AIDS vaccine falters in whites, may help blacks. (Mixed Results).

In its first large test, an AIDS vaccine has failed to shield an at-risk population from acquiring HIV, the virus that causes the disease.

The results, released by the biotechnology company VaxGen of Brisbane, Calif., may mark a turning point in AIDS-vaccine research, says Norman Letvin of Harvard Medical School in Boston. While VaxGen should be applauded for "a landmark effort" to get a vaccine to trial, he says, many scientists expected this vaccine to fail.

"This was a first-generation approach based on work done at a time when we had a very rudimentary understanding of the virus," says Letvin. The VaxGen vaccine, known as AIDSVAX, seeks to stimulate the generation of antibodies that attack HIV.

Newer, DNA-based approaches to an AIDS vaccine might work better because they deliver genes that encode proteins that could spur production of anti-HIV immune cells, Letvin says. Some such vaccines are entering early-stage human trials.

Meanwhile, VaxGen is continuing to pursue its antibody strategy and might even consider reformulating AIDSVAX to make it more potent, company officials say. Also, the company expects results late this summer from a trial of another antibody-based AIDS vaccine in Thailand.

In the study released this week, researchers found little difference in infection rates of 3,330 people who received the AIDSVAX vaccine and 1,679 people who received inert shots. Over a 3-year period, about 5.7 percent of people in each group developed an HIV infection.

More than 90 percent of the volunteers were homosexual or bisexual men. Participants were from North America, the Netherlands, Puerto Rico, and Australia. More than four-fifths were white.

The study did turn up one provocative result: The vaccine seemed to protect some African Americans. Among the 314 blacks in the study, the infection rate was 8.1 percent in those getting a placebo and only 2 percent among those vaccinated, VaxGen reports. Specifically, 9 of 111 black volunteers who got placebo injections contracted HIV, whereas 4 of 203 blacks receiving the vaccine became infected. Most of the black volunteers were women.

According to statistical standards, this apparent protection probably wasn't due to chance, says Michael F. Para of Ohio State University, one of the study investigators.

Nevertheless, basing a finding on so few cases gives some researchers pause. "The numbers [of infections] are incredibly small," says virologist Jack H. Nunberg, of the University of Montana in Missoula.

The key ingredient of AIDSVAX is a synthetic copy of a protein found on HIV. The vaccine is specifically aimed at HIV sub-type B, which is prevalent in North America, Australia, Japan, and Puerto Rico.

Blacks given the vaccine appeared to generate more anti-HIV antibodies than whites did. Antibody concentrations in blood might be influenced by gender, age, behavior, or even a person's location, says VaxGen's Phillip Berman, inventor of the vaccine. "We need to investigate all these possibilities before we can say there is a racial difference," he says.

In any event, a subtype B vaccine probably wouldn't work in Africa, where other HIV subtypes predominate, says Nunberg.

VaxGen has submitted its findings to the Food and Drag Administration.