目的:口腔鳞状细胞癌(OSCC)是口腔颌面部最常见的恶性肿瘤。乳酸在肿瘤微环境(TME)中的积累因其作为癌细胞的能量来源和对肿瘤进展至关重要的信号传导途径的激活剂的双重作用而受到关注。本研究旨在揭示乳酸相关基因(LRGs)对预后的影响,TME,和OSCC的免疫特性,最终目标是开发一种新的预后模型。
方法:对来自癌症基因组图谱数据库的OSCC患者的LRGs进行无监督聚类分析,以评估和比较TME,免疫功能,以及各种乳酸亚型的临床特征。通过应用Cox和最小绝对收缩和选择算子(LASSO)回归技术,开发了完善的预后模型。然后利用外部验证集来提高模型准确性,以及详细的药物敏感性相关分析。
结果:根据LRGs将癌症基因组Atlas-OSCC患者分为4种不同的乳酸亚型。值得注意的是,亚型1和亚型2的患者表现出最低和最有利的预后,分别。亚型1患者显示免疫检查点基因的表达水平升高。进一步分析确定了1086个在癌症和非癌症组织之间具有显着表达差异的基因。以及亚型1和亚型2患者之间。预后模型的选择基因包括ZNF662,CGNL1,VWCE,ZFP42该模型定义的高风险组的预后明显较差(P<.0001),并作为独立的预后因素(P<.001)。准确预测1-,3-,5年生存率。此外,高危人群对AZ6102和维奈托克等化疗药物的敏感性增强.
结论:基于基因ZNF662,CGNL1,VWCE,ZFP42可以作为可靠的生物标志物,为OSCC患者提供准确的预后预测和药物干预的潜在机会。
OBJECTIVE: Oral squamous cell carcinoma (OSCC) is the most common malignant tumour in the oral and maxillofacial region. Lactic acid accumulation in the tumour microenvironment (TME) has gained attention for its dual role as an energy source for cancer cells and an activator of signalling pathways crucial to tumour progression. This study aims to reveal the impact of lactate-related genes (LRGs) on the prognosis, TME, and immune characteristics of OSCC, with the ultimate goal of developing a novel prognostic model.
METHODS: Unsupervised clustering analysis of LRGs in OSCC patients from The Cancer Genome Atlas database was conducted to evaluate and compare TME, immune features, and clinical characteristics across various lactate subtypes. A refined prognostic model was developed through the application of Cox and Least absolute shrinkage and selection operator (LASSO) regression techniques. External validation sets were then utilised to improve model accuracy, along with a detailed correlation analysis of drug sensitivity.
RESULTS: The Cancer Genome Atlas-OSCC patients were categorised into 4 distinct lactate subtypes based on LRGs. Notably, patients in subtype 1 and subtype 2 exhibited the least and most favourable prognoses, respectively. Subtype 1 patients showed elevated expression levels of immune checkpoint genes. Further analysis identified 1086 genes with significant expression differences between cancer and noncancer tissues, as well as between subtype 1 and subtype 2 patients. Selected genes for the prognostic model included ZNF662, CGNL1, VWCE, and ZFP42. The high-risk group defined by this model had a significantly poorer prognosis (P < .0001) and functioned as an independent prognostic factor (P < .001), accurately predicting 1-, 3-, and 5-year survival rates. Additionally, individuals in the high-risk category exhibited heightened sensitivity to chemotherapy drugs such as AZ6102 and Venetoclax.
CONCLUSIONS: The predictive model based on the genes ZNF662, CGNL1, VWCE, and ZFP42 can serve as a reliable biomarker, providing accurate prognostic predictions for OSCC patients and potential opportunities for pharmaceutical interventions.