P751Potential clinical utility of intraprocedural three-dimensional integrated image guided transcatheter aortic valve implantation using novel automated computed tomography software

Abstract

Introduction: Multidetector computed tomography (MDCT) has an essential role in the preprocedural assessment of aortic annulus (AA) anatomy for transcatheter aortic valve implantation (TAVI). However, preceding MDCT software cannot be applied to intraprocedural assessment of TAVI. Novel automated CT software Valve Assist (GE healthcare) has been recently developed for TAVI procedure, which. not only provides 3D reconstruction of MDCT, but also enables intraprocedural real-time fusion of fluoroscopic and MDCT images. Purpose: In this study, we reported the preliminary experience of this novel integrated 3D image-guided TAVI. The aim of this study is to clarify the reproducibility and accuracy of Valve Assist in the AA assessment and verify the potential of intraprocedural integrated MDCT images for TAVI. Methods: We examined 50 patients with severe aortic stenosis underwent transfemoral TAVI. Before TAVI, patients underwent cardiac CT angiography. AA measurements were performed using 3mensio Structural Heart software (3mensio Structural Heart), and Valve Assist. For the intraprocedural imaging, 3D reconstruction images of the MDCT aortography was overlapped with fluoroscopic aortography using Valve Assist (Figure A) to identify the aortic root anatomy including coronary orifices (in green), distribution of calcification (in blue), and virtual AA plane (in yellow) during the TAVI procedure. Results: Transcatheter heart valve was successfully implanted in all cases. Sapien 3 (Edwards Lifesciences) was implanted in 45 patients and Evolut R (Medtronic Inc) was implanted in 4 patients. Lotus was implanted in 1 patient. Valve Assist provided excellent reproducibility for the measurement of AA area (intraobserver intraclass correlation coefficients (ICC) 0.959 [95\% confidence interval (CI) 0.929-0.976], interobserver ICC 0.941 [95\% CI 0.899-0.966]), and perimeter (intraobserver ICC 0.915 (95\% CI 0.856-0.951), interobserver ICC 0.912 [95\% CI 0.850-0.949]). Excellent correlation was found between Valve Assist and 3mensio (ICC (0.952 [95\% CI 0.917-0.972]) for AA area, and perimeter (ICC 0.923 [95\% CI 0.869-0.955]). Correlation figure and Bland-Altman plot also indicated the good correlation between Valve Assist and 3mensio for AA area. Intraprocedural fusion image of CT aortography and fluoroscopic aortic root aortography generated by Valve Assist identified the position of coronary orifices (green circles) and the distribution of aortic valve (in blue) during the transcatheter heart valve (THV) positioning (Figure B). Fusion image displayed the position of coronary orifices after the device implantation (Figure C). Conclusion: Valve Assist showed excellent reproducibility and accuracy in the assessment of aortic root anatomy prior to TAVI procedure. Furthermore, the integrated fusion image might have a potential as an intraprocedural imaging modality for TAVI to contribute to the development of a safer procedure. (Figure Presented).