Background
Hepatitis B virus (HBV) infection is a global threat to human health, especially in developing countries with high prevalence. Individuals with detectable surface antigen at six months interval (HBsAg) are considered carriers of HBV chronic infection. In China, the prevalence of chronic HBV carriers declined from 9.5 to 7.2 % in the general population aged 1–59 years. In children <5 years between 1992 and 2006 HBsAg prevalence was 1 % following nationwide implementation of universal HBV vaccination at birth [
1]. However, although HBV vaccine compliance progressively increased between 1992 and 2000 from 30 to 76 % then stabilized up to 96 % in 2005, recent reports indicated an overall compliance of >90 % in general populations <20 years of age corresponding to a significant decline in HBsAg prevalence between 3.6 % below 20 and 10.2 % between 20 and 60 years of age [
2,
3]. Screening of HBsAg in blood donors massively decreased the risk of HBV transmission by transfusion, but did not identify pre-seroconversion window period infection (WPI) [
4]. In recent years, occult HBV infection (OBI) was defined as an absence of detectable HBsAg in circulation but presence of HBV DNA in blood or liver tissue [
5]. Implementation of nucleic acid testing (NAT) for HBV DNA detection yielded 1:1000-1:20,000 donor OBI carriers [
4]. However, a small number of blood donors with low-level of HBV DNA load could not be identified by current screening NAT due to insufficient sensitivity [
6,
7].
Over the past decades, universal vaccination has led to a 70–90 % decrease in chronic HBV carrier rates worldwide [
8], and has resulted in a substantial decline in disease burden, hepatitis B-related morbidity and mortality [
9‐
11]. Since infant vaccination started nationwide in 1992, HBV vaccinees are gradually becoming the majority of blood donors in China and are expected to significantly improve HBV safety in blood donations. However, in a previous study, we reported cases of HBV recovered, chronic and occult infections in vaccinated blood donors [
12]. The current trend of HBV infection in the blood donor population has not been systematically analyzed. In this study, we conducted a comprehensive survey and analysis of HBsAg screening results during the past 10 years (2005 to 2014) in donors stratified according to type, geographical origin, gender, age, and assumed vaccination status.
Discussion
Shenzhen is a modern city close to Hong Kong, located in the Eastern part of Guangdong province, south China. Over 10 millions of people live in Shenzhen, 88.41 % of them being 15–59 years old. According to age distributions of blood donors, young adults with ages 18–35 years are the majority of the blood donor population, who contribute nearly 80 % of blood donations (Fig.
2). Shenzhen is also a younger city and approximately 70 % of blood donors are migrant citizens (Table
1) somewhat reflective of Chinese population as a whole [
15].
In this study, half a million blood donors were screened for chronic, window period or occult HBV infections between 2005 and 2014 in Shenzhen, China (Tables
1 and
4, Figs.
1 and
3). An overall prevalence of HBsAg in blood donors was 2.9 % in 2005, declined to 1.8 % in 2010, then rebounded to 2.8 % in 2014 during past 10 years (Fig.
3). Observed HBV prevalence was significantly higher in first-time donors than in repeat-donors, in male than in female and in native than in donors of other origin, respectively (
P < 0.0001). As expected, HBsAg prevalence in first-time donors significantly increased with age (Table
2). Low and stable prevalence of HBsAg in repeat donors reflected either new infections occurring between blood donation or higher sensitivity of assays over time.
The most surprising result in this study was the higher HBsAg prevalence in assumed vaccinated first-time donors (3.89 %) aged 18–22 between 2010 and 2014 than in presumed non-vaccinated donors in the same age range (3.51 %) tested at the same age between 2005 and 2009 (Table
3,
P < 0.005). HBsAg prevalence in first-time donors born after 1992 tended to increase with age while in donors born earlier, no such difference in prevalence according to age was observed (Table
3). These apparently surprising results can be interpreted by examining the expected rate of vertical HBV transmission before 1992 and after 1992 when compliance was settling in and the changes in HBV infection in vaccinated Chinese over the age of 14. Between 1987 and 1992, the rate of HBV vaccination in China was estimated at 30 % [
3,
16]. Assuming a prevalence of HBsAg of 15 % in the Guangdong province [
16] and a 40 % rate of vertical transmission [
17,
18] and 70 % of the population not vaccinated, the predicted incidence of HBsAg in children was 4.3 %, expected to rise somewhat when these children are tested between age 18 and 22. The observed HBsAg prevalence in such adults having become blood donors was 3.5 %, lower than anticipated. A possible explanation might be that among the 30 % vaccinated population, HBsAg positive women were selected for offspring vaccination.
Applying the same calculations to children born in 1992–1996 (blood donors of 18–22) during the ramp up of vaccination compliance from 30 to 70 %, the expected HBsAg rate is predicted to decline from 4.25 to 1.8 % (mean 3.0 %). This estimate is considerably lower than the 3.9 % observed (Table
3), suggesting that infections acquired by other means than MTCT played a role. In two large studies conducted in China and Taiwan, respectively, the prevalence of HBsAg between age 10 and 20–24 progressively increased in presumably or effectively vaccinated general populations from less than 1 % to approximately 3 % [
3,
9]. In the Taiwanese study, evidence of contact with HBV after age 14 when the prevalence of anti-HBs has declined below 20 % is provided by increasing prevalence of HBsAg from 1 to 3 %, anti-HBc from 2 to 11 % and anti-HBs from 19 to 68 % by age 24 [
9]. This data strongly suggests that protection to HBV genotype B infection provided by genotype A1 vaccine declines with age and translates into seroconversion to HBsAg and/or anti-HBc or natural anti-HBs boost when in contact with HBV in a high prevalence of chronic infection area [
12,
19]. The HBsAg prevalence observed in our study is quite compatible with this observation. This different distribution might be related to the progressive decline of anti-HBs vaccine-related protection leaving these vaccinated donors at higher risk of infection as they grow older and as previously reported [
12]. Other factors may also interfere such as the type of vaccine utilized (plasma derived or recombinant) but here are probably insignificant. However, between age 18 and 22, repeat donors presumably vaccinated had 0.3 % incidence of HBV infection, nearly half the rate observed in older donors, presumably not vaccinated (Table
4). This incidence of HBsAg seroconversion did not significantly increase with age (Tables
3 and
4) [
12]. This observation is compatible with frequently abortive infection when vaccinees with low level of anti-HBs are exposed sexually to high levels of HBV [
19]. The contrast in donors born since 1992 between the high prevalence of HBsAg and relatively few recent infections suggests chronic infections being acquired early in life presumably out of non compliance to vaccination or failure of the immunization and relatively effective protection against new infection likely related to sexual activity. In the cohort of donors born before 1992, it is possible that a fairly high proportion of them had been vaccinated ahead of the generalization of the intervention explaining the relatively low prevalence of chronic HBV infection.
In recent years, occult hepatitis B virus infection (OBI) attracted considerable attention with regard to blood safety [
5,
20], and was found the major remaining residual risk of transfusion-transmitted viruses [
14,
15]. In Shenzhen blood donor population, the yield of 1:3,302 was detected for HBV DNA+/HBsAg-, included relatively frequent OBIs (1:3,835) and few cases of window period infection (1:67,934), although these frequencies might be modified if all yield samples had been categorized (Table
4). The molecular biological features of OBIs in blood donors have been well characterized in East Asia [
12‐
14,
21]. This transfusion risk could be minimized but not completely eliminated by NAT screening in blood donation due to insufficient sensitivity of current assays [
6,
7].
Since hepatitis B vaccination became mandatory for all newborns within 24 h of birth nationwide in 1992, this universal vaccination program decreased the HBV prevalence to 1 % in vaccinated children aged <5 years and led to a reduction of HBV prevalence to 7.2 % in the general population of China [
1]. HBV prevalence reported from Shenzhen blood donors including vaccinated first-time donors is clearly lower than that from the general Chinese population, but still remains much higher than expected in vaccinated children. The potential causes of HBV infection in blood donors, particularly in vaccinated population may originate from the following options [
22]. (1) Low level or undetectable anti-HBs. Previous study confirmed that there is among adequately vaccinated newborn a small portion (3-7 %) of non- or low-responders [
23]. In addition over 50 % vaccinated children no longer carry detectable anti-HBs when reaching 11–17 years [
9,
24]. In our previous study of 1494 vaccinated blood donors aged 18–21 years in Shenzhen, Southern China, we found approximately 29 % of donors with no detectable HBV markers and 40 % of them carrying anti-HBs levels <100 IU/L [
12]. Those vaccinees with low level of anti-HBs are susceptible to infection associated with breakthrough or leading to occult HBV infections as described previously [
12,
25,
26]. (2) HBV genotype A2 vaccine efficacy. Current hepatitis B recombinant S protein vaccines are of genotype A2. A study found that 6 of 9 vaccinated blood donors were identified as OBIs [
19]. Five of these donors had anti-HBs <100 IU/L and had been infected with non-A2 (genotype B, C, F or D) or mixed HBV strains, which suggested that protection offered by genotype A2 hepatitis B vaccine might not be fully effective for individuals exposed to non-A2 strains such as genotype B or C prevalent in China by close contact with sexual partners carrying high HBV load. (3) A previous study conducted in vaccinated blood donors revealed a prevalence of anti-HBc increasing with age consistent with an increasing cumulative HBV exposure [
12], suggesting that those with low level immune response were insufficiently protected when in contact with high HBV DNA load mainly through sexual activity [
19]. Many studies showed the decline and the high percentage of undetectable of anti-HBs in vaccinated people over 14 years of age, including blood donors [
9,
12,
27]. A vaccine boost in adolescents has been considered and its efficacy was demonstrated [
28‐
30]. However the justification of such strategy remains controversial and its implementation still under consideration.
Acknowledgements
The authors would like to thank Dr. Yongshui Fu (Guangzhou blood center, China) for his critical comments on the manuscript.