Stereotactic diffusion tensor imaging tractography for Gamma Knife radiosurgery.

OBJECTIVE The integration of modern neuroimaging into treatment planning has increased the therapeutic potential and safety of stereotactic radiosurgery. The authors report their method of integrating stereotactic diffusion tensor imaging (DTI) tractography into conventional treatment planning for Gamma Knife radiosurgery (GKRS). The aim of this study was to demonstrate the feasibility of this technique and to address some of the technical limitations of previously reported techniques. METHODS Twenty patients who underwent GKRS composed the study cohort. They consisted of 1 initial test case (a patient with a vestibular schwannoma), 5 patients with arteriovenous malformations, 9 patients with cerebral metastases, 1 patient with parasagittal meningioma, and 4 patients with vestibular schwannoma. DT images were obtained at the time of standard GKRS protocol MRI (T1 and T2 weighted) for treatment, with the patient's head secured by a Leksell stereotactic frame. All studies were performed using a 1.5-T magnet with a single-channel head coil. DTI was performed with diffusion gradients in 32 directions and coregistered with the volumetric T1-weighted study. DTI postprocessing by means of commercially available software allowed tensor computation and the creation of directionally encoded color-, apparent diffusion coefficient-, and fractional anisotropy-mapped sequences. In addition, the software allowed visualized critical tracts to be exported as a structural volume and integrated into GammaPlan as an "organ at risk" during shot planning. Combined images were transferred to GammaPlan and integrated into treatment planning. RESULTS Stereotactic DT images were successfully acquired in all patients, with generation of correct directionally encoded color images. Tract generation with the software was straightforward and reproducible, particularly for axial tracts such as the optic radiation and the arcuate fasciculus. Corticospinal tract visualization was hampered by some artifacts from the base of the stereotactic frame, but this was overcome by a combination of frame/MRI volume adjustment and DTI seeding parameters. Coregistration of the DTI series with the T1-weighted treatment volume at the time of imaging was essential for the generation of correct tensor data. All patients with the exception of the vestibular schwannoma cases had treatment pathology in the vicinity of eloquent tracts and/or the cortex. No new neurological deficits due to radiation were recorded at the short-term follow-up. CONCLUSIONS Recent reports in the medical literature have suggested that white matter tracts (particularly the optic radiation and arcuate fasciculus) are more vulnerable to radiation during stereotactic radiosurgery than previously thought. Integration of stereotactic tractography into GKRS represents a promising tool for preventing GKRS complications by reduction in radiation doses to functional organs at risk, including critical cortical areas and subcortical white matter tracts.