Is FLAIRectomy Directly Correlated with Prolonged Survival in Glioblastoma? A Prospective National Multicenter Study on Correlation Between Extent of Tumor Resection and Clinical Outcome

BACKGROUND AND OBJECTIVES: Several articles have demonstrated a positive correlation between glioblastoma supramarginal resection, based on MRI fluid-attenuated inversion-recovery (FLAIR) sequences (ie, FLAIRectomy), and prolonged survival. This study analyses the efficacy, safety, and reliability of FLAIRectomy in a multicentric cohort of patients, correlating the extent of FLAIR resection (EOFR) with clinical outcome and survival. METHODS: One hundred fifty glioblastoma or grade IV astrocytoma patients (82 men), with a mean age of 58.2 years (range 36-82 years), from 3 neurosurgical centers were included. In all cases, supramarginal resection was deemed feasible preoperatively; multicentric neoplasms or tumors with enhancing nodule involving eloquent areas were excluded. Analysis of EOFR was based on comparison between preoperative and postoperative 3-dimensional FLAIR images. EOFR was compared with extent of tumor resection (EOTR) based on gadolinium-enhanced T1 sequences; theses data were also statistically correlated with survival parameters as well as with clinical and biomolecular data. RESULTS: EOFR rate was 78.8% in the entire cohort, whereas EOTR based on T1 sequences was 98.3%. Mean progression free survival (PFS) and overall survival (OS) were 16.33 and 28.4 months, respectively. Adjusted Cox-regression models showed that a higher EOTR based on T1 sequences and EOFR were both associated with improved OS in individuals with either isocytrate dehydrogenase-1 wild-type or isocytrate dehydrogenase-1 mutated tumors. After adjustment, only the EOFR retained a statistically significant association with OS. Specifically, the risk of mortality decreased by 6.8% and 12.1% with each one-unit increase in EOFR, respectively. Further analysis based on artificial intelligence demonstrated that the cluster of patients with higher values of PFS and OS received greater rate of FLAIRectomy. CONCLUSION: This multicenter study demonstrates that EOFR is a more reliable predictor of PFS and OS than extent of resection based on gadolinium-enhanced T1 sequences, if supramarginal resection is performed according to specific preoperative planning. 3-dimensional FLAIR navigation-guided resection may represent the optimal strategy to achieve a real FLAIRectomy.