April 25, 2013
Cardiovascular disease is the most common cause of mortality in the U.S. It is thus, almost by definition, a major topic of concern for the HIV-infected people we count among our patients and clients -- many of whom are intimately associated with traditional cardiovascular risk factors, such as cigarette smoking, low physical fitness, hyperlipidemia and hypertension.
Yet when it comes to the incorporation of cardiovascular risk management into HIV care, we haven't seen much in the way of consistent or comprehensive guidance. As we accumulate more data regarding the intersection of HIV and cardiovascular disease, a clearer picture of the issue is developing. But that picture has yet to be properly framed for many of the health care professionals who actively treat people infected with HIV in the U.S.
For this episode of HIV Management Today, we called in an expert to paint -- and frame -- that picture for us. David Alain Wohl, M.D., is an associate professor of medicine in the Division of Infectious Diseases at the University of North Carolina (UNC), the site leader of the UNC AIDS Clinical Trials Unit at Chapel Hill, and the director of the North Carolina AIDS Education and Training Center. He is a widely respected clinician-researcher who has extensively spoken and conducted research on cardiovascular complications in HIV-infected populations.
Table of Contents
Dr. Wohl, how has our knowledge regarding the relationship between HIV and cardiovascular risk deepened over the past several years?
I think our understanding has evolved, and the ways that we've come to understand things have been informed by more and more data.
There are a lot of disease states where they have these large cohorts of people. If you're talking about diabetes, hypertension, cardiovascular disease, cancer: You can look at these large cohorts of people who get these diseases because they're fairly prevalent. For a long time, we didn't have that with HIV. But we've got more and more of that.
On the other side, in addition to these relatively large cohort studies that we now can enjoy within the context of HIV, we have more and more pathogenesis studies -- studies that look at what is going on at the level of the blood vessel, and what the connection is between humeral markers, inflammation markers, traditional risk factors and novel, putative risk factors for cardiovascular disease.
What we've come to understand is: It's not as simple as we may have thought, where you take drug X for your HIV, that raises your lipids, and thus your risk for heart disease is higher. It's much more complicated: There may be some action of that drug on your cardiovascular risk, but what about the virus itself? What about the underlying immune activation that occurs in anyone who is HIV infected, to some extent -- some people more than others? How do you tease that apart from this higher prevalence of traditional risk factors that we see among people living with HIV -- where, instead of 20 percent of people smoking, 60 percent of HIV-infected people smoke? And hard living: What about cocaine? There are a lot of these hard-to-measure confounders.
I think our understanding has now encompassed an appreciation that this is pretty complicated and multifactorial, and that there's not a single line that you could connect from risk factor to risk in the HIV-infected person.
It feels like you just opened seven or eight different cans of worms. As disgusting as it is to explore worms, it would be really nice to examine those more closely. Let's start by focusing specifically on what we're talking about when we say "cardiovascular risk": What particular risks, and clinical manifestations of those risks, are we talking about when we talk about people living with HIV?
For some, it can be a composite between having a heart attack and a stroke; having to have some intervention to open up a blocked artery, such as angioplasty; or, even some definitions include peripheral vascular disease, peripheral artery disease. There are different ways to slice it.
In the purest sense -- maybe in the most absolute -- it's whether or not there's a myocardial infarction, a heart attack. Because you either have one or you don't have one.
What is our current knowledge regarding the relative risk of MI in HIV-infected people compared with the population as a whole?
Looking at heart attack and cardiovascular disease, in general, we see that people with HIV tend to have more than you would expect, even when you account for a higher prevalence of traditional risk factors. When you try to get rid of the confounding of more smoking, age and gender, things like that --
Some do correct for that, and some don't. I think that's an important point that I'd like to get back to, because I think that it is hard to account for every single confounder. That's sort of this lingering question.
But you do see multiple studies coming from multiple places, done in multiple ways, that are pretty much coming out with a fairly consistent picture of a somewhat higher risk [of MI among HIV-infected people].
The good news is -- you can look at this as the glass being half full or half empty -- but the good news is, this is not like a huge, amazing, oh-my-God-everyone's-going-to-get-a-heart-attack kind of increase in risk. It is almost like HIV is another risk factor.
There's a lot of other risks that exist, and quantifying this risk depends upon which study you look at. Some of the more recent studies are showing that the risk is there, but it's not whopping: There's a slight to a moderate risk increase among people living with HIV for this to happen, but it still is a fairly rare or unusual event in people living with HIV, especially the middle-age range.
When you get older, all of us have an increased risk for cardiovascular disease; and it becomes much more prevalent. This is where we start to see the difference. We see younger people with HIV not really having as much risk. After about age 40-plus is where you start to see more of this excess risk. I think that's intriguing, as well.
Are those parallel curves? If you're comparing HIV-infected to HIV-uninfected populations and you're watching the MI risk or cardiovascular disease risk as people age, do they increase at the same rate? Or does the rate of, let's say, MI among HIV-infected people increase at a faster rate as people age?
Some of the trend data that we saw most recently are from the D:A:D cohort, the large European cohort. That includes only people living with HIV. And when they compared coronary heart disease from age 40 onward to other cohort data, you can see that there may be an increased excess, but it's fairly parallel to most of the data from other cohorts.
So, you do have this higher relative risk. But, over time, it doesn't seem like there is an accumulation of even greater piling on of excess risk over time, such that they splay divergently.
I think what we are seeing is this increased excess risk during the early part of middle age that may be increasing because it started where there was really no excess risk, but hit some sort of a plateau, where you start to get higher rates but they don't seem to be wildly accelerating.
What do we know so far about what's to blame for this higher level of MI risk among people with HIV?
There is clearly this conceptual model of immune activation going on in people living with HIV. That immune activation leads to the expression of all these inflammatory chemicals that we have now come to appreciate -- and are even measuring routinely in our clinical research studies, if not sometimes even in clinic.
These inflammatory markers, or chemicals, have been known to have effects on people's risk for cardiovascular disease and the pathogenesis of cardiovascular disease, such that those who live with more inflammation -- whether it be from rheumatoid arthritis or other disease states that are chronic and inflammatory -- have higher rates of problems with their arteries and with atherosclerosis.
So it follows suit that you can imagine -- connect the dots, in a way -- that among people living with HIV, the same sort of thing happens.
There certainly is a component, I think, of the therapies, and what they do. How big a component that is, how big that drop is in the bucket, is up for debate. And different medicines have different activities; how much of that is active or operational, when HIV therapy reduces immune activation by controlling the virus? Is any effect of HIV medicine trumped by what it does to HIV itself? That, again, is debatable -- and it's all relative: If you have a drug that raises your LDL cholesterol, we believe that's going to raise your risk for cardiovascular disease -- even if it does reduce your immune activation -- compared to another drug that reduces your immune activation to the same extent but doesn't increase your LDL. Somebody will potentially have an event due to that LDL increase over time, if you look at enough people.
The thing that's not measured, and that I really am concerned about in some of these cohort studies, is that they're looking at people from a long period in the past -- 10 years or more, sometimes. I wonder whether or not these data will hold up over time.
By that, I mean: If we're looking at people in the D:A:D cohort, or these cohorts from Boston or the U.S. in general, we're looking at people with HIV infection who may have not been treated for five, six, seven, 10 years. These are people who oftentimes were diagnosed and were managed before guidelines became more aggressive, and who lived with HIV replicating in their body, with that immune activation.
How much of the groundwork for coronary artery disease and cardiovascular disease was paved by [studies involving people with] untreated HIV infection for a long time? With earlier treatment -- and with drugs that are not as offensive to lipids and cardiovascular risk -- will we still see this excess when we account for the traditional risk factors?
That's an intriguing question for me. I wonder whether or not some of the data we're seeing now will be relevant even just a decade from now, when we have a different paradigm for treating HIV.
Is the determination of risk really that subjective? Are there objective tools that we can use? How reliable is Framingham for HIV-infected people, for instance?
It looks like Framingham is pretty good. If anything -- like I think the D:A:D folks have been trying to tell us for a while, but we now have better data [to support] -- I think it probably underestimates risk for cardiovascular disease over a 10-year period in people living with HIV.
The way I look at [a person's Framingham score] is: Here's your bare-minimum risk. It's at least this -- or, if not, somewhat higher.
But if you ever play with the Framingham Risk Score -- which is really fun to do; I play with it all the time with my patients. I'll look at it and I'll say, "Well, here's what happens if I lower your LDL. And here's what happens if I raise your HDL. Here's what happens if you stop smoking" -- which, of course, has the biggest influence, and it's dramatic. If you change the things that you think you can change, you can get an incremental sense for what kind of differences you're seeing.
I think these risk assessments are useful. They help people quantify what range they are in. But they're not exact. And I think the D:A:D study from many years ago showed that there was a little bit of an added effect of HIV. I think that probably is the case.
Have you used the Veterans Aging Cohort Study Index?
That's really fascinating, too. And I think that we should be thinking more about it. One thing that may prevent some of us from using it more widely is it's a little complicated. I think there probably is an app for that already. As that kind of thing gets marketed, and becomes as ubiquitous as Framingham Risk and as easy to plug people into, I think it will be useful.
For a lot of my patients that I would probably calculate, I already know that they've got problems. If you already have anemia, which is a big driver of the frailty risk, that counts for a lot in that index. I can tell whether or not you have anemia, and if you have anemia, you have a whole bunch of other stuff going on.
It's less useful for my 50-year-old guy who gave up smoking years ago, who never smoked, and goes to the gym all the time. For me, I probably wouldn't reach for the frailty calculator.
But I do see how it's a useful tool. I think that the people who've developed it have done a really great job of validating it, and I think it should get more attention. I like how they've used it on a population basis that really helps us understand where people get into the danger zone.
Do you feel that inflammation markers are worth testing for in your patients, or does that not really inform the decision that you might otherwise make? Is it more for research purposes?
I very rarely order inflammatory markers on my patients. I usually don't have motivation to do so. I will sometimes do it: I have a patient I'm seeing now on whom I just ordered an hs-CRP -- a C-reactive protein level -- and a D-dimer. She's not on ritonavir; she didn't tolerate it, so I needed to put her on a friendly, effective regimen. She's doing great on abacavir/3TC and unboosted atazanavir, but does have hypertension and some dyslipidemia, and I've worried a little bit about that. So I did check a CRP and a D-dimer on her, to see if there was an increase. Whether or not any increase is related to the abacavir, I can't tell -- unless I stop the abacavir, replace it with something else and see what happens to those inflammatory markers. I'd probably repeat [the marker testing] before I did anything.
But that's a pretty unusual circumstance. Normally, I don't do it. If my person's on tenofovir/FTC and efavirenz, and doing great for years, I'm not going out and ordering a CRP or a D-dimer to check coagulation -- and certainly not an IL-6, or any of these things. I don't do that yet. I think we have to learn more about how to use those things.
You can certainly look at platelet count as a poor man's marker of inflammation. But it's unreliable.
I don't think it's ready for prime time to do an inflammatory panel. I do check lipids; I think those are important to check. I reduce the LDL, to the extent that I can, in people whose LDLs are not anywhere near where they should be.
Where does this leave us, then, practically speaking, in terms of what to do with the confusing array of data? From the standpoint of any clinician who is treating an HIV-infected patient today and may be concerned about their cardiovascular risk, how should they be approaching this?
Well, I think there is data you can look at and say, "Here is excess risk." Like I said, it's pretty clear that we're seeing this consistently from study to study. There's greater risk for cardiovascular disease among people living with HIV, even when you correct for the bad stuff -- smoking, whatever.
When you don't correct for that -- which we don't; in real life, we don't say, "Hmm, you know, since you're a smoker, I'm going to discount that." We count that. People with HIV unadjusted for these factors have a lot of cardiovascular disease, and much of that is driven by lifestyle stuff. Your smoking is a big deal; that really matters a lot. Uncontrolled blood pressure. Obesity.
When you look at U.S. cohort studies like the groups have done repeatedly that have been published, they match on a whole bunch of different things. But do they match on region? Because I happen to know that in the region that I live in, in the South, there is a higher prevalence of obesity and diabetes compared to other places. I bet that translates into the HIV population, too. If you have more people living in the South with HIV and they are more represented in some of these cohorts, let's say -- and I'm not saying that's the case -- does that mean we have an imbalance, just because of geography, and how the geographic considerations factor in?
It's hard to tease out, but what I will say is people with HIV do have an increased risk of cardiovascular disease that's evident to anyone who is looking. A lot of that is driven by the behaviors, the lifestyle, of people living with HIV. Take that away, though, and you're still left with something.
For us, that means figuring out: If immune activation paves the way for cardiovascular disease, let's reduce immune activation to the extent that we can. The number one way to remove immune activation from the equation is to treat HIV, because the immune system won't be activated to respond to HIV if the HIV level is so low. So: treating earlier; treating more effectively; engaging people in care; keeping them on therapy; no breaks in therapy, no holidays; using therapies that are as innocuous as possible for cardiovascular health, and we're getting more and more and more of those; and, for people who have innate cardiovascular risk, due to genetics or whatever, treating that aggressively.
There was a really nice simulation that was done by the D:A:D cohort, where over-40 people had an increased risk compared to all these cohorts. They manipulated the data to say, "What happens if you drop blood pressure? What happens if you drop cholesterol? What happens if you make people who are smokers stop smoking?" While most things had incremental benefits, stopping smoking was huge.
It's easy to get up on your soapbox and complain about smoking, but so many people in our population living with HIV smoke that it's worth doing that. It sounds somewhat hackneyed and redundant, but it is a really big deal.
We could talk about how many angels dance on the head of a pin, in terms of one antiretroviral versus another; but probably, the biggest bang for the buck will be just doing the stuff that we do outside of HIV to reduce cardiovascular risk -- these lifestyle changes -- which are really hard to do in a lot of our patients, but will make a big difference.
If you can't do that, or if there's still some remnant of risk due to some immune factors, maybe there's something else you could do. Many people are studying anti-inflammatories, whether it be traditional aspirin -- which is underutilized, by the way, in people living with HIV; study after study shows that we are not as present in thinking about aspirin as an intervention as we should be for people living with HIV.
We do great with statins. But there may be other things that we can do to reduce immune activation and inflammation beyond that that people are studying, when you've done everything else.
Why are we good with statins, but not aspirin?
I think statins have become pretty easy to use. And I think we've gotten great data about how to use them with HIV therapy. I think the guidelines are very clear about when to apply statins or lipid-lowering therapy.
I think the aspirin data have been more confusing for clinicians: who should be on them, who should not be on them, the risks involved. Most of us are really good at keeping up with things that we're told about that are clear and concise. When there's murkiness, things start to see some falling off.
I think what we need to do is start thinking more and more about traditional methods for reducing coronary artery disease, including using aspirin; using statins; implementing effective, evidence-based interventions to get people to change their habits that are bad, whether that be smoking -- and there are methods and there are pharmaceutical aids that help, as well -- or getting people into gear to eat well and exercise.
Again, these are tough things, even without HIV. But when you have a population that's at increased risk, it becomes really important to apply those in a way that's going to work.
I want to come back to antiretroviral therapy for a second. Do you feel that the choice of drugs is not a really important part of the conversation?
I think the choice of drugs is less a part of the conversation than many people make it out to be.
A decade ago, this was a huge thing. A decade ago, this was all we were talking about, when the first D:A:D results were coming out.
I think that's right; the D:A:D results implicating abacavir and ddI were concerning. Overall, even before that, results showed that people exposed to HIV therapy had an increased risk of myocardial infarction and cardiovascular disease over time. [However,] I think it's a little messy, and it involves people who have been treated and then went for a long period of time without treatment -- a whole mess of different therapies, including older therapies.
Most of our therapies raise LDL a little bit. Most of our therapies increase triglycerides. The increase in HDL that some of our therapies give us: We don't know whether or not that's as meaningful as HDL that rises from exercise. Not all HDL is created equally. So there is probably an effect of being on therapy that's a little hard to quantify and tease out, but is there.
The abacavir question still is really controversial. I think that abacavir, for the great majority of people who will take it, will not do them any harm. The question is in someone who already has risk factors: Is that the drug that you should use? When there are alternatives, most people are turning to the alternatives. And when there's not -- well, then you need to use it and just be aware; maybe try to compensate by being aggressive about other risk factors.
Remember -- and this is an important point -- there's absolute risk and relative risk. The relative risk [of MI] is increased. That means that, compared to people not on the drug, you have an increased risk that may be several fold higher, or whatever: 1.17, or 1.2, or 1.5.
But the absolute risk -- how many people out of a thousand are going to get it? -- is pretty low. The vast majority of people are not going to have this outcome. Even among people who have risk factors, not all of them are going to get a heart attack. So, I think that we have to be realistic.
There's a taint of abacavir right now that hasn't been completely dispelled by the data that's accumulated since then. We don't really have a great pathogenic mechanism, although there are some papers that purport to describe some stuff. But there are other papers that don't show the same thing, and some that don't even get published because there are negative results. We don't really have a full answer.
I think that, for most patients, it's a fine drug to use if you cannot use tenofovir -- and there are still people who can't use tenofovir, so I don't hesitate to put them on abacavir, if it means that's what I've got to get their virus under control. Because that's the number one thing: getting their virus under control. The worst thing you could do is not have their virus under control.
We just focused a lot on abacavir, but I want to make sure that we don't ignore the other drugs. If we look across the armamentarium of currently recommended drugs in the U.S. DHHS guidelines, are we not really looking at any particular noteworthy cardiovascular risk there?
No: I think Norvir is a problem, too. I think that we have to boost our protease inhibitors with Norvir is problematic. I think we know that Norvir has its own inherent issues, as far as cardiovascular risk -- and for lipid increases. It would be nice not to have that.
With cobicistat, we don't get much improvement, although we see lower triglyceride increases -- most notably, smaller triglyceride increases than we see with ritonavir in the head-to-head study that looked at it with atazanavir. That will be nice.
Efavirenz doesn't have that much of a benefit, as far as lipids go, compared to the boosted protease inhibitors. There are myths that it does; it doesn't. It leads to very similar sorts of lipid profiles, so I don't think you could get a benefit there.
Raltegravir is really innocuous when it comes to lipids, and it looks like it shouldn't cause anything. But there's baggage with it being a twice-a-day drug.
With dolutegravir, will that become another factor? It could. That may be more heart-safe, especially if we use it with tenofovir and FTC. It's going to be co-formulated with abacavir and 3TC, so that may be tantalizing for some people -- but then you have to deal with the abacavir issue.
I think we should be thoughtful about our medicines, vis-à-vis cardiovascular risk. That makes a lot of sense. But that is, to me, secondary or tertiary to dealing with the big-ticket items that we've talked a lot about.
To tie it back into what you had said very early in this conversation: It's also about being well aware of what your local epidemic is like, what the cardiovascular risk factors are among the population of people that you're treating. There are so many of these potentially regional, potentially socioeconomic associations with particular risk factors that are really worth paying attention to, and really worth addressing with your patient, even though some of those topics might be a little bit uncomfortable; and then being able to properly assess and try to do something about it.
Absolutely. I wish that we had more that we can offer our patients. I am really, really, really good at writing antiretroviral prescriptions. I am really, really, really not as good at recommending what would be a winning nutritional program for a patient who comes in with obesity, hypertension and diabetes, or pre-diabetes.
We need to have more of a streamlined, accessible, clinically useful approach to that.
Do we need a cardiovascular risk version of the U.S. DHHS guidelines panel? Do we need a panel that specifically covers this?
I think it would be great. The thing we need to do, though, is understand that even outside of HIV there's a lot of controversy here. Just go to your local bookstore and you'd find all these incredible diets. Someone's on a paleo diet, where they're munching down a lot of fat; and you have other people who are vegans; and you have other people who are doing the Mediterranean diet -- which, to me, naturally just seems to make a lot of sense, and a recent New England Journal of Medicine article showed that it seems to be very effective.
But diet is very cultural. Diet is very ingrained. There's a lot of psychological investment in your diet. It's very hard to change. But I do believe that dietary changes can help.
As I kind of intimated, it's not only just reducing the ingestion of these foods that can lead to promotion of atherosclerosis; there may be other effects on the microbiome, on the flora, that exist within us that change as well, that may also have benefits -- again, with inflammation and with atherosclerosis.
We're just in the infancy of understanding that part of it. But trust me: We are all going to be hearing a lot more about that, in ways that are only starting to get out now.
But there are some interventions that I think we're going to have to look at that may be beneficial to our patients, and may not be magic bullets, but may augment those types of things we're talking about, as far as lifestyle.
We'll explore those issues in more detail in the future, but for this discussion, I think we need to leave it there. Thank you so much, Dr. Wohl.
This transcript has been edited for clarity, grammar and length.
Visit our HIV Management Today home page for more episodes in this series.
Myles Helfand is the editorial director of TheBody.com and TheBodyPRO.com.
Follow Myles on Twitter: @MylesatTheBody.