G537 A retrospective analysis of recurrent paediatric ependymoma: poor survival and ineffective therapy necessitates a better understanding of relapse biology to develop improved therapeutic approaches

Background Ependymoma is the third most common brain tumour of childhood. Relapse occurs in approximately 50% of cases and is associated with a very poor prognosis. Treatment at relapse lacks a standard approach and currently employed interventions are often unsuccessful. Over the last few years numerous authors have described and validated molecularly defined ependymoma subgroups but few have evaluated these groups in exclusively paediatric ependymoma cohorts in the context of relapsed disease. Methods We analyzed a cohort of 302 paediatric ependymoma cases collated over a 25 year time period (1990–2015). Tumor, demographic and treatment variables were investigated for association with relapse risk and time to recurrence as well as survival of relapsed patients. DNA methylation profiling of 135 cases allowed the analysis to be placed in the context of molecular subgroups. Tissue samples were obtained through the Children’s Cancer and Leukaemia Group (CCLG) and our research into paediatric ependymoma has ethical approval. Results Our cohort was representative of paediatric ependymoma in terms of tumour location, age at presentation, molecular subgroup and treatment approach. 62% cases relapsed, on average suffering two recurrences. The majority relapsed within two years. However, a small proportion relapsed much later, with 3% beyond 10 years after initial diagnosis. Tumor grade, extent of surgery and radiotherapy significantly affected relapse risk and time to first relapse. However, once a relapse occurred surgery and irradiation delayed disease progression but had minimal impact on long-term survival. When analyzing ependymoma by molecular subgroup, EPN_PFA cases had an increased risk of relapse compared to EPN_RELA cases. However, outcome for relapsed cases was equally poor for both subgroups. Conclusions Recurrent paediatric ependymoma is an aggressive disease with a very poor outcome, for which current treatments are destined to fail. The advent of molecular subgrouping for ependymoma is exciting but also produces challenges to developing evidence based treatment approaches by narrowing the number of cases available for clinical trials and other research. A deeper understanding of the underlying biology at relapse and identification of novel therapeutic approaches relevant to molecular subgroups is urgently needed.