Safety and Immunogenicity of a DNA Vaccine With Subtype C gp120 Protein Adjuvanted With MF59 or AS01B: A Phase 1/2a HIV-1 Vaccine Trial

Supplemental Digital Content is Available in the Text.


INTRODUCTION
Despite progress in HIV prevention and treatment, an estimated 1.3 million people were newly infected with HIV in 2022, 1 highlighting the urgent need for an effective vaccine.To date, the RV144 trial remains the only HIV vaccine trial that has demonstrated partial efficacy against acquisition. 2he Pox-Protein Public-Private Partnership (P5) was established with the aim of improving RV144 by developing a vaccine capable of protecting against a broader diversity of HIV strains and achieving a better understanding of immune responses associated with preventing HIV infection. 3Vaccine concepts in the P5 program have focused on clade C immunogens, targeting predominant strains of East and Southern Africa, where approximately half of the 39 million people living with HIV reside. 1 The RV144 regimen, originally designed to protect against subtype B/E strains, was adapted to incorporate clade C antigens and adjuvanted with MF59. 4 This regimen demonstrated adequate immunogenicity in the HIV Vaccine Trials Network (HVTN)100 phase 1/2a trial 5 and was further evaluated in the HVTN702 efficacy trial in South Africa, but ultimately discontinued due to nonefficacy. 6In parallel, the P5 designed the correlates program: a series of phase 1/2a trials to evaluate vaccine candidates based on favorable immune profiles of putative correlates of protection.These trials employed novel prime-boost and co-administration regimens, varied protein doses, and used new adjuvants and vaccine delivery systems, with an emphasis on shared immunological endpoints to allow for cross-study comparisons.
Preclinical studies have shown promising immune responses using DNA/protein combination vaccines. 7,8A comparison of responses between HVTN100 (canarypox viral vector (ALVAC)) and HVTN111 (DNA) trials indicated that DNA priming with a protein boost led to increased antibody and cellular responses compared with priming with the canarypox vector. 9In the HVTN105 trial, both a DNA prime-protein boost and a co-administration regimen induced potent and durable V1/V2 binding antibody responses (a known correlate of lower HIV-1 infection risk in RV144), with co-administration inducing early antibody responses. 10urthermore, in the HVTN096 trial, including gp120 Env protein at the priming stage, co-administered with either vaccinia virus vaccine vector (NYVAC) or DNA, elicited earlier and even greater antibody responses. 11he adjuvant system 01 (AS01) has been successfully tested in vaccine trials for other infectious diseases including malaria, 12 shingles, 13,14 and tuberculosis. 15Some HIV vaccine studies have also used AS01 and have shown that it contributes to the induction of robust and persistent cellular and humoral responses. 16,17MF59 has likewise been used in several licensed vaccines and preclinical studies, 18 inducing strong and durable T-cell memory and humoral responses.MF59 was also used in HVTN studies with ALVAC 5 and was therefore chosen for comparison with AS01 B in this trial.
Thus, the aim of the HVTN108 trial was to evaluate the safety and immunogenicity of the DNA vaccine with different HIV clade C protein doses, adjuvanted with MF59 or AS01 B , and dosed in prime-boost or co-administration regimens.

Study Design
HVTN108 was a multicenter, phase 1/2a, randomized, double-blind, placebo-controlled trial.We randomly allocated participants to 1 of 7 treatment groups or placebo at 17 clinical research sites in the United States and South Africa.Vaccinations were administered at enrollment and Months 1, 3, and 6.Participants were followed for 12 months.Vaccine regimens included DNA priming at enrollment and Month 1 with DNA/protein/adjuvant boosts at Months 3 and 6; DNA/ protein/adjuvant co-administration at enrollment and Months 1 and 6; and low-dose protein/AS01 B alone at enrollment and Months 1 and 6 (Fig. 1, see Table 1, Supplemental Digital Content, http://links.lww.com/QAI/C284).Safety was assessed by a collection of reactogenicity and adverse events (AEs).Humoral and cellular responses were measured 2 weeks (peak, Month 6.5) and 6 months (durability, Month 12) after the Month 6 injection.HVTN111 was a randomized, double-blind, placebo-controlled trial in Zambian, Tanzanian, and South African sites comparing the safety and immunogenicity of DNA prime followed by DNA/protein boost with DNA/protein co-administration injected intramuscularly through either needle/syringe or biojector 2 weeks after the final (Month 6) vaccination. 19As prespecified in the study protocol, data from 66 HIV-negative adult HVTN111 participants who received identical regimens to one of the HVTN108 treatment or placebo groups were included in the immunogenicity analysis.

Participants
HVTN108 enrolled 334 HIV-negative adults aged 18-40 years of good general health.Participants were assessed as having a low likelihood of HIV acquisition, agreed to all study requirements, and provided written informed consent.Good general health was determined by medical history, physical examination, and laboratory tests.All participants assigned female sex at birth agreed to consistent use of contraception; pregnant or breastfeeding persons were excluded (full eligibility criteria in

Randomization
Randomization was determined by computer-generated sequences provided to sites through a web-based system and performed in blocks to ensure balance across groups and was stratified by geographical region.At each institution, a designated pharmacist was responsible for dispensing study products and maintaining the security of the product assignments.Participants and other site staff were blinded to group assignments.

Safety Assessments
Participants were followed for 12 months after the initial vaccination, with safety evaluations and procedures per the schedule in the study protocol.AEs were reported over 30 days after each vaccination visit, and a subset of AEs, including AEs of special interests (AESIs) and serious AEs, were reported throughout this study.

Laboratory Procedures
The immunogenicity objectives were to determine differences in vaccine-induced immune responses between prime-boost and co-administration regimens, regimens adjuvanted with AS01 B or MF59, and regimens using low-or high-dose protein.All laboratory assays (described below) were performed blinded to the treatment group with validated or qualified methods assessing peak immunogenicity (Month 6.5) and durability (Month 12).Specific antigens used in immunogenicity assays are presented in Table 3, Supplemental Digital Content, http://links.lww.com/QAI/C284.

Binding Antibody Multiplex Assay
HIV-1-specific IgG and IgG3 binding antibody responses were measured by Binding Antibody Multiplex assay, as described previously, 9,19-24 at a 1:50 dilution.Tested antigens and assay reagents included vaccinematched subtype C 96ZM651.C gp140, V1V2 antigens 1086.C V1V2 and CaseA2_gp70_V1V2.B, and heterologous antigen to assess breadth (Clade A. 00MSA gp140).All assays were conducted according to Good Clinical Laboratory Practice guidelines, including tracking of controls with Levey-Jennings charts.

ADCP and ADCC
The ability of vaccine-induced antibodies to engage Fc receptors and mediate antibody-dependent cellular phagocytosis (ADCP) by monocytes was measured as previously described. 25,26A phagocytic score was determined based on the ratio of the experimental sample to the no-antibody control.The mean phagocytosis score was calculated as follows: (% bead positive for participant • mean fluorescence intensity bead positive for participant)/(% bead positive for the no-antibody control • mean fluorescence intensity bead positive for the no-antibody control).Samples were run in duplicate within each assay and the average scores of the replicates were reported.

Statistical Analysis
Safety data were analyzed regardless of how many vaccinations participants received.Study enrollment was simultaneous with the first vaccination, thus all participants received at least 1 vaccination and provided safety data.Participants who discontinued vaccination were encouraged to remain in this study for safety follow-up.
Immune responses were summarized by the proportion of participants with a positive response to individual antigens at each time point, with boxplots showing the distributions of the immune response magnitudes among positive responders.Barnard exact and Wilcoxon 32 rank sum tests were used to compare the response rates and magnitudes for responders, respectively, between the 2 groups.Two-sided 95% confidence intervals for binomial proportions were calculated using the Wilson score method. 33All tests were two-sided with no adjustment for multiple comparisons; differences were considered statistically significant at P , 0.05.SAS (version 9.4; SAS Institute, Cary, NC) and R statistical software (version 4.0.4;R Foundation for Statistical Computing, Vienna, Austria) were used for statistical analysis.

RESULTS
A total of 400 participants were enrolled at 20 United States and African clinical research sites between 23 June 2016 and 25 July 2018.The median age at enrollment was 25 years (interquartile range 22-28), 214 (53.5%) were assigned female sex at birth, and participants had diverse racial backgrounds (Table 1).Vaccinations were completed in 369 (92.3%) participants, and 343 (85.8%) reached study completion (Fig. 1).
There were 48 grade 3 and 3 grade 4 reactogenicity events among 39/400 (9.8%) participants.Of the grade 3 events, there were 22 local and 26 systemic events.All local reactogenicity events occurred in the deltoid region where the protein and adjuvant were injected.There were 14 grade 3 erythema events, 7 grade 3 induration events, and 1 grade 3 tenderness event.Of note, all grade 3 local reactogenicity events were in participants that received the AS01 B adjuvant (Fig. 2, see Table 5, Supplemental Digital Content, http:// links.lww.com/QAI/C284).Of the 3 grade 4 events (all fevers), 2 occurred in group T7 (low-dose protein/AS01 B co-administration), and 1 occurred in group T4 (DNA/ protein/MF59 co-administration).All fevers were selflimiting and resolved within 1 day.Twelve participants (3.0%) discontinued vaccinations due to reactogenicity events (7 due to erythema and/or induration, 2 due to fever, and 3 due to systemic reactogenicity events).Eleven of these received AS01 B .
An additional 32 product-related AEs were reported in 23/400 (5.8%) participants.These included 28 grade 1 AEs and 4 grade 2 AEs, 87.5% of which started within 4 days of vaccination (see Table 6, Supplemental Digital Content, http://links.lww.com/QAI/C284).No AEs of special interests or related serious AEs were reported.There were no clinically significant differences in AEs between treatment groups.
ADCP activity among participants was increased at Month 6.5 in the AS01 B -adjuvanted co-administration group (T5) compared with the MF59-adjuvanted co-administration group (T4) (see Figure 4A, Supplemental Digital Content, http://links.lww.com/QAI/C284).Similarly, for ADCC functionality, response rates were significantly higher in the AS01 B -vs the MF59-adjuvanted group at Month 6.5, as were response magnitudes among all participants based on the luciferase assay with cells infected with either vaccinematched infectious molecular clones (see Figure 4B, Supplemental Digital Content, http://links.lww.com/QAI/C284).Of note, by contrast to what was observed with infected target cells, rates and magnitudes of ADCC responses at Month 6.5 did not differ between AS01 B -or MF59-adjuvanted groups when the gp120-coated target cells were used as targets, indicating a selective effect on epitope-specific functions by the adjuvants.
T-cell responses to the vector insert, Env-ZM96.C gp140, and to the vaccine-matched protein, 1086.C gp120, were evaluated, specifically comparing responses between the same regimens adjuvanted with MF59 or AS01 B .HIVspecific CD4 + T cells expressing IFN-g and/or IL-2 and/or CD40L were induced in most vaccine recipients at all timepoints and in all treatment groups.In the prime-boost high-dose protein regimen groups (T1 vs T2), CD4 + T-cell response rates to Env-ZM96 gp140 were higher at Month 6.5 and 12 in the AS01 B -vs MF59-adjuvanted group (P , 0.001 and P = 0.0003, respectively); response magnitudes among positive responders were higher in the AS01 B -adjuvanted groups at Month 6.5 (P = 0.021), but comparable at Month 12 (Figs.4A, B).The 1086.C gp120-specific CD4 + T-cell response rates were higher in the AS01 B -adjuvanted group at Month 6.5 and 12 (P = 0.0001 and P , 0.001, respectively), while the response magnitude was only higher in the AS01 B -adjuvanted group at Month 6.5 (P = 0.001), but not at Month 12 (Figs.4C, D).
In the AS01 B -adjuvanted prime-boost regimens (T2 vs T3), there were no differences in the Env-ZM96 gp140 specific and 1086.C gp120-specific CD4 + T-cell response rates at Month 6.5 and 12 between the low-and high-dose protein.By contrast, the Env-ZM96-specific CD4 + T-cell response magnitudes were higher at Month 6.5 (P = 0.008), and the 1086 gp120-specific CD4 + T-cell response magnitudes were higher at Months 6.5 and 12 in the low-vs high-dose protein groups (P = 0.005 and P = 0.002, respectively).
In the AS01 B -adjuvanted co-administration groups (T5 vs T6), the ENV-ZM96 gp140-and 1086 gp120-specific CD4 + T-cell response rates were higher in the low-dose protein groups at Month 12 (P = 0.0075 and P = 0.0205, respectively).There were no significant differences in the Env-ZM96-specific CD4 + T-cell response magnitudes, but the 1086.C gp120-specific CD4 + T-cell response magnitudes were higher in the low-dose vs high-dose protein group at Months 6.5 and 12 (P = 0.003 and P = 0.025).
Of note, the Env-ZM96 gp140-and 1086 gp120specific CD4 + T-cell response rates and magnitudes in the low-dose protein co-administration regimen (AS01 B adjuvant) did not show significant differences compared with the low-dose protein/adjuvant only regimen (T3 vs T6, Fig. 4).
Overall, there were no significant changes in the CD4 + T-cell response rates to all antigens between Month 6.5 and Month 12, but the response magnitudes decreased from Month 6.5 to Month 12 in all treatment groups regardless of adjuvant, protein dose, or regimen (P , 0.001, see Figure 2A, Supplemental Digital Content, http://links.lww.com/QAI/C284).Of note, while statistical comparisons were not done, the same regimen induced comparable CD4 + T-cell response rates and magnitudes whether administered as prime-boost or co-administered.Very few CD8 + T-cell responses were induced across groups (see Figure 5, Supplemental Digital Content, http://links.lww.com/QAI/C284).

DISCUSSION
In this phase 1/2a HIV vaccine trial, we assessed a DNA/protein vaccine with varying dosage regimens, protein doses, and adjuvants.All study groups had acceptable safety profiles, although more reactogenicity events were reported in the AS01 B -adjuvanted groups.Overall, all vaccine groups showed high IgG response rates and magnitude to gp120 and gp140, and moderate-to-high response rates and magnitude to Env V1V2.The AS01 B -adjuvanted DNA/protein coadministration regimen induced more durable antibody responses than the other regimens and showed higher phagocytosis scores than the MF59-adjuvanted coadministration regimen.The AS01 B -adjuvanted regimens induced higher CD4 + T-cell responses that persisted even 6 months after the last vaccination.Furthermore, we found that prime-boost or co-administration regimens including the lower protein dose induced immune responses comparable with or better than those induced with the higher dose.As humoral and cellular responses were strong, and antibodies were more durable, the co-administration regimen merits further evaluation.
0][11] However, these studies included a variety of vaccine candidates in different combinations, doses, and injection schedules.Furthermore, immunological assessments were performed in a variety of different laboratories using different assays.HVTN108, as part of P5, was optimized to compare regimens directly.
HVTN108 aimed to characterize immune responses elicited by regimens containing DNA and adjuvanted protein without a poxviral vector to down-select vaccine candidates for efficacy testing.In preclinical models, co-administration of DNA/protein elicited more robust humoral immunity than DNA alone or a prime-boost strategy. 7,8In humans, using a DNA prime-gp140 protein boost regimen yielded high levels of Env-binding antibodies and homologous neutralizing antibodies compared with protein alone, as well as robust and highly polyfunctional CD4 + T-cell responses to Env antigens.In HVTN111, DNA prime followed by DNA/ protein boost was compared with DNA/protein coadministration; the co-administration regimen was associated with an increased HIV-1 V1/V2 antibody response rate, a known correlate of decreased HIV-1 infection risk in RV144. 19e showed that DNA prime alone followed by DNA and protein/adjuvant boosts elicited robust Env-specific CD4 + T-cell, antibody, and ADCP/ADCC responses.While coadministration of DNA and protein/adjuvant induced comparable CD4 + T-cell responses to the prime-boost regimens, binding antibody responses were considerably higher and more durable.Of interest, the AS01 B -adjuvanted 20 mg protein administered alone induced comparable humoral and cellular response rates to the 20 mg DNA/protein coadministration regimen.These results highlight the potential potency of protein-based vaccines combined with an immunogenic adjuvant, but further assessment of the quality of immune responses is required to understand the potential impact of this regimen.
Adjuvants enhance the quality and durability of vaccine-induced immune responses.The MF59 adjuvant is used in flu vaccines due to its ability to improve antibody affinity maturation, targeted epitope breadth, and binding affinity and to elicit balanced Th1/Th2 responses. 34AS01 B has been used in several non-HIV vaccine candidates because of its ability to enhance the induction of durable immune responses. 12,35,36The superior CD4 + T-cell induction associated with AS01 B supports this adjuvant system for further HIV vaccine evaluation.Consistent with previous studies, we have shown that both cellular and humoral responses were significantly higher in the AS01 B -adjuvanted groups and that these responses were durable.Given that IgG3 response rates were also increased with AS01 B and that IgG3 is associated with improved ADCP function, 25 AS01 B likely influenced the Fc region of antibodies resulting in modified interactions with cellular Fc receptors.
We assessed whether a high-dose protein coadministered with AS01 B may overstimulate the immune system and thereby dampen or suppress responses.In HVTN041, a combination vaccine (NefTat and gp120W61D) formulated with AS02A was administered with varying doses (5, 20, and 100 mg) of gp120 protein.While participants developed durable gp120-specific binding antibodies, a dampening effect on CD4 + T-cell responses occurred at the highest dose. 37We evaluated similar regimens containing either 20 or 100 mg of the bivalent Env proteins adjuvanted with AS01 B and found similar results for cellular and humoral responses.
Strengths of our study included the randomized controlled trial design with a relatively large sample size, including diverse participants recruited on 2 continents, thereby increasing generalizability.We also compared multiple vaccine strategies with different schedules, adjuvants, and components in the same trial.A limitation of prime-boost and co-administration group comparisons was that coadministration participants received 3 protein doses while prime-boost participants received 2. Furthermore, HVTN108 was a phase 1/2a trial, meaning that it did not include HIV infection endpoints, so extrapolating whether the elicited immune responses in the promising regimens translate into protection remains unknown and would require a larger trial.Similar to previous HIV vaccine trials, 19 few CD8 + T-cell responses were elicited by any of the regimens in this trial, indicating that new approaches may be needed to activate this important line of defense.
While the induction of broadly neutralizing antibodies is a priority in HIV vaccine design, evidence is building for the role of Fc effector functions and nonneutralizing antibodies in HIV prevention.9][40] IgG3 antibodies have demonstrated enhanced effector functions, including ADCC and ADCP due to the longer hinge region compared with other IgG subclasses.This study provided insights into the elicitation of IgG and IgG3 antibodies and effector functions in regimens utilizing MF59 and AS01 B , in addition to the varied dosing regimens.Analyses are ongoing to determine potential correlations between observed responses and to compare these data to those of the HVTN702 trial, where significant correlations between humoral and cellular responses and HIV-1 acquisition were observed. 41VTN108 showed that prime-boost and DNA/protein/ adjuvant co-administration vaccination strategies were generally well tolerated.Combination administration of DNA/ protein/AS01 B elicited the strongest humoral responses and AS01 B -adjuvanted regimens elicited stronger CD4 + T-cell responses and antibody functions compared with MF59, providing important new insights into these vaccine products and suggesting that they may be valuable components of vaccine regimens evaluated in future trials.

FIGURE 1 .
FIGURE 1. HVTN 108 (with 66 included HVTN 111 participants) CONSORT diagram.Enrollment and follow-up of participants in HVTN 108 and selected participants in HVTN 111, including availability of samples for immunologic testing.Overlap may exist between "Vaccination Discontinued" and "Early Termination" participants, as those lost to follow-up during vaccination could fall into both categories.Ab, antibody; ICS, intracellular cytokine staining; M, month.

FIGURE 3 .
FIGURE 3. Antibody responses 2 weeks (Month 6.5) and 6 months (Month 12) after completion of the primary vaccine regimen.IgG response rate (bar charts) and magnitude (boxplots) 2 weeks or 6 months after the final immunization by treatment arm for various antigens.C-A, co-administration; P-B, prime-boost.

TABLE 1 .
Baseline Demographics of Participants in the HVTN 108 and HVTN 111 Trials