March 8, 2010
This is part one of a two-part interview discussing highlights from CROI 2010 related to antiretroviral therapy for HIV-infected patients. Part one focuses on antiretrovirals in development; part two focuses on already-approved antiretrovirals, the question of when to initiate antiretroviral therapy and the evolution of the CROI meeting itself. You can use the table of contents below to quickly jump from section to section.
Table of Contents
Myles Helfand: Although conferences such as CROI are research meetings at their core, the findings presented there have an undeniable power to alter clinical practice. As CROI 2010 slips into our rearview mirror, let's take a look at some of its key studies for signs of how the treatment landscape may evolve over the months to come.
Joining me for this examination is Joel Gallant, a professor of medicine and epidemiology at the Johns Hopkins University School of Medicine. Dr. Gallant, thanks for joining us.
Joel Gallant: My pleasure, Myles.
Myles Helfand: We've got a lot of research to get to, and obviously, we'd like to put it into as small an amount of time as possible. So let's dive right in.
Let's begin with what is always one of the more interesting topics in antiretroviral therapy: the drug development pipeline. What did CROI 2010 bring us?
Joel Gallant: We hear a lot of complaints that there's not much going on in the pipeline. The pipeline is certainly not as full as it used to be. But there were a few drugs discussed, including a new CCR5 inhibitor; a new integrase inhibitor; and, of course, the so-called Gilead [Sciences, Inc.] "quad" that consists of tenofovir [TDF, Viread], FTC [emtricitabine, Emtriva], elvitegravir [GS 9137, JTK-303], which is their integrase inhibitor, and their new booster, GS 9350.
Myles Helfand: Which has a new name, I believe: cobicistat.
Joel Gallant: That's right. Cobicistat. You need to learn that and be the first on your block to know it.
Myles Helfand: Or the first in your clinic, as the case may be.
Myles Helfand: Let's get into the quad tablet a little bit. How did this hold up in the study that was presented at CROI?
Joel Gallant: Cal Cohen presented three studies in 10 minutes, which was a tour de force. Because the organizers like to combine related topics, and make people have to rush through a lot of data. So he presented the study of the quad compared to Atripla [efavirenz/tenofovir/emtricitabine, EFV/TDF/FTC]. Then he also presented a study where the cobicistat, or GS 9350, was used in comparison with ritonavir [RTV, Norvir] to boost atazanavir [ATV, Reyataz].
In both studies, the results looked great. These are smaller, phase 2 studies. But basically, the quad looked comparable in efficacy to Atripla. The cobicistat looked comparable to ritonavir, in terms of its ability to boost atazanavir, in that study. But the unexpected finding, especially in the quad versus Atripla study, was that the serum creatinine was slightly higher in the cobicistat arm, or the quad arm. This resulted in a modest decline in estimated creatinine clearance, which is the way we measure kidney function.
This was concerning, of course, because we already know that tenofovir can sometimes affect kidney function. And the idea of developing a quad tablet that had two drugs that were toxic to kidneys was not very appealing. So they then did a quick study in healthy volunteers, where they gave cobicistat to people, and they measured their creatinine and their estimated GFR [glomerular filtration rate], and then they used iohexol clearance as a test of true GFR. It actually measures creatinine clearance, instead of estimating it.
What they found was that, while the estimated GFR went down a little bit, the actual GFR didn't change. Their theory is that this is not a kidney-toxic drug; it's simply that it inhibits the secretion of creatinine from the kidneys, and therefore, it elevates creatinine without actually doing any harm.
So that's the hypothesis. They're going to go ahead with the phase 3 clinical trials. The hope is that this is just a tiny decrease in apparent kidney function that occurs within the first week or two and then, from then on, remains stable. So if you needed to monitor for, for example, tenofovir nephrotoxicity, you'd do it just the way we do now. This was an unexpected finding; hopefully, not a big problem.
In terms of the ritonavir versus cobicistat comparison: Again, as I said, the efficacy was fine. I would say one of the disappointments was that it didn't look better, in terms of gastrointestinal side effects, than ritonavir. I think we were all hoping that this would be the booster that had none of the baggage that comes with ritonavir. But, at least in this small study, that didn't appear to be the case.
Myles Helfand: The cobicistat creatinine findings: Was that a substudy, or was it a separate study that was done in people who received cobicistat as monotherapy?
Joel Gallant: The study where they did the iohexol clearance was a completely separate study in HIV-negative volunteers.
Myles Helfand: OK. And they didn't receive tenofovir, as well?
Joel Gallant: No, this was just cobicistat.
Myles Helfand: Is it known whether there's a potential interaction between cobicistat and tenofovir?
Joel Gallant: They've looked at that. There's certainly no pharmacologic interaction. But if you look at the curve that they presented, two weeks of looking at kidney function after seven days of monotherapy, the actual GFR is completely stable across those weeks -- whereas, in the quad versus Atripla study, within a week, you saw this decline in creatinine. So it's a pretty convincing argument.
And, you know, there are other drugs we use that do this, too. Trimethoprim [Proloprim] is a good example. It inhibits creatinine excretion -- and you do see a very small increase in creatinine. It's something that we don't even notice, or look for. But things are a little different in this case, because of the fact that tenofovir is on board.
The examples I remember are cimetidine [Tagamet], which used to be used for peptic ulcer disease, before we had drugs that had fewer interactions; and trimethoprim. There may have been another example [Dr. Cohen] gave that I'm not remembering.
Myles Helfand: Is he right in that comparison?
Joel Gallant: Well, he's certainly right that those drugs do that. And it sounds like he's right that cobicistat doesn't affect true creatinine clearance. Whether the hypothesis that this is due to inhibition of tubular secretion is correct, I don't think we know. The assumption is based on the fact that other drugs that have this effect on creatinine, without actually damaging kidneys, do it in this way. So it's a reasonable hypothesis, but it probably needs to be established more firmly.
Myles Helfand: I imagine phase 3 trials will help.
Joel Gallant: Phase 3 trials will definitely help to determine whether this is indeed a problem that occurs quickly and then doesn't progress. That's what we care about. The last thing we want to do is have something that is causing even just an apparent alteration in creatinine clearance. If it goes on over time, then the problem with that is that it makes it very difficult to monitor for tenofovir toxicity. But if it's just something that happens within a week and then doesn't change, then it's not difficult to monitor for tenofovir toxicity the way we always do.
Myles Helfand: The quad versus Atripla comparison, was that in treatment-naive patients?
Joel Gallant: Yes. That's what this combination is being studied for: as a potential replacement for Atripla.
Myles Helfand: Thus far, from what you've seen, how is it stacking up?
Joel Gallant: So far, I think it looks very good. It certainly was associated with a lot fewer neuropsychiatric side effects. If you look at things like dreams, dizziness, fatigue, sleepiness, those things were much lower. So, at least in a phase 2 study, this seems to fulfill the promise that it's a combination that's as good as Atripla, but without the side effects [that are unique to Atripla].
Now, of course, the quad is not the only combination that is going for that Atripla place. There are several other competitors that are doing the same thing. So it's just a matter of time, I think, before we have something that's going to be as good as Atripla, but better tolerated. It's a question of which of these various regimens will win.
Myles Helfand: This may be an overly provocative question, but given that Atripla and quad share two ingredients, how much of a furthering of the evolution of treatment options is the quad drug, really? Half of its components are already being utilized in first-line treatment.
Joel Gallant: I think that where it's an advantage is for patients starting therapy. When I start somebody on Atripla, which is certainly something I use a lot of, I have to spend several minutes sitting and talking about how to deal with getting used to the drug, and the changes they are going to go through, in terms of the neurologic side effects, the dizziness and the dreams, the difficulty concentrating, and the potential for rash. I can't just hand over a prescription and say, "Take this once a day."
We would love to have a drug, or a combination pill, that we could just hand a patient a prescription to and say, "Take this once a day." We need something without a lot of baggage. I would say this is looking good for that.
Atripla has always had -- as good as it is, and as effective as it is -- it's always had its downside. So while you could say that this isn't all that new -- two of the drugs are the same -- I think from a patient standpoint, it's very important that something [be available to potentially] replace Atripla. And in fact, all three of the drug regimens that we're looking at now that could potentially replace Atripla have tenofovir and FTC in them. It's just a question of the third drug.
Myles Helfand: What are the other ones?
Joel Gallant: There's the coformulation of tenofovir, FTC and rilpivirine [TMC278], which is a new NNRTI [non-nucleoside reverse transcriptase inhibitor]; and then there's the idea of using tenofovir, FTC and raltegravir [RAL, Isentress] once a day, which is also being studied.
Myles Helfand: How important do you feel is the "one-pill, once-a-day" scheme? There are a number of other regimens that are on the preferred list on the U.S. DHHS [Department of Health and Human Services] guidelines -- boosted darunavir [TMC114, Prezista], for instance, boosted raltegravir -- that a person could turn to, but that involve more than one pill. Why would they not already be preferable to, say, Atripla, or a quad drug?
Joel Gallant: I think once-a-day dosing is quite important to many, many patients. Not all: There are some that would prefer, for example, Truvada [tenofovir/emtricitabine, TDF/FTC] plus Isentress, even though it's twice a day, just because of the tolerability. We have patients who are popping seven or eight vitamins a day. They don't care if their meds are one, two, three or four pills a day.
It really comes down to side effects and some of the other considerations that we take into account when we start a regimen. I have many patients who come in saying, "I want the one pill, once a day. I want Atripla." I give them Atripla. And there's always going to be a small number of them who just don't tolerate it, and who switch over to, for example, Truvada plus a boosted PI [protease inhibitor]. They are happy with the drug, because it doesn't have the side effects that they had with Atripla. So they don't mind that it's three or four pills a day.
It just depends on patient preference. For people who have a real pill phobia, or who are paying enormous copays on medications, sometimes the number of pills can really make a difference.
Myles Helfand: OK, fair enough. It seems like, on balance, the further development of the quad tablet is good news. We're looking at something that, at least for now, looks like it has the potential to enhance treatment options for naive patients.
Joel Gallant: Oh, definitely. It was concerning that elvitegravir was not going to get developed because, as we'll come to, it's becoming very, very difficult to develop a new drug for a treatment-experienced patient population. So the development of this booster, cobicistat, was a big plus, because it allowed them to do studies in naive patients. If you remember, this drug was going to need to be boosted by ritonavir, and they couldn't study a ritonavir-containing combination in a naive patient unless it included a protease inhibitor, because there was concern that even low-dose ritonavir could lead to drug resistance. So they were kind of stuck with studying this in an experienced population. The booster came along and rescued them, allowing them to use this up front.
Myles Helfand: Your comment about the difficulty in developing drugs for treatment-experienced patients leads pretty nicely into some news that is not as good as the quad tablet: the results that were presented about the vicriviroc [SCH 417690; SCH-D] phase 3 study.
Joel Gallant: Yeah. This was a disappointment. This is the presentation by Joe Gathe on the VICTOR-E3 and VICTOR-E4 studies, looking at vicriviroc, which is a CCR5 antagonist, versus placebo plus an optimized background regimen in patients who've had extensive treatment experience and have R5 tropism.
Just to remind people: This is the same study design we have used ever since the TORO studies led to the approval of Fuzeon [enfuvirtide, T-20]. It's been a standard study design, where you take people who are experienced; you pick the best combination for them, using available drugs; and then they are randomized to either take that combination with the investigational drug or without it. It's served us well. It led to the approval of Fuzeon, of Prezista, of Selzentry [maraviroc, MVC, Celsentri], of Isentress. All of these recent drugs were approved in this way.
The problem is, I don't think this design works anymore. And this study was a classic example, because what happened is, even though the drug worked virologically, they couldn't really determine a difference between the people who were on just the background regimen arm versus those who were on the vicriviroc arm, because the people on the background, the control arm, did so well. That just reflects the fact that we have lots of drugs we can use now, and that most people will do fine with existing drugs. Therefore, the ability to detect a difference, using an experimental drug, is very low.
If you looked at the subset of patients who had zero, one or two drugs that were active in their background regimen, you did see that vicriviroc helped, that they had a better response. But overall, the people in the background regimen arm did just as well as the vicriviroc arm.
I think this is telling us what we probably already knew: This TORO-style study design is not going to work anymore. We're going to have to figure out a new way to get investigational drugs approved, if they're being used for treatment-experienced patients. Because it's certainly not ethical to restrict the number of active drugs people get. And yet, without doing that, there's almost no way to see the benefit of your investigational drug.
Myles Helfand: For people who had two or fewer active drugs in their background regimen, we're talking about a pretty large difference: It was something like 70% in the vicriviroc arm, versus 55% or so in the placebo arm, obtained viral loads below 50 copies/mL. Why was that not enough to define superiority?
Joel Gallant: Because it was too small of a number. In the overall numbers, you're looking at almost 500 patients randomized to get vicriviroc, and over 200 randomized to get the placebo. The group that had two or fewer active drugs was a subset, and the numbers simply weren't great enough to allow the overall study results to reflect that difference.
Essentially, you'd have to either do the study over or somehow continue it out using only patients who met that criterion. It would be very expensive, and it would take a very long time to get the study done.
Myles Helfand: In this case, the maker of the drug has already announced that it's not going to continue to develop it in treatment-experienced patients.
Joel Gallant: Yeah. And you know, look at the way maraviroc is going: Very few people use it, even though it's a really great drug. I think right now the problem is that the CCR5 antagonists are probably not very exciting to investors, CEOs and boards of directors because they see that the first drug in that class is hardly selling at all. So why come up with a competitor that just has a modest advantage over the one that's already out there and kind of languishing?
I think that's not a function of the drug; it's a function of the fact that we're dependent on a very expensive Trofile assay to be able to use the drug. There were some data presented suggesting that we might be able to move towards less-expensive genotypic tropism assays in the future. That would be great, because we would potentially combine them with our standard genotypes that we're doing, and would, for a modest increase in cost, get tropism results. That could really move us to use these drugs more often than we're doing now. But right now, it doesn't look like a very exciting field to get into.
Myles Helfand: Stepping back a little bit to the broader view of drug development: Is this a regulatory issue? Or is this a study design issue?
Joel Gallant: I think, in this case, it's hard to know. But I would say it's a study design issue. And I think the writing was on the wall. I think a lot of people knew that we really weren't going to be able to use this design anymore. There were so many drugs and so few people who really were desperate for brand-new drugs.
I look at my own patient population: I think I have one patient out of 350 who has truly untreatable virus with current drugs. It's pretty hard to do a study when the numbers are so few.
I think the question, then, is, "OK, what do you do instead?" The way that Gilead has approached this with studying elvitegravir for treatment-experienced patients is to abandon that type of study design and instead compare elvitegravir with raltegravir. You simply compare it against another drug in its class; that way, you're not denying patients access to anything. And you'll at least be able to say, well, this works as well as raltegravir. I would assume that they wouldn't be doing this if they hadn't talked to the FDA [U.S. Food and Drug Administration] and made sure that it was going to be acceptable to allow them to get this drug approved.
I think that's probably the way we're going to move. Maybe the way that a CCR5 antagonist should be tested in the future is just a head-to-head comparison with maraviroc.
Myles Helfand: So ultimately, then, the same general kind of study that we just talked about with the quad versus Atripla -- but obviously for treatment-experienced patients.
Joel Gallant: With the quad versus Atripla, it's easy: You're taking people who have never been treated, and you're just comparing; you're comparing two things that suppress their virus. So that's pretty straightforward.
The problem with this new design that I'm mentioning is, let's say you give people a combination of a background regimen plus raltegravir, versus a background regimen plus elvitegravir. And everybody does great. How do you know that the integrase inhibitor had anything to do with their doing great? How do you know that they wouldn't have done great without the integrase inhibitor? Because you don't have an arm that's showing that.
So the FDA could come back and say, well, maybe they were doing great because they were on a boosted PI and some nukes [nucleoside/tide reverse transcriptase inhibitors]. You haven't shown me that your drug contributed to the success, whereas, with the quad versus Atripla, it's obvious that it did.
It's tricky. I don't know what the answer is. But it's clearly not going to be a TORO-style design.
Myles Helfand: All right. Let's talk briefly about some of the other drugs in development that were discussed at CROI. Since we just came off of vicriviroc, let's begin with another CCR5 antagonist in development: TBR-652.
Joel Gallant: TBR-652 is another CCR5 antagonist, as you mentioned. There was a study presented, again by Cal Cohen, showing excellent results in a short-term study. It was very well tolerated, as most of these drugs seem to be.
What was interesting is, they commented on the fact that it also seems to inhibit CCR2, which is a coreceptor that most of us haven't spent much time thinking about until now. The argument that they made was that CCR2 has been associated with a number of inflammation-related diseases, so perhaps this [drug] has some kind of anti-inflammatory activity as well. An interesting concept, but I don't know what it will mean yet.
Myles Helfand: Is there any actual proof that something that can potentially inhibit the CCR2 receptor on CD4 cells is going to help reduce any of those signs of inflammation that research increasingly indicates is happening even in fully treated people with HIV?
Joel Gallant: No; I don't think there's that kind of proof. We know that the biggest thing we do to decrease inflammation is give people antiretroviral therapy. And you see a big decrease in inflammation when you do that. But it's true that you don't get to normal levels; people who are HIV negative have even lower rates of inflammation than people who are fully suppressed on therapy. Is it possible that blocking CCR2 could further decrease that level of inflammation? It's certainly worth looking into, but I don't think we know that yet.
And you know, whenever you block things that are part of us, you have to wonder what could be the downstream implication. So far, blocking CCR5 seems to be pretty safe, but I don't think we can just say that blocking all of these coreceptors is going to be a safe thing to do. They must be there for a reason.
Myles Helfand: Also, like you said, despite all of the hype before maraviroc was approved, it has not turned out to be a particularly popular drug. I think price may be part of it. The Trofile assay, the need to use that, is certainly part of it. I also think the fact that it was approved for treatment-experienced patients, many of whom are CXCR4 tropic or dual tropic, and can't benefit from the use of the drug -- that might have had a role, too. Plus, given that I think the results of this particular study found that the drug [TBR-652] did decently well, in terms of suppressing viral load, but it wasn't outstanding; it didn't knock the ball out of the park, necessarily. There wasn't a 2-log reduction in viral load.
Joel Gallant: No. It was a little more than a 1.5-log reduction.
Myles Helfand: What does that mean we're looking at, as this drug moves through the development process?
Joel Gallant: I think, if you want to think of the best possible outcome for this class of drugs, it would be that, for an extra 50 bucks, when you do a baseline genotype in somebody who has just been diagnosed, you get a tropism [finding] in addition to the resistance test results. If they are R5 tropic, you say, hey, this is a really well tolerated, very safe class of drugs, and if I use it now, I'll get a benefit from it. If I wait till later, I might not -- they may have lost their R5 tropism. So why not use it first?
I think that makes a lot of sense, to use this incredibly well-tolerated class of drugs as early as possible. But we just can't do that right now with a $2,000 tropism assay.
I'm not dismissing this class. I think that, with improvements in the laboratory tests that we use, we could definitely move these drugs up sooner. And if we had a once-a-day, very safe, well-tolerated drug, it would be a logical approach to do this early. So I think that it's worth continuing to develop these drugs as we work on the laboratory issue in tandem.
Myles Helfand: What other pipeline drugs did CROI tell us about?
Joel Gallant: We heard a little more about the Glaxo[SmithKline]-Shionogi integrase inhibitor, which I don't think has a name yet.
Myles Helfand: It does have a very long denotation, though: S/GSK1349572.
Joel Gallant: Yeah. It probably has a nickname to its friends, but I'm not privy to it. This is supposed to be the first second-generation integrase inhibitor, one that will be active if you fail Isentress and elvitegravir. They didn't present a big clinical trial using this drug, but they did present some resistance data that suggest that, in fact, no single integrase mutation will knock it out, and the common integrase mutations that we see with Isentress failure will not cause cross-resistance to this drug. So that does look promising.
Myles Helfand: What else have we got?
Joel Gallant: That's about it. There were other drugs out there, and some other ones that were talked about. But on my radar screen, those were the ones that I paid attention to.
Myles Helfand: As a whole, then, this is not a terribly rich drug pipeline, when you consider what we were looking at when we went to CROI five, six, seven years ago. Then it was a very rich, rich field of candidates. This is not quite that wealth of options.
Admittedly, that may just be part of the nature of having about 30 drugs that are currently approved for use. But given what you have said about vicriviroc, the nature of studying treatment-experienced patients and the nature of trying to develop a drug that will help these people, what kind of situation is playing out for the future of HIV treatment? You'd said only one out of your 350 patients have untreatable virus, meaning extreme treatment experience, significant drug resistance across the board. But presumably, over the next few years, as our newer candidates spend more time out on the market, people are going to increasingly develop that level of resistance.
Given that it's becoming more difficult to study drugs for treatment-experienced patients, and that the pipeline does not appear to be rich, what kind of situation are we facing a few years down the road? What do we say to patients who develop extreme drug resistance?
Joel Gallant: I worry about that. Because if you put yourself in the shoes of some pharmaceutical company's CEO -- very expensive shoes, I'm sure -- and you look out and say, "Should I develop a new HIV drug?" Well, you'd be crazy to do that, because it costs billions of dollars to do it. And everybody's doing just fine.
Now, as you say, that may not last. We may be in a little honeymoon period where, four or five years from now, we're going to be facing integrase resistance and extensive resistance to non-nukes and PIs. And we're going to be in trouble, without a lot of stuff to bail us out. I don't know. When I put people on so-called salvage regimens, I really explain that. I say, "You know, this could be it for five or more years. So you've really got to make this work."
On the other hand, you could look at it in a more positive way and say, you know, given the situation right now, it's amazing that we have any drugs at all being developed. And I would also point out that not every drug that's in the pipeline gets presented at every meeting. For example, we didn't hear about the Progenics [Pharmaceuticals, Inc.] CCR5 antagonist at this one, or some of the other second-generation NNRTIs. We didn't hear about rilpivirine, which is in development. There are drugs out there that don't get presented every time. So the pipeline is a little bigger than it would appear, just looking at CROI. But I do worry about this.
I'm very glad that there is, for example, a second-generation integrase inhibitor in development. Because I'm sure that we're going to see Isentress resistance starting to pop up. And that's going to be a hard one, unless we have a drug that will rescue us from that.
Myles Helfand: It sounds like that just gets more to the point that you raised just a little bit ago, about how maybe this raises the importance of counseling patients on the importance of adherence, and making sure that they don't face obstacles to adherence -- whether it's social, psychological, financial, emotional -- that may prevent them from taking their drugs as often as they need to for them to retain their effectiveness.
Joel Gallant: Oh, it's absolutely the case. When you look at patients who are on complicated, multidrug salvage regimens these days, they can kind of be divided into two groups. There's the group that developed all that resistance because they've been on drugs since the late '80s or early '90s and, through no fault of their own, they took what we gave them. And because it wasn't fully suppressive, they developed lots of resistance. That's a very common scenario. Now they are having to take these difficult regimens.
But there's also a big group who have started therapy in the HAART [highly active antiretroviral therapy] era, after 1996, and who have failed because they weren't taking their medications the way they should. And they are now on these regimens. The problem is, of course, if they thought the first regimens were hard, imagine what it's like taking these.
So we really have to make sure that people understand that a first regimen these days is as easy as it's going to get. And if they want to continue on something like that, they need to really make sure that they take it right, and don't get to the point where they are having to take salvage therapy.
This transcript has been lightly edited for clarity.
Copyright © 2010 Body Health Resources Corporation. All rights reserved.