中晚期肝癌靶向免疫治疗病人预后风险预测模型的构建与验证OA
Construction and validation of a prognostic risk prediction model for patients with advanced liver cancer receiving targeted-immunotherapy
目的:构建并验证中晚期肝癌病人靶向免疫治疗预后的列线图预测模型,为临床制订个体化治疗决策提供可靠依据.方法:选取2022年5月—2025年1月江西省吉安市中心人民医院肿瘤科收治的150例接受靶向免疫联合治疗的中晚期肝癌病人为研究对象,通过单因素分析初步筛选预后相关因素,进一步采用Logistic回归分析确定独立预测因子,并构建可视化列线图预测模型.采用受试者工作特征(ROC)曲线、校准曲线、Bootstrap内部验证(1000次重复抽样)及决策曲线分析(DCA)等方法系统评估模型性能.结果:中晚期肝癌病人靶向免疫治疗后6个月内预后不良发生率为20.67％(31/150).巴塞罗那临床肝癌分期(BCLC)C期[OR=4.171,95％CI(1.918,9.072)]、门静脉侵犯[OR=3.337,95％CI(1.609,6.923)]、中性粒细胞与淋巴细胞比值(NLR)≥3[OR=4.889,95％CI(2.059,11.607)]、甲胎蛋白(AFP)≥400 ng/mL[OR=3.228,95％CI(1.517,6.869)]、白蛋白(ALB)＜35 g/L[OR=2.835,95％CI(1.343,5.984)]及治疗中断[OR=4.463,95％CI(1.895,10.509)]为独立危险因素(P＜0.05).模型验证结果显示,区分度良好,曲线下面积(AUC)为 0.823[95％CI(0.774,0.872)];校准度良好(Hosmer-Lemeshow检验x2=6.854,P=0.335;平均绝对偏差=0.025);内部验证校正C统计量0.812[95％CI(0.758,0.866)];决策曲线分析显示在20％～80％阈值概率内具有明显临床净获益,最高达41.25％.结论:中晚期肝癌病人靶向免疫治疗预后的列线图预测模型整合临床特征、实验室指标和治疗因素,具有良好的判别能力、校准精度和临床实用性,可为中晚期肝癌病人靶向免疫治疗的预后评估和个体化干预提供客观、量化的决策支持.
Objective:To construct and validate a Nomogram prediction model for the prognosis of targeted-immunotherapy in patients with advanced liver cancer,so as to provide a reliable basis for the formulation of individualized treatment decisions in clinical practice.Methods:A total of 150 patients with advanced liver cancer who received targeted-immunotherapy combined therapy and were admitted to the department of oncology,Ji'an Central People's Hospital,Jiangxi province from May 2022 to January 2025 were selected as the research objects.Prognostic factors were preliminarily screened through univariate analysis,and independent predicators were further determined by Logistic regression analysis,and a visual Nomogram prediction model was constructed.The model performance was systematically evaluated by using methods such as receiver operating characteristic(ROC)curve,calibration curve,Bootstrap internal validation(1 000 repeated samples),and decision curve analysis(DCA).Results:The incidence of poor prognosis within 6 months after targeted-immunotherapy in patients with advanced liver cancer was 20.67％(31/150).Barcelona Clinic Liver Cancer(BCLC)C[OR=4.171,95％CI(1.918,9.072)],portal vein invasion[OR=3.337,95％CI(1.609,6.923)],neutrophil-to-lymphocyte ratio(NLR)≥3[OR=4.889,95％CI(2.059,11.607)],alpha-fetoprotein(AFP)≥400 ng/mL[OR=3.228,95％CI(1.517,6.869)],albumin(ALB)＜35 g/L[OR=2.835,95％CI(1.343,5.984)]and treatment interruption[OR=4.463,95％CI(1.895,10.509)]were independent risk factors(P＜0.05).The model validation results showed that the discrimination was good,and the area under the curve(AUC)was 0.823[95％CI(0.774,0.872)].Good calibration(Hosmer-Lemeshow test x2=6.854,P=0.335,mean absolute deviation=0.025).Internal validation adjusted C-statistic 0.812[95％CI(0.758,0.866)].Decision curve analysis showed that there was a significant clinical net benefit within the threshold probability range of 20％to 80％,with the highest reaching 41.25％.Conclusions:The Nomogram prediction model for the prognosis of targeted-immunotherapy in patients with advanced liver cancer integrates clinical features,laboratory indicators and therapeutic factors.It has good discriminative ability,calibration accuracy and clinical practicability,and can provide objective and quantitative decision support for the prognosis assessment and individualized intervention of targeted-immunotherapy in patients with advanced liver cancer.
高佳丽;张晓鑫;刘艳
343000,江西省吉安市中心人民医院肿瘤科343000,江西省吉安市中心人民医院肿瘤科343000,江西省吉安市中心人民医院肿瘤科
肝细胞癌靶向治疗免疫治疗预后预测列线图
hepatocellular carcinomatargeted therapyimmunotherapyprognostic predictionNomogram
《全科护理》 2026 (10)
1824-1830,7
评论