Poster Presentation ESA-SRB Conference 2015

Effect of denosumab on glucose control in subjects with diabetes or pre-diabetes from the FREEDOM study (#243)

Nicola Napoli 1 , Eric Vittinghoff 2 , Nicola Pannacciulli 3 , Daria B Crittenden 3 , Andrea Wang 3 , Rachel B Wagman 3 , Ann V Schwartz 2 , Karl Peters 4
  1. Campus Bio-Medico , University of Rome, Rome, Italy
  2. University of California San Francisco, San Francisco, USA
  3. Amgen Inc., Thousand Oaks, USA
  4. Amgen Australia, North Ryde, NSW, Australia

High serum RANKL concentration was a predictor of incident type 2 diabetes (T2DM) in a population-based study, and blockage of RANKL signalling improved glucose intolerance by enhancing hepatic insulin sensitivity in mouse T2DM models (Kiechl et al. Nature Med 2013;19(3):358–366).  Denosumab is a fully human monoclonal antibody that binds with high affinity and specificity to RANKL and prevents the formation, function, and survival of osteoclasts, and is associated with vertebral and nonvertebral fracture risk reduction.  In a prior posthoc analysis of the FREEDOM trial, denosumab had no effect on incident diabetes or fasting serum glucose (FSG) in women without diabetes at baseline.  Based on the favourable effect of RANKL blockage on glucose tolerance in mouse T2DM models, we hypothesised that denosumab decreases FSG in FREEDOM subjects with diabetes or prediabetes.

Baseline diabetes status was by self-report, use of antidiabetic medication (ADM), or an FSG≥126mg/dL and prediabetes by FSG 100–125mg/dL on no ADM.  Average postbaseline FSG across visits was estimated using a repeated measures model including treatment group, baseline FSG, BMI, and age; visit; ADM use; treatment-by-visit interaction; and ADM use-by-visit interaction as fixed effects.

Baseline characteristics were similar between denosumab and placebo in both diabetes and prediabetes subpopulations.  Estimated average postbaseline FSG across visits was not significantly different between denosumab and placebo in women with either diabetes or prediabetes (p=0.20 and p=0.42, respectively); however, when censoring FSG values after ADM use in women with diabetes, estimated average postbaseline FSG across visits was lower with denosumab than placebo (p=0.02).

In this posthoc analysis, denosumab did not appear to affect FSG in subjects with diabetes or prediabetes.  There was evidence of FSG lowering with denosumab in diabetic women not currently using any ADM.  It remains to be determined whether blockage of RANKL has a clinically important effect on glucose metabolism.