Prevalence and correlates of low-flow state by 3-dimensional automated echocardiography in aortic stenosis with normal ejection fraction

Abstract

Abstract Funding Acknowledgements Type of funding sources: None. Introduction Measurement of left ventricular (LV) stroke volume normalized for the body surface area (SVi) is required in aortic stenosis (AS) to calculate aortic valve area (AVA) and determine flow status. However, the LV outflow tract in most patients is hourglass rather than cylindrical, leading to a significant overestimation of the low-flow state and AS severity. Therefore, an accurate LV volumetric approach may potentially outperform the standard linear method in SVi measurement. Aims To investigate the prevalence of low-flow state (SVi ≤35 ml/m2) in patients with AS and normal ejection fraction (EF) by automated machine learning 3D echocardiography based on an adaptive analytics algorithm. Materials and Methods Consecutive patients with AS (peak velocity>2.5 m/s) and EF>50% underwent Dynamic Heart Model (DHM) using the larger settings of the boundary detection sliders (end-diastolic position = 60/60; end-systolic position = 30/30). Exclusion criteria were > moderate aortic or mitral regurgitation, rheumatic valve disease or endocarditis, and previous valve repair or replacement. Results We included 68 patients (mean 75 ± 12 years; 60% men, mean AVA 1,12 ± 0,40 cm2). On DHM, 10/68 patients (15%) demonstrated low SVi. The LVEDVi, LVESVi, AVA, LV mass, and EFLA were significantly lesser in patients with low SVi (Table). Conclusions When quantified with the modern 3D volumetric method, the prevalence of low-flow state in patients with AS and normal EF may be significantly lower than previously reported. The phenotypic profile of patients with low SVi can be refined when new metrics by DHM are employed.