A longitudinal study on the natural history of serum hepatitis B surface antigen changes in chronic hepatitis B†
Article first published online: 11 JUN 2010
Copyright © 2010 American Association for the Study of Liver Diseases
Volume 52, Issue 4, pages 1232–1241, October 2010
How to Cite
Chan, H. L.-Y., Wong, V. W.-S., Wong, G. L.-H., Tse, C.-H., Chan, H.-Y. and Sung, J. J.-Y. (2010), A longitudinal study on the natural history of serum hepatitis B surface antigen changes in chronic hepatitis B. Hepatology, 52: 1232–1241. doi: 10.1002/hep.23803
Potential conflict of interest: Henry L. Y. Chan is an advisory board member of Hoffmann La-Roche, Abbott Diagnostics, Novartis Pharmaceutical, and Bristol-Myers Squibb.
- Issue published online: 11 JUN 2010
- Article first published online: 11 JUN 2010
- Manuscript Accepted: 3 JUN 2010
- Manuscript Received: 18 JAN 2010
- Direct Grant for Research, The Chinese University of Hong Kong. Grant Number: 2041344
Serum hepatitis B surface antigen (HBsAg) quantification has been suggested to reflect the concentration of covalently closed circular DNA in the liver. We aimed to investigate the HBsAg levels at different stages of chronic hepatitis B and the changes in HBsAg level during the natural progression of disease. One hundred seventeen untreated patients with chronic hepatitis B were studied with longitudinal follow-up for 99 ± 16 months. HBsAg quantification was performed at the first visit, the last visit, and three visits at each quartile during the follow-up. At the first visit, HBsAg level was higher among patients who were hepatitis B e antigen (HBeAg)-positive (N = 49) than those who were HBeAg-negative (N = 68) (4.01 ± 0.91 log IU/mL versus 2.73 ± 1.25 log IU/mL, P < 0.001). HBsAg level was persistently high at approximately 5 log IU/mL among patients in the immune tolerance phase (N = 7). The HBsAg levels among patients with HBeAg-positive active disease (N = 25) or sustained HBeAg seroconversion (N = 17) were comparable at approximately 3-4 log IU/mL. The HBsAg levels among patients who were HBeAg-negative tended to be higher among patients with active (N = 46) than those with inactive disease (N = 22). The median HBsAg levels decreased in HBeAg-negative patients with active and inactive disease by 0.041 log IU/mL/year and 0.043 log IU/mL/year, respectively. Twenty-two (17%) patients had HBsAg reduction >1 log IU/mL at the last visit; most of them showed reduced hepatitis B virus DNA, and eight had HBsAg loss. Conclusion: HBsAg remained stable in HBeAg-positive patients and tended to reduce slowly in HBeAg-negative patients. Reduction of HBsAg for >1 log IU/mL could reflect improved immune control. (HEPATOLOGY 2010)
Chronic hepatitis B virus (HBV) infection is the most common cause of liver cirrhosis and hepatocellular carcinoma in most parts of Asia.1 Patients who have persistently active hepatic necroinflammation and active viremia have a higher risk of disease progression and liver-related complications. Antiviral therapies, including peginterferon and nucleos(t)ide analogues, can suppress viral replication, which can lead to biochemical remission and improvement in liver histology.2 Recent evidence suggests that response to antiviral therapy can be extrapolated to a reduction in occurrence of liver-related complications and hepatocellular carcinoma.3, 4
The current antiviral therapies are not without limitations. Peginterferon is limited by its relatively low response rate (30%-40%) and multiple side effects.5 Nucleos(t)ide analogue treatment is limited by the need for long-term therapy and the emergence of drug resistance. Hence, predictors of treatment response, both before treatment and during treatment, have been investigated to guide the choice and regimen of antiviral therapy.6 One important indicator of viral persistence is covalently closed circular DNA (cccDNA), which serves as the template for viral replication inside hepatocytes.7 Reduction in cccDNA level after antiviral therapy is associated with sustained virologic response.8 Serum hepatitis B surface antigen (HBsAg) quantification has recently been evaluated as a surrogate marker of cccDNA. Good correlations have been found between the absolute levels as well as the changes of serum HBsAg and cccDNA before and after antiviral therapy.9, 10 Furthermore, reduction of serum HBsAg during and after peginterferon therapy has good predictive values for response to peginterferon therapy.11, 12
Nonetheless, controversies exist on the interpretation of serum HBsAg quantification in clinical practice. The relevance of a particular HBsAg level in the context of viral activity and disease progression is uncertain. Although a 0.5 log or 1.0 log reduction of HBsAg has been proposed as on-treatment predictors of response to peginterferon, the magnitude of HBsAg fluctuation in untreated individuals has not been evaluated.11 Before a particular HBsAg level or a magnitude of change in HBsAg level can be recommended to predict treatment outcome, one must first understand the meaning of these parameters in the treatment-free setting. In this study, based on a cohort of untreated patients with chronic hepatitis B with long-term follow-up, we aimed to investigate the HBsAg levels at different stages of chronic hepatitis B. We also aimed to investigate the changes in HBsAg level during the natural progression of disease. Our results would provide important information on the meaning of serum HBsAg quantification in relation to the natural history of chronic hepatitis B.
Patients and Methods
One hundred seventeen patients with chronic hepatitis B in the cohort recruited since 1997 with longitudinal follow-up in the outpatient clinic, Prince of Wales Hospital, Hong Kong, China, were studied.13 All patients received no antiviral therapy during the entire follow-up period. Coinfection by hepatitis C virus was excluded. These patients were followed at an interval of 6 months, or more frequently as clinically indicated. Hepatitis B e antigen (HBeAg) and antibody, liver biochemistry, and alfa-fetoprotein were monitored at every visit. Residual serum samples were stored at −80°C for HBV DNA and HBsAg quantification.
Patients were classified into five groups for analysis. Group 1 consisted of patients with persistently positive HBeAg and normal alanine aminotransferase (ALT), i.e., patients in the immune tolerance phase. Group 2 consisted of patients who had positive HBeAg at the first visit and persistently positive or fluctuating HBeAg during follow-up with intermittent elevation of ALT levels, i.e., unsuccessful immune clearance. Group 3 consisted of HBeAg-positive patients who underwent sustained HBeAg seroconversion. Group 4 consisted of HBeAg-negative patients with intermittent elevation of ALT levels and/or HBV DNA > 2000 IU/mL, i.e., HBeAg-negative active chronic hepatitis B. Group 5 consisted of patients who were HBeAg-negative with persistently normal ALT with HBV DNA ≤ 2000 IU/mL throughout the entire follow-up, i.e., HBeAg-negative inactive chronic hepatitis B.
For Groups 1, 2, 4, and 5, HBV DNA and HBsAg were measured at the first visit, the last visit, and visits at each quartile during the follow-up (second, third, and fourth visit). For Group 3 patients, the second visit was taken as the visit immediately before HBeAg seroconversion, the third visit as the visit immediately after HBeAg seroconversion, and the fourth visit as the visit midway between the third and last visits. Some of the patients had liver stiffness measurement assessed by transient elastography in our previous study and the last visit was taken as the day of transient elastography.14 Otherwise, the last visit was taken as the last follow-up visit with serum sample available.
Detailed analyses were also performed to investigate the changes of HBsAg levels at the time of HBeAg seroconversion and hepatitis flare. For Group 3 patients, the HBsAg levels immediately before and after HBeAg seroconversion were measured and compared. Hepatitis flare was defined as an abrupt elevation of ALT to > 200 IU/L or >3 times the baseline level, whichever was higher.15 HBsAg levels were determined in the available serum samples at the visits immediately before the flare, at the flare, and immediately after the flare.
Quantitative HBsAg Assay.
HBsAg was quantified by Architect HBsAg QT (Abbott Diagnostic, Germany) according to the manufacturer instructions.16 The sensitivity of Architect assay ranged from 0.05 to 250 IU/mL. Samples with HBsAg titer higher than 250 IU/mL were diluted to 1:500 to 1:1000 to bring the reading within the range of the calibration curve.
HBV DNA Assay.
HBV DNA was quantified by TaqMan real-time polymerase chain reaction assay as described.17 This assay was standardized by serial dilution of EUROHEP genotype D HBV standard, which contained 2.7 × 109 viral copies per mL and was validated by the World Health Organization HBV standard. The range of HBV DNA detection was from 102 to 109 copies/mL with correlation coefficient of the standard curve routinely greater than 0.990. For samples with HBV DNA > 109 copies/mL, HBV DNA assay would be repeated after dilution to 1:100. In this assay, 4.86 copies/mL equaled 1 IU/mL.
HBV genotyping was determined by restriction fragment length polymorphism and was confirmed by direct sequencing in case of doubt in the residual serum sample at initial visit, as described.18
Statistical analysis was performed by SPSS (version 15.0; SPSS, Inc., Chicago, IL). Continuous variables were expressed as mean ± standard deviation or median (range) as appropriate. HBV DNA (IU/mL) and HBsAg (IU/mL) were logarithmically transformed for analysis. For patients with undetectable HBV DNA and negative HBsAg, the results were taken as the lower limit of detection (20.6 IU/mL for HBV DNA and 0.05 IU/mL for HBsAg) for calculation. Ratio of HBsAg (log IU/mL) to HBV DNA (log IU/mL) was determined to reflect the proportion of subviral particles to virions. Continuous variables including HBV DNA and HBsAg were compared by Student t test or Mann-Whitney U test as appropriate. Wilcoxon signed rank test was used to compare the reduction in log HBV DNA and log HBsAg levels during the longitudinal follow-up. The annual decline of HBsAg was computed by dividing the HBsAg decline from the first to the last follow-up visit by the total duration of follow-up. Categorical variables were compared by Pearson's chi-squared test or Fisher's exact test as appropriate. Pearson's correlation coefficient was used to assess the correlation between HBV DNA and HBsAg. All statistical tests were two-sided. Statistical significance was taken as P < 0.05.
HBsAg and HBeAg Status.
Overall, 49 patients had positive HBeAg and 68 patients had negative HBeAg at the first visit (Table 1). HBeAg-negative patients were generally older than HBeAg-positive patients. All patients had compensated disease and only four patients had ultrasonic features of liver cirrhosis. HBeAg-positive patients had higher HBV DNA and ALT levels than HBeAg-negative patients. There was a relative predominance of genotype C HBV infection among HBeAg-positive patients. The mean duration of follow-up was 99 months in both HBeAg-positive patients (median 106, range 47-127 months) and HBeAg-negative patients (median 102, range 47-117 months).
|Age (years)||36 ± 13||28 ± 10||42 ± 11||<0.001|
|Male||73 (62%)||31 (63%)||42 (62%)||0.87|
|Hemoglobin (g/dL)||14.6 ± 1.4||14.7 ± 1.4||14.4 ± 1.4||0.33|
|White cell count (×109/L)||6.9 ± 1.5||7.1 ± 1.6||6.7 ± 1.4||0.23|
|Platelet (×109/L)||195 ± 50||198 ± 50||193 ± 51||0.57|
|Creatinine (mmol/L)||79 ± 17||77 ± 19||81 ± 16||0.24|
|Albumin (g/L)||40 ± 3||40 ± 3||40 ± 3||0.46|
|Bilirubin (μmol/L)||9 ± 6||9 ± 5||9 ± 6||0.81|
|Alanine aminotransferase (IU/L)||47 (14-447)||64 (17-447)||43 (14-326)||0.009|
|Cirrhosis by ultrasound (%)||4 (3%)||2 (4%)||2 (3%)||1.00|
|Genotype C HBV (%)||66 (58%)*||35 (71%)||31 (48%)*||0.011|
|HBV DNA (log IU/mL)||5.31 ± 2.26||7.22 ± 1.26||3.94 ± 1.77||<0.001|
|HBsAg (log IU/mL)||3.27 ± 1.28||4.01 ± 0.91||2.73 ± 1.25||<0.001|
|HBsAg/HBV DNA||0.70 ± 0.59||0.56 ± 0.10||0.80 ± 0.59||0.001|
|Follow-up (months)||99 ± 16||99 ± 20||98 ± 14||0.75|
|Number of visits||17 (8-49)||17 (8-37)||17 (8-49)||0.17|
At the first visit, HBeAg-positive patients had higher serum HBsAg levels than HBeAg-negative patients (Table 1). All HBeAg-positive patients had HBsAg > 1 log IU/mL, 41 (84%) patients had HBsAg > 3 log IU/mL, and 28 (57%) patients had HBsAg > 4 log IU/mL. On the other hand, 60 (88%) of the HBeAg-negative patients had HBsAg > 1 log IU/mL, 40 (59%) patients had HBsAg > 3 log IU/mL, and only three (4%) patients had HBsAg > 4 log IU/mL. HBeAg-positive patients also had lower HBsAg/HBV DNA than HBeAg-negative patients (Table 1).
Longitudinal Follow-Up of HBeAg-Positive Patients.
There was no statistical difference in the age, sex ratio, and HBV genotypes among HBeAg-positive patients with persistently normal ALT (Group 1, N = 7), elevated ALT (Group 2, N = 25), and sustained HBeAg seroconversion (Group 3, N = 17). At the first visit, Group 1 patients had higher HBV DNA and HBsAg levels than those in the other two groups. There was no difference in HBV DNA and HBsAg levels between patients in Group 2 and Group 3.
Patients in Group 1 had the highest HBV DNA and HBsAg levels among all five groups of patients throughout the entire follow-up period (Table 2). The mean HBV DNA level of Group 1 patients stayed at approximately 8 log IU/mL and HBsAg at approximately 5 log IU/mL at all time points of assessment (Fig. 1A). The median reduction in HBsAg per year was −0.006 (range −0.02 to 0.03) log IU/mL. Patients in Group 2 and Group 3 had lower HBV DNA and HBsAg levels than Group 1 patients throughout the follow-up, but there was no difference in the HBsAg levels between Group 2 and Group 3 at all time points of assessment (Table 2). The mean HBV DNA of patients in Group 2 stayed at approximately 6-7 log IU/mL and HBsAg at approximately 3-4 log IU/mL at all time points of assessment (Fig. 1A). Although there was some fluctuation in serum HBsAg levels during the follow-up, the median reduction of HBsAg per year was 0.021 (range −0.21 to 0.19) log IU/mL. Twenty-two (88%) patients in Group 2 had HBsAg reduction less than 1 log IU/mL at the last visit (Fig. 2). The age at HBeAg seroconversion among patients in Group 3 was 29 ± 9 (range 13-47) years. Among patients in Group 3, the decline of HBV DNA was most dramatic at the time of HBeAg seroconversion (6.27 ± 1.31 log IU/mL to 3.69 ± 1.26 log IU/mL; P < 0.001; Table 2). On the other hand, there was no obvious decline in HBsAg (3.58 ± 0.53 log IU/mL to 3.50 ± 0.62 log IU/mL; P = 0.64) at the time of HBeAg seroconversion after an interval of 6.5 ± 4.4 months (Fig. 1B). The ratio of HBsAg/HBV DNA increased significantly at the time of HBeAg seroconversion (0.60 ± 1.18 to 1.07 ± 0.46; P < 0.001). Six of 17 (65%) patients in Group 3 had HBsAg reduction more than 1 log IU/mL at the last visit as compared to the first visit (Fig. 2). There was no correlation between HBsAg level and age of HBeAg seroconversion (r = −0.087, P = 0.74).
|N||Group 1||Group 2||Group 3||P (Group 1 vs Group 2)||P (Group 1 vs Group 3)||P (Group 2 vs Group 3)|
|HBV DNA (log IU/mL)||8.46 ± 0.33||6.86 ± 1.26||7.23 ± 1.22||<0.001||0.001||0.34|
|HBsAg (log IU/mL)||4.97 ± 0.16||3.78 ± 0.98||3.95 ± 0.73||<0.001||<0.001||0.54|
|HBsAg/HBV DNA||0.59 ± 0.02||0.55 ± 0.14||0.55 ± 0.06||0.52||0.023||0.87|
|Follow-up (months)||31 (17-43)||28 (2-63)||15 (3-56)||0.41||0.081||0.026|
|HbeAg-positive (%)||7 (100%)||23 (92%)||17 (100%)||1.00||1.00||1.00|
|HBV DNA (log IU/mL)||7.58 ± 2.11||6.84 ± 1.54||6.27 ± 1.31||0.31||0.075||0.21|
|HBsAg (log IU/mL)||4.92 ± 0.14||3.69 ± 1.06||3.58 ± 0.53||0.005||<0.001||0.68|
|HBsAg/HBV DNA||0.75 ± 0.43||0.56 ± 0.22||0.60 ± 0.18||0.11||0.23||0.49|
|Follow-up (months)||61 (34-74)||57 (27-81)||23 (5-84)||0.37||0.007||0.001|
|HbeAg-positive (%)||7 (100%)||18 (72%)||0 (0%)||0.29||<0.001||<0.001|
|HBV DNA (log IU/mL)||8.33 ± 0.31||6.55 ± 1.48||3.69 ± 1.26||<0.001||<0.001||<0.001|
|HBsAg (log IU/mL)||5.01 ± 0.20||3.79 ± 0.94||3.50 ± 0.62||<0.001||<0.001||0.23|
|HBsAg/HBV DNA||0.60 ± 0.02||0.59 ± 0.12||1.07 ± 0.46||0.77||0.001||0.001|
|Follow-up (months)||88 (54-99)||78 (38-110)||66 (33-91)||0.37||0.02||0.025|
|HbeAg-positive (%)||7 (100%)||24 (96%)||0 (0%)||1.00||<0.001||<0.001|
|HBV DNA (log IU/mL)||7.55 ± 1.77||6.27 ± 1.70||3.33 ± 1.60||0.091||<0.001||<0.001|
|HBsAg (log IU/mL)||4.92 ± 0.18||3.68 ± 0.77||3.36 ± 0.80||<0.001||<0.001||0.19|
|HBsAg/HBV DNA||0.71 ± 0.29||0.64 ± 0.28||1.23 ± 0.64||0.56||0.011||0.002|
|Follow-up (months)||109 (67-122)||103 (53-127)||106 (47-116)||0.47||0.43||0.50|
|HbeAg-positive (%)||7 (100%)||17 (68%)||0 (0%)||0.15||<0.001||<0.001|
|HBV DNA (log IU/mL)||8.18 ± 0.35||6.32 ± 1.37||3.33 ± 1.30||<0.001||<0.001||<0.001|
|HBsAg (log IU/mL)||4.99 ± 0.18||3.63 ± 0.70||3.19 ± 1.29||<0.001||<0.001||0.16|
|HBsAg/HBV DNA||0.61 ± 0.02||0.59 ± 0.15||0.99 ± 0.63||0.78||0.024||0.02|
Longitudinal Follow-Up of HBeAg-Negative Patients.
There was no difference in the age, sex ratio and HBV genotypes between HBeAg-negative patients with active disease (Group 4, N = 46) and those with inactive disease (Group 5, N = 22). The HBV DNA levels among patients in Group 4 (approximately 4-5 log IU/mL) were persistently higher than that among patients in Group 5 (approximately 2 log IU/mL) throughout the follow-up (Table 3). There was a trend of higher HBsAg levels among Group 4 patients than Group 5 patients in the first visit, and the difference in HBsAg levels between the two groups became more evident during subsequent follow-up (Table 3, Fig. 1A). There was a tendency of decline in HBsAg levels during the follow-up (Fig. 1A). The median reduction of HBsAg from the first to the last visit in Group 4 was 0.29 (range −2.26 to 4.83) log IU/mL (P = 0.009 versus the first visit) and the decline per year was 0.041 (range −0.26 to 0.76) log IU/mL. The median reduction of HBsAg in Group 5 was 0.32 (range −0.19 to 4.90) log IU/mL until the last visit (P = 0.006 versus the first visit) and the decline per year was 0.043 (range −0.020 to 0.53) log IU/mL. Six of 40 (13%) patients in Group 4 and 7 of 15 (32%) patients in Group 5 had HBsAg reduction more than 1 log IU/mL until the last visit (P = 0.098; Fig. 2). The ratio of HBsAg/HBV DNA was comparable between the two groups at all time points of assessment, and there was no significant change in this ratio during the course of follow-up in either group of patients (Table 3).
|N||Group 4||Group 5||P|
|HBV DNA (log IU/mL)||4.68 ± 1.63||2.39 ± 0.74||<0.001|
|HBsAg (log IU/mL)||2.98 ± 0.88||2.24 ± 1.61||0.054|
|HBsAg/HBV DNA||0.73 ± 0.43||0.96 ± 0.81||0.23|
|Follow-up (months)||24.5 (7-42)||28 (10-38)||0.32|
|HBV DNA (log IU/mL)||4.62 ± 1.56||2.54 ± 0.96||<0.001|
|HBsAg (log IU/mL)||2.88 ± 0.87||2.02 ± 1.70||0.037|
|HBsAg/HBV DNA||0.68 ± 0.37||0.80 ± 0.84||0.54|
|Follow-up (months)||52 (28-71)||55.5 (22-66)||0.65|
|HBV DNA (log IU/mL)||4.291 ± 1.57||1.92 ± 0.65||<0.001|
|HBsAg (log IU/mL)||2.80 ± 0.96||1.83 ± 1.78||0.023|
|HBsAg/HBV DNA||0.75 ± 0.59||0.90 ± 1.01||0.52|
|Follow-up (months)||78 (52-96)||81 (35-92)||0.40|
|HBV DNA (log IU/mL)||4.12 ± 1.48||2.09 ± 0.84||<0.001|
|HBsAg (log IU/mL)||2.65 ± 1.02||1.52 ± 2.00||0.020|
|HBsAg/HBV DNA||0.71 ± 0.52||0.66 ± 1.17||0.87|
|Follow-up (months)||101.5 (74-117)||105 (47-116)||0.18|
|HBV DNA (log IU/mL)||3.86 ± 1.51||1.94 ± 0.64||<0.001|
|HBsAg (log IU/mL)||2.56 ± 1.14||1.47 ± 2.23||0.026|
|HBsAg/HBV DNA||0.74 ± 0.60||0.63 ± 1.19||0.69|
Pooling the HBsAg levels of all HBeAg-negative patients at five visits, the area under the receiver operating characteristic curve for HBsAg to differentiate patients with active (Group 4) and inactive (Group 5) was 0.63 (95% confidence interval [CI] = 0.56, 0.70; P < 0.001). A cutoff HBsAg at 1.5 log IU/mL has maximum sum of sensitivity and specificity. HBsAg > 1.5 log IU/mL has a sensitivity of 93% (95% CI = 90%, 95%), specificity of 40% (95% CI = 34%, 45%), positive predictive value of 76% (95% CI = 74%, 78%) and negative predictive value of 72% (95% CI = 61%, 81%) for active disease in HBeAg-negative chronic hepatitis B. No HBsAg cutoff can accurately exclude HBeAg-negative active hepatitis.
Clinical Features of Patients With Significant HBsAg Reduction.
Overall, 22 (17%) patients had HBsAg reduction by greater than 1 log IU/mL from the first to the last visit. Among them, 10 patients had HBsAg reduction by greater than 1 log IU/mL at the second visit (Fig. 2). The baseline HBV DNA, HBsAg, ratio of HBsAg/HBV DNA, and ALT levels had no relationship with chance of greater HBsAg reduction at subsequent follow-up (Table 4). In general, an HBsAg reduction of greater than 1 log IU/mL could reflect a better viral control. The reduction in HBV DNA among patients with greater HBsAg reduction (median 2.30, range 0.04-6.59, log IU/mL) was more dramatic than that among patients with HBsAg reduction <1 log IU/mL (median 0.58, range −3.50 to 5.53, log IU/mL; P < 0.001). Patients with greater HBsAg reduction also tend to have lower HBsAg/HBV DNA from the second to the last visit. Three patients (two in Group 2 and one in Group 4) developed hepatocellular carcinoma; all of them had HBsAg reduction <1 log IU/mL.
|N||HBsAg reduction > 1 log IU/mL||HBsAg reduction ≤ 1 log IU/mL||P|
|HBeAg-positive (%)||9 (41%)||40 (42%)||0.92|
|HBsAg (log IU/mL)||3.12 ± 1.60||3.31 ± 1.17||0.60|
|HBV DNA (log IU/mL)||5.57 ± 2.39||5.25 ± 2.24||0.56|
|HBsAg/HBV DNA||0.56 ± 0.29||0.73 ± 0.49||0.13|
|Alanine aminotransferase (IU/L)||50 (19-230)||52 (14-447)||0.70|
|HBeAg-positive (%)||9 (41%)||38 (40%)||0.94|
|HBV DNA (log IU/mL)||4.79 ± 2.38||5.20 ± 2.10||0.43|
|Reduction in HBV DNA (log IU/mL)||0.46 (-0.28 – 3.57)||−0.05 (−2.92 – 5.31)||0.01|
|HBsAg/HBV DNA||0.46 ± 0.29||0.72 ± 0.48||0.015|
|Alanine aminotransferase (IU/l)||50.5 (18-410)||48 (13-738)||0.49|
|HBeAg-positive (%)||3 (13%)||22 (23%)||0.33|
|HBV DNA (log IU/mL)||3.88 ± 2.36||4.62 ± 2.18||0.16|
|Reduction in HBV DNA (log IU/mL)||1.43 (−0.20 – 5.16)||0.26 (−4.03 – 5.32)||0.003|
|HBsAg/HBV DNA||0.40 ± 0.53||0.87 ± 0.60||0.001|
|Alanine aminotransferase (IU/L)||42.5 (16-171)||39 (13-699)||0.85|
|HBeAg (%)||3 (13%)||28 (30%)||0.13|
|HBV DNA (log IU/mL)||3.09 ± 2.10||4.57 ± 2.08||0.003|
|Reduction in HBV DNA (log IU/mL)||1.65 (0.10 to 6.83)||0.31 (–3.29 to 5.91)||<0.001|
|HBsAg/HBV DNA||0.38 ± 0.88||0.85 ± 0.60||0.027|
|Alanine aminotransferase (IU/L)||35.5 (12-121)||37 (11-219)||0.21|
|HBeAg (%)||2 (9%)||22 (23%)||0.14|
|HBV DNA (log IU/mL)||2.69 ± 1.59||4.56 ± 2.12||<0.001|
|Reduction in HBV DNA (log IU/mL)||2.30 (0.04 to 6.59)||0.58 (–3.50 to 5.53)||<0.001|
|HBsAg/HBV DNA||0.14 ± 0.87||0.85 ± 0.56||0.001|
|Alanine aminotransferase (IU/L)||29 (10-712)||34 (10-180)||0.054|
A good immune control was most obvious among patients who were negative for HBeAg while having an HBsAg reduction of >1 log IU/mL (Table 5). There was a trend, though not statistically significant, of higher proportion of low HBV DNA and HBsAg loss among patients who underwent HBeAg seroconversion (Group 3) if they had HBsAg reduction by >1 log IU/mL. Among HbeAg-negative patients (Group 4 and Group 5), HBsAg reduction >1 log IU/mL was associated with higher chance of HBsAg loss and better viral suppression. Among nine patients who experienced HBsAg loss, eight had HBsAg reduction by >1 log IU/mL at the last visit. The remaining patient had HBsAg level lower than 0.5 IU/mL at the first visit and had undetectable HBsAg (<0.05 IU/mL) starting the second visit.
|HBsAg reduction > 1 log IU/mL||HBsAg reduction ≤ 1 log IU/mL||P|
|HBV DNA ≤ 2000 log IU/mL||0 (0%)||1 (5%)||1.00|
|HBsAg loss||0 (0%)||0 (0%)||1.00|
|HBV DNA ≤ 2000 log IU/mL||4 (67%)||5 (45%)||0.62|
|HBsAg loss||1 (17%)||0 (0%)||0.35|
|HBV DNA ≤ 2000 log IU/mL||5 (83%)||9 (23%)||0.007|
|HBsAg loss||2 (33%)||0 (0%)||0.014|
|HBV DNA ≤ 2000 log IU/mL||7 (100%)||15 (100%)||1.00|
|HBsAg loss||5 (71%)||1 (7%)||0.004|
Change of HBsAg at Hepatitis Flare.
Twenty patients had a total of 27 hepatitis flares (12 HBeAg-positive, 14 HBeAg-negative, and one with absent HBeAg result) during the follow-up period. The ALT levels at the visits before flare, during flare, and after flare were 53 (24-185) IU/L, 330 (214-1066) IU/L, and 46 (24-128) IU/L, respectively. There was a significant increase in HBV DNA level at hepatitis flare (from 5.27 ± 1.89 log IU/mL to 6.79 ± 1.09 log IU/mL; P < 0.001) but the HBsAg level remained relatively static (from 3.32 ± 1.17 log IU/mL to 3.28 ± 0.98 log IU/mL; P = 0.72) (Fig. 1C). The HBsAg/HBV DNA ratio decreased from 0.68 ± 0.34 before flare to 0.48 ± 0.14 at ALT flare (P < 0.001) and increased back to 0.74 ± 0.33 after flare (P < 0.001).
Correlation of HBV DNA and HBsAg.
Overall, with the 585 samples from 49 HBeAg-positive and 68 HBeAg-negative patients at five time points of assessment, the HBsAg levels had moderate correlation with HBV DNA (r = 0.61, P < 0.001). The correlation of HBsAg with HBV DNA was higher in the 176 HBeAg-positive samples (r = 0.66, P < 0.001) than that in the 409 HBeAg-negative samples (r = 0.41, P < 0.001; Fig. 3).
Based on an 8-year follow-up of patients at different stages of chronic HBV infection, we found that serum HBsAg quantification was closely related to the HBeAg status and HBV DNA levels. Patients in immune tolerance phase had highest HBsAg levels at approximately 5 log IU/mL. After HBeAg seroconversion, the HBsAg level reduces progressively with time. HBeAg-negative patients who have inactive disease tend to have lower HBsAg levels than those who have active disease, but no cutoff value can confidently predict the disease activity due to significant overlap of HBsAg levels among patients with active and inactive diseases. HBsAg tend to fluctuate within a range of 1 log IU/mL in most (83%) patients throughout the follow-up period. A reduction of HBsAg for greater than 1 log IU/mL usually indicated improved immune control of the virus, particularly among patients who underwent HBeAg seroconversion or were negative for HBeAg.
The natural history of chronic hepatitis B is usually defined by the HBeAg status, HBV DNA, and ALT levels.19 In our longitudinal cohort, patients in the immune tolerance phase (Group 1) had persistently highest HBV DNA levels and positive HBeAg. On immune clearance, patients who failed to undergo HBeAg seroconversion had persistently elevated HBV DNA (Group 2), whereas those who had successful HBeAg seroconversion had significant reduction in viral load (Group 3). HBeAg-negative patients tend to have persistently higher HBV DNA among patients with active disease (Group 4) than those who have inactive disease (Group 5). One important finding of our study was that the difference in HBsAg levels was much less apparent among different stages of chronic HBV infection. At HBeAg seroconversion, a rapid reduction of HBV DNA was not accompanied by a significant change in HBsAg level. During ALT flare, the biochemical activity was largely related to increased viral replication but could hardly influence the HBsAg production. HBsAg is composed of both Dane particle, which contains the viral genome, and subviral particles. The mechanisms that regulate the production of HBsAg, particularly the subviral particles, are largely unclear. In a study investigating the serum HBsAg and the intrahepatic preS/S HBV RNA normalized for cccDNA between HBeAg-positive and HBeAg-negative patients, no relationship between the efficiency of HBsAg production and the HBeAg status could be observed.20 One potential bias of this study was that only patients with elevated ALT levels were recruited for liver biopsy. Nonetheless, together with our findings, it highlighted that the regulation of HBsAg production was a more complicated process than host immune viral clearance.
The correlation of HBV DNA and HBsAg was an interesting observation. A better correlation of HBV DNA and HBsAg was found in the HBeAg-positive phase than the HBeAg-negative phase. The HBsAg/HBV DNA ratio was very stable in HBeAg-positive patients (Group 1 and 2) and before HBeAg seroconversion (Group 3). It might reflect a very stable relationship between viral replication and HBsAg production (and possibly the amount of cccDNA) before a successful immune clearance.9, 10 The HBsAg/HBV DNA ratio increased dramatically with a relatively unchanged HBsAg level immediately after HBeAg seroconversion. Apparently, immune control over viral replication was the first step of immune clearance. In HBeAg-negative patients, the HBsAg/HBV DNA ratio fell slowly, particularly among patients with inactive disease, due to the progressive reduction of HBsAg level. It might indicate clearance of cccDNA and some patients eventually developed HBsAg loss.
In the current study, HBsAg levels were generally lower among patients who had negative HBeAg. In other words, a lower HBsAg level, to a certain extent, could reflect the tendency of immune clearance of HBeAg. However, because of the significant overlap in the HBsAg levels among patients with different disease stages, we could not identify any cutoff HBsAg level that could reliably differentiate patients with active or inactive hepatitis or those who will or will not spontaneously clear HBeAg. Among HBeAg-negative patients, an HBsAg cutoff of 1.5 log IU/mL (approximately 32 IU/mL) has a high sensitivity (93%) for active disease. In other words, HBeAg-negative patients who have higher HBsAg levels were more likely having active disease. On the other hand, both patients with active and inactive HBeAg-negative disease could have HBsAg level below 1.5 log IU/mL, and no HBsAg cutoff could offer a high specificity to confirm active disease.
Reduction of serum HBsAg level has been repeatedly shown to have good correlation with the reduction in cccDNA during antiviral therapy.9, 10 Based on our current natural history study, an 1 log IU/mL reduction of HBsAg rarely occurred spontaneously even after a follow-up of over 8 years. Among patients who had greater than 1 log IU/mL reduction in HBsAg, most of them showed improved immune control over the viral replication and normalization of ALT levels. In a Greek study among HBeAg-negative patients being treated with interferon alfa-2b versus lamivudine, rapid HBsAg reduction could only be observed among patients who received interferon treatment particularly among the sustained responders.21 Although studies with different designs from different ethnic groups should be interpreted with caution, a rapid reduction of HBsAg was likely signifying an enhanced immune viral clearance.
Our study has a few limitations. Although we have a long follow-up with serial samplings, we could not preclude the possibility of missing some elevated ALT episodes among patients in Group 1 and Group 5, leading to misclassification of the disease categories. Furthermore, because liver biopsy was not a routine procedure, we did not have histologic proof on the disease status. This potential error could partly be compensated by the availability of serial data on HBV DNA. The persistently high HBV DNA in Group 1 patients would support an immune tolerance state.22 The persistently low HBV DNA in Group 5 patients would support a low replicative state.23 Another limitation was the relative small number of patients in this study. With the rapid development of antiviral drugs in the last decade, many patients in the original cohort who had active liver disease were recruited into clinical trials or started on antiviral therapy.24-27 Because of a very restricted reimbursement policy in Hong Kong, some patients who had active hepatitis in the original cohort but remained untreated could be recruited in the current study.28
In summary, we have demonstrated the natural course of serum HBsAg changes in different stages of chronic HBV infection in this longitudinal study. HBsAg levels tend to be very stable in the HBeAg-positive phase and decreases slowly in the HBeAg-negative phase if the disease was untreated. No single HBsAg level could accurately predict the disease activity or viral clearance. On the other hand, a reduction of HBsAg greater than 1 log IU/mL seemed to indicate improved immune viral control. Future studies should be conducted to evaluate the use of HBsAg reduction as an on-treatment predictor of response, particularly with interferon-based therapy in which immune enhancement is the key mechanism of viral clearance.
- 18Viral genotype and hepatitis B virus DNA levels are correlated with histological liver damage in HBeAg-negative chronic hepatitis B virus infection. Am J Gastroenterol 2002; 97: 406-412., , , , , , et al.Direct Link:
- 23Evaluation of alanine transaminase and hepatitis B virus DNA to predict liver cirrhosis in hepatitis B e antigen-negative chronic hepatitis B using transient elastography. Am J Gastroenterol 2008; 103: 3071-3081., , , , , , et al.Direct Link: