Safety Comparison of Acetylcholinesterase Inhibitors Using FAERS Spontaneous-Reporting Data: Drug-Specific Reporting Signals and Time-to-Onset Dynamics.

Abstract

Introduction/objective: Acetylcholinesterase Inhibitors (AChEIs) are widely used for symptomatic treatment of Alzheimer's disease, but their post-marketing reporting profiles may differ. This study compared FAERS disproportional reporting signals and Time-To-Onset (TTO) patterns for donepezil, galantamine, and rivastigmine.

Methods: FAERS reports from Q1 2004 to Q2 2025 were analyzed. Reports listing donepezil, galantamine, or rivastigmine as the primary suspect drug were retained. Unique case-drug-Preferred Term (PT) triplets were constructed after case-level and drug-event-level de-duplication. Positive signals were defined conservatively by the intersection of ROR, PRR, BCPNN, and MGPS/EBGM criteria. TTO was evaluated using cumulative onset-distribution curves, Weibull models, and Accelerated Failure-Time (AFT) models.

Results: The analysis included 9,474 donepezil reports, 3,042 galantamine reports, and 14,666 rivastigmine reports. The four-method intersection identified 286 positive PT reporting signals for donepezil, 143 for galantamine, and 259 for rivastigmine. Donepezil showed prominent cardiac reporting signals, including sinus bradycardia (ROR = 49.0) and electrocardiogram QT prolongation (ROR = 16.75). Rivastigmine showed application-site signals, including application-site erythema (ROR = 21.36) and application-site pruritus (ROR = 18.07). Galantamine showed signals including bradycardia (ROR = 15.37), fall (ROR = 3.44), and decreased appetite (ROR = 3.57). Median TTOs were 38, 44, and 64 days for donepezil, galantamine, and rivastigmine, respectively. Rivastigmine showed a later reported onset than donepezil in the AFT model (TR = 1.08; p = 0.008).

Discussion: The findings suggest shared cholinergic reporting patterns and drug-specific signal profiles. Because FAERS lacks exposure denominators, these results indicate disproportional reporting rather than incidence, causality, or comparative clinical risk.

Conclusion: Donepezil, galantamine, and rivastigmine showed distinct FAERS reporting profiles and early post-initiation TTO patterns. These findings are hypothesis-generating and require confirmation in independent pharmacovigilance or clinical datasets.