Pooled safety profiles of bispecific antibodies targeting PD-1/CTLA-4 or PD-1/VEGF in non-small cell lung cancer: a systematic review and single-arm meta-analysis.

Abstract

Objective: This study aimed to systematically synthesize and separately quantify the pooled safety profiles of bispecific antibodies (BsAbs) targeting PD-1/CTLA-4 or PD-1/VEGF in patients with non-small cell lung cancer (NSCLC), and to descriptively summarize class-specific safety patterns.

Methods: Following PRISMA guidelines, we systematically reviewed prospective clinical trials published through January 2026. A total of 17 studies, encompassing 1802 patients, were included. Single-arm proportion meta-analyses were conducted separately for each BsAb class using random-effects models. No formal between-class statistical comparison was performed. Given the clinical heterogeneity across included cohorts, including differences in study phase, treatment line, molecular subtype, and treatment regimen, pooled estimates were interpreted descriptively. Subgroup analyses by treatment regimen, treatment line, and individual BsAb agent were considered exploratory. Because of the limited number of cohorts within each phase stratum, the impact of study phase was assessed descriptively rather than through formal stratified meta-analysis.

Results: In separate pooled analyses, the incidence of grade 3 or higher treatment-related adverse events (TRAEs) was 41.79% for PD-1/CTLA-4 BsAbs and 43.51% for PD-1/VEGF BsAbs. Exploratory subgroup analyses suggested different patterns of heterogeneity across the available cohorts. In PD-1/CTLA-4 BsAb studies, AE rates appeared to vary across individual agents, whereas in PD-1/VEGF BsAb studies, higher AE rates were observed mainly in chemotherapy-containing regimens. These observations should be interpreted cautiously because they may reflect differences in the distribution of included agents, patient populations, study phases, and treatment strategies rather than inherent class-specific toxicity mechanisms. Regarding tolerability, the pooled treatment discontinuation rate due to TRAEs was 10.26% in the PD-1/CTLA-4 analysis and 4.44% in the PD-1/VEGF analysis. The pooled incidence of immune-related adverse events (irAEs) was 46.68% in the PD-1/CTLA-4 analysis and 27.10% in the PD-1/VEGF analysis, whereas the pooled rates of grade 3 or higher irAEs were low in both analyses.

Conclusion: This systematic review provides a descriptive synthesis of available safety data for PD-1/CTLA-4 and PD-1/VEGF BsAbs in NSCLC. Because the included cohorts differed substantially in study phase, treatment line, molecular background, and concomitant therapies, the pooled AE estimates should not be interpreted as direct comparative evidence. Instead, the findings should be viewed as hypothesis-generating signals that require validation in more homogeneous prospective studies or head-to-head trials.