Bone, a fertile soil for tumor development

Introduction

Bone is a common metastatic site for a number of malignancies, accounting for 80%–90% of metastasis from multiple myeloma, 70%–80% of metastasis from breast and prostate cancer, 30%–40% of metastasis from lung and kidney, and evidence suggests that thyroid, colon, and gynecological cancers also develop metastases in bone [ ]. Not only do the majority of tumors metastasize to the bone but other tumors such as osteosarcoma develop in this site, suggesting that bone provides a favorable environment for tumor growth . Although the exact mechanisms are not fully elucidated, it is believed that the constant turnover of bone matrix along with the rich milieu of growth factors present in bone provide a “fertile soil” attracting tumor cells to colonize this site and promoting the development of tumor growth in bone. There is strong evidence demonstrating that once in bone tumor cells stimulate increased resorption of the bone matrix, through activation of osteoclasts, resulting in the release of matrix-bound cytokines, including interleukin 6 (IL-6) and tumor necrosis factor α (TNFα), which feedback onto the tumor cells further stimulating their growth. These interactions between tumor cells and bone cells cause a positive feedback loop known as the vicious cycle of bone metastasis [ ]. In this book chapter we will discuss aspects of the bone microenvironment that make it an attractive site for tumor cell metastatic spread and/or growth. We will focus on why tumor cells specifically home to bone (osteotropism) as well as how interactions between tumor cells and different cell types in bone (including osteoclasts, osteoblasts, osteocytes, adipocytes, and immune cells) promote tumor growth in this metastatic site.

Osteotropism

It is clear that metastasis is not a random event. Different tumor types demonstrate specific tissue tropism (e.g., estrogen receptor–positive breast cancers, prostate cancers, and multiple myeloma preferentially metastasize to bone, whereas estrogen receptor–negative breast cancers, colorectal, and melanomas predominantly spread to lung). This tissue-specific tropism is not a new phenomenon, it was first described by Stephen Paget in 1889 when he proposed a “seed and soil” hypothesis in which he suggested that metastases could only be established if tumor cells seeded into environments that are permissive for their colonization and growth, thus providing a “fertile soil” [ ]. Evidence from preclinical models suggests that tumor cells are able to “prime” the bone making it more receptive to the arrival of metastasizing tumor cells. The bone matrix serves as an important reservoir, storing calcium and growth factors. When needed, the body stimulates osteoclasts to resorb bone and release these factors into the circulation. Evidence suggests that tumor cells at the primary site can also stimulate bone resorption in this way: Breast, prostate, lung, and colon tumor cells all secrete lysyl oxidase [ ] and secretion of this enzyme has been shown to stimulate osteoclastic bone resorption “priming” the bone microenvironment for the arrival of tumor cells [ , ] and causing the release of previously matrix-bound growth factors, including transforming growth factor β (TGFβ), insulin-like growth factors (IGFs), fibroblast growth factors (FGFs), platelet-derived growth factor (PDGF), and bone morphogenic protein (BMP), which not only have cancer-promoting properties but also increase the production and release of cytokines and other bone-resorbing factors from tumor cells [ , ]. Evidence from preclinical lung cancer and melanoma models have also indicated that the bone microenvironment can act as a premetastatic niche, priming distant organs to be more receptive to metastasizing tumor cells during the initial stages of the metastatic process. Vascular endothelial growth factor receptor 1 (VEGFR1)-expressing bone marrow–derived hematopoietic stem cells (HSCs) have been demonstrated to arrive at sites of future metastasis before the metastasizing tumor cells. It is thought that arrival of these HSCs leads to increased production of fibronectin and/or inflammatory chemoattractants in the future site of metastases [ , ].

Once in bone tumor cells lodge in the metastatic niche(s) which comprises of the endosteal niche (containing the hematopoietic stem cell niche and osteoblasts) and the perivascular niche (on the outside of blood vessels) [ ]. In metastatic cancers tumor cells are thought to spread to the metastatic niche(s) through chemotactic mechanisms whereby chemokines in bone attract tumor cells expressing chemokine receptors into this environment. Evidence from mouse models confirm that metastatic breast, prostate, and myeloma tumor cells are usually located in the endosteal and the perivascular niches, and indicate that most disseminated tumor cells associate with the endosteal niche, although special dynamics in bone make it impossible to differentiate between these niches with certainty [ , ]. Interactions between C-X-C chemokine receptor type 4 (CXCR4)/C-X-C motif chemokine 12 (CXCL12) are thought to be key factors in the homing and adhesion of cancer cells to these metastatic niches within bone [ ]. CXCR4 expressed on cancer cells mediates the attachment to CXCL12 (a homeostatic chemokine produced by the osteoblasts and the vasculature in bone) [ , , ]. Interactions between receptor activator of nuclear factor-κB (RANK) on tumor cells and receptor activator of nuclear factor-κB ligand (RANKL) on osteoblasts also play a role in tumor cell homing and adhesion to the bone metastatic niche [ ]. Importantly, the RANKL/RANK/osteoprotegerin (OPG) system is also a critical component of the “vicious cycle” that drives disease progression and interactions between tumor cells and bone cells either through direct contact or via secreted molecules are key to determining which tumor cells will remain dormant and which will outgrow into overt metastasis in bone.

Outgrowth of metastatic tumors in bone

When tumor cells arrive in the bone metastatic niches they either die by apoptosis, become dormant as single cells, or form micrometastasis which are maintained in a state of equilibrium (in which rate of proliferation is equal to the rate of cell death) by immune regulation. Tumor cells are often maintained in a quiescent or dormant state by the bone niches for long periods of time, before “reawakening” and proliferating into overt metastasis. Indeed, disseminated tumor cells can be detected in the bone of 30%–40% of breast cancer patients [ ] and approximately 20% of patients thought to have recovered from breast cancer relapse in bone up to 15 years later [ ]. Furthermore, it is believed that tumor cells can transit through the bone to form metastases in other organs [ ]. Mechanisms of dormancy/tumor outgrowth are not fully understood and detailed; up-to-date information can be found in other book chapters/review articles [ , ]. It is broadly accepted, however, that tumor cells are stimulated to develop into overt metastases through cellular processes that result in the expansion of the metastatic niche(s), stimulate proliferation of previously dormant disseminated tumor cells, and/or reduce anticancer immunity in bone. The proinflammatory cytokine Interleukin 1β (IL-1β) is involved in all of these processes and is therefore a fundamental driver of metastatic outgrowth in bone. In breast cancer models direct contact of tumor cells with bone cells (osteoblasts or bone marrow cells) results in increased expression and secretion of IL-1β from tumor cells with osteoblasts and the bone marrow [ ]. Increased IL-1β in the bone microenvironment stimulates proliferation of preosteoblasts and endothelial cells leading to expansion of the bone metastatic niches [ ]. In addition, IL-1β from the bone microenvironment activates proliferation of disseminated tumor cells through activation of Wnt-mediated mechanisms [ ] and increases the activity of bone resorbing osteoclasts resulting in the release of growth factors from the bone matrix helping further fuel the growth of tumor cells in bone [ ]. Pharmacological inhibition of IL-1 signaling has been shown to inhibit metastatic outgrowth in mouse models of breast and prostate cancer [ ] and improve survival of patients with early-stage multiple myeloma [ ] suggesting that this cytokine may be involved in the early stages of bone colonization from multiple tumor types. However, in osteosarcoma, although IL-1β inhibition has been shown to reduce symptoms of hyperalgesia in mouse models no reduction in tumor growth has been observed further suggesting a role for IL-1β in metastasis rather than primary tumor growth in this site [ ].