Premium
Esophageal cancer lymph node metastasis–associated gene signature optimizes overall survival prediction of esophageal cancer
Author(s) -
Cai Weiyang,
Li Yanyan,
Huang Bo,
Hu Changyuan
Publication year - 2019
Publication title -
journal of cellular biochemistry
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.028
H-Index - 165
eISSN - 1097-4644
pISSN - 0730-2312
DOI - 10.1002/jcb.27416
Subject(s) - esophageal cancer , kegg , proportional hazards model , oncology , metastasis , cancer , survival analysis , univariate , lymph node , stage (stratigraphy) , medicine , gene signature , framingham risk score , gene , biology , computational biology , multivariate statistics , disease , gene expression , transcriptome , computer science , genetics , paleontology , machine learning
Esophageal cancer (EC) is characteristic of early regional lymph node metastasis (LNM) and most patients with metastasis have a poor prognosis. However, the current diagnostic techniques do not enable precise differentiation of EC LNM, prognostic stratification, and individual survival estimation. To identify potential molecular biomarkers for EC patients with LNM, we explored differently expressed genes in The Cancer Genome Atlas database between 77 non‐LNM cases and 88 LNM cases by limma package R. Then, according to univariate and multivariate Cox regression analyses, we constructed an 8‐messenger RNA (mRNA) prognostic signature model, which could predict the outcome in a more exact way. The area under the curve of the risk score is significantly higher than other clinical information, indicating that the 8‐mRNA–based risk score is a good indicator for prognosis. Then, combined with other individual risk factors, such as age, sex, T stage, M stage, etc, we could precisely calculate the individual 1‐, 3‐, and 5‐year survival rates. The Gene Set Enrichment Analysis, Gene Ontology, and Kyoto Encyclopedia of Genes and Genomes analysis indicate that the risk model is mainly associated with cancer‐related pathways, such as cell division, cellular meiosis, and cell cycle regulation. In summary, the 8‐mRNA–based risk score model that we developed successfully predicts the survival of EC. It is independent of clinical information and performing better than other clinical information for prognosis.