A Real-World Pharmacovigilance Analysis of the Safety Profiles Associated with Anti-MRSA Agents Using the Japanese Adverse Drug Event Report (JADER) Database

Background: Anti-MRSA agents are essential for treating severe infections, yet their use is constrained by distinct toxicity profiles. However, comparative real-world data remain scarce. Methods: This nationwide pharmacovigilance study used the Japanese Adverse Drug Event Report (JADER) database (2004–2025). Disproportionality analyses (proportional reporting ratio PRR) were performed at the Standardized MedDRA Query and Preferred Term levels, complemented by Weibull-based time-to-onset modeling, to characterize AE patterns associated with vancomycin (VCM), teicoplanin (TEIC), arbekacin (ABK), daptomycin (DAP), linezolid (LZD), and tedizolid (TZD). Results: Distinct agent-specific AE profiles were observed. VCM showed disproportionate reporting of acute renal failure (PRR 6.66) and severe cutaneous reactions. TEIC displayed fewer renal signals but relatively higher reporting of hematologic events (PRR 3.51). ABK demonstrated high disproportionality in acute and chronic renal failure, reflecting aminoglycoside nephrotoxicity. DAP showed a high reporting signal for eosinophilic pneumonia (PRR 23.30), interstitial lung disease, and creatine kinase elevation/rhabdomyolysis, with wear-out hazard patterns suggesting a possible time-dependent reporting tendency. LZD exhibited hematopoietic signals (PRR 6.13) and additional associations with hyponatremia, lactic acidosis, and optic neuropathy, consistent with marrow suppression and mitochondrial toxicity. Weibull analysis indicated cumulative “wear-out” risks for renal, hepatic, and hematologic events, whereas hypersensitivity and many pulmonary events followed random-failure patterns. Conclusions: This large-scale JADER analysis delineated the distinct safety profiles of the six anti-MRSA agents. The key findings included DAP pulmonary and muscle toxicities, LZD hematological events, and VCM nephrotoxicity. Time-to-onset modeling indicates potential cumulative versus random risk patterns, suggesting the need for individualized monitoring and cross-validation.