The last day of the Vienna International AIDS Conference featured
the results of the second-ever completed randomised controlled trial of
pre-exposure prophylaxis (PrEP) – giving antiretrovirals to HIV-negative people
at high risk of HIV to prevent them becoming infected.
The trial, which randomised 400 gay men to take either a daily
tenofovir (Viread) or a daily placebo pill, was too small to serve as an efficacy trial.
Lisa Grohskopf of the US Centers for Disease Control (CDC)
told the conference that none of the seven men who became infected with HIV
during the trial was taking tenofovir but, because of the size of the trial,
this could possibly be a random effect
– nothing could be read into it about
the efficacy of once-daily tenofovir as a means of preventing HIV.
This is a similar situation to the only other previous PrEP trial, from Ghana, which a reported a non-significant reduction in HIV infections in women taking PrEP (see aidsmap report: First hint that PrEP might work in humans).
Inevitably, though, it will be taken as a promising signal
in advance of the next trials to report – the CDC 4370 trial in Thai drug users
(see aidsmap report: Early data from the
Thai trial ) and the predominantly South Amercian iPrEX trial
in gay men (see more on community consultation
on aidsmap), which are due to report at the beginning of next year.
This trial, however, was intended as a safety trial, looking
closely at side-effects of the tenofovir medication and especially at risks
already associated or suspected of association with tenofovir, such as kidney
failure and bone mineral loss.
The trial recruited gay men from San Francisco, Atlanta and
Boston, and randomised them into four groups of 100. Two groups took either
tenofovir or a dummy pill from the start, for two years. The other two waited
for nine months before starting either tenofovir or placebo, and took them for
the remaining 15 months. This was in order to have data from a control group
that took no daily pills at all, in order to find out if taking a daily pill
changed sexual and risk-taking behaviour.
In the end, 373 men started taking tenofovir or placebo and 323
completed the study, a retention rate of 86%.
The average age of the participants was 38.5 and there was a
nearly significant difference in the racial balance of the trial, with 18% of
those on placebo being African-American and 12% of those on tenofovir (p=0.07).
Participants had had, on average, four male partners in the
last three months; 29% of those on tenofovir and 33% on placebo had had
unprotected sex in the last three months (not a significant difference); and
12.5% of the participants had had risky sex, i.e. unprotected sex with someone
who was or could have been HIV-positive.
There was virtually no difference in the rate of adverse
events seen between patients on tenofovir and those on placebo. The only side-effect in which a significant difference was seen was back pain, suffered by
13% on tenofovir and 6% on placebo (p=0.04), and there were no differences in
the rate of any serious side-effect.
In particular, there was no elevation whatsoever in
creatinine levels (an indicator of kidney problems) and the rate of low blood
phosphate, an indicator of possible bone mineral loss, was if anything higher
in participants taking placebo, although not significantly so (p=0.12).
One person died during the trial, of an accidental overdose
of opiate drugs with alcohol.
Seven participants tested HIV-positive during the trial. One
of these was clearly someone recruited in the window period for HIV infection;
he was HIV-negative when screened, but HIV-positive on the first day of
enrolment, with a viral load of 1770 copies/ml.
Of the other six, three were on placebo when they contracted
HIV and the other three were in one of the delayed-treatment arms and had not
started taking either tenofovir or placebo. No tenofovir resistance was seen.
Sexual risk behaviour did not change significantly during
the trial. At baseline, the proportion of men who said they had had unprotected
anal sex during the last year was 60% in the immediate-treatment arm and 53% in
the delayed-treatment arm (note that this was over a longer period than the
baseline questionnaire, which asked about sexual behaviour in the last three
It went down from the beginning of the trial to the next quarterly visit, an
effect often seen in trials, and went down more steeply in those randomised to
immediately take a tenofovir or placebo pill, to 46% – suggesting that taking a
daily pill may serve as a reminder to use condoms rather than a disincentive.
It rose very slightly and non-significantly thereafter to about 55% in the
immediate-treatment arm, but did not rise at all in the delayed-treatment arm.
Further analyses are continuing, which will look into risk
behaviour more deeply, though as a member of the audience commented, it’s much
less likely that people would take more sexual risks if they knew they might be
taking an inactive placebo than if they were sure they were on tenofovir.
Analyses will also look directly at bone mineral density and assess adherence.