Background: Hepatitis B e antigen (HBeAg) seroconversion is a treatment endpoint for HBeAg-positive CHB, and a necessary precursor to HBsAg loss. Biomarkers that predict serological outcomes would be useful.
Aim: To evaluate the utility of measuring HBeAg levels for predicting HBeAg seroconversion and HBsAg loss under long-term tenofovir (TDF) therapy.
Methods: A total of 266 patients were enrolled into a phase III study of TDF vs adefovir (ADV) for 48 weeks in HBeAg-positive patients, followed by open-label TDF up to 384 weeks. Serum HBeAg levels were measured for subjects with samples available at both baseline and week 24 of treatment (n = 200). Analysis compared subjects who achieved HBeAg seroconversion by week 384 vs no HBeAg seroconversion.
Results: HBeAg seroconversion rate was 52% by week 384. Time to HBeAg seroconversion was 80 weeks (IQR: 36-162). HBeAg decline at week 24 was associated with HBeAg seroconversion (1.63 vs 0.90 log10 PEIU/mL, P = .002). The optimal threshold for identifying HBeAg seroconversion was HBeAg decline ≥2.2 log10 PEIU/mL at week 24, with HBeAg seroconversion achieved by 76% of patients, compared to 44% if HBeAg decline <2.2 log10 (P < .0001). HBeAg decline ≥2.2 log10 PEIU/mL at week 24 was associated with HBsAg loss in genotype A or D patients (38% vs 15%, P = .03). Precore/basal core promotor variants were associated with lower baseline HBeAg levels, but not HBeAg seroconversion.
Conclusion: Decline in HBeAg levels by week 24 was associated with HBeAg seroconversion and HBsAg loss in HBeAg-positive chronic hepatitis B patients treated with long-term TDF.
© 2017 John Wiley & Sons Ltd.