We examined prognostic factors for adult patients with primary intracranial ependymoma using SEER data from 2004–2017. Younger age, female sex and grade 2 tumor histology emerged as favorable prognostic factors through multivariable analysis. In univariate subgroup analyses, GTR was significantly associated with improved OS in infratentorial cases but not supratentorial cases. Other factors including supratentorial vs. infratentorial location, race and receipt of adjuvant chemotherapy or radiation therapy, were not found to be significant prognosticators in our analysis.
Extent-of-resection has been well established as an important prognostic indicator for children and adults with ependymoma in the spinal, posterior fossa or supratentorial compartment19–21. In our cohort, GTR was only significantly associated with improved OS in patients with infratentorial ependymoma. Notably, our study was limited by the lack of specific anatomical detail needed to determine candidacy for GTR (e.g., brainstem involvement, deep vs. superficial location, eloquence). Our results corroborate findings from the Collaborative Ependymoma Research Network, who also found significantly improved progression free survival (PFS) and OS in patients undergoing GTR22. An earlier SEER-based study of older patients (≥ 60) similarly found GTR to be a significant prognostic indicator across all anatomic compartments23. Notably, the infratentorial cohort had most grade III tumors, which may have influenced the strong relationship with extent of resection in this subgroup. Though our subgroup analysis did not show a statistically significant association between GTR and OS in patients with supratentorial ependymoma, our results reaffirm the evidence supporting the importance of resection in adult patients with infratentorial ependymoma.
Despite a higher proportion of STR in patients with infratentorial ependymoma (44% vs. 65%), as well as a higher incidence of higher-grade tumors (Grade III, 71% vs. 30%), OS was shorter in patients with supratentorial ependymoma than in the infratentorial tumor cohort. Our findings resonate with those of a recent molecularly defined single-center analysis of over 200 adult intracranial ependymoma cases, which found similarly poorer survival in patients with supratentorial disease21. We know from their findings this may be attributable to biological factors such as RELA fusion status and proliferative index. Our findings highlight the currently accepted notion that supratentorial and infratentorial ependymoma represent two distinct disease entities and support the conclusion that adult posterior fossa ependymoma may be a surgically curable entity while RELA-fused supratentorial ependymoma is not10,15.
The prognostic value of histopathological grade in patients with ependymoma remains a contentious topic, in part because of recognized inter-observer inconsistency in assigning tumor grade24. The updated 2021 WHO Classification of Tumors of the Central Nervous System includes acknowledgement of this controversy and formally changes ependymoma histopathological grading by moving away from ‘anaplastic’ terminology, though WHO grade II and III may still be decided by the reporting pathologist1. In this study, we found significantly favorable survival among adult ependymoma patients with WHO grade II histology in both the supratentorial and infratentorial cohorts. Similar findings have been shown in other population-based studies9,22,25,26. These results underscore the importance of continued representation of histological criteria in addition to molecular markers for ependymoma classification1. Future trials should employ a centralized tissue review and include both histopathology and molecular profiles to best categorize intracranial ependymoma.
Neither adjuvant chemotherapy nor radiotherapy were associated with a significant change in OS within the SEER cohort following univariable or multivariable analysis. Current consensus guidelines for adult intracranial ependymoma advise maximal safe surgical resection followed by radiotherapy3,10. Based on the European Association of Neuro-Oncology guidelines, adjuvant radiotherapy is the mainstay of therapy for all grade 3 ependymoma, regardless of extent of resection, and in grade 2 ependymoma following STR27–29. This was reflected in our data, whereby the cohort undergoing adjuvant radiotherapy was enriched with Grade 3 tumors (65%), compared to the cohort who did not receive (28%). This discrepancy likely confounded the finding of no impact on OS among radiated patients. Controversy remains around optimal therapy for adults with WHO grade 2 ependymoma following GTR and this will likely be clarified by integrating molecular tumor information5. Trials are ongoing to determine the safety of radiation dose de-escalation in patients with YAP1 fusion supratentorial ependymoma and posterior fossa group B ependymoma, both favorable molecular subtypes30–32. Very few patients in our cohort underwent adjuvant chemotherapy and the indication or regimen is unclear from SEER database records. The role of adjuvant chemotherapy in adult intracranial ependymoma is not known and remains the subject of active investigation, especially in disseminated or multiply recurrent cases 26,29,33,34.
Our study has limitations. First, the molecular classification of ependymoma has transformed our understanding and management of this disease entity across central nervous system compartments. The current WHO classification system recognizes nine ependymoma subgroups through three anatomic compartments, of which the supratentorial subtypes are importantly distinguished by RELA/ZFTA and YAP1 fusion genes. Our study was unable to incorporate or account for ependymoma molecular profiles among included patients. Second, the SEER registry lacks granular information on adjuvant therapy. For example, we were unable to garner information on dose, fractionation, or modality of adjuvant radiotherapy. Further, the SEER data do not clarify whether radiation therapy took place at a time of initial diagnosis or recurrence or for what reason. Furthermore, data on the chemotherapy regimen administered and indications of chemotherapy were unavailable.