Introduction

Certain solid tumors, such as breast, prostate cancer and lung cancer, tend to metastasize to bone [1, 2]. When skeleton metastases happen, usually, there are two types of lesions: osteoblastic or osteolytic, which result from an imbalance between osteoblast-mediated bone formation and osteoclast-mediated bone resorption [3]. Osteoblastic bone metastases are usually seen in prostate cancer. In contrast, many solid tumors, including cancers of breast, lung, thyroid and renal, tend to form osteolytic metastases which are much common than osteoblastic lesions. It must be noted that both bone resorption and formation can be observed in most cancer types, with patients exhibiting both components to different degrees [1]. This difference in bone lesions indicates the different underlying mechanisms in forming osteoblastic or osteolytic lesions, leading to different treatment targets. In addition, there are many differences in clinical features between two bone metastases. In this review, differences in clinical features and molecular characteristics between osteoblastic bone metastases and osteolytic bone metastases will be discussed to help clinicians to understand the underlying mechanisms, behaviors and therapies in development and currently available for bone metastases.

Differences of bone metastases in clinical features

Incidence of bone metastases

The incidences of dysregulated osteolysis and abnormal bone formation both vary according to the primary tumor. A postmortem examination in different cancers showed that up to 70 percent of patients with advanced breast or prostate cancer suffered from this complication [4]. Other cancers including thyroid, kidney and lung also bear an incidence of 30–40 % to form bone metastases. Gastrointestinal cancer rarely produces bone metastases with an incidence only <10 % [4] (Table 1).

Table 1 Differences of osteoblastic bone metastases and osteolytic bone metastases in clinical features
Full size table

Prognosis

Notably, once tumor metastasizes to bone, it is incurable. The survival time after diagnosis varies among different tumor types. A study that analyzed the prognostic factors for survival in patients with spinal metastases demonstrated that the average survival time for patients with prostate, thyroid, breast and rectal cancer was longer than that of stomach and lung cancer, and it also suggested that type of primary tumor was the most powerful prognostic factor [5]. The 5-year survival rate is only 33 % for patients with distant prostate cancer metastases and 26 % for those with breast cancer [6]. Median survival of breast cancer or prostate cancer after the diagnosis of bone metastases is approximately 2–3 years [7]. However, the median survival of lung cancer with bone metastases is 6–7 months [8]. Recently, a retrospective observational study showed that the median overall survival after diagnosis of prostate cancer with spinal metastases was 24 months with an estimated 1-year overall survival of 73 % [9]. The median survival time after diagnosis of breast cancer with bone metastases was 32 months [10].

SREs

Both osteolystic and osteoblastic bone lesions tend to metastasize to the axial skeleton, such as rib sternum, pelvis and vertebrae, leading to skeletal-related events (SREs), which mean pathologic fractures, spinal cord compression, severe pain requiring radiotherapy or surgery and hypercalcemia [1]. Among all tumor types, patients with breast cancer have the highest incidence of skeletal complications, approximately 70 % [11]. The incidence of SREs of prostate cancer is 50 %, close to lung cancer and other cancers. The most common SREs in all tumor types are radiation to bone and pathologic fracture.

Bone biomarkers

Many bone metabolism markers elevated in bone metastases are closely associated with disease progression and treatment efficacy [12]. Some of them may have potential prognostic value and help clinicians diagnose bone metastases, determine treatment and monitor efficacy. For example, the bone resorption markers N-telopeptide of type I collagen (NTX) and bone alkaline phosphatase (BALP) are associated with higher rates of death and SREs in prostate cancer bone metastases and also provide prognostic information in patients receiving zoledronic acid and denosumab [13, 14]. Similar results are also found in lung cancer and other solid cancers [15]. Although bone markers are divided into bone formation markers and bone resorption markers, it must be noted that both of them increase in most bone metastases cases (Table 1). Usually, bone resorption happens before bone formation. Therefore, bone resorption markers could be an earlier detection tool in bone metabolism than bone formation markers.

Differences of bone metastases in molecular mechanisms and therapeutic implications

Bone metastasis is a multi-step process, involving the interplay between tumor cells and the bone microenvironment where various signaling pathways and molecules participate in to form a vicious circle and promote tumor cell and bone metastases [16]. Tumor-induced osteoblastic and osteolytic activities are different and play different roles in supporting their growth and survival, leading to different therapeutic implications [16]. It is important to better understand molecular mechanisms to break the vicious cycle.

Molecular mechanisms and therapeutic implications in osteoblastic bone metastases in prostate cancer

Osteoblastic bone metastases in prostate cancer are caused by tumor-derived factors that lead to osteoblast proliferation, differentiation and bone formation. Compared to models of osteolytic metastases, osteoblastic models are rare. The mechanisms that determine a metastatic lesion being osteoblastic or osteolytic remain unclear.

Endothelin-1 (ET-1) has been suggested to be a central mediator of osteoblastic metastases which stimulates the new bone formation via the endothelin A receptor (ETAR) in mice and humans [17]. ET-1 has been found increased in patients with bone osteoblastic lesions, especially in androgen-independent advanced prostate cancers [18]. Downstream genes of ET-1 with possible roles in osteoblast function include IL-6, Wnt5a, connective tissue growth factor (CTGF) and receptor activator of nuclear factor kappa-B ligand (RANKL) [19]. Importantly, ET-1 significantly suppresses the DKK-1 which is a negative regulator of the Wnt signaling pathway [20]. ET-1 can also enhance the mitogenic effect of other growth factors such as insulin-like growth factor 1 (IGF-I), platelet-derived growth factor (PDGF) and epidermal growth factor (EGF) [21]. In a mouse model, the ETAR antagonist atrasentan prevents osteoblastic bone metastases and reduces skeletal morbidity [22, 23]. Clinical trials indicate atrasentan seems to prolong time to disease progression (TTP), prostate-specific antigen (PSA) and BALP increase, although TTP does not reach statistical significance compared to control group [24]. However, a recent clinical trial indicated that ZD4054, another ETAR antagonist, did not prolong OS or TTP in patients with castrate-resistant prostate cancer (CRPC) with bone metastases [25, 26]. The negative results were also found in a phase III clinical trial with combined ZD4054 and docetaxel [27]. Unexpectedly, these therapies are not proven efficacious in limiting disease progression. However, it is possible that these drugs could be used to combat the development of osteoblastic lesions where there is a clear osteoblastic event which may impact life quality, becoming progressively worse if not treated. Furthermore, identification of patient subgroups based on either clinical characteristics or biomarkers may help better define people most likely to benefit from ETAR-targeted therapy in future clinical trials.

As mentioned above, DKK-1 is a negative regulator of Wnt proteins which are tumorigenic and osteogenic potential [28]. Elevated levels of Wnt proteins play an important role in the development of osteoblastic metastases [29]. DKK-1 inhibits Wnt canonical signaling by binding to the lipoprotein receptor-related protein (LRP) receptor, leading to internalization of the receptor. Moreover, it has been reported that DKK-1 inhibits the secretion of osteoprotegerin (OPG) and induces RANKL, which are important for osteoclastogenesis [30]. One study showed that blocking DKK-1 expression in osteolytic PC-3 prostate cancer cells induced osteoblastic activity and, conversely, inducing DKK-1 in the mixed osteoblastic/osteolytic prostate cancer C4-2B cell line resulted in experimental osteolytic bone metastases [31]. DKK-1 might determine prostate cancer bone metastases transit from osteolytic to osteoblastic [32]. However, a recent study demonstrated that DKK-1 significantly promoted prostate cancer growth and increased the incidence of bone metastases which has also been confirmed in myeloma and breast cancer [33, 34]. The probable mechanism is DKK-1 increases phospho46 JNK by the Wnt noncanonical pathway. Different types of cancer and different stages of cancer may be intrinsically caused by differing effects of the Wnt pathway [32]. Another explanation is that DKK1 might bind other proteins and have distinct, unknown mechanisms of action. The role of DKK-1 in bone metastases is not fully understood.

Bone metastases of prostate cancer tend to be osteoblastic, while the osteolytic factor parathyroid hormone-related protein (PTHrP) is also highly expressed in prostate cancer. A proposed explanation is that PTHrP can also stimulate bone formation by activating the ETAR with NH2-terminal fragments of PTHrP which share strong sequence homology with ET-1 [35]. It has also been shown that there are strong anabolic responses to PTHrP fragments 1–20 and 1–23. The PSA, an important indicator in prostate cancer diagnosis and treatment, may inactivate the osteolytic effects of PTHrP [36].

Bone morphogenetic proteins (BMPs) are well recognized to promote bone growth by directly simulating differentiation of osteoblast precursors to mature mineral-producing osteoblasts [37, 38]. Several researches indicate an autocrine effect of BMPs on prostate cancer cells, promoting cell invasion and migration [39, 40]. BMP4 is found to promote osteogenesis in the progression of prostate cancer in bone by modulating BMP4-mediated paracrine signaling [41]. Prostate cancer also promotes osteoblastic activity through BMP-6 and BMP-6 antagonist, noggin, could be a novel strategy treatment of osteosclerotic bone metastases of prostate cancer [42]. Noteably, a recent study suggested that the combined application of RANK-Fc, a recombinant RANKL antagonist and RN, a retroviral vector expressing noggin effectively inhibited the progression of mixed osteolytic/osteoblastic prostate cancer lesions in bone [43]. RANK-Fc alone inhibited osteolysis, while RN alone inhibited the osteoblastic component in a mixed lytic/blastic lesion. As the role of RANKL in osteolytic bone lesions is widely studied, this study confirms the role of BMP mainly in promoting osteoblastic lesions. This study also indicates the different applications of bone-targeted drugs based on the type of bone metastases.

Other factors such as PDGF, IGF, urokinase-type plasminogen activator (uPA) are also implicated in increasing osteoblastic bone metastases [36, 44].

Molecular mechanisms in osteolytic bone metastases in breast cancer

Breast cancer cells secrete many osteotropic cytokines to stimulate osteolysis, such as PTHrP, tumor necrosis factor α (TNF-α), interleukins, leukemia inhibitory factor (LIF) and receptor activator for RANKL [45, 46].

PTHrP is an important osteoclast-activating factor which is elevated in 90 % of bone metastases samples [47]. PTHrP activates osteoclasts and promotes bone resorption by binding to its receptor present on osteoblasts [48]. PTHrP upregulates RANKL and downregulates OPG by interacting with parathyroid hormone receptor 1. Also, PTHrP can promote tumor cells proliferation by autocrine action and tumor angiogenesis [49, 50]. Linforth et al. demonstrated that co-expression of PTHrP in early breast cancer indicated poor prognosis, while a recent study found that PTHrP expression in primary tumor was associated with improved survival and fewer bone metastases [51, 52]. Antibodies against PTHrP are reported to suppress osteolytic bone metastases and humoral hypercalcemia of malignancy (HHM) [53].

Transforming growth factor-β (TGF-β) is a critical factor in the formation of bone metastases in breast cancer. TGF-β released by activated osteoclasts can further increase production of PTHrP through Smad-dependent and the p38 mitogen-activated protein kinase signaling pathway [54]. The elevated PTHrP increases RANKL to stimulate osteoclast formation and activity and promotes bone metastases. Subsequently, bone matrix factors are produced which in turn, influence cancer cells to maintain a vicious cycle. Expression of a TGF-beta ligand trap, which neutralizes TGF-beta1 and TGF-beta3 significantly, decreases osteolytic lesions in MDA-MB-231 breast cancer cells and also in vivo [55, 56]. Recently, Notch ligand Jagged1 which is an important mediator of bone metastases by activating the Notch pathway in bone cells could also be activated by TGF-β [57]. In addition, TGF-β can stimulate cyclooxygenase-2(COX-2) expression, leading to prostaglandin E2 (PGE2) production in MDA-MB-231 breast cancer cells [58]. The elevated PGE2 increases RANKL production through binding to its receptor EP4 on the surface of the osteoblasts, resulting in osteoclastogenesis [59].

A number of genes that are selectively upregulated in aggressive bone metastatic clones are identified by microarray analysis, including IL-11, chemokine receptor type 4 (CXCR4), CTGF and matrix metalloproteinase-1 (MMP-1) [60]. These genes are also contributed to bone metastases in breast cancer, and they are directly or indirectly regulated by TGF-β. TGF-β could also stabilize hypoxia-inducible factor-1 (HIF-1α) by inhibiting its degradation which promotes osteolysis by stimulating angiogenesis, osteoclastogenesis and inhibition of differentiation of osteoblasts. TGF-β1 biological signals are also produced by hypoxic metastatic cells [61–63]. Inhibition of TGF-β pathway could inhibit both bone and lung metastases in breast cancers [64–66]. Inhibitors targeting TGF-β pathway are mainly ligand traps, antisense oligonucleotides (ASO), small molecule receptor kinase inhibitors and peptide aptamers [66]. A pan-neutralizing anti-mouse TGF-β monoclonal antibody 1D11 reduces osteolytic lesions in vivo and vitro [68]. Type I receptor kinase inhibitor, Ki26894, YR-290 and a dual inhibitor of TβRI/II, LY2109761 also show their promising therapeutic roles in bone metastases in breast cancer. Although these targeted drugs result in a significant reduction in metastases in mouse models, it seems they do little on primary tumors. These data indicate a combined therapy may strengthen the efficacy of TGF-β-targeted drugs (Fig. 1).

Fig. 1
figure 1

Role of TGF-β in osteolytic breast cancer bone metastases. Drugs targeted TGF-β blocks TGF-β-regulated bone metastases

Full size image

Other factors produced by breast cancer cells can also induce osteoclastic activation such as IL family, IGFs which are also implicated its role in proliferation of metastatic breast cancer cells in the bone [45, 46].

Common bone-targeted drugs both in prostate cancer and in breast cancer

In addition to drugs targeted to specific molecules in osteolytic or osteoblastic bone metastases, there are common bone-targeted therapies in prostate cancer and breast cancer. Currently, bisphosphonates (BPs) are the standard therapy for the prevention and treatment of malignant bone disease [69]. The second-generation nitrogen-containing BPs (N-BPs) (e.g., zoledronic acid, pamidronate) have better effects in reducing SREs compared with the first-generation BP compounds (e.g., clodronate) [70]. Zoledronic acid significantly reduces skeletal complications in breast cancer with bone metastases [71]. In addition, recent clinical trials indicated that adding zoledronic acid to adjuvant endocrine therapy improved disease-free survival (DFS) and overall survival (OS), especially in premenopausal women with early breast cancer and in women in established menopause at trial entry, indicating its potential antitumor effects [72]. Also, clinical trials indicate that adding zoledronic acid to adjuvant endocrine therapy improves bone mass density and prevents bone loss caused by endocrine therapy [73, 74]. Data with other bone-targeted agents are limited. Oral clodronate seemed to reduce distant metastasis in older postmenopausal women but no significance in OS and DFS in total population [75]. Also, no statistically significant differences in DFS or OS between the ibandronate and placebo-treated groups were seen. As in breast cancer, zoledronic acid also decreased SREs in hormone-refractory metastatic prostate cancer, although it did not improve OS and DFS [76]. However, pamidronate disodium failed to reduce bone pain or SREs compared with placebo [77]. Similarly, clodronate study did not demonstrate the efficacy for palliation of bone metastases symptoms in CPRC, although another trial of oral clodronate showed that OS was statistically significant in the men who received clodronate with long-term follow-up [78]. Recently, a meta-analysis compared the efficacy of BPs (zoledronic acid, ibandronate and pamidronate) in metastatic breast cancer, prostate cancer and multiple myeloma. As a result, zoledronic acid was clearly the best bisphosphonate for all three tumor types [79]. Another important bone-targeted drug is denosumab which is a fully humanized monoclonal antibody against RANKL approved by FDA for prevention of SREs. Recently, several clinical trials compared denosumab with zoledronic acid both in breast cancer and in prostate cancer and other solid tumors. Denosumab seems to be more effective in delaying or preventing SREs than zoledronic acid in patients with bone metastases from solid tumors potentially represents a novel treatment option for bone metastases [80, 81].

Conclusions and future perspectives

The cross-talk between bone microenvironment and different cancer cells activates and promotes various molecules and pathways, leading to the formation of different bone lesions, resulting in diverse clinical features.

Nowadays, targeted therapy is emerging as a new option in cancer treatment. Although the exact underlying mechanisms are not fully clarified in either osteoblastic or osteolytic bone lesions, it can provide therapy implications for bone metastases according to their unique molecules. In addition, although now common bone-targeted drugs such as BPs and denosumab play a main role in treating bone metastases, further research is needed to determine the optimal duration of treatment or treatment interval for patients who receive these medications for years. The potential benefits and adverse effects of these drugs should also be estimated according to more clinical trials. More specific targets based on cancer types may be also needed to improve current targeted therapies outcomes and reduce adverse effects. Further research into the molecular mechanisms of bone remodeling and bone metastases will provide more effective therapies with greater clinical efficacy than the therapies currently available.