Cn

ZYa

b

c

a

ARR2A

KHHAN

1

i[atstc

8

1h

Journal of Clinical Virology 59 (2014) 228–234

Contents lists available at ScienceDirect

Journal of Clinical Virology

jou rn al hom epage: www.elsev ier .com/ locate / j cv

linical implications of hepatitis B surface antigen quantitation in theatural history of chronic hepatitis B virus infection

haoxia Tana,c, Maoshi Lia,c,1, Xuemei Kuanga,c, Yu Tanga,c, Yi Fana,c, Guohong Denga,c,uming Wanga,c, Dengming Hea,b,c,∗

Institute of Infectious Disease, Southwest Hospital, Third Military Medical University, ChinaLiver Disease Diagnosis and Treatment Center, The 88th Hospital of Chinese PLA, ChinaThe Chongqing Key Laboratory for Research of Infectious Diseases, China

r t i c l e i n f o

rticle history:eceived 6 September 2013eceived in revised form0 December 2013ccepted 19 January 2014

eywords:BsAg quantitationBV DNAgeatural history

a b s t r a c t

Background: HBsAg quantitation may be useful for managing patients with hepatitis B virus (HBV) infec-tion.Objectives: We explored the clinical implications of HBsAg quantitation for patients with HBsAg levels>250 IU/ml (Abbott Diagnostics).Study design: Two hundred and thirty-three HBV-infected patients comprising 29 immune tolerancecases, 49 treatment-naïve HBeAg-positive chronic hepatitis B (CHB) cases, 91 inactive HBV carrier cases,and 64 treatment-naïve HBeAg-negative CHB cases were analyzed. HBsAg was quantified by the ArchitectHBsAg assay (Abbott Diagnostics) after a 1:500 automated dilution.Results and conclusions: HBsAg (log 10 IU/ml) was established for immune tolerance (4.50 ± 0.43),HBeAg-positive CHB (4.17 ± 0.66), inactive HBV carrier (3.32 ± 0.44), and HBeAg-negative CHB(3.23 ± 0.40); (p = 4.92 × 10−35). No significant difference was observed between inactive HBVcarrier and HBeAg-negative CHB (p = 0.247). The proportions of HBsAg <2000 IU/ml for inac-tive HBV carrier and HBeAg-negative CHB were 51.6% and 59.3%, respectively (p = 0.341).Positive correlations between HBsAg and HBV DNA were observed for immune tolerance(p = 1.23 × 10−4) and HBeAg-positive CHB (p = 0.003), but not for HBeAg-negative CHB (p = 0.432)

. A negative correlation between HBsAg and age was observed for immune tolerance (p = 0.030), HBeAg-positive CHB (p = 0.016), and inactive HBV carrier (p = 0.001), but not in HBeAg-negative CHB (p = 0.249).No significant differences between HBsAg and ALT for HBeAg-positive (p = 0.338) or HBeAg-negative CHB(p = 0.564) were observed. For patients with HBsAg quantitation >250 IU/ml, HBsAg may reflect HBV DNAreplication for HBeAg-positive cases. HBsAg is not a suitable marker for evaluating hepatitis activity anddistinguishing between cases of HBeAg-negative CHB and inactive HBV carrier state.

. Background

The detection of hepatitis B surface antigen (HBsAg) in serums crucial to the diagnosis of hepatitis B virus (HBV) infections1]. HBsAg is produced by the translation of transcriptionallyctive covalently closed circular (ccc) DNA molecules, and by theranslation of viral genes transcribed from integrated HBV DNA

equences in the host genome [2–4]. HBsAg seroclearance reflectshe immunological control of HBV infection, and is considered thelosest thing to a cure for chronic hepatitis B (CHB). However, the

∗ Corresponding author at: Liver Disease Diagnosis and Treatment Center, The8th Hospital of Chinese PLA, China. Tel.: +86-538-8839960.

E-mail address: hedengming@gmail.com (D. He).1 Contributed equally to this work.

386-6532/$ – see front matter © 2014 Elsevier B.V. All rights reserved.ttp://dx.doi.org/10.1016/j.jcv.2014.01.013

© 2014 Elsevier B.V. All rights reserved.

rates of spontaneous HBsAg seroclearance are extremely low, withannual seroclearance rates 0.72% and 0.07% for European and Asianpatients, respectively [2].

Recently, studies on the quantitation of HBsAg during naturalhistory of HBV infection and with antivirus therapy have emerged.Significant differences in serum HBsAg titers were reported acrossthe different phases of chronic HBV infection [3,5]. Several groupshave proposed cutoff levels of HBsAg and HBV DNA that, when usedtogether, reliably identify patients with inactive disease [5,6]. How-ever, correlations between HBsAg quantitation and HBV DNA werenot in agreement for several studies [3,5,7].

HBsAg serum levels are the resultant of the complex equilibrium

between the virus and the host’s immune system as well as tran-scription products of specific mRNAs rather than of viral replication[8]. HBsAg levels have been found to correlate well with the prog-nosis of CHB [9–12]. According to the natural history, chronic HBV

ical Virology 59 (2014) 228–234 229

iraactn

ao(hHH1

2

ltH

3

3

3>nttlTHinaPsvtCo

3

tpTqwtatAodtdS

Z. Tan et al. / Journal of Clin

nfection was divided into five phases: immune tolerance, immuneeactive phase, inactive HBV carrier state, HBeAg-negative CHB,nd HBsAg-negative phase [13]. Thus, HBsAg quantitation providesn alternative to, and complementary information, to assist in theharacterization of a patient’s infection status. In addition, quanti-ation of HBsAg offers promise as a prognostic marker during theatural history of chronic HBV infection.

The quantitation of HBsAg titters may be a useful tool in man-ging patients with chronic HBV infections [14]. The quantitationf serum HBsAg levels using the traditional Architect HBsAg assayAbbott Diagnostics) demonstrated that the majority of patientsad HBsAg levels >250 IU/ml. The recently developed ArchitectBsAg assay (commercially available since 2013) can quantitateBsAg levels within the range 175–125,000 IU/ml by an automatic:500 dilution.

. Objectives

We aimed to quantitate HBsAg levels in patients with HBsAgevels >250 IU/ml, and to explore the clinical utility in the charac-erization of infection status during the natural history of chronicBV infection.

. Study design

.1. Patients and groups

The study included 233 patients (163 male, 70 female, mean age7 ± 12 years) with chronic HBV infection and serum HBsAg levels250 IU/ml. Patients were classified according to four phases of theatural history of chronic HBV infection: immune tolerance state,reatment-naïve HBeAg-positive CHB, inactive HBV carrier state, orreatment-naïve HBeAg-negative CHB. The diagnostic criteria fol-owed the guidelines of the European Association for The Study ofhe Liver (EASL) [13]. The commonly accepted criteria for inactiveBV carrier state was adopted for the current study; that is, a min-

mum follow-up of 1 year, and the measurement of persistentlyormal alanine aminotransferase (ALT) levels (approx. 40 IU/ml)nd serum HBV DNA levels <2000 IU/ml, at least every 3–4 months.atients were excluded from the study if they had: (a) liver cirrho-is and hepatic cell carcinomas; (b) co-infections with hepatitis Cirus, hepatitis D virus or human immunodeficiency virus; (c) a his-ory of antiviral treatment for HBeAg-positive and HBeAg-negativeHB patients; (d) a history of autoimmune diseases; or (e) a historyf alcohol intake.

.2. Serological assays

Routine biochemical tests were performed using automatedechniques. Serum HBV DNA was quantified using the TaqManolymerase chain reaction (PCR) assay on a LightCycler 480 Real-ime PCR System (Roche, Basel, Switzerland), with a lower limit ofuantification of 500 IU/ml. HBsAg, HBeAg, anti-HBe and anti-HBcere detected by the Architect i2000SR System (Abbott Labora-

ories, Abbott Park, IL, USA). HBsAg quantitation was performedfter 1:500 automated dilution by a chemiluminescent micropar-icle immunoassay (CMIA) (Architect HBsAg, Abbott Laboratories).rchitect HBsAg CMIA detects 25 of the 26 HBsAg mutant samples,f which 24 mutant samples contain mutations spanning the “a”

eterminant region, whilst the two remaining samples have muta-ions outside of “a” determinant region [15]. HBV genotypes wereetermined using a HBV gene type PCR fluorescence kit (Fosun,hanghai, China).

Fig. 1. The distribution of HBsAg quantitation according to HBV DNA levels.

3.3. Statistical analysis

Continuous variables were expressed as the mean ± SD (range)for normal distributions or the median (interquartile range, IQR)for abnormal distributions. Pearson’s Chi-square test (r2), ANOVA,and Pearson’s (r) or Spearman’s (�) correlations were carried outas appropriate. All tests for significance and resulting p valueswere two-sided, with a level of significance of 0.05. The statisticalsoftware used for this analysis was SPSS (version 18.0; SPSS Inc.,Chicago, IL, USA).

4. Results

4.1. HBsAg quantitation of patients

The demographic data, clinical characteristics, and serumHBsAg levels for the 233 participants in the current study arequantitation of the presented in Table 1. As shown, HBsAg lev-els did not exceed the upper limit of 5.10 log 10 IU/ml, andthe average HBsAg levels for the sample population were3.61 ± 0.68 log 10 IU/ml (2.51–5.10). The average HBsAg levels formale and female patients were 3.56 ± 0.67 log 10 IU/ml (2.51–5.10)and 3.74 ± 0.68 log 10 IU/ml (2.78–5.10), respectively, and werenot considered significantly different (p = 0.063). The averageHBsAg levels for genotype B and C were 3.80 ± 0.74 log 10 IU/ml(2.61–5.10) and 3.81 ± 0.75 log 10 IU/ml (2.51–5.10), respectively,and were not considered significantly different (p = 0.952).The average HBsAg levels for HBeAg-positive patients andHBeAg-negative patients were 4.29 ± 0.60 log 10 IU/ml (2.93–5.10)and 3.28 ± 0.43 log 10 IU/ml (2.51–4.52), respectively, and werenot considered significantly different (p = 5.53 × 10−25). Theaverage levels of HBsAg quantitation for immune tolerancestate, 4.50 ± 0.43 log 10 IU/ml (3.59–4.96); HBeAg-positive CHB,4.17 ± 0.66 log 10 IU/ml (2.93–5.10); inactive HBV carrier state,3.32 ± 0.44 log 10 IU/ml (2.54–4.51); and HBeAg-negative CHB3.23 ± 0.40 log 10 IU/ml (2.51–4.52) were significantly different(p = 4.92 × 10−35). However, no significant difference was observedbetween HBsAg levels for the inactive HBV carrier state and HBeAg-negative CHB (p = 0.246).

4.2. Distribution of HBsAg quantitation

The distribution of HBsAg levels according to HBV DNA lev-els is presented in Fig. 1. The proportion of serum samples withHBsAg levels: <3 log 10 IU/ml, 3–4 log 10 IU/ml, 4–5 log 10 IU/ml,

230 Z. Tan et al. / Journal of Clinical Virology 59 (2014) 228–234

Table 1Demographic data, clinical characteristics, and HBsAg levels of patients.

Characteristics Tolerance HBeAg (+) CHB Inactive HBeAg (−) CHB

Number 29 49 91 64Male- no. (%) 13 (44.8) 35 (71.4) 66 (72.5) 49 (76.6)Genotype B–no. (%) 13 (44.8) 21 (42.9) NAa 36 (56.3)Age – ys 28 ± 10 32 ± 12 39 ± 10 44 ± 12HBV DNA – log 10 IU/ml 7.1 ± 0.7 7.0 ± 0.8 <3.3 4.8 ± 1.0ALT – ×ULN <1 2.1 (1.5–5.5) <1 1.6 (1.2–2.7)HBsAg levels – log 10 IU/ml 4.50 ± 0.43 4.17 ± 0.66 3.32 ± 0.44 3.23 ± 0.40

a(liHmiCpppbC

pHuTH(

4

fHaHTp(ais((

TD

a HBV DNA undetectable or under the lowest limit of HBV genotype detection.

nd >5 log 10 IU/ml, were 19.7% (46/233), 51.9% (121/233), 25.8%60/233), and 2.6% (6/233), respectively. The distribution of HBsAgevels according to sex, genotype, and natural history are shownn Table 2. There was no significant difference in the frequency ofBsAg levels based on sex (p = 0.079) or genotype (p = 0.515). Theajority of low HBsAg levels were for HBeAg-negative patients;

nclusive of both inactive HBV carrier state and HBeAg-negativeHB. The majority of high HBsAg levels were for HBeAg-positiveatients; inclusive of both immune tolerance state and HBeAg-ositive CHB. No significant difference was observed in theroportion of HBsAg levels <3 log 10 IU/ml and at 3–4 log 10 IU/mletween cases inactive HBsAg carrier state and HBeAg-negativeHB (p = 0.996).

HBsAg levels of 3.0–3.3 log 10 IU/ml (1–2 × 103 IU/ml) are theroposed cutoff levels for the identification of the true inactiveBsAg carrier state. As such, these values were applied in the eval-ation of HBeAg-negative CHB and inactive HBV carrier state cases.he proportion of HBsAg levels <3.3 log 10 IU/ml (2000 IU/ml) inBeAg-negative CHB and inactive HBsAg carrier state were 59.3%

38/64) and 51.6% (47/91), respectively (p = 0.341).

.3. Correlation between HBsAg and HBV DNA levels

Correlation between HBsAg and HBV DNA levels were per-ormed for a total of 142 patients. Patients with an inactiveBsAg carrier state were excluded from this analysis. Over-ll, a significant positive correlation was observed betweenBsAg and HBV DNA levels (r = 0.637, p = 4.21 × 10−19) (Fig. 2A).he same positive correlation was observed for male (r = 0.652,

= 8.80 × 10−14) (Fig. 2B) and female (r = 0.605, p = 1.60 × 10−4)Fig. 2C), for genotype B (r = 0.669, p = 2.48 × 10−10) (Fig. 2D)nd genotype C (r = 0.739, p = 1.23 × 10−13) (Fig. 2E). A signif-

cant positive correlation was observed for immune tolerancetate (r = 0.653, p = 1.23 × 10−4) (Fig. 2F) and HBeAg-positive CHBr = 0.416, p = 0.003) (Fig. 2G), but not for HBeAg-negative CHBr = 0.100, p = 0.432) (Fig. 2H).

able 2istribution of different HBsAg quantitation.

Groups

<3 (n = 46) 3–

SexMale – no. (%) 35 (76.1) 88Female – no. (%) 11 (23.9) 33

GenotypeB – no. (%) 11 (23.9) 29C – no. (%) 9 (19.6) 36NA – no. (%) 20 (56.5) 56

Natural historyTolerance – no. (%) 0 5HBeAg (+) CHB – no. (%) 1 (2.2) 19Inactive – no. (%) 26 (56.5) 56HBeAg (−) CHB – no. (%) 19 (41.3) 41

4.4. Correlation between HBsAg levels and age

A significant negative correlation was observed for all 233patients with regard to HBsAg levels and age (r = −0.484,p = 4.61 × 10−15) (Fig. 3A). The same negative correlation wasobserved for male (r = −0.500, p = 1.02 × 10−11) (Fig. 3B) andfemale (r = −0.515, p = 5.16 × 10−6) (Fig. 3C), for genotype B(r = −0.584, p = 1.09 × 10−7) (Fig. 3D) and genotype C (r = −0.490,p = 1.024 × 10−5) (Fig. 3E). A significant negative correlation wasobserved for immune tolerance state (r = −0.404, p = 0.03) (Fig. 3F),HBeAg-positive CHB (r = −0.343, p = 0.016) (Fig. 3G), and inactiveHBsAg carrier state (r = −0.333, p = 0.001) (Fig. 3H), but not forHBeAg-negative CHB (r = −0.146, p = 0.249) (Fig. 3I).

4.5. Correlation between HBsAg and ALT levels

Of the 113 patients with elevated ALT levels included in thisanalysis, 49 were HBeAg-positive CHB patients and 64 were HBeAg-negative patients. Overall, no significant correlation was observedbetween HBsAg and ALT levels (� = 0.050, p = 0.439) (Fig. 4A). Fur-ther to this, no significant correlation was observed for male(� = 0.100, p = 0.363) (Fig. 4B) and female (� = −0.039, p = 0.839)(Fig. 4C), for genotype B (� = −0.227, p = 0.090) (Fig. 4D) andgenotype C (� = −0.161, p = 0.235) (Fig. 4E), for HBeAg-positiveCHB (� = −0.140, p = 0.338) (Fig. 4F) and HBeAg-negative CHB(� = −0.073, p = 0.564) (Fig. 4G).

5. Discussion

Increasing evidence shows that a knowledge of HBsAg levelscan further develop our understanding of the natural history ofHBV infection [16]. Unfortunately, only low HBsAg levels have par-

ticular clinical significance. For HBeAg-negative patients infectedwith HBV genotype B or C infection and with HBV DNA levels<2000 IU/ml, HBsAg levels <10 IU/ml proved to be the strongestpredictor of HBsAg loss [12]. An HBsAg level <100 IU/ml predicted

HBsAg quantitation (log 10 IU/ml)

4 (n = 121) 4–5 (n = 60) >5 (n = 6)

(72.7) 35 (58.3) 5 (83.3) (27.3) 25 (41.7) 1 (16.7)

(24.0) 27 (45%) 3 (50.0) (29.8) 24 (40.0) 3 (50.0) (46.3) 9 (15.0) 0

(4.1) 24 (40.0) 0 (15.7) 23 (38.3) 6 (100) (46.3) 9 (15.0) 0 (33.9) 4 (6.7) 0

Z. Tan et al. / Journal of Clinical Virology 59 (2014) 228–234 231

F ith def 0−13);H

Hs4Hbmwd

tcsoA

mctie

ig. 2. Correlation between HBsAg quantitation and HBV DNA levels in patients wemale (p = 1.60 × 10−4); D: genotype B (p = 2.48 × 10−10); E: genotype C (p = 1.23 × 1BeAg-negative CHB (p = 0.432).

BsAg loss within 6 years with a diagnostic accuracy of 91.5%, sen-itivity of 83.3%, specificity of 92.1%, a positive predictive value of5.5%, and a negative predictive value of 98.6% in patients with anBV DNA level <200 IU/ml [9]. HBsAg levels <1000 IU/ml in com-ination with low levels of HBV DNA and ALT help to define theinimal-risk HBV carriers [10]. Among HBeAg-negative patientsith low viral loads, the risk of hepatocellular carcinoma (HCC) isetermined by levels of HBsAg, ALT and age, but not HBV DNA [11].

A decline of serum HBsAg levels represents a reduction inhe translation of mRNAs produced from transcriptionally activeccDNA or integrated sequences [8]. Given that, for most patient’serum HBsAg levels exceed 250 IU/ml during the natural historyf HBV infection, we aimed to explore the application of the newrchitect HBsAg assay for use in clinical settings.

We found that the new Architect HBsAg assay, with an auto-ated 500-fold sample dilution, fully met the clinical for use in

linical applications. Reductions in HBsAg levels, determined quan-itatively using this assay, coincide the natural history of HBVnfection. However, no significant difference in serum HBsAg lev-ls were observed between HBeAg-negative CHB and inactive HBV

tectable HBV DNA levels. A: total (p = 4.21 × 10−19); B: male (p = 8.80 × 10−14); C: F: immune tolerance state (p = 1.23 × 10−4); G: HBeAg-positive CHB (p = 0.003); H:

carrier state cases. In addition, the proportions of HBsAg belowthe proposed HBsAg cutoff levels [1–2 × 103 IU/ml) representingthe true inactive HBV carrier state were not significantly differ-ent between cases of HBeAg-negative CHB and inactive HBV carrierstate. This result suggests that the sole use of HBsAg quantitationfor the identification of the true inactive HBV carrier state may beinsufficient.

The quantitation of HBV DNA by PCR is routinely applied in thediagnostic workup for CHB. A decline of serum HBV DNA reflects areduction in viral replication. In this study, for patients with HBVDNA levels >2000 IU/ml, a strong positive correlation was observedbetween HBsAg and HBV DNA levels for immune tolerance stateand for HBeAg-positive CHB cases, but not for HBeAg-negative CHBcases. These results suggest that HBsAg levels are a strong indica-tor of HBV DNA replication in patients with chronic HBV infectionand who are HBeAg-positive. These results are not consistent with

the work of others in this area. A cross-sectional study demon-strated there was no significant correlation between HBsAg andHBV DNA levels for HBeAg-negative patients [17]. A European per-spective study demonstrated a strong correlation between HBsAg

232 Z. Tan et al. / Journal of Clinical Virology 59 (2014) 228–234

Fig. 3. Correlation between HBsAg quantitation and age. A: total (p = 4.61 × 10−15); B: male (p = 1.02 × 10−11); C: female (p = 5.16 × 10−6); D: genotype B (p = 1.09 × 10−7);E ositive(

aHiAweowpiptHwcd

us

: genotype C (p = 1.024 × 10−5); F: immune tolerance state (p = 0.03); G: HBeAg-pp = 0.249).

nd HBV DNA levels only during acute hepatitis B. Correlation ofBsAg and HBV DNA levels was weak or absent when analyz-

ng different phases of persistent HBV infection separately [5]. Ansian perspective showed that serum HBsAg titers only correlatedith HBV DNA levels in the immune clearance phase [3]. A mod-

rate correlation between serum HBsAg and HBV DNA levels wasbserved in untreated HBsAg carriers; however, the correlationas greater for HBeAg-positive patients than for HBeAg-negativeatients [7]. HBsAg was positively correlated with serum HBV DNA,

ntrahepatic cccDNA, and total HBV DNA for HBeAg-positive CHBatients. HBsAg correlated poorly with serum HBV DNA, and failedo correlate with either intrahepatic cccDNA or total HBV DNA forBeAg-negative CHB patients [4]. HBsAg was weakly correlatedith HBV DNA, irrespective of the HBsAg assay used [18]. Dis-

repancies in the outcomes of these studies may be the result of

ifferences in patient selection criteria.

With the exception of HBeAg-negative CHB, all remaining nat-ral history of HBV infection phases analyzed in the current studyhowed a strong negative correlation between HBsAg levels and

CHB (p = 0.016); H: inactive HBV carrier state (p = 0.001); I: HBeAg-negative CHB

age. A study showed that age was independently associated withHBsAg levels, and was dependent on HBeAg status. The authorsobserved a general reduction in HBsAg levels with an increasing inage [7]. Perinatally acquired HBV is the main model of chronic HBVinfection in China, hence the age of patients indicates the durationof infection for the majority of the chronic HBV infection and isconsidered an important factor associated with the progression ofHBV-related diseases. The negative correlation between HBsAg andage, and the decline in HBsAg levels associated with the progressivephases of the natural history of HBV infection, demonstrate thatHBsAg levels are reflective of antivirus immunity and the durationof HBV infection, particularly for HBeAg-positive patients.

There was no correlation between HBsAg and ALT levels forHBeAg-positive CHB cases or HBeAg-negative CHB cases. This resultis in agreement with other study [19], and suggests that HBsAg

quantitation may have no special value in evaluating the degree ofactive hepatitis.

Differences in HBV genotypes were considered in the presentstudy. We found no significant differences between genotype B and

Z. Tan et al. / Journal of Clinical Virology 59 (2014) 228–234 233

Fig. 4. Correlation between HBsAg quantitation and ALT in patients with elevation of ALT. A: total (p = 0.439); B: male (p = 0.363); C: female (p = 0.839); D: genotype B( negati

CbAttItd

HHiopscHd>

p = 0.090); E: genotype C (p = 0.235); F: HBeAg-positive CHB (p = 0.338); G: HBeAg-

with regard to HBsAg levels, distribution of HBsAg, or correlationsetween HBsAg levels and age, HBV DNA levels, or ALT levels. Ansian study showed that HBsAg titers were similar between geno-

ypes B and C [3]. Another study showed that HBsAg levels tendedo correlate with HBV DNA for genotype A, not for genotype D [18].n contrast, another study demonstrated this association for geno-ype D and not for genotype A [5]. Difference in HBsAg levels forifferent HBV genotypes needs to be investigated further.

It is worth noting that no correlation was observed betweenBsAg and HBV DNA levels, or age for HBeAg-negative CHB cases.BeAg-negative CHB may follow HBe seroconversion during the

mmune reactive phase or may develop after years or decadesf the inactive carrier state. Furthermore, HBeAg-negative CHBatients have a high risk of progression to advanced hepatic fibro-is, cirrhosis and subsequent complications such as decompensated

irrhosis and HCC [20]. Our study suggested that the value of exactBsAg quantitation should be limited for evaluating the actualegree of lesions in HBeAg-negative patients with HBsAg levels250 IU/ml.

ve CHB (p = 0.564).

We believe there are three limitations to our study. First, abnor-mal distribution of ALT levels may incur some degree of statisticalbias. Second, a greater number of phenotypes representing chronicHBV infection, such as liver cirrhosis and HCC, were not included.Finally, the absence of follow-up data is a particular shortcomingof this cross-sectional study.

In summary, HBsAg levels may reflect HBV DNA replication, butonly in HBeAg-positive patients. HBsAg is not a suitable marker forevaluating hepatitis activity, or for distinguishing between HBeAg-negative CHB and inactive HBV carrier state. In conclusion, theclinical utility of HBsAg quantitation is limited to evaluating thestate of the natural history of HBV infection in patients with HBsAglevels >250 IU/ml.

Funding

Financial support for this study came from grants 81270563from the Natural Science Foundation of China; 2012ZX10002004from the State Key Project specialized for HBV-related severe

2 ical V

hC

C

i

E

pfwlae

A

e

R

[

[

[

[

[

[

[

[

[

34 Z. Tan et al. / Journal of Clin

epatitis of China; 2012XLC05 from the TMMU Key Project forlinical Research.

ompeting interests

None of the authors has to declare any financial or non-financialnterests.

thical approval

The analysis was conducted on anonymized data, collected asart of routine patient care. No additional investigations were per-ormed. Therefore, no prior informed consent from the patientsas required. The study protocol conformed to the ethical guide-

ines of the 1975 Declaration of Helsinki as reflected in a prioripproval by the ethics committee of the Southwest Hospital. Ourthics committee waived the need for informed consent.

cknowledgement

We wish to acknowledge helpful suggestions, discussions andxcellent technical assistance by Prof. Qing Mao.

eferences

[1] Blumberg BS, Alter HJ, Visnich SA. New antigen in leukemia sera. JAMA1965;191:541–6.

[2] Chu CM, Liaw YF. Hepatitis B surface antigen seroclearance during chronic HBVinfection. Antivir Ther 2010;15(2):133–43.

[3] Nguyen T, Thompson AJ, Bowden S, Croagh C, Bell S, Desmond PV, et al. HepatitisB surface antigen levels during the natural history of chronic hepatitis B: a

perspective on Asia. J Hepatol 2010;52(4):508–13.

[4] Thompson AJ, Nguyen T, Iser D, Ayres A, Jackson K, Littlejohn M, et al. Serumhepatitis B surface antigen and hepatitis B e antigen titers: disease phase influ-ences correlation with viral load and intrahepatic hepatitis B virus markers.Hepatology 2010;51(6):1933–44.

[

[

irology 59 (2014) 228–234

[5] Jaroszewicz J, Calle Serrano B, Wursthorn K, Deterding K, Schlue J, Rau-pach R, et al. Hepatitis B surface antigen (HBsAg) levels in the naturalhistory of hepatitis B virus (HBV)-infection: a European perspective. J Hepatol2010;52(4):514–22.

[6] Brunetto MR, Oliveri F, Colombatto P, Moriconi F, Ciccorossi P, Coco B, et al.Hepatitis B surface antigen serum levels help to distinguish active from inactivehepatitis B virus genotype D carriers. Gastroenterology 2010;139(2):483–90.

[7] Jang JW, Yoo SH, Kwon JH, You CR, Lee S, Lee JH, et al. Serum hepatitis B surfaceantigen levels in the natural history of chronic hepatitis B infection. AlimentPharmacol Ther 2011;34(11–12):1337–46.

[8] Brunetto MR. A new role for an old marker, HBsAg. J Hepatol 2010;52:475–7.[9] Tseng TC, Liu CJ, Su TH, Wang CC, Chen CL, Chen PJ, et al. Serum hepatitis

B surface antigen levels predict surface antigen loss in hepatitis B e antigenseroconverters. Gastroenterology 2011;141(2):517–25, 25.e1–2.

10] Tseng TC, Liu CJ, Yang HC, Su TH, Wang CC, Chen CL, et al. Serum hepatitis Bsurface antigen levels help predict disease progression in patients with lowhepatitis B virus loads. Hepatology 2013;57(2):441–50.

11] Tseng TC, Liu CJ, Yang HC, Su TH, Wang CC, Chen CL, et al. High levels of hepatitisB surface antigen increase risk of hepatocellular carcinoma in patients with lowHBV load. Gastroenterology 2012;142(5), 1140-9.e3; quiz e13-e14.

12] Tseng T-C, Liu C-J, Yang H-C, Su T-H, Wang C-C, Chen C-L, et al. Determinantsof spontaneous surface antigen loss in hepatitis B e antigen-negative patientswith a low viral load. Hepatology 2012;55(1):68–76.

13] EASL clinical practice guidelines: management of chronic hepatitis B virusinfection. J Hepatol 2012;57:167–85.

14] Moucari R, Marcellin P. Quantification of hepatitis B surface antigen: a newconcept for the management of chronic hepatitis B. Liver Int 2011;31(Suppl.1):122–8.

15] Coleman PF, Chen YC, Mushahwar IK. Immunoassay detection of hepatitis Bsurface antigen mutants. J Med Virol 1999;59(1):19–24.

16] Liaw YF. Clinical utility of hepatitis B surface antigen quantitation in patientswith chronic hepatitis B: a review. Hepatology 2011;54(2):E1–9.

17] Ganji A, Esmaeilzadeh A, Ghafarzadegan K, Helalat H, Rafatpanah H, Mokhtari-far A. Correlation between HBsAg quantitative assay results and HBV DNA levelsin chronic HBV. Hepat Mon 2011;11(5):342–5.

18] Tuaillon E, Mondain AM, Nagot N, Ottomani L, Kania D, Nogue E, et al.Comparison of serum HBsAg quantitation by four immunoassays, and rela-tionships of HBsAg level with HBV replication and HBV genotypes. PLoS ONE2012;7(3):e32143.

19] Manesis EK, Papatheodoridis GV, Tiniakos DG, Hadziyannis ES, Agelopoulou OP,Syminelaki T, et al. Hepatitis B surface antigen: relation to hepatitis B replicationparameters in HBeAg-negative chronic hepatitis B. J Hepatol 2011;55(1):61–8.

20] Hadziyannis SJ, Papatheodoridis GV. Hepatitis B e antigen-negative chronichepatitis B: natural history and treatment. Semin Liver Dis 2006;26(2):130–41.