- Top of page
Autoantibodies to surface proteins that influence neuronal excitability are increasingly found in different forms of epilepsy or encephalitis in adults, and are also beginning to be identified in children. The conditions are often refractory to traditional antiepileptic drugs. Detection of these antibodies can help to identify forms of epilepsy that may respond to immunotherapies.
Antibodies against both voltage-gated and ligand-gated ion channels have been recognized in neurologic diseases for some time. Classically, antibodies to acetylcholine receptors and voltage-gated calcium channels cause, respectively, the neuromuscular junction disorders myasthenia gravis and the Lambert Eaton myasthenic syndrome. In addition, patients with acquired neuromyotonia, which results from peripheral nerve hyperexcitability, often have immunoglobulin G (IgG) autoantibodies that can immunoprecipitate voltage-gated potassium channel (VGKC) complexes labeled with 125I-dendrotoxin (a snake toxin that binds specifically to different subunits of VGKCs) from digitonin extracts of mammalian brain tissue.
More recently, the role of antibodies in neurologic diseases has expanded to include involvement in diseases of the central nervous system (Table 1). VGKC-complex antibodies have also been associated with Morvan’s syndrome, a rare condition that includes neuromyotonia, confusion, autonomic dysfunction, sleep, and circadian rhythm disturbance (e.g., Liguori et al., 2001). The antibodies defined by the same assay are increasingly identified in patients with limbic encephalitis (LE), which presents with memory loss, seizures, hyponatremia, and often high signal in the medial temporal lobes on magnetic resonance imaging (MRI) (Vincent et al., 2004; Irani et al., 2010a). It is now clear that the VGKC complexes contain additional proteins—such as CASPR2, LGI1, and contactin-2—that are the actual targets for many of the patients’ antibodies (Irani et al., 2010a; Lai et al., 2010). CASPR2 antibodies are most often found in patients with Morvan’s syndrome or neuromyotonia. By contrast, LGI1 antibodies are particularly frequent in patients with limbic encephalitis. Of particularly relevance to seizures, mutations in the gene for LGI1 are found in autosomal dominant lateral temporal lobe epilepsy (Morante-Redolat et al., 2002), and IgG from one patient with LGI1 antibodies and limbic encephalitis increased the probability of release at the CA3 mossy fiber/pyramidal cell synapses in rodent brain slices (Lalic et al., 2011). The effects seen were similar to those found with low concentrations of dendrotoxin that binds specifically to VGKC Kv1.1, 1.2, and 1.6 subunits and inhibits potassium channel function. Because potassium channels modulate neurotransmitter release, the results suggest that the antibodies against LGI1 reduce the function of the VGKCs with which LGI1 is complexed, and leading to increased transmitter release.
Table 1. The new antibodies to CNS proteins and their clinical associations
|Clinical syndrome||Clinical features||Investigations||Antibody||Response to antiepileptic drugs||Response to immunotherapies|
|Morvan’s syndrome||Peripheral nerve hyperexcitability, autonomic dysfunction, insomnia and confusion, usually adult males||EMG evidence of neuromyotonic discharges||VGKC complex, predominantly CASPR2||Not required||Generally good|
|Limbic encephalitis||Amnesia, seizures, confusion or other psychological disturbance, usually older adults but some children||MRI high signal in medial temporal lobes and low plasma sodium in around 50%, CSF often normal||VGKC complex, predominantly LGI1||Partial response||Generally very good; also expedites recovery of the cognitive impairment|
|Faciobrachial dystonic seizures||Brief dystonic movements of face and ipsilateral arm (+/−leg), can affect either side. Progresses to limbic encephalitis in most cases||MRI and CSF most commonly normal||VGKC-complex, almost always LGI1||Often poor; adverse reactions common||Very good and prompt treatment may prevent subsequent progression to limbic encephalitis|
|NMDAR antibody encephalitis||Psychiatric disturbance, seizures, amnesia, confusion progressing to movement disorder, autonomic dysfunction and reduced consciousness||Often no MRI changes but CSF usually cellular; oligoclonal bands later||NMDA receptor, NR1 subunit||Partial||Response may be slow and intensive care prolonged.|
|GAD-antibody limbic encephalitis||Temporal lobe seizures with cognitive disturbance||MRI high signal in medial temporal lobes||GAD||Partial control only in most cases||Generally poor response but a few exceptions|
More evidence for the association of seizures with these antibodies is the increasing number of patients with the VGKC-complex/LGI1 antibodies and identification of a previously unrecognized syndrome of faciobrachial dystonic seizures (FBDS) (Irani et al., 2011). FBDS are characterized by acult-onset frequent dystonic attacks affecting one arm and ipsilateral hemiface. Impairment of consciousness is seen in ∼60%. Most of these patients progress to a typical LE, but a small number have been identified and treated successfully, with immunotherapies without such progression. Interestingly, FBDS in these patients respond well to immunotherapies and less so to antiepileptic drugs, which are often associated with adverse reactions (Irani et al., 2011). Conversely, there are patients who have limbic encephalitis with subsequent development of hippocampal sclerosis with temporal lobe epilepsy (TLE) (Bien et al., 2007). In addition, some patients with isolated TLE and VGKC-complex antibodies have been identified and may respond to immunotherapies (Adcock J, Buckley CJ, and Vincent A, unpublished observations).
Other central nervous system (CNS) diseases associated with specific antibodies to ligand-gated ion channels, N-methyl-d-aspartate (NMDA), α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid, and γ-aminobutyric acid (GABA)B receptors have now been identified and are also associated with seizures (Lai et al., 2009; Lancaster et al., 2010; Dalmau et al., 2011). The most common is the complex encephalopathy associated with NMDA receptor antibodies. The patients usually present with neuropsychiatric disturbance and seizures but progress to bizarre movement disorders including choreoathetoid movements or catatonia and orofacial dyskinesias, reduced consciousness, and autonomic disturbance. A few patients have forme frustes of the disease with predominant seizures or psychosis, sometimes with subtle cognitive involvement, without the typical movement disorders (Irani et al., 2010b; Dalmau et al., 2011). Particularly important is the number of young children with this condition. They can present from any age, even younger than one year old, usually with behavioral changes and, often, bizarre movements. Experiments have shown that the antibodies reduce the numbers of NMDA receptors on the surface of hippocampal neurons in culture. Injection of the antibodies intraventricularly into mice results in reduced hippocampal NMDA receptor expression, which was also seen in the few patients examined postmortem (for a review, see Dalmau et al., 2011).
The NMDA receptor–antibody encephalopathy was first reported in association with ovarian teratomas, mainly in teenage or young adult women, and very occasionally with other tumors, but it is now clear that there are many patients, particularly younger children and both sexes older than 40 years of age, who have a nonparaneoplastic form (Florance et al., 2009; Irani et al., 2010b). Indeed, this now appears to be the more common form of the syndrome. Removal of the teratoma, if identified, and immunotherapies results in substantial or complete improvement in most patients. In the nonparaneoplastic cases, immunotherapy is important and should be started early, but improvement is often slow and may require multiple drugs to suppress NMDA receptor-antibody levels (Irani et al., 2010b).
High titers of GAD antibodies are typically found in stiff person syndrome, but are also proving useful in the clinical assessment of patients with adult onset seizures, although the GAD 65 antibodies are unlikely to be directly pathogenic as they target an intracellular protein. Patients with these antibodies tend to be female and have temporal lobe epilepsy as a presenting feature, but cognitive features are also evident (Malter et al., 2010). Although there are few formal studies, GAD antibodies have also been identified in patients with epilepsy without clear cognitive involvement (e.g., Giometto et al. 1998), and may be a marker for immune-mediated disease.
All the above antibodies, except GAD antibodies, are directed against the extracellular domains of membrane proteins and are highly likely to be pathogenic. As a result it is not surprising that the patients do well with immunotherapies in combination with symptomatic treatments, although there have been no clinical trials as yet. Steroids, initially intravenous and then oral high dose, and plasma exchange followed by intravenous immunoglobulins are often used, although for the less severe cases, steroids or intravenous immunoglobulins may be sufficient. Second-line treatments with cyclosporin A or rituximab are used if the response is poor (Dalmau et al., 2011). Recovery is not necessarily fast but the majority of patients can resume relatively normal lives, although memory may remain impaired. Patients with GAD antibodies, however, have not yet shown good treatment responses (Malter et al., 2010) and these patients need to be studied further.
Although autoimmunity is unlikely to be a major cause of idiopathic epilepsy, studies of cohorts of patients from epilepsy clinics are beginning to find patients with the antibodies described here who may well benefit from immunotherapies if identified promptly (e.g., McKnight et al., 2005). The existence of these antibodies in both children (Florance et al., 2009; Haberlandt et al., 2011; Suleiman et al., 2011) and adults with a variety of different forms of encephalitis (Granerod et al. 2010), or epilepsy, suggests that further work on the role of antibodies in defining immunotherapy-responsive forms of these diseases is important.
Finally, these CNS conditions are raising many challenging issues. How do the antibodies access the CNS compartment, where do they act, and what are the mechanisms? Will this knowledge help in the design of better symptomatic treatments and which are the most suitable immunotherapies? Animal models are now being developed in attempt to answer some of these questions.