Submitted to: Radiology
Publication Type: Peer reviewed journal
Publication Acceptance Date: 12/15/2005
Publication Date: 3/1/2006
Citation: Moselewski, F., Ferencik, M., Achenbach, S., Brady, T.J., Booth, S.L., Hoffmann, U., Ik-Kyung, J., Cury, R., Abbara, S. 2006. Threshold-dependent variability of coronary artery calcification measurements -- implications for contrast-enhanced multi-detector row-computed tomography. European Journal of Radiology. 57(3):390-5. Interpretive Summary: This study was conducted to determine the effect of contrast-enhanced multi-detector computerized tomography (MDCT) compared with non-contrast-enhanced scans on the measurement of the coronary artery, and the variability of coronary artery calcification (CAC) measurements that occurs when using different threshold levels for measurements in non-contrast enhanced MDCT scans. Contrast enhancement was measured in the major coronary arteries in 30 adult patients who underwent contrast-enhanced MDCT. In 50 subjects who underwent non-contrast MDCT, CAC was measured using five different approaches at four different thresholds. The absolute and relative measurement differences between the four thresholds were calculated for each of the different approaches to CAC measurement. The effect of contrast-enhanced multi-detector computerized tomography (MDCT) was similar compared with non-contrast-enhanced scans on the measurement of the coronary artery. All measurements varied significantly among thresholds. CAC measurements varied significantly with different threshold levels. Therefore reliable detection and accurate measurements of CAC may not be achievable in contrast-enhanced coronary computerized tomography exams using current evaluation techniques.
Technical Abstract: The present study investigated the threshold-dependent variability of coronary artery calcification (CAC) measurements and the potential to quantify CAC in contrast-enhanced multi-detector row-computed tomography (MDCT). METHODS: We compared the mean CT attenuation of CAC to luminal contrast enhancement of the coronary arteries in 30 patients (n = 30) undergoing standard coronary contrast-enhanced spiral MDCT. The modified Agatston score [AS], calcified plaque volume [CV], and mineral mass [MM]) at four different thresholds (130, 200, 300, and 400 HU) were measured in 50 patients who underwent non-contrast-enhanced MDCT. RESULTS: Mean CT attenuation of CAC was similar to the attenuation of the contrast-enhanced coronary lumen (CAC 297.1 +/- 68.7 HU versus 295 +/- 65 HU (p < 0.0001), respectively). Above a threshold of 300 HU CAC measurements significantly varied to standard measurements obtained at a threshold of 130 HU (p < 0.0001). The threshold-dependent variation of MM measurements was significantly smaller than for AS and CV (130 HU versus 400 HU: 63, 75, and 81, respectively; p < 0.001). These differences resulted in a change of age and gender based percentile category for AS in 78% of subjects. DISCUSSION: We demonstrated that CAC measurements are threshold dependent with MM measurements having significantly less variation than AS or CV. Due to the similarity of mean CT attenuation of CAC and the contrast-enhanced coronary lumen accurate quantification of CAC may be difficult in standard coronary contrast-enhanced spiral MDCT.