HGH Therapy in Normal but Short Children
ABSTRACT & COMMENTARY
Synopsis: Long-term administration of recombinant human growth hormone to children with idiopathic short stature increased adult height to a level greater than the predicted adult height.
Source: Hintz RL, et al. Effect of growth hormone treatment on adult height of children with idiopathic short stature. N Engl J Med 1999;340:502-507.
Hintz and associates and the genentech collaborative Group studied 121 children with idiopathic short stature. All had initial heights below the third percentile, low growth rates, and no evidence of growth hormone deficiency. Treatment consisted of 0.3 mg/kg/week of human growth hormone (HGH) for 2-10 years. Eighty of these children have reached adult height. The differences between the predicted adult height before treatment and the attained adult height was compared in treated children and with control groups consisting of similar but untreated children.
In the 80 HGH treated children who reached adult height, the difference between predicted and attained adult heights was 5.0 ± 5.1 cm for boys and 5.9 ± 5.2 cm for girls.
COMMENT BY MYRON GENEL, MD, FAAP
The ample availability of biosynthetic, recombinant HGH has, not unexpectedly, raised questions regarding potential benefit for children with short stature who do not fulfill classic criteria for growth hormone deficiency. Indeed, there are a large number of children—statistically the lowest 1-2.5% of the population—who can be regarded as having significant short stature, defined as greater than two standard deviations from the mean for the age. Do these children benefit from growth hormone treatment?
It is impossible to give a simple yes or no answer to this question as we lack definitive data that clearly define, either on biochemical or auxologic criteria, which of these children are likely to show sufficient improvement to warrant the emotional and financial cost of therapy. Neither of these is trivial. In an earlier comment for this newsletter,1 I noted the paucity of reported trials that had been carried out to completion of adult height and further noted the difficulties in adequately controlling these studies. The report by Hintz et al of results of a long-term multicenter trial supported by Genentech now provides data in 80 children who had achieved adult height, using what is now a fairly standard dosage of 0.3 mg/kg/week.
The study, widely cited in the media, draws attention to the difficulties in obtaining adequate control data. To begin with, 38 of the original 121 children enrolled (almost one-third) did not complete the study for a variety of reasons. Moreover, as noted by Hintz et al, this is a heterogenous group and responses to therapy were highly variable, a finding amply demonstrated by the wide standard deviations in the data depicted. More significantly, the reported improvements in final adult height were based on somewhat tenuous height predictions, themselves heavily dependent upon the interpretation of skeletal age at the hand and wrist. Hintz et al observe that the method used in the study typically overestimates adult height for short boys by an amount almost equal to the reported gain, while it underestimates for short girls. In discussing the use of predicted height with families considering growth hormone therapy, I invariably compare the accuracy of these predictions to weather forecasts. Both are reasonably reliable but subject to considerable variation. Height predictions may indeed be reasonably accurate within a few inches of the final height, but it is precisely these few inches that are important to families considering growth hormone therapy outside of the accepted indications of classic growth hormone deficiency, Turner’s syndrome, or chronic renal insufficiency.
Ultimately, as I noted in my July 1998 comment, a decision to use growth hormone in these circumstances must be made on an individual basis after thorough discussion with the family of the implications of therapy and the limitations of available studies.1 It is clear that some children with idiopathic, nondeficient short stature do benefit, but by no means all, and it is virtually impossible to identify responders in advance of a therapeutic trial. For the present, identifying these children is as much an art as a science and requires the full understanding of the families (and to the extent possible of the short child), before treatment begins. In the meantime, we continue to await the definitive study. The position of the AAP on this matter is not helpful. Growth hormone treatment is recommended for normal, but short, children whose "extreme short stature keeps them from partcipating in basic activities of daily living," and who have a condition for which growth hormone is effective.2 The only conditions for which growth hormone treatment is regularly effective and recommended are classic GH deficiency, end-stage renal disease awaiting transplantation, and Turner’s syndrome. In a balanced editorial accompanying the Hintz study,3 Sharon Oberfield, MD, points out that the results of the Hintz study does show some increase in stature of HGH-treated children; however, their adult heights were still significantly shorter than expected on the basis of midparental, sex-adjusted target heights. In addition, Hintz et al did not report the percentage of children who either exceeded or were less than 5.0 cm of their predicted heights. Studies have not shown lower self-esteem or body image as a result of height in short children, and HGH therapy was found to have no positive psychosocial effects.4
1. Genel M. Growth hormone treatment for short normal girls? Ped Adolesc Med Rep 1998;3:43-44.
2. Committee on Drugs and Committee on Bioethics, American Academy of Pediatrics. Considerations related to the use of recombinant human growth hormone in children. Pediatrics 1997;99:122-129.
3. Oberfield SE. Growth hormone use in normal, short children—A plea for reason. N Engl J Med 1999;340: 557-558.
4. Downie AB, et al. Psychological response to growth hormone treatment in short normal children. Arch Dis Child 1996;75:32-35.