New directions in HIV prevention: circumcision and PREP

One study already reported from Rakai in Uganda found that female partners of circumcised men had a 30% lower risk of acquiring HIV and several sexaullly transmitted infections (STIs) (HPV, genital herpes, trichonomiasis and bacterial vaginosis) than partners of uncircumcised men Although this difference was not statistically significant, in a plenary on prevention, Professor Tom Quinn of Johns Hopkins University said that this finding is not new and made some forecasts as to how increasing the 20-25% of the world’s male population who are circumcised might slow the spread of HIV.

Epidemiological studies have found that in both Africa and Asia there was a strong association between circumcision and HIV prevalence. In countries with high rates of male circumcision (over 20%), HIV prevalence is an order of magnitude lower than in countries where it is uncommon.

One study from as long ago as 1988 showed that African men who visited sex workers were four to six times less likely to acquire HIV if they were circumcised. The same study showed that the presence of genital ulcer disease (GUD) in the men accentuated this effect so that HIV prevalence in circumcised men without GUD was 26 times lower (2.5%) than it was in uncircumcised men with GUD (52.6%). A meta-review in 2005 had found that circumcision had between a 60% and 80% protective effect.

The finding that women were at lower risk if their partners were circumcised was not new either, he pointed out. In 2000 he had authored a study himself from Rakai that showed an overall protective effect to the female partner of 26%.

But this depended on the man’s viral load. If men had viral loads of over 50,000 copies/ml, circumcision had no protective effect on women. But among men with viral loads under 50,000 copies/ml, there were zero transmissions from circumcised men, compared with 26 transmissions from circumcised men.

What would the effect be of widespread circumcision? In a best-case scenario if 100% of men were circumcised and circumcision prevented 70% of male infections, HIV incidence in the whole population would be reduced by one-third even if it had no protective effect on women. However if the protective effect on women was included, HIV incidence would fall by two-thirds if all men were circumcised and by 40% if half of them were.

However, Quinn warned, if circumcision produced a perception in men that they were protected from HIV and could go back to having multiple partners, all benefits of the measure would be lost. Using the same projections, if the average number of partners men had increased by 50% HIV prevalence would increase by 40%; if the average number of partners doubled, prevalence would more than double and would be higher than it is today.

The biggest news at one of the scientific conferences last year, the IAS Conference in Rio, was the finding that circumcision may reduce the chances of men acquiring HIV from women by up to 75%. There was no similar revelation at CROI this year, but a track specifically devoted to prevention, two plenaries on the subject and a number of studies clarified the directions HIV prevention may take in the 21^st century to become more effective.

Pre-exposure prophylaxis (PREP) involves the use of oral antiretroviral treatment to prevent HIV infection. Although studies of PREP using tenofovir (Viread) have been halted in Cambodia and Cameroon due to local pressure, there has been no suspension of scientific interest in PREP, as this year’s Conference on Retroviruses and Opportunistic Infections showed.

In terms of pre-exposure prophylaxis, the study reported here, which successfully used a combination of tenofovir and FTC to prevent infection in monkeys given 14 rectal challenges of monkey-adapted HIV, revived interest in the concept, after a similar trial using tenofovir alone presented at last year’s CROI ended up with all monkeys eventually infected. It should not be surprising that ‘comboPREP’ should be of greater efficacy than single drugs, though using FTC alone in this trial also produced superior results to the 2005 one.

Another study already reported here found that FTC, tenofovir and 3TC (lamivudine, Epivir) achieved concentrations four to six times higher in genital fluids than in blood, far in excess of most other antiretrovirals, thus showing that whether by luck or judgement the drugs used in PREP studies were very likely the most effective ones.

As a prevention intervention PREP is controversial for two reasons. First is the perception that it will be prohibitively costly as a widely-used prevention measure. A poster by UK researchers (Hill) disputed this. They took the cheapest current annual prices for antiretroviral drugs in the developing world (about $50 for 3TC and $200 for tenofovir), and estimated the average lifetime treatment cost for a developing-world patient at $10,000. Using these figures, they calculated that 3TC PREP would be cost-effective (that is, would cost less than $10,000 per infection averted) in areas where annual incidence was over 1% (as in most of sub-Saharan Africa) and tenofovir/3TC would be cost-effective where incidence was over 3% (as in southern Africa and vulnerable populations in many other areas).

In situations of extremely high incidence (as in young women in South Africa), 3TC PREP might cost as little at $600 per infection averted a year.

The second reason PREP has been controversial is due to the perception within communities and community leaders that the approach may do more harm than good, whether through toxicity, resistance, behavioural disinhibition or unethical practice by researchers, and two trials have already been halted because of these perceptions.

The largest PREP trial yet to be considered is among 1600 men who have sex with men in Lima, Peru. It will start to enrol this spring after considerable delay.

A poster (Goicochea) explained how the Peruvian Institute for Health and Education, in collaboration with the trial investigators and informed by the failure of previous PREP trials, has engaged in an exhaustive series of consultations with different stakeholders prior to recruitment. These included opinion leaders such as academics and gay and HIV activists, but also every group within the MSM community including men who sell sex, ‘straight-identified’ MSM, transvestites and transsexuals as well as gay-identified men.

The findings are too long to be summarised but revealed divergent opinions among the affected groups. Academics and gay activists, for instance, thought that reimbursement for study participants would distort the results, whereas HIV activists and male sex workers thought payment to trial participants was essential.

The Lima consultation, especially in its disinclination to view one group as representative of ‘the community’ and its attempt to consult all possible stakeholders, is probably a model for the kind of pre-recruitment community consultation that prevention trials need to undertake.

Lastly, a trial of valaciclovir as anti-herpes prophylaxis to prevent HIV co-infection (reported here), which found that herpes treatment roughly halved the number of women who shed HIV, reminds us that PREP need not just involve HIV drugs.

Myron Cohen of the University of North Carolina told the conference that he thought that treating people with HIV would, in the end, provide the biggest contribution to controlling the epidemic globally. He said he had been ‘charged’ by the National Institutes of Health to prove that antiretroviral therapy could reduce transmission. The result is HPTN052 – a randomised trial comparable, in its length and ambition, to the SMART Study whose closure provided the Conference’s biggest news.

It will be a seven-year study with patients actually on ARVs for five years. It will study 1,750 serodiscordant couples in eight sites in Malawi, India, Thailand, Zimbabwe, Brazil and Boston, Massachusetts. The enrolment criteria for the HIV positive partner will include a CD4 count of between 300 and 500 and the two arms of the study will compare the rate of HIV acquisition in the partners of people who are immediately given antiretrovirals on enrolment with the rate in partners of people who are only given ARVs when they develop AIDS (which includes the CDC definition of a CD4 count under 200 being AIDS). Infections will be compared by genotype to make sure they really are from the primary partner. The study will have a 90% power to detect a 35% reduction in HIV transmission.

Cohen acknowledged that there were many unanswered questions begged by this study, similar to ones posed by the PREP studies.

“Will it end up suggesting a strategy too expensive to afford?” he asked. “Or will we just end up exchanging sensitive for resistant virus?”

However, he defended the study’s concept by saying “We have no option but to study the effect of antiretroviral therapy on HIV transmission. To do otherwise would be total folly. We should have studied this when HAART came along, and we need to start the work that hasn’t been done in the last 10 years.”