Two Percent of HIV-Positive U.S. Blood Donors Are Infected with Non-subtype B Strains ABSTRACT

AIDS RESEARCH AND HUMAN RETROVIRUSES Volume 19, Number 12, 2003, pp. 1065 1070 Mary Ann Liebert, Inc. Two Percent of HIV-Positive U.S. Blood Donors Are Infected with Non-subtype B Strains ERIC L. DELWART, 1,2 SHARYN ORTON, 3 BHARAT PAREKH, 4 TRUDY DOBBS, 4 KENNETH CLARK, 5 and MICHAEL P. BUSCH 1,6 ABSTRACT To estimate the prevalence of HIV strains other than the predominant HIV-1B subtype in the U.S. blood donor population we genetically and serologically characterized HIV in infected blood donations collected throughout he United States from 1997 to mid-2000. Using a combination of DNA heteroduplex mobility and DNA sequence analyses of the env and gag regions of HIV-1 we determined that 285 of 312 infections were caused by HIV-1B and six by non-subtype B HIV-1 (four HIV-1C, one HIV-1AE, one HIV-1A). Genetic distances of greater than 14% in the envelope V3 V5 region of the four HIV-1C strains indicated that they did not share a recent common origin. HIV-1 group M, N, and O, and HIV-2 specific peptide serological testing of the 20 PCR-negative samples determined that one infection was caused by HIV-2 and none by HIV-1 group N and O. The major risk factor for infection with a non-hiv-1b strain was sex with an HIV-infected person from Africa although three of seven non-hiv-1b-infected subjects did not fit that category. For four of seven non-hiv-1b-infected subjects the subtype detected was consistent with the African country of origin of the infected person or of their sexual partner. The frequency of genetically confirmed non-subtype-b HIV infection in a geographically dispersed group of infected U.S. blood donors in 1977 2000 was therefore 2.0% (6/312). INTRODUCTION THE GENETIC DIVERSITY OF HIV within the United States is of interest for tracking the spread of imported variant strains, ensuring the sensitivity of serological screening and viral RNA quantitation methods, and may possibly influence the efficacy of HIV-1B-based vaccine immunogen. 1 7 Unlike Europe, where multiple subtypes of HIV-1 cocirculate, 8 10 a single subtype (subtype B) of the main (M) group of HIV-1 (HIV-1B) still predominates in the United States. 11 An early and rapid spread of HIV-1B, possibly from a point source introduction in the midto late 1970s, 12 may account for the predominance of a single HIV-1 subtype in the United States. Nevertheless recent analyses have detected nonsubtype B HIV-1 in the United States. Subtypes A, C, D, E, F, G, A/J, as well as HIV-1 group O infections have been reported in non-national U.S. residents and in U.S. citizens likely to have been infected abroad. 13 18 Transmissions of non-hiv-1b that appeared to have occurred in the United States include 3A, 1F, 1E, 1B/F, 1A/B, and 1 A/B/E recombinant subtype viruses. 18 22 As part of a continuing effort to monitor the genetic diversity of HIV in the United States we attempted to genetically subtype 316 HIV strains from blood donors identified in U.S. blood collection sites as anti-hiv-1 and/or HIV-2 seropositive on the basis of routine laboratory screening of whole-blood donations given between 1997 and 2000, and subsequently enrolled in the CDC HIV Donor Study. 23 Subjects MATERIALS AND METHODS A total of 7.2 million blood donations were collected in 1997 2000 from the 18 participating blood centers located 1 Blood Systems Research Institute, San Francisco, California 94118. 2 Department of Medicine, University of California, San Francisco, California. 3 Transmissible Diseases Department, American Red Cross Holland Laboratory, Rockville, Maryland. 4 HIV Immunology Branch, Centers for Disease Control and Prevention, Atlanta, Georgia. 5 Division of HIV/AIDS Prevention, Centers for Disease Control and Prevention, Atlanta, Georgia. 6 Department of Laboratory Medicine, University of California, San Francisco, California. 1065

1066 throughout the United States. Seven hundred thirty-two donors (0.01%) were confirmed as HIV positive using standard blood banking criteria, i.e., repeat reactive by an HIV-1/HIV-2 enzyme immunoassay (3A77 EIA, Abbott Laboratories, Abbott Park, IL) and confirmed by Western Blot (Biotech; Rockville, MD) using CDC/FDA criteria (two bands from p24, gp41, and/or gp120/160). These donors were counseled and invited to participate in the study. If the donor was willing to participate (n 5 428), consent was obtained at that time for the following activities: willingness to participate in an interview about their sexual practices, drug use, and knowledge of AIDS and HIV transmission; blood draw for a complete blood count (CBC), white blood cell (WBC), and lymphocyte subset determination; and storage of blood for future testing when new research tests are developed. All the available demographic characteristics of the enrolled and nonenrolled subjects were identical (p $ 0.05 for age, gender, first time or repeated blood donor, year of donations, regions, hepatitis B core, hepatitis B surface antigen, syphilis, hepatitis C virus (HCV), and human T cell leukemia virus (HTLV) I/II antibody reactivities). Because the race/ethnicity of donors was not routinely determined by most participating blood banks during the initial blood donation and was removed from the record of the nonenrolled subjects a comparison of the race/ethnicity of the enrolled and nonenrolled subjects was not possible. Subjects whose HIV were genetically and serologically analyzed here did not significantly differ from the other enrolled subjects in any of the above demographic characteristics including race/ethnicity. The study protocol was reviewed and approved by the Committee on Human Research at the University of California, San Francisco. PCR, genetic analysis, and serology DNA was extracted from dimethyl sulfoxide (DMSO)-preserved buffy coat or peripheral blood mononuclear cell (PBMC) preparations using a DNA extraction kit (QIAGEN, San Clarita, CA). One microgram of genomic DNA was used to initiate nested polymerase chain reaction (npcr) using previously described primer pairs (ED5 ED12 followed by ES7 ES8) amplifying a 700-bp V3 V5 envelope fragment. 24 When amplifications of env genes remained negative following multiple attempts, a nested set of primers directed at the gag region designed to amplify all subtypes of HIV-1 M group was used. 25 Plasma was obtained from the subjects whose PBMC DNA remained PCR negative. Viral RNA was extracted from 140 ml of plasma using a viral RNA extraction kit (QIAGEN, San Clarita, CA) and resuspended in 30 ml of water with 3 U of RNasin (Promega, Madison, WI); 11.5 ml RNA was reverse transcribed using RNase H negative reverse transcriptase (Gibco BRL, Gaithersburg, MD) and 1.5 mg of six-nucleotide-long random oligonucleotides (Gibco BRL Life technologies, Rockville, MD) to prime reverse transcription. Five of the 20 ml of cdna so generated was used to initiate env or gag npcrs. HMA subtype classification was based on the electrophoretic mobility of DNA heteroduplexes formed between the PCR amplicons from each infected blood donor and amplicons derived from prototypes of HIV-1 from each of the major HIV- 1 subtypes. 24 26 Maximum likelihood phylogenetic trees were constructed with the PHYLIP dnaml program with global rearrangements and randomized input orders (http://evolution. genetics.washington.edu/phylip.html). 27 Evidence for intersubtype recombination in the untypable variant (U-US35) was tested using the subtyping http://www.ncbi.nlm.nih.gov/retroviruses/ subtype/subprehiv1.html) and RIP (http://hiv-web.lanl.gov/rip/ RIPsubmit.html) programs on the NCBI and Los Alamos National Laboratory HIV database website. Sequence information is reported as GenBank accession numbers AF327735 to AF327737 and AY039749 to AY039751. V3 peptides from HIV- 1 group M, N, and O were used to detect HIV-1 M-, N-, or O- specific antibodies in non-pcr amplifiable samples. Immunodominant gp41 or gp36 peptides from HIV-1 or HIV-2 were also used for serotyping HIV-1 and HIV-2 infections. 21,28 Peptides were coated on microwells of 96-well plates at 0.5 mg/well. 28 Plasma specimens were diluted 1:400 and used to detect peptide-specific antibodies. RESULTS DELWART ET AL. Blood samples from a total of 316 HIV-positive donors enrolled in this study were tested for PCR amplifiable HIV-1 sequences. PBMC DNA was first analyzed by PCR using primer pairs amplifying an env V3 V5 region fragment. 24 These nested env primers pairs used have been shown to amplify all known subtypes of the M (main) group of HIV-1. 24,29 The genomic DNA samples that tested negative for HIV-1 env were further analyzed by npcr using gag specific primers also designed to amplify all subtypes of the HIV-1 group M. 25 When genomic DNA was negative for both env and gag, cdna generated from plasma was tested using the nested npcr primer sets for both regions. Using these methods 291 (92.0%) of the 316 cases yielded PCR amplicons (284 env and 7 gag), which were then subtyped by env and gag heteroduplex mobility analysis (HMA) 24 26 ; 285 HIV-1 strains belonged to subtype B, four to subtype C, and two to subtype A. All four viruses subtyped as HIV-1C in the env HMA were also subtyped in the gag region. HMA showed these four viruses to have subtype C gag sequences and therefore they did not appear to be recombinant viruses. Env PCR fragments were also purified and directly sequenced. Sequence alignment followed by gap stripping and maximumlikelihood distance measurement between the four subtype C variants indicated that they diverged by.14% (Fig. 1). One subtype A virus could be amplified only in the gag region while the other subtype A virus was PCR positive only in the env region. Both PCR fragments were directly sequenced. The 10 sequences most closely related to the gag A variant in the Los Alamos HIV database belonged to the common recombinant form HIV-1CRF01-AE from Thailand. The subtype A env sequence (using HMA) was phylogenetically most closely related to subtype A env sequences from Africa. Because of its large degree of divergence from other subtype A env variants (including the new subsubtype A2 30 ), and its outlier status on the phylogenetic tree, this U.S. strain was labeled as an untyped (U) HIV-1 strain (Fig. 1). Trees constructed using the PHYLIP dnadist and neighbor programs produced trees with similar topologies (data not shown). The large degree of genetic divergence between these six non-b subtype HIV-1 strains indicated that they did not share a recent common ancestry and were therefore independently introduced into the United States. Based on V3 loop sequences

NON-HIV-1B SUBTYPES IN U.S. BLOOD DONORS 1067 FIG. 1. HIV-1 env V3 V5 sequences from non-hiv-1b-infected donors were aligned with sequences representative of different subtypes of group M HIV-1 and a phylogenetic tree derived by the maximum likelihood method. The 4 HIV-1C and the HIV-1U U.S. strains are highlighted in bold. all non-b viruses were inferred to be nonsyncytium-inducing, CCR5-dependent strains. 31 Plasma samples from the 25 subjects from which neither HIV-1 M env nor gag regions could be amplified from DNA or plasma RNA were tested for antibodies to HIV-1 group M, N, and O and HIV-2 peptides using an EIA serotyping assay. 28 Twenty of 25 plasma samples showed the presence of antibodies to HIV-1 group M peptides while one other sample reacted only to HIV-2 peptides. No antibody reactivity to HIV-1 groups N and O peptides were detected. Serotyping using HIV- 1 group M peptides from different subtypes 28 suggested the possible presence among the PCR-negative samples of nine subtype B serotypes, one subtype D serotype, one subtype C serotype, and nine samples whose HIV-1 group M serotype was indeterminate. Serotyping of the PCR-negative plasma samples therefore indicated the possible presence of one more subtype C and one subtype D infection. The inability to amplify HIV- 1 DNA and RNA from these 20 group M peptide reactive samples may reflect proviral and plasma viral loads that were too low to permit their amplification, sequence polymorphisms at PCR primer binding sites, or sample degradation. The five plasma samples that tested negative for HIV-1 group M, N, and O peptide antibodies were further tested for HIV antibody reactivity using the Genetic Systems combined HIV- 1/HIV-2 antibody EIA (Bio-Rad Laboratories, Redmond, WA), HIV-1 Western blots (Bio-Rad Laboratories, Hercules, CA), and for the presence of HIV-1 RNA using the Roche Amplicor HIV- 1 Monitor assay (v1.5 with a sensitivity of 400 copies/ml). Four of the five samples were negative on HIV-1/2 EIA and all were negative for HIV-1 viral RNA. The one HIV-1/2 EIA reactive sample was the same sample that was also reactive to the HIV- 2 peptide EIA. That sample had an indeterminate HIV-1 Western blot banding pattern, and was further classified as HIV-2 infected by specific serology using Select-1 and Select-2 EIAs (Biochem Immunosystems, Montreal, Canada), and an HIV-2 Western blot (GeneLabs, Singapore). Therefore, by all serological criteria tested, this blood donor appeared infected with HIV-2. The four remaining HIV-1 PCR-negative samples that tested negative by all serological (HIV-1 M, N, O group and HIV-2 peptides serotyping, combined HIV-1/2 peptide EIA, and anti-hiv- 1 Western blot) and virologic tests (HIV-1 amplicor RNA load, gag and env npcr with DNA and cdna) are therefore likely to represent falsely identified HIV-positive blood donations. Overall one infected donor (2%) of 50 tested from 1997 was found to be infected with a non-hiv-1b (one HIV-1C), two (1.8%) of 113 from 1998 were non-hiv-1b (one HIV-1C and one HIV-2), two (2.0%) of 97 from 1999 were non-hiv-1b (two HIV-1C), and two (6.0%) of the 33 from 2000 were non-hiv- 1B (one HIV-1CRF01_AE and one HIV-1U). A further two infections from 1998 and 2000 may also have been caused by non- HIV-1B viruses using the less stringent serotyping criterion. 32 34 The available demographics of the HIV-1B-infected, PCRnegative and non-hiv-1b-infected blood donors are shown in Table 1. The predominant risk factors for non-hiv-1b infection were having sex with a person from Africa. The HIV-2 infection was found in a male donor born in the Ivory Coast who acknowledged having sex with someone from the Ivory Coast, a country with a high rate of HIV-2 infection. 35 Subtype C infection was detected in a U.S.-born black male blood donor who acknowledged having sex with someone from Namibia, a black female born in Zimbabwe who acknowledged having had sex with someone from Zimbabwe, and a black female born in Ethiopia who reported having had sex with someone from Ethiopia. HIV-1C is the dominant subtype in these three African countries (http://hiv-web.lanl.gov/geography/). The fourth subtype C was found in a U.S.-born black female whose only identified risk factor was having sex with an intravenous drug user. The HIV-1U (or subtype A outlier variant) was found in a U.S.- born female of unknown race who acknowledged having sex with a bisexual man, while the HIV-1CRF01_AE virus was found in a U.S.-born white male with no identified risk factors.

1068 TABLE 1. TEMPORAL, DEMOGRAPHIC, AND RISK FACTOR CORRELATION OF GENETICALLY CONFIRMED HIV-1B AND NON-HIV-1B INFECTION IN U.S. BLOOD DONORS a HIV-1B PCR negative Non-HIV-1B b Year donation 1997 50 0 1 1998 114 5 2 1999 92 7 2 2000 29 8 2 Demographic Gender Male 185 10 3 Female 100 10 4 Race Black 138 7 5 Hispanic 70 7 0 White 65 1 1 Other 4 0 0 Unknown 8 4 1 Age,20 29 0 0 20 30 73 9 3 30 40 96 6 3 40 50 63 3 1 50 60 23 0 0 60 70 1 2 0 Location East 18 0 2 North 24 1 1 North East 127 6 0 Puerto Rico 35 5 0 South 14 1 0 South East 55 4 3 West 12 3 1 Risk factors No identified 128 11 1 risk Sex with IDU 40 2 1 MSM 68 4 0 Sex with HIV 1 23 1 0 Sex with 4 1 bisexual IDU 9 0 MSM 1 IDU 2 0 Sex with 6 1 4 someone from Africa Sex with 5 1 0 someone from Haiti a Of samples analyzed 312 from a total of 732 HIV-1/2-positive infections identified by participating blood banks. Four subsequently identified HIV-seronegative samples not included. Association between demographic groups and non- HIV-1B infection was measured using Fisher s exact test. Only sex with someone from Africa was correlated with non-hiv- 1B infection (p, 0.05). IDU, intravenous drug user; MSM, men who have sex with men. b Serologically identified HIV-2 case included in non-hiv- 1B column. DISCUSSION DELWART ET AL. This study reports a genetically confirmed frequency of non- HIV-1B strains of HIV of 2.0% in infected blood donations collected from throughout the United States in 1997 2000. Four of the seven cases of non-hiv-1b infections were likely acquired following sex with an HIV-positive African in the United States or abroad. For the remaining three cases the infected persons were U.S. born and did not report sex with someone from Africa, indicating that their viral strain may have already been transmitted within the United States. HIV-positive samples analyzed were detected using Food and Drug Administration (FDA)-approved blood bank serological tests and interpretive criteria. Blood banks employ the latest generation and highest sensitivity EIAs available, which are based on selected immunodominant HIV-1 and HIV-2 recombinant proteins detected using an antigen sandwich EIA configuration. These assays are validated to detect antibodies to all the different subtypes of the main (M) group of HIV-1, to HIV- 2, and to 90% of sera from HIV-1 group O-infected subjects, but they carry no FDA approved claim regarding their sensitivity to HIV-1 group N. 7,36 Hence it is possible that some infected donors with highly divergent HIV with poor cross-reactivity to antigens represented in FDA-licensed donor screening assays were not detected and therefore not included in our study. This would result in an underestimate of the frequency of non- HIV-1B-infected donors. A prior HIV-1 subtyping study of 142 hemophiliacs and blood donors identified as seropositive in 1985 showed the presence of only HIV-1B strains. 18,37 Analysis of 383 samples also collected throughout the United States from infected blood donors in 1993 1996 showed a non-hiv-1b frequency of 0.8% (one HIV- 1A, one HIV-1CRF02_AG, and one HIV-1C). 18,37 In this 1997 1999 study we identified seven non-hiv-1b among 292 samples subtyped (2.3%). When comparing the frequency of non- HIV-1B strains among hemophiliacs and/or blood donors in either 1982 1985 and 1993 1996 with that found in 1997 2000, a trend toward increasing frequency of non-hiv-1b strains was seen; however, this was not statistically significant with p-values of 0.06 and 0.08, respectively (Fisher s exact test). The 2.3% rate of non-hiv-1b viruses collected from blood donors throughout the United States in 1997 to 2000 was marginally higher than the 1.2% rate of non-hiv-1b reported for AIDS patients collected in 1997 1998 in the Bronx, NYC 21 (Fisher s exact test p 5 0.08) and the 0.6% rate of non-hiv- 1B detected at Stanford University Hospital in 1997 2000 38 (Fisher s exact test p 5 0.005). Continued monitoring of HIV genetic diversity in blood donors and other populations from throughout the United States will determine whether increasingly diverse strains of HIV are detected and transmitted, as has been seen in other regions of the world. Such surveillance is important for several reasons. Serological and nucleic acid screening and diagnostic monitoring assays need to be sensitive for detecting all circulating strains. Most tests currently licensed in the United States are based on early prototype U.S. HIV-1B isolates. The sensitivity of these tests to non-hiv-1b strains depends on the extent of serological cross-reactivity and nucleic acid sequence homology. A number of recent studies have documented failure of some an-

NON-HIV-1B SUBTYPES IN U.S. BLOOD DONORS 1069 tibody and RNA assays to sensitively detect or accurately quantify non-b HIV infections. 1,3 7,39 As these infections become more common it will be critical to ensure that assays with enhanced sensitivity for these infections are licensed and their use implemented. Second, the extent of diversity of HIV in a population is an important consideration in the development of candidate HIV vaccines. Similar to diagnostic assays, the earliest prototype HIV vaccines were based on envelope proteins from T cell line adapted U.S. HIV-1B isolates. These vaccines have shown to have very limited capacity to induce cross-neutralizing antibodies for primary strains of HIV-1B. 40 More recent vaccine candidates incorporate or express proteins from additional regions of the HIV-1B genome, as well as non-hiv-1b epitopes. It is possible that different vaccines may be needed to induce protective immunity for different subtypes of HIV-1 and HIV- 2, and therefore that regional HIV strain distributions will need to be monitored to appropriately target vaccines to each population. It is worth noting that the present study documented four cases in which donors were misdiagnosed as HIV infected. Follow-up testing of cells and plasma from these four donors showed no evidence of HIV antibodies or nucleic acids using a battery of serological and nucleic acid amplification assays. These donors also denied risk factors for HIV infection. False positive HIV results may be due to sample labeling error, sample cross-contamination, or nonspecificity of assays and interpretive criteria (i.e., the 1992 revision of Western blot interpretive criteria deleted the requirement for a p31 Western blot band.) 41 The rate of false-positive HIV-1 test results in the current study (four out of 320 HIV positive or 1.25%) is lower than the 4.8% (20 of 421) rate previously reported in a study of HIV-1 Western blot-positive U.S. blood donors identified from 1991 1995. 41 Nonetheless, each case of misdiagnosed HIV infection may have traumatic consequences for the donor. Hence, we agree strongly with the recommendation that repeat testing of a follow-up sample should occur prior to final diagnosis of HIV infection. 41,42 This is particularly important when testing low prevalence populations, when the Western blot band pattern is indeterminate, and when the subject denies risk factors following disclosure of the initial positive test results. In conclusion, we have documented a low but persistent level of non-hiv-1b infections in the U.S. blood donor population. Given the inherent propensity of HIV toward increasing diversity and recombination and the extent of viral global trafficking, it is important that we maintain a high level of vigilance for divergent variants and develop diagnostics and vaccines that are effective against all HIV strains. ACKNOWLEDGMENTS We thank Dr. Brian Foley of the Los Alamos National Laboratory HIV database for performing the phylogenetic analysis, Maya Royz for technical assistance, and Dr. Teresa Finlayson for demographic analyses. This study was supported by CDC cooperative agreements U64/CCU902948-14, NIH R01- AI43895, and a grant from the Blood Systems Foundation. REFERENCES 1. Emery S, Bodrug S, Richardson BA, et al.: Evaluation of performance of the Gen-Probe human immunodeficiency virus type 1 viral load assay using primary subtype A, C, and D isolates from Kenya. J Clin Microbiol 2000;38:2688 2695. 2. Jenny-Avital ER, Beatrice ST: Erroneously low or undetectable plasma human immunodeficiency virus type 1 (HIV-1) ribonucleic acid load, determined by polymerase chain reaction, in West African and American patients with non-b subtype HIV-1 infection. Clin Infect Dis 2001;32:1227 1230. 3. Chew CB, Herring BL, Zheng F, et al.: Comparison of three commercial assays for the quantification of HIV-1 RNA in plasma from individuals infected with different HIV-1 subtypes. J Clin Virol 1999;14:87 94. 4. Parekh B, Phillips S, Granade TC, Baggs J, Hu DJ, and Respess R: Impact of HIV type 1 subtype variation on viral RNA quantitation. AIDS Res Hum Retroviruses 1999;15:133 142. 5. Gobbers E, Fransen K, Oosterlaken T, et al.: Reactivity and amplification efficiency of the NASBA HIV-1 RNA amplification system with regard to different HIV-1 subtypes. J Virol Methods 1997;66:293 301. 6. Apetrei C, Loussert-Ajaka I, Descamps D, et al.: Lack of screening test sensitivity during HIV-1 non-subtype B seroconversions. AIDS 1996;10:F57 60. 7. Schable C, Zekeng L, Pau CP, et al.: Sensitivity of United States HIV antibody tests for detection of HIV-1 group O infections. Lancet 1994;344:1333 1334. 8. Simon F, Loussert-Ajaka I, Damond F, Saragosti S, Barin F, and Brun-Vezinet F: HIV type 1 diversity in northern Paris, France. Aids Res Hum Retroviruses 1996;12:1427 1433. 9. Parry JV, Murphy G, Barlow KL, et al.: National surveillance of HIV-1 subtypes for England and Wales: Design, methods, and initial findings. J Acquir Immune Defic Syndr 2001;26:381 388. 10. Fleury H, Recordon-Pinson P, Caumont A, et al.: HIV type 1 diversity in France, 1999 2001: Molecular characterization of non- B HIV type 1 subtypes and potential impact on susceptibility to antiretroviral drugs. AIDS Res Hum Retroviruses 2003;19:41 47. 11. Myers G: Tenth anniversary perspectives on AIDS. HIV: Between past and future. Aids Res Hum Retroviruses 1994;10:1317 1324. 12. Foley B, Pan H, Buchbinder S, and Delwart EL: Apparent founder effect during the early years of the San Francisco HIV type 1 epidemic (1978 1979). AIDS Res Hum Retroviruses 2000;16:1463 1469. 13. Gao F, Yue L, Hill SC, et al.: HIV-1 sequence subtype D in the United States. AIDS Res Hum Retroviruses 1994;10:625 627. 14. Brodine SK, Mascola JR, Weiss PJ, et al.: Detection of diverse HIV-1 genetic subtypes in the USA. Lancet 1995;346:1198 1199. 15. Sullivan PS, Do AN, Robbins K, et al.: Surveillance for variant strains of HIV: Subtype G and group O HIV-1. JAMA 1997;278: 292. 16. Sullivan PS, Do AN, Ellenberger D, et al.: Human immunodeficiency virus (HIV) subtype surveillance of African-born persons at risk for group O and group N HIV infections in the United States. J Infect Dis 2000;181:463 469. 17. Rayfield MA, Sullivan P, Bandea CI, et al.: HIV-1 group O virus identified for the first time in the United States. Emerg Infect Dis 1996;2:209 212. 18. de Oliveira CF, Diaz RS, Machado DM, et al.: Surveillance of HIV-1 genetic subtypes and diversity in the US blood supply. Transfusion 2000;40:1399 1406. 19. Womack C, Roth W, Newman C, et al.: Identification of non-b human immunodeficiency virus type 1 subtypes in rural Georgia. J Infect Dis 2001;183:138 142. 20. Artenstein AW, Ohl CA, VanCott TC, Hegerich PA, and Mascola

1070 JR: Transmission of HIV-1 subtype E in the United States [letter]. JAMA 1996;276:99 100. 21. Weidle PJ, Ganea CE, Irwin KL, et al.: Presence of human immunodeficiency virus (HIV) type 1, group M, non-b subtypes, Bronx, New York: A sentinel site for monitoring HIV genetic diversity in the United States. J Infect Dis 2000;181:470 475. 22. Irwin KL, Pau CP, Lupo D, et al.: Presence of human immunodeficiency virus (HIV) type 1 subtype A infection in a New York community with high HIV prevalence: A sentinel site for monitoring HIV genetic diversity in North America. Centers for Disease Control and Prevention-Bronx Lebanon HIV Serosurvey Team. J Infect Dis 1997;176:1629 1633. 23. Petersen LR and Doll LS: Human immunodeficiency virus type 1- infected blood donors: Epidemiologic, laboratory, and donation characteristics. The HIV Blood Donor Study Group. Transfusion 1991;31:698 703. 24. Delwart EL, Shaper EG, McCutchan FE, et al.: Genetic relationships determined by a heteroduplex mobility assay: Analysis of HIV env genes. Science 1993;262:1257 1261. 25. Heyndrickx L, Janssens W, Zekeng L, et al.: Simplified strategy for detection of recombinant human immunodeficiency virus type 1 group M isolates by gag/env heteroduplex mobility assay. Study Group on Heterogeneity of HIV Epidemics in African Cities. J Virol 2000;74:363 370. 26. Delwart EL, Herring BL, Rodrigo AG, and Mullins JI: Genetic subtyping of human immunodeficiency virus using a heteroduplex mobility assay. PCR Methods Appl 1995;4:S202 S216. 27. Felsenstein J: PHYLIP-Phylogency inference package. Cladistics 1989;5:164 166. 28. Pau CP, Lee-Thomas S, Auwanit W, et al.: Highly specific V3 peptide enzyme immunoassay for serotyping HIV-1 specimens from Thailand. AIDS 1993;7:337 340. 29. Lasky M, Perret JL, Peeters M, et al.: Presence of multiple non-b subtypes and divergent subtype B strains of HIV-1 in individuals infected after overseas deployment. AIDS 1997;11:43 51. 30. Gao F, Vidal N, Li Y, et al.: Evidence for two distinct sub-subtypes within the HIV-1 subtype A radiation. AIDS Res Hum Retroviruses 2001;17:675 688. 31. Xiao L, Owen SM, Goldman I, et al.: CCR5 coreceptor usuage of non-syncytium-inducing primary HIV-1 is independent of phylogenetically distinct global HIV-1 isolates: Delineation of consensus motif in the V3 domain that predicts CCR-5 usage. Virology 1998;240:83 92. 32. Nkengasong JN, Willems B, Janssens W, et al.: Lack of correlation between V3-loop peptide enzyme immunoassay serologic subtyping and genetic sequencing. AIDS 1998;12:1405 1412. 33. Plantier JC, Damond F, Lasky M, et al.: V3 serotyping of HIV-1 infection: Correlation with genotyping and limitations. J Acquir Immune Defic Syndr Hum Retrovirol 1999;20:432 441. DELWART ET AL. 34. Cheingsong-Popov R, Osmanov S, Pau CP, et al.: Serotyping of HIV type 1 infections: Definition, relationship to viral genetic subtypes, and assay evaluation. UNAIDS Network for HIV-1 Isolation and Characterization. AIDS Res Hum Retroviruses 1998;14: 311 318. 35. Kanki P, M Boup S, Marlink R, et al.: Prevalence and risk determinants of human immunodeficiency virus type 2 (HIV-2) and human immunodeficiency virus type 1 (HIV-1) in West African female prostitutes. Am J Epidemiol 1992;136:895 907. 36. Burgisser P, Simon F, Wernli M, Wust T, Beya MF, and Frei PC: Multicenter evaluation of new double-antigen sandwich enzyme immunoassay for measurement of anti-human immunodeficiency virus type 1 and type 2 antibodies. J Clin Microbiol 1996;34:634 637. 37. Machado DM, Delwart EL, Diaz RS, et al.: Use of the sensitive/less-sensitive (detuned) EIA strategy for targeting genetic analysis of HIV-1 to recently infected blood donors. AIDS 2002;16: 113 119. 38. Gonzales MJ, Machekano RN, and Shafer RW: Human immunodeficiency virus type 1 reverse-transcriptase and protease subtypes: Classification, amino acid mutation patterns, and prevalence in a northern California clinic-based population. J Infect Dis 2001;184: 998 1006. 39. Phillips S, Granade TC, Pau CP, Candal D, Hu DJ, and Parekh BS: Diagnosis of human immunodeficiency virus type 1 infection with different subtypes using rapid tests. Clin Diagn Lab Immunol 2000; 7:698 699. 40. Connor RI, Korber BT, Graham BS, et al.: Immunological and virological analyses of persons infected by human immunodeficiency virus type 1 while participating in trials of recombinant gp120 subunit vaccines. J Virol 1998;72:1552 1576. 41. Kleinman S, Busch MP, Hall L, et al.: False-positive HIV-1 test results in a low-risk screening setting of voluntary blood donation. Retrovirus Epidemiology Donor Study. JAMA 1998;280:1080 1085. 42. Hausler WJ Jr. and Getchell JP: Report of the Fourth Consensus Conference on HIV Testing sponsored by the Association of State and Territorial Public Health Laboratory Directors (ASTPHLD). Infect Control Hosp Epidemiol 1989;10:354 367. Address reprint requests to: Eric L. Delwart Blood Systems Research Institute 270 Masonic Avenue San Francisco, California 94118 E-mail: delwarte@medicine.ucsf.edu