FOSUN HEALTH Logo

Autoimmune Encephalitis Causes

catalogue

Autoimmune Encephalitis Causes

Autoimmune Encephalitis Causes

In autoimmune encephalitis the immune system targets a protein that a nerve cell needs in order to work. Everything that follows — how severe the illness is, whether a tumour has to be hunted for, whether the treatment works, and how much comes back — is shaped by one question: is that protein on the surface of the cell, or inside it? This page sets out the mechanism in plain language, the two families of antibody and why the distinction governs treatment, what is known about triggers including tumours and infections, and — because families ask it in every consultation — what this means for other members of the family.

What is actually happening in the brain

A nerve cell communicates with the next one across a junction, using chemical signals that are released on one side and received by receptors on the other. Those receptors are proteins, and they are what the immune system targets in this illness.

When an antibody binds a receptor that sits on the cell surface, three things can happen. It can block the receptor, so the signal is not received. It can cross-link and internalise it, pulling the receptor off the surface so there is less of it there to receive anything. Or it can trigger inflammation around the junction, altering how the cell handles the signal. All three change how the circuit behaves without necessarily killing the cell — which is the reason recovery is possible at all, and the reason the damage is not simply proportional to how long the illness ran.

Where the target is inside the cell, the mechanism is different and the antibody is not the weapon. The antibody is a marker: its presence tells you that immune cells — T cells — have been activated against that protein and are attacking the cells that contain it. That attack kills cells, which is why the intracellular group tends to produce more fixed deficits and responds less well to treatments that remove antibody.

Two consequences follow, and they explain most of the management. Surface antibodies are treatable targets. Removing the antibody, or stopping the cells that make it, changes the disease. Intracellular disease is driven by cells and, usually, by a tumour. Removing the antibody changes less, and finding the tumour changes more.

And the piece that explains why this illness presents as behaviour rather than as weakness: many of these proteins are concentrated in the circuits that serve memory, emotion and judgement — the medial temporal lobes and their connections — rather than in the motor pathways. An immune attack on those circuits produces a person who is frightened, forgetful and changed, with normal strength. Why the symptoms look the way they do follows from this directly.

Two families of antibody, and why it matters

If one distinction from this whole guide is worth carrying away, it is this one. It decides how hard to look for a tumour, how aggressively to treat, and what to tell a family about the likely course.

Cell-surface and synaptic antibodies Intracellular antibodies
Examples NMDA receptor, LGI1, CASPR2, AMPAR, GABAA receptor, DPPX, IgLON5 Hu / ANNA-1, Yo, Ri, Ma2, CRMP5, amphiphysin, GAD65
Is the antibody harmful itself? Usually yes — it binds something it can reach and alters how the cell signals No — it cannot reach its target. It is a marker of an immune-cell attack
What does the damage The antibody, and the inflammation around it T cells, which kill the cells containing the protein
Tumour association Variable, and often absent; a teratoma in NMDA receptor disease is the clearest example Usual. The tumour is the driver in most cases
Response to immunotherapy Generally good, and often the defining feature of the illness Generally poorer; tumour removal is the dominant intervention
Typical course Subacute onset, often severe, with a long but real recovery over months Often more gradual onset and a more fixed deficit; progression depends on the tumour
What the result changes Start immunotherapy; tumour search is real but usually bounded Search thoroughly and repeat; treat the tumour as the priority alongside immunotherapy

Two honest caveats. The boundary is not absolute — IgLON5, which is classified with the surface group, responds less predictably than the others; and GAD65, in the intracellular group, is often not associated with a tumour. And more than one antibody can be present, in which case the clinical picture decides which is treated first rather than the laboratory list. How the panel is run and read is set out on the diagnosis page.

What triggers it — and why usually nothing is found

The most likely answer, statistically, is that no trigger will be identified. That is the commonest outcome of a thorough search and it is not a failure of the search.

Where a trigger is found, it is usually one of three things.

  • A tumour expressing the same protein. The clearest documented mechanism. The immune response begins against the tumour and cross-reacts with the nervous system. Ovarian teratoma in younger women with NMDA receptor disease is the best-known example; small cell lung cancer, thymoma, breast and testicular tumours account for most of the intracellular group.
  • A preceding infection. Best documented for herpes simplex encephalitis in the brain, where a second illness can emerge weeks after recovery, with psychiatric symptoms and abnormal movements rather than with a return of fever and headache. Other viral illnesses are reported as precedents far more often than they are proven, and the distinction is worth keeping.
  • Immune-modulating cancer therapy. Checkpoint inhibitor drugs release the brakes on the immune system to attack a tumour, and occasionally the released response turns on the nervous system. The neurological illness can appear at any point during or after that treatment, and recognising it early matters because the immunotherapy used to treat it is given alongside decisions about the cancer treatment rather than instead of them.

What does not cause it. Stress, parenting, personality, diet, and anything the person did or did not do. This is worth stating in as many words as possible, because the question is asked in nearly every consultation and the answer is genuinely reassuring rather than merely polite.

What raises the background probability. Another autoimmune disease in the person, autoimmune disease in a first-degree relative, and — for specific antibodies — age and sex patterns: NMDA receptor disease is more often seen in younger adults and children and more often in women, LGI1 disease more often in men over fifty. None of these is a requirement, none is a screen, and the absence of all of them does not make the diagnosis less likely in a person whose clinical picture fits.

The tumour link, explained properly

The word paraneoplastic describes an illness that is caused by the immune response to a tumour rather than by the tumour itself. It is one of the more useful ideas in neurology, and one of the more frightening words to hear — which is a mismatch worth correcting, because in this illness finding the tumour is usually the better news rather than the worse.

How it works. A tumour arises that happens to make a protein normally found only in nerve cells. The immune system recognises that protein as foreign, mounts a response, and the response does not distinguish between the protein in the tumour and the same protein in the brain. The nervous system is damaged by an immune response that was, in a sense, aimed correctly at something else. That is why removing the tumour is the intervention most likely to change the course, and why immunotherapy alone — suppressing the response while the stimulus remains — is the weaker half of the treatment.

Why the tumour is often small and hard to find. Because the immune response is very effective at controlling it. A tumour that provokes a strong neurological immune response is frequently small, slow, and confined to a lymph node or an organ edge, which is why cross-sectional imaging can miss it and why PET-CT and repeat screening earn their place in the intracellular group. A negative first screen in that group is the beginning of the search rather than the end of it.

What it means for the family. That a tumour was found does not mean the neurological illness was misdiagnosed, or that the person has cancer in the way that word usually lands. It means the immune attack has an identifiable source, that removing it is a concrete step, and that the neurological treatment has a mechanism to work with. Where no tumour is found after a search appropriate to the antibody, treatment proceeds on the same basis and the outcome is not determined by the absence of that finding.

What is known, what is not, and what is being studied

This is a disease described in detail only in the last two decades, which means the map is genuinely incomplete in places and it is worth being honest about where.

Well established. That these illnesses are immune-mediated and can be diagnosed on clinical criteria with antibody confirmation; that the cell-surface and intracellular distinction predicts both the tumour association and the treatment response; that tumour removal is the dominant intervention in paraneoplastic disease; that earlier treatment is associated with better recovery; and that recovery, when it comes, is measured in months.

Reasonably well established, still being refined. Which second-line agent to use and when; how long maintenance treatment should continue and how it should be tapered; which people can safely stop immunosuppression entirely; and what the optimal intensity of the tumour search is for each antibody. These are areas where practice varies between centres for defensible reasons rather than through neglect.

Not established. Why most cases have no identifiable trigger. Why the same antibody produces different illnesses in different people. Whether anything can be done before the illness starts to reduce the chance of it, which is the reason the recovery page is about protecting the recovery rather than about preventing the disease — no intervention has been shown to prevent autoimmune encephalitis, and it would be dishonest to imply otherwise.

Where the research is going. Better and faster antibody assays, including on blood where a lumbar puncture is not possible; treatments aimed at the long-lived plasma cells that sustain antibody production; biomarkers that predict relapse so that maintenance can be individualised rather than standardised; and structured rehabilitation trials, which is arguably where the largest gain for patients remains untapped.

The closing sentence of this page is the one that matters most for the person reading it: the incompleteness of the map is not a reason to delay the parts that are certain. Aciclovir while infection is excluded, immunotherapy once the criteria are met, tumour removal where a tumour is found, and rehabilitation throughout — none of these waits for the research.

Frequently Asked Questions

Did something cause this — a virus, stress, or something we did?

In most cases no cause is ever found, and the absence of one is the commonest outcome rather than an incomplete investigation. Where a trigger is identified it is most often a tumour expressing the same protein as the nerve cell, or a preceding infection — best documented after herpes simplex encephalitis, where a second illness can appear weeks later. Stress does not cause this illness, and neither does anything the person or the family did. It is worth saying plainly, because families carry that question for a long time and the honest answer is that nobody gave themselves this.

Is it hereditary? Will my other children get it?

It is not inherited in the way a genetic condition is. What runs in families is a tendency to autoimmunity rather than this illness: a first-degree relative with an autoimmune disease is a recognised background factor, but it does not make the disease predictable and it does not mean screening asymptomatic relatives is useful. What is worth doing instead is making sure the family knows the early signs, because the thing that changes outcome in this illness is speed of recognition rather than genetic risk.

Why did it happen after a herpes encephalitis that had already been treated?

Because infection can be the trigger rather than only the mimic. Herpes simplex encephalitis damages brain tissue and releases the very proteins that the immune system then learns to target; a proportion of people — children more visibly than adults — develop a second illness some weeks after they appeared to be recovering, with new psychiatric symptoms, abnormal movements or a fall in consciousness rather than with a return of the fever and headache. The practical rule: any neurological or psychiatric deterioration in the weeks after herpes encephalitis should be assessed for anti-NMDA receptor encephalitis, not assumed to be a relapse of the infection or a delayed recovery.

If the antibody is found, does that mean it is causing the damage?

It depends entirely on which antibody, and the distinction is the single most important thing the result tells you. Antibodies against proteins on the cell surface or at the synapse are, in most cases, directly pathogenic — they bind something accessible and interfere with how the nerve cell signals. Antibodies against proteins inside the cell cannot reach their target; they are markers of an immune-cell-mediated attack, and the damage is being done by cells rather than by the antibody. That is why the intracellular group responds less well to treatments that remove antibody, and why it points so strongly towards a tumour.

Find out whether the trigger search has been adequate

Send the antibody report with the fluid on which it was run, every imaging study done for tumour screening and its date, a note on any infection preceding the illness, and a list of cancer treatments received including immunotherapy. Our team will return a written opinion on whether the search has been appropriate to the antibody, whether it needs to be repeated and when, and what the findings mean for the treatment sequence.

Want to see the whole process first? Your full medical journey, step by step →

Book an Appointment Request a Video Consultation

Follow Fosun Health

Apply for an Appointment

Do you have test results*