Abstracts

Rationale: The effectiveness and prognostic value of ECoG remains controversial in literature. Additional studies are needed to determine the utility of ECoG and better inform clinical practices for pediatric patients undergoing epilepsy surgery. We hypothesized that no epileptiform activity measured on ECoG after the resection of the EZ would be associated with long-term seizure freedom.

Methods: We conducted a retrospective cohort study of 116 pediatric patients who underwent ECoG-guided respective epilepsy surgery between 2005 and 2020 at the UPMC Children’s Hospital of Pittsburgh. Standard institutional pre-surgical evaluation including MRI, long-term video EEG, neuropsychological testing, interictal PET scan, ictal/interictal SPECT, and functional neuroimaging were completed. Patients were divided into three groups, Group A: post-resection ECoG demonstrated no epileptiform activity, no extension of resection beyond the margins of the pre-operatively defined area. Group B: post-resection ECoG demonstrated continued epileptiform activity, resection was extended until the final ECoG recording demonstrated no epileptiform activity. Group C: post-resection ECoG demonstrated continued epileptiform activity, however, the resection was not extended because the epileptiform activity was in eloquent cortex and/or was otherwise contraindicated.

Results: Our cohort included 115 patients (n=51 female). The average epilepsy diagnosis age was 6.6 years and the surgery age was 10.9 years. Breakdown of patient categorization is as follows: Group A: n=69 (60%), Group B: n=30 (26%), and Group C: n=16 (14%). A significantly greater number of patients received long-term preoperative intracranial monitoring in Group A (n=50 [73%]) compared to group C (n=7 [44%]) OR:3.4; [1.07-9.7]. Seizure freedom was achieved in 74 (64%) of all patients postoperatively. There was no significant difference seen in seizure freedom at 1-year follow up between the three surgical approaches (A: 65%, B: 70%, C:56% χ2 =1.88; p=0.65). Residual ECoG findings did not influence seizure freedom by comparing groups A and B (both without ECoG findings) vs. Group C (66/99 [67%], vs. 9/16 [56%]; OR: 1.56; [0.58-4.5]; p = 0.41). There was no significant difference in post-operative anti-seizure medication decrease in all groups. Multivariate logistic regression controlling for demographics (age at diagnosis, sex, MRI-located lesion, seizure semiology) and surgical strategy (i.e. ECoG findings and use in surgery) did not have a significant effect on 1-year seizure freedom (OR: 1.6 [0.29-9.6]; p=0.5835). Neoplastic pathology was highly predictive of seizure freedom at 1-year in our multivariate model (OR: 5.5 [1.9-19]; p = 0.003).

Conclusions: Our study showed that the absence of epileptiform activity after initial or extended resection was not a significantly associated with one-year seizure freedom. Furthermore, the modification of the surgical strategy based on ECoG was not significantly associated with one-year seizure freedom, nor did it have an impact on the complication rate.

Funding: None