1. Immunotherapy is the Preferred Initial Treatment for Most Patients with Metastatic BRAF V600 Mutant Melanoma Michael B. Atkins, M.D. Deputy Director Georgetown-Lombardi Comprehensive Cancer Center

2. The case for immunotherapy---  Immunotherapy produces durable unmaintained tumor responses  Immunotherapy works as well against BRAFV600 mutant melanoma as WT tumors  Immunotherapy is getting better  Better drugs  Better patient selection  BRAF inhibitors work as well in patients with prior immune therapy; the converse is not so

3. High Dose IL-2 Therapy*  RR: 16% (43 / 270)  Some large volume and visceral  Most soft tissue and lung  Durable responses  Median 8.9 mos  CR: not reached  Survival  Median 12 mos  11% >@ 5yrs *Atkins et al JCO, 1999 (N=270)

4. Pooled data from phase II studies CA184-008 and CA184-022: ipilimumab monotherapy 10 mg/kg (N=227) mWHO CR/PR/SD = 27.8% irRC CR/PR/SD with WHO PD = 9.7% Other PD or unknown = 62.6% 1.0 0.8 0.6 0.4 0.2 0 Proportion alive 02468101214161820222426 Wolchok, Hodi et al

5. Overall survival for 51-patient cohort Rx’d w/ 10 mg/kg x 4 + maintenance ~26% 2 yr OS estimate also seen in 165-pt cohort from expanded access trial Kevin Heller, BMS—data presented in part at ESMO 2009, ASCO 2010, 2011 posters

6. Relationship of MAPKinase pathway mutations and response to HD IL-2 MutationAllCR/PRSD/PDP-value BRAF6014(23%)46 (77%) 0.05 NRAS157 (47%)8 (53%) WT263 (12%)23 (88%) Joseph, Sullivan et al- JIT 2011 A significantly larger proportion of patients with BRAF or NRAS mutant tumors achieved CR/PR compared to those with WT tumors.

7.  Produces durable unmaintained responses; not restricted by BRAF status  Activity powerful enough to work in the CNS and overcome concurrent immunosuppression  Activity seen in patients with prior IL-2  Activity seen in patients with elevated LDH, liver metastases, etc  Treatment more widely applicable Ipilimumab Therapy

8. BRAF inhibitor Therapy - Limitations  Median PFS of only 6-7 months  Median OS 13-16 months  Forrest plot suggests most of the benefit confined to patients with M1c disease Sosman et al NEJM 2012, Chapman et al ASCO 2012

9. 100 90 80 70 60 50 40 30 20 10 0 Overall survival (%) 06121824 Vemurafenib (n=337) Median f/u 12.5 months Dacarbazine (n=338) Median f/u 9.5 months 338 337 173 280 79 178 24 44 0101 244 326 111 231 50 109 4747 9.713.6 Overall survival (February 01, 2012 cut-off) censored at crossover Hazard ratio 0.70 (95% CI: 0.57–0.87) p<0.001 (post-hoc) Time (months) No. at risk Dacarbazine Vemurafenib 15.9 BRIM2

10. All patients675 Age:<65 years ≥65 years 514 161 Sex:Female Male 294 381 ECOG status: 0 1 459 216 Disease stage: IIIc M1a M1b M1c 33 74 127 441 LDH:Normal Elevated 391 284 Factor Number of patients 0.20.40.61.024106 Favors vemurafenib Favors dacarbazine 20 Hazard ratio and 95% confidence interval Overall survival by baseline characteristic (February 01, 2012 cut-off) censored at crossover

11. BRAF inhibitor Therapy - Limitations  Median PFS of only 6-7 months  Median OS 13-16 months  Forest plot suggests most of the benefit confined to patients with M1c disease  Combination BRAF-MEK inhibition may offer some advantages… Sosman et al ASCO 2012, Chapman et al ASCO 2012

12. BRAF inhibitor Therapy - Limitations  Median PFS of only 6-7 months  Median OS 13-16 months  Forest plot suggests most of the benefit confined to patients with M1c disease  Combination BRAF-MEK inhibition may offer some advantages, but median PFS still only 7- 10 months Sosman et al ASCO 2012, Chapman et al ASCO 2012, Weber et al, ASCO 2012

13. BRAF inhibitor Therapy - Limitations Thus, BRAF (+/- MEK) inhibitor treatment postpones but does not prevent the tragedy of metastatic melanoma Sosman et al ASCO 2012, Chapman et al ASCO 2012, Weber et al, ASCO 2012

14. Treatment for BRAF Mutant Melanoma Years after stage IV diagnosis Proportion Surviving Ipilimumab 01234 5 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 BRAF i PRESENTED BY: Michael B. Atkins Which is preferred?

15. Immunotherapy is getting better

16. Clinical Activity of BMS-936558 in Melanoma Patients  4 melanoma patients had a persistent reduction in baseline target lesions in the presence of new lesions but were not classified as responders for the ORR calculation PopN ORR n (%) Median Duration of Response (95% CI) [individual pt response] SD  24 wk n (%) PFSR at 24 wk (%) All MEL 106 33 (31) Not Reached Range: 1.8+ to 25.7 6 (6)42 MEL 176 (35) Not Reached [3.7+, 4.2+, 5.6, 5.6, 5.6+, 11.2+] 041 185 (28) Not Reached [1.8+, 4.2, 7.4+, 7.6+, 9.2+] 1 (6)33 3411 (32) 24 months (22.9- NR) [1.9+, 5.5+, 7.5, 7.5, 11.1+, 13.4+, 18.4+, 22.9, 23.2+, 24, 24.9+] 4 (12)48 177 (41) Not Reached [9.2+, 9.3, 11.1, 12.9, 18.8+, 22+, 22.4+] 1 (6)55 204 (20) 25.7 months (17.0-25.7) [17, 18+, 24.6+, 25.7] 030 3mg/kg

17. Changes in Target Lesions Over Time in Melanoma Patients (3mg/kg)  Of 33 patients with OR (all dose levels)  29 were treated  1 year (before July 3, 2012) and 14 had responses of  1 year  4 were treated <1 year and 4 had responses ranging from 1.8-5.3 months

18. MK-3475 (Unconfirmed + Confirmed Responses) in Advanced MEL Patients Complete Response (N, 95% CI) Objective Response (N, 95% CI) Disease Control Rate (N, 95% CI) All MEL N=83 5% (4; 2%-13%) 47% (39; 34% - 56%) 60% (50; 48%- 70%) IPI Naïve N=58 7% (4; 2%-18%) 50% (29; 35% -61 %) 67% (39; 51%-76%) IPI Treated N=25 0 % 40% (10; 17% -59%) 44% (11; 24%-68%) All patients were dosed at 10 mg/kg Includes all patients who received first dose as of April 25, 2012. Centrally available response information as of Oct 19, 2012 Objective response= confirmed and unconfirmed complete and partial response Disease control rate= objective response + stable disease Hamid SMR 11/11/12

19. Clinical Activity in a Melanoma Patient-1 Baseline: 13/Apr/2012 27/July/2012 72 yrs old male with melanoma after progressing on bio-chemotherapy, HD IL-2, and ipilimumab Patient was on oxygen support due to progressive lung disease burden and pleural effusion After 3 months of MK-3475, the patient is off the oxygen support and continues to respond Courtesy of A. Ribas M.D.

20. Baseline: 13/Apr/201227/July/2012 Clinical Activity in a Melanoma Patient-1

21. MK-3475 Characteristics of Responses (irRC): Time to Respond & Duration for 43 Patients with Objective Response On Treatment Duration in Weeks Median duration of treatment, 7.6 months + (3.3-11+) one patients discontinued due to PD and four patients discontinued due to AEs

22. Immunotherapy for Melanoma PRESENTED BY: Michael B. Atkins Years after stage IV diagnosis Proportion Surviving Ipilimumab Anti-PD1 Anti PD1 + x 01234 5 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 ??? IL-2

23. Checkpoint blockade vs oncogene- mutation targeted therapy for melanoma # Annual Cases Proportion c\ mutation % RRMedian Durability of remission Pt years BRAF + cKIT 90000.5 0.6 yrs1350 pt yrs Anti- CTLA4/Anti -PD1 9000N/A0.332 years6000 pt yrs Annual pt-years of remission Adapted from Drew Pardoll

24. Treatment Selection Opportunities

25. NSCLC Melanoma RCC * *2 pts still under evaluation 42 pts include 18 MEL, 10 NSCLC, 7 CRC, 5 RCC, and 2 CRPC. Correlation of PD-L1 expression in pretreatment tumor biopsies with clinical outcomes in Anti-PD-1 Therapy Association Between Pretreatment Tumor PD-L1 Expression and Clinical Response Response StatusPD-L1 Positive no. (%) PD-L1 Negative no. (%) Total no. (%) CR/PR9 (36)09 (21) Nonresponder16* (64)17 (100)33 (79) All Patients251742 Topalian S, et al. NEJM 2012;366:2443-2454. Proportion of patients p=0.006 9/25 16*/25 17/17 0/17 † analysis not pre-planned and based on subset of subjects'. †

26. Sequencing of Treatment

27. Baseline characteristics Overall response rate (%) 30 40 50 70 60 80 BRIM2- ORR by pre-defined subgroups 20 0 Age <65  65 All treated patients Sex FM LDH at enrollment 1.0–1.5x ULN >1.5x ULN Normal ECOG PS 01 Stage M1a/ M1b M1c # prior therapies 1>1 Previous IL-2 YesNo 10 Overall ORR of 53% (IRC) RR (size proportional to the number of patients in the subgroup) 95% Confidence intervals Ribas et al ASCO 2011

28. BRAF inhibition +/- prior immunotherapy MAPKi PFS IT initially PFS 6.7 mo (CI 4.3-9.1 mo) MAPKi initially PFS 5.6 mo (CI 4.7-6.8 mo) p-value 0.43, log rank MAPKi OS IT initially OS 19.6 mo (CI 10.0- mo) MAPKi initially OS 13.4 mo (CI 10.1-17 mo) p-value 0.40, log rank OS (probability) Time (mo) Ackerman/Sullivan-BIDMC/MGH-SITC 2012

29. Immunotherapy following MAPKI: DFHCC/MIA Retrospective Data –193 patients discontinued MAPKi therapy (176 with disease progression) Median OS 2.9 mos (CI 1.8-4.4 mos) from last dose of MAPKi Single agent ipilimumab treatment (n=34 pts) No tumor responses (2 SD) Median PFS 2.7 mos, median OS 5 mos 50% of patients received < 4 doses All Pts alive > 1 year- are back on MAPK inhibitors inhibitors –Summary: Patients progressing on BRAFi appear unlikely to respond to ipilimumab Those alive either had slow growing disease and short period of RAFi treatment due to toxicity or are back on a RAFi Ackerman/Sullivan-BIDMC/MGH-SITC 2012

30. OS 5.0 mo (CI 3.0-8.8 mo) long term survivors all treated with additional MAPKi OS (probability) Time (mo) PFS 2.7 mo (CI 1.8-3.1 mo) Ipilimumab following BRAF inhibitor Therapy Ackerman/Sullivan-BIDMC/MGH-SITC 2012

31. Treatment Selection in Patients with BRAF Mutant Melanoma: Conclusions  Current data suggests that for many patients with BRAFV 600E melanoma (asymptomatic, immune infiltrate), starting with immunotherapy offers them a chance for longterm benefit without compromising their benefit from subsequent BRAFi therapy

32. Two shots on goal are better than one

33. E1612: Ipi vs Vemurafenib ECOG PS 1.0 2.1 Stage 1.St III or M1a/b 2.M1c Prior therapy 1.No prior Rx 2.Prior Rx RANDOMIZERANDOMIZE Arm 1: Ipi 3 or 10mg/kg q 3wks x 4 +/- maint q12 wks Arm 2: Vemurafenib 960mg BID ECOG and SWOG protocol – Atkins, Chmielowski Tumor measurements q12 wks Ipi 3 or 10mg/kg q 3wks x 4 +/- maint q12 wks Vemurafenib 960mg BID PD

34. Treatment Selection in Patients with BRAF Mutant Melanoma: Conclusions  Current data suggests that for many patients with BRAFV 600E melanoma (asymptomatic, immune infiltrate), starting with immunotherapy offers them a chance for longterm benefit without compromising their benefit from subsequent BRAFi therapy  Newer immunotherapies are approaching the efficacy (RR and PFS) of BRAFi with more durability

35. Immunotherapy may be the better of the two shots