Article
Integrative analysis of magnetic resonance spectroscopy (MRS) and genome-wide expression
Search Medline for
Authors
-
Dieter H. Heiland
- Department of Neurosurgery, University Medical Center Freiburg, Freiburg, Germany
-
D. Pfeifer
- Department of Hematology, Oncology and Stem Cell Transplantation, University Medical Center Freiburg, Freiburg, Germany
-
Irina Mader
- Department of Neuroradiology, University Medical Center Freiburg, Freiburg, Germany
-
Astrid Weyerbrock
- Department of Neurosurgery, University Medical Center Freiburg, Freiburg, Germany
| Published: | June 8, 2016 |
|---|
Outline
Text
Objective: High-grade gliomas (HGG) are the most common brain tumors with an average survival of 14 months. Proton magnetic resonance spectroscopy (H-MRS) is a quantitative MR imaging technique often used to complement conventional MR imaging with specific metabolic information. Our aim was to understand metabolic peaks of the MRS and find correlated genetic-pathway activity. Further, we want to identify clinical relevant parameters to improve further neuroradiology diagnostics.
Method: We retrospectively analyzed 15 Patients with a glioblastoma between 2010 and 2014. All patients underwent a presurgical MRS. Spectroscopic data were analysed and normalised by the department of Neuroradiology. In a following surgery we took tumour samples at a predefined region (controlled by neuro-navigation). RNA was extracted and controlled by Bioanalyser. Expression analyses were done by gene-expression array (Affymetrix HuGene 2.0) in the department Hematology, Oncology and Stem Cell Transplantation. Integrative analysis were performed using individual R-software pipelines. Networks were created by gene-set-enrichment analysis (GSEA r-tool) and Cytoscape 2.0. Differentially gene expression analysis were done by paired t-test and clustered in Cluster 3.0. Data visualization was performed by Java Treeviewer.
Results: All identified metabolite had specific expression profiles. We mapped all biological function and associated pathways to describe each metabolite influence on genome-wide expression changes. By unsupervised clustering of positively or negatively correlated metabolite signal intensity (r>0.8/r>-0.8) we found interesting subgroups that were significantly different in progression free survival (p<0.05 for NAA INO). We identified N-acetyl-aspatate (NAA) as significantly positively correlated (cor=0.45) to progression-free survival. We found pro-oncogenetic pathways as KRAS (p<0.001), JAK/STAT(p<0.001) and EGF(p<0.001) in patients with low NAA signal intensity. Patients with a high peak of NAA showed an enrichment of oligo-genes and activation of pro-apoptotic pathways as p53 (p<0.001).
Conclusions: We show that MRS data could improve basic neuroradiology diagnostics and helps to identify individual genetic tumor subtypes. Further analysis with an increased number of patients should be done in the future.
