Progression From Esophageal Thermal Asymptomatic Lesion to Perforation Complicating Atrial Fibrillation Ablation: A Single-Center Registry.

BACKGROUND Up to 40% of patients demonstrate endoscopically detected asymptomatic esophageal lesions (EDEL) after atrial fibrillation ablation. METHODS AND RESULTS Patients undergoing first atrial fibrillation ablation and postinterventional esophageal endoscopy were included in the study. Occurrence of esophageal perforating complications during follow-up was related to documented EDEL (category 1: erythema/erosion; category 2: ulcer). In total, 1802 patients underwent first atrial fibrillation ablation procedure between January 2013 and August 2016 at our institution. Out of this group, 832 patients (506 male patients, 61%; 64.0±10.0 years) with symptomatic paroxysmal (n=345; 42%) or persistent atrial fibrillation underwent postprocedural esophageal endoscopy. Patients were ablated using single-tip ablation with conventional or surround flow irrigation and circular ablation catheters with open irrigation (nMARQ). In 295 of 832 patients (35%), a temperature probe was used. EDEL occurred in 150 patients (18%; n=98 category 1 EDEL, n=52 category 2 EDEL). In 5 of 832 patients (0.6%), an esophageal perforation (n=3) or an esophagopericardial or atrioesophageal fistula (n=2) occurred 15 to 28 days (19±6 days) after ablation. Two patients (1 atrioesophageal fistula and 1 esophagopericardial fistula) died. Esophageal perforation occurred only in patients with category 2 lesions (absolute risk, 9.6%). In a logistic regression analysis, ulcers were identified to be a significant predictor for esophageal perforating complications. CONCLUSIONS Postablation endoscopy seems to identify patients at high risk of esophageal perforating complications only occurring in patients with category 2 EDEL. One out of 10 postablation esophageal ulcers progressed to perforation, and no patient without esophageal thermal ulcers showed the occurrence of perforating esophageal complications.