Related Experiment Video
Updated: Apr 22, 2026
Translational Orthotopic Models of Glioblastoma Multiforme
Published on: February 17, 2023
Prospective biopsy-controlled validation of an AI model for predicting glioblastoma infiltration: Results from the
Santiago Cepeda1,2, Elena Hernando-Pérez3,4,5, Enrique Pérez-Riesgo3
1Department of Neurosurgery, Río Hortega University Hospital, Valladolid, Spain.
Background:
Glioblastoma recurrence is driven by diffuse microscopic infiltration beyond the contrast-enhancing tumor margin. GlioMap is an open-access AI model predicting voxelwise infiltration and recurrence risk from multiparametric MRI. This prospective study aimed to validate GlioMap's biological accuracy and prognostic relevance through histopathological assessment, transcriptomic profiling, and survival analysis within the SupraGlio trial (NCT05735171).
Methods:
Patients with newly diagnosed glioblastoma underwent neuronavigated biopsies targeting AI-predicted high-risk (HRoR) and low-risk of recurrence (LRoR) regions beyond the contrast-enhancing tumor. Histopathological infiltration served as the ground truth, and transcriptomic profiling characterised each region's molecular phenotype. Model performance was evaluated using accuracy and area under the receiver operating characteristic curve (AUC). Survival analyses assessed the prognostic value of postoperative HRoR volume.
Results:
Fifty-eight biopsies from 27 patients were analyzed. GlioMap achieved 0.81 accuracy (95% confidence interval [CI], 0.71-0.91) and 0.84 AUC (95% CI, 0.73-0.93) for histologically confirmed infiltration. Transcriptomic analysis of 48 samples from 16 patients revealed progressive upregulation of invasion- and angiogenesis-related genes (CD44, CHI3L1, STAT3, VEGFA) and downregulation of neuronal markers (MBP, GABRA1) from LRoR to HRoR regions and the tumor core, confirming a neural-to-mesenchymal gradient. Postoperative HRoR volume >1.6 cm³ predicted shorter overall survival (P = .04) and progression-free survival (P = .008).
Conclusions:
To our knowledge, this study provides the first prospective, biopsy-controlled, molecular validation of an AI model for mapping glioblastoma infiltration. By accurately identifying histologically and transcriptionally infiltrated regions, GlioMap offers a biologically grounded imaging biomarker that could guide extended resection and personalized radiotherapy planning, potentially improving tumor control and patient outcomes.
Insights
GlioMap, an AI tool, accurately maps glioblastoma infiltration using MRI, validated by biopsies and gene expression. This AI biomarker can guide personalized treatment for better patient outcomes.
Area of Science:
- Artificial Intelligence in Oncology
- Neuro-oncology Imaging Biomarkers
- Molecular Profiling of Brain Tumors
Background:
- Glioblastoma recurrence is driven by diffuse microscopic tumor infiltration.
- Current imaging methods struggle to delineate the full extent of tumor spread.
- GlioMap is an AI model designed to predict infiltration and recurrence risk from MRI scans.
Purpose of the Study:
- To prospectively validate the biological accuracy of GlioMap.
- To assess GlioMap's prognostic relevance in glioblastoma patients.
- To correlate AI-predicted infiltration with histopathology and transcriptomics.
Main Methods:
- Prospective study within the SupraGlio trial (NCT05735171).
- Neuronavigated biopsies of AI-identified high-risk (HRoR) and low-risk (LRoR) regions.
- Histopathological assessment for ground truth infiltration.
- Transcriptomic profiling to characterize molecular phenotypes.
- Survival analysis based on postoperative HRoR volume.
Main Results:
- GlioMap achieved 0.81 accuracy and 0.84 AUC for predicting infiltration.
- Transcriptomic analysis revealed a neural-to-mesenchymal gradient from LRoR to HRoR regions.
- Upregulation of invasion/angiogenesis genes (e.g., VEGFA) and downregulation of neuronal markers observed.
- Postoperative HRoR volume >1.6 cm³ predicted shorter overall and progression-free survival (P=.04 and P=.008, respectively).
Conclusions:
- This study provides the first prospective, biopsy-controlled validation of an AI model for glioblastoma infiltration mapping.
- GlioMap accurately identifies histologically and transcriptionally infiltrated regions.
- GlioMap serves as a biologically grounded imaging biomarker for guiding surgical and radiotherapy planning.
