Identification of novel candidate genes for ependymoma progression in chromosomal regions over-represented at recurrence and detected by CGH array

Abstract

2008 Background: The biology and genetics of ependymomas remain poorly understood. As opposed to other pediatric tumors, few recurrent chromosomal abnormalities or relevant pathways have been identified. Method: In order to gain new insight into the molecular mechanisms involved in ependymoma progression, we performed BAC array-based comparative genomic hybridization (aCGH) in a series of 59 tumors at diagnosis and at relapse (42 posterior fossa and 17 supratentorial). Specific chromosomal imbalances were confirmed by FISH. Expression of candidate genes was evaluated by RT-PCR and immunohistochemistry. Results: As compared to CGH (Cancer Genet Cytogenet 2002), aCGH detected more abnormalities since 18/33 samples at diagnosis and 22/26 at recurrence showed chromosomal imbalances. Three distinct patterns were observed : no imbalance (15 at diagnosis, 4 at relapse), large and numerous imbalances (7 at diagnosis, 12 at relapse) and small and rare imbalances (11 at diagnosis, 10 at relapse). At diagnosis, absence of chromosomal imbalance was correlated with young age (p=0.034) and shorter progression-free survival (p=0.014). At relapse, patients with large and numerous imbalances had a shorter survival (p=0.026). Specific chromosomal imbalances were significantly more frequent at recurrence: 9 qter gain (54% vs 21%), 1q gain (12% vs 0%) and 6q loss (27% vs 6%). Samples were available both at diagnosis and relapse for 16 patients. Whole genome changes were towards increase of genomic abnormalities (n=10) but also simplification (n=6), the former being associated with a higher risk of death from disease (Fisher exact, p=0.03). A candidate gene strategy was focussed on the frequently gained 9qter locus. Significant overexpression compared to normal brain was identified for Tenascin-C (median 35 fold, max 220 fold, p<0.0001) and Notch1 (median 6 fold, max 45 fold, p<0.0001). Notch pathway analysis by RT-PCR revealed potential regulatory loops between JAG1/2 or DLL1/3 and Notch1/2. Conclusion: aCGH pattern may identify subgroups of ependymomas with respect to prognosis. Notch pathway and Tenascin-C are important regulatory genes in ependymoma progression and may represent interesting new targets for therapy. No significant financial relationships to disclose.