Keywords

Introduction

Metastatic osteosarcoma (OS) carries a poor prognosis and options for successful treatment and eventual cure are few. Despite dramatic progress in the 1970s and 1980s in the treatment of non-metastatic OS, the outcomes have not changed in several decades. The exact molecular mechanisms underlying drug resistance and development of metastatic disease remain unknown. Furthermore, the contribution of the organ microenvironment remains unexplored. Novel therapeutic approaches for OS lung metastasis and refractory/recurrent disease are desperately needed [1,2,3,4,5,6,7,8,9].

Similar to other cancer types, targeted therapy and immunotherapy are potential treatment alternatives which have yet to be fully evaluated in OS. Immunomodulatory agents have long been considered for OS as a way to enhance the immune response [2, 3]. In fact, several studies suggest that OS may be amenable to treatment with immune-based therapies including immune checkpoint inhibitors [4,5,6,7,8,9]. Furthermore, there are several ongoing clinical trials which focus on the use of targeted therapies for recurrent and refractory OS. These include denosumab (anti-NFκB ligand), glembatumumab vedotin (anti-glycoprotein NMB), dinutuximab (anti-GD2), sirolimus (mTOR inhibitor), and VEGFR inhibitors (apatinib, lenvatinib, cabozantinib). In the present chapter, we provide the rationale for an alternative combination therapy using a specific histone deacetylase (HDAC) inhibitpr, entinostat/Syndax 275 in combination with the nucleoside analog, gemcitabine for the treatment of OS.

Fas and the Fas Signaling Pathway

Fas (CD95) is a cell surface death receptor that belongs to the tumor necrosis factor receptor (TNFR) superfamily. Interaction of Fas with its cognate ligand, FasL (CD95L), induces apoptosis in Fas-expressing cells. Fas is expressed on several different cell types including tumor cells, whereas FasL expression is restricted to immune cells (activated T and NK cells) and privilege organs, such as the lung [10]. The Fas/FasL signaling pathway is involved in immune homeostasis and immune and tumor surveillance.

As with all death receptors, Fas has a conserved death domain (DD) in its cytoplasmic tail that is crucial for the initiation of Fas-induced apoptosis. Fas and FasL ligation results in oligomerization and aggregation of the Fas receptor, which then leads to death-inducing signaling complex (DISC) assembly at the cellular membrane. DISC consists of Fas receptor, Fas associated with a death domain (FADD) adaptor molecule, procaspase-8, procaspase-10, and the cellular FLICE-like inhibitory protein (c-FLIP). DISC formation results in procaspase-8 activation, which later leads to cleavage of various intracellular proteins and ultimately apoptosis.

Fas Expression and Its Role in OS Lung Metastasis Formation

Fas-induced apoptosis is involved in tumor cell death and regulation of tumor development. Multiple studies have demonstrated that the absence of the Fas signaling pathway in primary tumors is associated with poor prognosis [11,12,13,14]. Tumor cells downregulate their Fas expression to escape from FasL-mediated apoptosis induced by activated immune cells [11, 14, 15]. Altered Fas expression can also affect a tumor’s metastatic potential [16, 17].

OS most commonly metastasizes to the lungs. Metastases to the lungs are often resistant to salvage chemotherapy [18]. Our laboratory has previously demonstrated an inverse correlation between the metastatic potential of human OS cells with Fas expression [19]. The LM7 cell, a subline of the SAOS human OS cell line obtained by recycling the cells seven times through the lungs of nude mice, expresses low levels of Fas [20], whereas the SAOS cells express high levels of Fas. SAOS cells cannot induce pulmonary metastasis when injected intravenously (i.v.), whereas LM7 cells form metastasis in the lung when injected i.v [21]. Similarly, K7 mouse OS cells, which express high levels of Fas, are not metastatic whereas K7M3 cells, derived from K7 after recycling the cells through the lungs, express low levels of Fas and form lung metastases when injected i.v. In addition, K7M3 cells form primary tumors in the bone if injected into the tibia and metastasize to the lung spontaneously. The primary bone tumor that develops in the tibia homogeneously expresses Fas, while lung metastases have low to no Fas expression [22]. Because FasL is constitutively expressed in the lung, we hypothesized that when OS cells express a functional Fas receptor, they will undergo cell death due to Fas/FasL-mediated apoptosis as they approach the lung microenvironment. On the other hand, Fas− OS cells will survive and form lung metastasis. We also showed that LM7 cells transfected with the full-length Fas gene expressed a higher level of Fas and formed significantly fewer and smaller pulmonary nodules compared to control-transfected LM7 cells [21]. Conversely, blocking the Fas signaling pathway in K7M3 and K7 mouse OS cells by transfection with Fas-associated death domain (FADD) dominant-negative (FDN) plasmid resulted in lower sensitivity to FasL-mediated apoptosis in vitro and enhanced metastatic potential to the lungs. Lung nodules from mice injected with the FADD_DN-transfected cells contained both Fas-positive and Fas-negative cells [15, 22]. These results support our hypothesis that Fas expression influences OS cells metastatic potential. A functional and intact Fas/FasL signaling pathway is key to the development of OS lung metastases. We further confirm these findings by injecting wild-type K7M3 and K7 cells into an FasL-deficient gld mice and found an increase in the number of lung tumors with both Fas-positive and Fas-negative cells [15, 22] suggesting that in the absence of FasL in the pulmonary epithelium, Fas+ tumor cells can survive and grow in the lungs. Subsequent analysis of patient samples supported our pre-clinical findings. Immunohistochemistry staining for Fas expression of 38 OS lung metastatic patient samples revealed 60% of the samples to be Fas negative, 32% to be weakly positive, and 3.2% (only one sample) to be strongly positive. Fas-positive expression was only detected in patients who had received chemotherapy prior to lung metastasis resection suggesting that treatment may contribute to Fas upregulation in OS tumors. Indeed, we further demonstrated that gemcitabine [23], interleukin -12 [24], entinostat/syndax275 [25], and 9-Nitrocamptothecin [26] upregulated Fas expression on OS cells which then resulted in the regression of established lung metastases.

Taken together, our findings address the importance of the Fas/FasL signaling pathway in the metastatic potential of OS and suggest that therapies able to upregulate Fas expression may add benefit in the treatment of OS lung metastases.

Gemcitabine and Its Effect on Osteosarcoma

Gemcitabine (2′,2′-difluorodeoxycytidine, dFdC) is a chemotherapeutic agent that has been approved for the treatment of various solid tumors including non-small-cell lung carcinoma, pancreatic, breast, and ovarian cancers. Gemcitabine is a deoxycytidine analog and its antitumor activity is the result of its ability to inhibit DNA replication and ultimately lead to cell death [27]. It has been tested in multiple pre-clinical and clinical settings [28,29,30,31,32,33], including OS [34,35,36,37,38,39,40,41,42,43,44,45]. Gemcitabine in combination with docetaxel remains a standard well-tolerated salvage chemotherapy regimen in the treatment of multiple sarcomas. However, it has only shown modest efficacy in relapsed/refractory OS [46,47,48,49,50]. Ofer Merimsky and colleagues reported gemcitabine treatment prolongs disease stabilization in 70% of patients with bone sarcomas resistant to doxorubicin [35]. A phase II clinical trial of the combination of gemcitabine and sirolimus demonstrated promising results in patients with relapsed and progressing OS [43]. Other gemcitabine combinations have not been as successful, however. Specifically, the addition of gemcitabine to carboplatin, for example, did not show benefit as compared to carboplatin alone in dogs with OS [37].

Based on our preliminary findings in the laboratory that the Fas-FasL pathway is implicated in the metastatic potential of OS, we hypothesized that agents that upregulate Fas expression could provide therapeutic benefit as the presence of FasL in the lung microenvironment will lead to cell death. Indeed, we demonstrated [23,24,25,26] in vitro that gemcitabine upregulated Fas expression in various OS cell lines and enhanced cell sensitivity to FasL in the lung. Inhibition of the Fas/FasL signaling pathway abolished the gemcitabine therapeutic effect, suggesting that an intact Fas pathway is important to the therapeutic efficacy of gemcitabine [22]. Other groups have similarly reported that gemcitabine induced growth inhibition, cell cycle arrest, and apoptosis in canine OS cell lines [36, 38]. Consistent with our findings, in vitro culture with relatively low concentrations of gemcitabine significantly increased functional Fas receptor expression in lung, colon, breast, and pancreatic tumor cell lines [51, 52].

Using two OS mouse models (K7M3 and LM7), we demonstrated aerosol gemcitabine to have therapeutic effect. Gemcitabine therapy resulted in significant increase in Fas expression, enhanced apoptosis, and subsequent regression of lung metastases. Aerosol gemcitabine further inhibited the growth of a subcutaneous OS primary tumor [22, 53]. We also confirmed in vivo the importance of the Fas/FasL pathway in the therapeutic efficacy of gemcitabine as aerosol gemcitabine therapy given to gld mice whose FasL function is impaired, resulted in increased Fas expression in OS lung metastasis but no therapeutic effect [22]. Similarly, dogs with OS lung metastasis treated with aerosol gemcitabine demonstrated increased Fas expression, apoptosis, and percentage of tumor necrosis [54]. Takashi Ando and colleagues have also demonstrated that systemic administration of gemcitabine results in a decrease in primary tumor growth, increased cell apoptosis, and decreased pulmonary metastasis in an OS mouse model [38]. Taken together, these results provide a rationale for the use of gemcitabine in combination with other agents shown to upregulate Fas expression to further enhance gemcitabine therapeutic effect against OS.

Histone Deacetylase (HDAC) Inhibitors

Epigenetic modifications, such as DNA methylation and acetylation, induce chromatin remodeling and altered gene expression. Defects in epigenetic regulation may result in loss or gain of gene function and lead to onset and progression of human diseases including cancer [55].

Histone acetyltransferases (HATs) and histone deacetylases (HDACs) are responsible for histone modifications. HAT stimulates gene transcription through the transferring of acetyl moieties to histone’s N-terminal lysine residues, which results in a less compact chromatin state. The opposing activity of the HDAC enzymes contributes to transcriptional repression by removing the acetyl moieties, creating a more compact chromatin leading to less gene expression. 18 HDACs have been identified in humans and are classified into four groups. Class I contains HDAC 1, 2, 3, and 8; Class II contains HDAC 4, 5, 6, 7, 9, and 10; Class III contains sirtuins and Class IV contains HDAC 11. Studies suggest that aberrant function of HAT and HDAC is often linked to tumorigenesis and poor prognosis in cancer [56]. Therefore, targeting these two enzymatic activities may provide therapeutic means to treat several malignancies associated with faulty epigenetic modifications [57, 58].

Several HDAC inhibitors have been shown to have anti-cancer effects. HDAC inhibitors regulate gene transcription by limiting the accessibility of transcription factors and RNA polymerase activities at the promoter level. HDAC inhibitors belong to four structural classes: (I) hydroxamic acids (hydroxamates); (II) benzamides; (III) short-chain fatty (aliphatic) acids; (IV) cyclic tetrapeptides; and (V) sirtuin inhibitors. In recent years, several HDAC inhibitors, with various target specificities and pharmacokinetics, have been under evaluation in clinical and preclinical studies. Thus far, four have received FDA approval for cancer treatment: vorinostat (SAHA), Belinostat (PXD-101), panobinostat (LBH589), and Istodax (romidepsin) [59, 60].

HDAC inhibitors have demonstrated a broad range of effects on tumor cells including cell death, growth arrest, and cell cycle suppression. In the clinical setting, tumor debulking and differentiation, prevention of angiogenesis, and enhancement of host immune response have been attributed to HDAC inhibitors [58]. Studies demonstrated that HDAC inhibitors are selectively more cytotoxic to cancer cells than normal cells, suggesting a potential therapeutic benefit of these drugs for the treatment of cancer [61, 62]. It has been shown that class I HDACs (1, 2, 3 and 8) play a key role in the pathogenesis of OS [63, 64]. Entinostat/syndax-275, a member of the benzamide group, is a narrow-spectrum HDAC inhibitor and affect HDAC class I with limited effect on HDAC 8 [65]. Entinostat/syndax-275 is in several phase I/II clinical trials for the treatment of solid and hematologic malignancies.

Entinostat/Syndax-275 and Its Effect on Osteosarcoma

It is well known that HDAC inhibitors can inhibit human and canine OS cell growth by promoting apoptosis, mostly through Fas-mediated or caspase-dependent pathways. For example, treatment with valproic acid prior to incubation with doxorubicin resulted in less cell growth and more apoptosis both in canine and human OS cells. In addition, valproic acid and doxorubicin combination therapy in a canine OS subcutaneous xenograft model led to significantly less tumor growth compared to either alone [66]. Further combination of two epigenetic modifying drugs, the DNA methylation inhibitor, Zebularine, and the HDAC inhibitor suberoylanilide hydroxamic acid (SAHA) showed significant human and canine OS cell growth inhibition. Inhibition was more effective in cell lines with a more aggressive gene expression profile [67]. Similarly, co-treatment with a DNA methyltransferase inhibitor, 5-Aza-dC, and HDAC inhibitor trichostatin A effectively reduced cell proliferation of the multi-drug resistance OS cell line HosDXR150, whereas single treatment had only a minor effect on cell viability [27]. Lastly, SAHA in combination with cisplatin decreased cell proliferation and enhanced OS cell apoptosis via caspase activation [68, 69].

HDAC Effect on the Fas/FasL Apoptotic Pathway

HDAC inhibitors can sensitize tumor cells to Fas-mediated apoptosis using different mechanisms. For example, apicidin and depsipeptide (FR901228) increased apoptosis in acute promyelocytic leukemia cells and uveal melanoma by inducing upregulation of Fas/FasL expression [70,71,72]. Another study demonstrated the HDAC inhibitor PCI-24781 to induce apoptosis in acute leukemia cells through activation of caspase-8 and FADD [73]. In OS cells, FR901228 inhibited cell growth both in vitro and in xenograft mouse models. FR901228 upregulated FasL mRNA and cell surface expression, activated caspase-8 and -3 and ultimately induced Fas-mediated apoptosis [74].

We also have demonstrated that therapeutically achievable doses of entinostat/syndax-275 while having limited cytotoxic effect on OS cell growth in vitro activate the Fas pathway and enhance Fas mRNA and protein expression. Combination treatment entinostat/syndax-275 and FasL significantly increased OS cells’ sensitivity to FasL as demonstrated by enhanced caspase cleavage/activity and reduced clonogenic growth. Blocking the Fas pathway reversed this effect [25, 75]. Intranasal administration of entinostat/syndax-275 at a dose of 0.13 mg/kg (which is approximately 200-fold less than the therapeutically effective oral dose described before) in mice with established OS lung metastasis resulted in reduced metastatic tumor growth [13]. In addition, oral administration of entinostat/syndax-275 in mice with OS pulmonary metastasis resulted in tumor growth inhibition and increased survival rate. Histopathological examination showed a higher level of apoptosis and lower level of cellular FLICE inhibitory protein (c-FLIP) expression in the lung tissues of treated mice. No evidence of drug toxicity was observed in the treated group of mice [75].

Despite sufficient evidence to demonstrate that entinostat/syndax-275 activates the Fas pathway in OS, studies in our lab demonstrated that this HDAC inhibitor did not increase the expression of Fas on the cell surface. Instead, entinostat/syndax-275 treatment led to redistribution of Fas to membrane lipid rafts and downregulation of cellular c-FLIPmRNA and protein expression. c-FLIP knockdown in OS cells resulted in the redistribution of Fas to lipid rafts and enhanced sensitivity to FasL-induced cell death [75, 76]. Our findings were consistent with other studies demonstrating that the HDAC inhibitor FR901228 downregulated c-FLIP in both chronic lymphocytic leukemia cells and Fas-resistant OS cells and enhanced their sensitivity to Fas-mediated apoptosis [77, 78]. Entinostat/syndax-275 has also been shown to downregulate c-FLIP in chronic lymphocytic leukemia (CLL) cells and induce caspase-dependent apoptosis [79]. Similarly, 7 days of treatment with valproic acid sensitized OS cells to Fas-mediated cell death without enhancing Fas expression on the cell surface [80].

c-FLIP is a key regulator of Fas-mediated apoptosis. c-FLIP, a catalytically inactive caspase-8/-10 homolog, interferes with activation of procaspase-8 at the death-inducing signaling complex (DISC) level and prevents Fas-induced apoptosis [81]. Many studies showed that c-FLIP was overexpressed in various cancer cells and its expression is liked with tumorigenesis and poor survival [75, 82,83,84], highlighting a potential mechanism by which cancer cells resist to death receptor-induced apoptosis. The expression of c-FLIP has also been correlated with resistance to several chemotherapy drugs [70, 85]. We also evaluated c-FLIP expression in patient primary and pulmonary OS samples using immunohistochemistry. C-FLIP expression was significantly higher in pulmonary nodules than in primary tumors. Similar results were observed in our human xenograft models [76]. Taken together, these findings suggest that the overexpression of c-FLIP as an inhibitor of the Fas-signaling pathway may contribute to the survival and growth of OS cells in a FasL+ lung microenvironment. Therefore, the downregulation of c-FLIP in entinostat/syndax-275-induced Fas signaling may be therapeutically beneficial for the treatment of OS lung metastasis.

Gemcitabine and Entinostat/Syndax-275 as Potential Salvage Regimen for Osteosarcoma Lung Metastasis

The above preclinical data suggests that the use of therapeutic agents able to upregulate Fas expression, increase Fas localization to lipid rafts, or decrease cFLIP expression may offer benefit in the treatment of OS. The combination of gemcitabine and entinostat/syndax-275 – both of which have been shown to enhance Fas expression in OS cells – has not been studied in pediatric patients with refractory or recurrent pulmonary OS. Therefore exploitation of the Fas/FasL pathway as a potential therapeutic option for patients with refractory OS seems appropriate. A phase I/II clinical trial of the combination is under development at MD Anderson Cancer Center. This clinical trial will evaluate feasibility and safety of the combination therapy entinostat/syndax-275 and gemcitabine and determine whether there is potential utility for patients with refractory/relapsed OS. To this end, the primary objective of the study is to determine the maximum tolerated dose (MTD) of entinostat/syndax-275 when it is given in combination with gemcitabine in pediatric patients with recurrent sarcoma and recommend a phase 2 dose of the combination therapy. Secondary objectives include: 1) To determine the disease control rate at 4 months for pediatric patients with recurrent unresectable pulmonary OS when treated with gemcitabine in combination with entinostat/syndax-275 and 2) To estimate the disease-free survival for the subset of pediatric patients with recurrent pulmonary OS that has been fully resected after treatment with gemcitabine in combination with entinostat/syndax-275. It is expected that this trial will serve as a potential therapeutic alternative for patients with refractory OS. Several approaches have been taken to treat OS. However, none have shown significant benefit as there has been no impact in survival. It is of paramount importance that therapies that move into clinical trials have a scientific rationale. Here we present enough preclinical evidence to support combination therapy gemcitabine and entinostat/syndax275 for refractory OS. Therefore, results from this study holds promise as an alternative to treat patients with OS.