异常凝血酶原联合增强CT预测肝癌微血管侵犯价值研究

Study on the value of des-γ-carboxy-prothrombin combined with enhanced CT in predicting microvascular invasion of hepatocellular carcinoma

目的探讨术前异常凝血酶原(DCP)联合增强CT对肝细胞癌(HCC)微血管侵犯(MVI)的预测价值。方法回顾性分析2016-01-01至2019-03-01西南医科大学附属医院收治的行手术切除的105例HCC病人的临床资料。根据病理学检查分为MVI组(30例)和无MVI组(75例),分析HCC发生MVI的影响因素,采用受试者工作曲线(ROC)评估DCP联合增强CT预测HCC发生MVI的价值。结果单因素分析显示,血清DCP水平(t=4.667,P<0.001)、肿瘤大小(t=9.427,P<0.001)、包膜类型(χ~2=18.456,P<0.001)和病理类型(χ~2=11.914,P=0.003)是影响HCC发生MVI的危险因素(P<0.05)。多因素分析显示,肿瘤大小(OR=12.610,P=0.001)、包膜类型(OR=14.363,P=0.008)、血清DCP水平(OR=19.254,P=0.014)是影响HCC发生MVI的独立危险因素(P<0.05)。ROC分析结果显示,血清DCP联合增强CT预测HCC发生MVI的敏感度、特异度及曲线下面积(AUC)分别为90%、74.7%和0.856。结论术前DCP水平检测联合增强CT检查有助于预测HCC是否存在MVI。

Objective To explore the predictive value of preoperative enhanced CT combined with des-γ-carboxy-prothrombin(DCP)level for microvascular invasion(MVI)of hepatocellular carcinoma(HCC).Methods The clinicaldata of 105 cases of HCC diagnosed by pathology after surgical resection from January 1,2016 to March 1,2019 in thethe Affiliated Hospital of Southwest Medical University were analyzed retrospectively.According to the pathologicalresults,they were divided into two groups:experimental group(with MVI):30 cases,control group(without MVI):75 cases,using univariate and multivariate logistic regression analysis to predict influencing factors of MVI in HCC.Thereceiver working curve(ROC curve)was used to evaluate the predictive value of DCP combined with enhanced CT forMVI in HCC.Results Univariate analysis showed that serum DCP level(t=4.667,P<0.001),tumor size(t=9.427,P<0.001),capsule type(χ~2=18.456,P<0.001)and pathological type(χ~2=11.914,P=0.003)were risk factors for MVI in HCC,with statistically significant differences(P<0.05).Multivariate logistic regression analysis showed that tumor size(OR=12.610,P=0.001),capsule type(OR=14.363,P=0.008)and serum DCP level(OR=19.254,P=0.014)were independentrisk factors of MVI in HCC,with statistically significant differences(P<0.05).ROC curve analysis showed that thesensitivity,specificity and area under the curve(AUC)of serum DCP and enhanced CT model for predicting MVI in HCCwere 90.0%,74.7%and 0.856,respectively.Conclusion Preoperative serum DCP level combined with enhanced CThelps predict MVI in HCC.