Background-Standard mapping and ablation of focal sources of atrial fibrillation are associated with very long procedure times and low efficacy. An anatomic approach to complete pulmonary vein isolation could overcome these limitations. Methods and Results-Fifteen patients with atrial fibrillation refractory to medication underwent circumferential isolation of the pulmonary veins by using a novel catheter, with an ultrasound transducer (8-MHz) mounted near the tip, in a saline-filled balloon. Twelve atrial foci and/or atrial fibrillation triggers were identified in 9 patients (pulmonary vein locations: left upper, 3; right upper, 6; right middle, 1; right lower, 1; and left inferior, 1). In 5 patients, lesions were placed in the absence of any mapped triggers. Irrespective of trigger mapping, circumferential isolation of both upper pulmonary veins was attempted in all patients. The lower pulmonary veins were ablated when sinus rhythm activation mapping revealed evidence of a sleeve of atrial muscle in the vein. The median number of lesions per patient required to isolate 1 pulmonary vein was 4 (range, 1 to 29). After ablation, no evidence of narrowing was seen with repeat venography or follow-up computed tomography scan. After a mean follow-up of 35Ϯ6 weeks, 5 patients had recurrence of atrial fibrillation. Three responded to drugs that were previously ineffective, and 2 remained in atrial fibrillation. Conclusions-This novel ultrasound ablation system can successfully isolate multiple pulmonary veins. At early follow-up, this approach seems to be effective in preventing recurrent atrial fibrillation in a significant number of patients.

Ostial anatomy of the veins may affect delivery of ultrasound energy to achieve circumferential lesions. Energy delivery at the ostium with a temperature > 60 degrees C may be important to maximize success. Reconfiguration of the system to overcome the shortcomings identified in the initial experience could increase its performance.

The cause of death and clinical characteristics of 26 patients that died after implantable cardioverter defibrillator placement were reviewed and compared to the 145 patients still living after a mean follow-up of 17 months. Operative mortality was 4% (7/171) and resulted from postoperative ventricular arrhythmias (four patients), heart failure (two patients), and respiratory failure (one patient). Operative mortality was significantly higher (1.7% vs 9.6%, P less than 0.05) following concomitant surgical procedures. Total late mortality was 11% (18/171). Thirteen deaths (75%) occurred in-hospital from progressive deterioration of left ventricular function (nine patients), arrhythmia (two patients), and noncardiac causes (two patients). Outpatient mortality was 3.5% (6/171) and resulted from presumed sudden cardiac death in five of six patients; two of the five had devices that were inactive, one had high defibrillation thresholds, and two had suspected bradyarrhythmic deaths. One postoperative death and one late in-hospital death were also considered sudden cardiac deaths for a total of seven patients with defibrillation system failures. By multivariant analysis, preoperative clinical characteristics associated with a worse prognosis following defibrillator implantation were identified: presentation as ventricular tachycardia (P less than 0.02), induction of sustained monomorphic ventricular tachycardia (P less than 0.05), poor left ventricular performance (P less than 0.01), poor functional status (P less than 0.001), and the use of diuretics (P less than 0.01). Frequent device discharges (P less than 0.001) and concomitant antitachycardia pacing systems (P less than 0.001) were markers for greater arrhythmia recurrence and were potent predictors of a worse prognosis and particularly sudden death.