Study on the correlation between false‐positive filling defect in LAA CT and LAA structure in patients with atrial fibrillation based on TEE

OBJECTIVE: This study aims to explore the actual meaning of “false positive filling defect” in left atrial appendage (LAA) computed tomography (CT) in patients with atrial fibrillation (AF), with transesophageal echocardiography (TEE) as the gold standard. METHODS: Patients with AF undergoing cardiac CT angiography and TEE examinations for proposed radiofrequency catheter ablation between October 2020 and October 2021 were selected as the study subjects. Transesophageal echocardiography was taken as the “gold standard,” and spontaneous echocardiographic contrast (SEC) and thrombus events were defined as positive events. The CT manifestations were classified into three groups (true positive, false positive, and true negative) to evaluate the differences in left atrium (LA) anterior–posterior diameter (LAAP), LA anterior wall thickness, and LAA orifice long diameter and short diameter, area, and depth between the three groups. RESULTS: (1) There was no statistical difference in LA anterior wall thickness between the three groups (p > .05); there was a statistical difference in LAAP (only) between the true‐positive group and the true‐negative group (p < .05). (2) There was a statistical difference in LAA orifice long diameter, short diameter, and area between the true‐positive group and the true‐negative group as well as between the false‐positive group and the true‐negative group (p < .05). (3) There was a statistical difference in LAA depth between the true‐positive group and the false‐positive group as well as between the true‐positive group and the true‐negative group (p < .05). (4) The area under the receiver operator characteristic curve (AUC) of LAA depth affecting the LAA thrombus and SEC was 0.863 (confidence interval = 0.718–1.000), the sensitivity was 77.8%, and the specificity was 90.6% for predicting the occurrence of LAA thrombus and SEC in patients with nonvalvular AF (NVAF) and an LAA depth of ≥50.84 mm. CONCLUSIONS: There was a difference in LAA diameter between the TEE‐based CT false‐positive group and the other groups. A “CT false positive” is an objectively existing state, and CT might be able to identify the LAA hemodynamic disorder earlier than TEE. Furthermore, a CT + TEE combined application could more accurately evaluate LAA hemodynamics in patients with AF.