gms | German Medical Science

Objective: Glioblastoma multiforme (GBM) WHO°IV is the most frequent and at the same time prognostically most unfavorable brain tumor with a heterogeneous appearance in MRI. Recent publications have shown that GBMs grouped into three phenotypically different subtypes are associated with significant differences in overall survival and signaling activity. We assume, that GBM subtypes differ not only phenotypically and in terms of pathway activity, but also in terms of functional neuroplasticity.

Methods: The sample consists of 55 retrospectively selected patients with glioblastoma who have been examined by MRI. The patient sample was then subdivided into two prognostically different subgroups based on the morphological appearance of the tumor in T1C-weighted MR-images according to defined criteria based on recent publications. We analyzed the percent signal change of fMRI language mapping using six language-relevant Regions of Interest (ROIs) as well as two control ROIs in the occipital lobe and compared them between both GBM growth pattern groups. In addition, we calculated group ROI-to-ROI analyses for different implemented network ROIs (e.g. Default Mode network, Salience network, etc.) using the CONN toolbox to detect the average effect of the tumor/lesion on the functional connectivity.

Results: When analyzing the percent signal changes in six language-related ROIs and two control ROIs, there was no significant difference between the two GBM growth pattern groups in any of the regions of interest. In addition, functional connectivity was compared between both GBM growth pattern groups. This revealed several differences within and between functional networks. The results showed that there were correlations that were more pronounced in patients with uniform contrast enhancement as well as correlations that were more pronounced in patients with diffuse contrast enhancement. However, these were not statistically significant once corrected for multiple testing.

Conclusion: Contrary to the assumption that the prognostically less favorable group of tumors with diffuse contrast enhancement should also show a lower BOLD signal reflecting the clinical picture of this tumor type, these patients showed a higher percent signal change some areas compared to the group of tumors with uniform contrast enhancement. Data of a healthy control group will be additionally analyzed to better evaluate the magnitude of the individual differences.