Abstracts

Rationale: In patients with a single seizure or those with established epilepsy, current practice relies mainly on clinical information and tests such as MRI and EEG to gain limited insight regarding questions such as the likelihood of seizure recurrence, response to specific medications, likelihood of intractability, etc. Decisions are often based on probabilities in the range of 25-75% with broad confidence intervals. Any tool that provides additional information regarding these questions could be of immense clinical utility, allowing patients and physicians to make more accurate decisions. In this pilot study, we investigated whether microRNA (miRNA) or cytokine expression profiles in peripheral blood could serve as potential biomarkers by evaluating differences between patients with cryptogenic focal epilepsy, healthy controls and patients with migraines (as a neurological control). Methods: Adult subjects with cryptogenic, focal epilepsy without other significant medical history were recruited from the Neurology Clinic at Duke University Hospital. Whole blood stabilized in RNA paxgene tubes and EDTA blood samples for DNA and plasma were collected. Demographic and clinical data including imaging, EEG, and medications were also collected. For the neurological control group, similar samples and data were collected from subjects with migraines without other significant medical history. Matched healthy control samples were obtained from previously banked samples. MicroRNA sequencing and targeted 30-plex MSD assays were performed on total RNA and plasma, respectively, and candidate miRNA and cytokine biomarkers were profiled in each group. Following normalization and batch correction programs, linear and regularized logistic regression models were used to identify differential expression patterns among the focal epilepsy and control groups. Results: Thirty subjects with cryptogenic focal epilepsy and 8 subjects with migraine were enrolled in the study. Thirteen healthy control samples were obtained from the biobank. The collected samples met internal quality control criteria; however, there was significant batch effect from the RNA data of previously banked healthy control samples compared to the RNA data from migraine and epilepsy samples specifically collected for this study.  Since statistical significance could not be achieved following batch correction, additional migraine subjects are being enrolled to improve the analysis with a larger sample size.  Still, our preliminary analysis suggests expression differences between healthy control and epilepsy groups including observed increases in mir484 and mir556 as well as in mir199a and mir128 consistent with previous reports. Further, analysis of plasma cytokines revealed expression changes in certain inflammatory cytokines in particular increased IL-7, IL-6 and IL-1a levels in epilepsy patients and increased IL-7, IL-8, IL-1a, and VEGF levels relative to healthy subjects. Additional analyses are ongoing to improve the signal-to-noise ratio on the collected data. Conclusions: The preliminary data suggest that genomic and inflammatory biomarkers have the potential to distinguish patients with cryptogenic focal epilepsy from other populations. Further work is ongoing and needed to identify specific profiles (both individual markers and combinations) that may be clinically useful. Funding: This work was supported by a pilot grant from Center for Applied Genomics and Personalized Medicine and by a donation from Ms. Barabara Rosenblatt.