Does your adjuvant radiation approach differ for ependymomas with RELA fusion?

Ependymomas (EPN) can occur either in the cord, the posterior fossa (PF), or supratentorially, and are found both in adults and children. At a very high level, the standard management consists of resection followed by radiotherapy. Although various chemotherapy regimens have been used in several settings, no definitive drug, drug combination, patient subpopulation, or clinical setting has been identified where definitive benefit from chemotherapy can be contended. Historically, grading schemas have been described, but none of these have reliably permitted treatment selection. Similarly, great advances have been made regarding molecular interrogation, but currently, no marker-driven treatment selection paradigms exist for EPN. 

In the 2016 WHO classification of CNS tumors, ependymomas (EPN) were categorized into five subtypes presented here in ascending grade order: (1) subependymoma (SE; Grade I), (2) myxopapillary EPN (MPE; Grade I), (3) EPN, (4) EPN, RELA fusion-positive (EPN-RELA) (Grade II or III), and (5) anaplastic EPN (AEPN). In the updated WHO 4th edition, the supratentorial molecular variant, ST-EPN-RELA is defined as a distinct disease entity, but no specific treatment recommendations distinct from traditional therapeutic approaches are recommended.

The EPN-RELA variant accounts for approximately 70% of all childhood supratentorial ependymomas, and some reports suggest that a significant proportion, possibly up to one-quarter of EPN-RELA occur in adults. In general, in most studies, patients with EPN-RELA have poorer outcomes and worse prognosis with 5- and 10-year PFS of under 30 and 20%, respectively, and 19%. However, not all published reports are consistent in this regard, and there is a growing understanding, that even in EPN-RELA, certain variants, such as clear cell EPN might exhibit a more favorable prognosis. Also, a higher proportion of supratentorial EPN-RELA tend to get classified as AEPN WHO grade III. This simply underscores the heterogeneity and fluidity of current classification schemes, and therefore, greater reliance is being placed on molecular classification approaches. Global DNA methylation patterns distinguish nine molecular groups, some of which tightly overlap with histopathological subgroups, but not all. A 2017 Acta Neuropathol consensus paper recommended that management decisions should not be based on grade, and despite the increasing value of molecular stratification, this does not translate to clinical decision-making today.

The standard treatment decision today is maximal resection followed by radiotherapy, typically 54–59.4 Gy for non-disseminated ependymoma to lower the risk of local recurrence. In general, there has been a trend to decrease margins down to about 1 cm, as most relapses remain central. These recommendations are considered legacy and none were developed with prospective molecular stratification. 

Unquestionably, there is a trend, without level 1 evidence to withhold radiotherapy in some patients. For those undergoing gross total resection (GTR), sometimes such an approach is advocated. Additionally, given the superior outcomes of GTR, some have advocated interweaving chemotherapy as a “bridge” to response and second resection, although this approach should remain in the clinical trial realm. There is considerable interest in pursuing clinical trials withholding radiotherapy after gross total resection in one sub-entity, PF-EPN-B, where, retrospective data suggest low rates of local recurrence after GTR alone; this, however, needs to be done in the context of a clinical trial, and not ad-hoc. Of note, RELA fusions are usually not described in this entity, and they tend to occur in adolescents and young adults. For the specific situation of supratentorial RELA fusion tumors, there are no prospective data supporting the delay of radiotherapy. However, there are some retrospective data regarding withholding RT in gross totally resected ST-EPN (without the RELA fusion status being known).  For example, the MDACCC group (Ghia et al., PMID 24085643) reported on 92 patients with gross totally resected non-anaplastic supratentorial ependymomas and contended that they did not find loss of progression-free or overall survival when RT was deferred. In contradistinction, Pajtler et al. recently published an elegant report in Cancer Cell where molecular characteristics were determined and correlated with outcomes. 122 ST-EPN were classified according to their DNA methylation profiles as ST-EPN-RELA, ST-EPN-YAP1, and ST-subependymoma (SE). A significant majority, 70% harbored C11ORF95 gene fusions to RELA and these were specifically associated with worse outcomes, and although they could not answer the specific question regarding omission of RT, doing less for patients with worse prognosis, in the absence of prospective data is a bit cavalier, in my opinion. In fact, in the aforementioned 2017 Acta Neuropathol collaborative thinktank report, the authors “concluded that there was not enough evidence yet to recommend distinct treatment approaches for ST-EPN-RELA ependymoma.”

So, in our practice, we continue to recommend RT following resection for ST-EPN-RELA.