Zhang Di, Li Yongjian, Liu Tingting, Liu Xiaomin, Zhang Jingru
Department of Hematology, Qilu Hospital of Shandong University, Jinan, Shandong, People's Republic of China.
Blood Lymphat Cancer. 2025 Aug 18;15:115-132. doi: 10.2147/BLCTT.S529074. eCollection 2025.
Acute myeloid leukemia (AML) prognosis remains challenging due to limited biomarkers integrating tumor microenvironment (TME) dynamics. Neutrophils, key mediators of immune regulation, exhibit dual roles in cancer progression, yet their prognostic significance in AML is poorly defined. This study aimed to construct a neutrophil-related gene signature for AML risk stratification and explore its clinical and immunological implications.
Utilizing transcriptomic and clinical data from TCGA (The Cancer Genome Atlas), GEO (Gene Expression Omnibus), and OHSU cohorts (n=1537), we identified 148 neutrophil-related genes through literature mining. Prognostic genes were selected via univariate Cox regression and LASSO regression (R packages: survival, glmnet). A 5-gene model (CSF3R, BRAF, FFAR2, CD300A, CD37) was validated across internal (TCGA) and external cohorts (GSE10358, GSE14468, OHSU). Immune profiling, drug sensitivity analysis (GDSC database), and TIDE scoring were performed to assess immunotherapy relevance.
The neutrophil-based model stratified AML patients into high- and low-risk groups with distinct overall survival (OS, <0.0001 in TCGA). Multivariate Cox analysis confirmed its independence from age, FLT3, and TP53 mutations (HR=2.14, =0.015). CD37 emerged as the strongest prognostic marker (AUC 5-year=0.680, =0.0026), correlating with immunosuppressive TME features: elevated myeloid-derived suppressor cells (MDSCs, <0.01), Treg infiltration ( <0.05), and upregulated immune checkpoints (PD1, CTLA4, LAG3; <0.001). High CD37 expression predicted immunotherapy responsiveness (TIDE score, =0.004) and interacted with 146 potential therapeutic agents (eg, BCL2 inhibitors).
This study advances a novel 5-gene prognostic model integrating neutrophil biology into AML risk stratification. CD37, a key regulator of immune evasion, serves as a dual biomarker for prognosis and immunotherapy prediction. While validated across multiple cohorts, experimental studies are warranted to unravel CD37's mechanistic role. Our findings highlight the potential of neutrophil-centric biomarkers in guiding personalized AML therapy.
由于整合肿瘤微环境(TME)动态变化的生物标志物有限,急性髓系白血病(AML)的预后仍然具有挑战性。中性粒细胞是免疫调节的关键介质,在癌症进展中发挥双重作用,但其在AML中的预后意义尚不明确。本研究旨在构建一个与中性粒细胞相关的基因特征用于AML风险分层,并探讨其临床和免疫学意义。
利用来自癌症基因组图谱(TCGA)、基因表达综合数据库(GEO)和俄勒冈健康与科学大学队列(n = 1537)的转录组和临床数据,我们通过文献挖掘确定了148个与中性粒细胞相关的基因。通过单变量Cox回归和LASSO回归(R包:survival,glmnet)选择预后基因。一个包含5个基因的模型(CSF3R、BRAF、FFAR2、CD300A、CD37)在内部队列(TCGA)和外部队列(GSE10358、GSE14468、俄勒冈健康与科学大学队列)中进行了验证。进行免疫谱分析、药物敏感性分析(GDSC数据库)和TIDE评分以评估免疫治疗的相关性。
基于中性粒细胞的模型将AML患者分为高风险和低风险组,两组的总生存期(OS)明显不同(TCGA中P<0.0001)。多变量Cox分析证实其独立于年龄、FLT3和TP53突变(HR = 2.14,P = 0.015)。CD37成为最强的预后标志物(5年AUC = 0.680,P = 0.0026),与免疫抑制性TME特征相关:髓源性抑制细胞增加(MDSCs,P<0.01)、调节性T细胞浸润(P<0.05)和免疫检查点上调(PD1、CTLA4、LAG3;P<0.001)。高CD37表达预测免疫治疗反应性(TIDE评分,P = 0.004),并与146种潜在治疗药物相互作用(例如,BCL2抑制剂)。
本研究提出了一种新的包含5个基因的预后模型,将中性粒细胞生物学纳入AML风险分层。CD37是免疫逃逸的关键调节因子,是一种用于预后和免疫治疗预测的双重生物标志物。虽然在多个队列中得到验证,但仍需要进行实验研究来阐明CD37的机制作用。我们的研究结果突出了以中性粒细胞为中心的生物标志物在指导AML个性化治疗方面的潜力。