Method to Overcome CAC Score Variability Hones Risk Stratification

A vendor-neutral Agatston score that addresses variability in coronary artery calcium (CAC) scoring across CT scanner types improves classification of an individual’s future risk of cardiovascular events, according to an analysis of data from the Multi-Ethnic Study of Atherosclerosis (MESA).

Participants reclassified to a higher risk level based on the vendor-neutral score experienced roughly double the rate of coronary heart disease events over a median follow-up of 16.7 years compared with those who remained in the low-risk category (15.3% vs 7.0%; P = 0.008), researchers report in a paper published online recently in JACC: Cardiovascular Imaging.

Moreover, the number needed to treat with new or intensified statin therapy to prevent CVD events was lower among individuals reclassified to a higher risk level using the vendor-neutral score.

“Going forward, I think there’s pretty clear evidence now that it would be a good thing to calculate this vendor-neutral score and use that as a basis for decision-making versus just grabbing the value that the scanner and its software gives you,” senior author Tim Leiner, MD, PhD (University Medical Center Utrecht and Mayo Clinic, Rochester, MN), told TCTMD.

Variability in CAC Scoring

CAC scoring has increasingly been included in practice guidelines as a way to assess patients’ risk for CV events on top of traditional risk factors, Leiner said. Prior work for his group, however, has shown that Agatston scores provided after a CT scan can vary widely between scanners from different manufacturers and even across machines of different types from the same vendor.

And this could have clinical implications. The value of CAC scoring and the thresholds that confer risk have been established in analyses of large cohort studies with long-term follow-up, but these studies used CT scanners with electron beam tomography, which is obsolete technology, Leiner explained. Now, multidetector CT systems are the standard.

“What was missing was a translation between the values you got from electron-beam CT scanners and the calcium scores we measure today,” Leiner said.

For the current study, led by Niels van der Werf, PhD (University Medical Center Utrecht, the Netherlands), and Magdalena Dobrolinska, MD (University Medical Center Groningen, the Netherlands), he and his colleagues first scanned anthropomorphic phantoms meant to simulate the human body using an older electron-beam CT scanner (as the reference) and five state-of-the-art multidetector CT machines. They also obtained data on two older multidetector CT scanners used in MESA.

From these scans, they calculated a conversion factor to translate the Agatston score obtained on the newer CT scanners to what it would have been if using the older electron-beam system—the result was the vendor-neutral Agatston score.

Better Risk Classification

The investigators then evaluated whether the vendor-neutral score led to better classification of risk of future cardiovascular events using data from MESA, which enrolled asymptomatic individuals from six US centers—baseline data were collected between July 2000 and September 2002, with follow-up lasting through December 2018.

The current analysis included data on 3,181 MESA participants (mean age 62 years; 52.5% women). Of those originally placed in the low-calcium group (Agatston score < 100), 11% were reclassified into a higher risk level based on a vendor-neutral Agatston score ≥ 100. The reclassified patients had a significantly greater risk of coronary heart disease events compared with those who remained in the low-calcium group and compared with those with a CAC score of 0 (HR 3.39; 95% CI 1.82-6.35).

The researchers then evaluated the potential impact of using the vendor-neutral score on initiation or intensification of statin therapy in 889 MESA participants (mean age 64 years; 47% women) deemed to have an intermediate cardiovascular risk. Of those with an Agatston score below 100, which would not trigger statin therapy, 11% were reclassified above that threshold with the vendor-neutral score, making them candidates for statin therapy. The reclassified individuals also had a higher rate of atherosclerotic cardiovascular disease (ASCVD) events during follow-up.

Among participants with an Agatston score of 100 to 299, already in the range for statin therapy, 28% were bumped into a higher risk category by the vendor-neutral score, making them candidates for high-intensity statins. These reclassified individuals, too, had a heightened risk of ASCVD events during follow-up.

In addition, the number needed to treat for statins to prevent CVD dropped in both groups of reclassified participants compared with those who remained at lower risk levels.

The findings indicate “that the vendor-neutral score better reflects the risk of these patients,” Leiner said. “That has implications for their treatment because when a cardiologist sees these patients, the decision of whether or not to initiate statin therapy and blood pressure-lowering medication and things like that is going to be partially dependent on this calcium score value.”

Integration Into Clinical Practice

Leiner noted that the conversion factors that allow for calculation of the vendor-neutral Agatston score, found in Figure 1 of the paper, are freely available to anyone who wants to use them. His team also is in discussions to get the conversion factors placed into the online MESA risk calculator.

In addition to getting a more-accurate onetime CAC score, use of the vendor-neutral score will allow physicians to better assess CAC progression for a single patient, who may be tested using different scanners over time, he explained. “Our tool is going to solve this problem because you can calculate the score back to basically a vendor-neutral score, so it takes into account those differences between scanners and corrects for them.”

Overall, Leiner said, “if you get your coronary calcium measured, I think you want to use this tool to calculate it back to a value that better reflects your real risk, and that can serve as a guide to clinical decision-making.”

This may be particularly important in parts of the world that don’t have access to the latest-generation CT scanners, Leiner said, estimating that about 90% of centers are still using older technology. “Especially in middle- and lower-income countries . . . this is going to be very relevant because they typically have those older scanners for measuring the coronary calcium score and that’s where currently the burden of cardiovascular disease is highest, in those regions of the world,” he added.