Molecular epidemiological study of hepatitis B virus in blood donors from five Chinese blood centers

Hepatitis B virus (HBV) is an aetiological agent of acute and chronic liver disease, including fatal fulminant hepatitis, cirrhosis and hepatocellular carcinoma, which is one of the most common human cancers and causes of death worldwide [1]. It has been estimated that more than 2 billion of the global population have been infected with HBV. Of these, approximately 360 million people are chronically infected, and an estimated 500,000 to 700,000 people die from complications of HBV infection each year worldwide [2]. HBV infection is highly endemic in China, with 5.84 % prevalence of HBV surface antigen (HBsAg) in the population of 1-59 years of age in 2007 [3].

HBV demonstrates remarkable genetic variability. Historically, using hepatitis B surface antigen (HBsAg) subtyping techniques based on specific antibodies, the genetic variability of the HBV envelope allowed classification into ten serologic subtypes, designated ayw1, ayw2, ayw3, ayw4, ayr, adw2, adw3, adw4, adrq+, and adrq– [4]. More recently, HBV has been classified into eight genotypes (A to H, in the order of discovery) with each genotype differing from the others by 8 % nucleotide divergence in the complete genome and 4 % in the sequence of the S gene [4]. Currently, HBV subtypes can be deduced from amino acid sequences at positions 122, 127, 134, 159, 160, 177 and 178 [5]. HBV genotypes have a distinct geographical distribution [6, 7]. For genotype A, subgenotype A2 is found mainly in northwestern Europe, North America and in Australians of European origin, while the more prevalent A1 is found in East and South Africa and along the coast of the Indian Ocean; genotypes B and C are prevalent in Southeast Asia, China and Japan; genotype D is spread worldwide, but it is predominant in the Mediterranean region and the Middle East; genotype E is almost entirely restricted to Africa; genotype F is found in Central and South America; genotype G has been reported in France and North America; and genotype H predominates in Central America. Recently, a ninth genotype (I) was tentatively proposed for HBV strains detected in Laos and a tenth genotype (J) was proposed for a HBV strain detected in a Japanese HCC patient [8, 9]. In China, genotypes C and B are predominant in patients with chronic liver disease, followed by genotypes D, E and A [10].

Besides genotype diversity, HBV S-gene mutations have been reported to affect nearly all amino acid positions of the major immunogenic region, the “a” determinant, which spans residues 124–147 of HBsAg. These mutations can lead to a false negative result when testing for HBsAg in blood donations and thus impact blood safety, especially in countries, including China, where screening of antibody to hepatitis B core antigen (anti-HBc) or HBV DNA is not routinely done [10].

Because the prevalence and sequelae from chronic HBV infections are very high in China, the Chinese government has taken effective measures to establish a universal infant immunization program to prevent HBV infections since 1992. Since then, hepatitis B vaccine coverage has reportedly increased from 30.0 % for newborns in 1992 to 93.4 % in 2005 [11]. However, nearly 20 years after the vaccination policy was put into effect in China, no comprehensive study of the HBV genotype prevalence or S-gene mutations in Chinese blood donors has been conducted. Little is known about mutations in the MHR region of HBV from blood donors. In this study, we analyzed the HBV genotype/subtype and mutations in the S region from voluntary blood donors who attempted to donate in five geographically diverse Chinese blood centers that participated in the Retrovirus Epidemiology Donor Study-II (REDS-II) International-China Program. The REDS-II International Program was developed to improve international blood safety by studying infectious and non-infectious risks of blood transfusion and mechanisms to improve the successful recruitment of low-risk volunteer blood donors. All blood donors participating in the REDS-II International China studies were provided with information on the objectives and nature of the studies, and they gave their written consent before enrolling in the studies. All REDS-II study protocols were reviewed and approved by the IRB at Johns Hopkins University and Chinese Academy of Medical Sciences prior to the implementation of the protocols.

A total of 228 (93.1 %) of the 245 randomly selected HBsAg-positive samples were successfully genotyped. Of these, 146 samples were sequenced directly, and the mean donor age was 28.3 ± 8.8 years. Eighty-two samples were genotyped by multiplex PCR, and the mean donor age was 32.0 ± 10.2 years. The donors from the second group were significantly older than those from the first group (p = 0.004). The sequences from the second group samples were found by direct sequencing to have heterozygous loci, which accounted for 36.0 % of the successfully genotyped samples. Overall, 24.4 % of the samples with heterozygous loci were found to contain mixtures of genotypes (n = 20), while the rest represented mixtures of variants of a single genotype (n = 62). Of the donors with genotype mixtures, genotype B and C mixtures represented 75.0 % of the total (n = 15), followed by genotype A and B mixtures (n = 3) and genotype B and D mixtures (n = 2).

Of the 228 blood donors genotyped successfully, we found nine donors infected with genotype A (3.95 %, 95 % CI: 1.4 %-6.4 %), 134 with HBV genotype B (58.8 %, 95 % CI: 52.4 %-65.2 %), 50 with genotype C (21.9 %, 95 % CI: 16.6 %-27.3 %), 15 with genotype D (6.6 %, 95 % CI: 3.4 %-9.8 %), 20 (8.8 %, 95 % CI: 5.1 %-12.4 %) with two genotypes (genotype mixture), and no subjects with genotypes E to H.

Molecular epidemiological studies provide valuable information to help understand the prevalence and characteristics of HBV genotypes and mutations. However, most of the molecular epidemiological studies of HBV in China have focused on patients, not on blood donors, who are generally asymptomatic and healthy individuals. Knowledge of the HBV genotypes and mutations in HBsAg-positive donors is important for developing effective donor screening for HBV. Studying healthy donors also provides a window into the molecular epidemiological evolution of HBV in the general population, and therefore, screening healthy blood donors should be a part of the comprehensive surveillance program. We filled this gap with a multi-center study of the geographical and demographic distribution of genotypes, subtypes and mutations in the S region of the HBV genome in donors from five blood centers, where the blood donors were geographically, socially, economically and ethnically diverse.

Although sequencing of the whole genome or S region is the gold standard for determining the HBV genotype, it is difficult to detect genotype mixtures by this method. In our study, a multiplex PCR assay was used to detect genetic mixtures in samples with heterogeneous sequences. Using this method, we detected 20 samples with genotype mixtures.

Genotypes C and B were the major HBV genotypes endemic in Mainland China. As shown in our study, HBV genotype B is more prevalent than genotype C in blood donors, while other studies have reported that genotype C has a higher prevalence in patients than genotype B. This difference may be due to the greater virulence of HBV genotype C than B [19‐21]. Genotype C is associated with the development of cirrhosis and hepatocellular carcinoma as well as a lower response rate to interferon therapy. It also has a lower rate of seroconversion from HBeAg to anti-HBe and a higher HBV DNA level compared to genotype B [22]. Individuals infected with HBV genotype B may be more likely to remain asymptomatic and become part of the donor pool, posing a substantial threat to the safety of the blood supply.

Genotypes of HBV are generally subtype specific, although some subtypes are heterogeneous. In general, subtype adw is usually found in genotypes A and B, while adr occurs in genotype C [22]. In this study, we found that all of the genotype A strains and most of the genotype B HBV strains belonged to subtype adw, and most of the genotype C strains belonged to subtype adr, which is similar to previously published reports [22].

HBsAg is the main target for virus neutralization, either by natural or vaccine-induced anti-HBs. The basic working model is that of a protein with four transmembrane helices in which several residues at the N- and C-termini and a central major hydrophilic region (MHR) from residues 103-173 are exposed at the surface of viral particles. The MHR is composed of five regions: HBs1, HBs2, HBs3, HBs4 and HBs5. Mutations in this region occurring naturally or under immunization pressure could affect HBsAg test results or an individual’s reaction to vaccination [17]. In our study, about 17 % of samples were found to have mutations, and most of the mutations were located at loci that might be related to the failure of immunization and HBsAg detection. The mutation rate was lower than those reported elsewhere, including Japan [23] (24 %), Korea [24] (50 %), France [25] (28 %) and Spain (40 %) [26]. However, the lower mutation rate might also be due to differences in the length of the fragment that was analyzed. It is also possible that our mutation rate is an underestimation, because this study only included HBsAg-positive donors, while there may be HBsAg-false-negative donors who were missed by the screening due to the mutations in the “a” determinant region. In addition, the sequences were cloned, and only one clone was selected for sequencing for the samples that showed heterogeneity by direct sequencing. This approach may miss some mutations and decrease the rate of detection of mutations. The most important and best-documented mutation in MHR was G145R, which is also the most critical substitution to prevent HBsAg detection [27]. We detected this mutation in only one sample from the Urumqi Blood Center. Since G145R is associated with false negative HBsAg screening, it may be reasonable to assume that there may be infected donors with this mutation who were not included in our study because they had a false negative HBsAg screening result.

There are several potential limitations to this study. Due to low viral loads or mutations in the primer-recognition sites, a total of 17 (6.9 %) of the 245 selected HBsAg-positive samples were not genotyped successfully. Although it is possible that some of these 17 samples may be false positive for HBsAg, new primers and a more sensitive genotyping method may be useful in further studies. Because complete genome sequencing is not applicable for routine investigations, we have limited our analysis to a part of the S gene that has been described as reliable for genotyping [7]. Sequencing of a larger region will be necessary to identify subgenotypes accurately. In our study, the genotype of an isolate from Urumqi Blood Center could not be determined by sequencing because of the low bootstrap value with any genotype from A to G, but it was genotyped successfully as genotype B using multiplex PCR. However, complete genome sequencing is needed to clarify the genotype of this subject.

This study is the first multi-center study of HBV genotypes, subtypes and mutations in voluntary Chinese blood donors from representative regions throughout China. The study results were analyzed in the context of donors’ demographic and geographic characteristics. We have identified viral envelope mutants in this study that may compromise HBsAg detection in HBV-infected donors. However, HBV with envelope mutations can be detected by the nucleic acid test (NAT), which currently is not included as a test for routine blood donor screening. Our data suggest that the addition of the NAT to the routine donor screening process would minimize false negative HBV results due to S region mutations and would further increase blood safety in China. Meanwhile, large-scale molecular epidemiological studies of HBV in Chinese donors as well as in the general population will help to understand the diversity and distribution of HBV mutants and are essential to reduce the risks of transfusion-transmitted HBV.