Clinical Impact of Continuation Versus Cessation of Antiviral Therapy in Chronic Hepatitis B: A Modelling Study With Implications for Hepatitis B Cure

Discontinuing antivirals in chronic hepatitis B virus (HBV) 'e' antigen negative infection can enhance HBV surface antigen (HBsAg) loss but risks complications. We modelled the clinical impact of discontinuing antivirals in chronic HBV. We developed a Markov state model with Monte Carlo simulation of chronic HBV to compare continuation of antiviral therapy with 3 strategies of cessation and reinitiation for: (1) virologic relapse, (2) clinical relapse, or (3) hepatitis flare. We simulated the probability of virologic relapse as an exponential decay function from the time of antiviral cessation. We used literature-based estimates for input probabilities following virologic relapse: clinical relapse (60%, conditional on virologic relapse), hepatitis flare (57%, conditional on clinical relapse) and HBsAg-loss (6%-8%). We projected HBsAg loss, cirrhosis, HCC, and survival. In 10 years, cessation strategies would increase cumulative incidence of HBsAg loss from 4.6% to 12.9%-17.3% but would not appreciably change survival (from 90.6% with continuation to 88.1%-89.0%). In an undifferentiated population, continuation would be a preferred strategy to increase average life expectancy (by 0.75-1.05 years) unless HBsAg loss following treatment cessation was > 46%. Sensitivity analyses showed that the decision to continue or stop antivirals would depend on the off-treatment rates of cirrhosis and HCC for people who remain HBsAg-positive but do not fulfil retreatment criteria. Careful selection of people for antiviral cessation using quantitative HBsAg levels could improve survival compared with continuation. Clinical practice guidelines should emphasise selective application of antiviral cessation to persons most likely to lose HBsAg without experiencing liver-related complications.