基于免疫相关长链非编码RNA构建肝癌预后模型

Based on immune-related long non-coding RNAs risk model for prognosis prediction in hepatocellular carcinoma

目的探讨免疫相关长链非编码RNA(lncRNA)预测肝癌(HCC)患者预后生存情况。方法从癌症基因组图谱(TCGA)及Imlnc数据库获取与肝癌中差异表达且免疫相关的lncRNAs,构建加权的共表达网络后使用模块挖掘法挖掘肝癌相关模块,从模块中筛选肝癌预后相关的lncRNA。将肝癌患者样本等量分为训练集和测试集,基于训练集数据构建风险评分模型,基于训练集和测试集样本验证模型的有效性,并结合临床特征分析验证模型的稳定性。使用Lasso回归构建预后模型,Kaplan-Meier生存分析对比高低风险组预后、单因素和多因素Cox回归鉴定肝癌预后危险因素。结果本研究构建了由7个lncRNA构成的肝癌预后模型,根据模型风险评分将患者分为高风险组和低风险组,对肝癌患者1、3、5年生存预测的AUC值分别为0.76、0.68和0.62。高风险组总体生存时间在年龄>60岁[风险比(HR)=2.432,P<0.05]、男(HR=2.383,P<0.05),女(HR=2.396,P<0.05)、肿瘤T1~T2分级(HR=2.320,P<0.05)、肿瘤分期Ⅰ期与Ⅱ期(HR=2.223,P<0.05),Ⅲ期与Ⅳ期(HR=2.310,P<0.05)方面低于低风险组,风险分层分析结果表明模型的风险评分是独立预后指标(HR=2.243,P<0.05)。结论本研究构建的lncRNAs预后预测模型具有较好可信度和稳定性,表明这些lncRNA在肝癌发生发展中起重要作用。

Objective To investigate the prognostic survival of hepatocellular carcinoma(HCC)patients by immune-related long non-coding RNA(lncRNA).Methods The differentially expressed and immune-related lncRNAs in HCC were obtained from the Cancer Genome Atlas(TCGA)and Imlnc databases.After the weighted co-expression network was constructed,the module mining method was used to mine the HCC related modules,and the lncRNAs related to the prognosis of HCC were screened from the modules.The samples of hepatocellular carcinoma patients were divided into training set and test set,and the risk score model was constructed based on the data of the training set.The effectiveness of the model was verified based on the samples of training set and test set,and the stability of the model was verified by analyzing the clinical characteristics.Kaplan-meier survival analysis was utilized to compare the prognosis between high and low risk group.Univariate and multivariate Cox regression were used to identify the prognostic risk factors of HCC.Results A HCC prognosis model was constructed by 7 lncRNAs,which could divide patients into the high-risk group and the low-risk group according risk score.Predictive AUC values of this model for predicting the 1-,3-and 5-years survival were 0.76,0.68 and 0.62,respectively.The overall survival time of high-risk group was in age≥60 years[hazard ratio(HR)=2.432,P<0.05],male(HR=2.383,P<0.05),female(HR=2.396,P<0.05),tumor grade T1-T2(HR=2.320,P<0.05),tumor stageⅠandⅡ(HR=2.223,P<0.05),ⅢandⅣ(HR=2.310,P<0.05)were lower than those of low-risk group.Risk stratification analysis showed that the risk score of the model was the independent prognostic indicator(HR=2.243,P<0.05).Conclusion The prognostic prediction model of lncRNAs constructed in this study has well reliability and stability,which also indicate that these lncRNAs play an important role in the occurrence and development of HCC.