Anti-LGI1 encephalitis

Sudarshini Ramanathan MD (

Dr. Ramanathan of the University of Oxford and University of Sydney has no relevant financial relationships to disclose.

Sarosh R Irani MD (Dr. Irani of Oxiford University Hospitals received royalties from Nil as a patent co-applicant.)
Francesc Graus MD PhD, editor. (Dr. Graus of the University of Barcelona has no relevant financial relationships to disclose.)
Originally released May 8, 2018; expires May 8, 2021

This article includes discussion of anti-LGI1 encephalitis, autoimmune encephalitis, voltage-gated potassium channel complex antibodies, and limbic encephalitis. The foregoing terms may include synonyms, similar disorders, variations in usage, and abbreviations.


In 2010, antibodies to leucine-rich glioma inactivated 1 (LGI1) were first described as a distinct serological entity. In this review, we evaluate the historical and diagnostic evolution of this condition as well as the potential immunological and genetic contributors to disease pathogenesis. Firstly, we discuss the common epilepsy presentations in these patients, with particular attention to the faciobrachial dystonic seizures that are pathognomonic of this condition, as well as other associated seizure semiologies. Although cognitive impairment may be a presenting feature or, more commonly, follow the development of faciobrachial dystonic seizures, it also remains the most significant residual disability that can impair patients ability to return to functional independence. Therefore, secondly, we also explore the cognitive profiles of these patients. Thirdly, although there are characteristic peripheral, CSF, and MRI findings in a subgroup of patients with LGI1-antibody autoimmune encephalitis, in many patients these investigations can be normal and, hence, noncontributory to diagnosis. Finally, the majority of patients are responsive to immunotherapy. Early recognition and institution of immunotherapy and achieving seizure control is crucial to optimize long-term outcomes and should be a major future therapeutic goal.

Key points


• Testing directly for LGI1 and contactin-associated protein 2 (CASPR2) antibodies should be undertaken in patients with clinical features suggestive of faciobrachial dystonic seizures, limbic encephalitis, Morvan syndrome, and neuromyotonia.


• VGKC-antibody positivity, in the absence of antibodies to LGI1 and CASPR2, is unlikely to be of clinical significance and should be interpreted with caution.


• LGI1-antibody seropositivity is most commonly associated with limbic encephalitis, with a high incidence of seizures and cognitive impairment. There is significant overlap in clinical manifestations with CASPR2 seropositivity, with neuromyotonia, pain, and sleep disturbance occurring across both clinical entities, albeit at different frequencies.


• Faciobrachial dystonic seizures are exclusively associated with LGI1 antibodies and may predate the onset of cognitive impairment. These are brief dystonic events often involving the hemiface and ipsilateral upper limb and may occur hundreds of times a day. They are often refractory to therapy with antiepileptic drugs alone yet immunotherapy-responsive.


• Although typical findings on investigation include hyponatremia and medial temporal lobe hyperintensity or swelling on MRI, more than half the patients may have unremarkable findings on investigation, frequently with bland CSF and normal imaging.


• The majority of patients appear to be immunotherapy-responsive to first-line treatment, including one or more of the following: corticosteroids, intravenous immunoglobulin, and plasma exchange. This is particularly true in achieving control of faciobrachial dystonic seizures and other seizure semiologies and perhaps most clear with regards to the use of corticosteroids. Improvement in cognitive impairment is usually more gradual. Efficacy of second-line therapy remains to be systematically evaluated.


• Delay in initiation of immunotherapy, delay in controlling faciobrachial dystonic seizures, and the presence of relapses are associated with poorer outcomes, particularly cognitive dysfunction.

Historical note and terminology

Clinico-serological correlations. Autoimmune encephalitis is an increasingly recognized and understood clinical entity; many patients are found to have antibodies targeting domains of extracellular neuronal surface antigens, such as the N-methyl-D-aspartate (NMDA) receptor, the alpha-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) receptor, the gamma-aminobutyric acid (GABA) receptor, the glycine receptor, leucine-rich glioma inactivated 1 (LGI1) protein, and contactin-associated protein 2 (CASPR2) as well as numerous other less common targets (Binks et al 2018). These antigens are thought to play a critical role in central nervous system synaptic transmission; hence, disruption of these targets is likely to result in cognitive and psychiatric manifestations as well as seizures (Lai et al 2010). These neuronal surface antibody syndromes share some common paradigms: the target epitope is extracellular in location, and autoantibody binding is conformationally-sensitive. Hence, the native antigen is accessible to circulating autoantibodies. Also, the antibodies may alter the density and function of receptors; the clinical manifestations may be severe but are typically responsive to immunotherapy; and finally, the clinical phenotypes may be reminiscent of animal models of pharmacological or genetic dysfunction of the specific antigen in question (Lai et al 2010).

Antibodies targeting the voltage-gated potassium channel, a transmembrane ion channel, were previously identified in 3 primary neurologic syndromes: limbic encephalitis, neuromyotonia, and Morvan syndrome (Irani et al 2010). Limbic encephalitis is a CNS condition characterized by confusion and psychiatric changes, amnesia, seizures, and in a number of cases, hippocampal swelling on brain MRI. Neuromyotonia is a peripheral nervous system disorder characterized by peripheral nerve hyperexcitability that manifests as muscle cramps and stiffness. Morvan syndrome has features of both PNS hyperexcitability and CNS disturbance with encephalopathy as well as autonomic dysfunction and insomnia (Irani et al 2010). Neuromyotonia and Morvan syndrome have paraneoplastic associations, mostly with thymomas, particularly in Morvan syndrome (Irani et al 2010). These 3 disorders share a commonality in that they are often subacute in onset, patients share overlapping features, and they often respond to immunotherapy (Irani et al 2010; Machado et al 2012).

Revision of the underlying biochemistry. Antibody binding to the VGKC was previously detected using the pattern of immunostaining in rodent brain tissue and, most commonly, a radioimmunoprecipitation assay employing alpha-dendrotoxin-labelled VGKC channels from mammalian brain lysates (Irani et al 2010; Lai et al 2010; van Sonderen et al 2016b). Alpha-dendrotoxin binds with high affinity to 3 specific VGKC subunits: Kv1.1, 1.2, and 1.6 (Irani et al 2010), and these were initially thought to be the antigenic targets (Kleopa et al 2006). However, cell-based assays, using direct transfection of a native full-length human antigen into a live human cell line revealed that the majority of patients in fact had antibodies that targeted 2 main proteins complexed with the VGKC; specifically, LGI1 and CASPR2 (Irani et al 2010; Lai et al 2010). Hence, the term VGKC-complex antibodies was used to describe this collective group. Patients with LGI1 antibodies were more likely to have a limbic encephalitis presentation with prominent seizures, whereas those with CASPR2 antibodies more commonly showed neuromyotonia, dysautonomia, and neuropathic pain as well as typical limbic encephalitis. A direct comparison between LGI1- and CASPR2-antibody encephalitis is outlined in Table 1.

Table 1. Comparison of LGI1- and CASPR2-Antibody Autoimmune Encephalitis





64 kDa secreted protein expressed in hippocampus and neocortex

148 kDa cell adhesion molecule and a member of the neurexin intravenous superfamily, colocalizes with Kv1.1 and Kv1.2 at neural juxtaparanodes

Most commonly associated syndromes

CNS involvement: limbic encephalitis, faciobrachial dystonic seizures, other seizure types

CNS and PNS involvement: limbic encephalitis, Morvan syndrome, neuromyotonia


Adults, median age of onset 60 to 70 years of age (range, 20s to 90s)

2:1 male: female ratio

Extremely rare in children, with no support for testing in pediatric patients with encephalopathy and seizures

Adults, median age of onset 60 to 70 years of age (range 20s to 90s)

80% to 90% males

Extremely rare in children

Antibody subtype

IgG4 greater than IgG1

IgG4 greater than IgG1

Clinical features

Seizures, amnesia, confusion, hyponatremia (in 60% to 75%)

Cognitive impairment (in over 90%) (particularly verbal and visuospatial domains)

Seizures in up to 90%; faciobrachial dystonic seizures 50%; focal seizures 65% (dyscognitive/autonomic); generalized tonic-clonic seizures 60%

Sleep disturbance, insomnia in up to 50%

Less commonly, pain (15%), cerebellar ataxia, chorea, parkinsonism, dystonia, severe bradycardia, and other arrhythmias

Seizures in over half the patients, amnesia, hyponatremia more rare

Neuromyotonia, neuropathic pain (up to 40%), insomnia, dysautonomia, hyperhidrosis, and weight loss are more common

More chronic cognitive impairment (sometimes resembling dementia phenotype)

Sleep disturbance (agrypnia excitata most distinctively)

Less commonly, hyponatremia, cerebellar ataxia, chorea, orthostatic myoclonus, arrhythmias

CASPR2-antibody patients with thymomas often have AChR-antibody myasthenia gravis


Acutely, medial temporal lobe (hippocampal changes) on MRI, may have contralateral basal ganglia involvement with faciobrachial dystonic seizures

May be normal in around 50%

Long-term, hippocampal atrophy and whole brain atrophy more common; mesial temporal sclerosis present; reduction in pallidal volume with faciobrachial dystonic seizures

Often T2 hippocampal hyperintensities

Can be normal in around 70%


Often unremarkable in close to 75%

Minority have CSF pleocytosis or elevated protein

Rare to have intrathecal IgG synthesis

Normal in 60% to 70%

Paraneoplastic associations

Less common, 0% to 11%; thymoma, lung cancer, neuroendocrine pancreas tumor, abdominal mesothelioma

More common, 20% to 30%: mainly thymomas; also lung cancer, endometrial adenocarcinoma

Therapeutic response

Improvement with immunotherapy, particularly with regard to seizure cessation and control in response to corticosteroids

Improvement with immunotherapy in those without tumor associations, less of a response in those with tumor association

Management of underlying neoplasia essential in those with tumor

Outcomes and prognosis

Memory impairment and cognitive impairment present as long-term sequelae, particularly in the absence of early and appropriate immunotherapy; mortality rate reported between 6% to 19%

Over 70% have a favorable outcome with immunotherapy

10% fatality rate

Relapse rate

0% to 35%

0% to 32%

(Irani et al 2010; Irani et al 2013; Lai et al 2010; Malter et al 2014; van Sonderen et al 2016b; van Sonderen et al 2016c; van Sonderen et al 2017a; Bastiaansen et al 2017; Binks et al 2018; Thompson et al 2018; Varley et al 2018)

Although the pathogenic role of LGI1 or CASPR2 antibodies is increasingly clear, the role of VGKC-complex antibodies without antibodies to either LGI1 or CASPR2 is important to consider, not least because these may account for more than 60% of the VGKC antibody-positive results (van Sonderen et al 2016a). These “double-negative” patients do not form a uniform subgroup; rather, they are clinically heterogenous with presentations including epilepsy, neuropathic pain, sleep disorders, or Creutzfeldt-Jacob disease, and there appear to be very limited correlations between VGKC-complex antibody levels and clinical severity (van Sonderen et al 2010a; Malter et al 2014; Celicanin et al 2017; Lang et al 2017). Indeed, up to 5% of community controls may be positive for VGKC-complex antibodies (Vincent et al 2004). Furthermore, it was additionally found that antibodies to CASPR2, and less commonly LGI1, were also detected in patients with absent VGKC-complex antibodies (Irani et al 2013; van Sonderen et al 2016a). A study identified that double negative VGKC-complex antibodies very rarely bind the surface of live neurons in culture, and many are definitively proven to target cytosolic epitopes of Kv1 subunits or the non-mammalian alpha-dendrotoxin (Lang et al 2017). Therefore, the presence of VGKC-complex antibodies in the absence of antibodies to LGI1 or CASPR2 is of doubtful clinical or biological relevance; evidence for pathogenic potential is lacking; and these antibodies do not seem to correlate with a distinct immunotherapy-responsive syndrome.

For these reasons, a firm move should be made toward testing specifically for LGI1 and CASPR2 antibodies in the appropriate clinical setting in order to enhance the sensitivity and specificity of diagnosis and optimize clinical translatability.

The content you are trying to view is available only to logged in, current MedLink Neurology subscribers.

If you are a subscriber, please log in.

If you are a former subscriber or have registered before, please log in first and then click select a Service Plan or contact Subscriber Services. Site license users, click the Site License Acces link on the Homepage at an authorized computer.

If you have never registered before, click Learn More about MedLink Neurology  or view available Service Plans.