Anatomical proximity between ganglionated plexi and epicardial adipose tissue in the left atrium: implication for 3D reconstructed epicardial adipose tissue-based ablation

Purpose We sought to better understand the association between and clinical implications of ganglionated plexi (GPs), epicardial adipose tissue (EAT) in the left atrium (LA), and the initiation and maintenance of atrial fibrillation (AF). Methods Three-dimensional (3D) computed tomography (CT) reconstruction images of the LA, PVs, and LA-EAT were merged with the LA geometry for 25 with paroxysmal AF (PAF) and 15 with persistent AF (PerAF) scheduled for ablation. High-frequency stimulation (20 Hz, 25 mA, 10 ms) was performed at three sites within each of the five major anatomical LA GPs (superior left, inferior left, anterior right, inferior right, and Marshall tract GPs) to elicit vagal responses. Correspondence between the five GP areas and vagal response sites, LA-EAT, and complex fractionated atrial electrograms (CFAEs) was examined. The long-term outcomes of adjuvant LA-EAT-based ablation were assessed in 31 patients with PAF and 102 with PerAF. Results LA-EAT overlapped 93 ± 14 % of five major anatomical GP areas, and the vagal response sites and CFAE sites corresponded to 77 ± 23 and 76 ± 25 %, respectively ( p < 0.05 for both vs. LA-EAT locations). The EAT-based ablation eliminated 97.6 % of the vagal response sites and lengthened the AF cycle from 160 ± 26 to 193 ± 27 ms ( p < 0.001). The 2-year arrhythmia-free rate after EAT-based ablation was 72 % in the PAF group and 73 % in the PerAF group ( p = 0.614). Conclusions LA-EAT tends to overlie the major anatomical GP areas including most of the active GP response sites and CFAEs associated with AF. Ablation of GPs and CFAEs may explain the long-term efficacy of EAT-based ablation following extensive encircling pulmonary vein isolation (EEPVI) for AF.