Multimodality vascular imaging in CKD: divergence of risk between measured parameters

High cardiovascular risk in chronic kidney disease (CKD) patients appears only partly attributable to atherosclerosis, with much of the remaining risk being ascribed to other vasculature abnormalities, including endothelial dysfunction, arterial stiffness and vascular calcification (VC). To date, these factors have been primarily studied in isolation or in dialysis patients. This study performed a global vascular assessment in moderate CKD and assessed the relationships with both traditional and novel risk factors.

This was a prospective cross-sectional analysis of 120 patients (age 60 ± 10 years; estimated glomerular filtration rate 25-60 mL/min/1.73m(2)). Demographic, clinical and biochemical characterization was performed. VC was characterized by lateral lumbar radiograph; arterial stiffness by aortic pulse-wave velocity (PWV); atheroma burden by carotid intima-media thickness (cIMT) and endothelial function by flow-mediated dilation (FMD) of the brachial artery.

VC was highly prevalent (74%), and FMD generally poor (FMDΔ 3.3 ± 3.3%). There were significant correlations between all vascular parameters; although these were predominantly explained by age. cIMT was independently associated with classical risks and also PWV (adjusted standardized β = 0.31, P = 0.001). However, traditional risks showed almost no independent associations with other vascular measurements. In contrast, serum phosphate and 1,25-dihydroxyvitamin D (1,25-OHD) correlated with PWV and the presence of VC, respectively. After adjustment, every 1 pg/mL increase in 1,25-OHD was related to a 3% reduction in the chance of VC (odds ratio 0.97; 95% confidence interval 0.94-1.00, P = 0.03). Medication use, HOMA-IR and C-reactive protein did not correlate with any of the vascular measures.

This study demonstrates extensive vascular disease across multimodality imaging in moderate CKD. Atherosclerotic burden correlated with traditional risks and PWV, while higher 1,25-OHD was associated with less VC. The lack of association between renal function and imaging indices raises the possibility of a threshold, rather than graded uraemic effect on vascular health that warrants further exploration.