Serious adverse events after cessation of nucleos(t)ide analogues in individuals with chronic hepatitis B: A systematic review and meta-analysis

BACKGROUND & AIMS: The risk of serious clinical outcomes following cessation of nucleos(t)ide analogues (NUCs) in individuals with chronic hepatitis B remains poorly characterized. This systematic review and meta-analysis aimed to evaluate current literature on this issue. METHODS: We searched PubMed, Embase, and Web of Science for NUC stop studies that noted clinical outcomes published between January 1, 2006 and August 18, 2022. We performed meta-research analyses to examine the relationships of reported outcomes with study designs and characteristics and also pooled studies with non-overlapping populations to provide risk estimates for the proportions of (1) severe hepatitis flares or hepatic decompensation or (2) hepatitis flare-related death or liver transplantation. RESULTS: The meta-research analysis included 50 studies of highly heterogeneous designs and characteristics. We found that reporting of safety outcomes varied widely according to outcome definition, follow-up duration, and sample size. Only ten studies prespecified safety events as the study outcome, and only four had an outcome definition to include hepatic insufficiency, a follow-up duration >12 months, and a sample size >100 patients. We further pooled 15 studies with 4,525 individuals and estimated that severe hepatitis flares or decompensation would occur in 1.21% (95% CI 0.70–2.08%), with significant heterogeneity (I(2) = 54%, p <0.01), while hepatitis flare-related death or liver transplantation would occur in 0.37% (95% CI 0.20–0.67%), without significant heterogeneity (I(2) = 0.00%, p = 1.00). CONCLUSIONS: Current literature on the risk of serious clinical outcomes following NUC cessation is very limited and highly heterogeneous. Pooled analyses of available data found approximately 1% of patients who stopped NUCs developed severe flares or hepatic decompensation. IMPACT AND IMPLICATIONS: Current literature regarding the safety concerns surrounding NUC cessation for individuals with chronic hepatitis B is limited and heterogeneous in designs and characteristics, and thus should be interpreted with great caution. Based on currently available data, the proportion of patients that develop severe hepatitis flares or hepatic decompensation was estimated at 1.21% and that of flare-related death or liver transplantation at 0.37%. Our findings are important for individuals receiving nucleos(t)ide analogues for hepatitis B virus infection because we not only pooled currently available data to estimate the risk of serious clinical adverse events following treatment cessation but also uncovered critical limitations of existing literature regarding the safety of finite therapy.