Warfarin provides greater net benefit than rivaroxaban for patients with atrial fibrillation and rheumatic heart disease

Reference

Connolly SJ, Karthikeyan G, Ntsekhe M, et al, for the INVICTUS Investigators. Rivaroxaban in rheumatic heart disease–associated atrial fibrillation. N Engl J Med 2022;387:878-888.

Synopsis

Studies that have compared direct oral anticoagulants with warfarin have excluded patients with valvular atrial fibrillation (AF) like that caused by rheumatic heart disease (RHD), which is still common in low- and middle-income countries. This study, performed in Africa, Asia and Latin America, identified 4565 adults with echocardiographically confirmed RHD and atrial fibrillation or flutter. All patients had a CHA2DS2VASc score of 2 or higher, mitral stenosis, or evidence of a left atrial thrombus. Although the trial was open label, outcomes were assessed by researchers masked to treatment assignment and analysis was by intention to treat. Participants were randomized to receive rivaroxaban 15 mg or 20 mg daily on the basis of renal function, or warfarin titrated to an international normalized ratio (INR) of 2.0 to 3.0 measured at least monthly. Groups were balanced at baseline, with a mean age of 50 years, 72% were women, and mitral stenosis was present in 85% of participants. During the trial, INR was in the therapeutic range for 56% of patients taking a vitamin K antagonist at 6 months, and for 64% to 65% at 2 to 3 years. Approximately 20% of patients discontinued using rivaroxaban during the study, in part because they were undergoing valve replacement. Although patients who discontinued using rivaroxaban were usually put on a vitamin K antagonist, those who discontinued taking a vitamin K antagonist were usually left un-anticoagulated. After a mean duration of just over 3 years, the composite of stroke, embolism, myocardial infarction, or death from vascular causes was less common in the warfarin group (6.3% vs 8.1%; P = .002; number needed to treat = 56). Individual outcomes of stroke (0.9% vs 1.4%) and death from vascular causes (4.7% vs 6.0%) were both significantly less likely with warfarin. Although there was no significant difference in major bleeding events between groups, fatal bleeding was less likely with rivaroxaban (4 vs 15 deaths; hazard ratio 0.29; 95% CI 0.10 - 0.88). For approximately every 4 vascular deaths prevented by using warfarin, there was one additional fatal bleeding event. The benefit of warfarin only began to appear after 2 years, around the time when the percentage time in range increased for participants in the warfarin group.