The recent American College of Cardiology/American Heart Association (ACC/AHA) guidelines on managing blood cholesterol for primary prevention emphasizes the need for individual cardiovascular (CV) risk stratification (1). According to society guidelines, the categorization of CV risk is mainly based on the pooled cohort equations risk calculator (PCE) (1, 2), which significantly broadens the indication for statin therapy (3, 4). However, long-term statin treatment can expose subjects to unwanted side effects (5, 6) and hence, should be avoided in low-risk subjects. Guidelines suggest using risk modifiers to improve statin therapy allocation for primary prevention (1). Coronary artery calcium (CAC) is among the most powerful prognostic tools for CV events and mortality (7–21). It has been shown that the presence and severity of CAC helps to identify patients who are most likely to benefit from statins for primary prevention (22). Furthermore, subjects with zero CAC have excellent CV prognosis (23–29), and most of them are recategorized into the lower CV risk category (<5%). Consequently, CAC was accepted by both the ACC/AHA (1) and the European Society of Cardiology (6) as a risk modifier for statin treatment.

The multi-ethnic study of atherosclerosis-calcium (MESA-C) score is the first to include CAC, in addition to traditional risk factors, to estimate the 10-year cardiac risk (30). CAC has also been shown to be an independent predictor of cerebrovascular accidents (CVA) in asymptomatic adults (31). The MESA-C score has been validated in the Heinz Nixdorf Recall (HNR) study (32) and the Dallas Heart Study (DHS)(33). Recently, a study on a large retrospective cohort of asymptomatic patients has shown that the MESA-C score and PCE plus CAC showed the best CV risk discrimination among patients with an estimated CV risk of 5–20% and that the addition of CAC also modestly improved risk discrimination among those estimated as low or high CV risk (34).

We aimed to identify the proportion of asymptomatic patients for which using the MESA-C score will result in reallocation to statin treatment by recategorization to a higher or lower risk category.

A total of 1,850 subjects underwent annual screening as noted above. Of them, 745 asymptomatic consecutive men age > 40 and women age > 50 years who were CAD and/or stroke free consented to perform a screening cardiac CT for CAC. A total of 113 subjects were subsequently excluded from the final cohort due to missing or incomplete data. Thus, our final cohort consisted of 632 participants (84% male) with a mean age of 56 ± 7 years (Table 1). Subjects with low risk by PCE (<7.5%) and high risk by MESA-C (≥7.5%) had a higher CAC and a higher prevalence of CAC above zero than subjects with high PCE scores (≥7.5%) and low MESA-C scores (<7.5%) (Table 1). All other basic characteristics are outlined in Table 1.

In our cohort of asymptomatic adults, we demonstrated that the MESA-C risk score can improve allocation to statin treatment. The ability of the MESA-C score to improve our discriminative capacity is mainly due to the inclusion of the CAC score into the MESA algorithm. Interestingly, the CAC appears to be of importance when the MESA-C scores and PCE scores are non-concordant. This is of particular importance in light of the various professional society guidelines recommendations on cholesterol treatment which significantly increases the eligibility for statin therapy (3, 4, 6). Our data show that when the MESA-C score is applied, almost half of study subjects are recategorized to a different CV risk class, and the majority of the reallocations are to a lower CV risk category than previously defined by the PCE. Consequently, statin eligibility also changed. Approximately 85% of subjects with borderline PCE risk (i.e., 5–7.5%), who are questionably eligible for statins, were recategorized according to their MESA-C score. The majority of these participants were down-categorized with statin treatment no longer be indicated. In addition, almost two-thirds of our study population had a PCE risk score > 5% (i.e., candidate for statin therapy). However, almost half of them had a MESA-C score of < 5%, and thus statins were eventually not indicated. Our findings concur with previously published data from a larger study by Nasir et al. (37).

In addition, we also present data regarding survival calculation by these scores. Our Kaplan–Meier’s curves for survival according to the MESA-C and PCE risk scores show that compared to both scores being low, a high PCE score predicts low survival, and compared to both scores being high, a low PCE score predicts better survival. These data suggest that both scores are essential for more complete survival analysis.

CAC is a well-established surrogate marker for the total burden of coronary artery atherosclerosis. Its major advantage lies in that it measures sub-clinical CHD that reflects the entire genetic, metabolic, and acquired risk factors that determine the coronary atherosclerotic burden and progression. This explains the incremental prognostic value of CAC for the prediction of CV events and mortality beyond traditional risk factors (7, 8, 10–14, 22, 32, 37–41). The MESA-C study was a population-based study with a multiethnic composition and availability of a 10-year follow-up for CHD event incidents (38). The MESA-C risk score is the only risk score that optimizes 10-year CHD risk prediction by including CAC in addition to traditional risk factors (36), allowing it to be a potent adjunct tool in risk-based treatment decisions in the clinical practice (42, 43). Polonsky et al. (11) demonstrated that adding CAC to the MESA risk prediction model significantly alters the risk classification that was previously based solely on traditional risk factors. The main clinical contribution of CAC is its impact on statin therapy for primary prevention. Mitchell et al. (24) followed patients without baseline CVD for a median of 9.4 years and found that statin therapy was associated with a reduced risk of CV events in patients with measurable CAC, but not in patients with zero CAC and that the effect of statin use on CV events was significantly related to the severity of CAC. Most recently, a large retrospective cohort study of 53,487 patients with a mean follow-up of 12 years supports the use of CAC for further risk assessment in the borderline and intermediate risk groups. Furthermore, this study has shown that even in select subjects at low or high estimated risk, CAC modestly improves risk discrimination (34). Our findings agree with the aforementioned studies. However, we provide additional data regarding the discriminative capacity of the MESA-C score in a different patient population, which further contributes to the clinical validation of the MESA-C score. Moreover, our data emphasize the significance of risk recategorization and reallocation to an appropriate CV risk group, thereby obtaining a more informed decision regarding statin treatment.

The absence of CAC indicates an excellent CV prognosis and has a high negative predictive value for CV event occurrences (23, 25–29, 44). We found that almost one third of those who were classified by the PCE to the higher risk category of ≥ 7.5% had zero CAC, and consequently were recategorized by the MESA-C risk calculator to the lowest CV risk category. Nasir et al. evaluated the implications of zero CAC on risk score reclassification (37) and demonstrated that the absence of CAC reclassifies 44% of those who were defined as eligible for statins by the PCE, as not eligible. Similarly, we found that almost half of those who were classified by the PCE as eligible for statins had zero CAC and were therefore not eligible for statins by the MESA-C categorization. Furthermore, we found that zero CAC resulted in a MESA-C risk of < 5% in 99% of the cases and was associated with a very low event rate. We showed that when CAC is included in the MESA score, the discrimination capacity of MESA-C, as reflected by the ROC analysis, increases from a C-statistics of 0.653 to 0.770. Our findings concur with those suggested by Budoff et al. (10), who found that the 10-year event rates in those with zero CAC were almost exclusively below 5%, while those with CAC ≥ that the 10-year event rates in those with zero CAC were almost exs on cholesterol treatment (1) suggest statin treatment for all adults 40–75 years with CAC ≥ 100.

Our study has several limitations. First, this is a retrospective analysis with all of its inherent biases. Second, the information regarding statin treatment was available only at baseline but the use or initiation of statin during follow-up is lacking. Nevertheless, cholesterol levels at baseline were not statistically different between the study groups and they are also included in the MESA-C score algorithm. Third, our study is limited to subjects within the age range of the MESA-C study (45–80 years). Therefore, our findings cannot be applied to younger than 45 years. However, it has been previously shown that CAC is a risk modifier in the younger population as well (45), particularly, in those with a family history of CHD (46). Fourth, the MESA-C risk algorithm was designed to estimate only the cardiac risk, while the PCE also includes a composite vascular risk assessment. Notably, it has been previously demonstrated that CAC is an independent predictor of CVA events in the asymptomatic MESA cohort and improves the discrimination afforded by current stroke risk factors (31). Accordingly, we were able to demonstrate the relevance of the MESA-C risk categorization for overall CV risk, including CVA events. Finally, we excluded one hundred and thirteen subjects from our initial cohort due to incomplete data, thus our final cohort may also be at risk for selection bias.

In conclusion, our study provides further clinical validation for CAC and emphasizes the advantage of the MESA-C risk score as a discriminative tool for CV risk stratification and thus eligibility for statin use. Based on our data and current literature, the MESA-C score may serve as an important adjunct tool in our armamentarium for patient-physician decision-making and primary CV prevention.

The original contributions presented in the study are included in the article/supplementary material, further inquiries can be directed to the corresponding author.

The studies involving human participants were reviewed and approved by the Chaim Sheba Medical Center Helsinki Committee. Written informed consent for participation was not required for this study in accordance with the national legislation and the institutional requirements.

GS, JS, SS, and EG performed the material preparation, data collection, and analysis. GS and JS wrote the first draft of the manuscript. NK-M performed the statistical analysis. All authors commented on previous versions of the manuscript, contributed to the study conception and design, and read, and approved the final manuscript.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.