1716P - The molecular landscape of fusion genes in endometrial stromal sarcomas include three nosological entities with different natural history

Abstract

BackgroundAbout half of ESS do not harbor a JAZF1 or YWHAE fusion and other rearrangements have already been reported. With the aim to provide a comprehensive description of the molecular landscape of endometrial stromal sarcomas (ESS), we investigated by RNA-sequencing a series of tumors diagnosed as ESS based on the morphology but negative for JAZF1 and/or YWHAE in FISH experiments. MethodsThis study was performed between September 2017 and December 2018 as a Centre Léon Bérard monocentric retrospective translational research program on tumor samples from an investigational cohort of 43 ESS patients. For clustering analyses, we used a control cohort composed of 19 patient samples of BCOR-rearranged sarcomas (n=8), low-grade ESS with JAZF1-SUZ12 fusion (n=5), high-grade ESS carrying a BCOR internal tandem duplication (n=1), uterine leiomyoma (n=2) and uterine leiomyosarcomas (n=3). RNA sequencing was performed from FFPE material in all cases using TruSeq RNA Access Library Prep Kit (Illumina®). ResultsA chromosomal rearrangement was identified in 74% of the cases (n=32). We identified biologically homogeneous groups of tumors such as the JAZF1-fused (to PHF1 and SYNGAP1) and the BCOR-rearranged (BCOR-ITD, ZC3H7B-BCOR, CREBBP-BCOR). Using a clustering approach on transcriptomic data, tumors were divided in 3 subgroups: one mainly composed of BCOR-rearranged ESS (n=7), one mainly composed of JAZF1-fused ESS (n=14) and one composed of various molecular subtypes (n=22). This 3 subgroups display significantly different survivals (Log-rank test, p=0.004). Five cases which harbored a YWAHE-NUTM2B fusions clustered separately: 3 cases in the group of indolent tumors (JAZF1-fused ESS) and 2 cases in the group of aggressive tumors (BCOR-rearranged sarcomas). Of note, those 2 cases harbored also a single nucleotide variant, one on BCOR and one on BCORL1. Results of RNA-seq and clustering analysis have allowed a better classification in 6 cases. ConclusionsThis is the largest series of RNAseq performed in a very rare subtype of tumors: ESS. We report new and recurrent molecular variants of ESS including previously unreported fusions. RNA-sequencing may support the diagnosis of ESS and help a better classification. Legal entity responsible for the studyThe authors. FundingInterSARC grants. DisclosureAll authors have declared no conflicts of interest.