CardioExchange Archive

(Note to readers: CardioExchange invited experts with divergent opinions about this study to participate in a debate. The first part of the debate, including comments from PK Shah, Matthew Budoff, and Rita Redberg, is presented at the bottom of this news story.)
A calcium score of zero does not completely rule out significant coronary disease, according to the surprising results of a substudy of the CORE64 multicenter trial, in which patients referred for angiography were also asked to undergo a calcium scan. Ilan Gottlieb and colleagues report in the Journal of the American College of Cardiology on 291 patients enrolled in the study, of whom 72 had a calcium score of zero. 14 of these patients (19%) had at least one lesion with > 50% stenosis. In the study as a whole there were 64 totally occluded vessels, of which 13 (20%) had no calcium. 9 patients with no calcium underwent revascularization.

The authors concluded that “the absence of coronary calcification does not exclude obstructive CAD or the need for clinically indicated coronary revascularization… among patients with a high enough suspicion of CAD prompting an indication for CCA. The absence of coronary calcification should not be used as a gatekeeper and should not prevent a symptomatic patient from undergoing angiography. Furthermore, a large percentage of totally occluded vessels had no evidence of calcium by CT, emphasizing that calcification is not indispensable for plaque rupture and acute coronary events.”
In their report, the authors note that the AHA/ACC Expert Consensus Document  states that “for the symptomatic patient, exclusion of measurable coronary calcium may be an effective filter before undertaking invasive diagnostic procedures or hospital admission.”
In an accompanying editorial comment, Rita Redberg notes that calcium scans and angiography “measure different stages of the atherosclerotic process” and that coronary calcification occurs late in the process, so “it is not surprising that significant CAD can occur in the absence of calcification.” Redberg writes: “This apparent lack of predictive value of a CS should be enough to give a clinician pause,” especially in light of the radiation burden associated with the scan.
Until there are more outcomes data for CS with populations at different levels of risk, Redberg argues that “a CS of zero cannot be interpreted as a reassurance of the absence of CAD.” Calcium screening, she writes, “may yet have its place in the clinician’s arsenal for evaluation of patients with chest pain, but until its benefits are clearly established, we must take great care when subjecting patients to it.”
 
Response
PK Shah, MD
 
“This is an interesting study confirming prior reports that in a symptomatic population with a high pretest probability of CAD (high enough to warrant invasive coronary angiography) , a negative coronary calcium scan should not be considered reassuring in terms of ruling out obstructive CAD caused by a non-calcified plaque.This is quite different from usefulness of screening coronary calcium scan in asymptomatic  low or intermediate pretest probability group where zero coronary calcium predicts an extremely favorable intermediate term prognosis. We should be careful not to extrapolate the results of the current study to the asymptomatic low-intermediate risk group where coronary calcium scoring is likely to be most incrementally useful.”
Response
Matthew Budoff, MD
Coronary Calcium remains an effective filter for invasive angiography. In the paper by Gottlieb et al,  the authors question the guidelines and suggest that their small study trumps 25 years of published literature on coronary artery calcium. The guidelines were written based upon multiple studies that were 5-10 times larger than the cohort studied in this issue of JACC. Furthermore, the authors used a pre-determined cutpoint that would, by definition, limit the ability of CAC to predict stenosis. They excluded the 89 patients who were enrolled in CORE64 who had CAC >600.  Imagine a paper being published that states that hypertension does not correlate with left ventricular hypertrophy, but the study eliminated the 1/3 of patients who had significant hypertension from the study. Those 89 CORE64 patients, who had CAC scores >600 and angiography, should have been included in the analysis, and clearly that would have significantly changed both the sensitivity and specificity of the study findings.
Furthermore, the authors should not mention prognosis or implications of a negative score, as the literature, with studies published including over 100,000 patients with CAC, clearly demonstrate that a zero score carries an excellent long term prognosis. We followed patients for 8 years after admission to the emergency room, and no patients with a score of zero suffered coronary events.
A bigger issue is how the CORE 64 study is so divergent from almost all published literature on CAC.  Almost every published study of CAC, including multicenter trials involving over 2000 participants undergoing both CAC and invasive angiography, demonstrate a high sensitivity (>90%) and lower specificity (<50%).  Knez et al. studied 2,115 consecutive symptomatic patients (n = 1,404 men, mean 62±19 years old) with no prior diagnosis of CAD, finding CAC in over 99% of patients with obstructive CAD. No calcium was found in 7 of 872 men (0.7%) and in 1 of 383 women (0.02%) who had significant luminal stenosis on coronary angiography. Seven of these 8 patients with missed obstructive disease and scores of zero were <45 years old. However, the authors present almost exactly the opposite in their small study (a sensitivity of 45% and specificity of 91%), calling into question either the study design or equipment used. The scanner used in CORE64 was demonstrated in the Multi-Ethnic Study of Atherosclerosis to have a significantly worse reproducibility for CAC than all other scanners in the study, and these scanners have been systematically excluded from multicenter trials of lung disease, due to similar technical concerns. To reiterate the concern, results from the ACCURACY multi-center CT trial (using different CT equipment) demonstrated CAC to have a sensitivity of 94% and specificity of 42% for >50% stenosis by quantitative coronary angiography.
Coronary calcium with an effective radiation dose that approaches mammography, remains an effective filter for emergency room patients and those referred for invasive angiography in low to moderate pre-test probability patients. Given rising health care costs, we must strive for cost-effective and easy tests to stratify patients. The literature, with over 1000 published papers on CAC, is clear and even with one exception, consistent. CAC testing should remain a mainstay in both diagnosis and prognosis of the cardiac patient, with more attention on the type of CT scanner used to acquire the data.
 

Response
Rita Redberg, MD

As I stated in my JACC editorial, none of these articles cited by Budoff give us any information on how the calcium scoring adds incremental information to the traditional predictors for coronary artery disease, such as clinical assessment and stress testing. This apparent lack of predictive value of a calcium score should be enough to give a clinician pause. When combined with the significant radiation risks of coronary artery calcium scans, however, clinicians must use extreme caution when ordering such scans. Whether for a asymptomatic patien for “screening” or for risk stratification or diagnosis in a symptomatic patient, it is essential that the benefits of any test, eg CAC outweigh the risks for patients. That remains to be demonstrated for calcium scoring. CAC does not give us actionable information for screening, or for diagnosis, has the risk of radiation and additional testing to follow up on false positives. The best way to provide this data is via a randomized trial of calcium scoring versus traditional assessment in the populations in question. I hope we can see this soon. None of the many studies that Budoff refers to are randomized trials. The only randomized trial of CAC showed NO benefit (O’Malley JAMA).

Response
PK Shah, MD

I do believe that in selected asymptomatic subjects , identification of subclinical athero by carotid US or EBCT does provide useful information, but I do also agree that the only way to settle this issue once and for all is with a randomized trial and I believe the NHLBI should consider supporting such a trial.

Response
Matthew Budoff, MD

Coronary calcium scanning has been studied in more patients than any other modality in cardiology, save exercise testing. We have more robust data with CAC scanning than stress nuclear testing, stress echocardiography, and invasive angiography. It is true that there is a paucity of randomized controlled trials using one diagnostic strategy against another, but that is true for EVERY modality in cardiology, including Framingham risk assessment, exercise testing, and imaging. The only test validated to improve outcomes is abdominal aortic screening for aneurysms. So, should we abandon all tests due to lack of evidence, or continue to use best practices to take care of our patients? The argument can never be that there are no randomized trials, or cardiology would be paralyzed without any evidence that imaging of any sort improves outcomes. Any clinician recognizes the importance of imaging in practice, to sort out types of chest pain. Without imaging, we would be treating blindly. So far, we have over 16 studies that demonstrate that there is INCREMENTAL value to CAC testing, in that it predicts events better than Framingham, better than nuclear testing, better than stress echocardiography. This incremental value is key to our decision-making. Radiation dose is not high. Dr. Redberg and I published together a guideline in 2004 (Mieres et al.), in which we carefully outlined the low radiation doses with CAC testing (0.7 mSev for EBCT, 1.5 for MDCT). This is comparable to mammography (0.7 mSev per year for annual testing). Since CAC testing is not recommended to repeat over 5 years, the radiation dose for a female undergoing annual mammography is 3.5 mSev, significantly more than a single CAC test. Since more women die of heart disease than breast cancer, clearly the evidence is in favor of screening for heart disease with CAC testing. Remember, Dr. Redberg and I published in 2004 “Coronary calcium, in addition to traditional risk factors, provides independent incremental information in the prediction of cardiovascular outcomes…However for the asymptomatic women with risk factors and an intermediate Framingham risk score of 0.6%-2.0% or higher, screening with EBT can accurately provide evidence of subclinical atherosclerotic coronary disease.” Since data are now more robust after another 6 years of data accumulation, CAC testing remains a robust method of risk stratification, with low doses of radiation and ample clinical evidence.