Coronary artery calcium scoring on 320 detector row CT demonstrates low interscan variability at standard and 70% reduced radiation dose

Introduction: Coronary artery calcium (CAC) is a well-established predictor of coronary heart disease events and is an important factor for individualized cardiac risk assessment. In particular, interscan variability in CAC imaging may affect proper clinical risk stratification and requires assessment. This study evaluates reproducibility of CAC performed with a 320-detector row CT at both standard clinical and low radiation dose using conventional filtered back projection (FBP) and iterative reconstruction (IR). Methods:200 consecutive pts (59 % male, 60 ± 9 yrs, 28 ± 5 BMI) prospectively underwent 2 low and 2 standard dose scans (800 total scans, 1600 total reconstructions) in randomized order under an Institutional Review Board approved study. For each imaging method (standard or low radiation, FBP or IR), rescan reproducibility was assessed through the coefficient of reproducibility, the kappa statistic for standard Agatston categories defined by Agatston scores 0, 1-10, 11-100, 101-400, and >400 and Bland-Altman methods. Results: The coefficient of reproducibility was similar among all of the reconstructions: standard IR was 0.177 (95% CI: 0.161 - 0.194), standard FBP 0.192 (95% CI: 0.177-0.207), low FBP was 0.177 (95% CI: 0.153 - 0.200) and low IR was 0.169 (0.150 - 0.188) respectively. When stratified by standard Agatston categories, agreement was uniformly excellent. For standard FBP the agreement was 93% (K=0.91, 95% CI 0.86 - 0.95), for standard IR the agreement was 92% (K=0.89, 95% CI: 0.84 - 0.94), for low FBP the agreement was 91% (K = 0.88, 95% CI: 0.82 - 0.93) and for low IR the agreement was 91% (K = 0.88, 95% CI (0.83 - 0.93). By the Bland-Altman method, re-scan variability was small at low values and increased as the mean scores increased to a similar degree (Low IR and Standard FBP shown in Figure). Overall, radiation exposure was 70% lower (0.45 mSv ± 0.3 mSv vs. 1.5 mSv ± 0.9 mSv) for low versus standard radiation dose scans. Conclusions: In a 320-detector row CT scanner, inter-scan variability for all imaging methods (standard or low radiation, FBP or IR) is low. This study supports 70% reduced radiation dose for CAC imaging. (Figure Presented).