Figure 1A. BAP1 inacvaon, as determined by immunohistochemistry Figure 1B. BAP1 mutaons detected in MM paents The Cancer Genome Atlas DEL, deleon; INS, inseron; SNV, single nucleode variant. Figure 2. BAP1 inacvaon sensizes to EZH2 inhibion in preclinical in vivo models a a Reprinted by permission from Springer Nature: Springer Nature Medicine “Loss of BAP1 funcon leads to EZH2-dependent Transformaon” by Lindsay M LaFave, Wendy Béguelin, Richard Koche, Ma Teater, Barbara Spitzer et al., Copyright 2015. Figure 4. EZH-203 study design a All paents will have completed at least the week 12 assessment, completed the final study visit, or terminated early from the study, whichever is sooner; b Disease Control: RECIST 1.1 or modified RECIST –confirmed objecve response or stable disease at week 12. RECIST, Response Evaluaon Criteria In Solid Tumors; R/R, relapsed or refractory. Figure 5. Durability of disease control in paents with relapsed or refractory MM Number in bars indicates lines of prior an-cancer therapy excluding maintenance as its own line. Data as of Jan 16, 2018. *Dosing records incomplete. Figure 6. 64 Year Old Male with Pleural MM Shows Slow Steady Reducon in Tumor Burden with 1 st PR at Week 42 a Baseline as of February, 2017; b Last scan as of January 16, 2018. MM, malignant mesothelioma; PR, paral response; RECIST, Response Evaluaon Criteria In Solid Tumors. Paent Vignee Figure 3. EZH2-mediated repression of transcripon Phase 2 Mulcenter Study of the EZH2 Inhibitor Tazemetostat as Monotherapy in Adults With Relapsed or Refractory Malignant Mesothelioma With BAP1 Inacvaon (NCT02860286) Marjorie G. Zauderer 1 , Peter Szlosarek 2 , Sylvestre Le Moulec 3 , Sanjay Popat 4 , Paul Taylor 5 , David Planchard 6 , Arnaud Scherpereel 7 , Thierry Jahan 8 , Marianna Koczywas 9 , Marn Forster 10 , Robert B. Cameron 11 , Tobias Peikert 12 , Carly Campbell 13 , Inbal Sapir 13 , Alice McDonald 13 , Coreen Oei 13 , Alicia Clawson 13 , Maria Roche 13 , Dean A. Fennell 14 1 Memorial Sloan Keering Cancer Center, New York, NY; 2 St. Bartholomew’s Hospital, London, UK, 3 Instut Bergonie, Bordeaux, France; 4 Royal Marsden Hospital, London, UK; 5 Wythenshawe Hospital, Manchester University NHS Foundaon Trust, Manchester, UK; 6 Instut Gustave Roussy, Villejuif, France; 7 Hospital of the University (CHU) de Lille, Lille, France; 8 University of California, San Francisco, CA; 9 City of Hope Naonal Medical Center, Duarte, CA; 10 University College Hospital, London, UK; 11 University of California, Los Angeles, CA; 12 Mayo Clinic, Rochester, MN; 13 Epizyme, Inc., Cambridge, MA; 14 University of Leicester and University Hospitals of Leicester, Leicester, UK Presented at the 2018 Annual Meeng of the American Society of Clinical Oncology (ASCO); June 1–5, 2018; Chicago, IL, USA 121 BACKGROUND Mesothelioma • Malignant mesothelioma (MM) is caused by occupaonal and environmental exposure to asbestos fibers and other elongate mineral parcles 1 – The latency period from exposure to MM ranges from 20 to 40 years • Due to the locaon and characteriscs of the malignancy, diagnosis is oſten at advanced stage of MM 1,2 – Once diagnosed, paents have a median survival of approximately 1 year, experiencing rapid and aggressive progression with poor prognosis • Available treatment for these paents is limited and in dire need of improvement both in efficacy and safety. All paents relapse on average 6 months following first-line treatment with chemotherapy 2 • No drugs have been shown to improve survival as second-line therapy, and to date there is no standard of care or approved therapy in second-line 1,2 • There is an urgent need for safe and effecve targeted therapies for MM BAP1 • BRCA1-associated protein 1 (BAP1), a nuclear deubiquinase, is commonly inacvated in MM tumors 3,4 (Figure 1A-B) – Up to 66% of MM cases exhibit BAP1 inacvaon, as measured by a loss of nuclear protein 5 – Approximately 23% of MM cases harbor BAP1 genec alteraons 3,4 – Inacvaon is most common in epithelioid & biphasic subtypes compared with sarcomatoid 5 EZH2 • BAP1 inacvaon is linked to a dependency on enhancer of zeste-homolog 2 (EZH2) acvity 6 (Figure 2) • EZH2 acvity is implicated as an oncogenic driver and preclinical data suggest that elevated EZH2 acvity may play a role in MM in the context of BAP1 loss 6 Table 3. Paent demographics and baseline disease characteriscs Clinical Characteriscs Total Number of Paents N=74 Age (years), mean (SD) 65 (10) Male/Female, n (%) 49 (66)/25 (34) ECOG status PS 0/1, n (%) 19 (26)/55 (74) Prior lines of therapy, median (range) 2 (1–9) Therapy seng, n (%) Neoadjuvant 7 (9) Adjuvant 18 (24) Therapeuc for advanced/metastac disease/Palliaon 61 (83) Maintenance 5 (7) Unknown 6 (8) Time (months) from last PD to study entry, median (range) 1 (0–14) Primary tumor locaon; pleura/peritoneum, n (%) 68 (92)/6 (8) Histology, n (%) Epithelioid 65 (88) Sarcomatoid 1 (1) Biphasic 6 (8) Unknown 2 (3) Stage at diagnosis, n (%) I 3 (4) II 14 (19) III 13 (18) IV 28 (38) Unknown 16 (22) ECOG, Eastern Cooperave Oncology Group; PD, progressive disease; PS, performance status. Table 4. TEAEs occurring in ≥10% of paents, regardless of relaonship to study drug Paents (N=74) Adverse Events All Grades ≥ 10% n (%) Grade ≥3 n (%) Anemia 12 (16) 4 (5) Dyspnea 21 (28) 3 (4) Asthenia 9 (12) 2 (3) Cancer pain 19 (26) 2 (3) Fague 24 (32) 2 (3) Nausea 20 (27) 2 (3) Weight decreased 8 (11) 2 (3) Cough 16 (22) 1 (1) Decreased appete 21 (28) 1 (1) Voming 16 (22) 1 (1) Diarrhea 11 (15) 0 Lower respiratory tract infecon 8 (11) 0 Table 5. Best response to tazemetostat treatment Endpoint Category, n (%) Part 2 a N=61 Total N=74 Primary DCR (CR + PR + SD) at week 12 95% CI b 31 (51) 38-64% 35 (47) 36-59% Secondary DCR (CR + PR + SD) at week 24 16 (26) 17 (23) Best Response c Complete Response (CR) 0 0 Paral Response (PR) d 2 (3) 2 (3) Stable Disease (SD) 38 (62) 46 (62) Progressive Disease (PD) 16 (26) 21 (28) Non-evaluable 5 (8) 5 (7) a Success at stage 2 was based on results from Part 2; b Exact binomial; c Based on modified RECIST for thoracic disease and RECIST 1.1 elsewhere; d Confirmed responses. DCR, disease control rate. Table 1. Paent key inclusion criteria Inclusion Age ≥18 years ECOG performance status of 0 or 1 Life expectancy >3 months MM of any histology that is relapsed or refractory aſter treatment with at least 1 pemetrexed-containing regimen Documented and confirmed (CLIA/CAP) local diagnosc pathology of original biopsy Evidence of BAP1 protein loss by local pathology (for part 2 of the study) Measureable disease CAP, College of American Pathologists; CLIA, cerfied through Clinical Laboratory Improvement Amendments; ECOG, Eastern Cooperave Oncology Group. Table 2. Paent key exclusion criteria Exclusion Prior exposure to tazemetostat or other inhibitor(s) of EZH2 History of known central nervous system metastasis Prior malignancy other than the malignancies under study Excepon: A paent who has been disease-free for 5 years, or a paent with a history of a completely resected non-melanoma skin cancer or successfully treated in situ carcinoma is eligible Major surgery within 3 weeks prior to enrollment Cardiovascular impairment, history of congesve heart failure greater than New York Heart Associaon Class II, uncontrolled arterial hypertension, unstable angina, myocardial infarcon, or stroke within 6 months prior to the planned first dose of tazemetostat; or ventricular cardiac arrhythmia requiring medical treatment Tazemetostat • EZH2 is an epigenec regulator of gene expression and plays a crical role in mulple forms of cancer (Figure 3) • Tazemetostat is a potent, selecve, oral EZH2 inhibitor that is generally well tolerated and has demonstrated preclinical and clinical acvity in several tumor types • Here, we report preliminary data from the phase 2 trial examining tazemetostat in paents with measurable relapsed or refractory MM METHODS • Study EZH-203 (NCT02860286) is a phase 2, mulcenter, open-label, 2-part study of tazemetostat 800 mg administered orally twice daily for relapsed or refractory MM (Figure 4) – Response assessment was evaluated aſter 6 weeks of treatment and every 6 weeks thereaſter while on study • Key inclusion and exclusion criteria are presented in Tables 1 and 2 • Primary endpoint part 2: Disease control rate (DCR): CR+PR+SD at week 12 • 2-Stage Green Dahlberg Design; H 0 : DCR ≤25%, H a : DCR ≥35% RESULTS (JANUARy 16, 2018 DATA-CUT) Paent Characteriscs • Enrollment of 74 paents to EZH-203 was completed in 10 months (Table 3) • A total of 70 of 74 tumor samples (95%) centrally confirmed BAP1-deficient CONCLUSIONS AND NEXT STEPS • Tazemetostat was generally well tolerated in paents with relapsed refractory MM • Tazemetostat monotherapy showed promising an-tumor acvity in paents with tumors harboring BAP1 inacvaon, including confirmed responses and long-term disease control • Studies to fully characterize paent tumor samples are ongoing – Exploratory prognosc indices – RNA-seq – Immune profiling by mulplex immunofluorescence • Further clinical evaluaon of tazemetostat in combinaon studies is warranted REFERENCES 1. Mazurek JM, Syamlal G, Wood JM et al. MMWR Morb Mortal Wkly Rep. 2017;66:214–8. 2. Baas P, Fennell D, Kerr KM et al. Ann Oncol. 2015; 26:v31–9. 3. Bo M, Brevet M, Taylor BS et al. Nat Genet. 2011;43:668–72. 4. Bueno R, Stawiski EW, Goldstein LD et al. Nat Genet. 2016;48:407–16. 5. Cigogne M, Lonardi S, Fisogni S et al. Mod Pathol. 2015;28:1043–57. 6. LaFave LM, Béguelin W, Koche R et al. Nat Med. 2015;21:1344–9. ACKNOWLEDGMENTS We thank all of the physicians, study coordinators, site staff and most of all, the paents, caregivers, and families who have contributed to this study Third-party wring assistance was provided by Ashfield Healthcare and funded by Epizyme, Inc. Copies of this poster obtained through QR (Quick Response) code are for personal use only and may not be reproduced without wrien permission from the author of this poster. Safety • A total of 71/74 (96%) paents had a treatment-emergent adverse event* (TEAE) – The most frequent TEAEs included fague (32%), decreased appete (28%), and dyspnea (28%), and were generally grade 1 or 2 in severity (Table 4) • No paents disconnued the study due to treatment-related TEAEs • Five (7%) paents had a dose reducon due to TEAEs – One paent experienced a dose reducon on 2 separate occasions • Ten (14%) paents died in this study; no deaths were related to tazemetostat as assessed by invesgator *Any AE that occurred aſter paent iniated treatment Paent Disposion • Overall, 67 of 74 paents disconnued study – Disease progression: 61 – Death caused by disease under study: 5 – Refused further treatment: 1 • Seven paents are ongoing – Paral response (PR): 2 – Stable disease (SD): 3 – Progressive disease (PD): 2 • Connue on study per protocol Efficacy Objecve Response • Confirmed PRs were observed in 2 paents – Both paents with a PR are ongoing treatment at 42 and 48 weeks – An example of clinical acvity is shown in Figure 6A – Inial responses observed at weeks 36 and 42 (Figure 6B), consistent with emergence of late responses noted across the tazemetostat clinical development program Weeks since treatment initiation 0 4 8 12 16 20 24 28 32 36 40 44 48 52 56 Patients Part 1 Part 2 Progressive disease Partial response – confirmed Ongoing * * * * 6 5 2 3 2 2 1 3 1 3 2 1 2 1 3 2 1 1 1 4 1 3 4 1 1 3 1 2 5 3 3 1 1 3 2 2 5 2 1 1 1 2 1 1 2 1 3 2 1 1 1 3 2 2 3 2 1 1 1 3 1 1 1 1 2 1 1 2 2 1 3 7 3 24 weeks 48 weeks b Baseline a 2016 Tazemetostat: week 48+ Pemetrexed cisplan 2015 Mar July May Aug Feb 2017 PR SD 2018 PR at week 42 Diagnosed ES stage 3 Vinorelbine PD 0% 1.3% 3.9% -9.2% -12.7% -14.9% -16.2% -33.3% -34.6% -35% -30% -25% -20% -15% -10% -5% 0% 5% Baseline Week 6 Week 12 Week 18 Week 24 Week 30 Week 36 Week 42 Week 48 Change from baseline in sum of lesions Time on tazemetostat treatment PR criteria met per RECIST 1.1 Paents with R/R MM STAGE 1 ANALYIS Targeted enrollment = 30 a STAGE 2 ANALYIS Targeted enrollment = 55 <5 paents with Disease Control b No further enrollment in cohort b Pharmacokinec profile of tazemetostat and its metabolite Response assessment evaluated aſter 6 weeks of treatment and then every 6 weeks thereaſter Enrollment connues to n=55 ≥5 paents with Disease Control b PART 2 BAP1 status: tumor loss of nuclear posivity PART 1 BAP1 status: undefined Targeted enrollment = 12 ≥17 paents with Disease Control b <17 paents with Disease Control b BAP1 nuclear & cytoplasmic negave BAP1 nuclear negave, cytoplasmic posive BAP1 nuclear posive, cytoplasmic posive Nuclear Loss Nuclear Posi ve UCH Frequency of mutaon HBM NLS BARD1 HCF-1 DEL SNV INS BRCA1 1 250 363 656 717 729 p.W5G p.I47Lfs*21 p.Q280* p.S10R p.Q85Sfs*2 p.I210DEL p.X218_splice p.C91G p.X220_splice p.IA644Efs*11 p.IN645Kfs*3 p.V654Qfs*3 p.K651del p.E284* p.P293Gfs*35 p.X218x p.H169Tfs*18 p.N251Tfs*6 p.Q392* p.Q456* p.Y418Wfs*9 p.E182K p.I72Cfs*7 BAP1– BAP1 + 5 Vehicle 1 4 7 11 14 ** ** Time of EPZ011989 treatment (d) 17 21 24 H226 500 mg/kg EPZ011989 4 3 Tumorfold change (Normalized to day 1) 2 1 0 5 Vehicle 1 4 7 11 ** Time of EPZ011989 treatment (d) H2452 500 mg/kg EPZ011989 Tumorfold change (Normalized to day 1) 1 0 15 Vehicle 1 4 7 11 ns Time of EPZ011989 treatment (d) 14 Meso10 500 mg/kg EPZ011989 10 Tumorfold change (Normalized to day 1) 5 0 Vehicle Time of EPZ011989 treatment (d) MSTO-211H 500 mg/kg EPZ011989 Tumorfold change (Normalized to day 1) 5 4 3 2 1 0 1 4 7 11 14 17 21 24 PRC2 EZH2 K27me3 K27me3 K27me3 Compacted chromatin Transcriptional repression K27me3 A B On study scans Tazemetostat treatment meline