Non-invasive coronary artery evaluation with multidetector computed tomography.

Objectives: To evaluate the diagnostic accuracy of 16-slice multidetector spiral computed tomography in the non-invasive detection of coronary artery disease compared to invasive coronary angiography.

Design: Prospective study.

Patients: We studied 35 patients, 21 men and 14 women, mean age 63.6+/-13.8 years, who underwent elective invasive coronary angiography one week prior to undergoing 16-slice multidetector computed tomography.

Methods: We used a Toshiba Aquilion 16-slice system. Helical acquisition was performed with breath holding after injection of 140 ml of a contrast agent into a peripheral vein. Cross-sectional images were reconstructed with a slice thickness of 0.5 mm or 1 mm. We set 15 anatomical segments for the coronary tree and the evaluation was performed independently for both techniques. Coronary artery lesions with over 50% reduction in diameter were considered significant. Findings from both techniques were evaluated using ROC curve analysis.

Results: It was possible to evaluate 87.7% of the coronary tree segments; heavy calcification, enability to perform breath holding, motion artefacts and stents made evaluation of the other segments impossible. Overall sensitivity and specificity were 85.9% and 98.2% respectively with a 95 % confidence interval; positive predictive value was 87% and negative predictive value 97% For the different anatomical segments, sensitivity and specificity were: proximal 88.9% and 97.8%; medial 90.9% and 95.8%; distal 86.7% and 99.1% respectively.

Conclusions: Multidetector computed tomography is a promising non-invasive technique for detecting or ruling out significant coronary lesions.