By MARK SCHOOFS And PETER WONACOTT

In a potential breakthrough that opens a new way to protect against HIV and other sexually transmitted diseases, researchers found that a gel applied by women before and after sex cut the chance of acquiring the AIDS virus by 39% and the genital herpes virus by 51%.

It is the first time an HIV-prevention method controlled by women, who bear the brunt of the epidemic in Africa, has been shown to work. About a third of women in the study said their partners didn't know they were using the clear, odorless gel.

The findings come at a time when donor nations have been balking at continued large increases for funding AIDS treatment, intensifying pressure on finding new ways to prevent people from contracting the virus in the first place.

The study was randomized, double-blind and placebo-controlled—the gold standard for clinical research—and the results were deemed statistically significant. "It's a really well-done study," said Bruce Walker, director of the Massachusetts-based Ragon Institute, which is dedicated to immunology and HIV vaccine research. The herpes virus, HSV2, renders women more susceptible to HIV, so blocking it could add to the gel's ability to prevent HIV.

"This is a potential game changer," said Dr. Walker, who was briefed on the results but wasn't involved in the study.

Still, because it is the first trial to show the gel works, "I would like to see a confirmatory study," said one of the principal investigators, Salim Abdool Karim.

The gel, which the researchers estimate could prevent half a million infections over the next 10 years in South Africa, contained the antiretroviral drug tenofovir, provided free by California-based Gilead Sciences Inc., which markets it in the U.S. as Viread.

Administered as a pill in drug combinations, the drug is widely used to treat people already infected with the AIDS virus. According to Gilead, test-tube studies of tenofovir showed no known effect against the herpes virus, HSV2, so how the gel worked against that virus is unknown.

Gilead referred questions to Gilead Foundation President and Chairman Howard Jaffe, who said Gilead would support continuing research and attempts to win regulatory approval.

According to United Nations estimates, 18.1% of South Africans aged 15-49 were infected with HIV in 2007, one of the highest rates in the world. About 60% of those infected adults were women.

There are caveats. Many women at risk for HIV become pregnant. There is no evidence the drug causes birth defects, but the researchers want it tested.

There was a wide margin of error in the results, in part because the trial was only medium-size. It followed 889 women at two sites, one urban and one rural, for between one and two and a half years. Sixty of the 444 women using the placebo gel contracted HIV, versus 38 among the 445 on the tenofovir gel, which works out to a 39% lower infection rate per year for women using the gel. With the margin of error, the gel could have cut infection rates by as much as 60% or as little as 6%.

The effectiveness of the intervention seemed to wane over time, dropping from 50% at one year to 39% at two and a half years. One possible reason: Some women may have used the gel less often as time passed.

Those who used it consistently were better protected than those who used it off and on. In consistent users, efficacy didn't wane over time, according to the researchers, hovering around 54%.

The trial boosts a new push to use antiretroviral drugs to prevent transmission of HIV, including having uninfected people take them orally. The gel—applied on the skin and containing less drug than pills—may cause fewer side effects. The South African trial found only a slight increase of mild diarrhea in patients using the tenofovir gel compared with placebo, and no evidence of drug resistance. This gel is the first one armed with an antiretroviral drug to complete an efficacy study.

Results of the study, led by the South African husband-and-wife team of Quarraisha and Salim Abdool Karim at the Centre for the AIDS Programme of Research in South Africa, were scheduled to be presented Tuesday at the XVIII International AIDS Conference in Vienna and in the online edition of the journal Science.

Major funding for the approximately $18 million study was provided by the U.S. Agency for International Development.

At 12, 18, 24 and 30 months, the gel showed statistically significant protection against HIV. Still, the trial wasn't designed to have enough statistical power to win regulatory approval for the gel, said Anthony Fauci, director of the U.S. National Institute of Allergy and Infectious Diseases, who wasn't involved in the study. It is "a proof of concept that needs to be validated" by a second trial, he said.

A larger trial of about 5,000 women, using the same gel but with a different dosing regimen, is under way in Africa. Results aren't expected until 2013.

Some want a quicker confirmatory trial because infection rates in parts of Africa are shocking. The rural site of the trial, accounting for more than two-thirds of participants, was Vulindlela, a district of cows and corrugated metal shacks where those with AIDS were once considered bewitched. Only when the local healer could do nothing, said nurse Muke Mlotshina, were the sick taken to the health clinic, often in a wheelbarrow.

From 2005 to 2008, HIV testing in a Vulindlela community of 90,000 people found that more than a tenth of pregnant women ages 16 and under were infected. By age 24, more than half carried the AIDS virus.

In the study, each dose of the gel cost 32 cents, of which 30 cents went to the plastic applicator and packaging. Economies of scale would likely slash that price, and the drug itself would probably only cost a penny per dose, according to the Clinton Health Access Initiative.

Gilead licensed the gel for no royalty to two not-for-profit organizations. It has given licenses to manufacture the drug itself to Indian and South African generics companies, from which Gilead collects a 5% royalty.