Cardiovascular Risk Assessment Based on Quantification of Coronary Calcium in CCTA

Cardiology Review® Online, April 2012, Volume 28, Issue 2

Navin C. Nanda, MD


Bischoff B, Kantert C, Meyer T, et al. Cardiovascular risk assessment based on the quantification of coronary calcium in contrast-enhanced coronary computed tomography angiography. Eur J Echocardiol. 2011. doi:10.1093/ejechocard/jer261.

Non—contrast-enhanced calcium scoring is often performed simultaneously with contrastenhanced coronary computed tomography angiography (CCTA) as an added way to assess risk. With CCTA radiation dose continuing to decrease, the need for simultaneous performance of non–contrast-enhanced scans in coronary

artery assessment is being questioned. In their study of coronary artery calcification (CAC) risk classification in CCTA, Bischoff and colleagues sought to determine whether CCTA can reliably reproduce the expected CAC score as determined by a conventional, nonenhanced scan.

The investigators performed a retrospective analysis on a total of 600 patients who underwent both conventional non-enhanced multidetector CT scan for CAC scoring and a contrast-enhanced CCTA to evaluate for suspected coronary artery disease. One hundred of the subjects were randomly assigned to the first cohort, which was used to develop the model for CAC risk classification by CCTA. The second cohort of 500 patients was used to validate the prediction model.

All scans were performed on a dualsource CT system. Radiation dose—reduction protocols were in place, including tube voltage settings of 100 kV or 120 kV for body weights of less than 90 kg and greater than 90 kg, respectively. Patients received up to 20 mg of intravenous metoprolol, dosed in 5-mg increments, with a target heart rate of less than 60 beats per minute (bpm). CCTA scan protocol selection (sequential or spiral) was at the discretion of the performing physician, although sequential scanning was limited to patients with steady heart rates of less than 60 bpm. CCTA images were reconstructed using

a commercially available hard kernel. An Agatston score equivalent (ASE) score was used to stratify the patients as very low risk (score 0), low risk (1-99),

intermediate risk (100-399), and high risk (>400).

CAC has a characteristically high attenuation value, measured in Hounsfield units (HU). Iodinated contrast also has high attenuation values, albeit typically less than that of calcium. In order to estimate CAC in the setting of intravascular contrast, a method of identifying calcium despite the presence of contrast had to be established. Attenuation values above that threshold are identified as calcium, whereas values below are ignored. To identify an optimal threshold, the mean density of the contrast opacified aorta is determined. The HU threshold was set to 120% of the aorta’s intravascular density and subsequently increased in 10% increments up to 160%. Semi-auto mated scoring software was then used to identify which threshold correctly identified CAC without misidentifying contrast as calcium. From this analysis, it was determined that 150% was felt to be optimal. Higher attenuation values of the intravascular contrast make it more likely the software will miss calcified plaques, resulting in underestimation of the calculated CAC. To correct for these errors, a calibration factor, which takes into account intravascular HU, was developed by comparing the CAC score derived from the CCTA of the first cohort to their corresponding nonenhanced CT.

The mean HU of the study cohort was 448 + 107, resulting in an HU threshold of 672 + 154. The mean calculated calcium score was 94 + 162 ASE, which was similar to the corresponding conventional scan value of 108 + 170 ASE. Use of the calibration factor resulted in calcium scores comparable to conventional calcium scoring. Calculation of CAC using CCTA images was highly effective more than 90% of the time for appropriately stratifying patients as very low, low, intermediate, and high risk. If the 75th age- and gender-specific percentile is used as a metric, this method was more than 95% successful. Dose-reduction techniques as well as acquisition type, either helical or sequential, did not seem to affect the accuracy of risk stratification.


1. Glodny B, Helmel B, Trieb T, et al. A method for calcium quantification by means of CT coronary angiography using 64-multidetector CT: very high correlation with

Agatston and volume scores [published online ahead of print February 24, 2009]. Eur Radiol. 2009;19:1661-1668.

2. Otton JM, Lønborg JT, Boshell D, et al. A method for coronary artery calcium scoring using contrast-enhanced computed tomography [published online ahead of print

November 20, 2011]. J Cardiovasc Comput Tomogr. 2012;6:37-44.

3. Achenbach S, Goroll T, Seltmann M, et al. Detection of coronary artery stenoses by low-dose, prospectively ECG-triggered, high-pitch spiral coronary CT angiography.

JACC Cardiovasc Imaging. 2011;4:328- 337.

4. Stolzmann P, Leschka S, Betschart T, et al. Radiation dose values for various coronary calcium scoring protocols in dualsource CT [published online ahead of print

Devember 12, 2008]. Int J Cardiovasc Imaging. 2009;25:443-451.


CAC and Cardiovascular Risk Assessment in CCTA

Despite the promising findings, the fact that these data are acquired from a retrospective study cannot be overlooked. As with any provocative retrospective study, most will agree that a well-designed prospective study is required to confirm these findings. As such, we will not belabor the point.

In this study, an initial cohort of 100 subjects was used to define a calibration factor that was then validated in a similar but much larger cohort. Overall, the demographics and risk factors were fairly representative of the population generally referred for coronary CTA. Patients with very high calcium scores were excluded from the analysis

as it is the investigators’ clinical practice to routinely, and reasonably, defer CCTA when the calcium score is greater than 800. Unfortunately, as seen previously in other studies, accuracy in the estimation of CAC was diminished at very high levels,1,2 and there was a trend toward a degree of underestimation. But, as the authors mentioned, this is not likely clinically relevant in the setting of very high calcium scores.

What is clinically relevant are the cases that registered calcium on the CCTA but not on the native scan. In the setting of presumably normal-appearing coronary arteries on the CCTA, these false-positives would favor more aggressive therapy where perhaps none is required. What is more concerning are the false-negatives—the cases where identification of coronary calcium would stratify the patient into more aggressive, potentially life-saving therapy. Although false-negatives were seen relatively infrequently (1.6%), their impact should be considered. Absent coronary artery calcification seen on a CCTA should not necessarily place the patient into the lowest risk stratum. These patients may at least warrant consideration for further risk stratification if clinical suspicion suggests it.

There are also logistical concerns in this study that need to be addressed. For example, CCTA reconstruction to evaluate for calcium content was performed with a hard kernel to assess CAC. It has become commonplace to interpret CCTA images through a softer kernel, which may not be adequate for the evaluation of CAC. Also, using multiple reconstructions is more labor intensive and requires additional computer memory. In addition, some groups use the non-enhanced scan to optimize planning of the CCTA and ensure adequacy of the z-coverage. Insufficient z-coverage can result in “clipping” of vessels, resulting in unevaluable segments and an inadequate

exam. On the other hand, too much z-coverage results in excess radiation dose.

Several recent high-profile journal articles have rightfully shined a light on the potential dangers of cumulative medical radiation doses.3,4 Cardiac imaging sits at the epicenter of the discussion. As a result, there has been a drive toward progressive dose reduction. The stateof- the-art CCTA can be obtained with a radiation dose

approaching that of a conventional non-enhanced CAC scan,3,4 and the addition of a conventional calcium score nearly doubles the total radiation dose. The importance of acquiring the coronary anatomical data provided by the CCTA as well as the additional risk stratification of coronary calcium scoring without additional radiation cannot be overstated.

About the Authors

Navin C. Nanda, MD, is Distinguished Professor of Medicine and Cardiovascular Disease and director of the Heart Station/Echocardiography Laboratories at the University of Alabama at Birmingham. He is also director of the Echocardiography Laboratory at the Kirklin Clinic, University of Alabama Health Services Foundation. Dr Nanda has received many awards for service and achievement in medicine and is the author of numerous publications. Dr Nanda was assisted in the writing of this article by O. Julian Booker, MD, assistant professor of medicine, Division of Cardiovascular Disease, University of Alabama at Birmingham. Dr Booker’s MD is from Baylor

College of Medicine, Houston, Texas, where he also completed his internship, a residency in internal medicine, and a fellowship in cardiovascular disease.