Skull bone marrow injury caused by MR-guided focused ultrasound for cerebral functional procedures

Abstract

OBJECTIVE One patient for whom an MR-guided focused ultrasound (MRgFUS) pallidotomy was attempted was discovered to have multiple new skull lesions with the appearance of infarcts on the MRI scan 3 months after his attempted treatment. The authors conducted a retrospective review of the first 30 patients treated with MRgFUS to determine the incidence of skull lesions in patients undergoing these procedures and to consider possible causes. METHODS A retrospective review of the MRI scans of the first 30 patients, 1 attempted pallidotomy and 29 ventral intermediate nucleus thalamotomies, was conducted. The correlation of the mean skull density ratio (SDR) and the maximum energy applied in the production or attempted production of a brain lesion was examined. RESULTS Of 30 patients treated with MRgFUS for movement disorders, 7 were found to have new skull lesions that were not present prior to treatment and not visible on the posttreatment day 1 MRI scan. Discomfort was reported at the time of treatment by some patients with and without skull lesions. All patients with skull lesions were completely asymptomatic. There was no correlation between the mean SDR and the presence or absence of skull lesions, but the maximum energy applied with the Exablate system was significantly greater in patients with skull lesions than in those without. CONCLUSIONS It is known that local skull density, thickness, and SDR vary from location to location. Sufficient energy transfer resulting in local heating sufficient to produce a bone lesion may occur in regions of low SDR. A correlation of lesion location and local skull properties should be made in future studies.