WASHINGTON (Reuters) - Dr. Donald Francis is feeling combative these days.

His company, VaxGen, is about to find out whether what could prove to be the first-ever AIDS vaccine actually works. Results from the trial, which started in June 1998, will be unveiled and analyzed early next year.

Hardly anyone thinks it will work.

AIDS vaccine experts have been cautioning for years that no one has come up with the right formula for fighting the AIDS virus, which has killed 25 million people around the world and which infects 40 million more.

At best, doctors now hope that one or more of the vaccines may simply help people to live a little longer with the virus, or perhaps to reduce the ease with which it is transmitted from person to person.

But Francis, a career vaccine specialist who helped wipe out smallpox in India and Bangladesh in the 1970s when he worked for the U.S. Centers for Disease Control and Prevention, has little patience with sniping from academic researchers.

VaxGen, he said, is testing a real vaccine on real people while university-based researchers do theoretical tests involving animals.

"The crime is not to lead the country. Put your money where your mouth is and go test it, for God's sake," Francis said in an interview. Having just turned 60, he feels he has earned the right to speak bluntly.

FIRST EFFORT

VaxGen's AIDSVAX vaccine is based on some of the earliest knowledge of HIV, the virus that causes AIDS. It is the only HIV vaccine currently in Phase III clinical trials -- the last step before a vaccine or drug maker can seek approval from the U.S. Food and Drug Administration.

About 30 other vaccines are in earlier stages of human testing.

Many vaccines use a live but weakened virus, or a "killed" virus, to stimulate the body's immune response. But this is considered to be to dangerous a thing to do with HIV, a retrovirus that integrates its genetic material right into the human immune system cells it infects.

AIDSVAX uses two proteins, based on the gp-120 protein found on the outside "envelope" of the virus. The hope is the body's immune system can become sensitized to anything carrying gp-120 and will mount a response to the virus.

VaxGen has immunized two-thirds of the 5,400 volunteers in its first Phase III trial -- 5,000 homosexual or bisexual men and 400 women considered to be at high risk of getting HIV.

They have been watched since 1998 to see who becomes infected with HIV. No one knows who got the real vaccine and who got a placebo shot, but this information will be "unblinded" early next year.

The hope is, clearly, that fewer vaccinated people will have developed HIV than those who got dummy jabs.

Francis said trials in chimpanzees showed it worked well.

He dismisses some of the newer vaccine approaches as "fad science" and doesn't buy the argument that HIV is a unique virus that will require a whole new vaccine approach.

"I'm confident from the chimpanzees that the vaccine will be efficacious. The question is, how efficacious," he said.

"If it's straightforward nothing, it's easy." But if the results show at least some protection, then the race will be on to find what the correlates of protection are -- something that can be measured in the blood that will show a patient is protected from the virus.

A correlate would be a big shortcut to what VaxGen has just had to do -- vaccinate thousands of people considered to be at high risk of getting infected, and then waiting several years to see how many become infected.

"Everybody who gets the vaccine has a good antibody response," Francis said. He hopes measurements of antibody levels will work as a correlate to show whether the vaccine is working.

Even if just 20 percent of those vaccinated are immune to becoming infected by HIV, Francis believes this could have an impact on the AIDS epidemic.

The volunteers who do become HIV infected and who got the vaccine will be followed for two years to see if the vaccine at least makes the virus grow more slowly in the body.

DIFFERENT VERSIONS OF VIRUS

HIV comes in various versions called A, B, C, D and E. Subtype B is found mostly in Europe, the Americas and Japan, while A, C, D and E are spreading in Africa and Asia.

This first version of AIDSVAX, tested in the United States, Canada and the Netherlands, works against the B subtype. A B/E subtype vaccine is being tested in Thailand in trials that will finish at the end of next year.

A big question, Francis said, is whether the vaccine will work across these subtypes, or whether a specific vaccine will have to be built for each subtype.

Because the HIV epidemic is worse in Africa than anywhere else, the hope is for a vaccine that will work in Africa. Francis said it is easy to get funding for a vaccine that can be sold in the United States and Europe. VaxGen's investors have been enthusiastic for this approach.

As for an African vaccine, "There's not support for it," Francis said. "We've been limping along with it, much to our dissatisfaction." Francis has asked the government and private groups like the Bill and Melinda Gates Foundation for funding.

If it does work, then the real labor starts. VaxGen is a spinoff from biotech giant Genentech, which was originally going to gear up to make the vaccine if it worked. But Francis said Genentech is now too busy with its own products, which include the cancer drugs Herceptin and Rituxan.

"We are going to have to build a facility," Francis said.

They have made a deal with a group of South Korean investors and formed a joint venture called Celltrion, which has started work on facilities to make vaccines and other products in South San Francisco and in Inchon, South Korea.