The Clinical Research About The CT Imaging Features Of HBV-related HCC And Its Correlation With Microvascular Invasion

Background:Hepatocellular carcinoma(HCC)is a malignant tumor with high clinical morbidity and mortality,surgical treatment including hepatectomy and liver transplantation is the most important method for long-term survival of HCC patients,the current surgical skills are mature and perfect,but the recurren ce rate is still very high,the long-term effect is not good;Previous studies have strongly confirmed that the presence of microvascular invasion(MVI)is an imp ortant cause of recurrence after HCC surgery,The treatment directed against MVI will be an im portant new entry point to improve the long-term efficacy of HCC patients;The Code of Practice for the Diagnosis and Treatment of Primary Hepatocellular Carcinoma(2019 Edition)states that patients with liver cancer who can be surgically removed or prepared for liver transplantation are not advised to unde rgo a preoperative liver biopsy to reduce the risk of liver tumor dissemination,the heterogeneity of tumor restricts the molecular detection base d on puncture biopsy,At present,the diagnosis of MVI can only rely on the path ological examination of the specimen after hepatectomy,which leads to the lag of the diagnosis of MVI;Therefore,accurate preoperative prediction of MVI has become the focus of clinical attention and urgent need to solve the problem.M ulti-slice spiral CT enhanced scanning is widely used in the early detection and diagnosis of HCC,suggesting that we should consider whether it is possible to establish a model to predict MVI based on preoperative enhanced CT imaging signs and general clinical feature ?Nomogram is a statistical model for predicti ng the probability of individualized occurrence of clinical events,It makes a sco re according to the contribution of each factor to the dependent variable in the regression model,Finally calculate the total score of each patient to assess the probability of outcome.this study aims to establish a nomogram scoring model that can accurately predict the occurrence of HBV related HCC with MVI to improve clinical decision(patient individualized treatment options,follow-up plan and efficacy prediction,etc).Objective:To discuss the association of MVI of HBV-related HCC with general clinical risk factors and enhanced CT imaging signs,and establish a No mogram model for predicting the risk of occurrence of MVI before operation.Methods:One hundred and thirty two patients after partial hepatectomy and diagnosed with HCC by postoperative pathological immunohistochemistry were collected from our hospital since May 2016 to September 2018.According to the inclusion and exclusion criteria,eighty eligible HCC patients were selected in our study finally.The CT contrast-enhanced imaging features(inclu ding tumor max-diameter、tumor margin、tumor capsule、peritumoral enhance ment、intratumoral hemorrhage necrosis、internal arteries and etc)and the general clinical factors(including age、sex、child-pugh、alpha-fetoprotein lev els and etc)were compared with MVI.statistically significant factors in univaria te logistic regression analysis were included in the multivariate logistic regressi on analysis to screen the independent predictors of MVI.Finally,a nomogram model including AFP,tumor max-diameter,tumor margin,and intratumoral artery was established.The predictive ability of the model was evaluated using the Receiver operator characteristics and compared with the TTPVI(Two-trait predictor of venous invasion,Internal Arteries/Hypodense Halo)and the RVI(Radiogenomic venous invasion,Internal Arteries/Hypodense Halo/Tumer-Liver Difference)model.Results:Pathological results showed that there were 47 cases of MVI positive and 33 cases of MVI negative in 80 HCC patients,60 cases of male and20 cases of female,age range:25-81,mean age 55.7;AFP,tumor margin,tumor max-diameter,internal arteries,peritumor enhancement and hypodense halo were statistically significant(p<0.05)in univariate logistic regression analysi s;multivariate regression analysis showed that AFP,tumor max-diameter,tumor margin and internal arteries were independent predictors of MVI(p<0.05);therefore,a Nomogram model was constructed with AFP,tumor max-diam eter,tumor margin and internal arteries as predictors of MVI.The risk rate ranges from 0.1 to 0.95,The higher total score,the greater risk of MVI in HCC patients.On the basis of the Youden index,the optimal cut off value of the Nomogram scoring model was defined to be 116.38,the sensitivity was 0.957,and the specificity was 0.848.The area under ROC of the Nomogram model was 0.93,(95% CI: 0.86-0.99),and both significant higher than that of TTPVI and RVI(P<0.001).Conclusion:AFP,tumor max-diameter,tumor margin and internal arteries are independent predictors of MVI.For the first time,we established a Nomogr am scoring model for predicting MVI in HBV-related HCC based on preoperati ve CT imaging findings.It shows good accuracy for predicting MVI in HBV-rel ated HCC patients,and higher than the TTPVI and RVI models.it Can be a noninvasive preoperative method for predicting MVI.