Introduction
Cardiac imaging by computed tomography (CT) is a remarkable technological accomplishment and, yet CT imaging for coronary heart disease (CHD) risk assessment in asymptomatic subjects still has not achieved full acceptance. Hundreds of research reports support the prognostic usefulness of coronary artery calcium (CAC) testing1 for risk assessment. Practice guidelines from the American Heart Association/American College of Cardiology (AHA/ACC),2 from Canada,3 Europe,4 Australia, and New Zealand5 support the use of CAC in selected patients, but these recommendations lack Class Ia Level of Evidence. In the absence of convincing trial data, insurance coverage for CAC tests in asymptomatic people is variable, despite strong guideline endorsement. Clinical trials are urgently needed to determine whether CAC scores should be routinely used for risk assessment based on evidence of improved patient outcomes.6
Several ongoing trials may finally answer whether CAC scanning improves clinical outcomes. The Risk Or Benefit IN Screening for CArdiovascular disease (ROBINSCA) trial began in the Netherlands in 2014 and is due to conclude soon.7,8 The ROBINSCA trial enrolled 43,447 people, and the primary endpoints were coronary disease morbidity and mortality after 5 years. These results are eagerly awaited, and, because of the size and breadth of the study, could result in practice changes if results are convincing.
The Coronary Artery calcium score: Use to Guide management of Hereditary Coronary Artery Disease (CAUGHT-CAD) trial9,10 was a randomized trial that compared management using CAC scoring versus usual care in approximately 700 intermediate-risk participants with a family history of CHD who did not satisfy current Australian guidelines for primary prevention. The primary endpoint was change of coronary plaque volume at 3 years, and the secondary endpoint was clinical events after 9 years. Because of the size and scope of the study, it is unlikely that any results from this study would convince naysayers of the value of CAC testing for risk assessment.
The SCOT-HEART 2 (Computed Tomography Coronary Angiography [CTCA] for the Prevention of Myocardial Infarction) trial11 is a randomized controlled trial of at least 6,000 subjects at risk for atherosclerotic cardiovascular disease (ASCVD) and will test whether CTCA-guided management will be associated with better targeted intervention, prevent overmedication, and result in fewer cardiac events than traditional risk assessment. This trial is large enough to yield valuable clinical endpoint data, and is expected to report clinical outcomes in 2027, but the use of CTCA instead of simpler and cheaper CAC imaging may deter patients from participating and physicians from using the results, even if results are significant.
A trial was recently funded by the National Heart, Lung, and Blood Institute called the CAC PREVENTABLE (Coronary Artery Calcium in the Pragmatic Evaluation of Events and Benefits of Lipid Lowering in the Elderly) ancillary study.12 This is an ancillary study of a large pragmatic clinical trial that is currently randomizing 20,000 adults age 75 years or older to atorvastatin 40 mg or placebo and following them for ASCVD events. At trial conclusion, outcome analyses will be stratified by CAC to determine heterogeneity of statin effects by CAC status. These results will only be applicable to people age 75 years and older, whereas the greatest need for statin use is undoubtedly in younger people.
In addition, the CorCal (Effectiveness of a Proactive Cardiovascular Primary Prevention Strategy, With or Without the Use of Coronary Calcium Screening, in Preventing Future Major Adverse Cardiac Events) trial is ongoing and plans to end in 2024. CorCal is a pragmatic trial assessing the effectiveness of a proactive cardiovascular primary prevention strategy, with or without CAC screening, conducted among 9,000 patients within the Intermountain Healthcare System in the United States. This open-label study is randomizing eligible participants to receive a statin recommendation based on either CAC screening results or on standard risk assessment results using the AHA/ACC recommended pooled cohort equation (PCE). The final decision regarding treatment is in the hands of the patient and their personal clinician.
In this issue of iJACC, Muhlestein et al13 report the results of a feasibility study conducted to assist in the overall design of the CorCal study. Patients (n = 601) without known ASCVD, diabetes, or previous statin use, were enrolled from primary care clinics and randomized to a CAC-guided clinical recommendation for statin prescription or to a recommendation based on the PCE risk estimate. Patients and physicians made the final decision about treatment in both groups. A statin was recommended in 35.9% of the CAC arm and in 47.9% of the PCE arm, indicating more restricted use because of reclassification to lower risk with CAC testing. There was also evidence of greater statin adherence at 12 months in the CAC-guided patient group than with usual care (63.3% vs 45.6%).
However, the feasibility study also pointed to some potential pitfalls in the larger CorCal study. Approximately 30% of the participants in the CAC arm who qualified for a statin based on their score never started the medication; 54.5% stated that it was because their physicians advised against a statin or never prescribed it. Leaving treatment decisions to patients and providers reduces the cost and complexity of the study. However, when the study is finished, we may find no difference in event rates between the 2 treatment arms, perhaps because too many patients and providers did not follow the advice.
Furthermore, a large proportion of patients will have zero scores, allowing them to forego statin therapy. The ability of CAC to sort patients into higher and lower risk groups is one of its most appealing characteristics as a screening tool. However, the incidence of ASCVD is directly associated with low-density lipoprotein cholesterol (LDL-C) levels. This represents an inherent challenge when designing trials of CAC testing. Although adherence to statins was greater in the CAC arm of Muhlestein et al13 study, the absolute difference in LDL-C levels between groups was only 13.6 mg/dL, in part because adherence rates were still relatively low, and the boost in adherence with CAC was offset by fewer patients needing statins. We struggled with these design challenges when we and others proposed a large clinical trial of CAC testing several years ago.14 The sample size ranged from 20,000 to 30,000 participants depending on our assumptions about several variables, including statin adherence and baseline risk of the population. We also chose to have the study control statin prescriptions and take a more active role to enhance statin adherence. Such factors will be important to consider when interpreting the results of the full CorCal study.
There are many lessons in this 40-year story of technical advances,15 large population studies, improved image quality, reduced radiation exposure, and lower test costs. All of those accomplishments are incredibly important and have influenced clinical practice across the globe to some degree. However, at this time, trial evidence is long overdue, and we are eager to learn the results of all of the ongoing trials. We congratulate the team led by Muhlestein et al13 for their tenacity and courage to answer a question 40 years in the making.
ARYAN'S BLOG 