The AIDS Vaccine Research Subcommittee (AVRS) met in public session on December 12, 2007, North Building of The William F. Bolger Center, 9600 Newbridge Drive, Potomac, MD.
AVRS members present: Eric Hunter (chair), James Bradac (executive secretary), Jay Berzofsky (ex officio), Susan Buchbinder, Lawrence Corel (ex officio), Kevin Fischer, Barton Haynes (ex officio), R. Paul Johnson, Jeffrey Lifson, Margaret Liu, Timothy Mastro (ex officio), M. Juliana McElrath, Nelson Michael (ex officio), Gary Nabel (ex officio), Louis Picker, Nina Russell, Jerald Sadoff, Bruce Walker, Jay Whitescarver (ex officio), and Ian A. Wilson.
Ad hoc consultants present: R. Gordon Douglas, John Moore, Stanley Plotkin, William Snow, and David Watkins.
Eric Hunter, Chairperson, opened the meeting at 8:30 a.m. and asked each committee members—both attending in-person and via telephone—to introduce him/her and share any perceived or real conflicts of interest.
Anthony Fauci, Director of NIAID, welcomed the committee members and others present. He shared with the group that he was attending the meeting for the purpose of listening and learning. Please do not let my presence, Dr. Fauci remarked, affect your willingness to share candid remarks.
Dr. Fauci noted the importance of the day’s discussion. He planned to listen with an ear for the potential impact of STEP trial findings on the NIAID research agenda, as well as broader issues on AIDS research.
Dr. Fauci highlighted the key themes that the group would be addressing: (1) the impact of the data and STEP on plans for PAVE-100, and (2) if STEP results indicate a failure of the product or a failure of the concept. NIAID, Dr. Fauci noted, will examine and reassess, as they should, vaccine research in light of learnings from the STEP program.
Peggy Johnston welcomed the group and, recognizing the large number of external participants, provided an overview of NIAID HIV vaccine R&D programs.
Dr. Johnston started with a review of the external oversight bodies—one of which is the AIDS Research Advisory Committee. As part of this committee, the AIDS Vaccine Research Working Group (AVRWG) meets about three times per year—in 2007, the group convened four meetings. NIAID HIV vaccine R&D programs bring major decisions to the AVRS and get input about advancing candidates and suitability.
At the January 30, 2007 meeting of the AVRWG, the committee conducted a technical assessment of the PAVE-100 protocol, including the scientific rationale for the proposed trial, supporting data, trial design, and proposed timeline. The AVRWG also provided comment on whether this was a good use of public funds. Some recommendations generated at this meeting were considering a phase 3 instead of 2b design, ensuring TRIAD CD8+ T-cell response against circulating clades has lower bound >30%, and using 9-mers to evaluate response to gag.
At today’s meeting, the subcommittee is discussing STEP trial implications and research priorities. Specific input is sought on four key areas:
Morning presentations reviewed STEP trial results, including primary efficacy outcomes, laboratory investigations done to help understand trial results, and the impact of the STEP trial results on HIV vaccine policy.
Michael Robertson reviewed key efficacy results from the STEP trial. He reminded participants that STEP was not a licensure trial—it was a concept trial designed to find out if the vaccine worked and so, when the placebo was favored over the vaccine, researchers averted the trial.
The primary efficacy hypothesis of STEP was that subjects with baseline Ad5 titers £200 who receive the MRKAd5 HIV-1 gag/pol/nef vaccine are less likely to acquire HIV-1 infection (compared to those who receive a placebo). And/or, the viral load endpoint among subjects with baseline Ad54 titers £200 who subsequently become HIV-1 infected—those who receive the MRKAd5 HIV-1 gag/pol/nef vaccine will have a smaller average viral load set-point compared to those who receive placebo.
Researchers averted the trial when, at first interim analysis, futility cutoffs were met for EACH endpoint coupled with trend forward placebo for EACH endpoint. Findings strongly suggest that the vaccine neither prevents HIV infection nor reduces the amount of virus in those who become infected. Given these results, the data and safety monitoring board (DSMB) recommended not administering further injections. At this time, volunteers were also encouraged to return for all protocol visits and tests so that the investigators might fully evaluate whether there is an increased risk of infection in vaccine recipients over time. Working closely with the trial oversight committee, appropriate steps and timing of release of trial results (to volunteers, investigators, those conducting related trials, relevant agencies, and the public) was determined.
Design and execution of the study allowed for timely assessment of both primary endpoints. There was also no evidence that vaccination prevented infection or lowered viral set point. Of note is the finding that there were more infections in vaccines compared to placebo recipients—this trend was more pronounced in participants with high baseline Ad5 titers.
Analysis of study randomization found good matching among vaccine and placebo groups within strata (males). However, the lower Ad5 group was more likely to be white, circumcised males from the U.S. and the high Ad5 group was more likely to be nonwhite, not circumcised males from outside the U.S.
Drilling into the data, the team evaluated risk behaviors at baseline and every 6 months thereafter—no real differences emerged in risk behaviors. In fact, risk declined over time for most behaviors due to encouragement and risk reduction counseling. On some measures, the researches found more risk in the lower Ad5 than the higher Ad5 group—indicating that the populations may have differences unrelated to the vaccine. The team found no evidence of differences in knowledge of treatment assignment or risk in high Ad5 vaccine versus placebo. If anything, there were just modest differences seen only in the low Ad5 group. Looking at HIV infections by circumcision (modified intent-to-treat (MITT) cases, males), there were no differences in the number of infections in vaccinees versus placebos; however, the data reveals that non-circumcised males favored the placebo over the vaccine.
Dr. Robertson opened the floor for general discussion among committee members. Most of the questions and comments focused on the finding that circumcism appeared to play a key role in participant response. However, Dr. Robertson noted, the team still needs to look at other factors.
Committee members offered feedback on the following queries:
The group agreed to continue the discussion after completion of the next presentation.
Juliana McElrath discussed the biological basis for the STEP vaccine efficacy results with the goal of addressing two key questions: (1) where things are and (2) what needs to be done.
Investigators compared STEP data with findings of a Phase 1 study collaboratively conducted by the HTVN and Merck laboratories.
The research team sought answers to the following questions:
In summary, Dr. McElrath noted that at week 30 (4 weeks after 3 doses of MRK Ad5), no difference is observed in blood CD4+ T-cell immune activation or CCR5 expression between Ad5 £200 vaccinees and placebos, and Ad5 >200 vaccinees and placebos. However, a greater number of activated blood CD4+ T-cells expressing CCR5 are observed in the Ad5 >200 group than in the Ad5 £200 group. These findings suggest the need for additional studies. Specifically, to:
Dr. McElrath opened the floor for discussion among committee members. The committee initially discussed Dr. McElrath’s findings and clarified the scope of her team’s work.
Lawrence Corey noted that his task was to transition the focus of the discussion from data to policy. Dr. Corey stated that the STEP trial has seismic implications for developing HIV vaccines, whether the construct is directed primarily at including neutralizing antibodies, T-cell responses to HIV or both—and an optimal HIV vaccine may not induce neutralizing antibodies and assume that T-cells are needed to achieve a successful HIV vaccine.
Dr. Corey noted that there is much more to learn about the reasons for the MRK Ad5 gag/pol/nef vaccine’s failure. He shared the following lessons learned:
For policy reasons, Dr. Corey recommended cutting the data differently. Specifically, for policy implementation, he suggested using Ad seronegative (<18) versus Ad seropositive (>18). While the “split” between Ad5 seronegative and Ad5 seropositive was not a pre-defined study strata, it offers a common and traditional split for defining immune status to Ad5 or other infectious pathogen as compared to Ad5<200 versus Ad5>200. In recommending proceeding with future Ad5 vector studies using the <18 cutoff, Dr. Corey pointed out that the <18 group showed equal acquisition rates and that at >18 the split between vaccine and placebo begins.
Another consideration for the field is adopting transfer experiments in an experimental model system when attempting to determine what type(s) of T-cell responses are required for reducing viremia or acquisition—there is a need for the field to adopt this classical approach for HIV-1 experiments.
During the next few years, and moving forward, an additional issue is defining what immune responses influence acquisition or post infection viremia in the context of human efficacy trials.
There is also a need, as a field, to look at tissue-specific response—and evaluate if immunization alters cell populations that are involved in acquisition of HIV-1. For example, does boosting of persons with prior Ad5 immunity create a significant increase in Ad5 CD-4 cells that persist for a prolonged period? If so, do we detect such cells in PBMC and/or are they present only in the tissues, especially in the rectal or gut mucosal? A corollary is that we might see an association between the magnitude and the degree of persistent activation and the number of injections with the vaccine or the dose (number of particles) administered with each dose of vaccine.
How to handle the demonstration of T-cell activation after immunization may not be easy. After the first dose of vaccine, everyone acquires anti-Ad5 immunity and very high levels after the second dose—even the Ad5< 18 are seropositive to Ad5 immunity after the second dose. If increased acquisition is due to adenovirus and specific CD-4 T-cell populations, such populations may be limited to a specific Ad5 protein. There may well be differences in “homing to mucosal surfaces” between those infected naturally versus those receiving intramuscular inoculation with a vector. Dr. Corey noted the need to be sanguine about the ability to answer these queries.
Lastly, policy must balance clinical with laboratory investigation, requiring clarification on the:
Clearly, defining the acquisition issue requires both laboratory and clinical investigation. Detailed clinical trials in humans—of both the adenovirus and non-adenovirus based vectors—need to answer the increased risk of acquisition in Ad5 seropositives. However, deciding what to move forward will be difficult. Animal models and human safety and immunogenicity trials will provide us with lots of things to do and measure; but if we do not conduct other efficacy trials, we cannot discover improvements in response or design.
We should, as a field, and the HVTN, as an organization, will continue to perform clinical trials of candidate HIV vaccines devoted to eliciting T-cell response to HIV. Evaluations of candidate vaccines will need to involve a greater emphasis on vector induced immunity and mucosal immune response. We will need to establish a process and program to perform such selective studies—likely resulting in a more detailed phase 2A program for most vaccine products.
Drugs “flush out” all the time; however, to recover the field must optimize learning from the results and renew commitment to finding an answer to this important research question. We also need to continue to work together to address the uncertainties that arise from the STEP results to implement both NHP and human clinical studies that provide new scientific insights.
Gary Nabel introduced the afternoon sessions. Dr. Nabel began by highlighting the premises of the VRC multiclade vaccine. VRC leverages vaccine-induced T-cell immunity to control HIV infection by preventing acquisition of infection, delaying disease progression, and reducing spread within a population. Envelope antigens are critical components in addition to internal viral proteins, gag, pol, and nef—they provide additional T-cell epitopes and a basis on which to build neutralizing antibody response in future trails. Additional premises of this vaccine include DNA priming and rAd boosting offer a highly effective platform for inducing CD8+CTL, increasing the breadth of immunity by vaccinating with multiple genes to diminish immune escape, and provision of a vaccine that addresses viral diversity.
The VRC vaccine, with ~1200 vaccinees and ~500 placebo recipients participating in Phase I and II studies of VRC candidate HIV DNA and rAd vaccines, offers a solid safety record. There have been >700 person years of follow-up in TRIAD studies. Research has found no vaccine-related serious adverse events and no clinical trends of concern. Compiling findings of all of the Phase 1 and 2 protocols evaluating VRC candidate HIV vaccines, 10 subjects have acquired HIV infection—6 subjects after placebo and 4 after DNA only or DNA/Ad5.
The next talks review the immunologic characteristics of the VRC vaccine, characterization in non-human primates, differences from the Merck vaccine, and clinical protocol considerations.
Richard Koup spoke about the immunologic characteristics of the VRC vaccine, including the distinct character of DNA/Ad versus Ad immune responses. He also discussed Phase I and II immunogenicity and Ad responses and shared preclinical, T-cell, and antibody response data, as well as T-cell and antibody responses.
DNA priming was found to alter the magnitude and quality of the immune response induced by Ad5 (higher magnitude CD8+ response; CD4+ and CD8+ T-cells produce other cytokines in addition to IFNy—greater memory potential). The VRC multiclade DNA/rAd5 vaccine induces:
Daniel Barouch discussed characteristics in NHP including improved T-cell potency and expanded breadth increases efficacy in the SIVmac251 infection model. He also shared that the DNA/DNA/DNA/Ad vaccine is superior to Ad/Ad/Ad vaccines in this model.
In Merck preclinical studies, the SIV challenge model (Mamu-A*01-negative Indian-origin rhesus monkeys, high-dose i.v. SIVmac251 challenge) was utilized to assess protective efficacy of Merck rAd5 alone and DNA/rAd5 regimens. Results include:
Dr. Barouch also discussed the initial findings of peak and early post-acute viremia studies (long-term studies are still in progress) sponsored by the Integrated Preclinical/Clinical AIDS Vaccine Development program.
The first study looked at the protective efficacy of heterologous rAd prime-boost regimens against SIVmac251 in rhesus monkeys. Researchers found that protective efficacy correlates with immunogenicity and that heterologous rAd26/rAd5 regimen affords significant protection against SIVmac251 where the homologous rAd5/rAd5 regimen essentially fails—and likely reflects improved quality, magnitude, and breadth of responses elicited by rAd26/rAd5 compared with RAd5/rAd5.
A second study looked at the protective efficacy of adding Env to gag/pol/nef antigens against SIVmac251. Analysis found that rAdHVR48 alone regimen expressing SIV gag/pol/nef afforded minimal protection, which is consistent with findings of previous studies. Addition of homologous Env improved early protection. Broadening antigen coverage appears beneficial; however, evaluation did not include the impact of homologous Env and the role of Env cellular versus humoral immunity remains unclear.
Preliminary data from ongoing SIV challenge studies shows that heterologous rAd prime-boost regimens afford greater protective efficacy than homologous rAd5 regimens. The addition of homologous Env to gag/pol/nef antigens improves protective efficacy. Noral rAd vectors may represent a potential solution to the safety and immunogenicity concerns of utilizing rAd5 vectors in the presence of anti-Ad5 immunity including heterologous DNA prime, novel rAd boost regimen, and heterologous regimen using two novel rAds.
Dr. Nabel reviewed the differences from the MRK vaccine, including the implications of STEP and NHP analyses, and alternative vector planning.
From a product perspective, Dr. Nable noted that the VRC and MRK vaccines are very different. The only common constituent is Ad gag-pol (no nef). In addition to the number of components (<1 of 22 injected components), there are differences in the biology of the products. Dr. Nable shared the following summary of distinctions between the VRC and MRK vaccines:
NHP modeling of increased acquisition, Dr. Nabel noted, may inform us about the mechanism of adverse effects in STEP (cellular versus humoral versus cytokine). NHP modeling also offers an opportunity to test novel vaccines in a model that assesses safety and to examine the vaccine on acquisition—separate from studies that look at viral load. Alternatively, limitations exist for NHP modeling of increased acquisition including not knowing if the modeling can be successful, can prompt SIVmac251 to behave similarly to HIV-1 in this regard, or can achieve sufficient viremia at appropriate mucosal sites to model prior Ad5 immunity in humans. Dr. Nabel estimates that modeling work will take 1 ½ to 2 years plus more time to analyze the mechanism.
Noting that HIV infections—and associated mortality—continue to occur unabated throughout the world, Dr. Nabel advocated proceeding with PAVE-100 for several reasons. PAVE-100 will test a distinct T-cell vaccine concept, as well as provide lessons to advance the field. PAVE-100 also offers the first opportunity to test a vaccine platform that affects SIVmac251 disease progression and survival. Limiting the trial to Ad seronegatives addresses an imputed potential safety concern and at the same time gives the vaccine its best opportunity to work.
STEP results offer key implications for new vector development. It is necessary to understand the mechanisms of enhanced infection in the STEP trial, specifically whether it is related to rAd vectors or other factors, to inform vaccine strategies to control or prevent HIV infection. In addition, alternative Ad serotype viruses provide an alternative to circumvent pre-existing immunity to Ad5 if it should prove to be the cause of the adverse events in the STEP trial. The critical question going forward, emphasized Dr. Nabel, is whether alternative stereotype rAds are closely or distantly related to Ad5 and thus likely to circumvent problems associated with rAd5 immunity.
If there is efficacy in PAVE-100, and/or informative data from NHP modeling in STEP, then there is a rationale for utilizing alternative vectors already in development. Potential choices include:
Dr. Nabel concluded by supporting going forward with PAVE-100 because PAVE is a different immunologic probe—while STEP looked at intracellular cytokine induction, PAVE investigates CD4 memory. He suggested AVRS think about the optimal design of a trial in Ad seronegatives, threshold, plus process for proceeding into Ad seropositives, prioritization of alternative vectors, additional studies of TRIAD and STEP clinical specimens, suggestions about the clinical protocol.
Dr. Nabel opened the floor for discussion among committee members, asking the group to focus on the following question:
From a potential efficacy presentation, is the VRC candidate sufficiently different from prior candidates and promising to warrant further investigation? Committee members agreed that the VRC candidate is sufficiently different from prior candidates—such that it warrants further investigation.
Magdalene Sobieszczyk covered clinical protocol considerations including safety data, STEP implications for the PAVE protocol, and protocol redesign. Dr. Sobieszczyk opened with a review of the purpose of the PAVE-100 trial—to test the concept that the VRC multiclade, multigene DNA prime/rAd5 vector boost preventive vaccine approach provides efficacy in preventing HIV acquisition or modulating viremia (and disease progression) once HIV infected.
Issuance of protocol version 1.1 occurred on May 16, 2007—CRFs completed, regional trainings begun, and the study scheduled to open in North America on September 28, 2007. This launch date was set based on accumulating safety and immunogenicity data form TRIAD and other studies of the VRC products involving more than 900 volunteers and reviews by several groups.
PAVE-100, a phase IIB test-of-concept trial, is a multi-center, randomized, blinded, placebo-controlled, and event driven study. To determine vaccine efficacy, 180 endpoints provide 90% power for detecting acquisition endpoint, viremic endpoint, and primary evaluation time, and encourage heterogeneity. The primary analysis weighted intent-to-treat with infections at ³26 weeks after randomization. The original study population for PAVE-100 included 8,500 healthy, HIV-uninfected males and females, ages 18-45 years, at risk for HIV-1 infection through sexual exposure, from three regions of the world with circulating A, B, and C subtypes.
Dr. Sobieszczyk noted that with the many unanswered questions swirling around STEP, it is premature to abandon the “T-cell based vaccine” concept. The VRC vaccine regimen, with its distinct preclinical and human safety/immunogenicity profiles, should continue to move forward in development. However, the results of the STEP trial demand re-examination of the PAVE-100 study design and amendments to insure participant safety, optimal science, and maximal clinical trials efficiency.
Dr. Sobieszczyk said that, looking at efficacy testing of the VRC regiment in Ad5 seronegative populations, there are many advantages with moving forward expeditiously in Ad5 negatives—including equipoise. The trial design is simplified (fewer subgroups) and researchers would be testing the VRC regimen in the population most likely to see the efficacy signal. Arguably, the VRC regimen is the most efficient way to move the regimen, concept, and field forward with maximal safety. In contrast, acceptability of this approach in regions of high Ad5 seroprevalence is a major concern. The accrual process is inherently inefficient given the need to screen out Ad5 positive at baseline—subgroups could be efficiently studied would be limited.
Given that there is no clear way forward to efficacy testing as the next immediate step, it is necessary to proceed with the seronegative trial first to overcome the hurdle and get answers on how to handle the seropositive population. If the Ad5 seropositive trial proceeds, work must continue on developing a plan for Ad5 seropositives such as alternative Ad (26, 35, chimeras) or other viral vectors, or engage sites, communities, and country oversight groups in the process. The scientific opportunity presented by this curmudgeon highlights the need to understand the natural history of Ad5 titers/serostatus, alternative Ad5 serostatus, risk behavior, and HIV-1 acquisition among Ad seropositives.
The new design focuses on Ad5 seronegative subjects only, enrolls a smaller number of subjects are followed for a longer period and evaluated more frequently—DSMB review after the first 30 infections and a 3 month instead of 6 month testing schedule (shortening the trial by about six weeks). With only two regions (Africa and Americas) engaged, there are less numbers of subgroups.
The new design also preserves essential elements of the original design, including the event-driven design, regional differences, both MSM and heterosexual transmission cohorts, and evaluation of viral and host genetics. However, the redesigns reduce the specified number of infections:
Lively discussion occurred about Dr. Sobieszczyk’s proposed revisions to the PAVE study design. Debate transpired about the third redesign. One participant noted that the “almost complete lack of infection in women” is a reason to abandon this design. Another participant retorted that, in East Africa, the cohort includes women at risk and both seronegative men and women at high risk.
Comments about the proposed redesign of PAVE-100 also focused on the 90% power goals. Since PAVE-100 is not a licensure trial, one committee member questioned the need for 90% power goals. The recommendation was to consider an 80% power design in order to produce findings with greater speed and efficiency.
The committee also discussed the importance to the field of learning quickly about the efficacy of a T-cell vaccine. The need to design the study to get to the result in the shortest timeframe was highly debated. Some felt it was best to focus the study on the highest risk subgroup or to go forward with the simplest design in order to set up the trial to maximize testing of the concept. Others noted the need for a vaccine to be effective across diverse subgroups in order to ensure vaccine efficacy.
Is there equipoise with proceeding with a phase 2b test of concept trial of the VRC regimen in Ad5 seronegative individuals in 2008? The vast majority of committee members agreed that the VRC vaccine differed significantly from the MRK vaccine, and supported going forward with the VRC regimen. One member raised the importance of conducting a “head-to-head” compare of the MRK and VRC data—using the same method, same lab, and the same everything. Dr. McElrath noted that her group would like to do that and have the ability to do that, but need to get the go-ahead from the committee that this task is the highest priority. She asked for clarification on what is more important—comparing the data or analyzing what happened in the STEP trial.
Four individuals offered public comment. The first speaker noted that in Africa, there was an understanding that VRC is a very different vaccine and she advocated the importance of continuing with the PAVE-100 trials. The next two speakers, both from Africa, echoed the plea to continue with PAVE-100 studies. They also both highlighted the importance of bringing IRBs into the conversations—and emphasized the need to educate IRBs about the science and biological differences of the vaccines. One speaker also reiterated the importance of performing a head-to-head analysis of MRK versus VRC.
Drs. Hunter, Johnston, and Fauci thanked committee members and other participants for attending and sharing their insights. At 4:15 p.m., the meeting adjourned.
back to top
Last Updated June 27, 2011