paediatrics Brussels 17

Articles

The original aim of avoiding or delaying radiotherapy in these children without compromising outcome has been achieved. Our results confirm a role for primary chemotherapy in young children with intracranial ependymoma. The results reported here will contribute further to the impetus for collaborative studies in Europe and the US in this very young age group. The establishment of a clinical scientific consensus on risk stratification factors is the first, and most important, next step. Despite these advances, the long-term outlook for children with ependymoma remains unacceptably poor and further therapeutic advances will only come through a better understanding of the underlying tumour biology. Contributors RGG, SW, and MWE were responsible for data analysis and data interpretation. CW and DM undertook the statistical analysis and interpretation. KR was the trial coordinator, and was responsible for data management. JI and DWE undertook the central neuropathology review. TC and WKC did the central radiological review. JP and CM did the neurosurgical review and assessment. RG, RHAC, DAW, JP, CCB, and LSL designed the trial. NT did the trial radiotherapy review. RG and LSL wrote the report. SP, MWE, DAW, and LSL reviewed the report. LSL also contributed to data assessment. Participating centres Coordinating centre : CCLG Data Centre, University of Leicester, UK. Clinical centres : Denmark : University Hospital, Copenhagen. Eire : Our Lady’s Hospital for Sick Children, Dublin. England : Addenbrooke’s Hospital, Cambridge; Birmingham Children’s Hospital; Bristol Children’s Hospital; St James’ University Hospital, Leeds; Great Ormond Street Hospital for Children, London; The Royal Manchester Children’s Hospital; Queen’s Medical Centre, Nottingham; Royal Victoria Infirmary, Newcastle upon Tyne; John Radcliffe Hospital, Oxford; Sheffield Children’s Hospital; Southampton General Hospital; Royal Marsden Hospital, Sutton. Northern Ireland : The Royal Hospital for Sick Children, Belfast. Scotland : Royal Hospital for Sick Children; Edinburgh. Sweden : Queen Silvia’s Hospital for Children, Gothenburg. The Netherlands : Emma Kinderziekenhuis, Amsterdam. Wales : The Children’s Hospital for Wales, Cardiff. Conflicts of interest The authors declared no conflicts of interest. Acknowledgments We thank Tai Bee Choo for help with the competing risks analysis, Charles Stiller for critical reading of the manuscript and helpful suggestions, and Diane Gumley for data on neurocognitive outcome. The Children’s Cancer and Leukaemia Group (CCLG) is supported by Cancer Research-UK and this study was also funded by the Samantha Dickson Brain Tumour Trust. References 1 Bouffet E, Perilongo G, Canete A, Massimino M. Intracranial ependymomas in children: a critical review of prognostic factors and a plea for cooperation. Med Pediatr Oncol 1998; 30: 319–29. 2 Grill J, Le Deley MC, Gambarelli D, et al. Postoperative chemotherapy without irradiation for ependymoma in children under 5 years of age: a multicenter trial of the French Society of Pediatric Oncology. J Clin Oncol 2001; 19: 1288–96. 3 Duffner PK, Horowitz ME, Krischer JP, et al. Postoperative chemotherapy and delayed radiation in children less than three years of age with malignant brain tumors. N Engl J Med 1993; 328: 1725–31. 4 Merchant TE, Mulhern RK, Krasin MJ, et al. Preliminary results from a phase II trial of conformal radiation therapy and evaluation of radiation-related CNS effects for pediatric patients with localized ependymoma. J Clin Oncol 2004; 22: 3156–62.

(103), verbal (105) IQ, and performance (non-verbal) (99) IQ within the normal range. There is a clear association between methotrexate and acute and late neurotoxicity, the severity and nature of which are dependent on the dose and mode of administration of the drug, folinic acid rescue, and the concomitant use of radiotherapy. 39–41 Evidence of late CNS damage by high-dose methotrexate comes from the presence of leucoencephalopathy. Only two of our long- term survivors had subtle white-matter changes on MRI. Similar low incidence rates of this abnormality have been reported by Kellie and co-workers. 42 The risk factors for leucoencephalopathy due to methotrexate include highest doses (>10000 mg/m²) and frequent administration (7–10 day intervals). 40,41 The methotrexate dose (8000 mg/m²) and interval (8 weeks) in our study were less than this. The report describing the use of intraventricular and intravenous methotrexate in children younger than 3 years with medulloblastoma is also reassuring as although leucoencephalopathy was detected, its presence did not predict for worse neurocognitive outcomes within the treatment cohort, except when cranial radiotherapy was also used. The cognitive outcomes were, however, worse than a normal comparative group. 43 The risk of neurocognitive late effects from high-dose methotrexate in children with brain tumours would therefore seem to be acceptable, 40,42,44 but requires monitoring prospectively in future studies with neuropsychological assessment. We conclude that the risk of neurotoxicity from this protocol is acceptable given the serious nature of the presenting clinical problem and the multimodal therapy required for successful outcomes. The extent of surgical resection is the most consistently reported prognostic factor affecting both progression- free and overall survival both in single centre, 4,27,30,45 and multi-centre studies. 2,3,5 A few single-centre retrospective studies have found no survival advantage to complete resection. 29,46,47 However, the proportion of cases in which a complete surgical resection is obtained varies from around 50% in most studies, 2,3,27,34 to 85%. 4 We have shown that the neurosurgical assessment of the extent of surgery more closely reflected outcome than did radiology review. Our study showed that whilst there was an indication of a better event-free survival for children who had a complete resection compared with those with less complete resection, this did not translate into an improved overall survival. The lack of evidence for surgical resection predicting outcome could be due to the confounding effect of surgical toxicity compromising delivery of effective chemotherapy, or effective chemotherapy and stratified radiotherapy diluting out the effect of enhanced surgical resection. There is no doubt that optimised uncomplicated primary resection is an excellent start for the management of childhood ependymoma. 1 Whether centralised specialist surgical centres or vigorous training and multicentre audits can best deliver low surgical toxicity rates in health systems is yet to be established.

704

http://oncology.thelancet.com Vol 8 August 2007