icon_folder.gif   Conference Reports for NATAP  
  3rd International Workshop on Adverse Drug Reactions and Lipodystrophy in HIV
Athens, Greece - October 2001
Back grey_arrow_rt.gif
Written for NATAP by Michael Dube, MD, University of Indiana and the ACTG
  No major break-throughs in therapy for lipodystrophy was reported at this year's 3rd International Workshop.
There was a relative paucity of new information presented at this year's workshop, either regarding the pathogenesis, or the treatment, of lipid disorders. Kumar et al confirmed the results seen in switch studies, reporting a lack of significant effect of Trizivir on cholesterol and triglycerides, as compared to nelfinavir-based ART. Increased total and LDL-C occurred only in the PI containing arms. The Trizal study, of switching to Trizivir compared to continued HAART (mostly PI-based, but some on NNRTI as well), showed improvements in triglycerides and cholesterol.
Burnside et al and Dube et al reported moderate lipid effects with amprenavir-based therapy. Total cholesterol increased by 45-50 mg/dl, LDL-C by ~30 mg/dl, and triglycerides by 30-90 mg/dl among naive subjects initiating treatment. These results did not support the expectation that amprenavir might have lesser effects on lipid metabolism than other PIs. The magnitude of cholesterol increase is of concern, but Dube also reported that HDL-C increased and resulted in an unchanged total cholesterol:HDL-C ratio, and LDL particles did not shift to the more atherogenic small, dense LDL. Severe hypertriglyceridemia, such as is common with ritonavir- containing regimens, was not reported in either trial. Burnside further reported that changing from amprenavir 1200 mg to combination amprenavir 600 mg/ritonavir 100 mg caused a further increase in cholesterol and triglycerides that was of smaller magnitude. Hulse et al confirmed what has been reported in other cohort studies, that RTV-containing regimens tend to have greater lipid disturbances.
(editorial note from Jules Levin: in Dube study of amprenavir-based regimen, he reported finding no significant change in fasting glucose, but fasting insulin increased. Insulin sensitivity (resistance) did not fall significantly by week 8 or week 24, but was decreased at week 48. Six out of 14 study patients experienced new or worsening glucose tolerance by week 24, but fasting hyperglycemia (sugar in blood) did not occur. Interestingly, patients did gain a good deal of weight in the trunk as well as in the periphery. He did not measure visceral fat. Ten of the 14 patients were Hispanic and 2 were Black. This raised some questions by the audience suggesting that Hispanics and Blacks may have different responses in terms of developing body changes or metabolic abnormalities. Dube reported insulin resistance appeared late following weight gain, particularly trunk fat, suggesting that perhaps late develoment of insulin resistance may have resulted from fat accumulation in belly. Loss of limb fat or facial lipoatrophy did not occur. Non-significant body composition changes seen by DEXA were: body weight increased 2 pounds; total body fat increased; lean body mass increased; trunk fat increased; limb fat increased (p=0.054). Trunk:limb fat ratio increased significantly (p=0.005). Using patient reports of body changes as assessed by questionaire: at week 48, 5/11 (45%) reported no change in the amount of flesh in the face and 6/11 (55%) noticed a gain. Similarly, 7/11 (64%) noted no change in leg size, 4/11 (36%) reported an increase. No subject reported a decrease in the size of the legs or a decrease in the amount of flesh (lipoatrophy) in the face since starting study medications.
Fasting glucose was unchanged throughout the 48 week study. Fasting insulin increased from 9.6 to 13. 120-minute glucose was 114 at baseline, 139 at week 8, 146 at week 24, and 122 at week 48 (p=.35). Total cholesterol, triglycerides (by 90%), LDL-c (bad Cholesterol), HDL-c (good cholesterol) increased significantly.
In a related paper Jacqueline Capeau reported on an in vitro study "Differential Effects of 3 Protease Inhibitors (indinavir, nelfinavir, amprenavir) on Insulin Sensitivity, and Fat Cells & Lipoatrophy". By use of a cell system, which I think was a new cell syatem, the objective of the study was to examine the relative impact and mechanism of action of indinavir, nelfinavir, and amprenavir on adipogenesis (fat cell production), response to insulin, and apoptosis (cell death). Capeau found:
  • The 3 protease inhibitors inhibited adipocyte differentiation with a rank order gradually decreasing from indinavir to nelfinavir and amprenavir

  • Long-term treatment with indinavir and nelfinavir promoted cell apoptosis

  • Amprenavir did not lead to apoptosis

  • Rosiglitazone prevents the effect of protease inhibitors on cell differentistion, resistance to insulin, and programmed cell death. Capeau suggested that rosiglitazone could represent a therapeutic possibility based on these test tube studies

  • Insulin sensitivity was measured by the activation of MAPK and Akt/PKB. PIs inhibited insulin action on MAP and Akt/PKB kinases with half maximally effective concentrations for indinavir, nelfinavir, and amprenavir. This suggests to me that PIs inhibited insulin effectiveness. In personal conversation, Capeau told me that amprenavir did not lead to insulin resistance

  • Capeau summarized that indinavir, nelfinavir, and amprenavir affected several adipocyte (fat cells) key functions. Their affects differ in intensity in accordance with preliminary clinical data.

Use of Lipid Lowering Drugs in HIV
Visnegarwala et al reported an uncontrolled, observational experience with lipid-lowering drugs in a VA clinic in Houston. Only modest effects of drugs were seen, with an 11% mean (40% median) decrease in triglycerides among patients who were prescribed fibrates (fenofibrate and gemfibrozil). With a variety of different statin agents, an 11% mean (14% median) decrease in cholesterol was reported. However, many patients also discontinued PIs. In a multivariate analysis, only stopping PIs was associated with improvements in cholesterol. There were many problems with the observational design, and it would not be appropriate to conclude that HIV-infected patients respond less well to these drugs. However, some other reports have suggested a poorer response than in the general population, while others have not. Clearly, more data is needed from prospective, controlled studies of well-defined populations with lipid disorders.
Risk of Heart Disease in HIV
The link between HIV infection, lipid disorders, PI use, and risk of cardiovascular disease continues to be debated. Mooser, in a plenary presentation, discussed the expectation of increased cardiovascular disease as the treated population ages and continues longer-term treatment. On the other hand, Mauss and colleagues reported that the types of lipoprotein abnormalities commonly seen under HAART would be expected to be relatively less atherogenic (increasing lipid deposits on innermost layer of artery walls resulting in thickening or calcification of the arteries) than might normally be expected. What is needed is large cohort studies with systematic collection of treatment, lipid, cardiovascular risk, and cardiovascular disease events over a period of many years in order to answer these questions more definitively.
Increased Fat Plaque Found in Artery of Women with HIV & Fat Redistribution
A particularly troublesome report from Wilkins et al found a significantly increased carotid intimal media thickness (IMT) (fat plaque in the artery) by carotid ultrasound, among women with fat redistribution as compared to normal women matched for age and race from the general population. Carotid IMT can reflect asymptomatic atherosclerosis, and has been shown in multiple studies to correlate with the risk of coronary heart disease. These women, with a mean age of 45 years, had carotid IMT values comparable to those expected in 55 year old women. IMT correlated with variables such as age, blood pressure, and fasting insulin levels, similar to the general population. Although preliminary in nature, this study adds to the growing evidence that those individuals suffering from fat redistribution, who experience a number of adverse cardiovascular risk factors, may indeed be at increased cardiovascular risk. The variables that mediate this risk and their relative contribution to possible premature cardiovascular disease, require further study. Future studies need to take into account the contribution of HIV disease itself, immune reconstitution, drug effects (such as insulin resistance, abnormal lipids, endothelial dysfunction, mitochondrial toxicity), as well as classical risk factors such as smoking, hypertension, family history, physical inactivity, and diet.
More detailed report on this study:
Do Lipid Decreases During Intermittent HIV Therapy Help?
Strategies of intermittent antiretroviral therapy may help mitigate cardiovascular risk. Milinkovic et al reported that among 12 subjects with moderate to severe lipodystrophy, that triglycerides fell approximately 50 mg/dl and cholesterol fell approximately 30 mg/dl during treatment interruptions of 2 months duration. But exactly how much this "on-again, off-again" dyslipidemia represents an improvement over continuous elevations remains speculative. Ultimately, intermittent strategy trials will need to collect objective measures of cardiovascular disease to prove that this approach translates into reduced risk compared to continuous therapy.
So what does this all mean?
Well, it is difficult to come to a conclusion other than "more work is needed". Certainly, much will be learned from large cohort studies underway that are exposed to different types of HAART in terms of risk of cardiovascular disease. Until that time, more intensive measures of pre-clinical disease such as carotid IMT or tests of endothelial function such as brachial ultrasound, will hopefully begin to give some insight into what risk factors play the dominant role in determining cardiovascular risk. In terms of finding the "kinder, gentler" antiretroviral regimen for cardiovascular risk, a more thorough understanding of the specific lipid effects of different agents, weighed against virologic efficacy, will be helpful. It is encouraging that the new BMS protease inhibitor atazanavir (reported at the European Conference that immediately followed the Lipodystrophy Workshop) appears to have little effect on lipids. However, its effects on insulin resistance and other non-lipid cardiovascular effects awaits further study. Finally, interventions, both drug and non-drug, for treatment of dyslipidemia in patients with HIV deserve further systematic study.