A small human study may point the way to a cure for AIDS.

Behind the stunning results is a totally new approach to HIV treatment. It makes use of an epilepsy drug -- valproic acid -- that flushes HIV out of its most remote hiding places in the body.

Combined with powerful HIV drugs, the approach might totally eliminate the AIDS virus from the body. That promises a cure for AIDS, says study leader David M. Margolis, MD.

"This might lead to a therapy that would clear virus from an infected person, or reduce it to such a low level they would not need treatment for a very long time," Margolis tells WebMD. "But that is only a hope for the future. It will require a lot more work. Today, HIV is still a problem."

The Margolis team's groundbreaking report appears in the Aug. 13 issue of The Lancet. Accompanying it is an editorial by McGill University researcher Jean-Pierre Routy, MD, PhD.

"This is very serious data -- a very exciting piece of information," Routy tells WebMD. "For the first time, the disease may be attacked in the last cells where it is hiding. That is a sign of hope. We can expect one day to remove all these infected cells."

Read WebMD's "Statins May Help Curb Spread of HIV in Body"

HIV in Hiding

A decade ago, there was great optimism that powerful AIDS drug combinations would eliminate HIV from the bodies of infected people. These highly active antiretroviral therapies -- HAART -- put the AIDS virus on the defensive. The drugs kept HIV levels so low, it was hoped that the virus would soon vanish from the body.

That didn't happen. The reason: latent HIV. Latent HIV is inactive virus planted deep in the DNA of human cells. There it lies in wait, beyond the reach of anti-HIV drugs. This tiny, smoldering infection is enough to burst into full-fledged AIDS once a person stops taking HAART.

Researchers have tried everything, from years of toxic drugs to bone marrow transplants, to get rid of latent HIV. None of it worked. Making this even more maddening were recent findings that this last reservoir of HIV infection is very small.

"The total amount of infected cells is very small: just 1 gram in a person with normal body weight," Routy says. "But until a few weeks ago it was impossible to attack these cells."

Read WebMD's "Study: Many in U.S. Misunderstand HIV Vaccine"

Unmasking HIV

What changed? Until now, most efforts focused on turning the latent infection into active infection. The idea was to get HIV out in the open where HAART drugs and immune responses could get at it. But these approaches had the disadvantage of making HIV replicate faster throughout the body.

Margolis' team was one of several taking a different approach. Instead of activating the HIV gene in latent cells, they changed the cells' DNA architecture. To do this, they took advantage of the ability of valproic acid to inhibit an enzyme called HDAC.

"We thought a way to unmask latent virus would be to inhibit HDAC so you are not activating the virus but sort of taking the brake off," Margolis says.

Read WebMD's "More Than 1 Million Americans Living With HIV"

Patients Respond

It worked in the test tube. So Margolis and colleagues tested their idea at the University of Texas Southwest Medical Center in Dallas.

They recruited four volunteers with HIV infection. Each of them once had AIDS, but had their virus brought under control by HAART drugs. However, none of these volunteers had immune responses capable of fighting HIV.

Margolis and colleagues first intensified the patients' treatment by adding a new kind of HIV drug, Fuseon, to their treatment regimens. Then they gave them valproic acid for three months.

The results were unprecedented. Three of the four volunteers had an average 75 percent decrease in their latent HIV pool.

That's pretty amazing -- but none of the patients were actually cured of HIV infection. Margolis says it's going to take time to learn the best way to use valproic acid as an anti-HIV drug. And even if it's 100 percent effective at eliminating latent virus, he notes that current HAART treatment does not totally eliminate HIV replication. Until that happens, the pool of latent virus could always be replenished.

But that isn't always going to be an obstacle, Margolis predicts.

"If I look backward five to 10 years to the relative impotency of treatment, and look forward to how simple and effective treatment might be in five or 10 years, you will see the field does a good job of inhibiting HIV," he says. "If that is the only problem to overcome, we can definitely cure infection."

Read WebMD's "U.S. HIV-Infected Youth Take More Risks"

Warning to Patients: Don't Try This at Home

Margolis warns patients not to try taking valproic acid. It's a drug with serious possible side effects that include liver damage and, if taken by pregnant women, birth defects. Valproic acid also has dangerous interactions with AIDS drugs. In fact, one of the four patients in the study developed serious anemia because of an interaction with one of the drugs in his HAART regimen.

"It is totally irrational for patients to add valproic acid to therapy, outside clinical trials, at this point," Margolis says. "How much valproic acid would they take? How long would they take it for? How could you measure when they had enough? How do they know when to stop? There are plenty of reasons not to take valproic acid at this point."

Routy says that his team is organizing a clinical trial to try to find the best way to use valproic acid as an adjunct to HAART. He also expects the findings to give new impetus to therapeutic AIDS vaccines.

Meanwhile, the world AIDS pandemic continues unabated.

"We have to focus on HIV treatment today," Margolis says. "Right now we must continue to prevent, detect, and treat HIV infection. Finding a vaccine to prevent AIDS is still very important. But that goal is no more challenging than the goal of eradicating infection. And that is a worthy goal as well."

"Eradicating HIV from the body is a big hope," Routy. "But the hope is not today."

By Daniel J. DeNoon, reviewed by Brunilda Nazario, MD

SOURCES: Lehrman, G. The Lancet, Aug. 13, 2005; vol 366: pp 549-555. Routy, J.-P. The Lancet, Aug. 13, 2005; vol 366: pp 523-524. David M. Margolis, MD, professor of medicine, microbiology and public health, University of North Carolina, Chapel Hill. Jean-Pierre Routy, MD, PhD, associate professor, division of hematology, McGill University, Montreal.