A Early Predictive Model Of Microvascular Invasion Based On Systemic And Hepatic Inflammation Markers In HBV-associated Hepatocellular Carcinoma Patients

Objective: To evaluate the correlation between microvascular invasion(MVI)and systemic inflammation markers [(Neutrophil-to-lymphocyte ratio(NLR),Platelet-to-lymphocyte ratio(PLR),Systemic immune-inflammation index(SII),Prognostic Nutritional Index(PNI)] and hepatic inflammation markers [Aspartate aminotransferase to platelet ratio index(APRI),γ-glutamyl transferase/ Alanine aminotransferase(γ-GT/ALT)] in patients with hepatitis B virus(HBV)-associated hepatocellular carcinoma(HCC).On this base,the MVI early predictive model of HBV-associated HCC patients was established.Method: A total of 245 cases with HBV-associated HCC who underwent hepatectomy as therapy were analyzed retrospectively in hepatopancreatobiliary surgery clinic of the First Affiliated Hospital of Fujian Medical University from January 2015 to February 2020.205 cases of data were used as the modeling group,and 40 cases were used as the validation group.Taking whether to merge MVI as two-category dependent variable in the modeling group,to determine independent risk factors of HBV-associated HCC with univariate and multivariate analysis.According to the selected independent risk factors,a early model for predicting MVI of HBV-associated HCC was established.Include systemic inflammation markers/hepatic inflammation markers establish submodel.According to the Receiver Operating Characteristic curve(ROC)we can verify the predictive effectiveness of the models in the validation group.Result: Univariate analysis results show that: age、BMI、albumin、ALT、AST、γ-GT、AFP、the number of tumor、tumor location、tumor maximum diameter、NLR、PLR、SII、PNI、APRI、γ-GT/ALT were risk factors of MVI in HBV-associated HCC.Multivariate analysis results show that: NLR,APRI,AFP,albumin,tumor maximum diameter were independent risk factors of MVI in HBV-associated HCC.We established predictive model P1 of MVI combining all independent risk factors and predictive model P4 of MVI combining NLR AFP and APRI.Early predictive model was constructed according to the above factors : P1 = 1/[1 +e^(2.53-1.91X1-2.36X2-1.70X3-2.57X4+1.41X5)];P4 = 1/[1 +e^(2.62-2.37X1-1.73X2-1.82X3)](X1-X5 represent NLR 、 AFP 、 APRI 、 tumor maximum diameter、albumin)。In the modeling group,the Area Under the Curve(AUC)of model P1、P4 is 0.906、0.868.(P=0.072)In the validation group,the AUC of model P1、P4 is 0.928、0.881.(P=0.124)Conclusion: NLR 、 AFP and APRI are independent risk factors of MVI in HBV-associated HCC patients.The predictive model based on systemic and hepatic inflammation markers is suitable for the MVI prediction in patients with HBV-associated HCC.It is helpful to guide individual therapy and have certain clinical value.