Several of the microbicide efficacy studies, such as the Cellulose Sulfate (CS) and Savvy trials, have tried to keep the sample size manageable by trying to enrol the women they believe are most at risk of contracting HIV — but this can be challenging.
CS #1, for example, is recruiting women who have at least three vaginal acts per week, plus have had three different partners in the last three months and who expect to continue this behaviour, while CS #2 is looking at women between the ages of 18 and 35 who have at least three sexual acts per week and more than one partner in the last three months. The Savvy trials in Nigeria and Ghana both used similar enrolment criteria. All of these studies have a target accrual of under 3,000 women.
But there are downsides to this approach.
“Many studies focus on high risk women and this is good but it's not that easy,” said Dr Elof Johansson of the Population Council. “The problem we have found is that the high risk women are the most difficult to maintain.”
In other words, losses to follow-up, poorer adherence and pregnancy could both be more common among higher risk women.
Also, looking only for such high risk women can take longer. According to Dr Vera Halpern of Family Health International, who gave an update on the trial, CS #2 has had to screen more than twice the number of women to meet its target. A portion are already positive, or pregnant, but most of those screened but not enrolled simply do not return for enrolment. The slower than expected accrual has put back the study timeline, originally stated to be completed in 2007, by a year.
It can also pose dangers to the study team. According to a poster presentation, the Savvy Ghana team specifically tried to recruit as many women in high transmission areas and as many commercial sex workers as possible. They identified high transmission areas such as night clubs, beaches, hostels for commercial sex workers (CSWs) and internet cafes. Field staff were sent to visit some of these spots late at night in potentially dangerous areas, and were sometimes harassed by “pimps who claimed to be responsible for the ‘protection and welfare’ of some CSWs, hindered access to these potential participants, creating uncomfortable situations for the field staff.”
The team overcame these barriers by having female field staff work in pairs — with the aid of male informants — and by this method eventually recruited about 70% of the women who were screened at that site.
Higher risk women are often more financially disadvantaged and may have a low literacy level which can make the informed consent process more difficult. At one site in Nigeria, trial staff tried to tell whether women could read or not by handing them pamphlets upside down. "In Cellulose Sulfate trial #1", said Dr Lut Van Damme, "it's always a question of what do people really understand. The big problem in our study is illiteracy in regard to informed consent. The solution is very simple — you just have a [literate non-study staff] witness.” But she said that at some of the sites, the illiteracy rate is so high, they need a full-time witness, and there is a problem with reimbursement, because it is not a need for which they had budgeted.
But does high risk accurately predict HIV incidence?
However, high risk doesn’t always play out the way it is expected to (in terms of seroconversions).
CS #2, for instance, began enrolling patients at two sites in Lagos and two sites in Port Harcourt in Nigeria. Women at the two clinics in Port Harcourt reported having much more sex (seven - nine times within the past week) with far more partners in the last three months (30 partners on average) than in Lagos. In Lagos, women at one clinic (Ikeja) reported three sex acts per week, with two partners in the past three months, while women at the Apapa clinic reported four sex acts in the previous week with seven partners in the past three months.
When the researchers looked at the HIV and STIs prevalences reported at those same clinics, “it came as no surprise that the fewest infections were observed at the Ikeja site in Lagos,” said Dr Halpern. However, much to their surprise, the other site in Lagos (Apapa) had the highest prevalence of HIV and other STIs, more than even the Port Harcourt sites. A possible explanation is that the more sexually active women in Port Harcourt are more accustomed to detecting or preventing STIs through self-medication than in Lagos, and that this in turn has an impact on HIV transmission.
As a result, the investigators have closed the Ikeja site, slowed recruitment in Port Harcourt and are now focusing on recruiting more women from Apapa, which, according to Dr Halpern appears to be a ‘hot spot.’
And as the Savvy study in Ghana illustrates, just because the women are high risk when you enrol them, doesn’t mean that they will contract HIV at the predicted rate after they enter the clinical trial. In fact, simply by conducting the trial, and offering its participants free medical care, researchers are reducing women’s risk of HIV acquisition, even in the placebo arm.
One partner is risk enough
Meanwhile, in South Africa, there is a different definition of high risk. According to Dr Salim Karim, who presented the progress on HPTN 035: “In terms of the sexual risk profile, we are looking for women who... have vaginal intercourse at least once in the last three months, but on average not more than twice per day in the last two weeks. The idea behind this is try to make the results relevant to one South African population which has particular need for this sort of intervention — women with migrant partners who may not come home very often, but when they do, they pose a substantial risk.”
Likewise, many of the participants in the MDP301 trial of Pro 2000 will be from South Africa. The weighted average risk of transmission predicted by feasibility studies is 6.2 infections per 100 women years, with 12.6 infections per 100 women years at one South African site (Mtubatuba). “But these are not exactly high risk [women] per se,” investigator Dr Sheena McCormack said. “The median number of partners reported in the feasibility studies was one, so not perhaps the classic definition of high risk behaviour. These are general populations and these are staggeringly high incidences.”
And yet, at one site, this study too is enrolling higher risk women for a substudy — women whose partners are known to be HIV-positive in Masaka, Uganda. According to the feasibility study, HIV incidence at this site is 12.6 infections per year — the same as for women in monogamous relationships (of unknown serostatus) in Mtubamuba, South Africa.