基于肿瘤浸润免疫细胞的甲状腺癌患者甲状腺半切术后预后模型的构建及初步评价  

Construction and preliminary evaluation of a prognostic model for thyroid cancer patients after hemithyroidectomy based on tumor-infiltrating immune cells

作　　者：杨小丽 张勇 陈敏 YANG Xiaoli;ZHANG Yong;CHEN Min(The First Ward of General Surgery,the First People's Hospital of Shuangliu District of Chengdu,Sichuan University,West China Airport Hospital,Chengdu 610200,China)

机构地区：[1]成都市双流区第一人民医院,四川大学华西空港医院普外科一病区,成都610200

出　　处：《中国免疫学杂志》2025年第4期925-930,共6页Chinese Journal of Immunology

摘　　要：目的:探讨构建及初步评价甲状腺癌患者甲状腺半切术后肿瘤浸润免疫细胞的预后模型。方法:从癌症基因库(TCGA)中选取2020年11月至2022年11月的152例甲状腺癌甲状腺半切术后患者的基因表达谱与随访参数;采用ssGSEA量化肿瘤组织中免疫细胞浸润的情况;采用LASSO筛选重要预测因子以验证预后与不同浸润免疫细胞之间的关系;以合适的细胞构建预后风险评分模型,并将甲状腺癌患者分为高、低风险组,使用Kaplan-Meier生存曲线进行验证;以风险模型为基础构建列线图模型以预测甲状腺癌患者的存活率及治疗无效率;使用受试者工作特征(ROC)曲线以及AUC验证列线图准确性;采用校准曲线比较列线图模型的预测结果与观察结果;使用临床决策曲线鉴别预测模型的可靠性。结果:采用ssGSEA量化免疫细胞浸润数据得到22个浸润免疫细胞,包括活化肥大细胞、幼稚B细胞、未活化肥大细胞、浆细胞、CD8^(+)T细胞、γδT细胞、幼稚CD4^(+)T细胞、活化记忆CD4^(+)T细胞、Treg细胞、滤泡辅助性T细胞、记忆B细胞、活化NK细胞、未活化NK细胞、未活化记忆CD4^(+)T细胞、M0巨噬细胞、单核细胞、M1巨噬细胞、嗜中性粒细胞、未活化树突状细胞、活化树突状细胞、M2巨噬细胞和嗜酸性粒细胞;选取的预测因子为6个免疫细胞,分别是M0巨噬细胞、M2巨噬细胞、活化树突状细胞、未活化肥大细胞、CD8^(+)T细胞、单核细胞;根据这6个浸润的免疫细胞,计算样本的风险评分,风险评分与高风险组甲状腺癌患者和预后不良人数成正比;采用Kaplan-Meier制作生存曲线,结果为低风险组总生存时间明显长于高风险组(Log-Rankχ^(2)=4.524,P=0.024);预后列线图模型总分425分,死亡风险79.12%;治疗效果列线图模型总分406分,对应治疗无效的风险为77.97%,经过模型验证结果较可靠。结论:对肿瘤浸润的免疫细胞在甲状腺癌患者中的评分�Objective:To explore the prognostic model for construction and preliminary evaluation of tumor invasion of immune cells after hemithyroidectomy in patients with thyroid cancer.Methods:The gene expression profiles and follow-up parameters of 152 patients with thyroid cancer after hemithyroidectomy from November 2020 to November 2022 were selected from The Cancer Genome Atlas(TCGA);ssGSEA was used to quantify immune cell infiltration in tumor tissues.LASSO was used to screen for key predictors to verify the relationship between prognosis and different infiltrating immune cells.A prognostic risk score model was constructed using appropriate immune cells,and thyroid cancer patients were stratified into high-and low-risk groups.Kaplan-Meier survival curves were used for validation.Nomogram models were developed based on the risk model to predict survival rates and treatment inefficiency in thyroid cancer patients.Model accuracy was verified using ROC curve and AUC value.Calibration curves were employed to compare predicted and observed outcomes,and clinical decision curves were used to assess model reliability.Results:Immune cell infiltration data were quantified by ssGSEA to obtain 22 infiltrating immune cells,including activated mast cells,naive B cells,unactivated mast cells,plasma cells,CD8^(+)T cells,γδT cells,naive CD4^(+)T cells,activated memory CD4^(+)T cells,Treg cells,follicular helper T cells,memory B cells,activated NK cells,unactivated NK cells,unactivated memory CD4^(+)T cells,M0 macrophages,monocytes,M1 macrophages,neutrophils,unactivated dendritic cells,activated dendritic cells,M2 macrophages and eosinophils.Six immune cells were selected as predictors,M0 macrophages,M2 macrophages,activated dendritic cells,unactivated mast cells,CD8^(+)T cells,and monocytes.Risk scores calculated from these six immune cells were positively correlated with poor prognosis in high-risk patients.Kaplan-Meier analysis showed significantly longer overall survival in the low-risk group compared to the high-risk group(Log-

关 键 词：免疫细胞浸润 预后模型 甲状腺癌 治疗与预后 

分 类 号：R736.1[医药卫生—肿瘤]