基于自噬相关基因的肝细胞癌患者预后预测模型

Prognosis prediction model of hepatocellular carcinoma patients based on autophagy-related genes

目的:通过自噬相关基因表达水平来构建一个肝细胞癌患者预后预测模型并进行验证。方法:从TCGA及ICGC数据库下载肝细胞癌患者基因表达数据及相应临床数据。在TCGA组中通过单因素Cox分析找到肿瘤组织与正常组织间差异表达的自噬相关基因,再筛选出与患者总体生存期显著相关的差异表达基因作为候选建模基因,通过Lasso回归分析建立预后模型,并在ICGC组中验证模型。结果:232个自噬相关基因中有25.4%的基因在肿瘤组织和邻近的正常组织之间显著差异表达。在单因素Cox分析中,找到36个与患者总体生存期显著相关的差异表达基因。构建了一个囊括13个自噬相关基因的模型,并以风险值高低将患者分成两个组。低风险组患者的总体生存期相较于高风险组明显更长(在TCGA和ICGC中均P<0.001)。在单因素Cox分析和多因素Cox分析中风险值均与患者预后显著相关(P<0.001,HR>1),验证了风险值的独立预后因素价值。在时间依赖ROC曲线中肯定了模型的预测能力。功能分析表明,基因在细胞分裂和细胞周期相关的途径富集,且高低风险组间的免疫相关细胞及途径存在差异。结论:构建了一个新的基于自噬相关基因的肝细胞癌患者预后预测模型,可以帮助改善个体患者的预后预测。

Objective: To construct and verify a prognostic model for hepatocellular carcinoma patients which based on the expression levels of autophagy-related genes. Methods: Gene expression data and corresponding clinical data of patients with hepatocellular carcinoma from TCGA and ICGC databases were downloaded. In the TCGA group, the autophagy-related genes differentially expressed between tumor tissues and normal tissues were subjected to univariate Cox regression analysis, and differentially expressed genes that were significantly related to the overall survival of the patients were screened as candidate modeling genes. Then through Lasso regression analysis to establish a prognostic model,we verified the model in the ICGC group. Results: 25. 4% of the 232 autophagy-related genes were significantly differentially expressed between tumor tissues and adjacent normal tissues. In univariate Cox analysis, 36 differentially expressed genes that were significantly related to the overall survival of the patient were found. A model containing 13 autophagy-related genes was constructed, and patients were divided into two groups based on the risk score. The overall survival of patients in the low-risk group was significantly longer than that of the high-risk group(P<0. 001 in both TCGA and ICGC).In the univariate Cox analysis and the multivariate Cox analysis, the risk score was significantly related to the patient’s prognosis(P<0. 001, HR>1), which verified the independent prognostic value of the risk score. The predictive ability of the model was confirmed in the time-dependent ROC curve.Functional analysis showed that genes were enriched in pathways related to cell division and cell cycle, and there were differences in immune-related cells and pathways between high and low risk groups. Conclusion: A novel prognostic prediction model for hepatocellular carcinoma patients based on autophagy-related genes is constructed, which can help improve the prognosis prediction of individual patients.