Background:
Ablation of complex fractionated atrial electrograms (CFAEs) is controversial, primarily because of difficulty in visually distinguishing CFAEs representing an active site of driver activity from a passive site of double potentials, wave break, and/or slow conduction. We hypothesized that CFAEs within rotors in atrial fibrillation (AF) would exhibit highly recurrent behavior compared with CFAEs remote from these driver regions.
Methods:
Active and passive mechanisms of CFAE formation were simulated in several 2D 7.5 x 7.5 cm modified Luo-Rudy 1 models. CFAEs within areas of rotors were considered active, while those caused by wave break, slow conduction or double potentials remote from rotors were considered passive. Clinical signals were also collected during catheter ablation of paroxysmal AF (n=8 patients). An active driver CFAE site was defined by termination of AF with ablation followed by non-inducibility. A passive site was defined as CFAE occurring remotely. Detection of CFAEs was based on mean cycle length (MCL) calculated from 4 second windows using -dV/dt for detection (40ms refractory period/10ms maximum EGM width for simulations; 45ms/15ms respectively for clinical signals). Recurrence quantification analysis (RQA) was performed on discrete time series of simulated and clinical CFAE activations.
Results:
RQA was performed on 20 simulated EGMs. MCL was similar in both active and passive CFAEs (74±11ms and 78±6ms respectively, p=NS), but recurrence was significantly higher in active compared to passive sites (%recurrence: 61±22% active, 4±6% passive, p<0.01). In patients with AF, the driver sites were all located within the pulmonary vein antra while passive CFAEs were remote. The MCL of CFAEs at active driver sites was similar to that of passive sites (100±13ms active, 98±17ms passive, p=NS), but recurrence was significantly higher in the active driver sites (%recurrence: 18±15% active, 2±1% passive, p=0.02).
Conclusion:
CFAEs may occur due to either active or passive mechanisms. Sites within rotors or focal drivers of AF are more likely to exhibit recurrent patterns. RQA may be a powerful tool to differentiate driver from bystander CFAEs enabling more efficient targeting for ablation.
Characterization of human persistent atrial fibrillation electrograms using recurrence quantification analysis
