Incremental value of coronary computed tomography angiography-derived wall shear stress over stenosis grade in detecting significant coronary artery disease
Y B O Somsen, K Sadowski, P Siogkas, R W de Winter, R Hoek, A Nap, S A Kleijn, J W R Twisk, Fotiadis Di, P Knaapen, R S DriessenAbstract
Background
Wall shear stress (WSS) is a tangential hemodynamic force acting on the arterial endothelium and is a key modulator of early atherosclerosis development. Although WSS has traditionally been computed using invasive intracoronary imaging (e.g., intravascular ultrasound), recent advances in coronary computed tomography angiography (CCTA)-based computational fluid dynamics now enable non-invasive derivation of WSS. However, data on the incremental value of CCTA-WSS for detecting hemodynamically significant CAD beyond conventional CCTA-derived stenosis assessment are scarce.
Purpose
The present study aimed to investigate the incremental value of CCTA-WSS for the detection of significant CAD, the latter defined as an invasive fractional flow reserve (FFR) ≤0.80.
Methods
This post-hoc analysis of the PACIFIC-I trial considered 208 participants with suspected CAD who underwent routine 256-slice CCTA followed by invasive coronary angiography with three-vessel FFR interrogation. Using in-house arterial 3D reconstruction software and a dedicated commercial computational fluid dynamics package, patient-specific 3D coronary models were generated and segmented into equidistant 0.5-mm contours to derive cross-sectional mean and maximum WSS. For reader orientation, WSS is considered low at 0–1 Pascal (Pa), normal at 1–7 Pa, and elevated at >7 Pa [1]. CCTA stenosis grade was assessed and reported as a categorical variable (0%, 1–25%, 25–49%, 50–69%, 70–100%). Generalized estimating equations were applied to determine the incremental value of CCTA-WSS beyond stenosis grade. Model discrimination was assessed using receiver operating characteristic curves (AUC); AUCs were compared with DeLong’s test.
Results
CCTA-WSS was successfully computed in 170 patients (417 vessels; mean age 59 ± 9 years; 63% male); 68 (40%) subjects had significant CAD. On a per-vessel level, patients with significant CAD showed higher mean WSS compared to those with non-significant CAD (6.2 [4.6–9.3] vs 3.7 [2.5–5.6] Pa; p<0.01). Similarly, maximum vessel WSS was higher in the significant CAD group (40.6 [19.3–80.4] vs 16.9 [9.0–34.1] Pa; p<0.01). Figure 1 shows the CCTA-derived coronary models used for prediction. Adding WSS parameters to stenosis grade improved discrimination by increasing the AUC, suggesting incremental value beyond stenosis grade alone.
Conclusion
CCTA-derived WSS computation was feasible in the majority of patients and was higher in those with hemodynamically significant CAD. On a per-vessel level, mean and maximum WSS improved model discrimination beyond stenosis grade alone, albeit with a subtle increase in AUC. These findings support a potential adjunctive role for CCTA-derived WSS for the detection of significant CAD.Figure 1.ROC analysis