This study compared image quality and evaluability of coronary artery disease (CAD) in routine preparatory imaging for transcatheter aortic valve replacement using 64-slice (first-generation) to 192-slice (third-generation) dual-source computed tomography(DSCT).
Materials and Methods:
The CT data sets of 192 patients (122 women, median age 82 y) without severe renal dysfunction or known CAD were analyzed retrospectively. Half were examined using first-generation DSCT (June 2014 to February 2016) and the other half with third-generation DSCT (April 2016 to April 2017). Per protocol, contrast material (110 [110 to 120] vs. 70 [70 to 70] mL, P<0.001) and radiation dose of multiphasic retrospectively gated thoracic CT angiography (dose-length-product, 1001 [707 to 1312] vs. 727 [474 to 1369] mGy×cm, P<0.001) were significantly lower with third-generation DSCT. Significant CAD was defined as CAD-RADS ≥4 by CT. Invasive coronary angiography served as the reference standard (CAD is defined as ≥70% stenosis or fractional flow reserve ≤0.80).
In comparison with first-generation DSCT, third-generation DSCT showed significantly better subjective (3 [interquartile range 2 to 3] vs. 4 [3 to 4.25] on a 5-point scale, P<0.001) and objective image quality (signal-to-noise ratio of left coronary artery 12.8 [9.9 to 16.4] vs. 15.2 [12.4 to 19.0], P<0.001). Accuracy (72.9% vs. 91.7%, P=0.001), specificity (59.7% vs. 88.3%, P<0.001), positive (61.0% vs. 83.3%, P<0.001), and negative predictive value (91.9% vs. 98.2%, P=0.045) for detecting CAD per-patient were significantly better using third-generation DSCT, while sensitivity was similar (92.3% vs. 97.2%, P=0.129).
Coronary artery evaluation with CT angiography before TAVI is feasible in selected patients. Compared with first-generation DSCT, state-of-the-art third-generation DSCT technology is superior for this purpose, allowing for less contrast medium and radiation dose while providing better image quality and improved diagnostic performance.