Molecular Subtypes and CD4+ Memory T Cell-Based Signature Associated With Clinical Outcomes in Gastric Cancer

• Main Authors: Zhi-Kun Ning, Ce-Gui Hu, Chao Huang, Jiang Liu, Tai-Cheng Zhou, Zhen Zong
• Format: Article
• Language: English
• Published: Frontiers Media S.A. 2021-03-01
• Series: Frontiers in Oncology
• Subjects: gastric cancer; prognostic signature; CD4+ memory T cell; tumor microenvironment; weighted gene co-expression network analysis
• Online Access: https://www.frontiersin.org/articles/10.3389/fonc.2020.626912/full

BackgroundCD4+ memory T cells are an important component of the tumor microenvironment (TME) and affect tumor occurrence and progression. Nevertheless, there has been no systematic analysis of the effect of CD4+ memory T cells in gastric cancer (GC).MethodsThree datasets obtained from microarray and the corresponding clinical data of GC patients were retrieved and downloaded from the Gene Expression Omnibus (GEO) database. We uploaded the normalize gene expression data with standard annotation to the CIBERSORT web portal for evaluating the proportion of immune cells in the GC samples. The WGCNA was performed to identify the modules the CD4+ memory T cell related module (CD4+ MTRM) which was most significantly associated with CD4+ memory T cell. Univariate Cox analysis was used to screen prognostic CD4+ memory T cell-related genes (CD4+ MTRGs) in CD4+ MTRM. LASSO analysis and multivariate Cox analysis were then performed to construct a prognostic gene signature whose effect was evaluated by Kaplan-Meier curves and receiver operating characteristic (ROC), Harrell’s concordance index (C-index), and decision curve analyses (DCA). A prognostic nomogram was finally established based on the CD4+ MTRGs.ResultWe observed that a high abundance of CD4+ memory T cells was associated with better survival in GC patients. CD4+ MTRM was used to stratify GC patients into three clusters by unsupervised clustering analysis and ten CD4+ MTRGs were identified. Overall survival, five immune checkpoint genes and 17 types of immunocytes were observed to be significantly different among the three clusters. A ten-CD4+ MTRG signature was constructed to predict GC patient prognosis. The ten-CD4+ MTRG signature could divide GC patients into high- and low-risk groups with distinct OS rates. Multivariate Cox analysis suggested that the ten-CD4+ MTRG signature was an independent risk factor in GC. A nomogram incorporating this signature and clinical variables was established, and the C-index was 0.73 (95% CI: 0.697–0.763). Calibration curves and DCA presented high credibility for the OS nomogram.ConclusionWe identified three molecule subtypes, ten CD4+ MTRGs, and generated a prognostic nomogram that reliably predicts OS in GC. These findings have implications for precise prognosis prediction and individualized targeted therapy.