Reduction of the pituitary GH releasable pool in short children with GH neurosecretory dysfunction

The classical 'GH neurosecretory dysfunction' (GHNSD) refers to slowly growing children with normal GH responses to classical provocative tests but impaired spontaneous GH secretion over 24 h frequently leading to low IGF-I levels. Thus it has been assumed that these subjects have insufficiency of spontaneous GH secretion due to neuroendocrine abnormalities in spite of a normal releasable pool of GH. However, classical provocative tests do not reliably assess the maximal somatotroph capacity; thus it is still unclear if the GH pool is really preserved or not. GHRH + arginine test is more potent than the classical tests and evaluates the maximal secretory capacity of somatotroph cells. The GH response to this stimulus is reproducible and also independent of age and puberty.

We studied the GH response to GHRH (1 microgram/kg iv) + arginine (ARG, 0.5 g/kg iv) in 19 short children with GHNSD (14 boys and 5 girls, age: 12.1 +/- 0.7 years, pubertal stages I-III, HV-SDS between -1.6 and -4.9; GH peak > 10 micrograms/l after classical stimuli but mean GH concentration (mGHc) < 3 micrograms/l). The results in GHNSD were compared with those in 38 short children with idiopathic or organic severe GHD (GHD, 29 boys and 9 girls, age: 11.2 +/- 0.6 years, pubertal stages I-III, HV-SDS between -1.8 and -4.4; GH peak < 10 micrograms/l after 2 classical provocative tests) and in 83 children with normal or familial short stature (NC, 59 boys and 24 girls, age: 11.5 +/- 0.3 years., pubertal stages I-III; HV-SDS > 25th centile, normal IGF-I levels).

Mean IGF-I levels in GHNSD (121.9 +/- 20.3 micrograms/l) were lower (P < 0.001) than those in NC (270.3 +/- 13.8 micrograms/l) but higher (P < 0.001) than those in GHD (72.0 +/- 4.0 micrograms/l). The mean GH concentration (mGHc) in GHNSD (2.1 +/- 0.1 micrograms/l) was lower (P < 0.01) than that in NC (4.9 +/- 0.5 micrograms/l) but higher (P < 0.01) than that in GHD (1.5 +/- 0.2 micrograms/l). On the other hand, the mean peak GH response to GHRH + ARG in GHNSD (43.7 +/- 3.7 micrograms/l) was markedly higher (P < 0.001) than that in GHD (8.2 +/- 0.9 micrograms/l) but significantly lower (P < 0.01) than that in NC (60. 4 +/- 2.7 micrograms/l). All GHD patients had peak GH responses to GHRH + ARG below the 3rd centile limit of normality (20 micrograms/l), while all GHNSD patients had peak GH responses within the normal range. No significant correlation was found between GH peak after GHRH + ARG, mGHc and IGF-I levels in each group.

Our study demonstrates that short children with 'GH neurosecretory dysfunction' show reduction in the GH releasable pool evaluated by the provocative and potent GHRH + arginine test. However, the peak GH response to a single GHRH + arginine test in GH neurosecretory dysfunction is always within the normal range indicating that this test as well as classical stimuli does not distinguish normal subjects from GH neurosecretory dysfunction.