Long-term prognosis with or without nucleot(s)ide analogue therapy in hepatitis B virus-related decompensated cirrhosis.

Abstract

The development of nuleos(t)ide analogues (NAs) has dramatically changed the natural history of chronic hepatitis B virus (HBV) infection. In this study, we compared patients with HBV-related decompensated cirrhosis with and without NA therapy in terms of hepatocarcinogenesis and all-cause, liver-related, and non-liver-related mortality. This study enrolled 160 patients with decompensated cirrhosis, 78 of whom were treated with NA therapy (NA group) and 82 of whom were not (non-NA group). Propensity score matching and inverse probability weighting were performed to adjust the baseline characteristics in the NA and non-NA groups. Liver-related and non-liver-related mortality were analysed using the competing risks IPW cumulative incidence functions estimator. The Cox proportional hazards model and the Fine and Gray proportional hazards model were used to analyse factors associated with hepatocarcinogenesis and all-cause, liver-related, and non-liver-related mortality. HBV DNA ≥20,000IU/ml (adjusted hazard ratio [aHR], 8.440) and dyslipidemia (aHR, 0.178) were independently associated with hepatocarcinogenesis. HBV DNA ≥20,000IU/ml (aHR, 4.360) and non-NA group (aHR, 4.802) were independently associated with all-cause mortality. Diabetes mellitus (aHR, 4.925), FIB-4 score >3.6 (aHR, 4.151), non-NA group (aHR, 9.180), presence of dyslipidemia (aHR, 0.182) and male gender (aHR, 3.045) were independently associated with liver-related mortality. HBV DNA ≥20,000IU/ml (aHR, 3.216) and high age (aHR, 2.692) were independently associated with non-liver-related mortality. Although the cumulative incidence rate of hepatocarcinogenesis and non-liver-related mortality was not reduced by NA therapy, viral suppression reduced liver-related mortality in patients with DC.