Background/Objectives: The purpose of this study was to investigate the association between cardiovascular adverse drug events (ADEs) and the use of COVID-19 medicines. Methods: The analyses were conducted by leveraging pharmacovigilance data from the Food and Drug Authority (FDA) Adverse Event Reporting System (FAERS) and TriNetX electronic health records (EHRs). Transcriptomic data from human embryonic stem cell-derived cardiomyocytes (hESC-CMs) exposed to remdesivir were analyzed to provide supportive biological context for the observed cardiovascular safety signals. Results: Comparative analysis of three approved COVID-19 therapies revealed that COVID-19 patients treated with remdesivir had a higher risk of cardiovascular events than those treated with Paxlovid or REGEN-COV. FAERS analysis further indicated that bradycardia, hypotension, and cardiac arrest were the most frequently reported cardiovascular events associated with remdesivir, which was validated by propensity score-matched EHR data. These findings suggest an association between remdesivir exposure and increased cardiovascular ADEs relative to other COVID-19 therapies. Sex-stratified analysis using FAERS and EHR did not show strong sex-dependent patterns for remdesivir-associated cardiovascular ADEs. Age-stratified analyses of EHR data showed age-associated variation across the three cardiovascular ADEs. Bradycardia displayed a non-uniform pattern with higher prevalence in the youngest and oldest age groups, hypotension showed an overall age-associated increase, and cardiac arrest showed only a weak age-associated effect. Pathway enrichment analysis on transcriptomic data revealed that the “cGMP-PKG signaling pathway”, “dilated cardiomyopathy”, and “calcium signaling pathway” were enriched among genes up-regulated by remdesivir exposure. Conclusions: In summary, our integrated analysis of pharmacovigilance, EHR, and transcriptomic data provides convergent evidence for associations between remdesivir and cardiovascular ADEs and offers biological context into these associations.
Zhu et al. (Thu,) studied this question.