Methylation of RASSF1A and TRAIL pathway-related genes is frequent in childhood intracranial ependymomas and benign choroid plexus papilloma

Ependymomas (EP) represent the third most frequent type of central nervous system (CNS) tumor of childhood, after astrocytomas and medulloblastomas. No prognostic biological markers are available, and differentiation from choroid plexus papilloma (CPP) is difficult. The present objective was, for a sample of 27 children with intracranial EP and 7 with CPP, to describe and compare the methylation status of 19 genes (with current HUGO symbol, if any): p15INK4a (CDKN2B), p16INK4a and p14ARF (both CDKN2A), APC, RB1, RASSF1A (RASSF1), BLU (ZMYND10) FHIT, RARB, MGMT, DAPK (DAPK1), ECAD (CDH1), CASP8, TNFRSF10C, TNFRSF10D, FLIP (CFLAR), INI1 (SMARCB1), TIMP3, and NF2. Three adult corteses were used as a control. We detected a similar percentage of methylated tumors in both groups (71% in CPP and 77% in EP). No gene was methylated in that control group. RASSF1A was the most frequently methylated gene in both benign tumors (66%) and EP (56%). The genes associated with apoptosis were methylated in both groups of tumors. The percentages of TRAIL pathway genes (CASP8, TFRSF10C, and TFRSF10D) methylated were 30, 9.5, and 36.4%, respectively, in ependymomas and 50, 50, and 16.7%, respectively, in choroid plexus papillomas. No other gene was methylated in the benign tumors, whereas FHIT was methylated in 22%, RARB in 14.8%, BLU in 13.6%, p16INK4a in 11.1%, TNFRSF10C in 9.5%, and DAPK in 7.4% of ependymomas. Although we did not observe a statistical relationship between methylation and clinical outcome, the methylation pattern does not appear to be randomly distributed in ependymoma and may represent a mechanism of tumor development and evolution.