Determinants of re‐compensation in patients with hepatitis B virus‐related decompensated cirrhosis starting antiviral therapy

Summary Background Despite antiviral therapy, liver function often fails to recover in patients with hepatitis B virus (HBV)‐related decompensated cirrhosis. Aim To establish a prognostic model to predict re‐compensation in patients starting potent nucleos(t)ide analogue (NUC) therapy Methods We analysed 311 consecutive patients with HBV‐related decompensated cirrhosis treated with entecavir or tenofovir. The primary outcome was re‐compensation, defined as recovery to a Child–Pugh score of 5. The BC2AID score was developed from a cohort of 152 subjects based on competing risk models and validated in another cohort of 159 subjects. Results Re‐compensation occurred in 57.2% and 66.7% of the subjects in the derivation and validation cohorts, respectively. Six independent predictors for re‐compensation were identified in the derivation cohort and these comprised the BC2AID score: bilirubin ≤5 mg/dL (adjusted sub‐distribution hazard ratio [aSHR] 2.18), absence of severe complications (aSHR 2.78), alpha‐fetoprotein (AFP) ≥50 ng/mL (aSHR 2.54), alanine aminotransferase ≥200 IU/L (aSHR 2.62), international normalised ratio ≤1.5 (aSHR 2.37) and ≤6 months from initial decompensation until initiation of NUCs (aSHR 4.79). In the validation cohort, the area under the receiver operating characteristic curve of the BC2AID score for re‐compensation within 1 year of NUC therapy was significantly higher than that of the Child–Pugh, MELD, MELDNa and BE3A scores (0.813 vs 0.691, 0.638, 0.645 and 0.624, respectively; all P