Clinical and transesophageal echocardiographic variables for prediction of thromboembolic events in patients with nonvalvular atrial fibrillation at low-intermediate risk

Abstract

Background: There is no clear consensus about antithrombotic treatment in atrial fibrillation (AF) patients at low-intermediate thromboembolic risk. Transesophageal echocardiography (TEE) is useful for prediction of thromboembolic events in AF.

Methods and results: Of 498 patients with nonvalvular AF, incidence of stroke, cardiac events, and mortality was investigated in 280 patients with CHADS(2) score 0 or 1 (mean age 64 years, mean follow-up 6.4 ± 3.1 years). Left atrial abnormality (low left atrial appendage flow, spontaneous echo contrast, or thrombi), complex aortic plaque (mobile, ulcerated, pedunculate, or thickness ≥ 4mm), or both were defined as TEE risk. The incidences of ischemic stroke, cardiovascular events, and death were higher in patients with TEE risk than in those without the risk (2.0%/year vs. 0.5%/year, p<0.05; 4.7%/year vs. 1.9%/year, p<0.01; and 4.7%/year vs. 2.0%/year, p<0.01, respectively). This was also true for patients with CHADS(2) score of 0 (1.7%/year vs. 0.3%/year, p<0.05; 4.1%/year vs. 1.6%/year, p<0.05; and 3.9%/year vs. 1.4%/year, p<0.01; respectively). On multivariate analysis, TEE risk predicted ischemic stroke, cardiovascular events, and mortality independently of clinical variables or CHADS(2) score.

Conclusions: TEE could be useful for further stratification of patients with nonvalvular AF stratified at low-intermediate risk (CHADS(2) score 0 or 1) and could indicate who should receive anticoagulation treatment.