<p>Recent genomic studies have allowed the subdivision of intracranial ependymomas into molecularly distinct groups with highly specific clinical features and outcomes. The majority of supratentorial ependymomas (ST-EPN) harbor <i>ZFTA-RELA</i> fusions which were designated, in general, as an intermediate risk tumor variant. However, molecular prognosticators within ST-EPN ZFTA-RELA have not been determined yet. Here, we performed methylation-based DNA profiling and transcriptome RNA sequencing analysis of 80 ST-EPN ZFTA-RELA investigating the clinical significance of various molecular patterns. The principal types of <i>ZFTA-RELA</i> fusions, based on breakpoint location, demonstrated no significant correlations with clinical outcomes. Multigene analysis disclosed 1892 survival-associated genes, and a metagene set of 100 genes subdivided ST-EPN ZFTA-RELA into favorable and unfavorable transcriptome subtypes composed of different cell subpopulations as detected by deconvolution analysis. <i>BGN</i> (biglycan) was identified as the top-ranked survival-associated gene and high <i>BGN</i> expression levels were associated with poor survival (Hazard Ratio 17.85 for PFS and 45.48 for OS; log-rank; p-value &lt; 0.01). Furthermore, BGN immunopositivity was identified as a strong prognostic indicator of poor survival in ST-EPN, and this finding was confirmed in an independent validation set of 56 samples. Our results indicate that integrating <i>BGN</i> expression (at mRNA and/or protein level) into risk stratification models may improve ST-EPN ZFTA-RELA outcome prediction. Therefore, gene and/or protein expression analyses for this molecular marker could be adopted for ST-EPN ZFTA-RELA prognostication and may help assign patients to optimal therapies in prospective clinical trials.</p>

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Biglycan-driven risk stratification in ZFTA-RELA fusion supratentorial ependymomas through transcriptome profiling

  • Konstantin Okonechnikov,
  • David R. Ghasemi,
  • Daniel Schrimpf,
  • Svenja Tonn,
  • Martin Mynarek,
  • Jan Koster,
  • Till Milde,
  • Tuyu Zheng,
  • Philipp Sievers,
  • Felix Sahm,
  • David T.W. Jones,
  • Andreas von Deimling,
  • Stefan M. Pfister,
  • Marcel Kool,
  • Kristian W. Pajtler,
  • Andrey Korshunov

摘要

Recent genomic studies have allowed the subdivision of intracranial ependymomas into molecularly distinct groups with highly specific clinical features and outcomes. The majority of supratentorial ependymomas (ST-EPN) harbor ZFTA-RELA fusions which were designated, in general, as an intermediate risk tumor variant. However, molecular prognosticators within ST-EPN ZFTA-RELA have not been determined yet. Here, we performed methylation-based DNA profiling and transcriptome RNA sequencing analysis of 80 ST-EPN ZFTA-RELA investigating the clinical significance of various molecular patterns. The principal types of ZFTA-RELA fusions, based on breakpoint location, demonstrated no significant correlations with clinical outcomes. Multigene analysis disclosed 1892 survival-associated genes, and a metagene set of 100 genes subdivided ST-EPN ZFTA-RELA into favorable and unfavorable transcriptome subtypes composed of different cell subpopulations as detected by deconvolution analysis. BGN (biglycan) was identified as the top-ranked survival-associated gene and high BGN expression levels were associated with poor survival (Hazard Ratio 17.85 for PFS and 45.48 for OS; log-rank; p-value < 0.01). Furthermore, BGN immunopositivity was identified as a strong prognostic indicator of poor survival in ST-EPN, and this finding was confirmed in an independent validation set of 56 samples. Our results indicate that integrating BGN expression (at mRNA and/or protein level) into risk stratification models may improve ST-EPN ZFTA-RELA outcome prediction. Therefore, gene and/or protein expression analyses for this molecular marker could be adopted for ST-EPN ZFTA-RELA prognostication and may help assign patients to optimal therapies in prospective clinical trials.