Logarithmic and Conventional MRI Biomarkers of the Tumor-Edema Relationship in Meningioma Derived From Reproducible
Antonio Navarro-Ballester1, Rocío Fleta-Gascón1
1Department of Radiology, General University Hospital of Castellón, Castellón, Spain.
Rationale And Objectives:
Peritumoral edema is a common imaging finding in intracranial meningiomas and has been associated with histologic grade and tumor aggressiveness. However, quantitative magnetic resonance imaging (MRI) biomarkers of the tumor-edema relationship remain insufficiently explored. This study evaluated volumetric tumor-edema metrics as predictors of meningioma grade.
Materials And Methods:
In this retrospective study, 187 consecutive patients (median age, 61 years; 118 women) with histopathologically confirmed intracranial meningioma who underwent preoperative MRI between 2003 and 2025 were included. Tumor and peritumoral edema volumes were obtained using manual volumetric segmentation, with interobserver reproducibility evaluated in a randomly selected subset of 20 cases independently segmented by both readers. Six quantitative tumor-edema indices were calculated. Diagnostic performance for differentiating grade I from grade II-III meningiomas was assessed using receiver operating characteristic analysis and logistic regression.
Results:
Histologically, 145 of 187 tumors (77.5%) were grade I and 42 (22.5%) were grade II-III. Most edema-derived metrics showed modest but consistent discriminatory performance (area under the receiver operating characteristic curve [AUC] range, 0.61-0.64). The infiltration efficiency index showed the highest discriminatory performance (AUC, 0.64; P = .006), whereas the allometric index yielded an AUC of 0.62 (P = .014). Edema fraction was significantly associated with tumor grade at logistic regression (odds ratio, 3.15; P = .042). Absence of measurable peritumoral edema occurred in 51 tumors (27.3%), including 46 grade I tumors.
Conclusion:
Volumetric MRI biomarkers describing the tumor-edema relationship provide modest but consistent discrimination between grade I and grade II-III meningiomas. Scaling-based metrics showed slightly higher diagnostic performance than conventional edema-tumor indices, although overall discriminatory performance remained limited.
