Abstract
Targeted therapies have revolutionized cancer treatment; however, progress lags behind in alveolar (ARMS) and embryonal rhabdomyosarcoma (ERMS), a soft-tissue sarcoma mainly occurring at pediatric and young adult age. Insulin-like growth factor 1 receptor (IGF1R)-directed targeted therapy is one of the few single-agent treatments with clinical activity in these diseases. However, clinical effects only occur in a small subset of patients and are often of short duration due to treatment resistance. Rational selection of combination treatments of either multiple targeted therapies or targeted therapies with chemotherapy could hypothetically circumvent treatment resistance mechanisms and enhance clinical efficacy. Simultaneous targeting of distinct mechanisms might be of particular interest in this regard, as this affects multiple hallmarks of cancer at once. To determine the most promising and clinically relevant targeted therapy–based combination treatments for ARMS and ERMS, we provide an extensive overview of preclinical and (early) clinical data concerning a variety of targeted therapy–based combination treatments. We concentrated on the most common classes of targeted therapies investigated in rhabdomyosarcoma to date, including those directed against receptor tyrosine kinases and associated downstream signaling pathways, the Hedgehog signaling pathway, apoptosis pathway, DNA damage response, cell-cycle regulators, oncogenic fusion proteins, and epigenetic modifiers. Mol Cancer Ther; 17(7); 1365–80. ©2018 AACR.
Introduction
Rhabdomyosarcoma is the most common type of soft-tissue sarcoma (STS) observed in young patients with the most frequent subtypes being embryonal (ERMS) and alveolar rhabdomyosarcoma (ARMS). ERMS represents approximately 70% of childhood rhabdomyosarcoma and is most often observed in the head and neck region and genitourinary track. ARMS is seen in approximately 30% of rhabdomyosarcoma cases and usually occurs in the deep tissue of the extremities. The majority of ARMS tumors are characterized by a fusion between PAX3 or PAX7 on chromosome 2 and FOXO1 on chromosome 13 (∼80%). The remaining 20% are fusion negative. Although generally ARMS have a poorer outcome compared with ERMS, fusion-negative ARMS show a genetic profile similar to ERMS and an equally favorable clinical outcome. Multimodality treatment including surgery, chemotherapy, and radiotherapy has increased the 5-year overall survival (OS) to approximately 70%–90% for intermediate- and low-risk rhabdomyosarcoma, respectively. However, patients with high-risk rhabdomyosarcoma still have a 5-year OS of <40%. In addition, treatment-related toxicities severely decrease quality of life (1, 2). In an attempt to increase survival and improve quality of life, the field of targeted therapy has gained substantial interest in rhabdomyosarcoma, and its potential is supported by various lines of (pre)clinical research, which are mainly centered on targeted therapies originally developed for other tumor types. In the clinic, however, intrinsic and acquired resistance mechanisms have emerged as common pitfalls in rhabdomyosarcoma treatment. As such, increasing evidence exists that single-agent targeted therapy will not be sufficient to reach clinical efficacy in patients with rhabdomyosarcoma. The current hypothesis is that combination therapy could enhance clinical efficacy and/or decrease treatment-associated toxicities. In this regard, simultaneous targeting of different mechanisms of action could be more effective as opposed to combining inhibitors of similar classes, as the characteristic hallmarks of cancer illustrate that tumor progression is regulated by a wide variety of different processes.
To determine the most promising and clinically relevant combination treatments for rhabdomyosarcoma, we reviewed the preclinical and (early) clinical trial data addressing combinations of targeted therapies or targeted therapy combined with chemotherapy. We focused on the most common classes of targeted therapies investigated in rhabdomyosarcoma to date, including those directed against receptor tyrosine kinases (RTK) and associated downstream signaling pathways, the Hedgehog signaling pathway, apoptosis pathway, DNA damage response, cell-cycle regulators, fusion proteins, and epigenetic modifiers.
RTKs
RTKs are membrane-bound proteins involved in signal transduction to the tumor cell. Activation by extracellular ligand binding or genetic mutations can lead to constitutive activity. Intracellular signaling pathways, including the PI3K/AKT/mTOR, RAS/MEK/ERK, and JAK/STAT3 pathway, are subsequently activated. Several RTKs have been identified as possible targets for therapy in rhabdomyosarcoma, including the insulin-like growth factor 1 receptor (IGF1R), anaplastic lymphoma kinase (ALK), platelet-derived growth factor receptors α and β (PDFGRα/β), VEGFR, EGFR, and the fibroblast growth factor receptor 4 (FGFR4; Fig. 1A; ref. 3). Despite the promising preclinical effects, clinical efficacy is limited and observed in small subsets of patients (4, 5). Combination treatment might enhance the clinical efficacy of RTK-targeted therapies (Table 1).
Overview of the cellular processes used as targets in the targeted therapy–based combination treatments in ARMS and ERMS. A, Membrane-bound growth factor receptors IGF1R, ALK, PDGFR, FGFR4, EGFR, VEGFR, Patched 1 (PTCH1), SMO, and TRAILR1/2; ligands IGF1/2, VEGF and (Sonic, Indian, Desert) Hh; intracellular signaling proteins of the PI3K/AKT/mTOR, JAK/STAT3, RAS/MEK/ERK, Hh and apoptosis pathway (4E-BP1, eukaryotic translation initiation factor 4E-binding protein 1; eIF4E, eukaryotic translation initiation factor 4E; JAK, Janus kinase; FADD, Fas-associated protein with death domain; CASP, caspase; RIP1, receptor-interacting serine/threonine-protein kinase 1; BID, BH3 interacting-domain death agonist; BAX, Bcl-2-associated X protein; BAK, Bcl-2 homologous antagonist killer; BCL-2, B-cell lymphoma 2; MCL-1, induced myeloid leukemia cell differentiation protein; BCL-XL, B-cell lymphoma-extra large; Smac, second mitochondria-derived activator of caspases; Cyt C, cytochrome C, IAP, inhibitor of apoptosis protein). B and C, Intranuclear processes: DNA damage response (DDR) (PARP), the epigenome and the cell cycle that are used as therapeutic targets in (pre)clinical combination treatments in ARMS and ERMS.
Preclinical targeted therapy–based combination treatments in ARMS and ERMS
IGF1R/ALK.
One way of optimizing IGF1R treatment might be its combination with other (R)TK inhibitors. As coexpression of the RTKs IGF1R and ALK has been described in rhabdomyosarcoma, this may present a rational combination. In vitro, combined anti-IGF1R antibody R1507 and ALK inhibitor TAE684 treatment showed synergism in ARMS cell lines. In ERMS, however, no enhanced effect was observed (6). In ARMS, the characteristic PAX3-FOXO1 protein can enhance IGF1R and ALK transcription, possibly explaining this difference in sensitivity. However, we, among others, could not find intrinsic ALK activity in rhabdomyosarcoma cells (7–9). In addition, the antitumor effects observed with the ALK inhibitor ceritinib were mostly explained by its capacity to inhibit IGF1R signaling (9). Combined treatment of ceritinib with the multikinase inhibitor sorafenib showed synergistic effect in vitro (8). In addition, we observed high activity of the signaling protein Src post ceritinib treatment in both subtypes, and combined treatment of ceritinib and the Src inhibitor dasatinib was synergistic in vitro (9). Prior to our finding, increased activity of the Src family tyrosine kinase (SFK) YES was observed in IGF1R treatment–resistant cell lines. Combination treatment of IGF1R antibodies and SFK inhibitors led to superior in vitro apoptosis induction and in vivo tumor growth reduction compared with the monotherapies (10). One phase I/II trial is recruiting ARMS and ERMS patients to investigate the combined effects of the IGF1R antibody ganitumab and dasatinib (NCT03041701; Table 2). No results have yet been reported; however, based on the consistent preclinical findings concerning IGF1R targeting and an increase in Src signaling, the results are eagerly awaited.
Clinical trials examining targeted therapy-based combination treatments in patients with (soft tissue) sarcoma
In addition to an increase in Src activity, ERMS showed enhanced PDGFRβ activity as a resistance mechanism to IGF1R treatment. The combination of IGF1R- and PDGFRβ-targeted therapies increased growth inhibition in IGF1R-resistant ERMS cell lines. Three PDGFRβ inhibitors were tested, and combined therapy with pazopanib (VEGFR, PDGFR, c-KIT) or crenolanib (PDGFRα/β) was the most beneficial in vivo. The combinations delayed tumor growth compared with each monotherapy, although no complete tumor regressions were achieved (11).
Resensitization of IGF1R treatment–resistant tumor cells was also examined by combinations of IGF1R ligand antagonists and downstream signaling protein inhibitors. Insulin-like growth factor binding proteins (IGFBP) regulate binding of insulin-like growth factor 1 and -2 (IGF1/2) to IGF1R. In ARMS, IGF1R-resistant cell lines showed reduced IGFBP2 expression and a limited decrease of downstream AKT activity upon IGF1R antibody treatment. IGF1R antibodies combined with recombinant IGFBP2, the PI3K inhibitor BKM130, or the mTOR inhibitor Ku-0063784 resensitized the resistant cell lines to IGF1R targeting (12). The possible low bioavailability of recombinant IGFBP2 could affect the clinical potential of this particular combination treatment. Clinical trials did show a partial response (PR) in a patient with IGF1R-positive soft-tissue sarcoma and stable disease (SD) in 2 of 11 pediatric, adolescent, and young adult (AYA) patients with rhabdomyosarcoma for the combination of IGF1R inhibitors and mTOR inhibitors (Table 2; refs. 13–16). This shows that only a small group responded to treatment and one study showed that the addition of temsirolimus led to increased toxicity without a clear increase in efficacy in most patients (15). Nevertheless, these data do show that combined targeted therapy can overcome primary treatment resistance and suggest that, when given simultaneously, might delay or prevent treatment resistance altogether.
Because chemotherapy remains fundamental in rhabdomyosarcoma treatment, combinations with chemotherapy have been investigated. In young patients with ARMS and ERMS, the combination of the IGF1R antibody cixutumumab with conventional chemotherapy was compared with the combination of temozolomide with conventional chemotherapy. The combination with cixutumumab led to a higher percentage of patients reaching an 18-month event-free survival (EFS; cixutumumab 68% vs. temozolomide 39%; NCT01055314; Table 2). Of note, the combination with cixutumumab had more reports of high-grade toxicity compared with the combination with temozolomide. High-grade toxicities were, however, only observed in a very small group (3/97, 3%), leaving the combination with cixutumumab preferential to the combination with temozolomide (17). Although preclinical research and limited clinical research suggests that combinations of targeted and cytostatic agents could render success, phase I studies should first investigate the optimal dose and schedule of new combinations in clinical practice.
VEGFR.
VEGFs bind to VEGFRs and induce angiogenesis. This makes both VEGFs and VEGFRs a possible target for treatment. The camptothecin analogue namitecan was shown to reduce angiogenesis in an ERMS mouse model. Moreover, combined treatment of low-dose namitecan with either the VEGF antibody bevacizumab or the VEGFR inhibitor sunitinib led to an enhanced tumor growth reduction compared with the monotherapies. Sunitinib was in favor of bevacizumab, possibly due to the effect of camptothecins on VEGF expression or the multikinase inhibition of sunitinib (18). VEGF can also be targeted via inhibition of heparanase. Heparanase is an enzyme necessary to generate heparin sulfate–bound growth factors, including VEGF. Heparanase activity is increased in rhabdomyosarcoma and combination treatment of the heparanase inhibitor SST0001 and bevacizumab or sunitinib reduced angiogenic growth factor expression and decreased cellular invasion in vitro. In vivo SST0001 monotherapy effectively decreased ARMS and ERMS tumor volumes. No further examination of the combination was performed in vivo, precluding examination of the clinical potential of combinations with SST0001 (19). Clinical trials have not (yet) shown convincing effects of anti-VEGF(R) combination treatments in young patients with rhabdomyosarcoma. Bevacizumab combined with standard chemotherapy did not significantly increase the median EFS in patients with rhabdomyosarcoma (NCT00643565; Table 2; ref. 20). The combination of bevacizumab with sorafenib and low-dose cyclophosphamide did lead to a PR in a patient with rhabdomyosarcoma. However, only 2 patients with rhabdomyosarcoma were included in the study, making the clinical efficacy of this combination in patients with rhabdomyosarcoma difficult to determine (21).
EGFR.
Similar to IGF1R and VEGF(R), combined therapy of the anti-EGFR antibody cetuximab and the chemotherapeutic dactinomycin was superior to dactinomycin treatment alone in EGFR-expressing ARMS and ERMS cell lines (22). Despite the fact that EGFR overexpression is present in 37%–76% of ERMS and 16%–50% of ARMS (23), no in vivo or clinical trials have been described for EGFR-based combination treatment in rhabdomyosarcoma; hence, no solid conclusions can be drawn regarding the potential of this combination in the clinic. A novel way of targeting EGFR in combination with cytotoxic compounds was recently introduced, by creating antibody–drug conjugations. Niesen and colleagues used single-chain fragment variables (scFv) with high tumor penetration capacities in combination with truncated exotoxin A to create immunotoxins (IT) based on the anti-EGFR antibodies cetuximab and panitumumab. EGFR binding led to the internalization of the compound and reduced cell viability in a number of EGFR-expressing cell lines. Further analysis showed binding of the IT to EGFR-expressing rhabdomyosarcoma cell lines and reduced cell viability with induction of apoptosis. Moreover, the IT was shown to bind to EGFR-expressing rhabdomyosarcoma tumor tissue (23). Niesen and colleagues more recently developed fully human cytolytic fusion proteins (hCFP) by conjugating serine protease granzyme B to EGFR scFvs. hCFPs were shown to be superior to ITs since the generation of neutralizing antibodies against the bacterial toxins was prevented. EGFR-directed hCFPs showed similar effects as the IT, including binding to rhabdomyosarcoma tumor tissue, and the effects were increased when combined with potency-enhancing chloroquine (24). Despite the lack of in vivo experiments and clinical trials in patients with rhabdomyosarcoma, these hCFPs remain an interesting compound, especially because in vivo experiments with ITs in breast cancer xenografts, and a phase I study with anti-CD22 immunotoxin moxetumomab pasudotox in childhood acute lymphoblastic leukemia showed good tumor penetration and well-tolerated toxicity levels, respectively (25, 26).
FGFR4.
Similar to IGF1R, FGFR4 is a transcriptional target of the PAX3-FOXO1 fusion protein leading to increased protein expression (27). In rhabdomyosarcoma, FGFR4 targeting is expected to be most effective in those tumors expressing activated FGFR4 due to amplification or mutation (28). Amplification is more likely to occur in ARMS as a result of the PAX3-FOXO1 fusion protein regulating FGFR4 expression. Indeed, higher activated FGFR4 expression levels were observed in ARMS as compared with ERMS cell lines (7). In contrast, whole-exome/transcriptome sequencing revealed activated FGFR4 mutations in 4 of 60 patients with rhabdomyosarcoma, all of which were ERMS (28). The only study describing a combination treatment with anti-FGFR4 therapy showed that FGFR4 activity functions as a compensatory mechanism to the effects of combined IGF1R and PI3K/mTOR inhibitor treatment in ARMS and ERMS cell lines. In addition, combined IGF1R and FGFR4 inhibitor treatment showed synergism in an ARMS cell line (29). This suggests that combinations of IGF1R and/or PI3K/mTOR and FGFR4 inhibitors might be able to prevent acquired treatment resistance. The Pediatric MATCH study is an ongoing trial, in which the pan-FGFR inhibitor JNJ-42756493 is tested in pediatric patients with relapsed and refractory advanced solid tumors, including STS (NCT032107140). No combination studies with an FGFR4 inhibitor are currently in the clinic for rhabdomyosarcoma.
PDGFRα.
Expression of both PDGFRα and PDGF ligands has been described in rhabdomyosarcoma tumors (30). However, single-agent treatment ultimately leads to treatment resistance. Imatinib-resistant murine ARMS cell lines no longer showed a decrease of PDGFRα activity, and demonstrated increased Src activity following imatinib (Abl, c-KIT, PDGFR) treatment. Combination therapy of imatinib with the SFK inhibitor PP2 enhanced cell viability reduction in the resistant cell line. However, this combined treatment was not more effective than monotherapy with the PDGFR/RAF inhibitor sorafenib. Both in absence and presence of PDGFR, sorafenib reduced cell viability more effectively than imatinib and/or PP2. A similar reduction in cell viability was seen in both naïve and imatinib-resistant cell lines following sorafenib treatment. This suggests that PDGFRα activity is not as important for cell viability as RAF activity, questioning the importance of PDGFRα targeting in ARMS (31). One clinical trial compared the combination of the multikinase inhibitor pazopanib (VEGFR, PDGFR) with the MEK inhibitor trametinib to pazopanib monotherapy in advanced STS. One sinonasal ERMS was included, which showed a PR to the combined treatment. However, this patient was diagnosed with a PIK3CA E542K aberration and previously progressed on PI3K inhibitor monotherapy. Resistance to PI3K treatment can be related to an increase in MEK activity, the specific target of trametinib; thus, the PR might not necessarily be linked to the combination with pazopanib and might merely be the result of the administration of trametinib (32). This again questions the influence of PDGFRα inhibition on the observed effect. Nevertheless, the anti-PDGFRα antibody olaratumab has gained FDA approval for treatment of STS and olaratumab combined with doxorubicin showed an enhanced OS compared with doxorubicin alone in a group of mixed STS. No inclusion of rhabdomyosarcoma was mentioned and PDGFRα expression did not correlate with outcome (33–35). The ANNOUNCE trials will further examine the use of olaratumab combined with chemotherapeutics in a larger group of STS (NCT02451943; NCT02659020). Results are not yet available, leaving the clinical efficacy of olaratumab in patients with rhabdomyosarcoma unknown.
Downstream signaling pathways
PI3K/AKT/mTOR.
The PI3K/AKT/mTOR pathway shows aberrant activation in rhabdomyosarcoma, either by mutations in the PIK3CA gene or by high levels of growth factor signaling. PI3K/AKT/mTOR signaling promotes gene transcription, cell growth, metabolism, cell motility, and invasion (Fig. 1A). Combination treatment of the PI3K inhibitor buparlisib with the IGF1R inhibitor AEW541, the mTOR inhibitor rapamycin, or the MEK inhibitor trametinib showed synergism in vitro (36). Other MEK inhibitors (including selumetinib) showed similar synergism with multiple PI3K/mTOR inhibitors (37, 38). In vivo, the most effective treatment was the mTOR inhibitor AZD8055 combined with selumetinib, showing an enhanced reduction in downstream protein activity. This combination was hence suggested for further clinical evaluation (38). Clinical trials did find an increased efficacy of dual PI3K and MEK inhibition compared with the monotherapies in patients with advanced cancer; however, this was at the cost of increased toxicity (39). Combination treatment based on inhibiting these two main downstream signaling pathways might therefore not be feasible in patients with rhabdomyosarcoma. The combination of mTOR inhibitors with chemotherapy might have more potential, as combinations with liposomal doxorubicin, irinotecan, temozolomide, vinblastine, or cyclophosphamide and topotecan were well tolerated in pediatric, AYA patients (40–43). Two clinical studies reported a response in a small subset of patients with rhabdomyosarcoma (41, 42).
RAS/MEK/ERK pathway.
Similar to PIK3CA, RAS mutations are present in a subset of ERMS leading to higher activity of the pathway (28). Both tyrosine and serine residues are phosphorylated and monotherapy with the MEK inhibitor PD98059 and the chymotrypsin-like serine protease inhibitor TPCK significantly delayed tumor growth in a KRAS-mutated rhabdomyosarcoma zebrafish model without affecting normal behavior and growth (44). The monotherapies reduced proliferation in an NRAS-mutated ERMS cell line, but did not induce apoptosis. Combination of suboptimal concentrations significantly reduced in vitro cell proliferation and in vivo tumor growth, indicating that this combination could be of interest for RAS-mutated rhabdomyosarcoma. However, with the above-mentioned downstream protein inhibitor combination in mind, the pharmacokinetics and treatment-related toxicities should be closely monitored.
The inhibition of the RAS/MEK/ERK pathway was also tested in combination with radiotherapy. Cancer stem cells (CSC) are involved in self-renewal and migration of the tumor, and high expression of CD133 is a CSC marker in ERMS. CD133-positive ERMS spheres treated with the MEK inhibitor U0126 showed reduced sphere formation and combined treatment with radiotherapy enhanced antitumor effects. This suggests that MEK is involved in ERMS CSCs and that CSC may be vulnerable to combined radiotherapy and MEK treatment (45). These are, however, preliminary data; thus, further research is necessary before any potential clinical implications can be made.
JAK/STAT3 pathway.
Persistent STAT3 activity has been reported in rhabdomyosarcoma and monotherapy with the STAT3-specific LY5 or the upstream STAT3 GP130 inhibitor bazedoxifene led to decreased cell migration and induced apoptosis in pSTAT3-positive ARMS in vitro. The effects of both compounds could be enhanced by the addition of doxorubicin, cisplatin, or the MEK inhibitor AZD6244 (46). These results emphasize again that combinations of downstream protein inhibitors with either chemotherapy or other downstream inhibitors can be very capable of improving antitumor effects as compared with monotherapies. These are encouraging findings, although STAT3 inhibition combined with doxorubicin or cisplatin requires further in vivo evaluation to determine whether clinical evaluation is worthwhile.
Hedgehog signaling
Several well-known developmental pathways are involved in tumorigenesis, including the Hedgehog (Hh) signaling pathway. Hh signaling is involved in embryogenesis and ligand binding leads to GLI transcription factor (GLI) activation and subsequent target gene transcription. GLIs can also be activated by the PI3K/AKT/mTOR pathway, and a crosstalk between both pathways has been identified in several tumor types. The subsequent GLI-1 and -2 activity has been shown to have oncogenic consequences (Fig. 1A; ref. 47).
Hh signaling can be constitutively active in rhabdomyosarcoma and the GLI1/2 inhibitor GANT-61 significantly reduced cell growth in rhabdomyosarcoma xenograft models, although no complete responses (CR) were achieved. Combined treatment of GANT-61 with temsirolimus, rapamycin, or vincristine increased these effects, with a preference for temsirolimus over rapamycin (48, 49). Similar effects were observed in primary rhabdomyosarcoma cells, in which multiple PI3K and/or mTOR inhibitors combined with GANT-61 led to enhanced apoptosis. In addition, combination treatment of GANT-61 with the PI3K inhibitor PI103 showed a decreased clonogenic survival and growth in vivo. The combination was tested in a chicken embryo model, making additional in vivo examination in a fully formed organism necessary before a translation to the clinic can be made (50). Arsenic trioxide (ATO), an active component of Chinese medicine with anti-GLI1/2 effects, in combination with vincristine, vinblastine, and eribulin also showed synergism in both rhabdomyosarcoma subtypes in vitro (51). In addition, the dual inhibition of GLI1/2 and Hh-related protein smoothened (SMO) or glycogen synthase kinase 3 (GSK3) led to a significant reduction in colony formation in both ARMS and ERMS cells. However, 3D-spheroid culture, as a better representation of the in vivo situation, showed limited additive effects of these combinations (52, 53). The current in vitro and in vivo models do not allow for direct translation of these combinations to a clinical setting. However, with additional in vivo studies in more complex (mouse) models, the combination of Hh inhibitors with either chemotherapy or PI3K/AKT/mTOR inhibitors could have potential for rhabdomyosarcoma treatment.
Apoptosis pathway
Most targeted therapies are capable of inducing apoptosis. However, some treatments specifically activate the apoptosis pathway. Apoptosis is triggered via an extrinsic death receptor pathway and the intrinsic mitochondrial pathway. TNF-related apoptosis-inducing ligands (TRAIL) activate membrane-bound TRAIL receptors (TRAILR) and subsequently activate a caspase cascade or induce intrinsic mitochondrial pathway activation. Upon activation of the intrinsic pathway, mitochondrial cytochrome c, and Smac are released into the cytosol. Cytochrome c subsequently induces the activation of caspase-9, whereas Smac antagonizes the inhibitor of apoptosis (IAP) proteins (including survivin; Fig. 1A; ref. 54).
TRAILR1- and TRAILR2-specific agonistic antibodies, either alone or in combination with IAP inhibitors, were examined in both rhabdomyosarcoma subtypes. TRAILR1 and IAP inhibitor monotherapies were not effective. TRAILR2 therapy did show a dose-dependent cell viability reduction, although this was independent of the level of TRAILR2 present in the cell line. The TRAILR2 effects could be enhanced by addition of IAP inhibitors in both ARMS and ERMS (54).
A specific survivin inhibitor, YM155, was recently tested as a monotherapy and in combination with cisplatin in ERMS. YM155 reduced survivin levels and cell viability in vitro. Both in vitro and in vivo, combination treatment with cisplatin led to an enhanced effect on cell viability and caspase levels, but apoptosis only slightly increased as compared with cisplatin monotherapy. No examination of YM155 in the ARMS subtype was performed, leaving its use in ARMS to be evaluated (55).
Members of the antiapoptotic Bcl-2 family are overexpressed in rhabdomyosarcoma (Fig. 1A). Combined targeting of the Bcl-2 family inhibitor ABT737 and the mTOR inhibitor AZD8055 showed synergy in rhabdomyosarcoma cell lines. The added effects of AZD8055 with ABT737 were dependent on the inhibition of both mTOR complexes and a decrease in MCL-1 levels. Of note, combination of AZD8055 with chemotherapy was not synergistic (56).
All of the abovementioned studies suggest targeting of apoptosis pathway–associated proteins to be a potential therapeutic option. However, only one study examined in vivo effects and further research is necessary. Nonetheless, the combination of anti-BcL-2 and mTOR inhibitors does give a good example of the potential of a combination treatment that targets multiple processes in the tumor cell at once. Through the inhibition of multiple, distinct cellular processes, we might be able to generate a more robust antitumor effect without having to resort to combinations with systemic chemotherapy. One clinical trial is comparing the combination of a TRAILR2 agonist and chemotherapy with TRAILR2 and an anti-IGF1R antibody in adult solid tumors, including sarcomas (NCT01327612; Table 2). This trial might give more insights into the clinical efficacy of targeting multiple cellular processes simultaneously and whether it is preferential to combinations with systemic chemotherapy.
The abovementioned studies all describe targets consisting of membrane-bound growth factor receptors and associated intracellular signaling proteins. However, other intracellular processes such as DNA damage repair, cell-cycle regulation, and gene expression regulation are likewise targets for treatment (Table 1; Fig. 1B and C).
DNA damage response (DDR): PARP1
The DNA damage response (DDR) plays a crucial role in the defense against the deleterious effects of DNA damage (57, 58). Key to the DDR is PARP1. PARP1 is involved in single-strand break repair, where its binding to damaged DNA leads to poly (ADP-ribose) (pADPr) chain synthesis and recruitment of repair proteins. pADPr chains are also involved in PARP1 release from the DNA to ensure access of repair proteins to the damaged site (Fig. 1B). In rhabdomyosarcoma, monotherapy of the PARP inhibitor olaparib showed intermediate effects on cell viability. Combination of PARP inhibitors with multiple chemotherapeutics showed enhanced in vitro effects for combined treatment with irinotecan, melphalan, doxorubicin, and temozolomide (59–61). Lower synergism was seen for the combination with carboplatin or vincristine, and addition of topotecan did not enhance effects (60, 61). Moreover, the PARP inhibitor talazoparib combined with temozolomide showed response in 2 of 3 ARMS models in vivo, and one model showed a maintained CR until the end of the study (61). In addition, PARP and DNA repair enzyme tyrosyl-DNA phosphodiesterase (tdp1) can form a DNA repair complex. Genetic knockdown of tdp1 combined with irinotecan or the PARP1 inhibitor rucaparib led to enhanced antitumor effects in vitro (59).
A phase Ib trial investigating the effects of olaparib combined with trabectedin in STS reported a PR in 18% and SD in 23% of patients (NCT02398058; Table 2; ref. 62). A phase I trial in pediatric and AYA solid tumors is investigating the efficacy of combined treatment of talazoparib with irinotecan, with or without temozolomide. Preliminary data encouragingly showed response in a subset of patients for the combination of talazoparib and irinotecan (NCT02392793; Table 2; ref. 63). None of these studies have yet mentioned a response in rhabdomyosarcoma. One other trial is examining effects of olaparib combined with concomitant radiotherapy in locally advanced STS (NCT02787642; Table 2).
Cell cycle
In addition to the DDR, the cell cycle can be inhibited to affect cell viability. The cell cycle is a strictly regulated cellular process and multiple kinases are involved in its regulation, including the cyclin-dependent kinases (CDK), polo-like kinase 1 (PLK1), and Wee1 kinase (64). CDKs play a crucial role throughout the whole cell cycle and act at different stages of cell-cycle progression. PLK1 actively regulates the transition from G2–M phase by phosphorylating Wee1, triggering Wee1 degradation. Inhibition of PLK1 can induce a mitotic arrest leading to cell death. Wee1 negatively regulates entry into mitosis by inducing an inhibitory phosphorylation of CDK1, leading to a G2–M arrest necessary for DNA repair. Inhibition is thought to prevent the G2–M arrest and subsequent DNA repair resulting in a premature entry into mitosis and induction of cell death (Fig. 1C).
PLK1.
Rhabdomyosarcoma has higher levels of PLK1 compared with healthy tissue (65, 66). Multiple preclinical studies have shown ARMS to be more sensitive to PLK1 inhibitors compared with ERMS (65–67). This sensitivity might be related to the role of PLK1 in the activation and expression of PAX3-FOXO1 (68). All studies showed synergistic effects for the combination of PLK1 inhibitors and antimicrotubule agents (65–67). No synergism was observed for the combination with etoposide, doxorubicin, or paclitaxel (65, 66). The working mechanism of etoposide, doxorubicin, and paclitaxel is either before (etoposide) or after (doxorubicin, paclitaxel) the mitotic phase, which might explain the observed ineffectiveness of these combinations (66). In contrast to PLK1-catalytic domain inhibitors, polo-box domain inhibitors in combination with vincristine were less effective; although it still induced apoptosis to a larger extend than respective monotherapies (67). One phase I study with monotherapy volasertib in pediatric solid tumors has been concluded (NCT01971476); however, results are not yet available. Despite these promising preclinical results, there are no clinical trials examining combination treatments with PLK1 inhibitors in (pediatric) patients with rhabdomyosarcoma. However, based on the preclinical data and the tolerable toxicity of the PLK1 inhibitor NMS-1286937 in adult patients with solid tumor, such trials would be of interest (69).
Wee1.
An in vitro drug screen identified compounds with potential antitumor effects in rhabdomyosarcoma. The monotherapeutic and combined effects of cyclophosphamide, irinotecan, etoposide, dactinomycin, virorelbine, the Wee1 inhibitor AZD1775, the multi-RTK inhibitor cabozantinib, and the protease inhibitor bortezomib were tested. The results showed that for both rhabdomyosarcoma subtypes, AZD1775 combined with cabozantinib or bortezomib was most effective (70). In line with these findings, a multi-drug screen of pediatric orthotopic patient-derived xenograft (O-PDX) models also showed high sensitivity of ARMS and ERMS for AZD1775. Moreover, the combination of AZD1775 with irinotecan and vincristine improved tumor response (71). In line with the preclinical findings, the combined effects of AZD1775 and irinotecan hydrochloride will be examined in pediatric and AYA recurrent solid tumors, including rhabdomyosarcoma (NCT02095132; Table 2). In addition, AZD1775 combined with chemotherapy will be examined in adult solid tumor patients (NCT00648648; Table 2; ref. 72).
CDK.
CDK inhibitors could also be of interest for combination treatment in rhabdomyosarcoma. This could particularly be the case in ARMS, because activation of cyclin D1/CDK4 leads to PAX3-FOXO1 activation. Preliminary data described synergism for the combined treatment of CDK inhibitors with IGF1R or Wee1 inhibitors in rhabdomyosarcoma cell lines (73, 74). In contrast, combined treatment of the CDK4/6 inhibitor palbociclib with doxorubicin showed antagonism in an ERMS cell line. A possible biomarker for response could be expression of the cell-cycle–associated retinoblastoma protein (Rb). High Rb expression correlated with a higher response to the combination of palbociclib with either doxorubicin or AZD1775, whereas knockdown of Rb led to antagonistic effects (74). One phase I study will examine the combined effects of the CDK4/6 inhibitor ribociclib and doxorubicin in adult unresectable STS and this might give more insight into the clinical potential of this combination in rhabdomyosarcoma (NCT03009201; Table 2).
PAX3-FOXO1
Compared with the abovementioned processes, targeting of the PAX3-FOXO1 fusion in ARMS could lead to better results with a lower risk of treatment resistance. Direct targeting of this fusion protein has, however, proven difficult, hence multiple studies aimed at targeting pathways or proteins regulated by this oncogenic driver. However, as the fusion orchestrates multiple cellular processes and influences expression of many target genes, which are difficult to target simultaneously, robust antitumor responses in ARMS are difficult to achieve in this way. In this light, direct silencing of the fusion with PAX3-FOXO1-silencing RNA (siRNA) could be a more effective approach, as it would instantly deprive ARMS of its driver. Delivery of the siRNA to the tumor cell remains, however, an obstacle. A promising way of delivery could be a liposome-protamine-siRNA particle (LRP) that carries siRNA over the lipid barrier of the tumor cell. It has already been shown that ARMS can be targeted with LRPs when conjungated to cyclic RGD (arginine-glycine-aspartate) peptides, targeting the overexpressed αBβ3 integrin receptor. RGD-LRP particles loaded with anti-PAX3-FOXO1 siRNA (RGD-P3F-LRP) were effectively delivered to ARMS cell lines and led to reduction of both PAX3-FOXO1 and PAX3-FOXO1-target gene expression. Some effect on cell viability was seen, but no clear induction of apoptosis. In vivo, tumor outgrowth was delayed compared with the control, but effects on established tumors were limited, possibly due to the incomplete reduction of PAX3-FOXO1 expression following treatment (75). Because no other treatments are currently capable of directly targeting this fusion protein, the developments in nanomedicine are of interest, although the current data do not suggest fast implementation in the clinic. However, once the antitumor effects and bioavailability are optimized for clinical use, this could be a novel and, likely, highly effective treatment for patients with ARMS.
Epigenome
The epigenome, alongside PAX3/7-FOXO1 gene transcription, is responsible for the expression profile in rhabdomyosarcoma. Epigenetic changes affect chromatin, resulting in a more open or more closed conformation. Open conformations allow for gene transcription while closed conformations repress gene expression (Fig. 1B). In tumors, epigenetic alterations lead to the repression or induction of cancer-related genes. Several processes, including DNA methylation and histone modification, are involved in epigenetic regulation (76).
Histone deacetylase.
Histone deacetylases (HDAC) regulate the structure of chromatin around histone proteins by deacetylating lysine residues, leading to a compact structure. In many tumor types high levels of HDACs were shown to repress tumor suppressor gene expression (76). In rhabdomyosarcoma, monotherapy with the HDAC inhibitor vorinostat had limited effects in vivo, which was similar to the effects observed in a phase I study in children with recurrent solid tumors. Nevertheless, HDAC inhibitors combined with multiple chemotherapeutics were shown to have added effects in rhabdomyosarcoma cell lines (77–79). The combination with doxorubicin showed the highest synergism and reduced rhabdomyosarcoma tumor growth in vivo (78).
Dual and triple combinations, including vorinostat, were examined in ERMS. Multiple targeted therapies (vorinostat, 17-DMAG, sorafenib, abacavir) were combined with each other and/or with doxorubicin. Dual combinations of vorinostat with the multi-RTK inhibitor sorafenib, and sorafenib with the HSP90 inhibitor 17-DMAG were synergistic. In contrast, triple combinations with doxorubicin, 17-DMAG, and vorinostat or sorafenib did not show a clear increase in antitumor effects (80). One study examined multiple combinations in an ARMS PDX model based on the genomic and proteomic profile of the tumor tissue, showing among others high Sirt1 (NAD+ HDAC) levels. In vivo combination treatments were designed and the combination of the HDAC inhibitor entinostat and the chemotherapeutic docetaxel was most effective, with 65% tumor regression. Monotherapy entinostat led to a 43% tumor regression (81).
In addition to HDAC, lysine-specific demethylase 1 (LSD1) is overexpressed in rhabdomyosarcoma. Combination treatment of LSD1 and HDAC inhibitors showed synergism in ARMS and ERMS cell lines (82). HDAC inhibitors in combination with either targeted therapies or chemotherapy are in clinical trials (NCT01106872, NCT00937495, NCT01132911, NCT01294670; Table 2) and a subset of patients with advanced STS responded to monotherapy and combination treatments (NCT00878800; Table 2; refs. 83, 84). Both the preclinical and early clinical trial data show potential for epigenetic inhibitors combined with chemotherapy. It would be interesting to see whether epigenetic inhibitors alter the expression of repressed genes as this could lead to new combination treatments, possibly enhancing the antitumor effects of epigenetic inhibitors.
Epigenetic modification and immunotherapy.
In line with this hypothesis, demethylating 5-aza-2′-deoxytidine (DAC) enhanced the expression of tumor antigens in rhabdomyosarcoma cell lines. DAC enhanced mRNA expression of immunogenic cancer-testis antigens MAGE-A1, MAGE-A3 and NY-ESO1. Of note, mRNA expression did not necessarily correlate with protein expression, as only MAGE-A1 and –A3 were expressed in ARMS, whereas mRNA levels increased in both subtypes. The recognition-related proteins MHC class I–II molecules and costimulatory ICAM-1 were also increased, as was the reactivity of cancer-testis antigen-specific CTLs. This indicated that, despite a low cancer testis-antigen increase, treatment of demethylating agents can enhance the T-cell response against the tumor (85). No in vivo examination has been performed. However, these findings do suggest that targeting of the epigenome can alter protein expression and make tumor cells more vulnerable to other treatments, such as immunotherapy.
Discussion
As can be appreciated from the abovementioned studies, the potential of combination therapy has been examined in many different fields of targeted therapy. Recent studies show encouraging data, suggesting that a variety of targeted therapy–based combinations are capable of increasing antitumor effects in rhabdomyosarcoma. The majority of these studies focused on the potential of combinations with standard-of-care chemotherapeutics. In this regard, targeting of RTKs, downstream proteins, Hedgehog signaling, DDR, cell-cycle proteins and the epigenome shows promise, especially as in many cases synergy was achieved with low-dose chemotherapy, possibly lowering adverse events in patients. Because chemotherapy is and will for the foreseeable future remain a vital part of rhabdomyosarcoma treatment, these drug combinations are of importance.
However, as chemotherapy-based treatments remain toxic, noncytotoxic-based targeted combination regimens have been explored as well, although to a lesser extent. Some combinations targeting associated pathways, such as combined IGF1R and mTOR inhibition, show some efficacy in rhabdomyosarcoma. Simultaneous targeting of multiple cancer hallmarks that are tumor-specific and have complementary roles in tumor cell survival could hypothetically lead to more robust antitumor effects with lesser adverse effects on healthy tissue. Indeed, combinations targeting multiple distinct signaling pathways or different cellular processes show potential for rhabdomyosarcoma treatment, including combined Hedgehog and mTOR inhibition, and HDAC inhibitors combined with multi-RTK inhibitors. The combination of certain epigenetic modifiers with immunomodulating drugs are also worthy of further assessment. More research into this field is eagerly awaited. In addition to the targeted therapies currently used in combination treatments, other targeted therapies, such as those directed against checkpoint kinase 1 (CHK1) and enhancer of zeste homolog 2 (EZH2), could be of interest for future combination regimens. Preclinical data have shown promising single-agent efficacy and combination treatments might be able to reduce treatment resistance and/or enhance the antitumor effects (86, 87).
This review also highlights some recent developments in nanomedicine and drug conjugates. Even though the current compounds are not yet ready for clinical implementation, these could elicit high antitumor effects or even eliminate the ARMS driver. Further optimization of these compounds is therefore needed and examination of their potential in rhabdomyosarcoma treatment should be continued.
Of note, most studies, both preclinical and clinical, showed their effects to be present in a subset of cell lines, xenografts or patients, underlining the heterogeneous response these combination treatments can generate. Discovery of predictive biomarkers could decrease unnecessary treatment of patients and might prevent unnecessary toxicities. Identification of biomarkers in rhabdomyosarcoma remains difficult, however, and more research is necessary. Worthy of comment are the recent insights from a variety of genomic, epigenomic, and (phospho)proteomic projects. Although the majority of patients with rhabdomyosarcoma have a low mutational burden, a recent report did reveal a number of activating mutations and epigenetic alterations that could have an effect on pathway activity (28). Also a phosphoproteomics screen in rhabdomyosarcoma cell lines, and a combined genomic and morphoproteomic screening of an ARMS sample identified and validated adequate treatment strategies (7, 81). This exemplifies that molecular genomics, epigenomics, and (phospho)proteomics might have a place in rhabdomyosarcoma diagnosis to provide patients with the best, most personalized treatment available.
Altogether, a variety of targeted therapy–based combination treatment regimens show promise for patients with rhabdomyosarcoma. Although more research is required into the most suitable combinations accompanied with studies identifying predictive biomarkers, adequate implementation of such combined regimens in future clinical trials could improve outcome and reduce side effects of patients with rhabdomyosarcoma.
Disclosure of Potential Conflicts of Interest
No potential conflicts of interest were disclosed.
Footnotes
Note: Supplementary data for this article are available at Molecular Cancer Therapeutics Online (http://mct.aacrjournals.org/).
- Received November 15, 2017.
- Revision received February 27, 2018.
- Accepted May 1, 2018.
- Published first July 2, 2018.
- ©2018 American Association for Cancer Research.
References
- 1.↵
- 2.↵
- 3.↵
- 4.↵
- 5.↵
- 6.↵
- 7.↵
- 8.↵
- 9.↵
- 10.↵
- 11.↵
- 12.↵
- 13.↵
- 14.↵
- 15.↵
- 16.↵
- 17.↵
- 18.↵
- 19.↵
- 20.↵
- 21.↵
- 22.↵
- 23.↵
- 24.↵
- 25.↵
- 26.↵
- 27.↵
- 28.↵
- 29.↵
- 30.↵
- 31.↵
- 32.↵
- 33.↵
- 34.↵
- 35.↵
- 36.↵
- 37.↵
- 38.↵
- 39.↵
- 40.↵
- 41.↵
- 42.↵
- 43.↵
- 44.↵
- 45.↵
- 46.↵
- 47.↵
- 48.↵
- 49.↵
- 50.↵
- 51.↵
- 52.↵
- 53.↵
- 54.↵
- 55.↵
- 56.↵
- 57.↵
- 58.↵
- 59.↵
- 60.↵
- 61.↵
- 62.↵
- 63.↵
- 64.↵
- 65.↵
- 66.↵
- 67.↵
- 68.↵
- 69.↵
- 70.↵
- 71.↵
- 72.↵
- 73.↵
- 74.↵
- 75.↵
- 76.↵
- 77.↵
- 78.↵
- 79.↵
- 80.↵
- 81.↵
- 82.↵
- 83.↵
- 84.↵
- 85.↵
- 86.↵
- 87.↵