At least five major studies have reported molecular subtypes of high-grade, serous ovarian cancer, and these have the potential to guide development of targeted treatments. However, each study has proposed overlapping but different subtype algorithms, and have used small and differing subsets of publicly available data for validation. At least two other major ovarian cancer studies could not identify discrete subtypes in their gene expression data at all. Uncertainty in the clinical relevance of proposed subtypes is compounded by recent reports that most ovarian cancer tumors are multi-clonal, raising the possibility that multiple subtypes exist within a single tumor. Thus the nature of ovarian cancer subtypes remains controversial, even as they have spurred further investment in retrospective analysis of clinical trial specimens. This study aims to resolve controversy and uncertainty in the nature of ovarian cancer subtypes by two main approaches. First, subtype robustness and association to patient outcome are assessed by comparative meta-analysis using all relevant publicly available data. This will provide clarity on which definition of subtyps future research should focus on. Second, the likely ordering of subtype differentiation in tumor evolution is determined by analysis of the allele frequencies of subtype- associated DNA short variants in data from The Cancer Genome Atlas. This application moves the field of ovarian cancer research forward by resolving controversy around proposed subtypes using publicly available data, and by providing standardized definitions of subtype algorithms with documentation for their application to new patients.
This project employs meta-analysis to resolve uncertainty around the robustness and relationship to patient outcome of proposed molecular subtypes of ovarian cancer. It establishes whether genetic heterogeneity within tumors is likely to be an obstacle to subtype-specific therapies. It provides open- source implementations of all proposed subtypes to facilitate their usage by other translational researchers. These objectives will facilitate adoption of the most clinically relevant molecular subtypes to inform precision treatment of ovarian cancer.
| Ma, Siyuan; Ogino, Shuji; Parsana, Princy et al. (2018) Continuity of transcriptomes among colorectal cancer subtypes based on meta-analysis. Genome Biol 19:142 |
| Chen, Gregory M; Kannan, Lavanya; Geistlinger, Ludwig et al. (2018) Consensus on Molecular Subtypes of High-Grade Serous Ovarian Carcinoma. Clin Cancer Res 24:5037-5047 |
| Spratt, Daniel E; Chan, Tiffany; Waldron, Levi et al. (2016) Racial/Ethnic Disparities in Genomic Sequencing. JAMA Oncol 2:1070-4 |
| Waldron, Levi; Riester, Markus; Ramos, Marcel et al. (2016) The Doppelgänger Effect: Hidden Duplicates in Databases of Transcriptome Profiles. J Natl Cancer Inst 108: |
