Phase Ib Study Results: Subgroup Efficacy Analysis of Atezolizumab + - PowerPoint PPT Presentation

Phase Ib Study Results: Subgroup Efficacy Analysis of Atezolizumab + Bevacizumab in Patients With Previously Untreated, Unresectable Hepatocellular Carcinoma Kyung-Hun Lee, 1 Baek-Yeol Ryoo, 2 Chih-Hung Hsu, 3 Kazushi Numata, 4 Stacey Stein, 5 Wendy Verret, 6 Steve Hack, 6 Jessica Spahn, 6 Bo Liu, 6 Heba Abdullah, 6 Aiwu Ruth He, 7 Michael S Lee 8 1 Seoul National University Hospital, Seoul, South Korea; 2 Asan Medical Center, University of Ulsan College of Medicine, Seoul, South Korea; 3 National Taiwan University Hospital, Taipei, Taiwan; 4 Yokohama City University Medical Center, Yokohama, Japan; 5 Yale School of Medicine, New Haven, CT, USA; 6 Genentech, Inc., South San Francisco, CA, USA; 7 Georgetown University Medical Center, Washington, DC, USA; 8 Lineberger Comprehensive Cancer Center, University of North Carolina at Chapel Hill, Chapel Hill, NC, USA 13 th Annual Conference 20 ► 22 September 2019 13 th ILCA Annual Conference Chicago, USA 1 https://bit.ly/2ZCXqHc 20 ► 22 September 2019 │ Chicago, USA Lee KH. Atezo + Bev in HCC

Disclosures • Advisory board: Bayer, Ono Pharmaceutical, Samsung Bioepis, Roche, Eisai, AstraZeneca • Honoraria: Roche, AstraZeneca • This study is sponsored by F. Hoffmann-La Roche, Ltd 13 th ILCA Annual Conference 2 https://bit.ly/2ZCXqHc 20 ► 22 September 2019 │ Chicago, USA Lee KH. Atezo + Bev in HCC

Background Approximately 80% of patients with HCC present with unresectable cancer, 1 usually • complicated by underlying liver disease 1,2 • VEGFR tyrosine kinase inhibitors are the first-line systemic standard of care for unresectable or metastatic HCC 2,3 – Associated with modest survival benefits and considerable toxicities 2,3 – Unconfirmed ORRs per RECIST 1.1 were 19% with lenvatinib and 7% with sorafenib 3 • Single-agent PD-L1/PD-1 immune checkpoint inhibitors have shown clinical activity against HCC, but not yet superiority over standards of care, in randomised studies 4,5 • Many HCC tumors are hypervascularised and overexpress VEGF and PD-L1 6,7 – The anti-VEGF monoclonal antibody bevacizumab showed modest single-agent activity in HCC 8 – In addition to anti-angiogenic activity, bevacizumab’s immunomodulatory effects alter the tumour microenvironment and may boost anti–PD-L1-mediated anti-tumour response 9-11 – Dual blockade of PD-L1 and VEGF has shown clinical benefit in other tumour types 12-14 13 th ILCA Annual Conference 1. Lau WY, et al. An Surg . 2001; 2. Villanueva A, et al. NEJM. 2019; 3. Kudo M, et al. Lancet . 2018; 4. Finn RS, et al. JCO. 2019 (abstract 4004); 5. BMS press release. https://www.clinicaltrialsarena.com/news/bms-checkmate-459-trial/. 2019; 6. Morse MA, et al. CCR . 2019; 3 https://bit.ly/2ZCXqHc 7. Gao Q, et al. CCR. 2009; 8. Wattenberg MM, et al. Cancer Med. 2019; 9. Motz GT, et al. Nat Med. 2014; 10. Roland CL, et al. Mol Can Ther . 2009; 20 ► 22 September 2019 │ Chicago, USA 11. Voron T, et al. J Exp Med. 2015; 12. Reck M, et al. Lancet Respir Med. 2019; 13. Wallin JJ, et al. Nat Commun. 2016; Lee KH. Atezo + Bev in HCC 14. McDermott DF, et al. JCO 2016.

Phase Ib GO30140 Study Design (NCT02715531) Arm A: Unresectable HCC (n = 104) Eligibility Criteria: • Up to Child-Pugh score B7 Measurable disease per RECIST 1.1 • Atezolizumab 1200 mg IV q3w + ECOG PS 0/1 • Until disease bevacizumab 15 mg/kg IV q3w Adequate haematologic and • progression, Survival organ function unacceptable follow-up No prior systemic therapy • toxicity or loss of Arm B: Gastric cancer o No prior treatment with anti– clinical benefit Arm C: Pancreatic cancer CTLA-4, anti–PD-1 or anti–PD-L1 Arm E: Oesophageal cancer antibodies Arm F: Randomized first-line HCC (atezolizumab + bevacizumab vs atezolizumab) Primary endpoints (Arm A) • IRF-assessed ORR per RECIST 1.1, safety Key secondary endpoints (Arm A) • DOR and PFS per IRF RECIST 1.1 • ORR, DOR and PFS per IRF HCC mRECIST • ORR, DOR and PFS per INV-assessed RECIST 1.1 • OS At clinical data cutoff (14 June 2019), 104 patients with HCC treated with atezolizumab + bevacizumab in Arm A were evaluable for safety and efficacy with a median follow-up of 12.4 months 13 th ILCA Annual Conference 4 https://bit.ly/2ZCXqHc HCC mRECIST, HCC modified Response Evaluation Criteria in Solid Tumours; INV, investigator; IRF, independent review facility. 20 ► 22 September 2019 │ Chicago, USA Lee KH. Atezo + Bev in HCC

Baseline Demographics and Clinical Characteristics Arm A: Atezolizumab + bevacizumab Characteristic N = 104 Median age (range), years 62 (23-82) Sex, male , n (%) 84 (81) Region , n (%) Asia (excluding Japan) 59 (57) Rest of world a 45 (43) ECOG PS 1 , n (%) 52 (50) Child-Pugh class , n (%) A5 | A6 | B7 77 (74) | 21 (20) | 6 (6) Aetiology of HCC, n (%) HBV | HCV | Non-viral 51 (49) | 31 (30) | 22 (21) AFP ≥ 400 ng/mL , n (%) 37 (36) EHS , n (%) 74 (71) MVI , n (%) 55 (53) EHS and/or MVI , n (%) 91 (88) Prior TACE , n (%) 56 (54) Prior radiotherapy , n (%) 20 (19) 13 th ILCA Annual Conference a Includes the United States, Australia, New Zealand, and Japan. 5 https://bit.ly/2ZCXqHc Data cutoff: 14 June 2019. 20 ► 22 September 2019 │ Chicago, USA Lee KH. Atezo + Bev in HCC Lee MS, et al. Liver Cancer 2019;8:207 (abstract O-034).

Clinical Activity with Atezolizumab + Bevacizumab Arm A: Atezolizumab + Bevacizumab (N = 104) IRF RECIST 1.1 IRF HCC mRECIST INV RECIST 1.1 Confirmed ORR, n (%) 37 (36) 41 (39) 34 (33) (95% CI) (26 – 46) (30 – 50) (24 – 43) CR, n (%) 12 (12) 16 (15) 3 (3) PR , n (%) 25 (24) 25 (24) 31 (30) DCR, n (%) 74 (71) 74 (71) 78 (75) n = 37 n = 41 n = 34 Ongoing response, n (%) 28 (76) 28 (68) 24 (71) Median DOR, mo NE NE NE (95% CI) (11.8 – NE) (11.8 – NE) (11.7– NE) DOR range, mo 1.6+ to 31.0+ 1.6+ to 31.0+ 3.5 to 31.0+ ≥ 9 mo, n (%) 20 (54) 25 (61) 21 (62) ≥ 12 mo, n (%) 11 (30) 11 (27) 12 (35) 13 th ILCA Annual Conference DCR, CR + PR + SD; NE, not estimable; +, censored. 6 https://bit.ly/2ZCXqHc Data cutoff: 14 June 2019. 20 ► 22 September 2019 │ Chicago, USA Lee MS, et al. Liver Cancer 2019;8:207 (abstract O-034). Lee KH. Atezo + Bev in HCC

Progression Free Survival and Overall Survival Summary Arm A: Atezolizumab + Bevacizumab (N = 104) Progression Free Survival IRF RECIST 1.1 IRF HCC mRECIST INV RECIST 1.1 No. of events (%) 69 (66) 69 (66) 75 (72) Median PFS (95% CI), mo 7.3 (5.4 – 9.9) 7.3 (5.4 – 9.9) 7.4 (5.6 – 10.7) 6-month PFS rate (%) 54 55 56 12-month PFS rate (%) 35 35 38 Overall Survival No. of deaths (%) 47 (45) Median OS (95% CI), mo 17.1 (13.8 – NE) 6-month OS rate (%) 82 12-month OS rate (%) 63 13 th ILCA Annual Conference 7 https://bit.ly/2ZCXqHc Data cutoff: 14 June 2019. 20 ► 22 September 2019 │ Chicago, USA Lee KH. Atezo + Bev in HCC

Subgroup Analysis: Demographic Characteristics ORR, a % Patients, n Responders, n 95% CI b All 104 37 36 26–46 Age < 65 years 67 25 37 26–50 ≥ 65 years 37 12 32 18–50 Sex Male 84 30 36 26–47 Female 20 7 35 15–59 Region Asia (excluding Japan) 59 19 32 21–46 Rest of world 45 18 40 26–56 ECOG PS 0 52 23 44 31–59 1 52 14 27 16–41 0% 20% 40% 60% 80% 100% ORR 13 th ILCA Annual Conference a ORR by IRF RECIST 1.1. b 95% CIs were constructed using Clopper Pearson method. 8 https://bit.ly/2ZCXqHc 20 ► 22 September 2019 │ Chicago, USA Data cutoff: 14 June 2019. Lee KH. Atezo + Bev in HCC

Subgroup Analysis: HCC Characteristics ORR, a % Patients Responders 95% CI b All 104 37 36 26–46 HCC etiology Hepatitis B 51 17 33 21–48 Hepatitis C 31 16 52 33–70 Non-viral c 22 4 18 5–40 AFP level < 400 ng/ml 60 23 38 26–52 ≥ 400 ng/ml 37 13 35 20–53 Extra-hepatic spread Yes 74 21 28 19–40 No 30 16 53 34–72 Macrovascular invasion Yes 55 16 29 18–43 No 49 21 43 29–58 EHS and/or MVI Yes 91 29 32 23–43 No 13 8 62 32–86 High-risk disease d Yes 26 9 35 17-56 No 78 28 36 25-48 0% 20% 40% 60% 80% 100% ORR 13 th ILCA Annual Conference a ORR by IRF RECIST 1.1. b 95% CIs were constructed using Clopper Pearson method. c Non-viral HCC etiology includes unknown non-hepatitis B and C cause. 9 https://bit.ly/2ZCXqHc d Includes patients with bile duct invasion, main portal vein invasion and/or liver occupancy ≥ 50%. Data cutoff: 14 June 2019. 20 ► 22 September 2019 │ Chicago, USA Lee KH. Atezo + Bev in HCC

Subgroup Analysis: PD-L1 Status ORR, a % Patients Responders 95% CI b All 104 37 36 26–46 PD-L1 status c 1% cut-off TC or IC ≥ 1% 61 25 41 29–54 TC and IC < 1% 25 7 28 12–49 5% cut-off TC or IC ≥ 5% 37 17 46 30–63 TC and IC < 5% 49 15 31 18–45 10% cut-off TC or IC ≥ 10% 30 15 50 31–69 TC and IC < 10% 56 17 30 19–44 0% 20% 40% 60% 80% 100% ORR 13 th ILCA Annual Conference IC, tumour-infiltrating immune cells; TC, tumour cells. a ORR by IRF RECIST 1.1. b 95% CIs were constructed using Clopper Pearson method. c PD-L1 testing was based on IHC SP263; 10 https://bit.ly/2ZCXqHc 20 ► 22 September 2019 │ Chicago, USA 86 patients were evaluable for PD-L1 status. Data cutoff: 14 June 2019. Lee KH. Atezo + Bev in HCC