Dieter Henrik Heiland1, Thomas Lange2, Ralf Schwarzwald3, Dietmar Pfeifer4, Karl Egger3, Horst Urbach3, Astrid Weyerbrock1, and Irina Mader3
1Dept. of Neurosurgery, University Medical Center Freiburg, Freiburg, Germany, 2Dept. MR Physics, Dept. of Radiology, University Medical Center Freiburg, Freiburg, Germany, 3Dept. of Neuroradiology, University Medical Center Freiburg, Freiburg, Germany, 4Department of Hematology, Oncology and Stem Cell Transplantation, University Medical Center Freiburg, Freiburg, Germany
Synopsis
The purpose of this work
was to search for a connection between metabolites observed by proton magnetic
resonance spectroscopy of glioblastomas and tumor genetics. Two specific
pathways could be identified, one belonging to NAA and discriminating an astroglial versus oligo/neural subgroup. Another one was related to Cr also
distinguishing between two subgroups, one attributed to apoptosis and another one to the
PI3K-AKT-mTOR signaling cascade. Purpose
Connections between magnetic
resonance imaging of glioblastomas and tumor genetics have previously been described
1, 2. A correlation between the choline concentration
as measured with proton magnetic resonance spectroscopy (MRS) and a
calcium-sensing receptor expression in breast cancer has also recently been
shown
3. The purpose of this study was to find out
whether metabolite concentrations obtained by proton MRS are also related to
tumor genetics of glioblastomas.
Methods
Fifteen patients with glioblastomas receiving
presurgical proton MRS (2D CSI, PRESS, TR=1.5s, TE=30ms,
voxel size 1x1x1.5cm
3) were investigated on a 3T system.
Quantification was performed by LCModel
4,
e.g. Figure 1,
and ipsilateral/contralateral metabolite ratios were determined. Tumor
samples were obtained at a predefined region (contrast-enhancing tumor, controlled by
neuro-navigation). RNA was extracted and analyzed by gene-expression array
(Affymetrix HuGene 2.0). Positively and negatively correlated genes (r>0.8,
p<0.05) were extracted and unsupervised clustered to identify
expression-based subgroups. Gene ontology analysis was performed and visualized
in networks build by Cytoscape 2.0, Figure 2.
Gene-set-enrichment analysis (GSEA) was performed to identify subgroup-specific
pathway activation (Figures 3 and 4). All analysis
was done in individually designed r-software based pipelines including bioconductor packages.
As clinical data, the progression-free
survival (PFS) was taken and analyzed by
Kaplan-Meier statistics. All p-values given are corrected by false discovery
rate.
Results
N-acetylaspartate [“NAA”, sum of (NAA +NAAG)]
in contrast-enhancing tumor was positively correlated (r=0.45, p<0.05) to
progression-free survival. Cluster analysis and Gene Set Enrichment Analysis (GSEA)
identified an expression subgroup corresponding to lower NAA signal intensity.
This subgroup had a significant enrichment of astroglial genes and
pro-oncogenetic pathways as KRAS (p<0.01), JAK/STAT (p<0.01) and EGF (p<0.01).
A subgroup with higher NAA signal intensities corresponding to gene expression showed
a high enrichment of oligodendrocyte and neural signature genes and activation
of pro-apoptotic pathways as p53 (p<0.001). Kaplan-Meier statistics of both
subgroups with corresponding low/high NAA signal intensity showed significant PFS differences of both groups (p=0.01, Figure 3).
According to creatine (Cr) in contrast-enhancing
tumor, two different subgroups of gene expression could be assigned. Lower Cr was
accompanied by an activation of apoptosis, inflammation, and p53 as "the
guardian of the genome", whereas higher Cr came along with MYK, PI3K-AKT-mTOR
activation as response to necrosis. Survival analysis could not detect any
significant difference (p=0.21, Figure 4).
Discussion
The fact
that high levels of NAA correspond to an enrichment of oligo-neural genes is
corroborated by the integrative work of
Baslow showing oligodendrocytes as target of NAA in a tri-cellular metabolism
pathway of NAA and NAAG 5.
Creatine
comes into the brain by passing the blood-brain-barrier, where Cr transporters
(CrT) are expressed and localized in the endothelial cells. Cr is actively
transported into the extracellular space by neurons and oligodendrocytes (cells
expressing CrT), but not by astrocytes. Besides this specific preference of
cells, Cr is also thought to have a neuroprotective effect 6.
In
conclusion, the gene expression of the investigated glioblastomas is mirrored
by the metabolites NAA and Cr, and their abundance to neurons, oligodendrocytes
and astrocytes.
Acknowledgements
The authors thank Mr. Hansjörg Mast for his help with the acquisition of the data.References
1 Diehn, M., et al., Identification of noninvasive imaging surrogates for brain tumor
gene-expression modules. Proc Natl Acad Sci U S A, 2008. 105(13): p. 5213-8.
2 Jamshidi, N.,
et al., Illuminating radiogenomic
characteristics of glioblastoma multiforme through integration of MR imaging,
messenger RNA expression, and DNA copy number variation. Radiology, 2014. 270(1): p. 1-2.
3 Baio, G., et
al., Correlation between Choline Peak at
MR Spectroscopy and Calcium-Sensing Receptor Expression Level in Breast Cancer:
A Preliminary Clinical Study. Mol Imaging Biol, 2015. 17(4): p. 548-56.
4 Provencher,
S.W., Estimation of metabolite
concentrations from localized in vivo proton NMR spectra. Magn Reson Med,
1993. 30(6): p. 672-9.
5 Baslow, M.H., Evidence that the tri-cellular metabolism of
N-acetylaspartate functions as the brain's "operating system": how
NAA metabolism supports meaningful intercellular frequency-encoded
communications. Amino Acids, 2010. 39(5):
p. 1139-45.
6 Andres, R.H.,
et al., Functions and effects of creatine
in the central nervous system. Brain Res Bull, 2008. 76(4): p. 329-43.