This is the fourth article in a six-part series on the challenges of HIV-prevention research in South Africa. On Wednesday 10 October IRIN/PlusNews will look at the impact of a failed trial on the participants, followed by the challenges of designing an HIV-prevention trial on Thursday 11 October.

A telephone call on a Friday afternoon. That's how South African researcher Dr Roshini Govinden found out about a major setback in the field of microbicide research this year. The clinical trial she was head of in the country's east-coast KwaZulu-Natal Province would have to be stopped after results showed that the risk of HIV infection could increase.

When conducting clinical trials of a product that could potentially protect women from HIV infection, the ideal scenario would be getting it right the first time, without harm, and as quickly as possible, given that far greater numbers of women are at risk from HIV infection than men.

But the reality is much more complex. The road to finding an effective microbicide - applied via a range of products like gels, films and sponges - that could help prevent the transmission of HIV and other sexually transmitted infections, has been long and not very successful.

In 2000, a large full-scale trial showed that the over-the-counter spermicide, nonoxynol-9, considered a potential microbicide, was unsafe after women in the study developed a higher risk of HIV infection.

Seven years later, microbicide research was dealt another blow when the US-based reproductive health research organisation, CONRAD, announced the premature end of trials of a cellulose sulphate-based microbicide after the data safety and monitoring committee found a higher number of infections in the active group compared to the placebo group.

This was a huge and unexpected disappointment, as preclinical testing of the microbicide had given no indication of a potential for harm, had caused minimal side effects when used in the vagina, and appeared acceptable to women.

CONRAD was conducting the trial in Benin (West Africa), India, South Africa and Uganda. In Nigeria, a cellulose sulphate trial by US-based Family Health International, was also stopped after the CONRAD findings.

Govinden, the principal investigator of the South African study, and Dr Gita Ramjee, director of the HIV Prevention Unit at the Medical Research Council (MRC), which conducted the trials, were catapulted into a storm of controversy when media reports suggested that trial participants had been used as "human guinea pigs", and that researchers had encouraged them to seek out sexual partners at local taverns and have unprotected sex.

Almost 10 months after the trial stopped, Govinden is still a bit bruised from the negative publicity, which alluded that the trials had not been conducted in an ethical manner. "It was very stressful for me ... it was very difficult."

What went wrong?

In July 2007, Dr Lut Van Damme, lead investigator of the CONRAD trials, told delegates attending the fourth International AIDS Society conference on HIV Pathogenesis and Treatment, in Sydney, Australia, that the reasons for the larger number of new HIV infections in the cellulose sulphate arm of the trial were not clear.

According to Dr Van Damme, inflammatory reactions, or disruption of the normal vaginal flora with frequent cellulose sulphate use, could be contributing factors. CONRAD is conducting more tests to find an adequate scientific explanation for the higher number of seroconversions (HIV infection) in the cellulose sulphate group.

One of the biggest challenges in microbicide research is that the only way to check whether the product is working, is to measure new HIV infections as an outcome.

The efficacy of a treatment is often assessed by using surrogate markers like viral load (the level of HI-virus in the bloodstream) and CD4 cell count (measuring the strength of the immune system), which are used to predict whether the treatment is preventing disease progression. By using these surrogate markers, the trial investigator doesn't have to wait for opportunistic infections or deaths to occur.

HIV prevention trials, on the other hand, must enrol large numbers of women living in resource-limited settings, who are also at risk of HIV infection. Researchers place some of them in a placebo-control arm and wait for an event that no one wants - for some of them to become HIV-infected.

In all prevention trials, some participants are likely to become infected - irrespective of the intervention - but getting the public to understand this was difficult, Govinden said.

"You are not trying to make people become HIV positive, or increase their risk; we make sure that all participants are counselled, have an adequate supply of condoms, and receive other prevention measures," she told IRIN/PlusNews.

Problems and pitfalls

A week after the CONRAD announcement, public outrage erupted following sensational media coverage and health minister Manto Tshabalala-Msimang ordered a probe into the trials involving just over 600 women in KwaZulu-Natal Province.

"To make matters worse, people interviewed [in the article] weren't in our trial - not one of the trials that we as the unit had implemented ... it was just mischief making, and bad reporting," Govinden commented.

She said it was unfortunate that the health department and the national ethics committee "actually chose to believe" the media, despite the "blatant inconsistencies" in the reports; the health department had received regular updates on the progress of the trials, and had been informed of the closure before the news broke.

"As the principal investigator of the Durban site, you may think I'm biased, but the trial was conducted in the best way possible," Govinden said.

Nevertheless, there is always room for improvement. In their report looking at South Africa's experience of the closure of the trials, Govinden and Ramjee suggested that quarterly meetings between the department of health, other governing bodies and researchers be held more often, as the current frequency of meetings was "clearly not sufficient".

The ethical review of clinical trials also needed to be strengthened, so that there would be "no doubts created about the trial conduct when there are unexpected trial outcomes". According to the report, regulatory bodies approve clinical trials, but do not conduct site reviews often enough, mainly because of a lack of human capacity.

Ntokozo Madlala, an advocacy officer at the Gender AIDS Forum, a non-governmental organisation monitoring microbicide trials in South Africa, acknowledged that "they [the MRC] were very efficient in getting us the results and tracing the participants," but had underestimated the power of the media, and had not worked closely enough with community advocates.

Many South Africans found the media accusations plausible because they appeared to confirm long-held suspicions of Western medicine; despite attempts to ensure that the correct facts were published, the public was more attracted to the sensational articles.

"I think the most important lesson that we learned from the cellulose sulphate trial is that when embarking on such major research, communities, the media, researchers and the population involved have to be prepared for any outcome that can occur," said Dr Khatijah Ahmed, a principal investigator of a South African trial of the microbicide gel, Carraguard, run by the Population Council, an international non-profit organisation.

Picking up the pieces

After the damage caused by the negative publicity of the trial closure, organisations working to mobilise communities to participate in microbicide trials now have their work cut out for them.

"It's so important that there's education, especially of the media ... the news was disappointing, but the greater disappointment was the way in which the media handled the news," said Ahmed, whose trial results are expected to be released at the end of the year. She admitted that efforts to motivate the community would be "so much more difficult" next time around.

According to Govinden, many people in the communities they were currently working in were under the incorrect impression that the minister had called for all microbicide trials to be stopped, and also believed that the gel inserted had contained HIV.

Nomusa (last name withheld), 33, a participant in the cellulose sulphate trial, finished her year-long course of the microbicide gel in November 2006. A peer educator who had recruited many women to enrol in the trial, Nomusa was often being stopped in the street and asked to account for why she had made people use "the gel with HIV".

The mother of four told IRIN/PlusNews that she would not encourage any more women to participate in other microbicide trials because it would be too difficult to get them to join. "People are always asking me how I am; they think I have HIV ... they don't believe me, they still believe the papers."