Bone Disease in Multiple Myeloma

The pathogenesis of OBD in MM is primarily associated with generalized OC activation. BM biopsies from MM patients show a correlation between tumor burden, OC numbers, and resorptive surface [22, 23]. Furthermore, OC activity has positive correlation with disease activity [24, 25]. The main cytokines involved in OC differentiation and activity in MM OBD are RANKL/OPG, decoy receptor 3 (DcR3), CCL3 (also known as macrophage inflammatory protein (MIP)-1α), MIP-1β, tumour necrosis factor-alpha (TNFα), interleukin (IL)-3, IL-6, IL-11, Stromal derived factor-1 alpha (SDF-1a), B-cell activating factor (BAFF), activin A, and VEGF.

MM cells stimulate OC differentiation by producing IL-3 [26], DcR3 [27, 28], CCL3, MIP-1β [29‐31], VEGF [32], TNFα, [33, 34] and RANKL [35‐38]. MM cells also adhere to BMSCs via very late antigen (VLA)-4 and vascular cell adhesion molecule (VCAM)-1 interaction leading to the secretion of cytokines including RANKL, SDF-1a, IL-6, BAFF, VEGF, and activin A which in turn positively affect OC differentiation and activation [9, 14, 32, 39‐43]. MM cells stimulate not only RANKL expression, but also inhibit OPG expression, leading to an increase in RANKL/OPG ratio in BMSCs and OBs which in turn strongly stimulate OC differentiation [24, 44]. In addition to BMSCs and OBs, MM cells also stimulate CCL3 and pro-osteoclastogenic cytokine, IL-11 in osteocytes [45]. Moreover, OCs secrete CCL3 and activin A, and stimulate OC differentiation and activation by themselves [9, 31]. BM macrophages stimulated by IL-3 also secrete activin A [46]. All these cytokines stimulate OC differentiation and activity, and contribute to the development of MM OBD.

Besides OCs, BMSCs and OBs derived from BMSCs, play an important role in the development of OBD in the presence of MM cells. MM cells stimulate OC differentiation directly by secreting OC-activating factors (OAFs) and, indirectly, by stimulating OAFs secretion such as RANKL, Activin A and VEGF in BMSCs and OBs [14, 35, 36, 59, 60]. Adhesion of MM to BMSCs leads to RANKL and VEGF secretion by BMSCs via p38 MAPK activation [59, 60]. Moreover, the sequestosome 1, p62 is an upstream regulator of p38 MAPK and NF-κB signaling pathway, activated in BMSCs by MM cell adhesion. Inhibition of p62 in BMSCs represses OC differentiation and MM cell proliferation [61]. These data suggest that p62 is a novel promising target in MM OBD. Adhesion of MM to BMSCs and immature OBs also leads to IL-6 secretion via NF-κB signaling [42, 43, 62] and X-box-binding protein 1 (XBP1) signaling [63] pathway. IL-6 stimulates MM cell proliferation and inhibition of MM plasma cell apoptosis [64] in addition to OC differentiation. Moreover, adhesion of MM cells also stimulates BAFF expression in BMSCs via NF-κB signaling [41]. BAFF is a MM cell survival factor and it rescues MM cells from apoptosis induced by IL-6 deprivation and dexamethasone via activation of NF-κB, phosphatidylinosiol-3 (PI-3) kinase/AKT, and MAPK pathways in MM cells and induction of a strong upregulation of Mcl-1 and Bcl-2 antiapoptotic proteins [65, 66]. Secreted IL-6 and BAFF also stimulates the serine/threonine kinase Pim-2 expression in MM cells via activation of NF-κB and JAK2/STAT3 pathway, resulting in MM cell survival [67]. Furthermore, MM cells stimulate activin A expression in BMSCs via Jun N-terminal kinase-dependent (JNK) activation [9]. Importantly, high activin A levels in MM patients are associated with advanced bone disease and advanced features of MM [68]. Secreted Activin A inhibits OB differentiation in addition to the growth stimulatory effects on OCs. MM cells also stimulate Pim-2 expression in BMSCs/OBs by IL-3, IL-7, TNF-a, TGF-β and activin A secretion, and inhibit OB differentiation [69].

Osteocytes act as main regulators of bone homeostasis in healthy bone [16]. A recent study showed that MM patients have a significantly lower number of viable osteocytes than healthy controls, and that osteocyte death correlates with the presence of bone lesions [45]. Besides a lower number of viable osteocytes has been observed in the MM patients, no significant difference in the expression of sclerostin, an osteocyte marker, in biopsies of MM patients bone and healthy controls osteocyte was observed [45]. On the other hand, higher circulating levels of sclerostin have been found in newly diagnosed MM patients as mentioned before [74]. These data suggest that there might be other alternate sources of sclerostin in addition to osteocytes in MM. Moreover, MM cells stimulate osteoclastogenic cytokines, CCL3 and IL-11 expression in pre-osteocytes leading to increased OC differentiation [45]. Further investigations regarding the role of osteocytes in MM OBD are underway.

Bisphosphonates represent the standard of care for MM OBD. Nitrogen-containing bisphosphonates such as pamidronate (PAM) or zoledronic acid (ZA), more potent than PAM, reduce osteoclast activity through inhibiting farnesyl pyrophosphate synthase (FPPS) [77]. Bisphosphonates prevent OB and osteocyte apoptosis with a different mechanism from the effect on OCs [78‐80]. Bisphosphonates induce ERK activation without nuclear accumulation in OBs and osteocytes. Activated ERK stimulates p90RSK and induces phosphorylation of the cytoplasmic substrates, BAD and C/EBP, which are required for OB and osteocyte survival [81].

Bisphosphonates have a well-established role in the treatment of osteoporosis [82, 83] and metastatic bone involvement from solid tumors [84‐86]. In MM, treatment with bisphosphonate significantly reduces pain related to OBD and prevents SREs. Monthly infusion of PAM reduces bone pain and SREs compared with placebo [87]. PAM also significantly improved quality of life, with decreases in pain scores seen within a month. Moreover, Major et al. reported that ZA was superior to PAM for the treatment of hypercalcemia of malignancy including MM [88] although Rosen et al. reported the efficacy of ZA in preventing SREs in MM was comparable to that of PAM [84].

In addition to their role in OBD, bisphosphonates may also have an antitumor effect. The Austrian Breast and Colorectal Cancer Study Group 12 (ABCSG-12) trial showed that the administration of zoledronic acid every 6 months for 3 years reduced the risk of disease recurrence in estrogen-receptor—positive breast cancer patients [89] although no improvement was seen in the rate of disease-free survival in another study [90]. In MM, The MRC Myeloma IX trial compared ZA and oral clodronate in newly diagnosed patients and found that ZA reduced mortality by 16 % and increased median overall survival from 44.5 to 50.0 months (P = 0.04) [91]

Denosumab is an OC inhibitor that plays a role in the supportive care of MM OBD. It is a monoclonal antibody, given subcutaneously, that inhibits OC activity through targeting RANKL. Denosumab is approved for increasing bone density in patients with osteoporosis and for preventing SREs in patients with metastatic bone disease [100]. It has been recently reported that denosumab causes osteosclerosis [101], and hypercalcemia has been observed following discontinuation of denosumab [102] in children. In MM, although a favorable trend was observed, denosumab was equivalent to ZA in delaying time to first on-study SRE [103]. Denosumab is not currently FDA approved for use in patients with MM; a larger, ongoing phase III study (ClinicalTrials.gov identifier: NCT01345019) is comparing it with ZA in this disease setting.

OPG is a decoy receptor for RANKL, and it blocks OC differentiation and activation. In MM patients, BM plasma levels of OPG is decreased compared with normal volunteers and patients with MGUS [35]. Importantly, low levels of OPG in serum correlate with advanced OBD in MM [55]. Treatment with OPG or OPG-like molecules prevented both bone destruction and MM growth in vivo [36, 56]. A Phase I study of a recombinant OPG construct (AMGN-0007) was conducted in MM patients with OBD, and decreased NTX/creatinine levels was observed [104].

The CCL3/CCR1 pathway stimulates OC differentiation, MM cell survival and migration, and inhibits OB differentiation suggesting that CCL3/CCR1 is a relevant target in MM OBD. Both antisense sequence and neutralizing antibody against CCL3 effectively inhibited tumor growth and restored bone remodeling in a mouse model of MM OBD [15, 105]. Similar results have been shown with a clinical grade small molecule CCR1 antagonist, MLN3897 (Millennium Pharmaceuticals) [51]. In addition to these molecules, several CCR1 antagonists were evaluated for MM OBD [106, 107]. Future clinical trials using CCR1 inhibition strategies in patients with MM OBD will help to confirm these promising preclinical results.

In MM, BAFF is expressed by monocytes, macrophages, dendritic cells, T cells, neutrophils, MM cells, and OCs [65, 108‐111]. BAFF is a MM cell survival factor and rescues MM cells from apoptosis induced by IL-6 deprivation and dexamethasone via activation of NF-kB, PI-3 kinase/AKT, and MAPK kinase pathways and induction of a strong upregulation of the Mcl-1 and Bcl-2 antiapoptotic proteins [65]. In vivo—neutralizing antibodies against BAFF (LY2127399, Eli Lilly) significantly inhibit tumor burden and, importantly, reduce OBD and OC differentiation in preclinical setting [66]. On the basis of these results, a clinical trial combining BAFF-neutralizing antibody with proteasome inhibitor, bortezomib is currently ongoing, preliminary results from Raje et al. reported the treatment was well tolerated and 22 of the 48 patients enrolled achieved a partial remission or better (https://​ash.​confex.​com/​ash/​2012/​webprogram/​Paper52052.​html).

Activin A is secreted by BMSCs and OCs in MM OBD. Activin A stimulates OC differentiation and inhibits OB formation in MM OBD. Activin A can be targeted by a chimeric antibody RAP-011 (Acceleron Pharma), derived from the fusion of the extracellular domain of type II activin receptor (ActRIIA) and the constant domain of the murine IgG2a [112]. RAP-011 enhances OB mineralization and increases bone density in an osteoporotic mouse model. In MM, RAP-011 reversed OB inhibition, improved MM bone disease, and inhibited tumor growth in an in vivo humanized MM model [9]. In human, ACE-011 which is the humanized counterpart of RAP-011 effectively decreased bone resorption markers, C-terminal type 1 collagen telopeptide (CTX) and TRACP-5b and increased bone formation marker, serum levels of bone-specific alkaline phosphatase (BSALP) in postmenopausal women [113]. It has been shown in vitro that lenalidomide, a well known and approved treatment strategy for relapsed MM, stimulates activin A secretion on BMSCs via an Akt-mediated increase in JNK signaling [14]. Clinical trials for ACE-011 with Lenalidomide + Dexamethasone are ongoing and evaluating its role in MM (ClinicalTrials.gov identifier: NCT01562405).

Dkk-1 plays one of the key roles in mediating OB inhibition in MM [71]. Therefore, treatment strategies to block Dkk-1 activity have been developed. In vitro assays show that inhibition of Dkk-1 via a specific neutralizing antibody promotes OB differentiation and function and reverses the negative effect of MM cells on OB differentiation [114, 115]. Moreover, in vivo studies using both murine and humanized murine models of MM-induced bone disease showed increased bone formation, OB numbers, and improvement of osteolytic lesions by Dkk-1 inhibition [115‐117]. Importantly, blocking Dkk-1 also resulted in reduction of tumor growth, mainly as an indirect effect via modification of the tumor microenvironment [115]. Therefore, Dkk-1 inhibition via a neutralizing antibody restores bone homeostasis and may have an inhibitory effect on tumor growth. Currently, ongoing clinical trials combining Dkk-1 neutralizing antibody and bisphosphonates will test these promising preclinical results. In particular, ZA in combination with the proanabolic agent BHQ880, a fully human anti-Dkk-1 monoclonal antibody, has being studied in a phase I clinical trial (ClinicalTrials.gov identifier: NCT00741377). BHQ880 was also tested in a phase II clinical trial in smoldering MM (ClinicalTrials.gov identifier: NCT01302886) and preliminary results showed that BHQ880 significantly stimulated the vertebral strength by qCT from a baseline of 3 % (P = 0.002) (https://​ash.​confex.​com/​ash/​2012/​webprogram/​Paper48568.​html).

Several studies have already demonstrated the importance of sclerostin in osteoporosis [118, 119], and inhibition of sclerostin represents an important strategy in the treatment of bone conditions with high catabolism. In fact, clinical trials with sclerostin neutralizing antibodies, romosozumab and blosozumab for the treatment of postmenopausal osteoporosis are ongoing and preliminary results have shown increase of bone mineral density [120‐122]. In MM, higher circulating levels of sclerostin have been found in newly diagnosed MM patients, and it correlated with MM disease stage and fractures [74]. These data underscore the importance of targeting sclerostin for treatment of MM OBD. However, the source and role of sclerostin in MM OBD still remains unclear. Further studies about sclerostin’s role in MM and application of sclerostin neutralizing antibody to MM OBD are expected.

Bortezomib is a proteasome and NF-kB signaling pathway inhibitor with potent anti-MM activity. Bortezomib also inhibits MM-BMSC interactions, impairs osteoclastogenesis, and stimulates mesenchymal stem cell differentiation to OB and, therefore, actively modulates bone remodeling in MM [123‐125]. The anabolic effects of bortezomib are associated with Runx2 upregulation via inhibition of proteasomal degradation. Runx2 is a critical transcription factor in early OB differentiation and modulates the expression of the OB-specific transcription factor osterix [125, 126]. The anti-OC effects of bortezomib are mediated by p38 inhibition at early time points and, at later time points, by impairment of NF-kB signaling and AP1 inhibition [123]. These effects have been confirmed in the clinical setting by upregulation of OB activation markers (BSALP and osteocalcin) and downregulation of bone resorption markers (CTX and TRACP-5b) as well as decrease of Dkk-1 and sRANKL in patients treated with bortezomib [127].

In contrast to bortezomib, carfilzomib is a new proteasome inhibitor that is associated with a very low incidence of peripheral neuropathy. Carfilzomib is a structural analog of the microbial natural product epoxomicin that selectively inhibits the chymotrypsin-like activity of both the constitutive proteasome and the immunoproteasome [128]. It was recently approved in July 2012 for patients with MM experiencing disease progression after prior therapy with bortezomib and an immunomodulatory drug. Carfilzomib strongly stimulates OB calcification and inhibits OC differentiation in addition to the antitumor effect [129‐131]. Moreover, we showed carfilzomib reversed OB inhibition, improved MM bone disease, and inhibited tumor growth in an in vivo disseminated MM model [131]. Interestingly, we could not see upregulation of OB differentiation marker in OBs in the presence of higher concentration of carfilzomib although the concentration of carfilzomib strongly stimulates OB calcification. Further studies are necessary to evaluate the detailed mechanism of carfilzomib effect on OBs.

Bruton’s tyrosine kinase (Btk) belongs to the Tec family of tyrosine kinases. The activation of Btk regulates B-cell development and antibodies production. Thus, Btk pathway is a potential therapeutic target in a variety of B-cell malignancies, including Waldenström’s macroglobulinemia, diffuse large B-cell lymphoma, follicular lymphoma, mantle cell lymphoma and chronic lymphocytic leukemia [132]. In MM, we showed that Btk inhibitor, CC-292 strongly inhibits OC activity and improves MM OBD [131]. It decreased only INA-6 MM cell line viability in higher concentration, however, had negligible direct in vitro effects on other MM cells viability or in animal models. On the other hand, the other Btk inhibitors, PCI-32765 (ibrutinib) and LFM-A13 have shown to display some antitumor effect in MM xenograft mouse model when INA-6 MM cells were used [133, 134]. More investigations are needed to reveal the role of Btk inhibitors in the MM OBD.

MM cells upregulate Pim-2 expression in BMSCs/OBs and inhibit OB differentiation [69]. Meanwhile, IL-6, produced by BMSCs, BAFF, and APRIL, produced by OCs, stimulate Pim-2 expression in MM cells via activation of NF-κB and JAK2/STAT3 pathway, resulting in MM cell survival [67]. Importantly, Pim inhibitor prevents bone destruction while suppressing MM tumor burden in MM model mouse [69]. Pim-2 may become a new target for not only MM OBD but also MM treatment.