Evaluation of a mosaic HIV-1 vaccine in a multicentre, randomised, double-blind, placebo-controlled, phase 1/2a clinical trial (APPROACH) and in rhesus monkeys (NHP 13-19)

Abstract

Background: More than 1·8 million new cases of HIV-1 infection were diagnosed worldwide in 2016. No licensed prophylactic HIV-1 vaccine exists. A major limitation to date has been the lack of direct comparability between clinical trials and preclinical studies. We aimed to evaluate mosaic adenovirus serotype 26 (Ad26)-based HIV-1 vaccine candidates in parallel studies in humans and rhesus monkeys to define the optimal vaccine regimen to advance into clinical efficacy trials.

Methods: We conducted a multicentre, randomised, double-blind, placebo-controlled phase 1/2a trial (APPROACH). Participants were recruited from 12 clinics in east Africa, South Africa, Thailand, and the USA. We included healthy, HIV-1-uninfected participants (aged 18-50 years) who were considered at low risk for HIV-1 infection. We randomly assigned participants to one of eight study groups, stratified by region. Participants and investigators were blinded to the treatment allocation throughout the study. We primed participants at weeks 0 and 12 with Ad26.Mos.HIV (5 × 1010 viral particles per 0·5 mL) expressing mosaic HIV-1 envelope (Env)/Gag/Pol antigens and gave boosters at weeks 24 and 48 with Ad26.Mos.HIV or modified vaccinia Ankara (MVA; 108 plaque-forming units per 0·5 mL) vectors with or without high-dose (250 μg) or low-dose (50 μg) aluminium adjuvanted clade C Env gp140 protein. Those in the control group received 0·9% saline. All study interventions were administered intramuscularly. Primary endpoints were safety and tolerability of the vaccine regimens and Env-specific binding antibody responses at week 28. Safety and immunogenicity were also assessed at week 52. All participants who received at least one vaccine dose or placebo were included in the safety analysis; immunogenicity was analysed using the per-protocol population. We also did a parallel study in rhesus monkeys (NHP 13-19) to assess the immunogenicity and protective efficacy of these vaccine regimens against a series of six repetitive, heterologous, intrarectal challenges with a rhesus peripheral blood mononuclear cell-derived challenge stock of simian-human immunodeficiency virus (SHIV-SF162P3). The APPROACH trial is registered with ClinicalTrials.gov, number NCT02315703.

Findings: Between Feb 24, 2015, and Oct 16, 2015, we randomly assigned 393 participants to receive at least one dose of study vaccine or placebo in the APPROACH trial. All vaccine regimens demonstrated favourable safety and tolerability. The most commonly reported solicited local adverse event was mild-to-moderate pain at the injection site (varying from 69% to 88% between the different active groups vs 49% in the placebo group). Five (1%) of 393 participants reported at least one grade 3 adverse event considered related to the vaccines: abdominal pain and diarrhoea (in the same participant), increased aspartate aminotransferase, postural dizziness, back pain, and malaise. The mosaic Ad26/Ad26 plus high-dose gp140 boost vaccine was the most immunogenic in humans; it elicited Env-specific binding antibody responses (100%) and antibody-dependent cellular phagocytosis responses (80%) at week 52, and T-cell responses at week 50 (83%). We also randomly assigned 72 rhesus monkeys to receive one of five different vaccine regimens or placebo in the NHP 13-19 study. Ad26/Ad26 plus gp140 boost induced similar magnitude, durability, and phenotype of immune responses in rhesus monkeys as compared with humans and afforded 67% protection against acquisition of SHIV-SF162P3 infection (two-sided Fisher's exact test p=0·007). Env-specific ELISA and enzyme-linked immunospot assay responses were the principal immune correlates of protection against SHIV challenge in monkeys.

Interpretation: The mosaic Ad26/Ad26 plus gp140 HIV-1 vaccine induced comparable and robust immune responses in humans and rhesus monkeys, and it provided significant protection against repetitive heterologous SHIV challenges in rhesus monkeys. This vaccine concept is currently being evaluated in a phase 2b clinical efficacy study in sub-Saharan Africa (NCT03060629).

Funding: Janssen Vaccines & Prevention BV, National Institutes of Health, Ragon Institute of MGH, MIT and Harvard, Henry M Jackson Foundation for the Advancement of Military Medicine, US Department of Defense, and International AIDS Vaccine Initiative.

Copyright © 2018 Elsevier Ltd. All rights reserved.

Figures

Figure 1.. APPROACH and rhesus monkey study screening, enrollment, vaccination, termination and continuation.

Participants in the APPROACH study were randomized by region (US, Africa and Asia) into eight vaccine regimen groups. Groups 1–7 were administered prime immunizations with trivalent Ad26 followed by different boost immunizations: Ad26+gp140 HD, Ad26+gp140 LD, Ad26, MVA+gp140 HD, MVA+gp140 LD, MVA, gp140 HD, respectively. Group 8 received placebo for both prime and boost. 86% of participants received all four vaccinations, 92% of participants received the first 3 vaccinations, 4% of participants received only 2 vaccinations and 3% received only the first vaccination. The full analysis set (n=393) consisted of all participants who were randomized and who received at least one dose of study vaccine. This was the primary population for all analyses (except immunogenicity). The per protocol immunogenicity population (PPI [n=358]) consisted of all participants who received at least the first three vaccinations, according to the protocol-specified vaccination schedule (+/− 2 weeks), have at least one measured post-dose blood sample collected and were not diagnosed with HIV during the study. Week 50/52 samples from participants in the PPI population who missed the 4th vaccination or did not receive the 4th vaccination in the protocol-specified time window (+/− 2 weeks) were excluded from the analysis (Panel A). Rhesus monkeys were also randomized into regimens that consisted of two prime immunizations with Ad26, followed by boost immunization with either: Ad26, MVA, gp140 or a combination thereof. Rhesus monkeys were then challenged with SHIV-SF162P3, and viral loads determined by qualified viral load assay. All monkeys were correctly immunized and survived to the end of the study period (Panel B).

Figure 1.. APPROACH and rhesus monkey study screening, enrollment, vaccination, termination and continuation.

Participants in the APPROACH study were randomized by region (US, Africa and Asia) into eight vaccine regimen groups. Groups 1–7 were administered prime immunizations with trivalent Ad26 followed by different boost immunizations: Ad26+gp140 HD, Ad26+gp140 LD, Ad26, MVA+gp140 HD, MVA+gp140 LD, MVA, gp140 HD, respectively. Group 8 received placebo for both prime and boost. 86% of participants received all four vaccinations, 92% of participants received the first 3 vaccinations, 4% of participants received only 2 vaccinations and 3% received only the first vaccination. The full analysis set (n=393) consisted of all participants who were randomized and who received at least one dose of study vaccine. This was the primary population for all analyses (except immunogenicity). The per protocol immunogenicity population (PPI [n=358]) consisted of all participants who received at least the first three vaccinations, according to the protocol-specified vaccination schedule (+/− 2 weeks), have at least one measured post-dose blood sample collected and were not diagnosed with HIV during the study. Week 50/52 samples from participants in the PPI population who missed the 4th vaccination or did not receive the 4th vaccination in the protocol-specified time window (+/− 2 weeks) were excluded from the analysis (Panel A). Rhesus monkeys were also randomized into regimens that consisted of two prime immunizations with Ad26, followed by boost immunization with either: Ad26, MVA, gp140 or a combination thereof. Rhesus monkeys were then challenged with SHIV-SF162P3, and viral loads determined by qualified viral load assay. All monkeys were correctly immunized and survived to the end of the study period (Panel B).

Figure 2.. APPROACH immune response to vaccination regimens.

All Panels show responder rates for each vaccine group for baseline (BL) and weeks post third and fourth vaccination (week 26 or 28, and week 50 and 52, respectively) beneath the graph. All groups (except for placebos) were primed twice with Ad26 vectors and then boosted with the regimens shown at the top of the graph. Vaccine response was defined as titer >threshold (if baseline is 6 PBMC) are shown. The dotted horizontal line shows the threshold for each assay performed: lower limit of quantification (LLOQ) for Panel A, and the limit of detection (LOD) for Panel B, and the 95th percentile (P95) of the overall baseline values in Panel C. Panel D shows the number of ELISPOT subpools with vaccine-induced T-cell responses for a subset of participants in Ad26/Ad26+gp140 LD and Ad26/MVA+gp140 LD vaccine groups, respectively. The dotted line shows the median number of subpools recognised. PBMC denotes peripheral blood mononuclear cells; Resp % is the percentage response, and SFU are spot forming units.

Figure 2.. APPROACH immune response to vaccination regimens.

All Panels show responder rates for each vaccine group for baseline (BL) and weeks post third and fourth vaccination (week 26 or 28, and week 50 and 52, respectively) beneath the graph. All groups (except for placebos) were primed twice with Ad26 vectors and then boosted with the regimens shown at the top of the graph. Vaccine response was defined as titer >threshold (if baseline is 6 PBMC) are shown. The dotted horizontal line shows the threshold for each assay performed: lower limit of quantification (LLOQ) for Panel A, and the limit of detection (LOD) for Panel B, and the 95th percentile (P95) of the overall baseline values in Panel C. Panel D shows the number of ELISPOT subpools with vaccine-induced T-cell responses for a subset of participants in Ad26/Ad26+gp140 LD and Ad26/MVA+gp140 LD vaccine groups, respectively. The dotted line shows the median number of subpools recognised. PBMC denotes peripheral blood mononuclear cells; Resp % is the percentage response, and SFU are spot forming units.

Figure 3.. Immune response to vaccination regimens in rhesus monkeys.

All groups (except for placebos) were primed twice with Ad26 vectors and then boosted with the regimens shown at the top of the graph. Humoral immune responses to each vaccine group were measured by ELISA and ADCP assays (Panel A and B), and cellular immune response was measured by ELISPOT assay (Panel C). In Panels A and B, shown beneath the data points are the geometric mean titers (GMT) of binding antibody responses for each vaccine group at baseline (BL) and weeks post third and fourth vaccination (week 26 or 28, and week 54 and 56, respectively). In panel C, median spot forming units per million PBMC (SFU / 106 PBMC) responses are shown. In Panels A and C, the dotted horizontal line shows the lower limit of quantification. Vaccine response was defined as titer >threshold (if baseline is <threshold or is missing); otherwise, was a titer with a 3-fold increase from baseline (if baseline is ≥threshold). PBMC denotes peripheral blood mononuclear cells; Resp % is percentage response, and SFU are spot forming units.

Figure 4.. Protection and correlates in rhesus monkeys.

Six weekly intrarectal challenges were administered to rhesus monkeys at weeks 0, 1, 2, 3, 4, 5. Panel A shows Kaplan-Meier plot of the level of protection each vaccine regimen offered to the 12 rhesus monkeys at risk, assessed one week after each challenge. No animals were censored. Panel B shows humoral and cellular immune response measured by clade C ELISA at week 28 and PTEg Env ELISPOT at week 26, the week following each of six challenges (at weeks 77–84) of monkeys from groups: Ad26/Ad26, Ad26/gp140, Ad26/Ad26+gp140. Blue dots represent pooled uninfected monkeys per timepoint, and red dots indicate pooled monkeys that were infected following each challenge. The diagonal lines display model-derived probabilities of infection, modeled on ELISA and ELISPOT responses.

Figure 5.. Human and rhesus monkey data comparison.

Comparisons of the magnitude of immunological responses between rhesus monkey and human studies. Panels A and C display transformed rhesus monkey and human immunological responses to each vaccine regimen, by ELISA and ELISPOT assay data, respectively. Rhesus monkey ELISA data in panel A and B have been transformed to human ELISA units. Panel B shows longitudinal comparison of clade C gp140 ELISA data for Ad26+gp140 groups. Panel D shows comparisons of Env PTEg ELISPOT binding response between humans and monkeys at post third and fourth vaccinations in the Ad26+gp140 group. PBMC denotes peripheral blood mononuclear cells, SFU are spot forming units; data in panels A-C are represented as GMT ± SD.