ORIGINAL ARTICLE

Faciobrachial Dystonic Seizures PrecedeLgi1 Antibody Limbic Encephalitis

Sarosh R. Irani, DPhil,1 Andrew W. Michell, PhD,2 Bethan Lang, PhD,1 Philippa Pettingill, BSc,1

Patrick Waters, PhD,1 Michael R. Johnson, PhD,3 Jonathan M. Schott, MD,4

Richard J. E. Armstrong, PhD,1,4 Alessandro S. Zagami, MD,5 Andrew Bleasel, PhD,6

Ernest R. Somerville, FRCAP,5,7 Shelagh M. J. Smith, FRCP,8 and Angela Vincent, FRCPath1,9

Objective: To describe a distinctive seizure semiology that closely associates with voltage-gated potassium channel(VGKC)-complex/Lgi1 antibodies and commonly precedes the onset of limbic encephalitis (LE).Methods: Twenty-nine patients were identified by the authors (n ¼ 15) or referring clinicians (n ¼ 14). The temporalprogression of clinical features and serum sodium, brain magnetic resonance imaging (MRI), positron emissiontomography/single photon emission computed tomography, and VGKC-complex antibodies was studied.Results: Videos and still images showed a distinctive adult-onset, frequent, brief dystonic seizure semiology thatpredominantly affected the arm and ipsilateral face. We have termed these faciobrachial dystonic seizures (FBDS).All patients tested during their illness had antibodies to VGKC complexes; the specific antigenic target was Lgi1 in89%. Whereas 3 patients never developed LE, 20 of the remaining 26 (77%) experienced FBDS prior to thedevelopment of the amnesia and confusion that characterize LE. During the prodrome of FBDS alone, patients hadnormal sodium and brain MRIs, but electroencephalography demonstrated ictal epileptiform activity in 7 patients(24%). Following development of LE, the patients often developed other seizure semiologies, including typical mesialtemporal lobe seizures. At this stage, investigations commonly showed hyponatremia and MRI hippocampal high T2signal; functional brain imaging showed evidence of basal ganglia involvement in 5/8. Antiepileptic drugs (AEDs)were generally ineffective and in 41% were associated with cutaneous reactions that were often severe. By contrast,immunotherapies produced a clear, and often dramatic, reduction in FBDS frequency.Interpretation: Recognition of FBDS should prompt testing for VGKC-complex/Lgi1 antibodies. AEDs often produceadverse effects; treatment with immunotherapies may prevent the development of LE with its potential for cerebralatrophy and cognitive impairment.

ANN NEUROL 2010;000:000–000

Limbic encephalitis (LE) associated with antibodies (Abs)

to the voltage-gated potassium channels (VGKCs) is a

well-established immunotherapy-responsive condition, usu-

ally without an associated tumor.1–3 The Abs are now rec-

ognized to bind to different components of VGKC com-

plexes such as leucine-rich glioma inactivated-1 (Lgi1)3,4 or

contactin-associated protein 2 (Caspr2),3 and we will refer

to them as VGKC-complex Abs. Similar Abs have also

been identified in around 10% of unselected patients with

otherwise unexplained and drug-resistant epilepsy5,6 and in

4 patients with late onset epilepsy with an unusual semiol-

ogy.7,8 These 4 patients presented with very frequent, brief,

dystonic, seizure-like episodes.

Here, we describe 29 adult patients with this distinc-

tive antiepileptic drug (AED)-refractory, but immunother-

apy-responsive, seizure semiology, which we refer to as

faciobrachial dystonic seizures (FBDS). The majority of

patients subsequently developed a typical VGKC-complex

View this article online at wileyonlinelibrary.com. DOI: 10.1002/ana.22307

Received Sep 3, 2010, and in revised form Sep 27, 2010. Accepted for publication Oct 15, 2010.

Address correspondence to Dr Vincent, Nuffield Department of Clinical Neurosciences, West Wing, Level 6, John Radcliffe Hospital, Oxford, OX3 9DU,

United Kingdom. E-mail: angela.vincent@imm.ox.ac.uk

From the 1Department of Clinical Neurology, Oxford University, Oxford, UK; 2Department of Clinical Neurosciences, Addenbrooke’s Hospital, Cambridge

University, Cambridge, UK; 3Centre for Neurosciences, Imperial College, London, UK; 4Dementia Research Centre, National Hospital for Neurology and

Neurosurgery, Queen Square, London, UK; 5Institute of Neurological Sciences and Clinical School, Prince of Wales Hospital and University of New South

Wales, Sydney, Australia; 6Westmead Hospital, University of Sydney, Sydney, Australia; 7Comprehensive Epilepsy Service, Prince of Wales Hospital, Sydney,

Australia; 8Department of Clinical Neurophysiology, National Hospital for Neurology and Neurosurgery, Queen Square, London, UK; and 9Institute of

Neurology, University College London, London, UK.

Additional supporting information can be found in the online version of this article.

VC 2010 American Neurological Association 1

Ab-positive nonparaneoplastic LE. Because early treatment

of the LE syndrome may limit the duration or severity of the

illness and the degree of cognitive disability, recognition of

these seizures should offer a window of opportunity for early

treatment and possible prevention of permanent disability.

Patients and Methods

In total, 29 patients with FBDS are described. Referring neurol-

ogists communicated with the authors verbally, via questionnaires

(as part of Irani et al,3 a study of patients with VGKC-complex

Abs >400pM) and/or clinic letters. In addition, 26 of 29

patients/relatives subsequently underwent retrospective telephone/

personal interviews. Emphasis was placed on the nature of

the seizures, their timing in relation to development of LE, and

the treatment responses. Overall, 17 patients were seen by the

authors (n ¼ 6), referring neurologists (n ¼ 6), or general practi-

tioners (n ¼ 5) while FBDS alone were present. In another 6

patients, timing of the FBDS was obtained from the relatives. In

the remaining 6 patients, FBDS were recognized by the authors

(n ¼ 3) or the referring neurologists (n ¼ 3) during the LE.

The FBDS in 3 of these patients were described previ-

ously.7 Other patients were included in studies of LE23 or Mor-

van syndrome,10 but their FBDS were not described in those

publications. Videos were available for review in 16 of the

patients.

VGKC-complex Abs were determined by routine immu-

noprecipitation by patient serum immunoglobulin G (IgG) of

VGKC complexes from 125I-adendrotoxin–labeled rabbit brain

extract (normal range <100pM, as in Vincent et al1). Cell-

based assays for Lgi1, Caspr2, and contactin-2 Abs used human

embryonic kidney 293 cells transfected with cDNAs encoding

membrane-anchored Lgi1, Caspr2, or contactin-2, which were

incubated with sera (1:20-1:100). The cells were washed, fixed,

incubated with fluorescent secondary Abs, and visualized under

the fluorescence microscope (for methods see Irani et al3).

Results

A Distinctive FBDS Semiology before andduring Lgi1 Ab Limbic EncephalitisTwenty-nine patients were identified with FBDS either

during the prodromal stage or during LE. The age range

was from 36 to 83 years (median, 64 years), with a mal-

e:female ratio of around 2:1. The distinctive FBDS are

shown in 5 illustrative still images (Fig 1A–E) and

FIGURE 1: Faciobrachial dystonic seizures (FBDS). (A–E) Ictal stills of 5 patients included in this series are shown. Theipsilateral face grimacing and arm posturing are visible in all cases. Videos are available, with the patients’ consent, in theonline supplementary data. (F) Percentage of patients with dystonic posturing of specified body part(s) during FBDS (whitebars) and whether FBDS remained strictly unilateral or sometimes alternated in an individual patient (black bars).

ANNALS of Neurology

2 Volume 000, No. 000

corresponding videos (Supporting Information Videos A–

E). The FBDS were very brief (usually <3 seconds) and very

frequent, occurring at a median of 50� per day at their peak

(Table 1). FBDS always involved the arm and commonly

also the ipsilateral face (76%) and leg (34%). The trunk

was involved in 28% (see Fig 1F). Hand involvement

caused many patients to drop items within their grasp. In

69% of patients, FBDS could involve either side but were

always unilateral on any occasion. Twenty-four percent

produced vocalizations at the start of the FBDS. Ictal loss

of awareness was noted in a minority of seizures by either

the patient or carer/relative in 66% of patients. In 28% of

patients, the FBDS could be triggered by auditory stimuli

or high emotion. In addition, unexplained drop attacks,

often with backward falls, were reported by 62% of sub-

jects unrelated to leg ictal dystonia. Overnight recordings,

when performed, identified some FBDS during sleep. No

patient had a family history of similar episodes.

Fifteen of the 64 patients (23%) with LE and

VGKC-complex Abs in our recent study had FBDS.3

Investigations When FBDS Alone Were PresentDuring the period of FBDS without cognitive impair-

ment (see Table 1), both serum sodium levels (13 of 13

patients) and brain magnetic resonance imaging (MRI)

were normal (9 of 9 patients). The FBDS were predictive

of VGKC-complex Abs in each of the 4 patients with

FBDS who were identified by the authors and tested at

this stage; Lgi1 was identified as the specific antigenic

target within the VGKC complex in 3 of these. Electro-

encephalography (EEG) demonstrated ictal epileptiform

abnormalities in 7 patients with a frontotemporal (n ¼2, for example Fig 2), frontal (n ¼ 2), or temporal (n ¼3) focus. Two patients also developed temporal lobe sei-

zures as well as FBDS before the onset of the amnesia.

FBDS as a Clinically Identifiable Prodrometo Limbic EncephalitisThree patients (10%) never experienced FBDS. Twenty

of the other 26 (77%) patients developed FBDS before

the onset of amnesia and confusion, with a median lag

of 36 days (Table 2). There were no clinically identifiable

triggers for the apparent switch from a period of FBDS

alone to the development of (often dense) amnesia. In

the remaining 6 patients, the seizures were only recog-

nized during LE, after the onset of amnesia.

The majority of patients had FBDS that increased

in frequency until they reached a maximum at around the

time of onset of LE (shown later in Fig 3A–C). During

the period of isolated seizures, most patients were able to

perform daily activities independently. At the onset of

memory impairment/confusion, clear functional disability

was evident (see modified Rankin Scale, Fig 3C).

The LE was similar to that described previously (see

Table 2),1,3,11 with amnesia (100%), confusion (88%), hal-

lucinations (35%) and sleep disturbances (31%), including

rapid eye movement sleep behavior disorder, hypersomno-

lence, and insomnia. One case had persistent foot dystonia

(see Supporting Information Video D). Principally during

the period of LE, 18 of 26 (70%) patients also developed

other nondystonic seizure types, including generalized

tonic-clonic seizures (n ¼ 13), typical complex partial

TABLE 1: Clinical Features of All Patients withFBDS and Results of Investigation at Times WhenNo Cognitive Impairment Was Apparent

Characteristic Value

Clinical features of FBDS n ¼ 29

Age, median yr (range) 64 (36–83)

Sex, male:female 19:10

Number of faciobrachialseizures per day,median (range)

50 (6–360)

Alteration in consciousness 19 (66%)a

Falls 18 (62%)

Stimulus-triggered seizures 8 (28%); auditoryin 6, highemotion in 2

Ictal vocalizations 7 (24%)

Investigation results at time of FBDS withoutcognitive impairment

Serum sodium level>135mmol

13 of 13 (100%)

Normal brain MRI 9 of 9 (100%)b

VGKC-complex antibodies,mean pM (range)

1,962 (639–5,409)c

Lgi1 antibody positive 3 of 4 (75%)

Caspr2 antibody positive 0 of 4 (0%)

Contactin 2 antibody positive 0 of 4 (0%)

In 6 of 29 patients, FBDS were first recognized after onsetof amnesia/confusion.aAltered awareness was observed in 66% of patients but notseen during all seizures in any individual.bOne of these MRIs was later judged to have subtle bilateralhigh hippocampal signal by 2 neuroradiologists with an in-terest in limbic encephalitis.cVGKC-complex antibody testing was only requested on 4samples during the period of FBDS alone. Three of thesepatients were given immunotherapies and never developedamnesia; only 1 progressed to develop amnesia/confusion(this case is also included in Table 2).FBDS ¼ faciobrachial dystonic seizures; MRI ¼ magnetic res-onance imaging; VGKC ¼ voltage-gated potassium channel.

Irani et al: Lgi1 Ab Limbic Encephalitis

Month, 2010 3

medial temporal lobe seizures (n ¼ 12), and simple partial

seizures with piloerection (n ¼ 1), which have all previously

been reported in VGKC-complex Ab-associated LE.1,3,12,13

Investigations during or after the Periodof Limbic Encephalitis (n ¼ 26)All 24 patients with active LE had serum VGKC-com-

plex Abs by radioimmunoprecipitation (see Table 2 and

Supporting Information Fig). Lgi1 Abs were present in

22 (2 of these also had Abs against Caspr2, and 1 also

against contactin-2; for further information about these

antigens, see Irani et al3). We were unable to identify the

specific target in 2 VGKC-complex Ab-positive patients.

Two patients who were tested >4 years after the illness

were negative for VGKC-complex Abs, but 1 of these

was positive for Lgi1 Abs by the cell-based assay.

Interictal EEG abnormalities included diffuse mild

slowing (n ¼ 9), bilateral frontotemporal slowing (n ¼ 6), or

temporal sharp waves (n ¼ 2). In 9 patients, no EEG abnor-

malities were detected. At this stage in the illness, hyponatre-

mia (<135mmol) was seen in 23 (88%) patients.

Twenty-five patients had cerebrospinal fluid (CSF)

sampled. Analysis of white cells, protein, glucose, and

CSF unmatched oligoclonal bands were unremarkable

in 20 (80%). Five patients had a mild lymphocytosis (6–

11 cells/mm3), and/or modestly elevated protein (0.63–

1.1g/l). All patients were found to be negative for onco-

neural Abs by immunoblotting (ravo kit, ravo Diagnos-

tika, Freiburg, Germany).

Imaging during Limbic EncephalitisDuring the period of LE, 12 of 26 patients had normal brain

MRI. Of the other 14 patients, 10 showed bilateral (an

example shown in Fig 4A-C) and 3 showed unilateral

medial temporal lobe high T2 signal change, and 1 showed

high signal in the right caudate/putamen. Four patients, all

with initial bilateral high signal, developed hippocampal at-

rophy (after median follow-up of 8 months [range, 6–10

months], as illustrated in Fig 4C and D), and the individual

with caudate high signal had caudate atrophy at 3 months.

Eight patients underwent fluorodeoxyglucose (FDG)-

positron emission tomography (PET), and 2 underwent sin-

gle photon emission computed tomography (SPECT) (1 had

both investigations). Four of the PET/SPECTs were abnor-

mal at times when the MRI was normal. PET in 6 of 8

patients showed altered glucose metabolism in the temporal

region (increased in 4 and decreased in 2, Fig 4E and F).

More surprisingly, 5 of 8 patients showed abnormal glucose

metabolism in the basal ganglia (4 hypermetabolic and 1

hypometabolic, see Fig 4H-J for examples). The 2 patients

who underwent SPECT showed hypometabolic and hyper-

perfusion in the temporal lobe (hypoperfusion shown in

Fig 4G). Whole body computed tomography and/or PET

were unremarkable in the 28 patients tested. To date, no

FIGURE 2: Electroencephalography. (A) Interictal and (B) ictal electroencephalograms (EEGs) were recorded during afaciobrachial dystonic seizure in a 39-year-old woman. On this occasion, ictal 2 to 4Hz spike-wave activity was noted of maximalamplitude over the left frontotemporal region. The interictal EEG was normal. Ictal epileptiform activity was seen in 7 of 29 cases.Bar 5 1 second.

ANNALS of Neurology

4 Volume 000, No. 000

patient has developed a tumor after a median follow-up of

2.75 years (range, 0.5–8 years).

Response to Treatments in All 29 PatientsFor each patient, a mean of 2.6 AEDs were administered

(range, 1–6). The effect on FBDS frequency was generally

poor, and only 4 patients showed a good (20–50%) or

excellent (>50%) reduction in FBDS frequency within 1

month of treatment (see Fig 3D). Significant side effects

were seen in 12 patients (41%) (see Fig 3D), with a local-

ized rash in 8 patients (carbamazepine in 4, phenytoin in 2,

lamotrigine in 1, and levetiracetam in 1) and erythroderma

in 2 patients (both phenytoin). In addition, 2 patients

developed Stevens-Johnson syndrome after receiving carba-

mazepine; 1 required intensive care unit admission.

By contrast, the FBDS response to immunotherapy

was excellent (see Fig 3). Two patients were not offered

immunotherapies, because they were originally seen before

widespread knowledge of VGKC-complex Abs. They prog-

ressed from FBDS only to develop amnesia/confusion

with hyponatremia. With AEDs, but without immuno-

therapy, these 2 patients made some symptomatic recovery

TABLE 2: Timing of FBDS, TLS, and LE from the 26 Patients Who Developed LE

Characteristic Value

Timings n ¼ 26a (unless otherwise stated)

Patients with FBDS prior to amnesia, No. 20 (77%)

Days from FBDS onset to amnesia/confusion, median (range) 36 (�150 to 730)a

Additional seizure semiologies Total ¼18 (70%); generalized tonicclonic ¼ 13; complex partial ¼ 12(all MTLS); simple partial ¼ 1

Patients with onset of TLS after onset of FBDS, No. 10 of 12 (83%)

Days from onset of FBDS to onset of temporallobe seizures, median (range)

12.5 (�15 to 455)

Features associated with LE

Amnesia 26 (100%)

Confusion 23 (88%)

Hallucinations 9 (35%)

Sleep disturbance 7 (31%)

Depression 5 (19%)

Dysautonomia 4 (15%)

Pain 4 (15%)

Cerebellar features 2 (8%)

Interictal limb dystonia 1 (4%)

Investigations during LE n ¼ 26

VGKC-complex antibodies, mean pM (range) 2,281 (0–8,800)b

Lgi1antibody positive 23 (88%)b

Lgi1 and Caspr2 or contactin 2 antibodies 3 (12%)

No VGKC-complex–specific target determined 3 (12%)

Serum sodium <135mmol 23 (88%)

Normal brain MRIa 12 (46%)aThree other patients never developed amnesia or confusion and had normal MRI throughout their illness.bTwo patients tested >4 years after their illness had negative VGKC-complex antibodies as assessed by the routine radioimmuno-precipitation assay (Lgi1 antibodies were detected by cell-based assay in 1 of these patients); 1 patient administered intravenousimmunoglobulins 3 weeks before antibody testing had low positive VGKC-complex antibody levels (238pM).FBDS ¼ faciobrachial dystonic seizures; TLS ¼ temporal lobe seizures; LE¼ limbic encephalitis; MTLS ¼ mesial temporal lobe seizuresas defined clinically or by electroencephalography; VGKC ¼ voltage-gated potassium channel; MRI ¼ magnetic resonance imaging.

Irani et al: Lgi1 Ab Limbic Encephalitis

Month, 2010 5

at 4-year follow-up, although 1 case still experiences daily

FBDS. Of the other 27 patients, steroid therapy (oral or

intravenous) was used in 24 of 27, 14 received intravenous

immunoglobulin (IvIg), 13 were given plasma exchange,

and 1 received rituximab. Fourteen patients (52%) showed a

>50% reduction in seizure frequency, 12 (44%) showed

a 20 to 50% reduction, and only 1 case (4%) had

<20% reduction over the first month of administration (see

Fig 3D). Usually, there was an excellent correlation between

seizure reduction and fall in VGKC-complex Abs, often after

unsuccessful AED therapies (see Fig 3A–C), although this

was not universal. Adverse events with immunotherapies

were uncommon, although a fatal infection and a steroid-

induced psychosis occurred in 2 patients during their LE.

In 4 of 6 patients given IvIg before prednisolone, the IvIg

was considered ineffective. The 3 patients with FBDS who

did not develop amnesia were not administered immuno-

therapy earlier than those who did develop amnesia (data not

shown).

Relapses are not common in VGKC-complex

Ab-associated LE. However, 4 of these patients relapsed

after weaning prednisolone from 50 to 60mg down to

30mg over 3 months (n ¼ 3) or after IvIg only (n ¼ 1).

The FBDS were highly responsive to reinstitution of

prednisolone (example in Fig 3B).

Discussion

The distinctive adult-onset, high-frequency, very brief,

AED-resistant FBDS described in this international

cohort of patients have a very unusual seizure semiology,

but appear to be closely associated with raised serum

VGKC-complex Abs, almost always of the Lgi1 subtype,

and a high probability of development of nonparaneoplas-

tic LE. Recognition of these brief seizures, and their associ-

ation with VGKC-complex/Lgi1 Abs, should prompt con-

sideration of immunotherapies and may prevent the onset

and long-term consequences of limbic encephalitis.

The semiology is distinctive and was predictive of

VGKC-complex Abs in the 4 patients recently seen

by the authors, 3 of whom were treated and did not

develop LE. However, most of the patients were identified

FIGURE 3: Clinical and serological responses to immunotherapies. (A–C) voltage-gated potassium channel (VGKC)-complexantibody (Ab) titers, seizures per day, and modified Rankin Scores (mRS, green in C) are shown in individual cases. The purpleasterisk indicates the onset of amnesia/confusion. Pred 5 prednisolone; IvIg 5 intravenous immunoglobulins; Px 5 plasmaexchange; Aza 5 azathioprine; AEDs 5 antiepileptic drugs. (D) The responses of the patients to AEDs (n 5 29) and/orimmunotherapies (ITs, n 5 27) and the adverse events attributed to these medications (blue bar) are illustrated.

ANNALS of Neurology

6 Volume 000, No. 000

retrospectively from those who had already developed LE

with high levels of VGKC Abs (>400pM), and it is not

yet clear whether FBDS are 100% predictive of VGKC

Abs and whether, if untreated, FBDS inevitably proceed

to LE. One patient sampled after recovery was negative

for VGKC Abs but positive for Lgi1 Abs; this indicates

that not all Lgi1 is associated with VGKCs and that Abs to

Lgi1, and by inference to other associated proteins such as

Caspr2, may be found in patients who are not positive for

VGKC-complex Abs by the immunoprecipitation assay;

but that also requires further study. For the time being, we

continue to screen sera initially with the radioimmunopre-

cipitation assay for VGKC-complex Abs, but assays for

Lgi1 and Caspr2 Abs are being developed for wider use.

The description here significantly extends the origi-

nal report of 3 patients with these seizurelike episodes,7

and argues against a paroxysmal movement disorder. The

epileptic nature of FBDS is supported by the associated

brief loss of awareness and the highly stereotyped semiol-

ogy, although a coincident epileptiform change on scalp

EEG was only demonstrated in the minority. The ictal

onset zone has not been determined, but the high fre-

quency of brief attacks with a startle component and

trunk turning are suggestive of frontal lobe involve-

ment.14 However, ictal head version and predominant

nocturnal episodes, also typical of frontal lobe seizures,

were not apparent. Furthermore, typical temporal lobe

seizures have a longer duration, occur less frequently, and

commonly show a sustained (often >10 seconds) contra-

lateral arm dystonia with automatisms and absence of

face involvement.15,16 Irrespective of the location of the

ictal onset zone, it seems likely that the ictal dystonia in

these patients reflects basal ganglia involvement. This is

supported by the surprisingly frequent demonstration of

abnormal basal ganglia metabolism on FDG-PET, which

has previously been shown in typical temporal lobe seiz-

ures associated with upper limb dystonia.17,18 The few

previous VGKC-complex Ab LE studies using PET have

not detected basal ganglia abnormalities.12,19,20 Thus,

this may be a relatively specific feature within the subset

of patients with FBDS and may be related to their ictal

dystonia.

These seizures are likely to have been seen by

others. Descriptions of very frequent brief twitches affect-

ing the shoulder and ipsilateral face appear in recent case

reports of patients with seizures and VGKC-complex

Abs.8,10,21,22 Moreover, given that around 20% of our 64

patients with VGKC-complex Ab LE3 had FBDS, it is

interesting that 2 surveys of VGKC Ab-positive patients

noted a 22% and 40% frequency of myoclonus.4,20

It is now clear that VGKC-complex Abs, at least in

patients with high titers,3 are rarely directed against Kv1-

potassium channels themselves, but bind to other pro-

teins that are tightly complexed with Kv1 subunits in

situ and in detergent extracts.3,4 Almost all patients

sampled during their illness, before immunotherapies

began, had Abs directed against Lgi1, which is a secreted

protein that modulates neuronal morphology and

FIGURE 4: Examples of brain imaging in faciobrachial dystonic seizures (FBDS). (A–C) Typical medial temporal lobe high signalfrom magnetic resonance imaging (MRI) is shown during periods of amnesia in 3 patients. Hippocampal atrophy was seen insome cases at follow-up MRI (eg, change from C to D). Fluorodeoxyglucose-positron emission tomography (PET) and singlephoton emission computed tomography (SPECT) images in 7 patients showed basal ganglia and temporal lobe abnormalities,which included (E) temporal lobe PET bilateral hypermetabolism, (F) right hypometabolism, or (G) SPECT left hypoperfusion.Three examples of metabolic changes within the basal ganglia are shown in H and I (bilateral hypermetabolism) and J (lefthypermetabolism). These images were acquired during the limbic encephalitis phase, when the patients were also having manybrief FBDS per hour, and it was difficult to identify whether they were directly related to the FBDS.

Irani et al: Lgi1 Ab Limbic Encephalitis

Month, 2010 7

synaptic excitability.23 Abs to Lgi1 are likely to induce sim-

ilar complex effects depending on where and how they act

and will require detailed investigations. One study has

found epileptiform activity in mouse hippocampal slices

exposed to purified IgG from a patient with LE and Lgi1

Abs, likely as a result of the Abs enhancing neurotransmit-

ter release at mossy fiber/CA3 pyramidal cell synapses, and

consistent with a reduction of VGKC expression or func-

tion.24 Of relevance, Lgi1 mutant mice show increased sei-

zure susceptibility, with prominent dystonic, myoclonic,

and generalized seizures,25–27 and Lgi1 mutations have

been discovered in humans with autosomal dominant lat-

eral temporal lobe epilepsy.28–30 However, by contrast to

the patients we describe, patients with Lgi1 mutations of-

ten have an excellent response to AEDs and infrequent

seizures, although 1 family has been described with a high

frequency of drug-resistant seizures.31

Recognition of the very brief adult-onset, frequent

faciobrachial dystonic seizures should prompt testing for

VGKC-complex or Lgi1 Abs. Our data suggest that this

syndrome is unlikely to be associated with an underlying

tumor, although this should still be excluded, and that

caution should be used with trials of multiple AEDs,

which are frequently ineffective and associated with an

unusually high risk of adverse reactions. As, when pres-

ent, the characteristic FBDS often predate the onset of

amnesia/confusion, their recognition may provide a ther-

apeutic window of opportunity during which time

immunotherapies may prevent the LE and possibly the

potential sequelae of cerebral atrophy and cognitive

impairment.1,3,11,32

Acknowledgments

S.R.I. was supported by the National Institute for Health

Research (NIHR), Department of Health, UK. A. W. M.

receives support from the Cambridge NIHR Biomedical

Research Centre. J.M.S. is a UK Higher Education

Funding Council for England Clinical Senior Lecturer.

Some of this work was undertaken at University College

London Hospitals/University College London, which

received a proportion of funding from the Department

of Health’s NIHR Biomedical Research Centres funding

scheme. B.L. receives funding from Epilepsy Research

UK. P.P. is supported by a Medical Research Council

Clinician Scientist Fellowship to Dr C. Buckley. P.W.

and A.V. receive support from the Oxford NIHR Bio-

medical Research Centre. The Dementia Research Centre

is an Alzheimer’s Research Trust Coordinating Centre.

We thank Medical Illustration (John Radcliffe Hos-

pital) for help with the video editing, and particularly the

patients, their relatives/carers, and the following neuro-

logists for the clinical details: Drs M. Bogdanovic, D.

Hilton-Jones, K. Nithi, J. Palace, T. Andrews, O. C.

Cockerell, C. Everett, M. Husain, P. Jarman, M. Lunn,

P. Rudge, P. Maddison, J. Bowen, M. Lawden, J. Stone,

N. Lawn, K. Fuller, and W. Huynh.

Authorship

S.R.I. and A.W.M. are first authors. E.R.S., S.M.J.S., and

A.V. are senior authors.

Potential Conflicts of Interest

A.V. and the Department of Clinical Neurology in Oxford

receive royalties and payments for Ab assays. A.V. is the

inventor on patent application WO/2010/046716 entitled

‘‘Neurological Autoimmune Disorders.’’ The patent has

been licensed to Euroimmun AG for the development of

assays for Lgi1 and other VGKC-complex Abs. A.V. has

done paid consultancy for Athena Diagnostics, and is

employed by Oxford University and University College

London. A.V.B.L. has a grant pending from Wellcome

Trust and received payment for a lecture from Euro-

immune. A.V., B.L., P.W., and S.R.I. may receive royalties

for testing of VGKC complex Abs.

References1. Vincent A, Buckley C, Schott JM, et al. Potassium channel antibody-

associated encephalopathy: a potentially immunotherapy-responsiveform of limbic encephalitis. Brain 2004;127:701–712.

2. Thieben MJ, Lennon VA, Boeve BF, et al. Potentially reversibleautoimmune limbic encephalitis with neuronal potassium channelantibody. Neurology 2004;62:1177–1182.

3. Irani SR, Alexander S, Waters P, et al. Antibodies to Kv1 potas-sium channel-complex proteins leucine-rich, glioma inactivated 1protein and contactin-associated protein-2 in limbic encephalitis,Morvan’s syndrome and acquired neuromyotonia. Brain 2010;133:2734–2748.

4. Lai M, Huijbers MG, Lancaster E, et al. Investigation of LGI1 asthe antigen in limbic encephalitis previously attributed to potas-sium channels: a case series. Lancet Neurol 2010;9:776–785.

5. McKnight K, Jiang Y, Hart Y, et al. Serum antibodies in epilepsyand seizure-associated disorders. Neurology 2005;65:1730–1736.

6. Majoie HJ, de Baets M, Renier W, et al. Antibodies to voltage-gated potassium and calcium channels in epilepsy. Epilepsy Res2006;71:135–141.

7. Irani SR, Buckley C, Vincent A, et al. Immunotherapy-responsiveseizure-like episodes with potassium channel antibodies. Neurol-ogy 2008;71:1647–1648.

8. Barajas RF, Collins DE, Cha S, Geschwind MD. Adult-onset drug-refractory seizure disorder associated with anti-voltage-gatedpotassium-channel antibody. Epilepsia 2009;51:473–477.

9. Reid JM, Foley P, Willison HJ. Voltage-gated potassium channel-associated limbic encephalitis in the West of Scotland: casereports and literature review. Scot Med J 2009;54:27–31.

10. Toosy AT, Burbridge SE, Pitkanen M, et al. Functional imagingcorrelates of fronto-temporal dysfunction in Morvan’s syndrome.J Neurol Neurosurg Psychiatry 2008;79:734–735.

ANNALS of Neurology

8 Volume 000, No. 000

11. Graus F, Saiz A, Lai M, et al. Neuronal surface antigen antibodiesin limbic encephalitis: clinical-immunologic associations. Neurol-ogy 2008;71:930–936.

12. Wieser S, Kelemen A, Barsi P, et al. Pilomotor seizures and statusin non-paraneoplastic limbic encephalitis. Epileptic Disord 2005;7:205–211.

13. Malter MP, Helmstaedter C, Urbach H, et al. Antibodies to glu-tamic acid decarboxylase define a form of limbic encephalitis. AnnNeurol 2010;67:470–478.

14. Manford M, Fish DR, Shorvon SD. An analysis of clinical seizurepatterns and their localizing value in frontal and temporal lobeepilepsies. Brain 1996;119:17–40.

15. Kotagal P, Luders H, Morris HH, et al. Dystonic posturing in com-plex partial seizures of temporal lobe onset: a new lateralizingsign. Neurology 1989;39:196–201.

16. Bleasel A, Kotagal P, Kankirawatana P, Rybicki L. Lateralizing valueand semiology of ictal limb posturing and version in temporallobe and extratemporal epilepsy. Epilepsia 1997;38:168–174.

17. Newton MR, Berkovic SF, Austin MC, et al. Dystonia, clinicallateralization, and regional blood flow changes in temporal lobeseizures. Neurology 1992;42:371–377.

18. Dupont S, Semah F, Baulac M, Samson Y. The underlying patho-physiology of ictal dystonia in temporal lobe epilepsy: an FDG-PET study. Neurology 1998;51:1289–1292.

19. Fauser S, Talazko J, Wagner K, et al. FDG-PET and MRI inpotassium channel antibody-associated non-paraneoplastic limbicencephalitis: correlation with clinical course and neuropsychology.Acta Neurol Scand 2005;111:338–343.

20. Tan KM, Lennon VA, Klein CJ, et al. Clinical spectrum of voltage-gated potassium channel autoimmunity. Neurology 2008;70:1883–1890.

21. Park DC, Murman DL, Perry KD, Bruch LA. An autopsy case of lim-bic encephalitis with voltage-gated potassium channel antibodies.Eur J Neurol 2007;14:e5–e6.

22. Geschwind MD, Tan KM, Lennon VA, et al. Voltage-gated potas-sium channel autoimmunity mimicking Creutzfeldt-Jakob disease.Arch Neurol 2008;65:1341–1346.

23. Zhou YD, Lee S, Jin Z, et al. Arrested maturation of excitatory syn-apses in autosomal dominant lateral temporal lobe epilepsy. NatMed 2009;15:1208–1214.

24. Lalic T, Pettingill P, Vincent A, Capogna M. Human limbic enceph-alitis serum enhances hippocampal mossy fiber-CA3 pyramidalcell synaptic transmission. Epilepsia (in press).

25. Chabrol E, Navarro V, Provenzano G, et al. Electroclinical charac-terization of epileptic seizures in leucine-rich, glioma-inactivated1-deficient mice. Brain 2010;133:2749–2762.

26. Fukata Y, Lovero KL, Iwanaga T, et al. Disruption of LGI1-linkedsynaptic complex causes abnormal synaptic transmission and epi-lepsy. Proc Natl Acad Sci U S A 2010;107:3799–3804.

27. Yu YE, Wen L, Silva J, et al. Lgi1 null mutant mice exhibit myo-clonic seizures and CA1 neuronal hyperexcitability. Hum MolGenet 2010;19:1702–1711.

28. Morante-Redolat JM, Gorostidi-Pagola A, Piquer-Sirerol S, et al. Muta-tions in the LGI1/Epitempin gene on 10q24 cause autosomal dominantlateral temporal epilepsy. HumMol Genet 2002;11:1119–1128.

29. Gu W, Brodtkorb E, Steinlein OK. LGI1 is mutated in familial tem-poral lobe epilepsy characterized by aphasic seizures. Ann Neurol2002;52:364–367.

30. Kalachikov S, Evgrafov O, Ross B, et al. Mutations in LGI1 causeautosomal-dominant partial epilepsy with auditory features. NatGenet 2002;30:335–341.

31. Di Bonaventura C, Carni M, Diani E, et al. Drug resistant ADLTEand recurrent partial status epilepticus with dysphasic features in afamily with a novel LGI1 mutation: electroclinical, genetic, andEEG/fMRI findings. Epilepsia 2009;50:2481–2486.

32. Bien CG, Urbach H, Schramm J, et al. Limbic encephalitis as aprecipitating event in adult-onset temporal lobe epilepsy. Neurol-ogy 2007;69:1236–1244.

Irani et al: Lgi1 Ab Limbic Encephalitis

Month, 2010 9