Development and validation of a novel prognostic nomogram for hepatitis B virus-related acute-on-chronic liver failure patients receiving artificial liver therapy

Abstract

Background Hepatitis B virus-related acute-on-chronic liver failure (HBV-ACLF) is frequently accompanied by short-term morbidity and mortality. However, there have been no studies on the associations between baseline clinicopathologic characteristics at hospital admission and clinical prognosis after receiving artificial liver therapy. Therefore, the current study aimed to develop a prognostic nomogram for predicting the outcomes of patients with HBV-ACLF following artificial liver support. Methods A retrospective study of 110 consecutive patients who were diagnosed with HBV-ACLF between January 2018 and August 2022 was conducted. First, univariate and multivariate logistic regression analyses were performed to determine the independent prognostic factors significantly associated with patient outcomes. Moreover, a predictive nomogram model underlying the prognostic factors was established and further evaluated. The area under the curve (AUC) was used to gauge the predictive accuracy. The calibration curve and decision curve analysis (DCA) were employed to assess the discriminability and clinical effectiveness, respectively. Results In patients with HBV-ACLF, multivariate logistic analysis revealed that age ≥ 40 years (OR 6.76, p = 0.025), middle-stage liver failure (OR 49.96, p < 0.001), end-stage liver failure (OR 19.27, p = 0.002), hepatic encephalopathy (OR 7.06, p = 0.032), upper gastrointestinal hemorrhage (OR 47.24, p = 0.047), and artificial liver therapy consisting of plasma exchange (PE) + plasma exchange double plasma molecular adsorption system (DPMAS) (OR 0.26, p = 0.04) were identified as prognostic factors. Then, we established and evaluated a predictive nomogram with an area under the curve (AUC) of 0.885, which showed better predictive accuracy than the model for end-stage liver disease (MELD) score (AUC of 0.634) and the Child‒Pugh score (AUC of 0.611). Moreover, the calibration curve showed good agreement between the ideal and bias-corrected curves. Decision curve analysis confirmed the better clinical utility of this approach. Conclusion We developed and evaluated a unique nomogram that was more accurate than conventional prognostic models for predicting the clinical prognosis of HBV-ACLF patients receiving artificial liver therapy. As a result, the nomogram may be a helpful tool in clinical decision-making to predict the outcomes of patients with HBV-ACLF.