September 30, 2006
|Listen (8.7MB MP3, 25 min.)|
Calvin J. Cohen, M.D., M.S., Community Research Initiative of New England and Harvard Medical School, Boston, Mass.
So, why would we even want to treat immediately? One of the main reasons is that it's been pretty well documented that the initial onslaught of HIV in the body creates this profound immune destruction, particularly noted in gut lymph nodes. As a result of this loss of lymph tissue, and [because of] a sense that if we intervene really early we could prevent this massive onslaught and loss, there's been interest in that phenomenon of protecting the body.
There's also been interest that if we could protect the body during this initial onslaught, we might be able to actually shorten the amount of time that people needed therapy. If somebody could emerge from this initial exposure with their immune cells intact, could we, in fact, turn those people into long-term nonprogressors and allow their immune system to control HIV? A small number of people [can do this]. Could we somehow manipulate the immune system in people who are intact, though HIV positive, and then allow their immune system to do the work without the need of treatment?
Unfortunately, those research questions remain research questions with very few data to guide us. So randomized trials are called for as a next step in the research because there's an interest; there are intriguing data to support that one or both of those factors may play a role, in terms of why it's a good idea to treat; and because [now] we have better regimens to treat with. So [we have] a sense that we've got good enough regimens that make it worth it, and a sense of the data to suggest that organized research is warranted, but [we] still [have] no answers.
If you find somebody in acute infection, do you treat?
Well, I would certainly endorse the main answer to that, which is, if there is research going on, the main thing I'd do is let the person know that there are research options available to them to help answer the question, since I don't know what to do, I don't know what to offer. But not everybody has research going on in their community and not every patient is interested in participating, even if there is. So what I do is ask the person how they feel about the pros and cons. [I ask things such as:] Do you feel miserable enough that you want to be on treatment just to get rid of the symptoms? Do you feel that, if you started treated, that you're starting down a pathway you would just as soon avoid? Because maybe [the patient thinks:] "I'll get through this with my immune system intact, and maybe I'll have five, ten years less treatment overall. I'd rather see if that's my status." I present the pros and the cons, and then, hopefully, we come to a conclusion. Even though it's based on intuition and guesswork, it's the best conclusion we can come up with.
Thanks. So, on to the next question. First line therapy: Do you think starting HIV treatment at 350 CD4 cells is the new 200 CD4 cells?
What we mean, of course, by the "new 200" is, below 200 is where badness happens, where HIV's immune destruction is so severe that, of course, that's where we see the AIDS events. I think what you're referring to is a recognition that, from the beginning, there are clinical bad events that happen between 200 and 350, so that 200 was the thickest line in the sand, but not the only line we could draw; and that above 200, sometimes bad things happened. And we could prevent those things by treating above it.
Certainly in prior years, we avoided treating at close to 350 because [HIV} treatment had complications. There were many issues: [including] toxicities, as well as inconvenience. As those are addressed, it's certainly appropriate to revisit why we delay treatment at all. So is 350 a line, or is 450 the line?
The SMART Study, which was presented a couple of times in the past year [for The Body's coverage, click here], has clarified that one of the reasons the SMART result happened the way it did -- in terms of the risk of treatment interruption -- was a recognition that there are events between 250 and 350 that are worth avoiding if treatment is good enough. I think there's an appreciation these days that treatment is good enough, and those events are worth avoiding.
So, yes. I do think the pendulum is swinging up. I think that really, the pendulum of the moment is [that] -- if resources aren't a problem -- how far above [a CD4 count of] 200, 250, 350 do we go? Is there a point above which, in fact, we shouldn't treat HIV? Is there a point at which the damage done by HIV is so little, and the benefits of the meds are so slim, that the toxicities still outweigh the benefit? I think that's a very interesting question.
The good news is that the regimens have improved enough that we can even consider that question these days. Certainly, it's fair to say that there are recognitions that there are problems with these meds. There certainly are people who have them [problems]. But you're absolutely right that the clinical research question of the moment is to revisit the 200. It's really how far above 350 should we go, given the improvements in meds and the appreciation that below 350, things did happen [to patients in the SMART trial].
For patients in your practice, do you regularly recommend treatment at 350 or some other CD4 count?
Once again, it's clear to say that in the absence of knowing what to do, all we can do is discuss the pros and cons. Various groups around the country are actually organizing to see if the funders, the NIH [U.S. National Institutes of Health], would be willing to sponsor a study to answer this with data. Now, they may or may not choose to take this on. Even if they did, it will be years before we'll have the answer in front of us. Nevertheless, I think it's fair, in the absence of those data, and the absence of trials, to raise the option for treatment [earlier] because, once again, we do know that below 350 things do go wrong. We also know that, for some people, [at a CD4 count of] 375 they don't feel well, or that they know somebody who developed KS [Kaposi's Sarcoma] at 340. So there may be, absolutely, reasons to revisit it. And that's what I do -- I raise the option of treating [earlier] with anybody.
One other finding of the SMART Study that I think is relevant is that, above 350, people who were viremic had more clinical events happen than people who had virologic suppression. This virus is bad for us. And [although] the absolute risk of events is very low: 1 percent of people or ½ percent of people develop things that we can avoid. But we do know that people who are viremic have more events. And while the absolute risk is low, the benefit of antivirals could still be measured in a trial large enough, like the SMART trial. So, if we can do benefit by treating we should at least propose it.
Related to your study of the SMART Study which looked at structured treatment interruptions -- do you think in the end that a treatment interruption is a good idea, or not? And does another CD4-guided trial, the STACCATO trial [whose results were recently published in the Lancet], with a more positive outcome, impress you?
Let's start with the whole background of treatment interruption. Probably it's worth pointing out that there are two large categories of treatment interruption at the moment: interruptions [that are] long enough so that viremia comes back, [i.e.,] that the virus returns. That's what most people think of as [a] treatment interruption, of a considerable period of time -- weeks, to months, if not years. There is a small subset of treatment interruption that we can come to at the end, in which [the] treatment interruption is a very short period of time, so short that we don't see a rebound in viremia, or that the intent [of the interruption] is to avoid viremia rather than allow it.
Nevertheless, what most people think about when they think "treatment interruption" is: I'm going to stop treatment for weeks, months, for as long as my CD4 can protect me from bad events. So if my CD4 count is 800 and if I can stop meds for a year or two and still have a CD4 count of 400, then why not? And that's a very valid question. Trials like the SMART Study were organized to try to answer it.
What we learned in the SMART trial, [which was] a trial of 5,500 people around the world, is that, first of all, the clinical event rate -- [i.e.,] bad things happening to people -- happen predominantly with a CD4 count below 350; and certainly below 250 it was even worse. So, the lower the CD4, the more the risk of bad events. No doubt about it: 250 is worse, and we don't want people to go down to those numbers anymore, if we can avoid it.
But what about at higher CD4s? Is there a problem of having a high CD4 count, and being off treatment? That's where the SMART Study had some important contributions. It's because we did see about a 1 percent rate of bad events with CD4 counts above 350 in viremic patients. It was ½ percent rate in people who had virologic suppression. So, you could say it's a 50 percent reduction -- which, indeed, it's about a 50 percent reduction. You could also say that it's 1 percent versus ½ percent.
In both groups, most people are doing fine. So, do you need treatment at high CD4 counts? Well, I think that's where people have to decide on their own. If I can reduce my risk, if treatment is that easy, that desirable, that safe...then ½ percent is definitely less than 1 percent. On the other hand, both groups, for the most part, are doing well. And the roulette wheel doesn't have too many bad slots in it. So some people would rather be off of treatment because they don't want to be on treatment. They don't like it; it's a burden; there are toxicities, and so on. Other people would rather be on treatment because the benefits are measurable.
That's really, I think, the best we can do. Studies like STACCATO, in some ways, remind us that there are many people [who are] off treatment, with high CD4s, and things don't happen. And that's true. In fact, [for] most people in a year's time, nothing bad will happen. Our CD4s protect us. The problem is that if you are that one person in 200 who develops lymphoma -- and it might have been less likely to happen on treatment -- [than you may think, if you were that patient] "maybe I should have been on treatment." We don't know who that one person's going to be. So that's why, as a policy statement, we review this [information] in kind of a global way, realizing that people need to make up their own minds.
It's not that different from anything in life. Certainly, there are things that are risky that people may choose to do. The nature of choices when we're alive is doing things that have some degree of risk, either for recreation, or because you want to avoid something. People still smoke [for instance].
But nevertheless, I think the SMART Study does point out that there is a measurable risk of being off treatment. It isn't zero. This virus does cause a problem.
OK. Now, I'd like to ask you about the clinical management of HIV-infected people who are treatment experienced. There has been a lot of good news for treatment-experienced people recently, and I think some of the studies that were released at ICAAC were very interesting. Can you talk a little bit about the hope for HIV-positive people who are treatment experienced, and what kind of treatment regimens you think can be made from the new treatments?
During the past couple of years where our field has made some dramatic progress is in the arrival of drugs that are specifically designed to address treatment-experienced patients -- to address the viruses that have become resistant to previous drugs. Perhaps we have made the most progress with regards to at least approved drugs in the protease arena. The protease inhibitors [PIs] darunavir [Prezista] and tipranavir [Aptivus] were both developed and approved because of their outstanding activity against these PI-resistant viruses.
At this conference, we saw some additional data about both of these drugs, mostly about the durability of response, and then various subsets. The data presented here at ICAAC corroborated everything we've seen in previous conferences: that these drugs do reestablish suppression, and they do it in a durable way. In those who get suppressed, [we now have study results that show] they often stay there -- [at least] one year [for darunavir] or, in the case of tipranavir now, two years later. So what these studies also teach us is that these drugs, on their own, aren't good enough; that they do better with company. Again, it's a lesson we've known for about four to five years, as well, and both of these studies corroborate that an active background allows these drugs to achieve the majority of patients with controlled virus, rather than the minority.
So we do look towards new drugs to support these currently available drugs, because the rules of treatment-experienced management are the same they have always been: identify the most active drugs you can, and start at least two of them at the same time. Now it's a question of: Can we identify the two or three active drugs? Certainly, within current classes, we can. Darunavir and tipranavir do have the preserved activity despite the typical PI mutations.
And on to new antiretrovirals. I think there was particularly interesting news about the new Merck integrase inhibitor [MK-0518] at ICAAC. Can you talk a little bit about that drug, and some other drugs in development?
The week 24 results of the Merck integrase inhibitor [MK-0518] were presented here for the first time [for our coverage of this study, click here]. This study had week 16 data from treatment-experienced patients, triple-class experienced, who started one of three doses of the Merck integrase [with an optimized background regimen] versus a placebo and an optimized background, containing whatever drugs were perceived to be the most active. The 16-week data [were] impressive enough, and now we're out to week 24; and various subsets were presented here. The three doses of the integrase inhibitor performed similarly, so they can be spoken of as one essential outcome. The study was large enough so that some of these subsets did have some interesting numbers, enough to understand what happens in various backgrounds.
Perhaps the most important background that was presented to us is the activity of the background regimen. For example, a small number [of patients] -- about a third -- had started enfuvirtide [T-20, Fuzeon] for the first time during this study. And so one of the focuses here was: What's the activity of the Merck integrase when enfuvirtide is started at the same time in our patients? What they presented is that [after 24 weeks] 90 to 100 percent of them had a viral load of less than 400 copies, with that combination and whatever compromised nucleosides they might have added to it. But when just those two drugs are fully intact, we are reestablishing suppression in essentially everybody.
Even more impressive, perhaps, is that if we don't have enfuvirtide available to us, but we just have compromised nucleosides that have a suggestion of activity, based on phenotypic testing, at least; that still, about 70 percent get to [a viral load of] less than 400.
Perhaps the most dramatic demonstration of this drug is in those patients in whom the phenotypic score was zero for the background -- meaning that whatever drugs were used might have had some residual activity, but clearly compromised. And in that setting, what we've learned is that almost half of the patients get a viral load less than 400, just with the integrase inhibitor doing most of the work. It's a very impressive demonstration so that, even 24 weeks later, their viral loads are less than 400.
Excuse me for interrupting, but how does this compare to other drugs in the history of HIV antiretrovirals?
Well, I think it's fair to say that if, based on the results at Toronto, of protease inhibitor monotherapy in people who have no protease resistance, who started on a protease inhibitor even without any other drugs, [i.e.,] just as monotherapy; the majority of patients will achieve virologic suppression. So that a single drug is able to reestablish suppression in the majority of patients isn't what's extraordinary. What's extraordinary is now we have yet another chance to do this with our highly triple-class-experienced patients, in whom one class has emerged fully intact and, of course, even more important, the safety was presented as being as safe as any of the drugs we use. Indeed, it didn't seem that there were any safety signals of additional problems, as compared to the placebo or control arm. If we can reestablish control in these patients, it isn't because it's never been seen before, but it's because it can be done in patients in whom the chances of resuppression at that high rate are often much lower than what we have seen in this trial.
Certainly, it's notable. I think in the study you're referring to: This was a study of treatment-naive patients who were starting either the integrase inhibitor or efavirenz, [both] with a nucleoside backbone of tenofovir [Viread] and 3TC [lamivudine, Epivir]. What was reported is the usual lipid changes of tenofovir or 3TC in efavirenz. Certainly nothing all that terrible, but nevertheless, you can see a small increase in cholesterol, and some other changes.
Nevertheless, the profound observation was that the Merck integrase inhibitor [MK-0518] with these two other drugs had no change in cholesterol, no change in LDL, no change in triglycerides, and a small increase in HDL, which is, in fact, a good thing, in terms of cardiac protection.
Now, it's hard to, in some ways, even understand that. Because we think that untreated HIV infection causes perturbations in lipids that should normalize if we go on treatment. So why the cholesterol and other LD fractions don't improve, or don't change, is a little bit of a mystery. But nevertheless, certainly the lack of lipid responses in this combination add to the safety profile that this drug is impressive, both in potency as well as in safety, and will certainly play a role in how interested clinicians are as this drug becomes more widely available.
Are there other drugs currently in development that you think are interesting for 2006, 2007? Maraviroc or anything else?
Well, the easiest way, probably, to look at what's coming up in the next year is to look at the expanded access programs. Because we do have, in fact, the arrival of a second expanded access program now, in the end of 2006, with a drug called etravirine (or, formally, TMC125), a "second generation" non-nucleoside [NNRTI]. The reason that it's become available is that in all of the studies so far, even in patients who have NNRTI resistance, this drug does contribute activity. In some patients, it's as much as a log and a half of additional activity. So, we now have an important non-nucleoside entering clinical practice, at least through research trials, and now the expanded access programs.
Then next year we all do expect that maraviroc, the entry inhibitor, is going to become more widely available. Of course, that's more dependent on the data and data safety review boards of pending trials. But nevertheless, that might be the third drug to emerge into more wide use, based on the observations of the data.
Now, those studies may, of course, limit themselves, in the case of maraviroc, to earlier stage patients because R5 tropism, in terms of what viruses these drugs target, are more common in people with intact CD4s. But nevertheless, even at low CD4s, perhaps half of the patients still have R5-only tropic virus. So, yes; I do think we're about to have a wave of one, two, and now even three, drugs for a range of our patients.
So it's a real exciting time for people with HIV, and for the people who treat them. I guess one thing that's still an ongoing mystery is metabolic complications. I know there was just a conference that discussed that issue. I know that at every conference, they find a little more, or they find a little bit more about this complication. I was wondering if there's anything new in terms of metabolic complications, and understanding how they happen, or understanding anything that can be done to change the complications.
What we have learned in some of these studies around metabolic complications is that drugs do differ in their metabolic profile. For example, in our group in Boston, we did one study in which we switched ten people from Kaletra [lopinavir] to boosted atazanavir [Reyataz], in a background regimen containing saquinavir and about half were on nucleosides. And while it's a small study, we focused not on lipids -- because, predictably, the lipid fractions improved or dropped when you go from Kaletra to atazanavir; that's been recorded many times in the past year or more. But what we reported was the result of DEXA scans, and particularly, abdominal CT scans, focusing on the loss of abdominal fat. And what we were able to measure is that there was a favorable trend of the reduction of abdominal fat in people who have been on Kaletra or lopinavir, and now we're seeing a trend of some improvements, in at least some of the patients, when they go to atazanavir.
Now, this is just a pilot, and randomized larger studies are already underway. But it does give us, certainly, yet more evidence that these drugs differ in their toxicities, their tolerabilities, and switching is part of what we try to do to see if we can have people find the safer drugs -- for them.
Another piece of the puzzle that we saw at this meeting is trying to take the guesswork out, and realizing that the genetics of how we differ partly explains people's different responses [to treatment]. Now, we've known that for years. At this meeting, we didn't see necessarily more data about the relevance of this in HIV management. But there were certainly yet more demonstrations of the importance and the interaction of toxicities of various treatments, of various diseases, and genetics. As we hopefully [now] understand that, we can take somebody who's having miserable toxicities on some drugs and have a better, clearer sense of what other drugs might be safer or better for them.
This meeting also did address a little bit of some of the standard "antidote" approaches. In other words, by using drugs like glitazones or uridine supplementation to address the lipoatrophy. And there were some data to support some modest benefits of pioglitazone and, again, some data on uridine supplementation. But these data are the beginning of the story; they are not yet conclusive, but certainly do suggest that we do have some leads of what we might be able to do.
Thank you. One more question: I think a lot of people are still worried about cardiovascular disease and wondering ... Is it HAART? Is it HIV? Is it smoking cigarettes? What's your opinion?
Well, part of the reason the SMART Study was so striking when we revealed the results is because there's no doubt about it that people who are on various treatments had understood -- and there were cohorts to describe -- that there were complications associated with treatment that do increase the risk of heart disease. Certainly, that's the D:A:D cohort, which documents that for every year you're on therapy, there's about a 16 percent increase in the risk of cardiovascular complications.
So, part of what made looking at that so important is: Can we understand, then, can we stop therapy in a way that decreases risk of complications? Since heart attacks obviously are worth avoiding as we all get older. Now, it's fair to say that there are treatment interruption studies, again, that allow viremia, and those that don't. We mentioned this earlier. But the ones that don't allow viremia do have perhaps some role to play, given the data I'm about to describe since, what's interesting in the SMART Study is, when we stop drugs that we thought increased the risk of heart attacks, what we saw when we stopped treatment was an increase in heart attacks. So even though we thought the drugs were causing it, when we stopped the meds we saw yet more of an increase. So we're stuck with a little bit of a paradox -- and our study team is, in fact, trying to help understand what explains this. Because if the treatments are making it [heart disease] worse, then removing the treatments should make it better. But they got worse again.
So it looks like, perhaps that, at least from the 30,000-foot view we have now, stopping treatment isn't the way to reduce the risk of heart attacks. So that's why some of us are doing, for example, treatment interruption studies in which we try to allow people to have a couple of days off treatment that doesn't allow viremia. For example, our what we call FOTO study -- five days on, two days off -- of people who are on the very long half-life drugs like efavirenz, tenofovir, FTC [Emtriva] -- drugs that stay in the body more than 24 hours. What we saw in our pilot study that we presented is that patients who took two days off -- particularly, weekends off -- even a year later, maintained suppression.
So part of what our field is doing is trying to see: Can we simplify therapy and avoid the complications, as well as avoid the burden of daily therapy? Are there ways to make therapy even safer? Unfortunately, treatment interruption may not be the way to make things safer since, as we observed, sometimes things get worse ... even things that should have gotten better.
Thank you very much.