#3539 calcium isotope ratios in serum are the strongest predictor of bone calcium balance in patients on dialysis

Abstract Background and Aims Dysregulated mineral homeostasis is common in chronic kidney disease (CKD) and associated with bone demineralization and vascular calcification. The balance between bone formation and resorption, which reflect the bone calcium (Ca) balance (BCaB), cannot be determined without bone biopsy which is invasive and not easily repeatable. Recently, we have shown that stable (i.e. non-radioactive) Ca isotopes, 42Ca and 44Ca, can be measured in serum and their ratio (δ44/42Caserum) quantitatively determines net bone gain or loss of Ca. Thus, when bone formation exceeds bone resorption, the net BCaB is positive and δ44/42Caserum is high, and when bone resorption is the predominant process δ44/42Caserum is low compared to age-matched controls. In this study we compared δ44/42Caserum against δ44/42Cabone and arterial biopsy samples (δ44/42Caartery) and the sensitivity of δ44/42Ca in predicting changes in bone histology. Method Adults receiving chronic dialysis who underwent bone and arterial biopsies at the time of kidney transplantation were recruited. Patients who had parathyroidectomy or received cinacalcet or anti-resorptive agents were excluded. All participants had Dual Energy X-ray Absorptiometry (DXA) of the hip and lumbar spine. Ca44 and Ca42 measurements were performed in serum and bone and arterial biopsy samples using a multi-collector inductively-coupled plasma mass spectrometer (Thermo Fisher Scientific, Germany). Results Nineteen patients, median age 59.8 years, 84% male, median time on dialysis 3.3 years were included. δ44/42Ca was significantly higher in serum compared to bone or arterial biopsy samples (p < 0.0001), with the lowest isotope ratios in bone (Fig. 1A). δ44/42Cabone was significantly lighter than δ44/42Caartery (p = 0.0002; Fig. 1B). δ44/42Cabone correlated positively with the osteoblastic markers BAP and P1NP (p = 0.0006, R2 = 0.51 and p = 0.009, R2 = 0.31) and inversely with PTH and the osteoclastic marker RANKL (p = 0.0017, R2 = 52 and p = 0.02, R2 = 0.29 respectively; Fig. 2). Both the DXA hip and lumbar spine T-scores and z-scores correlated positively with δ44/42Cabone. δ44/42Caserum showed an inverse correlation with the osteoid area (p = 0.04, R2 = 0.22) and a positive correlation with the absolute mineralized area and the trabecular thickness (p = 0.0004, R2 = 0.58 and p = 0.013, R2 = 0.34 respectively. The were no significant correlations with δ44/42Caartery. On multivariable linear regression analysis significant predictors of δ44/42Caserum were δ44/42Cabone (p = 0.018, 95%CI −1.35 to −0.16), age (p = 0.02, 95%CI 0.002 to 0.02) and BAP (p = 0.019, 95%CI 0.04 to 0.38), together predicting 71% of the variability in δ44/42Caserum. The only significant predictor of δ44/42Cabone was the δ44/42Caserum: p = 0.004, 95%CI −1.6 to −0.37, model R2 = 69%. Conclusion δ44/42Caserum is a significant and independent maker of BCaB, correlating with bone histology measures, and may provide a more sensitive measure than DXA or bone biomarkers. Further studies are required to determine the clinical utility of using δ44/42Caserum to guide management of mineral bone disease in CKD.