Comprehensive analysis of lncRNAs reveals candidate prognostic biomarkers in multiple cancer types

Abstract

Long non-coding RNAs (lncRNAs) are increasingly recognized as functional units in cancer pathways and powerful molecular biomarkers, however most lncRNAs remain uncharacterized. Here we performed a systematic discovery of prognostic lncRNAs in 9,326 patient tumors of 29 types using a proportional-hazards elastic net machine-learning framework. lncRNAs showed highly tissue-specific transcript abundance patterns. We identified 179 prognostic lncRNAs whose abundance correlated with patient risk and improved the performance of common clinical variables and molecular tumor subtypes. Pathway analysis revealed a large diversity of the high-risk tumors stratified by lncRNAs and suggested their functional associations. In lower-grade gliomas, discrete activation of HOXA10-AS indicated poor patient prognosis, neurodevelopmental pathway activation and a transcriptomic similarity to glioblastomas. HOXA10-AS knockdown in patient-derived glioblastoma cells caused decreased cell proliferation and deregulation of glioma driver genes and proliferation pathways. Our study underlines the pan-cancer potential of the non-coding transcriptome for developing molecular biomarkers and innovative therapeutic strategies.