Abstract
IMPORTANCE: Patients with atrial fibrillation (AF) who have ischemic stroke despite taking oral anticoagulation therapy (OAT) have a very high risk of recurrence. Left atrial appendage occlusion (LAAO) is a mechanical stroke prevention strategy that may provide additional protection in patients with thromboembolic events under OAT. OBJECTIVE: To compare percutaneous LAAO with continuing OAT alone regarding stroke prevention in patients with AF who had a thromboembolic event despite taking OAT. DESIGN, SETTING, AND PARTICIPANTS: This cohort study was a propensity score-matched comparison of the STR-OAC LAAO cohort, an international collaboration of 21 sites combining patients from multiple prospective registries of patients who underwent LAAO between 2010 and 2022. STR-OAC LAAO cohort patients who had follow-up longer than 3 months were propensity score-matched to a previously published control cohort comprising patients from an established international collaboration of investigator-initiated prospective studies. This control cohort included patients with nonvalvular AF, recent ischemic stroke or transient ischemic attack, and follow-up longer than 3 months who were taking OAT before the index event. Analyses were adjusted for imbalances in gender, age, hypertension, diabetes, and CHA2 DS2-VASc score. EXPOSURE: Left atrial appendage occlusion vs continuation of oral anticoagulation therapy alone (control group). MAIN OUTCOMES AND MEASURES: The primary outcome was time to first ischemic stroke. RESULTS: Four hundred thirty-three patients from the STR-OAC LAAO cohort (mean [SD] age, 72 [9] years; 171 [39%] females and 262 [61%] males; mean [SD] CHA2 DS2-VASc score, 5.0 [1.6]) were matched to 433 of 1140 patients (38%) from the control group. During 2-year follow-up, 50 patients experienced ischemic stroke: an annualized event rate of 2.8% per patient-year in the STR-OAC LAAO group vs 8.9% per patient-year in the control group. Left atrial appendage occlusion was associated with a lower risk of ischemic stroke (hazard ratio, 0.33; 95% CI, 0.19-0.58; P < .001) compared with the control group. After LAAO, OAT was discontinued in 290 patients (67%), and the remaining 143 patients (33%) continued OAT after LAAO as an adjunctive therapy. CONCLUSIONS AND RELEVANCE: In patients with nonvalvular AF and a prior thromboembolic event despite taking OAT, LAAO was associated with a lower risk of ischemic stroke compared with continued OAT alone. Randomized clinical trial data are needed to confirm that LAAO may be a promising treatment option for this population with a very high risk of stroke.