Classification of Peritumoral Fiber Tract Alterations in Gliomas Using Metabolic and Structural Neuroimaging

The aims of this study were to investigate and categorize peritumoral fiber tract alterations while considering changes in metabolism and integrity of fiber structures using multimodal neuroimaging—that is, PET with O-(2-18F-fluoroethyl)-l-tyrosine and diffusion tensor imaging evaluated by fiber density mapping—and to correlate categories of fiber alterations with preoperative neurologic deficits and postoperative course. Methods: We examined 26 patients with cerebral gliomas. Fiber density data were used to segment peritumoral fiber structures and were coregistered to anatomic MR images and PET data. Fiber density and O-(2-18F-fluoroethyl)-l-tyrosine uptake values were evaluated as ipsilateral-to-contralateral ratios. Four metabolic categories were defined on the basis of O-(2-18F-fluoroethyl)-l-tyrosine values: tumor-infiltrated tissue, reactive tissue (astrogliosis and microglial activation), normal brain tissue, and tissue with attenuated amino acid metabolism. Fiber density values were grouped in 3 categories for structural integrity: compressed, normal, and attenuated fibers. Results: We evaluated and classified 103 peritumoral fiber structures with 10 patterns of fiber tract alterations. Fiber structures in tumor-infiltrated, reactive, and normal brain tissue showed compressed fibers, displaced fibers, and (partly) destroyed fibers, respectively. Attenuated amino acid metabolism was associated only with attenuated fiber density. Thirteen patients showed white matter–related neurologic deficits (paresis, hypoesthesia, aphasia, or anopia) as initial symptoms. Three patients showed tumor infiltration in the corresponding fiber tracts; all the others had reactive or normal brain tissue. Fiber structures were compressed or attenuated but not normal. The 3 patients with tumor infiltration in the corresponding fiber tracts and 1 with compressed fibers in normal brain showed no improvements or worsening of the deficits in the postoperative course. Eight patients with the corresponding fiber tracts in reactive or normal brain areas showed improvement of deficits. One patient underwent biopsy only. Conclusion: Our multimodal neuroimaging approach provides complementary information and more detailed understanding of peritumoral fiber tract alterations in gliomas which are more complex as described so far. We presented a classification model for systematic assessment of these alterations that may be helpful for treatment planning and prediction of patients’ prognoses.