Structured Treatment Interruptions Workshop Summary
January 31, 2001
Interlude Talks - II
Aaron Diamond AIDS Research Center, New York, NY
Antigen requirements for CD8 stimulation
Rafi Ahmed, PhD
Professor Ahmed, an investigator of the dynamics of the murine antiviral immune response who works outside of the HIV field, gave some much needed guidance for dealing with the myriad combinations of TI schedules, durations and restart triggers.
"What you're trying to do here," he said, "is reset the clock." The goal is to reset the balance between the number of HIV-1-specific T and B cells a patient has when starting HAART at peak viral load, and the number of HIV-1-specific cells they have when they interrupt treatment after a period of suppression. Ideally, there will be a greater number of HIV-1-specific T and B cells at interruption that control the viral rebound "You don't want to recapitulate the primary infection or you're back at square one."
Ahmed outlined some rules for performing effective prime/boost vaccinations.
A strong booster is essential, and the key to a strong boost is the amount of antigen. "The greater the antigen, the greater the boost." But, Ahmed warned, "The duration of the boost must be limited. For T-cells, the ideal is unlimited antigen for a limited period of time. If you continue the antigen too long you're going to destroy what you activated."
The amount of antigen present during the boost determines the extent of memory and naive T-cell recruitment. Studies that measure the number of memory cells recruited to become effector cells show that complete recruitment only occurs at the highest antigen doses.
The duration of antigen stimulation is governed by proliferation. "When you recruit a T-cell to divide -- be it naive or memory -- it goes through the entire cell cycle in about six hours. Very fast." At these rates of division, proliferation can't continue indefinitely or you would explode! After 10 to 20 divisions, the cells are driven to apoptosis or some other form of non-functionality. For mice, the limit is 6 to 8 days.
The ideal is to stimulate cells capable of massive expansion and rapid proliferation. You don't want to stimulate activated effector cells. If you are trying to auto-vaccinate with antigen during a treatment interruption, you need to interrupt when you have memory cells, not activated effector cells. Memory CD8 T-cells are not found during acute infection but appear only after a period of rest, during chronic infection. Effector CD8 T-cells have already gone though many cycles of replication and will not continue to proliferate very much. Memory CD8 T-cells, however, will proliferate for 4 to 6 cycles, slightly better than naive cells. The transition from activated effector to memory cell happens best about 15 days after antigenic stimulation subsides.
"So clearly a period of rest is necessary for cells to re-acquire the ability to proliferate in response to antigen." Continued cellular differentiation depends on the upregulation of new and additional genes.
How would this understanding influence the design of STI schedules? If HAART is initiated during primary infection, the system is not pushed too far. HAART removes the antigenic stimulation. This allows the cells rest and lets some of them to turn into memory cells. The next time therapy is stopped; the memory cell population is boosted, which helps to expand it. Some of the most successful results with STI appear to have followed this scenario.
In chronic infection, the situation is more complex. Effector cells may be retained in the presence of antigen but, if they are functionally exhausted, they may not go on to become memory cells, even if HAART subsequently reduces antigen. For a person with chronic infection on a first interruption, the initial response may be a naive cell response rather than a boosted memory response. "If a first STI produces a naive response, then you are not starting off that much better than when the person was first infected. If the response improves after the second or third TI, then you may be generating some memory cells. Since you don't want to push the system too far in chronic infection, it may be best to keep the durations of TI shorter. If the TI is too long you risk over-stimulating the effector cells to the point they are not able to become memory cells. "A prolonged, chronic stimulus is the worst thing you can do if you want to generate a good memory response."
The meeting became activated as participants peppered Dr. Ahmed with questions about the implications of his data on the design and feasibility of STI protocols.
Q: In the mouse you find the peak of viral replication at 5-7 days and the T-cell response a bit later. A mouse lives about two years if it's lucky. How does that timing relate to humans? Would two days look like four weeks?
Dr. Rafi Ahmed: Look at the case of EBV in infectious mononucleosis. A tremendous CD8+ response is generated, which thereafter controls the infection for life.
Q: So one suggestion for planning STIs might be a short first interruption with longer 2nd and 3rd interruptions?
RA: Yes. For the first interruption, restart treatment after maybe 2 or 3 weeks, even if the RNA hasn't gone up that much. You may see the effect on the 2nd or 3rd STI.
Q: Is there any hope for those with chronic infection?
RA: This would be more difficult because the pool of memory cells is going to be lower. There could also be destruction of the dendritic cells, a lack of costimulation. Have you considered pulsing them with DCs during the STI?
Q: Would it be wise to select CHI patients who have regenerated naive cells during HAART in the hope that you could stimulate new memory cells during an STI?
RA: Well, is the repertoire the same as it was before infection or have you been left with holes in the repertoire? In animals we've found that the most effective effectors are deleted during chronic stimulation.
Q: It seems that those with a broader immune reconstitution would have an improved ability to restimulate. But, people have been on effective suppression for varying periods of time. Your work has shown that LCMV memory cells can persist for a year (which is half the life of a mouse), but is the memory component stable?
RA: If you look at antigen-specific cells in a vaccine setting they remain stable.
Q: So would the duration of effective HAART prior to entry into the protocol be relevant? Is a shorter duration of HAART better?
Q: The rest period is important for expanding the memory subset. But coming into a study with the effector cells engaged is also important. Would the best strategy be to do small interruptions at the start followed by longer ones?
RA: If the effector cells are too activated you won't be able to expand them.
Q: Could you use cytokines to stimulate memory cell production?
RA: If you have the same frequency for a given epitope within naive and memory cells, you'll have more memory cells after cytokine stimulation because memory cells out-compete naive cells, which take longer to get started. We don't have a memory cell-specific cytokine at this time.
This article was provided by Project Inform. Visit Project Inform's website to find out more about their activities, publications and services.