Background: Insulin-induced hypoglycaemia (ITT) is considered the gold standard method for the evaluation of the hypothalamo-pituitary-adrenal axis. The standard short Synacthen test (SST) has been proposed as an alternative, but it may be associated with falsely reassuring results and potentially serious sequelae.
Aim: To compare the serum cortisol response achieved in the SST with the one obtained during the ITT in patients with hypothalamo-pituitary disease and to investigate whether the former may obviate the need to perform the latter.
Patients and methods: The results from thirty patients (22 males, median age 39, range 1763) who underwent both tests (ITT: 0.150.3 IU of soluble insulin/kg body weight, i.v. SST:250 mcg Synacthen i.m.) were analysed retrospectively. In no subject the two tests were separated by any therapeutic intervention.
Results: Twelve patients had peak serum cortisol >580 nmol/l (time 30) in the SST and 17 had peak >580 nmol/l in the ITT. There was a positive correlation between peak serum cortisol in both tests (Spearmans correlation coefficient=0.70, P<0.01). Among 12 subjects with peak >580 in the SST, 1 (8.3%) did not achieve peak >580 in the ITT. Among 18 with peak <580 in the SST, 6 (33.3%) achieved peak >580 in the ITT. These 6 subjects had median peak in SST of 541 (range 505575). Among 17 with peak >580 in the ITT, 6 (35.3%) did not achieve peak >580 in the SST. Among 13 with peak in the ITT <580, 1 (7.7%) achieved peak >580 in the SST.
Conclusions: Using the cut-off of 580 nmol/l, the SST may overestimate cortisol deficiency, but the chance of missing this diagnosis is small. Cases with borderline peak serum cortisol in the SST require re-evaluation with an ITT; otherwise, and particularly when other pituitary hormone deficits are present, the SST suffices.