The XVII International AIDS Conference (AIDS 2008)

Antiretroviral Strategy Update: Highlights From the XVII International AIDS Conference

August 13, 2008

Multimedia Toolbox

Listen to Audio (47 min.)

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

Table of Contents

Note: This CME/CE activity expired on August 13, 2009. For a list of currently available activities, click here.


The XVII International AIDS Conference [AIDS 2008] is wrapping up now, and in this last week, we heard a lot of presentations originating from all over the world. People come together at this conference for many reasons: There are those whose interest is treatment. There are those here who are interested in community advocacy. There are politicians. It's clearly the most important get-together dedicated to the HIV epidemic that occurs on any regular basis.

For me, as a clinician, it's a challenge to identify pieces of information that I can take back and use in my practice. Partly, that's because this conference is just so massive. As I said, there are people who come from all over the world to share their experiences, and finding the tidbits that are relevant to my own particular practice can be challenging in the midst of all of that.

However, this conference is also an opportunity to learn about life in other parts of the world, and also to understand how different people approach the HIV epidemic, whether it be from a prevention standpoint or from a therapeutic standpoint or from an advocacy standpoint.

What I'd like to do here is highlight what was really relevant for me -- thinking mostly about what I will take back to my practice, and how I will use what I learned here at the conference in my clinic -- and hopefully, I'll be able to convey some information to you that you too can use to help your patients and your management of people living with HIV infection.

Antiretroviral Trials

The topic that gets the most attention amongst clinicians is that of the treatment trials. There were several major treatment trials that were discussed here in Mexico City. One of the biggest headlines here has been the abacavir [ABC, Ziagen] story. That's because over the last year there have been at least two to three major clinical studies that have examined the potency and the safety of abacavir,1,2 and updates to the data from these studies were presented here.

More Patients With High Viral Loads Experience Virologic Failure When Starting Therapy With an Abacavir-Based Regimen Than a Tenofovir-Based Regimen

ACTG 5202: Study Design
Click to enlarge
Arguably one of the most important studies discussed here at Mexico City was AIDS Clinical Trials Group [ACTG] study 5202, presented by Paul Sax, M.D.3 It is a very large treatment trial, with more than 1,800 individuals, and it is currently ongoing. The study attempts to address what's the best medication to start with when treating HIV infection.

ACTG 5202 is somewhat complicated in its study design. There are four different study arms. Half of the participants in the study are randomized to receive tenofovir/emtricitabine [TDF/FTC, Truvada]; the other half is randomized to receive abacavir/lamivudine [ABC/3TC, Epzicom, Kivexa]. In addition, there's a second randomization to either efavirenz [EFV, Sustiva, Stocrin] or atazanavir [ATV, Reyataz] boosted with ritonavir [RTV, Norvir].

Study participants are blinded as to whether they're on tenofovir/emtricitabine or abacavir/lamivudine. They do know whether they're on efavirenz or ritonavir-boosted atazanavir.

The study is of treatment-naive patients who are 16 years of age or older and have a detectable viral load of more than 1,000 copies/mL. Study participants could have any CD4+ cell count.

ACTG 5202: NIAID Therapeutic DSMB Findings and Recommendations
Click to enlarge
As many people may be aware, in January of 2008, a data and safety monitoring board [DSMB] performed a routine review of the data from ACTG 5202 and found that there was concerning information regarding the potency of abacavir.2 What the DSMB at that time recommended to the study team was that there should be an unblinding of the study for those study participants who had a viral load of more than 100,000 copies/mL when they initiated the trial.

It was within this stratum that there was a difference noted between abacavir/lamivudine and tenofovir/emtricitabine, in terms of viral load suppression. The study participants in this stratum were unblinded to their nucleosides assignment, and were then able to decide whether to continue on their assignment, or switch to the alternative. For those who started the study with a viral load of less than 100,000 copies/mL, they remain blinded to their NRTI [nucleoside reverse transcriptase inhibitor]-treatment assignment.

What we heard at the conference was the first official presentation of the data that the DSMB reviewed.3 The study enrolled from 2005 to 2007. The follow-up is planned for 96 weeks after the last person is enrolled, which is expected to be in the fall of 2009.

ACTG 5202: Primary Study Endpoints
Click to enlarge
Interesting points that have to be made about this study are the way that the study was designed and the endpoints that the study uses. For efficacy, the time to virologic failure is used and, somewhat uniquely in this ACTG trial, the failure is broken up into two different categories: There's early failure and later failure. Early failure is defined as occuring between weeks 16 and 24, when there's a confirmed viral load of 100,000 copies/mL or greater. Later failure is defined as occuring 24 weeks and on, when there's confirmed virologic failure, as evidenced by a viral load of 200 copies/mL or greater.

Another co-endpoint is safety, for which the time to the first grade 3 or 4 sign or symptom, or a lab abnormality that's at least one grade higher than baseline, is used. Tolerability is also being assessed in ACTG 5202, and it is based on the time to modification of the original assignment that the study participant received.

ACTG 5202: Timing and Type of Virologic Failure
Click to enlarge
The results that were presented at this conference only address the 797 study participants who initiated the study with a viral load of 100,000 copies/mL or greater. They had a median follow-up of 60 weeks.

This cohort is very representative of HIV-infected patients who are starting therapy in the United States nowadays. The study participants had fairly high viral loads and low CD4+ cell counts, less than 200 cells/mm3. A quarter of the study participants already had a history of AIDS.

Importantly, genotype resistance testing was not mandated by the study at the time it was designed. However, about half of the patients in the study did have genotypic resistance data available, which were obtained as just a part of standard of care at the clinical sites where the study participants were enrolled.

The major motivation for the unblinding of the higher viral load stratum was that the time to virologic failure was shorter for those study participants who were assigned to abacavir/lamivudine compared to those who had been assigned to tenofovir/emtricitabine.

There were 57 virologic failures that occurred in the abacavir/lamivudine arm of the study, compared to 26 in the tenofovir/emtricitabine study arm. This was highly significant, with a P value of .0003.

When we look at the time-to-virologic-failure endpoint, there was clearly excess virologic failure with abacavir/lamivudine, whether we're talking about early failure or later failure.

If we look at the study participants who had a viral load of more than 200 copies/mL at 24 weeks or beyond (i.e., patients who previously had a viral load of less than 200 copies/mL), there were more persons in the abacavir/lamivudine study arm who experienced virologic failure compared to those in the tenofovir/emtricitabine study arm.

Reassuringly, study participants who achieved a confirmed viral load of less than 50 copies/mL did not experience virologic rebound on abacavir/lamivudine to any greater extent than what was seen with tenofovir/emtricitabine.

In addition, the time to regimen completion -- that means the time to the first virologic failure or the first time that the nucleosides had to be modified -- was also shorter for trial participants who had been assigned to abacavir/lamivudine, with 114 events in that group of patients, compared to 68 events in the tenofovir/emtricitabine study arm. Again, highly significant, with a P value of .001.

ACTG 5202: Proportion < 50 Copies/mL (95% CI)
Click to enlarge
A secondary analysis looked at the intent-to-treat proportion of study participants with a viral load of less than 50 copies/mL out to 96 weeks. One has to appreciate that, in this intent-to-treat analysis, switching does not equal failure, such that if you switch from abacavir/lamivudine to tenofovir/emtricitabine, you're still considered to be in the abacavir/lamivudine group. Which is why, when this was presented, we didn't see differences between abacavir/lamivudine and tenofovir/emtricitabine in terms of the proportion of study participants with a viral load of less than 50 copies/mL at week 48.

This is a confusing point for most people, because at week 48, even if 98% of the study participants on abacavir/lamivudine had already switched to tenofovir/emtricitabine, they'd still be considered abacavir/lamivudine-assigned patients. It's really important that people examine this very carefully. When looking at the secondary analysis of the proportion of study participants who had a viral load of less than 50 copies/mL at week 48, it's important to recognize that a significant proportion of these people on the abacavir/lamivudine arm were no longer on abacavir/lamivudine. That's an important point that I think sometimes gets missed.

There was concern that perhaps the abacavir hypersensitivity reaction [HSR] that occurs with abacavir in some proportion of patients might have been driving the shorter time to virologic failure or regimen completion in this study, but that does not seem to be the case.

Fortunately, there were not very many suspected drug hypersensitivity reactions. Actually, there was an identical number in each of the study groups: 27 study participants were suspected of having drug hypersensitivity reaction to abacavir in the abacavir/lamivudine group, and 27 study participants in the tenofovir/emtricitabine-assigned group also had a suspected HSR. Thus, it doesn't seem as if this was a major concern. Importantly, screening with HLA-B*5701 was not required in this study.

ACTG 5202: Grade 3 or 4 Safety Events Selected Events Among Those Reported in ≥ 5%
Click to enlarge
When looking at the safety endpoints, there was also a difference, and it did not favor abacavir/lamivudine. There seem to be more adverse events, in general, in the abacavir/lamivudine study arm. Again, this is not driven by abacavir hypersensitivity reaction. What we see is just a variety of different aches and pains that seem to be more concentrated in the abacavir/lamivudine group.

The Bottom Line

ACTG 5202: Conclusions
Click to enlarge
The take-home messages from this important study are that when looking at potency, and perhaps safety, there is a difference between abacavir/lamivudine and tenofovir/emtricitabine in patients who start antiretroviral therapy with a very high viral load.

I don't think that the unique ways in which the ACTG chooses their endpoints detracts from the differences that are being seen in this very large study. Those who quibble with the choice of ACTG endpoints have to recognize that there is a valid reason for why these endpoints are chosen. Oftentimes, the ACTG will not use a less-than-50 outcome due to the inherent variability in the assay, and also so as not to get fooled by blips. Thus, often what we see is less than 200 copies/mL used as a cutoff, as it was with the later failures in this trial.

It's important to recognize that these data are robust. There are more than 780 people involved in this analysis; and thus, it's hard to refute. There are those who have questioned these results, but I personally do think that they are real, and that we have to have some appreciation for these differences when we're treating our patients.

There are some caveats, of course. We don't yet understand the baseline resistance patterns, as resistance testing wasn't mandated as a part of this study. Thus, some of the differences in response may be due to perhaps a disproportionate number of individuals who had baseline resistance that would confer decreased susceptibility to abacavir/lamivudine. But again, we apparently didn't see this type of difference in people with lower viral loads at baseline, so I'm not sure how that would pan out. But that's something that I think would be important to look at.

Given these results, clinicians and patients are going to have to think hard about the role of abacavir, especially given other data that try to refute this.

Abacavir/Lamivudine Noninferior to Tenofovir/Emtricitabine in the HEAT Study

Kimberly Smith presented the results of the HEAT study, which was sponsored by the maker of abacavir.4 In this study -- which was much smaller than ACTG 5202 -- tenofovir/emtricitabine was again pitted against abacavir/lamivudine. This time, however, all study participants also received lopinavir/ritonavir [LPV/r, Kaletra].

A difference in potency was not seen in this study, whether in people with a high viral load or a low viral load. But the HEAT study was smaller in size compared to ACTG 5202 and had lopinavir/ritonavir on board, and it's possible that boosted PI [protease inhibitor] was able to overcome any differences, or mask any differences, between the nucleoside selections.

What was interesting during this conference, and what permeated at least the treatment aspect of it, was this he-said, he-said sort of back-and-forth between the ACTG study team and the investigators from the manufacturer of abacavir. Clearly, the maker of abacavir has good reason to be somewhat defensive, given recent findings regarding abacavir.

Virologic Suppression Rate of Abacavir/Lamivudine Across Six Trials Regardless of Baseline Viral Load

Responding to ACTG 5202: Clinical Trials With Third Drugs (N ≥ 100)
Click to enlarge
Thus, on the heels of the presentation of the ACTG data, was a presentation from investigators for the maker of abacavir, spelling out their experience with the drug in their treatment trials.5

What they did was clever, and very necessary. They went back into their database and looked at clinical trials of abacavir. What they tried to do was apply the same endpoint analysis that was utilized in the ACTG study to their data. What they wanted to answer was whether abacavir/lamivudine was less effective in patients who had a viral load of greater than 100,000 copies/mL in a review of six different trials, and also looking at the HEAT data that I mentioned previously.

The six different trials used different companion drugs. In some of the trials, efavirenz was used. In one trial, ritonavir-boosted atazanavir was used -- similar to ACTG 5202 -- but in others, lopinavir/ritonavir or fosamprenavir [FPV, Lexiva, Telzir] was the third companion drug. In addition, in some of these trials, abacavir was used once a day, as it was in ACTG 5202. In others, it was utilized twice a day.

Responding to ACTG 5202: 48-Week Protocol Defined Virologic Survival by Baseline VL Using A5202 Efficacy Endpoint
Click to enlarge
The bottom line from their presentation was that -- across the different treatment trials -- the virologic suppression rate using the ACTG 5202 endpoints was very, very, very high -- 89% to 95%. There didn't seem to be any difference at all between study participants who started therapy with a higher viral load of more than 100,000 copies/mL, versus those who initiated therapy with a viral load of less than 100,000 copies/mL. In fact, typically, in these studies, the difference between the two viral load strata was only 1 or 2 percentage points with regards to the proportion of study participants who achieved virologic success by achieving an undetectable viral load.

Likewise, the analysis highlighted the results from the HEAT study where, again, there was no major difference seen between study participants with viral loads greater than 100,000 copies/mL and those with viral loads less than 100,000 copies/mL, and abacavir/lamivudine and tenofovir/emtricitabine seemed to be very equivalent.4

The Bottom Line

ACTG 5202: Why Might These Results Differ From Other Studies?
Click to enlarge
What are we to make of these data? There are two different analyses that we're asked to consider. ACTG 5202 is a very large, randomized clinical trial;3 the other is an analysis of pooled data from six different clinical trials.5 I think that we have to look closely at the populations. There are clearly going to be differences between the population enrolled in one very recent trial versus the disparate populations enrolled over a longer period of time in different treatment trials sponsored by the manufacturer of abacavir. I think that we'd want to understand more -- and this topic came up in the question-and-answer period after the presentation -- about the population vis-?-vis the gender makeup, the racial makeup, CD4+ cell count, viral load, etc., and whether this could have influenced some of the differences we're seeing between these two analyses.

For myself, the cleaner way of approaching this is the way that the ACTG study is doing things. It's a prospective, randomized trial, and we have infinite respect for that study design. I can certainly understand the motivation for looking back at previously collected data in studies that were conducted by the manufacturer, but I think it's going to be difficult to override the findings of a large ACTG study.

In my clinical practice, the use of abacavir/lamivudine as an initial regimen is really, at this point, just an alternative option, especially for people who have higher viral loads.

We still have a lot to learn from ACTG 5202. Much of what we understand now is what was released by the DSMB.2 There are still a lot more data to parse.

We don't fully understand, for instance, the influence of efavirenz versus ritonavir-boosted atazanavir as a companion drug on the potency of these two different nucleoside combinations.

I think caution is probably wise at this point. Plus, I think that the attractiveness of tenofovir/emtricitabine as a fixed-dose combination is really hard to beat. So, at this point, we have to ask ourselves, given the data that we're seeing here, what would be the reason for using abacavir/lamivudine, especially in patients who have higher viral loads.

As is apparent to anyone who looks at the images that have emerged from ACTG 5202, the differences between the two nucleoside combinations are not subtle. It really does give pause to the enthusiastic use of abacavir, especially, again, in patients such as those who were studied in the ACTG 5202 analysis.

That said, it's important that abacavir continue to be an option; I think it's a good thing that abacavir is available for people who cannot, or will not, take tenofovir [TDF, Viread], for whatever reason. But again, I think we have to be open to the possibility that this drug is less potent, and perhaps less safe, than tenofovir.

A question that comes up all the time, though, is: "What do I do with my patient who has already started on abacavir, and is doing well?" This is a difficult question, and it's one that I will come back to after I discuss the data from the D:A:D study6 and the SMART study,7 which I'll talk about in a few minutes. But first, for most people -- if they are doing well -- there's a real reluctance to try to change their therapy. But it's important, and necessary, that we have a conversation with these patients about that data. I don't think we have to go into great detail, but they should be made aware that there is an indication that there are alternative drugs -- in this case, tenofovir/emtricitabine -- that might be a more potent option for them if their viral load was more than 100,000 copies/mL when they initiated antiretroviral therapy. In addition, there may be questions about the longer-term tolerability and safety of abacavir, especially when it comes to cardiovascular disease.

This is a conversation that our patients need to have. For patients at low risk for the tolerability problems I'll soon describe, and who have already proven themselves virologically by doing fantastically well on abacavir, there's less motivation to change. However, the discussion should still be had.

But that's not all, the abacavir story continues. It does seem like every other study presentation of interest at the conference had to do with abacavir. That in some way is unfortunate, because it may detract from other studies that were presented here that also merit some discussion.

Most people reading this should appreciate by now that an association between abacavir and myocardial infarction was found in a study called the D:A:D study,6 which is a very large collection of observational cohort studies. The study was published in April 2008 in The Lancet.6 The researchers found about a 90% increased risk of myocardial infarction among those study participants recently treated with abacavir, compared to patients who were treated with other nucleosides.

Review of 54 Trials Finds No Association Between Cardiovascular Disease Events and Abacavir

Responding to D:A:D/SMART: Methods
Click to enlarge
In a perhaps necessary, but somewhat defensive posture, the maker of abacavir presented data in a late breaker session to try to put into context some of the findings from that study.8 Again, the company went back to its collective data to try to further understand whether there is an association between its drug and the outcomes that were described in the D:A:D study.

In this case, the investigators went back to their repository and examined 54 clinical trials. This included 13 studies of adults who were randomized to abacavir versus control, 33 trials that included abacavir in the background regimen, and eight trials that did not include abacavir at all. This includes data from more than 14,600 study participants, with more than 12,000 person-years of follow-up. Importantly, almost all of the study participants were adults; 509 were children.

They looked back at their data for any report or description of coronary artery disorders or ischemic coronary artery disorders. Specific terms included coronary artery disease, or acute MI [myocardial infarction], angina; these were terms that they were able to search for within their database of collected data from these trials.

Responding to D:A:D/SMART: CV Outcomes - Exposure to ABC or No ABC (12 Adult Randomized Trials)
Click to enlarge
Importantly, these trials involved treatment-naive patients, as well as patients who were treatment experienced. The bottom line is, the investigators looked really hard and, fortunately, found that myocardial infarction and ischemic coronary artery disease were pretty rare. The rates were on the order of 2.04 to 2.36 per thousand person-years when looking at myocardial infarction, and a little bit higher, about 3.45 to 5.82 per thousand person-years, when looking at the broader category of ischemic coronary artery disease.

If anything, there did seem to be a little bit more instances of coronary artery disease among those study participants who were not receiving abacavir, compared to those who did receive abacavir. Thus, they found that when it came to cardiovascular disease events such as MIs or other coronary artery disorders, there was no significant difference at all between patients who were taking abacavir and those who were not taking abacavir.

Of course, the limitations of such an analysis are pretty obvious. For one thing, all the investigators are able to look at are events that have been reported, or detected. There may as well have been events that did not get reported. There was also a short duration of follow-up, compared to that of a longer observational cohort.

There could be other biases, of course, with regards to the study participants who are enrolled in these particular clinical trials. Theirs is a very different population than the large group that was studied in D:A:D, which consisted of more than 30,000 patients, who were followed longitudinally.6

In any case, these data are somewhat reassuring and, in my opinion, indicate that we're talking about a fairly rare event that's made slightly less rare with the use of abacavir.

Abacavir and Myocardial Infarction Link Confirmed in SMART Study

What makes it even harder to refute the D:A:D findings now, was another presentation, this one by the SMART study group, although the analyses were motivated by the investigators from the D:A:D cohort to try to validate their findings from that cohort in a very large clinical trial called the SMART study.7

SMART: Primary and Major Secondary End Points
Click to enlarge
The SMART study is one of the most important trials that has ever been conducted in the realm of HIV therapy. This was a large clinical trial of more than 5,000 HIV-infected individuals -- most of whom were on HIV therapy when they entered the study -- who were randomized to either continue on their HIV therapy, or stop their HIV therapy until their CD4+ cell count had fallen to 250 cells/mm3, at which point they would reinitiate their HIV therapy, and then stop treatment once again when their CD4+ cell count had risen to at least 350 cells/mm3.9

As most people reading this are aware, this study was stopped prematurely due to excess adverse outcomes. In patients who discontinued their therapy, there were excess events of death and morbidities, including cardiovascular disease, renal disease and hepatic disease, as well as AIDS-related events.

SMART: Comparison of Hazard Ratios for "ABC (No ddI)" and for "ddI (w/wo ABC)"  vs. "Other NRTIs"
Click to enlarge
What the investigators presenting the data at this conference did was look at the study participants who were on HIV therapy, and specifically analyze cardiovascular disease events occurring among those who were receiving abacavir versus those who were not receiving abacavir -- so, all the others.7

When you look at different definitions of cardiovascular disease (i.e., being more inclusive, being a little bit less inclusive, and as narrow as just myocardial infarction), the data were pretty consistent in that patients who were not on abacavir seemed to be more protected when it came to cardiovascular disease compared to those who were taking abacavir. In other words, those who were on abacavir had a greater risk of having cardiovascular disease and myocardial infarction compared to those who were not on abacavir.

During the presentation, the presenter, Jens Lundgren, also described results for didanosine [ddI, Videx], and they were not as robust as those from the analyses of abacavir. There did not seem to be as strong a signal, although there certainly was an effect, as well, for didanosine.

What wasn't demonstrated, but came out later in the question-and-answer period, was that tenofovir did not seem to have any problem at all as far as cardiovascular disease or myocardial infarction, in distinction, of course, to abacavir in this analysis.

SMART: Hazard Ratios for Using "ABC (No ddI)" vs. Using "Other NRTIs" According to CV Risk Status at Study Entry
Click to enlarge
Importantly, the excess cardiovascular events seemed to be really concentrated in patients who had five or more cardiovascular risk factors. In fact, for those who had fewer risk factors, there wasn't a significant effect of being on abacavir, as far as the development of cardiovascular disease. These are very consistent data with the D:A:D study, where those who were at moderate or high risk for cardiovascular disease seemed to really have the greatest impact from being on abacavir, as far as cardiovascular disease events, compared to those who had a lower risk of cardiovascular disease.6

Thus, the D:A:D data and the SMART data are very complementary, which leads to the question of whether there was some overlap in the populations. In fact, there was. Individuals in D:A:D could enroll into the SMART study. However, Dr. Lundgren indicated that more than 90% of the individuals who were in the SMART study were not in the D:A:D cohort, and that analysis, excluding those who were, did not change the findings.7

SMART: Adjusted Mean Differences in Biomarker Levels At Study Entry for Using "ABC (No ddI)" or "ddI (w/wo ABC)" vs. Using "Other NRTIs"
Click to enlarge
To try to understand a mechanistic explanation for why abacavir might be associated with cardiovascular disease, there was an analysis of cytokines and inflammation, looking specifically at IL-6 [interleukin-6]. IL-6 is a cytokine that had been associated previously in the SMART cohort with adverse outcomes, including mortality. The investigators also looked at high-sensitivity CRP (C-reactive protein), as well as D-dimer, amyloid P and amyloid A. Most significantly, study participants on abacavir, and not didanosine, had higher levels of IL-6, compared to the other nucleosides taken as a group; the same goes for high-sensitivity CRP. Thus, there may be more inflammation or more cytokine release, for some reason, that we're seeing in people who are on abacavir, compared to those on other nucleosides.

The Bottom Line

Where does this leave us? I think this is a difficult situation. Abacavir is a nice alternative to tenofovir, since not everyone can take tenofovir. As we discussed previously regarding potency, we have to be very cautious when using abacavir, given these data. There are some caveats and there is some instruction from these studies. In both the SMART study7 and the D:A:D study,6 the risk of cardiovascular disease was much higher among those who not only received abacavir, but who also had higher baseline cardiovascular risk factors. That's the group that I'd be most concerned about using abacavir. And regarding patients already on abacavir who have cardiovasular risk factors, that's another group that I'd also be concerned about. For these patients, who have significant risk factors for cardiovascular disease and who are doing well on abacavir, there's a real strong pull to talk to them about perhaps switching.

However, if these individuals also have mutable risk factors -- such as smoking -- things we can change, I think that it's really important to try to act on those as much as we can. If those are not things that we can change in these patients (eg., we can't control their blood pressure better, or we can't stop them from smoking), then these are the patients who I'm most worried about continuing on abacavir.

What we still need to understand more about is why this occurs. What's the mechanism behind abacavir's role in cardiovascular disease, if this does pan out? It's going to be hard to get much more conclusive data. There are not too many opportunities to look at an association between abacavir and cardiovascular disease in clinical trials. You would need a much larger clinical trial than those that we design now, and you would have to follow people for quite a long time to detect those events. However, that's not going to happen. There aren't too many observational cohort studies that can actually look at this. Perhaps the U.S. Veterans Administration or perhaps some HMO [health maintenance organization] with a large collection of patient data can do this. But I don't think we're going to get a whole lot more information that's going to help us understand what's going on here beyond what we have right now, unless we have a mechanism that's demonstrated pathogenically.

So there's an attempt to try to understand what's going on, and discover how this drug interacts with our immune system to create cytokines or other inflammatory chemicals that may propagate and promote atherosclerosis. Till that time, we have to be cautious. There is enough information now that we should make rational and evidence-based choices when selecting antiretroviral therapy, and we have to be, again, somewhat circumspect about using abacavir as a part of first-line therapy.

For me, abacavir is going to continue to be an alternative to tenofovir. I don't think that there's going to be much in the way of new data that's going to change that for quite some time, if ever.

Raltegravir Equivalent to Efavirenz at 96 Weeks in Treatment-Naive Patients

Protocol 004: Part II Design
Click to enlarge
But it wasn't all just about abacavir in Mexico City. There were other drugs that were worth discussing, and there was some new information that is going to be useful to those of us in the clinic.

Raltegravir [MK-0518, Isentress] has become a really exciting new drug that many people are looking at for a variety of different situations clinically. An interest that comes up pretty consistently when we discuss this drug is whether it can be used in treatment-naive patients. We have limited data on the use of raltegravir in treatment-naive patients. There is one protocol from which we get most of these data, if not almost all the data, and that's Protocol 004. This study has been presented previously,10 and an update on the outcomes in this study was presented here in Mexico City.11

This was a study of treatment-naive individuals who were randomized to receive raltegravir versus efavirenz in combination with tenofovir and lamivudine. Patients had to demonstrate susceptibility to efavirenz, tenofovir and lamivudine at baseline, and have a viral load of more than 5,000 copies/mL. They also had to have a CD4+ cell count of more than 100 cells/mm3.

Protocol 004: 96 Weeks Percent of Patients (95% CI) With HIV RNA < 50 Copies/mL [Non-Completer = Failure]
Click to enlarge
Twenty-four week12 and 48-week data10 had been presented previously. Here we heard about the 96-week data.11 It's a small study. This is not a large clinical trial; 160 people were randomized to raltegravir, 38 to efavirenz. What we saw early on that was quite remarkable -- and caused quite a bit of buzz when the data were first presented -- was that, during the first few weeks of the study, there was a dramatic decline in viral load with the different doses of raltegravir that were used compared to efavirenz. There was a hope or a promise that perhaps that quick viral load decay that occurred with raltegravir would lead to even more robust and potent virologic outcomes.

However, as far as viral suppression, the differences over time between efavirenz and raltegravir dissipated, and both drugs continued to be equivalent after 48 weeks. What we saw here is that they are almost identical, even up to 96 weeks, with 84% of individuals in both study arms -- those who received raltegravir versus those who received efavirenz -- achieving a viral load of less than 400 copies/mL.

When you look at the results for the number of study participants who achieved a viral load of less than 50 copies/mL, the results were almost identical to the number of study participants who achieved a viral load of less than 400 copies/mL: 83% and 84% of individuals, respectively, getting undetectable.

So clearly, raltegravir is a drug -- in this limited study, with a small number of patients -- that appears to be as potent, or nearly as potent, as efavirenz, a drug that is the most potent antiretroviral anchor that we currently have. CD4+ cell counts were exactly similar in the two study arms, again reassuring us that this is going to be a drug that has a future, perhaps as a therapy for individuals who are treatment naive.

Protocol 004: Treatment-Emergent Mutations
Click to enlarge
There were some failures, of course, in the study. The good news was that we did not see the emergence of widespread raltegravir resistance. There's been concern that there's a lower genetic barrier for raltegravir than protease inhibitors, and that resistance can be acquired relatively easily. But again, in this treatment-naive study, that was not a major concern.

Percent of Patients With Grade 3 / 4 Laboratory Abnormalities
Click to enlarge
There were some adverse events, and they are notable. What was seen in the study, of course, were the central nervous system problems that are associated with efavirenz and that were concentrated in the efavirenz arms of the study. However, there were more CPK [creatine kinase] elevations in the raltegravir arm: 6.3% versus 2.6% in the efavirenz arm. Ten individuals had a high, grade 3 or grade 4, elevation of CPK in the raltegravir arm. Elevations of CPK have also been described in patients receiving raltegravir in the BENCHMRK studies,13 which looked at treatment-experienced patients, and is a cause for concern.

Elevated lipase levels also were seen in the treatment studies looking at treatment-experienced patients. Here there was, again, a slight difference between the two arms,11 although this is a small study, and so we should be cautious about over-interpreting the cases of the two patients in the raltegravir arm who had an elevated lipase level, versus none in the efavirenz arm.

The Bottom Line

In summary, I think that this is an important first step in understanding a role for raltegravir in individuals who are treatment naive. Further studies will emerge that will look at this in a much more rigorous way.

Efavirenz Virologically Superior to Lopinavir/Ritonavir in Mexican Head-to-Head Trial of Treatment-Naive Patients

EFV vs. LPV/r: Study Design
Click to enlarge
Another study that is worth mentioning that involves drugs that are approved for HIV therapy is what's being lovingly called the Mexican A5142 study.14 Now, remember: A5142 was a nice study done by the ACTG.15 It was the 5202 study3 of its time. It looked at efavirenz versus lopinavir/ritonavir as initial therapy for HIV.15 It was really a blockbuster of a clinical trial that helped us understand the relative merits and disadvantages of efavirenz versus lopinavir/ritonavir.

Well, this trial was a Mexican version. What J. Sierra-Madero et al did -- in a more compact and more limited way -- was to look at efavirenz versus lopinavir/ritonavir, this time in combination with zidovudine/lamivudine [AZT/3TC, Combivir].14 What they also wanted to do was look at patients with a little bit more advanced disease, those who had a CD4+ cell count of less than 200 cells/mm3.

This is a study that interestingly enrolled people in different parts of Mexico, including people living in fairly rural areas. So it really is an important study, especially in Mexico and other places where these therapies might be used that are not super cosmopolitan or, let's say, resource rich.

EFV vs. LPV/r: Patient Disposition At Week 48
Click to enlarge
What I think is interesting about this study is that it really confirmed what ACTG 5142 found, namely that efavirenz beats anything that's compared to it. There is no trial as of yet that has ever found a therapy that has beaten efavirenz head-to-head in treatment-naive patients. Here again, we see that efavirenz had better virologic response rates, compared to lopinavir/ritonavir.

TMC278?C204: Potent Antiviral Efficacy at Week 48 Sustained to Week 96
Click to enlarge
The good news for lopinavir/ritonavir is that the differences were not significant when looking at who achieved a viral load of less than 400 copies/mL. It was only when looking at the less than 50 copies/mL viral load endpoint that a significant difference was observed. Thus, maybe some of the virologic efficacy deficit was made up of those study participants who had low-level viremia. Again, this study looked at the use of soft-gel capsules of lopinavir/ritonavir that, in many settings, are hard to use, especially, I think, in rural Mexico.

EFV vs. LPV/r: Change of Median CD4+ Cell Counts
Click to enlarge
Nonetheless, CD4+ cell counts were found to be higher in those study participants who were on lopinavir/ritonavir, just as was demonstrated in the larger U.S. study (ACTG 5142). So these data are important in that they validate what was really, if we recall, a pretty surprising result when the ACTG 5142 study first came out in 2006. This is an important piece of data, and certainly confirms mostly what we saw in that larger study.

TMC278 Advances in Development

TMC278?C204: Study Design
Click to enlarge
As far as studies of new drugs that are not yet approved, the one study that I'd like to highlight is a study of rilpivirine -- that's TMC278.

As most people probably remember, this is a new once-a-day non-nucleoside reverse transcriptase inhibitor [NNRTI], so it has some hope of perhaps being a new and improved efavirenz. The data that were presented here focused on a dose ranging clinical trial that was done.16 It's the TMC278-C204 study that was sponsored by the manufacturer of this particular compound and, like most other studies, pits itself against efavirenz.

There were three different doses of TMC278: a 150-mg dose, a 75-mg dose and a 25-mg dose, all compared against efavirenz taken with two nucleosides. All 368 patients were treatment naive. All had to have a viral load of more than 5,000 copies/mL.

TMC278?C204: High Response Rate and Sustained Virologic Response Over 96 Weeks Similar to EFV
Click to enlarge
When looking out to 96 weeks, as far as virologic efficacy, getting to less than 50 copies/mL by an intent-to-treat TLOVR [time to loss of virologic response] analysis, we see that the differences between the different treatment arms are fairly minimal. Even the lower dose 25-mg/day TMC278 arm had a response of 76%, compared to the efavirenz arm, which was around 71%. Thus, we're not seeing any big differences between these different study arms -- different doses of TMC278 -- or between efavirenz and TMC278.

TMC278?C204: Potent Antiviral Efficacy at Week 48 Sustained to Week 96
Click to enlarge
Even when looking at pure virologic failure, there weren't any significant differences between the different study arms. In the efavirenz arm, there were five pure virologic failures by week 48. It was up to seven at week 96. In the TMC278 groups, the number of failures ranged from six to 10, when combining them altogether. Again, we're seeing that this is a potent non-nucleoside in treatment-naive patients.

Much of the attention during the presentation focused on the 25-mg dose. Previous to this conference, it was announced [in an advisary note to TMC278 study investigators from the agent's maker] that there were problems with EKG [electrocardiogram] changes among patients who were taking TMC278 at the 75-mg and 150-mg doses. Thus, the manufacturer of the drug had gone back to look at the 25-mg dose, and has now launched clinical trials of the drug using a lower 25-mg dose.17 That dose was found to be less involved in any changes of the EKG,16 specifically, a prolongation of the QTc interval had been detected for TMC278 during earlier studies.

When looking at the 25-mg dose of TMC278, there were eight virologic failures. Unfortunately, the presenter didn't detail the resistance mutations that might have developed in these nine individuals who developed resistance on the 25-mg dose of TMC278.

TMC278?C204: Doses Were Safe and Well Tolerated With No Consistent Association Between Safety Assessments and TMC278 Dose
Click to enlarge
TMC278?C204: Additional Investigations
Click to enlarge
In addition, there was some confusion about the way in which the adverse events were described. It wasn't clear whether all the particular adverse events associated with TMC278 were accounted for in the presentation.

In addition, during this discussion, there was a simple bullet point regarding increases in the QTc interval that were seen with TMC278 and with efavirenz. However, there were no data presented, and I think that was unfortunate.

Overall, however, I think that this was a positive study that indicated that TMC278, even at a lower dose, seemed to go toe-to-toe with efavirenz, and that further study at the 25-mg dose of the drug, in larger clinical trials, was warranted. However, I think the presentation does leave some gaps in our understanding regarding the resistance that develops to this drug, as well as our understanding about the potential toxicity of the agent. Hopefully, that will be reconciled in later presentations at other conferences.

Interim Results From the GRACE Study

GRACE: Study Design and Baseline Characteristics
Click to enlarge
The last treatment study I want to mention is the GRACE [Gender, Race and Clinical Experience] study, presented by Judith Currier.18 This study is pretty important. It looks at the use of antiretrovirals in combination for individuals who are treatment experienced. Patients did not have to be deeply treatment experienced; it could be a first failure.

Specifically, the study is an open-label, phase 3b trial designed to assess the gender and racial differences with regards to tolerability to ritonavir-boosted darunavir [TMC114, Prezista] at 600 mg and 100 mg BID [twice a day], plus an optimized background regimen that could include nucleosides and/or non-nucleosides, over the course of a year of therapy. Study participants could also be on etravirine [TMC125, Intelence].

At this conference, the 24-week interim analysis was presented. This involved 203 patients, out of the total of 429 individuals. This study, again, is directed towards trying to understand more about the utility of darunavir in combination with an investigator-selected background regimen in African Americans and women. Thus, 65% of the persons who were analyzed here are African American; 22% are Latino.

GRACE: 24-Week Interim Analysis: Efficacy and Safety
Click to enlarge
The virologic responses of the female study participants, compared to that for all study participants, were presented. Looking at a viral load of less than 50 copies/mL, in an intent-to-treat TLOVR analysis, found that 50.6% of female study participants achieved virologic suppression to less than 50 copies/mL, compared to 51.2% of all study participants. There were identical rates of suppression among female study participants. Likewise, we saw an increase in CD4+ cell counts in female study participants of around 86 cells/mm3, compared to 83 cells/mm3 for all study participants taken in total. Thus, we're seeing similar sorts of responses immunologically in female study participants compared to the other study participants.

This is a nice analysis, and it's a good effort at trying to better understand how these drugs work in the populations that are likely to be exposed to them. So I think that this is reassuring.

The overall rate of response was only 50%, which could be seen as glass half-full, or glass half-empty. I think this could be a more difficult-to-treat population. It's hard to compare this study to other studies that had different populations and different success rates. We'll see what happens as the study progresses, but I think it's a very good first step towards understanding that this is totally appropriate therapy for this group, and that it seemed to be well tolerated and efficacious.


We had a slew of different studies: studies of old drugs, studies of newer drugs and studies of drugs that haven't even come out yet. What are we to make of all of this?

At this conference there was an overlay, as well, of drama. We had the ACTG 5202 data.3 We had rebuttals from the manufacturer of the medication. Then, on the other hand, we had the D:A:D data6 and the SMART data,7 and rebuttals8 to that information. So there was all this back and forth, and I think there was concern that we're at loggerheads here.

I don't think that's the case. I think that the system is working. We are doing investigations. We find things. We try to validate them to the best of our ability. That's what research is about. We can't answer every question. I know that's frustrating to our patients. It's frustrating to us as clinicians. After this conference, people are probably going to be saying, "Well, what am I to do?"

What I think we need to do, and what we have to tell our patients to understand is that, we need to look at the data and make our own decisions. When ACTG 5142 came out at the last International AIDS Conference in Toronto15 and demonstrated that efavirenz was better virologically than lopinavir/ritonavir, but lopinavir/ritonavir was better immunologically than efavirenz, together with the resistance story and the lipoatrophy story, there was no bottom line, no single easy one-liner, no take-home message for people to understand. People had to look at the data and decide for themselves what was important. We have to do the same thing with the data we're receiving now. These data will be published soon. These analyses were done very rapidly. What we're going to have to decide is, for our individual patient, what is best for him or her.

Should patients use abacavir? Should they not use abacavir? Should they use any of the other drugs that are being studied in the different studies that I talked about? Should they not? What's the role of the newer drugs? How much information does it take to change your practice?

We're all different in that way. We have to take these messages, digest them and then translate them for our clinical practices in a way that we think is best for our patients, and that's all a conference can provide us. I think this conference, at least from the therapeutic side of it, was successful, even with the limited number of presentations that were available here.

I think this is good to know. I think each of us should look carefully at the data that were presented here and, again, make our own decisions. I'm sure that our patients will benefit from that.

This transcript has been lightly edited for clarity.


  1. Sabin C, Worm S, Weber R, et al, and the D:A:D Study Group. Do thymidine analogues, abacavir, didanosine and lamivudine contribute to the risk of myocardial infarction? The D:A:D Study. In: Program and abstracts of the 15th Conference on Retroviruses and Opportunistic Infections; February 3-6, 2008; Boston, Mass. Abstract 957c.
  2. NIAID modifies HIV antiretroviral treatment study: combination therapy that includes ABC/3TC found less effective in subgroup of antiretroviral-naive individuals. Bethesda, Md: National Institute of Allergy and Infectious Diseases; February 28, 2008.
  3. Sax P, Tierney C, Collier A, et al. ACTG 5202: shorter time to virologic failure (VF) with abacavir/lamivudine (ABC/3TC) than tenofovir/emtricitabine (TDF/FTC) as part of combination therapy in treatment-naive subjects with screening HIV RNA ≥ 100,000 c/mL. In: Program and abstracts of the XVII International AIDS Conference; August 3-8, 2008; Mexico City, Mexico. Abstract THAB0303.
  4. Smith KY, Fine DM, Patel P, et al. Similarity in efficacy and safety of abacavir/lamivudine (ABC/3TC) compared to tenofovir/emtricitabine (TDF/FTC) in combination with QD lopinavir/ritonavir (LPV/r) over 96 weeks in the HEAT Study. In: Program and abstracts of the XVII International AIDS Conference; August 3-8, 2008; Mexico City, Mexico. Abstract LBPE1138.
    View poster: Downolad PDF
    View slides: Download PDF
  5. Pappa K, Hernandez J, Ha B, Shaefer M, Brothers C, Liao Q. Abacavir/lamivudine (ABC/3TC) shows robust virologic responses in ART-naive patients for baseline (BL) viral loads (VL) of ≥ 100,000c/mL and <100,000c/mL by endpoint used in ACTG5202. In: Program and abstracts of the XVII International AIDS Conference; August 3-8, 2008; Mexico City, Mexico. Abstract THAB0304.
    View slides: Download PDF
  6. D:A:D Study Group, Sabin CA, Worm SW, et al. Use of nucleoside reverse transcriptase inhibitors and risk of myocardial infarction in HIV-infected patients enrolled in the D:A:D study: a multi-cohort collaboration. Lancet. April 26, 2008;371(9622):1417-1426.
  7. Lundgren J, Neuhaus J, Babiker A, et al, and the SMART/INSIGHT and D:A:D Study Groups. Use of nucleoside reverse transcriptase inhibitors and risk of myocardial infarction in HIV-infected patients enrolled in the SMART study. In: Program and abstracts of the XVII International AIDS Conference; August 3-8, 2008; Mexico City, Mexico. Abstract THAB0305.
    View slides: Download PDF
  8. Cutrell A, Hernandez J, Yeo J, Brothers C, Burkle W, Spreen W. Is abacavir (ABC)-containing combination antiretroviral therapy (CART) associated with myocardial infarction (MI)? No association identified in pooled summary of 54 clinical trials. In: Program and abstracts of the XVII International AIDS Conference; August 3-8, 2008; Mexico City, Mexico. Abstract WEAB0106.
    View slides: Download PowerPoint
  9. Strategies for Management of Antiretroviral Therapy (SMART) Study Group, El-Sadr WM, Lundgren JD, et al. CD4+ count-guided interruption of antiretroviral treatment. N Engl J Med. November 30, 2006;355(22):2283-2296.
  10. Markowitz M, Nguyen B-Y, Gotuzzo E, et al, and the Protocol 004 Part II Study Team. Rapid onset and durable antiretroviral effect of raltegravir (MK-0518), a novel HIV-1 integrase inhibitor, as part of combination ART in treatment HIV-1 infected patients: 48-week data. In: Program and abstracts of the 4th International AIDS Society Conference on HIV Pathogenesis, Treatment and Prevention; July 22-25, 2007; Sydney, Australia. Abstract TUAB104.
    View slides: Download PowerPoint
  11. Markowitz M, Nguyen B-Y, Gotuzzo E, et al, and the Protocol 004 Part II Study Team. Sustained antiretroviral efficacy of raltegravir as part of combination ART in treatment-naive HIV-1 infected patients: 96-week data. In: Program and abstracts of the XVII International AIDS Conference; August 3-8, 2008; Mexico City, Mexico. Abstract TUAB0102.
    View slides: Download PowerPoint
  12. Markowitz M, Nguyen B-Y, Gotuzzo F, et al. Potent antiretroviral effect of MK-0518, a novel HIV-1 integrase inhibitor, as part of combination ART in treatment-naive HIV-1 infected patients. In: Program and abstracts of the XVI International AIDS Conference; August 13-18, 2006; Toronto, Canada. Abstract THLB0214.
    View slides: Download PowerPoint
  13. Steigbigel RT, Cooper DA, Kumar PN, et al, for the BENCHMRK Study Teams. Raltegravir with optimized background therapy for resistant HIV-1 infection. N Engl J Med. July 24, 2008;359(4):339-354.
  14. Sierra-Madero J, Villasis A, Mendez P, et al. A prospective, randomized, open label trial of efavirenz versus lopinavir/ritonavir based HAART among antiretroviral therapy naive, HIV-infected individuals presenting for care with CD4 cell counts <200/mm3 in Mexico. In: Program and abstracts of the XVII International AIDS Conference; August 3-8, 2008; Mexico City, Mexico. Abstract TUAB0104.
    View slides: Download PowerPoint
  15. Riddler SA, Haubrich R, DiRienzo G, et al, and the AIDS Clinical Trials Group 5142 Study Team. A prospective, randomized, phase III trial of NRTI-, PI-, and NNRTI-sparing regimens for initial treatment of HIV-1 infection -- ACTG 5142. In: Program and abstracts of the XVI International AIDS Conference; August 13-18, 2006; Toronto, Canada. Abstract THLB0204.
    View slides: Download PowerPoint
  16. Santoscoy M, Cahn P, Gonsalez C, et al. TMC278 (rilpivirine), a next-generation NNRTI, demonstrates long-term efficacy and tolerability in ARV-naive patients: 96-week results of study C204. In: Program and abstracts of the XVII International AIDS Conference; August 3-8, 2008; Mexico City, Mexico. Abstract TUAB0103.
    View slides: Download PowerPoint
  17. TMC278-TiDP6-C215: A clinical trial in treatment naive HIV-subjects patients comparing TMC278 to efavirenz in combination with 2 nucleoside/nucleotide reverse transcriptase inhibitors. Accessed August 9, 2008.
  18. Currier J, Squires K, Averitt Bridge D, et al, on behalf of the GRACE Study Group. Safety, tolerability and efficacy of darunavir/ritonavir in treatment-experienced women with HIV infection: 24-week interim analysis of GRACE (Gender, Race, And Clinical Experience). In: Program and abstracts of the XVII International AIDS Conference; August 3-8, 2008; Mexico City, Mexico. Abstract THPDB202.
    View poster: Download PDF
    View slides: Download PowerPoint

Copyright © 2008 Body Health Resources Corporation. All rights reserved.

This article was provided by It is a part of the publication The XVII International AIDS Conference.

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.

The content on this page is free of advertiser influence and was produced by our editorial team. See our content and advertising policies.