Background: Data regarding the outcome of hepatitis B virus (HBV)-related cirrhosis after the onset of decompensation is scanty.
Method: From January 1998 to December 2008, a retrospective-prospective inception cohort study involving HBV-related decompensated cirrhotics was performed. Predictors of death and clinical events after the onset of decompensation were evaluated. Patients with co-infection with hepatitis C virus and/or human immunodeficiency virus, alcohol consumption to any degree and diabetes diagnosed before the detection of liver disease were excluded.
Result and analysis: Two hundred and fifty-three patients (231 males, 139 e-negative), including 102 untreated patients, were analysed. The mean (+/-SD) age was 43.0 (+/-12.0) years. The mean (+/-SD) follow-up period was 47 (+/-47) months. Decompensation was the first presentation of liver disease in 210 (83%) patients. Ascites (70%) and variceal bleed (28%) were predominant modes of decompensation. Forty-three (17%) patients died (22 vs 14% in untreated and treated cohort, respectively; P=0.002). Type 2 hepato-renal syndrome was the commonest cause of death (32%). Survival was independent of e-antigen status. In the total cohorts, predictors of death were occurrence of sepsis with systemic inflammatory response (SIRS), ascites as the initial mode of decompensation, absence of antiviral therapy and events of high-grade hepatic encephalopathy [hazards ratios (HR) of 4.4, 3.6, 2.2 and 1.7 respectively]. In the untreated cohort, initial decompensation with ascites and development of sepsis with SIRS were independent predictors of death (HR 8.5 and 2.3 respectively), while 5-year survival was higher in patients having initial decompensation with variceal bleed vs ascites (29 vs 16%, respectively, P=0.002).
Conclusion: Decompensation with ascites and sepsis with SIRS predict reduced survival. Antiviral therapy beyond 6 months improves outcome.