July 11, 2002
So -- we can stop. Should we?
This series of presentations, followed by a series of summary lectures, presented the newest information available about the different approaches to treatment interruption. This work continues on in three different circumstances, with differing goals:
There were data presented relevant to all three circumstances.
Very recent infection ("acute"): this approach was first identified through work done by Drs. Franco Lori and Bruce Walker, who showed that treating extremely early -- just after initial infection -- and then doing one or more treatment interruptions, can result in some patients maintaining low viral loads (usually defined as less than 5,000 copies) for months or even years when off antivirals.
The first presentation (ThOrB1437) in this symposium session, however, pointed out that the window of opportunity for intervention and for a potentially dramatic outcome may be small. In this study, 12 people were treated about nine weeks after they were diagnosed with "primary" or acute HIV infection. They all had evidence of seroconversion as defined by positive antibodies by Western Blot HIV testing.
Here the strategy was to treat for as long as two years in order to suppress HIV, and then interrupt treatment repeatedly for periods of several weeks over the next year, restarting antiviral treatment after several weeks off treatment. There was some evidence of improvement in the immune response to HIV -- in that the peak viral load off treatment did decrease somewhat over time. However, only 33 percent were defined as responders, meaning they had a viral load less than 5,000 copies off antivirals one year after this protocol.
This is similar to other studies done in chronic HIV infection, with 20 to 30 percent responding to this type of intervention. This contrasts with the more optimistic data, presented by Dr. Walker, who suggested the rates he sees when starting treatment after infection but before seroconversion on antibody testing is closer to 80 percent success in terms of control of viremia off treatment.
Dr. Markowitz, in a debate session (ThOrB260), also discussed concerns raised about treatment interruption done for people treated soon after infection, with the goal of augmenting immune function. In his small study, he treated a group of people and later stopped treatment, testing whether a vaccine could delay the viral load rebound. The vaccine tested had no impact on the rate of rebound or the viral load set point after stopping. He also noted that while there were some people in the study who did have initial control of viremia after the stop, there was a loss of this control over the next year.
Specifically, he noted that 11/15 had a viral load below 5,000 copies off treatment at three months, but only 4/12 remained this low after one year. There are also general concerns that interruptions might increase the size of the latent reservoir of HIV in the body, although Dr. Walker has not found this to occur with his approach. There is also the general concern that stopping treatment leads to a return of viremia and that HIV targets HIV-specific T helper cells for destruction, so that this approach of stopping to augment immunity must be done carefully, otherwise there is the risk of losing the very immune cells one is trying to expand. There was agreement that learning how to best stop and monitor, in order to augment rather than decrease these cells is ongoing, thus leading to the general recommendation that this not be done outside of clinical trial centers where monitoring these cells can occur.
All agreed that there must be care when stopping medications that have a low "genetic barrier" to resistance -- meaning medications that need only one mutation to result in high levels of resistance. Moreover, medications that have a long "half-life" in the body are of particular concern when stopping, since these meds may remain in the body at low amounts during the interruption, and when viral load rebounds there is a risk of resistance occurring.
Thus, interrupting regimens containing either of the commonly used non-nucleosides -- nevirapine or efavirenz, or lamivudine (3TC) -- should consider substituting other agents for these while maintaining viral suppression for as long as a week prior to the interruption so as to ensure these drugs are not present when HIV rebounds. It is not clear how often resistance will happen when stopping regimens with these agents, but there were cases presented where it did occur, suggesting that a cautionary approach is needed to preserve these medications for future use when restarting.
Joel Gallant from Johns Hopkins University in Baltimore (ThOrB1439) presented an update on the second type of interruption, where people who initiated antiviral treatment for a variety of reasons were monitored. Here, the group was more mixed, but he described a group of 101 people whose stop in treatment was intended to be for a few months, with the intent of restarting once there was a CD4+ count drop or other factors forced them to start treatment again.
In this group, at the time of the analysis, about a third (33) were back on treatment. Gallant's analysis focused on which factors influenced the time someone could remain off treatment. The one factor with the largest impact was the pre-treatment CD4+ count, noting that people whose CD4+ count nadir was relatively high prior to starting treatment were able to remain off treatment longer than those who had a lower CD4+ count. In other words, those who started treatment at a high CD4+ count in the past, now higher as a result of treatment, were able to remain off treatment longer than those whose CD4+ counts were low when treatment was started. Viral load was less predictive of how long one could remain off of treatment. However, Gallant did note that there were a significant number of people who initially began treatment at CD4+ counts anywhere over 200, who did have a substantial amount of time off of antivirals before going back on.
A group in Argentina (ThOrB1440) reported on a randomized study exploring this same question. Here they stopped or continued therapy on a small group of people all of whom had initiated antiviral treatment with a CD4+ count greater than 350.
The patients in this study also had to have a pre-treatment viral load below 60,000 and had to have been on treatment for at least six months. The plan was to initiate treatment when and if a patient's CD4+ count fell below 350 on two occasions. At this point, 36 people were randomized in this study.
Like researchers in other treatment interruption trials, they noted three different patterns of viral load rebounds after stopping treatment -- some people had a peak viral load soon after stopping treatment, which then came down to a lower value after a few weeks. Others had a more gradual rise back to their pre-treatment setpoint. Still others had a lower return of viral load and it remained below their set point over time. They also noted that the stop was associated with a typical pattern of CD4+ counts, with an expected initial drop in the counts, but that after a few more weeks, some would bounce back up and reequilibrate at a higher level, sometimes near what it was at the time of the interruption. At this point, their study is ongoing, with the intent to describe how long those who have stopped treatment are able to remain off of treatment. At this point, almost all the patients who did stop treatment remain off of treatment six months later.
Mark Dybul updated us on the outcome of his research (ThOrB261). His treatment interruptions are geared for patients who started with chronic infection. Here the goal of interrupting was to have it be brief enough so as to avoid any return of viral load, and then retreat for short periods of time to reduce medication use overall while avoiding HIV rebound.
The study approach is to enroll patients who have a viral load below 50 copies on a combination of d4T, 3TC, and indinavir "boosted" by low-dose ritonavir. The plan is to stop treatment for seven days, and then restart this regimen for seven days, repeating this cycle over the next year or more. He continues to note success with the approach in that no one has had loss of viral suppression over a year later. There was a trend for lower lipid levels as well over time.
Finally, Dr. Dybul reported the initial results on a smaller group of people who did this same schedule while on the combination of ddI, 3TC, and efavirenz. Again he noted that this was successful, with no viral rebounds noted over the next several months. Of concern, however, was a contrasting report from the SSITT study, where they did note a reasonable rate of viral rebound in people who stopped as early as four days off treatment. It remains unclear why Dybul is not seeing return of viral load at this early timepoint in his studies, although there are some who do have very low level "blips" that later resuppress while on treatment.
Finally, Dr. Katlama presented her results from a study called GigaHAART. Her study focused on testing whether there was any benefit in stopping treatment prior to switching to another combination, here done for people with high viral loads despite treatment (e.g., treatment failure).
In her approach, people were randomized to either simply switch to a new regimen, or stop for about eight weeks, and then start a new combination. The new regimen was standardized, containing four NRTIs (d4T, ddI, abacavir, 3TC), a non-nucleoside, and three protease inhibitors.
At week 24, she noted a benefit to the interruption, reporting that there were twice as many people (50 percent) who had a one log viral drop on the new combination in those in the "stop and switch" arm, while only 24 percent achieved this goal in the "switch" arm, a finding that was statistically different.
In contrast, there was a report from a Spanish trial comparing this same approach of randomizing to a switch versus stopping and switching, which found absolutely no difference in outcome as a result of the interruption. It remained unclear what accounts for the different outcomes in these two studies and other research continues to sort out these contrasting results.
In sum, there is ongoing research in the different ways that treatment interruption may improve care. There is agreement that there are some initial promising results, especially with the approach taken by Dr. Walker in extremely early infection, Dr. Dybul in the brief interruption approach while maintaining suppression in chronic infection, Dr. Gallant in the "long cycle" approach -- taking time off of treatment for months, and Dr. Katlama for improving response to a "salvage" regimen as a result of interruption. These and other studies will be updated in future meetings to provide some insight in how this approach can improve on the overall goal of treating HIV infection, which is a long and healthy life whether on or off antiviral medication.