The short Synacthen and insulin stress tests in the assessment of the hypothalamic–pituitary–adrenal axis

Abstract

The best dynamic test for the assessment of the hypothalamic-pituitary-adrenal axis and the interpretation of the cortisol levels, remain a matter of controversy. We aimed to establish normal ranges with current assays, for both the short Synacthen (SST) and insulin stress tests (IST) and then to use these data to examine whether the SST can satisfactorily substitute for the IST in assessment of the hypothalamic-pituitary-adrenal axis. Thirty SSTs and 27 ISTs were performed on different healthy volunteers. The results of all paired tests performed on patients in the last three years are reviewed. Programmed Investigation Unit. Fifty-seven healthy volunteers and 166 patients. Basal serum cortisol concentration and cortisol values obtained at 30 and 60 minutes during the SST compared to the maximum obtained with adequate hypoglycemia (plasma glucose < 2 mmol/l) during an IST. From normal data the mean-2SD 30-minute value during the SST was 392 nmol/l and 60-minute value was 497 nmol/l. The maximal cortisol response (mean - 2SD) during the IST was 519 nmol/l. Sixty patients failed the IST, none of whom had a basal cortisol > 450 nmol/l and only six (10%) had a 30-minute cortisol value > 600 nmol/l. The 30-minute value provided a better index than the 60-minute value. The basal, 30 and 60-minute values during the SST all correlated positively and significantly with the maximal cortisol on IST. The correlations persisted for all microadenomas and macroadenomas secreting prolactin, gonadotrophins or growth hormone, patients undergoing either pre or post-adenomectomy evaluation, and in those patients who had received long-term steroids provided that the medication had been reduced and stopped two days prior to admission. Using a 30-minute cortisol value > 600 nmol/l as a cut-off, the short Synacthen test provides a suitable substitute for the insulin stress test. Adopting this policy will decrease the number of insulin stress tests performed by one-quarter and thus provide a substantial saving without detriment to patient care.