Vaccines and Adjuvant

ALVAC vCP205The recombinant canarypox virus, ALVAC-HIV vCP205 (Aventis Pasteur), expressed the products of several HIV-1 genes, including the p55 polyprotein expressed by the gag gene of HIV-1 LAI strain, a portion of the pol gene sufficient to express protease activity from HIV-1 LAI strain, gp120 expressed by a part of the env gene of HIV-1 MN strain, and the anchoring transmembrane region of gp41 of HIV-1 LAI strain. The gag gene, which codes for virus core antigens, is a relatively well-conserved gene. The gag gene encodes a polyprotein (55 kDa) that matures into the core proteins p24, p17, and p15, the last of which is then split into p9 and p6. This maturation process is catalyzed by the HIV protease encoded by a part of the pol gene and is the reason for the inclusion of the open-reading frame of the protease gene in the construction of ALVAC-HIV vCP205. Genotypic and phenotypic analyses of the ALVAC-HIV vCP205 recombinant were confirmed by nucleotide sequence and restriction analyses, as well as by Western blot and immune precipitation analyses. Derivation of the vector has been described elsewhere for similar vaccines [13, 14]

The recombinant vaccine virus was grown on specific pathogen–free chick embryo fibroblasts, and the vaccine was suspended in a solution of serum-free, antibiotic-free culture medium, containing virus stabilizers, and was lyophilized. Each 0.5 mL dose of the vaccine contained 10^6.0 TCID₅₀ of virus

HIV-1 SF-2 rgp120Subunit vaccine was an intact glycosylated form of HIV-1 SF-2 rgp120 (Chiron Corporation) and was given as 0.5 mL doses containing 50 μg of antigen in MF59 adjuvant. The immunogen was derived through recombinant technology in CHO cells under the control of the cytomegalovirus immediate early–1 promoter. It was purified by ion exchange, hydrophobic interaction, and gel chromatography, as described elsewhere [15]. The immunogen was formulated in a vehicle composed of buffered saline

AdjuvantThe adjuvant for the HIV-1 SF-2 rgp120 (MF59; Chiron Corporation) consisted of 0.5% polysorbate 80 (Tween 80 [Sigma] and polyoxyethylene sorbitan mono-oleate) and 0.5% sorbitan trioleate in a citrate buffer [15]. Squalene (5%), a metabolizable lipid, constituted the oil phase. Emulsification by high-pressure homogenization resulted in a physically stable emulsion with a mean droplet size of <300 nm

PlacebosThe placebo ALVAC (Aventis-Pasteur) was a mixture of 10 mM Tris-HCI buffer at pH 9.0, virus stabilizer, and freeze-drying medium. Saline was used as a placebo control for HIV-1 SF-2 rgp120

Volunteers

This multicenter, double-blind, randomized phase 2 safety and immunogenicity trial was conducted at 6 AIDS Vaccine Evaluation Group (AVEG) and 8 HIV Network for Prevention Trials (HIVNET) sites sponsored by the National Institute of Allergy and Infectious Diseases. Healthy HIV-1–uninfected adults aged 18-60 years were recruited into this study. Risk status for HIV infection was assessed at entry by a series of questions designed to identify risk for HIV-1 infection. The collaborating clinical trials groups in this study, AVEG and HIVNET, used group-specified standard criteria for defining the risk categories

The definition of the low-risk group follows that used in the AVEG phase 1 studies, since the rationale for including a low-risk group in the phase 2 study was to ensure that immune responses to the new lot and new dose of vCP205 were comparable to what has been seen in phase 1 studies. Briefly, low-risk subjects were persons with ⩽2 sex partners presumed to be HIV negative in the past 6 months, no injection drug use, and no newly acquired sexually transmitted diseases in the past 6 months [16]

The higher-risk participants were individuals from groups who may be asked to participate in phase 3 efficacy trials, including gay or bisexual men, injection drug users of either sex, and higher-risk heterosexual women. Participants at HIVNET sites were eligible for the study if they met the eligibility criteria for the HIVNET Vaccine Preparedness Study [16]. In brief, higher-risk subjects in HIVNET were, for men, persons reporting anal intercourse with another man in the last year, vaginal or anal intercourse with an HIV-infected woman in the last year, or injection of illicit drugs in the last 6 months. For women, entry criteria included injection of illicit drugs within the last 6 months, having a current HIV-infected male partner or a partner who had sex with other men, having ⩾5 male sex partners within the last year, or exchanging sex for money or drugs within the last year. AVEG volunteers were defined as higher risk if they met the HIVNET criteria for defining higher risk, with the exception that men who had sex with men and were in long-term mutually monogamous relationships with a presumed HIV-negative partner were considered lower risk

After informed consent was obtained, a history and physical examination were done, and screening laboratory assessments were performed. Eligibility criteria included a normal complete blood cell count (white blood cell count, 3500–12,000 cells/mm³; total lymphocyte count, ⩾800 cells/mm³; platelets, 125,000–550,000/mm³; and differential within institutional normal limits); hematocrit ⩾30% for women, ⩾38% for men; alanine aminotransferase ⩽3 times the institutional upper normal limit; creatinine ⩽1.6 mg/dL; normal urine dipstick test results, including esterase and nitrite; negative ELISA for HIV within 8 weeks of immunization; and availability for follow-up for a planned duration of ⩾24 months

Exclusion criteria included a history of immunodeficiency, chronic illness, autoimmune disease, a psychiatric condition, receipt of live attenuated vaccines within 60 days of the study, receipt of blood products or immunoglobulin in the past 6 months, active syphilis or tuberculosis (TB; but volunteers with a positive purified protein derivative test result and a normal chest x-ray showing no evidence of TB, and not requiring isoniazid therapy, were eligible), immediate type hypersensitivity reaction to egg products or neomycin (used to prepare ALVAC vaccines), prior receipt of HIV-1 vaccines or receipt of placebo in a previous HIV vaccine trial, and, in women, pregnancy or lactation

Vaccine Schedule, Risk Reduction Counseling, Self-Reported Risk-Taking Behavior, Benefits and Risks of Participation, and Monitoring for Unblinding

Volunteers received 2 injections each at times 0, 1, 3, and 6 months (table 1). vCP205 (groups A and B) or placebo ALVAC (group C) was given intramuscularly in the left arm, and rgp120 (group A) or saline (groups B and C) was given intramuscularly in the right arm. Volunteers were evaluated 1 day after each injection for systemic and local reactions, and self-reported data were collected for 4 days. Data on serious adverse events were collected for 2 years of study

View larger version:

• In this window
• In a new window

• Download as PowerPoint Slide

Figure 1

Top Safety of recombinant canarypox vector expressing human immunodeficiency virus type 1 (HIV-1) antigens (vCP205), with or without recombinant envelope glycoprotein (gp120), in higher- and lower-risk persons. Bars indicate percentages of volunteers reporting pain or tenderness after the indicated dose of vaccine among those who received placebo, the combination vCP205 + gp120, or vCP205 alone. Adverse reactions were classified as mild (hatched bars) moderate (gray bars) or severe (black bars) as defined elsewhere [23]. Bottom Percentages of subjects, divided into placebo group, combination vaccine group, and vCP205 alone, reporting systemic symptoms, defined as fever, nausea, malaise, myalgia, or headache, within 3 days of vaccination, after the indicated dose of vaccine

View larger version:

• In this window
• In a new window

• Download as PowerPoint Slide

Figure 2

Lymphocyte proliferation responses to recombinant human immunodeficiency virus type 1 (HIV-1) antigens (rp24 antigen [top panel] or rgp120 SF-2 [bottom panel]) among 44 subjects given placebo (open bars) live recombinant canarypox vector (vCP205) expressing HIV-1 antigens (hatched bars) or vCP205 + rgp120 (solid bars). Mean stimulation indices (SIs) are shown above the columns. A positive SI was defined as ⩾5. Significant differences in the proportion of subjects with positive lymphoproliferative responses to rgp120 SF-2 after dose 4 were present among the group receiving the combination vector vCP205 + rgp120 vaccination, compared with those receiving only vCP205 (P=.01, Fisher’s exact test)

View larger version:

• In this window
• In a new window

• Download as PowerPoint Slide

Figure 3

Percentages of higher-risk volunteers participating in this study of human immunodeficiency virus type 1 (HIV-1) vaccines with increases or decreases (bars) in self-reported high-risk behavior for 7 indicated behaviors. Data at baseline to 1 year are illustrated

View larger version:

• In this window
• In a new window

• Download as PowerPoint Slide

Table 1

Vaccine schema and distribution of volunteers according to human immunodeficiency virus (HIV) vaccine and recruitment groups (AVEG or HIVNET)

At the time of each vaccination and every 3 months during the 24 months of follow-up, volunteers were given risk reduction counseling. These times were also used for monitoring for intercurrent HIV infection, collection of risk-taking behavior data, and collection of data on events related to participation. These latter observations included volunteers’ perceived benefits and social harms of participation in the study. To monitor for possible unblinding of the study, volunteers were asked if HIV antibody assays had been done outside the study and the reasons for testing

Immunogenicity

Serum was obtained from all subjects before initial vaccination, 2 weeks after doses 3 and 4 (3- and 6-month injection times), and at month 12. Humoral immune responses to HIV-1 were assessed at those times, including binding antibody reactivity by Sanofi HIV-1/HIV-2 peptide EIA kit (Genetic Systems), HIV-1/-2 ELISA (Abbott Laboratories), Western blot (commercial test kits varied by location), and EIA to gp120 (SF-2), as described elsewhere [17]. Neutralizing activity to HIV-1 MN was determined among the subset of volunteers who were also assayed for CTL (see below), as described elsewhere [18, 19]. Neutralization of primary isolates of HIV-1 was not assessed in this study; little neutralizing of primary isolates was detected in a phase 1 study [20]

A subset of volunteers had peripheral blood cells drawn at the time serum was drawn (as described above), to assess cellular immune responses to HIV-1. These assays included HIV-specific env and gag CD8⁺ CTL activity. Ninety volunteers each in groups A and B and 30 volunteers in group C were randomized into the CTL substudy. To maintain blinding of the treatment assignments, acid citrate dextrose–anticoagulated blood was collected from all AVEG volunteers (180 individuals) and from all volunteers at the HIVNET trial sites participating in the CTL/Neutralization substudy (∼90 volunteers per vaccine group and 30 placebo recipients). The peripheral blood mononuclear cells (PBMC) from the day of initial immunization were cryopreserved at the AVEG Central Immunology Laboratory for future studies, for quality control and assay comparisons. CD8⁺ CTL were determined at the central laboratory, using fresh PBMC (K. Weinhold, Duke University), as described elsewhere [20]. Lymphocyte proliferation assays were performed on the 44 Saint Louis volunteers at the Saint Louis University AIDS Vaccine Evaluation Unit, using cryopreserved PBMC, to assess T cell memory for gag and env by stimulation with rp24 and rgp120 SF-2 antigens, using methods described elsewhere [21, 22]

Statistical Analysis

χ² tests were used to compare frequency distributions, with Fisher’s exact test used for 2×2 tables. Wilcoxon tests were used to test for treatment differences in vaccine reactogenicity scores. Antibody responses were compared by repeated measures analysis of variance, and reciprocal antibody titers below the level of detection were assigned a value of half the lower limit (e.g., a reciprocal titer of <10 was expressed as 5). The proportion of subjects who were CD8⁺ CTL positive were examined with adjustment for repeated testing, using generalized estimating equation (GEE) methods. Analysis was done using the SAS version 6.12 software package (SAS Institute)

Volunteer Accrual and Demographic Data

The first volunteer was entered on 22 May 1997, and the last volunteer was entered on 22 January 1998; 435 volunteers (4% above target) were enrolled. Table 2 summarizes the demographics for the volunteers. Sixty-nine volunteers were considered to be higher risk due to drug abuse and 306 due to sexual risk factors. Among the higher-risk men, 77 gave a history of sex with an HIV-infected partner, 213 had sex with a partner of unknown HIV status, 12 had receptive anal sex with an HIV-infected partner, and 61 had receptive anal sex with a partner of unknown HIV status. Among higher-risk women, 6 reported unprotected vaginal sex with an HIV-infected partner

View larger version:

• In this window
• In a new window

• Download as PowerPoint Slide

Table 2

Demographic characteristics of the 435 volunteers enrolled and stratified into the indicated arms of the study of human immunodeficiency virus type 1 (HIV-1) recombinant canarypox virus vaccine expressing HIV-1 antigens (vCP205), with or without the recombinant gp120 SF2 strain: 2 vaccine regimens versus 1 placebo regimen

Vaccination and Vaccine Safety

Four hundred thirty-five volunteers received the first immunization, 432 received the second, 427 received the third immunization, and 416 received the fourth immunization. Reasons for not receiving all vaccinations included intercurrent HIV infection (5 subjects; one between screening and dose 1, two before dose 3, and two before dose 4). Two volunteers died: one due to homicide and the other due to cocaine overdose with subdural haemorrhage; each had received 3 vaccinations. Three volunteers were incarcerated, one each after the first, second, and third vaccinations. Three volunteers were lost to follow-up, two after the second vaccination and one after 3 vaccinations. Four volunteers either moved (one after the second vaccination) or refused to continue further in the study (1 volunteer each after the first, second, and third vaccinations). Two volunteers had seizures believed to be unrelated to the vaccine but were not given additional vaccine. Another volunteer was unable to continue due to an unrelated adverse event (uterine leiomyomas) after 3 vaccinations

The profiles of local and systemic reactions to the vaccines or placebo were not different between higher- and lower-risk subjects (data not shown). Moderate local pain or tenderness and erythema or induration at the injection site was significantly more common among subjects receiving one or both vaccines than among those receiving placebo (figure 1) [23]. Severe systemic events or high fever was uncommon. Severe systemic symptoms (severe malaise, myalgia, nausea, and/or headache) were reported by 9 volunteers (two with dose 1, two with dose 2, one with dose 3, four with dose 4); one was in the placebo group, five were in the vCP205 group and three were in the vCP205 + gp120 group. Severe local reactions (severe pain and/or tenderness) were reported by 6 volunteers (one reported severe pain and tenderness after both dose 1 and dose 2, one with dose 1, two with dose 2, and two with dose 3); none were in the placebo group, three were in the vCP205 group, and three were in the vCP205 + gp120 group. No volunteers had both severe systemic and severe local reactions, and volunteers were not excluded from additional vaccinations because of these reactions. Moderate systemic events (malaise, headache, or myalgia) were more likely to occur in vaccinated subjects than in subjects receiving placebo (figure 1). Among volunteers receiving vCP205 (left arm) + gp120 (right arm), there were no significant differences between left and right arms in reaction severity with respect to pain, tenderness, erythema, or induration (data not shown). However, among volunteers receiving vCP205 (left arm) + saline (right arm), there were significant differences between arms in reaction severity with respect to pain and tenderness (pain and tenderness greater in left arm than right arm in each case; P=.001), but there were no significant differences with respect to erythema or induration. Among the volunteers receiving ALVAC placebo + saline, there were no significant differences between left and right arms in reaction severity with respect to pain, tenderness, erythema, or induration. Overall, there was not an increase in local or systemic events with dose 4, compared with dose 1; rather, there was a trend toward decreased local and systemic events (figure 1)

Immunogenicity

Antibody responsesSubjects given vector plus gp120 developed significantly more neutralizing antibody than did subjects given vector alone (table 3). Vector alone induced low titers of neutralizing antibody in 56% of volunteers, and the addition of rgp120 SF-2 significantly increased the proportion of volunteers with neutralizing antibodies. However, compared with historic data from a previous study, in which vCP205 vector alone was given at times 0 and 1 month, followed by vCP205 + rgp120 SF-2 at 3 and 6 months (geometric mean neutralizing titer [GMT], 1:232; 95% confidence interval [CI], 127–427 to MN strain), the GMT of volunteers in the present study (1:48; 95% CI, 38–60 to MN strain), after simultaneous administration of vaccines vCP205 + rgp120 SF-2 at all 4 time points (0, 1, 3, and 6 months), was significantly lower

View larger version:

• In this window
• In a new window

• Download as PowerPoint Slide

Table 3

Development of human immunodeficiency virus type 1 (HIV-1) binding or neutralizing antibodies 2 weeks after the fourth vaccination with recombinant canarypox vector expressing HIV-1 antigens (vCP205), with or without recombinant gp120 antigen

The majority of subjects vaccinated with either the vCP205 vector alone (73% positive) or vCP205 + gp120 subunit (72% positive) developed a reactive EIA by the Abbott HIV-1/-2 test by 2 weeks after dose 4 (table 3). The Sanofi EIA, an assay based on detecting antibody to a gp41 peptide not contained in the vaccine, infrequently (<2%) was positive. Western blotting detected env bands in 15% and 40% of subjects who received either vector alone or the vector and subunit, respectively (P=.07, Fisher’s exact test). Binding antibody assay results to SF-2 vaccine antigen 2 weeks after the second, third, and fourth vaccinations were 0% at all times for placebo recipients, <5% at all times for vCP205 recipients, and 53%, 92% and 93%, respectively, for the vCP205 + gp120 vaccine group. There was no difference in vaccine immunogenicity between higher- and lower-risk subjects, as indicated by antibody to immunogen (data not shown)

Cellular immunityCTL and lymphocyte proliferation assay results are summarized in table 4 and figure 2, respectively. At some time point during the study, 33% of vaccinated subjects developed CD8⁺ CTL to env and/or gag HIV antigen. As expected, env or gag CD8⁺ CTL activity was detected infrequently in volunteers receiving placebo. Only 2 positive responses were observed among placebo recipients, and these are believed to have been false positives, since HIV infection was not present in these subjects. The proportion of volunteers with CD8⁺ CTL was not significantly different between the vector alone– (32 of 84 volunteers) and the vector + gp120–vaccinated group (24 of 86 volunteers; P=.8, GEE method). No difference in the frequency of CTL responses was found between higher- and lower-risk subjects (table 4; P=.5, GEE method). Addition of rgp120 to the vaccine regimen did not increase the CD8⁺ CTL response to env in the vaccine group (table 4)

View larger version:

• In this window
• In a new window

• Download as PowerPoint Slide

Table 4

Development of CD8+ cytotoxic T lymphocytes (CTL) to gag or env in higher and lower risk volunteers according to vaccine groups given either recombinant canarypox vector (vCP205) expressing human immunodeficiency virus type 1 (HIV-1) antigens alone or with recombinant gp120 HIV-1 subunit antigens or placebo

There was no correlation between development of binding antibody to p24 antigen and gag CD8⁺ CTL results 2 weeks after the fourth vaccination in the subgroup of volunteers in whom both assays were conducted. Among 35 vaccinees who were antibody negative to p24, 11 had CD8⁺ CTL to gag and among 19 vaccinees with positive p24 antibodies, 5 had CD8⁺ CTL to gag (P value not significant)

Lymphocyte proliferation was detected in 8 (28%) of 29 volunteers after the fourth immunization with vCP205 alone or with rgp120 (figure 2). Proliferation to envelope glycoprotein was significantly more frequent among recipients of the combination rgp120 + vCP205 than among recipients of vCP205 alone (P=.01, Fisher’s exact test)

Social Benefits and Risks of Participation

Volunteers were asked at 3-month intervals about the beneficial and negative consequences of study participation in their lives. Ninety-three percent of volunteers reported a beneficial impact of study participation at ⩾1 visits. The most commonly noted beneficial effects generally were a sense of helping to find a vaccine to prevent HIV infection, educational, or receiving risk reduction counseling. Volunteers were also asked whether study participation had a negative impact on 9 specific aspects of their lives (personal relationships, health insurance, life insurance, travel/immigration, military/other federal agencies, educational programs, employment, housing, and medical/dental treatment). They were given the opportunity to report any other negative impact not covered by the 9 categories. Overall, 106 volunteers (27%) reported ⩾1 negative impact within 24 months of entry. There was a statistically significant decrease in the number of reports between the first time (day 84 visit) participants were asked about negative impact and the second time this inquiry was made (day 168; P<.001). Reports of negative social impact were largely on personal relationships (117 reports) and less frequently involved health insurance (2 reports) or life insurance (5 reports) coverage, attitudes or behaviors of medical providers (3), loss of employment (2), or problems with a landlord (1)

Neither of the individuals reporting problems with health insurance was denied coverage; one was waiting for the study to conclude before applying for coverage, and the other had job-related changes in insurance unrelated to the trial. Two of the persons reporting life insurance problems were denied coverage, but only one denial of coverage was directly related to trial participation; in this situation, the participant had incorrectly reported to the insurance company that the participant was HIV positive, despite a negative test result. In both situations, staff at the sites have written letters to the insurance companies to clarify that participants were uninfected. Two volunteers reported being fired when their employers learned of their study participation. In both instances the vaccine study site staff wrote letters to the employers, but the volunteers did not want to pursue any further action and found other employment

Potential Unblinding

Forty-seven volunteers had a total of 59 HIV antibody tests done outside the study; only one was specifically done to attempt to unblind the study. Reasons for HIV tests outside this study included participants’ being in another research study (n=15), medical admission or health examination (n=17), insurance examination (n=7), jail or detoxification center (n=9), fear of HIV infection (n=4), health care worker needlestick injury (n=2), selling blood (n=2), other reasons (n=2), and attempting to find out whether vaccine was received (n=1)

Change in Risk Behavior over Time and HIV Infections among Volunteers

Selected self-reported risk-taking behaviors are summarized for year 1 of the study in figure 3. Overall, the level of self-reported risks decreased. Within each category there were some subjects who reported an increase in risk taking, but for each high-risk behavior there were as many or more subjects reporting a decrease in risky behavior, compared with those who reported an increase

Intercurrent HIV infections and other severe adverse events are summarized in table 5. Fourteen HIV infections occurred among the volunteers during 2 years of follow-up. Of the 14 infections, 5, 2, and 2 infections occurred in the first year of study in the placebo, vCP205, and vCP205 + gp120 arms, respectively. Five infections occurred in the second year of follow-up in 1 placebo, 2 vCP205, and 2 vCP205 + gp120 vaccinees, respectively. One subject was found to be infected at the time of study. Excluding this case, the higher-risk group had 12 HIV infections in 837 follow-up years, for an attack rate of 1.43 per 100 years (95% CI, 0.74–2.50). One low-risk subject in 151 follow-up years was infected (attack rate, 0.66 per 100 years; 95% CI, 0.02–3.68). The single lower-risk subject was subsequently found to be in a mutually monogamous relationship with an HIV-infected partner who had been presumed to be HIV uninfected

View larger version:

• In this window
• In a new window

• Download as PowerPoint Slide

Table 5

Occurrence of intercurrent natural infection with human immunodeficiency virus type 1 (HIV-1) or death or life-threatening event other than HIV Infection, by study group