The correction of BMD measurements for bone strontium content

Strontium ranelate (SR) is a new oral treatment for osteoporosis associated with large increases in bone mineral density (BMD) compared with alternative therapies such as bisphosphonates. Much of the BMD increase during SR treatment is a physical effect caused by the increased attenuation of X-rays due to the accumulation of strontium in bone tissue. The aim of this study was to assess the contribution made by bone strontium content (BSC) to the overall BMD increase by evaluating the percentage F of the BMD change explained by the physical presence of strontium in bone. A value of F less than 100% would provide evidence of the anabolic effect of SR as an additional factor contributing to the overall BMD increase. Studies of mixtures of strontium hydroxyapatite (SrHA) and calcium hydroxyapatite (CaHA) scanned on a variety of dual-energy X-ray absorptiometry (DXA) systems show that a 1% molar ratio of SrHA/(CaHA+SrHA) causes a 10% overestimation of BMD. The correction of spine BMD measurements for the physical effects of strontium depends on knowledge of 2 further factors: (1) bone biopsy measurements of iliac crest BSC and (2) the ratio R of BSC at the DXA site to BSC at the iliac crest measured in animal studies. We used clinical trial data and values of R(spine) measured in studies of monkeys and beagle dogs to determine values of F(spine) for 1, 2, and 3 yr treatment with SR. Based on the average value of R(spine) approximately 0.7 for male and female monkeys, we found values for F(spine) approximately 75-80% for 1, 2, and 3 yr of treatment. Using the value of R(spine) approximately 1.0 from the beagle study gave values of F(spine) approximately 100%. Although values of F(spine) as low as 40% are possible, we conclude that the most likely figure is 75% or greater. However, it is apparent that there are large uncertainties in the correction of BMD results for the effect of bone strontium and that the most important of these is the inference of BSC values at DXA scan sites from measurements of iliac crest bone biopsy specimens.