Members presentNancy L. Haigwood, ChairJames A. Bradac, Executive SecretaryDeborah L. Birx Dennis R. BurtonJeffrey D. LifsonDouglas NixonBali PulendranNina D. RussellJerald SadoffGeorge M. ShawWilliam Snow
Ex Officio Members PresentJay A. BerzofskyCol. Nelson MichaelGary Nabel
Lawrence Corey via phone
SpeakersRama Rao AmaraDan BarouchRonald C. DesrosiersVanessa HirschBrandon KeeleJohn MascolaNancy MillerDavid C. MontefioriMichael PensieroMatthew R. ReynoldsHarriet L. RobinsonJohn K. RoseRuth RuprechtAlan SchultzNancy A. Wilson
Dr. James Bradac welcomed Dr. Nancy Haigwood as the new AVRS chair, and thanked Dr. Alan Schultz, who helped put together the day 1 workshop.
Dr. Haigwood welcomed the subcommittee members and others in attendance. She introduced the two new members, Dr. Satya Dandekar, who could not attend the meeting, and Mr. William Snow, who was present.
Dr. Haigwood then introduced the topic for day 1 of the meeting, which was the use of challenge models for AIDS vaccine research in non-human primates (NHPs).
NHP models have contributed to vaccine research in several critical areas:
Vaccine design has gone through ebbs and flows.
Some major issues are causing investigators to rethink the use of NHP. Most earlier studies were performed on small numbers of animals per group utilizing challenge viruses with a variety of pathogenic potential, making it difficult to truly understand what these studies mean outside of the lab. Vaccines that provided some protection against moderately replicating viruses did not stand up well against more aggressive viruses. In addition, the challenges were all high-dose and in many cases intravenous and might not be relevant to heterosexual exposure. Finally, the challenges were generally homologous with the vaccines and might not be predictive in the context of actual transmission environments.
Some key issues remain:
Dr. Haigwood concluded by noting the convergence of goals. Collaborations on vaccine analyses have served as a paradigm for effective cooperation. Clinicians and basic scientists have found common ground in standardized, comparative immunological analyses.
Dr. Alan Schultz of DAIDS spoke about NHP lentivirus models to aid in the identification and development of an HIV vaccine.
Dr. Schultz began by noting that there is not a single NHP model, but rather a series of models. The best model may not be the same for all questions. NHP models require researchers to think.
The fact remains that macaques are not human and SIV is not HIV, thus there is no perfect model for the human situation. However, the NHP model is the only place where researchers can study in vivo/host interactions in a system similar to HIV/human. In addition, highly invasive sampling is possible with NHPs. The NHP experiment involves a vaccine platform, vaccine antigen, choice of animals, and a challenge virus system that involves virus that is either homologous or heterologous to the vaccine antigen and various routes of challenge (intravenous or mucosal, the latter more recently involving repeated administration).
Dr. Schultz briefly reviewed the challenge stocks. A diagram of the SIVmac251 challenge stocks illustrated how they had come about. The “evolution” of the challenge stocks would inform some of the meeting’s other presentations.
There are a number of pros and cons in terms of the repetitive versus single mucosal challenges. Among the benefits of the repetitive challenges are that they model the human condition and can attain 100 percent infection in controls. However, immune responses wane during challenge, the repetitive challenges are logistically onerous and consume more of the challenge stocks, and statisticians insist that these studies require a larger number of subjects, affecting costs and logistics. Nonetheless, if the money and resources are available, the repetitive challenges may be more informative.
Dr. Schultz believed the meeting would involve a lot of discussion of transmitted viruses, including the following questions:
Investigators have agreed that NHP studies should focus on mucosal transmission. It will be important to determine whether or not rectal and vaginal challenges are different with respect to vaccine protection. Rectal challenge is more often done because male macaques are easier to obtain. Is it clear how best to employ the transmitted viruses and challenge stocks? Is it true that vaccine groups of 25 are necessary? How can investigators manage the demand for the resource, and how can they provide consistent challenge stocks? Much of the session was to center on the E660 virus as this is the virus that has been used in several recent studies where protection from acquisition was demonstrated.
In a brief question and answer session, Dr. Berzofsky said that another problem with repetitive challenges is that every repetitive challenge is a boost to the vaccine. This muddles the issue of what is truly at work, rather than being a straightforward challenge involving the vaccine versus the virus.
Dr. Brandon Keele introduced the topic by stating that mucosal transmission of HIV is inefficient, and most people are infected with a single variant. His research team uses a single genome application of early samples, which allows for the identification and enumeration of transmitted founder viruses.
Non-human primates are essential for HIV modeling. Among the advantages for transmission studies is that the challenge stock can be genetically defined, the dose is controlled, the infection route and time of infection are known, and early blood and tissue sampling can occur.
Dr. Keele proposed the question of whether the current mucosal challenge system is a good model. He showed a model of mucosal infection. The questions raised concern the reproductive ratio of viruses before they become systemic, whether there are lethal RT errors, and how much is blocked by mucous. However, investigators do know that there are RT mutations over time in the virus. The sequences can be taken back to the founder virus.
Dr. Keele concluded that intrapatient diversity is roughly equivalent to that of either SIVmac251 or E660 challenge stocks. The overarching conclusion is that mucosal infections can replicate many features of HIV infection. The end result will be a challenge system that closely replicates human infection.
In regard to founder viruses, Dr. Keele was asked if this is a stochastic phenomenon in which a virus can get across or if it is something special about the specific virus. He replied that that was still being determined. It was added that there has been tremendous progress, but investigators have to be humble and acknowledge their limitations. What happens is still a black box. In conceptualizing and testing a model, researchers are not yet at the point where they can characterize it and understand how to take it to humans.
Dr. Dan Barouch began by discussing the earliest virologic events following intrarectal SIV infection in rhesus monkeys. Defining the earliest virologic events following HIV-1 transmission may be critical for the design of vaccine strategies that block acquisition of HIV-1 infection. A prophylactic vaccine must block infecting viruses in the mucosa to prevent systemic infection. In particular, the length of the eclipse phase and the number of transmitted virus variants are critical virologic parameters that define the window in which a prophylactic vaccine must act.
The aim of his study was to assess the impact of the virus inoculum dose on key virologic parameters following intrarectal SIV challenge. The study used 24 rhesus macaques divided evenly into four groups of six. The animals were inoculated once with dilutions of 1:1, 1:10, 1:100, or 1:1000 of the SIVmac251 challenge stock. They were bled on Days 0, 1, 2, 4, 7, 10, 14, 21, and 28 for fine resolution analysis of the eclipse and acute phases.
It was first observed that low-dose SIV infection reduces infectivity but does not impact SIV RNA levels, as measured on Day 21. For the dilutions, the eclipse phase occurred at an average of 4 days for 1:1 and 1:10, 7 days for 1:100, and 8.5 days for 1:1000.
When looking at transmitted founder (T/F) viruses, the investigators found that the 1:1 and 1:10 solutions had more than 10 T/F viruses and the 1:100 dilution had a median of two T/F viruses. Two animals infected with the 1:1000 dilution became infected with a single founder variant. These data indicate that there are significantly fewer T/F virus variants in lower dose groups when compared with the higher dose groups. Lower dose groups also had lower levels of innate serum cytokine and chemokine.
In discussing conclusions, it is important to keep in mind the overarching question: Is this challenge model relevant to human infections? In that context, it is also important to note the study conclusion that low-dose intra-rectal SIV infection resulted in a longer eclipse phase, fewer T/F virus variants, and reduced innate immune activation as compared with high-dose intrarectal SIV infection. These parameters may critically impact the capacity of a vaccine to block acquisition of infection, because the eclipse period defines the window of time in which vaccine-elicited immune responses must act, and the number of T/F virus variants defines the diversity of viruses that must be blocked. It therefore is reasonable to conclude that it would be considerably easier for a vaccine to block a single transmitted virus in a longer timeframe than a swarm of genetically diverse viruses in a shorter timeframe.
It is possible that high-risk HIV transmission in humans is characterized by increased frequency and higher dose per exposure. The data suggest a mechanism by which it might be easier for a vaccine to protect against low-risk as compared to high-risk HIV transmission. Dr. Barouch cautioned that this is speculation. Nonetheless, the study findings might have implications in the design of HIV-1 vaccine efficacy studies for humans.
Dr. Barouch concluded with a brief discussion of an ongoing study of the earliest mucosal events following intra-rectal SIV infection in rhesus monkeys. Using 12 monkeys in two groups of six, with staggered biopsies and bleeds on Days 0, 2, 4, 7, 10, 14, 21, and 28, the investigators have seen preliminary data indicating that the early eclipse-phase mucosal gag-specific CD8+ T lymphocyte responses may precede systemic responses. Important events occur at the mucosal site of transmission during the eclipse phase of infection and are not well understood. The local eclipse-phase immune response could be exploited by vaccine strategies to contain or block local infection following mucosal virus exposure.
Dr. Nancy Wilson reported on macaque studies testing a DNA prime, Ad5 boost strategy with vaccine inserts that encoded all SIVmac239 proteins except env. The study team had previously used SIVmac239 challenge, where they saw some protection. However, they wanted a more relevant model this time, and therefore used E660. The genetic differences between 239 and E660 appear similar to the differences between intraclade HIV infection. Animals were selected to be Mamu-A*02 positive so that investigators could take advantage of epitopes and tetramers associated with this allele to follow development of immune responses. At 61 weeks, both vaccinated group and control group animals were challenged with a low dose of E660 five times. If the animals were not infected after five challenges, the amount was increased until they were infected. There were good vaccine-induced T-cell responses, and all animals had responses to at least one epitope. There were some peptides to which all of the animals responded.
The investigators observed control of both acute and chronic phase vRNA however the vaccine did not provide any protection against acquisition of infection. The mucosal challenge resulted in only a few variants transmitted – between one and four. There was no statistical difference in the number of variants transmitted between vaccinees (avg 1.6) and naïve animals (avg 2.3). The number of variants transmitted was similar to human mucosal transmission. Neither peak viremia nor chronic viremia correlated to the number of variants.
There was no pattern to the viral loads. About one to four viruses crossed the mucosal barrier, and they were not necessarily related to each other. The same was true with the control animals.
Two years after the challenges, five of the eight control animals had died of SAIDS, but only two of the eight vaccinees had succumbed to SAIDS. All of the surviving vaccinees have viral loads that are either undetectable or near the limit of what can be detected, while two of the three surviving control animals control viremia. Some animals seem to be able to control E660 without vaccination.
There were no NAbs to E660 seen at 8 weeks post-infection in any of these animals.
In summary, SIVsmE660 as a heterologous mucosal challenge may simulate human transmission. About one-third of the animals spontaneously controlled in the E660 challenge, but the difference between the vaccine-induced control and the spontaneous control is obvious: the former have virtually no detectable load, while the others do have a detectable load.
SIV239 DNA/MVA Vaccine with Repeat E660 Challenge
Dr. Harriet Robinson discussed her most recent vaccine study utilizing a repeat E660 challenge. Her vaccines contained SIVmac239 inserts and consisted of DNA, MVA, and DNA expressing GM-CSF. (D=DNA alone, Dg=DNA + DNA GM-CSF, M=MVA). The immunization regimen was at 8-week intervals using eight rhesus monkeys per group. All were A*01 negative, with one B*08 and one B*17 positive per group. Group 1 received two doses of DNA followed by two doses of MVA (DDMM), group 2 was similar but received DNA GM-CSF with the DNA immunizations (DgDgMM), and group 3 received three doses of MVA alone (MMM).
The investigators found that the DNA-primed groups had the highest CD4 T-cells, while all groups had similar CD8+ T-cells. The MVA-only group had the highest binding antibodies and avidity. The best NAbs were associated with GM-CSF-adjuvanted DNA. Finally, MMM and DgDgMM presented the best anti-env rectal IgA.
The study involved 12 intrarectal SIVE660 weekly challenges with a dose of 1.8x107 RNA copies (5x103 TCID50). The number of infected monkeys at the 12-week mark were six out of eight for both the DDMM and the MMM, while only two of the seven DgDgMM vaccinees were infected at that point. All controls were infected. The investigators are adding more monkeys to the DDMM and MMM groups to get more power.
The analysis indicated that there were factors that did not correlate for protection: CD4 and CD8 T cells, mucosal IgA, and NAb for E660.11. There was a mild correlation with avidity.
What the investigators have learned from the repeat E660 challenge is that MMM is better or at least as good as DDMM at preventing acquisition. They also found that co-expressed GM-CSF in the DNA prime is a strong adjuvant for prevention of infection.
Dr. Rama Amara spoke about CD40L DNA as an adjuvant to enhance protection of mucosal surfaces against HIV transmission. In his NHP study he tested the same DNA vaccine reported by Dr. Robinson, expressing SIV239 genes, versus the DNA vaccine co-expressing CD40L along with the SIV239 genes, both boosted with MVA expressing SIV239 genes. There were 3 arms; 15 control animals, 8 receiving DDMM and 12 receiving DD(with CD40L)MM. Animals were challenged 20 weeks following the last immunization with repeated doses of SIVE660 IR. All controls were infected following 11 exposures, whereas 2/8 remained uninfected in the DDMM arm and 7/12 uninfected in the DD(CD40L)MM arm. Protection did not correlate with IFNgamma+ CD8 or CD4 T cells, no correlation with vaccine-elicited binding antibodies or neutralizing antibodies. There was an inverse correlation observed with avidity against native form of E660 Env. It was also observed that CD40L DNA blunted the virus peak load in animals that did become infected.
Dr. John Rose discussed his research on VSV/FSV. The investigation had three goals:
The research team tested VSV and SFV vectors expressing SIVE660 Env and Gag proteins. The investigators used six rhesus monkeys each in two vaccine groups. The first vaccine group received VSV G(NJ) on day 0; SFV replicon on day 49; VSV on day 112; and the rectal challenge of SIVsmE660 high dose (4000 TCID50) on Day 147. The second vaccine group received same vectors as the first vaccine group as well as VSV expressing rGM-CSF (10 percent of total pfu) in the prime only. A control group of six rhesus monkeys received vectors of an irrelevant antigen.
All six controls became infected, with peak viral loads between 106 and 2 x 108. The CD4 cells did not change much until the monkeys got AIDS, but the high-load animals were losing all CD4s in the gut. In the vaccine group, four animals showed no sign of the virus. The infected vaccinees show four log reduction in peak viral load. Inclusion of a virus expressing rGM-CSF in prime inhibited protection. On average, these animals have a reduced load compared to the controls.
No vaccine group animals had detectable serum NAbs to the E660 challenge swarm before or after the challenge. They also lacked NAb to a Nab-resistant SIV E660 virus envelope (CR54-PK-2A5) at the time of the challenge. All of the vaccine group animals developed NAb to Nab-sensitive E660 env pseudotype (smE660.11). They also had NAb to the E660 env used in the vaccination.
The team also found that the pre-challenge cellular immune responses did not correlate with protection from infection.
In summary, the VSV prime, SFV-G propagating replicon boost is a potent vector combination that deserves additional study. The research team observed significant protection in the vaccine groups. The non-protected animals in the vaccine group showed very low peak viral loads and their challenge virus load quickly dropped to below detectable levels.
The mechanism of protection is not yet clear. It may not be NAb because none of the protected animals had measurable NAb to the E660 challenge swarm. The study team will look at local gut T-cell responses and ADCC as alternatives, but the current hypothesis is that some type of mucosal antibody protects against infection in some of the animals. The study team hopes to repeat the challenge in order to examine the longevity of protection and learn what is required for protection.
Dr. John Mascola spoke for his colleague, Dr. Norman Letvin, who was unable to attend the meeting. Their study team has developed and worked with a low-dose challenge model for several years now. Recently, the team initiated three related SIV challenge studies, and also did some modeling.
For all studies, there was an SIV low-dose rectal challenge with two arms of 20-25 monkeys per group, with gene inserts of SIVmac239 Gag/Pol and gp140 Env. The vaccine group was immunized with DNA (4mg), I.M. at Weeks 0, 4, and 8, with rAd5 (1 x1011) particles i.m. at Week 26. About 4 months later, they were rectally challenged with either E660 or SIVmac251.
The first of the three studies involved two groups of 20 animals each, all Mamu A*01/B*08/B*17 allele negative, with homologous SIVmac251 as the challenge. The second had two groups of 25 animals each, also Mamu A*01/B*08/B*17 allele negative, and challenged with heterologous E660. The third study comprised two groups of 20 Mamu A*01 positive animals each, challenged with the heterologous E660.
For the SIVmac251 study, all of the animals were infected over five challenges. The difference in peak virus load between vaccinees and controls was statistically significant, with the controls between 107 and 108 and the vaccine group below 107. For the first SIVE660 challenge study 22 of 25 controls became infected, but 13of the vaccine group were protected from acquisitionfollowing 18 challenges. There was no statistically significant difference in the peak viral loads in animals that became infected.
The third study, with Mamu A*01 positive animals, presented essentially the same vaccine effect of about 50 percent, which again was statistically significant. In this study there IS a statistically significant difference in virus load between vaccinees that became infected and controls. When combining the two E660 studies, the vaccine group had an infection rate of 51 percent. Ongoing analyses are focusing on three factors:
The team has follow-up plans with RV144. They hope to use the existing E660 challenge model to evaluate the AVLAC/protein platform in a manner similar to the RV144 HIV-1 study.
DiscussionDuring discussion Dr. Mascola acknowledged that their group has also shown that vaccination of macaques with a DNA vaccine alone that expresses SIV env showed some degree of protection from acquisition in the SIVE660 challenge model.
History and properties of SIVsmE660
Dr. Vanessa Hirsch began her discussion with the derivation of E543 and E660 In 1992, the original virus was isolated from CEMx174 stock, which was then developed in 1996 into a number of PBMC stocks. E543 is a clone, and E660 is a quasi-species. Both are macrophage and T-cell tropic, with CD4-dependent entry. Both are relatively resistant to neutralization. They are pathogenic in rhesus and pigtail macaques, and display variable plasma viremia. Memory CD4 loss correlates with the virus load. The survival curves for E236, E543, and E660 are not significantly different.
E660 is the most sensitive in terms of resistance to neutralization. Virus loads in animals infected with either E543 or E660 are more variable when compared to SIV239 infection.. A study conducted in 2000 that examined the susceptibility of rhesus macaques to E543 found that two of six animals were extremely susceptible, while one was barely susceptible, and this was a stable property. Susceptibility correlated with viremia. Previous studies showed considerable heterogeneity in macaques infected with E660.
Recently Dr, Hirsch and collaborators have shown that control of virus load in macaques to these viruses is due to TRIM5α genotype. Dr. Hirsch showed the distribution of genotypes in 43 E543-infected rhesus monkeys, broken down by Trim5. There is a significant difference in virus replication and in plasma viral RNA in TFP/TFP animals. Trim5α has modest effects on SIVmac239. So the question comes up as to why mac239 is more resistant to Trim5 than SIVsm isolates. SIVmac239 has unusual substitutions in CA.
In summary, E660 and E543 are similar in pathogenicity to SIV239/251. There is more variability in viremia, especially early. The variation can be explained mainly by Trim5α allelic polymorphisms in rhesus. It appears to be due to differences in the cypA binding loop of CA of SIVsm versus mac.
Questions remain: Is it possible to define a more Trim5-resistant E543/E660. Does Trim5 restriction affect transmissibility? Could susceptibility to Trim5 restriction be why it is easier to observe protection against E660 in the low-dose i.r. model?
DiscussionAcquisition should take Trim5 into account, as it prolongs the eclipse phase. If there is a partly effective vaccine and the investigator adds an allele that delays eclipse and peak, it may be possible to the protect. Dr. Hirsch thought this seemed like an obvious feature to examine.
SIVsmE660 challenge stock population analysis
Dr. George Shaw addressed analyses of the challenge stocks. For challenge stocks, the rationale is that in examining the inoculums, researchers might see properties that distinguish the virus that is actually transmitted from the other viruses. Some would be fit, some would be less fit, and some would be dead.
Dr. Shaw explained that low dose i.r. and IVAG infection of macaques by SIV recapitulates key features of human HIV-1 mucosal transmission. Like HIV-1, early SIV diversification is star-like and conforms to a model of random virus evolution with sequences coalescing to transmitted/founder virus(es) in a timeframe consistent with model projections. Also like HIV-1, low-dose mucosal SIV inoculation generally results in one or few viruses establishing productive clinical infection. There is a similar rapid selection in SIV for CTL escape mutations, and a substantial SIV mucosal transmission bottleneck of 2,000 to 20,000-fold. Novel molecular clones are needed of genetically-diverse SIVsm strains that exhibit eclipse phase viral dynamics, set-point viremia, and neutralization susceptibility profiles comparable to HIV-1. Such viruses can potentially serve as new, heterologous challenge stocks for preclinical SIV-macaque vaccine trials.
Neutralization analysis of SIVsmE660 and variants
Dr. David Montefiori presented on his lab’s work on neutralization analysis of HIV and SIV. The main assays used at Duke for NAbs include: CEMx174;M7-Luc); TZM-bl; and PBMC.
Montefiori and others have found that neutralization results obtained with PBMC-based assays vary considerably with different donor PBMCs. His lab moved more toward the use of cell line based assays to obtain more consistent results. With the advent of the TZM-BL assay, the study team began cloning functional env genes about 6 years ago. The team made a number of pseudoviruses that contained tier 2 and 3 phenotypes. They made a clone from E660 which has a tier 1A profile, as well as tier 3 clones from SIVagm and SIVsab.
In looking at the neutralization of uncloned SIVsmE660 by sera from Rh628 (SIVsmE660 infection), the study team found that with the uncloned virus, it is super-sensitive to neutralization, especially assayed in M7-LUC cells. The team is now trying to determine the properties behind this.
Dr. Montefiori showed the neutralization profiles of SIVsmE660 env-pseudotyped viruses and full-length molecular clones, which are clones they make from the 1996 E660 stock. There is a great deal of variability, and all but one have tier 1 neutralization phenotypes.
In TZM-bl cells, there is a big difference in neut profile between the neutralization of E660.11 (tier 1) and the neutralization of E660/CR54-PK-2A5 (tier 2). A comparison of E543 and E660 clones shows a huge difference in neutralization sensitivity depending on the assay used.
So is the difference due to the cells or the virus? A number of variables can affect what is measured. It is still not clear which is the exact best way to assess NAbs, but the study team hopes to gain some insights into the variables.
Intravenous vs mucosal challenge with SIVsmE660
Dr. Matthew Reynolds discussed studies his group has carried out with live attenuated vaccines.. Live-attenuated SIV vaccines are the gold standard in HIV challenge and protection. Dr. Reynolds’ study team asked if SIVmac239∆nef-induced immune responses can control heterologous SIV replication.
To this end, the study team expanded on an earlier study and investigated 20 animals in a heterologous challenge, with 10 vaccinees and 10 naïve controls. Each group included two animals that expressed either Mamu-A*01, -A*02, -A*11, -B*08, or B*17. The team used the attenuated strain SIVmac239∆nef and challenged the study animals, along with the naïve controls, 6 months later intravenously with 100 TCID50 SIVsmE660.
The naive controls infected with SIV E660 had viral loads ranging from 10^5 to 10^7 throughout the course of infection, and the virus levels were relatively consistent in this group of animals. This group did not show variability of virus levels, which can be a concern in challenge studies involving a swarm.
The vaccinated animals demonstrated a level of control that was not seen in the naïve controls: 5 of 10 vaccinees were controlling. At 11 months, that number was four. At that point, the study team began to dissect the data. Two of the animals expressed Mamu-B*08, an MHC class I that has previously been linked to spontaneous control of SIVmac239. The vaccinated animals as a group were able to significantly reduce their viral load, which was quite low. Three of the four had undetectable levels at Week 2, which is typically peak. However, they all had gradually increasing viral loads.
However, the geometric mean gradually increased over time due to an increase in viremia in the B*17 animals. They had a gradual increase in plasma virus levels starting around 6 weeks post-infection. What caused the B*17 animals to lose control? After sequencing, they found that recombination had occurred between the challenge and the vaccine.
The team learned a number of important lessons from the i.v. E660 challenge. They found that naïve controls had consistent plasma virus concentrations between 10^5 and 10^7 throughout the course of the study. The SIVmac239∆nef vaccinated macaques significantly reduced plasma virus levels in comparison to the naïve controls. They also learned that control of acute plasma virus replication is associated with expression of Mamu-B*08 or Mamu-B*17. And recombination between vaccine and challenge strains may contribute to the loss of control of virus replication during the chronic phase of infection.
Questions remained: Is SIVmac239∆nef vaccination more effective against a low-dose mucosal challenge with a heterologous virus? And is the vaccine still effective in the macaques that are not expressing the protective MHC class I alleles Mamu-A*01, -B*08, or -B*17?
To address these issues, the study team carried out a study with low dose heterologous E660 challenge, waiting 6 months following immunization to challenge. Six of 8 naïve controls were infected after two challenges, and another at the fourth challenge. Among the vaccines, three were infected on the fourth challenge, one was infected on the fifth challenge, another was infected on the sixth challenge, and three were never infected at all across 10 challenges. The Kaplan- Meier analysis indicated that the vaccinated animals significantly delayed virus acquisition. Among the infected animals, the vaccinated ones were able to delay peak plasma virus levels, and four of the five vaccinees dropped down to a low viral load. Recombination could be playing a role in this.
The study concluded that SIVmac239∆nef vaccination significantly reduced the rate of acquisition of SIVsmE660 infection after repeated mucosal challenges and the induced immune responses significantly reduced peak acute SIVsmE660 replication. This suggests that a well-designed HIV vaccine might both reduce the rate of acquisition and control viral replication.
Challenge with mismatched env
Dr. Ronald Desrosier discussed key components of the protective immune response with live attenuated SIV. The study team tested 30 animals in six arms:
The study team looked at the NAbs in sera taken on the day of the challenge. None of the 12 animals receiving SIV239∆nef had neutralizing activity against 239.. However, a few of the animals receiving SIV239∆nef/EnvE543 could neutralize E543. The 12 animals that received the homologous challenge all had apparent sterilizing immunity, as did 10 of the 12 in the heterologous challenge group.
In conclusion, it appears that neutralizing antibodies play little role in protection afforded by live attenuated SIVΔnef when the challenge occurs at 22 weeks post immunization.
The group discussed the issue of how to define a “low dose” since there is no consistency between labs as to the amount of virus used in multiple exposure mucosal challenges. There seemed to be consensus that mucosal challenges should be done with the amount of virus needed to give infection of 50% of animals with each challenge. Therefore, rather than referring to challenges as “high dose” versus “low dose” investigators should refer to challenges as done with “100% single infectious dose” or “50% infectious dose.” The group also discussed the number of animals needed/arm in such studies. Several recent statistical papers have suggested 20-25 animals/arm in two-arm placebo controlled studies. The actual number of animals needed depends on a number of factors including the estimated degree of efficacy of the test vaccine.
STEP trial follow-up NHP
Dr. Reynolds began by presenting some background about the STEP trial. This was a Phase IIb study to test the ability of the Merck Ad5 trivalent vaccine to decrease HIV acquisition or reduce set-point viral loads. The vaccine consisted of three doses of 1:1:1 mixture of 3 Ad5 expressing HIV-1 gag/pol/nef, and participants were at-risk individuals.
The immunizations were given on Weeks 0, 4, and 26, with immunogenicity analysis at Weeks 8 and 30. Investigators found that 77 percent of participants had detectable responses by IFN-γ ELISPOT: 62 percent recognized two of three proteins and 45 percent recognized all three proteins. In addition, 42 percent had detectable HIV-specific CD4+ T-cell responses by ICS, and 72 percent had detectable HIV-specific CD8+ T cell responses by ICS.
Dr. Reynolds and his group designed an NHP study in an attempt to duplicate the STEP trial results. The hypothesis for the NHP study was that Ad5 gag/pol/nef would not result in broad cellular immune responses, and that the vaccine would fail to protect the animals against a heterologous virus challenge. In the study, which involved eight macaques, Dr. Reynolds has tried to match the STEP investigation as closely as possible, planning doses at Weeks 0, 4, and 26. Immunogenicity analyses were done at Weeks 3, 5, and 8. The study is currently at Week 14.
Thus far, the investigators have concluded that the cellular immune response induced by the STEP trial regimen may be broader than originally thought. They have observed that the regimen induces broad CD4+ T-cell responses in macaques, with a steep decline in magnitude and breadth of responses between Weeks 5 and 8.
There are continuing experiments in the works. These include further definition of SIV-specific responses by ICS; in vitro culture CD4+ T cells to confirm positive responses; monitoring the breadth and magnitude of SIV-specific responses after the third Ad5 boost; assessment of the levels of anti-Ad5 antibodies in the vaccinated animals; and challenges with repeated, low doses of SIVsmE660.
Planned RV144 follow-up NHP studies
Dr. Bradac presented two slides in order to discuss possible NHP protocols to follow-up on the RV144 human efficacy trial. The Military HIV Research Program has put together a working group for designing such a study. Other groups are also collaborating to conduct similar studies. There is discussion about what the env component of the vaccine should be and whether the adjuvant used should match what was used in RV144 (alum), since alum has not been shown to be optimal in NHPs. There are also questions about what the challenge virus should be and which route of challenge to use. There will be more about these plans in the near future.
Challenge stock viruses in progress
Dr. Nancy Miller discussed SIVmac251 and E660 challenge stocks in preparation. Repeated low-dose mucosal challenge models have increased the need for well-characterized and in vivotitered SIV virus stocks produced in large enough quantities so that the stocks can be available for challenges across several vaccine studies. It is more economical to do only one set of titrations on one large virus stock, and use of the same challenge virus in multiple vaccine studies makes it easier to compare results.
A large stock of SIVmac251 has been prepared. It is less diverse than some others, but this will help determine whether the cell in which the virus is grown matters. Dr. Desrosiers has agreed to prepare the large stock, and will do single infectious dose titrations. The new SIVmac251 challenge stock will essentially be the same as the Desrosiers 2006 stock. The E660 stock has also been expanded and will be fully characterized. Both viruses replicate well in vivo.
Dr. Ruth Ruprecht spoke about the development of Clade C SHIVs. SHIV-1157ipd3N4 is an infectious molecular clone, pathogenic in rhesus and pigtailed macaques, and tier 2 neutralization phenotype.. The second virus is SHIV-1157ipEL-p, a tier 1 biological isolate. The third virus is SHIV-2873Nip, a tier 2 pathogenic biological isolate.
Three of five animals given SHIV-1157ipd3N4 progressed to AIDS and were consistently viremic. The first animal that ever got the virus is a progressor and developed AIDS at week 140.
The first monkey to receive the infectious parental clone (SHIV-1157i) developed high titers of NAbs against the early virus. The late infectious molecular clone, SHIV-1157ipd3N4, remained exclusively R5 tropic and was found to be a neutralization-escape virus. Although SHIV-1157ipd3N4 is more difficult to neutralize than the early virus, it is not neutralization resistant. So Dr. Ruprecht’s team decided to develop a new challenge virus, SHIV-1157ipEL-p. This is an engineered new tier 1, R5 Clade C SHIV that transmits rapidly. There are no long-term studies yet.
The meeting was adjourned for the day at 5:28 p.m.
Dr. Nancy Haigwood called the meeting to order at 8:44 a.m. She gave a brief summary of the previous day’s discussion.
Drs. Brandon Keele and Dan Barouch addressed SIV stock analyses:
Regarding the transmitted variants, it looks like investigators are closer to getting a sense of the mucosal route and breakpoint where the virus is transmitted. A new finding was the evidence of mucosal responses earlier than serum appearance.
Five speakers addressed different experiments on mucosal challenges with E660:
Next was a discussion of SIVsmE660:
The next speakers were Drs. Reynolds and Desrosiers, who discussed heterologous challenges after SIVmacΔnef:
Finally, there were presentations on comparison work and reagents:
Dr. Nelson Michael talked about RV144 in terms of immunogenicity, vaccine effect, efficacy, correlates, lessons learned, and the Phase III trial in Thailand. He began by showing a comparison between RV135 and RV144. Immunogenicity was comparable in all measurable parameters except ADCC where levels measured were much lower in RV144.This result could be methodological. Many different ADCC assays are about to be launched by different investigators, who will pay more attention to methodology.
The vaccine effect appears to be transient, although the study was not powered to ask this question. Nonetheless, the data indicate a vaccine effect of 60 percent at 12 months, and 36 percent at 24 and 30 months.
In discussing the search for correlates, Dr. Michael noted that 75 percent of the Military HIV Research Program (MHRP) post-trial project funding came from NIAID, with the rest from the Army. It was therefore decided to throw the project open to the field, not just for transparency but also for efficiency.
After launching a website for open solicitation, MHRP received lots of proposals and have approved 32 so far. The biggest stumbling block were the MTAs, half of which had been executed at the time of this meeting. Altogether, there were 32 approved proposals from 20 institutions and 35 investigators. Many investigators are working on multiple projects.
Dr. Michael presented a list of lessons learned:
MHRP would like to follow up with the following projects:
The guiding principles are to minimize the number of variables that change in the next Phase IIb study. The key variables for the vaccines and study population include the nature and basic schedule of the prime and boost vaccines; incidence; exposure route; subtype and complexity of circulating HIV variants; and geographical location confounded with genetic background and co-morbid conditions.
There are also pharmaceutical imperatives, such as a consistent platform. Scientists need to consider this, and also keep in mind the need for a reliable pharmaceutical partner appropriate for what the specific research.
RV305 is a late boost target of opportunity. The primary point is to characterize the innate, humoral, and cellular immune responses following late boosts for each strategy tested in both the systemic and mucosal compartments. Can there be additional boosts to some vaccines that have been out for 7 years? This research must wait until it is known which current vaccines can be re-qualified. The primary purposes of an RV306 immunogenicity trial are to characterize systemic and mucosal immunity of the ALVAC/AIDSVAX combination or AIDSVAX alone or ALVAC alone; and to characterize the innate, humoral, and cellular immune responses after late boosting with ALVAC/AIDSVAX or AIDSVAX alone or ALVAC. These are some of the things investigators might wish they had done with RV144.
The best geographic areas to work in for the Phase IIb/III studies appear to be Asia, for low-incidence heterosexual and high-incidence MSM transmission, and southern Africa for high-incidence heterosexual transmission. A proposed reference study on low-incidence heterosexual transmission would be large, with 27,000 subjects and would cost $100+ million. In an Asian MSM study, investigators could keep some variables the same. The expensive Thai heterosexual study would be a pathway to licensure, as would the Thai MSM group. The latter would also provide a bridge to countries with such as the Americas, Europe, and Australia. The African study would be a pathway to a pivotal study.
Cost is always an issue and must be considered. The Phase IIb trials in high-incidence populations are efficient, but they also represent more variable changes from RV144. The Phase IIb trial in low-incidence Thais is costly and cumbersome, but presents less variability from RV144.
Dr. Michael concluded by saying that RV 144 suggests an early, non-durable protective effect, and the hunt for correlates has begun. Extension to new routes of infection, geography, and subtypes is reasonable. The next phase IIb/III trials of pox + protein must balance the risk of over-extension from RV 144 with the efficient development of a vaccine relevant to global health, and should not completely absorb the field’s capability to test other promising candidates until 2017.
Dr. Michael Pensiero gave a brief update on the DAIDS/VRP HIV vaccine pipeline, focusing on activities that DAIDS has funded. Funding comes from grants, contracts, and pre-clinical master contract (PCMC). Near-term vaccine candidates/reagents in progress include the following:
Dr. Pensiero quickly reviewed a number of individual project.s AAV vectors have been developed to determine whether different serotype AAV vectors can be effective prime-boost regimens for T cell responses.Other projects include VSV vectors which asks if different serotype VSV vectors can be effective prime-boost regimens for T cell responses.
For the Poly I:C/L:C adjuvant + Protein study, investigators want to determine whether poly I:C/L:C adjuvanted protein is capable of inducing both CD4+ and CD8+ T cell responses. The adenovirus vectors based on alternative serotypes project aims to develop Ad vectors that avoid high baseline vector-specific NAbs that can be combined in a prime-boost regimen.
The Replication Competent Ad4-Gag project will address whether mucosal delivery of replication competent Ad vectors can induce strong mucosal responses. Investigators will conduct the DNA+ GM-CSF adjuvant/MVA study in order to ask if adjuvanting DNA with GM-CSF followed by MVA can increase Ab avidity.
Comparative Centralized HIV Env is Phase I study to compare two in silico centralized approaches as a means of enhancing breadth of T cell responses. DNA + electroporation is a comparison of high concentration PennVax-GP adjuvanted DNA delivered by IM/EP vs ID/EP.
Long-term vaccine products studies awarded include:
Dr. James Bradac discussed the preclinical portfolio at DAIDS. He noted that at the previous meeting, Dr. Peggy Johnston of DAIDS provided an update on vaccine portfolio from the financial standpoint. His presentation focused on the science.
Standard grant mechanisms include R01 Research Project Grants, R43 Small Business Innovation Research (SBIR), and R21 Exploratory Developmental Research. The DAIDS solicited programs fall into the areas of innovation, like the Phased Innovation Awards (PIA) (R21); basic vaccine research, such as Basic HIV Vaccine Discovery (BVD) (R01); specific research areas, such as B-cell Immunology for Protective HIV-1 Vaccines (R21, U01); and translational research, an example of which is Integrated Preclinical/Clinical AIDS Vaccine Development (IPCAVD) (U19). The examples are all from recent solicited grant programs, and some are standard programs that are ongoing.
There are also research support contracts, such as the HIV databases program. DAIDS created portfolio buckets to arrange the grants. There are 11 areas of emphasis: B-cell, antibody, Mab and env structure, env immunogen design, T-cell, vector, NHP, mucosal, adjuvant/immune modulator, innate, and host targets. Dr. Bradac gave examples of funded studies from each area of emphasis.
There are many grants, with about 60 targeted to B-cell and env. B-cell and T-cell have always received a lot of attention and funding. Areas that need strengthening include mucosal, adjuvant, and innate immunity, though there are other NIAID branches that sponsor this research.
DAIDS will continue its strong support for env/antibody research, including MPER and trimeric forms. The Division will also continue supporting research on T-cell immunity, especially new vectors, with much less emphasis on inserts. Finally, there are growing programs in innate, mucosal immunity, and the application of systems biology approaches.
Dr. Alan Fix discussed the current scientific priorities at VRP and NIAID regarding HIV vaccine research. Most protocols now include a mucosal assessment and innate response component. There are efforts to build on the RV144 trials by trying to identify correlates of immune protection, maximize induction of relevant responses for any correlates found, and conduct trials of improved candidates. NIAID also seeks to extend the research to additional populations, continue to explore other concepts in smaller clinical trials, and address basic questions through further research.
Dr. Fix gave some examples of NIAID-supported clinical trials, such as heterologous vectors (±DNA) alone or in combination, heterologous prime-boost regimens with shared vs heterologous epitopes, and others. He also discussed applying lessons from the STEP, in which investigators observed limited breadth with rAd5-Gag/Pol/Nef vaccine. Considerations include the number of epitopes recognized and needed to cover global diversity (breadth), and the number of epitope variants recognized and needed to limit T-cell escape (depth). Mosaic candidates are in development as a strategy to improve the breadth and depth. A number of proposals are moving forward.
At the previous AVRS meeting, in February 2010, several groups discussed the outlook for Phase IIb trials. The HIV Vaccine Trials Network (HVTN) is looking at NYVAC or ALVAC mosaic + gp120 boost, Ad26/35 mosaic +/- gp120 boost, and a passive immunoprophylaxis trial. The International AIDS Vaccine Initiative (IAVI) is also investigating Ad26/35, and the MHRP is studying Ad26/MVA mosaic and ALVAC/AIDSVAX.
Setting priorities for products is always a concern; there must be a way of focusing. Dr. Fix ran through the process employed at NIAID, which starts with developing a hypothesis and thoroughly characterizing response phenotypes in preclinical studies, then moves to evaluating safety and response phenotypes in Phase I and II trials. Should those trials either improve the correlate or provide sufficient information to address a new hypothesis regarding correlates, the Institute supports Phase IIb trials. Phase IIb trials producing efficacy greater than 50 percent are candidates for advancement to Phase III trials.
Dr. Fix showed a grid of VRP/DAIDS-supported research, noting that in the 2008, immediately following the STEP trial, there was a relative paucity of activity. Now, in 2010, many more trials are ramping up, with about 10 ready to start but not yet begun. Awards for NIAID’s six HIV/AIDS clinical trials network Leadership Groups are scheduled to expire in 2013, necessitating re-competition. Three awards will cover prevention, and three will be for treatment. Awards for the clinical trial units and clinical research sites are scheduled to expire in 2014. Re-competition considerations include expansion of the research focus, additional infectious diseases, and additional co-morbidities. NIAID wants to leverage resources across NIH to the maximum extent possible. Many sites face major challenges when activities drop off, so multiple streams of funding might help avoid the problems attendant to less activity, such as staff layoffs and ramp-up for the next study.
Reorganization can be difficult, so NIAID is developing new options for input, comments, and suggestions. These include meetings with individual networks, outreach PIs and the community, and a revised website. Clinical collaboration meetings are held to provide an informal forum for funding/implementing organizations engaged in preventive HIV vaccine clinical research, allowing them to share information about current/long-term clinical research priorities and plans, and to discuss areas for additional collaboration in scientific planning, including clinical trials and the use and development of reagents. There were two of these meetings in 2009, which helped NIAID go in the direction of more collaborations, closer work with FDA, and specific issues. NIAID is also interested in multiple modalities and ethical considerations of HIV research. There will be a small workshop on this.
The clinical collaboration meetings revealed that there are two parallel streams of organizations looking at the same thing: HVTN and IAVI are both, as noted earlier, studying Ad26 in Africa on the Mosaic platform. The result is a new emphasis on coordination and working together rather than in parallel.
The Early Stage Investigator (ESI) scholar award program gave out its first nine awards in two cycles. Another program focuses on rapid HIV point-of-care (POC) diagnostics in resource-limited settings, with the goal of developing rapid diagnostics at a low-cost, enabling users to distinguish between HIV-1 infection and vaccine-induced seroconversion, to identify acutely infected individuals, and to diagnose infants and initiate MTCT prevention. Three awards have been made for this program.
The Mucosal Immunity Working Group aims to: develop standardized protocols for mucosal sample collection, storage, and transportation for use in clinical trials; develop standardized assays to measure and characterize the major effector and memory mucosal immune responses in the GI and GU tracts; and provide standard operating procedures. To that end, there are now three subgroups, focused on gastrointestinal, genitor-urinary, and systems biology issues.
Dr. Birx said that it seems there could be some way to bring together those working downstream and upstream, and that maybe as NIAID is reaching out, they could discuss the continuum of prevention activities. It seems like there is a gap in the information. What holds the field back is that scientists are working individually with the network rather than as a group. This prevents looking at a cohort in an integrated way. There is a need to integrate across funding streams and possibly share funding. Each year, $800 million is going into unknowns. Coordination could strengthen the investigators and bolster impact evaluation.
Another meeting participant asked if there will be work on vaccines relevant to children and pregnant women. Dr. Fix said that in the re-competition, the criteria will be to cover the entire spectrum. There are obvious constraints in looking at children, but the intent has always been to move to all age groups. Mr. Snow asked whether the geographic focus of products, the frequent use of Clade B, and other factors are affecting the balance of research and the impact on future trials. Obviously, some work is done in the United States, but on the surface it appears that there is an exclusion. It was noted that the STEP study showed a small rate of infection of women. That is being addressed by two studies to identify high-risk women in the United States and the Caribbean. Mr. Snow asked how efficiency is maintained when early decisions based on Phase I studies conducted here with Clade B lead to products that are reengineered for Asia and Africa. Dr. Fix said that not all of these products are developed just for local populations, as investigators and funders want products for global populations. Some will be specific to certain populations, however. Dr. Johnston added that the goal is to maintain flexibility for later decision-making for both Clades and risk groups.
It was noted that, when talking about spending $105 million, it should be noted that $240 million was spent to license rotovirus, making $105 million in line with trial costs. In addition, there still seems to be a huge time issue from concept approval to execution. There should be a performance metric for this, as it should not take more than a month after IND approval to move forward. The failure to make vaccines has put DAIDS personnel in the position of planning for failure rather than success. One group will only do Phase IIb on the condition that they could manufacture for a Phase III trial soon thereafter. Phase IIb should be preparation for Phase III. The time to get started makes it all difficult. Dr. Fix said that there are increasing standards for moving things forward, but there are also issues like a centralized IRB. A commenter said that IRB issues should be dealt with while the IND is in process.
Dr. Haigwood asked those present to note their thoughts about the meeting.
Dr. Haigwood thanked everyone for a productive and interesting meeting.
The meeting was adjourned at 12:41. p.m.
Last Updated June 27, 2011
Last Reviewed June 20, 2011