Incidence of Hepatic Decompensation After Nucleos(t)ide Analog Withdrawal: Results From a Large, International, Multiethnic Cohort of Patients With Chronic Hepatitis B (RETRACT-B Study).

ABSTRACT

INTRODUCTION: Despite improvements in the management of chronic hepatitis B (CHB), risk of cirrhosis and hepatocellular carcinoma remains. While hepatitis B surface antigen loss is the optimal end point, safe discontinuation of nucleos(t)ide analog (NA) therapy is controversial because of the possibility of severe or fatal reactivation flares. METHODS: This is a multicenter cohort study of virally suppressed, end-of-therapy (EOT) hepatitis B e antigen (HBeAg)-negative CHB patients who stopped NA therapy (n = 1,557). Survival analysis techniques were used to analyze off-therapy rates of hepatic decompensation and differences by patient characteristics. We also examined a subgroup of noncirrhotic patients with consolidation therapy of ≥12 months before cessation (n = 1,289). Hepatic decompensation was considered related to therapy cessation if diagnosed off therapy or within 6 months of starting retreatment. RESULTS: Among the total cohort (11.8% diagnosed with cirrhosis, 84.2% start-of-therapy HBeAg-negative), 20 developed hepatic decompensation after NA cessation; 10 events were among the subgroup. The cumulative incidence of hepatic decompensation at 60 months off therapy among the total cohort and subgroup was 1.8% and 1.1%, respectively. The hepatic decompensation rate was higher among patients with cirrhosis (hazard ratio [HR] 5.08, P < 0.001) and start-of-therapy HBeAg-positive patients (HR 5.23, P < 0.001). This association between start-of-therapy HBeAg status and hepatic decompensation remained significant even among the subgroup (HR 10.5, P < 0.001). DISCUSSION: Patients with cirrhosis and start-of-therapy HBeAg-positive patients should be carefully assessed before stopping NAs to prevent hepatic decompensation. Frequent monitoring of viral and host kinetics after cessation is crucial to determine patient outcome.