Endocrine and metabolic evaluation of human immunodeficiency virus-infected patients with evidence of protease inhibitor-associated lipodystrophy

Multidrug antiretroviral regimens that include human immunodeficiency virus-1 (HIV-1) protease inhibitors are associated with distinct lipodystrophy, hypertriglyceridemia, hyperinsulinemia, and deposition of visceral abdominal adipose tissue. To determine whether these findings are related to abnormalities of adrenal function, we compared the hypothalamic-pituitary-adrenal axes of HIV-positive patients who had evidence of protease inhibitor-associated lipodystrophy (PIAL), control volunteers (CON), and patients with Cushing's syndrome (CS). To elucidate the metabolic consequences of the observed lipodystrophy, we measured basal serum lipids and compared glucose and insulin concentrations during an oral glucose tolerance test. Spontaneous plasma cortisol showed normal diurnal variation in PIAL. Cortisol levels were similar in CON and PIAL, and levels in these groups were less than those in CS at all times of the night or day (P < 0.005). Ovine CRH-stimulated morning plasma cortisol levels were similar in PIAL and CON. ACTH was significantly greater in PIAL than CON (P < 0.05) at 0, 15, and 30 min after CRH stimulation. Urinary free cortisol in PIAL(mean +/- SD, 76 +/- 51 nmol/day) was significant lower than those in CON (165 +/- 64 nmol/day; P < 0.001) and CS (1715 +/- 1203 nmol/day; P < 0.001). However, 17-hydroxycorticosteroid excretion was significantly greater in PIAL (43 +/- 23 mu mol/day) than in CON (17 +/- 8 mu mol/day; P < 0.001), although lower than that in CS (74 +/- 47 mu mol/day; P < 0.01). Scatchard analysis revealed normal glucocorticoid receptor number and affinity in PIAL. Serum triglycerides were significantly greater in PIAL (6.57 +/- 5.63 mmol/L) than in CS (1.78 +/- 0.83 mmol/L; P < 0.001) or CON (1.36 +/- 0.84 mmol/L; P < 0.001). Although triglyceride levels were significantly correlated with body mass index for CON and CS, these were not correlated for PIAL. During an oral glucose tolerance test, similar glucose and insulin values were found in PIAL and CS that were greater (P < 0.05) than CON values at 30, 60, 90, and 120 min. We conclude that the lipodystrophy associated with use of HIV-1 protease inhibitors is a syndrome of increased intraabdominal adiposity with concomitant dyslipidemia and insulin resistance, but without total body weight gain and is distinct from any known form of hypercortisolism. Although urinary cortisol disposition seems to be altered in HIV-infected patients who are being treated with multidrug regimens that include protease inhibitors, the decreased free cortisol and increased 17-hydroxycorticosteroid excretion appear to be unlikely explanations for the observed lipodystrophy. The cause remains to be elucidated.