Friday, July 23, 2010

How Convinced Were You by the Microbicide Story?

The six day Vienna Aids Conference has come to an end and the freeloaders will have to finish up their last free drinks until the next junket. The press has excelled itself by repeating, almost word for word, everything they were told to publish. Searches for any of the conference catch phrases, such as 'Treatment 2.0', 'treatment as (or 'is') prevention', 'rights here, right now' or 'Tenofovir gel', etc, will yield thousands of results, all equally uninformative.

Outside the mainstream press, the Huffington Post has an article that questions the big story of the week, the announcement that a vaginal gel, using the antiretroviral drug Tenofovir, has shown some promise in trials. The author, John R Talbott, points out that the results are not really as exciting as they seem and that using this gel will only give a relatively small amount of protection. Over a lifetime, the majority of people in a population like the one in the trial will still eventually become infected, even if they use the gel all the time.

The paper entitled 'Effectiveness and Safety of Tenofovir Gel, an Antiretroviral Microbicide, for the Prevention of HIV Infection in Women' is indeed very unconvincing. Even the authors must feel pretty uncomfortable about the results of their trial and the many unanswered questions thrown up. In the introduction, they mention that in the past three decades "Only five of 37 randomized controlled trials, which tested 39 HIV prevention strategies, have demonstrated protection against sexual transmission of HIV infection". Three of the five are circumcision trials, which were well publicized but not very encouraging. The fourth was a vaccine in Thailand, which was also disappointing.

The fifth was a sexually transmitted infection (STI) treatment trial in Tanzania. Similar trials were carried out in other locations at the same time but they didn't result in any reduction in HIV transmission. The trial in Tanzania seemed to, except that there was an injection safety trial going on at the same time in the same place. It is highly likely that this affected the results of the STI trial. Yet, even though those working on it knew about the injection safety trial, they failed to mention it until many years after publishing their initial results.

The history of HIV prevention is notable not so much for its failures as for the way those failures have been dressed up as successes. The Vienna Conference has pushed the Tenofovir gel trial as a great breakthrough and the press have bounced the story around the world (and the www). The authors of the paper may well be regretting that their work has been chosen to be hyped above all others this time around. They, of all people, are probably painfully aware that we are still a long way from a pharmaceutical product that will have a significant impact on HIV transmission.

The trial itself is quite tentative, involving a small number of people. And the scope of the research was very narrow. One of the most shocking things is how high incidence is, in both the intervention (5.6%) and the control (9.1%) groups. If you had hundreds of people whose sexual behavior you were studying, you might want to investigate their sexual behavior. This trial didn't really do that. They didn't test the participants' partners. They don't even know if those who contracted HIV did so through sexual behavior! Or, if they did check the partners, they don't mention this in the paper.

Another extraordinary thing about the trial is that condom use was high and steady throughout the trial, in both intervention and control groups. Condom manufacturers must be wondering exactly what the participants did with the condoms to have such poor results with them.

Over the trial period, the number of sex acts involved was not particularly high. Even if none of the participants had used condoms, incidence of between 5.6 and 9.1% would be hard to explain. The authors speculate about the women having sex with 'migrant workers' but they don't say if the women did so. The trial selected women that were at low risk of being infected with HIV, so why did so many turn out to be at high risk?

Not only did the women in the trial not have sex particularly often but they had sex less and less often as the trial went on. The number of sex acts per month went from 7.2 to 3.1 per month. The more people were having sex, the less likely they were to use the gel and those who were having sex least frequently were the most likely to use the gel. But frequency of sex doesn't seem to change the risk of becoming infected. It's as if sexual behavior is not especially relevant to the results of the trial.

Talbott touts his own theory of why HIV rates are so high in Africa. He concludes that it is because of the "numerous and informal sexual affairs common in Southern African nations, found both among married and unmarried men, especially with a much smaller highly infected group of very sexually active women who trade sexual favors for material goods and money". There is, in fact, no evidence that 'Southern Africans' have more 'informal sexual affairs' than many people in other continents (or countries). But the Tenofovir trial itself even finds that sexual behavior is not very frequent. Nor does it appear to involve a lot of 'unsafe' sex.

And, like many people who rush to such conclusions, Talbott also doesn't explain how large groups of sexually promiscuous men and small groups of sexually promiscuous women can give rise to far more women being infected than men. Perhaps he would, along with others, say that all these men go home and infect their wives. But about half of the HIV positive married women have HIV negative husbands. His argument suggests, as all arguments do that try to explain high HIV prevalence by reference to sexual behavior alone, that Africans have a lot more unsafe sex than non-Africans and that many of the women are exceptionally promiscuous.

Firstly, we know that Africans don't have a lot more unsafe sex than non-Africans, secondly, we know that African women are not exceptionally promiscuous and thirdly, it has also been shown that even very high rates of unsafe sexual behavior do not explain the levels of HIV prevalence found in some African countries. I applaud Talbott for criticizing the dubious claims being made about the Tenofovir gel trials. But he doesn't do much for his credibility, or that of the Huffington Post, by falling back on the racist and sexist views of the HIV  industry.

Talbott calls for testing. What is needed is investigation. Why are so many women whose lifestyle is not risky becoming infected with a virus that is difficult to transmit? Why were their partners not tested, or if they were, why were the results not published? Who cares whether the gel works or not when we don't even know how the women were becoming infected? Like all the other 'major breakthroughs' discussed at the conference, this gel will only be useful if HIV is mostly sexually transmitted. This is far from evident.


No comments: