/ (BUP/SAM) NDA 210,417

Psychopharmacologic Advisory Committee and the Safety and Risk Management Advisory Committee , Inc.

1 Introduction

Lisa von Moltke, MD Senior Vice President, Clinical Development Alkermes, Inc.

2 Agenda Introduction Lisa von Moltke, MD Senior Vice President, Clinical Development Alkermes, Inc. The Unmet Need in MDD George Papakostas, MD Harvard Medical School Challenges in MDD Clinical Trials Massachusetts General Hospital

Clinical Efficacy Jerald Schindler, DrPH Vice President, Biostatistics Alkermes, Inc. Clinical Safety Gary Bloomgren, MD Vice President, Drug Safety and Pharmacovigilance Risk Mitigation Strategies Alkermes, Inc.

Clinical Perspective and Benefit-Risk Profile Sanjay Mathew, MD Baylor College of Medicine Conclusion Lisa von Moltke, MD 3 Additional Participants External: Aparna Anderson, PhD Vice President and Chief Operating Officer Statistics Collaborative, Inc. Sponsor: Craig Hopkinson, MD Bhaskar Rege, PhD Chief Medical Officer Vice President, Clinical Georgianna Harris, PhD & Translational Medicine Senior Vice President, Regulatory Affairs Asli Memisoglu, ScD Barry Lubarsky, PhD Senior Director, Biostatistics Senior Director, Medical Affairs Arielle Stanford, MD Sanjeev Pathak, MD Medical Director, Clinical Development Vice President, Clinical Development Psychiatry 4 Key Topics

• Significant unmet need and treatment challenges for patients with MDD – Need for new differentiated mechanisms • Significant development challenges of studying MDD – Utilization of Sequential Parallel Comparison Design (SPCD) methodology • Positive Benefit-Risk for Buprenorphine/Samidorphan (BUP/SAM or ALKS 5461) – Substantial evidence of efficacy – Well characterized and manageable safety profile – Low potential for abuse • Approving a new modulator in the midst of an opioid crisis – Committed to responsible risk mitigation and commercial distribution 5 Buprenorphine/Samidorphan (BUP/SAM)

• and BUP in particular may have effects – SAM mitigates the abuse and dependence potential • Proposed for adjunctive treatment of major depressive disorder (MDD) – Patients not responding to available antidepressant therapies • Proposed therapeutic dose – 2 mg BUP/ 2 mg SAM (BUP/SAM 2/2) sublingual tablet • Following a 1-week titration (0.5/0.5 for 3 days, 1/1 for 4 days) – 1 mg BUP/ 1 mg SAM (BUP/SAM 1/1) for special populations 6 BUP/SAM Regulatory History Through Development

• Pre-IND Meeting - February 2011 – Initial discussions regarding study designs (SPCD) • End-of-Phase 2 Meeting - October 2014 – Clinical study designs • Fast Track Designation - October 2014 – BUP/SAM has the potential to address an unmet need for a serious condition based on results of the Phase 2 study • Two Scientific Exchange Meetings - September 2016 & February 2017 – Share results of Phase 3 studies • Pre-NDA Meeting – July 2017 – NDA submission content 7 Major Depressive Disorder (MDD) • Major source of morbidity and disability – Carries a risk of suicide • Significant percentage of patients do not achieve adequate symptom relief with standard • All approved antidepressants work via mechanisms – Serotonin, norepinephrine, dopamine • Only antipsychotics are approved as adjunctive therapies – Can have significant side effects • New approaches to treat depression are urgently needed 8 BUP/SAM Program Founded on Evidence of Antidepressant Effect of Opioids

• Opioid receptors highly expressed in brain regions associated with emotional regulation1 • Clinical history of efficacy with opioids, particularly BUP, in patients with MDD2 – Broader use limited by abuse and dependence potential

1. Nummenmaa and Tuominen. Br J Pharmacol. 2017.doi:10.1111/bph.13812 2. Emrich et al. Neuropsychopharmology. 2015;40(6):1448-1455; Bodkin et al. J Clin Psychopharmacol. 1995;15(1):49-57; Nyhuis et al. J Clin Psychopharmacol. 2008;25(5):593-595; Karp et al. J Clin Psychiatry. 2014;75(8):e785-e793; & Yovell et al. Am J Psychiatry. 2016; 173(5):491-498. 9 BUP/SAM for Adjunctive Treatment of MDD

BUP SAM BUP/SAM

Buprenorphine (BUP) Samidorphan (SAM) Co-formulation – New molecular in a single – Partial µ-opioid entity sublingual tablet and κ- – Potent µ-opioid – Bioavailable sublingually antagonist – Optimized for: • High potency • High bioavailability with oral and sublingual administration

10 Extensive Safety and Efficacy Evaluation of BUP/SAM

• 34 clinical studies – 19 conducted with BUP/SAM and 15 with SAM – Four placebo-controlled studies in MDD patients • Studies 202, 205, 206, & 207 – Long-term study (Study 208)

• 2165 subjects received BUP/SAM – 1531 MDD patients treated with 2/2 dose • 947 patients treated ≥6 months • 743 patients treated ≥12 months

11 Study Population for the Development Program • Diagnosed with MDD – Mean lifetime number of major depressive episodes (MDEs): 4-7 – Cycled through multiple therapies – Median duration of current MDE: 9 – 10 months

• 1-2 inadequate responses to an antidepressant therapy (ADT) in current MDE

• Continued background ADT – SSRI, SNRI, or bupropion

12 Concluding Remarks • Efficacy – Two studies (202 and 207) met the pre-specified primary endpoint – Supportive evidence of efficacy from a third study (205) – Clinically meaningful efficacy • Safety – Generally well-tolerated and common adverse events were gastrointestinal or sedation related – Low abuse potential • Favorable benefit-risk profile for adjunctive treatment of MDD • Robust risk mitigation plan due to presence of BUP

13 The Unmet Need in MDD

George Papakostas, MD Associate Professor, Harvard Medical School Director, Treatment-Resistant Depression Studies Depression and Clinical Research Program Massachusetts General Hospital

DISCLOSURE: Consultant to Alkermes. Compensated for time and travel.

14 MDD Is a Serious, Life-threatening Disease • Lifetime prevalence of MDD in adults in the US is 16.6%1 • Significantly impacts home & work life, relationships, social life2 • 59.3% of people with 12-month MDD reported either severe or very severe role impairment3 • ~35 days a year unable to work or participate in normal activities3 • MDD is the #1 contributor to disability worldwide4 • Burden of MDD includes increases in the relative risk of various diseases5,6 • Such as heart disease, diabetes mellitus, and cancer • Patients with MDD are at increased risk for suicide7

1. Kessler RC et al. Epidemiol Psychiatr Sci 2015;24:210-226. 2. Nierenberg AA, et al. J Clin Psychiatry 1999;60:221–225. 3. Kessler RC et al. JAMA 2003;289:3095-3105. 4. World Health Organization. The global burden of disease. http://www.who.int/healthinfo/global_burden_disease/GBD_report_2004update_full.pdf Accessed August 16, 2017. 5. Otte C et al. Nat Rev Dis Primers;2:16065. 6. Pennix BWJH et al. BMC Med 2013;11:129. 7. Ferrari AJ et al. PLOS Med 2013;10:e1001547 15 Switching Antidepressant Treatments Is Increasingly Futile in Addressing Symptoms Proportion of Patients Remaining Symptomatic By Line of Therapy

100 91.7 87.1 84.0 85.5 80 75.0 74.5 73.4 63.2 60

Rate, % Rate, 40

20

0 citalopram venlafaxine-XR bupropion-SR sertraline venlafaxine-XR tranylcypromine mirtazapine Step 1 Step 2 Step 3 Step 4

Rush AJ et al. Am J Psychiatry 2006;163:1905-1917 16 Persistent Symptoms Can Increase Risk of Relapse, Among Other Serious Consequences

Patients with persistent symptoms Patients who remain symptomatic despite are at greater risk for early relapse treatment …

1.0 Without persistent symptoms (n=197) Weeks Well, With persistent symptoms (n=125) Median (95% CI) • Are ~2X more likely to be hospitalized 220 (151-383) 0.8 88 (60-121) • Have ~2X higher odds of suicide attempt

Function 2 Log Rank X1 =22.30;P <.0001 0.6 • Use up to 3X more psychotropic medications, including antidepressants 0.4 val val Distribution 0.2 … compared to the overall MDD patient Survi population 0.0 0 5 10 15 20 25 30 Years Crown WH, et al. J Clin Psychiatry 2002;63:963-71 Judd LL et al. J Clin Psychiatry 2016;77:1065-1073 Souery D, et al. J Clin Psychiatry 2007;68:1062-70 17 Reasons for Incomplete Symptom Control Unknown • MDD is a heterogeneous disease likely encompassing multiple pathologies1 • Currently approved pharmacotherapies for MDD all target monoamine neurochemical pathways2 Year Serotonin Norepinephrine Dopamine Introduced

SSRI 1987 (eg, fluoxetine, escitalopram)

SNRI 1993 (eg, venlafaxine, duloxetine)

TCA / MAOI 1950s (eg, nortriptyline, selegiline) 1. Fried EI, et al. J Affect Disord 2015;172:96-1022. 2. Hillhouse TM, et al. Exp Clin Psychopharmacol 2015;23:1-21. 18 Pharmacologic Augmentation Is Often the Next Approach • Approved adjunctive options are atypical antipsychotics – Also work through the monoamine pathway – Demonstrated efficacy in MDD • Atypical antipsychotics are associated with potentially significant side effects that seriously discourage their use – Risk of weight gain and new-onset diabetes – Often associated with significant sedation – Frequently accompanied by movement disorders • Akathisia • Parkinsonism – Tardive dyskinesia – Potentially life-threatening neuroleptic malignant syndrome

REXULTI (brexpiprazole) Prescribing Information. https://www.accessdata.fda.gov/drugsatfda_docs/label/2015/205422s000lbl.pdf. REXULTI (brexpiprazole) Summary Basis of Approval. https://www.accessdata.fda.gov/drugsatfda_docs/nda/2015/205422Orig1Orig2s000MedR.pdf. D’Abreu A, Akbar U, Friedman JH. J Neurological Sci 2018;389;17-20. Sarkar S, Gupta N. BJ Psych Bulletin 2017;41:211-216. 19 Summary of Unmet Need • Treatment of MDD is challenging1 • Despite available therapies, many patients experience persistent symptoms1,2 • Patients with persistent symptoms have a higher burden of illness than other patients with MDD3 • Current therapies for MDD all target monoamine signaling,4 underscoring the need for alternative mechanisms of action • Desired characteristics of such a therapy include – Clinically meaningful symptom improvement – Safety profile with good tolerability and manageable adverse effects 1. Rush AJ, et al. Am J Psychiatry 2006;163:1905-1917; 2. Trivedi MH, et al. Am J Psychiatry 2006;163:28-40; 3. Crown WH, et al. J Clin Psychiatry 2002;63:963-71; 4. Sanacora G, et al. Neuropharmacology 2012;62:63-77 20 Challenges in MDD Clinical Trials

George Papakostas, MD Associate Professor, Harvard Medical School Director, Treatment-Resistant Depression Studies Depression and Clinical Research Program Massachusetts General Hospital

21 High Historical Failure Rate of MDD Trials FDA Analysis

Nearly half of MDD trials of approved drugs have been unsuccessful

Successful trials Failed trials 47% 53%

1983 - 2008 Khin et al. J Clin Psychiatry 2011;72(4):464-472 22 Evolution of Trial Design to Improve Efficiency of Signal Detection in Adjunctive MDD Clinical Trials

Sequential Parallel Comparison Design (SPCD) Placebo run-in (eg, BUP/SAM) (eg, aripiprazole, brexpiprazole) Simple randomization SPCD combines elements of both simple randomization and placebo (eg, quetiapine XR) run-in designs 23 Placebo Run-in Design (eg, aripiprazole, brexpiprazole)

Response Placebo + Background ADT

R Active + Background ADT Placebo + Background ADT a n d No o Response m i z e Placebo + Background ADT

ADT: Antidepressant therapy. 24 Sequential Parallel Comparison Design (SPCD)

Stage 1 Stage 2

Active + Background ADT Active + Background ADT R a n d o (All-comers m population) Response Placebo + Background ADT i z e R Active + Background ADT Placebo + Background ADT a n d No o Response m i z e Placebo + Background ADT ADT: Antidepressant therapy. Fava M et al; Psychother Psychosom 2003:72:115-27 25 SPCD Is Becoming Increasingly Utilized in Studies of CNS Disease, Particularly MDD Intervention Indication Context N Completion Sponsor(s) Reference L-methylfolate MDD Adjunct 148 2008 MGH; Pamlab NCT00321152 Aripiprazole MDD Adjunct 225 2009 MGH; BMS NCT00683852 Ziprasidone MDD Monotherapy 120 2010 MGH; Pfizer NCT00555997 L-methylfolate MDD Adjunct 75 2011 MGH; Pamlab NCT00955955 JNJ-40411813 MDD Adjunct 121 2013 Janssen NCT01582815 EB-1010 MDD Adjunct 342 2013 Euthymics NCT01318434 CERC-301 MDD Adjunct 1357 2014 Cerecor NCT01941043 AVP-923 Alzheimer's —— 220 2014 Avanir NCT01584440 Riluzole MDD Adjunct 104 2015 Yale; NIMH NCT01204918 Esketamine MDD Adjunct 108 2015 Janssen NCT01998958 PF-06372865 Generalized Anxiety Adjunct 90 2015 Pfizer NCT02310568 Vortioxetine ADHD —— 227 2016 Lundbeck NCT02327013 AVP-786 MDD Adjunct 206 2016 Avanir NCT02153502 Low field magnetic stimulation MDD Adjunct 84 2016 MGH; NIMH NCT01654796 Sodium nitroprusside Single IV 60 2017 MGH NCT02164981 NSI-189 phosphate MDD Monotherapy 220 2017 Neuralstem NCT02695472 Pimavanserin MDD Adjunct 207 2018 ACADIA NCT03018340 26 Summary of Challenges • Demonstrating antidepressant efficacy in a clinical trial setting is difficult • Evolution in clinical trial design is needed to better enable signal detection in MDD • SPCD, which combines simple randomization and placebo run-in designs, is increasingly becoming a valuable methodology in psychiatric disease trials

27 Clinical Efficacy

Jerald Schindler, DrPH Vice President, Biostatistics Alkermes, Inc.

28 BUP/SAM 2/2 is Effective for the Adjunctive Treatment of MDD • Substantial evidence of efficacy demonstrated across the clinical development program

• Four randomized, double-blind, placebo-controlled studies of BUP/SAM + antidepressant therapy (ADT) vs. placebo + ADT

• Two of the four studies (202 and 207) met the primary endpoint

• One study provided supportive evidence of efficacy (205)

• One study did not support efficacy (206) 29 BUP/SAM Clinical Efficacy

• Study Designs and Analysis Methods

• Primary Efficacy Results

• Comparison Across Studies with Common Endpoints

• Conclusions

30 BUP/SAM 2/2 Evaluated in Four Double-blind Placebo-controlled Studies

Primary Number Phase Study Design Doses Assessment Randomized 2/2 2 202 SPCD HAM-D17 142 8/8 0.5/0.5 3 205 SPCD MADRS-10 385 2/2

3 206 Placebo Run-In 2/2 MADRS-10 297

1/1 MADRS-6* 3 207 SPCD 407 2/2 MADRS-10

* Basis for concluding efficacy

SPCD: Sequential Parallel Comparison Design HAM-D17: 17-item Hamilton Depression Rating Scale MADRS: Montgomery-Asberg Depression Rating Scale 31 202, 205, and 207 Study Design (SPCD) STAGE 1 (2:2:9) STAGE 2 (1:1:1) BUP/SAM 2/2 + ADT BUP/SAM 2/2 + ADT R a BUP/SAM Other + ADT BUP/SAM Other + ADT n d o m i Response z Placebo + ADT e R Placebo + ADT a BUP/SAM 2/2 + ADT n No d Response o BUP/SAM Other + ADT m i z e Placebo + ADT Week Week 0 1 2 3 4* 5 0 1 2 3 4* 5 6 Note: For Study 202, patients who received BUP/SAM in Stage 1 were switched to placebo in Stage 2. * 202 End of Treatment 32 Methods for Statistical Analysis • Longitudinal studies – Weekly efficacy assessments • Mixed Models Repeated Measures (MMRM) – Uses all available longitudinal data without imputation – At each time point, model estimates: • Mean change from baseline by treatment group • Mean difference in change from baseline between BUP/SAM vs. Placebo – Statistical test to determine if difference between BUP/SAM vs. Placebo = 0 • SPCD studies – Difference estimates combined across stages – Weights pre-specified for Stage 1/Stage 2 – Uses data from all subjects 33 202, 205, and 207 Study Design (SPCD) STAGE 1 (2:2:9) STAGE 2 (1:1:1) BUP/SAM 2/2 + ADT R a BUP/SAM Other + ADT n d o m i z e R Placebo + ADT a BUP/SAM 2/2 + ADT n No d Response o BUP/SAM Other + ADT m i z e Placebo + ADT Week Week 0 1 2 3 4* 5 0 1 2 3 4* 5 6 Note: For Study 202, patients who received BUP/SAM in Stage 1 were switched to placebo in Stage 2. * 202 End of Treatment 34 202, 205, and 207 Study Design (SPCD) STAGE 1 (2:2:9) STAGE 2 (1:1:1) BUP/SAM 2/2 + ADT R a BUP/SAM Other + ADT n d o m i z e R Placebo + ADT a BUP/SAM 2/2 + ADT n No d Response o BUP/SAM Other + ADT m i z e Placebo + ADT Week Week 0 1 2 3 4* 5 0 1 2 3 4* 5 6 Note: For Study 202, patients who received BUP/SAM in Stage 1 were switched to placebo in Stage 2. * 202 End of Treatment 35 BUP/SAM Clinical Efficacy

• Study Designs and Analysis Methods

• Primary Efficacy Results

• Comparison Across Studies with Common Endpoints

• Conclusions

36 Primary Endpoint BUP/SAM vs. Placebo Difference Consistent Improvement Across Studies Pre-specified Primary Endpoint P-value

202 HAM-D17EOT 0.014

SPCD 205 MADRS-10Week 5 0.109

207 MADRS-6AVG 0.018

Placebo 206 MADRS-10 0.782 Run-In EOT

-6 -5 -4 -3 -2 -1 0 1 2 Favors Favors  BUP/SAM Placebo  +ADT +ADT BUP/SAM 2/2 vs. Placebo Least Squares Mean Difference (95% CI) 37 Study 202 (Phase 2) • Randomized, double-blind, placebo-controlled SPCD study

• Two BUP/SAM doses (2/2 and 8/8)

• Two 4-week treatment periods

• 142 patients randomized

• Primary endpoint: HAM-D17 (Week 4)

• MADRS-10 included as secondary endpoint

38 Study 202: Primary Efficacy Results – HAM-D17EOT Stage 1 Stage 2

e Placebo n i l

e +ADT s ) a n

HAM-D17 B BUP/SAM 2/2 a

e Change from Baseline m +ADT M o

r f S by Treatment L e ( g n a h C

Week Week

Placebo, n 95 94 92 93 90 20 20 20 20 20

BUP/SAM, n 20 20 17 17 17 23 23 19 18 18

BUP/SAM 2/2 Least Squares Mean Difference vs. Placebo (95% CI) BUP/SAM 8/8

39 Study 202: Multiplicity Adjustment • No pre-specified multiplicity adjustment • Comparison of each dose of BUP/SAM vs. placebo – 2/2: P = 0.014 – 8/8: P = 0.699 • Post-hoc Bonferroni multiplicity adjustment splits the alpha in half – 2/2 dose still significantly different from placebo • Bonferroni adjustment very conservative and requires no prior assumption • No prior suggestion that either BUP/SAM dose was more likely to be successful, an ordered hierarchical adjustment not appropriate – Complex pharmacology of BUP/SAM (agonist/antagonist) – Uncertain dose response 40 Study 202: Single Responder Does Not Drive Efficacy Results • FDA briefing book presents analysis excluding a single responder – No legitimate reason to remove this patient – Contrary to intent-to-treat principle • Analyses show consistent results without this patient’s data HAM-D17 Change from Baseline to Week 4

BUP/SAM 2/2 vs. Placebo Least Squares Mean Difference (95% CI) 41 Study 205 (Phase 3) • Randomized, double-blind, placebo-controlled SPCD study

• Two BUP/SAM doses (2/2 and 0.5/0.5)

• Two treatment periods – Stage 1: 5 weeks – Stage 2: 6 weeks

• 385 patients randomized

• Primary endpoint: MADRS-10 (Week 5) 42 Study 205: Primary Efficacy Results – MADRS-10Week 5 Stage 1 Stage 2

Placebo MADRS-10 +ADT Change from Baseline BUP/SAM 2/2 by Treatment +ADT

Week Week

Placebo, n 256 256 255 253 250 247 54 54 53 53 53 53 51

BUP/SAM, n 59 59 57 54 53 52 54 54 53 52 50 49 49

P-value

BUP/SAM 2/2 0.109 Least Squares Mean Difference vs. Placebo (95% CI) BUP/SAM 0.5/0.5 0.975

-5 -4 -3 -2 -1 0 1 2 3 Favors Favors BUP/SAM Placebo +ADT +ADT 43 Insights from Study 205

• Supportive evidence of BUP/SAM 2/2 efficacy

• Average (multiple time points) vs. end of treatment (EOT, single time point) – Reduces influence of week-to-week variability – Reflects patient’s experience over time

• MADRS-6 vs. MADRS-10 – MADRS-6 is a sensitive measure of depression symptom severity

44 Study 207 (Phase 3) • Randomized, double-blind, placebo-controlled SPCD study • Two BUP/SAM doses (2/2 and 1/1) • Two treatment periods – Stage 1: 5 weeks – Stage 2: 6 weeks • 407 patients randomized • Primary Endpoints

− MADRS-6AVG (Week 3 to EOT) – basis for concluding efficacy

− MADRS-10AVG (Week 3 to EOT)

− MADRS-10EOT − Hypothesis testing BUP/SAM 2/2 then 1/1 45 Study 207: Primary Efficacy Results Stage 1 Stage 2

Placebo MADRS-6 +ADT Change from Baseline BUP/SAM 2/2 by Treatment +ADT

Week Week Placebo, n 273 273 270 265 263 257 60 60 60 59 58 57 56 BUP/SAM, n 63 63 59 53 49 49 63 63 61 59 58 58 58

Endpoint P-value

BUP/SAM 2/2 Least Squares Mean Difference vs. Placebo (95% CI) BUP/SAM 1/1

46 Study 207: Durability of Effect MADRS-10 Change From Baseline by Visit Stage 1 Stage 2 0 Placebo aseline -2 BUP/SAM 2/2 -4

-6

(Mean) -8 nge from Stage 1 B fromStage 1 nge -10 10 Cha 10 - -12

-14

MADRS Baseline 1 2 3 4 5 1 2 3 4 5 6 Week

Placebo, n 273 273 270 265 263 257 BUP/SAM 2/2, n 63 63 59 53 49 49 48 46 45 44 44 42 47 Study 206 (Phase 3) • Randomized, double-blind, placebo-controlled, placebo run-in study

• One BUP/SAM dose (2/2)

• Two treatment periods – Stage 1: 4-week placebo run-in – Stage 2: 6-week treatment period

• 297 patients randomized

• Primary endpoint: MADRS-10 (EOT) 48 Study 206: Primary Efficacy Results – MADRS-10EOT

Placebo e n i

l +ADT e s )

a BUP/SAM 2/2 n

MADRS-10 B a +ADT e m M o

Change from Baseline r f S

L e ( by Treatment g n a h C

Week

Placebo, n 146 146 144 140 140 140 138

BUP/SAM, n 142 142 141 138 136 134 131

Least Square Mean Difference BUP/SAM 2/2 vs. Placebo

49 BUP/SAM Clinical Efficacy

• Study Designs and Analysis Methods

• Primary Efficacy Results

• Comparison Across Studies with Common Endpoints

• Conclusions

50 Common Endpoints Used for Comparison Across Studies • All four studies included weekly MADRS-10 assessments

• MADRS-10AVG – Based on multiple time points – Average of differences from MMRM model – Reduces influence of week-to-week variability – Reflects patient’s experience over time

• MADRS-10EOT – Based on single time point – Differences from MMRM model – Conventional endpoint 51 Consistent Improvement for BUP/SAM Across All SPCD Studies and Stages - MADRS-10 Stage 1 Stage 2 Placebo +ADT 202 BUP/SAM 2/2 +ADT

MADRS-10 Change from Baseline 205 by Treatment

207

Note: Only Stage 1 placebo non-responders are included in Stage 2 Week Week 52 BUP/SAM vs. Placebo Difference Consistent Improvement Across Studies Common Endpoint and Analysis - MADRS-10AVG

202

205

207

206

Meta-Analysis

-7 -6 -5 -4 -3 -2 -1 0 1 2 Favors Favors BUP/SAM Placebo +ADT +ADT BUP/SAM 2/2 vs. Placebo Least Squares Mean Difference (95% CI) 53 BUP/SAM vs. Placebo Difference Consistent Improvement Across Studies Common Endpoint and Analysis - MADRS-10EOT

54 BUP/SAM Clinical Efficacy

• Study Designs and Analysis Methods

• Primary Efficacy Results

• Comparison Across Studies with Common Endpoints

• Conclusions

55 Conclusion: Consistent Efficacy for BUP/SAM 2/2

Pre-specified • 2 of the 4 trials met their primary Primary endpoint endpoint

• 3 of the 4 trials demonstrate MADRS-10 AVG evidence of efficacy

• 2 of the 4 trials demonstrate MADRS-10EOT evidence of efficacy

Overall, these show that BUP/SAM 2/2 is effective for the treatment of MDD 56 Clinical Safety

Gary Bloomgren, MD Vice President, Drug Safety and Pharmacovigilance Alkermes, Inc. 57 BUP/SAM Safety Overview

• Exposure and Pooling Strategy

• Adverse Event (AE) Profile

• Laboratory/Vital Signs/ECGs

• Topics of Special Interest

• Abuse Potential Assessment

• Summary 58 BUP/SAM Patient Exposure

All Populations ≥1 Dose ≥6 Months ≥12 Months Any Dose 2165 2/2 Dose 1860

MDD Population 2/2 Dose 1531 947 743

Total MDD patient exposure >1100 years

59 Safety Pooling Strategy

• Four placebo-controlled studies

• Each study had two randomizations – 1st at treatment start (Stage 1) – 2nd mid study (Stage 2, Stage 1 Placebo non-responders)

• Patients pooled by randomization time point

60 BUP/SAM Safety Overview

• Exposure and Pooling Strategy

• Adverse Event (AE) Profile

• Laboratory/Vital Signs/ECG

• Topics of Special Interest

• Abuse Potential Assessment

• Summary 61 Treatment Emergent Adverse Events (TEAE) Were Generally Tolerability-Related

Stage 1 Stage 2 Placebo-Controlled BUP/SAM 2/2 Placebo BUP/SAM 2/2 Placebo Studies Incidence ≥5% (N=289) (N=286) and >Placebo (N=162) (N=658) n (%) n (%) n (%) n (%) Any TEAE 111 (68.5) 358 (54.4) 137 (47.4) 119 (41.6) Nausea 43 (26.5) 46 (7.0) 36 (12.5) 5 (1.7) Dizziness 21 (13.0) 27 (4.1) 9 (3.1) 6 (2.1) Constipation 20 (12.3) 18 (2.7) 11 (3.8) 2 (0.7) Headache 17 (10.5) 59 (9.0) 10 (3.5) 13 (4.5) Vomiting 16 (9.9) 11 (1.7) 14 (4.8) 4 (1.4) Fatigue 12 (7.4) 10 (1.5) 10 (3.5) 6 (2.1) 11 (6.8) 22 (3.3) 3 (1.0) 1 (0.3) Sedation 11 (6.8) 6 (0.9) 4 (1.4) 0 Dry mouth 10 (6.2) 29 (4.4) 7 (2.4) 4 (1.4) • Majority were mild or moderate in severity and tended to occur with treatment initiation • No meaningful differences by gender, age, race, concomitant antidepressant or benzodiazepine use • No new findings for long-term study 62 Few Adverse Events (AEs) Leading to Discontinuation

Stage 1 Stage 2

BUP/SAM 2/2 Placebo BUP/SAM 2/2 Placebo Placebo-Controlled Studies (N=162) (N=658) (N=289) (N=286) Incidence ≥2% n (%) n (%) n (%) n (%)

Any AE Leading to Study 22 (13.6) 13 (2.0) 11 (3.8) 4 (1.4) Discontinuation

Nausea 8 (4.9) 2 (0.3) 2 (0.7) 0

Vomiting 5 (3.1) 0 3 (1.0) 0

Dizziness 5 (3.1) 0 1 (0.3) 0

• Similar findings for long-term study 63 Few Serious Adverse Events (SAEs)

• Placebo-controlled studies – 1.9% with BUP/SAM 2/2 vs. 0.5% in placebo – No deaths • Long-term study – 3.2% of patients – Most common events: depression and suicidal ideation (each 0.2%) – Two deaths (not related) – Patient with COPD who died of respiratory arrest 47 days after last dose – Patient with hypertension and CHF who died of a cerebral hemorrhage (Day 87 of study). Family history of cerebrovascular hemorrhage

64 BUP/SAM Safety Overview

• Exposure and Pooling Strategy

• Adverse Event (AE) Profile

• Laboratory/Vital Signs/ECG

• Topics of Special Interest

• Abuse Potential Assessment

• Summary 65 No Clinically Relevant Laboratory, Vital Sign, Weight, or ECG Changes

Placebo-controlled and long-term studies:

• No meaningful post-baseline changes or outliers

• No evidence of risk of QT prolongation associated with BUP/SAM

66 BUP/SAM Safety Overview

• Exposure and Pooling Strategy

• Adverse Event (AE) Profile

• Laboratory/Vital Signs/ECG

• Topics of Special Interest

• Abuse Potential Assessment

• Summary 67 Topics of Special Interest

• CNS Depression – Mild to moderate sedation/somnolence with treatment initiation, resolved with continued use • Precipitated opioid withdrawal – One case occurred with first dose of BUP/SAM in patient with undisclosed pre-existing opioid dependence, event was serious and related

68 Additional Topics of Special Interest No Evidence of: • Respiratory Depression

• Hypotension/Orthostatic Hypotension Associated with BUP

• Hepatic Injury

• Hypomania/Mania

Associated with • Sexual Dysfunction Antidepressants • Suicidal Ideation or Behavior 69 Less Suicidal Ideation or Behavior with BUP/SAM Treatment vs. Placebo

Placebo-Controlled Studies

Stage 1 Stage 2

BUP/SAM 2/2 Placebo BUP/SAM 2/2 Placebo Long-term Study Post-Baseline C-SSRS (N=162) (N=658) (N=289) (N=286) (N=1454) Categories n (%) n (%) n (%) n (%) n (%)

Suicidal Behavior 0 2 (0.3) 0 1 (0.3) 1 (0.1)

Suicidal Ideation 16 (9.9) 107 (16.3) 27 (9.3) 42 (14.7) 152 (10.5)

Self-injurious Behavior 0 2 (0.3) 0 0 6 (0.4) without Suicidal Intent

Columbia-Suicide Severity Rating Scale (C-SSRS) 70 BUP/SAM Safety Overview

• Exposure and Pooling Strategy

• Adverse Event (AE) Profile

• Laboratory/Vital Signs/ECG

• Topics of Special Interest

• Abuse Potential Assessment

• Summary 71 Evaluation of Abuse Potential

• Dedicated Human Abuse Potential Study

• In MDD Studies – Evaluation of Adverse Events of Special Interest (AESI) related to abuse potential, dependence and withdrawal – Clinical Opioid Withdrawal Scale following BUP/SAM Discontinuation

72 Human Abuse Potential of BUP/SAM is Low At the Moment Drug Liking

2/2 *Margin of clinical significance for BUP/SAM vs. Placebo

• 6-way crossover study BUP/SAM 8/8 • Non-dependent recreational opioid users 16/16 • 38 subjects

8 mg BUP 16 mg

-5 0 5 1 0 1 5 2 0 2 5 3 0 3 5 4 0 Difference vs. Placebo (90% CI)

Emax Visual Analogue Scale

*Assessment of Abuse Potential of Drugs Guidance for Industry 2017 (FDA) Chen and Bonson; J Biopharm Stat. 2013;23(2):294-306. 73 Endpoints Predictive of Real-world Abuse Liability Show BUP/SAM Similar to Placebo

2/2 2/2 BUP/SAM 8/8 BUP/SAM 8/8 16/16 16/16 8 mg 8 mg BUP BUP 16 mg 16 mg

• Both measures assessed global drug effects (i.e., the entire drug experience) • Further supports assessment of low abuse potential with BUP/SAM

White et al. J Opioid Manage. 2015; 3:199-210. Miller et al. J Opioid Manage. 2017;13(6):379-389. Setnik et al. Pain Res Manag 2013;18(4):e55-62. 74 Consistent Evidence of Low Abuse Potential in MDD Studies • Abuse potential terms query – Majority of events non-specific to abuse potential (i.e., dizziness, somnolence, and sedation) – Low incidence of -related events: • BUP/SAM 2/2 (1.6%) vs. Placebo (0.2%) – Stage 1 and 2 combined • Incidence in long-term study similar (1.2%) • Majority associated with first dose, none recurred • No abuse behavior • No evidence of dependence 75 Little Evidence of Withdrawal: Post-discontinuation Change in COWS Scores • Mean post-discontinuation scores were ≤1 in all treatment groups in placebo-controlled studies and the long-term study

Placebo-controlled Studies Long-term Study BUP/SAM 2/2 Placebo BUP/SAM 2/2 COWS Score Category* (N=113) (N=148) (N=831) No Withdrawal (0-4) 96.5% 97.3% 94.3% Mild Withdrawal (5-12) 2.7% 2.7% 4.9% Moderate Withdrawal (13-24) 0.9% 0 0.7% Moderate-severe or Severe 0 0 0 Withdrawal (25-48)

Abbreviation: COWS=Clinical Withdrawal Scale * Patients with score of “No Withdrawal” at end of treatment period 76 BUP/SAM Safety Overview

• Exposure and Pooling Strategy

• Adverse Event (AE) Profile

• Laboratory/Vital Signs/ECG

• Topics of Special Interest

• Abuse Potential Assessment

• Summary 77 BUP/SAM General Safety Summary • Common AEs

– Gastrointestinal or sedation-related

– Mild/moderate in severity

– Typically occurred with treatment initiation • No clinically meaningful changes in laboratory, vital signs, weight or ECGs • No evidence of increased treatment-emergent suicidal ideation or behavior 78 BUP/SAM Abuse Potential is Low • Human Abuse Potential Study – Abuse potential with BUP/SAM 2/2 is similar to placebo – 4-8x the therapeutic dose of BUP/SAM • Slightly greater than placebo • Significantly less than equivalent dose of BUP alone • MDD data consistent – Low incidence of euphoria, typically with first dose/none recurrent – No evidence of dependence during treatment – Little evidence of withdrawal, discontinuation well-tolerated 79 Risk Mitigation Strategy

Gary Bloomgren, MD Vice President, Drug Safety and Pharmacovigilance Alkermes, Inc. 80 Alkermes’ Commitment to Responsible Use • Product Characteristics – Selection of lowest effective dose of BUP (2 mg) and addition of SAM to mitigate risk of abuse and dependence – Co-formulated as a homogenous mixture of BUP and SAM – Packaged in cards with individual tablets blistered – Packaged to limit unintentional pediatric exposure • Continuous monitoring at multiple points in the distribution channel (wholesaler, third party logistics provider and manufacturer) • Educate prescribers, pharmacists and patients • Implement appropriate post-marketing safety initiatives and risk mitigation strategies 81 Alkermes’ Proposed Education Plan • Objective: Alert health care professionals and patients to important safety information and appropriate use − Risks and precautions associated with concomitant use of opioids − Instructions for safe use, storage and disposal − Risks of abuse, misuse and diversion − Risks of addiction, overdose and death

Prescribers, Nurses & Pharmacists Patients • Product Label • Medication Guide • Product REMS • Medication ID Wallet Card • Selection of appropriate patients • Product Website • Patient educational materials • Patient Provider Discussion Guide – Patient Counseling Tool • Ongoing product information for patients – Placebo Training Tool who opt in • Trained medical staff available to respond to queries via call center or in person • Healthcare professional product website • Multi-media information/awareness campaigns 82 Proposed BUP/SAM Risk Evaluation and Mitigation Strategy (REMS) Objectives: Mitigate risk of misuse and accidental exposure BUP/SAM Medication Guide ✔ REMS Website & Call Center ✔ Communication Materials Health Care Provider (HCP) Letter ✔ HCP Brochure ✔ Appropriate Use Checklist ✔ Monitoring for misuse, abuse, dependence, or diversion (RADARS*) ✔ Identify new HCPs annually For 3 years At 18 months, 3 Submission of REMS assessments to FDA years, and 7 years *RADARS = Researched Abuse, Diversion and Addiction-Related Surveillance – Managed by the Rocky Mountain Poison and Drug Center 83 Clinical Perspective and Benefit-Risk Profile

Sanjay Mathew, MD Professor of Psychiatry and Behavioral Sciences Marjorie Bintliff Johnson and Raleigh White Johnson, Jr Chair for Research in Psychiatry Menninger Dept of Psychiatry & Behavioral Sciences Director, Mood Anxiety Disorders Program Baylor College of Medicine DISCLOSURE: Consultant to Alkermes. Compensated for time and travel. 84 BUP/SAM Would Be for Patients With MDD Who Are Candidates for Adjunctive Treatment General Patient Characteristics • Long-standing depression • Little or no success from several monotherapy antidepressants despite adequate dose & duration • Current antidepressant is addressing some symptomatology • Persistent symptoms significantly impact work or personal life • Patient is willing to add another treatment to their regimen

85 BUP/SAM Would Be for a Patient Like J.D.

• Mid-40s experiencing her 2nd major depressive episode • Antidepressant treatment history in current episode – Escitalopram 10-20mg for about 3 months • QIDS-SR improvement from 21 to 16 – Venlafaxine XR up to 225 to 300mg over about 3 months • Minimal to no improvement in QIDS-SR • Experienced concerning BP elevations at highest dose 86 Augmentation with BUP/SAM Gives J.D. an Opportunity to Address Her Persistent Symptoms • Depressive symptoms continue to significantly impact her home and personal life – Anhedonia, social withdrawal – Affects entire family • Unable to meet needs of 2 school-age children

87 BUP/SAM Offers Efficacy Comparable to a Recently Approved Adjunctive Treatment for MDD

BUP/SAM 2/2 (4 studies) Brexpiprazole (4 studies)

Standardized Effect Size (Hedges' g) (95% CI)

Favors Treatment And does so through a different mechanism of action Yoon S et al. J Clin Psychopharmacol 2017;37:46-53 88 BUP/SAM Would Be an Attractive Therapeutic Option for J.D. • Considered augmentation with aripiprazole or quetiapine • Her main concern is weight gain – Current BMI of 30 – Borderline diabetic • Also worried about somnolence – Caring for 2 young children

89 BUP/SAM Safety Summary

• Common adverse events are GI or sedation – Mainly associated with treatment initiation – Generally mild or moderate in severity

• Low potential for abuse – Human abuse potential study • BUP/SAM 2/2 similar to placebo – Clinical trial patients with MDD • Low incidence of euphoria • No dependence • Little evidence of withdrawal upon discontinuation 90 BUP/SAM Has a Positive Benefit-Risk Profile

Risk Considerations Benefit Considerations • Abuse liability related to BUP • Consistent antidepressant addressed effects • Adverse effects (eg, nausea, • Clinically meaningful efficacy in vomiting, constipation) are difficult-to-treat patients manageable • Efficacy consistent with • Lack of weight gain, metabolic available adjunctive therapies impact, and movement side effects – Accomplished via a different • No apparent precipitation of mechanism of action suicidality

91 Conclusions

• BUP/SAM has a positive benefit-risk profile • Efficacy comparable to current adjunctive medications, but through a different mechanism of action – Offers an alternative approach in patients who have not found adequate relief through monoaminergic therapies • Favorable safety profile • BUP/SAM would be an important new treatment option for patients who urgently need more options – Patients like J.D. 92 Conclusions

Lisa von Moltke, MD Senior Vice President, Clinical Development Alkermes, Inc.

93 Conclusions • Significant unmet need and treatment challenges for patients with MDD – Need for new differentiated mechanisms

• BUP/SAM works through opioid system modulation – Maintains the efficacy of BUP while mitigating the risk for abuse and dependence • Significant development challenges of studying MDD – Utilization of SPCD methodology • Positive Benefit-Risk for BUP/SAM – Substantial evidence of efficacy – Well characterized and manageable safety profile • Committed to responsible risk mitigation and commercialization 94 Backup Slides Shown BUP/SAM Offers Efficacy Comparable to a Recently Approved Adjunctive Treatment for MDD

BUP/SAM 2/2 (4 studies) Brexpiprazole (4 studies)

-0.2 0.0 0.2 0.4 0.6 Standardized Effect Size (Hedges' g) (95% CI)

Favors Treatment And does so through a different mechanism of action Yoon S et al. J Clin Psychopharmacol 2017;37:46-53 88 Sensitivity Analysis with Multiple Imputation 202 HAM‐D17EOT and MADRS‐10EOT

HAM-D17EOT MADRS-10EOT

MI

BUP/SAM 2/2 vs. Placebo Least Squares Mean Difference (95% CI) E-160 Single Patient with High Response (202) Efficacy Assessments

Stage 1 Stage 2 60 (BUP/SAM 2/2) (Placebo) Stage 1 Stage 2 HAM-D17 6 (BUP/SAM 2/2) (Placebo) CGI-S 50 MADRS-10 5 40 4 30

3 20

10 2

0 Baseline 1 2 3 4 Baseline 1 2 3 4 1 Baseline 1 2 3 4 Baseline 1 2 3 4

Started taper Started taper Started taper Started taper Week Week

E-166 Consistent Efficacy Results Regardless of Number of Weeks Included in Average – 207 MADRS‐10

Note: EOT was Week 5 in Stage 1 BUP/SAM 2/2 vs. Placebo and Week 6 in Stage 2 Least Squares Mean Difference (95% CI) E – 12 Study 207: Primary Efficacy Results Stage 1 Stage 2

Placebo MADRS-6 +ADT Change from Baseline BUP/SAM 2/2 by Treatment +ADT

Week Week Placebo, n 273 273 270 265 263 257 60 60 60 59 58 57 56 BUP/SAM, n 63 63 59 53 49 49 63 63 61 59 58 58 58

Endpoint P-value

BUP/SAM 2/2 Least Squares Mean Difference vs. Placebo (95% CI) BUP/SAM 1/1

46 Switching Antidepressant Treatments Is Increasingly Futile in Addressing Symptoms Proportion of Patients Remaining Symptomatic By Line of Therapy

100 91.7 87.1 84.0 85.5 80 75.0 74.5 73.4 63.2 60

Rate, % Rate, 40

20

0 citalopram venlafaxine-XR bupropion-SR sertraline nortriptyline mirtazapine venlafaxine-XR tranylcypromine mirtazapine Step 1 Step 2 Step 3 Step 4

Rush AJ et al. Am J Psychiatry 2006;163:1905-1917 16 BUP/SAM vs. Placebo Difference Consistent Improvement Across Studies Common Endpoint and Analysis ‐ MADRS‐10EOT

54 Study 205: Primary Efficacy Results – MADRS‐10Week 5 Stage 1 Stage 2

Placebo MADRS-10 +ADT Change from Baseline BUP/SAM 2/2 by Treatment +ADT

Week Week

Placebo, n 256 256 255 253 250 247 54 54 53 53 53 53 51

BUP/SAM, n 59 59 57 54 53 52 54 54 53 52 50 49 49

P-value

BUP/SAM 2/2 0.109 Least Squares Mean Difference vs. Placebo (95% CI) BUP/SAM 0.5/0.5 0.975

-5 -4 -3 -2 -1 0 1 2 3 Favors Favors BUP/SAM Placebo +ADT +ADT 43 Consistent Efficacy Equal Number of Weeks Included in Average – 207 MADRS‐10

BUP/SAM 2/2 vs. Placebo a Includes Weeks 3-5 in Stage 1 Least Squares Mean Difference (95% CI) and Weeks 3-6 in Stage 2 E –107 Summary of Documents Confirming Eligibility of Subject in 202 (Provided to FDA) Record Type Content Includes (but not limited to) Signed and Dated Clinical Note Depression history, Zoloft switch to Prozac in August 2012, Signed and dated by MD 10/24/12 remaining signs and symptoms of depression Psychiatric Evaluation Dates of depression diagnoses and episodes, Prozac 20 mg Signed and dated by coordinator of Depression 8/13/12 to current 10/24/12 Zoloft 100 mg June 2011-June 2012 MGH ATRQ Site Supports treating of current episode Signed by MD version Checks present for at least 10 wks for Prozac and Zoloft, No ECT Document dated 10/24/12 therapy, % improved <25% SAFER Interview form Fluoxetine x 1 week and Zoloft at least 8 wks (see PDF page 14) Signed and dated by CTNI rater with 3rd party ATRQ 10/29/12 performed by CTNI

Addendum dated 10/31/12 Addendum note (see PDF page 13) indicating Fluoxetine x 10 wks, made by Chang CTNI rater following discussion and confirmation with the site Concomitant Fluoxetine listed (among others), 20 mg, start 8/13/12 with No overall signature and date for Medication Log “ongoing” checked the log. However some corrections were noted with AA initials and dates for 2012 and 2013. E – 201 Study 202 HAMD‐17 Change from Baseline By Patient 10 10 5 Stage 15 Stage 2 0 0 -5 -5 -10 -10 -15 -15 -20 -20 -25 -25 -30 -30 -35 -35 BUP/SAM BUP/SAM 10 10 5 5 0 0 -5 -5 -10 -10 -15 -15 MADRS-10 Score Change from Baseline Change from MADRS-10 Score MADRS-10 Score Change from Baseline Change from Score MADRS-10 -20 -20 -25 -25 -30 -30 -35 -35 Placebo Placebo E –177 BUP/SAM Offers Efficacy Comparable to a Recently Approved Adjunctive Treatment for MDD

BUP/SAM 2/2 (4 studies) Brexpiprazole (4 studies)

-0 .2 0 .0 0 .2 0 .4 0 .6 Standardized Effect Size (Hedges' g) (95% CI)

Favors Treatment And does so through a different mechanism of action Yoon S et al. J Clin Psychopharmacol 2017;37:46-53 88 Study 202 HAMD-17 Percent Change from Baseline By Patient 100 Stage 1100 Stage 2 80 80 60 60 40 40 20 20 0 0 -20 -20 -40 -40 -60 -60 -80 -80 -100 -100

BUP/SAMALK5461202-129-002 ALK5461202-131-006 ALK5461202-112-004 ALK5461202-112-008 ALK5461202-118-003 ALK5461202-116-005 ALK5461202-111-011 ALK5461202-106-019 ALK5461202-132-003 ALK5461202-102-007 ALK5461202-117-001 BUP/SAMALK5461202-106-012 ALK5461202-106-004 ALK5461202-112-007 ALK5461202-111-008 ALK5461202-129-003 ALK5461202-127-004 ALK5461202-111-010 ALK5461202-106-007 ALK5461202-104-009 ALK5461202-120-008 ALK5461202-118-005 ALK5461202-106-003 100 100 80 80 60 60 40 40 20 20 0 0 -20 -20 -40 -40 MADRS-10 Percent Change Baseline Score from MADRS-10 MADRS-10 Baseline Change Score Percent from -60 -60 -80 -80 -100 -100 Placebo Placebo E –178 Discontinuation Rates with BUP/SAM Similar to Approved Adjunctive ADTs

Discontinuation Rate

Short-Term Studies Long-Term Studies (6-11 weeks) (52 weeks)

BUP/SAM 8-14% 51%

Brexpiprazole 6-12%a 51%b

a Thase et al, J Clin Psychiatry. 2015 Sep;76(9):1232-40; Thase et al, J Clin Psychiatry. 2015 Sep;76(9):1224-31; Hobart et al, Curr Med Res Opin. 2018 Apr;34(4):633-642 b Nelson et al, Curr Psychiatry Reviews. 2016 Aug; 12: 278-90 E –110 BUP/SAM Treatment Effect (MADRS-10EOT) is Comparable to Approved Adjunctive Treatments for MDD

Dose Study 1 mg 2 Brexpiprazole 2 mg 1* 3 mg 2

Aripiprazole 5-20 mg 1 5-20 mg 2

150 mg 1 Quetiapine 150 mg 2 200 mg 2 300 mg 1 -9 -8 -7 -6 -5 -4 -3 -2 -1 0 1 * Based on post-hoc analyses Favors Favors Treatment Placebo Rexulti USPI, 2018 +ADT +ADT Abilify USPI, 2016 Seroquel XR USPI, 2017 Least Squares Mean Difference (95% CI) E –167