| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Department of Pediatrics, University of Bologna, S. Orsola-Malpighi Hospital, 40138 Bologna, Italy
Address all correspondence and requests for reprints to: Stefano Zucchini, Department of Pediatrics, via Massarenti 11, 40138 Bologna, Italy. E-mail: zucchini{at}med.unibo.it.
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Objectives: The objectives of this study are to establish the characteristics and the percentage of the subjects with isolated GHD who normalized secretion at puberty and to compare their statural outcomes with those of the subjects with persistent deficiency treated also after retesting.
Design and Setting: This was a prospective, nonrandomized, open-label study conducted in a university research hospital.
Patients and Intervention: Sixty-nine subjects (40 male, 29 female) with a diagnosis before puberty of isolated GHD by means of arginine and l-dopa tests were reevaluated with the same tests after at least 2 yr of therapy and after puberty onset. If GH peak at retesting was more than 10 µg/liter, therapy was withdrawn.
Main Outcome Measures: Percentage and characteristics of normalized subjects at retesting, outcome of treatment in the subjects treated or untreated to adult height, and factors predictive of growth outcome were measured.
Results: At retesting, 44 subjects (63.7%) confirmed a GH peak less than 10 µg/liter (24 of 40 male and 20 of 29 female). Apart from a less delayed bone age at diagnosis in females, the subjects with confirmed GHD were not different at diagnosis from the other group for height deficit at diagnosis, first year growth response to GH, age and height at puberty onset, height, and IGF-I at retesting. Mean adult height was 165.1 ± 4.5 cm in the male group treated until adult height vs. 164.0 ± 3.4 cm in the group who suspended therapy at retesting. Mean adult height was 153.2 ± 4.1 cm in the female group treated until adult height vs. 152.9 ± 5.2 cm in the group that suspended therapy at retesting. As regards the parameters expressing the final outcome, the only difference was found in the mean increment adult height-target height SD score in favor of the male group treated until adult height. In both sexes, therapy duration and GH levels at diagnosis and at retesting were unrelated to adult height parameters and to height increments during the period of observation.
Conclusions: One third of our GHD subjects diagnosed before puberty presented a normal secretion at puberty. The withdrawal of GH therapy in these subjects after retesting was not associated with a catch down growth, and they obtained an adult height similar to those obtained by the GHD subjects treated until adult height. It seems convenient, in subjects with nonsevere GHD, to retest GH secretion at midpuberty and to withdraw treatment for the subjects that are no longer deficient.
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
In 1995, we started a prospective study in which all subjects with a diagnosis before puberty of isolated and idiopathic GHD were treated with GH, retested at puberty and, if found normalized, left untreated until the end of growth, whereas the others with persistent deficiency continued the treatment. In this study, we present the adult height data of 69 patients, treated or not after retesting. Primary objectives of our study were to establish the characteristics and the percentage of the subjects who normalized secretion at puberty and the comparison of their statural outcome with that of the subjects with persistent deficiency treated also after retesting. A secondary objective was the examination of the possible predictive parameters influencing adult height in the whole group of subjects.
|
Subjects and Methods |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
At the time of retesting, all parents and subjects were informed about the results of the tests and about the consequent decision to withdraw treatment if GH secretion was normal and a written consent was obtained.
Inclusion criteria
At diagnosis, inclusion criteria were: 1) height less than third percentile and height velocity less than 25th percentile; alternatively height between the third and the 10th percentile, height velocity less than 25th percentile with target height more than 50th percentile; 2) GH peak less than 10 µg/liter after arginine and l-dopa tests; 3) no signs of puberty; and 4) normal magnetic resonance imaging (MRI) of the sellar region in cases of severe GHD (GH peak < 3 µg/liter).
At final evaluation, inclusion criteria were: 1) GH therapy for at least 24 months before GH retesting; 2) spontaneous puberty; and 3) isolated GH deficiency.
Exclusion criteria
Exclusion criteria were: 1) severe GHD associated with MRI abnormalities of the pituitary area; 2) multiple pituitary hormone deficiency; 3) nonidiopathic GHD; 4) congenital syndromes or bone diseases; 5) birth weight less than third percentile, preterm delivery; and 6) age at diagnosis less than 6 yr.
The latter was included to obtain a more homogeneous group of subjects in terms of age and therapy duration before puberty. All subjects with a GH peak less than 3 µg/liter at diagnosis or at retesting had a MRI of the pituitary area. The others did not undergo MRI because it is widely accepted that morphological abnormalities of the pituitary are unlikely except severe isolated GHD.
Table 1
shows the characteristics of the 69 patients examined. Forty were males, and 29 were females. In nine cases, birth weight was between the third and 10th percentile (five were in the group with persistent deficiency). Mean therapy duration in males with persistent deficiency was 87.3 ± 20 months as a whole and 44.2 ± 21 months before retesting, whereas in males with transient deficiency, it was 46.4 ± 18 months. Mean therapy duration in females with persistent deficiency was 61.8 ± 15 months as a whole and 35.7 ± 15 months before retesting, whereas in females with transient deficiency, it was 41.6 ± 16 months. Therapy duration before retesting was not statistically different in the two groups of subjects subdivided into males and females. At retesting, testosterone levels were measured in 27 of the 40 males, as well as estradiol levels in 19 of the 29 females. The choice of the subjects was due to technical reasons without any specific bias.
|
Height was measured with a Harpenden stadiometer. Height data were expressed as SD score (SDS) using the recently published Italian data (8). Adult height and growth during puberty were also provided in centimeters. Testicular volume was measured with a Prader orchidometer. Onset of puberty was defined by a testicular volume 4 ml or more in boys and by the presence of breast stage 2 in girls. If, at patient examination, these parameters had already been reached, the onset was calculated as mean height between that examination and the previous one. Target height was defined as sex-corrected midparental height (father’s height + mother’s height/2 + 6.5 cm for males and –6.5 cm for females) and was expressed in SDS units. Bone age was evaluated according to the method of Greulich and Pyle.
Serum GH was measured by a commercial solid-phase RIA assay (Technogenetics, Milan, Italy). The intraassay and interassay coefficients of variation were 5.8 and 9.7%, respectively, at the level of 1.2 µg/liter and 5.3 and 8.9% at the level of 12.5 µg/liter. Sensitivity of the assay was 0.1 µg/liter, as determined by the mean + 2 SD of the zero standard. IGF-I levels were evaluated both at diagnosis and at retesting, but unfortunately, our laboratory changed the assay between the two evaluations. Therefore, only values at retesting were considered in the analysis. Serum IGF-I was measured by a commercial solid-phase, enzyme-labeled chemiluminescent assay by DPC (Los Angeles, CA). The interassay coefficient of variation at the level of 150 µg/liter was 6.0%. Sensitivity of the assay was 20 µg/liter. Testosterone and estradiol were measured with commercial immunochemiluminometric assay kits (Bayer Healthcare LLC, Diagnostic Division, Tarrytown, NY).
Statistical analysis
For statistical analysis, the computer program Statistical Package for Social Science (SPSS Inc., Chicago, IL) was used on an IBM computer. Data distribution was analyzed with skewness and kurtosis coefficients. Results were expressed as means ± SD. Within groups comparisons were made using Student’s t test. 
2 Test was used to compare the proportions of subjects with adult height above or below target height and below the third percentile. Outcome of treatment was evaluated as adult height, as difference between adult height-target height SDSs, difference between adult height-height at diagnosis SDSs, percentage of subjects with adult height above target height, and number of subjects with adult height below the third percentile. R Pearson coefficient was used to correlate the variables analyzed. Multiple linear regression analysis evaluated the influence of explanatory variables on adult height and on the difference between initial height and adult height. For the latter, a prediction model was developed by means of multiple regression analysis. A hierarchy of prediction factor was derived for ordering predictive factors. All results nominally significant at P < 0.05 were indicated.
|
Results |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
shows the flow chart of the subjects subdivided according to type of GH peak after arginine and l-dopa tests (less than 3, between 3 and 10, above 10 µg/liter). Most of the subjects with severe deficiency (response < 3 µg/liter) at diagnosis or at retesting were also retested at the end of treatment. Among the nine cases with a severe GH deficiency at diagnosis, five displayed a normal secretion at retesting, two cases had a peak GH peak between 3–10 µg/liter, and two cases remained severely deficient. One of these remained severely deficient also at the end of therapy. Six of the 60 cases with peak at diagnosis of 3–10 µg/liter had a peak less than 3 µg/liter at retesting: four of these were retested at the end of treatment and, in one case, the peak was still less than 3 µg/liter.
|
shows the auxological characteristics at diagnosis, at puberty onset, and at retesting of the groups with persistent or transient deficiency subdivided into males and females. Apart from bone age delay at diagnosis in the female subjects, the values obtained were not statistically different between the various groups. Table 3 shows the final outcome expressed as adult height, height increment SDS from retesting to adult height, height increment SDS from initial to adult and target height, percentage of subjects with adult height above target height and with adult height less than third percentile. The only statistical difference was in the mean increment adult height-target height SDS, higher in the male group with persistent deficiency.
|
|
Males. Mean testicular volume at retesting was 9.7 ± 2.3 ml in subjects of group 1 and 9.6 ± 2.6 in group 2. Mean testosterone levels were 1.7 ± 0.9 ng/ml (range 0.25–3.5) in the 16 subjects of group 1 and 2.4 ± 0.9 ng/ml (0.5–3.5) in the 11 subjects of group 2 (NS). Testosterone levels were correlated with testicular volume (r = 0.58; P = 0.001), although they showed no correlations with GH peaks after pharmacological tests and IGF-I levels at retesting.
Females. Estradiol levels were 54.9 ± 30 pg/ml (range 20–110) in the 12 females of group 1 and 80.0 ± 80 pg/ml (range 20–248) in the seven females of group 2 (NS). There was a correlation at the limit of statistical significance between estradiol and IGF-I levels at retesting (r = 0.52; P = 0.06) and no correlation with GH peaks after pharmacological tests.
Correlations, multiple regression analysis, and prediction model
Males and females displayed different characteristics in the variables influencing adult height and were evaluated separately. All male and female results were calculated as a whole group, irrespective of the group of diagnosis at retesting. In both sexes, therapy duration and GH levels at diagnosis or at retesting were unrelated both to adult height parameters and to height increments during the period of observation. IGF-I levels at retesting were not correlated with GH peaks either at diagnosis or at retesting.
Males. Including in the equation whole height gain since diagnosis, height SDS at diagnosis, and bone age delay at diagnosis, adult height was influenced by total height gain SDS (R2 = 0.50; P = 0.0001) and by height SDS at diagnosis (R2 change = 0.39; P = 0.0001). When we examined the variables influencing height gain SDS from diagnosis to adult height introducing in the equation bone age delay at diagnosis, height SDS at diagnosis, GH peaks at diagnosis, and first year growth velocity, the significant variables were bone age delay at diagnosis (R2 = 0.20; P = 0.01) and height SDS at diagnosis (negative correlation; R2 change = 0.23; P = 0.002). In other words, the best statural increments were independently obtained with greater bone age delay at diagnosis and greater height deficit at diagnosis. The equation describing the predicted height gain was as follows: height gain = –1.84 + [bone age delay (yr) x 0.32] + [height SDS at diagnosis x –0.82]. The model explained 43% of the variability of the height gain with a total error SD of 0.83 SDS.
Females. By introducing in the equation whole height gain SDS since diagnosis, birth weight, height SDS at diagnosis, and age at puberty onset, the variables influencing adult height were height gain SDS from diagnosis to adult height (R2 = 0.80; P = 0.0001), height SDS at diagnosis (negative correlation; R2 change = 0.18; P = 0.0001), and age at puberty onset (R2 change = 0.01; P = 0.002). None of the parameters studied had an independent influence on whole height gain SDS, despite the fact that birth weight alone had a significant influence on whole height gain SDS (r = 0.40; P = 0.03). In females, it therefore was not possible to elaborate a prediction model.
|
Discussion |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Our data confirmed, first of all, that a consistent percentage of subjects diagnosed as GHD were no longer so when reexamined at puberty, after a mean period of GH therapy before retesting of 3–4 yr. In agreement with our previously published data (5) showing that puberty is a significant time for GH normalization, the percentage of our subjects that were persistently deficient (63.7%) is similar to that of Darendeliler et al. (7) (57%) and less than that found by Thomas et al. (6) (81%) in the subjects with isolated GHD examined 1 yr after the onset of treatment. Therefore, it appears that the percentage of subjects reaching a normal secretion at retesting increases progressively from prepuberty to midpuberty and end of growth, when most subjects with isolated GHD have been described to present a normal secretion. This is further confirmed by the fact that our subjects showing a GH peak less than 3 µg/liter at diagnosis or at retesting seldom confirmed their severe deficiency (Fig. 1
).
The evaluation of testosterone levels at retesting in some of our male subjects showed a tendency toward higher mean value despite their similar testicular volume. This may confirm the fact that testosterone, probably through its conversion to estrogens, was determinant in the normalization of GH secretion in some of our cases. This may require different GH cutoff values, depending on age and pubertal phase. However, we also confirmed our previous experience in which a few patients showed a worsened GH secretion despite the intervention of puberty (10).
Loche et al. (11) questioned the concept of transient GHD. The authors, in fact, found a normalized test in almost all cases only a few months after the first diagnosis and suggested the possibility that this type of patient may be falsely diagnosed as GHD. Obviously, we are aware that the reliability of most pharmacological tests has been debated, above all in the subjects with nonsevere GHD (3, 12, 13, 14), that the GH cutoff limits have been arbitrarily established without taking into account the physiological rise of GH secretion during puberty, and that the apparent normalization may simply be due to a regression to the mean effect. However, a fixed GH cutoff limit of 10 µg/liter is still the required finding to prescribe GH therapy in childhood in Italy, and it is also the value used by the majority of pediatric endocrinologists to prescribe GH therapy (15). If we retrospectively consider our decision to suspend GH treatment in some subjects based on casualty, the present study results could be examined as the final outcome of two groups of short stature subjects with similar auxological and hormonal characteristics at diagnosis, but with treatment withdrawal at puberty in one of the two groups. The analysis of the clinical characteristics of our subjects at diagnosis confirms the latter hypothesis, because, apart from bone age delay in the female group, only minimal differences were found between the subjects with persistent or transient deficiency. This is in agreement with some findings (16), but not with others (7) and probably reflects the clinical homogeneity of our study sample: most of our subjects, in fact, sooner or later presented a normal GH secretion after the initial diagnosis.
IGF-I levels at retesting were comparable in the two groups and close to the mean of our normal range, showing no correlation with the GH peak obtained at retesting. On the other hand, Loche et al. (11) in prepubertal patients and Hoeck et al. (17) in adolescents found no correlation between IGF-I values and GH peaks in subjects with nonclassical GHD. We also observed that IGF-I levels of our patients were similar at retesting in all groups and not consistent with those found in severe GHD. However, it has already been found that IGF-I levels show a good sensitivity only in detecting subjects with severe GHD, whereas the overlap between normal and GHD subjects is significant when a GH cutoff of 10 µg/liter is used (7, 18, 19). Furthermore, Tauber et al. (20) found IGF-I levels around 0 SD in adolescents with partial GHD at the end of treatment. However, we would underscore that a certain diagnosis of GHD could be made if more restrictive criteria are applied, i.e. a lower GH cutoff value and imaging data.
Considering the statural outcome of our subjects, both in terms of adult height and height increments on initial and target heights, the results obtained were not statistically different between the subjects who suspended the therapy and those who continued. The final difference between the groups of about 1 cm is a demonstration of their very similar statural outcome. Certainly a larger study sample may have detected more significant differences, but we would like to emphasize the fact that our data were obtained in more homogeneous subjects than those described in the literature. A weakness in our present study is that it was not designed to continue randomly or withdraw the treatment at retesting irrespective of the GH secretion obtained. This would have provided a theoretical solution concerning whether to withdraw treatment in the subjects with a normalized GH secretion. However, our results showed first of all that the withdrawal of GH therapy in the subjects with "transient" GHD determines minor auxological consequences both in terms of height velocity and residual growth after retesting, adult height, percentage of subjects overcoming target height, and number of subjects with adult height below the third percentile. This can be regarded from different points of view. The first would suggest that it is safe to withdraw GH treatment in the subjects showing a normalized secretion at midpuberty (hypothesis of transient GHD). The other would indicate that GH treatment, at a dose of 7 mg/m2·wk, has little influence on growth once the pubertal sex steroid secretion has started (hypothesis of false diagnosis of GHD). On the other hand, the role of GH treatment during the entire period of pubertal growth is still being debated (21), and the studies demonstrating a positive effect of GH on pubertal growth used higher GH doses (22). It is interesting that Reiter et al. (23) reported no increase in height SDS after the start of puberty in subjects with isolated GHD treated until adult height, without considering the possibility that some subjects could be no longer deficient during the last years of treatment. Therefore, part of our subjects displayed a normal height velocity also without GH, without the usual catch down phenomenon when GH is suspended before puberty as in the study by Darendeliler et al. (7). From a practical point of view, we recommend to reevaluate GH secretion at puberty in patients with GHD without anatomical abnormalities of the pituitary and subsequently to withdraw treatment if the deficiency is not confirmed. This would minimize national health system costs and reduce discomfort to patients and their families.
The height gain from diagnosis, ranging from 0.7 SD (4.6 cm) in the males treated until retesting to 1.0 SD (6 cm) in the females, obtained during the period of observation with our GH dose of 7 mg/m2·wk (similar to 0.3 mg/kg·wk) is comparable to our previous data (24), similar to other previously published studies on subjects with isolated GHD (25, 26, 27) or with idiopathic short stature (28) and slightly worse than in the two more recently published studies (23, 29). The difference between nonclassical GHD and idiopathic short stature, whose responses to GH therapy often overlap, has been questioned frequently (30, 31), and we cannot exclude that a percentage of our subjects, whose target heights were below the median of the normal population, may also be affected by familial idiopathic short stature.
It is also noteworthy that 13 subjects, variously distributed in all the groups (Table 3), presented an adult height below the third percentile despite several years of GH therapy. However, apart from severe GH deficiency, the criteria for deciding which subjects could be treated successfully have not yet been clarified, and our protocol was only useful in shortening GH treatment in a group of subjects.
Our data, in agreement with our previous investigation (11), but not with others (6), showed that first year response to therapy did not differ between subjects with persistent deficiency and those with normalized secretion. Also Tauber et al. (2) found no differences in adult heights between subjects with normal GH at final retesting and those with insufficient GH secretion.
Finally, the analysis of the parameters predicting growth and adult height, despite the relatively small study sample compared with others (25, 32), gave us some indication as to the usefulness of GH treatment in our subjects. First of all, GH levels at diagnosis and therapy duration were not predictive of future growth. Therefore, the height increment from diagnosis seemed more dependent on other factors possibly genetically determined and with an influence already described in the literature (27, 32). These characteristics differed between the two genders (in males, bone age delay at diagnosis and greater height deficit at diagnosis predicted 43% of the height increment) and had no relationship with GH response to pharmacological tests. As previously reported by our group (23), female subjects showed the least predictable response, possibly due to the prevalent secretion of estrogens, whose levels affect GH and IGF-I function (33).
In conclusion, the present study showed that the subjects with a normal secretion at puberty, about one third of our cases, were clinically indistinguishable at diagnosis from those with persistent deficiency. Withdrawal of GH therapy at a dose of 7 mg/m2·wk in the subjects who were no longer deficient led to a final result similar on average to that obtained by the other groups with longer treatment, with no catch down phenomenon. Therefore, it seems useful, in subjects with isolated GHD without anatomical abnormalities of the pituitary, to retest GH secretion at midpuberty and a safe practice to withdraw treatment in the subjects that are no longer deficient.
|
Footnotes |
|---|
Abbreviations: GHD, GH-deficient; MRI, magnetic resonance imaging; SDS, SD score.
Received February 21, 2006.
Accepted August 8, 2006.
|
References |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  G. Gelwane, C. Garel, D. Chevenne, P. Armoogum, D. Simon, P. Czernichow, and J. Leger Subnormal Serum Insulin-Like Growth Factor-I Levels in Young Adults with Childhood-Onset Nonacquired Growth Hormone (GH) Deficiency Who Recover Normal GH Secretion May Indicate Less Severe but Persistent Pituitary Failure J. Clin. Endocrinol. Metab., October 1, 2007; 92(10): 3788 - 3795. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. di Iorgi, A. Secco, F. Napoli, C. Tinelli, A. Calcagno, N. Fratangeli, L. Ambrosini, A. Rossi, R. Lorini, and M. Maghnie Deterioration of Growth Hormone (GH) Response and Anterior Pituitary Function in Young Adults with Childhood-Onset GH Deficiency and Ectopic Posterior Pituitary: A Two-Year Prospective Follow-Up Study J. Clin. Endocrinol. Metab., October 1, 2007; 92(10): 3875 - 3884. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Radovick and S. DiVall Approach to the Growth Hormone-Deficient Child during Transition to Adulthood J. Clin. Endocrinol. Metab., April 1, 2007; 92(4): 1195 - 1200. [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 |
