Abstract Background Patient sex is an important factor in determining cancer incidence and prognosis; however, little is known of the catalog of sex-linked molecular alterations in cancer and how these may impact treatment response. To begin to address these questions, we developed a computational framework to identify sex-linked molecular alterations and to predict therapies to target oncogenic expression patterns identified in male or female tumors. Methods We utilized DNA copy number, DNA methylation, and RNA-Seq data from both normal and tumor tissues across a total of 17 cancer types, profiled as part of The Cancer Genome Atlas (TCGA). For each cancer type and each molecular data type, we performed supervised statistical analyses to identify molecular alterations significantly associated with patient sex. We then used the tumor and normal samples to separately identify RNA expression signatures associated with carcinogenesis in tumors from males and females. Based on these results, we performed sex-stratified pathway analyses and sex-stratified Connectivity Map (C-Map) analyses using lincscloud to identify biological pathways differentially activated in each sex and to prioritize drugs to target the oncogenic signatures in each sex. Results Integrating results from across the 17 cancer types in a meta-analytic framework, we identified 6,469 genes with significantly sex-disparate expression, 3,217 genes with significantly sex-disparate copy number alterations, and 275 genes with significantly sex-disparate methylation (All FDR < 0.05). Of these genes, 13 exhibited significantly (FDR < 0.05) sex-differentiated RNA expression, DNA copy number, and DNA methylation, including: CAV2, a potential tumor suppressor gene; AIFM2, a p53-dependent apoptosis-inducing gene; and DDX43, which induces resistance to MEK inhibitors. Next, utilizing both tumor and normal samples stratified by sex, we identified sex-linked oncogenic pathways, including the Liver X Receptor (LXR) and Retinoid X Receptor (RXR) activation pathways, both potential therapeutic targets, which were significantly associated with male but not female neoplastic signatures in hepatocellular carcinoma. Lastly, we constructed sex-specific sensitivity and resistance signatures from RNA-Seq data and used the lincscloud CMAP resource to calculate sex-specific connectivity scores for thousands of drugs and small molecule compounds, leading to the identification of known sex-linked drug sensitivity associations (e.g. Tamoxifen in lung cancer), as well as the prediction of several novel sex-linked drug sensitivity associations. Conclusion In summary, we developed an integrative computational approach for large-scale discovery of sex-differentiated molecular alterations and for rational prediction of therapeutic strategies to target sex-linked oncogenic signatures. Citation Format: Jonathan Ma, Sadhika Malladi, Andrew H. Beck. Systematic identification of sex-linked molecular alterations and therapeutic strategies in cancer. [abstract]. In: Proceedings of the 106th Annual Meeting of the American Association for Cancer Research; 2015 Apr 18-22; Philadelphia, PA. Philadelphia (PA): AACR; Cancer Res 2015;75(15 Suppl):Abstract nr 2985. doi:10.1158/1538-7445.AM2015-2985
Read full abstract