Abstract 13595: Association of Costal Cartilage Calcification With Cardiovascular Events and Risk Factors in Multi-Ethnic Study of Atherosclerosis (MESA); The MESArthritis Study

Abstract

Introduction: Costal cartilage calcification (CCC) is an evident yet usually neglected finding in calcium scoring CT scans. Although CAC score is a well-known predictive marker for CVD, little is known about the potential association between CCC and CVD risk factors (RFs) and events. Hypothesis: CCC is associated with CVD RFs and is predictive of CVD events independent from CAC score. Methods: Cardiac CAC score images from Johns Hopkins Hospital field center acquired within the fifth exam of the MESA cohort (2010-2012) were re-analyzed as part of the MESArthritis ancillary study. The CCC was measured bilaterally for the first pair of cartilages at the superior end of image FOVs using calcium scoring package (VScore, Vitrea 7.11, Vital Images) with a 180 HU cut-off. Outlier values were excluded using the IQR-based outlier fence estimates. Association of CCC with CVD RFs, and CVD and mortality time-to-event was investigated by multivariable linear regression and Cox proportional hazard models, respectively. Results: After exclusion of 3 outliers, 329 participants (53% female) with a mean age of 70.1 (±8.75) were included. In cross-sectional analysis, in addition to age (β=8.86, p:.003) and gender (376.0, p<.001), CCC was associated with diabetes (141.42, p:.019) and glucose level (2.85, p:.006). CCC was correlated with Framingham global CVD risk score (FRS) (coefficient:0.39, p<.001), but not with CAC when adjusted for age and gender. In time-to-event analysis, adjusted CCC was positively associated with CHD (HR:1.17, p:.039) and CVD (1.14, p:.012) risk. Compared with CAC score (C-index: 0.76) and FRS (0.74) individually, models with CCC addition to FRS (0.85) and both CAC and FRS (0.84) had higher C-index for CHD prediction. Conclusions: CCC is associated with diabetes and can be used to predict CHD and CVD events and augment the predictive power of the highly validated CAC and FRS for CHD events. However, this scoring should be validated in other larger cohorts.