乙肝肝硬化患者肝细胞癌发生风险预测模型的构建  

Establishment of a model for predicting hepatocellular carcinoma occurence of patients with hepatitis B cirrhosis

作　　者：秦进 岳宁宁 王双凤[1] 吴芃诺 陈香宇[1] QIN Jin;YUE Ningning;WANG Shuangfeng;WU Pengnuo;CHEN Xiangyu(Digestive Endoscopy Center,the First Affiliated Hospital,Zhengzhou University,Zhengzhou 450052)

机构地区：[1]郑州大学第一附属医院消化内科镜中心,郑州450052

出　　处：《郑州大学学报（医学版）》2023年第2期217-223,共7页Journal of Zhengzhou University(Medical Sciences)

基　　金：河南省高等学校重点科研项目计划(20A320081)。

摘　　要：目的:构建乙肝肝硬化患者肝细胞癌(HCC)发生风险预测模型。方法:收集乙肝肝硬化确诊病例,按照2∶1随机分为训练集(344例)和验证集(172例)。收集病例信息,随访HCC发生情况。利用训练集数据,采用Cox回归建立乙肝肝硬化患者HCC发生风险预测模型,绘制列线图。分别利用两个数据集,采用ROC曲线评价预测效能;通过决策曲线分析(DCA)评估模型的临床效益;根据列线图评分的中位数将患者分为高风险组和低风险组,采用Kaplan-Meier法绘制两个风险组的生存曲线并进行Log-rank检验。结果:最终列线图纳入6个因素,分别为年龄、是否吸烟、是否抗病毒治疗、食管胃底静脉曲张程度、有无腹水、血清AFP水平。模型在训练集和验证集中预测乙肝肝硬化患者5 a内发生HCC的AUC(95%CI)分别为0.863(0.815~0.911)和0.873(0.812~0.934),一致性指数分别为0.850(0.820~0.876)和0.837(0.804~0.870)。DCA曲线表明模型具有较好的临床效益。Kaplan-Meier法生存分析结果显示,在训练集和验证集中,高风险组患者HCC发生风险均明显高于低风险组(P<0.001)。结论:成功建立了准确预测乙肝肝硬化患者5 a内HCC发生风险的模型。Aim:To develop a model for predicting hepatocellular carcinoma(HCC)occurence in patients with hepatitis B cirrhosis(HB-cirrhosis).Methods:A total of 516 patients with HB-cirrhosis were collected and randomly allocated into training set(n=344)and validation set(n=172)according to the ratio of 2∶1.Clinical information and follow-up of occurrence of HCC were recorded.Based on the training set data,Cox regression was used to establish a prediction model for the occurrence of HCC of HB-cirrhosis patients,and the nomogram was drawn.ROC curve was adopted to evaluate the prediction efficiency using the two data sets,respectively,and clinical benefits were evaluated through decision curve analysis(DCA)of the model.The patients were divided into high-risk and low-risk groups according to the median of nomogram scores,and survival curves for the two risk groups were plotted using the Kaplan-Meier method and subjected to the Log-rank test.Results:A prediction model was established based on the Cox regression,which including 6 factors(age,smoking,antiviral treatment,esophageal and gastric varices,ascites and AFP).AUC(95%CI)of the nomogram to predict the 5-year risk of HCC in training set and validation set were 0.863(0.815-0.911)and 0.873(0.812-0.934),C-index(95%CI)were 0.850(0.820-0.876)and 0.837(0.804-0.870),respectively.DCA demonstrated that the nomogram was clinically useful.Kaplan-Meier survival analysis results showed that the risk of HCC was significantly higher in the high-risk group than in the low-risk group in training set and validation set(P<0.001).Conclusion:The nomogram could accurately predict the 5-year risk of HCC in patients with HB-cirrhosis.

关 键 词：乙肝肝硬化 预测模型 肝细胞癌 列线图 

分 类 号：R735.7[医药卫生—肿瘤]