From 30 July to 1 August 1999, a diverse, international group of biomedical researchers, statisticians, clinicians, research administrators and community treatment advocates met to discuss and develop plans for research on structured treatment interruptions (STIs) in the context of highly active antiretroviral therapy (HAART). Participants reviewed observations to date, currently available virologic, immunologic and clinical hypotheses, and reviewed studies now underway or in the planning stages. They evaluated STI research in the context of fully-virally suppressed patients with primary or chronic HIV infection, as well as multi-drug resistant (MDR) patients who are failing to achieve full viral suppression. In a series of intradisciplinary and interdisciplinary breakout-groups, participants identified gaps in current STI research and developed several proposals and mechanisms to address these gaps, coordinate and expedite the overall STI research effort. Among the conclusions and follow-up steps to emerge by consensus from the STI Workshop were the following:
- The need to establish a prospective observational STI cohort study to pool observations regarding patients who elect to undergo an STI and to assess its safety, efficacy, and virologic, immunologic and clinical impact;
- The need to establish an STI Laboratory Working Group to pool resources and improve the ability of researchers to take advantage of new virologic, immunologic and pathologic assays;
- The need to develop and promulgate a carefully-worded and thought-out clinical practice guideline outlining the potential risks & benefits, and the knowns and (far more) the unknowns about undertaking an STI at various stages of HIV disease;
- The need to address pharmacologic and quality-of-life considerations in STI research; and
- The need to coordinate STI research, particularly vis-à-vis studies of STIs in heavily pre-treated patients with few treatment options and (possibly) low CD4 T cell counts, or those at risk for a CD4 T cell plunge or clinical progression during an STI.
Workshop participants discussed these and other objectives. The following draft report summarizes the workshop proceedings and the discussions that led to the workshop conclusions and follow-up recommendations. The STI Steering Committee will undertake to facilitate their implementation.
The Structured Treatment Interruptions (STI) Workshop was held on July 30-August 1, 1999 at the Boston Marriott Newton Hotel. This workshop was co-sponsored by the Foundation for AIDS & Immune Research (FAIR), Project Inform and Treatment Action Group (TAG).
- Some individuals who discontinue highly active antiretroviral therapy (HAART) continue to maintain viral suppression during an STI.
- In some cases, prolonged suppression of plasma viral load to beneath 500 copies/ml in the absence of drug therapy has been observed-is this related to start/stop therapy?
- There are observational indications that the time to rebound of viral load increases in some people with each subsequent start/stop STI cycle.
Researchers from the Aaron Diamond AIDS Research Center (ADARC) have shown that slower decay rates in the latently infected cell compartment (L cell) occur in people who have intermittent episodes of plasma viremia (>50 but <500 copies/ml) while on HAART. Those with more than two episodes of intermittent viremia per year had non-decaying L cell slopes.Treatment interruptions have been studied in several settings:
Primary HIV infection
Chronic HIV infection with full suppression
- Comet study (Results reported quick viral rebound, but retreatment was successful, with no evidence of the development of resistance.)
- NoHRT (Rich Davey/NIAID; 22 treated; HAART + IL-2, STIs, different patterns)
- Gatell/Garcia (Barcelona, one or two STIs, all patients. below 20 copies/ml, viral load went up to 200 copies/ml, re-initiation, then 2nd STI. Evidence of immune activity in some patients; some interesting viral load rebound curves -- evidence of containment of viral load by immune system)
- Phillips (AIDS, compared rebound in STIs; patients with low CD4 T cell cells had double rebound kinetics vs. those with high CD4 T cell counts - 5 in each group)
- Several additional reports made since the STI meeting, including a case of 3 patients in Brussels who discontinued therapy with no (or low) rebound of virus.
Late salvage/MDR patients
- Frankfurt HIV Cohort (Miller et al.)
- Royal Free Hospital, London (Youle et al.)
These patients were later challenged with mega-HAART regimens. Patients who had experienced an STI seemed to have an independently elevated chance of going beneath the limit of quantitation when resuming therapy.
There were important differences in the Frankfurt patients between those whose virus shifted to wild-type and those whose virus did not shift (see September '99 TAGline). Those who experienced a treatment interruption and had a viral shift back towards wild-type had a five-fold greater chance of having their viral load go beneath the limit of quantitation (500 copies/ml) than those who did not revert to wild-type during the interruption.
However, there is a major unresolved risk/benefit equation in the disconnect between rising viral load and persistently elevated CD4 T cell counts in patients experiencing partial suppression. The immunologic benefit may persist, but at the cost of the further evolution of multi-drug resistance. The disconnect between viral load and CD4 T cells in partial suppression-disconnect vs. further evolution of multi-drug resistant virus. How far will those viruses evolve?
Why is viral replication contained in some individuals during an STI? Several observational studies have observed temporary containment of HIV replication during treatment interruptions. The Berlin patient's HIV-specific CD4 T cell response increased despite a lack of significant viral activity.
Why is viral load contained (for variable-to-indefinite periods) in some individuals who interrupt treatment?
- It could be stochastic (due to random variation).
- The individuals could actually be long-term non-progressors.
- It could be due to immunologically mediated suppression, change in virologic type, both or other causes.
- It could reflect direct HIV-specific immune responses or other immune responses.
- There may be other unexplained factors.
Primary HIV infection (PHI)
We know that in most PHI patients there is a quick rebound in virus levels, but in some individuals there is a delayed rebound. Usually the subsequent response to therapy has been good with no evidence of drug resistance.
Chronic HIV infection (CHI), suppressed viral load
In the untreated state, the HIV/CD4 T cell interaction typically leads to advanced stage disease ("AIDS"). When intervening with HAART, suppressing HIV replication reduces the HIV/CD4 interaction and reverses disease. When measurable plasma viremia returns, disease progression eventually resumes (although possibly in different forms). So the question of stopping therapy immediately raises the danger that the patient will experience an immediate recurrence of the HIV/CD4 interaction that could once again lead to progression. It is unclear whether the timing and pace of progression, however, follow the same patterns as seen in natural history data.
What causes the rapid declines in CD4 T cells observed in some people during an STI? Is it T cell destruction or redistribution? How functional are those cells? Why is this drop (perhaps less often) seen in people who never achieved a large gain in CD4 T cells from HAART?
- Will viral replication after STI stimulate an antiviral immune response that can keep viral replication in check?
- How can the immune response be broadened assuming that broader equals better?
- In chronically-infected patients who are suppressed, can re-exposure to the virus after an STI lead to a stronger immune response to HIV?
- In salvage patients who are not suppressed, can an STI convert the drug-resistant virus into a drug-responsive one?
Do individuals, particularly those in late-stage disease, lack the residual capacity to mount an immune response against HIV, possibly due to insufficient CD4 or CD8 T cells, defects in antigen presenting cells (APCs), defective microenvironments? Can anything be done to help?
Proposed STI Studies
Several different trials were proposed and would involve different entry criteria:
- Observational studies -- to gather information regarding the incidence, safety, virologic, immunologic and clinical impact of STIs occurring due to decisions of people with HIV and/or their clinicians;
- Safety trials -- open to all comers, regardless of the duration of suppression;
- Studies in primary HIV infection and chronic HIV infection, suppressed virologically -- enter if viral load was suppressed for at least one year;
- Studies in multi-drug resistant/salvage patients-comparing various strategies to encourage the virus to revert to wild-type before undertaking a mega-HAART or salvage regimen;
- Quality-of-life studies in people experiencing poor quality of life as a result of ART or experiencing disease progression.
Study design considerations
- Randomization -- will patients be willing to be randomized to such studies?
- In patients who stopped treatment, would they be willing to go back on treatment?
- Care has to be taken to ensure that OI prophylaxis or maintenance therapy is resumed, as needed, prior to study randomization.
- Individual CD4 T cell nadirs should be considered in the study design
- Durability -- Short term responses will not give us adequate answers; we need long-term follow-up.
- Access to medications. When a patient has discontinued, to re-get the medications in some states could be a problem... Could you get that PI again if you go to the bottom of the list?
- It's hard to plan trials today with the development of new, easier drugs and regimens and immune-based therapies.
Studies in People Chronically Infected/Suppressed
There is a need for two studies, one for those with viral load below 50 copies/ml (profoundly suppressed) and those with 50-5,000 copies/ml (somewhat suppressed). One study design would randomize individuals to continual vs. intermittent HAART or HAART with one (or more) STIs. Therapy would be resumed in people in the STI group when the CD4 T cell count dropped below 300/mm3. The endpoints would be:
- The proportion of people who responded to HAART after re-initiation of therapy, or
- Both viral load declines and CD4 T cell rebounds post-re-challenge.
The studies should be stratified by pre-treatment and baseline viral load. The studies also need to incorporate an evaluation of the impact of hydroxyurea, which blunts CD4 T cell responses. There needs to be an evaluation of whether the studies are doing patients harm.
Studies for Chronically Infected/Unsuppressed (viral load detectable)
There are profound differences between people with unsuppressed viral load and high CD4 counts and those with similar viral load but falling CD4 counts, for whom the drugs might be doing no good at all. While return to wild type virus may increase the chances of future response to therapy, if it fails to result in renewed ability to suppress virus, it may do harm. Wild type is the pathogenic virus which originally led to immune depletion and the decision to initiate therapy. It's quite possible that the drug-resistant, but possibly less fit, virus is less pathogenic and hence more desirable. Some researchers would prefer to understand pathogenesis better in this subgroup before generating new hypotheses.
An important control arm in the heavily pre-treated population with few treatment options might be people who are continued on partially suppressive regimens. Data suggest that partial viral suppression still has an effects on prolonging health and life. Their results could be compared to those of people who change to a new, more aggressive mega-HAART regimen either with or without an intervening STI:
- Continue on partially-suppressive regimen;
- Initiate mega-HAART immediately; or
- Take an STI, then mega-HAART
It is important to develop and promulgate a statement about what is known and issues to consider for researchers, clinicians and people with HIV when considering a structured treatment interruption.
The issue of STIs for people taking drugs with long half-lives such as efavirenz and nevirapine is complicated by the fact that patients may have to stop their NNRTIs before stopping their nucleoside analogues and/or protease inhibitors. Based on these basic principles, can all nucleosides and protease inhibitors be stopped at the same time (after all, within 24 hours all will be undetectable in plasma or near undetectable)? Or should nevirapine and efavirenz (the only two with significantly longer half-lives) be stopped two to three days earlier than the others? Many of