O-007Nivolumab monotherapy in patients with advanced gastric or gastroesophageal junction (GEJ) cancer and 2 or more prior treatment regimens: Sub-analysis of the CheckMate 032 study

Introduction: In the phase 3 ATTRACTION-2 (ONO-12) trial, nivolumab demonstrated an OS benefit vs placebo (primary endpoint) in Asian patients with advanced or recurrent gastric/GEJ cancer with ≥2 prior treatments (median OS, 5.3 vs 4.1 months; HR, 0.63; P < 0.0001; Kang YK, et al. ASCO-GI 2017 [abstract 2]). Prior randomized controlled trials have observed improved survival outcomes for Asian vs Western patients. In the gastric cohort of the phase 1/2 CheckMate 032 study (NCT01928394), nivolumab showed favorable clinical activity in Western patients with advanced gastric/GEJ/esophageal cancer refractory to ≥ 1 prior treatment (Janjigian YY, et al. ASCO 2016 [abstract 4010]). Here we report efficacy of nivolumab in a patient subset from CheckMate 032 with characteristics similar to the patient population in ATTRACTION-2. Methods: Fifty-nine patients enrolled in the CheckMate 032 gastric cohort were treated with nivolumab 3 mg/kg. For this sub-analysis, a subset of patients with gastric/GEJ cancer and ≥2 prior treatment regimens were analyzed (n = 42). Minimum follow-up was 9.5 months (March 2016 database lock). Endpoints included ORR (per investigator [INV; primary] and blinded independent central review [BICR]), BOR, DCR, time to response (TTR), duration of response (DOR), PFS, OS, and safety. Results: In this 42-patient subset, the median age was 58.5 years, 62% of patients (n = 26/42) had GEJ cancer, and 38% (n = 16/42) had gastric cancer. Ninety-three percent of patients (n = 39/42) had prior systemic treatment in the metastatic setting, most commonly fluoropyrimidine, platinum, or taxane therapies; 43% (n = 18/42) had 2 prior regimens, and 57% (n = 24/42) had ≥3 prior regimens. The ORR per INV was 16.7% (n = 7/42; CR, n = 2; PR, n = 5). Per BICR, ORR was 7.1% (n = 3/42; all PRs). Stable disease per BICR was reported for 31.0%; DCR was 38.1% (n = 16/42). In the 3 patients with a response per BICR, median TTR was 1.38 months (range, 1.2-1.4 months) and median DOR was not reached. DOR was ≥6 months in 2 of 3 patients. BOR, TTR, and DOR per BICR and INV are provided in the Table. Median PFS was 1.49 months (95% CI, 1.31-2.76) per BICR and 1.38 months (95% CI, 1.25-2.27) per INV. PFS rates per BICR at 3 and 6 months were 30.7% (95% CI, 16.8%-45.6%) and 13.9% (95% CI, 5.1%-27.1%), respectively. Updated median OS (≈7.5 months of additional follow-up [November 2016 database lock]) was 8.48 months (95% CI, 3.35-15.01); OS rates at 6 and 12 months were 58.1% (95% CI, 41.4%-71.6%) and 44.3% (95% CI, 28.4%-59.0%), respectively. Safety was determined in all 59 patients treated with nivolumab 3 mg/kg. Grade 3/4 treatment-related AEs (TRAEs) were reported in 10 patients (16.9%). One grade 3/4 TRAE occurred in ≥ 2 patients (AST increased: n = 3; 5.1%).O-007 Table Open table in a new tab Conclusion: Nivolumab monotherapy demonstrated favorable clinical activity in this subset of Western patients with gastric/GEJ cancer with ≥2 prior systemic treatments, with a 44.3% 12-month OS rate and durable responses in some patients. Clinical activity was consistent with that observed in the phase 3 ATTRACTION-2 study, suggesting that nivolumab may be similarly effective in Asian and Western patients with heavily pretreated advanced gastric/GEJ cancer.