Survival benefit and toxicity trade-offs of first-line chemoimmunotherapy in advanced gastric and gastroesophageal junction cancer: a meta-analysis of randomized trials.

BACKGROUND: First-line chemoimmunotherapy has become a major therapeutic strategy for advanced gastric and gastroesophageal junction cancer, yet the magnitude of survival benefit, regional consistency, and toxicity trade-offs remain important considerations. We conducted an systematic review and meta-analysis of randomized controlled trials evaluating PD-1/PD-L1 inhibitor plus chemotherapy versus chemotherapy alone or placebo plus chemotherapy. METHODS: PubMed/MEDLINE, Embase, Cochrane CENTRAL, Web of Science, and ClinicalTrials.gov were searched from inception to March 1, 2026. Eligible studies enrolled adults with previously untreated unresectable, recurrent, locally advanced, or metastatic gastric or gastroesophageal junction cancer. Hazard ratios were pooled for overall survival and progression-free survival, and risk ratios were pooled for objective response rate and safety outcomes using random-effects models. Risk of bias was assessed using RoB 2, and certainty of evidence was evaluated using GRADE. RESULTS: Seven randomized controlled trials including 6,517 patients were analyzed. PD-1/PD-L1 inhibitor plus chemotherapy significantly improved overall survival and progression-free survival. Objective response rate was also increased (RR, 1.24; 95% CI, 1.18-1.31), with negligible observed heterogeneity (I² = 0.0%). Overall survival effects were similar in global trials (HR, 0.79; 95% CI, 0.75-0.85) and Asian-only or China-only trials (HR, 0.81; 95% CI, 0.73-0.91; P for subgroup difference = 0.70). The progression-free survival effect was greater in Asian-only or China-only trials (HR, 0.66 vs. 0.78; P for subgroup difference = 0.02), although this analysis was exploratory. Combination therapy increased grade ≥3 treatment-related adverse events (RR, 1.15; 95% CI, 1.08-1.23), serious treatment-related adverse events (RR, 1.53; 95% CI, 1.29-1.83), and treatment discontinuation (RR, 1.53; 95% CI, 1.37-1.72). The pooled estimate for treatment-related death was statistically inconclusive (RR, 1.54; 95% CI, 0.75-3.16); the wide confidence interval indicated substantial imprecision and could not exclude clinically important increases or decreases in treatment-related mortality. CONCLUSIONS: First-line PD-1/PD-L1 inhibitor + chemotherapy provides a consistent survival and response benefit in advanced gastroesophageal junction or gastric cancer, but with increased clinically relevant toxicity. These findings support chemoimmunotherapy as a preferred first-line strategy for appropriately selected patients, with treatment decisions guided by biomarker status, patient fitness, and toxicity risk.