You are leaving Medscape Education
Cancel Continue
Log in to save activities Your saved activities will show here so that you can easily access them whenever you're ready. Log in here CME & Education Log in to keep track of your credits.


HAART, Dyslipidemia, and CV Risk: An Expert Interview With Grace A. McComsey, MD


This feature requires the newest version of Flash. You can download it here.

Editor's Note:
Metabolic abnormalities associated with HIV and its treatment have emerged as one of the most vexing challenges in the management of patients with HIV infection. Dyslipidemias -- increased triglycerides and LDL cholesterol, and decreased HDL cholesterol -- are key concerns for most HIV clinicians, and effective interventions for these problems have been lacking. At the 13th Conference on Retroviruses and Opportunistic Infections (CROI) in Denver, Colorado, Scott Williams, Editorial Director of Medscape HIV/AIDS, spoke with Dr. Grace McComsey, Chief of Pediatric Infectious Diseases and Rheumatology, and Associate Professor of Pediatrics and Medicine at Case Western Reserve University in Cleveland, Ohio. Dr. McComsey is an expert in metabolic abnormalities in HIV infection, and she offered her clinical and scientific insights regarding this complex issue as well as key data from the meeting on this topic.

Medscape: We don't have a head-to-head prospective trial that compares the lipid effects of all 4 protease inhibitors (PIs) that are most often ritonavir-boosted-- lopinavir, fosamprenavir, atazanavir, and saquinavir. If ritonavir is taken out of the equation, what is your thought on the contribution of each of these agents to dyslipidemias?

Dr. McComsey: It's clear now that every PI besides atazanavir has some effect on lipids, although this effect does vary by PI. A question addressed at this meeting was whether adding a small dose of ritonavir (100 mg/day) would affect the lipid effects of atazanavir.[1] When 100 mg of ritonavir was added to atazanavir [in a prospective, randomized study], it did lead to an increase in fasting lipids. After 48 weeks there was a 26% increase in triglycerides and about a 15% increase in total cholesterol; both of these increases were significantly different from baseline values. Given these data, we now know that boosted atazanavir, even the naive setting, does lead to some increases in lipid levels. Are these as bad as with other boosted PIs? I do not believe so, and they're still in the range of mild increases.

Prior to this randomized study with atazanavir, we had some studies in treatment-naive patients with each of these agents that can tell us something about the lipids. For instance, with saquinavir, the older studies have shown mild effects on lipids -- not as bad as indinavir or ritonavir. I would put fosamprenavir in the middle. We know from the NEAT study[2] that used fosamprenavir without ritonavir that it does affect cholesterol and triglycerides, but the important thing in that study, if you look at the changes in HDL cholesterol, is that it also increased. So it's not only LDL and it's not only triglycerides; you have to look at the whole picture. With fosamprenavir, after 48 weeks the ratio of total cholesterol to HDL cholesterol was unchanged, so that was an acceptable result for fosamprenavir-treated patients.

Medscape: What about the real-world clinical situation of using unboosted atazanavir? Aren't most clinicians using it in a boosted fashion at this point?

Dr. McComsey: That's correct. I think that unless a patient has significant dyslipidemias to start with, most clinicians would use a boosted PI. In some patients who have a family history of coronary artery disease or high lipids before starting medications, some physicians would use unboosted atazanavir. Some people don't trust [unboosted atazanavir] in terms of potency, but in one study[3] it was as good as efavirenz, so some physicians do trust it in treatment-naive patients.

Medscape: The type of dyslipidemia, such as elevated triglycerides or LDL or a low HDL, can influence whether a clinician first uses a statin or a fibrate to improve the lipid profile. Does the particular type of dyslipidemia influence the switch to another boosted PI or other drug class if a regimen switch is in order?

Dr. McComsey: Well, I think it's a good question. The answer to that is that we don't know. I think that so far the switch studies that showed a positive effect on lipids showed, in general, positive effects on triglycerides as well as cholesterol, so it's hard from the data to conclude that certain switches are better than others for specific dyslipidemias. The switches that have been [generally] successful are PI to NNRTIs [nonnucleoside reverse transcriptase inhibitors], PIs to atazanavir, as well as d4T to tenofovir. Notice that I didn't mention switching to boosted atazanavir, as I am still not convinced of that data yet.

Medscape: ACTG 5186 tested the safety and efficacy of fish oil plus fenofibrate in antiretroviral-treated patients with hypertriglyceridemia. What did their results show, and what implications does this have in the management of this condition?

Dr. McComsey: We've been using fish oil in the clinic for patients with high lipids, specifically triglycerides, and it's been used outside of HIV. But this is the first large study[4] that has a slightly different design than the 2 prior studies that tested fish oil in the setting of HIV infection. The 2 prior studies just looked at fish oil vs placebo and found that fish oil decreased triglycerides and didn't have any effect on cholesterol.

In this study, the ACTG looked at a different approach. They enrolled patients into a study where they got randomized to "Step 1," where they received either fish oil or fenofibrate, 2 agents that are known to affect triglycerides outside of HIV infection. After Step 1, if there was not a good response of triglycerides (to ≤ 200 mg/dL), then patients were enrolled in "Step 2," where they received both medications together. This tested the efficacy of each agent alone and then in combination, which might lead to a better attainment of NCEP [National Cholesterol Education Project] goals.

In Step 1, very few people overall reached the NCEP goal. About 75% went on to Step 2 and about 23% did respond to the combination by the most rigorous intent-to-treat analysis. The decreases in the levels of triglycerides were pretty impressive: from about 700 to 350 mg/dL-- approximately a 50% decrease

Medscape: Can we say that that's a clinically significant decrease?

Dr. McComsey: We don't know. We're learning more and more that triglyceride level is a significant independent risk factor for cardiovascular disease. Decreasing the risk of triglycerides not only will decrease pancreatitis risk, but also probably decrease long-term cardiovascular risk.

A potentially negative finding was that they noted mild increases in LDL cholesterol, which I doubt are clinically significant. Some of the studies outside of HIV that have assessed fish oil have noted decreased triglycerides and increased LDL cholesterol, but the increase in LDL was very mild. Other than that, it was very safe, including for gastrointestinal side effects, and well tolerated, which is something that our patients sometimes worry about when we suggest fish oil.

Medscape: And there was no negative effect on antiretroviral drug levels?

Dr. McComsey: No, and in a subset of patients, they looked at lopinavir levels before and after addition of the fish oil; there was no difference in lopinavir levels and there were no virologic failures.

Outside of HIV studies, there has been a suggestion that fish oil might be immunosuppressive. They looked at that in this study and found no effect. The fish oil they used is of good quality, which also likely helped the good response that they observed.

Medscape: We also saw additional data from the D:A:D [Data Collection on Adverse Events of Anti-HIV Drugs] study, which is a very large observational trial that is measuring cardiovascular risks and events in antiretroviral-treated patients. What did those latest data show?

Dr. McComsey: We've seen published data from D:A:D before,[5] but this is the first time that we've heard results from that study presented by type of antiretroviral therapy.[6] For the first time, they compared data on patients who received NNRTI [nonnucleoside reverse transcriptase inhibitor]-containing regimens with data on patients who received PI-containing regimens. What we learned is that NNRTIs are not associated with increased risk of myocardial infarction, but their data still show that PIs are associated with some increase in the incidence of myocardial infarction (MI). This was not a huge absolute increase, and it's a relatively small increase in the relative risk of MI. It's interesting to see that the NNRTIs do not affect the incidence of MI, and when they adjusted for the NRTIs, that did not affect the results in the NNRTI or the PI groups.

Medscape: Many of the classic risk factors for cardiovascular disease did pan out in this trial as well, and those should be aggressively managed by clinicians when possible.

Dr. McComsey: Absolutely. In their initial analysis and every time they look at the data, classic risk factors still pan out as independent predictors. Unfortunately, some of those we cannot change, but obviously smoking and aggressive management of dyslipidemia, among others, are factors that we could affect positively.

Medscape: Certainly the SMART study[7] and data from a number of other structured treatment interruption trials[8-10] were obviously the major story at this meeting. SMART investigators saw a significant increase in major adverse events (AEs) in the intermittent-treatment arm, including cardiovascular AEs, as compared with the continuous-treatment arm. Now this is contrary to the expectation that less treatment would result in less AEs. Why do you think this is the case, and what might have accounted for the increase in cardiovascular AEs among patients who received less antiretroviral therapy?

Dr. McComsey: Obviously that was one of the major findings from CROI and something that was surprising to most of us. I think that if you ask HIV treaters what they would suspect in terms of cardiovascular events, they would say that if you give more drugs or continuous antiretroviral therapy, then you're going to see a more significant increase in cardiovascular disease. We learned at the meeting not only from the SMART study, but even in ACTG 5170 (another treatment-interruption study), that some patients had significant cardiovascular events, including mortality from an MI after interruption of antiretrovirals. We don't know why that is the case, but what we know is that the virus itself has an effect on the heart, so we shouldn't always blame antiretroviral therapy. It could be HIV itself, it could be immune activation that is the result of HIV, and perhaps intermittent therapy may be the worse thing that we can do to the heart. When viral load goes up and down with stopping and starting ART, immune activation may indeed turn out to be a worse problem than treatment-associated AEs, even dyslipidemias that we commonly see.

Even before highly active antiretroviral therapy (HAART) we knew that HIV itself can affect triglycerides and lower HDL very significantly. There were a lot of discussions about what would be the best test for measuring immune activation [in this type of study]. I know that people are going to be looking at that issue and at the details of these cases.

Medscape: Were there any other studies related to cardiovascular risk and dyslipidemia that you thought were particularly important?

Dr. McComsey: We saw more data on carotid intima media thickness, which is a surrogate marker for underlying atherosclerosis.[11] We saw a study in children[12] who had no other risk factor for cardiovascular disease, other than HIV infection and antiretroviral therapy. These children had a significantly thicker intima media, which indicated more atherosclerosis risk compared with an HIV-negative control group. Even in children who had been on treatment for a median duration of 8 years, there was already evidence of premature atherosclerosis.

Of interest, both in the study by Judy Currier and colleagues[13] as well as this study in children, intima media thickness correlated with insulin resistance, something that we don't check in the clinic on a regular basis, but more and more we're learning that it's an independent risk factor for cardiovascular disease.

The session on metabolic complications also had a lot of data on metformin and the glitazones.[14-16] To sum it up very quickly, metformin was disappointing. It did not lead to any change in central fat accumulation, which is a little different from what we saw in a prior smaller study. Pioglitazone actually did pretty well: Patients with lipoatrophy did have improvement of their peripheral fat after the addition of pioglitazone. As in the various studies that switched antiretrovirals in an attempt to improve lipoatrophy, improvement was very slow and fat was only partially restored. Lipoatrophy did not resolve completely in this new study, but there was about a 10% improvement in limb fat after a year of treatment with pioglitazone.

Medscape: This differs from some of the other glitazone studies that have been reported previously.

Dr. McComsey: Absolutely. Some have shown no significant gain in fat, such as the Rosey study from Andrew Carr's group.[17] But pioglitazone is a better PPAR-agonist, so that may be a reason why pioglitazone did better than rosiglitazone. They're not the same; pioglitazone is more potent. In addition, pioglitazone was more effective in people not receiving d4T; this may turn out to be a very significant factor in predicting efficacy to the glitazones.

I also wanted to mention the ACTG testosterone study.[18] In the clinic, a lot of males ask for testosterone because there's a belief that if you give testosterone as shots or as gel, that patients will improve their lean body mass as well as decrease their central fat accumulation. The ACTG did a placebo-controlled study of testosterone in subjects with central fat accumulation and mild or moderately low levels of testosterone, which is what we commonly see in the clinic. What they found was, to everybody's surprise, that testosterone did not improve central fat accumulation, and even worse, it decreased limb fat. For patients who have lipoatrophy, we should be cautious to give testosterone because lipoatrophy can worsen. That's something that has an impact on our clinical practice.

Medscape: For a hypogonadal male with lipoatrophy, would you suggest not giving testosterone, even if he has very low levels?

Dr. McComsey: If you have a very low level or suggestive symptoms, then you have to replace it, but the study specifically looked at people with what we call mildly decreased levels.

Medscape: And how would you define that?

Dr. McComsey: They defined it as a level between 125 and 400 mg/dL or patients with testosterone level of > 400, but a low free testosterone level. That's something we see a lot in our clinic. The question is what to do with these patients.

If you need to replace testosterone because the patient is having symptoms, that would be different. You just have to watch and hope that there is not decreased limb fat. These data do give us insight, however, that testosterone should not be used for increased central fat accumulation. It just doesn't work, and it may worsen lipoatrophy.

Medscape: You delivered a review talk at CROI on the metabolic effects of antiretroviral therapy in children and adolescents.[19] Your colleagues and you have expressed concern that the long-term AEs of antiretroviral therapy may pose an even greater threat to HIV-infected children because of the potentially longer cumulative exposure to HAART compared with adults with HIV. Because of this, how does your risk-benefit analysis and management of dyslipidemias differ for your young patients?

Dr. McComsey: Overall, it's about the same as adults, except that we have limitations in the antiretrovirals that we can use. For example, because of concerns about bone disease, tenofovir is not approved for use in children. Sometimes a clinician might want to switch our young patients to a certain NRTI or PI, but you cannot because it's not available in suspension form or it's not approved. Other than these limitations, we usually think about children the same way that we think about adults, in terms of selection of NRTIs and PIs.

Medscape: How does selection of antiretroviral drugs for a first regimen or a drug switch differ for your young patients, whether they are children, adolescents, or even young adults, compared with your adult patients?

Dr. McComsey: I would say that overall there is a trend towards using NRTIs with low potential for mitochondrial toxicity. More and more pediatricians are switching away from the d-drugs. d4T [stavudine] is still used in children because it's well tolerated and it tastes decent, so a lot of pediatricians like to prescribe it. However, now more pediatricians are using abacavir, for example, as part of a first-line treatment, often combined with 3TC [lamivudine]. Tenofovir, as I mentioned, is not approved in children so that's a drug that we do not give except in salvage regimens.

Medscape: How do lipid-lowering treatment options differ for children vs adults?

Dr. McComsey: The problem of treating dyslipidemia is even more complex in children because there are FDA-approved statins for children 10 years and older, and pediatricians know from several cohorts that younger age is a risk factor for dyslipidemia, so it's very hard to know what to do for younger children with high cholesterol and triglycerides. This is why in practice you see pediatricians switching drugs vs adding hypolipemics. For example, regarding even the use of fish oil for the treatment of high triglycerides in children, we have no data. There are good data in adults to show that it does decrease triglycerides, as discussed above, but nothing yet in children.

Medscape: Are studies under way looking at lipid-lowering interventions for children?

Dr. McComsey: Not to my knowledge. It's very hard because of the small sample size. For instance, in Cleveland we have 35 HIV-infected children, and we're considered the largest center in the area. In the United States, there are so few HIV-infected children that it's very hard to do any studies in children without wide-scale collaborations.