FAERS Study Links EV-Pembro to Broader Toxicity Signals

Key Takeaways

• In a FAERS analysis focused on the enfortumab vedotin with or without pembrolizumab safety question relevant to locally advanced or metastatic urothelial carcinoma, the combination was associated with a broader adverse-event reporting spectrum than enfortumab vedotin alone.
• Excess reporting with the combination clustered mainly in endocrine, immune-mediated, hepatic, gastrointestinal, and pulmonary toxicities rather than appearing evenly across organ systems.
• At the preferred-term level, hepatitis, immune-mediated enterocolitis, adrenal insufficiency, interstitial lung disease, and pneumonitis were among the signals that stood out with enfortumab vedotin plus pembrolizumab.
• A supportive validation analysis within the study also described higher pneumonitis rates with enfortumab vedotin plus pembrolizumab than with enfortumab vedotin monotherapy.
• Overall adverse-event onset appeared similar between regimens, while some organ-system events arose earlier with the combination and possible hearing and dental or gingival findings did not remain robust after correction.

In advanced urothelial carcinoma, combining an antibody-drug conjugate with checkpoint inhibition can produce overlapping toxicities that span several organs and may emerge early in treatment. Clinicians therefore need to know whether adding pembrolizumab changes the postmarketing safety landscape of enfortumab vedotin beyond what is reported with the antibody-drug conjugate alone. Investigators addressed that question by comparing spontaneous adverse-event reports in the Food and Drug Administration Adverse Event Reporting System (FAERS).

In the FAERS comparison of enfortumab vedotin with or without pembrolizumab, investigators used a retrospective case/non-case design covering reports from 2020 Q4 through 2025 Q3. They imported 8,649,232 reports, removed 1,274,325 duplicates, and identified 3,004 reports for enfortumab vedotin alone and 2,265 for enfortumab vedotin plus pembrolizumab. Combination exposure required both drugs to be listed as suspect drugs in the same report. Seven FAERS files were linked, reactions were mapped across Preferred Term, High-Level Group Term, and System Organ Class levels in the Medical Dictionary for Regulatory Activities (MedDRA), and analyses incorporated logistic regression, age and sex adjustment, Benjamini-Hochberg false discovery rate correction, a primary-suspect restriction, shrinkage sensitivity testing, and Kaplan-Meier onset analysis.

Compared with enfortumab vedotin alone, the combination showed higher reporting odds across nine system organ classes, with endocrine disorders as the clearest system-organ-class signal at OR 5.88 (95% CI 3.64-9.48). Immune system and respiratory disorders were the other leading categories, and the broader toxicity pattern centered on endocrine, immune-mediated, hepatic, gastrointestinal, and pulmonary events. At the preferred-term level, hepatitis, immune-mediated enterocolitis, adrenal insufficiency, interstitial lung disease, and pneumonitis stood out. Adjusted and sensitivity analyses were directionally consistent with the primary findings.

In the authors’ supportive pooled literature review of five external studies (228 enfortumab vedotin monotherapy patients and 561 enfortumab vedotin plus pembrolizumab patients), any-grade treatment-related pneumonitis was 10% with the combination versus 3.5% with monotherapy, and grade 3 or higher pneumonitis was 3.7% versus 0.4%. Overall median adverse-event onset was 14.5 days in both groups. Several organ-system events nonetheless appeared earlier with the combination, while metabolism and nutrition disorders trended later.

These findings are non-causal because FAERS disproportionality reflects comparative reporting patterns rather than incidence, absolute risk, or proof of causality. The database lacks denominators and key clinical variables, may be shaped by under-reporting, duplicate or stimulated reporting, and regional reporting differences, and could not distinguish concomitant from sequential enfortumab vedotin plus pembrolizumab use. The timing analyses were exploratory because more than half of reports were excluded for missing or illogical dates, and possible hearing and dental or gingival signals did not remain statistically robust after correction.

According to the authors, enfortumab vedotin plus pembrolizumab carried a broader and more complex reported toxicity pattern than enfortumab vedotin alone, concentrated in endocrine, immune, hepatic, gastrointestinal, and pulmonary events.

Clinician Questions

Could the FAERS EV+P reports distinguish concomitant from sequential use of enfortumab vedotin and pembrolizumab?

Combination exposure in FAERS was defined by enfortumab vedotin and pembrolizumab both being recorded as suspect drugs in the same report, but the database could not separate truly concomitant treatment from sequential administration. That limits how precisely the reported signal maps onto real-world regimen timing in locally advanced or metastatic urothelial carcinoma.

Which pulmonary events appeared as positive EV+P signals, and which pulmonary events did not?

Positive pulmonary signals with enfortumab vedotin plus pembrolizumab centered on pneumonitis, interstitial lung disease, and immune-mediated lung disease, whereas pneumonia, aspiration pneumonia, metastases to lung, and nonspecific lung disorder did not show significant positive signals. Within FAERS, that pattern favored inflammatory lung toxicities in reported events rather than serving as proof of mechanism or causality.

Why can’t FAERS odds ratios for EV+P be interpreted as incidence or absolute risk in advanced urothelial carcinoma?

FAERS does not provide a reliable exposed-population denominator and is vulnerable to under-reporting, duplicate or stimulated reporting, regional reporting differences, and residual confounding. For enfortumab vedotin plus pembrolizumab versus enfortumab vedotin alone, the odds ratios therefore describe comparative adverse-event reporting patterns, not true incidence, absolute risk, or proof of a drug-drug interaction.