Abstract 19093: Assessing the Predictive Value of Traditional Risk Factors for Coronary Artery Disease Severity and Plaque Morphology on Coronary Computed Tomography Angiography: Individual vs Population
Objectives: To assess the predictive value of traditional cardiac risk factors (CRFs) for coronary artery disease (CAD) burden and plaque morphology (PM) as reported by Coronary CT Angiography (CCTA).
Methods: Consecutive CCTAs performed at a quaternary center in Vancouver, Canada between January 2008 and March 2010 were reviewed. Patients with known CAD or un-interpretable scans were excluded. Patients' demographics and self reported CRFs were obtained. Experienced physicians interpreted CCTAs using an 18-segment model. The degree of stenosis in each segment was quantified. Segment scores were summed to generate a total stenosis score (TSS). Each segments' PM was reported as either calcified (CP), non calcified (NCP) or mixed plaque (MP) and patients' dominant PM was identified. Independent sample T tests and linear regression were utilized to compare the mean TSS and prevalence of various PM dominance by risk factor profile at the population level and to assess the correlation of each CRF with TSS and type of PM dominance at the individual level, respectively.
Results: The cohort consisted of 421 patients with 257 (61%) male. Mean age was 56.2 (+/− 13.8) years. Coronary arteries were normal in 32% of patients. Population analysis revealed older age, male gender, hypertension, dyslipidemia, diabetes mellitus (DM), current smoking and sedentary lifestyle were associated with higher mean TSS (p<0.01). All of the abovementioned CRFs with the exception of DM were associated with higher prevalence of CP or MP dominance (P<0.03). DM was associated with higher prevalence of NCP dominance (p<0.01). At individual level, CRFs were only weakly correlated with TSS (Pearson Coefficient (PC) range: 0.073–0.211, mean: 0.142) and type of PM (PC range = −0.04 − 0.182).
Conclusions: Traditional CAD risk factors have a strong association with the CAD burden and type of PM dominance at the population level. However, their correlations were only weak at individual level. These results highlight the limitations of the application of population level analysis to individual patients. With the rapid evolution of low radiation dose non-invasive imaging modalities such as CCTA, refinement of CAD risk and assessment of CAD burden and PM for the individual patient warrants further study.
- © 2010 by American Heart Association, Inc.