Zhang Tongxin, Liu Jingyu, Wang Meihuan, Liu Xiao, Qu Jia, Zhang Huawei
Department of Ultrasound, Shandong Provincial Hospital Affiliated to Shandong First Medical University, Jinan, Shandong, China.
Front Oncol. 2024 Jan 4;13:1288909. doi: 10.3389/fonc.2023.1288909. eCollection 2023.
Breast cancer (BC) is the most common malignant tumor in the female population. Despite staging and treatment consensus guidelines, significant heterogeneity exists in BC patients' prognosis and treatment efficacy. Alterations in one-carbon (1C) metabolism are critical for tumor growth, but the value of the role of 1C metabolism in BC has not been fully investigated.
To investigate the prognostic value of 1C metabolism-related genes in BC, 72 1C metabolism-related genes from GSE20685 dataset were used to construct a risk-score model via univariate Cox regression analysis and the least absolute shrinkage and selection operator (LASSO) regression algorithm, which was validated on three external datasets. Based on the risk score, all BC patients were categorized into high-risk and low-risk groups. The predictive ability of the model in the four datasets was verified by plotting Kaplan-Meier curve and receiver operating characteristic (ROC) curve. The candidate genes were then analyzed in relation to gene mutations, gene enrichment pathways, immune infiltration, immunotherapy, and drug sensitivity.
We identified a 7-gene 1C metabolism-related signature for prognosis and structured a prognostic model. ROC analysis demonstrated that the model accurately predicted the 2-, 3-, and 5-year overall survival rate of BC patients in the four cohorts. Kaplan-Meier analysis revealed that survival time of high-risk patients was markedly shorter than that of low-risk patients (p < 0.05). Meanwhile, high-risk patients had a higher tumor mutational burden (TMB), enrichment of tumor-associated pathways such as the IL-17 signaling pathway, lower levels of T follicular helper (Tfh) and B cells naive infiltration, and poorer response to immunotherapy. Furthermore, a strong correlation was found between MAT2B and CHKB and immune checkpoints.
These findings offer new insights into the effect of 1C metabolism in the onset, progression, and therapy of BC and can be used to assess BC patients' prognosis, study immune infiltration, and develop potentially more effective clinical treatment options.
乳腺癌（BC）是女性人群中最常见的恶性肿瘤。尽管有分期和治疗的共识指南，但BC患者的预后和治疗效果仍存在显著异质性。一碳（1C）代谢的改变对肿瘤生长至关重要，但1C代谢在BC中的作用价值尚未得到充分研究。
为了研究1C代谢相关基因在BC中的预后价值，我们使用来自GSE20685数据集的72个1C代谢相关基因，通过单变量Cox回归分析和最小绝对收缩和选择算子（LASSO）回归算法构建了一个风险评分模型，并在三个外部数据集上进行了验证。根据风险评分，将所有BC患者分为高风险和低风险组。通过绘制Kaplan-Meier曲线和受试者工作特征（ROC）曲线，验证了该模型在四个数据集中的预测能力。然后对候选基因进行了基因突变、基因富集途径、免疫浸润、免疫治疗和药物敏感性分析。
我们确定了一个与预后相关的7基因1C代谢特征，并构建了一个预后模型。ROC分析表明，该模型准确预测了四个队列中BC患者的2年、3年和5年总生存率。Kaplan-Meier分析显示，高风险患者的生存时间明显短于低风险患者（p<0.05）。同时，高风险患者具有更高的肿瘤突变负担（TMB），肿瘤相关途径如IL-17信号通路富集，T滤泡辅助细胞（Tfh）和幼稚B细胞浸润水平较低，对免疫治疗的反应较差。此外，发现MAT2B和CHKB与免疫检查点之间存在强相关性。
这些发现为1C代谢在BC的发生、发展和治疗中的作用提供了新的见解，可用于评估BC患者的预后、研究免疫浸润，并开发可能更有效的临床治疗方案。
