Patients with lipodystrophy require individualised care, emphasise Dutch investigators writing in the September 6th edition of the Annals of Internal Medicine, after they found that although rosiglitazone may be able to partially correct fat loss (lipoatrophy), metformin improved fat gain, blood lipids, and vascular function.
Lipodystrophy is one of the major complications of anti-HIV therapy. The body fat changes which can characterise the condition are often disfiguring and have been shown to reduce adherence to antiretroviral therapy. Metabolic changes caused by anti-HIV drugs could increase an individual’s risk of diabetes and heart disease.
Rosiglitazone and metformin are used to improve glycemic control in patients with type 2 diabetes and both have been examined independently as potential treatments for lipodystrophy, although no studies have directly compared the effects of rosiglitazone and metformin for treating lipodystrophy.
Investigators in Utrecht randomised 39 HIV-positive men to receive either rosiglitazone (8mg/day) or metformin (2g/day) and compared the impact of the two drugs on body fat (both abdominal lipoatrophy and abdominal adiposity), lipid profile and vascular function (measured by endothelial function, blood flow through the brachial artery in the arm). The study was conducted between 2003 and 2004 with study medication provided for six months.
The patients had a mean age of approximately 48 years, and had been taking anti-HIV therapy for a mean of five and a half years.
Metabolic effect of treatment
Both rosiglitazone and metformin decreased total glucose and insulin. The mean per two-hour change in glucose for rosiglitazone was – 1.9mmol/l (p = 0.04) from baseline with the corresponding change for metformin being - 1.1mmol/l (p = 0.05). The mean per two-hour change in insulin for rosiglitazone was – 258pmol/l (p = 0.01) and the mean per two-hour change for metformin was – 231pmol/l (p = 0.01).
Metformin had a greater effect on fasting lipids (mean total cholesterol – 0.4mmol/l from baseline; mean triglycerides – 0.6mmol/l) than rosiglitazone (mean increase in total cholesterol from baseline, 0.4mmol/l; mean increase in triglycerides, 0.5mmol/l).
Aminotransferase levels, an indication of inflammation in the liver, decreased in both arms of the study.
Body fat distribution
Compared with metformin, rosiglitazone increased both subcutaneous abdominal fat (p = 0.007) and visceral abdominal fat (p =0.01) significantly.
Median body weight increased by 2kg in patients taking rosiglitazone compared to a median fall of 1kg in individuals randomised to receive metformin. Median body mass index increased slightly in patients taking rosiglitazone (0.4kg/m2), against a small fall in patients taking metformin (median – 0.4kg/m3). Total body fat mass increased by a median of 3kg in the rosiglitazone-treated patients, compared to a median fall of 1kg in patients taking metformin.
A total of 47% of individuals taking rosiglitazone said that they thought that their had been an improvement in their lipodystrophy compared to 22% of individuals receiving metformin.
Vascular function
Compared with baseline, flow mediated vasodilatation improved significantly in patients taking metformin (mean change, 2% increase), but not with rosiglitazone (mean change
Side-effects
A total of eight patients reported side-effects, and all of these side-effects had been previously reported with treatment with either of the study medications.
There were no reports of interactions between antiretroviral or lipid-lowering drugs and either rosiglitazone or metformin.
“Our findings emphasise the importance of individualised care in HIV-infected patients. Although rosiglitazone may partially correct lipoatrophy, metformin improves visceral adiposity, lipid profile, and endothelial function”, comment the investigators.
They caution that as they observed “a detrimental effect on plasma lipid levels with rosiglitazone in some patients…[it] should be used with caution in patients with HIV lipodystrophy and possibly should be avoided in patients who are already hyperlipidimic or given in conjunction with lipid-lowering agents.”
Regarding the use of metformin, they suggest that it might be appropriate “for viscerally obese, overweight, dyslipidemic patients”, but they warn it “might not be appropriate for patients with predominant lipoatrophy since they may have a further loss of subcutaneous fat.”
The investigations acknowledge that their study had limitations. These include the study population consisting entirely of men who did not have severe insulin resistance. The investigators comment, “whether the results can be extrapolated to women or to a more hyperinsulinemic group remains to be shown.” In addition, the study was not double blinded, was not placebo-controlled, and did not measure clinical outcomes.
van Wijk JPH et al. Comparison of rosiglitazone and metformin for treating HIV lipodystrophy. Ann Intern Med 143: 337 – 346, 2005.