Artikel
Comprehensive transcriptomic characterization of long-term epilepsy associated tumors (LEAT)
Suche in Medline nach
Autoren
-
Dieter Henrik Heiland
- Universitätsklinikum Freiburg, Klinik für Neurochirurgie, Freiburg, Deutschland
-
Oliver Schnell
- Universitätsklinikum Freiburg, Klinik für Neurochirurgie, Freiburg, Deutschland
-
Ori Staszewski
- Universitätsklinikum Freiburg, Institut für Neuropathologie, Freiburg, Deutschland
-
Josef Zentner
- Universitätsklinikum Freiburg, Klinik für Neurochirurgie, Freiburg, Deutschland
-
Daniel Delev
- Universitätsklinikum Freiburg, Klinik für Neurochirurgie, Freiburg, Deutschland
| Veröffentlicht: | 18. Juni 2018 |
|---|
Gliederung
Text
Objective: Long-term Epilepsy Associated Tumors (LEAT) are a heterogeneous group of mostly benign brain tumors associated with an early-onset, drug-resistant epilepsy. However, some LEAT can recur and become malignant. Because of their rare occurrence, the molecular background of LEAT has been seldom investigated. Therefrom, this study presents a comprehensive transcriptional characterization of 70 LEATs and the validation of novel biomarkers for the first time.
Methods: Transcriptome profiles of public datasets (Gene Expression Omnibus) and the local tumor database were pooled and analyzed. 21 ganglioglioma (GGs), 4 (pleomorphic xanthoastrocytoma (PXAs) and 20 (dysembryoplastic neuroepithelial tumors (DNTs) were selected by input quality criteria (outlier detection based on batch effects). 25 samples (13 DNTs/ 10GG/ 2PXA) were analyzed by transcriptome array (HG2.0, Affymetrix). A "partitioning around medoids" (PAM) cluster analysis was performed and identified four robust transcriptional clusters. Validation of clinical biomarkers was performed by immunostaining and qRT-PCR.
Results: The novel transcriptional classification of LEAT showed four distinct subgroups, irrespective of the histopathological morphology. The first discovered subclass contained predominantly DNTs and was characterized by its close alignment with the histological diagnosis. The second subclass was defined by metabolic alterations in the Krebs cycle and included both DNTs (n=9) and GGs (n=6). Next subclass was determined by specific alterations in the MAPK pathway. Of note, GGs and PXAs containing BRAF mutation/fusion or FGFR-fusions occupied this subclass and showed a significantly increased recurrence rate. The forth cluster contained only GGs and showed an increased expression of immune response related genes.
Conclusion: This study reports a novel molecular classification based on integrative transcriptional, genetic and histopathological morphology. The LEAT subclass defined by MAPK pathway alterations presents a potential biomarker for future molecular stratification of recurrent LEATs.
