Left Atrial Appendage Closure or Anticoagulation for Atrial Fibrillation

Left atrial appendage closure (LAAC) has emerged as an alternative to long-term anticoagulation for preventing stroke or systemic embolism in selected patients with atrial fibrillation (AF) based on several randomized controlled trials (RCTs) that collectively enrolled >7,000 participants. In this JACC Evidence Review, we interpret the major RCTs as distinct experiments, clarify what each can and cannot establish, and offer a phenotype-based framework for interpreting the evidence. The most recent RCTs, OPTION, CLOSURE-AF, and CHAMPION-AF, have prompted efforts to pool the evidence into a single summary estimate. However, these trials are not one coherent evidence base; they are distinct clinical experiments that asked different questions. They differ in patient phenotype, comparator, device generation, postprocedure antithrombotic regimen, endpoint construction, and follow-up duration. Study-level meta-analysis of published reports remains methodologically valid for estimating the broad direction of effect of a device class, but when the component trials differ materially in patient phenotype, comparator, and outcome construction, the pooled estimate does not map to a common patient group or treatment decision. Patient eligibility as well as thromboembolic and bleeding risk varied markedly across trials. In addition, device technology has progressed, and comparators have evolved from warfarin to direct oral anticoagulants. Heterogeneity and insufficient details also exist about the antithrombotic therapy regimen in LAAC arms and adjunct antiplatelet therapy in control arms. Methodological issues further complicate interpretation, including the use of bleeding events (ie, the putative ancillary benefit of LAAC compared with anticoagulation) within composite endpoints tested for noninferiority, varying noninferiority margins, analytical approaches that may undermine the interpretation of noninferiority, and exclusion of periprocedural bleeding from primary safety endpoints in some trials. Finally, the standard of care for anticoagulation in AF has evolved, obviating concomitant aspirin use in stable individuals on therapeutic anticoagulation, thereby reducing bleeding risk compared with that of the control arm of LAAC RCTs. For most anticoagulation-eligible patients, anticoagulation remains the preferred first-line strategy. LAAC is a reasonable option when long-term anticoagulation is unsafe or unsustainable provided that procedural risk, adjunct antithrombotic feasibility, surveillance needs, and competing mortality are carefully considered. RCT evidence in this latter population is limited, and RCTs in other populations have features that limit their support for LAAC as an alternative to anticoagulation. Because LAAC and anticoagulation offer different types of risk, aligning the most applicable evidence with patient values through shared decision-making is central to care.