Prospective Randomized Evaluation of High Power During CLOSE-Guided Pulmonary Vein Isolation The POWER-AF Study

Background: CLOSE-guided atrial fibrillation (AF) ablation is based on contiguous (intertag distance <= 6 mm), optimized (Ablation Index >550 anteriorly and >400 posteriorly) point-by-point radiofrequency lesions. The optimal radiofrequency power remains unknown. Methods: The POWER-AF study is a prospective, randomized controlled monocentric study including patients with paroxysmal AF, planned for first CLOSE-guided pulmonary vein isolation using a contact force radiofrequency catheter (Thermocool SmartTouch, Biosense Webster, Inc, Irvine, CA). A total of 100 patients were randomized into 2 groups (1:1). The control group received AF ablation using the standard CLOSE protocol (35 W), whereas in the experimental group, pulmonary vein isolation was performed using high power (45 W). Endoscopic evaluation was performed in patients with intraesophageal temperature rise >38.5 degrees C. Results: The resulting sample size was 96 (48+48) patients. In the high power group, shorter procedure time (80 versus 102 minutes, P<0.001), shorter total radiofrequency application time (16 versus 26 minutes, P<0.001), and radiofrequency time per application (26 versus 37 s anteriorly, P<0.001 and 13 versus 17 s posteriorly, P<0.001) were observed. Endoscopic evaluation (performed in 19/48 versus 25/48 patients respectively, P=0.31) showed an ulcerative perforation in a high power group patient (treated by endoscopic stenting and normalization after approximate to 4 months) and a superficial ulcerative lesion in a control group patient (conservative treatment). Both occurred following excessive Ablation Index applications (up to 460 and 480, respectively) with excessive contact force (30 g on average, with peaks up to 50 g). Six-months AF recurrence was not significantly different (10% in high power versus 8% in control, P=0.74). Conclusions: This randomized controlled study shows that a 45 W radiofrequency power CLOSE protocol in patients with paroxysmal AF significantly increases the global procedural efficiency with similar midterm efficacy. However, our study showed a narrower safety margin and a limited increased efficiency at the posterior wall using high power. This advocates against the use of high power in the region neighboring the esophagus.