Autoimmune Encephalitis Antibodies

Autoimmune encephalitis antibodies are now identified by phenotype-driven Mayo panels rather than a single legacy reflexive test. This page is the lab-side reference: what to order, what assay, what sample, titer interpretation, and how the result changes treatment. Clinical phenotype detail (psychiatric course, dyskinesia patterns, seizure semiology) is left for the Neuroimmunology / clinical chapters; here, the focus is on the assay itself and its decision impact.

🔹 Bottom Line: Autoimmune Encephalitis Antibodies

  • Cell-surface antibodies (NMDAR, LGI1, CASPR2, AMPAR, GABA-B, DPPX, glycine-R, mGluR5) are pathogenic and treatable; cell-based assay (CBA) is the gold standard.
  • Intracellular antibodies (Hu, Yo, Ri, Ma2, CV2, amphiphysin) are paraneoplastic markers — see Paraneoplastic Antibodies page.
  • PAVAL is gone (retired mid-2024). Order Mayo phenotype panels in parallel serum + CSF versions: encephalopathy (ENS2/ENC2), epilepsy (EPS2/EPC2), dementia (DMS2/DMC2), movement (MDS2/MDC2).
  • CSF more sensitive than serum: anti-NMDAR, GFAP-IgG. Serum more sensitive than CSF: LGI1, CASPR2, intracellular onconeural antibodies. Send both when uncertain.
  • Test BEFORE immunotherapy when possible — steroids/IVIG/PLEX can transiently lower titers and cause false negatives.
  • “VGKC-complex” positivity without LGI1/CASPR2 specificity is uninterpretable — do not treat on the basis of VGKC-complex alone.
  • Graus criteria support empirical immunotherapy in compelling syndromes before antibody results return.

Two Mechanistic Categories — Why It Matters for Testing

Cell-Surface (Synaptic) Antibodies — Pathogenic

  • Targets: ion channels, receptors, adhesion molecules on neurons (NMDAR, LGI1, CASPR2, AMPAR, GABA-B-R, DPPX, glycine receptor, mGluR5, IgLON5).
  • The antibody is itself pathogenic — directly disrupts receptor function or causes internalization.
  • Methodology: cell-based assay (CBA) is the gold standard. Older indirect immunofluorescence (IIF) misses some antibodies and is non-specific at low titers.
  • Generally respond well to immunotherapy — positive result mandates aggressive treatment.

Intracellular (Onconeural) Antibodies — Paraneoplastic Markers

  • Targets: nuclear or cytoplasmic neural antigens (Hu, Yo, Ri, Ma2, CV2/CRMP5, amphiphysin).
  • The antibody is a marker, not the mechanism — T-cell-mediated injury causes the damage.
  • Methodology: IIF screen + line-blot or immunoprecipitation confirmation.
  • Generally poor response to immunotherapy — tumor treatment is the cornerstone. Covered on the Paraneoplastic Antibodies page.

Antibody Profiles — Testing Focus

Anti-NMDA Receptor (NMDAR)

  • Sample: CSF more sensitive than serum — order BOTH. Negative serum does not exclude anti-NMDAR encephalitis.
  • Methodology: cell-based assay (gold standard).
  • Tumor link: ovarian teratoma in ~50% of young women; lower in men and older patients. Pelvic MRI / TV-US mandatory in women.
  • Treatment impact: positive → steroids + IVIG or plasmapheresis (first-line); rituximab + cyclophosphamide (second-line); tumor removal when present. Empirical immunotherapy is justified before antibody results return in compelling clinical syndromes.

Anti-LGI1

  • Sample: serum is sufficient (CSF less sensitive than for anti-NMDAR).
  • Methodology: cell-based assay. Do not rely on “VGKC-complex” testing alone — VGKC-complex without LGI1/CASPR2 specificity is uninterpretable.
  • Tumor link: low (~10%, mostly thymoma).
  • Treatment impact: steroids + IVIG/PLEX (often dramatic FBDS response). Hyponatremia (SIADH) is a clue to expect with positive LGI1.

Anti-CASPR2

  • Sample: serum.
  • Methodology: cell-based assay.
  • Tumor link: thymoma in ~20–30%.
  • Treatment impact: steroids + IVIG/PLEX; rituximab for severe / refractory disease.

Anti-GABA-B Receptor

  • Sample: serum; CSF can be added for confidence.
  • Methodology: cell-based assay.
  • Tumor link: strong SCLC association (~50%) — mandatory tumor screening when positive.
  • Treatment impact: immunotherapy + tumor treatment.

Anti-AMPA Receptor

  • Sample: serum and CSF.
  • Methodology: cell-based assay.
  • Tumor link: thymoma, SCLC, breast cancer.
  • Treatment impact: immunotherapy + tumor treatment.

Anti-DPPX

  • Sample: serum and CSF.
  • Methodology: cell-based assay.
  • Tumor link: usually NOT paraneoplastic; rarely B-cell lymphoma.
  • Treatment impact: immunotherapy. Distinctive GI prodrome (diarrhea + weight loss) is a clue.

Anti-Glycine Receptor

  • Sample: serum and CSF.
  • Methodology: cell-based assay.
  • Tumor link: rarely paraneoplastic.
  • Treatment impact: immunotherapy. PERM (progressive encephalomyelitis with rigidity and myoclonus) and exaggerated startle are clues.

Anti-mGluR5 (Ophelia Syndrome)

  • Sample: serum and CSF.
  • Tumor link: Hodgkin lymphoma — order in young patients with limbic encephalitis + lymphoma history or risk.
  • Treatment impact: immunotherapy + lymphoma treatment.

Anti-IgLON5

  • Sample: serum and CSF.
  • Methodology: cell-based assay.
  • Tumor link: generally none.
  • Treatment impact: variable response to immunotherapy. Tauopathy on autopsy — partial overlap with neurodegenerative mechanism.

Anti-GFAP IgG

  • Sample: CSF more sensitive than serum. Order both, with CSF as primary.
  • Methodology: cell-based assay.
  • Tumor link: ~25% paraneoplastic — ovarian teratoma, others.
  • Treatment impact: often dramatically steroid-responsive. Linear perivascular radial enhancement on MRI is the imaging clue.
  • Covered in detail on the CNS / Demyelinating Antibodies page.

Anti-GAD65 (Encephalitis Context)

  • Sample: serum (CSF for intrathecal synthesis adds confidence).
  • Methodology: radioimmunoassay or ELISA.
  • Titer interpretation — critical: low titers are common in T1DM, autoimmune thyroid, and healthy controls — not clinically significant for neurologic disease at low titer. Very high titers (typically >100× threshold) required to attribute neurologic syndromes.
  • Treatment impact: generally poor immunotherapy response — intracellular antibody mechanism. Treatment trial is reasonable but disappointing in most cases.

How to Order Antibody Panels

The Shift to Phenotype-Driven Panels

Mayo Clinic Laboratories — the dominant US reference lab — retired the broad PAVAL serum panel in mid-2024 in favor of phenotype-specific evaluations. Pick the panel that matches the leading clinical phenotype, in parallel serum + CSF versions.

🔹 Clinical Relevance: Match the Panel to the Phenotype

Before ordering, decide which clinical syndrome best describes the patient — encephalopathy, epilepsy, dementia, movement disorder, myelopathy, peripheral/autonomic. Then send the matching phenotype panel in both serum AND CSF unless one is clearly preferred.

The APE2 (Antibody Prevalence in Epilepsy & Encephalopathy) scorecard helps decide pre-test whether antibody testing is likely to be informative.

Mayo Phenotype Panels (2024–2025 catalog)

Phenotype Serum CSF When to choose
Encephalopathy — limbic encephalitis, FBDS, behavioral / amnestic / psychiatric ENS2 ENC2 Most autoimmune encephalitis presentations
Epilepsy — adult new-onset focal seizures, refractory epilepsy without structural cause EPS2 EPC2 Adult new-onset focal epilepsy, especially with autonomic features or rapid cognitive decline
Dementia — rapidly progressive or atypical dementia, “young” dementia DMS2 DMC2 Rapidly progressive dementia <65 years; atypical cognitive decline; after CJD workup
Movement disorders — chorea, ataxia, parkinsonism, brainstem encephalitis MDS2 MDC2 Subacute ataxia / chorea / parkinsonism with cancer risk or atypical features

Each panel uses cell-based assays for surface antigens and IIF + line blot for intracellular onconeural antibodies. Equivalent phenotype panels are available from other US labs (Quest / Athena, ARUP) — codes differ; the paradigm applies.

CSF vs Serum — Which Is Better for Which Antibody

  • CSF more sensitive than serum: NMDA-R, GFAP-IgG, some glycine-R and DPPX. Don’t rule out anti-NMDAR encephalitis on a negative serum alone.
  • Serum more sensitive than CSF: LGI1, CASPR2, most intracellular onconeural antibodies (Hu, Yo, Ri, Ma2, CV2, amphiphysin).
  • Send both when uncertain.
  • Cell-based assays are the modern standard for surface antibodies; older IIF results without CBA confirmation should be interpreted with caution.

Practical Tips

  • Order paired serum + CSF phenotype panels rather than legacy “send everything” approaches.
  • APE2 scorecard helps gauge pre-test probability — useful in epilepsy / encephalopathy.
  • Collect samples before immunotherapy when possible.
  • Repeat borderline / low-titer surface antibody 8–12 weeks apart — persistence supports significance.
  • The phenotype panels include the major intracellular onconeurals plus cell-surface targets — you generally don’t need a separate paraneoplastic panel.

Graus Criteria — Empirical Treatment Before Results Return

The 2016 Graus criteria support starting immunotherapy in compelling clinical syndromes before antibody results return:

  • Possible autoimmune encephalitis (all 3): subacute (<3 months) onset of working memory deficits / altered mental status / psychiatric symptoms; at least one of (focal CNS findings, seizures, CSF pleocytosis, MRI features); reasonable exclusion of alternative causes.
  • Definite autoimmune encephalitis: above + specific antibody OR classic syndrome features (e.g., anti-NMDAR phenotype + teratoma + CSF pleocytosis).

Antibody results often take weeks. In compelling cases, start steroids + IVIG (or plasmapheresis) without waiting — the rapid response can be diagnostic, and delays worsen long-term outcomes.

Interpreting Equivocal or Low-Positive Results

  • Anti-GAD65 low titer: very common; usually not neurologically significant.
  • “VGKC-complex” without LGI1/CASPR2 specificity: uninterpretable; do NOT treat on this alone.
  • Anti-MOG low-positive: low specificity; high titer required for confidence in MOGAD (see CNS / Demyelinating Antibodies page).
  • Serum negative + CSF negative in a classic clinical syndrome: consider Hashimoto encephalopathy / SREAT (anti-TPO) or treat empirically when the syndrome is compelling.

Pitfalls and Pearls

  • NMDAR: CSF > serum; image the ovaries in women; empirical immunotherapy in compelling syndromes.
  • LGI1: serum sufficient; SIADH/hyponatremia is a clue; FBDS phenotype is classical.
  • CASPR2: 20–30% thymoma — image the chest.
  • GABA-B: ~50% SCLC — chest CT mandatory.
  • AMPA: thymoma / SCLC / breast — multi-organ tumor workup.
  • DPPX: distinctive GI prodrome; usually NOT paraneoplastic.
  • mGluR5: Hodgkin lymphoma (Ophelia).
  • GAD65: high titer only; poor immunotherapy response (intracellular mechanism).
  • GFAP: CSF > serum; linear perivascular radial enhancement on MRI; steroid-responsive.
  • PAVAL retired (mid-2024): order Mayo phenotype panels (ENS2/ENC2, EPS2/EPC2, DMS2/DMC2, MDS2/MDC2) instead.
  • “VGKC-complex” without LGI1/CASPR2: uninterpretable.
  • Cell-based assay is the modern standard for surface antigens.
  • Send before immunotherapy when possible.
  • Graus criteria: empirical immunotherapy in compelling syndromes BEFORE results return.
  • Empty results in compelling syndrome: consider Hashimoto / SREAT (anti-TPO).

References

  1. Graus F, Titulaer MJ, Balu R, et al. A clinical approach to diagnosis of autoimmune encephalitis. Lancet Neurol. 2016;15(4):391-404.
  2. Dalmau J, Graus F. Antibody-mediated encephalitis. N Engl J Med. 2018;378(9):840-851.
  3. Titulaer MJ, McCracken L, Gabilondo I, et al. Treatment and prognostic factors for long-term outcome in patients with anti-NMDA receptor encephalitis. Lancet Neurol. 2013;12(2):157-165.
  4. Irani SR, Alexander S, Waters P, et al. Antibodies to Kv1 potassium channel-complex proteins (LGI1, CASPR2) in limbic encephalitis, Morvan’s syndrome and acquired neuromyotonia. Brain. 2010;133(9):2734-2748.
  5. Mayo Clinic Laboratories. Autoimmune Neurology Testing — phenotype-specific evaluations (ENS2/ENC2, EPS2/EPC2, DMS2/DMC2, MDS2/MDC2, MAS2/MAC2, AIAES). news.mayocliniclabs.com/neurology/autoimmune-neurology/.
  6. Mayo Clinic Laboratories. PAVAL retired mid-2024 in favor of phenotype-driven panels.
  7. Dubey D, Singh J, Britton JW, et al. Predictive models in the diagnosis and treatment of autoimmune epilepsy. Epilepsia. 2017;58(7):1181-1189 (APE2 score).
  8. Dubey D, Pittock SJ, Kelly CR, et al. Autoimmune encephalitis epidemiology and a comparison to infectious encephalitis. Ann Neurol. 2018;83(1):166-177.