Identification of Early Recurrence Factors in Childhood and Adolescent B-Cell Acute Lymphoblastic Leukemia Based on Integrated Bioinformatics Analysis.

Over the past 50 years, great progress has been made in the diagnosis and treatment of acute lymphoblastic leukemia (ALL), especially in pediatric patients. However, early recurrence is still an important threat to the survival of patients. In this study, we used integrated bioinformatics analysis to look for biomarkers of early recurrence of B-cell ALL (B-ALL) in childhood and adolescent patients. Firstly, we obtained gene expression profiles from the Therapeutically Applicable Research to Generate Effective Treatments (TARGET) database and the Gene Expression Omnibus (GEO) database. Then, we identified differentially expressed genes (DEGs) based on whether the disease relapsed early. LASSO and Cox regression analysis were applied to identify a subset of four genes: HOXA7, S100A11, S100A10, and IFI44L. A genetic risk score model was constructed based on these four optimal prognostic genes. Time-dependent receiver operating characteristic (ROC) curves were used to evaluate the predictive value of this prognostic model (3-, 5-, and 10-year AUC values >0.7). The risk model was significantly associated with overall survival (OS) and event-free survival in B-ALL (all p < 0.0001). In addition, a high risk score was an independent poor prognostic risk factor for OS (p < 0.001; HR = 3.396; 95% CI: 2.387–4.832). Finally, the genetic risk model was successfully tested in B-ALL using an external validation set. The results suggested that this model could be a novel predictive tool for early recurrence and prognosis of B-ALL.

在过去的50年中，急性淋巴细胞白血病（ALL）的诊断和治疗取得了巨大进展，尤其是在儿童患者中。然而，早期复发仍然是对患者生存的一个重要威胁。在本研究中，我们利用综合生物信息学分析来寻找儿童和青少年B细胞急性淋巴细胞白血病（B - ALL）早期复发的生物标志物。首先，我们从治疗应用研究以产生有效治疗（TARGET）数据库和基因表达综合数据库（GEO）获取基因表达谱。然后，我们根据疾病是否早期复发确定差异表达基因（DEGs）。应用套索回归（LASSO）和考克斯回归分析确定了四个基因的子集：HOXA7、S100A11、S100A10和IFI44L。基于这四个最佳预后基因构建了一个基因风险评分模型。采用时间依赖性受试者工作特征（ROC）曲线评估该预后模型的预测价值（3年、5年和10年的曲线下面积值>0.7）。该风险模型与B - ALL的总生存期（OS）和无事件生存期显著相关（所有p < 0.0001）。此外，高风险评分是OS的一个独立的不良预后风险因素（p < 0.001；风险比 = 3.396；95%置信区间：2.387 - 4.832）。最后，利用外部验证集在B - ALL中成功测试了该基因风险模型。结果表明，该模型可能是B - ALL早期复发和预后的一种新型预测工具。