| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Endocrine Care |
Departments of Pediatrics (R.C., S.R., J.M.L.) and Radiology (D.L.), University Hospital, 49000 Angers, France; and Department of Pediatrics, University Hospital (F.D., N.M.), 37000 Tours, France
Address all correspondence and requests for reprints to: Dr. Régis Coutant, Department of Pediatrics, University Hospital, 4 rue Larrey, 49000 Angers, France. E-mail: recoutant{at}chu-angers.fr
Abstract
We analyzed the final height of 146 short children with either nonacquired GH deficiency or idiopathic short stature. Our purpose was 1) to assess growth according to the pituitary magnetic resonance imaging findings in the 63 GH-treated children with GH deficiency and 2) to compare the growth of the GH-deficient patients with normal magnetic resonance imaging (n = 48) to that of 32 treated and 51 untreated children with idiopathic short stature (GH peak to provocative tests >10 µg/liter). The mean GH dose was 0.44 IU/kg·wk (0.15 mg/kg·wk), given for a mean duration of 4.6 yr.
Among the GH-deficient children, 15 had hypothalamic-pituitary abnormalities (stalk agenesis), all with total GH deficiency (GH peak <5 µg/liter). They were significantly shorter and younger at the time of diagnosis than those with normal magnetic resonance imaging, had better catch-up growth (+2.7 ± 0.9 vs. +1.3 ± 0.8 SD score; P < 0.01), and reached greater final height (-1.1 ± 1.0 vs. -1.7 ± 1.0 SD score; P < 0.05). Among patients with normal magnetic resonance imaging, there was no difference in catch-up growth and final height between partial and total GH deficiencies.
GH-deficient subjects with normal magnetic resonance imaging and treated and untreated patients with idiopathic short stature had comparable auxological characteristics, age at evaluation, and target height. Although they had different catch-up growth (+1.3 ± 0.8, +0.9 ± 0.6, and +0.7 ± 0.9 SD score, respectively; P < 0.01, by ANOVA), these patients reached a similar final height (-1.7 ± 1.0, -2.1 ± 0.8, and -2.1 ± 1.0 SD score, respectively; P = 0.13).
Pituitary magnetic resonance imaging findings show the heterogeneity within the group of nonacquired GH deficiency and help to predict the response to GH treatment in these patients. The similarities in growth between the GH-deficient children with normal magnetic resonance imaging and those with idiopathic short stature suggest that the short stature in the former subjects is at least partly due to factors other than GH deficiency.
GH DEFICIENCY (GHD) is diagnosed on the basis of auxological data, namely short stature and growth deceleration, in addition to low serum GH levels in response to stimulation tests (1, 2). GHD has usually been considered to be idiopathic when no acquired cause impairs hypothalamic or pituitary function (3), and several studies have reported the long-term results of GH treatment in children with idiopathic (nonacquired) GHD (reviewed in Refs. 4 and 5). Recently, magnetic resonance imaging has proved valuable for the investigation of apparently idiopathic GHD by showing minor abnormalities of the hypothalamus or pituitary (anterior pituitary hypoplasia, pituitary stalk agenesis, and ectopic posterior pituitary) (6, 7, 8, 9). To date, no report has been published assessing the long-term results of GH treatment in GHD according to magnetic resonance imaging (MRI) findings.
Several recent studies have found that 40–67% of children diagnosed with GHD normalized their GH secretion after growth completion (10, 11, 12, 13). Guyda (5) suggested that the initial diagnosis of GHD was uncertain in most of them; a significant number of GH-treated children previously labeled as GHD may have idiopathic short stature. Maghnie et al. (10), reassessing GHD patients after growth completion, found that all subjects with normal pituitary MRI or isolated small pituitary gland normalized their GH secretion. They proposed that these patients should be retested well before the attainment of adult height. No study has yet addressed this by comparing the growth of GHD children with normal pituitary MRI to that of children with idiopathic short stature (ISS; GH peak >10 µg/liter).
We report the growth and final height of 146 children with either GHD or ISS. The first aim of this study was to analyze the response to treatment in children with nonacquired GHD (n = 63) according to pituitary MRI findings. We showed that auxological characteristics at evaluation as well as response to treatment were different between GHD patients with abnormal and normal pituitary MRI. The second aim was to examine the role of GHD in the short stature of patients with GHD and normal pituitary MRI (n = 48) by comparing their growth and final height to those of 32 treated and 51 untreated non-GHD short children (idiopathic short stature).
Subjects and Methods
Patients
Sixty-three patients with nonacquired GHD were included based on the following criteria: 1) height at the time of evaluation below -3 SD or growth velocity below -2 SD and height -2 to -3 SD below the mean (2), 2) diagnosis of GHD defined as peak serum GH below 10 µg/liter after two conventional stimulation tests, and 3) no acquired cause (trauma, neoplasm, or irradiation) impairing hypothalamic or pituitary function. Criteria for exclusion were chromosomal abnormalities, dysmorphic syndromes, skeletal dysplasia, chronic illness, and primary hypothyroidism. All of the GHD subjects were treated with GH. Eighty-three short children (ISS) had height at the time of evaluation below -2 SD. They all had a peak serum GH of 10 µg/liter or more. Exclusion criteria were similar. Thirty-two of the ISS patients volunteered for GH treatment, and 51 were followed to final height without treatment, primarily because of refusal of GH therapy. The SD scores of birth weight, birth height, height, growth velocity, and body mass index for age and sex were calculated (14, 15, 16). Pubertal stage was recorded (17, 18). At evaluation, 115 subjects were prepubertal, and the remaining 31 were at Tanner stage 2. Genetic target height (midparental height) was calculated as previously described (19), using height standards obtained during the approximate parental generation (20). Predicted height was calculated using the Bayley-Pinneau method (21, 22).
Diagnosis of GHD
The stimuli used for pharmacological testing of GH secretion were ornithine (86% of children), arginine and insulin hypoglycemia (65% of children), and glucagon and betaxolol (49% of children). Serum GH was measured by polyclonal RIA using kits obtained from BioMerieux (Marcy-l’Etoile, France) calibrated against the First International Reference Preparation (Medical Research Council 66/217). The intra- and interassay coefficients of variation were 2% and 5.5%, respectively, at a level of 10 µg/liter. GH results are expressed in Medical Research Council 66/217 units (2 µIU = 1 ng).
MRI
Pituitary MRI was performed with a 0.5-tesla superconductive system with a head coil, a 256 x 256 reconstruction matrix, and a 20-cm field of view. T1-weighted images (spin-echo repetition time, 400; echo time, 200) were obtained in the sagittal and coronal planes using 3-mm sections. The presence or absence of the pituitary stalk was recorded before and after gadolinium injection in all cases.
Classification of nonacquired GHD
Total GHD was diagnosed when peak serum GH was less than 5 µg/liter (23 patients), and partial GHD when peak serum GH was between 5–10 µg/liter (40 patients). In addition, MRI of the brain identified congenital hypothalamic-pituitary abnormalities in 15 of the 23 patients (65%) with total GHD (pituitary stalk interruption, ectopic or nonlocated posterior pituitary, and hypoplasia of the anterior pituitary), whereas all patients with partial GHD had normal pituitary MRI.
Treatment and follow-up
A total of 95 children started GH treatment between 1983 and 1994 (92% of the patients after 1987). Ninety-one received exclusively recombinant GH (6 times weekly), and 4 patients were treated with human-extracted GH for 6 months to 2 yr, then with recombinant GH. Mean doses were given after correction for weight gain between study visits. Treatment lasted at least 2 yr. The mean GH dose was 0.44 ± 0.11 IU/kg·wk (0.15 ± 0.04 mg/kg·wk). Patients were recruited and followed in the Pediatric Department of Angers University Hospital (n = 63) or Tours University Hospital (n = 32); all were seen every 3 months. Untreated children were followed in Angers University Hospital during the same study period.
The criteria for discontinuation of GH treatment was a growth velocity less than 2 cm/yr in the previous 6 months with a bone age greater than 13 yr for girls and 15 yr for boys. The mean delay between end of treatment and the last visit was 2.2 ± 1.9 yr.
Fifteen children with GHD had TSH deficiency and received appropriate T4 replacement. Seven boys and two girls had gonadotropin deficiency. Boys received T heptylate, and girls received ethinyl E2 after 15 and 14 yr of age, respectively. Dosages were progressively increased over a 3-yr period.
Final height
Final height was measured in boys at Tanner stage 5 with a bone age of 17.5 yr or more (99.4% of final height according to Bayley-Pinneau) or a growth velocity less than 0.5 cm/yr. Chronological age and bone age were 19.1 ± 1.8 and 18.0 ± 1.4 yr (99.6% of final height), respectively. Final height was measured in girls at Tanner stage 5 with a bone age of 16 yr or more (99.6% of final height according to Bayley-Pinneau), or a growth velocity less than 0.5 cm/yr. Chronological age and bone age were 18.0 ± 2 and 16.7 ± 1.4 yr (99.7% of final height), respectively. These combinations of criteria adequately select patients with adult height according to the National Cooperative Growth Study (23).
Statistical analysis
Results are expressed as the mean ± SD. Variance analyses followed by Tukey’s procedure and t tests were used for comparison between groups. P < 0.05 was considered significant. Statistical tests were performed with the SPSS 9.0 statistical package (SPSS, Inc., Chicago, IL).
Results
Baseline characteristics of GHD patients according to MRI findings
(Table 1
)
Auxological characteristics were significantly different between the GHD patients with abnormal and those with normal pituitary MRI findings. Chronological age, bone age, and height SD score at evaluation were lower in patients with abnormal pituitary MRI as a result of their lower postnatal growth (P < 0.05 for each comparison). By contrast, birth length and target height were lower in the patients with normal pituitary MRI, indicating a genetic contribution to their short stature (P < 0.05 for each comparison). Pretreatment growth velocities were similar between the GHD patients with abnormal pituitary MRI and those with normal pituitary MRI. However, none of the 15 subjects with abnormal pituitary MRI had pretreatment growth velocity over -0.5 SD score, whereas 9 of 48 GHD subjects with normal MRI had growth velocity between 0 and -0.5 SD score (the other 39 subjects had growth velocity below -0.5 SD score). All patients with abnormal MRI had total GHD.
|
Overall, treated patients received an average GH dose of 0.44 ± 0.1 IU/kg·wk (0.15 ± 0.03 mg/kg·wk) for 5.0 ± 2.4 yr. Adult height was -1.6 ± 1.0 SD score, and final catch-up growth (difference between height SD score at final height and that at first evaluation) was +1.6 ± 1.0 SD score.
However, age at diagnosis was lower and duration of treatment was
longer in the patients with abnormal MRI (P < 0.01;
Table 1
). The 1-yr catch-up growth was significantly higher in this
group (+1.0 ± 0.6 vs. +0.4 ± 0.4
SD score, abnormal vs. normal MRI;
P < 0.01), as were the final catch-up growth
(+2.7 ± 0.9 vs. +1.3 ± 0.8
SD score; P < 0.01) and the
final height (-1.1 ± 1.0 vs. -1.7 ± 1.0
SD score; P < 0.05).
Baseline characteristics and evolution of height in GHD patients according to serum peak GH
Patients with total GHD had a lower height SD score at evaluation (-3.5 ± 1.2 vs. -2.8 ± 0.7 SD score; P < 0.05) and a better final catch-up growth (+2.1 ± 1.5 vs. +1.3 ± 0.7 SD score; P < 0.01) than patients with partial GHD.
All patients with partial GHD had normal pituitary MRI, whereas 15 of
the 23 patients with total GHD had abnormal pituitary MRI. Only these
latter subjects had a lower height SD score at evaluation
(-3.8 ± 1.2 vs. -2.8 ± 0.7
SD score; P < 0.05) and a better
final catch-up growth (+2.7 ± 0.9 vs. +1.3 ± 0.7
SD score; P < 0.01) than
patients with partial GHD (Table 1 and Fig. 1). Conversely, the subjects with
total GHD and normal MRI had pretreatment auxological characteristics,
final catch-up growth, and final height in the same range as that
observed in subjects with partial GHD (Fig. 1 and Table 2). Three of the 8 subjects with total
GHD and normal MRI had pretreatment growth velocity between 0 and -0.5
SD score (Table 2; the other subjects had growth
velocity below -0.5 SD score).
|
|
)At the time of evaluation, age, height SD score, and pubertal status were similar in these patients. All subjects had equally reduced target height.
|
As auxological characteristics were similar in patients with ISS
and GHD with normal MRI, we compared the evolution of height in these
subjects. The 1-yr catch-up growth was similar between treated patients
(+0.4 ± 0.4 vs. +0.4 ± 0.4
SD score). The final catch-up growth was
significantly different among untreated children with ISS, treated
children with ISS, and treated children with GHD and normal MRI
(+0.66 ± 0.89, +0.90 ± 0.57, and +1.27 ± 0.81
SD score; P < 0.01, by ANOVA);
post-hoc analysis showed that it was significantly better in
treated children with GHD and normal MRI than in untreated subjects
with ISS, but was not different between treated subjects with ISS and
the two other groups (Table 3). However, untreated patients with ISS,
treated patients with ISS, and treated subjects with GHD and normal MRI
reached similar adult height SD scores
(-2.1 ± 1.0, -2.1 ± 0.8, and -1.7 ± 1.0
SD score, respectively; P = 0.13;
Fig. 2).
|
Forty-two of the 63 subjects with GHD (67%) underwent pharmacological testing after completion of growth, and 52% had a peak serum GH over 10 µg/liter. Twenty-two of 35 retested subjects with GHD and normal MRI had a peak serum GH of 10 µg/liter or more compared with none of 7 retested subjects with GHD and abnormal MRI. We also checked that both the baseline auxological characteristics and the response to treatment in GHD patients who were not reassessed were not different from those of retested subjects (not shown).
Discussion
In this report we analyzed the growth data of GH-treated patients with nonacquired GHD according to the pituitary MRI findings. We observed that auxological characteristics, growth response to treatment, and final height were different between GHD patients with abnormal and those with normal pituitary MRI. Among the patients with normal pituitary MRI, there was no difference in catch-up growth and final height between subjects with total and partial GHD. These facts underline the heterogeneity of the nonacquired GHD group and the relevance of pituitary MRI findings to predict the growth response to treatment. Auxological characteristics, evolution of height under GH treatment, and final height were similar in the patients with ISS and GHD with normal MRI, suggesting that the growth defect in the latter group may be at least partly the result of factors other than GHD, such as genetic factors.
Several studies have evaluated the long-term benefits of GH treatment in children with nonacquired GHD (4, 5). Summarizing the growth results of 4529 subjects whose treatment ended after 1987, Guyda (5) found a mean catch-up growth of +1.5 SD score for a mean treatment duration of 6.2 yr, leading to a mean final height of -1.4 SD score. In our study, catch-up growth was +1.6 ± 1.0 SD score, duration of treatment was 5.0 ± 2.4 yr, and final height was -1.6 ± 1.0 SD score for the GHD subjects, indicating that our auxological criteria and provocative tests used for GHD diagnosis selected patients similar to those in published series. However, the results varied widely between studies; for GH doses of 0.30–0.60 IU/kg·wk (0.10–0.20 mg/kg·wk), catch-up growth ranged from +0.5 up to +2.7 SD score, suggesting that GHD is a heterogeneous group. Evidence of the level of GH secretion as a predictor of growth response to GH treatment has been inconclusive (24, 25, 26, 27, 28). In their predictive models of the growth response to treatment in GHD, Ranke et al. (28) found that the maximum GH level during provocative testing was the most important predictor, but it was relevant only when the GH level was below 5 µg/liter. Whereas some researchers found a more favorable effect of GH in patients with total GHD than with partial GHD, others found the effect to be similar (29, 30, 31, 32). In our study we observed a significant difference in catch-up growth between total and partial GHD. However, this difference was entirely explained by the subjects with abnormal MRI, who all had total GHD, whereas subjects with normal MRI and total GHD had pretreatment growth, growth response to treatment, and final height similar to those of patients with partial GHD (who all had normal pituitary MRI). These results may explain the apparent discrepancies between studies of total and partial GHD. They point out that pituitary MRI findings are more accurate than the GH peak to provocative tests to indicate true profound GHD and to predict the response to treatment in patients with nonacquired GHD. The better response to GH treatment in children with abnormal pituitary MRI may suggest that GHD was their main cause of shortness, whereas it was only part of the reason why children with normal pituitary MRI were short.
The GHD children with abnormal pituitary MRI had features indicative of a more profound GHD than those with normal MRI; they were shorter and younger at the start of GH treatment, had a more delayed bone age, and had a longer treatment duration. In addition, they had multiple pituitary hormone deficiencies, as is usually described (6, 7, 8, 9). Sixty percent of these subjects (9 of 15) had gonadotropin deficiency, and their pubertal development was induced artificially at a later age than that of physiological pubertal development. All of these factors have been described in previous studies as significantly associated with the height gain in GH-treated children with GHD (5, 24, 25, 26, 27, 28, 33, 34, 35) and may contribute to the higher final height in this group.
We observed that clinical characteristics and genetic target height were similar in the patients with ISS and GHD with normal MRI. Therefore, we compared their evolution of height using ANOVA. We found a significant difference in catch-up growth among untreated patients with ISS, treated patients with ISS, and treated patients with GHD and normal MRI, suggesting a progressive increase in catch-up growth from the untreated ISS children to the treated children with GHD and normal MRI. At a mean dose of 0.46 IU/kg·wk (0.15 mg/kg·wk), GH treatment was associated with a mean height gain of 0.25 SD score for ISS (difference between catch-up growth in treated and untreated ISS). This height increase, although not significant, is consistent with the results of controlled studies of GH treatment for ISS, which found a height gain attributed to treatment ranging from 0.30–1.2 SD score for GH doses ranging from 0.5–1 IU/kg·wk (0.16–0.33 mg/kg·wk) (4, 5). The GH dose used to demonstrate an increase in the height of patients with ISS is somewhat higher than that currently used for children with GHD. We could not really assess the height gain due to GH treatment in the subjects with GHD and normal MRI because there was no control (untreated) group, and the spontaneous growth of these patients is unknown. However, the children identified as GH deficient with normal pituitary MRI had an evolution of height during treatment and a final height slightly, but not significantly, better than those of treated ISS subjects. These results indicate that the short stature of children diagnosed with nonacquired GHD and normal MRI shares common determinants with that of children with ISS. These considerations justify prospective controlled randomized trials of the effect on adult stature of higher GH doses in children diagnosed with GHD and normal pituitary imaging; these patients, like those with ISS, might benefit from higher doses of GH than those used in GHD.
As GH secretion is a continuous spectrum, a number of the GHD children with normal MRI may be short children with GH secretion in the lower part of the normal range (2, 5, 36, 37, 38). After completion of growth, 63% of our reevaluated GHD subjects with normal MRI (22 of 35) were retested as normal, compared with 100% in the study by Maghnie et al. (10). Percentages have ranged from 40–67% in the series of subjects with nonacquired GHD regardless of the MRI findings (10, 11, 12, 13). This indicates that initial testing resulted in an erroneous diagnosis of GHD or that the low GH production measured at the time of evaluation in several of our patients may not have lasted to the end of growth. More stringent criteria may need to be used to select children with true profound GHD who will benefit from GH treatment at current dosages. It has been emphasized that many of the problems associated with the diagnosis of GHD are the result of inappropriate testing of children who do not have genuine growth failure: specifically, a low growth velocity is certainly necessary to diagnose true GHD (2). In addition, low serum IGF-I and IGF-binding protein-3 measurements, sex steroid priming before GH testing, and a lower cut-off level of GH peak to provocative tests have been proposed to overcome the limitations of GHD diagnosis (2, 36, 39). In any case, the above criteria were not generally used to ascertain GHD diagnosis at the time of evaluation in most of our patients and have not yet been used in published series of final heights of GHD patients. It is likely that these criteria will select a high percentage of subjects with abnormal pituitary MRI, and it has been suggested that only a minority of patients have true profound GHD with normal pituitary MRI (10, 40). In fact, normal pituitary MRI and profound GHD have been observed in rare cases of subjects with a genetic cause of GHD (41).
In conclusion, our data indicate that pituitary MRI findings help to predict the evolution of height under GH treatment in children with nonacquired GHD. With mean GH doses of 0.44 IU/kg·wk (0.15 mg/kg·wk), children with abnormal MRI have a significantly higher final height than those with normal MRI. These latter children reach a final height close to that of treated children with ISS. Their short stature seems to be at least partly the result of factors other than GHD. Prospective randomized trials are needed to determine whether the children with GHD and normal pituitary MRI will benefit from higher GH doses than that currently used in GHD.
Acknowledgments
Footnotes
Abbreviations: GHD, GH deficiency; ISS, idiopathic short stature.
Received July 26, 2000.
Accepted July 16, 2001.
References
for predicting
adult height from skeletal age. Revised for use with the Greulich-Pyle
hand standards. J Pediatr 40:423–441[CrossRef][Medline]
This article has been cited by other articles:
 |
|
 |
|
  L. Dunkel Management of children with idiopathic short stature Eur. J. Endocrinol., November 1, 2006; 155(suppl_1): S35 - S38. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. M. Lee, M. M. Davis, S. J. Clark, T. P. Hofer, and A. R. Kemper Estimated Cost-effectiveness of Growth Hormone Therapy for Idiopathic Short Stature Arch Pediatr Adolesc Med, March 1, 2006; 160(3): 263 - 269. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. W. Leschek, S. R. Rose, J. A. Yanovski, J. F. Troendle, C. A. Quigley, J. J. Chipman, B. J. Crowe, J. L. Ross, F. G. Cassorla, W. F. Blum, et al. Effect of Growth Hormone Treatment on Adult Height in Peripubertal Children with Idiopathic Short Stature: A Randomized, Double-Blind, Placebo-Controlled Trial J. Clin. Endocrinol. Metab., July 1, 2004; 89(7): 3140 - 3148. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. G. F. Osorio, S. Marui, A. A. L. Jorge, A. C. Latronico, L. S. S. Lo, C. C. Leite, V. Estefan, B. B. Mendonca, and I. J. P. Arnhold Pituitary Magnetic Resonance Imaging and Function in Patients with Growth Hormone Deficiency with and without Mutations in GHRH-R, GH-1, or PROP-1 Genes J. Clin. Endocrinol. Metab., November 1, 2002; 87(11): 5076 - 5084. [Abstract] [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 |
