Endocrine Abstracts

Introduction: The insulin tolerance test (ITT) is commonly used for the evaluation of the corticotropic axis. There are different protocols depending on the number and timing of blood samples. The aim of our study was to investigate the predictive factors of a late response during ITT.

Methods: Our study was cross-sectional and descriptive. One hundred patients with suspected adrenal insufficiency (AI) were evaluated. An intravenous rapid insulin bolus of 0.1 to 0.2 units/kg was administrated. Eight blood samples for glucose and cortisol measurements were taken: before the insulin injection, at the time of hypoglycemia and samples every 10 minutes for 60 minutes after hypoglycemia. An adequate response was defined as a serum cortisol level above or equal to 18 μg/dl after a serum glucose level below 40 mg/dl. we considered a late response if it occurred 30 minutes after hypoglycemia.

Results: The mean age was 41.4±15.12 years. The gender ratio (F/H) was 0.2. Forty-two patients (42%) had an adequate response and 58 patients (58%) had an inadequate response. Twenty-six patients (62%) had an early response and 16 patients (38%) had a late response. Mean age was comparable between the two groups (39.5±16.1 years vs 44.8±12.5 years; P=0.26). We didn’t find a difference between early and late responders patients regarding weight, BMI, hypoglycemia onset time and fasting blood glucose (P=0.48, P=0.60, P=0.24 and P=0.34 respectively). Basal cortisol level was lower in late responders patients than in early responders patients (P=0.001). A basal cortisol level ≤ 11.2 μg/dl had a sensitivity of 62% and a specificity of 82% to predict a late response.

Conclusions: Our study showed that the occurrence of late response is frequent. Therefore, it is necessary to multiply the number of cortisol level samples after hypoglycemia, especially in patients with basal cortisol level below 11.2 µg/dl.