Journal of Practical Hepatology ›› 2026, Vol. 29 ›› Issue (5): 745-748.doi: 10.3969/j.issn.1672-5069.2026.05.027

• Liver transplantation • Previous Articles     Next Articles

Construction and validation of predicting model by combination of serum IL-34, ADR and MELD in recipients with liver transplant for early allograft dysfunction

Zhao Yuhao, Zhang Hezhao, Zhang Rui, et al   

  1. Department of Hepatobiliary Surgery, First Hospital, Shanxi Medical University, Taiyuan 030000, Shanxi Province, China
  • Received:2026-02-24 Published:2026-09-14

Abstract: Objective This study aimed to construct and validate a predicting model by combination of serum IL-34 (IL-34), alanine aminotransferase/direct bilirubin ratio (ADR) and model for end -stage liver disease (MELD) score in recipients with liver transplant (LT)for early allograft dysfunction (EAD). Methods A total of 156 patients were encountered in our hospital between January 2020 and December 2024, and all underwent allogeneic liver transplantation (ALT). Serum IL-34 level was assayed by ELISA, routine biochemical parameters were detected, and ADR and MELD score were calculated. Multivariate Logistic regression analysis was performed to identify independent predictors for EAD occurrence, predicting models were established and predicting performance were evaluated by receiver operating characteristic (ROC) curve analysis. Results By end of seven days after LT, EAD occurred in 27 cases (17.3%) and didn’t in 129 cases (82.7%) in our series; baseline percentage of Child-Pugh class C and MELD score in patients with EAD were much higher than in those without EAD(P<0.05); serum IL-34, ALT, direct bilirumin levels and ADR in patients with EAD were (87.6±15.3)pg/mL, (1865.4±320.8)U/L, (58.3±12.4)μmol/L and(31.9±7.8), all significantly higher than [(45.2±10.7)pg/mL, (642.3±185.6)U/L, (23.7±8.9)μmol/L and (27.0±6.5), respectively, P<0.001] in those without EAD; multivariate Logistic regression analysis showed that baseline MELD score, Child-Pugh classification, serum IL-34 level and ADR were all the independent risk factors for EAD occurrence after LT (OR>1,all P<0.05); the AUC was 0.913(95%CI:0.865-0.961), with sensitivity of 88.9% and specificity of 85.3%, when MELD, Child-Pugh classification, serum IL-34 level were combined with ADR to predict EAD happening after LT. Conclusion Combination of serum IL-34 and ADR with liver reservoir index has a promising efficacy in predicting EAD after LT, which warrants further clinical verification.

Key words: Liver transplantation, Early allograft dysfunction, Interleukin-34, Alanine aminotransferase/direct bilirubin ratio, Model for end -stage liver disease, Prediction