Potential of pre-operative MRI features in glioblastoma to predict for molecular stem cell subtype and patient overall survival

Franziska Eckert; Katrin Ganser; Benjamin Bender; Jens Schittenhelm; Marco Skardelly; Felix Behling; Ghazaleh Tabatabai; Elgin Hoffmann; Daniel Zips; Stephan M Huber; Frank Paulsen

doi:10.1016/j.radonc.2023.109865

Potential of pre-operative MRI features in glioblastoma to predict for molecular stem cell subtype and patient overall survival

Radiother Oncol. 2023 Nov:188:109865. doi: 10.1016/j.radonc.2023.109865. Epub 2023 Aug 22.

Authors

Affiliations

¹ Department of Radiation Oncology, University of Tübingen, Germany; Medical University Vienna, Department of Radiation Oncology, Comprehensive Cancer Center Vienna, Vienna, Austria. Electronic address: franziska.eckert@meduniwien.ac.at.
² Department of Radiation Oncology, University of Tübingen, Germany.
³ Department of Diagnostic and Interventional Neuroradiology, University of Tübingen, Tübingen, Germany.
⁴ Department of Pathology and Neuropathology, University of Tübingen, Germany.
⁵ Department of Neurosurgery, University of Tübingen, Germany; Centre for Neurooncology, University of Tübingen, Germany.
⁶ Centre for Neurooncology, University of Tübingen, Germany.
⁷ Department of Radiation Oncology, University of Tübingen, Germany; Department of Radiation Oncology, Charité Universitaetsmedizin Berlin, Germany.

PMID: 37619660
DOI: 10.1016/j.radonc.2023.109865

Abstract

Aim of the study: A molecular signature based on 10 mRNA abundances that characterizes the mesenchymal-to-proneural phenotype of glioblastoma stem(like) cells (GSCs) enriched in primary culture has been previously established. As this phenotype has been proposed to be prognostic for disease outcome the present study aims to identify features of the preoperative MR imaging that may predict the GSC phenotype of individual tumors.

Material/methods: Molecular mesenchymal-to-proneural mRNA signatures and intrinsic radioresistance (SF₄, survival fraction at 4 Gy) of primary GSC-enriched cultures were associated with survival data and pre-operative MR imaging of the corresponding glioblastoma patients of a prospective cohort (n = 24). The analyzed imaging parameters comprised linear vectors derived from tumor volume, necrotic volume and edema as contoured manually.

Results: A necrosis/tumor vector ratio and to a weaker extent the product of this ratio and the edema vector were identified to correlate with the mesenchymal-to-proneural mRNA signature and the SF₄ of the patient-derived GSC cultures. Importantly, both parameter combinations were predictive for overall survival of the whole patient cohort. Moreover, the combination of necrosis/tumor vector ratio and edema vector differed significantly between uni- and multifocally recurring tumors.

Conclusion: Features of the preoperative MR images may reflect the molecular signature of the GSC population and might be used in the future as a prognostic factor and for treatment stratification especially in the MGMT promotor-unmethylated sub-cohort of glioblastoma patients.

Keywords: Glioblastoma; MGMT promoter methylation; MRI; Mesenchymal; Proneural; Stem cells.