Seizures fail to respond to medication in 30-40% of patients with epilepsy, and epilepsy surgery is an effective alternative. We need better methods to identify epileptogenic cortex that must be removed to render patients seizure free. Chronic intracranial EEG (iEEG) telemetry recording is often required, and this poses risk, discomfort, inconvenience, and cost. The proposed project aims to develop a new method to more safely plan epilepsy surgery by reducing or eliminating the need for chronic intracranial monitoring. High-frequency oscillations (HFOs) are a candidate biomarker of epileptogenic brain and consist of brief (25-200 msec) EEG events with a spectral content ranging between 80-600 Hz. HFO rates are elevated in the seizure onset zone (SOZ), but HFO rates may also be elevated in the epileptogenic zone (EZ) a hypothetical concept constituting the area that must be excised to attain postoperative seizure freedom. The SOZ does not always indicate the location of the EZ. The accuracy of HFO rates for determining the location of the SOZ is not established, and whether HFO rates in resected regions have different prognostic value than HFO rates outside resected regions, also requires elucidation. I hypothesize that inter-ictal HFO rates measured during acute intra- operative recording has sufficient accuracy for localizing the EZ to justify reducing or eliminating inpatient iEEG telemetry for patients with temporal lobe epilepsy (TLE), but not for patients with extra-temporal focal epilepsy. We will test this hypothesis in an observational clinical study of diagnostic accuracy in patients with medically refractory focal epilepsy.
Aim 1 of the study will determine and compare the accuracy of HFO rates to identify the SOZ and the EZ in chronic recordings during sleep and acute recordings at the time of depth electrode implantation in patients with focal epilepsy. This will be accomplished by 1) Assessing the concordance between HFO biomarker predictions and clinical outcomes to distinguish classifiers of SOZ from EZ using a receiver operating characteristic (ROC) approach, and 2) Assessing the accuracy of the rates of occurrence of the HFO biomarkers in specific neuroanatomical structures and locations.
Aim 2 will determine the predictive value of resected and residual HFO biomarker rates recorded with electrocorticography (ECoG) during surgery for post-resection seizure freedom. The clinical endpoint of this aim is that the predictive value of residual HFO rates for seizure recurrence will exceed the predictive value of resected HFO rates for seizure freedom. This research study is closely integrated with my career development plan that focuses on course work in biostatistics, clinical trial design, and neuroengineering. Four mentors who are recognized leaders in clinical and translational epilepsy research, with a specific expertise in the study of HFOs, and a dedicated statistician will provide the teaching and technical expertise to assure that the study is conducted properly and that I meet the aims of my career development plan.
Medically refractory focal epilepsy affects over 100,000 Americans and surgery is often the best treatment option for these patients. This project aims to determine the diagnostic accuracy of high-frequency oscillations, a neurophysiological biomarker, to determine the location of brain regions that generate seizures. If the accuracy of this biomarker is sufficient new safer, less expensive, and more effective surgical methodologies could be developed to treat epilepsy.
