Longitudinal spatial changes of the contralesional corticospinal tract in high-grade glioma: a within-patient DTI tractography study

Abstract Purpose The contralesional corticospinal tract (CST) is often used as an internal reference in glioma tractography, but its longitudinal stability remains poorly understood. We explored within-patient changes in contralesional CST characteristics between primary and recurrent high-grade glioma. Methods Twenty-nine patients with paired preoperative diffusion tensor imaging at primary and recurrent assessments were retrospectively analyzed. The contralesional CST was reconstructed using anatomically guided deterministic tractography at fractional anisotropy thresholds (FATs) of 0.10 and 0.15. Tract-averaged fractional anisotropy (FA), tract length, and tract volume were compared longitudinally. Spatial changes were assessed using patient-level slice-wise centroid trajectories and paired voxel-wise analyses in standardized space, with HCP842 used as a fixed normative spatial reference. Results No longitudinal differences in FA, tract length, or tract volume remained significant after BH-FDR correction. In contrast, slice-wise analyses showed a predominantly anterior CST displacement at recurrence, with mean paired Y-axis differences of 2.576 mm at FAT 0.10 and 2.078 mm at FAT 0.15; 57 and 21 slices, respectively, remained significant after correction. Mediolateral changes were smaller and less consistent, and no voxel-wise differences survived correction. Relative to HCP842, both assessments were posterior to the reference trajectory, while the recurrent CST was less posterior. Conclusion The contralesional CST showed longitudinal spatial displacement without corresponding corrected changes in tractography-derived FA, tract length, or tract volume. These findings indicate that its spatial position may not remain fixed between primary and recurrent glioma assessments.