Ge Mingqin, Niu Jie, Hu Ping, Tong Aihua, Dai Yan, Xu Fangjiang, Li Fuyuan
Department of Endocrinology, Linyi Central Hospital, Linyi, China.
Front Med (Lausanne). 2021 Apr 13;8:637743. doi: 10.3389/fmed.2021.637743. eCollection 2021.
This study aimed to construct a prognostic ferroptosis-related signature for thyroid cancer and probe into the association with tumor immune microenvironment. Based on the expression profiles of ferroptosis-related genes, a LASSO cox regression model was established for thyroid cancer. Kaplan-Meier survival analysis was presented between high and low risk groups. The predictive performance was assessed by ROC. The predictive independency was validated multivariate cox regression analysis and stratified analysis. A nomogram was established and verified by calibration curves. The enriched signaling pathways were predicted GSEA. The association between the signature and immune cell infiltration was analyzed by CIBERSORT. The ferroptosis-related genes were validated in thyroid cancer tissues by immunohistochemistry and RT-qPCR. A ferroptosis-related eight gene model was established for predicting the prognosis of thyroid cancer. Patients with high risk score indicated a poorer prognosis than those with low risk score ( = 1.186e-03). The AUCs for 1-, 2-, and 3-year survival were 0.887, 0.890, and 0.840, respectively. Following adjusting other prognostic factors, the model could independently predict the prognosis ( = 0.015, HR: 1.870, 95%CI: 1.132-3.090). A nomogram combining the signature and age was constructed. The nomogram-predicted probability of 1-, 3-, and 5-year survival approached the actual survival time. Several ferroptosis-related pathways were enriched in the high-risk group. The signature was distinctly associated with the immune cell infiltration. After validation, the eight genes were abnormally expressed between thyroid cancer and control tissues. Our findings established a prognostic ferroptosis-related signature that was associated with the immune microenvironment for thyroid cancer.
本研究旨在构建甲状腺癌的预后铁死亡相关特征，并探究其与肿瘤免疫微环境的关联。基于铁死亡相关基因的表达谱，建立了甲状腺癌的LASSO Cox回归模型。对高风险组和低风险组进行了Kaplan-Meier生存分析。通过ROC评估预测性能。通过多变量Cox回归分析和分层分析验证预测独立性。建立了列线图并通过校准曲线进行验证。通过GSEA预测富集的信号通路。通过CIBERSORT分析特征与免疫细胞浸润之间的关联。通过免疫组织化学和RT-qPCR在甲状腺癌组织中验证铁死亡相关基因。建立了一个用于预测甲状腺癌预后的铁死亡相关八基因模型。高风险评分的患者预后比低风险评分的患者差（P = 1.186e-03）。1年、2年和3年生存率的AUC分别为0.887、0.890和0.840。在调整其他预后因素后，该模型可以独立预测预后（P = 0.015，HR：1.870，95%CI：1.132-3.090）。构建了一个结合特征和年龄的列线图。列线图预测的1年、3年和5年生存率接近实际生存时间。高风险组中富集了几条铁死亡相关通路。该特征与免疫细胞浸润明显相关。验证后，这八个基因在甲状腺癌组织和对照组织之间异常表达。我们的研究结果建立了一个与甲状腺癌免疫微环境相关的预后铁死亡相关特征。
