Mechanism of complex fractionated electrograms recorded during atrial fibrillation in a canine model

Complex fractionated atrial electrograms (CFEs) have been described as a target during atrial fibrillation (AF) ablation; however, the mechanism leading to CFEs is poorly understood. We used noncontact mapping in a canine model of AF to determine the activation patterns in areas of CFEs.

Sustained AF was induced in 10 canines with 10-12 weeks of atrial tachy-pacing at 440 ppm. A roving mapping catheter and noncontact multielectrode array (MEA) were deployed in the left atrium (LA). NavX software was used to construct a contact bipolar CFE LA map. The MEA was then used to reconstruct wavefront propagation in proximity to CFE regions. Wavefront propagation was assessed during three separate recording segments for each site.

There were 34 CFE regions identified (3.4/dog) and 102 noncontact CFE regional activation sequences studied. The CFE regions were stereotypically located at the junctions of (1) the left pulmonary vein (PV)/posterior LA, (2) right inferior PV/posterior LA, (3) right superior PV/anterior LA, and (4) the LA roof. The majority (47%) of CFE recordings were characterized by wavefront collision, usually between circulating LA wavefronts and entry/exit from the PVs. Thirty-eight (38%) CFE recordings were noted to be the central functional barrier of a reentrant wavefront. Ablation through CFE regions due to reentry led to AF termination and noninducibility in 3/5 animals.

In this pacing-induced AF model, common causes of CFEs include: (1) wavefront collision, (2) conduction through channels of functional block, (3) reentry. The vast majority of these CFE regions were caused by wavefront collision rather than true "drivers" of AF.