BACKGROUND: The complex fractionated atrial electrogram (CFAE) region has been suggested to contribute to the maintenance of atrial fibrillation (AF), but its role for the initiation of AF has not been clarified. OBJECTIVE: We analyzed the mechanisms of the initiation of random reentrant wave propagation at AF onset, especially in relation to CFAE region. METHODS: Endocardial mapping of the left atrium using a 3-dimensional noncontact mapping system was performed in 19 patients. RESULTS: Thirty-two spontaneous AF onset episodes, which were initiated by the focal repetitive discharges (9 ± 9 beats), deriving from the pulmonary veins (PV) (n = 17) and from non-PV CFAE regions (n = 15) were observed. The coupling intervals of the focal discharges that initiated AF (AF-D) were significantly shorter than those that did not initiate AF (non-AF-D) (179 ± 33 ms vs. 217 ± 45 ms, P = .0005). After the AF-D, localized conduction blocks occurred in the CFAE region. Subsequently, the waves propagated to the remainder of the atrium, accompanying the anchored activation around the localized conduction block lines in the CFAE regions. Left atrial activation times of AF-D were significantly longer than those of non-AF-D (151 ± 35 ms vs. 83 ± 17 ms, P < .0001). These longer activation times after AF-D enabled the waves to reenter the previously blocked CFAE region from the opposite direction, and thus the meandering reentrant wave propagation was initiated. CONCLUSION: Unidirectional conduction block in the CFAE region and subsequent prolonged left atrial activation time following short coupled premature discharge were the underlying mechanisms of AF initiation, suggesting the importance of the CFAE region as the substrate for AF onset.
BACKGROUND: The complex fractionated atrial electrogram (CFAE) region has been suggested to contribute to the maintenance of atrial fibrillation (AF), but its role for the initiation of AF has not been clarified. OBJECTIVE: We analyzed the mechanisms of the initiation of random reentrant wave propagation at AF onset, especially in relation to CFAE region. METHODS: Endocardial mapping of the left atrium using a 3-dimensional noncontact mapping system was performed in 19 patients. RESULTS: Thirty-two spontaneous AF onset episodes, which were initiated by the focal repetitive discharges (9 ± 9 beats), deriving from the pulmonary veins (PV) (n = 17) and from non-PV CFAE regions (n = 15) were observed. The coupling intervals of the focal discharges that initiated AF (AF-D) were significantly shorter than those that did not initiate AF (non-AF-D) (179 ± 33 ms vs. 217 ± 45 ms, P = .0005). After the AF-D, localized conduction blocks occurred in the CFAE region. Subsequently, the waves propagated to the remainder of the atrium, accompanying the anchored activation around the localized conduction block lines in the CFAE regions. Left atrial activation times of AF-D were significantly longer than those of non-AF-D (151 ± 35 ms vs. 83 ± 17 ms, P < .0001). These longer activation times after AF-D enabled the waves to reenter the previously blocked CFAE region from the opposite direction, and thus the meandering reentrant wave propagation was initiated. CONCLUSION: Unidirectional conduction block in the CFAE region and subsequent prolonged left atrial activation time following short coupled premature discharge were the underlying mechanisms of AF initiation, suggesting the importance of the CFAE region as the substrate for AF onset.