Real-world (RW) effectiveness of nivolumab plus chemotherapy (NIVO+chemo) in patients (pts) with advanced or metastatic gastric carcinoma, gastroesophageal junction carcinoma, or esophageal adenocarcinoma (GC/GEJC/EAC).

Abstract

295 Background: In the CheckMate 649 trial, NIVO+chemo demonstrated superior efficacy vs chemo alone in pts with previously untreated, advanced, non-human epidermal growth factor receptor 2 positive GC/GEJC/EAC. Our study was the first to evaluate the RW adoption patterns and effectiveness of NIVO+chemo as a first-line (1L) treatment of advanced GC/GEJC/EAC in the immediate period after NIVO approval in this setting. We report the 18-mo follow-up results. Methods: Pts in the United States ≥ 18 years of age with previously untreated advanced or metastatic GC/GEJC/EAC receiving NIVO+chemo or chemo alone (FOLFOX or CAPEOX) between 04/01/2021 and 12/31/2022 were identified in the Flatiron Health database. Pts were followed from the start of 1L therapy to death, loss of data availability, or the end of data abstraction, whichever occurred first. Results: 528 pts were eligible for the analysis (231 [44%] NIVO+chemo, 297 [56%] chemo). The prevalence of PD-L1 combined positive score (CPS) ≥ 5 before the 1L therapy was 55% for NIVO+chemo and 27% for chemo. Baseline demographic and clinical characteristics were comparable between treatment groups in the overall cohort and the CPS subgroups. Median follow-up (interquartile range) was 6.0 (2.4−9.8) mo for NIVO+chemo and 6.1 (2.5−9.7) mo for chemo. There were 87 (37.7%) deaths in the NIVO+chemo cohort and 138 (46.5%) in the chemo cohort. NIVO+chemo demonstrated superior overall survival (OS) in the overall cohort, with a 12-mo OS rate of 53.3% vs 38.1% for chemo and a 2.6-mo improvement in median OS (12.6 [95% confidence interval (CI), 10.2−17.9] mo vs 10.0 [8.9−11.5] mo, P = 0.047). Univariate Cox regression analysis of OS yielded a hazard ratio (HR) of 0.76 (95% CI, 0.58−1.00; P = 0.048) and multivariate Cox regression yielded an HR of 0.74 (95% CI, 0.56−0.98; P = 0.033) for NIVO+chemo vs chemo. There was also a consistent trend toward OS benefit with NIVO+chemo vs chemo in all CPS subgroups (Table). Multivariate analysis in the CPS ≥ 1 subgroup yielded a HR of 0.51 (95% CI, 0.30−0.89) for NIVO+chemo vs chemo but was not feasible in other subgroups due to sample size. Conclusions: NIVO+chemo showed superior OS vs chemo alone in previously untreated pts with advanced GC/GEJC/EAC in the RW setting. A trend toward OS benefit was also observed in all CPS subgroups, which will be further examined in future analyses with larger samples and longer follow-up. Clinical trial information: NCT02872116 . [Table: see text]