February 11, 2008
Listen to Audio (42 min.)
Please note: These files can be quite large. Allow some time for them to download.
There are a lot of conferences now that are dedicated to specific issues within HIV, including metabolic complications. Some of the metabolic-related material gets diluted a little bit across the different conferences. Thus, it's not too surprising that we don't have any earth-shattering data regarding metabolic complications, especially about body shape changes, that were presented at this year's Conference on Retroviruses and Opportunistic Infections [CROI]. There are some important data, which I'll mention in a little bit, about cardiovascular disease risk that I think was really one of the top stories of this conference.
|Click to enlarge|
I'm going to start with body shape changes, and talk a little bit about a study that looked at growth hormone.1 Before you roll your eyes -- oh no, not another growth hormone study -- this one's a little bit different. I think that this study demonstrates the way that growth hormone studies should have been done. In fact, some people who really knew a lot about this subject had been advocating for it to be done this way, but it just hadn't happened until now.
Specifically, I'm talking about a very nice study presented by Steve Grinspoon in which he used growth hormone, although he didn't use it in the really high doses that have been explored in previous studies2 among HIV-infected patients. He used physiologic doses.
|Click to enlarge|
Steve is a really smart endocrinologist, and endocrinologists like to use hormones in doses that replicate normal cycles that occur in the body. So Steve did what I think he's been dreaming of doing for a long time -- he used low doses of growth hormone. We're talking about 2 to 6 µg/kg per day. These are the kinds of doses of growth hormone that are given to people who are deficient in growth hormone in order to replicate and mimic what would be a natural growth hormone level in the body.
Steve conducted an 18-month, double-blind, placebo-controlled study of this strategy. First, people were screened to see whether they had some evidence of growth hormone deficiency. Then, using a sophisticated test, called the growth hormone-arginine stimulation test, that endocrinologists who know what they're doing can use, he was able to select patients who really could benefit from growth hormone replacement and study them over time.
It's a really interesting study design that is a little bit different than what we've seen before. The participants were adults. They had to be HIV infected and on stable antiretroviral therapy. To get in the study, patients had to have a waist-to-hip ratio that was .9 for men and .85 for women, or greater, which are pretty standard. Patients also had to have evidence of increased abdominal girth, or loss of fat in the extremities, or loss of fat in the face. Again, growth hormone response, given the arginine testing, had to be low.
There were exclusion criteria. Patients with diabetes were excluded since growth hormone can make diabetes worse. Patients couldn't be pregnant. They couldn't have carpal tunnel syndrome, because it also could be aggravated by this hormone.
There were 65 patients who were randomized: 72 to growth hormone and 29 to placebo. At the end of the day, 21 completed the protocol in the growth hormone arm, versus 27 in the placebo arm. So, again, not a very large study, but still an important one.
When assessing body shape using imaging techniques, we see that over the time period of 18 months, in the placebo arm, there was no change in visceral adiposity. However, among patients who received growth hormone at the low doses that were described previously, we do see a decrease over time. It ended up being an approximately 9% decrease in visceral adiposity over the course of the study.
In addition, there were no major changes in extremity fat and that has been a concern, because growth hormone is lipolytic, meaning it can lyse adipocytes. It doesn't care if the adipocyte is located in the visceral abdomen or the face or the arms. At low doses, we didn't see much of a problem with growth hormone wasting fat in the periphery.
Lean mass did seem to increase a little bit, and significantly. That's a positive thing, because lean body mass improvements have been associated with improved health.
Lipid profiles were not a concern with growth hormone in this study at these low doses. Triglycerides actually decreased in people who got on the growth hormone, compared to in those who received the placebo, where we saw basically no change in triglycerides. No effects were seen on other lipoprotein profiles.
There was not really any change seen in terms of carotid intima-media thickness with growth hormone, despite the lipid changes that I talked about previously. One explanation for that may be that -- despite using near physiologic growth hormone -- glucose tolerance did worsen with growth hormone. This was a surprising finding, because if we really were replicating physiologic doses of growth hormone, one wouldn't expect that there would be a problem with glucose tolerance. But there was a significant difference in glucose tolerance over the course of the study, with patients who were taking the growth hormone having worse glucose tolerance, and those who were on placebo having no change. This may have something to do with the administration of growth hormone, since it is not being secreted in little amounts over a period of time and in cycles. The administration of the drug -- I.V. or, in this case, subcutaneously -- may have some effect on glucose levels that are poorly explained.
Otherwise, adverse effects were fairly minimal. The adverse effects that have plagued this drug in other trials, including the carpal tunnel syndrome, the arthralgias, and the edema, really weren't seen here. Thus, low doses of growth hormone over a long period of time were well tolerated and led to modest decreases in visceral adiposity.
When you compare this to other studies of growth hormone,2 not surprisingly, you see more bang for the buck when you give higher doses, but there is also more toxicity. In other studies, a higher 4 mg per day dose of growth hormone resulted in a 20% or more decrease in visceral adiposity. In this study, with much more modest doses averaging around .3 mg per day over a longer period of time, there was about a 9% to 10% decrease in visceral adiposity. So there's a slower loss of visceral adipose tissue, but at the same time, there's better tolerability.
Growth hormone releasing factor, or growth hormone releasing hormone, is another compound that's under study. This looks to be a very exciting drug. There was nothing major presented on that drug at this conference, but at previous conferences, it looked very promising.3 It also can produce decreases in visceral adiposity, approaching 15% or more, and seems to be better tolerated than higher dose growth hormone.
At the end of the day, whether we're going to be using low-dose growth hormone or some alternative is still unclear. The role for low-dose growth hormone is not very certain at this point, but what this study points out is that we've basically been overdosing people with growth hormone in these previous studies.
If I was forced to use growth hormone for lack of an alternative, I would probably use a low dose of growth hormone, as was done in this study, and I would do so for a longer period of time. It may be that another 18 months of study beyond the 18 months that were looked at here would produce even further gains in visceral adipose tissue.
So I think that this is an option for some folks, but not something that most of us are going to be able to do. Of course, lower doses of growth hormone are less expensive than higher doses and this may make its use a little bit more palatable.
Click to enlarge|
For additional images from this study, click here.
Another study at CROI that was interesting, in terms of body shape change, is an analysis that was done by Spanish investigators looking at just how much loss of fat patients have to experience in order to perceive the loss of this fat.4
In many of the studies that have been conducted of late, a 20% loss in extremity fat has been considered a significant loss. Many people justify this by saying that that's a clinically significant amount of fat loss. However, there's really no data to indicate that that is a clinically significant amount of fat loss. It is sort of a line in the sand that has been developed somewhat arbitrarily.
In this one analysis, loss of fat in the extremities by DEXA scan was correlated to perceptions of fat loss among people who had experienced some of that loss. What these investigators found was that people started to notice a change when they had experienced fat loss in the extremities of around 30%.
This is useful to know. It helps us to understand what the threshold is before fat loss becomes really clinically evident. In most of the studies that we're doing, we don't want it to get that bad. Thus, using a 20% fat loss as a line to not be crossed, or a line to compare across study arms, does still have some clinical utility.
It is interesting to know that, when you think about patients and what they perceive, it's going to be even greater fat losses that are noticeable to them and others. So I think this is another little presentation that helps us understand a little bit more about fat loss.
Lastly, in this category of body shape changes, there was an interesting poster looking at supplementation with chromium. Chromium is a micronutrient that purportedly has activity to improve some insulin parameters, glucose metabolism parameters, and maybe some other things.5 It was thought that, if this was given to people who have HIV infection, it might reduce visceral adiposity and it might improve lean fat loss, and also glucose and insulin parameters.
Click to enlarge|
For additional images from this study, click here.
To make a pretty short story even shorter, in a smaller study from Canada, the supplement seemed to have some activity.6 The poster was a little unclear about some of the measurements in terms of the degree of decline in fat, but there seemed to be at least a trend for fat loss in the abdomen by DEXA scan. DEXA scan would look at both subcutaneous and visceral fat, and not be able to distinguish the two. Chromium seemed to improve glycemic parameters, as well.
Chromium may be worthy of future study, perhaps in a more rigorous way, with larger numbers. Perhaps also utilizing a standardized compound. There was not much mentioned about the type of chromium that was used, nor what quality assurance went into the use of this particular supplement. So, while I think it is intriguing, it is not ready for prime time; it is something that we could probably explore in another way.
Let's look now at cardiovascular issues. There were a number of presentations at this conference that were, at the very least, intriguing, and at most, provocative. The most provocative, and I think the presentation that generated the most buzz at CROI, was an interesting report from the D:A:D cohort. This is a very interesting study and an interesting group of investigators. As many of you are aware, D:A:D is a meta-cohort of 11 different cohorts across Europe and North America that combine their data prospectively to examine outcomes among people who have HIV infection, largely people who are under care and taking HIV therapies.
In this particular analysis, the researchers were trying to understand the risk of myocardial infarction and nucleoside therapy. This group has published quite extensively about the risk for myocardial infarction and stroke amongst their cohort, a very robust cohort of more than 25,000 people. At the time of this particular analysis, 33,000 individuals had been enrolled in this particular cohort. The researchers published in The New England Journal of Medicine recently that protease inhibitor exposure was linked to increased risk of myocardial infarction,7 and that there had been a trend for NNRTI [non-nucleoside reverse transcriptase inhibitor] exposure to also lead to increased risk of myocardial infarction at different time points, but not at the most recent time point.
Click to enlarge|
For additional images from this study, click here.
What the researchers wanted to do was also look at the nucleosides. So they did an analysis under the hypothesis that the thymidine analogs of stavudine [d4T, Zerit] and zidovudine [AZT, Retrovir] would likely be linked to myocardial infarction.8
Lo and behold, what did they find? They found that it was actually abacavir [ABC, Ziagen] and, to a lesser extent, didanosine [ddI, Videx] that was linked to myocardial infarction in their cohort.
Interestingly, they've done a number of different analyses to try to understand this and make sure it wasn't a mistake. Even after double-checking their numbers, the same relationship comes out.
When you look at the people who did get abacavir, and those who didn't, some things strike you right away. One is that the cardiovascular risk profile of people who were not on abacavir was a little bit better than that of those on abacavir. This may suggest something called "channeling bias." That's where you channel people to a particular therapy based upon your perceptions about their risk for an adverse event or some other outcome. In this case, you might say to someone who has cardiovascular risks, "I'm going to put you on abacavir, because I don't want to put you on stavudine. Maybe I don't even want to put you on zidovudine." In this way, you would concentrate people at higher risk for cardiovascular disease in your abacavir arm.
The D:A:D investigators totally appreciated that, and while controlling for a number of different cardiovascular disease risk factors, tried to adjust for that channeling bias in their analysis. What they were able to find is that the association between abacavir and myocardial infarction still persisted. Even when controlling for a number of different factors, including hypertension, dyslipidemia, coronary heart disease risk -- as predicted by the Framingham cohort, etc. -- there's this persistent effect of abacavir being linked to the risk of myocardial infarction.
The relative risk here is on the order of about 1.6. This is an increased risk that's statistically significant, even after trying to control for confounders.
Other interesting analyses were also done by the D:A:D investigators, including looking at exposure time, because perhaps it's the use of abacavir that's a marker for longer term NNRTI exposure. What they found is that it wasn't cumulative or past exposure to abacavir that made the biggest impact; it was more recent use of abacavir.
Furthermore, when abacavir was discontinued, the risk of myocardial infarction -- that excess risk -- dissipated, suggesting that this effect existed during treatment with the drug and went away when the drug was no longer taken.
One can hypothesize that abacavir could have some direct toxicity on cardiovascular disease. Stroke was not increased in this particular analysis; it was only myocardial infarction. A cumulative effect doesn't seem to be likely. Didanosine, as I mentioned earlier, also had an effect, but it was much more modest.
This study has generated, as you can imagine, a lot of concern. Abacavir is now becoming increasingly popular. We can now determine who is at risk for developing abacavir hypersensitivity reaction with the use of HLA-B*5701 testing. This test has opened up a new venue for abacavir where it didn't have one before. Furthermore, the U.S. Department of Health and Human Services has recently upgraded abacavir and lamivudine [3TC, Epivir] to a preferred NNRTI status.9
What does this all mean? I think it's important to recognize a few things. This is a cohort study. This is not a randomized clinical trial. The D:A:D investigators even did an unusual thing by issuing a statement, almost like a press release, along with their poster, indicating that we should be very cautious about interpreting these data,10 and I think that was very wise of them.
Cohort data are not the same thing as clinical trial data. A rigorous clinical trial can maybe help tease this out much better than a cohort study, which inherently has to be subjected to increased bias, compared to a randomized study that tries to negate such bias.
I also think that there's no underlying mechanism that's been postulated yet. So we have to be very cautious about interpreting exactly what's going on here.
Certainly others will go into their databases and look at abacavir exposure, myocardial infarction and other cardiovascular disease risks. The problem is, there's probably not too many cohorts that have the power that this one does, with as many outcomes. In this case, there were 33,000 individuals in their cohort and over 500 individuals had had myocardial infarctions.
One thing that I think also has to be explored is whether there are any other confounders that were not corrected for. You can try to adjust for as many of these confounders as possible, but you can't get all of them. It's just the ones you think are most likely to be operative. Could there be other things that are going on? Was there a preferential statin used in those not on abacavir? Were more hypertension medicines used in those not on abacavir? These are things that I think have to be teased out, but this is the start of, I think, an investigation that is going to need to be done. Hopefully, we'll have a better sense of this not too long from now.
Importantly in this study, the risk was pretty low for those who had minimal cardiovascular disease risk factors. Taking abacavir slightly increased a patient's risk of having a myocardial infarction, but the relative risk, while there, was nothing like the absolute risk. The absolute risk remained fairly low. It's those people who did have high risk of cardiovascular disease, based upon the Framingham calculation, who seemed to have the biggest impact from having abacavir added to their regimen.
At this point, this is very exploratory. I'm not sure what most people are going to do with it. If I were to look at a patient who had significant cardiovascular risks, and I could choose between abacavir and tenofovir [TDF, Viread] or another drug, I might go with the non-abacavir nucleoside.
Now, I say tenofovir; however, tenofovir wasn't looked at in this study. There aren't enough people yet in the D:A:D analysis database that are on tenofovir to explore this. They think that will take about another year. That also gives you a sense of the data that are in the D:A:D that are being evaluated. Clearly, there is some lag time, given that they don't have enough people on tenofovir just yet.
With this minimal information, my bottom line would be that this is not going to impact my practice all that much. It may give me second thoughts among my patients who are smokers and have dyslipidemia, hypertension and such. However, this clearly is a very, very interesting and provocative study that, unfortunately, was a poster and not an oral presentation, but in many ways, I think, it generated more buzz than most of the oral presentations at this conference.
Another story at CROI focused on inflammation. We know that HIV can cause inflammation. In the SMART study, one of the leading hypotheses as to why people did so poorly when they stopped their therapy was that, when you stop HIV therapy, the viral load shoots up and you get inflammatory chemicals that are coursing through your body that cause all sorts of badness,11 but that has not really been documented. We certainly know that, as people progress with their HIV infection and even develop AIDS, there are a number of cytokines and lymphokines that increase; TNF-alpha, for instance.
But what we wanted to find out from SMART is whether, when people stopped therapy, there was this release of cytokines and lymphokines that are known to be pro-inflammatory and which ones were at work at this time point.
|Click to enlarge|
There was a very nice discussion at CROI from the SMART investigators looking at what happened when people stopped their HIV therapy. Specifically, the investigators looked at a number of different markers for inflammation and coagulation.12 As far as coagulation, D-dimer was looked at. D-dimer is a marker of thrombosis. Inflammatory reactions, C-reactive protein, IL-6, amyloid A and serum amyloid P were also evaluated.
What was obvious was that, even before people stopped their HIV therapy, across the cohort there already was evidence of increased inflammation going on at baseline. Specifically, IL-6 and D-dimer levels were elevated.
|Click to enlarge|
What you see, though, is that once people stop their HIV therapy, their viral load goes up and D-dimer also goes up. D-dimer, outside of HIV, has been linked to cardiovascular disease. Increases in D-dimer levels might double, or a little bit more than double, your risk for cardiovascular disease. In this particular study, we saw fantastic increases in D-dimer levels as HIV levels climbed. Patients who had greater than 50,000 increases in their HIV viral load level by one month experienced close to a 28-fold increase in their D-dimer level. This is off the charts compared to people who are not HIV infected, indicating that these people are at increased risk for thrombosis-related events, including cardiovascular disease.
|Click to enlarge|
In addition, when the inflammatory markers were looked at, similarly, an inflammatory marker such as IL-6 did increase after cessation of HIV therapy. Importantly, when you look at IL-6, it was closely linked to mortality, well beyond and above viral load or CD4+ cell count. There was a very strong and robust link to mortality with this particular inflammatory cytokine.
|Click to enlarge|
These data help to close the loop on our understanding of what happened when people stopped their HIV therapy. Clearly, there are people who died in this discontinuation therapy arm, but not of cardiovascular disease. I think we have to understand more about the linkage between inflammation and some of these other conditions. At this time though, I think it's very clear that stopping HIV therapy increases inflammation and that can't be a good thing.
There were a couple of posters at CROI looking at pro-inflammatory markers and correlates with intimal thickness of the carotid that did indicate a correlation there.13,14 Other studies have not found the same thing. So there are some discrepancies there, but I think the well-controlled studies and the studies that were presented here are more fairly impressive.
In addition, there was one nine-patient study of pentoxifylline [Pentoxil, Trental], a drug that's used oftentimes for claudication of the legs.15 This particular compound was also found to reduce inflammatory markers and improve endothelial function when given to patients with HIV infection. This might be another thing to explore to try to reverse some of the inflammation that occurs during HIV infection, even in people who have fairly well controlled HIV, as was seen in this study.
|Click to enlarge|
I think that this is an emerging picture. Importantly, on the heels of the results from SMART, there were other data provided by investigators from the STACCATO study, another treatment interruption study.16 The analysis that was presented at CROI was restricted to 145 individuals who were started on HIV therapy in Thailand.
You have to think very hard about the setting in which the study was done, as it differs tremendously from studies that would be done, let's say, here in the United States, in that we're dealing with a population that has a lower BMI [body mass index] and lower cardiovascular disease at baseline.
In fact, in this study, there really weren't too many inflammatory markers that were elevated in patients at baseline or during their therapy. What the researchers did see, however, was that once you stopped HIV therapy during a strategic treatment interruption, pro-inflammatory cytokines did change; they increased, and endothelial function markers also worsened when people were off of therapy.
Importantly, when people went back on therapy, much like we saw with the SMART study, there was residual damage. So the cytokine levels improved somewhat when you restarted your HIV therapy, but did not return to baseline levels, such as those achieved when viremia was under control.
What I think this means is that HIV therapy reduces pro-inflammatory markers and the inflammatory state by reducing antigenemia and by reducing the viral load, and when you stop this, you get this burst of inflammation. This burst of inflammation is not something you can turn on and off. Even when patients start back on HIV therapy, some damage has been done. Some residual inflammation continues, maybe because of some continued antigenic stimulation that we're not measuring or because there's a cascade that's been unleashed.
More work has to be done. This is still very preliminary. We don't have many opportunities to look at people on and off HIV therapy, and will have fewer opportunities in the future, since treatment trials that involve interruptions are pretty much a nonstarter. But I think there's an evolving story here regarding the link between HIV, HIV therapy and inflammation that we're going to all hear a lot about from here on in.
Another important theme from this conference dealt with mortality and aging. It's clear that the HIV-infected population is getting older and that there are diseases, such as the cardiovascular diseases we've mentioned, that are becoming more and more important to the population, and more and more of a concern for the clinicians who care for them.
So it's only proper that we start thinking about HIV in the long term and about people who have HIV infection living into their 50s, 60s, 70s, 80s and beyond. There have been a number of analyses now that show us that HIV-infected people can expect to live as long as three decades after their diagnosis, and that's assuming that there are no major advances in HIV therapy from now until then.17
It's not inconceivable that the majority of people who are living with HIV infection and doing well may well succumb to something other than HIV itself. That's not to say that everyone's off the hook. It's clear that there could be relationships between having HIV and the risk for non-HIV-related events. A number of the cohort studies we've mentioned have already found links between immunosuppression and diseases that are not classically considered to be HIV related, including some malignancies that we don't think about as being necessarily related to immunosuppression, i.e., lung cancer. Lung cancer is increasing among people with HIV infection.18 That's a consequence of some lifestyle issues, such as smoking, but it's also likely that there is an immune component to the control of malignancies that people with HIV infection are more likely to develop. I think that's a consequence of the immune suppression that they may have experienced either in the past or recently.
A CD4+ cell count of 400 cells/mm3 is great, but it's not an 800 or a 1,200 cells/mm3 CD4+ cell count. Does having a CD4+ cell count of 400 cells/mm3 over a long period of time put someone at risk for developing diseases that can be kept at bay with a higher CD4+ cell count?
I think that's an important part of the ongoing research in the realm of HIV. Here, we saw several different studies that have tried to look at that. One of the most important studies looked at what happens when people present with opportunistic infections. These are people who come in to care either knowing they have HIV infection but not dealing with it or, more likely, not knowing they are HIV infected. They come in with a devastating opportunistic condition, such as PCP [Pneumocystis carinii pneumonia], cryptococcal meningitis or disseminated MAC [Mycobacterium avium complex]. These are problems that we are continuing to see. A major concern regarding people who are not on HIV therapy and who develop one of these opportunistic infections is the question of when's the right time to start them on HIV therapy. Should we start HIV therapy in the hospital while that person is getting treated for his or her opportunistic infection, or should we wait? Most providers that I know like to wait: Treat the opportunistic infection, get them out of the hospital, have them come back to the clinic, and then start their HIV therapy.
When you talk about this with some providers, that's the way they do it. Talk to other providers and they say, "No. Let's start the therapy right away, in the hospital. We can monitor the side effects. We can work out drug interactions, etc."
|Click to enlarge|
There's this equipoise between these two strategies, and that's a perfect set up for doing a clinical trial. It's what the AIDS Clinical Trials Group [ACTG] did. They launched an ambitious study, called A5164, that randomized to one of two study arms people who had opportunistic infections -- specifically PCP; a serious bacterial infection such as pneumonia, sepsis, a deep-seated infection, endocarditis, cryptococcal disease, disseminated MAC or other atypical microbacterial infections; toxoplasmosis of the central nervous system [CNS]; and CMV [cytomegalovirus] end-organ disease.19 Patients who had any of those, and who were not on HIV therapy, were eligible for this study.
|Click to enlarge|
Participants were randomized to either:
The HIV regimens that were provided by the study included lopinavir/ritonavir [LPV/r, Kaletra], tenofovir/emtricitabine [TDF/FTC, Truvada] and stavudine. This was the menu from which people could choose. The study recommended that if one of these drugs were not used, that the regimen could be crafted by the clinician, but that a protease inhibitor that's boosted with ritonavir [RTV, Norvir] or an NNRTI be used along with two nucleosides. The management of HIV therapy during this trial was totally up to the clinician.
|Click to enlarge|
Importantly, people with TB [tuberculosis] were not eligible to partake in this clinical trial. There are other studies looking at the initiation of HIV therapy around the diagnosis of TB. This study was not designed to look at that particular question.
The study enrolled 282 people. It took a long time. This was a tough study to do, and it took a lot of work. It's hard to get clinicians and patients to think about a study when the patient is hypoxic or has a CNS lesion from toxoplasmosis.
One hundred and forty-one patients entered the immediate treatment arm, versus 141 in the deferred therapy arm. The patient population was largely male. About a third of the patient population was black. About another third was Hispanic. The CD4+ cell count, as you'd expect, was quite low. On average, it was 29 cells/mm3. The viral load was high. PCP was the leading diagnosis that led to entry into the study, followed by bacterial infections or other opportunistic infections. Cryptococcus was seen in about 16% to 18% of people in the study; toxoplasmosis was less commonly seen. Disseminated MAC and CMV were also rare, with only 1% to 3% of people coming into the study with those. Interestingly, about a third of the participants had more than one opportunistic infection.
Defying some of the oddsmakers, perhaps, those people who immediately started HIV therapy had a significant decrease in their mortality, or progression to AIDS. This was very important, and it developed early on into the study. The differences between the two arms separated out quite early. The survival advantage persisted, even out to 48 weeks after the initiation of the study.
Here is a really clear answer to a question whose answer had evaded us clinically until now. We still continue to see, and continue to argue about what to do with, patients who are not on HIV therapy and who present with an opportunistic infection; this is not some interesting tidbit that we could look at as being a remnant from days of HIV past. These data resoundingly tell us that when we see someone with a treatable opportunistic infection, delaying HIV therapy puts the patient at risk. So we should be aggressive in starting HIV therapy at the time of their diagnosis.
Importantly, immune reconstitution syndrome, which is a huge concern for those who are thinking about starting therapy immediately, was not seen as being a problem in the immediate therapy arm. There were 10 cases in the immediate therapy arm, versus 13 in the deferred arm.
However, we have to recognize that PCP was a leading diagnosis here, and a majority of those patients probably received prednisone [Prednisone Intensol, Sterapred]. Prednisone may have had an impact on that. That's not to say we shouldn't look for immune reconstitution syndromes, or not be concerned about them. We should be concerned about them, but the overall benefit of starting on HIV therapy outweighs the risk of immune reconstitution being a major problem in this setting.
This is a very important study. It is one of the studies at this conference that I think can and will change clinical practice.
When we look at studies, such as that ACTG study, that help us understand how we can improve patient survival, there are other considerations that need to be taken into account. I mentioned before that a number of studies consistently find that African Americans don't do as well as whites in clinical trials, and in clinical care.20,21 There wasn't a whole lot of illumination about why that is. There's not a lot of data about adherence in African Americans versus whites, and that may be part of the explanation, but I think even that is just the tip of the iceberg.
A number of different posters here basically reiterated the same point regarding these ethnic and racial differences. There was another study that I thought was worth mentioning, only because of its shocking take-home lesson.
Researchers in San Francisco have been following a very interesting cohort of HIV-infected people who are homeless, or almost nearly so. David Bangsberg and his group have been doing this for a number of years, and generated a lot of important information, including about adherence.
Click to enlarge|
For additional images from this study, click here.
In this particular study they looked at food insecurity.22 This is defined as having an uncertain or limited availability of nutritionally adequate or safe food. Therefore, the participants of this study are people who are marginally housed and who often are hungry.
The researchers looked to see whether food insecurity played a role in virologic outcomes. I don't think you have to think too long or too hard to figure out that, yes, it did. And that's shocking. It's all well and good to think about the latest inhibitor of some sort of viral mechanism, but when people can't even get enough food to eat, it's unlikely to have a tremendous effect.
This is an important study that tells us that there are people at fantastic risk. The very ambitious and accomplished group of investigators of this study point out some major concerns that we have to have regarding the care of people with HIV and regarding care in general, in our country, for disadvantaged people. I think that this is a very noteworthy poster in the context of all these discussions about mortality, and it shouldn't be lost on those who are thinking about these things.
I want to talk just briefly again about aging. There were a number of different plenary sessions regarding aging.23-26 The bottom line is that HIV accelerates the aging process. We can see this biologically, and we can see this in a number of different ways, even cognitively. The call here was for further studies to look at what exactly is this nexus between HIV and the aging process. But a consensus was reached that this is a concern -- that there are indications that people with HIV infection, even those whose HIV infection is under control, may be at increased risk for conditions that occur at advanced ages. A lot of research is going to go into this, including research for interventions to delay some of these aging processes. I expect there will be more research to come, and that most conferences from now on will have some section dedicated to aging and HIV.
In summary: I think what we are really trying to concentrate on is: How do we prevent diseases that are preventable? How do we prevent deaths that are preventable? There are some deaths we can't prevent. All of us are eventually going to die. However, with the advent of highly active antiretroviral therapy [HAART], along with other medical advances, death rates have decreased significantly among people with HIV infection.
As clinicians, we're trying to understand what are the preventable conditions and what we can do to optimize the chances that our patients will live as long as possible. That's what we do in the health field for everybody. With people with HIV infection, there are these additional challenges. We have to understand more about reducing cardiovascular disease risk, reducing malignancy risk, keeping patients mobile and not frail, and maintaining bone health. There were a number of different studies presented at CROI showing that bone density does decrease during HIV infection.27-29 There is some good indication that this is going to be a serious concern, and that that might lead to increased funding for this area of research.
These are the areas we have to work on and think about for the long haul. I'm excited about this, because I think we're going to learn a lot over the next few years in order to do exactly what we want to do, which is to keep people living a long time.
This transcript has been edited for clarity.
|Please note: Knowledge about HIV changes rapidly. Note the date of this summary's publication, and before treating patients or employing any therapies described in these materials, verify all information independently. If you are a patient, please consult a doctor or other medical professional before acting on any of the information presented in this summary. For a complete listing of our most recent conference coverage, click here.|