Aims: The accuracy of transcatheter aortic valve replacement (TAVR) sizing using three-dimensional transoesophageal echocardiography (3D-TEE) compared with the gold-standard multi-slice computed tomography (MSCT) remains unclear. We compare aortic annulus measurements assessed using these two imaging modalities.
Methods and results: We performed a single-centre prospective cohort study, including 53 consecutive patients undergoing TAVR, who had both MSCT and 3D-TEE for aortic annulus sizing. Aortic annular dimensions, expected transcatheter heart valve (THV) oversizing, and hypothetical valve size selection based on CT and TEE were compared. 3D-TEE and CT cross-sectional mean diameter (r = 0.69), perimeter (r = 0.70), and area (r = 0.67) were moderately to highly correlated (all P-values <0.0001). 3D-TEE-derived measurements were significantly smaller compared with MSCT: perimeter (68.6 ± 5.9 vs. 75.1 ± 5.7 mm, respectively; P < 0.0001); area (345.6 ± 64.5 vs. 426.9 ± 68.9 mm(2), respectively; P < 0.0001). The percentage difference between 3D-TEE and MSCT measurements was around 9%. Agreement between MSCT- and 3D-TEE-based THV sizing (perimeter) occurred in 44% of patients. Using the 3D-TEE perimeter annular measurements, up to 50% of patients would have received an inappropriate valve size according to manufacturer-recommended, area-derived sizing algorithms.
Conclusion: Aortic annulus measurements for pre-procedural TAVR assessment by 3D-TEE are significantly smaller than MSCT. In this study, such discrepancy would have resulted in up to 50% of all patients receiving the wrong THV size. 3D-TEE should be used for TAVR sizing, only when MSCT is not available or contraindicated. The clinical impact of this information requires further study.
Keywords: multi-slice computed tomography; three-dimensional (3D) transoesophageal echocardiography; transcatheter aortic valve replacement.
Published on behalf of the European Society of Cardiology. All rights reserved. © The Author 2015. For permissions please email: email@example.com.