The idea that HIV can be eradicated from the body with current treatments is now widely regarded with scepticism. The theory of eradication enjoyed a degree of acceptance in 1996 and 1997 when there was great optimism about the potential effects of antiretroviral therapy (Penissi 1996). It was thought that treatment could completely block new cycles of HIV replication. However, there is now considerable evidence of viral persistence, despite several years of suppressive anti-HIV therapy.

The reasons for viral persistence are contested by leading researchers. Some claim that HIV continues to replicate at very low levels, infecting new cells despite antiretroviral therapy. Others have argued that cell division protects the pool of infected cells. According to this group, release of HIV from the viral reservoirs, rather than replication, may explain the presence of virus.

More recently, however, there have been developments that may soon allow for the 'flushing out' of latently infected virus from viral reservoirs. This might not totally eradicate HIV, but could reduce the amount of HIV in the body and extend the amount of time that people can spend off treatment.

A theoretical approach

In 1996, AIDS researchers began to speculate for the first time about the possibility of eradicating HIV from an infected person's body altogether. This remarkable optimism was inspired by studies which showed that powerful multi-drug combinations appeared to block HIV reproduction and the infection of new cells. The theory contended that over time, the cells that are already infected with HIV should reach the end of their natural life-span and die, to be replaced by freshly produced, uninfected cells. Eventually, all the HIV-infected cells might die, leaving the individual uninfected.

Eradication theory constructed the following scenario. In the first few weeks of treatment, an effective combination of drugs can reduce virus levels in the blood by more than 2 log₁₀ or 99%. In this phase, it is mainly the production of HIV from activated HIV-infected CD4 T-cells with short lifespans that is being blocked. However, other infected cells with longer life-spans such as macrophages, together with 'resting' CD4 T-cells, may continue to produce HIV. As these cells gradually die off in a second phase that takes many months, virus production gradually falls even further.

On the basis of mathematical models, researchers at the Aaron Diamond Research Centre in New York led by David Ho initially estimated that this process might require three years or more of totally suppressive anti-HIV therapy. Subsequent research has repeatedly increased this estimate over the intervening years. In mid-1999, the latest best guess had reached sixty years of treatment.

Time to eradication

The initial optimism of researchers concerning the Aaron Diamond team's eradication theory has been tempered following more detailed investigation of the speed at which the reservoir of latently infected cells is depleted.

Data published in 1999 strongly suggested that it may be impossible to eradicate HIV from some individuals using antiretroviral drugs, because some latently infected cells persist for up to 60 years. Resting CD4 T-cells were isolated from 35 people on antiretroviral therapy. By correlating the HIV burden in these cells with the length of time on therapy, the researchers were able to estimate the half-life for these cells. After repeated testing over an average period of 14 months, their analysis suggested that it could take anywhere between 21 and 73 years for the reservoir to disappear completely. However, there were wide variations between individuals in decay rates observed, and no clear relationship between the stage of infection at which treatment commenced and the half-life. Several patients who started treatment in the early stages of primary infection had detectable latently-infected CD4 T-cells and minimal evidence of decay during the follow-up period (Finzi 1999).

Still optimistic about the prospects of eradication, David Ho has argued that the major obstacle lies in the inability of current drugs to shut down HIV replication in all compartments.

Alternative theory I

David Hos theory of HIV dynamics and eradication has been challenged by a team of scientists from the National Institutes of Health (NIH) in the United States led by Zvi Grossman.

The NIH group refute the Aaron Diamond team's theory, based on the observations that viral decay is variable during the first few days of treatment, and that decay rates vary with the efficacy of a particular combination.

According to the NIH group, intermittent bursts of local HIV replication continue despite treatment. Consequently, overall production of HIV in the body continues. Drawing on evidence from clinical trials, the NIH group argue that:

• While antiretroviral therapy does not completely block new infection of cells, it does reduce the size and frequency of these bursts.

• The second phase of viral decay is the result of less frequent bursts of viral production, due to reduced antigenic stimulation.

• The decline in viral load slows over time as the proportion of HIV in tissue (where it is less accessible to drugs) increases and more susceptible target cells are selected.

• On treatment, the viral load reaches a new set point which may be below the level of detection.

The NIH team claims that eradicating HIV with antiretroviral drugs alone is a formidable task. Furthermore, they argue that stimulating latently-infected cells may actually lead to infection of new cells. They suggest that attempts to eradicate HIV using immune stimulation should be accompanied by additional means to eradicate infected cells, such as treatments which stimulate cytotoxic responses to HIV or neutralising antibodies.

Ongoing replication

The way in which HIV can insert its genetic material into cells and then remain dormant is highly problematic for any approach to eradication. These latent or inactive HIV-infected cells are not affected by most current anti-HIV drugs. The only way to get rid of them is to wait for them to die or be activated. Once activated, HIV-infected cells become vulnerable to therapy. However, as described above, the NIH team has suggested ongoing bursts of HIV replication in response to antigen make eradication of HIV infection virtually impossible with current treatments.

A number of research groups have now examined this pool of latently infected CD4 T-cells and discovered evidence of ongoing replication. Manohar Furtado's team from the Northwestern University measured cell-associated viral DNA and messenger RNA (mRNA; necessary for HIV replication) in the blood of five men who had sustained undetectable viral load for over 20 months. After an initial decline in viral DNA and mRNA, levels plateaued, indicating ongoing HIV replication. These researchers concluded that HIV cannot be eradicated from the body with current antiretroviral treatments (Furtado 1999).

Many, but not all experts accept that low level HIV replication occurs despite undetectable plasma viral load. In addition to the Northwestern team, several other research teams have reported evidence of ongoing viral replication (Haase 1999; Sonza 2001). For some experts, the presence of molecular debris known as 2-long terminal repeat (LTR) circles is proof of ongoing replication. Others have highlighted that undetectable viral load does not mean an absence of virus (Dornadula 1999; Mathez 1999). One group suggested that a person could have 51,000 HIV RNA copies in their body, despite a viral load below 50 copies/ml in the blood plasma, while more recent studies have confirmed that low-level viral replication can occur in many patients, albeit below the level of detection with conventional viral load tests.

David Ho initially suggested that much proviral DNA was 'archival dead sequences' of DNA-fragments of HIV's genetic material that are defective and incapable of producing new virus particles. However, other researchers have infected cells with proviral DNA and successfully stimulated these cells to produce infectious HIV. Even research into men treated during seroconversion found evidence of ongoing viral replication and evolution, despite two to three years of undetectable HIV in their blood plasma. The Aaron Diamond team now estimates the half-life of latent, infected CD4 T-cells to be six months. However, the half-life of six months does not apply to individuals who are unable to maintain undetectable viral load on therapy. The Aaron Diamond team reported that intermittent bursts of viraemia were common over three years in people on antiretroviral therapy, and that the decay of replication competent HIV in resting CD4 T-cells in these people was slower, perhaps due to replenishment of the virus pool (Ramratnam 2000). They have concluded that suppression of HIV may only be possible in some individuals and that eradication may takes years to achieve (Zhang 1998). According to David Hos latest calculations, the pool of latently infected cells amounts to somewhere between 4000 and 100,000 latently infected memory CD4 T-cells, and up to 200 memory cells are activated each day to produce new virus particles.

What is unclear is the extent to which these activated cells contribute to an ongoing cycle of new infection or remain as a pool of extremely long-lived cells which will only be activated by contact with very rare antigens. To date, researchers have not found genotypic changes (alterations in the genetic structure of HIV) which could lend support to the view that new cycles of HIV infection are continuing to take place at a very low level.

Alternative theory II

Dr Robert Siliciano has put forward the theory that low level HIV replication is not occurring during viral suppression (Siliciano 2001; Blankson 2002). Like the NIH team, he has argued that the pool of HIV-infected CD4 T-cells is maintained by the bodys own homoeostatic mechanism. That is, the bodys automatic process of producing lifelong immunity to particular infectious agent by dividing cells ensures a pool of HIV-infected CD4 T-cells.

These claims are based on several pieces of evidence:

• The pool of inactive, infected CD4 T-cells remains static over five years despite viral suppression.

• The pool of 2-LTR circles is not decaying and so is not indicative of viral replication.

• Viral evolution does not affect the HIV in these cells.

On the last point, Silicianos team added to work published in June 1999 by a group from California. The California team found that people whose viral load fell below 50 copies/ml fastest did not experience evolution in RNA sequences from lymph node tissue, which suggests this pool was not subject to ongoing replication but instead represents 'trapped virus'. They presented preliminary information from a small study investigating the effects of therapy on residual viral reservoirs. Six participants in the Merck 035 study of indinavir (Crixivan), AZT (zidovudine, Retrovir) and 3TC (lamivudine, Epivir) underwent lymph node biopsies one and two years after starting treatment. All had sustained viral load below 50 copies/ml for two years. Significant evolution was seen in those patients whose viral load fell more slowly, or who experienced bursts of viraemia whilst on treatment (Gunthard 1999). Silicianos team reported no viral evolution among people with full viral suppression for three to four years.

The source of viral rebound

There is now doubt about whether latently infected, resting CD4 T-cells are the exclusive or primary source of viral rebound. This doubt arose when a group of researchers found that virus cultured from these resting infected cells did not match rebound virus in seven of nine individuals who ceased anti-HIV treatment. In the other two individuals, the rebound virus and the viral culture were identical (Chun 2000). This finding indicates that sources other than resting CD4 T-cells must be considered as the origin of viral rebound.

Is antiretroviral therapy strong enough?

David Ho believes that two of the underlying reasons for the inability of antiretroviral therapy to significantly reduce the reservoirs of latent HIV thought to undermine attempts at eradication are the persistent low levels of replication seen with current regimens, characterised by viral blips and evolution of drug resistant mutants in individuals with viral load below 500 copies/ml, and the relative slowness with which viral replication is brought under control after beginning treatment. Ho has argued that the reservoir of latently infected cells is not static, but is constantly replenished by new infection during 'blips in replication, as well as replication which continues to occur below the level of detection of current viral load tests (20 to 50 copies/ml).

Using a combination of ritonavir-boosted lopinavir (Kaletra), tenofovir (Viread), 3TC and efavirenz (Sustiva) in 22 individuals, Ho found that the speed of viral decay was significantly increased in comparison to a control group treated with saquinavir, ritonavir (Norvir), AZT and 3TC. A 1.4 log₁₀ drop was seen within five days in the five-drug group, and the decay rate was calculated to be twice as high in these patients. In this 14-day study, in which patients were hospitalised in order to permit very frequent viral load measurements, tolerability of the regimen was not reported.

The objective of the study was to test whether using five agents each known to individually reduce HIV replication by 1.5 log₁₀ could produce a quicker decay in HIV replication than seen hitherto in studies of antiretroviral therapy.

Assuming that the efficacy of the five drug combination was 100% (in itself questionable), the four-drug combination had an efficacy of 79% and previous viral dynamics studies suggested an efficacy of 67% for ritonavir monotherapy, based on the earliest analyses of viral dynamics published in 1996 and 1997.

Eradicating other viral reservoirs?

If, as some experts argue, the pool of latent, HIV-infected CD4 T-cells is perpetuated by the bodys own homoeostatic mechanism for maintaining memory CD4 T-cells, then eradicating HIV will be impossible.

For there to be any hope of eradication, anti-HIV drugs would need to be able to suppress the virus throughout the body's tissues and blood cells. Key viral reservoirs include long-lived, infected immune cells, lymphatic tissue, including the gastrointestinal tract, tonsils and rectal mucosa, lymph nodes, the central nervous system, the thymus and the testicles.

Lymphatic tissue is a major reservoir for HIV. Pilot studies suggest that regimens that do not produce at least a 1 log₁₀ decline in plasma viral load will not affect lymph node viral load. One study found that the triple combination of AZT, ddI (didanosine, Videx / VidexEC) and 3TC did reduce lymph node viral load, whereas AZT and ddI combination therapy did not (Lafeuillade 1996). Other studies using protease inhibitor-based combinations have demonstrated substantial reductions in viral load in the lymphatic tissue (Cavert 1997; Kotler 1998; Wong 1997). However, in many cases ongoing viral replication in these tissues can be detected even after plasma viral load has been suppressed below the limit of detection for many months.

Even if drugs can block HIV replication in the blood and lymphatic system, it is possible that they do not penetrate as well into other parts of the body where HIV can live, such as the thymus, the central nervous system or the testicles.

It seems that treatments are not as effective against HIV in these 'viral reservoirs'. For example, a recent study showed that HIV viral load declined in both the blood plasma and the central nervous system as a result of treatment during the early phases of HIV disease. However, during late stage disease, HIV in the brain fluid persisted and did not decline in line with virus in the blood (Ellis 2001).

HIV also has the ability to infect B-cells which bear the CXCR4 co-receptor, making these another possible reservoir of HIV infection (Moir 1999). These places could provide HIV with 'sanctuary sites' from which it could emerge to reinfect the rest of the body whenever treatment is discontinued.

Naive CD4 T-cells also form a hidden reservoir of HIV. While memory CD4 cells hold more proviral material than naive cells, the naive cells also contain proviral material which can be cultured into replication competent HIV. Naive cells may be more vulnerable to infection by HIV which uses the CXCR4 co-receptor (Ostrowski 1999). Long-living infected macrophages also form a viral reservoir.

Implications for the timing of treatment

If eradication of HIV is at all possible, most researchers believe that it is most plausible among newly infected people. For this reason, almost all the experiments which have been conducted so far have taken place in groups of people infected for fewer than six months who have never received drug treatment before.

It is suggested that treatment within the first few months of infection has the best chance of success for three reasons:

• Individuals in the early stages of infection are less likely to have developed a wide variety of strains of HIV, some of which may prove to be resistant to anti-HIV drugs.

• It might prevent the 'sanctuary sites' from becoming infected.

• People with very early stage infection are likely to have strong immune responses, and are less likely to have suffered immune damage that the body cannot repair.

However, there is evidence of ongoing viral replication among people treated during seroconversion, as discussed above (Zhang 1998). No seroconverter to date has been shown to have eradicated HIV following early treatment. Nevertheless, several studies have shown a significantly greater reduction in the burden of proviral DNA when compared to people treated in chronic infection if treatment begins within months of infection.

A review of individuals who started treatment during primary infection showed that after 12 months on treatment, the rate of proviral DNA decay was greater amongst this group than amongst chronically infected individuals who had received the same amount of treatment with a three-drug regimen (Ngo-Giang-Huong 2001).

Treatment during primary infection or within one year of infection reduced the burden of proviral DNA significantly compared with treatment that began more than one year after infection, and after 36 weeks of treatment the primary infection group had a level of proviral DNA comparable to that of a long-term progressor control group (Pires 2004).

Testing the theory

To accelerate the process and target the longer-lived cells and 'sanctuary sites', researchers are trying new approaches such as:

• Stimulating the immune system to activate latently-infected cells and accelerate their demise.

• Temporarily suppressing the immune system to kill off latently infected cells, allowing them to be replaced by uninfected ones.

• Using transplants of thymus tissue to produce new T-cells.

• Using other treatments to target latent cells, such as CD8 T-cells modified through gene therapy.

In September 2003, two new developments took place that reawakened interest in the possibility of reducing HIV reservoirs in the body:

• University of California researchers found that a combination of two agents reduced the burden of HIV-infected cells by up to 80%. One agent activates HIV lying dormant in cells, and the other is a targeted antibody that kills newly produced HIV before it leaves these cells.

• In the same week, two United States research groups reported that the scale of HIV infection of lymph nodes in early and asymptomatic HIV infection is far more limited than hitherto believed.

Pioneers of both techniques are suggesting that they could eventually be used to reduce the amount of HIV in the body and extend the amount of time that people can spend off treatment.

The extent of viral reservoirs

In 2003, important new findings from two groups of researchers challenged accepted wisdom about the extent of viral reservoirs in HIV-infected people. Previous research had suggested that the lymph nodes in the gut are the biggest reservoir of HIV in most people with HIV.

Both groups used a technique called positron emission tomography (PET) scanning to identify lymphoid tissue that was packed full of activated lymphocytes. These white blood cells are drawn to lymph nodes that are focal points of HIV infection. PET scanning is able to identify activated lymphocytes by tracking levels of radioactively labelled glucose, since activated lymphocytes increase their glucose uptake 20-fold.

Lymph nodes are the main reservoir of HIV infection, and PET scanning has revealed that people with early or non-progressive HIV infection (either recent seroconversion or a CD4 cell count above 400 cells/mm³) have little sign of HIV activity in lymph nodes below the chest. Outside the lymph nodes, PET scanning could not detect substantive levels of lymphocyte activation, indicating that the vast majority of virus activity in the body is concentrated in the lymph nodes and lymphoid tissue such as the spleen. The authors of both studies suggest that surgical removal of lymph nodes may be an option.

A group from Johns Hopkins University School of Medicine reported on twelve recent seroconverters and eleven HIV-positive untreated individuals with CD4 cell counts above 400 cells/mm³ and viral loads ranging from undetectable to 52,000 copies/ml. All the latter group had stable viral loads, fluctuating less than 0.5log₁₀ over three measures. Whilst recent seroconverters had wide-ranging lymph node involvement, the asymptomatic group showed little involvement of deeper lymph nodes in the torso. Instead, lymph nodes in the neck and upper torso tended to be much more strongly activated than lower torso lymph nodes.

One individual was scanned prior to therapy, and again during a one-month treatment interruption that took place after six months of therapy and one month of viral suppression below 50 copies/ml. Viral load peaked at around 10,000 copies/ml after one month off treatment, and the pattern of lymph node involvement during the treatment interruption matched the one seen in recent seroconverters. Of particular note, no lymph nodes unactivated at the time of the first PET scan became activated during the treatment interruption (Scharko 2003).

In a second study carried out by the University of Wisconsin, Dr David Pauza evaluated lymph node activation by PET scan in 15 HIV-positive patients at various stages of disease progression. He found a clear anatomical pattern. As HIV disease progressed, the number of lymph nodes activated began to increase, and the focus of activation moved from the head and neck down the torso, until in late stage disease the lymphoid tissue in the gut became heavily activated. In one patient close to death with a CD4 cell count of 0 cells/mm³, the only sites of activation were the rings of lymphoid tissue at the ileocacal junction, the point at which the large intestine meets the small intestine. The authors of this study speculate that these rings of lymphoid tissue are the last reservoir of HIV in the body (Iyengar 2003).

Flushing out viral reservoirs

It may be possible to stimulate HIV-infected cells to express latent virus to express some components of HIV without releasing new virus particles. This would allow infected T-cells to be targeted with immunotoxins, in order to kill them selectively.

Immunotoxins are monoclonal antibodies with a toxin attached to them. The principle is that the monoclonal antibody binds to a specific molecule in the surface of the cell, which is then taken into the cell by a process called phagocytosis. Following ingestion of the antibody and the toxin, the target cell is killed.

This feasibility of this approach was demonstrated in 2003. The investigators implanted human thymus tissue into mice that lack an immune system, before infecting the tissue with HIV. The mice responded by producing human T-cells infected with latent HIV. They then stimulated the T-cells strongly enough to prompt the cell to express latent virus but not to trigger other cellular functions, revealing the hidden HIV.

The research team used a diphtheria immunotoxin called Pseudomonas exotoxin linked to an antibody that targets the CD4 binding site of HIVs gp120 protein. The researchers also used the cytokine interleukin-7 (IL-7) to induce latently infected cells to produce virus without stimulating activation of all T-cells (a measure that would have upset immune function). Prostratin was also tested for the same purpose (Brooks 2004). The model successfully exposed and killed hidden HIV without affecting the rest of the immune system. The T-cells in the model also did not divide, indicating that they were able to produce virus without behaving as if they were confronting a foreign particle.

In another possible scenario, physicians might first administer a therapeutic vaccine to enhance the ability of the patient's T-cells to kill HIV-infected cells. This would help the two-step approach rid the body of latent virus more efficiently.

A team of American researchers recently demonstrated in the test tube that an anti-CD45RO immunotoxin eliminated latently-infected CD4 T-cells obtained from individuals with an undetectable viral load (below 400 copies/ml), making this a potential candidate for human trials (Saavedra-Lozano 2004).

It has also been demonstrated that the anticonvulsant drug valproic acid (Convulex / Depakote), was an effective, if relatively weak inhibitor of histone deacetylase, a human enzyme which contributes to HIVs ability to hide in resting CD4 T-cells in viral reservoirs. Use of this drug, which is currently prescribed to people with bipolar disorder or epilepsy, appears to activate provirus hidden within resting T-cells without fully activating cells, and may be used in future clinical trials seeking to flush out HIV from viral reservoirs.

This treatment produced much excitement after publication of a small proof-of-concept study in August 2005, which claimed that the use of valproic acid could eventually point the way towards a 'cure' for HIV infection. The study tested the use of the drug in four patients who had had undetectable viral loads on antiretroviral therapy for over two years. After 16 to 18 weeks of taking valproic acid plus T-20 (enfuvirtide, Fuzeon) alongside their existing anti-HIV drug combination, three of the four patients had a decline in the number of resting, infected cells of between 68 and 84%, with the fourth patient showing a decline of 29%. While this proves that valproic acid can reduce the number of infected cells, experts have called into question the applicability of this approach in the future, since a reduction close to 100% would be necessary to prevent HIV returning as soon as the drugs are stopped. Larger, randomised studies are required to assess whether this approach will be of benefit to patients in the future (Lehrman 2005).

Yet another approach has identified genes that control the latency and activation of cells infected by HIV. A compound called resveratrol activates egr1, a gene whose product causes cell growth to slow, creating favourable conditions for HIV replication. A large number of other genes are expressed at different levels according to whether a cell is active or resting, suggesting that there are many avenues to active the reservoir of latently infected cells (Krishnan 2004).

Eradication using immune stimulants?

It has now been established that infected CD4 T-cells that are 'resting' or inactive constitute a reservoir of HIV production. Researchers are investigating ways of activating these 'resting' CD4 cells with antibodies or cytokines, and making them 'visible' to the immune system. They are also examining whether therapeutic vaccines may improve the immune response against infected cells.

Ongoing research by Anthony Fauci and Tae-Wook Chun has provided some encouraging evidence that interleukin-2 (IL-2) may play a role in eradicating HIV from the body, although the researchers urge caution. Fourteen people with undetectable virus in their blood plasma received IL-2 and HAART. Subsequently, no replication competent HIV could be cultured from the CD4 T-cells of six participants. In addition, researchers could not find HIV in the lymph nodes of two out of three of these individuals. However, the results do not prove that HIV has been eradicated in these people, because a number of other viral reservoirs were not investigated.

In contrast, a group from Spain have reported that low doses of IL-2 did not reduce HIV proviral DNA in peripheral blood mononuclear cells amongst eight people with CD4 cell counts below 250 cells/mm³, who had suppressed viral load below 200 copies/ml for more than 24 weeks on HAART (Ruiz 1999). Furthermore, a study which ceased anti-HIV treatment in 18 people who had undetectable viral load for about two years found that previous exposure to IL-2 did not affect the rate of viral rebound (Davey 1999).

However, less encouraging results have been reported by a British-Dutch team who found that the rates of depletion and replenishment of infected cells were determined by the potency of antiretroviral therapy and the degree of T-cell stimulation. The group led by Dr Christopher Fraser from the University of Oxford found that high level T-cell stimulation was likely to produce negative results. However, repeated, low-level immune stimulation was associated with expansion of the CD4 T-cell pool and a reduction in the reservoir of long-lived HIV-infected T-cells.

Additionally, IL-7 has been shown to induce HIV-1 replication from resting CD4 T-cells in patients on HAART. It is suggested that the immunomodulatory effects of IL-7 could be combined with its ability to stimulate HIV-1 replication from resting CD4 T-cells to purge HIV-1 reservoirs. These events could be counterbalanced by the intensification of HAART in the attempt to reduce the size of HIV-1 reservoirs (Wang 2004).

Macrophages are another important viral reservoir. A number of therapies that target infected macrophages are being researched. Granulocyte-macrophage colony stimulating factor (GM-CSF), IL-2 and tumour necrosis factor are therapies currently being tested. Researchers hope that these therapies will stimulate infected cells into action, and trigger an immune response which targets infected macrophages.

For more information, see Colony stimulating factors and Interleukin-2 (IL-2) - overview in Drugs used by people with HIV: Immune-modulating drugs.

References

Blankson JN et al. The challenge of viral reservoirs in HIV-1 infection. Ann Rev Med 53: 557-593, 2002.

Brooks DG et al. Molecular characterization, reactivation and depletion of latent HIV. Immunity 19: 413-424, 2003.

Cavert W et al. Kinetics of response in lymphoid tissues to antiretroviral therapy of HIV-1 infection. Science 276: 960-964, 1997.

Chun TW et al. Quantification of latent tissue reservoirs and total body viral load in HIV-1 infection. Nature 387: 183-188, 1997.

Chun TW et al. Effect of interleukin-2 on the pool of latently infected, resting CD4+ T cells in HIV-1-infected patients receiving highly active anti-retroviral therapy. Nat Med 5: 651-655, 1999.

Chun TW et al. Relationship between pre-existing viral reservoirs and the re-emergence of plasma viremia after discontinuation of highly active anti-retroviral therapy. Nat Med 6: 757-761, 2000.

Davey R et al. HIV-1 and T cell dynamics after interruption of highly active antiretroviral therapy (HAART) in patients with a history of sustained viral suppression. Proc Natl Acad Sci 96: 15109-15114, 1999.

Dornadula G et al. Residual HIV-1 RNA in blood plasma of patients taking suppressive highly active antiretroviral therapy. JAMA 282: 1627-1632, 1999.

de Jong MD et al. Overshoot of HIV-1 viraemia after early discontinuation of antiretroviral treatment. AIDS 11: 79-84, 1997.

Ellis R et al. Dynamics of HIV replication and persistence in the CNS after antiretroviral therapy. Eighth Conference on Retroviruses and Opportunistic Infections, Chicago, abstract S19, 2001.

Fauci AS et al. Host factors in the pathogenesis of HIV disease. 36th Infectious Diseases Society of America Conference, Denver, abstract S94, 1998.

Finzi D et al. Identification of a reservoir for HIV-1 in patients on highly active antiretroviral therapy. Science 278: 1295-1300, 1997.

Finzi D et al. Latent infection of CD4+ T cells provides a mechanism for lifelong persistence of HIV-1, even in patients on effective combination therapy. Nat Med 5: 512-517, 1999.

Fraser C et al. Reduction of the HIV-1-infected T-cell reservoir by immune activation treatment is dose-dependent and restricted by the potency of antiretroviral drugs. AIDS 14: 659-669, 2000.

Furtado MR et al. Persistence of HIV-1 transcription in peripheral-blood mononuclear cells in patients receiving potent antiretroviral therapy. N Engl J Med 340: 1614-1622, 1999.

Grossman Z et al. Ongoing HIV dissemination during HAART. Nat Med 5: 1099-1104, 1999.

Gunthard H et al. Evolution of envelope sequences of human immunodeficiency virus type 1 in cellular reservoirs in the setting of potent antiviral therapy. J Virol 73: 9404-9412, 1999.

Haase AT et al. Kinetics of response in lymphoid tissues to antiretroviral therapy of HIV-1 infection. Science 276: 960-964, 1997.

Haase AT et al. Sexual transmission and propagation of SIV and HIV in resting and activated CD4+ T cells. Science 286: 1353-1357, 1999.

Iyengar S et al. Anatomical loci of HIV-associated immune activation and association with viremia. Lancet 362: 945-950, 2003.

Krishnan V et al. Host cell gene expression during HIV-1 latency and reactivation, and effects targeting genes differentially expressed in viral latency. J Virol 78: 9458-9473, 2004.

Kotler DP et al. Effect of combination antiretroviral therapy upon rectal mucosal HIV RNA burden and mononuclear cell apoptosis. AIDS 12: 597-604, 1998.

Lafeuillade F et al. Human immunodeficiency virus type 1 kinetics in lymph nodes compared with plasma. J Infect Dis 174: 404-407, 1996.

Lehrman G et al. Depletion of latent HIV-1 infection in vivo: a proof-of-concept study. Lancet 366: 549-555, 2005.

Lisziewicz J et al. Control of HIV despite the discontinuation of antiretroviral therapy. N Engl J Med 340: 1683-1684, 1999.

Margolis DM et al. Coaxing HIV-1 from Resting CD4+ T Cells: Valproic Acid Induces Latent Viral Expression. Eleventh Conference on Retroviruses and Opportunistic Infections, San Francisco, abstract 427c, 2004.

Markowitz M et al. The effect of commencing combination antiretroviral therapy soon after human immunodeficiency virus type 1 infection on viral replication and antiviral immune responses. J Infect Dis 179: 527-537, 1999.

Mathez D et al. Continuous HIV replication under effective HAART: immediate HIV recurrence off treatment irrespective of latently infected cell titres. Seventh European Conference on Clinical Aspects and Treatment of HIV-Infection, Lisbon, abstract 114, 1999.

Moir S et al. CD40-mediated induction of CD4 and CXCR4 on B lymphocytes correlates with restricted susceptibility to human immunodeficiency virus type 1 infection: potential role of B lymphocytes as a viral reservoir. J Virol 73: 7972-7980, 1999.

Ngo-Giang-Huong N et al. Proviral DNA in subjects followed since primary HIV-1 infection who suppress plasma viral load after one year of highly active antiretroviral therapy. AIDS 15: 665-673, 2001.

Ostrowski MA et al. Both memory and CD45RA+/CD62L+ naive CD4(+) T cells are infected in human immunodeficiency virus type 1-infected individuals. J Viorl 73: 6430-6435, 1999.

Penissi E et al. Antiviral therapy: eradicating HIV from a patient: not just a dream? Science 272: 1884, 1996.

Perelson AS et al. Decay characteristics of HIV-1-infected compartments during combination therapy. Nature 387:188-191, 1997.

Pires A et al. Initiation of antiretroviral therapy during recent HIV-1 infection results in lower residual viral reservoirs. J Acquir Immune Defic Syndr 36: 783-790, 2004.

Ramratnam B et al. The decay of the latent reservoir of replication-competent HIV-1 is inversely correlated with the extent of residual viral replication during prolonged anti-retroviral therapy. Nat Med 6: 82-85, 2000.

Ruiz L et al. Effect of interleukin-2 at low doses on HIV-1 total and integrated DNA in advanced HIV infection. Antivir Ther 4: S121, 1999.

Saavedra-Lozano J et al. An Anti-CD45RO Immunotoxin Kills Latently Infected, Replication-competent, CD4+ T Cells Purified from HIV+ Individuals on HAART with Viremia below the Limit of Detection. Eleventh Conference on Retroviruses and Opportunistic Infections, San Francisco, abstract 427b, 2004.

Scharko AM et al. Whole body positron emission tomography in patients with HIV-1 infection. Lancet 362: 959-961, 2003.

Siliciano R et al. Viral reservoirs and ongoing virus replication in patients on HAART: implications for clinical management. Eighth Conference on Retroviruses and Opportunistic Infections, Chicago, abstract L5, 2001.

Sonza S et al. Monocytes harbour replication-competent, non-latent HIV-1 in patients on highly active antiretroviral therapy. AIDS 15: 17-22, 2001.

Wang FX et al. Purging HIV-1 Viral Reservoirs: The Role of IL-7. Eleventh Conference on Retroviruses and Opportunistic Infections, San Francisco, abstract 424, 2004.

Wong JK et al. Reduction of HIV-1 in blood and lymph nodes following potent antiretroviral therapy and the virologic correlates of treatment failure. Proc Natl Acad Sci 94: 12574-12579, 1997.

Zhang L et al. Quantifying residual HIV-1 replication in patients receiving combination antiretroviral therapy. N Engl J Med 340: 1605-1613, 1999.

Zhang ZQ et al. Kinetics of CD4+ T cell repopulation of lymphoid tissues after treatment of HIV-1 infection. Proc Natl Acad Sci 95: 1154-1159, 1998.

Zhang ZQ et al. Reversibility of the pathological changes in the follicular dendritic cell network with treatment of HIV-1 infection. Proc Natl Acad Sci 96: 5169-5172, 1999.