ORIGINAL ARTICLE High prevalence of anti-hepatitis B core antigen in hepatitis B virus–vaccinated Chinese blood donors suggests insufficient protection but little threat to the blood supply Xin Zheng,1,2* Xianlin Ye,1* Peng Du,2 Jinfeng Zeng,1 Weigang Zhu,1 Baocheng Yang,1 Chengyao Li,2 and Jean-Pierre Allain2,3
BACKGROUND: In East Asia, individuals systematically vaccinated at birth to hepatitis B virus (HBV) are an increasing part of the blood donor population. Their environment presents a high risk of contact with HBV. HBV vaccine efficacy and potential safety risk carried by vaccinated donors were examined. STUDY DESIGN AND METHODS: A total of 2028 vaccinated blood donors were recruited in 2012 and 2013 and tested for serologic (hepatitis B surface antigen [HBsAg], antibody to hepatitis B surface antigen [antiHBs], and antibody to hepatitis B core antigen [antiHBc]) and molecular (HBV DNA) markers of HBV. HBsAg, anti-HBs, and viral load were quantified. RESULTS: Donors 18 to 21 years systematically vaccinated at birth and 22 to 25 years and older donors had both 30.0% negative serology and 1.8% anti-HBc only but the latter group carried significantly higher prevalence of anti-HBc (p < 0.0001). Anti-HBc, mostly associated with anti-HBs, increased from 10.7% at age 18 to 31.5% at age 25. The level of anti-HBs was significantly higher in anti-HBc–positive donors than in anti-HBs– only donors (p < 0.0001). Samples from 24 donors contained low viral load (25 ± 22 IU/mL), half of them undetected by standard nucleic acid testing (NAT), and were classified as four recent infections, 17 occult HBV infections (OBI), and three primary OBIs. Eighteen of 24 carried anti-HBs; 14 of 15 strains were wild-type Genotype B and one was Genotype C. CONCLUSIONS: In an environment of frequent high Genotype B or C viremia, blood donors vaccinated at birth are frequently but mildly infected: asymptomatic and normal alanine aminotransferase level, identified by anti-HBc seroconversion and boosting of anti-HBs. Low viral load and frequent anti-HBs limit transfusion risk.
I
n China, chronic hepatitis B was ranked first among the 27 infectious diseases reported by the Chinese government for more than 10 years. Approximately 60% of the Chinese population has laboratory evidence of contact with hepatitis B virus (HBV) and 7.2% are chronic carriers of hepatitis B surface antigen (HBsAg).1 HBV is the main threat to the safety of the blood supply. HBsAg screening was first implemented in 1978 but became mandatory in 1993 and massively decreased the risk of HBV transfusion transmission. HBsAg screening assays considerably improved over time but remain unable to detect the pre-seroconversion window period (WP) or samples with occult HBV infection (OBI).2 The availability of HBV nucleic acid testing (NAT) in blood donations enabled the identification of HBV DNA carriers in asymptomatic blood donors negative for HBsAg with a prevalence ranging between 1 in 1000 and 1 in 20,000 in
ABBREVIATIONS: BCP/PC = basic core promoter/precore; OBI = occult hepatitis B virus infection; qPCR = quantitative polymerase chain reaction; WP = window period. From the 1Shenzhen Blood Center, Shenzhen, China; the 2 Department of Transfusion Medicine, Southern Medical University, Guangzhou, China; and the 3Department of Haematology, University of Cambridge, Cambridge, United Kingdom. Address reprint requests to: J.-P. Allain, Cambridge Blood Centre, Long Road, Cambridge CB2 0PT, UK; e-mail: [email protected]; or C.Y. Li, Department of Transfusion Medicine, Southern Medical University, Guangzhou 510515, China; e-mail: [email protected]. *XZ and XY contributed equally to this work. This work was supported by the National Natural Science Foundation of China (81071348 and 81371801). Received for publication July 23, 2014; revision received September 4, 2014, and accepted September 4, 2014. doi: 10.1111/trf.12902 © 2014 AABB TRANSFUSION **;**:**-**. Volume **, ** **
TRANSFUSION
1
ZHENG ET AL.
donated blood untested for antibody to hepatitis B core antigen (anti-HBc).3 However, a small number of chronic low-level carriers could not be identified even by individual-donation genomic amplification due to insufficient sensitivity.4 The total number of blood donations collected in China in 2013 was 12.78 million, of which 32% were tested for HBV DNA (Chinese CDC). The residual risk of HBV transfusion transmission from OBI can be predicted on the basis of 1-in-1000 OBI, 50% of recipient susceptible (anti-HBc negative) and 20% infectivity to approximately 1 in 8000. Anti-HBc screening can eliminate nearly all OBIs but not WP cases.5 Rare cases of vaccinated anti-HBs–positive but anti-HBc–negative blood donors carrying HBV DNA have been reported that corresponded to acute infection occurring in low-anti-HBs, sexually exposed individuals.6 Nonetheless, in areas where anti-HBc prevalence is greater than 2% to 5%, the exclusion of anti-HBc–positive donors is impractical as it would too severely impact on the blood supply.7 In China, HBV NAT added to HBsAg screening remains the main option for improved HBV blood safety. To control HBV infection, the Chinese government implemented mandatory infant vaccination with three doses of hepatitis B vaccine at 0, 1, and 6 months as the highest priority in 1992 which resulted in a significant reduction of carrier rate in children from 10% to less than 1% over two decades.1 Before 1990, 95% of HBV vaccines were locally produced plasma derived and after that recombinant protein Genotype A2.8 Between 1992 and 2001, compliance was overall 64% reaching 75% in urban areas.9 A vaccinated population has enrolled as blood donors since 2010, and numbers increasing every year should impact the HBV blood safety positively. However, the previously mentioned report suggested that HBV vaccine made of Genotype A2 recombinant protein might not be fully efficient when used in areas where other genotypes are prevalent such as Genotypes B and C in China.6 To test this hypothesis, a large cohort of vaccinated Chinese blood donors recruited by Shenzhen Blood Center was identified and tested for HBV serologic and molecular markers. This study was intended to examine the effectiveness of the vaccination program and the potential threat of vaccination failures to the blood supply.
MATERIALS AND METHODS Population The target population was HBV-vaccinated blood donors born either between January 1, 1992, and April 1996 (age 18-21 years) with a history of neonatal HBV vaccination or April 1988 to December 1991 (age 22-25 years). Samples were collected in two separate periods: January to December 2012 and April to June 2013. The sample collection was limited to donors 18 to 21 years in 2012 and extended to 18 to 25 years in 2013. Hepatitis B vaccination history was 2
TRANSFUSION Volume **, ** **
collected from consenting donors with a questionnaire to all donors and a review of vaccination records for 18- to 21-year-old donors. In all cases, three doses of vaccine were injected within 24 hours of birth and at 1 and 6 months of age. HBV vaccine efficacy was defined as absence of HBV markers (HBsAg, anti-HBc, and HBV DNA). All blood donors were voluntary and nonremunerated and passed the predonation questionnaire as well as the rapid predonation testing for HBsAg and alanine aminotransferase (ALT) at the collection sites (colloidal gold strip method, Abon Diagnostics, Hangzhou, China; and Roche Refletron, Roche Diagnostics, Mannheim, Germany, respectively). The sensitivity of these HBsAg assays was 2.5 ng/mL and specificity was greater than 99%. Qualified donors (negative for the predonation tests) proceeded to donate blood. Whole blood donations were then tested by two enzyme immunoassays (EIAs) at the Shenzhen Blood Center, for HBsAg (Diasorin, Dartford, UK; and Xinchang Diagnostics, Xiamen, China), for antibody to hepatitis C virus (anti-HCV; Ortho Clinical Diagnostics, High Wycombe, UK; and Lizhu Diagnostics, Zhuhai, China), for antibody to human immunodeficiency virus Types 1 and 2 (anti-HIV-1/2; Bio-Rad, Marnes-la-Coquette, France; and Wantai Diagnostics, Beijing, China), for syphilis (Diasorin and Lizhu Diagnostics), and for ALT with a kinetic method (AusBio Biotech, Fuzhou, China). Only donors nonreactive for serologic tests and with normal ALT level were screened by NAT for HBV, HCV, and HIV-1 genomes with a multiplex assay (Procleix Ultrio, individual-donation NAT, Novartis Diagnostics, Emeryville, CA). Donors negative for all tests were enrolled randomly in this study. In both 2012 and 2013, only a subpopulation of donors in the 18- to 21-year (12.1% of total) or 22- to 25-year (22.5% of total) age group was enrolled in a proportion of 93% for 18- to 21-year-old donors and 77.6% for 22- to 25-year-old donors, respectively, after checking for evidence of vaccination. Anti-HBc was tested with a commercially available EIA (Kehua Biotech, Shanghai, China). Anti-HBc–reactive samples were further tested with an alternative commercial EIA (Wantai Biotech). Samples reactive with both assays were considered anti-HBc positive. Anti-HBs was quantified with a commercial EIA (Wantai Biotech) according to the manufacturer’s instructions. Values below 10 IU/L were considered negative. The assay was quantitative within the range 10 to 327 IU/L and higher levels were given as more than 327 IU/L. HBV DNA was extracted from 2 mL of plasma from anti-HBc–positive samples with a viral nucleic acid largevolume kit (HighPure, Roche Diagnostics) and was tested with a combination of quantitative real-time polymerase chain reaction (PCR; sensitivity, 5 IU/mL) and nested PCR procedures amplifying the basic core promoter/precore (BCP/PC), the S region, the pre-S/S region, and full genome as previously described.10 Samples providing an
HBV INFECTION IN VACCINATED BLOOD DONORS
TABLE 1. Distribution of HBV serologic markers according to age and sex in 2028 vaccinated blood donors screened for HBsAg and ALT* Age (years) 2012, Markers Sex, male/female Anti-HBs+/anti-HBc+ Anti-HBs+/anti-HBc– Anti-HBs–/Anti-HBc+ Anti-HBs–/Anti-HBc– Total (%)
2013
18-21
18-21
22-25
Total
560/589 195 (17.0) 630 (54.8) 17 (1.5) 307 (26.7) 1149 (56.7)
186/159 52 (15.1) 156 (45.2) 8 (2.3) 129 (37.4) 345 (17.0)
309/225 150 (28.1) 200 (37.5) 11 (2.1) 173 (32.4) 534 (26.3)
1055/973 397 (19.6) 986 (48.6) 36 (1.8) 609 (30.0) 2028
They were distributed in 48% females and 52% males. In total, 1494 donors were 18 to 21 years of age (1149 in 2012, 345 in 2013) and 534 were 22 to 25 years in 2013. The first age group had been included in the systematic vaccination program at birth initiated in January 1992; donors of the second age group were born in the years preceding such implementation but presented evidence of having been vaccinated.
* Data are reported as number (%).
amplicon in two regions were considered confirmed positive. Samples amplified in only one region but for which a sequence was obtained were considered positive when no element suspect of contamination was present. To avoid contamination, extraction and amplification were performed in separate rooms; each run included a negative control since the extraction step and sequences (except for BCP/PC) were scrutinized for identity and subsequent exclusion.
HBV DNA sequencing and genotype determination To confirm the HBV DNA specificity, the amplified products, obtained from the BCP/PC (295 nucleotides) and S (496 nucleotides) regions, pre-S/S (1433 nucleotides) or full genome (3162 nucleotides) were purified with a DNA gel extraction kit (AxyPrep, Axygen, Union City, CA) to remove excess primers. The purified amplicons were sent for sequencing to Shanghai Invitrogen (Guangzhou, China). Nucleotide sequences and reference genotyped sequences from Genebank were aligned and manually edited using a biologic sequence alignment editor program (BioEdit, http://www.mbio.ncsu.edu/bioedit/ bioedit.html). Genotype determination was performed by phylogenetic analysis using a computer program (MEGA5.1, Informer Technologies, Inc., Redwood City, CA). The neighbor-joining method with 1000 bootstrap replications was used and over 75% bootstrap value was considered significant.
Statistical analyses Ninety-five percent confidence interval for the observed yield rate was derived using binomial exact proportion method. Categorical variables were compared using Fisher’s exact test and, for continuous variables, the nonparametric Mann-Whitney test. A p value of less than 0.05 was used as the cutoff level for significance.
RESULTS Blood donor population The 2028 vaccinated blood donors were collected by the Shenzhen Blood Center: 1149 in 2012 and 879 in 2013.
Distribution of HBV serologic markers in vaccinated blood donors The distribution of HBV serologic markers (anti-HBs and anti-HBc) in 2028 blood donor samples is shown overall and according to donor age in Table 1 and Fig. 1. Thirty percent of vaccinated donors did not carry detectable anti-HBs 18 to 25 years after vaccination. The percentage of donors with anti-HBs only tended to decrease with age (p = 0.0001 between age 18 to 25 groups) while donor samples carrying both anti-HBs and anti-HBc increased (Fig. 1). Confirmed anti-HBc was present in 21.4% of vaccinated donors, the vast majority of samples also carrying anti-HBs (91.7%). Overall, 68.2% of the population carried quantifiable anti-HBs. The distribution of anti-HBs levels in donors age 18 to 21 years is shown in Fig. 2. The data are stratified according to the presence or absence of anti-HBc together with anti-HBs. The anti-HBc–positive samples carried significantly higher levels of anti-HBs than those anti-HBs only (p < 0.001, Fig. 2). The median level of anti-HBs in antiHBs–only samples was 136 IU/L with 17.4% of values of higher than 327 IU/L; the median level of samples that were anti-HBc and anti-HBs positive was 241 IU/L with 23% of samples more than 327 IU/L (p < 0.001). At age 18, 10.7% of vaccinated donors carried anti-HBc and 31.5% by age 25 (Fig. 1; p = 0.008 between frequency of anti-HBc in 18- and 25-year groups). This trend in distribution was independent of sex (data not shown). In donors age 22 to 25 years, the anti-HBs levels between samples that were anti-HBs and anti-HBc positive was also significantly higher than in those that were anti-HBs only (p < 0.0001; data not shown). The potential difference in efficacy of HBV vaccine between donors born before and after the universal neonatal vaccination was examined. No significant difference between these two groups was found whether in frequency of detectable anti-HBs or anti-HBs levels or sex distribution. In donor group below and above age 21, only the percentage that were anti-HBc positive significantly differed (18.2 and 30.1%, respectively; p < 0.001). There was a small population of anti-HBc– confirmed-reactive but anti-HBs–negative donors Volume **, ** **
TRANSFUSION
3
ZHENG ET AL.
confirmed positive (Table 2, first 12 samples). In addition, 433 samples carrying anti-HBc were tested by quantitative PCR (qPCR) and confirmed by nested PCR and 12 samples were positive (Table 2, last 12 samples). In the Ultrio-detected group, 11 of 12 were quantified (27 ± 66 IU/mL); in the qPCR-detected group, seven of 12 were quantified (16 ± 8 IU/mL). All samples had normal levels of ALT. Three of these samples did not contain serologic markers of HBV. One of them (V11) was shown on follow-up to be a WP infection and two (V12 and V13) remained DNA positive and serology negative 4 months after index sample and were classified as primary OBI (Table 3). Of the remaining Fig. 1. Distribution of HBV serologic markers according to age and sex in 2028 vacci21 samples 15 carried both anti-HBs nated blood donors. Open bars = no anti-HBs, no anti-HBc; closed bars = anti-HBc and anti-HBc or anti-HBs only (three only; light gray bars = anti-HBs only; dark gray bars = both anti-HBs and anti-HBc cases) or anti-HBc only (three cases; positivity. The percentage of anti-HBc in anti-HBs–positive samples increases with Table 2). Eighteen HBV DNA carriers age, irrespective of donor sex. For each age year, the number of donors is 18 years, were younger than 22 years (75%) and 150; 19 years, 351; 20 years, 427; 21 years, 570; 22 years, 109; 23 years, 162; 24 years, 15 were females (62.5%). 117; and 25 years, 146. qPCR results were confirmed by nested PCR and amplicon sequencing in each of the 24 cases. Amplicons of BCP/PC and S (Table 1, Fig. 1). Three of these samples (V3, V126, and regions were found to be positive in 21 and 15 samples, F161) contained detectable HBV DNA and sequences were respectively. The latter amplicon was sufficiently informaobtained (Table 2). The frequency of this situation was tive to identify the strain genotype (14 Genotype B and similar in 18- to 21- and 22- to 25-year-old donors (1.7 and one Genotype C; Table 2). The only Genotype C case was 2.1%, respectively). in the WP with a viral load of 243 IU/mL. Only amplicon of To obtain a complete picture of HBV infections in the BCP/PC region was obtained in nine samples and the blood donors during the period of interest, HBsAgsequences were not informative for genotyping. The alignpositive samples identified either with the predonation ment of sequences from either the six full genome or the rapid test screening or secondarily by EIA and NAT (see two pre-S/S sequences showed less than 99% identity Materials and Methods) were retrieved and tabulated. In between strains, excluding contamination. In the S region, 7512 age 18 to 21 donors seen during the 2012 period of all 15 sequences were wild type; none of the mutations sample collection 267 HBV-infected samples were identifound in OBI cases of nonvaccinated blood donors were fied (221 HBsAg reactive predonation, 34 HBsAg positive present.7 postdonation, and 12 NAT positive/HBsAg negative). All EIA HBsAg–positive samples were anti-HBc positive. The median level of EIA HBsAg was 28 ± 76 IU/mL and the Follow-up of HBsAg-negative and HBV median level of viral load was 1204 ± 5615 IU/mL. When DNA–positive vaccinated blood donors extrapolating the 18.5% anti-HBc–positive samples detected in the 1149 subgroup of age 18 to 21 donors colFor 14 HBV DNA–positive index cases, one or more lected in 2012 (Table 1), the whole donor population in follow-up samples obtained 47 to 560 days later were that age group included 1340 samples anti-HBc positive tested (Table 3). In 10 cases, index sample serologic or (7512 − 267 = 7245 × 18.5% = 1340). The total number of molecular markers remained unchanged confirming the donors exposed to or infected by HBV was 1607 initial diagnosis of OBI or primary OBI. V11, as previously (1340 + 267 = 1607)/7512 or 21.4%. indicated, presented as a WP infection, seroconverting to anti-HBc and anti-HBs at 4 and 6 months postindex sample. Donor V78 who was low positive for anti-HBc and HBV DNA in vaccinated blood donors anti-HBs in the index sample was negative for both markers 18 months later although HBV DNA remained HBV DNA was tested by NAT (Ultrio, Novartis Diagnostics) detectable. Follow-up samples from donors V5 and V48 in HBsAg-negative samples, and 12 samples were 4
TRANSFUSION Volume **, ** **
HBV INFECTION IN VACCINATED BLOOD DONORS
mandatory vaccination at birth in China (Table 1). For donors over 21 years, requirements for evidence of vaccination are less stringent than in donors 18 to 21 years (questionnaire instead of checking the vaccination booklets). The absence of a significant difference in the distribution of anti-HBs prevalence and level between the two groups ultimately justified a global analysis of the cohort. During the period of the 2012 study, a total population of 7512 18- to 21year-old donors was tested for HBV markers. HBsAg- and HBV DNA only– positive donors accounted for 3.4 and 0.15%, respectively. Adding the 18.45% prevalence of anti-HBc–positive obtained with a subset of 1149 donors (Table 1), the predicted prevalence of HBV infection markers reaches 22%. This percentage is similar to 20.9% Fig. 2. Distribution of anti-HBs levels (IU/L) in vaccinated donors (18-21 years old as found in The Gambia in a 20- to 24-yearin Table 1) carrying anti-HBs only or both anti-HBs and anti-HBc. Light gray bars = old population of vaccinees11 but higher samples carrying only anti-HBs; dark gray bars = samples carrying both anti-HBs than 14.7% found in a 19- to 21-year and anti-HBc; anti-HBs levels are stratified in five categories. The difference in antipopulation of Northern China12 and 6% HBs levels between samples with and without anti-HBc was significant (p < 0.001). reported in Taiwan.13 However, it might be underestimated since only anti-HBc– positive samples were retested for HBV DNA and the charshowed an increased level of anti-HBs, a low level of antiacterization of the Ultrio DNA–positive samples were not HBc 3.5 months after the index sample and HBV DNA all anti-HBc positive. becoming undetectable 3.5 months after the index Examining the 2028 donors 18 to 25 years tested for sample, suggesting a recent contact with HBV. anti-HBs and anti-HBc, the relatively large proportion of vaccinated donors without detectable anti-HBs (31.8%; Identifying the potential origin of HBV infection Table 1) is consistent with the 39.3% found in Northern China,11 37.3% Guangxi province by age 15,14 41.5% in 18To attempt to identify the origin of the HBV infection in vaccinated donors, parents and sexual partners were to 21-year-old Taiwanese students,13 and 25% found in called upon, informed consent was obtained, and tests The Gambia.12 In Thailand, between age 15 and 20, the were performed. Parents were available in four cases and a percentage of seronegative vaccinees ranged between 33.3 sexual partner in one case. Both parents of four donors and 36.4%.15 Figure 1 shows no significant changes of (V1, V9, V11, and V78) were tested. Only V1 mother and V9 antibody-negative donors between age 18 and 25 but a father were viremic but available sequences indicated no relatively large range (20.9 and 38.9%) year by year. more than 98% homology excluding them as source of The efficacy of HBV vaccine can be examined from infection. Among the seven other parents, all but one had different angles: the occurrence of acute hepatitis and the evidence of contact with HBV but without detectable viral occurrence of chronic, often asymptomatic, infection indiload. cated by circulating HBsAg or contact with HBV not prevented by vaccine-induced anti-HBs indicated by antiHBc. The occurrence of acute hepatitis cannot be DISCUSSION estimated since no detailed questionnaire was administered to candidate donors and ALT level was prescreened The cohort of 2028 vaccinated blood donors was considand excluded donors with abnormal levels without ered suitable for donation after negative screening with attempting to find its origin. HBsAg was detected in 3.4% of two rapid HBsAg assays, normal ALT level, and absence of 18- to 21-year-old blood donors but the presence of liver clinical acute hepatitis. These donors were further tested disease was not investigated. This percentage appears with HBsAg EIA and NAT with triplex Ultrio was higher than previously reported in China such as in Long assembled. This population was 48% female and 73% of An (2.4%) in subjects less than 20 years old, in Jiangsu donors were born after 1992, the official date of Volume **, ** **
TRANSFUSION
5
ZHENG ET AL.
TABLE 2. Characterization of HBV DNA–positive, HBsAg-negative samples from vaccinated donor ID V129 V8 V5 V7 N01 V10 V1 V4 V3 V12 V13 V11 V9 F09 F63 V48 V85 F82 F48 V76 F71 V78 V126 F161
Sex Female Male Female Female Female Female Male Male Female Female Female Female Male Male Male Male Female Male Female Female Female Female Female Male
Age (years) 21 19 20 20 23 21 20 21 21 21 18 21 20 25 23 21 21 25 24 19 25 20 21 22
Anti-HBs (IU/L) 789 400 23 23 18 86 14 10 − − − − >1000 >300 >300 890 825 315 310 55 39 25 − −
Anti-HBc + + + + + − − − + − − − + + + + + + + + + + + +
NAT Ultrio Ultrio Ultrio Ultrio Ultrio Ultrio Ultrio Ultrio Ultrio Ultrio Ultrio Ultrio qPCR qPCR qPCR qPCR qPCR qPCR qPCR qPCR qPCR qPCR qPCR qPCR
VL (IU/mL) 45 83 69 25 59 21 9 26 17 27
BACKGROUND: In East Asia, individuals systematically vaccinated at birth to hepatitis B virus (HBV) are an increasing part of the blood donor population. Their environment presents a high risk of contact with HBV. HBV vaccine efficacy and potential safety risk carried by vaccinated donors were examined. STUDY DESIGN AND METHODS: A total of 2028 vaccinated blood donors were recruited in 2012 and 2013 and tested for serologic (hepatitis B surface antigen [HBsAg], antibody to hepatitis B surface antigen [antiHBs], and antibody to hepatitis B core antigen [antiHBc]) and molecular (HBV DNA) markers of HBV. HBsAg, anti-HBs, and viral load were quantified. RESULTS: Donors 18 to 21 years systematically vaccinated at birth and 22 to 25 years and older donors had both 30.0% negative serology and 1.8% anti-HBc only but the latter group carried significantly higher prevalence of anti-HBc (p < 0.0001). Anti-HBc, mostly associated with anti-HBs, increased from 10.7% at age 18 to 31.5% at age 25. The level of anti-HBs was significantly higher in anti-HBc–positive donors than in anti-HBs– only donors (p < 0.0001). Samples from 24 donors contained low viral load (25 ± 22 IU/mL), half of them undetected by standard nucleic acid testing (NAT), and were classified as four recent infections, 17 occult HBV infections (OBI), and three primary OBIs. Eighteen of 24 carried anti-HBs; 14 of 15 strains were wild-type Genotype B and one was Genotype C. CONCLUSIONS: In an environment of frequent high Genotype B or C viremia, blood donors vaccinated at birth are frequently but mildly infected: asymptomatic and normal alanine aminotransferase level, identified by anti-HBc seroconversion and boosting of anti-HBs. Low viral load and frequent anti-HBs limit transfusion risk.
I
n China, chronic hepatitis B was ranked first among the 27 infectious diseases reported by the Chinese government for more than 10 years. Approximately 60% of the Chinese population has laboratory evidence of contact with hepatitis B virus (HBV) and 7.2% are chronic carriers of hepatitis B surface antigen (HBsAg).1 HBV is the main threat to the safety of the blood supply. HBsAg screening was first implemented in 1978 but became mandatory in 1993 and massively decreased the risk of HBV transfusion transmission. HBsAg screening assays considerably improved over time but remain unable to detect the pre-seroconversion window period (WP) or samples with occult HBV infection (OBI).2 The availability of HBV nucleic acid testing (NAT) in blood donations enabled the identification of HBV DNA carriers in asymptomatic blood donors negative for HBsAg with a prevalence ranging between 1 in 1000 and 1 in 20,000 in
ABBREVIATIONS: BCP/PC = basic core promoter/precore; OBI = occult hepatitis B virus infection; qPCR = quantitative polymerase chain reaction; WP = window period. From the 1Shenzhen Blood Center, Shenzhen, China; the 2 Department of Transfusion Medicine, Southern Medical University, Guangzhou, China; and the 3Department of Haematology, University of Cambridge, Cambridge, United Kingdom. Address reprint requests to: J.-P. Allain, Cambridge Blood Centre, Long Road, Cambridge CB2 0PT, UK; e-mail: [email protected]; or C.Y. Li, Department of Transfusion Medicine, Southern Medical University, Guangzhou 510515, China; e-mail: [email protected]. *XZ and XY contributed equally to this work. This work was supported by the National Natural Science Foundation of China (81071348 and 81371801). Received for publication July 23, 2014; revision received September 4, 2014, and accepted September 4, 2014. doi: 10.1111/trf.12902 © 2014 AABB TRANSFUSION **;**:**-**. Volume **, ** **
TRANSFUSION
1
ZHENG ET AL.
donated blood untested for antibody to hepatitis B core antigen (anti-HBc).3 However, a small number of chronic low-level carriers could not be identified even by individual-donation genomic amplification due to insufficient sensitivity.4 The total number of blood donations collected in China in 2013 was 12.78 million, of which 32% were tested for HBV DNA (Chinese CDC). The residual risk of HBV transfusion transmission from OBI can be predicted on the basis of 1-in-1000 OBI, 50% of recipient susceptible (anti-HBc negative) and 20% infectivity to approximately 1 in 8000. Anti-HBc screening can eliminate nearly all OBIs but not WP cases.5 Rare cases of vaccinated anti-HBs–positive but anti-HBc–negative blood donors carrying HBV DNA have been reported that corresponded to acute infection occurring in low-anti-HBs, sexually exposed individuals.6 Nonetheless, in areas where anti-HBc prevalence is greater than 2% to 5%, the exclusion of anti-HBc–positive donors is impractical as it would too severely impact on the blood supply.7 In China, HBV NAT added to HBsAg screening remains the main option for improved HBV blood safety. To control HBV infection, the Chinese government implemented mandatory infant vaccination with three doses of hepatitis B vaccine at 0, 1, and 6 months as the highest priority in 1992 which resulted in a significant reduction of carrier rate in children from 10% to less than 1% over two decades.1 Before 1990, 95% of HBV vaccines were locally produced plasma derived and after that recombinant protein Genotype A2.8 Between 1992 and 2001, compliance was overall 64% reaching 75% in urban areas.9 A vaccinated population has enrolled as blood donors since 2010, and numbers increasing every year should impact the HBV blood safety positively. However, the previously mentioned report suggested that HBV vaccine made of Genotype A2 recombinant protein might not be fully efficient when used in areas where other genotypes are prevalent such as Genotypes B and C in China.6 To test this hypothesis, a large cohort of vaccinated Chinese blood donors recruited by Shenzhen Blood Center was identified and tested for HBV serologic and molecular markers. This study was intended to examine the effectiveness of the vaccination program and the potential threat of vaccination failures to the blood supply.
MATERIALS AND METHODS Population The target population was HBV-vaccinated blood donors born either between January 1, 1992, and April 1996 (age 18-21 years) with a history of neonatal HBV vaccination or April 1988 to December 1991 (age 22-25 years). Samples were collected in two separate periods: January to December 2012 and April to June 2013. The sample collection was limited to donors 18 to 21 years in 2012 and extended to 18 to 25 years in 2013. Hepatitis B vaccination history was 2
TRANSFUSION Volume **, ** **
collected from consenting donors with a questionnaire to all donors and a review of vaccination records for 18- to 21-year-old donors. In all cases, three doses of vaccine were injected within 24 hours of birth and at 1 and 6 months of age. HBV vaccine efficacy was defined as absence of HBV markers (HBsAg, anti-HBc, and HBV DNA). All blood donors were voluntary and nonremunerated and passed the predonation questionnaire as well as the rapid predonation testing for HBsAg and alanine aminotransferase (ALT) at the collection sites (colloidal gold strip method, Abon Diagnostics, Hangzhou, China; and Roche Refletron, Roche Diagnostics, Mannheim, Germany, respectively). The sensitivity of these HBsAg assays was 2.5 ng/mL and specificity was greater than 99%. Qualified donors (negative for the predonation tests) proceeded to donate blood. Whole blood donations were then tested by two enzyme immunoassays (EIAs) at the Shenzhen Blood Center, for HBsAg (Diasorin, Dartford, UK; and Xinchang Diagnostics, Xiamen, China), for antibody to hepatitis C virus (anti-HCV; Ortho Clinical Diagnostics, High Wycombe, UK; and Lizhu Diagnostics, Zhuhai, China), for antibody to human immunodeficiency virus Types 1 and 2 (anti-HIV-1/2; Bio-Rad, Marnes-la-Coquette, France; and Wantai Diagnostics, Beijing, China), for syphilis (Diasorin and Lizhu Diagnostics), and for ALT with a kinetic method (AusBio Biotech, Fuzhou, China). Only donors nonreactive for serologic tests and with normal ALT level were screened by NAT for HBV, HCV, and HIV-1 genomes with a multiplex assay (Procleix Ultrio, individual-donation NAT, Novartis Diagnostics, Emeryville, CA). Donors negative for all tests were enrolled randomly in this study. In both 2012 and 2013, only a subpopulation of donors in the 18- to 21-year (12.1% of total) or 22- to 25-year (22.5% of total) age group was enrolled in a proportion of 93% for 18- to 21-year-old donors and 77.6% for 22- to 25-year-old donors, respectively, after checking for evidence of vaccination. Anti-HBc was tested with a commercially available EIA (Kehua Biotech, Shanghai, China). Anti-HBc–reactive samples were further tested with an alternative commercial EIA (Wantai Biotech). Samples reactive with both assays were considered anti-HBc positive. Anti-HBs was quantified with a commercial EIA (Wantai Biotech) according to the manufacturer’s instructions. Values below 10 IU/L were considered negative. The assay was quantitative within the range 10 to 327 IU/L and higher levels were given as more than 327 IU/L. HBV DNA was extracted from 2 mL of plasma from anti-HBc–positive samples with a viral nucleic acid largevolume kit (HighPure, Roche Diagnostics) and was tested with a combination of quantitative real-time polymerase chain reaction (PCR; sensitivity, 5 IU/mL) and nested PCR procedures amplifying the basic core promoter/precore (BCP/PC), the S region, the pre-S/S region, and full genome as previously described.10 Samples providing an
HBV INFECTION IN VACCINATED BLOOD DONORS
TABLE 1. Distribution of HBV serologic markers according to age and sex in 2028 vaccinated blood donors screened for HBsAg and ALT* Age (years) 2012, Markers Sex, male/female Anti-HBs+/anti-HBc+ Anti-HBs+/anti-HBc– Anti-HBs–/Anti-HBc+ Anti-HBs–/Anti-HBc– Total (%)
2013
18-21
18-21
22-25
Total
560/589 195 (17.0) 630 (54.8) 17 (1.5) 307 (26.7) 1149 (56.7)
186/159 52 (15.1) 156 (45.2) 8 (2.3) 129 (37.4) 345 (17.0)
309/225 150 (28.1) 200 (37.5) 11 (2.1) 173 (32.4) 534 (26.3)
1055/973 397 (19.6) 986 (48.6) 36 (1.8) 609 (30.0) 2028
They were distributed in 48% females and 52% males. In total, 1494 donors were 18 to 21 years of age (1149 in 2012, 345 in 2013) and 534 were 22 to 25 years in 2013. The first age group had been included in the systematic vaccination program at birth initiated in January 1992; donors of the second age group were born in the years preceding such implementation but presented evidence of having been vaccinated.
* Data are reported as number (%).
amplicon in two regions were considered confirmed positive. Samples amplified in only one region but for which a sequence was obtained were considered positive when no element suspect of contamination was present. To avoid contamination, extraction and amplification were performed in separate rooms; each run included a negative control since the extraction step and sequences (except for BCP/PC) were scrutinized for identity and subsequent exclusion.
HBV DNA sequencing and genotype determination To confirm the HBV DNA specificity, the amplified products, obtained from the BCP/PC (295 nucleotides) and S (496 nucleotides) regions, pre-S/S (1433 nucleotides) or full genome (3162 nucleotides) were purified with a DNA gel extraction kit (AxyPrep, Axygen, Union City, CA) to remove excess primers. The purified amplicons were sent for sequencing to Shanghai Invitrogen (Guangzhou, China). Nucleotide sequences and reference genotyped sequences from Genebank were aligned and manually edited using a biologic sequence alignment editor program (BioEdit, http://www.mbio.ncsu.edu/bioedit/ bioedit.html). Genotype determination was performed by phylogenetic analysis using a computer program (MEGA5.1, Informer Technologies, Inc., Redwood City, CA). The neighbor-joining method with 1000 bootstrap replications was used and over 75% bootstrap value was considered significant.
Statistical analyses Ninety-five percent confidence interval for the observed yield rate was derived using binomial exact proportion method. Categorical variables were compared using Fisher’s exact test and, for continuous variables, the nonparametric Mann-Whitney test. A p value of less than 0.05 was used as the cutoff level for significance.
RESULTS Blood donor population The 2028 vaccinated blood donors were collected by the Shenzhen Blood Center: 1149 in 2012 and 879 in 2013.
Distribution of HBV serologic markers in vaccinated blood donors The distribution of HBV serologic markers (anti-HBs and anti-HBc) in 2028 blood donor samples is shown overall and according to donor age in Table 1 and Fig. 1. Thirty percent of vaccinated donors did not carry detectable anti-HBs 18 to 25 years after vaccination. The percentage of donors with anti-HBs only tended to decrease with age (p = 0.0001 between age 18 to 25 groups) while donor samples carrying both anti-HBs and anti-HBc increased (Fig. 1). Confirmed anti-HBc was present in 21.4% of vaccinated donors, the vast majority of samples also carrying anti-HBs (91.7%). Overall, 68.2% of the population carried quantifiable anti-HBs. The distribution of anti-HBs levels in donors age 18 to 21 years is shown in Fig. 2. The data are stratified according to the presence or absence of anti-HBc together with anti-HBs. The anti-HBc–positive samples carried significantly higher levels of anti-HBs than those anti-HBs only (p < 0.001, Fig. 2). The median level of anti-HBs in antiHBs–only samples was 136 IU/L with 17.4% of values of higher than 327 IU/L; the median level of samples that were anti-HBc and anti-HBs positive was 241 IU/L with 23% of samples more than 327 IU/L (p < 0.001). At age 18, 10.7% of vaccinated donors carried anti-HBc and 31.5% by age 25 (Fig. 1; p = 0.008 between frequency of anti-HBc in 18- and 25-year groups). This trend in distribution was independent of sex (data not shown). In donors age 22 to 25 years, the anti-HBs levels between samples that were anti-HBs and anti-HBc positive was also significantly higher than in those that were anti-HBs only (p < 0.0001; data not shown). The potential difference in efficacy of HBV vaccine between donors born before and after the universal neonatal vaccination was examined. No significant difference between these two groups was found whether in frequency of detectable anti-HBs or anti-HBs levels or sex distribution. In donor group below and above age 21, only the percentage that were anti-HBc positive significantly differed (18.2 and 30.1%, respectively; p < 0.001). There was a small population of anti-HBc– confirmed-reactive but anti-HBs–negative donors Volume **, ** **
TRANSFUSION
3
ZHENG ET AL.
confirmed positive (Table 2, first 12 samples). In addition, 433 samples carrying anti-HBc were tested by quantitative PCR (qPCR) and confirmed by nested PCR and 12 samples were positive (Table 2, last 12 samples). In the Ultrio-detected group, 11 of 12 were quantified (27 ± 66 IU/mL); in the qPCR-detected group, seven of 12 were quantified (16 ± 8 IU/mL). All samples had normal levels of ALT. Three of these samples did not contain serologic markers of HBV. One of them (V11) was shown on follow-up to be a WP infection and two (V12 and V13) remained DNA positive and serology negative 4 months after index sample and were classified as primary OBI (Table 3). Of the remaining Fig. 1. Distribution of HBV serologic markers according to age and sex in 2028 vacci21 samples 15 carried both anti-HBs nated blood donors. Open bars = no anti-HBs, no anti-HBc; closed bars = anti-HBc and anti-HBc or anti-HBs only (three only; light gray bars = anti-HBs only; dark gray bars = both anti-HBs and anti-HBc cases) or anti-HBc only (three cases; positivity. The percentage of anti-HBc in anti-HBs–positive samples increases with Table 2). Eighteen HBV DNA carriers age, irrespective of donor sex. For each age year, the number of donors is 18 years, were younger than 22 years (75%) and 150; 19 years, 351; 20 years, 427; 21 years, 570; 22 years, 109; 23 years, 162; 24 years, 15 were females (62.5%). 117; and 25 years, 146. qPCR results were confirmed by nested PCR and amplicon sequencing in each of the 24 cases. Amplicons of BCP/PC and S (Table 1, Fig. 1). Three of these samples (V3, V126, and regions were found to be positive in 21 and 15 samples, F161) contained detectable HBV DNA and sequences were respectively. The latter amplicon was sufficiently informaobtained (Table 2). The frequency of this situation was tive to identify the strain genotype (14 Genotype B and similar in 18- to 21- and 22- to 25-year-old donors (1.7 and one Genotype C; Table 2). The only Genotype C case was 2.1%, respectively). in the WP with a viral load of 243 IU/mL. Only amplicon of To obtain a complete picture of HBV infections in the BCP/PC region was obtained in nine samples and the blood donors during the period of interest, HBsAgsequences were not informative for genotyping. The alignpositive samples identified either with the predonation ment of sequences from either the six full genome or the rapid test screening or secondarily by EIA and NAT (see two pre-S/S sequences showed less than 99% identity Materials and Methods) were retrieved and tabulated. In between strains, excluding contamination. In the S region, 7512 age 18 to 21 donors seen during the 2012 period of all 15 sequences were wild type; none of the mutations sample collection 267 HBV-infected samples were identifound in OBI cases of nonvaccinated blood donors were fied (221 HBsAg reactive predonation, 34 HBsAg positive present.7 postdonation, and 12 NAT positive/HBsAg negative). All EIA HBsAg–positive samples were anti-HBc positive. The median level of EIA HBsAg was 28 ± 76 IU/mL and the Follow-up of HBsAg-negative and HBV median level of viral load was 1204 ± 5615 IU/mL. When DNA–positive vaccinated blood donors extrapolating the 18.5% anti-HBc–positive samples detected in the 1149 subgroup of age 18 to 21 donors colFor 14 HBV DNA–positive index cases, one or more lected in 2012 (Table 1), the whole donor population in follow-up samples obtained 47 to 560 days later were that age group included 1340 samples anti-HBc positive tested (Table 3). In 10 cases, index sample serologic or (7512 − 267 = 7245 × 18.5% = 1340). The total number of molecular markers remained unchanged confirming the donors exposed to or infected by HBV was 1607 initial diagnosis of OBI or primary OBI. V11, as previously (1340 + 267 = 1607)/7512 or 21.4%. indicated, presented as a WP infection, seroconverting to anti-HBc and anti-HBs at 4 and 6 months postindex sample. Donor V78 who was low positive for anti-HBc and HBV DNA in vaccinated blood donors anti-HBs in the index sample was negative for both markers 18 months later although HBV DNA remained HBV DNA was tested by NAT (Ultrio, Novartis Diagnostics) detectable. Follow-up samples from donors V5 and V48 in HBsAg-negative samples, and 12 samples were 4
TRANSFUSION Volume **, ** **
HBV INFECTION IN VACCINATED BLOOD DONORS
mandatory vaccination at birth in China (Table 1). For donors over 21 years, requirements for evidence of vaccination are less stringent than in donors 18 to 21 years (questionnaire instead of checking the vaccination booklets). The absence of a significant difference in the distribution of anti-HBs prevalence and level between the two groups ultimately justified a global analysis of the cohort. During the period of the 2012 study, a total population of 7512 18- to 21year-old donors was tested for HBV markers. HBsAg- and HBV DNA only– positive donors accounted for 3.4 and 0.15%, respectively. Adding the 18.45% prevalence of anti-HBc–positive obtained with a subset of 1149 donors (Table 1), the predicted prevalence of HBV infection markers reaches 22%. This percentage is similar to 20.9% Fig. 2. Distribution of anti-HBs levels (IU/L) in vaccinated donors (18-21 years old as found in The Gambia in a 20- to 24-yearin Table 1) carrying anti-HBs only or both anti-HBs and anti-HBc. Light gray bars = old population of vaccinees11 but higher samples carrying only anti-HBs; dark gray bars = samples carrying both anti-HBs than 14.7% found in a 19- to 21-year and anti-HBc; anti-HBs levels are stratified in five categories. The difference in antipopulation of Northern China12 and 6% HBs levels between samples with and without anti-HBc was significant (p < 0.001). reported in Taiwan.13 However, it might be underestimated since only anti-HBc– positive samples were retested for HBV DNA and the charshowed an increased level of anti-HBs, a low level of antiacterization of the Ultrio DNA–positive samples were not HBc 3.5 months after the index sample and HBV DNA all anti-HBc positive. becoming undetectable 3.5 months after the index Examining the 2028 donors 18 to 25 years tested for sample, suggesting a recent contact with HBV. anti-HBs and anti-HBc, the relatively large proportion of vaccinated donors without detectable anti-HBs (31.8%; Identifying the potential origin of HBV infection Table 1) is consistent with the 39.3% found in Northern China,11 37.3% Guangxi province by age 15,14 41.5% in 18To attempt to identify the origin of the HBV infection in vaccinated donors, parents and sexual partners were to 21-year-old Taiwanese students,13 and 25% found in called upon, informed consent was obtained, and tests The Gambia.12 In Thailand, between age 15 and 20, the were performed. Parents were available in four cases and a percentage of seronegative vaccinees ranged between 33.3 sexual partner in one case. Both parents of four donors and 36.4%.15 Figure 1 shows no significant changes of (V1, V9, V11, and V78) were tested. Only V1 mother and V9 antibody-negative donors between age 18 and 25 but a father were viremic but available sequences indicated no relatively large range (20.9 and 38.9%) year by year. more than 98% homology excluding them as source of The efficacy of HBV vaccine can be examined from infection. Among the seven other parents, all but one had different angles: the occurrence of acute hepatitis and the evidence of contact with HBV but without detectable viral occurrence of chronic, often asymptomatic, infection indiload. cated by circulating HBsAg or contact with HBV not prevented by vaccine-induced anti-HBs indicated by antiHBc. The occurrence of acute hepatitis cannot be DISCUSSION estimated since no detailed questionnaire was administered to candidate donors and ALT level was prescreened The cohort of 2028 vaccinated blood donors was considand excluded donors with abnormal levels without ered suitable for donation after negative screening with attempting to find its origin. HBsAg was detected in 3.4% of two rapid HBsAg assays, normal ALT level, and absence of 18- to 21-year-old blood donors but the presence of liver clinical acute hepatitis. These donors were further tested disease was not investigated. This percentage appears with HBsAg EIA and NAT with triplex Ultrio was higher than previously reported in China such as in Long assembled. This population was 48% female and 73% of An (2.4%) in subjects less than 20 years old, in Jiangsu donors were born after 1992, the official date of Volume **, ** **
TRANSFUSION
5
ZHENG ET AL.
TABLE 2. Characterization of HBV DNA–positive, HBsAg-negative samples from vaccinated donor ID V129 V8 V5 V7 N01 V10 V1 V4 V3 V12 V13 V11 V9 F09 F63 V48 V85 F82 F48 V76 F71 V78 V126 F161
Sex Female Male Female Female Female Female Male Male Female Female Female Female Male Male Male Male Female Male Female Female Female Female Female Male
Age (years) 21 19 20 20 23 21 20 21 21 21 18 21 20 25 23 21 21 25 24 19 25 20 21 22
Anti-HBs (IU/L) 789 400 23 23 18 86 14 10 − − − − >1000 >300 >300 890 825 315 310 55 39 25 − −
Anti-HBc + + + + + − − − + − − − + + + + + + + + + + + +
NAT Ultrio Ultrio Ultrio Ultrio Ultrio Ultrio Ultrio Ultrio Ultrio Ultrio Ultrio Ultrio qPCR qPCR qPCR qPCR qPCR qPCR qPCR qPCR qPCR qPCR qPCR qPCR
VL (IU/mL) 45 83 69 25 59 21 9 26 17 27
