A retrospective study comparing the severe cutaneous adverse reactions in patients receiving triazole antifungal therapy and literature analysis.

Triazole antifungal drugs (TADs) are widely used in clinical practice but carry a risk of severe cutaneous adverse reactions (SCARs). However, comprehensive real-world evidence regarding the comparative safety profiles of individual TADs remains limited. This research aimed to investigate and compare the real-world risk, clinical characteristics, and time-to-onset of SCARs associated with individual TADs. The SCAR cases with TADs as the primary suspect drug were identified in the FDA Adverse Event Reporting System. Disproportionality analysis and Bayesian methods were employed to detect safety signals with Reporting Odds Ratio (ROR), Proportional Reporting Ratio (PRR), Bayesian Confidence Propagation Neural Network (BCPNN), and Empirical Bayes Geometric Mean (EBGM). Differences of clinical features, outcomes, time-to-onset (TTO), and signal strength at the preferred term (PT) level were analyzed. A total of 1429 SCAR cases with five TADs were identified. Fluconazole showed the strongest and most consistent statistical association across all four algorithms (ROR = 5.2), followed by itraconazole. Voriconazole yielded mixed signals (positive by ROR and BCPNN only), while Posaconazole and Isavuconazole exhibited no significant SCAR signals. Voriconazole was associated with the highest mortality proportion (12.8%), followed by posaconazole (10%). The median TTO for SCARs was 8 days, with voriconazole having the longest delay (16 days). At the PT level, 45 distinct signals were identified, including high-risk manifestations such as fixed eruption (ROR = 24.43), mucosal ulceration, and SJS/TEN overlap. This study identified significant variability of SCARs signals among triazole antifungals. These findings support early risk recognition and informed antifungal selection in clinical practice.