NEW YORK (Reuters Health) – A normal coronary artery calcium (CAC) scan in middle-age appears to have a “warranty period” of about four years, during which the patient is unlikely to develop coronary calcium, U.S. investigators report.
Therefore, they say, “Repeat CAC scoring should not be performed for a minimum of 4 years in individuals with a baseline score of CAC = 0.”
As first author Dr. James K. Min, from Weill Medical College of Cornell University, New York, and his colleagues point out in the March 16th issue of the Journal of the American College of Cardiology, little is known about the timing of CAC conversion and factors that affect progression.
Their prospective study involved 422 patients, free of known coronary artery disease, who had no CAC on an initial screening scan. They had annual scanning (with electron beam computed tomography) for 5 years or until a scan showed calcium. The researchers described the extent of CAC in terms of the Agatston Score, which is based on the area and density of the calcified plaques.
Overall, 106 (25.1%) subjects converted to a positive CAC, but the conversion rate was not linear. Conversion at 1, 2, 3, 4, and 5 years occurred in 0.5%, 1.2%, 5.7%, 6.2%, and 11.6%, respectively. The average time to conversion was 4.1 years, with no link between time to conversion and any cardiovascular risk factor, or patient age. “Thus,” they write, “it is unlikely that clinical factors” will identify patients with an initially normal scan for whom early repeat CAC scanning may be warranted.”
Dr. Min’s team also studied patterns of progression in 621 individuals with known coronary artery calcium who had repeat CAC scans. During a mean follow-up of nearly 2 years, CAC progressed in 497 (80%) patients, by an average of 45.5 points. (A 2003 report in Circulation by Hoffman et al from the Massachusetts General Hospital notes that absolute Agatston scores of less than 10, 11 to 99, 100 to 400, and above 400 have been proposed to categorize coronary calcium as minimal, moderate, increased, or extensive, respectively.)
In univariate analysis, the risk of progression increased with increasing baseline CAC. Hazard ratios for progression ranged from 1.22 for CAC > 100, to 1.81 for CAC > 600 and 1.64 for CAC > 1000).
In multivariate analysis, only baseline CAC score was independently associated with risk of progression, leading the authors to conclude that “a one-time CAC score may be misleading” if it was positive.
Finally, the research team compared conversions to higher CAC scores in propensity-matched individuals with and without initially normal CAC scans. The 266 patients in both groups were similar in age, sex, hyperlipidemia, hypertension, smoking and diabetes.
In multivariate analysis, a CAC score greater than zero emerged as the strongest predictor of progression (HR 14.96), followed by diabetes (HR 2.07) and smoking (HR 1.29).
“The presence and absolute amount of CAC seems to trump other risk factors for CAC progression,” the authors say.
So although patients with an initial negative scan don’t need another scan soon afterward, repeat scanning may be in order for those with a positive scan, for “more accurate updating of CHD risk,” the authors suggest.
Limitations of the study include lack of information about treatment of cardiovascular risk factors, which could cause treatment bias, and about potential effects of subclinical chronic kidney disease.
Also, the researchers acknowledge, they had no “true” baseline measurements, but they still believe clinicians “can use these data as a benchmark for middle-age, low- to intermediate-risk individuals.”
In an editorial, Dr. Harvey S. Hecht, from Lenox Hill Heart and Vascular Institute, New York, points out that with a zero CAC score, asymptomatic patients can be assigned to the lowest risk category for heart disease and be spared “a potentially harmful and costly drug” for treatment of elevated lipid levels.
Consequently, he maintains, widespread CAC scanning “will offer an opportunity for implementing a safe, cost-saving paradigm.”
J Am Coll Cardiol 2010;55:1110-1120.