Dexamethasone in the diagnostic work-up of growth hormone deficiency

Abstract

OBJECTIVE Acute administration of dexamethasone (dexa) has recently been shown to induce growth hormone (GH) release. To ascertain the efficacy of this stimulus in assessing GH secretory status in children, we tested it in a large group of patients with short stature. METHODS We administered dexamethasone at the dose of 2 mg/m2 to 44 short normal children and 19 GH deficient (GHD) children, either orally or intravenously and compared the results of the dexa-test to the more classical clonidine test. RESULTS The oral dexa-test induced a GH peak similar to the clonidine test (clo) (controls clo: 23.8 ± 7.8 mU/l, median 22.8, range 15.2–45.4 vs. control dexa: 20.6 ± 10.8, median 16.8, range 8–47, P = 0.2. GHD clo: 9.8 ± 2.6, median 9.2, range 6.4–13.4 vs. GHD dexa: 9.4 ± 3.4, median 10.2, range 4.6–14, P = 0.8). Its sensitivity and specificity with respect to the clonidine test were 91% (10/11 GHD) and 65% (15/23 controls), respectively. The GH peak after i.v. dexa was smaller than that after clonidine (control clo: 30.6 ± 14 μg/l, median 24.8, range 14.2–62.4 vs. control dexa: 21.6 ± 5.4, median 21.6, range 11.2–33, P = 0.01. GHD clo: 7.4 ± 4.2, median 8.8, range 0.4–11.8 vs. GHD dexa: 6.4 ± 5.6, median 5.8, range 0.4–16.2, P = 0.17) with sensitivity and specificity of 87% (7/8 GHD) and 90% (19/21 controls), respectively. The lower potency of dexamethasone could account for these figures, since when a different cut-off was used (12 mU/l and 11 mU/l for the oral and i.v. route) both sensitivity and specificity were improved. More data are needed to support these findings and establish a clear cut-off. In the control group, no difference was found between GH peak after oral or i.v. dexa but GH-area under the curve (AUC) was larger for i.v. than for oral dexa. No side effects were noted. CONCLUSIONS Intravenous dexamethasone appears to be a promising stimulus for the detection of GH deficiency in children, particularly for use in outpatients.