Abstract

Abstract The hallmark gene expression signatures encapsulate oncogenic and tumor suppressor phenotypes; however, the overview of these phenotypes and their underlying genetic changes, particularly those shared across tumor types, remain unknown. Although a large-scale pan-cancer analysis tends to produce many false discoveries, we can detect reliable relationships by identifying those consistently observed across tumor types with robust statistical methods. Thus, we scored 50 hallmark signatures using single-sample GSEA in approximately 9,000 primary tumors from 33 tumor types in TCGA. To identify the genetic mechanisms underlying tumor phenotypes, we then tested signature scores with and without DNA alterations, including 96 recurrently mutated genes, 1,002 significant focal copy number alterations (CNAs), and 78 arm-level CNAs, confirming hits across multiple tumor types through both a random-effects linear model and a robust permutation framework. First, we confirmed our method by identifying well-described relationships in the mutation analysis, including CTNNB1 mutation with WNT signaling and RB1 mutation with E2F pathway. Then, we identified many significant associations which have not been reported in a large cohort of clinical samples. A considerable proportion of these associations consisted of TP53 mutation with various signatures, including MTORC signaling and glycolysis.Arm-level CNAs also demonstrated previously reported associations, such as 8q gain and MYC activation, as well as a number of novel relationships. The most significant association was 5q deletion and glycolysis signature. Interestingly, although TP53 mutation correlated with a variety of aneuploidies, 5q deletion was the most consistently enriched arm-level alterations in TP53-mutated cases above 17p deletion containing TP53. In addition, the association of TP53 mutation and 5q deletion with glycolysis was shared across five tumor types: breast cancer (BRCA) and various squamous carcinomas. Among BRCA subtypes, these features were most prominent in the basal-like subtype. Clustering of the hallmark signatures by tumor subtypes remarkably showed hormone-negative basal-like BRCA grouping with squamous tumors including esophageal, HPV- head and neck, and lung cancers, far from the hormone-positive BRCA subgroups. This cluster was defined by elevated proliferation, inflammatory signaling, and glycolysis. Our findings provide new evidence of DNA alterations beyond tissue of origin which function to execute parallel oncogenic changes. These results identify common mechanisms in basal-like BRCA with squamous carcinomas across mutations, selected CNAs, and phenotypes. Further investigation is warranted to identify actionable therapeutic target in this aggressive population of tumors. Citation Format: Marni B. McClure, Yasunori Kogure, Yuki Saito, Mariko Tabata, Katherine A. Hoadley, Charles M. Perou, Keisuke Kataoka. TP53 mutation and 5q deletion define common phenotypic changes regardless of tissue of origin [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 5904.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call