The effect of recombinant human growth hormone with or without rosiglitazone on hepatic fat content in HIV-1-infected individuals: a randomized clinical trial

Donald P Kotler, Qing He, Ellen S Engelson, Jeanine B Albu, Marshall J Glesby, Donald P Kotler, Qing He, Ellen S Engelson, Jeanine B Albu, Marshall J Glesby

Abstract

Background: Hepatic fat is related to insulin resistance (IR) and visceral adipose tissue (VAT) in HIV+ and uninfected individuals. Growth hormone (GH) reduces VAT but increases IR. We evaluated the effects of recombinant human GH (rhGH) and rosiglitazone (Rosi) on hepatic fat in a substudy of a randomized controlled trial.

Methods: HIV+ subjects with abdominal obesity and IR (QUICKI≤0.33) were randomized to rhGH 3 mg daily, Rosi 4 mg twice daily, the combination or double placebo. Hepatic fat was measured by magnetic resonance spectroscopy, visceral fat by MRI and IR by frequently sampled intravenous glucose tolerance tests at baseline and week 12.

Results: 31 subjects were studied at both time points. Significant correlations between hepatic fat and VAT (r=0.41; P=0.02) and QUICKI (r=0.39; P<0.05) were seen at baseline. IR rose with rhGH but not Rosi. When rhGH treatment groups were combined, hepatic fat expressed as percentage change decreased significantly (P<0.05) but did not change in Rosi (P=0.71). There were no correlations between changes in hepatic fat and VAT (P=0.4) or QUICKI (P=0.6). In a substudy of 21 subjects, a trend was noticed between changes in hepatic fat and serum insulin-like growth factor-1 (IGF-1; P=0.09).

Conclusions: Hepatic fat correlates significantly with both VAT and IR, but changes in hepatic fat do not correlate with changes in VAT and glucose metabolism. Hepatic fat content is reduced by rhGH but Rosi has no effect. These results suggest an independent effect of GH or IGF-1 on hepatic fat. The study was registered at Clinicaltrials.gov (NCT00130286).

Figures

Figure 1a
Figure 1a
Relative change in liver fat (as percent of liver fat + water) after 12 weeks of treatment (p=0.32 by general linear model (controlled for baseline)
Figure 1b
Figure 1b
Absolute change in liver fat as percent of baseline, after 12 weeks of treatment (p=0.26 by general linear model (controlled for baseline)
Figure 2
Figure 2
Relationship between the changes in serum IGF-1 concentration and hepatic fat content

Source: PubMed

3
Prenumerera