Heterogeneity phenotypes in recurrent glioblastoma: a multimodal MRI-based spatial mapping framework for precision...

Area of Science:

• Neuroimaging
• Oncology
• Radiomics

Background:

• Recurrent glioblastoma (rGBM) presents significant treatment challenges due to its heterogeneity.
• Accurate characterization of intratumoral heterogeneity is crucial for predicting treatment response and patient outcomes.

Purpose of the Study:

• To develop a multimodal magnetic resonance imaging (MRI)-based spatial mapping framework for characterizing rGBM intratumoral heterogeneity.
• To identify distinct imaging subregions and classify heterogeneity phenotypes predictive of treatment response and survival.

Main Methods:

• 140 rGBM patients underwent diffusion-weighted imaging (DWI) and dynamic contrast-enhanced MRI (DCE-MRI).
• Multimodal Imaging Subregions (MIS) were identified by colocalizing apparent diffusion coefficient (ADC) and DCE-MRI features.
• Hierarchical clustering defined imaging phenotypes, and heterogeneity was quantified using entropy and Moran's I.

Main Results:

• A treatment-resistant subregion (MIS4) characterized by low ADC and slow enhancement was identified.
• Three distinct imaging phenotypes demonstrated significant prognostic stratification across different treatment regimens (CCNU, bevacizumab-based therapies).
• Imaging phenotypes independently predicted progression-free survival (PFS) and overall survival (OS).

Conclusions:

• The developed framework allows for spatially resolved, phenotype-based analysis of rGBM heterogeneity using routine MRI.
• Identified imaging phenotypes serve as non-invasive biomarkers for personalized treatment planning and outcome prediction in rGBM.