Abstracts

Rationale: Supratentorial gliomas commonly present with seizures especially in low grade gliomas as compared to in high grade gliomas particularly in isocitrate dehydrogenase 1 mutant (IDH1mut) gliomas. Intraoperative electrocorticography (ioECoG), using interictal epileptiform spike and sharp wave discharges, and more recently using high frequency oscillations (HFOs), has been extensively studied in non-tumor related epilepsy. Little is known about the clinical utility of HFOs in the setting of tumor-related epilepsy (TRE). The relationship between HFOs and the presence of preoperative seizures, World Health Organization (WHO) grade, and IDH1 mutant status has not been examined. Methods: We retrospectively studied ioECoG recorded in a cohort of patients with glioma who underwent awake craniotomy and surgical resection. Clinical data including demographics, presence of preoperative seizures, tumor location, WHO grade, 1p/19q codeletion and IDH1 mutant status were assessed. The number and distribution of HFOs in the ioECoG were visually determined and quantified. HFOs were defined as discrete oscillations observed in the raw, unfiltered recording lasting less than 400msec, with a dominant oscillation frequency greater than 30Hz.  Automated HFO analysis was also performed using a MATLAB based program. The occurrence and rate of HFOs (high gamma [60-120 Hz] and ripples [120– 250 Hz]) in each recording was compared to the presence or absence of preoperative seizures, WHO grade, tumor location, and IDH1 mutant status. Results: Seventeen patients (mean age 46.7 years, 11 males) were included in the study. Twelve patients had preoperative clinical seizures (TRE) and five did not. Six patients had low grade and eleven had high-grade gliomas, with lesions located in the temporal lobe in six, frontal lobe in six, and parietal lobe in five patients. IDH mutation was found in seven (41%) and 1p/19q codeletion in four (23%) patients. Twelve patients underwent ioECoG using an 8X8 high density grid +/- depth electrodes and five were monitored using strips +/- depth electrodes (mean duration of recording= 10 minutes). Patients with preoperative seizures were more likely to have HFOs than those without preoperative seizures (9/12 vs. 0/5, p=0.009). The occurrence of HFOs, however, was not significantly different between high grade (5/11) and low grade gliomas (4/6). When HFOs were detected, the rate of HFOs was significantly higher in patients with IDHmut (mean= 18/min) than IDH-wild (mean=1/min) genotype (p=0.017). Intraoperative focal seizure occurred during the recording in 3 cases: cortical stimulation induced in two and spontaneous in one. Patients with TRE were maintained on antiepileptic drugs (AEDs) postoperatively, and all except two patients become seizure free (mean follow-up=4 months). Conclusions: HFOs are common in TRE and could be useful measures of epileptogenecity in gliomas and may have a potential role in tailoring glioma resection. Our findings further support the notion that IDH1mut genotype is more epileptogenic than IDH1 wild genotype gliomas. The latter observation could have implications on the selection and duration of postoperative AED therapy in TRE. Funding: None