Bone is constantly being remodeled throughout adult life, which results in the removal of old bone and its replacement by new bone. In general, the processes of bone resorption and formation are "coupled" so that there is no net change in bone mass. The exceptions to this are during growth and development (where bone formation outstrips bone resorption, leading to a net increase in bone mass), and diseases such as postmenopausal osteoporosis (where bone resorption outstrips bone formation, leading to a net decrease in bone mass). Thus, while BMD provides a static picture of the skeleton, the hope has been that biochemical markers of bone formation and resorption may provide dynamic measures of bone remodeling and thus potentially be useful in predicting the course of changes in bone mass. Moreover, there is some evidence that in the adult, increased bone remodeling may be structurally destabilizing for the skeleton, leading in and of itself to an increase in fracture risk. While currently available markers have not fulfilled all of their promises, there are nonetheless certain situations where assessment of bone turnover using these markers may be of clinical utility.

Table 7 lists the currently available markers of bone formation and resorption. Since increases in bone resorption (relative to formation) are typically of the greatest concern in postmenopausal and other forms of osteoporosis, bone resorption markers are theoretically more attractive. However, until recently, these could only be measured in the urine, which had its problems in terms of sample collection and requirement for normalization to creatinine excretion (which introduced additional variability into an already highly variable measurement). The availability of serum markers for bone resorption should circumvent some of these problems with the urine resorption markers.

A detailed discussion of all of these markers is beyond the scope of this chapter, and likely not of particular use to the practicing clinician. Suffice it to say that for assessing bone formation, either serum BSAP or osteocalcin are perhaps the most widely used, and for bone resorption, urine NTx or CTx and, more recently, serum NTx or CTx appear to be most useful. BSAP and osteocalcin are both produced by osteoblastic cells in the process of bone formation, and NTx and CTx are cross-linked telopeptides that are released from type I collagen in bone during the process of bone resorption (19).

While markers of bone turnover are generally inversely correlated with BMD (20), these correlations are not strong enough to have any value in terms of predicting bone mass for a given individual. Thus, they cannot be used in lieu of a BMD measurement.

Perhaps the most intriguing use of these markers lies in their potential for predicting fracture risk, independent of BMD. There are now several studies that have shown that bone turnover may be an independent predictor of fracture risk (20, 21, 22). Thus, in a prospective cohort study of elderly (age > 75 yrs) French women, urinary CTx and free Dpd excretion above the upper limit of the premenopausal range (i.e. mean + 2 SD) was associated with an increased risk of hip fracture (see Figure 1, below), even after adjusting for femoral neck BMD (22). In recent population-based studies in women, bone resorption markers were negatively correlated with BMD of the hip, spine and forearm, and women with osteoporosis were more likely to have high bone turnover (20). Moreover, a history of osteoporotic fractures of the hip, spine, or distal forearm was associated with reduced hip BMD and with elevated biochemical markers of bone resorption (20). The mechanisms by which increased bone turnover adversely affects fracture risk include exacerbation of rates of bone loss (23) (see below), microarchitectural deterioration of the skeleton due to perforation of trabeculae and loss of structural elements of bone (24), or a reduction in bone strength due to an enlarged remodeling space (21, 25). Thus, bone turnover, as assessed by biochemical markers, appears to have a significant impact on the risk of fracture independent of BMD. However, until more prospective data are available, particularly in younger women than those studied in the French cohort (22), the routine use of bone biochemical markers to complement BMD measurements for prediction of fracture risk cannot be recommended at this time.

Estrogen deficiency at the menopause increases the rate of bone remodeling, which results in high turnover bone loss. This is reflected by a significant increase in the mean value of markers of resorption and formation from before to after menopause. Moreover, the individual variability in the bone turnover markers also increases following the menopause, reflecting a variable skeletal response among different individuals to estrogen deficiency. This is also reflected in the variable rates of bone loss observed among women following the menopause. Several studies now indicate that, at least for groups of individuals, bone biochemical markers can be used to predict the rate of bone loss. Thus, Hansen et al. (23) measured the bone mineral content of the forearm at baseline and 12 years later and attempted to predict the observed rate of bone loss by using a biochemical model that included fat mass, serum alkaline phosphatase activity, fasting urinary calcium to creatinine ratio, and fasting urinary hydroxyproline to creatinine ratio. Using these relative crude bone turnover markers, they were able to predict the observed bone mass 12 years later with a high degree of accuracy. Thus, these and other data (26) do suggest that bone turnover markers, either individually or in combination, may be able to predict rates of bone loss, thus complementing the static measurement of BMD. However, while this appears to be the case for groups of individuals, whether bone turnover markers can predict the rate of bone loss in a given individual, particularly in light of the day to day variability in some of the markers, remains to be established.

Several studies indicate that individuals with the highest levels of bone turnover appear to have the best response to antiresorptive therapy (i.e. with estrogen, calcitonin, or bisphosphonates). Thus, in a prospective 2 year study of hormone replacement therapy (HRT), Chesnut et al. (27) found that subjects in the highest quartiles for baseline urinary NTx excretion demonstrated the greatest gain in BMD in response to HRT (Figure 2).

Perhaps the best established clinical use of bone biochemical markers at present is in monitoring the effectiveness of antiresorptive therapy. Considerable data now indicates that following initiation of antiresorptive therapy, there is a significant reduction in markers of bone resorption within 4 to 6 weeks (28, 29), and in markers of bone formation in 2 to 3 months (28). Thus, bone turnover markers can be used to determine when therapy is ineffective. Antiresorptive agents should produce a reduction in the markers of resorption of between 20% to 80%, depending on the agent and the markers. Thus, despite the potential technical and biologic variability of the markers, changes of this magnitude should be clinically meaningful. For most treatments, the nadir will be reached between 2 to 3 months after initiation and will remain constant as long as the patient is on therapy (27).

Failure to show the expected reduction in resorption markers could indicate non-compliance with therapy or the possible need to change the dose or type of therapy. This use of bone biochemical markers offers a marked advantage over using BMD to assess the effectiveness of therapy, since the interval between serial measurements of BMD must be at least 12 (and possibly 24) months before significant changes in BMD can be documented or, more importantly, the lack of change in BMD can be established with any certainty.

In summary, the potential role of biochemical markers of bone turnover in the evaluation and management of the patient with osteoporosis is still evolving. A baseline measurement prior to initiation of antiresorptive therapy, followed by a repeat measurement in 2 to 3 months, may be useful in some patients, particularly those that require relatively early feedback on the possible efficacy of the treatment that they are receiving. While bone turnover markers cannot be accurately used to predict rates of bone loss in an individual patient, their independent association with fracture risk suggests that they may provide additional information beyond BMD on the subsequent likelihood of fracture.

Thus, a case could be made for assessing bone turnover using a formation and resorption marker in a patient with osteopenia to help determine whether treatment should be instituted. While defensible based on existing data, current evidence is insufficient to recommend this approach under routine circumstances.