Optimal Factors of Diffusion Tensor Imaging Predicting Corticospinal Tract Injury in Patients with Brain Tumors

OBJECTIVE: To identify the optimal factors in diffusion tensor imaging for predicting corticospinal tract (CST) injury caused by brain tumors. MATERIALS AND METHODS: This prospective study included 33 patients with motor weakness and 64 patients with normal motor function. The movement of the CST, minimum distance between the CST and the tumor, and relative fractional anisotropy (rFA) of the CST on diffusion tensor imaging, were compared between patients with motor weakness and normal function. Logistic regression analysis was used to obtain the optimal factor predicting motor weakness. RESULTS: In patients with motor weakness, the displacement (8.44 ± 6.64 mm) of the CST (p = 0.009), minimum distance (3.98 ± 7.49 mm) between the CST and tumor (p < 0.001), and rFA (0.83 ± 0.11) of the CST (p < 0.001) were significantly different from those of the normal group (4.64 ± 6.65 mm, 14.87 ± 12.04 mm, and 0.98 ± 0.05, respectively) (p = 0.009, p < 0.001, and p < 0.001). The frequencies of patients with the CST passing through the tumor (6%, p = 0.002), CST close to the tumor (23%, p < 0.001), CST close to a malignant tumor (high grade glioma, metastasis, or lymphoma) (19%, p < 0.001), and CST passing through infiltrating edema (19%, p < 0.001) in the motor weakness group, were significantly different from those of the patients with normal motor function (0, 8, 1, and 10%, respectively). Logistic regression analysis showed that decreased rFA and CST close to a malignant tumor were effective variables related to motor weakness. CONCLUSION: Decreased fractional anisotropy, combined with closeness of a malignant tumor to the CST, is the optimal factor in predicting CST injury caused by a brain tumor.