In this article, Dr. Walker describes a symposium that occurred at the 39th Interscience Conference on Antimicrobial Agents and Chemotherapy (ICAAC), which took place in September 1999 in San Francisco. ICAAC is the largest annual international conference on infectious diseases for researchers and physicians; coverage of the 40th ICAAC will appear in the upcoming Winter issue of BETA_._
With the permission of Medscape, Inc., BETA is presenting Dr. Walker’s piece because it neatly summarizes many of the basic ideas underlying current research underway at the National Institutes of Health (NIH) and other sites into cutting-edge concepts in HIV disease management. Such concepts include structured treatment interruption (STI) and immunotherapy with interleukin 2 (IL-2) and antiretroviral treatment. Please see the article by Mark Dybul, MD, for an inside look at NIH STI research.
The treatment of HIV infection has been revolutionized by the advent of increasingly potent combination drug therapies. Whereas one drug had a beneficial effect, combinations of three or four complementary drugs can lower the level of virus in the blood stream to levels that cannot be detected by the best assays currently available. This has resulted in a dramatic decrease in the number of AIDS-related deaths in persons fortunate enough to have access to these medications. However, there are increasing concerns about how long these dramatic treatment effects can be sustained. The impact of anti-HIV drugs is already being eroded by cumulative drug toxicities and by the gradual development of antiretroviral drug resistance. Bold new approaches are clearly needed, and the most promising on the horizon is the prospect of harnessing the body’s natural immune defense system to play a more potent role in keeping HIV in check. At a symposium on "The Role of Immunotherapy in HIV Infection" at the 39th ICAAC, new developments and new challenges in the emerging field of immune reconstitution in HIV infection were reviewed by a panel of experts. Some of the data look promising indeed.
Eradication with HAART Is Unlikely
As Michael Saag, MD, from the University of Alabama at Birmingham (UAB) explained, HAART is not likely to lead to viral eradication. One reason is that only about half of the persons started on potent combination therapies are still able to sustain undetectable viral loads one to two years later. This is likely due to two factors: side effects of the drugs, which can be disabling, and the development of mutant virus species within the body that are able to resist the antiretroviral effects of the drugs. Mutations in HIV occur every time the virus reproduces itself. In the absence of therapy, 10 billion virions are produced each day, providing ample opportunity for mutations to occur. As long as the drugs are not keeping 100% of viruses from reproducing themselves, progressive mutations will arise which can result in the virus escaping the effects of the drugs.
HAART Cannot Eliminate All Infected Cells
Another reason why eradication is unlikely is that the virus persists in a quiescent form in infected cells, such that neither drugs nor the immune system can see it. Antiretroviral therapy only affects viruses that are being actively produced, making them defective so that they are less infectious. It does nothing to destroy cells that are already infected. These cells can be actively producing virus, or alternatively can be silent, containing virus that has inserted itself into the host genetic material where it can effectively hide from the immune system and from drugs. These so-called latently infected cells can persist in the body for prolonged periods of time. It was initially thought that these cells would die off at a regular rate, because cells have a finite life span. Thus, even though there might be latently infected cells in the body, these would gradually diminish in number in the presence of effective therapy until all were gone and eradication was achieved. Although it is likely that these cells do gradually die, they do so at such a slow rate that it is estimated that it would take 12-60 years for all of them to die. Moreover, this would occur only if the drug therapy were so effective that no new cells became infected over this period of time.
Since the block provided by the current anti-HIV drug regimens is not 100% effective, it is unrealistic to expect viral eradication to occur. Were eradication ever to be achieved, it would be only under very unique circumstances. These include an improved block in virus production by better drugs, some way to shorten the life span of latently infected cells, and lastly a method to flush virus from the latent infected reservoirs. As these issues cannot be addressed effectively by currently available approaches, one has to assume that viral eradication is not presently a realistic goal.
HAART-Induced Increases in CD4 Cells: Decreased Turnover and Redistribution
Dr. Saag also addressed important issues related to immune reconstitution in HIV infection. It is well known that antiretroviral therapy leads to rapid increases in CD4 cell counts in the blood stream, but there has been dispute over what drives this increase. In studies performed by his collaborators at UAB, Dr. Saag showed that a significant portion of the increase in CD4 lymphocytes is due to these cells leaving lymph nodes and returning to the blood stream when treatment is initiated. This so-called redistribution model suggests that HAART is resulting in a decrease in overall immune activation. The immune cells that have been targeting sites of viral replication within the nodes are no longer needed when virus replication is curtailed by therapy. With this model, one would expect that not just CD4 cells but also CD8 cells and B-cells would increase, and in fact this is what is observed.
However, Dr. Saag pointed out that some of the increase in CD4 and CD8 cell number when HAART is initiated is likely to be due to a decrease in cell turnover. According to this model, cell production is working overtime owing to increased destruction of CD4 cells by virus. When HAART is initiated, the overproduction of CD4 cells at this steady state results in an overall increase in CD4 cell number since suddenly far fewer cells are dying. In reality both increased production and redistribution are likely to be contributing to the encouraging increases in lymphocyte number with initiation of therapy.
HAART Limits the Number of Virus-Producing Cells
Although the overall goal of drug therapy is to stop virus replication, Dr. Saag pointed out that lowering viral load to below the limits of detection might have some previously unanticipated adverse effects. In elegant studies from UAB, investigators have shown that a single infected cell is making about 3,000-4,000 virions at any point in time. Somewhat surprisingly, this number is constant regardless of viral load or CD4 cell count -- i.e., it is the same for persons in early-stage infection as for those in late-stage infection. The number of such virus-producing cells will be influenced by the use of antiretroviral drugs; these drugs protect new cells from becoming infected, thus decreasing the number of cells producing virus, and thereby reducing the level of virus in the blood stream. However, the duration of time that a given cell is producing virus is also an important determinant of how high the viral load will be. The longer a cell can produce virus before dying, the greater the chances that these viruses will go on to infect other cells.
HAART Diminishes HIV-Specific Immune Responses
What then controls how long an infected cell is able to produce virus, which is a critical determinant of the overall viral load? This is determined by the immune response, which is able to target infected cells and eliminate them. With a strong cytotoxic T-lymphocyte (CTL) response (also called killer-cell response), infected cells will be eliminated earlier, and therefore the total number of viruses produced will be lower. With a weak CTL response, on the other hand, an infected cell will be likely to produce more virions, resulting in a higher viral load. A problem with highly potent antiretroviral therapy is that it actually leads to a decrease in the levels of CTLs, which decline in number because they encounter dramatically fewer virus-infected cells in the HAART-treated person. Thus, this represents a paradox with antiretroviral therapy: the more potent therapy is, the less vigorous the CTL response will be, and the longer an infected cell will be able to produce virus. If the therapy were 100% effective (i.e., all viruses produced would no longer be infectious), this decrease in CTL function would not be a problem because no new cells would become infected, and those already infected would produce only defective viruses. However, increasing evidence indicates that low levels of virus growth still occur in most persons on HAART, and these cells will then be harder to eliminate because the immune response has been blunted in the presence of HAART.
As Michael Lederman, MD, from Case Western Reserve University discussed, HIV infection is associated with a number of defects in the natural defense mechanisms that are supposed to protect the body against chronic viral infections. Not only is there a deficiency in the number of cells to fight HIV, but the ones that are present are not fully functional. In addition, there is growing evidence that this dysfunctional immune system may be overstimulated to the point that it has destructive side effects. New insights into the immune deficiency associated with HIV infection have come with the advent of HAART; in seeing how HAART restores immune function, one is given insights into the immune deficiencies produced by HIV infection.
Generation of Effective Immune Responses
Proper immune function relies on the proper development of those cells that mediate immune surveillance -- particularly CD4 and CD8 cells. T-lymphocytes are generated in the bone marrow and then migrate to the thymus, where they become educated to respond to invading pathogens. Naive cells migrate from the thymus, and when they encounter viral proteins (antigens) they respond by expanding and coordinating an effective immune response. Some of these cells go on to become memory cells, which can then rapidly expand in response to repeat encounters with the pathogen, resulting in more rapid deployment of effective immunity. HAART has the beneficial effect of increasing both naive and memory cells in treated persons, providing more naive cells to respond to new pathogens and more memory cells to rapidly defend against previously encountered pathogens.
HAART-Induced Increases in Immune Function
The improvements in CD4 cell number and function on therapy are variable, and are predicted by a number of factors. The rise in CD4 cell number is greater for persons who have higher baseline viral loads, possibly because these persons have higher baseline CD4 cell turnover rates, and thus production rates are higher. If the CD4 cell count is falling rapidly prior to treatment, then CD4 cells also rise more quickly with treatment. The size of the thymus also turns out to predict the rise in CD4 cells. Although the reasons are not clear, it appears that the bigger the thymus, the more rapid the increase in naive cells.
HAART also has an effect on cell function beyond its effect on cell number. HAART is associated with increases in delayed-type hypersensitivity responses (also called skin-test responses). These increases occur in response to pathogens other than HIV that are present in the body, such as cytomegalovirus (CMV) and Candida. In contrast, there is no increase in HIV-specific immune responses, because there is insufficient HIV antigen to drive these responses. The good news is that persons treated with HAART are able to respond to vaccinations such as hepatitis B vaccine, and thereby increase specific immune responses. This suggests that the immune system on HAART can be functional and that it may be possible to boost HIV-specific immunity through therapeutic vaccination with HIV antigens. Interestingly, the larger the thymus, the less likely a person is to respond to immunization, perhaps because the large thymus means that the immune system is already working at maximal output at baseline.
Remaining Questions Regarding Immune Reconstitution with HAART
HIV infection is associated with progressive and ultimately profound immune suppression. This raises the important question as to what the limits of immune reconstitution may be for persons treated with HAART, and whether important components of the immune repertoire may be lost in early stages of infection such that they cannot be restored with HAART. Among persons with current CD4 cell counts greater than 200 cells/mm3, those whose CD4 cell counts were lower at the time therapy was started have a greater risk of disease progression, suggesting that important components of the immune repertoire may have been irreversibly lost. For those in whom HAART is able to completely suppress virus, it remains unknown what the limits of immune reconstitution with HAART might be. Answers to these critical questions are likely to be forthcoming in the context of numerous clinical trials presently under way.
Eric Rosenberg, MD, from Massachusetts General Hospital next discussed the potential for restoring HIV-specific immunity with HAART. He observed that many chronic viral infections are not eradicated by the immune system but rather are held in check by a persistent and effective immune response. Notable examples include most herpesvirus infections, such as CMV and Epstein-Barr virus (EBV). The components of immunity that are required for this control are becoming better understood, and particularly include cellular immune responses rather than antibody responses.
Components of Effective Immunity in HIV Infection: CTLs
Emerging data in HIV-infected persons indicate that cellular immune responses are pivotal in providing immune containment. Cytotoxic T-cells are among the main mediators of this control; they are able to kill virus-infected cells by recognizing small viral peptides that are presented at the surface of infected cells. Viruses go through a very vulnerable stage when they infect a cell, because they lose their protective outer coating. During this time the virus is making proteins to construct new viruses, which will bud from the cell surface and go on to infect other cells. At the same time some of these new viral proteins are captured by the cell and delivered to the cell surface as a signal that the cell is infected and needs to be eliminated. Although there is very little time before new viruses are produced, in practice CTL can kill infected cells before this happens. They can also kill cells after the production of new viruses has begun and thus limit the number produced, thereby also affecting viral load.
Dependence of CTLs on T-Helper Cells
If CTLs are so important in controlling viral load, what controls the effectiveness of CTLs? It turns out that virus-specific T-helper cells are the critical factor. This has been amply demonstrated in mouse models in which CD4 cells have been removed; in these cases, CTLs are no longer able to control viremia. Dr. Rosenberg and colleagues have recently shown that HIV-infected individuals who spontaneously control viremia without the need for antiretroviral therapy have potent HIV-specific T-helper cell responses directed against the viral Gag protein. In fact, the stronger these responses, the lower the viral load. These data provide a correlate of immune protection.
Why Are T-helper Cells Absent in Most HIV-Infected Persons?
Interestingly, the lack of virus-specific T-helper cell responses in most infected persons is the most blatant defect in the immunologic repertoire observed in HIV disease. The reasons for this may be related to the unique property of HIV, namely that it selectively infects activated CD4 cells. Thus, at the onset of infection, when HIV-specific T-helper cells are being generated to help orchestrate an effective immune response against HIV, these cells may become selectively infected. Their number may then be below a critical threshold needed to allow for effective CTL responses.
Effect of HAART on HIV-Specific T-Helper Cell Responses
The hypothesis that HIV-specific T-helper cells are generated in early infection can be tested by treating persons in the very earliest stage of HIV infection with potent HAART. This would be predicted to prevent the activated CD4 cells from becoming infected. Dr. Rosenberg presented data on a cohort of over 20 such persons, and showed that control of viremia was associated with the gradual development of strong and persistent Gag-specific T-helper cell responses. In only two persons were these responses not detected, and both were infected with multidrug-resistant viruses. These data indicate that all persons infected with HIV attempt to mount a T-helper cell response, but in the absence of therapy the vast majority of these responses are lost.
Strategic Treatment Interruption in Persons with Treated Acute HIV Infection
The finding that persons with treated acute infection were able to mount strong virus-specific T-helper cell responses led Dr. Rosenberg to conduct a trial of treatment interruption, to determine whether the immune responses generated in these persons would be sufficient to control viremia. Although viremia recurred, it was delayed, suggesting that the immune response may be limiting viral spread. Perhaps more strikingly, re-treatment was associated with a large boost in virus-specific T-helper cells. These data indicate that HAART-treated persons are able to produce an immune response to their own virus when they re-encounter it, and suggest that such autologous immunization deserves to be further explored. Encouragingly, some persons first treated in the chronic phase of infection were also able to restore virus-specific T-helper cell responses, suggesting that meaningful immune reconstitution may still be a realistic goal for persons who are not identified in the earliest stages of acute infection.
Ronald Mitsuyasu, MD, from the University of California, Los Angeles, delivered the final lecture. There is clearly a myriad of immunologic abnormalities associated with HIV infection, and the challenge is to develop treatment strategies that will reverse these. As pointed out earlier by Dr. Rosenberg, a major defect in immune function in persons with chronic HIV infection is the lack of virus-specific T-helper cells. In addition, in the later stages of infection there is also a quantitative defect in the number of CD4 cells. Both of these defects might be addressed by IL-2 therapy. CD4 cells produce IL-2, which is a TH1 type cytokine that promotes cellular immunity. Recombinant IL-2 therapy has been associated with profound increases in CD4 cell number in persons treated in the earlier stages of HIV infection.
Biology of IL-2
IL-2 is a protein produced by CD4 cells and natural killer cells, and promotes the growth of T-cells. It also promotes the release of secondary cytokines, and has potentiating effects on antigen presentation. IL-2 can be made as a recombinant protein that retains all of the biologic activity of native IL-2. In light of data indicating that HIV infection is associated with a relative deficiency of IL-2, it has been used as treatment in HIV-infected persons. IL-2 might be expected to reverse a number of defects observed in HIV infection, including increasing CD4 cell number, increasing both naive and memory cells, and improving T-helper cell responses to invading pathogens, including HIV.
Controlled Trials of IL-2 Therapy
To date there have been a number of controlled trials of IL-2 therapy. These have demonstrated dramatic and sustained effects on overall CD4 cell number, but a clinical benefit is yet to be shown despite treatment of dozens of persons in these various trials. The drug has been given by continuous intravenous (IV) infusion to persons with CD4 cell counts over 300 cells/mm3 for five days every two months. This regimen led to frequent toxicities, mainly due to constitutional symptoms such as malaise and fever. Although the mean increase in CD4 cell number was 412 cells/mm3 (vs. a loss of 48 cells/mm3 in controls), there was no consistent change in viral load and there was no broadening of the CD4 cell repertoire to include subsets that had already been lost. However, IL-2 receptor expression increased over time and the number of activated CD4 cells (as determined by HLA-DR expression [Human Leukocyte Antigen, a group of D-related human leukocyte antigens found on lymphoid cells]) decreased with treatment. In the final analysis of this study, IL-2 appeared to increase CD4 cell number but did not expand the CD4 cell repertoire or have an effect on viral load.
More recent trials of IL-2 therapy have used subcutaneous infusion of drug, which can be self-administered and thus is much easier to provide to patients. Although IL-2 was associated with transient increases in viral load in this study, these elevated levels rapidly returned to baseline. A rise of over 200 cells/mm3 was observed in 44% of the 18 participants, with less robust increases in another third of study subjects, and the rise in CD4 cell count correlated with the baseline CD4 cell count. Similar results have been found in other trials of subcutaneous IL-2, and it has been shown to be well tolerated for prolonged periods of follow-up. However, constitutional toxicities remain a limiting factor.
Recently a number of Phase III trials of IL-2 therapy have been published. In all of these, therapy was associated with increases in CD4 cell number, but there have been no in-depth studies of CD4 cell function to determine whether there are also qualitative changes in the ability of these expanded CD4 cells to respond to antigen. A consistent finding in all of these studies has been a lack of decline in viral RNA, although in one study, treatment was associated with a decline in cell-associated proviral DNA. The combination of IL-2 with HAART has been reported to result in a higher proportion of persons with undetectable viral loads compared with HAART alone, although subsequent discontinuation of HAART was associated with viral rebound in all cases.
IL-2 in Advanced HIV Infection
A number of studies are now evaluating the long-term effects of IL-2 the-rapy in persons with more advanced HIV infection. These studies will evaluate both qualitative and quantitative changes in CD4 cells with therapy, and the effect of IL-2 on the latent cell reservoir. It is theoretically possible that IL-2 stimulation of latently infected cells will lead to increased purging of virus from these reservoirs, and the intermittent stimulation of virus production by IL-2 could in theory lead to an immunologic boosting of HIV-specific immune responses.
The advent of HAART has opened new possibilities for meaningful immune reconstitution in HIV infection. In addition, there are increasingly strong data indicating that the cellular immune response is likely to be key to enhancing immunologic control. The next great advances in treatment can be expected to come from a combination of both antiretroviral and immune-boosting therapies, which together can be hoped to greatly enhance the degree and durability of virus suppression in HIV-infected persons.
Bruce D. Walker, MD, is Director of Partners AIDS Research Center at Massachusetts General Hospital and Associate Professor of Medicine at Harvard Medical School.
Back to the SFAF BETA Autumn, 2000 contents page.