Abstract 13775: Novel Risk Model of Quantitatively Predicting Recurrence of Atrial Fibrillation After the Cox-Maze IV Procedure

Abstract

Introduction: This study developed a novel risk score model for quantitative prediction of the rate of atrial fibrillation (AF) recurrence after the bipolar radiofrequency Cox-maze IV procedure (CMP-IV) in patients with AF and chronic valvular disease. Methods: We enrolled 408 consecutive patients who underwent the CMP-IV concomitant with valve surgeries in our hospital between 2014 and 2021. We randomly divided the cohort into two groups. A ‘development cohort’for model generation (204 patients; 50%) and a ‘test cohort’for model validation (204 patients; 50%) were calculated by Kaplan-Meier survival curve analysis and receiver operating characteristic curve analysis. Results: The median follow-up was 3.5 (interquartile range: 0.9-6.8) years. The 1-, 3-, 5- and 6-year rates of freedom from AF recurrence in the entire cohort were 91.3%, 87.1%, 80.9% and 76.3%, respectively. Risk factors for AF recurrence examined by multivariate Cox regression analysis included preoperative AF duration >4.4 years, postoperative neutrophil/lymphocyte ratio >9.23, anteroposterior diameter of left atrium >51mm before discharge and AF rhythm at discharge(panel A). Points were assigned to each risk factor according to its hazard ratio. A novel risk score model was developed with a range up to 13 points(panel B). High score (≥7) predicted significantly higher rates of AF recurrence after the CMP-IV than low score (0-3) and medium score(4-6) did in both development cohort(panel C) and test cohort(panel D). The area under the receiver operating characteristic curve of the novel risk model score at 1-, 3- and 5-year was 0.72, 0.80 and 0.89 (95% confidence interval: 0.49-0.95, 0.70-0.90 and 0.81-0.96) respectively in the test cohort(panel E). Conclusions: Surgeons should be cautious to apply the CMP-IV in patients with high model score due to higher risk of AF recurrence.