The Journal of Clinical Endocrinology & Metabolism Vol. 89, No. 12 6100-6104
Copyright © 2004 by The Endocrine Society
Growth Hormone Therapy in Children after Cranial/Craniospinal Radiation Therapy: Sexually Dimorphic Outcomes
Shulamit E. Lerner,
Guo Jun Michael Huang,
Donald McMahon,
Charles A. Sklar and
Sharon E. Oberfield
Division of Pediatric Endocrinology (S.E.L., G.J.M.H., S.E.O.), Childrens Hospital of New York-Presbyterian, New York, New York 10032; Division of Endocrinology (D.M.), New York Presbyterian Hospital, New York, New York 10032; and Division of Pediatric Endocrinology (C.A.S.), Memorial Sloan Kettering Cancer Center, New York, New York 10021
Address all correspondence and requests for reprints to: Dr. Sharon E. Oberfield, Division of Pediatric Endocrinology, Childrens Hospital of New York-Presbyterian, 630 West 168th Street, PH-5E-522, New York, New York 10032. E-mail: seo8{at}columbia.edu.
 |
Abstract
|
|---|
Radiation therapy (RT) to the craniospinal region in childhood affects final height. The use of GH treatment (GHRx) in children after cranial or craniospinal RT results in variable improvement in final height. Nineteen children (12 males and 7 females) with tumors of the head, treated with cranial or craniospinal RT and subsequently with GHRx, were assessed for final height. Two outcome measures of efficacy of GHRx were used: Y1 = final height SD score (SDS) corrected for genetic potential, using midparental sex-adjusted target height (SATH) SDS, and Y2 = change in height SDS from predicted final height SDS pre-GHRx to actual final height SDS post-GHRx. The median age at diagnosis was 5.4 yr, the median RT to the hypothalamic-pituitary axis was 40 Gy, the median spinal RT dose in 13 of 19 of the subjects treated was 36 Gy, and the median years post-RT to GHRx was 4.8 yr. Adjuvant chemotherapy was used in 12 of 19 patients. All but one (optic glioma) had a lesion anatomically distant from the suprasellar region. The effects of age at diagnosis, sex, L-T4 or GnRH agonist use, conventional vs. hyperfractionated RT, spinal RT, dose of spinal or cranial RT, chemotherapy, peak stimulated GH, dose and duration of GHRx, age at GHRx, time interval between RT and GHRx initiation, bone age, and height SDS at the start of GHRx were also assessed. Y1girls best correlated with younger age at diagnosis and im vs. sc GHRx. Y2girls best correlated with delayed bone age and younger age at diagnosis [Y1girls = 9.95 + 0.38 (age in years at diagnosis) + 3.11[GH method (1 = im; 2 = sc)]; r2 = 0.898; P = 0.02; Y2girls = 3.54 + 1.8 (bone age age in years) + 0.334 (age at diagnosis in years); r2= 0.956; P = 0.02]. Both Y1boys and Y2boys were strongly associated with spinal RT and younger age at diagnosis or treatment [Y1boys = 11.22 + 4.65 [spinal RT (1 = yes; 2 = no)] + 0.396 (age in years at diagnosis); r2= 0.64, P = 0.01; Y2boys = 6.32 + 0.23 (age in years at GH start) + 1.75 [spinal RT (1 = yes; 2 = no)]; r2= 0.646; P < 0.01]. This small historical cohort underscores that final stature is significantly reduced when immature bones are exposed to ionizing radiation. Intramuscular vs. sc use of GHRx is likely to be simply a surrogate marker for earlier methods of treatment. Of note, spinal RT did not significantly impact girls final heights, whereas in boys, spinal RT strongly predicted ultimate short stature and a reduced response to GHRx. This sexually dichotomous response may be due in part to the greater percentage of spinal growth remaining for boys vs. girls throughout childhood.
 |
Introduction
|
|---|
FINAL HEIGHT IN survivors of childhood head tumors is often less than expected based on midparental sex-adjusted target height (SATH). This reduction in final height is attributed to a variety of factors, including the impact of serious illness and poor nutrition, the direct impingement of the tumor on the hypothalamus or pituitary gland, and the effects of treatment of the tumor (1). Furthermore, GH deficiency (GHD) may occur as the result of a tumors direct effect on hypothalamic and pituitary structures or as the result of surgical or radiation treatment (2). Radiation therapy is an important component in the treatment of many tumors of the head, for primary tumors that may not be easily reached through conventional surgery, or for postsurgical ablation of residual tumor. Craniospinal radiation is included in therapeutic protocols when there is concern about metastasis throughout the central nervous system, particularly to the spinal column (3).
Although exogenous GH is frequently used to treat the resultant short stature, these patients are recognized to have great variability in their response to GH therapy (GHRx) (4, 5, 6, 7, 8, 9, 10). It is not clear which factors in a patients history and clinical course cause this variability. Therefore, we studied a historical cohort to ascertain the predictive factors prior to GHRx that may have influenced these patients post-GHRx final height.
 |
Subjects and Methods
|
|---|
A chart review was performed on patients with a history of tumor of the head, who were treated with cranial or craniospinal radiation therapy and subsequently were treated with GH. This review was approved by the institutional review board of New York-Presbyterian Medical Center. Patients excluded were those who died, who had recurrence of their brain tumor before completion of growth, who were not yet at final height, or who failed to return for follow-up.
Two outcome measures were used for the efficacy of GHRx. First, to assess final height relative to genetic influence, final height for midparental SATH was measured as an SD score (SDS). This first outcome was termed Y1. Midparental SATH was computed as: [maternal height (centimeters) + paternal height (centimeters)]/2 + 13 cm (male) or 13 cm (female), where absolute final height (centimeters) SDS was computed as a multiple of ±5 cm/SD for SATH (centimeters) (11). Second, to evaluate the change in SDS for height over the course of GHRx, we calculated any change by subtracting the final height SDS post-GHRx from the predicted height SDS pre-GHRx. This second outcome was termed Y2. Predicted height was calculated with the Adult Height Predictor (Genentech, Inc., South San Francisco, CA), using bone age, relation of bone age to chronological age, and present height (12, 13); final height SDS was calculated with the Palm Systems Height Predictor (Eli Lilly & Co., Indianapolis, IN) (14).
Statistics
Assessment of the impact of the following variables was performed for outcomes Y1 and Y2: sex, type of tumor diagnosed, weight at start of GH treatment (kilograms), height at start of GH treatment (centimeters), height at start of GH treatment (SDS), age at diagnosis of cancer (years), radiation dose during treatment (Gray), spinal radiation exposure during radiation treatment (yes or no), years treated with radiation, treatment with chemotherapy (yes or no), age at start of GH treatment (years), bone age at start of GH treatment (years), age at start of GH treatment minus bone age (years), GH formulation (im or sc), GH deficiency (yes or no), peak GH response to stimulation test (nanograms per milliliter), frequency of GH administration, initial GH dose (milligrams per kilogram per week), and gonadotropin agonist therapy or synthetic thyroid hormone replacement (yes or no). Multiple regression with stepwise selection of predictors meeting a value of P < 0.15 for entry into the model and a value of P < 0.15 for staying in the model was used to generate preliminary models. For each model generated, colinearity analysis among predictors, partial correlations, semipartial correlations, Akaikes information criterion, and stepwise improvement in total model r were examined. Models with predictors with least colinearity, highest partial and semipartial correlations, and lowest Akaikes information criterion were selected. Given the small sample size, subsequent models were restricted to the two best predictors.
 |
Results
|
|---|
The clinical parameters of the 19 patients who met inclusion criteria and were studied are shown in Table 1
. Table 2
summarizes initial and final height parameters. Both tables show parameters for girls and boys separately. Final outcomes were measured as the SDS for Y1, the difference of final height from midparental SATH, and Y2, the change from predicted height SDS to final height SDS.
For girls, Y1 best correlated with younger age at diagnosis and im vs. sc GHRx [Y1girls = 9.95 + 0.38 (age in years at diagnosis) + 3.11(GH method (1 = im; 2 = sc)]; r2 = 0.898; P = 0.02; Fig. 1A
]. Y2 best correlated with delayed bone age and younger age at diagnosis [Y2girls = 3.54 + 1.8 (bone age age in years) + 0.334 (age at diagnosis in years); r2= 0.956; P = 0.02; Fig. 1B
]. For boys, spinal RT and younger age at diagnosis or treatment were associated with Y1 [Y1boys = 11.22 + 4.65 [spinal RT (1 = yes; 2 = no)] + 0.396 (age in years at diagnosis); r2= 0.64; P = 0.01; Fig. 2A
], and Y2 [Y2boys = 6.32 + 0.23 (age in years at GH start) + 1.75 [spinal RT (1 = yes; 2 = no)]; r2= 0.646; P < 0.01; Fig. 2B
].
 |
Discussion
|
|---|
Survivors of childhood tumors of the head who have received RT are known to have decreased final height. The degree to which final height is affected is variable, as is the response to GHRx, which is used to improve the resultant short stature. We therefore analyzed a cohort of survivors of tumors who were treated with cranial or craniospinal RT and were subsequently treated with GH to determine which treatment and demographic parameters had the greatest predictive value for growth response to GHRx.
GHD or a defect in GH secretion will diminish the potential for growth. All patients in our cohort either tested as GHD or were presumed to have a secretory defect after receiving high doses of cranial irradiation, with final height predictions significantly shorter than their predicted midparental SATHs, which account for height potential based on genetic factors (11). These findings are consistent with other studies demonstrating GHD after cranial RT of 1825 Gy or more (1, 15, 16, 17, 18) within 12 yr after RT (19, 20), due to direct injury to hypothalamic neurons rather than damage to the pituitary itself (15, 20). In addition, a pattern of low growth velocity, despite a stimulated GH level of 6 ng/ml or more, has been suggested to signify a neurosecretory defect within the hypothalamus (21), particularly because IGF-I levels do not consistently discriminate between subjects with and without GHD (15).
Previous studies have attempted to correlate patient characteristics and treatment parameters with final height outcome. Chemotherapy has been associated with a variable risk of short stature (4, 22, 23, 24, 25, 26). Puberty has been noted to take place at a significantly younger age after RT, but does not significantly affect final height (4, 27), whereas delayed puberty has also been associated with decreased upper segment to lower segment ratio, presumably due to decreased sex steroid hormone production (5). Recent studies suggest that in the case of early puberty, the use of GnRH agonist therapy alone or in combination with GHRx may result in an improvement in final height in this population or in other cases of GHD (6, 28, 29, 30, 31).
Other parameters known to influence final height outcome in patients who have received GHRx include the patients genetically determined height potential (32), GHRx dose and frequency (22), and other concomitant hormone insufficiencies, such as hypothyroidism (33). Additionally, younger age at the time of RT has consistently been associated with a strong risk for short adult height (1, 25, 26, 34), as has treatment with RT during rapid pubertal growth (35). Similar to these studies, we showed in both boys and girls that irradiation had a significantly negative impact on final height with younger age. The additional negative effect of a delayed bone age seen in the girls cohort likely represents a biological surrogate marker for the risk factor of younger age. For girls, im administration of GH also significantly decreased the salutary effect of GHRx on final height. The reduced effect of semiweekly im vs. daily sc administered GHRx has previously been noted (36).
More than any other factor, however, irradiation of the spine has been implicated in a high risk of short adult height. This is probably due to damaged growth plates after high doses of radiation (37). Patients treated with spinal RT are consistently shorter than those treated with cranial RT alone (1, 7, 19, 25, 26, 35), with reduced response to GHRx or normal endogenous GH levels after spinal RT (8). The upper segment to lower segment ratio is decreased in all reviewed studies (5, 21, 38, 39, 40), with a greater negative effect on growth of spinal RT vs. cranial RT despite an overall final height improvement with GHRx (9, 38, 41, 42), whereas both final and sitting heights in children treated with cranial RT have been shown to improve with GHRx (10, 42).
The striking negative effect of spinal RT for boys, but not for girls, may be due to the difference between boys and girls in potential spinal growth. Growth curves of sitting height (43, 44) demonstrate that the percentage of height remaining is greater in boys than in girls at every age until final height and may be used to predict final height after RT (45). The median ages at the time of irradiation of our male and female cohorts were 5.3 and 5.5 yr, respectively. Using growth charts of sitting height, the median potential of spinal growth remaining at the time of RT was 31.3 cm for boys and 24.6 cm for girls, a difference of 6.7 cm. We therefore suggest that although this study included only a small and heterogeneous group of patients, sex-dependent growth may account for some of the significant difference between the responses of boys and girls to GHRx after spinal RT.
Conclusions
This small historical cohort underscores that radiation damage to less mature spines, as evidenced by younger age and delayed bone age, is highly correlated with greater loss of statural growth. We observed an effect of spinal irradiation on final height in our cohort of boys, but not in our cohort of girls. We suggest that this may reflect a significantly greater percentage of growth remaining for boys vs. girls of similar age until achievement of final height. We therefore support prior conclusions that a major contributing factor to loss of height despite GHRx is direct damage to growing bones. Furthermore, although we observed a sexually dichotomous response to treatment, we acknowledge that our groups had considerable heterogeneity, including different methods of radiation treatment. We therefore suggest that similar analyses of larger cohorts of similar patients are needed to verify the findings we observed in our small historical cohort.
 |
Footnotes
|
|---|
Abbreviations: GHD, GH deficiency; GHRx, GH treatment; RT, radiation therapy; SATH, sex-adjusted target height; SDS, SD score.
Received July 30, 2004.
Accepted September 23, 2004.
 |
References
|
|---|
- Oberfield SE, Sklar CA 2002 Endocrine sequelae in survivors of childhood cancer. Adolesc Med 13:161169[Medline]
- Merchant TE, Williams T, Smith JM, Rose SR, Danish RK, Burghen GA, Kun LE, Lustig RH 2002 Preirradiation endocrinopathies in pediatric brain tumor patients determined by dynamic tests of endocrine function. Int J Radiat Oncol Biol Phys 54:4550[Medline]
- Donahue B 1992 Short- and long-term complications of radiation therapy for pediatric brain tumors. Pediatr Neurosurg 18:207217[Medline]
- Ogilvy-Stuart AL, Shalet SM 1995 Growth and puberty after GH treatment after irradiation for brain tumours. Arch Dis Child 73:141146[Abstract]
- Burns EC, Tanner JM, Preece MA, Cameron N 1981 Growth hormone treatment in children with craniopharyngioma: final growth status. Clin Endocrinol (Oxf) 14:587595[Medline]
- Adan L, Sainte-Rose C, Souberbielle JC, Zucker JM, Kalifa C, Brauner R 2000 Adult height after growth hormone (GH) treatment for GH deficiency due to cranial irradiation. Med Pediatr Oncol 34:1419[CrossRef][Medline]
- Gleeson HA, Stoeter R, Ogilvy-Stuart AL, Gattamaneni HR, Brennan BM, Shalet SM 2003 Improvements in final height over 25 years in growth hormone (GH)-deficient childhood survivors of brain tumors receiving GH replacement. J Clin Endocrinol Metab 88:36823689[Abstract/Free Full Text]
- Sulmont V, Brauner R, Fontoura M, Rappaport R 1990 Response to growth hormone treatment and final height after cranial or craniospinal irradiation. Acta Paediatr Scand 79:542549[Medline]
- Clayton PE, Shalet SM, Price DA 1988 Growth response to growth hormone therapy following craniospinal irradiation. Eur J Pediatr 147:597601[CrossRef][Medline]
- Clayton PE, Shalet SM, Price DA 1988 Growth response to growth hormone therapy following cranial irradiation. Eur J Pediatr 147:593596[CrossRef][Medline]
- Luo ZC, Albertsson-Wikland K, Karlberg J 1998 Target height as predicted by parental heights in a population-based study. Pediatr Res 44:563571[Medline]
- Post E, Richman R 1981 A condensed table for predicting adult stature. J Pediatr 98:440[CrossRef][Medline]
- Bayley N, Pinneau SR 1952 Tables for predicting adult height from skeletal age: revised for use with the Greulich-Pyle hand standards. J Pediatr 40:423431[CrossRef][Medline]
- Kuczmarski RJ, Ogden CL, Grummer-Strawn LM, Flegal KM, Guo SS, Wei R, Mei Z, Curtin LR, Roche AF, Johnson CL 2000 CDC growth charts: United States. Adv Data 314:127[Medline]
- Popovic V, Pekic S, Golubicic I, Doknic M, Dieguez C, Casanueva FF 2002 The impact of cranial irradiation on GH responsiveness to GHRH plus GH-releasing peptide-6. J Clin Endocrinol Metab 87:20952099[Abstract/Free Full Text]
- Adan L, Trivin C, Sainte-Rose C, Zucker JM, Hartmann O, Brauner R 2001 GH deficiency caused by cranial irradiation during childhood: factors and markers in young adults. J Clin Endocrinol Metab 86:52455251[Abstract/Free Full Text]
- Brennan BM, Rahim A, Mackie EM, Eden OB, Shalet SM 1998 Growth hormone status in adults treated for acute lymphoblastic leukaemia in childhood. Clin Endocrinol (Oxf) 48:777783[CrossRef][Medline]
- Costin G 1988 Effects of low-dose cranial radiation on growth hormone secretory dynamics and hypothalamic-pituitary function. Am J Dis Child 142:847852[Abstract]
- Brauner R, Rappaport R, Prevot C, Czernichow P, Zucker JM, Bataini P, Lemerle J, Sarrazin D, Guyda HJ 1989 A prospective study of the development of growth hormone deficiency in children given cranial irradiation, and its relation to statural growth. J Clin Endocrinol Metab 68:346351[Abstract]
- Merchant TE, Goloubeva O, Pritchard DL, Gaber MW, Xiong X, Danish RK, Lustig RH 2002 Radiation dose-volume effects on growth hormone secretion. Int J Radiat Oncol Biol Phys 52:12641270[CrossRef][Medline]
- Oberfield SE, Allen JC, Pollack J, New MI, Levine LS 1986 Long-term endocrine sequelae after treatment of medulloblastoma: prospective study of growth and thyroid function. J Pediatr 108:219223[CrossRef][Medline]
- Kanev PM, Lefebvre JF, Mauseth RS, Berger MS 1991 Growth hormone deficiency following radiation therapy of primary brain tumors in children. J Neurosurg 74:743748[CrossRef][Medline]
- Gurney JG, Kadan-Lottick NS, Packer RJ, Neglia JP, Sklar CA, Punyko JA, Stovall M, Yasui Y, Nicholson HS, Wolden S, McNeil DE, Mertens AC, Robision LL 2003 Endocrine and cardiovascular late effects among adult survivors of childhood brain tumors. Cancer 97:663673[CrossRef][Medline]
- Olshan JS, Gubernick J, Packer RJ, DAngio GJ, Goldwein JW, Willi SM, Moshang Jr T 1992 The effects of adjuvant chemotherapy on growth in children with medulloblastoma. Cancer 70:20132017[CrossRef][Medline]
- Noorda EM, Somers R, van Leeuwen FE, Vulsma T, Behrendt H 2001 Adult height and age at menarche in childhood cancer survivors. Eur J Cancer 37:605612
- Schriock EA, Schell MJ, Carter M, Hustu O, Ochs JJ 1991 Abnormal growth patterns and adult short stature in 115 long-term survivors of childhood leukemia. J Clin Oncol 9:400405[Abstract]
- Oberfield SE, Soranno D, Nirenberg A, Heller G, Allen JC, David R, Levine LS, Sklar CA 1996 Age at onset of puberty following high-dose central nervous system radiation therapy. Arch Pediatr Adolesc Med 150:589592[Abstract]
- Saggese G, Federico G, Barsanti S, Fiore L 2001 The effect of administering gonadotropin-releasing hormone agonist with recombinant-human growth hormone (GH) on the final height of girls with isolated GH deficiency: results from a controlled study. J Clin Endocrinol Metab 86:19001904[Abstract/Free Full Text]
- Mericq MV, Eggers M, Avila A, Cutler Jr GB, Cassorla F 2000 Near final height in pubertal growth hormone (GH)-deficient patients treated with GH alone or in combination with luteinizing hormone-releasing hormone analog: results of a prospective, randomized trial. J Clin Endocrinol Metab 85:569573[Abstract/Free Full Text]
- Job JC, Toublanc JE, Landier F 1994 Growth of short normal children in puberty treated for 3 years with growth hormone alone or in association with gonadotropin-releasing hormone agonist. Horm Res 41:177184[Medline]
- Marx M, Schoof E, Grabenbauer GG, Beck JD, Doerr HG 1999 Effects of puberty on bone age maturation in a girl after medulloblastoma therapy. J Pediatr Adolesc Gynecol 12:6266[CrossRef][Medline]
- Rochiccioli P, David M, Malpuech G, Colle M, Limal JM, Battin J, Mariani R, Sultan C, Nivelon JL, Simonin G 1994 Study of final height in Turners syndrome: ethnic and genetic influences. Acta Paediatr 83:305308[Medline]
- Nanto-Salonen K, Muller HL, Hoffman AR, Vu TH, Rosenfeld RG 1993 Mechanisms of thyroid hormone action on the insulin-like growth factor system: all thyroid hormone effects are not growth hormone mediated. Endocrinology 132:781788[Abstract]
- Helseth E, Due-Tonnessen B, Wesenberg F, Lote K, Lundar T 1999 Posterior fossa medulloblastoma in children and young adults (019 years): survival and performance. Childs Nerv Syst 15:451456[CrossRef][Medline]
- Probert JC, Parker BR, Kaplan HS 1973 Growth retardation in children after megavoltage irradiation of the spine. Cancer 32:634639[CrossRef][Medline]
- Kastrup KW, Christiansen JS, Andersen JK, Orskov H 1983 Increased growth rate following transfer to daily scadministration from three weekly im injections of hGH in growth hormone deficient children. Acta Endocrinol (Copenh) 104:148152[Medline]
- Eifel PJ, Donaldson SS, Thomas PRM 1995 Response to growing bone to irradiation: a proposed late effects scoring system. Int J Radiat Oncol Biol Phys 31:13011307[CrossRef][Medline]
- Shalet SM, Gibson B, Swindell R, Pearson D 1987 Effect of spinal irradiation on growth. Arch Dis Child 62:461464[Abstract]
- Chin HW, Maruyama Y 1984 Age at treatment and long-term performance results in medulloblastoma. Cancer 53:19521958[CrossRef][Medline]
- Clayton PE, Shalet SM 1991 The evolution of spinal growth after irradiation. Clin Oncol 3:220222
- Darendeliler F, Livesey EA, Hindmarsh PC, Brook CG 1990 Growth and growth hormone secretion in children following treatment of brain tumours with radiotherapy. Acta Paediatr Scand 79:950956[Medline]
- Xu W, Janss A, Moshang T 2003 Adult height and adult sitting height in childhood medulloblastoma survivors. J Clin Endocrinol Metab 88:46774681[Abstract/Free Full Text]
- Katz JR, Bareille P, Levitt G, Stanhope R 2001 Growth hormone and segmental growth in survivors of head and neck embryonal rhabdomyosarcoma. Arch Dis Child 84:436439[Abstract/Free Full Text]
- Dangour AD, Schilg S, Hulse JA, Cole TJ 2002 Sitting height and subischial leg length centile curves for boys and girls from southeast England. Ann Hum Biol 29:290305[CrossRef][Medline]
- Silber JH, Littman PS, Meadows AT 1990 Stature loss following skeletal irradiation for childhood cancer. J Clin Oncol 8:304312[Abstract]