What We Can Learn from the Elite Controllers

Today's calm is a far cry from the late 1980s and early 1990s, when these rooms were the front line in an epidemic that infected a quarter of the gay men in San Francisco. In those days, these halls teemed with people seeking rescue from near-certain death. The sense of desperation was palpable.

Skip Ordway, who has a checkup with Deeks today, remembers those times vividly. "My partner was diagnosed and died four months later. I thought I was one month behind him," expecting to die. The story is told again and again by the men who survived. Brian Whitford thought he was a year behind his partner. As a volunteer at an AIDS hotline, he had talked with hundreds, if not thousands, of men in his situation, and familiar with the medical research and the scant options of the day, he says, "I let my credit go to hell, sold my life insurance, had no income."

But against unfavorable odds these men did survive, and researchers such as Deeks are trying to find out why.

In the first years of the 1990s, while Ordway and Whitford were grappling with their diagnoses, Deeks was finishing his residency in inter-nal medicine at SFGH and had accepted a one-year position in outpatient HIV primary care under the direction of Paul Volberding. An oncologist, Volberding had been among the first physicians to see AIDS patients, who usually presented with the tell-tale purple lesions of Kaposi's sarcoma. Deeks enjoyed the work, and one year turned into several. By 1994 he had begun to make a name for himself through outreach programs and his tendency for taking risks (see "Jeff Getty: Lessons in desperate measures").

It was about this time that Whitford's partner died. "I hunted Steven Deeks down in an elevator at a conference," he recalls, and he's been in his care for 15 years. Whitford, today, is one of about 600 patients in Deeks' care, who participate in the Study of the Conse-quences of the Protease Inhibitor Era (SCOPE) umbrella study, making him "among the best described patients in the world," according to Peter Hunt, a collaborator and assistant professor of medicine at the University of California, San Francisco. SCOPE was designed to follow long-term outcomes along with virologic, immunologic, and clinical progression in patients on highly active antiretroviral therapy (HAART). Participants are seen every four months, at which point their HIV plasma RNA levels and CD4⁺ T cell counts are taken, along with extensive interviews.

HAART was hailed a breakthrough. "In 1996, everybody here went on this three-drug cocktail," Deeks recalls. "People who were ly-ing on their deathbeds in the hospitals got up and walked out." The gay community was euphoric, but, then in 1997, "We realized the tremendous benefit of the drug [cocktail] was only happening in half our patients," Deeks says.

The issue was that most of SFGH's patients weren't able to start three new drugs simultaneously because they had undergone mono-therapy, starting with AZT (zidovudine) in 1987 and progressing to other drugs in that class as they were introduced: notably, DDI (dida-nosine) in 1991, and 3TC (lamivudine) in 1995. By using monotherapy, the AIDS treatment centers had created large groups of people from 1996 to today, who would always be one step behind the optimal therapy.

No one has proven that elite controllers have not cleared the virus, in which case this becomes an extremely important group to study — Steven Deeks and his colleagues presented these observations as a "late-breaker" abstract at the 37th annual meeting of the Interscience Conference on Antimicrobial Agents and Chemotherapy in late September 1997 in San Diego. The study followed 136 patients who stayed on their triple-drug cocktails between March 1996 and March 1997. At the time, clinical trials of three-drug cocktails were claiming great successes, yet Deeks' results showed that only 47% responded successfully to HAART.¹

The abstract was highly controversial. "This shocked a lot of people," recalls Hunt, then in medical school. Because the drugs were clearly effective and were saving lives, "many people thought the job was done, that HIV was no longer a problem."

Benigno Rodriguez, at Case Western Reserve University, published similar results from his work at about the same time. "There was skepticism," he says, noting that one of the most important outcomes was the understanding that "clinical trials don't reflect what's happening in real life." Eventually, Deeks determined that the cocktail does work, "if you can start the three new drugs at the same time." 

In the meantime, Deeks noticed that several dozen patients who were failing on successive rounds of antiretrovirals - so called virologic failures, with viral loads above 500 RNA copies per milliliter of plasma in at least two successive tests six months apart - weren't getting ill despite increasing viral loads. The resulting retrospective study showed that the 143 virologic failures followed at SFGH for 18 months actually had increased levels (from baseline) of CD4⁺ T cells, and there were no new cases of cytomegalovirus or mycobacterium avium complex.²

That study did two things. First, it showed that viral load was not predictive of CD4⁺ T-cell decline, pointing clinical care in a new direction. It also launched Deeks into the study of what he calls partial controllers on antiretroviral therapy (PCATs). These patients have a weakened form of the virus. They can still spread it and they can progress through the disease, "but it takes years and years," Deeks says.

Whitford is a PCAT. His plasma HIV RNA levels generally remain between 500 and 10,000 copies per milliliter. These individuals are distinct from patients who control the virus without medication. In these patients, including Ordway, the virus is difficult to find. They have fewer than 50 copies of the virus per ml, and they don't take antiretroviral medications.

These people are sometimes called elite controllers. "Everyone hates the name, but now that it's out there, it's hard to go back," Deeks says. Their numbers are few. "One in 200 at most," guesses Deeks who has combed the Bay area to find them. Studies of elite controllers are rare as well. "Only a handful of researchers, most of whom are working together, are looking at them," says Volberding, now at the VA Medical Center in San Francisco.

Deeks works with researchers throughout the world to learn everything possible about these remarkable patients. He's been working for the past decade, along with Whitford, Ordway, and others, to answer the big question: why these patients are doing so well. It's a question that's garnered some attention from the American Foundation for AIDS Research (amFAR), which awarded Deeks $120,000 this past June as part of a $1.5 million gamble that elite controllers might serve as a model for HIV eradication.

Eradication is a big jump from control, and the wording is in part political. No one wants to develop a treatment that will create thousands of typhoid Marys, who spread the infection without suffering its effects. Deeks' response is that this is a public health issue and not his concern when dealing with individual patients. That said, HIV transcript levels have remained so undetectable in some of these elite controllers, he says, that they might indeed have eradicated it. "No one has proven elite controllers have not cleared their virus, in which case this be-comes an extremely important group to study." Deeks says. He has some patients that, using conventional blood tests, show no HIV virus. "But, you can't prove a negative."

The task now is to assemble groups of elite controllers and identify any qualitative differences in how their immune systems work, Volberding says. By studying the broad category of nonprogressors, Deeks and others may learn more about the pathogenesis of those who are progressing, and then take that data to the vaccine field, he says. Deeks and Bruce Walker at Harvard have the two largest groups of nonprogressors, and they are working together to expand their cohorts internationally. 

Today, the question as to why some people with HIV don't develop AIDS is open-ended. "It's a bit of a fishing expedition," Volberding says.

Viral factors, host factors, or a combination of the two must be causing a subpopulation of HIV-infected patients to resist the onset of disease. The simplest explanation is that the virus transmitted to so-called resistant patients is a heavily compromised, mutated version of HIV. For example, the virulence factor nef limits surface-expression of CD4.  A known mutation in this gene may reduce the ability of the virus to activate T cells, which is a critical factor in maintaining high levels of viral replication.

Nef-depleted virus was transmitted to several people in Australia by blood transfusion before 1985. Those patients in the Sydney Blood Bank Cohort were asymptomatic in 1999 - 14 to 18 years after infection, without any HIV therapy.³ Research revealed a deletion where the nef gene overlaps with a regulatory long-terminal repeat. Regardless, attenuated viruses such as this may not underlie con-trol, and how the mutation arose in the original blood donor is a mystery.

"Since the sexual transmission of HIV is relatively inefficient," says Deeks, "it's hard to imagine why a heavily-mutated, replication-incompetent virus would have been transmitted in the first place." A posttransmission host response probably changes the nature of the virus. Anti-HIV drugs exert a strong selective pressure on the virus. Mutants escape complete control but may end up with reduced rep-lication efficiency. Successive rounds of drug therapy may result in the emergence of progressively less virulent strains.

The recruitment of large numbers of elite controllers is important to testing this idea, and so is collecting a lot of virus. It's a possibly futile quest, Deeks says, because elite controllers have so few HIV copies in their bodies. "Our current focus is on developing the tech-niques necessary to identify and isolate viruses that are likely to be exceedingly rare - less than one in a million cells."

To increase the chances of obtaining viruses capable of replication, elite controllers will undergo large-volume leukophoresis. The procedure will produce a large number of purified CD4⁺ T cells, which will be activated in vitro with cytokines and then co-incubated with a pool of activated CD8⁺ depleted CD4⁺ T cells. The viruses obtained in this way will then be compared, in terms of fitness, to other pa-tients' viruses that replicate well in vivo.

"We will also sequence the integrated proviral DNA obtained from both peripheral blood and tissue CD4⁺ T cells," Deeks says, "to determine to what degree the viruses harbor genetic deletions that may result in a virus that is not able to replicate efficiently." Deeks and most researchers suspect that host response factors are most important.

The CCR5 D32 deletion is probably the most well-known protective host factor. It decreases the risk of acquiring HIV, and if infec-tion occurs, slows the rate of disease progression. HIV normally binds to CD4 and to a coreceptor, CCR5. The D32 deletion is found only in a few of the elite controllers.

It's possible that elite controllers inherited a unique set of HLA (human leukocyte antigen) molecules that sit on the surface of the T cells and identify the virus to the immune system, which then can kill the virus (and the cell).

Elite controllers are enriched for certain HLA types, "so it's reasonable to assume that HIV-specific T-cell responses are important fac-tors in at least some of our controllers … and has, in fact been shown by our groups and others," says Rodriguez. But there are many HLA variants, he adds, and some are associated with a much lower rate of infection. Even in elite controllers, however, protective HLA types "account for only a very small percentage of patients. There are other mechanisms in place," and we don't know what most of them are, he says.

A good bet, based on Deeks' just-published work, is that neutralizing antibodies play a role. They may limit the ability of the virus to develop escape mutations, thus contributing to but not necessarily causing long-term control of HIV in some patients.4 Neutralizing-antibody responses are lower in patients with acute infections than among those with chronic infections, and titers are es-pecially high in long-term nonprogressors.

Michael Lederman, director of the Center for AIDS Research at Case Western Reserve University, contends that viruses in persons with low-level viremia are less likely to be able to escape. "A two-way interaction is not out of the question," he writes via E-mail. "It is likely that several factors are at play in vivo, and that viral and host factors may drive the replication rate of virus populations. How this all balances out is a mystery to me, but I wouldn't be surprised to learn that the function is not a linear one and cannot be explained by any single element."

Deeks' access to a huge group of patients sets him apart and gives him the opportunity to learn something useful. This is not just because he can look for population trends. Findings in individual patients can be every bit as vital.

Another PCAT, and possibly an elite controller, Rob Rosenthal was infected in August 1986. Rosenthal joined the SCOPE program only recently. For most of the last 20 years his viral load was undetectable or very low, and his CD4⁺ T-cell count has stayed in the normal range.

Indeed, Rosenthal's viral load was consistently at or below 500 until about four years ago. Then, he says, "it started doubling … probably every four months. It doubled up to about 64,000." Although nervous about the viral load, Rosenthal didn't begin HAART. "I came to the conclusion that nobody knew what my viral load was during the 15 years before they came up with the test, and the only thing they looked at was T cells." Researchers were astonished when his viral load started halving, dropping to between 1,000 and 2,000. Later it started to climb; most recently it was measured at 71,000, the highest it's been.

As to why the viral load is cycling, Deeks says he has ideas, "but they're far-fetched ideas." He guesses that Rosenthal is, essentially, a unique elite controller whose immune system is learning how to respond to the virus. As the viral load increases, the immune system learns to recognize and attack it. The viral load decreases, mutates, and then increases again. "It's the predator/prey hypothesis and is standard in evolution," Deeks says, "… and in all the "Terminator' movies," Rosenthal adds.

"We see it in populations, but we've never documented it inside a person. That's our best educated guess of what's happening here," Deeks says. For the near future, "an educated guess" is where things are likely to remain for Rosenthal. He will be monitored like the others in the SCOPE group, but there are no plans to develop a specific protocol to address his situation.

Like Whitford and Ordway, Rosenthal is doing what he can, participating in studies that, one piece at a time, are elucidating the pathogenesis of AIDS. Although continued survival is Rosenthal's fondest dream, he says it's also a double-edged sword. "Talking with other HIV-positive men who tell me about their drug regimens, I was a little proud that I never reacted to the virus," People were supportive, he says, "I was lucky."

Whitford is aware of his own luck, too, but he doesn't discount Deeks' role. "He was cutting-edge. I'd go to the ends of the earth for him. Write that down."