Abstract and Introduction
Abstract
Objectives This study identified the incidence and predictors of conversion of a normal to abnormal coronary artery calcium (CAC) scan during serial CAC scanning over 5 years.
Background Although a normal CAC scan signifies absence of significant atherosclerosis and is used to identify individuals at low clinical risk, the "warranty period" of a normal CAC scan relative to its ability to predict sustained absence of coronary atherosclerosis remains unknown.
Methods We assessed frequency of and time to progression, as well as proportional increase of CAC in 422 individuals with normal CAC scan (CAC = 0) undergoing annual CAC scanning for 5 years. Results were compared with those of a referent cohort of 621 individuals with baseline CAC scan (CAC >0).
Results A total of 106 (25.1%) patients with CAC = 0 developed CAC during follow-up at a mean time to conversion of 4.1 ± 0.9 years. Incidence of conversion to CAC >0 was nonlinear and was highest in the fifth year. In multivariable analysis, progression to CAC >0 was associated with age, diabetes, and smoking (p < 0.01 for all). Among the 621 individuals with baseline CAC >0, only the presence of CAC itself, rather than CAD risk factors, was predictive of CAC progression. Among propensity score-matched individuals with CAC >0 versus CAC = 0, baseline CAC >0 emerged as the strongest predictor of CAC progression (hazard ratio [HR]: 12.50, 95% confidence interval [CI]: 9.31 to 16.77), followed by diabetes (HR: 2.07, 95% CI: 1.47 to 2.90) and smoking (HR: 1.29, 95% CI: 1.02 to 1.63, p < 0.05 for all).
Conclusions Among individuals with CAC = 0, conversion to CAC >0 is nonlinear and occurs at low frequency before 4 years. No clinical factor seems to mandate earlier repeat CAC scanning.
Introduction
Coronary artery calcium (CAC) scoring has been proposed as a useful atherosclerosis imaging method for stratification and reclassification of risk of coronary heart disease (CHD). Coronary atherosclerotic lesions often contain calcified components, which can be precisely measured using either electron beam computed tomography or multidetector computed tomography by the Agatston scoring method. Increasing degrees of CAC predict adverse CHD events and all-cause mortality, whereas a normal CAC scan is a reliable predictor for low risk and is used clinically to signify the absence of any major atherosclerosis. Relatively little study has been performed, however, to examine the temporal conversion of CAC and the clinical factors that influence CAC progression. Three recent cohort studies have addressed the issue of CHD risk factors and CAC progression, and in each study, a constant conversion and progression rate for CAC has been assumed.
To address this issue prospectively, in the present study we uniquely followed up a cohort of patients who had normal CAC scans at baseline with yearly CAC scans for 5 years. Our goals were both to test the assumption of a linear conversion rate for CAC and to evaluate the factors that might predict a more accelerated rate of conversion from a normal to an abnormal CAC scan. Comparisons were made to a contrasting group of patients with an abnormal CAC scan for this purpose.