INTRODUCTION — Bone is the third most frequent site of tumor metastasis; it is estimated that in the United States approximately 350,000 people die each year from bone metastases [1]. Analysis of medical records has demonstrated that 8.4 percent of patients with solid tumors have bone metastases by ten years from the initial diagnosis [2].
The incidence of bone metastases varies based on age, sex, and the primary tumor site [3,4]. Bone is the preferred site for breast and prostate cancer metastasis, with 50 to 70 percent of patients with advanced disease having bone involvement [5,6]. Many other cancers, including lung cancer [7], renal cell cancer, and melanomas, also colonize the skeleton prior to metastasizing to other organs [2]. For example, in autopsy series, 30 to 40 percent of patients with carcinomas of the thyroid, kidney, and bronchus have bone metastases [8]. Although hematologic malignancies do not frequently cause bone disease, multiple myeloma is the most frequent cancer to involve bone, with 70 percent of patients presenting with bone involvement at diagnosis and over 80 percent of patients having bone involvement with advanced disease [9,10]. Importantly, the presence of cancer in bone has been termed "the lethal phenotype" [11] as patients are rarely curable once bone is involved.
Normal bone remodeling is markedly disturbed in bone metastasis, with imbalances in both osteoclast and osteoblast numbers and activity [12]. These disturbances in the bone remodeling process can result in debilitating skeletal-related events (SREs), which are common complications (31 to 47 percent of patients with breast or prostate cancer or myeloma [13]) that have catastrophic sequelae for patients. SREs include severe bone pain, pathologic fractures, spinal cord and nerve compression syndromes, and derangements of calcium and phosphate homeostasis that can result in life-threatening hypercalcemia [14]. Because bone involvement is the main source of cancer-related pain and is the major cause of severe pain for patients with advanced malignancies [15], SREs not only greatly increase morbidity and mortality [16], but also diminish quality of life for patients, which further impacts survival.
In addition to SREs, a significant proportion of patients with bone metastases have systemic muscle weakness that results from cancer-related bone destruction [17], which increases the risk of falls, which can result in fractures, and negatively impacts performance status, survival, and quality of life. Advances in our understanding of the mechanisms responsible for bone metastasis have provided the basis for development of mechanistic-based therapies to treat these patients and potentially reverse or prevent the severe sequelae of cancer in bone. Several of these therapies are in clinical trials. (See "Overview of cancer pain syndromes", section on 'Multifocal bone pain' and "Clinical presentation and evaluation of complete and impending pathologic fractures in patients with metastatic bone disease, multiple myeloma, and lymphoma" and "Clinical features and diagnosis of neoplastic epidural spinal cord compression" and "Hypercalcemia of malignancy: Mechanisms".)
This topic will review the mechanisms by which bone metastases develop and affect normal bone. Control of normal bone remodeling is discussed separately. (See "Normal skeletal development and regulation of bone formation and resorption".)
OSTEOLYTIC VERSUS OSTEOBLASTIC BONE METASTASES — Normal bone constantly undergoes a remodeling process that involves the resorption of bone by osteoclasts and the deposition of new bone by osteoblasts. The bone remodeling process, including osteoclast and osteoblast activity, is regulated by the osteocyte [18].
Bone metastases are generally classified as osteolytic or osteoblastic based upon a radiologic appearance that demonstrates predominant bone destruction or deposition of new bone, respectively. However, this distinction is not absolute, and many patients have a mixed picture of both osteolytic and osteoblastic metastases (table 1). Both types of bone metastases are characterized by dysregulation of the normal bone remodeling process. (See "Normal skeletal development and regulation of bone formation and resorption", section on 'Bone modeling and remodeling'.)
The spectrum of osteolytic and osteoblastic bone involvement is illustrated by three malignancies that are frequently associated with skeletal metastases:
●Multiple myeloma – The classic bone lesions in multiple myeloma are purely osteolytic due to increased bone destruction and suppressed bone formation (image 1) [19]. These lesions are often not detectable on bone scans, which rely upon technetium uptake to identify areas of new bone formation. (See "Multiple myeloma: Clinical features, laboratory manifestations, and diagnosis" and "Epidemiology, clinical presentation, and diagnosis of bone metastasis in adults", section on 'Bone scan'.)
●Prostate cancer – Males with bone metastases from prostate cancer predominantly have osteoblastic lesions with increased numbers of irregular bone trabeculae (image 2) [20]. Although these metastases are osteoblastic, bone resorption is also increased [21,22]. Thus, agents that block bone resorption can decrease bone pain and the risk of skeletal-related events in patients with metastatic prostate cancer [23]. (See "Bone metastases in advanced prostate cancer: Management", section on 'Osteoclast inhibitors'.)
●Breast cancer – Although bone metastases due to breast cancer are predominantly osteolytic, osteoblastic areas are also usually present, and at least 15 to 20 percent of cases have predominantly osteoblastic lesions [20,24]. In addition, the bone destruction in osteolytic lesions can induce secondary new bone formation.
BONE AS A PREFERENTIAL SITE FOR METASTASIS — Multiple mechanisms have been identified that increase the potential for tumor cells to metastasize to bone. The intrinsic properties of the tumor cells, changes induced by the tumor cells or their products in the bone microenvironment, and the bone microenvironment itself all contribute to the preferential colonization and growth of cancer cells in bone [25]. Stephen Paget in 1889 published the "seed and soil" hypothesis [26], in which he proposed that cancer cells only grow in "congenial" secondary sites, to explain his autopsy findings that bone was a preferred site for breast cancer metastasis. This hypothesis went against the prevailing belief that when cancer cells spread from the primary site, they would grow similarly in any tissue that they "seeded."
Intrinsic tumor cell properties — In preclinical studies of mice injected with human breast cancer cells, breast cancer clones with a high propensity to induce bone metastasis were found to express a specific four-gene signature that increased their bone metastatic potential [27]. These genes included CXCR4 (CXC motif chemokine 4), which binds CXCL12 (the product of the CXC motif chemokine ligand 12 gene) on bone cells to increase bone homing, fibroblast growth factor 5 (an angiogenic factor), interleukin 11 (IL-11; an osteoclast stimulatory factor), and osteopontin. Breast cancer cells that expressed at least three of these four genes had an increased propensity to metastasize to bone. Validation studies are needed to confirm these findings.
The investigators also found that increased expression of both matrix metalloproteinase 1 (MMP1) and A disintegrin-like and metalloproteinase with thrombospondin type 1 motif 1 (ADAMTS1) in breast cancer cells further enhanced their bone metastatic potential. MMP1 and ADAMTS1 cooperated to cleave amphiregulin (AREG), an epidermal growth factor family member, from the breast cancer cell membrane. AREG in turn increased bone resorption by suppressing osteoprotegerin (OPG) expression by osteoblasts [28]. OPG is a decoy receptor for receptor activator of nuclear factor kappa-B ligand (RANKL) that blocks its capacity to induce osteoclast formation and activity and thereby increase bone resorption. Increased bone resorption enhances bone metastasis in breast cancer [29] and prostate cancer [30].
Changes induced by tumor cells or their products in the bone microenvironment — Several changes induced by tumor cells or their products in the bone microenvironment have been implicated in the propensity of tumor cells to metastasize to bone. As examples:
●Increased transforming growth factor (TGF)-beta signaling in the tumor cells plays a crucial role in prostate cancer and breast cancer bone metastasis. TGF-beta controls expression of multiple genes that promote bone metastasis, including CXCR4, MMP1, IL-11, Jagged 1 (JAG1), and parathyroid hormone-like hormone, in breast and prostate cancer cells [31,32]. Others have found that TGF-beta also induces expression of an inhibitor of TGF-beta signaling in prostate cancer cells [33]. PMEPA1 (prostate transmembrane protein, androgen induced) expression was higher at the primary tumor site than in bone metastases, and decreased expression of PMEPA1 in the bone metastases was associated with decreased survival and cancer recurrence in patients with breast, lung, and prostate cancer. The decreased expression of PMEPA1 in prostate cancer cells in bone most likely resulted from methylation of the PMEPA1 gene that inhibited PMEPA1 expression. Finally, they showed that increased TGF-beta signaling was required for formation of osteolytic bone metastases by the prostate cancer cells.
●Tumor-associated stromal cells in the primary tumor site also can increase the bone metastatic potential of breast cancer cells [34]. One study found that cancer-associated fibroblasts bias the heterogeneous population of breast cancer cells present in the primary tumor toward clones that express SRC (the product of the SRC oncogene), which allows the cancer cells to respond to the CAF-derived factors CXCL12 and insulin-like growth factor 1 (IGF1) [34,35]. Because limiting concentrations of these factors are available at the primary tumor site, cancer cells with high SRC activity are primed for metastasis to the CXCL12-rich bone microenvironment that is present in the bone marrow. SRC activity promotes breast cancer cell growth and survival in bone marrow [36].
●Increased expression of adhesive molecules, cytokine receptors, and receptor ligands on the surface of tumor cells also plays an important role in bone metastasis. CXCR4, which binds to CXCL12 and is expressed on the surface of pericytes and bone marrow stromal cells in bone, is a homing receptor for breast cancer, prostate cancer, and multiple myeloma cells to bone [37]. Tumor cells expressing alpha-v beta-3 integrin or E-cadherin can also home to the bone marrow by binding osteopontin, bone sialoprotein, vitronectin, or N-cadherin, respectively [38]. Similarly, multiple myeloma cells express alpha-4 beta-1 integrin to bind vascular cell adhesion molecule 1 on marrow stromal cells [39]. Additionally, breast cancer, melanoma, and prostate cancer cells express RANK (the receptor for RANKL), which promotes their homing to the marrow [40].
●Multiple products of tumor cells also prepare the premetastatic niche in bone for tumor cell colonization [41]. These include lysyl oxidase, which increases osteoclast activity; extracellular vesicles (EVs)containing micro-RNAs that induce angiogenesis or increase osteoclast activity [42]; and factors that induce immune suppression in the bone microenvironment by increasing regulatory T cells and myeloid suppressor cells that enhance tumor colonization of bone [43,44].
Tumor cell-derived EVs, including exosomes and microvesicles, contain a variety of macromolecules such as proteins, mRNA, and miRNA. EV release from the primary tumor have been shown to target the bone marrow and establish a premetastatic niche that promotes targeting and growth of cancer cells in the bone microenvironment [45-47].
Cross talk between macrophages and cancer cells in the bone microenvironment promote metastasis progression [48]. Macrophages in the bone marrow and primary tumor site can be activated by tumor cells to change from M1 (antitumorigenic) to M2 (protumorigenic) macrophages [49]. This increase in M2 macrophages in bone contributes to the growth of prostate cancer bone metastases [50]. Further, TGF-beta can induce expression of the notch receptor ligand JAG1 on breast cancer cells to activate notch signaling in osteoblasts and increase production of interleukin 6 (IL-6) [32]. IL-6 enhances tumor growth in bone and increases the size of lytic lesions. In keeping with the importance of this pathway, investigators have developed a therapeutic antibody targeting JAG1 on tumor cells that successfully blocked breast cancer bone metastasis in a preclinical model and sensitized the bone metastases to chemotherapy [51].
Cytokines produced by the primary tumor or the bone microenvironment, such as TGF-beta, also upregulate expression of adhesion molecules on bone marrow stromal cells and osteoblasts, and interact with hypoxia-inducible factor 1 alpha to increase vascular endothelial growth factor and CXCR4 expression in the bone microenvironment, which increases bone metastasis [52].
Physical characteristics of bone — The physical characteristics of bone may also contribute to tumor cell colonization.
The bone extracellular matrix is extremely rigid. Exposure to the high matrix rigidity in bone increases levels of the GLI-Kruppel family member 2 (GLI2) gene transcription factor in breast cancer cells. GLI2 in turn increases parathyroid hormone-related protein promoter activity and TGF-beta signaling, which promote cancer growth in bone [53].
Osteocytes mediate mechanotransduction (ie, the ability to sense and induce cell signaling secondary to mechanical forces). As tumor cells grow, they can induce pressure in the bone microenvironment. The increase in pressure is exacerbated by the rigidness of bone. Intraosseous pressure has been shown to stimulate osteocytes to produce the C-C motif chemokine ligand and MMPs that promote cancer growth and invasion in bone [54].
CONTRIBUTION OF BONE CELLS TO TUMOR CELL DORMANCY, REACTIVATION, AND TUMOR PROGRESSION IN BONE — In the skeleton, the multistep process of metastasis development begins with colonization, when circulating tumor cells enter the bone marrow compartment and engage in specialized microenvironments or niches [55,56]. The second step involves survival and dormancy in that the colonizing tumor cells adapt to their new microenvironment, evade the immune system, and reside in a dormant state for long periods of time [57]. The third step, reactivation and development, requires the ability to escape from the dormant state to actively proliferate and form micrometastases [58]. The final step occurs when cells grow uncontrollably, become independent of the microenvironment, and ultimately modify bone as the metastases flourish.
The molecular mechanisms responsible for both osteolytic and osteoblastic metastases are being identified. The use of gene arrays and proteomics, and the availability of appropriate animal models of bone metastasis have permitted the identification of factors, produced by the tumor cells themselves or by the bone cells or bone microenvironment in response to the tumor, that mediate dormancy and reactivation of tumor cells in bone and the bone destructive process. Advances in understanding these pathways have revealed new potentially beneficial therapeutic approaches to target bone metastases [59,60].
Hematopoietic stem cell niches — When tumor cells colonize bone, they appear to localize to the same specific areas in bone as hematopoietic stem cells (HSCs). This area is the osteoblastic niche, which regulates and supports HSC self-renewal, quiescence, and differentiation [60,61]. In support of this concept, investigators have shown that prostate cancer cells directly compete with HSCs for niche support and can displace HSCs from the osteoblastic niche, releasing HSCs into the circulation [62]. Similarly, when myeloma cells home to bone, they also interact with the osteoblastic niche and are retained in a dormant state [63], which protects the myeloma cells from the effects of chemotherapy. Furthermore, metastatic breast cancer cells localize to the hematopoietic stem cell niche that is heavily populated with osteoblasts [64]. This hematopoietic stem cell mimicry is regulated, in part, by Notch 2 in breast cancer cells [65].
The molecular mechanisms responsible for tumor cell dormancy are not well defined, but they may involve binding of the annexin II receptor on myeloma and prostate cancer cells to annexin II expressed on osteoblastic lineage cells in bone [55]. Annexin II binding to myeloma and prostate cancer cells can regulate tumor cell growth [66,67]. This interaction also increases AXL expression in prostate cancer cells, increasing transforming growth factor (TGF)-beta-2 signaling, which is also associated with tumor cell dormancy [68-70]. Others have shown that dormant breast cancer cells in bone reside on the vasculature of metastatic sites, and that tumor cell dormancy in this case is mediated by endothelial-produced thrombospondin-1 [69] and induction of Notch 2 signaling in the breast cancer cells [65].
Our understanding of how dormant tumor cells escape control of the niche is very limited. Osteoclastic bone resorption can reactivate and release dormant myeloma and breast cancer cells from their niches, analogous to the release of HSCs from their niche [42,71,72]. Other data suggest that reactivation may be under tumor cell control, which may be intrinsic to the colonizing tumor cells or acquired once the cells enter the niche [55,73]. In addition, inflammatory activity appears to promote tumor cell plasticity and dormancy escape [74]. A model depicting the various bone marrow niches, cell types, and cues that might regulate the dormancy of disseminated tumor cells is illustrated in the figure (figure 1) [59].
Osteoclasts — Studies in preclinical models have shown that higher levels of osteoclast activity increase tumor metastasis and growth in bone, while blocking osteoclast bone resorption decreases osteolytic metastasis [75,76]. In clinical studies, blocking bone resorption through the use of osteoclast inhibitors reduces skeletal-related events (SREs) and, at least in multiple myeloma, increases the overall survival of patients with bone metastases. By contrast, while antiresorptive agents decrease SREs and improve quality of life, they appear to have limited impact on overall survival in men with prostate cancer bone metastases [77]. (See "Osteoclast inhibitors for patients with bone metastases from breast, prostate, and other solid tumors" and "Multiple myeloma: The use of osteoclast inhibitors", section on 'Indications'.)
Multiple factors that are induced or produced by tumor cells in the bone microenvironment enhance osteoclastic bone resorption [12]. The resulting increased bone destruction releases growth factors, such as TGF-beta, insulin-like growth factor 1 (IGF1), and others, to increase tumor cell growth. This results in a "vicious cycle" (figure 2) whereby tumor cells in bone induce osteoclastic bone resorption that releases growth factors from the bone matrix to further increase tumor growth.
RANK/RANKL signaling pathway and bone remodeling — Products of tumor cells can activate osteoclastic bone resorption in a number of ways. The receptor activator of nuclear factor kappa-B (RANK)/RANK ligand (RANKL) signaling pathway is a major regulator of both normal and pathologic bone remodeling. Many of the cytokines, chemokines, and hormones produced by tumor cells that induce osteolytic metastasis do so by increasing RANKL expression.
RANKL is a type II homotrimeric (a trimer derived from three identical monomers) transmembrane protein that exists as a membrane-bound or secreted soluble protein that is produced by cleavage of the full-length form on the cell surface [78]. In bone, RANKL is mainly expressed by bone marrow stromal cells, osteoblasts, and osteocytes and is also secreted by activated T lymphocytes [12]. Studies suggest that osteocytes are the major source of RANKL production in bone and produce 10-fold higher levels of RANKL than do osteoblasts [79]. RANKL binds to RANK, a member of the tumor necrosis factor (TNF) receptor superfamily, which is expressed on the surface of osteoclast precursors and mature osteoclasts (figure 3). The binding of RANKL to RANK stimulates a number of signaling cascades that are vital for osteoclast differentiation, survival, and activity [80]. (See "Normal skeletal development and regulation of bone formation and resorption", section on 'Signaling pathways'.)
Osteoprotegerin (OPG), another member of the TNF receptor superfamily, is a soluble decoy receptor for RANKL that is produced by mature osteoblasts and osteocytes and normally blocks the interaction between RANKL and its receptor on osteoclasts [81]. (See 'Osteoblasts' below.)
The RANKL/OPG ratio is crucial for regulating osteoclast formation and activity. Many of the cytokines, chemokines, and hormones produced by tumors cells that induce osteolytic metastasis (eg, parathyroid hormone-related protein [PTHrP], 1,25D3, prostaglandins, interleukin 1 beta [IL1B], macrophage inflammatory protein 1 alpha [MIP-1-alpha], and TNF-alpha) do so by increasing RANKL expression [19].
It is unclear if tumor cells in bone directly produce RANKL or only induce its production by acting on cells (predominantly osteoblasts and osteocytes) in the bone microenvironment. (See 'Osteoblasts' below.)
Several studies report that RANKL is expressed by multiple myeloma cells in bone marrow biopsies and by human and murine multiple myeloma cell lines [82]. Others have shown that multiple myeloma cells from patients expressed RANKL and that RANKL was responsible for the release of TNF-alpha, interleukin 6 (IL-6), and interleukin 8 (IL-8) via an autocrine/paracrine mechanism that increased the survival and growth of malignant cells and exacerbated bone destruction [83]. However, still other investigators have found that certain (syndecan-1 [CD138] positive) multiple myeloma cells and multiple myeloma cell lines do not express RANKL, and that an increased RANKL/OPG ratio occurred only when human multiple myeloma cells were cocultured with bone marrow stromal cells [84]. The increased RANKL/OPG ratio in multiple myeloma patients correlated with poor prognosis and reduced survival [81,85]. Importantly, preclinical models of multiple myeloma bone disease demonstrate that blocking RANKL-induced osteoclast formation, via administration of recombinant OPG or RANK-Fc, significantly decreases osteolytic lesions and tumor growth in mice [86].
The RANK/OPG axis has also been demonstrated to impact progression of prostate cancer bone metastases. Inhibition of RANKL using either OPG [87] or a solubilized form of the RANKL receptor [88] decreases growth of prostate cancer in bone in preclinical models.
Denosumab, a human antibody to RANKL, is highly effective for preventing SREs in patients with bone metastases from prostate cancer, breast cancer, and other solid tumors, and in multiple myeloma. It also prevents treatment-related bone loss in men treated with androgen deprivation therapy for prostate cancer and in women with breast cancer who are treated with aromatase inhibitors. In spite of the efficacy of denosumab for decreasing skeletal-related events, it is accompanied by rare, but serious toxicities such as osteonecrosis of the jaw suggesting that dose optimization and additional therapeutic strategies need consideration. (See "Osteoclast inhibitors for patients with bone metastases from breast, prostate, and other solid tumors" and "Multiple myeloma: The use of osteoclast inhibitors" and "Use of osteoclast inhibitors in early breast cancer" and "Medication-related osteonecrosis of the jaw in patients with cancer" and "Risks of therapy with bone antiresorptive agents in patients with advanced malignancy".)
In addition to inducing RANKL or decreasing OPG, products of tumor cells can activate osteoclastic bone resorption in other ways:
●Jagged 1 (JAG1) expressed on breast cancer cells can activate notch signaling in osteoclast precursors to induce osteoclast formation [32].
●Metalloproteinases, such as matrix metalloproteinase (MMP) 7, produced by osteoclasts in response to prostate cancer cells can cleave membrane-bound RANKL into a secretable soluble protein that increases bone destruction [89].
●Multiple tumor types produce PTHrP including prostate cancer [90,91], multiple myeloma [92], and breast cancer [93]. PTHrP induces osteoclast activity through induction of RANKL.
●MMP13 produced by myeloma cells can increase osteoclast precursor fusion to enhance osteolysis. This effect of MMP13 is independent of its proteolytic activity [94].
Osteoblasts — Cells of the osteoblast lineage primarily contribute to bone metastasis through production of RANKL and OPG. Immature osteoblasts/bone marrow stromal cells produce RANKL, while more mature osteoblasts and osteocytes produce OPG. As noted above, multiple tumor-derived factors can induce RANKL expression by immature osteoblasts or block OPG expression by mature osteoblasts or osteocytes. For example, direct interactions between myeloma cells and osteocytes increase RANKL expression and decrease OPG expression by osteocytes [95]. In addition, myeloma cells inhibit osteoblast precursor differentiation, thereby increasing the RANKL/OPG ratio. The RANKL/OPG ratio is crucial for regulating osteoclast formation and activity. (See 'RANK/RANKL signaling pathway and bone remodeling' above.)
As noted above, cells of osteoblast lineage also contribute to maintaining tumor cells in the dormant state within HSC niches. (See 'Hematopoietic stem cell niches' above.)
Advances in understanding the molecular mechanisms underlying these osteoblast functions suggest the potential for new therapeutic approaches to target bone metastases. As an example, it has been reported that hepatocyte growth factor (HGF) and vascular endothelial growth factor (VEGF) alpha, which are present in the tumor microenvironment, induce expression of RANKL and macrophage colony-stimulating factor (MCSF) by mature osteoblasts through transactivation of c-Met, the receptor for HGF, and that this increases osteolysis [96]. These same investigators then showed that a dual kinase inhibitor of c-Met and the VEGF receptor 2 on osteoblasts could significantly reduce the growth of prostate cancer in bone. Thus, targeting these pathways in osteoblasts could represent a new approach to treating prostate cancer metastatic to bone.
Additionally, bone matrix proteins produced by osteoblasts, such as osteopontin, bone sialoprotein, and decorin, affect bone metastasis. High levels of osteopontin increase bone metastasis, while overexpression of decorin inhibits bone metastasis [97,98]. These results suggest that blocking osteopontin-breast cancer cell interaction or overexpressing decorin may provide new ways to treat bone metastases. The bone matrix microstructure itself may regulate bone metastasis. Typically bone matrix is an organized structure, but dysregulated remodeling induced by inflammation or tumors can lead to disorganized crystalline structure. In the case of melanoma and breast cancer, the disorganized crystalline structure is associated with increased homing of cancer cells to bone [99,100].
Tumor cells in bone can also regulate osteoblast differentiation. Tumor cells that cause osteoblastic metastasis produce osteoblast differentiation factors, such as endothelin-1, bone morphogenic proteins (BMPs), insulin-like growth factors, platelet-derived growth factors, and fibroblast growth factors, that drive the development of osteoblastic metastases and increase tumor cell growth [41]. In one study, administration of a selective endothelin-1 receptor antagonist decreased both osteoblastic metastases and tumor burden in an animal model, although it had no effect on tumor growth at orthotopic sites [101].
BMPs, including BMP-6 and BMP-2, also increase osteoblast differentiation in osteoblastic metastasis and appear to act independently of endothelin-1 [102]. As an example, the administration of an anti-BMP antibody inhibited prostate cancer growth in bone in a preclinical model [102]. Both BMP-6 and BMP-2 also increase the local invasiveness of prostate cancer cells in the bone microenvironment. Overproduction of urokinase-type plasminogen activator (uPA) by prostate cancer cells increases in bone metastasis [103], and cells transfected with an antisense DNA to uPA had a threefold decrease in bone metastases compared with empty-vector transfected cells. An anti-urokinase receptor antibody also blocked bone metastasis by prostate cancer cells [104]. These observations suggest that blocking the production of osteoblast-inducing activity by tumors may decrease tumor growth. Thus, as with osteolytic metastasis, a feedback cycle may be present in which tumors induce osteoblast activity and osteoblasts in turn release growth factors that increase tumor growth. (See 'Osteoclasts' above.)
Other soluble factors can modulate osteoblast differentiation and impact bone metastasis. For example, activation of the retinoic acid receptor inhibits the endothelial cell transition to osteoblasts resulting in decreased prostate cancer-induced bone formation [105].
By contrast, tumor cells that induce osteolytic lesions may also inhibit osteoblast differentiation, with the most profound inhibition occurring in myeloma (figure 4). Multiple factors act to suppress osteoblast differentiation in myeloma. These include soluble factors such as interleukin 7 (IL-7) and TNF-alpha, the Wnt signaling antagonists Dickkopf-1 (DKK1) and sclerostin (produced by multiple myeloma cells and osteocytes), as well as adhesive interactions between myeloma cells and bone marrow stromal cells [106]. DKK1 levels are also increased in osteolytic breast cancer bone metastasis and early in the course of prostate cancer bone metastasis [107,108]. However, DKK1 levels decrease in response to the PTHrP produced by prostate cancer cells, allowing formation of the osteoblastic metastases that are characteristic of prostate cancer [107] (see 'Osteolytic versus osteoblastic bone metastases' above). In addition to direct effects on bone, DKK1 may promote bone metastasis through regulation of an antitumor immune response [109].
Importantly, the suppression of bone formation in multiple myeloma persists even when multiple myeloma cells are eradicated and can no longer affect osteoblast differentiation through direct interactions or release of osteoblast inhibitory soluble factors. Epigenetic changes in RUNX2 (runt-related transcription factor 2) and SP7 (also called Osterix), master genes required for osteoblast differentiation, occur in bone marrow stromal cells exposed to myeloma cells [110]. These epigenetic changes play a critical role in the protracted suppression of osteoblast differentiation in multiple myeloma.
The transcriptional repressor growth factor-independent 1 (GFI1) is upregulated in bone marrow stromal cells from myeloma patients and can induce long-term suppression of osteoblast differentiation in multiple myeloma. GFI1 binds to the RUNX2 promoter and recruits chromatin co-repressors to RUNX2 to induce epigenetic changes at the RUNX2 locus that block RUNX2 gene transcription [110]. These repressive chromatin changes persist even after removal of the multiple myeloma cells. Importantly, knockdown of GFI1 or selective pharmacologic inhibition of these co-repressors in bone marrow stromal cells prevented repression of RUNX2 transcription induced by multiple myeloma and restored osteoblast differentiation. These data suggest that treatment of multiple myeloma patients with clinically available co-repressor inhibitors might be beneficial for reversing the profound osteoblast suppression that is associated with multiple myeloma, allowing the repair of osteolytic lesions [110].
Osteocytes — The important contributions of osteocytes to cancer progression in bone are just becoming appreciated. Osteocytes comprise more than 95 percent of all bone cells (compared with 1 to 2 percent for osteoclasts and 3 to 4 percent for osteoblasts) and are central regulators of physiologic bone remodeling [111]. The osteocyte dendritic network allows direct cell-to-cell contact not only between osteocytes but also with cells on the bone surface and in the bone marrow. This network also distributes osteocyte-secreted molecules within the bone/marrow microenvironment and into blood vessels, so that they can enter the general circulation. Osteocytes produce RANKL, OPG, sclerostin, and DKK1, which regulate generation and activity of osteoclasts and osteoblasts. In addition, apoptotic osteocytes target bone remodeling to specific areas in bone [112]. Induction of osteocyte apoptosis is sufficient to increase resorption and bone loss [113], suggesting that osteocyte apoptosis underlies pathologic conditions involving enhanced bone resorption. The mechanisms underlying osteoclast recruitment and differentiation driven by osteocyte apoptosis are not fully understood.
Osteocyte apoptosis is increased in the bone lesions of patients with multiple myeloma, and the number of apoptotic osteocytes positively correlates with the number of osteoclasts in bone [114]. Multiple myeloma cells interact directly with osteocytes, and these interactions activate bidirectional notch signaling between multiple myeloma cells and osteocytes, which increases multiple myeloma cell proliferation and induces osteocyte apoptosis, which is further increased by the TNF-alpha secreted by the multiple myeloma cells [95]. Osteocyte apoptosis induced by multiple myeloma cells in turn increases RANKL expression in osteocytes and stimulates osteoclast recruitment. Inhibiting osteocyte apoptosis prevents these effects. In addition, multiple myeloma cells increase expression of sclerostin in osteocytes, which inhibits osteoblast differentiation. Deletion of the sclerostin gene or neutralization of its product with an anti-sclerostin antibody stimulates bone formation and reduces the bone resorption induced by multiple myeloma cells [115]. Furthermore, myeloma promotes osteocyte production of VEGF alpha and subsequent angiogenesis [116].
There are some potential therapeutic implications of these findings, at least in the setting of multiple myeloma:
●Bortezomib, a proteosome inhibitor with antitumor activity in myeloma, decreases osteocyte autophagic/apoptotic death in vitro and the number of dead osteocytes in the bones [117]. Strong immunoexpression of DKK1 is associated with the response to bortezomib [118].
●Antibody-based inhibition of DKK1 prevented the development of osteolytic bone disease in a murine model of multiple myeloma [119], and at least one inhibitor (BHQ880) may be beneficial for patients with relapsed or refractory myeloma and prior SREs [120].
●Antisclerostin antibody has been shown to decrease myeloma-associated bone loss in a preclinical model [121].
In addition to myeloma, others have also shown a role for osteocytes in prostate cancer bone metastasis by showing that tumor-generated pressure in bone promotes prostate cancer bone metastasis acting through osteocytes [54]. Application of pressure to osteocytes, the main mechanotransducing cells in bone, increased C-C motif chemokine ligand 5 (CCL5) and metalloproteinases in osteocytes, which enhanced prostate cancer growth and invasion. Furthermore, it has been suggested that mechanical stimulation of osteocytes promotes proliferation and migration of breast cancer cells through CXC motif chemokine ligand 12 (CXCL12) [122]. (See 'Physical characteristics of bone' above.)
Finally, in addition to promoting bone metastatic progression, there is evidence that stimulation of Wnt signaling pathway in osteocytes may protect against bone metastasis [123].
Bone marrow adipocytes — Bone marrow adipocytes are one of the most abundant cells in adult bone marrow, and they increase with age, so that by age 65, adipocytes comprise 60 percent of marrow volume [124]. Adipocytes are derived from the same common mesenchymal stromal precursor as osteoblasts, which is induced to differentiate to a particular lineage by differentiation factors in the bone marrow microenvironment [125].
Marrow adipocytes promote bone metastasis and enhance the growth of cancers through the production of adipokines and chemokines that increase tumor cell survival, growth, and osteoclastogenesis [126-128]. The role of adipocytes in the development and progression of bone metastases is supported by the following observations:
●Cell-to-cell interactions between cancer cells and adipocytes induce morphologic and phenotypic changes in the adipocytes that decrease their expression of adiponectin, a suppressor of tumor growth [129], decrease their lipid content, and decrease expression of adipocytic genes, while increasing expression of inflammatory cytokines. Inflammatory cytokines and chemokines, such as IL-6, TNF-alpha, CXCL12, and leptin, can enhance multiple myeloma cell growth and migration while preventing apoptosis of myeloma cells [130].
●Adipocyte-derived CXC motif chemokine receptor 1 and CXCL12 increase the osteoclast activity that enhances metastatic prostate cancer growth and survival [131].
●Secretion of IL1B and leptin by bone marrow adipocytes attracts breast cancer cells to bone [132].
●A high-fat diet increases adipocytes in bone marrow with an associated increased caprylic acid, a medium-chain fatty acid, that promoted prostate cancer invasion in a preclinical model [133].
Platelets and megakaryocytes — Platelets also play an important role in bone metastasis. Alpha-2 beta-3 integrin on platelets controls the homing of melanoma cells to bone [134]. Additionally, platelet-derived lipid lysophosphatidic acid induces growth of breast cancer cells in bone and production of the osteoclastogenic factors IL-6, IL-8, and monocyte chemoattractant protein 1 [135]. Consistent with these findings, investigators have shown that antiplatelet therapy decreases breast and melanoma bone metastasis in mice [136].
In contrast to platelets, megakaryocytes (the hematopoietic cells that produce platelets) act to decrease bone metastasis by suppressing osteoclast formation through the production of OPG [137], as well as by increasing bone formation [138]. Increasing megakaryocytes in the marrow through the use of recombinant thrombopoietin has been shown to block prostate cancer bone metastasis [139].
Macrophages modulate bone metastases — The following evidence supports the view that macrophages modulate bone metastases:
●Ablation of macrophages inhibits breast cancer growth in bone [140].
●Furthermore, it was shown that the macrophage promoted metastasis through interleukin-4 receptor. In prostate cancer CCL5 derived from tumor-associated macrophages promotes metastasis via b-catenin/STAT 3 pathway activation [141].
●Bone marrow macrophages have been shown to clear apoptotic cancer cells (a process called efferocytosis) with a resulting increase in production of inflammatory cytokines, that could enhance cancer cell growth in bone [142].
Taken together these studies underscore the importance of future studies delineating the roles of macrophages as mediators and potential therapeutic targets for bone metastasis.
INFORMATION FOR PATIENTS — UpToDate offers two types of patient education materials, "The Basics" and "Beyond the Basics." The Basics patient education pieces are written in plain language, at the 5th to 6th grade reading level, and they answer the four or five key questions a patient might have about a given condition. These articles are best for patients who want a general overview and who prefer short, easy-to-read materials. Beyond the Basics patient education pieces are longer, more sophisticated, and more detailed. These articles are written at the 10th to 12th grade reading level and are best for patients who want in-depth information and are comfortable with some medical jargon.
Here are the patient education articles that are relevant to this topic. We encourage you to print or e-mail these topics to your patients. (You can also locate patient education articles on a variety of subjects by searching on "patient info" and the keyword(s) of interest.)
●Basics topics (see "Patient education: Bone metastases (The Basics)")
SUMMARY
●Morbidity and mortality – Bone is a frequent site of metastasis, and bone metastases cause significant morbidity and increased mortality for patients. (See 'Introduction' above.)
●Osteoblastic versus osteolytic metastases – Normal bone constantly undergoes a remodeling process that includes the resorption of bone by osteoclasts and the deposition of new bone by osteoblasts. Bone metastases are generally classified as osteolytic or osteoblastic based upon the radiologic appearance of predominant bone destruction or deposition of new bone. However, this distinction is not absolute, and many patients have a mixed picture of both osteolytic and osteoblastic metastases. Both types of bone metastases are characterized by dysregulation of the normal bone remodeling process. (See 'Osteolytic versus osteoblastic bone metastases' above.)
●Bone is a preferential site for metastases – Multiple mechanisms increase the potential for tumor cells to preferentially metastasize to bone, including the intrinsic properties of the tumor cells, changes induced by the tumor cells or their products in the bone microenvironment, and the bone microenvironment itself. (See 'Bone as a preferential site for metastasis' above.)
●Interaction of bone cells with metastases and molecular mechanisms – In the skeleton, the multistep process of metastasis development begins with colonization, when circulating tumor cells enter the bone marrow compartment and engage in specialized microenvironments or niches. The second step involves survival and dormancy in that the colonizing tumor cells adapt to their new microenvironment, evade the immune system, and reside in a dormant state, sometimes for long periods of time. The third step, reactivation and development, requires the ability to escape from the dormant state to actively proliferate and form micrometastases. The final step occurs when cells grow uncontrollably, become independent of the microenvironment, and ultimately modify bone as the metastases flourish. (See 'Contribution of bone cells to tumor cell dormancy, reactivation, and tumor progression in bone' above.)
The molecular mechanisms responsible for both osteolytic and osteoblastic metastases are being identified. The use of high throughput genomics, proteomics, and single-cell analytic technology; the availability of appropriate animal models of bone metastasis; and in vitro models have permitted the identification of several factors, produced by the tumor cells themselves or by the bone cells or bone microenvironment in response to the tumor, that mediate dormancy and reactivation of tumor cells in bone and the bone destructive process.
Advances in understanding these pathways have revealed new potentially beneficial therapeutic approaches to target bone metastases in a variety of malignancies, especially multiple myeloma. (See 'Osteocytes' above.)
ACKNOWLEDGMENT — The UpToDate editorial staff acknowledges David G Roodman, MD, PhD, who contributed to an earlier version of this topic review.
آیا می خواهید مدیلیب را به صفحه اصلی خود اضافه کنید؟