术前CT定量参数与肝动脉化疗栓塞术后肝癌患者预后的相关性及预后预测列线图模型的构建

Correlation between preoperative CT quantitative parameters and prognosis of patients with liver cancer treated with transcatheter arterial chemoembolization and construction of nomogram model for prognostic prediction

目的探讨术前CT定量参数与肝动脉化疗栓塞术(TACE)后肝癌患者预后的相关性,并构建预后预测列线图模型。方法选择2020年5月至2022年5月于濮阳市第五人民医院医学影像科行TACE治疗的120例肝癌患者为研究对象。术前均完善螺旋CT平扫及灌注成像扫描,测定肝血流量(HBF)、肝动脉灌注量(HAP)、门静脉灌注量(PVP)、肝动脉灌注指数(HAI)、平均通过时间(MTT)等CT定量参数。术后随访至2023年5月,根据临床转归将患者分为预后良好组(n=90)和预后不良组(n=30),比较2组患者的性别、年龄、体质量指数(BMI)、吸烟酗酒史、TNM分期、肿瘤最大径、肿瘤数目、包膜是否完整、白蛋白、血小板、总胆红素、白细胞计数、丙氨酸氨基转移酶(ALT)、天冬氨酸氨基转移酶(AST)、甲胎蛋白(AFP)及HBF、HAP、HAI、MTT;采用logistic多因素回归分析影响经TACE治疗肝癌患者预后的危险因素,应用R软件基于筛选出的CT定量参数构建经TACE治疗肝癌患者预后预测列线图模型,应用rms程序包和Bootstrap内部验证法验证该模型预测经TACE治疗肝癌患者预后不良的效能。结果预后不良组患者TNM分期Ⅲ~Ⅳ期占比、包膜不完整占比及AFP、HBF、HAP、HAI水平显著高于预后良好组,MTT显著短于预后良好组(P<0.05);预后不良组与预后良好组患者的性别构成、年龄、BMI、吸烟史比例、肿瘤最大径>5 cm占比、肿瘤数目>1个占比、白蛋白水平、血小板计数、总胆红素水平、ALT水平、AST水平,PVP比较差异无统计学意义(P>0.05)。Logistic多因素回归分析结果显示,包膜不完整[比值比(OR)=2.202,95%置信区间(CI):1.884~2.880,P=0.000]、AFP水平高(OR=2.187,95%CI:1.818~2.787,P=0.000)、HBF高(OR=2.177,95%CI:1.859~2.862,P=0.000)、HAP高(OR=2.192,95%CI:1.907~2.909,P=0.000)、HAI高(OR=2.212,95%CI:1.926~2.816,P=0.000)、MTT短(OR=2.208,95%CI:1.927~2.954,P=0.000)是影响经TACE治疗肝癌患者预后的危险因素。基于HBF、HAP、HAI、MTT构建经TACE治疗肝癌患者预后预测列线图模型,该模式实测值与预测值结果基本一致,具有良好的预测性能;C-index为0.908(95%CI:0.787~0.968),该模型区分度及准确度较好。结论HBF、HAP、HAI、MTT是影响TACE后肝癌患者预后的危险因素,基于HBF、HAP、HAI和MTT构建的列线图模型对经TACE治疗肝癌患者预后预测效能较高。

Objective To investigate the correlation between preoperative CT quantitative parameters and the prognosis of patients with liver cancer treated by transcatheter arterial chemoembolization(TACE),and to construct a nomogram model for prognostic prediction.Methods A total of 120 patients with liver cancer who received TACE treatment in the Department of Medical Imaging,the Fifth People′s Hospital of Puyang City from May 2020 to May 2022 were selected as the research subjects.Spiral CT plain scan and perfusion imaging were performed before surgery,and CT quantitative parameters such as hepatic blood flow(HBF),hepatic artery perfusion(HAP)volume,portal vein perfusion(PVP)volume,hepatic arterial infusion(HAI)index and mean transit time(MTT)were measured.These patients were followed up to May 2023,and according to clinical outcomes,they were divided into the good prognosis group(n=90)and the poor prognosis group(n=30).Gender,age,body mass index(BMI),history of smoking and alcoholism,TNM stage,maximum tumor size,tumor number,complete or incomplete envelope,albumin,platelets,total bilirubin,white blood cell count,alanine aminotransferase(ALT),aspartate aminotransferase(AST),alpha-fetoprotein(AFP),HBF,HAP,HAI,and MTT were compared between the two groups.The risk factors affecting the prognosis of liver cancer patients treated with TACE were analyzed by logistic multivariate regression.CT quantitative parameters were screened out by R software to construct a nomogram model for predicting the prognosis of liver cancer patients treated by TACE.The efficiency of this model in predicting the poor prognosis of liver cancer patients treated by TACE was verified by the rms package and Bootstrap internal validation method.Results The proportion of TNM stagesⅢ-Ⅳ,the proportion of incomplete envelope and the levels of AFP,HBF,HAP and HAI in the poor prognosis group were significantly higher than those in the good prognosis group,while MTT in the poor prognosis group was significantly shorter than that in the good prognosis group(P<0.05).There was no significant difference in gender composition,age,BMI,smoking ratio,ratio of maximum tumor diameter>5 cm,ratio of tumor number>1,albumin level,platelet count,total bilirubin level,ALT level,AST level and PVP between the poor prognosis group and the good prognosis group(P>0.05).Logistic multivariate regression analysis showed that the incomplete envelope[odds ratio(OR)=2.202,95%confidence interval(CI):1.884-2.880,P=0.000],high AFP level(OR=2.187,95%CI:1.818-2.787,P=0.000),high HBF(OR=2.177,95%CI:1.859-2.862,P=0.000),high HAP(OR=2.192,95%CI:1.907-2.909,P=0.000),high HAI(OR=2.212,95%CI:1.926-2.816,P=0.000),and short MTT(OR=2.208,95%CI:1.927-2.954,P=0.000)were the risk factors affecting the prognosis of patients with liver cancer treated by TACE.The measured results were basically consistent with the predicted results based on the nomogram model constructed based on HBF,HAP,HAI and MTT,indicating that the model had good predictive performance.The C-index was 0.908(95%CI:0.787-0.968),indicating that the model had good differentiation and accuracy.Conclusion HBF,HAP,HAI and MTT are risk factors that affect the prognosis of patients with liver cancer treated by TACE.The nomogram model constructed based on the above four CT quantitative parameters has high predictive efficacy.