ESPE Abstracts

Background: Insulin Tolerance tests (ITT) have long been considered the gold standard for dynamic function testing of the hypothalamo-pituitary-adrenal (HPA) axis. Sub-optimal cortisol responses during an ITT in children may be found in children during ITT without a previous clinical suspicion of Adrenal insufficiency (AI). It is not clear what the clinical significance of this is and whether all of these children require formal synacthen testing.

Methods: A 9-year retrospective case review of cortisol responses to insulin during ITT was conducted for children and young people at a tertiary paediatric endocrine centre in the United Kingdom (UK). The primary objective was to establish the outcome of patients who attended for an ITT with a sub-optimal cortisol (<430nmol/L) response and how many had further testing with a standard synacthen test (SST). An optimal cortisol response was considered to be any result greater than 430nmol/l on both ITT and on SST.

Results: 223 patients (72.5% male, 27.5% female) were identified from 2013-2022. The median age was 13.74 (5.74-20.1) years. 204 (91.5%) were included, 19 (8.5%) were excluded due to missing data. 87/204 (42.6%) patients had a sub-optimal (cortisol <430nmol/L) cortisol response to ITT. Of those with a sub-optimal response to ITT, 55/87 (63.2%) proceeded to SST. Only 2/55 (3.6%) patients that underwent an SST had a sub-optimal response. Of those who did not have a SST, 2/32 (6.3%) were commenced on hydrocortisone, 7/32 (21.9%) had a previous normal response to SST, 2/32 (6.3%) had an acceptable early morning cortisol, 8/32 (25%) it was deemed not medically required, there was no documentation for 7/32 (21.9%) patients and 6/32 (18.7%) patients were already on hydrocortisone. Thus, only 4/204 (2%) new patients were diagnosed with AI following initial referral for ITT. Six patients that had a sub-optimal response to ITT were already known to have AI. Therefore, in total, ITT identified 10/204 (5%) patients as having AI despite 87/204 (42.6%) patients having a sub-optimal cortisol response to Insulin.

Conclusion: Low cortisol levels during ITT were identified in nearly half of children undergoing ITT, however these children are rarely confirmed as having AI on synacthen testing. Additional synacthen testing, results in further hospital visits, cannulations, use of resources and parental concerns for a large number of patients. Data on the long-term outcome of cases that have sub-optimal cortisols on ITT, but adequate levels on synacthen testing are lacking.