The effectiveness and safety of proton beam radiation therapy in children and young adults with Central Nervous System (CNS) tumours: a systematic review.

BACKGROUND: Central nervous system (CNS) tumours account for around 25% of childhood neoplasms. With multi-modal therapy, 5-year survival is at around 75% in the UK. Conventional photon radiotherapy has made significant contributions to survival, but can be associated with long-term side effects. Proton beam radiotherapy (PBT) reduces the volume of irradiated tissue outside the tumour target volume which may potentially reduce toxicity. Our aim was to assess the effectiveness and safety of PBT and make recommendations for future research for this evolving treatment. METHODS: A systematic review assessing the effects of PBT for treating CNS tumours in children/young adults was undertaken using methods recommended by Cochrane and reported using PRISMA guidelines. Any study design was included where clinical and toxicity outcomes were reported. Searches were to May 2021, with a narrative synthesis employed. RESULTS: Thirty-one case series studies involving 1731 patients from 10 PBT centres were included. Eleven studies involved children with medulloblastoma / primitive neuroectodermal tumours (n = 712), five ependymoma (n = 398), four atypical teratoid/rhabdoid tumour (n = 72), six craniopharyngioma (n = 272), three low-grade gliomas (n = 233), one germ cell tumours (n = 22) and one pineoblastoma (n = 22). Clinical outcomes were the most frequently reported with overall survival values ranging from 100 to 28% depending on the tumour type. Endocrine outcomes were the most frequently reported toxicity outcomes with quality of life the least reported. CONCLUSIONS: This review highlights areas of uncertainty in this research area. A well-defined, well-funded research agenda is needed to best maximise the potential of PBT. SYSTEMATIC REVIEW REGISTRATION: PROSPERO-CRD42016036802.

Outcomes of non-anaplastic stage III and 'inoperable' Wilms tumour treated in the UKW3 trial.

BACKGROUND AND PURPOSE: To describe the outcome of patients with stage III Wilms tumours (WT) treated in the UKW3 trial. MATERIAL AND METHODS: Patients with a pathologically confirmed stage III non-anaplastic WT at nephrectomy (Group A) or with an 'inoperable' tumour at diagnosis managed by biopsy and pre-operative chemotherapy (Actinomycin D-Vincristine-Doxorubicin) but stage I or II at subsequent nephrectomy (Group B) were included. RESULTS: The 4-year overall (OS)/event free survival (EFS) for Group A (n = 117) patients was 90%(95%CI:83-94)/81%(CI:73-87) and for Group B (n = 32) 94%(CI:77-98)/88%(CI:70-95). The 4-year OS/EFS of patients with pathological stage III WT according to whether they received flank/abdominal radiotherapy (95 patients) or not (37 patients, 22 from UKW3 pooled with 17 patients from UKW2) were 91%(CI:83-95)/82%(CI:73-89), and 84%(CI:67-92)/78%(CI:61-89), respectively. The 4-year OS/EFS for patients having one reason to be stage III versus two or three was 92%(CI:84-96)/83%(CI:73-90) and 85%(CI:70-93)/78%(CI:61-88), respectively. CONCLUSION: Our findings question the inclusion of biopsy or pre-operative chemotherapy as sole criterion for assigning a tumour stage III. Selected patients with pathological stage III WT can survive without radiotherapy. Whilst cautious interpretation is needed due to the post hoc nature of these analyses, further biological studies may better characterise those who could benefit from reduced therapy.

The effectiveness and safety of proton beam radiation therapy in children with malignant central nervous system (CNS) tumours: protocol for a systematic review.

BACKGROUND: The aim of this study is to use a systematic review framework to identify and synthesise the evidence on the use of proton beam therapy (PBT) for the treatment of children with CNS tumours and where possible compare this to the use of photon radiotherapy (RT). METHODS: Standard systematic review methods aimed at minimising bias will be employed for study identification, selection and data extraction. Twelve electronic databases have been searched, and further citation, hand searching and reference checking will be employed. Studies assessing the effects of PBT used either alone or as part of a multimodality treatment regimen in children with CNS tumours will be included. Relevant economic evaluations will also be identified. The outcomes are survival (overall, progression-free, event-free, disease-free), local and regional control rates, short- and long-term adverse events, functional status measures and quality of survival. Two reviewers will independently screen and select studies for inclusion in the review. All interventional study designs will be eligible for inclusion in the review. However, initial scoping searches indicate the evidence base is likely to be limited to case series studies, with no studies of a higher quality being identified. Quality assessment will be undertaken using pre-specified criteria and tailored to study design if applicable. Studies will be combined using a narrative synthesis, with differences in results between studies highlighted and discussed in relation to the patient population, intervention and study quality. Where appropriate, if no studies of a comparative design are identified, outcomes will be compared against a range of estimates from the literature for similar populations and treatment regimens from the best available evidence from studies that include the use of advanced conventional photon therapy. DISCUSSION: The evidence base for the use of PBT in children with CNS tumours is likely to be relatively sparse, highly heterogeneous and potentially of a low quality with small sample sizes. Furthermore, selection and publication biases may limit the internal and external validity of studies. However, any tentative results from the review on potential treatment effects can be used to plan better quality research studies that are of a design appropriate for outcome comparison with conventional therapy. SYSTEMATIC REVIEW REGISTRATION: PROSPERO CRD42015029583.

Radiotherapy for benign disease; assessing the risk of radiation-induced cancer following exposure to intermediate dose radiation.

Most radiotherapy (RT) involves the use of high doses (>50 Gy) to treat malignant disease. However, low to intermediate doses (approximately 3-50 Gy) can provide effective control of a number of benign conditions, ranging from inflammatory/proliferative disorders (e.g. Dupuytren's disease, heterotopic ossification, keloid scarring, pigmented villonodular synovitis) to benign tumours (e.g. glomus tumours or juvenile nasopharyngeal angiofibromas). Current use in UK RT departments is very variable. This review identifies those benign diseases for which RT provides good control of symptoms with, for the most part, minimal side effects. However, exposure to radiation has the potential to cause a radiation-induced cancer (RIC) many years after treatment. The evidence for the magnitude of this risk comes from many disparate sources and is constrained by the small number of long-term studies in relevant clinical cohorts. This review considers the types of evidence available, i.e. theoretical models, phantom studies, epidemiological studies, long-term follow-up of cancer patients and those treated for benign disease, although many of the latter data pertain to treatments that are no longer used. Informative studies are summarized and considered in relation to the potential for development of a RIC in a range of key tissues (skin, brain etc.). Overall, the evidence suggests that the risks of cancer following RT for benign disease for currently advised protocols are small, especially in older patients. However, the balance of risk vs benefit needs to be considered in younger adults and especially if RT is being considered in adolescents or children.

Hyperfractionated Accelerated Radiotherapy (HART) with maintenance chemotherapy for metastatic (M1-3) Medulloblastoma--a safety/feasibility study.

BACKGROUND AND PURPOSE: To evaluate feasibility and toxicity of Hyperfractionated Accelerated Radiotherapy (HART) 1.24Gy b.i.d. followed by chemotherapy for M1-3 Medulloblastoma (MB). The aim of HART was to use hyperfractionation to improve therapeutic ratio combined with acceleration to minimise tumour cell repopulation during radiotherapy (RT). MATERIALS AND METHODS: Between February 2002 and May 2008, 34 eligible patients (22 male, 12 female) aged 3-15years (median 7) with metastatic MB (M1-9; M2-3, M3-22) received HART with a craniospinal radiotherapy (CSRT) dose of 39.68Gy followed by 22.32Gy boost to the whole posterior fossa and 9.92Gy metastatic boosts. The 8th and subsequent patients received vincristine (VCR) 1.5mg/m(2) weekly×8 doses over 8weeks starting during the 1st week of RT. Maintenance chemotherapy comprised 8 six-weekly cycles of VCR 1.5mg/m(2) weekly×3, CCNU 75mg/m(2) and cisplatin 70mg/m(2). RESULTS: Median duration of HART was 34days (range 31-38). Grade 3-4 toxicities included mucositis (8), nausea (10), anaemia (5), thrombocytopaenia (2), leucopaenia (24). With 4.5-year median follow-up, 3-year EFS and OS were 59% and 71%, respectively. Of 10 relapses, 1 was outside the central nervous system (CNS), 1 posterior fossa alone and 8 leptomeningeal with 3 also associated with posterior fossa. CONCLUSION: HART with or without VCR was well tolerated and may have a place in the multi-modality management of high-risk MB.

DNA methylation profiling of medulloblastoma allows robust subclassification and improved outcome prediction using formalin-fixed biopsies.

Molecular subclassification is rapidly informing the clinical management of medulloblastoma. However, the disease remains associated with poor outcomes and therapy-associated late effects, and the majority of patients are not characterized by a validated prognostic biomarker. Here, we investigated the potential of epigenetic DNA methylation for disease subclassification, particularly in formalin-fixed biopsies, and to identify biomarkers for improved therapeutic individualization. Tumor DNA methylation profiles were assessed, alongside molecular and clinical disease features, in 230 patients primarily from the SIOP-UKCCSG PNET3 clinical trial. We demonstrate by cross-validation in frozen training and formalin-fixed test sets that medulloblastoma comprises four robust DNA methylation subgroups (termed WNT, SHH, G3 and G4), highly related to their transcriptomic counterparts, and which display distinct molecular, clinical and pathological disease characteristics. WNT patients displayed an expected favorable prognosis, while outcomes for SHH, G3 and G4 were equivalent in our cohort. MXI1 and IL8 methylation were identified as novel independent high-risk biomarkers in cross-validated survival models of non-WNT patients, and were validated using non-array methods. Incorporation of MXI1 and IL8 into current survival models significantly improved the assignment of disease risk; 46 % of patients could be classified as 'favorable risk' (>90 % survival) compared to 13 % using current models, while the high-risk group was reduced from 30 to 16 %. DNA methylation profiling enables the robust subclassification of four disease subgroups in frozen and routinely collected/archival formalin-fixed biopsy material, and the incorporation of DNA methylation biomarkers can significantly improve disease-risk stratification. These findings have important implications for future risk-adapted clinical disease management.

MYC family amplification and clinical risk-factors interact to predict an extremely poor prognosis in childhood medulloblastoma.

The MYC oncogenes are the most commonly amplified loci in medulloblastoma, and have previously been proposed as biomarkers of adverse disease prognosis by us and others. Here, we report focussed and comprehensive investigations of MYCC, MYCN and MYCL in an extensive medulloblastoma cohort (n = 292), aimed to define more precisely their biological significance and optimal clinical application to direct improved disease risk-stratification and individualisation of therapy. MYCC and MYCN expression elevations were multifactorial, associated with high-risk (gene amplification, large-cell/anaplastic pathology (LCA)) and favourable-risk (WNT/SHH molecular subgroups) disease features. Highly variable cellular gene amplification patterns underlay overall MYC copy number elevations observed in tumour biopsies; we used these alternative measures together to define quantitative methodologies and thresholds for amplification detection in routinely collected tumour material. MYCC and MYCN amplification, but not gain, each had independent prognostic significance in non-infants (≥3.0-16.0 years), but MYCC conferred a greater hazard to survival than MYCN when considered across this treatment group. MYCN's weaker group-wide survival relationship may be explained by its pleiotropic behaviour between clinical disease-risk groups; MYCN predicted poor prognosis in clinical high-risk (metastatic (M+) or LCA), but not standard-risk, patients. Extending these findings, survival decreased in proportion to the total number of independently significant high-risk features present (LCA, M+ or MYCC/MYCN amplification). This cumulative-risk model defines a patient group characterised by ≥2 independent risk-factors and an extremely poor prognosis (<15% survival), which can be identified straightforwardly using the reported MYC amplification detection methodologies alongside clinical assessments, enabling targeting for novel/intensified therapies in future clinical studies.

Treatment of recurrent central nervous system primitive neuroectodermal tumours in children and adolescents: results of a Children's Cancer and Leukaemia Group study.

BACKGROUND: The treatment of previously irradiated patients with recurrent central nervous system primitive neuroectodermal tumours (PNETs) is a considerable challenge. A study was undertaken to attempt to improve the outcome for such patients using a high dose chemotherapy (HDCT) based strategy. METHODS: Between 2000 and 2007, 40 patients with relapsed medulloblastoma (MB) and 5 with relapsed supratentorial PNETs (StPNETs) were accrued. All but one had received prior craniospinal radiotherapy. Patients were initially treated with cyclophosphamide (4 g/m(2)) together with surgery or local radiotherapy where appropriate. If complete or near complete remission was achieved, the patient proceeded to receive two sequential courses of HDCT with stem cell rescue. The first course consisted of thiotepa (900 mg/m(2)) and the second carboplatin (AUC 21). RESULTS: All five patients with StPNET died of tumour progression with a median OS of 0.4 years. Nineteen of the 40 patients with relapsed MB underwent surgery. Radiotherapy was administered to eight patients. All patients received at least one course of cyclophosphamide. Only 22 MB patients progressed to the HDCT phase; 10 patients received thiotepa only and 12 thiotepa and carboplatin. At a median follow-up of 7.4 years (Range 2.8-8.2 years), only three MB patients are still alive, one following a further relapse. Three and 5 year OS was 22.0% and 8.2%, respectively and 3 and 5 year EFS was 14.6% and 8.7%, respectively. CONCLUSION: This national study based on a strategy including a particular tandem HDCT regimen showed no benefit for previously irradiated patients with relapsed StPNET and very limited benefit for patients with relapsed medulloblastoma.

Medulloblastoma: clinicopathological correlates of SHH, WNT, and non-SHH/WNT molecular subgroups.

Medulloblastoma is heterogeneous, being characterized by molecular subgroups that demonstrate distinct gene expression profiles. Activation of the WNT or SHH signaling pathway characterizes two of these molecular subgroups, the former associated with low-risk disease and the latter potentially targeted by novel SHH pathway inhibitors. This manuscript reports the validation of a novel diagnostic immunohistochemical method to distinguish SHH, WNT, and non-SHH/WNT tumors and details their associations with clinical, pathological and cytogenetic variables. A cohort (n = 235) of medulloblastomas from patients aged 0.4-52 years was studied for expression of four immunohistochemical markers: GAB1, ß-catenin, filamin A, and YAP1. Immunoreactivity (IR) for GAB1 characterizes only SHH tumors and nuclear IR for ß-catenin only WNT tumors. IRs for filamin A and YAP1 identify SHH and WNT tumors. SHH, WNT, and non-SHH/WNT tumors contributed 31, 14, and 55% to the series. All desmoplastic/nodular (D/N) medulloblastomas were SHH tumors, while most WNT tumors (94%) had a classic phenotype. Monosomy 6 was strongly associated with WNT tumors, while PTCH1 loss occurred almost exclusively among SHH tumors. MYC or MYCN amplification and chromosome 17 imbalance occurred predominantly among non-SHH/WNT tumors. Among patients aged 3-16 years and entered onto the SIOP PNET3 trial, outcome was significantly better for children with WNT tumors, when compared to SHH or non-SHH/WNT tumors, which showed similar survival curves. However, high-risk factors (M+ disease, LC/A pathology, MYC amplification) significantly influenced survival in both SHH and non-SHH/WNT groups. We describe a robust method for detecting SHH, WNT, and non-SHH/WNT molecular subgroups in formalin-fixed medulloblastoma samples. In corroborating other studies that indicate the value of combining clinical, pathological, and molecular variables in therapeutic stratification schemes for medulloblastoma, we also provide the first outcome data based on a clinical trial cohort and novel data on how molecular subgroups are distributed across the range of disease.

Definition of disease-risk stratification groups in childhood medulloblastoma using combined clinical, pathologic, and molecular variables.

PURPOSE: Medulloblastomas are heterogeneous and include relatively good-prognosis tumors characterized by Wnt pathway activation, as well as those that cannot be successfully treated with conventional therapy. Developing a practical therapeutic stratification that allows accurate identification of disease risk offers the potential to individualize adjuvant therapy and to minimize long-term adverse effects in a subgroup of survivors. METHODS: Using formalin-fixed paraffin-embedded (FFPE) tissue for immunohistochemistry, fluorescent in situ hybridization, and direct sequencing to identify tumors with a Wnt pathway signature and those harboring copy number abnormalities (CNAs) of potential prognostic significance (MYC/MYCN amplification, CNAs of chromosome 6 and 17), we evaluated clinical, pathologic, and molecular outcome indicators and stratification models in a cohort (n = 207) of patients with medulloblastoma 3 to 16 years of age from the International Society of Pediatric Oncology CNS9102 (PNET3) trial. RESULTS: Metastatic disease and large-cell/anaplastic (LC/A) phenotype were the clinicopathologic variables associated with poor progression-free survival (PFS). Nuclear immunoreactivity for ß-catenin, CTNNB1 mutation, and monosomy 6 all identified a group of good-prognosis patients. MYC amplification was associated with poor outcome, but other CNAs were not. Low-risk medulloblastomas were defined as ß-catenin nucleopositive tumors without metastasis at presentation, LC/A phenotype, or MYC amplification. High-risk medulloblastomas were defined as tumors with metastatic disease, LC/A phenotype, or MYC amplification. Low-risk, standard-risk, and high-risk categories of medulloblastoma had significantly (P < .0001) different outcomes. CONCLUSION: Integrating assays of molecular biomarkers undertaken on routinely collected diagnostic FFPE tissue into stratification schemes for medulloblastoma alongside clinical and pathologic outcome indicators can refine current definition of disease risk and guide adjuvant therapy.

The addition of high-dose tamoxifen to standard radiotherapy does not improve the survival of patients with diffuse intrinsic pontine glioma.

The study aimed to examine the tolerability of the combination of radiotherapy and tamoxifen and the effect on median and event free survival as well as collecting data on the use of steroids in this population. 31 patients with diffuse intrinsic pontine glioma, diagnosed on clinical and radiological criteria, were treated with high-dose oral tamoxifen (120 mg/m(2)/day) given concomitantly with standard dose radiotherapy (54 Gy in 1.8 Gy fractions over 6 weeks). Results Tamoxifen was well tolerated with no grade 3 or 4 CTC toxicity reported. At 1 year, the progression free and event free survival were 3.2% (95% CI: 0.2-14.1%), and at 6 months 19.4% (CI: 7.9% to 34.6%). The overall survival at 1 year was 16.1% (CI: 5.9-30.9%) with median survival 6.32 months. In this study, in which tamoxifen was used in conjunction with radiotherapy, progression free survival was shown to be less good when compared with historical data HR = 3.1 (CI: 1.7-5.7). There was no significant reduction in overall survival. The addition of high-dose tamoxifen, although well tolerated, confers no clinical benefit to patients treated with diffuse intrinsic pontine glioma treated with standard radiotherapy.

Survival after second primary neoplasms of the brain or spinal cord in survivors of childhood cancer: results from the British Childhood Cancer Survivor Study.

PURPOSE: Survival after brain or spinal cord neoplasms is poor and varies by diagnostic group, age, grade, treatment and pretreatment factors, and location and size of tumor. We carried out a study to investigate survival and factors affecting survival of all diagnostic types of second primary brain or spinal cord neoplasms. PATIENTS AND METHODS: The British Childhood Cancer Survivor Study (BCCSS) is a long-term population-based follow-up study of 17,980 5-year survivors of childhood cancer. We used relative survival and multivariate Cox regression analysis to determine 5-year relative survival and factors affecting survival in second primary meningiomas and gliomas that developed in survivors included in the BCCSS. RESULTS: There were 247 second primary brain or spinal cord neoplasms, including 137 meningiomas and 73 gliomas in a young adult population. Five-year relative survival after meningiomas was similar for males (84.0%; 95% CI, 72.6% to 91.1%) and females (81.7%; 95% CI, 69.9% to 89.3%). For gliomas, 5-year relative survival was 19.5% (95% CI, 8.6% to 33.7%) for males and females. Multivariate analysis showed significant heterogeneity by decade of treatment (P = .04), grade (P = .03), and genetic risk (P = .03) for rate of mortality after a meningioma. For gliomas, survival was significantly affected by grade (P < .001). CONCLUSION: Our results indicate survival is poor after second primary glioma in this young adult population, although survival after second primary meningioma is good. Our study has clinical implications for the surveillance of childhood cancer survivors at risk of developing second primary brain tumors, in particular survivors of childhood acute lymphoblastic leukemia or childhood brain tumors.

Impact of radiotherapy parameters on outcome for patients with supratentorial primitive neuro-ectodermal tumours entered into the SIOP/UKCCSG PNET 3 study.

BACKGROUND AND PURPOSE: To evaluate the impact of radiotherapy (RT) parameters on outcome in the SIOP/UKCCSG study of pre-RT chemotherapy for Supratentorial Primitive Neuro-ectodermal Tumours. METHODS AND MATERIALS: Sixty-two patients aged 2.9-16.6 (median 6.4 years) were eligible. Forty-eight (77%) had non-pineal sites and 14 (23%) had pineal sites. Eleven were randomized to RT alone (6) and five to pre-RT Vincristine, Etoposide, Carboplatin and Cyclophosphamide. Fifty-one were not randomized, 15 receiving RT alone and 36 receiving pre-RT chemotherapy. Craniospinal RT (CSRT) 35 Gy/21 fractions were followed by 20 Gy/12 fractions to primary tumour. RESULTS: Mean CSRT dose was 34.7 Gy and mean total primary dose was 53.4 Gy for those who received radiotherapy. Of 30 relapses, 18 (60%) were local only and 5 (16.7%) were combined local and leptomeningeal. There was no significant impact on Overall Survival (OS) or Event-Free Survival (EFS) of surgery-RT interval for patients treated by pre-RT chemotherapy or RT alone, or duration of RT (completing within 50 days). Planning films were received for 42/54 (77.8%) patients. Fourteen (33%) had one or more targeting deviations (10 cribriform fossa, 11 base of skull). There was a statistically significant increase in the risk of recurrence for patients with cribriform fossa targeting deviations (p=0.033), but not for patients with base of skull targeting deviations (p=0.242). There was no statistically significant difference in OS (p=0.0598) or EFS (p=0.0880) for patients who had one or more targeting deviations compared to those who had none. CONCLUSIONS: This study has not demonstrated a statistically significant impact of radiotherapy duration or targeting deviations on OS or EFS, possibly due to small patient numbers. However, multi-institutional SPNET trials should incorporate quality assurance programs including analysis of relapse pattern in relation to primary target volume coverage.

Nodule formation and desmoplasia in medulloblastomas-defining the nodular/desmoplastic variant and its biological behavior.

Among the variants of medulloblastoma in the current WHO classification of nervous system tumors, the desmoplastic variant, which has been reported to constitute 5%-25% of pediatric medulloblastomas, is defined by its nodular collections of neurocytic cells bounded by desmoplastic internodular zones. We have studied the frequency, morphological features and biological behavior of medulloblastomas in two contemporaneous SIOP/UKCCSG trial cohorts of children with medulloblastomas, CNS9102 (n = 315) and CNS9204 (n = 35), focusing on tumors with nodular and desmoplastic phenotypes. In children aged 3-16 years (CNS9102), the nodular/desmoplastic medulloblastoma represented 5% of all tumors, while in infants aged <3 years (CNS9204) this variant represented 57% of medulloblastomas. Using iFISH to detect molecular cytogenetic abnormalities in medulloblastomas with a nodular architecture, we demonstrated distinct genetic profiles in desmoplastic and non-desmoplastic (classic and anaplastic) tumors; in particular, abnormalities of chromosome 17 occurred in the latter, but not the former. Significantly different outcomes were demonstrated for classic, nodular/desmoplastic and large cell/anaplastic medulloblastomas in both cohorts. In conclusion, the nodular/desmoplastic medulloblastoma appears to have clinical, genetic and biological characteristics that set it apart from other variants of this tumor.

Current developments in radiotherapy for paediatric brain tumours.

This review summarises current developments in radiation oncology and how they impact on the management of children with brain tumours. Improved understanding of radiobiology has led to attempts to improve the therapeutic ratio with hyperfractionated radiotherapy. Recent advances in planning and delivery of radiotherapy, including three-dimensional conformal radiotherapy, intensity modulated radiotherapy, and proton therapy allow a more precise localisation of the maximum dose region with maximum sparing of normal brain. Increasingly interactions between drugs and radiotherapy are exploited, but it is important to evaluate toxicity of combined modality therapy. The introduction of models to predict the impact of radiotherapy dose-volume parameters on long-term neuropsychological function will hopefully lead to further benefit with respect to sparing of normal tissue morbidity.

Analysis of patients with supratentorial primitive neuro-ectodermal tumours entered into the SIOP/UKCCSG PNET 3 study.

The SIOP PNET 3 study was designed to determine whether 10 weeks of moderately intensive chemotherapy given after surgery and before radiotherapy (RT) would improve the outcome for patients with primitive neuroectodermal tumours (PNETs) compared with RT alone. Patients with a histological diagnosis of supratentorial PNET (StPNET) and no radiological evidence of metastatic disease were initially eligible for randomisation to either chemotherapy followed by craniospinal RT 35 Gy in 21 fractions with a boost of 20 Gy in 12 fractions to the primary site, or RT alone. In respect of the increasing recognition that StPNET were high-risk tumours, randomisation for this group closed in November 1999. This analysis includes both randomised and non-randomised patients with StPNET entered into the study database. Sixty-eight patients aged 2.9-16.6 years (median 6.5 years) were included in the analysis (chemotherapy+RT: 44, RT alone: 24). Fifty-four patients (79%) had a non-pineal and 14 (21%) a pineal site. At a median follow-up of 7.4 years, for all patients overall survival (OS) at 3 and 5 years was 54.4% and 48.3%, respectively. Event-free survival (EFS) at 3 and 5 years was 50.0% and 47.0%, respectively. There was no statistically significant difference in OS or EFS according to treatment received. OS (P=0.05) and EFS (P=0.03) were significantly better for patients with pineal primary sites. EFS for pineal tumours were 92.9% at 3 years and 71.4% at 5 years and for non-pineal primaries 40.7% at 3 years and 40.7% at 5 years. This study confirmed the relatively good survival for non-metastatic pineal PNETs but poor survival of non-pineal StPNETs. There was no evidence that pre-radiation chemotherapy improved outlook. Future treatment programs should be directed at the particular natural history of these tumours, to further define prognostic factors and to explore further biological characteristics.