摘要
目的本研究基于多模态临床数据构建AFP阴性肝细胞癌(ANHC)患者术后复发及生存风险的动态列线图模型,通过机器学习整合肿瘤生物学特征与治疗响应参数,揭示ANHC特异性预后标志物组合,为突破传统血清学标志物局限提供个性化风险评估工具。方法回顾性纳入2012年4月28日—2018年12月31日于东方肝胆外科医院接受肝切除术的421例ANHC患者,随机分为训练组(210例)和验证组(211例)。通过单因素及多因素Cox比例风险回归筛选独立预后因素,构建列线图模型,并采用受试者操作特征曲线(ROC曲线)、校准曲线和决策曲线分析(DCA)验证性能。检测指标包括前白蛋白(PA)、白细胞(WBC)、肿瘤大小、微血管侵犯等。计数资料两组间比较使用χ^(2)检验或Fisher精确检验,计量资料两组间比较使用成组t检验或Mann-Whitney U检验。结果通过多因素分析确定的无病生存期(DFS)的独立危险因素包括多发肿瘤(HR=3.30,P<0.001)、WBC(HR=1.05,P=0.005)、血糖(HR=1.15,P=0.026)、糖类抗原19-9(CA19-9)(HR=1.01,P=0.005)和肿瘤大小(HR=1.17,P<0.001),PA为保护因素(HR=0.99,P=0.022)。总生存期(OS)的独立危险因素包括肿瘤包膜不完整(HR=0.60,P=0.009)、年龄(HR=1.02,P=0.035)、凝血酶原时间(PT)(HR=1.27,P=0.023)、CA19-9(HR=1.01,P<0.001)和肿瘤大小(HR=1.15,P<0.001)。DFS列线图在训练组和验证组ROC曲线下面积(AUC)分别为0.74(95%CI:0.64~0.84)和0.67(95%CI:0.57~0.77),OS列线图的AUC分别为0.76(95%CI:0.64~0.88)和0.73(95%CI:0.60~0.87)。校准曲线和DCA表明模型具有良好的预测性能和临床实用性。结论肿瘤数目、PA、WBC、肿瘤大小等术前指标可有效预测ANHC患者术后复发,而肿瘤包膜完整性、年龄、PT等与OS显著相关。构建的列线图模型验证性能良好,可为个体化预后评估提供依据。
Objective To establish dynamic nomogram models for postoperative recurrence and survival risk of patients with AFP-negative hepatocellular carcinoma(ANHC)based on multimodal clinical data,to identify ANHC-specific prognostic biomarker combinations by integrating tumor biological characteristics and treatment response parameters through machine learning,and to provide an individualized risk assessment tool for overcoming the limitations of traditional serum biomarkers.Methods A retrospective analysis was performed for 421 ANHC patients who underwent hepatectomy in Eastern Hepatobiliary Surgery Hospital from April 2012 to December 2018,and they were randomly divided into training group with 210 patients and validation group with 211 patients.The univariate and multivariate Cox proportional-hazards regression models were used to identify independent prognostic factors and establish a nomogram model,and the receiver operating characteristic(ROC)curve,the calibration curve,and the decision curve analysis were used to assess the performance of the model.Related indicators were measured,including prealbumin(PA),white blood cell count(WBC),tumor size,and microvascular invasion.The chi-square test or the Fisher’s exact test was used for comparison of categorical variables between two groups,and the independent-samples t test or the Mann-Whitney U test was used for comparison of continuous variables between two groups.Results The multivariate analysis showed that multiple tumors(hazard ratio[HR]=3.30,P<0.001),WBC(HR=1.05,P=0.005),blood glucose(HR=1.15,P=0.026),CA19-9(HR=1.17,P=0.005),and tumor size(HR=1.17,P<0.001)were independent risk factors for disease-free survival(DFS),while PA(HR=0.99,P=0.022)was a protective factor.Incomplete tumor capsule(HR=0.60,P=0.009),age(HR=1.02,P=0.035),prothrombin time(PT)(HR=1.27,P=0.023),CA19-9(HR=1.01,P<0.001),and tumor size(HR=1.15,P<0.001)were independent risk factors for overall survival(OS).The DFS nomogram achieved an AUC of 0.74(95%confidence interval[CI]:0.64—0.84)in the training group and 0.67(95%CI:0.57—0.77)in the validation group,while the OS nomogram had an AUC of 0.76(95%CI:0.64—0.88)and 0.73(95%CI:0.60—0.87),respectively.The calibration curve and the decision curve analysis showed that the models had good predictive accuracy and clinical practicability.Conclusion Preoperative indicators,including tumor number,PA,WBC,and tumor size,can effectively predict postoperative recurrence in ANHC patients,while tumor capsule integrity,age,and PT are significantly associated with OS.The nomogram models established have good performance and can provide a basis for individualized prognostic assessment.
作者
李慧明
武烨晔
郭永庆
饶春美
刘军
汪玲
LI Huiming;WU Yeye;GUO Yongqing;RAO Chunmei;LIU Jun;WANG Ling(Department of Clinical Laboratory,The First Affiliated Hospital of Nanchang University,Nanchang 330006,China;Second Department of Hepatic Surgery,Eastern Hepatobiliary Surgery Hospital,Navy Medical University,Shanghai 200438,China;Department of Clinical Laboratory,Eastern Hepatobiliary Surgery Hospital,Navy Medical University,Shanghai 200438,China)
出处
《临床肝胆病杂志》
北大核心
2025年第8期1606-1614,共9页
Journal of Clinical Hepatology
基金
江西省教育厅重点项目(GJJ210135)
江西省卫生健康委员会(202310277)。
关键词
甲胎蛋白类
癌
肝细胞
列线图
预后
alpha-Fetoproteins
Carcinoma,Hepatocellular
Nomogram
Prognosis
