Conference Coverage: Giving It Their Best Shot

Researchers from many areas of vaccine development gathered recently to discuss the challenges of developing and delivering life-saving vaccines

By Kristen Jill Kresge

You've probably heard the parable about the man who was upset that he had no shoes until he met someone without feet. This came to mind during a meeting held from October 8-13 in Cape Town, South Africa that brought together vaccine researchers from different disciplines to discuss developing and delivering life-saving vaccines throughout the world. Commiseration, as well as a sense of shared commitment, pervaded the meeting as researchers, many of whom don't usually attend the same conferences, shared ideas and approaches to developing vaccines against three of the world's biggest killers—tuberculosis (TB), malaria, and HIV/AIDS.

This inaugural Keystone Symposium on the Challenges of Global Vaccine Development was an experiment in its own right, according to conference co-chair Margaret Liu of the Karolinska Institute. It explored many of the common challenges and creative approaches, as well as some of the overlap in the strategies being investigated to combat all three diseases. The conference, which was also held in conjunction with the annual meeting of the Gates Foundation's Grand Challenges in Global Health initiative, had an added focus on efforts to successfully deliver vaccines. Tachi Yamada of the Gates Foundation says that although the foundation has always been committed to discovery, "we also have to think about how to deliver these exciting new products."

One thing that is certain is the massive public health benefit that vaccines can have. Since the genesis in 2000 of the Global Alliance for Vaccines and Immunization (GAVI), now the GAVI Alliance, the World Health Organization (WHO) estimates that the introduction of vaccines in developing countries has prevented 2.6 million deaths.

Just one year after the vaccine against haemophilus influenzae type B (Hib) was introduced in Mali through the GAVI Alliance, there was a 67% reduction in mortality due to Hib and a 32% reduction in all hospitalizations in the country. But these dramatic effects come with a hefty price tag. The WHO and the United Nations Children's Fund (UNICEF) estimate that GAVI will require between US$226 and $778 million between 2011 and 2015 to continue funding vaccination programs in its target countries.

Boosting spirits of AIDS vaccine researchers

The gathering for the Keystone conference occurred just a few weeks after the announcement that Merck and the National Institute of Allergy and Infectious Diseases (NIAID) stopped immunizations in a large Phase IIb test-of-concept trial, known as the STEP study, because Merck's adenovirus serotype 5 (Ad5)-based AIDS vaccine candidate (MRKAd5) was not effective. At the same time, enrollment and immunizations in the Phambili or HVTN 503 trial, which was taking place in South Africa, were suspended—they have since been stopped entirely (seeStopping a steam train). These were some of the most hotly discussed issues both in and out of the meeting.

Carolyn Williamson of the University of Cape Town told the audience assembled for her plenary session, "We really have to go back to the drawing board." But while many AIDS vaccine researchers were still reeling from the news, those from other disciplines were able to provide some fresh perspective. "I wouldn't be too downbeat," says Adrian Hill of Oxford University, who is currently developing possible vaccine candidates against malaria. "We've had candidates fail for malaria about 15 times."

Data analysis for the STEP trial is proceeding rapidly, but for now it remains unclear why MRKAd5 was not effective (see A STEP back?). The immunogenicity data analyzed so far shows the vaccine induced the expected levels of immune responses. But Hill expressed doubt that the low quantity of T cells generated by this vaccination strategy could ever be sufficient to fend off HIV. "Three hundred or 400 T cells against HIV aren't enough to give you protection," he says. "That's not enough to give you protection against malaria either." Hill thinks a vaccine candidate should induce three to four times as many T cells to provide some protection against HIV.

Many researchers at the meeting spoke favorably about the pursuit of a prime-boost strategy for AIDS vaccines. "The Merck vaccine was in fact three shots of a single structure," says Gustav Nossal of the University of Melbourne and conference co-chair. For some time now the idea of using a heterologous prime-boost combination of different vaccine constructs has been gaining favor in the AIDS vaccine field, and after this latest development it seems to have become the battle cry. "Prime-boost got a big boost from this negative result," added Nossal, referring to the STEP study.

The prime-boost strategy furthest along in development is a DNA candidate followed by a boost with an Ad5 vector-based candidate, which is slightly different from the one developed at Merck, both encoding multiple immunogens (gag/pol fusion construct and env) from several HIV clades. This candidate was developed at the Vaccine Research Center at NIAID. Gary Nabel presented data on this prime-boost regimen and summarized the differences between it and MRKAd5 (see What next?).

But Nabel faced some tough questions. Hill pointed out that the immune responses, as measured using the interferon (IFN)-g ELISPOT assay, induced by the VRC's DNA/Ad5 candidates against the HIV genes gag, pol, and nef, were of a similar magnitude to those induced by MRKAd5. He therefore questioned whether just the inclusion of the HIV env gene would really be sufficient to induce a significantly different magnitude of immune responses. Nabel says he expects inclusion of env will make a difference, not only in the magnitude of immune responses but also by improving the quality of these responses. He also argued that this is the reason to test these candidates in the planned Phase IIb test-of-concept trial, known as PAVE 100, which has been delayed until more data from the STEP study is analyzed and researchers can sort out any possible association between the level of pre-existing immunity to Ad5 and enhanced susceptibility to HIV infection. "The clinical trial now becomes the experiment," says Nabel.

Other prime-boost AIDS vaccine regimens were also discussed. Williamson presented data in nonhuman primate studies with a prime-boost combination of the VRC's DNA plasmid vaccine encoding Gag, Rev, Tat, Nef, and Env clade C HIV proteins and a modified vaccinia Ankara (MVA)-based vaccine, developed at the University of Cape Town, carrying the same immunogens. All eight baboons that received three immunizations with the DNA candidate followed by two booster immunizations with the MVA vector had immune responses to SIV and were considered responders based on results from the IFN-g ELISPOT assay. All of the baboons generated responses directed towards gag and nef, while six of the eight had responses to env. Williamson says this prime-boost regimen elicited a high magnitude and breadth of both CD4+ and CD8+ T cells, and therefore warrants evaluation in human volunteers. A Phase I safety trial of this DNA/MVA prime-boost regimen conducted by the South African AIDS Vaccine Initiative (SAAVI) was scheduled to begin later this year, pending approval from the US Food and Drug Administration and the Medical Research Council of South Africa, but all vaccine trials have been placed on hold by the South African Health Minister following the news on the Phambili trial (see Stopping a steam train).

A different DNA/MVA prime-boost regimen is also being tested in an ongoing Phase I trial in Dar es Salaam, Tanzania by researchers at the Karolinska Institute in Sweden, in collaboration with the Walter Reed Army Institute of Research (WRAIR). Britta Wahren of the Karolinska Institute provided a progress report on this trial and says that so far 57 volunteers have already received at least one dose of the DNA vaccine candidate or placebo and 21 have already completed all three doses. Administration of the MVA boost, which was developed by researchers at WRAIR, is expected to begin in the first volunteers later this year.

Presentations on these similar, but different, prime-boost regimens prompted multiple questions on how these combinations would stack up in a head-to-head comparison. Without comparative studies it is difficult to know how any of these regimens would compare to MRKAd5, because they all include different immunogens and many also use different nonhuman primate models for preclinical evaluation. Researchers also use different criteria to determine immunogenicity. This makes it nearly impossible to establish which of the current strategies is the most immunogenic, and on several occasions during the meeting, researchers outside the AIDS vaccine field asked why comparative studies aren't conducted as part of preclinical evaluation. "Some people really don't want to compare their vaccine to each other's," says Liu. Nabel says there are several complications to doing these types of studies, but added that the VRC has created a standardized HIV Env insert that is available to all researchers and can be used to eliminate at least one of the variables between experiments.

The bandwagon

It's not only HIV researchers who are exploring prime-boost regimens—they are currently in vogue in other fields as well. Recently, there was some good news in the malaria vaccine field. The most advanced of a slew of candidates is being developed by GlaxoSmithKline Biologics and a recently completed Phase II safety study in Mozambique showed that it was 65% effective at protecting infants from malaria (Lancet 370, 1523, 2007). Phase III efficacy studies with the candidate, known as RTS,S or Mosquirix, will begin next year and if similar results are observed, the first potentially licensable malaria vaccine may be available as early as 2011.

But over the last few years, researchers working on malaria vaccines have also developed a heightened interest in using viral vectors to target the disease during a different stage of the parasite's lifecycle—when it is released from the liver into the blood. At this stage of malaria, cellular immune responses are a critical component in controlling disease progression or eliminating the parasite.

Researchers, including Hill, have tested DNA plasmid vaccines alone and in combination with MVA vector-based candidates, as well as a heterologous prime-boost combination of a fowlpox vector-based candidate and an MVA candidate. In mouse studies, the fowlpox/MVA prime-boost regimen was the most protective, generating CD8+ T cells that correlate with protection against malaria, using the IFN-g ELISPOT assay. Clinical trials conducted in the UK and the Gambia also showed high levels of immune responses in human volunteers, but when this prime-boost regimen was tested in a Phase IIb clinical trial in Kilifi, Kenya, it showed no efficacy. Hill says the immunogenicity of the vaccines was markedly lower in areas where malaria transmission occurs more frequently. Children with high quantities of parasite in their blood had the lowest immune responses to the candidates. Hill speculates that this may be a recurring problem for malaria vaccines in high-burden areas, where the vaccines could also have the greatest impact.

Following this failure, researchers set out to find a better heterologous prime-boost regimen and because more T cells correspond to better protection, Hill says, he and others tried to identify vectors that could induce even higher levels of cellular immune responses. This led them to explore using adenovirus as a vector. "Adenovirus vectors have in many ways been the high-flying vectors," says Myron Levine of the University of Maryland. "This [the outcome of the STEP study] does not mean adenovirus would not be a great vector encoding other antigens."

Hill's group at Oxford compared the immunogenicity of different serotypes of human adenoviruses with simian versions and found that chimpanzee adenovirus serotype 63 (AdCh63) was even more immunogenic than human Ad5. A prime-boost regimen with the AdCh63, followed by an MVA-based candidate encoding TRAP (a multiple-epitope fusion protein from the sporozoite stage of the parasite) induced 3,000 IFN-g producing T cells per million peripheral blood mononuclear cells (PBMCs) in rhesus macaques.

Hill is currently preparing to begin a Phase I trial to test this AdCh63/MVA prime-boost regimen in humans. This will mark the first time a chimpanzee Ad has ever been tested in human volunteers. "There's a lot of interest in adenovirus vectors for malaria at this moment," says Hill.

Chimpanzee adenoviruses have also been of keen interest to AIDS vaccine researchers, but as of yet no candidates have been advanced into clinical trials. This vector may hold even greater interest in the future, based on the results from the STEP trial.

Before and after

Without question, there are still substantial scientific challenges facing the development of new vaccines against the most pervasive global health threats. "Science is the critical ingredient for success," says Regina Rabinovich of the Gates Foundation in the opening keynote address of the Keystone conference. "You can't get there without it."

But science is not the only barrier. Along with the scientific challenges, there are others that occur after effective vaccines are licensed for public use, including manufacturing capacity and vaccine production, as well as vaccine delivery and administration. "Finding a new way of creating a vaccine is only half the issue," says Duncan Steele of the World Health Organization (WHO). Despite high-flying success stories of late, like the licensure of effective vaccines against human papillomavirus (HPV) and rotavirus, there are still many issues to resolve about how best to deliver these vaccines to the world's poorest people. Many of these issues aren't resolved before vaccines are licensed and this accounts for the sometimes lengthy lag time between the introduction of vaccines in rich and poor countries.

Immune responses to vaccines can also vary in different populations, so even when a vaccine is delivered successfully, it still may not provide optimal protection to everyone. For this, critical lessons can be learned from the delivery of already licensed vaccines. There is documented evidence of vaccines inducing varying levels of antibody responses in different regions of the world. Overall, vaccines that are administered orally tend to induce greater immune responses in industrialized nations in agreement with the hygiene hypothesis, which postulates that in richer countries there aren't as many enteric viruses or bacteria competing for the immune system's attention.

"The 'normal' gut is very different in developing and industrialized countries," says Levine. "In kids in developing countries the innate immune system is turned on full volume," he adds. As a result, the vaccine gets "laughed at." The responses induced by the live oral cholera vaccine are just one example of this phenomenon. Greatly diminished immune responses to this vaccine have been observed in Brazil, in children of low socioeconomic status in Peru, and in Indonesia, where a higher dose of the vaccine is required to achieve similar levels of immunity. For rotavirus, several of the earlier live oral candidates failed to work at all when tested in developing country populations.

But so-called conjugate vaccines—those made by joining an antigen to a protein—typically work better in developing countries, Levine says. The Hib vaccine is one example of this phenomenon. While only 10% of US infants reach the required level of serum antibodies against Hib after a single vaccination, 29% of infants in Chile reached the same level after one shot. Based on this observation, the government funded a study to evaluate fractional or partial doses of the vaccine, which at its full dosage cost more than all of the vaccines that were currently part of the country's expanded program on immunization.

This study showed that in Chile there was no difference between administering a third, a half, or a full dose of the Hib vaccine. The Chilean government never introduced these fractional doses of Hib vaccine because its cost was eventually covered by GAVI. But this case suggests it may be possible to get equivalent protection in some populations with less vaccine and, as the cost of newly-licensed vaccines soars, this could translate into a substantial savings. Levine suggested that studies to quantify the level of antibody required for protection for new and expensive vaccines, like those against HPV, are vital so that determinations about the dose required for protection can also be made.