Asymptomatic cerebral lesions during pulmonary vein isolation under uninterrupted oral anticoagulation.

AIMS: Left atrial radiofrequency ablation has been shown to carry a risk of asymptomatic cerebral lesions. No data exist in patients under continued oral anticoagulation during the ablation procedure. The aim of this study was to quantify the amount of silent cerebral lesions assessed by pre-procedural and post-procedural magnetic resonance imaging (MRI) in patients under therapeutic international normalized ratio (INR) and to identify clinical or procedural parameters that correlate with cerebral embolism.

METHODS AND

RESULTS

A total of 131 consecutive patients undergoing catheter ablation for paroxysmal (n = 80, 61.1%) or persistent (n = 51, 38.9%) atrial fibrillation were included in the study. Pulmonary vein antrum isolation (PVI), roofline, mitral isthmus line, and complex fractionated atrial electrogram (CFAE) ablation using 3.5 mm open-irrigated tip catheters were performed, as needed. All patients underwent pre-procedural and post-procedural cerebral MRI. Post-procedural MRI revealed new embolic lesions in 16 patients (12.2%), all of them asymptomatic. Clinical parameters showing a significant correlation with cerebral embolism in univariate analysis were age (P = 0.027), persistent atrial fibrillation (vs. paroxysmal; P = 0.039), and spontaneous echo contrast in transesophageal echocardiography (P = 0.029). Significant procedural parameters were electric cardioversion (P = 0.041), PVI only (P = 0.008), and ablation of complex atrial electrograms (P = 0.005). Independent risk factors in multivariate analysis were age (P = 0.009), spontaneous echo contrast (P = 0.029) and CFAE ablation (P = 0.006).

CONCLUSION

Radiofrequency ablation in patients under continued oral therapeutic anticoagulation is associated with a substantial risk of silent embolism detected by cerebral MRI. Therefore, continuation of oral anticoagulation is not able to prevent cerebral embolism. A variety of different clinical and procedural factors seem to contribute to the risk of cerebral lesions.