Left atrial appendage thrombus detection in routine workflow for patients with atrial fibrillation under oral anticoagulation: transoesophageal echocardiography vs. cardiac computed tomography

Aims The presence of a left atrial appendage (LAA) thrombus is an absolute contraindication for ablation. Despite the use of oral anticoagulant (OAC) therapy, LAA thrombi may still occur. The objective of this study is to identify the incidence of LAA thrombi in transoesophageal echocardiography (TEE) (the gold standard) and computed tomography (CT) scans and to investigate any correlation in thrombus detection between the two methods. Additionally, the study aims to investigate whether LAA flow velocity or volume influences thrombus detection. Methods and results Between May 2018 and October 2023, patients with atrial fibrillation/atrial tachycardia (AF/AT) under OAC, which were scheduled for AF catheter ablation, were included. TEE and CT were carried out at intervals of up to 7 days. LAA flow velocity was measured in the TEE. The volume of the LA, including LAA, excluding the pulmonary veins (PV), was obtained from a 3D reconstruction of the LA from CT. The study included 283 patients (pts), mean age of 66y (± 9 years), 182 male, mean CHA₂DS₂-VA score 3 [0–7]. All pts were orally anticoagulated with the majority under non-vitamin K antagonist oral anticoagulant (NOAC) (n = 265). Mean LA volume (LA + LAA) was 158 mL (± 45 mL). LAA flow velocity was reduced (<40 cm/s) in 119 pts (42%). In the majority of cases TEE and CT were performed at the same day (51%) or with an interval of 1 day (22%). A LAA thrombus was identified in 35 (12%) patients at least in one of the two methods. CT detected a thrombus in 27 patients, while TEE identified a thrombus in 16 patients, with both methods agreeing in 8 cases. A significant association between LAA flow velocity and thrombus detection by TEE was observed (P < 0.001; r = 0.36). In contrast, no significant relationship was observed between left atrial volume and thrombus presence on CT (P = 0.964). Conclusion In 12% of OAC-treated patients, a thrombus was detected in at least one exam, with concordant TEE and CT diagnoses in only 25%. Both methods excluded thrombi in 80% of cases. LAA flow velocity correlated with TEE findings, whereas LA volume showed no association with CT. Discrepancies highlight diagnostic challenges: TEE is operator-dependent, and CT is limited by flow and timing.