opioid-Negative Urine Samples Study Arm N Week B/W Arm Description Data B/W Arm Comparison Description Data Comparison Treatment reporting IV (1): 93 0.13 (0.02, 0.78), 4.6 (2.1, 10), 17 24 as usual drug use (2): NR p<0.03 p<0.0001 % (95%CI) of 0.8 (0.0, Vivitrol 126 participants 2.3) relapsed to physiologica tx diff (95%CI) Krupitsky 201186 24 l 17.3 (2.3, 127.8), NR 13.7 (7.7, Placebo 124 dependence p<0.0001 19.8) (positive naloxone test) Continuing 73.7 67 Krupitsky 2013 90 Vivitrol mean (SD) % of negative (33.2)* 52 NR NR (Krupitsky 2011 OLE) Inducted monthly samples 81.0 47 on Vivitrol (28.6)* median 137 Vivitrol 51 [range] days [0 to 168] to first relapse NEW HOPE88 24 p=0.03 NR based on 29 Placebo 23 self- [0 to 168] reported opioid use ©Institute for Clinical and Economic Review, 2018 Page 133 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents *Urine test results confirmed with self-report; † Relapse defined as the use of non-study opioids any time after 20-day randomization (at the start of four consecutive opioid use weeks or at the start of seven consecutive days of self-reported opioid use days). A use week was defined as any week where the participant reported at least one day of non-study opioids (buprenorphine, methadone, morphine, heroin, codeine, oxycodone) or did not provide a urine sample; ‡ A relapse event was defined as 10 or more days of opioid use in a 28-day (four-week) period as assessed by self-report or by testing of urine samples obtained every two weeks; a positive or missing sample was computed as five days of opioid use; §Relapse defined as ≥10 of 28 days of self-reported opioid misuse following jail release or two or three positive of the three urine samples during weeks two, three and four. A single positive or missing urine result counted as seven opioid misuse days. 95% CI: 95% confidence interval; bup/nal: buprenorphine/naloxone; b/w: between; CDF: cumulative distribution function; HR: hazard ratio: IQR: interquartile range; IV: intravenous; mg: milligram: mo.: month(s); N: number of participants; NI: noninferiority; NR: not reported; NS: not significant; OR: odds ratio; SE: standard error; SD: standard deviation; tx diff: treatment difference ©Institute for Clinical and Economic Review, 2018 Page 134 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Table D5. Key Abstinence and Relapse Outcomes in Included Studies II Responders, n (%) Other Abstinence Outcomes Study Arm N Week B/W Arm B/W Arm Description Data Description Data Comparison Comparison CAM2038 no evidence of illicit CAM2038 213 opioid use at 37 (17.4)* tx diff (95%CI) Lofwall 2018 76 1-24 prespecified time 3.0 (-4.0, 9.9), NR points† assessed via p(NI)<0.001 SL bup/nal 215 31 (14.4)* urine tests CAM2038 114 18 (15.8)* tx diff (95%CI) CPDD Injection See Lofwall 201876 1-24 7.3 (0.2, 16.8) NR Poster154 description SL bup/nal 110 NR p=0.047 tx diff (95%CI) CAM2038 152 24 (15.8)* See Lofwall 2018 76 11.2 (4.5, CPDD Heroin Poster155 1-24 NR description 17.9) SL bup/nal 151 7 (4.6)* p<0.001 Sublocade Sublocade vs. placebo: mean (SE) # of vs. placebo: 194 55 (28.4)* 8.5 (0.68)* 300mg/100mg p<0.0001 weeks of p<0.0001 ≥80% of negative Sublocade 5-24 vs. placebo: abstinence vs. placebo: 196 urine samples 57 (29.1)* 8.5 (0.68)* 300mg/300mg p<0.0001 assessed via urine p<0.0001 Placebo 99 2 (2.0)* tests 1.0 (0.84)*  Trial 13-000177 Sublocade vs. placebo: 194 # of participants 71* 300mg/300mg p<0.0001 abstinent Sublocade 24 NR vs. placebo: 196 assessed via urine 87* 300mg/300mg p<0.0001 tests Placebo 99 2*  Probuphine ©Institute for Clinical and Economic Review, 2018 Page 135 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Responders, n (%) Other Abstinence Outcomes Study Arm N Week B/W Arm B/W Arm Description Data Description Data Comparison Comparison tx diff 8.8%; Probuphine 84 ≥4 of 6 mo. w/o 81 (96.4)* one-sided % of participants 85.7* tx diff (95%CI) evidence of illicit 97.5%CI abstinent over 13.8% (0.018, Rosenthal 201679 24 opioid use assessed (0.009, ∞); mo. 1-6 assessed 0.258), SL bup/nal 89 via urine tests 78 (87.6)* p(NI)<0.001 ; via urine tests 71.9* p=0.027 p(S)=0.03 Vivitrol Vivitrol 283 median (IQR) self- 39 (1, 144) Lee 201883 3-24 NR reported opioid- p<0.0001 (X-BOT) SL bup/nal 287 abstinent days 81 (16, 144) Continuing 54 % of patients 53.7 Solli 2018 (Tanum 89 Vivitrol 36 NR reporting NR 2017 OLE) Inducted on 63 abstinence 44.4 Vivitrol % of 2-week Vivitrol 153 71.1* intervals with no OR (95%CI) Lee 201692 24 NR opioid use as 2.50 (1.66, Treatment as 155 assessed via urine 49.5* 3.76), p<0.001 usual test Vivitrol 16 50* OR (95%CI) 4 NR % of participants 7.5 (1.3, 44), Treatment as 17 13* p<0.03 usual abstinent Lee 201591 assessed via urine Vivitrol 16 50* OR (95%CI) tests Treatment as 8 NR 16 (1.7, 151), 17 7* p<0.007 usual ©Institute for Clinical and Economic Review, 2018 Page 136 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Responders, n (%) Other Abstinence Outcomes Study Arm N Week B/W Arm B/W Arm Description Data Description Data Comparison Comparison (1): 90.0 (69.9, (1): median % 92.4)* Vivitrol 126 (95% CI) weeks tx diff (95% CI) (2): 35.7 with abstinence (1): 55.0 (15.9, (27.4, (2): % (95%CI) of 76.1), Krupitsky 201186 24 NR 44.1)* patients p=0.0002 (1): 35.0 abstinent (2): 1.58 (1.06, (11.4, assessed via urine 2.36), p=0.02 Placebo 124 63.8)* test (2): 22.6 (15.2, 29.9) (1): % of patients (1): 49.3* Continuing 67 abstinent (2): 80.6 Vivitrol assessed via urine (29.7) Krupitsky 201390 52 NR test NR (Krupitsky 2011 OLE) (1): 53.2* Inducted on (2): mean (SD) % 47 (2): 87.4 Vivitrol of reported (23.8) opioid-free days % of participants Vivitrol 66 19.7 abstinent NEW HOPE 88 24 NR NR assessed via urine Placebo 27 18.5 tests *Urine test results confirmed with self-report; †Phase 1 at week 12 and for at least 2 of 3 assessments at weeks 9 to 11 and in phase 2 for at least 5 of 6 assessments from weeks 12 to 24, including month 6 (i.e., weeks 21- 24) 95% CI: 95% confidence interval; bup/nal: buprenorphine/naloxone; b/w: between; IQR: interquartile range; mg: milligram: mo.: month(s); N: number of participants; NI: non- inferiority; NR: not reported; NS: not significant; OR: odds ratio; S: superiority SE: standard error; SD: standard deviation; tx diff: treatment difference ©Institute for Clinical and Economic Review, 2018 Page 137 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Table D6. All-Cause Discontinuation and Treatment Retention in Included Studies N at Discontinued, Number of Days/Weeks Study Arm Randomization n (%)* Retained CAM2038 CAM2038 215 89 (41) NR Lofwall 201876 SL bup/nal 213 92 (43) NR Sublocade Sublocade 300 mg/ 203 78 (38) NR 100 mg Trial 13-000177 Sublocade 300 mg/ 201 72 (36) NR 300 mg Placebo 100 66 (66) NR De Novo 412 206 (50) NR Trial 13-000378 Roll Over 257 57 (22) NR Probuphine Probuphine 87 6 (7) NR Rosenthal 201679 SL bup/nal 90 5 (6) NR Probuphine 114 41 (36) NR Placebo Rosenthal 201381 54 40 (74) NR implant SL bup/nal 119 43 (36) NR Probuphine 108 37 (34) NR Ling 201080 Placebo 55 38 (69) NR Implant Vivitrol Vivitrol 283 78 (28) NR Lee 201883 (X-BOT) SL bup/nal 287 62 (22) NR Vivitrol 80 24 (30) 69.3 days† Tanum 201784 SL bup/nal 79 30 (38) 63.7 days† ©Institute for Clinical and Economic Review, 2018 Page 138 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents N at Discontinued, Number of Days/Weeks Study Arm Randomization n (%)* Retained Continuing Solli 201889 (Tanum 54 26 (48) 25.6 weeks† XR-NTX 2017 OLE) Inducted 63 33 (52) 25.4 weeks† on XR-NTX Vivitrol 153 34 (22) NR Lee 201692 Treatment 155 29 (19) NR as usual Vivitrol 16 7 (41) NR Lee 201591 Treatment 17 10 (59) NR as usual Vivitrol 126 59 (47) >168 days‡ Krupitsky 201186 Placebo 124 77 (62) 96 days‡ Continuing Krupitsky 201390 67 28 (42) NR XR-NTX (Krupitsky 2011 OLE) Inducted 47 15 (32) NR on XR-NTX Vivitrol 66 0 NR NEW HOPE88 Placebo 27 0 NR *Percentage of participants who discontinued was calculated from numbers reported in each trial; †Mean; ‡Median bup/nal: buprenorphine/naloxone; mg: milligram; n: number of participants; OLE: open-label extension; SL: sublingual ©Institute for Clinical and Economic Review, 2018 Page 139 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Table D7. Opioid Craving – VAS Scores* Duration of Mean VAS Over Duration Mean VAS Change from Study Arm N p-Value Follow-Up of Follow-Up Baseline CAM2038 CAM2038 213 17.3 (SD: 25.5)† NR Lofwall 201876 24 weeks NR SL bup/nal 215 17.3 (SD: 25.5)† NR Sublocade Sublocade 300mg/100mg 192¤ NR 2.1 (SE: 1.63) vs. placebo: p=0.0003 Trial 13-000177 24 weeks Sublocade 300mg/300mg 193¤ NR -0.9 (SE:1.63) vs. placebo: p<0.0001 Placebo 96¤ NR 11.5 (SE: 2.48)  Probuphine -2.3 (SD: 11.15)‡; Probuphine 84 NR -2.7 (SD: 12.58)† Rosenthal 201679 24 weeks NS for both -2.8 (SD: 19.57)‡; SL bup/nal 89 NR -1.9 (SD: 18.97)† vs. placebo: p<0.0001 Probuphine 114 10.2 NR vs. SL bup/nal: p=0.054 Rosenthal 201381 24 weeks Placebo Implant 54 21.8 NR NR SL bup/nal 119 7.1 NR Probuphine 108 9.9 (95% CI: 7.8 to 12.0) NR Ling 201080 24 weeks p<0.001 Placebo Implant 55 15.8 (95% CI: 12.7 to 18.9) NR Vivitrol Vivitrol 56 0.83 (95% CI: -0.81 to 2.43)§ NR Tanum 201784 12 weeks NR SL bup/nal 49 2.69 (95% CI: 1.77 to 3.60)§ NR Vivitrol 126 NR -10.1 (95% CI: -12.3 to -7.8)† Krupitsky 201186 24 weeks p<0.0001 Placebo 124 NR 0.7 (95% CI: -3.1 to 4.4)† VAS increase: 18.8% Vivitrol 32¤ NR No change: 37.5% VAS decrease: 43.8%# NEW HOPE88 24 weeks NR VAS increase: 20.0% Placebo 15¤ NR No change: 46.7% VAS decrease: 33.3%# ©Institute for Clinical and Economic Review, 2018 Page 140 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents 95% CI: 95% confidence interval; bup/nal: buprenorphine/naloxone, mg: milligram; NR: not reported; NS: not significant; SL: sublingual, VAS: visual analog scale *Opioid craving measured on 100 mm scale, where 0=no craving and 100=strongest craving, unless otherwise noted; †Mean VAS need-to-use score, where 0=no need and 100=strongest need; ‡Mean VAS desire-to-use score, where 0=no desire and 100=strongest desire; §Craving for heroin, rated on a scale of 0=no craving to 10=very strong; #Data reported are percentage of patients reporting opioid craving increases, decreases, or no changes compared to baseline, where 0=no craving and 10=strongest craving; ¤Number of participants analyzed for outcome. ©Institute for Clinical and Economic Review, 2018 Page 141 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Table D8. Opioid Withdrawal – Clinical Opiate Withdrawal Scale (COWS) and Subjective Opiate Withdrawal Scale (SOWS) Mean Change Duration of Mean Change Study Arm N p-Value in SOWS p-Value Follow-Up in COWS Score Score CAM2038 CAM2038 213 3.3 (SD: 3.5)† NR Lofwall 201876 24 weeks NR NR SL bup/nal 215 2.7 (SD: 4.0)† NR Sublocade 191/ Sublocade 300mg/100mg -0.5 (SE: 0.22) vs. placebo NS -0.9 (SE: 0.51) vs. placebo NS 192* 192/ vs. placebo: Trial 13-000177 24 weeks Sublocade 300mg/300mg -1.1 (SE: 0.21) vs. placebo: p=0.01 -2.0 (SE: 0.51) 193* p=0.0028 96/ Placebo -0.1 (SE: 0.35)  0.7 (SE: 0.8)  96* Probuphine Probuphine 84 -0.1 (SD: 1.51) -0.6 (SD: 4.63) Rosenthal 201679 24 weeks NS NS SL bup/nal 89 -0.1 (SD: 1.69) 0.1 (SD: 5.26) Rosenthal 201381 vs. placebo vs. placebo p<0.0001 p<0.0001 Probuphine 114 2.49 (NR)† 5.3 (NR)† vs. bup/nal vs. bup/nal 24 weeks p=0.0005 p=0.0006 Placebo implant 54 4.52 (NR)† 8.42 (NR)† NR NR SL bup/nal 119 1.71 (NR)† 2.83 (NR)† Ling 201080 2.3 (95% CI: 1.9 4.1 (95% CI: Probuphine 108 to 2.7)† 3.1 to 5.1)† 24 weeks p<0.001 p=0.004 Placebo implant 55 3.4 (95% CI: 2.8 6.5 (95% CI: to 4.0)† 5.1 to 7.9)† 95% CI: 95% confidence interval; bup/nal: buprenorphine/naloxone, COWS: clinical opiate withdrawal scale, mg: milligram, N: number of participants, NR: not reported, NS: not significant SD: standard deviation, SE: standard error; SL: sublingual, SOWS: subjective opioid withdrawal scale *Number of p analyzed for COWS and SOWS, respectively; †Mean score, not change ©Institute for Clinical and Economic Review, 2018 Page 142 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Table D9. Serious Adverse Events and Adverse Events Leading to Discontinuation in Included Studies Serious Adverse Discontinuation Due to At Least One Opioid Overdose Fatal Overdoses, Study Arm Death, n (%) Event, n (%) Adverse Event, n (%) Event, n (%) n (%) CAM2038 CAM2038 5 (2.3) 7 (3.3) 0 0 1 (0.5) Lofwall 201876 SL bup/nal 13 (6.0) 3 (1.4) 5 (2.3) 0 0 CPDD Injection CAM2038 2 (1.8) NR 0 0 NR Poster154 SL bup/nal 16 (14.5) NR 5 (4.5) 0 0 Sublocade Sublocade 4 (2.0) 7 (3.4) 0 0 0 300mg/100mg Trial 13-000177 Sublocade 7 (3.5) 10 (5.0) 0 NR 1 (0.5) 300mg/300mg Placebo 5 (5.0) 2 (2.0) 1 (1.0)* 0 0 Trial 13- Roll-over 9 (3.5) 5 (2.0) 0 0 0 000378,156 De Novo 16 (3.9) 12 (3.0) 2 (0.5)* 0 0 Probuphine Rosenthal Probuphine 2 (2.3) 1 (1.1) NR NR NR 201679 SL bup/nal 3 (3.4) 0 NR NR NR Probuphine 6 (5.3) 0 NR 0 0 Rosenthal Placebo implant 3 (5.6) 0 NR 0 0 201381 SL bup/nal 7 (5.9) 1 (0.8) at least 1 (0.8) 1 (0.8) 1 (0.8) Probuphine 2 (1.9) 4 (3.7) NR NR NR Ling 201080 Placebo implant 4 (7.3) 0 NR NR NR Vivitrol Lee 201883 Vivitrol 29 (14.0) 6 (2.1) 15 (5.3) 2 (0.7) 3 (1.1) X-BOT SL bup/nal 29 (11.0) 8 (2.8) 8 (2.8) 3 (1.0) 4 (1.4) Vivitrol 6 (8.5) 4 (5.6) 0 0 0 Tanum 201784 SL bup/nal 3 (4.2) 6 (8.3) 1 (1.4) 0 0 Solli 201889 Continuing Vivitrol 1 (1.9) 4 (7.4) 0 0 0 (Tanum 2017 Inducted on Vivitrol 4 (6.4) 3 (4.8) 0 0 1 (1.6) OLE) ©Institute for Clinical and Economic Review, 2018 Page 143 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Serious Adverse Discontinuation Due to At Least One Opioid Overdose Fatal Overdoses, Study Arm Death, n (%) Event, n (%) Adverse Event, n (%) Event, n (%) n (%) Vivitrol 16 (10.5) 5 (3.3) 0 0 0 Lee 201692 Treatment as usual 45 (29.0) NA 5 (3.2) 2 (1.3) 2 (1.3) Vivitrol 0 NR 0 0 0 Lee 201591 Treatment as usual 0 NR 0 0 0 Krupitsky Vivitrol 3 (2) 2 (2) 0 0 0 201186 Placebo 4 (3) 2 (2) 0 0 0 Krupitsky Continuing Vivitrol 3 (4.5) 0 0 0 0 201390 (Krupitsky 2011 Inducted on Vivitrol 0 1 (2.1) 0 0 0 OLE) Vivitrol 0 0 0 0 0 NEW HOPE88 Placebo 0 0 0 0 0 *Trial reports accidental overdoses only bup/nal: buprenorphine/naloxone; mg: milligram; n: number of participants; OLE: open-label extension; SL: sublingual ©Institute for Clinical and Economic Review, 2018 Page 144 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Table D10. Adverse Events≥5% in Included Studies Injection/ Nervous Gastrointestinal Psychiatric Fatigue/ Infections and Study Arm Implant Site Headache, % System Upset, % Issues, % Insomnia, % Infestations, % Reaction, % Disorders, % CAM2038 6.1 (pruritus); 7.5 (constipation); CAM2038 5.6 (erythema); 7.5 5.6 (insomnia) 7.0 (nausea) Lofwall 8.9 (pain) 201876 6.0 (pruritus); 7.4 (constipation); SL bup/nal 5.6 (erythema); 7.9 2.8 (insomnia) 7.9 (nausea) 7.9 (pain) 5.9 (severe 6.6 (constipation); 5.3 (upper respiratory CAM2038 reaction); 6.6 5.9 (nausea) tract) CPDD Heroin 5.9 (pain) Poster155 2.0 (severe 4.0 (constipation); 3.3 (upper respiratory SL bup/nal reaction); 2.0 2.6 (nausea) tract) 5.3 (pain) Sublocade 9.4 (constipation); 5.4 (nasopharyngitis); Sublocade 6.4 (pruritus); 3.9 (fatigue); 8.9 (nausea); 9.4 7.4 (upper respiratory 300mg/100mg 4.9 (pain) 6.4 (insomnia) 9.4 (vomiting) tract) 8.0 (constipation); 5.0 (nasopharyngitis); Trial 13- Sublocade 9.5 (pruritus); 6.0 (fatigue); 8.0 (nausea); 8.5 6.0 (upper respiratory 000177 300mg/300mg 6.0 (pain) 8.5 (insomnia) 5.5 (vomiting) tract) 0 constipation); 1.0 (nasopharyngitis); 4.0 (pruritus); 3.0 (fatigue); Placebo 5.0 (nausea); 6.0 1.0 (upper respiratory 3.0 (pain) 11 (insomnia) 4.0 (vomiting) tract) ©Institute for Clinical and Economic Review, 2018 Page 145 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Injection/ Nervous Gastrointestinal Psychiatric Fatigue/ Infections and Study Arm Implant Site Headache, % System Upset, % Issues, % Insomnia, % Infestations, % Reaction, % Disorders, % 3.5 (constipation); Roll-over 2.7 (pain); 2.0 3.9 (insomnia) 2.3 (nasopharyngitis) 3.9 (nausea) Trial 13- 2.0 (erythema) 000378 9.5 (pain); 11.4 (constipation); De Novo 7.5 6.6 (insomnia) 5.8 (nasopharyngitis) 5.3 (erythema) 9.0 (nausea) Probuphine 6.9 Probuphine 13.8 (any) 8.0 (any) 6.9 9.2 8.0 (nasopharyngitis) (depression) Rosenthal 201679 2.2 SL bup/nal 7.9 (any) 1.1 (any) 3.4 3.4 4.5 (nasopharyngitis) (depression) 27.2 (any); 8.8 5.3 (nasopharyngitis); 6.1 (nausea); Probuphine 7.0 (hematomas); 13.2 (depression); 7.9 (insomnia) 8.8 (upper respiratory 6.1 (vomiting) 5.3 (pain) 1.8 (anxiety) tract) Rosenthal 201381 25.9 (any); 5.6 5.6 (nasopharyngitis); 1.9 (nausea); Placebo implant 11.1 (hematomas); 9.3 (depression); 14.8 (insomnia) 7.4 (upper respiratory 1.9 (vomiting) 9.3 (pain) 5.6 (anxiety) tract) ©Institute for Clinical and Economic Review, 2018 Page 146 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Injection/ Nervous Gastrointestinal Psychiatric Fatigue/ Infections and Study Arm Implant Site Headache, % System Upset, % Issues, % Insomnia, % Infestations, % Reaction, % Disorders, % 3.4 10.1 (nasopharyngitis); 6.7 (nausea); SL bup/nal NA 16.0 (depression); 13.4 (insomnia) 9.2 (upper respiratory 4.2 (vomiting) 5.9 (anxiety) tract) Continuing 14.0 (any) Probuphine Beebe 2012 82 Placebo (Study 2 of 8.2 (upper respiratory implant 12.5 (any) 11.8 Rosenthal tract) Probuphine 201381) SL bup/nal 15.0 (any) Probuphine 56.5 (any); 25.0 (erythema); 13.9 (constipation); 13.9 (nasopharyngitis); 13.0 (edema); Probuphine 5.6 (diarrhea); 25.0 10.2 (anxiety) 21.3 (insomnia) 13.0 (upper 25.0 (itching); 13.9 (nausea) respiratory tract) 22.2 (pain); 12.0 (bleeding) Ling 201080 52.7 (any); 21.8 (erythema); 5.5 (constipation); 5.5 (nasopharyngitis); 9.1 (edema); Placebo implant 12.7 (diarrhea); 18.2 9.1 (anxiety) 21.8 (insomnia) 10.9 (upper 14.5 (itching); 12.7 (nausea) respiratory tract) 10.9 (pain); 12.7 (bleeding) Vivitrol 10.6 Lee 201883 Vivitrol 16.3 (any) 12.0 (any) (psychiatric 7.8 (any) (X-BOT) disorder) ©Institute for Clinical and Economic Review, 2018 Page 147 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Injection/ Nervous Gastrointestinal Psychiatric Fatigue/ Infections and Study Arm Implant Site Headache, % System Upset, % Issues, % Insomnia, % Infestations, % Reaction, % Disorders, % 10.1 SL bup/nal NA 20.6 (any) (psychiatric 9.4 (any) disorder) 16.9 (anxiety or Vivitrol 5.6 (any) 11.3 (insomnia) Tanum depression) 201784 8.3 (anxiety or SL bup/nal N/A 4.2 (insomnia) depression) 9.3 Continuing 9.3 (any) 9.3 (psychological 3.7 (insomnia) Solli 201889 Vivitrol reactions) (Tanum 2017 12.7 OLE) Inducted on 3.2 (any) 11.1 (psychological 9.5 (insomnia) Vivitrol reactions) Vivitrol 27.5 (mild reaction) 18.3 (any) 19 7.2 (insomnia) 9.8 (nasopharyngitis) Lee 201692 Treatment as NA 1.9 (any) 8.4 5.2 (insomnia) 11.0 (nasopharyngitis) usual Vivitrol 5.0 (pain) 6.0 (insomnia) 7.0 (nasopharyngitis) Krupitsky 201186 Placebo 1.0 (pain) 1.0 (insomnia) 2.0 (nasopharyngitis) 15.2 (immediate Vivitrol 7.6 9.1 (fatigue) reaction) NEW HOPE88 7.4 (immediate Placebo 0 3.7 (fatigue) reaction) bup/nal: buprenorphine/naloxone; mg: milligram; n: number of participants; OLE: open-label extension; SL: sublingual ©Institute for Clinical and Economic Review, 2018 Page 148 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Appendix E. Comparative Value Supplemental Information Table E1. Impact Inventory Included in This Analysis Notes on Sources (if Type of Impact from… Perspective? quantified), Likely Sector (Add additional domains, as relevant) Health Care Magnitude & Impact Societal Sector (if not) Formal Health Care Sector Longevity effects X X Health Health-related quality of life effects X X outcomes Adverse events   Paid by third-party payers X X Paid by patients out-of-pocket X X Medical costs Future related medical costs   Future unrelated medical costs   Informal Health Care Sector Patient time costs NA  Health-related Unpaid caregiver-time costs NA  costs Transportation costs NA  Non-Health Care Sectors Labor market earnings lost NA X Cost of unpaid lost productivity due to NA X Productivity illness Cost of uncompensated household NA  production 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 X Impact of intervention on educational Education NA  achievement of population Housing Cost of home improvements, remediation NA  Production of toxic waste pollution by Environment NA  intervention Other Other impacts (if relevant) NA  NA: not applicable Adapted from Sanders et al., 2016.157 ©Institute for Clinical and Economic Review, 2018 Page 149 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents hēRo3 hēRo3 compiles information and data that users enter into a browser describing the structure and estimated parameters of a model, sends it to the cloud-based platform where necessary calculations are performed in heRomod, and then parses information received from the modeling package to various output displays, including Markov traces, bar charts, area charts, tornado diagrams, waterfall charts, efficiency frontiers, and hexbin and contour plots, as well as tabular displays. hēRo3 effectively allows users to build and run models in the programming language, R, even if they have had limited or no experience programming in R. hēRo3 also generates an Excel workbook with every model that provides a detailed listing of all input variables, intermediate calculations, and final output on a cycle-by-cycle basis to facilitate model checking and auditing. Table E2. Akaike-Information-Criterion (AIC) for Parametric Curve Functions Fit to Treatment Discontinuation/Relapse Parametric Curve Distributions Log- Exponential Weibull Log-Normal Gamma Logistic CAM2038117 876.7731 876.8574 871.4067* 890.3406 876.1517 Generic SL 867.3982 867.1221 871.8161 883.9106 866.8735* Buprenorphine/Naloxone117 Vivitrol114 652.9479 633.2282 630.9630* 651.0424 634.6471 Generic SL 891.1924 874.8861 867.0644* 900.6312 877.4627 Buprenorphine/Naloxone114 Probuphine116 115.5127 115.5183 120.7789 131.0022 115.4177* Generic Buprenorphine/Naloxone116 198.8468 198.9383 198.0326* 213.8443 199.2321 *Distribution chosen for the model Table E3. Disutility Associated with HIV Infection Utility Multiplier Calculation PWID 135 0.90 Symptomatic HIV 135 0.81 ART135 1.15 PWID with Symptomatic HIV treated with 0.838 0.90*0.81*1.15 ART Disutility in PWID with Symptomatic HIV 0.069 -(1-(0.838/0.90)) treated with ART PWID: persons who inject drugs, HIV: human immunodeficiency virus, ART: anti-retroviral therapy ©Institute for Clinical and Economic Review, 2018 Page 150 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Table E4. Productivity Loss Calculations Inflated/Deflated Parameter Original Value Notes Value Deflated from 2009 to 2007 to match Total Annual Workplace $25,582,000,000 $23,995,751,389 SAMHSA number of persons abusing Productivity Cost145 prescription opioids in 2007 Annual Number of SAMHSA number of persons abusing Persons Abusing 1,707,000 prescription opioids in 2007 Prescription Opioids146 Annual Workplace Inflated using OECD Hourly Earnings Index, Productivity Cost per $14,058 $17,405 2007 Annual to 2018 Q1-Q2 Average158 Person Table E5. Criminal Justice and Incarceration Calculations Parameter Original Value Notes Per Day Cost when on OAT 147 $35 Inflated using General CPI 2014 Annual Value to 2018 Per Day Cost Post-Treatment 147 $175 January -June Average Value.159 Inflated Value multiplied to Calculate Cost per Cycle One-Way Sensitivity and Probabilistic Analyses Inputs for Treatment Discontinuation Inputs for One-Way Sensitivity Analyses When available, varied base case inputs by 95% CIs or published ranges. • All drug costs were varied by ±25% and non-drug health care costs were varied by ±20% of the base case estimate. • All utilities were varied by their 95% CIs or assumed/calculated ranges Only estimates for which ranges were not presented in the main report are presented in the table below. Table E6. One-Way Sensitivity Analyses Inputs Estimate Base Case Estimate Range Notes PWID % 49.34% 39.47% to 59.21% Assumption (±20%) Incidence of HCV Infection 26.7% 0.017 to 0.517 Assumption (±25% Points) Probability of Abstinence 95% CI calculated using reported over a 24-week period – 34.2% 29.3% to 39.1% standard error (2.46%) CAM2038 Probability of Abstinence over a 24-week period – 95% CI calculated using reported 27.4% 22.5% to 32.3% Generic SL standard error (2.45%) Buprenorphine/Naloxone ©Institute for Clinical and Economic Review, 2018 Page 151 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Estimate Base Case Estimate Range Notes Proportion Permanently Abstained from Illicit Use of 10% 0% to 20% Assumption (±10% Points) Opioids Proportion of Discontinuation from Health States With And Without 46% 41.1% to 50.3% Assumption (±10%) Illicit Use of Opioids While On MAT Opioid Overdose-Related Range based on lowest and Mortality Rate (per 100,000 13.3 2.4 to 43.4 highest US national rates Illicit Users of Opioids) Range based on lowest and Physician’s Office Visit Cost $21.96 $19.77 to $27.63 highest US national non-facility (CPT: 99211) price Range based on lowest and Cost of Probuphine Insertion $145.80 $129.09 to $179.43 highest US national non-facility (CPT: price Range based on lowest and Cost of Probuphine Removal $163.08 $144.99 to $202.43 highest US national non-facility (CPT: price Range based on lowest and Cost of SC/IM Injection $20.88 $18.78 to $26.26 highest US national non-facility Administration (CPT: price For all relapse/discontinuation parameters for all MATs and their respective comparators (except Sublocade and its comparator), 95% confidence interval estimates for the parametric curve functions were used in the one-way sensitivity analyses, and are presented in Tables E9 to E14. Since discontinuation/relapse is a function of two parameters (mean & SD for lognormal distributions OR shape & scale for gamma distributions), we jointly varied the parameters in the “one-way” sensitivity analyses, taking into account their correlation. Table E7. Time to Discontinuation Parameter Estimates for the Lognormal Model for CAM2038117 Mean Estimate 95% CI Lower Bound 95% CI Upper Bound Standard Error Mean Log 3.5041796 3.1923378 3.8160215 0.15910590 SD Log 0.5862179 0.4347833 0.7376525 0.07726397 CI: confidence interval ©Institute for Clinical and Economic Review, 2018 Page 152 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Table E8. Time to Discontinuation Parameter Estimates for the Gamma Model for Generic SL Buprenorphine/Naloxone (versus CAM2038)117 Mean Estimate 95% CI Lower Bound 95% CI Upper Bound Standard Error Shape -0.1949288 -0.4402407 0.05038306 0.1251614 Scale -4.1432573 -4.6410451 -3.64546956 0.2539780 CI: confidence interval Table E9. Time to Discontinuation Parameter Estimates for the Lognormal Model for Vivitrol114 Mean Estimate 95% CI Lower Bound 95% CI Upper Bound Standard Error Mean Log 3.0975681 2.7731938 3.4219424 0.16550014 SD Log 0.7018769 0.5193503 0.8844035 0.09312753 CI: confidence interval Table E10. Time to Discontinuation Parameter Estimates for the Lognormal Model for Generic SL Buprenorphine/Naloxone (versus Vivitrol)114 Mean Estimate 95% CI Lower Bound 95% CI Upper Bound Standard Error Mean Log 2.9434162 2.6977650 3.1890673 0.12533454 SD Log 0.5740773 0.4143103 0.7338443 0.08151526 CI: confidence interval Table E11. Time to Discontinuation Parameter Estimates for the Gamma Model Probuphine116 Mean Estimate 95% CI Lower Bound 95% CI Upper Bound Standard Error Shape -0.4962139 -1.199059 0.2066317 0.3586013 Scale -6.6180437 -9.290763 -3.9453238 1.3636576 CI: confidence interval Table E12. Time to Discontinuation Parameter Estimates for the Lognormal Model for Generic SL Buprenorphine/Naloxone (versus Probuphine)116 Mean Estimate 95% CI Lower Bound 95% CI Upper Bound Standard Error Mean Log 4.2138822 3.6446007 4.7831637 0.2904551 SD Log 0.6124004 0.3520751 0.8727258 0.1328215 CI: confidence interval Inputs for Probabilistic Analyses Triangular distributions were used for all cost parameters with the base case assumed as the “peak,” and lower and upper bound of ranges assumed as the “lower” and “upper” bounds of the distribution. Additional probabilistic analyses inputs are presented below. ©Institute for Clinical and Economic Review, 2018 Page 153 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Table E13. Probabilistic Analyses Inputs Parameter Distribution Parameters Utility in “MAT with NO Illicit Use of Opioids” α = 648.61 Beta Health State β = 198.14 Utility in “MAT with Illicit Use of Opioids” Health α = 502.49 State Among Patients Illicitly Using Prescription Beta β = 215.36 Opioids Utility in “MAT with Illicit Use of Opioids” Health α = 420.08 Beta State Among PWID β = 259.66 Utility in “OFF MAT with Illicit Use of Opioids” α = 503.81 Health State Among Patients Illicitly Using Beta β = 222.14 Prescription Opioids Utility in “OFF MAT with Illicit Use of Opioids” α = 404.15 Beta Health State Among PWID β = 299.94 Peak = 0.852 Utility in “OFF MAT with NO Illicit Use of Opioids” Triangular Lower = 0.736 Health State Upper = 0.901 Peak = 26.7% Incidence of HCV Infection Triangular Lower = 1.7% Upper = 51.7% α = 32.17 Opioid-Related Overdose Mortality Rate Beta β = 58463.17 Proportion of PWID diagnosed with HCV with α = 74.80 Beta Spontaneous Clearance of HCV Infection β = 231.77 Probability of Abstinence over a 24-week period – α = 182.20 Beta CAM2038 β = 3014.36 Probability of Abstinence over a 24-week period – α = 119.32 Beta Generic SL Buprenorphine/Naloxone β = 2493.47 Proportion of Discontinuation from Health States Prob = 45.71% With And Without Illicit Use of Opioids While On Binomial Size = 70 MAT Probability of Discontinuation of Sublocade over a Prob = 35.8% Binomial 24-week period Size = 196 Peak = 6.9% HIV Disutility Multiplier Triangular Lower = 1% Upper = 19.5% Peak = 10% Proportion Permanently Abstained from Illicit Use Triangular Lower = 0% of Opioids Upper = 20% α = 3833.92 HCV Disutility Multiplier - Post SVR Beta β = 3.84 ©Institute for Clinical and Economic Review, 2018 Page 154 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Parameter Distribution Parameters α = 47.47 HCV Disutility Multiplier - F0 to F3 Liver Disease Beta β = 3.57 Odds Ratio – Discontinuation of Generic SL Mean = 0.67 Lognormal Buprenorphine/Naloxone vs. Sublocade SD Log = 0.45 For all relapse/discontinuation parameters for all MATs and their respective comparators, normal distributions were used for the relevant parametric curve functions in the probabilistic analyses, with distribution parameters being mean, standard deviation, and correlation presented in Tables E16 to E21. Table E14. Covariance of Discontinuation Parameters for Lognormal Model for CAM2038117 Mean Log SD Log Mean Log 0.025314689 0.005866107 SD Log 0.005866107 0.005969721 Correlation coefficient: 0.4771849; SD: standard deviation Table E15. Covariance of Discontinuation Parameters for Gamma Model for Generic SL Buprenorphine/Naloxone (versus CAM2038)117 Shape Scale Shape 0.01566538 0.02794661 Scale 0.02794661 0.06450484 Correlation coefficient: 0.8791488 Table E16. Covariance of Discontinuation Parameters for Lognormal Model for Vivitrol114 Mean Log SD Log Mean Log 0.027390297 0.004508215 SD Log 0.004508215 0.008672736 Correlation coefficient: 0.2925016; SD: standard deviation Table E17. Covariance of Discontinuation Parameters for Lognormal Model for Generic SL Buprenorphine/Naloxone (versus Vivitrol)114 Mean Log SD Log Mean Log 0.015708746 0.003050788 SD Log 0.003050788 0.006644737 Correlation coefficient: 0.2986086; SD: standard deviation ©Institute for Clinical and Economic Review, 2018 Page 155 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Table E18. Covariance of Discontinuation Parameters for Gamma Model for Probuphine116 Shape Scale Shape 0.1285949 0.4618465 Scale 0.4618465 1.8595621 Correlation coefficient: 0.9444533 Table E19. Covariance of Discontinuation Parameters for Lognormal Model for Generic SL Buprenorphine/Naloxone (versus Probuphine)116 Mean Log SD Log Mean Log 0.08436416 0.02151929 SD Log 0.02151929 0.01764156 Correlation coefficient: 0.5578025; SD: standard deviation ©Institute for Clinical and Economic Review, 2018 Page 156 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents One-Way Sensitivity Analyses Figure E1. Tornado Diagram – Vivitrol versus Generic SL Buprenorphine/Naloxone ©Institute for Clinical and Economic Review, 2018 Page 157 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Figure E2. Tornado Diagram – Probuphine versus Generic SL Buprenorphine/Naloxone Relapse to illicit use of opioids when on generic SL buprenorphine/naloxone was also a key driver of the results. However, that has not been included here since varying this estimate changed not only the costs and QALYs in the comparator arm, but also in the Probuphine arm since the comparator is the treatment choice in those abstinent from illicit use at the time of removal of Probuphine implant. ©Institute for Clinical and Economic Review, 2018 Page 158 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Probabilistic Analyses Figure E3. Probabilistic Analyses: Vivitrol versus Generic SL Buprenorphine/Naloxone – Incremental Cost-Effectiveness Ratio HexBin Figure E4. Probabilistic Analyses: Vivitrol versus Generic SL Buprenorphine/Naloxone – Incremental Cost-Effectiveness Ratio Acceptability Curve ©Institute for Clinical and Economic Review, 2018 Page 159 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Figure E5. Probabilistic Analyses: Probuphine versus Generic SL Buprenorphine/Naloxone – Incremental Cost-Effectiveness Ratio HexBin Figure E6. Probabilistic Analyses: Probuphine versus Generic SL Buprenorphine/Naloxone – Incremental Cost-Effectiveness Ratio Acceptability Curve Modified Societal Perspective Table E20. CAM2038 versus Generic SL Buprenorphine/Naloxone Criminal Justice & Treatment Lost Productivity Costs Incarceration Costs CAM2038 $13,600 $81,700 Generic SL Buprenorphine/Naloxone $17,200 $96,700 ©Institute for Clinical and Economic Review, 2018 Page 160 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Table E21. Vivitrol versus Generic SL Buprenorphine/Naloxone Criminal Justice & Treatment Lost Productivity Costs Incarceration Costs Vivitrol $17,700 $101,000 Generic SL Buprenorphine/Naloxone $15,500 $91,700 Table E22. Probuphine versus Generic SL Buprenorphine/Naloxone Criminal Justice & Treatment Lost Productivity Costs Incarceration Costs Probuphine $9,700 $67,700 Generic SL Buprenorphine/Naloxone $10,300 $70,200 Shorter Time Horizons Table E23. One-Year Time Horizon Incremental QALYs Incremental Costs Incremental Cost per QALY CAM2038* 0.004 - - Vivitrol -0.020 $3,200 More costly, less effective Probuphine 0.002 $2,400 $963,000 QALY: quality-adjusted life year Each intervention was compared to its relevant generic buprenorphine/naloxone comparator. *No incremental costs or cost per QALY is reported since CAM2038 currently does not have a list or net price. Table E24. Two-Year Time Horizon Incremental QALYs Incremental Costs Incremental Cost per QALY CAM2038* 0.014 - - Vivitrol -0.028 $5,500 More costly, less effective Probuphine 0.005 $2,500 $465,000 QALY: quality-adjusted life year Each intervention was compared to its relevant generic buprenorphine/naloxone comparator *No incremental costs or cost per QALY is reported since CAM2038 currently does not have a list or net price Population Cohort Comprising Only PWID Seeking MAT for OUD Table E25. CAM2038 versus Generic SL Buprenorphine/Naloxone in an OUD Population Comprising 100% PWID Treatment Total Costs QALYs CAM2038 - 3.224 Generic SL Buprenorphine/Naloxone $73,500 3.221 QALY: quality-adjusted life year ©Institute for Clinical and Economic Review, 2018 Page 161 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Table E26. Vivitrol versus Generic SL Buprenorphine/Naloxone in an OUD Population Comprising 100% PWID Treatment Total Costs QALYs Incremental Cost per QALY Gained Vivitrol $84,100 3.208 More costly, less effective Generic SL $73,600 3.241 Buprenorphine/Naloxone QALY: quality-adjusted life year Table E27. Probuphine versus Generic SL Buprenorphine/Naloxone in an OUD Population Comprising 100% PWID Treatment Total Costs QALYs Incremental Cost per QALY Gained Probuphine $79,400 3.353 $233,000 Generic SL $76,400 3.342 Buprenorphine/Naloxone QALY: quality-adjusted life year Analyses Excluding the “Permanently Abstained from Illicit Use of Opioids” Health State Table E28. CAM2038 versus Generic SL Buprenorphine/Naloxone in a Scenario That Excludes Permanent Abstinence from Illicit Use of Opioids Treatment Total Costs QALYs CAM2038 - 3.255 Generic SL Buprenorphine/Naloxone $70,500 3.193 QALY: quality-adjusted life year Table E29. Vivitrol versus Generic SL Buprenorphine/Naloxone in a Scenario That Excludes Permanent Abstinence from Illicit Use of Opioids Treatment Total Costs QALYs Incremental Cost per QALY Gained Vivitrol $81,900 3.239 More costly, less effective Generic SL $71,800 3.268 Buprenorphine/Naloxone QALY: quality-adjusted life year ©Institute for Clinical and Economic Review, 2018 Page 162 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Table E30. Probuphine versus Generic SL Buprenorphine/Naloxone in a Scenario That Excludes Permanent Abstinence from Illicit Use of Opioids Treatment Total Costs QALYs Incremental Cost per QALY Gained Probuphine $78,300 3.359 $279,000 Generic SL $75,600 3.369 Buprenorphine/Naloxone QALY: quality-adjusted life year “Protocol” Approach to Treatment Table E31. CAM2038 versus Generic SL Buprenorphine/Naloxone Treatment Total Costs QALYs CAM2038 - 3.261 Generic SL Buprenorphine/Naloxone $70,100 3.202 QALY: quality-adjusted life year Table E32. Vivitrol versus Generic SL Buprenorphine/Naloxone Treatment Total Costs QALYs Incremental Cost per QALY Gained Vivitrol $88,300 3.310 $1,100,000 Generic SL $71,600 3.295 Buprenorphine/Naloxone QALY: quality-adjusted life year Table E33. Probuphine versus Generic SL Buprenorphine/Naloxone Treatment Total Costs QALYs Incremental Cost per QALY Gained Probuphine $79,500 3.416 $64,700 Generic SL $76,500 3.404 Buprenorphine/Naloxone QALY: quality-adjusted life year Consecutive Use of Probuphine per Prescribing Label Table E34. Probuphine versus Generic SL Buprenorphine/Naloxone when Patients Are Administered Two Probuphine Implants Consecutively Treatment Total Costs QALYs Incremental Cost per QALY Gained Probuphine $81,100 3.395 $236,000 Generic SL $75,100 3.370 Buprenorphine/Naloxone QALY: quality-adjusted life year ©Institute for Clinical and Economic Review, 2018 Page 163 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Appendix F. 2014 APA Clinical Guideline The following guideline was summarized in ICER’s 2014 report on opioid dependence. This guideline has not been updated since the previous report was issued. American Psychiatric Association Practice Guideline for the Treatment of Patients with Substance Abuse Disorders (2010) Buprenorphine or Buprenorphine/Naloxone (Suboxone) The APA clinical guidelines state that buprenorphine may be effective on a less than daily schedule and as a bridging agent to naltrexone. Therefore, the guidelines recommend that clinicians administer higher, but less frequent doses. Buprenorphine may be best suited for patients with less severe physical dependence. Although the rate of overdose is lower compared to methadone, combining buprenorphine and a benzodiazepine is more likely to be fatal. Naltrexone and Vivitrol (Injectable Naltrexone) The APA clinical guidelines recommend naltrexone as a maintenance agent as it is highly effective in blocking short-acting opioids. However, retention is generally poor and treatment with naltrexone poses a high risk of relapse. As such, the APA states that naltrexone should be utilized in particularly motivated patients who are willing to participate in ancillary services, such as psychosocial and behavioral counseling. ©Institute for Clinical and Economic Review, 2018 Page 164 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Appendix G. Public Comments This section includes summaries of the public comments prepared for the New England CEPAC Public Meeting on November 8, 2018 in Newton, Massachusetts. These summaries were prepared by those who delivered the public comments at the meeting and are presented in order of delivery. One speaker did not submit the summary of his public comments. A video recording of all comments can be found here, beginning at minute 1:14:00. Conflict of interest disclosures are included at the bottom of each statement for each speaker who is not employed by a pharmaceutical manufacturer. Ted Buckley, PhD Vice President, Government Affairs and Advocacy Braeburn Braeburn acknowledges that ICER did include the data which shows the superiority Of CAM2038 versus sublingual buprenorphine with naloxone on the Cumulative Distribution Function with respect to the proportion of opioid-negative urine samples for weeks 4 to 24 . Braeburn believes that there are two areas on which ICER should focus to improve reports. 1. Inclusion of more data that is relevant to the issue being studied: Braeburn believes that there are some gaps with the information included in the report including the issue of diversion. There are multiple published studies in peer-reviewed journals which highlight diversion frequently occurring with prescribed oral buprenorphine which should be mitigated with a HCP-injected formulation. Braeburn believes that ICER should have included this and provided a reasonable assumption of benefits. 2. Improvements to the process: Braeburn suggests some improvements to the process that would increase stakeholder input and its utilization. These could include: a. Lengthening the time period for commenting on various steps of the process. b. Eliminating limitations on the length of the submissions and increasing the opportunities for discussion of draft content. Braeburn found its discussions with ICER to be helpful – resulting in a better product. We believe that increasing these interactions can only be beneficial. For instance, there was a collaborative dialogue between ICER and Braeburn at the CEPAC meeting which may lead to value enhancing additions to the report. Implementation of these two ideas could result in more robust, thoughtful discussions which would lead to a better, more thorough report. Conflicts of interests: Ted Buckley is a full-time employee of Braeburn. ©Institute for Clinical and Economic Review, 2018 Page 165 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Ponni Subbiah, MD, MPH Chief Medical Officer Indivior At this critical time in the opioid crisis, access to all medication-assisted treatments and counseling is critical to ensure that patients get the medical care they deserve and need, just like the care provided to patients with other chronic diseases. Opioid use disorder is a chronic disease that is characterized by long-lasting changes in the brain, which leads to intense drug craving and compulsive use with loss of control.1,2 In addition, each individual’s treatment journey is different and uniquely impacted by societal considerations such as stigma, lack of access to prescribers and treatment, and difficulty adhering to treatment plans, all of which make it challenging to fully quantify the value of these new treatment options until more data is generated through real-world practice. Long-acting formulations of buprenorphine have the potential to improve adherence, as well as address treatment challenges like medication abuse, misuse, and diversion. Accordingly, the conclusions of ICER’s evidence report could negatively impact patient access to these important new options. Subsequently, Indivior offers the following suggestions: -Payers, prescribers and policymakers recognize the potential value of long-acting injectables and acknowledge that it is premature to reach any conclusions on the comparative cost-effectiveness of these treatments until more evidence is generated in real-world practice. -ICER considers recommending to payers that they exercise their own judgement and consider all available data before making coverage decisions. -Manufacturers should be held accountable to generate data to support the value proposition of their treatments. 1. Murthy VH. Ending the Opioid Epidemic – A Call to Action. N Engl J Med 2016; 375:2413-2415. 2. National Institute on Drug Abuse. (2007). Drugs, Brains, and Behavior: The Science of Addiction. Retrieved from https://d14rmgtrwzf5a.cloudfront.net/sites/default/files/soa_2014.pdf. Accessed October 23, 2017. Conflicts of interests: Ponni Subbiah is a full-time employee of Indivior. Maria Sullivan, MD Senior Medical Director Alkermes We are at a critical time in this crisis. Over two million people are battling opioid use disorder (OUD) and only a small percentage are getting the care that they need.1,2 Most patients living with this devastating disease do not receive treatment that includes medication-assisted treatment (MAT), primarily because of limited awareness of treatment options and poor access to healthcare providers.2,3,4 In fact, less than 10% of individuals with OUD receive any form of treatment and less than 3% of treatment programs in the U.S. offer all three MAT options.1,5 ©Institute for Clinical and Economic Review, 2018 Page 166 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents OUD is too complex a disease to use methods of comparison that lack real-world context and patient choice.6 The three FDA-approved medications to treat OUD have distinctly different roles in a complex treatment paradigm and are not interchangeable.7,8 Any comparative analysis of these medications must take into consideration that each treatment is fundamentally different, and patients seeking each type of medication may vary in their clinical presentation, preferences, and where they are in their recovery journey. Only broad awareness and access to all evidence-based medications will allow people with opioid dependence to engage with their providers to find the right treatment plan to meet their evolving needs. The urgency for accessible, comprehensive, and patient-centered care for people with OUD has never been more timely or pressing. We believe that we must all come together with even more determination and collaboration than ever before to demand open and equal access for all evidence-based, FDA-approved medications. 1. Ahrnsbrak, Rebecca, et al. “Key Substance Use and Mental Health Indicators in the United States: Results from the 2016 National Survey on Drug Use and Health.” Key Substance Use and Mental Health Indicators in the United States: Results from the 2016 National Survey on Drug Use and Health, SAMHSA, 2016, www.samhsa.gov/data/sites/default/files/NSDUH-FFR1-2016/NSDUH-FFR1-2016.htm. 2. Saloner B, Karthikeyan S. Changes in substance abuse treatment use among individuals with opioid use disorders in the United States, 2004-2013. JAMA 2015;314:1515-7. 3. Volkow ND, Frieden TR, Hyde PS, Cha SS. Medication-assisted therapies--tackling the opioid-overdose epidemic. The New England Journal of Medicine 2014;370:2063-6. 4. Treatment Center for Substance Abuse. SAMHSA/CSAT treatment improvement protocols. Medication-assisted treatment for opioid addiction in opioid treatment programs. Rockville (MD): Substance Abuse and Mental Health Services Administration (US); 2005. 5. Roman PM, Abraham AJ, Knudsen HK. Using medication-assisted treatment for substance use disorders: Evidence of barriers and facilitators of implementation. Addictive Behaviors 2011;36:584-9. 6. American Society of Addiction Medicine. “Treating Opioid Addiction as a Chronic Disease.” ASAM, 2014, Accessible at: www.asam.org/docs/default-source/advocacy/cmm-fact-sheet---11-07-14.pdf. 7. Tanum L, Solli KK, Latif ZE, et al. Effectiveness of Injectable Extended-Release Naltrexone vs Daily Buprenorphine-Naloxone for Opioid Dependence: A Randomized Clinical Noninferiority Trial. JAMA Psychiatry 2017;74:1197-205. 8. Lee JD, Nunes EV, Novo P, et al. Comparative effectiveness of extended-release naltrexone versus buprenorphine-naloxone for opioid relapse prevention (X:BOT): A Multicentre, Open-label, Randomised Controlled Trial. Lancet (London, England) 2018;391:309-18. Conflicts of interests: Maria Sullivan is a full-time employee of Alkermes. James Andersen, MD Principal Investigator Meridien Research As a provider of all the various pharmacologic therapies reviewed I can see the likely areas where each is probably going to be preferred, sometimes by the provider, sometimes by the patient. People do self-select due to previous experiences (good or bad) and the experiences of others. With regards to drop out rates, it is clear that addiction predisposes to this because of the brain changes and need of a thorough change of lifestyle to remain in treatment and recovery. Our ©Institute for Clinical and Economic Review, 2018 Page 167 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents current state is similar to trying to design better lifeboats for the Titanic. The life preservers were like naloxone, hoping to make it to a lifeboat. Some of the lifeboats were filled with passengers who had not been in the water and took only dry people (naltrexone ER). Others, equipped with officers and crew took the injured or panicked who needed supervision (methadone), and those who were better off ended up in available seats (buprenorphine). But what about a lifeboat that was covered, stabilized, heated, and supplied with regular meals and beverages, that could make one feel almost normal? That is the response I received commonly from participants in our open label study on buprenorphine injectable (Sublocade). They felt “normal” as in feeling they were not using. Longer term studies will be needed (and are ongoing or planned) to determine if a better designed lifeboat will carry more people farther, even to the point of reaching shore (remission leading to meaningful recovery). Think about cost as the amount a passenger would pay to have that form of lifesaving available. Conflicts of Interest: James Andersen received manufacturer support while serving as principal investigator on RB 6000 (Sublocade) at Meridien Research Maddy Reinert, MPH Policy and Programs Associate Mental Health America In conducting the cost-effectiveness modeling, Mental Health America (MHA) asks that the Institute for Clinical and Economic Review (ICER) separately consider cost-effectiveness from the perspective of the government as a public payer – Medicaid and disability Medicare are the largest payers for adults experiencing opioid addiction in the United States. We know that poverty and disability contribute to the development of behavioral health conditions, and behavioral health conditions create burdens that can cause poverty and disability. Effective treatment and management of behavioral health conditions can break this cycle and allow individuals to reach or maintain a level of community participation that positions them to maintain or be able to purchase commercial insurance, which dramatically reduces costs and increases cost- effectiveness from a public payer perspective. With Medicaid and disability Medicare, increases in productivity beyond a threshold uniquely reduce health care costs as the individual disenrolls or never requires coverage in the first place, impacting ICER’s cost-effectiveness calculations for these public payers. This threshold can be estimated through scenario analysis. Further, by making such analyses common practice, it can shift the paradigm for how the Centers for Medicare and Medicaid Services (CMS) and state Medicaid agencies view costs and benefits, away from trimming health care costs and toward making investments in human thriving that alleviate poverty and disability. Finally, we believe there are key clinical differences between opioid agonist and antagonist medications. This is especially important as this report has implications for access to MATs, and if ICER assesses the comparative clinical effectiveness and value of these medications, we ask that ©Institute for Clinical and Economic Review, 2018 Page 168 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents they are compared to comparable medications in terms of clinical indication and patient population. Conflicts of Interest: Mental Health America receives more than 25% of its funding from Alkermes. ©Institute for Clinical and Economic Review, 2018 Page 169 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Appendix H. Conflict of Interest Disclosures Tables G1 through G3 contain conflict of interest (COI) disclosures for all participants at the November 8, 2018 Public meeting of NE CEPAC. Table G1. ICER Staff and Consultant COI Disclosures Name Organization Disclosures Foluso Agboola, MBBS, MPH ICER None Reiner Banken, MD, MSc ICER None Rick Chapman, PhD, MS ICER None Laura Cianciolo, BA ICER None Sarah Emond, MPP ICER None Katherine Fazioli, BS ICER None Noemi Fluetsch, MPH ICER None Serina Herron-Smith, BA ICER None Varun Kumar, MBBS, MPH, MSc ICER None Jerry Gurwitz, MD ICER None Madeline O’Grady, BS ICER None Ifeoma Otuonye, MPH ICER None Steven Pearson, MD, MSc ICER None David Rind, MD, MSc ICER None Matt Seidner, BS ICER None Milon Waththuhewa, PharmD, MSc ICER None Remziye Zaim, MSc, MS ICER None ©Institute for Clinical and Economic Review, 2018 Page 170 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents Table G2. New England CEPAC COI Disclosures Name Organization Disclosures Robert H. Aseltine, Jr., PhD UCONN Health * Stacey L. Brown, PhD University of Connecticut School of Medicine * Austin Frakt, PhD Boston University School of Medicine and School of * Public Health Marthe Gold, MD, MPH New York Academy of Medicine * Claudia B. Gruss, MD, FACP, FACG Western Connecticut Medical Group. Wilton, * Connecticut Claudio W. Gualtieri, JD AARP * Stephen Kogut, PhD, MBA, RPh University of Rhode Island College of Pharmacy * Stephanie Nichols, PharmD, BCPS, BCPP Husson University * Julia Prentice, PhD VA Boston Healthcare System * Jeanne Ryer, MSc, EdD New Hampshire Citizens Health Initiative * Jason Wasfy, MD, MPhil Massachusetts General Hospital * Edward Westrick, MD, PhD Comprehensive Community Action Program * Rev. Albert Whitaker, MA American Diabetes Association * * 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. Table G3. Policy Roundtable Participant COI Disclosures Policy Roundtable Title and Affiliation Conflict of Interest Participant Barbara Henry, RPh Lead Pharmacy Specialist, Harvard Pilgrim Health Care Full-time employee of Harvard Pilgrim Health Care. Kimberly Lenz, PharmD Clinical Pharmacy Manager, MassHealth None declared. Richard Malamut, MD Chief Medical Officer, Braeburn Full-time employee of Braeburn. Lewis Nelson, MD Professor and Chair, Department of Emergency None declared. Medicine; Chief, Division of Medical Toxicology, Rutgers New Jersey Medical School Amy O’Sullivan, PhD Head of Health Economics and Outcomes Research, Full-time employee of Alkermes. Alkermes Maria Schiff, MPH Senior Officer, Substance Use Prevention and Treatment None declared. Initiative, The Pew Charitable Trusts Ann Wheeler, PharmD, National Director of Managed Care Medical Affairs; Head Full-time employee of Indivior. BCPP of Behavioral Health Medical Affairs, Indivior Joe Wright, MD Medical Director, Boston Health Care for the Homeless Received consultancy fees from Program; Clinician, CareZone Massachusetts League of Community Health Centers. ©Institute for Clinical and Economic Review, 2018 Page 171 Final Evidence Report - MAT in Patients with OUD Return to Table of Contents