基于术前临床及超声特征的预测肝细胞癌切除术后复发的列线图模型构建

A nomogram to predict the risk of postoperative recurrence of hepatocellular carcinoma based on preoperative clinical indicators and ultrasound features

目的基于临床及超声特征,构建肝细胞癌(HCC)患者根治性切除术后2年内复发的列线图预测模型。方法回顾性收集2021年1月至2021年12月在复旦大学附属中山医院行根治性肝肿瘤切除术的405例HCC患者的临床资料,其中男性327例,女性78例,年龄60(53,66)岁。将患者分为两组:训练组(n=283)和测试组(n=122)。根据术后2年内是否复发,283例训练组患者分为复发组(n=73)和未复发组(n=210),122例测试组患者中,2年内复发33例,2年内未复发89例。收集患者的年龄、微血管侵犯、甲胎蛋白、甲胎蛋白异质体(AFP-L3)、异常凝血酶原(PIVKA-II)、肿瘤数目、增强均匀度等资料。在训练组中利用logistic回归分析筛选出与术后复发相关的危险因素,并联合各项因素构建预测HCC复发的列线图模型。校准曲线用于比较训练组及测试组中预测情况与实际情况的一致性。结果多因素logistic回归分析结果显示,年龄越小(OR=0.976,95%CI:0.953~1.000,P=0.004)、AFP-L3(OR=1.066,95%CI:1.014~1.120,P=0.012)和PIVKA-II(OR=1.000,95%CI:1.000~1.001,P=0.042)越高、肿瘤数目多发(OR=0.399,95%CI:0.225~0.706,P=0.038)、增强不均匀(OR=0.472,95%CI:0.243~0.916,P=0.045)的HCC患者,肝部分切除术后复发的风险增加。基于上述变量构建的列线图在训练组中模型C指数为0.87(95%CI:0.81~0.93),测试组中模型C指数为0.83(95%CI:0.77~0.89)。列线图模型在训练组及测试组中预测HCC患者肝部分切除术后2年内复发的校准曲线拟合程度高,预测情况与实际情况基本一致。结论基于术前临床和超声特征构建的列线图模型可对HCC患者根治性切除术后2年内的复发风险进行预测。

ObjectiveTo establish a nomogram prediction model for recurrence within 2 years after radical resection of hepatocellular carcinoma(HCC)based on clinical and ultrasonographic characteristics.MethodsClinical data from 405 HCC patients(including 327 males and 78 females),aged 60(53,66)years old,who underwent radical hepatectomy in the Zhongshan Hospital,Fudan University,from January to December 2021,were retrospectively collected.The patients were divided into two groups:the training group(n=283)and the validation group(n=122).Based on recurrence within 2 years after surgery,the 283 patients in the training group were further categorized into the recurrence group(n=73)and the non-recurrence group(n=210).Among the 122 patients in the validation group,33 had recurrence within 2 years,while 89 did not.Data on age,microvascular invasion,alpha-fetoprotein(AFP),AFP lentil lectin-reactive fraction(AFP-L3),protein induced by vitamin K absence or antagonist-II(PIVKA-II),tumor number,and enhancement homogeneity were collected.Logistic regression analysis was performed on the training group to identify risk factors associated with postoperative recurrence,and a nomogram model for predicting HCC recurrence was constructed based on these factors.Calibration curves were used to compare the consistency between predicted and actual outcomes in both the training and validation groups.ResultsMultivariate logistic regression analysis revealed that younger age(OR=0.976,95%CI:0.953-1.000,P=0.004),higher AFP-L3(OR=1.066,95%CI:1.014-1.120,P=0.012),higher PIVKA-II(OR=1.000,95%CI:1.000-1.001,P=0.042),multiple tumors(OR=0.399,95%CI:0.225-0.706,P=0.038),and heterogeneous enhancement(OR=0.472,95%CI:0.243-0.916,P=0.045)were significant risk factors for recurrence after partial hepatectomy in HCC patients.The nomogram constructed based on these variables had a C-index of 0.87(95%CI:0.81-0.93)in the training group and 0.83(95%CI:0.77-0.89)in the validation group.The calibration curves for predicting recurrence within 2 years after partial hepatectomy in HCC patients showed a high degree of fit in both the training and validation groups,indicating a good agreement between predicted and actual outcomes.ConclusionThe nomogram model constructed based on preoperative clinical and ultrasonographic characteristics can effectively predict the risk of recurrence within 2 years after radical resection of HCC.