Slouching Toward a Cure?
After 15 years, we've turned a corner. Mark Harrington reports on the prospects of long-term HIV remission-and the "C" word.

"Eradication may have been an unreal expectation. Perhaps a durable virologic remission is a more realistic goal. It exists in nature, with long-term non-progressors, though it's rare. Perhaps early aggressive therapy and some other manipulations can do the same. We'll find out over the next year or so, I hope."
-Martin Markowitz, M.D. Aaron Diamond AIDS Research Center, New York

Late every summer, the once and future greats (and some not so greats) of AIDS research pay homage to HIV codiscoverer Robert C. Gallo at his annual lab meeting, held this year at the harborside Hyatt Regency in sweltering Baltimore, Maryland. Researchers were reluctant to use the "C" word ("cure") in their formal presentations, but the halls were abuzz with slightly surreal gossip about toxic, potentially fatal, possibly curative regimens. There was more buzz, however, than data.

Immunologist Cecil Fox, who's never been inclined to hide his opinion, approached me in the hallway. "Most people here have no idea what's really going on," he said. "Some people are going to be cured this year."

I had to check my instinctive rejection of this notion; after 10 years of treatment advocacy, I've become all too familiar with the seductive allure of false hopes and easy predictions. But Fox's opinions aren't so easy to dismiss. He's viewed as a brilliant researcher by his peers, one with a sharp bullshit detector. Years ago, Fox made his HIV reputation as one of the first to hunt for the virus in lymphoid tissues. Today he works for several top laboratories, including Anthony Fauci's lab at the National Institute of Allergy and Infectious Diseases (NIAID), which took the pictures of HIV in my lymph nodes that I showed at the international AIDS conference in Amsterdam in 1992 and at Geneva six years later. I've been HIV positive since 1985.

We talked about what a cure might really mean. "It's the cancer model, not the infectious-disease model," Fox explained. Whereas in the infectious-disease model one seeks eradication of every last microbe, in the cancer model one seeks remission. Until recently, eradication was being championed, but as Fox points out, remission is the more realistic goal. And that shift in our thinking is reflected in how we approach long-term treatment of what many now view as a chronic, but potentially manageable, disease.

Several teams-in Amsterdam, Baltimore and Bethesda, Maryland, Boston, and New York-are following groups of recently and chronically infected individuals on triple and drug therapy. After many months on highly active antiretroviral therapy (HAART), many have no detectable levels of HIV in their blood, as measured by ultrasensitive commercial viral-load tests. A further search has found only rare traces of viral particles (called provirus) in a very small pool of dormant immune cells in their bodies. Some of these proviral particles are infectious (capable of reproducing), but others are noninfectious-like viral junk. That means the virus isn't completely wiped out but is virtually trapped in these resting cells. As long as such a cell lies dormant, so does HIV. When the infected cell dies, so does the virus.

Once these latently infected cells are impossible to isolate, patients will be offered the chance to stop taking antiretroviral therapy. They will be closely monitored thereafter for the reemergence of any HIV infection. For many people, there is a restoration of immune responses. The hope is that their immune systems, bolstered by HIV therapies, might be strong enough to contain any escaping virus. Our approach to long-term HIV therapy might then look very different.

The goal of remission is modeled on one group that is successfully controlling HIV: long-term non-progressors. These individuals make up no more than 5 percent of individuals who have been infected with the virus. They are not HIV free but remain healthy, with undetectable or very low viral loads.

To achieve remission, we'll have to figure out the places HIV may be hiding out. Studies suggest there are distinct "third compartments" where HIV sequesters itself, including areas such as the thymus, testes, spleen, and genital tracts (see "Vaginal Dreams," page 20). We now know that even when people have no detectable virus in their blood, researchers can often find it in long-lived cells, such as macrophages and dendritic cells, that are found in tissues throughout the body, including the brain, which is seeded with HIV early on. Unfortunately, drugs like protease inhibitors have a limited ability to penetrate the brain. On the other hand, HIV brain researchers say what goes into the brain doesn't easily come out. But we can't rule out the risk of a virus escaping from some hidden compartment that might trigger a new infection.

So how are we faring on the road to HIV remission?

The Eradication Hypothesis and the Third Compartment
A lot has happened since David Ho revolutionized HIV research at the 1996 World AIDS Conference in Vancouver by suggesting that HIV could be eradicated from an infected person within one and a half to three years by a combination of HIV drugs including protease inhibitors. Ho, a leading researcher at the vanguard Aaron Diamond AIDS Research Center in Manhattan, presented a mathematical model that was based on measurements of how quickly HIV decays in people on drug therapy. This first version of the eradication hypothesis relied on observed rates of HIV decay in the blood.

Ho described two phases of decay: a first, rapid decay of free virus (virus outside or on the surface of cells) and actively infected CD4 T-cells with a half-life of roughly 36 hours; and a second, slower phase that, he speculated, involved chronically infected cells (like macrophages and dendritic cells) with a half-life of 8 to 28 days. Ho was careful to point out that the three-year period-to-eradication estimate would no longer apply if additional infected compartments with slower decay rates were discovered.

By fall 1997, three teams discovered the existence of such a "third compartment," a small population of resting, latently infected CD4 memory T-cells that harbor HIV genetic material and which, when activated, produce infectious HIV. The pioneers of this discovery were Robert Siliciano and Tae-Wook Chun from Johns Hopkins Medical Institutions, whose work has been confirmed by Doug Richman's laboratory in San Diego and Tony Fauci's at NIAID. The number of these cells, which Siliciano calls "L cells" (for latently infected), ranges from perhaps 10,000 to 1 million in people on potent HIV (HAART) therapy. A fraction of those cells contain infectious viral particles. But it may take only one to reignite an infection. It's estimated that an infected cell produces up to 200 new copies of HIV. That's actually a low estimate, according to Siliciano; some researchers put the number much higher.

Today there is debate over the size of the latently infected cell pool and whether it is growing or decreasing. Recent studies show the third compartment is established very early in infection, before seroconversion. (Sero-conversion is measured by the appearance of antibodies to HIV, typically occurring three to six months after exposure.) Siliciano tracked 22 HIV-positive people taking potent HIV therapies who had no detectable level of HIV in their blood based on sensitive viral-load tests. In a sobering report last year, he declared that the latently infected pool, though minuscule, doesn't appear to decline over time; instead, it remains steady. That means L-cells aren't going away any time soon. And since they may harbor infectious virus, they pose a threat of reinfection.

Over at Aaron Diamond, David Ho and Marty Markowitz have observed a measurable, but slow, decay of L-cells. Markowitz estimates the number of L-cells, after months of HAART, to range between one per million and one per more than 100 million CD4 T-cells. That's still very small, compared with the overall number of cells in the body. In Geneva, however, Fauci, Ho, and immunologist Brigitte Autran's group in Paris all presented data showing residual HIV replication in spite of ongoing HAART therapy. That suggests that the latent L-cell reservoir is being refilled at a slow but steady rate. Says NIAID's Fauci, "Virus continues to replicate even with potent drugs on board. You may not be able to get rid of the reservoir if you keep feeding it at a low level."

In an update at the August Gallo lab meeting, Siliciano reworked his numbers. He now estimates that the half-life of decay for L-cells might be on the order of 240 days, or 8.8 months. If so, and with perhaps 100,000 such cells in the body, it would take up to 12 years on HAART for the last latently infected cell to die off. Obviously this imposes an enormous burden on individuals to adhere perfectly to demanding drug regimens for over a decade. Even a single drug holiday could be enough to refill the reservoir, and the timeline for eradication would start all over again.

Researchers are now looking at several approaches to achieve virologic remission and attack the remaining latent pool.

HAART Intensification Alone
The first approach involves "hitting harder, faster"-piling on HIV drugs in order to drive viral activity down to zero-if that's possible. This is the basic chemotherapy model: kill the sucker. In Geneva, Joep Lange and colleagues reported using four and five antivirals together to speed up viral decay. In their Amsterdam study, 35 drug-náive patients were given a four-drug regimen, and eight took a five-drug cocktail. The five-drug regimen significantly sped up the time in which viral load decreased beneath the limit of detection (below 50 RNA copies per milliliter). In results from another group on the five-drug regimen, the Amsterdam researchers found that the amount of virus in the lymph nodes dropped by two logs (100-fold) in eight weeks. But side effects were more pronounced than with those on triple-drug therapy.

What's more, such intensification still hasn't eradicated the virus, which suggests that the residual compartment may somehow be inaccessible to current drugs. Thus, intensifying HAART may not help. At Aaron Diamond, Ho and Markowitz are planning to pull out the big guns, trying five-drug regimens to further suppress viral activity. Meanwhile, the greater side effects, increased expense, and unknown longer-term tolerability of five-drug regimens doesn't warrant their use, unless they prove clearly superior in maintaining viral suppression.

HAART Plus Immune Activation
The second approach involves turning on the immune system by flushing out viral particles inside resting T-cells by using pro-inflammatory molecules called cytokines (see "Immune-Based Therapies," page 40). Leading contenders include the cytokines like interleukin-2 (IL-2), interleukin 6 (IL-6) , and tumor necrosis factor (TNF); and a mouse monoclonal antibody to T-cells called OKT3 that targets and activates all the body's T cells-both CD4 and CD8 cells. By doing so, OKT3 produces a condition akin to toxic shock; the cells release inflammatory cytokines, and many die.

The idea is to use cytokines to flush lingering HIV from cells and antivirals to prevent the emerging virus from targeting new cells. At Geneva, NIAID's Fauci presented results from a test-tube study comparing IL-2 alone, a three-cytokine cocktail (IL-2, IL-6, TNF), and OKT3. All of them accelerated HIV's decay but none eliminated the virus.

At the Gallo lab meeting, Joep Lange presented a study of three individuals on HAART with undetectable viral loads who were given OKT3 to activate their immune systems. While the patients became very sick (requiring admission to the intensive care unit), infectious virus could no longer be cultured from two of the three. One promptly declared himself cured and disappeared from follow-up studies. Over at NIAID, Tae-Wook Chun also has encouraging new data relating to the latent reservoir, based on a comparative study of 26 patients taking potent HIV-therapies. Fourteen of these patients also added IL-2 to the mix. As of November, no signs of HIV provirus could be found in the blood or lymphoid tissue of six patients taking IL-2 when they looked at 10 to 20 million cells per person. But HIV provirus was found in the 12 people who didn't IL-2. Obviously the "limit of detection" for eradication experiments will be bound by the sensitivity of tests to detect HIV in tissue reservoirs. But it's good news.

At press time, National Institute's of Health's Dr. Cliff Lane and his colleague Dr. Richard Davey planned to open a study for people on HAART who have a viral load below 50 copies and are willing to go off therapy. Cells will be extracted for identification of latent infection, and participants will stop HAART therapy. Researchers will then follow these brave volunteers for evidence of viral reemergence. (Some say this approach does not seem to make sense unless the patients have received immuno-based therapy with IL-2 or other cytokines or have evidence of strong HIV-specific CD4 and CD8 immune responses to HIV.)

Meanwhile, the champions of eradication are taking a more cautious approach. "We are grappling with the issue of ongoing, very low level virus replication," says Aaron Diamond's Marty Markowitz. "Until this is resolved, we will hold off on activation strategies to avoid increasing the latent pool rather than decreasing it. It is easier to infect an activated cell than to promote the death of a latently infected one."

Intensify-Activate-Immunize
A third approach is to immunize people on HAART with a vaccine made of HIV antigens in order to generate an HIV-specific immune response to HIV. The hope here is that the immune system will program new naive CD4 T-cells emerging from the thymus to recognize and kill HIV-infected cells as they awaken from latency. HAART therapy would be stopped when the body's own immune system has learned to control the virus-the long-term non-progressor model.

At Harvard, Bruce Walker is hopeful about this strategy. His groundbreaking studies show that by treating people with potent HIV drugs during acute primary infection, one can preserve the critical HIV-specific T-cell responses seen in long-term non-progressors. Unfortun-ately, these responses don't appear to return in most of the chronically infected individuals he's studied who started treatment with HAART long after acute primary infection. Walker is following one patient treated early who went off his drugs a year and a half later.

Then there is the famous Berlin patient. This is a young man who developed acute primary HIV infection and was treated with HAART therapy 10 days later with a combination of ddI, Crixivan, and the cancer fighter hydroxyurea, a drug that targets cellular enzymes rather than the viral ones aimed at by HIV drugs. After three weeks, he developed a testicular infection and stopped taking his medication, then went back on drugs, developed hepatitis A, and eventually quit his regimen altogether. As of October-19 months later-the Berlin patient remains healthy, with no detectable viral load. He has strong HIV-specific CD4 responses to two HIV proteins (p24 and gp160).

But he's not HIV free. Using a supersensitive test, virus hunter Bob Siliciano has found infectious provirus in cells taken from the Berlin patient. That means, theoretically, anything that stimulates his immune system-a hit of the flu or some other bug-could reignite his HIV infection. Researchers think the Berlin patient will probably have to be followed for life and is likely still infectious. But he is clearly in remission. Whether or not hydroxyurea played a key role in that outcome is hotly debated.

A cautionary counterpoint to the Berlin case was published in May in the Annals of Internal Medicine by Eric Daar and Janis Giorgi, two bright lights of HIV research. They followed an infected patient treated with HAART before seroconversion who stopped taking his HIV therapy after six months. His viral load rebounded and he had a second bout of symptomatic primary infection-flu, swollen glands, the whole bit. Daar and Giorgi, among others, think this rebound is a more predictable outcome for most people who are treated early during acute infection. Others have found many more infectious proviruses in the infected cells of acutely infected people. That means it's very risky for patients who are treated early to go off their drugs.

Waiting for Volunteers
We've clearly reached an exciting crossroads. It's clear that brave-or foolhardy-individuals are likely to go off their HIV regimens in the coming months or years. It will be interesting to follow what happens to them and determine whether there are, as is likely, fourth or fifth reservoirs for HIV in the body. If HIV can be put into true long-term remission, the next step will be figuring out how to duplicate that success on a global level. These questions will take years to answer. In the meantime and for the foreseeable future, we are stuck with HAART and all its difficulties.