Predictive factors of recurrent atrial tachyarrhythmia after multiple procedures of radiofrequency catheter ablation for paroxysmal atrial fibrillation: Kansai Plus Atrial Fibrillation Registry (KPAF)

Abstract

Abstract Background The predictors of arrhythmia recurrence after radiofrequency catheter ablation (RFCA) for paroxysmal atrial fibrillation (PAF) have not yet been fully evaluated. Purpose The aim of this study was to develop and validate a risk scoring system to predict the incidence of recurrence of atrial tachyarrhythmia after the final RFCA for PAF. Methods The study population consisted of 3223 consecutive patients undergoing first-time RFCA for PAF from November 2011 to March 2014 in 26 cardiovascular centers in Japan who were enrolled in the Kansai Plus Atrial Fibrillation (KPAF) registry. We developed a scoring system in a derivation cohort with 2149 patients and assessed its reproducibility in a validation cohort with 1074 patients. The primary endpoint was recurrent atrial tachyarrhythmia lasting for ≥30 seconds after 91 days post the final ablation. Results During a median follow-up period of 3.1 years, 404 (18.8%) patients of the derivation cohort had AF recurrence after the final RFCA. The baseline patient characteristics of the derivation cohort were as follows: mean age 64.7 years, male 1480 (68.9%), mean body mass index (BMI) 23.6 kg/m2, hypertension 1122 (52.2%), prior heart failure 182 (8.5%), diabetes mellitus 203 (9.5%), prior stroke and/or transient ischemic attack 21 (1.0%), prior vascular disease 209 (9.7%), prior valvular disease 105 (4.9%), median CHADS2 score 1.1, median CHA2DS2-VASc score 2.1, mean number of ineffective antiarrhythmic drugs (AAD) 0.80, median duration of history of AF episodes 2.1 years, mean left atrial diameter (LAD) 38.2 mm, mean left ventricular ejection fraction (LVEF) 65.3%, and mean eGFR 68.7 mL/min/1.73m2. There was no significant difference in the baseline characteristics between derivation and validation cohorts. The results of the multivariate logistic regression models identified 5 independent variables of recurrent atrial tachyarrhythmia after the final RFCA: female (odds ratio (OR) = 1.45, p=0.0017), BMI <25 kg/m2 (OR=1.40, p=0.0081), duration of AF history 3 years≤ (OR=1.39, p<0.0034), chronic kidney disease (CKD) (OR=2.1, p=0.005, for stage 2/3CKD, OR=2.6, p=0.018 for stage 4/5 CKD), and LVEF (OR=2.1, p=0.039 for LVEF <50%, OR=1.5, p=0.022 for LVEF 50–60%). The predictive score for each factor was 3 points for CKD stage 4/5, 2 for CKD stage2/3 and LVEF <50%, and 1for the others (11 points in total). The arrhythmia-free rates after the final RCFA in the derivation cohort according to the score were as follows: 0–2 points = 91.7%, 3–4 = 80.7%, 5< = 72.6%, respectively. The similar results were reproduced in the validation cohort (Figure 1). Conclusion Our newly developed scoring system, composed of female, BMI, AF duration, CKD, and LVEF, could reproducibly predict arrhythmia recurrence after the final RFCA for PAF. Funding Acknowledgement Type of funding source: None