Table4_A Novel Integrated Metabolism-Immunity Gene Expression Model Predicts the Prognosis of Lung Adenocarcinoma Patients.XLSX

Background: Although multiple metabolic pathways are involved in the initiation, progression, and therapy of lung adenocarcinoma (LUAD), the tumor microenvironment (TME) for immune cell infiltration that is regulated by metabolic enzymes has not yet been characterized.
Methods: 517 LUAD samples and 59 non-tumor samples were obtained from The Cancer Genome Atlas (TCGA) database as the training cohort. Kaplan-Meier analysis and Univariate Cox analysis were applied to screen the candidate metabolic enzymes for their role in relation to survival rate in LUAD patients. A prognostic metabolic enzyme signature, termed the metabolic gene risk score (MGRS), was established based on multivariate Cox proportional hazards regression analysis and was verified in an independent test cohort, GSE31210. In addition, we analyzed the immune cell infiltration characteristics in patients grouped by their Risk Score. Furthermore, the prognostic value of these four enzymes was verified in another independent cohort by immunohistochemistry and an optimized model of the metabolic-immune protein risk score (MIPRS) was constructed.
Results: The MGRS model comprising 4 genes (TYMS, NME4, LDHA, and SMOX) was developed to classify patients into high-risk and low-risk groups. Patients with a high-risk score had a poor prognosis and exhibited activated carbon and nucleotide metabolism, both of which were associated with changes to TME immune cell infiltration characteristics. In addition, the optimized MIPRS model showed more accurate predictive power in prognosis of LUAD.
Conclusion: Our study revealed an integrated metabolic enzyme signature as a reliable prognostic tool to accurately predict the prognosis of LUAD.

背景：尽管多种代谢通路参与肺腺癌（lung adenocarcinoma, LUAD）的发生、进展与治疗，但由代谢酶调控的、与免疫细胞浸润相关的肿瘤微环境（tumor microenvironment, TME）尚未得到明确阐释。
方法：从癌症基因组图谱（The Cancer Genome Atlas, TCGA）数据库获取517份LUAD样本与59份非肿瘤样本作为训练队列。采用卡普兰-迈耶（Kaplan-Meier）分析与单变量Cox分析，筛选与LUAD患者生存率相关的候选代谢酶。基于多变量Cox比例风险回归分析构建预后代谢酶特征模型，命名为代谢基因风险评分（metabolic gene risk score, MGRS），并在独立测试队列GSE31210中对该模型进行验证。此外，按风险评分对患者分组，分析其免疫细胞浸润特征。进一步通过免疫组化在另一独立队列中验证这4种酶的预后价值，并构建优化后的代谢-免疫蛋白风险评分（metabolic-immune protein risk score, MIPRS）模型。
结果：构建了包含4个基因（TYMS、NME4、LDHA及SMOX）的MGRS模型，用于将患者划分为高风险组与低风险组。高风险评分患者预后较差，且表现出碳代谢与核苷酸代谢激活的特征，这两种代谢通路均与TME免疫细胞浸润特征的改变相关。此外，优化后的MIPRS模型在LUAD预后预测中展现出更精准的预测能力。
结论：本研究揭示了整合的代谢酶特征可作为可靠的预后工具，精准预测LUAD患者的预后。