OBJECTIVE: HBsAg loss is rare in chronic hepatitis B patients, even in the patients with long-term nucleos(t)ide analogue therapy; therefore information about serum HBsAg kinetics will be of value in understanding this unusual occurrence. METHODS: Forty-five consecutive patients were studied, which were all HBeAg positive and never had antiviral therapy prior to lamivudine treatment; they then achieved rapid and good viral responses (defined as undetectable HBV DNA [Roche Lightcycler, less than 1000 copies/ml] at treatment week 24 and they remained so until week 156). Abbott Architect HBsAg assay was used to quantify serum HBsAg and HBV genotypes were determined by direct sequencing. RESULTS: Twenty-six (57.8%) patients had HBeAg loss during the observation and one patient had HBsAg loss following his HBeAg seroconversion. Serum HBsAg levels decreased to 39.5% (median) of their baseline values at week 12, but no further significant reductions of serum HBsAg were found afterwards. Changes of serum HBsAg were comparable between patients with or without HBeAg loss. Serum HBsAg levels at their baselines were higher in HBV genotype B (HBV/B, n = 21) patients than in genotype C (HBV/C, n = 24) patients. HBV/B patients achieved many more HBsAg reductions than HBV/C ones (75.5 vs. 26.0%, median, P less than 0.05) in the first 12 treatment weeks, however HBsAg levels at week 156 were comparable between these two subgroups. HBsAg changes mainly showed two distinct patterns: a biphasic pattern (HBsAg levels were less than 60% of baseline ones at week 12 and 24, n = 25) and a maintaining pattern (HBsAg levels were greater than 80% of the baseline ones at week 12 and 24, n = 14). Logistic regression analysis showed that low serum HBsAg at baseline (odds ratio 0.020, 95% confidence interval 0.002-0.743, P less than 0.05) and HBV/C infection (odds ratio 8.206, 95% confidence interval 1.070-62.948, P less than 0.05) were the determinants of the occurrences of the maintaining pattern. CONCLUSION: In patients we examined, their HBsAg changes were mainly presented as either a biphasic pattern or a maintaining pattern, which were associated with HBV genotypes (B/C) but not with HBeAg loss. This might explain that why HBsAg loss is a rare occurrence even with long-term lamivudine therapy.
OBJECTIVE: HBsAg loss is rare in chronic hepatitis Bpatients, even in the patients with long-term nucleos(t)ide analogue therapy; therefore information about serum HBsAg kinetics will be of value in understanding this unusual occurrence. METHODS: Forty-five consecutive patients were studied, which were all HBeAg positive and never had antiviral therapy prior to lamivudine treatment; they then achieved rapid and good viral responses (defined as undetectable HBV DNA [Roche Lightcycler, less than 1000 copies/ml] at treatment week 24 and they remained so until week 156). Abbott Architect HBsAg assay was used to quantify serum HBsAg and HBV genotypes were determined by direct sequencing. RESULTS: Twenty-six (57.8%) patients had HBeAg loss during the observation and one patient had HBsAg loss following his HBeAg seroconversion. Serum HBsAg levels decreased to 39.5% (median) of their baseline values at week 12, but no further significant reductions of serum HBsAg were found afterwards. Changes of serum HBsAg were comparable between patients with or without HBeAg loss. Serum HBsAg levels at their baselines were higher in HBV genotype B (HBV/B, n = 21) patients than in genotype C (HBV/C, n = 24) patients. HBV/Bpatients achieved many more HBsAg reductions than HBV/C ones (75.5 vs. 26.0%, median, P less than 0.05) in the first 12 treatment weeks, however HBsAg levels at week 156 were comparable between these two subgroups. HBsAg changes mainly showed two distinct patterns: a biphasic pattern (HBsAg levels were less than 60% of baseline ones at week 12 and 24, n = 25) and a maintaining pattern (HBsAg levels were greater than 80% of the baseline ones at week 12 and 24, n = 14). Logistic regression analysis showed that low serum HBsAg at baseline (odds ratio 0.020, 95% confidence interval 0.002-0.743, P less than 0.05) and HBV/C infection (odds ratio 8.206, 95% confidence interval 1.070-62.948, P less than 0.05) were the determinants of the occurrences of the maintaining pattern. CONCLUSION: In patients we examined, their HBsAg changes were mainly presented as either a biphasic pattern or a maintaining pattern, which were associated with HBV genotypes (B/C) but not with HBeAg loss. This might explain that why HBsAg loss is a rare occurrence even with long-term lamivudine therapy.