The utility of intraoperative magnetic resonance imaging in recurrent glioblastoma surgery: A volumetric analysis

Objective: Intraoperative magnetic resonance imaging (iMRI) for the resection of newly diagnosed glioblastoma has shown efficacy comparable to fluorescence guidance; however, its role in recurrent glioblastoma (recGB) remains unclear. This study evaluated the efficacy and safety of iMRI-guided recGB surgery. Methods: Consecutive patients with recGB undergoing iMRI-guided re-resection between 2015 and 2023 were retrospectively analyzed. Contrast-enhancing (CE) tumor volumes from preoperative (3T), intraoperative (1.5T), and early postoperative (epMRI, 3T) scans were assessed per RANO Resect criteria. The impact of iMRI on intraoperative decision-making for additional resection (AR), functional outcomes, and survival was evaluated. Results: A total of 150 patients were included. Complete CE tumor resection was intended in 85% (n = 127). iMRI prompted AR in 77% (n = 115), increasing RANO class 1-2 resections from 58% on iMRI to 85% on epMRI. A "RANO-switch" (conversion from class 3 on iMRI to class 1-2 on epMRI) occurred in 27% (n = 41). Preoperative CE tumor volume was the only independent predictor of AR (p = 0.001), with 6.9 ml identified as the cut-off above which AR was significantly more likely to achieve a "RANO-switch." Neurological deterioration was not associated with AR (transient: 15% vs 16%; permanent: 2% vs 3%). Median survival after re-resection was 11 months. Gross total resection (residual tumor volume ≤0.175 ml), but not RANO class, independently predicted improved survival (HR 0.50, p = 0.006). Conclusion: In recGB, where a real-time intraoperative information about RTV could be relevant, iMRI significantly improved the extent of resection without increasing the risk of neurological deficits. Larger preoperative CE tumor volumes were associated with a greater resection benefit from iMRI, supporting its efficacy and safety in recGB surgery.