Author + information
- Received November 24, 2008
- Revision received June 18, 2009
- Accepted June 24, 2009
- Published online October 1, 2009.
- Boaz D. Rosen, MD⁎,
- Veronica Fernandes, MD, PhD⁎,
- Robyn L. McClelland, PhD‡,
- Jeffrey J. Carr, MD, MSCE§,
- Robert Detrano, MD∥,
- David A. Bluemke, MD, PhD† and
- João A.C. Lima, MD⁎,†,⁎ ()
Reprint requests and correspondence:
Dr. João A. C. Lima, Cardiology Division, Blalock 52469, Johns Hopkins Hospital, 600N Wolfe Street, Baltimore, Maryland 21287-6568
Objectives The MESA (Multi-Ethnic Study of Atherosclerosis) is a population-based study of 6,814 men and women. We sought to analyze the relationship between the extent of coronary artery calcium (CAC) at baseline and the severity of coronary stenoses in clinically indicated coronary angiography studies during follow-up.
Background CAC is an established predictor of major cardiovascular events. Yet, the relationship between CAC and the distribution and severity of coronary artery stenoses has not been widely explored.
Methods All MESA participants underwent noncontrast enhanced cardiac computed tomography during enrollment to determine baseline CAC. We analyzed 175 consecutive angiography reports from participants who underwent coronary catheterization for clinical indications during a median follow-up period of 18 months. The relationship between baseline CAC and the severity of coronary stenosis detected in coronary angiographies was determined.
Results Baseline Agatston score was 0 in only 7 of 175 (4%) MESA participants who underwent invasive angiography during follow-up. When coronary arteries were studied separately, 13% to 18% of coronary arteries with ≥75% stenosis had 0 calcium mass scores at baseline. There was close association between baseline calcium mass score and the severity of stenosis in each of the coronary arteries (test for trend, p < 0.001). For example, mean calcium mass scores for <50%, 50% to 74%, and ≥75% stenosis in the left anterior descending coronary artery were 105.1 mg, 157.2 mg, and 302.2 mg, respectively (p < 0.001). Finally, there was a direct relationship between the total Agatston Score at baseline and the number of diseased vessels (test for trend, p < 0.001).
Conclusions The majority of patients with clinically indicated coronary angiography during follow-up had detectable coronary calcification at baseline. Although there is a significant relationship between the extent of calcification and mean degree of stenosis in individual coronary vessels, 16% of the coronary arteries with significant stenoses had no calcification at baseline.
Most coronary artery disease (CAD) events occur in the population deemed to be at “intermediate risk” (1). Therefore it is assumed that refined methods of risk prediction might be useful.
The use of computed tomography (CT) imaging to detect coronary artery calcium (CAC) has been proposed as a tool to allow improved detection of at-risk individuals (2). However, not all coronary events are caused by calcified plaque (3,4), and although the relationship between CAC and coronary events on a per-patient basis has been thoroughly studied, little is known about the relationship between calcification of individual coronary arteries and the degree of coronary artery stenosis in invasive coronary angiography.
The extent of calcified plaques has been shown to correlate moderately closely to overall plaque burden in histology (5,6). Coronary calcifications measured by CT are usually expressed as the “Agatston score” (2,7). In numerous trials, CAC has been shown to be predictive of major cardiovascular events and to modify the cardiovascular risk predicted by the Framingham risk score, especially in the intermediate risk group (2,8–13).
The Agatston score has been shown to be related to the severity of CAD and the number of stenosed vessels determined by coronary angiography (7,14–17). However, the association between the extent of coronary calcification and the severity of coronary stenosis in individual coronary beds has not been widely explored.
We studied individuals enrolled in the MESA (Multi-Ethnic Study of Atherosclerosis) study (18). All MESA participants were asymptomatic for cardiovascular disease at the time of enrollment. At the time of this analysis, 6 years had elapsed from enrollment, and a substantial number of participants had developed symptoms of CAD or were diagnosed to have myocardial ischemia by stress testing. As a result, individuals underwent diagnostic coronary catheterizations, and some proceeded to have surgical or percutaneous coronary revascularization to treat obstructive CAD. The purpose of this study is to report the relationship between the baseline extent of CAC measured by cardiac CT and the anatomic findings in subsequent clinically indicated coronary angiography.
The characteristics of subjects enrolled in MESA have been described elsewhere (18). In short, the MESA study is a multicenter community-based study designed to investigate the mechanisms underlying the development and progression of subclinical cardiovascular disease.
Individuals with symptoms or a history of cardiovascular disease were excluded; 6,814 men and women, 45 to 84 years of age, from different ethnic origins (White, African American, Hispanic, and Chinese) were enrolled. As of June 2005, 6 years after enrollment, the reports of clinically driven invasive coronary angiography in 175 participants had been received. Data from these subjects are described in this report. The study protocol had been approved by the institutional review boards in each of the participating centers, and informed consent was obtained from each of the participants.
Measurement of CAC
As a part of the MESA protocol, all participants underwent measurement of CAC with either multidetector row CT or electron beam computed tomography (EBCT) at baseline. Detailed scan protocols have previously been published (19). Image settings are described in Supplement A in the Online Appendix.
Agatston score (7) is an important and well-established measure of CAC that has been used in multiple studies. Therefore, we used this parameter to evaluate the total CAC score/patient. Calcium mass score is less commonly used as an index of CAC but has been found to be more reproducible in assessing CAC (20). In view of the multiple types of CT scanners, including multidetector row CT and EBCT, and the multiple sites participating in the MESA study and the smaller amounts of calcium in a single coronary vessel as opposed to the entire coronary artery tree, we opted to use calcium mass score as a measure of CAC in individual coronary beds.
Standard calcium scoring software tools were used for calculating the Agatston score and calcium mass score. For analysis, we used calcium mass scores (mg) determined for each of the coronary arteries and the total Agatston score for the entire coronary bed.
Results obtained from coronary angiography reports
As of June 2005, the MESA co-coordinating center had received incident coronary angiography reports for 175 individuals. The indications for coronary angiography were clinical and not part of the MESA protocol. All angiographic studies had been performed after baseline CT exams. Only the first catheterization report was used for analysis if an individual underwent more than 1 procedure.
The studies were performed in different catheterization laboratories, and quantitative analysis (quantitative coronary angiography) of luminal stenosis was not routinely available. In most cases, the extent of stenosis was assessed by visual assessment. Therefore, the following categories were used to define the degree of stenosis for individual coronaries: 0% to 49%, 50% to 74%, and ≥75% luminal diameter stenosis to indicate nonsignificant, intermediate, and significant stenosis, respectively. For the analysis of the extent of coronary disease (1-, 2-, and 3-vessel disease), significant stenosis was defined as: ≥50% left main coronary artery (LMCA); and ≥75% in the left anterior descending (LAD), left circumflex (LCX), or right coronary artery (RCA) or their branches.
Definition of risk factors
Hypertension was defined as: diastolic blood pressure ≥90 mm Hg, systolic blood pressure ≥140 mm Hg, or receiving treatment for hypertension. Dyslipidemia was defined as total cholesterol ≥240 mg/dl, low-density lipoprotein cholesterol ≥160 mg/dl, triglycerides ≥150 mg/dl, high-density lipoprotein cholesterol <45 mg/dl, or receiving treatment for dyslipidemia. Diabetic individuals were defined as either having fasting plasma glucose ≥126 mg/dl or receiving treatment. Smoking status was defined as current smoking, former smoking, or never smoked.
All analyses were performed with STATA-8 software (College Station, Texas). To study the association between average baseline calcium scores in each coronary artery and the degree of stenosis (0% to 49%, 50% to 74%, and ≥75%) in the corresponding vessel or its branches, we used analysis of variance. Due to the skewed distribution of calcium scores, a log-transformation of calcium scores was used when testing trends. Even in this symptomatic subset, a considerable portion of baseline coronary artery calcium plaque (CACP) scores were 0. Hence the value (CACP score + 1) was used for log transformation. Chi-squared tests (or Fisher exact tests where indicated) were performed to compare the distribution of stenosis across quartiles of CACP. Significance was defined as p < 0.05.
Demographic and clinical characteristics
The demographic characteristics and risk factor profile of the 175 individuals with available angiographic data are shown in Table 1. Compared with the other participants of the MESA cohort, the 175 individuals with available angiographic data were older, more likely to be men, and Caucasian. There were no significant differences between the 2 groups in socioeconomic parameters, including educational status, yearly income, and health insurance coverage (Online Table A).
Compared with those who did not undergo coronary angiography, patients who had invasive angiography had lower high-density lipoprotein cholesterol (p = 0.006) and were more likely to have a history of hypertension, dyslipidemia, and diabetes mellitus (21% vs. 9%, p < 0.001).
The total Agatson score was higher in the patients who underwent invasive angiography as compared with control subjects (528.4 ± 772.2 vs. 136.3 ± 339.4, p < 0.0001). Similarly, the average calcium mass score was approximately 100 mg greater for each of the coronary arteries (Table 2, p < 0.001 for all vessels). The detailed distribution of calcium scores in the MESA participants who underwent angiography is shown in the Online Table B. As expected, calcium mass score in the LMCA was lower than the other arteries. Distribution of CAC in the LAD, LCX, and RCA was similar, except for a slightly greater percentage of patients with a calcium mass score ≥200 mg in the LAD.
Results of coronary angiography
The indications for coronary angiography and the angiographic findings are shown in Table 3.
The majority of participants (73%) underwent coronary angiography because of angina or acute myocardial infarction. Eight (5%) participants underwent coronary angiography as a direct result of the CT study, and 11 (6%) individuals underwent catheterization due to a positive stress test. Indications for catheterization are unknown in 3 patients. The median time interval between baseline cardiac CT and the clinical angiographic procedure was 18 months (range <1 to 52 months).
Among the MESA participants who underwent coronary angiography, 41%, 32%, and 31% individuals had ≥75% stenosis in the LAD, LCX, and RCA, respectively. Importantly, 20 individuals (12%) had significant LMCA disease (≥50% stenosis), whereas 21% had either 3-vessel or LMCA disease.
Relationship between baseline calcium scores and obstructive CAD in individual vessels
For each of the coronary vessels, there was a significant relationship between baseline calcium scores and the severity of coronary artery stenosis determined by angiography (Fig. 1). The baseline score was markedly higher in coronary arteries that were found to be stenotic. For example, in the LAD, baseline calcium mass scores were 105.1 ± 167.4 mg (n = 83), 157.2 ± 216.6 mg (n = 20), and 302.2 ± 329.8 mg (n = 71) in the presence of 0% to 49%, 50% to 74%, and ≥75% stenosis, respectively (p < 0.001). Similar relationships were noted for the LCX and RCA (p < 0.001) (Online Table C). For the LMCA, differences were of lower magnitude yet significant (p < 0.05).
When the distribution of stenotic segments in each of the coronary arteries was studied with calcium score quartiles, there was a gradual increase in the severity of stenosis among those who had greater calcium scores at baseline (Fig. 2). This pattern was seen in each of the coronary arteries except for the LMCA (p < 0.001 for each of the coronaries).
Seven of 176 participants (4%) with significant flow-limiting CAD (≥75% stenosis in at least 1 of the coronary beds) had a 0 calcium mass score at baseline. Similarly, 11 of 176 participants (6.25%) with at least 1 stenosis ≥50% had a 0 mass calcium score. Importantly, the rate of a 0 calcium mass score in spite of the presence of a ≥75% stenosis was 12.7% (9 of 71) for the LAD, 17.9% (10 of 56) for the LCX, and 17% (9 of 53) for the RCA, when analyzing individual arteries. Of vessels with ≥50% stenoses, 16.3% (15 of 92), 15.7% (11 of 70), and 21.8% (17 of 78) had 0 baseline calcium mass scores in the LAD, LCX, and RCA territories, respectively. Moreover, among 20 patients with ≥50% stenosis in the LMCA, 8 individuals (35%) had a 0 mass score in the LMCA at baseline, and 2 patients had a total 0 Agatston score.
Finally, results for patients who underwent coronary angiography earlier than the median time interval (18 months after enrollment) were similar to patients whose angiographic studies were performed later (data not shown).
Relationship between baseline calcium scores and the number of diseased coronary vessels
There was a direct relationship between the total Agatston score at baseline and the number of diseased vessels (Fig. 3). For patients without significant stenosis as well as for patients with 1-, 2-, or 3-vessel disease or LMCA disease, average Agatston scores were 161.3 ± 268.2 U, 462.7 ± 608.5 U, 961.7 ± 986.9 U, 1,351.4 ± 1,180.1 U, and 658.3 ± 607.4 U, respectively (p < 0.001). The relationship between the number of diseased vessels and the categories of the Agatston score is displayed in Figure 4. Notably, 66% of patients with scores <100 Agatston units at baseline did not have significant coronary stenoses, whereas only 13% of the participants with baseline scores ≥400 had no coronary artery stenoses.
In contrast, 9% of the participants with Agatston score <100 had 3-vessel or LMCA disease, compared with 41% of those with a baseline score ≥400 units (p < 0.001).
The main purpose of the present study was to explore the relationship between the extent of coronary calcification detected by cardiac CT and the severity of coronary stenosis documented in clinically indicated coronary angiography. The study population consisted of 175 individuals who were asymptomatic at enrollment but subsequently developed symptoms of CAD or had positive stress tests requiring coronary angiography.
We could demonstrate that, on a per-patient basis, the severity of coronary calcification was directly related to the extent of obstructive CAD in those individuals who underwent coronary angiography due to clinical indications after a median time interval of 18 months after enrollment. Only 4% of individuals who had significant stenosis in angiography had a baseline coronary score of 0.
Importantly, this does not indicate the likelihood of an individual with a 0 calcium score to develop stenoses or a coronary event during follow-up. Of the entire group of 3,563 MESA participants who had baseline calcium score of 0, only 11 individuals (0.3%), were referred for coronary angiography due to clinical indications and were found to have significant stenosis after a median of 18 months.
Calcium scores in each of the major coronary beds (except for the LMCA) predicted the severity of coronary stenosis as well as the extent and distribution of CAD among those MESA participants who underwent clinically driven coronary angiography, similar to the relationship on a per-patient level. However, and as expected, calcium was less predictive on a per-vessel than on a per-patient level: between 13% and 18% of individual coronary arteries demonstrated stenoses ≥75% in spite of the absence of CAC at baseline. For left main stenoses ≥50%, absence of calcium was noted in 35% of cases.
In previous studies, CAC has been shown to closely correlate to the severity of coronary stenosis and number of vessels with significant stenoses as determined by coronary angiography (14,15,17). Budoff et al. (15) evaluated 1,851 patients who underwent both EBCT and coronary angiography due to suspected CAD. The overall sensitivity for CAC to predict obstructive disease on angiography was 95%, and the specificity was 66%. A strong correlation (r = 0.92) has been demonstrated between the amount of CAC detected by CT as well as the extent of calcium deposits and the severity of coronary stenosis determined by histomorphometry (5,6). However, coronary plaque area was approximately 5 times greater than the area of calcium. Although calcium is correlated to the amount of plaque, it is not necessarily correlated to the degree of stenosis of individual plaques. Similarly, the relationship between plaque calcification and the risk of plaque rupture is not well established. Recurrent subclinical episodes of plaque rupture with repeated hemorrhage and healing might constitute an important mechanism of plaque growth and worsening of luminal stenosis and might predispose plaques to calcification (21). However, plaques can rupture in the absence of calcium. A recent study used CT to demonstrate that, in patients presenting with suspected acute coronary syndromes, there is a relatively high prevalence of noncalcified plaques. Moreover, 33% of the patients had a 0 calcium score. This study, although relatively small (n = 40), clearly demonstrated that the absence of calcium does not absolutely exclude the existence of a vulnerable plaque (22).
However, in asymptomatic individuals and on a per-patient level, coronary calcification remains an excellent predictor of future CAD events. CAC has been found to predict major coronary events independently of standard risk factors and C-reactive protein and has been found to be superior to Framingham risk score in predicting events (8,10,23). As mentioned earlier, of the 3,563 MESA participants who had total 0 calcium score, only 11 individuals (0.3%), were referred for coronary angiography due to clinical indications and were found to have significant stenosis (≥50%) after a median of 18 months. These findings are consistent with previous large studies, including the overall results of the MESA trial, indicating that a 0 score is associated with a low incidence of future cardiovascular events (9,10,24,25).
In summary, although the calcium score is a useful surrogate of the total atherosclerosis burden in the coronary vessels, only a small portion of the atherosclerotic plaque is calcified, and the relationship to the severity of stenosis is less close.
The main limitation of the study is the availability of data in only 175 individuals who underwent coronary angiography for clinical indications. This number constitutes <3% of the MESA cohort. Moreover, the database includes only catheterization reports received by the MESA coordinating center and does not represent complete ascertainment of all the angiography procedures, despite extensive efforts dedicated to track all clinical events.
The indications for performance of coronary angiography were entirely clinical. It is possible that individuals who were aware of their positive CACP scores were more likely to seek medical attention when becoming symptomatic. Five percent of individuals underwent angiography as a result of the CT results, as can be seen in Table 2. Therefore, ascertainment bias is a possibility and might have affected the findings. As a result, any conclusions about the predictive value or the performance of baseline cardiac CT (e.g., sensitivity and specificity) cannot be made. These factors might have been more important in individuals of a higher socioeconomic status and better access to medical care. No differences were observed in various socioeconomic parameters, implying that access to medical care was probably not a major factor distinguishing between individuals who did and those who did not undergo catheterization.
Angiographic studies have been performed in different laboratories, and thus quantification of coronary luminal area (quantitative coronary angiography) was not performed. Therefore, it was necessary to express stenoses with a categorical scale. Finally, the time between the baseline CT study and angiography was dictated by clinical indications and was not uniform (ranging from 1 to 52 months).
We opted to define 75% stenosis in the major coronary arteries and 50% stenosis in the LMCA as significant stenoses, because these are routinely used for the performance of either percutaneous or surgical coronary interventions. Similar results were obtained if 50% were chosen to define significant luminal stenosis.
Agatston score (7) is an important and well-established indicator of coronary calcification and has been used in multiple studies. Therefore we used this parameter for assessing the total calcium score. Calcium mass score is less commonly used to assess CAC. However, mass score has been found to be more reproducible in assessing CAC (20), and in view of the multiple centers and the use of different CT scanners, we opted to use the mass score for individual arteries.
In a previous study, also based on the MESA study, Brown et al. (26) underscore the importance of the spatial distribution of calcified plaque. They show that “calcium coverage score” was a better predictor of cardiovascular events than calcium mass or the “Agatston score.” Such an analysis was beyond the scope of our study, and we could not obtain exact per-segment correlations between calcium deposits and the stenotic lesions determined by angiography.
The extent of coronary calcification in individual coronary arteries is related to the severity of coronary stenosis in these vessels. All the same, a considerable number of arteries have significant stenosis, despite having no calcification at baseline. On a per-patient level, the presence of significant stenoses in the absence of calcium is rare.
The authors thank the other investigators, the staff, and the participants of the MESA study for their valuable contributions. A full list of participating MESA investigators and institutions can be found at http://www.mesa-nhlbi.org.
For supplementary tables and material, please see the online version of this article.
This study was supported by the National Heart, Lung, and Blood Institute grants (RO1-HL66075-01 and RO1-HL6607-01) and the Multi-Ethnic Study of Atherosclerosis study contracts (N01-HC-95159 through N01-HC-95166, NO1-HC-95168, and NO1-HC-95169). Dr. Lima is also supported by the Johns Hopkins Reynolds Center. Dr. Rosen is currently affiliated with the University of Utah, Salt Lake City, Utah.
- Abbreviations and Acronyms
- coronary artery calcium
- coronary artery calcium plaque
- coronary artery disease
- computed tomography
- electron beam computed tomography
- left anterior descending
- left circumflex
- left main coronary artery
- left ventricle
- right coronary artery
- Received November 24, 2008.
- Revision received June 18, 2009.
- Accepted June 24, 2009.
- American College of Cardiology Foundation
- Greenland P.,
- Bonow R.O.,
- Brundage B.H.,
- et al.
- Schmermund A.,
- Erbel R.
- Rumberger J.A.,
- Simons D.B.,
- Fitzpatrick L.A.,
- Sheedy P.F.,
- Schwartz R.S.
- Agatston A.S.,
- Janowitz W.R.,
- Hildner F.J.,
- Zusmer N.R.,
- Viamonte M. Jr..,
- Detrano R.
- Arad Y.,
- Goodman K.J.,
- Roth M.,
- Newstein D.,
- Guerci A.D.
- Raggi P.,
- Callister T.Q.,
- Cooil B.,
- et al.
- Budoff M.J.,
- Georgiou D.,
- Brody A.,
- et al.
- Budoff M.J.,
- Diamond G.A.,
- Raggi P.,
- et al.
- Kaufmann R.B.,
- Peyser P.A.,
- Sheedy P.F.,
- Rumberger J.A.,
- Schwartz R.S.
- Rumberger J.A.,
- Sheedy P.F. III.,
- Breen J.F.,
- Schwartz R.S.
- Bild D.E.,
- Bluemke D.A.,
- Burke G.L.,
- et al.
- Carr J.J.,
- Nelson J.C.,
- Wong N.D.,
- et al.
- Burke A.P.,
- Kolodgie F.D.,
- Farb A.,
- et al.
- Henneman M.M.,
- Schuijf J.D.,
- Pundziute G.,
- et al.
- Guerci A.D.,
- Spadaro L.A.,
- Goodman K.J.,
- et al.
- Lamonte M.J.,
- Fitzgerald S.J.,
- Church T.S.,
- et al.