Erroneous Measurement of the Aortic Annular Diameter Using 2-Dimensional Echocardiography Resulting in Inappropriate CoreValve Size Selection: A Retrospective Comparison With Multislice Computed Tomography

Study Questions:

Does annular sizing using transesophageal echocardiography (TEE) and computed tomography (CT) result in differences in valve size selection and paravalvular aortic regurgitation (PAR) in patients treated with transcatheter aortic valve replacement (TAVR)?


This retrospective study examined 157 patients with severe aortic stenosis treated with a self-expanding TAVR device (CoreValve) using TEE for valve sizing; retrospective analysis of CT images was used to determine the annular size based on CT, and annular sizes were compared between TEE and CT to determine how frequently CT would result in selection of a different size of valve. Further, the incidence of significant (≥ grade 2) PAR was compared between patients with or without appropriate oversizing on CT and TEE.


There were significant differences between the annular diameter by TEE and the diameter derived from the perimeter on CT (23.2 vs. 25.3 mm, p < 0.001), as well as percentage oversizing between TEE and CT (20.1 ± 8.2% vs. 10.4 ± 7.8%, p < 0.001). Valve sizing followed manufacturer recommendations in 81.0% of patients based on TEE, but only 51.0% by CT perimeter size (p < 0.001). The presence or absence of significant PAR could not be accurately discriminated by the percentage oversizing on TEE (19.0 ± 8.6% vs. 20.5 ± 8.1%, p = 0.32), although oversizing based on the CT annular perimeter demonstrated a significant difference between groups (6.2 ± 7.1% vs. 11.7 ± 7.5%, p < 0.001). On multivariable analysis, adherence to CT perimeter-based oversizing was associated with a reduction in significant PAR (odds ratio, 0.36; 95% confidence interval, 0.14-0.90; p = 0.03). Using receiver-operator curves to predict PAR based on adherence to oversizing criteria, CT annular perimeter (area under the curve [AUC] 0.65) was superior to TEE annular diameter (AUC 0.51).


TEE frequently underestimates annular size as compared to CT, and resulted in an incorrect valve size in approximately 50% of patients. Inadequate valve oversizing on retrospective review of CT images was associated with increased PAR.


Appropriate valve oversizing reduces PAR in patients treated with TAVR, and multiple studies have demonstrated that 3D imaging using methods such as CT result in larger annular sizes than 2D measurements by TEE, which is not surprising given the typically elliptical shape of the annulus and the 3D structure of the aortic valve and annulus. The results of this paper and prior studies demonstrate that 2D TEE is inadequate for valve sizing in TAVR—whether for a SAPIEN or CoreValve device—despite its use in early TAVR. This study did not address the use of other 3D techniques such as 3D TEE or magnetic resonance imaging, which are other potential methods to define the 3D anatomy of the aortic valve annulus.

Clinical Topics: Cardiac Surgery, Invasive Cardiovascular Angiography and Intervention, Noninvasive Imaging, Valvular Heart Disease, Aortic Surgery, Cardiac Surgery and VHD, Interventions and Imaging, Interventions and Structural Heart Disease, Computed Tomography, Echocardiography/Ultrasound, Magnetic Resonance Imaging, Nuclear Imaging

Keywords: Heart Valve Prosthesis, Aortic Valve, Multidetector Computed Tomography, Aortic Valve Stenosis, Confidence Intervals, Magnetic Resonance Imaging, Echocardiography, Aortic Valve Insufficiency

< Back to Listings