Atrial thrombus detection on transoesophageal echocardiography in patients with atrial fibrillation undergoing cardioversion or catheter ablation: A pooled analysis of rates and predictors
Jean Jacques Noubiap, Thomas A. Agbaedeng, Aude Laetitia Ndoadoumgue, Ulrich Flore Nyaga, André Pascal Kengne
Abstract
Abstract Objective To summarize data on the rates and predictors of left atrial thrombus/left atrial appendage thrombus (LAT/LAAT) detection by transoesophageal echocardiography (TEE) before electrical cardioversion (ECV) or catheter ablation (CA) for atrial fibrillation (AF). Methods EMBASE, MEDLINE, and Web of Science Core Collection were searched to identify all studies providing relevant data and published by October 7, 2020. A random‐effects meta‐analysis method was used to pool effect size estimates. Results A total of 85 studies were included, reporting data from 56 660 patients with AF. In patients undergoing CA and ECV, the pooled prevalence of LAT/LAAT was 1.8% and 7.5% in those not on oral anticoagulation (OAC), 1.8% and 5.5% in those taking OAC, and 1.3% and 4.9% in case of adequate OAC, respectively. According to the type of OAC, the prevalence was 2.0% and 7.6% for vitamin K antagonist, 1.3% and 3.5% for direct oral anticoagulant. Predictors of LAT/LAAT detection were nonparoxysmal AF (odds ratio [OR]: 3.6, 95% confidence interval: 2.4–5.2), hypertension (OR: 2.9, 1.2–7.0), previous stroke (OR: 3.0, 1.6–5.63), heart failure (OR: 4.3, 2.7–6.8), and CHADS 2 score ≥2 (OR: 3.3, 1.9–5.8) for patients undergoing CA; and heart failure (OR: 2.8, 1.3–6.2) and the CHA 2 DS 2 ‐VASc score (OR: 2.55, 1.5–4.5) for those undergoing ECV. Conclusion The prevalence of LAT/LAAT in AF patients undergoing ECV or CA varies widely, mainly due to differences in patient risk profiles and OAC types. Further research should determine whether the predictors of LAT/LAAT detection identified by this study could be used to select patients who require preprocedural TEE.