The Predictive Model And Graft Molecular Characteristic Of Early Allograft Dysfunction After Liver Transplantation

Background:Liver transplantation is the most curative therapy for various end-stage liver diseases.Early allograft dysfunction(EAD)is one of the common complications after liver transplantation which accounts for 20%-50% 。 The occurrence of EAD is associated with decresed graft and recipient survival and prolonged ICU stay,resulting in a waste of valuable medical resources.EAD can deteriorate into primary nonfunction.So it is pretty important to predict the incidence of EAD before it occurred.The etiology of EAD is multiple factors,including both donor and recipient factors.Among them,donor liver quality is the most essential factor for EAD occurrence.Analysing the association between donor liver molecular characteristics and EAD is significant for early predicting or treating EAD and evaluating the graft quality.In the end,it could be helpful for improving graft and recipient survival rates.Aims:1.Based on the analysis of liver transplant cases,the predictive model for EAD after liver transplantation would be constructed and validated in a multi-center and large cohort.2.Through the research of the molecular biological characteristics of donor liver,donor liver molecules related to the occurrence of EAD would be found,and the role and potential mechanism of the molecular characteristics in the development of EAD would be studied.Methods:1.Clinical data of 321 donation after circulatory death liver transplantation from the first affiliated hospital of Zhejiang University School of Medicine were analysed.The continuous variables were demonstrated as mean(± standard deviation)or the median(interquartile range).Categorical variables were presented as frequency(percent).Graft survival rate and recipient survival rate of different types of EAD were further analyzed through classify the severity of EAD after liver transplantation.Multivariate analysis through logistic regression was used to identify the risk factors of EAD after liver transplantation.A predictive nomogram model was established to predict the occurrence of EAD after liver transplantation based on the risk factors.Bootstrap was applied for internal validation,and a multi-center 501 liver transplant cases were used for the external validation cohort.Further,conjoint analysis combining inflammatory cytokine and lipid metabolites was used to explore their role in the development of severe EAD.2.Circular RNA sequencing was conducted in donor liver tissues(3 EAD vs 3 Non-EAD)and gene enrichment analysis(GO Ontology,GO)and pathway analysis(Kyoto Encyclopaedia of Genes and Genomes,KEGG)was used to explore the relationship between EAD and graft circ RNAs.QPCR was used to validate the expression of candidate graft circ RNAs in EAD and non-EAD group.The EAD risk score was constructed based on donor liver molecular characteristics to predict the incidence of EAD.Three databases(RNAhybrid,Target Scan and mi RWalk)was used to predict the downstream target mi RNAs and m RNAs,and a circ RNA-mi RNA-m RNA regulatory network was constructed to explore the potential mechanism of EAD.Additionally,we also explored the role of circ NECTIN3 in mouse liver ischemia reperfusion injury to provide a theoretical basis for the treatment of EAD.Results:1.A retrospective analysis of 321 cases of DCD liver transplantation clinical data in our center found that the incidence rates of EAD was 46.4%,and in the validation cohort(n=501),40.5% liver transplant cases developed EAD.EAD can be further divided into EAD type A and type B according to the ALT / AST diagnostic creteria.EAD type B had relatively poor graft and recipient survival rate than those of EAD type A.Multivariate analysis found that graft weight,cold ischemic time,donor age,and recipient MELD score were independent risk factors for EAD.The nomogram model for predicting EAD was constructed based on the above 4 risk factors.And,among the multi-center 501 liver transplant cases,the nomogram model demonstrated a good performance with a high discrimination(c-index: 0.712,95% CI: 0.666-0.758)and calibration(Hosmer-Lemeshow: P = 0.384).Similarly,a nomogram model was constructed based on graft weight,cold ischemia time,and recipient MELD score to predict EAD type B which had poor graft and recipient survival.The nomogram for predicting EAD had a good discrimination(c-index: 0.707,95% CI: 0.641-0.773)and calibration(Hosmer-Lemeshow: P = 0.425)in the multi-center 501 liver transplant cases.Further analysis found that if the EAD nomogram model score was high,the incidence of EAD and EAD type B increased significantly(P = 0.013,0.008)in the long-term surgery group(> 6 hours).It was also validated in the multi-center 501 liver transplant cases.Analysis combined inflammatory cytokine and lipid metabolites showed that IL-6 was significantly correlated with phosphatidylcholine_PC(34:2),phosphatidylinositol_PI(38:3)and phosphatidylinositol_PI(38:4),and G-CSF was significantly related to free fatty acid_FFA(14:0)and triglyceride_TG(54:1)(Mantel test,P < 0.05).They play an important role in the development of EAD type B.The above findings indicated that the predictive model of EAD could be used to predict the occurrence of EAD preoperatively.EAD could be divided into two subtypes,among which the survival rates of type B was significantly decreased.Inflammatory cytokines and lipid metabolites were associated with the development of EAD type B.2.Compared with those without EAD,there were 442 differentially expressed circular RNAs in donor liver tissue in the EAD group(Fold change> 2,P <0.05).Among these circ RNAs,223 circular RNAs were up-regulated,and 219 circular RNAs were down-regulated.The top 3 Gene Ontology(GO)terms were listed below;binding,protein binding and catalytic activity in the molecular functions;intracellular part,organelle and intracellular in the cellular components;cellular process,metabolic process and organic substance metabolic process in the biological processes.The top 3 KEGG pathways associated with host genes of the differential expressed circ RNAs consist of arginine biosynthesis,tryptophan metabolism and glyoxylate and dicarboxylate metabolism.Furthermore,among the 115 liver transplant cases,circ FOXN2 and circ NECTIN3 in the EAD group were significantly down-regulated(P = 0.038;P = 0.024).According to the expression level of two circular RNA molecules(circ FOXN2 and circ NECTIN3)in the graft,the donor liver was classified into circular RNA signature A,B and C.The incidence rates of EAD in circular RNA signature A,B and C was 3.1%,21.2% and 42.0%,respectively.Multivariate analysis found that cold ischemic time and donor liver circular RNA feature subtype were independent risk factors for the occurrence of EAD.In addition,the established EAD risk score based on donor liver characteristics had a good performance(AUC: 0.870;95% CI: 0.797-0.942)).The circ FOXN2 / circ NECTIN3-mi RNA-m RNA regulatory network was constructed based on the databases predicting the downstream targets of circular RNA.Further studies showed that circ NECTIN3 overexpression in mouse liver tissue reduced serum liver enzymes,Suzuki’s pathological score and cellular apoptosis in the ischemia reperfusion injury group.In conclusion,based on the expression of circ FOXN2 and circ NECTIN3 in donor liver tissue,the recipients could be divided into circ RNA signature A,B and C.The risk score based on circ RNA signature had good performance in predicting EAD.Circ NECTIN3 alleviated liver ischemia reperfusion injury in mouse model.Conclusions:1.A nomogram model based on graft weight,cold ischemic time,donor age,and MELD score could predict EAD with good performance.2.According to the severity,EAD could be divided into type A and type B,in which type B survival rate was significantly decreased.It could be important for clinical evaluation.3.Donor liver circ FOXN2 and circ NECTIN3 are associated with the development of EAD.Risk score based on the two circular RNA signature has a good performance predicting EAD.4.Circ NECTIN3 has a protective effect on hepatic ischemia reperfusion injury,which provides an important theoretical basis for the treatment of EAD in the future.