Search

Atrioesophageal fistula (AEF) is a rare but devastating complication of radiofrequency catheter ablation (RFCA) of atrial fibrillation (AF). Surgical repair and esophageal stents are available treatment options for AEF. We report outcomes of these 2 management strategies.Nine patients with AEF post-RFCA for AF were included in this study. AEF was diagnosed based on symptoms and chest CT imaging. Of the 9 patients, 5 received stents and 4 underwent surgical repair of fistula.AF ablation was performed under general anesthesia (n = 4) or conscious sedation (n = 5). During ablation, RF power was maintained between 25 and 35 Watts in areas close to the esophagus and energy delivery discontinued when esophageal temperature reached 38 °C. Seven patients underwent ablation with 3.5-mm open-irrigated catheter, 1 with 8-mm nonirrigated catheter, and 1 had surgical epicardial ablation. Seven patients received proton pump inhibitor and sucralfate before and after procedure. AEF symptoms developed within 2–6 weeks from ablation. Esophageal stenting was performed in 5 patients (median age 58 years, median time from RFCA 4 weeks) and 4 underwent surgical repair (median age 54 years, median time from RFCA 4 weeks) within 2–4 hours from diagnosis. All 5 patients receiving stents died within 1 week of the procedure due to cerebral embolism, septic shock, or respiratory failure. On the other hand, the 4 patients that received surgical repair were alive at median follow-up of 2.1 years (P = 0.005).Esophageal stenting should be discouraged and prompt surgical repair is crucial for survival in patients with atrioesophageal fistula.

Thromboembolic events represent one of the most serious complications of atrial fibrillation (AF). Left atrial appendage (LAA) clots are the dominant cause of such events in AF patients. Therefore, LAA closure devices have been developed. We sought to evaluate the follow-up and management of early

Results: During 18±9 months of follow-up, 6 (38%) patients received appropriate ICD shock for rapid sustained VT. No baseline clinical and noninvasive inding predicted the occurrence of ICD shock at follow-up. Electroanatomic scar was present in 14 (88%) patients, and no signiicant correlation was found between the presence and extension of RV scar the occurrence of ICD shocks at follow-up. Ten (63%) patients presented slow conduction channels (i.e., fractionated, isolated and very late potentials) within scar. Reduced HRV parameters were found in patients with slow conduction channels within the scar (standard deviation of RR intervals [SDNN] 87.8±35.6 vs. 140.2±33.5 ms, P=0.01 in the time-domain; low frequency amplitude [LF] 19.2±15.2 vs. 34.5±13.5 ms, P=0.04 in the frequency-domain). Presence of slow-conduction channels within scar (P=0.034) and reduced HRV parameters (P=0.007 for the SDNN; P