Feasibility of ONCOhabitats for Surgical and Treatment Planning in IDH-Wildtype Glioblastoma (SINUE)

Imaging biomarkers-including relative cerebral blood volume (rCBV), volume (cm³), and radiomic vascular heterogeneity metrics-will be extracted from the peritumoral infiltrated edema (IPE) habitat using the ONCOhabitats platform on preoperative MRI. These biomarkers will be quantitatively assessed for their association with overall survival. Overall survival is defined as the time from surgical resection to death from any cause, with a follow-up of up to 24 months. Statistical analysis will include Kaplan-Meier survival estimation and Cox proportional hazards regression models.

Gene expression analysis will be performed on tissue samples obtained from the peritumoral infiltrated edema (IPE) region during surgical resection. RNA sequencing (RNA-seq) or targeted transcriptomic panels will be used to identify differential expression of immunosuppressive and immune-regulatory genes (e.g., PD-L1, CTLA-4, TGF-β, IDO1) in patients with IDH-wildtype glioblastoma. These molecular profiles will be correlated with clinical response to immunotherapy, as defined by progression-free survival (PFS) and overall survival (OS),

Histopathological analysis will be conducted on tissue samples obtained from four MRI-derived habitats segmented using the ONCOhabitats software platform:

• HAT (High Angiogenic Tumor) and LAT (Low Angiogenic Tumor), corresponding to tumor regions
• IPE (Infiltrated Peritumoral Edema) and VPE (Vasogenic Peritumoral Edema), corresponding to peritumoral edema regions

Tissue sections will be evaluated using hematoxylin and eosin (H&E) staining and immunohistochemistry to characterize features such as cellularity, necrosis, microvascular proliferation, immune cell infiltration, and other relevant histopathological traits. The aim is to describe and compare the biological composition of tumor versus edema habitats.

Tissue samples will be collected from three MRI-derived habitats defined by the ONCOhabitats platform in patients with IDH-wildtype glioblastoma:

• IPE (Infiltrated Peritumoral Edema)
• HAT (High Angiogenic Tumor)
• VPE (Vasogenic Peritumoral Edema)

Immunohistochemical and transcriptomic analyses will be performed to quantify the expression of immunosuppressive markers, including but not limited to PD-L1, CTLA-4, TGF-β, and IDO1. The level of immunosuppressive activity in each habitat will be characterized to explore spatial heterogeneity in the tumor microenvironment.

Quantitative perfusion biomarkers-such as relative cerebral blood volume (rCBV), cerebral blood flow (CBF), and Ktrans-will be obtained from preoperative MRI processed with the ONCOhabitats platform. These imaging-derived parameters will be correlated with molecular markers obtained from tumor and edema tissue samples, including MGMT promoter methylation, IDH mutation status, and expression of immune-related genes (e.g., PD-L1, CTLA-4). The aim is to explore associations between vascular imaging features and the underlying molecular profile of IDH-wildtype glioblastoma.

Multimodal biomarkers-including perfusion imaging features (e.g., relative cerebral blood volume [rCBV], cerebral blood flow [CBF]), histopathological traits (e.g., vascular proliferation, immune infiltration), molecular alterations (e.g., MGMT promoter methylation, gene expression profiles), and cellular markers (e.g., CD3+, CD8+ lymphocyte density)-will be integrated using multivariate models and artificial intelligence (AI) techniques to predict clinical outcomes in patients with IDH-wildtype glioblastoma. These outcomes include diagnosis, prognosis (e.g., overall survival), and treatment response-based patient stratification.

Histopathological, molecular, and imaging biomarkers will be analyzed to identify features that allow for patient stratification regarding two therapeutic approaches: (1) surgical resection of the infiltrated peritumoral edema (IPE) region, and (2) administration of local therapies such as immunotherapy in the resection cavity. Biomarkers may include, among others, perfusion imaging parameters (e.g., rCBV), gene expression profiles related to immune activation or suppression, and histopathological indicators of tumor infiltration and immune cell presence. These variables will be evaluated for their potential to guide individualized treatment decisions in IDH-wildtype glioblastoma.

The association between MGMT promoter methylation status and relative cerebral blood volume (rCBV)-measured from preoperative dynamic susceptibility contrast (DSC) MRI using the ONCOhabitats platform-will be evaluated to stratify patients with IDH-wildtype glioblastoma according to their response to temozolomide (TMZ) treatment. Response will be assessed based on progression-free survival (PFS) and overall survival (OS) over a follow-up period of up to 24 months. The predictive value of the combined biomarkers will be analyzed using multivariate models and ROC curve analysis.

MGMT promoter methylation status will be assessed in tumor tissue samples using two methods: immunohistochemistry (IHC) and multiplex ligation-dependent probe amplification (MLPA). The results of both techniques will be compared to evaluate their concordance and their respective predictive value for response to temozolomide (TMZ) in patients with IDH-wildtype glioblastoma. Response will be measured using progression-free survival (PFS) and overall survival (OS) over a 24-month follow-up period.