Demographic Factors That Influence the Neutralizing Antibody Response in Recipients of Recombinant HIV-1 gp120 Vaccines

Transcription

1 MAJOR ARTICLE Demographic Factors That Influence the Neutralizing Antibody Response in Recipients of Recombinant HIV-1 gp120 Vaccines David C. Montefiori, 1 Barbara Metch, 2 M. Juliana McElrath, 3 Steve Self, 2 Kent J. Weinhold, 1 and Lawrence Corey, 3 for the HIV Vaccine Trials Network 1 Duke University Medical Center, Durham, North Carolina; 2 Statistical Center for HIV/AIDS Research and Prevention and 3 University of Washington, Seattle We compared neutralizing antibody responses in human immunodeficiency virus (HIV) type 1 gp120 vaccine recipients by age, sex, and race. Four phase 1 or 2 trials involving 505 vaccinated subjects were analyzed. Age and sex had no detectable effect on neutralizing antibody responses. However, race influenced the response to one vaccine, MN gp120, in alum. Four inoculations with this vaccine generated higher serum titers of neutralizing antibodies in African Americans than in whites. Despite potent neutralization of T cell line adapted HIV-1, serum from these African Americans failed to neutralize primary HIV-1 isolates. Neutralizing antibody responses did not differ between races when SF2 gp120 in MF-59 was administered either alone or with recombinant canarypox vcp205; they also did not differ when vcp1452 was administered either alone or with AIDSVAX B/B in alum. These data indicate that race may affect the neutralizing antibody response to some gp120 immunogens. To fully evaluate immunogenicity, clinical trials of candidate vaccines should enroll diverse populations of subjects. The envelope glycoproteins of HIV-1 have been incorporated into a variety of candidate AIDS vaccines that have been tested in clinical trials for the purpose of generating neutralizing antibodies. Some candidates are composed of gp120 or noncleaved gp160 formulated in adjuvants [1 8]. Other candidates express gp120 in vivo using recombinant vaccinia or canarypox vectors that are administered with or without subunit protein boosting [8 14]. Neutralizing antibodies have been detected in most cases and are produced at much higher levels with gp120 and gp160 proteins than with recombinant virus vectors as single modalities [15 17]. As dual modalities, recombinant virus vectors prime B cells for sec- Received 8 March 2004; accepted 15 June 2004; electronically published 28 October Presented in part: AIDS Vaccine 2003, New York, September (abstract 121). Financial support: National Institute of Allergy and Infectious Diseases (grants AI46705 and AI48017). Reprints or correspondence: Dr. David C. Montefiori, Dept. of Surgery, Laboratory for AIDS Vaccine Research and Development, PO Box 2926, Duke University Medical Center, Durham, NC (monte@acpub.duke.edu). The Journal of Infectious Diseases 2004; 190: by the Infectious Diseases Society of America. All rights reserved /2004/ $15.00 ondary neutralizing antibody production after subunit protein boosting [18]. Unfortunately, the specificity of neutralizing antibodies generated by candidate HIV-1 vaccines has been mostly limited to T cell line adapted strains of virus that closely resemble the vaccine strain [12, 19 21]. To date, no vaccine candidate has generated neutralizing antibodies with desirable potency and crossreactivity against primary HIV-1 isolates. In early 2003, the first phase 3 clinical trial in the United States and Europe of a candidate HIV-1 gp120 vaccine, AIDSVAX B/B (VaxGen), was completed. Overall, no effectiveness in the reduction of either the acquisition of infection or levels of plasma viremia after acquisition was noted. However, a subset analysis purported to show differences in the levels of neutralizing antibodies in serum between African Americans and US whites involved in the trial [22]. Moreover, the geometric mean titer (GMT) of neutralizing antibodies was reported to be higher in women than in men [22]. In this case, neutralizing antibody responses were measured against a T cell line adapted strain, HIV-1 MN, whose gp120 was used in the vaccine. The unexpected disparity in neutralizing antibody responses in this 1962 JID 2004:190 (1 December) Montefiori et al.

2 phase 3 trial prompted our retrospective analyses of neutralizing antibody responses in clinical trials conducted by the National Institutes of Health sponsored AIDS Vaccine Evaluation Group (AVEG) and HIV Vaccine Trials Network (HVTN). Results of these analyses are reported here. SUBJECTS, MATERIALS, AND METHODS Table 1. Trial designs and demographics. Study Product Dose Adjuvant Schedule, months Vaccine protocols. Study subjects were enrolled in 1 of 4 phase 1 or 2 clinical trials in the United States that were conducted by either AVEG or HVTN (table 1). Racial category in all trials was self-reported. Subjects in AVEG 201 (phase 2) received either SF2 gp120 (Chiron; 50 mg/dose) or MN gp120 (Genentech; 600 mg/dose) administered intramuscularly at 0, 1, and 6 months, with a fourth inoculation at either 12 or 18 months [23]. Subjects in AVEG 202 (phase 2) received ALVAC vcp205 at 0, 1, 3, and 6 months, with simultaneous inoculations with SF2 gp120 (50 mg/dose) [24]. Subjects in HVTN 203 (phase 2) received ALVAC vcp1452 at 0, 1, 3, and 6 months, with simultaneous inoculations with AIDSVAX B/B (containing equal amounts of MN gp120 and GNE8 gp120; 600 mg/dose) at either 3 and 6 months or 0, 1, and 6 months. Subjects in HVTN 039 (phase 1) received either a high or low dose of ALVAC vcp1452 at 0, 1, and 3 months. Gp120 proteins were formulated in either alum (MN gp120 and AIDSVAX B/B) or MF-59 (SF2 gp120). ALVAC vcp205 and vcp1452 (Aventis Pasteur) are recombinant canarypox vectors that express membrane-anchored MN gp120 [25]. ALVAC vcp205 also contains the entire gag and a portion of pol sufficient to encode protease (p15) from strain IIIB; this vector has been shown to produce virus-like particles in avian and mammalian cells in vitro [26]. ALVAC vcp1452 is a modified version of vcp205 that contains the same gagpol-env sequences in addition to a synthetic peptide with known cytotoxic T lymphocyte epitopes in Nef and Pol; vcp1452 also differs by the addition of vaccinia-derived coding sequences that improve RNA translation and expression of HIV-1 gene products. AIDSVAX B/B is an equal mixture of 2 clade B gp120 proteins (MN and GNE8). Informed consent was obtained from all study subjects, as approved by local institutional review boards and biosafety committees. The present study followed the human-experimentation guidelines of the US Department of Health and Human Services. Neutralizing antibody assays. For neutralizing antibody assays, blood was drawn from subjects before immunization and 2 weeks after boosting. Neutralization of HIV-1 MN, HIV- 1 SF2, HIV-1 W61D (T cell line adapted), and SHIV-89.6 was assessed in MT-2 cells as described elsewhere [27]. Neutralization titers are reported as the serum dilution at which 50% of cells were protected from virus-induced killing, as measured by neutral red uptake. This 50% cutoff corresponds to an 90% reduction in p24 Gag antigen synthesis [19]. Assay stocks of HIV- 1 MN and HIV-1 SF2 were generated in H9 cells, and assay stocks of HIV-1 W61D and SHIV-89.6 were generated in SupT-1 cells and human peripheral-blood mononuclear cells (PBMCs), respectively. Virus-containing culture supernatants were made cell free by filtration (0.45-micron pore size) and were stored at 80 C until use. Neutralization assays with HIV-1 MN and HIV-1 SF2 were performed blinded with respect to race, sex, and intraprotocol assignment group. Neutralization of primary HIV-1 isolates was measured in 5.25.EGFP.Luc.M7 (M7-Luc) cells (gift from N. R. Landau, the Salk Institute for Biological Studies, La Jolla, CA). M7-Luc is a genetically engineered clone of CEMx174 that expresses mul- White Race and sex, no. of women:no. of men African American AVEG 201 Arm 1 SF2 gp mg MF-59 0, 1, 6, and 12 or 18 28:69 2:21 0:1 Arm 2 MN gp mg Alum 0, 1, 6, and 12 or 18 25:65 13:17 0:1 AVEG 202, arm 1 vcp TCID 50 None 0, 1, 3, and 6 11:43 15:13 2:6 SF2 gp mg MF-59 0, 1, 3, and 6 HVTN 203 Arm 1 vcp TCID 50 None 0, 1, 3, and 6 12:19 4:5 0:2 AIDSVAX B/B 600 mg Alum 3 and 6 Arm 2 vcp TCID 50 None 0, 1, 3, and 6 7:32 0:3 0:1 AIDSVAX B/B 600 mg Alum 0, 1, and 6 HVTN 039 Arm 1 vcp TCID 50 None 0, 1, and 3 8:16 3:6 1:4 Arm 2 vcp TCID 50 None 0, 1, and 3 13:26 4:5 0:2 Other a NOTE. AVEG, AIDS Vaccine Evaluation Group; HVTN, HIV Vaccine Trials Network. a Includes Hispanic, Asian, Native American, and multiracial individuals incorporated into the age and sex analyses. Demographics of gp120 Immunogenicity JID 2004:190 (1 December) 1963

3 tiple entry receptors (CD4, CXCR4, and GPR15/Bob) and is transduced to express CCR5 [28]; it also possesses Tat-responsive reporter genes for luciferase and green fluorescent protein (GFP). Cells were maintained in growth medium (RPMI 1640, 12% heat-inactivated fetal bovine serum, and 50 mg/ml gentamicin) containing puromycin (0.5 mg/ml), G418 (300 mg/ ml), and hygromycin (200 mg/ml), to preserve CCR5 and reporter-gene plasmids. For neutralizing antibody assays, cellfree virus (5000 TCID 50 ) was incubated in triplicate with a 1:15 dilution of serum samples in 96-well flat-bottom culture plates (Costar) for 1 h at 37 C. Next, 100 ml of cells ( cells/ ml of growth medium containing 25 mg/ml DEAE dextran but lacking puromycin, G418, and hygromycin) was added to each well. One set of wells received cells plus virus (virus control), and another set of wells received cells only (background control). Plates were incubated until 10% of cells in the virus control wells were positive for GFP expression by fluorescence microscopy ( 3 days). At this time, 100 ml of cell suspension was transferred to a 96-well white solid plate (Costar) for measurements of luminescence by use of Bright Glo substrate solution (Promega), in accordance with the manufacturer s instructions. Neutralization was considered positive when relative luminescence units (RLUs) were reduced by either 50% or 80%, compared with the RLUs measured in a 1:15 dilution of the corresponding preimmune serum sample. Cell-free stocks of primary isolates were generated in PBMCs as described elsewhere [29]. ELISAs. Binding antibodies against MN gp120, SF2 gp120, and peptides corresponding to the V3 loop of MN (CYNKRK- RIHIGPGRAFYTTKNIIG) and SF2 (CNNTRKSIYIGPGRAF- HTTGRIIG) were assessed by ELISA, as described elsewhere [23]. Statistical analyses. Sample sizes for data analysis are shown in table 1. Differences in binding and neutralizing antibody responses were compared between (1) men and women, (2) African Americans and whites, and (3) African Americans and whites within each sex. Neutralizing antibody responses were further analyzed for differences between 2 age groups, and years. Comparisons were made within each subgroup according to virus strain (neutralization), antigen (ELISA), vaccine, and number of inoculations. Differences in the distribution of ELISA values and titers of neutralizing antibody were compared by use of the Wilcoxon rank sum test. An exact Wilcoxon test was used for comparisons involving groups with a sample size 10. Comparisons involving groups with a sample size 5 were not included. Because many tests were done for each demographic comparison, the probability of falsely rejecting the null hypothesis of no difference was increased. To control the overall probability of reaching a false conclusion regarding gp120 vaccines when all tests for a demographic comparison were considered, the nominal Wilcoxon P values were adjusted by use of the false-discovery-rate method [30]. This method limits the expected proportion of falsely rejected null hypotheses to 5%. Differences in positive neutralization response rates (percentage of subjects with positive neutralizing activity) were compared by use of Fisher s exact test. For comparisons of titers of neutralizing antibodies, a titer value of 5 (one-half the lowest dilution tested) was assigned to nonresponders. RESULTS Neutralizing antibody responses after immunization with gp120 alone. In AVEG 201, neutralizing antibody responses against HIV-1 MN and HIV-1 SF2 were assessed after the third and fourth inoculations in both arms of the trial. To increase the power of the statistical analyses, subjects inoculated at 0, 1, 6, and 12 months and at 0, 1, 6, and 18 months were combined into a single group in each arm of the trial. Arm 2 (MN gp120) used the same dose and adjuvant as did the VaxGen phase 3 trial of AIDSVAX B/B, but the vaccine contained only 1 of the 2 gp120 components in bivalent AIDSVAX B/B. In this arm, sex alone had no influence on neutralizing antibody responses after 3 and 4 inoculations (table 2). Race had a moderate effect on neutralizing antibody responses after 3 inoculations and a stronger effect after 4 inoculations. Specifically, neutralizing antibody responses after 3 inoculations were weak overall but were significantly higher against HIV-1 SF2 in African Americans than in whites (GMT, 40 and 21; adjusted P p.02). Neutralizing antibody responses increased in all groups after 4 inoculations but were significantly stronger against both HIV-1 MN and HIV-1 SF2 in African American men and women (table 2). Overall, the GMT of neutralizing antibodies after 4 inoculations was times higher in African Americans than in whites. Serum samples obtained 2 weeks after final MN gp120 boosting (peak immunity) in 10 African Americans were examined for neutralizing antibody response against 10 heterologous clade B primary HIV-1 isolates. These primary isolates all possessed an R5 phenotype, were genetically and geographically diverse, and represented a spectrum of neutralization sensitivities ranging from above average (strains Bal, JR-FL, Bx08, and 1196) to average (strains 6101, 5768, 0515, 0692, 1168, and Pvo) when tested with serum samples from infected individuals (D.C.M., unpublished data) [19]. Of the 100 virus/serum combinations tested, 16 were positive for 50% neutralization, but none were positive for 80% neutralization. These same serum samples possessed potent neutralizing antibody activity against 3 T cell line adapted strains of HIV-1 (GMT against HIV-1 MN, 588 [range, ]; GMT against HIV-1 SF2, 421 [range, ]; GMT against HIV-1 W61D, 770 [range, ]) but did not neutralize SHIV Because assays with primary isolates were performed on serum samples that had been stored frozen for 9 years, antibody stability was examined in repeat assays with HIV-1 MN and showed no loss of neutralizing activity JID 2004:190 (1 December) Montefiori et al.

4 Table 2. Geometric mean titers (GMTs) of neutralizing antibodies against HIV-1 MN and HIV-1 SF2 after 4 inoculations with MN gp120 in alum, by groups based on race and sex (AIDS Vaccine Evaluation Group 201). Group No. of subjects HIV-1 MN GMT (95% CI) P Nominal Wilcoxon FDRM adjusted HIV-1 SF2 No. of subjects GMT (95% CI) P Nominal Wilcoxon FDRM adjusted Race African American ( )! ( )! White ( ) (48 92) Sex Women ( ) (76 238) Men ( ) (55 107) Race/sex subgroup Women African American ( ) ( ) White ( ) (44 165) Men African American ( ) ( ) White ( ) (42 88) NOTE. CI, confidence interval; FDRM, false discovery rate method. No significant differences in titers of neutralizing antibody against HIV-1 MN and HIV-1 SF2 were observed between either race or sex after 3 and 4 inoculations with SF2 gp120 (data not shown). The number of African American women who received this vaccine was too small for race/sex subgroup analysis. Neutralizing antibody responses after immunization with recombinant ALVAC plus gp120. In AVEG 202, neutralizing antibody responses against HIV-1 MN were assessed after the third and fourth inoculations, and neutralizing antibody responses against HIV-1 SF2 were assessed after the fourth inoculation only. GMTs against HIV-1 MN were 24 (95% confidence interval [CI], 18 32) and 48 (95% CI, 38 60) after the third and fourth inoculations, respectively. The GMT against HIV- 1 SF2 was 52 (95% CI, 41 66) after the fourth inoculation. No significant differences between races, sexes, and race/sex subgroups were observed. In HVTN 203, neutralizing antibody responses against HIV- 1 MN were assessed after final boosting. GMTs were 143 (95% CI, ) and 324 (95% CI, ) in the groups receiving 2 and 3 inoculations of AIDSVAX B/B, respectively. Responses were compared between whites and African Americans in arm 1 and between men and women in arms 1 and 2. No significant differences were observed. The number of African Americans enrolled in this trial was too small to make comparisons between races in arm 2 and between races within each sex in arm 1. Neutralizing antibody responses after immunization with recombinant ALVAC alone. In HVTN 039, neutralizing antibody responses against HIV-1 MN were assessed 2 weeks after final boosting. GMTs in arm 2 (high dose) were 110 and 57 for African Americans and whites, respectively (nominal P p.15). GMTs in arm 1 (low dose) were 68 and 41 for African Americans and whites, respectively (nominal P p.49). The number of African Americans enrolled in this trial was too small ( n p 9 for each dose) to make comparisons between races within each sex. Positive neutralization response rates. Positive neutralization response rates, defined as the percentage of vaccine recipients whose antibody titer was 10 in antibody neutralization assays, are shown in table 3. In HVTN 039, rates were 92% for both arms. Rates were lowest after 3 inoculations with SF2 gp120 in AVEG 201 (69% for HIV-1 MN and 75% for HIV- 1 SF2 ) and AVEG 202 (74% for HIV-1 SF2 ) but were 180% in AVEG 201 and 190% in AVEG 202 after the fourth inoculation. In HVTN 203, rates were 90% (HIV-1 MN ) after 2 and 3 inoculations. Differences between the same groups and subgroups described above for titers of neutralizing antibodies were compared. In AVEG 201, men receiving SF2 gp120 had a higher rate against HIV-1 SF2 than did women after the fourth inoculation (nominal P p.01). After adjustment for multiple comparisons, no significant differences were observed in any trial. Age-related differences in neutralizing antibody responses. In each study, titers of neutralizing antibodies were compared between 2 age groups that were separated on the basis of the median age of 35 years. When data for race and sex were combined into single sets, no significant differences between vaccine recipients and years old were observed (data not shown). Thus, age was not an influential factor as an independent variable. The size of AVEG 201 allowed a refined analysis between African Americans and whites as dependent variables within each age group. No significant differences were observed in arm 1 (SF2 gp120) of the trial. However, in arm 2 (MN gp120), significant differences were Demographics of gp120 Immunogenicity JID 2004:190 (1 December) 1965

5 Table 3. Positive neutralization response rates. No. positive/total (% positive [95% CI]) a Study, vaccine HIV-1 MN HIV-1 SF2 AVEG MN gp /121 (98 [94 100]) 113/121 (93 [87 97]) 4 MN gp120 91/91 (100) 84/91 (92 [85 97]) 3 SF2 gp120 83/121 (69 [60 77]) 91/121 (75 [67 83]) 4 SF2 gp120 74/87 (85 [76 92]) 71/87 (82 [72 89]) AVEG 202 vcp SF2 gp120 66/89 (74 [64 83]) NA vcp SF2 gp120 83/89 (93 [86 97]) 83/89 (93 [86 97]) HVTN 203 vcp AIDSVAX B/B at 3 and 6 months 38/42 (90 [77 97]) NA vcp AIDSVAX B/B at 0, 1, and 6 months 43/43 (100 [92 100]) NA HVTN 039 vcp1452, low dose 35/38 (92 [81 98]) NA vcp1452, high dose 46/50 (92 [79 98]) NA NOTE. AVEG, AIDS Vaccine Evaluation Group; HVTN, HIV Vaccine Trials Network; NA, not assayed. a No. of samples positive for neutralization in MT-2 cells (titer 10). observed in both age groups. In arm 2, after 4 inoculations, titers of neutralizing antibodies were significantly higher in African Americans than in whites when measured against both HIV-1 MN (18 35 years: GMT, 426 for African Americans and 180 for whites [adjusted P p.03]; years: GMT, 714 for African Americans and 218 for whites [adjusted P p.02]) and HIV-1 SF2 (18 35 years: GMT, 280 for African Americans and 71 for whites [adjusted P p.009]; years: GMT, 333 for African Americans and 62 for whites [adjusted P p.03]). After 3 inoculations, significant differences between African Americans and whites years old were observed when neutralization was measured against HIV-1 SF2 (GMT, 46 for African Americans and 20 for whites [adjusted P p.006]). No significant age-related differences in positive neutralization response rates were observed. Binding antibodies. Serum samples from all vaccine recipients in AVEG 201 were tested for the presence of binding antibodies before immunization and 2 weeks after each inoculation. The majority of vaccine recipients tested positive against all test antigens after 3 and 4 inoculations [23]. No significant differences between races, sexes, and race/sex subgroups were observed (data not shown). DISCUSSION We have analyzed neutralizing antibody responses by age, race, and sex in clinical trials of candidate HIV-1 vaccines that enrolled sufficient numbers of HIV-1 negative women and minority subjects to make statistical comparisons possible. Our analyses indicated that age and sex had no detectable effect on neutralizing antibody responses to multiple gp120 vaccines. However, we did observe significant differences in the magnitude of neutralizing antibody responses between African Americans and whites, and we discovered that these differences were dependent on vaccine type. Thus, among subjects who received monomeric MN gp120 in alum, African Americans were more likely to generate high titers of neutralizing antibodies than were whites. A similar vaccine, AIDSVAX B/B, given with ALVAC produced no discernible differences in neutralizing antibody responses between races. Moreover, SF2 gp120, when given in MF-59 and at a lower dose, also did not produce discernible differences in neutralizing antibody responses between races. At present, it is not known what mechanisms account for the disparate neutralizing antibody responses between races, and the consequences for future HIV-1 vaccine candidates are also not apparent. The disparate responses could reflect elevated serum levels of total IgG and, to a lesser extent, IgA and IgM in African Americans than in whites [31 34], as is governed genetically [35 37]. Little is known about the effect of race on other vaccines. Most vaccine studies fail to include race in their reporting and data analyses [38], and those that do are often complicated by such confounding variables as age, nutritional status, environmental factors, previous exposure to microorganisms, and the absence of a valid classification system for race [39, 40]. Interestingly, after vaccination with pertussis vaccines, 2-fold higher antibody levels were reported in African American infants, compared with those in white infants [41]. This difference reached statistical significance for multiple antibody specificities and could not be discounted on the basis of sex, parental education, study site, or preimmunization antibody levels. Notably, these racial differences in antibody levels were much less significant for whole-cell vaccines than for acellular pertussis vaccines. In one study, elevated levels of neutralizing antibodies were 1966 JID 2004:190 (1 December) Montefiori et al.

6 reported in whites versus African Americans and in men versus women after yellow fever vaccination [42]. Also, after a single dose of Haemophilus influenzae type b polysaccharide mutant diphtheria toxin vaccine, 2-year-old Apache children (despite having higher levels of total IgG, IgM, and IgA) had lower levels of vaccine-specific antibodies before vaccination and a 5-fold diminished response after vaccination than did white children of a similar age [43]. Impaired responses in these Apache children were associated with abnormally low levels of total IgG2 and IgG4, which did not affect the response to protein toxoids. In another study, antibody responses were significantly lower in whites and African Americans than in Asians and Hispanics after 3 doses of H. influenza type b polysaccharide tetanus toxoid vaccine [44]. Also, in Australia, lower antibody responses were observed in Aboriginal infants than in white infants after 3 doses of H. influenzae type b polysaccharide Neisseria meningitidis outer-membrane protein conjugate vaccine [45]. Finally, in northern Newfoundland, higher seropositive rates for measles and higher mean antibody levels were observed in Innu and Inuit schoolchildren than in white schoolchildren among those who received a single dose of measles-mumps-rubella vaccine [46]. Several genetic factors might influence the antibody response to vaccines [47]. For example, homozygosity in certain HLA class II and TAP2 alleles is strongly associated with weak antibody responses to vaccines for hepatitis B [48 50] and measles [51 54]. Also, the Km(1) and G2m immunoglobulin allotypes are significant markers of the antibody response to bacterial capsular polysaccharide vaccines, including H. influenzae type b, pneumococcus, meningococcus, and N. meningitidis [55 60]. In one case, conjugation of polyribosylribitol phosphate with a carrier protein (outer-membrane protein conjugate) reversed the association of the Km(1)-negative allotype with low antibody responsiveness to H. influenzae type b vaccine in African American children [61]. In another case, the Km(1)- positive allotype was associated with a decreased risk of Haemophilus meningitis in African Americans but not in whites [62]. Interestingly, the Km(1)-positive allotype has been associated with higher total IgG2 levels in African Americans, compared with those in whites [37]; this finding might, in part, explain the altered immune response of African Americans to certain polysaccharide vaccines. Linkages among race, genetics, and immunogenicity with respect to non HIV-1 vaccines appear to be sensitive to minor differences in vaccine composition and delivery [41, 61]. That we did not always see an effect of race on gp120 immunogenicity suggests that the same may be true for HIV-1 vaccines. Possible influential variables include vaccine strain, dose, adjuvant, schedule, and whether gp120 is administered in combination with a recombinant vector, such as ALVAC. It must be emphasized that the stronger neutralizing antibody response observed here in African Americans did not translate into an improved ability to neutralize primary isolates. The neutralizing antibody response generated by most HIV-1 Env vaccines targets linear epitopes in the V3 loop of gp120 that are readily exposed on T cell line adapted strains but are poorly exposed on primary isolates [12, 19 21]. In this regard, the racial effect observed here could not be attributed to differences in the titers of antibodies that bind either V3 peptides or gp120. The neutralization assay might allow a finer discrimination of quantitative differences in antibodies directed against the V3 loop. Alternatively, the antibodies might have differed in either specificity or affinity, depending on race. Regardless, the inability of antibodies from African American vaccine recipients to neutralize primary isolates makes it highly unlikely that this response would protect against a virus that is as genetically variable and structurally complex as HIV-1 [63, 64]. Identification of whether broadly reactive neutralizing antibody production would be influenced by race or sex requires the creation of a vaccine that is capable of generating such antibodies. It is uncertain how our findings relate to the general population of both races. Heterogeneity within races, especially selfreported racial designations, can be expected and might, in part, explain the wide distribution of neutralizing antibody responses within subgroups. Future studies of specific alleles associated with high and low antibody responsiveness within and between races and sexes seem warranted. Specifically, in AVEG 201, low responsiveness after 3 inoculations with either MN gp120 or SF2 gp120 was predictive of low responsiveness after the fourth and final inoculation, regardless of race (data not shown). Studies that use different immunogens, formulations, and modes of vaccine delivery in diverse genetic backgrounds, including low responders, might reveal strategies for the generation of optimal responses in a majority of individuals. Finally, our data indicate that phase 2 clinical trials of candidate HIV-1 vaccines should evaluate a wide variety of racial groups and include enough men and women to define the issues outlined above. Acknowledgments We gratefully acknowledge the individuals who participated in these clinical trials. We also thank José Esparza, Mary Allen, and George Siber, for helpful discussions. References 1. Dolin R, Graham BS, Greenberg SB, et al. The safety and immunogenicity of a human immunodeficiency virus type 1 (HIV-1) recombinant gp160 candidate vaccine in humans. Ann Intern Med 1991; 114: Keefer MC, Graham BS, Belshe RB, et al. Studies of high doses of a human immunodeficiency virus type 1 recombinant glycoprotein 160 candidate vaccine in HIV type 1 seronegative humans. AIDS Res Hum Retroviruses 1994; 10: Demographics of gp120 Immunogenicity JID 2004:190 (1 December) 1967

7 3. Belshe RB, Clements ML, Dolin R, et al. Safety and immunogenicity of a fully glycosylated recombinant gp160 human immunodeficiency virus type 1 vaccine in subjects at low risk of infection. J Infect Dis 1993; 168: Schwartz DH, Gorse G, Clements ML, et al. Induction of HIV-1 neutralising and syncytium-inhibiting antibodies in uninfected recipients of HIV-1 IIIB rgp120 subunit vaccine. Lancet 1993; 342: Belshe RB, Graham BS, Keefer MC, et al. Neutralizing antibodies to HIV-1 in seronegative volunteers immunized with recombinant gp120 from the MN strain of HIV-1. JAMA 1994; 272: Gorse GJ, McElrath MJ, Matthews TJ, et al. Modulation of immunologic responses to HIV-1 MN recombinant gp160 vaccine by dose and schedule of administration. Vaccine 1998; 16: Gorse GJ, Corey L, Patel GB, et al. HIV-1 MN recombinant glycoprotein 160 vaccine induced cellular and humoral immunity boosted by HIV- 1 MN recombinant glycoprotein 120 vaccine. AIDS Res Hum Retroviruses 1999; 15: Evans TG, McElrath MJ, Matthews T, et al. QS-21 promotes a dosesparing adjuvant effect during HIV-1 subunit immunization. Vaccine 2001; 19: Graham BS, Gorse GJ, Schwartz DH, et al. Determinants of antibody response after recombinant gp160 boosting in vaccinia-naivevolunteers primed with gp160-recombinant vaccinia virus. J Infect Dis 1994; 170: Corey L, McElrath MJ, Weinhold K, et al. Cytotoxic T cell and neutralizing antibody responses to human immunodeficiency virus type 1 envelope with a combination vaccine regimen. J Infect Dis 1998; 177: Clements-Mann ML, Weinhold K, Matthews TJ, et al. Immune responses to human immunodeficiency virus (HIV) type 1 induced by canarypox expressing HIV-1 MN gp120, HIV-1 SF2 recombinant gp120, or both vaccines in seronegative adults. J Infect Dis 1998; 177: Belshe RB, Gorse GJ, Mulligan MJ, et al. Induction of immune responses to HIV-1 by canarypox virus (ALVAC) HIV-1 and gp120 SF-2 recombinant vaccines in uninfected volunteers. AIDS 1998; 12: Evans TG, Keefer MC, Weinhold KJ, et al. A canarypox vaccine expressing multiple human immunodeficiency virus type 1 genes given alone or with rgp120 elicits broad and durable CD8 + cytotoxic T lymphocyte responses in seronegative volunteers. J Infect Dis 1999; 180: Corey L, Mulligan M, Goepfert P, et al. Cellular and humoral immune responses to a canarypox vaccine containing HIV-1 Env, Gag, and Pol in combination with rgp120. J Infect Dis 2001; 183: Graham BS, Matthews TJ, Belshe RB, et al. Augmentation of human immunodeficiency virus type 1 neutralizing antibody by priming with gp160 recombinant vaccinia and boosting with rgp160 in vaccinianaive adults. J Infect Dis 1993; 167: Cooney EL, McElrath MJ, Corey L, et al. Enhanced immunity to human immunodeficiency virus (HIV) envelope elicited by a combined vaccine regimen consisting of priming with a vaccinia recombinant expressing HIV envelope and boosting with gp160 protein. Proc Natl Acad Sci USA 1993; 90: Excler J-L, Plotkin S. The prime-boost concept applied to HIV preventive vaccines. AIDS 1997; 11(Suppl A):S Montefiori DC. HIV-1 specific neutralizing antibodies. In: Pantaleo G, Walker B, eds. Retroviral immunology. Totowa, NJ: Humana Press, 2001: Bures R, Gaitan A, Tuofu Z, et al. Immunization with recombinant canarypox vectors expressing membrane-anchored gp120 followed by gp160 protein boosting fails to generate antibodies that neutralize R5 primary isolates of human immunodeficiency virus type 1. AIDS Res Hum Retroviruses 2000; 16: Beddows S, Lister S, Cheingsong R, Bruck C, Weber J. Comparison of the antibody repertoire generated in healthy volunteers following immunization with a monomeric recombinant gp120 construct derived from a CCR5/CXCR4-using human immunodeficiency virus type 1 isolate with sera from naturally infected individuals. J Virol 1999; 73: Mascola JR, Snyder SW, Weislow OS, et al. Immunization with envelope subunit vaccine products elicits neutralizing antibodies against laboratory-adapted but not primary isolates of human immunodeficiency virus type 1. J Infect Dis 1996; 173: Gilbert P, Popovic V, Gurwith M, et al. Immunologic responses to rgp120 vaccine and correlation with the risk of HIV infection: results from the world s first HIV vaccine efficacy trial [abstract 147]. In: Program and abstracts for AIDS Vaccine 2003 (New York). Alexandria, VA: Foundation for AIDS Vaccine Research and Development, 2003: McElrath MJ, Corey L, Montefiori D, et al. A phase II study of two HIV-1 envelope vaccines comparing their immunogenicity in populations at risk for acquiring HIV-1 infection. AIDS Res Hum Retroviruses 2000; 16: Belshe RB, Stevens C, Gorse GJ, et al. Safety and immunogenicity of a canarypox-vectored human immunodeficiency virus type 1 vaccine with or without gp120: a phase 2 study in higher- and lower-risk volunteers. J Infect Dis 2001; 183: Ferrari G, Berend C, Ottinger J, et al. Replication-defective canarypox (ALVAC) vectors effectively activate anti human immunodeficiency virus 1 cytotoxic T lymphocytes present in infected patients: implications for antigen-specific immunotherapy. Blood 1997; 90: Fang Z-Y, Kuli-Zade I, Spearman P. Efficient human immunodeficiency virus (HIV) 1 Gag-Env pseudovirion formation elicited from mammalian cells by a canarypox HIV vaccine candidate. J Infect Dis 1999; 180: Montefiori DC, Robinson WE Jr, Schuffman SS, Mitchell WM. Evaluation of antiviral drugs and neutralizing antibodies to human immunodeficiency virus by a rapid and sensitive microtiter infection assay. J Clin Microbiol 1988; 26: Brandt SM, Marian R, Holland AU, Hope TJ, Landau NR. Association of chemokine-mediated block to HIV entry with coreceptor internalization. J Biol Chem 2002; 277: Montefiori DC, Pantaleo G, Fink LM, et al. Neutralizing and infectionenhancing antibody responses to human immunodeficiency virus type 1 in long-term nonprogressors. J Infect Dis 1996; 173: Benjamini Y, Hochberg Y. Controlling the false discovery rate: a new and powerful approach to multiple testing. J R Stat Soc [Ser B] 1995;57: Buckley RH, Dees SC, O Fallon WM. Serum immunoglobulins. I. Levels in normal children and in uncomplicated childhood allergy. Pediatrics 1968; 41: Buckley CE III, Dorsey FC. The effect of aging on human serum immunoglobulin concentrations. J Immunol 1970; 105: Maddison SE, Stewart CC, Farshy CE, Reimer CB. The relationship of race, sex, and age to concentrations of serum immunoglobulins expressed in international units in healthy adults in the USA. Bull World Health Organ 1975; 52: Lucey DR, Hendrix CW, Andrzejewski C, et al. Comparison by race of total serum IgG, IgA, and IgM with CD4 + T-cell counts in North American persons infected with the human immunodeficiency virus type 1. J Acquir Immune Defic Syndr 1992; 5: Allansmith M, McClellan B, Butterworth M. The influence of heredity and environment on human immunoglobulin levels. J Immunol 1969; 102: Grundbacher FJ. Heritability estimates and genetic and environmental correlations for the human immunoglobulins G, M, and A. Am J Hum Genet 1974; 26: Shackelford PG, Granoff DM, Nahm MH, et al. Relation of age, race and allotype to immunoglobulin subclass concentrations. Pediatr Res 1985; 19: Poirier MK, Poland GA, Jacobson RM. Parameters potentially affecting interpretations of immunogenicity and efficacy data in vaccine trials: are they adequately reported? Vaccine 1996; 14: Centers for Disease Control and Prevention. Use of race and ethnicity in public health surveillance: summary of the CDC/ATSDR workshop. MMWR Morb Mortal Wkly Rep 1993; 41(RR-10): JID 2004:190 (1 December) Montefiori et al.

8 40. Haga SB, Venter JC. FDA races in wrong direction. Science 2003; 301: Christy C, Pichichero ME, Reed GF, et al. Effect of gender, race, and parental education on immunogenicity and reported reactogenicity of acellular and whole-cell pertussis vaccines. Pediatrics 1995; 96: Monath TP, Nichols R, Archambault WT, et al. Comparative safety and immunogenicity of two yellow fever 17D vaccines (ARILVAX and YF-VAX) in a phase III multicenter, double-blind clinical trial. Am J Trop Med Hyg 2002; 66: Siber GR, Santosham M, Reid GR, et al. Impaired antibody responses to Haemophilus influenza type b polysaccharide and low IgG2 and IgG4 concentrations in Apache children. N Engl J Med 1990; 323: Greenberg DP, Vadheim CM, Partridge S, et al. Immunogenicity of Haemophilus influenza type b tetanus toxoid conjugate vaccine in young infants. J Infect Dis 1994; 170: Guthridge S, McIntyre P, Isaacs D, Hanlon M, Patel M. Differing serologic responses to an Haemophilus influenza type b polysaccharide Neisseria meningitidis outer membrane protein conjugate (PRP-OMPC) vaccine in Australian Aboriginal and Caucasian infants: implications for disease epidemiology. Vaccine 2000; 18: Poland GA, Jacobson RM, Colbourne SA, et al. Measles antibody seroprevalence rates among immunized Inuit, Innu and Caucasian subjects. Vaccine 1999; 17: Poland GA, Jacobson RM. The genetic basis for variation in antibody responses to vaccines. Curr Opin Pediatr 1998; 10: Alpers CA, Kruskall MS, Marcus-Bagley D, et al. Genetic prediction of nonresponse to hepatitis B vaccine. N Engl J Med 1989; 321: Kruskall MS, Alper CA, Awdeh Z, Yunis EJ, Marcus-Bagley D. The immune response to hepatitis B vaccine in humans: inheritance patterns in families. J Exp Med 1992; 175: Mineta M, Tanimura M, Tana T, Yssel H, Kashiwagi S, Sasazuki T. Contribution of HLA class I and II alleles to the regulation of antibody production to hepatitis B surface antigen in humans. Int Immunol 1996; 8: Hayney MS, Poland GA, Jacobson RM, Schaid DJ, Lipsky JJ. The influence of the HLA-DRB1*13 allele on measles vaccine response. J Investig Med 1996; 44: Hayney MS, Poland GA, Dimanlig P, Schaid DJ, Jacobson RM, Lipsky JJ. Polymorphisms of the TAP2 gene may influence antibody responses to live measles vaccine virus. Vaccine 1997; 15: Tan P-L, Jacobson RM, Poland GA, Jacobsen SJ, Pankratz VS. Twin studies of immunogenicity: determining the genetic contribution to vaccine failure. Vaccine 2001; 19: Poland GA, Ovsyannikova IG, Jacobson RM, et al. Identification of an association between HLA class II alleles and low antibody levels after measles immunization. Vaccine 2001; 20: Pandey JP, Virella G, Loadholt CB, et al. Association between immunoglobulin allotypes and immune responses to Haemophilus influenzae and meningococcus polysaccharides. Lancet 1979; 1: Ambrosino DM, Barrus VA, DeLange GG, Siber GR. Correlation of the Km(1) immunoglobulin allotype with anti-polysaccharide antibodies in Caucasian adults. J Clin Invest 1986; 78: Ambrosino DM, Schiffman G, Gotschlich EC, et al. Correlation between G2m(n) immunoglobulin allotype and human antibody responses and susceptibility to polysaccharide encapsulated bacteria. J Clin Invest 1985; 75: Rodriguez-Barradas MC, Groover JE, Lacke CE, et al. IgG antibody to pneumococcal capsular polysaccharide in human immunodeficiency virus infected subjects: persistence of antibody in responders, revaccination in nonresponders, and relationship of immunoglobulin allotype to response. J Infect Dis 1996; 173: Sarvas H, Rautonen N, Sipinen S, Mäkelä O. IgG subclasses of pneumococcal antibodies: effect of allotype G2m(n). Scand J Immunol 1989; 29: Musher DM, Groover JE, Watson DA, et al. Genetic regulation of the capacity to make immunoglobulin G to pneumococcal capsular polysaccharides. J Investig Med 1997; 45: Weinberg GA, Granoff DM. Immunogenicity of Haemophilus influenzae type b polysaccharide-protein conjugate vaccines in children with conditions associated with impaired antibody responses to type b polysaccharide vaccine. Pediatrics 1990; 85: Granoff DM, Pandey JP, Boies E, Squires J, Munson RS Jr, Suarez B. Responses to immunization with Haemophilus influenzae type b polysaccharide-pertussis vaccine and risk of Haemophilus meningitis in children with the Km(1) immunoglobulin allotype. J Clin Invest 1984; 74: McCutchan FE, Salminen MO, Carr JK, Burke DS. HIV genetic diversity. AIDS 1996; 10(Suppl 3):S Wyatt R, Sodroski J. The HIV-1 envelope glycoproteins: fusogens, antigens, and immunogens. Science 1998; 280: Demographics of gp120 Immunogenicity JID 2004:190 (1 December) 1969