Percutaneous left atrial appendage occlusion versus oral anticoagulation in nonvalvular atrial fibrillation.

Abstract

Percutaneous left atrial appendage occlusion (pLAAO) has emerged as an alternative to oral anticoagulation (OAC) for stroke prevention in atrial fibrillation, yet no individual trial has been powered to detect meaningful differences in ischemic stroke or systemic embolism. We evaluated the comparative efficacy and safety of pLAAO vs OAC in nonvalvular atrial fibrillation using pooled analysis of randomized controlled trials. We searched PubMed, Embase, and Cochrane Central through April 3, 2026. Risk ratios were pooled using a random-effects model. Bayesian and trial sequential analyses were performed to assess non-inferiority using a prespecified margin of RR = 1.40, and power of available evidence. The primary efficacy endpoint was ischemic stroke or systemic embolism; primary safety endpoints were major bleeding and hemorrhagic stroke. Seven RCTs enrolling 7353 patients were included. Compared with OAC, pLAAO was associated with significantly more ischemic stroke or systemic embolism (3.3% vs 2.0%; RR 1.48, 95% CI 1.06-2.07). Hemorrhagic stroke was numerically lower with pLAAO but non-significant (0.6% vs 0.9%; RR 0.65, 95% CI 0.39 - 1.07). Major bleeding did not differ (8.3% vs 8.1%; RR 0.98, 95% CI 0.84-1.15). Bayesian analysis yielded a posterior probability of non-inferiority of 45.3%, failing the prespecified margin. Trial sequential analysis confirmed insufficient evidence to establish non-inferiority, and the major bleeding Z-curve entered the futility zone. Compared with OAC, pLAAO was not non-inferior for the outcome of ischemic stroke or systemic embolism, and did not significantly reduce major bleeding, though hemorrhagic stroke was numerically lower with pLAAO.