基于EPSTI1表达的临床病理特征列线图预测肾透明细胞癌预后

Construction of prognostic nomogram based on clinicopathological characteristics and epithelial-stromal interaction 1 expression for clear cell renal cell carcinoma

目的:构建基于上皮间质相互作用蛋白1(epithelial-stromal interaction protein 1,EPSTI1)预后列线图预测肾透明细胞癌的预后。方法:回顾性分析2012年1月至2015年12月于福建医科大学附属第一医院221例接受手术治疗的肾透明细胞癌患者和TCGA数据库中533例肾透明细胞癌患者数据,对癌旁正常组织和癌组织标本进行免疫组织化学(immunohistochemistry,IHC)染色,分析EPSTI1的表达差异及与临床病理特征的相关性。对EPSTI1高表达与低表达患者的总生存期(overall survival,OS)和无病生存期(disease-free survival,DFS)进行Kaplan-Meier生存分析,采用单因素和多因素Cox比例风险模型分析OS的预后因素,进一步构建列线图模型并验证。结果:与癌旁正常肾组织比较,肾透明细胞癌组织中EPSTI1的IHC评分和m RNA表达水平均显著高于正常组织(均P<0.001),且在高T分期的癌组织中表达更高(P=0.036,P=0.006);EPSTI1蛋白表达与肿瘤最大径、TNM分期相关(P=0.002,P=0.032);EPSTI1低表达组OS、DFS均优于高表达组(P=0.046,P=0.003,P=0.001);单因素和多因素Cox回归分析结果显示,EPSTI1蛋白高表达、WHO/ISUP分级、AJCC/TNM分期是影响肾透明细胞癌患者预后不良的独立危险因素(P=0.009,P=0.039,P<0.001);基于上述变量构建的预后列线图模型对患者5年OS预测能力优于AJCC/TNM分期,校准曲线显示模型预测值与实际值间具有良好的一致性。结论:基于EPSTI1、AJCC/TNM分期和WHO/ISUP分级建立的列线图模型对肾透明细胞癌预后具有较强的预测能力。

Objective:To construct a prognostic nomogram based on epithelial-stromal interaction protein 1(EPSTI1)and predict the prognosis of clear cell renal cell carcinoma(ccRCC).Methods:A retrospective analysis was performed from January 2012 to December 2015 at The First Affiliated Hospital of Fujian Medical University,on 221 patients with ccRCC who underwent surgical treatment in our center and 533 patients with ccRCC in The Cancer Genome Atlas(TCGA)database.Immunohistochemical(IHC)staining was performed on adjacent normal and cancerous tissues to analyze the expression level of EPSTI1 and its correlation with clinicopathological characteristics.Kaplan-Meier survival analysis was performed for the overall survival(OS)and disease-free survival(DFS)of patients with high and low EPSTI1 expression levels.Univariate and multivariate Cox proportional hazards models were used to analyze the prognostic factors for OS,and a nomogram model was constructed and verified.Results:The IHC scores and mRNA expression levels of EPSTI1 were significantly higher in ccRCC tissues than in normal tissues(all P<0.001).EPSTI1 was expressed at higher levels in cancer tissues at higher T stages(P=0.036,P=0.006).The EPSTI1 protein expression level was related to the maximum tumor diameter and TNM stage(P=0.002,P=0.032,respectively).The OS and DFS were higher in the low-EPSTI1-expression group than the high-EPSTI1-expression group(P=0.046,P=0.003,P=0.001).Univariate and multivariate Cox regression analyses showed that a high EPSTI1 protein expression level,WHO/ISUP grade,and AJCC/TNM stage were independent risk factors for poor prognosis(P=0.009,P=0.039,P<0.001).The prognostic nomogram model constructed based on the above variables was superior to the AJCC/TNM stage in predicting the 5-year OS,and the calibration curve showed that the predicted value of the model was con sistent with the actual value.Conclusions:The nomographic model based on EPSTI1,AJCC/TNM staging and WHO/ISUP staging has a strong predictive ability for the prognosis of renal clear cell carcinoma.