Medical Student University of California, San Francisco University of California, San Francisco
Introduction: Despite recent developments in meningioma classification systems, WHO grades remain a robust predictor of tumor recurrence. Accurate pre-operative identification of Grade 1 meningiomas is crucial for informed management decisions and effective patient counseling. This study aimed to establish clinically single-slide MR and CT, sensitive and specific markers for low-grade meningioma to develop automated image recognition software.
Methods: Patients with newly diagnosed meningioma from a single center between 2021 and 2024 were included. Radiographic features were manually curated from CT and MRI sequences, including location, size (maximal tumor dimension), T1 enhancement pattern, T2 signal change, bone involvement, presence of osteolysis, and hyperostosis type (none,Type 1,Type II). Type 1 hyperostosis was defined as hyperostosis without normal cortical-cancellous architecture, while Type 2 referred to the preservation of normal bone structure. Classical statistics and three machine learning models—Random Forest, Support Vector Machine, and Gradient Boost Machine—were employed to assess nonparametric relationships between features.
Results: 268 Grade 1 meningiomas were compared to 116 Grade 2-3 meningiomas. In comparison, Grade 1 to meningiomas were more likely in females (73.9%vs.51.7%,p < 0.001), located at the skull base (69.8%vs.42.2%,p < 0.001), had smaller size (3.4±1.5vs.4.6±1.9 ,p < 0.001), had less sinus invasion (21.6vs.36.9,p=0.006) were homogenous enhancing (78.7%vs52.6%,p < 0.001), and exhibited less T2-signal brain change (42.5%vs72.4%,p < 0.001). After controlling for interactions, multivariate logistic regression analysis found skull base location (OR:2.12,P=0.006), and size (OR 0.76,p=0.001) to be significantly associated with Grade 1 meningiomas. The RF model achieved the highest performance on the test set (AUC:0.94,Accuracy: 0.86). The top features of the RF model included tumor size, enhancement pattern, and location. In subgroup analysis 92.2% of meningiomas located at the skull base with no signal change and homogenous enhancement were Grade 1.
Conclusion : In addition to the known skull base location and smaller size, Grade 1 meningioma pathology is associated with the absence of T2 signal change and homogenous enhancement pattern. The presence these features is highly predictive of a Grade 1 meningioma and can help making decisions regarding surgical removal, counseling and follow-up.
.jpg)
