Correction to This Article
A Feb. 7 article about a study of the effects of interrupting anti-retroviral treatment in people infected with the AIDS virus incorrectly stated that a normal CD4 cell count is above 700. The normal range in most laboratories is 500 to 1,500 cells per microliter of blood. The cells are a gauge of the immune system's health.

NIH Warns AIDS Patients Against Stopping Therapy

By David Brown
Washington Post Staff Writer
Tuesday, February 7, 2006

People infected with the AIDS virus who periodically interrupt their drug treatment run a higher risk of falling ill and dying of both AIDS and other diseases compared with people who stay on the medicines.

That is the conclusion of the largest and most expensive AIDS treatment study ever conducted, and it comes as a surprise and bitter disappointment to thousands of people who flocked to it in hopes of finding a way around lifelong use of the drug combinations.

Although the hazards are small -- about a 5 percent chance of falling ill, and a 1.5 percent chance of dying, over a little more than a year -- they are large enough that patients probably should never stop their medicines if they can help it.

The study, called SMART, was unexpectedly ended last month even before it finished enrolling volunteers. Virtually no details were released at the time, other than that people cycling on and off anti-retroviral AIDS drugs were faring worse.

Details of the experiment -- which had 5,472 subjects in 33 countries -- are being presented this morning at the 13th Conference on Retroviruses and Opportunistic Infections in Denver. They were previously presented to the AIDS Research Advisory Committee at the National Institutes of Health, which has spent $73 million on the study.

The findings -- and whether they are the last word on this controversial treatment strategy -- are already under debate.

"I am not ready to throw out the concept of treatment interruptions. That may not be scientific or rational. But I'm just not there, and I'm not alone," said Claire Rappoport, 46, a Californian who represented AIDS patients on SMART's steering committee.

Her feelings are shared by Richard Jefferys, a leader of the Treatment Action Group, an AIDS activist organization in New York.

"Our concern is that this will be considered the end of any studies of treatment strategies. We think that would be premature," he said. "The assumption that people are just going to be on therapy for the rest of their lives is not a practical assumption."

The physicians running SMART, which stands for Strategies for Management of Antiretroviral Therapy, also are struggling to come to terms with the results. The study not only reached the opposite conclusion of what many expected, it reached it in one-quarter of the anticipated time.

"We all wanted so much for it to work. But that's how science goes," said Mauro Schechter, a professor at the Federal University of Rio de Janeiro and the head of the Brazilian study site. "You do the studies to get the answer. It is not always the answer you like."

"We think we answered the question. That we answered it sooner than we thought was a surprise," said Calvin Cohen, an AIDS physician at two clinics in Massachusetts that enrolled 80 patients as part of the Community Programs for Clinical Research on AIDS network.

Anti-retroviral therapy (ART) built on the simultaneous use of three or more drugs revolutionized AIDS care in wealthy countries 10 years ago. However, the idea that a person should start treatment as soon as possible -- the "hit hard, hit early" strategy -- lost favor in the late 1990s when it became clear that the drugs could not cure the infection. That strategy was further questioned when some people suffered side effects such as changes in body fat, diabetes, high cholesterol and an apparently higher risk of heart attack.

The idea behind SMART was that people with HIV might need the drugs only when their immune systems were clearly losing the battle with the virus. They would take the drugs until their immunity rebounded, as it almost always does, and then stop until the battle turned again.

The study's organizers theorized that even if patients interrupting therapy had minor complications, they would do better over the long run because they would take their medicine more faithfully, develop less drug resistance and have a wider range of treatment options if their infection "progressed." The researchers also expected these patients would have fewer non-AIDS problems such as heart attacks, strokes and diabetes, and would reap the psychological benefits of throwing off the yoke of daily pill-taking, at least for a while.

SMART's organizers had more difficulty convincing doctors than patients that the study was worth running. But eventually 318 sites -- in places as diverse as Norway, Peru, South Africa, Morocco and Thailand -- signed on.

One of the many motivations for the trial was to find a more economical form of ART when the world is committed to bringing AIDS drugs to patients in poor countries.

"The idea is that a country in Africa only has so many dollars to treat people with HIV. Can you make those dollars go further by only treating periodically?" said Fred Gordin, an infectious-disease specialist at the Veterans Affairs Medical Center in the District, who helped run SMART.

The first patient was enrolled on Jan. 8, 2002. The target was 6,000. "Patients really wanted to participate. We had to slow down because we didn't have the manpower to randomize them as quickly as we could have," said Schechter, who ran the Brazilian site.

More than 80 percent of the volunteers were on treatment when they enrolled; all but 5 percent had taken the drugs at some point. All had CD4 cell counts above 350 per milliliter of blood, a gauge of the immune system's health. A normal count is above 700.

Once in the study, half were assigned at random to use ART continuously. The other half took the drugs only if their CD4 cells dropped below 250. But once the count got over 350, they would interrupt treatment until the count once more fell below 250.

When an independent scientific monitoring board looked at the interim results in January, it saw that patients in the interrupted treatment group had higher rates of both HIV and non-HIV illness and death.

With the average patient in the study only 15 months -- out of a projected five to seven years -- treatment interruption was not only conferring no benefit, it appeared to be causing harm.

The number of people with bad outcomes was very small, and more than 95 percent of patients were doing well. Nevertheless, 5.4 percent of those interrupting treatment had died or developed a serious complication, compared with 3.6 percent of those on continuous treatment.

The monitoring board advised SMART's leaders to stop enrolling new patients, and after a few days of discussion they reluctantly did.

The study's abrupt end is raising questions that will take months, maybe years, to answer.

Could the results be a fluke? It appears not. Nearly every category of problem -- not just AIDS, but heart, liver and kidney disease -- is higher in the interruption group.

Is there something about having HIV circulating in the blood that is dangerous in general? Could it be setting up an inflammatory state that is worsening other diseases?

Having a high CD4 count when not taking the drugs did seem to protect people. But if a patient does not need to start the drugs until CD4 drops below 350, why couldn't the patient safely stop when it is 700?

And what is the best way to measure success or failure?

Patients rushed to SMART because they were willing to accept some risk in exchange for the emotional relief of not taking medicines. But only the risk-- and not the relief -- was counted when the leaders decided to stop it.

Anecdotally, the word from the study sites is that many patients who had stopped treatment do not want to restart the drugs even though SMART's leaders have told them it is prudent to do so.

Marvin Bell, 49, a volunteer from Baltimore, is one of them.

He started taking ART in 2000. In November, he signed up for SMART and was randomly assigned to stop. He says he now sleeps more soundly, has his diabetes under control and has fewer bouts of depression.

"My attitude is pretty much always positive now. I just feel better," he said last week.

He realizes there is a good chance this is all psychological. Whatever the reason, he is holding off going back on ART.

His last CD4 count was 597. His strategy is to not restart until it dips below 350 -- and his doctor agrees.

The researchers will continue to follow SMART's patients for at least a couple of years. Several AIDS advocacy groups wrote the NIH last week asking that patients be followed much longer, perhaps indefinitely.

That way, more information about the effects of treatment interruption might yet be squeezed from the study even though patients will no longer be assigned to one strategy or another.

© 2006 The Washington Post Company