Article
MRI angiogenesis parameters (VSI, CBV) are associated with specific genome-wide gene expression patterns in glioblastoma multiforme
Search Medline for
Authors
-
Dieter H. Heiland
- Department of Neurosurgery, Medical Center, University of Freiburg, Freiburg, Germany
-
D. Pfeifer
- Department of Hematology, Oncology and Stem Cell Transplantation, Medical Center, University of Freiburg, Freiburg, Germany
-
Irina Mader
- Department of Neuroradiology, Medical Center, University of Freiburg, Freiburg, Germany
-
Astrid Weyerbrock
- Department of Neurosurgery, Medical Center, University of Freiburg, Freiburg, Germany
| Published: | June 8, 2016 |
|---|
Outline
Text
Objective: Vessel size Imaging (VSI) and quantification of cerebral blood volume (CBV) are well-established MRI techniques to assess tumor vascularization in tumors. Currently no studies exist which link these MRI-based parameters of tumor vascularization with genetic profiles in malignant gliomas and to study a potential impact on progression-free (PFS) and overall survival (OS) in patients with glioblastoma multiforme.
Method: We retrospectively analyzed 25 patients with a glioblastoma between 2010 and 2014. In a presurgical MRI CBV and VSI were calculated in addition to standard imaging. RNA of tumor samples was extracted and analysed by gene-expression array (Affymetrix HuGene 2.0) and integrative analysis was performed using individual R-software pipelines. Networks were created by gene-set-enrichment analysis (GSEA r-tool). Differentially gene expression analysis was done by paired t-test (limma r-tool) and clustered in Cluster 3.0. PFS and OS of these subgroups were compared by Kaplan-Meier statistics.
Results: We analyzed VSI/CBV values and correlated (Spearman-ranked) them to genome-wide expression. Three main groups were identified that represent subgroups with high, medium and low VSI by unsupervised clustering of genes correlating with VSI (r>0.8). GSEA of genes ranked by positive or negative correlation to VSI was done for biological functions or pathway activation. We found VSI to be significantly associated to regulation of immune response and angiogenesis (p<0.05) and a negative enrichment of cell-cycle associated pathways (p<0.05). All three groups showed a significant (p<0.05) difference in PFS and a non-significant trend for a difference in OS. Interestingly, the subgroup with high VSI values had a better clinical outcome with longer PFS, possibly due to a better response to treatment. Correlation of CBV and genome-wide expression showed two main subgroups separated by unsupervised clustering. In patients with high CBV we detected an activation of NFκB-related pathways as KRAS (p<0.05) and a negative correlation to mitotic functions and DNA-repair mechanism (p<0.05). The subgroups did not differ with regard to PFS or OS.
Conclusions: We show that CBV/VSI-based data could be a valuable adjunct to basic neuroradiology diagnostics and may present a surrogate imaging marker for a specific genetic tumor subtype with a distinct clinical course. Further analysis with an increased number of patients to confirm these findings is pending.
