Objective: The insulin tolerance test (ITT) is the gold standard for assessing GH and cortisol production in pituitary disease. However, areas of uncertainty remain regarding its safety in older people, the optimal duration of the test and its performance in insulin resistant states. Whether basal cortisol concentration can reliably predict an adequate adrenal response to hypoglycaemia remains to be determined.
Patients and design: We performed a cross-sectional retrospective examination of 197 consecutive patients (mean age 41.8 +/- 16.4, range 13-76 years) with pituitary disease who underwent the ITT over 18 months. The standard intravenous insulin dose administered was 0.15 units/kg body weight. Patients with acromegaly or with type 2 diabetes received 0.2 units/kg; those with basal cortisol < 100 nmol/l received 0.1 units/kg.
Results: Adequate hypoglycaemia (nadir glucose < 2.2 mmol/l) was achieved in 87% of patients, amongst whom 17% did not achieve peak cortisol until 120 min. There were no significant adverse events, even in those >or= 65 years. Of the 18 patients who had basal cortisol < 100 nmol/l, 78% achieved adequate hypoglycaemia; 29% of those had an adequate peak cortisol response >or= 500 nmol/l. A receiver operating characteristics curve identified a basal cortisol of >or= 393 nmol/l to predict a normal peak cortisol response to hypoglycaemia.
Conclusions: In specialized units, the ITT is safe even in elderly patients and those with hypocortisolaemia. The standard 90-min. cut-off for the ITT misses one in six peak cortisol responses. Some patients with basal cortisol < 100 nmol/l had a normal cortisol response to hypoglycaemia.