GHRH plus arginine in the diagnosis of acquired GH deficiency of childhood-onset

We evaluated the GH-releasing effect of GHRH plus arginine (ARG) in 36 patients (22 males and 14 females) with acquired GH deficiency including idiopathic inflammatory pituitary stalk thickness (n = 15), Langerhans cell histiocytosis (LCH) affecting the hypothalamic-pituitary area (n = 11), and craniopharyngioma (n = 10). All of the patients (mean age, 9.6 ± 3.1 yr; range, 5.6-20.8) showed GH response less than 10 μg/liter after 2 pharmacological stimuli and were tested with GHRH plus ARG at a mean age of 11.2 ± 4.1 yr. Twenty-nine patients had vasopressin deficiency, 10 had TSH deficiency, 8 had gonadotropin deficiency, and 4 had ACTH deficiency. The median peak GH response to insulin test was 2.1 μg/liter (range, 1.1-2.9), whereas it was 1.5 μg/liter (range, 1.3-2.4) after ARG. The median peak GH response to insulin was significantly lower in the patients with craniopharyngioma (1.4 μg/liter; range, 0.8-1.7) than in the patients with idiopathic pituitary stalk thickness (2.2 μg/liter; range, 1.0-2.4) or with LCH (2.6 μg/liter; range 2.0-4.3, P = 0.02). The median peak GH response to ARG was significantly lower in the patients with idiopathic inflammatory pituitary stalk thickness (1.3 μg/liter; range, 0.8-1.8) than in those with craniopharyngioma (1.5 μg/liter; range, 1.1-1.6) or with LCH (2.8 μg/liter; range, 1.9-3.2, P = 0.00007). The median peak GH response after GHRH plus ARG was significantly lower in the overall patient population (8.3 μg/liter; range, 4.4-28.4) than in the age-matched controls (49.8 μg/liter; range, 39.9-81.6, P < 0.00001). The median peak GH response was significantly lower in the patients with craniopharyngioma (4.6 μg/liter; range, 3.6-6.3) than in those with LCH (8.9 μg/liter; range, 4.4-28.4) or with idiopathic pituitary stalk thickness (12.6 μg/liter, range, 6.4-24, P = 0.07). Ten patients had a GH response of more than 20 μg/liter after GHRH plus ARG. There was a trend toward a decrease in peak GH response to GHRH plus ARG (r = -0.57, P = 0.06) as patient age increased. For cut-off values of 20 μg/liter, the sensitivity of GHRH plus ARG was 75% (95% CI, 57.8-87.9%) and the specificity was 96.4% (95% CI, 89.9-99.2%); whereas, for cut-off values of 24.2 μg/liter, sensitivity was 86.1% (95% CI, 70.5-95.3%), and specificity was 95.2% (95% CI, 88.2-98.7%). The median IGF-I level did not differ between the children with idiopathic pituitary stalk thickness (57 μg/liter; range, 46-68), those with LCH (55 μg/liter; range, 34-63), and those with craniopharyngioma (41 μg/liter; range, 39-49). The present study confirmed the diagnostic potential of the GHRH-plus-ARG test in children with acquired GH deficiency caused by hypothalamic-pituitary lesion. It stimulates GH secretion to a greater extent in those patients with GH deficiency with primary involvement of the hypothalamic area, e.g. patients with idiopathic pituitary stalk thickness or LCH, than in those with both hypothalamic and pituitary lesion, as in craniopharyngioma. In some patients, the GHRH-plus-ARG test stimulates GH response to a so-called: normal value, suggesting that pituitary responsiveness to GHRH plus ARG may fail to recognize acquired GHD. Finally, the number of pituitary hormone deficits and the patient's age affect the GH response to GHRH plus ARG.