The immune system is designed to protect our bodies from infection, and from 'foreign' bodies that have somehow entered (for example viruses or bacteria that have entered through a cut in the skin). One of the ways in which the immune system works is to produce molecules known as antibodies, which bind to and destroy the unwanted agent. Sometimes, however, the body accidentally produces antibodies against one of its own proteins, for example a particular receptor, hormone or enzyme, which can often result in illness. This is known as autoimmunity and the antibodies are called auto-antibodies. Type 1 diabetes, rheumatoid arthritis and multiple sclerosis are all examples of autoimmune conditions, and they are diagnosed from both their clinical effects and the detection of particular auto-antibodies in the blood stream.

Researchers suspect that autoimmunity is involved in some forms of epilepsy, and since 1980 there have been many attempts to prove the existence of auto-antibodies to brain proteins. The hope is that, if autoimmune forms of epilepsy are discovered, these might respond to immunotherapy where anti-epileptic drugs (AEDs) have been unsuccessful.

Limbic encephalitis (LE) is acute inflammation of the limbic system, a set of structures (including the hippocampus) that controls emotion, behaviour, long-term memory and smell. Symptoms of LE include memory problems, confusion, depression, agitation, dementia, hallucinations and personality changes. It can also result in seizures, both as a direct result of the inflammation and through long-term damage to the hippocampus known as sclerosis (this is a common cause of temporal lobe epilepsy). LE is often caused by a viral infection, or in response to certain cancers including lung and breast. In the last five years, evidence has shown that the cause of LE can also be autoimmune, i.e. seizures can result from the direct effects of auto-antibodies.

The auto-antibodies that have been linked to LE include one against an enzyme involved in GABA production, called glutamic acid decarboxylase (GAD) and one against a type of potassium ion channel referred to as VGKC. Antibodies to two glutamate receptors, N-methyl-D-aspartate (NMDA) and glutamate receptor epsilon 2 have also been identified by different groups; and the epsilon 2 receptor has been classed as a subunit of the NMDA receptor. It is known as NR2B NMDA.

The four auto-antibodies described are already being tested for in some centres, in the diagnosis of LE. It is important that autoimmune conditions are detected as early as possible, so that the appropriate treatment can be administered before the damage is too severe.

These antibodies are also increasingly being detected in some people for whom epilepsy is their presenting, dominant or only clinical feature. Does this mean that there is a case for routine auto-antibody testing in epilepsy?

This month a review, bringing together current evidence for autoimmunity in epilepsy and LE, was published in the journal Current Opinion Neurology. Epilepsy Research UK is particularly interested in this, because we have partly funded this research. In 2008 we awarded one of the authors of the review, Dr Bethan Lang
(University of Oxford), £55,258 over 12 months, for a project entitled 'Detection of auto-antibodies in patients with epilepsy'.

According to the review, studies in which the blood of large numbers of people with different forms of epilepsy have been screened for VGKC and GAD antibodies have shown some promise. In one, 16% of the 139 people examined had VGKC antibodies. These people represented a variety of forms of epilepsy, including a few with recent onset LE. VGKC antibodies were less common, however, in people who had had drug-resistant epilepsy for a long time (6% of the people in this category).

GAD antibodies were not often detected in this study (in only 1-2% of people); however they were found in 6% of children with myoclonic epilepsy. In addition, two more recent studies revealed significant levels of GAD antibodies in 2-6% of participants who had temporal lobe epilepsy.

Looking at the full range of studies that have been performed, the review authors conclude that there is some evidence to support the existence of autoimmune mechanisms in a proportion of epilepsy syndromes, but 'not (enough) to establish a role for these particular antibody tests in the routine assessment of epilepsy, unless there are features suggesting an autoimmune basis.' So what features should doctors be looking for?

The review describes several studies that have examined the characteristics of LE caused by each of the four auto-antibodies, VGKC, NMDA receptor, NMDA NR2B and GAD. A summary of the findings can be viewed in table 1.

On the basis of these observations, the authors recommend that every person with an acute/subacute onset of seizures, amnesia or psychological disturbance, which cannot be explained by other causes such as viral infections, tumours, or drugs, be tested for antibodies to VGKC, NMDA receptors and GAD; particularly if the individual shows any other signs of autoimmune LE. They also highlight the importance of early treatment in order minimise the damage to the hippocampus and maximise seizure control.

The study of autoimmune epilepsy associated with specific antibodies is still in its infancy, but the authors hold hope that, in the future, auto-antibody testing will play an important role in the diagnosis and treatment of forms of epilepsy in patients who may otherwise prove to be AED unresponsive.

Reviews such as this are important, because they bring together a wide range of existing evidence, allowing scientists to build a clearer picture of a given field. This can influence the scope of future research projects, maximising their long-term impact on patients. We look forward to receiving further findings from Dr Lang and her colleagues.

Read more here