Exercise echocardiographic determinants of prognosis in aortic stenosis

Abstract Introduction A variation of 20 mmHg in the mean aortic gradient (MAG) in asymptomatic aortic stenosis (AS) was initially proposed as an indicator of severity during exercise echocardiography. Objective to identify exercise echocardiographic predictors of prognosis in AS. Our primary endpoint was a combined outcome of cardiac cause of death or aortic valve intervention. Methods prospective study of patients (pts) referred for testing with asymptomatic AS. We performed exercise echocardiography using the modified Bruce Protocol. Baseline, peak and recovery imaging was performed. Exercise tolerance was expressed as METS and % of predicted METS for age and sex. Population: 23 pts, 13 males, age 72+-9 years, body mass index (BMI): 28+-4 Kg/m2, etiology: bicuspid 5 pts, degenerative tricuspid 18 pts, 19 pts were in sinus rhythm, 4 in atrial fibrillation. Results baseline parameters: indexed aortic valve area (AVAi): 0,51+-0,13 cm2/m2, MAG 37+-12 mmHg, left ventricle: ejection fraction (LVEF): 63+-9%, cardiac index L/min/m2(CI) 3,2+-0,6, global longitudinal strain (LVGLS) -16+-3%, VTI ratio: 0,26+-0,05, valvulo-arterial impedance (Zva): 3,9+-0,9 mmHg.ml-1.m2. Peak parameters: AVAi: 0,54+-0,14 cm2/m2, MAG 53+-17 mmHg, LVEF: 69 (IQR 10)%, CI 3,2+-0,6, LVGLS -19+-5%, VTI ratio: 0,28+-0,06, Zva: 4,3+-1,4. Exercise tolerance: METS 6+-2, representing 86+-26% of the predicted exercise tolerance for age and sex. During a mean follow-up of 32+-20 months, 9 deaths occurred (causes: 2 sudden deaths, 6 noncardiac, 1 unknown), 5 patients were submitted to percutaneous valve implantation and 10 pts to surgical aortic valve replacement. Due to small sample size we applied bootstrapping (1000 samples). Patients who reached the endpoint had a higher MAG at peak exercise (57+-17 vs 39 +-17 mmHg, p=0.04, effect size Hedges’ g 0.99) and higher variation of MAG with exercise (20+-9 vs 6+-10 mmHg p=0.008, effect size 1.41). We found no significant differences in the other echocardiographic variables studied. The ROC analysis for peak MAG showed an area under the curve of 0,74 (CI: 0,53-0.95) and the best cutoff 51 mmHg and for MAG variation an AUC 0,85 (CI 0.66-1.04) for a cutoff of 8.4 mmHg. The Kaplan Meyer analysis showed a longer event free survival time in patients with peak MAG inferior to 51 mmHg (42+-9 vs 32+-7 months). The Cox regression analysis identified the best predictor the MAG variation of 8.4 mmHg with exercise with a hazard ratio of 21 (CI 22-29, p=0.002). Conclusions in exercise echocardiography the main prognostic predictors were the peak mean aortic gradient and the mean aortic gradient variation with exercise, which was, in this cohort, inferior to previously reported.