Risk of serious bleeding with concomitant use of apixaban or rivaroxaban with amiodarone compared to flecainide or sotalol in patients with atrial fibrillation

Abstract BACKGROUND Direct oral anticoagulants (DOACs), including apixaban and rivaroxaban, are widely used for stroke prevention in patients with atrial fibrillation (AF), a population characterized by multimorbidity and polypharmacy. Because apixaban and rivaroxaban are primarily metabolized by cytochrome P450 3A4 (CYP3A4) and P-glycoprotein (P-gp), co-administration with inhibitors of these pathways may increase drug exposure and bleeding risk. Amiodarone, a potent inhibitor of both CYP3A4 and P-gp, is commonly prescribed to maintain sinus rhythm in patients with AF. Preclinical studies have demonstrated that co-administration of amiodarone with CYP3A4 substrates, such as DOACs, increases systemic drug levels, suggesting that combining amiodarone with apixaban or rivaroxaban may elevate the risk of serious bleeding. OBJECTIVE To compare the risk of serious bleeding complications among AF patients who used apixaban or rivaroxaban in combination with amiodarone compared to flecainide or sotalol. METHODS We conducted a retrospective cohort study using de-identified patient-level claims data from the IBM MarketScan Commercial and Medicare Supplemental Databases (January 1, 2015, to December 31, 2023). We emulated a hypothetical target trial comparing bleeding risks in AF patients treated with apixaban or rivaroxaban concurrently with amiodarone (exposed group) versus flecainide or sotalol (referent group). Eligible patients were: a) new users of apixaban or rivaroxaban; b) aged ≥18 years at cohort entry; c) had a prescription fill for amiodarone, flecainide, or sotalol after initiating a DOAC; d) continuously enrolled in medical and pharmacy benefits for ≥12 months before cohort entry (lookback period); and e) had a diagnosis of AF based on ≥1 inpatient or ≥2 outpatient claims within the prior year. The cohort entry date was defined as the first date of the combined therapy. The primary outcome was hospitalization for serious bleeding, identified using principal discharge diagnosis codes based on validated algorithms. Follow-up continued from cohort entry through the first occurrence of: a study outcome, discontinuation of regimen, insurance disenrollment, switching to another regimen, death or end of study. Covariates included sociodemographics, comorbidities, medications and healthcare utilization measures assessed during the lookback period. We used propensity score (PS) matching to balance covariates between exposure groups and estimated marginal hazard ratios (HRs) and 95% confidence intervals (CIs) using Cox proportional hazards models with robust variance estimators. RESULTS We identified 23,968 users of amiodarone and 27,939 users of flecainide or sotalol among 169, 213 patients with AF who had initiated DOACs. Prior to PS matching, baseline characteristics were relatively similar between groups: mean age (57 vs. 56 years), prevalence of liver disease (1% vs. 1%), asthma (6% vs. 6%), chronic kidney disease (1.5% vs. 1%), nonsteroidal anti-inflammatory drug use (16% vs. 15%), and proton pump inhibitors use (18% vs. 18%). After 1:1 PS matching, the final cohort included 16,297 patients in each group. Covariates were well balanced post-matching (standardized differences <0.1). After matching, 253 patients in the amiodarone group experienced serious bleeding (25.8 events per 1,000 person-years) compared to 175 in the comparator group (16.4 events per 1,000 person-years) (HR, 1.53; 95% CI, 1.26 to 1.86). CONCLUSION In this comparative safety study, among patients with AF treated with apixaban or rivaroxaban, concomitant use of amiodarone was associated with a higher risk of serious bleeding compared to use of flecainide or sotalol. These findings highlight the importance of considering drug-drug interactions when managing anticoagulated patients with AF to promote safe prescribing and minimize the risk of serious bleeding complications.