Drug-induced agranulocytosis: a disproportionality analysis and umbrella review.

Abstract

Background: Drug-induced agranulocytosis (DIA) is a rare but life-threatening hematologic disorder that demands increased clinical and research attention. This study aimed to provide the current overview of DIA for clinical guidance.

Methods: Using real-world data from FDA Adverse Event Reporting System (FAERS), we performed a disproportionality analysis to identify the drugs associated with agranulocytosis, employing the information component and reporting odds ratio algorithms. Logistic analysis was conducted to explore the confounding factors of DIA. Time-to-onset analysis was implemented to compare the adverse event onset time among different drugs. To comprehensively supplement and corroborate our disproportionality findings, we further conducted an umbrella review of systematic reviews (SRs). Five electronic databases were searched with SRs addressing DIA as an included adverse event. Two independent reviewers performed literature screening, data extraction, and quality assessment according to the preferred reporting items for systematic reviews and meta-analysis statement. The results of the included SRs were synthesized using qualitative analysis.

Results: The disproportionality analysis revealed that most identified DIA signals were for anticancer drugs. The top-five drugs with DIA signals by case number were methotrexate (6,462 cases), lenalidomide (5,722 cases), rituximab (5,691 cases), doxorubicin (4,391 cases), and carboplatin (4,371 cases). High-risk drugs (e.g., deferiprone), old age, and abnormal weight were strongly associated with DIA based on multivariate logistic regression. Time-to-onset analysis showed that clozapine has the longest median of onset time (1,121.3 days), while azithromycin has the shortest time (8.1 days). The umbrella review included seven systematic reviews, with five focusing on anticancer therapy. Their findings on DIA-associated drugs, including protein kinase inhibitors and immune checkpoint inhibitors, were consistent with those from the disproportionality analysis. Antibiotics, antithyroid drugs, and psychotropic drugs were also identified as causative drugs of DIA.

Conclusion: This study systematically reviewed the FAERS database and existing literature on DIA to identify a spectrum of associated drugs. Anticancer drugs were predominant, with targeted therapies comprising a large proportion, while non-chemotherapy drugs were also identified as suspect drugs. These findings underscored the need for heightened clinical vigilance toward suspected drugs and highlighted the importance of future efforts to validate high-risk mechanisms and explore DIA monitoring strategies.