The Value Of Nomogram Based On PALBI Score And FIB-4 Score In Predicting Post Hepatectomy Liver Failure Grade B-C

BackgroundWith the rapid increase in hepatocellular carcinoma(HCC),HCC has become the world’s sixth-largest invasive malignant tumor and the second large number of cancer-related deaths.Among existing treatment strategies,surgery removal is still the most effective treatment for HCC patients.Although the level of surgical techniques,perioperative care,and preoperative comprehensive assessment has been significantly improved,the post-hepatectomy liver failure(PHLF)is still the main drive factor of complications and death after hepatectomy,the incidence of PHLF is especially high in patients with lower liver regeneration and liver metabolism capacity in residual liver.The risk of predicting the occurrence of PHLF is critical,which is essential for surgeons to choose individualized treatment.Therefore,a more comprehensive evaluation method is needed to predict the occurrence of PHLF.Although the current evaluation system has a certain clinical value for the prediction of PHLF,it has not comprehensively considered the factors affecting the occurrence of PHLF,such as unbounded liver reserves capacity and fibrosis level.Hence,this study mainly explores the combination of Platelet-Albumin-Bilirubin Score(PALBI score)and Fibrosis-4 Score(FIB-4 score)to construct a nomogram,to predict the occurrence of PHLF more accurately,to identify high-risk patients as soon as possible,and carry out early intervention and treatment to reduce postoperative complications in patients with f occurrence and improve the overall prognosis of HCC patients.ObjectivesIn this study,univariate and multivariate logistic regression analysis was used to obtain independent influencing factors of the occurrence of PHLF grades B-C in HCC patients,and based on this,a nomogram was constructed.By evaluating and verifying the performance of its nomogram in predicting the occurrence of PHLF grades B-C,a predictive model with more diagnostic performance than the existing scoring system was proposed.MethodsThe clinical data of 574 patients with HCC who underwent radical partial hepatectomy and postoperative pathologically confirmed HCC from January 2014 to December 2018 in Guangdong Provincial People’s Hospital,Hepatobiliary Surgery,Pancreatic Surgery were retrospectively collected.These include the last relevant blood draw of the patient before surgery,intraoperative and postoperative conditions.The above-enrolled patients were randomly assigned to the training group(383 cases in total)and the validation group(191 cases in total)according to the ratio of 1:2.Patients in both groups were defined as PHLF grade 0-A and PHLF grade B-C using the International Study Group of Liver Surgery(ISGLS)criteria for liver failure.Based on the last blood draw before surgery,the PALBI score and FIB-4 score were calculated.In the training group,univariate and multivariate logistic regression models were used to analyze the independent influencing factors of the occurrence of PHLF grade B-C and establish a visual prediction model—nomogram.The value of the nomogram in predicting the occurrence of PHLF grade B-C was compared with the current scoring system by receiver operating characteristic(ROC).Finally,the predictive value of the above nomogram for PHLF grade B-C has been verified again in the validation group.Results574 HCC patients were collected in this study,all the patients conform to the both inclusion and exclusion criteria.And randomly assigned to the training set(383 cases)and the internal validation set(191 cases)in a ratio of 2:1.In the above population,85 patients(14.8%)had PHLF grade B-C in the overall study population,of which 59(15.4%)patients in the training set had PHLF grade B-C,and 26(13.6%)patients in the validation set had PHLF grade B-C.The incidence rate of PHLF grade B-C between the two groups was not statistically significant(P=0.619).The results of Univariate and multivariate logistic regression analysis in the training set showed major hepatectomy(P=0.031,OR=2.211),liver cirrhosis(P=0.026,OR=2.296),ascites(P=0.014,OR=0.014),intraoperative blood loss>400ml(P<0.001,OR=4.683),PALBI score grade 2-3(P=0.005,OR=3.609),FIB-4 score grade 2-3(P<0.001,OR=2.077)were independent impact factors(all P<0.05),and all were risk factors affecting the occurrence of PHLF grade B-C(all OR values were higher than 1).The new nomograms constructed on the above independent influencing factors have AUC values of 0.832(95%Cl:0.777-0.886)in the training set,C-index value of 0.832,and mean absolute error of the calibration curve of 0.028.The AUC values of the nomograms in the training set are all larger than the existing scoring systems(AUC values:0.595～0.758).To further verify the predictive ability of the nomogram,the AUC value of the nomogram in the validation set was calculated to be 0.803(95%Cl:0.723-0.883),the C-index value was 0.808,and the mean absolute error of the calibration curve was 0.022.The predictive performance of nomograms in the validation set is also greater than that of existing scoring systems.ConclusionLiver cirrhosis,Major hepatectomy,ascites,intraoperative blood loss>400ml,PALBI score grade 2-3,FIB-4 score grade 2-3 were independent risk factors associated with PHLF grade B-C in HCC patients and can be used as early predictors clinically relevant indicators for PHLF B-C grade.The predictive ability of the nomogram model built on those independent risk factors is higher than the Child-Pugh score,MELD score,ALBI score,APRI score,PALBI score,and FIB-4 score in both the training set and the validation set,which is a reliable new predictive PHLF B-C class model.