Pb2409: clinical prognostic model of liver failure after allogeneic haematopoietic stem cell transplantation

Topic: 22. Stem cell transplantation - Clinical Background: As a rare but life-threatening critical complication following allogenic haematopoietic stem cell transplantation (allo-HSCT), liver failure is significantly related to poor survival and increased non-relapse mortality. Therefore, identifying patients at high risk of death who should receive specific therapeutic management is key to improving their survival. Here, we established the first clinical prognostic model to predict post-transplant liver failure prognosis to facilitate physician decision-making. Aims: This study was aimed at establishing a prognostic model for selecting post-transplant patients who are at a higher risk of liver failure and disease mortality. Methods: We retrospectively evaluated 113 consecutive liver failure patients who underwent allo-HSCT at a single centre between 2010 and 2022. These patients were further randomly divided into a derivation cohort of 57 patients and a validation cohort of 56 patients. Pearson’s chi-square test and Fisher’s exact test were applied to compare the categorical variables. Outcomes described by the cumulative incidence of overall survival (OS), non-relapse mortality (NRM) and relapse were evaluated by Kaplan-Meier method and the log-rank test. We assessed the prognostic model performance by evaluating the discrimination [area under the curve (AUC)], calibration (calibration plot), and net benefit [decision curve analysis (DCA)]. Results: In total, 113 of 16819 patients (0.67%) were diagnosed with liver failure after transplantation. Disseminated intravascular coagulation (DIC), viral hepatitis, acute kidney injury (AKI), gastrointestinal disease or bleeding, poor platelet implantation, and chronic graft versus host disease (cGHVD) were determined to be significant risk factors for the occurrence of liver failure after allo-HSCT. Because of the higher mortality rate after liver failure, these factors are also significantly related to mortality in patients with liver failure after allo-HSCT. The AUCs of the model were 0.887 (95% CI 0.833-0.940) and 0.770 (0.701-0.839) for the training and validation cohorts, respectively. The calibration plots showed high consistency between the predicted and observed outcomes. Patients were further classified into low-risk and high-risk categories while considering the observed death distribution of the total cohort by tertiles of 2-year outcome probability according to the score. Significant differences in Kaplan-Meier estimations of overall survival were obtained. Summary/Conclusion: The predictive models developed for the occurrence and mortality of liver failure performed well in terms of discrimination, and they might assist clinicians with personalized strategies for liver failure prevention and treatment in patients after allo-HSCT. Keywords: Liver, Allogeneic hematopoietic stem cell transplant, Prognostic factor