Updates From the European Lung Cancer Conference: Immunotherapy and Non-Small Cell Lung Cancer

Transcription

1 Updates From the European Lung Cancer Conference: Immunotherapy and Non-Small Cell Lung Cancer Benjamin Besse, MD, PhD Chair, EORTC Lung Group Chair, Gustave Roussy Thoracic Unit Villejuif, France

2 What Is This? TRADITIONAL ONCOLOGY VIEW IMMUNO- ONCOLOGY VIEW A CANCER THAT GROWS A BODY THAT LET A CANCER GROW

3 Courtesy of S. Champiat Immune Checkpoint : Mechanism of Action

4 Know Your Immune Checkpoint Antibodies Anti-CTLA-4 Anti-PD-1 Anti-PD-L1 Ipilimumab (BMS) Tremelimumab (AZ) Approved Nivolumab (BMS) Pembrolizumab (MSD) Approved Atezolizumab (MPDL3280A) Durvalumab (MEDI4736) Avelumab (PF /MSB C) Approved Courtesy of S Champiat

5 Hirsch FR, et al. Presented at: 2016 AACR Annual Meeting; April 16-20, 2016: New Orleans, Louisiana. Abstract RSP05. Blueprint PD-L1 IHC Pembrolizumab Nivolumab Atezolizumab

6 PD-1/PD-L1 Antibodies in NSCLC Second Line and + Nivolumab data

7 CheckMate 017 (NCT ) - Study Design Stage IIIb/IV Squamous NSCLC Stage IIIb/IV SQ NSCLC 1 prior platinum doubletbased chemotherapy ECOG PS 0 1 Pretreatment (archival or fresh) tumor samples required for PD-L1 analysis N = 272 Randomize 1:1 Nivolumab 3 mg/kg IV q2w until PD or unacceptable toxicity n = 135 Docetaxel 75 mg/m 2 IV q3w until PD or unacceptable toxicity n = 137 Primary Endpoint: OS Additional Endpoints: Investigator-assessed ORR Investigator-assessed PFS Correlation between PD-L1 expression and efficacy Safety Quality of life (LCSS) Patients stratified by region and prior paclitaxel use One preplanned interim analysis for OS At time of DBL (December 15, 2014), 199 deaths were reported (86% of deaths required for final analysis) The boundary for declaring superiority for OS at the preplanned interim analysis was P<.03 LCSS, Lung cancer symptom scale Brahmer J, et al. N Engl J Med. 2015;373(2):

8 Updated Overall Survival mos, mos (95% CI) Nivolumab n = (7.33, 12.62) Docetaxel n = (5.29, 7.39) OS, % # events HR = 0.62 (0.48, 0.81); P =.0004 Nivolumab Docetaxel Number of Patients at Risk Time, Months Nivolumab Docetaxel Minimum follow-up for survival: 18 months Based on August 2015 DBL. Symbols refer to censored observations. Adapted from: Reckamp K, et al. J Thorac Oncol. 2015;10(9 Suppl 2): Abstract ORAL Brahmer J, et al. N Engl J Med. 2015;373(2):

9 Survival Benefit By PD-L1 Expression PD-L1 positive expression PD-L1 Expression OS Patients, n Nivolumab Docetaxel Unstratified HR (95% Cl) <1% (0.37, 0.92) 1% (0.45, 1.05) <5% (0.47, 1.02) 5% (0.31, 0.89) <10% (0.48, 1.01) 10% (0.28, 0.89) Not quantifiable (0.19, 0.82) PFS <1% (0.43, 1.00) 1% (0.44, 1.01) <5% (0.52, 1.08) 5% (0.32, 0.90) <10% (0.49, 0.99) 10% (0.33, 1.02) Not quantifiable (0.23, 0.89) Interaction P value PD-L1 negative expression Not quantifiable Nivolumab benefit was independent of PD-L1 expression 83% of patients (225/272) had quantifiable PD-L1 expression Based on December 2014 DBL Nivolumab Docetaxel Adapted from: Reckamp K, et al. J Thorac Oncol. 2015;10(9 Suppl 2): Abstract ORAL Brahmer J, et al. N Engl J Med. 2015;373(2):

10 CheckMate 057 (NCT ) Study Design Stage IIIB/IV Nonsquamous NSCLC Stage IIIB/IV non-sq NSCLC Pretreatment (archival or recent) tumor samples required for PD-L1 ECOG PS 0 1 Failed 1 prior platinum doublet Prior maintenance therapy allowed a Prior TKI therapy allowed for known ALK translocation or EGFR mutation N = 582 Randomize 1:1 Nivolumab 3 mg/kg IV q2w until PD or unacceptable toxicity n = 292 Docetaxel 75 mg/m 2 IV q3w until PD or unacceptable toxicity n = 290 Primary Endpoint OS Additional Endpoints ORR b PFS b Safety Efficacy by tumor PD- L1 expression Quality of life (LCSS) Patients stratified by prior maintenance therapy and line of therapy (second- vs third-line) PD-L1 expression measured using the Dako/BMS automated IHC assay 14,15 Fully validated with analytical performance having met all predetermined acceptance criteria for sensitivity, specificity, precision, and robustness a Maintenance therapy included pemetrexed, bevacizumab, or erlotinib (not considered a separate line of therapy); b Per RECIST v1.1 criteria as determined by the investigator. Par-Ares L, et al. J Clin Oncol. 2015;33(suppl): Abstract LBA 109. Borghaei H, et al. N Engl J Med. 2015;373(17):

11 Par-Ares L, et al. J Clin Oncol. 2015;33(suppl): Abstract LBA 109. Borghaei H, et al. N Engl J Med. 2015;373(17): OS, % Overall Survival 1-yr OS rate = 39% 1-yr OS rate = 51% Nivolumab n = 292 Docetaxel n = 290 mos, mo HR = 0.73 (96% CI: 0.59, 0.89); P =.0015 Nivolumab 20 Number of Patients at Risk Nivolumab Time, Months Docetaxel Docetaxel

12 OS by PD-L1 Expression (Nonsquamous) 10% PD-L1 Expression Level <10% PD-L1 Expression Level mos, mo Nivo 19.4 Doc mos, mo Nivo 9.9 Doc HR (95% CI) = 0.40 (0.26, 0.59) Time (months) 10 0 HR (95% CI) = 1.00 (0.76, 1.31) Time (months) (53% pts PD-L1 1%) Par-Ares L, et al. J Clin Oncol. 2015;33(suppl): Abstract LBA 109. Borghaei H, et al. N Engl J Med. 2015;373(17):

13 PD-1/PD-L1 Antibodies in NSCLC Second Line and + Nivolumab data Pembrolizumab data

14 Pembrolizumab in NSCLC OS = 12.5 mo OS PD-L1+ : not reached Garon EB, et al. N Engl J Med. 2015;372(21):

15 KEYNOTE-010 Study Design Herbst RS, et al. Ann Oncol. 2015;26(suppl_9): Abstract LBA3_PR. Herbst RS, et al. Lancet. 2016;387(10027):

16 OS: PD-L1 TPS 50% Stratum Analysis cutoff September 30, 2015 Herbst RS, et al. Ann Oncol. 2015;26(suppl_9): Abstract LBA3_PR. Herbst RS, et al. Lancet. 2016;387(10027):

17 PD-1/PD-L1 Antibodies in NSCLC Second Line and + Nivolumab data Pembrolizumab data Atezolizumab data

18 What was the median overall survival of the atezolizumab arm in the OAK trial? months months months months

19 Phase III OAK Study Design Locally Advanced or Metastatic NSCLC Stratification factors PD-L1 expression Histology Prior chemotherapy regimens Atezolizumab 1200 mg IV q3w PD or loss of clinical benefit 1 2 prior lines of chemo including at least 1 platinum based Any PD-L1 status N = 1225 enrolled a R 1:1 Docetaxel 75 mg/m 2 q3w PD Primary Endpoints (first 850 enrolled patients): OS in the ITT population OS in patients with PD-L1 expression on 1% TC or IC Secondary Endpoints: ORR, PFS, DoR, Safety a A prespecified analysis of the first 850 patients provided sufficient power to test the co-primary endpoints of OS in the ITT and TC1/2/3 or IC1/2/3 subgroup ( 1% PD-L1 expression). TC, tumor cells; IC, tumor-infiltrating immune cells Barlesi F, et al. Ann Oncol. 2016;27(Suppl 6): Abstract LBA44. Rittmeyer A, et al. Lancet. 2017;389(10066):

20 Overall Survival, ITT (N = 850) HR, 0.73 a (95% CI, 0.62, 0.87) P =.0003 Overall Survival, % Atezolizumab Docetaxel Minimum follow-up = 19 months Median 9.6 mo (95% CI, 8.6, 11.2) Median 13.8 mo (95% CI, 11.8, 15.7) Months a Stratified HR Barlesi F, et al. Ann Oncol. 2016;27(Suppl 6): Abstract LBA44. Rittmeyer A, et al. Lancet. 2017;389(10066):

21 Overall Survival by PD-L1 Expression On-Study Prevalence 16% 31% 55% 45% Subgroup 0.41 TC3 or IC3 TC2/3 or IC2/3 TC1/2/3 or IC1/2/3 a TC0 and IC Median OS, mo Atezolizumab Docetaxel n = 425 n = % ITT a Hazard Ratio a a Stratified HR for ITT and TC1/2/3 or IC1/2/3. Unstratified HR for subgroups. TC, tumor cells; IC, tumor-infiltrating immune cells; OS, overall survival In favor of atezolizumab In favor of docetaxel Barlesi F, et al. Ann Oncol. 2016;27(Suppl 6): Abstract LBA44. Rittmeyer A, et al. Lancet. 2017;389(10066):

22 Results From the OAK Subgroup Analyses: A Randomized Phase III Study of Atezolizumab vs Docetaxel in Patients With Advanced NSCLC Cortinovis D, Gadgeel SM, Rittmeyer A, Barlesi F, Cobo Dols M, Hida T, He P, Ballinger M, Gandara DR, von Pawel J Cortinovis D, et al. Presented at: European Lung Cancer Conference; 5-8 May 2017: Geneva, Switzerland. Abstract 89PD.

23 Cortinovis D, et al. Presented at: European Lung Cancer Conference; 5-8 May 2017: Geneva, Switzerland. Abstract 89PD. Overall Survival By Histology And PD-L1 Subgroup OS benefit in patients with nonsquamous or squamous histology, regardless of PD-L1 expression status, treated with atezolizumab vs docetaxel

24 Cortinovis D, et al. Presented at: European Lung Cancer Conference; 5-8 May 2017: Geneva, Switzerland. Abstract 89PD. OS and PFS by Presence of CNS Metastases and History of Tobacco Use CNS Metastases Tobacco Use Improved OS in patients with baseline CNS metastases who received atezolizumab vs docetaxel A similar pattern was seen with PFS in patients with baseline CNS metastases, but not in patients without baseline CNS metastases OS advantage seen with atezolizumab vs docetaxel irrespective of tobacco use Conclusion: Subgroup analyses in the OAK study demonstrated there was broad clinically-relevant efficacy with atezolizumab vs docetaxel with respect to OS

25 Safety and Efficacy Analyses of Atezolizumab in Advanced Non-Small Cell Lung Cancer Patients With or Without Baseline Brain Metastases Lukas R, Gandhi M, O Hear C, Hu S, Ballinger M, Lai C, Patel JD Lukas R, et al. Presented at: European Lung Cancer Conference; 5-8 May 2017: Geneva, Switzerland. Abstract 81O.

26 Patient Source and Subgroups for Analyses Pooled Safety Analyses OAK Efficacy Analyses a Patients with and without baseline BM (PCD4989g, BIRCH, FIR, POPLAR, OAK) N = 1452 Patients with and without baseline BM (OAK) N = 850 Atezolizumab Atezolizumab 1200 mg IV q3w Docetaxel 75 mg/m 2 IV q3w With BM (n = 79) Without BM (n = 1373) With BM n = 38 Without BM n = 387 With BM n = 47 Without BM n = 378 AEs, SAEs Treatment-related AEs Neurologic AEs OS Time to development of new brain lesions a Includes the first 850 randomized patients evaluated in primary analysis (from a total of 1225 randomized in OAK) BM, brain metastases; AE, adverse event; IV, intravenous; OS, overall survival; q3w, every 3 weeks; SAE, serious adverse event PCD4989g (NCT ); BIRCH (NCT ); FIR (NCT ); POPLAR (NCT ); OAK (NCT ) Lukas R, et al. Presented at: European Lung Cancer Conference; 5-8 May 2017: Geneva, Switzerland. Abstract 81O.

27 CNS-Specific Patient Criteria For pooled safety analysis, patients with treated asymptomatic CNS metastases were included For efficacy analysis in OAK, CNS metastases-related eligibility criteria included: Measurable disease outside the CNS Only asymptomatic supratentorial metastases allowed (no infratentorial, spinal, or leptomeningeal disease) No history of intracranial hemorrhage No ongoing requirement for corticosteroids No stereotactic radiation within 7 days or whole-brain radiation within 14 days prior to cycle 1, day 1 No evidence of interim progression between the completion of CNS-directed therapy and the screening radiographic study Patients with new asymptomatic CNS metastases detected at the screening scan must have received radiation therapy and/or surgery for CNS metastases Lukas R, et al. Presented at: European Lung Cancer Conference; 5-8 May 2017: Geneva, Switzerland. Abstract 81O.

28 Baseline Characteristics Pooled Safety Population OAK Efficacy Population Baseline BM Without Baseline BM With Baseline BM Without n = 1373 With n = 79 Atezolizumab n = 387 Docetaxel n = 378 Atezolizumab n = 38 Docetaxel n = 47 Median age (range), y 63.0 (24-85) 59.0 (42-74) 63.0 (33-82) 64.0 (34-85) 61.5 (42-74) 62.0 (40-83) Male 62% 51% 61% 62% 63% 53% ECOG PS, % 0 /1 34% / 66% 33% / 67% 36% / 64% 37% / 63% 37% / 63% 45% / 55% History of tobacco use, % Never 18% 13% 20% 16% 13% 21% Current/previous 82% 87% 80% 84% 87% 79% Histology, % Nonsq / sq 71% / 29% 87% / 13% 72% / 28% 73% / 27% 87% / 13% 83% /17% PD-L1 expression status, a % TC3 or IC3 29% 24% 17% 16% 18% 11% TC1/2/3 or IC1/2/3 78% 67% 57% 53% 63% 51% a POPLAR and OAK were in PD-L1 unselected patients, FIR and BIRCH were in PD-L1 selected patients, PCD4989g was in PD-L1 selected and unselected patients. BM, brain metastases; ECOG PS, Eastern Cooperative Oncology Group Performance Status; IC, tumor-infiltrating immune cell; Nonsq, nonsquamous; Sq, squamous; TC, tumor cell Lukas R, et al. Presented at: European Lung Cancer Conference; 5-8 May 2017: Geneva, Switzerland. Abstract 81O.

29 AE, n (%) Pooled Safety Analysis Without Baseline BM n = 1373 With Baseline BM n = 79 Any AE 1261 (92%) 75 (95%) Any neurologic AE 414 (30%) 37 (47%) Treatment-related AEs 876 (64%) 55 (70%) Any treatment-related neurologic AEs 128 (9%) 14 (18%) Treatment-related Gr3 neurologic AEs 10 (1%) 2 (3%) Any SAE 492 (36%) 26 (33%) Any neurologic SAE 36 (3%) 5 (6%) Treatment-related SAEs 135 (10%) 7 (9%) Treatment-related neurologic SAE 7 (0.5%) 0 Discontinued treatment due to AEs 99 (7%) 8 (10%) Following atezolizumab monotherapy as 2L+ treatment, the incidences of all AEs were similar in patients with or without brain metastases As expected, numerically higher rates of neurologic AEs were reported in patients with vs those without baseline brain metastases (47% vs 30%) AE, adverse event; SAE, serious adverse event Lukas R, et al. Presented at: European Lung Cancer Conference; 5-8 May 2017: Geneva, Switzerland. Abstract 81O.

30 Efficacy Analysis From OAK Overall Survival Patients With Baseline Brain Metastases In patients with pretreated brain metastases, mos was longer in those treated with atezolizumab vs docetaxel Lukas R, et al. Presented at: European Lung Cancer Conference; 5-8 May 2017: Geneva, Switzerland. Abstract 81O.

31 Efficacy Analysis From OAK Overall Survival Patients Without Baseline Brain Metastases In patients without brain metastases, mos was longer in those treated with atezolizumab vs docetaxel Lukas R, et al. Presented at: European Lung Cancer Conference; 5-8 May 2017: Geneva, Switzerland. Abstract 81O.

32 Time to Development of New Brain Lesions Patients With Baseline Brain Metastases Atezolizumab n = 38 New brain lesion free rate, % Docetaxel n = 47 6 months 84% 66% 12 months 73% 28% 18 months 73% NE ( 28%) 24 months 73% NE ( 28%) HR based on log-rank test. HR, hazard ratio; NE, not estimable P values for descriptive purposes only Lukas R, et al. Presented at: European Lung Cancer Conference; 5-8 May 2017: Geneva, Switzerland. Abstract 81O.

33 Time to Development of New Brain Lesions Patients Without Baseline Brain Metastases Atezolizumab n = 387 Docetaxel n = 378 New brain lesion free rate, % 6 months 94% 93% 12 months 90% 89% 18 months 86% 81% 24 months 86% NE ( 81%) HR based on log-rank test. HR, hazard ratio; NE, not estimable P values for descriptive purposes only Lukas R, et al. Presented at: European Lung Cancer Conference; 5-8 May 2017: Geneva, Switzerland. Abstract 81O.

34 Conclusions This pooled analysis, including 1452 atezolizumab-treated patients from 5 phase I-III studies in advanced NSCLC, demonstrated that atezolizumab has an acceptable safety profile in patients who have asymptomatic and previously-treated stable brain metastases Treatment-related grade 3 neurologic AEs were observed in 2 (3%) of patients with baseline brain metastases No treatment-related grade 4/5 neurologic AEs or SAEs were observed in patients with baseline brain metastases in the pooled safety population Data from OAK suggest that in patients with pretreated brain metastases, treatment with atezolizumab showed longer OS compared with docetaxel The risk of developing new CNS lesions in patients with baseline brain metastases appears to be lower with atezolizumab than with docetaxel These results support further investigation of atezolizumab in NSCLC patients with CNS metastases Lukas R, et al. Presented at: European Lung Cancer Conference; 5-8 May 2017: Geneva, Switzerland. Abstract 81O.

35 PD-1/PD-L1 Antibodies in NSCLC Second Line and + Nivolumab data Pembrolizumab data Atezolizumab data Durvalumab data

36 Papadimitrakopoulou VA, et al. Presented at: European Lung Cancer Conference; 5-8 May 2017: Geneva, Switzerland. Abstract 83O. A Phase II Study of Durvalumab for Previously Treated Patients with Stage IV Squamous NSCLC Papadimitrakopoulou VA, Redman MW, Borghaei H, Naheed Qaqar S, Robert F, Kiefer GJ, McDonough S, Herbst R, Kelly K, Gandara DR

37 Lung-MAP Schema Biomarker- Driven Substudies Nonmatch Substudy Single-Arm Phase II S1400B PI3K+ GDC-0032 S1400C CCGA+* Palbociclib S1400D FGFR+ AZD4547 S1400A Checkpoint Naive Randomized Phase III Durvalumab Docetaxel GDC vs TBD Palbociclib vs TBD AZD4547 vs TBD Progression following 1 year of treatment Retreatment with durvalumab (optional) ELIGIBILITY: 1. Pts must have received at least one line of systemic therapy at least one platinum-containing chemotherapy. 2. Pts could register to the nonmatch sub study if they didn t satisfy substudy-specific eligibility for the matched substudies. 3. Pts must have adequate tumor available for genomic screening. 4. EGFR mut, ALK gene fusion excluded * CCGA, cell cycle gene alteration Papadimitrakopoulou VA, et al. Presented at: European Lung Cancer Conference; 5-8 May 2017: Geneva, Switzerland. Abstract 83O.

38 Papadimitrakopoulou VA, et al. Presented at: European Lung Cancer Conference; 5-8 May 2017: Geneva, Switzerland. Abstract 83O. Baseline Characteristics Characteristic Durvalumab n = 68 Docetaxel n = 30 Median (range) age, years 66 (35.3, 91.8) 71.2 (50.2, 83) Male, % White / Black / Asian, % 81 / 9/7 90 / 3/3 Zubrod PS 0 / 1, % 26 / / 53 Never smoker, % 6 10 Current or former smoker, % PS, performance status;

39 Papadimitrakopoulou VA, et al. Presented at: European Lung Cancer Conference; 5-8 May 2017: Geneva, Switzerland. Abstract 83O. PD-L1 Expression Status and Concurrent Alterations Among Patients Randomized to Durvalumab (N = 68) N (%) Pts with known PD-L1 status * 43 (63%) PD-L1 Neg/Low (0-24%) PD-L1 High ( 25%) 29/43 (67%) 14/43 (33%) Distribution of PD-L1 Expression Levels (N = 43) Pts could register to the nonmatch substudy if they didn t satisfy substudy-specific eligibility for the matched substudies

40 Papadimitrakopoulou VA, et al. Presented at: European Lung Cancer Conference; 5-8 May 2017: Geneva, Switzerland. Abstract 83O. Antitumour Activity by PD-L1 Status Durvalumab Arm Docetaxel Arm PD-L1 low (<25%) n = 29 PD-L1 high ( 25%) n = 14 All patients n = 68 n = 30 Best objective response, N (%) Complete response 0 (0%) 1 (7.1%) 1 (1.5%) 0 (0%) Partial response 2 (6.9%) 1 (7.1%) 10 (14.7%) 2 (6.7%) Stable disease 11 (37.9%) 6 (42.9%) 26 (38.2%) 16 (53.3%) Progressive disease 15 (51.7%) 6 (42.9%) 30 (44.1%) 9 (30%) Not assessable 1 (3.5%) 0 (0%) 1 (1.5%) 3 (10%) ORR, % (95% CI) 6.9% (0%, 16.1%) 14.3% (0%, 32.6%) 16.2% (7.4%, 24.9%) 6.7% (0%, 15.6%) Median DoR, months (95% CI) (4.4, 18.5) - PFS 6 months, % (95% CI) 13.8% (1.2%, 26.3%) 21.4% (0%, 42.9%) 25% (14.7%, 35.3%) 13.3% (1.2%, 25.5%) ORR, objective response rate; CI, confidence interval; DoR, duration of response;

41 Papadimitrakopoulou VA, et al. Presented at: European Lung Cancer Conference; 5-8 May 2017: Geneva, Switzerland. Abstract 83O. Objective Response Rate Odds Ratios According to Subgroups Characteristic Odds Ratio Lower 95% CI Upper 95% CI P value Age: 65 and older vs less than Zubrod PS: 0/1 vs Smoking: current/former vs never TMB level: high vs intermediate/low PD-L1: high vs negative/low

42 Papadimitrakopoulou VA, et al. Presented at: European Lung Cancer Conference; 5-8 May 2017: Geneva, Switzerland. Abstract 83O. Overall Survival All Patients by Treatment Arm All Durvalumab Patients by PD-L1 Status

43 Garassino MC, et al. Presented at: European Lung Cancer Conference; 5-8 May 2017: Geneva, Switzerland. Abstract 82O. Durvalumab in 3rd-Line EGFR Mutant/ALK+, Locally Advanced or Metastatic NSCLC: Results From the Phase II ATLANTIC Study Garassino MC, Cho B-C, Gray JE, Mazières J, Park K, Soo RA, Dennis PA, Huang Y, Wadsworth C, Rizvi N

44 ATLANTIC: Phase II, Open-Label, Single-Arm Study NSCLC patients (stage IIIB/IV) 2 prior systemic treatment regimens, including 1 platinumbased CT and 1 TKI (EGFRmut/ALK+ patients) Recent ( 3 months) tumor biopsy and archived tumor tissue block for PD-L1 testing Protocol amendment restricted selection to patients with PD-L1 high tumors* Durvalumab IV 10 mg/kg q2w up to 12 months Cohort 1 (n = 111) EGFRmut/ALK+ PD-L1 high ( 25% tumor cells) and PD-L1 low/negative Cohort 2 (n = 265) EGFR/ALK wildtype PD-L1 high ( 25% tumor cells) and PD-L1 low/negative Cohort 3 (n = 68) EGFR/ALK wildtype PD-L1 high ( 90% tumor cells) Primary endpoint: ORR Secondary endpoints: DCR, DoR, TTR, PFS, OS Safety PK Immunogenicity Cohorts were independent ClinicalTrials.gov number: NCT *PD-L1 expression levels assessed by immunohistochemistry (VENTANA PD-L1 [SP263] Assay); ORR by independent central review (RECIST v1.1) CT, chemotherapy; DCR, disease control rate; DoR, duration of response; mut, mutated; NSCLC, non-small cell lung cancer; ORR, objective response rate; OS, overall survival; PFS, progressive-free survival; PK, pharmacokinetics; TKI, tyrosine kinase inhibitor; TTR, Time to response Garassino MC, et al. Presented at: European Lung Cancer Conference; 5-8 May 2017: Geneva, Switzerland. Abstract 82O.

45 Patient Baseline Characteristics Characteristic Cohort 1 (n = 111) Median age, years 61.0 Male, n (%) 41 (36.9) White / Asian, n (%) 44 (39.6) / 66 (59.5) WHO performance status 0 / 1, n (%) 45 (40.5) / 65 (58.6) Squamous / nonsquamous, n (%) 1 (0.9) / 110 (99.1) Never / former / current smoker, n (%) 65 (58.6) / 42 (37.8) / 4 (3.6) CNS metastases (asymptomatic ± previously treated), n (%) 25 (22.5) Mutation status, n (%) EGFRmut ALK+ EGFRmut and ALK+ 97 (87.4)* 15 (13.5)* 1 (0.9)* Number of prior regimens 2 / 3 / 4, n (%) 32 (28.8) / 33 (29.7) / 46 (41.4) Mean Median (range) PD-L1 expression status, n (%) <25% tumor cells 25% tumor cells Unknown Missing (2 11) 30 (27.0) 77 (69.4) 3 (2.7) 1 (0.9) *One patient is counted in all three subgroups; Recruited prior to the protocol amendment to include only PD-L1-high patients CNS, central nervous system; WHO, World Health Organisation Garassino MC, et al. Presented at: European Lung Cancer Conference; 5-8 May 2017: Geneva, Switzerland. Abstract 82O.

46 Antitumor Activity (Full Analysis Set*) Confirmed ORR, n (%) [95% CI] EGFRmut/ALK+ n = 74 9 (12.2) [ ] PD-L1 High ( 25%) PD-L1 Low/Neg (<25%) EGFRmut n = 64 9 (14.1) [ ] ALK+ n = 10 0 EGFRmut/ALK+ n = 28 1 (3.6) [ ] Stable disease 8 weeks, n (%) 23 (31.1) 21 (32.8) 2 (20.0) 5 (17.9) Progressive disease, n (%) 40 (54.1) 32 (50.0) 8 (80.0) 22 (78.6) Median DoR, months (95% CI) 7.4 (5.4, 9.2) 7.4 (5.4, 9.2) NC 7.9 (NC) *Patients evaluable for response per Independent Central Review; 2 patients were not evaluable due to incomplete post baseline assessments; All responses were partial responses. NC, not calculated (due to zero or one responders) Garassino MC, et al. Presented at: European Lung Cancer Conference; 5-8 May 2017: Geneva, Switzerland. Abstract 82O.

47 Garassino MC, et al. Presented at: European Lung Cancer Conference; 5-8 May 2017: Geneva, Switzerland. Abstract 82O. Best Change in Target Lesion Size (Full Analysis Set*) PD-L1 High ( 25%) PD-L1 Low/Negative (<25%) Best Change From Baseline in Target Lesion Size, % ALK+ only EGFRmut only Best Change From Baseline in Target Lesion Size, % ALK+ only EGFRmut only *Patients evaluable for response per Independent Central Review (only patients who had a post-baseline tumor assessment are shown on the graphs); Best objective response is progression, due to disease progression in nontarget lesions

48 Progression-Free Survival (Full Analysis Set*) Probability of PFS PD-L1 status: high ( 25%) vs low/neg (<25%) Median PFS (95% CI) PD-L1 25% (n = 77): 1.9 ( ) PD-L1 <25% (n = 30): 1.9 ( ) Probability of PFS EGFRmut vs ALK+: PD-L1 high ( 25%) Median PFS (95% CI) PD-L1 25% EGFRmut (n = 66) : 2.0 ( ) PD-L1 25% ALK+ (n = 12) : 1.8 ( ) Time From First Dose (months) No. of patients at risk: PD-L1 25% PD-L1 <25% Time From First Dose (months) No. of patients at risk: EGFRmut ALK *Based on Independent Central Review assessments according to RECIST v1.1; One patient who was EGFRmut and ALK+ is included in both subgroups Garassino MC, et al. Presented at: European Lung Cancer Conference; 5-8 May 2017: Geneva, Switzerland. Abstract 82O.

49 Garassino MC, et al. Presented at: European Lung Cancer Conference; 5-8 May 2017: Geneva, Switzerland. Abstract 82O. Overall Survival (Full Analysis Set) Probability of OS PD-L1 status: high ( 25%) vs low/neg (<25%) Median OS (95% CI)* PD-L1 25% (n = 77): 13.3 (8.1 NC) PD-L1 <25% (n = 30): 9.9 ( ) 1-year OS 54.8% ( ) 40.0% ( ) Time from first dose (months) No. of patients at risk: PD-L1 25% PD-L1 <25% Probability of OS EGFRmut vs ALK+: PD-L1 high ( 25%) Median OS (95% CI) PD-L1 25% EGFRmut (n = 66): NR (8.2 NC) PD-L1 25% ALK+ (n = 12): 6.3 (0.9 NC) 1-year OS 57.4% ( ) 35.7% ( ) Time from first dose (months) No. of patients at risk: EGFRmut ALK *Median follow-up for OS was 6.5 months (PD-L1 25%) and 8.2 months (PD-L1 <25%); NC, not calculated; NR, not reached Median follow-up for OS was 6.7 months (EGFRmut) and 5.3 months (ALK+)

50 Garassino MC, et al. Presented at: European Lung Cancer Conference; 5-8 May 2017: Geneva, Switzerland. Abstract 82O. Subsequent Anticancer Therapy (Safety Analysis Set)* Anticancer therapy Cohort 1 (n = 111) Post discontinuation anticancer therapy, n (%) 43 (38.7) Radiotherapy 15 (13.5) Erlotinib 9 (8.1) Docetaxel 8 (7.2) Gefitinib 8 (7.2) Pemetrexed 6 (5.4) Carboplatin 4 (3.6) Osimertinib 4 (3.6) Gemcitabine 3 (2.7) Paclitaxel 3 (2.7) Afatinib 2 (1.8) Gimeracil + oteracil potassium + tegafur 2 (1.8) *Subsequent therapies received by 2 or more patients are included in the table In Cohort 1, no patients went on to receive another immune checkpoint inhibitor

51 Garassino MC, et al. Presented at: European Lung Cancer Conference; 5-8 May 2017: Geneva, Switzerland. Abstract 82O. Conclusions First trial prospectively evaluating immune checkpoint blockade in EGFRmut/ALK+ patients Patients were heavily pretreated (means 4 prior regimens) ORR in this cohort (PD-L1 high, 12.2%) was comparable to cohort 2 (EGFR/ALK wild type PD-L1 high, 16.4%) 1 Higher PD-L1 expression (25% cutoff) appeared to be associated with higher ORR, although the small number of patients with PD-L1 low/negative tumors complicates such a comparison OS data are encouraging in PD-L1 high and low/negative patients but limited by the short duration of follow up Overall safety profile of durvalumab in the EGFRmut/ALK+ population did not differ from the EGFR/ALK wild type cohort Most adverse events (AEs) were low grade and immune-mediated AEs were manageable Ongoing phase III studies will clarify the role of durvalumab in stage III/IV NSCLC: MYSTIC and NEPTUNE: 1 st -line durvalumab ± tremelimumab vs standard of care in metastatic NSCLC ARCTIC: 3 rd -line durvalumab ± tremelimumab vs standard of care in metastatic NSCLC PACIFIC: durvalumab monotherapy following chemoradiation in stage III NSCLC 1. Garassino MC, et al. Presented at: IASLC 17 th World Conference on Lung Cancer; 4-7 December 2016: Vienna, Austria. Abstract PL04a.03.

52

53 Anti-PD-1 Phase III Trials in 1 st -Line Advanced NSCLC (>10,000 Patients) Nivolumab CHECKMATE 227 Pembrolizumab KEYNOTE-189 Nivolumab CHECKMATE 026 Pembrolizumab KEYNOTE-042 Treatment-naïve or recurrent NSCLC N = 1980 Treatment-naïve nonsquamous NSCLC N = 580 Treatment-naïve nonsquamous NSCLC PD-L1 positive NSCLC N = 495 Treatment-naïve nonsquamous NSCLC PD-L1 positive NSCLC N = 1240 Nivolumab Nivolumab + ipilimumab Platinum-based chemotherapy Pembrolizumab + pemetrexed/platinum Pemetrexed/platinum Nivolumab 3 mg/kg IV q2w ICC a with potential for crossover Pembrolizumab 200 mg IV q3w SOC chemotherapy Primary endpoints: OS, PFS Primary endpoints: PFS Primary endpoint: PFS Primary endpoint: OS Durvalumab MYSTIC Advanced NSCLC N = 675 Durvalumab Durvalumab + tremelimumab SOC chemotherapy Primary endpoint: PFS Durvalumab NEPTUNE First-line metastatic NSCLC N = 800 Durvalumab + Tremelimumab SOC chemotherapy Primary endpoint: OS Anti-PD-L1 Atezolizumab Impower 110 Atezolizumab Impower 111 Atezolizumab Impower 130 Stage IV nonsquamous PD-L1+ NSCLC N = 400 Stage IV squamous PD-L1+ NSCLC N = 400 Stage IV nonsquamous NSCLC N = 550 Atezolizumab Carboplatin or carboplatin + pemetrexed Atezolizumab Gemcitabine + cisplatin or carboplatin Atezolizumab + carboplatin + nab-paclitaxel Carboplatin + nab-paclitaxel Primary endpoint: PFS Primary endpoint: PFS Primary endpoint: PFS Atezolizumab Impower 131 Stage IV squamous NSCLC N = 1200 Atezolizumab + carboplatin + nab-paclitaxel Atezolizumab + carboplatin + paclitaxel Carboplatin + nab-paclitaxel Primary endpoint: PFS Atezolizumab Impower 150 Avelumab JAVELIN Lung 100 Stage IV nonsquamous NSCLC N = 1200 Treatment-naïve nonsquamous NSCLC PD-L1 positive NSCLC N = 420 Atezolizumab + carboplatin + paclitaxel Atezolizumab + bevacizumab + paclitaxel + carboplatin Bevacizumab + paclitaxel + carboplatin Avelumab 10 mg/kg IV q2w Platinum-based chemotherapy Primary endpoint: PFS Primary endpoint: PFS Courtesy of S. Peters

54 PD-1/PD-L1 Antibodies in NSCLC Second Line and + Nivolumab data Pembrolizumab data Atezolizumab data Durvalumab data First Line Nivolumab data 50

55 Phase III CheckMate-026 Study Design: Nivolumab vs Chemotherapy in First-Line NSCLC Key eligibility criteria: Stage IV or recurrent NSCLC No prior systemic therapy for advanced disease No EGFR/ALK mutations sensitive to available targeted inhibitor therapy 1% PD-L1 expression a CNS metastases permitted if adequately treated at least 2 weeks prior to randomization Randomize 1:1 Nivolumab 3 mg/kg IV q2w n = 271 Tumor scans q6w until wk 48 then q12w Disease progression or unacceptable toxicity Chemotherapy (histology dependent) b Maximum of 6 cycles n = 270 Disease progression Crossover nivolumab c (optional) Stratification factors at randomization: PD-L1 expression (<5% vs 5%) a Histology (squamous vs nonsquamous) Primary endpoint: PFS ( 5% PD-L1+) d Secondary endpoints: PFS ( 1% PD-L1+) d OS ORR d a Dako 28-8 validated; archival tumor samples obtained 6 months before enrollment were permitted; PD-L1 testing was centralized b Squamous: gemcitabine 1250 mg/m 2 + cisplatin 75 mg/m 2 ; gemcitabine 1000 mg/m 2 + carboplatin AUC 5; paclitaxel 200 mg/m 2 + carboplatin AUC 6; Nonsquamous: pemetrexed 500 mg/m 2 + cisplatin 75 mg/m 2 ; pemetrexed 500 mg/m 2 + carboplatin AUC 6; option for pemetrexed maintenance therapy c Permitted if crossover eligibility criteria met, including progression confirmed by independent radiology review d Tumor response assessment for PFS and ORR per RECIST v1.1 as determined by independent central review Socinski MA, et al. Ann Oncol. 2016;27(Suppl 6): Abstract LBA7_PR

56 Primary Endpoint (PFS per IRRC in 5% PD-L1+) CheckMate-026: Nivolumab vs Chemotherapy in First-Line NSCLC PFS, % Median PFS, months (95% CI) Nivolumab n = (3.0, 5.6) Chemotherapy n = (5.4, 6.9) 1-year PFS rate, % HR = 1.15 (95% CI: 0.91, 1.45), P = Nivolumab Chemotherapy Months No. of patients at risk: Nivolumab Chemotherapy All randomized patients ( 1% PD-L1+): HR = 1.17 (95% CI: 0.95, 1.43) Socinski MA, et al. Ann Oncol. 2016;27(Suppl 6): Abstract LBA7_PR

57 OS ( 5% PD-L1+) CheckMate-026: Nivolumab vs Chemotherapy in First-Line NSCLC 100 Nivolumab n = 211 Chemotherapy n = Median OS, months (95% CI) 14.4 (11.7, 17.4) 13.2 (10.7, 17.1) 60 1-year OS rate, % OS, % 40 HR = 1.02 (95% CI: 0.80, 1.30) Chemotherapy 20 Nivolumab No. of patients at risk: Months Nivolumab Chemotherapy All randomized patients ( 1% PD-L1+): HR = 1.07 (95% CI: 0.86, 1.33) Socinski MA, et al. Ann Oncol. 2016;27(Suppl 6): Abstract LBA7_PR

58 PD-1/PD-L1 Antibodies in NSCLC Second Line and + Nivolumab data Pembrolizumab data Atezolizumab data Durvalumab data First Line Nivolumab data Pembrolizumab data 54

59 Based on the results of the KEYNOTE- 024 study, what patient characteristic would suggest front-line treatment with pembrolizumab? 1. ALK translocation 2. EGFR-sensitizing mutation 3. PD-L1 expression in 50% tumor cells (compared with >1% cutoff for secondline treatment) 4. All of the above

60 KEYNOTE-024 Study Design (NCT ) Key Eligibility Criteria Untreated stage IV NSCLC PD-L1 TPS 50% ECOG PS 0-1 No activating EGFR mutation or ALK translocation No untreated brain metastases No active autoimmune disease requiring systemic therapy R (1:1) N = 305 Pembrolizumab 200 mg IV Q3W (2 years) Platinum-Doublet Chemotherapy (4-6 cycles) Key Endpoints Primary: PFS (RECIST v1.1 per blinded, independent central review) Secondary: OS, ORR, safety Exploratory: DOR Reck M, et al. Ann Oncol. 2016;27(Suppl6): Abstract LBA8.

61 Efficacy Data Clear and strong signal of activity Imaging was every 9 weeks 45% ORR is the best RR ever reported in 1 st -line setting (and with a monotherapy!) PFS is improved by 4.3 months (HR of 0.50) Strongest signal of PFS benefit observed in SCC (HR of 0.35) Reck M, et al. Ann Oncol. 2016;27(Suppl6): Abstract LBA8.

62 Survival Data Clear survival benefit Estimated rate of 12 months: 70% (Pembro) vs 54% (Chemo) HR for death: 0.60 But crossover was limited to 44% of the patients Reck M, et al. Ann Oncol. 2016;27(Suppl6): Abstract LBA8.

63 What was the mpfs of cohort 1 in the BIRCH trial investigating atezolizumab in PD-L1-selected advanced NSCLC? months months months months

64 PD-1/PD-L1 Antibodies in NSCLC Second Line and + Nivolumab data Pembrolizumab data Atezolizumab data Durvalumab data First Line Nivolumab data Pembrolizumab data Atezolizumab data 54

65 Atezolizumab As First-Line (1L) Therapy for Advanced Non-Small Cell Lung Cancer (NSCLC) in PD-L1- Selected Patients: Efficacy Data From the BIRCH Trial Peters S, Carcereny Costa E, Garassino MC, Christoph D, Kurata T, Chaft J, Johnson ML, Mocci S, Gettinger S, Felip E Peters S, et al. Presented at: European Lung Cancer Conference; 5-8 May 2017: Geneva, Switzerland. Abstract 84O.

66 BIRCH: Phase II Trial of Atezolizumab in PD-L1 Selected Advanced NSCLC Locally advanced or metastatic NSCLC Tumor PD-L1 expression by IHC (TC2/3 and/or IC2/3) ECOG PS 0 or 1 No brain metastases N = 667 Cohort 1 (1L) No prior chemo n = 138 Cohort 2 (2L) 1 prior platinum chemo n = 271 Cohort 3 (3L+) 2 prior chemos (including 1 platinum) n = 254 PD Until loss of clinical benefit Atezolizumab dosed at 1200 mg IV q3w in all cohorts Primary endpoint: Objective response rate (ORR) per RECIST v1.1 Secondary endpoints: IRF-assessed progression-free survival (PFS), duration of response (DOR) per RECIST v1.1 Investigator (INV)-assessed ORR, PFS and DOR per RECIST v1.1 and modified RECIST Overall survival (OS) Safety Baseline PD-L1 expression was scored by IHC in tumor cells (TC) as percentage of PD-L1 expressing TC and in tumor-infiltrating immune cells (IC) as percentage of tumor area TC2/3 or IC2/3 = TC or IC 5% PD-L1 expression assayed by IHC using VENTANA SP142 antibody. Peters S, et al. Presented at: European Lung Cancer Conference; 5-8 May 2017: Geneva, Switzerland. Abstract 84O.

67 Baseline Characteristics Age, median (range) Patients n = 138 b 67 y (35-88) Male 51% ECOG PS 1 57% Current/previous tobacco use 84% Nonsquamous histology 77% EGFR mutation, no. mutated/tested [n = 117] a 13 (9%) KRAS mutation, no. mutated/tested [n = 100] a 33 (24%) TC3 or IC3 status 47% PD-L1 Prevalence c Subgroup TC3 or IC3 TC2 or IC2 ITT (TC2/3 or IC2/3) IC, tumor-infiltrating immune cells; TC, tumor cells TC3 or IC3 = TC 50% or IC 10% PD-L1 expressing cells; TC2 or IC2 = TC2/3 or IC2/3 excluding TC3 or IC3; TC2/3 or IC2/3 = TC or IC 5% PD- L1 expressing cells, respectively. a Mutational status not required at enrollment; limited data available. EGFR mutant includes at least one of the following: exon 19 deletion, exon 20 insertion, G719X, L858R, L861Q or S768I. b Treated patients. c Cohort 1. Data cutoff date: August 1, 2016 Peters S, et al. Presented at: European Lung Cancer Conference; 5-8 May 2017: Geneva, Switzerland. Abstract 84O.

68 Response Rates (INV) and Drug Exposure Frequency, % % 25% TC3 or IC3 TC2/3 or IC2/3 TC2 or IC2 18% 34% 42% 49% Atezolizumab Exposure Mean (SD) Treatment duration, mo 8.4 (8.0) Dose intensity, % 98.5 (7.0) No. of doses 12.7 (11.1) 0 CR/PR SD n = 22 n = 35 n = 13 n = 22 n = 58 n = 36 CR, complete response; INV, investigator assessed; PR, partial response; SD, stable disease TC3 or IC3 = TC 50% or IC 10% PD-L1 expressing cells; TC2 or IC2 = TC2/3 or IC2/3 excluding TC3 or IC3; TC2/3 or IC2/3 = TC or IC 5% PD-L1 expressing cells, respectively. Error bars correspond to 95% CIs Data cutoff date: August 1, 2016 Peters S, et al. Presented at: European Lung Cancer Conference; 5-8 May 2017: Geneva, Switzerland. Abstract 84O.

69 Duration of Response and Progression- Free Survival in PD-L1 Subgroups Endpoint TC2/3 or IC2/3 n = 138 TC3 or IC3 n = 65 TC2 or IC2 n = 73 mdor (95% CI) 16.5 mo (9.9, NE) NE (8.5, NE) 12.3 mo (8.3, 17.9) mpfs (95% CI) 7.3 mo (5.7, 9.7) 7.3 mo (4.9, 12.0) 7.6 mo (4.0, 9.7) 12-mo PFS rate (95% CI) 32.5% (24.2, 40.8) 36.5% (24.0, 48.9) 28.9% (17.9, 40.0) N/A, not available; NE, not estimable TC3 or IC3 = TC 50% or IC 10% PD-L1 expressing cells; TC2/3 or IC2/3 = TC or IC 5% PD-L1 expressing cells, respectively; TC2 or IC2 = TC2/3 or IC2/3 excluding TC3 or IC3. Data cutoff date: August 1, 2016 Peters S, et al. Presented at: European Lung Cancer Conference; 5-8 May 2017: Geneva, Switzerland. Abstract 84O.

70 ORR, DOR and PFS by Mutation and PD-L1 Status Endpoint EGFR Mutant a n = 13 EGFR Wildtype n = 104 KRAS Mutant n = 33 KRAS Wildtype n = 67 INV ORR TC2/3 or IC2/3, n (%) INV ORR TC3 or IC3, n (%) 4 (31%) 23 (22%) 10 (31%) 15 (22%) 1 (25%) 15 (31%) 6 (38%) 10 (30%) mdor, mo (95% CI) NE (5.6, NE) 8.5 mo (5.6, 12.3) 10.0 mo (6.9, NE) 7.1 mo (5.2, 12.3) mpfs, mo (95% CI) 5.5 mo (2.6, 8.3) 5.5 mo (3.0, 6.9) 8.2 mo (4.2, 17.9) 7.1 mo (3.9, 9.9) Investigator assessments a Patients were considered to be EGFR mutant for the analysis if their tumors tested positive for at least one of the following: exon 19 deletions or insertions, L858R, exon 20 insertion, G719X, L861Q or S768I. Three patients with a T790M mutation were not included in this analysis; two of those also had an exon 19 deletion. TC3 or IC3 = TC 50% or IC 10% PD-L1 expressing cells; TC2/3 or IC2/3 = TC or IC 5% PD-L1 expressing cells, respectively Data cutoff date: August 1, 2016 Peters S, et al. Presented at: European Lung Cancer Conference; 5-8 May 2017: Geneva, Switzerland. Abstract 84O.

71 Overall Survival Median duration of survival follow-up = 22.5 months mos (95% CI) TC3 or IC3 (n = 65) 26.9 mo (12.0, NE) 12-mo OS rate (95% CI) 61.5% (49.0, 74.0) mos (95% CI) TC2/3 or IC2/3 (n = 138) 23.5 mo (18.1, NE) 12-mo OS rate (95% CI) 66.4% (58.1, 74.6) mos (95% CI) TC2 or IC2 (n = 73) 23.5 mo (18.1, NE) 12-mo OS rate (95% CI) 70.7% (59.8, 81.6) EGFR a mos (95% CI) Mutant 26.0 mo (20.1, 26.0) Wildtype 20.1 mo (15.5, NE) KRAS Mutant Wildtype mos (95% CI) 24.1 mo (20.3, NE) 23.5 mo (15.5, NE) NE, not estimable TC3 or IC3 = TC 50% or IC 10% PD-L1 expressing cells; TC2 or IC2 = TC2/3 or IC2/3 excluding TC3 or IC3; TC2/3 or IC2/3 = TC or IC 5% PD-L1 expressing cells, respectively. a Patients were considered to be EGFR mutant for the analysis if their tumor tested positive for at least one of the following: exon 19 deletions or insertions, L858R, exon 20 insertion, G719X, L861Q or S768I. Three patients with a T790M mutation were not included in this analysis; two of whom also had an exon 19 deletion. Data cutoff date: August 1, 2016 Peters S, et al. Presented at: European Lung Cancer Conference; 5-8 May 2017: Geneva, Switzerland. Abstract 84O.

72 PD-1/PD-L1 Antibodies in NSCLC Second Line and + Nivolumab data Pembrolizumab data Atezolizumab data Durvalumab data First Line Nivolumab data Pembrolizumab data Atezolizumab data Combo with chemotherapy 54

73 KEYNOTE-021 Cohort G Key Eligibility Criteria Untreated stage IIIB or IV nonsquamous NSCLC No activating EGFR mutation or ALK translocation Provision of a sample for PD-L1 assessment a ECOG PS 0-1 No untreated brain metastases No ILD or pneumonitis requiring systemic steroids R (1:1) a N=123 Pembrolizumab 200 mg Q3W for 2 years + Carboplatin AUC 5 mg/ml/min + Pemetrexed 500 mg/m 2 Q3W for 4 cycles Carboplatin AUC 5 mg/ml/min + Pemetrexed 500 mg/m 2 Q3W for 4 cycles PD Pemetrexed 500 mg/m 2 Q3W permitted as maintenance therapy Pembrolizumab 200 mg Q3W for 2 years Endpoints Primary: ORR (RECIST v1.1 per blinded, independent central review) Key secondary: PFS Other secondary: OS, safety, relationship between antitumor activity and PD-L1 TPS a Randomization was stratified by PD-L1 TPS <1% vs 1% b Indefinite maintenance therapy with pemetrexed 500 mg/m 2 Q3W permitted. Langer CJ, et al. Ann Oncol. 2016;27(Suppl 6): Abstract LBA46. Langer CJ, et al. Lancet Oncol. 2016;17(11):

74 Survival Data Clear PFS benefit and no OS advantage Median PFS improved by 4.1 months PFS HR is 0.53 No difference for OS In chemo arm crossover is 48% to PD1 therapies (pembro & others) Langer CJ, et al. Ann Oncol. 2016;27(Suppl 6): Abstract LBA46. Langer CJ, et al. Lancet Oncol. 2016;17(11):

75 Second Line and + Nivolumab data Pembrolizumab data Atezolizumab data Durvalumab data First Line Nivolumab data Pembrolizumab data Atezolizumab data Combo with chemotherapy Combo with CTLA-4 54 PD-1/PD-L1 Antibodies in NSCLC

76 CheckMate 012: Safety and Efficacy of First Line Nivolumab and Ipilimumab in Advanced NSCLC Hellmann MD, Gettinger SN, Goldman J, Brahmer J, Borghaei H, Chow LQ, Ready NE, Gerber DE, Juergens R, Shepherd FA, Laurie SA, Young T, Geese WJ, Agrawal S, Li X, Antonia SJ Hellman MD, et al. J Clin Oncol. 2016;34(suppl): Abstract 3001.

77 Nivolumab Plus Ipilimumab in First-Line NSCLC: Summary of Efficacy Nivo 3 Q2W + Ipi 1 Q12W n = 38 Nivo 3 Q2W + Ipi 1 Q6W n = 39 Nivo 3 Q2W n = 52 Confirmed ORR, % (95% CI) 47 (31, 64) 39 (23, 55) 23 (13, 37) Median duration of response, mo (95% CI) NR (11.3, NR) NR (8.4, NR) NR (5.7, NR) Median length of follow-up, mo (range) 12.9 ( ) 11.8 ( ) 14.3 ( ) Best overall response, % Complete response Partial response Stable disease Progressive disease Unable to determine Median PFS, mo (95% CI) 8.1 (5.6, 13.6) 3.9 (2.6, 13.2) 3.6 (2.3, 6.6) 1-year OS rate, % (95% CI) NC 69 (52, 81) 73 (59, 83) NC, not calculated (when >25% of patients are censored); NR, not reached Combination data based on a February 2016 database lock; monotherapy data based on a March 2015 database lock except for OS data, which are based on an August 2015 database lock Hellman MD, et al. J Clin Oncol. 2016;34(suppl): Abstract 3001.

78 Second Line and + Nivolumab data Pembrolizumab data Atezolizumab data Durvalumab data First Line Nivolumab data Pembrolizumab data Atezolizumab data Combo with chemotherapy Combo with CTLA-4 Toxicities 54 PD-1/PD-L1 Antibodies in NSCLC

79 Which of the following would you NOT consider an immune-related adverse event in a patient receiving treatment for NSCLC with a checkpoint inhibitor? 1. Pneumonitis 2. Rash 3. Hypertension 4. Colitis 5. Hepatotoxicity

80 Time to Onset of First Treatment-Related Select Adverse Events by Category (Any Grade) Number of Patients With First Event in Category Patients still on study, n Patients still on treatment, n Total patients with a first event, n Months Skin Gastrointestinal Pulmonary Endocrine Renal Hypersensitivity/Infusion reaction Hepatic 0-3 >3-6 >6-12 > July 2014 DBL. Includes events reported between first dose and 30 days after last dose of study therapy Within each time interval, patients with 1 event were counted only once in each category but could be classified into more than one category Horn L, et al. J Thorac Oncol. 2015;9(Suppl 2): Abstract ORAL02.03.

81 Nivolumab Plus Ipilimumab in First-Line NSCLC: Treatment-Related Select AEs Nivo 3 Q2W + Ipi 1 Q12W (n = 38) Nivo 3 Q2W + Ipi 1 Q6W (n = 39) Nivo 3 Q2W (n = 52) Grade 1 2 Grade 3 4 Total Patients With an Event, % All treatment-related pulmonary events were pneumonitis Grade 1 2 hypersensitivity/infusion reaction occurred in 5% and 6% of patients in the nivo 3 q2w + ipi 1 q12w and monotherapy groups, respectively Combination data based on a February 2016 database lock; monotherapy data based on a March 2015 database lock Select AEs are those with potential immunologic etiology that require frequent monitoring/intervention Hellman MD, et al. J Clin Oncol. 2016;34(suppl): Abstract 3001.

82 What do you consider to be the ideal place for PD-1/PD-L1 inhibitors in NSCLC? 1. First line in combination with chemotherapy in patients without targetable mutations 2. First-line monotherapy in patients without targetable mutations 3. First line in combination with CTLA-4 inhibitor in patients without targetable mutations 4. Second-line therapy (regardless of mutation status) 5. In combination with targeted therapy (either first-line or second-line) 6. After all other treatment options have been exhausted

83 Conclusion Anti PD-1 & PD-L1 Standard in 2 nd -line NSCLC Nivolumab all comers Pembrolizumab if PDL1 + ( 1% TC) Atezolizumab all comers Standard in 1 st -line NSCLC Pembrolizumab if PDL1 strong + (>50%) Future standard in 1 st -line NSCLC Pembrolizumab + CT all comers? Combinations are challenging and toxic