Background: Cerebral diffusion-weighted MRI lesions have been observed after catheter ablation of atrial fibrillation. We hypothesized that conditions predisposing to microembolization could be identified using a porcine model of pulmonary vein ablation and an extracorporeal circulation loop.
Methods and results: Ablations of the pulmonary veins were performed in 18 swine with echo monitoring. The femoral artery and vein were cannulated and an extracorporeal circulation loop with 2 ultrasonic bubble detectors and a 73-μm filter were placed in series. Microemboli and microbubbles were compared between ablation with an irrigated radiofrequency system (Biosense-Webster) and a phased radiofrequency multielectrode system (pulmonary vein ablation catheter [PVAC], Medtronic, Inc, Carlsbad, CA) in unipolar and 3 blended unipolar/bipolar modes. Animal pathology was examined. The size and number of microbubbles observed during ablation ranged from 30 to 180 μm and 0 to 3253 bubbles per ablation. Microbubble volumes with PVAC (29.1 nL) were greater than with irrigated radiofrequency (0.4 nL; P=0.045), and greatest with type II or III microbubbles on transesophageal echocardiography. Ablation with the PVAC showed fewest microbubbles in the unipolar mode (P=0.012 versus bipolar). The most occurred during bipolar energy delivery with overlap of proximal and distal electrodes (median microbubble volume, 1744 nL; interquartile range, 737-4082 nL; maximum, 19 516 nL). No cerebral MRI lesions were seen, but 2 animals had renal embolization.
Conclusions: Left atrial ablation with irrigated radiofrequency and PVAC catheters in swine is associated with microbubble and microembolus production. Avoiding overlap of electrodes 1 and 10 on PVAC should reduce the microembolic burden associated with this procedure.