A comparison of the naloxone test with ovine CRH and insulin hypoglycaemia in the evaluation of the hypothalamic‐pituitary‐adrenal axis in normal man

Abstract

It has been suggested that naloxone might be useful in clinical testing of the hypothalamic-pituitary-adrenal (HPA) axis. We have therefore evaluated this non-selective opioid receptor antagonist, as a test of HPA axis function, and compared the results to ovine corticotrophin-releasing hormone (oCRH) and the insulin tolerance test (ITT). Following i.v. administration at time zero of naloxone 20 mg (n = 12) on day 1, and either oCRH 1 microgram/kg (n = 6) or soluble insulin 0.15U/kg (n = 6) on day 2, venous blood was sampled at times 120, 0, 15, 30, 45, 60, 90 and 120 minutes for cortisol, ACTH and AVP. Peripheral CRH was also measured following naloxone and insulin hypoglycaemia. Twelve normal males (age 20-57 years) with no history of hypothalamic-pituitary-adrenal axis disease. Peptide hormones in plasma samples were measured by radioimmunoassay and cortisol by ELISA. Results are expressed as mean +/- SEM. Following naloxone, there was a highly significant overall rise in ACTH (P < 0.0005) and cortisol(P < 0.0001), but 1 out of the 12 subjects failed to respond. This subject had a normal ACTH and cortisol response to oCRH, indicating normal pituitary-adrenal function. Peripheral levels of CRH also increased significantly following naloxone (P < 0.002), while AVP did not alter significantly (P = 0.38). Maximal levels of CRH were seen following the ACTH peak however, at a time when ACTH was returning to baseline. All six subjects who received oCRH had an increase in ACTH and cortisol, and the ACTH response to oCRH was greater that that to naloxone (P < 0.05). One subject who developed nausea and hypotension following oCRH had a large rise in AVP and very high levels of ACTH and cortisol. Following insulin each subject had symptomatic hypoglycaemia and significant rises in cortisol (P < 0.0001), ACTH (P < 0.0001), AVP (P < 0.0005) and CRH (P < 0.01) were seen. Both cortisol and ACTH responses to ITT were significantly greater than those to naloxone (P < 0.05 for each). The HPA axis response to naloxone is smaller in magnitude overall compared to oCRH or insulin hypoglycaemia and is variable in normal subjects. This variability probably reflects changes in central opioid tone rather than alterations in pituitary responsiveness to CRH. It is unlikely that the naloxone test will replace currently used clinical tests of HPA axis function, particularly in the setting of a possible ACTH deficiency, because some subjects wit ha normal HPA axis appear not to respond to naloxone. As the mechanism involved in the ACTH response to naloxone has not yet been defined with certainty, the naloxone test should not be regarded simply as a test of endogenous CRH release.