Absence of Coronary Calcium Reclassifies Nearly 50% of Statin-Eligible Patients: MESA Analysis

October 06, 2015

MIAMI, FL — Two new analyses of the Multi-Ethnic Study of Atherosclerosis (MESA) provide evidence that the use of coronary artery calcium (CAC) screening can reclassify patients at risk from cardiovascular disease and better identify those who would most benefit from statin therapy compared with current guidelines.

In the first study[1], which was led by Dr Khurram Nasir (Baptist Health South Florida, Miami), investigators found that 50% of 4758 MESA participants would be recommended for moderate- or high-intensity statin therapy based on the 2013 American College of Cardiology/American Heart Association (ACC/AHA) clinical guidelines. Moreover, another 12% of patients could be "considered" for statin therapy based on their 10-year 5% to 7.5% risk of atherosclerotic cardiovascular disease (ASCVD).

Taken together, the researchers say that nearly two-thirds (62%) of the MESA cohort would be recommended or considered for statin therapy.

Yet among those recommended statins—this included patients with an LDL-cholesterol level >190 mg/dL, diabetic patients with an LDL-cholesterol level 70–189 mg/dL, and nondiabetic patients with an LDL-cholesterol level 70–189 mg/dL and an estimated 10-year ASCVD risk >7.5%—more than 40% of these individuals had a CAC score of zero and an ASCVD event rate of 5.2 per 1000 person-years of follow-up. Among those who could be considered for statin therapy, 57% of these individuals had a CAC score of zero and an ASCVD event rate of 1.5 per 1000 person-years of follow-up.

Of the entire group eligible for statins, either recommended or considered, 44% of the 2966 patients had a CAC score of zero at baseline and a 10-year ASCVD event rate of 4.2 per 1000 person-years.

"Overall, our results showed that nearly 50% of individuals who are statin candidates, if they undergo a test that costs between $75 and $100, would have a calcium score of zero, and their 10-year risk would be below the threshold in which the guidelines recommend statins," Nasir told heartwire from Medscape. "This is most important for the patients in the middle, those who have a 10-year risk of 5% to 20%."

In contrast, 56% of patients considered statin eligible based on the guidelines had coronary calcification identified on the noncontrast computed tomography (CT) scan (CAC>0) and had an ASCVD event rate of 11.2 events per 1000 person-years.

The results of the study, along with another analysis led by Dr Robyn McClelland (University of Washington, Seattle) that included CAC scores in a risk-prediction algorithm, are published in the October 13, 2015 issue of the Journal of the American College of Cardiology.

Deep Dive Into the Benefits of Adding CAC

Speaking with heartwire , Nasir said the recent change in statin guidelines—one that moved away from LDL-cholesterol targets to focus on risk—expanded the number of eligible patients. In their analysis, of the 2377 MESA participants who are candidates for statin therapy, 77% would be recommended the drugs because their 10-year risk of ASCVD was 7.5% or greater. While he said the focus on patient risk is commendable, statins, despite being cost-effective and safe, are a lifelong therapy for patients.

"For me, I don't want to be on a pill, however safe, cost-effective, or cheap it may be, if I'm not at high risk," said Nasir.

In explaining the rationale for their study, Nasir said the hypothesis was based on data from previous studies, including work they've done in the past, showing that a significant proportion of individuals, irrespective of their estimated 10-year risk of cardiovascular events, had no detectable calcium on the CT scan. Their goal was to assess whether the absence of CAC could help reclassify patients who might not need statin therapy.

Practically speaking, Nasir said a patient with a calculated 10-year ASCVD risk of 12% would be recommended to start statin therapy based on the ACC/AHA clinical guidelines. However, based on their analysis, one out of every two such patients would have a CAC score of zero, meaning their actual risk of cardiovascular disease is threefold less, on average, at approximately 4%. Such a patient, given their lower adjusted risk with the addition of CAC, might not choose to take a statin.

"Our thought is that in 2015, when a large proportion of US adults are at least candidates for statins, the value of screening to find additional individuals to treat becomes less compelling," he said. Instead, information provided by the absence of CAC to assure patients that their risk of events is not as high as their 10-year risk score would indicate is more valuable, said Nasir.

Speaking anecdotally, Nasir said that nearly every patient they encounter with a CAC score of zero opts to avoid statins, based on evidence of low risk, instead focusing on diet and lifestyle. "This is very different from a screening test where you identify a disease to treat," he said. "Here, we're using CAC to support uncertain patients to make an informed decision for avoiding pharmacotherapy."

He added that the value of CAC is not in universal screening—which would needlessly test low- and high-risk patients—but rather as a sequential process in intermediate-risk patients where testing is used as part of a shared decision-making tool. Used in this way, CAC can provide information to identify the 50% of patients who are truly at lower risk to allow greater flexibility in treatment (such as emphasizing diet and lifestyle changes instead of drugs), avoid overtreatment, and maximize the yield of the intervention in higher-risk groups.

"It's testing as a gatekeeper where we provide some choice for the patient," said Nasir.

Improving Risk Prediction With CAC

In the second study[2], investigators created and validated a novel risk-score algorithm within MESA to assess the 10-year risk of coronary heart disease (CHD) using traditional risk factors and CAC. After 10.2 years of follow-up, there were 422 CHD events, including 68 CHD deaths, 190 fatal MIs, 149 angina-driven revascularizations, and 15 resuscitated cardiac arrests.

Using traditional risk factors alone, the researchers first developed a risk score without CAC. For this internally developed risk-scoring algorithm, the area under the survival receiver-operating characteristic (ROC) curve was 0.76. When CAC was added, the area under the survival ROC curve was 0.81, which investigators say indicates excellent discrimination between events and nonevents and a significant improvement over the risk model that used traditional risk factors alone.

Evaluation of MESA Risk Model

Risk model MESA (n=6726) HNR (n=3692) DHS (n=1080)
Model (risk factors only)
C statistic 0.75 0.72 0.78
Model (risk factors plus CAC)
C statistic 0.80 0.78 0.82
HNR= Heinz Nixdorf Recall Study
DHS=Dallas Heart Study

The researchers also measured the discrimination slope for their risk models. With CAC, the discrimination slope was 0.086, meaning those with CHD events had a 10-year predicted risk 8.6% higher than those who did not have events. In the risk model without CAC, the discrimination slope was 0.052, meaning individuals with events had a 10-year predicted risk only 5.2% higher than those without events.

The researchers validated the risk model in the Heinz-Nixdorf Recall (HNR) study and the Dallas Heart Study (DHS).

CAC Not Without Cost or Risk, Says Editorial

In an accompanying editorial[3], Dr Javier Sanz (Mount Sinai School of Medicine, New York) says the CAC study by Nasir et al addresses a critical issue and provides solid direction.

"In an environment with limited resources, efforts should be made to identify not only newer indications for therapy but also those who may not significantly benefit from it," he writes. The absence of CAC may represent such a tool to achieve that goal, adds Javier.

Despite this, Javier argues against measuring CAC screening in all intermediate-risk patients, noting CAC testing is not without cost or risk. Also, if intermediate-risk patients are willing to take statins, there is little reason to stop them from doing so. He also notes that the use of screening tools are "justified" when it triggers an intervention of proven benefit in a given context. In this case, investigators appeared to use CAC screen to screen out patients for therapy.

In a second editorial[4], Dr Donald Lloyd-Jones (Northwestern University Feinberg School of Medicine, Chicago, IL) notes that CAC does a good job of identifying those likely to have events in low- and intermediate-risk subgroups because CAC represents the actual disease of interest rather than risk factors or surrogates.

"But, to date, no national evidence-based guidelines have adopted a 'disease-screening' approach," he writes. "The current paradigm remains fixed on 'risk-based' screening." One of the concerns about including CAC scores into risk-prediction algorithms would be the implicit assumption of universal screening.

"The lack of a randomized screening trial demonstrating the efficacy, utility (including potential adverse events), cost-effectiveness, and net clinical benefit of CAC screening, particularly in intermediate-risk patients when there is uncertainty regarding decisions, is a substantial barrier to widespread adoption," writes Lloyd-Jones. "Without these data, current clinical practice guidelines cannot provide strong evidence-based recommendations to guide practice."

To heartwire , Dr Harlan Krumholz (Yale University School of Medicine, New Haven, CT), the senior investigator of the trial led by Nasir, is confused by the criticisms that point to a lack of randomized trial data with CAC screening. He noted that there was no randomized trial data testing any of risk calculators and added that CAC is not being used as an outcome, one that would need to be studied to determine the effects of treatment. "In this study, we are testing it as a biomarker that can add substantially to risk stratification," noted Krumholz.

"And even people with low risk may choose to be treated with statins, and for those people the scan would not be useful," he added. "The study suggests that for people who would make different decisions about medications if their risk were lower, then this might be a useful strategy."

Nasir is on the advisory board for Quest Diagnostic and is a consultant for Regeneron. Krumholz is the recipient of research agreements from Medtronic and Johnson & Johnson (Janssen) through Yale University to develop methods of clinical-trial data sharing and is chair of a cardiac scientific advisory board for UnitedHealth. Disclosures for the coauthors are listed in the paper. McClelland reports no relevant financial relationships; disclosures for the coauthors are listed in the article. Lloyd-Jones is a MESA investigator and reports no disclosures relevant to the contents of the paper. Sanz reported no relevant financial relationships.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.