Identification of long non-coding RNA expression patterns useful for molecular-based classification of type I endometrial cancers

Endometrial cancer represents the most frequent gynecologic malignant disease. Although several genetic alterations have been associated with increased risk, to date diagnosis and prognosis still rely on morphological features of the tumor, such as histological type, grading and invasiveness. As molecular-based classification is desirable for optimal treatment and prognosis of these cancers, , we explored the potential of lncRNAs as molecular biomarkers. To this end, we first identified by RNA sequencing (RNA-Seq) a set of lncRNAs differentially expressed in cancer vs normal endometrial tissues, a result confirmed also by analysis of RNA-Seq data of normal and cancerous endometrium from The Cancer Genome Atlas (TCGA). A significant association of a subset of these differentially expressed lncRNAs with tumor grade was then determined in 405 TCGA endometrial cancer profiles. Integrating endometrial cancer-specific expression profiles of long and small non-coding RNAs a functional association network was then identified. These results described for the first time a functional 'core' network, comprising small and long RNAs, whose deregulation is associated with endometrial neoplastic transformation, representing a set of cancer biomarkers that can be monitored and targeted for diagnosis, follow-up and therapy of these tumors.