HMGB1: A Novel Mechanistic Biomarker in Epilepsy

LE Walker1, K Tse1, GJ Sills1, DJ Antoine1, T Thippeswamy2, AG Marson1, M Pirmohamed1. 1The Wolfson Centre for Personalised Medicine, University of Liverpool, Liverpool, Merseyside, UK, 2College of Veterinary Medicine, IOWA State University, IOWA, USA

Brain inflammation has been described in both human epilepsy and experimental models of seizures. High Mobility Group Box-1 (HMGB1) is a “cellular alarmin”; alerting the immune system to sterile or infectious cellular insults. Through interaction with TLR4 it switches on the transcription of pro-inflammatory cytokine genes [1]. In rodent models of epilepsy, treatment of mice with Box A (antagonist fragment of HMGB1) or antagonists of TLR4 significantly reduce both the time spent in seizures and seizure severity [2]. The development of targeted immunomodulatory agents as antiepileptogenic strategies has tremendous potential, but this may come at a cost. Targeting a fundamentally protective pathway such as inflammation may have diverse and significant consequences. Careful patient selection is required to maximise the risk-benefit balance. HMGB1 has the potential to both stratify patients whilst providing insights into the mechanisms of drug-resistant epilepsy.

1) Epilepsy model: Adult male C57BL/6J mice (/~30g) were given repeated intraperitoneal injections of kainic acid until the onset of convulsive status epilepticus (SE), terminated by diazepam after 2 hours. Brain and blood was harvested 3, 6, 24, 72 hours, 7 and 14 days after SE. 2) Seizure model: Tonic seizures were induced in adult male CF1 mice (~25g) by supramaximal electrical stimulation delivered via corneal electrodes. Brain and blood was harvested 1, 4, 8, 16 and 24 hours after seizure. 3) Epilepsy Patients: Patients with chronic drug refractory epilepsy were admitted to the Walton Centre for Neurology & Neurosurgery for video EEG telemetry. Blood samples were taken 1, 4, 8 and 12 hours following seizure.

Peripheral blood HMGB1 was significantly elevated 3-4 hours following seizures in rodent models and in an analogous pattern in humans with drug resistant epilepsy. Baseline HMGB1 in patients with drug-resistant epilepsy was significantly higher compared to controls (median 7.69 vs 0.74; p<0.0002).

Future studies will determine whether HMGB1 levels at diagnosis can predict the development of drug resistance in a cohort of 200 newly diagnosed epilepsy patients.

[1] Bianchi & Manfredi. Science 2009; 323: 1683-1684.

[2] Maroso et al, Nat Med 2010; 16: 413-419.