New perspective on the risk markers for left atrial thrombosis in patients with atrial fibrillation

Abstract

Abstract Background Anticoagulant therapy is one of the important aspects of atrial fibrillation (AF) management, which can effectively reduce the formation of left atrial thrombosis (LAT) and the occurrence of embolic events. The CHA2DS2-VASc score is a commonly used risk assessment tool for embolic events, and it has guiding significance for anticoagulant therapy. However, a large number of recent studies have clearly shown that some of the markers that are not included in the score affect the formation of LAT. Objective This single-center study probed for risk markers for LAT by analyzing the clinical features of patients who experienced AF. Methods We reviewed patients with AF who had undergone a transesophageal echocardiography exam over the past 6 years and used binary logistic regression analysis to identify risk markers other than CHA2DS2-VASc score. For the risk markers found, the propensity score matching (PSM) was used to further evaluate whether it was an independent risk marker for LAT. The newly discovered markers were added to the score, and receiver operating characteristic analysis was used to evaluate whether the ability of the model to predict LAT was improved. Results A total of 2246 patients were included in the study. In total, 838 of them were anticoagulated (314 with rivaroxaban, 57 with dabigatran, and 467 with warfarin) and 30 patients (1.33%) had LAT. Regression analysis revealed abnormal uric acid metabolism (abUA) and obesity were risk markers for LAT. Further PSM analysis found that abUA was an independent risk marker for LAT. After including abUA, the CHA2DS2-VASc score was more accurate for LAT prediction (area under the curve difference is 0.0651, 95% confidence interval: 0.0247, 0.1050, Z = 3.158, P = 0.0016). Conclusions AbUA is an independent risk marker for LAT. After considering abUA, the CHA2DS2-VASc score for LAT is more accurate.