Introduction: Tri-dimensional imaging techniques such as CT angiography (CT) and 3D transesophageal echocardiography (3D TEE) have been used to study the anatomy, size and calcification of the aortic annulus before transcatheter aortic valve implantation (TAVI).
Aim: To compare the measurements and agreement of different imaging modalities in valve selection: CT (using two different softwares, the Osirix 32bit version and 3Mension (3M) 2.0 version) and 3D TEE (Philips - 3DQ - MPR).
Methods: Single centre retrospective analysis of 20 patients in whom both CT and 3D TEE were used to study the annulus before TAVI. Three independent observers were blinded to each other and to the type of implanted prosthesis (16 auto-expandable and 4 balloon expandable). The following annulus variables were compared: orthogonal diameters in the sagittal and coronal planes, area, perimeter and mean diameter. The correlation analysis between the different methods was performed using the Spearman coefficient. The inter-method measurement differences were assessed by the Wilcoxon and Sign tests.
Agreement was assessed by the construction of Bland-Altman plots.
Results: There was a good correlation between CT measurements using Osirix and 3M (r >.74 p < 0.02). The correlation was moderate between the measurements of CT and 3D TEE (r 0.46/0.66 for sagittal diameter/area and intermediate in the remaining variables p <0,05) and it was slightly superior for the 3M measurements. The 3D TEE measurements were significantly inferior than the CT ones, with the exceptions for the sagittal diameter in all comparisons (p<0.142) and for the coronal diameter when compared to the 3M results (p=0.064) (Table)
Conclusion: The aortic annular diameters and areas measured by CT (Osirix, 3M) and 3D TEE (MPR) showed a fair correlation. Measurements by 3D TEE were significantly inferior, except for sagittal diameter (measured by CT- 3M and 3D TEE – MPR), which was the most consistent variable among the three modalities.