Endocrine Abstracts

Background: The glucagon tolerance test (GTT) is used to diagnose growth hormone deficiency (GHD) in younger children or in patients where the insulin tolerance test is contraindicated. We assessed GTTs carried out over five years in a tertiary paediatric centre to assess growth hormone (GH) and cortisol axes. The aim was to assess at which time points the GH peak was observed, and assess whether any predictive value is gained from demographics or IGF-1.

Methods: 55/58 tests (3 excluded due to incomplete data) from 52 patients were analysed. We assigned risk groups based on indication; high-risk referring to oncology patients (3/55) and low risk to patients with isolated short stature and syndromic patients. Z-scores were used for height, weight, BMI and IGF-1 concentration (sex, age adjusted). A binary logistic regression was performed to analyse height, weight, BMI and IGF-1 Z-scores for both normal and abnormal GH results.

Results and Discussion: In 19/55 (35%) tests the GH response was suboptimal (<6.7 mg/l). The peak sample precluding GHD occurred at 120 minutes from glucagon administration. No diagnostic samples occurred before 60 and after 180 minutes. Cortisol response was suboptimal (<420 nmol/l) in 10/55 tests. The lowest glucose levels were recorded at 90 (26/55 tests) and 120 (21/55 tests) minutes. Hypoglycaemia (<2.2 mmol/l) occurred in 9/55 (16%) tests. All three high-risk patients had GHD. Patients with GHD had significantly lower IGF-1 (mean Z-score = -1.21 for GHD, -0.61 for non-GHD, P=0.009). No significant difference was found when comparing height, weight and BMI Z-scores between GHD and non-GHD groups.

Conclusion: The mode single peak sample precluding GHD occurred at 120 minutes from glucagon administration followed by the 90 minute time point. IGF-1 was lower in the GHD group. We suggest adding a time-point at 105 minutes to reduce the false positive rate. We also note that a spontaneous peak prior to glucagon administration doesn’t preclude a second peak later during the test.