Annual Meeting Abstracts: View

<< Back to Search Results

(Abst. 1.069), 2017

Expanded proteomic screen and identification of new novel biomarkers in seizure
Authors: John M. Gledhill, Evogen, Inc.; Elisa A. Waxman, Evogen, Inc.; Elizabeth J. Brand, Evogen, Inc.; Richard D. St. Clair, Evogen, Inc.; John R. Pollard, Christiana Care; Todd M. Wallach, Evogen, Inc.; and Peter B. Crino, University of Maryland, School of Medicine
Content: Rationale: To aid in elucidating the role of inflammation in epilepsy we undertook an extensive proteomic screen to identify novel inflammatory proteins that may be associated with seizures. Recent studies have established that brain inflammation is a prevalent etiological process in focal epilepsies and may contribute to seizure generation. Further, blocking this inflammation reduces or prevents seizures in preclinical models. We applied our findings to devise a predictive “seizure score” algorithm to aid in the rapid diagnosis of epilepsy. Methods: Blood samples (n=31) were obtained from epilepsy patients admitted to the epilepsy monitoring unit (EMU) for pre-surgical evaluation. All patients had EEG confirmed complex partial or generalized tonic-clonic seizures. All patients exhibited interictal epileptiform EEG discharges (assessed by 2 epileptologists). EMU patients gave a blood sample within 24 hours of a clinical event captured by video-EEG. Normal controls (n=29, no history of seizures) gave a single blood sample. Plasma was isolated from centrifuged whole blood. Concentrations of 51 biomarkers were assayed across 6 multiplex ELISA panels and quantitated using an electrochemiluminescent detection system (Meso Scale Discovery Platform). Results: Analysis of the proteomic survey of 51 markers identified 11 markers (IL-16, TARC, TRAIL, IL-7, MCP-4, P-Cadherin, Osteoactivin, M-CSF, ICAM-1, MMP-3 and MCP-2) with a significant difference in mean plasma concentrations between control and seizure patients within 24 hours after a captured seizure (p < 0.05). An optimal set of biomarkers was selected to create a “seizure score” algorithm through additional statistical analyses on the interactions, correlation, error analysis, and the influence of anti-epileptic drugs on the biomarkers. In the end, a set of 7 biomarkers were chosen to create a diagnostic model (IL-16, TARC, TNF-a, MIP-1B, TRAIL, MMP-3 and P-Cadherin) to yield a “seizure score” at 24 hours post-seizure. The "seizure score" algorithm resulted in Receiver Operating Characteristic (ROC) area-under-the-curves (AUC) of 0.954 (95% C.I. 0.899 – 1.01) within 24 hours of a confirmed seizure. The diagnostic performance metrics were: sensitivity 100%, specificity 90%, positive predictive value (PPV) 91%, negative predictive value 100% and overall "seizure score" accuracy of 95%. Conclusions: This discovery of novel combinations of biomarkers provides a diagnostic tool with significant clinical utility and insights into inflammatory proteins linked to epilepsy.  Data support differential expression of inflammatory markers between epilepsy patients and controls with high predictive markers falling generally into three categories: chemotactic signals, cell death signals, and extracellular matrix support. Multiple markers are also expressed in the choroid plexus and involved in brain injury, suggestive of blood brain barrier alterations and damage associated with seizures.  These markers show strong diagnostic performance for epilepsy. Funding: 1R43NS079029-01A1 and Evogen, Inc.