Professional Documents
Culture Documents
1 s2.0 S1386653218300970 Main
1 s2.0 S1386653218300970 Main
A R T I C LE I N FO A B S T R A C T
Keywords: Background: To avoid false negative results, hepatitis B surface antigen (HBsAg) assays need to detect samples
Hepatitis B with mutations in the immunodominant ‘a’ determinant region, which vary by ethnographic region.
HBsAg mutations Objective: We evaluated the prevalence and type of HBsAg mutations in a hepatitis B virus (HBV)-infected East-
Detection and Southeast Asian population, and the diagnostic performance of the Elecsys® HBsAg II Qualitative assay.
Screening
Study design: We analyzed 898 samples from patients with HBV infection from four sites (China [Beijing and
Immunoassay
Immunodominant ‘a’ determinant region
Guangzhou], Korea and Vietnam). HBsAg mutations were detected and sequenced using highly sensitive ultra-
deep sequencing and compared between the first (amino acids 124–137) and second (amino acids 139–147)
loops of the ‘a’ determinant region using the Elecsys® HBsAg II Qualitative assay.
Results: Overall, 237 distinct amino acid mutations in the major hydrophilic region were identified; mutations
were present in 660 of 898 HBV-infected patient samples (73.5%). Within the pool of 237 distinct mutations, the
majority of the amino acid mutations were found in HBV genotype C (64.8%). We identified 25 previously
unknown distinct mutations, mostly prevalent in genotype C-infected Korean patients (n = 18) followed by
Chinese (n = 12) patients. All 898 samples were correctly identified by the Elecsys® HBsAg II Qualitative assay.
Conclusions: We observed 237 distinct (including 25 novel) mutations, demonstrating the complexity of HBsAg
variants in HBV-infected East- and Southeast Asian patients. The Elecsys® HBsAg II Qualitative assay can reliably
detect HBV-positive samples and is suitable for routine diagnostic use in East and Southeast Asia.
Abbreviations: HBsAg, hepatitis B surface antigen; HBV, hepatitis B virus; MHR, major hydrophilic region
⁎
Corresponding author.
E-mail address: wolfgang.kaminski@bioscientia.de (W.E. Kaminski).
1
Professor Nauck is recently deceased.
https://doi.org/10.1016/j.jcv.2018.04.005
Received 22 January 2018; Received in revised form 29 March 2018; Accepted 4 April 2018
1386-6532/ © 2018 Roche Diagnostics International Ltd. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/BY-NC-ND/4.0/).
H.S. Kim et al. Journal of Clinical Virology 103 (2018) 48–56
Sum A B C D E F G D144E 5 0 0 5 0 0 0 0
S136Y 5 0 0 5 0 0 0 0
K122R 201 0 186 15 0 0 0 0 G119R 5 0 1 4 0 0 0 0
I126T 187 0 0 187 0 0 0 0
I126S 97 0 0 97 0 0 0 0
Y161F 55 0 55 0 0 0 0 0
differences of > 8% in genome sequence; 10 HBV genotypes (A–J) have
M133T 46 0 12 34 0 0 0 0
T131N 41 0 6 35 0 0 0 0 been identified to date, with distinct geographic distributions [3]. For
T140I 41 0 39 2 0 0 0 0 example, genotype C is primarily found in South-East Asia, and geno-
R160K 38 0 0 38 0 0 0 0 type I has been reported in Vietnam and Laos [3]. In China, genotype B
A159V 35 0 13 22 0 0 0 0 is more prevalent in the south of the country, whereas genotype C is
G145R 34 0 6 28 0 0 0 0
Q101R 31 0 3 28 0 0 0 0
more prevalent in the north [4,5]. These genotypes are clinically im-
T126A 29 0 29 0 0 0 0 0 portant, as they are associated with differences in genetic mutations
I110L 28 0 28 0 0 0 0 0 (defined herein as amino acid substitutions that differ from the geno-
P127T 28 0 19 9 0 0 0 0 type-specific consensus sequence), disease progression and treatment
V168A 26 0 5 20 0 0 0 1
response [3].
Y100C 26 0 3 23 0 0 0 0
L109P 21 0 16 5 0 0 0 0 HBV is transmitted via blood and other bodily fluids and so can
I126V 20 0 0 20 0 0 0 0 threaten the safety of blood transfusion services, particularly in coun-
Q101K 18 0 0 18 0 0 0 0 tries where HBV is endemic [6,7]. The hepatitis B surface antigen
P120T 18 0 6 12 0 0 0 0 (HBsAg) is the primary serologic marker for HBV detection [8,9]. The
M133L 17 0 14 3 0 0 0 0
P120S 17 0 15 2 0 0 0 0
majority of HBsAg immunoassays target the immunodominant ‘a’ de-
T113S 16 0 0 16 0 0 0 0 terminant region, which is located in the major hydrophilic region
I126N 15 0 0 15 0 0 0 0 (MHR) and is divided into two loops of amino acid residues (the first
T123A 15 0 8 7 0 0 0 0 loop comprises amino acids 124–137; the second loop comprises amino
T123N 14 0 0 14 0 0 0 0
acids 139–147) [8,10]. Commercially available HBsAg assays must be
M103I 14 0 0 14 0 0 0 0
G145A 14 0 0 14 0 0 0 0 able to accurately detect samples with known mutations in the ‘a’ de-
Q101H 13 0 1 12 0 0 0 0 terminant region, such as the common vaccine-induced immune-escape
Y100S 12 0 4 8 0 0 0 0 variant G145R [8], to avoid false-negative results [10–12].
T125M 12 0 1 11 0 0 0 0 In a recent global study of 1553 HBsAg-positive blood samples from
E164G 12 0 10 2 0 0 0 0
L110I 12 0 0 12 0 0 0 0
20 countries across Africa, America, Asia and Europe, we observed
S114A 11 0 1 10 0 0 0 0 HBsAg MHR mutations in 72.8% of sequenced samples, three times
G130R 11 0 1 10 0 0 0 0 higher than previously reported. The highest prevalence of mutations in
F170S 11 0 1 10 0 0 0 0 the ‘a’ determinant region was observed among samples from Asia [13].
F134I 10 0 2 8 0 0 0 0
We also demonstrated that the Elecsys® HBsAg II Qualitative assay can
F134L 10 0 1 9 0 0 0 0
T116N 10 0 1 9 0 0 0 0 reliably detect HBV-positive samples with known in vivo mutations in
A166V 10 0 6 4 0 0 0 0 the ‘a’ determinant region.
M133I 9 0 2 7 0 0 0 0
T126S 8 0 8 0 0 0 0 0
G130N 8 0 0 8 0 0 0 0 2. Objectives
Q129N 8 0 1 7 0 0 0 0
W156G 8 0 0 8 0 0 0 0 We sought to evaluate the prevalence of HBsAg mutations in an
S114T 7 0 2 5 0 0 0 0
East- and Southeast Asian chronic HBV-infected population. This con-
G145E 7 0 0 7 0 0 0 0
F134S 7 0 3 4 0 0 0 0 sisted of Korean and Vietnamese patients from the global study [13],
T118K 7 0 0 7 0 0 0 0 and a new cohort of patients from China. A further objective was to
T131P 7 0 1 6 0 0 0 0 determine the diagnostic performance of the Elecsys® HBsAg II Quali-
E164D 7 0 0 7 0 0 0 0 tative assay in these patients. This East- and Southeast Asian population
T140S 7 0 1 6 0 0 0 0
T131I 6 0 1 5 0 0 0 0
was chosen due to the high rates of endemic HBV infection in the re-
Y100F 6 0 4 2 0 0 0 0 gion; a systematic review of articles published between 1965 and 2013
S167L 6 0 3 3 0 0 0 0 reported an estimated prevalence of HBsAg seropositivity in China,
F134V 6 0 1 5 0 0 0 0 Korea and Vietnam of 5.49%, 4.36% and 10.79%, respectively, com-
G130E 6 0 1 5 0 0 0 0
pared with a global prevalence of 3.61% [1].
Q129H 6 0 5 1 0 0 0 0
W156L 6 0 2 4 0 0 0 0
Q129R 6 0 4 2 0 0 0 0 3. Study design
Y161S 6 0 6 0 0 0 0 0
S154P 6 0 1 5 0 0 0 0
P111L 6 0 0 6 0 0 0 0 The global study included samples from 1553 patients with HBV
A128V 5 0 0 5 0 0 0 0 infection obtained between January 2007 and February 2015 (Africa,
S143T 5 0 0 5 0 0 0 0 n = 435; Asia, n = 653; Europe, n = 72; Latin America, n = 35; Middle
S113T 5 0 4 0 1 0 0 0
East, n = 79; USA, n = 234; unknown, n = 45) excluding the samples
T113N 5 0 0 5 0 0 0 0
T143M 5 0 5 0 0 0 0 0 (n = 407) collected in China [13]. Of these, 562 samples were from
T116A 5 0 0 5 0 0 0 0 patients with chronic HBV infection (diagnosed with HBV > 6 months
C147Y 5 0 0 5 0 0 0 0 before enrollment) and 991 samples were defined as unselected random
S136F 5 0 0 5 0 0 0 0 HBV (samples with HBV DNA [ > 100 IU/ml] or HBsAg positivity ob-
tained from clinical settings in Europe, South Africa and the USA). The
49
H.S. Kim et al. Journal of Clinical Virology 103 (2018) 48–56
Fig. 1. Distribution of location-specific mutations in the ‘a' determinant region (amino acids 124–147) according to HBV genotype in A) Beijing, B) Guangzhou, C)
Korea and D) Vietnam, respectively.
50
H.S. Kim et al. Journal of Clinical Virology 103 (2018) 48–56
Fig. 1. (continued)
global study did not include the 407 samples from Chinese HBV-in- from the global study of patients from Korea (Severance Hospital,
fected patients that are reported in this current analysis. Seoul, Korea) and Vietnam (Ho Chi Minh City University Medical
We report the results for a combined population with chronic HBV Center, Ho Chi Minh City, Vietnam); and a new cohort of HBV-infected
infection from four sites in East- and Southeast Asia: two subgroups patients from two sites in China (You’an Hospital, Capital Medical
51
H.S. Kim et al. Journal of Clinical Virology 103 (2018) 48–56
Table 2
Selected mutations with clinical impact and their within one-sample mutation frequency by region/country.
Location of mutation Beijing Guangzhou Korea Vietnam
University Beijing and Guangzhou No. 8 People’s Hospital, Guangzhou, were genotype C. The vaccine-induced immune-escape mutation G145R
China). Blood samples were collected with informed consent from each was identified in 17 samples (Beijing, n = 13; Guangzhou, n = 4). Se-
patient between January 2007 and February 2015. The study protocol lected mutations associated with a clinical impact, and their within one-
was approved by the Ethics Committees of the participating hospitals. sample mutation frequency, are shown in Table 2. Occult HBV infec-
The detection and sequencing of HBsAg mutations was conducted as tion-associated Y100C mutations, which are located outside the ‘a’
previously described using highly sensitive ultra-deep sequencing [13]. determinant region, were identified in 12 samples (Beijing, n = 10;
The HBV genotypes of samples were determined by phylogenetic ana- Guangzhou, n = 2). The distribution of 317 mutations (Beijing,
lysis [14,15]. The mutation rates of the first and second loops of the n = 164; Guangzhou, n = 153) outside the ‘a’ determinant region in the
immunodominant ‘a’ determinant region (amino acids 124–137 and Chinese samples, according to amino acid position, is shown in Fig. 2A
139–147, respectively) were compared to explore location-dependent and B.
aspects of the mutations; genotype-specific reference sequences used for All of the samples from Korean patients were genotype C, except for
bioinformatic analyses are listed in Supplementary Table 1 [16]. The one, which was genotype B. Of the 233 samples analyzed, 146 (63%)
frequency of occult HBV infection-associated Y100C mutations was also contained a total of 604 mutations (Supplementary Table 5); 323 mu-
assessed. tations were located in the ‘a’ determinant region (Fig. 1C) and 272
Samples with and without mutated HBsAg were tested using the mutations were found outside the ‘a’ determinant region (Fig. 2C). The
fully automated Roche Elecsys® HBsAg II Qualitative assay (Roche most common variant was I126T, which was observed in 74 samples, all
Diagnostics GmbH, Mannheim, Germany) according to the manu- with genotype C. The vaccine-induced immune-escape mutation G145R
facturer’s recommendations. The results were interpreted using the was identified in nine samples and occult HBV infection-associated
following cut-off reference values: < 0.9, negative; ≥1.0, positive; Y100C mutations were identified in six samples. Selected mutations
≥0.9 to < 1.0, borderline. Confirmatory testing was not performed as associated with a clinical impact, and their within one-sample mutation
all samples were obtained from HBV DNA-positive patients. frequency, are shown in Table 2.
All statistical analyzes were conducted independently by the AIT The majority of samples from Vietnamese patients were genotype B
Austrian Institute of Technology (Vienna, Austria) using a multistep (196 [76%]), followed by genotype C (61 [24%]). Of the 258 samples
approach (R, Python version 2.7.6 scripts using the NumPy package). analyzed, 233 (90%) contained a total of 493 mutations
The matplotlib package was used for graphical visualizations. (Supplementary Table 6), of which 152 mutations were observed in the
‘a’ determinant region (Fig. 1D) and 260 mutations were located out-
4. Results side the ‘a’ determinant region (Fig. 2D). The most common variant was
K122R, which was observed in 179 samples; all of these samples were
A total of 898 blood samples from patients with chronic HBV in- genotype B. G145R and Y100C mutations were each identified in eight
fection were included: 407 samples from China (Beijing, n = 281; samples. Some selected mutations associated with a clinical impact, and
Guangzhou, n = 126); 233 samples from Korea; and 258 samples from their within one-sample mutation frequency, are shown in Table 2.
Vietnam. Mutations were present in 660 of these samples (73.5%). In Comparison of immunodominant ‘a’ determinant region mutations
total, 237 distinct amino acid mutations in the MHR were identified. of genotype B- and C-infected subjects revealed some differences de-
The most frequently occurring mutations in the whole East- and pending on the country of origin. Investigation of the individual gen-
Southeast Asian cohort (≥5 samples) are presented in Table 1 and a otype-B immunodominant ‘a’ region mutations revealed that positions
comprehensive list of all mutations identified, and those identified in 131, 133, 134 and 143 were more frequently mutated in Vietnamese
each region/country cohort, are provided in Supplementary Tables 2–6. subjects than in Chinese subjects. The individual mutations’ rates were
Three patients (two from Guangzhou and one from Vietnam) were co- 7.1%, 22.6%, 13.1% and 4.8%, respectively, in Vietnamese patients
infected with multiple HBV genotypes containing mutations, these pa- compared with 2.0%, 8.8%, 1.0% and 2.9%, respectively, in Chinese
tients were counted in results for both genotypes and details of variants patients. However, amino acid mutations in position 126 were higher in
are presented in Supplementary Table 7. Chinese genotype B-infected patients, with a rate of 30.4%, compared
The majority of samples from Chinese patients were HBV genotype with 8.3% in Vietnamese genotype B-infected patients.
B or C (Beijing: 67 [24%] and 212 [75%], respectively; Guangzhou: 82 Genotype C-infected Korean subjects had a higher rate of amino acid
[65%] and 45 [36%], respectively). Of the 407 samples analyzed, 281 changes at positions 131 (8.3%), 133 (10.8%) and 136 (3.1%) com-
(69%) contained a total of 664 mutations (Supplementary Tables 3 and pared with genotype C-infected Chinese subjects. Compared with gen-
4). In the Beijing and Guangzhou samples, respectively, 231 and 112 otype C-infected Vietnamese subjects, Korean subjects had a higher
mutations were located in the immunodominant ‘a’ determinant region. mutation rate at positions 133 (10.8%), 134 (6.5%) and 136 (3.1%).
The distribution of mutations in the ‘a’ determinant region in the The classical G145R immune escape mutant [17,18] was observed
Chinese samples according to genotype is shown in Fig. 1A and B. The in 4.2%, 3.8% and 3.1% of Chinese, Korean and Vietnamese subjects,
most common variant identified was I126T, which was observed in 90 respectively, and mostly in genotype C-infected subjects. In general, the
samples (Beijing, n = 61; Guangzhou, n = 29); all of these samples mutation rate was higher in loop 1, with no significant difference
52
H.S. Kim et al. Journal of Clinical Virology 103 (2018) 48–56
Fig. 2. Distribution of MHR mutations located outside the ‘a' determinant region (amino acids 99–119 and 148–170) in A) Beijing, B) Guangzhou, C) Korea and D)
Vietnam.
53
H.S. Kim et al. Journal of Clinical Virology 103 (2018) 48–56
Fig. 2. (continued)
between the genotypes B- and C-infected subjects. The sensitivity of the Elecsys® HBsAg II Qualitative assay was 100%
A total of 25 novel mutations were identified in three countries (95% CI: 99.59–100.0) for all 898 samples with and without mutations.
(Table 3). These were most prevalent in genotype C-infected Korean All 898 samples were correctly identified by the Elecsys® HBsAg II
patients (n = 18) followed by Chinese (n = 12) patients. Qualitative assay, including samples with common diagnostic and
54
H.S. Kim et al. Journal of Clinical Virology 103 (2018) 48–56
55
H.S. Kim et al. Journal of Clinical Virology 103 (2018) 48–56
sequencing methods and providing technical support to study sites in structural diversity of hepatitis B surface antigen mutations in a global population,
Germany. Editorial support, in the form of development of a draft PLoS One 12 (5) (2017) e0172101.
[14] H. Norder, et al., Genetic diversity of hepatitis B virus strains derived worldwide:
manuscript in consultation with the authors and preparation of the genotypes, subgenotypes, and HBsAg subtypes, Intervirology 47 (6) (2004)
tables and figures, was provided by Chloe Fletcher, MSc, and Lucy 289–309.
Carrier, PhD, Gardiner-Caldwell Communications, Macclesfield, UK, [15] M. Neumann-Fraune, et al., Hepatitis B virus drug resistance tools: one sequence,
two predictions, Intervirology 57 (3–4) (2014) 232–236.
and was funded by Roche Diagnostics. [16] H. Norder, A.M. Courouce, P. Coursaget, J.M. Echevarria, S.D. Lee, I.K. Mushahwar,
et al., Genetic diversity of hepatitis B virus strains derived worldwide: genotypes,
Appendix A. Supplementary data subgenotypes, and HBsAg subtypes, Intervirology 47 (6) (2004) 289–309.
[17] C. Wu, et al., Amino acid substitutions at positions 122 and 145 of hepatitis B virus
surface antigen (HBsAg) determine the antigenicity and immunogenicity of HBsAg
Supplementary data associated with this article can be found, in the and influence in vivo HBsAg clearance, J. Virol. 86 (8) (2012) 4658–4669.
online version, at https://doi.org/10.1016/j.jcv.2018.04.005. [18] M.P. Cooreman, et al., Characterization of the reactivity pattern of murine mono-
clonal antibodies against wild-type hepatitis B surface antigen to G145R and other
naturally occurring a loop escape mutations, Hepatology 30 (5) (1999) 1287–1292.
References [19] X.Y. Wang, et al., The prevalence of mutations in the major hydrophilic region of
the surface antigen of hepatitis B virus varies with subgenotype, Epidemiol. Infect.
[1] A. Schweitzer, et al., Estimations of worldwide prevalence of chronic hepatitis B 143 (16) (2015) 3572–3582.
virus infection: a systematic review of data published between 1965 and 2013, [20] D.M. Yu, et al., N-glycosylation mutations within hepatitis B virus surface major
Lancet 386 (10003) (2015) 1546–1555. hydrophilic region contribute mostly to immune escape, J. Hepatol. 60 (3) (2014)
[2] G. Borgia, et al., Treating chronic hepatitis B: today and tomorrow, Curr. Med. 515–522.
Chem. 13 (23) (2006) 2839–2855. [21] Y. Yong-Lin, et al., Hepatitis B surface antigen variants in voluntary blood donors in
[3] M. Sunbul, Hepatitis B virus genotypes: global distribution and clinical importance, Nanjing China, Virol. J. 9 (2012) 82.
World J. Gastroenterol. 20 (18) (2014) 5427–5434. [22] Y. Shi, et al., Mutations in the major hydrophilic region (MHR) of hepatitis B virus
[4] Y. Cui, et al., Update on epidemiology of hepatitis B and C in China, J. genotype C in North China, J. Med. Virol. 84 (12) (2012) 1901–1906.
Gastroenterol. Hepatol. 28 (Suppl. 1) (2013) 7–10. [23] Y. Liu, et al., Molecular epidemiological study of hepatitis B virus in blood donors
[5] G. Zeng, et al., Geographic distribution, virologic and clinical characteristics of from five Chinese blood centers, Arch. Virol. 157 (9) (2012) 1699–1707.
hepatitis B virus genotypes in China, J. Viral Hepat. 12 (6) (2005) 609–617. [24] T. Bian, et al., Change in hepatitis B virus large surface antigen variant prevalence
[6] Y. Song, et al., Prevalence and trend of major transfusion-transmissible infections 13 years after implementation of a universal vaccination program in China, J. Virol.
among blood donors in Western China, 2005 through 2010, PLoS One 9 (4) (2014) 87 (22) (2013) 12196–12206.
e94528. [25] H. Kim, et al., Naturally occurring mutations in large surface genes related to occult
[7] I.W. Apata, et al., Progress toward prevention of transfusion-transmitted hepatitis B infection of hepatitis B virus genotype C, PLoS One 8 (1) (2013) e54486.
and hepatitis C infection–sub-Saharan Africa, 2000–2011, MMWR Morb. Mortal. [26] B.C. Song, et al., Prevalence of naturally occurring surface antigen variants of he-
Wkly. Rep. 63 (29) (2014) 613–619. patitis B virus in Korean patients infected chronically, J. Med. Virol. 76 (2) (2005)
[8] N. Coppola, et al., Clinical significance of hepatitis B surface antigen mutants, 194–202.
World J. Hepatol. 7 (27) (2015) 2729–2739. [27] T.T. Thuy le, et al., Distribution of genotype/subtype and mutational spectra of the
[9] P. Caligiuri, et al., Overview of hepatitis B virus mutations and their implications in surface gene of hepatitis B virus circulating in Hanoi, Vietnam, J. Med. Virol. 76 (2)
the management of infection, World J. Gastroenterol. 22 (1) (2016) 145–154. (2005) 161–169.
[10] P.F. Coleman, Surveillance for hepatitis B surface antigen mutants, J. Med. Virol. 78 [28] S. Louisirirotchanakul, et al., Comparison of the technical and clinical performance
(Suppl. 1) (2006) S56–58. of the Elecsys HBsAg II assay with the Architect, AxSym, and Advia Centaur HBsAg
[11] P.F. Coleman, Detecting hepatitis B surface antigen mutants, Emerg. Infect. Dis. 12 screening assays, J. Med. Virol. 82 (5) (2010) 755–762.
(2) (2006) 198–203. [29] A. Muhlbacher, et al., Multicenter study of a new fully automated HBsAg screening
[12] A. Servant-Delmas, et al., Variable capacity of 13 hepatitis B virus surface antigen assay with enhanced sensitivity for the detection of HBV mutants, Med. Microbiol.
assays for the detection of HBsAg mutants in blood samples, J. Clin. Virol. 53 (4) Immunol. 197 (1) (2008) 55–64.
(2012) 338–345. [30] S. Louisirirotchanakul, et al., Mutation of the a determinant of HBsAg with dis-
[13] M. Gencay, et al., Ultra-deep sequencing reveals high prevalence and broad cordant HBsAg diagnostic kits, Viral. Immunol. 17 (3) (2004) 440–444.
56