Ki-67 ≥ 13% and Local Failure in WHO grade 2 and 3 Meningiomas: A Single-Institution Cohort Study

Ki-67 labeling index predicts recurrence in meningiomas, but proposed thresholds vary widely and lack validation in high-grade cohorts. We evaluated whether Ki-67 ≥ 13%, a literature-informed threshold positioned between mean proliferative indices for WHO grade 2 and grade 3 tumors, is associated with inferior local failure-free survival (LFFS) in grade 2 and 3 meningiomas.

This retrospective cohort study included 82 patients with WHO grade 2 (n = 70) or grade 3 (n = 12) meningiomas treated between 2005-2024. Ki-67 was dichotomized at 13% based on prior literature. Preoperative tumor volume was quantified using AI-assisted MRI autosegmentation. Kaplan-Meier and multivariable Cox regression analyses assessed associations between Ki-67, clinicopathologic features, and outcomes.

Median follow-up was 30.3 months. Higher Ki-67 was associated with inferior LFFS on univariable analyses. On multivariable analysis, Ki-67 ≥ 13% retained independent association with LFFS (HR 20.97, 95% CI 1.79-246.41, p = 0.015). A bivariable sensitivity analysis adjusting for Ki-67 ≥ 13% and WHO grade alone yielded a more stable estimate (HR 4.47, 95% CI 1.92-10.37, p < 0.001). Two-year LFFS was 41.9% for Ki-67 ≥ 13% versus 100% for Ki-67 < 13% (p < 0.001). Combined stratification revealed grade 2 tumors with Ki-67 ≥ 13% had outcomes (2-year LFFS 43.4%) approaching grade 3 disease (32.1%), while grade 2 tumors with Ki-67 < 13% demonstrated excellent control (100%). Preoperative tumor volume was not prognostic. Exploratory analysis showed that adjuvant radiotherapy was associated with improved LFFS and overall survival following both subtotal and gross total resection.

Ki-67 ≥ 13% was independently associated with local failure in this single-institution high-grade meningioma cohort. This literature-informed threshold identified a subset of grade 2 tumors with outcomes approaching grade 3 disease. Ki-67 may complement WHO grade for postoperative risk stratification, warranting prospective validation in independent cohorts.