A novel predictive model of recurrence after surgical ablation of valvular atrial fibrillation

Background: Radiofrequency ablation(RFA) has emerged as a common approach to treatment of atrial fibrillation(AF) coupled with valve surgery. However, risk factors were rarely reported on recurrence after RFA of valvular atrial fibrillation . Methods: Retrospective study of a single-center database was performed in patients underwent valvular surgery combined with radiofrequency ablation. Follow-up was conducted from 3 months to 1 year after procedure. AF recurrence was defined as any episode of atrial arrhythmia lasting more than 30 seconds. Risk factors were analysis using Logistic regression and a predictive model was established. Results: 3-month follow-up was completed in 319 patients, 195 patients (61.1%) had AF recurrence. During the 1-year follow-up, 74 /166 patients(44.6%) had AF recurrence. Compared with sinus rhythm maintenance group, overweight(p=0.022), high BMI(p=0.024), large preoperative LAD(p<0.001), small MI area(p=0.019), large MI diameter(p<0.001) and other liver function parameter like low ALT level(p=0.023), high total bilirubin(p=0.013) and direct bilirubin level(p=0.033) were significantly correlated with AF recurrence after RF. In multivariate analysis, BMI(OR 1.12, 95% CI 1.04-1.21, p=0.003), LAD(OR 1.49, 95% CI 1.09-2.07, p=0.015), MI area(OR 0.41, 95% CI 0.18-0.85, p=0.023), MI diameter(OR 2.45, 95% CI 1.16-5.34, p=0.021), ALT(OR 0.98, 95% CI 0.97-0.99, p=0.002) and eGFR(OR 1.01, 95% CI 1.09-1.02, p=0.042) were significantly independent risk predictors of AF early recurrence. A predictive model was created with AUC of 0.720. Conclusions: BMI、LAD、MI area、MI diameter、ALT and eGFR were risk predictors of AF early recurrence. We established a predictive model and it devote to preoperative clinical health management of individual with valvular AF.