Abstracts

Rationale: We previously found that the combination of plasma concentrations of soluble Intercellular Adhesion Molecule 5 (sICAM5) and Thymus and Activation Regulated Chemokine (TARC) can differentiate epilepsy patient plasma from normal controls. To extend biomarker discovery and to begin evaluation, we assayed plasma from consecutive patients admitted to the Epilepsy Monitoring Unit and compared the results to the gold standard of clinical diagnosis. Methods: Each patient gave one sample of blood per day while admitted, plus an additional sample within 12 hours after a seizure or seizure-like event. Blood was collected and processed in Na-citrate, and stored at −80°C. Patients were diagnosed as having epilepsy, psychogenic non-epileptic seizures (PNES) or an unclear diagnosis by routine clinical care. We subsequently utilized the Meso Scale Discovery (MSD) high throughput ELISA platform to test the levels of TARC, TNFalpha, GMCSF, IFNgamma, IL10, IL12 p70, IL1, IL2, IL6, IL8 and sICAM5 levels in human plasma. Results: 59 consecutive patients were enrolled, and 148 samples were collected . By using the median values for the patients and excluding the patients who had clinical non-diagnostic studies, we found that the combination of TARC and TNFalpha was able to differentiate between epileptic and PNES samples (AUC=0.89). The number of antiepileptic drugs the patients were taking correlated with TNFalpha levels but not TARC. Analyzing all the samples together we found that TARC is significantly higher in epilepsy samples than in PNES samples (p= 0.0401), and that sICAM5 was significantly lower in both epilepsy and PNES categories compared to those with unclear diagnoses and normal controls (p = 0.0166). The TARC /sICAM5 ratio was lower in epilepsy and PNES patients compared to those with an unclear diagnosis (p=0.0189) in plasma obtained 24 hours or less after seizure/event. The TARC /TNFalpha ratio was significantly higher in epilepsy samples than in PNES samples (p= 0.0014). We also found that the ICAM5/TNFalpha ratio is lower in epilepsy and PNES patients than in patients with an unclear diagnosis (p=0.0268). Finally, in all patients, ranging from 18 to 87 years old, we found that there was no correlation with biomarkers tested and age. Conclusions: We conclude that while our original biomarker ratio of TARC /sICAM5 can be used to distinguish patients with seizures from people without any kind of seizure, a subsequent TARC/TNFalpha ratio must be used to distinguish epilepsy from PNES. A caveat of this method is that a medication effect may be implied due to circulating TNFalpha levels correlating with antiepileptic drug use. Because diagnosis of epilepsy is a significant problem, a new blood-based biomarker would help to streamline the path from incident to diagnosis. To further evaluate this biomarker we plan to obtain plasma from the time of outpatient presentation and compare to their eventual diagnosis. This work was supported by GlaxoSmithKline and the Center for Neuroscience and Regenerative Medicine.