This cross-sectional analysis sought to determine whether immune checkpoint inhibitors (ICIs) modify associations between other drugs and adverse events, and whether these modifications differ by phenotype and over time. The authors aimed to define the organization of ICI-associated toxicity across organ systems and to quantify whether drug-associated adverse events and documented onset times vary with ICI exposure and by checkpoint pathway.
ICIs are established causes of diverse immune toxicities. The central question addressed here is whether checkpoint blockade acts more broadly as a modifier of drug-associated toxicity beyond events attributed directly to the ICI itself.
The analysis used deduplicated reports from the FDA Adverse Event Reporting System (FAERS) covering 2016 through 2025; analyses were performed in 2026. From 13,701,106 deduplicated reports, 2,365,269 reports were cancer associated and 256,940 reports contained an ICI in any reported drug role.
Demographics reported for cancer-associated reports with observed age included a median age of 66 years (IQR, 56–75 years). Sex distribution among cancer reports was 1,031,999 (43.6%) female and 1,003,154 (42.4%) male. The investigators evaluated ICI exposure overall, exposure to individual primary-suspect drugs, and exposure by checkpoint pathway.
Primary outcomes and analytic approaches included calculation of reporting odds ratios (ORs), identification of cross-organ adverse-event communities, adjusted primary-suspect drug × ICI interaction ORs for selected phenotypes (Stevens–Johnson syndrome/toxic epidermal necrolysis [SJS/TEN], drug reaction with eosinophilia and systemic symptoms [DRESS], acute generalized exanthematous pustulosis [AGEP], interstitial nephritis, drug-induced liver injury [DILI], and vomiting [VOM]), and accelerated failure-time model time ratios for documented onset.
Out of 3,001 eligible Preferred Terms in cancer-associated reports, 2,091 differed with a false discovery rate (FDR) less than .05, indicating widespread differences in reported adverse-event patterns associated with ICI exposure. Analyses identified four cross-organ toxicity communities, reflecting structured groupings of adverse events across organ systems in the context of checkpoint blockade.
These findings portray a reproducible, organized landscape of ICI-associated toxicity signals within the FAERS cancer-report subset, rather than isolated or random associations.
The authors evaluated 138 eligible drug–phenotype pairs for interaction effects between primary-suspect drugs and ICI exposure. Of these, 65 pairs had FDR-significant interactions, demonstrating phenotype-specific modification of drug-associated adverse events by ICI exposure.
Examples highlighted in the source text include:
Amplification of SJS/TEN risk with moxifloxacin when reported with ICI exposure (interaction OR, 101.72; 95% CI, 39.11–264.55).
Attenuation of SJS/TEN reporting with enfortumab vedotin in the context of ICI exposure (interaction OR, 0.17; 95% CI, 0.13–0.23).
Amplification of interstitial nephritis reporting with omeprazole when ICIs were co-reported (interaction OR, 10.35; 95% CI, 7.62–14.05).
These drug–phenotype interaction estimates indicate that checkpoint blockade can both amplify and attenuate drug-associated adverse-event signals, and that effects are phenotype specific.
Temporal analyses used 60,324 reports with documented onset data. ICI exposure was associated with longer adjusted documented time to onset for five of six phenotypes examined, with time ratios ranging from 1.37 to 1.59, indicating delayed reported onset in those contexts. AGEP was the exception, with a time ratio of 0.99 (95% CI, 0.67–1.46), showing no evidence of delay in documented onset with ICI exposure.
The analyses also found that temporal associations differed across checkpoint pathways, suggesting that specific checkpoint blockade mechanisms may influence not only the presence but also the timing of drug-associated adverse events.
The study concludes that ICIs are associated with a structured cross-organ toxicity landscape, phenotype-specific modification of drug-associated adverse events, and distinct temporal patterns across checkpoint pathways. These results support the concept that checkpoint blockade acts as a modifier of drug-associated toxicity and underscore the need for longitudinal and mechanistic validation.
Data produced in the study are available upon reasonable request to the authors, and funding sources and competing interests were disclosed in the source. The provided text indicates that the analyses relied on FAERS, an openly available postmarketing safety database.
Notes on limitations and next steps: The authors recommend longitudinal and mechanistic validation; specific additional limitations inherent to FAERS-based cross-sectional analyses were not detailed in the supplied text. Future work should prospectively validate these interaction and temporal findings and investigate underlying biological mechanisms.