HIV/AIDS

AIDS Vaccine Research Subcommittee Meeting - May 30, 2008

The AIDS Vaccine Research Subcommittee (AVRS) met in public session on May 30, 2008 at the Bethesda North Marriott Hotel and Conference Center in Bethesda, MD.  

This regular meeting of the AVRS focused on PAVE 100A. Presentations elucidated the findings from the STEP study, described the design and rationale behind the PAVE study and reflected the community’s perspective. The PAVE 100A team emphasized that the STEP study should be interpreted as a failure of a product rather than a failure of a concept, and delineated the similarities and differences between the VRC vaccine and the vaccine used in STEP. The VRC vaccine that would be used in PAVE 100A has a different profile in non-human primate studies and the targeted population for PAVE 100A should reduce safety risks.

1. Is the main conclusion from the 12/12 AVRS meeting still valid?
• that the VRC DNA+Ad5 candidate is sufficiently different from the Merck Ad5 candidate to warrant further testing
2. Will the proposed PAVE 100 protocol advance the field? Is the design optimal?
3. Does the committee have further suggestions?

STEP Trial Results, Comparative Data with VRC Samples 

• The AVRS was extremely impressed with the amount of data generated since the December 12, 2007 consultation aimed at the ability to compare the Merck versus VRC vaccines utilizing standardized reagents and assays.
• A very limited data set from STEP suggested an inverse correlation between cellular immune response as measured by gamma IFN Elispot and viral load in the Ad<18 titer group but not the Ad>18 titer group.
• PAVE 100A would permit a closer look at this observation, and this provided some committee members with a rationale for moving forward with the proposed PAVE 100A trial
• However, one member expressed concern that, if this were the motivation for moving forward, then the two vaccines are “too different” and thus there may be limited ability to reproduce the observation seen with the Merck vaccine.
• Some committee members questioned whether the Merck and VRC vaccines were different enough to warrant expectation that the end results would be different. A majority, however, acknowledged that:
• The backbone and biology of the two vectors are different.
• The regimens are different (3xDNA prime/Ad boost vs 3x Ad immunizations).
• The magnitude of overall CD4 response is enhanced with the VRC vaccine.
• This is due in most part to CD4 responses to Env (not present in the Merck vaccine).
• CD4 responses to Gag are similar for both vaccines.
• The magnitude of the overall CD8 response was greater against PTE peptides with the Merck vaccine, and this is seen also with each individual antigen (Gag, Nef, and Pol).
• The VRC vaccine includes Env antigens.
• This provides an additional target for CD8 responses, and the point was made that this could help decipher the role of Env in a vaccine, since in natural infection CTL responses to Env are associated with higher VL.
• Also raised the question of whether Env-binding antibodies might play a protective role in the context of a robust cellular response.
• There is no significant difference between the two vaccines in terms of breadth of response. This is a major concern to some members of the committee.
• NHP studies with the VRC vaccine demonstrated that use of a DNA prime resulted in an altered SIV disease course as determined by a decrease in virus load for the first 120 days and decreased rate of mortality. Merck has subsequently presented results of similar magnitude with respect to viral load with use of a DNA prime, but this effect was not observed with the Merck Ad5 that was tested in the STEP trial.
• The VRC experiment was performed in MamuA01 negative animals with IV challenge, whereas Merck experiments included MamuA01 positive animals and low dose intrarectal challenge.
• Challenges in all experiments used a homologous SIV so breadth of protection could not be assessed.
• Differences could be due to Env antigen that is present in the VRC product, which was homologous to the challenge virus.
• Some members of the committee expressed concern about whether NHP studies should be used to prioritize clinical trials of different vaccine platforms and stressed the need to test the VRC vaccine in humans.

PAVE 100 Protocol Redesign

The AVRS was impressed by the PAVE 100 team’s responsiveness to the AVRS recommendations of December 12, 2007 and made the following observations:

• The study design for PAVE 100A is improved over the previous version.
• Much of what was learned from STEP has been incorporated into the PAVE 100A design.
• Due to safety concerns, at the present time an Ad5-based vaccine can only be tested in a small subset of the population (Ad-5 seronegative males who are circumcised). For this reason, the study results will not be applicable to a broad population and most certainly could not lead to product licensure should results prove positive. Therefore the PAVE 100A trial must be viewed more as a clinical scientific experiment rather than a pathway to product development. Because of this, several points were stressed by committee members:
• Testing the VRC vaccine in a focused population could provide an opportunity to more definitively address the possibility of an inverse association between cellular immune responses and virus load without some of the confounding variables (i.e. Ad5 serostatus, circumcision status) identified in the STEP study.
• Expectations would need to be clearly articulated to both the trial participants and the general public.
• What will be learned if the trial shows no impact on viral load or HIV acquisition?
• A strategic plan for moving forward after PAVE 100A would be needed.
• PAVE 100A would be a proof of principle; an immune correlate study, not a test of product nor a test of concept.
• If PAVE 100A is intended as an immune correlate study, the types of assays used to measure responses would need to be further evaluated.
• More extensive genomics analyses would also be needed.
• The study may actually be underpowered to look at immune correlates but this could be addressed with longer term follow up to yield more endpoints.
• In terms of safety, the statisticians should develop an algorithm to monitor acquisition in an unblinded fashion for the ability to determine if there is enhancement in the vaccine arm as early as possible; a continuous monitoring algorithm was suggested.
• Additional thought should be given to the number of trial participants needed to answer the most important scientific question – whether the vaccine impacts viral load.
• One member suggested PAVE 100A should be redesigned as a smaller Screening- Test-Of-Concept (STOC) trial. Others argued that a STOC trial would determine impact on viral load but would not generate sufficient endpoints to thoroughly evaluate potential immune correlates should a difference in viral load be observed. A STOC trial would also not be sufficiently powered to look at an effect on HIV acquisition.
• The idea of a clinical experiment to determine if the VRC candidate could provide a lead on a correlate of vaccine-induced control of viral load was acceptable to a majority of AVRS members.

Community Input

Community members recognized the challenges of implementing PAVE 100A given the proximity of the STEP trial; and while there was a general perception that the trial could be successfully implemented, it was recognized that if the trial goes forward considerable effort must be made to engage and educate communities as to the rationale and importance of the protocol to the field.

Public Input

Input from two members of the public was obtained. The first argued that given the limited differences between the Merck and VRC vaccines and the limited chances for a successful outcome that the PAVE 100A trial should not go forward. The second also argued against the trial with concerns about the limited data for differences, and the potential negative impact of an unsuccessful trial.

Summary:

• While there were a variety of opinions on the merits of the VRC vaccine versus the Merck vaccine, a majority of the committee expressed support for the modified PAVE 100A trial moving forward based on the conclusions above. A minority of members were opposed to the trial going forward (see below)
• Many on the committee expressed the opinion that PAVE 100A should be considered as a clinical experiment that could provide a lead on a correlate of vaccine-induced control of viral load. If the trial goes forward, this would need to be communicated clearly to the potential participants and the broader community.

Minority Conclusion:

Some committee members felt that the comparison of the Merck and VRC immunogenicity studies indicated that there is no reason to hypothesize that the VRC vaccine will have any chance of efficacy, and that the modified PAVE 100 A trial should not go forward.

In terms of the 3 questions posed to the AVRS one member proposed the following:

1. Is the main conclusion from the 12/12 AVRS meeting still valid - that the VRC DNA+Ad5 candidate is sufficiently different from the Merck Ad5 candidate to warrant further testing? No, not in an efficacy trial, since there are no data to suggest a better outcome than achieved with the Merck product
2. Will the proposed PAVE 100 protocol advance the field? Is the design optimal?  No, but further testing in humans of vaccines that are not effective may provide more insight into why these vaccines are failing
3. Does the committee have further suggestions?  Consider further testing of the VRC vaccine, but do not do this as a modified PAVE 100A trial, but consider a trial designed to answer the new questions that arise from the conclusion that this vaccine induces T cell responses that are remarkably similar to the Merck product

back to top