Multiple myeloma is characterized by a unique form of destructive bone disease which occurs in the majority of patients. The bone destruction, which is progressive, is responsible for the most prominent and distressing clinical features of this disease, namely intractable bone pain, fractures occurring either spontaneously or following trivial injury, and hypercalcemia with its attendant symptoms and signs. Although myeloma is a disease with protean features resulting from the effects of the disease on multiple organ systems, perhaps its most important clinical manifestation, and certainly the one which most often heralds the onset of the disease, is the bone disease with which it is associated.
Nature of the Bone Disease
The bone lesions in myeloma, which typically appear on radiographs as radiolucent areas, occur in several patterns. Occasionally, patients develop single osteolytic lesions that are associated with solitary plasmacytomas. Some patients have diffuse bone loss (osteopenia), which mimics the appearance of osteoporosis, due to the dissemination of the myeloma cells throughout the axial skeleton. However, in most patients there are multiple discrete lytic bone lesions occurring at the site of deposits or nests of myeloma cells. Rarely, patients with myeloma have an increase in the formation of new bone around myeloma cells rather than lytic lesions or bone loss, a rare condition known as osteosclerotic myeloma. The extent of the bone disease is an important factor in the prognosis of a patient with myeloma.
Bone remodeling is profoundly impaired in almost all patients with myeloma. Although some patients with other hematopoietic malignancies such as B-cell lymphoma and adult T cell leukemia occasionally have skeletal manifestations, they are rarely as severe and certainly not as common as in myeloma. There is evidence that excessive bone resorption is an early tumor-induced event in myeloma and is often associated with progression from quiescent monoclonal gammopathy of undetermined significance (MGUS) or smoldering myeloma to active myeloma. Bone pain is a major cause of morbidity in myeloma and 75-80% of patients present with bone pain as a predominant symptom. The bone pain is often unremitting, but occasionally fluctuates in intensity for reasons that are unknown. Patients with myeloma bone disease are susceptible to fractures occurring either spontaneously or following trivial injury. These pathological fractures mainly involve the vertebrae, ribs, and long bones (most commonly), but occasionally occur in other sites such as the sternum and pelvis. Hypercalcemia, which occurs mainly in patients with advanced disease, is accompanied by characteristic distressing symptoms and signs. Occasionally, hypercalcemia is exacerbated by concomitant acute renal failure that frequently complicates the course of the disease.
Pathophysiology of Bone Lesions in Myeloma
Although the precise molecular mechanisms mediating myeloma bone disease remain unclear, observations over the past 30 years have revealed:
- Osteolysis in myeloma is due to an increase in the number and activity of osteoclasts, the only cells known to have the capacity to resorb mineralized bone. This is the only cellular mechanism for bone destruction which is clearly evident in myeloma.
- Excessive osteoclast activity in myeloma almost always occurs adjacent to foci of myeloma tumor cells. In some patients, discrete lytic lesions occur adjacent to nests of myeloma cells. Myeloma cells may be spread more diffusely throughout the marrow, resulting in osteopenia. Thus, it appears that the predominant mechanism by which osteoclasts are stimulated in myeloma is a local one, whereby myeloma cells (or host cells) produce local factors (cytokines) responsible for increasing osteoclast formation and activation in an autocrine, paracrine or juxtacrine fashion.
- It has been known now for many years that cultures of human myeloma cells in vitro express and secrete several osteoclast activating factors (OAFs) into conditioned media. Earlier work had identified several soluble mediators with bone-resorptive activity including lymphotoxin (tumor necrosis factor-β), tumor necrosis factor-α, interleukin-1α/β and interleukin-6, all of which are capable of acting locally. Most of these putative mediators were identified in conditioned media from myeloma cells cultured independently. However, these in vitro studies have not clarified the nature of the critical mediator in vivo and our understanding of how and to what extent each of these soluble mediators influence bone resorption in myeloma remains limited. More recent information suggests that other putative factors with substantial bone resorbing activity are produced only when myeloma cells are in contact with marrow stromal cells. One such critical mediator identified recently is RANK Ligand (RANKL) in the myeloma bone microenvironment; RANKL is expressed by tumor cells and is overexpressed by osteoblasts and bone marrow stromal cells. Concomitantly, the production of its naturally occurring decoy receptor, OPG (osteoprotegrin) by bone marrow stromal cells is markedly down-regulated. The change in the RANKL/OPG ratio in favor of RANKL is responsible for the exaggerated osteoclast formation and activity seen in patients. RANK.Fc and OPG.Fc, synthetic chimeric antagonists of RANKL/RANK interactions, each block osteolytic lesions in preclinical mouse models of multiple myeloma. Another mediation that has received a lot of attention recently is MIP-1α. MIP-1α is overexpressed in bone marrow of myeloma patients compared with other hematological neoplasms and has been implicated in myeloma-induced osteolysis. Neutralizing anti-MIP-1α antibodies and MIP-1α antisense block osteolytic lesions and myeloma tumor progression in preclinical mouse models of multiple myeloma
- Hypercalcemia occurs in many patients with myeloma at some time during the course of the disease. Hypercalcemia is almost always associated with markedly increased bone resorption and frequently with impaired renal function that is fixed and due to direct effects of the disease on renal function. Glomerular filtration may be further compromised by concurrent volume depletion and hypercalcemia.
- The increase in osteoclastic bone resorption in myeloma is usually associated with impaired osteoblast function and the rate of new bone formation is often markedly reduced. In contrast to other types of osteolytic bone disease such as breast cancer, serum alkaline phosphatase activity is decreased or within the normal range, and radionuclide scans do not always show evidence of increased uptake.
- Recent evidence suggests that the frequency of osteolytic bone disease may be correlated with the pattern of infiltration of the marrow by myeloma cells, the highest frequency of lytic lesions being in seen in patients with nodular or diffuse, rather than insterstitial involvement.
- Rarely, patients with myeloma show an predominant increase in new bone formation with subsequent osteosclerosis. This is often associated with the POEM's Syndrome.
Experimental Animal Models of Human Myeloma Bone Disease: Opportunity to Study Myeloma Cells in their Normal Microenvironment
Until a decade ago, a major drawback to studying the mechanisms responsible for myeloma bone disease has been lack of suitable animal models of the human disease. Human myeloma cells do not home to the bone marrow in naïve nude mice even when injected systemically. This has meant that it has been difficult to establish an acceptable animal model of the human disease to study not only the pathogenetic mechanisms but also to determine the efficacy of various novel therapies. The importance of cell-cell interactions between myeloma cells and resident cells in the bone microenvironment in facilitating not only growth and survival of the tumor cells but also in promoting osteolysis, spurred a number of groups to develop experimental animal models of human myeloma bone disease that are useful for studying the mechanisms involved. These include models in which human myeloma cells are xenografted in irradiated or non-irradiated athymic (T-cell deficient) or SCID mice (B- and T-cell deficient). In addition to being immunodeficient, most of these models are further compromised by the requirement for irradiation to ensure tumor engraftment in medullary cavities, a procedure that almost certainly affects the bone microenvironment. To date, the only myeloma model utilizing immunocompetent, non-irradiated hosts is a murine syngeneic model originally characterized by Radl et al (1988), who described a myeloma bone disease which occurs spontaneously in aging mice of the C57BL/KaLwRij strain. These series of myelomas, designated 5T myelomas, developed at the rate of approximately 1 in every 200 aged mice, and caused a monoclonal gammopathy with features reminiscent of the human disease, including infiltration of the bone marrow by the myeloma cells and most importantly, characteristic myelomatous skeletal lesions. Osteolytic lesions were found in most tumor-bearing mice, and as in humans, very infrequently some mice developed osteosclerotic lesions. Either freshly dispersed myeloma cells from the bone marrow or involved spleens of myeloma-bearing mice can be serially transplanted by tail vein or intraperitoneal injections into fresh naïve recipients of the same strain, or by direct bone marrow inoculation with the disease being faithfully transmitted from mouse to mouse. In order to develop a more convenient animal model of the human myeloma bone disease, we and others have developed cell lines from one of these myelomas, the 5T33 myeloma, which can be studied both in vitro and in vivo. The cell line that we established, designated 5TGM1, which also reproducibly causes characteristic osteolytic bone lesions in young C57BL/KaLwRij mice when injected via the tail vein, produces the monoclonal IgG2b paraprotein and IL-6, hitherto but grows independent of exogenous IL-6 in vitro. Identical results are obtained when 5TGM1 cells are inoculated into immunodeficient (bg/nu/xid) mice. Some but not all mice carrying these myelomas become mildly hypercalcemic, again reminiscent of human myeloma bone disease. Importantly, histomorphometric analysis show that the osteolytic bone lesions in tumor-bearing mice are associated with an increase in osteoclast numbers and activity confirming the validating of this model for studies of pathogenetic mechanisms in myeloma and for preclinical evaluation of novel anti-tumor and anti-osteolytic strategies.
Hypercalcemia in Myeloma Bone Disease
Hypercalcemia in myeloma is due primarily to increased osteoclastic bone resorption caused by local cytokines released by the myeloma cells, which in turn leads to efflux of calcium into the extracellular fluid. This overwhelms the patient’s capacity to maintain normal calcium homeostasis resulting in elevated serum calcium levels. However, the pathogenesis of hypercalcemia in myeloma is probably more complex than this. Firstly, not all patients with significant myeloma bone disease develop hypercalcemia. Until the last 5 years, approximately one third of patients developed hypercalcemia, usually late in the course of the disease. This frequency is now decreasing with the advent of bisphosphonates as standard therapy for cancer-induced osteolysis. Hypercalcemia is most common in patients who have the largest tumor volume. The reasons for this are unclear, but may be related to the amount of bone-resorbing activity produced by the myeloma cells, as well as the status of glomerular filtration and renal function. Measurements of total body myeloma cell burden together with production of bone resorbing activity by cultured bone marrow myeloma cells in vitro do not correlate closely with hypercalcemia, although they do correlate somewhat with extent of osteolytic bone lesions. Thus, there are clearly other factors which are involved in the pathogenesis of hypercalcemia in addition to those that promote osteoclast formation and induce osteoclast activation. Probably the most important of these is the impairment of renal function which occurs frequently complicates the course of the disease in patients with myeloma. In addition to impaired glomerular filtration, increased renal tubular calcium reabsorption may also be a contributing factor to the pathophysiology of hypercalcemia and elevated serum PTH-rP levels may play a role in this regard in a subset of patients. In patients with hypercalcemia due to myeloma, there is almost always impaired renal function and an increase in serum phosphate that is associated with decreases in glomerular filtration rate. Markers of bone formation such as serum alkaline phosphatase are usually not increased in patients with myeloma, since bone formation is often not increased and in fact may be suppressed for reasons which are not entirely clear. Patients with hypercalcemia due to myeloma usually respond very rapidly to treatment with corticosteroids, unlike patients with humoral hypercalcemia due to solid tumors. However, efficacy of steroids in this setting may be due in part to their rapid suppression of myeloma tumor growth.
Bone Markers and Bone Mineral Density (BMD) for Monitoring Myeloma Bone Disease
Bone markers may eventually be useful for non-invasive monitoring of the response of cancer-induced osteolytic bone disease to therapy. Earlier markers for osteoclastic bone resorption include the measurement of deoxypyridinoline (Dpyr) crosslinks of collagen, but the usefulness of this and other markers in assessing and monitoring patients with myeloma remains to be unequivocally established. Dpyr crosslinks can be readily measured in the urine (u-Dpyr) by chemical assays or by ELISA, and sensitive as well as specific serum-based assays are becoming increasingly available. Dpyr measurement is much improved over previous markers such as urinary hydroxyproline, urinary pyridinoline or fasting urine calcium. Moreover, a number of studies have suggested that u-Dpyr is more sensitive in distinguishing between treated and untreated patients and may help not only in differentiating between multiple myeloma and MGUS, but may also serve as a marker of the underlying osteolytic bone disease activity. Other studies have suggested that serum levels of the carboxy-terminal telopeptide of type I collagen (ICTP), but not the carboxy-terminal propeptide of type I collagen (PICP), may be a sensitive prognostic marker of bone disease in myeloma. More information is needed with later generation assays, including measurement of N-terminal and C-terminal collagen telopeptides, and serum TRAP (tartrate-resistant acid phosphatase) measurements. Further studies are needed to show correlation between the levels of these markers, alone or in combination, and evidence of changes in the markers that correlate with bone disease progression in myeloma patients treated with or without bisphosphonates. It remains to be determined if these markers could facilitate determination of the efficacy of dose, determination of optimal mode of administration as well facilitate follow-up of bone turnover in patients with myeloma bone disease especially after specialized treatments such as autologous transplantation.
Parameters of bone formation such as serum alkaline phosphatase are not increased in patients with myeloma, unless the patient has an active fracture undergoing repair. In myeloma patients, measurements of osteocalcin (bone GLA protein) a marker of bone formation show a large scatter. Serum osteocalcin is usually decreased in myeloma patients with advanced disease and more extensive bone lesions consistent with the impaired bone formation. By contrast, serum osteocalcin levels may be normal or even increased earlier in the disease and in patients who have less aggressive or no obvious bone disease. More recently, bone mineral density has been shown to increase in myeloma patients in sustained remission after chemotherapy who have never been on bisphosphonates.
Management of Myeloma Bone Disease: Bisphosphonates and Other Potential Therapeutics
The major symptoms and the high morbidity and mortality rates associated with myeloma are due largely to the progressive bone destruction. Since patients may survive for many years post-diagnosis, clinicians have attempted to devise therapeutic approaches in myeloma that would relieve disabling symptoms, and in particular the severe bone pain, thereby improving quality of life. An early approach was the use of fluoride and later calcium and fluoride, although this combination was ineffective and, in fact, probably detrimental because of associated side effects. More recently, bisphosphonates, which inhibit osteoclastic bone resorption mainly by inducing osteoclast apoptosis, have become the standard anti-osteolytic therapy for myeloma-induced bone disease. Several groups have shown that the more potent second generation (Pamidronate; Aredia) and third generation (Zoledronic acid; Zometa) bisphosphonates not only relieve bone pain, they also produce a rapid, sustained, and significant decrease in the urinary excretion of calcium and hydroxyproline indicating decreased bone turnover. This was first shown with pamidronate (Van Breuklen et al, 1979), and then with clodronate (Siris et al, 1980). Pamidronate and zoledronic acid have since been approved by the FDA in the United States for myeloma patients with bone disease with or without hypercalcemia. These approvals were based primarily on studies evaluating efficacy and safety which showed that in several hundred patients with myeloma, there was a satisfactory response in bone pain, reduced need for radiation therapy, and reduction in vertebral and non-vertebral fractures (Berenson et al, 1995). These studies are consistent with a number of smaller studies in Europe and the much larger study sponsored by the Medical Research Council in the United Kingdom, in which clodronate was the bisphosphonate used (McCloskey et al, 1998). The more recent approval of the zoledronic acid third generation bisphosphonate zoledronic acid for use in bone disease was based in part on studies showing a marked reduction in skeletal-related events in myeloma patients (Berenson et al, 2001, Rosen et al, 2001). This agent is more potent than pamidronate, and this can be used by infusion over a shorter period of time (but not less than 15 minutes). Although the majority of patients with myeloma are now being treated with one of these agents above, it is important to recognize that not all bisphosphonates have been demonstrated to be effective in myeloma. There is no information that alendronate is useful in this disease or what the dose and timing of administration should be, yet there is clearly much off-label use of this compound. The newer bisphosphonate ibandronate has not been shown to be effective in clinical studies, possibly because the doses studied were not optimal (Menssen et al, 2002). Although there have been recent reports that bisphosphonates may be cytotoxic or cytostatic to tumor cells based on in vitro studies (Shipman et al, 1998), there has been little or no beneficial effects of these drugs on tumor burden in patients or in experimental animal models of myeloma. There are still a number of outstanding issues regarding the use of bisphosphonates in myeloma bone disease, including whether an efficacious orally available bisphosphonate can be developed, whether bisphosphonates should be given to patients early in the course of the disease, as well as in a preventative manner to patients before they have obvious bone disease or with MGUS, and most importantly which any of these drugs have a beneficial effect on survival. Research in this area is intensive and is likely to lead to the eventual introduction of new bisphosphonates or related drugs that are safe, can be given to patients orally and lead to powerful and beneficial effects on morbidity and survival as well as mortality.
In patients with myeloma, even when conventional chemotherapeutic regimens reduce tumor burden, the underlying bone disease may remain refractory to treatment. There is therefore a compelling need for the development of novel biologically based therapies that may concurrently improve both the bone lesions and reduce the tumor burden in myeloma. Two such agents that block RANKL effects are OPG.Fc. RANK.Fc and OPG.Fc both inhibit osteoclast activation, and may be more effective anti-osteolytic agents in myeloma because they also effectively prevent osteoclast formation. They have been shown to produce impressive beneficial effects in preclinical models. Importantly, emerging independent data from our group and others suggest that the two novel molecules have additional effects in myeloma in vivo, beyond those on inhibiting osteolysis, notably a significant reduction in tumor burden, assessed histomorphometrically and by monoclonal paraprotein titers, in different preclinical models of human bone myeloma disease. Although the mechanistic basis of this unanticipated reduction in myeloma tumor load remains to be elucidated, these findings potentially constitute a major advance as bisphosphonates have not been shown to date to exert an anti-tumor effect either in preclinical models or in myeloma patients. Although reduction in tumor load as a primary endpoint is not being examined, in on-going Phase I studies a single subcutaneous dose of OPG.Fc has been shown to be as effective as an intravenously administered bolus of pamidronate. It remains to be determined whether anti-tumor efficacy can be demonstrated with these novel molecules in Phase II and III clinical trials in myeloma patients.
References available upon request.