Aim: High-risk patients are currently being evaluated for various catheter-based aortic valve replacement (AVR) techniques. To identify an individual patient's risk, scores such as the EuroSCORE or STS risk calculator (RC) are used. The aim of the present study was to evaluate the surgical results in patients who underwent isolated AVR via a median re-sternotomy after prior CABG.
Patients and methods: Between 01/96 and 01/08, 349 patients underwent AVR as a redo procedure. One hundred and thirty patients had undergone previous CABG; in 39 patients (29 male, median age 75 (60-84)) preoperative coronary angiography revealed open grafts with no need for additional revascularization (30 had LIMA grafts). These patients underwent isolated AVR. Operative mortality was calculated using the standard and logistic EuroSCORE, and the STS RC.
Results: Operative (30-day mortality) was 5% (2 patients). Mean calculated predicted mortality rates for the cohort were: 12+/-3% for the standard, and 32+/-21% for the logistic EuroSCORE, and 10+/-4% according to the STS RC. Receiver operated characteristics (ROC) analysis revealed 100% specificity for standard EuroSCOREs up to 12.5%, logistic EuroSCOREs up to 39.7% and up to 17.45% for STS RC, with a sensitivity of 69.5%, 75% and 97.2%, respectively. The STS RC showed significant better prediction of mortality than the EuroSCOREs (p=0.006).
Conclusions: Conventional AVR as a redo procedure after CABG with patent grafts can be performed with excellent results and lower mortality than estimated. Results of newer catheter-based AVR approaches should not to be compared with artificial scores to justify high morbidity rates.