Prevalence and mechanism of rotor activation identified during atrial fibrillation by noncontact mapping: Lack of evidence for a role in the maintenance of atrial fibrillation

Background It remains unclear whether atrial fibrillation (AF) is maintained by the rotor. Objective We evaluated the role of the rotor and examined its mechanism. Methods Among 75 patients with AF (60 paroxysmal, 15 persistent AF) who underwent 3-dimensional noncontact left atrial mapping during AF, we examined the prevalence and location of rotor activation and elucidated its mechanism. Catheter ablation was performed in a stepwise fashion (linear roof lesion and complex fractionated atrial electrogram ablation after pulmonary vein [PV) isolation) until AF termination. Results Rotor activation was observed in 11 patients (14.7%; 10 paroxysmal and 1 persistent AF) (tachycardia cycle length 160.0 ± 19.8 ms). Rotors were observed transiently (duration 6128 ± 9094 ms) during AF at the roof (n = 5), septum (n = 3), and ostium of the left superior PV (n = 3). Five rotors circulated in clockwise and 6 in counterclockwise directions. The length of the block line at the center of the rotor was 15.2 ± 6.9 mm. The electrograms at the block line showed low-amplitude multiple deflections (n = 7) or double potentials (n = 4), and the amplitudes during rotor activation were significantly lower than those during sinus rhythm (0.27 ± 0.18 mV vs 1.22 ± 0.92 mV; P < .01). No conduction disturbances were found during sinus rhythm, suggesting that the central line of block was formed functionally. AF was terminated by PV isolation alone without additional lesions in patients with rotors. Conclusion Functionally formed rotor activation was observed during AF in a limited number of patients. These rotor activations may not be related to AF maintenance, but rather may reflect a transient organization of random propagation.