Animal retroviral infections suggest third kind of potential treatment: HIV harm reduction.

All the approved HIV treatments so far are antiretrovirals--drugs that directly target some step in viral replication. In the future we may have another kind of treatment, immune-based therapy, which strengthen the immune system's ability to control HIV, instead of attacking the virus directly.

A third approach, less talked about so far, might be called HIV harm-reduction treatment--preventing the virus from causing harm despite a continuing high viral load. This could work because HIV seems to cause most of its damage indirectly--by the toxic tat protein, for example, or by dysregulation of immune responses leading them to kill normal cells--rather than by killing infected cells, which the body could normally replace. If so, then ways to block the indirect damage might become a new kind of treatment. (We distinguish HIV harm reduction from immune-based therapies because the former would not necessarily target the immune system at all--and also because it would have to be tested differently, since it might not decrease viral load, which immune-base therapies might be expected to do.)

One observation supporting this way of thinking is described in an abstract at the recent conference of the Institute of Human Virology, September 9-13, 2002, in Baltimore. (1) Mark Feinberg of Emory University noted that in monkeys and other primates, all known retroviral infections in their natural hosts did not cause AIDS-like disease. But the same viruses, in primates that are not natural hosts, do cause persistent infection, loss of CD4+ T-cells, and susceptibility to opportunistic infections. And in these animals that do get sick, low viral loads and strong cellular immune responses predict slower disease progression--as they do in humans with AIDS.

But at least some animals naturally infected with SIV (simian immunodeficiency virus) successfully control the infection in a very different way. In the sooty mangabey monkey, for example, the immune system does not suppress viral load, which stays high, yet the animal does not become ill. From the abstract:

"... Surprisingly we have found that SIV-infected sooty mangabey monkeys do not develop AIDS despite high level virus replication, short longevity of infected cells and limited anti-SIV specific cellular immune responses.... Interestingly, an attenuated host immune response to the infection is manifest from early times during primary infection, suggesting that sooty mangabey evolution has selected for a limited, rather than an aggressive, host response. In all, these data suggest that the direct consequences of high level virus replication alone cannot account for the progressive CD4+ T cell depletion leading to AIDS, and that active antiviral cellular immune responses may not always be beneficial. Indeed, SIV-infected sooty mangabeys may be spared, by their failure to mount significant antiviral immune responses, much of the indirect bystander damage seen in pathogenic primate lentivirus infections that both contributes to accelerated CD4 depletion and compromises host immune regenerative capacity. In contrast, following zoonotic transmission of SIV to non- natural hosts, the generation of active but incompletely effective immune responses may indirectly both increase the magnitude of overall T cell destruction and reduce the host immune regenerative capacity, thereby leading to the development of progressive immune deficiency as T cells lost to cumulative direct and indirect consequences of virus infection are not replaced." (1)

How to Proceed

A treatment that prevents AIDS by reducing damage from HIV might be hard to recognize. It might not decrease viral load at all, or even increase it. The ultimate proof would be that people would not get sick over a long period of time. But it would probably be impossible to conduct clinical trials in the straightforward way--randomly assigning patients to antiretrovirals with or without the new treatment--because the effectiveness of antiretrovirals has made it almost impossible to run clinical-endpoint trials. Instead, new drugs today are approved by their effect on viral load, an endpoint that would not work in this case. (In fact, if an existing drug for some other medical purpose happened to work this way and prevent the development of AIDS without lowering viral load, we would probably not know it, even if many patients with HIV had used the drug coincidentally.)

How then might it be possible to get a handle on the development of this kind of drug? Here are some possible approaches:

* Many patients have "discordant" viral load and CD4 counts. Instead of fitting the usual pattern of having higher CD4 counts if they have low viral load or vice versa, they either have a high CD4 count despite a high viral load, or a low CD4 count even though their viral load is low.

These two kinds of discordant patients could be compared to each other, to look for differences in how they respond to HIV infection. What could be learned from patients who can tolerate a high viral load and still maintain a high CD4 count--especially those who remained healthy despite having the high viral load for a long time? If the mechanisms could be identified, perhaps some kind of pharmaceutical intervention could help other patients do likewise.

If there are some patients who, like the sooty mangabeys, are long-term non-progressors despite having a high viral load, we probably would not have recognized them. Instead we would have treated their viral load, and attributed non-progression to the treatment. But these patients might be identified by careful examination of their medical records.

* Basic research could look for the mechanisms involved, either in sooty mangabeys, other animals, or in any humans known to tolerate a high HIV viral loads and remain healthy.

Of course differences in the virus as well as the host could be responsible for reduced disease progression despite high viral load. But still the host somehow avoids disease even though the virus reproduces well and does cause disease in nonnative hosts.

* For many years some physicians and researchers were interested in immune suppressive drugs to treat HIV. Their experience should be reviewed--especially since sooty mangabeys seem to have evolved an effective defense against AIDS that includes a notably unaggressive immune response. Existing drugs may be too non-selective or have too many side effects for widespread use. More selective immune-suppressive drugs could be developed.

* Once a candidate harm-reduction drug is identified, it could be tested to see if it improves the health of patients who cannot control the virus with any existing treatment. Since their viral load cannot be controlled in any case, an experimental treatment to reduce viral damage could ethically be tested while the viral load stays high (necessary to see if the new treatment worked). A placebo control would be used since there is no approved HIV harm-reduction treatment. The volunteers could either take antiretrovirals or not as they chose. The study could look for T-cell count increases, reduction of symptoms believed to be caused by the high viral load, and/or other evidence of clinical improvement. Such endpoints could show significant change quickly in a small number of patients (unlike the "clinical endpoint" of disease progression, which requires hundreds if not thousands of volunteers because it counts low-probability, all-or-nothing events instead of measuring continuous data on everybody).

A treatment that prevented viral damage without reducing viral load would not have the public-health advantage of antiretroviral treatment in making patients less likely to transmit the virus to others. But in practice, this kind of treatment would probably be combined with antiretrovirals for maximum benefit, so the risk of transmission would still be reduced.

A possible advantage of HIV harm reduction is that HIV develops resistance to all known antiretrovirals--and to the body's immune responses as well. But a harm-reduction treatment would create different evolutionary incentives, as HJV variants would not need to evade either the therapy or the body's defenses in order to survive. They could do best by not provoking the immune system. And in the sooty mangabey example the viral load does not increase without limit until it kills the animal; there is still a setpoint, still a limit, and the animal remains healthy. So a harm-reduction treatment may also allow relatively harmless viruses (which would have an advantage here) to help crowd out more dangerous ones.

Perhaps such reasons explain why animals apparently evolved a strategy of maintaining health by preventing harm, even from continuous high levels of viruses still able to cause disease in other species. Human long-term non-progressors (at least those who have been identified) use a different strategy, of aggressive immune defense that keeps viral replication low enough to greatly delay escape from immune control. It seems likely that the former approach is the better one for controlling a virus that can mutate so rapidly. Possibly some patients are already benefiting from it, but under current medical and research practices we do not see them. For where viral load testing is available, treatment is available too, and almost no one gets viral load tests repeatedly unless they plan to treat a high viral load. Usually antiretroviral treatment would reduce the viral load and be credited for non-progression. And experimental HIV therapies that fail to lower viral load are not studied today.

As a result, a new kind of potential treatment for AIDS may have been overlooked.

References

(1.) Feinberg M. Ignorance is bliss: how the natural hosts for SIV infection remain healthy despite long-term, high-level virus replication. Journal of Human Virology. 2002; volume 5, number 1, abstract #8.