Epicardial Atrial Fat at Cardiac Magnetic Resonance Imaging and AF Recurrence after Transcatheter Ablation.

The relationship between epicardial adipose tissue (EAT) and atrial fibrillation (AF) has gained interest in recent years. The previous literature on the topic presents great heterogeneity, focusing especially on computed tomography imaging. The aim of the present study is to determine whether an increased volume of left atrial (LA) EAT evaluated at routine pre-procedural cardiac magnetic resonance imaging (MRI) relates to AF recurrences after catheter ablation. A total of 50 patients undergoing AF cryoballoon ablation and pre-procedural cardiac MRI allowing quantification of LA EAT were enrolled. In one patient, the segmentation of LA EAT could not be achieved. After a median follow-up of 16.0 months, AF recurrences occurred in 17 patients (34%). The absolute volume of EAT was not different in patients with and without AF recurrences (10.35 mL vs. 10.29 mL; -value = 0.963), whereas the volume of EAT indexed on the LA volume (EATi) was lower, albeit non-statistically significant, in patients free from arrhythmias (12.77% vs. 14.06%; -value = 0.467). The receiver operating characteristic curve testing the ability of LA EATi to predict AF recurrence after catheter ablation showed sub-optimal performance (AUC: 0.588). The finest identified cut-off of LA EATi was 10.65%, achieving a sensitivity of 0.5, a specificity of 0.82, a positive predictive value of 0.59 and a negative predictive value of 0.76. Patients with values of LA EATi lower than 10.65% showed greater survival, free from arrhythmias, than patients with values above this cut-off (84% vs. 48%; -value = 0.04). In conclusion, EAT volume indexed on the LA volume evaluated at cardiac MRI emerges as a possible independent predictor of arrhythmia recurrence after AF cryoballoon ablation. Nevertheless, prospective studies are needed to confirm this finding and eventually sustain routine EAT evaluation in the management of patients undergoing AF catheter ablation.