Implications of Coronary Artery Calcium Testing Among Statin Candidates According to American College of Cardiology/American Heart Association Cholesterol Management Guidelines

MESA (Multi-Ethnic Study of Atherosclerosis)

Discussion

When applying the 2013 ACC/AHA cholesterol treatment guidelines (1), we noted that nearly two-thirds of MESA participants 45 to 75 years of age without known CVD are potentially eligible for statins (recommended or considered). Approximately one-half of these statin-eligible candidates had CAC = 0 and, as a group, had a lower 10-year observed ASCVD risk than the threshold recommended for treatment. The majority of ASCVD events occurred among those with detectable CAC, consistent with 10-year risk levels suggested by ACC/AHA cholesterol management guidelines for statin therapy.

Since the release of the ACC/AHA cholesterol management guidelines in 2013, the central role of 10-year ASCVD risk estimation in guiding statin eligibility has received considerable attention (3–5). Recent studies have demonstrated that the impact of the updated recommendations will be higher treatment rates among those expected to have future cardiovascular events (19), such as those with increased atherosclerotic burden, as compared with prior guidelines (20). However, we are also challenged to treat a significant proportion of patients in low-risk groups, who may derive minimal benefit from these therapies. Accurate risk assessment that can identify these lower-risk subjects can potentially have a profound impact in facilitating appropriate resource allocation and shared decision-making to allow flexible treatment choices.

Our study results suggest that the absence of CAC can potentially overcome these challenges by providing incremental information that may move many people from risk levels that are recommended for treatment to risk levels that are not. This observation is consistent with prior reports highlighting the value of the absence of CAC (power of zero). In a pooled analysis of more than 71,000 patients with CAC = 0, only 0.5% of participants suffered an adverse event in a 4.2-year follow-up (8). In the prospective MESA and Heinz Nixdorf Recall cohorts, only 1% of participants had a hard CHD event during 5 years of follow-up (6,7). Furthermore, CAC = 0 has also been associated with a much more favorable prognosis, even among groups traditionally considered at higher risk (9–11,21). Our current study adds to the literature by demonstrating that the absence of CAC is associated with a meaningfully lower ASCVD risk among those deemed as statin candidates in a follow-up extending to 10 years. Among 41% who were recommended moderate- to high-intensity statins and who had no detectable CAC, only 4.9% experienced an ASCVD event. Furthermore, in those considered for statins (10-year risk of 5% to 7.5%), the absence of CAC was noted in a much larger proportion (57%), with an extremely low 10-year ASCVD risk of 1.5%.

These findings may have important implications. The patient-centered emphasis within the guidelines recognizes that thresholds identified for treatment do not mandate a statin prescription, but rather call for a discussion between providers and patients to foster informed decisions regarding initiation of statin therapy (1). The process by which individual patients reach an informed choice when considering a statin medication for the next 10 years, in which they balance concerns of side effects, costs, and burden of use, is also influenced by the estimate of the absolute risk reduction likely to be achieved (1,22–24). As the absolute ASCVD risk decreases, so does the net benefit of any intervention with a relative risk reduction that does not increase with lower patient risk. In these circumstances, the absence of CAC can afford significant value in promoting shared decision-making and better informing patients, who may consider avoiding statins to focus on prudent lifestyle changes, of their choices.

Our study findings also suggest that CAC testing may have limited impact on decisions regarding statin utilization in a meaningful manner at the extremes of calculated 10-year ASCVD risk. For example, among MESA participants with a 10-year ASCVD risk <5%, we noted an extremely low 10-year event rate of 1.6%. The majority (79%) of these participants have CAC = 0. The presence of mild CAC (1 to 100) did not result in reclassifying ASCVD risk above 5%, with only 4% of participants with CAC >100 having an observed ASCVD risk above the threshold for statin eligibility. Although the current guidelines suggest that CAC testing can be considered in selected patients among those with ASCVD risk <5% to inform treatment decision-making, because 25 patients will need to be scanned to influence the treatment decision, its utility in this group is likely to be limited. At the other end of the risk spectrum (estimated 10-year risk >20%), although the presence of CAC = 0 is associated with a much lower observed event rate of 11% than predicted, it may not impact the decision to avoid statins because the risk still remains above the threshold suggested by the guidelines for treatment (1). By avoiding testing at the extremes of risk (<5% and >20%), 49% of study participants with 10-year ASVCD risk estimates of 5% to 20% can be reclassified by the absence of CAC to a risk threshold below that suggested for statin therapy (Central Illustration). In addition, among participants not considered for statins by current guidelines, such as those >75 years of age and with LDL-C <70 mg/dl, CAC testing may have limited impact because it provided meaningful risk reclassification in approximately 10% of study participants to influence the decision regarding statin use.

Consideration of CAC testing for identifying appropriate candidates for statin therapy is not a straightforward decision; the pros and cons of this strategy deserve further discussion. First, with current guidelines, nearly two-thirds of U.S. adults are already candidates (recommended or considered) for statins. In this setting, screening with CAC testing to identify additional candidates for preventive treatments is questionable. We believe that the value of CAC testing in the current era may be in limiting the scope of statin therapy to more selective use, rather than in expanding it. Wijns et al. termed this risk-reduction approach “interventional prevention” (25). Second, it is important to realize that CAC testing adds to health care costs. The average national cost for CAC testing is around $100, which may be equivalent to 18 months of a generic statin priced at $5 per month. Noncardiac findings on CAC testing, such as nodules, can generate recommendations for follow-up imaging in approximately 5% of adults without a history of smoking. However, a recent study from MESA has demonstrated that even accounting for this, it may still be cost-effective to treat only those with CAC >0, compared with strategies from established risk-assessment guidelines (26). There are also concerns that CAC scoring may lead to additional cardiac testing and interventions, although a recent report from the EISNER (Early Identification of Subclinical Atherosclerosis by Noninvasive Imaging Research) study showed no such trends, with fewer downstream tests noted among those with CAC = 0 (27). In addition, the literature supports significant lifestyle optimization as well as improved preventive medication adherence among those with increasing CAC burden (28). Third, CAC testing is associated with an average radiation dose of 0.89 mSv compared with background radiation of 3 mSv per year, which should be discussed with the patient to allow informed decision-making. Finally, it is also important to note that our study does not address the issue of whether a CAC-based strategy versus guideline-based recommendations for statin selection is better or will have a favorable impact on outcomes; this vital question needs to be critically tested by well-designed comparative-effectiveness clinical studies.