GGCT作为乳腺癌患者预后的预测因子

GGCT Serves as a Prognostic Predictor for Patients with Breast Cancer

摘要: 目的：乳腺癌是全球发病率最高的癌症，并且占女性肿瘤死亡率第一。准确预测预后和提供个体化治疗尤为重要。许多证据表明，遗传因素和临床病理特征与癌症的发病和进展有关。γ-谷氨基环转移酶(GGCT)是谷胱甘肽代谢的主要酶之一，可催化γ-谷氨酰肽产生5-氧脯氨酸和游离氨基酸。在许多癌症中，GGCT的表达水平显著升高并且与临床预后不佳有关。因此，本研究旨在探讨GGCT的表达水平与乳腺癌预后之间的关系，并开发一种诺模图来预测乳腺癌的生存和复发。方法：利用Oncomine、Timer2.0、GEPIA2.0、HPA、UALCAN、bc-GenExMinerv4.7、Kaplan-Meier plotter、Linked Omics、TCGA等在线数据库，分析GGCT的表达和乳腺癌的情况。结果：与正常组织相比，GGCT在乳腺癌组织中表达上调；GGCT的表达和乳腺癌不同分型以及组织分级相关；GGCT高表达的乳腺癌患者预后较差。利用Cox回归分析TCGA-BRCA RNA-seq数据，年龄、T、M、ER是整体生存的重要独立风险因素，M、临床分期和GGCT是无病生存的重要独立因素。两种预测模型分别预测OS和DFS的3年生存率和5年生存率。3年总生存为0.744，5年总生存为0.689，3年无病生存为0.744，5年无病生存为0.742。而TNM模型预测3年总生存为0.638，5年总生存为0.628，3年无病生存为0.659，5年无病生存为0.66。对时间依赖的ROC曲线进行头对头比较，新模型比TNM模型具有较高的预测精度。结论：GGCT可作为乳腺癌患者的预后标志物，与临床相结合建立的诺模图能更好地预测复发。

Abstract: Objective: Breast cancer is the most common cancer in the world and causes the most cancer deaths among women. It is important to accurately predict prognosis and provide individualized treatment. There is much evidence that genetic factors and clinicopathological features are associated with the onset and progression of cancer. γ-glutamylcyclotransferase (GGCT) is one of the major enzymes in glutathione metabolism, which catalyzes the production of 5-oxo-proline and free amino acids from γ-glutamyl peptide. GGCT expression levels are significantly elevated in many cancers and are associated with poor clinical prognosis. Therefore, this study aims to explore the relationship between GGCT expression level and prognosis of breast cancer, and to develop a nomogram to predict survival and recurrence of breast cancer. Methods: Oncomine, Timer2.0, GEPIA2.0, HPA, UALCAN, bc-GenExMinerv4.7, Kaplan-Meier plotter, Linked Omics, TCGA and other online databases were used to analyze the expression of GGCT and breast cancer. Results: Compared with normal tissues, GGCT expression was up-regulated in breast cancer tissues. The expression of GGCT was correlated with different types of breast cancer and tissue grading. Breast cancer patients with high GGCT expression have a poor prognosis. Cox regression analysis of TCGA-BRCA RNA-seq data showed that age, T, M and ER were important independent risk factors for global survival, while M, clinical stage and GGCT were important independent factors for disease-free survival. Two prediction models predicted the probability of 3-year survival and 5-year survival for OS and DFS, respectively. The concordance indexes were 0.744 for 3-year overall survival, 0.689 for 5-year overall survival, 0.744 for 3-year disease-free survival and 0.742 for 5-year disease-free survival. However, TNM model predicted that 3-year overall survival was 0.638, 5-year overall survival was 0.628, 3-year disease-free survival was 0.659, and 5-year disease-free survival was 0.66.The head-to-head comparison according to time-dependent ROC curves indicated that the new model has higher prediction accuracy than the TNM model. Conclusion: GGCT can be used as a prognostic marker for breast cancer patients, and the nomogram established by combining GGCT with clinical data can better predict recurrence.