Cardiac Output Reserve Predicts Exercise Capacity and Need for Surgery in Aortic Stenosis

Background: The development of symptoms in patients with aortic stenosis (AS) relates inconsistently with echocardiographic measures of stenosis severity. This is a reflection of the dynamic interaction between the valve and ventricle on exercise, which is not captured by any single parameter. We hypothesised that cardiac output reserve (COR), the ratio of cardiac output at peak exercise to cardiac output at rest, would predict exercise capacity (EC) and the need for aortic valve surgery. Methods: 40 patients with moderate-severe AS, were prospectively recruited. All patients underwent resting transthoracic echocardiography, modified Bruce exercise treadmill testing (ETT), B-type natriuretic peptide measurement and bicycle exercise stress echocardiography. EC was defined by exercise duration during the ETT. The decision to refer the patient for surgery was made by the specialist valve team who had access to all data but were blinded to COR results. Regression analysis was performed using independent variables shown to correlate with each outcome measure on univariate analysis. Results: The mean±SD of the EC was 645±294s, aortic valve area (AVA) 0.92±0.28cm 2 and COR 1.86±0.39. 22 of the 40 patients were referred for surgery. Exercise Capacity: On univariate analysis, resting AVA or mean/peak pressure gradient (AVG) did not correlate with EC. EC significantly correlated with COR (r = 0.71, p < 0.000001), age (r = -0.63, p < 0.00005), BNP (r = -0.49, p = 0.03), peak AVG at maximal exercise (peak AVG ex ) (r = 0.32, p = 0.04) and tissue Doppler lateral mitral valve annulus systolic velocity, Sa (r = 0.32, p = 0.05). By multiple linear regression, only COR (β = 0.5, p = 0.0004) and peak AVG ex (β = 0.3, p = 0.09) were predictive of EC. Referral for Surgery: By ROC analysis, a COR of 1.8 (AUC 0.87, 95% CI 0.77-0.98) best predicted the need for surgery. Energy loss index at maximal exercise, meanAVG at rest and AVA at maximal exercise were also predictors of need for surgery. By binary logistic regression, only COR was predictive of the need for surgery (OR 0.02, 95% CI 0.00001-0.53, p = 0.03) Conclusion: Cardiac Output Reserve is a novel index that predicts exercise capacity and the need for surgery in moderate to severe AS.