Impact of aortic root geometry and degree of aortic calcification on outcomes of patients undergoing TAVI procedure

TL;DR

This study introduces a new method to better measure aortic valve calcification in CT scans, improving predictions of post-TAVI complications like paravalvular leak.

Contribution

A novel method for quantifying aortic valve calcification using contrast-enhanced CT with variable iodine content and individual HU thresholds.

Findings

Calcification volume and cusp-specific calcification were independent predictors of postoperative paravalvular leak.

The new method improved differentiation of calcium from contrast agent, enhancing preoperative risk assessment.

37% of patients experienced trace to moderate paravalvular leak after TAVI.

Abstract

Adequate differentiation of calcifications in contrast-enhanced CT scans remains difficult to assess TAVI parameters. The size of the aortic leaflets has not been taken into account so far in present studies. The aim of our study was to establish a new method for optimized quantification of the aortic valve calcification degree in contrast-enhanced CT scans for better preoperative prediction of postoperative paravalvular leak after TAVI.

We retrospectively analyzed preoperative contrast-enhanced CT scans of patients who underwent TAVI in our institution between 2014 and 2017. Calcium volume was quantified by a method using contrast enhanced computer tomography (3mensio-Structural Heart-7.2 software) with different iodine contents for better discrimination of contrast agent from calcium and by an individually set Houndsfield Unit (HU) threshold with 50HU above the individually determined reference value. Calcium volume was correlated with surface area of each aortic cusp. Perioperative variables were analyzed. All patients (n = 150) with severe aortic stenosis were treated with TAVI implantation. Overall incidence of postoperative trace to moderate PVL was 37%. The amount of calcium correlated with the incidence of PVL. In a logistic regression analysis total volume of calcification (p = .032) as well as calcification of each aortic cusp (NC_p = .001; RC_p < .001; LC_p = .001) were independent predictors.

Calcification degree as well as its correlation with the surface area of each aortic cusp significantly influence incidence of PVL. Our new method improves preoperative quantification of the calcification degree by use of contrast agents with different iodine contents and thereby helps to improve patients’ outcomes.

Figures