Abstract 2308: Conventional Aortic Valve Replacement in Very Elderly Patients

Suboptimal early and late results of percutaneous AVR have been justified by its use in patients claimed to be at high risk for conventional surgery, particularly the very elderly. We performed an analysis to identify which risk factors in conventional AVR patients are most predictive of mortality and to quantify those risk factors in order to provide a realistic baseline for comparison. We analyzed the outcome of surgery in 1168 patients operated on over a 31 year period on one service. Of those patients, 50 suffered perioperative mortality. Of those, 54% (27/50) were male vs. 62% (699/1118) who did not suffer perioperative mortality and 32% (16/50) vs 22% (242/1118) were >76 yrs old; 21% (9/43) vs. 25% (241/982) had diabetes, and 9.3% (9/43) vs. 5.5 (54/982) had bacterial endocarditis (all p=NS); 22% (11/50) vs. 11% (122/1118) had a prior MI (p=0.0152); 26% (13/50) vs 11% (118/1118) had a prior CAB (p=0.0007); 46% (23/49) vs 29% (322/1118) had a concurrent CAB (p=0.0063); mean EF of 44.50±18.26 vs. 53.33±13.90 (p<0.0001); and mean replacement valve size of 20.69±2.14 vs. 21.75±2.46 (p=0.0031). Logistic regression analysis identified prior CAB (odds ratio (OR) 1.65. p<0.01), concurrent CAB (OR 1.63, p<0.01), bacterial endocarditis (OR 1.85, p<0.05) and replacement valve size (OR 0.81, p<0.01 – larger size is protective) as predictors of perioperative mortality. Age was not a predictor of perioperative mortality. Cox regression analysis for factors predictive of overall mortality identified age (relative risk (RR)1.046, 4.6%/yr, p<0.0001) preop EF (RR 0.982, decrease risk of 1.8%/1% increase EF, p<0.0001), diabetes (RR 1.254, p=0.0031, 25% increase risk with diabetes), and replacement valve size (RR 0.889, p=0.0004, 11.1% decreased risk/mm valve size.) These data suggest that even in high-risk patients, perioperative mortality is relatively low and is not predicted by age alone; therefore conventional surgery should be seriously considered in almost all patients. Age is only one of several risk factors which should be evaluated.