A major clinical feature in multiple myeloma is the development of osteolytic bone disease. The increase in bone destruction is due to uncontrolled osteoclastic bone resorption. Until recently the factors responsible for mediating the increase in osteoclast formation in myeloma have been unclear. However, recent studies have implicated a number of factors, including the ligand for receptor activator of NFκB (RANKL) and macrophage inflammatory protein-1α. The demonstration that increased osteoclastic activity plays a central role in this process and the identification of molecules that may play a critical role in the development of myeloma bone disease have resulted in studies aimed at identifying new approaches to treating this aspect of myeloma.
Studies have been performed to determine the ability of recombinant osteoprotegerin (Fc.OPG), a soluble decoy receptor for RANKL, and potent new bisphosphonates to inhibit the development of myeloma bone disease in the 5T2MM murine model of multiple myeloma.
Fc.OPG was shown to prevent the development of osteolytic bone lesions in 5T2MM bearing animals. These changes were associated with a preservation of the cancellous bone loss induced by myeloma cells and an inhibition of osteoclast formation. Bisphosphonates, including ibandronate and zoledronic acid, were also shown to inhibit the development of osteolytic bone lesions in the 5T2MM model and alternative models of myeloma bone disease.
Multiple myeloma is a disease characterized by the infiltration and growth of myeloma cells in the local bone marrow microenvironment. Patients with multiple myeloma commonly develop osteolytic bone disease, which predominantly affects the skull, ribs, vertebrae, pelvis, and proximal long bones. Characterized by the presence of bone pain, pathologic fractures, and hypercalcemia, this aspect of myeloma represents a major cause of morbidity. Indeed, more than 50% of patients will have evidence of vertebral fractures at diagnosis and a significant proportion of patients may also have nonvertebral fractures.1, 2 Thus, the majority of patients will present with evidence of pathologic fractures or develop lytic bone lesions during the course of their disease. Although the development of osteolytic bone lesions is a major cause of morbidity, our understanding of this aspect of the disease and the means by which to treat it has, until recently, received relatively little attention. However, new studies have bought about an improvement in our understanding of both the cellular and molecular mechanisms that may contribute to the development of lytic bone lesions. This knowledge has resulted in the identification of new therapeutic approaches to treating myeloma bone disease.
Cellular Mechanisms of Bone Loss in Myeloma
Information on the cellular mechanisms responsible for the bone disease that occurs in patients with multiple myeloma has been generated largely from histomorphometric studies of trans-iliac crest bone biopsies, taken from patients with myeloma. These studies have shown that in biopsies with evidence of myeloma cell infiltration, the proportion of bone surface undergoing resorption is increased and the numbers of osteoclasts present on bone surfaces may also be elevated.3–5 Furthermore, such studies have also suggested that the amount of bone resorbed within sites of bone resorption is increased, which is consistent with an increase in the number and/or activity of osteoclasts at these locations.6 Osteoclastic bone resorption has not only been shown to be increased in patients with active disease but is also increased in patients defined as having early myeloma4 or with a low tumor infiltration of the bone marrow.5 Bataille et al.7 have also shown that a proportion of patients with monoclonal gammopathy of underdetermined significance have elevated histologic indices of bone resorption. These patients were those most likely to develop overt myeloma, which suggests that activation of bone resorption may be an early event in the disease process. In addition to the increase in bone resorption seen in patients with myeloma, histomorphometric studies have shown that bone formation may also be affected. In the early stages of myeloma, bone formation may also be increased, reflecting the tight coupling of resorption to formation.4 However, as the disease progresses, bone formation is decreased, which leads to an increased negative remodeling balance, an uncoupling of resorption and formation, and the rapid bone loss that is seen in the osteolytic bone disease characteristic of myeloma.4, 5
Bisphosphonates as Inhibitors of Myeloma Bone Disease
Since increased bone resorption is a key event in the development of osteolytic bone disease, blocking osteoclast activity has been the subject of substantial research effort. The most well characterized inhibitors of osteoclastic resorption are bisphosphonates. Bisphosphonates are analogues of inorganic pyrophosphate in which the central oxygen atom is replaced by a carbon atom to give the P-C-P motif. This motif is responsible for the high affinity of these compounds for hydroxyapatite and bone. The presence of a hydroxyl group in the R1 position increases the affinity of bisphosphonates for hydroxyapatite, whereas modifications to the R2 position are responsible for determining anti-resorptive potency. These changes have resulted in the development of a number of bisphosphonates with a broad range of anti-resorptive activities.
Over the last decade clinical studies have shown that bisphosphonates can be effective in treating bone disease found in patients with myeloma.1, 8, 9 Although early studies with etidronate suggested that this bisphosphonate was unable to have a significant effect on the development bone disease, studies of clodronate have revealed that this compound is able to reduce the incidence of skeletal related events in patients with myeloma.1, 8 Pamidronate, when administered intravenously, is also effective in the management of myeloma bone disease,9, 10 although when given orally this compound was reported to be less effective.11 Zoledronic acid, which is more potent than pamidronate, has also been shown to be effective in treating the bone disease observed in patients with myeloma.12
In all of these studies, bisphosphonate treatment has been shown to have no effect on overall patient survival. However, an increase in survival was reported in a subgroup of patients treated with pamidronate and on salvage therapy.10 In a long-term followup of patients with multiple myeloma, clodronate treatment was associated with an increase in survival in those patients who had no skeletal fractures at presentation.13 These observations have raised the possibility that bisphosphonates may modify the local bone microenvironment to create a less favorable environment for the growth of myeloma cells. However, these data require further investigation.
The effect of several of the new, more potent bisphosphonates has now been investigated in murine models of myeloma bone disease (Table 1). The use of such animal models has also allowed the effect of bisphosphonates on tumor activity to be investigated. The first of such studies were performed in the 5T2MM murine model of myeloma. This model involves injecting 5T2MM murine myeloma cells into the tail vein of C57BL/KaLwRij mice.14 The tumor cells home to the bone marrow, where they develop an osteolytic bone disease characterized by the presence of lytic bone lesions, a decrease in cancellous bone volume, and an increase in the number of osteoclasts.15, 16 Pamidronate was able to reduce radiographic and histologic indices of bone destruction in the 5T2MM model.17 Although treatment had no effect on tumor burden, pamidronate did promote an increase in survival when given from the time of tumor cell injection.18 Pamidronate has also been investigated in SCID mice bearing human bone and injected with human tumor cells.19 Treatment with pamidronate prevented further development of bone disease and, in two out of three experiments, decreased serum paraprotein concentrations.19
Table 1. Summary of the Effect of Bisphosphonates on the Development of Bone Disease and Indices of Tumor Activity in Murine Models of Multiple Myeloma
Dallas et al. have investigated the effect of ibandronate in the 5TGM1 model of myeloma.20 Daily subcutaneous injections of ibandronate reportedly prevented the development of osteolytic lesions. Treatment had no effect on serum paraprotein concentrations or tumor burden.20 Cruz et al.21 have investigated the effect of ibandronate treatment in the ARH-77/SCID model. Ibandronate reduced the number of osteolytic bone lesions and the proportion of bone surface undergoing resorption. A significant increase in the time to paraplegia was observed in animals treated with ibandronate for seven days prior to injection of ARH-77 cells. No effect was seen if treatment began once tumor cells had been injected. The effect of zoledronic acid has also been investigated in murine models of myeloma. Yaccoby et al.19 have shown that zoledronic acid can prevent the bone loss and the increase in osteoclast numbers induced by human myeloma cells in SCID-hu mice. Treatment of myeloma bearing animals reduced serum paraprotein concentration, whereas pre-treatment appeared to prevent the appearance of a serum paraprotein altogether.19 In the 5T2MM model of myeloma, zoledronic acid also prevents the development of osteolytic bone lesions, the decrease in cancellous bone loss, and the loss of bone density induced by the presence of tumor cells.22 In this study zoledronic acid treatment was also able to reduce serum paraprotein concentrations and to promote a significant increase in survival.22
Molecular Mechanisms of Osteoclast Activation in Myeloma
Targeting osteoclastic bone resorption with bisphosphonates is an effective means of managing the development of myeloma bone disease. However, it is anticipated that an improved understanding of the molecular mechanisms of bone loss will lead to the rational design of new therapeutic strategies.
Histomorphometric studies have shown that the increase in bone resorption in multiple myeloma is associated with tumor cell infiltration and can be correlated with tumor burden.5, 23 These observations suggest that the myeloma cells stimulate the resorption process directly, either by producing soluble factors that promote bone resorption, by interacting directly in a cell contact dependent manner to stimulate osteoclastic bone resorption, or by modifying expression of factors that indirectly upregulate osteoclastic activity. In a search for the factors responsible for mediating osteoclastic bone resorption in myeloma, Mundy et al.24, 25 reported the production of an osteoclast activating activity by human myeloma cell lines and primary cells isolated from patients with multiple myeloma. However, the specific identity of this factor(s) remains unclear. A number of early studies have implicated interleukin-1β or lymphotoxin in this process; however, as yet there are little functional data to support a causal role. More recently new candidate molecules have been identified, including macrophage inflammatory protein 1-α (MIP 1-α) and the ligand for receptor activator of NFkB (RANKL).
Macrophage inflammatory protein 1-α
Macrophage inflammatory protein 1-α is a member of the chemokine family with the ability to induce the formation of osteoclasts in bone marrow cultures and to be chemotactic for osteoclasts. MIP 1-α has been reported to be produced by myeloma cells and to be elevated in the bone marrow plasma of patients with active disease.26 Antibodies to MIP 1-α inhibit the osteoclast forming activity present in the bone marrow plasma of patients with myeloma,26 and antisense constructs inhibit the development of osteolytic lesions in the ARH-77 model of lytic bone disease.27
Receptor activator of NFkB
RANKL,28 which is also known as osteoprotegerin ligand,29 osteoclast differentiation factor,30 or TNF-related activation induced cytokine (TRANCE),31 has recently been shown to play a key role in the normal development of osteoclasts.32 RANKL has been shown to be expressed by stromal cells and osteoblasts in the local bone marrow microenvironement, where it can bind to its receptor RANK on the surface of osteoclast precursors.28 The interaction between RANKL and RANK plays a critical role in promoting osteoclast differentiation and bone resorption and may also be responsible for activating mature osteoclasts to increase bone resorption.33 A soluble decoy receptor, called osteoprotegerin (OPG), has also been identified,34 which binds to RANKL, inhibiting its interaction with RANK and blocking osteoclast formation. Overexpression of OPG in mice, or mice deficient in RANKL, has been reported to reduce osteoclast formation and results in the development of osteopetrosis.32, 34 In contrast, mice deficient in OPG develop osteoporosis.35, 36
Since the RANKL system plays such a critical role in normal osteoclast development, it is possible that aberrant regulation of this system by myeloma cells may contribute to the development of myeloma bone disease. Immunocytochemical studies have shown increased expression of RANKL in bone marrow stromal cells from patients with myeloma; however, in these studies, RANKL expression was not detected in myeloma cells.37–39 This contrasts with other reports that have detected expression of RANKL by flow cytometry in murine myeloma cells and primary cells isolated from the bone marrow of patients with myeloma.16, 40, 41 Although it is not clear why there are discrepancies between these studies, they do suggest that expression of RANKL is upregulated in the cells found in the local bone marrow microenvironment in myeloma. In addition to upregulation of RANKL expression, OPG expression appears to be down-regulated in cells found in the bone marrow microenvironment. Myeloma cells inhibit production of OPG in stromal cells and osteoblasts in a contact dependent manner.38 Furthermore, serum concentrations of OPG are significantly lower in patients with myeloma when compared to healthy controls.42 Taken together, these data suggest that the RANKL/OPG axis may be disregulated in myeloma.
Osteoprotegerin as an Inhibitor of Myeloma Bone Disease
Since the RANKL system may be abnormally regulated in myeloma, targeting this system may represent a novel therapeutic approach. Inhibiting the actions of RANKL could be achieved in a number of ways, including by utilizing a soluble form of RANK to prevent the association of RANKL with the membrane-bound form of RANK. This strategy has been employed by a number of groups in several different models of myeloma bone disease. Oyajobi et al.43 have shown that a recombinant form of RANK, RANK.Fc, is able to prevent the development of osteolytic bone lesions in the 5TGM-1 model of myeloma. Furthermore, Pearce et al.37 and Yaccoby et al.19 have reported that RANK.Fc can block the development of bone disease in the ARH-77 and SCID/Hu models of myeloma, respectively. Treatment of SCID/Hu mice bearing human myeloma with RANK.Fc was also associated with a decrease in paraprotein.19
An alternative strategy for blocking the activity of RANKL is to manipulate expression of OPG. As a soluble decoy receptor OPG can bind RANKL, which prevents it from binding to RANK and inhibits osteoclastic bone resorption. Recombinant OPG has been shown to inhibit the development of hypercalcemia induced by Colon-26 cells injected subcutaneously into normal mice.44 The development of hypercalcemia was associated with increased osteoclastic bone resorption, which was blocked by treatment with OPG.44 OPG is also able to prevent osteoclast formation and the development of lytic bone lesions induced by MDA-MB-231 cells in athymic mice.45 The ability of OPG to inhibit the development of bone disease has also been investigated in the 5T2MM and 5T33MM murine models of myeloma.16, 46 Treatment of mice with established growth of 5T2MM cells (i.e., with a detectable serum paraprotein) with recombinant OPG was able to prevent the development of radiographically detectable osteolytic bone lesion in the long bones (Fig. 1). Histologic assessment of the long bones of 5T2MM-bearing animals showed a significant decrease in cancellous bone volume, which could be partially prevented by treatment with OPG. 5T2MM bearing animals also had increased numbers of osteoclasts lining bone surfaces (Fig. 2). This was most apparent on the cortical-endosteal surface, as tumor bearing animals had little cancellous bone. Treatment with OPG completely prevented osteoclast formation in this model. Treatment of 5T2MM-bearing animals with OPG was associated with a 25% decrease in serum paraprotein concentration, although this was not statistically significant.16 Conversely, in the 5T33MM murine model of myeloma, which develops more rapidly than the 5T2MM model (4–5 weeks rather than 12–14 weeks), OPG treatment caused significant decreases in both serum paraprotein and the proportion of tumor cells in the bone marrow.46 The effect of manipulating OPG expression has also been investigated in the ARH-77/SCID model.47 ARH-77 cells engineered to overexpress OPG promoted an increase in bone mineral density when injected in vivo and were associated with an increase in survival when compared to parental cells. An osteoprotegerin construct has also been investigated in patients with multiple myeloma with confirmed osteolytic bone disease. Treatment was associated with a decrease in a biochemical marker of bone resorption, the N-telopeptide of collagen, suggesting that this approach may have clinical utility.48
The bone disease that develops in patients with myeloma is clearly a major cause of morbidity. Bisphosphonates are potent inhibitors of osteoclastic bone resorption and are effective in preventing the development of myeloma bone disease in vivo. Improvements in our understanding of the cellular and molecular mechanisms responsible for the development of myeloma bone disease have led to the identification of new therapeutic targets for managing this aspect of myeloma. One such approach has been to target RANKL with osteoprotegerin. This approach, along with others that target RANKL, has been shown to inhibit the development of bone disease in vivo in murine models of myeloma. A number of these studies have suggested that, by targeting bone resorption and/or the molecules that mediate the increase in bone resorption, it may be possible to modify the myeloma disease itself. One of the challenges in the future will be to establish the importance of these observations.