In addition, we have always claimed that our strength is in our numbers. When the Consortium takes a stance on an issue and, for example, writes a letter to an elected official, it behooves us to say that we are an organization of XXX HIV care providers in the San Francisco Bay Area. Obviously, the larger the number that fills in the blank, the more clout we have. The Community Consortium has much to be proud of. We have been a model for community based clinical trials and provider education in this country for over 15 years. We intend to continue our efforts for as long as they are necessary. We do need and appreciate your support! Please take a moment now to fill out and send back the enclosed membership application if you have not already done so since July. Thanks again for your continued support!

Return to Top

Forensic AIDS Erratum

It was bound to happen eventually. After 16 years of synopsizing Community Consortium educational programs, an error was committed in the June 2001 discussion of forensic AIDS. Page 4 column 3 states that "Dr. Estes queried the audience." In fact, the query and all the following information through page 5 column 1 ending with "There are 33 prisons in California" were contributed by Guy Vandenberg, R.N., M.S.W.. Apologies to Guy and Milton for this transcription error. Hey, you try reading my handwriting!

Return to Top

The panel was moderated by Donald Abrams, M.D., who had also made the journey south. He opened the program by saying that although two of the opening night speakers focussed on the problem of HIV in the developing world, the rest of the conference was much less oriented in that direction than the Durban meeting in July 2000. He reviewed some of the frightening global statistics. In the year 2000, 3 million people died of AIDS. Many African nations have now trimmed 15 years from their life expectancies. It had taken a century for many to attain a life expectancy of 60 years. A child born in Botswana today has a life expectancy of just 40 years -- the same as it was in the time of the Roman Empire, two thousand years ago!

Donald went on to provide an overview of what he perceived as some of the dominant themes of the meeting. The "Hit Early, Hit Hard" mantra of yesteryear is now being replaced by a "less and later may be better" approach, especially with regard to resource poor settings. The jury is still out on the clinical use of structured treatment interruptions. The rationale for interleukin-2 use appears to be gaining. There are, in fact, some new drugs in the pipeline, but effective vaccines are not on the immediate horizon. He concluded that it does take two to tango as assessed from the numerous tango shows available for conference attendees.

PEP STUDIES

The panelists agreed that the conference was strong on clinical science presentations with a paucity of real basic science or pathogenesis. John Stansell presented results of an animal trial of early post exposure prophylaxis (LeGrand et al). Twelve macaques were inoculated with 50AIDS 50SHIV intravenously. They were randomized 1:1:1 to receive HAART (AZT/3TC/IDV) at 4 or 72 hours post-exposure or placebo. Treatment was given for 21 days and the animals were followed for one year. The HAART treated animals seroconverted significantly faster. There was only a mild and transient CD4+ cell decline following infection observed in the HAART treated animals. Response to recall antigens was preserved and MIP-1 beta and RANTES were reduced in the HAART treated animals. At one year, CD4+ cell counts appeared to be preserved in the treated groups. There was, however, no effect of HAART on the viral "set point."

Turning to human experience with post-exposure prophylaxis, Dr. Stansell reviewed a presentation on PEP for occupational exposure, summarizing the experience at 47 U.S. hospitals from 1996-2000 (Beltrami et al). Over this time period there were 11,784 health care worker exposures. The mean time to the start of PEP was 2 hours. Eighty percent of the exposures were percutaneous and 14% were mucosal. Of the index cases, 6% were HIV positive, 75% negative and 20% status unknown. Of those exposed to positive index cases, 63% initiated PEP and 54% completed the course. Fourteen percent of those exposed to what ultimately was found to be a negative index case and 12% of the status unknowns commenced therapy, with 3% and 21% of those completing the course. The median duration of treatment after a negative exposure was only one day. Three drugs were used in two-thirds of the PEP recipients. Seventy-five percent of the discontinuations were due to adverse drug experiences.

PI UPDATES

Steve Becker turned to some of the indinavir studies presented at the meeting. The Merck 035 data is now mature to 5 years showing persistent virologic control in the 29 individuals still being followed. Four have discontinued therapy because of nephrotoxicity. There were a number of posters on the indinavir/ritonavir combination that he summarized by stating the optimal dose combination remains unclear. There appears to be increased toxicity, especially nephrotoxicity, of the combination compared to the single agents in all of the studies presented. A study by Cahn et al suggested that virologic control was better and side effects less with indinavir q8hr compared to the RTV/IDV 100/800 mg bid combination. A pharmacokinetic evaluation of RTV/IDV 100/400 mg bid in 47 patients suggested excellent virologic control with a 2-3 fold increase in Cmin and a 2-3 fold decrease in Cmax (Katlama et al).

A study described an induction regimen of RTV/IDV 400/400 mg bid with NRTI initially followed by the dual protease alone (Workman et al). Of 32 patients initiating this program, 28 remained fully suppressed. Of the 4 breakthroughs, three had wild type virus.

Dr. Becker also reviewed data on RTV/IDV salvage regimens. In a Johns Hopkins study of 61 heavily pre-treated individuals (56% previously exposed to IDV, 56% previously exposed to RTV), 80% had to discontinue the regimen. Thirty-nine percent achieved HIV RNA < 400 copies/mL on the intent to treat analysis. A smaller trial suggested that when using phenotypic resistance information, one might consider >25-fold as the cut point for use of the RTV/IDV combination in salvage. A salvage study looking at RTV and amprenavir showed a decrease in RTV levels. He remarked that precise PK studies are necessary when devising complex multi-drug salvage regimens to fully understand the potential drug-drug interactions.

Steve next discussed results of 4 small lopinavir expanded access experiences from Spain, Argentina, Italy and the U.K.. In general, 35-60% of the heavily pre-treated individuals were driven to less than 400-500 copies with fewer patients achieving levels <50 copies/mL. Donald reviewed data from a 60 week trial comparing Kaletra to nelfinavir in 653 antiretroviral naïve subjects. Patients received 3TC/d4T and either nelfinavir 750 mg tid or Kaletra. CD4+ cell counts and HIV RNA were equivalent at baseline in both groups at 260/mm3 and 4.9 log₁₀. Treatment was well tolerated in both arms with 4% in each group discontinuing therapy because of side effects. Diarrhea was reported by 17-18% in each arm. The Kaletra arm had statistically better HIV RNA compared to the nelfinavir recipients with similar CD4+ cell increases (246 and 224, respectively). At the 60 week point, 65 of the 326 Kaletra recipients and 106 of the 327 nelfinavir recipients harbored viruses with resistance mutations, with 0 and 37% respectively demonstrating PI resistance.

NEW DRUGS IN PIPELINE

Claire Borkert tackled the task of reviewing data presented on some of the new drugs in the pipeline. Tipranavir represents a new class of non-peptidic protease inhibitor that is reportedly active against >90% of HIV-1 strains that are resistant to other PIs. It provides a potent 1.5 log₁₀ reduction in HIV RNA after 15 days of montherapy. The drug induces CYP450 3A4; hence it will decrease RTV levels when given together, but the RTV leads to markedly increased levels of tipranavir. A newer formulation (SEDDS) allows for decreased dosing and better overall tolerance. The drug is highly protein bound. A phase 2 open-label study looked at RTV/TPV in 41 patients who were naïve to NNRTIs and had failed at least two PIs. Baseline CD4+ counts were ~300/mm3, with HIV RNA ~4.5 log₁₀. At 24 weeks, 50-60% of the subjects were suppressed to <50 copies/mL on an intent-to-treat analysis. CD4+ cells increased 140-200 cells/mm3. The most common side effects were gastrointestinal -- nausea/vomiting and diarrhea, which may be related to the formulation.

Atazanavir (formerly BMS232632) is a once a day protease inhibitor with a resistance profile not quite as good as tipranavir. A virus that is resistant to all other currently available PIs will also be resistant to ATV. The main side effect with ATV appears to be hyperbilirubinemia, similar to indinavir. Lipid elevations appear to be less with ATV than other PIs. A phase 2, 24 week trial in 461 drug naïve patients compared two doses (400 and 600 mg qd) of ATV with nelfinavir on the background of d4T/3TC. Baseline CD4+ counts were 270/mm3, with HIV RNA 4.7 log₁₀. At 24 weeks, all three arms had ~40% individuals with HIV RNA < 50 copies/mL. The NFV group had a 30mg/dl increase in lipids that was not seen in the ATV groups which did, though, have increased hyperbilirubinemia compared to NFV.

Mozenavir (DMP450) is a water soluble PI that could conceivably be combined with other drugs into one pill. It is metabolized in the liver, but other PIs do not appear to affect its metabolism. Development is currently on hold in the U.S. because of QT prolongation on ECGs seen with higher doses. The safety data otherwise looks good. A phase 1/2 trial of 50 drug naïve individuals was conducted in Mexico and Europe comparing 3 doses of mozenavir (750 mg tid, 1250 mg bid, 1250 mg tid) with IDV on a background of d4T/3TC. On an intent-to-treat analysis at 48 weeks, 80% of the mozenavir patients and 70% of the IDV group had undetectable HIV RNA levels.

Emtricitabine (FTC) is a nucleoside analogue similar to 3TC, but more potent in vitro with a longer intracellular half- life. Over 900 patients have been enrolled in trials to date. Licensing is currently on hold because of data from South African trials where 36 patients developed severe hepatotoxicity. All were on nevirapine in addition to 3TC. Thirty-three of the 36 developed liver toxicity within the first two weeks of therapy. Women had twice the incidence rate of hepatotoxicity. Two patients have died as a consequence.

Drugs that are becoming more widely available at this time include tenofovir and the fusion inhibitors. Tenofovir is a nucleotide analogue active against HIV-1, HIV-2 and hepatitis B virus. It has a half-life of 17 hours, which allows for once a day dosing. A 1.5 log₁₀ decrease in HIV RNA is seen when tenofovir is given as monotherapy to drug naïve patients. The drug is active against isolates resistant to all nucleosides including those with the Q151 mutation. It is well tolerated with a good safety profile and is currently available in expanded access. It is hoped that T-20, the subcutaneously dosed fusion inhibitor, will be available in expanded access sometime in the near future. This agent can lead to a 1.5 log₁₀ reduction in individuals harboring multi-drug resistant virus. T-1249 is the next generation fusion inhibitor with activity against strains resistant to T-20.

RESISTANCE UPDATES

Steve Deeks reported that the Resistance Meeting in Arizona really had not been that scintillating, with a dearth of new data. He said much of the meeting was spent evaluating comparisons of genotyping vs phenotyping. He thought that data presented on subtype non-B virus was of particular interest. Subtype B virus constitutes only 4% of HIV worldwide, but 96% of the resistance data is describing this strain. There had been an impression that subtype B virus might be particularly sensitive to protease inhibitors compared to other subtypes. In London, many African patients with non-B subtype viruses are under treatment. A case control study by Loveday et al from the Royal Free Hospital matched the non-B patients with subtype B patients according to CD4+ and HIV RNA at baseline. At 48 weeks, 70% of the subtype B and 78% of the non-B subtype patients had achieved HIV RNA levels <400 copies/mL with comparable CD4+ cell increases (169/mm3 and 124/mm3, respectively). These results suggest that non-B subtype virus may respond comparably to the B strains.

TREATMENT INTERRUPTIONS

There were a number of presentations on structured treatment interruption, called by many different names. Although some results were quite provocative, per Dr. Stansell, none of the studies were very large. Anthony Fauci reviewed the ostensible goals of structured interrupted therapy (SIT) as being an increase in wild-type virus, "autoimmunization", and allowing the patient to spend less time on therapy with a resultant improved adherence, decreased toxicity and decreased cost. The Fauci group at the NIAID is investigating two SIT strategies: two months, one month off and seven days on, seven days off. The first strategy was investigated in 22 chronically infected patients. It led to decreased susceptibility to HAART with frequent emergence of resistance mutations, especially K103N and M184V. A negative effect on the immune system was also reported. The 7/7 strategy has been investigated in 10 patients with results given through 32 cycles for a mean 48 weeks of follow-up. No patient experienced a sustained viral rebound; one had a "blip." There were no effects seen on proviral DNA, CD4+/CD8+ pools or markers of immune activation. Significant declines in cholesterol and triglyceride levels were noted.

Continuing along the treatment interruption lines, John next reviewed data on the effects of prolonged discontinuation of successful antiretroviral therapy (Tebas et al). In this report of 72 subjects, baseline data included CD4+ 272/mm3 and HIV RNA 108,000 copies/mL. Individuals had an average of 36 weeks of undetectable HIV RNA with a median CD4+ cell count of 571/ mm3 at the time of discontinuation of therapy. After a median duration of follow-up of 45 weeks, the mean monthly CD4+ cell decline was 16 + 5 cells/mm3.

The slope of decline was correlated with CD4+ increase on therapy (r=-.28, p=0.02) and duration of undetectable virus (r=-.25, p=0.04). A regression analysis showed that only the increase on HAART was significant with no effect of the type of HAART, nadir CD4+, baseline VL, CD4+ at discontinuation of therapy or gender.

Diane Havlir summarized the risks of treatment interruptions. There is the potential for the development of resistance, especially when one factors in the differential half-lives of discontinued drugs. There is concern about reseeding of reservoirs, especially the central nervous system. The impact of interruptions on the patient's ultimate adherence to therapy is also unclear -- will it be better or worse. Finally she mentioned the potential for increased HIV transmission, both vertical and horizontal.

HIV AND HCV CO-INFECTION

Claire Borkert next reviewed presentations dealing with the issue of HIV and hepatitis C co-infection. HIV infection is estimated at 1 million in the U.S. and 40 million globally; for HCV these estimates are 40 and 40-125 million. It is felt that 10-40% of people with HIV-1 are co-infected with HCV and 10% of the people with HCV are co-infected with HIV-1. Fifty to ninety percent of injection drug users with HIV-1 are presumed co-infected with HCV. With controlled HIV disease, HCV co-infection clearance rates are 15% - similar to clearance rates with HCV alone. There really is no definitive data that HCV has a negative effect on HIV progression; however there is some evidence to suggest that the level of HCV replication may affect HIV progression. An Italian study of 416 HIV patients followed for 30 months where 50% were HCV co-infected showed no increase in progression to AIDS or CD4+ cell decline. There was more convincing evidence that progression to cirrhosis in patients with HCV is accelerated in HIV co-infection (7 years vs 23 years with HCV mono-infection). More rapid progression to cirrhosis is associated with lower CD4+ cell counts, any alcohol use and older age at time of HCV infection.

Dr. Borkert than reviewed some of the HCV treatment trials. Currently there are two large scale Peg-IFN trials in people co-infected with HIV and HCV. Results from both will be available in 2003. Soriano et al from Madrid presented results from 31 co-infected patients treated with PEG-IFN 150 mg subq weekly and ribavirin 400 mg bid. All patients were on HAART with CD4+ counts > 300/mm3 with HIV RNA < 5000 copies/mL. None of the patients had decompensated cirrhosis or active alcohol use. At baseline, all patients had transaminases > 1.5 upper limit of normal. HCV RNA was > 800,000 IU/mL in 60%. At 24 weeks, transaminases had normalized in 85% and HCV RNA was undetectable in 65%. There was no significant impact on CD4+ cell counts or HIV RNA levels. Treatment resulted in mild flu-like symptoms; only four patients stopped therapy due to adverse experiences (fever, pancreatitis, hypersensitivity). She summarized the recommendations on who should be treated. In patients with CD4+ > 200/mm3, with any HIV RNA level and any biopsy result, treatment should be offered with a goal of eradication of HCV. Patients with CD4+ cell counts in the range of 100-200/mm3 with HIV RNA < 10,000 copies/mL and any ALT should be treated with a goal of limiting HCV disease progression.

PHENOTYPING UPDATE

Dr. Deeks presented the results of ACTG 575 which compared phenotyping to standard of care in patients failing a regimen. Inclusion criteria were more than six months of antiretroviral therapy including one PI and a viral load > 400 copies/mL Two hundred fifty-six people were enrolled with a median CD4+ cell count of 277/ mm3 and a median HIV RNA of 10,000 copies/mL. Three quarters were NNRTI naïve. Most participants had been treated with IDV or NFV with 50% primary NFV failures. The primary endpoint of the study was the proportion with HIV RNA < 400 copies/mL at week 48 via intent to treat analysis. The results for the phenotyping and SOC groups were identical at both 6 and 12 months of follow-up (48% and 46%, respectively). In a post-hoc analysis, it was demonstrated that the subset of patients with baseline resistance to more than three PIs did significantly better with phenotyping compared with standard of care (50% vs 17%, p=0.02).

Dr. Deeks mentioned that for the most part patients can be salvaged with knowledge of their prior treatment history; heavily pretreated patients are most likely to benefit from phenotypic testing.

"Resistance" is defined as > 2.5 fold decrease in drug susceptibility for all drugs. Dr. Deeks suggested that perhaps each drug may, in fact, have different sensitivities. For example, he suggested that the point of decreasing response (PDR) is likely to be >1.7 fold decrease for d4T and ddI, but may be up to 4.5 for abacavir. He then presented a "Clinical Cut-Offs" consensus chart summarizing the above information and suggesting that > 4 may be the PDR for NNRTIs and > 10 may be appropriate for Kaletra and RTV/IDV.

INTERLEUKIN-2

John Stansell reviewed some of the IL-2 presentations. Katlama et al reported on 72 HAART treated patients with CD4+ cell counts < 200/mm3 and HIV RNA < 1000 copies/mL on therapy for more than six months, randomized to add IL-2 vs placebo to their regimen. IL-2 was dosed at 4.5 MIU bid for five days for 4 cycles every 6 weeks. After 24 weeks, the placebo recipients were crossed over to active drug. At 24 weeks, the IL-2 treated cohort had a highly statistically significant 65 cell/mm3 increase in their CD4+ counts compared to an 18 cell/mm3 increase in the placebo group. Baseline CD4+ was 149/mm3 in the 36 IL-2 recipients and 138/mm3 in the 36 placebo recipients with 81% and 33% respectively rising to counts over 200/mm3 at week 24 (p<0.0001). On these trajectories, the time to reach 250 cells was 6 months in the IL-2 group compared to 3 years in the placebo recipients; the time to 400 cells, one year and 5-6 years respectively. There was no information on clinical endpoints as the trial was not conducted for long enough to reach them. Typical flu-like IL-2 symptoms were seen in the majority of patients. There was a significant increase in both naïve and memory lymphocyte populations. No effect on proliferative responses to recall antigens was seen, nor was there any effect on HIV RNA levels. The French investigators believe that IL-2 increases thymic production of CD4+ lymphocytes where as the Lane group at the NIAID demonstrates both peripheral expansion of the CD4+ cell pool as well as prolonged survival of CD4+ cells in IL-2 treated patients.

Wu et al presented results of quality of life evaluations in an ACTG IL-2 trial. In this study patients received HAART alone (51), HAART plus subcutaneous IL-2 (54) or HAART and continuous infusion IL-2 (53). Quality of life measurements were obtained at weeks 12, 16, 28 and 52. At the end of one year, the group receiving HAART and subcutaneous IL-2 had a significant increase in their QOL measures compared to the two other cohorts (p<0.05).

TDM

Dr. Becker concluded the evening's presentation by discussing the results of the ATHENA therapeutic drug monitoring trial. This was a trial where patients who were naïve and commencing therapy with IDV or NFV containing HAART regimens were randomized to receive TDM or routine follow-up. Participants were evaluated after 12 months. There were 55 IDV patients, of whom 28 received TDM, and 92 NFV bid patients. In the IDV cohort, 50% of the controls discontinued IDV compared with 25% of the TDM group because of patient request or toxicity. In the NFV group the corresponding numbers were 35% vs 12%, largely due to virologic failure (18% vs 2%). Seventy-eight percent of the TDM group had HIV RNA < 500 copies at 12 months compared to 55% of the SOC group. Steve pointed out that there are 21 laboratories that do TDM of antiretroviral drugs in Europe compared to only three in the United States. This is because Europeans have more readily accepted the test's utility, where we are demanding controlled clinical trials in our evidence-based medicine climate. It is likely that we will hear more about TDM at upcoming meetings.

THANKS

The Consortium is grateful for the unprecedented collaborative efforts and support of: Abbott Laboratories, Agouron Pharmaceuticals, Boeringer Ingelheim, Bristol-Myers Squibb, DuPont Pharmaceuticals, Gilead Sciences, GlaxoSmithKline, Merck & Co., Ortho Biotech, Roche, Serono, Vertex Pharmaceuticals and ViroLogic, Inc. for making the evening's educational dinner meeting possible! Thanks to you all!!!

In addition, a meeting of Community Consortium clinician/investigators who may be interested in participating in the upcoming CPCRA SMART trial is scheduled for October 17th. Presentations will be made by Fred Gordin, M.D., from the Washington D.C. VA Medical Center, chair of the CPCRA Executive Committee; Jim Neaton, Ph.D., from the CPCRA Statistical Center at the University of Minnesota and Cal Cohen, M.D., from the Community Research Initiative of New England and member of the SMART protocol team. The SMART trial hopes to enroll 6000 participants and follow them over 6 to 8 years to assess two different treatment strategies in HIV infection -- a Drug Conservation (DC) strategy and a Viral Suppression (VS) strategy. The study will be open to any individual with CD4+ count > 350/mm³, regardless of their antiretroviral therapy status. Subjects randomized to the DC strategy will be asked to discontinue their antiretroviral regimen until such time that their CD4+ count falls to < 250 cells/mm3. Those naïve to treatment will be asked to remain so until reaching that CD4+ threshold. Individuals randomized to the VS arm will continue on or begin antiretroviral therapy with a goal of maintaining viral suppression as low as possible for as long as possible. The SMART study will be piloted by the current CPCRA and additional U.S. and international sites to assess the feasibility of the study and will then expand further to include additional participating sites. The goal is to launch the trial at the October 17th dinner meeting. We will be inviting our current clinician/investigators to the dinner, where space is limited. If you are interested in learning more about SMART and would like to attend the dinner, please call Carroll Child, R.N., Community Consortium Research Director, at 415-476-9554, ext 22.

The University of California Center for Medicinal Cannabis Research (CMCR) has approved its second set of proposals for the study of cannabis as treatment for specific medical conditions. Eight additional studies from various California research institutions were recommended for funding by the CMCR's independent Scientific Review Board following rigorous scientific review. A study submitted by Donald Abrams and collaborators at the UCSF Pain Clinical Research Center, to evaluate the effect of marijuana on pain from bone metastases in patients with breast and prostate cancer, is undergoing application to the numerous regulatory bodies at this time. The study will be similar to the HIV peripheral neuropathy study that we hope to begin this summer. Both pilot trials will be short General Clinical Research Center inpatient studies. Additional information will be available as the study gets closer to launch.

The EAB also noted the construction going on at the Consortium offices at 18th and Folsom. Pioneers in our building, the Consortium is being joined by the remainder of the Positive Health Program clinical research programs at our site. The Adult AIDS Clinical Trials Group (AACTG) moved to the facility over a year ago. Now the remaining clinical research investigation staffs are joining us, moving from cramped quarters in various locations on the San Francisco General Hospital campus to consolidate our resources and efforts at what is being affectionately called "PHP West". This is a time of change for the Consortium staff as well as our physical plant. Bear with us if you are trying to call or visit during this transition.

The Community Consortium is now officially one of the National Trial Coordinating Centers (NTCC) of the ESPRIT trial! The Evaluation of Subcutaneous Proleukin in a Randomized International Trial is very similar to CPCRA 059, the evaluation of two different dosages of interleukin-2. However, in ESPRIT, participants will only be randomized to receive either interleukin-2 7.5 MIU subcutaneously twice a day for 5 days every 8 weeks or no treatment. As well, instead of monthly visits as in the CPCRA study, ESPRIT data collection only occurs every 4 months. The study is designed to assess clinical benefit of IL-2 and hence will enroll 4000 patients worldwide and follow them for disease progression events for a minimum of five years. Participants randomized to the IL-2 arm will repeat cycles of therapy to maintain their CD4+ cell counts at twice baseline or above 1000 cells/mm3. If you have patients who may be interested in ESPRIT, please call Paula Pell, R.N., at 415-476-9554, ext 24.

This study of a STI in patients with multi-drug resistant (MDR) HIV is a randomized clinical endpoint study to determine whether a prescribed four-month STI will delay clinical disease progression and death compared with a strategy of immediately initiating a new antiretroviral regimen in patients with MDR virus. Subjects will initiate the salvage regimen based on information generated from both genotypic and phenotypic resistance testing that is provided as part of the protocol. As there is little existing information on the optimal amount of time to wait before resuming therapy following an STI, the first 150 participants in the CPCRA trial have had point mutation assays performed. These are more sensitive than routine resistance assays. These initial patients will also undergo more frequent HIV RNA testing, which should enable the investigators to know early in the course of the trial whether four months is too long or too short of an STI.

Four-hundred-eighty subjects will be randomized in the trial over a 12-month accrual period. Participants will be followed monthly for the first 8 months, and then every 4 months. Should CD4 cell counts drop below threshold levels, prophylaxis for opportunistic infections will be resumed. This study will provide much information on the kinetics or the return of CD4 cells following the reinstatement of an antiviral regimen, as well as much clinical information over its planned two years of follow-up. The eligibility criteria have recently been modified in Version 2.0 of the protocol, which has recently been approved. Entry criteria now includes age > 13 years with HIV RNA >5000 copies/ml.