The tumor microenvironment (TME) is of great clinical significance for predicting the therapeutic effect of tumors. Nonetheless, there was no systematic analysis of cellular interactions in the TME of head and neck cancer (HNSC). This study used gene expression data from 816 patients with HNSC to analyze the scores of 22 immune cells. On this basis, we have established a novel TMEscore-based prognostic risk model. The relationship between TMEscore and clinical and genomic characteristics was analyzed. The sample was divided into risk-H and risk-L groups based on the prognosis risk model of TMEscore, with significant differences in overall survival between the two groups (log rank p < 0.001). In terms of clinical features, the TMEscore is closely related to the T staging, Grade, and HPV. As for genomic characteristics, the genomic features of the Risk-H samples are a low expression of immune-related genes and high-frequency mutations of TP53 and CEP152. This model was validated in an external test set, in which the prognosis for Risk-H group and Risk-L group was also significantly different (log rank p = 0.017). A quantitative method of TME infiltration pattern is established, which may be a potential predictor of HNSC prognosis.

肿瘤微环境（TME）对于预测肿瘤的治疗作用具有重要的临床意义。尽管如此，尚无系统分析头颈癌（HNSC）TME中的细胞相互作用。这项研究使用了来自816例HNSC患者的基因表达数据来分析22个免疫细胞的评分。在此基础上，我们建立了基于TMEscore的新型预后风险模型。分析了TMEscore与临床和基因组特征之间的关系。根据TMEscore的预后风险模型，将样本分为高危H组和高危L组，两组的总体生存率差异显着（对数秩p <0.001）。就临床特征而言，TMEscore与T分期，等级和HPV密切相关。至于基因组特征，Risk-H样品的基因组特征是免疫相关基因的低表达以及TP53和CEP152的高频突变。在外部测试集中验证了该模型，其中Risk-H组和Risk-L组的预后也显着不同（对数秩p  = 0.017）。建立了定量的TME浸润模式方法，可能是HNSC预后的潜在预测指标。