Repeatability limits derived for CT coronary artery calcification
 | |
| Image courtesy of Philips. |
In a study published this month in the American Journal of Roentgenology, researchers from the University of Washington in Seattle, Wake-Forest University School of Medicine in Winston-Salem, N.C., and the University of California in Los Angeles have established a 95 percent repeatability limit for coronary artery calcification (CAC) as measured by CT.
“For variability on repeated scans, the main challenges are continuous heart motion, varying plaque size, large patient size, and intraobserver and interobserver variability in calcium scoring,” the authors wrote. “The degree of agreement between measurements obtained in quick succession, called repeatability, is important not only in comparisons of scanner types and CAC scoring systems but also in the evaluation of change in CAC over time.”
The team used data from the Multi-Ethnic Study of Atherosclerosis, a prospective cohort study with 6,814 United States participants free of clinical cardiovascular disease, for its calculations. Nearly all patients (6,742) in the study had two sequential CT scans at enrollment, either with a 4-slice multidetector CT (MDCT) or an electron beam CT (EBCT).
The study employed the Imatron C150 EBCT (GE Healthcare) at three study sites, and either the 4-slice LightSpeed (GE) or 4-slice Volume Zoom (Siemens Medical Solutions) MDCT at three sites to conduct its baseline studies.
Follow-up CAC measurements were performed on one half of the cohort that was randomly selected an average of 1.6 years after the baseline exam and the other half of the cohort had a third examination an average 3.2 years after baseline.
“The Multi-Ethnic Study of Atherosclerosis affords an excellent opportunity to compare rescan variability across CT type and participant characteristics (body size, sex, race) and to explore temporal change in CAC exceeding that expected on the basis of baseline rescan variability,” the authors noted.
For the patients with a positive CAC score (3,380), the scientists examined detectable CAC change from baseline to follow-up, and a point estimate of the amount of calcium in each participant was presented as an average CAC score. A baseline repeatability limit of 95 percent was established with a quantile regression model.
A detectable change in CAC during follow-up was defined by the researchers as an increase or decrease beyond the baseline repeatability limit. They noted that they found no difference in rescan variability for the scanner types (EBCT or 4-slice MDCT) used to conduct the baseline examinations.
“Among 5,757 Multi-Ethnic Study of Atherosclerosis participants with a follow-up CAC score between the second and third examinations, 2,832 participants had a positive CAC score at baseline,” the authors wrote. “Among these 2,832 participants, the follow-up CAC score was greater than the baseline score in 2,392 (84.5 percent) of the participants.”
The scientists observed that the prognostic utility of tracking change in coronary calcified plaque remains unclear with limited evidence; however, the use of baseline rescan repeatability information may be one way to increase the sensitivity and specificity of change in calcification.
The authors cautioned that a CT CAC score is but one part of a comprehensive cardiovascular workup. As such, a positive change in a CAC score over time may serve to focus clinician attention on examining CT cardiovascular imaging studies for additional anatomic changes.
“A CAC score captures only part of the available pathobiologic information on calcified plaque available from cardiac CT images,” they wrote. “Global CAC scores of the entire coronary circulation can mask or minimize substantial specific changes in vessels and plaque.”