The Addition of Vascular Calcification Scores to Traditional Risk Factors Improves Cardiovascular Risk Assessment in Patients with Chronic Kidney Disease

Abstract

Background: Although a variety of non-invasive methods for measuring cardiovascular (CV) risk (such as carotid intima media thickness, pulse wave velocity (PWV), coronary artery and aortic calcification scores (measured either by CT scan or X-ray) and the ankle brachial index (ABI)) have been evaluated separately in chronic kidney disease (CKD) cohorts, few studies have evaluated these methods simultaneously. Here, we looked at whether the addition of non-invasive methods to traditional risk factors (TRFs) improves prediction of the CV risk in patients at different CKD stages.

Methods: We performed a prospective, observational study of the relationship between the outputs of non-invasive measurement methods on one hand and mortality and CV outcomes in 143 patients at different CKD stages on the other. During the follow-up period, 44 patients died and 30 CV events were recorded. We used Cox models to calculate the relative risk for outcomes. To assess the putative clinical value of each method, we also determined the categorical net reclassification improvement (NRI) and the integrated discrimination improvement.

Results: Vascular calcification, PWV and ABI predicted all-cause mortality and CV events in univariate analyses. However, after adjustment for TRFs, only aortic and coronary artery calcification scores were found to be significant, independent variables. Moreover, the addition of coronary artery calcification scores to TRFs improved the specificity of prediction by 20%.

Conclusion: The addition of vascular calcification scores (especially the coronary artery calcification score) to TRFs appears to improve CV risk assessment in a CKD population.

Fig 1. Linear regression curve.

Relationship between the aortic calcification scores based on a CT scan or on X-ray data (n = 143, r² = 0.761, p>0.001).

Fig 2. Unadjusted Kaplan-Meier cumulative event curves for all-cause mortality and the first non-fatal CV event, as a function of the median value of the (A) IMT, (B) PWV, (C) aortic calcification score (CT scan), (D) aortic calcification score (X-ray), (E) coronary artery calcification score (CT scan), and ABI tertile (F).

Fig 3. Harrell’s C-index for the various Cox models.

Model 1, traditional risk factors (TRFs): 0.67 ± 0.35; Model 2, TRFs + PWV: 0.67 ± 0.036; Model 3, TRFs + IMT: 0.69 ± 0.036; Model 4, TRFs + aortic calcification score (CT scan): 0.70 ± 0.036; Model 5, TRFs + aortic calcification score (X-ray): 0.71 ± 0.05; Model 6, TRFs + coronary artery calcification score: 0.70 ± 0.036; Model 7, TRFs + ABI (<1.3 vs. >1.3) 0.69 ± 0.035. All differences were non-significant.