Abstract 14556: Comparison of Endothelial Shear Stress and Lumen Area Measurement with CT Angiography vs Coronary Angiography/IVUS: Preliminary Results from the PREDICTION Study

Abstract

Background: The PREDICTION Study demonstrated that large plaque burden and low regional endothelial shear stress (ESS), derived from coronary angiography/intravascular ultrasound (IVUS) 3D reconstruction, are independent and additive predictors of plaque progression, but that these predictors occur in only about 6% of patients with an acute coronary syndrome. It would be invaluable to assess these vascular predictors with non-invasive imaging to enable large scale screening to identify patients with a high-risk coronary plaque suitable for invasive intervention. Very few studies investigated assessment of local ESS using CT angiography (CTA) techniques. Our purpose was to compare lumen area and ESS by CTA and by angio/IVUS imaging performed at the same time.

Methods: Coronary arteries were 3D reconstructed using vascular profiling (VP) techniques of angiography & IVUS images and local ESS was calculated with computational fluid dynamics (CFD) as part of the PREDICTION Study. We identified patients who had undergone coronary CTA using a 64 row multi-detector CT at the same time frame and derived cross-sectional anatomy from 3D CTA model and then performed CFD as with angio/IVUS images. Corresponding distal and proximal reference branches were identified in both CTA and VP models for comparison. Each reconstructed artery was divided into sequential 3-mm segments. We compared cross-sectional areas and ESS of corresponding segments in the CTA 3D models with those from the VP 3D models, considering VP as the gold standard.

Results: 9 vessels (1 LAD, 5 Cx and 3 RCA) were analyzed from 7 patients. CTA and angio/IVUS data were acquired a mean of 15±12 days apart. We compared 133 segments for cross-sectional area and 123 segments for ESS. Cross-sectional area was very similar between two methods (r=0.89, p<0.01; mean difference 0.10 ± 2.25 mm²) as was local ESS (r=0.78, p<0.01; mean difference 0.16 ± 0.88 Pa). Sensitivity and specificity of detecting high-risk segments with low ESS (<1.0 Pa) with CTA model were 53% and 96%, respectively.

Conclusion: Coronary 3D geometry reconstructed from coronary CTA correlates with VP based 3D geometry with high accuracy. Coronary CTA is a promising modality for estimation of ESS and identification of the high-risk plaque.