(P 381) THE GENOMIC LANDSCAPE OF CLEAR CELL SARCOMA

Objective:

Background: Clear cell sarcoma of soft tissue (CCS) is a rare soft tissue tumour of tendons and aponeuroses mostly affecting the lower extremities. CCS tumours are quite aggressive with local recurrence rates approaching 40% and metastasis rates of 70%. Since it is a rare cancer, there is limited understanding of the factors leading to disease progression and poor overall survival in CCS. Previous studies have established EWSR1 gene fusions as the genetic hallmarks of CCS. However, the processes which initiate these fusion events are not fully explored in CCS unlike tumours like Ewing sarcoma where chromoplectic loops are known to generate EWS-ERG fusion. All the CCS cases are driven by the same fusion event yet show variation in their survival outcome. The mechanism of cancer evolution in CCS genome is little-known and hence, it becomes imperative to investigate the disease progression in CCS. In this study, we aim to investigate the longitudinal whole genome data of CCS in combination with bulk transcriptomics and methylation data to gain a comprehensive understanding of the CCS genome.

Hypothesis: We hypothesize that selection of well-characterized cases along with comprehensive molecular analysis will identify novel correlations between molecular markers and clinical/pathologic features of CCS, potentially leading to identification of therapeutic targets.

Methods: Careful pathological evaluation and clinical annotation was undertaken for 45 CCS samples (28 primaries, 4 local recurrences and 13 metastases) where fresh frozen tumour with matching germline DNA were available. These were sequenced at high resolution and depth using whole genome sequencing (>70X tumour and >30X germline). Similarly, frozen tissue of 37 cases (23 primaries, 2 local recurrences and 12 metastases) underwent bulk RNA sequencing. The resulting fastq files were QC’d ,aligned to the reference hg38 using the fully automated pipeline of bcbio-nextgen toolkit. Bam files from the WGS data were further processed to get the somatic single nucleotide variants, indels, copy number variants and structural variants which underwent consensus mutation calling and manual curation. For RNA-seq data, differential gene expression analysis was done using DESeq2 package on R Studio. For methylation analysis, frozen DNA of 46 CCS samples (30 primaries, 2 local recurrences and 14 metastases) were run on an EPICv2.0 array. The resulting idat files were processed for QC and differential methylation analysis using the sesame package.

Results: Analysis revealed the prevalence of diploidy and low tumour mutation burden across all CCS cases ranging from 0.4 to 3.3 mutations per Mb. There was absence of any specific recurrent non-synonymous mutation, however, the spectrum of somatic non-synonymous mutations identified converged on common signaling pathways like MAPK and PI3K-Akt. Frequent copy number alterations were observed in chromosomes 8, 7, 9 and 17 and unexpectedly identified significant subclonality. Significantly altered recurrent focal copy number variations were identified at loci- 1p22.1, 1q23.2, 8q24.23, 9p21.3 and 17q25.1. These positions harbor critical genes like MYC, CDKN2A and CDKN2B. >90% samples harbored the EWSR1::ATF fusion followed by EWSR1::CREM and EWSR1::CREB fusions. Unexpectedly, complex structural events of chromothripsis and chromoplexy were identified in 8 CCS patients. Integration of the genomic, transcriptomic and methylation data revealed hitherto unknown patterns of genetic and transcriptional intra and inter tumour heterogeneity.

Conclusion: Despite all CCS showing recurrent driver fusion genes and a paucity of recurrent point mutations, the spectrum of genetic heterogeneity underpinned by subclonal copy number changes, complex structural events such as chromothripsis and chromoplexy is unexpected. Differential gene expression analysis between primary and metastatic tumours showed insights into potential mechanisms associated with metastasis. Evaluation of temporal (primary, recurrence and metastatic) whole genome data will further reveal the evolution patterns in CCS that could provide insights into how to advance treatment strategies, and thereby improving survival in CCS patients.

This research was supported by the generous philanthropic gift to UCL by the Edward Showler Foundation. https://www.edwardshowlerfoundation.com/