hen highly active antiretroviral therapy was introduced in 1996, there was widespread expectation that it could eradicate HIV. At the time, research pioneer David Ho, director of the Aaron Diamond AIDS Research Center in New York City, estimated that such treatment could cure infection in two to three years. By the time of the international AIDS conference that summer in Vancouver, Canada, rumors had spread of patients who were already deciding to abandon their treatment regimens.

But that early optimism over undetectable viral loads achieved with drug "cocktails" proved unfounded as the individuals who had stopped taking their drugs quickly discovered that the virus still lurked within them.

People nonetheless began stopping therapy on their own more frequently than ever. Many, like Hank Wilson of San Francisco, were just plain tired of an endless regimen of drugs endowed with powerful toxicities. "I had friends with lipodystrophy," Wilson says, "and some who died suddenly despite being on HAART with viral loads under control and high T-cell counts. One [went into] kidney failure."

Not seeing the full benefit of HAART, Wilson went through periods where he decided he would not take his medication. He is now the veteran of five self-administered treatment breaks lasting one to four weeks each. But he admits the results were "scary." "The first time off," he says, "my CD4 count went down from 487 to 180 and my viral load soared to 180,000. I had to go on Septra to prevent [pneumocystis pneumonia]." His viral load still takes a leap when he stops his medications but falls back to undetectable levels when he restarts. His latest T-cell count was 463.

Research is mounting, though, to show that a majority of HIV patients do not return to optimum levels of viral load and T-cell count after periodically abandoning their medications.

Douglas Nixon of the Aaron Diamond AIDS Research Center has been studying people who interrupt treatment the way Wilson does. Nixon is deeply troubled by the "drug holiday" type of haphazard starting and stopping occurring among many HIV-positive individuals. "It's not safe to stop and start on your own," he says. "HAART has been fantastic about keeping the virus down for long periods, but long-term toxicities make drug regimens hard to follow. If there is a way to help people come off drugs, it would be very good."

The Quest for a New Strategy
As time passed and more and more patients on HAART started to show signs that they carried drug-resistant strains of HIV, researchers theorized that structured treatment interruptions of drug therapy could help rebuild failing immune systems.

"WE ARE FAR FROM UNDERSTANDING WHETHER INTERRUPTION IS BENEFICIAL AND IN WHOM AND WHAT IS THE CORRECT RECIPE." —Franco Lori of the Research Institute for Genetic and Human Therapy

One of the problems with successful HAART is that it perversely reduces immune system defense against HIV, because there just is not enough HIV around anymore to activate the necessary cells. The theory of STIs is that they may function as a kind of self-vaccination. The patient stops treatment, lets the viral load climb for a short period to generate immune activity against HIV, and then restarts treatment to suppress the virus before it can do any significant damage to the immune system. And, indeed, in the patients Nixon has looked at, those stopping therapy have shown some signs of renewed anti-HIV immunity.

Theoretically, a person could halt and resume HAART several times with the goal of eventually going off drugs and having the immune system be ready for HIV—keeping it permanently in remission without the need for further drug therapy. But STI pioneer Franco Lori of the Research Institute for Genetic and Human Therapy at Georgetown University in Washington, D.C., says, "We are far from understanding whether interruption is beneficial and in whom and what is the correct recipe."

The Tricky Thing About HIV
There are several hazards associated with periodic STIs. The HIV that survives each round of therapy can evolve to have less and less susceptibility to the regimen used to control it. "The stop-and-go game could lead to drug resistance," warns Anthony Fauci, director of the National Institute of Allergy and Infectious Diseases. "Close monitoring of patients in a trial would limit the expansion of resistant clones during STIs."

Then there is the inevitable expansion of the pool of infected T cells that would occur as the HIV population soars in the period without therapy. Those newly infected T cells set the clock back to zero as far as the gradual decay of HIV under HAART is concerned. When a patient restarts therapy, he or she will have lost any reduction in latent HIV affected by the previous years on HAART.

Worse yet, the surge in viral load during treatment interruption may damage immune function rather than rebuild it. Once treatment is restarted, it could be difficult to reestablish the lower levels of viral load that had been achieved on HAART.

The Tricky Thing About Research
There have been initial reports from several studies of people who do well during treatment interruptions. People are reading about them in all forms of the media, and not-completely-informed HIV patients are telling others about these "successes" with STIs they have heard of.

Nearly all of the studies of STIs are agonizingly small, though, and data are coming in very slowly. There is one large trial under way, the Spanish Swiss Intermittent Therapy Trial, whose first results are available, and the figures are not encouraging. Bernard Hirschel of the Hôpital Cantonal in Geneva described this trial at the international AIDS conference, held summer 2000 in Durban, South Africa. The Spanish Swiss trial has enrolled 122 volunteers with viral loads below 50 while on HAART. After each eight weeks on therapy, the volunteers take a two-week break.

At week 40, following four STI cycles, the volunteers are supposed to halt therapy indefinitely unless their viral loads rebound to over 5,000. By the time of the conference, 14 had begun this open-ended interruption, and the results bore no pattern. Two had viral loads that stabilized under 5,000 and they remained off treatment, but four ended up with viral loads above their pretreatment levels and six had viral loads close to pretreatment levels. Viral loads of another two exceeded 5,000 but remained lower than before they started treatment. In addition, 56 trial participants had gone through their fourth two-week STI by the time of the conference, and they showed no trend toward reduced HIV with successive treatment interruptions.

Georgetown University's Lori and others have criticized the Spanish Swiss study as lacking the right approach. "Two weeks off therapy is not enough to produce immunological improvement," Lori says. "HIV does not come back enough. There is a critical window—not too much, not too little—to get response."

Others contend that the rationale behind STIs is itself suspect. They think treatment interruptions actually suppress anti-HIV immune responses. When scientists at the NIAID monitored seven people who stopped HAART, they noticed a curious result. Using newly introduced assays, they could detect the existence of anti-HIV T cells both before and after the interruption. These cells became dysfunctional as HIV levels rebounded. They still produced immune-stimulating chemicals, such as gamma interferon, when exposed to HIV proteins, but they no longer proliferated to create the usual immune activity in response to that exposure.

"There is something about the presence of HIV that by itself depresses immunity," says Andrew McNeil, the study's main investigator. "You have to be very careful about accepting studies reporting more immune response during STIs. Things are not necessarily what they appear."

"IT COULD JUST AS EASILY TURN OUT THAT STAYING ON A FAILING REGIMEN IS MORE SENSIBLE THAN STOPPING WHEN NO EFFECTIVE ALTERNATIVE DRUGS ARE AVAILABLE."

The Twist About the "Evidence"
One population in which STIs do seem to show promise, though, is people who began HAART within a few months of contracting HIV—especially in the first weeks, when they were still experiencing the mononucleosis-like primary infection syndrome. Bruce Walker and colleagues at Boston's Massachusetts General Hospital have just reported on 12 such patients. Three had drug-resistant HIV and could not suppress their viral loads. The other nine began treatment interruptions after HAART had kept their viral loads below 50 for at least eight months. Five of these nine were able to control HIV on their own after one or two STIs. These five volunteers have now maintained viral loads below 5,000 for up to nine months without drug therapy.

Walker thinks the lucky people whose HIV was detected and successfully treated at its onset may represent a special case. "You have to consider viral diversity," he says. "The more a virus replicates, the more diverse it becomes and the harder for the immune system to contain. This is the case in chronic infection. Those treated during primary infection have more-intact immune systems, and small amounts of viral replication do stimulate immune responses."

A viral load in the thousands still represents substantial HIV replication. Can human immunity contain the level of infection seen in Walker's successful treatment interrupters, or will cell death and viral evolution eventually create the conditions for runaway HIV? Walker is not sure.

"The long-term response is still in question, so be careful not to overinterpret the data," he cautions. In the end, his study's major significance may not lie in its demonstration of the value of STIs. What could be more important is that it makes it possible to identify the elements of immunity that are key to controlling HIV. These features could then be stimulated in chronically infected, HAART-treated individuals through vaccines or other immune-boosting techniques that do not rely on actively replicating HIV.

"Grasping for Options"
Still, though, the debate continues over the benefits of interrupting therapy—even as another STI theory has fallen by the wayside. This theory was that it would help people experiencing treatment failure caused by drug-resistant HIV. Veronica Miller, an HIV specialist and researcher in Frankfurt, Germany, introduced the idea a year and a half ago. She hoped that stopping therapy in people with multi-drug-resistant HIV would remove the evolutionary pressures that favored drug-resistant mutations. Stopping therapy, Miller believed, could allow the nonresistant (or "wild type") HIV to again become the major HIV population in a patient's body, since this HIV frequently replicates faster than drug-resistant strains. HAART might then have a second chance to gain the upper hand, she theorized.

Miller's initial observations in 39 patients (later expanded to 165) suggested that switches from drug-resistant to wild type HIV often occurred when interrupting therapy. After such switches, most of her patients found that their drugs had renewed effectiveness against the virus within them.

This success proved short-lived, though: two months. Also troublesome were the surges in viral load that took place during the treatment hiatus. These led to steep drops in CD4 counts and to AIDS-defining conditions in 15 of the 165 patients. CD4 counts never—or only slowly—recovered after therapy resumed. Like Walker, Miller now worries about becoming too excited over preliminary data.

"The enthusiasm with which patients were asking to be taken off therapy and with which some physicians were jumping into this "strategy" was alarming to me," Miller says. "In my opinion, the fact that STI was advocated as a 'treatment strategy' as soon as we presented our data represents the phenomenon of grasping for options when there are few."

As with patients on successful HAART, the logic behind using STIs to reverse treatment failure may be fatally flawed. It could just as easily turn out that adhering to a failing regimen is more sensible than stopping when no effective alternative drugs are available—especially if the patient is able to stay in good health. Viral replication and viral loads may remain lower if the drug-resistant HIV is maintained and immune defenses improve as a consequence.

Several studies, including one by Steve Deeks at San Francisco General Hospital and another by Robert Chaisson at Johns Hopkins University, have found that patients who stick to supposedly failing regimens actually can achieve long-term stability. Their outcomes are not so much inferior to those attained by people whose therapy has reduced HIV to undetectable levels.

When Deeks took 19 of his clinic's male patients off their failing regimens for 12 weeks, CD4 counts plummeted from an average of 256 to 179—below the threshold for an AIDS diagnosis. Viral loads went up fivefold. And patients' blood lipids (triglyceride and cholesterol) normalized and lean tissue increased. Sugar metabolism and fat distribution remained abnormal, though. NIAID researchers made similar observations in 17 patients who stopped therapy when their viral loads were undetectable.

On the current outlook for STIs, NIAID's Fauci says, "The goal of completely suppressing HIV is not possible in chronic infection. If your end point is not retraining the immune system, then shifting on and off at least reduces drug exposure. If you are on drugs for half the amount of time, then compliance, cost, and toxicity problems will be better."

To this end, NIAID is now trying different schedules for what Fauci calls structured intermittent therapy, or SIT. Trial participants who have viral loads below 50 on HAART are now taking their drugs only (a) two months out of three, (b) one week out of two, or (c) two days out of seven. So far, 14 of the first 16 patients have resuppressed their HIV during the periods on therapy. (The two exceptions were following the cycle of two days out of seven.) These volunteers have followed their regimens for about only six months so far, though.

Fauci says it will take many more volunteers and one to two years to thoroughly test intermittent therapy's potential to reduce toxicities as well as its likelihood of triggering drug-resistant HIV. The latter is a major worry since SIT—even more than STIs—gives HIV frequent opportunities for evolution in this direction. Thus, the hopes for intermittent therapy are limited, and its validation remains elusive. Yet pauses in therapy are so attractive to patients that the idea refuses to die. As Fauci puts it: "People will be all over this [when they hear of even the smallest level of success]."

Back in San Francisco, the daring Hank Wilson is pleased with his personal results, though, and is not waiting for the trials to show that intermittent therapy works. He has had no obvious side effects with his most recent HAART combination. Still, he says, "I love not being on medication. I don't have to worry about the time or about carrying my meds around. I love just being able to sleep late and not having to wake up and take those pills."