– Georg Behrens, Reinhold E.Schmidt –
The HIV lipodystrophy syndrome including metabolic complications and altered fat distribution is a possible side effect of HIV therapy. Fortunately, modern therapy regimens are much less likely to lead to fat tissue abnormalities. The metabolic abnormalities may harbor a significant risk of developing cardiovascular disease. In addition, several studies report a reduced quality of life in patients with body habitus changes leading to a reduced treatment adherence. Despite the impact of lipodystrophy syndrome on HIV management, little is known about the pathogenesis, its prevention, diagnosis and treatment. Current data indicate a rather multifactorial pathogenesis where HIV infection, ART, and patient-related factors are all major contributors. The lack of a clear and easy definition reflects the clinical heterogeneity, limits a clear diagnosis and impairs the comparison of results among clinical studies. Therapeutic and prevention strategies have so far been of only limited clinical success, where avoiding the use of thymidine analogues appears to be most effective in avoiding peripheral fat loss. General recommendations include dietary changes and lifestyle modifications, altering antiretroviral therapy (replacing protease inhibitors with NNRTIs or replacing d4T and AZT with abacavir or tenofovir), and finally, the use of metabolically active drugs. Here we summarize the pathogenesis, diagnosis and treatment options of the HIV lipodystrophy syndrome.
Lipodystrophy was originally described as a condition characterized by regional or generalized loss of subcutaneous fat. The non-HIV-associated forms, such as congenital or familial partial lipodystrophy, have a very low prevalence. Generally, these forms are associated with complex metabolic abnormalities and are difficult to treat. The term “lipodystrophy syndrome” was introduced to describe a complex medical condition including an apparent abnormal fat redistribution and metabolic disturbances in HIV patients receiving protease inhibitor therapy (Carr 1998). Now, years after its first description, there is still no consensus on a case definition for lipodystrophy syndrome in HIV. Thus, the diagnosis of lipodystrophy in clinical practice often relies on a more individual interpretation than on an evaluated classification. Finally, changes in fat distribution have to be considered as being part of a rather dynamic process. In most cases, peripheral lipoatrophy is clinically diagnosed when significant fat loss of about 30% has already occurred.
HIV-associated lipodystrophy includes both clinical and metabolic alterations. The most prominent clinical sign is a loss of subcutaneous fat (lipoatrophy) in the face (periorbital, temporal), limbs, and buttocks. Prospective studies in patients on thymidine alanogues have demonstrated an initial increase in limb fat during the first months of therapy, followed by a progressive decline over the ensuing years (Mallon 2003), which is mostly persistent (Grunfeld 2010). Peripheral fat loss can be accompanied by an accumulation of visceral fat, which can cause mild gastrointestinal symptoms. Initially truncal fat increases on therapy and then remains stable (Mallon 2003). Visceral obesity, as a singular feature of abnormal fat redistribution, appears to occur in only a minority of patients. Fat accumulation may also be found as dorsocervical fat pads (buffalo hump) or within the muscle and the liver. Female HIV-positive patients sometimes complain about painful breast enlargement, attributed to the lipodystrophy syndrome. Whether gynecomastia in male patients is a component of the syndrome remains unclear. There is now accumulating evidence that the major clinical components – lipoatrophy, central adiposity and the combination of both – result from different pathogenetic developmental processes.
The prevalence of a clinically evident lipodystrophy syndrome was estimated to be between 30 and 50% based on cross-sectional studies before 2005. A prospective study over an 18-month period after initiation of therapy revealed a prevalence of 17% but current HIV therapy combinations can be expected to lead to a lower incidence. More recent studies estimated the annual incidence rates of detectable but clinically inapparent peripheral fat loss (-20%) with about 5-10% in patients receiving a nuke-backbone with tenofovir, abacavir, 3TC or FTC. Lipodystrophy, and in particular lipoatrophy, has been observed most frequently in patients receiving a combination regimen of nucleoside analogues (particularly thymidine analogues) and protease inhibitors, although almost all antiretroviral drug combinations can be associated with fat redistribution. The risk of the syndrome increases with the duration of treatment, the age of the patient and the level of immunodeficiency. Children can be affected, like adults, with clinical fat redistribution shortly after initiation or change of ART. The evolution of the individual clinical components of the lipodystrophy syndrome is variable. The nucleoside analogue linked most strongly to lipoatrophy is d4T, particularly when used in combination with ddI. Tenofovir combined with 3TC and efavirenz is associated with less loss of limb fat than d4T in a similar combination in therapy-naïve HIV patients (Gallant 2004).
Frequently, complex metabolic alterations are associated with these body shape alterations. These include peripheral and hepatic insulin resistance, impaired glucose tolerance, type 2 diabetes, hypertriglyceridemia, hypercholesterolemia, increased free fatty acids (FFA), and decreased high-density lipoprotein (HDL). Often these metabolic abnormalities appear or deteriorate before the manifestation of fat redistribution. The prevalence of insulin resistance and glucose intolerance has been reported in the literature at 20 to 50% depending on the study design and measurement methods and it increases with age (Hasse 2011). Frank diabetes is less frequent with a prevalence of between 1 and 6%. The incidence rate were highest at the time when PIs were introduced around 2000 but remain elevated as compared to seronegative control groups (Capeau 2011). Lipodystrophic patients present with the highest rates of metabolic disturbances.
Hyperlipidemias are a frequently observed side effect of antiretroviral therapy, especially in combinations that include PIs. Newer drugs such as maraviroc or raltegravir and also second-generation NNRTIs such as rilpivirine seem to cause only minor disturbances in lipid metabolism (DeJesus 2010). Given that many HIV-infected patients present with already decreased HDL levels, these are not further reduced by antiretroviral drugs, but usually improve to some degree, particularly when NNRTIs such as nevirapine are used. Hypertriglyceridemia, especially in patients with evidence of body fat abnormalities, is the leading lipid abnormality either alone or in combination with hypercholesterolemia. Several weeks after initiation or change of HIV therapy, lipid levels usually reach a plateau and remain stable. Part of this increase can be considered as reconstitution of health, as some patients return to the lipid levels they had before seroconversion. All protease inhibitors can potentially lead to hyperlipidemia, although to different extents. For example, atazanavir and darunavir appear to be less frequently associated with dyslipidemia and insulin resistance. In contrast, ritonavir often leads to hypertriglyceridemia correlating to the drug levels. Lopinavir leads to an approximate 18% mean increase in total cholesterol and 40% mean increase in triglycerides in patients on first line therapy.
The therapy-induced dyslipidemias are characterized by increased low-density lipoproteins (LDL) and triglyceride-rich very low density lipoproteins (VLDL). Detailed characterization revealed an increase of apoplipoprotein B, CIII and E. Raised levels of lipoprotein(a) have been described in protease inhibitor recipients. Mild hypercholesterolemia can occur during therapy with efavirenz but is not typical with nevirapine. D4T-based ART is associated with early and statistically significant increases in total triglycerides and cholesterol or NRTIs. Several studies suggest that tenofovir exerts a moderate lipid-lowering effect on both total and LDL cholesterol as well as HDL cholesterol (Randell 2010). It is important to note that HIV infection itself is associated with disturbed lipid metabolism. During disease progression, total cholesterol, LDL, and HDL levels decline and the total triglyceride level rises. The latter is presumably caused by increased cytokine concentrations (TNFα, IFNg) and an enhanced lipogenesis in addition to impaired postprandial triglyceride clearance.
ART, lipodystrophy syndrome and cardiovascular risk
The fat redistribution and disturbances in glucose and fat metabolism resemble a clinical situation that is known as the “metabolic syndrome” in HIV-negative patients. This condition includes symptoms such as central adipositas, insulin resistance and hyperinsulinemia, hyperlipidemia (high LDL, Lp(a) hypertriglyceridemia and low HDL) and hypercoagulopathy. Given the well-established cardiovascular risk resulting from this metabolic syndrome, there is growing concern about a potential therapy-related increased risk of myocardial infarction in HIV-positive patients. These fears are further sustained by reports of arterial hypertension on ART, a high rate of smoking among HIV patients and increased levels of tissue plasminogen activator (tPA) and plasminogen activator inhibitor-1 (PAI-1) in patients with lipodystrophy. Although many of the mainly retrospective studies dealing with this issue are inconclusive, data from a large international study (the D:A:D study) provide evidence for an increased relative risk of myocardial infarction during the first 7 years of ART (Friis-Møller 2003, El-Sadr 2006). The incidence of myocardial infarction increased from 1.39/1,000 patient years in those not exposed to ART, to 2.53/1,000 patient years in those exposed for < 1 year, to 6.07/1,000 patient years in those exposed for ≥ 6 years (RR compared to no exposure: 4.38, p = 0.0001). After adjustment for other potential risk factors, there was a 1.17-fold increased risk of myocardial infarction per additional year of combined ART exposure. It is, however, of note that older age, male gender, smoking, diabetes mellitus, and pre-existing coronary artery disease were still associated with a higher risk of sustaining cardiovascular events than ART in this study. Analysis of this cohort later provided evidence that PIs (particularly indinavir and lopinavir), abacavir and triglycerides contribute to this increased risk (Friis-Moller 2007, D:A:D Study Group 2008, Worm 2010, Worm 2011). Several other cohort studies, although not all (Lang 2011), confirmed the association of abacavir use and myocardial infarction (Behrens 2010). According to a meta-analysis (Cruciani 2011) considering prospective, controlled studies about the use of abacavir, the rate of myocardial infarction in these mostly young patients with a rather low cardiovascular risk profil was not significantly different to control regimens. Currently, it seems reasonable to consider alternatives for abacavir only in patients with a high cardiovascular risk (Framingham risk score >20%).
Although the CHD risk profile in D:A:D patients worsened over time, the risk of myocardial infarction decreased over time after controlling for these changes. Several other studies used ultrasonography to measure the thickness of the carotid intima media or endothelial function to predict the cardiovascular risk. Some of these investigations found abnormal test results (e.g., reduced flow-mediated dilation) that correlated either with the use of PIs or the presence of dyslipidemia (Currier 2005). Interestingly, HIV infection may lead to endothelial dysfunction and an unfavorable pro-atherogenic profile (Grunfeld 2009). While there is some indication of an increased rate of coronary artery disease with ART, particularly in combinations containing abacavir (D:A:D 2008), the benefit of suppressed viral replication and improved immune function resulting in reduced morbidity and mortality, clearly argues for the use of antiretroviral drugs according to current international guidelines. It seems obvious however, that pre-existing cardiovascular risk factors in individual patients need to be considered more carefully before starting or switching ART.
Recommendations such as the National Cholesterol Education Program (NECP) have been proposed for non-HIV-infected patients with similar risk profiles. These guidelines are being considered for HIV patients as well (Schambelan 2002, Grinspoon 2005). According to these recommendations, the overall cardiovascular risk in HIV-infected patients can be determined from specific risk factors by using the Framingham equation. Prediction of coronary heart disease using this equation, however, may have some limitations. A 10-year CHD risk estimation at any time point is determined by the individual’s past and expected future lipid levels (best assessed as area under the curve). Hyperlipidemia in many treated HIV-infected patients, however, does not follow the 10-year time course seen in the uninfected population due to therapy changes that may lower total cholesterol, increase HDL, and improve atherogenic risk (Behrens 2005). Thus, the validity of this calculation for long-term cardiovascular risk assessment in young patients with changing lipid levels and medication regimens requires further studies, but it seems helpful to identify patients with increased myocardial risk.
Clearly, more clinical studies are necessary to assess whether these recommendations are also applicable in the presence of HIV and to determine the clinical value of lipid lowering drug therapy in these patients. Most importantly, the information about drug interactions of lipid lowering and antiretroviral drugs is still incomplete. The accumulation of pre-existing and drug-related risk factors will get more clinical attention, because, by improving the HIV-associated morbidity and mortality, ART consequently increases an additional relevant cardiovascular risk factor: the age of patients who are effectively treated with antiretroviral drugs.
For a better understanding of the pathogenesis of complex metabolic abnormalities, it is useful to separate individual aspects of the lipodystrophy syndrome: adipocytes/fat redistribution, lipid metabolism, and carbohydrate metabolism. This is because it is very likely that the lipodystrophy syndrome is not a stereotypic syndrome but rather an amalgam of miscellaneous clinical features, with perhaps multifactorial causes. Studies published in recent years provide evidence for two fundamental assumptions: firstly, lipoatrophy and lipoaccumulation result from divergent or only partially overlapping pathogenetic reasons. Secondly, NRTIs, NNRTIs, PIs, and even drugs within each class contribute to the lipodystrophy syndrome and its individual features by different possibly overlapping mechanisms.
NRTIs and lipodystrophy
The patterns of fat redistribution in patients who are only taking NRTIs are unlike those observed in patients on PI therapy. Peripheral fat loss is the major symptom observed in NRTI therapy (particularly d4T and AZT), although a few clinical studies have described a minimal intra-abdominal fat increase in these patients, which is clearly less than on PIs. In ACTG 5142, in which patients receiving efavirenz (+2 NRTIs) or lopinavir (+2 NRTIs) where compared, a 20% percent reduction in peripheral fat tissue was significantly more frequently found in patients receiving d4T or AZT (Haubrich 2009). Given that, commonly, only a mild increase in triglycerides has been observed, exclusive NRTI therapy seems to be of minor impact on lipid metabolism, but is also not a preferred choice of therapy. Postprandially elevated FFA in patients with lipodystrophy, together with in vitro experiments, have led to the hypothesis that NRTIs could impair fatty acid binding proteins (FABP) that are responsible for cellular fat uptake and intracellular fat transport.
It is well established that long-term NRTI therapy can cause mitochondrial toxicity. The clinical manifestation of this presents in symptoms such as hepatic steatosis, severe hyperlactatemia, and polyneuropathy. As an explanation for these symptoms, the “pol-γ hypothesis” has been proposed, which was later extended to reveal the lipoatrophy observed under NRTIs (Brinkman 1999). To maintain an adequate bioenergetic level for accurate cell function, all metabolically active cells depend on a persistent polymerase γ-mediated mitochondrial (mt) DNA synthesis. Mitochondria require a constant supply of nucleosides for this process. The mitochondrial DNA polymerase γ retains both DNA- as well as RNA-dependent DNA polymerase activity. The latter is perhaps responsible for the HIV reverse transcriptase activity and therefore its susceptibility for interactions with NRTIs. Experimental data reveals that, for NRTI uptake into mitochondria, the subsequent phosphorylation and then incorporation into the DNA, certain pharmacodynamic requirements need to be fulfilled. These requirements, including thymidine kinase activity and deoxynucleotide transport specificity of the mitochondrial membrane, are apparently different for AZT and d4T, which partially explains the prevailing association between lipoatrophy and d4T therapy. The postulated mechanisms of NRTI-induced mitochondrial dysfunction consist of competitive inhibition, incorporation into the mtDNA resulting in mtDNA depletion, impairment of mitochondrial enzymes, uncoupling of oxidative phosphorylation and induction of apoptosis. Depletion of mtDNA and structural changes in the mitochondria, resulting in increased rates of apoptosis in subcutaneous adipocytes, have been confirmed in other studies. Despite the experimental link between mitochondrial toxicity and fat tissue as one potential target organ, the degree to which mitochondrial damage contributes to fat distribution abnormalities and its specificity remains unknown. Most likely, additional factor may be relevant given that mtDNA loss and mitochondrial dysfunction has been found in fat tissue of therapy-naïve patients (Garrabou 2011). Also, mt DNA measurement in PBMCs from the blood may be of little relevance for toxicity in fat tissue. In contrast, mitochondrial damage is widely believed to be responsible for other NRTI-related side effects, such as myopathy, hyperlactatemia, microvesicular steatosis, and steatohepatitis with lactic acidosis.
Protease inhibitors and lipodystrophy
PIs account for the majority of metabolic abnormalities associated with the lipodystrophy syndrome. Numerous studies report increases in the levels of total triglycerides and triglyceride-rich lipoproteins (VLDL) accompanied by raised LDL levels after initiation of PI therapy (Walli 1998, Behrens 1999). Conversely, these parameters improve substantially in most studies after discontinuation of the PI or on switching to abacavir or nevirapine. The hyperlipidemic changes are frequently associated with hyperinsulinemia and/or insulin resistance.
It has been proposed, based on in vitro experiments, that PIs such as saquinavir, indinavir, and ritonavir are able to inhibit proteasomal degradation of apolipoprotein B leading to intracellular stockpiling of this lipoprotein and excessive release in response to FFA (Liang 2001). Using stable isotopes in vivo, other authors demonstrate a dramatic increase in FFA turnover together with increased lipolysis and decreased clearance of triglyceride-rich VLDL and chylomicrons (Shekar 2002). These conditions point towards an impaired postprandial insulin-mediated lipid metabolism, since insulin, on the one hand, normally inhibits lipolysis and, on the other hand, increases uptake of FFA, triglyceride synthesis, and fat oxidation in favor of glucose oxidation.
It remains unclear whether impaired insulin action eventually leads to dyslipidemia, or whether hyperlipidemia is responsible for reduced insulin function and insulin resistance in the periphery. Presumably, both mechanisms are important given that some PIs (e.g., indinavir) have been shown to induce insulin resistance without changes occurring in lipid metabolism after short-term administration (Noor 2001, Noor 2002), whereas other PIs (e.g., ritonavir) have been demonstrated to cause mainly hypertriglyceridemia due to increased hepatic synthesis without major changes occurring in glucose metabolism (Purnell 2000).
It is reasonable to speculate that lipid abnormalities and, in particular increased FFA levels, contribute substantially to the peripheral and central insulin resistance of skeletal muscles and the liver, presumably due to the increased storage of lipids in these organs (Gan 2002). Given this hypothesis, the visceral adiposity could reflect the adaptation of the body in response to raised FFA concentrations and an attempt to minimize the lipotoxic damage to other organs.
Several in vitro experiments have indicated that almost all PIs can potentially lead to insulin resistance in adipocytes. Short-term administration of indinavir caused an acute and reversible state of peripheral insulin resistance in healthy volunteers, which was determined in an euglycemic-hyperinsulinemic clamp. These effects are most likely caused by the inhibition of glucose transport mediated by GLUT-4, the predominant transporter involved in insulin-stimulated cellular glucose uptake in humans (Murata 2002). A common structural component found in most PIs has been proposed to cause GLUT-4 inhibition. In some patients with lipodystrophy, additional impairment of glucose phosphorylation may contribute to insulin resistance (Behrens 2002). This is presumably due to an impaired insulin-mediated suppression of lipolysis and subsequently increased FFA levels (Behrens 2002, van der Valk 2001) and accumulation of intramyocellular lipids. Peripheral insulin resistance may also account for an increase in the resting energy expenditure in HIV lipodystrophy and a blunted insulin-mediated thermogenesis.
Indinavir may also induce insulin resistance by inhibiting the translocation, processing or phosphorylation of the sterol regulatory element-binding protein 1c (SREBP-1c). Either directly or via the peroxisome proliferator activated receptor g (PPARg), SREBP-1 regulates FFA uptake and synthesis, adipocyte differentiation and maturation, and glucose uptake by adipocytes. Similarly, the function of these factors has been proposed to be disturbed in inherited forms of lipodystrophy. Finally, hypoadiponectinemia, as found in patients with abnormal fat distribution, may contribute to insulin resistance (Addy 2003). Genetically, host factors interfering with drug metabolism (Domingo 2011) and additional predisposing factors in mechanistically plausible and other genes (Montes 2010, Pinti 2010 Wangsomboonsiri 2010) appear to contribute.
Both the lack of a formal definition and uncertainty about the pathogenesis and possible long-term consequences leads to a continuing discussion about appropriate guidelines for the assessment and management of HIV lipodystrophy syndrome and its metabolic abnormalities. Outside clinical studies, the diagnosis relies principally on the occurrence of apparent clinical signs and the patient reporting them. A standardized data collection form may assist in diagnosis (Grinspoon 2005). This appears sufficient for the routine clinical assessment, especially when the body habitus changes develop rather rapidly and severely. For clinical investigations however, especially in epidemiological and interventional studies, more reliable measurements are required. A recent multicenter study to develop an objective and broadly applicable case definition proposes a model including age, sex, duration of HIV infection, HIV disease stage, waist-to-hip ratio, anion gap, serum HDL cholesterol, trunk-to-peripheral-fat ratio, percentage leg fat, and intra-abdominal to extra-abdominal fat ratio. Using these parameters, the diagnosis of lipodystrophy had a 79% sensitivity and 80% specificity (Carr 2003). Although this model is largely for research and contains detailed body composition data, alternative models and scoring systems, incorporating only clinical and metabolic data, also gave reasonable results (for more information, see http://www.med.unsw.edu.au/nchecr).
Despite individual limitations, several techniques are suitable for measuring regional fat distribution. These include dual energy x-ray absorptiometry (DEXA), computer tomography (CT), magnetic resonance imaging (MRI) and sonography. Anthropometric measurements are safe, portable, cheap and much easier to perform than imaging techniques. Waist circumference or sagittal diameter are more sensitive and specific measures than waist-to-hip ratio. Repeated measurements of skin fold thickness can be useful for individual long-term monitoring but need to be performed by an experienced person.
The main imaging techniques (MRI, CT, DEXA) differentiate tissues on the basis of density. Single-slice measurements of the abdomen and extremities (subcutaneous adipose tissue = SAT, visceral adipose tissue = VAT) and more complex three-dimensional reconstructions have been used to calculate regional or total body fat. Limitations of these methods include most notably their expense, availability and radiation exposure (CT). Consequently, CT and MRI should only be considered in routine clinical practice for selected patients (e.g., extended dorso-cervical fat pads, differential diagnosis of non-benign processes and infections).
DEXA is appropriate for examining appendicular fat, comprised almost entirely of SAT, and has been successfully employed in epidemiological studies. However, SAT and VAT cannot be distinguished by DEXA, which therefore limits the evaluation of changes in truncal fat. Application of sonography to measure specific adipose compartments, including those in the face, requires experienced investigators and has been minimally applied in HIV infection so far. Bioelectrical impedance analysis estimates the whole body composition and cannot be recommended for measurement of abnormal fat distribution.
Patients should routinely be questioned and examined for cardiovascular risk factors, such as smoking, hypertension, adiposity, type 2 diabetes, and family history. For an accurate assessment of blood lipid levels, it is recommended to obtain blood after a fasting of at least 8 hours. Total cholesterol and triglycerides together with LDL and HDL cholesterol should be obtained prior to the initiation of, or switch to, any new antiretroviral therapy and repeated 3 to 6 months later. Fasting glucose should be assessed with at least a similar frequency. The oral glucose tolerance test (OGTT) is a reliable and accurate instrument for evaluating insulin resistance and glucose intolerance. An OGTT may be indicated in patients with suspected insulin resistance such as those with adipositas (BMI > 27 kg/m2), a history of gestational diabetes and a fasting glucose level of 110 to 126 mg/dl (impaired fasting glucose). The diagnosis of diabetes is based on fasting glucose levels > 126 mg/dl, glucose levels of > 200 mg/dl independent of fasting status, or a 2-hour OGTT glucose level above 200 mg/dl. Screening of HbA1c appears to be less reliable as in sero-negative patients (Kim 2009, Eckhardt 2011). Additional factors that could lead to or assist in the development of hyperlipidemia and/or insulin resistance always need to be considered (e.g., alcohol consumption, thyroid dysfunction, liver and kidney disease, hypogonadism, concurrent medication such as steroids, b-receptor blockers, thiazides, etc.).
So far, most attempts to improve or even reverse the abnormal fat distribution by modification of the antiretroviral treatment have shown only modest clinical success. In particular, peripheral fat loss appears to be resistant to most therapeutic interventions. The metabolic components of the syndrome may be easier to improve (Table 1). Thus, preventing lipoatrophy by avoiding thymidine analogues (AZT, d4T) is the main goal (Behrens 2008). For more detailed recommendations for improving fat redistribution and treating dyslipidemia, please see the guidelines of the European AIDS Clinical Society (www.eacs.eu). These guidelines emphasize that all traditional cardiovascular risk factors, such as arterial hypertension, hyperlipidemia and type 2 diabetes should be assessed and considered for intervention.
Dietary interventions are commonly accepted as the first therapeutic option for hyperlipidemia, especially hypertriglyceridemia. Use of NCEP guidelines may reduce total cholesterol and triglycerides by 11 and 21%, respectively. Whenever possible, dietary restriction of total fat to 25-35% of the total caloric intake should be a part of any treatment in conjunction with lipid-lowering drugs. Consultation with professional and experienced dieticians should be considered for HIV-infected patients and their partners. Patients with excessive hypertriglyceridemia (>1,000 mg/dl) may benefit from a very low fat diet and alcohol abstinence to reduce the risk of pancreatitis, especially if there is a positive family history or concurrent medications that may harbor a risk of developing pancreatitis. Regular exercise may have beneficial effects, not only on triglycerides and insulin resistance, but probably also on fat redistribution (reduction in truncal fat and intramyocellular fat) and should be considered in all HIV-infected patients (Driscoll 2004a). All patients should be advised and supported to give up smoking in order to reduce cardiovascular risk. Cessation of smoking is more likely to reduce cardiovascular risk than any choice or change of ART or use of any lipid-lowering drug (Petoumenos 2010).
Table 1: Therapeutic options for HIV-associated lipodystrophy and related metabolic complications
|Lifestyle changes (reduce saturated fat and cholesterol intake, increase physical activity, stop smoking)|
Change antiretroviral therapy: replacement of PI, d4T (Zerit®) or AZT (Retrovir™)
Statins: Atorvastatin (Sortis®), Pravastatin (Pravasin®), Fluvastatin (Lescol®)
Fibrates: Gemfibrozil (Gevilon®) or Bezafibrat (Cedur®)
Thiazolidinediones: pioglitazone (Actos®)
Recombinant human growth hormones (e.g., Serostim®) or analogues (e.g. Tesamorelin®)
Given the extensive indications that PIs are the culprits that substantially contribute to metabolic side effects, numerous attempts have been made to substitute the PI component of a regimen with nevirapine, efavirenz, or abacavir. Similarly, given the close association of d4T-based therapy with lipoatrophy, replacement of this thymidine nucleoside analogue by, for example, abacavir or tenofovir has been evaluated in several studies. Indeed, these “switch studies” have demonstrated substantial improvement, although not normalization, of serum lipids (total and LDL cholesterol, triglycerides) and/or insulin resistance in many patients. In patients with hyperlipidemia, substitution of PIs with alternative PIs that have less metabolic side effects (e.g., atazanavir) has also been proven to be a successful strategy (Mallolas 2009, Moebius 2005). Protease inhibitor cessation has not been shown to improve lipoatrophy. However, stopping administration of the thymidine nucleoside analogue d4T or AZT usually leads to a slow recovery (over months and years) measured by DEXA and moderate clinical increase in limb fat (Moyle 2006, Tebas 2009). Under restricted inclusion criteria and study conditions, most patients maintain complete viral suppression after changes to the ART regimen, but not all of these studies included control groups with unchanged antiretroviral therapy. Recently, a pilot study evaluating the effect of uridine (NucleomaxX®) on lipoatrophy in HIV-infected patients continuing their ART regimen described a significant increase in subcutaneous fat after only three months, but this effect was not confirmed in a larger randomized trial (MyComsey 2010). .
The most advantageous changes of metabolic parameters have been observed by replacement of the PI with nevirapine or abacavir. This option is, however, not always suitable, and the clinical benefit of effective viral suppression and improved immune function needs to be considered in view of drug history, current viral load, and resistance mutations. When options are limited, antiretroviral drugs that may lead to elevation of lipid levels should not be withheld for fear of further exacerbating lipid disorders.
Lipid-lowering agents should be considered for the treatment of severe hypertriglyceridemia, elevated LDL or a combination of both. The clinical benefit, however, of lipid lowering or insulin-sensitizing therapy in HIV patients with lipodystrophy remains to be demonstrated. In light of the potentially increased cardiovascular risk to recipients of antiretroviral therapy, the AIDS Clinical Trials Group (ACTG) published recommendations based on the National Cholesterol Education Program (NCEP) for primary and secondary prevention of coronary artery disease in seronegative patients. In addition, more detailed recommendations by the European AIDS Clinical Society have been published to provide guidelines for physicians actively involved in HIV care that will be regularly updated (www.eacs.eu). However, these recommendations should be considered as being rather preliminary, given the so far limited numbers, size and duration of the clinical studies they are based on. It appears reasonable to measure fasting lipid levels annually before and 3-6 months after ART is initiated or changed. Whenever possible, the ART least likely to worsen lipid levels should be selected for patients with dyslipidemia.
The decision on lipid-lowering therapy can be based on estimating the 10-year risk for myocardial infarction according to the Framingham equation (http://hin.nhlbi.nih.gov/atpiii/calculator.asp). In case of a more than 20% risk, dietary interventions and change of antiviral therapy should be considered. In patients with frank diabetes or coronary heart disease (CHD), lipid lowering drugs are recommended. The EACS guidelines (2011) recommend to target total cholesterol levels of < 190 mg/dl (<5 mmol/l) and for patients with type 2 diabetes and CHD levels < 155 mg/dl (<4 mmol/l). For LDL-cholesterol levels one should aim for levels < 115 mg/dl (<3 mmol/l), and if possible, in patients with type 2 diabetes or CHD, levels < 80 mg/dl (< 2mmol/l).
HMG-CoA reductase inhibitors have been successfully used in combination with dietary changes in HIV-positive patients with increased total and LDL cholesterol. These drugs may decrease total and LDL cholesterol by about 25% (Grinspoon 2005). Many of the statins (as well as itraconazole, erythromycin, diltiazem, etc.) share common metabolization pathways with PIs via the cytochrome P450 3A4 system, thereby potentially leading to additional side effects due to increased plasma levels of statins which can then cause liver and muscle toxicity. They can be combined with ezetimibe in order to improve their lipid lowering effect (Negredo 2006). Based on limited pharmacokinetic and clinical studies, atorvastatin (Sortis®), fluvastatin (Lescol®), and pravastatin (Pravasin®), carefully administered at increasing doses, are the preferred agents for a carefully monitored therapy in HIV-infected patients on ART. Lovastatin (Mevinacor®) and simvastatin (Zocor®) should be avoided due to their potential interaction with PIs.
Fibric acid analogues such as gemfibrozil or fenofibrate are particularly effective in reducing the triglyceride levels by up to 50% (Rao 2004, Miller 2002) and should be considered in patients with severe hypertriglyceridemia (>1000 mg/dl). Fibric acid analogues retain a supportive effect on lipoprotein lipase activity and can thereby lower LDL levels. Despite their potentially synergistic effect, co-administration of fibric acid analogues and statins in patients on ART should only be used carefully in selected individuals, since both can cause rhabdomyolysis. Niacinic acid has been shown to only minimally improve the hyperlipidemia induced by ART. It does, however, increase peripheral insulin resistance (Gerber 2004). Extended-release niacin (Niaspan®) has been shown to have beneficial effects mainly on triglycerides and was well tolerated at a dose of 2,000 mg daily in a study with 33 individuals (Dube 2005). Similarly, polyunsaturated fatty acids could be beneficial in patients with hypertriglyceridemia (De Truchis 2007). Finally, it should be stressed that the long-term effects of lipid-lowering agents and their impact on cardiovascular outcomes, especially in HIV-positive patients with moderate or severe hypertriglyceridemia, are unknown.
Metformin has been evaluated for the treatment of lipodystrophy syndrome. Some studies have revealed a positive effect on the parameters of insulin resistance and the potential reduction of intra-abdominal (and subcutaneous) fat, although not clinically obvious. Together with exercise training, metformin has been described to reverse the muscular adiposity in HIV-infected patients (Driscoll 2004b). Metformin, like all biguanides, can theoretically precipitate lactic acidosis and should thus be used with caution. Use of metformin should be avoided in patients with creatinine levels above 1.5 mg/dl, increased aminotransferase levels, or hyperlactatemia. Thiazolidinediones, such as rosiglitazone (Avandia®) or pioglitazone (Actos®), exhibit the potency to improve insulin sensitivity via stimulation of the PPARγ and other mechanisms. However, both drugs have been associated with significant side effects and can currently not be recommended for HIV-infected patients. Rosiglitazone has been successfully used to treat abnormal fat distribution in genetic lipodystrophies. Three published studies on HIV patients, however, revealed no or only minimal improvement in abnormal fat distribution. But, insulin sensitivity was increased at the expense of increased total cholesterol and triglycerides (Carr 2004, Hadigan 2004, Sutinen 2003, Cavalcanti 2005, Sheth 2010). Thus, rosiglitazone cannot be recommended for general treatment of lipoatrophy in HIV (Grinspoon 2005). It also reduces the bioavailability of nevirapine, although not of efavirenz and lopinavir (Oette 2005). A randomized double-blind placebo-controlled trial (ANRS 113) revealed a significant increase in subcutaneous fat 48 weeks after treatment with pioglitazone 30 mg once daily without demonstrating negative effects on lipid parameters (Slama 2008). The peripheral fat increase was most pronounced in patients, which stopped thymidine analogue therapy (Tungsiripat 2010), but because of side effects pioglitazone cannot be recommended to treat HIV-associated lipoatrophy.
Recombinant growth hormone (Serostim®) at doses of 4-6 mg/d sc over 8-12 weeks has been demonstrated in small studies to be a successful intervention for reducing visceral fat accumulation, but it also reduces subcutaneous fat (Kotler 2004). Unfortunately, these improvements have been shown to consistently reverse after the discontinuation of growth hormone therapy. Studies with lower maintenance doses have not been performed yet. The possible side effects associated with growth hormone therapy include arthralgia, peripheral edema, insulin resistance and hyperglycemia. Alternatively, a stabilized analogue of the growth hormone-releasing factor (Tesamorelin®), administered subcutaneously, can lead to reduction in visceral fat accumulation with less side effects (Falutz 2007) and was recently approved by the FDA.
Surgical intervention (liposuction) for the treatment of local fat hypertrophy has been successfully performed, but appears to be associated with an increased risk of secondary infection (Guaraldi 2011), and recurrence of fat accumulation is possible. For the treatment of facial lipoatrophy, repeated subcutaneous injection of agents such as poly-L-lactic acid (Sculptra®, New-Fill®), a resorbable molecule that promotes collagen formation, has been effectively used in HIV patients (Valantin 2003, Lafaurie 2003¸ Guaraldi 2005, Mest 2004, Casavantes 2004, Behrens 2008). In 2004, Sculptra®was approved by the FDA as an injectable filler to correct facial fat loss in people with HIV. We recommend consultation with experienced specialists for surgical treatments and injection therapy. Further evaluation in long-term follow-up studies is necessary to fully assess the value of these methods.
Addy CL, Gavrila A, Tsiodras S, et al. Hypoadiponectinemia is associated with insulin resistance, hypertriglyceridemia, and fat redistribution in HIV–infected patients treated with HAART. J Clin Endocrinol Metab 2003; 88:627-36.
Behrens GMN. Treatment option for lipodystrophy in HIV-positive patients. Expert Opin Pharmacother 2008; 9:39-52.
Behrens G, Dejam A, Schmidt H, et al. Impaired glucose tolerance, beta cell function and lipid metabolism in HIV patients under treatment with protease inhibitors. AIDS 1999; 13:F63-70.
Behrens GMN. Cardiovascular risk and body-fat abnormalities in HIV-infected adults. N Engl J Med 2005; 352:1721-1722.
Behrens GMN, Boerner AR, Weber K, et al. Impaired glucose phosphorylation and transport in skeletal muscle causes insulin resistance in HIV-1 infected patients with lipodystrophy. J Clin Invest 2002, 110:1319-1327.
Behrens GM, Reiss P. Abacavir and cardiovascular risk. Curr Opin Infect Dis 2010;23:9-14.
Brinkman K, Smeitink JA, Romijn JA, Reiss, P. Mitochondrial toxicity induced by nucleoside-analogue reverse-transcriptase inhibitors is a key factor in the pathogenesis of antiretroviral-therapy-related lipodystrophy. Lancet 1999; 354:1112-1115.
Calza L, Manfredi R, Chiodo F. Statins and fibrates for the treatment of hyperlipidaemia in HIV-infected patients receiving HAART. AIDS 2003; 17: 851-859.
Capeau J, Bouteloup V, Raffi F et al. Diabetes Mellitus in Treated HIV-infected Patients: Incidence over 10 Years in 1046 patients from the ANRS CO8 APROCO-COPILOTE Cohort. Abstract 850, 18th CROI 2011, Boston.
Carr A, Emery S, Law M, Puls R, Lundgren JD, Powderly WG. An objective case definition of lipodystrophy in HIV-infected adults: a case-control study. Lancet 2003; 361:726-35.
Carr A, Samaras K, Burton S, et al. A syndrome of peripheral lipodystrophy, hyperlipidaemia and insulin resistance in patients receiving HIV protease inhibitors. AIDS 1998; 12:F51-8.
Carr A, Workman C, Carey D, et al. No effect of rosiglitazone for treatment of HIV-1 lipoatrophy: randomised, double-blind, placebo-controlled trial. Lancet 2004; 363:429-38.
Casavantes LC, Gottlieb M. Bio-Alcamid, a high-volume injectable posthesis for facial reconstruction in HIV-related lipoatrophy: a report on 100 patients. Antivir Ther 2004; 9:L37.
Cavalcanti R, Kain K, Shen S, Raboud J, Walmsley S. A randomized placebo controlled trial of rosoglitazone for the treatment of HIV lipodystrophy. Abstract 854, 12th CROI 2005, Boston.
Cruciani M, Zanichelli V, Serpelloni G, et al. ABACAVIR use and cardiovascular disease events: a meta-analysis of published and unpublished data. AIDS. 2011 Jun 29. PMID:21716077
Currier JS, Kendall MA, Zackin R, et al. Carotid artery intima-media thickness and HIV infection: traditional risk factors overshadow impact of protease inhibitor exposure. AIDS 2005; 19: 927-33.
D:A:D Study Group. Use of nucleoside reverse transcriptase inhibitor and risk of myocardial infarction in HIV-infected patients enrolled in the D:A:D study: a mulit-cohort collaboration. Lancet 2008; 371: 1417-26.
De Truchis P, Kirstetter M, Perier A, et al. Reduction in triglyceride level with N-3 polyunsaturated fatty acids in HIV-infected patients taking potent antiretroviral therapy: a randomized prospective study. J Acquir Immune Defic Syndr 2007, 44: 278-85.
DeJesus E, Cohen C, Lennox J, et al. Metabolic profiles and body composition changes in treatment-naive HIV-infected patients treated with raltegravir 400 mg twice-daily vs efavirenz 600 mg each bedtime combination therapy: 96-week follow-up. Abstract 720, 17th CROI 2010b, San Francisco
Domingo P, Cabeza MC, Pruvost A et al. Association of thymidylate synthase genep olymorphisms with stavudine triphosphate intracellular levels and lipodystrophy. Antimicrob Agents Chemother 2011; 55:1428-35.
Driscoll SD, Meininger GE, Lareau MT, et al. Effects of exercise training and metformin on body composition and cardiovascular indices in HIV-infected patients. AIDS 2004, 18: 465-473.
Eckhardt B, Holzman R, Kwan C, Baghdadi J, Aberg J. Glycated Hemoglobin A1C as Screening for Diabetes Mellitus in HIV-infected Individuals. Abstract 849, 18th CROI 2011, Boston.
El Sadr WM, Lundgren JD, Neaton JD, et al. CD4+ count-guided interruption of antiretroviral treatment. N Engl J Med; 2006; 355:2283-2296.
Falutz J, Allas S, Blot K, et al. Metabolic effects of a growth hormone-releasing factor in patients with HIV. N Engl J Med 2007; 357:2359-70.
Friis-Moller N, Sabin CA et al. for the Data Collection on Adverse Events of Anti-HIV Drugs (DAD) Study Group.Combination antiretroviral therapy and the risk of myocardial infarction. N Engl J Med 2003; 349: 1993-2003
Friis-Moller N, Reiss P, El-Sadr W et al. Class of antiretroviral drugs and the risk of myocardial infarction. N Engl J Med. 2007; 356:1723-35.
Gan SK, Samaras K, Thompson CH, et al. Altered myocellular and abdominal fat partitioning predict disturbance in insulin action in HIV protease inhibitor-related lipodystrophy. Diabetes 2002; 51:3163-9.
Garrabou G, López S, Morén C et al. Mitochondrial damage in adipose tissue of untreated HIV-infected patients. AIDS 2011, 25:165-70.
Gerber MT, Mondy KE, Yarasheski KE, et al. Niacin in HIV-infected individuals with hyperlipidemia receiving potent antiretroviral therapy. Clin Infect Dis 2004; 39: 419-25.
Grinspoon S, Carr A. Cardiovascular risk and body-fat abnormalities in HIV-infected adults. N Engl J Med 2005; 352:48-62.
Grunfeld C et al. HIV infection is an independent risk factor for atherosclerosis similar in magnitude to traditional cardiovascular disease risk factors. Abstract 146, 16th CROI 2009, Montreal.
Grunfeld C, Saag M, Cofrancesco J Jr et al. Study of Fat Redistribution and Metabolic Change in HIV Infection (FRAM). Regional adipose tissue measured by MRI over 5 years in HIV-infected and control participants indicates persistence of HIV-associated lipoatrophy. AIDS 2010; 24:1717-26.
Guaraldi G, Orlando G, De Fazio D, et al. Comparison of three different interventions for the correction of HIV-associated facial lipoatrophy: a prospective study. Antivir Ther 2005; 10: 753-9.
Guaraldi G, Fontdevila J, Christensen LH et al. Surgical correction of HIV-associated facial lipoatrophy. AIDS 2011; 25:1-12.
Hadigan C, Yawetz S, Thomas A, et al. Metabolic effects of rosiglitazone in HIV lipodystrophy: a randomized, controlled trial. Ann Intern Med 2004; 140:786-94.
Hasse B, Ledergerber B, Egger M et al. Aging and Non-HIV-associated Co-morbidity in HIV+ Persons: The SHCS. Abstract 792, 18th CROI 2011, Boston.
Haubrich RH, Riddler S, Dirienzo G, et al. Metabolic outcomes in a randomized trial of nucleoside, nonnucleoside and protease inhibitor-sparing regimens for initial HIV treatment. AIDS 2009, May 4. [Epub ahead of print]
Kim PS, Woods C, Georgoff P, Crum D, Rosenberg A, Smith M, Hadigan C. A1C underestimates glycemia in HIV infection. Diabetes Care. 2009; 32:1591-3.
Kotler DP, Muurahainen N, Grunfeld C, et al. Effects of growth hormone on abnormal visceral adipose tissue accumulation and dyslipidemia in HIV-infected patients. J Acquir Immune Defic Syndr 2004; 35: 239-52.
Lafaurie M, Dolivo M, Boulu D, et al. Treatment of facial lipoatrophy with injections of polylactic acid in HIV-infected patients. Abstract 720, 10th CROI 2003; Boston.
Liang JS, Distler O, Cooper DA, et al. HIV protease inhibitors protect apolipoprotein B from degradation by the proteasome: A potential mechanism for protease inhibitor-induced hyperlpidemia. Nat Med 2001; 7:1327-1331.
Mallolas J, Podzamczer D, Milinkovic A, et al. Efficacy and safety of switching from boosted lopinavir to boosted atazanavir in patients with virological suppression receiving a LPV/r-containing HAART: the ATAZIP study. J Acquir Immune Defic Syndr 2009; 51: 29-36.
Mallon PWG, Miller J, Cooper DA, Carr A. Prospective evaluation of the effects of antiretroviral therapy on body composition in HIV-1-infected men starting therapy. AIDS 2003; 17:971-79.
McComsey G, Walker U, Budhathoki C, et al. Uridine supplementation in the management of HIV lipoatrophy: Results of ACTG 5229. Abstract 131, 17th CROI 2010, San Francisco
Mest DR, Humble G. Safety and efficacy of intradermal poly-L-lactic acid (SculptraTM) injections in patients with HIV-associated facial lipoatrophy. Antivir 2004; 9:L36.
Miller J, Brown D, Amin J, et al. A randomized, double-blind study of gemfibrozil for the treatment of protease inhibitor-associated hypertriglyceridaemia. AIDS 2002; 16: 2195-2200.
Möbius U, Lubach-Ruitman M, Castro-Frenzel B, et al. Switching to atazanavir improves metabolic disorders in patients with severe hyperlipidemia. J Acquir Immune Defic Syndr 2005; 39:174-180.
Montes AH, Valle-Garay E, Suarez-Zarracina T et al. The MMP1 (-16071G/2G) single nucleotide polymorphism associates with the HAART-related lipodystrophic syndrome. AIDS 2010; 24:2499-506.
Moyle GJ, Sabin CA, Cartledge J, et al. A randomized comparative trial of tenofovir DF or abacavir as replacement for a thymidine analogue in persons with lipoatrophy. AIDS 2006; 20: 2043-50.
Murata H, Hruz PW, Mueckler M. Indinavir inhibits the glucose transporter isoform Glut4 at physiologic concentrations. AIDS 2002; 16:859-863.
Negredo E, Molto J, Puig J, et al. Ezetimibe, a promising lipid-lowering agent for the treatment of dyslipidaemia in HIV-infected patients with poor response to statins. AIDS 2006; 20: 2159-64.
Noor MA, Lo JC, Mulligan K, et al. Metabolic effects of indinavir in healthy HIV-seronegative men. AIDS 2001; 15:F11-F18.
Noor MA, Seneviratne T, Aweeka FT, et al. Indinavir acutely inhibits insulin-stimulated glucose disposal in humans: a randomized, placebo-controlled study. AIDS 2002; 16:F1-F8.
Oette M, Kurowski M, Feldt T, et al. Impact of rosiglitazone treatment on the bioavailability of antiretroviral compounds in HIV-positive patients. J Antimicrob Chemother 2005; 56:416-419.
Petoumenos K, Worm S, Reiss P, et al. Rates of cardiovascular disease following smoking cessation in patients with HIV infection: results from the D:A:D study. Abstract 124, 17th CROI 2010, San Francisco
Pinti M, Gibellini L, Guaraldi G et al. Upregulation of nuclear-encoded mitochondrial LON protease in HAART-treated HIV-positive patients with lipodystrophy: implications for the pathogenesis of the disease. AIDS 2010, 24:841-50.
Rao A, D’Amico S, Balasubramanyam A, Maldonado M. Fenofibrate is effective in treating hypertriglyceridemia associated with HIV lipodystrophy. Am J Med Sci 2004; 327: 315-318.
Randell PA, Jackson AG, Zhong L, Yale K, Moyle GJ. The effect of tenofovir disoproxil fumarate on whole-body insulin sensitivity, lipids and adipokines in healthy volunteers. Antiviral Therapy 2010; 15:227-233
Schambelan M, Benson CA, Carr A, et al. Management of metabolic complications associated with antiretroviral therapy for HIV-1 infection: recommendations of an International AIDS-Society-USA panel. J AIDS 2002; 31:257-275.
Sekhar RV, Jahoor F, White AC, et al. Metabolic basis of HIV-lipodystrophy syndrome. Am J Physiol Endocrinol Metab. 2002;283:E332-7.
Sheth SH, Larson RJ. The efficacy and safety of insulin-sensitizing drugs in HIV-associated lipodystrophy syndrome: a meta-analysis of randomized trials. BMC Infect Dis. 2010;10:183.
Slama L, Lanoy E, Valentin MA et al. Slama L, Lanoy E, Valentin MA, et al. Effect of pioglitazone on HIV-1-related lipodystrophy: a randomized double-blind placebo-controlled trial (ANRS 113). Antivir Ther 2008, 13:67-76.
Sutinen J, Hakkinen AM, Westerbacka J, et al. Rosiglitazone in the treatment of HAART-associated lipodystrophy – a randomized double-blinded placebo controlled study. Antivir Ther 2003; 8:199-207.
Tebas P, Zhang J, Hafner R, et al. Peripheral and visceral fat changes following a treatment switch to a non-thymidine analogue or a nucleoside-sparing regimen in HIV-infected subjects with peripheral lipoatrophy: results of ACTG A5110. J Antimicrob Chemother 2009; 63: 998-1005.
Tungsiripat M, Bejjani DE, Rizk N et al. Rosiglitazone improves lipoatrophy in patients receiving thymidine-sparing regimens. AIDS 2010;24:1291-8.
Valantin MA, Aubron-Olivier C, Ghosn J, et al. Polylactic acid implants (New-Fill) to correct facial lipoatrophy in HIV-infected patients: results of the open-label study VEGA. AIDS 2003; 17: 2471-7.
van der Valk M, Bisschop PH, Romijn JA, et al. Lipodystrophy in HIV-1-positive patients is associated with insulin resistance in multiple metabolic pathways. AIDS 2001; 15:2093-2100.
Walli R, Herfort O, Michl GM, et al. Treatment with protease inhibitors associated with peripheral insulin resistance and impaired oral glucose tolerance in HIV-1-infected patients. AIDS 1998; 12:F167-73.
Wangsomboonsiri W, Mahasirimongkol S, Chantarangsu S et al. Association between HLA-B*4001 and lipodystrophy among HIV-infected patients from Thailand who received a stavudine-containing antiretroviral regimen. Clin Infect Dis 2010; 50:597-604.
Worm SW, Kamara DA, Reiss P et al. Elevated triglycerides and risk of myocardial infarction in HIV-positive persons. AIDS. 2011;25:1497-504
Worm SW, Sabin C, Weber R, et al. Risk of myocardial infarction in patients with HIV infection exposed to specific individual antiretroviral drugs from the 3 major drug classes: the data collection on adverse events of anti-HIV drugs (D:A:D) study. J Infect Dis 2010;201:318-30.