Atrial fibrillation (AF) ablation can result in oesophageal injuries that lead to atrio-oesophageal fistulae, a life-threatening complication. This study aimed to evaluate whether oesophageal cooling could prevent oesophageal lesions complicating AF ablation.Methods and Results
We randomly assigned 100 patients with drug-resistant AF to an oesophageal cooling group or a control group. In the oesophageal cooling group, we injected 5 mL of ice water into the oesophagus prior to radiofrequency (RF) energy delivery adjacent to the oesophagus. If the oesophageal temperature reached 42°C, the RF energy delivery was stopped, and the ice water injection was repeated. In the control group, oesophageal cooling was not applied. Oesophageal endoscopy was performed 1 day after the catheter ablation, and lesions were qualitatively assessed as mild, moderate, or severe. The numbers of ablation sites with an oesophageal temperature of >42°C were 1.7 ± 1.4 and 2.6 ± 1.7 in the oesophageal cooling group and the control group, respectively (P = 0.04), and the maximal oesophageal temperature at those sites was 43.0 ± 0.6 and 44.7 ± 0.9°C (P < 0.0001). Oesophageal lesions occurred almost equally between the oesophageal cooling group [10 of 50 patients (20%)] and the control group [11 of 50 patients (22%)]. However, the severity of the oesophageal lesions was slightly milder in the oesophageal cooling group (three moderate, seven mild) than in the control group (three severe, one moderate, seven mild).Conclusion
Oesophageal cooling may alleviate the severity of oesophageal lesions but does not reduce the incidence of this complication under the specific protocol evaluated here.