| Objective To investigate the effect of serum prealbumin-bilirubin score(PALBI)to predict posthepatectomy liver failure(PHLF)after hepatectomy for HBV related hepatocellular carcinoma(HCC).Methods Data on 919consecutive patients who underwent liver resection with curative intent for HBV-related HCC during September 2013 to December 2016 were included in this retrospective study.These patients were divided into a training cohort(n=689)and a validation cohort(n=230)using the 3:1 matching principle.The training cohort was divided into the control group(n=546)and the PHLF group(n=143)according to whether PHLF occurred.The independent risk factors of PHLF in the training cohort were identifed by Logistic multivariate regression analysis.Based upon the independent risk factors,PALBI score for predicting PHLF of HCC was established.The accuracy of model for predicting PHLF was respectively detected in two groups by area under the receiver operating characteristic curve(AUC)and compared with commonly predictive systems,such as Child-Pugh score,model for end-stage liver disease(MELD)score and serum albumin-bilirubin(ALBI)score.The cutoff value of PALBI score in predicting PHLF was determined according to the Youden index.PALBI score were stratifed by dichotomy to analyze correlations with incidence and grade of PHLF.The validation cohort validates the results of the training cohort.Results In this study,there were 194 patients developed PHLF(21.1%),143 patients developed PHLF(20.8%)in the training cohort,and 51 patients developed PHLF(22.2%)in the validation cohort.The baseline characteristics between the training cohort and the validation cohort were not significantly different.Compared with the control group,patients with PHLF in the training cohort had older age,higher incidence of cirrhosis,CSPH,ascites,multiple tumors,vascular invasion and extrahepatic metastasis,amount of blood loss,lager extent of resection and HBV-DNA≥10~3 IU/mL,higher level of prothrombin time,aspartate aminotransferase,tumor diameter,and total bilirubin,lower level of prealbumin,albumin and platelet count.All these differences are statistically significant(P<0.05).Multivariate Logistic regression analyses showed the factors including HBV-DNA≥10~3 IU/mL,total bilirubin,prealbumin,platelet count,aspartate aminotransferase,prothrombin time,intraoperative blood loss≥400 mL and major liver resection were closed related to PHLF.The ability of PALBI score(AUC 0.733)to assess PHLF preoperatively was superior to that of the Child-Pugh score(AUC 0.56),MELD score(AUC 0.652)and ALBI score(AUC 0.683)in the training cohort.Similar results were obtained in the entire validation cohort(AUC:0.752 vs.0.599-0.678).In addition,in order to eliminate the effect of a small residual liver volume on PHLF,the ability of each scores in the training and validation cohorts to predict PHLF was calculated respectively in these two cohorts of patients who underwent minor liver resection,and also obtained the similar results.With an optimal cutoff value of0.39,the sensitivity and specificity of the PALBI score for predicting PHLF were 75.4%and 63.7%,respectively,and the incidence and grade of PHLF in patients with high risk group(PALBI score>0.39)was significantly higher than in the low-risk group(PALBI score≤0.39),and the incidence and grade of PHLF increased with the increase of PALBI score Conclusion PALBI score predicted PHLF in patients with HBV-related HCC undergoing liver resection with curative intent more accurately than the Child-Pugh,MELD or ALBI scores.The PALBI score is a simple,non-invasive,reliable novel model in predicting PHLF. |