Early prediction model for prognosis of patients with hepatitis-B-virus-related acute-on-chronic liver failure received glucocorticoid therapy

Abstract

Abstract Background Early prediction for short-term prognosis is essential for the management of hepatitis B virus (HBV)-related acute-on-chronic liver failure (ACLF). In this study, we aim to establish a noninvasive model for predicting the 90-day mortality in patients with HBV–ACLF received glucocorticoid therapy. Methods Two hundred and eighty patients with HBV–ACLF were enrolled from July 2010 to June 2022. All patients received routine medicine treatment and 204 of them received additional glucocorticoid treatment. Then, the patients who received glucocorticoid treatment were randomly divided into a training cohort and a validation cohort. An early prediction model for 90-day mortality of HBV–ACLF was established in the training cohort and then validated in the validation cohort. Results HBV–ACLF patients received glucocorticoid treatment showed significantly better survival that those not (P < 0.01). In the training cohort, a noninvasive model was generated with hepatic encephalopathy grade, INR, total bilirubin, age and SIRS status, which was named HITAS score. It showed significantly better predictive value for 90-day mortality of HBV–ACLF than MELD score and Child–Turcotte–Pugh score in both the training cohort and validation cohort. Using the Kaplan–Meier analysis with cutoff points of 2.5 and 3.47, the HITAS score can classify HBV–ACLF patients into different groups with low, intermediate and high risk of death after glucocorticoid therapy. Conclusions We proposed a HITAS score, which was an early prediction model for the prognosis of HBV–ACLF. It might be used to identify HBV–ACLF patients with favorable responses to glucocorticoid treatment.