February 11, 2009
There's nothing like hearing the results of studies directly from those who actually conducted the research. In this interview, you'll meet two of these impressive HIV researchers and read their explanation of studies they presented at CROI 2009. After their explanation, they will answer several questions from the audience. This discussion was moderated by Judith Currier, M.D.
Carl Grunfeld: As death from HIV-related factors has gone down, a series of theses have emerged that are of interest in HIV infection. One is the issue of cardiovascular disease in HIV infection.
Carl Grunfeld, M.D., Ph.D.
Now, why is this important? It's important because, when you're sitting in a doctor's office, the doctor is supposed to look at your cardiovascular disease risk factors -- "Do you smoke?" "Do you have diabetes?" "What's your good cholesterol?" "What's your bad cholesterol?" "What's your blood pressure?" -- and, assessing those factors by a series of guidelines, come to a conclusion as to whether or not you should be treated [for your cardiovascular disease risk], and how rigorously you should be treated.
If all of the effects of HIV are associated with the traditional risk factors, then that's simple. But if there is an effect of HIV above traditional risk factors, you have to know that to advise patients with HIV infection. And that's true in other diseases -- for instance, rheumatoid arthritis, where the risk is above those of traditional risk factors.
But unfortunately, the previous studies in HIV have been fairly limited. There are database studies where the rate of myocardial infarction is higher, but they can't adjust for traditional risk factors, because they don't have full measurements of: How many people smoke? Who has diabetes? What are the lipids? What's the blood pressure? You can't pull that out of a database.
The best surrogate marker for years has been something called carotid IMT [intima-media thickness]. Those studies were ambiguous in HIV infection for a variety of reasons, one of which is that many of them were small, and another of which is that there's an easy way to do it, and there's a more complicated way to do it. What we did was to do both the easy way and the more complicated way, with a very large set of control subjects, so that we really could adjust for traditional risk factors.1
When we did that, we found that in both ways of measuring the carotid, HIV infection was clearly associated with more atherosclerosis after adjusting for traditional risk factors. So, yes, there is an additive of HIV. It was more obvious in the difficult way than in the easy way, but we're big enough to say it's clear. HIV is an additional risk factor. So that when you advise patients in your office, you can say, "Here's your standard risk. But because you have HIV, you are at higher risk."
We had 33,000 individuals, and in these 33,000 individuals we saw 2,192 deaths. This gave us a rate of death of 13.8 per 1,000 person-years of follow-up. We found the most common causes of death were AIDS-related deaths -- despite this being a cohort who were attending regularly for care, and despite the fact that they had freely available antiretroviral therapy, 32 percent of the deaths were attributable to AIDS-related causes. The next most common causes of death were liver, non-AIDS malignancies and cardiovascular-related diseases, which were between 14 and 11 percent of all of the deaths for each of cohorts.
When we started to look at the factors that were associated with this, we found that smoking was associated with an increased risk of death from cardiovascular disease and non-AIDS malignancies. We find that lower BMIs [body mass indexes] were associated with all-cause mortality, and with AIDS- and non-AIDS malignancy-related death. We found that those who were hypertensive were at an increased risk of liver-related or cardiovascular-related disease. Those who were hepatitis C or hepatitis B positive had an increased risk of liver disease.
We also generally found that those who had lower CD4 counts were at a higher risk of death from all of the causes, although the effect was marginal when looking at cardiovascular endpoints.
Finally, we found that the effect of the current viral load was such that, as the viral load increased, the risk of death from AIDS-related and liver-related causes also increased.
We found that there were a number of important traditional risk factors that were associated with death in our population, such as smoking and hepatitis. And we also found that some more specific risk factors, such as CD4 and viral load, were associated both with AIDS-related deaths and some of the specific non-AIDS-related causes.
We'd hope for further research looking into whether modifying the CD4 count and the viral load could perhaps lead to reductions in mortality from these causes.
Reporter #1: We've seen some studies that suggest certain drugs -- protease inhibitors and abacavir -- raise the risk of cardiovascular disease, of heart attacks. And it's been suggested that patients who are at risk of heart attacks shouldn't use those drugs. Dr. Grunfeld, do your findings combine with those in any way? In other words, should everybody with HIV avoid those drugs?
Carl Grunfeld: I think we would all hesitate to make a recommendation for everybody with HIV. What's clear, particularly from our study, is that HIV infection itself has bad consequences. The effect of HIV infection on atherosclerosis -- above and beyond the effects of HIV and the drugs' own risk factors -- is the equivalent of being male, smoking, or having diabetes.
I believe that the effect of HIV infection is much larger than the small signals seen with those drugs -- which, by the way, may affect later steps than the atherosclerosis that we measure. So I think one has to be cautious. Given the fact that we know treating HIV works, the decision should be modified a bit by what drug you put someone on, based upon what we know. But it's clear that the major thing to do is to treat the infection effectively.
Judith Currier: I think another clarification with Dr. Grunfeld's presentation is, the analysis didn't look at the role of drugs, or didn't have the ability to look at individual drugs. So, it was a summary effect of HIV and/or its treatment.
Colette Smith: Similarly, with our study, we didn't actually look at specific drugs. We just looked at the modifiable risk factors, rather than specific drugs. So I'm not sure that the data would enable you to answer that question. But there's been a lot on specific drugs and cardiovascular risk at other sessions, as well.
Reporter #2: Dr. Smith, the D:A:D study has had several publications. I'm having difficulty parsing this out. Can you tell us what's new in what you presented today, compared with what has been published, and what's been reported earlier?
Colette Smith: Well, the point of the study really was to look at deaths -- not only cardiovascular deaths, but other specific causes of deaths. The aim of the study also was to look at an important public health message, to look at what factors are associated with death -- potentially modifiable factors that perhaps would lead to interventions, and with other causes other than just cardiovascular disease, so we could look at these further.
Reporter #2: So it's kind of summing up what else happens aside from the CVD [cardiovascular disease] stuff?
Colette Smith: Yes, and looking at what factors might be potentially modifiable.
Reporter #3: Dr. Grunfeld, could you give us a better sense as to the impairment of HIV, how it compares with other risk factors, such as smoking, et cetera? Is it bigger? Smaller? Give us a context of how it stacks up.
Carl Grunfeld: It is virtually the same as being male. It is virtually the same as being a current smoker. It may be slightly higher than diabetes, or the same, depending upon which area of the carotid we are looking at. And it's the equivalent of five to ten years of age. So it is in the same ballpark as the major, traditional, dichotomous risk factors that we worry about: smoking, male gender.
Reporter #3: So in essence, one can construct an algorithm of adding up the risk factors and seeing how things stand?
Carl Grunfeld: I don't think we're at the stage where we can mathematically construct the formula and add it up. But we have a signal that it's very similar to those, and we could say yes; it's the equivalent of getting diabetes.
Reporter #4: From the beginning, there has been -- I don't know how to put it gently -- confusion about how to measure IMT, and arguments about: Was it measured the right way? Was it the best measurement? From your perspective, is there any movement towards standardization of carotid IMT measurements in trials?
Carl Grunfeld: There is movement, in part based upon the FRAM study, in part based upon a recent working group of many of the grants that are studying cardiovascular disease going forward, with the idea of looking more at the [carotid] bulb region, the "more difficult" region, that actually is better at picking up the signals for risk factors. But clearly what we have shown is that, in both regions, you can find the effect if you have got a large enough group of HIV-infected and control subjects to be able to adequately adjust. But there definitely is a movement to begin to look at the other region.
Reporter #4: For us non-experts, can you give us a quick summary of bulb region? What's the other region?
Carl Grunfeld: The common carotid goes up, and it's a straight flow. It's less susceptible to risk factors. At the point that you divide into the internal carotid and the external, the exact point of division is called the bulb, and then there's the internal. That's an area of turbulence. Areas of turbulence are actually the first areas to get atherosclerosis, and they are much more susceptible to things like lipids. It's why early lesions in the heart are proximal; they're near, in the periphery. But that's a more difficult area to measure. Many people have preferred the common, because it's easier to measure.
But issue you raised, which is an important one, the NIH [U.S. National Institutes of Health] already has moved on, in recommending in a conference of the people doing many of the going-forward IMT studies; we've come to an agreement to measure common and bulb.
Reporter #5: Dr. Grunfeld, did you have any data on HIV viral loads? Were you able to see any?
Carl Grunfeld: That analysis is not yet ready. It is being done. The point that we needed to make is the point for clinicians, which is that, in the context of traditional risk factors, HIV is an extra one. My prediction, knowing the data, is that the effect of HIV is not going to be able to be accounted for by things like viral load. But we don't have those analyses finished yet.
Judith Currier: I think, importantly -- being a collaborator on this project -- that we don't have enough untreated patients to be able to contrast the difference between suppressed and unsuppressed [viral load].
Carl Grunfeld: That's for sure.
Judith Currier: So that may be a limitation of what we're going to be able to find.
Carl Grunfeld: The FRAM study is representative statistically of HIV in the United States, compared to the Hicks study, which attempted to be representative. We're very, very similar. But we're talking about 97 percent of patients having been treated, and 94 percent having HAART. So we can't make that comparison.
On the other hand, the reality is that in the United States, and in most developed countries, a very high percentage of patients are treated. And these are the patients that we have to worry about cardiovascular disease in.
Reporter #6: Dr. Grunfeld, as many as 50 to 80 percent of the patients in the study being smokers to start with are people with HIV, smoking. Do you have enough patients left over, once you correct for smokers, to really have high enough numbers to see differences in risk factors?
Carl Grunfeld: Yes. We're below. We're about at 40 percent. And the two are independent. There is a risk with smoking. There is a risk with HIV. Obviously, smoking is a modifiable risk factor, but a very difficult one to modify. And the take-home lesson from patients with diabetes is, you work as hard as you can to modify the modifiable risk factors. And I would urge everyone with HIV infection, and everyone taking care of patients with HIV infection, to do the same thing: modify the modifiable risk factors until we understand what in HIV does this.
Reporter #6: Would smoking and having HIV basically double your risk of IMT growth?
Carl Grunfeld: I can't quote the exact number. But smoking and HIV clearly significantly increase your risk of -- OK, the smoking risk: 36 percent, versus 15 percent in controls. Smoking clearly significantly, maybe in that ballpark, increases your risk of having atherosclerosis. Now, her [Colette Smith's] data show, and our data from FRAM mortality, which was shown in a poster, all showed that smoking increases deaths in HIV. It's a very important risk factor.
This transcript has been edited for clarity.