By Michael H. Crawford, MD

Professor of Medicine, Chief of Clinical Cardiology, University of California, San Francisco

Dr. Crawford reports no financial relationships relevant to this field of study.

SYNOPSIS: A subgroup analysis of the PROMISE trial showed that CT coronary calcium scores in symptomatic patients at low to intermediate risk for coronary artery disease are more sensitive but less specific for major adverse cardiac events over a two-year follow-up period than stress testing. Consequently, both approaches exhibited similar but modest discriminatory ability.

SOURCES: Budoff MJ, Mayrhofer T, Ferencik M, et al. Prognostic value of coronary artery calcium in the PROMISE Study (Prospective Multicenter Imaging Study for Evaluation of Chest Pain). Circulation 2017;136:1993-2005.

Newby DE. Computed tomography or functional stress testing for the prediction of risk: Can I have my cake and eat it? Circulation 2017;136:2006-2008.

The diagnostic accuracy of stress testing for detecting significant coronary artery disease (CAD) in low-risk patients that clinicians encounter frequently is reduced compared to that in intermediate-risk patients. Thus, investigators from the PROMISE trial (Prospective Multicenter Imaging Study for the Evaluation of Chest Pain) hypothesized that CT coronary artery calcium score (CAC) would be superior to stress testing for predicting major adverse cardiac events (MACE) in symptomatic but low to intermediate risk patients. Among the more than 10,000 patients enrolled in PROMISE at 193 North American medical centers, 4,209 received CAC as their first test and 4,602 received stress testing first, as randomized. Stress testing included exercise or pharmacologic stress plus either ECG alone, nuclear myocardial perfusion imaging, or echocardiography. CAC was determined on at least a 64-slice multidetector CT machine. Studies were classified as normal (CAC = 0), mildly abnormal (CAC = 1-99), moderately abnormal (CAC = 100-400), and severely abnormal (CAC > 400). The stress tests were rated similarly based on the perceived extent of ischemia. There were no clinically meaningful differences in baseline characteristics between the patients in each group. The average age was 61 years, slightly more than half were women, and about half were low risk. Median follow-up was 26 months. Moderate or severely abnormal results in both tests robustly predicted MACE (moderate CAC = hazard ratio, 3.14; 95% confidence interval, 1.81-5.44, and stress test 2.65 [1.46-4.83]; severe CAC = 3.56; 1.99-6.36 and stress test, 3.88 [2.58-5.85]). Any CAC abnormality detected 84% of the patients experiencing a MACE, whereas a positive stress test only detected 43%. However, an abnormal stress test was significantly more specific in predicting MACE (78.6% vs. 35.2%; P < 0.001). Thus, overall discriminatory ability was similar and modest for both tests (C statistic for CAC = 0.67 and for stress testing = 0.64). In a separate analysis of those in the CAC group who also underwent CT angiography (CTA), CTA demonstrated better discriminatory ability (C = 0.72).

The authors concluded that in symptomatic outpatients at low to intermediate risk of CAD, most with subsequent events registered a CAC > 0, but only less than half exhibited an abnormal stress test. On the other hand, an abnormal stress test was much more specific for predicting events. Consequently, both tests demonstrated a similar but modest discriminatory ability.


This post-hoc subgroup analysis of the PROMISE trial, which compared CTA to stress testing, is really an apple-oranges comparison. CAC detects the presence of advanced atherosclerosis (young plaques are not detected) and is not directly related to luminal stenosis or plaque status. In fact, the presence of calcium may be a stabilizing event in atherosclerosis. So, it is not surprising that CAC > 0 detects 84% of patients who experience a MACE over two years. Presumably, the other 16% exhibited non-calcified plaques, which could be detected by CTA. However, CTA is not perfect, as it cannot image the vessel lumen through heavy calcification and there are contradictions to its use, such as advanced kidney disease and tachycardia, that cannot be eliminated safely. CAC imaging has no contraindications, but it carries a low specificity for predicting MACE, especially if the CAC value of > 0 is used (35%). Specificity increased with higher cutoff values (> 100, 67%; > 400, 85%) but at the expense of sensitivity (61% and 31%, respectively). Stress testing detects the functional consequences of atherosclerosis but does not diagnose atherosclerosis because myocardial ischemia can be caused by other conditions such as microvascular disease or left ventricular hypertrophy. Also, it cannot detect non-flow-limiting plaques. Most cardiac events occur with the disruption of such plaques. However, stress testing is superior for detecting those at the highest risk and those who may benefit from revascularization. CAC scores > 400 are associated with a yearly MACE rate of 6% vs. 10% with a markedly positive stress test. Also, a positive stress test is more specific (79% vs. 35%) and a normal stress test is associated with a MACE rate of < 1% per year.

So, how should cardiologists deploy these tests in a low- to intermediate-risk patient with symptoms suggestive of CAD? The authors noted that CAC is a rapid test that can be performed on any CT scanner with low radiation exposure. Also, it is relatively inexpensive and produces no real contraindications. Additionally, studies have shown that knowledge of the presence of calcium can motivate patients to improve their risk profile. So CAC may represent an ideal first screening test, which then could be followed by a second test if CAC crosses some threshold. Traditionally, the second test has been stress testing, but the results of the main PROMISE study suggest it could be CTA, since compared to stress testing it was better at predicting future MACE. CTA may perform better at the lower CAC score range (1-400), but a CAC score > 400 will make lumen visualization by CTA problematic. Perhaps stress testing would be better if a CAC score is > 400. We will need further studies to sort this out.