Clinical Gastroenterology and Hepatology 2019;17:182–191

Serum Level of Antibodies Against Hepatitis B Core Protein Is

Associated With Clinical Relapse After Discontinuation of
Nucleos(t)ide Analogue Therapy
Heng Chi,*,a Zhandong Li,‡,a Bettina E. Hansen,*,§ Tao Yu,‡ Xiaoyong Zhang,‡ Jian Sun,‡
Jinlin Hou,‡ Harry L. A. Janssen,*,§,b and Jie Peng‡,b

*Department of Gastroenterology and Hepatology, Erasmus MC University Medical Center Rotterdam, Rotterdam,
the Netherlands; ‡Department of Infectious Diseases, Guangdong Provincial Key Laboratory of Viral Hepatitis Research,
Nanfang Hospital, Southern Medical University, Guangzhou, China; and §Toronto Centre of Liver Disease, University Health
Network, Toronto General Hospital, University of Toronto, Toronto, Ontario, Canada

BACKGROUND & AIMS:
Levels of antibodies against the hepatitis B virus (HBV) core protein (anti-HBc) have been
associated with response to nucleos(t)ide analogue and (peg)interferon therapy in patients
with chronic HBV infection. We performed a prospective study to determine whether the total
serum level of anti-HBc level (immunoglobulins M and G) is associated with clinical relapse
after discontinuation of nucleos(t)ide analogue-based therapy.

METHODS: We collected data from patients with chronic HBV infection who discontinued nucleos(t)ide
analogue therapy according to pre-specified stopping criteria, recruited from November 2012
through July 2016 at an academic hospital in Guangzhou, China. Patients were followed through
February 2017. We performed comprehensive biochemical and virologic tests at every visit, and
anti-HBc was quantified for 2 years after treatment cessation (at baseline and weeks 4, 8, 12,
24, 48, and 96). The primary endpoint was clinical relapse, defined as level of HBV DNA >2000
IU/mL and level of alanine aminotransferase more than 2-fold the upper limit of normal—these
were also the criteria for retreatment.

RESULTS: We followed 100 patients (71% positive for HB e antigen [HBeAg] at the start of nucleos(t)ide
analogue therapy, 43% treated with entecavir or tenofovir) for a median of 2.5 years after
stopping therapy. Clinical relapse occurred in 39 patients (in 46% of patients at year 4 after
discontinuation). High level of anti-HBc at the end of treatment (hazard ratio [HR], 0.31 per log
IU/mL; P [ .002) and low level of HB surface antigen (HBsAg) at the end of treatment (HR, 1.71
per log IU/mL; P [ .032) were associated with a reduced risk of clinical relapse after adjusting
for age, start of nucleos(t)ide analogue therapy, HBeAg-status, and consolidation therapy
duration. At year 4, 21% of patients with anti-HBc levels at the end of treatment ‡1000 IU/mL
developed a clinical relapse compared to 85% of patients with levels <100 IU/mL (P < .001).
An HBsAg level at the end of treatment £100 IU/mL was associated with a reduced risk of
relapse (HR 0.30; P [ .045). However, 82% of patients had levels of HBsAg above 100 IU/mL;
for these patients, level of anti-HBc at the end of treatment could be used to determine the risk
of relapse (HR 0.39 per log IU/mL; P [ .005).

CONCLUSION: In a median 2.5-year follow-up study of patients with chronic HBV infection who stopped
nucleos(t)ide analogue therapy, total serum level of anti-HBc at the end of treatment was
associated with risk of clinical relapse. Serum level of anti-HBc might be used to select patients
suitable for discontinuing nucleos(t)ide analogue therapy.

Keywords: Prognostic Factor; Treatment Cessation; Antiviral Agents; CHB.

a
Authors share co-first authorship. bAuthors share co-senior authorship.
Abbreviations used in this paper: ALT, alanine aminotransferase; anti-
HBc, antibody against the hepatitis B virus core protein; CHB, chronic Most current article
hepatitis B; HBeAg, hepatitis B e antigen; HBsAg, hepatitis B surface
antigen; HBV, hepatitis B virus; HCC, hepatocellular carcinoma; Ig, © 2019 by the AGA Institute
immunoglobulin; IQR, interquartile range; LLOD, lower limit of detection; 1542-3565/$36.00
NA, nucleos(t)ide analogue; ULN, upper limit of normal. https://doi.org/10.1016/j.cgh.2018.05.047
January 2019
See editorial on page 39. See related article on
page 172.
ucleos(t)ide analogue (NA) therapy is widely
N used for patients with chronic hepatitis B virus
(HBV) infection.1 NAs inhibit the reverse transcriptase
activity of the HBV polymerase and thus suppress viral
replication. It is generally suggested that NA should be
continued until the occurrence of hepatitis B surface
antigen (HBsAg) seroclearance because stopping NA
treatment before achieving this endpoint is associated
with relapse.2–4 We have recently shown that relapses
were absent in NA-treated patients who discontinued NA
after HBsAg seroclearance,5 but HBsAg seroclearance is
rare even after many years of NA treatment.6 Therefore,
the benefit of long-term treatment should be weighed
against the burden of life-long medication, monitoring,
and adherence. In addition, the long-term risk for devel-
opment of resistance and adverse events remains unclear.
There is an increasing amount of studies investigating
the outcomes of NA discontinuation.1,7 It has also been
shown that despite the frequent occurrence of HBV DNA
elevations after stopping treatment, a substantial amount
of patients were able to achieve sustained response,
sometimes accompanied by HBsAg loss.4,8,9 The recently
updated chronic hepatitis B (CHB) clinical guidelines are
increasingly recommending the possibility of NA discon-
tinuation in selected patients.1,7 Nevertheless, few consis-
tent predictors have been found that can identify patients
with the smallest risk of relapse. As a result, no new pre-
dictors have been recommended by the new guidelines.1,7
Hepatitis B core protein is a subunit of the inner
nucleocapsid that encapsulates the viral DNA.10 This pro-
tein is highly immunogenic, and the antibody against this
antigen, immunoglobulin (Ig) G antibody against the HBV
core protein (anti-HBc),11 is broadly used to identify in-
dividuals who were ever infected with HBV, whereas IgM
anti-HBc is used to identify acute infection.1,12,13 Brunetto
et al14 have used a semiquantitative IgM anti-HBc assay to
differentiate between the stages of chronic HBV infection.
Despite this, anti-HBc testing has mainly been used as a
qualitative assay. Recently, a quantitative anti-HBc immu-
noassay has been developed that measures the total serum
anti-HBc (IgG and IgM).15,16 With this assay, subsequent
studies have shown that a high anti-HBc level was associ-
ated with response to NA as well as peg-interferon
treatment.17–19 Importantly, anti-HBc is produced by
HBV-specific B cells activated by T-cell–dependent as well
as T-cell–independent pathways.20 It has also been sug-
gested that the serum anti-HBc level may be correlated with
the host’s adaptive immune response.16,17,19 Also, HBsAg
level is correlated with the amount of covalently closed
circular DNA and has been associated with sustained
response after NA discontinuation.2,4,21 It is unknown
whether anti-HBc or the combination of anti-HBc and
HBsAg levels can reliably identify patients at low risk of
relapse after NA treatment discontinuation.
Anti-HBc Level and Nucleos(t)ide Analogue Discontinuation 183
What You Need to Know
Background
Nucleos(t)ide analogue discontinuation may be
attempted in selected chronic hepatitis B patients, yet
no predictors of relapse have been recommended.
Total anti-HBc level predicted response to nucleos(t)
ide analogue and peginterferon treatment, but it is
unknown whether it can predict relapse after
nucleos(t)ide analogue discontinuation.
Findings
High anti-HBc level at treatment discontinuation was
associated with a lower risk of clinical relapse.
Low HBsAg level was also associated with a reduced
risk of clinical relapse, and the combined use of
HBsAg and anti-HBc level stratified the risk of
relapse further.
Implications for patient care
Total anti-HBc level may be used in the selection of
patients suitable for NA discontinuation but warrants
further investigation.
In this prospective cohort of patients with chronic HBV
infection who discontinued NA therapy according to pre-
specified criteria, the aim was to investigate whether
quantitative anti-HBc level alone or in combination
with HBsAg level is able to identify patients with high
chance of sustained response after stopping NA. In addi-
tion, we aimed to investigate the association between off-
treatment anti-HBc kinetics and sustained response.
Methods
Setting and Patient Population
This was a prospective observational study of
patients with chronic HBV infection (HBsAg positive >6
months) who discontinued NA according to pre-specified
stopping criteria and consented to standardized study
follow-up and retreatment criteria. From November
2012 until July 2016, patients were recruited from an
academic hospital (Nanfang Hospital, Southern Medical
University, Guangzhou, China). This study is part of an
on-going effort to prospectively investigate NA therapy
discontinuation.4 All follow-up data until February 2017
were analyzed in this study.
Both start-of-treatment hepatitis B e antigen
(HBeAg)-positive as well as HBeAg-negative patients
who fulfilled the pre-defined stopping criteria were eligible
for inclusion. HBeAg-positive patients were required to
achieve HBeAg seroconversion and undetectable HBV DNA,
followed by at least 12 months of consolidation therapy,
consistent with the criteria of international guidelines.1,7,22
HBeAg-negative patients were required to achieve unde-
tectable HBV DNA, followed by at least 18 months of
consolidation therapy, in line with the Asian-Pacific
184 Chi et al
stopping criteria.22 At time of treatment discontinuation, all
patients were HBsAg-positive and HBeAg-negative and had
undetectable HBV DNA.
Exclusion criteria were coinfection with human
immunodeficiency virus, hepatitis C virus, or hepatitis D
virus; radiologic suspicion of hepatocellular carcinoma
(HCC) or alpha-fetoprotein >20 ng/mL; biopsy-proven
cirrhosis or liver stiffness >9 kPa with FibroScan (Echos-
ens, Paris, France); other concurrent liver disease; alcohol
abuse; immunosuppression; malignancy; history of liver
transplantation; history or presence of decompensated
liver disease (ascites as confirmed by ultrasound, bleeding
esophageal varices, hepatic encephalopathy, albumin <32
g/L, prothrombin time prolongation
3 seconds, or total
bilirubin level
2 the upper limit of normal [ULN]).
This study was approved by the ethics committee
(NFEC-201209-K3) and was performed according to
the Declaration of Helsinki.
Follow-up, Endpoints, and Retreatment Criteria
After NA therapy discontinuation, patients were
followed monthly during the initial 3 months. Thereafter,
the patients were followed every 3 months. After 2 years
of follow-up, the patients were evaluated every 6
months. Biochemical and virologic tests were performed
at every visit. Quantitative anti-HBc testing was done up
until 2 years of post-treatment follow-up (at baseline,
weeks 4, 8, 12, 24, 48, and 96).
The primary study endpoint was clinical relapse
defined as HBV DNA level >2000 IU/mL and alanine
aminotransferase (ALT) elevation >2 ULN. Patients
with clinical relapse were restarted on NA treatment and
discontinued from the study protocol. Secondary end-
points included HBsAg seroclearance and off-treatment
anti-HBc kinetics.
Laboratory Testing
Serum HBV DNA at start of treatment was tested with
a polymerase chain reaction HBV assay with a lower limit
of detection (LLOD) of 1000 copies/mL (Daan Gene Co,
Ltd, Sun Yat-sen University, Guangzhou, China). For
conversion of copies to IU, a conversion factor of 5 copies
per IU was applied. HBV DNA levels at NA discontinua-
tion and off-treatment follow-up visits were assessed
by using the Cobas TaqMan polymerase chain
reaction HBV assay with a LLOD of 20 IU/mL (Roche
Diagnostics, Basel, Switzerland). Quantitative HBsAg
(LLOD, 0.05 IU/mL) and qualitative HBeAg and anti-HBe
were determined by using the ARCHITECT (Abbott Lab-
oratories, Chicago, IL). Quantitative total serum anti-HBc
(IgG and IgM) was tested with a double-sandwich immu-
noassay (Wantai, Beijing, China; LLOD, 0.1 IU/mL).15 This
assay has previously been shown to have good reliability
and reproducibility.17,23 The ULN of ALT was 40 U/L for
men and 35 U/L for women.
Clinical Gastroenterology and Hepatology Vol. 17, No. 1
Statistical Analysis
Baseline was defined as the time of NA therapy
discontinuation. Follow-up time was calculated from
baseline until the occurrence of a particular study
endpoint, loss to follow-up, or last follow-up when
appropriate. Kaplan-Meier analysis and the log-rank test
were used to analyze longitudinal data. Cox proportional
hazards regression analysis was used to study the
association between variables and study endpoints.
Multivariable Cox regression models were adjusted for
start-of-treatment HBeAg status and covariates that
differed significantly between HBeAg-positive and
HBeAg-negative patients. Pearson correlation analysis
was used to study the correlation between anti-HBc and
other covariates.
The increase in anti-HBc level was calculated from
baseline level to the last available level. The Student
t test was used to compare the difference in increase. In
addition, we analyzed whether the off-treatment anti-
HBc level or the change in anti-HBc level at off-
treatment weeks 4, 8, 12, and 24 predicted clinical
relapse. For this analysis, follow-up time was calcu-
lated from the respective off-treatment week until
clinical relapse or last follow-up, and events occurring
at the respective off-treatment were excluded. Similar
analysis was performed for the change in off-treatment
HBsAg level. Because the kinetics of viral biomarkers
is usually non-linear, we modeled the off-treatment
anti-HBc kinetics with a biphasic exponential model
that considers both exponential changes and steady
state kinetics.24
Statistical tests were two-sided, with P value
<.05 considered as statistically significant. Tests were
performed with SPSS version 21.0 (IBM Corp, Armonk,
NY) and SAS software version 9.3 (SAS Institute Inc,
Cary, NC).
Results
Study Population
A total of 107 patients were included. At time of
treatment discontinuation, 7 patients were tested as
HBsAg-negative and were excluded from analysis.
Therefore, 100 patients were analyzed who were HBsAg-
positive, HBeAg-negative, and HBV DNA undetectable at
time of treatment discontinuation.
Seven patients were lost to follow-up. Patients who
did not develop a clinical relapse (thus continuing off-
treatment follow-up) were followed for a median of
2.5 years (interquartile range [IQR], 1.0–3.5). No patients
developed decompensated liver disease or HCC after
treatment discontinuation.
At start of NA treatment, 71 patients (71%) were
HBeAg-positive, whereas 29 patients (29%) were HBeAg-
negative (Table 1).
January 2019 Anti-HBc Level and Nucleos(t)ide Analogue Discontinuation 185

Table 1. Patient Characteristics (n ¼ 100) According to

HBeAg Status at Start of Treatment

HBeAgþ at HBeAg– at
start of start of
treatment treatment P
Characteristics N ¼ 71 N ¼ 29 value

Baseline (end of treatment)

Age, ya 33  7.5 41  8.0 <.001
Male sex 59 (83%) 27 (93%) .34
Transient elastography, 5.3 (4.8–6.5) 5.9 (4.6–6.3) .66
kPab
Peg-interferon 17 (24%) 4 (14%) .26
experienced
NA therapyc .83
First-line 31 (44%) 12 (41%)
Second-line 40 (56%) 17 (59%) Figure 1. Cumulative rate of clinical relapse (n ¼ 39) after NA
Therapy duration, yb 3.9 (2.6–5.8) 4.8 (2.5–6.4) .36 discontinuation in 100 chronic hepatitis B patients according
Consolidation therapy 2.2 (1.2–3.4) 2.9 (2.0–4.9) .019 to start-of-treatment HBeAg status.
duration, yb
Lab (serum)
ALT, xULNb 0.5 (0.4–0.7) 0.6 (0.5–0.8) .031 and HBeAg-negative patients. Further stratification of
HBV DNA, log IU/mLa UD UD —
HBsAg, log IU/mLa 2.8  0.9 2.5  1.0 .19
end-of-treatment anti-HBc levels showed that level
Anti-HBc, log IU/mLa 2.8  0.5 2.5  0.6 .050
1000 IU/mL was associated with the lowest risk of
Start of treatment clinical relapse (n ¼ 6/36, 21% at year 4) as compared
Lab (serum) with level between 100 and 999 IU/mL (n ¼ 21/48, 50%
ALT, xULNb 6.3 (2.7–9.2) 5.0 (3.0–10.9) .85 at year 4), whereas patients with anti-HBc level <100
HBV DNA, log IU/mLa 6.1  1.3 5.4  1.6 .014
IU/mL (n ¼ 11/13, 85% at year 4) harbored the highest
risk of clinical relapse (all P values <.05; Figure 2).
ALT, alanine aminotransferase; HBeAg, hepatitis B e antigen; NA, nucleos(t)ide The relation between factors and virologic relapse is
analogue; HBV, hepatitis B virus; HBsAg, hepatitis B surface antigen;
presented in Supplementary Table 1.
UD, undetectable (<20 IU/mL).
a
Mean standard deviation.
b
Median (IQR).
c
First-line: entecavir, tenofovir. Second-line: lamivudine, adefovir, telbivudine. Combined Use of Hepatitis B Surface Antigen
and Antibody Against the Hepatitis B Virus Core
Protein Level to Predict Relapse
Serum Antibody Against the Hepatitis B Virus
Core Protein Level and Relapse An end-of-treatment HBsAg level 100 IU/mL was
associated with lower risk of clinical relapse, yet only 18
A total of 39 patients developed a clinical relapse
patients (18%) had level 100 (Figure 3A). Therefore,
resulting in year 1, 2, and 4 rates of 22%, 32%, and 43%,
the majority of patients (82%) were at high risk of
respectively, for start-of-treatment HBeAg-positive
relapse. Among these patients, end-of-treatment anti-
patients and 24%, 39%, and 58%, respectively, for start-
HBc level could further identify patients at low risk of
of-treatment HBeAg-negative patients (Figure 1). Viro-
clinical relapse (per log IU/mL increase: hazard ratio,
logic relapse (HBV DNA elevation >2000 IU/mL
0.39; 95% confidence interval, 0.21–0.75; P ¼ .005). At
measured once) was observed in 76 patients (85% at
year 4, 27% of the patients with anti-HBc level
1000
year 4). In total, 6 patients (3 start-of-treatment HBeAg-
IU/mL developed clinical relapse as compared with 64%
positive and 3 start-of-treatment HBeAg-negative
of the patients with anti-HBc level <1000 IU/mL
patients) developed HBsAg seroclearance after NA ther-
(P ¼ .009; Figure 3B and C).
apy discontinuation, resulting in a cumulative rate of 10%
In contrast, among patients with end-of-treatment
at year 4 (Supplementary Figure 1). At time of clinical
anti-HBc level between 100 and 999 IU/mL who had
relapse, the mean HBV DNA level was 5.9 log IU/mL, and
intermediate risk of relapse (Figure 2A), HBsAg level
the median ALT level was 4.0 times the ULN.
100 IU/mL was not significantly associated with
Younger age, low end-of-treatment HBsAg level, and
increased likelihood of sustained response (P ¼ .24).
high end-of-treatment anti-HBc level were associated
with significantly reduced risk of clinical relapse in
multivariable analysis (Table 2). All 3 factors remained Off-Treatment Antibody Against the Hepatitis
statistically significant in sensitivity analysis adjusted for B Virus Core Protein Kinetics
other characteristics (end-of-treatment ALT level and
start-of-treatment HBV DNA level), which differed Patients with clinical relapse showed a strong
significantly between start-of-treatment HBeAg-positive increase in anti-HBc level after treatment
186 Chi et al Clinical Gastroenterology and Hepatology Vol. 17, No. 1

Table 2. Cox Proportional Hazards Regression Analysis of Clinical Relapse

Clinical relapse (n ¼ 39)

Univariable Multivariable

HR 95% CI P value HR 95% CI P value

End of treatment
Age, per year 1.06 1.02–1.10 .005 1.06 1.01–1.11 .025
Age 35 y 0.44 0.23–0.86 .017
Female sex 1.00 0.35–2.83 1.00
First-line NAa 0.99 0.52–1.87 .97
Consolidation duration, per month 1.00 0.98–1.01 .66 0.98 0.96–1.00 .071
ALT, xULN 0.72 0.26–2.00 .53
HBsAg level, per log IU/mL 1.38 0.93–2.05 .11 1.71 1.05–2.81 .032
HBsAg 100 IU/mL 0.30 0.09–0.97 .045
HBsAg 200 IU/mL 0.50 0.22–1.14 .099
HBsAg 1000 IU/mL 1.11 0.59–2.10 .74
Anti-HBc level, per log IU/mL 0.38 0.21–0.70 .002 0.31 0.15–0.65 .002
Anti-HBc level
<100 IU/mL Reference
100–999 IU/mL 0.37 0.18–0.77 .008

1000 IU/mLb 0.14 0.05–0.37 <.001
Start of treatment
HBeAg positivity 0.72 0.37–1.40 .33 1.08 0.50–2.33 .85
HBV DNA, log IU/mL 0.98 0.77–1.24 .85
ALT, xULN 0.99 0.95–1.03 .61

ALT, alanine aminotransferase; NA, nucleos(t)ide analogue; CI, confidence interval; HBeAg, hepatitis B e antigen; HBsAg, hepatitis B surface antigen;
HBV, hepatitis B virus; HR, hazard ratio; ULN, upper limit of normal.
a
Entecavir/tenofovir.
b
End-of-treatment anti-HBc level
1000 IU/mL vs 100–999 IU/mL: HR, 0.37; 95% CI, 0.15–0.91; P ¼ .031.

discontinuation, whereas the anti-HBc level remained
stable in patients with sustained response (Figure 4A).
Patients who developed a clinical relapse had an esti-
mated off-treatment anti-HBc increase of 3.6 log IU/mL
per year compared with an increase of 0.5 log IU/mL per
year in patients with sustained response (Figure 4B).
Off-treatment anti-HBc kinetics was not significantly
associated with relapse, except for the anti-HBc level at
off-treatment week 4 (Supplementary Table 2).
Discussion
This prospective study demonstrates the role of
quantitative anti-HBc levels in patients with chronic HBV
infection who discontinued NA therapy. A high anti-HBc
level at discontinuation (
1000 IU/mL) was associated
with a lower risk of clinical relapse, also after adjustment
for other important co-factors. In addition, an end-of-
treatment HBsAg level 100 IU/mL was associated with
a lower risk of clinical relapse, yet the majority of patients
(82%) in our study had higher levels. In this subgroup
with end-of-treatment HBsAg level >100 IU/mL, the end-
of-treatment anti-HBc level was able to further stratify the
risk of relapse. Our findings underline the promising po-
tential of quantitative anti-HBc as a serum marker that
may partially reflect the host’s immune response.
Our results suggest that the end-of-treatment anti-HBc
level may be used to inform the decision of NA treatment
discontinuation. This is in line with previous studies in
which a high anti-HBc level was associated with a higher
likelihood of response to (peg)interferon or NA treat-
ment.17,18,25 Because of this, quantitative anti-HBc testing
may have a broad utility potential in clinical practice. In
addition, HBsAg level is related with the amount of
infected hepatocytes and intrahepatic covalently closed
circular DNA,26,27 and we have demonstrated that a low
end-of-treatment HBsAg level 100 IU/mL was associ-
ated with a lower risk of clinical relapse, consistent with
results of our previous studies.2,4,21
Our results suggest that a high level of anti-HBc may
be correlated with a more enhanced HBV-specific
immune response. It has been shown that HBV-specific
B cells can be activated through T-cell–dependent and
T-cell–independent pathways, which in turn produce
anti-HBc.20 Activated HBV-specific B cells were shown to
have a highly potent antigen-presenting function that
also activates naive T-helper cells, augmenting viral
control. High anti-HBc level may represent a larger
amount of activated HBV-specific B cells. Accordingly, a
high level of anti-HBc has been associated with a better
response to both peg-interferon and NA treatment.17,18,25
In addition, studies have suggested that the recognition
and targeting of hepatitis B core antigen by the adaptive
immune system play an important role in viral clearance.
For instance, there is evidence suggesting that HBeAg, a
secreted viral antigen with no clear function in the virus’
replication cycle, diverts the attention of the adaptive
January 2019
Figure 2. (A) Cumulative
rate of clinical relapse
after NA discontinuation
according to end-of-
treatment anti-HBc level.
(B) Probability of clinical
relapse after NA discon-
tinuation according to
end-of-treatment anti-HBc
level.
immune response from hepatitis B core antigen to
HBeAg.11,28,29
After treatment discontinuation, a strong increase in
anti-HBc level was seen in patients with a clinical relapse
as compared with those without. We hypothesize that
this is most likely due to the reactivation of viral repli-
cation and subsequent immune-mediated cytolysis of
hepatocytes, resulting in the release of a substantial
amount of hepatitis B core protein and concurrent pro-
duction of anti-HBc by B cells.30 Previously, it has been
shown that the ALT level is positively correlated with the
anti-HBc level in untreated patients with chronic HBV
infection.19,23 During off-treatment follow-up, we found
that the anti-HBc level was only correlated with the ALT
level in patients who developed a clinical relapse (data
not shown), also suggesting that the off-treatment in-
crease in anti-HBc level is more likely to be related to
cytolytic activity. This may explain why the off-treatment
anti-HBc level did not predict relapse. It is plausible that
IgM anti-HBc is mainly responsible for the increase of
Anti-HBc Level and Nucleos(t)ide Analogue Discontinuation 187
total anti-HBc during a relapse because IgM anti-HBc is
associated with acute exacerbation, whereas during NA
suppression, IgG anti-HBc is primarily detected.11,13,14
The strength of our study is the well-defined cohort of
patients who were prospectively and frequently assessed
after stopping NA with standardized retreatment criteria.
Nevertheless, our study also has limitations. First, anti-
HBc levels before treatment discontinuation were not
available. It would be valuable to investigate the role of
the anti-HBc level at start of NA and its kinetics during
NA therapy on treatment discontinuation. Also, the
used cutoffs of anti-HBc level were not pre-specified.
However, cutoffs as used in previous studies were
considered in the analysis.17,23 HBV genotype was not
available in our study. However, because all patients
were Asian, the genotypes were likely to be B or C. A
previous study did not find these 2 genotypes to be
associated with relapse after treatment discontinua-
tion.31 In addition, HBeAg-positive and HBeAg-negative
patients may differ in characteristics, potentially
188 Chi et al Clinical Gastroenterology and Hepatology Vol. 17, No. 1

Figure 3. (A) Cumulative rate

of clinical relapse after NA
discontinuation according to
end-of-treatment HBsAg level.
(B) Cumulative rate of clinical
relapse after NA discontinuation
according to end-of-treatment
anti-HBc levels in 82 patients with
end-of-treatment HBsAg level
>100 IU/mL. (C) Probability of
clinical relapse after NA discon-
tinuation according to the end-of-
treatment HBsAg and anti-HBc
levels.
January 2019 Anti-HBc Level and Nucleos(t)ide Analogue Discontinuation 189

Figure 4. (A) Serum off-

treatment anti-HBc levels
in sustained responders
and clinical relapsers
as estimated by a
biphasic exponential
model. P values describe
the difference in mean
anti-HBc level between
sustained responders and
clinical relapsers. (B) Indi-
vidual off-treatment serum
anti-HBc levels in (top)
sustained responders and
(bottom) clinical relapsers.
*P < .001 for the compar-
ison of sustained response
vs clinical relapse.

contributing to heterogeneity. On the other hand, HBeAg
status was not associated with relapse in multivariable
analysis, and sensitivity analyses were performed to
account for the difference in characteristics. Also, it
is important to consider that anti-HBc level had a limited
predictive value on the endpoint virologic relapse
(HBV DNA elevation >2000 IU/mL measured once).
Currently, however, there is no consensus on the defi-
nition of relapse after treatment discontinuation.
However, an increasing number of studies have shown
that HBV DNA elevation after cessation can be tran-
sient,4,8,9 and that some patients may achieve sustained
response, an increased decline of HBsAg level, or even
HBsAg loss after an initial virologic reactivation.9,32
Nonetheless, at time of clinical relapse, some patients
in our study had high levels of HBV DNA and ALT.
Therefore, it remains important to follow patients
carefully and to not attempt discontinuation in patients
with cirrhosis. We have shown in another study that
patients with very high HBV DNA (>200,000 IU/mL) or
patients with persistent elevated HBV DNA (>2000 IU/
mL) after treatment cessation may benefit from early
retreatment because they were likely to develop a
biochemical relapse.4 Finally, the investigated anti-HBc
assay measured the total anti-HBc and not IgM and
IgG anti-HBc levels separately. From an immunologic
190 Chi et al
aspect, it would be interesting to study the levels of
IgM and IgG anti-HBc separately because the level of
each type of antibody may have different predictive
property.
In conclusion, end-of-treatment anti-HBc level pre-
dicted clinical relapse in patients with chronic HBV
infection who discontinued NA therapy. Low end-of-
treatment HBsAg level was associated with a lower
risk of relapse, but the majority of patients in our
cohort remained at high risk of relapse according
to HBsAg level
100 IU/mL, whereas the anti-HBc
level was able to further stratify the risk of relapse in
this group of patients. Anti-HBc alone or in combina-
tion with HBsAg appears to play an important role
in the selection of patients who are suitable for
treatment discontinuation and therefore warrant
further validation.
Supplementary Material
Note: To access the supplementary material accom-
panying this article, visit the online version of Clinical
Gastroenterology and Hepatology at www.cghjournal.org,
and at https://doi.org/10.1016/j.cgh.2018.05.047.
References
1. European Association for the Study of the Liver. Electronic
address eee, European Association for the Study of the Liver:
EASL 2017 Clinical Practice Guidelines on the management of
hepatitis B virus infection. J Hepatol 2017;67:370–398.
2. Chi H, Hansen BE, Yim C, et al. Reduced risk of relapse
after long-term nucleos(t)ide analogue consolidation therapy
for chronic hepatitis B. Aliment Pharmacol Ther 2015;
41:867–876.
3. Seto WK, Hui AJ, Wong VW, et al. Treatment cessation of
entecavir in Asian patients with hepatitis B e antigen negative
chronic hepatitis B: a multicentre prospective study. Gut 2015;
64:667–672.
4. Cao J, Chi H, Yu T, et al. Off-treatment hepatitis B virus (HBV)
DNA levels and the prediction of relapse after discontinuation
of nucleos(t)ide analogue therapy in patients with chronic
hepatitis B: a prospective Stop study. J Infect Dis 2017;
215:581–589.
5. Chi H, Wong D, Peng J, et al. Durability of response after
hepatitis B surface antigen seroclearance during nucleos(t)ide
analogue treatment in a multiethnic cohort of chronic hepatitis B
patients: results after treatment cessation. Clin Infect Dis 2017;
65:680–683.
6. Zoutendijk R, Hansen BE, van Vuuren AJ, et al. Serum HBsAg
decline during long-term potent nucleos(t)ide analogue therapy
for chronic hepatitis B and prediction of HBsAg loss. J Infect Dis
2011;204:415–418.
7. Terrault NA, Bzowej NH, Chang KM, et al. AASLD guidelines
for treatment of chronic hepatitis B. Hepatology 2016;
63:261–283.
8. Hadziyannis SJ, Sevastianos V, Rapti I, et al. Sustained
responses and loss of HBsAg in HBeAg-negative patients with
chronic hepatitis B who stop long-term treatment with adefovir.
Gastroenterology 2012;143:629–636 e621.
Clinical Gastroenterology and Hepatology Vol. 17, No. 1
9. Berg T, Simon KG, Mauss S, et al. Long-term response
after stopping tenofovir disoproxil fumarate in non-cirrhotic
HBeAg-negative patients: FINITE study. J Hepatol 2017;
67:918–924.
10. Seeger C, Mason WS. Molecular biology of hepatitis B virus
infection. Virology 2015;479–480:672–686.
11. Diepolder HM, Ries G, Jung MC, et al. Differential antigen-
processing pathways of the hepatitis B virus e and core pro-
teins. Gastroenterology 1999;116:650–657.
12. Grob P, Jilg W, Bornhak H, et al. Serological pattern “anti-HBc
alone”: report on a workshop. J Med Virol 2000;62:450–455.
13. Colloredo Mels G, Bellati G, Leandro G, et al. Role of IgM
antibody to hepatitis B core antigen in the diagnosis of hepatitis
B exacerbations. Arch Virol Suppl 1993;8:203–211.
14. Brunetto MR, Cerenzia MT, Oliveri F, et al. Monitoring the natural
course and response to therapy of chronic hepatitis B with an
automated semi-quantitative assay for IgM anti-HBc. J Hepatol
1993;19:431–436.
15. Li A, Yuan Q, Huang Z, et al. Novel double-antigen sandwich
immunoassay for human hepatitis B core antibody. Clin Vaccine
Immunol 2010;17:464–469.
16. Yuan Q, Song LW, Cavallone D, et al. Total hepatitis B core
antigen antibody, a quantitative non-invasive marker of
hepatitis B virus induced liver disease. PLoS One 2015;
10:e0130209.
17. Fan R, Sun J, Yuan Q, et al. Baseline quantitative hepatitis B
core antibody titre alone strongly predicts HBeAg seroconver-
sion across chronic hepatitis B patients treated with peginter-
feron or nucleos(t)ide analogues. Gut 2016;65:313–320.
18. Xu JH, Song LW, Li N, et al. Baseline hepatitis B core
antibody predicts treatment response in chronic hepatitis
B patients receiving long-term entecavir. J Viral Hepat 2017;
24:148–154.
19. Yuan Q, Song LW, Liu CJ, et al. Quantitative hepatitis B core
antibody level may help predict treatment response in chronic
hepatitis B patients. Gut 2013;62:182–184.
20. Milich DR, Chen M, Schodel F, et al. Role of B cells in antigen
presentation of the hepatitis B core. Proc Natl Acad Sci U S A
1997;94:14648–14653.
21. Chen CH, Lu SN, Hung CH, et al. The role of hepatitis B surface
antigen quantification in predicting HBsAg loss and HBV relapse
after discontinuation of lamivudine treatment. J Hepatol 2014;
61:515–522.
22. Liaw YF, Kao JH, Piratvisuth T, et al. Asian-Pacific consensus
statement on the management of chronic hepatitis B: a 2012
update. Hepatology International 2012;6:531–561.
23. Song LW, Liu PG, Liu CJ, et al. Quantitative hepatitis B core
antibody levels in the natural history of hepatitis B virus infec-
tion. Clin Microbiol Infect 2015;21:197–203.
24. Wolters LM, Hansen BE, Niesters HG, et al. Viral dynamics
during and after entecavir therapy in patients with chronic
hepatitis B. J Hepatol 2002;37:137–144.
25. Hou FQ, Song LW, Yuan Q, et al. Quantitative hepatitis B core
antibody level is a new predictor for treatment response in
HBeAg-positive chronic hepatitis B patients receiving peginter-
feron. Theranostics 2015;5:218–226.
26. Manesis EK, Papatheodoridis GV, Tiniakos DG, et al. Hepatitis B
surface antigen: relation to hepatitis B replication parameters in
HBeAg-negative chronic hepatitis B. J Hepatol 2011;55:61–68.
27. Thompson AJ, Nguyen T, Iser D, et al. Serum hepatitis B surface
antigen and hepatitis B e antigen titers: disease phase
January 2019
influences correlation with viral load and intrahepatic hepatitis B
virus markers. Hepatology 2010;51:1933–1944.
28. Chen MT, Billaud JN, Sallberg M, et al. A function of the hepatitis
B virus precore protein is to regulate the immune response to
the core antigen. Proc Natl Acad Sci U S A 2004;
101:14913–14918.
29. Milich DR, Chen MK, Hughes JL, et al. The secreted hepatitis B
precore antigen can modulate the immune response to the
nucleocapsid: a mechanism for persistence. J Immunol 1998;
160:2013–2021.
30. Maini MK, Boni C, Lee CK, et al. The role of virus-specific
CD8(þ) cells in liver damage and viral control during persistent
hepatitis B virus infection. J Exp Med 2000;191:1269–1280.
31. Jeng WJ, Sheen IS, Chen YC, et al. Off-therapy durability of
response to entecavir therapy in hepatitis B e antigen-negative
chronic hepatitis B patients. Hepatology 2013;58:1888–1896.
32. Honer Zu Siederdissen C, Rinker F, Maasoumy B, et al. Viral and
host responses after stopping long-term nucleos(t)ide analogue
therapy in HBeAg-negative chronic hepatitis B. J Infect Dis
2016;214:1492–1497.
Anti-HBc Level and Nucleos(t)ide Analogue Discontinuation 191
Reprint requests
Address requests for reprints to: Jie Peng, MD, Hepatology Unit, Nanfang
Hospital, Southern Medical University, 1838 Guangzhou Avenue North,
Guangzhou, Guangdong Province, China. e-mail: pjie138@163.com;
fax: þ86-20-8771-9653; Harry L. A. Janssen, MD, PhD, University Health
Network, Toronto General Hospital, 200 Elizabeth Street, Eaton Building 9th
Floor Room 234 (9EB234), Toronto, ON, Canada M5G 2C4. e-mail: harry.
janssen@uhn.ca; fax: 0086 20 62786530.
Acknowledgments
The authors thank all the patients in this study and the nurses who assisted in
the patient management and the collection of serum samples.
Conflicts of interest
These authors disclose the following: B. E. Hansen has received grants from
and is consultant for Intercept Pharmaceuticals. H. L. A. Janssen has received
grants from and is consultant for Abbott, Bristol-Myers Squibb, Gilead Sci-
ences, Merck, Novartis, Roche, and Janssen. J. Hou has received consulting
fees from Roche, Novartis, GSK, and Bristol-Myers Squibb and has received
grant/research support from Roche, Novartis, and GSK. The remaining authors
disclose no conflicts.
Funding
Supported by grants from the National Science Foundation of China (no.
81572000) and Foundation for Liver Research (Rotterdam, the Netherlands).
191.e1 Chi et al Clinical Gastroenterology and Hepatology Vol. 17, No. 1

20 Supplementary Table 2. Association Between Off-Treatment

HBsAg seroclearance (%)

Anti-HBc Level and Clinical Relapse
Cumulative rate of
15
Univariable Cox regression
10%
10 Off-treatment anti-HBc kinetics HR 95% CI P value

Week 4
5 Anti-HBc level, per log IU/mL 0.36 0.20–0.67 .001
Change in anti-HBc level, 0.64 0.07–5.70 NS
0 per log IU/mL
Anti-HBc increase, vs decrease 0.64 0.32–1.27 NS
0 1 2 3 4
Week 8
Follow-up (years) Anti-HBc level, per log IU/mL 0.58 0.28–1.19 NS
No. at risk 100 63 43 26 12 Change in anti-HBc level, 1.61 0.82–3.15 NS
per log IU/mL
Supplementary Figure 1. Cumulative rate of HBsAg loss
Anti-HBc increase, vs decrease 1.43 0.60–3.37 NS
after NA discontinuation. In total, 6 patients developed
Week 12
HBsAg loss, of whom 3 were HBeAg-positive and 3 HBeAg-
Anti-HBc level, per log IU/mL 1.02 0.54–1.93 NS
negative at start of treatment.
Change in anti-HBc level, 1.40 0.80–2.45 NS
per log IU/mL
Anti-HBc increase, vs decrease 0.68 0.29–1.58 NS
Week 24
Anti-HBc level, per log IU/mL 1.47 0.69–3.13 NS
Change in anti-HBc level, 1.71 0.87–3.36 NS
per log IU/mL
Anti-HBc increase, vs decrease 2.30 0.53–10.06 NS

NOTE. The change in anti-HBc at each respective week was compared with
the end-of-treatment anti-HBc level.
CI, confidence interval; HR, hazard ratio; NS, not significant (P > .05).

Supplementary Table 1. Univariable Cox Proportional

Hazards Regression Analysis of
Virologic Relapse

Virologic relapse (n ¼ 76)

Univariable

HR 95% CI P value

End of treatment
Age, per year 1.02 0.99–1.04 .21
Age 35 y 0.61 0.38–0.96 .033
Female sex 1.09 0.54–2.20 .80
First-line NAa 0.60 0.38–0.95 .030
Consolidation duration, per month 1.00 0.99–1.02 .50
ALT, xULN 1.25 0.62–2.49 .53
HBsAg level, per log IU/mL 1.30 0.99–1.69 .06
HBsAg 100 IU/mL 0.52 0.27–1.02 .055
HBsAg 200 IU/mL 0.56 0.32–0.98 .041
HBsAg 1000 IU/mL 1.11 0.71–1.75 .64
Anti-HBc level, per log IU/mL 0.69 0.45–1.06 .088
Anti-HBc level
<100 IU/mL Reference
100–999 IU/mL 0.79 0.41–1.52 .49

1000 IU/mLb 0.51 0.25–1.02 .055
Start of treatment
HBeAg positivity 0.88 0.54–1.44 .62
HBV DNA, log IU/mL 0.86 0.73–1.03 .10
ALT, xULN 0.99 0.96–1.02 .36

CI, confidence interval; HR, hazard ratio.

a
Entecavir/tenofovir.
b
End-of-treatment anti-HBc level
1000 IU/mL vs 100–999 IU/mL: HR, 0.64;
95% CI, 0.38–1.06; P ¼ .082.