Expanding the Use of Rivaroxaban for Patients With Atrial Fibrillation: The RIVER Trial

Patients with atrial fibrillation (AF) require long-term anticoagulation therapy for stroke prevention. The pivotal randomized trials in AF comparing the direct oral anticoagulants (DOACs) to vitamin K antagonists (VKAs) demonstrated a significant reduction in stroke and systemic embolism, with a reduction in major bleeding with DOACs.¹  However, these trials mostly excluded patients with moderate-to-severe mitral stenosis and prosthetic heart valves. For this reason, current guidelines from the American Heart Association recommend using VKAs over DOACs for patients with moderate-to-severe mitral stenosis or mechanical heart valves.² 

There remains uncertainty about the efficacy and safety of DOACs in patients with AF with bioprosthetic mitral valves. These patients were not included in the ROCKET-AF study (rivaroxaban); however, 162 such patients were included in the Aristotle (apixaban) and ENGAGE AF-TIMI 48 (edoxaban) trials.^3,4  A recent meta-analysis of patients with AF and valvular heart disease aside from significant mitral stenosis and mechanical heart valves demonstrated a reduction in stroke and systemic embolism with DOACs compared with VKA. However, there is a paucity of randomized comparisons of DOACs compared with warfarin in patients with bioprosthetic valves.⁵  The RIVER trial provided this comparison specifically in those with AF and bioprosthetic mitral valves.

This was a randomized, open-label, noninferiority study completed at 49 sites in Brazil. Patients with AF or flutter and bioprosthetic mitral valve were randomly assigned to receive rivaroxaban 20 mg daily (or 15 mg daily if creatinine clearance is 30-49 mL/min) or warfarin (international normalized ratio target, 2.0-3.0). The primary outcome was a composite of major cardiovascular events (including transient ischemic attack [TIA], stroke, valve thrombosis, systemic embolism, or hospitalization for heart failure), death, or major bleeding at 12 months. The study included primarily younger individuals (mean age 59 years) with a mean CHA₂DS₂-VASc score of 2.5. Fourteen percent had prior stroke or TIA, and 35 percent had their mitral valve prosthesis implantation within one year prior to randomization.

The mean time to primary outcome was similar in the rivaroxaban and warfarin groups (347.5 days vs. 340.1 days; 95% CI, –1.4 to 16.3; p<0.001 for noninferiority). Death from cardiovascular causes or thromboembolic events was numerically lower in the rivaroxaban group (3.4%) compared with warfarin (5.1%), and incidence of stroke was 0.6 percent and 2.4 percent, respectively. The incidence of valve thrombosis was similar between the two groups (1.0% vs. 0.6%). Major bleeding occurred less frequently with rivaroxaban (1.4%) compared with warfarin (2.6%). In the subgroup of patients who had undergone randomization within three months of valve implantation (n = 189), the incidence of the primary outcome appeared to be numerically lower with rivaroxaban (6.4%) compared with warfarin (18.9%).