February 6, 2008
Listen to Audio (12 min.)
Please note: These files can be quite large. Allow some time for them to download.
There's nothing like hearing the results of studies directly from those who actually conducted the research. It is these women and men who are transforming HIV treatment and care. In this interview, you'll meet one of these impressive HIV researchers and read an explanation of the study he is presenting at CROI 2008. The interview was conducted by Gerald Pierone, M.D., an HIV clinician/researcher and the founder and executive director of the AIDS Research and Treatment Center of the Treasure Coast in Fort Pierce, Fla.
I'm Brian Porter. I'm from the National Institutes of Health [NIH]. I work with a group at the NIH that's been studying IL-2 [interleukin-2] for about two decades now. It seems like the people in the IL-2 field are really dedicated to it and have been with it for a long time.
Brian Porter, M.D., Ph.D., M.P.H.
You probably know that IL-2 has been looked at for a lot of different indications in HIV. We're all familiar with the concerns about treatment interruption that many large studies have shown us -- SMART,1 STACCATO,2,3 etc.-- that there are problems with that. One question that we had was: Could IL-2 serve as a HAART-sparing therapy? In other words, during periods of treatment interruption, if those subjects received IL-2 as needed -- based on CD4+ count -- could they maintain their CD4+ count at the end of a six-month period compared to [their CD4+ count at] baseline?
Essentially, what we did is look at a group of 41 patients who had all received IL-2 in the past as part of previous therapy.4 We gave them a baseline IL-2 cycle while they were relatively immunocompetent, with good T-cell counts around 950 and, for the most part, suppressed viral loads less than 50. From that point on, we randomized them to either stop their HAART or to continue HAART for six months. Any of those subjects could get additional IL-2 cycles -- again, as needed, for CD4+ counts falling below a threshold of 500 cells per microliter. At the end of that six-month period, we asked ourselves, "Did those patients maintain their CD4+ counts, and were they considered treatment successes because they had CD4+ counts that were at least 90% of what their baseline counts were?"
Before you go into the results, do you know why the decision was made to include people that had previously received IL-2?
That's a great question. A lot of that was because, again, the subjects were drawn from other IL-2 studies that had already been ongoing at the NIH. Most of these patients had participated actively in NIH protocols using IL-2 as a supplement to their HAART therapy already.
The two groups were randomized. On average, the subjects had [received] a median of about nine IL-2 cycles, but with a relatively wide range. Some had as few as four. Others had as many as 26; I think that was our highest. Our criteria was that we wanted them to all have at least three cycles of IL-2 as a way of normalizing that. We actually did bring in some people from outside of the NIH cohort if we could establish that they had received IL-2 previously.
The best way to summarize the answer to [your] question would be that we were also looking for the impact of IL-2 with continuous HAART on a relatively immunocompetent person. Giving the immune boosting [that can occur] with the IL-2, could that protect you during treatment interruption?
What did you find?
The primary endpoint really showed that the IL-2 supplementation could not be considered to be non-inferior to just continuing HAART. In other words ... what it amounts to is they couldn't maintain their CD4+ cell counts at six months compared to baseline at the same level as those patients who simply continued HAART. What you can see from Table 2 [of my poster] is that with the treatment interruption group, eight of the subjects needed to get an additional IL-2 cycle because their T-cell counts were waning while they were undergoing treatment interruption. [That is] compared to the continuous group, none of whom needed any IL-2 while they were still continuing HAART.
That was one difference there. You can see a treatment success rate of just under 50% for the interruption group and almost 100% -- 92%, or basically everyone except for one subject -- in the continuous group.
Can you tell me how treatment success was defined in the study?
Treatment success on an individual basis is having a month-six CD4+ cell count of at least 90% of what your baseline count was. We did allow people to fall a little bit -- within 10% -- but we didn't want their T-cell count to fall below 10% [of their baseline level]. Remember that these were relatively healthy patients -- that's the best way to characterize them.
It sounds like it's a fairly stringent threshold for defining treatment failure: 10% [drop in CD4+ count] for people starting with a CD4+ count of almost 1,000.
But again, I think if you're putting it in the context of treatment interruption, we want to make sure that our interruptions are not in any way harming our patients. That's why we were going to call them a success if they could maintain [a CD4+] count at least 90% [of baseline]. Remember that we're starting at baseline with relatively high counts. The two groups are randomized, and they both started with median [CD4+ cell counts] of about 950 T cells at baseline.
Again, they had high T-cell counts and suppressed viral loads, for the most part, at baseline. When you look at the actual changes over time, the treatment interruption group fell by about 90 [CD4+] cells per microliter in terms of the delta -- the change -- whereas the continuous group, who continued on their HAART, actually went up even further, by 400 [CD4+] cells at month six. And again, that's probably due to that initial baseline IL-2 cycle that they had at the start of the study.
That was really the primary endpoint, but there were a couple of very interesting secondary endpoints in the study as well. One is that at the end of the six-month period, not only could the treatment interruption group continue to interrupt for another six months -- so basically have a one-year interruption, which all of the 27 in that group decided to do -- but the continuous HAART group had the option of crossing over into [the] treatment interruption [group]. Now we could compare what the difference is between early treatment interrupters, who interrupted at the start of the study after that first baseline [IL-2] cycle, versus late treatment interrupters, who basically had, after the baseline [IL-2] cycle, six months of HAART, and then interrupted.
What was interesting about that is, treatment success-wise, [there was] no difference. Those late interrupters also only had about a 50% treatment success rate. What did differ is that their CD4+ count seemed to decline at an even more rapid rate [than the early interrupters]. This is what Figure 3 [of my poster] is showing: the slopes of CD4+ count changes for the two groups. The late treatment interrupters, after stopping HAART, seemed to decline even faster than the early treatment interrupters, and ultimately had the same rate of treatment success or failure, about 50%.
To put this into perspective: Someone in your study could be defined as a treatment failure if their CD4+ count, starting from a level of 1,000, went down to 900.
That's correct. That is exactly right. Because we were looking for no more than a 10% decline.
I think in clinical practice, many times, patients with a very high CD4+ count will vary from month to month by over 100 CD4+ cells. I think out in clinical practice, that would not be considered a clinical failure, it would be considered more of a natural variation in CD4+ levels from time to time.
That's really a good point. I couldn't tell you [offhand] the standard deviation in terms of the month-to-month variability [of a patient's CD4+ count]. Definitely, though, you could look at the two groups as a whole, and look at their median [CD4+] counts, and really see a trend. You could definitely see a trend [in my poster]. But I think that is a good point, that there definitely is month-to-month variation.
Again, I think the reason for having somewhat of a strict criteria is because we knew that we were doing an intervention with treatment interruptions, so you want to make sure you're adhering to that "do no harm" standpoint.
Also, the group on continuous HAART received, in effect, an IL-2 boost right at the initiation of the study, so they were given even more of a buffer, perhaps.
That's exactly right. One other point: We all know the concerns about treatment interruption that we have mentioned, but the one that we were really trying to address in this study was whether or not that interruption period would give you a benefit in terms of metabolic indices. Even though it's not reflected as much in the data presented here, we did a really extensive metabolic analysis of these two groups at baseline and month six, including both external and internal measures of body fat. The ones we're familiar with -- lymph thicknesses, skin fold thicknesses -- but also internal measures: visceral and subcutaneous fat measures by CT scan, and then a whole host of biochemical parameters -- liver tests, thyroid tests, lipids, etc.
Out of all those parameters, about 50 different parameters, the only ones that really showed a significant difference at month six between the two groups were shown here [on the poster]: total cholesterol, apolipoprotein A-I (ApoA-I) and hemoglobin A1c.
If you notice the P values for the first two, cholesterol and ApoA-I, they're less than .05. But if you do a Bonferroni correction <> for the number of comparisons we did, honestly I'm not that impressed by those [P values]. I really think that those were just randomly found to be significant, because those P levels really are too high. A good threshold P level would be .0005 or less, which hemoglobin A1c actually did achieve. Then again, is there are any clinical relevance to those absolute values?>
Just to clarify, you saw that the hemoglobin A1c levels in patients with treatment interruption were --
Slightly higher, and at a significant level compared to the continuous treatment group. But again, I'm not so concerned about the clinical impact of that, because we're talking 5.4% [in the treatment interruption group] versus roughly 5% in continuous [treatment]. Good hemoglobin A1c is below 6%.
So the difference was statistically significant, but --
Exactly, but maybe not clinically relevant. What's interesting is the fact that a six-month treatment interruption -- and these were all adequately powered at 80%, using the sample size for the metabolic indices -- had just a minimal impact on those metabolic indices between the two groups.
Any side effects from IL-2?
We saw some adverse effects, which are familiar from all of our IL-2 dosing studies.5,6,7 These are basically Grade 3 events: headache, fatigue, some abdominal pain. Not in many patients. What people have normally seen [is that patients] will often get Tylenol or ibuprofen and they'll be able to get through those side effects. Nothing that made anyone stop their IL-2 cycle prematurely.
The side effects were manageable.
Again, these were patients preselected for ability to tolerate IL-2.
Exactly right, that's a great point. There was one Grade 4 adverse event, a [case of] hyperbilirubinemia, but we think that this was consistent with this patient having pre-existing Gilbert's syndrome -- in other words, problems with processing unconjugated bilirubin.
I will point out that there were four clinical events that were significant in the study: acute retroviral syndrome, syphilis, Kaposi's [sarcoma] and non-Hodgkin's lymphoma. These were all in patients who underwent treatment interruption -- either during the study itself, as part of the protocol, or after they completed the study, during their enrollment in the study. We still followed the patients to see if our study could have impacted them in the future. We saw that there were four significant clinical events, but there didn't seem to be any relationship [in terms of] which group they were from.
Going forward, the take-home message for clinicians is: IL-2 in this context is not useful.
I think staying in this context is really important. With this particular context, it couldn't maintain your T-cell count during treatment interruption. Ongoing studies are looking at IL-2 being used before clinical indications for HAART in more immunocompetent individuals, [or] even using it in early HIV infection. I don't know if we could make the blanket statement that [IL-2] could never be a HAART-sparing therapy, but in this context I feel that we have pretty convincing data that it wasn't able to do that.
To add to that point, there's another poster at this meeting, using IL-2 to perhaps delay an initiation of HAART.8
OK, very good. Thank you.
This transcript has been lightly edited for clarity.
|Please note: Knowledge about HIV changes rapidly. Note the date of this summary's publication, and before treating patients or employing any therapies described in these materials, verify all information independently. If you are a patient, please consult a doctor or other medical professional before acting on any of the information presented in this summary. For a complete listing of our most recent conference coverage, click here.|