Abstract
Background: The prevalence of neural autoantibodies in epilepsy of unknown etiology varies among studies. We aimed to conduct a systematic review and meta-analysis to determine the pooled global prevalence and the prevalence for each antibody. Methods: A systematic search was conducted for studies that included prospectively patients ≥16 years old with epilepsy of unknown etiology and systematically determined neural autoantibodies. A meta-analysis was undertaken to estimate pooled prevalence in total patients with a positive result for at least one neural autoantibody in serum and/or cerebrospinal fluid (CSF) and for each autoantibody. Results: Ten of the eleven studies that met the inclusion criteria and a total of 1302 patients with epilepsy of unknown etiology were included in themeta-analysis. The global pooled prevalence (IC95%) was 7.6% (4.6–11.2) in a total of 82 patients with a positive result for any neural autoantibody. None of the controls available in the studies had a positive result. Individual pooled prevalence for each autoantibody was: glycine receptor (GlyR) (3.2%), glutamic acid decarboxylase (GAD) (1.9%), N-methyl-d-aspartate receptor (NMDAR) (1.8%), leucine-rich glioma inactivated-1 protein (LGI1) (1.1%), contactin-2-associated protein (CASPR2) (0.6%) and onconeuronal (0.2%). Conclusions: The pooled prevalence of neural autoantibodies in patients with epilepsy of unknown etiology is small but not irrelevant. None of the controls had a positive result. There was high heterogeneity among studies. In the future, a homogeneous protocol for testing neural autoantibodies is recommended.