IMG-102. Radio-pathomic assessment of tumor dynamics following radiotherapy in glioblastoma patients
Year: 2025
Type: conference-abstract
Abstract: Following initial surgical resection, glioblastoma patients typically receive radiotherapy with adjuvant temozolomide to treat the resection cavity and non-enhancing areas beyond its margin. However, radiotherapy’s impact on traditional imaging signatures can confound the interpretation of post-therapy scans, challenging treatment response assessment, particularly in non-enhancing regions. This study uses previously developed radio-pathomic maps (RPMs) sensitive to non-enhancing tumor presence to test the hypothesis that prognostically valuable pathological response patterns can be visualized following radiation. RPMs were generated for clinical imaging timepoints following radiotherapy through initial radiographic recurrence for 30 primary glioblastoma patients. These maps were developed using autopsy tissue samples aligned to clinical MRI (i.e., pre- and post-contrast T1, FLAIR, and ADC) to predict cell and ECF densities. Contrast enhancement and non-enhancing FLAIR hyperintensity masks were manually drawn, and mean radio-pathomic values were extracted for each region of interest. Linear mixed effect models assessed longitudinal changes post-radiotherapy using a random effect for subject to account for repeated measures. Initial assessments for sex-related differences showed increased ECF density growth in both enhancing (B=2.35, p=0.022) and non-enhancing regions (B=1.99, p=0.05) in female patients relative to male patients. Adjusting for sex and contrast-enhancing volumes as covariates, shows reduction in cell density within the non-enhancing region following treatment (B=-2.31, p=0.023). Cross-sectional assessment on the scan prior to radiation revealed a correlation between non-enhancing ECF density and progression-free survival (R=0.47, p=0.021), and analysis of the scan of first radiographic progression identified a positive association between ECF density within enhancement and overall survival. These results indicate that radiation-related changes in the non-enhancing tumor margin can be visualized using radio-pathomic mapping, highlighting a potential dependence on sex and other demographic differences. Additionally, protective prognostic factors were identified both prior to therapy and at radiographic recurrence, with high ECF density within enhancement acting as a potential biomarker for pseudo-progression.
Source: Neuro-Oncology
Institution: Medical College of Wisconsin
Language: English
Subfield: Genetics
Field: Medicine
Domain: Health Sciences
Open Access status: bronze
OpenAlex ID: W4416085344