A prognostic 3-long noncoding RNA signature for patients with gastric cancer

Abstract

Current studies showed that long noncoding RNAs (lncRNAs) may act as prognostic biomarkers in a variety of cancers. The aim of this study was to identify and assess a prognostic lncRNA signature in patients with gastric carcinoma (GC). LncRNAs expression profiles and corresponding clinicopathological data for 350 patients with GC were obtained from The Cancer Genome Atlas (TCGA), and the least absolute shrinkage and selection operator Cox (LASSO Cox) regression model was used to identify the lncRNA signature. Finally, 3 lncRNAs (CYP4A22-AS1, AP000695.6, and RP11-108M12.3) were identified using LASSO Cox. The prognostic score was imputed as follows: (0.354 × expression level of AP000695.6) + (- 0.899 × expression level of CYP4A22-AS1) + (0.881 × expression level of RP11-108M12.3). The nomogram that integrated independent prognostic factors (age, American Joint Committee on Cancer-lymph node status, and residual tumor and risk score) was constructed. In TCGA cohort, the area under the curve (AUC) for the predictive nomogram was 0.737 (19-month survival). The calibration curve also demonstrated satisfactory agreement between predictive values and observation values in the probabilities of 1-, 3-, and 5-year overall survival. We also established prognostic model using machine learning technique, and the corresponding AUC for the predictive model was 0.756 (19-month survival). The Kyoto Encyclopedia of Genes and Genomes pathway analysis showed that the Hippo signaling pathway was the main pathway associated with the 3-lncRNA signature.

Keywords: gastric carcinoma; long noncoding RNA; nomogram; prognosis.