1310P - Survival and safety of atezolizumab by best overall response (BOR) in the phase III NSCLC OAK study

Abstract

Background: Atezolizumab (atezo; anti–PD-L1) inhibits binding of PD-L1 to PD-1 and B7.1, thereby restoring tumor-specific T-cell immunity. OAK, the first randomized Ph III study of atezo vs docetaxel (doc) in 2L/3L NSCLC demonstrated a superior OS benefit of atezo (HR 0.73; 95% CI: 0.62, 0.87; P = 0 .0003) in patients (pts) regardless of PD-L1 expression levels on tumor cells (TC) or tumor-infiltrating immune cells (IC). Here we present efficacy and safety analyses in the OAK primary population (n = 850) by BOR subgroups. Methods: Previously treated pts were randomized 1:1 to atezo (1200 mg) or doc (75 mg/m2) IV q3w. Co-primary endpoints were OS in ITT and PD-L1 expression subgroup (≥ 1% PD-L1 on TC or IC). Secondary endpoints included ORR and safety. BOR subgroups were defined based on RECIST v1.1 response determined by investigators. Time to response (TTR) was based on tumor assessment every 6 weeks. Data cutoff, July 7, 2016. Results: Baseline demographics were generally similar across BOR subgroups except for PD-L1 expression status. TC1/2/3 or IC1/2/3 prevalence was 74% in CR/PR, 53% in SD and 55% in PD subgroups, respectively. The survival benefit of atezo vs doc was observed across BOR subgroups with greatest benefit occurring in the CR/PR subgroup (HR, 0.32; 95% CI 0.16, 0.63; see Table). Among pts in the CR/PR subgroup, the median TTR was comparable between atezo and doc arms (2.8 mo each). The median duration of response was longer in the atezo arm pts (16.3 mo, 95% CI: 10.0, NE) vs doc arm pts (6.2 mo, 95% CI: 4.9, 7.6). OS benefit with atezo was also observed in patients with SD and PD as BOR (see Table). No new safety findings were observed among BOR subgroups.Table1310P Efficacy of atezolizumab vs docetaxel by BOR subgroupsPatient PopulationAtezolizumabDocetaxelHR (95% CI)aStratified HR for ITT and unstratified HR for subgroups. 95% CI for HR were estimated using Cox regression.nmOS (95% CI), monmOS, (95% CI) moITT (N = 850)42513.8 (11.8, 15.7)4259.6 (8.6, 11.2)0.73 (0.62, 0.87)BOR subgroupsCR/PR58NE (NE, NE)5720.0 (15.9, NE)0.32 (0.16, 0.63)SD15017.6 (15.7, 20.2)17713.0 (11.5, 14.7)0.70 (0.53, 0.92)PD1877.3 (6.7, 9.4)1176.4 (5.6, 7.3)0.72 (0.56, 0.93)NE, not estimable.a Stratified HR for ITT and unstratified HR for subgroups. 95% CI for HR were estimated using Cox regression. Open table in a new tab NE, not estimable. Conclusions: Atezo responses were durable. Atezo responders had more than two-thirds reduction in the risk of death compared with doc responders. In addition, the improved OS with atezo vs doc seen in pts with SD and PD suggests that clinical benefit also extended to patients who did not have a radiographic response. Clinical trial identification: NCT02008227 Legal entity responsible for the study: F. Hoffmann - La Roche Ltd. Funding: F. Hoffmann - La Roche Ltd. Disclosure: F. de Marinis: Consultation fees received from Roche/BMS/Boehringer/Novartis/Pfizer/MSD/Astrazeneca. F. Barlesi: Honarium from Roche. A. Rittmeyer: Grants as an advisor or speaker by: Astra Zeneca, BMS, Boehringer Ingelheim, Eli Lilly, Pfizer and Roche Genentech. J. von Pawel: Adboard: AbbVie, Pfizer, Bristol Myers Squibb, Novartis money paid to the institution. A. Spira: Research sponsored by Roche/Genentech (payable to institution), Speakers bureau. D.R. Gandara: Consultant: Genentech Clinical trial grant: Genentech. W. Yu: Genentech Employee. P. He: Employee of Roche/Genentech, and have stocks for Roche, Amgen. Husband has stocks for Allergan and Gilead. C. Yun: Employee of Genentech, Roche stock, Research funding from Genentech M. Ballinger: Employee of Genentech, Roche stock. M. Gandhi: Employee of Genentech. S. Gadgeel: Speaker's bureau: Astra‐Zeneca, Genentech/Roche Advisory Boards‐ Astra‐Zeneca, Ariad, Pfizer, Bristol Myers‐ Squibb and Genentech/Roche. All other authors have declared no conflicts of interest.