Pic: Â© Ciao-Chow/Flickr
The recent news that for the first time an HIV vaccine had shown some protective effect generated widespread excitement, until it emerged that the results were based on the most promising of three different analyses of the trial findings.
The trial team in Bangkok, Thailand’s capital, announced on 24 September that a combination of two vaccines had reduced the rate of HIV infection by 31 percent in about 8,200 volunteers, compared to around the same number who were given a placebo.
A few weeks later, researchers who had seen full data from the trial told Science magazine that an analysis based only on participants who had received all six doses of the vaccine at the right times did not show a statistically significant protective effect.
It was hoped that the release of more details from the trial to coincide with the AIDS Vaccine 2009 conference taking place in Paris this week would settle the question of whether the vaccine results were really as significant as the initial announcement had suggested or a mere fluke. Instead, full results of the study, published online yesterday in the New England Journal of Medicine (NEJM) raised more questions than they answered.
The 31 percent efficacy in the initial announcement was based on a “modified intention-to-treat” analysis that included all the 16,402 trial participants, except for seven who were found to have contracted HIV before receiving any vaccinations.
A second analysis included those seven, while a third “per-protocol” analysis involving 12,452 participants – the one cited in Science magazine – found that the vaccine was only 26 percent effective. This was not enough to be statistically significant, meaning that the difference between the vaccine and the placebo arms of the trial was so small that it could have been a coincidence.
Dr Jerome Kim of the US Military HIV Research Programme, who helped lead the trial, yesterday told reporters at the vaccine conference in Paris that the modified intention-to-treat analysis was the most accurate, but others disagreed.
A statistician quoted in a New York Times report placed more emphasis on the analysis that included the seven HIV-positive participants, while another did not believe that any of the analyses provided sufficient evidence the vaccine worked.
In an editorial accompanying the article, NEJM editor Raphael Dolin said that “although the merits of each type of analysis can be debated, all three yielded a possible, albeit modest, effect of the vaccine in preventing HIV infection.”
The study authors also argued that, taken together, the three different analyses of the results were “consistent with a modest protective effect of vaccine”, but could not explain why other findings from the trial indicated that the vaccine’s efficacy appeared to decrease over time, or why it was less effective among participants at high risk of infection.
They were also unsure whether it was one of the two vaccines that produced a potentially protective effect, or the combination of the two. Dolin noted that the findings raised “a number of questions that have important implications for future directions in vaccine research”, and recommended that the duration of the vaccine’s effect be addressed by following up the trial participants, as well as by future trials.
According to a report by a South African news service, Health-e, Colonel Nelson Michael of the US Military HIV Research Programme, another lead investigator of the Thai trial, told a press conference in Paris that a further study of the vaccine may be conducted in South Africa, which has a much higher HIV prevalence than Thailand. The vaccine would have to be modified to contain the strain of HIV most common in sub-Saharan Africa.