Abstracts

RATIONALE: Seizures occur in 30-80% of patients with supratentorial tumors. Total or subtotal resection often ameliorates seizures, but anti-epileptic drugs (AEDs) are commonly used after surgery, even in those who have not had seizures. Duration and efficacy of AED use in these patients is controversial, as is the role of EEG in this setting. Furthermore, cranial surgery alters EEG, complicating interpretation of interictal epileptiform activity. METHODS: We reviewed reports and selected tracings of EEGs recorded between 1997 and 1999 in 76 patients after resection of supratentorial tumors. Of these, 38 (age range 21-89; 18 male, 20 female; 13 high grade gliomas 12 low grade, 5 metastases, and 8 meningiomas) were selected on the basis of known post-surgical seizure history and consistent approach to EEG interpretation. Sharply contoured waveforms were graded as "spikes" or "sharp waves" by standard criteria (Neidermeyer, 1993), or as "sharp components/transients" if the waveform stood out from the background but did not meet these criteria. Degree of focal slowing (theta vs. delta) and presence of a breach rhythm were also analyzed, as were pathology and pre-operative seizure history. Univariate analysis using Chi-square was performed for each variable in relation to post-operative seizure status. RESULTS: Nineteen patients were seizure free and and 19 had seizures after surgery. Pre-operative seizures were significantly associated with post-operative seizures (p<.001). Patients with meningiomas were more likely to be seizure free after surgery (p<.025). Among EEG variables, the presence of any type of sharp transient was significantly associated with post-operative seizures (p<.01), but this association was lost if only sharp waves and spikes were analyzed (p>.50) Breach rhythm and focal delta were also not predictive of post-operative seizures. CONCLUSIONS: Visual analysis of routine EEG can yield useful information about seizures after tumor resection; standard criteria for epileptiform discharges may be too resrictive in this population. Prospective studies using of patients weaned off AEDs post-operatively would further clarify this.