Preventing metastases to bone: Denosumab or bisphosphonates?

Bone is a preferred site for metastases owing to local signals that promote skeletal colonization.1, 2 As metastases occur, the normally tempered osteoclasts are transformed into unruly bullies, proliferating and resorbing bone with abandon. For cancer patients, the unrelenting progression of bone metastases results in hypercalcemia, immobilizing fractures, loss of independence, and costly hospitalizations. Fortunately for patients, a continuing scientific effort has advanced our understanding of osteoclast biology with the identification of pathways that promote the differentiation, survival, and function of these cells.3 This progress from basic science to clinical application revealed targets for treatment and the promise of reduced suffering from skeletal-related events (SREs). More than two decades ago, bisphosphonate therapy provided an advance to older, less efficient, and untargeted therapies for SREs by reducing osteoclast survival.4 Bisphosphonates quickly became the therapy of choice, but they are not uniformly effective because some cancer patients continue to develop SREs despite therapy.

Later, a remarkable discovery identified the receptor activator of NF-κB-inducing ligand (RANKL), a powerful stimulus to osteoclast differentiation.5–7 This signal, when bound to its cognate RANK receptor, promotes differentiation of hematopoietic precursors toward a mature bone-resorbing phenotype. The greater the RANKL stimulus, the greater is the number of resorbing osteoclasts. This process is held in check only by a coexisting soluble decoy, osteoprotegerin.7–9 RANKL was immediately viewed as a bull's eye for the treatment of excess bone resorption for a number of reasons. First, the position of the protein in the osteoclastogenic pathway provided specificity, sparing hematopoietic cells that spin off this lineage upstream. Second, the necessity of RANKL for osteoclastogenesis suggested that blockade would lead to potent inhibition of bone resorption. The development of denosumab (AMG 162), a human neutralizing monoclonal antibody against RANKL, followed and has so far proven an effective therapy for patients with cancer, osteoporosis, and inflammatory arthritis. These clinical conditions share a pathophysiologic role of RANKL in bone destruction. A number of phase 2 and 3 trials have so far shown denosumab to be an effective therapy for these disorders.10–26

Questions remain as to whether denosumab should be a first-line treatment to prevent SREs in cancer patients and if patients who previously failed bisphosphonates would respond to denosumab. In this issue of the Journal of Bone and Mineral Research, Body and colleagues make an important contribution by comparing the efficacy of denosumab and bisphosphonates to prevent SREs in two studies.27 In the first study, patients with breast cancer were randomized to denosumab or intravenous bisphosphonates for their very first treatment, and the outcome was assessed over 25 weeks by measuring bone biochemical markers and SREs. In these patients, both treatments reduced bone marker levels similarly, and the overall rate of SREs was not different between groups.

The story was much different in the second study. Here, the patients were selected because of persistently elevated bone markers despite prior therapy with bisphosphonates, suggesting that they represented a group at high risk for SREs. Categorization of such patients as high risk is reasonable because urinary N-telopeptide (uNTX) > 50 nM BCE/mM creatinine, the chosen response cutoff, had been linked to a higher rate of SREs, cancer progression, and death.28 This more heterogeneous group included patients with a variety of solid tumors and multiple myeloma but predominantly cancers of the breast and prostate. Here, denosumab lowered uNTX by 80% compared with a 56% reduction in patients who remained on bisphosphonates. TRAP5b fell by 73% after denosumab compared with 11% in the bisphosphonate group. Other surrogate markers of bone turnover, including bone-specific alkaline phosphatase, osteocalcin, and P1NP, did not differ between groups, perhaps because they are not as informative about osteoclasts. More important, the number of SREs was 8% in the denosumab group compared with 17% in the bisphosphonate group. Although the overall rate of SREs in the cohort was low, the results supported the conclusion that denosumab is an effective treatment to block bone resorption and prevent SREs in cancer patients who had not responded adequately to bisphosphonate therapy.

Furthermore, the similar outcome from denosumab and bisphosphonates in the first study of treatment-naive patients suggested an efficient clinical pathway in which denosumab might be used as a second-line agent for those who failed to respond adequately to bisphosphonates. Additional data from larger trials will be required to confirm this approach. Nevertheless, in the study of Body and colleagues, the efficacy of denosumab as a first-line therapy in breast cancer was similar to that of bisphosphonates.

This and other studies have allowed an opportunity to weigh the relative merits of denosumab and bisphosphonates. One advantage of denosumab is the rapid onset of action, a fact attributed to its distinct mechanism as an inhibitor of osteoclastogenesis. Another benefit is its long duration of action and lack of tachyphylaxis in studies to date. The long duration of denosumab may be advantageous when compared with bisphosphonates, as shown by the study of Body and colleagues, in which some patients received denosumab every 12 weeks.

Other trials for metastatic bone disease also have shown efficacy with denosumab treatment every 12 or 24 weeks.17, 18, 24, 29 Another consideration is osteonecrosis of the jaw (ONJ), a complication observed in some patients after bisphosphonate treatment.30 It is possible that ONJ will be less common after denosumab treatment because the mechanism of action is different from that of bisphosphonates, and studies in which denosumab was administered with no prior exposure to bisphosphonates have not reported ONJ. Still, ONJ occurs sporadically and should be expected to be seen in cancer patients regardless of treatment.30 Again, time will tell.

Other potential adverse responses, including abnormalities in serum calcium, renal function, and the formation of antidenosumab antibodies, were not observed in the phase 2 trial of Body and colleagues. A residual concern about sporadic infections in denosumab-treated patients continues to be expressed by the Food and Drug Administration (FDA), but such infections were not observed in the study of Body and colleagues. The current data also show that denosumab efficacy is independent of the hormonal therapy used for breast cancer. Together this information suggests good tolerability and safety for denosumab, which will need to be confirmed in large trials for cancer, as has been done for treatment of osteoporosis.15

Therapy that inhibits osteoclast numbers and activity should provide relief for cancer patients; however, blockade of this one cell type is not all that could be done to avoid a spinal compression fracture or hypercalcemia. In addition to treatment of the cancer itself, we must be reminded that the osteoclast is only one of the cell types that regulates skeletal turnover. Many cancers have properties that inhibit the differentiation of osteoblasts or their ability to synthesize the matrix proteins needed to fill the resorption defect.31 In myeloma, for example, secretion of DKK1 and other factors inhibits the few remaining osteoblasts, allowing only feeble attempts to heal resorbed areas.32–36 This leaves bone at the mercy of the osteoclast and the invading cancer. Perhaps future progress in cancer-related bone disease will address the other side of this equation, the preservation and renewal of damaged bone.

Cancer patients and their doctors have made another step forward in preventing the pain and suffering that ensues from unrestrained osteoclasts. The availability of denosumab is an added weapon in the struggle to tame these bullies of bone.

Disclosures

The author states that he has no conflicts of interest.