Markers of Bone Turnover and Postmenopausal Hormone Therapy

ABSTRACT & COMMENTARY

Synopsis: Bone markers correlate with bone loss but have little practical clinical use.

Source: Nilas L, et al. Menopause 1996;3:190-196

Nilas and colleagues from Denmark prospectively studied 75 women randomized to treatment either with a daily combination of 2 mg estradiol valerate and 1 mg cyproterone acetate or a sequential regimen with 2 mg estradiol valerate and 75 mcg of levonorgestrel on days 17-28 or placebo. During the three-year study, the placebo group demonstrated a steady loss of bone in the forearms and the spine, while bone mass increased in both treatment groups. The biochemical markers of bone turnover included bone Gla protein (osteocalcin), an indicator of bone formation, and fasting urinary crosslinks, an indicator of bone resorption. Changes in bone density correlated with the biochemical markers; a loss of bone was associated with a decrease in the urinary crosslinks during the first 3-6 months, which then remained at stable levels. The blood levels of the bone Gla protein demonstrated a gradual decline over the first 1.5 years and then remained at stable levels. Thus, in the hormone-treated groups, indicators of bone resorption declined rapidly, and this was followed by a slower reduction in the indicator of bone formation. The response in the forearm was significantly related to changes in the urinary crosslinks after six and 12 months, but only the three-year changes in the spine were related to the blood level of the Gla protein.

COMMENT BY LEON SPEROFF, MD

While a correlation between the biochemical markers (the newer urinary crosslinks and the older blood Gla protein [osteocalcin]) can be demonstrated with bone density in a study such as this, the critical question is whether these biochemical markers have clinical utility. These tests that have emerged in recent years use the measurement of bone-related proteins that are 100% specific for bone. Bone formation can be assessed by measuring serum osteocalcin, serum bone-specific alkaline phosphatase, and serum procollagen I peptides. Bone resorption is assessed by measuring the following: urinary proteins, pyridinoline, deoxypridinoline, n-telopeptides of collagen, and c-telopeptides of collagen. After menopause, the markers for both bone formation and bone resorption increase, and, with estrogen therapy, these markers return to normal levels. The higher the level of the bone markers, the greater the risk of fractures.

The clinical use of these markers is affected by the steady state of the bone. With acute changes, either the removal of estrogen or the initiation of estrogen therapy, the resorption markers change quickly within days or weeks, while the formation markers change 3-6 months later when a new steady state is reached. Although there was a correlation between the highest rate of bone loss and the least change in the urinary resorption marker in this story, the individual variation was so great that the predictive value is low. The bottom line is that these markers cannot be substituted for the measurement of bone density. They can be used to assess response to treatment. In this case, repeated measurements would be necessary. The full utility and cost-effectiveness of such management are not yet demonstrated. In my view, the utilization of bone markers is still limited to research protocols until firm conclusions are available for clinicians and patients.