USPSTF: Insufficient Evidence to Support CAC Score, hsCRP, or ABI for Risk Assessment
The group’s decision, however, left some doctors “bewildered,” particularly when it comes to using CAC screening for risk stratification.
There is insufficient evidence to currently assess the benefits and risks of adding nontraditional cardiovascular risk factors—ankle brachial index (ABI), high-sensitivity C-reactive protein (hsCRP), and coronary artery calcium (CAC) screening—to traditional methods of risk assessment in asymptomatic adults, according to new recommendations from the United States Preventive Services Task Force (USPSTF).
The USPSTF concluded there aren’t enough sufficiently powered clinical trials to evaluate the incremental effect of the three methods for assessing risk and starting preventive medical therapy. In their review of the data, the task force found that adding ABI, hsCRP levels, or the CAC score to existing risk assessment models resulted in only small improvement in discrimination and reclassification and that the clinical meaningfulness of these changes is largely unknown.
“The USPSTF found inadequate evidence to assess whether treatment decisions guided by the ABI, hsCRP level, or CAC score, in addition to risk factors in existing CVD risk assessment models, leads to reduced incidence of CVD events or mortality,” according to the task force members in a recommendation statement published online July 10, 2018, in JAMA.
Those conclusions, however, struck a nerve with several cardiologists, particularly when it came to the task force’s stance on CAC screening.
“I’m bewildered,” Michael Blaha, MD (Johns Hopkins Medicine, Baltimore, MD), told TCTMD. He noted that the latest USPSTF report is an update to the 2009 recommendations on the use of nontraditional risk factors for coronary heart disease, but fails to account for new data showing CAC scoring can reclassify patients when used with traditional risk assessment tools, such as American College of Cardiology/American Heart Association (ACC/AHA) pooled cohort equations. “It’s almost as if nothing has happened in the last 9 years,” he said.
The CAC score can be an important tool for refining patient risk, argued Blaha. Data from the Heinz Nixdorf Recall study, the Jackson Heart study, the Multiethnic Study of Atherosclerosis (MESA), and BioImage, among others, have shown that noninvasive imaging with CAC testing has the potential to stratify cardiovascular risk in statin-eligible patients. Moreover, Blaha said it’s important to emphasize CAC is not a universal screening tool, but rather an additional means to help physicians make informed decisions about patient risk.
“Screening is something you do in the general population at large,” said Blaha. “Risk assessment is something you do in a select group of patients for whom you need more information. If somebody is very young and very low risk, or if they have already had heart disease, I’m not going to order a calcium score. I order it for the purpose of risk stratification. To me, the [USPSTF] doesn’t differentiate between screening and risk assessment.”
Blaha agreed there is no evidence to support CAC scoring as a widespread screening test for the general population, and that it should not be used in this manner, but said “there is plenty of evidence that says if you’re unclear about the risk, and it might change your management, then this test is helpful.”
Khurram Nasir, MD (Yale University School of Medicine, New Haven, CT), who also is frustrated with the USPSTF recommendations, said several studies published in the last few years have shown that the ACC/AHA pooled cohort equations overestimate risk in certain populations. Additionally, data suggests that as many as 33% to 50% of patients classified as candidates for statin therapy (those with a 10-year risk of atherosclerotic cardiovascular disease > 7.5%) have a CAC score of zero. With the addition of CAC, these patients have a much lower risk of clinical events and do not need to be treated with a statin.
“This is well established now,” said Nasir. “An absence of calcium—a calcium score of zero—is powerful data showing that you can stop statins. Nothing means something.”
USPSTF Approaches Evidence
For the USPSTF evidence review, which was also published in JAMA, Jennifer Lin, MD (Kaiser Permanente, Portland, OR), and colleagues analyzed 43 studies that included 267,244 individuals to assess the benefits and harms of ABI, hsCRP, and CAC score for cardiovascular risk assessment.
Specifically, the group sought to address several questions: 1) Does risk assessment of asymptomatic adults using nontraditional risk factors, followed by treatment specific to that risk, reduce the incidence of cardiovascular events compared with the pooled cohort equations or Framingham risk score? 2) Does use of nontraditional risk factors improve measures of calibration, discrimination, and risk reclassification? 3) Does treatment guided by nontraditional risks factors, in addition to traditional risk factors, reduce the incidence of cardiovascular events?
In their evidence review, Lin and colleagues found that ABI, hsCRP, and CAC score can improve various risk assessment performance measures, including discrimination and risk reclassification, but the magnitude and consistency varies by risk factor. They concluded that adding CAC score had the largest improvement in discrimination (how well a test predicts who will and will not get CVD) and reclassification of risk, but found that CAC scoring might inappropriately reclassify individuals not having CVD into higher-risk categories. With hsCRP, they found a minimal effect on CVD risk prediction.
Regarding clinical endpoints, the USPSTF noted there were no trials comparing treatments guided by nontraditional risk factors (in addition to traditional risk factors) versus usual care or no treatment.
To TCTMD, Nasir said if CAC was used as a universal screening test, then a randomized clinical trial would be needed to prove that its use to guide treatment would improve clinical outcomes, such as MI, stroke, or mortality, but that’s not how the test is used by practicing physicians. “Unfortunately, the USPSTF is still stuck in the old screening paradigm,” said Nasir. “We have moved on from there. I think it’s time they move on from there, too. It’s a decision aid, not a screening tool, and if they look at it through that prism, they’d understand the value it imparts.
Paul Ridker, MD (Brigham and Women’s Hospital, Boston, MA), who has been at the forefront of investigating the role of hsCRP in cardiovascular disease, told TCTMD the USPSTF report addresses an “outdated question and leads to an unfortunate clinical paradox.”
“We have definitive evidence from CANTOS that lowering hsCRP in secondary prevention lowers cardiovascular risk, and we have definitive evidence from JUPITER in primary prevention that patients with elevated hsCRP benefit markedly from statin therapy,” he said. “By contrast, we have no data in primary or secondary prevention that raising HDL cholesterol improves outcomes. And yet since HDL cholesterol it is part of a ‘traditional’ risk panel, universal HDL screening is fully endorsed by the USPSTF without even asking the crucial questions of ‘how much does it add?’ and ‘is there biologic evidence that altering it matters for patient care?’”
For Ridker, the only way to ensure progress in cardiology is to make an “apple to apples comparison,” which would include comparing the performance of newer versus older cardiovascular risk factors.
More on CAC Scoring
In an editorial, John Wilkins, MD, and Donald Lloyd-Jones, MD (Northwestern University Feinberg School of Medicine, Chicago, IL), write that the USPSTF’s conclusions are understandable from a policy perspective but don’t fully address issues faced by individual patients and doctors when trying to decide on the relative merits of preventive therapy.
Like Nasir and Blaha, they argue that when used selectively, and not universally, in patients with an estimated 10-year risk around the treatment threshold (approximately 5% to 15%), there is good evidence CAC scores reclassify many patients to accurate risk levels.
“I think USPSTF has stringent criteria for their recommendations, and they are answering policy-level questions about universal screening with these tests,” Lloyd-Jones told TCTMD. “What clinicians and patients want to know is: will any of these tests change decisions in a meaningful way for an individual patient? In that context, I agree that CRP and ABI do not assist in reclassifying risk meaningfully after use of a risk score for primary prevention of atherosclerotic CVD. But CAC scoring has consistently shown it can reclassify risk in meaningful ways to sharpen our decision-making around the need for preventive medications like statins.”
In a second editorial, which was published in JAMA Cardiology, Tamar Polonsky, MD (University of Chicago, IL), and Philip Greenland, MD (Northwestern University Feinberg School of Medicine), also make the case for the use of CAC scoring in selected patients. Like others, they agree that CAC scoring should not be used as a widespread screening tool, noting the noninvasive test is less informative in patients deemed low- or high-risk based on traditional risk scores.
“Coronary artery calcium testing is more helpful when used selectively for patients in whom treatment decisions are unclear,” they write.
Polonsky and Greenland also point out that the ACC/AHA clinical guidelines for the assessment of cardiovascular risk highlight shared decision-making. The estimated risk score using the pooled cohort equations are designed to start a conversation between physicians and patients about initiating statin therapy, and the CAC score can help in that regard.
To TCTMD, Lloyd-Jones said that an elevated CAC score may solidify a recommendation for statin therapy, but a CAC score of zero can provide reasonable justification for avoiding medication for the time being in patients who might have otherwise been prescribed a statin based on their 10-year risk score.
Wilkins and Lloyd-Jones add that with respect to CAC scoring, it is difficult to determine whether a given clinical strategy will change cardiovascular event rates in the population, but the latest recommendations from the USPSTF should spur researchers to conduct further studies, particularly an endpoint-driven trial. Blaha, however, argued there is no such randomized, controlled trial evidence for the Framingham risk score or ACC/AHA pooled cohort equations. Regardless, physicians still use these tools to identify a patient’s risk and then treat accordingly.
Blaha and Nasir both pointed out that the 2013 ACC/AHA guidelines for the assessment of cardiovascular risk are expected to be updated later this year and those recommendations will likely provide a more definitive take on the use of CAC screening, as well as on the use of ABI and hsCRP, for cardiovascular risk assessment.
US Preventive Services Task Force. Risk assessment for cardiovascular disease with nontraditional risk factors: US Preventive Services Task Force recommendation statement. JAMA 2018;Epub ahead of print.
Lin JS, Evans CV, Johnson E, et al. Nontraditional risk factors in cardiovascular disease risk assessment: updated evidence report and systematic review for the US Preventive Services Task Force. JAMA 2018;Epub ahead of print.
Wilkins JT, Lloyd-Jones DM. USPSTF recommendations for assessment of cardiovascular risk with nontraditional risk factors: finding the right tests for the right patients. JAMA 2018;Epub ahead of print.
Polonsky TS, Greenland P. Viewing the value of coronary artery calcium testing from different perspectives. JAMA Cardiology 2018;Epub ahead of print.
- Nasir reports serving on the advisory board of Quest Diagnostics and consulting for Regeneron.
- Blaha reports receiving grants from the NIH, FDA, AHA, and Aetna Foundation and personal fees from Amgen, Novartis, Sanofi/Regeneron, MedImmune, and Akcea.
- Ridker reports receiving research support from Novartis to conduct the CANTOS study. He also reports consulting for Novartis, Pfizer, and Sanofi. He is a co-inventor of patents held by the Brigham and Women’s Hospital and licensed to AstraZeneca and Siemens that relate to the use of inflammatory biomarkers in CVD and diabetes.
- Wilkins, Lloyd-Jones, Polonsky, and Greenland report no relevant conflicts of interest.