Autoimmune encephalitis in the age of neuronal surface antigens.

Since anti-NMDA receptor (NMDAR) encephalitis was first described almost a decade ago, the family of disorders associated with antibodies against neuronal surface antigens is one of the most rapidly expanding categories of neurologic disease. With this growth comes a paradigm shift in neurology—autoimmune encephalitis has evolved from an often vaguely defined clinical syndrome to a group of precisely delineated disorders associated with specific autoantibodies. Forms of autoimmune encephalitis, even those associated with long-established neuronal autoantibodies, are now identified as having distinct subtypes likely caused by distinct pathogenic antibodies directed to neuronal surface antigens. For example, some patients with GAD65 antibodies harbor likely disease-causing antibodies against glycine receptors or GABA receptors, and the encephalitides associated with voltagegated potassium channel (VGKC) antibodies are now subdivided into multiple unique disorders with antibodies against potassium channel associated proteins, such as leucine-rich glioma-inactivated 1 (LGI1) and Caspr2. A more precise understanding of these previously nebulous conditions may lead to important changes in clinical diagnosis and treatment. In this month’s issue of Neurology®, van Sonderen et al. report a nationwide study of the incidence and clinical features of anti-LGI1 encephalitis, providing a comprehensive view of this recently defined form of autoimmune encephalitis. The authors report a current annual incidence of anti-LGI1 encephalitis in the Netherlands of 0.83 cases per million. Although still rare, this incidence marks a steep rise in recognition of the disorder, with a nearly 4-fold increase over the known annual incidence in the preceding 3 years. As the authors point out, this incidence places anti-LGI1 encephalitis in the same ranks as Creutzfeldt-Jakob disease or Lambert-Eaton myasthenic syndrome, both well-known and clinically important disorders. Furthermore, anti-LGI1 encephalitis is almost certainly still underdiagnosed. It is likely that many potential autoimmune encephalitis patients, particularly those with atypical or mild presentations, are never even tested for the disorder. Additionally, given the low sensitivity of CSF testing alone (only a 53% detection rate in this study), diagnosis may be missed if a serum sample is not also evaluated. Therefore, as awareness of this disorder increases, the known incidence will likely rise. As neurologists increasingly make a specific diagnosis of anti-LGI1 encephalitis, how might clinical practice change? First, by defining the clinical syndrome more precisely, we can begin to recognize symptom sets in need of specific management, and start to delineate specific approaches to this management. For example, establishment of antiNMDAR encephalitis as a distinct clinical syndrome has led to work describing its associated particular psychiatric features, with the goal of rationally targeted psychopharmacologic therapies. Similarly, definition of the movement disorder in patients with anti-NMDAR encephalitis has led to successful treatment of some patients with tetrabenazine. Likewise, in anti-LGI1 encephalitis, approximately half of patients have an unusual seizure type, termed faciobracial dystonic seizures. These seizures may lack ictal EEG findings, leading to underrecognition and undertreatment. Knowledge that a patient has anti-LGI1 encephalitis may facilitate prompt appreciation and management of such seizures. Second, an emerging story is that specific forms of autoimmune encephalitis may have distinct responses to immunotherapy. Here, the authors report an 80% initial efficacy for first-line immunotherapy (corticosteroids alone, IV immunoglobulin alone, or corticosteroids plus IV immunoglobulin or plasma exchange). A similar marked responsiveness to first-line therapy, particularly corticosteroids, has been reported for anti-LGI1 encephalitis. Although it is difficult to compare across modalities, this initial response rate appears more robust than that reported for anti-NMDAR encephalitis (approximately 50% by 1 month after first-line therapies). In contrast, there may be persistent neurocognitive deficits even in those anti-LGI1 patients who have made substantial recoveries. In the current study, neuropsychological assessment, obtained for a subset of patients, showed disturbed spatial recognition memory. Malter et al. recently evaluated neurocognitive outcomes