Researchers at the AIDS Vaccine Conference in Bangkok, Thailand, Sept. 12-15 announced discovery of the first correlates of protection of a vaccine that protects against HIV infection. They hope to use this information to shape a better, more effective vaccine. However, a product that might prove useful in the US is still at least a decade away.
Only one of the many HIV vaccine trials has shown any sign of working. That was the RV144 study conducted in Thailand, which showed a modest 31 percent protection. The results, made public at the Paris conference in 2009, surprised many researchers, who did not believe the approach would work and had tried to stop it from going forward.
It has taken two years of additional analysis of stored blood samples from the 1600 trial participants to begin to figure out "clues to why it might have worked. That is something we haven't had over the past 30 years," said Barton Haynes, the Duke University researcher who led the effort.
The key findings involve antibodies IgG and IgA, two different types or classes of molecules produced by the body's B cells that fight invading pathogens, locking on to them and preventing HIV from entering cells.
"We found IgG antibodies that matched the scaffolded-V1V2 of the recombinant protein [ of the study vaccine ] , that correlated inversely with the infection rate. That means, the higher the IgG antibody, the lower the infection rate," said Haynes.
Second, they found that levels of IgA in the blood that binds to the outside shell of HIV "correlated directly with infection rates. The higher the IgA to envelop, the higher the infection rate."
He believes that the IgG is protective but when the IgA binds to a site, it prevents the IgG from doing so. That probably means that future vaccines should trip to generate a lot of IgG but little or no IgA.
When asked if he was surprised to have found correlates of protection, Haynes replied simply: "Yes."
Advancing the research
Jerome Kim, a US Army researcher deeply involved with the Thai vaccine trial, cautioned that the correlates they have identified may only apply to this particular vaccine when used against the strain of HIV common in Thailand.
HIV is a very diverse virus and mutates rapidly. There are a handful of major clades or strains of HIV, plus different hybrid combinations of those clades, and up to 30% variability in the genetic sequence within a particular clade.
And other vaccine constructs may stimulate production of different antibodies that target other portions of the virus, and hence other correlates of protection that are specific to that vaccine.
Sanofi Pasteur developed one part of the vaccine used in the RV144 Thai trial. Sanjay Gurunathan told the conference his company is moving forward in collaboration with other partners—including the Bill and Melinda Gates Foundation, and US, Thai, and South African government agencies — to conduct additional trials building on this information.
He noted that protection was as high as 60% in the Thai trial at one year, but it quickly declined. The collaboration plans to tinker with various components of the two different products used in the series of injections to try to get a stronger, more sustained antibody response.
Dr. Gurunathan said the collaboration partners are planning trials of regional vaccine candidates in Thailand in high-risk populations such as men who have sex with men and female sex workers, and in South Africa in high-risk heterosexuals.
Modeling has shown that a vaccine with 50% efficacy, administered to 30 or 60 percent of the highest risk populations, could have a significant public health impact. It could reduce the number of new infections by 5 to 15 percent over ten years and be cost effective.
Complexity
Vaccine trials are built upon a certain number of infections occurring; once that predetermined endpoint is met the trial is unblinded and researchers figure out if fewer infections occurred in the group who received the vaccine compared with those who received a placebo.
But that is getting harder to do with each passing year because more people have access to treatment and hence are less infectious. And prevention activities have improved and may do so even more during the course of the not yet started vaccine trials.
Voluntary circumcision is being rolled out in much of sub-Saharan Africa and the latest studies have shown that its effectiveness can be as high as 70% on a population basis when the "herd immunity" protective effect is taken into account.
Microbicides and PrEP studies also have looked promising and may be additional prevention tools put into place over the next few years.
"If you halve the incidence [ of new infections ] , you have to double the sample size" of the study, said James Kublin, with the HIV Vaccines Treatment Network. That can be very expensive and the NIH budget is not likely to increase.
Carl Dieffenbach, NIH's principle administrator for HIV research, said, "There are a series of contingency plans that are very robust, that indicate we can do these trials, assuming a level NIH budget."
Mombassa Sex Workers
Researchers have been studying a cohort of female sex workers in Mombassa, Kenya for many years. One of the surprising early findings was that despite having likely been exposed to HIV numerous times they either did not become infected with the virus or they were controlling their infection extraordinarily well.
Reports some years ago suggested that they remained protected so long as they were sufficiently exposed to the virus but became vulnerable to it after returning from an extended holiday from the sex business.
University of Washington researcher Valerie Cortez looked at the antibodies of these women over time in stored samples of blood. She found that "superinfection" with a second version of HIV produced an enhanced neutralizing antibody response that provided both broader coverage and cross-reactivity to the virus.
Only those who were "superinfected" developed those responses, suggesting that mere exposure to another strain of the virus was not sufficient to develop antibodies — which often take months if not years to add residues and mature — only a durable infection provided sufficient exposure to generate such a response.
Geographic location also might be important. Mombassa is where clades A and C of the virus overlap and where many mosaic recombinations of the virus clades occur. Cortez agreed with a suggestion from the audience that perhaps exposure to a sufficiently different virus is necessary to trigger a "superinfection" that generates the type of neutralizing antibodies she observed. It might not occur if a person were to becomes "superinfected" with different variants of the clade B virus that is prevalent in the US and Europe.
