The authors reviewed all published (English language) manuscripts and manually searched all issues of the JAMA, Journal of Pediatrics, Pediatrics, and Acta Paediatrica, and the meeting abstract books of the Lawson Wilkins Pediatric Endocrine Society and the Endocrine Society between 1985-2000 for publications (N=761) that reported primary effects of recombinant human growth hormone (rhGH) on the growth of children.  From this group, the authors culled those papers reporting adult stature in more than 5 healthy children with “idiopathic” short stature treated with rhGH whose heights were below the 10th percentile at the initiation of treatment and who had “normal” GH secretion (³10 ng/mL during provocative testing) and in which more than 50% of the starting population completed the study.  From this pool, 19 articles describing 10 controlled studies (N=434) and 34 articles reporting 28 uncontrolled studies (N=655) were selected for more thorough analysis.  In both groups the mean age at the beginning of treatment was 10-11 years, baseline growth rates were approximately 4.3 cm/year, and therapy with rhGH was maintained for approximately 5 years. 

In the controlled studies, adult stature of rhGH-treated children exceeded that of the control group by 0.84 SD (5-6 cm) with the treated group achieving an adult stature of -1.51 SDs and the control group -2.29 SDs.  The adult stature of the rhGH-treated group exceeded their pretreatment predicted adult height by 0.54-0.65 SDs (+3.6-4.6 cm).  In the uncontrolled studies, the adult stature of the rhGH-treated group exceeded their pretreatment predicted adult height by 0.56-0.63 SDs (+3.8-4.5 cm). The authors concluded that administration of rhGH can modestly increase the adult stature of children with idiopathic short stature.  They estimated the cost of treatment to be approximately $14,170/cm ($35,000/in).  The authors discuss the limitations of this meta-analysis (such as the heterogeneity of the populations treated; absence of data on those children who did not complete the course of treatment with rhGH) and point out that there are no data demonstrating any beneficial effect of treatment on psychological well-being, educational achievement, or vocational advancement.

Editor’s Comment: The authors are to be complimented on the completion of an arduous task.  Of concern to this reviewer is the inclusion criterion for short stature of height below the 10th percentile.  This reviewer cannot imagine that there are any pediatric endocrinologists who prescribe rhGH for otherwise normal children with heights between the 3rd-10th percentiles.  One would very much like to see the data reanalyzed to include only children with heights below the 3rd percentile (or -2 SD) at the initiation of therapy.  Among the questions that would be of interest to answer are: 1) Was the growth promoting effect of rhGH more apparent in those with the shortest stature? 2) Did the children with familial (intrinsic/genetic) short stature respond more/less favorably than did those with non-familial short stature? 3) Did pre-treatment skeletal maturation influence the linear growth response to rhGH?

In addition to the data analyzed by Finkelstein et al, two additional reports of the effect of rhGH on adult stature in children with idiopathic short stature have been published.  Lopez-Siguero et al¹ observed a mean gain in adult height of 4.5 cm in 30 boys treated with rhGH compared to an historical control group of 42 lads.  Wit and Rekers-Mombarg² reported that treatment with rhGH (0.17-0.32 mg/kg/week for approximately 7 years) resulted in a gain in adult height SD score of 1.3 versus baseline height in 53 patients with idiopathic short stature (12 born small for gestational age) as compared to a gain of 0.7 SD in an historical control group of 64 subjects.  There was an increment of 4 cm in adult height over pretreatment predicted adult height in those children receiving rhGH (+0.8 cm in controls).  In children who received the highest dose of rhGH (0.32 mg/kg/week) throughout the study, the increment in adult stature over pretreatment predicted adult height was 7 cm.  These authors concluded that higher doses of rhGH led to greater increments in gain in adult height.  However, they also concluded that in the absence of proven benefit of greater stature on well-being, the ethical controversy about the administration of rhGH to healthy children, and the high cost of rhGH treatment, “rhGH treatment for (idiopathic short stature) cannot be advised in general.”  This reviewer would agree with this conclusion.

Allen W. Root, MD

	1.       Lopez-Siguero JP, et al. J Pediatr Endocrinol Metab 2000.13:1595-1602.
	2.       Wit JM, Rekers-Mombarg LTM. J Clin Endocrinol Metab 2002;87:604-611.