Antibody-mediated central nervous system diseases: disease recognition and treatment challenges

Antibody-associated disorders of the central nervous system (CNS) can be divided broadly into three groups based on the location of the target antigens 1. In the classic paraneoplastic (cancer-associated) disorders (e.g. anti-Hu, Ri, Yo) the antigens are intracellular (nuclear or cytoplasmic) neuronal proteins. Studies of patients' brain tissue obtained at biopsy or autopsy, systemic tumour, cerebrospinal fluid (CSF) and blood suggest that these disorders are mediated by cytotoxic T cell mechanisms and that the antibodies are markers of paraneoplasia 2. Treatment of the tumour and immunotherapy may stabilize the neurological syndrome, but significant improvements are rare.

In the second group of disorders, antibodies target intracellular synaptic proteins [e.g. 65 kDa glutamic acid decarboxylase (GAD65) and amphiphysin] that might be vulnerable to antibody-mediated disruption during synaptic vesicle fusion and reuptake. However, it is unclear if antibodies or T cell mechanisms mediate the neuronal dysfunction.

The third and largest group, and the focus of this section, is the autoimmune encephalitis syndromes associated with antibodies to synaptic or neuronal cell-surface antigens, such as the N-methyl-_D-aspartate (NMDA), alpha-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) or gamma-aminobutyric acid (GABA) receptors, among others (Table 1) 1. In contrast to the previously mentioned groups, which affect mainly older patients, this group of disorders frequently affect young individuals, and may occur with or without a cancer association. Prior to the elucidation of the underlying immune pathogenesis, many of these disorders were known by descriptive terms such as dyskinetic encephalitis lethargica, post-partum psychosis and juvenile acute non-herpetic encephalitis. The presentation is commonly, but not exclusively, with psychosis, catatonia, memory deficits, cognitive decline, movement disorders and/or seizures 3. Patients often develop intrathecal synthesis of antibodies, and the antibody levels in CSF correlate with symptoms and outcome 4. Evaluation of human brain demonstrates deposits of antibodies without complement, reduced levels of the target antigens and the presence of B and/or plasma cells but rare T cell infiltrates. An antibody-mediated pathogenesis is supported by studies demonstrating that patients' antibodies have functional and/or morphological effects on the target antigens. For example, patients' NMDA and AMPA receptor antibodies cause a titre-dependent decrease of synaptic and extrasynaptic NMDA receptors through a mechanism of capping, cross-linking and internalization 5. Patients' GABA_A receptor antibodies disrupt receptor signalling by reducing receptor density in synapses through relocation of receptors from synaptic to extrasynaptic sites 6. These effects are reversible in all cases, and are likely to explain that although patients may be severely impaired or comatose for weeks or months, most are responsive to immunotherapy aimed at antibody depletion and tumour removal (when present). Increasing awareness of these disorders has led to the identification of patients with less severe or partial syndromes, including patients with pure or predominant psychosis, predominant refractory seizures or abnormal movements. This suggests that pathogenic antibodies are likely to play a role in a wider group of neurological disorders. This is supported by the recent description of antibodies to IgLON5, a neuronal cell adhesion molecule, in patients with a non-rapid eye movement (NREM) and rapid eye movement (REM) sleep behaviour disorder with pathological findings of a novel tauopathy 7.

Table 1.

Autoimmune encephalitis associated with antibodies to the neuronal cell surface or synaptic antigens 1

Antigen	Syndrome and main features
NMDA	Anti-NMDAR encephalitis (complex neuropsychiatric syndrome)
AMPA	Limbic encephalitis, psychosis
GABAB	Limbic encephalitis with prominent seizures, status
GABAA	Status epilepticus, refractory seizures
LGI1	Limbic encephalitis, myoclonus, hyponatraemia
Caspr2	Encephalitis and/or neuromyotonia
GlyR	Stiff-person syndrome, hyperekplexia, PERM
mGluR1	Cerebellitis (± Hodgkin's disease)
mGluR5	Limbic encephalitis + Hodgkin's disease (Ophelia syndrome)
Dopamine-2	Basal ganglia encephalitis, Sydenham's chorea
DPPX*	Hallucinations, agitation, myoclonus, tremor, seizures, diarrhoea
IgLON5	NREM/REM parasomnia and sleep breathing disorder

^*A subunit of the Kv4·2 potassium channel; AMPA = alpha-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid receptor; Caspr2 = contactin-associated protein-like 2; DPPX = dipeptidyl-peptidase-like protein-6; GABA_A = gamma-aminobutyric acid-A receptor; GABA_B = gamma-aminobutyric acid-B receptor; GlyR = glycine receptor; LGI1 = leucine-rich glioma inactivated protein-1; mGluR1 = metabotrophic glutamate receptor 1; mGluR5 = metabotrophic glutamate receptor 5; NMDA = N-methyl-_D-aspartate receptor; NREM/REM = non-rapid eye-movement/rapid eye movement; PERM = progressive encephalomyelitis with rigidity and myoclonus.

The mechanisms that initiate and maintain the autoimmune responses in paraneoplastic neurological disorders (PND) and the autoimmune encephalitis are unclear. In cancer-associated disorders, the immune response is likely to be initiated against neuronal antigens expressed by the tumour. In autoimmune encephalitis, the occurrence of a viral-like prodrome in many patients suggests that an infectious process may play a role in activating the immunological system. The description of patients developing anti-NMDA receptor encephalitis and autoimmune responses to other neuronal cell surface antigens after herpes simplex viral encephalitis supports this concept 8. There are some patients in whom autoimmune encephalitis overlaps with demyelinating disorders, but it remains to be established whether there is a relationship between the two syndromes.

The diagnosis of classic PND and autoimmune encephalitis is based on the recognition of the neurological syndrome, the detection of the specific antibodies in serum and/or CSF and the identification of the underlying cancer (if paraneoplastic). Generally, physicians should consider autoimmune encephalitis with neuronal cell surface or synaptic antibodies in any patient, especially if young, with a rapidly progressive encephalopathy of unclear aetiology. Many patients are initially suspected of having a ‘viral’ aetiology, although viral studies are negative. For some disorders, such as NMDA receptor encephalitis, patients may initially be given a primary psychiatric diagnosis, and the associated abnormal movements or fever ascribed erroneously to the use of anti-psychotic medication. Patients with autoimmune encephalitis can experience relapses, and this diagnosis should be considered in patients with a past history of encephalitis or a relapsing encephalopathy.

Ancillary studies may show mild to moderate pleocytosis in the CSF, but these studies can be normal. In some cases, oligoclonal bands are the only CSF abnormality found. Neuroimaging can be useful to rule out other aetiologies, but is often normal. The disorders associated with antibodies to LGI1, and GABA_B and AMPA receptors more commonly show magnetic resonance imaging (MRI) fluid attenuated inversion recovery (FLAIR)/T2 changes in limbic structures that suggest the diagnosis.

One study has shown that up to 13% of serum samples can be negative, and therefore CSF should be evaluated during initial tests for antibodies 4. If antibodies are found only in serum but not in CSF, the possibility of a false positive result should be considered, and the CSF should be retested. The relevance of following antibody titres is questionable, as titres can remain elevated even after patients have improved; however, a rise in titres can help to ascertain the cause of recurrent symptoms.

The general treatment approach, based on studies with anti-NMDA receptor encephalitis, includes first-line immunotherapy with intravenous immunoglobulins (IVIg) and corticosteroids and tumour treatment when appropriate. It was observed that approximately 50% of patients treated using this approach showed a response within 4 weeks 3. For non-responders, second-line therapy with rituximab and cyclophosphamide is often effective. As 50% of patients do not respond to first-line therapy, there is increasing support for the upfront use of second-line therapies. Furthermore, rituximab and cyclophosphamide target the antibody-producing cells and there are data suggesting that patients treated with these therapies may experience fewer relapses than patients not treated with these agents. There are few data regarding the need or benefit of long-term maintenance of immunosuppression at this time.

Disclosures

M. R. R. has a patent for the use of NMDAR antibodies as a serological test.