Coronary Artery Calcium: Score? or No More?

Maybe it's not the $64,000 question, but increasingly cardiologists are asking: What is the role of coronary artery calcium (CAC) scoring in 2014? On the surface, it's not a tough question to answer. After all, CAC has been around since at least 1990, and we've seen dozens of studies trumpeting its predictive value. At the 2014 ACC Annual Scientific Sessions, an hour-long press conference was dedicated to CAC—not to announce a new way of measuring it or a change in clinical guidelines, but to present yet more research demonstrating its value as a predictor of future events across a range of individuals.

And the Winner Is...
There's no mystery about CAC score's popularity as a high-tech screen. Cardiovascular disease stubbornly stays the leading cause of mortality worldwide—and now the new #1 cause of years of life lost—despite the fact that effective preventive methods exist.

Yet cardiovascular risk prediction remains an imprecise science. Data from the Framingham Heart Study and other population-based cohorts have convincingly demonstrated that age, sex, cigarette smoking, low-density lipoprotein cholesterol (LDL-C) and total cholesterol levels, diabetes, and blood pressure can be combined in predictive models to estimate risk of future coronary heart disease (CHD) events. However, conventional risk factor assessment using algorithms such as the Framingham Risk Score (FRS) and Framingham-like risk scores underestimate risk, particularly in women and in younger patients. Since a substantial proportion of cardiovascular death first manifests as acute myocardial infarction or sudden cardiac death, high-risk patients not captured by current risk scores are left unacceptably vulnerable. Enter CAC.

CAC can be measured using chest computed tomography (CT) in about 5 minutes.1 Radiation exposure can be limited to about 0.9-1.1 mSv, although lacking careful controls, the radiation dose often surges significantly higher.2 But what's a little radiation exposure in the presence of calcification that is pathognomonic for atherosclerosis? Conversely, a score of zero is clear assurance of an extremely low CHD event risk. Wouldn't the patient—almost any patient—want to know?

Coronary calcium rules as a power player in assessing risk. "It's not a risk factor, it's the actual disease," said Robert O. Bonow, MD, in an interview with CardioSource WorldNews. "So, it's not surprising that it turns out to be a good predictor of events." For every calcified plaque found in a patient's coronaries, Dr. Bonow said there are many noncalcified plaques, too, such that the calcium burden measured by CAC probably equals about one-fifth of the total atherosclerosis burden.

Dr. Bonow should know: not only is he the Max and Lilly Goldberg Distinguished Professor of Cardiology at Northwestern University Feinberg School of Medicine, Chicago, and a past-president of the American Heart Association, he was a member of the writing committee of the ACCF/AHA 2007 Clinical Expert Consensus Document on CAC scoring.3

And how many other tests can make this claim: there are no false positives with CAC scanning and very little measurement error. "Using properly calibrated instruments and standard software, if I test my own CAC on machine A, machine B, and machine C, I should get pretty much the same answer from all three tests," Philip Greenland, MD, told CSWN.

Dr. Greenland's qualifications in CAC are among the best, too. Also from Northwestern University, he is a professor of cardiology and chaired the aforementioned ACCF/AHA 2007 Consensus Document on CAC scoring3 as well as the 2010 ACCF/AHA Guidelines for Assessment of Cardiovascular Risk in Asymptomatic Adults.4 For the most recent 2013 ACC/AHA Guideline on the Assessment of Cardiovascular Risk, Dr. Greenland was a writing committee member.5 He has studied and written about CAC extensively, and has used it in his own practice for many years.

In short, no matter which way you slice it, most agree that CAC offers the best discrimination of all the second-tier assessment techniques available—an opinion that is echoed in the recent guidelines. In several studies, CAC provided additional risk stratification when added to traditional risk scores, testing better than C-reactive protein (CRP) and carotid intima-media thickness (CIMT).

"It feels to me that we can say pretty definitively that if we were going to line up all the currently available risk markers that are not traditional risk factors, the winner is still CAC," said Dr. Greenland. Ding, ding, ding—we have a winner. But a pyrrhic victory, it would seem. Right after calling it a winner, Dr. Greenland questioned the value of CAC in today's heath care environment where many deem medical imaging seriously overdone—including those who pay for this high-tech insight. Cost remains one concern, of course, but so is radiation risk. That's two strikes against a technology where any possible large clinical outcomes trial proving efficacy is unlikely to be funded, and where the threshold for preventive action (namely statin therapy) has been lowered by the newest guidelines. Who's left to fall through the cracks when such a large share of the adult population are now considered "at risk" and candidates for preventive therapy?

Not Worth the Cost or Risk
Coronary calcium scanning draws naysayers as well. Steven Nissen, MD, chairman of cardiovascular medicine at the Cleveland Clinic in Ohio, has long thought CAC scanning does more harm than good. "Perfectly healthy people are getting CAC scans all over the country and their doctors then say, 'Well, you've got all this calcium in your coronaries, maybe we better do a heart catheterization, or we should do a thallium scan,' and the next thing you know they're in the cath lab, sometimes even getting stents," said Dr. Nissen. "I see patients continuously in my clinic who have been harmed by getting an inappropriate CAC scan that leads to unnecessary procedures, and it's just bad medicine."

What about "seeing is believing"? Dr. Nissen categorically rejects the idea of using CAC scanning as a means of prompting patients to adhere more closely to lifestyle advice or preventive treatments. "Exposing people to radiation in order to motivate them... take a deep breath and think about that," he said.

Dr. Nissen's preferred method of motivation: one-on-one discussions about individual risk factors and what patients need to do to avoid disease. "I don't have to stick them in a calcium scanner and expose them to possibly fatal radiation in order to have that conversation."

To be fair, Dr. Nissen does believe CAC is a good predictor of future events and has advocated for the test in at least one setting: he was part of a committee of cardiologists who recommended to NASA that CAC be included in the evaluation protocol for astronauts entering the Astronaut Corps.

"Our government invests millions of dollars in training these astronauts to go on a few space missions 10 and 20 years down the line. Before you make that investment, you need to have a good idea that their coronaries are likely to be clean. In this setting, the benefits of CAC exceed the risk, particularly when you consider that the amount of radiation they get in one spaceflight dwarfs what you get from a scan."

Another complaint lobbed against CAC is the lack of a randomized clinical trial showing that its use actually improves outcomes. Several groups, including one headed by Dr. Greenland, have tried to fill this gap, but the necessary trial would be both large and prohibitively expensive. "We tried to get a study like that funded but we estimated we'd need to enroll about 30,000 patients and it would cost between 80 and 100 million dollars to conduct," said Dr. Greenland. "In the current environment, that is unlikely to get funded."

Blue Light Special?
Everyone interviewed for this article agreed that CAC has no role as a widespread screening modality in individuals at low risk. The fact that numerous centers in the United States persist in offering cheap calcium scans without appropriate pre-screening has clearly damaged the test's reputation in the wider cardiology community.

"It's terrible medicine," Dr. Nissen flatly declared, especially when centers use billboard ads to pitch wives on a unique Valentine's Day gift for their husbands' hearts: Get his coronaries scanned for 99 bucks. Think of it as medicine's "blue light special." "It's a loss leader to get people into the cath lab in order to have unnecessary and inappropriate procedures," said Dr. Nissen—an opinion echoed by Dr. Greenland. Although these centers require, by law, a physician's referral, concerns persist that the referrals are rubber stamps and the scans are being massively overused.

Or not: Dr. Bonow suggests that cardiologists may be overstating the case. In the early days, CAC scans were being used "in some circles" to drive more cardiovascular imaging and more procedures, but this is not really the case anymore. "With time, I think the field has matured and we realize the value of this test when used properly," he said.

"Of course, it's not the test itself that changes things; it's what the doctor or the patient does with the information that changes things," added Dr. Bonow. Evidence suggests that even when confronted with relatively advanced atherosclerosis, it's hard to motivate people in a free-living, asymptomatic, otherwise healthy population to change their behavior. Getting a CAC score starts the process to help an individual patient.

CAC in the Guidelines
Ultimately, we look to the guidelines to guide, and even here it's hard to decipher whether or not you should send your at-risk patients for a CAC scan.

In the 2010 ACCF/AHA Guideline for Assessment of Cardiovascular Risk in Asymptomatic Adults, CAC received a Class IIa recommendation in intermediate-risk patients (10% to 20% 10-year risk) and a Class IIb recommendation in low-to-intermediate risk patients (6% to 10% 10-year risk based on traditional risk factors).

The new 2013 ACCF/AHA Guideline for the Assessment of Cardiovascular Risk does away with the "intermediate-risk" category and lowers the bar for treatment to a lower risk threshold of 7.5% 10-year risk for a first hard atherosclerotic cardiovascular disease event (which includes stroke, as opposed to just cardiovascular risk, as in previous guidelines), and drops CAC scanning to a Class IIb recommendation.

The new guideline states: "If, after quantitative risk assessment, a risk-based treatment decision is uncertain, assessment of one or more of the following—family history, hs-CRP, CAC score, or ankle brachial index—may be considered to inform treatment decision making."

Dr. Greenland explained that the guidelines "downgraded" CAC not because the authors uncovered new information suggesting that CAC was not a useful test. To the contrary, he feels that the data supporting the role of CAC for risk prevention have "only gotten stronger and stronger" in recent years and that CAC's predictive value has actually been very well established. "People who know the risk factor prediction data would agree that it is the strongest additional marker and the only one that really substantially adds to the traditional risk factors," he said.

Essentially, the downgrade occurred courtesy of a technicality, explained Dr. Greenland: "The new guideline came out with a recommendation for a new risk score and that risk score was derived from a number of datasets that needed to have all the markers measured in those datasets, including the ARIC study, MESA, and the Framingham study. The ARIC study did not have CAC, so by definition CAC could not have possibly gotten into the new risk score."

He doesn't place "too much weight on it," given that the new risk score doesn't really recommend any novel risk factors. CIMT and CRP were also excluded—not because of the same technicality but because new data showed that they actually failed to add predictive value.

Said Dr. Greenland, "From my point of view, the biggest challenge facing any new risk marker today—and CAC is a good example—is that if you have a recommendation based on a traditional risk factor score that says that pretty much anybody with a risk score above 7.5% should be treated with a statin, then why exactly do you need any additional risk marker?"

He said some clinicians may continue to use CAC, as he has in the past, as a "tie-breaker." And while he thinks that's an appropriate use, it's harder to justify with the new guidelines.

Just Why Are We Testing CAC Again?
An intriguing cost-effectiveness analysis published in March 2014 (with Dr. Greenland as senior author) offered a slightly different view on the CAC question.6 In short, the dollars don't add up.

He and his team estimated the costs and effectiveness of using CAC scanning to guide primary prevention with statin therapy using an established decision model (The University of North Caroline-Research Triangle Institute Cardiovascular Disease Prevention Model). The prevalence of CAC and expected distribution of the CAC score was estimated using MESA data and also offered a range of assumptions about statin therapy and CAC scan cost, effectiveness, and adverse effects.

With that, they found that 10 years of statin treatment for 10,000 55-year-old women with high cholesterol and a 10-year CHD risk of 7.5% would prevent 32 myocardial infarctions, cause 70 cases of statin-induced myopathy, and add 1,108 years to total life expectancy. Conversely, measuring CAC and targeting statin treatment only to the 2,500 women with CAC >0—the estimated prevalence in this MESA cohort—would provide 45% of the benefit (+501 life-years), but at a steep price tag: $2.25 million for testing and 9 radiation-induced cancers.

"What we showed, quite robustly I think, was that if you're in an environment where statins are cheap and don't cause a lot of adverse effects, you can skip the [CAC] test and just treat everyone, even down to a risk threshold of 2.5%—well below the 7.5% treatment threshold given in the new guideline," noted Dr. Greenland.

Here's the rub: This is based on favorable statin assumptions (i.e., the drugs cost $0.13/pill, are available in discount pharmacies across the United States, and there was no measurable quality of life [QOL] penalty from the treatment). That is a serious set of assumptions, given that a good number of patients will have QOL issues on statin therapy and may require a statin that costs more than a few cents per pill in order to limit these drug-related issues.

If you put it that way, how fair are those highly favorable assumptions? What about less favorable "assumptions?"

Dr. Greenland and colleagues looked at that, too. They considered a single-dose cost of $1.00 and a QOL hit that was equivalent to trading away 2 weeks of perfect health to avoid 10 years on statins. Using these less favorable parameters, CAC screening with targeted treatment only for persons with CAC >0 was cost effective (<$50,000 per quality-adjusted life-year) in most scenarios where CHD risk was intermediate (5% to 10%), but with some variation by age and gender.

Dr. Greenland explained that statin costs actually range from free to probably a few dollars a day. Several discount pharmacy outlets offer $3 or $4 per month statin drug programs, although most Americans do not avail themselves of them. Going one better, some supermarket chains actually dispense generic statin drugs for free.

"They do this because the generics are so inexpensive and there are so many patients on statins, that if they give it away for free, they end up getting the rest of that patient's pharmacy business and it's worth it for them," he said.

After being both a guidelines author and a CAC proponent, perhaps Dr. Greenland's conclusions are as clear an answer as we're likely to get without a large-scale clinical trial of CAC: "I've spent a lot of time looking at CAC and have written a lot of papers on it and I've been very enthusiastic about it. But when you, number one, lower the threshold for treatment to the point that almost everybody becomes a statin candidate as we've done in the new guidelines, and, number two, look at the data on cost effectiveness, where, for just about everyone above a 2.5% risk threshold, statins become cost effective under most favorable conditions—in this environment, I think it starts to become very challenging to make a case for any additional tests."


  1. Polonsky RS, Greenland P. Nat Rev Cardiol. 2012;9:599-604.
  2. Smith-Bindman R, Lipson J, Marcus R, et al. Arch Intern Med. 2009;169:2078-86.
  3. Greenland P, Bonow RO, Brundage BH, et al. J Am Coll Cardiol. 2007;49:378-402.
  4. Greenland P, Alpert JS, Beller GA, et al. J Am Coll Cardiol. 2010;56:e50-103.
  5. Goff Jr DC, Lloyd-Jones DM, Bennett G, et al. J Am Coll Cardiol. 2014;63:2935-59.
  6. Pletcher MJ, Pignone M, Earnshaw S, et al. Circ Cardiovasc Qual Outcomes. 2014;7:276-84.

< Back to Listings