Quantification of atherosclerotic coronary plaque components by submillimeter computed tomography

Abstract

Although several investigations have shown that multi-detecor row computed tomography (MDCT) of the coronary arteries can detect noncalcified atherosclerotic plaque, it has remained unresolved if the method also determines features of a rupture-prone plaque. We set out to correlate the size of atherosclerotic plaque components with cardiac MDCT with histology. In 30 autopsy cases, hearts were isolated, coronary arteries filled with contrast agent, and depicted with a clinical 16-row detector CT with a slice thickness of 0.63 mm. Transections of the three main coronary arteries were reconstructed and compared with histopathologic sections using light microscopy. MDCT measurements of total plaque area (r = 0.73, P < 0.0001) and calcified plaque area (r = 0.83, P < 0.0001) correlated well with histopathology, while measurements of non-calcified plaque area (r = 0.53, P < 0.0001) and lipid core size (r = 0.43; P < 0.0001) correlated less well. MDCT overestimated all plaque areas except lipid core size, which was underestimated. Coronary CT provides an accurate and reproducible method for the quantitative assessment of total plaque and calcified plaque areas. However, the method is less accurate for the quantification of non-calcified plaque area and lipid core size, which is ascribed to limited spatial and contrast resolution. With the present technique, the detection of vulnerable plaques by MDCT remains uncertain.