Report 1: 7th Conference on Retroviruses and Opportunistic Infections

Here is the first report of more to come from guest writers for NATAP.

Fat Redistribution and Metabolic Disturbances

Dr Graeme Moyle, MD, Associate Director of HIV Research, Chelsea and Westminster Hospital, London, UK. 

The introduction of potent three or more drug antiretroviral therapy regimens has dramatically improved health and life expectancy for people with HIV-1 infection. However, several clinical and metabolic phenomena have been observed in individuals receiving combination antiretroviral therapy that represent significant management challenges. The metabolic abnormalities include altered fat handling leading to elevations of both cholesterol and triglycerides and altered sugar handling sometimes leading to glucose intolerance or diabetes mellitus. Physical changes include altered body fat distribution, both loss of subcutaneous fat, mostly in the face and periphery (sometimes called lipoatrophy) and localised fat gain, most often inside the abdomen, but also localised fatty masses and breast enlargement. The presentation of problems vary with some individuals manifesting only a single change whilst others may have multiple problems. In general, metabolic disturbances are noted prior to body shape changes. 

The cause?

The causative mechanism of these metabolic and clinical phenomena remains speculative. Several hypotheses exist. Both protease inhibitors and nucleoside analogues have been suggested to be key to the causation of these changes. However, the clinical and/or metabolic changes have been reported in persons receiving PI-containing, PI-sparing and nucleoside analogue-sparing regimens. As these changes have not been reported in untreated persons, some physicians have also speculated that these changes are a consequence of therapy. (note from Jules Levin--Don Kotler has reported observing similar body habitus changes in people with HIV as we see now in the early days of HIV prior to the use of antiretroviral therapy).

Studies with a mice model of fat redistribution suggested that a high fat diet may increase the effects of PI therapy on metabolism. However, differences in effects were noted between PIs suggesting that fat and sugar handling problems may not simply be a class effect. Individual genetics may also be influential in determining responses. For example Amprenavir and saquinavir actually reduced triglycerides in mice fed a low fat diet. In mice on a high fat diet indinavir, nelfinavir and, to a lesser extent saquinavir, were all associated with rises in triglycerides (abstract 37). Indinavir was found to have specific effects on retinoid receptors, a key part of the fat handling pathway (abstract 38). This may not only provide a direct mechanism for the impact of indinavir on fat handling but also explain why dry skin and lips, hair loss and nail problems are more common with indinavir. Other data, looking at the influence of PIs on liver cells found nelfinavir and ritonavir stimulated increased production of triclycerides from these cells (abstract 39). These data all provide support for the impact of PI on fat handling but do not necessarily say anything about fat redistribution. Little data supporting or refuting the hypothesis regarding nucleoside analogues and mitochondrial toxicity in fat cells were presented. However, in persons with abnormal triglycerides on a NRTI+PI therapy lipid oxidation was noted to be increased relative to PI nave HIV+ persons with normal fasting triglycerides, implying that mitochondrial function, responsible for fat oxidation, was not effected (abstract 42).

Risk Factors

The prevalence of fat redistribution is not known an appears dependent on duration of therapy. In a large survey in Australia, involving mainly male patients, 54% reported body shape changes. Changes were more common in persons on PI-based regimens (73%) compared to PI-nave (33%). Surprisingly, 12% of treatment nave patients also reported body shape changes mostly peripheral fat loss (abstract 201). Several studies including the large HOPS survey of 1077 patients across the US (abstract 23) have suggested risk factors for the development of metabolic and clinical abnormalities including age, body shape prior to therapy, duration of both HIV infection and therapy, and a low CD4 count in the past. Many of these factors were also observed in the Australian survey which also found risk was increased in the presence of virological control (abstract 201). The reasons for this association are unclear but may simply relate to greater drug exposure in persons who achieve viral control. More controversially, it has been suggested that differences exist between the protease inhibitors, and the thymidine analogues. Clearly insufficient information exist at present to make treatment recommendations based on what is generally conflicting evidence. In HOPS survey, the use of stavudine (d4T) and indinavir were associated with a greater risk of fat redistribution. Studies from France (abstracts 19, 20) and Spain (abstract 15) also observed an association with greater risk of clinical fat redistribution in patients currently receiving d4T rather than zidovudine (AZT). However, many of the patients in these studies had had prior treatment with AZT and may have had other factors that biased the analysis. Indeed, other studies could not determine differences between nucleoside analogues, although the cumulative exposure to these drugs appeared to remain a risk factor for fat redistribution (abstracts14, 17, 46). (note from Jules Levin--In abstract predictive factors of at least one sign of lipodystrophy at months one and four were looked at-- demographics, HIV/AIDS history, prior treatment with NRTIs, baseline body mass index -BMI- CD4 count, CD8 count, HIV RNA, PI and NRTI prescribed, CD4 and HIV-RNA responses to HAART. At month 20, there were no predictive factors for lipodystrophy. At month 12 lipodystrophy was associated with host factors.) An evaluation of cholesterol and triglyceride changes in 138 UK patients receiving their first treatment regimen found no differences between the choice of NNRTI (efavirenz or nevirapine) or between the choice of d4T and AZT (abstract 21). Gender and race may also be factors determining risk and presentation of fat redistribution. Body fat accumulation may be more common than fat wasting in women (abstracts 24,26), the breasts, abdomen and neck being most often involved (abstracts 24-26, 28). African-Americans appear to have a lower risk of fat redistribution than other ethnicities (abstract 24). 

Making the diagnosis

Self report with confirmation by physician examination appears the most popular means of assessing body shape changes, the two assessments show a high level of correlation (abstract 21). Dual X-ray absorptiometry (DEXA) scans are also able to track body fat (and bone density) changes, although difference between machine calibration makes comparison of data from different centres challenging. However, DEXA appears to be able to detect differences between treated patients experiencing fat redistribution, where ratios of truncal to peripheral fat are higher than either HIV negative controls or untreated individuals. Differences between PI treated and PI nave patients were not observed (abstract 202). Similarly, single slice CT scan demonstrates marked differences in subcutaneous and visceral (intraabdominal) fat between persons presenting with fat redistribution and HIV negative or HIV positive untreated controls. Again, fat redistribution with PI nave and experienced patients were indistinguishable by this method (abstract 22).

Consequences of high cholesterol

It remains unclear as to the extent to which lipid disturbances associated with HIV infection (low HDL, the good cholesterol) and, in particular some PIs (increases in LDL, the bad cholesterols) with have on the risk of future vascular disease. Several small studies evaluating the thickness of the carotid artery (a marker of risk for stroke) provided conflicting information. Overall, however, the impression was that persons with HIV infection may show trends to greater carotid wall thickness (abstracts 30-32). However, there remains little evidence that at least the short-term risk of coronary artery disease is increased by PI therapy (abstract 33, 34). 

Switching studies

Switching studies that are now providing data were mostly initiated in the era when PIs were seen as the chief culprits for fat redistribution. In general these studies involve switching to a PI-sparing regimen without changing the background nucleoside analogue thearpy. Unfortunately, in many studies, often for recruitment reasons, no control group of continued therapy is included. Switching to nevirapine appears to be associated with significant decreases in triglycerides, improvements in insulin resistance and HDL cholesterol but more modest changes in total cholesterol (abstract 45, 206). Fat mass measurements did not improve over 24 weeks. The frequency of adverse events varied between reports. In one study of 39 patients switched to nevirapine only 10% of patients reported any rash, possibly because antihistamines were used at nevirapine initiation. However a further 6 patients experienced hepatitis. (note from Jules Levin-- At week 48, quality of life improvement was reported, no significant improvements in body-shape were reported . But, she reported there was a stabilization in the body changes in the NVP group showed by DEXA measures. Ruiz suggested body shape improvements may take longer) (abstract 206). In a study of 40 patients switched to nevirapine, 6 (15%) developed severe rash, 1 hepatitis and 1 person required switching back to the PI due to viral rebound (abstract 45). In an Australian study where PI was switched for an intensive combination of nevirapine, abacavir, adefovir and hydroxyurea a high rate of adverse events but modest benefits on lipids but not clinical fat redistribution (note from Jules Levin-- Switching led to reduced central fat and improved lipid parameters, but decline in peripheral fat and overall muscle mass) (abstract 205). Benefits in terms of body shape and fat mass were not observed with switching to efavirenz and both cholesterol and triglyceride value tended to remain stable after switching to this agent (abstracts 46-50), although modest improvement in HDL cholesterol was reported by one group (abstract 48). Maintenance of virological control after switching ranged from 90-100%, with several investigators reporting benefits in terms of adherence and patient well being. (In one study -abstract 48- 76% of patients subjectively reported improvement in their body shape, but not back to prior status. In study in abstract 50, 11 patients -55%- reported a partial improvement in body changes that was related to a significant decrease in WHR and intra-dominal fat thickness. Perihperal fat was reported unchanged. Severe adverse events occurred in 3 patients in whom EFV was substituted by NVP.) 

Switching to abacavir looks a promising approach for switching. In a sub-study of a larger controlled substitution trial, no viral breakthroughs were seen in either the 17 patients who switched from PI to abacavir or in the 17 continuation patients (note from Jules Levin-- in the larger study 200 people, individuals were randomized to abacavir or continuing PI. There were 3 virological failures in abacavir arm (>400) and 2 in PI arm). Cholesterol fell markedly in the switch patients (a median 0.75mmol/l fall) compared with no change in the PI group. Resolution in breast enlargements was seen in 2 of 4 patients with this problem at baseline, reduction in abdominal enlargement in 5 of 10, and in leg vein prominence in 3 of 4 were reported after switching (abstract 51). 

Switching of d4T to AZT or abacavir was also found in an uncontrolled cohort to lead to significant falls in triglycerides and increases in subcutaneous fat fold thickness after switching (abstract 52). Also, 11 patients out of 36 in study evaluated over 9 months reported a major improvement in their fat redistribution and 21 reported a partial improvement (note from Jules Levin-- ".abdominal and midthigh subcutaneous fat area increased by 32% and 36%.biceps, triceps, and suprailiac skinfold increased by 38%, 23%, 27%, 36%, respectively").

>From a lifestyle stand point whilst a lower fat diet appears to help mice cope with PI effects in fats, a 16 week 1 hour 3 days/week exercise regimen was found to not only increase muscle strength but also reduce triglycerides in HIV infected men on antiretroviral therapy. 

Lactic acidosis

Lactic acidosis is a potentially serious metabolic disturbance associated with nucleoside analogues. It is thankfully rare, but remains challenging to diagnose, particularly in initial stages when symptoms include fatigue, loss of appetite and stomach upsets The ideal way to perform the key diagnostic test, lactate, remains controversial. One report at this meeting suggested that raised lactate was present at a rate of 1.4 episode for ever 1000 months of therapy (abstract 56). Prompt cessation of therapy in the 20 patients this group found with raised lactate may have been important in the absence of fatal episodes in this cohort. Whilst lactic acidosis has been reported with a range of treatment regimens several studies suggested that it may be most common with d4T containing regimens (abstracts 55, 56, 57). (note from Jules Levin-- its connection to lipodystrophy was questioned at a BMS meeting in SF prior to HRC where it was reported from one study that people with lipodystrophy were not generally found to have elevated lactic acidosis. It was suggested that the liver may play a role in this finding and in lipodystrophy symptomology).


Two years after the recognition of fat redistribution the mechanism remains unknown. Pre-clinical data suggest PIs have clear effects on fat handling through a range of mechanisms. All PIs do not appear the same. However, PIs are clearly not the whole story. Whilst the impact on blood fats does not appear to differ between nucleoside analogues some surveys suggest differences may exist for rapidity or risk of developing physical changes. Presentation of the syndrome(s) is varied and no standardised diagnostics exist. Special scans such as DEXA and CT scans appear to be the most useful. A range of interventions have been tried. Modest benefits appear to be gained in switching from PI- based regimens although it is possible that some of these benefits relate to improved well being and dietary changes. Switching between nucleosides, in uncontrolled studies, also suggest benefits. Diet and exercise appear under evaluated at the present time but represent something everyone should have access to. More expensive therapies such as growth hormone, for which no new data were presented, have also suggested benefits in small studies, predominately in fat accumulations (note from Jules Levin-- fat depletion in periphery is not reversed by growth hormone, and glucose can increase). The most challenging clinical problem remains repleting lost fat, no study having convincingly shown regain of peripheral fat.