Author information

• ¹Senior Research Associate, The Forum for Collaborative Research, University of California, Berkeley, School of Public Health, Washington, District of Columbia.
• ²Graduate Student, Toronto Centre for Liver Disease, University Health Network, Toronto, Ontario, Canada.
• ³Scientific Director, Janssen Research and Development, Antwerp, Belgium.
• ⁴AIDS Clinical Trial Group, Chicago Illinois.
• ⁵Program Director, Toronto Centre for Liver Disease, University Health Network, Toronto, Ontario, Canada.
• ⁶Deputy Division Director for Safety, US Food and Drug Administration, Silver Spring, Maryland.
• ⁷Clinical Virology Reviewer, US Food and Drug Administration, Silver Spring, Maryland.
• ⁸Senior Clinical Assessor, Swedish Medical Products Agency, Uppsala, Sweden; Researcher, Department of Medical Sciences, Section of Infectious Diseases, Uppsala University, Uppsala, Sweden.
• ⁹Clinical Assessor, Federal Institute for Drugs and Medical Devices, Bonn, Germany.
• ¹⁰Executive Director, The Forum for Collaborative Research, University of California, Berkeley, School of Public Health, Washington, District of Columbia.
• ¹¹Senior Biostatistician, Institute of Health Policy, Management and Evaluation, University of Toronto, Toronto, Ontario, Canada. Electronic address: Bettina.Hansen@utoronto.ca.

Abstract

Background & aims: Seroclearance of hepatitis B surface antigen (HBsAg) is the desired endpoint of treatment for chronic hepatitis B virus (HBV) infection, according to guidelines. We performed a systematic review and meta-analysis to evaluate the strength of association between HBsAg seroclearance and long-term clinical outcomes.

Methods: We performed a systematic review of the PubMed, EMBASE, and Cochrane Library databases for articles that assessed HBsAg status and reported the incidence of hepatocellular carcinoma (HCC), liver decompensation, liver transplantation, and/or all-cause mortality during follow up. We performed a meta-analysis of rate ratios (RR) using a random effects model independently for each endpoint and for a composite endpoint.

Results: We analyzed data from 28 studies, comprising a total of 188,316 patients with chronic HBV infection (treated and untreated), and 1,486,081 person-years (P-Y) of follow up; 26 reported data on HCC, 7 on liver decompensation, and 13 on liver transplantation and/or death. The composite event rates were 0.19/1000 P-Y for the HBsAg seroclearance group and 2.45/1000 P-Y for the HBsAg-persistent group. Pooled RRs for the HBsAg seroclearance group were 0.28 for liver decompensation for liver decompensation (95% CI, 0.13-0.59; P=.001), 0.30 for HCC (95% CI, 0.20-0.44; P<.001), 0.22 for liver transplantation and/or death (95% CI, 0.13-0.39; P<.001), and 0.31 for the composite endpoint (95% CI, 0.23-0.43; 95% CI, .023-0.43; P<.001). No differences in RR estimates were observed among subgroups of different study or patient characteristics.

Conclusions: In a systematic review and meta-analysis, we found seroclearance of HBsAg to be significantly associated with improved patient outcomes. The results are consistent among different types of studies, in all patient subpopulations examined, and support the use of HBsAg seroclearance as a primary endpoint of trials of patients with chronic HBV infection.