Christophe Tribouilloy1, Yohann Bohbot2, Sylvestre Maréchaux2, Nicolas Debry2, Quentin Delpierre2, Marcel Peltier2, Momar Diouf2, Michel Slama2, David Messika-Zeitoun2, Dan Rusinaru2. 1. From the Department of Cardiology (C.T., Y.B., Q.D., M.P., D.R.), Division of Clinical Research and Innovation (M.D.), and Intensive Care Unit, Department of Nephrology (M.S.), University Hospital Amiens, France; INSERM U-1088, Jules Verne University of Picardie, Amiens, France (C.T., S.M., M.S., D.R.); Groupement des Hôpitaux de l'Institut Catholique de Lille/Faculté libre de médecine, Université Lille Nord de France (S.M.); Department of Cardiology, Centre Hospitalier Universitaire de Lille, F-59000, France (N.D.) and Department of Cardiology, Cardiovascular Division, AP-HP, Bichat Hospital, Paris, France (D.M.-Z.). tribouilloy.christophe@chu-amiens.fr. 2. From the Department of Cardiology (C.T., Y.B., Q.D., M.P., D.R.), Division of Clinical Research and Innovation (M.D.), and Intensive Care Unit, Department of Nephrology (M.S.), University Hospital Amiens, France; INSERM U-1088, Jules Verne University of Picardie, Amiens, France (C.T., S.M., M.S., D.R.); Groupement des Hôpitaux de l'Institut Catholique de Lille/Faculté libre de médecine, Université Lille Nord de France (S.M.); Department of Cardiology, Centre Hospitalier Universitaire de Lille, F-59000, France (N.D.) and Department of Cardiology, Cardiovascular Division, AP-HP, Bichat Hospital, Paris, France (D.M.-Z.).
Abstract
BACKGROUND: Current guidelines define severe aortic stenosis in patients with aortic valve area normalized to body surface area (AVA/BSA) <0.6 cm2/m2; yet, this cutoff has never been validated. Moreover, it is not known whether AVA normalization to other body size indexes allows improved outcome prediction. We aim to test the value of AVA normalized to body size for outcome prediction in asymptomatic aortic stenosis. METHODS AND RESULTS: We included 289 patients with asymptomatic aortic stenosis, preserved ejection fraction, and AVA<1.3 cm2 at diagnosis. The outcome measure was the occurrence of aortic valve replacement or all-cause death or during follow-up. AVA was normalized to BSA, height, weight, and body mass index. For each normalized index, patients in the lowest tertile were at high risk of events whereas outcome was similar for the other tertiles. High risk of events was observed with AVA/BSA <0.4 cm2/m2 (adjusted hazard ratio [HR], 3.42 [2.09-5.60]), AVA/height <0.45 cm2/m (adjusted HR, 3.99 [2.42-6.60]), AVA/weight <0.01 cm2/kg (adjusted HR, 3.37 [2.07-5.49]), and AVA/body mass index <0.029 cm2/kg per meter square (adjusted HR, 3.23 [1.99-5.24]). Mortality risk was high with AVA/height <0.45 cm2/m (adjusted HR, 2.18 [1.28-3.71]), followed by AVA/BSA <0.40 cm2/m2 (adjusted HR, 1.84 [1.09-3.11]), AVA/weight <0.01 cm2/kg (adjusted HR, 1.78 [1.07-2.98]), and AVA/body mass index <0.029 cm2/kg per meter square (adjusted HR, 1.75 [1.04-2.93]). AVA/height showed better predictive performance than AVA/BSA with improved reclassification and better discrimination (net reclassification improvement: 0.33 versus 0.28; integrated discrimination improvement: 0.10 versus 0.08; C statistic: 0.67 versus 0.65), whereas AVA/weight and AVA/body mass index showed lower predictive capacity. CONCLUSIONS: Among AVA normalization methods, AVA/height <0.45 cm2/m followed by AVA/BSA <0.40 cm2/m2 seem as robust parameters for defining high risk in asymptomatic aortic stenosis. The prognostic value of AVA/height deserves future research.
BACKGROUND: Current guidelines define severe aortic stenosis in patients with aortic valve area normalized to body surface area (AVA/BSA) <0.6 cm2/m2; yet, this cutoff has never been validated. Moreover, it is not known whether AVA normalization to other body size indexes allows improved outcome prediction. We aim to test the value of AVA normalized to body size for outcome prediction in asymptomatic aortic stenosis. METHODS AND RESULTS: We included 289 patients with asymptomatic aortic stenosis, preserved ejection fraction, and AVA<1.3 cm2 at diagnosis. The outcome measure was the occurrence of aortic valve replacement or all-cause death or during follow-up. AVA was normalized to BSA, height, weight, and body mass index. For each normalized index, patients in the lowest tertile were at high risk of events whereas outcome was similar for the other tertiles. High risk of events was observed with AVA/BSA <0.4 cm2/m2 (adjusted hazard ratio [HR], 3.42 [2.09-5.60]), AVA/height <0.45 cm2/m (adjusted HR, 3.99 [2.42-6.60]), AVA/weight <0.01 cm2/kg (adjusted HR, 3.37 [2.07-5.49]), and AVA/body mass index <0.029 cm2/kg per meter square (adjusted HR, 3.23 [1.99-5.24]). Mortality risk was high with AVA/height <0.45 cm2/m (adjusted HR, 2.18 [1.28-3.71]), followed by AVA/BSA <0.40 cm2/m2 (adjusted HR, 1.84 [1.09-3.11]), AVA/weight <0.01 cm2/kg (adjusted HR, 1.78 [1.07-2.98]), and AVA/body mass index <0.029 cm2/kg per meter square (adjusted HR, 1.75 [1.04-2.93]). AVA/height showed better predictive performance than AVA/BSA with improved reclassification and better discrimination (net reclassification improvement: 0.33 versus 0.28; integrated discrimination improvement: 0.10 versus 0.08; C statistic: 0.67 versus 0.65), whereas AVA/weight and AVA/body mass index showed lower predictive capacity. CONCLUSIONS: Among AVA normalization methods, AVA/height <0.45 cm2/m followed by AVA/BSA <0.40 cm2/m2 seem as robust parameters for defining high risk in asymptomatic aortic stenosis. The prognostic value of AVA/height deserves future research.