LncRNA profile study reveals a seven-lncRNA signature predicts the prognosis of patients with colorectal cancer

BACKGROUND: The prognosis of colorectal cancer (CRC) is still challenging to evaluate or predict. Recently, long non-coding RNAs (lncRNAs) have been found to play an important role in tumorigenesis and prognosis, however, few lncRNAs have been identified in CRC progression. We aimed to establish a lncRNA signature to improve prognosis prediction of CRC. METHODS: In the present study, we profiled lncRNA expression with a lncRNA-mining approach in two CRC data sets from Gene Expression Ominus (GEO) (GSE39582, N = 557 and GSE17538, N = 200). LncRNAs were analyzed to determine a prognostic signature by Cox regression and Robust likelihood-based survival model. We identified seven lncRNAs that significantly associated with the disease free survival (DFS) in the training group. A risk score formula was constructed to evaluate the performance of this lncRNA panel. RESULTS: A seven-lncRNA signature was established to predict prognosis of CRC patients. The prognostic value of this signature was verified in the training group, internal validation group and external validation cohort, respectively. Receiver operating characteristic (ROC) analysis suggested a powerful discrimination ability of the seven-gene signature. Finally, Cox regression analyzed this signature as an independent influencing factor and subsequent pathway or network analysis implicated a potential mechanism of these lncRNAs. CONCLUSIONS: In summary, the seven-lncRNA signature we identified can effectively classify patients. This risk score model could serve as an independent biomarker to predict prognosis of CRC patients.