Seizures associated with antibodies against cell surface antigens are acute symptomatic and not indicative of epilepsy: insights from long-term data

Abstract

Background Clinicians have questioned whether any disorder involving seizures and neural antibodies should be called “(auto)immune epilepsy.” The concept of “acute symptomatic seizures” may be more applicable in cases with antibodies against neural cell surface antigens. We aimed at determining the probability of achieving seizure-freedom, the use of anti-seizure medication (ASM), and immunotherapy in patients with either constellation. As a potential pathophysiological correlate, we analyzed antibody titer courses. Methods Retrospective cohort study of 39 patients with seizures and neural antibodies, follow-up ≥ 3 years. Results Patients had surface antibodies against the N-methyl-d-aspartate receptor (NMDAR, n = 6), leucine-rich glioma inactivated protein 1 (LGI1, n = 11), contactin-associated protein-2 (CASPR2, n = 8), or antibodies against the intracellular antigens glutamic acid decarboxylase 65 kDa (GAD65, n = 13) or Ma2 (n = 1). Patients with surface antibodies reached first seizure-freedom (88% vs. 7%, P < 0.001) and terminal seizure-freedom (80% vs. 7%, P < 0.001) more frequently. The time to first and terminal seizure-freedom and the time to freedom from ASM were shorter in the surface antibody group (Kaplan–Meier curves: P < 0.0001 for first seizure-freedom; P < 0.0001 for terminal seizure-freedom; P = 0.0042 for terminal ASM-freedom). Maximum ASM defined daily doses were higher in the groups with intracellular antibodies. Seizure-freedom was achieved after additional immunotherapy, not always accompanied by increased ASM doses. Titers of surface antibodies but not intracellular antibodies decreased over time. Conclusion Seizures with surface antibodies should mostly be considered acute symptomatic and transient and not indicative of epilepsy. This has consequences for ASM prescription and social restrictions. Antibody titers correlate with clinical courses.

Inflammation, ictogenesis, and epileptogenesis: An exploration through human disease

“…Less than 5% to 33% of patients with anti-NMDAR antibody-mediated AE have ongoing seizures after >2 years of follow-up 19,29,41 and, although up to 30%-40% of anti-LGI1 AE patients had ongoing anti-seizure medication use, only 9%-22% had ongoing seizures after ≥2 years of follow-up. 2,19,21,29,42,43 This suggests, at least in the most part, a reversible mechanism of seizure generation. Anti-GABA B R AE also has a high incidence of acute symptomatic seizures, but unlike anti-NMDAR and anti-LGI1 antibody-mediated AEs, it may have a higher risk for epilepsy, with one study including 11 patients with anti-GABA B R AE reporting a 45% incidence of ongoing seizures at 2 years.…”

“…The pathogenicity of these antibodies vary: Those targeting neuronal cell surface antigens or synaptic proteins are thought to be directly pathogenic, whereas those directed against intracellular antigens are postulated to be an epiphenomenon of an underlying autoimmune process in which T cells potentially play a dominant role 13,15–18 . Treatment responsiveness as well as the occurrence of seizures during AEs and the development of post‐AE epilepsy varies between syndromes, with those featuring antibodies directed against cell surface antigens being more responsive to immunotherapy and less likely to develop subsequent epilepsy than those with antibodies directed against intracellular antigens 2,19 . The development of subsequent epilepsy may be due to ongoing inflammation, or result from irreversible changes to neuronal networks persisting after the inflammatory process has resolved and leading to spontaneous seizure occurrence 20 …”

“…Despite the high risk of acute symptomatic seizures with these two conditions, the risk of chronic epilepsy is low. Less than 5% to 33% of patients with anti‐NMDAR antibody–mediated AE have ongoing seizures after >2 years of follow‐up 19,29,41 and, although up to 30%‐40% of anti‐LGI1 AE patients had ongoing anti‐seizure medication use, only 9%‐22% had ongoing seizures after ≥2 years of follow‐up 2,19,21,29,42,43 . This suggests, at least in the most part, a reversible mechanism of seizure generation.…”

“…Once within the CNS, anti‐NMDAR antibodies are thought to act via binding to and crosslinking of NMDARs on the neuronal cell surface, ultimately leading to their internalization and a loss of function 33,69 (Figure 2). This process appears to be concentration dependent and reversible, 69 and the CSF antibody titers generally parallel the disease course 19,33,70 . In postmortem pathologic studies, immunoglobulin G (IgG) deposits are seen throughout the CNS, with a predominance in the hippocampus, basal forebrain, basal ganglia, and cervical spine, and are mostly of the IgG1 subclass 63 —a subclass that is induced in response to soluble and membrane protein antigens 71 .…”

“…• Elevated proinflammatory cytokines such as IL-6, tumor necrosis factor α (TNF-α), and IL-1β are found in new-onset refractory status epilepticus (NORSE) and febrile infection-related epilepsy syndrome (FIRES), and therapies targeting these cytokines have been efficacious in these conditions | 305 of post-AE epilepsy varies between syndromes, with those featuring antibodies directed against cell surface antigens being more responsive to immunotherapy and less likely to develop subsequent epilepsy than those with antibodies directed against intracellular antigens. 2,19 The development of subsequent epilepsy may be due to ongoing inflammation, or result from irreversible changes to neuronal networks persisting after the inflammatory process has resolved and leading to spontaneous seizure occurrence. 20 The risk of seizures is highest with anti-leucine rich glioma inactivated 1 (LGI1), γ-aminobutyric acid A receptor (GABA A R) and GABA B R antibody-mediated AE, with around 74%-90% of patients with anti-LGI1 AE 21,22 and >90% of those with anti-GABA A R AE [23][24][25] and anti-GAB-A B R AE [26][27][28] developing seizures.…”