Glioblastoma pseudoprogression and true progression reveal spatially variable transcriptional differences

Area of Science:

• Neuro-oncology
• Cancer immunology
• Computational pathology

Background:

• Post-surgical monitoring for glioblastoma (GBM) recurrence relies on imaging, but treatment-induced pseudoprogression mimics these signs.
• Accurate differentiation between true progression and pseudoprogression is vital for appropriate patient management.
• Current diagnostic methods face challenges due to similar imaging features and admixed histology.

Purpose of the Study:

• To validate current approaches for differentiating GBM progression from pseudoprogression.
• To characterize molecular and morphologic differences between true progression and pseudoprogression.
• To identify novel biomarkers for objective stratification of these entities.

Main Methods:

• Analysis of RNA sequencing data from recurrent GBM samples.
• nCounter RNA expression analysis of 48 clinical samples from second neurosurgical resections.
• Automated image processing and spatial expression analysis of histologic images.
• Unsupervised clustering of segmented histologic images.

Main Results:

• Gene expression pathways in pseudoprogression are characterized by immune activation, while progression shows increased cell cycle activity.
• Automated image analysis revealed challenges in differentiating cases with admixed histology.
• Unsupervised clustering of histologic images identified novel morphologic differences between progression and pseudoprogression.
• Spatial data indicated myeloid cell polarization associated with tumor recurrence.

Conclusions:

• Distinct molecular signatures and spatial immune profiles differentiate GBM progression from pseudoprogression.
• Morphologic analysis using unsupervised clustering offers new insights for pathological stratification.
• Understanding tumor-immune microenvironment evolution is key to targeting GBM recurrence.