| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Departments of Endocrinology, Pediatric Oncology and Clinical Oncology, Christie Hospital, Manchester, United Kingdom M20 4BX; and Department of Pediatrics, Addenbrooke’s Hospital (A.L.O.-S.), Cambridge, United Kingdom CB2 2QQ
Address all correspondence and requests for reprints to: Prof. S. M. Shalet, Department of Endocrinology, Christie Hospital, Wilmslow Road, Withington, Manchester, United Kingdom M20 4BX. E-mail: helena.gleeson{at}christie-tr.nwest.nhs.uk.
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
GHD occurs secondary to damage to the hypothalamic-pituitary axis due to either the proximity of the brain tumor or after cranial radiation. GH secretion is the most frequent and often the only pituitary hormone affected by irradiation. Cranial irradiation doses as low as 18 Gy may adversely affect GH status (1). The majority of children receiving a dose in excess of 27 Gy to the hypothalamic-pituitary axis, a dose used in the treatment of brain tumors, will become GHD within 2–5 yr (2).
A brain tumor itself, cerebrospinal fluid obstruction leading to hydrocephalus, and irradiation to the hypothalamic area may affect the timing of puberty. Cranial irradiation (CI) with a dose in excess of 18 Gy administered to a prepubertal child can induce early or even precocious puberty, particularly in girls (3, 4). The age of onset of puberty is also related to the age at irradiation (4).
GH replacement has been used to improve FH in childhood cancer survivors with radiation-induced GHD since 1975. Repeated single center and multinational database studies (5, 6, 7, 8, 9) have demonstrated that there is no increased risk of brain tumor recurrence in those treated with GH. As a consequence, GH therapy is instituted earlier after completion of cancer therapy. More recently, GnRH analogs (GnRHa) have been introduced to delay early or rapidly progressing puberty to allow more time for linear growth. The impact of these latter changes in management on the growth outcome of brain tumor survivors with radiation-induced GHD is unknown.
Studies to FH in childhood brain tumor survivors are important to determine the effectiveness of growth-promoting strategies. We have performed a retrospective study of patients who have reached FH and were treated with GH therapy for radiation-induced GHD secondary to brain tumor therapy. Our aim was to assess whether FH outcome is improving and to determine whether GnRHa therapy has contributed auxologically.
|
Subjects and Methods |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Sixty-eight children with a diagnosis of a brain tumor were identified as starting GH replacement between 1975–2000. Ten children were excluded: 8 were GHD as a consequence of a tumor affecting the hypothalamic pituitary axis, 1 had an additional reason for poor growth documented before diagnosis and treatment of the brain tumor, and 1 had neurofibromatosis with a long history of untreated multiple gliomas that had affected growth. Growth data were examined in 58 children (31 males and 27 females) with radiation-induced GHD. This included 7 children in whom the tumor was in close proximity to the hypothalamic-pituitary axis, but GH status was established as normal at the time of irradiation, thereby implying that the tumor itself had not rendered the patient GH deficient. All children had received at least 1 yr of GH and reached FH.
Cancer therapy
All had received CI or craniospinal irradiation (CSI) with or without CT. Seventeen children with the histological diagnosis of locally invasive tumor received a fractionated course of radiotherapy (median total dose, 42.5 Gy in 16 fractions over 21 d). One patient had local irradiation to the posterior fossa only, 2 had local irradiation to the orbit or optic nerve, and the remaining 14 patients received radiation to the hemispheres. Thirty-eight children with a tumor with the propensity to disseminate throughout the central nervous system received an initial fractionated course of radiotherapy to the tumor (median dose, 15 Gy in 10 fractions over 11 d), followed by a course of radiotherapy to the whole head and spine (median dose, 30 Gy in 20 fractions over 27 d). Thirty-four patients had local radiation to the posterior fossa, 2 had local radiation to the pineal gland, and the remaining 2 received radiation to the hemispheres. Three children treated before 1972 received multiple field irradiation (2 fields), including the spine without a separate dose to the tumor site (median dose, 27 Gy in 17 fractions over 22 d). In most children spinal irradiation was given as oblique wedge pairs, but in very young children (<2 yr of age) it was given as a single posterior field with a reduced dose.
In the treatment of the original disease, 20 children received 12–18 months of CT, which consisted of vincristine alone (1.5 mg/m2; n = 3) or in combination with carmustine (100 mg/m2; n = 4) given every 6 wk for 12 months (Manchester protocol before 1975), or vincristine (1.5 mg/m2) in combination with lomustine alone (40 mg/m2; n = 2) or with procarbazine (100 mg/m2; n = 11) given for a period of 18 months (United Kingdom Children’s Cancer Study Group protocol after 1979). Three additional children received other chemotherapy for a shorter time interval: two courses of mustine, vincristine, prednisolone, and procarbazine (n = 1); cisplatinum (n = 1); or actinomycin (n = 1).
GH therapy
GHD was diagnosed biochemically if the peak GH response to two provocative tests (arginine test plus insulin tolerance test or glucagon stimulation test) was less than 6.7 ng/ml (20 mU/liter). GH therapy was started after biochemical confirmation of GHD. Before May 1985, children received 5 mg pituitary-derived GH three times a week by im injection. Thereafter, synthetic GH was used, initially in a dose of 4 IU three times a week regardless of weight. Since 1988 children have received 0.5 IU/kg·wk, administered by daily sc injection.
GnRHa therapy
There is no set protocol for the introduction of GnRHa. The decision to start GnRHa therapy was clinically based on the onset and tempo of puberty, the confirmation of a pubertal pattern of gonadotropin response to a GnRH stimulation test, and the individual child’s growth prediction at that time. Once started, the GnRHa schedule was altered according to both progression through puberty and evidence of adequate suppression of the gonadotropin response to a GnRH stimulation test. The decision to discontinue GnRHa therapy was based on age, height prediction, and the views of the child and parents.
Sex steroid replacement
At the discretion of the clinician, sex steroids were instituted in children who failed to start or progressed poorly through puberty in the presence of biochemical evidence of hypogonadism.
Auxology
Anthropometry was carried out every 3–4 months on wall- and table-mounted stadiometers by trained observers to determine standing (StandH) and sitting height (SitH) measurements. Subischial leg length (LL) measurements have been extrapolated from these direct measurements. Data were collected from clinic visits close to the time of irradiation (<6 months) and at FH. The raw data have been transformed to SD scores using UK90 reference data (10).
FH has been defined as the height achieved when measurements over 12 months differ by less than 1 cm. Optimization of FH was assessed by comparing it with midparental height [height loss (HL) = FH minus midparental height].
Disproportion at FH has been calculated by the formula: (LL - SitH) - (µLL - µSitH), where µ is the population mean.
Puberty
Pubertal status was assessed by the method of Marshall and Tanner (11, 12). The onset of puberty was taken as G2/B2. The age at which this occurred was taken as the midage between the clinic visit previously when the child was still prepubertal and the clinic visit when the child was G2/B2. Duration of puberty was taken as time from G2/B2 to G4/B4. Age at G4/B4 was assessed similarly by taking the midage between the clinic visit previously when the child was G3/B3 and the clinic visit when the child was G4/B4. Data from patients requiring sex steroid replacement for the initiation of puberty and those requiring GnRHa to halt precocious puberty secondary to a lesion directly affecting the hypothalamus were excluded from the analysis of pubertal onset. Data from patients requiring sex steroid replacement during puberty and those requiring GnRHa to halt puberty were excluded from the analysis of puberty duration.
Statistics
All statistics were performed using the package SigmaStat (Jandel Scientific, San Rafael, CA). The normality of the studied variables was tested using the Komolgorov-Smirnov test. Given the non-Gaussian distribution of lag (time from irradiation to starting GH), these were natural log transformed (ln(lag)). All data are expressed as the mean (±SD) unless otherwise stated. P < 0.05 was taken as significant.
Unpaired t tests were performed to compare the characteristics of the patients treated with GH before and since 1988. Pearson’s correlation was used to examine for associations. Forward stepwise regression analyses were performed to determine predictors of outcome in the whole group (A; n = 58) and the group treated with GH after 1988 excluding those who were treated with GnRHa therapy (B; n = 28). FH outcome [HL and StandH SD score (SDS)], segmental growth and disproportion at FH, and onset and duration of puberty were dependent variables, and clinical variables [age at irradiation, lag time (ln), modes of cancer therapy, and use of GnRHa] were used as independent variables.
Outcome with GnRHa usage was assessed using the predictive models of FH outcome (both HL and StandH SDS), segmental growth, and disproportion at FH created from the data of those patients in group B, as this time period covers the use of the GH regimen, i.e. daily injections of synthetic GH, that was used in patients treated with GnRHa and therefore may more accurately predict outcome. Paired t tests were used to compare the actual height outcome with the predicted outcome in the patients treated with GnRHa.
|
Results |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
).
|
The mean age at irradiation was 6.1 (2.9) yr, and at initiation of GH therapy it was 10.1 (2.7) yr. The lag time between irradiation and GH therapy was 4.0 (2.1) yr, and the duration of GH therapy was 5.2 (2.7) yr. All patients achieved FH, with no evidence of brain tumor recurrence at a mean time after irradiation of 9.2 (2.9) yr.
There were no differences in age at irradiation between those subjects treated with GH before 1988 and since 1988 [mean (±SD) 5.8 (3.0) vs. 6.2 (2.9) yr; P = 0.6], but the earlier treated cohort did experience a more prolonged time interval from completing irradiation to starting GH [5.4 (2.4) vs. 3.3 (1.6) yr; P < 0.001]. The age at irradiation correlated inversely with ln(lag) (r = -0.4; P < 0.001), i.e. those treated with irradiation at a younger age had a longer lag time until starting GH.
Other hormone replacement
Seven patients (1 male and 6 females) required sex steroids to initiate or complete puberty (Table 1
). Five of the 7 patients required long-term sex steroid replacement. At the most recent clinic visit: 12 patients (2 males and 10 females) required sex steroid replacement; 6 patients had primary gonadal failure; 15 patients required T4 replacement, of whom 13 patients had received direct radiation damage to the thyroid; and ACTH deficiency requiring hydrocortisone replacement was diagnosed in 2 patients in adult life.
Assessment of GH status
The time after completion of radiation until first biochemical investigation of GH status was 2.8 (2.1) yr. At first assessment 45 patients were GHD, and 13 were not GHD; in those who proved to be GHD, tests were performed at 3.2 (2) yr after irradiation compared with 1.5 (1.8) yr in those who were not GHD.
At the time at which biochemical GHD was confirmed in all patients, 46 patients underwent an insulin tolerance test, 10 underwent a glucagon stimulation test, and 56 underwent an arginine stimulation test, with peak GH levels achieved of 2.5 (1.2), 2.7 (1.5), and 2.7 (1.8) ng/ml, respectively.
Puberty
The timing of onset and duration of puberty data are presented in Tables 2 and 3. Pubertal onset in the girls was earlier than that in the normal population (10.2 vs. 11.2 yr); however, in the boys the age of onset was similar to that in the normal population (11.6 vs. 12 yr) (11, 12). Using forward stepwise regression, the onset of puberty can be predicted by gender (P = 0.02) and age at irradiation (P = 0.003), i.e. the younger the age at irradiation the earlier the onset of puberty in both males and females.
|
|
Changes in outcome over the time course of the study
Throughout the study period there were improvements in HL and StandH, SitH, and LL SDS at FH (Fig. 1). There was a trend toward improved FH (a reduction in HL) both in patients treated with CI (r = 0.5; P = 0.03) and CSI (r = 0.6; P < 0.001) over time.
|
HL and StandH SDS in each treatment group are presented in Table 4. In children treated with GH before 1988, the detrimental effect of spinal irradiation on FH is clear; however, after 1988, those treated with spinal irradiation achieved similar FH outcomes as those who received only CI. The effect of CT on FH is apparent in those treated after 1988.
|
. In group A, HL was predicted by age at irradiation (P < 0.001), CSI (P = 0.02), CT (P < 0.001), and use of GnRHa (P < 0.001). The same result was achieved in the analysis of StandH SDS at FH (model not shown). However, when the FH data from group B were analyzed, CSI was no longer predictive of HL or StandH SDS at FH. HL was predicted by age at irradiation (P = 0.03) and CT (P = 0.02).
Segmental growth
Multiple linear regression models of outcome for groups A and B are presented in Table 3. SitH SDS at FH in groups A and B (excluding the effect of GnRHa therapy) can be predicted by age at irradiation, CSI, CT, GnRHa, and SitH SDS at irradiation. LL SDS at FH in group A can be predicted by ln(lag) (P = 0.005), CT (P < 0.001), GnRHa (P < 0.001), and LL SDS at irradiation (P = 0.003). However, when LL SDS at FH data in group B were analyzed, the patients who had CSI had a significant increase in LL SDS of approximately 0.8 (P = 0.01). CT (P = 0.009) and LL SDS at irradiation (P = 0.003) were also significant factors.
Disproportion at FH in both groups A and B (excluding the effect of GnRHa) can be predicted by CSI (P < 0.001) and GnRHa (P < 0.001), i.e. disproportion at FH is worsened not only by CSI but also by the use of GnRHa.
Use of GnRHa
Eleven patients received GnRHa (Tables 1 and 5). Actual and predicted HL, StandH SDS, SitH SDS, and LL SDS at FH and disproportion for patients who received GnRHa are presented in Table 6 and Fig. 2, A–E.
|
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Our current study provides FH data in the largest cohort of brain tumor survivors reported from a single center. It incorporates FH data previously reported in 1988 and 1995 (13, 19, 20). Due to the time span of the data collection we have been able to assess differences in practice over the last 25 yr. We have also devised a predictive model to determine the benefit of GnRHa in the 11 treated patients, taking into consideration other factors.
We have demonstrated improvements in FH over the last 25 yr in a cohort of 58 children treated with GH therapy for radiation-induced GHD after therapy for brain tumors. This particularly applies to the patients who received CSI. This is most likely related to earlier testing for GHD, a reduction in the lag time from completing irradiation to starting GH therapy, and improved GH regimens. We have demonstrated significant improvements in FH outcome with the use of GnRHa, particularly in patients who were treated with CI (+18 cm; +2.7 SDS). The effect of GnRHa in patients receiving CSI is more modest (+3.2 cm; +0.9 SDS), but is consistent with some benefit. These data support the use of combination GnRHa and GH in those children with radiation-induced GHD and early puberty.
Many of the auxological outcomes are affected by age at irradiation, with the youngest children having the worst prognosis (13). There are a number of potential reasons for this finding; the younger child may be more sensitive to radiation damage to the hypothalamic-pituitary axis and therefore may become more severely GHD earlier; a younger age at irradiation results in a longer period of growth to be potentially affected; and spinal irradiation in a younger child results in greater deficit in SitH at FH (21, 22). From the aspect of clinical management, however, age at irradiation was also correlated with a longer period of time until first assessment of GH status and consequently a longer time until GH therapy was started. Those children irradiated in the latter time period of the study were first assessed for GHD and started on GH therapy significantly earlier after irradiation. This change in practice may in part explain the improvements in FH outcome.
Another reason for improvement in FH is the effect of the GH treatment protocol. Adan et al. (14) commented that the FH results of a previous study using extracted GH given as 0.3–0.4 IU/kg im injections 3 times weekly (15) were worse than the results achieved using daily sc injections of GH (0.5 IU/kg·wk) (14). FH gain of 9.7 cm has been observed in prepubertal children treated for 4 yr with daily sc injections of GH compared with those treated with 3 times weekly injections (23). Our patient cohort includes 19 children who were irradiated between 1965 and 1981 and received either extracted GH or 3 times weekly injections; these 19 children also had a longer lag time until GH was initiated (5.4 vs. 3.3 yr; P < 0.001) and received more CT. Therefore, although some of the improvement in FH may be accounted for by the GH regimen, the other differences between the cohorts make the detection of key contributory factors difficult.
Interestingly, our FH results in patients who did not receive GnRHa demonstrated that only CT, and not spinal irradiation, as a component of cancer therapy was a significant cause of standing HL, contrary to the findings of previous studies (13, 15, 16, 19, 24). One explanation is that with improvements in GH regimens and reduction in lag time, the effect of spinal irradiation on HL is minimized due to improvements in LL, as demonstrated by the increased LL SDS at FH (+0.8 SDS) when the CSI patients treated with GH after 1988 were analyzed separately. Confirmatory evidence of limited spinal growth in patients receiving CSI is found in the analysis of SitH SDS at FH and disproportion at FH. A younger age at spinal irradiation is known to be more deleterious to final SitH SDS (21, 22). Spinal growth exceeds LL growth during puberty, and it is during this time that spinal growth is particularly impaired, resulting in an exaggeration of disproportion at FH (25). No effect of CI with or without CT was seen on skeletal disproportion at FH.
As previously (13) we have demonstrated a clear effect of CT on growth resulting in a reduction in FH of approximately 7 cm in addition to the effect of CI or CSI alone. Cytotoxic drugs may amplify the damage to the hypothalamic-pituitary axis by irradiation (26), directly affect the production of IGF-I by the liver (27), and impair the action of IGF-I on the growth plate. It has been shown that growth in the first 4 yr after treatment with CSI is more profoundly affected in children who have received adjuvant CT than in those receiving CSI alone, suggesting potentiation of radiation-induced growth failure by CT (28).
In children with radiation-induced GHD, the onset of puberty is earlier than in the normal population, particularly in girls (10.2 vs. 11.2 yr), but also in boys (11.6 vs. 12 yr), but puberty duration in both boys (1.8 yr) and girls (2.2 yr) is similar to that seen in the normal population (1.8 and 2 yr for normal children, respectively). This has been previously reported (4, 24, 29, 30, 31). We also observed an association between onset of puberty and duration of puberty; for example, those who enter puberty earlier have an extended duration of puberty, this has not previously been reported in brain tumor survivors. A similar pattern is seen in healthy children who enter puberty early (32).
Eleven patients in our study were treated with GnRHa, 3 of whom had precocious puberty, presumably secondary to a hypothalamic lesion, with an average onset of puberty of 4.8 yr, all had received CI. These 3 patients remained in the study cohort, as repeated endocrine assessment suggested that their GHD was radiation induced and not due to tumor mass per se. The remaining 8 had early-onset puberty, with an average onset of puberty of 8.8 yr. By using a predictive model created from the 28 subjects who were treated after 1988 with GH therapy but did not receive GnRHa therapy, improvements were demonstrated in HL, StandH SDS, SitH SDS, and LL SDS at FH, compared with the predicted values, in patients who received CI. In the 6 patients who received CSI, gain was seen in StandH SDS at FH, but HL reduction failed to reach significance. The magnitude of the gain was +18 cm and +2.7 SDS in patients who received CI and +3.2 cm and +0.9 SDS in patients receiving CSI. The effect of GnRHa in children with precocious puberty without organic pathology is clear, with the results from numerous studies demonstrating an improvement in FH compared with untreated patients (33). However, there is little information about its impact on FH if used from early puberty onward. Three studies have investigated the use of GnRHa to delay puberty in GHD children and have reported that FH outcome is improved nearly to target height (34, 35, 36). One previous study has looked at the use of GnRHa for the treatment of early puberty in cancer survivors treated with GH for radiation-induced GHD (14). They reported improvements in HL and SitH SDS in 21 cancer survivors who had received CI with GnRHa compared with a group of survivors who did not receive GnRHa. The groups, however, were not matched for diagnosis, age at irradiation, dose of irradiation, or use of chemotherapy, and curiously, the investigators actually found that the age at irradiation was younger in those entering puberty at a normal age compared with those entering puberty earlier (14).
In conclusion, improved GH replacement schedules and the increasing use of GnRHa have contributed to better auxological outcomes in children with radiation-induced GHD. In addition, safety data have allowed an earlier and bolder approach to cancer survivors. This has resulted in a reduction in time from completing irradiation to starting GH therapy, with an improvement in FH. At retest, 80–90% of these teenagers are likely to prove to be severely GHD and therefore potentially they will benefit from GH replacement in adult life (37).
|
Footnotes |
|---|
Received March 3, 2003.
Accepted May 8, 2003.
|
References |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  S. J. Laughton, T. E. Merchant, C. A. Sklar, L. E. Kun, M. Fouladi, A. Broniscer, E. B. Morris, R. P. Sanders, M. J. Krasin, J. Shelso, et al. Endocrine Outcomes for Children With Embryonal Brain Tumors After Risk-Adapted Craniospinal and Conformal Primary-Site Irradiation and High-Dose Chemotherapy With Stem-Cell Rescue on the SJMB-96 Trial J. Clin. Oncol., March 1, 2008; 26(7): 1112 - 1118. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. Ergun-Longmire, A. C. Mertens, P. Mitby, J. Qin, G. Heller, W. Shi, Y. Yasui, L. L. Robison, and C. A. Sklar Growth Hormone Treatment and Risk of Second Neoplasms in the Childhood Cancer Survivor J. Clin. Endocrinol. Metab., September 1, 2006; 91(9): 3494 - 3498. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. E. Lipshultz, S. A. Vlach, S. R. Lipsitz, S. E. Sallan, M. L. Schwartz, and S. D. Colan Cardiac Changes Associated With Growth Hormone Therapy Among Children Treated With Anthracyclines Pediatrics, June 1, 2005; 115(6): 1613 - 1622. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. H. Darzy, S. S. Pezzoli, M. O. Thorner, and S. M. Shalet The Dynamics of Growth Hormone (GH) Secretion in Adult Cancer Survivors with Severe GH Deficiency Acquired after Brain Irradiation in Childhood for Nonpituitary Brain Tumors: Evidence for Preserved Pulsatility and Diurnal Variation with Increased Secretory Disorderliness J. Clin. Endocrinol. Metab., May 1, 2005; 90(5): 2794 - 2803. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H K Gleeson and S M Shalet The impact of cancer therapy on the endocrine system in survivors of childhood brain tumours Endocr. Relat. Cancer, December 1, 2004; 11(4): 589 - 602. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. E. Lerner, G. J. M. Huang, D. McMahon, C. A. Sklar, and S. E. Oberfield Growth Hormone Therapy in Children after Cranial/Craniospinal Radiation Therapy: Sexually Dimorphic Outcomes J. Clin. Endocrinol. Metab., December 1, 2004; 89(12): 6100 - 6104. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. J. Kim, A. Janss, D. Shanis, S. Homan, and T. Moshang Jr. Adult Heights Attained by Children with Hypothalamic/Chiasmatic Glioma Treated with Growth Hormone J. Clin. Endocrinol. Metab., October 1, 2004; 89(10): 4999 - 5002. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. M. Brownstein, A. C. Mertens, P. A. Mitby, M. Stovall, J. Qin, G. Heller, L. L. Robison, and C. A. Sklar Factors That Affect Final Height and Change in Height Standard Deviation Scores in Survivors of Childhood Cancer Treated with Growth Hormone: A Report from the Childhood Cancer Survivor Study J. Clin. Endocrinol. Metab., September 1, 2004; 89(9): 4422 - 4427. [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 |
, CSI.