Search

The urgent need to improve therapies for people with drug-resistant epilepsy (DRE) requires improved understanding of the aetiopathophysiology of epilepsy and the mechanisms responsible for pharmacoresistance. Autoantibody-mediated processes may explain DRE in a proportion of cases. Autoantibodies can be detected in people with isolated epilepsy (without encephalitis) but their prevalence rate, associated clinical features and overall clinical significance are unknown. Glutamic acid decarboxylase 65 (GAD65) autoantibodies are the commonest such example described in the literature and GAD65 autoantibody-positive epilepsy is thought to be particularly treatment-resistant. Laboratory techniques for detecting GAD65 autoantibodies are calibrated for the concentrations observed in patients with type 1 diabetes mellitus (T1DM) and it is unclear how well different assays perform at the higher levels of autoantibodies found in autoimmune neurological conditions. It is also unknown whether other immune factors, such as cytokines or lymphocyte subsets, are associated with DRE. Studies of these parameters offer novel potential therapeutic insights. The aims of this work were to: • estimate the prevalence of serum autoantibodies in epilepsy, and search for distinguishing clinical and paraclinical features of seropositive patients; • develop and optimise a novel cell-based assay (CBA) for the detection of GAD65 IgG autoantibodies in neurological conditions; • apply the novel CBA to cohort screening to identify DRE patients with GAD65 autoantibodies for clinical phenotyping; • characterise plasma cytokine profiles and immunocellular phenotypes in DRE. Clinical and experimental data from two large prospective studies of adults with epilepsy are presented. The first focussed on new-onset focal epilepsy (NOFE) and the second on DRE. The observed autoantibody prevalence rate in adult outpatients with epilepsy was 9% (42/479). Neuronal surface-directed autoantibody (NSA) status was best predicted by older age, ictal piloerection, epilepsy risk factors, mood problems, attention domain of the Addenbrooke's Cognitive Examination-Revised (ACE-R) and limbic system changes on MRI. A fixed and permeabilised CBA was developed to detect GAD65 IgG autoantibodies in serum, plasma, CSF and supernatants of cultured peripheral blood mononuclear cells (PBMCs). Serum from 170 healthy controls tested negative. High endpoint titre GAD65 autoantibodies were present in 7/206 (3.4%) patients with DRE but in none of 213 with NOFE. The CBA demonstrated inter-rater reliability and compared favourably to commercial GAD65 autoantibody radioimmunoprecipitation and enzyme-linked immunosorbent assays. To try and enhance the CBA, molecular cloning techniques for antigen tagging and cell surface expression for use in a live and non-permeabilised CBA were applied. However, antigen surface expression was unsuccessful. Our novel assay identified seven patients with DRE and high titre GAD65 autoantibodies. All were female and demonstrated continuous GAD65 autoantibody production throughout long-term follow-up. Five of seven had temporal lobe epilepsy, three had T1DM and two had had bilateral hippocampal sclerosis. Pain was reported as a seizure trigger in three patients, an unusual finding in epilepsy. Cytokine concentrations measured by a multiplex bead-based immunoassay combined with flow cytometry showed GAD65 autoantibody-positive DRE patient plasmas were enriched for IL1b, IL4, IL10, IL15 and TNFa, consistent with multifaceted immune system involvement. Mass cytometry of PBMCs suggested DRE GAD65 autoantibody-positive patients with high GAD65 autoantibody titres may represent a distinct subgroup. This work provides valuable insights into potential clinical and immunological aspects of epilepsy, with implications for clinical practice and future epilepsy research.