Immunotherapy for autoimmune psychosis

Treatment that removes or suppresses the antibodies behind a rare form of psychosis caused by the immune system.

At a glance

Type
Medicine: immune treatment
Used for
Psychosis caused by antibodies against nerve cells, as in anti-NMDA receptor encephalitis
Status
In treatment guidelines, for autoimmune encephalitis; in trials for psychosis with antibodies alone
First used
2007, for anti-NMDA receptor encephalitis, when the disease was first described
Evidence
EmergingHow evidence is rated
Administration
In hospital. First corticosteroids, immunoglobulin or plasma exchange; then rituximab or cyclophosphamide if these are not enough
Starts to work
About half of people with anti-NMDA receptor encephalitis improve within 4 weeks; recovery can continue for 18 months
Common side effects
Infections, a brief worsening of psychiatric symptoms with high-dose corticosteroids, blood clots with immunoglobulin
Checks needed
At diagnosis, a brain scan, an EEG, blood and spinal fluid tests for antibodies, and a search for a tumour
In Europe
Given for autoimmune encephalitis in neurology and psychiatry departments. Tested for psychosis with antibodies alone in the SINAPPS2 trial in England
Clinical summaryFor health professionals

What it is

In a small number of people, psychosis is caused by the immune system. Antibodies attack receptors on the surface of nerve cells, most often the NMDA receptor for glutamate, and the condition is then called autoimmune psychosis.1 Its best-known form is anti-NMDA receptor encephalitis, an inflammation of the brain. Among the first 100 patients described in detail, every one began with psychiatric symptoms or memory problems, and most went on to develop seizures, involuntary movements and reduced consciousness.2

Immunotherapy removes these antibodies or stops the body from making them. It is the standard treatment for autoimmune encephalitis, although no medicine is approved for this use and the evidence comes from large observational studies rather than placebo-controlled trials.1,3 Whether it helps people who have the antibodies and psychosis but no other sign of encephalitis is being tested.4

How it works

The antibodies attach to NMDA receptors and make the nerve cell draw them inside, so fewer receptors remain to receive signals. The loss grows with the amount of antibody and reverses when the antibodies are removed.5

Treatment therefore works in two steps. The first lowers the antibodies in the blood: plasma exchange filters them out, intravenous immunoglobulin (IVIG), a concentrate of antibodies from blood donors, is given for the same purpose, and a tumour that drives their production is removed. The second suppresses further production, with corticosteroids or medicines such as azathioprine or mycophenolate. Rituximab (opens in a new tab) removes the B cells that mature into antibody-producing cells.1 In some patients, B cells in the blood and in an ovarian teratoma keep producing the antibodies.6 If the first treatments bring no improvement within two to four weeks, experts advise moving on to rituximab or cyclophosphamide.7

Recognising it

In 2020, an international group of 28 experts in psychiatry and neurology agreed on when to suspect it. Possible autoimmune psychosis means psychotic symptoms that began abruptly and developed within three months, together with at least one of the following:1

  • a tumour, now or recently
  • catatonia or involuntary movements
  • a severe reaction to antipsychotics, with rigid muscles, fever or a raised level of the muscle enzyme creatine kinase in the blood
  • problems with thinking that are severe or out of proportion
  • reduced consciousness
  • seizures without a known cause
  • marked swings in blood pressure, temperature or heart rate

In that case, the consensus advises an EEG, an MRI scan, and tests of the blood and the spinal fluid for antibodies against nerve cells, together with a search for a tumour. Antibodies found in the blood alone, with normal results on every other test, are not enough reason for immunotherapy; the decision is made case by case with neurologists.1

Such antibodies are uncommon. In European studies of 1,114 people with psychosis, 41 had antibodies against the NMDA receptor in their blood, and their symptoms were on average milder than those of people without them.8 In a Danish study of 104 people with recent-onset psychosis, none had antibodies against the NMDA receptor.9

How well it works

The largest study of anti-NMDA receptor encephalitis followed 577 patients. Almost all received first-line immunotherapy or had a tumour removed, and about half improved within four weeks. Among those who did not, the patients given second-line treatment did better than the rest. Within two years, 79 in 100 had recovered or were left with no more than slight disability, and 30 had died. Recovery continued for up to 18 months, 12 in 100 relapsed within two years, and patients treated early fared better.3 A pooled analysis of 1,550 published patients linked going without immunotherapy in the first month to a worse outcome.10

For psychosis with antibodies but no encephalitis, the evidence so far comes from case series. In a German series of 91 patients, 83 in 100 improved after immunotherapy, but the patients were partly selected because they had improved.11 The first placebo-controlled trial, SINAPPS2, is under way in England. People whose psychosis began within the previous two years, and who have antibodies against nerve cells but no encephalitis, receive IVIG followed by rituximab, or a placebo.4 Recruitment closed on 31 March 2026, and the results have not yet been published.12

Side effects

High-dose corticosteroids can at first worsen psychiatric symptoms. IVIG can cause blood clots and lower the sodium level in the blood. Plasma exchange needs a large line into a vein and is less suitable for agitated patients. Rituximab raises the risk of infections, sometimes severe ones, and can cause reactions during the infusion. Cyclophosphamide can suppress the bone marrow and reduce fertility.7

In encephalitis, antipsychotics can bring on a reaction resembling neuroleptic malignant syndrome. Experts advise starting them at a low dose, and preferring newer antipsychotics to older ones.1

What we don’t know yet

  • Whether immunotherapy helps people who have antibodies in the blood but no other sign of encephalitis.
  • How to tell antibodies that cause illness from antibodies that do no harm.

Sources

  1. Pollak TA, Lennox BR, Müller S, et al. Autoimmune psychosis: an international consensus on an approach to the diagnosis and management of psychosis of suspected autoimmune origin. The Lancet Psychiatry. 2020;7(1):93–108.
    DOI (opens in a new tab)PubMed (opens in a new tab)
  2. Dalmau J, Gleichman AJ, Hughes EG, et al. Anti-NMDA-receptor encephalitis: case series and analysis of the effects of antibodies. The Lancet Neurology. 2008;7(12):1091–1098.
    DOI (opens in a new tab)PubMed (opens in a new tab)
  3. Titulaer MJ, McCracken L, Gabilondo I, et al. Treatment and prognostic factors for long-term outcome in patients with anti-NMDA receptor encephalitis: an observational cohort study. The Lancet Neurology. 2013;12(2):157–165.
    DOI (opens in a new tab)PubMed (opens in a new tab)
  4. Lennox B, Yeeles K, Jones PB, et al. Intravenous immunoglobulin and rituximab versus placebo treatment of antibody-associated psychosis: study protocol of a randomised phase IIa double-blinded placebo-controlled trial (SINAPPS2). Trials. 2019;20(1):331.
    DOI (opens in a new tab)PubMed (opens in a new tab)
  5. Hughes EG, Peng X, Gleichman AJ, et al. Cellular and Synaptic Mechanisms of Anti-NMDA Receptor Encephalitis. The Journal of Neuroscience. 2010;30(17):5866–5875.
    DOI (opens in a new tab)PubMed (opens in a new tab)
  6. Makuch M, Wilson R, Al‐Diwani A, et al. N‐methyl‐D‐aspartate receptor antibody production from germinal center reactions: Therapeutic implications. Annals of Neurology. 2018;83(3):553–561.
    DOI (opens in a new tab)PubMed (opens in a new tab)
  7. Mulic-Al Bunni S, Gschwind M, Finkener S, et al. The Treatment of Antibody-Mediated Encephalitis: Current, Future Therapies, Unmet Need and Patient Management. Advances in Therapy. 2026;43(5):1891–1925.
    DOI (opens in a new tab)PubMed (opens in a new tab)
  8. Luykx JJ, Visscher R, Winter-van Rossum I, et al. Clinical symptoms and psychosocial functioning in patients with schizophrenia spectrum disorders testing seropositive for anti-NMDAR antibodies: a case–control comparison with patients testing negative. The Lancet Psychiatry. 2024;11(10):828–838.
    DOI (opens in a new tab)PubMed (opens in a new tab)
  9. Jeppesen R, Nilsson AC, Sørensen NV, et al. Antineuronal antibodies in cerebrospinal fluid and serum of 104 patients with psychotic disorders compared to 104 individually matched healthy controls. Schizophrenia Research. 2023;252:39–45.
    DOI (opens in a new tab)PubMed (opens in a new tab)
  10. Nosadini M, Eyre M, Molteni E, et al. Use and Safety of Immunotherapeutic Management of N-Methyl-d-Aspartate Receptor Antibody Encephalitis. JAMA Neurology. 2021;78(11):1333.
    DOI (opens in a new tab)PubMed (opens in a new tab)
  11. Endres D, Lüngen E, Hasan A, et al. Clinical manifestations and immunomodulatory treatment experiences in psychiatric patients with suspected autoimmune encephalitis: a case series of 91 patients from Germany. Molecular Psychiatry. 2022;27(3):1479–1489.
    DOI (opens in a new tab)PubMed (opens in a new tab)
  12. ISRCTN registry. SINAPPS2: Immunotherapy for psychosis associated with neuronal membrane autoantibodies (ISRCTN11177045). 2026.
    isrctn.com (opens in a new tab)

Not medical advice. Talk to a doctor before starting or changing any treatment. Disclaimer

Spotted an error? Tell us. Corrections are dated and listed on Corrections.

Page last reviewed: 5 November 2026Next review due: May 2027

The Commissure newsletter

Monthly research updates.