基于淋巴细胞比值相关免疫-炎症反应评分的网络计算器预测肝细胞癌患者总生存率  

A network calculator based on lymphocyte ratio-associated immune-inflammatory response score predicts survival in patients with hepatocellular carcinoma

作　　者：施理敏 张明星 蔡思齐 杨悦仪 杨春忆 张雯琼 Shi Limin;Zhang Mingxing;Cai Siqi;Yang Yueyi;Yang Chunyi;Zhang Wenqiong(Department of Laboratory Medicine,Shanghai Public Health Clinical Center,Shanghai 201508,China)

机构地区：[1]上海市公共卫生临床中心检验医学科,上海201508

出　　处：《肝癌电子杂志》2024年第4期19-26,共8页Electronic Journal of Liver Tumor

基　　金：国家自然科学基金面上项目(82072281)。

摘　　要：目的:构建淋巴细胞比值相关免疫-炎症反应评分(immune-inflammatory response score,IRS),并构建预测肝细胞癌(hepatocellular carcinoma,HCC)患者预后风险的网络计算器。方法:回顾性分析2018年1月至2023年12月于上海市公共卫生临床中心进行手术治疗的HCC患者。收集中性粒细胞与淋巴细胞比值(neutrophil-lymphocyte ratio,NLR)、血小板与淋巴细胞比值(platelet-lymphocyte ratio,PLR)、单核细胞与淋巴细胞比值(monocyte to lymphocyte ratio,MLR)和C反应蛋白与淋巴细胞比值(C-reactive protein to lymphocyte ratio,CLR)。使用多因素Cox回归模型构建IRS。采用单因素和多因素Cox回归分析筛选HCC患者死亡风险因素。使用rms、foreign、readxl、Hmisc和rmda等R包构建和评估列线图预测的准确性。使用Dyn Nom包开发HCC患者死亡风险网络计算器。结果:154例患者随访1~60个月,中位随访时间为17个月,其中1年、3年和5年死亡率分别为30.5%、50.6%和61.0%。根据多因素Cox回归模型系数β构建IRS=0.213×NLR+0.005×PLR+0.042×CLR。死亡HCC患者IRS高于生存HCC患者,且差异有统计学意义(P<0.05)。高风险组(IRS≥2.88分)HCC患者总生存率低于低风险组(IRS<2.88分)HCC患者,且差异有统计学意义(P<0.05)。时间-受试者操作特征(receiver operating characteristic,ROC)曲线显示,IRS预测HCC患者1年、3年、5年总生存率的曲线下面积(area under the curve,AUC)分别为0.829[95%置信区间(95%confidence interval,95%CI)为0.760~0.885]、0.901(95%CI为0.842~0.943)和0.898(95%CI为0.839~0.941)。多因素Cox回归分析结果显示,肿瘤最大直径、γ-谷氨酰转移酶(γ-glutamyl transferase,GGT)、碱性磷酸酶(alkaline phosphatase,ALP)、甲胎蛋白(alpha-fetoprotein,AFP)水平和IRS是HCC患者死亡独立风险因素(P均<0.05)。基于IRS构建预测HCC患者1年、3年及5年死亡风险的列线图。校准曲线显示,列线图预测1年、3年和5年死亡风险的C-index分别为0.836(95%CI为0.818~0.912Objective:To develop a lymphocyte ratio-associated immune-inflammatory response score(IRS)and construct a network calculator to predict prognostic risk in patients with hepatocellular carcinoma(HCC).Method:HCC patients who were surgically treated in Shanghai Public Health Clinical Center between January 2018 and December 2023 were selected.Neutrophil to lymphocyte ratio(NLR),platelet to lymphocyte ratio(PLR),monocyte to lymphocyte ratio(MLR)and C-reactive protein to lymphocyte ratio(CLR)were collected.IRS was constructed using multifactorial Cox risk regression.Univariate and multifactorial Cox regression models were used to analyse risk factors for mortality in HCC patients.Constructed and assessed nomogram prediction accuracy using R packages such as rms,foreign,readxl,Hmisc and rmda.Development of a mortality risk network calculator for HCC patients using the DynNom package.Result:154 patients were followed up from 1 month to 60 months,with a median follow-up of 17 months,in which the 1-,3-and 5-year mortality rates were 30.5%,50.6%and 61.0%,respectively.IRS=0.213×NLR+0.005×PLR+0.042×CLR was constructed based on the multifactorial Cox risk regression coefficientβ.IRS was significantly higher in patients who died than in those who survived(P<0.05).Patients in the high-risk group(IRS≥2.88)had significantly lower survival than patients in the low-risk group(IRS<2.88)(P<0.05).Time-receiver operating characteristic(ROC)curves showed areas under the curve(AUCs)of 0.829(95%confidence interval[95%CI]:0.760-0.885),0.901(95%CI:0.842-0.943),and 0.898(95%CI:0.839-0.941)for 1-year,3-year,and 5-year risk of death among patients with IRS prediction,respectively.The results of multifactorial Cox risk regression analysis showed that maximum tumour diameter,γ-glutamyl transferase(GGT),alkaline phosphatase(ALP),alpha-fetoprotein(AFP)and IRS were independent factors for the risk of death in HCC patients(P<0.05).Nomogram were constructed based on IRS to predict the risk of death in HCC patients at 1,3 and 5 years.The calibr

关 键 词：肝细胞癌 淋巴细胞 炎症 免疫 预后 网络计算器 

分 类 号：R73[医药卫生—肿瘤]