Coronary Calcium Controversy Continues

From Heartwire

Reed Miller

August 9, 2010 (Rio de Janeiro, Brazil) — Experts continue to debate the significance of a Coronary Evaluation Using Multidetector Spiral Computed Tomography Angiography Using 64 Detectors (CORE64) substudy, in which the absence of coronary calcification did not exclude the possibility of obstructive coronary disease in symptomatic patients [1].

On the basis of the results of the 291-patient study, authors Dr Ilan Gottlieb (Federal University of Rio de Janeiro, Brazil) and colleagues concluded that a zero coronary artery calcium score (CACS) with computed tomography (CT) should not be "used as a gatekeeper" to prevent symptomatic patients from undergoing angiography. As reported by heartwire , the results of the study sparked a debate about the utility of CT calcium scoring after they were published in the February 16, 2010 issue of the Journal of the American College of Cardiology.

That debate continues in five letters to the editor published in the August 10, 2010 issue of that journal. The letters are both critical of and favorable toward the CORE64 substudy and the accompanying editorial by Dr Rita Redberg (University of California, San Francisco), which argued that the incremental value of calcium scoring over traditional coronary disease predictors alone has not been demonstrated [2].

In their letter, Drs Luis Correia and Fabio Esteves (Medical School of Bahia, Salvador, Brazil) argue that the low negative predictive value of 68% found in the CORE64 study is "strikingly different from that reported in most trials" [3]. Likewise, Dr Matthew Budoff (Los Angeles Biomedical Research Center, CA) points out that Gottlieb et al's conclusions appear to contradict current professional guidelines for coronary calcium scoring and more than 1000 studies, including several that were five to 10 times larger than this CORE64 substudy [4].

Correia and Esteves contend that Gottlieb et al's calculations of sensitivity, specificity, negative predictive value, and positive predictive are in error. They argue that Gottlieb et al defined a negative test result as a CACS >0, when negative results should suggest only no disease. "If a negative test result was appropriately defined as a zero CAC score, the actual negative predictive value (number of patients without stenosis >50% divided by the number of patients with zero calcium score) would be 81% (58 of 72).

In their response [5], Gottlieb et al explain that the predictive values in their paper "have slightly different meanings than commonly utilized in other trials, [because] we chose to take a different perspective." Gottlieb et al state that their goal was to determine whether a CACS of zero could predict the absence of obstructive coronary artery disease, whereas the other trials referred to by the letter writers examined whether the presence of calcium increases the likelihood of chest pain being related to significant stenosis.

"Using our approach of calling a zero CACS a positive scan, the positive predictive value refers to the ability of zero calcium to rule out obstructive CAD. This is in fact the same message of a negative predictive value using the 'conventional' approach." So the predictive value was low in their study (68%) and the sensitivity of zero calcium to detect the absence of disease and rule out obstructive CAD was also low, at 45%, the authors explain. "When our results are interpreted from this perspective, they are clearly consistent with previously published studies."

In their letter to the editor, Dr Michael Blaha, Dr Roger Blumenthal, and Dr Khurram Nasir (Johns Hopkins University, Baltimore, MD) defend the approach used by Gottlieb et al and disagree with the statement in Redberg's editorial that the results are "starkly" different from those of previous studies [6]. They point out that when the differences in patient population are taken into account and the predictive values are calculated in the same way, the negative and positive predictive values in the CORE64 data are not much different than those found in previous studies.

Gottlieb et al argue that their approach is a more accurate test of the utility of calcium scoring when used specifically to rule out obstructive disease in symptomatic patients in order to discharge them from the emergency department or to guide outpatient investigation of their chest pain.

What Is the Value of Calcium Scoring?

The CORE64 authors stress that it would "be a grave mistake" to generalize their findings for all subgroups, because their study focused only on symptomatic patients. Therefore, Budoff's contention that their study contradicts more than 1000 studies with more than 100 000 patients "regrettably misses the fact that the vast majority of the published CACS literature refers to asymptomatic patients."

Redberg's editorial "took a broader view than our data warrant," Gottlieb et al argue in their response. "We acknowledge the role CACS has for risk stratification in selected symptomatic populations as well as in epidemiologic studies of atherosclerotic disease."

Gottlieb told heartwire that "it's very inappropriate to use the calcium score to rule out obstructive coronary disease, especially among acute chest-pain patients . . . because it depends very heavily on the prevalence of disease that you're looking at."

"We're not talking about asymptomatic patients. [Those patients] benefit a lot from calcium scores. It's a really great thing to do among certain groups of asymptomatic people," Gottlieb said, observing that some of the letters to the editor appear to conflate the asymptomatic and symptomatic populations. "This is a major mistake. It is a major division. If the patient is asymptomatic, the investigation algorithm should be very different than if the patient is symptomatic."

In her reply to the letters to the editor [7], Redberg points out that none of the letters "address the key clinical point of whether an imaging test such as coronary artery calcium will give us new information that leads to better patient care and improved outcomes. . . . Despite the use of CACS for the past 20 years, there are still no data for either the asymptomatic or symptomatic group to show that this information benefits our patients."

She points out that the most recent US Preventive Services Task Force recommendation statement on congestive heart disease risk assessment concludes that there is still not enough evidence to weigh the benefits and harms of using "nontraditional risk factors" to screen asymptomatic men and women with no history of coronary disease, and that the task force stated that there remains a critical gap in the evidence for screening with coronary artery calcium scoring because of the lack of information on how this type of screening will ultimately reduce the risk for coronary adverse events.

"Before subjecting healthy men and women to a test with significant radiation--2 to 7 mSv or 100 chest roentgenograms--one must be able to tell patients that there is a benefit from having this test," Redberg argues. "With no known benefit, CACS fails this essential criterion, and the harm, including cancer risk from radiation, and incidental findings prevail."

References