NIAID's announcement that it would not support the 2,400-person PAVE100 trial left open the possibility of an even smaller trial, solely evaluating the impact on postinfection viral load set point.
A little over a year ago, the HIV vaccine field was shaken by the news that Merck's adenovirus serotype 5 (Ad5)-based candidate had failed to protect against HIV infection or lower viral load among recipients who became infected. Even worse, the vaccine appeared to increase susceptibility to HIV infection among a subset of trial participants with preexisting antibody responses to Ad5 (see TAGLine 15, no. 1). At the time these results were announced, the National Institutes of Health (NIH) was gearing up to launch a large, international, 8,500-person efficacy trial of another Ad5-based HIV vaccine candidate designed at the NIH's Vaccine Research Center (VRC); inevitably this study -- dubbed PAVE100 -- was put on hold as scientists trawled through the data from the Merck vaccine trial to try and better understand what happened. After a series of meetings, consultations, and finally an executive decision from Anthony Fauci (director of the National Institutes of Allergy and Infectious Diseases at NIH), PAVE100 is now being redesigned as HVTN 505: a small, U.S.-based "test of concept" trial that will involve around 1,400 people.
The original design of PAVE100 was intended to provide a robust answer regarding the protective efficacy of the VRC vaccine candidate, and to also provide data that -- in combination with the results from the Merck vaccine trial -- would help guide future HIV vaccine research. The VRC approach involves a series of initial immunizations with a DNA priming vaccine followed by a single shot of an Ad5 vector. This is significantly different from Merck's Ad5-only strategy. The VRC vaccines also contain additional HIV antigens: the envelope proteins from three different HIV-1 clades (A, B, and C) are included along with the Gag, Pol, and Nef proteins that were also in the Merck construct. It was hoped that the results from trials of the two vaccines would show if these differences impacted protective efficacy. However, when it emerged that Merck's Ad5 vector had not only failed show any efficacy but was also associated with an increased risk of HIV infection among participants with anti-Ad5 antibodies, it quickly became obvious that PAVE100 could not proceed as originally designed. The trial population was focused on Southern Africa, where the vast majority of individuals (~80%) have antibodies against Ad5, and the potential for the VRC's vector to have a similarly adverse impact on susceptibility could not be ruled out.
The task of mulling whether PAVE100 could be redesigned in a way that would allow the efficacy of the VRC vaccines to be safely studied fell to a NIAID advisory body called the AIDS Vaccine Research Subcommittee (AVRS, formerly known as the AIDS Vaccine Research Working Group). At a meeting on December 12, 2007, AVRS members listened to presentations on the Merck results and the PAVE100 design and offered a series of recommendations to the PAVE100 protocol team, led by principle investigator (PI) Scott Hammer. The recommendations included reducing the sample size and focusing on a single population, as well as excluding individuals with Ad5 antibodies and uncircumcised men (because analyses of the Merck results indicated that circumcision protects against the enhancement effect observed in the trial).
The PAVE100 protocol team then spent several months designing a new trial called PAVE100A, incorporating all the AVRS recommendations and ending up with a proposal for a 2,400-person, U.S.-based trial that would enroll only circumcised gay men lacking anti-Ad5 antibodies. The co-primary endpoints were slated to be a lack of enhancement of HIV acquisition (e.g., confirming the safety of the Ad5 vector in this population) and a reduction in set point viral load among vaccine recipients who became infected. Multiple secondary immunological analyses were also included in the hopes of maximizing what could be learned from the study, particularly in terms of looking for correlations with either prevention of HIV infection or control of postinfection viral load.
The proposal for PAVE100A was formally presented by Scott Hammer at a specially convened meeting of the AVRS that took place on May 30, 2008. TAG released a statement to coincide with the event, noting that the decision was a "tough call" and outlining TAG's view that the safety cloud hanging over Ad5 vectors and the inability of the VRC vaccine to induce broader responses to the HIV Gag protein than Merck's argued against conducting PAVE100A. The AVRS meeting was presided over by Anthony Fauci, who explained that the final decision on whether to move forward with the trial rested solely on him. AVRS members also heard the latest news from ongoing analyses of the Merck trial data. Julie McElrath presented immunological analyses showing an inverse correlation between the magnitude of the Gag-specific CD8 T-cell response induced by Merck's vaccine and the set point viral load in the subset of trial participants lacking anti-Ad5 antibodies; the numbers were small, however, and the statistical significance fragile. The majority of AVRS members ended up voting in favor of conducting PAVE100A, but a variety of different and sometimes mutually incompatible reasons were offered in justification. Some members felt that the VRC vaccine was sufficiently different from Merck's that it stood a better chance of working. Among the differences cited were better CD4 T cell responses, the inclusion of envelope antigens and possible induction of envelope-binding antibodies, and potentially more functional CD4 and CD8 T-cell responses due to the DNA/Ad5 prime-boost approach. Conversely, other AVRS members supported PAVE100A on the basis that the VRC's vaccine was sufficiently similar to Merck's that it might be able to confirm the associations between vaccine-induced T-cell responses and postinfection viral load levels that McElrath had described at the meeting.
Several weeks after the AVRS meeting, Anthony Fauci announced that NIAID had decided not to conduct PAVE100A. However, NIAID's announcement left open the possibility of an even smaller trial of the VRC vaccine, solely evaluating the impact on postinfection viral load set point. This type of trial has been proposed by the International AIDS Vaccine Initiative (IAVI) and dubbed a "screening test-of-concept" for T-cell-based vaccines. Once again, the PAVE100/100A protocol team and PI Scott Hammer -- who, it is probably fair to say, have not been well-served by the extended and rather muddy decision-making process surrounding the trial -- are having to go back to the drawing board. In the last few weeks, NIAID's vaccine communications staff has announced that a 1,400-person trial (now designated HVTN 505) is in the works, involving the same population proposed for PAVE100A: U.S. based men who have sex with men who are circumcised and lack anti-Ad5 antibodies. NIAID is beginning a series of consultations with stakeholders in the trial, including community-based organizations. More details are expected to emerge as a formal protocol for the study is developed. The trial is not expected to begin until sometime in 2009.
Individuals and community-based organizations interested in participating in the HVTN 505 consultation process can join the AIDS Vaccine Advocacy Coalition's Advocates Network: www.avac.org.