Abstract 9536: Computed Tomography Derived Cardiovascular Risk Markers, Incident Cardiovascular Events and All- Cause Mortality In Adults Free of Clinical Cardiovascular Disease. The Multi Ethnic Study of Atherosclerosis

Abstract

Introduction: Cardiac CT derived CV risk markers such as coronary artery calcium (CAC), thoracic wall calcium (TAC), aortic valve calcium (AVC), mitral valve calcium (MVC), pericardial adipose tissue volume (PAT) and liver attenuation (LA) have all been associated with CV risk and events. However, there is no head to head comparison of these markers in a single sample using the same statistical approach, and it remains unclear whether a combination of these markers can further improve CV risk prediction. We assessed the predictive value of CAC, TAC, AVC, MVC, PAT, LA and a combination of these markers for incident CVD/CHD/mortality in MESA.

Methods: After 9 years of follow up, 327 had adjudicated CHD defined as MI, angina if followed by PTCI/CABG, cardiac arrest or CHD death, 457 had CVD events defined as CHD, stroke/TIA or CVD death and 398 died. Cox proportional hazard, area under curve(auc) and net reclassification improvement (NRI) analyses were used adjusting for age, gender, race, SBP, total and HDL cholesterol, triglycerides, BMI, smoking, BP med and statin use.

Results: 6591 MESA participants mean age and BMI, 62 years and 28.4 kg/m², 53% females, 38% Caucasians, 28% African Americans, 22% Hispanics and 11% Chinese. Unlike PAT and LA; CAC, TAC, AVC and MVC were all significantly associated with incident CVD/CHD/ mortality in our multivariable Cox model with CAC having the strongest association. CAC had the highest AUC when used alone to predict incident CHD/CVD whilst LA had the least. The addition of CAC to Framingham Risk Score (FRS) resulted in the most improvement in AUC for incident CVD (0.721 vs. 0.766, p<0.0001) and CHD (0.712 vs. 0.766, p<0.0001). The addition of CAC to the FRS resulted in an NRI of 0.191 for incident CVD and 0.229 for incident CHD. The addition of TAC, AVC, MVC, PAT and LA individually to FRS + CAC resulted in little further improvement in AUC/NRI for the prediction of CHD/CVD. Similar results were obtained when the analysis was limited to participants with intermediate FRS. Any other combination of risk markers yielded inferior results.

Conclusion: CAC is superior for improving CV risk prediction over and beyond the FRS. The addition of TAC, AVC, MVC, PAT and LA only minimally improves the predictive accuracy afforded by the combination of FRS and CAC.