Abstracts

Rationale: 30 to 40% of all epilepsy patients have drug-resistant epilepsy (DRE). For them, surgery might be the best option. To determine the area to be removed, patients must undergo a costly pre-surgical evaluation. Interestingly, genomic evaluation is not a part of the pre-surgical assessment. The impact of genomic evaluation on surgical candidacy and outcome is still underestimated.

Methods: Patients older than 18 years old, with a diagnosis of DRE, that had received an evaluation for surgical candidacy and had a pathogenic variant on genetic testing were included. If surgery was done, at least 2 years of follow up was needed. Patients were divided into 3 groups: resective surgery, palliative surgery (VNS, DBS or callosotomy), or no surgery.

Results: 89 patients were included. Of these, 36 (40%) had a surgical procedure, 7 of them had resective surgery, including one with intracranial electrodes implanted before the resective surgery. Four patients presented a good outcome after the resective surgery (SCN1B in 2 patients, KCNA2 in 1 patient, and TSC2 in 1 patient) and 3 patients had poor outcomes after the resective surgery (SCN1A in 2 patients, TSC1 in 1 patient, and TSC2 in 1 patient). Twenty out of the 30 patients (66%) in the palliative group had good outcomes, with more than 50% reduction in the most disabling type of seizure, including 3 patients with genetic variants in SLC6A1, KAT6B and KMT2A that were almost seizure free. Out of 19 patients who received VNS, 10 (52%) patients had a good outcome. Five out of 7 (71%) DBS, and 7 out of 11 (64%) who had a 2/3 anterior callosotomy, had good outcomes. 53 patients had a surgical evaluation but were not deemed adequate for a resective surgery and declined palliative procedures.

Conclusions: We report the first case of a good resective surgical outcome on a patient with KCNA2 pathogenic variant. Our study combined with a previous report of 2 patients with SCN1B variant, suggest variants in this gene can be associated with good resective surgical outcomes. The finding of 2 patients with SCN1A pathogenic variants (and Dravet phenotype) and poor resective surgical outcomes is concordant with the literature. Our overall numbers for response to neuromodulation therapy were similar to the literature, with 71% responder rate (50% or more reduction in seizures) in the DBS group, 64% in the callosotomy group, and 52% in the VNS group. For Dravet syndrome, 6 out of 8 (75%) had a good outcome, 5 with VNS and 1 with DBS. One patient with an ATP1A3 variant had an excellent outcome, regarding his seizures and the episodes of alternating hemiplegia, a finding that has not been yet reported. This study represents a step towards a better knowledge of surgical outcomes in patients with genetically determined DRE. By understanding the implications of genomic abnormalities on surgical investigation and outcomes, genomic findings can be incorporated as a prediction tool on evaluating surgical candidacy, aiding to efficiently selecting eligible surgical candidates. Moreover, it can avoid the use of already scarce resources on patients that would have a poor response to surgery.

Funding: Please list any funding that was received in support of this abstract.: No funding received.