Immunology -- IL-2, CD4 changes during STI, CTLs, Chemokins receptors
     Written for NATAP by David Margolis, MD, Associate Professor, University of Texas Southwestern Medical Center

A variety of reports immunomodulatory therapies, and of the immune responses to and consequences of HIV infection were presented at the 8th Conference on Retroviruses and Opportunistic Infections. While no groundbreaking findings were unveiled, attendees saw continuing evidence of incremental progress in the understanding of HIV pathogenesis and immune therapies. Many findings that should inform rational therapeutic choices were presented, and directions for future studies outlined.

(Comments from Jules Levin: the main appeal for IL-2 may be for people who get a good viral response from HAART to undetectable but still have low CD4s. A number of studies have shown that even with CD4s as low as 100-150 you still can see CD4 increases. The remaining question that many doctors still have--are these cd4 increases clinically beneficial? ESPRIT and another study will try to answer this question. Data from several studies, including ones below, suggest that the cd4 increases from IL-2 give clinical benefit. The studies below are encouraging in suggesting further that cd4 increases from IL-2 are beneficial. But we have not yet had results from a well designed study whose objective was to answer this question. ESPRIT & the other study are designed to answer the question. Unfortunately, we won't have an answer from these studies for several years).

Interleukin 2: clearly good for your T cells, but is it good for you?
A variety of studies performed in different patient populations and clinical settings, demonstrated that adjunctive IL-2 therapy could result in substantial rises in CD4 cell count. The toxicities and side effects of IL2 are well known. Whether this therapy is worth its cost and toxicity awaits the outcome of ongoing clinical endpoint studies in both early (ESPRIT) and later-stage patients (SILCAAT). If these studies are favorable, the broad experience outlined below will validate the use of IL2 in a variety of settings.

Slight Hint of Less OIs.
Ron Mitsuyasu (#17) reported the final 84 week results of ACTG 328, a long awaited randomized study of the effects of HAART with or without continuous IV (CIV) or subcutaneous (SC) IL-2. Protease inhibitor-naive patients with CD4 cell counts 50*350 were treated with indinavir + 2 NRTIs for 12 weeks. At week 12, patients with HIV RNA <5000 c/ml were randomized to HAART alone (n = 52), HAART + CIV IL-2 at 9 million units (MIU) qd X 5d q8wks (n = 54), or HAART + SC IL-2 at 7.5 MIU bid X 5d q8wks (n = 55) for up to 84 wks. As allowed by protocol, most patients on CIV (76%) switched to SC IL-2 after 3 or 6 cycles of IL-2. As dose rducton was allowed for toxicities or poor tolerance the average daily doses of IL-2 were 9.3 MIU/d in CIV and 9.8 MIU/d in SC arms.

While CD4 counts rose progressively on HAART alone, increases were substantially larger in patients receiving IL-2. Median CD4 at week 84 and changes from baseline to wk 84 were as follows:

HAART 396 and 121; CIV 800 and 480; SC 614 and 302. While 41% of subjects receiving HAART had >50% rise in CD4, those proportions were 86% in CIV and 77% in SC.

There was no overall change in HIV RNA across the arms of the study. Mild side effects were troublesome, but serious Grade 3/4 toxicities were infrequent and likely related to IL-2: nausea, vomiting, fever, and fatigue. Some patients on SC IL2 had an asymptomatic increase in bilirubin, thought to be related to increased indinavir levels induced by IL-2.

In this largest prospective randomized study of IL-2 in advanced HIV patients to date, significant increases in CD4 counts were seen with both CIV and SC IL-2 compared to HAART alone after 60 wks of therapy, which continued to increase and grow larger to week 84. Only 1 AIDS-related infection was seen in both of the IL-2 arms, and 5 in the HAART arm, giving a hint that IL2 may have clinical benefits, but direct proof of this still awaits the completion of ongoing clinical endpoint studies.

Real Immune Resconsitution?
Corroborating this study, Saint-Mezard et al (poster 350) reported evidence of increase of thymic production of naïve T cells with scIL2 therapy in advanced HIV patients treated with HAART. 14 patients with CD4 <200/µl received adjuvant scIL2 4.5 MIU bid in 4 cycles every 6 weeks. The proportion of T cells recently produced by the thymus was measured with a PCR assay that detects a genetic DNA recombination event that occurs during development of T cells in the thymus. IL2 therapy induced a significant increase in the percentage and the number of naive CD4+T cells and of memory CD4+T cells, paralleled by an increase of the number of TRECs in the CD4 T cells subset. This study provided evidence that IL2 can do more than just increase T cell number, it can contribute to real immune reconstitution.

Nadir CD4 Predicts IL-2 Response Suggesting Immune Defects Persist Even After HAART; Nice CD4 Increases Even When Nadir CD4 <250.
Markowitz (poster 343) reported a community study of IL2 in patients who had responded to HAART. CPCRA 059 was a multicenter, open-label, randomized trial of SC rIL-2 at 7.5 or 4.5 MIU twice daily administered for 5 consecutive days every 8 wks for at least 3 cycles. 126 and 130 pts were assigned the 7.5 and 4.5 MIU doses, respectively: 192 pts completed >3 SC rIL-2 cycles and 245 had a month 12 CD4+cell count. At BL viral load was <50 copies/ml in 60% of pts. Median pre-enrollment nadir and BL CD4+cell counts were 293 cells/µl and 538 cells/µl, and median time on ART at entry was nearly 4 years.Nadir CD4+was the strongest predictor of SC rIL-2 response, suggesting that immune defects may persist. However, longer followup and treatment of these poor responders may demonstrate that immune reconstitution with IL2 takes more than 12 months in some subjects. Body mass index also correlated with CD4 response, suggesting that BMI should be considered in selecting SC rIL-2 dosing. Even in the worst cohort, those with CD4 nadirs < 250/µl and baseline CD4 counts < 480/µl, CD4 response at 12 months was +363 cells. If IL2 can be shown to confer a clinical benefit, this study would suggest that it may work in community settings, even in patients with long histories of HIV infection and therapy.

(Comments: Mike Youle, a researcher/doctor in London has presented data on treating people with IL-2 earlier in disease progression before they receive their initial HAART regimen. The purpose is to delay starting HAART by keeping cd4s elevated. Another study planned in the ACTG will look at boosting cd4s with IL-2 & conducting therapy interruptions. HAART would be resumed after cd4s decline to certain level. This might permit longer therapy interruptions).

5 MU Twice Daily For 5 days Every 4 Weeks More Tolerable.
For the French ANRS 079 study group, Levy (poster 344) reported patients with CD4+ cell counts of 200-550 cells/µl without prior AIDS events and naive to PIs, randomized to start new therapy with D4T + 3TC + Indinavir or HAART combined with SC IL-2 cycles (5 MIU, twice a day for 5 days) every 4 weeks for 3 cycles and then every 8 weeks for 7 cycles. A scheduled analysis of the first 80 patients at week 74 weeks showed +865 cells/µl in the IL-2 arm vs +240 cells/µl with HAART alone (P<.0001). Nearly twice as many patients on IL2 achieved an 80% increase in CD4 count; no difference in viral load wereseen. A subgroup analysis (poster 345) found a marked increase in naive (CD45RA+/CD62L+) CD4+ cells in the IL-2 group compared with contols (+475/µl vs +83 cells/µl; P<.0001). This intermediate dose of IL-2 combined with HAART allowed a greater increase in CD4 cells than HAART alone, with similar control in HIV replication and fewer dose reductions.

7.5 MU Less Tolerable.
A meta-analysis (poster 346) of the CD4 cell response to 3 doses of SC IL-2 in 3 Vanguard studies performed in Houston, Argentina and Thailand again found a significant dose-dependent increase in CD4. Again scIL-2 was given bid in 5-day cycles q8 weeks for 3 cycles to subjects on ART with CD4 >350 cells/µl. While the analysis supported the ability of the 7.5 MIU dose to most rapidly achieve CD4 counts > 100/µl, this dose was also associated with higher dropout and toxicity rates.

Long Term IL-2 Followup Receive IL-2 Every 26 Months.
The NIAID group also reported extremely long-term followup of subjects on iL2 for extended periods (Chiatt et al, poster 347). Of 97 patients originally enrolled in IL2 studies between 1993*1997, 73 entered the extension phases of these trials and 63 remain on active follow-up. The mean duration of study participation for these 63 patients is 4.6 (range: 3.4 *7) years. Their mean CD4 cell count and CD4% rose from values of 532 cells/µL and 27% at study entry to present levels of 1030 cells/µl and 40%, respectively. Patients tolerated a mean total daily scIL-2 dose of 11.9 MIU (range: 3.0*15.0) during their most recent cycle. No AIDS-defining OIs have occurred in this cohort. Median plasma virus levels dropped from a baseline of 3.40 log10copies/ml to 1.69 log10copies/ml. These subjects, now on study for a mean of 84 months, are now given IL2 every 26 months, on average, to maintain CD4 counts > 1000/µl.

The nettlesome side effects of IL2, prompted therapeutic attempts to blunt IL2 toxicity. Sha et al. (poster 66) tested the ability of recombinant tumor necrosis factor (TNF) receptor (Etanercept) to blunt the symptoms caused by IL-2 therapy. TNF is a proinflammatory cyokine, produced in response to IL-2. 11 patients within ACTG 328 were studied: HAART only controls (5), HAART + SC IL-2 (3), HAART + CI IL-2 (3). Etanercept pretreatment attenuated increases in the cytokine IL-6 and the acute phase inflammatory response plasma protein CRP following IL-2 treatment. An effect on TNF bioactivity could not be assessed as all measurements were below limits of detection. No changes were seen in temperature or plasma levels of the cytokines IL-4, IL-10, IL-12, or IFN-gamma. Unfortunatgely, etanercept did not decrease the incidence of grade 3 toxicities when administered before a single cycle of IL-2 in 6 subjects studied. The incidence of fever was also unaffected by etanercept. Etanercept did not appear to be a particularly effective strategy for increasing the tolerability of IL-2.

Prednisone Reduces Toxicity But Blunted CD4 Increases.
A successful approach to blunt the side effects of IL2 came at the cost of CD4 cell response to IL2 therapy. Tavel reported for the Lane group at the NIAID (poster 348) a study of subjects on HAART with baseline CD4+cell counts over 350 cells/µl were randomized to receive rIL-2 7.5 MIU twice a day subcutaneously for 5 days every 8 weeks in combination with placebo or prednisone dosed at 0.5 mg/kg divided into two daily doses during and two days after rIL-2 administration. rIL-2 dose could be adjusted in decrements of 1.5 MIU per dose for toxicity management. Analyses were performed on an intent-to-treat basis. Ten patients completed a total of 37 cycles of rIL-2 with prednisone or placebo with a mean follow-up of 7 months. Fewer toxicities were seen in the group receiving prednisone (31 grade 3 or 4 toxicities in 3 patients receiving placebo, only 3 grade 3 or 4 toxicities in 2 patients receiving prednisone; p < 0.001). HIV-RNA decreased or remained undetectable in all patients, with no significant difference between the groups. However, while fewer grade 3 and 4 toxicities are observed in rIL-2 recipients receiving prednisone, this trend does not permit higher rIL-2 dosing. In addition, rI-2-associated CD4+cell count increases are significantly blunted by prednisone coadministration (mean increase of 0.7 CD4 cells per day in the prednisone group and 2.4 cells per day in the placebo group; p = 0.02).

Immune Defects Persist.
In contrast to the difficulties of IL2, two posters highlighted the limitations of HAART with regard to immune reconstitution, and the risks of delayed HAART. Valdez (#372) analyzed data from ACTG 375, a prospective, open-label study of the immunologic effects of HAART on protease-inhibitor-naïve patients. Most patients with VL under 1000 c/mL at 154 weeks of HAART who began HAART with moderately advanced HIV infection do not normalized their CD4 counts after 3 years of therapy.

When To start Therapy.
Lange (#380) showed that fewer CD4 cells capable of responding to immune signals return when HAART is started with lower CD4 counts. This argument for starting therapy earlier must be balanced with the recognized difficulties and toxicities of therapy.

Immune depletion and reconstitution: why do we care how to T cells return?
The conference provided several further chapters in the lively and entertaining ongoing debate as to the mechanism of CD4 cell loss and repletion. Although these friendly arguments may appear arcane, their outcome will eventually define a central issue in HIV therapeutics: how is the immune system damaged by HIV infection, and what must we do to repair damaged immune systems. In simpler terms, it is a debate over whether AIDS develops because of the tap of CD4 cells runs dry (cd4s are not produced any longer), the drain of T cell destruction is opened (cd4s are killed off), or whether the truth lies in the nuanced in-between.

(Comments: the data suggest that with HAART CD4 increases functional effective cd4s are produced but immune defects persist. So deferring therapy risks less effective cd4 cells).

The session began with data presented by H. Mohri (#269) who use radio-labelled glucose to document elevated T-cell turnover as a result of HIV infection in 7 patients. This marker can be safely given to people, and the incorporation of label into DNA measured to document cell proliferation and replication. Prior to HAART, an 8-fold increase of CD4 cell replication over seronegative controls was found. 4 drug HART regimen resulted in a significant reduction (5 fold) in T-cell turnover. Mohri concluded that CD4+ destruction was the result of increased cell turnover (cell death) and not the result of reduced thymic production (t-cells could still be produced).

A further chapter in the ongoing debate of HIV pathogenesis as T cell destruction vs. decreased T cell production was provided by D.C. Douek (#270). He compared the reaction by the body to low lymphocyte counts induced by radiation/chemotherapy used to allow bone marrow transplantation, to immune reconstitution in HIV-1 disease. Dr Douek presented data documenting levels of recent thymic emigrants, i.e. naïve T-cells, by measurement of T cell receptor rearrangement excision circles (TREC). This PCR assay detects a DNA remnant left in a new T cell that has developed in the thymus. A cell that is TREC positive is a new T cell. As a cell ages and divides into multiple generations of daughter cells, its TREC signal wanes by dilution.

In HIV infection, they found no evidence of proliferation of naïve T-cells, decreases in thymic output (TREC) and an expansion of memory cells.
Once HAART was initiated immune reconstitution mirrored that observed for patients after marrow transplant: new TRECs were seen in CD4s and CD8s, suggesting the body cannot tell the difference between the levels of CD4s or CD8s, but reconstitutes by producing new cells.

Lewin and coworkers (poster #370) also found that reduction in TREC in HIV-1-infected individuals may be partially explained by increased lymphocyte proliferation, but that increases in TREC number following HAART may also indicate the production of new T cells.

While this debate is not over, a role for both HIV-induced T cell loss and decreased production of new T cells in the development of AIDS, and of the ability of HAART to reverse both processes is supported. The difficulty of complete reconstitution in late stage disease again emphasizes that while there are now reasons not to want to "hit hard and hit early," there are also reasons not to wait too long.

Cytotoxic T cell response: a critical parameter controlling HIV infection
Much of a Tuesday slide session concentrated on analysis of and functionality of cytotoxic T-lymphocytes (CTL), highlighting the importance of an HIV specific CTL response and how HIV has managed to evade this by creating a blockade in the CTL maturation pathway.

An elegant presentation by Michael Betts (#272) covered a huge amount of data regarding total HIV-1 specific CD4+ and CTL response from 23 drug naïve patients. In an exhaustive and methodical approach, Betts for the first time examined the CTL response to every HIV protein. His study covered the whole HIV-1 protein sequence by using 756 peptides which cover the entire HIV protein sequence 10 amino acids at a time. The data presented demonstrated that current strategies of investigation of single or few restricted protein fragments does not provide an accurate picture of the HIV-1 specific immune response. Betts presented data indicating that no single protein response was representative of an individual's total anti-HIV response.

The most important point made in this presentation was the observation that CD8+ CTL response was positively correlated with viral load. The author put forward the hypothesis that more CTL will be present at higher concentrations of virus due to the presence of more antigens. This contradicted previous studies from other groups which stated that CTL was inversely correlated with viral load, i.e. better the CTL response = lower viral load. This result may have been due to the investigation of only a few epitopes, which were not representative of the total CTL response by that group. These findings have considerable import for the study of CTL response and vaccine development.

A poster (#166) by Casazza and colleagues looked at this picture from the opposite perspective. They measured the decline in HIV-specific CD8+T-cell response in patients receiving treatment with HAART. A rapid decline in HIV-specific CD8+T-cell reponse occurred as the viral load became undetectable. Use of data after an undetectable viral load was achieved resulted in a constant decrease in HIV-specific CD8+T-cell response that was well described by first-order kinetics. In the one patient who stopped treatment a rapid increase in viral load and HIV-specific CD8+T-cells occurred. In contrast to the response to HIV, response to CMV rapidly increased during therapy, reflecting either the return of the ability to respond to this common viral antigen. These findings reinforce the idea that the maintenance of immune response to HIV CTL during HAART is likely to require strategies such as vaccination.

Two poster from Bruce Walker's group (#167 and 168) provided further corroboration of the above studies. Again through the use of overlapping peptides and the interferon-gamma production assay, this group found previously unappreciated frequencies of CTL response to the HIV proteins Tat, Rev, Vif, and Vpr. Together these findings may be important for the design of future HIV vaccines, suggesting that broad response to many parts of HIV other than its envelope may be important for a protective or therapeutic vaccine.

The virus strikes back: evading CTL
Although nearly every HIV-infected person mounts an immune response that controls HIV replication for a time, immune control usually wanes over time. Three presentations discussed dysfunction within the CD8+ population induced by HIV. P. Champagne (#274) investigated the size of naïve, effector and memory CTL cells. The author proposed that HIV-1 somehow block the maturation of CD8+ from naïve to fully functional effector (killer) cells at an early stage. In HIV+ subjects 70% of CD8+ cells exhibited the cell surface markers of a immature (CD45RA- / CCR7-) phenotype. This phenomenon was not observed in CMV infection, suggesting a novel mechanism through which HIV may blunt the immune response that contains it.

This work was supported by data presented by D. Van Baarle (#275). She also measured the size of naïve, effector and memory CTL pools, comparing HIV specific, Epstein Barr virus specific, and total CTL pools. HIV/EBV co-infected individuals had consistently low levels of HIV specific effector CTL but EBV specific effector cells were present. L. Papagano (#276) also presented data demonstrating that long term progressors displayed numerous HIV-specific CTL that had not developed into mature effector cells capable of mediating an antiviral response. As these subjects had low CD4+ cell numbers, it was suggested that CD4 cell helper function needed to develop an effective CD8 response had been lost.

These findings may fit into the model proposed in recent publications by the Walker group and others, which suggest that antiretroviral therapy soon after acute HIV infection protects CD4 helper function, and allows a mature and durable antiviral response to develop. The challenge is now to develop vaccines or therapies which promote the ability to control HIV replication prior to infection, or in chronic infection.

Chemokines receptors and HIV infection: more keys to the lock
Since the discovery of the second class of cell surface molecule used by HIV to enter cells, the chemokine receptors have been an intense area of study. The conference offered new insights into the importance of these entry receptors, the molecules that bind them, and the cell signals they transmit.

Valenzuela Fernandez (#67) found that the activity of the chemokine SDF-1, which binds to and signals cells through the HIV co-receptor CXCR4, interacts with cell surface molecules known as Heparan Sulfate (HS). The authors and other groups have shown that chemokines associate with cellular or matrix extracellular glycosaminoglycans (GAGs), and are secreted from cells packaged with GAGs. The group made mutants of SDF-1 that could bind receptors and signal cells, but could not bind to GAGs. These mutants retain less than 30% of the capacity normal SDF to inhibit HIV-1 envelope-mediated virus-cell fusion. The authors concluded that binding of SDF-1a is packaged by HS, increasing its concentration near CXCR4 receptors, and permitting sustained occupancy and downregulation of the HIV coreceptor. These findings could be important to the formulation of chemokine analogs that might be used to block HIV infection in selected settings, such as use in vaginal microbicides.

Oral Contraceptives in Women & Risk of HIV-Infection.
In a poster presentation, Prakash et al (#76) suggested that oral contraceptive use might increase the levels of CCR5 receptor expression in T lymphocytes in the cervix of women taking combined oral contraceptives. In theory, this could lead to an increased risk of HIV infection. This was a very limited and preliminary study, and required significant expansion and further controls. However, the findings were of concern, and suggests there is yet another reason to implement effective strategies to prevent HIV transmission.

Carlos Paya's group presented (#77) their recently published findings that viruses that access cells via the R5 or X4 chemokine receptor may be able to kill CD4 cells via different mechanisms. They found by several assays that HIV using the X4 receptor, previously known as syncytia-inducing, and often seen predominantly in late stage disease, killed noninfected "bystander" cells without triggering the cell suicide mechanism known as apotosis. In contrast, R5, or non-syncytia-inducing HIV, often seen early in infection, required the activation of death receptor signaling pathways to kill CD4 cells. These findings suggest that approaches to selectively interrupt cell suicide signals may be therapeutically useful.

Other assorted immuno-tidbits of interest:

B-Cell Function reduced When Viral Load >50,000.
Moir (Late Breaker 2) presented new evidence to explain B cell dysfunction during periods of high-level viremia. HIV virions directly bind to B cells via complement receptor 2 or CD21. Chronically viremic patients have also been shown to express reduced levels of CD21 on their B cells. Despite high levels of serum antibodies, HIV patients have poor B cell function. For example, antibody responses to vaccines are often poor.

Moir found that both the number of CD21 receptors and the percentage of CD21-positive B cells were markedly reduced in patients with viral loads greater than 50,000/ml. High viremia significantly decreased he numbers of B cells able to proliferate and express cell surface molecules that enable B cells to perform antigen-presenting functions when stimulated. Further, the remainder of the B cell population that was CD21-negative were dysfunctional, less able to proliferate and more prone to spontaneous cell death.

Critically, lowering of viremia was associated with normalization of CD21 expression and enhanced capacity to proliferate and express B cell functional surface molecules when stimulated. These studies provided further evidence that high-level viremia induces important B cell dysfunction, may predispose to bacterial infection, and provides further rationale for antiretroviral therapy in the setting of high-level (> 50,000 copies) viremia.

Rate of CD4 Decline During Treatment Interruptions.
For those considering structured, strategic, spontaneous, or other forms of treatment interruption, Tebas (poster #355) provided useful background information. The rate of decay of the CD4 cell count in patients who discontinued antiretroviral therapy for at least 20 weeks after being fully suppressed with potent antiretroviral therapy in two clinical practices was measured retrospectively. 31 subjects were identified (43% males, 40% white, 32+2 yr of age). The most frequent reasons for discontinuation of therapy were patient preference in 56% (most frequently in females in the post-partum), and drug toxicity (43%). The median nadir CD4 (before therapy) was 383 cells/µl (interquartile range, IQR 244-588) and the baseline VL was 29,845 HIV RNA copies/ml. Patients gained on therapy an average 202+32 CD4 cells. The median time with an undetectable VL before stopping therapy was 19 weeks (IQR, 9-38 weeks). The median CD4 cell count at stop was 635 cells/mm3(IQR 355-832). The mean follow-up after disontinuation of therapy was 50 weeks (range 20 to 119 weeks). Mean CD4 decay was 16+4 cells/month. The slope of the CD4 decay inversely correlated with the magnitude of change of CD4 on therapy (r = -0.527, p = 0.008) but did not correlate with nadir CD4 cell count, aseline viral load, type of potent antiretroviral regimen (PI based vs others), CD4 at the time of stop or gender. No patient developed an AIDS-defining event during follow-up. Two patients restarted therapy and reached an undetectable viral load again. Prolonged (but not repetitive or periodic) discontinuation of therapy while fully suppressed was clinically safe in this cohort with nearly normal CD4 counts prior to treatment interruption.

Additional Report From Jules Levin

Tebas says the predicted average time to reach cd4 of 250 in this group of patients is 24 months (again, cd4 was 635 when patients stopped therapy). Patients who gained more cd4s on therapy tended to lose them faster.

Comments from Jules Levin: Tebas concludes that discontinuation of therapy was safe but patients were followed only for 50 weeks on average ( 20-119 weeks). If there are long-term adverse effects this may not be enough time to see them. This study did not look at people who conducted repeated interruptions. The patients in this study did one prolonged interruption. Since Tebas reported on only two patients who restarted therapy, I'd like to see longer term follow-up on these patients. I still discourage interruptions & remained concerned about resistance emerging).

In poster 361, Andrew Phillips reports that after an interruption CD4 count remained above pre-HAART baseline for over 24 weeks in a substantial proportion of patients studied. In the program abstract he reports on 68 patients from Royal Free Clinic in London who started HAART (³3 or more drugs, PI, NNRTI or abacavir) with median cd4 of 157. And interrupted therapy for an average of 24 weeks or more. Median viral load when starting HAART was 5.3 log (200,000 copies/ml) and when they interrupted therapy viral load was 3.95 log (9000 copies/ml). They were on HAART for median 44 weeks and median cd4 count was 350 when they interrupted HAART. This must have been a retrospective analysis, and there was a 24% cumulative probability of returning to HAART by 12 weeks, 46% by 24 weeks, and 60% by 48 weeks after interruption. Phillips reported the median off-therapy cd4 count (last value before re-starting or before week 24--whichever was first) was 210 (the median reduction during interruption was 121), a value taken taken a mean of 15 weeks after interruption. After interruption, viral load increased to a median of 5.4 log (250,000 copies/ml) which was a median rise of 1.21 log (16,000 copies/ml). So, on average patients had 157 cd4s before HAART and 210 at interruption.

The Kaplan-Meier estimate for the median time for cd4 to declineback to pre-HAART level was 24 weeks. Therapy was more likely to be restarted in those patients with lower cd4s. Phillips said the rate of cd4 decline varies from patient to patient and rapid declines occur frequently.

(Comments: we have seen in studies that interruptions where cd4s decline below 200 can result in opportunistic infections. Regardless of cd4 count viral load increases).

Loss of T-cell Functionality During Therapy Interruption
Van Lunzen reported in poster 363 that t-cell functionality declined after an interruption. He studies 15 patients from 2 studies who stopped HAART ± IL-2 or hydroxyurea after viral load was <25 copies/ml for 13-23 months. T-cells were normal and there was little or no productive HIV replication in lympf tissue and viral load was below detection in the CSF. He measured t-cell activation & proliferation. Additional tests evaluating t-cell functionality were performed. Tests were performed to detect HIV-specific CTL in 7 patients. TREC analysis was performed to detectrecent thymic immigrants. All patients returned to baseline viral load by 4-8 weeks. CD4 & cd8s appeared to be more activated which is a sign of viral activity. T cell responses to antigens in vitro decreased, indicating recurrence of functional defects. The CD4/CD8 ratio inverted or declined after stopping therapy due to rapid t-cell activation and death. Van Lunzen concludes that functional t-cell defects reappear despite previous supressed viral replication in chronic HIV. IL-2 nor HU seemed to have postive effects in staving off the functional t-cell declines. I would add that the actual clinical effects on a person is unknown. In other words, do these loss of functionality really affect the person. We don't know but of course this effect should be considered when thinking about taking therapy interruption.