2.1. Insulin-like Growth Factor 1 Receptor (IGF1R) in ES

IGF1R is a transmembrane receptor composed by two subunits, ⲁ and β, connected by disulphide bonds. The first component, the ⲁ subunit, is responsible for ligand binding whereas the other, the β subunit, upon autophosphorylation leads to the activation of several intracellular pathways involved in cell proliferation and survival such as the mitogen-activated protein kinase (MAPK or MAP kinase) pathway and the phosphatidylinositol-3-kinase (PI3K)/Akt and the mammalian target of rapamycin (mTOR) pathway. Both pathways drive oncogenic transformation, proliferation, and survival [11].

The relation between the fusion protein EWSR1-ETS and IGF1R pathway is complex and not completely understood.

Preclinical evidence has shown that IGF1R is ubiquitously expressed in neuroectodermal-originating tumors and its activation is sustained by the autocrine production of insulin-like growth factor-1 (IGF-1), one of the IGF1R ligands, by the cancerous cells themselves [12].

Furthermore, EWSR1-ETS acts both as a transcription factor of the IGF-1 and IGF-2 encoding genes and as a negative regulator of the sequester of the circulating IGF-1, the insulin-like growth factor binding protein 3 (IGFBP3). In fact, EWSR1-ETS once binding IGFBP3 promoter reduces its expression and as a consequence IGF-1 can bind IGF1R [13,14,15].

The prognostic role of IGF1R expression has been proved in clinical samples of ES. Indeed, a better prognosis and a lower tumor proliferative rate when IGF1R is less expressed has been demonstrated [16].

The existence of this crosstalk has paved the way to the investigation, as has the development, of antibodies and small molecules, both in preclinical and clinical stages, that could interfere with this biological interaction (Figure 1A).

Open in a separate window

Figure 1

(A) EWSR1-ETS transcription activity enhances the transcription of IGF1 and reduces the transcription of IGFBP3; IGF1 binds to IGF1R and activates MAPK pathway leading to proliferation and survival; antibodies and small molecules that inhibit IGF1R signaling. (B) The mRNA, derived from the transcription of the fusion gene, is translated into the oncogenic fusion protein EWSR1-ETS that elicits its downstream effect; the mRNA translation process is stopped by oligos, and siRNAs and the fusion protein is no longer produced. (C) EWSR1-ETS depends on RHA to enhance its transcriptional activity; YK-4-279 can alter the interaction between EWSR1-ETS and RNAH resulting in transcription stop. (Created with BioRender.com).

• Monoclonal antibodies directed against IGF1R

R1507, a monoclonal antibody (mAb) directed against IGF1R, has proven to be active in the growth inhibition of ES cells that expressed high levels of IGF-2 and has shown encouraging activity and a durable benefit in a phase II trial, with acceptable toxicities. In this trial the OS was 6, 9 months with an overall response rate (ORR) of 10% and a duration of response (DOR) of 28 weeks [17].

IGD1R inhibitors have gained momentum and apart from R1507, other antibodies have demonstrated activity against IGF1R: dalotuzumab, known as MK-0646, also showed in vitro and in vivo activity against ES cells, as well as IMC-A12 (cixutumumab), AMG 479 (ganitumab), CP-751, 871 (figitumumab), and SCH-717454 [18,19].

In a phase II trial ganitumab has demonstrated a median PFS of 7.9 months in patients with metastatic ES [20].

In the light of the potential synergism between IGF1R inhibitors (IGF1Ri) and conventional chemotherapy, the addition of ganitumab to chemotherapy in newly diagnosed metastatic ES patients was investigated in a phase III trial [21].

Unfortunately, no event-free survival (EFS) benefit with the addition of the anti-IGFR1 mAb to interval-compressed chemotherapy, consisting of vincristine/doxorubicin/cyclophosphamide alternating once every two weeks with ifosfamide/etoposide (CDV/IE), was seen [21].

Increased toxicity, mainly febrile neutropenia and the elevation of alanine aminotransferase, was associated with ganitumab.

Moreover, figitumumab, another mAb directed against IGF1R, has shown preliminary encouraging results in a phase II trial in patients with ES, particularly when elevated blood IGF-1 levels were detected at the baseline. As a matter of fact, a longer OS was observed in subjects with baseline IGF-1 levels > 110ng/mL compared to patients with lower IGF-1 levels at the baseline. However, the mPFS in the overall population was very poor reaching only 1.9 months [22,23]. This finding suggests that a better selection of patients should be pursued to observe better results. As an example, preliminary analysis of IGF-1 levels may represent a strategy to understand which patients would benefit the most from IGF1R inhibition.

The modest activity of these drugs, emerging from the above-described trials, could probably be due to the heterogeneity of IGF1R distribution on the cells surface and to the overexpression of IGFBP3 which represent a secondary resistance mechanism.

In order to increase the clinical activity of IGF1Ri, their association with other drugs, such as mTOR inhibitors, have been investigated. In fact, pre-clinical data suggest that a deletion of the phosphatase and tensin homolog on chromosome 10 (PTEN), involved in the mTOR pathway, can be identified in 25% of patients with ES [24]. This genetic alteration confers resistance to IGF1Ri explaining the rationale of combining these two drugs.

Furthermore, the IGF1R overexpression represents a resistance mechanism to different conventional and not conventional drugs.

As an example, IGF1R is a mediator of resistance to target drugs, such as cyclin dependent kinase 4/6 inhibitors (CDK4/6i).

The CDK4/6 pathway is active in several cancers and shows a dependency in ES on the IGF1R pathway.

In the cell cycle regulation both CDK4 and CDK6 mediate the transition from the G1 phase to the S phase in order to prime DNA replication. Upon activation by the CDK-activating kinase (CAK) complex, both CDK4 and CDK6 inhibit, by phosphorylation, retinoblastoma (RB), a tumor suppressor protein. RB exerts an inhibitory function on the E2F family of transcription factors whose target genes are implicated in cell proliferation [25].

The consequences of CDK4/6i are the blockade of cell proliferation caused by an arrest in the G1 phase.

Ganitumab, and other IGF1Ri, have shown to be synergistic with CDK4/6i in vitro and active in ES mouse models. For this reason, ganitumab is under investigation in combination with palbociclib, a CDK4/6i. The aim of the trial is to verify if this combination is safe and active against ES [26].

Furthermore, it was demonstrated that IGF1R signaling could protect cancerous cells from the cytotoxic effect of several chemotherapeutic agents active in sarcomas.

Indeed, the efficacy of vincristine, trabectedin, and doxorubicin was proven to increase when IGF1R is inhibited [27,28,29].

• Tyrosine kinase inhibitors against IGF1R

Apart from mAbs, IGF1R can also be inhibited by small molecules, namely inhibitors of the intracellular tyrosine kinase domain of IGF1R.

One established resistance mechanism to IGF1R mAbs is represented by the compensatory activation of other RTKs as described below.

Linsitinab, a dual inhibitor of both IGF1R and insulin receptor (IR), whose activation is recognized as one resistance mechanism to anti-IGF1R drugs, has demonstrated to decrease the phosphorylation of the IGF1R thus inducing a reduction of ES cell proliferation.

Its good pharmacokinetic profile made this drug well tolerated [30].

Besides Linsitinib (OSI-906), ADW742 and NVP-AEW54, alone or in combination with conventional CT, have been demonstrated to inhibit cell growth in in vitro and in vivo studies [30,31,32].

In particular, NVP-AEW541 has demonstrated to selectively inhibit IGF1R by distinguishing between IGF1R and other RTKs leading to a more sustainable toxicity profile. Its effects on tumor growth were noticed when NVP-AEW541 was combined with vincristine, actinomycin D, and ifosfamide [32].

It is worth remembering that toxicity profiles of these agents must be considered.

Although IGF-1R mAbs and small inhibitors are well tolerated, particular attention should be paid to mild and not mild adverse effects.

Hyperglycemia was commonly described in all the pre-mentioned trials, probably because IGF-1R is involved in glucose metabolism. Two mechanisms could explain this expected toxicity: compensatory insulin secretion and, consequently, insulin resistance [33].

To summarize, although large amounts of preclinical and clinical data support the use of agents targeting the IGF1R pathway, the real benefit of these compounds is questionable and seems not to apply to every patient. This data suggests that it is possible that combined therapies may represent a suitable strategy to enhance the, at least for now, disappointing anti-tumor activity of anti-IGF1R agents in monotherapy. Predictive biomarkers of the response to IGF1Ri are not available but there is a need to better select patients that may benefit from these drugs.

3.1. Decreasing EWSR1-FLI1 Expression

In order to decrease the expression of the fusion protein both small antisense oligodeoxynucleotides (oligos) and small interfering RNAs (siRNAs) have been investigated in in vitro experiments. The results have shown that the use of these agents against ES cells is safe, and it generates a positive tumor response.

The oligos bind complementary mRNA sequences, in this case the mRNA that codes for EWSR1-FLI1, to primarily stop its translation and subsequently induce its degradation by ribonuclease H [41].

The siRNAs are also able to switch off EWSR1-FLI1 by gene-silencing but the exact mechanism of knockdown is still an area of investigation [42].

It is widely accepted that, upon recognition of targeted sequences, siRNAs mediate a specific and selective cleavage of the target mRNAs interfering with their translation into functional proteins capable of driving oncogenic features such as proliferation, tumor development and progression, and apoptosis evasion [43].

One of these molecules is currently being evaluated in a clinical trial (NCT02736565).

Both oligos and siRNAs interfere with the translation of the fusion gene into the fusion protein and consequently reduce the downstream activity of the same. Although potentially promising agents, both oligos and siRNAs have a major limitation: because of their short half-life, they require repeated administration to achieve and sustain EWSR1-FLI suppression; an expensive and time-consuming procedure [44] (Figure 1B).

4.1. Histones Demethylation Inhibitors (HDMi)

EWSR1-FLI1 uses a specific enzyme, named lysine-specific demethylase 1 (LSD1) to repress some critical tumor suppressors, such as lectin-like oxidized low-density lipoprotein (LDL) receptor-1 (LOX1) and transforming growth factor beta receptor 2 (TGFBR2) [50].

LSD1 is known to be overexpressed in ES cancer cells [51] and can be inhibited by HCI2509.

Upon the inhibition of LSD1, a reduction of the expression of the pre-mentioned genes, LOX and TGFBR2, both involved in tumor suppression, occurs. The inhibition of LSD1 leads to an impairment of tissue culture cell viability in multiple ES cell lines as verified by some preclinical trials.

This drug was found to disrupt the transcriptional signature of EWSR1-FLI and EWS-ERG and to silence malignant characteristics of ES cells by reversal of global transcriptional activity of both fusion genes [50] (Figure 2A).

Open in a separate window

Figure 2

(A) EWSR1-ETS uses LDS1 to repress the expression of two target genes LOX and TGFBR2 that act as tumor suppressors; HCL2509 can inhibit LSD1 activity and reverse transcriptional activity of EWSR1-ETS. (B) EWSR1-ETS induces Id2 that releases E2f from RB inhibition; E2f induces cell cycle progression and PARP1 expression; replication and transcriptional stress lead to DNA damage that cannot be repaired either by PARP1, targeted by PARPi, nor by BRCA1 that is functionally absent in ES cells. (Created with BioRender.com).

A first report of an LSD1 inhibitor, seclidemstat (SP-2577), in a phase I trial focused exclusively on ES, has shown promising results with a manageable safety profile and a proof-of-concept of preliminary activity in heavily pretreated relapsed or refractory patients.

The most common adverse events related to the treatment were vomiting, abdominal pain, hypokalemia, and hematologic disorders [52].

Seclidemstat deserves more investigation in an already planned phase II expansion trial of seclidemstat, as a single agent and in combination with CT, topotecan and cyclophosphamide, in ES and other sarcomas that share similar translocations (NCT03600649).

Among other molecules that have shown to regulate epigenetic profiling in ES, second-generation histone demethylases (HDM) inhibitors, such as the pan-Jumonji inhibitor JIB-04, have proven activity in preclinical models of ES. This drug targets HDM3, a vector of a significant slowdown in the in vitro and in vivo growth of ES cells [53].

4.2. Histones Methyltransferase Inhibitors (HMTi)

An enhancer of Zeste Homologue 2 (EZH2) is involved in a wide range of biological processes, including cell cycle regulation, cell differentiation and proliferation, division, and senescence/apoptosis.

EZH2, the catalytic subunit of Polycomb Repressive Complex 2 (PRC2), is also crucial for maintaining the epigenetic state of the cells, thanks to the modulation of the chromatin structure and, consequently, the regulation of gene expression [54].

From a biological point of view, EZH2 is a histone methyltransferase that can be mutated across different cancers, both hematologic and solid.

A very small percentage of soft tissue sarcomas (STS), and among them ES, can exhibit a mutation in EZH2 or in another component of its belonging complex [55].

Tazemetostat, a EZH2 inhibitor that has shown to be active in epithelioid sarcoma (EpS) [56], has been also investigated in other solid tumors harboring alterations in EZH2 or the SWI/SNF complex.

Unfortunately, the study did not produce significant results, showing an ORR of only 5% [57].

6.2. Forkhead Box O (FOXO)1

Other two transcription factors have shown a potential role in the inhibition of ES cells when specific drugs were used to antagonize their activities.

Consolidated evidence suggests that the EWSR1 domain confers transcription activating properties to the EWSR1-FLI1 chimeric protein; whereas it is also known that the fusion protein has transcription repressing properties. EWSR1-FLI1-repressed genes show enrichment of forkhead box (FOX) recognition motifs.

Forkhead box O (FOXO)1 is a transcription factor involved in cell proliferation, differentiation, and apoptosis; it acts as a tumor suppressor that limits the growth of cancerous cells inducing cell-cycle arrest, apoptosis, and DNA repair [78].

FOXO1 is transcriptionally and post-transcriptionally repressed by EWSR1-FLI1 meaning that the re-activation of FOXO1 may represent a promising strategy for ES. Methylseleninic acid (MSA), a chemical compound that has proved to reactivate FOXO1 in prostate cancer [79], was investigated also in ES. A proof of principle study was undertaken demonstrating a dose- and time-dependent reactivation of endogenous FOXO1 after MSA administration. Furthermore, MSA has proved to induce cell death in a FOXO1 expression-dependent manner with a subsequent reduction in tumor growth in vivo [80].

6.3. Glioma-Associated Oncogene Homolog 1 (Gli1)

Moreover, Glioma-Associated Oncogene Homolog 1 (Gli1), a transcription effector of the canonical Hedgehog (HH) pathway, shown to be highly expressed in ES, was known to mediate the downstream effects of the fusion protein. Gli1 is a direct target of EWSR1-FLI1 and seems to mediate the developmental process and the maintenance of the stem-cell-reservoir confirming the aggressiveness of ES cells [81].

When Gli1 levels were reduced by Arsenic trioxide (ATO) in combination with other chemotherapeutic drugs, such as etoposide and paclitaxel, tumor growth seems to be controlled in 75% of cases [82].

Gli1 can also be pharmacologically inhibited by two compounds, GANT58 and NCS75503, which act as Gli1 antagonists [83].

7.1. Heat-Shock Protein 90 (HSP90)

The modulation of EWSR1-FLI1 protein expression may be reached with the inactivation of chaperones, small molecules that participate in protein homeostasis in normal cells. Among chaperones, Heat-shock protein 90 (HSP90) plays a crucial role in the response to protein misfolding and aggregation. When EWSR1-FLI1 proteins are produced, HSP90 works to guarantee the pre-mentioned homeostasis.

Zelavespib, a HSP90 inhibitor, disrupts this equilibrium and therefore there is an increase in EWSR1-FLI1 turnover and a decrease in ES cell viability, a mechanism that was preclinically proved in in vitro studies both alone and in combination with bortezomib [87].

The aim of this study was to investigate the activity of PU-H71, named zelavespib, in multiple patient-derived ES cell lines.

The drug demonstrates the inhibition of ES proliferation via activation of pro-apoptotic mechanisms.

HSP90 needs a chemical reaction in order to be activated; this reaction involves acetylation.

Taking into consideration what was previously said, entinostat or other histone deacetylase inhibitors, not only could be useful as epigenetic modulators, as previously mentioned, but also as a target of the protein homeostasis system [88] (Figure 3B).

7.2. Proteolysis Targeting Chimeric Molecules (PROTACs)

Degradation of EWSR1-FLI1 proteins may be also achieved through the interaction with Proteolysis Targeting Chimeric Molecules (PROTACs), even if no compound of this class has been tested in clinical trials yet.

PROTACs are bi-functional compounds comprising two parts: one portion has the role of binding specific target proteins while the other acts as a chaperone, conducting the targets to poly-ubiquitination and degradation [89].

Preclinical evidence shows promising activity of BETd, a PROTAC directed against BRD4, a transcriptional and epigenetic regulator essential during ES development [90].

7.3. Unfolded Protein Response and IRE1α-XBP1 Inhibitors

Biologically speaking, the synthesis, the correct folding and maturation of proteins, occurs in the endoplasmic reticulum (ER).

When misfolded or unfolded proteins are produced, they induce the so-called ER stress that enhances the activation of the unfolded protein response (UPR). In the signaling pathway of UPR, inositol-requiring enzyme 1 (IRE1α) plays a major role as it is a sensor protein which serves as a kinase and endoribonuclease located in the ER membrane [91].

During ER stress, IRE1α undergoes oligomerization and when tetramers are formed, the X-box binding protein 1 (XBP1) splicing reaction occurs.

In fact, upon activation of IRE1α, its ribonuclease (RNase) domain triggers the unconventional splicing of XBP1 (XBP1s) mRNA and, therefore, the homeostatic transcription factor XBP1s is produced. This mechanism promotes cell survival via upregulation of pro-survival pathways in cancer cells [92].

Proteomic studies of ES have showed that XBP1, a protein playing a critical role in the UPR in the ER, is highly expressed in surgical samples of ES.

IRE1α-XBP1 inhibitors, such as toyocamycin, 2-hydroxy-1-naphthaldehyde (HNA), STF-083010 (STF), and 3-ethoxy-5 6-dilbromosalicylaldehyde (3ETH), seem to suppress cell viability both in vitro and in vivo (Figure 3C).

As a matter of fact, IRE1α-XBP1 inhibitors may be useful therapeutic options for ES patients [93].

8.1. Immune Checkpoint Inhibitors (ICIs)

Immune checkpoint inhibitors (ICIs), such as pembrolizumab, a mAb directed against PD-1, have failed in demonstrating a significant clinical activity in adults with ES. The reasons why this agent did not confirm the same exciting results as in other solid tumors could be due to various factors: the low mutational burden, the lack of PD-L1 expression, the HLA loss, and the absence of reactive T-cell in tumor microenvironment that characterize ES cells.

Pembrolizumab failure stems from SARC028, the first prospective open-label phase II multicenter study where none of the thirteen enrolled ES patients showed any clinical benefit from anti-PD-1 blockade [99]. This failure in the ES cohort could be related to the highly suppressive immune microenvironment in the tumor as previously underlined. The fact that only one patient with ES had demonstrated stable disease (SD) suggests that immune biomarkers could help to select patients that may benefit, at least in part, from ICI single agent treatment. Furthermore, poor activity in the monotherapy of ICIs could be increased with combination therapies as discussed below.

In a phase I/II trial nivolumab, another PD-1 inhibitor, was investigated as monotherapy in terms of safety and anti-tumor activity in children and young adults with recurrent or refractory tumors including ES [100].

As for SARC028, ICI monotherapy with nivolumab did not show significant anti-tumor activity although it was well tolerated in terms of toxicities.

In order to increase the efficacy of ICIs in sarcomas, the immunological and immunomodulatory properties of conventional CT and target therapies (TT) have been investigated. Metronomic cyclophosphamide has been described as being able to restore T-cells and NK effectors while inhibiting T-regs [101].

Although apparently promising, this strategy has not led to important results in terms of OR and PFS rates [102]. In a single arm phase II trial axitinib, a selective TKI, was combined with pembrolizumab in patients with advanced or metastatic sarcoma. Most patients with partial response (PR) belonged to the alveolar soft part sarcoma (ASPS) group.

The benefit in the non-ASPS group, in terms of PFS and OR, was limited, raising the question of whether the addition of an ICIs has any value in non-ASPS histology [103]. It is worth noting that only one patient in the trial had a diagnosis of ES.

8.2. Tumor-Reactive T-Cells

In recent unsuccessful trials, tumor-reactive T-cells have been used but there is a main limit: ES cells downregulate MHC-1 on their surfaces reducing the recognition of intracellular antigen from T-cells as a defense mechanism to avoid the immune response.

MHC-1 expression may be induced by IL-12 or INF-γ, two cytokines that could play a fundamental role in future trials dealing with immunotherapy in ES [104]. Tumor-reactive T-cells can be obtained in two ways: one requires the selection of the autologous adoptive T-cell which recognizes the tumor followed by its expansion in vitro and eventually its transfer into the patient. Although intriguing, this process could be challenging, time consuming, and possibly inefficient. The other way consists in isolating TCRs from T-cells with proven tumor reactivity and using them to transduce polyclonal autologous T-cells, transferred after expansion into patients whose tumors express the given antigens. Several pre-clinical trials have tried to translate these approaches into practice with disappointing results [105,106,107]. It is possible that the identification of combination therapy and the implementation of strategies with the aim of rescuing HLA downregulation could increase ES cells sensitivity to these strategies.

8.3. Chimeric Antigen Receptor T-Cell (CAR-T)

In order to overcome these limits, the chimeric antigen receptor T-cell (CAR-T) has been used because of its independence from MHC-1 recognition. CAR-T cells are T-cells engineered to express CARs on their surfaces. CARs are made up by an antigen binding domain and a T-cell activation domain.

CAR-T cells have been developed against the vascular endothelial growth factor receptor 2 (VEGFR2), IGF1R [108], ganglioside2 (GD2) [109], hepatocellular receptor tyrosine kinase class A2 (EphA2) [110]; all these antigens have presented as highly expressed on ES cells’ surface. Preclinical studies targeting GD2 demonstrated a reduction in tumor volume and clinical trials with the more promising CAR-T cells have been initiated [111]. (Figure 4A,B)

Open in a separate window

Figure 4

(A) e (B) Immunotherapy in ES. (Created with BioRender.com).

This review was part of research activity of the Rare Tumors Coordinating Center of Campania Region (CRCTR), recognized as a full member of the European Reference Network (ERN-EURACAN). Illustrations have been realized with BioRender.com. The authors have obtained the licenses for each illustration for this publication.