Authors: Hart RG, Sharma M, Mundl et al. Rivaroxaban for stroke prevention after embolic stroke of undetermined source. N Engl J Med
Reviewed by: Kyle Hobbs, MD, Assistant Professor, Wake Forest School of Medicine
Read the article*
*You will need to log in or register for the The New England Journal of Medicine in order to read this article.
Embolic stroke of undetermined source (ESUS) is a group of cryptogenic ischemic strokes not associated with proximal arterial stenosis or a cardioembolic source, and that are not lacunar. This industry-funded, international, randomized phase 3 trial investigated whether anticoagulation with a direct factor Xa inhibitor (rivaroxaban) would be more effective than aspirin to prevent recurrent ESUS. Patients were stratified according to country and age (<60 versus >60 years) and randomized 1:1 to receive either rivaroxaban 15mg plus placebo or aspirin 100mg plus placebo daily. Patients were included if they had ischemic stroke between seven days and six months; qualifying strokes were not lacunar, could not be associated with extracranial vessel atherosclerosis causing >50 percent stenosis in arteries supplying the area of ischemia, and did not occur in patients with a cardiac source of embolism, with no other cause of stroke found. Patients were also excluded if they had intracranial vessel imaging finding >50 percent stenosis due to atherosclerosis, history of atrial fibrillation, severely disabling stroke (mRS > 4 at screening), specific indication for anticoagulant or antiplatelets, regular NSAID use, major bleeding within six months, or prior intracerebral hemorrhage. Patients had to be >49 years old, and if they were 50-59 years old, had to have an additional vascular risk factor. Cardiac workup included at least 20 hours of cardiac rhythm monitoring for atrial fibrillation of >six minutes and echocardiography. Primary efficacy outcome was the first recurrent stroke (ischemic, hemorrhagic, or undefined) or systemic embolism in a time-to-event analysis. Secondary efficacy outcomes were a composite of death from cardiovascular causes, recurrent stroke, systemic embolism, or myocardial infarction; death from any cause; disabling stroke (mRS 4-5 at discharge); or fatal stroke. The primary safety outcome was major bleeding at any site in the body.
The trial was planned to enroll 7,000 patients followed for a mean of two years, but was terminated at the second interim analysis due to an excess risk of bleeding among patients in the rivaroxaban group without an offsetting benefit of reduction in the rate of stroke. 7,213 patients were randomly assigned to a treatment group. Participants had been followed for a median of 11 months (IQR 5-17) at the time of trial termination. Mean patient age was 67 years (62 percent men). Median time from qualifying stroke to randomization was 37 days (IQR14-88). Median NIHSS after initial stroke was one. The trial drug was discontinued early in 15 percent of rivaroxaban and 12 percent of aspirin patients. The primary efficacy outcome occurred in 172 patients in the rivaroxaban group (annualized rate 5.1 percent) and in 160 patients in the aspirin group (annualized rate, 4.8 percent) (hazard ratio, 1.07; 95 percent CI, 0.87-1.33; p=0.52); 95 percent of these events were ischemic strokes (158 rivaroxaban and 156 aspirin), 5 percent were hemorrhagic strokes, and 1 percent systemic embolism. There were 13 hemorrhagic strokes in the rivaroxaban group versus two in the aspirin group (hazard ratio, 6.50; 95 percent CI, 1.47-28.8). There was no difference in effect of rivaroxaban vs aspirin regarding the other secondary efficacy outcomes. Major bleeding occurred in 62 patients in the rivaroxaban group and 23 patients in the aspirin group (hazard ratio, 2.72; 95 percent CI, 1.68-4.39; p<0.001), and the rate of life-threatening or fatal bleeding, symptomatic ICH, and nonmajor bleeding were also significantly higher in the rivaroxaban group.
For embolic strokes of undetermined source, rivaroxaban did not decrease the rate of stroke recurrence compared to aspirin, and did lead to more clinically significant bleeding. Of note, the rate of intracerebral hemorrhage in patients taking aspirin was lower than in previous trials. One possible reason for the lack of benefit seen with rivaroxaban is that the cryptogenic strokes in this trial were likely of heterogeneous etiology, with a number of them not responding to rivaroxaban for secondary prevention. Limitations of the trial include the use of a 15mg dose of rivaroxaban (the usual dose in most countries for stroke prevention is 20mg), although the authors felt there was enough therapeutic overlap for the 15mg dose to be sufficient; in addition, the higher 20mg dose could theoretically have led to more bleeding. In summary, anticoagulation with rivaroxaban for embolic-appearing cryptogenic strokes confers no benefit above aspirin.
#NEWSReview #KyleHobbs #LiteratureWatch