Three-dimensional versus two-dimensional transthoracic echocardiography for left ventricular outflow tract measurements in severe aortic stenosis. A cross-sectional study using computer tomography and Haegar sizers as reference.

Objectives. In grading of aortic stenosis, two-dimensional transthoracic echocardiography (2D TTE) routinely results in underestimation of the left ventricular outflow tract (LVOT) area, and hence the aortic valve area (AVA). We investigated whether three-dimensional (3D) TTE measurements of the LVOT would be more accurate. We evaluated the feasibility, agreement and inter-observer variability of 3D TTE LVOT measurements with computed tomography (CT) and Haegar sizers as reference. Design. Sixty-one patients with severe aortic stenosis were examined with 2D and 3D TTE. 41 had CT and 13 also had perioperative Haegar sizing. Pearson's correlation and Bland-Altman plots were used to compare methods. Inter-observer variability was tested for 2D and 3D TTE. Trial registration: Current research information system in Norway (CRISTIN). Id: 555249. Results. Feasibility was 67% with 3D TTE and 100% with 2D TTE and CT. Mean LVOT area for 2D, 3D, CT and Haegar sizers were 3.7 +/- 0.6 cm2, 4.0 +/- 0.9 cm2, 5.2 +/- 0.8 cm2 and 4.4 +/- 1.0 cm2 respectively. Bias and limits of agreements for 2D TTE was 1.5 +/- 1.3 cm2, compared with CT and 0.4 +/- 1.5 cm2 with Haegar sizers. Corresponding results for 3D TTE were 1.2 +/- 1.6 cm2 and 0.2 +/- 1.8 cm2. Intraclass correlation coefficients for LVOT area were 0.62 for 3D and 0.86 for 2D. Conclusions. 2D TTE showed better feasibility and inter-observer variability in measurements of LVOT than 3D TTE. Both echocardiographic methods underestimated LVOT area compared to CT and Haegar sizers. These observations suggest that 2D TTE is still preferable to 3D TTE in the assessment of aortic stenosis.