Diffusion Anisotropy Predicts Directional Progression After Complete Resection of Frontal Glioblastoma
Jawad Fares, Yonghao Li, Yizhou Wan, Roxanne Mayrand, Stephen J. PriceBACKGROUND AND OBJECTIVES:
Despite complete resection of contrast-enhancing tumor, glioblastoma frequently progresses due to microscopic invasion beyond surgical margins. In frontal glioblastoma, functional tolerance of anterior and lateral regions enables consideration of supramaximal resection when invasion direction can be anticipated. Imaging biomarkers that predict postresection progression direction may thus inform surgical planning.
METHODS:
We performed a retrospective analysis of a prospectively recruited multicenter cohort of patients with newly diagnosed isocitrate dehydrogenase–wildtype frontal glioblastoma who underwent complete resection of contrast-enhancing tumor. Eligible patients had preoperative diffusion tensor imaging (DTI) and longitudinal MRI follow-up. Imaging biomarkers included fluid-attenuated inversion recovery abnormality, isotropic diffusion tensor component, and diffusion anisotropy (DTI-q). Directional progression was classified on follow-up imaging. Predictive performance was assessed using accuracy and McNemar tests, and associations with progression-free survival and overall survival were evaluated.
RESULTS:
Of 229 screened patients, 44 had frontal glioblastoma, and 29 underwent complete resection of contrast enhancement (n = 29). Medial (52%) and posterior (48%) progression were most common. DTI-q demonstrated the highest predictive accuracy across directions (median 0.83), outperforming fluid-attenuated inversion recovery and isotropic diffusion tensor component (median 0.69), with significantly higher paired accuracy for lateral, superior, and inferior progression (
CONCLUSION:
Preoperative DTI-q more accurately predicts directional progression of frontal glioblastoma than conventional MRI after complete resection and identifies surgically relevant invasion patterns. Visualization of spatial invasion patterns along white-matter tracts provides a practical framework for directionally tailored resection planning within surgical navigation workflows.