ES is a fast-growing tumor, forming osteolytic lesions, which can lead to bone pain and sometimes to pathological fractures. The clinical picture of ES is not very distinctive, which often leads to a delay in diagnosis, ranging from several weeks to several months. The first signs that may lead to consultation are the appearance of swelling in the affected bone, associated with a slight pain that may become more pronounced at night or following physical activity. In more advanced cases, pain and the presence of a mass are accompanied by less specific symptoms, such as fever, fatigue or weight loss, which may be a sign that the tumor has become metastatic (Biswas et al., 2014; Pizzo et al., 2015).

Diagnosis includes imaging studies before confirmation with surgical biopsy and histological and molecular analyses. Imaging consists of conventional radiography, through which it is possible to see the osteolytic damage created by the tumor mass in the diaphyseal-metaphyseal bone (Riggi et al., 2021). Magnetic Resonance Imaging (MRI) is then generally prescribed, which can be combined with a tomographic examination. Thanks to better spatial resolution and contrast, this allows better visualization of calcifications and tumor extension into adjacent bone and soft tissue. A technique combining 18F-fluorodeoxyglucose (FDG), PET-scan and tomography can also be used to assess tumor regression or progression upstream of MRI (Gerth et al., 2007). It has been shown that this technique can also be used to assess the presence of spinal cord metastases, in order to avoid spinal cord puncture (Newman et al., 2013; Kopp et al., 2015; Kasalak et al., 2018).

Histological features of the sample include the observation of small, round, undifferentiated cells. These cells have a prominent nucleus and sparse cytoplasm with glycogen deposits. Classically, the marker CD99 (Cluster of Differentiation 99), a transmembrane glycoprotein, is used for the diagnosis of ES (Ambros et al., 1991). Other markers can also be used, such as CD57 (Cluster of Differentiation 57) and synaptophysin, which are neuronal markers (Riggi and Stamenkovic, 2007). However, it should be noted that CD99 is not specific to ES as it can be a marker in other round cell sarcomas and even leukemias (Baldauf et al., 2018). The definitive diagnosis relies on the identification of the fusion protein by fluorescent in situ hybridization (FISH) or even quantitative Polymerase Chain Reaction (qPCR) techniques.

During the progression of ES, 20%–25% of patients may develop metastases at diagnosis (Grünewald et al., 2018; Strauss et al., 2021) to the lung (10%), bone (10%) or other sites (5%), spreading via the bloodstream (Casali et al., 2018) (Figure 2A).

Although many diverse processes are described to drive the metastatic development of ES, multiple studies report the crucial role of the EWS-FLI1 fusion protein in intra-tumor heterogeneity and in the ability of cells to migrate to metastatic sites (Franzetti et al., 2017; Sheffield et al., 2017; Aynaud et al., 2020). Indeed, cells with high EWS-FLI1 expression are rather undifferentiated and proliferate rapidly, whereas cells with low EWS-FLI1 expression have a more mesenchymal phenotype and are inclined to migrate and metastasize (Figure 2B). In this context, several studies have shown that EWS-FLI1 controls the metastatic potential of cells by regulating the organization of the actin cytoskeleton (Amsellem et al., 2005; Chaturvedi et al., 2012; 2014). Cells with low EWS-FLI1 expression show a loss of E-cadherin expression (Jolly et al., 2019) in favor of high expression of mesenchymal markers such as N-cadherin and Slug (Chaturvedi et al., 2012). EWS-FLI1 silencing also leads to decreased expression of cell-cell adhesion proteins such as tight junctions (CLD1, Claudin-1; OCL, Occludin) or desmosomes (DSP, Desmoplakin; PKP1, Plakophilin-1) (Franzetti et al., 2017).

The presence of metastases at the time of diagnosis is the most important prognostic factor, significantly reducing the probability of survival from around 70% at 5 years when the tumor is localized to less than 30% for patients with metastases (Takenaka et al., 2016; Li et al., 2022). Other factors, such as the location of metastases, the patient’s age and the location of the primary tumor (Hu et al., 2021), also play a part in prognosis. Indeed, the presence of bone rather than lung metastases (Casali et al., 2018; Strauss et al., 2021), adolescents/young adults rather than children aged 0–14 (Desandes and Stark, 2016), and a primary tumor located in the pelvis, sacrum or coccyx rather than in the long bones have a worse prognosis (Hu et al., 2021).

According to European guidelines, the therapeutic management of ES is as follows (Strauss et al., 2021): whether the tumor is localized or metastatic, the patient will undergo initial neoadjuvant chemotherapy, followed by local tumor therapy through surgical removal of the tumor and then adjuvant chemotherapy. Until recently, standard treatment in Europe combined 4 agents: Vincristine, Ifosfamide, Doxorubicin and Etoposide (VIDE) (Juergens et al., 2006; Ladenstein et al., 2010). If the primary tumor is large, or if the response to treatment is poor, the patient may be prescribed consolidation treatment with high-dose chemotherapy based on Vincristine, Actinomycin D and Ifosfamide (VAI) or Cyclophosphamide (VAC) or Busulfan and Melphalan. Because ES is a radiation-sensitive tumor (Gaspar et al., 2015), surgical resection can be supplemented with postoperative radiation therapy or even replaced by radiation therapy if the tumor is inoperable because in a site difficult to reach. Nonetheless, patients who have undergone surgical excision as well as radiation therapy have a decreased risk of recurrence compared to patients who have only undergone radiation therapy, although there does not appear to be a significant effect on overall patient survival (Foulon et al., 2016).

New therapeutic strategies are being explored to improve the prognosis and survival of ES patients, which have not improved over the past 20 years.

The EuroEwing 2012 clinical trial was aimed to compare the efficacy, survival, and toxicity of European standardized chemotherapy with the protocol used in the United States from the Children’s Oncology Group (AEWS0031 clinical trial, (Womer et al., 2012; Anderton et al., 2020). Patients on this protocol receive alternating cycles of Vincristine - Doxorubicin - Cyclophosphamide and Ifosfamide - Etoposide (VDC/I.E.,) as induction chemotherapy and then alternating cycles of Ifosfamide - Etoposide and Vincristine - Cyclophosphamide (I.E.,/VC). Initial results from this clinical trial indicate that the US protocol is more effective in terms of overall survival and event-free survival, less toxic, and shorter in duration for newly diagnosed cases of ES (Brennan et al., 2020; Brennan et al., 2022). Since this study, the interval compressed V(D)C/I.E., schedule is now standard of care across Europe and the treatment backbone for the upcoming academic trials in ES. Other clinical trials aimed at improving response to chemotherapy using other chemotherapeutic agents, alone or in combination with various drugs, are currently being studied or enrolled (Supplementary Table S1). For example, the Phase II/III multi arm multi stage clinical trial for relapsed refractory ES (rEECur) was evaluating the effect of Topotecan and Cyclophosphamide (TC), Irinotecan and Temozolomide (IT), Gemcitabine and Docetaxel (GD), or high-dose Ifosfamide (IFOS) in the treatment of recurrent and refractory ES (RR-ES) (ISRCTN36453794). The first results showed that IFOS is more effective in prolonging survival than TC, having previously overcoming GD and IT, and should be considered as a control arm in future randomized phase II/III studies in RR-ES if combination with IFOS makes sense. This trial is currently recruiting patients to IFOS and Cyclophosphamide/Etoposide regimens, and an additional arm with a molecularly targeted agent is planned, such as Multi-Tyrosine Kinase Inhibitors (MTKI) (McCabe et al., 2022).

The chromosomal translocation t (11; 22) (q24,q12) generates the EWS-FLI1 fusion gene (under the control of the EWS promoter), composed of the N-terminal domain of the EWS gene and the C-terminal domain of FLI1 (Figure 3A). Depending on the position of the breakpoints in the different genes, over 10 different EWS-FLI1 transcripts have been described in the literature (Zucman et al., 1993), the majority being between exon 7 of the EWS gene and exon 6 of the FLI1 gene (Delattre et al., 1992), or between exon 7 of the EWS gene and exon 5 of the FLI1 gene (Delattre et al., 1992; Zucman et al., 1993). It should be noted that the reciprocal transcript (FLI1-EWS) also exists, but until recently was considered to be little or not expressed in ES cells (Zucman et al., 1993). Indeed, EWS-FLI1 is expressed under the control of the EWS promoter, whereas FLI1-EWS is expressed under the control of the FLI1 promoter, which is not very active in ES (Zucman et al., 1993). However, a recent study shows that FLI1-EWS expression in some ES cell lines is involved in the regulation of ES cell proliferation (Elzi et al., 2015).

This EWS-FLI1 fusion gene leads to the formation of the EWS-FLI1 fusion protein, whose activator domains are more potent than those of FLI1 alone as a result of N-terminal domain substitution during chromosomal translocation (May et al., 1993). This property enables EWS-FLI1 to act as an aberrant transcription factor, thus activating the expression of numerous oncogenes. After translocation into the cell nucleus, EWS-FLI1 binds to its consensus sequences located at the promoters of its target genes. Using chromatin immunoprecipitation experiments, several studies have identified the DNA-binding motif of EWS-FLI1 in vitro and in vivo (Gangwal et al., 2008; Guillon et al., 2009; Boeva et al., 2010). EWS-FLI1 is known to bind to the canonical DNA-binding motif of the ETS protein family, the GGAA sequence (Boeva et al., 2010), but it is also capable of binding to GGAA microsatellites, according to the following rule: “the greater the number of repeats, the more the target gene will be expressed”, within the limit of 20 GGAA repeats (Guillon et al., 2009; Johnson et al., 2017) (Figure 3B). In this context, a 2014 study demonstrates that the regulation of target genes by EWS-FLI1 depends closely on the number of repeats of the GGAA motif at the DNA-binding domain: genes activated by EWS-FLI1 are associated with GGAA microsatellites (motif repeated at least 4 times), while genes repressed by EWS-FLI1 are associated with the canonical DNA-binding motif of the ETS protein family (1 GGAA sequence) (Riggi et al., 2014).

As this ability to bind to GGAA microsatellites is not observed for FLI1 alone, this confers on EWS-FLI1 specific functions enabling the activation of de novo enhancers through the recruitment of BAF (BRG1-or BRM-Associated Factors), a chromatin remodeling complex (Boulay et al., 2017). This modulates the expression of numerous target genes, such as LOXHD1 (Deng et al., 2022), or activates the expression of genes not expressed in healthy tissues or other tumor tissues outside ES (Vibert et al., 2022).

EWS-FLI1 expression can be regulated at transcriptional, translational and post-translational levels, and via protein-protein interactions (Yu et al., 2023).

At the transcriptional level, EWS-FLI1 expression is promoted by methylation or acetylation of histone 3 (H3K4me3, H3K9ac or H3K27ac) (Montoya et al., 2020) and by the binding of the transcription factor SP1 (Specificity Protein 1) to its promoter. Conversely, expression of miR-145 inhibits EWS-FLI1 transcription (Ban et al., 2011).

EWS-FLI1 translation is modulated by compounds such as Lovastatin or Tunicamycin, which reduce the protein level of EWS-FLI1, thus reducing primary tumor growth (Wang et al., 1999; Herrero-Martin et al., 2011), but increasing a migratory phenotype (Franzetti et al., 2017). Numerous post-translational modifications can take place on EWS-FLI1, resulting in overexpression or degradation of the fusion protein. For example, phosphorylation of threonine 79 (T79) (Klevernic et al., 2009) or O-GlcNAcylation of EWS-FLI1 (Bachmaier et al., 2009) stimulates the oncogenic function of EWS-FLI1, while its ubiquitination leads to its degradation by the proteasome (Gierisch et al., 2016).

Finally, EWS-FLI1 is known to interact with numerous protein players regulating its transcriptional activity. For example, it can bind with RHA (RNA Helicase A) to increase its transcriptional activity (Toretsky et al., 2006), or with PARP-1 to facilitate transcription (Brenner et al., 2012). Conversely, its interaction with CIMPR (Cation-Independent Mannose 6-Phosphate Receptor) leads to its degradation via a lysosome-dependent pathway (Elzi et al., 2015).

Through its function as a transcription factor, EWS-FLI1 regulates the expression of numerous oncogenes involved in many key tumorigenesis processes (cell proliferation, migration, apoptosis, etc.). In particular, EWS-FLI1 has been shown to regulate the expression of numerous transcription factors (Cidre-Aranaz and Alonso, 2015). For example, EWS-FLI1 stimulates the expression of c-Myc (Dauphinot et al., 2001), Gli1 (Glioma-associated Oncogene Homolog 1) (Merchant et al., 2009; Mullard et al., 2020) or NR0B1 (Nuclear Receptor subfamily 0 group B member 1) (Kinsey et al., 2006; Gangwal et al., 2008), genes involved in tumor progression and metastatic development. Conversely, EWS-FLI1 reduces the expression of genes involved in tumor suppressor mechanism (ex: apoptosis), such as FOXO1 (Forkhead Box O1) (Yang et al., 2010), or IER3 (Immediate Early Response 3) (Tsafou et al., 2018), tumor suppressor genes regulating mechanisms such as DNA repair, cell cycle arrest and apoptosis.

By recruiting the BAF complex (Boulay et al., 2017), EWS-FLI1 plays a major role in chromatin accessibility at the promoter and enhancer regions of various genes (Tolstorukov et al., 2013) and its role in regulating chromatin remodeling is diverted, like p300 (Riggi et al., 2014), towards aberrant transcription of oncogenes. The GGAA microsatellites therefore appear to be a potential therapeutic target for Ewing sarcoma. Their epigenetic silencing (H3K9me3 labelling) inhibited EWS-FLI1 binding to the SOX2 enhancer (SRY-box transcription factor 2), thereby reducing its expression and altering tumor growth in vivo (Boulay et al., 2018). More recently, EWS-FLI1 has been shown to be directly involved in changes in chromatin configuration (Showpnil et al., 2022). Indeed, through its binding to DNA, via GGAA microsatellites, EWS-FLI1 plays a major role in reprogramming the 3D structure of chromatin, by disrupting, for example, TADs (Topological Associated Domains, highly self-interacting genomic regions, delimited by regions enriched in CCCTC binding factor (CTCF) and involved in the regulation of gene expression by limiting enhancer-promoter interactions within the same TAD (Dixon et al., 2012)), or by modulating chromatin loops, with the direct consequence of altering transcription in ES (Showpnil et al., 2022).

Finally, EWS-FLI1 is also involved in ES tumorigenesis by inducing genomic instability. Indeed, it has been shown that EWS-FLI1-induced transcriptional regulation promotes an accumulation of R-loops (Gorthi et al., 2018). These triple-stranded nucleic acid structures (DNA-RNA complex associated with a single-stranded DNA molecule) are rare products of transcription, with a particular conformation and influencing many cellular processes (Aguilera and García-Muse, 2012). In ES, these structures sequester the BRCA1 (Breast cancer type 1 susceptibility) gene, preventing its expression, and its crucial role in the response to DNA damage (Gorthi et al., 2018).