基于钆塞酸二钠增强MRI列线图预测MVI阴性肝细胞癌术后复发

A nomogram based on Gadoxetic Acid-enhanced MRI for prediction of postoperative recurrence of microvascular invasion-negative hepatocellular carcinoma

目的:探讨基于钆塞酸二钠(Gd-EOB-DTPA)增强MRI列线图模型预测微血管浸润(MVI)阴性肝细胞癌(HCC)患者根治性切除术后无复发生存(RFS)的价值。方法:回顾性搜集2015年2月-2021年5月在我院接受根治性肝切除术且经病理证实为MVI阴性的HCC患者125例,分析患者的术前临床资料、MRI资料及术后病理资料。通过单因素和多因素Cox比例风险模型分析获得患者术后RFS的独立危险因素,并构建预测患者1、3和5年RFS率的列线图模型,采用C指数和受试者工作特征(ROC)曲线评价模型的区分度,采用校准曲线评价模型的一致性。结果:随访过程中,37例HCC患者复发,中位RFS时间为27.1个月[95%置信区间(CI):43.3~54.1个月],术后1、3和5年RFS率分别为88.0%、71.9%、45.9%。多因素分析结果显示病理分级[风险比(HR)=4.807,95%CI:1.126~20.526,P=0.034]、肝胆期瘤周低信号(HR=3.755,95%CI:1.736~8.122,P=0.001)、肝胆期病灶-肝实质信号强度比(RIR)(HR=0.019,95%CI:0.001~0.396,P=0.010)是HCC患者术后RFS的独立危险因素。整合上述3个变量构建的列线图预测模型的曲线下面积(AUC)为0.819(95%CI:0.740~0.882),同时经Bootstrap内部验证,该模型的C指数为0.795(95%CI:0.723~0.867),且校准曲线验证结果显示该模型预测概率与实际观察概率吻合度较好。结论:基于Gd-EOB-DTPA增强MRI构建的列线图可以有效预测MVI阴性HCC患者的术后RFS。

Objective:To investigate the value of a nomogram based on Gadoxetic Acid(Gd-EOB-DTPA)enhanced MRI in predicting recurrence-free survival(RFS)after radical hepatectomy in patients with microvascular invasion(MVI)-negative hepatocellular carcinoma(HCC).Methods:A total of 125 patients with pathologically proved MVI negative HCC who underwent radical hepatectomy from February 2015 to May 2021 were retrospectively enrolled in this study.The preoperative clinical data,MRI images and postoperative pathological data were analyzed.Univariate and multivariate Cox proportional hazards models were used to analyze postoperative independent risk factors for RFS,and a nomogram was constructed to predict 1,3,and 5-year RFS.C-index and receiver operating characteristic(ROC)curve were used to evaluate the predictive ability of the nomogram,and calibration curve was used to evaluate the model's consistency.Results:During follow-up,37 HCC patients relapsed,with a median RFS of 27.1 months[95%confidence interval(CI):43.3~54.1 months].RFS were 88.0%,71.9%and 45.9%at 1,3 and 5 years after operation,respectively.Multivariate analysis showed that HCC pathological grade[hazard ratio(HR)=4.807,95%CI:1.126~20.526,P=0.034],peritumoral hypointensity on hepatobiliary phase(HR=3.755,95%CI:1.736~8.122,P=0.001)and the relative intensity ratio(RIR)in hepatobiliary phase(HR=0.019,95%CI:0.001~0.396,P=0.010)were independent risk factors for postoperative RFS.A nomogram integrating the above three variables,the area under the curve(AUC)was 0.819(95%CI:0.740~0.882,P<0.001),in addition,with bootstrap internal verification,the model C-index was 0.795(95%CI:0.723~0.867),and the calibration curve verification result shows that the predicted probability and the actual observed probability of the model agree well.Conclusion:The nomogram model based on Gd-EOB-DTPA-enhanced MRI can effectively predict the postoperative RFS of MVI-negative HCC patients.