Software-automated multidetector computed tomography-based prosthesis-sizing in transcatheter aortic valve replacement: Inter-vendor comparison and relation to patient outcome

Background Correct sizing of the aortic annulus and determination of coronary height is crucial for the efficacy and safety of trans-catheter aortic valve replacement (TAVR). We aimed to compare 3 different software solutions and manual measurements with respect to MDCT-derived annulus sizing parameters, reproducibility and processing time for TAVR planning and to evaluate the impact on patients' outcome. Methods One-hundred-seventeen patients with severe aortic stenosis prior to TAVR were included. Prosthesis sizes were determined separately using 3 different platforms: i) 3mensio, ii) IntelliSpacePortal V7 (ISP7), and iii) ISP9. In addition, manual measurements were performed. Intra- and inter-observer reproducibility as well as systematic differences were assessed. Finally, hypothetical prosthesis sizing was related to patient outcome using 3mensio as the clinical reference standard. Results Intra- and inter-observer reproducibility were excellent for all annulus parameters (ICCs 0.787–0.994). Distances to coronary ostia showed excellent intra-observer reproducibility except for 3mensio (ICC LCO: 0.655; RCO: 0.020), while inter-observer reproducibility was poor except for ISP9. For annulus parameters, slightly larger values were observed for ISP9 when compared to the other platforms. ISP9 allowed for the fastest and most reproducible measurement of aortic annulus parameters (although it did result in slightly larger annulus sizing). Conclusions Reproducible measurement of aortic annulus parameters can be obtained by all investigated post-processing solutions. All post-processing platforms are likely to result in similar and reliable measurements and thus can be used interchangeably.