To Establish A Preoperative Nomogram Model For Predicting The Risk Of Hepatocellular Carcinoma With Microvascular Invasion Based On Lymphocyte Subsets

Objective:Primary liver cancer is the sixth most common cancer in the world and the third leading cause of cancer death,among which hepatocellular carcinoma(HCC)accounts for 75% to 85%.Hepatectomy is an important means for patients with HCC to obtain long-term survival,but the recurrence rate after 5 years is 50% to70%.Microvascular invasion(MVI)is a related factor for postoperative recurrence of liver cancer,and lymphocytes are the main effector cells of anti-tumor immunity.Since MVI is the result of postoperative pathological examination and its presence or absence will affect the choice of treatment options,this study intends to establish and validate a nomogram model for predicting the risk of MVI in HCC patients before surgery based on lymphocyte subsets,so as to provide references for clinicians in diagnosis and treatment.Methods:This study included HCC patients who underwent liver resection in the Affiliated Hospital of Qingdao University from January 2013 to October 2021 and had complete clinical data,and collected general clinical data,preoperative imaging and serological indicators of the patients,including: visiting time,gender,age,history of liver cirrhosis,maximum tumor diameter,tumor number,lymphocyte count,total bilirubin,direct bilirubin,aspartate aminotransferase(AST),alanine aminotransferase(ALT),De Ritis ratio(AST/ALT),γ-glutamyl transferase(GGT),alpha-fetoprotein(AFP),lymphocyte subsets,etc.According to the visiting time,patients were divided into a modeling group(the first 170 cases)and a validation group(the last 40 cases).The chi-square test was used to compare whether there is a statistical difference in the data between the two groups.Based on the clinical data of the patients in the modeling group,the indicators with statistical differences between MVI positive patients and negative patients were screened by the rank sum test,the cutoff-value of the indicators to predict MVI was determined by the receiver operating characteristic(ROC)curve,and the univariate and multivariate Logistic Regression analysis screened independent factors associated with MVI.R software was used to establish a preoperative nomogram model for predicting the risk of HCC with MVI,and the clinical data of patients in the validation group were used for external validation.The data were statistically analyzed by SPSS 22.0 software,and P<0.05 was considered statistically significant.Results:1.A total of 210 HCC patients were enrolled,including 169 males and 41 females,aged 57(12)years(range: 30-80 years).Compared with the data of patients in the modeling group and the validation group,only direct bilirubin had a statistical difference between the two groups(P=0.024),while age,gender,history of liver cirrhosis,total bilirubin,De Ritis ratio,AFP,GGT,the opposite number of activated peripheral blood T cells ratio(-aPBTLR),and maximum tumor diameter had no statistical difference(P>0.05).2.Based on the data of the patients in the modeling group,the indicators with statistical differences between the MVI positive and negative groups were screened out through the rank sum test,including: GGT concentration,De Ritis ratio,-aPBTLR and the largest diameter of the tumor.The cutoff-values of De Ritis ratio,GGT concentration,-aPBTLR,and maximum tumor diameter predicted by ROC curve were 0.95(AUC:0.634,95%CI: 0.549-0.719),38.2 U/L(AUC: 0.604,95%CI: 0.518-0.689),-6.05%(AUC: 0.660,95%CI: 0.578-0.742),4cm(AUC: 0.618,95%CI:0.533-0.703).The results of univariate and multivariate Logistic regression analysis showed that De Ritis ratio ≥ 0.95,GGT concentration ≥ 38.2 U/L,-aPBTLR ≥-6.08%,and tumor diameter ≥ 4 cm were independent factors associated with MVI in patients with HCC(P< 0.05).3.Using R software to establish a nomogram prediction model based on the above four indicators has good prediction performance,and the C indices of the modeling group and the vadilation group are 0.758 and 0.751,respectively.Both Decision Curve Analysis(DCA)and Clinical Impact Curve(CIC)showed that the nomogram model has good clinical benefits.Conclusions:Compared with MVI-negative HCC patients,MVI-positive patients had higher De Ritis ratio,GGT concentration,maximum tumor diameter,and-aPBTLR.These four data are also valuable indicators for preoperative prediction of HCC with MVI.Based on this,this study established a visualized MVI nomogram prediction model,and evaluated the prediction performance and clinical effectiveness of the model through internal and external validation to guide clinicians in choosing treatment options.