Hypothalamus-pituitary-adrenal axis evaluation in patients with hypothalamo-pituitary disorders: comparison of different provocative tests

Summary Background The insulin tolerance test (ITT) is the gold standard test to evaluate hypothalamic–pituitary–adrenal (HPA) axis in suspected ACTH insufficiency. When contraindicated, alternative tests have been proposed such as metyrapone and ACTH stimulation test. 250 µg ACTH is a supramaximal dose and unreliable in this setting. The diagnostic reliability of 1·0 µg ACTH test is controversial and very low doses have been proposed. Design In 31 patients with hypothalamo–pituitary disorders and normal basal cortisol, we compared the diagnostic sensitivity, specificity and accuracy of metyrapone [metyrapone test (MET) 30 mg/kg p.o.], high (HDT, 250 µg i.v.), low (LDT, 1·0 µg i.v.) and very‐low (VLDT, 0·06 µg i.v.) dose ACTH tests. Receiver operator curve (ROC) analysis was applied with ITT as reference test. Results MET approached the best pairs of values for highest sensitivity (71·4% and 64·3%) and highest specificity (100% and 82·4%) using ACTH and 11‐deoxycortisol (11‐DOC) cut‐off of 17·3 pmol/l and 144·3 nmol/l. Either HDT or LDT sensitivity approached 71·4% with a specificity of 82·4% or 73·3% with a specificity of 80% for cortisol cut‐off of 582·1 or 477·3 nmol/l. VLDT approached the highest sensitivity (57·1%) and highest specificity (88·2%) for a cortisol cut‐off of 364·2 nmol/l. Conclusion Neither MET nor ACTH test can be considered completely reliable for the diagnosis of secondary hypoadrenalism, when compared with ITT that remains the best test. Either MET or ACTH stimulation test, at both high and low dose, show an overall similar reliability, provided that appropriated cut‐off values were considered; testing with very low ACTH doses seems to be misleading.