Search

Several animal models, as well as a number of cell culture experiments, indicate a prophylactic effect of bisphosphonates in respect of subsequent bone metastasis. Moreover, in preliminary clinical trials involving patients with advanced breast cancer and local or remote metastases, biophosphonates produced a reduction in new skeletal metastases. This overview summarizes and discusses the results of the latest investigations. It opens with a section on the pathophysiology of bone metastasis, which is followed by a report on animal models and first studies of bisphosphonate treatment as a new approach in systemic adjuvant therapy. © 2000 Cancer Research Campaign

There does not exist an entirely satsifactory drug for patients with established osteoporosis. No currently available drug is either universally acceptable or leads to substantial formation of new bone. Moreover, in spite of the recent introduction of newer bisphosphonates, intranasal calcitonin and estrogen-related compounds, there is still room for new preventive treatments for bone loss associated with the menopause and hormone withdrawal. There is an even greater need for acceptable and efficacious therapies for patients with established osteoporosis who have severe bone deficits. This is the reason that the pharmaceutical industry is now spending billions of dollars in a search for better modalities of treatment than those currently available. Specific problems can be found with all existing therapies, and drugs under development. The task that confronts the industry in identifying new therapies is particularly difficult. Although bringing a new drug to the market is very expensive in any pharmaceutical arena, it is particularly so in osteoporosis. This is because the Food and Drug Administration in the United States and many other regulatory agencies require evidence of beneficial effects on fracture rates. This problem is compounded somewhat by the expense involved in clinical studies in osteoporosis (requiring large numbers of patients from multiple centers), and the decline in research and development budgets associated with reduced profits due to managed care and price controls on drugs in some countries. Establishing practice guidelines for osteoporosis may also limit the use of drugs, particularly those that are expensive, or require treatment in many patients to see benefits in one. In the European Community, price controls on drugs already exist in many countries. A pharmaceutical company entering this field must consider the size of the therapeutic market, as well as the competition. For example, the bisphosphonate market is likely to be shared by a number of similar compounds during the next decade, and the same appears likely for estrogenrelated compounds. The lag time between discovery of a new lead for osteoporosis and the introduction of a new drug in the clinic may be 10–15 years. This certainly seems to have been the experience with resorption inhibitors such as alendronate and raloxifene. This means that there must be an enormous investment by the pharmaceutical company, possibly in the order of $500 million in today’s dollars, so that the eventual market must be substantial to accommodate this risk.

We have reported that P388D1 cell line murine macrophages stimulated with lipopolysaccharide (LPS) from Actinobacillus actinomycetemcomitans release interleukin-1 (IL-1) inhibitor. The IL-1 inhibitor was purified from conditioned media of P388D1 cells stimulated with A. actinomycetemcomitans LPS for 72 h to homogeneity by a four-step procedure: acetic acid extraction from conditioned media; Bio-Gel P-60 gel filtration chromatography; DEAE-Sepharose CL-6B column chromatography; and reverse-phase high-performance liquid chromatography on a C18 hydrophobic support. The purified IL-1 inhibitor gave a single band of protein with a molecular mass of 26 kDa on sodium dodecyl sulfate-polyacrylamide gel electrophoresis. The purified IL-1 inhibitor was a heat- and acid-stable protein that was inactivated by digestion with trypsin and reduction with dithiothreitol. This inhibitory factor suppressed the proliferation of C3H/HeJ mouse thymocytes and the proliferation of IL-1-dependent cell lines, D10.G4.1 and RPMI 1788, induced by IL-1. However, this inhibitor did not affect the proliferation of IL-2-dependent CTLL-2 cells induced by IL-2, the proliferation of C3H/HeJ mouse thymocytes stimulated with a mitogenic dose of concanavalin A, and the proliferation of IL-6-dependent B9 cells induced by IL-6. Furthermore, the IL-1 inhibitor significantly blocked stimulation of bone resorption in organ cultures of newborn mouse calvaria and inhibited the osteoclast-like cell formation in mouse marrow cultures. A monoclonal antibody prepared against the purified IL-1 inhibitor reacted with mouse recombinant IL-1 receptor antagonist (rIL-1ra), and a polyclonal antibody to mouse rIL-1ra reacted with the IL-1 inhibitor by Western blot (immunoblot) analysis. These results indicate that the IL-1 inhibitor is an identical molecule to rIL-1ra, suggesting that the IL-1 inhibitor (IL-1ra) released by macrophages stimulated with LPS from A. actinomycetemcomitans may play an important mediative role in the development of periodontal disease.

An ideal therapeutic agent for osteoporosis — one that both inhibits osteoclastic bone resorption and stimulates new bone formation — is not currently available. Currently available resorption inhibitors (estrogen, bisphosphonates, and calcitonin) prevent further bone loss and stabilize bone mass. However, they have a negligible to very modest effect on bone formation over prolonged periods, and do not cause prolonged increases in bone mass. A universally acceptable bone formation stimulator is not currently available. Future research is likely to focus on the use of peptide growth factors to stimulate appositional bone growth, or on the use of low molecular weight compounds that may selectively stimulate bone cells to produce these factors in the local bone remodeling microenvironment.

Bisphosphonate (BP), a specific inhibitor of osteoclasts, has been widely used as a beneficial agent for the treatment of bone metastases in patients with breast cancer. It is well recognized that BP reduces osteolysis by promoting apoptosis in osteoclasts. However, recent animal and human data suggest that BPs not only reduce osteolysis associated with metastatic breast cancer, but also decrease tumor burden in bone. The mechanisms by which tumor burden is decreased following BP administration are unknown. Here we examined the effects of the BP ibandronate on MDA-231 human breast cancer cells in bone metastases in a well-characterized animal model of bone metastasis. Ibandronate, which was administered (s.c. daily; 4 microg/mouse/day) after bone metastases were established, inhibited the progression of established osteolytic bone metastases as assessed by radiographic analysis. Histological and histomorphometrical examination revealed that ibandronate reduced osteoclastic bone resorption, with increased apoptosis in osteoclasts. Furthermore, ibandronate also significantly decreased the MDA-231 tumor burden, with increased apoptosis in MDA-231 breast cancer cells in bone metastases. In contrast, ibandronate failed to inhibit MDA-231 tumor formation with no effects on apoptosis in MDA-231 breast cancer cells in the orthotopic mammary fat pads. These data suggest that the effects of ibandronate on apoptosis in MDA-231 breast cancer cells are restricted in bone in which ibandronate selectively deposits. Consistent with these in vivo results, a relatively high concentration of ibandronate (100 microM) increased caspase-3 activity and induced DNA fragmentation in MDA-231 breast cancer cells in culture. Moreover, a caspase inhibitor, z-Val-Ala-Asp-fluoromethyl ketone, blocked ibandronate-induced DNA fragmentation in MDA-231 cells, suggesting an involvement of caspase-3 in ibandronate-induced apoptosis. Our results suggest that BP suppresses bone metastases through promotion of apoptosis in metastatic cancer cells as well as in osteoclasts. However, it still remains open whether BP has direct anticancer actions in vivo.

Myeloma is a unique hematologic malignancy that exclusively homes in the bone marrow and induces massive osteoclastic bone destruction presumably by producing cytokines that promote the differentiation of the hematopoietic progenitors to osteoclasts (osteoclastogenesis). It is recognized that neighboring bone marrow stromal cells influence the expression of the malignant phenotype in myeloma cells. This study examined the role of the interactions between myeloma cells and neighboring stromal cells in the production of osteoclastogenic factors to elucidate the mechanism underlying extensive osteoclastic bone destruction. A murine myeloma cell line 5TGM1, which causes severe osteolysis, expresses alpha(4)beta(1)-integrin and tightly adheres to the mouse marrow stromal cell line ST2, which expresses the vascular cell adhesion molecule-1 (VCAM-1), a ligand for alpha(4)beta(1)-integrin. Co-cultures of 5TGM1 with primary bone marrow cells generated tartrate-resistant acid phosphatase-positive multinucleated bone-resorbing osteoclasts. Co-cultures of 5TGM1 with ST2 showed increased production of bone-resorbing activity and neutralizing antibodies against VCAM-1 or alpha(4)beta(1)-integrin inhibited this. The 5TGM1 cells contacting recombinant VCAM-1 produced increased osteoclastogenic and bone-resorbing activity. The activity was not blocked by the neutralizing antibody to known osteoclastogenic cytokines including interleukin (IL)-1, IL-6, tumor necrosis factor, or parathyroid hormone-related peptide. These data suggest that myeloma cells are responsible for producing osteoclastogenic activity and that establishment of direct contact with marrow stromal cells via alpha(4)beta(1)-integrin/VCAM-1 increases the production of this activity by myeloma cells. They also suggest that the presence of stromal cells may provide a microenvironment that allows exclusive colonization of myeloma cells in the bone marrow. (Blood. 2000;96:1953-1960)

Kaposi's sarcoma-associated herpesvirus (KSHV, also known as HHV-8) has been found in patients with multiple myeloma (MM) and postulated to be aetiologically associated with the development of this common plasma cell malignancy. A murine model of MM was previously established in which intravenous transfer of 5T myeloma cells into C57BL/KaLwRij mice resulted in characteristic features of human MM. In the present study, we sought to identify herpesvirus DNA sequences in this murine model of MM through polymerase chain reaction (PCR) analysis using primers specific for KSHV, murine herpesvirus 68 (MHV68) and murine cytomegalovirus (MCMV) as well as consensus primers designed from the highly conserved DNA polymerase genes of the Herpesviridae family. None of the DNA samples from whole bone marrow (n = 6) or dendritic cells enriched by long-term culture (n = 8) of 5T myeloma-bearing mice as well as the 5T myeloma cell lines (n = 3) maintained in long-term culture yielded specific amplification products in any of the PCR assays. Two KSHV-specific serological assays measuring antibodies to KSHV latent and lytic antigens also failed to detect the presence of anti-KSHV antibodies in mice that developed MM. These results suggest that the development of 5T murine MM is unlikely to be involved with KSHV or a KSHV-like murine herpesvirus.

Calcium homeostasis in the extracellular fluid is tightly controlled and defended physiologically. Hypercalcemia always represents considerable underlying pathology and occurs when the hormonal control of calcium homeostasis is overwhelmed. The major hormones that are responsible for normal calcium homeostasis are parathyroid hormone and 1,25-dihydroxyvitamin D; these hormones control extracellular fluid calcium on a chronic basis. Over- or underproduction of these hormones or the tumor peptide, parathyroid hormone-related peptide, are the major causes of aberrant extracellular fluid calcium concentrations. These hormonal defense mechanisms are reviewed here.

Solid cancers metastasize to bone by a multistep process that involves interactions between tumor cells and normal host cells. Some tumors, most notably breast and prostate carcinomas, grow avidly in bone because the bone microenvironment provides a favorable soil. In the case of breast carcinoma, the final step in bone metastasis (namely bone destruction) is mediated by osteoclasts that are stimulated by local production of the tumor peptide parathyroid hormone-related peptide (PTH-rP), whereas prostate carcinomas stimulate osteoblasts to make new bone. Production of PTH-rP by breast carcinoma cells in bone is enhanced by growth factors produced as a consequence of normal bone remodeling, particularly activated transforming growth factor-beta (TGF-beta). Thus, a vicious cycle exists in bone between production by the tumor cells of mediators such as PTH-rP and subsequent production by bone of growth factors such as TGF-beta, which enhance PTH-rP production. The metastatic process can be interrupted either by neutralization of PTH-rP or by rendering the tumor cells unresponsive to TGF-beta, both of which can be accomplished experimentally. The osteoclast is another available site for therapeutic intervention in the bone metastatic process. Osteoclasts can be inhibited by drugs such as the new-generation bisphosphonates; as a consequence of this inhibition, there is a marked reduction in the skeletal events associated with metastatic cancer to bone, such as pain, fracture, and hypercalcemia. However and possibly even more importantly, there is also a reduction of tumor burden in bone. In experimental situations, this has clearly been shown to affect not only morbidity but also survival. The precise mechanism by which bisphosphonates inhibit osteoclasts is still unclear and may represent a combination of inhibition of osteoclast formation as well as increased apoptosis in mature osteoclasts. However, studies with potent bisphosphonates such as ibandronate, pamidronate, and risedronate have clearly documented that reduction of bone turnover and osteoclast activity leads to beneficial effects not only on skeletal complications associated with metastatic cancer, but also on tumor burden in bone.

Humoral hypercalcemia of malignancy results from the effects of tumor-produced factors on bone, kidney, and intestine that disrupt normal calcium homeostasis. Although parathyroid hormone-related protein (PTHrP) is a major mediator of the syndrome, tumors also produce other hypercalcemic factors, such as tumor necrosis factor (TNF), which may modulate the effects of PTHrP. It has been postulated that TNF may counteract the stimulatory effects of PTHrP on bone formation. To examine the effects of TNF on PTHrP-induced changes in calcium and bone metabolism, a murine tumor model of hypercalcemia was used. Nude mice were inoculated with Chinese hamster ovarian (CHO) cells expressing human TNF (CHO/TNF) or nontransfected CHO cells (CHO/-) and further treated with injections of human PTHrP(1-34) or vehicle. The effects of TNF, PTHrP, and the combination of the two factors on blood ionized calcium, osteoclast recruitment, and bone histomorphometry were evaluated. Mice bearing CHO/TNF tumors that were injected with PTHrP had significantly higher calcium concentrations, increased committed osteoclast progenitors, and mature osteoclasts as well as enhanced bone resorption compared with mice bearing CHO/TNF tumors injected with vehicle or those bearing CHO/- tumors injected with PTHrP or vehicle. A 2-fold increase in new woven bone formed in the calvaria at sites of previous bone resorption was observed in CHO/TNF mice treated with PTHrP. Bone formation rates in the vertebrae were similar in both CHO/- and CHO/TNF mice treated with PTHrP. These data demonstrate that the hypercalcemic effects of PTHrP are enhanced by TNF and that this effect is due to the increased production of committed osteoclast precursors with a subsequent increase in osteoclastic bone resorption. Furthermore, PTHrP caused a coupled increase in osteoclastic bone resorption and new bone formation that was not inhibited by TNF. These findings highlight the complex interactions that may occur between tumor-produced factors on bone that result in malignancy-associated hypercalcemia and suggest that TNF may not be responsible for the decreased bone formation seen in some patients with this condition.

The molecular mechanisms by which human cancer cells spread to bone are largely unexplored. The process likely involves cell adhesion molecules (CAMs) that are responsible for homophilic and heterophilic cell-cell interactions. One relevant CAM may be the calcium-dependent transmembrane glycoprotein E-cadherin. To investigate the involvement of E-cadherin in breast cancer metastasis to bone, we used an in vivo model in which osteolytic bone metastases preferentially occur after injections of cancer cells directly into the arterial circulation through the left ventricle of the hearts of nude mice. We have found that E-cadherin-negative human breast cancer cells MDA-MB-231 (MDA-231) develop radiographically detectable multiple osteolytic bone metastases and cachexia in this model. However, MDA-231 breast cancer cells that were transfected with E-cadherin cDNA showed a dramatically impaired capacity to form osteolytic metastases and induce cachexia. Histological and histomorphometrical analyses of bones of mice bearing mock-transfected MDA-231 revealed aggressive metastatic tumor, whereas metastatic tumor burden was significantly decreased in the bones of mice bearing E-cadherin-expressing MDA-231. Nude mice bearing E-cadherin-transfected MDA-231 breast cancer cells survived longer than mice bearing mock-transfected MDA-231 breast cancer cells. Anchorage-dependent and -independent growth in culture and tumor enlargement in the mammary fat pad of nude mice were unchanged between mock-transfected and E-cadherin-expressing MDA-231, suggesting that these differences in metastatic behavior are not due to an impairment of cell growth and tumor-igenicity. Our results show the suppressive effects of E-cadherin expression on bone metastasis by circulating breast cancer cells and suggest that the modulation of expression of this CAM may reduce the destructive effects of breast cancer cells on bone.

Osteolytic bone destruction and its complications, bone pain, pathologic fractures, and hypercalcemia, are a major source of morbidity and mortality in patients with multiple myeloma. The bone destruction in multiple myeloma is due to increased osteoclast (OCL) activity and decreased bone formation in areas of bone adjacent to myeloma cells. The mechanisms underlying osteolysis in multiple myeloma in vivo are unclear. We used a human plasma cell leukemia cell line, ARH-77, that has disseminated growth in mice with severe combined immunodeficiency (SCID) and expresses IgG kappa, as a model for human multiple myeloma, SCID mice were irradiated with 400 rads and mice were injected either with 10(6) ARH-77 cells intravenously (ARH-77 mice) or vehicle 24 hours after irradiation. Development of bone disease was assessed by blood ionized calcium levels, x-rays, and histology. All ARH-77, but none of control mice that survived irradiation, developed hind limb paralysis 28 to 35 days after injection and developed hypercalcemia (1.35 to 1.46 mmol/L) a mean of 5 days after becoming paraplegic. Lytic bone lesions were detected using x-rays in all the hypercalcemic mice examined. No lytic lesions or hypercalcemia developed in the controls. Controls or ARH-77 mice, after developing hypercalcemia, were then killed and bone marrow plasma from the long bones were obtained, concentrated, and assayed for bone-resorbing activity. Bone marrow plasma from ARH-77 mice induced significant bone resorption in the fetal rat long bone resorption assay when compared with controls (percentage of total 45Ca released = 35% +/- 4% v 11% +/- 1%). Histologic examination of tissues from the ARH-77 mice showed infiltration of myeloma cells in the liver and spleen and marked infiltration in vertebrae and long bones, with loss of bony trabeculae and increased OCL numbers. Interestingly, cultures of ARH-77 mouse bone marrow for early OCL precursors (colony-forming unit-granulocyte-macrophage [CFU-GM]) showed a threefold increase in CFU-GM from ARH-77 marrow versus controls (185 +/- 32 v 40 +/- 3 per 2 x 10(5) cell plated). Bone-resorbing human and murine cytokines such as interleukin-6 (IL-6), IL-1 alpha or beta, TGF-alpha, lymphotoxin, and TNF alpha were not significantly increased in ARH-77 mouse sera or marrow plasma, compared with control mice, although ARH-77 cells produce IL-6 and lymphotoxin in vitro. Conditioned media from ARH-77 cells induced significant bone resorption in the fetal rat long bone resorption assay when compared with untreated media (percentage of total 45Ca released = 22% +/- 2% v 11% +/- 1%). This effect was not blocked by anti-IL-6 or antilymphotoxin (percentage of total 45Ca released = 19% +/- 1% and 22% +/- 1%, respectively). Thus, we have developed a model of human multiple myeloma bone disease that should be very useful to dissect the pathogenesis of the bone destruction in multiple myeloma.

Tumors occasionally stimulate bone formation and cause osteoblastic metastases. Although this occurs most frequently in widespread prostate cancer, human prostate cancer cells are difficult to grow in culture without changing their phenotype, and the few available prostatic cancer lines do not increase bone formation in vivo. To identify tumor-derived osteoblast-stimulatory factors, we studied a long-established human tumor cell line derived from human amnion that has, in the past, been reported to cause bone formation in vivo when inoculated into nude mice. Tumor cells were inoculated into nude mice and induced extensive new bone formation. To characterize osteoblast growth factors produced by these tumor cells, solid tumor was isolated from the mice and extracted at neutral pH. Biological activity, assessed by stimulation of proliferation of MG-63 osteoblast-like cells, was used to monitor purification after heparin-Sepharose column chromatography, Mono-S, and C4 reverse-phase high-performance liquid chromatography. An extend amino-terminal form of basic fibroblast growth factor (FGF) was purified by its capacity to stimulate proliferation in MG-63 cells and partially sequenced. Basic FGF is also known to stimulate proliferation in MG-63 cells and other osteoblasts in vitro and bone formation in vivo. In summary, these human tumor cells stimulate new bone formation in vivo and produce an osteoblast stimulating activity in vitro, which has been identified as a form of basic FGF.

Bone morphogenetic protein-4 (BMP-4) is one of a member of related polypeptides that are important in bone formation and other developmental processes. We isolated the BMP-4 gene from a mouse genomic library and characterized the exon-intron structure and one of the candidate promoters. Two alternative 5'-noncoding exons, 1A and 1B, were identified by reverse transcription polymerase chain reaction assays. Quantitative competitive polymerase chain reaction using Exon 1A, Exon 1B, and Exon 3 primers indicate the 1A-containing transcript is the primary BMP-4 mRNA expressed in bone cell cultures. Primer extension analysis supports that 1A is the major promoter utilized in bone cell cultures as well as in 9.5-day mouse embryos. 1A promoter activity indicate selective DNA regions functional in bone cells. We found potential regulatory response regions in the 1A 5'-flanking region of the BMP-4 gene for the chicken ovalbumin upstream-transcription factor I (COUP-TFI). Specific binding to the COUP-TFI response regions in the BMP-4 1A promoter was demonstrated. By co-transfection of a COUP-TFI expression plasmid with the BMP-4 1A promoter in fetal rat calvarial osteoblasts, we demonstrated that COUP-TFI inhibits the BMP-4 promoter activity. This suggests that COUP-TFI could act as a silencer for BMP-4 transcription in vivo.

Human breast cancer frequently metastasizes to the skeleton to cause osteolysis and subsequent pain, pathological fracture, and hypercalcemia. Because bone continuously releases growth factors stored in bone matrix by bone resorption during physiological remodeling and, thus, possibly provides a favorable microenvironment for metastatic breast cancer cells to proliferate, inhibitors of bone resorption used either prophylactically or in patients with established disease, therefore, would seem likely to be useful adjuvant therapy in patients with breast cancer. However, the parameters for monitoring progressive osteolytic bone disease in humans are imprecise. We examined the effects of the third generation bisphosphonate, risedronate, which is a specific inhibitor of osteoclastic bone resorption, in a bone metastasis model in nude mice in which intracardiac injection of the human breast cancer cell line MDA-231 leads to osteolytic bone metastases. Risedronate (4 micrograms/animal/day) was given s.c. to animals (a) after radiologically small but defined osteolytic metastases were observed; (b) simultaneously with MDA-231 cell inoculation through the entire experimental period; or (c) by short-term prophylactic administration before inoculation of MDA-231 cells. In all experiments, risedronate either slowed progression or inhibited the development of bone metastases assessed radiographically. Furthermore, mice treated continuously with risedronate showed significantly longer survival than did control mice. Histomorphometrical analysis revealed that osteoclast numbers were diminished at metastatic tumor sites. Unexpectedly, there was also a marked decrease in tumor burden in bone in risedronate-treated animals. In contrast, the growth of metastatic breast cancer in soft tissues surrounding bones was not affected by risedronate. Moreover, risedronate had no effects on the local growth of s.c. implanted MDA-231 breast cancers in nude mice or on MDA-231 cell proliferation in culture. These data demonstrate that risedronate decreases metastatic MDA-231 breast cancer burden selectively in bone, as well as suppresses progression of established osteolytic lesions and prevents the development of new osteolytic lesions; thus, the data suggest that inhibition of osteoclastic bone resorption may be a useful adjunctive therapy for the treatment of cancers that have colonized in bone.

Osteolytic bone destruction and its complications, such as hypercalcemia, pathologic fractures and nerve compression, are the major source of morbidity in patients with multiple myeloma (MM). The bone destruction in MM is due to increased osteoclast activity, but the mechanisms responsible are not entirely clear. We have utilized a human plasma cell leukemia cell line, ARH-77, that has disseminated growth in mice with severe combined immunodeficiency (SCID) and expresses immunoglobulin G kappa (IgG kappa), as a model for human MM. Fifteen SCID mice were irradiated with 400R and 10 of these were injected with 10(6) ARH-77 cells i.v., 24 h after irradiation. Five mice were used as a control group. Development of bone disease was assessed by blood calcium levels, x-rays and histology. Seven out of seven mice that survived irradiation and received ARH-77 cells developed hind limb paralysis 28-35 days after injection. One hundred percent of these mice developed hypercalcemia (1.35-1.46 mmol/l), a mean of five days after becoming paraplegic. Lytic bone lesions were detected by x-ray in all the hypercalcemic mice examined. No lytic lesions or hypercalcemic developed in the controls. Mice were then sacrificed after developing hypercalcemia. Histologic examination of the ARH-77 mice showed infiltration of myeloma cells in the liver and spleen. Marked infiltration by the tumor was found in vertebrae and long bones, with loss of bony trabeculae and increased osteoclast numbers.(ABSTRACT TRUNCATED AT 250 WORDS)

Although bone formation traditionally has been difficult to study in vitro, recent improvements in cell culture techniques have made it possible to study primary cultures of fetal rat calvarial osteoblasts, and to examine specific gene events associated with osteoblast proliferation, differentiation, and mineralization of extracellular matrix to form bone nodules. Results of studies show that extracellular matrix proteins such as Type I collagen and osteocalcin are expressed. Also expressed are growth regulatory factors such as the bone morphogenetic proteins, which are presumably secreted and stored in the bone matrix or involved in additional osteoblast differentiation in an autocrine/paracrine role. The bone matrix is 1 source of these growth regulatory factors for bone formation; another is those animal and human tumors associated with osteoblastic metastases. In some of these tumors, an extended form of basic fibroblast growth factor has been identified, and various bone morphogenetic proteins have been found in others.

Most patients with cancer die because the cancer metastasizes to other organs. Bone is 1 of the 3 most common sites of tumor metastasis, and is a catastrophic complication of cancer, because it is associated not only with intractable pain, but also with other morbidity that usually hastens an uncomfortable death. Despite its importance, to date no specific therapies to prevent this complication of cancer or treat it once it occurs have been demonstrated. In this article, current understanding of the processes involved in the pathophysiology of bone metastasis will be reviewed, and approaches aimed at preventing the development of a bone metastasis and treating bone metastasis once it has occurred will be discussed.

All bone is remolded by bone resorbing (osteoclasts) and bone forming (osteoblasts) cells that are under the regulatory control of local factors generated in the bone cell microenvironment. These local factors are comprised of peptides and nonpeptides, including arachidonic acid metabolites such as prostaglandins of the E series and the leukotrines. Most is known of the peptides the stimulate or inhibit osteoclastic bone resorption, but information is also now accumulating on the peptide growth factors that are incorporated into the bone matrix, released in an active form when bone is resorbed, and stimulate those cellular events that result in bone formation. In this article these control mechanisms are reviewed.

Porcine pancreatic extracts (PXs) have previously been shown to decrease blood ionized calcium in BALB/c mice (T. Yoneda, Y. Takaoka, and G. R. Mundy. FEBS Lett., 278: 171-174, 1991). In the present study, we show that the PX is effective in preventing progression of hypercalcemia and decreasing osteoclastic bone resorption associated with a human squamous carcinoma in nude mice. PX inhibited osteoclast-like cell formation in mouse bone marrow cultures and bone resorption in organ cultures of fetal rat long bones which had been stimulated by serum-free culture supernatants of this cancer. In addition, PX increased food intake, decreased weight loss, and prevented development of cachexia. In parallel with these effects, PX prolonged survival of tumor-bearing animals. PX might have therapeutic potential for management of hypercalcemia and cachexia associated with malignancy.

We have previously shown that bone organ cultures produce large amounts of latent transforming growth factor beta (TGF beta), which lacks latent TGF beta-binding protein (LTBP). In this study we used the known osteoblast-like cell lines UMR-106, ROS 17/2.8, and MG63 as models to further examine latent TGF beta expression in bone. We found that the osteosarcoma cell line UMR-106 secreted latent TGF beta almost exclusively as a 100-kDa complex lacking LTBP. ROS 17/2.8 cells produced both the 100-kDa complex and also a 290-kDa complex containing the fibroblastic (190 kDa) form of LTBP. MG63 cells (like human foreskin fibroblasts) expressed almost exclusively the 290-kDa complex. To investigate the regulation of latent TGF beta complexes in bone cells we assessed the effects of TGF beta 1 treatment on expression of active and latent TGF beta. TGF beta 1 induced secretion of latent but not active TGF beta in all cell types examined. In human foreskin fibroblast cells, TGF beta 1 and LTBP mRNA were expressed concomitantly. In contrast, in osteosarcoma cell lines autoinduction of TGF beta 1 mRNA was associated with either a delayed increase or no change in LTBP mRNA. In UMR-106 cells LTBP message was virtually undetectable. We postulate that the expression of different latent TGF beta forms by osteoblast-like cells may reflect their maturation states and that different latent TGF beta complexes may have different functions, for example as secretory forms or as matrix storage forms.

Bone resorption requires cooperation between osteoclasts and mononuclear accessory cells by mechanisms which have not been elucidated. Since multinucleated cells in giant cell tumors of bone have many phenotypic and functional characteristics of normal osteoclasts, we have examined the interaction between the bone-resorbing multinucleated cells and the distinct mononuclear stromal cells from these tumors. We have found that these mononuclear cells produce an activity which stimulates both giant cells from giant cell tumors and rodent osteoclasts to resorb bone in vitro. We have identified the activity and found that it represents several products of the 5-lipoxygenase pathway of arachidonic acid metabolism, namely 5-hydroxyeicosatetraenoic acid and the leukotrienes. These data indicate that 5-lipoxygenase metabolites stimulate isolated osteoclasts to resorb bone in vitro and may represent a mechanism by which mononuclear stromal cells in human giant cell tumors communicate with the giant cells. In addition, these results may explain a possible mechanism for communication between accessory cells and osteoclasts involved in normal bone resorption.

Bone resorbing cytokines may be associated with abnormalities in bone cell function and calcium homeostasis in a number of pathological conditions. One of these cytokines is interleukin 6 (IL-6), which is a multifunctional cytokine which has been shown to be associated with increased formation of bone resorbing osteoclasts in vitro. In this report, we demonstrate that neutralizing antibodies to human IL-6 lower the blood calcium in nude mice carrying a human tumor associated with increased IL-6 production. There was no effect on blood calcium in hypercalcemic nude mice carrying tumors not associated with increased IL-6 production and normocalcemic tumor-bearing nude mice. These results suggest that increased production of IL-6 may contribute to disturbances in calcium homeostasis in some malignancies, and suggest that neutralization of IL-6 effects can lower the serum calcium in these tumors.

Tumors frequently affect the skeleton and alter the function of normal bone cells. In most patients, they increase the activity and probably the formation of bone-resorbing osteoclasts. The result is a destructive or osteolytic bone lesion. Occasionally they lead to an increase in activation of osteoblasts or bone-forming cells. In a relatively small number of patients, this increase in osteoblastic activity is profound and the result is an area of newly formed bone around the tumor cell deposit which can be detected radiologically and is referred to as an osteoblastic or sclerotic metastasis. In this chapter, we plan to review what is known of the mechanisms responsible for development of osteolytic and osteoblastic bone lesions.

The activity of bone cells is controlled by circulating systemic factors and by locally produced factors or cytokines. This distinction is not absolute, since some cytokines may circulate when produced in large amounts. The production of the three major systemic factors (parathyroid hormone, calcitonin, and 1, 25-dihydroxyvitamin D) is regulated by extracellular fluid calcium, and their primary role in maintenance of calcium homeostasis rather than control of bone cell function. Nevertheless, to maintain calcium homeostasis, bone cell function may be altered. In this chapter, the effects of these systemic factors on bone cell activity will be reviewed.

In recent years, it has become increasingly obvious that the cellular events involved in bone remodeling are likely modulated by a group of local factors. These local factors, or osteotropic cytokines, have extremely potent effects on bone cells in both in vitro and in vivo systems.

Parathyroid hormone-related protein (PTH-rP) has been implicated as a causative factor in the humoral hypercalcemia of malignancy. Animal models of hypercalcemia of malignancy that have traditionally utilized human or animal tumors or injections or infusions of hypercalcemic factors have limitations that may prevent exact delineation of the biologic effects of tumor-produced PTH-rP.To assess the effects of tumor-produced PTH-rP in vivo, we have transfected Chinese hamster ovarian (CHO) cells with cDNA encoding human preproPTH-rP-(1-141) which then stably express only PTH-rP. We inoculated these tumor cells into nude mice, and compared tumor-bearing nude mice with similar tumor-bearing nude mice inoculated with non-transfected CHO cells using standard parameters of calcium homeostasis.The nude mice carrying tumors expressing PTH-rP became hypercalcemic over 12 to 18 days. Blood ionized calcium values correlated positively with plasma PTH-rP concentration and tumor volume. Plasma PTH-rP concentrations in hypercalcemic mice were similar to those reported in humans with hypercalcemia of malignancy. Bone histology and histomorphometry from hypercalcemic nude mice demonstrated increased bone resorption when compared with mice bearing nontransfected CHO tumors.We have produced an animal model of tumor-produced PTH-rP by transfecting CHO cells with the cDNA for PTH-rP and inoculating these tumor cells into nude mice. Hypercalcemia in this model is mediated in part by the effects of PTH-rP to increase osteoclastic bone resorption. The model is advantageous because PTH-rP alone is secreted in a prolonged, constitutive fashion with pharmacokinetics similar to naturally occurring tumors. It should prove to be a useful model to study the effects of tumor-produced PTH-rP in vivo.

The mechanisms by which tumor cells metastasize to bone are not well understood. We have investigated the role of the basement membrane glycoprotein, laminin, in bone metastasis, since antagonists to laminin have been shown to inhibit the formation of lung metastases. We studied the formation of osteolytic metastases caused by a human tumor which is known to cause osteolysis and hypercalcemia in nude mice. We found that tumor-bearing nude mice developed hypercalcemia, cachexia, and characteristic osteolytic lesions throughout the skeleton after injection of this human melanoma cell line (A375) into the left ventricle. When we gave injections to nude mice with A375 cells which had been exposed to C(YIGSR)3-NH2, a laminin-derived synthetic peptide containing three linear sequences of YIGSR with an amino-terminal cysteine which competes with laminin for its receptor, we found a decrease in the formation of detectable osteolytic bone metastases. The tumor cells were incubated with the antagonist and then inoculated into nude mice which were administered the antagonist i.p. Hypercalcemia and cachexia were also decreased in tumor-bearing mice treated with the laminin antagonist. In contrast, laminin itself increased the number of osteolytic bone metastases, as has been shown for other tumor cells. These data suggest that laminin plays a role in the formation of osteolytic bone metastases in this model and that laminin antagonists may be useful in the prevention of bone metastases in some human tumors.

Human breast cancer cell proliferation is regulated by growth factors that bind to receptors with intrinsic tyrosine kinase (TK) activity, including the epidermal growth factor (EGF) receptor. To determine whether inhibition of receptor TK activity inhibits tumor growth, we studied the effects of a tyrosine kinase inhibitor, RG-13022, on cultured human breast cancer cells. RG-13022 represents a class of compounds which have been shown to inhibit preferentially the TK activity of the EGF receptor in a cell-free system and also to inhibit EGF-stimulated growth of cultured cells. RG-13022 significantly inhibited EGF-stimulated autophosphorylation of its receptor in two breast cancer cell lines that have abundant, although not amplified, EGF receptor content (MDA-231 and T47D). RG-13022 also inhibited EGF-stimulated DNA synthesis and proliferation of T47D and MCF-7 breast cancer cells in a reversible and dose-dependent manner. Inhibition was observed at 0.1 microM, and it was maximal at 10 microM. The effect was rapid (within 3 h), persisted for 18 h, and was partially reversed by 24 h at 1 microM. At 5 microM, inhibition persisted for more than 50 h. Inhibitory effects were also observed in a panel of estrogen receptor-positive and estrogen receptor-negative breast cancer cell lines. RG-13022 inhibited not only EGF-induced growth but also growth stimulated by insulin, insulin-like growth factor I, insulin-like growth factor II, or transforming growth factor alpha. RG-13022 also totally blocked estrogen-stimulated phosphorylation of the EGF receptor, as well as estrogen-induced cell proliferation, suggesting that functioning TK pathways are required for estrogen action. The TK inhibitor RG-13022 is a potent inhibitor of hormonally regulated growth of human breast cancer. Tyrosine kinase inhibitors have the potential of providing a new strategy for the "endocrine therapy" of breast cancer.

Tumors are frequently associated with factors that affect bone cell function and calcium homeostasis. As a result, many patients with cancer develop paraneoplastic syndromes. The commonest is hypercalcemia of malignancy. Other related paraneoplastic syndromes are hypocalcemia, hypophosphatemia, and osteomalacia.

Many human tumors of epithelial origin contain cells overexpressing the epidermal growth factor (EGF) receptor, and there is convincing evidence that cancer cell growth is correlated with the loss of the normal regulation of the EGF receptor signal transduction pathway. Some cancers are clearly dependent on activation of the EGF receptor for their proliferation. Recently, a class of compounds, tyrphostins, which inhibit the protein tyrosine kinase activity of the growth factor receptor, have been described. In this report, we have examined the antiproliferative effects of potent new tyrphostins on a well-characterized human squamous cell carcinoma in vitro and in vivo. We found that two of these compounds (RG-13022 and RG-14620) suppressed not only EGF-stimulated cancer cell proliferation in vitro but also tumor growth in nude mice. RG-13022 also increased the life span of these tumor-bearing nude mice. When administered to tumor-bearing nude mice together with monoclonal antibodies to the EGF receptor at a suboptimal dose which had no effect alone, inhibition of tumor growth was markedly enhanced. These data suggest that tyrphostins have potential as anticancer agents.

Increased levels of epidermal growth factor receptor (EGFR) have been shown on squamous cell carcinomas. Recently, we described a squamous cell carcinoma (MH-85) derived from the oral cavity which was associated with several paraneoplastic syndromes including hypercalcemia and cachexia. This tumor induced the same paraneoplastic syndromes in nude mice (BALB/c, nu/nu, male, 4-6 weeks old). Scatchard analysis revealed that there are two classes of EGFR in MH-85. The dissociation constant and number of binding sites for the high affinity receptors were 38 pM and 5 x 10(4)/cell, respectively, and 2.2 nM and 6 x 10(5) cell, respectively, for the low affinity receptors. Growth of MH-85 in culture was stimulated by epidermal growth factor (EGF) and inhibited by monoclonal antibody 108 to human EGFR, which recognizes the extracellular domain of the EGF receptor. Surgical removal of submandibular glands from male nude mice resulted in a dramatic decrease in plasma EGF levels and a significant reduction of tumor growth, hypercalcemia, and cachexia. When EGF (5 micrograms/mouse, every 2 days for 6 weeks, i.p.) was administered to these sialoadenectomized animals, tumor growth increased, with a parallel increase in hypercalcemia. When monoclonal antibody 108 (1 mg/mouse, i.p.) was given 1, 5, and 10 days after MH-85 tumor implantation, tumor formation was retarded, which resulted in delayed onset of hypercalcemia and cachexia. Moreover, when the antibody was injected 6 times in nude mice exhibiting large tumors and profound hypercalcemia and cachexia, there was a striking decrease in tumor growth, which was accompanied with a reversal of hypercalcemia and cachexia. These results indicate that growth of the human squamous cell carcinoma MH-85 is dependent on the EGFR pathway and that subsequent development of hypercalcemia and cachexia is dependent on tumor growth. They also suggest that agents which interfere with the EGFR pathway may have therapeutic potential as anticancer agents in some human tumors.

The integrity of the skeleton is maintained by the continued cellular remodelling of bone that occurs throughout life and is characterized by an orderly sequence of events beginning with osteoclastic bone resorption followed by osteoblastic new bone formation to repair the localized defects made by osteoclasts. Bone resorption and formation are closely coupled by mechanisms which, until recently, have been poorly defined in both normal physiological and most pathological conditions, but are probably due to the generation during resorption of local 'coupling' factors. TGF-beta promises to be one of the key factors involved in coupling bone formation to previous bone resorption. This potent osteotropic polypeptide is abundant in the bone matrix, and is produced in response to factors that stimulate osteoclastic bone resorption. It is a very potent stimulator of osteoblastic bone formation, causing chemotaxis, proliferation and differentiation in committed osteoblasts. TGF-beta has complex effects on bone resorption: it inhibits osteoclast formation and osteoclast activity. TGF-beta is released from bone in a biologically inert state that is due to the presence of at least two proteins which appear to regulate TGF-beta activity. Release of active TGF-beta from these latent complexes occurs during bone resorption and is mediated by osteoclasts. Knowledge of the mechanisms responsible for these activation processes may be vital to understanding the role of TGF-beta in bone remodelling.

The rat Leydig cell tumor is a well characterized model of the humoral hypercalcemia of malignancy. The studies reported here were provoked by the observation that tumor-bearing rats become extremely cachectic and develop hypertriglyceridemia as they become hypercalcemic. Since the bone resorbing cytokine tumor necrosis factor (TNF)/cachectin is associated with cachexia and hypertriglyceridemia, we examined hypercalcemic tumor-bearing rats for evidence of increased TNF production using a TNF radioimmunoassay. We found that immunoreactive TNF was increased in the plasma of tumor-bearing rats. The increase in plasma TNF was comparable to that previously shown in hypercalcemic nude mice bearing Chinese hamster ovarian cell tumors transfected with the human TNF gene. There was no detectable TNF activity in tumor culture media which suggested that the tumor itself was not the source of excess TNF production. However, we found that tumor cell conditioned media enhanced the production of TNF activity by normal macrophages in vitro, indicating that increased TNF production in vivo may result from a tumor factor(s) which stimulates TNF production by normal immune cells. When TNF was added together with tumor products to organ cultures of fetal rat long bones, osteoclastic bone resorption was potentiated. These data are consistent with the concept that in this model of the humoral hypercalcemia of malignancy, increased TNF production by normal immune cells is increased, has systemic effects as suggested by cachexia and hypertriglyceridemia, and may work in concert with factors produced directly by tumor cells to overwhelm normal calcium homeostasis.

Demineralized bone matrix contains a number of growth factors for osteoblast-like cells. Two of these, the novel glycoprotein osteoinductive factor (OIF) and transforming growth factor-beta (TGF beta), act together to cause ectopic bone formation in vivo. Since OIF, like TGF beta, is likely released from bone when the matrix is resorbed, we examined the effects of homogeneous OIF and TGF beta on osteoclast function. Osteoclast function was tested in isolated avian osteoclasts and was measured in terms of tartrate-resistant acid phosphatase (TRAP) activity, oxygen-derived free radical production, and formation of characteristic resorption lacunae on slices of sperm whale dentine. OIF (50-100 ng/ml) inhibited the capacity of these osteoclasts to form lacunae whether assessed by the number of excavations per slice or by the total area resorbed. OIF (10-100 ng/ml) or TGF beta (10-20 ng/ml) caused a decrease in TRAP activity as well as a reduction in oxygen-derived free radical generation detected by nitroblue tetrazolium staining. TGF beta had no effect on the resorption capacity of isolated osteoclasts in concentrations that inhibited TRAP activity and nitroblue tetrazolium staining. These results suggest that growth regulatory factors, such as OIF and TGF beta, released during the resorption of bone may be endogenous inhibitors of continued osteoclastic activity. This cessation of osteoclast activity may be an essential preliminary step to the new bone formation that occurs at resorption sites during bone remodeling.

It has long been known that complex interactions occur between tumors and normal host immune cells. The human melanoma cell line A375 has been used previously as an indicator cell for tumor cell cytotoxicity mediated by monocytes. During other studies on this tumor cell line, we noted that the conditioned media harvested from A375 cultures induced both the human monocytoid cell line U937 and human blood monocytes to release the cytokine tumor necrosis factor (TNF). We characterized this tumor factor which induced TNF release by monocytic cells. Purification was performed using ammonium sulfate precipitation, ion exchange (DEAE) chromatography, gel filtration, and reversed-phase high performance liquid chromatography. The factor copurified with granulocyte-macrophage colony-stimulating factor (GM-CSF). The purified material caused the release of TNF by U937 cells and stimulated formation of granulocyte-macrophage colonies in methyl cellulose. TNF release by U937 cells in response to A375-conditioned medium was inhibited by neutralizing antibodies to GM-CSF. The TNF-inducing activity in A375-conditioned medium was completely removed by an anti-GM-CSF affinity column. Western blotting using antibodies to GM-CSF confirmed a single Mr27,000 band in A375-conditioned medium. We found that recombinant human GM-CSF stimulated TNF production by the same cells as the tumor-conditioned medium. These data show that A375 human melanoma cells produce GM-CSF, which in turn causes TNF production by cells in the monocyte lineage. The combination of GM-CSF production by the tumor and TNF production by immune cells may influence not only tumor growth but also some of the paraneoplastic syndromes associated with malignancy such as hypercalcemia, cachexia and leukocytosis.

Osteoinductive factor (OIF) is a glycoprotein in bone that induces ectopic bone formation. Implantation of OIF plus transforming growth factor beta (TGF-beta) type 1 or 2 into subcutaneous tissues of rats induces formation of bone at the implantation site. Since TGF-beta is also present in bone matrix and inhibits formation of multinucleated cells that express an osteoclast phenotype in long-term human marrow cultures, we tested the effects of OIF on formation of these osteoclast-like cells to determine the effects of OIF on cells in the osteoclast lineage. We found that OIF inhibited total multinucleated cell (MNC) formation in a dose-dependent fashion and preferentially inhibited formation of MNCs that react with monoclonal antibody 23c6 (23c6-positive MNCs), an antibody that identifies osteoclasts. In addition, low concentrations of OIF in combination with low concentrations of TGF-beta acted synergistically to inhibit 23c6-positive MNC formation. The inhibition of 23c6-positive MNC formation by OIF was not mediated by prostaglandin synthesis. These data suggest that regulatory growth factors, such as OIF or TGF-beta, that are stored within the bone matrix and released when bone is resorbed can serve as natural inhibitors of osteoclast activity by inhibiting osteoclast formation.