Half-dose rivaroxaban vs antiplatelet therapy for preventing silent cerebral embolism after left atrial appendage occlusion: the HALO-SCE trial

Vollständiger Abstract

Worum geht es in dieser Arbeit?

Abstract Background and aims The optimal long-term antithrombotic strategy after left atrial appendage occlusion (LAAO) remains undetermined. The present study aimed to investigate whether half-dose rivaroxaban (10 mg daily) could better reduce silent cerebral embolic lesions (SCEs) and preserve cognitive function compared to antiplatelet therapy after successful LAAO. Methods In this investigator-initiated, prospective, multicenter, randomized controlled trial, patients with successful LAAO confirmed 45 days post-procedure were assigned 1:1 to half-dose rivaroxaban or antiplatelet therapy group. Diffusion-weighted magnetic resonance imaging and cognitive assessments were repeated at 90, 180 and 365 days after LAAO. The primary outcome was the patient-level incidence of any newly detected SCE during follow-up. Secondary outcomes included cognitive trajectories, SCE burden, and a composite of all-cause mortality, clinical thromboembolic events and major bleeding. Results Between December 2022 and February 2025, 164 patients were randomized. The patient-level incidence of new SCEs was significantly lower in the half-dose rivaroxaban group than in the antiplatelet therapy group (10/82 [12.2%] vs. 26/82 [31.7%]; P = 0.005). At 365 days, model-derived between-group differences favored the half-dose rivaroxaban group for both Mini-Mental State Examination (2.56; 95% confidence interval [CI] 1.11-4.01; P < 0.001) and Montreal Cognitive Assessment (2.67; 95% CI 1.07-4.26; P = 0.001) scores. The composite clinical outcome occurred in 2.4% of the half-dose rivaroxaban group vs. 11.0% of the antiplatelet therapy group (P = 0.057). Conclusions In patients eligible for oral anticoagulation after successful LAAO, rivaroxaban 10 mg daily significantly reduced SCEs and better maintained cognitive function compared with antiplatelet therapy, with numerically fewer composite clinical events.