Abstract
Background
Aortic stenosis (AS) grading is mainly based on aortic valve area (AVA) calculation by 2D transthoracic echocardiography (2D-TTE), using continuity equation (CE). However, 2D-TTE shows several limits, mainly due to left ventricular outflow tract (LOVT) underestimation. Different 3D imaging modalities have been proposed to overcome 2D-TTE limitations, including 3D transesophageal echocardiography manual and software measurements (3D-TEEm and 3D-TEEs) and multidetector computed tomography (MDCT). The AVA cut-off value generally used to define severe AS has been established and validated by outcome studies in which AVA was measured by 2D-TTE. This cut-off value cannot be directly extrapolated to the 3D-TEE combined approach that systematically measures larger LVOT compared with 2D-TTE.
Purpose.To evaluate the diagnostic accuracy of 3D transesophageal echocardiography manual and software measurements (3D-TEEm and 3D-TEEs) in AS grading, compared with multidetector computed tomography (MDCT) as gold standard, and to identify a new cut-off for AS severity assessment.
Methods
218 patients (81 ± 5.4 years, 54% male) with symptomatic normal-flow AS underwent 2D-TTE, 3D-TTEm, 3D-TEEs and MDCT within the same hospitalization. 3D-TEE LVOT reconstruction was performed manually and with semi-automated software (EchoPAC version 201). 3D-TEEm, 3D-TEEs and MDCT LVOT areas were combined with 2D-TTE Doppler parameters to calculate AVA by CE. Using Doppler parameters (Vmax >4 m/s and MPG >40 mmHg) to define AS severity, a receiving-operating curve (ROC) was calculated for AVA obtained with different 3D imaging modalities.
Results
There was a good correlation between both 3D-TEEm and 3D-TEEs and MDCT measurements (r = 0.800 and r = 0.814, respectively) and excellent agreement between 3D-TEEm and 3D-TEEs with minimum bias. 2D-TTE significantly underestimated AVA compared to 3D-TEEm, 3D-TEEs and MDCT. On the other hand, both 3D-TEEm and 3D-TEEs underestimated AVA compared to MDCT (mean AVA difference = 0.13 and =0.06 cm2, respectively). ROC curve analysis demonstrated 91% sensibility and 34% specificity for 2D-TTE AVA using a cut-off of 1 cm2 (AUC 0.732). For 3D-TEEm and 3D-TEEs, a 1 cm2cut-off resulted in 74% sensibility and 59% specificity, while a 1.2 cm2cut-off resulted in 91% sensibility and 31% specificity (AUC 0.715). MDCT showed 59% sensibility and 70% specificity using a 1 cm2 cut-off and 83% sensibility and 45% specificity using a 1.2 cm2 cut-off (AUC 0.708).
Conclusion
3D-TEE represents a valuable tool for AS grading using a combined approach incorporating 3D LVOT measurements and 2D Doppler parameters in the CE. Both 3D-TEEm and 3D-TEEs AVA measurements demonstrated good correlation with MDCT and excellent reproducibility. 3D-TEE measurements underestimate AVA compared to MDCT. Given the multiparametric assessment of AS severity, a 1.2 cm2 AVA cut-off could be considered to define AS severity with emerging 3D imaging modalities.
P755 Aortic valve area calculation by 3D transesophageal echocardiography: new insights in severity grading of aortic stenosis