A novel approach to quantification of aortic valve calcifications in patients undergoing transcatheter aortic valve implantation.

Background: Precise calcium evaluation in the aortic complex may be complicated. We aimed to assess the usefulness of a novel semi-automatic algorithm for multi slice computed tomography-derived (MSCT) quantitative estimation of aortic valve calcifications (AVC) in patients with severe aortic stenosis (AS) undergoing transcatheter aortic valve implantation (TAVI).

Methods: Ten patients with severe AS who underwent TAVI with Edwards Sapien S3 26 mm prosthesis and had a pre-procedural MSCT scan were included. Data on baseline characteristics, procedural and long-term outcomes were collected prospectively. Pre-procedural MSCT data were used for AVC evaluation with 3D modeling (calcium volume, thickness, area, density, and distribution) in a dedicated program.

Results: Mean calcium thickness was 4.6 (3.6-5.8) mm. Median calcium are 333.6 (274.7-386.7) mm2. We found a significant correlation between larger maximal calcium layer thickness and PVL occurrence after TAVI (P=0.039). The radial representation of the calcium distribution allowed to divide aortic valve into 3 zones and to compare each zone to parallel zone on TTE images. In zones with PVL ≥2 mean AVC was higher than in zones with PVL <2 (7354.6±4020.4 pixels vs. 4325.1±1790.6 pixels; P=0.018). Based on ROC analysis, the optimal cut-off value of AVC to predict PVL ≥2 was >6506 pixels with 57.1% sensitivity and 90.5% specificity (AUC 0.762 [95% CI: 0.564 to 0.901], P=0.029).

Conclusions: Multiplane AVC quantitative evaluation provided details on total calcium amount, pattern and distribution in aortic valve. Established AVC parameters allowed better visualization of an operating area and prediction of PVL after TAVI.