| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Department of Pediatric Endocrinology and Institut National de la Santé et de la Recherche Médicale U561 (J.-C.C., J.-L.C.), Groupe Hospitalier Cochin-Saint Vincent de Paul and Faculté Cochin-Université Paris V, 75014 Paris, France; Department of Pediatric Endocrinology (P.C.), Hôpital Debrousse, 69005 Lyon, France; and Department of Pediatric Endocrinology (P.R.), Hôpital des Enfants, 31059 Toulouse, France
Address all correspondence and requests for reprints to: Prof. Jean-Claude Carel, Pediatric Endocrinology and Institut National de la Santé et de la Recherche Médicale U561, Groupe Hospitalier Cochin-Saint Vincent de Paul, 82 av Denfert Rochereau, 75014 Paris, France. E-mail: carel{at}cochin.inserm.fr.
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
A total of 168 short children born SGA (age, 10.5 yr for girls and 12.5 yr for boys) were randomly assigned to receive either 0.067 mg/kg·d GH until attainment of AH or no treatment. In this per-protocol analysis, 91 of 102 patients in the treated group and 33 of 47 patients in the control group were followed to AH.
Mean height at inclusion was -3.2 SD score (SDS). Treatment duration was 2.7 ± 0.6 yr. AH was -2.7 ± 0.9 and -2.1 ± 1.0 SDS in the control and treated groups, respectively (P < 0.005). The groups differed by 0.6 SDS units (95% confidence interval, 0.2–0.9). Height gain was 0.5 ± 0.8 and 1.1 ± 0.9 SDS in the control and treated groups, respectively (P = 0.002). Multivariate analyses confirmed the independent effects of treatment (0.6 SDS) and treatment duration (0.4 SDS/yr). All potential biases would tend to decrease the estimate of the treatment effect. Treatment tolerance was excellent.
We concluded that the potential for spontaneous catch-up in short adolescents born SGA is limited. GH treatment increases AH by at least 0.6 SDS in this population.
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
In young prepubertal children with idiopathic short stature, a more heterogeneous patient population, one small controlled study (15) and several open trials (16, 17, 18) concluded that GH increases adult height by about 0.6 to 1.3 SD scores (SDS) after treatment for 3–6 yr. However, several other studies concluded that treatment resulted in little or no increase in height (reviewed in Ref. 19). All of these studies suffer from methodological limitations: the absence of a control group in most cases, limited patient numbers, and large numbers of patients lost to follow-up, resulting in analysis focusing on good responders (17, 19).
Although the need for earlier referrals has been stressed, children with short stature often seek medical advice around puberty. In general, children short at this age have a favorable adult height prognosis, because most display a constitutional delay in growth and puberty (20). However, those who are born short have a high risk of remaining short (2, 21). Studies investigating whether GH increases the adult height of early pubertal children born small for gestational age have reached no definitive conclusions, due to methodological limitations similar to those indicated above (21, 22).
We present here the results of the first long-term, controlled trial of GH treatment in children born small for gestational age, presenting short stature at around puberty.
|
Subjects and Methods |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
The principal criteria for inclusion in the study were: birth length less than -2 SDS for gestational age (23) and term greater than 30 wk; then at study inclusion, height -2.5 SDS for age or less (24), chronological age greater than 10.5 yr for girls and greater than 12.5 yr for boys, bone age at least 9 yr for girls and at least 10 yr for boys (25), and peak plasma GH concentration after pharmacological stimulation at least 10 µg/liter to exclude GH deficiency. Participants had to be at Tanner stage I or II, with a testicular volume less than 8 ml or uterus length less than 50 mm on ultrasound examination. Exclusion criteria included chromosomal abnormalities in girls (karyotyping), constitutional bone diseases (x-rays of the lumbar spine, pelvis, knee, and hand in all patients), any chronic disease interfering with growth, steroid or sex steroid treatment, and dysmorphic syndromes other than Russell-Silver syndrome (n = 4 in the study). Parents and most participants gave informed consent for participation in the study, which was approved by Comité Consultatif de Protection des Personnes dans la Recherche Biomédicale Paris-Cochin. The study was conducted at the French pediatric endocrinology and pediatrics centers listed in the Acknowledgments.
Objectives of the protocol and intervention
The aim was to investigate whether GH treatment during puberty significantly increased adult height compared with untreated controls. Participants were randomly assigned to two groups, one receiving daily GH injections [Maxomat, Sanofi-Synthelabo, Paris, France; 0.2 IU/kg·d (0.067 mg/kg·d)] and the other receiving no treatment until the attainment of adult height. The allocation sequence was generated by Méthodologie et Recherche Clinique, Cernay les Reims, France, and communicated by fax to the study participants. Group assignment was not masked, and the treated group was twice as large as the control group. The principal outcome criterion was adult height expressed in SDS, which was compared in the two groups. The secondary criterion was gain in SDS units between height at inclusion and adult height.
Data collected and criteria used to define adult height
At baseline (start of treatment) and follow-up visits (every 3 months for the treatment group and every 6 months for the control group), the following data were recorded: height, weight, chronological age, pubertal stage (26, 27), dose, and tolerance. Bone age was analyzed yearly (25). The per-protocol criteria for stopping treatment or follow-up were less than 1 cm growth over the last 6 months and a bone age of at least 15 yr for girls or at least 16 yr for boys. Because only 4% of patients met these criteria when treatment was stopped, we considered treatments to be almost complete, for analytical purposes, if growth velocity was 2 cm or less over the last 6 months or bone age was at least 13 yr for girls or at least 15 yr for boys. Patients who had discontinued regular follow-up before reaching adult height were contacted for a final visit in 2000 to record adult height. Adult height was considered attained if growth velocity was 2 cm/yr or less or if bone age was at least 15 yr for girls or at least 16 yr for boys (99 and 98%, respectively, of adult height; Ref. 28). If not available, bone age was estimated from previous measurements as described (29). Three patients with heights recorded in 2000 had not reached adult height, i.e. the tallest boy in the control group and two girls of the treated group with a growth potential of a few centimeters. They were maintained in the analysis without correction. Clinical tolerance was recorded at each visit and coded according to World Health Organization Adverse Reaction Terminology (30). We assessed the following indicators of tolerance in the treated group every 6 months: fasting plasma glucose, cholesterol, triglyceride, creatinine, and free T4 concentrations, blood count, erythrocyte sedimentation rate, and hemoglobin A1c. Plasma IGF-I concentrations were measured yearly in the treated group in a centralized laboratory, as described (31). Values were classified as low, normal, or high according to age- and sex-matched reference values.
Analysis of growth and statistical methods
SDS of height and weight for gestational age, age and sex, and target height were calculated (32, 33). Means and SD values are presented. We used the Mann-Whitney U test to compare groups. An 
risk of 5% was set as the significance threshold. We also performed multivariate analysis of height gain (in SDS units, baseline to adult; Ref. 33), although this was not specified in the initial protocol. We used the SAS statistical package (SAS Institute, Inc., Cary, NC) for calculations (34).
|
Results |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
The flow of participants is presented in Fig. 1
. Four patients in the treatment group were excluded from analysis due to severe diseases interfering with growth (sickle cell anemia, pulmonary hypertension, type 1 neurofibromatosis, and severe prematurity). Five patients assigned to the treatment group refused GH treatment but remained in the study and were analyzed as part of the control group. Fifteen patients (14 in the control group, 1 in the treated group) left the study early. Therefore, we included a total of 102 treated and 47 control patients in this per-protocol analysis. The initial characteristics of these patients are presented in Table 1. These children generally had short parents, severe growth retardation (height less than -4 SDS in 16 patients), a 2.0-yr retardation in bone age, and delayed puberty (only 31% of boys and 13% of girls having reached puberty at 13.7 and 11.8 yr, respectively). Several of the patients deviated slightly from the inclusion criteria. Four of the 149 patients had minor signs of skeletal dysplasia (1 control, 3 treated), 11 of 149 patients had birth lengths of at least -2 SDS (2 control, 9 treated; range, -1.8 to -2.0 SDS), 1 had an abnormal karyotype (ring chromosome 15), and 2 had a stimulated peak plasma GH concentration less than 10 ng/ml (8.0 and 7.4 ng/ml, respectively). Altogether, group reassignment or protocol deviations concerned 12 and 5 patients followed to adult height in the treated and control groups, respectively. No difference was found between the initial characteristics of the per-protocol and intention-to-treat populations, and the analysis was therefore performed on the per-protocol population (Fig. 1).
|
|
Description of outcome
The changes in height of treated and control patients are recorded in Table 2. Adult height was available for 70% of control patients and 89% of treated patients. At baseline in the treated group, patients followed to adult height were 0.4 SDS units taller than those lost to follow-up (P = 0.04). Other baseline characteristics were similar for patients followed and not followed to adult height. At the time of adult height measurement, control patients were a mean of 1 yr older than treated patients, bone age was 1 yr higher in boys of the control group than in boys of the treated group (17.9 ± 1.3 vs. 17.0 ± 1.1 yr; P < 0.01) but was similar in control and treated girls (15.9 ± 1.3 yr).
|
Model for adult height
Growth is a multifactorial process. We therefore designed a multivariate model to analyze outcome after adjustment for baseline variables that might affect growth (Table 3). The model incorporated all potential constitutive, baseline, and treatment variables (33), but the final model only retained two variables in addition to those representing regression toward the mean. A 1-yr variation of bone age retardation was associated with a 0.5 SDS unit greater gain in adult height. Treatment was associated with a 0.6 SDS unit (95% CI, 0.3–0.9) increase in adult height. The analysis was also performed on the intention-to-treat population, which gave identical results.
|
). In the treated group, adult heights and height gains were similar for patients with incomplete and almost complete treatment.
|
Adverse events recorded in the treated group included all intercurrent symptoms and illnesses. They were reported in 44% of treated patients, 10% having 4 or more events. The most frequently reported events involved the respiratory system (19%), osteomuscular system (14%), central nervous system (9%), and digestive tract (8%). All were mild, reversible, benign conditions, unlikely to be related to GH treatment. Sixteen adverse events, considered severe by the investigators, were recorded in 14 treated patients. Of those, trauma, psychiatric symptoms, abdominal symptoms, otitis, asthma, varicocele, striae, and migraine were unlikely to be due to GH treatment. Two adverse events are probably causally related to GH treatment: one patient had an episode of slipped capital epiphysis after 1.5 yr of treatment, and one had a single seizure episode 10 min after the first injection of GH.
Mildly high plasma glucose concentrations, probably nonfasting, were recorded in 8% of patients, but with no consistent pattern. Seven patients had hemoglobin A1c values above 6% at one point, but values always returned to normal. Before treatment, 89% of patients had plasma IGF-I values below the normal range for age and sex, probably reflecting the delay in puberty (Fig. 3). During GH treatment, plasma IGF-I levels increased but did not exceed the upper limit of normal in most cases. None of the other biological factors showed consistent variation with treatment.
|
|
Discussion |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Unlike previous studies, this study directly investigated whether GH increased adult height by comparing outcome after random assignment to treatment or observation groups. The mean spontaneous catch-up in the control group was 0.5 SDS units, similar to our previous observations (21) and considerably lower than that generally observed in patients with idiopathic short stature, irrespective of birth characteristics (35). Thus, decreased birth length is a marker for poor growth prognosis in short adolescents. Direct comparison showed that treatment had a positive effect on outcome, with 0.6 SDS units difference between the two groups, corresponding to +4 cm or +18% more growth in treated than control patients during the study. Treated children were closer to target height and more likely to have adult heights in the normal range than controls. Multivariate analysis, taking into account regression toward the mean and baseline characteristics, confirmed the independent effect of GH treatment. Tolerance was carefully monitored and was excellent.
This study has several limitations. Firstly, it was not blinded, because it was considered unethical to perform a placebo-controlled study in this clinical setting. Secondly, the per-protocol population described here differed from the intention-to-treat population generated by randomization because some patients refused their assignment to the control (25%) or treatment (5%) group. However, the patients who left the study early had similar baseline characteristics to those who remained. Thirdly, more control than treated patients were lost to follow-up (30% vs. 11%), but initial characteristics were similar in the two subgroups (Tables 1
and 2). These deviations from the initial protocol illustrate the difficulties in performing growth studies with untreated control groups. Fourthly, as in other studies, a surrogate variable for adult height was used, making it possible to complete the study within a reasonable time frame, avoiding excessive losses to follow-up. The criteria used are compatible with a maximum further gain of less than 3 cm, but actual bone age values were a mean of 1–2 yr higher than the threshold. The controls of both sexes were older than treated patients at attainment of adult height, and boys of the control group had a higher bone age. Therefore, treated individuals had a greater potential for further growth than controls. Fifthly, treatment duration was variable, ranging from 6 months to 3.2 yr, and few patients completed treatment as planned. Treatment duration was associated with adult height in univariate and multivariate analysis, but we analyzed all treated patients, irrespective of treatment duration or completion. Overall, the potential biases identified would all decrease the magnitude of the treatment effect. Therefore, our results not only demonstrate the growth-promoting effect of GH in this population but also provide minimal estimates of its magnitude.
This study differs noticeably from other studies on the effect of GH in patients with idiopathic short stature or born small for gestational age (Table 4). The endpoint was adult height rather than change in height during a course of treatment of fixed duration (11, 14, 36). As in the study by McCaughey et al. (15), we used a long-term control group rather than comparison with historical controls (16, 18, 21) or predicted heights, which are unreliable (17). We were also able to follow adult height in a larger proportion (83%) of patients prospectively included in a trial than was the case in most other studies (Table 4). Retrospective studies or studies in which a substantial proportion of patients are lost to follow-up are likely to overestimate treatment effects because they concentrate on good responders (19, 33). We selected children born small for gestational age rather than with idiopathic short stature and excluded GH deficiency in contrast to others (14). We also selected prepubertal or early pubertal adolescents rather than younger children. Finally, the dose of GH used was about 50% higher than in most other studies. Overall, the other studies presented in Table 4 may have overestimated treatment effect (19), whereas we may have underestimated it.
|
Several aspects of our study should be investigated further. Although the magnitude of the effect might be higher than our estimation, its clinical significance is unclear. Short stature is associated with dissatisfaction with height at around 12 yr of age (37) but does not affect self-esteem (38) or self-perception (37). These issues were not investigated in the study. In addition, at the current market price of GH, the average cost of such a treatment would be 98,000 euros, raising the issue of its cost effectiveness. The reason why bone age delay in patients born small for gestational age results in limited catch-up also deserves further study. Finally, additional drug treatments with GnRH agonists (39) or aromatase inhibitors (40) have been suggested as a means of increasing the magnitude of the effect of GH in similar patient populations but should be evaluated through methodologically satisfactory controlled trials.
|
Acknowledgments |
|---|
We also thank Drs. Souad Nafa, Isabelle Fermé, and Jean-Pascal Ducret, and Ms. Annie Serrière (Sanofi-Synthélabo), Mrs. Jean-Christophe Charpentier, and Renaud Fay (Méthodologie et Recherche Clinique).
|
Footnotes |
|---|
Abbreviations: AH, Adult height; CI, confidence interval; SDS, SD score; SGA, small for gestational age.
Received July 18, 2002.
Accepted January 8, 2003.
|
References |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  A. Carrascosa, L. Audi, C. Esteban, M. Fernandez-Cancio, P. Andaluz, M. Gussinye, M. Clemente, D. Yeste, and M. A. Albisu Growth Hormone (GH) Dose, But Not Exon 3-Deleted/Full-Length GH Receptor Polymorphism Genotypes, Influences Growth Response to Two-Year GH Therapy in Short Small-for-Gestational-Age Children J. Clin. Endocrinol. Metab., January 1, 2008; 93(1): 147 - 153. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. M. N. Bannink, J. van Doorn, P. G. H. Mulder, and A. C. S. Hokken-Koelega Free/Dissociable Insulin-Like Growth Factor (IGF)-I, Not Total IGF-I, Correlates with Growth Response during Growth Hormone Treatment in Children Born Small for Gestational Age J. Clin. Endocrinol. Metab., August 1, 2007; 92(8): 2992 - 3000. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Argente, R. Gracia, L. Ibanez, A. Oliver, E. Borrajo, A. Vela, J. P. Lopez-Siguero, M. L. Moreno, F. Rodriguez-Hierro, and on behalf of the Spanish SGA Working Group Improvement in Growth after Two Years of Growth Hormone Therapy in Very Young Children Born Small for Gestational Age and without Spontaneous Catch-Up Growth: Results of a Multicenter, Controlled, Randomized, Open Clinical Trial J. Clin. Endocrinol. Metab., August 1, 2007; 92(8): 3095 - 3101. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Saenger, P. Czernichow, I. Hughes, and E. O. Reiter Small for Gestational Age: Short Stature and Beyond Endocr. Rev., April 1, 2007; 28(2): 219 - 251. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Carrascosa, C. Esteban, R. Espadero, M. Fernandez-Cancio, P. Andaluz, M. Clemente, L. Audi, H. Wollmann, L. Fryklund, L. Parodi, et al. The d3/fl-Growth Hormone (GH) Receptor Polymorphism Does Not Influence the Effect of GH Treatment (66 {micro}g/kg per Day) or the Spontaneous Growth in Short Non-GH-Deficient Small-for-Gestational-Age Children: Results from a Two-Year Controlled Prospective Study in 170 Spanish Patients J. Clin. Endocrinol. Metab., September 1, 2006; 91(9): 3281 - 3286. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. B. Allen Growth Hormone Therapy for Short Stature: Is the Benefit Worth the Burden? Pediatrics, July 1, 2006; 118(1): 343 - 348. [Full Text] [PDF]
|
|
|
|
 |
|
 S. K. Varma Long-Term Growth Hormone Treatment of Idiopathic Short Stature AAP Grand Rounds, January 1, 2006; 15(1): 9 - 10. [Full Text] [PDF]
|
|
|
|
 |
|
 M. Rosilio, J.-C. Carel, E. Ecosse, J.-L. Chaussainon, and on behalf of the 0908 Lilly Study Group Adult height of prepubertal short children born small for gestational age treated with GH Eur. J. Endocrinol., June 1, 2005; 152(6): 835 - 843. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J.-C. Carel Growth Hormone in Turner Syndrome: Twenty Years after, What Can We Tell our Patients? J. Clin. Endocrinol. Metab., June 1, 2005; 90(6): 3793 - 3794. [Full Text] [PDF]
|
|
|
|
 |
|
 N. Bridges New indications for growth hormone Arch. Dis. Child. Ed. Pract., June 1, 2005; 90(1): ep7 - ep9. [Full Text] [PDF]
|
|
|
|
|
|
J.-C. Carel, E. Ecosse, I. Bastie-Sigeac, S. Cabrol, M. Tauber, J. Leger, M. Nicolino, R. Brauner, J.-L. Chaussain, and J. Coste Quality of Life Determinants in Young Women with Turner's Syndrome after Growth Hormone Treatment: Results of the StaTur Population-Based Cohort Study J. Clin. Endocrinol. Metab., April 1, 2005; 90(4): 1992 - 1997. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. Brodsky and H. Christou Current Concepts in Intrauterine Growth Restriction J Intensive Care Med, November 1, 2004; 19(6): 307 - 319. [Abstract] [PDF]
|
|
|
|
 |
|
 L B Johnston and M O Savage Should recombinant human growth hormone therapy be used in short small for gestational age children? Arch. Dis. Child., August 1, 2004; 89(8): 740 - 744. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. R. Rose Growth Hormone Therapy Benefits Former SGA Infants AAP Grand Rounds, August 1, 2003; 10(2): 14 - 14. [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
) and girls (
); the horizontal dashed line represents -2 SDS; the univariate correlation between treatment duration and adult height is also presented (sloping dashed line).
) and girls (•). B, Disposition of values classified according to sex- and age-matched normative data as low (light gray), normal (dark gray), or high (black).