Vascular Headache Epilepsy Immunology Movement Neuromuscular Dementia Other Topics Neurosurgery

Autoimmune Epilepsy

What Do You Need to Know?

Prevalence: Autoimmune etiologies account for ~5–7% of all epilepsies and up to 20% of epilepsies of unknown cause
Two antibody classes: Cell-surface antibodies (NMDA-R, LGI1, CASPR2, GABA-B, GABA-A, AMPA) are directly pathogenic and treatment-responsive; intracellular antibodies (GAD65, ANNA-1/Hu, CV2/CRMP5, amphiphysin) are T-cell mediated, poorly immunotherapy-responsive, and strongly paraneoplastic
Clinical clues: Subacute onset, drug-resistant seizures, new psychiatric symptoms, FBDS, hyponatremia, autonomic dysfunction, temporal FLAIR changes
APE2 score ≥4: 97% sensitivity for predicting positive neural-specific antibodies — validated screening tool
Treatment paradigm: Immunotherapy (steroids, IVIg, PLEX) targets the cause; ASMs alone are often insufficient; early treatment improves outcomes
Seronegative cases: Up to 40–50% of suspected autoimmune epilepsy is antibody-negative; empiric immunotherapy trial warranted if APE2 ≥4

Antibody Classification

Cell-Surface vs. Intracellular Antibodies

Feature	Cell-Surface Antibodies	Intracellular Antibodies
Targets	NMDA-R, LGI1, CASPR2, GABA-B, GABA-A, AMPA	GAD65, ANNA-1 (Hu), CV2/CRMP5, amphiphysin
Pathogenic mechanism	Directly pathogenic — receptor internalization, blockade, or complement-mediated damage	T-cell mediated cytotoxicity — antibodies are biomarkers, not direct effectors
Cancer association	Variable: NMDA-R (ovarian teratoma 20–50% in women); GABA-B (SCLC ~50%); AMPA (SCLC, breast, thymoma); LGI1 rarely paraneoplastic	Strongly paraneoplastic: ANNA-1 (SCLC >80%); CV2 (SCLC, thymoma); amphiphysin (breast, SCLC)
Immunotherapy response	Often excellent — >70% improve with first-line immunotherapy	Usually limited — neuronal damage is irreversible; tumor removal is primary treatment
Prognosis	Generally favorable with prompt immunotherapy	Guarded — depends on tumor status and extent of neuronal loss

Key Antibodies & Clinical Syndromes

Antibody	Seizure Prevalence	Hallmark Feature	Cancer Association
NMDA-R	70–80%	Stereotyped progression: psychiatric onset → seizures → movement disorder (orofacial dyskinesias) → autonomic instability → decreased consciousness/coma; extreme delta brush on EEG	Ovarian teratoma (20–50% in women); rare in men/children
LGI1	>90%	FBDS pathognomonic — <3 sec, up to 100/day, ASM-resistant, immunotherapy-responsive; hyponatremia 60–70%; may prevent limbic encephalitis if treated early	Rare (<5%); thymoma reported
CASPR2	30–50%	Morvan syndrome: insomnia + autonomic dysfunction + neuromyotonia + encephalopathy; older males predominate	Thymoma (20–40%)
GABA-B	>90%	Early refractory seizures, often temporal; status epilepticus common; older adults	SCLC (~50%)
GABA-A	>90%	Refractory status epilepticus; multifocal cortical FLAIR abnormalities on MRI; rapidly progressive encephalopathy	Thymoma (occasional); often non-paraneoplastic
AMPA	40–60%	Limbic encephalitis with relapsing course; memory impairment	SCLC, breast, thymoma (~70%)
GAD65	Variable	Only high titers ≥20 nmol/L meaningful; associated with stiff-person syndrome, cerebellar ataxia, drug-resistant TLE; poor immunotherapy response	Rarely paraneoplastic

Faciobrachial Dystonic Seizures (FBDS)

Virtually pathognomonic for LGI1 antibody encephalitis
Brief (<3 sec), very frequent (up to 100/day) dystonic contractions of face + ipsilateral arm
Often precede cognitive decline by weeks to months — critical early treatment window
ASM-resistant but respond rapidly to immunotherapy (corticosteroids, IVIg)
EEG may be normal in >50% — frequently misdiagnosed as myoclonus, tics, or PNES
Prompt immunotherapy may prevent progression to full limbic encephalitis

Post-HSV NMDA-R Encephalitis

NMDA-R antibodies develop 2–6 weeks after HSV encephalitis due to exposure of neuronal antigens during viral tissue destruction
Presents as clinical worsening weeks after initial HSV improvement
Always test for HSV before attributing relapse solely to autoimmune encephalitis
Requires immunotherapy, not additional antiviral treatment

APE2 Score & RITE2 Score

APE2 Score: Antibody Prevalence in Epilepsy and Encephalopathy

Clinical Feature	Points
New-onset seizure (within 1 year of evaluation)	+1
Neuropsychiatric changes (cognitive or behavioral)	+1
Autonomic dysfunction	+1
Viral prodrome	+2
Faciobrachial dystonic seizures	+3
Facial dyskinesias or orolingual dyskinesias	+2
Refractory status epilepticus (no prior epilepsy history)	+2
CSF findings (pleocytosis, elevated protein, or oligoclonal bands)	+2
Brain MRI suggesting autoimmune encephalitis (mesial temporal T2/FLAIR hyperintensity)	+2
History of autoimmune disease (personal)	+1
Total score ≥4: autoimmune evaluation recommended	—

APE2 Score Performance

APE2 ≥4 = 97% sensitivity, 14% specificity for positive neural-specific antibodies
Excellent screening tool with high negative predictive value
When APE2 ≥4 is combined with immunotherapy response: 78% sensitivity, 81.4% specificity, 88.1% PPV

RITE2 Score (Response to ImmunoTherapy in Epilepsy)

RITE2 ≥7 = 93% sensitivity, 60% specificity for predicting seizure response to immunotherapy
Incorporates APE2 components plus additional clinical and paraclinical features
Patients with APE2 ≥4 but negative antibodies are candidates for empiric immunotherapy trial

💎 Board Pearl

APE2 ≥4 = send autoimmune panel. FBDS alone scores +3 — add any one additional feature (new-onset seizure, neuropsychiatric change, autonomic dysfunction) and you meet the threshold

Clinical Clues Suggesting Autoimmune Etiology

When to Suspect Autoimmune Epilepsy

New-onset refractory seizures in a previously healthy person (especially young adult)
Faciobrachial dystonic seizures — virtually pathognomonic for LGI1
Limbic encephalitis features: subacute memory loss, psychiatric symptoms, MRI mesial temporal T2/FLAIR hyperintensity
Multifocal seizures without a clear structural cause
Rapid progression over days to weeks with drug-resistant seizures
Associated autoimmune disease (thyroid disease, type 1 diabetes, celiac disease)
Ovarian teratoma or other malignancy in setting of new neurological symptoms
Psychiatric symptoms disproportionate to seizure burden (psychosis, personality change)
Movement disorders (orofacial dyskinesias, chorea) accompanying seizures
Hyponatremia without other clear cause — think LGI1

Diagnostic Workup

Systematic Evaluation

Testing should be performed before immunotherapy whenever possible — treatment may reduce antibody titers and cause false negatives

Test	Details	Key Considerations
Serum antibody panel	NMDA-R, LGI1, CASPR2, GABA-B, GABA-A, AMPA, DPPX; GAD65 (quantitative); ANNA-1/Hu, CV2/CRMP5, amphiphysin	Cell-based assays preferred for surface antibodies
CSF antibody panel	Same as serum + oligoclonal bands, IgG index, cytology	Always send BOTH serum and CSF — NMDA-R may be negative in serum but positive in CSF in ~15%
CSF routine	Cell count, protein, glucose, cultures, HSV/VZV PCR	Lymphocytic pleocytosis (10–100 cells) common; must exclude infectious encephalitis
Brain MRI	Epilepsy protocol with coronal FLAIR through temporal lobes	Mesial temporal T2/FLAIR hyperintensity (LGI1, GABA-B); multifocal cortical lesions (GABA-A); may be normal in up to 50%
EEG	Continuous video-EEG if encephalopathic; routine EEG otherwise	Extreme delta brush (NMDA-R); temporal IEDs; subclinical seizures
Cancer screening	Whole-body CT or PET/CT	Mandatory in all patients; repeat at 6–12 months if initial negative with cancer-associated antibody
Sex-specific imaging	Pelvic MRI/ultrasound (women); testicular ultrasound (men)	Ovarian teratoma (NMDA-R in women); testicular germ cell tumor (men with NMDA-R)

Treatment

First-Line Immunotherapy

IV methylprednisolone: 1 g/day × 5 days, then oral prednisone taper over 3–6 months
IVIg: 0.4 g/kg/day × 5 days (total 2 g/kg); may repeat monthly
PLEX: 5–7 exchanges on alternate days; most effective for cell-surface antibodies; consider first-line in fulminant presentations
First-line agents are typically combined (steroids + IVIg or steroids + PLEX)

Second-Line Immunotherapy

Rituximab: 375 mg/m² weekly ×4 or 1000 mg ×2 (2 weeks apart); B-cell depletion within 2–4 weeks; preferred first second-line agent
Cyclophosphamide: IV pulse 750 mg/m² monthly × 3–6; reserved for rituximab-refractory cases; significant toxicity
Consider within 7 days of onset in refractory cases

Tumor Removal

Essential when paraneoplastic etiology identified — immunotherapy alone is often insufficient
Ovarian teratoma resection in NMDA-R encephalitis
Thymectomy for thymoma-associated CASPR2 or AMPA
SCLC treatment for GABA-B, AMPA, ANNA-1 syndromes

Long-Term Immunosuppression

Rituximab: every 6 months, guided by CD19/CD20 recovery or clinical relapse
Mycophenolate mofetil: 1000–1500 mg BID; onset 2–3 months; steroid-sparing
Azathioprine: 2–3 mg/kg/day; onset 3–6 months; check TPMT genotype
Continue immunotherapy ≥1–2 years; taper ASMs cautiously after sustained seizure freedom

Relapse Rates

NMDA-R: 12–20% relapse; higher if tumor not removed or no second-line therapy given
LGI1: 20–35% relapse; most during steroid taper; many require prolonged immunosuppression
GAD65: poor immunotherapy response overall; chronic ASMs usually required

Seronegative Autoimmune Epilepsy

Up to 40–50% of suspected autoimmune epilepsy cases are antibody-negative
Does NOT exclude the diagnosis — novel antibodies continue to be discovered
Empiric immunotherapy trial warranted if APE2 ≥4 and clinical suspicion high
Response to immunotherapy supports diagnosis of probable autoimmune epilepsy
Trial should be of adequate duration (at least 3–6 months) with escalation through first- and second-line agents

Diagnostic Classification

Definite: APE2 ≥4 + positive validated neural antibody
Probable: APE2 ≥4 + negative antibodies + clinical response to immunotherapy
Possible: APE2 ≥4 + negative antibodies + immunotherapy not yet attempted or response inconclusive

Board Pearls & Clinical Pearls

💎 Board Pearl

FBDS = LGI1 until proven otherwise — brief (<3 sec), frequent (up to 100/day), ASM-resistant, immunotherapy-responsive; pathognomonic for LGI1 antibody encephalitis
Extreme delta brush on EEG = NMDA-R encephalitis — rhythmic delta activity at 1–3 Hz with superimposed bursts of beta activity; seen in ~30% of NMDA-R cases
Hyponatremia + seizures + older adult = think LGI1 — hyponatremia occurs in 60–70% of LGI1 encephalitis due to SIADH
Post-HSV worsening at 2–6 weeks = NMDA-R antibodies — not HSV relapse; requires immunotherapy, not more acyclovir
Multifocal cortical FLAIR + status epilepticus = GABA-A — unique MRI pattern distinguishing GABA-A from other autoimmune encephalitides
Cell-surface = treatable, intracellular = search for tumor — the single most important distinction in autoimmune epilepsy is the antibody target location
GAD65 titers <20 nmol/L are NOT specific for autoimmune epilepsy — found in 5–8% of the general population (type 1 diabetes); only high titers (≥20 nmol/L) are meaningful

Clinical Pearl

Do not delay immunotherapy while awaiting antibody results in patients with rapidly progressive encephalopathy, refractory status epilepticus, or high clinical suspicion (APE2 ≥4). Early treatment is associated with significantly better seizure and cognitive outcomes, particularly for NMDA-R and LGI1 encephalitis. Obtain samples before treatment whenever possible, but treatment should not be withheld for pending results.

Clinical Pearl

FBDS as an early treatment window: Faciobrachial dystonic seizures typically precede cognitive decline and full limbic encephalitis by weeks to months. Recognizing FBDS and initiating immunotherapy at this stage may prevent progression to overt encephalitis with memory impairment — one of the few scenarios where a specific seizure semiology directly dictates immune-targeted treatment that can alter disease course.

Clinical Pearl

Paraneoplastic cancer screening should not stop after one negative scan. Repeat imaging at 6-month intervals for at least 2 years in patients with high-risk antibody profiles (ANNA-1, CV2, GABA-B). Tumors may be occult at initial presentation. PET/CT has higher sensitivity than CT alone for small tumors.

References

Dubey D, Pittock SJ, Kelly CR, et al. Autoimmune encephalitis epidemiology and a comparison to infectious encephalitis. Ann Neurol 2018;83(1):166–177.
Dubey D, Singh J, Britton JW, et al. Predictive models in the diagnosis and treatment of autoimmune epilepsy. Epilepsia 2017;58(7):1181–1189.
Dalmau J, Graus F. Antibody-mediated encephalitis. N Engl J Med 2018;378(9):840–851.
Graus F, Titulaer MJ, Balu R, et al. A clinical approach to diagnosis of autoimmune encephalitis. Lancet Neurol 2016;15(4):391–404.
Irani SR, Michell AW, Lang B, et al. Faciobrachial dystonic seizures precede LGI1 antibody limbic encephalitis. Ann Neurol 2011;69(5):892–900.
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.
Toledano M, Britton JW, McKeon A, et al. Utility of an immunotherapy trial in evaluating patients with presumed autoimmune epilepsy. Neurology 2014;82(18):1578–1586.
de Bruijn MAAM, Bastiaansen AEM, Thijs RD, et al. Antibody prevalence in epilepsy and encephalopathy score: increased immunotherapy utilization and improved outcome. Neurology 2022;98(11):e1167–e1178.
Bien CG, Holtkamp M. Autoimmune epilepsy: encephalitis with autoantibodies for neurologists. Epilepsy Curr 2017;17(3):134–141.
Scheffer IE, Berkovic S, Capovilla G, et al. ILAE classification of the epilepsies: position paper. Epilepsia 2017;58(4):512–521.
Finke C, Pruss H, Heine J, et al. Evaluation of cognitive deficits and structural hippocampal damage in encephalitis with leucine-rich, glioma-inactivated 1 antibodies. JAMA Neurol 2017;74(1):50–59.
Falip M, Rodriguez-Bel L, Casasnovas C, et al. Prevalence and immunological spectrum of temporal lobe epilepsy with glutamic acid decarboxylase antibodies. Eur J Neurol 2012;19(6):827–833.