Preserving Normal Bone Health
Predicting Who Will Develop Bone Metastases 
There is increased interest in personalized medicine whereby biological factors are utilized to predict the probability or distribution of recurrence following treatment of early cancer. In breast cancer, gene expression profiling has identified genetic signatures that are strongly associated with prognosis52,53 possibly outperforming conventional panels of clinical and pathological criteria. However the risk of metastasis per se does not necessarily predict for metastasis to specific sites. Recently a 69 gene signature was described that predicted specifically for metastasis to bone as opposed to metastasis in other sites.54 The set of genes expressed suggested involvement of the fibroblast growth factor receptor (FGFR) pathway in the process of metastasis to bone. Subsequently, a classifier of 31 genes was constructed which was able to predict all tumors relapsing in bone with a specificity of 50%. Much more work is required to confirm the utility of genetic profiling in clinical management but this approach holds much promise.
There are also an increasing number of individual tissue-based markers that predict for metastasis to bone.The association between estrogen receptor (ER) expression and bone metastasis has been appreciated for over 20 years.55 More recently other markers, especially the non-collagenous bone proteins, have been suggested as predictors of bone recurrence (Table 5). These have included bone sialoprotein (BSP), which in addition to appearing to predict for bone metastasis in breast cancer,56 may be relevant in other tumors with a propensity to metastasize to bone.57 Again, further study is required but, assuming bone specific therapies do have an impact on the development of metastases, identification of patients at high risk of bone metastasis, especially by simple tissue based assays such as BSP, would be very valuable for patient selection over and above the traditional clinico-pathological characteristics.
Prioritizing Bisphosphonate Use in Advanced Cancer
Despite the obvious clinical benefits of bisphosphonates, data from recent studies suggest that only a proportion of SREs are prevented. In addition, because some patients do not experience a SRE despite the presence of metastatic bone disease, it is often not possible to predict accurately whether an individual patient needs or will benefit from bisphosphonate treatment. Bisphosphonates are a relatively costly additional intervention in cancer care that is now potentially applicable to a very large proportion of patients with advanced malignancy. The cost effectiveness of routine long-term treatment has been questioned, and prioritization of bisphosphonate use is needed.58,59
A report on the use of the bone resorption marker n-telopeptide of type I collagen (Ntx) suggested that biochemical monitoring could be useful to identify patients at increased risk of bone complications.58 In this study of 121 patients, all skeletal complications, hospital admissions for control of bone pain, and deaths during the period of observation were recorded. Ntx was strongly correlated with the number of skeletal events and/or deaths (p<0.001). Patients with Ntx values above 100 nmol/mmol creatinine were many times more likely to experience an SRE or death than those with Ntx below this level (p<0.01).Thus, a more cost-effective use of bisphosphonates, particularly in patients with additional extensive visceral metastases or solid tumors associated with a short life expectancy, might be to prioritize them in patients with raised Ntx levels.
Further clinical evidence of correlations between bone marker levels and patient outcomes has come from retrospective analyses of large randomized trials with zoledronic acid.60,61 Elevated on-study bone marker levels correlated with negative clinical outcomes in patients with metastatic breast cancer, prostate cancer, lung cancer, or other solid tumors.52 Specifically, patients with elevated urinary Ntx levels either at baseline or at their most recent assessment had an approximate 2-fold increase in their risk of disease progression and an approximate 2- to 3-fold increase in their relative risk of skeletal-related events (SREs) compared with patients with low Ntx levels.60,61 Elevated Ntx levels clearly have negative prognostic implications, and there is a continuum of risk among all patients with bone lesions.62
It has been proposed that the aim of bisphosphonate treatment should be to adjust the dose and schedule to maintain a normal rate of bone resorption. Randomized trials to assess this approach are in progress, including the BISMARK trial (cost-effective use of BISphosphonates in metastatic bone disease: a comparison of bone MARKer-directed zoledronic acid therapy to a standard schedule). This study is planned to enroll 1,500 patients with bone metastases from breast cancer. Patients will be randomized to receive either zoledronic acid (4 mg) every 3 to 4 weeks or zoledronic acid (4 mg) on a marker-directed schedule based on changes from baseline Ntx.The primary endpoint is the frequency and timing of SREs. Secondary endpoints include the evaluation of pain, quality of life, incidence of new bone metastases and overall survival.
Uncertainty remains as to the appropriate duration and schedule of treatment. Bone marker-directed therapy is under evaluation in randomized clinical trials and the value of maintenance therapy after one to two years of treatment may become clear from the Optimize trial ongoing in the United States.
Treatment-Induced Bone Loss
Breast Cancer 
Tamoxifen has been the cornerstone of adjuvant endocrine therapy for breast cancer for several decades. Recent research indicates, however, that the use of aromatase inhibitors (AIs), anastrozole, letrozole and exemestane, may further reduce the risk of recurrence after a diagnosis of estrogen receptorpositive breast cancer when given either in place of tamoxifen or after a few years of tamoxifen therapy.63-67 Overall, Als have a favorable side-effect profile but, due to the known relationships between residual estrogen levels and fracture risk,68 the associated marked reduction in estradiol would be expected to have significant effects on bone physiology. This has been confirmed in a range of studies and the use of Als is indeed associated with increased bone turnover leading to a loss of bone mineral density (BMD) and an increased rate of fracture (Table 6) .
In comparison to naturally occurring bone loss, Al-associated bone loss is more marked with particularly rapid rates of loss in the immediate postmenopausal period.69-71 However the risk for fractures is more complex than changes in BMD alone, and, as in postmenopausal osteoporosis, other factors including family history, lifestyle, concomitant medication, and nutrition, all influence the risk for bone loss and may interfere with assessing the effect of therapy-induced bone loss.71
The bisphosphonates are the preferred treatment for Alinduced bone loss.The results of several intervention studies with zoledronic acid have been recently published,69-75 and preliminary data are also available from smaller studies using oral bisphosphonates at standard osteoporosis doses, such as ibandronate in the ARIBON trial74 and risedronate in the SABRE trial (Table 7). The data from these studies demonstrate the ability of bisphosphonates to prevent bone loss when given concurrently with Als.
Prostate Cancer
Many men with prostate cancer are at risk of developing osteoporosis, largely because of the androgen deprivation therapy (ADT) they receive for their cancer. ADT is being introduced earlier and earlier in the course of the disease with the result that men may experience many years of androgen suppression. ADT results in substantially reduced serum concentrations of testosterone (to less than 5% normal level) and estrogen (to less than 20% of normal level).
It might be expected that androgen deprivation would lead to increased bone loss and increased fracture rate. As yet there appear to have been no large prospective studies of the relationship between fracture rate and androgen deprivation therapy but retrospective studies indicate a significantly increased fracture rate with an estimated 10 year fracture free survival of only 60%.76 As the potential scale of the problem is being realized, management guidelines to assess those at risk are in development and therapeutic options such as bisphosphonates to prevent or treat therapy related bone loss in prostate cancer are under investigation. Pamidronate given on a three monthly schedule has been shown to prevent loss in BMD in patients with locally advanced prostate cancer.77 A multi-centre prospective study has evaluated zoledronate in locally advanced or recurrent prostate cancer.78 At 1 year, in the zoledronic acid arm, BMD increased by 5.3 at the lumbar spine and 1.1% in the total hip. In the placebo arm, the BMD decreased by 2% in the lumbar spine and 2.8% in the total hip.%
A number of other phase II, III, and IV trials are ongoing to evaluate the safety and efficacy of zoledronic acid to prevent bone loss caused by cancer treatment induced bone loss and to prevent further bone loss in osteoporotic patients with prostate cancer.