BoneKEy-Osteovision | Commentary
Surrogate markers for the effect of antiresorptive therapies on non-vertebral fractures - a critique
DOI:10.1138/2002057
Commentary on: Hochberg MC, Greenspan S, Wasnich RD, Miller P, Thompson DE, Ross PD. Changes in bone density and turnover explain the reductions in incidence of nonvertebral fractures that occur during treatment with antiresorptive agents. J Clin Endocrinol Metab. 2002 Apr;87(4):1586-92.
Hochberg et al. () have carried out a meta-analysis of 18 randomized controlled trials and conclude that the reduction in the risk of non-vertebral fractures is related to the changes in bone mineral density and in bone turnover markers. They claim that the association of change in bone mineral density is just as good a prediction as the change in bone turnover markers and those gains in spine and hip BMD of 6% and 3%, respectively, are associated with a 46% decrease in non-vertebral fractures.
This conclusion sounds plausible, but there are a number of shortcomings to this analysis that should be taken into account in the interpretation of this paper. First, they consider all bone resorption markers as interchangeable. Garnero () pointed out that bisphosphonate therapy for osteoporosis results in no change (or only a small change) in free deoxypyridinoline excretion and a very large change in excretion of N-telopeptides. It is remarkable that the authors did not recognize this important observation, especially as the finding has already been reported by one of the authors (). This makes the analyses on bone turnover markers uninterpretable. Before leaving bone turnover markers (erroneously referred to as biochemical markers - BCM), the authors report r-squared values of around 0.80 for the relationship between changes in these markers and change in bone mineral density. These are far higher than any previous report, and point to fundamental flaws in the analysis approach.
Two of the authors have previously published a similar analysis on the relationship between change in spine BMD and vertebral fracture risk reduction () and came up with an estimate of 41 to 54%. There have been three recent analyses of individual studies of the relationship between spine BMD and vertebral fracture risk reduction. These have reported estimates between 4 and 28% for the proportion of vertebral fracture risk reduction explained by change in BMD of the spine (). This suggests that there is a problem with these meta-analyses. One possibility for this is the non-linear relationship between change in surrogate marker and fracture risk reduction, an observation made by Eastell et al.() for the relationship between urinary resorption markers and vertebral fracture risk on risedronate therapy. There appears to be a bone resorption level below which there is no further benefit.
Finally, it is likely that there are different pathogenetic mechanisms for the different non-vertebral fractures. For example, slender stature is a risk factor for hip fracture and obesity a risk factor for ankle fracture (). It would be helpful to see separate analyses of these different types of fractures.
It is too bad that the analysis of the bone turnover markers is flawed. It is becoming apparent that these are more useful surrogates than bone mineral density and we are gaining real insights into key determinants of fracture risk and particularly bone quality ().
This work is licensed under a
Creative Commons Attribution-Noncommercial-No Derivative Works 3.0 United States License.