Toward an integrated histomolecular diagnosis of supratentorial ependymoma

In the past, ependymomas from across the neuro-axis have been considered a single disease entity, with 3 histologically defined grades used to predict patient outcome. Predictions for patient prognosis based on ependymoma histology have, however, not been reliable, since patients with histologically similar tumors often have distinct clinical outcomes.¹ Recent genetic, transcriptional, and epigenetic profiling efforts have found that histologically defined ependymoma represents a diverse set of entities comprising clinically, demographically, and molecularly distinct subgroups.^2–6 Three of these ependymoma subgroups, defined based on DNA methylation analyses, were identified in the supratentorial compartment. One subgroup was enriched for World Health Organization (WHO) grade I tumors, and termed “molecular” subependymoma (ST-SE), while the other 2 were characterized by mutually exclusive prototypic fusion genes involving YAP1 and RELA, and named ST-EPN-YAP1 and ST-EPN-RELA, respectively.^5,7,8 Parker et al⁵ first demonstrated that approximately two thirds of supratentorial ependymomas occurring in childhood harbor fusions between RELA, the principal effector of canonical nuclear factor-kappaB (NF-κB) signaling, and an uncharacterized gene, C11orf95, resulting from chromothripsis. The C11orf95-RELA fusion protein translocates spontaneously to the nucleus, where it activates targets associated with NF-κB signaling. Parker et al showed that these RELA fusion proteins alone have high oncogenic potential and are sufficient to transform neural stem cells to form tumors in mice.⁵ Although the mechanisms by which RELA fusion proteins exert their oncogenic potential are not yet completely understood, NF-κB signaling must be regarded as a promising new therapeutic target for this subgroup of supratentorial pediatric ependymomas.

The study by Figarella-Branger et al published in this issue of Neuro-Oncology focuses on supratentorial clear cell ependymoma. Clear cell ependymoma represents one of the 4 main histological ependymoma subtypes defined by the 2007 WHO brain tumor classification.⁹ The authors report on a rare series of 20 supratentorial clear cell ependymomas that all share an additional histopathological attribute, the appearance of branching capillaries. The cohort showed a predominance of male patients with a median age at surgery of 10.4 years (range, 0.8–68.8 y). Data from MRI or CT scans revealed contrast enhancement as well as cystic and calcified regions for most of the tumors and a localization to the frontal cortex. Interestingly, nuclear immunostaining with a p65/RELA antibody was detected in all cases. The authors identified a causative C11orf95-RELA fusion using reverse transcriptase (RT)-PCR primers for the most common subtype, RELA fusion type I, in 19 of 20 samples. The remaining sample is thought to harbor one of the other 6 described RELA fusion subtypes which were not tested for. Based on these findings, the authors concluded that supratentorial ependymomas, phenotypically appearing as clear cell tumors with branching capillaries, represent a subset of the molecularly defined ST-EPN-RELA group.

Figarella-Branger et al are the first to report a correlation between a defined histological appearance of supratentorial ependymoma and a recently described distinct molecular subgroup. Interestingly, 10-year overall survival in patients with this histologically homogeneous RELA-positive series identified by immunohistochemistry and/or RT-PCR was better compared with survival rates in 122 patients with supratentorial ependymomas from a recent molecular classification study that were subgrouped according to their DNA methylation profiles as ST-EPN-RELA (72.2% vs 49%).⁸ Clinical outcome for patients with supratentorial ependymomas has mainly been analyzed using retrospective series that were either not further molecularly characterized^10,11 or, apart from the WHO grade, did not account for the exact histological subtype.⁸ At this stage, it is therefore difficult to infer whether RELA-positive clear cell ependymoma with branching capillaries represents a subset of histomolecularly defined tumors that have a more favorable outcome compared with the cellular subtype harboring the same fusion event. To date, diagnosis of ependymoma based on the WHO brain tumor classification (the latest version from 2007 is currently being updated) is still based mainly on histological and immunohistochemical findings.⁹ However, given the emerging knowledge of the molecular basis of these tumors, there is a pressing need to also incorporate molecular analyses.¹² Tumors characterized by C11orf95–RELA fusions, which were shown to drive a specific oncogenic NF-kB transcription program, will be even introduced as a new entity in the revised version of the brain tumor classification. An integrative approach including routine testing for the RELA fusion as a diagnostic marker in histologically identified supratentorial ependymomas and in cases with diagnostic uncertainty or ambiguous features will undoubtedly facilitate and improve tumor classification. The study by Figarella-Branger et al has now set the stage to further explore if and to what extent distinct morphological appearances reflect biological subgroups of supratentorial ependymoma. Analyses of bigger cohorts are needed to confirm whether clear cell morphology with branching capillaries is always associated with ST-EPN-RELA tumors or if this histology might also occur in ST-EPN-YAP1 tumors. Specialties involved in the treatment of these tumor types should now move forward quickly to elicit the prognostic as well as predictive impact of the molecular subgroups ST-EPN-RELA and ST-EPN-YAP1 in more detail. Integrated histomolecular analyses of supratentorial ependymomas from histopathologically and clinically well-annotated international trial cohorts as well as from single series with long-term follow-up, as in the Figarella-Branger et al study, are essential to answer this question. An interconnection of tissue-based information with more precise diagnostic aids gained from novel, unbiased, and observer-independent molecular analyses as well as data from imaging methods, including central pathology and radiological review, is required to finally converge at a refined diagnostic setting and improve treatment strategies for ependymoma. For RELA-positive ependymoma there is a good chance that this theory will be turned into action, since these tumors will be considered as an entity in the update of the WHO classification.