Pharmacovigilance assessment of drug-induced aplastic anemia: analysis of the FDA adverse event reporting system.

Abstract

Introduction: Aplastic anemia is a rare but life-threatening disorder often triggered by drug exposure. Given its low incidence, identifying drug-associated risks requires large-scale real-world data. The FDA Adverse Event Reporting System (FAERS) is a valuable pharmacovigilance resource for detecting rare and serious drug reactions.

Aim: This study aimed to evaluate the association between a broad range of medications and the risk of drug-induced aplastic anemia using FAERS data obtained through disproportionality analysis, regression modeling, time-to-onset analysis, and predictive modeling.

Method: A retrospective pharmacovigilance study was conducted using FAERS reports from January 1, 2004, to December 31, 2024. Aplastic anemia cases were identified using five Preferred Terms from the Medical Dictionary for Regulatory Activities. Signal detection was performed using ROR, PRR, BCPNN, and MGPS. Logistic regression analyses with weighted LASSO variable selection identified the independent risk factors. A predictive model was developed and validated using receiver operating characteristic (ROC) curve analysis. Time-to-onset (TTO) and pharmacological classification were also conducted.

Results: A total of 4493 drug-related aplastic anemia cases were identified. Disproportionality analysis revealed 593 significant drugs, of which 16 met stringent inclusion criteria for multivariate analysis. Temozolomide was most frequent (n = 148), followed by methotrexate (n = 126), busulfan (n = 100), and linezolid (n = 90). Other common drugs included ribavirin, nivolumab, pembrolizumab, fludarabine, carboplatin, etoposide, and cyclophosphamide. Male sex was a significant risk factor (OR = 1.63), while older age (> 42 years) and higher weight (> 60 kg) were protective. The predictive model showed good discrimination (AUC = 0.777). Median TTO was 299 days, with most cases occurring within six months. The 16 implicated drugs fell into categories including antineoplastics, antibacterials, antivirals, antiepileptics, and antigout agents.

Conclusion: This study identified key drugs and patient factors associated with aplastic anemia, providing a data-driven framework for pharmacovigilance. These findings support early detection and informed clinical and regulatory decision making, but prospective studies are required to confirm causality and refine individualized risk predictions.