Autoimmune Encephalitis: When New-Onset Psychiatric Symptoms or Seizures Are Actually a Brain Antibody Attack

Research-informed explainer — last updated April 12, 2026

Autoimmune encephalitis is a group of brain inflammatory disorders in which the immune system produces antibodies that directly attack neuronal proteins, causing psychiatric symptoms, seizures, and cognitive changes that are often mistaken for psychiatric illness, viral encephalitis, or epilepsy — but can be reversed with immunotherapy if diagnosed promptly. The field has expanded rapidly in the past two decades, identifying more than 30 distinct antibody-mediated subtypes, each with its own clinical profile, tumor associations, and treatment response.

This article draws on research from five neurologists. Lawrence Honig, MD, PhD, Professor of Neurology at Columbia University Irving Medical Center, is a co-author on the landmark observational study of treatment outcomes and prognostic factors in anti-NMDA receptor encephalitis (3,041 citations) — the largest outcomes study defining long-term recovery patterns. Eric Lancaster, MD, Associate Professor of Neurology at the University of Pennsylvania, co-authored the foundational anti-NMDAR clinical experience study (2,395 citations), identified LGI1 as the target antigen in a major form of limbic encephalitis previously misattributed to potassium channels (1,351 citations), published a comprehensive update on anti-NMDAR mechanisms (829 citations), and authored a review of neuronal autoantigen pathogenesis and antibody testing (492 citations). David Lynch, MD, Professor of Neurology at the Perelman School of Medicine, co-authored the original anti-NMDAR encephalitis case series (3,427 citations), published antibody titer kinetics research (1,270 citations), and characterized the cellular and synaptic mechanisms by which NMDA receptor antibodies reduce receptor surface density (1,116 citations). Daniel Friedman, MD, Professor at NYU Grossman School of Medicine and Director of the Division of Epilepsy, was first author on the seminal study of GABA-B receptor antibodies in limbic encephalitis with seizures (905 citations). Barbara Dworetzky, MD, Professor of Neurology at Brigham and Women's Hospital, has expertise in the psychiatric-neurological differential that delays autoimmune encephalitis recognition, including functional neurological disorder (436 citations) and psychogenic nonepileptic seizures (121 citations).

What autoimmune encephalitis is and why it happens

In autoimmune encephalitis, the immune system generates IgG antibodies targeting proteins on the surface of neurons or at the synapse. These antibodies bind to their targets and either block receptor function, reduce receptor surface density through internalization, or trigger complement-mediated damage. The result is disrupted neural signaling that produces the clinical symptoms.

The best-characterized and most common subtype is anti-NMDA receptor (anti-NMDAR) encephalitis, in which antibodies target the GluN1 subunit of the NMDA glutamate receptor. Lynch's cellular mechanism study demonstrated that patients' antibodies cause a selective and reversible decrease in NMDA receptor surface density at synapses, directly correlated with antibody titer — confirming that the antibodies are not merely markers but the actual pathogenic agents. This mechanism explains why symptoms can resolve when antibodies are cleared by immunotherapy.

Other important subtypes include:

• LGI1 encephalitis: Antibodies targeting leucine-rich glioma-inactivated 1 protein, a synaptic scaffolding protein. Lancaster's case series proved that what had previously been attributed to potassium channel antibodies was actually LGI1 antibody-mediated. LGI1 encephalitis is most common in older men and presents with characteristic faciobrachial dystonic seizures.
• GABA-B receptor encephalitis: Antibodies targeting the GABA-B receptor, causing severe limbic encephalitis with particularly prominent seizures. Friedman's foundational paper in The Lancet Neurology characterized the antigen and clinical features, finding a strong association with small cell lung cancer that makes tumor evaluation essential.
• CASPR2, AMPA receptor, GABA-A receptor encephalitis: Rarer subtypes, each with distinct clinical profiles and tumor associations.

The symptoms that should raise suspicion

Anti-NMDAR encephalitis typically follows a recognizable progression: a brief prodrome of headache and fever, followed by rapid onset of psychiatric symptoms — psychosis, paranoia, bizarre behavior, visual or auditory hallucinations — that are indistinguishable from acute schizophrenia. Within days to weeks, patients develop seizures, decreased consciousness, movement disorders (orofacial dyskinesias, abnormal limb movements), autonomic instability (fluctuating heart rate, blood pressure, and temperature), and hypoventilation.

Lancaster's clinical experience study of 400 patients with anti-NMDAR encephalitis found that more than 75% of patients were initially admitted to psychiatric units and received antipsychotic medications before the correct diagnosis was considered. The median time from symptom onset to diagnosis has historically been several months, during which patients continue to deteriorate.

Dworetzky's expertise in the differential between functional neurological disorder and autoimmune encephalitis is clinically important: psychogenic nonepileptic seizures (PNES) and functional movement disorders can superficially resemble autoimmune encephalitis, but the pattern of escalating psychiatric and cognitive decline with autonomic instability should prompt urgent neurological evaluation for an antibody-mediated cause.

Tumor association: when to look for cancer

Many autoimmune encephalitis syndromes are paraneoplastic — triggered by a tumor that expresses the neuronal protein, prompting an antibody response that then attacks the brain. In anti-NMDAR encephalitis, approximately 38% of women have an associated ovarian teratoma. Removal of the tumor is an important component of treatment, as it eliminates the ongoing antigenic stimulus. In LGI1 encephalitis, tumor association is less common. GABA-B encephalitis is strongly associated with small cell lung cancer.

Current guidelines recommend comprehensive tumor screening in all newly diagnosed autoimmune encephalitis patients, including CT of the chest, abdomen, and pelvis; transvaginal ultrasound or MRI of the pelvis in women; and, in selected cases, PET-CT for occult malignancy.

Treatment and what recovery looks like

The 2013 Lancet Neurology observational study by Honig and collaborators — the largest outcomes study of anti-NMDAR encephalitis — found that 81% of patients improved within 24 months with early immunotherapy. First-line treatment consists of IV methylprednisolone, IV immunoglobulin (IVIG), and/or plasma exchange. Second-line agents — rituximab and cyclophosphamide — are used for patients who do not respond to first-line therapy within 2–4 weeks. The study identified that early initiation of first-line treatment and early escalation to second-line therapy were the strongest predictors of favorable outcomes.

Lynch's antibody titer study found that antibody levels at diagnosis correlated with illness severity, and declining titers over treatment correlated with clinical improvement — providing a quantitative monitoring tool for treatment response.

Relapse occurs in approximately 12–24% of patients following anti-NMDAR encephalitis, and relapses are more common in patients who did not have a tumor identified or who did not receive second-line immunotherapy. Maintenance immunosuppression with mycophenolate mofetil or azathioprine is recommended for patients at high relapse risk.

Questions to ask your doctor

• Given my symptom progression — rapid psychiatric and cognitive decline followed by seizures — should autoimmune antibody testing be sent on blood and CSF?
• What panel of antibodies should be tested, and where should the samples be processed to maximize accuracy?
• Has a paraneoplastic workup been done, and did it include pelvic imaging (for women)?
• If first-line immunotherapy doesn't work within 2 weeks, is escalation to rituximab planned?
• What is the plan for monitoring antibody titers over the course of treatment?
• How long should maintenance immunosuppression continue, and what symptoms would indicate a relapse?

The bottom line

Autoimmune encephalitis is an underrecognized and treatable cause of acute psychiatric and neurological decline. The key to better outcomes is speed: earlier antibody testing, earlier immunotherapy initiation, and earlier escalation to second-line treatment when first-line agents fail. Any patient — particularly a young woman — with rapid-onset psychosis, unexplained seizures, and progressive cognitive decline should be evaluated for autoimmune encephalitis before assuming a primary psychiatric diagnosis. The disease is reversible when caught in time.