Research Article

Hepatitis B surface antigen levels during the natural historyof chronic hepatitis B: A perspective on Asia

Tin Nguyen1,2, Alexander J.V. Thompson1,2, Scott Bowden1, Catherine Croagh2, Sally Bell2,Paul V. Desmond2, Miriam Levy3, Stephen A. Locarnini1,*

1Department of Research and Molecular Development, Victorian Infectious Diseases Reference Laboratory, North Melbourne, Vic., Australia;2Department of Gastroenterology, St. Vincent’s Hospital, Fitzroy, Vic., Australia; 3Department of Gastroenterology and Infectious Diseases,

Liverpool Hospital, NSW, Australia

See Editorial, pages 475–477

Background & Aims: Data from clinical trials suggest a potential rithms for both immune-modulator therapy and oral nucle-

role for on-treatment monitoring of serum HBsAg titres duringinterferon-alpha (pegIFN) therapy in predicting virologicalresponses. However, baseline HBsAg titres during the natural his-tory of chronic hepatitis B (CHB) have not been well-character-ized. We aimed to define the serum HBsAg titres during thedifferent phases of CHB in a cohort of Asian patients infected witheither genotype B or C HBV.Methods: Two-hundred and twenty patients were classified intoimmune-tolerant (IT), immune-clearance (IC), non/low-replicative(LR) or hepatitis B e antigen negative hepatitis (ENH) phases. SerumHBsAg was quantified using the ARCHITECT platform (Abbott Lab-oratories, Chicago, USA). Correlation of HBsAg titre with HBV DNAand serum ALT within each phase of infection was performed.Results: Median HBsAg titres were different between each phaseof CHB (p = 0.001): IT (4.53 log10 IU/ml), IC (4.03 log10 IU/ml), LR(2.86 log10 IU/ml), and ENH (3.35 log10 IU/ml). HBsAg titres werehighest in the IT phase, and lowest in the LR phase. In general,median HBsAg titres were similar between genotypes B and CHBV. Serum HBsAg titres only correlated with HBV viral load inthe IC phase. No correlation between the serum HBsAg leveland ALT was observed.Conclusions: This study demonstrated significant differences inmedian baseline serum HBsAg titres across the different phasesof CHB. These results provide further insight into the HBV virallife cycle in the setting of the various phases of CHB. BaselineHBsAg quantification may help refine future treatment algo-

Journal of Hepatology 20

Keywords: Chronic hepatitis B; Hepatitis B virus; HBV DNA; HBsAg; HBeAg.Received 30 August 2009; received in revised form 20 October 2009; accepted 22October 2009; available online 16 February 2010* Corresponding author. Address: Victorian Infectious Diseases Reference Labo-ratory (VIDRL), 10 Wreckyn St, North Melbourne, Vic. 3051, Australia. Tel.: +61 39342 2637; fax: +61 3 9342 2666.E-mail address: stephen.locarnini@mh.org.au (S.A. Locarnini).Abbreviations: CHB, chronic hepatitis B; IT, immune-tolerant; IC, immune-clear-ance; LR, low-replicative; ENH, e antigen negative hepatitis; HBeAg, hepatitis B eantigen; HBV DNA, hepatitis B virus DNA; cccDNA, covalently closed circularDNA; HBsAg, hepatitis B surface antigen; HBV, hepatitis B virus; HIV, humanimmunodeficiency virus; ULN, upper limit normal.

os(t)ide analogue therapy.� 2010 European Association for the Study of the Liver. Publishedby Elsevier B.V. All rights reserved.

Introduction

The natural history of chronic hepatitis B (CHB) is typicallyregarded as consisting of four phases [1,2]; immune-tolerant(IT), immune-clearance (IC), non/low-replicative (LR), and hep-atitis B e antigen negative hepatitis (ENH). These phases havebeen classified by specific biochemical, serological and virolog-ical characteristics, including serum ALT levels, hepatitis B eantigen (HBeAg) serostatus, and hepatitis B virus DNA (HBVDNA) titre. It is important to note that these phases do notoccur in all individuals, and do not always necessarily occursequentially [3].

The understanding of the pathogenesis and natural historyof CHB continues to evolve. This has been facilitated by theimproved sensitivity of HBV DNA viral load assays, and thedevelopment of assays for the detection and measurement ofHBV intrahepatic replicating forms of the virus such as cova-lently closed circular DNA (cccDNA) and other replicative inter-mediates [4,5]. Recently, sensitive and reliable assays have alsobeen developed to quantify both serum hepatitis B surfaceantigen (HBsAg) and HBeAg.

HBsAg seroclearance represents the preferred endpoint oftherapy for CHB, as it is believed to represent successful immuno-logical control of active HBV replication. Recent data from clinicaltrials suggest a potential role for on-treatment monitoring ofserum HBsAg and HBeAg titres during interferon-alpha (pegIFN)therapy in predicting virological responses [6,7]. However, base-line HBsAg titres have not been well-characterized in each phaseof CHB, particularly in the IT and LR phases. The main objective ofthis study was to determine the serum HBsAg titres during thedifferent phases of the natural history of CHB in a cohort of Asianpatients infected with genotype B or C HBV.

10 vol. 52 j 508–513

Table 1. Baseline population characteristics.

Immune Tolerant (n = 32) Immune Clearance (n = 55) Low Replicative (n = 50) HBeAg negative hepatitis (n = 83) p Value

HBeAg status Positive Positive Negative Negative –Agea (Years) 30 (23, 33) 30 (24, 37) 36 (28, 45) 43 (35, 50) <0.001Sex M/F 9/23 17/38 18/32 48/35 0.002Asian Ethnicity % 100 100 100 100 –HBV DNAa# log10IU/ml 8.22 (7.32, 8.90) 8.02 (5.23, 9.39) <2.55 (<2.55, 3.18) 4.95 (3.41, 6.89) <0.001ALTa# IU/ml 18 (10, 30) 101 (51, 237) 23 (10, 55) 50 (20, 132) <0.001Genotype B/C (%) 65/35 49/51 60/40 61/31 <0.007

a Data expressed as the median (IQR) and as percentages, # 10–90% confidence interval.

JOURNAL OF HEPATOLOGY

Methods

Patients

A cross-sectional study was performed in patients with CHB from two tertiaryhospitals (St. Vincent’s Hospital Melbourne and Liverpool Hospital Sydney, Aus-tralia). Patients included from Liverpool hospital were a cohort of pregnantwomen. Only patients with genotype B or C HBV infection were included. Allpatients tested negative for markers of hepatitis C virus, hepatitis D virus, andhuman immunodeficiency virus (HIV). Markers for co-existent auto-immune ormetabolic liver disease were negative.

Patient demographics, liver biochemistries, qualitative HBsAg and HBeAg sta-tus (by standard qualitative enzyme immunoassay), HBV DNA, and HBV genotype[8] were recorded. Biochemical and virological data were obtained from patientserum samples collected on the same day.

Patients were classified into a phase of CHB at the centers involved in thestudy after a follow-up period of 3–6 months. The phase of CHB in each patientwas determined by HBeAg/anti-HBe serostatus, and measurement of HBV DNAand serum ALT levels according to the recently published European Associationfor the Study of the Liver (EASL) clinical practice guidelines [2]. The IT phasewas defined as: HBeAg positive, high viral load, serum ALT <2� upper limit nor-mal (ULN). The IC phase was defined as: HBeAg positive, elevated viral load,serum ALT >2� ULN. The LR phase was defined as; HBeAg negative, HBV DNA<2000 IU/ml, normal serum ALT. The ENH phase was defined as: HBeAg negative,HBV DNA >2000 IU/ml, serum ALT >2� ULN. For this study, the cut-off for a nor-mal ALT was 630 Units/L.

The study was conducted according to the guidelines of the Declaration ofHelsinki, and was approved by the local institutional ethics research committee.

Quantitative serum HBsAg assay

Serum HBsAg was quantified using the ARCHITECT platform (Abbott Laboratories,Chicago, USA), as per the manufacturer’s instructions. The ARCHITECT quantita-tive HBsAg assay is a chemiluminescent microparticle assay, internally calibratedusing the World Health Organisation (WHO) standard for HBsAg [9]. QuantitativeHBsAg levels are reported in IU/ml, with a dynamic range of 0.05–250 IU/ml.Given that most serum HBsAg titres are above this range, samples were initiallytested at dilutions of 1 in 100 or 1 in 1000. ARCHITECT HBsAg Manual Diluent(Abbott Diagnostics) was used to dilute patient sera. This diluent contains recal-cified human plasma that is non-reactive for HBsAg, HIV-1 RNA or HIV-1 Ag, anti-HIV-1/HIV-2, anti-HCV, or anti-HBs.

HBV DNA viral load

HBV DNA viral load testing was performed using the Versant HBV DNA 3.0 (BayerDiagnostics, Emeryville, CA, USA), according to the manufacturer’s instructions(dynamic range 3.5 � 102 – 1.8 � 107 IU/ml). Samples with a viral load abovethe upper limit of the dynamic range were diluted by 103 to obtain a defined titre.ARCHITECT HBsAg Manual Diluent (Abbott Diagnostics) was used to dilutepatient sera.

HBV genotyping

Genotyping (HBV genotypes A–H) was performed in 200/220 (91%) patients usingmethods as previously described [8]. Genotyping was not possible in 20 patients inthe LR phase due to insufficient HBV DNA. Briefly, HBV DNA was extracted from

Journal of Hepatology 201

200 ll of patient serum using the QIAamp DNA MiniKit (QIAGEN, CA, USA) accord-ing to the manufacturer’s instructions. Oligonucleotides were synthesised byGeneworks, Adelaide, Australia. For amplification of the polymerase gene the senseprimer was 1877a (nt 1877–1996*, CCT GCT GGT GGC TCC AGT TC) and the anti-sense primer 2996 (nt 2996–3014*, GCG TCA GCA AAC ACT TGG C) was used.The amplified HBV envelope/surface gene was purified using PCR purification col-umns from MO BIO Laboratories Inc. and directly sequenced using Big Dye termi-nator Cycle sequencing Ready Reaction Kit Version 3.1 (Perkin-Elmer, CetusNorwalk, CT). HBV consensus sequences were constructed using the DNA sequenceanalysis program Seqscape (Applied Biosystems, USA). HBV genotype was deter-mined using a web-based program, SeqHepB (http://www.seqvirology.com/genome7/index.htm) [10]. This program analyses HBV DNA to determine HBVgenotype, to identify key mutations associated with antiviral resistance as wellas other clinically important HBV variants by comparing the input sequence datawith known HBV reference sequences.

Statistical analysis

Continuous and categorical variables were compared between groups, using theMann–Whitney test and Kruskall–Wallis ANOVA for non-parametric continuousdata, and v2/Fisher’s exact test for categorial data. Pearson’s correlation coeffi-cient (r) was used to describe the correlation between two variables. Statisticalanalysis was performed using GraphPad Software, San Diego California USA,www.graphpad.com.

Results

Two-hundred and twenty treatment naïve patients wererecruited. Patients from Liverpool Hospital, Sydney comprised acohort of pregnant women (n = 70). The classification of patientsinto respective phases of CHB were: IT (n = 32), IC (n = 55), LR(n = 50), ENH (n = 83).

The baseline patient characteristics are presented in Table 1.All patients were of Asian ethnicity and were infected with eithergenotype B or C HBV. There were overall more females in thestudy group (56%). HBeAg positive patients were younger thanHBeAg negative patients (p < 0.001). In HBeAg positive patients,there was no significant age difference between those in theimmune-tolerant and immune-clearance phases. It should benoted that the participating hospitals did not include childrenor adolescents with CHB in this study. Patients in the ENH phasewere older than those in the LR (p = 0.02).

Population distribution of HBsAg titre

The distribution of serum HBsAg levels across the study popu-lation was evaluated (Fig. 1). HBsAg titres were differentbetween each phase of CHB (p = 0.001). The median HBsAgtitres in each phase of CHB were: IT (4.53 log10 IU/ml), IC(4.03 log10 IU/ml), LR (2.86 log10 IU/ml), and ENH (3.35log10 IU/ml), respectively. There was no difference in median

0 vol. 52 j 508–513 509

Immune Tolerant

2.5 3.0 3.5 4.0 4.5 5.0 5.56

7

8

9

10

11 r = 0.30p = 0.09

HB

V D

NA

log 1

0 IU

/ml

Low Replicative

0 2 4 61.5

2.0

2.5

3.0

3.5 r = 0.22p = 0.11

C

Serum HBsAg log10 IU/ml

Serum HBsAg log10 IU/ml

HB

V D

NA

log 1

0 IU

/ml

A

Fig. 2. (A–D) Correlation of serum HBsAg titres

IT IC LR ENH0

1

2

3

4

5

6p = 0.007 p = 0.003

Phase of CHB

HB

sAg

log 1

0 IU

/ml

Fig. 1. Distribution of serum HBsAg titres throughout the natural history ofCHB. Median values with 95% confidence interval (of median) represented. IT,immune-tolerant; IC, immune-clearance; LR, low-replicative; ENH, HBeAg neg-ative hepatitis.

Research Article

510 Journal of Hepatology 201

HBsAg titres between pregnant (n = 14) and non-pregnant(n = 18) immune-tolerant cohorts (p = 0.055). The medianHBsAg in the LR and ENH were significantly lower than bothHBeAg positive phases.

Correlation of serum HBsAg titres with serum ALT levels and HBVviral load

There was no observed statistical correlation between HBsAg titreswith serum ALT levels in any phase of CHB. The correlationbetween serum HBsAg titres with serum HBV DNA in each phaseof CHB is presented in Fig. 2A–D. There was a modest correlationobserved in the IC phase (r = 0.77, p = 0.0001), but poor correlationbetween serum HBsAg and HBV DNA in either the IT (r = 0.30,p = 0.09), LR (r = 0.22, p = 0.11) or ENH (r = 0.29, p = 0.008).

The ratio of HBsAg (log10 IU/ml) to HBV DNA (log10 IU/ml) ineach phase of CHB was also examined (Fig. 3). The HBsAg/HBVDNA ratio was significantly higher in the low-replicative phasecompared to immune-tolerant, immune-clearance and HBeAgnegative phases, respectively (1.05 vs 0.55, 0.55, 0.64.p < 0.0001).

Immune Clearance

1 2 3 4 5 6 72

4

6

8

10

r = 0.77p = 0.0001

BH

BV

DN

A lo

g 10

IU/m

l

Serum HBsAg log10 IU/ml

Serum HBsAg log10 IU/ml

HBeAg negative hepatitis

0 1 2 3 4 50

2

4

6

8

10 r = 0.29p = 0.008

D

HB

V D

NA

log 1

0 IU

/ml

and HBV DNA in different phases of CHB.

0 vol. 52 j 508–513

0.55 0.55 1.05 0.64

IT IC LR ENH0

0.5

1.0

1.5

2.0

2.5 p < 0.0001

HB

sAg

/ HB

VD

NA

(lo

g 10

IU/m

l)

Fig. 3. Ratio of HBsAg and HBVDNA in each phase of CHB. Median values with95% confidence interval (of median) represented. IT, immune-tolerant; IC,immune-clearance; LR, low-replicative; ENH, HBeAg negative hepatitis.

JOURNAL OF HEPATOLOGY

Serum HBsAg titres between genotypes B and C

The distribution of serum HBsAg titres in different phases of CHBwas evaluated based on viral genotype (Fig. 4). In each viral geno-type, median HBsAg titres were clearly different in each phase ofCHB. The median serum HBsAg titre in the LR phase was lower ingenotype B (2.24 log10 IU/ml) compared to genotype C (3.34log10 IU/ml). However, it should be noted that viral genotype wasnot available in all patients in this phase as some patients hadinsufficient HBV DNA to allow for genotype determination bysequencing.

Discussion

This study aimed to evaluate the baseline serum HBsAg titres indifferent phases of CHB infection. All patients were treatmentnaïve and chronically infected with either genotype B or C HBV,and were of Asian ethnicity. This study demonstrates that serumHBsAg titres differ between the four phases of CHB. This findingwas also clearly demonstrated in an accompanying article by Jar-

Genotype B

IT IC

A

LR ENH0

1

2

3

4

5

6

Phase of CHB

HB

sAg

log 1

0 IU

/ml

Fig. 4. Distribution of serum HBsAg titres in (A) genotype B CHB, and (B) ge

Journal of Hepatology 201

oszewicz et al., which examined HBsAg levels in a predominantlyEuropean cohort (HBV genotypes A and D). In our article, HBsAgtitres were higher in HBeAg positive compared to HBeAg negativeCHB. The lowest HBsAg titres were evident in the low-replicativephase. The difference between median lowest and highest HBsAgtitres in the different phases was less than 2 log10 IU/ml, far smal-ler than that observed with changes in HBV DNA levels. Otherthan in the low-replicative phase, there was no difference inmedian HBsAg titres between genotypes B and C HBV. Genotyp-ing was only performed in a subset of patients in the low-replica-tive phase due to insufficient HBV DNA.

Interest in quantitative HBsAg serology as a clinical bio-marker has been based upon studies which showed a positiveassociation with intrahepatic cccDNA levels [9,11] and HBVDNA [9,12]. HBV DNA quantification is currently the standardin selecting patients who are candidates for therapy, monitor-ing response to therapy, and detecting the emergence of drugresistance. Compared to HBV DNA, the assays for HBsAg quan-tification are less expensive, and are also fully automated witha high throughput capacity. However, the utility of HBsAg titresas a reliable surrogate for both cccDNA and HBV DNA remainsunclear, as studies have also shown a poor correlation withHBV cccDNA [13], and only a positive correlation with HBVDNA in HBeAg positive CHB [14].

An understanding of HBsAg titre changes throughout HBVinfection may provide some potentially useful insights into hep-atitis B pathogenesis and viral life cycle. The mechanisms under-lying HBsAg and viral replication during different phases of CHBare currently unclear. This current study observed a modest cor-relation of serum HBsAg with HBV DNA in the immune-clearancephase of CHB (r = 0.77, p = 0.0001). No correlation was observedin the IT, IC or ENH phases. Furthermore, the ratio of HBsAg toHBV DNA was significantly higher in the low-replicative phasecompared to all other phases (1.05 vs 0.55, 0.55, 0.64.p < 0.0001), a finding which is in accordance with previous stud-ies [4]. The apparent ‘‘disconnect” between HBsAg and HBV DNAat different phases may possibly be due to the processes of inte-grated viral envelope sequences. A second possible explanation isa different in the regulation of viral replication during differentphases of infection, resulting in altered ratios of HBV virion tosub-viral HBsAg particles [15].

Genotype C

IT IC

B

LR ENH0

1

2

3

4

5

6

Phase of CHB

HB

sAg

log 1

0 IU

/ml

notype C CHB. Median values with 95% confidence interval represented.

0 vol. 52 j 508–513 511

Uncoating

HBsAgPre-S truncation

GOLGI

ER

Spherical & Filamentous HBsAg

Mature HBV virion

Nuclear transport

RC-DNA

Transcription

Viral integration pathway DSL-DNADSL-DNA

TranslationReverse

transcription

Immaturenucleocapsid

Mature nucleocapsid

Fig. 5. The two separate pathways of HBsAg and HBVDNA production. RC-DNA, relaxed circular DNA; DSL DNA, double-stranded linear DNA; cccDNA, covalently closedcircular DNA; mRNA, HBsAg, hepatitis B surface antigen.

Research Article

HBsAg synthesis during the HBV viral life cycle is complex,and typically occurs at the endoplasmic reticulum (Fig. 5). Theenvelope open reading frame (ORF) contains three in frame‘‘start” codons which further divide it into preS1, preS2, andORF-S domains. Envelope proteins are generated from twoHBV mRNA transcripts, with subsequent translation resultingin small (ORF-S), medium (pres2 + ORF-S) and large surfaceenvelope proteins (preS1 + preS2 + ORF-S). These are also knownas L, M and S surface proteins, respectively. Newly synthesisedenvelope protein interacts with mature HBV nucleocapsids atthe endoplasmic reticulum prior to secretion from the hepato-cyte. However, HBsAg production far exceeds that required forvirion assembly, and excess surface envelope proteins are cova-lently linked by intermolecular disulphide bonds and secretedas non infectious filamentous or spherical sub-viral particles[15]. These sub-viral particles may play a role in evading thehost immune response [16], and may also co-exist with anti-HBs as part of circulating immune complexes [17]. It is impor-tant to appreciate that whilst HBsAg quantification detects allthree forms of systemic HBsAg (part of HBV virion, spherical, fil-amentous), differentiation between the relative proportions iscurrently not routine.

HBsAg may also be produced from HBV DNA integrated intothe host genome. Although viral integration is an essential com-ponent of the life cycle of retroviruses such as HIV, it is notrequired for normal productive hepadnaviral infection. Rather,integration of HBV DNA occurs illegitimately through recombina-

512 Journal of Hepatology 201

tion mechanisms using host enzymes from double-stranded lin-ear (DSL) HBV DNA (Fig. 5) [18,19]. In HBV infection, viralintegration does seem to occur early in infection. Whilst HBVintegration is believed to be a random event, a high preferencefor integration occurs at the DR1 and DR2 sequences on theHBV genome [20]. Integrated sequences cannot provide a tem-plate for productive viral replication as a complete genome isnot present [21]. However, given that sequences of the S genesof the enhancer I elements are often present in integrated seg-ments, HBsAg may be produced [21].

The role of quantitative HBsAg in predicting response topeginterferon therapy has been the focus of several recentstudies [7,22–24]. In HBeAg negative patients treated withpegylated interferon, the decline in HBsAg titre at week 12and 24 has been shown to be a useful predictor of achievingan undetectable viral load at 24 weeks post therapy [7]. Theresults of this current study may have implications for futuretreatment algorithms evaluating the on-treatment decline ofHBsAg titres in both HBeAg positive and negative patients.For example, the median baseline serum HBsAg titre in theLR phase was 2.86 log10 IU/ml. Future clinical trials couldevaluate whether an on-treatment HBsAg titre decline tothese levels is a predictor of durable suppression of viralreplication.

A limitation of this study was its cross-sectional design, as itwould have been useful to follow patients longitudinally throughdifferent phases of infection. However, such longitudinal follow-

0 vol. 52 j 508–513

JOURNAL OF HEPATOLOGY

up is difficult given patients can remain in either the immune-tolerant or non/low-replicative phases for years, and patients inthe immune-clearance and HBeAg negative hepatitis phases arepotential treatment candidates.

In conclusion, this study demonstrates significant differencesin the baseline serum HBsAg titres across the different phasesof CHB infection. Quantitative HBsAg assays are non invasive,easy to perform and relatively inexpensive. Understanding thechanging HBsAg titres throughout the natural history of CHB rep-resent a step forward in further investigating the HBV viral lifecycle and the influence of the host immune response. BaselineHBsAg quantification may help refine future treatment algo-rithms for both immune-modulator therapy and oral nucle-os(t)ide analogue therapy. Larger prospective studies are nowrequired to evaluate longitudinal changes in serum HBsAg, andevaluating their significance in predicting the ultimate goal ofantiviral therapy, HBsAg seroconversion.

Acknowledgements

The authors who have taken part in this study declared thatthey do not have anything to declare regarding funding fromindustry or conflict of interest with respect to this manuscript.This work was supported by a Postgraduate Scholarship fromthe Gastroenterological Society of Australia (GESA). SL hasbeen in receipt of a consultancy for Abbott Diagnostics Divi-sion. Abbott Diagnostics (Abbott Laboratories, Chicago, USA)provided the Architect Instrument for quantitative serologicaltesting.

References

[1] Liaw YF, Leung N, Kao JH, Piratvisuth T, Gane E, Han KH, Guan R, Lau G,Locarnini S. Asian-Pacific consensus statement of the management ofchronic hepatitis B: a 2008 update. Hepatol Int 2008;2:263–283.

[2] European Association for The Study of The L. EASL Clinical PracticeGuidelines: management of chronic hepatitis B. J Hepatol 2009;50:227–42.

[3] Hadziyannis SJ, Vassilopoulos D. Hepatitis B e antigen-negative chronichepatitis B. Hepatology 2001;34:617–624.

[4] Volz T, Lutgehetmann M, Wachtler P, Jacob A, Quaas A, Murray JM, et al.Impaired intrahepatic hepatitis B virus productivity contributes to low viremiain most HBeAg-negative patients. Gastroenterology 2007;133:843–852.

[5] Zoulim F. New insight on hepatitis B virus persistence from the study ofintrahepatic viral cccDNA. J Hepatol 2005;42:302–308.

[6] Brunetto MR, Moriconi F, Bonino F, Lau GK, Farci P, Yurdaydin C, et al.Hepatitis B virus surface antigen levels: a guide to sustained response topeginterferon alfa-2a in HBeAg-negative chronic hepatitis B. Hepatology2009;49:1141–1150.

[7] Moucari R, Mackiewicz V, Lada O, Ripault MP, Castelnau C, Martinot-Peignoux M, et al. Early serum HBsAg drop: a strong predictor of sustained

Journal of Hepatology 201

virological response to pegylated interferon alfa-2a in HBeAg-negativepatients. Hepatology 2009;49:1151–1157.

[8] Ayres A, Locarnini S, Bartholomeusz A. HBV genotyping and analysis forunique mutations. Methods Mol Med 2004;95:125–149.

[9] Deguchi M, Yamashita N, Kagita M, Asari S, Iwatani Y, Tsuchida T, et al.Quantitation of hepatitis B surface antigen by an automated chemilumines-cent microparticle immunoassay. J Virol Methods 2004;115:217–222.

[10] Yuen LK, Ayres A, Littlejohn M, Colledge D, Edgely A, Maskill WJ, et al.SeqHepB: a sequence analysis program and relational database system forchronic hepatitis B. Antiviral Res 2007;75:64–74.

[11] Rodella A, Galli C, Terlenghi L, Perandin F, Bonfanti C, Manca N. Quantitativeanalysis of HBsAg, IgM anti-HBc and anti-HBc avidity in acute and chronichepatitis B. J Clin Virol 2006;37:206–212.

[12] Chen CH, Lee CM, Wang JH, Tung HD, Hung CH, Lu SN. Correlation ofquantitative assay of hepatitis B surface antigen and HBV DNA levels inasymptomatic hepatitis B virus carriers. Eur J Gastroenterol Hepatol2004;16:1213–1218.

[13] Manesis EK, Papatheodoridis GV, Hadziyannis E, Nastos T, Karayannis P.HBsAg serum levels correlate with total liver HBV DNA but not with cccDNA.Hepatology 2008;48 (No. 4, Suppl.):371A, abstract 147.

[14] Thompson A, Nguyen T, Gane E, Abbott W, Lau GK, Desmond P, Bowden S,Locarnini S. Serum HBsAg concentration: Relationship to intrahepatic andserum markers of HBV replication. J Hepatol 2009, abstract.

[15] Stibbe W, Gerlich WH. Variable protein composition of hepatitis B surfaceantigen from different donors. Virology 1982;123:436–442.

[16] Vanlandschoot P, Leroux-Roels G. Viral apoptotic mimicry: an immuneevasion strategy developed by the hepatitis B virus? Trends Immunol2003;24:144–147.

[17] Madalinski K, Burczynska B, Heermann KH, Uy A, Gerlich WH. Analysis ofviral proteins in circulating immune complexes from chronic carriers ofhepatitis B virus. Clin Exp Immunol 1991;84:493–500.

[18] Kimbi GC, Kramvis A, Kew MC. Integration of hepatitis B virus DNA intochromosomal DNA during acute hepatitis B. World J Gastroenterol2005;11:6416–6421.

[19] Bill CA, Summers J. Genomic DNA double-strand breaks are targets forhepadnaviral DNA integration. Proc Natl Acad Sci USA 2004;101:11135–11140.

[20] Nagaya T, Nakamura T, Tokino T, Tsurimoto T, Imai M, Mayumi T, et al. Themode of hepatitis B virus DNA integration in chromosomes of humanhepatocellular carcinoma. Genes Dev 1987;1:773–782.

[21] Thomas HC, Lemon S, Zuckerman AJ. Viral Hepatitis. 3rd ed. Massachusetts,USA: Blackwell Publishing; 2005.

[22] Brunetto MR, Bonino F, Marcellin P, Button P, Batrla R. Kinetics of HBsAgdecline in patients with HBeAg-negative chronic hepatitis B treated withpeginterferon alfa-2a according to genotype and its association withsustained HBsAg clearance 4 years post-treatment. Hepatology 2008;48(No. 4, Suppl.):740A, abstract 965.

[23] Rijckborst V, Borg MJ, Akarca US, Grima P, Flisiak R, Vafiadis-Zouboulis I,et al. Early reduction of serum HBsAg levels in HBeAg-negative chronichepatitis B patients achieving sustained virological response after Peginter-feron alfa-2a +/� ribavirin treatment. Hepatology 2008;48 (No. 4, Sup-pl.):749A, abstract 986.

[24] Lau G, Marcellin P, Brunetto MR, Piratvisuth T, Kapprell H, Button P, et al.On-treatment HBsAg decline during peginterferon alfa-2a (40KD) +/�lamivudine in patients with HBeAg-positive CHB as a potential predictorof durable off-treatment response. Hepatology 2008;48 (No. 4, Suppl.):714A,abstract 910.

0 vol. 52 j 508–513 513