Immune Reconstituion, Viral Reservoirs and HIV Pathogenesis

Restored Immunity. Immune system restoration has started in individuals successfully taking HAART. Because of HAART, opportunistic infections are dramatically decreasing in incidence and some people have had OIs go into remission. These improvements in health and the dramatic decreases in death are correlated with suppression of viral load and improvements in immunity. Some believe the best prophylaxis and maybe treatment for OIs is HAART.

However, many unanswered questions remain. At the Retrovirus Conference, improvements in a number of parameters of immune function were reported. But, the immune system and its response to HAART is not well understood. A number of studies reported increases in naive CD4. Some researchers feel the significance of these increases may not be understood, and that increases may be slow and modest. Some researchers believe that at some point in adulthood people may lose the ability to regenerate naive CD4s from the thymus. Dr Pat Bucey however, does not think you completely lose that ability.

How much immune restoration is possible? Some individuals’ immune response appears better than others. How much restoration is enough to sustain significantly improved health and life? We don’t yet have the answer. Is full restoration of immune function possible with continued suppression of viral load below 50 copies/ml? We don’t know, but it may be too soon to judge. If suppressed viral load rebounds, the immune restoration that has occurred may be diminished or lost. The goal of therapy may have to be to sustain suppression of viral load below 50 copies/ml. The development of resistance is obviously an impediment to that goal. Non-adherence and treatment regimens not individualized to a patient’s particular situation contribute to failure to maintain low viral load.

The key is that improved clinical outcome or improved health and life extension is occurring. Research is trying to understand the immune system and its response to HAART. It is possible improved treatment for individuals not responding well to HAART can emerge from a better understanding of the immune response. Perhaps we can understand why some individual’s immune system responds better to HAART than others.

A number of studies reported that the expression of the CD38 activation marker on CD8 cells decreased on therapy. RL Hengel, of the CDC, reported the intensity of the CD38 expression on CD4 cells decreased by 45% and by 56% on CD8 cells by the fourth month. This reflects a decrease in viral replication because there is less virus around. The CD38 marker increases in number and intensity when the immune system is activated in response to virus. Significant recovery in B cell function was reported for individuals on HAART and it is related to decreased HIV viral replication. Several studies reported increased response to DTH skin testing, a measure of the bodies’ response to antigen.

There is controversy on the source of the increased CD4s resulting from HAART. The findings of several researchers suggest that the increases in absolute numbers of CD4 and naive CD4 are the result of redistribution from the lymph tissue rather than the regeneration of new naive CD4 from the thymus. It is also possible that increased CD4 cell counts could be the result of peripheral expansion (expansion of CD4s in peripheral blood), or reversion from memory cells, as seen with rats and mice. It is possible new naive CD4s can only be regenerated from the thymus. Although regeneration from the thymus might offer a better chance for regaining a full CD4 repertoire, some researchers believe the thymus cannot regenerate new naive CD4 for adults with HIV since age and HIV infection degenerated or destroyed the thymus. Mike McCune, of the Gladstone Institute, has suggested that his research shows that the abundance of thymus correlates with the amount of naive CD4. An increase in naive CD4 does in fact reflect a functional thymus. However, some question his premise indicating the presence of thymus for the individuals tested because he based it upon taking CT scans. It’s suggested that a CT scan may only be detecting a shadow and not an actual thymus.

Crystail MacKail, of the National Cancer Institute, has reported her findings that cancer patients under 16 years old who suffered T-cell depletion due to chemotherapy restored normal CD4 numbers via thymic-dependent pathways within 6 months. Meanwhile, older patients who regenerated CD4s primarily via thymic-independent pathways showed prolong CD4 depletion, decreased CD4/CD8 ratios, and diminished CD4 naive/memory cell ratios. They also showed a propensity for programmed cell death and a loss of T-cell repertoire.

A so-called naive CD4 is born from the thymus with a specificity to respond and mobilize an immune response to a specific infection such as CMV or MAC. After meeting an invading infection for the first time that CD4 expands in numbers into a pool of memory CD4s, capable of responding if that infection invades again. The presence of naive CD4s may be crucial to respond to an infection.

A number of researchers believe that disease progression associated with viral replication preferentially destroys naive CD4s rather than memory CD4s. Dr. Pat Bucey, of UAB, does not believe naive CD4s are destroyed by disease progression, but rather are hiding. He believes that active viral replication causes an inflammatory response which results in CD4s getting stuck in the spleen, lymph tissue, and possibly the intestine. HAART inhibits the inflammatory process and permits the release of these cells, which include naive CD4s. He believes the first phase of CD4 increase after starting HAART comes almost completely from redistribution and starts a functional recovery of the immune system. After six weeks, the second phase of CD4 increase begins. About 80% may be due to peripheral expansion or reversion of memory cells to naive cells and 20% may come from the thymus. Though others believe that people lose their thymus with age, Bucey maintains that a little remains even if it’s only 5% of the original size, which is capable of producing new CD4s.

RL Hengel of the CDC reported findings from a study of 20 individuals treated with protease inhibitors. At baseline, 8/10 individuals with >100 absolute CD4 had naive CD4, while only 2/10 individuals with <100 CD4 had naive CD4 at baseline. This suggests that individuals with low CD4 (>100) were less likely to have naive CD4s. The individuals with naive CD4s at baseline, regardless of their absolute CD4 count, immediately repopulated their naive CD4s at a greater or equal rate as memory cells. Hengel found that half of the individuals with no baseline naive CD4s did not repopulate naive CD4s at all, while the other half were able to reconstitute naive CD4s, but at a slower rate than the individuals who had naive CD4s at baseline.

Viral Reservoirs. Previously, Dr Robert Siliciano was the first to report a HIV reservoir of post integrated replication competent proviral DNA in resting memory CD4 cells. He reported this reservoir is in a non-productive or latent state but that it can be reversed to produce infectious virus if stimulated by antigen. Previously, he examined in a cross-sectional study a group of 22 patients receiving HAART for up to 30 months who were undetectable (<200 copies/ml). Replication competent virus was found in 18/22 individuals. They couldn’t isolate enough cells to complete the study in the other four. The size of the reservoir was found to be the same size for persons regardless if they were on therapy for 30 months or just several months. He concluded that this reservoir did not decay, which presents an obstacle to the possibility of virus eradication. At the Retrovirus Conference he reported new research. Based on preliminary longitudinal studies there was no evidence of decay in the size of the reservoir.

However, David Ho reported data suggesting that this reservoir may decay slowly, with a half-life of three months. He studied 18 individuals receiving HAART. Four patients were chronically infected and 14 were acutely infected. He found less replication competent virus than Siliciano. At 18 months on therapy, he found infectious virus in 43% of individuals and in 28% after 22 months. Although this difference was not statistically significant, it suggests decay in the reservoir. When looking at proviral DNA in PBMCs, his preliminary findings suggest a steady decay with a half-life of 110 to 120 days.

The Researchers used RT-PCR to measure different forms of mRNA within PBMCs and tissues. MS (multiply sliced) mRNA should only be found in productively infected cells. The US (unsliced) mRNA could be detected in productively infected cells as well as within virions trapped on the cell surface or as infectious core within the cytoplasm. They do not find MS mRNA. In some cases they are finding US mRNA. They are unsure why. Ho said, a more likely explanation is that all of these individuals are still proviral DNA positive and much of this is defective virus. It is possible a fraction of these cells would turn on and have US mRNA that we could detect. Ho stressed this is speculation.

Based on these cases, which have been mainly acutely infected individuals (14/18), they are making preliminary conclusions. They are seeing a range in pool size, in one case, of <4000 for the total body of infected cells (with proviral DNA) to about 100,000 with a mean of about 30,000. These numbers are lower than Siliciano described, but are taken from acutely infected patients. Proviral DNA copy numbers are about 10 fold lower in acute patients than in chronic patients. He has a few cases that seem to decay at a half-life of a few months. Other cases don’t seem to decay, they believe that the decay may just take longer in these cases.

Based on a worse case scenario with a pool size of 100,000 and a decay half-life of 18 months, it would take longer than 20 years for the compartment to decay. Ho suggests, a more realistic scenario would be a pool size of 30,000 with a decay half-life of 3-6 months based primarily on the proviral DNA decay data. In this situation it would take about 5 years or slightly more for the pool size to decay to less than one. I assume this implies an uninterupted "full" suppression of virus load.

Dr. Tae-Wook Chun reported his preliminary findings from his ongoing research using sensitive testing. In 4/6 individuals with undetectable HIV RNA in plasma and replication competent DNA found after 10 months of HAART, their reservoir of integrated proviral HIV DNA in resting CD4+ memory cells was decaying, but "very slowly." He looked at the individuals data twice, six months apart. Based on his research to date, Dr. Chun suggests the decay may not be due to antiretroviral therapy, but to the cells just dying off over time due to their half-life. However, these latently infected cells may not die but could hang around forever, and which would be a concern.

The half-life for virus was 108 days. Chun estimated that with 100% suppressive therapy it would take 5.8 to 7.8 years to achieve eradication only in this reservoir. This doesn’t include any other reservoir that may be sequestered such as the brain. If there is virus production in the brain there should be integrated proviral DNA present in the cell host. Without integration there is no virus production. HIV production has been found in the brain of HIV infected individuals upon autopsy after death. It is likely that macrophages in the brain are productively infected. The six patients mentioned above were taken from an original group of 13 patients who received HAART. Nine were <500 copies/ml and four were detectable. In all 13, integrated replication competent proviral DNA was found in their resting CD4 memory cells 10 months after starting therapy. The six patients mentioned were selected from the nine undetectables. Second samples were taken about eight months later from these six, and 4/6 had replication competent proviral DNA.

Latent virus in resting CD4 memory cells can become activated at any time by a number of sources such as an intercurrent illness, a vaccine, or an opportunistic infection. Once a person stops therapy that has been fully suppressive to undetectable in plasma, there is no protection from activated latent virus. It is recommended to remain on therapy if you are undetectable. If a person is on therapy and the latent virus is activated, therapy can target that virus. In fact, that may be what is occurring. Individuals with undetectable HIV RNA continue to be undetectable because their therapy targets activated latent virus.

The ACTG is starting a study in an attempt to activate this virus reservoir and accelerate its rate of clearance of HIV. A total of 24 individuals will be enrolled at the Aaron Diamond Research Center, University of Colorado, and University of California at San Diego. Participants will receive indinavir 1000 mg every 8 hours+nevirapine 200 mg twice daily+AZT/3TC. In addition, individuals will receive either GM-CSF or IL-12 to activate the reservoir in the cells. Some will not receive immune modulation as a comparison control group.

Dr. Bucey also reported on a viral reservoir of RNA. In data presented at the Chicago meeting, he showed data indicating the persistence of a few cells in lymph node tissue that appear to be continuously producing virus. The quantitative amount of viral RNA present in each of these cells was essentially the same as those cells making virus prior to HAART. The frequency of these cells decreased significantly with treatment, but not the apparent amount of virus produced in each infected cell. At a subsequent presentation at an ACTG retreat, Dr. Bucey showed data extending this analysis to more lymph node specimens and analysis of the blood in additional individuals on HAART. Results were similar, showing the persistence at very low levels of continual viral infection. After a period of years, which may differ individually, he suggests this virus could develop drug resistance and cause viral load rebound.

Dr. Janice Clements reported at the conference that her research of SIV in macaques monkeys strongly suggests replicating virus in the CNS can traffic back and seed the peripheral blood, which was observed to be a frequent occurrence. However, this is when virus replication is high because there is no ongoing treatment in these monkeys. The monkeys had a high viral load in the CNS. If replication is controlled by treatment, Clements said we don’t know if in the SIV infected macaques’ virus traffics into the peripheral blood, although she feels it may only be when there is active replication. If this trafficking occurs only when replication is active or high, latently infected DNA may not traffic back to the periphery. These processes are very difficult to study in people.

Viral Reservoirs in Genital Secretions for Men and Women

At the Retrovirus Conference, Dr. Roger Pomerantz discussed how blood-tissue barriers are also important sanctuary sites where a few latently infected cells may hide behind the tissue blood barriers. These sites, which include the brain, retina, testes, and cervical vaginal tract, each have something in common. They have a barrier, a tight junction in their endotherial lining, which leads to what is commonly referred to as a blood-brain barrier, a blood-retina barrier, or a blood-testes barrier. However, it’s important to remember there has not been a described barrier in the female genital tract.

Genital secretions can be a sanctuary for HIV because of the presence of latently infected cells. He discussed HIV in the seminal fluid of patients on HAART and HIV in the cervical-vaginal secretions of women who are both pregnant and non-pregnant. He also discussed why virions themselves might be different depending on the tissue reservoirs in which they are produced, particularly in the seminal fluid.

HIV in Seminal Fluid of Men. His research group looked for HIV in seminal fluid in both cell free virion RNA and infected monocytes/ macrophages (cells). Researchers also examined potentially infected sperm and spermatic progenitors. After findings of HIV in the seminal secretions of infected individuals from both his group as well as others, his group carefully looked at four men, all who are on various combinations of HAART. They have had undetectable virus RNA in the blood for four months to two years, with no spikes of detection when testing every three months.

Looking at the seminal fluid of these men using a sensitive test (RT-PCR in a quantitative analysis), they found very low levels of HIV RNA. The level of viral RNA was below the level of clear quantification in this analysis and clearly would not have been picked up if you used standard <400 copy/ml RNA PCR. When looking at the DNA of these seminal secretions in the sperm monocyte/ macrophages in the cellular component of the semen, they found quantifiable DNA in two of those four patients. These appear to be latently infected cells that have proviral DNA months or years after the patient has no viral RNA detectable in the blood (<400 copies/ml). Using a more sensitive test, they picked up low amounts of DNA in the seminal cells in each of the four patients. Upon quantification, they found the amount to be extremely low, but still quantifiable. In one patient they were able to detect replication competent DNA in their PBMC (peripheral blood mononuclear cells). In summary, these four patients show it is possible to have undetectable virus (<400 copies/ml) in peripheral blood, yet still have both DNA and RNA at very low levels in the seminal secretions. It is uncertain if these reservoirs will present a problem upon cessation of therapy or in transmission of HIV. Further studies are needed to address this question.

HIV in Cervical-Vaginal Secretions. Women could have undetectable virus (RNA) in blood plasma but still have virus (RNA) detected in cervical-vaginal secretions. Looking for HIV in cervical-vaginal secretions, they looked at non-pregnant women treated with various antiretroviral regimens with both detectable low virus RNA and undetectable virus RNA in the blood. In looking at the RNA from the cervical secretions (cell free, outside of cells) a number of patients had detectable and quantifiable RNA. About four women had no RNA detectable in the blood, however they did have low levels in the cervical secretions. These levels would not have been picked up by standard testing, but were picked up by sensitive assays. DNA was also detected. Pomerantz said a person could be DNA positive in their cervical secretions, though they may not necessarily find RNA. Their data showed a general trend that agreed with the findings in some of the different reports at Retrovirus. These reports suggest that when lowering viral RNA in blood, the amount in the cervical secretions in most patients decline.

At Retrovirus, U. Uvin and others reported in "HIV-1 RNA Levels in the genital tract of Women on Antiretroviral Therapy," that antiretroviral therapy reduces HIV-1 RNA levels in the genital tract. The level of HIV RNA in cervicovaginal lavage (CVL) is significantly correlated with the level of HIV RNA in plasma. Twenty women were studied. Almost all women (12/13 ) with detectable CVL HIV RNA had a decrease in CVL HIV RNA with the addition of a protease inhibitor or modification of nucleoside therapy.

Pomerantz’s study group has 51 pregnant women with various levels of HIV RNA in their blood. Once again, looking at the HIV RNA in the cervicovaginal lavage they were still able to detect viral RNA in a number of these women who had high, low or undetectable HIV RNA in their blood. DNA could also be found with some discordance between RNA and DNA. Although the levels of RNA found in CVL may be low, it can be present in both pregnant and non-pregnant women. The majority of the 51 women in this group have both RNA and DNA detectable and quantifiable in their CVL secretions, suggesting that reservoirs for HIV may not require a blood tissue barrier to be of potential significance. You’ll remember that earlier in the discussion Pomerantz mentioned several potential HIV reservoirs having a blood-tissue barrier (brain, retina, and testes). However, that the cervical-vaginal tract does not have a blood tissue barrier.

Virus in Seminal Fluid Could Be More Infectious than in Blood. Pomerantz said in vitro (lab) data suggests that seminal fluid make virion particles more infectious than those that come out of peripheral blood. However, it remains to be researched in vivo (humans). Intravirion HIV reverse transcription is a stage of the lifecycle. Remember, AZT is a reverse transcriptase inhibitor (the RT enzyme promotes virus reproduction). Pomerantz said it has been demonstrated in a number of articles over the last two years in various journals that reverse transcription starts in virion particles and not always after binding to the target cells.

Pomerantz’s research group found ways in which intravirion reverse transcription could be stimulated in seminal fluid before a virion binds to a target CD4. There seem to be pores within virion particles that allow the ingress of "chemicals" which allows virion particles to be stimulated towards reverse transcription. Due to this stimulation of intravirion reverse transcription in seminal fluid, Pomerantz said it makes these virion particles far more infectious.

Pomerantz finished by saying it is important to consider things other than latency of cells and blood tissue barriers. You should also think about the virions themselves. HIV virions are biochemical active particles that are altered by the microenvironment in which they sit. The virions themselves may be critically different depending on the tissue reservoir in which they are being produced.

Correlation Between Blood and Seminal Viral Load for Men

JS Evans reported findings at the Retrovirus Conference from a cross sectional study of 79 men participating in the USAF Natural History study called, "Correlates of HIV Detection by Viral Load and Culture in Semen."

The investigators concluded that HIV RNA in seminal plasma in these male study participants correlated strongly with blood plasma levels but were 5.5 fold lower. This suggests that if a person has a lower blood plasma VL they may also have lower seminal plasma VL. There was also a significant correlation between seminal plasma viral load and positive seminal plasma cultures. However, nine patients (11%) were an exception. These nine had three times higher blood plasma VL than the other 70 patients although their seminal plasma VL was >20 times greater than the others. So, their blood plasma viral load did not correlate with their seminal plasma viral load. An important message from this study is that these nine individuals could represent a concern for increased transmission of HIV, because their seminal plasma viral load is so disportionately higher than their blood plasma viral load. In other words, in some individuals you may not be able to judge the level of viral load in their semen based on their blood viral load. Thus, they may be more likely to transmit HIV.

Interestingly and with possible important implications, the therapy receiving group had more advanced HIV (CD4 322 vs 551 for those not on therapy), and a distinctly higher proportion of them showed undetectable seminal plasma HIV RNA than those not receiving treatment for HIV (60% vs 17%). In the study group, 37 (47%) were receiving antiretroviral therapies (19 receiving a PI therapy). This study suggests that reducing blood plasma viral load with antiretroviral therapy should reduce seminal plasma viral load.

The median seminal plasma viral load of the 79 men was 2,100 copies/ml, 63% of who had detectable viral load in their seminal plasma, and 9% were HIV culture positive. Viral load was detectable in the blood of 97% of the men. Their blood plasma viral load was 11,500 copies/ml, and 22% were HIV culture positive.

HIV Pathogenesis (Written by Larry Cohen and Jules Levin)

Dr. Bruce Walker’s research suggests that individuals identified with acute HIV infection should be immediately treated with HAART. He has reported findings that show a correlation between the presence of CD4 cells with HIV specific responses, a resulting CTL (cytotoxic T lymphocyte) response to HIV, and control of viral load. He reports that this response, which may be present after initial infection will wane, but can be preserved by HAART treatment during acute infection.

Walker studied an individual who was HIV+ for 18 years, had a normal CD4 count, a viral load of <400 copies/ml, and had never been treated with antiretroviral therapy (a so-called long-term non-progressor). He detected an "enormous" CTL response broadly directed against multiple HIV antigens, which was associated with a strong HIV specific CD4 response. Subsequently, he treated seven individuals with HAART during acute infection, prior to sero-conversion. He also examined four individuals treated during acute infection with hydroxyurea+ddI+a protease inhibitor in the Berlin study reported by F. Lori and Jessen. Walker reported that all 11 individuals generated strong and increasing virus specific CD4 proliferative responses, and that these responses were persistent for up to one year of follow-up.

Walker maintains that he has not found a person who has a strong virus specific CD4 response who does not also have a strong CTL response. He suggests that the CD4 response stimulates the CTL response, which translates into control of HIV. Thus, he maintains the host immune system is capable of contributing to the containment of HIV, but that disease progression depends on several additional factors. He says the immune response by long-term, non-progressors is similar to what he has observed in the 11 individuals treated during acute infection. However, he added that we’re uncertain how long-term non-progressors retain this immune response without therapy.

In Science (V 278, 21 Nov ‘97), Walker reported on his study of 10 untreated individuals with varying clinical histories and viral loads ranging from <400 to 300,000 copies/ml.

Viral load correlated with the presence of a virus specific immune response. Individuals with the strongest p24 specific response had the lowest viral load. Individuals with the highest viral load had markedly decreased ability to respond to p24, the antigen associated with HIV.

Walker’s research suggests that there is an initial CTL response following infection that wanes over time. Initiating HAART during acute infection preserves this response. Walker suggests virus specific CD4 cells may be selectively eliminated during primary infection. This could occur either because these cells would be activated during the period of increased virus production during acute infection, making the cells particularly susceptible to infection, or these cells might be selectively eliminated by activation induced cell death during acute infection.

Walker said at the Retrovirus Conference, "….People with acute infection should be treated immediately with potent antiviral therapy." His data makes a case for considering HAART during acute infection if you can identify that a person has been recently infected. He thinks individuals will probably have to be treated within six months of infection to preserve the HIV specific CD4 response, but he is uncertain about the time frame. Ongoing research may soon tell us how long that window of opportunity may remain open.

Chronic Infection. It is uncertain if this HIV specific CD4 response and the CTL response can be elicited in individuals with chronic infection. They may have lost the ability to mount such a response. In his Science article, Walker said that although therapy with potent antiviral drugs in persons with chronic infection has resulted in increases in CD4 cell counts, it has not resulted in the restoration of strong or persistent virus specific CD4 proliferative responses. Therapeutic vaccines, that is giving vaccines to individuals with HIV infection, got a bad name because studies were unable to show good responses by individuals receiving these vaccines. However, that was before we had the ability to control viral replication with HAART. Walker contends that controlling viral load and replication with HAART may allow for a maturation of the immune response. It is possible that DNA vaccines may be able to elicit the immune response. A human study of such a vaccine is ongoing. Another such vaccine is in an earlier stage of development.

However, a group at UCSD (S Little, D Havlir, D Richman and others) reported at Retrovirus that "profound abnormalities of circulating CD4 and CD8 cell populations" are still present in five individuals, despite being treated with HAART during acute infection. Among several abnormalities reported were that these individuals had reductions in naive CD4. Significant recovery occurred only in those patients with the least initial reductions. There was a selective and persistent loss of cells expressing the IL-2 receptor. Other abnormalities were reported (abstract 587).

At the Retrovirus Conference, Dr. Norman Letvin reported data from his studies of acute infection in rhesus macaques monkeys with SIV (simian immunodeficiency virus). In one experiment, a CTL response was detected in regional and distant lymph nodes (lymph nodes under the arm pit) within 72 hours of vaginal exposure to cell free virus. This suggests that either the virus traveled quickly from the area of exposure to the distant site or the local immune response traveled up to the distant area. In a second experiment in which monkeys were infected intravenously about 5% to 10% of all T cells in the regional and distant lymph nodes were replicating virus two weeks after infection. However, by day 16 this replication was starting to be contained. This indicates that the immune system is stimulated very early after infection and that CTLs are assisting in containing replication. Letvin says he and others have thought for a long time that a CTL response is not possible without the CD4 virus specific response, that these responses were required for an adequate immune response; but, the window period may only be a week after infection