Long-term safety of growth hormone replacement after CNS irradiation

CONTEXT

Radiotherapy is a central component in the treatment of many brain tumors, but long-term sequelae include GH deficiency and increased risk of secondary neoplasms. It is unclear whether replacement therapy with GH (GHRT) further increases this risk.

OBJECTIVE

The objective of the study was to assess the effect of GHRT on the incidence of secondary tumors and tumor recurrence after cranial irradiation.

DESIGN AND SETTING

We conducted a retrospective matched-pairs analysis of previously irradiated patients, with and without GHRT, attending a tertiary center between 1994 and 2009.

PATIENTS

We reviewed the records for all patients undergoing GHRT at our institution over the study period. PATIENTS were included if they had received cranial irradiation, GHRT for at least 12 months, and records of serial magnetic resonance imaging data and data for dose and fractionation of irradiation were available. GH-naïve control patients were selected from a radiotherapy database of patients attending the same hospital. PATIENTS were matched for date of radiotherapy, age, site of primary diagnosis, radiation dose, and fractionation.

MAIN OUTCOME MEASURE

The primary outcome measure was risk of tumor recurrence or secondary tumor.

RESULTS

Matched controls were identified for 110 GH-treated patients. Median follow-up was 14.5 yr. No significant differences were apparent in the number of tumor recurrences (six vs. eight, GHRT vs. control group) or secondary tumors (five vs. three, respectively) between groups.

CONCLUSIONS

Our study demonstrates no increased risk for recurrent or secondary neoplasms in patients receiving GHRT, thus supporting a high safety profile of GHRT after central nervous system irradiation.

Gonadal dysfunction due to cis-platinum

Gonadal function was studied in 15 patients 12 pubertal or postpubertal, and three prepubertal, who had been treated during childhood for nonmetastatic osteosarcoma of the long bones by chemotherapy regimens that included cis-platinum and adriamycin. Of seven postpubertal female patients assessed (mean age at diagnosis 16.5 years), three were amenorrhoeic and showed evidence of ovarian damage with raised gonadotrophin levels and a low serum oestradiol concentration. One patient who had regular periods had a raised luteal-phase follicle-stimulating hormone (FSH) concentration suggestive of gonadal dysfunction. Severe oligospermia or reduced testicular volumes in the presence of raised gonadotrophin levels were observed in three of the five pubertal males (mean age at diagnosis 13.25 years). A reliable assessment of gonadal function was not possible in three male patients who remained prepubertal at the time of study. The median total dose of cis-platinum received by those patients with gonadal damage (median dose, 490 mg) was significantly higher than in those patients with normal gonadal function (median dose, 300 mg) (P = 0.01). In the boys the damage to the testes was primarily directed at the germinal epithelium. Leydig cell function was intact and the males progressed spontaneously through puberty. In the girls, unlike the boys, there was evidence of reversibility of gonadal damage with time. This is the first study to show gonadal dysfunction due to cis-platinum and adriamycin therapy in childhood.

Symptom interval in young people with bone cancer

Symptom interval (SI), the time from first symptom/sign to diagnosis and initiation of treatment, appears to be principally influenced by tumour biology. Whether the age of the patient, patient delay, professional delay and access to health professionals influences the SI in bone tumours was investigated in this study. 115 patients with newly diagnosed osteosarcoma and Ewing's sarcoma were retrospectively reviewed. The median total SI for all bone tumours was 3.8 months (range 1-46 months). Patients older than 12 years had a longer SI (P = 0.05) and more patient delays (P = 0.02). Total SI and professional delays were longer if the General Practitioner was first seen compared with an Accident and Emergency Consultant (P = 0.02 and 0.02, respectively). However, SI did not influence overall and event-free survival in this series. Bone tumour patients have long SIs that are significantly affected by age and local health-care support systems. Early referral to specialists would help to alleviate anxiety and distress to the patient and family, even if currently delay does not influence outcome.

Reassessment of growth hormone status is required at final height in children treated with growth hormone replacement after radiation therapy

The most appropriate way to manage GH replacement in the transition period to adulthood in children treated with GH for GH deficiency (GHD) is controversial. The Growth Hormone Research Society suggests that the retesting of GH status at final height (FH) is unnecessary in the presence of severe organic GHD, and cranial irradiation falls into this etiological category. This recommendation has never been validated. To investigate whether patients diagnosed in childhood as GHD secondary to irradiation require retesting after FH, GH status has been reassessed in a large cohort of irradiated children treated with GH during childhood. Seventy-three children underwent biochemical assessment of GH status after irradiation and again at FH after GH therapy had been discontinued; 66 and 67 of the 73 patients underwent two provocative tests at the two time points, respectively. The characteristics of the cohort include a median age at irradiation of 5 yr (range, 1-11 yr), a median biological effective dose (BED) of irradiation to the hypothalamic pituitary axis of 54 Gy (range, 23-82 Gy), and a median time of GH status reassessment after FH of 0.4 yr (range, 0-8.4 yr). During childhood, patients with all degrees of GHD (peak GH responses to provocative test < 6.7 ng/ml) are treated, whereas in adulthood, only patients with severe GHD (peak GH responses to provocative test < 3 ng/ml) are considered for GH replacement. GH status has been grouped as follows: group 1, peak GH less than 3 ng/ml to both tests (severe GHD); group 2, one test with a peak GH less than 3 ng/ml and the other test with a peak of 3 ng/ml or greater; group 3, peak GH of 3-6.7 ng/ml to both tests; group 4, one test with a peak GH of 3-6.7 ng/ml and the other test with a peak of more than 6.7 ng/ml; and group 5, peak GH more than 6.7 ng/ml to both tests (normal GH status). In childhood, the number of patients in groups 1, 2, 3, and 4 were 33, 22, 17, and one, respectively. At retesting, severe GHD was diagnosed in 21 (64%) of 33 patients who were diagnosed in childhood with severe GHD (group 1) and 17 (44%) of 39 patients who were diagnosed in childhood with moderate GHD (groups 2 and 3). In total, 35 (48%) of 73 patients in the whole cohort and 12 (36%) of 33 patients with severe GHD in childhood did not fulfill the severe GHD biochemical criteria for GH replacement in adulthood. Using multiple linear regression, GH status at retesting is predicted by BED, age at irradiation, and use of chemotherapy. In conclusion, the diagnosis of severe GHD in childhood secondary to irradiation should not be taken as irrefutable evidence of permanent severe organic GHD, and our recommendation is that retesting of GH status at FH should be mandatory.

Improvements in final height over 25 years in growth hormone (GH)-deficient childhood survivors of brain tumors receiving GH replacement

Final height (FH) outcome is important in survivors of childhood brain tumors. GH replacement is indicated in those found to be GH deficient (GHD). More recently, GnRH analogs (GnRHa) have been introduced to delay early or rapidly progressing puberty to allow more time for linear growth. Studies to FH are important to determine the effectiveness of growth-promoting strategies. Our aim was to assess whether evolving endocrine strategies have improved FH outcome and to determine whether GnRHa therapy has contributed auxologically. FH data were examined in 58 children (31 males and 27 females) with radiation-induced GHD who had been treated with GH. All had received a combination of cranial (CI; n = 17) or craniospinal (CSI; n = 41) irradiation with or without chemotherapy for a brain tumor. Eleven patients received GnRHa therapy. Throughout the 25 yr of the study patients came closer to achieving target height (i.e. a reduction in height loss), both those receiving CI (r = 0.5; P = 0.03) and those receiving CSI (r = 0.6; P < 0.001). The patients receiving GH therapy before 1988 compared with from 1988 onward had a similar age at irradiation [mean (+/-SD), 5.8 (3.0) vs. 6.2 (2.9) yr; P = 0.6], but experienced a more prolonged time interval from completing irradiation to starting GH [5.4 (2.4) vs. 3.3 (1.6) yr; P < 0.001]. Forward stepwise regression analysis revealed that height loss is affected by age at irradiation (P < 0.001), previous spinal irradiation (P = 0.02), chemotherapy (P < 0.001), and exposure to GnRHa therapy (P < 0.001). In the 11 patients treated with GnRHa therapy FH SD scores were improved compared with FH predictions calculated from a model derived from the patients not treated with GnRHa [-0.8 (1.6) vs. -2.4 (0.8) SD score; P < 0.001]. We have demonstrated an overall improvement in FH in children treated with GH for GHD after therapy for brain tumors over the last 25 yr. In the subset of children in whom the growth prognosis was adversely affected by early puberty, the combination of GnRHa and GH improved their prospects of achieving target height. The improved auxological outcome may reflect 1) the use of more standardized GH schedules and better dosing regimens, 2) a reduction in the time interval between finishing radiotherapy and receiving GH replacement, and 3) the use of GnRHa in addition to GH replacement in carefully selected patients.

Investigation of dose homogeneity in paediatric anthropomorphic phantoms for a simple total body irradiation technique

The technique for treating total body irradiation patients used at the centre involves no compensation for the inhomogeneity of patient shape. Dose is prescribed to the lung, and monitor units are derived from standard data depending on the external dimensions of the patient at nipple level. Dose measurements were made during standard treatments on three paediatric anthropomorphic phantoms representing children of 5, 10 and 15 years of age. The results confirmed that the measured dose to the lung was within 4% of the prescribed dose, and dose homogeneity was within +/- 5%, excluding the neck, where the higher measured doses were still within tissue tolerance.

Ifosfamide/etoposide alternating with high-dose methotrexate: evaluation of a chemotherapy regimen for poor-risk osteosarcoma

Fifteen patients with relapsed osteosarcoma were treated with an intensive combination chemotherapy schedule. Ifosfamide 2.5 g m(-2) daily and etoposide 150 mg m(-2) daily coincidentally for 3 days and high-dose methotrexate 8 g m(-2) (with folinic acid rescue) on days 10-14 in a planned 21 -day cycle. Feasibility, toxicity and response to this alternative combination for the treatment of relapsed osteosarcoma was assessed. There were 98 evaluable cycles for toxicity and tolerability. The majority of cycles were well tolerated. Haematological toxicity of grade 3/4 (common toxicity criteria) was seen in all courses. Renal tubular loss of electrolytes, particularly magnesium, occurred in 71% of cycles. Thirteen per cent of cycles were repeated within 21 days and 61% within 28 days. In the thirteen patients evaluable for response, a partial response rate of 31% was seen after two cycles. However, patients with stable disease continued on therapy, and an overall consequent response rate of 62% was observed. Four patients were alive with no evidence of disease at 8-74 months. Three are alive with disease (at 8-19 months). There were six deaths, all disease related. This regimen exhibits an encouraging response rate in a group of children with poor prognosis disease, with a tolerable toxicity profile.

Spinal cord compression--do we miss it?

Four children with spinal cord compression due to malignant tumours are presented. The severity of the condition was not initially recognized by parents, or the nature of the likely cause by the initial physicians. Lower limb asymmetrical weakness, clear-cut sensory levels, and marked pain indicate need for urgent imaging and exclusion of a space occupying lesion. In 1997 diagnosis of Guillain-Barré syndrome should not be made without careful prior spinal imaging.

Malignancies in Down syndrome

Down Syndrome (DS) is associated with an increased incidence of malignancies, especially leukaemias. We came across 8 DS children presenting with malignancies and having trisomy 21 as the sole cytogenetic abnormality. Of these 8 DS cases, 4 presented with acute lymphocytic leukaemia, 2 with acute myeloid leukaemia and one case each with Hodgkin's disease and Wilms' tumour. There are contradictory reports regarding the distribution of myeloid versus lymphoid malignancies in DS children and their response to therapy. The exact mechanism by which patients with DS are predisposed to develop malignancies is unclear. However, presence of the extra chromosome no. 21 is presumed to disrupt the genetic balance which increases generalized susceptibility to genetic and environmental trauma. Furthermore, an increased methotrexate toxicity observed in these patients should also be taken into consideration in designing treatment for DS children with malignancies.

Carcinoma of the thyroid in children: a 25-year experience

Over the past 25 years, 23 children with carcinoma of the thyroid have been treated at the Christie Hospital, Manchester. Twenty-one cases were well-differentiated carcinoma, and two were medullary carcinoma. They were all treated by resection, 14 with total thyroidectomy and 9 with lobectomy or subtotal thyroidectomy. Sixteen children also had surgery for nodal disease. Two children presented with lung metastases. Sixteen children received post-operative radiotherapy (4 external beam, 12 131I). Median follow-up of 67 months (range 7-233), was the same for the 21 well-differentiated carcinomas and the whole group including the two medullary carcinomas. All 21 children with well-differentiated carcinomas are alive with no evidence of progressive disease. Two relapsed after total thyroidectomy, but both were salvaged, one with external beam radiotherapy, one with 131I. One child with medullary carcinoma died with progressive disease after 43 months, the other is alive, but with slowly progressive disease 145 months after diagnosis. Ten of 14 children experienced post-operative hypocalcaemia following total thyroidectomy, in 7 cases it persisted long-term. 131I and external beam radiotherapy were both well tolerated. The long-term results of treatment of well-differentiated carcinoma of the thyroid are excellent, but there remains disagreement over the extent of treatment required. Some authors believe the condition is multifocal and requires total thyroidectomy, others argue that lobectomy or subtotal thyroidectomy avoids the possible post-operative complications of total thyroidectomy and gives equal long-term cure rates. We agree with the latter view. Although a small series cannot be conclusive, we feel that our results are consistent with this. We also believe, that for children, radiotherapy can be reserved for relapse only, as long as regular follow-up is available.

Relapse of intracranial germinoma 23 years postirradiation in a patient given growth hormone replacement

There is no clear evidence that growth hormone replacement therapy for treatment-related growth hormone deficiency in patients with childhood intracranial malignancies has a role in tumour relapse or second malignancy. A 16-year-old girl with an intracranial germinoma was treated with local radiotherapy and subsequently received growth hormone replacement therapy as an adult. Three years after starting growth hormone therapy, 23 years after her radiotherapy treatment, the patient's tumour recurred. Surveillance requirements for patients receiving growth hormone in this setting are discussed.

Survival and late effects in medulloblastoma patients treated with craniospinal irradiation under three years old

Conventional treatment of medulloblastoma has involved surgery to the primary tumour and radiotherapy to the primary site and craniospinal axis. However CNS irradiation in a young child may result in significant side effects. Thus new treatment strategies have emerged which include chemotherapy, given in order to delay radiotherapy, to enable radiation dose reduction to the primary site and craniospinal axis, or even to eliminate radiotherapy completely. Such treatments have not yet been adequately evaluated in terms of survival and late effects. We report a retrospective study of 37 patients under the age of 36 months treated with postoperative craniospinal irradiation, in which the radiation dose to the neuroaxis was below conventional dosage. The overall actuarial 10-year survival rate was 44% and the actuarial 10-year relapse tree survival rate was 54%. Both radiotherapy and chemotherapy contributed to morbidity and mortality. Tour of 16 patients who survived longer than 10 years had no hard neurological signs; all but one patient have required extra support at school. Of nine patients available for work, two have obtained employment but only one has maintained this. No young adults have married. Despite lower doses of radiation, all but 1 survivor has significant spine shortening, and all who reached final height were short. Further work is needed to complete the profile of late effects in this group, which should include the survivors own perceptions of quality of life. It is hoped that multimodality treatment and supportive care can sustain acceptable survival rates but reduce the burden of late effects.

Pediatric oncology and hematology in Manchester, England

The Manchester pediatric oncology unit is the third largest unit in the United Kingdom, with approximately 120 new referred cases per annum (10% of the U.K. total). Research activities include a gene therapy program, peripheral blood stem cell studies, the genetic epidemiology of childhood cancer, late-effects research (growth, body composition, pulmonary, quality of life), psychosocial studies, and clinical trial organization. Both the clinical oncology service and research activities involve close team coordination and collaboration with scientists both within and outside Manchester. A comprehensive pediatric hematology service is provided. The unit contains the second largest children's hemophilia service in the United Kingdom, serving 200 patients with congenital blood disorders. Twenty-five bone marrow transplants are performed each year (allogeneic, unrelated donor, autologous, and peripheral stem cell) for malignant and nonmalignant disorders. These activities are closely related to local, national, and international research groups.

An intensive multiagent chemotherapy regimen for brain tumours occurring in very young children

Standard treatment for the majority of malignant brain tumours consists of surgery and radiotherapy. This treatment has late morbidity which is accentuated in the very young child. As part of a strategy to improve quality of life and overall survival of young children with brain tumours, members of the United Kingdom Children's Cancer Study Group (UKCCSG) have piloted an intensive chemotherapy regimen which aims to avoid or delay radiotherapy following surgery. Twenty eight children with a variety of malignant brain tumours have received the regimen, which contains carboplatin, vincristine, cyclophosphamide, methotrexate, and cisplatin. The treatment is toxic, resulting in one death from infection. The bulk of the toxicity was associated with the administration of carboplatin. All but three children eventually required adjuvant radiotherapy and this was given between 1.5 and 27 months from diagnosis (median delay to radiotherapy, 12 months). Using this treatment regimen, overall survival at four years is 35% (confidence intervals 10% to 60%). While there is no evidence from this study that radiotherapy can be abandoned in the management of malignant brain tumours, its introduction may be delayed using suitable chemotherapy, thus allowing time for further CNS development. This treatment strategy has been taken forward as an international clinical trial run through the International Society for Paediatric Oncology, but using a smaller dose of carboplatin to reduce toxicity.

Synovial sarcoma of the pharynx

Synovial sarcoma is rarely seen in the head and neck region. A case of synovial sarcoma of the pharynx in a child is presented.

Survival and quality of life of paediatric intracranial germ cell tumour patients treated at the Christie Hospital, 1972-1993

From 1972 to 1993, 25 patients under 16 years old were treated at the Christie Hospital for intracranial germ cell tumours (ICGCTs). A retrospective analysis of the case notes was undertaken. The cases comprised 10 germinomas, nine non-germinomatous germ cell tumours (NGGCTs), and six cases with no histology. Ten patients had either complete or incomplete removal of the tumour. All patients received radiotherapy (20 patients received craniospinal irradiation [CSI]). Thirteen patients received chemotherapy at presentation (six platinum-based). All marker-negative pure germinomas treated with CSI survived. The actuarial 5-year survival for NGGCTs was 44%. Although CSI resulted in spine shortening, the overall effect on growth was not marked and the neuropsychologic sequelae were minimal with good overall functional results.

Late carmustine lung fibrosis. Age at treatment may influence severity and survival

Previously, we have reported in 1990 that 35% of carmustine treated patients (6 of 17) who survived childhood brain tumors died of pulmonary fibrosis between 2 and 13 years after treatment. In addition, 8 patients studied in 1989 (13 to 17 years post treatment), had physiologic and biopsy or radiologic evidence of pulmonary fibrosis. We now report 3 more years of follow-up on these patients. Between 1989 and 1992, two more patients have died of pulmonary fibrosis, giving an overall mortality of 47%. Of the eight patients who died of pulmonary fibrosis, the median age at treatment was 2.5 years, whereas the nine long-term survivors had a median age at treatment of 10 years. All five patients treated below the age of 5 years have died of lung fibrosis. Analysis by the standard survival curve method indicated that patients treated at an age less than 6 years were more likely to die than those treated at an age older than 7 years (p = 0.03). Of the nine survivors, seven were observed over 3 more years. There was a gradual decline in mean forced vital capacity from 55% predicted (range, 44 to 81) to 51% predicted (range, 41 to 72) and total lung capacity fell from 65% predicted (range, 51 to 89) to 57% predicted (range, 47 to 77).

Leukemia, lymphoma, and related disorders in families of children diagnosed with Wilms' tumor

Leukemias and lymphomas occurring in a series of families with Wilms' tumor (WT) are described. One surviving case developed a large cell anaplastic Ki-1 lymphoma at age 20 years, and 23 second- and higher degree relatives were affected. In two instances leukemia/lymphoma occurred in the context of Li-Fraumeni syndrome (LFS) and two other families showed striking clusters of unusual and early-onset malignancies. In several cases, children had genitourinary abnormalities of the type associated with the WT1 gene on chromosome 11p13. Some of these families may provide important subjects for study of WT genes in hematologic disease and lymphomas and for investigation of interaction between different tumor-suppressor genes, e.g., WT1 and other candidate WT genes, and p53.

Pubertal growth in response to testosterone replacement therapy for radiation-induced Leydig cell failure

The adolescent growth pattern of eight boys, who had puberty induced with androgen replacement therapy following radiation-induced Leydig cell failure, was studied from induction of puberty at a mean age of 13.1 years (range 11.6-14.5) to final height at mean age of 18.8 years (range 17.7-20.3). The mean gains during puberty (SD) for standing height, sitting height, and sub-ischial leg length were 18.56 cm (3.98), 10.46 cm (2.39), and 8.1 cm (2.01) respectively, which were significantly reduced compared with normal Tanner standards (P < .001). The peak velocity for each parameter occurred in the 1st year of induced puberty in contrast to the pattern in normal adolescence, although the mean peak velocity for each auxological parameter was not significantly different from the normal Tanner standards. The mean adult standing height (SD), 167.5 cm (9.88), and mean adult leg length (SD), 80.8 cm (6.19), were not significantly different from the normal Tanner standards, whereas the mean adult sitting height (SD), 86.7 cm (4.78), was shorter (P < .001). Three of the eight patients had a leg length standard deviation score less sitting height standard deviation score in excess of +2.96 suggesting the presence of significant skeletal disproportion. Seven of the eight boys reached target genetic height, though in six, the final height was below mid-parental height (P < .05). The modest loss in height potential was mainly due to radiation-induced skeletal dysplasia attenuating the growth of the spine. The families of boys with radiation-induced Leydig cell failure requiring androgen replacement therapy can be reasonably optimistic about height prognosis as seven of the eight boys reached target genetic height.

Genitourinary tumors in the families of children with renal tumors

The occurrence of genitourinary tumors in the relatives of a population-based series of 218 children diagnosed with renal tumors was investigated. Family data on 92% (176 of 192) of Wilms' tumor (WT) patients and 77% (20 of 26) of other renal tumor patients were obtained. In all, 21 genitourinary tumors in first-degree relatives in 19 families were ascertained, together with 30 such tumors in second-degree relatives. Ten families were diagnosed with multiple genitourinary tumors, although none of these manifested familial WT. It is proposed that a small proportion of families of children with renal tumors has a genetic predisposition to develop genitourinary tumors and that these tumors may represent further manifestations of the pleiotropic effects of the WT1 gene or of other genes involved in WT predisposition.

Second primary neoplasms in a population-based series of patients diagnosed with renal tumours in childhood

Eight second malignant tumours developed in a population-based series of 218 patients diagnosed with renal tumours in childhood: renal cell carcinoma of the contralateral kidney, hepatocellular carcinoma, Hodgkin's disease, and 4 basal cell and 1 squamous cell carcinomas of skin. Excess risk of developing a second malignancy (excluding skin carcinomas but including a registrable spinal neurofibroma) was 14.7 (95% CI 4.0-37.7, P = 0.0003) for Wilms' tumour patients. Cumulative incidence of second malignant neoplasms (excluding skin carcinoma) was zero at 10 years, 5.0% at 20 years, and 10.2% at 30 years. The most common second neoplasms seen were benign osseous/chondromatous tumours and 4 of the 7 Wilms' tumour patients with malignant tumours had previous or synchronous tumours of this kind. Development of bony exostoses may be a marker for those patients at particularly high risk of subsequent malignancy.

Congenital anomalies and genetic syndromes in 173 cases of medulloblastoma

One hundred seventy-three consecutive cases of medulloblastoma recorded in the Manchester Children's Tumour Registry from 1954 to 1989 were studied. After review of case notes, X-rays, and health surveys the clinical outcome and incidence of congenital anomaly was determined. A previously unreported association with Rubinstein Taybi syndrome was found. Evidence of a genetic syndrome or congenital anomaly was found in 6.4%. These figures provide further evidence of the higher-than-expected incidence of congenital abnormalities.

Salivary gland carcinomas in children: a review of 15 cases

Salivary gland carcinomas are rare in childhood. We have reviewed the case records of 15 children aged 3-14 years (median 11) identified from children's tumour registries. Primary sites were parotid: 11, submandibular: 3, and base of tongue: 1. The range of histologies was similar to that occurring in adults. Six were treated by complete excision, with one given post-operative radiotherapy (RT). All six remain disease-free at 2 months to 21 years after completion of treatment. Five were treated by partial or sub-total excision. Four were given post-operative RT, of whom 3 are disease-free at 3 years, 6 months--18 years and 1 lost to follow-up (LTFA). One not given RT developed a local recurrence at 11 months and was given RT and LTFA. Four patients had a biopsy only. Three were treated by RT. One is disease-free at 8 years, one died of metastatic disease at 6 months, and one developed a local recurrence at 11 years and has remained disease-free following salvage surgery. One patient with advanced disease not suitable for RT died 3 months after diagnosis. Complete excision is the treatment of choice. Following sub-total or incomplete excision post-operative RT can prevent recurrence. Careful RT planning is necessary to minimise late effects.

Prognostic relevance of DNA ploidy in rhabdomyosarcomas and other sarcomas of childhood

A study of DNA content by flow cytometry revealed a significant difference between rhabdomyosarcomas, which were mainly non-diploid, and other sarcomas of children which were mainly diploid (p = 0.01). There was no association between DNA ploidy and survival or aggressive behaviour of the tumour as indicated for example by advanced clinical stage or unfavourable histology. While DNA ploidy correlated with age, it did not correlate with any other clinical characteristic. The apparent lack of prognostic value of DNA content may have been masked by some high CV values and overridden by the effect of chemotherapy which was the most significant variable in determining a patient's survival (p = 0.00005).

Growth hormone and tumour recurrence

OBJECTIVE

To determine whether using growth hormone to treat radiation induced growth hormone deficiency causes tumour recurrence.

DESIGN

Comparison of tumour recurrence rates in children treated with growth hormone for radiation induced deficiency and an untreated population. Computed tomograms from children with brain tumours were reviewed when starting growth hormone and subsequently.

SETTING

North West region.

PATIENTS

207 children treated for brain tumour, 47 of whom received growth hormone and 161 children with acute lymphoblastic leukaemia 15 of whom received growth hormone.

MAIN OUTCOME MEASURES

Tumour recurrence and changes in appearances on computed tomography.

RESULTS

Among children with brain tumour, five (11%) who received growth hormone had recurrences compared with 42 (26%) who did not receive growth hormone. Also adjusting for other variables that might affect tumour recurrence the estimated relative risk of recurrence was 0.82 (95% confidence interval 0.28 to 2.37). The only child with acute lymphoblastic leukaemia who relapsed while taking growth hormone had relapsed previously before starting treatment. Two of the five children with brain tumours who relapsed had abnormal appearances on computed tomography when growth hormone was started. 14 other children who remained relapse free and had follow up computed tomography showed no deterioration in radiological appearance during treatment.

CONCLUSIONS

In this population growth hormone did not increase the risk of tumour recurrence but continued surveillance is essential. Abnormal results on computed tomography are not a contraindication to treatment with growth hormone.

Paediatric craniopharyngiomas: a regional review

Forty paediatric craniopharyngioma cases treated between 1956 and 1987 by conservative surgery (15), radical surgery (10), conservative surgery and radiotherapy (9) and shunting (6) are reviewed. The conservative surgery and radiotherapy group's local control and survival (100%) is significantly better than that of any other group. This group also achieved the most consistent level of employment or tertiary education. Overall morbidity was high. Overall survival has improved since 1976. Whether given as an adjuvant or salvage a radiotherapy dose of TDF 83 or greater gave a significantly better survival (100%) than lower doses.

Thyroid function after treatment of brain tumors in children

In 134 children who had been treated for a brain tumor not involving the hypothalamic-pituitary axis, thyroid function was assessed up to 24 years after treatment with cranial or craniospinal irradiation. In addition, 78 children received up to 2 years of cytotoxic chemotherapy. Of 85 children who received craniospinal irradiation, 30 (35%) had abnormalities of thyroid function, and 10 (20%) of 49 who received cranial irradiation had such abnormalities. Frank hypothyroidism developed in three children and thyrotoxicosis in one. Thirty-six children had an elevated thyroid-stimulating hormone level in the presence of a normal thyroxine level; in 16 of them the thyroid-stimulating hormone level subsequently returned to normal. Twenty-eight children who were treated between 1960 and 1970 were excluded from the analysis. Of 34 children who received cranial irradiation, five had thyroid dysfunction and 24 of 72 who received craniospinal irradiation had such dysfunction (p = 0.013). Thyroid dysfunction was present in 4 of 35 children who received no chemotherapy and in 25 of 71 who received chemotherapy (p = 0.014). Direct irradiation plus chemotherapy was more damaging than irradiation alone. These data confirm the high incidence of thyroid dysfunction when the thyroid gland is included in the radiation field. However, in a high proportion, the thyroid abnormalities are minor and revert to normal with time; life-long replacement therapy with thyroxine may be unnecessary.

The incidence of Gorlin syndrome in 173 consecutive cases of medulloblastoma

We have investigated the incidence of Gorlin syndrome (GS) in patients with the childhood brain tumour, medulloblastoma. One hundred and seventy-three consecutive cases of medulloblastoma in the North-West Regional Health Authority between 1954 and 1989 (Manchester Regional Health Board before 1974) were studied. After review of case notes, X-rays and health surveys only 2/173 cases had evidence supporting a diagnosis of GS. A further case at 50% risk of GS died of a brain tumour aged 4 years. The incidence of GS in medulloblastoma is, therefore, probably between 1-2%. A population based study of GS in the region started in 1983 was used to assess the incidence of medulloblastoma in GS, which was found to be between 3-5%. This figure is lower than previous estimates, but this is the first population based study undertaken. In view of the early age of onset in GS (mean 2 years) children presenting with medulloblastoma, especially under 5 years, should be examined for signs of the syndrome. Those at high risk of developing multiple invasive basal cell carcinomata will then be identified.

Chronic lung fibrosis following carmustine (BCNU) chemotherapy: radiological features

Carmustine (BCNU) is a cytotoxic drug which is a recognized cause of acute pulmonary fibrosis. We describe the radiological findings in six patients who received carmustine in childhood for treatment of central nervous system tumours and were subsequently found to have pulmonary fibrosis 13-17 years after treatment. Patients were studied by chest radiography and high resolution computed tomography. The pattern of disease is novel, involving the upper zones in a predominantly peripheral pattern.

N-myc gene is amplified in alveolar rhabdomyosarcomas (RMS) but not in embryonal RMS

DNA from 13 (6 alveolar and 7 embryonal) childhood rhabdomyosarcomas (RMS) was examined to determine the incidence and prognostic relevance of N- and c-myc genes. Southern analysis showed 5- to 20-fold amplification of N-myc gene in 4 of 6 alveolar but in none of 7 embryonal RMS (p less than 0.04; Fisher's exact test). The number of children who died with multiple- and single-copy N-myc gene was 4/4 and 5/9 respectively (p greater than 0.05; Chi-squared test). There was no statistically significant correlation between N-myc amplification and age, gender, site, stage or survival time. There was no amplification or gross rearrangement of c-myc in any of the 13 RMS.

Effect of abdominal irradiation on growth in boys treated for a Wilms' tumor

To study the effect of abdominal irradiation on spinal growth in childhood we have measured final height, sitting height, and leg length in 30 male survivors of a Wilms' tumor. Twenty-one patients received whole abdominal irradiation by either megavoltage therapy (MV: n = 11) or orthovoltage therapy (OV: n = 10); the remainder received flank irradiation. To examine the effect of the adolescent growth spurt on the irradiated spine we have followed prospectively seven patients who received whole abdominal irradiation and nine patients who received flank irradiation through puberty. Compared to a normal population there is a modest reduction in median final standing height SDS (H.SDS: -1.15) accompanied by a marked reduction in median final sitting height SDS (S.HT SDS: -2.41) with no apparent effect on median subischial leg length SDS (SILL.SDS: 0.04). This reduction in spinal growth is reflected by a strongly positive disproportion score (DPS; [SILL SDS-S.HT SDS] + 2.81). The incidence of scoliosis after abdominal irradiation has been low (10%). During puberty there is a significant fall in median sitting height SDS after both whole abdominal (median fall: -0.9, P = 0.02) and flank irradiation (median fall: -1.85, P = 0.01), and this is reflected in a significant increase in disproportion (DPS: whole abdominal; median rise +1.4, P = 0.02: flank, median rise +1.34, P = 0.01). After MV irradiation there is a significant correlation between the degree of disproportion and the age at treatment (P less than 0.0005). The younger the patient is at treatment the more severe is the restriction on spinal growth and the shorter and more disproportionate they become as an adult. The estimated eventual loss in potential height from abdominal irradiation at the age of one is 10 cm and at five years is 7 cm.

Ovarian failure following abdominal irradiation in childhood: the radiosensitivity of the human oocyte

Ovarian function has been studied sequentially since 1975 in 19 patients treated in childhood for an intra-abdominal tumour with surgery and whole abdominal radiotherapy (total dose 30 Gy). Eleven patients received chemotherapeutic agents that are not known to cause gonadal dysfunction. All but one patient have developed ovarian failure with persistently elevated gonadotrophin levels (FSH and LH greater than 32 IU/litre) and low serum oestradiol values (less than 40 pmol/litre) before the age of 16 years. The majority (n = 12) did not progress beyond breast stage 1 without sex steroid replacement therapy. As the number of oocytes within the ovary declines exponentially by atresia from approximately 2,000,000 at birth to approximately 2000 at the menopause, we have been able to estimate that the LD50 for the human oocyte does not exceed 4 Gy.

Ovarian failure following abdominal irradiation in childhood: natural history and prognosis

Ovarian function has been reviewed sequentially since 1975 in 53 patients treated in childhood between 1942 and 1985 for an intraabdominal tumour with surgery and external abdominal radiotherapy (XRT). Of 38 patients who received whole abdominal XRT (20-30 Gy), 27 failed to undergo or complete pubertal development (pubertal failure) and a premature menopause (median age 23.5 years) occurred in a further ten. Of 15 patients who received flank XRT (20-30 Gy), ovarian function (median age at last assessment 15.2 years) was normal in all but one in whom pubertal failure occurred. In only one patient, who developed pubertal failure after whole abdominal XRT and required sex steroid replacement therapy (HRT) to achieve normal secondary sexual characteristics, has there been evidence of reversibility of ovarian function with a documented conception at the age of 22.7 years. Five patients who developed pubertal failure required bilateral augmentation mammoplasties despite sex steroid replacement therapy. Four patients have had documented conceptions, all received whole abdominal XRT (20-26.5 Gy) and subsequently developed a premature menopause. There have been no live births, with all miscarriages occurring in the second trimester. The outlook for normal ovarian function following whole abdominal XRT is poor, flank XRT introduced intermittently from 1972, has resulted in less pubertal failure but the possibility of a premature menopause may with time become a reality.

Sera of children with renal tumours contain low-molecular-mass hyaluronic acid

The molecular mass of hyaluronic acid (HA) rather than its serum concentration alone may be a hallmark of certain types of malignancy. A radiometric assay was used to measure HA levels in 35 children with renal tumours [33 Wilms' tumours and 2 bone metastasizing renal tumours of childhood (BMRTC)] and 20 normal siblings of children with cancer. The HA level in the sera of normal children was barely detectable and had a molecular mass of 1-5 x 10(5). In both Wilms' and BMRTC patients, very high levels of HA were found in preoperative serum samples; these fell dramatically following surgical excision of the tumours. A novel finding of our study was the presence of low-molecular-mass HA (similar to the angiogenic fragments of HA) in the sera of BMRTC patients. In contrast, high-molecular-mass HA (which is not angiogenic) was found in the sera of Wilms' patients (2 x 10(6) kDa). Following surgery in BMRTC patients, not only did serum HA levels fall to a value within normal ranges, but also the HA which remained was of high molecular mass.

A review of astrocytoma in childhood

Between 1954 and 1984, 282 children with astrocytoma were included in the Manchester Children's Tumour Registry (MCTR), giving an overall incidence of 9.3 per million person-years. There were 110 children with adult astrocytoma and 172 children with juvenile astrocytoma. The five-year survival for adult astrocytoma was 15% and 75% for juvenile astrocytoma. There were no significant improvements in survival with time. There were 21 children with neurofibromatosis (NF) and 4 children had tuberous sclerosis. Some children had other recognized syndromes and others had major or minor abnormalities. Nine children had second tumors, mainly associated with NF, and seven siblings had malignant tumors. A number of mothers of these children were found to have breast cancer. Some of these families may represent examples of the Li-Fraumeni cancer family syndrome. We conclude that astrocytomas is an important problem in childhood and that a proportion of cases may have a genetic origin.

Adrenal cortical carcinoma: survival after radiotherapy

This study describes 15 patients with histologically proven adrenal cortical carcinoma seen at one radiotherapy centre between 1968 and 1981. Nine patients had radiotherapy following surgery and their 10 year uncorrected survival was 33%. Prognostic factors included age at diagnosis, hormone production and complete surgical removal. Two of the three long term survivors developed second malignancies.

Does growth hormone cause relapse of brain tumours?

Tumour relapse rates in 14 patients with medulloblastoma, 8 with glioma, 2 with ependymoma, 6 with leukaemia, and 1 with T-cell lymphoma who received growth hormone (GH) treatment for growth failure secondary to cranial irradiation were compared with rates among patients treated with radical radiotherapy for the same types of tumour. Five relapses (in 5 patients) occurred (1 optic nerve glioma, 2 medulloblastomas, and 2 ependymomas), three during and two after completion of GH treatment. Patients with medulloblastoma and ependymoma who relapsed were older at tumour diagnosis, underweight at the start of GH therapy, and entered puberty later than similar relapse-free patients. The late relapse rate of medulloblastoma and glioma was unaltered by GH therapy. Ependymoma carries a poor prognosis, and of the 4 late survivors, the 2 who received GH relapsed. No leukaemic relapse has been associated with GH treatment. The findings indicate that GH therapy does not increase the relapse rate of medulloblastoma, glioma, and leukaemia.