Abstract
BACKGROUND: Complex fractionated atrial electrograms (CFAE) are targets of atrial fibrillation (AF) ablation. Serial high-density maps were evaluated to understand the impact of activation direction and rate on electrogram (EGM) fractionation. METHODS AND RESULTS: Eighteen patients (9 persistent) underwent high-density, 3-dimensional, left-atrial mapping (>400 points/map) during AF, sinus (SR), and CS-paced (CSp) rhythms. In SR and CSp, fractionation was defined as an EGM with ≥4 deflections, although, in AF, CFE-mean <80 ms was considered as continuous CFAE. The anatomic distribution of CFAE sites was assessed, quantified, and correlated between rhythms. Mechanisms underlying fractionation were investigated by analysis of voltage, activation, and propagation maps. A minority of continuous CFAE sites displayed EGM fractionation in SR (15+/-4%) and CSp (12+/-8%). EGM fractionation did not match between SR and CSp at 70+/-10% sites. Activation maps in SR and CSp showed that wave collision (71%) and regional slow conduction (24%) caused EGM fractionation. EGM voltage during AF (0.59+/-0.58 mV) was lower than during SR and CSp (>1.0 mV) at all sites. During AF, the EGM voltage was higher at continuous CFAE sites than at non-CFAE sites (0.53 mV (Q1, Q3: 0.33 to 0.83) versus 0.30 mV (Q1, Q3: 0.18 to 0.515), P<0.00001). Global LA voltage in AF was lower in patients with persistent AF versus patients with paroxysmal AF (0.6+/-0.59 mV versus 1.12+/-1.32 mV, P<0.01). CONCLUSIONS: The distribution of fractionated EGMs is highly variable, depending on direction and rate of activation (SR versus CSp versus AF). Fractionation in SR and CSp rhythms mostly resulted from wave collision. All sites with continuous fractionation in AF displayed normal voltage in SR, suggesting absence of structural scar. Thus, many fractionated EGMs are functional in nature, and their sites dynamic.