Search

Diffuse midline gliomas are uniformly fatal pediatric central nervous system cancers that are refractory to standard-of-care therapeutic modalities. The primary genetic drivers are a set of recurrent amino acid substitutions in genes encoding histone H3 (H3K27M), which are currently undruggable. These H3K27M oncohistones perturb normal chromatin architecture, resulting in an aberrant epigenetic landscape. To interrogate for epigenetic dependencies, we performed a CRISPR screen and show that patient-derived H3K27M-glioma neurospheres are dependent on core components of the mammalian BAF (SWI/SNF) chromatin remodeling complex. The BAF complex maintains glioma stem cells in a cycling, oligodendrocyte precursor cell–like state, in which genetic perturbation of the BAF catalytic subunit SMARCA4 (BRG1), as well as pharmacologic suppression, opposes proliferation, promotes progression of differentiation along the astrocytic lineage, and improves overall survival of patient-derived xenograft models. In summary, we demonstrate that therapeutic inhibition of the BAF complex has translational potential for children with H3K27M gliomas. Significance: Epigenetic dysregulation is at the core of H3K27M-glioma tumorigenesis. Here, we identify the BRG1–BAF complex as a critical regulator of enhancer and transcription factor landscapes, which maintain H3K27M glioma in their progenitor state, precluding glial differentiation, and establish pharmacologic targeting of the BAF complex as a novel treatment strategy for pediatric H3K27M glioma.

Melanoma-derived brain metastases (MBM) represent an unmet clinical need because central nervous system progression is frequently an end stage of the disease. Immune checkpoint inhibitors (ICI) provide a clinical opportunity against MBM; however, the MBM tumor microenvironment (TME) has not been fully elucidated in the context of ICI. To dissect unique elements of the MBM TME and correlates of MBM response to ICI, we collected 32 fresh MBM and performed single-cell RNA sequencing of the MBM TME and T-cell receptor clonotyping on T cells from MBM and matched blood and extracranial lesions. We observed myeloid phenotypic heterogeneity in the MBM TME, most notably multiple distinct neutrophil states, including an IL8-expressing population that correlated with malignant cell epithelial-to-mesenchymal transition. In addition, we observed significant relationships between intracranial T-cell phenotypes and the distribution of T-cell clonotypes intracranially and peripherally. We found that the phenotype, clonotype, and overall number of MBM-infiltrating T cells were associated with response to ICI, suggesting that ICI-responsive MBMs interact with peripheral blood in a manner similar to extracranial lesions. These data identify unique features of the MBM TME that may represent potential targets to improve clinical outcomes for patients with MBM.

Supplementary Table from BAF Complex Maintains Glioma Stem Cells in Pediatric H3K27M Glioma

Supplementary Figure from BAF Complex Maintains Glioma Stem Cells in Pediatric H3K27M Glioma

Supplementary Figure from Microenvironmental Landscape of Human Melanoma Brain Metastases in Response to Immune Checkpoint Inhibition

Supplementary Data from Microenvironmental Landscape of Human Melanoma Brain Metastases in Response to Immune Checkpoint Inhibition

Diffuse midline gliomas are uniformly fatal pediatric central nervous system cancers that are refractory to standard-of-care therapeutic modalities. The primary genetic drivers are a set of recurrent amino acid substitutions in genes encoding histone H3 (H3K27M), which are currently undruggable. These H3K27M oncohistones perturb normal chromatin architecture, resulting in an aberrant epigenetic landscape. To interrogate for epigenetic dependencies, we performed a CRISPR screen and show that patient-derived H3K27M-glioma neurospheres are dependent on core components of the mammalian BAF (SWI/SNF) chromatin remodeling complex. The BAF complex maintains glioma stem cells in a cycling, oligodendrocyte precursor cell–like state, in which genetic perturbation of the BAF catalytic subunit SMARCA4 (BRG1), as well as pharmacologic suppression, opposes proliferation, promotes progression of differentiation along the astrocytic lineage, and improves overall survival of patient-derived xenograft models. In summary, we demonstrate that therapeutic inhibition of the BAF complex has translational potential for children with H3K27M gliomas.Significance:Epigenetic dysregulation is at the core of H3K27M-glioma tumorigenesis. Here, we identify the BRG1–BAF complex as a critical regulator of enhancer and transcription factor landscapes, which maintain H3K27M glioma in their progenitor state, precluding glial differentiation, and establish pharmacologic targeting of the BAF complex as a novel treatment strategy for pediatric H3K27M glioma.See related commentary by Beytagh and Weiss, p. 2730.See related article by Mo et al., p. 2906.This article is highlighted in the In This Issue feature, p. 2711

Human tumors are composed of diverse malignant and nonmalignant cells, generating a complex ecosystem that governs tumor biology and response to treatments. Recent technological advances have enabled the characterization of tumors at single-cell resolution, providing a compelling strategy to dissect their intricate biology. Here we describe recent developments in single-cell expression profiling and the studies applying them in clinical settings. We highlight some of the powerful insights gleaned from these studies for tumor classification, stem cell programs, tumor microenvironment, metastasis, and response to targeted and immune therapies.Significance:Intratumor heterogeneity (ITH) has been a major barrier to our understanding of cancer. Single-cell genomics is leading a revolution in our ability to systematically dissect ITH. In this review, we focus on single-cell expression profiling and lessons learned in key aspects of human tumor biology.

Glioblastoma is an incurable malignancy that remains poorly understood. By large-scale single cell RNA-seq and follow up experiments we found that malignant cells in glioblastoma exist in four main cellular states that recapitulate distinct steps of neural development, are influenced by the tumor microenvironment, and exhibit plasticity. Importantly, we show that while each glioblastoma contains similar cellular states, their relative frequency varies between tumors and is influenced by chromosomal aberrations as well as by the tumor microenvironment. In particular, we demonstrate that macrophages induce a transition of glioblastoma cells into mesenchymal-like (MES-like) states. This effect is mediated, both in vitro and in vivo, by macrophage-derived ligands that interact with receptors on glioblastoma cells. Finally, we show that MES-like glioblastoma states is associated with altered T cells cytotoxicity and response to immunotherapies. Overall, this work dissects the cellular diversity and interactions within the glioblastoma microenvironment, with potential implications for therapies. Citation Format: Toshi Hara, Rony Chanoch, Mario Suva, Itay Tirosh. Dissecting glioblastoma by single cell RNA-seq [abstract]. In: Proceedings of the AACR Virtual Special Conference on Tumor Heterogeneity: From Single Cells to Clinical Impact; 2020 Sep 17-18. Philadelphia (PA): AACR; Cancer Res 2020;80(21 Suppl):Abstract nr IA24.

Human diffuse gliomas are incurable brain tumors, where cellular state diversity fuels tumor progression and resistance to therapy. Single-cell RNA-sequencing (scRNAseq) studies recently charted the cellular states of the two major categories of human gliomas, IDH-mutant gliomas (IDH-MUT) and IDH-wildtype glioblastoma (GBM), showing that malignant cells partly recapitulate neurodevelopmental trajectories. This raises the central questions of how cell states are encoded epigenetically and whether unidirectional hierarchies or more plastic state transitions govern glioma cellular architectures. To address these questions, we generated multi-omics single-cell profiling, integrating DNA methylation (DNAme), transcriptome and genotyping of 1,728 cells from 11 GBM and IDH-MUT primary patient samples. The assessment of DNAme intra-tumoral heterogeneity of malignant cells revealed that single-cell DNAme profiles within tumors span multiple bulk subtypes, are associated with important biological features of malignant cells, and may be confounded by the tumor micro-environment. Such sources of intra-tumoral heterogeneity in bulk profiles are important to recognize, as DNAme profiling is increasingly being utilized for bulk clinical brain tumor classification. The direct comparison of the methylomes of distinct glioma cellular states revealed Polycomb repressive complex 2 (PRC2) targets DNAme as a key switch in the differentiation of malignant GBM cells. In contrast, dissecting aberrant circuits of hypermethylation and gene expression in IDH-MUT gliomas, we observed a decoupling of the promoter methylation-expression relationship, with disruption of CTCF insulation and enhancer vulnerability which increases with cellular differentiation. To define cell state transition dynamics directly in patient samples, we generated high-resolution lineage histories of glioma cells using heritable DNAme changes, and projected the scRNAseq-derived cell states onto the lineage trees. This analysis demonstrated that cell states are heritable across malignant gliomas and, while in IDH-MUT differentiation far outpaces de-differentiation, GBM harbors a higher degree of cellular state plasticity allowing reversion of GBM cells from a differentiated to a stem-like state. Overall, our work provides detailed insights into gliomagenesis, dissecting the epigenetic encoding, regulatory programs, and dynamics of the cellular states that drive human gliomas. Importantly, it also carries significant translational implication, as the high degree of de-differentiation in GBM challenges the paradigm of therapeutically targeting glioma stem-like cells to deprive tumors of their ability to regenerate. Citation Format: Federico Gaiti, Ronan Chaligne, Dana Silverbush, Joshua S. Schiffman, Hannah R. Weisman, Lloyd Kluegel, Simon Gritsch, Sunil D. Deochand, Alyssa R. Richman, Johanna Klughammer, Tommaso Biancalani, Christoph Muus, Caroline Sheridan, Alicia Alonso, Franco Izzo, Orit Rozenblatt-Rosen, Aviv Regev, Mario L. Suvà, Dan A. Landau. Deciphering differentiation hierarchies, heritability and plasticity in human gliomas via single-cell multi-omics [abstract]. In: Proceedings of the AACR Virtual Special Conference on Tumor Heterogeneity: From Single Cells to Clinical Impact; 2020 Sep 17-18. Philadelphia (PA): AACR; Cancer Res 2020;80(21 Suppl):Abstract nr PO-019.

In situ transgenesis methods such as virus and electroporation can create somatic transgenic mice quickly, but they lack the exquisite control over copy number, zygosity, and locus specificity. We have recently established mosaic analysis by dual recombinase-mediated cassette exchange (MADR), which permits stable labeling of mutant cells expressing transgenic elements from precisely-defined chromosomal loci. MADR provides a toolkit of elements for combinatorial labeling, inducible/reversible transgene manipulation, VCre recombinase expression, and genetic manipulation of human cells. Further, we have demonstrated the versatility of MADR by creating glioma models with mixed, reporter-identified zygosity or with “personalized” driver mutations from pediatric glioma. For example, introducing H3f3a mutation variants with MADR regulates the spatiotemporal profile of glioma, and single-cell RNA and ATAC sequencing analysis demonstrates a recapitulation of developmental hierarchy seen in K27M-mutant human glioma. Moreover, we have generated novel models of supratentorial ependymoma using patient-derived oncofusion transgenes. These models display a high degree of fidelity and we now compare these models on a single-cell level with our previous models and human tumor cell transcriptomes. In addition, we now demonstrate the ability to generalize the MADR technology to other non-CNS tissues using local plasmid delivery. Finally, we have engineered human cells to allow for MADR transgenesis and somatic transgenic organoids. These combined approaches will enable researchers to discovery disease mechanisms and test therapeutics in more physiologically relevant cancer models.MADR is extensible to thousands of existing mouse lines and can be adapted to human cells, providing a flexible platform to democratize the generation of somatic transgenic disease models. Citation Format: Gi Bum Kim, David Rincon Fernandez Pacheco, David Saxon, Amy Yang, Sara Sabat, Marina Dutra-Clarke, Rachelle Levy, Ashley Watkins, Hannah Park, Aslam Abbasi Akhtar, Paul W. Linesch, Naomi Kobritz, Swasty S. Shandra, Katie Grausam, Alberto Ayala-Sarmiento, Jessica Molina, Kristyna Sedivakova, Daniel S. Gareau, Mariella G. Filbin, Serguei Bannykh, Jie Tang, Mario L. Suva, Bin Chen, Moise Danielpour, Joshua J. Breunig. MADR: Rapid generation of somatic mosaics to model cancer in mice and human organoids [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr NG01.

In situ transgenesis methods such as virus and electroporation can create somatic transgenic mice quickly, but they lack the exquisite control over copy number, zygosity, and locus specificity. We have recently established mosaic analysis by dual recombinase-mediated cassette exchange (MADR), which permits stable labeling of mutant cells expressing transgenic elements from precisely defined chromosomal loci. MADR provides a toolkit of elements for combinatorial labeling, inducible/reversible transgene manipulation, VCre recombinase expression, and genetic manipulation of human cells. Further, we have demonstrated the versatility of MADR by creating glioma models with mixed, reporter-identified zygosity or with “personalized” driver mutations from pediatric glioma. For example, introducing H3f3a mutation variants with MADR regulates the spatiotemporal profile of glioma, and single-cell RNA and ATAC sequencing analysis demonstrates a recapitulation of developmental hierarchy seen in K27M mutant human glioma. Moreover, we have generated novel models of supratentorial ependymoma using patient-derived oncofusion transgenes. These models display a high degree of fidelity, and we now compare these models on a single-cell level with our previous models and human tumor cell transcriptomes. In addition, we now demonstrate the ability to generalize the MADR technology to other non-CNS tissues using local plasmid delivery. Finally, we have engineered human cells to allow for MADR transgenesis and somatic transgenic organoids. These combined approaches will enable researchers to discover disease mechanisms and test therapeutics in more physiologically relevant cancer models. MADR is extensible to thousands of existing mouse lines and can be adapted to human cells, providing a flexible platform to democratize the generation of somatic transgenic disease models. This abstract is also being presented as Poster B46. Citation Format: Gi Bum Kim, David Rincon Fernandez Pacheco, David Saxon, Amy Yang, Sara Sabet, Marina Dutra-Clarke, Rachelle Levy, Ashley Watkins, Hannah Park, Aslam Abbasi Akhtar, Paul W. Linesch, Naomi Kobritz, Swasty S. Chandra, Katie Grausam, Alberto Ayala- Sarmiento, Jessica Molina, Kristyna Sedivakova, Daniel S. Gareau, Mariella G. Filbin, Serguei Bannykh, Jie Tang, Mario Suva, Bin Chen, Moise Danielpour, Joshua J. Breunig. MADR: Rapid generation of somatic mosaics with locus-specific, stably integrated transgenic elements for generation of “personalized” mouse models and human organoid tumor models [abstract]. In: Proceedings of the AACR Special Conference on the Evolving Landscape of Cancer Modeling; 2020 Mar 2-5; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2020;80(11 Suppl):Abstract nr PR11.

Single-cell RNA sequencing (scRNAseq) has been performed across a range of intermediate to high-grade gliomas that each harbor multiple driver mutations. Pilocytic astrocytoma (PA), the most common childhood brain tumor, is a low-grade glioma. PAs frequently harbor oncogenic KIAA1549-BRAF fusions, and exhibit low rates of other driver mutations. While PAs exhibit a favorable prognosis compared to the higher-grade gliomas, treatment morbidity and tumor recurrence can represent major challenges for some PA patients. We performed scRNAseq of both tumor and non-tumor cells in newly-diagnosed PAs that contained KIAA1549-BRAF rearrangements. To confidently distinguish tumor cells from nontumor cells, we sorted cells by glial progenitor marker A2B5 status and profiled KIAA1549-BRAF fusion status of cells using a sensitive quantitative PCR-based assay. Results were validated using RNA in situ hybridization. When compared to higher-grade gliomas, a higher proportion of the PA tumor cells exhibited a differentiated, astrocyte-like phenotype. A smaller proportion of cells exhibited a progenitor-like phenotype with evidence of proliferation. These progenitor-like tumor cells expressed a mitogen-activated protein kinase (MAPK) program that was absent from higher-grade gliomas. Similar patterns of expression of genes associated with the astrocyte-like and MAPK gene programs were also seen in formalin fixed, paraffin embedded PA tissues using RNA in situ hybridization. Immune cells, especially microglia, comprised almost half of all cells in the PAs and accounted for differences in bulk expression profiles between tumor locations and subtypes. These single cell transcriptional data reveal a transcriptional developmental hierarchy in a pediatric low grade glioma that is skewed towards more mature brain cells compared to higher-grade gliomas. The results indicate that future analyses of bulk PA tissues should attempt to account for considerable infiltration by nontumor cells. Finally, the finding that a MAPK gene program is not uniformly expressed in PA tumor cells has implications for ongoing clinical investigations of therapies directed at the MAPK pathway for the treatment of PA. Citation Format: Zachary J. Reitman, Brenton Paolella, Guillaume Bergthold, Kristine Pelton, Robert Jones, Sarah Becker, Claire E. Sinai, Haley Malkin, Ying Huang, Leslie Grimmett, Zachary T. Herbert, Yu Sun, Jessica Weatherbee, John Alberta, John Daley, Orit Rozenblatt-Rosen, Rosalind Segal, Daphne Haas-Kogan, Mariella G. Filbin, Mario L. Suva, Aviv Regev, Charles Stiles, Mark W. Kieran, Liliana Goumnerova, Keith L. Ligon, Alex K. Shalek, Pratiti Bandopadhayay, Rameen Beroukhim. Single cell RNA sequencing reveals mitogenic and progenitor gene programs inBRAF-rearranged pilocytic astrocytomas [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr 3647.

Personal vaccines directed at neoantigens arising from tumor mutations can induce neoepitope-specific T cell responses in patients with highly mutated tumors such as melanoma. It remains unknown if this approach can be successfully applied in tumors with low mutation frequency. We conducted a phase 1/1b study to determine the safety and feasibility of patient-specific neoantigen-targeting vaccination in patients with newly diagnosed, methylguanine methyltransferase (MGMT) unmethylated glioblastoma, administered following maximum surgical resection and conventional radiotherapy. Tumor-specific mutations were identified by whole-exome sequencing. Each vaccine, composed of up to 20 synthetic long peptides containing predicted tumor neoepitopes admixed with poly-ICLC, was administered subcutaneously on a prime-boost schedule. Among 8 treated patients, adverse events were limited to mild injection site reactions. Seven patients (88%) received a vaccine with ≥10 neoepitope peptides (median 12, range, 7-20), with median time to vaccine initiation of 18.6 weeks from surgery. All patients died from progressive disease. Median progression-free and overall survival was 7.5 months (95% CI: 6.2, 9.7) and 16.8 months (90% CI: 9.6, 21.3), respectively. 3 patients dropped out of the study due to disease progression. We analyzed vaccine responses on 5 patients that got at least one booster immunization. All 3 patients who required dexamethasone during vaccine priming failed to generate interferon-γ responses in peripheral blood mononuclear cells. In contrast, 2 patients who did not receive dexamethasone during vaccine priming, generated robust de novo immune responses against multiple personal neoantigens. Circulating vaccine-induced polyfunctional neoantigen-specific CD4+ and CD8+ T cell responses were enriched for memory and activated phenotypes, and increased numbers of tumor-infiltrating CD4+ and CD8+ T cells were detected in these 2 patients. T cell receptor analysis identified identical clonotypes isolated from post-vaccination glioblastoma tissue and peripheral blood including a clonotype specific for ARHGAP35, a neoantigen targeted by vaccination. To our knowledge we provide the first evidence that tumor specific T cells can traffic from the periphery into glioblastoma tumors and that neoantigen-targeting vaccines can favorably alter the tumor immune milieu of glioblastoma. In conclusion, individualized, multi-neoepitope vaccination is feasible, safe, and generates neoantigen-specific T cell responses in the periphery and intracranial tumors of patients with glioblastoma. Citation Format: Derin B. Keskin, Itay Tirosh, Annabelle Anandappa, Jing Sun, Nathan D. Mathewson, Sachet A. Shukla, Evisa Gjini, Shuqiang Li, Letitia Li, Anita Giobbie-Hurder, Phuong M. Le, Zhuting Hu, Wandi Zhang, Oriol Olive, Christine McCluskey, Heather Daley, Patrick Y. Wen, Antonio E. Chiocca, Maegan Harden, Niall Lennon, Stacey Gabriel, Gad Getz, Donna Neuberg, Jerome Ritz, Eric S. Lander, Aviv Regev, Kai Wucherpfennig, Mario Suva, Edward F. Fritsch, Rodig Scott, Keith L. Ligon, Kenneth J. Livak, Hacohen Nir, Patrick A. Ott, Catherine J. Wu, David A. Reardon. Personal neoantigen-targeting vaccination generates neoepitope-specific T cell responses in tumors of patients with glioblastoma [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr 5631.

Obtaining paired alpha- and beta-chain information of the T cell receptor (TCR) enables identification of which TCR binds to a particular antigen, providing critical knowledge for understanding how a tumor interacts with its immune microenvironment and for monitoring responses to immunotherapy. We have developed a rapid, cost-effective single-cell TCR sequencing protocol (scTCRseq) that processes single T cells flow sorted into 384-well plates. Exploiting the added specificity provided by RNase H-dependent PCR (rhPCR; Dobosy et al., BMC Biotechnology 11:80, 2011), our scTCRseq protocol mixes primers for all known productive alpha and beta TCR alleles and obtains specific amplification of only the alleles present in a single-cell lysate (typically one alpha allele and one beta allele) without the use of nested PCR. Applying this protocol to T cell samples from normal donors, we detect TCRs arising from all productive TRAV and TRBV genes, and observe the expected highly diverse repertoires. Our method also includes targeted sequencing of 96 (or more) marker transcripts in order to obtain cell phenotype information. Our scTCRseq protocol has been adapted to analyze low-input samples (100 pg-10 ng) of PBMC RNA for characterization of the TCR repertoire and to improve the sensitivity of paired TCR detection in Smart-seq2 libraries prepared from single-cell RNA. We recently applied this approach to analyze T cell clonality in the T cell infiltrate from a glioblastoma patient receiving neoantigen-specific peptide vaccination. Fresh tumor obtained post-vaccination at the time of recurrence was dissociated and single CD3+ T cells were isolated by flow cytometry. Smart-seq2 analysis suggested the presence of clones, but this was based largely on inference from one-chain data. Using our scTCRseq protocol to re-amplify the TCR segments, paired alpha- and beta-chain information was obtained from 277 of 363 (76%) single cells that passed quality control. Out of 231 distinct clonotypes detected, 25 were identified as clones, where a clone is defined as two or more cells with identical alpha- and beta-CDR3 segments. GLIPH analysis (Glanville et al., Nature 547:94-98, 2017) was used to cluster infiltrate TCRs by specificity. A representative TCR from each specificity cluster was cloned in a TCR-negative Jurkat cell line. Cellular assays were used to determine which specificity groups react with epitopes from peptides used in the vaccination. TCRs specific for neoantigens included in the vaccine were detected. We have demonstrated a robust workflow that combines sensitive detection of paired alpha- and beta-chain TCR sequences with TCR cloning and expression to determine which TCRs in a tumor infiltrate have neoantigen specificity. Citation Format: Shuqiang Li, Jing Sun, Annabelle Anandappa, Oliver Spiro, Ignaty Leshchiner, Krishnalekha Datta, Yun Bao, Zhuting Hu, Nir Hacohen, Nathan D. Mathewson, Itay Tirosh, Kai Wucherpfennig, Aviv Regev, Mario Suva, Derin B. Keskin, Catherine J. Wu, Kenneth J. Livak. Novel targeted single-cell TCR sequencing method used for analysis of a T cell infiltrate from a glioblastoma patient [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr 5387.

Ewing's sarcoma is a member of Ewing's family tumors (EFTs) and the second most common solid bone and soft tissue malignancy of children and young adults. It is associated in 85% of cases with the t(11;22)(q24:q12) chromosomal translocation that generates fusion of the 5′ segment of the EWS gene with the 3′ segment of the ETS family gene FLI-1. The EWS-FLI-1 fusion protein behaves as an aberrant transcriptional activator and is believed to contribute to EFT development. However, EWS-FLI-1 induces growth arrest and apoptosis in normal fibroblasts, and primary cells that are permissive for its putative oncogenic properties have not been discovered, hampering basic understanding of EFT biology. Here, we show that EWS-FLI-1 alone can transform primary bone marrow–derived mesenchymal progenitor cells and generate tumors that display hallmarks of Ewing's sarcoma, including a small round cell phenotype, expression of EFT-associated markers, insulin like growth factor-I dependence, and induction or repression of numerous EWS-FLI-1 target genes. These observations provide the first identification of candidate primary cells from which EFTs originate and suggest that EWS-FLI-1 expression may constitute the initiating event in EFT pathogenesis. (Cancer Res 2005; 65(24): 11459-68)