Morgan-Hughes G J, Marshall A J, Roobottom C A
Department of Cardiology, South West Cardiothoracic Centre, Plymouth NHS Trust, Plymouth, UK.
Clin Radiol. 2003 May;58(5):378-83. doi: 10.1016/s0009-9260(03)00051-5.
To evaluate the technique of coronary angiography with retrospectively electrocardiogram (ECG)-gated four-slice helical computed tomography (CT).
Within 1 month of undergoing routine day-case diagnostic coronary angiography, 30 consecutive patients also underwent retrospectively ECG-gated multislice CT coronary angiography. This enabled direct comparison of seven segments of proximal and mid-coronary artery for each patient by two blinded assessors. Each segment of coronary artery from the multislice CT image was evaluated initially for "assessability" and those segments deemed assessable were subsequently investigated for the presence or absence of a significantly (n=70%) stenotic lesion.
Overall 68% of proximal and mid-coronary artery segments were assessable. The sensitivity and specificity of four-slice CT coronary angiography in assessable segments for detecting the presence or absence (n=70%) of stenoses were 72 and 86%, respectively. These results correspond to a positive predictive value of 53% and a 93% negative predictive value. If the 32% of non-assessable segments are added into the calculation then the sensitivity and specificity fall to 49 and 66%, respectively.
Although multislice CT coronary angiography is a promising technique, the overall assessability and diagnostic accuracy of four-slice CT acquisition is not sufficient to justify routine clinical use. Further, evaluation should investigate the benefit of the reduction in temporal and spatial resolution offered by 16 and 32 slice acquisition.
