Abstracts

Rationale: Tumor-related epilepsy is a cause of drug-resistant focal epilepsy in children, and some investigators recommend resection of the tissues including the tumor and seizure onset zone (SOZ) identified in intracranial EEG (iEEG) recording. However, capturing seizure events and localizing the SOZ can be a time-consuming procedure. Thus, we aimed to determine the findings on cortico-cortical evoked potential (CCEP) characterizing the SOZ. Based on previous literature, we hypothesized that the SOZ would be characterized by increased CCEP amplitudes.




Methods: We studied a consecutive series of nine children with tumor-related epilepsy (4 girls; age range: 3-16 years) who had CCEP measurement during extraoperative iEEG recording. For measuring CCEP, we delivered 40 single-pulse electrical stimuli (SPES) at 5 mA, 1 Hz to adjacent pairs of iEEG electrodes. Peak latency was defined as peak voltage of evoked potential within 11 to 50 ms for N1 and within 51 to 799 ms for N2. The peak voltage was z-score normalized using the 50-200 ms pre-stimulus baseline period. Mixed model analyses determined how the SOZ would be characterized by the CCEP amplitude and latency measures, including N1, N2, and N1/N2 ratio.




Results: The N1 peak voltage was lower when the SOZ was stimulated (linear mixed model coefficient: -0.494; p=0.012) and when CCEP was measured at SOZ sites (linear mixed model coefficient: -0.362; p=0.006). The N2 peak voltage was likewise lower when the SOZ was stimulated (linear mixed model coefficient: -0.361; p=0.043) but not when CCEP was measured at SOZ sites (linear mixed model coefficient: -0.210; p=0.079). The N1/N2 latency ratio was higher when the SOZ was stimulated (linear mixed model coefficient: +0.013; p=0.004) but not when CCEP was measured at SOZ sites (linear mixed model coefficient: -0.003; p=0.322).




Conclusions: Our study failed to replicate the generally accepted notion that the SOZ excessively responds to SPES at surrounding areas or that SPES at the SOZ induces excessive neural responses at surrounding areas. Instead, we found that neural responses to SPES were diminished in the tumor-related epilepsy network. This observation may suggest that tumor-related epilepsy has a mechanistic setup different from those resulting from other causes. A higher N1/N2 latency ratio induced by SOZ stimulation suggests that either N1 peak is delayed and/or N2 is shortened. A possible explanation for this observation is that delayed N1 latency reflects impairment of direct cortico-cortical propagation and shortened N2 latency reflects excessive suppression occurring in areas surrounding the SOZ.




Funding:
NIH NS064033 (E.A.), JSPS JP22J23281 (N.K.), JSPS JP22KJ0323 (N.K.).