Abstract

Objective—To assess the accuracy of measurement of area of the lumen, and sensitivity, and specificity of detection of atheroma in coronary arteries in vitro with a commercially available 20 MHz intravascular ultrasound system.

Setting—A teaching hospital department of cardiology with the support of the department of cardiovascular pathology.

Procedure—10 segments of coronary artery were removed from cadaver hearts. Intravascular ultrasound imaging was performed at fixed levels and the vessels were then sectioned and photographed before histological preparation. An independent blinded observer measured luminal area and assessed the presence of atheroma on the intravascular ultrasound images of 76 vessel sections (304 quadrants). The sensitivity and specificity of detection of atheroma was assessed in comparison with the histologically prepared sections. Luminal areas from intravascular ultrasound, photographs of cross sections of the vessels and histological sections were compared with the technique of limits of agreement.

Results—Overall 36% of the 304 quadrants studied histologically had identifiable atheroma. Intravascular ultrasound sensitivity for atheroma was 0·593 and the specificity was 0·839. The positive predictive value was 0·674, and the relative risk 3·139. Values for area of the vessel lumen were on average 9·4 mm² (confidence interval (CI) 8·6–10·2 mm²) larger than those measured from photographs and 10·7 (CI 9·8–11·6 mm²) larger than those measured from the histological sections.

Conclusions—The intravascular ultrasound system assessed in this study significantly overestimated coronary vessel luminal area and had low sensitivity and specificity for detection of atheroma. Improvements in image resolution are required before this system can provide useful information on coronary artery size and morphology.