HIV-associated adipose redistribution syndrome (HARS): etiology and pathophysiological mechanisms

Active antiretroviral therapy has reduced mortality of AIDS (acquired immune deficiency syndrome) and increased both the quality of life and longevity in patients infected with HIV (human immunodeficiency virus). However, long-term effects of HIV infection are increasingly observed [1, 2]. Among the more apparent effects are changes in fat distribution or "lipodystrophy," which includes fat loss or "lipoatrophy" and/or an increase in fat accumulation, also known as "lipohypertrophy" [3]. The pathological aberrations of fat metabolism seen in some HIV-infected patients remain inadequately understood and controversial despite extensive studies[4]. These changes may be an outcome of HIV itself or that of active antiretroviral therapy. Lipoatrophy and lipohypertrophy may occur separately or together in an individual.

The term HIV-associated adipose redistribution syndrome (HARS) describes a fat accumulation disorder in patients with HIV, which has specific characteristics. The primary characteristic is an increase in the amount of visceral adipose tissue (VAT) [5], often observed in patients while they are undergoing active anti-retroviral therapy[6]. In addition, adipose tissue may accumulate subcutaneously in regions such as the trunk and the dorsocervical area – where a significant depot may be referred to as a "buffalo hump". Further, metabolic perturbations including insulin resistance, glucose intolerance, dyslipidemia and hypertension, as well as body image distress may accompany these changes. These metabolic disturbances may portend the development of other, more serious, medical conditions including diabetes mellitus and cardiovascular disease [7–9]. While the most commonly observed form of HIV lipodystrophy is lipoatrophy[10], patients with HARS may or may not present with lipoatrophy of subcutaneous adipose tissue (SAT), including abdominal SAT[11]. In this review, we address potential pathophysiologic mechanisms to explain HARS and its associated health risks.

Although the underlying mechanisms of HIV-associated lipodystrophy and HARS remain unclear, investigators show increasing interest in teasing apart the various aspects of HIV-lipodystrophy. While peripheral fat loss and central fat accumulation may be seen in the same patient, there is no statistically significant relationship between the two conditions[12], so a single metabolic process is unlikely to simultaneously decrease and increase fat in the same patient. Though the term "redistribution" is used in HARS and the Fat Redistribution and Metabolic Changes in HIV Infection Study (FRAM), it is not meant to imply the migration of fat from one depot to another. Rather, it refers to the overall re- or maldistribution of fat as adipose tissue atrophies in some areas, e.g., face, buttocks and limbs, and at the same time fat accumulates or hypertrophies in other areas, e.g. in the abdomen as VAT.

The etiology of HIV lipodystrophy, and HARS in particular, is certainly multifactorial. Intrinsic host factors, disease status, treatment duration and type, as well as other factors probably all play a role. Evidence summarized by Lichtenstein[13] suggests that significant risk factors for fat loss include exposure to and duration of thymidine analogues (particularly the older ones, stavudine, zidovudine), age, CD4-positive count, viral load, duration of therapy, and white race. In contrast, risk factors for abnormal fat accumulation include duration of therapy, CD4-positive count, viral load, age, protease inhibitor use (particularly the older ones, indinavir, nelfinavir), and female sex[14]. These differences in risk factors for the two manifestations of lipodystrophy suggest that the pathogenic mechanisms for fat redistribution, loss versus gain, are probably independent and the result of complex interactions of host, disease and drug factors.

Visceral and subcutaneous fat depots are genetically and metabolically different from each other. Surgical removal of VAT in animals and humans improved insulin resistance and glucose while removal of SAT did not [15–17]. Acute increases in glucose uptake into adipose tissue by either hyperglycemia or hyperinsulinemia increased the expression of several fat-derived peptides to a greater extent in VAT than in SAT[18].

It is known that the metabolic function of adipose tissue is controlled by numerous humoral, paracrine, and intracrine factors[19], as well as by sympathetic innervation (stimulating lipolysis) and parasympathetic innervation (parasympathectectomy can induce insulin resistance in the denervated fat depot) [20, 21]. It is likely that disruption of these complex interacting regulatory pathways, in various ways (e.g. by HIV itself or its treatment), accounts for at least some of the fat maldistribution and metabolic consequences seen in HARS. Chronic disruption of paracrine interactions between lymphoid and adipose tissue may play a role in HARS[19], while HIV or antiretroviral agents may adversely affect neural pathways to different fat depots[20].

Among the observed endocrine and cytokine alterations in HIV lipodystrophy are increased levels of insulin[22] and tumor necrosis factor-α (TNF-α)[19] and defects in the secretion of growth hormone (GH), all of which may alter adipocyte function. The problem in understanding the pathogenesis of HARS, however, is that the observed metabolic effects associated with HIV lipodystrophy are the sum of multiple metabolic disturbances, each a separate component.

Disruption of specific regulatory controls of VAT may account for HARS. VAT may differ from SAT tissue in function, however, recent data from FRAM suggests that both VAT and upper trunk SAT are independently associated with insulin resistance[23]. A critical VAT threshold may exist, whereby signs of metabolic syndrome do not appear until the VAT mass exceeds this level for a given individual[24].

In the fasted state, HARS is characterized by an excessive FFA release by adipocyte lipolysis, resulting in a greater net delivery of FFA to the plasma, ultimately resulting in greater hepatic delivery and export as triglyceride-rich very low density lipoprotein (VLDL) leading to hypertriglyceridemia[22, 41, 42]. In the fed state, a different mechanism results in a second contribution to the hypertriglyceridemia associated with HARS. Examination of the disposal of labeled triglyceride from the plasma chylomicron pool in HARS patients showed a marked retardation of the labeled fatty acids from the chylomicrons compared to controls. Further, of the small amount cleared, a greater proportion of the released FFA was found within the plasma rather than taken into the adipocyte in HARS patients suggesting that there was an adipocyte defect resulting in impaired fat storage [43].

The mechanism underlying lipodystrophy may involve the regulation of adipocyte hormone-sensitive lipase; the result is an accelerated rate of whole-body lipolysis that facilitates the "redistribution" of fat. Results of a study in men with both peripheral fat loss and central fat accumulation, as well as dyslipidemia suggested that a regulatory defect in adipocyte lipolysis could account for both the dyslipidemia and the peripheral fat loss [42]. Hyperlipolytic activity and increased release of FFAs would promote processes for which fatty acids are substrates, for example, the hepatic extraction and conversion to glycerolipids. This mechanism could explain the hypertriglyceridemia seen in HIV-infected patients with lipodystrophy, but does not explain the central adiposity seen in some of them. The authors conjecture that either decreased lipolysis or increased deposition of fatty acids outstripping lipolysis in visceral fat depots might account for the observed central adiposity [42]. This is further discussed below.

Sekhar and colleagues have posited a basic defect in fatty acid metabolism in peripheral adipocytes in HARS patients to account for: the acceleration in lipolysis (primarily in the femoral-gluteal region), release of fatty acids for hepatic re-esterification leading to hypertriglyceridemia, along with decreased clearance of chylomicron triglyceride[43]. The greater availability of fatty acids increases uptake by the visceral adipocytes, which have a higher lipid turnover rate than peripheral adipocytes, favoring greater net deposition of fat and the development of central adiposity. The underlying causal factor may be the HAART agents or proteins expressed by the virus itself. The net result is increased triglyceride deposition in the liver, central fat, and skeletal muscle; and an increase in proatherogenic lipoproteins or "systemic steatosis".

In this "systemic steatosis" model[44], increased uptake of fatty acids in the liver promotes the synthesis of triglycerides and apolipoprotein B, reduces degradation of apolipoprotein B, and leads to hypertriglyceridemia due to increased production of VLDL. Lipid uptake within the central fat depots is higher than in peripheral fat depots (femoral-gluteal regions). High fat diets cause overexpression of the endothelial cell enzyme lipoprotein lipase, which hydrolyzes lipids in lipoproteins. The differential lipid uptake in the "systemic steatosis" model may therefore be due to an increased sensitivity to lipoprotein lipase-activating hormones (e.g. cortisol in omental adipocytes), thus sequestering fatty acids as di- and tri-glycerides in abdominal visceral depots[44].

HARS may be considered a state of functional GH deficiency. Normally, GH is released in pulsatile fashion by the somatotrophic cells of the anterior pituitary in response to growth hormone-releasing hormone (GHRH), somatostatin, insulin-like growth factor (IGF-I), thyroid hormone, and glucocorticoids. Although GHRH is released primarily by the hypothalamus, it is also synthesized in the placenta, ovary, testis, lymphocyte, pancreas, and gastrointestinal tract. The complexity of GH regulation suggests the corresponding complexity of its function in the body. GH concentrations vary inversely with excess weight (obesity). Both reduced GH secretion[49, 50] and increased clearance have been associated with visceral adiposity in non-HIV patients[51].

Rietschel and colleagues have found normal concentrations of IGF-I but reduced GH concentrations in patients with HARS, suggesting greater receptor sensitivity to GH so that a smaller amount of GH stimulates a normal amount of IGF-I release[52]. Patients in this study were HIV-infected men in whom weight and total body fat were normal but visceral fat depots were enlarged while peripheral fat depots were reduced. The study found 33%-38% prevalence of deficient GH response in lipodystrophic subjects (peak GH stimulatory cut-off values 3.0 μg/mL to 5.0 μg/mL with arginine testing). Reduced mean overnight GH concentration (i.e., reduced basal GH concentration and reduced GH pulse amplitude associated with normal IGF-I) was observed in HAART-treated adults with excess VAT. VAT was the most significant predictor of GH secretion in these subjects.

The actions of the hypothalamic hormone, somatostatin, play a role in GH regulation as well as adipose metabolism. Release of somatostatin is counter regulatory in that it inhibits GH secretion. Koutkia and colleagues compared the difference in GH release by GHRH+arginine stimulation with GHRH alone in 13 HIV-infected men with lipodystrophy and 10 HIV-infected men without lipodystrophy after an overnight fast[27]. VAT in the lipodystrophy group was 197 ± 19 cm² vs 66 ± 10 cm² for the non-lipodystrophy men, and 94 ± 13 cm² for the comparison group. The data demonstrated GH deficiency in 18% of the lipodystrophy group vs 5.9% of the non-lipodystrophy group and 0% of the comparison group, using the stringent criterion of 3.3 ng/ml for peak GH response to GHRH-arginine. Among the lipodystrophy patients, the peak GH response to GHRH-arginine was significantly predicted by VAT (P = 0.008), FFA (P = 0.04) and insulin level (P = 0.007) in regression modeling. These data demonstrate increased frequency of GH deficiency in HIV lipodystrophy patients in conjunction with increased VAT[53]. The addition of arginine to GHRH in the lipodystrophy patients led to a 247% greater increase in GH secretion compared to the lipodystrophy patients who received only GHRH. Since arginine blocks somatostatin's inhibition of GH release, this suggests that patients with HIV lipodystrophy and increased visceral fat have elevated somatostatin tone.

Circulating FFAs may impair GH secretion; conversely, GH replacement demonstrates marked reductions in total and net lipolysis and the availability of FFAs for hepatic reesterification[39]. These results highlight the complex relationship between FFA and GH.

The underlying pathophysiologic mechanisms responsible for HARS are multifactorial and not well understood at present. These may include dysfunction of the GH axis, alterations in immune function, along with increased levels of insulin and various other hormones, cytokines and proteins. Factors contributing to the dysfunction of the GH axis may also be multifactorial. Further understanding of these pathophysiological factors is needed for a complete understanding of HIV lipodystrophy. We can expect that future research will gradually elucidate the complex pathogenic mechanisms and the disfiguring, as well as health-threatening, conditions associated with HARS.