膀胱癌代谢风险模型的构建及其应用

Construction and application of metabolic risk model for bladder cancer

目的:建立一个高效能的膀胱癌(BLCA)代谢风险模型,并应用多组学技术揭示不同风险人群的异质性特征。方法:利用TCGA-BLCA转录组数据,根据86条代谢通路中共1 660个代谢相关的基因,应用单因素COX分析、Lasso回归分析和多因素COX分析,构建BLCA代谢风险模型;应用转录组学、代谢组学和单细胞蛋白质组学技术揭示高风险和低风险人群的分子表达谱、代谢物表达谱及免疫微环境特征。结果:构建了一个由CYP2J2、GDPD3、IL4I1、ATP6V1B1、AKR1B1、ENGASE、CHPF、MBOAT7、CAD、FASN、AHCY、FLAD1和ALG3 13个基因组成的BLCA代谢风险模型;训练队列对应的受试者工作特征(ROC)曲线预测BLCA患者1年、3年和5年总生存期(OS)的曲线下面积(AUC)分别为0.796、0.702和0.717,广西队列对应的ROC曲线预测BLCA患者1年、2年和3年OS的AUC分别为0.866、0.810和0.816;Kaplan-Meier曲线分析显示,低风险组预后显著优于高风险组(P<0.001);高风险组和低风险组患者的代谢物存在显著改变,在高风险组中富集了焦谷氨酸、柠檬酸和尿酸等代谢物;高风险组患者募集了更多的免疫抑制性细胞亚群,免疫抑制程度更高。结论:本研究构建的代谢风险模型可较为准确地预测BLCA患者的临床预后;不同代谢风险模式的BLCA患者的代谢物和免疫微环境的异质性特征,为BLCA精准治疗方案的开发提供了潜在靶点。

Objective:To establish a high-efficiency metabolic risk model of bladder cancer(BLCA)and reveal the heterogeneity of different risk populations using multi-omics technology.Methods:Based on the transcriptome data of TCGA-BLCA,a total of 1,660 metabolism-related genes in 86 metabolic pathways were analyzed by univariate COX analysis,Lasso regression analysis and multivariate COX analysis to construct the metabolic risk model of BLCA.Transcriptomics,metabolomics and single-cell proteomics techniques were used to reveal the molecular expression profiles,metabolite expression profiles and immune microenvironment characteristics of high-risk and low-risk populations.Results:A BLCA metabolic risk model was constructed consisting of 13 genes:CYP2J2,GDPD3,IL4I1,ATP6V1B1,AKR1B1,ENGASE,CHPF,MBOAT7,CAD,FASN,AHCY,FLAD1 and ALG3.The areas under the curves(AUC)predicted by receiver operating characteristic(ROC)curves corresponding to the training cohort for 1-year,3-year and 5-year overall survival(OS)of BLCA patients were 0.796,0.702 and 0.717,respectively.The corresponding areas under the ROC curves for the Guangxi cohort predicting the 1-year,2-year,and 3-year overall survival(OS)of BLCA patients were 0.866,0.810 and 0.816,respectively.Kaplan-Meier curve analysis showed that the prognosis of low-risk group was significantly better than that of high-risk group(P<0.001).There were significant changes in metabolites between the high-risk group and the low-risk group,and metabolites such as pyroglutamic acid,citric acid and uric acid were enriched in the high-risk group.Patients in the high-risk group recruited more immunosuppressive cell subsets and had a higher degree of immunosuppression.Conclusion:The metabolic risk model constructed in this study can accurately predict the clinical prognosis of BLCA patients.The heterogeneity of metabolites and immune microenvironment in BLCA patients with different metabolic risk patterns provide potential targets for the development of precision treatment plans for BLCA.