Treatment as a means to prevent HIV infection has hit the media following a declaration by Brian Williams, professor of epidemiology at the South African Centre for Epidemiological Modelling and Analysis in Stellenbosch, at the AAAS in San Diego.

Whilst the HIV epidemic has shown some signs of stabilisation in the recent years, more than 7,000 people are still infected every day with the virus that causes AIDS.

Despite interesting results of a vaccine trial in Thailand, prevention is still limited to a small number of options many of which are not bullet-proof and biomedical interventions based on vaccines and microbicides are still a long shot away. Conversely, treatment is working very well in bringing HIV-infected people back to a normal life and potentially reducing the risk of HIV transmission by reducing their viral load.

But would it work? Would putting all those infected with HIV on ARV halt the epidemic?

Mathematical modelling suggests that treatment as prevention could deliver more rapidly and for a lower cost on the long term. We know from the prevention of mother to child transmission that if you treat an HIV-infected mother before she gives birth, the risk of her baby being infected is considerably reduced.

But there is dissent and concerns in the chorus of ecstatic voices. Elizabeth Pisani (author of the recommended Wisdom of Whores) expressed hers in a column for the Guardian.

One of her reasons to question the “Test and Treat” approach is that, so far, more treatment means more new infections and she explained why. Her comment received mixed responses but a strong rebuttal came from The Bathhouse who wrote “Pisani seeks media attention by making a point in an inflammatory and simplistic way.”

Who is right? Who is wrong? Is someone right?

As always, it is not simple, but it is certainly not simplistic.

Using gay men in the UK (a relatively contained population, which is important for such kind of epidemiological study) we can have a good idea of what happens when treatment is made more available. To date, 70% of HIV positive gay men know their status. 70% of these are on treatment which they started once their CD4 count went below a certain level, 350 cells/mm³.

This means that ~50% of all those gay men who should be under treatment according to the Test and Treat approach (where HIV+ receive treatment whatever their CD4 level) are currently receiving treatment. Is this enough to slow down the epidemic in gay men in the UK? The answer is in the data and is “no” (Thanks to EP for prompting this example).

Number of new HIV and AIDS diagnoses, numbers accessing care and deaths in HIV infected MSM, UK

Why, as suggested above, is because 30% of those who are HIV positive still don’t know they are infected and that only those with a CD4 count below 350 started receiving treatment (note that promiscuity as nothing to do with it as long as it is practised with condom).

Does this means that the Test and Treat approach does not work? No. It simply means that not enough people are receiving ARV and that those receiving ARV are receiving them too late.

Why is it important to receive treatment as soon as possible after infection? Because it is at this point in time that an infected individual’s viral load is at its highest and that (s)he is the most infectious.

HIV Time course. Viral load is at its highest in the first months following infection (red) and in the very late stage of the infection. This is when people are most infectious. Detecting HIV infection early and treating it immediately could prevent further infections.

The question “will treatment as prevention work if we could put everybody infected on treatment” then becomes a very different question: “Can we detect all HIV infections early enough and put everybody testing positive on treatment?” Should this be possible, then there is a whole new range of questions to answer including acceptability, human rights, and economics cost.

This being said, another of EP’s point needs to be addressed, “We don’t need to build computer models based on entirely unrealistic assumptions in order to justify the need for more treatment.”

We don’t. But money currently invested in programmes we know don’t work, such as abstinence-only programme, or programmes for which evidence that they work are still lacking, such as serosorting/strategic position and the likes (I am going to make some friends here), or programme that are going to screw up fragile health care systems, such as PrEP (more friends here) could be diverted toward building up capacities to test, treat and follow up HIV infected people so as to ensure that their viral load is controlled early enough to reduce the risk of HIV transmission through unprotected sex.

After all, EP makes a very good case in her book on how “The Widows and the Orphans” successfully attracted funding for HIV whilst those most in need where sex-workers, gay men and drug users…

Related posts brought to you by Yet Another Related Posts Plugin.