A microbicide gel that would have helped women protect themselves from HIV has failed its largest clinical trial to date, a finding the researchers call "bitterly disappointing" and that likely spells the end of research on this type of product.

It had been hoped that research would show that the microbicidal gel, inserted into the vagina before and after intercourse, would prevent transmission of HIV. It could have also given women better control over their health when their partners refused to use condoms.

But the study found the gel was no more effective than a placebo.

The findings, announced by England's Medical Research Council, were surprising, since a smaller study presented at an AIDS conference in Canada in February suggested the microbicide could block HIV transmission by 30 per cent.

This study involved a microbicide called PRO 2000, manufactured by specialty drugmaker Endo Pharmaceuticals of Chadds Ford, Penn.

Its active ingredient is a form of naphthalene sulfonate. Tests in the lab showed it binded strongly to receptors on the surface of HIV, preventing it from penetrating vaginal tissue, which suggested it would be effective in human use.

Researchers tested the gel on 9,385 African women in four countries. The women were given either the gel containing 0.5% of PRO 2000, a gel containing 2% PRO 2000 or placebo. The women were also counseled to use condoms whenever possible.

The study found that both men and women found the use of the gel acceptable. But the arm of the trial using the higher concentration was halted in February, 2008 when it became clear the microbicide would not work at that concentration.

Among the rest of the women, the researchers found a rate of 4.5 infections per 100 women-years in the group receiving the low-dose gel, and 4.3 infections in the placebo group.

"This result is disheartening, particularly in light of the results of a smaller trial sponsored by the U.S. National Institutes of Health," the study's chief investigator, Dr. Sheena McCormack of the Medical Research Council, said in a statement.

"Nevertheless, we know that this is an important result and it clearly shows the need to undertake trials which are large enough to provide definitive evidence for whether a product works."

Researchers are now turning their attention to gels containing microbicides and anti-HIV drugs, such as tenofovir, dapivirine and maraviroc.