BOSTON — Several promising, large-scale trials trying to prevent the spread of HIV have produced sobering results, as researchers discussed at a meeting last week, but longer-term data on new treatments are proving encouraging.

Much of the buzz at the 15th Conference on Retroviruses and Opportunistic Infections, the largest yearly scientific meeting on HIV and AIDS, centered on further analyses of a Merck & Co. vaccine trial known as STEP. The trial was stopped last fall because more people in the vaccine group got infected with HIV than in the placebo group.

"The whole field of HIV research is really going in the right direction, with I think the one exception of not making any tangible, major league headway in vaccines," said Dr. Anthony Fauci, director of the National Institute of Allergy and Infectious Diseases.

Scientists at Merck and collaborating institutions went back through demographic data of the 3,000 participants and ran extensive tests looking for differences in immune cell response but found no "smoking gun," said Dr. Susan Buchbinder, a San Francisco Department of Public Health researcher who presented results from the analysis.

They noticed that previous exposure to adenovirus, the virus modified to carry three parts of the HIV genome in the vaccine, increased the risk of infection. But, surprisingly, an even greater association was seen with circumcision status, Buchbinder said.

Volunteers who were not circumcised in the vaccine group were nearly four times more likely to get infected than those who got a placebo shot.

Buchbinder said her group has not figured out why circumcision appears to play a role, but previous research has shown that some cells on the foreskin are particularly vulnerable to HIV infection.

The STEP trial marks the third human vaccine that has failed. "I'm not sure having a fourth, fifth, sixth and seventh efficacy trial that doesn't work is very useful to us," said Ron Desrosiers, chairman of microbiology at Harvard Medical School. "Maybe now is the time to step back and make some creative discoveries and come up with novel ideas."

Some patient groups and doctors were also hoping for a success in reducing HIV infections by suppressing herpes simplex 2. But in a study of 3,251 people infected with that virus but not with HIV, there was no significant difference in the number of HIV infections in patients who received the herpes drug acyclovir and those taking a placebo after about 18 months, said the University of Washington's Dr. Connie Celum, lead author on the study.

"There was a strong basis to think having herpes increases risk for contracting HIV and transmitting it to partners," said Paul Dalton, of Project Inform, a treatment and advocacy group in San Francisco. "The herpes data I saw was very disappointing."

Even as these setbacks led to a subdued feeling about prevention, the mood on therapeutic treatments was optimistic, said Dr. Robert Schooley, a researcher from UC San Diego.

Longer-term data on drugs in new classes, such as Merck's Isentress, show them to be "just as good as we thought they were," said Schooley, who wasn't involved in Isentress research reported at the meeting.

Even after 48 weeks, about 60% to 65% of patients with a drug-resistant virus that added Isentress to their regimen reduced the amount of virus in their blood to undetectable levels. Of patients who added a placebo to their drug regimen, 31% to 35% reduced the virus in their blood to that level. Side effects were minimal.

Dr. John Mellors, an infectious-diseases specialist at the University of Pittsburgh who helped plan the scientific program, said that though there were fewer blockbuster studies to report at this year's conference, he didn't feel pessimistic about the future of AIDS. "Research is ongoing," he said.