Luminal B型乳腺癌预后相关基因模型的建立与临床应用

Establishment and clinical application of a prognostic gene model for Luminal B subtype breast cancer

目的基于Luminal B型乳腺癌的预后高度差异性建立预后风险评分模型,指导并预测Luminal B型乳腺癌患者预后生存能力。方法选取癌症基因组图谱(TCGA)数据库中194例Luminal B型乳腺癌数据集和119例癌旁正常乳腺组织进行差异基因表达分析。结合生物信息学方法,通过差异基因表达分析得到与预后相关的基因组,利用最小绝对收缩和选择算子(least absolute shrinkage and selection operator,LASSO)Cox回归分析构建预后模型,即对该基因组在正常乳腺组织与乳腺癌组织的差异表达构建与预后相关的风险模型,并对模型的临床应用进行验证。同时探讨该风险模型可能触发的分子机制。免疫组织化学法检测2015-01-01-2022-12-31南昌市人民医院乳腺中心64例三阳性乳腺癌组织相关基因的表达。结果利用LASSO Cox回归分析构建预后模型,当纳入11个基因时风险效能最高,数据集验证其95%置信区间差异有统计学意义,其1、3和5年的受试者工作特征(ROC)曲线下面积分别为0.975、0.829和0.885,且高危组预后较差,差异有统计学意义,均P<0.05。单因素Cox回归分析结果发现,风险评分与预后相关,HR=5.72,95%CI为3.34~9.81,P<0.001;多因素Cox回归分析结果显示,风险评分(HR=8.91,95%CI为4.23~18.72,P<0.001)和年龄(HR=1.03,95%CI为1.00~1.06,P=0.047)与预后相关。列线图直观有效地展示了不同变量对预后结局的影响,以及纳入与预后相关的因素构建列线图模型,评估列线图模型的C-index值结果显示,模型C-index值为0.84,95%CI为0.75~0.92,P<0.001;富集通路分析结果显示,高风险组显著富集在可变剪接氨酰-tRNA合成等通路,低风险组富集在细胞因子-细胞因子受体、原发性免疫缺陷和趋化因子受体等信号通路;肿瘤微环境与风险评估相关性显示,大部分免疫细胞在低风险组中得分高。免疫组织化学检测结果显示,ERCC6L、COL11A1、NEURL1、CXCL9、ROPN1、PPP1R16B和PIGR在三阳性乳腺癌组织标本中呈阳性表达,TLCD1、ABCA10、TMEM273和ARHGAP40呈弱阳性表达。结论构建的风险模型评分可为临床医生提供定量方法来预测预后,可以作为Luminal B型乳腺癌患者预后的生物标志物。

Objective Based on the highly different prognosis of Luminal B breast cancer,a prognostic risk scoring model was established to guide and predict the prognosis and survival of patients with Luminal B breast cancer.Methods Totally 194 cases of Luminal B breast cancer and 119 cases of normal breast tissue adjacent to cancer were selected from the database for differential gene expression analysis from the cancer genome atlas(TCGA).Combined with bioinformatics methods,the genome related to prognosis was obtained through differential gene expression analysis.The prognosis model was constructed by least absolute shrinkage and selection operator(LASSO)to conduct Cox regression analysis;that was,a risk model related to prognosis and constructed for the differential expression of the genome in normal breast tissue and breast cancer tissue,and the clinical application of the model was verified.Simultaneously we explored the molecular mechanisms that may trigger this risk model.Immunohistochemical method was used to detect the expression of related genes in 64triple positive breast cancer tissues from January 1,2015to December 31,2022in the Breast Center of Nanchang People's Hospital.Results LASSO Cox regression analysis was used to construct the prognosis model,when 11genes were included,the risk efficiency was the highest.The data set verified that its 95%confidence interval had significant statistical differences.The area under the ROC curve for 1,3,and 5years was 0.975,0.829,and 0.885,respectively.The prognosis of the high-risk group was poor,and the difference was statistically significant,all P<0.05.The results of univariate Cox regression analysis showed a correlation between risk score and prognosis,with HR=5.72,95%CIranging from 3.34to 9.81,P<0.001.The results of multivariate Cox regression analysis showed that risk score(HR=8.91,95%CI:4.23-18.72,P<0.001)and age(HR=1.03,95%CI:1.00-1.06,P=0.047)were associated with prognosis.The nomogram intuitively and effectively displayed the impact of different variables on prognosis outcomes,and included factors related to prognosis to construct a nomogramic model.The evaluation of the C-index value of the nomogramic model showed that the C-index value of the model was 0.84,with a 95%CI of 0.75-0.92,P<0.001.The analysis of enrichment pathways showed that the high-risk group was significantly enriched in pathways such as variable splicing and aminoacyl tRNA synthesis,while the low-risk group was enriched in signaling pathways such as cytokine cytokine receptors,primary immunodeficiency,and chemokine receptors.The correlation between tumor microenvironment and risk assessment showed that most immune cells scored high in the low-risk group.Immunohistochemical test results showed that ERCC6L,COL11A1,NEURL1,CXCL9,ROPN1,PPP1R16B,PIGR were positive in triple positive breast cancer tissue samples,and TLCD1,ABCA10,TMEM273,ARHGAP40were weakly positive.Conclusion The risk model score constructed can provide clinicians with a quantitative method to predict the prognosis,and can be used as a biomarker for the prognosis of Luminal B breast cancer patients.