Abstracts

Rationale: Germline genetic epilepsies present with complex phenotypes and are often antiseizure medication (ASM)-refractory. Non-pharmacological methods, such as vagus nerve stimulation (VNS) could bridge genetic epilepsy care until the development of precision therapies. Existing evidence on VNS use in genetic epilepsies is limited to case reports and series, of which a few ensure genetic variant credibility or factor in simultaneous ASM use. Here, we comprehensively describe a single-center genetic epilepsy cohort with VNS and analyze factors associated with attaining clinical response to stimulation.


Methods: We identified patients with presumed genetic epilepsy among 476 patients implanted with VNS in 1998-2023 by screening their electronic health records with a custom natural language algorithm. We reviewed the resulting records and included patients with pathogenic/likely pathogenic variants and VNS follow-up time of ≥ 6 months. We stratified the resulting cohort into channelopathies, other non-lesional, and lesional genetic epilepsies. We collected cross-sectional demographic, genetic, and epilepsy-related variables (age at epilepsy onset and VNS implantation, baseline seizure frequency, seizure types, presence of MRI lesions) and longitudinal data on ASMs and VNS settings. Strata were compared with Fisher exact test and Wilcoxon rank sum tests. A mixed effects binomial generalized linear model (GLMM) was used for the longitudinal analysis of factors affecting response (having a≥50% seizure decrease).


Results: We included 22 patients with genetic epilepsies with a total of 288 VNS interrogations. Of them 6 (27.3%) were channelopathies (SCN1A (N=3), SCN2A, SCN8A, CACNA1C), 4 - lesional epilepsy (TSC1/2, DCX, PTEN), and 10 - non-lesional epilepsy (other than channelopathies) genetic disorders (CDKL5, SLC2A1 (N=2), SHANK3, mosaic trisomy 21, MEIS2, tetrasomy 22q11, tetrasomy 5p, CHD2, GABRB3, KIAA2022, CSTB). The median follow-up time was 60.25 months (range 9.46-282.57). At the last follow-up, 12/22 (55%) patients were responders, among them all 6 channelopathies (50% of responders; p=0.027). All responders but CHD2 reached a response within 12 months. Notably, individuals with SCN1A variants had an initial increase in seizure frequency before achieving a sustained response. Eight patients (36%) reported non-seizure-related benefits, such as improved behavior. Five (22.7%) patients had transient side effects. At the last follow-up, there were no statistically significant differences in VNS settings or ASM burden between the three groups of genetic disorders. On longitudinal analysis with GLMM, having a channelopathy (OR 1.48, 95%CI 1.08-2), lower duty cycle (OR 0.99, 95%CI 0.98-1), and higher output current (OR 1.2, 95%1.09-1.3) were associated with increased odds of response.


Conclusions: The rate of response to VNS in our genetic epilepsy cohort was comparable to reports in the general population. VNS is a safe method for treating genetic epilepsy. Patients with channelopathies might respond to VNS better than patients with other genetic disorders, but further, larger-scale controlled studies are required to explore this observation.


Funding: The study received no funding.