A vaginal gel reduced HIV infection in South African women by more than half, the first hint of success in the long hunt for protection women can use against the virus, according to a study released Monday in Science Express magazine.

"I think it's a landmark paper," said Dave O'Connor, an HIV researcher at the University of Wisconsin-Madison School of Medicine and Public Health, who was not involved in the study. "It opens up the possibility that we'll have an intervention that women can use discreetly — that works."

Women who used an anti-retroviral gel faithfully before and after sex cut their chances of getting HIV by 54 percent, while those who had typical use cut their chances by 38 percent, according to the study. The findings were also presented at the International AIDS Conference in Vienna on Tuesday.

If follow-up studies show the same result, the gel could potentially save millions of lives and allow women to protect themselves from HIV — even when their partners refuse to wear condoms, said Quarraisha Abdool Karim, lead author of the study and infectious disease epidemiologist at Columbia University's Mailman School of Public Health, in a news conference Monday.

Researchers have long sought a drug that could allow women to stop HIV transmission, but six potential microbicides over the last 15 years failed to curb infection — and one actually increased transmission rates.

Worldwide, 33 million people have HIV, and in South Africa, about one in every three women is infected, said Barbara Jasny, the deputy editor of Science, in a Monday news conference about the trial.

Giving women the power to protect themselves is especially important in Sub-Saharan Africa, where many of the women who are infected cannot get their partners to use condoms, O'Connor said.

The study enrolled 889 rural and city-dwelling women, ranging in age from 18 to 40, who were not using condoms in KwaZulu-Natal, South Africa. Half of the group used a placebo gel, and half received a gel that contained tenofovir, an anti-retroviral drug that prevents the virus from replicating inside cells.

Both groups of women in the study applied the gel once up to 12 hours before sex, and again within 12 hours after sex. Researchers measured how faithfully they applied the gel by counting the number of used and unused gel applicators the women brought back at monthly check-ups.

In the 2 1/2-year study, 38 women using tenofovir were infected with HIV, compared with 60 women in the placebo group, said Salim Abdool Karim, pro-vice chancellor of research at the University of KwaZulu-Natal, who led the study along with wife Quarraisha Abdool Karim.

The findings still need to be confirmed by larger trials and clear regulatory hurdles before it can be marketed, the scientists said. But "they suggest that we could soon have a new method to help reduce the heavy toll of HIV among women around the world," said Kevin Fenton, the director of the National Center for HIV/AIDS, Viral Hepatitis, STD, and TB Prevention, in an e-mail statement.

The gel also slashed rates of herpes type 2 by 51 percent in the group that used it, although the researchers aren't sure why that worked.

The researchers don't know why this vaginal gel worked after the long string of failures that preceded it.

But Salim Abdool Karim, who conducted research on several of those failed drugs, said tenofovir works by a different mechanism. Tenofovir enters the cells that HIV targets and prevents the virus from replicating, while past candidates mostly tried to prevent HIV from getting into cells in the first place, he said.

In addition, the findings raise hopes that optimizing the dose could make the microbicide even more effective. In an ongoing trial, women are using tenofovir in combination with another anti-retroviral in a microbicide, O'Connor said.

"The impact of this could be felt worldwide," he said.