Peng Zhang, Liu Xin-Yuan, Cheng Zeng, Kai Wu, Song Zhao
Department of Thoracic Surgery, The First Affiliated Hospital of Zhengzhou University, Zhengzhou, China.
School of Basic Medical Sciences, Henan University, Kaifeng, China.
Ann Transl Med. 2021 Oct;9(20):1576. doi: 10.21037/atm-21-4756.
The incidence of esophageal cancer (ESCA) is increasing rapidly, and the 5-year survival rate is less than 20%. This study provides new ideas for clinical treatment by establishing a prognostic signature composed of immune-related genes (IRGs), and fully analyzing its relationship with target genes and the tumor microenvironment (TME).
We downloaded the ESCA expression matrix and clinical information from The Cancer Genome Atlas (TCGA) database. Differential expression genes (DEGs) were identified with the edgeR package and crossed with the IRGs we obtained from the ImmPort database to obtain differential IRGs (DEIRGs). The prognostic signature was then obtained through univariate Cox, LASSO-Cox, and multivariate Cox analyses. The receiver operating characteristic (ROC) curve was used to evaluate the prediction effect of the model. The immune cell infiltration abundance obtained by ssGSEA and therapeutic target genes was used to perform sufficient correlation analysis with the obtained prognostic signature and related genes.
A total of 173 samples were obtained from TCGA database, including 162 tumor and 11 normal samples. The 3,033 differential genes were used to obtain 254 DEIRGs by intersections with 2,483 IRGs (IRGs) obtained from the ImmPort Database. Finally, multivariate Cox regression analysis identified eight prognostic DEIRGs and established a new prognostic signature (HR: 2.49, 95% CI: 1.68-3.67; P<0.001). Based on the expression of the eight genes, the cohort was then divided into high and low risk groups and Kaplan-Meier (K-M) curves were plotted with the log-rank test P<0.0001 and 1-, 3-year area under the curve (AUC) >0.7. The K-M curves grouped according to high and low risks performed well in the two subgroup validation cohorts, with log-rank test P<0.05. There were differences in the degree of infiltration of 16 kinds of immune cells in tumor and normal samples, and the infiltration abundance of 12 kinds of immune cells was different in the high and low-risk groups.
An effective and validated prognostic signature composed of IRGs was established and had a strong correlation with immune cells and target genes of drug therapy.
食管癌（ESCA）的发病率正在迅速上升，其5年生存率低于20%。本研究通过建立由免疫相关基因（IRGs）组成的预后特征，并全面分析其与靶基因及肿瘤微环境（TME）的关系，为临床治疗提供新思路。
我们从癌症基因组图谱（TCGA）数据库下载了ESCA表达矩阵和临床信息。使用edgeR软件包鉴定差异表达基因（DEGs），并将其与我们从ImmPort数据库获得的IRGs进行交叉分析，以获得差异免疫相关基因（DEIRGs）。然后通过单因素Cox分析、LASSO-Cox分析和多因素Cox分析获得预后特征。采用受试者工作特征（ROC）曲线评估模型的预测效果。利用单样本基因集富集分析（ssGSEA）获得的免疫细胞浸润丰度和治疗靶基因，与获得的预后特征及相关基因进行充分的相关性分析。
从TCGA数据库共获取173个样本，其中包括162个肿瘤样本和11个正常样本。通过将3033个差异基因与从ImmPort数据库获得的2483个IRGs进行交叉分析，得到254个DEIRGs。最后，多因素Cox回归分析确定了8个预后DEIRGs，并建立了一个新的预后特征（HR：2.49，95%CI：1.68 - 3.67；P < 0.001）。根据这8个基因的表达情况，将队列分为高风险组和低风险组，并绘制Kaplan-Meier（K-M）曲线，对数秩检验P < 0.0001，1年和3年曲线下面积（AUC）> 0.7。在两个亚组验证队列中，根据高风险和低风险分组的K-M曲线表现良好，对数秩检验P < 0.05。肿瘤样本和正常样本中16种免疫细胞的浸润程度存在差异，高风险组和低风险组中12种免疫细胞的浸润丰度也不同。
建立了一个由IRGs组成的有效且经过验证的预后特征，其与免疫细胞及药物治疗的靶基因具有很强的相关性。
