Normal growth is the result of a complex interaction between genetic and environmental factors. The Infancy-Childhood-Pubertal (ICP) model of growth describes appropriately the growth of an individual. In growth during infancy, nutrition and growth hormone (GH) are the primary drivers. During childhood, GH and thyroxine are critical for normal growth, whilst during puberty, GH and sex steroid are needed. Growth hormone deficiency (GHD) has an incidence of 1 in 4000 - 1 in 10000. The causes of GHD include congenital and acquired factors. Congenital causes include genetic, structural abnormalities such as septo-optic dysplasia, or midline facial defects such as cleft lip and palate. Acquired causes include trauma, infection, tumours such as germinoma and craniopharyngioma, effects of treatment in childhood cancer survivors, or psychosocial factors. The symptoms of GHD include hypoglycaemia in infants with hypopituitarism, midline facial hypoplasia, short stature, fat accumulation, muscle hypotonia, high-pitched voice, and bone age delay. However, the diagnosis of GHD can be far from straightforward. Evidence suggests that the diagnosis is dependent upon a combination of clinical, auxological, biochemical and neuroradiological data. The detection of GHD in early life is challenging. One reason is that two GH stimulation tests are needed for diagnosis according to consensus guidelines, but there are several GH stimulation tests, and each test varies in its sensitivity, specificity and reliability. The GH cut-off level for GHD is defined according to the individual centre, but it varies according to the test used and the assay, so it needs careful evaluation. Additionally, patient status such as obesity and the lack of sex steroid priming in the peri-pubertal age influences GH secretion and the provocation test result. It is reported that BMI SD score was inversely associated with peak GH to provocative testing, so a higher BMI may lead to over-diagnosis of GHD. Sex steroid priming prior to GH stimulation test is effective at the pubertal age. Neuroradiology may be extremely useful in the diagnosis of GHD; in patients with congenital hypopituitarism, brain MRI may reveal ectopic posterior pituitary, often with an absent stalk and a small or normal anterior pituitary.
Since the recent advances in molecular genetics, it has been possible to make a diagnosis of GHD genetically in rare cases. Isolated GHD is classified into three groups. Type 1A and 1B are inherited in an autosomal recessive pattern, but type 2 is inherited in an autosomal dominant manner. Type 1A and 1B is caused by mutations in the GH1 gene, and type 1B is also caused by mutations in the GHRHR gene. Homozygous GHRHR mutations were first identified in patients from India and Brazil. The patients have severe dwarfism with early growth failure, hypoglycaemia, midline facial hypoplasia, and low but detectable GH/IGF-1 concentrations. They respond to GH therapy very well. Isolated GHD type 2 is caused by splice site or missense mutations in the GH1 gene that have a dominant-negative effect. Within the GH1 gene, exon 3 has a very weak 5’ acceptor, but exon 4 has a stronger 5’ acceptor. This leads to exon-skipping of exon 3, and results in the formation of a 17.5kDa GH molecule, which is toxic to somatotropes, and can also affect other cell types resulting in GH, ACTH and gonadotrophin deficiencies as well. Recently, in patients with severe isolated GHD and pituitary hypoplasia, RNPC3 gene mutations were detected. We are at an early stage of understanding the molecular basis of GHD and related phenotypes, and the use of genetics to diagnose these conditions is as yet not well-established.
Treatment of GHD entails the administration of recombinant human GH (rhGH) at a dose of 20~25μg/kg/day. Responsiveness to GH is highly variable. Mean final height SDS gains have been reported in large cohort studies such as the KIGS study, and earlier GH treatment is associated with improved final heights. For the study of safety of GH, it is important to monitor IGF-1/IGFBP-3. Short-term safety of growth hormone is generally considered satisfactory. In terms of long-term cohort studies, the Safety and Appropriateness of Growth hormone treatment in Europe (SAGhE) study has recently revealed new insights into the safety of GH treatment. In this study, French researchers reported that mortality was increased in a population of short children treated with recombinant GH (standardized mortality ratio 1.33), and high doses of GH were associated with increased mortality (adjusted SMR 2.94). Mortality due to cardiovascular disease, bone tumours, and cerebral hemorrhage were particularly increased. On the other hand, researchers in Sweden, Belgium, and the Netherlands concluded that the majority of deaths were caused by accidents or suicides in their cohorts. In addition, in the Swedish cohort, it was revealed that the ratio of observed/expected deaths was not increased in childhood rhGH-treated patients. In other studies, a carcinogenic effect of rhGH has not been supported, but second primary malignancy risk was increased in GH-treated patients with previous tumours. With respect to other side-effects, type 2 diabetes mellitus was more often observed in SGA and Turner syndrome. Recently, it has been revealed that untreated GHD may be associated with neurocognitive deficits. In brain MRI, volumetric analysis revealed reductions in volumes of certain structures in untreated GHD patients. In addition to height, other benefits of GH therapy include a change in body composition as well as cardiovascular effects. It has been reported that GH secretion recovers in 25-75% GHD patients, so it is important to retest GHD patients with stimulation testing after stopping GH therapy once final height has been attained.
To conclude, early diagnosis of GHD is still challenging but important to achieve a good final height. GH replacement therapy is widely beneficial, and is generally safe if used for treatment of appropriate indications at recommended doses.
（文責 山本賢一）
