Multimodal molecular analysis of astroblastoma enables reclassification of most cases into more specific molecular entities

Matthew D. Wood, Tarik Tihan, Arie Perry, Geeta Chacko, Clinton Turner, Cunfeng Pu, Christopher Payne, Alexander Yu, Serguei I. Bannykh, David A. Solomon

Research output: Contribution to journalArticlepeer-review

60 Scopus citations

Abstract

Astroblastoma is a rare and controversial glioma with variable clinical behavior. The diagnosis currently rests on histologic findings of a circumscribed glioma with astroblastomatous pseudorosettes and vascular hyalinization. Immunohistochemical studies have suggested different oncogenic drivers, such as BRAF p.V600E, but very few cases have been studied using genome-wide methodologies. Recent genomic profiling identified a subset of CNS embryonal tumors with astroblastoma-like morphology that harbored MN1 gene fusions, termed “CNS high-grade neuroepithelial tumors with MN1 alteration” (CNS-HGNET-MN1). To further characterize the genetic alterations that drive astroblastomas, we performed targeted next-generation sequencing (NGS) of 500 cancer-associated genes in a series of eight cases. We correlated these findings with break-apart fluorescence in situ hybridization (FISH) analysis of the MN1 locus and genome-wide DNA methylation profiling. Four cases showed MN1 alteration by FISH, including two pediatric cases that lacked other pathogenic alterations, and two adult cases that harbored other cancer-associated gene mutations or copy number alterations (eg, CDKN2A/B homozygous deletion, TP53, ATM and TERT promoter mutations). Three of these cases grouped with the CNS-HGNET-MN1 entity by methylation profiling. Two of four MN1 intact cases by FISH showed genetic features of either anaplastic pleomorphic xanthoastrocytoma (BRAF p.V600E mutation, CDKN2A/B homozygous deletion and TERT promoter mutation) or IDH-wildtype glioblastoma (trisomy 7, monosomy 10, CDK4 amplification and TP53, NRAS and TERT promoter mutations) and these cases had an aggressive clinical course. Two clinically indolent cases remained unclassifiable despite multimodal molecular analysis. We conclude that astroblastoma histology is not specific for any entity including CNS-HGNET-MN1, and that additional genetic characterization should be considered for astroblastomas, as a number of these tumors likely contain a methylation profile or genetic alterations that suggest classification as other tumor entities. Our heterogeneous molecular findings help to explain the clinical unpredictability of astroblastoma.

Original languageEnglish (US)
Pages (from-to)192-202
Number of pages11
JournalBrain Pathology
Volume28
Issue number2
DOIs
StatePublished - Mar 2018

All Science Journal Classification (ASJC) codes

  • General Neuroscience
  • Pathology and Forensic Medicine
  • Clinical Neurology

Fingerprint

Dive into the research topics of 'Multimodal molecular analysis of astroblastoma enables reclassification of most cases into more specific molecular entities'. Together they form a unique fingerprint.

Cite this