基于坏死性凋亡相关长链非编码RNA构建肝细胞癌预后模型及药物治疗反应分析  

Prognostic model construction and drugtherapy response analysis of hepatocellular carcino⁃ma based on necroptosis-related long noncoding RNA

作　　者：张玉俊 朱琳[2] 赵璇 地力亚尔·吾斯曼江 王岩[2] ZHANG Yujun;ZHU Lin;ZHAO Xuan;Diliyaer·Wusimanjiang;WANG Yan(Public Health Institute of Xinjiang Medical University,Urumqi,Xinjiang Uyghur Autonomous Region 830054,China;Afliated Tumor Hospital,Xinjiang Medical University,Urumqi,Xinjiang Uyghur Autonomous Region 830011,China)

机构地区：[1]新疆医科大学公共卫生学院,新疆维吾尔自治区乌鲁木齐830054 [2]新疆医科大学附属肿瘤医院,新疆维吾尔自治区乌鲁木齐830011

出　　处：《安徽医药》2023年第8期1595-1601,I0007-I0010,共11页Anhui Medical and Pharmaceutical Journal

基　　金：新疆维吾尔自治区自然科学基金项目(2021D01C379);省部共建中亚高发病成因与防治国家重点实验室开放课题项目(SKL-HIDCA-2020-ER3,SKL-HIDCA-2020-33)。

摘　　要：目的使用坏死性凋亡相关长链非编码RNA(NRLs)构建肝细胞癌(HCC)预后模型并分析不同风险组间药物敏感性差异,为HCC病人预后预测和临床个体化治疗提供理论依据。方法从癌症基因组图谱(TCGA)数据库中下载HCC病人的RNA测序数据和临床信息。采用共表达网络分析鉴定NRLs。使用单变量Cox回归和LASSO-Cox回归构建预后模型,并在测试集和整个集合中进行验证。运用生存分析、受试者操作特征(ROC)曲线、临床病理分层相关性分析、多变量Cox回归、列线图和校准曲线来评估预后模型。随后,采用基因集富集分析(GSEA)不同风险群体间生物过程和功能的差异。使用单样本基因集富集分析(ssGSEA)来探讨不同风险群体与肿瘤免疫、浸润之间的关系,并采用Pearson相关分析HCC病人预后特征与免疫检查点表达的相关性。最后,使用药物敏感性分析20种化疗药物在不同风险群体中的IC50值。结果构建了由4个NRLs(ZFPM2-AS1、MKLN1-AS、LINC01116、AP003390.1)组成的风险评分(NRLs risk-Score)预后特征,并根据风险评分中位值将病人划分为高风险组和低风险组。生存分析表明,NRLs低风险组的总生存期(OS)显著高于高风险组;与临床病理特征相比,NRLs risk-Score具有更高的诊断效率,受试者操作特征曲线下面积(AUC)为0.74;临床病理变量分层生存分析表明,高风险组病人OS显著低于低风险组;多变量Cox结果显示,分期(Stage)和NRLs risk-Score可作为HCC病人的独立预后因子,结合临床病理学特征和NRLs risk-Score的列线图显示出较好的预测性能。GSEA分析表明,癌症相关通路主要在高风险组中富集。ssGSEA结果显示,NRLs预测特征与HCC病人的免疫状态显著相关。免疫检查点分析结果显示,高风险组病人的免疫检查点表达较高,表明可能受益于检查点阻滞剂免疫疗法的高风险组HCC病人中免疫功能更为活跃。化疗药物敏感性分析结果显示,Objective To construct a prognostic model of hepatocellular carcinoma(HCC)using necroptosis-related long noncoding RNAs(NRLS)and analyze the difference of drug sensitivity among different risk groups,so as to provide a theoretical basis for prognosis prediction and clinical individualized treatment of HCC patients.Methods RNA sequencing data and clinical information of HCC patients were downloaded from the cancer genome atlas(TCGA)database.NRLS were identified by co-expression network analysis.A prognostic model was constructed using univariate Cox regression and lasso Cox regression and validated in the test set and whole set.Survival analysis,receiver operating characteristic curve(ROC),stratified correlation analysis with clinicopathology,multivariate Cox regression,nomogram and calibration curve were used to evaluate the prognostic model.Subsequently,gene set enrichment analysis(GSEA)was employed to investigate the differences in biological processes and functions among different risk groups.Single sample gene set enrichment analysis(ssGSEA)was used to explore the relationship between different risk groups and the tumor microenvironment,and Pearson correlation was employed to analyze the correlation between predictive features and immune checkpoint expression in HCC patients.Finally,the IC50 values of 20 chemotherapeutic agents in different risk groups were analyzed using drug sensitivity.Results A risk score(NRLS risk-score)prognostic signature consisting of four NRLS(ZFPM2-AS1,MKLN1-AS,LINC01116,AP003390.1)was constructed,and patients were assigned into high-and low-risk groups according to the median risk score values.Survival analysis showed that the overall survival(OS)of the low-risk group with NRLS was significantly longer than that of the highrisk group.Compared with clinicopathological variables,the NRLS risk score signature had higher diagnostic efficiency with an area under the receiver operating characteristic curve(AUC)of 0.74.Survival analysis stratified by clinicopathological variables showed t

关 键 词：癌 肝细胞 坏死性凋亡 长链非编码RNA 肿瘤免疫微环境 总生存期 

分 类 号：R735.7[医药卫生—肿瘤]