How excited should we be about the results of the HIV vaccine trial in Thailand? I argued in The Times today that the results are the worst type of good news. The combination vaccine is good enough to raise real hopes that we can find something that works. After the gloom of last year, that’s no small thing. But even Vaccine Cheerleader-in-Chief Anthony Fauci is circumspect about yesterday’s announcement. (NPR interviews Anthony Fauci here)
There a couple of things that no-one much has mentioned. The first is that the vaccine was designed to produce a response to sub-types E and B. Subtype E (and an AE recombinant) is a form of HIV-1 that is very common in heterosexual HIV infections in Thailand, but much less common in other parts of the world. In most Western epidemics (and among gay men in other parts of the world, including Thailand) subtype B is most common. But as you can see on the map, in east and southern Africa, where around half of people with HIV in the world live, subtype C has the upper hand.
So even if we improve the efficacy of the combination that was tried in Thailand, there’s no telling whether it will be useable in Africa, where it is most needed.
Second, those who got the real vaccine combination but went on to get infected with HIV anyway had just the same amount of virus in their blood, on average, as those who got the placebo injections. This was a surprise; researchers had hoped that people who had been vaccinated would at least have lower concentrations of virus. It was also a disappointment, because lower viral loads generally mean less damage to the immune system (which in turn means it takes longer to get sick from other things, or to develop AIDS). Lower viral loads also mean that an infected person is less likely to pass HIV on to other people.
Some of the better scientific analyses (such as this report from IAVI) raise questions about just how robust the results will turn out to be once they are analysed more carefully. Interestingly to epi-nerds, none of the press releases or official fact sheets about the trial gave a p-value — a measure of the probablility that the difference between the two groups was due to chance. It fell the the ever-reliable Jon Cohen at Science to dig that out (and to Roger to find it for me at a bleary-eyed hour; thanks.)
It’s always easier to pick holes in these trials than to do them; whatever the results, the Thai Ministry of Public Health and the US Army deserve credit for having had the balls to push ahead with this difficult and much criticised piece of research.