A novel score to predict left atrial appendage thrombus and sludge in patients with persistent atrial fibrillation on chronic anticoagulants

Abstract Introduction Left atrium (LA) and left atrial appendage (LAA) thrombus may occur during atrial fibrillation (AF) and atrial flutter (AFl), and potentially cause main embolic events. Sludge is a pre-thrombotic condition, associated with long term adverse outcomes. Current guidelines do not recommend to perform transesophageal echocardiography (TEE) before cardioversion (CV) of patients with persistent AF/AFl when at least 3 weeks of anticoagulation are administered with appropriate compliance. Instead, they suggest TEE before atrial ablations (Abl), even in anticoagulated patients. Evidence to support these approaches are weak and conflicting, due to not infrequent occurrence of peri-procedural stroke and several reports of not negligible prevalence of LAA thrombus finding. Purpose Aim of the study was to evaluate the real prevalence of LAA thrombus and sludge in a retrospective cohort of consecutive anticoagulated patients with persistent AF and AFl, who underwent TEE before a scheduled cardioversion or catheter ablation. Moreover, clinical and echocardiographic characteristics were analyzed to develop a score that could predict the individual risk of combined thrombus/sludge. Methods We conducted a retrospective, monocentric, observational study, recruiting 1037 consecutive patients with persistent AF/AFl referred to our center for CV or Abl. TEE had been always performed to rule out thrombus and sludge before procedures. We collected data about patients characteristics, medication and compliance, electrographic and echocardiographic findings and type of scheduled procedure. Data was compared in univariate and multivariate analysis to identify possible predictors of LAA thrombus and sludge, and we used the last significant results to create a predictive score with logistic regression analysis. We identified best cutoff value with ROC curve. Results 767 patients were included in final dataset. Mean CHA2DS2-VASc was 3.2. LA/LAA thrombus and sludge were identified in 63 subjects (8.21%), 17 thrombi (2.21%) and 45 sludge (5.86%). Presence of valvular AF (+1), previous stroke or TIA (+1.5), heart failure history (+1), ventricular systolic (LVEF < 45%)(+0.5) and diastolic (Lateral E/e’ > 9) dysfunctions (+1) were significantly associated to higher risk of thrombus/sludge occurrence, meanwhile AFl (-1) was associated to lower risk. A novel score, called HEELP score, was developed based on these variables, and a threshold of 1 point showed high sensitivity (89%) and remarkable negative predictive value (98%). Conclusion Thrombus and sludge prevalence is not infrequent in anticoagulated patients with persistent AF/AFl. A strategy to perform TEE before procedures in patients with high risk features according to our score could be effective in preventing main embolic events, but it has to be validated in further studies.LAA empty (1), sludge (2), thrombus (3)HEELP score