The subject was already being discussed before the main conference in the
satellite 'preconference' organised by the Global Forum for Men who have Sex with Men and HIV
(MSMGF) the day before the main conference opened.
Bob Grant, Principal Investigator of iPrEx, the
proof-of-concept study of PrEP in gay men, told the preconference that in his
view it was misleading to quote the 42% efficacy observed as if this was the
highest achievable in gay men. Drug
level studies in iPrEx (and in the Partners
PrEP study in different-status heterosexual couples) had shown that the
reduction in the risk of HIV infection in people with detectable drug in their
blood - implying adequate adherence levels - was in the order of 90-92%, and
that in iPrEx, in the relatively few subjects at US sites, drug had been
detectable in 94% of samples.
In addition, in iPrEx as a whole, adherence had correlated with risk:
participants who had unprotected receptive anal sex had higher adherence to
PrEP than ones who didn't and PrEP efficacy was somewhat higher in this group
as a result (53%). For Grant, this showed that PrEP is likely to be used by
those who need it most. He recommended that use of PrEP should be guided by the
person's request to have it more than the physician's judgement that they need
it, and that PrEP and other biomedical prevention interventions are likely to
work better in a non-judgmental atmosphere where stigmatising language is
In the main conference, in a symposium that
discussed papers published in a
special issue of The Lancet that
focused on men who have sex with men (MSM), Patrick Sullivan of Emory
University, Atlanta presented
a model of the likely reduction in HIV infections in men who have sex with
men (MSM) in four different countries (Kenya, the US, Peru and India) using
three different prevention programmes: one with an intensified emphasis on
condom use as its primary ingredient, one focusing on earlier treatment for MSM
with HIV, and one adding PrEP to existing prevention programmes.
The base model used assumed one of the following three scenarios:
- Uptake of PrEP in gay men needing it at rates
ranging from 20% to 80% and also at various adherence rates, ranging from 50%
increase, ranging from 20% to 80%, in the proportion of men taking ARVs with
CD4 counts above 200 cells/mm3 or (in the USA) 350 cells/mm3
(these being the ART thresholds at the time the model was done);
increase in condom use over baseline from 10% to 40%.
His model found that PrEP, under these scenarios, would be slightly more
effective than condom promotion in all countries other than India, though only
by a couple of percentage points in terms of the proportion of HIV infections
averted that would otherwise have happened. For instance this proportion would
be about 23% with 40% PrEP uptake, and about 20% with a 20% increase in condom
In all countries but Kenya, the model suggested that earlier treatment
would have less effect than the other two interventions. The reason condoms
would be more effective in India is because usage rates in MSM in that country
are, at least in studies, higher, so a percentage rise in condom use would involve a greater number of extra condoms being used and have a
greater effect on the remaining HIV transmissions. Conversely, in a country
like Kenya where people currently start ART later than elsewhere, the
consequence of starting it early is greater.
Sullivan said his model predicted that, using the most realistic
scenarios, 25% of infections could be
averted over the next ten years in MSM using a combination of these methods.
However this would happen only if the criminalisation of male/male sex, threats
and violence against MSM, lack of understanding and training among healthcare
workers, and barriers against implementation research were addressed.
In a poster session, Kate Mitchell
of the London School of Hygiene and Tropical Medicine found that the
effectiveness of PrEP would vary widely according to how carefully gay men were
targeted for it, using India as an example.
India has had quite a segregated MSM population historically, with
feminine, gay-identified ‘kothi’ men who take the receptive role in sex and
non-gay-identified ‘panthis’ who take the insertive role tending to stay in
their separate populations (this means panthis, who do not identify as gay, are harder
to reach). A third identity as ‘double decker’ or ‘dupli’, i.e. versatile, has
grown up amongst urbanised gay men more recently.
In her model a PrEP intervention targeted at kothis
or duplis, in which the intervention was 42% effective and was taken up by 60%
of targeted men, would prevent 25-30% of HIV infections, whereas if PrEP were
targeted at panthis, or selectively used by them, the result would be only a
5%-10% reduction in HIV infections.