PTHP-02. The clinical value of histopathological and molecular metrics in pediatric ependymoma - data from the ACNS0831 trial

The ACNS0831 Children’s Oncology Group clinical trial enrolled patients from 2010 to 2019 with the primary aim of determining whether post-irradiation chemotherapy offers a survival benefit to children with intracranial ependymoma. Histopathological metrics to inform grade were collected from tumors of 449 patients, 283 (63%) with a posterior fossa (PF) ependymoma, and 166 (37%) with a supratentorial (ST) ependymoma. A retrospective analysis of molecular group (ZFTA, YAP1, PFA, PFB), H3K27-trimethylation status (PF tumors), and CNVs on chromosomes 1q and 6q (PF tumors) and at the CDKN2A locus (ST tumors) was undertaken on prospectively collected samples using methylation profiling, fluorescence in situ hybridization, and immunohistochemistry. Survival data were used to assess the prognostic value of histopathological and molecular variables. Of ST and PF ependymomas, 75% and 54% were WHO grade 3, respectively, the remainder being WHO grade 2. Grade was significantly (p<0.01) associated with event-free and overall survival in all patients with a gross total resection. Molecular group was established in 422/449 (94%) of cases. Most ependymomas were classified by methylation profiling as PFA (59%), followed by ZFTA (31%), PFB (8%), and YAP1 (2%). A relatively good survival was associated with PFB tumors, when compared to ZFTA and PFA tumors. Gain of 1q and loss of 6q were associated with a worse survival among PF and PFA tumors, but homozygous deletion of CDKN2A was not an adverse prognostic factor among ST tumors. Utilizing the histopathological and molecular data from PF ependymomas, it was possible to create ‘tumor risk-groups’ with significantly different survival outcomes. However, ST ependymoma molecular markers did not carry prognostic value. This is the first clinical trial of new therapeutic approaches for intracranial ependymoma that has prospectively collected tissue for molecular analysis. Trial data suggest that histopathological and molecular variables can be used in future trials for therapeutic stratification.