The process of tumor metastasis involves multiple stages during which malignant tumor cells spread from a primary site to other parts of the body. Malignant tumors tend to be poorly encapsulated, invasive, and rapid growing. Bone is commonly involved in tumor metastasis because it is the third most common site of metastasis after lung and liver; certain malignancies are thought to be osteotropic. In clinical practice, the most commonly encountered osteotropic malignancies are prostate cancer, lung cancer, and breast cancer. It has been more than 100 years since Paget described metastatic disease from an analysis of autopsy reports of 735 women who died from metastatic breast cancer. The result of Paget’s research was the classic seed-and-soil hypothesis of metastasis in which metastasis is likened to the growth of vegetation. The seeds (cancer cells) are able to colonize and grow only in receptive soil (particular environments within the body). This hypothesis forms the basis of many of the principles of the biology of bone metastasis.

Bone is a highly mineralized tissue that contributes both metabolic function and mechanical support to the skeleton. Embryologically, bone is formed by endochondral or intramembranous ossification. At the ends of the long bones, bone formation occurs from a cartilaginous scaffold. In flat bones, such as those of the skull, bone can form from intramembranous ossification, which occurs when mesenchymal precursor cells differentiate into osteoblasts that stimulate bone production.

Bone cells are the basis of production in this metabolically active environment. Recent research and innovation into tumor metastasis to bone has focused on the products of these cells. Osteoclasts are derived from the same precursor as monocytes. Once activated, osteoclasts are responsible for the degradation of bone matrix. Macrophage colony-stimulating factor and receptor activator of nuclear factor kappa B ligand (RANKL) are essential to the production of osteoclasts. Macrophage colony-stimulating factor is produced by stromal cells within the bone marrow and is important in the beginning of osteoclast formation, whereas RANKL is involved in the maturation, differentiation, and activation of the osteoclast (Figure 1). Matrix metalloproteinases (MMPs) control the bioavailability and function of RANKL, and this function may make them key players in the process of bone metastasis.^1,2

Osteoblasts are the producers of bone. Differentiation of the osteoblast is fine-tuned by the parathyroid hormone-related protein (PTHrP), an important product that is a focus of current research into bone metastasis. As osteoblasts are embedded into the bone matrix, they undergo terminal differentiation into osteocytes, which further modulate changes in the bone microenvironment, as discussed in a 2023 study.³

Bone is composed of an organic matrix that is strengthened by calcium; most of the matrix is composed of type I collagen.⁴ Bone is a dynamic substance that turns over 20% of its mass each year through the process of remodeling. Because bone continuously remodels, bone-stored growth factors are continuously released by resorption of bone by osteoclasts.⁴ The release of growth factors provides a favorable environment for metastatic tumor cells to reproduce within the bone.

Tumor cells must survive to metastasize. The tumor cells that develop the appropriate genetic changes to survive are transported through the blood to the bone marrow. Within the bone marrow, the tumor cells must adapt to the new environment for continued survival. Tumor cells, and in particular, epithelial tumor cells, are well known to develop a mesenchymal phenotype. In essence, the relatively low cohesion that is characteristic of the mesenchymal phenotype allows cancer cells to migrate away from the main neoplasm to other areas of the body^4,5,6 (Figure 2). Bone is a site in which cancer cells can reproduce exceptionally well.

The first site of interaction between the metastatic tumor cell and the bone marrow is the endothelial cell.⁴ The combination of adhesion and chemoattractive molecules within the bone marrow endothelium makes it especially favorable for attracting circulating cancer cells. Stromal cell-derived factor 1 (SDF-1) induces hematopoietic stem cells of the bone marrow to undergo transmembrane migration mediated by P- and E-selectins. Some carcinoma cells use the same mechanism.⁴

Osteopontin is a major component of bone that mediates motility, survival, local adhesion, and growth by integrins. Cancer cell adhesion to osteopontin is integrin dependent. Therefore, integrins are critical not only in the movement of hematopoietic stem cells to hematopoietic sites but also in bone marrow colonization by cancer cells. Particularly, the integrins αIIbβ3 and αvβ3 play a role in bone marrow colonization in several cancers.^4,5

For malignant cells to enter the bone marrow, they must be able to penetrate the basement membrane and traverse the extracellular matrix. This movement can occur only if the normal balance between proteases and their inhibitors is disturbed.^2,5 The family of MMPs is central to the process of invasion. MMPs play a critical role in the process of bone remodeling, which is necessary for bone metastasis to occur. It is not surprising, then, that MMPs are upregulated in sites in which bone metastasis has occurred.² MMPs can be derived from several types of cells within the bone microenvironment, in addition to osteoclasts, osteoblasts, and tumor cells.² Metastatic tumor cells produce MMPs that are specific to individual tumor types; for example, metastatic breast cancer cells produce MMP-2 and MMP-13. Osteoblasts produce MMPs that are thought to be integral to skeletogenesis. The function of osteoclast MMPs is still unclear.²

Although these molecules are important in providing the conditions for metastatic lesions, an event must occur to allow metastasis.⁴ Vascular endothelial growth receptor 1 (VEGFR1)-positive, bone marrow-derived, hematopoietic progenitor cells were found to be mobilized by factors secreted by tumor cells to create niches in target organs. Remarkably, this premetastatic niche hypothesis assumes that long-range communication exists between the primary malignant cell and the future site of metastasis.⁴

Once cancer cells are in bone, they are required to undergo genetic changes that allow them to establish residence.⁵ During this process, called osteomimicry, metastatic tumor cells acquire genetic changes that allow them to produce bone matrix proteins. For example, bone sialoprotein is produced by metastatic breast cancer cells and assists in bone metastasis. Osteopontin is related to bone sialoprotein and also is important in the process of osteomimicry.⁵

Osteolytic bone resorption is key to the establishment of residence in bone. Active osteoclasts are associated with bone metastasis deposits and often are adjacent to these deposits. The ability of the cancer cells to promote the formation of active osteoclasts is a special property of tumors that produce bone metastases and is required for initiating and sustaining tumor expansion. Cancer cells promote the formation of osteoclasts by acting on host bone cells to induce the changes needed for osteoclast generation.⁵ Several cancers also directly produce RANKL, an important molecule in osteoclastogenesis.

The hypothesis of the vicious cycle of bone metastasis summarizes the steps in the process of bone metastasis (Figure 3). Much of the research into bone metastasis is based on the vicious cycle hypothesis.² In the first step of the cycle, bone-lining osteoblasts proliferate and/or differentiate through metastasis-derived signals. PTHrP is a mediator of this process. Bone morphogenetic proteins (BMPs); fibroblast growth factors (FGFs) such as FGF-9; endothelins; interleukins (ILs) such as IL-1, IL-6, and IL-8; wingless-type (Wnt) signaling pathways; and epidermal growth factor receptor (EGFR) ligands such as transforming growth factor alpha also have been implicated in the formation of metastatic bone lesions.² In the second step, the metastatic cancer cells produce factors that stimulate osteoblasts. The osteoblasts express factors that result in the activation of osteoclasts such as RANKL.² The presence of membrane-bound RANKL is an important step in the activation of osteoclast precursors. In the third step, RANKL promotes maturation of the osteoclast precursors into active multinucleated osteoclasts. Mature osteoclasts subsequently form a seal on the mineralized bone matrix surface and, through acidification and the secretion of acidophilic proteases, mediate the process of bone resorption.² Osteoclasts are major mediators of angiogenesis in the bone microenvironment via the regulation of the bioavailability of vascular endothelial growth factor-A (VEGF-A). In addition, bone destruction results in the release of transforming growth factor beta (TGF-β), which causes further production of PTHrP, completing the vicious cycle.²

Bone marrow acidosis has also been explored as a contributing factor to the physiology of bone metastasis. Tumor acidosis has been associated with tumor invasion in various types of cancer.⁷ The tumor microenvironment has an abnormal pH, which is thought to be mediated by ion-proton pumps expressed in tumor cells and in normal cells. Of these, vacuolar H⁺-adenosine triphosphatase is thought to be the most important because it is expressed on osteoclasts and tumor cells.¹ Various subunits of this adenosine triphosphatase have been implicated in aggressive forms of breast carcinoma and melanoma as well as in osteoclasts.⁷ Invadosomes also have a role in this process. Invadosomes are adhesive molecules that degrade extracellular matrix. Invadosomes form in cancer cells, whereas podosomes form in their normal osteoclast counterparts.⁷ In the acidic environment, invadosomes are able to break down basement membranes and allow for enhanced motility of tumor cells, enabling them to enter the bone microenvironment. In addition, the podosomes of osteoclasts are central for the extrusion of protons and proteases through the ruffled border into Howship lacunae. Within Howship lacunae, the acidic microenvironment allows for the dissolution of hydroxyapatite, and inorganic and organic bone matrix.⁷ These processes, which result in increased acidosis within the bone marrow, are thought to allow tumor cells to take hold and flourish within the bone microenvironment.

Hypoxia is commonly seen in solid tumors, and with it comes an associated risk for poor prognosis and metastasis.⁸ Research has shown that hypoxia encourages metastasis via transition from an epithelial phenotype to a mesenchymal phenotype, dormancy of tumor cells, angiogenesis, development of cancer stem cell-like phenotypes, and the release of extracellular vesicles.⁸ Hypoxia is thought to have a role in bone destruction by upregulating RANKL, thus resulting in osteoclast formation; hypoxia is also thought to inhibit the differentiation of osteoblasts. This combination increases
bone destruction and is thought to create a microenvironment conducive to metastasis.⁴ Despite these advances, the relationships are not entirely clear in many of these mechanisms and more research is warranted.