Aortic Valve Replacement: Results and Predictors of Mortality From a Contemporary Series of 2256 Patients
What are contemporary outcomes associated with isolated aortic valve replacement (AVR) in a large European population?
A prospective regional surgical registry including two academic and four private hospitals was queried for patients undergoing isolated AVR between 2003-2007. Outcomes assessed included predictors of in-hospital and mid-term (1- and 3-year) mortality, early and mid-term survival in those ages >80 years versus <80 years of age, and the value of the EuroSCORE in predicting in-hospital mortality in high-risk surgical patients.
Mean age was 70.4 ± 11.2 years; 48.3% were female. Four hundred and thirty patients (19.1%) were >80 years. Severe symptomatic aortic stenosis (AS) was the most common indication for operation. Overall in-hospital mortality was 2.2%. On multivariate analysis, New York Heart Association (NYHA) class III-IV (odds ratio [OR], 1.9; 95% confidence interval [CI], 1.0-3.7), Canadian Cardiovascular Society III-IV (OR, 4.3; 95% CI, 1.8-10.2), dialysis (OR, 9.8; 95% CI, 2.4-47.5), and severe chronic obstructive pulmonary disease (OR, 2.6; 95% CI, 1.6-5.9) were predictive of in-hospital mortality, but age >80 years was not. Mean length of stay was 14.3 ± 8.5 days. Patients >80 years had more preoperative comorbidities, higher in-hospital mortality (3.7% vs. 1.8%; OR, 2.1; p = 0.014), postoperative complication rates, and longer length of stay (15.7 ± 8.2 vs. 14.0 ± 8.5 days, p < 0.001) than younger patients. Patients >80 years with a logistic EuroSCORE of >15% (n = 114 patients) had an actual in-hospital mortality of 7% compared to the predicted mortality of 22.4% (p < 0.001). Mid-term survival in this group was 82.3% at 3 years, was improved in those with better NYHA functional classification (class I-II), and those with NYHA class I-II had comparable survival to age- and gender-matched regional controls.
The overall in-hospital mortality of isolated AVR in this registry was significantly less than predicted by the EuroSCORE, both overall and in those >80 years of age, and is comparable to other recent surgical registry data. The presence of more severe heart failure symptoms preoperatively seems to impact mid-term survival. Older age (>80 years) did not predict in-hospital mortality.
The overestimation of operative risk using the logistic EuroSCORE in this population is particularly relevant in this era of expanding interest in, and availability of, transcatheter aortic valve replacement (TAVR) therapy. Among other factors, the EuroSCORE is used in ongoing clinical trials to assess operative risk and determine candidacy for TAVR. In light of these data, the authors suggest this may not be the most accurate tool for risk assessment in this population. Overall, these data support early referral for consideration of surgical intervention for severe AS in all patients, despite advanced age.
Keywords: Heart Valve Prosthesis, Postoperative Complications, Registries, Hospital Mortality, Canada, Heart Failure, Risk Assessment, New York
< Back to Listings