INTRODUCTION: Sites of complex fractionated atrial electrograms (CFAE) and dominant frequency (DF) have been implicated in maintaining atrial fibrillation (AF); however, their relationship is poorly understood. METHODS AND RESULTS: Twenty patients underwent biatrial high-density contact mapping (507 +/- 150 points/patient) during AF. CFAE were characterized using software to quantify electrogram complexity (CFE-mean). Spectral analysis determined the frequency with greatest power and sites of high DF with a frequency gradient. CFE-mean was higher (less fractionated) for right compared with left atria (P < 0.001) and in paroxysmal compared with persistent AF (P < 0.001). DF was lower for right compared with left atria (P = 0.02) and in paroxysmal compared with persistent AF (P < 0.001). There was significant regional variation in DF in paroxysmal (P < 0.001) but not persistent AF. Highest DF points clustered together with 5.2 +/- 1.7 clusters/patient. Correlation between CFE-mean and DF was poor on a point-by-point basis (r =-0.17, P < 0.001), but moderate on an individual basis (r =-0.50, P = 0.03). Exploration of their spatial relationship demonstrated CFAE areas in close proximity (median 5 mm, IQR 2-10) to high DF sites; within 10 mm in 80% and 10-20 mm in 10%. Simultaneous activation mapping at these sites further supports this observation. CONCLUSION: Greater fractionation and higher DF are seen in persistent AF and left atria during AF. Preferential areas of high DF are observed in paroxysmal but not persistent AF. CFAE and DF correlate within an individual but not point-by-point. Exploration of their spatial relationship demonstrates CFAE in areas adjacent to high DF, and this is supported by activation mapping at these sites.
INTRODUCTION: Sites of complex fractionated atrial electrograms (CFAE) and dominant frequency (DF) have been implicated in maintaining atrial fibrillation (AF); however, their relationship is poorly understood. METHODS AND RESULTS: Twenty patients underwent biatrial high-density contact mapping (507 +/- 150 points/patient) during AF. CFAE were characterized using software to quantify electrogram complexity (CFE-mean). Spectral analysis determined the frequency with greatest power and sites of high DF with a frequency gradient. CFE-mean was higher (less fractionated) for right compared with left atria (P < 0.001) and in paroxysmal compared with persistent AF (P < 0.001). DF was lower for right compared with left atria (P = 0.02) and in paroxysmal compared with persistent AF (P < 0.001). There was significant regional variation in DF in paroxysmal (P < 0.001) but not persistent AF. Highest DF points clustered together with 5.2 +/- 1.7 clusters/patient. Correlation between CFE-mean and DF was poor on a point-by-point basis (r =-0.17, P < 0.001), but moderate on an individual basis (r =-0.50, P = 0.03). Exploration of their spatial relationship demonstrated CFAE areas in close proximity (median 5 mm, IQR 2-10) to high DF sites; within 10 mm in 80% and 10-20 mm in 10%. Simultaneous activation mapping at these sites further supports this observation. CONCLUSION: Greater fractionation and higher DF are seen in persistent AF and left atria during AF. Preferential areas of high DF are observed in paroxysmal but not persistent AF. CFAE and DF correlate within an individual but not point-by-point. Exploration of their spatial relationship demonstrates CFAE in areas adjacent to high DF, and this is supported by activation mapping at these sites.