729    

Background:  The thiazolidinedione, rosiglitazone (RSG), is an insulin-sensitising drug that has beneficial effects on arterial function in HIV-uninfected adults.  The aim of this study was to investigate the effects of rosiglitazone on endothelial function, as assessed by brachial reactivity, a marker of coronary artery disease in lipoatrophic HIV⁺ subjects taking combination antiretroviral therapy.

Methods:  Patients participating in the ROSEY study, a double-blinded, placebo-controlled trial in which subjects were randomised to receive either RSG 4mg or matched placebo twice daily for HIV lipoatrophy, were eligible to enrol. Recruited subjects had brachial artery reactivity ultrasound studies performed at weeks 0, 12, 24 and 48.  Scanning measurements were taken at rest and after two vasodilatory stimuli: reactive hyperaemia (flow-mediated dilatation, FMD) and sublingual glyceryl trinitrate (GTND).  Both FMD and GTND were expressed as a percentage of the resting brachial artery diameter (FMD% and GTND%).  Intention-to-treat analyses were performed.  Student's t-test or the Wilcoxon rank test were used to compare treatment groups, the Spearman's correlation coefffificent to assess associations between endothelial parameters and covariates of interest, and linear regression to identify independent predictors.

Results:  Of the 108 subjects recruited to the main ROSEY trial, 70 patients enrolled in this sub-study with 57 (81%) completing all four scans (32 RSG, 25 placebo).  There was no significant difference between patients treated with RSG or placebo over 48 weeks in FMD% (mean difference 0.66% [95% CI:  -1.20 to 2.53]; p = 0.89) or GTND% (mean difference 0.56% [95% CI:   -3.01 to 4.20]; p = 0.65).  Resting systolic blood pressure decreased by 3 mmHg with RSG and increased by 5mmHg with placebo (p = 0.03, RSG vs placebo) over the 48 weeks.  The % increase in FMD was independently correlated with both higher baseline diastolic blood pressure (r = 0.28; p = 0.04) and total cholesterol (r = 0.27; p = 0.05).

Conclusions:  RSG treatment did not alter endothelial function, as measured by FMD, in patients with HIV lipoatrophy .

Keywords: cardiovascular risk; lipodystrophy; rosiglitazone