Comparative Safety of Long-Acting Versus Short-Acting Erythropoiesis-Stimulating Agents (ESA): Disproportionality Analysis Using the FDA Adverse Event Reporting System

Objectives: Erythropoiesis-stimulating agents (ESAs), classified as long-acting or short-acting, is a primary treatment for anemia in chronic kidney disease and chemotherapy. Given ongoing safety concerns regarding cardiovascular events and potential immunogenic risks associated with PEGylation, this study aims to detect safety signals for long-acting versus short-acting ESAs. Methods: We conducted disproportionality analyses using the FDA Adverse Event Reporting System (FAERS) database. All adverse event (AE) reports related to long-acting and short-acting ESA use from Q1 2018 to Q4 2022 were analyzed, including darbepoetin alfa and methoxy polyethylene glycol-epoetin beta as long-acting ESAs and epoetin alfa and epoetin beta as short-acting ESAs. Data were divided into 2 data sets: ESA as the primary suspect (PS) and ESA as either the PS or secondary suspect (SS). Proportional reporting ratios, reporting odds ratios, and 95% CIs for information components were calculated to identify safety signals. Sensitivity analyses included AE reports with unspecified ESA types to confirm signal consistency. Results: A total of 5702 and 9772 reports for long-acting ESAs, 2487 and 3975 for short-acting ESAs, and 2 and 352 for ESAs with unspecified drug names that could not be classified as either long-acting or short-acting were identified in the PS and PS/SS data sets. In the PS-restricted primary analysis, 10 robust safety signals were identified for long-acting ESAs: back pain, chest pain, drug hypersensitivity, dyspnea, falls, hypotension, pruritus, urinary tract infection, urticaria, and vomiting. Notably, signals such as falls and hypotension were uniquely detected when isolating the PS role and failed to reach the signal threshold in the broader secondary analysis. Conclusions: Our analysis found disproportionately reported AEs for long-acting ESAs, particularly when ESAs were the PS, highlighting the importance of monitoring AEs not listed in the package label, such as falls. Furthermore, despite existing black box warnings regarding cardiovascular risks, our analysis found weak evidence for cardiovascular safety signals strictly attributable to long-acting ESAs. Further studies are needed to confirm the association between these signals and long-acting ESAs.