Bone metastases: When and how lung cancer interacts with bone

Abstract

Bone metastasis is a common and debilitating consequence of lung cancer: 30%-40% of patients with non-small cell lung cancer develop bone metastases during the course of their disease. Lung cancer cells find a favorable soil in the bone microenvironment due to factors released by the bone matrix, the immune system cells, and the same cancer cells. Many aspects of the cross-talk among lung tumor cells, the immune system, and bone cells are not clear, but this review aims to summarize the recent findings in this field, with particular attention to studies conducted to identify biomarkers for early detection of lung cancer bone metastases.

Key Words: Lung cancer, Bone metastases, Osteoclast, T cell, Bone microenvironment

Core tip: This review reports current knowledge on the cross-talk among lung tumor cells, the bone microenvironment, and the immune system, that lead to bone metastasis.

INTRODUCTION

Lung cancer is the most common tumor worldwide, and in Europe alone, it is responsible for the 20.6% of cancer mortality[1]. It has poor overall survival rates: the majority of advanced stage lung cancer patients die within 18 mo from diagnosis[2]. The predominant form of lung cancer is non-small cell lung cancer (NSCLC), and is responsible for 80%-85% of all lung tumors[3].

Lung cancer frequently metastasizes to bone, with 36% of patients presenting with bone lesions at autopsy[4] and 22%-60% showing bone marrow micro-metastases[5]. The early diagnosis of NSCLC is difficult, and 30%-40% of patients with NSCLC develop bone metastases during the course of their disease[6,7]. The prognosis of lung cancer patients with bone metastases is poor, with a median survival time from detection of lesions measured in months[4]. Even though NSCLC patients have a relatively short survival time, a large percentage develop skeletal-related events (SREs), thus clinicians show particular attention to the management of bone metastases to prevent debilitating skeletal complications. Indeed, bone metastases are often not diagnosed in NSCLC patients until they cause pain and SREs[8], but once a patient has a first SRE, he is likely to experience subsequent events, leading to debilitating bone lesions and a reduction in the quality of life. Lung cancer bone metastases normally appear as areas of radiolucency and are osteolytic, with poor margination, no matrix and cortical destruction[9]. They commonly affect the spine, ribs, pelvis and proximal long bones. At an early stage, bone metastases may occur easily at an axial bone through the vertebral vein system, then at appendicular bone in more advanced stages of the disease[10]. A unique feature of this tumor is the ability to spread to the bones of the hands and feet. This is probably due to the ability of lung cancer to shed malignant cells directly into the arterial blood flow, from where they can be seeded far and wide.

In this review, I summarize the current knowledge on the cross-talk among lung tumor cells, the bone microenvironment, and the immune system, that lead to bone metastasis formation.

BONE MICROENVIRONMENT IS A FERTILE SOIL FOR DORMANT AND PROLIFERATING TUMOR CELLS

Tumor dissemination is an early event and is not related to the size of the tumor mass, supporting the idea of concomitant progression of the primary tumor and metastasis[11]. Several studies have shown that disseminated cells, shed from a primary tumor, may lie dormant in distant tissues for long periods before they can be activated to form metastases. The skeleton has a large surface area and a microenvironment conducive and protective to tumor cell growth. Indeed, disseminated tumor cells (DTCs) can persist in a quiescent state in the bone marrow of cancer patients for years, particularly if they lie in the hematopoietic stem cell (HSC) niche[12], which expresses adhesion molecules and secretes factors contributing to tumor cell dormancy. At present, it is not clear how dormancy is induced or what leads to activation of the dormant cells. One hypothesis is that molecules that induce HSC dormancy are likely to induce dormancy of metastatic tumor cells[13]. For instance, osteoblasts (OBs) which are a crucial component of the stem cell niche, express stromal derived factor-1 (SDF-1) and annexin-II (Anxa2), that attract both HSCs and cancer cells expressing their receptors, CXCR4 and Anxa2r, respectively[14]. In this way, the niche attracts, protects and induces dormancy of DTCs, which may be released from quiescence and grow as a consequence of a stimulatory microenvironment, like bone. DTCs detected in bone marrow of breast cancer patients exhibited features of mesenchymal and cancer initiating cells (CICs)[15]. These latter cells are responsible for both primary tumor generation, maintenance, recurrence and drug resistance[16,17]. In lung cancer, a subset of cells expressing CD133 and CXCR4 has been shown to have CIC features and to be essential for tumor metastasis formation[18]. Moreover, chemotherapy may select resistant CICs[19], as also demonstrated by Bertolini et al[20] who showed that lung CICs developed resistance to cisplatin. The monitoring of DTCs is a potential method to follow the dissemination of tumors, but a recent work reported that in NSCLC patients, DTC detection is not particularly useful to determine the presence of the disseminated disease in its early stages[21].

Bone has physical properties such as low pH, hypoxia and high levels of extracellular calcium[22], which sustain the proliferating tumor cells that secrete high quantities of lactic acid, creating a local area of bone with a low pH, which stimulates osteoclasts (OCs) activity. On the other hand, tumor cells grow better with a low pH and release proteolytic enzymes, that maintain the low pH, thus tumor expansion is perpetuated[23]. The active bone resorption causes an increase in extracellular calcium level that stimulates the calcium-sensing receptors on surrounding cells and tumor cells, leading to an increased secretion of parathyroid hormone-related peptide (PTHrP), a potent stimulator of OCs[24,25]. Bone is a hypoxic tissue, thus it promotes the ability of cancer cells to grow under hypoxic conditions, stimulating the expression of hypoxia inducible factor-1 (HIF-1). HIF-1 activates the transcription of target genes which encode proteins that play important roles in many critical aspects of cancer biology[26]. For instance, HIF-1 is involved in many steps of breast cancer metastasis, including metastatic niche formation, recruitment of bone marrow-derived cells to the metastatic niche, OC formation and OB inhibition[27-29].

FACTORS PRODUCED BY BONE AND TUMOR CELLS MODULATE OC ACTIVITY

The interaction of tumor and bone cells induces the formation of a vicious cycle leading to bone metastases[30]. Indeed, tumor cells disrupt normal bone remodeling, a perfectly balanced activity of bone resorption by OCs and bone formation by OBs. Bone resorption induced by cancer cells creates a physical space for tumor expansion, and it induces the release of growth factors and cytokines supporting tumor metastasis[31].

Bone marrow produces factors, such as SDF-1, which attracts cancer cells expressing the chemokine receptors, CXCR4 and CXCR7, thus it represents a favorable soil for secondary tumor localization[32,33]. Moreover, activated OCs resorb bone and release growth factors enmeshed in the bone matrix, such as bone morphogenetic proteins, transforming growth factor-β (TGF-β), insulin-like growth factor (IGF), fibroblast growth factor and others that stimulate the growth of metastatic tumor cells in bone, and the production and release of bone resorbing factors from tumor cells[34,35]. In particular, TGF-β stimulates HIF-1 signaling within bone, contributing to the vicious cycle driving the development of metastatic osteolysis[36]. Malignant cells produce cytokines, such as interleukin-6 (IL-6), IL-8 and IL-11, which activate OCs[37]. IL-6 can stimulate tumor cell proliferation, migration and invasion of lung and other tumors[38,39], and it may induce bone resorption or formation according to the interactions with other factors like PTHrP, IL-1 and receptor activator of nuclear factor-kB ligand (RANKL). IL-8 stimulates OC maturation through binding with its receptor CXCR1 expressed on OC precursors[40]. IL-11 stimulates osteoclastogenesis and it has been reported to be a predictive factor for development of osteolytic bone metastasis[41]. Cancer cells also secrete molecules such as parathyroid hormone, PTHrP, prostaglandins, activated vitamin D and TNF, which stimulate RANKL expression on OBs and bone marrow stromal cells[42]. RANK, the RANKL receptor, is expressed by solid tumors, with high concordance between bone metastasis and corresponding primary tumors[43]. The activation of the RANK/RANKL axis leads to an increase in OC number, survival and activity, and it may induce migration and homing of tumor cells to the RANKL-rich bone microenvironment[44]. In clinical practice the use of target therapy to reduce bone metastases gave encouraging results: the fully human monoclonal antibody directed against RANKL reduces bone metastases in breast and lung tumors[45-47]. Nuclear factor-kappa B is the downstream target of RANKL and it is also activated in lung cancer cells by epidermal growth factor receptor (EGFR), implicating this pathway as pivotal in lung cancer bone metastases[46]. EGFR-tyrosine kinase inhibitors (TKIs), such as erlotinib, show dramatic anti-tumor activity in a subset of NSCLC patients with an active mutation in the EGFR gene. Some lung cancer patients with wild type EGFR respond to EGFR-TKIs, thus EGFR-TKIs have an effect on host cells as well as tumor cells, preventing bone metastases by affecting the host microenvironment irrespective of its direct effect on tumor cells[48].

PTHrP promotes OC bone resorption[49-51], and it stimulates OBs and stromal cells to express RANKL, which induces OC maturation[52,53]. Among the inhibitors of PTHrP, recent data report that the microRNA miR-33a is downregulated in lung cancer cells, and it directly targets PTHrP leading to decreased osteolytic bone metastasis[54]. Indeed, the downregulation of PTHrP, induced by miR-33a, causes a decreased secretion of IL-8, with consequent reduction in OC differentiation and bone resorption[54]. Novel data derived from a pre-clinical model of NSCLC reports that miR335 inhibits the expression of RANKL and IGF-1 receptor by lung cancer cells, thus skeletal metastasis is reduced[55].

In the bone marrow, a direct activation of osteolysis by cancer cells has been shown through the interaction between Notch and Jagged. Jagged1, a downstream mediator of the pro-metastatic TGF-β, promotes tumor growth through stimulation of IL-6 production from OBs, and directly activates OC differentiation[56]. Moreover, Jagged is overexpressed by bone metastatic tumor cells[57], whereas its receptor Notch is frequently expressed by progenitors and mature cells in the bone marrow[58]. In breast cancer, Notch-Jagged interactions activate biological responses in OCs and OBs, which promote both tumor invasion of bone and tumor cell growth in bone[56]. In NSCLC, the expression of Notch-3 receptor correlates with a poor prognosis and the stage of the disease[59].

ROLE OF IMMUNE SYSTEM CELLS IN REGULATION OF OCS AND TUMOR GROWTH IN BONE

Oxidative stress is implicated in the initiation and progression of lung cancer[60]. In particular, myeloid-derived suppressor cells (MDSC), which infiltrate different tumors, generate reactive oxygen species (ROS) and cytokines, that suppress host T cell responses, promoting tumor progression and metastasis[61]. Both the production of ROS and nitric oxide are involved in osteoclastogenesis, and bone marrow-derived MDSCs have been showed to be able to differentiate into OCs; thus, all these factors contribute to enhanced bone destruction in tumor osteolysis[62,63].

Direct involvement of T cells in regulating OC activity has been demonstrated in patients affected by multiple myeloma, lung cancer and other solid tumors with bone metastasis[64-67]. These bone metastatic patients showed an increase in circulating OC precursors compared with both healthy controls and cancer patients without bone metastases[66]. OC precursors differentiate into mature, multinucleated and bone resorbing OCs in vitro, without adding exogenous pro-osteoclastogenic factors, such as TNF-α and RANKL, which instead are released by T cells. Cell cultures of PBMCs derived from cancer patients without bone metastasis, and depleted of T cells, do not differentiate into OCs without adding macrophage colony stimulating factor and RANKL[65,66], confirming that T cells regulate osteoclastogenesis and play an important role in the cancer cycle of bone destruction.

Recently, it has been demonstrated that T cells are additional regulators of bone tumor growth. In particular, their activation diminishes bone metastases, whereas their depletion enhances them, even in the presence of zoledronic acid[68]. Indeed, some patients treated with anti-resorptive therapies develop further skeletal metastases, suggesting that T cells modulate bone tumor growth. Zoledronic acid can activate cytotoxic γ/δ-T cells and inhibit populations of myeloid-derived cells with T-cell-suppressor capabilities[69]. The anti-bone metastatic therapy based on the blockade of TGF-β at metastatic sites may locally activate an anti-tumor T cell response, because normally TGF-β, released in bone marrow by OC activity, inhibits T cell proliferation[70]. Tumor cell-derived IL-6, IL-1 and TGF-β can drive T-cell differentiation towards a Th17 secretory helper-cell phenotype able to induce RANKL production by OB and OC activation through IL-17 production[71]. All these data demonstrate the fundamental role of immune system cells in the control of bone metastatic disease.

SERUM MARKERS FOR EARLY DETECTION OF LUNG CANCER BONE METASTASES

The delayed demonstration of skeletal involvement may seriously affect survival, thus an early diagnosis of bone metastases is necessary. The early detection of asymptomatic bone disease due to lung cancer can be obtained through positron-emission tomography scans[72], but European guidelines recommend a bone scan only in the presence of bone pain[3], thus other systems for an early diagnosis are required. The sensitivity of common serum tumor markers is low, and they are used mainly for monitoring the efficacy of therapy and detection of recurrence. The use of serological markers is desirable, but unfortunately, the identification of potentially useful and specific biomarkers is difficult. According to data in the literature, serum markers of bone turnover may be able to determine the time to tumor progression, the metastatic potential, and the overall survival of NSCLC patients. Furthermore, they may contribute to a more accurate follow-up and tailored treatment options[73,74]. In particular, there is a statistically significant relationship between levels of biochemical markers of bone metabolism and clinical outcome: both N-telopeptide and bone-specific alkaline phosphatase levels were highly predictive of SRE recurrence, the time to a first SRE, and the occurrence of disease progression and death[7,75]. Also, the carboxy-terminal telopeptide of type I collagen and amino-terminal propeptide of type I collagen measurement can be useful in monitoring patients with NSCLC during follow-up, with the aim of detecting bone metastases early[76].

Among the molecules potentially involved in the pathogenesis of bone metastases from lung cancer, there is IL-7, which has been studied as serum marker for monitoring bone metastasis development in NSCLC patients. Indeed, IL-7 is an important regulator of the interaction between bone and immune system, and it has a role in bone homeostasis, in particular in bone loss in estrogen-deficient conditions[77,78], psoriatic arthritis[79] and periodontitis[80]. Other studies support an active role of IL-7 in promoting bone lesions from solid tumors[81,82] and multiple myeloma[83]. In culture of PBMCs derived from patients with bone metastases due to lung cancer and other solid tumors, IL-7 was mainly released by B cells, and it directly sensitized T cells to produce pro-osteoclastogenic factors, such as tumor necrosis factor-α and RANKL, which enhanced osteoclastogenesis[82,83]. Moreover, in lung cancer patients with bone metastases, IL-7 serum levels were found to be significantly higher than in non-bone metastatic patients and in healthy controls[84]. This increase in serum IL-7 directly depends on tumor production, and indeed strong IL-7 expression was detected in human tumor masses in a mouse model of bone metastases and in human bone metastatic biopsies[85]. Further studies on a large cohort of patients needs to be performed, but IL-7 could be very useful to monitor the progression and early development of lung cancer bone disease.

CONCLUSION

The current literature reports the existence of an important interaction among lung tumor cells, the bone microenvironment and immune system cells. Many factors are involved in this cross-talk, which may lead to the discovery of new biomarkers useful for early detection of lung cancer bone metastases.