Gammaknife Thamamotomy For Intractable Tremors: Clinical Outcome And Correlations With Neuroimaging Features

OBJECTIVE: To analyse MRI postoperative imaging features of radiosurgical thalamic lesion and establish correlations with clinical outcome. BACKGROUND: Clinical response to Gammaknife radiosurgical thalamotomy for severe tremor can be variable and unpredictable. DESIGN/METHODS: VIM Gammaknife unilateral thalamotomy was performed in 75 patients with severe tremors (Essential or Parkinsonian tremor). Follow-up included tremor assessment, neuropsychological assessment, impairment in ADL and serial MRI scans at baseline and at 3,6,12 months. Each MRI scan was analysed by a single investigator,who was blinded for clinical outcome. The presence or absence and the size of the post-gadolinium enhancement were analysed. The size of edema around the target was assessed according to a semi-quantitative scale (grade 1 to 5). According to the features of the lesion and irrespective of the clinical results, patients were classified into three subgroups of normo, hyper or hyporesponders. RESULTS: The percentage of patients presenting a good outcome (decrease in tremor score > 50%) was 74%. A workable MRI follow-up allowed the response profile to be classified into 3groups: Hyperresponders (4%): 3 patients had extensive MRI lesions with massive edema extending beyond internal capsule. Two patients presented with transient hemiparesia. Hyporesponders: 25% displayed minimal changes (no post-gadolinium enhancement, very limited T2 or flair lesion. Most patients (87.5%) without clinical benefit had this pattern of weak MRI response. There was no discrepancy between stereotactic coordinates of the target and the lesion center as depected on imaging studies. Normoresponders: 71%of the patients displayed significant MRI changes with stereotyped pattern (cocade like lesion) and had a good clinical response. CONCLUSIONS: Analysis of serial postoperative MRI scans demontrated a good correlation between the clinical outcome and neuroimaging features. Among the patients with poor clnical benefit, hyporesponder type pattern was frequent with almost no radiosurgical lesion. Factors encounting for this absence of radiosurgical response need to be clarified and further investigate. Disclosure: Dr. Witjas has recieved personal compensation for activities with GlaxoSmithKline, Medtronic, Boehringer, UCB Pharma, as a consultant and/or speaker. Dr. Carron has nothing to disclose. Dr. Eusebio has nothing to disclose. Dr. Azulay has received personal compensation for activities with GSK, Bohringer, UCB, Novartis, Teva and Lundbeck as a consultant. Dr. Regis has nothing to disclose.