Epilepsy*

Date/Time: Tuesday, September 12, 2023 - 11:00 AM – 12:30 PM
Track: Special Interest Group (SIG) Session
Room: Franklin Hall 3 (4th Floor)
Log in to Add to My Schedule

Description:

Session Evaluation Form: https://myana.org/form/ana2023-session-evaluation-epile

Chair: Colin Ellis, MD

Co-Chair: Elizabeth Gerard, MD, FANA

The management of epilepsy in patients with gestational capacity requires specialized knowledge topics seizures during pregnancy; impacts of antiseizure medications on pregnancy outcomes and childhood development; management of catamenial seizures; and the role of genetics in the recurrence risk of passing on epilepsy to children. There have been several recent, high-impact publications from the Maternal Outcomes and Neurodevelopmental Effects of Antiepileptic Drugs (MONEAD) study, which is a prospective, observational, multicenter investigation of pregnancy outcomes for people with epilepsy and their children. Catamenial seizure exacerbation is common in women with epilepsy, and may not be optimally treated because of uncertainty regarding which treatment works best and when in the menstrual cycle treatment should be taken, as well as the possible impact on fertility, the menstrual cycle, bone health, and cardiovascular health. Finally, the role of genetics in epilepsy, and its impact on family planning, is an important topic that can now be informed by molecular genetic testing. However, many neurologists are unfamiliar or uncomfortable with ordering genetic tests, interpreting their results, and counseling patients about genetic risks. In this SIG session focused on epilepsy in patients with gestational capacity, leaders in the field will speak about recent advances in these areas and their impacts on patient care.

Learning Objectives:

Understand the latest literature on pregnancy outcomes in patients with epilepsy and gestational capacity and how this data is obtained.
Discuss neuroendocrine aspects of catamenial epilepsy and how they inform potential treatments.
Review the implications of neurogenetic testing for family planning and how neurogenetic testing differs from prenatal genetic screening.

Studying and Optimizing Pregnancy Outcomes in Epilepsy

Speaker: Page Pennell, MD, FANA

Approaches to study maternal and fetal outcomes in females with epilepsy are limited by ethical constraints against randomized trials. Despite this, we have gained data to practice evidence-based medicine in selection of anti-seizure medications and dosing strategies from pre-conception through pregnancy and postpartum, with the goal of maintaining maternal seizure stability while limiting adverse neonatal and neurodevelopmental effects. We have also gained data to counsel women about breastfeeding and comorbidities.

Catamenial Epilepsy: Biological Basis and Treatment Options

Speaker: Paula Emmanuela Voinescu, MD, PHD

My presentation will provide evidence for how sex hormones influences neuronal excitability and explains catamenial patterns of seizures. It will review the basic science and clinical studies available. It will also review treatment options used so far for catamenial epilepsy. It will emphasize the importance for collaboration between bench scientists and clinicians to better understand the effects of sex hormones and design better treatment options for catamenial epilepsy.

Neurogenetic Testing in Epilepsy and Implications for Family Planning

Speaker: Elizabeth Gerard, MD, FANA

Neurogenetic testing is increasingly offered to adults with epilepsy and can have important implications for family planning. This session will describe how neurogentic testing differs from prenatal genetic screening and underscore the complexities of counseling prospective parents in the face of genetic disorders with variable penetrance and expressivity. It will review the epidemiology of epilepsy inheritance and specific situations when a more individualized counseling is needed.

Homeostatic Sleep Need Increases Seizure Risk

Oral Abstract Presenter: Vishnu Cuddapah, MD, PhD

Rationale: Interictal high-frequency oscillations (HFOs) are considered one of the promising neurophysiological biomarkers of the epileptogenic zone. However, distinguishing pathological HFOs from physiological ones presents a significant challenge, yet it's crucial for their clinical application. We hypothesize that the distinctive morphological features of pathological HFOs can be discerned from physiological HFOs using an unsupervised learning approach, negating the need for pre-assigned training labels. Methods: We used chronic intracranial electroencephalogram (iEEG) data through subdural grids from 18 pediatric patients with medication-resistant neocortical epilepsy. After identifying 92,860 HFOs using an automated detector, each HFO event's EEG time-series data was transformed into time-frequency analysis imaging data. This data served as the input for the deep learning model, specifically a variational autoencoder (VAE). During training, the model was tasked with reconstructing the input time-frequency plot, ensuring the latent space followed a Gaussian distribution. This unsupervised approach didn't require labels indicating whether an event was pathological. Post-training, the HFO events' latent codes, stratified from all training patients, were clustered by the Gaussian Mixture Model (GMM) with K = 2. The cluster with a higher association with resection in post-surgical seizure-free patients was deemed pathological. The GMM model was then used to assign predictions, pathological or physiological, on all HFOs' latent codes from test set patients. Results: The effectiveness of our unsupervised method was gauged through a patient-wise 5-fold cross-validation. We projected randomly selected HFOs' latent codes into a two-dimensional (2D) space, comparing the pathological predictions from the VAE model with HFO-with-spike. Our analysis reveals that the pathological prediction from the VAE closely aligns with HFO-with-spike. Moreover, pathological HFOs, as predicted by our VAE model, established a pattern in the time-frequency plot. This pattern closely resembled the structure of an inverted T-shaped template, exhibiting characteristics akin to the ones we identified in our prior research. Using the resection ratio of pathological HFOs, as predicted by the VAE model, to forecast postoperative seizure outcomes resulted in an AUC of 0.91 (p < 0.001), signifying an improvement compared to the AUC of 0.82 (p < 0.001) obtained using the resection ratio of unclassified HFOs. Additionally, the VAE model outperformed the AUC of 0.89 (p < 0.001) achieved using the resection ratio of HFOs with spikes. Conclusions: We have demonstrated the ability to classify pathological HFOs using unsupervised machine learning with VAE, eliminating the need for any labeling. This approach could significantly enhance the clinical utility of pathological HFOs, particularly in delineating the epileptogenic zone during epilepsy surgery.

An Unsupervised Learning Approach for Discovering Pathological High-Frequency Oscillations

Oral Abstract Presenter: Hiroki Nariai MD, PhD, MS