2026 Proffered Presentations
S054: INTEGRATED ANATOMIC AND MOLECULAR PREDICTORS OF RECURRENCE MENINGIOMAS
Patrick Bi1; Ivan Brown1; Kyle Montgomery1; Wenxuan Xiong2; Kristyn Galbraith, MD3; Jacob Ruzevick, MD4; Samuel Emerson, MD, PhD4; 1UW Medicine; 2UW Medicine Neurology & Neurosurgery; 3UW Medicine Neuropathology; 4UW Medicine Neurosurgery
Background: Meningiomas are the most common primary intracranial brain tumor. Accurate risk stratification, particularly for high risk or highly recurrent disease can guide early adjuvant therapy. Recent advancements in molecular profiling have led to enhanced diagnostic criteria, with the WHO classification of central nervous system tumors (CNS5) incorporating molecular biomarkers such as TERT promoter mutations and CDKN2A/B deletions for WHO grade 3 tumors. The cIMPACT-NOW update 8 suggests that additional molecular features including inactivation of BAP1, PBRM1 and SMARCE1 may be linked to specific meningioma subtypes, behavior and prognosis. However, while SMARCE1-mutant meningiomas demonstrate clear aggressive behavior, the impact of BAP1 and PBRM1 alteration remain uncertain. We assessed clinical, surgical, histopathological, and molecular predictors of recurrence, with particular focus on BAP1 loss and evaluated the cIMPACT-NOW risk classification in a single-center cohort.
Methods: We retrospectively reviewed 142 patients from a single center with WHO grade 1-3 intracranial meningiomas who underwent resection and molecular profiling between January 2015 and May 2025. Recurrence-free survival (RFS) was analyzed using Cox proportional hazards models. Clinicopathological features included BAP1 loss, Ki-67 proliferation index, brain invasion, and extent of resection. Sparse molecular markers such as ≥2 whole-arm copy number loss (chromosomes 6, 10, 14, 18) and codeletion 1p/22q were explored descriptively. Univariable analysis was conducted for all predictors. Multivariable analyses included a core model (cohort without brain invasion) and a secondary model (reduced cohort including brain invasion).
Results: The median age of the study population was 54.5 years (IQR 43 - 67 years), with 94 (66.2%) women. Median follow-up was 5.4 months. Overall recurrence rate was 12% (17 recurrences). Recurrence rates stratified by WHO grade were 4.3% (95% CI 0.5–14.5) for grade 1, 14.1% (95% CI 7.3–23.8) for grade 2, and 16.7% (95% CI 2.1–48.4) for grade 3 meningiomas. BAP1 loss was significantly associated with recurrence on univariable analysis (log-rank p = 0.009) but did not remain an independent predictor in multivariable models. In the core Cox model, Ki-67 was independently associated with shorter RFS (HR 1.07 per 1% increase, 95% CI 1.02–1.12, p < 0.01). The secondary Cox model demonstrated that brain invasion was also independently associated with recurrence (HR 8.78, 95% CI 1.74–44.2, p < 0.01), while the effect of Ki-67 was attenuated. ≥2 whole-arm copy number loss and 1p/22q codeletion were rare and descriptively associated with more recurrent tumors, but too infrequent for formal analysis.
Conclusion: BAP1 loss was associated with higher recurrence on univariable analysis, but did not retain independent significance, possibly as a result of the limited number of events. Ki-67 remained a consistent predictor of meningioma in the core model, reinforcing its prognostic value in meningioma. Brain invasion also emerged as an independent risk factor when included, though variable presence of brain parenchyma on pathology to detect invasion resulted in a reduced cohort size. Together, our findings support an integrated clinical–molecular model for recurrence risk in meningioma. Validation in larger, multicenter cohorts will be necessary to clarify the roles of BAP1, and PBRM1.