Acute esophageal safety of impedance-drop based left atrial point-by-point radiofrequency catheter ablation

Abstract Background Pulmonary vein isolation (PVI) is an established treatment for atrial fibrillation (AF). Atrio-esophageal fistula (AEF) is a rare (< 0.02%) major complication with high mortality rate following radiofrequency catheter ablation (RFCA). Previous studies suggested endoscopically detected esophageal lesions (EDEL) as a prerequisite for AEF after PVI. Incidence of EDEL following 50W high-power short-duration (HPSD) AF ablation with strict post-ablation endoscopy follow-up is reported in up to 6%. Studies are however limited to ablation-index guided lesion formation and no evidence regarding impedance-drop based ablation is available. Aim To evaluate the acute esophageal safety of impedance-drop based RFCA for PVI. Methods All patients undergoing left atrial ablation for AF or atypical/left atrial flutter using the IntellaNav StablePoint catheter and Rhythmia system (Boston Scientific Inc.) were retrospectively included. Ablation was performed according to HPSD protocol (50W, 30mL/min saline flush) aiming for an impedance drop of -17 Ohm posterior and -20 Ohm anterior (with maximum of -40 Ohm), with maximal interlesion distance of 6mm. The need for additional ablation lines (e.g. mitral line, roof line, box lesion) or left atrial substrate modification (LASM) was at the discretion of the operator. All patients underwent upper GI endoscopy the next working day. If no posterior ablation was performed, no endoscopy was performed. Results A total of 84 patients were identified, 53 patients underwent first-PVI and 31 underwent redo procedure. Five patients undergoing redo procedure did not undergo upper GI endoscopy, excluding them from further analysis. Baseline patient characteristics are described in Table 1. Patients were relatively young, predominantly male, with (pre-)obesity. In the redo PVI group, persistent atrial fibrillation was more common (76.9%). Procedural characteristics are described in Table 2. In the first-PVI group, only two patients (3.8%) received LASM on top of circumferential pulmonary vein isolation. Reisolation of the pulmonary veins (circumferential or focal) was necessary in 22 patients (84.6%) in the redo PVI group, and 12 patients (46.2%) needed additional linear ablation or LASM. Upper GI endoscopy was performed after a median of 1 day after the procedure. In only 1 patient (1.3% of total), a EDEL type I (erosion) was described, necessitating no further follow-up. Relevant gastric food retention was detected in 10 patients (18.5%) and 1 patient (3.8%) in the first-PVI and redo-PVI group respectively. Other periprocedural complications were in line with those reported in the literature. Conclusions The esophageal safety impedance drop based RFCA ablation was comparable to that of ablation-index based ablation methods, documenting only 1.3% EDEL type I in a retrospective analysis. Patients undergoing first PVI had more food retention upon endoscopy, possible related to more extensive posterior ablation.Baseline patient characteristics Procedural characteristics