Effectiveness and safety of rivaroxaban vs. warfarin in patients with non-valvular atrial fibrillation and coronary or peripheral artery disease

Aims There are scarce data evaluating the effectiveness and safety of rivaroxaban vs. warfarin in non-valvular atrial fibrillation (NVAF) patients with concomitant coronary artery disease (CAD) and/or peripheral artery disease (PAD) treated in routine practice. Methods and results Using MarketScan data from January 2012 to December 2017, we identified oral anticoagulant (OAC)-naive NVAF patients receiving rivaroxaban (15-20 mg once daily) or warfarin, with comorbid CAD and/or PAD and >= 12 months of insurance coverage before OAC initiation. Differences in baseline covariates between cohorts were adjusted using inverse probability-of-treatment weights based on propensity scores (standardized differences <0.1 achieved for all covariates after adjustment). Endpoints included a composite of major thrombotic vascular events (MTVEs) (including ischaemic stroke, myocardial infarction, or need for lower limb revascularization/major amputation) and major bleeding. Patients were followed until an event-of-interest, discontinuation/switch of index OAC, insurance disenrolment, or end-of-data availability. Hazard ratios (HRs) and 95% confidence intervals (CIs) were calculated using Cox regression. We identified 3257 rivaroxaban (30.4% received a 15 mg dose) and 5046 warfarin users with NVAF and comorbid CAD and/or PAD. Rivaroxaban was associated with a 32% (95% CI = 8-50%) reduction in the composite of MTVE. No significant difference in major bleeding was observed (HR = 1.13, 95% CI = 0.84-1.52). No statistical interactions were noted in subgroup analyses performed on the MTVE (P-interaction >= 0.35 for all) or major bleeding endpoints (P-interaction >= 0.09 for all). Conclusion Among patients with NVAF and comorbid CAD and/or PAD, rivaroxaban use was associated with a reduced risk of MTVEs vs. warfarin, without significantly increasing major bleeding risk.