Predictors of short and long-term mortality after surgical aortic valve replacement in 681 octogenarians: a monocentric experience

Abstract

Abstract Introduction In spite of potential frailty, very elderly patients are increasingly referred for surgical aortic valve replacement (SAVR), even after the introduction of transcatheter aortic valve implantation. Purpose What are the predictors for short and long-term mortality after surgical aortic valve replacement in octogenarians and are these avoidable? Patients and methods This is a retrospective file study of 861 octogenarians, who underwent SAVR with or without associated procedures between 1987 to 2017. Mechanical valves and valves in another position were excluded. Age, gender, cardiac and non-cardiac comorbid conditions (renal, pulmonary, diabetes, malignancy), operative data (cross-clamp and bypass time, associated procedures) and postoperative adverse events were investigated for their effect on 30-day and on long-time mortality. Significant factors in an univariate analysis (chi-square and Kaplan-Meier) were entered in a multivariate model (logistic regression, Cox' proportional hazard analysis). Results Hospital mortality was 68/681 (10%). The follow-up was 3817 years, with a mean of 6.2 years for the remaining 613 patients. Predictors (Odds Ratio, p-value) for hospital mortality are: – Need for SAVR <24h, 3.87 (p=0.001) – Pulmonary disease, 2.44 (p=0.002) – Atrial fibrillation, 2.30 (p=0.004) – Age >85 years, 2.02 (p=0.021) – Renal disease, 1.94 (p=0.025) – Coronary artery disease, 1.99 (p=0.046) Five-year survival is 66.4±1.9% (392 patients at risk); ten-year survival is 25.3±2.0 (89 patients at risk), 15-year survival is 4.0±1.3% (6 patients at risk). Its predictors (Odds Ratio, p-value) are: – Renal disease, 1.40 (p=0.005) – Congestive heart failure, 1.37 (p=0.006) – Postoperative Low Cardiac Output Syndrome, 1.82 (p=0.007) – Postoperative delirium, 1.45 (p=0.010) – Postoperative ventricular arrhythmia 1.84, (p=0.014) – Need for SAVR <24 h, 1.69 (p=0.021) – Diabetes mellitus, 1.30 (p=0.032) – Malignancy, 1.30 (p=0.046) – Pulmonary disease, 1.26 (p=0.047) – Preoperative carotid artery endarterectomy, 1.49 (p=0.052) Conclusions In this elderly population, short and long-term mortality are acceptable. The need for urgent SAVR (<24h) is the dominant predictor for 30-day mortality. Of the ten predictors for long-term survival, seven are of preoperative nature. Two postoperative cardiac events as well as need for urgent SAVR have an important effect on long-term outcome. No operative variables were identified as predictor. Need for urgent SAVR should be avoided by early referral. Funding Acknowledgement Type of funding sources: None.