Nucleic Acid Testing to Detect HBV Infection in Blood Donors

The n e w e ng l a n d j o u r na l of m e dic i n e original article Nucleic Acid Testing to Detect HBV Infection in Blood Donors Susan L. Stramer, Ph.D., Ulrike Wend, Daniel Candotti, Ph.D., Gregory A. Foster, B.A., F. Blaine Hollinger, M.D., Roger Y. Dodd, Ph.D., Jean-Pierre Allain, M.D., and Wolfram Gerlich, M.D. A bs t r ac t Background From the Scientific Support Office, American Red Cross, Gaithersburg, MD (S.L.S., G.A.F.), and the Holland Laboratory, American Red Cross, Rockville, MD (R.Y.D.); the Institute of Medical Virology, University of Giessen, Giessen, Germany (U.W., W.G.); the National Health Service Blood and Transplant (D.C.) and the Department of Haematology, University of Cambridge (J.-P.A.) — both in Cambridge, United Kingdom; and the Eugene B. Casey Hepatitis Research Center, Baylor College of Medicine, Houston (F.B.H.). Address reprint requests to Dr. Stramer at the American Red Cross, Scientific Support Office, 9315 Gaither Rd., Gaithersburg, MD 20877, or at stramers@usa.redcross.org. N Engl J Med 2011;364:236-47. Copyright © 2011 Massachusetts Medical Society. The detection of hepatitis B virus (HBV) in blood donors is achieved by screening for hepatitis B surface antigen (HBsAg) and for antibodies against hepatitis B core antigen (anti-HBc). However, donors who are positive for HBV DNA are currently not identified during the window period before seroconversion. The current use of nucleic acid testing for detection of the human immunodeficiency virus (HIV) and hepatitis C virus (HCV) RNA and HBV DNA in a single triplex assay may provide additional safety. Methods We performed nucleic acid testing on 3.7 million blood donations and further evaluated those that were HBV DNA–positive but negative for HBsAg and anti-HBc. We determined the serologic, biochemical, and molecular features of samples that were found to contain only HBV DNA and performed similar analyses of follow-up samples and samples from sexual partners of infected donors. Seronegative HIV and HCV-positive donors were also studied. Results We identified 9 donors who were positive for HBV DNA (1 in 410,540 donations), including 6 samples from donors who had received the HBV vaccine, in whom subclinical infection had developed and resolved. Of the HBV DNA–positive donors, 4 probably acquired HBV infection from a chronically infected sexual partner. Clinically significant liver injury developed in 2 unvaccinated donors. In 5 of the 6 vaccinated donors, a non-A genotype was identified as the dominant strain, whereas subgenotype A2 (represented in the HBV vaccine) was the dominant strain in unvaccinated donors. Of 75 reactive nucleic acid test results identified in seronegative blood donations, 26 (9 HBV, 15 HCV, and 2 HIV) were confirmed as positive. Conclusions Triplex nucleic acid testing detected potentially infectious HBV, along with HIV and HCV, during the window period before seroconversion. HBV vaccination appeared to be protective, with a breakthrough subclinical infection occurring with non-A2 HBV subgenotypes and causing clinically inconsequential outcomes. (Funded by the American Red Cross and others.) 236 n engl j med 364;3 nejm.org january 20, 2011 The New England Journal of Medicine Downloaded from nejm.org at UC SHARED JOURNAL COLLECTION on July 5, 2012. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved. Nucleic Acid Testing to Detect HBV in Blood T he transfusion of blood containing hepatitis B surface antigen (HBsAg) is associated with post-transfusion infection with hepatitis B virus (HBV). Blood that is free of HBsAg but has high-titer antibodies against hepatitis B core antigen (anti-HBc) in the absence of antibodies against hepatitis B surface antigen (anti-HBs) can also transmit HBV infection.1,2 In 1986, screening for anti-HBc was implemented in the United States to reduce HBV transmission and as a surrogate marker for non-A, non-B hepatitis (i.e., hepatitis C virus [HCV]).3,4 However, a small proportion of donors with anti-HBc in the absence of HBsAg have circulating HBV DNA and may have a risk of infectivity.1-6 Blood that is collected during the early window period of HBV infection is highly infectious, but this risk declines as anti-HBs develops.5,7-12 The estimated residual risk of HBV infection from donations to the American Red Cross ranges from 1 in 280,000 to 1 in 357,000 donations.13 After the introduction of nucleic acid testing for screening in minipools (pools of 6 to 16 donations), the estimated yield of HBV infection ranges from 1 in 830,000 to 1 in 2 million donations.14 However, these estimates do not capture the possibility of HBV infection in vaccinated donors who have acute infection with low or no expression of HBsAg. HBV-seronegative but infected donors have been identified by means of nucleic acid testing at a rate of approximately 1 in 600,000 donations screened15 but have not been further studied. Although the residual risk of transmission of HBV by transfusion has decreased progressively,13,16,17 it remains higher than the estimated risks for human immunodeficiency virus (HIV) and HCV (1 in 1,467,000 donations and 1 in 1,149,000 donations, respectively).18 We conducted a study to evaluate the use of nucleic acid testing in determining the number of seronegative donors who have HBV DNA and to characterize these donors according to their risk factors and progression of molecular, serologic, and biochemical markers. We used a triplex test combining the detection of HBV DNA with that of HIV and HCV RNA. We also evaluated donors who were seronegative for HIV and HCV but had positive results on nucleic acid testing. n engl j med 364;3 Me thods Screening Protocol During 2008, the American Red Cross implemented prospective screening of all blood donations, using the Procleix Ultrio assay and the TIGRIS automated platform10 (Gen-Probe and Novartis) at three of its five National Testing Laboratories. The American Red Cross collects about 42% of the U.S. blood supply from volunteer, unpaid donors in the continental United States and Puerto Rico. All donations were screened by nucleic acid testing of either individual donations or 16-sample minipools. Blood units that were released for transfusion were negative on triplex nucleic acid testing and for all standard blood-screening markers.18 All Ultrioreactive minipools were resolved to the reactive individual donation, followed by testing of the reactive individual donations (including donations screened individually by nucleic acid testing) by means of separate discriminatory assays for HBV, HIV, and HCV. Any individual donation with a positive result on nucleic acid testing was not used for transfusion (even if the donor was subsequently determined to have a false positive result), and associated donors were deferred from future donation. Seronegative donors with one or multiple reactive discriminatory tests were considered potentially infected donors for further study. For individual donations, the 95% lower limit of detection for the HBV portion of Ultrio is 10.4 IU per milliliter (approximately 50 copies per milliliter), as compared with approximately 30 copies per milliliter each for HIV and HCV. The study protocol was approved by the institutional review board of the American Red Cross. Written informed consent for follow-up studies was obtained from donors who were reactive on nucleic acid testing and from their partners. The study was conducted in accordance with the provisions of the protocol. HBV DNA Confirmation Donors who were positive for HBV DNA on the Ultrio assay and seronegative for HBsAg (lower limit of detection, 0.10 ng per milliliter) and for total (IgM and IgG) anti-HBc (HBsAg and HBcore PRISM, Abbott Laboratories) were enrolled in nejm.org january 20, 2011 The New England Journal of Medicine Downloaded from nejm.org at UC SHARED JOURNAL COLLECTION on July 5, 2012. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved. 237 The n e w e ng l a n d j o u r na l follow-up studies, which involved the collection of samples at intervals of 1 to 2 weeks until seroconversion to anti-HBc occurred. Each plasma unit from a reactive donation was retrieved for further testing. Additional testing that was repeated on all index donations, as well as performed on each follow-up sample, included duplicate Ultrio nucleic acid testing of individual donations and discriminatory testing if reactivity was observed, followed by confirmation by means of polymerase-chain-reaction (PCR) assay with the use of COBAS AmpliScreen HBV Test (Roche) with the 1-milliliter extraction method (95% lower limit of detection, 5 IU per milliliter). In addition, the retrieved plasma units from each donation with reactivity on discriminatory testing were separated into aliquots and tested in replicates of 10 to confirm reactivity. Reactivity in at least 1 of the 10 replicate tests was required for confirmation of positive results; donations with no reactivity were considered to have false positive results. Serologic testing was performed on retrieved plasma units and follow-up samples for HBsAg, total anti-HBc, qualitative and quantitative antiHBs (Monolisa, Bio-Rad Laboratories), IgM antiHBc (ARCHITECT, Abbott), and alanine aminotransferase (Quest Diagnostics). Donors whose HBV DNA reactivity could be confirmed in index samples or who had serologic evidence of HBV infection in follow-up samples were considered to have HBV infection. The sexual partner of each of these donors was contacted for follow-up, if possible. Screening of a subgroup of donors for anti-HBs was performed as an indicator of HBV vaccine penetrance in the donor population (for of m e dic i n e details, see the Supplementary Appendix, available with the full text of this article at NEJM .org). Molecular analyses of donor and partner samples were performed in the Geissen and Cambridge laboratories (see the Supplementary Appendix, available with the full text of this article at NEJM.org). HIV and HCV RNA confirmation Confirmation of the presence of HIV and HCV RNA included HIV and HCV qualitative and quantitative PCR and HCV genotyping performed at the National Genetics Institute. Antibodies against HIV types 1 and 2 (HIV-1/HIV-2 rDNA EIA, Abbott Laboratories) and HCV (HCV 3.0 ELISA, Ortho Clinical Diagnostics) were used for serologic testing of index and follow-up samples. Serologic confirmation of reactive samples was performed with tests licensed by the Food and Drug Administration.18 Sexual partners were contacted if possible, as described for HBV confirmation. Study Oversight The study was funded in part by Gen-Probe and Novartis Vaccines and Diagnostics, which provided the Ultrio test kits and TIGRIS instruments. The American Red Cross provided a clinical-trial site in support of an investigational protocol designed to determine the frequency of donors who test HBV-seronegative but DNA-positive so that an HBV DNA blood-donation-screening claim could be justified for the Ultrio assay performed on the TIGRIS automated platform. The study protocol was designed and executed by the American Red Cross and its collaborators. Table 1. Confirmed Results on Nucleic Acid Testing and Serologic Analysis of 3,694,858 Blood Donations.* Classification HBV HIV HCV no. frequency no. frequency no. frequency Positive on nucleic acid testing and serologic analysis 426†‡ 1 in 8673 231 1 in 15,995 1426 1 in 2591 Positive on nucleic acid testing only 9 1 in 410,540 2 1 in 1,847,429 15 1 in 246,324 178†§ 1 in 20,758 29 1 in 127,409 517 1 in 7147 Positive on serologic analysis only * HBV denotes hepatitis B virus, HCV hepatitis C virus, and HIV human immunodeficiency virus. † Positive results for HBV on serologic analysis indicate the presence of hepatitis B surface antigen (HBsAg). ‡ Of the 426 donations that were positive for HBV DNA and HBsAg (by neutralization), 407 (96%) were reactive to antibodies against hepatitis B core antigen (anti-HBc). § Of the 178 donations that were negative for HBV DNA and positive for HBsAg (by neutralization), 86 (48%) were reactive to anti-HBc. 238 n engl j med 364;3 nejm.org january 20, 2011 The New England Journal of Medicine Downloaded from nejm.org at UC SHARED JOURNAL COLLECTION on July 5, 2012. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved. Nucleic Acid Testing to Detect HBV in Blood Table 2. Characteristics of 15 Donors with HCV Infection and 2 with HIV Infection.* Donor No. Sex Donor Status Follow-up Age HCV RNA days yr copies/ml Genotype 55 <100 NA 3,400,000 3A 3,800,000 3A HCV 013–1 M Repeat Follow-up Partner 11 F NA 003–1 M Repeat 63 12,000 2 011–1 F First time 24 240,000 3A 012–1 M Repeat 77 380,000 1A 074–1 M First time 34 4,900,000 1A 074–2 F First time 24 6,600,000 1A 074–3 M First time 17 8,400,000 2B 011–2 M First time 38 11,000,000 3A 012–2 M Repeat 49 Follow-up 18 19,000,000 1B 370,000,000 1B 011–3 F Repeat 24 21,000,000 3A 074–4 M First time 21 31,000,000 1A 055–1 F Repeat 48 33,000,000 3A 012–3 M Repeat 52 41,000,000 3A 054–1 M Repeat 45 43,000,000 2 019–1 F First time 30 53,000,000 1A 074–5 M Repeat 16 44,000 NA 074–6 M Repeat 29 340,000 NA HIV * Donors with hepatitis C virus (HCV) or human immunodeficiency virus (HIV) are identified by a four-number designation. The first three numbers indicate the collection site, and the number after the dash indicates the donor-sequence number at the given site. Thus, at site 074, six seronegative, RNA-positive donors were identified. NA denotes not available. R e sult s Detection of HBV In 2008, a total of 3,694,858 donations from 2,137,275 donors were screened with the use of Ultrio. Of these samples, 576,940 donations (16%) were screened individually by nucleic acid testing and the remainder were screened in minipools. After virus-specific discriminatory testing, 9 HBVseronegative donors were confirmed as having HBV infection by additional testing (1 of 410,540 donations) (Table 1). Of the 9 infected donors, 7 (of whom 6 had been vaccinated) consented to participate in the follow-up study, as did the sexual partners of 4 donors. The remaining 2 donors were confirmed as having HBV infection at their index donation from the retrieved plasma unit by repeat nucleic acid testing. All but 1 of the 9 do- n engl j med 364;3 nors with HBV infection were identified by means of minipool nucleic acid testing; the remaining donor (019) required individual-donation nucleic acid testing for detection and tested nonreactive by intermediate dilutions. Of the 9 donors with HBV infection, 8 were repeat donors, and 1 was a first-time donor (055). The mean age of the infected donors was 26 years; 6 of the donors were men. Detection of HCV and HIV Fifteen HCV-seronegative donors had confirmed positive results for HCV RNA on PCR assay at their index donation. Repeat nucleic acid testing of the retrieved plasma unit further confirmed the index results, including identical genotypes and similar viral loads. The 15 HCV-infected donors are shown in order of increasing viral load at the nejm.org january 20, 2011 The New England Journal of Medicine Downloaded from nejm.org at UC SHARED JOURNAL COLLECTION on July 5, 2012. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved. 239 The n e w e ng l a n d j o u r na l of m e dic i n e Table 3. Characteristics of Nine Blood Donors with Hepatitis B Virus (HBV) Infection in Index and Follow-up Samples and in Samples from Sexual Partners.* Donor No. Sex Age Receipt of HBV Vaccine (yr) yr 001 Partner 003 Partner HBV DNA at Index Donation Duration of Detected Viremia Anti-HBs Level at Index Donation IU/ml days IU/liter M 22 Yes (1992) 11 45 to 69 Negative¶ F 23 No 1.9×108 NA Negative F 37 Yes (1981) 86 44 to 70 3 (ruled nonimmune) NA Negative M 38 No 3.6×108 F 17 Yes (1998) 13 137 to 168 11 M 19 No 2.6×108 NA Negative M 27 Yes (2001) 35 34 to 42 43 F 25 No 1.8×106 NA Negative 029** M 19 Yes (unknown)†† 30 NA 96 042 M 28 Yes (unknown)†† 27 72 to 107 33 019** F 44 No 21 NA Negative 055 M 20 No 47 70 to >125 Negative 074 F 24 No 18 73 to 115 Negative 011 Partner 013 Partner * Donors are identified by a three-number designation, indicating the collection site. At each site, there was only one seronegative, DNApositive donor. ALT denotes alanine aminotransferase, anti-HBc antibodies against hepatitis B core antigen, anti-HBs antibodies against hepatitis B surface antigen, HBsAg hepatitis B surface antigen, and NA not available. † The number preceding the genotype, or subgenotype in donors 029 and 042, indicates the number of clones of this genotype or subgenotype in the entire quasispecies. ‡ A value higher than 30 U per liter was considered to be elevated. § “Cons” indicates consensus sequence obtained by direct sequencing. The number preceding “x” indicates the number of clones carrying the substitution; “s” indicates the location of the amino acid substitution in the HBsAg gene. ¶ The donor became positive for anti-HBs (100 IU per liter) within 45 days, which was up to 24 days before seroconversion to IgM anti-HBc. ‖ The donor was reactive at the index donation. ** The donor declined enrollment in the follow-up study. †† The donor could not recall the year of vaccination. index donation (median, 11 million copies per milliliter) in Table 2. The rate of HCV RNA–positive, seronegative donations was 1 in 246,324 (Table 1). Of the 15 donors, there was a nearly even split between first-time and repeat donors, with a mean age of 40 years; 10 of the donors were men. The female sexual partner of donor 013-1 was available for testing; the two had the same 3A genotype, implying infection from a common source or through sexual transmission. Two donors had confirmed positive results for HIV RNA on PCR assay at their index donation, with viral loads of 44,000 and 340,000 copies 240 n engl j med 364;3 per milliliter (Table 2). Testing of the plasma unit confirmed the index test results. These two donors, who were both seronegative, were male repeat donors who were 16 and 29 years of age. The HIV RNA–positive donation rate was 1 in 1,847,429 donations (Table 1). In a combined analysis of study findings, among 75 samples from seronegative donors that were reactive on Ultrio and discriminatory testing, there were 26 confirmed infections (9 HBV, 15 HCV, and 2 HIV), or 1 in 153,952 donations. Thus, the positive predictive value for the Ultrio triplex assay in seronegative samples, in- nejm.org january 20, 2011 The New England Journal of Medicine Downloaded from nejm.org at UC SHARED JOURNAL COLLECTION on July 5, 2012. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved. Nucleic Acid Testing to Detect HBV in Blood Genotype or Subgenotype† Days to Seroconversion HBsAg Anti-HBc (IgM and total) ALT‡ Number of Clones Sequenced U/liter no. Amino Acid Substitutions on S Protein§ Anti-HBs days No 69 45 C2 7 to 16 20 20xsG145R, sV184G, W199 stop NA NA NA C2 31 14 Cons sG145R, 3xsG145R, sV184G No 70 20 A2 3 to 8 11 G102D, T134A 10 NA NA NA A2 17 108 168 Reactive‖ F1 4 to 7 12 Wild type NA NA NA F1 36 11 4xsY72 stop No No Reactive‖ B2 8 to 19 10 1xsG112R, 1xsA128T NA NA NA B2 38 9 Wild type NA NA Reactive‖ 7D, 4A2, 1A2/D 9 10 2xsQ54 frameshift 72 107 Reactive‖ 6D, 2A2 3 to 12 8 sY134H, sT143A NA NA NA A2 18 15 sG112R, sT118 M, sF134L 40 70 125 A2 9 to 640 18 C124Y, sD151L, sW156R No 73 115 A2 9 to 119 15 sT126A cluding virus-specific discrimination, was 35%, which is similar to the predictive value of other screening assays. Two of the 26 samples required individual-donation nucleic acid testing for detection, for a yield of 1 in 288,470 donations. The rates of detection on serologic analysis were higher than the rates on nucleic acid testing for all three viruses, underscoring the need for both nucleic acid testing and serologic screening (Table 1). Course of HBV Infection in Index Donors The HBV viral load was consistently low in all index samples from the nine infected donors (11 to 86 IU per milliliter) (Table 3). Five index samples contained only HBV DNA, with no serologic HBV markers, and four samples contained lowlevel anti-HBs (3 to 96 IU per liter), all from donors who had been vaccinated against HBV. Attempts were made to contact all nine donors. Two donors (019 and 029) did not respond; the other seven donors had viremia for 34 to at least 137 days but subsequently had undetectable lev- n engl j med 364;3 els of HBV DNA (Table 3). Samples from these seven donors were obtained regularly for 119 to 320 days. IgM anti-HBc (confirming acute infection) developed in all but one donor (013); this donor had the shortest period of viremia and did not have detectable HBsAg or total anti-HBc. One unvaccinated donor (055) had the classic pattern of window-period evolution accompanied by clinical signs of disease, with the presence of HBsAg from 40 to at least 70 days, corresponding to a peak viral load of 1.6×108 IU per milliliter and an alanine aminotransferase level of 640 U per liter. IgM and total anti-HBc appeared on day 70, and anti-HBs appeared on day 125 (Table 3 and Fig. 1A). Another unvaccinated donor (074) who did not have anti-HBs at baseline became reactive to IgM and total anti-HBc on day 73, with seroconversion to anti-HBs on day 115. Although HBsAg was never detected, there were clinical signs of disease, on the basis of an elevated alanine aminotransferase level (119 U per liter) on day 73. Follow-up studies were performed for five of the six donors with a history of HBV vaccination; nejm.org january 20, 2011 The New England Journal of Medicine Downloaded from nejm.org at UC SHARED JOURNAL COLLECTION on July 5, 2012. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved. 241 The n e w e ng l a n d j o u r na l of m e dic i n e Anti-HBc (total) CO/S, HBsAg and Anti-HBc (IgM) S/CO, Anti-HBs (IU/liter), HBV DNA (copies/ml) A Donor 055 109 108 107 106 105 104 103 102 HBV DNA Anti-HBs Anti-HBc (total) Anti-HBc (IgM) HBsAg 101 100 10−1 0 40 70 125 175 225 Days after Index Anti-HBc (total) CO/S, HBsAg and Anti-HBc (IgM) S/CO, Anti-HBs (IU/liter), HBV DNA (copies/ml) B Donor 013 103 Donor 042 105 Anti-HBs 104 HBV DNA 102 Anti-HBs 103 101 102 HBV DNA Anti-HBc (total) Anti-HBc (IgM) 101 100 HBsAg 10−1 Anti-HBc (total) 100 HBsAg 10−1 0 5 25 30 40 45 55 60 70 75 110 125 240 0 15 20 30 Days after Index Donor 011 105 50 75 105 110 125 130 Days after Index Donor 003 104 HBV DNA 104 HBV DNA 103 Anti-HBs Anti-HBs 103 102 102 Anti-HBc (IgM) 101 101 Anti-HBc (total) Anti-HBc (total) 100 10−1 0 80 100 105 120 135 165 100 Anti-HBc (IgM) HBsAg 10−1 185 0 Days after Index HBsAg 20 40 70 80 90 95 105 110 155 320 Days after Index Figure 1. Evolution of Serologic Markers for Hepatitis B Virus (HBV) after Index Donation from Five Donors with Acute HBV Infection. The results of follow-up analyses of virologic and serologic markers for HBV are shown for one donor who did not receive the HBV vaccine (Panel A) and for four vaccinated donors (Panel B). Included in the analyses are values for HBV DNA (in copies per milliliter). Values for hepatitis B surface antigen (HBsAg) and for IgM antibodies against hepatitis B core antigen (anti-HBc) are reported as signal-to-cutoff (S/CO) ratios, where values of 1.00 or more are considered to indicate reactivity. Values for anti-HBc (total) are reported as the reciprocal of S/CO since the assay format is based on competitive inhibition. Values of 10 or more for antibodies against hepatitis B surface antigen (anti-HBs) are considered to indicate immunity. Dashed lines indicate values below the threshold of detection. 242 n engl j med 364;3 nejm.org january 20, 2011 The New England Journal of Medicine Downloaded from nejm.org at UC SHARED JOURNAL COLLECTION on July 5, 2012. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved. Nucleic Acid Testing to Detect HBV in Blood four of the five were positive for anti-HBs and for HBV DNA at baseline. Transiently low levels of HBsAg were observed in two of these donors (011 and 042), whereas anti-HBs remained reactive (Table 3 and Fig. 1B). In these two donors, an increase in anti-HBs corresponded to the disappearance of HBsAg. IgM and total anti-HBc became positive on day 107 in one donor (042) and on day 168 in the other (011). Donor 013 did not have detectable HBsAg or anti-HBc for 9 months, but an increase in anti-HBs occurred on day 42, when HBV DNA became undetectable (Table 3 and Fig. 1B). The fourth donor (003) had received a plasma-derived HBV vaccine 27 years earlier; her level of anti-HBs in the index donation (3 IU per liter) was considered to be nonimmune. HBV DNA remained detectable for up to 70 days, at which time IgM and total anti-HBc appeared. HBsAg was never detected; however, an anamnestic response for anti-HBs was observed in the first follow-up sample (obtained on day 20). A fifth vaccinated donor (001) was negative for anti-HBs at baseline but had seroconversion (100 IU per liter) within 45 days, a finding that was consistent with an anamnestic response, since it preceded the detection of total and IgM anti-HBc by up to 24 days. Despite the detection of IgM antiHBc in four of the five vaccinated donors, no increased levels of alanine aminotransferase were noted, which further substantiated the inconsequential nature of these infections. HBV genotypes or subgenotypes of the 6 vaccinated donors were C2 (in donor 001), A2 (in donor 003), F1 (in donor 011), B2 (in donor 013), and sequences representing both D and A2 (in donors 029 and 042), as compared with A2 in all unvaccinated donors (P = 0.048 by Fisher’s exact test, two-tailed, for the comparison between the dominant strains in the 6 vaccinated donors and in the 3 unvaccinated donors) (Table 3 and Fig. 2, and the Supplementary Appendix). Figure 3 shows the levels of anti-HBs in 520 randomly selected donors who were not infected with HBV. Of the 231 immune donors in this subgroup (44%), more than 65% were 29 years of age or younger, which was consistent with the mean age of 26 years for the 9 donors with HBV infection, of whom 6 had also been vaccinated. have HBV infection. In all cases, the sequences of the presurface and surface (pre-S/S) regions of the HBsAg gene were nearly identical at the nucleotide level (<0.1% divergence) and the amino acid level in two laboratories, as were the full genome sequences in donor 013 and his sexual partner, which strongly suggests that infection occurred between vaccinated donors and their sexual partners who carried high viral loads (>1.0×106 IU per milliliter) (Table 3). All these partners were highly positive for HBsAg and anti-HBc but were negative for anti-HBs and IgM anti-HBc, indicating chronic carriage. Three of the four partners had mild alanine aminotransferase elevations (range, 31 to 38 U per liter), which were consistent with chronic hepatitis B. The transfer of a mutated minority variant to one donor (001) from his partner further confirms sexual transmission and immune selection. The fifth and sixth vaccinated donors were health care workers whose acute infections were probably the result of occupational exposure. Discussion Analysis of the results of blood-donor testing, along with follow-up analyses of infected donors, permits assessment of the details of early infection and modeling of the residual risk of infectivity from transfusion.13,18 In this study, we found some unexpected patterns of early HBV infection as a result of evaluating HBV nucleic acid testing among 3.7 million blood donations. On the basis of modeling studies, we anticipated two to four seronegative, HBV DNA–positive samples from minipool nucleic acid testing and one seronegative, HBV DNA–positive sample from single-donation testing.14 In fact, we found nine seronegative, HBV DNA–positive samples, and all but one were detected on minipool nucleic acid testing, for a total rate of 1 per 410,540 donations. Three of the infected donors appeared to have conventional window-period infection, which was consistent with expected findings. Unexpectedly, five of the other six infected donors had low-level anti-HBs (3 to 96 IU per liter) or a rapid anamnestic response, attributable to the receipt of HBV vaccine 7 to 27 years earlier. The five vaccinated donors had a brief, transient course of infection Origin of HBV Infection with no evidence of disease and very low or abThe sexual partners of infected donors were sent expression of HBsAg. Viremia, which was of traced when possible, and four were found to short duration, ranged from 11 to 86 IU per milli- n engl j med 364;3 nejm.org january 20, 2011 The New England Journal of Medicine Downloaded from nejm.org at UC SHARED JOURNAL COLLECTION on July 5, 2012. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved. 243 The n e w e ng l a n d j o u r na l of m e dic i n e ABO36920 74 61 100 100 90 M32138 Z35716 U9551 ARC042 cl6 ARC042 cl4 100 ARC042 cl3 ARC042 cl2 68 ARC042 cl7 ARC042-ind-cam ARC042 cl9 100 X02496 ARC029 cl11 ARC029 cl9 74 100 ARC029 cl4 ARC029 cl2 ARC029 cl6 ARC042 cl12 100 ARC042 cl8 ARC029 cl10 ARC029 cl8 M54923 D23678 D50522 98 Cons ARC013 ind 100 Cons ARC013 part ARC013-ind-cam ARC013-part-cam M38636 D23680 D50519 Cons ARC001 part 100 ARC001-part-cam 98 Cons ARC001 ind M57663 74 AY934772 AY934773 AY233290 AB194949 98 AB194950 Z35717 Z72478 AB116077 ARC042 cl5 ARC042 cl1 Cons ARC003 ind Cons ARC003 part 80 Cons ARC055 ind ARC055-ind-cam ARC003-ind-cam ARC003-part-cam 100 100 100 100 100 244 B C A1 A3 S50225 V00866 X70185 AY233286f Cons ARC074 ind 90 ARC074-ind-cam ARC029-ind-cam ARC029 cl7 ARC029 cl3 ARC029 cl1 75 Cons ARC019 ind ARC019-ind-cam ARC029 cl5 AB056513 AB064311 AB056515 70 97 AY090457 AY090454 Cons ARC011 part 100 ARC011-part-cam ARC011-ind-cam X75658 X69798 X75663 AB036910 D AB091256 AB091255 AB106564 X75664 X75657 A2 G E H F Scale bar indicates 0.005 nucleotide substitutions per site. n engl j med 364;3 nejm.org january 20, 2011 The New England Journal of Medicine Downloaded from nejm.org at UC SHARED JOURNAL COLLECTION on July 5, 2012. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved. Nucleic Acid Testing to Detect HBV in Blood Figure 2 (facing page). Phylogenetic Analysis of Nine Index Blood Donors and Their Sexual Partners in a 1200-Nucleotide Region, Including the Pre-S/S Gene. Individual strains that are included in the study are shown in red, and genotype references from GenBank are shown in black. Study samples are coded as ARC (indicating the American Red Cross), followed by a three-digit code number. The abbreviation “ind” indicates an index sample, and “part” indicates a sample from a sexual partner of an index donor. Sample codes ending with “cam” indicate consensus sequences obtained in the Cambridge laboratory. Sample codes preceded by “cons” indicate consensus of all clones obtained at the Giessen laboratory for a particular strain. For the two samples that were identified as being coinfected with two different genotypes, “cl” plus a number indicates the clone number within a sample quasispecies. In all cases, consensus sequences that were obtained either in Cambridge or in Giessen were identical or differed by fewer than three nucleotides. The letters to the right of the strains indicate the HBV genotype (B through H) or subgenotype (A1, A2, and A3). The numbers to the left of the strains indicate genomic distance calculations (bootstrap values in percentages), in which values of more than 70% are considered to be significantly different. liter at donation. Four of the five vaccinated donors who were followed were also positive for IgM anti-HBc, confirming acute infection. Smaller studies in the United States have typically shown approximately one HBV-infected donor per 600,000 donations screened.15 In countries where HBV infection is endemic, rates are much higher.10,19 The three acute infections among nonvaccinated donors were of subgenotype A2, which is the most frequent genotype in the United States and is the parent strain of the HBV vaccine. Elevations in alanine aminotransferase levels developed in two of these donors (119 and 640 U per liter) during follow-up. In contrast, only one of the six previously vaccinated donors was uniquely infected with the A2 strain; she had received the plasma-derived vaccine 27 years earlier. All other donors were infected exclusively or predominantly with non-A2 strains. None of the vaccinated donors had any signs of elevated levels of alanine aminotransferase. Overall, the data emphasize the protective effect of the vaccine, since the observed infections were transient, blunted, and without effect on liver function. Our findings thus suggest that the vaccine may be less effective for non-A2 infections. Of interest, four donors with acute HBV infection were discovered to have sexual partners who were chronic n engl j med 364;3 HBV carriers with high viral loads of wild-type (or in one case, mutated) HBV DNA. These infected donors had the same subgenotype and full genome sequences compatible with transmission. The blood of donors with acute HBV DNA– positive infection during the window period is likely to be highly infectious in transfusion recipients.11,12 The significance of infection in vaccinated donors is less clear. In one study, blood donations that were positive for HBV DNA with detectable levels of anti-HBs were not infectious in any of 22 recipients, as compared with a rate of infection of 27% among 37 recipients of blood that was devoid of anti-HBs.11 Similarly, the absence of infectivity in the presence of antiHBs has been observed in other studies.2,20-23 Conversely, blood containing HBV DNA with low-level anti-HBs (<75 IU per liter) may carry a risk of transmission leading to acute hepatitis.24 These study findings may be relevant to decisions about the need to implement screening for HBV DNA among blood donors. The rates of RNA-positive HIV and HCV infection among seronegative donations in our study (1 in 1.85 million for HIV and 1 in 246,000 for HCV) parallel those observed by the American Red Cross among more than 66 million donations tested in the 10 years since the introduction of nucleic acid testing for HIV and HCV (32 HIV-positive donations, or 1 in 2 million, and 244 HCV-positive donations, or 1 in 270,000)18 and are consistent with earlier reports.25 The characteristics of the 15 donors with HCV infection and the 2 donors with HIV infection in our study were similar to those reported previously.18 The infrequent detection of transfusiontransmitted infection contributes to the low cost-effectiveness of nucleic acid testing (for details, see the Supplementary Appendix).26,27 During 10 years of nucleic acid testing for HIV and HCV in the United States, five HIV-infected donors have transmitted the virus to six of eight living transfusion recipients. Two recipients were infected by one donor, and two other windowperiod donors each infected one recipient. In two cases, only one of two living recipients became infected; of these last two cases of differential transmission (i.e., in which one recipient became infected and the other did not), one case is unpublished.28-31 In contrast, there has been nejm.org january 20, 2011 The New England Journal of Medicine Downloaded from nejm.org at UC SHARED JOURNAL COLLECTION on July 5, 2012. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved. 245 The n e w e ng l a n d j o u r na l of m e dic i n e 1000 900 800 Anti-HBs (U/liter) 700 600 500 400 300 200 100 0 0 50 100 150 200 250 300 350 400 450 500 Donor Number Figure 3. Levels of Antibodies against Hepatitis B Surface Antigen (Anti-HBs) in 520 Randomly Selected Samples from U.S. Blood Donors. Shown are results for 520 blood donors who tested negative for hepatitis B virus (HBV) DNA, hepatitis B surface antigen (HBsAg), and antibodies against hepatitis B core antigen (anti-HBc), plotted in order of increasing level of anti-HBs antibodies. Each symbol represents an individual donor; 44% of donors were considered to have immunity against HBV (i.e., an anti-HBs value of 10 IU per liter or more). The arrows indicate anti-HBs levels that correspond to values for the six HBV-vaccinated study donors who were found to be positive for HBV DNA: negative, 3, 11, 33, 43, and 96 IU per liter. only one report of suspected transmission of HCV through transfusion.32 For HBV, there has been no documentation of transmission through transfusion in the United States since the introduction of ultrasensitive screening for HBsAg in 2006. The two cases of differential transmission of HIV that were observed by the American Red Cross, in which a high-volume plasma component transmitted the virus but packed red cells containing low amounts of plasma did not, indicate that the sensitivity of nucleic acid testing, even in minipools, may approach the threshold for infection. In any event, since the implementation of nucleic acid testing and sensitive serologic assays, viral transmission by transfusion from donors who previously had negative results on both serologic and nucleic acid testing is a rare finding27 (for details, see the Supplementary Appendix). In summary, this study showed a higherthan-expected rate of HBV infection with the use 246 n engl j med 364;3 of triplex nucleic acid testing, mainly in donors who had been vaccinated against HBV and who would not have been identified by routine screening for HBsAg or anti-HBc. However, these acute HBV infections rapidly resolved and are of inconsequential clinical significance, but their potential for transmission remains unresolved. Our findings show the efficacy of the HBV vaccine for the prevention of clinical disease but not infection, and the cost of interdicting donations that contain HBV DNA from seronegative donors is high in the face of unknown benefit. Because of the high cost of introducing new screening assays for blood donors and the inability to document cost-effectiveness similar to that of other medical interventions, the continued development of new tests either to replace older tests or to detect newly identified agents33 is in question. Supported by the American Red Cross, Novartis Diagnostics, and Gen-Probe and by grants from the Eugene B. Casey Foundation and the William and Sonya Carpenter Fund (to Dr. Hollinger), from the Robert Koch Institute (to Ms. Wend and Dr. nejm.org january 20, 2011 The New England Journal of Medicine Downloaded from nejm.org at UC SHARED JOURNAL COLLECTION on July 5, 2012. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved. Nucleic Acid Testing to Detect HBV in Blood Gerlich), and from the U.K. National Health Service Blood and Transplant (to Dr. Candotti). Abbott Laboratories and Bio-Rad Laboratories provided additional testing support. Disclosure forms provided by the authors are available with the full text of this article at NEJM.org. We thank Rebecca L. Townsend and David E. Krysztof of the Scientific Support Office of the American Red Cross and Jaye P. Brodsky of Quality Analytics for managing the study database and follow-up of donors and partners, the American Red Cross staff and staff at Community Blood Centers of Florida for their participation in the study, and Adrian Fisher for assistance in the preparation of the manuscript. References 1. Hoofnagle JH, Seeff LB, Bales ZB, Zimmerman HJ. Type B hepatitis after transfusion with blood containing antibody to hepatitis B core antigen. N Engl J Med 1978;298:1379-83. 2. Mosley JW, Stevens CE, Aach RD, et al. Donor screening for antibody to hepatitis B core antigen and hepatitis B virus infection in transfusion recipients. Transfusion 1995;35:5-12. 3. Stevens CE, Aach RD, Hollinger FB, et al. Hepatitis B virus antibody in blood donors and the occurrence of non-A, non-B hepatitis in transfusion recipients: an analysis of the Transfusion-Transmitted Viruses Study. Ann Intern Med 1984; 101:733-8. 4. Mosley JW, Huang W, Stram DO, et al. Donor levels of serum alanine aminotransferase activity and antibody to hepatitis B core antigen associated with recipient hepatitis C and non-B, non-C outcomes. Transfusion 1996;36:776-81. [Erratum, Transfusion 1997;37:109.] 5. Raimondo G, Allain JP, Brunetto MR, et al. Statements from the Taormina expert meeting on occult hepatitis B virus infection. J Hepatol 2008;49:652-7. 6. Hollinger FB. Hepatitis B virus infection and transfusion medicine: science and the occult. Transfusion 2008;48:1001-26. 7. Yoshikawa A, Gotanda Y, Minegishi K, et al. Lengths of hepatitis B viremia and antigenemia in blood donors: preliminary evidence of occult (HBsAg-negative) infection in the acute stage. Transfusion 2007;47:1162-71. 8. Manzini P, Abate ML, Valpreda C, et al. Evidence of acute primary occult hepatitis B virus infection in an Italian repeat blood donor. Transfusion 2009;49:75764. 9. Bremer CM, Saniewski M, Wend UC, et al. Transient occult hepatitis B virus infection in a blood donor with high viremia. Transfusion 2009;49:1621-9. 10. Phikulsod S, Oota S, Tirawatnapong T, et al. One-year experience of nucleic acid technology testing for human immunodeficiency virus Type 1, hepatitis C virus, and hepatitis B virus in Thai blood donations. Transfusion 2009;49:1126-35. 11. Satake M, Taira R, Yugi H, et al. Infectivity of blood components with low hepatitis B virus DNA levels identified in a lookback program. Transfusion 2007;47: 1197-205. 12. Tabuchi A, Tanaka J, Katayama K, et al. Titration of hepatitis B virus infectivity in the sera of pre-acute and late acute phases of HBV infection: transmission experiments to chimeric mice with human liver repopulated hepatocytes. J Med Virol 2008;80:2064-8. 13. Zou S, Stramer SL, Notari EP, et al. Current incidence and residual risk of hepatitis B infection among blood donors in the United States. Transfusion 2009; 49:1609-20. 14. Center for Biologics Evaluation and Research. Blood Products Advisory Committee, 94th Meeting, Gaithersburg, MD, April 1, 2009 (transcript). (http://www.fda .gov/downloads/AdvisoryCommittees/ CommitteesMeetingMaterials/Blood VaccinesandOtherBiologics/BloodProducts AdvisoryCommittee/UCM155628.pdf.) 15. Linauts S, Saldanha J, Strong DM. PRISM HBsAg detection of hepatitis B virus minipool nucleic acid testing yield samples. Transfusion 2008;48:1376-82. 16. O’Brien SF, Xi G, Fan W, et al. Epidemiology of hepatitis B in Canadian blood donors. Transfusion 2008;48:2323-30. 17. Daniels D, Grytdal S, Wasley A. Surveillance for acute viral hepatitis — United States, 2007. MMWR Surveill Summ 2009;58:1-27. 18. Zou S, Dorsey KA, Notari EP, et al. Prevalence, incidence, and residual risk of human immunodeficiency virus and hepatitis C virus infections among United States blood donors since the introduction of nucleic acid testing. Transfusion 2010;50:1495-504. 19. Hollinger FB, Sood G. Occult hepatitis B virus infection: a covert operation. J Viral Hepat 2010;17:1-15. [Erratum, J Viral Hepat 2010;17:600.] 20. Aach RD, Alter HJ, Hollinger FB, et al. Risk of transfusing blood containing antibody to hepatitis-B surface antigen. Lancet 1974;2:190-3. 21. Prince AM, Lee DH, Brotman B. Infectivity of blood from PCR-positive, HBsAgnegative, anti-HBs-positive cases of resolved hepatitis B infection. Transfusion 2001;41:329-32. 22. Dreier J, Kröger M, Diekmann J, Götting C, Kleesiek K. Low-level viraemia of n engl j med 364;3 nejm.org hepatitis B virus in an anti-HBc- and antiHBs-positive blood donor. Transfus Med 2004;14:97-103. 23. Gerlich WH. Breakthrough of hepatitis B virus escape mutants after vaccination and virus reactivation. J Clin Virol 2006;36:Suppl 1:S18-S22. 24. Levicnik-Stezinar S, Rahne-Potokar U, Candotti D, Lelie N, Allain JP. Anti-HBs positive occult hepatitis B virus carrier blood infectious in two transfusion recipients. J Hepatol 2008;48:1022-5. 25. Stramer SL, Glynn SA, Kleinman SH, et al. Detection of HIV-1 and HCV infections among antibody-negative blood donors by nucleic acid–amplification testing. N Engl J Med 2004;351:760-8. 26. Jackson BR, Busch MP, Stramer SL, AuBuchon JP. The cost-effectiveness of NAT for HIV, HCV, and HBV in whole-blood donations. Transfusion 2003;43:721-9. 27. Kleinman SH, Lelie N, Busch MP. Infectivity of human immunodeficiency virus-1, hepatitis C virus, and hepatitis B virus and risk of transmission by transfusion. Transfusion 2009;49:2454-89. 28. Delwart EL, Kalmin ND, Jones TS, et al. First report of human immunodeficiency virus transmission via an RNAscreened blood donation. Vox Sang 2004; 86:171-7. 29. Phelps R, Robbins K, Liberti T, et al. Window-period human immunodeficiency virus transmission to two recipients by an adolescent blood donor. Transfusion 2004; 44:929-33. 30. Stramer SL, Chambers L, Page PL, Wagner AG, Gibble J. Third reported US case of breakthrough HIV transmission from NAT screened blood. Transfusion 2003;43:Suppl:40A-41A. abstract. 31. Laffoon B, Crutchfield A, Levi M, et al. HIV transmission through transfusion — Missouri and Colorado, 2008. Morb Mortal Wkly Rpt 2010;59:1335-9. 32. Taylor C, Price TH, Strong DM. Possible HCV transmission from blood screened by pooled nucleic acid testing. Transfusion 2002;42:Suppl:9S. abstract. 33. Stramer SL, Hollinger FB, Katz LM, et al. Emerging infectious disease agents and their potential threat to transfusion safety. Transfusion 2009;49:Suppl 2:1S29S. Copyright © 2011 Massachusetts Medical Society. january 20, 2011 The New England Journal of Medicine Downloaded from nejm.org at UC SHARED JOURNAL COLLECTION on July 5, 2012. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved. 247