2026 Proffered Presentations
S248: HYPEROSTOSIS SUBTYPE AS A RADIOLOGIC SIGNATURE OF TRAF7 MUTATIONS IN MENINGIOMAS
Aymen S Kabir1; Abraham Dada1; Wesley Shoap, MD2; Rithvik Ramesh1; Daniel Quintana1; Michael A Torres-Espinosa1; Christian Jimenez1; Robert C Osorio, MD1; Kanish Mirchia, MD1; Charlotte D Eaton, PhD1; David R Raleigh, MD, PhD1; Ezequiel Goldschmidt, MD, PhD1; 1University of California, San Francisco; 2Louisiana State University
Introduction: Meningioma-induced hyperostosis (MIH) is a frequent radiographic finding in meningiomas, yet its underlying mechanisms remain poorly understood. Recent studies suggest a possible link between somatic mutations, particularly TRAF7 mutations, and bone remodeling. While hyperostosis has traditionally been treated as a binary phenomenon, this study investigates whether MIH represents a heterogeneous process with distinct radiological subtypes and genetic associations.
Objective: To define radiographic subtypes of MIH and evaluate their association with TRAF7 mutation status.
Methods: The authors retrospectively reviewed the records and imaging of patients with meningiomas with available genetic panels resected between 2021-2024 at a single institution. Patient characteristics examined in the analysis included patient demographics, tumor location and size, somatic DNA mutations through capture-based next-generation sequencing, and imaging characteristics including the presence and extent of peritumoral edema and enhancing pattern. CT images were evaluated for bone involvement, presence of osteolysis or hyperostosis, and type of hyperostosis. Type I hyperostosis was defined as hyperostosis with destruction of cortical architecture while Type II hyperostosis was defined by the preservation of cortical structure. Associations with TRAF7 mutations were analyzed using univariate tests and multivariate logistic regression. Furthermore, Random Forest, Gradient Boosting Machine (GBM), and Support Vector Machine (SVM) learning models were used to assess non-parametric relationships.
For a subset of patients, ImageJ was used to perform quantitative bone density analysis. Regions of interest (ROIs) were placed in the outer cortex, central zone, and inner cortex. Grayscale histograms were normalized to compare pixel intensity distributions across normal bone and hyperostosis subtypes.
Results: Among 384 tumors, 54 (14.1%) exhibited hyperostosis—23 (6.0%) Type I and 31 (8.1%) Type II. TRAF7 mutations were significantly enriched in Type I hyperostosis (78.3%) compared to Type II (25.8%) and non-hyperostotic tumors (17.0%) (p < 0.001). TRAF7 mutations were also more likely to be located in the skull base (82.9% vs 55.6%, p<0.001), have smaller maximal tumor dimension (3.2 vs 4.0 cm, p<0.001), and more likely to have large-volume T2 hyperintensity (21.6% vs 7.5%, p=0.016). Multivariable analysis identified skull base location (OR:3.11, p=0.002), smaller maximal tumor dimension (OR:0.72, p=0.006), homogenous contrast enhancement (OR:5.57, p=0.001), large-volume T2 hyperintensity (OR:7.64, p=0.001), and Type I hyperostosis (OR:18.73, p=0.001) as independent predictors of TRAF7 mutations. Machine learning models confirmed these predictors, with gradient boosting achieving the highest accuracy (AUC = 0.854). A decision tree classifier revealed that tumors with Type I hyperostosis and absence of T2 hyperintensity carried a 92.3% TRAF7 mutation rate.
Quantitative analysis of bone density showed that in normal bone, a characteristic bimodal pattern was observed with peaks corresponding to the cortical and cancellous bone. While type II hyperostosis preserved the bimodal pattern resembling normal bone, type I hyperostosis demonstrated a more homogenous distribution with broad curves and minimal distinction between cortical and cancellous ROIs highlighting the disruption of normal cortical-cancellous architecture.
.jpg)
Conclusions: Meningioma-induced hyperostosis represents a radiographically heterogenous phenomenon with differing associations to genetic mutations. Hyperostosis with disrupted cortical architecture is strongly associated with TRAF7 mutations and may represent a key feature of this mutation’s radiographic phenotype.