Abstract 3303: A comprehensive TCGA Pan-Cancer molecular study of gynecologic and breast cancers

The Pan-Gynecologic (Pan-Gyn) molecular study analyzed data from 2,579 tumors representing five TCGA tumor-type projects: high-grade serous ovarian cystadenocarcinoma (OV), invasive breast carcinoma (BRCA), uterine corpus endometrial carcinoma (UCEC), cervical squamous cell carcinoma and endocervical adenocarcinoma (CESC), and uterine carcinosarcoma (UCS). Principal aims were to identify shared and unique molecular features, clinically meaningful subtypes, and potential future therapeutic targets. We analyzed data of multiple types, including somatic copy number alterations (SCNAs), mutations, DNA methylation, mRNA expression (from mRNA-seq), microRNA (miRNA) expression (from miRNA-seq), long non-coding RNA (lncRNA) expression (from miRNA-seq), and protein expression (from reverse-phase protein arrays; RPPA). We found multiple genomic and epigenomic features that were shared across the Pan-Gyn tumor types and distinguished them from other cancer types. Sixty-one somatic copy number alterations (SCNAs) and 46 significantly mutated genes (SMGs) were found in the Pan-Gyn cohort; 11 of the SCNAs and 11 of the SMGs had not been reported in previous TCGA single-disease papers on the same tumor types. Ten predominant mutation signatures were found that correlated with COSMIC signatures, whereas 10% of the samples had no SMGs. Most of the Pan-Gyn tumor types shared similar miRNA profiles. The exception, ovarian cancer, was extremely different from the rest, and, unexpectedly, the miRNA profiles of basal-like breast cancers closely resembled those of cervical cancers. Protein expression analysis (RPPA) revealed some OV and UCEC samples with the “reactive” signature previously identified in BRCA and shown to be prognostically relevant. It also revealed a subtype with low ER and AR expression levels (important markers for hormone therapy) that spanned all 5 tumor types. Novel lncRNA analysis showed several functionally significant ER-regulated lncRNAs and gene/lncRNA interaction networks. Pathway analysis identified subsets of tumor samples that had high levels of leukocyte infiltration, with potential implications for immunotherapy. Using 16 key molecular features across the Pan-Gyn cancers, we identified five novel cross-tumor and prognostic subtypes with possible clinical implications. Finally, we developed a decision tree that can classify patients into the five subtypes based on just 6 of the 16 molecular features. Those 6 features are potentially assessable in clinical laboratories. Citation Format: Ashton Berger, Anil Korkut, Rupa S. Kanchi, TCGA Pan-Gynecologic Group and Research Network, Gordon B. Mills, Douglas A. Levine, Rehan Akbani. A comprehensive TCGA Pan-Cancer molecular study of gynecologic and breast cancers [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 3303.