Abstract 12: Volume of Calcium in the Descending Thoracic Aorta Predicts All Cause Mortality Beyond Coronary Artery Calcium: The Multi-Ethnic Study of Atherosclerosis
Introduction: Coronary artery calcium (CAC) volume and density differentially predict incident cardiovascular disease (CVD), with CAC density inversely associated with these outcomes. Whether similar associations exist between descending thoracic aortic calcium (DTAC) volume and density and all cause mortality (ACM) are unknown. We hypothesized that DTAC volume and density predict ACM independently of CAC. Methods: The Multi-Ethnic Study of Atherosclerosis enrolled 6,814 participants free of clinical CVD at baseline and followed them for incident adverse events. Cardiac CT at baseline visualized the segment of the descending thoracic aorta posterior to the heart. Only participants with prevalent DTAC were included (necessary to evaluate DTAC density). DTAC and CAC volumes were natural log transformed to adjust for skewness. Cox regression models estimated the associations of DTAC volume and density with ACM after adjustment for age, gender, ethnicity, CVD risk factors, statin use, and CAC volume and density. The incremental predictive values of DTAC volume and density were evaluated by area under receiver operating characteristic (AUC) curves. Results: Of the total cohort, 1,850 participants (27%) had prevalent DTAC and 491 deaths occurred over 10.3 years. In separate regression models, DTAC volume was independently associated with ACM after adjustment for CAC volume (HR 1.21 [95% CI 1.09-1.35]) and additional adjustment for CAC density (1.18 ([1.06-1.32]). After the same adjustments, DTAC density was not significantly associated with ACM (0.94 [0.84-1.06]). The AUC for the base Model 1 (risk factors + CAC volume) was 0.706 (0.680-0.732), which increased to 0.716 (0.690-0.742) with the addition of DTAC volume in Model 2 (p=0.03 compared to Model 1). Further addition of DTAC density in Model 3 did not improve the AUC significantly (0.717 [0.692-0.743], p=0.23 compared to Model 2). Conclusions: In a cohort free of baseline clinical CVD, DTAC visualized on cardiac CT was common. When DTAC was present, DTAC volume (but not density) was independently associated with ACM. DTAC volume also significantly improved ACM risk prediction beyond risk factors and CAC volume.