| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
CLINICAL STUDY |
Lilly Research Laboratories, Eli Lilly and Company, Suresnes, France and 1 Groupe Hospitalier Cochin-Saint Vincent de Paul, Université Paris V, 75014 Paris, France
(Correspondence should be addressed to M Rosilio, Lilly France 13 rue Pagès, 92158 Suresnes, Cedex, France; Email: rosilio_myriam{at}lilly.com)
 |
Abstract |
|---|
 Top Abstract IntroductionPatients and methodsResultsDiscussionReferences |
|---|
Design: Patients in this open-label, Phase III, multicenter study received a daily hGH dose of 0.067 mg/kg for 2 years, and then received no treatment for the following 2 years. After the fourth year on study, patients whose height had decreased more than 0.5 SDS but who still showed growth potential based on bone age were allowed to resume treatment until they reached adult height.
Methods: Height gain SDS was assessed for 11 girls and 24 boys (mean age±S.D. 9.6±0.9 years) at the end of the 2 years of hGH treatment, during the subsequent 2-year off-treatment period, and upon reaching adult height.
Results: At the end of the initial 2-year treatment period, 83% of patients had reached a height within the normal range, with a mean increase in height SDS vs baseline of 1.3±0.3 (P <0.001). Adult heights (n = 20) were within the normal range for 50% of patients, and mean height gain from baseline was statistically significant (0.7±0.8 SDS, P <0.001). Fasting glucose and glycosylated hemoglobin levels were not significantly modified during treatment.
Conclusions: High-dose hGH treatment for a minimum of 2 years in short children born SGA was well tolerated and resulted in a significant increase in adolescent and adult height.
|
Introduction |
|---|
|
TopAbstractIntroductionPatients and methodsResultsDiscussionReferences |
|---|
Clinical trials aimed at improving the growth of short children born SGA using human growth hormone (hGH) have been conducted. Short- to mid-term studies have demonstrated that catch-up growth can be induced by hGH administration, with height gains ranging from 1.1 to 2.6 SDS depending on the age of the children at treatment initiation, the hGH dosage used, and the duration of hGH therapy (7–13). However, the pattern of hGH administration is still a matter of debate, as some authors favor continuous treatment (10) while others favor discontinuous treatment (9). Long-term results on adult height are scarce and discordant, with height gains ranging from 0.1 to 1.8 SDS in this patient group (14–18).
In the exploratory study reported here, we hypothesized that hGH treatment in short children born SGA, begun near the start of puberty and continued for 2 years, would lead to a sufficient height gain at pubertal entry, and that the pubertal growth spurt would then enable these patients to maintain the treatment benefit and increase adult height.
|
Patients and methods |
|---|
|
TopAbstractIntroductionPatients and methodsResultsDiscussionReferences |
|---|
Study inclusion/exclusion criteria
The following inclusion criteria applied: chronological age (CA) 7 to <9 years for girls and 9 to <11 years for boys; height below –2 S.D. (19); bone age (BA) less than or equal to CA (20); birth length below –2 S.D. for gestational age (21); duration of pregnancy >36 weeks; parental height > or = 160 cm (–2.5 S.D.) for the father and > or = 147 cm (–2.9 S.D.) for the mother (19); growth documented by two or three measurements during the 2 years prior to the study; and a normal response to a GH stimulation test. Patients were not included if they had previously received GH treatment or if they had any chronic disease or malformation.
Treatment protocol
hGH was administered to all patients s.c. at a dose of 0.067 mg/kg per day (equivalent to 0.2 IU/kg per day) for 2 years, and treatment was then discontinued for the next 2 years; the period between initiation of hGH therapy and the end of the off-treatment period will henceforth be referred to as ‘Period I’. After the fourth year on study, patients who lost more than 0.5 SDS in height during the 2 year off-treatment period and who still presented growth potential (BA <12 years for girls or <14 years for boys) were allowed to resume hGH treatment at the same dose; the period between the point at which hGH treatment could be resumed and the last height measurement will henceforth be referred to as ‘Period II’. Patients were followed, whether they had resumed treatment or not, until they reached adult height, defined as a height velocity 
1 cm/year using two subsequent measurements taken at least 6 months apart (19), and/or BA 
16 years for girls or 18 years for boys (22). If BA was not assessed at the last height measurement, the time from the last BA to the last height measurement was added to the last BA to derive an estimated BA (23, 24). Adult height SDS was calculated either based on normative values for a CA of 18 years (referred to as SDS/age 18), regardless of the patient’s actual CA at the time of adult height measurement, or as SDS/CA when actual CA was taken into consideration.
Data collected
BA reading was performed centrally according to the method of Greulich & Pyle (20). Predicted height was calculated according to the method of Bayley & Pinneau (22). Pubertal development was evaluated according to the Tanner classification, and pubertal onset was defined as stable Tanner stage B2 for girls or Tanner stage G2 for boys (25). Insulin-like growth factor-I (IGF-I) was assayed centrally after acid/ethanol extraction at the Fondation de Recherche en Hormonologie (Fresnes, France). Mean normal IGF-I values for boys were 0.61 IU/ml for ages 8–10 years and 0.88 IU/ml for ages 10–14 years. Mean normal IGF-I values for girls were 0.69 IU/ml for ages 8–10 years and 1.28 IU/ml for ages 10–14 years. Fasting glucose and glycosylated hemoglobin (HbA1c) determinations were performed locally at each study center. Normal values ranged from 3.3 to 6.1 mmol/l for fasting glucose and 3.9 to 6.5% for HbA1c.
Patient disposition (Fig. 1
)
During study Period I, 35 patients (11 girls and 24 boys) recruited from seven centers in France were treated with hGH for 2 years. Of these patients, five presented with protocol violations: one patient did not meet the birth length criterion of <–2 S.D. for gestational age (birth length = –1.44 SDS); two patients did not meet the parental height criterion (one father’s height was unknown, and one mother’s height was equal to 146 cm); one patient presented with cardiac disease; and one patient had constitutional bone disease that was diagnosed later during the study. However, as we performed an intent-to-treat analysis, all those patients were kept in the efficacy and safety analyses.
|
for simplicity. Data analysis
TH was calculated according to the method proposed by Ranke to account for assortative mating (26). Parental adjusted height SDS was calculated as the difference between actual height SDS and TH SDS.
Statistical evaluations were performed using the SAS 8.1 statistical package. Continuous data are summarized by number of values, mean, S.D., minimum and maximum. Categorical data are presented as numbers and percentages. For normally distributed variables, comparisons were made using a paired t-test. Otherwise, comparisons were made using the signed-rank statistic.
Three linear multiple regression models were developed to explain height gain in SDS at three endpoints, using demographic and auxological information as independent variables. Before each multivariate analysis, bivariate analyses, in which Pearson’s correlation coefficient r, r 2, and significance of relationship (P) were calculated to select potential variables for the models, were performed.
The first model was developed using a step-wise backward regression with an 
risk of 0.25. This value was used instead of the commonly used 0.05 due to the low number of observations (n = 35).
Inspection of the distribution of Cook’s statistic (27) was performed to detect each observation’s influence on the estimated parameter. In cases of outlier observations distorting the model, these observations were excluded and a new model was evaluated.
|
Results |
|---|
|
TopAbstractIntroductionPatients and methodsResultsDiscussionReferences |
|---|
Height changes during Period I
Patients’ characteristics at birth and baseline are summarized in Table 1. During Period I, all patients received daily hGH therapy for 2.3±0.1 years (range: 2.1–2.6 years).
|
), which was very close to their mean TH (–1.5±0.9 SDS), and 83% of patients had reached a height within the normal range for CA by the end of the 2 year treatment period. Preliminary analyses of these data have been previously published (28).
|
All children were prepubertal at baseline, and most of them entered puberty during Period I. By the end of the second year of the study, 46% of boys and girls had entered puberty, and this proportion increased to 93% of boys and girls by the end of the fourth year (78% of the girls and 100% of the boys). Mean age at pubertal onset for the patients followed until the end of period I was 11.9±0.9 years for boys (n = 21; range: 10.0–13.3 years) and 10.5±0.8 years for girls (n = 8; range: 8.6–11.1 years). Mean BA maturation progressed 2.3 years during the 2 years of treatment and 2.5 years during the subsequent 2 year off-treatment period.
IGF-I concentrations increased during hGH treatment from a baseline value of 0.8±0.3 IU/ml to 1.7±0.6 IU/ml, and then to 1.5±0.4 IU/ml, at the end of the first and second years of therapy respectively. These values remained stable at 1.4±0.6 U/ml during the third year of the study, despite the fact that patients were untreated.
Model for height gain at 2 years (Model 1)
The model used to explain outcome after 2 years of hGH treatment is shown in Table 2 (r 2 = 0.6). Lesser baseline BA retardation, greater baseline height velocity, and lower baseline IGF-I levels were associated with an increased height gain at 2 years. Other variables, including gender, gestational age, birth length, birth weight, weight, age at pubertal onset, age at baseline, TH, mother’s height, father’s height, and parental adjusted height SDS at baseline, were not associated with outcome in this model.
|
(r 2 = 0.75). Baseline weight and male gender were positively associated with height SDS changes, while baseline BA retardation and height SDS gain at 2 years were inversely correlated with height SDS changes between years 2 and 4. As pubertal age is co-linear to gender, it was not included in Model 2. Other variables not independently associated with height SDS changes during this period included TH, parental adjusted height SDS, height SDS, height velocity, mother’s height, father’s height, and IGF-I at baseline.
|
Height changes for patients entering Period II Twenty-four patients entered study Period II (starting at year 5 of the study): seven resumed hGH therapy (three girls and four boys), while 17 did not (four girls and 13 boys). For patients who resumed hGH therapy, mean height SDS remained stable until last height measurement (–1.8±0.3 SDS at year 5 vs –1.7±1.0 SDS at last height measurement), which occurred 7.1±1.0 years after baseline measurement and after an additional treatment duration of 1.6–0.8 years (range: 0.5–2.4 years). For patients who did not resume hGH therapy, height SDS decreased from –1.8±0.7 at year 5 to –2.2±0.7 at last height measurement, which occurred 7.4±0.9 years after baseline measurement.
Adult height
Of 20 patients who reached adult height, four had resumed hGH therapy during Period II. For the adult height population, heights at the end of Period I were –1.8±0.6 and –1.5±0.6 SDS/CA for patients who resumed hGH therapy vs those who did not respectively. No difference in adult height was observed between the two groups (–2.1±1.2 and –2.1±0.7 SDS/age 18 for patients who resumed hGH therapy vs those who did not respectively). Adult heights were 0.7±1.2 SDS below the respective THs for both groups. Both groups have been analyzed together thereafter for adult height and their characteristics and outcomes are summarized by gender in Table 4. The overall height gains from baseline to adult height were 0.7±0.8 SDS/CA and 0.5±0.8 SDS/age 18 (P = 0.01; 95% CI: 0.1 to 0.9). Height gain tended to be greater in girls than in boys (0.7±1.1 vs 0.4±0.7 SDS/age 18, not significant). Height gain during puberty was 25.2±4.1 cm for the 12 boys and 21.3±3.9 cm for the eight girls.
|
). When compared with predicted height at baseline (Fig. 3b), adult height improved significantly in girls (by +1.2±1.1 SDS/age 18 (P = 0.018)) but not in boys (+0.2±0.9 SDS/age 18 (P = NS)). Altogether, 50% of these children reached an adult height within the normal range.
|
), only three independent variables – BA at baseline, age at pubertal onset (S.D.), and baseline IGF-I – were found to be positively associated with adult height gain (r 2 = 0.76). Other variables not significantly associated with adult height gain were parental adjusted height SDS, BA retardation, height velocity at baseline, weight at baseline, TH, father’s height and mother’s height.
|
Treatment-emergent adverse events and serious adverse events were reported for 30 and 13 patients respectively. One death from suicide occurred approximately 4 years after the patient had completed hGH treatment and was not considered by the investigator to be related to the study drug. The remaining adverse events reported during this study were typical for children and were not considered to be related to hGH treatment. Additionally, there were no relevant changes in clinical laboratory parameters that were considered to be related to hGH treatment. In particular, neither fasting blood glucose levels (baseline: 4.8±0.6 mmol/l; at year 2: 4.6±0.5 mmol/l; P = 0.132/n = 31) nor HbA1c values (baseline: 4.9±0.5%; at year 2: 5.1±0.6%; at year 4: 5.1±0.3%; P = 0.078/n = 33) were altered by hGH treatment. None of the children developed diabetes during the observation period (median = 5.2 years, range: 1.9–9.4 years).
|
Discussion |
|---|
|
TopAbstractIntroductionPatients and methodsResultsDiscussionReferences |
|---|
The present study is particularly valuable since it reports adult heights in 20 out of 35 children included in the trial. The gains in adult height compared with baseline were 0.7±0.9 SDS/CA and 0.5±0.9 SDS/age 18 for the adult height population. This modest result, which was achieved after a mean treatment duration of less than 3 years and after discontinuation of treatment several years before adult height measurement for 80% of the cohort, has questionable clinical significance and raises the question of the role of hGH treatment vs spontaneous catch-up. However, the design of the study with discontinuous therapy aimed at testing the effect of a relatively limited exposure to hGH, with attention to comfort for the patient, safety and overall costs of treatment. Therefore, our results have to be interpreted in the context of this framework if compared with other studies summarized in Table 6.
|
The absence of an untreated control group precludes definite conclusions regarding the impact of the limited hGH treatment period in this study on adult height, since the relatively small gain in height SDS at adult height may have been attributable to spontaneous catch-up growth.
When the trial was designed, a randomized control group was not considered feasible by investigators and ethics committees. Therefore, it was originally planned to compare the patients in this study with a historical control group of patients selected in the mid-1980s. However, when the analysis was performed, major baseline differences, such as older age, lower birth length, or onset of puberty before the beginning of the observation (25% of patients in the historical control group), became obvious, and therefore, a comparison was deemed not to be appropriate. These patients have been described by Coutant et al. (18) and are referenced in Table 6.
Contradictory results regarding adult height in untreated patients – ranging from no catch-up growth (14, 16) to a mean spontaneous height gain of 0.5 SDS (15) or 0.6 SDS (18) – have been published (Table 6). One reason for these apparent discrepancies might be the variable age at inclusion of untreated patients. In the present study, children were included in the study at a mean CA of 9.7 years and were therefore closer to the onset of puberty than children in other studies (14, 17).
Sixty percent of the initial catch-up growth attained during hGH therapy was lost after treatment discontinuation, while the children in this study were followed up to adult height (mean loss of 0.5 SDS vs an initial height gain of 1.3 SDS). In comparison, patients who started hGH treatment in early puberty and who received a similar dose administered continuously for a mean of 2.7 years have shown greater height gains (15). The comparison of these two studies (15) and the current study) suggests that if hGH treatment is begun close to or during early puberty, it must be continued until adult height to ensure greater adult height gains.
Based on the multivariate analysis, the variability in height gain after 2 years of hGH treatment was best explained by BA retardation, height velocity and IGF-I levels at baseline. Children who grew faster during the first 2 years of the study had less pronounced BA retardation and greater height velocity at baseline. Additionally, they were likely to enter puberty during the first 2 years of the study (almost 50% of children were Tanner stage 2 at the end of year 2), as expected considering the mean age of the children at study entry. Patients with lower baseline IGF-I levels attained greater adult heights. As GH deficiency (GHD) was excluded before study entry, this finding suggests that for a given auxological (height SDS/CA) and maturational status (BA and pubertal stage), low IGF-I levels at baseline indicate a greater potential to respond to exogenous GH. This finding is similar to that reported for patients with GHD (29).
Mean age at onset of puberty was normal for the patients in the study, consistent with what has been observed in other studies with hGH treatment (9, 10, 12, 14, 30), despite the fact that untreated SGA children have been reported in the past to have an earlier onset of puberty than normal children (19, 31, 32). The multivariate analysis in this study revealed that children who were older at the onset of puberty had a greater height gain at adult height. This finding, which goes against what is known about growth, is probably due to the fact that most of the children entered puberty during the first 4 years of the study. In comparison, van Pareren et al. (14) did not identify age at onset of puberty as an important parameter in their prediction model for adult height.
No disturbance in glucose metabolism was observed, during either hGH treatment or the follow-up period. This is reassuring, although the size of the sample studied and the duration of follow-up cannot exclude long-term effects on glucose metabolism. Most adverse events were characteristic for a population composed of children, and were unrelated to hGH treatment.
In conclusion, the short-term beneficial effects of high-dose hGH treatment observed in the present study were not sustained at adult height in most patients, probably due to the discontinuous pattern of hGH administration. Our present results are complementary to other evidence accumulated on hGH treatment in similar patient populations. In particular, they point to the need for a better evaluation of the spontaneous growth potential and long-term response to treatment in order to individualize treatment decision and protocol, aiming at further improvement of the overall outcome.
|
Acknowledgements |
|---|
The authors wish to thank all contributing investigators of the 0908 Lilly Study Group from the seven centers in France (P Chatelain and P Berlier, Lyon; M Colle, Bordeaux; C Jeandel and B Leheup, Nancy; Y Lebouc and S Cabrol, Trousseau Hospital, Paris; R Rappaport, Necker Hospital, Paris; M Tauber and C Pienkowski, Toulouse; and J-E Toublanc, St Vincent de Paul Hospital, Paris), as well as the Clinical Research staff of Lilly France (Nathalie Martin). We also wish to thank Gordon B Cutler Jr, Werner F Blum and Heike Jung for detailed critical review of the manuscript, and Anne M Wolka for editorial assistance.
|
References |
|---|
|
TopAbstractIntroductionPatients and methodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  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]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
) and with (
) GH treatment. Significant level of change in height SDS vs baseline at 2 and 4 years; P <0.001.
, adult height, predicted height and target height expressed as SDS/age 18.