| Objective:To explore the correlation analysis of the risk factors and pre-dictive value of serum AFP,DCP combined with preoperative Gd-EOB-DTPA MRI on MVI of HCC.Methods:The clinicopathological data and Gd-EOB-DTPA MRI imaging data of 105 patients with HCC who met the inclusion and exclusion criteria admitted to the Department of Hepatobiliary Surgery of the First Affiliated Hospital of Guangxi Medical University from June 2018 to December 2019were retrospectively analyzed.According to whether microvascular invasion(MVI)occurred in postoperative pathology,they were divided into two groups:group without MVI(n=71 cases)and group with MVI(n=34 cases).Research variables of single factor and multiple factors analysis using binary Logistic regression analysis,it is concluded that independent risk factors of MVI of HCC and establish a diagnostic model of MVI prediction of HCC diagnosis model for preoperative prediction about MVI by Gd-EOB-DTPA MRI combined with serum AFP and DCP.The ROC curve can be used to evaluate the diagnostic effectiveness of the diagnostic model,and the area under the curve(AUC),sensitivity and specificity can be obtained;Hosmer-Lemeshow goodness of fit test was used to conduct calibration analysis on the diagnostic model to evaluate whether the predicted risk of the diagnostic model was consistent with the actual risk.p<0.05 was considered to be significant.Results:among the 105 patients,34 had MVI(with MVI group)and 71 had no MVI(without MVI group).The results of univariate logistic regression analysis showed that there were no significant differences in age,gender,history of hepatitis B,hepatitis B DNA,tumor number,BCLC stage,edge shape,fast in and fast out,peripheral high signal of arterial stage and magnetic resonance diffusion weighted imaging(DWI)between the two groups(all p>0.05),The comparison of serum AFP,DCP,tumor diameter,capsule integrity,and peripheral low signal of hepatobiliary stage showed that all p<0.05,indicating that the differences were significant.Multivariate analysis results showed that:Serum AFP(p=0.038),DCP(p=0.048)and capsule integrity(p=0.034),peripheral low signal of hepatobiliary stage(p=0.047),all p<0.05,which proved that they were independent risk factors for MVI of HCC.The predictive diagnosis model of serum AFP,DCP combined with preoperative Gd-EOB-DTPA MRI for MVI was established.The results of ROC curve analysis showed that the AUC under the curve was 0.807,the sensitivity was94.1%,and the specificity was 63.4%,indicating that the diagnostic model had a good diagnostic efficiency;Then,the Hosmer-Lemeshow goodness of fit test was used to analyze the calibration degree of the established diagnostic model for predicting MVI,and the result was x~2=8.436,p=0.296.p>0.05 indicates that there is no significant difference between the actual measured value and the expected value of the predictive diagnostic model,indicating that the predictive diagnostic model has a good calibration ability,and it also indicates that the predictive risk of the diagnostic model has a high consistency with the actual risk.Conclusion:Serum AFP,DCP and capsule integrity,peripheral low signal of hepatobiliary stage were independent risk factors for MVI of HCC.Serum AFP and DCP combined with preoperative Gd-EOB-DTPA MRI can effectively predict MVI of HCC. |