基于M1/M2型巨噬细胞浸润相关特征基因构建乳腺癌预后风险预测模型

Construction of a breast cancer prognostic risk prediction model based on characteristic genes related to M1/M2 macrophage infiltration

目的:基于M1/M2型巨噬细胞浸润特征基因构建乳腺癌预后风险预测模型,识别新的乳腺癌预后标志物,以促进乳腺癌精准治疗策略的发展。方法:利用癌症基因图谱(TCGA)数据库中的乳腺癌RNA-seq数据和临床信息,结合加权基因共表达网络分析(WGCNA)、CIBERSORT、xCell、QUANTISEQ和单因素Cox回归筛选与乳腺癌预后相关的M1/M2型巨噬细胞特征基因。通过LASSO-Cox回归建立预后风险模型,使用ROC曲线、Kaplan-Meier生存曲线和列线图评估模型。同时,使用ESTIMATE算法和TIDE评分分析肿瘤微环境和免疫治疗响应性。结果:识别了91个与乳腺癌预后密切相关的特征基因。基于这些基因,构建了含39个关键基因的预后风险模型,同时计算了风险评分。该模型准确预测了乳腺癌患者1年、3年和5年生存预后,风险评分被证实为预后的独立预测因子。对患者进行风险分层,高风险组生存期显著短于低风险组(P<0.001)。列线图验证了风险评分的临床价值。低风险组患者相对于高风险组表现出更高的免疫分数和ESTIMATE分数,以及较低的肿瘤纯度和TIDE评分,这暗示他们可能更能从免疫治疗中获益。结论:本研究基于M1/M2型巨噬细胞浸润相关特征基因成功构建了乳腺癌预后风险预测模型。该模型不仅能够对乳腺癌患者预后进行有效的评估和精准预测,还能为免疫治疗的临床决策提供参考依据,有助于乳腺癌个体化精准治疗。

Objective:To construct a breast cancer prognostic risk prediction model based on the gene infil-tration characteristics of M1/M2 macrophages,identify new prognostic biomarkers for breast cancer,and promote the development of precision treatment strategies for breast cancer.Methods:Utilizing breast cancer RNA-seq data and clinical information from the Cancer Genome Atlas(TCGA)data-base,we combined weighted gene co-expression network analysis(WGCNA),CIBERSORT,xCell,QUANTISEQ,and univariate Cox regression to screen for M1/M2 macrophage characteristic genes associated with breast cancer prognosis.A prognostic risk model was established through LASSO-Cox regression,then the model was evaluated by ROC curves,Kaplan-Meier survival curves,and no-mogram.Additionally,the tumor microenvironment and immune therapy response were analyzed by using the ESTIMATE algorithm and TIDE scores.Results:A total of 91 characteristic genes closely related to breast cancer prognosis were identified.Based on these genes,a prognostic risk model con-taining 39 key genes was constructed,along with a risk score calculation.The model accurately pre-dicted 1-year,3-year,and 5-year survival prognoses,with the risk score being confirmed as an inde-pendent prognostic factor.Patients in the high-risk group had significantly shorter survival than those in the low-risk group(P<0.001).Forest plots validated the clinical value of the risk score.Patients in the low-risk group exhibited higher immune scores and ESTIMATE scores,and lower tumor purity and TIDE scores,suggesting they might benefit more from immunotherapy.Conclusion:This study successfully constructed a breast cancer prognostic risk prediction model based on M1/M2 macrophage infiltration-related characteristic genes.The model not only effectively assesses and accurately predicts the prognosis of breast cancer patients,but also provides a reference for clinical decision-making in im-munotherapy,contributing to the individualized precision treatment of breast cancer.