|Home | About | Journals | Submit | Contact Us | Français|
Morphologic and metabolic abnormalities, including subcutaneous adipose tissue wasting, central adipose tissue accumulation, dyslipidemia and disorders of glucose metabolism are common among HIV-infected patients receiving highly active antiretroviral therapy (HAART) and contribute to the risk of cardiovascular disease in this population. The pathogenesis of these disorders is due to complicated interactions between effects of chronic HIV infection, HAART medications, and patient factors, including genetic susceptibility. HAART has transformed HIV into a chronic condition for many patients and as a result the majority of HIV-infected patients in many areas of the developed world are ≥50 years. Since metabolic and cardiovascular diseases increase with aging, knowledge of the optimal management of these conditions is essential for practitioners caring for HIV-infected patients, including endocrine subspecialists. This Review highlights the clinical management of these disorders, focusing on the most recent evidence regarding the efficacy of treatment strategies, newly available medications and potential interactions between HAART medications and medications used to treat metabolic disorders.
With the introduction of highly active antiretroviral therapy (HAART) in the mid-1990s, morbidity and mortality have decreased dramatically among HIV-infected patients.1,2 In the late 1990s, however, multiple morphologic and metabolic abnormalities were observed among HAART-treated patients with well-controlled HIV, including subcutaneous adipose tissue wasting (lipoatrophy), central adipose tissue accumulation (lipohypertrophy), severe dyslipidemia and abnormalities of glucose metabolism.3 The pathogenesis of these disorders is multifactorial, with varying contributions from some HAART components, chronic HIV infection, as well as behavioral and genetic factors unrelated to HIV infection. Although the direct adverse effects of HAART have lessened over the past 15 years with the introduction of less toxic HAART regimens, metabolic disorders remain a major concern for HIV-infected patients, especially since their consequences, such as cardiovascular disease, are expected to increase with the aging of the HIV population.
This Review outlines current understanding of the pathogenesis, consequences and optimal management of metabolic abnormalities among HIV-infected patients. The article will highlight areas of particular interest to endocrinologists and other practitioners working outside the field of HIV medicine, including the role of HAART in metabolic disorders, the most recent information regarding specific therapies for metabolic complications, and potential interactions between drugs used to treat metabolic problems and HAART.
Body composition changes in HIV-infected individuals are often grouped under the term lipodystrophy, which includes both subcutaneous lipoatrophy and central lipohypertrophy. Despite sharing common risk factors, including older age and low nadir CD4 T-lymphocyte cell count,4 the pathogenesis of these two disorders is distinct and better discussed and evaluated separately. Co-occurrence of lipoatrophy and central lipohypertrophy in the same patient is best described as mixed lipodystrophy.
Diffuse loss of subcutaneous adipose tissue that is most apparent in the face, buttocks, legs and arms is the most frequent morphologic change among HIV-infected patients receiving HAART, and affects up to one third of HAART-treated patients.5–7 Facial lipoatrophy, which begins in the malar region and extends to the buccal and temporal regions with increasing severity (Figure 1), can reduce quality of life and compromise adherence to HAART.8 Subcutaneous adipose tissue wasting is also associated with insulin resistance,9 dyslipidemia10 and increased systemic inflammation.11
Exposure to the thymidine analogue nucleoside reverse transcriptase inhibitors (NRTIs) stavudine and zidovudine12 is the most important risk factor for the development of lipoatrophy. These medications cause mitochondrial dysfunction in adipocytes, which leads to adipocyte apoptosis.13 Other components of HAART may further worsen lipoatrophy, including certain protease inhibitors, such as nelfinavir.14,15
Controversy exists regarding the specific effect of the non-nucleoside reverse transcriptase inhibitor (NNRTI) efavirenz on limb adipose tissue. Findings of a multicenter clinical trial investigating the initiation of HAART in HIV-infected patients suggested that efavirenz was associated with a higher incidence of lipoatrophy than the protease inhibitor combination of lopinavir with ritonavir (as assessed by dual-energy x-ray absorptiometry [DXA]).15 However, this effect of efavirenz was not observed in another study of the initiation of HAART, as compared with the combination of atazanavir plus ritonavir.16 In addition, trials that have compared efavirenz with HAART regimens not based on protease inhibitor combinations, or switch studies that have compared the effects of a change to efavirenz versus continuation of protease inhibitor combinations have not demonstrated worsening lipoatrophy with efavirenz.17,18 Taken together, these findings argue against an independent effect of efavirenz on adipose tissue wasting.
Other patient factors, such as male sex, older age and more advanced HIV disease have also been associated with the development of lipoatrophy4,5,7 and certain mitochondrial haplotypes and nuclear genetic polymorphisms may increase susceptibility.19,20 The low incidence of lipoatrophy in HIV-infected individuals not receiving thymidine analogue NRTIs further supports the primary role of these medications in its pathogenesis.15
The most effective treatment for lipoatrophy is avoidance of or switching from treatment with thymidine analogue NRTIs.21–23 Fortunately, with the availability of other NRTIs without adverse effects on body composition (for example, abacavir and tenofovir disoproxil fumarate) use of thymidine analogue NRTIs has decreased considerably in the developed world.24
While an increase in subcutaneous adipose tissue occurs after switching from thymidine analogue NRTIs to abacavir or tenofovir disoproxil fumarate, most evidence suggests that full restoration does not occur.22,25,26 In the longest of these switch studies, for example, the mean gain in extremity fat mass over 104 weeks was 1.2 kg when stavudine was replaced with abacavir.21 Although this gain was a mean 35% increase in extremity fat mass over the baseline measurement, it represented only a fraction of the estimated 5 to 6 kg of extremity fat that had presumably been lost during stavudine use and self-perception of the extent of lipoatrophy did not improve with the gain. For these reasons, although incident cases of lipoatrophy have decreased markedly in the past decade as stavudine and zidovudine are no longer in widespread use, lipoatrophy remains a prevalent problem among patients who have been previously exposed to thymidine analogue NRTIs.
Pharmacologic treatment of lipoatrophy has had disappointing results (Table 1). Initial enthusiasm existed regarding the use of thiazolidinediones in the management of HIV-associated lipoatrophy, bolstered by their known effects on adipocyte differentiation and proliferation,27 their efficacy in congenital lipoatrophy28 and promising results in initial preclinical and clinical studies.29,30 Over the past 10 years, however, randomized clinical trials in HIV-infected individuals have had mixed results.30–34 A meta-analysis of randomized clinical trials of HIV-associated lipoatrophy published in 2010 showed overall no benefit in five studies that examined the effect of rosiglitazone and a modest increase in extremity fat mass (0.35 kg over 48 weeks) in the one study that examined pioglitazone.35,36 The primary endpoint in these studies was DXA, a sensitive measurement of extremity fat. Even among the trials that demonstrated an increase in extremity fat by DXA, however, the change was not perceptible to the patient. Given the inconsistent benefits of thiazolidinediones in clinical trials and the potential adverse effects of long-term use, thiazolidinediones cannot be recommended for the management of lipoatrophy in HIV-infected individuals without type 2 diabetes mellitus.
The benefits of other initially promising therapies have not been substantiated on further study. The pyrimidine uridine is depleted in HIV-infected patients with lipoatrophy37 and in in vitro models, uridine supplementation has been shown to prevent and reverse stavudine-induced mitochondrial toxicity.38 Despite these results and positive pilot data,39 two large, randomized clinical trials have shown no benefit of a uridine-rich supplement for the treatment of lipoatrophy in HIV-infected individuals.40,41 Similarly, the unexpected benefits of the 3-hydroxy-3-methyl-glutaryl-CoA reductase inhibitor, pravastatin, on extremity fat recovery in a small, randomized clinical trial42 were not confirmed in two more recent randomized clinical trials.40,43 Neither uridine nor pravastatin can, therefore, be recommended for the treatment of lipoatrophy.
For patients with facial lipoatrophy that causes psychological distress, facial fillers can improve quality of life and decrease anxiety and depression symptoms.44–46 Both permanent and absorbable compounds have been used and should be administered by an experienced plastic surgeon or dermatologist. The use of permanent filler may be associated with infection, which may necessitate antibiotic treatment or removal of the compound.47 Fat transplantation is another approach that has been used, but may be associated with fat hypertrophy of the transplanted fat and an undesired clinical result.48
Recombinant leptin has been investigated as an approach to improve metabolic abnormalities among HIV-infected patients with lipoatrophy.49,50 Similar to the effects of recombinant leptin in congenital lipoatrophy,51 its administration in HIV-infected patients with lipoatrophy and hypoleptinemia has been associated with increased insulin sensitivity and reduced triglyceride levels in two small pilot studies.49,50 One study showed a 32% reduction in visceral adipose tissue, without a change in subcutaneous adipose tissue.50 These findings provide a rationale for further investigation of the effects of recombinant leptin, particularly among patients with mixed lipodystrophy and metabolic abnormalities.
A subset of HIV-infected patients develop abnormal accumulation of adipose tissue, termed lipohypertrophy, typically in the setting of successful HAART. The cardinal manifestation of lipohypertrophy is a relative excess of abdominal fat in the visceral adipose tissue compartment (Figure 2). A minority of patients have coexisting dorsocervical fat pad enlargement without evidence of Cushing syndrome (Figure 1). Ectopic adipose tissue deposition in the liver, epicardium and muscle has also been associated with the quantity of visceral adipose tissue.52,53 Physiologically, the ectopic adipose tissue deposition may result in adverse metabolic effects, such as insulin resistance, as a result of lipotoxicity.54
Lipohypertrophy is generally a clinical diagnosis that can be supported by anthropometric measurements. Clinical trials that have evaluated agents in the treatment of lipohypertrophy have used waist circumference and waist-to-hip ratio criteria as entry criteria.55,56 However, the waist-to-hip ratio will also be affected by lipoatrophy in the hip region and, therefore, may not be an accurate assessment of central adiposity in an HIV-infected persons with mixed lipodystrophy. Waist circumference criteria used to define the metabolic syndrome, which differ by sex, country or ethnicity,57 may be reasonable cut-offs, but may underestimate the volume of visceral adipose tissue in a patient with subcutaneous lipoatrophy of the abdomen. Cross-sectional imaging by CT or MRI can quantify visceral adipose tissue in research settings, but a paucity of normative data limits the utility of imaging in a clinical setting. Similarly, DXA can measure changes in central versus appendicular adipose tissue in research settings but cannot distinguish between subcutaneous and visceral adipose tissue accumulation and hence is of limited clinical utility. Abdominal imaging may be useful in select cases, such as in the setting of liver disease, to distinguish ascites from lipohypertrophy.
The pathogenesis of lipohypertrophy remains elusive. No definitive evidence implicates specific antiretroviral drugs, although the AIDS Clinical Trials Group A5224s study recently reported greater increases in trunk adipose tissue and a trend towards greater increases in visceral adipose tissue in antiretroviral-naïve patients randomly allocated to atazanavir plus ritonavir compared with efavirenz.58 In this study, which had a factorial design, trunk and visceral adipose tissue did not differ by randomization to abacavir plus lamivudine versus tenofovir plus emtricitabine. Generalized weight gain and immune reconstitution by the control of HIV replication may contribute to the pathogenesis of lipohypertrophy. Debate exists about whether the gain of central adipose tissue after initiation of HAART represents a return to health or a pathologic process. Two specific pathogenic mechanisms have been hypothesized. The first involves the dysregulation of fatty acid metabolism due to HIV or antiretroviral agents, which leads to selective deposition in the visceral adipose tissue depot in the setting of blockage of the subcutaneous depot.3,59 The second postulates increased local cortisol concentrations owing to aberrant conversion from cortisone.60 Relative growth hormone deficiency may also contribute to excess visceral adipose tissue.61–63 Gene expression studies of dorsocervical fat pad tissue from small numbers of HIV-infected patients demonstrate increased expression of mitochondrial brown fat uncoupling protein 1 relative to that from healthy control individuals, which suggests that the adipose tissue has acquired some phenotypic features of brown fat.64 The potential consequences of lipohypertrophy in the general population include an increased risk of coronary heart disease and type 2 diabetes mellitus.65,66 These adverse consequences may relate in part to associated insulin resistance and dyslipidemia as part of a metabolic syndrome.67 Specific data on the long-term consequences of excessive visceral adipose tissue are limited in the HIV population. Abnormal adipose tissue distribution, however, has been linked to increased stigma, diminished quality of life, and possibly decreased adherence to HAART.68–70 An analysis of 922 HIV-infected individuals found that patients in the highest tertile of volume of visceral adipose tissue had twice the odds of death at 5 years of follow-up compared with the lowest tertile after adjustment for HIV-related factors, cardiovascular risk factors, inflammatory markers and renal function.71 Increased volume of visceral adipose tissue may also contribute to hepatic steatosis and decreased BMD.72–74 These consequences merit consideration of interventions to reduce visceral adipose tissue.
Randomized controlled studies of switching specific antiretroviral agents have not had favorable effects on the volume of abdominal adipose tissue.19 An exception is a switching study of 15 HIV-infected patients with elevated insulin, total cholesterol or triglyceride levels treated with the protease inhibitor combination of lopinavir and ritonavir. This study reported a 15% reduction in visceral adipose tissue at 6 months in patients randomly allocated to switch their protease inhibitor combination to atazanavir and ritonavir versus an 11% increase in patients who did not change regimens.75 These data require confirmation before such a strategy can be recommended.
Several small studies of lifestyle interventions, such as supervised aerobic and progressive resistance training, have reported modest, if any, reductions in abdominal adipose tissue but have yielded some improvements in associated metabolic abnormalities.77–79 While these trials have been disappointing, lifestyle interventions are safe and appropriate to recommend as general health measures for patients with lipohypertrophy.
A randomized placebo-controlled trial of metformin in HIV-infected patients with elevated waist-to-hip ratios and impaired glucose tolerance or hyperinsulinemia demonstrated modest reductions in abdominal adipose tissue.80 The loss of adipose tissue, however, is not selective for the visceral compartment, and subcutaneous lipoatrophy can worsen with use of metformin.80,81 Other trials of metformin in HIV-infected patients have not had favorable effects on visceral adipose tissue.34 While data do not support recommending metformin as a specific treatment for lipohypertrophy,81 it is a reasonable pharmacologic intervention for abnormal glucose homeostasis in patients with lipohypertrophy who have impaired fasting glucose, impaired glucose tolerance or type 2 diabetes mellitus and mild or no lipoatrophy.
Visceral adipose tissue area is inversely proportional to testosterone levels in men in the general population,81 and hypogonadism is common in HIV-infected men.83,84 However, a placebo-controlled trial of transdermal testosterone gel use in HIV-infected men with elevated waist-to-hip ratios or waist circumference and low testosterone levels reduced subcutaneous adipose tissue but not visceral adipose tissue,85 which suggests that testosterone repletion is not indicated solely to improve body composition in this population.
Recombinant human growth hormone (rhGH) has been used to treat AIDS-related wasting and was known to be lipolytic in this setting.86 Supraphysiologic doses of rhGH (4 mg daily or every other day) reduced visceral adipose tissue by 17–20% and improved lipid profiles in placebo-controlled trials in HIV-infected patients with abdominal obesity.55,87,88 Subcutaneous adipose tissue was reduced only modestly in these trials. The long-term safety of such high doses of rhGH, however, is uncertain given the associated worsening of insulin sensitivity, increased risk of developing type 2 diabetes mellitus and a high incidence of arthralgias and peripheral edema from increased tissue turgor. Lower doses of rhGH also reduce visceral adipose tissue in a dose-dependent fashion with improved tolerability.89,90 In a novel trial, titration of rhGH dosing to achieve insulin-like growth factor 1 (IGF-1) levels in the upper quartile of the normal range (average rhGH dose of 0.33 mg per day) resulted in an 8.5% reduction in visceral adipose tissue over 18 months with modest adverse effects on glucose homeostasis.91 Unfortunately, visceral adipose tissue re-accumulates within several months of stopping rhGH,55,92 and a maintenance dose of 2 mg every other day of rhGH in one study was only marginally effective.55
The most promising intervention for lipohypertrophy is tesamorelin, a synthetic analogue of growth hormone releasing hormone that yields more physiologic levels of IGF-1 than high-dose rhGH due to preservation of the negative feedback loop at the level of the pituitary. Two phase III, placebo-controlled trials of approximately 400 patients per study have reported 11–15% reductions in visceral adipose tissue at 6 months in the tesamorelin arms.93,94 In one of these trials, re-randomization of patients in the tesamorelin arm to continue tesamorelin for an additional 6 months versus placebo resulted in an overall reduction in visceral adipose tissue of 17.5% in the active arm while visceral adipose tissue returned to baseline in the placebo arm.94 Limb adipose tissue and subcutaneous adipose tissue were reduced only modestly by tesamorelin, while lipid profiles tended to improve. The most common adverse effects were injection site erythema, pruritus and hypersensitivity skin reaction. About half of the patients treated with tesamorelin in a phase II trial developed IgG antibodies against tesamorelin at week 26, the clinical significance of which is unknown.93 No differences were found in IGF-1 levels or changes in visceral adipose tissue by IgG status. Mild deterioration in glucose tolerance has also been observed with tesamorelin therapy.95
Many unanswered questions exist concerning the optimal use of tesamorelin, including whether intermittent therapy or an induction–maintenance approach could be clinically useful or whether long-term administration is required to maintain the reduction in visceral adipose tissue. The long-term safety of elevating IGF-1 levels in this setting is unknown, but the potential associations between IGF-1 levels and risk of malignancies in the general population96,97 and the increased predisposition of HIV-infected patients to certain malignancies are of concern.98 A summary of the treatment strategies for central lipohypertrophy is presented in Table 1.
Untreated, advanced HIV infection is characterized by hypertriglyceridemia, low concentrations of HDL cholesterol and LDL cholesterol and a predominance of small LDL particles.99,100 Untreated, less advanced HIV disease may be associated with lower concentrations of HDL particles101,102 in association with elevated inflammatory markers.102
Effective HAART seems to result in a return of total cholesterol and LDL cholesterol to the premorbid state, in what has been termed a return to health phenomenon.103 This phenomenon probably only constitutes a part of what is observed with initiation of HAART since specific drug regimens can result in different effects on lipids. HDL cholesterol levels may increase in patients treated with HAART, although the effects have been variable in different studies perhaps as a function of the treatment regimen.
The effects of individual antiretroviral agents on lipid parameters have been difficult to ascertain since drugs are typically given in three-drug combinations. Limited data from healthy volunteers who received short courses of single drugs indicate that several protease inhibitor drugs elevate triglyceride levels in the absence of HIV.104,105 Most protease inhibitors are taken with a low dose of ritonavir, a protease inhibitor that is used as a pharmacokinetic enhancer to boost levels of the co-administered active protease inhibitor. Ritonavir-boosted protease inhibitors commonly result in elevations of triglyceride levels. LDL cholesterol levels increase with most contemporary HAART regimens, which may largely be due to the patient’s improved health status.106 Use of the NNRTIs efavirenz and nevirapine is associated with significant elevations in levels of HDL cholesterol.107 Tenofovir, a nucleotide analogue, seems to reduce levels of non-HDL cholesterol, LDL cholesterol and total cholesterol even when added to a fully suppressive HAART regimen.108
Guidelines based on expert opinion advocate measurement of fasting lipids in HIV-infected patients annually and before and after changing HAART regimens.109 Management of dyslipidemia and targeting of lipid levels follow guidelines from the general population.109,110 Risk stratification based on the Framingham risk equation seems to underestimate coronary heart disease event rates in HIV-infected patients but only modestly;111 investigators have derived an HIV-specific equation that requires further validation.112 Whether HIV-infected patients should be managed more aggressively owing to their chronic inflammatory state and perceived increased risk of coronary heart disease is an open question.
If lifestyle intervention fails to achieve appropriate lipid goals, clinicians have the option of switching HAART regimens or adding lipid-lowering therapy. Randomized controlled trials have demonstrated beneficial effects on lipid profiles from switching certain protease inhibitors to alternative agents.23,113–118 Clinicians and patients, however, may be reluctant to switch effective HIV therapies due to concerns about potential loss of virologic control and new adverse effects. Some clinicians favor the addition of a lipid-lowering agent despite the added pill burden and cost since the clinical efficacy of some such agents are proven in the general population and some may have pleiotropic effects. The efficacy of lipid-lowering therapy in HIV-infected individuals may, however, be lower than that in the general population.119
When hypertriglyceridemia predominates, fibrates, fish oil supplementation or extended release niacin are all effective in HIV-infected patients.120–125 Similarly, statins are effective at lowering LDL cholesterol,120,126–128 with the caveat that important drug–drug interactions exist between some statins and antiretroviral drugs (Table 2). Lovastatin and simvastatin should not be co-administered with protease inhibitors because of the potential for substantial elevations in statin exposure.129 Pravastatin exposure does not increase with co-administered protease inhibitors, with the exception of darunavir; darunavir and pravastatin should, therefore, not be co-administered.130,131 Atorvastatin and rosuvastatin exposure increase modestly with some protease inhibitors and can be administered safely starting at low doses.129,132,133 Data on efavirenz use suggest that NNRTIs lower exposure to statins, which may necessitate the use of increased doses.134
The risk of insulin resistance and type 2 diabetes mellitus is increased in HIV-infected patients receiving HAART compared to HIV-seronegative control individuals in some cohorts.135,136 In the Multicenter AIDS Cohort Study, for example, the incidence of type 2 diabetes mellitus was more than four times greater in HIV-infected men receiving HAART than in HIV-seronegative control participants.135 Certain protease inhibitors have been shown to reversibly induce insulin resistance through the inhibition of glucose translocation through GLUT4.137 The thymidine analogue NRTIs, which lead to lipoatrophy as reviewed above, also have direct effects on glucose metabolism by inducing mitochondrial dysfunction.138,139 Exposure to HAART may also lead to insulin resistance indirectly, through effects on regional body adipose tissue changes, inflammation and adipokine and free fatty acid dysregulation.
Chronic HIV infection and the resulting systemic inflammation can also lead to insulin resistance and type 2 diabetes mellitus among HIV-infected individuals. In the Multicenter AIDS Cohort Study, higher levels of insulin resistance markers were observed in all groups of HIV-infected men than in HIV-uninfected control participants, even among those who were not receiving HAART, which suggests an effect of HIV infection itself.140 In another study, markers of systemic inflammation 48 weeks after HAART initiation, particularly increased tumor necrosis factor activity, were associated with an increased risk of incident type 2 diabetes mellitus.141
Screening for diabetes mellitus in HIV-infected patients is currently recommended before HAART initiation, from 3–6 months after initiation and annually thereafter.142,143 In the general population, HbA1c has recently been recommended for screening of diabetes mellitus.144 In HAART-treated, HIV-infected individuals, however, HbA1c may underestimate glucose exposure,145 which could potentially lead to underdiagnosis if HbA1c is used as the sole diagnostic criteria. In one study, HbA1c underestimated average blood glucose by approximately 1.7 mmol/l (approximately 1.0 HbA1c percentage points),146 although this was not confirmed in a study of HIV-infected women.147 Thereasons underlying the potential inaccuracy of HbA1c in HIV-infected individuals require further study. Factors associated with discordance between actual and predicted HbA1c include higher mean corpuscular volume, current NRTI use, and, in particular, abacavir use.146 Because of the potential inaccuracy of HbA1c, fasting plasma glucose test or 2-h plasma glucose values after an oral glucose challenge should continue to be the diagnostic tests of choice for HIV-infected patients.
The choice of treatment for type 2 diabetes mellitus should follow recommendations in the general population. Lifestyle modification is an essential part of diabetes therapy for HIV-infected individuals and results in a significant reduction in HbA1c.148 As in the general population, metformin should be considered first-line therapy in those without a contraindication, particularly in those with central lipohypertrophy in whom reductions in visceral adipose tissue may be enhanced.149 In patients with lipoatrophy and type 2 diabetes mellitus, pioglitazone may be a preferred option, since metformin may exacerbate lipoatrophy. However, the potential benefits of pioglitazone on glycemic control and modest effects on subcutaneous adipose tissue needs to be weighed against the possibility of fluid retention, increased risk of bladder cancer, and decreased BMD and fracture.150 The last two considerations are especially important in that the risk of osteoporosis and fracture is higher in HIV-infected individuals compared with HIV-uninfected individuals.151,152 No known pharmacologic interactions exist between HAART and anti-diabetic medications. One exception is the newly approved dipeptidyl peptidase IV inhibitor, saxagliptin, which is metabolized by cytochrome P450 3A4 and, therefore, should be avoided or the dosage should be reduced in ritonavir-treated patients.
In HIV-infected patients with impaired fasting glucose or impaired glucose tolerance, lifestyle modification is the first-line treatment. Similar to the general population, pharmacologic treatment of prediabetes with metformin can be considered, especially in those patients <60 years of age with a BMI≥35 kg/m2.153 Among HIV-infected patients, it is also reasonable to consider metformin in a patient with impaired fasting glucose or impaired glucose tolerance who also has central adipose tissue accumulation.
Metabolic abnormalities in HIV-infected patients may contribute to an increased risk of cardiovascular disease, compared with that in the general population.157 In addition, certain components of HAART and chronic HIV infection may have more direct effects on cardiovascular risk.155–157 Exposure to certain protease inhibitors and abacavir have been associated with cardiovascular events, but these data are controversial.158,159
Current data also indicate that cardiovascular events in HIV-infected patients are closely associated with traditional cardiovascular risk factors,160 which may be more common in the HIV-infected population. Cigarette smoking, for example, is highly prevalent in HIV-infected cohorts.161,162 and is closely associated with cardiovascular events. Importantly, smoking cessation also leads to a reduction in cardiovascular events among HIV-infected individuals and is, therefore, a crucial target for intervention.162 Hypertension is another modifiable cardiovascular risk factor that is more prevalent in some HIV-infected populations163 and should be managed similarly to that in the general population.164 Aspirin should be given for secondary prevention of cardiovascular disease and for primary prevention in accordance with current recommendation in the general population until more data are available.165 Further research is required to determine whether the risks and benefits of aspirin therapy are similar to those in the HIV-uninfected population.
Data in the past few years indicate that residual inflammation and immune activation may occur in patients with suppressed HIV viremia receiving HAART and in so-called ‘elite controllers’ who are able to control HIV replication without the use of antiretroviral therapy.166–168 The residual inflammation may be due to the host response to persistent viral reservoirs, the loss of integrity of the gut mucosa owing to CD4 depletion and resulting microbial translocation,169 or the reactivation of other latent viruses, such as cytomegalovirus.170 Whether this inflammatory response contributes to the pathogenesis of some of the metabolic complications discussed above is uncertain. Strategies aimed at mitigating end-organ complications, such as atherosclerosis, by targeting inflammation (for example with statins) are being studied.
Limited studies of drugs in new classes of antiretroviral agents, such as integrase inhibitors and CCR5 inhibitors, suggest that they do not adversely affect lipids or insulin sensitivity, although data are lacking on body composition. Ongoing clinical trials with metabolic and body adipose tissue assessments, including some with measurements of carotid intima–media thickness and brachial artery reactivity, will clarify whether contemporary HAART regimens have more favorable safety profiles than previously used HAART regimens.
The potential for uncontrolled HIV replication and immune dysregulation to contribute to the pathogenesis of complications such as atherosclerosis has contributed to changes in treatment guidelines that favor earlier initiation of HAART (namely, at a higher CD4 cell count than previously). An ongoing randomized trial of immediate versus deferred initiation of HAART for patients with CD4 cell counts >500 cells/mm3 may shed light on the risks versus benefits of earlier initiation of HIV therapy in relation to metabolic complications.
With the advent of effective antiretroviral therapy, the life expectancy of HIV-infected individuals is approaching that of their HIV-uninfected counterparts. Owing to effects of HAART and chronic HIV infection, body adipose tissue changes, dyslipidemia and glucose abnormalities are common among HIV-infected patients. However, HAART-specific metabolic toxicities have decreased with the availability of agents with fewer glucose and lipid effects. Nonetheless, the consequences of metabolic abnormalities, which include an increased risk of type 2 diabetes mellitus and cardiovascular disease, remain a major concern among HIV-infected patients, particularly since these diseases increase in frequency with aging. Current strategies to decrease cardiometabolic risk in HIV-infected patients mirror those in the general population and include aggressive treatment of modifiable risk factors. Potential interactions between HAART and medication to treat metabolic disease, such as certain statin medications, should also be kept in mind when treating HIV-infected individuals. The long-term benefits and risks of tesamorelin, the newly approved agent to treat central lipohypertrophy require clarification. Further research is needed to understand the role of chronic inflammation in the pathogenesis of metabolic and cardiovascular disease, to evaluate the effect of new antiretroviral agents and to identify other treatment strategies.
The Review was based on the authors’ personal collection of publications and conference abstracts concerning lipoatrophy, lipohypertrophy, insulin resistance, dyslipidemia and cardiovascular risk among HIV-infected individuals. Searches based on these search terms were made in PubMed up to December 2010. In addition, supplemental PubMed searches were performed using the terms “thiazolidinedione AND HIV” and “growth hormone AND HIV AND lipodystrophy”. Only English language articles were selected for this Review.
T. T. Brown declares associations with the following companies: Abbott, EMD Serono, Gilead, GlaxoSmithKline, Merck, Theratechnologies, Tibotec, ViiV Healthcare. M. J. Glesby declares associations with the following company: Pfizer. See the article online for full details of the relationships.
Todd T. Brown, Division of Endocrinology and Metabolism, Johns Hopkins University, 1830 East Monument Street, Baltimore, MD 21287, USA.
Marshall J. Glesby, Division of Infectious Diseases, Weill Cornell Medical College, 525 East 68th Street, New York, NY 10021, USA.