Department of Epidemiology and Biostatistics, College of Public Health, Zhengzhou University, Zhengzhou 450001, China.
Zhengzhou University Library, Zhengzhou University, Zhengzhou 450001, China.
Biomed Res Int. 2021 Mar 30;2021:8864436. doi: 10.1155/2021/8864436. eCollection 2021.
The development of immunotherapy has greatly changed the advanced-stage non-small-cell lung cancer (NSCLC) treatment landscape. The complexity and heterogeneity of tumor microenvironment (TME) lead to discrepant immunotherapy effects among patients at the same pathologic stages. This study is aimed at exploring potential biomarkers of immunotherapy and accurately predicting the prognosis for advanced NSCLC patients. RNA-seq data and clinical information on stage III/IV NSCLC were obtained from The Cancer Genome Atlas (TCGA) and Gene Expression Omnibus (GEO). In TCGA-NSCLC with stage III/IV ( = 192), immune scores and stromal scores were calculated by using the ESTIMATE algorithms. Univariate, LASSO, and multivariate Cox regression analyses were performed to screen prognostic TME-related genes (TMERGs) and constructed a gene signature risk score model. It was validated in external dataset including GSE41271 ( = 91) and GSE81089 ( = 36). Additionally, a nomogram incorporating TMERG signature risk score and clinical characteristics was established. Further, we accessed the proportion of 22 types of tumor-infiltrating immune cells (TIIC) from the CIBERSORT website and analyzed the difference between two risk groups. OS of patients with high immune/stromal scores were higher (log-rank = 0.044/log-rank = 0.048). Multivariate Cox regression identified six prognostic TMERGs, including CD200, CHI3L2, CNTN1, CTSL, FYB1, and SLC52A1. We developed a six-gene risk score model, which was validated as an independent prognostic factor for OS (HR: 3.32, 95% CI: 2.16-5.09). Time-ROC curves showed useful discrimination for TCGA-NSCLC cohort (1-, 2-, and 3-year AUCs were 0.718, 0.761, and 0.750). The predictive robustness was validated in the external dataset. The C-index and 1-, 2-, and 3-year AUCs of nomogram were the largest, which demonstrated the nomogram had the greatest predictive accuracy and effectiveness and could be used for clinical guidance. Besides, the increased infiltration of T cells regulatory (Tregs) and macrophages M2 in the high-risk group suggested that chronic inflammation may reduce survival probability in patients with advanced NSCLC. We conducted a comprehensive analysis of the tumor microenvironment and identified the TMERG signature, which could predict prognosis accurately and provide a reference for the personalized immunotherapy for advanced NSCLC patients.
免疫疗法的发展极大地改变了晚期非小细胞肺癌（NSCLC）的治疗格局。肿瘤微环境（TME）的复杂性和异质性导致同一病理阶段的患者免疫治疗效果存在差异。本研究旨在探索免疫治疗的潜在生物标志物，并准确预测晚期 NSCLC 患者的预后。从癌症基因组图谱（TCGA）和基因表达综合数据库（GEO）中获得了 III/IV 期 NSCLC 的 RNA-seq 数据和临床信息。在 TCGA-NSCLC 中，III/IV 期（=192），使用 ESTIMATE 算法计算免疫评分和基质评分。进行单因素、LASSO 和多变量 Cox 回归分析，筛选与预后相关的肿瘤微环境基因（TMERGs），构建基因特征风险评分模型。在外部数据集 GSE41271（=91）和 GSE81089（=36）中进行了验证。此外，还建立了一个包含 TMERG 特征风险评分和临床特征的列线图。进一步，我们从 CIBERSORT 网站获取了 22 种肿瘤浸润免疫细胞（TIIC）的比例，并分析了两个风险组之间的差异。高免疫/基质评分患者的 OS 更高（log-rank =0.044/log-rank =0.048）。多变量 Cox 回归鉴定了 6 个预后 TMERGs，包括 CD200、CHI3L2、CNTN1、CTSL、FYB1 和 SLC52A1。我们开发了一个六基因风险评分模型，验证其为 OS 的独立预后因素（HR：3.32，95%CI：2.16-5.09）。时间 ROC 曲线表明 TCGA-NSCLC 队列具有良好的区分能力（1、2 和 3 年 AUC 分别为 0.718、0.761 和 0.750）。该预测的稳健性在外部数据集得到验证。列线图的 C 指数和 1、2 和 3 年 AUC 最大，表明列线图具有最大的预测准确性和有效性，可用于临床指导。此外，高危组中 T 细胞调节性（Tregs）和巨噬细胞 M2 的浸润增加提示慢性炎症可能降低晚期 NSCLC 患者的生存概率。我们对肿瘤微环境进行了全面分析，确定了 TMERG 特征，该特征可以准确预测预后，为晚期 NSCLC 患者的个性化免疫治疗提供参考。
