AIDS answers and questions

VaxGen's AIDSVAX vaccine, based on a recombinant gp120 coat protein, failed to prevent HIV infection or even slow the development of disease in its second phase III clinical trial, conducted in Thailand and reported this month.The results were even worse than those from the first trial in the United States and elsewhere, reported in February. Nevertheless, many important lessons for future HIV vaccines can be learned from the two studies, said José Esparza, coordinator of the World Health Organization/United Nations Programme on HIV/AIDS HIV Vaccine Initiative. “These trials moved the field ahead,” Esparza told The Scientist this week. “It's a bit naive at this time to believe that we will do a trial and get a vaccine at the end. After 15 years of trials, we have had to modify our expectations, and rather than taking shots in the dark, hoping that this or that vaccine will work, we have to use them to learn.”In terms of the science of vaccine development, several lessons have now been learned. The VaxGen trials showed that antibody induced by this vaccine against gp120 is not protective, according to Esparza. “Now there are already several groups working on second generation gp120, or other envelope vaccines, using deglycosylation or tinkering with deletions, making trimers as in the natural protein and so on.” If, in phase I trials, those vaccines produce the same quality and quantity of antibodies as the vaccines just tested, there may be good reasons to believe that they won't work either, he said. “This is important information to guide design of envelope based vaccines.”The VaxGen vaccine was designed to produce antibodies—another approach is to induce cytotoxic T lymphocytes (CTLs). A number of candidates use a prime-boost combination, for example DNA followed by modified Vaccinia Ankara virus (MVA). Other CTL candidates are in the pipeline using prime boost with adenovirus, or boosting with Semliki forest virus or Venezuelan equine encephalitis replicons. Other groups, like Merck and Aventis, are now working together to combine vaccines.A further challenge to scientists is to determine if existing concepts, such as DNA with MVA, adenovirus, Semiliki forest virus, and so on, are effective in humans. “But we already know they are not highly immunogenic. For instance, with DNA/MVA, we don't see CTLs in all the volunteers. And the CTLs are not constant in time, from month to month they may vanish," Esparza said. "It's not very consistent. With canarypox, only 30% to 40% of volunteers show CTLs at any point. Now the question is—Is this enough?”One limitation may be in researchers' ability to measure immune responses in the lab, he said. “So people are tempted to move ahead to efficacy trials to find the answer, and in fact, last month the world's third efficacy trial started in Thailand. It's a prime-boost with canarypox boosted with gp 120, sponsored by the US National Institutes of Health and the US Walter Reed Army Institute of Research. It will enroll 16,000 people, with results available if everything goes to plan in 2008.” “But to get a useful HIV/AIDS vaccine, additional basic science has to be done. And we have learned it must be done in very close collaboration with clinical trials, in an iterative process: Do a trial, learn, go back,” Esparza stressed.One option is to develop more immunogenic antibody vaccines, possibly vaccines that induce antibodies that neutralize primary or clinical isolates of HIV, which the VaxGen candidate couldn't do. Lab isolates of the virus have been adapted to grow in tissue culture and tend to bind to CXCR4 (X4) coreceptors rather than CCR5 (R5), which are used by clinical isolates, Esparza noted. “Initially, most of the gp120 vaccines were based on X4 strains. Bob Gallo has some hybrid constructs, a chimeric protein of gp120 and CD4, that can induce antibodies that seem to be able to neutralize primary isolates and different subtypes. This is probably because gp 120 complexed with CD4 opens up and exposes cryptic regions. We can learn that kind of thing from structural biology. This is a booming field.”The failure of the VaxGen vaccine trials also teaches lessons about trial design. First, it shows that large-scale complex trials can be conducted with the proper scientific and ethical standards, Esparza said. Concerns that participants would increase—or decrease—their risk behaviors or be impossible to follow for 3 years proved unfounded, he said, even among 2500 drug users in the Thai trial. “We've also learned that the next trials will have to consider gender and race,” Esparza said. The first AIDSVAX trial offered hints that the vaccine may have protected blacks and perhaps women, but numbers of participants were small. The next phase III trial should have enough volunteers from these groups and others to allow statistically significant conclusions, he said.Also, drug users are attractive for trials because incidence is so high and because they'd benefit from a vaccine, Esparza said, but it may be more difficult to protect people who've been infected by injection than those infected sexually, in which the first challenge is mucosal. “We know that from monkeys—it's very reproducible, across several vaccines. That may explain why the efficacy in Thailand was so flat, zero, nothing,” he said. “A vaccine trial in drug users may show nothing, but could yet be effective in cases of sexual transmission.”

AIDS answers and questions

Interested in reading more?

Become a Member of