MO810: Hypocalcaemia Associated with Denosumab Antiresorptive Therapy is Independent of Bone Turnover Biomarkers

Abstract

Abstract BACKGROUND AND AIMS Background: The incidence of major fractures is increased in the setting of haemodialysis patients. This is related to the decrease in bone density and bone quality and the high prevalence of low bone turnover (LBT). The alteration of calcium-phosphorus metabolism (CKD-MBD), results in a high incidence of cardiovascular disease in patients with multiple comorbidities (inflammation, diabetes). The use of antiresorptive therapy is limited by side effects (hypocalcaemia and LBT). However, it is unclear whether hypocalcaemia is determined by therapy regardless of the prevalence of LBT and comorbidities. Aim: To evaluate the incidence of hypocalcaemia (albumin and mortality adjusted) in a setting of haemodialysis patients yes-therapy (denosumab) versus no-therapy. To evaluate the association of therapy with hypocalcaemia regardless of biomarkers of bone turnover. METHOD From a group of DXA-assessed haemodialysis patients, those who meet the standard eligibility criteria for osteoporosis are undergone on antiresorptive drug denosumab 60 mg every 6 months in subcutaneous administration (yes-therapy group). The remaining DXA-assessed constitutes the control group (no-therapy group), ratio 1: 2. Patients are followed prospectively from the date of execution of the DXA until 30 September 2021. Patients are censored for death or withdrawal of consent. At the DXA date, baseline variables of demographic, biochemical type, related to calcium phosphorus metabolism and DXA, comorbidities are recorded. Prospectively, the levels of calcium, phosphorus, PTH, phosphatase alkaline albumin Hb are recorded every 3 months. Basal therapy of Vit D, P-binders and Ca mimetics was recorded at baseline and could be modified by the doctor on duty for the optimization of parameters Statistical analysis: Data are expressed as mean and SD or as median and range, where required. The outcome is the incidence of hypocalcaemia (<2 quartile = <8.5 mg/dL). Calcium is corrected for albumin and mortality. Calcium adjustment for mortality is achieved by the survival Cox model and a mixed model (calcium as a result variable and random intercept, random slope and fixed effect for time). The slope of the calcemia is informed by the expected longitudinal value at the time of death (informative censoring). Bivariate cox models explore the association between baseline variables (propensity score), comorbidities and biomarkers of bone turnover (mixed model derived) and outcome (joint models). The parameterization of PTH and AF (mixed model derived) is AUC of PTH and of alkaline phosphatase, z-score of PTH and AF beta coefficients. P = .005 is significant. RESULTS Out of 31 DXA-assessed haemodialysis patients, 11 (35%) met the eligibility criteria for denosumab antirabsorptive therapy. In the yes-therapy group, 63% of patients received three subcutaneous doses during the observation period [mean follow-up 226 (179) days]. All-cause mortality did not differ between the yes-therapy versus no-therapy groups (9% versus 25%; P = 0.28). The two groups are homogeneous for all variables except the derived DXA ones. The incidence of hypocalcaemia (<2nd quartile = 8.5 mg/dL) is different for calcium corrected for albumin in the yes-therapy versus no-therapy group (81% versus 46%; P = .02) and for mortality adjusted calcium (81% versus 35%; P = .01). There is no difference in the incidence of hyperphosphorus (>3rd quartile = 5.8 mg/dL). In the JM analysis, denosumab therapy is associated with outcome regardless of biomarkers of bone turnover (PTH and AF) (parameterized as AUC): HR 4 (1.34–11.9); P = .01 and HR 4.88 (1.5–15); P = .006, respectively. Both interactions of PTH and AF with the therapy variable (yes) are not significant. The interaction between therapy and vitamin D is significant [HR 0.04 (0.004–0.40); P = 0.005]. CONCLUSION The incidence of hypocalcaemia is higher in patients on denosumab antiresorptive therapy but this association is independent of bone turnover biomarkers.