BACKGROUND: Quantification of HBsAg in serum may be of clinical importance in predicting HBsAg seroconversion and complete response to treatment. METHODS: Serum HBsAg was quantified by ADVIA Centaur in 63 patients with HBeAg-negative chronic hepatitis B (CHBe-). A total of 42 had received interferon-alpha2b (IFN-alpha2b) (median 12.1 months; 19 sustained responders including 12 HBsAg-seroconvertors; 23 non-sustained responders) and 21 were on lamivudine (LAM) (median 33.0 months). Measurements were done at baseline, during and at the end of treatment, and during and at the end of follow-up. RESULTS: Baseline median [interquartile range (IQR)] HBsAg levels in all patients were 3286 (1602-7458) IU/ml, not different between IFN- and LAM-treated (P = 0.139). IFN significantly depressed HBsAg in all patients except IFN non-responders, but HBsAg decline persisted only in sustained responders. Low pretreatment HBsAg level was the only significant prognostic variable of HBsAg seroconversion by multivariate analysis. LAM treatment also suppressed HBsAg levels but at a significantly slower rate compared with IFN (P = 0.022). The median (IQR) estimated time to HBsAg undetectability (ETU-HBsAg), derived from best curve fitting, was 127 (87.6-263.5) months for LAM virological responders and 65.3 (36.3-95.0) months for IFN sustained responders (P = 0.002). In 12 HBsAg seroconvertors, ETU-HBsAg was similar to the real time of HBsAg loss (P = 0.525) and seroconversion (0.056). CONCLUSIONS: In CHBe-, IFN induces a sharper decrease in serum HBsAg compared with LAM and low pretreatment levels are significantly associated with HBsAg serocon-version. Serial HBsAg measurements may be useful for prediction of HBsAg loss and our data suggest that to achieve this, 5.4 years of sustained response to IFN or 10.6 years of effective LAM therapy are probably needed.
BACKGROUND: Quantification of HBsAg in serum may be of clinical importance in predicting HBsAg seroconversion and complete response to treatment. METHODS: Serum HBsAg was quantified by ADVIA Centaur in 63 patients with HBeAg-negative chronic hepatitis B (CHBe-). A total of 42 had received interferon-alpha2b (IFN-alpha2b) (median 12.1 months; 19 sustained responders including 12 HBsAg-seroconvertors; 23 non-sustained responders) and 21 were on lamivudine (LAM) (median 33.0 months). Measurements were done at baseline, during and at the end of treatment, and during and at the end of follow-up. RESULTS: Baseline median [interquartile range (IQR)] HBsAg levels in all patients were 3286 (1602-7458) IU/ml, not different between IFN- and LAM-treated (P = 0.139). IFN significantly depressed HBsAg in all patients except IFN non-responders, but HBsAg decline persisted only in sustained responders. Low pretreatment HBsAg level was the only significant prognostic variable of HBsAg seroconversion by multivariate analysis. LAM treatment also suppressed HBsAg levels but at a significantly slower rate compared with IFN (P = 0.022). The median (IQR) estimated time to HBsAg undetectability (ETU-HBsAg), derived from best curve fitting, was 127 (87.6-263.5) months for LAM virological responders and 65.3 (36.3-95.0) months for IFN sustained responders (P = 0.002). In 12 HBsAg seroconvertors, ETU-HBsAg was similar to the real time of HBsAg loss (P = 0.525) and seroconversion (0.056). CONCLUSIONS: In CHBe-, IFN induces a sharper decrease in serum HBsAg compared with LAM and low pretreatment levels are significantly associated with HBsAg serocon-version. Serial HBsAg measurements may be useful for prediction of HBsAg loss and our data suggest that to achieve this, 5.4 years of sustained response to IFN or 10.6 years of effective LAM therapy are probably needed.