Immunotherapy plus chemotherapy versus immunotherapy alone in esophageal cancer: a real-world pharmacovigilance analysis of safety signals and fatal outcomes

Background Immune checkpoint inhibitors (ICIs) have become an important component of systemic therapy for esophageal cancer, but real-world safety data for combining ICIs with taxane- and platinum-based chemotherapy remain limited. This study compared adverse-event (AE) profiles, time to onset (TTO), and reported mortality between ICI-based combination therapy and ICI monotherapy in esophageal cancer. Methods Using the FDA adverse event reporting system (FAERS), we screened reports from Q3/2019 through Q1/2026 for esophageal cancer involving pembrolizumab or nivolumab as the primary suspect. Combination included immunotherapy with a taxane and platinum; monotherapy included immunotherapy alone. Disproportionality was assessed via reporting odds ratios ( RORs ) and proportional reporting ratios ( PRRs ). TTO was evaluated using Kaplan-Meier and Cox models , and logistic regression examined factors for reported mortality. Results Among 1,580 eligible reports (103 combination, 1,477 monotherapy), distinct AE signal profiles emerged. In combination, prominent signals were gastrointestinal, hematologic, and respiratory, with esophagitis showing the strongest signal ( ROR , 61.97; 95%CI , 22.80–168.48), followed by decreased lymphocyte count ( ROR , 59.68; 95%CI , 28.99–122.83) and behaviour disorder ( ROR , 42.85; 95%CI , 13.58–135.17). In monotherapy, signals were more immune-mediated, including immune-mediated cystitis ( ROR , 261.05; 95%CI , 113.92–598.18), immune-mediated cholangitis ( ROR , 87.65; 95%CI , 41.36–185.78), immune-mediated endocrinopathy ( ROR , 85.51; 95%CI , 27.19–268.89) and so on. Among 571 reports with TTO data, AEs occurred significantly earlier in combination than monotherapy, with median TTOs of 23 days ( IQR , 8-72) vs. 49 days ( IQR , 21–129) ( log-rank P < 0.001), consistent across age, sex, weight, and reporting-year subgroups. Reported mortality was 6.8% for combination vs. 21.1% for monotherapy. After multivariable adjustment, combination was associated with lower reported mortality odds than monotherapy ( adjusted OR , 0.27; 95%CI , 0.06–0.77; P = 0.033). Conclusion ICI-based combination therapy showed a distinct AE profile and significantly earlier AE onset than monotherapy, but lower reported mortality odds. These findings underscore the need for early monitoring of hematologic, gastrointestinal, and respiratory toxicities during combination therapy.