Citations

Importance Left atrial appendage occlusion (LAAO) is used in patients with contraindications to oral anticoagulation, who are at increased risk for major bleeding (MB). Objective To evaluate the incidence, predictors, management, and clinical outcomes of MB after LAAO in the Amulet IDE trial. Design, Setting, and Participants The randomized clinical Amulet IDE trial enrolled patients from 2016 through 2019 with 5-year follow-up. Procedures were performed at specialized LAAO centers. Of 1878 randomized patients (1:1 Amulet vs Watchman), 1833 underwent attempted LAAO implantation. These data were analyzed from May 2025 through November 2025. Main Outcomes and Measures Pooled (Amulet and Watchman) patients with vs without MB (Bleeding Academic Research Consortium [BARC] score ≥3: overt bleeding with transfusion and ≥3 g/dL hemoglobin drop) were compared. Results This study included a total of 1833 patients (57.7% male and 42.3% female; mean [SD] age, 75.8 [7.5] years). MB occurred in 331 patients (18.1%) over 5 years (annualized rate 5.9% per year). Patients with MB were older (76.8 vs 74.7 years) and had higher CHA 2 DS 2 -VASc (4.8 vs 4.6) and HAS-BLED (3.4 vs 3.2) scores, as well as greater prevalence of diabetes (38.7% vs 34.1%), prior MB (45.3% vs 28.0%), and kidney disease (9.4% vs 4.5%). MB risk was highest in the first 6 months after LAAO (20.5% per year), then decreased to 3.9% per year through 5 years. Most MB events were nonprocedural (88.8%) with gastrointestinal bleeding accounting for 252 of 438 events. Independent predictors of MB included increasing age (hazard ratio [HR], 1.04; 95% CI, 1.02-1.06), female sex (HR, 1.25; 95% CI, 1.01-1.56), diabetes (HR, 1.26; 95% CI, 1.01-1.58), prior MB (HR, 1.93; 95% CI, 1.55-2.40), and kidney disease (HR, 2.15; 95% CI, 1.48, 3.12). At first MB event, 47.1% of patients were not receiving antithrombotic therapy and 92 patients had recurrent events. Patients with MB had significantly higher rates of the composite of stroke, systemic embolism, or cardiovascular death (32.5% vs 18.0%; P < .001), driven by higher stroke (14.7% vs 6.6%) and cardiovascular death (24.4% vs 12.9%) rates. All-cause mortality was also higher (49.2% vs 25.4%; P < .001). MB was fatal (BARC score of 5) in 25 patients (1.4%). Conclusions and Relevance In this study, MB was common in this high-risk population and largely related to underlying patient factors. Early MB risk was highest 6 months post-LAAO, then dropped considerably. Most events were gastrointestinal and occurred even in patients on minimal or no antithrombotic therapy. MB after LAAO was strongly associated with increased stroke and mortality. Trial Registration ClinicalTrials.gov Identifier: NCT02879448

BackgroundPostmenopausal hormone therapy (HT) increases the risk of venous thrombosis and ischemic stroke.ObjectivesWe postulated that HT might increase the risk of ischemic stroke by promoting venous clots that travel to the brain through a right to left shunt (RLS).MethodsA total of 2,389 records were studied. After eliminating the premenopausal patients, and those with TIAs and non‐ischemic strokes, the medical records of 1846 postmenopausal women hospitalized at four institutions for ischemic stroke were reviewed to identify those who had undergone an adequate study to assess for RLS. The proportion of women with a shunt in users and non‐users of HT was compared in stroke patients and in a reference population consisting of postmenopausal women undergoing elective cardiac catheterization.ResultsThere were 363 (20%) records that had complete data and were included in the analysis. A shunt was more prevalent in patients with a cryptogenic stroke than in patients with a stroke of known cause (55/88 (63%) vs. 53/275 (19%), P < 0.001). In patients with a stroke of known cause, the frequency of a shunt was similar to that in reference women 31/136 (23%), and the proportion of women with a shunt was similar in non‐users and current users of HT (14% vs. 20%, P = 0.40). However, among patients with a cryptogenic stroke, the prevalence of a shunt was 1.5 times higher in current users than non‐users of HT (82% vs. 56%, P = 0.04).ConclusionsApproximately 23% of older women have a RLS. HT in these women may increase the risk of ischemic stroke by promoting paradoxical embolism. © 2014 Wiley Periodicals, Inc.