Nearing a decision on PAVE
Experts discuss new design of the Phase IIb PAVE 100 trial
By Andreas von Bubnoff
It is not uncommon in the AIDS vaccine field for there to be some controversy about the conduct of large trials. Prior to the start of the first Phase III efficacy trial of an AIDS vaccine candidate known as AIDSVAX, there was extensive public discussion about whether or not the US National Institute of Allergy and Infectious Diseases (NIAID) should fund the trial. And before the now ongoing Phase III trial (RV 144) began in Thailand—testing a canarypox-based vaccine developed by Sanofi Pasteur in combination with AIDSVAX—there was public scrutiny over whether the trial should take place at all.
The latest debate centers around whether to conduct a Phase IIb test-of-concept trial, known as PAVE 100, which was on target to begin last year but was delayed when Merck’s candidate vaccine MRKAd5 failed to show any efficacy in the STEP trial in either preventing HIV infection, or in lowering the amount of virus circulating in the blood of individuals who became HIV infected, despite vaccination, through natural exposure.
AIDS vaccine experts met on May 30 as part of the AIDS Vaccine Research Subcommittee (AVRS) to discuss the fate of the PAVE trial. After a day of discussion about the merits of the trial and the proposed design, a majority of committee members supported conducting the trial, albeit in a smaller form than originally planned. “The vast majority, it seemed, said to go ahead,” said NIAID director Tony Fauci after the meeting, which took place in Bethesda, Maryland. After taking into account the committee’s recommendations, Fauci will make the final decision.
The PAVE 100 trial would test a vaccine regimen developed by NIAID’s Vaccine Research Center (VRC). This regimen bears some similarities to the MRKAd5 candidate—both use a cold virus called adenovirus serotype 5 (Ad5) as a vehicle to introduce HIV genes that will stimulate an immune response. The vaccination regimen evaluated in the STEP trial involved three vaccinations with the same Ad5 candidate. In contrast, the prime-boost vaccine regimen developed at the VRC involves three vaccinations with DNA encoding HIV fragments, followed by a vaccination with the Ad5 component of the vaccine regimen.
A wrinkle in the plans for the PAVE 100 trial came when researchers involved with the STEP trial subsequently observed a trend toward increased susceptibility to HIV infection among some sub-groups of trial volunteers who received the vaccine candidate. These volunteers were uncircumcised men who have sex with men (MSM) who had pre-existing immunity to the Ad5 vector from natural exposure to that cold virus.
As a result, the PAVE 100 trial would likely only involve volunteers who were circumcised MSM with no pre-existing Ad5 immunity, according to Scott Hammer of Columbia University, who chairs the trial’s protocol team. Due to these restrictions, the size of the trial will likely be much smaller than originally planned. The original plans for PAVE 100 included 8,500 HIV-uninfected men and women from the Americas and southern and eastern Africa. But at the AVRS meeting Hammer suggested a much smaller trial called PAVE 100A that would enroll only 2,400 circumcised MSM in the US who have no pre-existing Ad5 immunity. Hammer said the other branch of the trial to be conducted in Africa, now known as PAVE 100B, is “deferred for now.” “It is unlikely going to happen in any meaningful way for a considerable period of time,” Fauci said.
PAVE 100A would be limited to the US because it would be easier to enroll circumcised volunteers without pre-existing Ad5 immunity. Prevalence of this serotype of cold virus is typically higher in developing countries—in South Africa, only about 20% of people are Ad5 seronegative, according to John Hural, associate director of laboratory operations at the HIV Vaccine Trials Network (HVTN). Still Hammer said it would be necessary to screen about 6,000 potential volunteers to successfully enroll the 2,400 proposed for the trial based on observations from the STEP trial in which 56% of the men screened for Ad5 in the US were seronegative and, of those, 83% were circumcised. “Screening and recruitment will be a challenge,” Hammer said, “but definitely feasible.”
Setting a new endpoint
Hammer also suggested that PAVE 100A should not focus on whether the vaccine candidates could prevent HIV infection altogether, but instead on their ability to lower the amount of virus circulating in people who become infected with HIV during the trial, despite vaccination. In the STEP trial, researchers looked at both of these endpoints and this was also the original plan for the PAVE trial. But Hammer noted that animal studies in nonhuman primates showed that versions of the VRC’s DNA/Ad5 vaccine regimen carrying genes from simian immunodeficiency virus (SIV), the monkey equivalent of HIV, do not provide protection against infection. “That’s what the science says,” said Hammer.
The studies in monkeys did show, however, that there was some impact on viral load. Some researchers are also intrigued by data from a very small group of volunteers in the STEP trial. Julie McElrath, director of laboratories at the HVTN and a professor of medicine at the University of Washington, presented preliminary data from the STEP trial looking at a small group of volunteers who were infected but had low levels of HIV in their blood. In about a dozen such volunteers who were Ad5 seronegative, those with higher T-cell immune responses to the MRKAd5 vaccine had lower viral loads. However, this analysis was in a tiny subgroup of the 3,000 volunteers who participated in the STEP trial and interpretation of this observation is severely limited, not only by the small number of people, but also because this analysis was not part of the original trial design. “We call that sort of the glimmer of hope,” Hural said.
Jerald Sadoff, who heads the AERAS Global TB Vaccine Foundation, supports conducting PAVE 100A in order to confirm this possible correlation between viral load and T-cell responses. “It’s the only positive finding of vaccine-induced protection in the entire field of HIV vaccine research,” said Sadoff, adding that the STEP trial analysis is only a preliminary finding. “[We need to] repeat the study to show that that’s correct.”
But others questioned whether the VRC vaccine regimen is sufficiently different from MRKAd5 to learn anything beyond what was observed in the STEP trial. To address this issue, McElrath presented a preliminary comparison of human immune response data collected from the STEP trial to data from a Phase II trial (HVTN 204) of the VRC vaccine regimen. While the VRC’s DNA/Ad5 combination showed stronger CD4+ T-cell responses, both the CD8+ T-cell responses and the breadth of responses appeared to be similar for the DNA/Ad5 and MRKAd5 candidates except the predominant responses were to different HIV proteins for the two regimens.
Not everyone agreed as to whether these differences were enough to justify conducting PAVE 100A. “There is a lot of subjectivity in how you view the lab data,” said Hural.
“I was impressed by the difference in the immune responses,” said committee member Deborah Birx of the US Centers for Disease Control and Prevention. “I am strongly supportive of PAVE 100A proceeding.” But others disagreed. “The data look more similar than different,” said Jeffrey Lifson, head of retroviral pathogenesis at the National Cancer Institute. He suggested that PAVE 100A may not provide any additional insights beyond the STEP trial.
Sadoff argued that the similarity between the two vaccines was actually a good thing because it will make it possible to confirm the preliminary observation from the STEP trial that immune responses induced by this type of Ad5-based vaccine candidate may impact viral load. He said that in vaccine development it is not unusual to conduct a trial to confirm preliminary results from a previous trial. “This is normal vaccine development,” he said.
Confirming results or providing insight into future development of AIDS vaccine candidates would be the main objective of the newly-proposed PAVE 100A trial. As a result of the restrictive trial design, it won’t be possible to generalize the results of PAVE 100A to the general population, said Hammer, and the purpose of the trial would not be identifying a candidate vaccine. Rather, it would be to test the hypothesis of whether this kind of vaccine is safe and can lower viral load in individuals who become HIV infected. “We have to be very clear about this as we reach out to potential participants,” said Hammer.
That may be not easy, said Mitchell Warren, executive director of the AIDS Vaccine Advocacy Coalition (AVAC). “[This] will be the most complicated AIDS vaccine trial any of us will ever have to explain,” Warren told the committee. Others expressed concern as to what would happen if the PAVE 100A vaccine, like the STEP candidate, also failed to have any effect. “What if another trial fails? Do you risk losing the trust and support of this community for future trials?” asked Enid Moore, associate director for community education at the HVTN, which would provide clinical sites for the PAVE 100A trial. But even a negative result would move the field forward, said Barton Haynes of Duke University and director of the Center for HIV-AIDS Vaccine Immunology (CHAVI). “If it doesn’t show any control of viral load, it will empty out a large number of the trials in the pipeline,” said Haynes, since several candidates are based on similar approaches.
Stacey Little, senior program officer of the Academy for Educational Development, which works to educate the community about AIDS vaccine research, pointed out another community concern—the exclusion of women from the PAVE 100A trial. But Fauci said that eventually, such a trial will move the entire field forward and that would benefit everyone.
At the conclusion of the day-long meeting, some argued that nothing is more valuable than human data. “Man is the measure of all things,” said Stanley Plotkin, an advisor to Sanofi Pasteur. “We have to do this clinical trial.”