DataSheet3_Identification and Validation of EMT-Related lncRNA Prognostic Signature for Colorectal Cancer.PDF

Background: This study aimed to explore the biological functions and prognostic role of Epithelial-mesenchymal transition (Epithelial-mesenchymal transition)-related lncRNAs in colorectal cancer (CRC).
Methods: The Cancer Genome Atlas database was applied to retrieve gene expression data and clinical information. An EMT-related lncRNA risk signature was constructed relying on univariate Cox regression, Least Absolute Shrinkage and Selector Operation (LASSO) and multivariate Cox regression analysis of the EMT-related lncRNA expression data and clinical information. Then, an individualized prognostic prediction model based on the nomogram was developed and the predictive accuracy and discriminative ability of the nomogram were determined by the receiver operating characteristic curve and calibration curve. Finally, a series of analyses, such as functional analysis and unsupervised cluster analysis, were conducted to explore the influence of independent lncRNAs on CRC.
Results: A total of 581 patients were enrolled and an eleven-EMT-related lncRNA risk signature was identified relying on the comprehensive analysis of the EMT-related lncRNA expression data and clinical information in the training cohort. Then, risk scores were calculated to divide patients into high and low-risk groups, and the Kaplan-Meier curve analysis showed that low-risk patients tended to have better overall survival (OS). Multivariate Cox regression analysis indicated that the EMT-related lncRNA signature was significantly associated with prognosis. The results were subsequently confirmed in the validation dataset. Then, we constructed and validated a predictive nomogram for overall survival based on the clinical factors and risk signature. Functional characterization confirmed this signature could predict immune-related phenotype and was associated with immune cell infiltration (i.e., macrophages M0, M1, Tregs, CD4 memory resting cells, and neutrophils), tumor mutation burden (TMB).
Conclusions: Our study highlighted the value of the 11-EMT-lncRNA signature as a predictor of prognosis and immunotherapeutic response in CRC.

背景：本研究旨在探讨上皮间质转化（Epithelial-mesenchymal Transition, EMT）相关长链非编码RNA（long non-coding RNA, lncRNA）在结直肠癌（Colorectal Cancer, CRC）中的生物学功能及预后价值。
方法：本研究通过癌症基因组图谱（The Cancer Genome Atlas, TCGA）数据库获取基因表达数据与临床信息。基于单变量Cox回归、最小绝对收缩和选择算子（Least Absolute Shrinkage and Selection Operator, LASSO）以及多变量Cox回归，结合EMT相关lncRNA表达数据与临床信息，构建EMT相关lncRNA风险特征模型。随后，基于列线图（nomogram）建立个体化预后预测模型，并通过受试者工作特征曲线（Receiver Operating Characteristic curve, ROC曲线）与校准曲线评估该列线图的预测准确性与区分能力。最后，通过功能分析、无监督聚类分析等一系列手段，探究各独立EMT相关lncRNA对结直肠癌的影响。
结果：训练队列中共纳入581例患者，通过对其EMT相关lncRNA表达数据与临床信息进行综合分析，筛选出11个EMT相关lncRNA构建风险特征模型。通过计算风险评分将患者分为高风险组与低风险组，卡普兰-迈耶曲线（Kaplan-Meier curve）分析显示，低风险组患者的总生存期（Overall Survival, OS）更优。多变量Cox回归分析表明，该EMT相关lncRNA风险特征模型与患者预后显著相关，该结果在验证队列中得到证实。随后，基于临床因素与风险特征模型，构建并验证了总生存期预测列线图。功能分析结果证实，该风险模型可预测免疫相关表型，并与免疫细胞浸润水平相关，包括M0型巨噬细胞、M1型巨噬细胞、调节性T细胞（Regulatory T cells, Tregs）、静息CD4记忆性T细胞以及中性粒细胞，同时与肿瘤突变负荷（Tumor Mutation Burden, TMB）存在关联。
结论：本研究证实，11个EMT相关lncRNA风险特征模型可作为结直肠癌患者预后与免疫治疗反应的预测指标，具有重要临床应用价值。