586P BRCA1, BRCA2 and RAD51C somatic RNAseq study in ovarian cancer: A description of physiological and pathogenic splicing patterns

Abstract

Genomic molecular profiling is becoming a key step in the management of advanced ovarian cancer. Currently, the main indication for tumor analysis is on DNA (BRCA1/2 mutations and HRD status); there is no indication for RNA tumor analysis. The aims of this work were to describe physiological splicing profiles of BRCA1, BRCA2 and RAD51C in ovarian tissue and then search for pathogenic splicing of these genes in patients with high-grade serous papillary adenocarcinomas stage III of FIGO classification with good response to neoadjuvant cisplatin. RNAseq (Agilent SureSelect XT HS/capture-based target enrichment and NextSeq 550 Illumina Sequencer) was performed on RNA extracted from ovarian cell lines OVCAR4 and COV318 and five pre-cisplatin frozen ovarian tumors with wild-type BRCA1 and BRCA2 somatic DNA screening. We obtained the Genomic Instability Score with Myriad approach and performed RNAseq on post-cisplatin FFPE ovarian block. We also checked BRCA1 methylation status by digital PCR before their inclusion to describe physiological splicing patterns, in addition to OVCAR4 and COV318. A pathogenic skip of BRCA1 exon 3 was found in one tumor at a high level, comparing to the normal transcript. The loss of exon 3 caused a truncation of BRCA1 protein at codon 27; no variant was identified on DNA. This splicing event explained the high GIS at 66 and the good clinical response to cisplatin. Exon 3 skipping was tumoral as it was absent on the post-cisplatin residual ovarian block. Another tumor was BRCA1 methylated. There were neither enrichment of alternative nor novel splicing event of BRCA1, BRCA2 and RAD51C, specific of ovarian tissue. Physiological splicing profiles were qualitatively and quantitatively close to those reported in lymphoblastoid cell lines. Unlike BRCA2 and RAD51C, five known reported transcripts of BRCA1 were missing in the ovarian tissue. We provided a complete description of physiological splicing profiles of BRCA1, BRCA2 and RAD51C in the ovarian tissue. We also showed that somatic RNAseq analysis in ovarian cancer may modify the therapeutic prognosis by allowing access to targeted therapies.