Dong Yunfei, Shang Tao, Ji HaiXin, Zhou Xiukou, Chen Zhi
Department of Proctology, The First Affiliated Hospital of Zhejiang Chinese Medical University (Zhejiang Provincial Hospital of Traditional Chinese Medicine), Hangzhou, China.
Front Genet. 2021 Aug 6;12:685371. doi: 10.3389/fgene.2021.685371. eCollection 2021.
The pathological stage of colon cancer cannot accurately predict recurrence, and to date, no gene expression characteristics have been demonstrated to be reliable for prognostic stratification in clinical practice, perhaps because colon cancer is a heterogeneous disease. The purpose was to establish a comprehensive molecular classification and prognostic marker for colon cancer based on invasion-related expression profiling.
From the Gene Expression Omnibus (GEO) database, we collected two microarray datasets of colon cancer samples, and another dataset was obtained from The Cancer Genome Atlas (TCGA). Differentially expressed genes (DEGs) further underwent univariate analysis, least absolute shrinkage, selection operator (LASSO) regression analysis, and multivariate Cox survival analysis to screen prognosis-associated feature genes, which were further verified with test datasets.
Two molecular subtypes (C1 and C2) were identified based on invasion-related genes in the colon cancer samples in TCGA training dataset, and C2 had a good prognosis. Moreover, C1 was more sensitive to immunotherapy. A total of 1,514 invasion-related genes, specifically 124 downregulated genes and 1,390 upregulated genes in C1 and C2, were identified as DEGs. A four-gene prognostic signature was identified and validated, and colon cancer patients were stratified into a high-risk group and a low-risk group. Multivariate regression analyses and a nomogram indicated that the four-gene signature developed in this study was an independent predictive factor and had a relatively good predictive capability when adjusting for other clinical factors.
This research provided novel insights into the mechanisms underlying invasion and offered a novel biomarker of a poor prognosis in colon cancer patients.
结肠癌的病理分期无法准确预测复发，迄今为止，尚未有基因表达特征被证明在临床实践中可用于可靠的预后分层，这可能是因为结肠癌是一种异质性疾病。目的是基于侵袭相关表达谱建立一种全面的结肠癌分子分类和预后标志物。
从基因表达综合数据库（GEO）中，我们收集了两个结肠癌样本的微阵列数据集，另一个数据集来自癌症基因组图谱（TCGA）。对差异表达基因（DEGs）进一步进行单变量分析、最小绝对收缩和选择算子（LASSO）回归分析以及多变量Cox生存分析，以筛选与预后相关的特征基因，并在测试数据集中进一步验证。
基于TCGA训练数据集中结肠癌样本的侵袭相关基因，鉴定出两种分子亚型（C1和C2），且C2预后良好。此外，C1对免疫治疗更敏感。共鉴定出1514个侵袭相关基因，具体而言，C1和C2中有124个下调基因和1390个上调基因被确定为差异表达基因。鉴定并验证了一个四基因预后特征，将结肠癌患者分为高风险组和低风险组。多变量回归分析和列线图表明，本研究中开发的四基因特征是一个独立的预测因素，在调整其他临床因素时具有相对较好的预测能力。
本研究为侵袭的潜在机制提供了新的见解，并为结肠癌患者预后不良提供了一种新的生物标志物。
