Prediction of ablation success in atrial fibrillation patients based on electrocardiographic imaging

Abstract

Abstract Funding Acknowledgements Type of funding sources: Public Institution(s). Main funding source(s): Instituto de Salud Carlos III, and Ministerio de Ciencia, Innovación y Universidades (supported by FEDER Fondo Europeo de Desarrollo Regional PI17/01106), Agencia Estatal de Investigación (RYC2018-024346B-750 and PID2020-119364RB-100), Generalitat Valenciana Grants (ACIF/2020/265) and EIT Health (Activity code 19600), EIT Health is supported by EIT, a body of the European Union). Introduction Success rates of pulmonary vein isolation (PVI) are still suboptimal for termination of Atrial Fibrillation (AF). Electrocardiographic imaging (ECGI) could be used to determine the state of the atrial substrate non-invasively. Objective Improve the current stratification protocol for AF patients candidates to PVI using ECGI metrics prior to the referral to an ablation procedure. Methods ECGI from 35 AF patients (22 male, 10 paroxysmal, 63.3 ± 9.5 years) were recorded prior to PVI plus rotor ablation. Patients were latter classified based on their 6 months follow-up (23 sinus rhythm vs. 12 arrhythmia recurrence). ECGI dominant frequencies (DF) and rotor metrics were calculated in each patient, together with their temporal variability. Multivariate logistic regression was computed on the metrics, their variability and AF type to study the prediction power on ablation outcome. Results Dominant frequency metrics present higher values in patients with arrhythmia recurrence after the ablation. Rotor metrics showed higher variability in patients with bad PVI outcome (Table 1). ECGI metrics alone allow better outcome prediction (Fig. 1) than standard AF classification (AUC = 0.85 vs. 0.65). ECGI metrics combined with the type of AF increased outcome prediction, AUC = 0.88. Conclusion Patients with poor ablation outcome presented higher DFs and reentrant activity variability than patients with good PVI outcome. ECGI derived metrics in combination with AF classification increase PVI follow-up prediction compared to traditional patient stratification showing potential to be an alternative to the standard of care.