| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
University Childrens Hospital (S.B., R.D.P., W.B., H.P.S.), Division of Endocrinology and Diabetology, D-80337 Munich, Germany; and Childrens Hospital for Rheumatology (P.R., R.H., H.M.), D-82467 Garmisch-Partenkirchen, Germany
Address all correspondence and requests for reprints to: S. Bechtold, M.D., Kinderklinik und Kinderpoliklinik, Im Dr. von Haunerschen Kinderspital, Pediatric Endocrinology, Lindwurmstrasse 4, D-80337 München, Germany. E-mail: Susanne.Bechtold{at}med.uni-muenchen.de.
| Abstract |
|---|
|
|
|---|
Methods: Thirty-one growth-retarded children with systemic and polyarticular idiopathic arthritis were enrolled in this controlled study. After a mean observational time of 8.4 yr, final height was reached in 13 patients (seven females and six males) treated with GH for a mean of 6.7 yr in a dose of 0.33 mg/kg body weight per week. Eighteen patients (12 females and six males) served as an untreated control group.
Results: Mean increment in height in the treatment group was 1.6 ± 0.8 SD, whereas the patients of the control group lost 0.7 ± 1.8 SD. Overall, mean final height in the treatment group was –1.6 SD and in the control group –3.4 SD. More GH-treated patients reached a final height within target height than untreated patients (11 of 13 vs. four of 18). Disease activity markers had a significant influence on height outcome. After adjustment for baseline and average disease activity, the difference between treatment and control group was still significant (mean 1.5 SD). Patients with a moderate overall disease activity profited most from GH treatment. No adverse events were noted throughout the study.
Conclusion: Our data suggest that long-term GH therapy has a beneficial effect on growth and final height in the majority of growth retarded children with severe forms of JIA.
| Introduction |
|---|
|
|
|---|
The aim of our randomized controlled study was to evaluate for the first time the efficacy and safety of GH up to final height in severely growth-retarded children with JIA. We compared growth rates and standardized height at final height between 13 GH-treated children with JIA and 18 controls, who did not receive GH therapy.
| Subjects and Methods |
|---|
|
|
|---|
Fifty prepubertal children with severe systemic or nonsystemic polyarticular JIA and growth retardation were enrolled within 10 months, fulfilling the International League of Associations of Rheumatologists classification criteria (14). All children had a SD score (SDS) for height of –2.0 or below and/or a height velocity below the 25th percentile during the year before inclusion. At the start of the study, all patients were prepubertal. They had been treated with GC on a relatively stable dose for more than 6 months before inclusion and continued with GC treatment during the study. Exclusion criteria were endocrinopathy or other metabolic or congenital disorders, renal failure, nephrotic syndrome, diabetes, heart failure, and previous treatment with GH.
All fifty children were tested for GH deficiency by pharmacological stimulation (clonidine or arginine). Nine patients were thus considered to be GH deficient and were excluded. The remaining 41 patients were randomly allocated either to the treatment (n = 19) or to the control group (n = 22).
Ten of 41 patients had to be excluded from the study (24% dropout rate). In the treatment group, four patients had compliance problems or interrupted GH treatment and two children relocated. In the control group, one boy relocated, one girl died after 2 yr, and two patients started GH therapy.
All patients had received daily GC for a mean duration of 3.9 ± 0.9 yr at the start of the study. Several types of GC were used and converted to prednisolone equivalent doses (mg/kg body weight per day). Regular drug therapy was modified as required by the disease state, including nonsteroidal antiinflammatory drugs, slow acting antiinflammatory agents, methotrexate, or cyclosporine A, and some were on etanercept for a short period of several months to a few years.
In all children, anthropometric measurements were obtained at 3- to 6-month intervals. Height was measured in a standing position, using a digital telescopic wall-mounted stadiometer (Ulmer Stadiometer; Prof. E. Heinze, Ulm, Germany). Weight was determined to the nearest of 0.1 kg using an electronic scale (Seca 753 E). Reference data were taken from the Zurich Longitudinal Growth Study (15). The genetic target height was calculated as the midparental height (the mean of the two parents heights) plus 6.5 cm for boys and minus 6.5 cm for girls (16). Radiographs of the hand were obtained at approximately 12-month intervals. Bone age was determined by the Greulich and Pyle method (17). In each patient, joint involvement was assessed clinically by the same experienced rheumatologist. Prepubertal and pubertal stages were graded with Tanner scores for breast development in girls and genital status in boys by the same experienced endocrinologist (18). Final height was defined by a height velocity of less than 1 cm/yr and by evidence of epiphyseal closure on radiography of the hand. The study protocol was approved by the local ethics committee. Written informed consent was obtained from the parents and oral or written consent from the children.
Treatment regimens
Patients of the treatment group received daily injection of GH (Genotropin; Pfizer, Karlsruhe, Germany) in a dose of 0.33 mg/kg body weight per week; patients of the control group were not treated.
Laboratory assessment
Every 3–6 months, blood samples were taken for measurement of routine blood cell count, erythrocyte sedimentation rate (ESR), C-reactive protein (CRP), fasting glucose, glycosylated hemoglobin (HbA1c), cholesterol, calcium, phosphate, and alkaline phosphatase. IGF-I, IGFBP-3, and thyroid hormones were measured at least every 6 months. Serum GH was measured using an immunoenzymetric assay (HGH EASIA; Biosource, Fleurus, Belgium), IGF-I was measured using an immunoenzymetric assay (OCTEIA IGF-I; IDS, Boldon, UK), and IGFBP-3 was measured using a RIA (Nichols Institute Diagnostics, San Juan Capistrano, CA).
Statistical assessment
Results in patients were converted into sex- and age-specific SDS. Growth rates are presented as annualized growth velocities (centimeters per year). Data for height, growth velocity, body mass index, and target height were expressed as mean ± SDS or mean and range. We used the nonparametric Mann Whitney U test to evaluate the comparisons of the mean. The nominal significance level was defined as 0.05 (
). Analysis of covariance was employed to adjust final height SD for prognostic factors. Prognostic factors were chosen a priori on clinical grounds as starting height SD, target height SD, and clinical parameters of disease activity (ESR, CRP, platelet count, hemoglobin, and daily prednisolone equivalent dose). Additionally, an exploratory analysis of covariance was calculated, adjusting additionally for mean levels of the aforementioned clinical parameters during the study. No stepwise procedures were implemented. Pearsons product-moment correlations were used to determine associations between variables. To obtain a synchronized growth curve, the individual smoothed growth curves were synchronized according to start of GH treatment or time of first observation in controls, minimal prespurt height velocity, pubertal peak height velocity, and the time of the end of the pubertal growth spurt. For this purpose, a synchronization program was applied that transfers the time scale of each individual curves to align the characteristic points with their respective means (Gompert Modeling) (19). Statistical analysis was performed using the SPSS 14.1 (SPSS, Chicago, IL) and the SAS (version 8.2; SAS Institute GmbH, Heidelberg, Germany) statistical software packages.
| Results |
|---|
|
|
|---|
The analysis is restricted to 31 children (12 males and 19 females), randomly allocated to the treatment or control group. Mean age at onset of the disease was 3.7 ± 1.2 yr. At the study start, there was no statistical difference between the treatment and control group regarding classification of JIA, age at onset of JIA, and duration of illness. There were more females in the control group. Anthropometric characteristics of treatment and control patients with JIA are given in Table 1
.
|
GH induced sustained catch-up growth in the treatment group, whereas progressive growth failure was observed in the control group (Figs. 1
and 2
). Therefore, net height SD increase in the treatment group in comparison with the control group was 2.3. In GH-treated patients, 11 of 13 reached a final height within target height range, whereas in the control group, only four of 18. Height loss throughout the observation period occurred in 15 patients of the control group vs. one patient of the treatment group.
|
|
Clinical, hormonal, and biochemical assessment
At the start of the study, there was no statistical difference between groups in overall joint status, the number of relapses, or joint destruction, and the development of mild scoliosis and lumbar lordosis due to hip destruction was similar in treated and control patients. There were five patients (three males and two females) in each group with severe joint destruction and consecutive development of lumbar lordosis and mild scoliosis of the spine.
Pretrial mean IGF-I SD and IGFBP-3 SD levels were low. Average IGF-I SD and IGFBP-3 SD were significantly higher with GH treatment but significantly reduced in all compared with reference values. Fasting glucose, HbA1c, cholesterol, electrolytes, and TSH were unchanged throughout the study in all patients. Laboratory variables of disease activity like ESR, CRP, platelet count, and hemoglobin were not significantly different at the study start but showed great variability throughout the study. The overall inflammatory activity during the observational time was significantly higher in the control group. Cumulative, lifelong prednisone equivalent dose was not statistically different between the treatment and control group (Table 2
).
|
In the treatment and control group, GC treatment could be reduced over time (data not shown). Because the prednisolone dose grossly reflects disease activity, there was a negative correlation between lifelong prednisolone equivalent dose and final height in patients of the treatment and control group (r = –0.57; r = –0.61; P < 0.05, respectively).
Patients of the control group had higher disease activity parameters throughout the observation period. Adjusting final height SD data for parameters of initial and average disease activity, height SD at start, and target height SD, the difference between the groups for final height SD was between 1.41–1.69 SD (Table 3
).
|
Multiple regression analysis revealed that final height in the treated and control patients was significantly associated with height at start of puberty (r = 0.83), average levels of IGF-I (r = 0.61), IGFBP-3 (r = 0.61), and average CRP (r = –0.61) and ESR (r =–0.69), with a cumulative R2 = 0.72.
| Discussion |
|---|
|
|
|---|
Following a control group is essential in a chronic disease with an unpredictable course. Because it was not clear at the study start whether GH treatment would have a long-term beneficial effect on height and whether it might have a negative impact on disease activity, we decided to follow the control group until final height. The parameters of disease activity increased or were persistently higher in the control patient throughout the study. It may be questioned whether similarly progressive growth failure would have developed in the control group with a disease activity comparable to the treatment group. Therefore, we adjusted final height data for baseline and average disease activity with and without GH treatment and could show a persistent difference between groups between 1.41–1.69 SD for final height SD. Moreover, the decrease in height SDS in the control group during follow-up was quantitatively similar to that found in several previous studies of spontaneous growth in children with JIA (4, 5).
The impressive increment in height SDS in response to GH therapy was maintained during prepubertal and pubertal years. Duration of puberty was similar in the study and control group. There was no indication for an earlier onset or a shorter duration of pubertal growth. Bone age was retarded in both groups. However, the predictive value of bone age in these children is compromised. The height increment during pubertal years of 1.04 SD in GH-treated patients and 0.12 SD in control patients was probably also due to a relative prepubertal loss in height SD as a result of delayed onset of puberty. In the treatment group, the total height gain during the observation period was not associated with the initial degree of growth retardation (23, 24), suggesting that efficacy of GH depends also on factors like disease activity markers allowing catch-up growth or not (12). Because there are no prognostic factors for the overall course of disease activity, the lower disease activity in the treatment patients could also be attributed to GH treatment and a positive effect on disease activity.
High disease activity and therefore high levels of inflammatory cytokines such as IL-6, TNF
, and IL-1ß are associated with low levels of IGF-I and IGFBP-3 in mice and human models (25, 26, 27, 28). Initial and average IGF-I and IGFBP-3 levels correlated significantly with disease activity parameters in treated and control patients but were significantly higher with GH administration. This finding is compatible with the concept that GH resistance in patients with severe forms of JIA is only partially overcome by exogenous GH. Thus, together with GH treatment, the necessity to minimize the disease activity persists (12).
No adverse events occurred throughout the study. Disease activity decreased in the treated group, excluding an adverse effect of GH on the underlying disease. There was no difference between groups in HbA1c or fasting glucose levels. Monitoring for potential metabolic side effects, especially in combination with GC treatment, must be performed carefully. Skeletal complications such as the development of scoliosis or lumbar lordosis was not increased in the treatment group compared with the control group.
The strength of our study is the long-term follow-up of patients with a low dropout rate, probably due to close contact of patients and physicians. A limitation is that it was not placebo controlled. However, we felt it unethical to conduct a placebo-controlled study until final height. Another limitation is the relatively small number and heterogeneity of the patients.
In conclusion, GH in children with JIA treated with GC results in a sustained catch-up growth leading to a final height within target height range for the majority of our patients. Even after adjustment for different courses of the disease with higher overall disease activity in the control patients, the difference between the groups persisted. GH therapy was most effective in those patients with a moderate disease activity. However, because the course of the disease is unpredictable, we would recommend GH treatment in all JIA patients with growth failure as soon as possible to avoid additional height loss.
| Acknowledgments |
|---|
| Footnotes |
|---|
S.B., P.R., R.D.P., W.B., R.H., and H.M. have nothing to declare. H.P.S. received grant support (2000–2005) from Pfizer.
First Published Online May 29, 2007
Abbreviations: CRP, C-reactive protein; ESR, erythrocyte sedimentation rate; GC, glucocorticoid; HbA1c, glycosylated hemoglobin; IGFBP-3, IGF-binding protein 3; JIA, juvenile idiopathic arthritis; SDS, SD score.
Received March 13, 2007.
Accepted May 22, 2007.
| References |
|---|
|
|
|---|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 |