| Objective:Due to the limitations of the current serum markers for diagnosis and prognosis assessment of hepatocellular carcinoma(HCC),it is necessary to continue to search for accurate and effective indicators for diagnosis and prognosis prediction of HCC.Golgi protein 73(GP73)is a transmembrane protein on the Golgi membrane.Serum GP73 levels are low in healthy people,but in patients with hepatitis,liver fibrosis,cirrhosis or HCC,serum GP73 levels will be increased to varying degrees,which is expected to be a new serological indicator of liver diseases.So far,on the one hand,the value of serum in the diagnosis of HCC remains controversial.On the other hand,the prognostic value of serum GP73 in HCC is also rarely explored.This study will explore the value of serum GP73 in the diagnosis and prognosis of HCC,in order to develop more effective and accurate biomarkers,and assist doctors to make clinical diagnosis,choose more favorable treatment strategies,and improve the prognosis of patients.Methods:(1)In the study of GP73 and the diagnosis of HCC,410 patients with HCC who were pathologically diagnosed after radical hepatectomy in the First Hospital of Jilin University from March 2009 to October 2021 were included.The high-risk population of HCC who participated in the "Live Cancer Screening Program of Jilin Province" from 2019 to 2022 and were included in the screening cohort as the source of the control population.The screening population and the HCC population were stratified with cirrhosis and hepatitis type as the main characteristics,and then the screening population were matched with HCC patients by 1:1 propensity score matching method.The basic characteristic and clinical data of the subjects,and the preoperative clinical and postoperative pathological data of patients with HCC were collected.The concentrations of serum GP73,alpha-fetoprotein(AFP),lens culinaris agglutininreactive(AFP-L3)and des-gamma-carboxy prothrombin(DCP)were measured by the magnetic microparticle chemiluminescence immunoassay method.Mann-Whitney U test was used to compare the levels of GP73,AFP,AFP-L3 and DCP between HCC group and non-HCC group.The receiver operating characteristic curve(ROC)and the area under the ROC curve(AUC)was calculated to compare the diagnostic value in HCC.(2)In the study of GP73 and the prognosis of the HCC,334 patients with HCC who were pathologically diagnosed after radical hepatectomy in the First Hospital of Jilin University from March 2009 to December 2020 were followed up for 3 months,6months,1 year and every year after operation.The general demographic data,clinical and pathological data of patients with HCC were collected,and the concentrations of serum GP73 were measured by the magnetic microparticle chemiluminescence immunoassay method.X-Tile software was used to identify the optimal cut-off value of quantitative serological indicators.The Kaplan-Meier method was used to draw the survival curves of GP73,and the log-rank test was used to assess the difference between survival curves.The cox regression was used to calculate hazard ratios(HR)with its95% confidence intervals(95%CI).The Harrell’ concordance index(C-index)was used to evaluate and compare the predictive value of GP73 and BCLC stage.(3)In the study of establishment of nomogram model,2/3 patients(223 cases)in the prognosis study were randomly selected as the training data set,and the remaining1/3 patients(111 cases)were selected as the validation data set,and the HRs with95%CIs was calculated by the univariate cox regression to find prognostic index.The multivariate cox proportional hazard model was used to establish prognostic model.The “rms” package of R software was used to establish nomogram to visualize the model.The C-index and calibration plot was used to evaluate the discrimination and calibration of different indicators and models.The integrated discrimination improvement(IDI)and the net reclassification index(NRI)were used to evaluate the discriminatory performance of the prognostic prediction models and the decision curve analysis(DCA)were used to evaluate the net benefit of different models.Result:(1)In the study of GP73 and the diagnosis of HCC,a total of 820 patients were included.There were 410 patients in HCC group and 410 patients in non-HCC group,and 249 patients(60.7%)in each group had cirrhosis.The distribution of liver cirrhosis and hepatitis in HCC group was exactly the same as that in non-HCC group.Different from AFP,AFP-L3% and DCP,there was no significant difference in the level of serum GP73 between the two groups(P>0.05).Compared with the AUC of AFP,AFP-L3%and DCP [0.823(95%CI: 0.794-0.851),0.762(95%CI: 0.729-0.796)and 0.805(95%CI:0.773-0.837)respectively] in the diagnosis of HCC,the minimum AUC of GP73 was0.525(95%CI: 0.485-0.564)(P<0.001).(2)In the study of GP73 and the prognosis of the HCC,a total of 334 patients with HCC were included,and the median concentrations of GP73 in HCC patients was75.94ng/ml(95%CI: 51.37-124.58ng/ml).The patients were divided into GP73-positive group(GP73≥65 ng/ml)and GP73-negative group(GP73<65ng/ml),with an optimal cut-off value of 65ng/ml for GP73 based on the X-Tile software.The results showed that HCC patients in GP73-positive group had poor survival after hepatectomy(Log-rank P<0.001).GP73-positive patients with HCC was associated with poor survival(HR=3.146,95%CI: 2.053-4.821).The C-index of GP73 for predicting the prognosis of patients with HCC was 0.614(95%CI: 0.576-0.652),which was higher than that of BCLC stage(C-index=0.608,95%CI: 0.563-0.653).(3)In the study of establishment of nomogram model,multivariate cox regression showed that GP73-positive(HR=1.892,95% CI: 1.117-3.204),largest tumor size(per each increase of 1cm)(HR=1.100,95% CI: 1.040-1.163),neutrophil / lymphocyte ratio(NLR)≥2(HR=1.867,95% CI: 1.184-2.945)and high TBIL≥18 μmol(HR=1.621,95%CI: 1.058-2.485)are also independent predictors of poor survival in patients with HCC.A nomogram model based on GP73,largest tumor size,TBIL,and NLR was constructed to predict the overall survival of HCC patients at 3,5,and 10 years.In the training cohort and validation cohort,the C-index of the nomogram model was 0.713(95%CI: 0.656-0.770)and 0.713(95%CI: 0.639-0.787),respectively.The calibration curves also matched well.In the training cohort and validation cohort,the C-index of the nomogram model was higher than that of BCLC stage(0.586 and 0.634)and GP73alone(0.592 and 0.642),with statistical differences(P<0.05).Compared with the BCLC stage and GP73 alone,the prediction efficiency of nomogram model showed positive improvements(the nomogram model vs BCLC stage,the IDI was 0.070 and0.150,and the NRI was 0.157 and 0.173,the nomogram model vs GP73 alone,the IDI was 0.074 and 0.093,and the NRI was 0.199 and 0.237 respectively)(P<0.05).Conclusion:The diagnostic value of serum GP73 in HCC is low,it is not suitable for diagnosis of HCC,however,as its predictive efficacy is comparable to that of BCLC stage,it can be used to predict the prognosis of patients with HCC.The nomogram model based on GP73,largest tumor size,NLR and TBIL is effective in predicting the prognosis of patients with HCC,and is better than BCLC stage or GP73 alone. |
