E

Nwd

KAA

a

b

c

Ed

RA

i(a

0h

pilepsy Research (2013) 105, 86—91

j ourna l h om epa ge: www.elsev ier .com/ locate /ep i lepsyres

ewly-diagnosed pediatric epilepsy is associatedith elevated autoantibodies to glutamic acidecarboxylase but not cardiolipin

adi Veria,1, Oivi Uibob,1, Tiina Talvikb,1, Inga Talvikb,1, Kaja Metskülac,2,ita Napad,1, Ulvi Vaherd,1, Eve Õiglane-Slikb,1, Reet Reind,1,nneli Kolkb,1, Aili Traatd,1, Raivo Uiboc,∗

Department of Pediatrics, University of Tartu, Lunini 6, Tartu 51014, EstoniaDepartment of Pediatrics, University of Tartu, Children’s Clinic of Tartu University Hospital, Lunini 6, Tartu 51014, EstoniaImmunology Group, Institute of Biomedicine and Centre for Translational Medicine, University of Tartu, Ravila 19, Tartu 50411,stoniaChildren’s Clinic of Tartu University Hospital, Lunini 6, Tartu 51014, Estonia

eceived 26 July 2012; received in revised form 4 February 2013; accepted 12 February 2013vailable online 25 March 2013

KEYWORDSEpilepsy;Autoantibodies;Glutamic aciddecarboxylase;Anti-cardiolipinautoantibodies;Children

Summary Glutamic acid decarboxylase autoantibodies (GADA) and anti-cardiolipin autoan-tibodies (ACA) have been detected in adult subjects with epilepsy, though the functionalimplications of these findings are a matter of debate. This study aimed to determine the preva-lence of GADA and ACA and to investigate their clinical significance in pediatric subjects withnewly-diagnosed epilepsy. For this purpose GADA and ACA were assessed by enzyme-linkedimmunosorbent assays in 208 pediatric patients with newly-diagnosed epilepsy and 128 controls.The clinical data (results of electroencephalography, magnetic resonance imaging, 6-month out-come etc.) was compared to antibody test results. Our study revealed GADA in 14 (6.7%) patientswith epilepsy and in 1 (0.8%) control, which was a statistically significant difference (P = 0.010;Chi-square test). The GADA-positive and -negative patients had similar clinical characteristics.

The prevalence of ACA in patients with epilepsy (6.3%) was not significantly different thancontrols (2.6%). These results snewly-diagnosed pediatric patiesignificance and pathogenic role© 2013 Elsevier B.V. All rights re

∗ Corresponding author. Fax: +372 7374232.E-mail addresses: kadi.veri@kliinikum.ee (K. Veri), oivi.uibo@

nga.talvik@kliinikum.ee (I. Talvik), kaja.metskyla@kliinikum.ee (K. MetU. Vaher), eve.oiglane-slik@kliinikum.ee (E. Õiglane-Slik), reet.reiili.traat@kliinikum.ee (A. Traat), raivo.uibo@ut.ee (R. Uibo).1 Fax: +372 7319608.2 Fax: +372 7374232.

920-1211/$ — see front matter © 2013 Elsevier B.V. All rights reserved.ttp://dx.doi.org/10.1016/j.eplepsyres.2013.02.009

uggest that GADA is associated with epilepsy in a subgroup of

nts. Further studies are required to determine the prognostic

of GADA among pediatric subjects with epilepsy.served.

kliinikum.ee (O. Uibo), tiina.talvik@kliinikum.ee (T. Talvik),sküla), aita.napa@kliinikum.ee (A. Napa), ulvi.vaher@kliinikum.een@kliinikum.ee (R. Rein), anneli.kolk@kliinikum.ee (A. Kolk),

87

Table 1 Clinical data from 208 children with epilepsy.

General features No. (%)

SexMales 109 (52.4)Females 99 (47.6)

Mean age (years) 7.8Coexisting diabetes 1 (0.5)Epilepsy type

Idiopathic 113 (54.3)Focal 69 (33.2)Generalized 44 (21.1)

Cryptogenic 25 (12.0)Symptomatic 34 (16.3)

Focal 32 (15.4)

yfnCana

aasLuAts(eait

sb

a

R

Gadco12p

Pediatric epilepsy associated with elevated autoantibodies

Introduction

Recent studies suggest an increased prevalence of vari-ous autoantibodies in subjects with epilepsy (Irani et al.,2011). Specifically, glutamic acid decarboxylase autoan-tibodies (GADA), anti-cardiolipin autoantibodies (ACA) orautoantibodies to other phospholipids, glutamate receptors,and voltage-gated potassium channels have been reportedto be associated with epilepsy. Among adult subjects withepilepsy, GADA and ACA have been shown to be the mostcommon and have therefore attracted considerable atten-tion for immunological characterization of the disease(Liimatainen et al., 2010; Falip et al., 2012).

Antibodies to glutamic acid decarboxylase (GAD), therate-limiting enzyme of gamma-aminobutyric acid (GABA)synthesis, have been largely studied in type 1 diabetes mel-litus (Baekkeskov et al., 1990) and in several neurologicalconditions, including stiff-person syndrome (Solimena et al.,1990; Vianello et al., 2005), cerebellar ataxia (Vianelloet al., 2003, 2005), palatal myoclonus (Nemni et al., 1994;Vianello et al., 2005) and limbic encephalitis (Matà et al.,2008). There are some reports of increased GADA levelsin idiopathic generalized epilepsies (Aykutlu et al., 2005;Striano et al., 2008), but GAD autoimmunity has mostly beenassociated with localization-related epilepsies, in particularwith temporal lobe epilepsy of unknown etiology (Giomettoet al., 1998; Errichiello et al., 2009; Liimatainen et al., 2010)and drug-resistant epilepsy (Peltola et al., 2000; McKnightet al., 2005; Yoshimoto et al., 2005; Vulliemoz et al., 2007).The clinical significance of elevated GADA levels in sub-jects with epilepsy is still widely debated (Bien and Scheffer,2011).

Although ACA are a common marker for rheumatic dis-eases, they have also been detected in subjects withepilepsy (Verrotti et al., 2003; McKnight et al., 2005).Increased prevalence of ACA in subjects with partial epilepsyhas been associated with a long duration of disease and poorseizure control (Ranua et al., 2004).

The aim of the current study was to determine the preva-lence of GADA and ACA, and to elucidate their clinicalsignificance among pediatric subjects with newly-diagnosedepilepsy.

Methods

This prospective study comprised 208 pediatric patients(mean age of 7.8 years; range from 1 month to 19 years),including 109 males, who were admitted to the Children’sClinic of Tartu University Hospital, one of two tertiary pedi-atric care centers of Estonia, between January of 2009and April of 2011. All these patients had newly-diagnosedepilepsy, which was classified by a neurologist and a neu-rophysiologist according to the recommendations of theInternational League Against Epilepsy (ILAE, 1989) andconfirmed. Neonatal seizures and cases with only febrileseizures were excluded. All subjects underwent an awakeand sleep-deprived video-electroencephalography (EEG),

and brain magnetic resonance imaging (MRI). The effec-tiveness of antiepileptic treatment was evaluated 6 monthslater. Clinical characteristics of the study subjects are pre-sented in Table 1.

4s

l

Generalized 2 (0.9)Unclassified 36 (17.3)

The control group [n = 128 (64 males), mean age 9.5ears; range from 2 to 18 years] included subjects withunctional urinary (enuresis) and gastrointestinal (abdomi-al pain, constipation) disorders admitted to the Children’slinic of Tartu University Hospital. A complete blood countnd C-reactive protein levels were used to exclude acute ill-ess. Clinical records were reviewed to exclude coexistingutoimmune and neurological disorders.

In all children venous blood samples for autoantibodynalyses were obtained. Serum was divided into aliquotsnd stored at −20 ◦C. GADA (GAD65 isoform) were mea-ured using enzyme-linked immunosorbent assay (ELISA, RSRtd., UK) where GAD isoform of 65 kD served as antigen. Val-es ≥5 U/ml were considered positive for GADA presence.ccording to the 2009 Diabetes Autoantibody Standardiza-ion Program (DASP), this method has high sensitivity andpecificity for GADA detection. ACA were detected by ELISAEuroimmun AG, Germany) with values ≥12 RU/ml consid-red positive. This assay detects IgM, IgG and IgA antibodiesgainst cardiolipin and was performed under UKNEQAS qual-ty assurance control. In patients with epilepsy blood wasaken before the antiepileptic treatment.

Written informed consent was obtained from all studiedubjects and/or their parents, and the study was approvedy the Research Ethics Committee of the University of Tartu.

Statistical analysis was performed using Chi-square testnd Fisher’s exact test.

esults

ADA were detected in 14 (6.7%) patients with epilepsynd in 1 (0.8%) control subject with statistically significantifference (Chi-square test P = 0.010). The GADA-positiveontrol subject (antibody level of 6 U/ml) was a 12-year-ld male with enuresis but no other signs of disease. Of the4 GADA-positive patients, 5 had focal idiopathic epilepsy,

had focal symptomatic epilepsy, 2 had generalized idio-athic epilepsy, 1 had generalized symptomatic epilepsy and

had unclassified epilepsy. The GADA-positive and -negativeubjects showed similar clinical characteristics.

Most patients with epilepsy (n = 11) displayed a low GADAevel (5—38 U/ml), but three had GADA values >50 U/ml,

8

atmamht(aa

p(t7c

(canGr(

D

Irodapuac

basoaasn2jsrMslG1tnam

ao

t1t(edw

tsjspmwoet

e(mGsiaaneef

etdpf

irngsaGrtrcmost

erws

8

nd their clinical data are presented in Table 2. Of thesehree patients, one had a history of coexisting autoim-une disease: a 9-year-old female (GADA > 2000 U/ml) with

diagnosis of focal symptomatic epilepsy after surgicalanagement of hypophyseal teratoma with concomitant

ypopituitarism and diabetes of unknown etiology. The otherwo patients with GADA > 50 U/ml included a 4-year-old male350 U/ml) with benign epilepsy with centrotemporal spikesnd a 15-year-old male (70 U/ml) with a diagnosis of gener-lized idiopathic epilepsy.

ACA were detected at a range of 12—120 RU/ml beingresent in 13 (6.3%) patients with epilepsy and in 32.6%) control subjects (non-significant difference). Two ofhe ACA-positive patients also had elevated GADA (5 and0 U/ml). ACA-positive and -negative patients had similarlinical characteristics.

Treatment response data were obtained from 156 of 20875.0%) patients after 6 months. Overall, seizures wereontrolled in 116 (74.8%) subjects, while 22 (14.2%) had

decrease in seizure frequency and 17 (11.0%) showedo significant response to antiepileptic treatment. In theADA-negative group, 16 of 142 (11.3%) patients showed no

esponse to treatment, while in the GADA-positive group one7.1%) patient did not respond (non-significant difference).

iscussion

n the last few years, a number of studies and caseeports have been published demonstrating the presencef GADA and ACA in subjects with different epileptic syn-romes. However, these studies were mostly performed indult subjects or in children with long-lasting epilepsy. Theresent study investigated a representative group of consec-tive, unselected children with newly-diagnosed epilepsy in

well-defined, ethnically-homogenous population using aase—control design.

The prevalence of GADA in the general population isetween 0.4% and 1.0%. A previous study by our group found

prevalence of 0.5% (Uibo et al., 2001), and the currenttudy found a prevalence of 0.8%. An increased prevalencef GADA has been reported in adults with refractory epilepsynd/or temporal lobe epilepsy. Higher GADA titers havelso been observed in other neurological conditions, such astiff-person syndrome, limbic encephalitis and other para-eoplastic syndromes (Solimena et al., 1990; Matà et al.,008). Peltola et al. (2000) found GADA in 8 of 51 sub-ects (15.7%) with refractory localization-related epilepsy; 2ubjects with high levels of GADA had long-standing therapy-esistant temporal lobe epilepsy, but did not have diabetes.cKnight et al. (2005) reported elevated GADA in 5 of 139

ubjects (3.6%); all had had drug-resistant focal epilepsyasting at least 3 years. Liimatainen et al. (2010) detectedADA in 15 of 253 subjects with epilepsy (5.9%, aged from6 to 76 years); seven (2.8%) had high GADA titers, and six ofhese had temporal lobe epilepsy. Kwan et al. (2000) foundo difference in GADA levels between seizure-free subjectsnd subjects with refractory epilepsy. Thus, evidence from

ultiple studies suggests that GADA are elevated in epilepsy.There is little data about the prevalence of GADA in pedi-

tric epilepsy. Errichiello et al. (2009) reported GADA in 6f 233 (2.6%) epileptic subjects (aged from 6 to 78 years),

eoms

K. Veri et al.

wo of whom had idiopathic generalized epilepsy and type diabetes mellitus, while the other four had cryptogenicemporal epilepsy with no history of diabetes. Verrotti et al.2003) studied 74 pediatric subjects with different types ofpilepsy and found no significant difference between chil-ren with epilepsy and controls, or between subjects whoere seizure-free and those with uncontrolled epilepsy.

We observed GADA in 6.7% of subjects across differentypes of newly-diagnosed childhood epilepsy. GADA-positiveubjects could not be distinguished from GADA-negative sub-ects by clinical epilepsy characteristics or EEG findings. Ourtudy did not show an association between GADA and tem-oral lobe epilepsy. Importantly, the effect of antiepilepticedication in this study was excluded, as all blood samplesere obtained prior to initiating treatment. On the basisf our 6-month follow-up, we conclude that initial pres-nce of GADA did not influence the efficacy of antiepilepticreatment.

In type 1 diabetes mellitus, an immune response to GADlucidates disease diagnosis, prognosis and pathogenesisUibo and Lernmark, 2008). Pertinent to our findings, GADAay have important effects on neural tissue. Thus, whereasAD regulates GABA production in vitro, GADA may have aignificant role through the entire GABAergic system, result-ng in unbalanced neurotransmission (Ali et al., 2011). It islso assumed that GADA could mimic the structure of GABAnd cross-react with GADA binding sites accessible on theeural surface (Vianello et al., 2002). In addition, the intrat-chal synthesis of GADA, demonstrated in GADA-positivepileptic patients (Saiz et al., 2008), may play a pivotal roleor GADA pathogenicity inside of the central nervous system.

Therefore, testing for GADA could identify subgroups ofpilepsy patients who may have differential responses toreatment (Irani et al., 2011). In our study, however, noifferences in epilepsy treatment efficacy between GADAositive and negative groups were found during 6 monthsollow-up period.

It may not seem plausible that autoantibodies couldnteract with intracellular enzymes like GAD. However, aecent study demonstrated that IgG antibodies can entereurons enabling the antibodies to bind intracellular anti-ens endocytosis (Hill et al., 2009). In addition, bothensitization of T cells against GAD and T cell reactionsgainst neural tissue compartments occur in parallel withADA development (Burton et al., 2010). Furthermore, the

oles of GADA in cerebrospinal fluid, as well as of GADA epi-ope specificity, are not fully understood, but could play aole. Notwithstanding the mechanism, the presence of GADAould be a valuable indicator of a specific underlying autoim-une process. However, in this connection the significance

f low titers of GADA (revealed in most of our patients) needspecial attention, both in association with epilepsy and withype 1 diabetes.

Notably, data on the presence of ACA in subjects withpilepsy are more contradictory. Verrotti et al. (2003)eported a significantly higher prevalence of ACA in childrenith epilepsy than in controls, but no difference between

ubjects who were seizure-free and those with uncontrolled

pilepsy. Ranua et al. (2004) reported that the prevalencef ACA was similar in a large epilepsy cohort to that ofatched controls, but that ACA had a higher prevalence in

ubjects with poor seizure control and a long duration of

Pediatric epilepsy

associated w

ith elevated

autoantibodies

89

Table 2 Clinical data from GADA-positive subjects.

Pt Epilepsy type GADA value(U/ml)

Age at onset,years/sex

EEG (side, lobe) MRI Treatment 6-Month outcome Diabetes

1 FI 7 10/F L C-T discharges Normal OXC Seizure control No2 FI 350 4/M Bilateral C discharges Minimal lateral ventricle

enlargement, L > RCBZ Seizure control No

3 FI 7 3/M R > L T, O and Pdischarges

Normal VPA Seizure control No

4 FI 6 6/F O discharges Minimal R lateral ventricleenlargement

OXC Seizure control No

5 FI 5 9/F R C-T discharges Normal OXC Seizure control No6 FS >2000 9/F P discharges, R > L Findings after surgical management

of hypophyseal teratomaTPM Seizure control Yes

7 FS 11 6/F Disturbed backgroundactivity; L F, T and Odischarges

Normal OXC Seizure control No

8 GI 70 15/M Generalized paroxysms Normal VPA Seizure control No9 GS (Aicardi

syndrome)6 0.3/F Multifocal

sharp-slow-wavedischarges

Hypoplasia of the corpus callosum;subependymal and subcortical graymatter heterotopia

TPM + VGB + LEV Drug-resistantepilepsy

No

10 GI 6 12/F Bilateral spike-waveparoxysms

Asymmetry of the lateral ventricles,L F gray matter heterotopia

VPA Seizure control No

11 UC 5 1/F Normal Ventriculomegaly; findings aftersurgical management ofinfratentorial subdural hemorrhage

VPA Seizure control No

12 UC 5 1/F Normal Bilateral periventricular whitematter lesions

VPA Seizure control No

13 UC 9 3/M Bilateral paroxysms Normal VPA Seizure control No14 UC 38 4/M Disturbed background

activityCyst of the septum pellucidum andcavum Vergae

VPA Seizure control No

ACA, anti-cardiolipin autoantibodies; GADA, glutamic acid decarboxylase autoantibodies; AI, autoimmune; epilepsy type (FI, focal idiopathic; FS, focal symptomatic; GI, generalized idio-pathic; GS, generalized symptomatic; UC, unclassified); lobe (C, central; C-T, centro-temporal; F, frontal; O, occipital; T, temporal; P, parietal); Pt, patient; treatment (CBZ, carbamazepine;LEV, levetiracetam; OXC, oxcarbazepine; TPM, topiramate; VGB, vigabatrine; VPA, valproate).

9

pisieoarmpd

gFaf

A

Ttf2U

R

A

A

B

B

B

E

F

G

H

C

I

K

L

M

M

N

P

R

S

S

S

U

U

V

V

V

0

artial epilepsy. In the present study, ACA were detectedn 13 (6.3%) subjects with epilepsy, similar to controls,uggesting that ACA might not be related to epilepsyn newly-diagnosed pediatric cases. However, we cannotxclude the role of ACA in modulation of the natural coursef epilepsy, especially in cases where ACA titers are high andntiphospholipid syndrome is present. These questions war-ant further studies to clarify the role of ACA in children byeasuring autoantibodies against neural phospholipids likehosphatidylcholine, which may be readily a target candi-ate for ACA in neural tissue.

In conclusion, GADA appears to be a marker for a sub-roup of newly-diagnosed pediatric patients with epilepsy.urther research is required to determine the prevalencend pathogenic role of GADA in pediatric subjects with dif-erent types of epilepsy.

cknowledgements

his study was supported by the Estonian Science Founda-ion, Grant Nos. 7749 and 8334; by targeted financial supportrom the Estonian Ministry of Education and Research, TARLA695, SARLA 11091 and SF0180035s08; and by the Europeannion through the European Regional Development Fund.

eferences

li, F., Rowley, M., Jayakrishnan, B., Teuber, S., Gershwin,M.E., Mackay, I.R., 2011. Stiff-person syndrome (SPS) andanti-GAD-related CNS degenerations: protean additions to theautoimmune central neuropathies. J. Autoimmun. 37, 79—87.

ykutlu, E., Baykan, B., Gürses, C., Gokyigit, A., Saruhan-Direskeneli, G., 2005. No association of anti-GM1 and anti-GADantibodies with juvenile myoclonic epilepsy: a pilot study.Seizure 14, 362—366.

aekkeskov, S., Aanstoot, H.J., Christgau, S., Reetz, A., Solimena,M., Cascalho, M., Folli, F., Richter-Olesen, H., De Camilli, P.,1990. Identification of the 64 K autoantigen in insulin-dependentdiabetes as the GABA-synthesizing enzyme glutamic acid decar-boxylase. Nature 347, 151—156.

ien, C.G., Scheffer, I.E., 2011. Autoantibodies in epilepsy. Epilep-sia 52, 18—22.

urton, A.R., Baquet, Z., Eisenbarth, G.S., Tish, R., Smeyne, R.,Workman, C.J., Vignali, D.A., 2010. Central nervous systemdestruction mediated by glutamic acid decarboxylase-specificCD4+ T cells. J. Immunol. 184, 4863—4870.

rrichiello, L., Perruolo, G., Pascarella, A., Formisano, P., Minetti,C., Striano, S., Zara, F., Striano, P., 2009. Autoantibodies toglutamic acid decarboxylase (GAD) in focal and generalizedepilepsy: a study on 233 patients. J. Neuroimmunol. 211,120—123.

alip, M., Carreno, M., Miró, J., Saiz, A., Villanueva, V., Quílez,A., Molins, A., Barceló, I., Sierra, A., Graus, F., 2012. Preva-lence and immunological spectrum of temporal lobe epilepsywith glutamic acid decarboxylase antibodies. Eur. J. Neurol. 19,827—833.

iometto, B., Nicolao, P., Macucci, M., Tavolato, B., Foxon, R.,Bottazzo, G.F., 1998. Temporal-lobe epilepsy associated withglutamic-acid-decarboxylase autoantibodies. Lancet 352, 457.

ill, K.E., Clawson, S.A., Rose, J.W., Carlson, N.G., Greenlee, J.E.,2009. Cerebellar Purkinje cells incorporate immunoglobulins andimmunotoxins in vitro: implications for human neurological dis-ease and immunotherapeutics. J. Neuroinflamm. 6, 31.

V

K. Veri et al.

ommission on Classification and Terminology of the InternationalLeague Against Epilepsy, 1989. Proposal for revised classifi-cation of epilepsies and epileptic syndromes. Epilepsia 30,389—399.

rani, S.R., Bien, C.G., Lang, B., 2011. Autoimmune epilepsies. Curr.Opin. Neurol. 24, 146—153.

wan, P., Sills, G.J., Kelly, K., Butler, E., Brodie, M.J., 2000.Glutamic acid decarboxylase autoantibodies in controlledand uncontrolled epilepsy: a pilot study. Epilepsy Res. 42,191—195.

iimatainen, S., Peltola, M., Sabater, L., Fallah, M., Kharazmi,E., Haapala, A.M., Dastidar, P., Knip, M., Saiz, A., Peltola,J., 2010. Clinical significance of glutamic acid decarboxy-lase antibodies in patients with epilepsy. Epilepsia 51,760—767.

atà, S., Muscas, G.C., Naldi, I., Rosati, E., Paladini, S., Cruci-atti, B., Bisulli, F., Paganini, M., Mazzi, G., Sorbi, S., Tinuper,P., 2008. Non-paraneoplastic limbic encephalitis associated withanti-glutamic acid decarboxylase antibodies. J. Neuroimmunol.199, 155—159.

cKnight, K., Jiang, Y., Hart, Y., Cavey, A., Wroe, S., Blank, M.,Shoenfeld, Y., Vincent, A., Palace, J., Lang, B., 2005. Serumantibodies in epilepsy and seizure-associated disorders. Neurol-ogy 65, 1730—1736.

emni, R., Braghi, S., Natali-Sora, M.G., Lampasona, V., Bonifacio,E., Comi, G., Canal, N., 1994. Autoantibodies to glutamic aciddecarboxylase in palatal myoclonus and epilepsy. Ann. Neurol.36, 665—667.

eltola, J., Kulmala, P., Isojärvi, J., Saiz, A., Latvala, K.,Palmio, J., Savola, K., Knip, M., Keränen, T., Graus,F., 2000. Autoantibodies to glutamic acid decarboxylasein patients with therapy-resistant epilepsy. Neurology 55,46—50.

anua, J., Luoma, K., Peltola, J., Haapala, A.M., Raitanen, J.,Auvinen, A., Isojärvi, J., 2004. Anticardiolipin and antinuclearantibodies in epilepsy — a population-based cross-sectionalstudy. Epilepsy Res. 58, 13—18.

aiz, A., Blanco, Y., Sabater, L., Gonzaléz, F., Bataller, L., Casamit-jana, R., Ramió-Torrentà, L., Graus, F., 2008. Spectrum ofneurological syndromes associated with glutamic acid decar-boxylase antibodies: diagnostic clues for this association. Brain131, 2553—2563.

olimena, M., Folli, F., Aparisi, R., Pozza, G., De Camilli, P., 1990.Autoantibodies to GABA-ergic neurons and pancreatic beta cellsin stiff-man syndrome. N. Engl. J. Med. 322, 1555—1560.

triano, P., Perruolo, G., Errichiello, L., Formisano, P., Beguinot, F.,Zara, F., Striano, S., 2008. Glutamic acid decarboxylase antibod-ies in idiopathic generalized epilepsy and type 1 diabetes. Ann.Neurol. 63, 127—128.

ibo, R., Lernmark, A., 2008. GAD65 autoimmunity — clinical stud-ies. Adv. Immunol. 100, 39—78.

ibo, R., Sullivan, E.P., Uibo, O., Lernmark, A., Salur, L., Kivik,T., Mandel, M., 2001. Comparison of the prevalence of glutamicacid decarboxylase (GAD65) and gliadin antibodies (AGA) in arandomly selected adult Estonian population. Horm. Metab. Res.33, 564—567.

errotti, A., Greco, R., Altobelli, E., Latini, G., Morgese, G.,Chiarelli, F., 2003. Anticardiolipin, glutamic acid decarboxylase,and antinuclear antibodies in epileptic patients. Clin. Exp. Med.3, 32—36.

ianello, M., Tavolato, N., Giometto, B., 2002. Glutamic acid decar-boxylase autoantibodies and neurological disorders. Neurol. Sci.23, 145—151.

ianello, M., Tavolato, B., Armani, M., Giometto, B., 2003. Cere-

bellar ataxia associated with anti-glutamic acid decarboxylaseautoantibodies. Cerebellum 2, 77—79.

ianello, M., Keir, G., Giometto, B., Betterle, C., Tavolato,B., Thompson, E.J., 2005. Antigenic differences between

YI., Uchimura, I., Hirata, Y., 2005. Type 1 diabetes mellitus and

Pediatric epilepsy associated with elevated autoantibodies

neurological and diabetic patients with anti-glutamic acid decar-boxylase antibodies. Eur. J. Neurol. 12, 294—299.

Vulliemoz, S., Vanini, G., Truffert, A., Chizzolini, C., Seeck, M.,2007. Epilepsy and cerebellar ataxia associated with anti-glutamic acid decarboxylase antibodies. J. Neurol. Neurosurg.Psychiatry 78, 187—189.

91

oshimoto, T., Doi, M., Fukai, N., Izumiyama, H., Wago, T., Minami,

drug-resistant epilepsy: presence of high titer of anti-glutamicacid decarboxylase autoantibodies in serum and cerebrospinalfluid. Intern. Med. 44, 1174—1177.