Abstract

BACKGROUND

Ependymoma is a malignant brain tumor, which is more often diagnosed in children, than in adult patients. Up to now, the classification was performed entirely histopathological. The mainstays of treatment are surgery and radiotherapy. Recently, a molecular classification approach distinguished distinct entities based on DNA methylation profiling in a mixed cohort of all age-groups in ependymal brain tumors. Herein, we characterized DNA methylation profiles and their clinical correlates in a prospectively assembled uniformly adult cohort of the German Glioma Network (GGN), as the first validation cohort to date.

METHODS

The GGN cohort was screened for patients diagnosed with ependymal central nervous system (CNS) tumors, including myxopapillary ependymoma, ependymoma, anaplastic ependymoma, and subependymoma for its clinical presentation. In total, 122 tumors were analyzed based on DNA methylation profiling using the Illumina Infinium HumanMethylation450 BeadChip platform. The assigned molecular subgroups and the identified chromosomal aberrations were correlated with histopathological characteristics and clinical patient data.

RESULTS

Each ependymal CNS tumor could be assigned to one of the previously defined molecular methylation subgroups. Molecular entities of ependymal brain tumors display distinct copy-number variations, including partial loss of chromosome 19 of intracranial and spinal subependymomas in about 50% of cases. PF-EPN-B and ST-EPN-RELA tumors tend to have a worse prognosis compared to other molecular subgroups of ependymal and subependymal tumors in adults.

CONCLUSIONS

Molecular subgrouping of ependymal CNS tumors should be integrated in routine diagnostics, to improve accuracy in diagnostic and prognosis, especially as a basis for risk stratification in future clinical trials. Patients with PF-EPN-B and ST-EPN-RELA subtypes might benefit from early adjuvant treatment beyond surgery.

This content is only available as a PDF.