AUTHOR=Li Jianxin , Han Ting , Wang Xin , Wang Yinchun , Chen Xuan , Chen Wangsheng , Yang Qingqiang 

TITLE=Construction of a Novel Immune-Related mRNA Signature to Predict the Prognosis and Immune Characteristics of Human Colorectal Cancer

JOURNAL=Frontiers in Genetics

VOLUME=Volume 13 - 2022

YEAR=2022

URL=https://www.frontiersin.org/journals/genetics/articles/10.3389/fgene.2022.851373

DOI=10.3389/fgene.2022.851373

ISSN=1664-8021

ABSTRACT=Background: Anti-cancer immunotherapeutic approaches have gained significant efficacy in multiple cancer types. However, not all patients with colorectal cancer (CRC) could benefit from immunotherapy due to tumor heterogeneity. The purpose of this study is to construct an immune-related signature for predicting the immune characteristics and prognosis of CRC.
Methods: RNA-sequencing data and corresponding clinical information of patients with CRC were obtained from The Cancer Genome Atlas (TCGA) and Gene Expression Omnibus (GEO), and immune-related genes (IRGs) were downloaded from the Immunology Database and Analysis Portal (ImmPort). Then, we utilized univariate, lasso regression, and multivariate cox regression to identify prognostic IRG and develop the immune-related signature. Subsequently, a nomogram was established based on the signature and other prognostic factors, and its predictive capacity was assessed by receiver operating characteristic (ROC) and decision curve analysis (DCA). Finally, associations between the signature and the immune characteristics of CRC were assessed.
Results: In total, 472 samples downloaded from TCGA were divided into training cohort (236 samples) and internal validation cohort (236 samples), and GEO cohort was downloaded as external validation cohort (122 samples). A total of 476 differently expressed IRG were identified and 17 of which were significantly correlated prognosis of CRC patients. Finally, 10 IRGs were filtered out to construct risk score signature and patients were divided into low- and high-risk groups according to the median of risk scores in training cohort. High risk score was significantly correlated with unfavorable survival outcome and aggressive clinic-pathological characteristics in CRC patients, and the results were further confirmed in internal validation cohort, entire TCGA cohort, and external validation cohort. Immune infiltration analysis revealed that patients in the low-risk group infiltrated with high tumor-infiltrating immune cells (TIIC) abundances compared to high-risk group. Moreover, we also found that the immune checkpoint biomarkers were significantly overexpressed in low-risk group.
Conclusions: The prognostic signature established by IRGs showed a promising clinical value for predicting the prognosis and immune characteristics of human CRC, which contribute to individualized treatment decisions.