Abstracts

Rationale: Neural autoantibodies have been classically associated with syndromes such as limbic encephalitis but also, more recently, focal epilepsy. The association with neural autoantibodies may pose an opportunity for epilepsy etiology-targeted interventions such as immune-modulating therapies. We performed a systematic review and meta-analysis of the frequency of neural autoantibodies in epilepsy. Methods: The following databases were searched systematically for relevant studies from inception to May 10, 2019: Medline, EMBASE, CINAHL and Scopus. Search terms included a combination of controlled vocabulary and free-text words around the key concepts of “Epilepsy”, “Autoantibodies”, “Autoimmunity”, “Antibodies”, “Prevalence” and “Epidemiology”. Two independent reviewers (LG, CS) screened the title and abstract of each record, and full texts of potentially relevant papers were obtained. Studies were included if neural autoantibodies were systematically assessed in a cohort of epilepsy patients. Studies were excluded if the study population included subjects with a pre-existing diagnosis of encephalitis. Case reports and conference abstracts with insufficient data were excluded. Data collection included study characteristics and prevalence of overall and individual antibodies. In addition, for each study, a clinically relevant overall antibody prevalence was calculated by excluding VGKC without LGI1 or CAPSR2, low titer GAD65 and antibodies to unclassified neural specific antibodies without characterized antigens. The Newcastle-Ottawa quality assessment scale (NOS) was used to determine risk of bias in case-control studies. Results: The search returned 1268 abstracts and 51 full text articles were identified for review, with 25 included in the final analysis. 15 of these studies were case-control and 10 were descriptive studies, published between 2000-2018 with a total of 3266 cases tested. We identified a high risk of bias in most case-control studies (mean NOS scale 4.3). Antibody testing methods and panels varied between studies and 17 separate antibodies were considered. The majority of studies included exclusive serum testing, with rare cerebrospinal fluid sampling. Overall reported antibody frequency was 10.0% (95% CI 9-11.11%), with a heterogeneity index I2 of 80% (indicating considerable heterogeneity). After the exclusion of nonspecific antibodies, clinically relevant antibody frequency was 6.1% (95% CI 5.3-7% - I2 87%). Clinically relevant antibody frequency in specific populations studied included: focal epilepsy of unknown cause- 9.6% (7.7-12.0%), temporal lobe epilepsy- 5.5% (3.6-8.5%), new onset epilepsy- 5.0% (3.4-7.4%), and refractory epilepsy- 4.0% (2.4-6.8%). High titer GAD antibodies were present in 1.0% of idiopathic generalized epilepsy cases (95% CI 0.3-2.7%). Among controls (ranging from healthy to neurological disease controls), 1.0% (95% CI 0.6-1.7%) were antibody positive. Frequency of antibody positivity varied according to individual antibody as follows: Gly-R- 2.9%, NMDA-R- 2.2%, high titer GAD65- 2.2%, CASPR2- 1.5%, LGI1- 1.4%, GABA-A- 0.5%, Hu- 0.4%, Ma2- 0.4%, AMPA-R- 0.2%, GABA-B- 0.1%. Conclusions: Neural autoantibodies are present in a significant minority of epilepsy cases, with highest frequency in focal epilepsy of unknown cause. High heterogeneity and risk of bias in prior studies limits interpretation of the current literature. Future prospective studies with appropriate epilepsy controls and blinding are necessary to determine an accurate frequency of this likely small but therapeutically significant epilepsy subgroup. Funding: CS received a RTFC grant from the AES.