Bethesda North Marriott Hotel & Conference Center, Bethesda, MD
The AIDS Vaccine Research Subcommittee (AVRS) met in public session on February 2–3, 2010, at the Bethesda North Marriott Hotel & Conference Center in Bethesda, MD.
AVRS members present: Eric Hunter (chair), James Bradac (executive secretary), Jay Berzofsky (ex officio), Deborah Birx, Dennis Burton, Larry Corey (ex officio), Kevin Fisher, Nancy Haigwood, Barton Haynes (ex officio), Jeffrey Lifson, Nelson Michael (ex officio), Gary Nabel (ex officio), Douglas Nixon, Michel Nussenzweig, Louis Picker, Bali Pulendran, Nina Russell, Jerald Sadoff, George Shaw, Bruce Walker.
Other NIH personnel participating:
James Bradac called the meeting to order at 8:30 a.m. and asked the committee members and observers to introduce themselves. Carl Dieffenbach recognized the service of three members who are rotating off the subcommittee: Kevin Fisher (3 years), Eric Hunter (6 years), and Bruce Walker (3 years).
Peggy Johnston explained the rules covering conflicts of interest and asked members, in turn, to reveal any conflicts.
Anthony Fauci welcomed participants to the meeting, and specified that its purpose was to solicit input on the way forward given the divergent views on the appropriate follow-up to RV144. He sought their opinion on the following questions:
Jerome Kim explained that final result will not be available until the end of 2010, but preliminary results show that the vaccine regimen produced modest protection against infection, albeit with no reduction in viral load (VL). Analysis of risk indicators showed that compared with baseline risk, the highest risk group (heterosexual contact with multiple partners) had three to four times the risk of baseline (almost 60 percent). Intravenous drug use (IDU) and rectal sex carry particular risk, as do coital risk in developing countries (for reasons that are not completely understood). One possibility, then, would be to restructure the population by “rising risk,” an analysis that was not specified in the original protocol.
Evidence suggests that vaccine efficiency (VE) was about 60 percent after 12 months, but that there was a higher VE early after vaccination that was not sustained over time. Both the frequency and the magnitude of binding antibody responses decline sharply at 24 weeks, compared with 2 weeks post-vaccination. There was no difference in VL between vaccinated and unvaccinated seroconverters, either before or after 600 days post-diagnosis. Key questions to be resolved include the following:
Nelson Michael explained that the samples collected during RV144, and the discoveries flowing from the trial, will be most efficiently exploited through large collaborative networks. As part of its post-trial activities, therefore, the US Military HIV Research Program has established two working groups to formally review proposals for release of samples and further work to identify correlates of protection (COPs)—a Scientific Advisory Committee, made up of four subcommittess, and a Product Development Advisory Committee. Interested parties can submit concepts for future studies at www.hivresearch.org.
Bart Haynes, speaking via telephone, brought two recent papers to the attention of the subcommittee. The first, by S.A. Plotkin (Clin. Inf. Dis. 47:401–409, 2008), offers useful definitions of absolute, relative and surrogate correlates of infection and proposes a three-step strategy for evaluating COPs in RV144:
The second paper, by Gilbert et al. (J. Inf. Dis. 196:1304–1312, 2007), has been applied to data from RV144 to include the biological variability of immune responses in the power calculations for assessing risk in the vaccine arm. This approach calls for evaluation of immune response both at week 26 and at a most recent sample, as well as vaccination and measurement of immune responses in 200 to 400 uninfected placebo recipients.
Haynes reported that 19 proposals have been received to date, looking not only at events in early transmission but also at cellular responses, genetic screening, and NHP models. Pilot studies for assay development will begin February 2010; design of case-control studies will take place in May–June 2010; case-control study will be implemented July 2010; and data analysis will begin in November 2010.
In the discussion that followed, members explained that many of the assays to be used in the studies have not been validated, and will be qualified in a short time frame for implementation in the studies. Only four assays to date have been validated for sensitivity and specificity. Cellular assays present a greater logistical challenge than molecular ones. Sequencing of 125 breakthrough infections, involving 105 viruses, will require 6 months, followed by evolutionary studies of any interesting viruses. RV152 will provide additional opportunities for study of breakthrough infections and linkages. One problem in the NHP model is the narrow eclipse phase – the short interval between transmission and peak viremia do not give the innate response much time to act. Mucosal samples are not from RV144; they are planned as part of future clinical trials.
At present, there are not enough endpoints to determine whether the vaccine effect is greater or more sustained for low-risk than high-risk populations, and it is not clear whether the difference between high- and low-risk populations can be modeled in NHPs. Activities in CAVD and CHAVI will be covered in existing budgets, and DAIDS will try to absorb other proposals without impact on the payline for unsolicited RO1 proposals. Several members suggested that there is a need for transparency in funding mechanisms.
Phil Gomez stressed the need to establish a timeline for scientific results in order to coordinate clinical outcomes with product development. The consensus of the advisory group was future clinical trials should adhere closely to the RV144 regimen not only to refine our understanding of the immune response but also advance this regimen and improve the vaccine. Improvements to this regimen should be compared with the results of other new vaccines. It will take about 36 months to complete this process, at which time it will be possible to conduct a new Phase 3 trial, either in high- and low-risk populations or in non-clade E populations.
In the discussion that followed, one member suggested that it would be desirable, at a minimum, to boost the vaccination in order to prevent the observed fadeout. This procedure was successful in the cholera vaccine; a 12-month boost looks appropriate. Another member suggested that three factors – risk, product performance, and immune response – would determine the best trial design. A third member suggested that it might be useful to conduct pilot studies in intravenous drug users (IDUs), who not only have higher risk of infection but also show a consistently higher rate of multiple infections (up to 16 founder viruses). Apparently, there are not enough infections in sex workers and men having sex with men, and not enough clade B vs. E infections, to compare the efficacy of the existing regimen.
Nelson Michael made it clear that MHRP is not comfortable changing too many variables at once. It’s not obvious whether they should use the same or a new (non-E) insert, or whether high risk lowers vaccine efficacy. It would be best, in any follow-up study, to simplify the regimen and focus on results at 12, 24, and 36 months. He urged members not rush to conclusions, or even to changes, but rather to wait for more mature scientific findings and discussions.
Peggy Johnston explained that the preceding presentations were designed to inform AVRS members on the current status and activities with regard to RV144. The following panel discussion will give stakeholders a chance to describe what they’re doing in response to RV144, with the goal of identifying additional things that NIAID should do.
HVTN – Larry Corey pointed out that the field is always surprised by the results of HIV vaccine efficacy trials in humans, largely because our predictive models (i.e., NHP studies) are flawed. Moreover, the answers to emerge from NHP studies on most of the questions that will help us design better vaccines—improved immunogens, host factors, synergy between cellular, innate and adaptive immunity, mucosal events and responses –will be less accurate and useful than answers from human trials. In his view, human trials should be the driver in vaccine biology, with NHP studies to focus on gaps, such as early events; certainly, there should be fewer studies in large numbers of animals.
Corey added that the current pace of vaccine efficacy trials is too slow; 3 to 5 years, plus another year of analysis to arrive at COPs, is unsatisfactory. Instead, he proposed a new strategy based on a sequential, adaptive design in which a series of coordinated Phase 2B trials are conducted in the same population or region over time with similar entry criteria and followup. In order to quickly eliminate ineffective regimens, identify regimens that are better than placebo, and alter those regimens (e.g., boosting) as indicated. Corey suggested South Africa as a logical host for this approach, on both operational and epidemiological grounds. Assumptions include 1500 total subjects, 4% incidence, and multiple comparative arms. One possible trial would involve three arms, all using pox vectors and clade C inserts; a second would evaluate passive immunoprophylaxis vs. placebo; a third would evaluate Ad26 and/or Ad35 vectors with mosaic inserts, with or without gp120 boost, vs. placebo. Results could be forthcoming by 2014.
PTVDC/CAVD. – Giuseppi Pantaleo described plans for a poxvirus T-cell vaccine plus protein boost. In view of the results of RV144, there is a need to augment the vaccine effect (>61 percent), prolong it (>12 months), and broaden it (by combining innate and cellular responses). CAVD has extensive experience with NYVAC, and in that experience three doses of NYVAC+DNA is better than two doses. Improvement to NYVAC itself might include insertions or deletions to enhance the insert (NYVAC-C KC appears promising). Clinical trials would be designed to identify the best combination of vector, insert and boost. Vector development is the critical path, requiring several years, but he believes CAVD would be ready to mount a Phase 2B clinical trial in sub-Saharan Africa in 2013–2014.
MHRP. – Nelson Michael posited that ALVAC vector is the “poxvirus to beat,” and added that DNA prime enhances the immune response to MVA in preclinical studies. Much could be accomplished with mosaic inserts, which deepen and broaden the immune response in mice, but there are no efficacy data in humans. Heterologous vectors (e.g., Ad26/MVA have shown superior immunogenicity in NHPs, but once again there are no human data. He suggested a series of trials to resolve some of these questions:
To achieve these results, however, will require the field to “enfranchise the discovery element of the collaborative networks.”
IAVI. – Rick King agreed on the need to improve upon the response elicited in RV144 and suggested that there are several paths toward that goal. IAVI’s new strategic plan identifies three priorities for its future research:
King added that, by 2013, the field would be further informed by the results of HVTN 505, follow-up studies from RV144, Phase 1 trials of new vector/antigen combinations, and preclinical research on COPs and immunogen design. He therefore suggested four areas where NIAID could support these and other HIV vaccine efforts:
VRC. – Gary Nabel indicated that VRC would focus on vaccine biology, as well as vaccine candidates. The next generation of B cell vaccines may be ready 1-2 years from now, which can go forward into clinical trials. Four planned VRC initiatives that are relevant to RV144 include:
Gates Foundation. – Jose Esparza reported that the Gates foundation is focused on the development and delivery of products, with the long-term goals of an effective (70 percent) vaccine by 2015 and a reduction in the overall incidence of AIDS by 2025. In response to RV144, Gates supports collaborative efforts to develop an improved poxvirus prime/protein boost for testing in a Phase 2B trial in South Africa in 2013. In parallel, it supports Phase 1 and 2 trials of alternative components and regimens, with the possibility of multi-arm Phase 2B testing of alternatives in Africa. Its recommendations for NIAID included the following:
CHAVI. – Bart Haynes suggested that investigators are right to look for COPs, if they exist, but it will be just as important to determine which vaccine component is responsible for protection, whether in human or NHP studies. Analysis of differences between high- and low-risk subpopulations is also valuable. However, the key question is how to improve the RV144 vaccine? This can be done on the T-cell side by using 2- or 4-valent mosaics; a comparative trial is planned for about January 2011. It can also be done by developing a better boost – comparing different proteins and analyzing the epitopes elicited in vaccines.
Peggy Johnston thanked the panelists for their presentations and recommendations, which provided a wealth of ideas in three streams:
She again requested subcommittee members and other participants to engage in a frank discussion of what NIAID should (or shouldn’t) do to advance this effort. Anthony Fauci focused on the second question by asking for the pros and cons of repeating RV144 vs. conducting a new clinical trial with a different vaccine. The two aren’t mutually exclusive, but they are very expensive.
In the discussion that followed, it emerged that the most cogent arguments for repeating RV144 is the need to confirm the original results, and if possible to extend those findings by doing things a bit more carefully. If possible, this might allow investigators to understand the results in hand and perhaps add an arm in which vaccination is boosted, to improve and extend the weak, transient protection observed. It also allows a second chance to collect “the right specimens at the right times.” It is not clear how large the follow-up needs to be to deconvolute all of these variables. MHRP is confident that it can handle the logistical challenges – “the Army is all about logistics.”
One argument against repetition is that RV144 cannot be repeated exactly in South Africa, where clade C is dominant. The new populations would have more STDs, different routes of infection, different risk behaviors, different Helminth infections and coinfections. Logistical difficulties aside, with an estimated incidence of only 4 percent, it would take a long time to get results even from a trial with 50,000 participants. More to the point, we cannot wait another 3 years for mere confirmation – it’s time for the field to move on and take some risks.
Another member suggested that half of the scientific community doesn’t believe the results of RV144, in part because these are the first positive results in 26 years, but also because they don’t understand why this combination of individually failed vaccines should produce these promising results. The new trial doesn’t have to be huge, or even identical. It would be relatively cheap to answer most of the remaining scientific questions in a small clinical trial that looks at high- and low-risk subpopulations in the same clade; if the durational value of a boost is also of interest, then another arm can be added for that. In addition, if it were vital to move something to Africa immediately, it would be relatively simple to do these same arms with a different (clade C) insert but the same vector, stratified by risk, with or without boost (also clade C). The goal should be to do something in South Africa that is as good as or better than RV144.
In further discussion, most participants supported the concept of adaptive design, as described initially by Larry Corey, because it tests many components rapidly and allows the field to drop “losers” and move “winners” forward as quickly as possible. In addition, it uses no more human subjects than necessary, using them more intelligently and making each volunteer more valuable – something that might well be added to the recruitment message. This approach also allows the field to integrate its portfolio, conserve resources, and respond quickly to new findings. This approach promises to bridge the gap between basic and clinical research, generating hypotheses and jump-starting product development. Agreement is needed in advance that this is an “experiment,” rather than a “trial,” in order to avoid some regulatory hurdles.
Participants were quick to point out that plans to repeat and extend RV144, and plans to adopt the adaptive design of clinical trials, does not affect NIAID’s balance between basic and clinical research. Significant investments continue to be made in basic research, and new discoveries have many routes to move forward. CHAVI and HVTN are working on the interface between human and NHP studies, with recommendations due soon. Much of what is needed for clinical trials can come from existing collaborations; if Gates wants to support a poxvirus trial in South Africa, then NIAID is glad to collaborate with them. Meetings of implementers can be used to work out the operational details; a second such meeting is coming soon. More discussion of funding mechanism would come the following morning.
Peggy Johnston summarized the day’s deliberations by saying that there was considerable enthusiasm for additional studies of immunogenicity and COPs coming out of RV144. This includes the proposed study of the RV144 regime, with new inserts and boosts, in South Africa. Feelings were more mixed about mounting a new study with new money to investigate new products in new cohorts, so this idea goes to the back burner. There was also enthusiasm for the idea of adaptive design, making small changes in sequential stages that are part experiment, part trial (questions remain about how to present this approach to volunteers and regulators). No such enthusiasm (and few specifics) emerged on passive transfer, so this will be held for future discussion. NIAID should preserve the present balance between basic and translational research, and should partner with others for trials that NIAID itself cannot afford.
There being no public comment, the meeting adjourned for the evening at 4:00 p.m. and reconvened the following morning at 8:00 a.m.
The chair, Eric Hunter provided a summary of the first day’s discussions, with particular attention to the subcommittee’s support for further clinical trials in Thailand to investigate immunogenicity and COPs—but its lack of support for large-scale trials in Thailand. Instead, the subcommittee supported sequential adaptive design of trials in high-incidence sites, keeping changes from RV144 regimen to a minimum. There was also support for studying passive protection from infection through Ab or gene transfer. However, there was agreement that NIAID should not tip the balance away from basic/discovery science.
In the discussion that followed, it became clear that the Gates Foundation is indeed shifting the bulk of its investment to translation and delivery, leaving basic research to NIAID. MHRP funding remains steady at $105 million per year, but no other DoD funds are likely to become available for basic biomedical science.
Several members raised questions about how the field will manufacture the reagents and adjuvants needed for this work, both in experimental quantities and in pilot and commercial lots. Big Pharma isn’t interested in these compounds, which can’t be followed to licensure or market; contract manufacturing organizations (CMOs) are available, but they require budget and organization.
Members suggested that, since RV144 is the first successful HIV vaccine trial to modest protection in 26 years, the field should consider carefully what it does next. For example, it might be desirable at this time to move the trial site to a region of high incidence, and to include an arm that includes low-risk heterosexuals. Both are included in the proposed RV144 follow-on trial in South Africa. Populations there are well enough characterized that this can be accomplished in a reasonably sized trial of perhaps 1,500 volunteers per arm, and meet the statistical requirements. Several members spoke against the assumption that a small trial cannot clarify scientific questions, if that trial is properly designed.
Because the goal is to prevent transmission as well as to suppress infection, there is a continuing need to talk about acquisition as well as viral load, and to better integrate efforts aimed at T cells, B cells, and passive protection.
Peggy Johnston reviewed the mission and scope of the DAIDS vaccine portfolio (summaries were distributed in member briefing books). The total budget increased by $30 million from FY 2008 to FY 2009, with most of the increase coming in solicited categories (including collaborative networks). Funding for contracts, which also supports clinical trials, can vary considerably from year to year, and the so-called “master contract” is being re-competed this year, which has constrained the agency’s ability to launch new activities. For this reason, DAIDS is gratified to see central laboratory capacity included in existing networks.
In response to questions, Johnston added that DAIDS has not yet evaluated the program of pilot grants from CFARs. Money for reagents is available through three different contract manufacturing organizations, of which the master contract (once in place) will be the most flexible. Total NIAID funding actually coded for vaccines is $48 million for RO1 and R21, plus $21.3 million for collaborative mechanisms like HIT-IT and the B-cell consortium. Total funding for five clinical trial networks is $280 million per year. Much of the FY 2009 increase was one time only; future funding looks essentially flat; staff would welcome suggestions on how to attract new investment and proposals.
Drs. Charles Hackett and Carole Heilmen reported on DAIT’s and DMID’s programs, respectively, with emphasis on items relevant to AIDS vaccine research.
The meeting adjourned at 12:30 p.m.
back to top
Last Updated June 27, 2011