基于焦亡相关基因构建乳腺癌预后风险模型

Construction of A Breast Cancer Prognostic Risk Model Based on Pyroptosis-related Genes

目的 利用公共数据库构建和评估乳腺癌焦亡相关预后模型。方法 在TCGA和GEO数据库中下载乳腺癌相关基因表达矩阵和临床信息,收集焦亡相关基因并探究差异表达基因及关键基因,通过LASSO回归构建基于TCGA数据库的焦亡相关风险评分模型,使用GSE20685和GSE42568数据集对模型进行验证,并探究风险评分与临床分子免疫特征的相关性,在高低风险亚组中分析功能富集途径及基因组改变状态,结合临床特征构建列线图模型并进行综合评估。结果 共收集35个焦亡基因,其中32个为差异表达基因,包括18个表达下调基因和14个表达上调基因,35个基因集中PYCARD、AIM2、IL18、CASP1、CASP5、CASP8、NLRC4、IL6、NLRP3和TNF为焦亡通路的关键基因;从32个焦亡相关差异表达基因中鉴定出5个具有预后价值的基因:CASP9、GSDMC、GZMA、IL18和PYCARD,使用LASSO-Cox回归分析构建评分模型:风险评分=0.107 146 60×GSDMC表达值-0.207 295 50×CASP9表达值-0.205 471 10×IL18表达值-0.038 234 58×PYCARD表达值-0.092 932 92×GZMA表达值;风险评分模型显示,CASP9、IL18、PYCARD、GZMA上调与预后良好相关,GSDMC上调与预后不良相关;高风险组预后较低风险组差(P<0.001),且免疫浸润水平较低;男性、导管癌、肿瘤大于2 cm、三阴亚型、高MSI和TMB状态的患者风险评分较高;功能分析显示,高风险组上调基因集中在上皮细胞生长、角质化及炎症因子介导的信号通路;高风险组下调基因富集于生物代谢反应;基因改变显示,TP53基因突变与乳腺癌发生细胞焦亡密切相关,且TP53-PIK3CA互斥突变;单因素和多因素Cox分析显示,焦亡风险评分可作为预测乳腺癌生存状态的独立预后指标,该列线图预测患者总体生存率性能良好(C-index=0.7618)。结论 本次构建的预后模型和对细胞焦亡的综合性探索有助于乳腺癌患者的预后风险预测和个体化治疗。

Objective To construct and evaluate a breast cancer pyroptosis-related prognostic model using public databases.Methods The expression matrix and clinical information of breast cancer-related genes were downloaded from TCGA and GEO databases. The pyroptosis-related genes were collected and the differentially expressed genes and key genes were explored. The pyroptosis-related risk scoring model based on TCGA database was constructed by LASSO regression. The GSE20685 and GSE42568 datasets were used to verify the model, and the correlation between risk score and clinical molecular immune characteristics was explored. The functional enrichment pathway and genomic change status were analyzed in high and low risk subgroups, and the nomogram model was constructed and comprehensively evaluated in combination with clinical characteristics.Results A total of 35pyroptosis genes were collected, of which 32 were differentially expressed genes, including 18 down-regulated genes and 14 up-regulated genes.PYCARD, AIM2, IL18, CASP1, CASP5, CASP8, NLRC4, IL6, NLRP3 and TNF were the key genes of pyroptosis pathway. Five genes with prognostic value were identified from 32 differentially expressed genes related to pyroptosis: CASP9, GSDMC, GZMA, IL18 and PYCARD. The scoring model was constructed by LASSO-Cox regression analysis: risk score=0.107 146 60×GSDMC expression value-0.207 295 50×CASP9 expression value-0.205 471 10×IL18 expression value-0.038 234 58 ×PYCARD expression value-0.092 932 92 ×GZMA expression value. The risk score model showed that the up-regulation of CASP9, IL18, PYCARD and GZMA was associated with good prognosis, and the up-regulation of GSDMC was associated with poor prognosis.The prognosis of the high-risk group was worse than that of the low-risk group(P<0.001), and the level of immune infiltration was lower. Patients with male, ductal carcinoma, tumor larger than 2 cm, triple negative subtype, high MSI and TMB status had higher risk scores. Functional analysis showed that the up-regulated genes in the high-risk group were concentrated in epithelial cell growth, keratinization and inflammatory factor-mediated signaling pathways;the down-regulated genes in the high-risk group were enriched in biological metabolic reactions;gene changes showed that TP53 gene mutation was closely related to pyroptosis in breast cancer, and TP53-PIK3 CA mutual exclusion mutation;univariate and multivariate Cox analysis showed that the pyroptosis risk score could be used as an independent prognostic indicator for predicting the survival status of breast cancer. The nomogram had a good performance in predicting the overall survival rate of patients(C-index = 0.7618).Conclusion The constructed prognostic model and the comprehensive exploration of pyroptosis are helpful for the prognostic risk prediction and individualized treatment of breast cancer patients.