Predictors of aortic stenosis severity reclassification using an imaging data fusion method in patients referred for transcatheter aortic valve implantation

Abstract

Background: The use of imaging data fusion method (IDFM) with multislice computed tomography (MSCT) and two-dimensional transthoracic echocardiography (2D-TTE) in patients with aortic stenosis (AS) may result in reclassification of AS severity from severe to non-severe.

Aim: We sought to establish potential predictors of AS severity reclassification using the IDFM method.

Methods: A total of 54 high-risk patients (mean age 79 ± 7.9 years; 40.7% male) with severe AS by 2D-TTE (indexed aortic valve area [AVAi] < 0.6 cm2/m2), referred for transcatheter aortic valve implantation, were included in the analysis. AVAi was subsequently recalculated using IDFM by replacing 2D-TTE left ventricular outflow tract (LVOT) measurements with MSCT LVOT parameters.

Results: Imaging data fusion method reclassified 20.4% patients into the potentially non-severe AS group. In a multivariable model including clinical variables, reclassification to non-severe AS by IDFM was independently associated with younger age and diabetes mellitus (DM), (odds ratio [OR] 0.864; 95% confidence interval [CI] 0.76–0.99; p < 0.035 and OR 19.259; 95% CI 2.28–162.41; p < 0.007, respectively). In a multivariable analysis of echocardiographic variables, reclassification was associ­ated with higher LVOT velocity time integral and lower aortic mean gradient (OR 1.402; 95% CI 1.07–1.84; p < 0.014 and OR 0.858; 95%: CI 0.760–0.968; p < 0.013, respectively). In addition, 24.1% of patients were reallocated from low-flow (< 35 mL/m2) to normal-flow AS.

Conclusions: Imaging data fusion method reclassified a substantial proportion of patients with severe AS into a potentially moderate AS group and from a low-flow to a normal-flow AS group. Such regrouping calls for increased diagnostic prudence in AS patients, especially those with specific clinical and echocardiographic predictors of reclassification, such as DM or low aortic mean gradient.