About 2.5 million new cases of single seizures and epilepsy occur each year. The unpredictability of seizure occurrence and severity is a major contributor to patient morbidity. Current practice relies heavily on clinical information and tests such as MRI and EEG and is only able to offer limited insight on several important questions:
- Recurrence of seizures in a patient with a single seizure;
- Response to a specific medication or treatment (including when medications can be safely discontinued);
- Inadequate seizure control with medications; and
- Seizure occurrence on a given day or time frame.
Decisions are often based on probabilities in the range of 25-75 percent, with broad confidence intervals. Developing tools that could provide additional information regarding these questions would allow patients and physicians to make more accurate decisions, significantly reducing morbidity and anxiety associated with seizures.
Expression Profiles for Epilepsy
Although the potential of genomics, next-generation detection platforms and novel analytical modeling to enhance biological understanding and improve clinical care are being realized for several human diseases, there is little application of these technologies to study epilepsy in humans.
Specifically, there is a great need for identification and development of new molecular biomarker tests for the diagnosis, prognosis and treatment response both for patients presenting with one seizure and those with recurrent epilepsy.
micoRNAs (miRNAs) are important modulators of gene expression and have a role in the pathogenesis of human disease including cancer, inflammatory disease, cardiovascular diseases, and several neurologic conditions, such as Alzheimer’s, multiple sclerosis, and Parkinson’s. More recently, miRNAs are implicated in the development of epilepsy as researchers have observed dysregulated miRNA patterns in brain tissue and blood mostly studied in model systems of induced epilepsy. Few recent studies used human samples to show altered miRNA expression in serum of epilepsy patients, and several overexpressed miRNAs are associated with the inflammatory process. Unlike the labile nature of messenger RNA (mRNA), miRNAs are stable in tissue and the circulation, and along with circulating cytokines, provide opportunity to develop a more tractable and rapid diagnostic test for epilepsy at the point of care.
Pilot Study to Evaluate Epilepsy Expression Profiles
This pilot study will evaluate peripheral blood of epilepsy patients for expression profiles that distinguish epilepsy from healthy individuals, patients with a single seizure episode and patients with neurologic disorder other than epilepsy, such as migraines. We hypothesize that molecular signals are detectable in blood and are associated with epilepsy, and along with standard clinical assessments (EEG, MRI), a blood signature can ultimately aid in the diagnosis and management of epilepsy and prognosis of recurrence following first seizure event.
- Develop a study protocol, enroll patients with focal epilepsy, single seizure or migraines, and collect relevant samples and clinical data.
- Investigate miRNA and cytokine profiles associated with epilepsy compared to a healthy control group.
- Perform subgroup analysis to evaluate time-dependent expression changes of miRNAs and cytokines following first onset seizure.
The ultimate goal is a noninvasive diagnostic test based on molecular markers of seizure activity that provides additional and accurate information to the patient and clinician to enhance care decisions.
Using 50 eligible study subjects from the Duke Neurology Clinic and Emergency Department, we will collect blood samples during or near routine phlebotomy collections and abstract relevant demographic and clinical information from subject health records. We will capture and store clinical data in the Institutionally-supported and secured REDCap database. Researchers will analyze and quantify expression patterns using single-dimensional regression models for individual markers or multi-dimensional factor modeling for combined molecular and clinical datasets. These analytical models will be applied in the context of clinical outcome variables.
Tim Veldman, Ph.D.