Author + information
- aDepartment of Imaging, Burns and Allen Research Institute, Cedars-Sinai Medical Center, Los Angeles, California
- bDepartment of Medicine, David Geffen School of Medicine, University of California Los Angeles, Los Angeles, California
- cDivision of Cardiology, Mount Sinai St. Luke’s Hospital, Mount Sinai Heart, and the Icahn School of Medicine at Mount Sinai, New York, New York
- ↵∗Address for correspondence:
Dr. Daniel S. Berman, Department of Imaging, Burns and Allen Research Institute, Cedars-Sinai Medical Center, 8700 Beverly Boulevard, Room 1258, Los Angeles, California 90048.
Coronary artery calcification (CAC) on electrocardiogram-gated noncontrast computed tomography (CT) provides a sensitive and specific marker of coronary atherosclerosis. The extent of CAC reflects the lifetime effect of known and unknown risk factors that cause coronary atherosclerosis in an individual patient. Nearly all outcome studies involving CAC scoring and virtually all clinicians who use CAC scanning in clinical practice rely on the Agatston score, first described by Agatston and Janowitz in 1991, and commonly referred to as the CAC score. This score has been consistently shown to provide strong predictive value for future cardiovascular events, providing incremental prognostic value over global risk scores (1).
The Agatston score is a derived variable that increases based on the area (which is directly related to volume) and the density of calcified coronary plaques. It uses an arbitrary threshold of 130 Hounsfield units (HU) that seemed to best separate calcium from statistical noise on electron-beam CT imaging. The weighting for density that is applied to each coronary plaque identified on CAC scanning is based on the maximal Hounsfield units within each plaque, assessed by a categorical 4-point multiplication: 1 = 130 to 199; 2 = 200 to 299; 3 = 300 to 399; and 4 = ≥400 HU. A prior study report has provided evidence that increased density might be associated with lower rather than higher risk of cardiac event (2).
In this issue of iJACC, Criqui et al. (3) provide additional evidence that CAC plaque density is inversely related to prognosis among 3,398 participants from MESA (Multi-Ethnic Study of Atherosclerosis). In the present study, the follow-up has been extended to a mean of 10.3 years, compared with 7.6 years in the prior study, resulting in 47% more events. In addition, new analyses are provided, including a comparison of the prognostic net reclassification improvement provided by various CAC parameters over clinical factors. As consistently shown in prior outcome studies, the present study confirms a proportional relationship between the magnitude of the Agatston score and the frequency of cardiac events. Similarly, the CAC volume score, which does not include the weighting of CAC for density, manifested a direct relationship.
The findings concerning CAC density, however, were more complex. Among the unadjusted data, cardiac events increased, rather than decreased, among more dense CAC lesions. In part, this is caused by a tendency toward greater density among larger volume CAC lesions. Because the overall magnitude of coronary atherosclerosis (as reflected by the CAC score) is the most potent predictor of cardiac outcomes (4), a greater density score may predict worse outcomes merely because it reflects the presence of larger plaques and correspondingly greater CAC scores. However, when the CAC density scores were adjusted for CAC volume and clinical variables, the hazard ratios flipped, and an inverse relationship between CAC density scores and cardiac outcomes was unmasked. Overall, the study found hazard ratios of 0.71 and 0.75 for coronary heart disease and total cardiovascular disease (CVD) events, respectively, with each SD increase of CAC density when adjusted for CAC volume and the ASCVD score, race/ethnicity, and statin use. The greatest impact of the density assessment when adjusted for volume, risk factors, race/ethnicity, and statin use was in the lowest quartile of CAC volume. Thus, these data suggest a clinical utility to measuring CAC density, particularly with an eye toward including decreased CAC density as an indicator of increased risk.
Incremental Value of CAC Density Measurements
On a practical basis, the Agatston score has multiple strengths, including its ubiquity; the presence of strong, reproducible outcome data gathered over 2 decades; and, perhaps most importantly, the convenience of a single score that can be widely understood by physicians and patients alike. Thus, the potential routine use of any alternative derived variables, which might incorporate volume, a different weighing of density, and potentially other features relating to CAC scans to provide a new prognostic index, needs to be weighed against the current convenient use of the Agatston score in clinical practice. Of most relevance to physicians in this regard is how much incremental prognostic information is provided by any new CAC scoring index. Most relevant in this regard was the authors’ assessment of the net reclassification improvement provided by the CAC density measurement. Compared with assessment by clinical risk factors alone, the magnitude of net reclassification improvement for CVD events was 25.4% for the Agatston CAC score, 27.2% for the CAC volume score, and 34.5% for the model combining CAC volume and density score.
Assessment according to the areas under the receiver-operating curve for predicting CVD events was modest: the area under the receiver-operating curve for predicting CVD events was 0.683 in a model that incorporated clinical factors and CAC volume, but only increased to 0.694 by further incorporating CAC density. Moreover, the area under the receiver-operating curve for the Agatston score was nearly comparable with the CAC volume score. The reason for the relatively comparable predictive value of the volume score versus the Agatston score, given the density hypothesis, is not clear. Because in the present density analysis the predictive value of calcified plaque density is weighted in the opposite direction than the Agatston score, it would be expected that the CAC volume score (not influenced by density) to quantify atherosclerotic burden should have greater prognostic value than the Agatston score.
Potential Pathophysiology of Low-Density CAC Plaque
The underlying pathophysiological process that results in calcified plaques detectable by CT is complex. Coalescence of microcalcification, which is too small to be detected by CT (10-μm diameter), is considered to arise in active plaques from proinflammatory stimuli and macrophage apoptosis (5). The importance of accelerated microcalcification with respect to plaque stability is supported by the recent novel application of 18F-NaF positron emission tomography to assess the rate of plaque calcification (6). When correlating positron emission tomography and CT findings in patients with recent myocardial infarction, foci of 18F-NaF uptake (evidence of active calcium deposition) were found in culprit plaques without relationship to CAC evident on CT scanning, and densely calcified plaques frequently had no 18F-NaF uptake (7). Pathophysiologically, it is considered that this active calcification would be occurring around plaques that contain a higher concentration of inflammatory cells than densely calcified plaques, and as such would be less stable. Furthermore, inflamed plaques are frequently associated with a necrotic core that contains high lipid content. Because lipid manifests as <30 HU on CT, its presence would decrease the HU of any given CAC containing plaque.
Limitations of Current Measurement of CAC Density
The assessment of CAC density used by Criqui et al. (3) is limited by various technical constraints, as pointed out by the authors. The CAC density measurement they used incorporates the Agatston score described nearly 3 decades ago (8). As such, the described CAC density approach used the discrete 4-point scale rather than a continuous HU scale, because a continuous variable of CAC density was not available in the MESA dataset. Thus, a plaque with a score of 199 would be considered to have one-half the density of a plaque with a score of 201. Furthermore, the arbitrary threshold of 130 HU was used, precluding the exploration of the assessment of plaques that have even lower density CAC, potentially in the ranges of 110 to 130. Additionally, for each plaque, the maximum density was used to characterize CAC of the entire plaque. Thus, plaques that were high risk because of having low overall calcium density may have been misclassified as low risk if they had even a small high density component. Finally, across the arterial tree, there could be a considerable heterogeneity in CAC density among coronary plaques, but the approach used by Criqui et al. (3) represents the average of all densities among the coronary plaques. It is possible that 2 patients could have relatively similar total average CAC density scores but differ in the number of low-density lesions, and therefore have different risk profiles for future events. Developing an approach that seeks to better identify and characterize the presence and number of low-density lesions is desirable.
The findings of Criqui et al. (3) regarding CAC plaque density provide an important insight into the potential drawbacks of the Agatston score and is deserving of further study. Future work might proceed along 3 lines. First, a considered replacement of the Agatston score by a CAC score that does not up-weight volume measurements according to increasing CAC density should be assessed. Second, approaches that explore different Hounsfield units for identification of CAC, characterize the heterogeneity of plaque density measurements on CAC scanning, and consider the presence and number of low-density plaques merit development and evaluation in prospective study. Third, recent studies suggest that there may be multiple aspects of CAC beyond the volume and density of CAC that may help characterize CAC-related risk in an individual patient. From the simplest to the most complex, these include the number of vessels with CAC, the number of calcified plaques, their location, the extent to which they cover the coronary artery tree, and plaque CAC density (9). Prospective study should seek to integrate these measurements in a way that maintains the current straightforward and highly convenient use of CAC scanning in clinical practice.
↵∗ Editorials published in JACC: Cardiovascular Imaging reflect the views of the authors and do not necessarily represent the views of JACC: Cardiovascular Imaging or the American College of Cardiology.
All authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- 2017 American College of Cardiology Foundation
- Hecht H.,
- Blaha M.J.,
- Berman D.S.,
- et al.
- Criqui M.H.,
- Knox J.B.,
- Denenberg J.O.,
- et al.
- Arbab-Zadeh A.,
- Fuster V.
- Otsuka F.,
- Sakakura K.,
- Yahagi K.,
- Joner M.,
- Virmani R.
- Dweck M.R.,
- Chow M.W.,
- Joshi N.V.,
- et al.
- Agatston A.S.,
- Janowitz W.R.,
- Hildner F.J.,
- Zusmer N.R.,
- Viamonte M. Jr..,
- Detrano R.
- Berman D.S.,
- Arnson Y.,
- Rozanski A.