Artikel
Identification of prognostic markers in primary glioblastomas with spatial relationship to the subventricular zone
Suche in Medline nach
Autoren
Veröffentlicht: | 13. Mai 2014 |
---|
Gliederung
Text
Objective: We recently reported that contact of primary glioblastoma (GBM) to the subventricular zone (SVZ), the predominant neurogenic region of the adult brain, is associated with increased stem cell marker expression and reduced overall survival (OS). Moreover, mRNA microarray analysis revealed a gene signature distinctive for GBM in contact with the SVZ that was enriched for genes associated with the immune system, developmental pathways and angiogenesis. This analysis aimed at validating single candidate genes with respect to their prognostic impact.
Method: mRNA microarray analysis was conducted from vital tumor samples of 36 GBM with or without contact to the SVZ. R software was used for data processing and statistical analysis. For identification of single genes distinctive of GBM in contact with the SVZ, we investigated genes with the strongest differential expression (p<0.01; fold change(log)>0.5). Validation of localization-dependent expression and evaluation of a potential prognostic impact was done by qPCR in an independent validation set of 145 patients with primary GBM, 98 of them having received intensified treatment (i.e. completion of Stupp regimen). The median was used as the cut-off for grouping into “low” and “high” expression. For survival analysis, a Cox proportional hazard model was employed (univariate (UV); multivariate (MV)).
Results: qPCR analysis of candidate genes in the validation set revealed the prognostic role of 3 genes: High expression of insulin-growth factor binding protein 5 (IGFBP5; linked to tumor dormancy) was associated with inferior OS (p=0.012; UV). In intensively treated patients, high expression of neurotrophic tyrosine kinase receptor type 2 (NTRK2; acting on cell differentiation) was associated with superior OS (p=0.033; UV) and high expression of delta-like ligand 3 (DLL3; a Notch ligand) with inferior OS (p=0.038; UV). In addition, localization-dependent expression of NTRK2 was confirmed in the validation set. After adjustment for known prognostic factors, DLL3 overexpression proved to be an independent prognosticator for inferior OS in a multivariate model (p=0.017; MV).
Conclusions: Validating our candidate genes in an independent study sample revealed the localization-dependent expression of NTRK2. Furthermore, DLL3 was shown to be an independent negative prognosticator. Interestingly, all these genes are associated with stem cell self-renewal, dormancy and differentiation, emphasizing the significance of (cancer) stem cells on the clinical course of GBM.