微创医学

目的通过生物信息学方法筛选雌激素受体（ER）阳性乳腺癌中与神经炎症相关的枢纽基因并构建预后预测模型，探讨ER阳性乳腺癌的潜在治疗靶点，筛选靶向药物。方法从癌症基因组图谱（TCGA）数据库获取840例ER阳性乳腺癌及正常对照样本，进行差异基因表达分析，结合GeneCards数据库的神经炎症基因，筛选与差异表达基因（DEGs）的共有基因；通过对共有基因进行免疫浸润分析、基因本体论（GO）功能和京都基因与基因组百科全书（KEGG）通路富集分析，探究其生物学功能；利用多因素COX回归模型构建预后预测模型并验证其预测效能；分析关键基因可能的微小RNA（miRNA）调控网络，通过比较毒理基因组学数据库（CTD）预测潜在治疗药物。结果共获得2 529个DEGs，包括1 383个上调基因和1 146个下调基因。共获得100个ER阳性乳腺癌与神经炎症相关的共有基因。共有基因显著富集于癌症相关信号通路、内环境稳态及免疫炎症反应等，涉及多种生物过程、参与细胞组分与分子功能，免疫浸润分析显示多种免疫细胞丰度改变。获得11个ER阳性乳腺癌患者预后相关的基因，基于11个预后相关基因构建风险评分预后预测模型具有良好的预测效能［3年的受试者操作特征曲线下面积（AUC）=0.77，95%CI（0.67，0.88）；5年的AUC=0.80，95%CI（0.72，0.87）］。11个预后相关基因受153个miRNA及43个转录因子调控，基于CTD数据库对与共有基因相互作用的化合物进行筛选，预测出多柔比星、白藜芦醇等29种潜在药物。结论神经炎症相关基因通过调控炎症通路、重塑免疫微环境参与ER阳性乳腺癌进展，筛选出11个ER阳性乳腺癌患者预后相关基因，基于此11个预后相关基因构建的预后预测模型具有较好的预测价值，多柔比星是临床常用化学治疗药物，而白藜芦醇的抗炎特性为联合用药治疗提供新思路。

Objective Bioinformatics analysis was performed to screen neuroinflammation-related hub genes and construct a prognostic model in estrogen receptor (ER)-positive breast cancer, exploring potential therapeutic targets for ER-positive breast cancer and screening targeted drugs. Methods The data of 840 ER-positive breast cancer and normal control samples were obtained from The Cancer Genome Atlas (TCGA) database for differentially expressed gene analysis. Combined with neuroinflammation-related genes retrieved from the GeneCards database, overlapping genes between neuroinflammation genes and differentially expressed genes (DEGs) were screened. Immune infiltration analysis, Gene Ontology (GO) functional enrichment analysis and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis were performed on these overlapping genes to explore their biological functions. A multivariate COX regression model was used to construct a prognostic prediction model, and its predictive efficiency was validated. The potential microRNA (miRNA) regulatory network of key genes was analyzed, and candidate therapeutic drugs were predicted via the Comparative Toxicogenomics Database (CTD). Results A total of 2,529 DEGs were identified, including 1,383 upregulated genes and 1,146 downregulated genes. One hundred overlapping genes associated with neuroinflammation in ER-positive breast cancer were obtained. These overlapping genes were significantly enriched in cancer-related signaling pathways, homeostasis, immune and inflammatory responses, participating in diverse biological processes, cellular components and molecular functions. Immune infiltration analysis revealed altered abundances of multiple immune cells. Eleven genes related to the prognosis of patients with ER-positive breast cancer were screened out. The risk score prognostic prediction model constructed based on these 11 prognostic-related genes exhibited favorable predictive efficiency (3-year AUC=0.77, 95%CI [0.67, 0.88]; 5-year AUC=0.80, 95%CI [0.72, 0.87]). The 11 prognostic-related genes were regulated by 153 miRNAs and 43 transcription factors. Compounds interacting with overlapping genes were screened via the CTD database, and 29 candidate agents such as doxorubicin and resveratrol were predicted as potential therapeutic drugs. Conclusion Neuroinflammation‑related genes participate in the progression of ER‑positive breast cancer by regulating inflammatory pathways and remodeling the immune microenvironment. A total of 11 prognosis‑related genes were screened, and the prognostic prediction model constructed based on these genes exhibited favorable predictive value. Doxorubicin is a widely used clinical chemotherapeutic agent, while the anti‑inflammatory properties of resveratrol provide novel insights for combined therapeutic strategies.