A novel survival prediction model after isolated surgical aortic valve replacement in the transcatheter aortic valve replacement era

OBJECTIVE: To examine the surgical outcomes of surgical aortic valve replacement in the transcatheter aortic valve replacement era and propose a novel patient-specific prognostic model. METHODS: We randomly divided 772 patients with aortic stenosis who underwent surgical aortic valve replacement in 2016-2021 into two cohorts (derivation, 515; validation, 257). In the derivation cohort, no data were missing for any patients for the candidate predictors including age, sex, body mass index, left ventricular ejection fraction, levels of albumin, hemoglobin, and serum creatinine, presence of chronic atrial fibrillation, and end-stage renal disease requiring hemodialysis. We developed possible scoring models using Cox proportional hazards regression with overall survival as the endpoint and calculated the cross-validated 5-year C-statistics to assess accuracy. RESULTS: The mean patient age was 74.2 years, and 46.9% were female. Kaplan-Meier analysis revealed overall 1- and 5-year survival rates of 96.6 and 88.7%, respectively. The 5-year C-statistic of the derivation cohort was 0.785 (95% confidence interval: 0.716-0.853), while the estimated 1-, 3-, and 5-year C-statistics of the validation cohort were 0.885 (0.806-0.965), 0.888 (0.824-0.953), and 0.801 (0.702-0.901), respectively. Calibration plots revealed good agreement between predicted and actual 5-year survival (intraclass correlation coefficient = 0.955; 95% confidence interval: 0.827-0.989). CONCLUSIONS: This novel survival prediction model after isolated surgical aortic valve replacement in the transcatheter aortic valve replacement era showed good survival prediction, and may guide the decision-making process for surgical aortic valve replacement versus transcatheter aortic valve replacement with lifetime management.