基于5个PI3K通路相关基因构建肺腺癌预后预测模型

Construction prognosis model for lung adenocarcinoma based on 5 PI3K pathway related genes

目的基于磷脂酰肌醇-3激酶(PI3K)通路相关基因构建肺腺癌患者预后预测模型。方法癌症基因组图谱(TCGA)数据库获取了497例肺腺癌及54例正常组织表达谱数据及相匹配的临床病理资料;并通过系统检索获取了PI3K通路关键基因集合。差异表达的PI3K相关基因经由肿瘤和正常组织比较产生。基于差异表达基因,采用单因素和多因素Cox回归分析构建模型。并根据风险评分,将患者分为高死亡风险组和低死亡风险组;采用生存分析(K-M)和受试者工作特征(ROC)曲线分析检验预测效能。结果分析共筛选出34个显著差异表达的PI3K通路相关基因用于单因素Cox回归分析。多因素Cox回归分析确定了5个与总体生存率显著相关的差异表达基因用于模型构建(F2RL1、TWIST、FGR、KLF4和CEP55)。K-M分析结果提示高危组患者生存百分比较低危组低,差异有统计学意义(P<0.05)。ROC曲线结果证实该模型较临床病理特征预测效能更强。此外,多因素Cox回归分析提示该模型为肺腺癌患者预后预测的独立风险因子(P<0.05)。结论本研究构建的以PI3K通路相关基因为基础的模型可用于肺腺癌患者的预后评估。

Objective To establish a prognostic model based on phosphatidylinositol 3-kinase(PI3 K)-related genes for lung adenocarcinoma(LUAD) patients.Methods The expression profiles of 497 LUAD cases and 54 normal controls were derived from The Cancer Genome Atlas(TCGA) database,as well as the matched clinicopathological data.We performed a systematic retrieve to identify the PI3 K-related genes.Subsequently,the differentially expressed PI3K pathway-related genes were determined by comparing the LUAD cases and normal controls.Based on these gene candidates,univariate and multivariate Cox regression analyses were used to construct a prognostic prediction model.According to the median risk score,the patients were assigned into the low-death-risk and high-death-risk subgroups.Kaplan-Meier(K-M) and Receiver Operating Characteristic(ROC) curve analyses were used to determine the predicting performance.Results A total of 34 differentially expressed PI3K pathway-related genes were identified for following univariate Cox regression analysis.Subsequently,multivariate Cox regression analysis identified five overall survival-related differentially expressed genes to establish the prognostic model(F2RL1,TWIST,FGR,KLF4, and CEP55).KM analysis showed that patients in the high-risk group had worse outcomes than those in the low-risk group(logrank P<0.05).The ROC curve results confirmed that the predictive value of the model,which was better than the existing clinicopathological indicators.Besides,multivariate Cox regression analysis proved that the model was an independent risk factor for predicting the prognosis of patients with lung adenocarcinoma(P<0.05).Conclusion The five PI3 K pathway-related gene-based signature is able to predict the overall survival of patients with lung adenocarcinoma.