It is now over 2 years since the widespread recognition of the metabolic and morphologic changes associated with antiretroviral therapy, now most commonly called lipodystrophy. Only now are studies to provide a clinical case definition being performed, which will enable comparison across intervention studies. The causes of peripheral lipoatrophy and other clinical and metabolic consequences of antiretroviral therapy remain enigmatic. Indeed, the publication of hypotheses, rather than investigation of the causes in a thorough and systematic scientific manner, has been the initial response to this phenomenon. This has left patients and many physicians feeling the frustrations of 15-20 years ago when research into treatments for HIV were taking a frustratingly long period of time to emerge from the laboratories. Only through understanding the cause can optimal management of these problems be established.

Evidence from cross sectional surveys has pointed to an interaction between disease and/or immune recovery and drugs. Both PI and nucleoside analogues have been associated with the changes and theories published hypothesizing how they may play a role. However, patients with the syndrome who have never received PIs or have never received nuceloside analogues have been reported suggesting that while these drug classes or specific members of these classes may accelerate the onset of clinical manifestations they are not sufficient alone to cause the problems. It is therefore not surprising that switching therapy away from whichever drug or drug class is the current fashion to blame has not led to resolution of the syndrome. In considering information from surveys or ‘cross-sectional studies’ it is reasonable to remember that associations found in these studies do not necessarily point to causation of the problem. For example, age over 40 years has been associated in several studies with lipodystrophy but, clearly, being over 40 is not the cause of lipodystophy.

A range of in vitro (test tube) and animal studies have provided some pieces of information which may be relevant to whole syndrome or specifically individual manifestations. Currently, we appear to have some pieces of a jigsaw but remain uncertain as to how many more pieces we need and how to start putting those pieces together to come up with a mechanism and hence therapy.

Suggested contributory factors are listed below:

• Interaction with enzymes involved in fat handling

• in vitro interaction with a glucose transporter found on fat cells

• in vitro blocking of chemicals which enable immature fat cells to become fat cells

• in vitro blocking release of triglyceride form fat cells

• in vitro blocking of normal cholesterol production in liver cells

• in vitro enhanced retinoid signaling (IDV) which may cause dry skin and hair loss
• in healthy volunteers increased hepatic VLDL production ( very low density protein- a bad cholesterol) with RTV but not IDV
• in healthy volunteers increased insulin resistance with IDV
• drug-hormone interactions: DHEA:Cortisol (hormones involved in fat and sugar metabolism
• drug-hormone interactions: sex hormones, Growth Hormone/Insulin like Growth Factor

• Mitochondrial toxicity. Differences in visceral and peripheral fat mitochondrial reserve. Mitochondria carry key apoptosis (cell death) factors
• Nucleotide integration in to nuclear DNA or Genotoxicity.
• Shortening of chromosomes end which reduce the life expectancy of cells
• Altered expression of chemicals which are involved in fat handling in mice

HIV Factors

• Altered cortisol response
• Increased fat cell sensitivity to glucocorticoids
• Vpr (a viral gene) activates glucocorticoid receptors
• Tumor necrosis factor a
• Potent inducer of adipocyte apoptosis (progammed cell death)
• Regulates bound leptin (the chemical that regulates appetite)

• Cytokine factors such as IFNa , TNFa , IFNg , IL-1, 6 and 12 are involved in both Immune reconstitution and lipid handling
• Dysregulation of TNFa and association of TNF producing CD8 cells with lipoatrophy
•  Serum corticotrophin, urinary 17-hydroxycorticosteroid, free cortisol (i.e these chemical which are involved in fat and sugar handling may be destroyed at an accelerated rate)

The consequences of lipoatrophy, have been investigated in genetically modified mice. Fat tissue plays a number of roles in mammals including insulation, temperature generation, and fat storage but also hormone production functions such as producing leptin, TNFa , free fatty acids (FFA) and other hormones. Fatless mice have raised FFA, triglycerides and type 2 (insulin resistant) diabetes mellitus with raised insulin; many of the same metabolic problems we see in persons with HIV-related lipodystrophy. It is thought that insulin resistance in hyperlipidaemic states relates to free fatty acid blocking transport of glucose into muscle and fat cells. In persons with HIV, protease inhibitors appear to further inhibit this glucose transporter. Fat free mice have low leptin levels and the animals have poor tolerance of fasting due to low energy stores. The circulating triglycerides, with no adipose tissue to be stored in, lodge in and around visceral organs in particular the liver where they cause hepatic steatosis (fatty liver).

This was an important reminder that hepatic steatosis, typically associated by HIV physicians with associated with lactic acidosis and mitochondrial dysfunction during nucleoside therapy, may be caused through mechanisms other than mitochondrial toxicity. Hepatitic steatosis is a pattern of liver injury that may be seen in both alcoholic and non-alcoholic liver disease. Risk factors for fatty liver include obesity, type II diabetes, hyperlipidemia, total parenteral nutrition, jejuno-ileal bypass surgery, and the use of certain drugs. It may also be observed in pregnancy. Thus, finding hepatic steatosis on a liver ultrasound or biopsy may reflect metabolic disturbances in persons with HIV infection, not mitochondrial injury by nucleoside analogues.

The first published theory focussed on how the HIV protease has similar amino acid sequences to enzymes involved in fat handling including a retinoid receptor (CRABP-1) and liver enzyme for clearing fats from the blood (LRP) and the liver cytochrome enzymes, several of which are known to be involved in PI metabolism. Several parts of this theory have been deconstructed (refuted) by subsequent investigation. For example, IDV has been shown to upregulate retinoid signaling rather than block it as proposed. Additionally, the blockade in fat cell maturation is not related to retinoid inhibition. Furthermore, the structural similarity of protease and CRABP-1 not present at tertiary level, the form in to which the enzyme is folded in humans. Knockout of Lipoprotein Receptor-related Protein in mice does lead to increased cholesterol rich particles but this is subsequently compensated for by other hepatic mechanisms. Ultimately, though the key factor in refuting these proposed mechanisms is that the syndrome occurs in absence of PIs, albeit at an apparently slower rate or lower incidence. PIs may therefore contribute, potentially by other mechanisms as listed above such as diminishing fat cell replacement, blocking fat cell glucose uptake and altering liver fat handling, but do not appear to be the primary cause. They may therefore be seen as ‘straws on the camels back’, adding an additional disruption to fat and sugar handling which accelerates the syndrome. However, as they are not the cause, their removal is not the treatment (although it may help in particular the insulin resistance and in some cases the lipid elevation).

Refutation of the Mitochondrial hypothesis

The second prominent hypothesis is that nucleoside analogues may damage the cellular energy producing mitochondria, by inhibition of DNA-polymerase gamma, the enzyme involved in their reproduction. However, there is little evidence to support mitochondrial toxicity as the mechanism of lipoatrophy. Indeed, considerable data have accumulated to refute this hypothesis.

Two fat biopsy studies recently reported demonstrated only modest reductions (mean 44%) in mitochondrial DNA even in samples from lipoatrophic patients. In inherited mitochondrial DNA disorders reductions in mtDNA to less than 20% are generally required for disease to occur. Importantly, samples from persons with lipoatrophy in some cases (13% of samples in one study) had normal levels of mtDNA and some control samples (from HIV negative individuals) showed diminished mtDNA. This underlines that loss of mtDNA is not necessary for, characteristic of or diagnostic of lipoatrophy in persons with HIV. Additionally, the mtDNA mutation associated with the inherited Madelung’s syndrome (where fat pads grow around the neck often associated with insulin resistance) was absent in all samples in one of the studies.

Furthermore, two separate groups have demonstrated that fat oxidation, a mitochondrial function, is normal or potentially increased in persons with metabolic disturbances on PI+NRTI based regimens. An additional study of muscle biopsies pre and post exercise and lactate levels found normal oxidative phosphorylation (a key mitochondrial function) with a similar recovery rate in lactate and pyruvate levels after exercise and no significant abnormalities of muscle mitochondria. The authors of this study suggested that hyperlactatemia seen in some patients, often suggested to be a marker of mitochondrial toxicity, may in fact be related to decreased clearance of lactate. Many of the patients studied, all of whom clinically had lipodystrophy, had hypertriglyceridaemia. 

Thus, one potential explanation for these findings is that hypertriglyceridaemia leads to hepatic steatosis (as in fatless mice) hence diminished hepatic function and reduced clearance of lactate. The association of lipoatrophy and hyperlactataemia observed in one study may be explained by this mechanism and enables understanding of why isolated mild hyperlactataemia may be relatively common and lactic acidosis remains rare.

Results of culturing fat cells with nucleoside analogues and/or PIs yielded different effects on triglyceride accumulation, fat release (lipolysis) and ATP (cellular energy) levels. Whilst PIs had effects at physiologic concentrations on triglyceride accumulation and at higher concentrations on lipolysis and ATP production, effects were not observed at concentrations normally seen in treated persons for either ZDV or d4T. Thus, NRTIs alone at current doses do not appear to affect adipocyte mitochondrial function. Interestingly, when PI and NRTI were given together to the fat cell culture a marked synergy was observed on a range of adipocyte functions. This may be an important observation, suggesting that the combining of PIs and NRTIs may have greater impact on fat cell function than either alone. This is in keeping with clinical observations.

A further study of d4T in mice, feeding very high doses of 100mg/kg/day of drug (human dose ~1.0mg/kg/day although mice metabolize d4T more rapidly than humans)over 6 weeks no effects on skeletal muscle or adipose tissue mtDNA were observed. Changes in liver mtDNA were observed only in week 1 samples but not at week 6 suggesting a compensatory recovery in mtDNA. This may also explain why mitochondrial toxiciity is reported with these drugs in short term in vitro studies, the problem is a short term one and is generally compensated for by the mechanisms which normally regulated mitochondrial production in the cell. The findings of these studies are perhaps not surprising for the thymidine analogues ZDV and d4T as adipocytes, and other resting cells, are unlikely to express thymidine kinase type 1 (TK1), the enzyme involved in the first step of activation of these drugs, and these drugs are poor substrates for the mitochondrially located TK2. Thus, adipocytes are unlikely to have active (i.e. potentially toxic) concentrations of activated thymidine analogues.

The end of the era of blaming PI exclusively for lipodystrophy came with the recognition that lipoatrophy and metabolic disturbances could occur in PI naïve individuals. However, NRTI-naïve but ART treated individuals are relatively rare making evaluation of NRTI-sparing more difficult. A first survey from the DP-006 study indicated that patients in the EFV-IDV arm of this study had been diagnosed with lipodystrophy despite having never received nucleoside analogues. Two further recent reports, evaluating patients on dual PI regimens (both RTV-SQV) observed lipodystrophy in NRTI spared individuals but higher rates when NRTIs were added to the regimen. In the Prometheus study, after 96 weeks follow-up, 22/88 (25%) of d4T/RTV/SQV patients but also 7/87 (8%) RTV/SQV alone treated patients were diagnosed with lipodystrophy, including 5/44 (10%) RTV/SQV treated patients who had never received NRTIs. A second study with 144 weeks of follow-up, mostly in persons who discontinued prior NRTI therapy and then received RTV/SQV found 6% of NRTI spared individuals had both buttock and facial wasting with 9% having increase in waist size. Patients who added NRTIs during the study were more likely to have lipoatrophy. These studies provide several important observations. Most importantly that lipodystrophy/lipoatrophy can occur in the absence of NRTIs. Previous studies have suggested a that use of PIs can accelerate the onset of or increase the incidence of lipoatrophy in NRTI treated individuals. These data suggest this interaction works both ways, that NRTIs also accelerate the onset or increase the incidence of lipoatrophy in PI treated individuals.

Interest in the potential role of cytokines in lipodystrophy is growing. This interest was first stimulated by the association found in some observational studies with either complete viral suppression, a history of low CD4 and, in particular, the rise in CD4 count with therapy and lipodystrophy. An association between a range of cytokines such as TNF, interferons and some interleukins with HIV-related wasting has been previously noted. Additionally, a study of HIV negative individuals who had undergone bone marrow transplant (hence were previously both immunodeficient and immunosuppressed) but were now mostly off immunosppressants has reported similar metabolic disturbances to those seen in HIV infection, possibly providing a corollary for the HIV syndrome. Obesity but not fat wasting were also reported from this study.

Recent data in HIV patients have indicated that dysregulation of TNFa production may occur in persons on HAART, with some cells continuing to produce high levels of TNFa . In particular, an association with the proportion of TNF producing CD8 cells and lipoatrophy was noted in one study. TNFa is a known inducer of apoptosis (programmed cell death) in adipocytes and may be produced by adipocytes or by T cells located in adipose tissue. In a study evaluating gene expression in liver cells of mice the use of high doses of nucleoside analogues resulted in increased expression in a number of genes including TNFa . Clearly, this study need to be repeated at exposures in mice which equate to normal human exposures and with drug combinations as, for example, low exposures of one drug may lead to one set of effects, higher exposures a different set of effects. The additional interesting finding from this study was that co-administration of the antioxidant vitamins C and E blocked the majority of drug effects on gene expression and associated metabolic disturbances. Many individuals with HIV (+ lipodystrophy) already take high doses of these antioxidant vitamins hence they are unlikely to hold the answer to the management of lipodystrophy. However, adequate intake of these vitamins may be prudent for individuals on nucleoside analogue therapy.

Broadly, management approaches to lipodystrophy depend on the current assumptions of ætiology (cause). The approaches tend to be dictated at present by ‘fashion’ and perhaps ‘marketing’ rather than fact and science. The risk versus benefit of these approaches have not been tested so individuals switching must consider that they may risk their long-term HIV management in exchange for an uncertain outcome with regard to their lipodystrophy. Switching is therefore best evaluated within clinical trials.

Current approaches include:

• Assume PI ætiology (cause) Switch to NNRTI/Triple NA regimen
• Assume Thymidine (AZT, d4T) analogue ætiology Switch to ddI, ABC based regimen
• Assume NA (nuke) ætiology switch to PI + NNRTI regimen
• Assume Cytokine ætiology Use SIT/pulse therapy, consider loose viral control
• Assume Multifactorial Treat individual manifestations
• Assume Adipocyte Apoptosis use Rosiglitazone.

Similarly, adding new agents in to the regimen risks interactions with the antiretrovirals, side effects and toxicities from the new drugs and often only addresses one aspect of the syndrome, for example cholesterol elevation, rather than addressing why the problem has arisen. This may be analogous to the old lady who swallowed the fly who then swallowed a spider to catch the fly and finished up swallowing more and more things to deal with the consequences of the last thing she swallowed. Coughing up the fly would have been a better solution. Ultimately, addressing the root cause of lipodystrophy will help provide the answers to management. Only recently have we begun to dig for that root.