Electrocardiographic imaging as preoperative tool in persistent and long-standing persistent atrial fibrillation: A prospective observational study

Surgical ablation for treating atrial fibrillation is currently performed mostly without preoperative electrophysiological imaging. This study aimed to investigate the use of non-invasive surface mapping as a preoperative tool to explore potential mechanisms and patterns involved in the electrophysiology of persistent and long-standing persistent atrial fibrillation. This prospective, observational study included cardiac surgery candidates without previous ablation. Bi-atrial epicardial activation sequences were obtained with electrocardiographic imaging and analyzed in an independent core lab. Statistical analyses included hierarchical clustering, which quantified three clusters based on the number of drivers to identify specific characteristics. All 51 patients [14 (27%) persistent; 37 (73%) long-standing persistent; atrial fibrillation, duration 42 months (IQR 14-120)] had bi-atrial electrophysiological abnormalities. Most rotors were harbored in the upper half of the right atrium, involved in almost all patients (50/51, 98%), followed by the pulmonary vein-areas and left-atrial backwall (48/51, 94%). Longer atrial fibrillation duration showed no association towards fewer rotor and focal activity (r=-0.08, p = 0.42; r=-0.06, p = 0.56, respectively). A significant correlation existed between larger left atrial-size and fewer rotors (r=-0.33, p < 0.001), but not focal activity (r = 0.01, p = 0.92). The clusters differed in atrial fibrillation duration and left atrial-size, and in their number of rotor and focal activities (p = 0.005, p < 0.001, respectively). The underlying electrophysiological mechanism was identified in all patients and consistently showed bi-atrial involvement irrespective of atrial fibrillation duration, left atrial-size, or concomitant heart disease. In larger left atria and longer atrial fibrillation duration, the observed lower number of rotors might be related to atrial tissue fibrosis and lower amplitudes. The results demonstrate the potential role of preoperative mapping to improve procedural planning and our understanding of patients' electrophysiology.