February 14, 2003
During the last few years, the Retrovirus poster area, typically devoted to opportunistic infections, has been losing ground and shrinking, while the new, uncharted territory of metabolic complications of HIV infection and its treatment has grown bigger than ever. Unfortunately, there has yet to be any earth-shattering presentation to dramatically affect our understanding of metabolic complications.
I will summarize what I think were the highlights of the conference, in the areas of epidemiological studies, cardiovascular complications, glucose metabolism and insulin resistance, lipid abnormalities, fat redistribution and bone disease.
This conference featured the first "official" presentation of FRAM.i, ii, iii The Fat Redistribution and Metabolic Change in HIV Infection Study (FRAM) is a cross-sectional survey of 1,200 randomly selected HIV-infected patients from 16 U.S. sites plus 300 HIV-negative subjects enrolled in the ongoing CARDIA study. FRAM seeks to estimate the prevalence of fat redistribution, insulin resistance and pro-atherogenic hyperlipidemia as a result of protease inhibitor (PI) therapy. Fat distribution is measured using head-to-toe wide-slice MRI and Dual Energy X-ray Absorptiometry (DEXA) and is then compared with standard clinical anthropometric measurements and metabolic changes from a fasting blood sample. By all measures, FRAM is a large and expensive study.
The FRAM investigators presented three posters at the conference. They had no oral presentations, probably because the results were already known, as they had been "advanced" during the Barcelona Conference by Dr. C. Grunfeld. The main conclusions of this study are: lipoatrophy is the most characteristic manifestation of fat redistribution in HIV-infected individuals; fat accumulation might occur in HIV-infected individuals, but it is not statistically associated with peripheral fat loss; buffalo humps are common in HIV and non HIV-infected patients (may be a little larger in HIV-infected subjects).
The main problem with FRAM -- and its investigators recognize this -- is that its cross-sectional nature does not allow us to get a sense of what is happening to patients over time. Other longitudinal studies have been presented this year that provide a much better sense of the evolution of fat distribution after the initiation of antiretroviral therapy. ACTG 5005 (A5005s), a longitudinal study that followed patients with HIV infection from the very beginning of their initiation of antiretroviral therapy, was presented during the 4th International Workshop on Adverse Drug Reactions and Lipodystrophy in HIV (ADRL 2002) in San Diego this past fall. This study showed that subjects do gain fat after the initiation of treatment and that over time they tended to lose fat in the extremities and maintain the fat they gained in the abdomen. This contradicts to a certain degree the FRAM findings that these processes do not occur at the same time.
The main goal of A5005s was to compare the effects of nelfinavir (NFV, Viracept), efavirenz (EFV, Sustiva), AZT (zidovudine, Retrovir) + 3TC (lamivudine, Epivir) and d4T (stavudine, Zerit) + ddI (didanosine, Videx) regimens on glucose and lipid metabolism. This study also evaluated the effects of these regimens on fat distribution. When the investigators looked at the percentage change in limb (arm plus leg) fat or trunk fat compared to baseline, they saw that limb fat increased early with both NRTI regimens and PI or NNRTI regimens. At weeks 48, 64 and 80, randomization to d4T plus ddI resulted in a greater percentage decrease in limb fat. At week 80, randomization to NFV resulted in a greater percentage of loss of limb fat as well. All groups tended to gain fat in the trunk, and maintain that gain, suggesting that both processes (peripheral fat loss and central fat gain) do go on at the same time.
Longitudinal studies, especially when patients are evaluated before the initiation of treatment, are the only way to clearly explore questions about what happens with fat distribution after the initiation of antiretroviral therapy. I think we have pretty much exhausted the information we are going to get from cross-sectional studies like FRAM, with the exception perhaps of the prevalence and frequency of these problems among women, a group in which large, high-quality studies are missing. Future studies will be longitudinal in nature. One of the most interesting longitudinal studies which will provide invaluable information during the next couple of years about the longitudinal effects of therapy on fat distribution is ACTG 5142, a study of approximately 600 patients who will be randomized to receive three different class-sparing regimens (two nucleoside analogs + lopinavir/ritonavir (LPV + RTV, Kaletra), two nucleoside analogs + EFV or EFV + LPV + RTV).
FRAM also includes a significant number of women; unfortunately, their data was not presented during this conference. Hopefully that will be done formally in the near future.
Although there was a whole section of posters dedicated to metabolic problems in women at Retrovirus, most of the studies presented were cross-sectional in nature. The main conclusion of these studies is that lipodystrophy and metabolic complications also happen in women, and that some of these metabolic complications have characteristics particular to this population. Over the last few years, hospital admission trends among women have mimicked that of men, with increased hospital admissions related to cardiovascular and liver problems.iv
An interesting section of the posters presented was devoted to the metabolic problems of children with HIV infection. This is a population that HIV doctors who treat only adults tend to forget. Because of the longer life expectancy, the development of metabolic complications can potentially have a dramatic impact in the future of these children. It is not difficult to imagine that if osteoporosis is very frequent among HIV-positive children, as several studies have suggested, this problem when these children reach adulthood will be even more severe than in those diagnosed in adulthood. For example, the prevalence of osteopenia/osteoporosis among children with HIV infection varied between 42 percentv and 74 percentvi. There were no differences found between those with bone loss and those without with respect to age, ethnicity or CDC classification. In one study, the longer a child received PI therapy, the greater the risk for osteopenia or osteoporosisvii, but not in the other study. This confusing situation is similar to what was originally described in adults. Unfortunately, we still do not understand the relative contribution of therapy and HIV infection itself in the pathogenesis of these problems.
Last year Sam Bozzette generated debate in the HIV medical community when he presented the data collected on a very large cohort of HIV-infected patients cared for by the Department of Veterans Affairs (VA).viii That study showed that during the years 1995 through 2001, overall mortality among HIV-infected VA patients decreased by over 75 percent, while rates of hospital admission or death due to cardiovascular or cerebrovascular disease remained constant or showed a slight decline. That study was published almost simultaneously with this conference in the New England Journal of Medicine with an accompanying editorial by Drs. Currier and Kuritzkes.
The conclusions are contra-intuitive, since, although it seems that all of us are seeing more cardiovascular events than before, Bozzette's study tells us that it just looks that way. With the decrease in opportunistic infections, we think we are seeing more cardiovascular problems, when in fact there is the same absolute number of them. The follow up, however, is short in that study, and one would expect that cardiovascular mortality would increase in our patients as a consequence of growing age and worsening "classic" cardiovascular risk factors like LDL cholesterol (LDL-C) secondary to antiretroviral treatment. So the question is: Does antiretroviral therapy increase cardiovascular risk even when all other variables associated with an increased risk are controlled? And the answer, at least according to what we saw in this conference, seems to be yes, although the increased risk is so far not very big.
One of the most important studies presented during this meeting was The Data Collection on Adverse Events of Anti-HIV Drugs know as the D:A:D study.ix It has its own Web site (www.cphiv.dk/dad/), if the reader wants more details about this large cohort. D:A:D is a prospective multi-cohort study of HIV-infected persons under active follow up. The purpose of the study is to assess the incidence of myocardial infarction among HIV/AIDS patients receiving antiretroviral therapy. And the study's main goal is to investigate whether treatment with antiretroviral drugs is associated with the development of cardiovascular disease. The study is still ongoing. Eleven cohorts worldwide are participating, with a total current enrollment of 23,468 patients who have contributed more than 36,000 person-years of follow-up. The study is projected to continue at least until 2005, and provide almost 90,000 person-years of data. Over 36,199 person-years of follow-up in the 23,468 patients, 126 developed a myocardial infarction. The overall rate of myocardial infarction was 3.5 events per 1,000 person-years; 28 percent of the events were fatal. A total of 7 percent of all deaths were caused by myocardial infarction. The rate of myocardial infarction gradually increased with extended exposure to combination antiretroviral therapy. For example, in patients exposed to treatment for less than one year, the rate was 2.2 events per 1,000 person-years, but was 5.5 for patients exposed for more than four years. Few patients were exposed for more than six years, and hence the study does not -- at the present time -- allow for an evaluation of the effect of exposure to combination antiretroviral therapy beyond five or six years. After adjustment for age, gender and other factors -- including smoking status -- the relative rate of myocardial infarction per additional year of exposure to combination antiretroviral therapy, was 1.26 (95 percent confidence interval: 1.12-1.41, p<0.0001), i.e., a 26 percent relative increase in the rate of myocardial infarction per year of exposure. Why are the results different from Bozzette's study? Probably the main difference is the length of the follow up and a better capturing of the events in this study.
Other important studies addressing cardiovascular risk were presented during the conference, also with opposing conclusions. Two studies looked at carotid intima thickness in patients with HIV infection receiving antiretroviral therapy.x, xi High-resolution carotid ultrasonography was used to obtain measurements of the thickness of the intima and media of the carotid arteries. The methodology was different between the two studies. Previous studies have shown cross-sectional associations between common-carotid-artery intima-media thickness (IMT) and cardiovascular risk factors, the prevalence of cardiovascular disease, and atherosclerosis.
In the ACTG cross-sectional study, three types of participants enrolled. The subjects were matched for age, sex, race/ethnicity, smoking status, blood pressure and menopausal status. The groups were as follows:
No difference was found among the three groups in intima thickness. This was a surprise for all of us. We all had expected to see some worsening in study participants exposed for more prolonged periods of time to PIs.
The UCSF study had two parts, one cross-sectional (106 patients) and one longitudinal (21 patients). In the cross-sectional part, carotid IMT was independently associated with classic coronary risk factors (age, LDL-C and hypertension) and nadir CD4 count <200. No big surprises. However, the most worrisome part of this study was its longitudinal section: 21 subjects were followed for over a year. Progression of the intima thickness in this subset accelerated by tenfold compared to non-HIV infected populations, and was associated with age and duration of PI use. This study made a big jump comparing the rates of progression with historical controls that were obtained with different methods of measurements almost a decade before and the subject population was a very high risk population. But the study is important because it is the first to provide longitudinal follow up of these measurements. As usual, we need more longitudinal studies to better understand what is happening.
A group from Washington University in Saint Louis presented data relating the amount of hepatic, skeletal muscle and intra-abdominal lipid deposition (measured using HMR spectroscopy) and insulin resistance.xii The amount of fat in those tissues is a strong predictor of insulin resistance in non-obese HIV-infected people. This suggests that lipid accumulation in non-adipose tissues (liver, muscle, heart) may mediate central and peripheral insulin resistance. The cause of hepatic and skeletal muscle lipid accumulation is under investigation.
Another study of this group looked at the effects of niacin in insulin sensitivity.xiii Niacin (Niaspan) administration reduces triglycerides and LDL cholesterol, and increases HDL cholesterol. However, because of concerns of increased insulin resistance, the use of niacin has been limited in patients with HIV/AIDS and hyperlipidemia. In this study, niacin and nutrition intervention significantly decreased serum triglycerides and total cholesterol levels, but did not normalize them. On the negative side, the use of niacin was associated with moderate increases in insulin resistance, which might limit the use of this drug for the treatment of hyperlipidemia.
There was a presentation last year concerning the use of rosiglitazone for the treatment of insulin resistance and fat redistribution. The results were not very promising. This year there was a paucity of presentations using this and other drug classes for the treatment of insulin resistance. Perhaps this summer, at the 2nd IAS Conference on HIV Pathogenesis and Treatment, the Australian ROSEY study will be presented. This study is evaluating the role of rosiglitazone for the treatment of lipoatrophy associated with HIV infection and its treatment. ACTG is also conducting a study comparing the use of metformin versus rosiglitazone or both for the treatment of insulin resistance and visceral adiposity (ACTG 5082).
Hyperlipidemia was also the focus of a few presentations. In one study researchers looked at a new formulation of amprenavir (APV, Agenerase) and in another study nevirapine (NVP, Viramune) was compared to EFV. In both, researchers examined the effects of a patient's first antiretroviral regimen on lipid parameters. It seems like pharmaceutical companies may now try to claim that their drug is more "lipid friendly" than another "competing" drug.
The new formulation of APV is called GW433908 or "908" and it seems to be easier on lipids than the classic NFV formulation.xiv This open-label, randomized study in ART-naive subjects compared the efficacy and safety of 908 to NFV over 48 weeks. The new APV formulation was slightly better than NFV in changes in lipid parameters like cholesterol and triglycerides, but both drugs increased the levels of these markers. 908 seemed to have an edge in antiviral potency.
In a different study, NVP use was associated with a better lipid profile than EFV use, although the data presented was evaluated in the non-fasting samples of the large van Leth "2NN" study.xv NVP and EFV are associated with slight increases in total cholesterol and triglycerides and marked increases in HDL cholesterol (the "good cholesterol"); these changes were more significant in the NVP-containing arms (that were analyzed together) than in EFV-containing arms. The ratio of total cholesterol to HDL cholesterol (a marker of cardiovascular risk) was better in the NVP-taking patients. The study confirmed the prevailing idea that NVP is more "lipid friendly" than EFV. As NVP and EFV seem to be pretty much equipotent, this might be a consideration at the time of initiation of an antiretroviral regimen, particularly in patients with multiple cardiovascular risk factors.
Dr. Grunfeld's group presented data on lipid and glucose metabolism in HIV-seronegative individuals receiving LPV + RTV, the most popular PI nowadays.xvi They used a similar design to their pivotal study that evaluated the effects on glucose metabolism of indinavir (IDV, Crixivan).xvii The results of the current study were somewhat surprising: The effects of LPV + RTV were clearly quite different from IDV. Treatment with four weeks of LPV + RTV in HIV-seronegative men caused an increase in triglycerides and free fatty acids, and a slight deterioration of glucose tolerance at two hours, but there was no significant changes in insulin mediated glucose disposal rate measured by CLAMP. The main conclusion of this study is that not all PIs are created equal, and that they have different effects in glucose and lipid metabolism. It is also clear that there is a theme in metabolic abnormalities associated with the use of PIs: They significantly affect glucose and lipid parameters. The design of the two studies utilizes non-infected individuals, which allows us to separate the effects of the antiretroviral drugs from the effects of HIV infection itself. But this "clean" approach also has problems. These drugs are never used alone in HIV-infected individuals. They are always used in combination with nucleoside analogs, and there is in vitro data that suggest that when these drugs are used in combination with NRTIs, they might have different metabolic effects. Nothing is as simple as it seems.
Very few evaluations of drugs for the treatment of hyperlipidemia were discussed. (See above for the preliminary evaluation of niacin.) The largest ACTG trial that compared pravastatin to fenofibrate was presented during the fall of 2002 at the 40th Annual Meeting of the Infectious Diseases Society of America (IDSA 2002) meeting and has already been discussed on The Body's site.
In an excellent summary (metabolic complications of antiretroviral therapy symposia), Dr. Reiss synthesized the current approach to hyperlipidemia associated with HIV and its treatment:
Two small studies evaluated the role of adiponectin as a marker of lipodystrophy and insulin resistance associated with HIV infection.xix, xx Discovered only a few years ago, diponectin is an adipocyte-derived hormone. Early after its discovery, it was found that the adiponectin gene is exclusively expressed in adipose tissue, which, therefore, represents the only source of the circulating protein. However, unlike leptin, its tissue expression and its plasma concentration are decreased (not increased) in obesity and type 2 diabetes. So, in general, lower levels are seen in individuals with insulin resistance. Adiponectin may prevent atherosclerosis by suppressing endothelial adhesion molecules. Adiponectin levels were markedly reduced among HIV-infected patients with lipodystrophy when compared to controls. This is a worrisome trend that is probably associated with insulin resistance and, in the long run, increased cardiovascular risk. Two years ago two studies in Nature Medicine showed that mice with insulin resistance treated with adiponectin improved dramatically. Several drug companies are working on analogs of adiponectin for the treatment of diabetes and obesity. These drugs might be useful in the future to treat lipodystrophy associated with HIV infection.
This year was surprisingly thin in studies evaluating interventions for the treatment of lipodystrophy. Several posters evaluated different plastic surgery interventions.xxi, xxii, xxiii, xxiv, xxv, xxvi The intervention that seems to have the most data to support it is the injection of polylactic acid. The aesthetic results of these injections are excellent, judging from the photographs presented during the meeting. This procedure is expensive for patients with limited economic resources and it obviously does not fix the underlying mechanism of the problem. Other interventions discussed during the meeting included aspirations of the "buffalo pads" using ultrasound and autologous transplant of fat from other areas of the body to the cheek, done by an Italian group in 60 patients. Unfortunately, some of these surgical solutions only offer transient benefits, with a significant proportion of patients returning to their baseline situation and requiring reinterventions.
The only proven intervention that seems to increase subcutaneous fat is the substitution of d4T by other nucleoside analogs such as abacavir (ABC, Ziagen). Several studies presented during last year's conference showed that this intervention has modest but significant effects in the treatment of patients with lipoatrophy.xxvii
Two postersxxviii, xxix and an oral presentationxxx evaluated what happened to the amount of mitochondrial DNA in different bodily compartments, including muscle and peripheral blood mononuclear cells in patients who switched off d4T. Once again, a relationship between the use of d4T and a decrease in mitochondrial DNA was shown. These are just associations and do not prove causality, but the evidence is mounting against d4T and its role in mitochondrial toxicity as a mediator in peripheral lipoatrophy. PIs originally blamed for this complication, do not seem to play a primary role, as yet another study showed that long term substitution of the PI component of an antiretroviral combination does not improve peripheral fat loss.xxxi
Another interesting study, from a Dutch group led by Dr. Reiss, evaluated what happened to peripheral fat distribution when patients switched off nucleoside analogs.xxxii The HIV-NAT 009 study was a single arm, open-label study of ritonavir (RTV, Norvir)-boosted IDV 800/100 mg plus EFV 600 mg QD in patients who had failed dual or triple combination NRTI therapy. Patients were evaluated by DEXA scanning at week 0 and 48.
This is the first study that has really examined the impact on body composition of discontinuing all nucleoside analogs. Switching from failing nucleoside therapy to boosted IDV and EFV led to significant increases in visceral and subcutaneous fat but also to significant decreases in lean body mass, an unexpected and disappointing finding that requires further study. The observed changes are consistent with the hypothesis that the NRTI class is associated with lipoatrophy, which improves after switching to an NRTI-sparing regimen. The changes also support the hypothesis that the PI class is associated, through a separate mechanism, with fat accumulation, because these patients had increases in intra-abdominal fat. Lipid values, not surprisingly, rose significantly after the switch.
Another study that I found interesting came from the University of Alabama and examined the economic costs of having lipodystrophy.xxxiii It showed that for any given CD4 count, the care of patients with lipodystrophy tended to be more expensive, because of the associated increased costs of medication and close monitoring. So, preventing lipodystrophy, choosing or developing antiretroviral regimens that are not, or will not be, associated with this problem, can potentially save money in the long run. And, as we all know, costs are an important issue in this era of limited health care resources.
Another small study from Washington University in Saint Louis evaluated the role of alendronate in the treatment of osteopenia/osteoporosis associated with HIV disease or its treatment.xxxiv Thirty-one patients with evidence of osteopenia or osteoporosis were randomized to receive vitamin D (400 IU every day) and calcium (1,000 mg every day) with or without alendronate 70 mg weekly. Those who received alendronate demonstrated significantly better improvements in bone density in the lumbar spine by week 48 of the trial (5.2 percent vs. 1.3 percent). The study was powered to detect differences greater than 3 percent within the arms of the trial. As expected, there were no significant changes in other regions like the hip. Alendronate was well tolerated. Markers of bone formation and resorption decreased. This study suggests that alendronate is useful in the management of osteopenia and osteoporosis, but larger studies that include a significant number of women, are necessary to evaluate the generalization of these findings.
|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.|