The Body: The Complete HIV/AIDS Resource
Follow Us Follow Us on Facebook Follow Us on Twitter Download Our App 
Professionals >> Visit The Body PROThe Body en Espanol
CROI 2008: Boston, Massachussetts; February 3-6, 2008

Key Links:

By Topic:


  • Email Email
  • Glossary Glossary
The Body Covers: The 15th Conference on Retroviruses and Opportunistic Infections
Atazanavir Noninferior to Lopinavir in Treatment-Naive HIV-Infected People

February 6, 2008

Multimedia Options

Listen to Audio (17 min.)

Please note: These files can be quite large. Allow some time for them to download.

Although atazanavir, also known as Reyataz, is not uncommonly used in first-line patients, it is not officially approved by the U.S. Food and Drug Administration [FDA] for such use. In a large study presented at CROI 2008, atazanavir was compared with lopinavir/ritonavir, a preferred first-line drug also known as Kaletra. To discuss the results I spoke with Dr. Donnie McGrath, one of the study authors. Dr. McGrath works at Bristol-Myers Squibb.

Bonnie Goldman: Could you tell me a little bit about the study results, Dr. McGrath?

Donnie McGrath: The CASTLE study is a 96-week study designed as a non-inferiority study to evaluate the comparative safety and efficacy of boosted atazanavir and ritonavir [Norvir] compared to boosted lopinavir given twice daily. The boosted atazanavir is given once daily. They are both given on a backbone of fixed-dose tenofovir and emtricitabine, otherwise known as Truvada. That's given once daily.

This was done because boosted atazanavir currently does not have a treatment-naive labeling in the United States, and it doesn't have a naive indication in the EU [European Union], either. It had never been compared in a head-to-head trial against the standard of care boosted protease inhibitor. So the study was done for that reason. The primary endpoint was at week 48, and it was the week 48 results that we presented here at CROI.1

Click to enlarge
Jean-Michael Molina et al. CROI 2008; abstract 37. Reprinted with permission.

In terms of the results themselves, the primary endpoint was met, and boosted atazanavir was determined to be non-inferior to boosted lopinavir. The primary endpoint was an HIV viral load less than 50 copies per mL, and that was looked at using a couple of different analyses. The principal one was a confirmed virologic response, in which patients had to have a confirmed viral load of less than 50 at week 48. Using that principal analysis, 78% of the patients on atazanavir were undetectable, and were successes, and 76% of the patients on lopinavir [were successes].

There is a pre-defined criterion, a delta of -10%, which is required by regulatory agencies for these non-inferiority studies. Using that delta, you get quite narrow confidence intervals around your result, and the lower bound of confidence interval for this result was -3.6, which is well below that -10%. So it easily met the criterion for non-inferiority and that's the primary result of the study.

In terms of the efficacy, the safety results were pretty much as expected. These drugs have been on the market for quite some time, and I think clinicians are familiar with the safety profiles of these drugs.

For boosted atazanavir, more jaundice was seen, and we know that hyperbilirubinemia is something we expect with this drug.

For the lopinavir arm, the GI [gastrointestinal] toxicity profile of that drug has been known for some time, obviously: There was more diarrhea and nausea on that arm. Lopinavir/ritonavir has a worse impact on the metabolics of the patients, and that particularly manifests itself in increases in total cholesterol, non-HDL and triglycerides for those three lipid parameters. Boosted atazanavir was significantly better in not increasing those parameters. It was very significant at .0001 for those three parameters.

There are some interests in this study, too -- because it's a very large study: 883 patients -- in terms of renal safety, because the backbone drug, tenofovir, has been associated in some studies with renal toxicity. But I think this study was reassuring overall in terms of renal safety. There was negligible decline in serum creatinine on both arms. In terms of renal adverse events, they were only 2% of all grade adverse events reported on both arms of the study.

I think overall it's encouraging that renal safety doesn't seem to be a big issue when using tenofovir with either of these boosted protease inhibitors in the treatment-naive population.

Bonnie Goldman: What were the gender breakdowns, and race?

Donnie McGrath: 31% of subjects on both arms were female, which is a pretty decent number. It's higher than has been seen in other recent studies and I think that's good. We want to get more women in our studies.

In terms of the racial breakdown: I don't know that data off the top of my head. But the study was stratified at baseline by region, as well as high and low viral loads. The regional breakdown: Again, I would have to look at these data. I think about 46% were recruited in South America, and then 15% in North America, in Asia, and in Europe, sort of making up the other 50% or so of the study. So it was a multinational study. Data on efficacy and safety by those subgroups -- male, female, and different regions and ethnic groups -- will be presented at future meetings.

Bonnie Goldman: Were there differences in resistance mutations?

Click to enlarge
Jean-Michael Molina et al. CROI 2008; abstract 37. Reprinted with permission.

Donnie McGrath: The resistance data has not been -- the analysis hasn't been completed yet. As you're probably aware, when patients fail and they rebound, sometimes the samples are not able to be run immediately. The viral loads may not rebound high enough. So we've been still getting in some genotypes of patients who have had virologic failure in week 48, up to the last month or so, and it's a large dataset. We're still analyzing that. That should be complete in the next month or so, hopefully in time for the International Workshop on HIV Drug Resistance and Treatment Strategies meeting in June, which is in Spain, I believe. We'll present it there.

Bonnie Goldman: I have here Dr. Gerry Pierone, and he's going to ask some additional questions.

Gerald Pierone: I think this is a trial that clinicians have been waiting for for a long time. Because, as you've alluded to, many clinicians out in the practice world have already started using boosted atazanavir frontline, even though the guidelines don't contain that yet.2 Were there any findings in the study that surprised you in any way?

Donnie McGrath: No, Gerry. It was pretty much as anticipated. I think that we anticipated that the drugs would be very similar. I think that that's probably what most people expected -- that these drugs are potent and if patients take them and can stay on therapy, they'll work. I think the ITT [intent-to-treat] analysis shows that.

The thing I should note about this that did surprise me a little bit was that the discontinuation rate was very low. In fact, it was low in both arms. It was only 9% for boosted atazanavir and 13% for lopinavir. I'm not aware of any treatment-naive study with lopinavir that's had discontinuation rates that low. So I think the investigators and the patients were very committed to the study. I think that validates, again, the results of the fact that we were able to keep patients on study and on drug, and as many as possible.

So that surprised me a little bit. But I wasn't surprised at the efficacy results. As a clinician, I have used both atazanavir and Kaletra, and I know that in the right patients, if they can tolerate the drugs, they'll do very well.

The safety results: I don't think anyone would be surprised there. Again, I think there was just some keen interest in seeing the renal safety data and I think that was reassuring; there wasn't really a signal that renal toxicity was a problem here.

Gerald Pierone: One of the findings that caught my eye that was a bit surprising was a post hoc analysis that looked at the drop-off in response rates with Kaletra with high viral load. Can you comment on that?

Click to enlarge
Jean-Michael Molina et al. CROI 2008; abstract 37. Reprinted with permission.

Donnie McGrath: Yes. You know, it was pre-specified in our analysis plan to look at response rates by baseline CD4+ strata. I think all large studies will do that to see if there's a pattern there. When we did that, when we looked at doing that prespecified analysis to look at those response rates, it was clear that there was an interesting pattern there. Or the pattern seemed different; there was this ladder-like step pattern on the lopinavir arm, and the response rates seemed to be consistent for atazanavir, regardless of the CD4+ count at baseline.

When we observed that, we then decided to do a post hoc statistical test on that. That was the post hoc part of the decision to do the actual test. The most valid way, and the scientifically rigorous way to do that, is not to compare one arm to the other because you're getting into issues of multiplicity and small sample sizes if you try that. The only valid way to do that is to do a Cochran-Armitage test, where you're testing within arm. We're not testing, comparing, the two arms here.

We're looking at the lopinavir pattern and saying: Is there a linear trend here? When you have a categorical variable, and you have a binary variable, the Cochran-Armitage test is the valid one to do. So we did that and that's why you'll see on that graphic there the P values above both groups, and the trend is positive, significant, for the lopinavir arm. I suppose that wasn't a great surprise when you looked at that. It's fairly obvious when you see those steps down that there may be something going on there. There obviously wasn't a positive trend for the atazanavir arm and I think that was a little bit surprising. But you know, I think the previous studies have had shown some differential response rates by CD4+ count. I think Study 863, the original registrational study in treatment-naive patients of Kaletra, showed differential response rates with low CD4+ count and high CD4+ count.2

I know other studies have not shown a difference so, it's what we found and again, Study 863, I think, also indicated a differential there. I think I was a little surprised at that. But I think it's important information. You know, I'm happy to comment further.

We did want to dig into this a little bit. So we looked at the treatment outcomes of those patients who came in with those low CD4+ counts and the rates of virologic failure in both arms are identical, which would suggest it's not a potency issue. It's not that they are failing. But what drives this difference is the discontinuations in patients with low CD4+ counts.

On the Kaletra arm, about half of those discontinuations were due to adverse events. The other half were to miscellaneous issues -- withdrew consent, doesn't want to be in the study -- normal things we see and that we saw in the atazanavir arm. So the difference seems to be that tolerability in patients with low CD4+ counts. That seems to be what's driving that and I think that's plausible from our clinical experience. At least, that's some of the feedback we've got.

Gerald Pierone: That's an interesting finding and I think, as you say, it is consistent with clinical experience. It makes sense intuitively that if someone is sicker to begin with, that if they have a better-tolerated regimen, they may have better clinical outcomes.

Donnie McGrath: Yes. The test itself was post hoc. There are all the caveats around that. But I think it was interesting to look back and see that even in Study 863, what was published -- the CD4+ results were published in the Journal of Infectious Diseases in 20043 -- was that difference in response rates by CD4+ count, which I had forgotten about, frankly. I think it's a real finding, but given the caveat -- again, it's post hoc -- I think that if it makes clinical sense to people then that's fine. People can make of it what they will.

Gerald Pierone: On the front end of the study, a greater number of patients withdrew consent on the lopinavir/ritonavir arm: 13 versus 4, 3% versus less than 1%. Do you think that was a reflection perhaps of the disappointment of subjects or patients that they were randomized to the twice a day arm?

Click to enlarge
Jean-Michael Molina et al. CROI 2008; abstract 37. Reprinted with permission.

Donnie McGrath: You know, that's always a chance. It was an open-label study, and so there's always a risk that you'll have a loss of some patients if they don't like what they've been randomized to, particularly when you have drugs that have been on the market for a while. Patients are aware of the profiles [of the drugs] and so forth.

We did, of course, look and dig into that number a little bit. The 3% is not really out of the norm too much for "withdrew consent" for these sorts of trials. You know, pretty much most of the treatment-naive trials, we see 1% or 2%. So 3% might be on the higher range here. When we looked at that, there were a couple of withdrew consents, and a few withdrew consents in patients who were having ongoing tolerability issues. All of those reasons for discontinuation from a study, as you know, are determined by the investigator. The protocol doesn't mandate what the reason is.

If a subject is having some tolerability issues and the investigator thinks they should be able to stick with it, and the subject says, "No, I really can't tolerate it," most investigators will put down, "Withdrew due to AE [adverse effect]," but some will say, "Withdrew consent." There's a little bit of a subjective opinion going on here as to why. But we did look at that, and there are certainly several of those -- that 13 subjects -- were having ongoing tolerability issues.

In fact, at week 96, after they were queried -- because whenever things don't quite add up, we do query -- and some of them have been reassigned by the investigator as being discontinued due to an adverse event. So that number is probably a bit less than that.

Indeed, there is a question -- and I'm going to proactively deal with this -- that the capsule formation of Kaletra was used in the study through week 48, and the Meltrex is now allowed after week 48. We stuck with the capsule because we felt that to change the formulations would impact the integrity of the primary 48-week endpoint. There's this question of whether or not there is any difference in tolerability with the newer tablets than the capsules.

We made that decision. I think the very low discontinuation rate was very reassuring for us that, really, patients were tolerating the lopinavir reasonably well. We only had two patients who withdrew consent and the specified reason was they wanted to switch to Meltrex. I think that, again, was very reassuring for us that only two out of the 443 patients randomized to lopinavir said, "I want to switch to the Meltrex." In those cases, I'm not sure if it's tolerability issues. I think it was more like they didn't want to have to stick their drug in the fridge anymore. But, there you go.

Gerald Pierone: So we can anticipate now a new 800-patient study comparing once-a-day Kaletra and Meltrex?

Donnie McGrath: [Laughing.] You can anticipate away, but I don't think we'll be doing it.

Bonnie Goldman: How does it work, in terms of FDA approval, for drugs to be used in first-line HIV treatment? Is this part of the procedure of getting that?

Donnie McGrath: Yes, obviously. As I said at the beginning, we do not have a treatment-naive indication for boosted atazanavir in the U.S., and as Gerry said, many clinicians are using it, anyway, as first-line therapy. But we do want to get it on our label. Obviously the FDA will look at this data. I can't really comment on our filing plans, but I think everyone's aware that they will be getting the data. Hopefully our label will change subsequently, and that's good for clinicians. They'll have the data in the label.

Bonnie Goldman: Thank you very much.

This transcript has been lightly edited for clarity.


  1. Molina J-M, Andrade-Villanueva J, Echevarria J, et al. Efficacy and safety of once-daily atazanavir/ritonavir compared to twice-daily lopinavir/ritonavir, each in combination with tenofovir and emtricitabine in ARV-naive HIV-1-infected subjects. The CASTLE study, 48-week results. In: Program and abstracts of the 15th Conference on Retroviruses and Opportunistic Infections; February 3-6, 2008; Boston, Mass. Abstract 37.
  2. Updated U.S. DHHS Guidelines for the Initial Treatment of HIV Infection. Guidelines for the use of antiretroviral agents in HIV-1-infected adults and adolescents By the Panel on Clinical Practices for Treatment of HIV Infection, January 29, 2008.
  3. Kempf DJ, King MS, Bernstein B, et al. Incidence of resistance in a double-blind study comparing lopinavir/ritonavir plus stavudine and lamivudine to nelfinavir plus stavudine and lamivudine. J Infect Dis. January 1, 2004;189(1):51-60.

See Also
More on HIV Medications
More News and Research on Atazanavir (Reyataz)

  • Email Email
  • Printable Single-Page Print-Friendly
  • Glossary Glossary

This article was provided by The Body PRO. Copyright © Body Health Resources Corporation. All rights reserved.

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.