Author + information
- James K. Min, MD⁎ (, )
- Allison Dunning, MS,
- Leslee J. Shaw, PhD,
- Daniel S. Berman, MD and
- Tracy Q. Callister, MD
- ↵⁎Departments of Medicine and Radiology, Weill Medical College of Cornell University, 520 East 70th Street, K415, New York, New York 10021
We thank Dr. McEvoy and colleagues for their interest in our study (1). They raise several pertinent points that deserve response. As noted, the mean 10-year coronary heart disease (CHD) Framingham risk score for our study cohort with coronary artery calcium (CAC) scores of zero was 9%, which comprised 62% low-risk (<10% 10-year risk), 34% intermediate-risk (10% to 20% 10-year risk), and 5% high-risk (>20% 10-year risk) patients. Although CAC scoring is not presently endorsed for low-risk patients, we observed relationships of CAC conversion and time to conversion in low-risk versus intermediate-risk patients that merit consideration.
Time to conversion to CAC >0 in low-risk patients did not differ compared with intermediate-risk patients (4.01 vs. 4.17 years, p = 0.99), and low-risk patients even trended toward higher rates of conversion from CAC = 0 to CAC >0 (11% vs. 8%, p = 0.08). These findings are in keeping with recent population-based studies of CAC, including the Multiethnic Study of Atherosclerosis, which demonstrated a prognostic value of CAC beyond the Framingham risk score across age, sex, and ethnic category, and the Dallas Heart Study, in which CAC scores >100 were observed in a substantial proportion of low-risk patients who would not otherwise have been candidates for statin-based therapy (2,3). Given the uneasiness on the part of many clinicians about relying solely on the Framingham risk score (particularly in younger women and men and those with strong family histories of premature CHD), the performance of CAC scoring in low-risk patients may represent not only a clinical reality but also an opportunity for improved stratification and reclassification of CHD risk.
Dr. McEvoy and colleagues also suggest that progression of CAC >0 to a higher CAC score may be due to the intermediate follow-up period of 1.9 years or interscan variability. Using the “clinically meaningful cutoff” of a >15% increase between CAC scans they recommend, the relationship of increase in CAC remains dependent solely on the baseline CAC score. In stepwise multivariate analyses, CAC >400 and CAC >600 were the only predictive factors for CAC rise in both low-risk (hazard ratio: 2.08; 95% confidence interval: 1.29 to 3.35; p = 0.003) and intermediate-risk (hazard ratio: 1.82; 95% confidence interval: 1.23 to 2.58; p = 0.001) patients.
We agree with Dr. McEvoy and colleagues that many questions related to the potential effects of lipid-lowering therapies on CAC remain unanswered, and we agree that CAC = 0 holds great potential for optimizing the classification of asymptomatic patients. We hope that our findings that demonstrate a 4-year “warranty period” of CAC = 0 across all strata of CHD risk will add to the foundation of scientific evidence on which future studies can be based, and in the interim, may inform physicians who use CAC scanning that retesting patients at intervals of <4 years is unlikely to be helpful.
- American College of Cardiology Foundation