Pro-angiogenic cellular and genomic expression patterns within glioblastoma influences dynamic susceptibility weighted perfusion MRI

Aim To investigate whether quantitative dynamic susceptibility-weighted contrast-enhanced (DSC) perfusion magnetic resonance imaging (MRI) metrics are influenced by cellular and genomic expression patterns of glioblastoma angiogenesis. Materials and methods Twenty-five stereotactic neurosurgical tissue samples were prospectively obtained from enhancing and non-enhancing tumour regions from 10 patients with treatment-naïve glioblastoma. Using monoclonal antibodies, histopathological features of angiogenesis were examined: total microvascular density, vascular morphology, and hypoxia. Angiogenic expression patterns of tissue samples were investigated using RNA microarrays. DSC perfusion MRI metrics were measured from the tissue sampling sites. MRI and histopathological variables were compared using Pearson's correlations. Microarray analysis was performed using false discovery rate (FDR) statistics. Results Thirteen enhancing and 12 non-enhancing MR image-guided tissue specimens were prospectively obtained. Enhancing tumour regions demonstrated a significant difference in DSC perfusion and histopathological metrics of angiogenesis when compared to non-enhancing regions. Four angiogenic pathways (vascular endothelial growth factor [VEGF], hypoxia inducible factor [HIF], platelet-derived growth factor [PDGF], fibroblast growth factor [FGF]; 25 individual genes) were significantly up-regulated within enhancing regions when compared to non-enhancing regions (adjusted p<0.05, FDR <0.05). A statistically significant correlation was observed between VEGF-A expression, microvascular density, microvascular morphology, and DSC perfusion MRI metrics (p<0.05). Conclusion Pro-angiogenic genomic and cellular expression patterns of treatment-naïve primary glioblastoma significantly influences morphological and physiological DSC perfusion metrics suggesting that expression levels of therapeutically relevant genetic signatures can be quantified using MRI.