Mechanism of action study to evaluate the effect of rosiglitazone on bone in postmenopausal women with type 2 diabetes mellitus: rationale, study design and baseline characteristics

Objectives: Post-hoc analyses have shown an increase incidence of fractures among type 2 diabetes (T2DM) patients treated with thiazolidinediones (TZDs). The mechanisms by which TZDs may be associated with increased fracture risk is not well understood. This article describes the study design and baseline characteristics for a prospective, randomized, double-blind, active-controlled trial to evaluate the effects of rosiglitazone on changes in measures of skeletal structure, surrogates of bone strength and metabolism.

Methods: Postmenopausal women without osteoporosis and diagnosed with T2DM were randomized in a double-blind design to either rosiglitazone or metformin for 52 weeks, then all subjects received open-label metformin for 24 weeks. Study endpoints included changes in bone mineral density (BMD), quantitative computed tomography (QCT), digitized hip radiography (HXR) and high resolution magnetic resonance imaging (hrMRI). Serum markers of bone metabolism and indices of glycemic control were assessed within and between treatment groups.

Results: A total of 226 subjects were randomized. Baseline characteristics included: age 63.8 ± 6.5 years; years postmenopausal 16.9 ± 8.4; duration of diabetes 3.5 (1.8-7.8) years; body mass index (BMI) 31.4 ± 5.9 kg/m(2); and glycated hemoglobin (HbA1c) 6.4 ± 0.65%. At baseline, mean T-scores were -0.95 ± 0.91 at the femoral neck, -0.02 ± 0.97 at the total hip and -0.55 ± 1.25 at the total spine. Since there are no well recognized techniques to determine bone mass and structure at the distal limbs (cortical bone sites where fractures were reported in RSG subjects), using the femoral neck as a surrogate for these areas may be a potential limitation of the study.

Conclusion: This is the first randomized trial utilizing multiple techniques to evaluate bone mass, structure, serum markers of bone remodeling, and potential reversibility of changes after discontinuation of rosiglitazone. This study will provide information about RSG bone effects in a population of postmenopausal women at risk for bone loss and subsequent fracture.

Clinicaltrialsgov number: NCT00679939.