Left Atrial Appendage Occlusion versus Oral Anticoagulation in Atrial Fibrillation: A Meta-Analysis of Randomized Trials
In brief
Appendage closure cuts nonprocedural major bleeding 29%, but stroke signal remains uncertain
In six randomized trials involving 7,004 people with atrial fibrillation, left atrial appendage closure reduced nonprocedural major bleeding by 29% - one fewer event for every 19 patients treated - and also reduced clinically relevant nonmajor bleeding. Ischemic strokes were more frequent, but that finding was fragile; overall stroke and mortality were similar, leaving the benefit-harm tradeoff central to treatment decisions.
- Journal
- JACC. Advances (Q1)
- Published
- 1 October 2026
- Study design
- Systematic review / meta-analysis of RCTs
- Evidence level
- Level 1, High (CEBM 1a)
- Authors
- Aravind Dilli Babu, Scott J Cameron, Haroon Gill, Asher Gorantla, Karan Bhanushali, Subhashini Gowda, et al.
- PMID
- 42954691
- DOI
- 10.1016/j.jacadv.2026.103251
Why clinicians should know about it
- Picked for Neurology (clinical) (top studies of the week, 1 November 2026): Compares atrial fibrillation stroke prevention strategies
Abstract
BACKGROUND: Emergence of direct oral anticoagulants as the first-line therapy for stroke prevention in nonvalvular atrial fibrillation has redefined the comparator benchmark for left atrial appendage occlusion (LAAO). OBJECTIVES: The objective of the study was to compare the efficacy and temporal safety of LAAO vs oral anticoagulation (OAC). METHODS: Randomized trials of LAAO vs OAC in nonvalvular atrial fibrillation were pooled for incidence rate ratios (IRRs) per patient-year; key endpoints were assessed with trial sequential analysis (TSA) and Grading of Recommendations Assessment, Development, and Evaluation (GRADE). RESULTS: Across 6 trials (N = 7,004), LAAO vs OAC was associated with higher ischemic stroke (130 vs 76 events; IRR: 1.43; 95% CI: 1.07-1.91; P = 0.01; number needed to harm: 101; GRADE: low); however, this was sensitive to individual trial exclusion and TSA indicated insufficient information (41% of required size). All-cause stroke (152 vs 114; IRR: 1.11; 95% CI: 0.86-1.43; P = 0.43) and all-cause mortality (403 vs 368; IRR: 0.93; 95% CI: 0.78-1.10; P = 0.39) were comparable. Total major bleeding did not differ (IRR: 0.91; P = 0.29), but LAAO significantly reduced nonprocedural major bleeding (194 vs 232; IRR: 0.71; 95% CI: 0.55-0.91; P = 0.01; number needed to treat: 19; GRADE: high) and clinically relevant nonmajor bleeding (145 vs 303; IRR: 0.48; 95% CI: 0.39-0.58; P < 0.001; number needed to treat: 15). CONCLUSIONS: LAAO significantly reduces nonprocedural major bleeding (TSA-confirmed) and clinically relevant nonmajor bleeding vs OAC. A trend toward higher ischemic stroke was not TSA-confirmed and was sensitive to trial exclusion; all-cause stroke and mortality were similar. These findings quantify the benefit-harm tradeoff central to LAAO decisions.
Abstract as published, via PubMed.
For healthcare professionals. The summary is generated by AI from the published abstract, and the evidence level is assigned automatically from the study design on the Oxford CEBM hierarchy. Neither is medical advice. Read the full paper before changing practice.