Editor's column

How should we test for growth hormone deficiency, and whom should we treat?

There are few data on the long-term outcome of growth-hormone treatment in short normal children. We assessed the impact of growth-hormone treatment on pubertal development and near-final height in girls.

In a randomised controlled trial, we studied ten girls, with a mean age of 8·07 years and height 2 SDs or more below the mean for their age, and eight short untreated controls matched for age, and 20 short untreated girls who did not give consent for randomisation. The girls received either 30 IU/m² somatropin per week as daily subcutaneous injections or no treatment. We assessed pubertal staging and height gain every 6 months.

Eight treated girls completed a mean of 6·2 years' therapy. By a mean age of 16·4 years, their mean height SD score had changed significantly from −2·42 to −1·14 (p=0·008) and they were, on average, 7·5 cm taller than the girls in the control group (height SD scores did not change significantly from −2·55) and 6·0 cm taller than the non-consent group. The timing of each pubertal stage, and the age and amplitude of peak height velocity were similar for all groups.

Growth-hormone therapy effectively increased height SD score among short normal girls started on treatment in early to mid childhood, with no untoward effect on pubertal progression.

Treatment with human growth hormone in growth hormone deficient patients will improve growth rate significantly, initially demonstrating catch-up growth and later bringing forth a normal growth rate. During childhood, a dose of 0.3–0.6 units/kg body weight per week (or 14 units/m² body surface area per week) is recommended, but during puberty the dose should be increased by 50–100%. The goal of therapy is the attainment of a normal final height, which in the past has often not been fulfilled. This was partly due to the inadequate supply of growth hormone. Since recombinant human growth hormone is now available in unlimited amounts, all patients can be treated continuously. The shorter the child is at time of presenting for therapy, the lower final height will be. It is mandatory to start therapy as early as possible. Concomitant hormonal deficiencies must be corrected by adequate therapy.

Despite the fact that growth hormone is diabetogenic, supplementary therapy will not induce diabetes mellitus. Subtle changes in the immune system can be detected but no clinical correlates, such as increased susceptibility to infection, exist. Induction of leukaemia has been suspected as a possible side-effect of human growth hormone treatment but so far there is insufficient evidence to prove that growth hormone is oncogenic.

The unlimited availability of recombinant DNA-derived human growth hormone has led to the broader use of this agent in short stature syndromes not due to classic growth hormone deficiency. Short- and long-term studies have shown that administration of growth hormone has resulted in improvements in the growth rates and predicted heights of children with isolated short stature, Turner syndrome, chronic renal failure, and other short stature syndromes. However, concerns remain regarding the long-term efficacy of this therapy, especially with respect to final, adult height.