HIV/AIDS

AIDS Vaccine Research Subcommittee (AVRS)Meeting Summary - December 12, 2007

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.

Speakers:

• Eric Hunter, Ph.D., Professor of Pathology and Laboratory Medicine, Georgia Research, Emory Vaccine Research Center
• Anthony Fauci, M.D., Director of the National Institute of Allergy and Infectious Diseases (NIAID)
• Peggy Johnston, Ph.D., Director, Vaccine Research Program, Division of Acquired Immune Deficiency Syndrome (AIDS), NIAID
• Michael N. Robertson, M.D., Director, HIV Vaccines Clinical Research, Merck & Co. Inc.
• M. Juliana McElrath, M.D., Ph.D., M.P.H., HIV Vaccines Trial Network (HTVN), Director of Laboratories and Professor of Medicine, University of Washington
• Lawrence Corey, M.D., Principal Investigator and Professor of Medicine and Laboratory Medicine, University of Washington
• Gary Nabel, M.D., Ph.D., Director, Vaccine Research Center, NIAID
• Richard Koup, M.D., Director, Human Immunology Program, Vaccine Research Center, NIAID
• Daniel Barouch, M.D., Ph.D., Associate Professor of Medicine, Division of Infectious Disease, Brigham and Women’s Hospital
• Magdalene Sobieszczyk, M.D., Ph.D., Assistant Professor of Clinical Medicine, Division of Infectious Diseases, Columbia University College of Physicians Data and Safety Monitoring Board (DSMB)

Introductions, Conflicts of Interest

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.

Welcome

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.

Charge to Subcommittee

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:

• Primary efficacy results from the STEP trial: What are the implications of the trial for future HIV vaccine development? What additional data analyses should be afforded a high priority?
• Laboratory investigations to understand STEP trial findings: Is the process for establishing scientific priorities and access to STEP trial specimens adequate? Are timelines acceptable?
• Scientific update: What are your thoughts on the concept of T-cell based vaccines and the role of clinical trials, immunology of the VRC regimen, enhanced protection in NHP with Env immunogens, and distinctive features of the VRC regimen and future vector planning?
• Clinical protocol development: Is there equipoise with proceeding with a phase 2b test of concept trial of the VRC regimen in Ad5 seronegative individuals in 2008?

STEP Trial Presentations

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.

Primary efficacy results from the STEP trial

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.

Group Discussion

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:

• What are the implications of this trial for future HIV vaccine development?  Discussion about the pros and cons of developing vaccines for targeted subgroups (e.g., Ad5 levels) transpired. One participant raised the risk of waiting on complete analysis of the STEP trial before proceeding with other HIV vaccine development efforts—and the need for parallel activities, specifically going forward with alternative protocols while analyzing STEP data. Another committee member noted, in the context of HIV vaccine development and no viable animal model, the only way to go forward is through a clinical trial.
• What additional analyses should be afforded a high priority?  One participant noted the need to evaluate site differences and the potential phenomenon of transition clusters. Another committee member suggested adopting a more in-depth analytics approach, which entails conducting prospective case control or logistic regression analysis. Also raised was the need for more analytics regarding bio mechanisms such as T-cell response and other factors such as at what level antibody T-cell is a risk.

The group agreed to continue the discussion after completion of the next presentation.

Laboratory investigations to understand the STEP trial findings

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:

• Did the vaccine induce expected T-cell responses? Were these similar to findings in previous phase 1 studies? The team concluded “yes” to both of these questions based on ELISPOT responder data on vaccine immunogenicity evaluated for week 8—and found that both response frequency and geometric mean (GM) to be similar, across the board.
• Were the magnitude and frequencies of the immune responses lower in the vaccine cases than in the overall study cohort? How do the magnitudes compare with those in HIV-infected cohorts? Magnitudes and frequencies of responses were lower in vaccine cases; however, comparison of magnitudes is pending. Dr. McElrath noted that the compare of cases with non-cases yielded very similar responses.
• Was the breadth of T-cell responses adequate? In the cases, did the vaccine provide antigenic coverage against the infected virus? Dr. McElrath said the answers to these questions are pending. Planned studies include evaluating responses by protein, epitope mapping (at week 8 after 2 doses), HLA typing, viral sequencing (from early and later-stage infection), and confirmatory immunologic analyses.
• Is the vaccine-induced immunity long lived? Data from phase 1 studies represent that the vaccine-induced immunity is long lived. Looking at immune responses, frequencies stay high once solicited—and remain high for 78 weeks.
• Was the Ad5 HIV vaccine unable to provide the optimal quality of T-cell response to induce HIV protection? The answer to this question remains unclear. Initial studies show no significant differences for envelope or non-envelope. Though additional statistics are required, the functional profile yielded similar patterns for cases and noncases for both CD4+ and CD8+ responses. Dr. McElrath highlighted the need for further investigations to define immune function and anti-viral activities of vaccine-induced T-cells including functional phenotype of epitope specific T-cells, anti-HIV activities, and transcriptional and proteomic analysis. 
• Can vaccine-induced T-cells migrate to sites of HIV transmission? The male genital tract produced an affirmative response. Findings also revealed gag-specific T-cell responses in blood and semen (weeks 4 and 28) after the third MRK Ad5 vaccination. Looking at innate immune responses, after one dose of MRK Ad5 vaccine, the team found peak response within 24 hours, starting within 6 hours. Response began to fall back to normal 3 to 7 days after immunization.
• In the Ad5>200 group, does Ad5-HIV vaccination enhance the susceptibility of target T-cells to HIV? Does the vaccine enhance susceptibility by increasing CD4+ T-cell activation and CCR5 expression? Researchers looked at CD4+ cells at week 30 comparing vaccinees and placebos, and found a higher medium expression with high Ad5 titer for both groups. However, the team is seeing a CD8+ response for the high Ad5 titer group with vaccination and continuing their studies.

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:

• Evaluate the potential biological mechanisms for apparent increased acquisition in the Ad5-immune vaccine group:
• Define Ad5-specific immune responses and possible association with increased acquisition
• Examine potential effect of immune responses with repeated doses
• Examine effect of immunization of CD4+ T-cell number, activation, CCR5 expression in mucosal sites where HIV is transmitted (rectum and lower genital tract)
• Explore the relationship of specific Ad5 gene deletions with increased CD4+ T-cell activation
• Assess in vitro susceptibility of CD4+ T-cells (DCs, macrophages) in Ad5 low versus Ad5 high titer vaccinees
• Determine what potential biological mechanisms for apparent increased acquisition in the Ad5-immune vaccine group:
• Determine if this observation is specific to Ad5 or also characteristic of other adenovirus vectors
• Evaluate if there will be a similar issue with use of highly immunogenic viral vectors in previously immune study participants
• Learn if the vaccination altered the course of HIV infection
• Find out if HIV-1 epitopes recognized with vaccination present in the infecting strain—if not, learn if variants in the infecting strain are recognized
• Explore if immunization with MRKAd5 HIV vaccine alters the early immune adaptive response to HIV-1 and the rate of viral diversification 
• Determine if there is an anamnestic immune response to vaccine-induced T-cells and how it impacts viral load (VL) set point 
• Determine if CD4+ T-cell memory cells were preserved at set point
• Explore if escape mutation occurs early within the vaccine-induced T-cell epitopes, and if this is associated with higher VL set point
• Find out if some epitope-specific T-cells better control viremia

Group Discussion

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.

• What additional analyses should be afforded a high priority?  Discussion included the need for additional analysis of the impact of geographic location on response to vaccine—and the phenomenon that where one lives may be a driver for immune profile and that the immunology appears to be far more insightful than the biology. One participant raised the need to look at long-term progression by subgroup.
• Is the process for establishing scientific priorities and access to STEP trial specimens adequate? One committee member noted the challenge of getting access to trial samples. He strongly advised being proactive and advocated going out actively to find people who can obtain the best assays.
• Are timelines acceptable?  Dr. McElrath shared with the committee the goal to pull together a report with some major findings by February 2, 2008. She noted that the team also hopes to complete laboratory work by early spring and to have a larger data set by late summer.

Pursuing the concept of T-cell based vaccines

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:

• There is a need to better integrate the non-human primate (NHP) and clinical trials fields—and to ask overlapping questions.
• STEP results reveal that dual failures were experienced—however, looming questions remain about the reasons for the lack of vaccine efficiency on viral load set point and the potential biological mechanisms for the apparent increased rate of HIV-1 acquisition in those men with serologic evidence of naturally acquired Ad5 infection prior to vaccination.
• All cases, to date, who acquired HIV after vaccination, induced HIV specific T-cell responses following two doses of the vaccine—unanswered questions include why there was no effect on viral load set point despite the pre-acquisition elicitation of HIV specific CD-8 T-cell responses and HIV specific CD-4 T-cell responses.
• Researchers must do more work on why the vaccine did not alter HIV replications post acquisition—there are not apparent differences in set point viremia between the Ad5 seropositive and Ad5 seronegative. Data analysis is pending on CD-4 cell loss or slope, memory cell populations, and major clinical endpoint (time to initiation of ART).
• Evaluation is pending on the effects of pre-exiting immunity on post HIV infection (replication of CD-4 cell loss. Studies are underway to learn if the vaccine induced any immune pressure on the virus and to find out if the immune escape occurred very early and hence “erase” vaccine-induced immunity. The team is also determining if there was matching in the epitope specific immune responses and if CD-8+ and CD-4+ cell responses to HIV were generated after infection.
• Benchmarking against MRK Ad5 gag/pol/nef vaccine regime may prove useful for defining and prioritizing future vaccine candidates—with improved quality, magnitude, breadth, and protective efficacy in NHP, as well as evaluation of the post infection responses from vaccinated persons who become infected, may provide further insights. The vaccine field needs to develop a real heterogeneous challenge model and NHP geared to defining a correlate of post infection viremia.
• The NHP model needs to evolve and improve, and STEP reinforces the need to increase the types and size of studies in the NHP further. However, the timeline is long (years), success not assured, and human clinical trials requisite to both improve the model and test the validity that the concept of T-cell based vaccines can actually work in humans.
• Co-variate analysis in the Ad5 seropositive men revealed several factors that potentially influence the acquisition rate data—circumcision rates, race/ethnicity, age, and location of enrollment differed between Ad5 seropositive and Ad5 seronegative men.
• Covariates may eventually influence final interpretation; however, at this juncture the hypothesis that the trend in acquisition in men was due to Ad5 seropositive status requires exploration—to assume that these data were a chance event seems not scientifically correct or organizationally prudent.

Translating Research into Policy

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:

• Role of Ad5 versus an, as of yet, unidentified covariate
• Likelihood of identifying the mechanism of increased acquisition from ongoing or planned studies
• Probability that all Ad5 vaccines are the same
• Certainty that one dose of Ad5 will produce the same trend in increased acquisition rate of HIV as the three doses used in the MRK vaccine.

A Clear Path Forward

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.

PAVE Trial Presentations

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.

Immunology of the multiclade vaccine

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:

• High frequency (~70%) and magnitude of T-cell responses (sustained, polyfunctional, and frequently targeted conserved epitopes, including Env)
• High titer antibody to HIV Env (1:10,00–1:100,00)

Enhanced protection in NHP and ENV immunogens

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:

• Monovalent rAd5-gag challenge found no reduction of SIV viremia afforded by rAD5 alone or DNA/rAd5 regimen. There was transient 0.8 log reduction in peak SIV viremia for 90 days (Mamu-A*01-positive monkeys) in the context of the DNA/rAd5 boost regimen.
• Multivalent gag/tat/rev/nef (Mamu-A*01-positive monkeys) offered superior protection superior compared with monovalent gag: 1.1 log reduction in peak and 1.4 log reduction in set point SIV viremia. These data suggest that increasing antigen coverage improves protection.
• VRC DAN prime, rAd5 boost gag/pol/nef/Env found transient 1.1 log reduction in peak SIV viremia for 112 days (Mamu-A*01-negative monkeys) and significant survival advantage of vaccinated animals.

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.

Distinctive features of the vaccine and future vector planning

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:

• Different type of immunization—DNA/DNA/DNA/Ad versus Ad/Ad/Ad
• Differences in inserts in the vaccine-inclusion of Env A, B, C, which expand the breadth of the T-cell immune response
• Qualitative and quantitative differences in the type of T-cell immunity induced by the vaccine—VRC offers expanded potency, expanded immunogenicity production
• Alternative biology/construction of Ad5 vector
• Prolonged survival and reduction in viral load for >100 days in the more rigorous SIVmac239/251 challenge model in contrast to the Merck vaccine that showed no efficacy

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.

Why proceed with PAVE-100

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:

• Ad5—if no evidence for adverse effect in NHP models, or HVR48—if evidence that adverse effect is due to anti-hexon Ab but not cellular immunity
• Alternative or chimeric Ads—focus efforts on menu or at least four different alternatives such as Ad14, Ad26, Ad48, or Ad35/5 fiber
• Poxvirus
• Alternative vectors—mycobaterial, alphavirus, etc.
• Protein conjugates

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.

Group Discussion

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.

Perspectives of clinical protocol development

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.

PAVE-100: Post-STEP results

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.

Potential PAVE-100 Redesigns in Ad5 Seronegative Populations

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:

• Redesign 1: Groups all infections irrespective of geographic region with follow up at 60, 90 or 120 infections
• Redesign 2: Same as design 1, but requires at least ³ 30 or ³ 45 infections per region
• Redesign 3: Develops efficacy data in women by revising design 2 to require ³ 30 or ³ 45 infections in women, in addition to minimum infections per region.

Group Discussion

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.

Public Comment

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.

Adjournment

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