Anti-GAD Antibodies and Periodic Alternating Nystagmus

OBSERVATION Anti-GAD Antibodies and Periodic Alternating Nystagmus

Caroline Tilikete, MD; Alain Vighetto, MD; Paul Trouillas, MD; Jérome Honnorat, MD

Background: Autoantibodies directed against glu- Intervention: Baclofen, a GABAergic medication, was tamic acid decarboxylase (GAD-Ab) have recently been given to the patient. described in a few patients with progressive cerebellar ataxia, suggesting an autoimmune physiopathologic Main Outcome Measures: Eye movement recording mechanism. of spontaneous nystagmus and postrotatory vestibular responses. Objective: To determine the exact role of GAD-Ab and ␥-aminobutyric acid (GABA)–ergic neurotransmission in Results: Baclofen was effective in suppressing PAN and the pathogenesis of cerebellar ataxia. improving postrotatory vestibular responses but not for improving cerebellar ataxia. Design: Case report. Conclusion: The presence of PAN and the response to Setting: University neurological hospital. baclofen provide a unique opportunity to suggest a direct role of GAD-Ab in cerebellar dysfunction in this Patient: We report the case of a patient with subacute patient. cerebellar ataxia associated with GAD-Ab showing periodic alternating nystagmus (PAN). Arch Neurol. 2005;62:1300-1303

LUTAMIC ACID DECARBOX- ebellar ataxia associated with periodic al- ylase (GAD) is a major ternating nystagmus (PAN). Precise enzyme of the nervous knowledge of the physiopathologic mecha- system that catalyzes the nism of this nystagmus and its response conversion of glutamate to treatment may help to better under- toG␥-aminobutyric acid (GABA). Antibod- stand that of neurological symptoms as- ies against GAD (GAD-Ab) have been iden- sociated with GAD-Ab. tified in association with neurological manifestations such as stiff-person syndrome and more recently late-onset cer- REPORT OF A CASE ebellar ataxia.1-4 The frequent association of the latter with other autoimmune dis- Author Affiliations: A 76-year-old woman was admitted to the Neuro-Ophthalmology Unit eases as well as cerebrospinal fluid (CSF) hospital for a 3-year history of slowly pro- (Drs Tilikete and Vighetto) and inflammation suggests that late-onset cer- gressive cerebellar ataxia, blurred vision, and Neurology B (Drs Trouillas and ebellar ataxia could have an autoimmune- cognitive dysfunction. She had age-related Honnorat), Université Claude mediated pathogenesis when associated macular degeneration but no history of in- Bernard Lyon-I, Hospices Civils with GAD-Ab.1,5 sulin-dependent diabetes mellitus. de Lyon, and INSERM U534 The pathogenic role of GAD-Ab is con- At examination, she presented with (Drs Tilikete and Vighetto), troversial.5,6 Whether GAD-Ab can be pres- marked ataxia of stance and gait, mild Bron, France; and Institut ent as an epiphenomenon caused by de- ataxia of the right side of the body, and hy- Fédératif des Neurosciences Lyon, Hôpital struction of the nervous system or whether potonia. Adiadochokinesis was marked in Neuro-Cardiologique neurological manifestations can be re- her right hand. She had a score of 37 of (Drs Tilikete, Vighetto, lated to impaired GABA synthesis due to 100 on the International Cooperative Trouillas, and Honnorat), GAD-Ab is debated. We report the case of Ataxia Rating Scale.7 A neuropsychologi- Lyon, France. a patient presenting with late-onset cer- cal examination disclosed memory and ex-

(REPRINTED) ARCH NEUROL / VOL 62, AUG 2005 WWW.ARCHNEUROL.COM 1300

©2005 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 10/01/2021 ecutive dysfunction. There were no signs of stiff-person syndrome. An oculomotor examination disclosed a hori- A

zontal jerk-form nystagmus in the primary position of 15 R gaze, changing direction every 2 minutes, which was di- 10 agnosed as PAN. Gaze-evoked nystagmus, impaired 5 smooth-pursuit movements, and absent suppression of 0 the vestibulo-ocular reflex (VOR) by fixation were also –5 observed. Visual acuity, measured when the nystagmus –10 L –15 stopped before reversing, was 20/50 OD and 20/100 OS 0 2 4 6 8 10 at distance and 20/60 OD and 20/200 OS at near. An oph-

thalmogical examination disclosed retinal signs of macu- B lar degeneration. Magnetic resonance imaging of the brain 15 showed marked cerebellar atrophy with moderate cor- R 10 ° tical and subcortical atrophy of both cerebral hemi- 5 spheres. Although the CSF cell count and protein con- 0 tent were normal, numerous oligoclonal IgG bands were –5 Eye Position, found. The complete blood cell count and standard bio- –10 L logical test results were normal. Infections with human –15 immunodeficiency virus and borreliosis were excluded. 33 34 35 36 37 38 39 40 41 42 43 There was no evidence of any malignancy or paraneo- plastic cerebellar degeneration; the patient tested nega- C 15 tive for anti-Yo, anti-Hu, anti-Ri, antiamphiphysin, and R anti-CV2 antibodies in the serum and CSF. There was 10 no diabetes mellitus; thyroid hormone and thyroid an- 5 tibody levels were normal. There was no evidence of any 0 –5 other autoimmune disease; results of tests for anti- –10 L nuclear, anticytoplasm of the neutrophile polynuclear, –15 antiendomysium, and antigliadin antibodies were nega- 50 52 54 56 58 60 tive. Vitamin E, B1, B6, B12, and folate deficiencies were Time, s excluded. The GAD-Ab were measured by immunohis- tochemical analysis on frozen sections of paraformalde- Figure 1. Recording of periodic alternating nystagmus in the patient. 2 Positive values of horizontal eye position are assigned to rightward (R) hyde-fixed rat cerebellum as previously described ; movements, negative values to leftward (L) movements. A, Nystagmus is GAD-Ab were found at high levels in both the serum right beating. B, Nystagmus is changing direction. C, Nystagmus is left sample (1:16000) and the CSF (Ͻ1:500). beating. Absent data are due to blinking. Our patient was treated with baclofen (30 mg/d), and PAN was abolished within 3 weeks of progressive treat- ping before reversing. The time constant of postrotatory ment. Visual acuity slightly improved at near: 20/50 OD VOR was about 25 seconds and was abnormally long for and 20/100 OS. The ataxia did not improve, although a both directions (Figure 2A and B). During treatment with slight improvement on the ataxia scale was noted: 30 of baclofen, PAN was abolished and the postrotatory time con- 100. Cognitive functions were not modified, and higher stant returned to normal (Figure 2C and D). doses were not tolerated by the patient.

EYE MOVEMENT RECORDING COMMENT

Low-frequency (50-Hz sampling) 2-dimensional infra- Periodic alternating nystagmus is the best-understood red video-oculography was used (VNG Ulmer; Synapsys, form of acquired central nystagmus, with an animal Marseille, France). The patient wore a mask, with a cam- model,8 a mathematical model,9 and drug treatment (bac- era fixed in front of the right eye. Using contrast image lofen) that works in both animals8 and humans.9,10 There- detection of the eye and iris, horizontal and vertical po- fore, PAN in this patient provides a unique opportunity sitions of the eye were automatically calculated online. Af- to better understand the mechanisms of cerebellar dys- ter calibration, spontaneous nystagmus was recorded in function in GAD-Ab disorders. the primary eye position. Eye movement recording in our Acquired PAN is a spontaneous horizontal nystag- patient showed PAN, with a peak slow-phase velocity of mus present in primary gaze that reverses its direction 15°/s in darkness and an oscillatory period of 4 minutes about every 2 minutes.9 In humans and monkeys, PAN (Figure 1). The patient gave informed consent for eye has been reported with lesions of the caudal and medial movement recording in accordance with the require- parts of the cerebellum, involving mostly the nodulus and ments of the hospital’s ethics committee and the Declara- uvula.11,12 The nodulus and uvula appear to control the tion of Helsinki. time course of postrotational VOR, leading to pro- Right and left postrotatory VORs (progressive ramp of longed VOR as found in our patient. Normal vestibular 60 °/s of constant velocity for 1 minute followed by 100 repair mechanisms act to reverse the direction of the nys- °/s2 of deceleration step) were elicited in the horizontal tagmus, determining its periodic oscillations.12,13 Such os- plane. The step was applied when the nystagmus was stop- cillations would normally be suppressed by visual fixa-

(REPRINTED) ARCH NEUROL / VOL 62, AUG 2005 WWW.ARCHNEUROL.COM 1301

100 50

50 0

0 –50

–50 –100

–100 –150 ° –150 –200 0 10 20 30 40 50 60 0 10 20 30 40 50 60

B D

0 50 Cumulative Eye Position,

–50 0

–100 –50

–150 –100

–200 –150

–250 –200 0 10 20 30 40 50 60 0 10 20 30 40 50 60 Time, s Time, s

Figure 2. Recording of postrotatory vestibulo-ocular reflex (VOR) before (A and B) and after (C and D) treatment by baclofen in the patient. Postrotatory VOR was elicited in the horizontal plane, either to elicit a leftward response (A and C) or a rightward response (B and D). The cumulative slow-phase eye position is presented (y-axis) as a function of time (x-axis) in postrotatory conditions. Note the longer duration and slope of postrotatory responses before as compared with after treatment for both rotations. The time scale is equivalent on the 4 plots.

tion. In our patient, poor visual acuity due to age- drome without GAD-Ab.16 Furthermore, it has been shown related macular degeneration as well as floccular that immunoglobulins present in the CSF of a patient with involvement, known to control smooth-pursuit move- ataxia associated with GAD-Ab selectively suppress pre- ments and fixation, might explain the occurrence of synaptic GABA-mediated transmission from the basket cells nystagmus. In most patients with acquired PAN, a to the Purkinje cells in the cerebellum.17 GABAergic drug, baclofen, abolishes nystagmus.9,10 Phar- Our study supports the hypothesis of GABAergic neu- macological evidence suggests that the nodulus and uvula rotransmission impairment by showing that at least some maintain inhibitory control of the duration of vestibular neurological manifestations are the result of a GABAer- rotational responses by using GABA.14 In our patient, we gic neurotransmission deficit, suggesting a pathogenic role suggest that acquired PAN is linked to a functional of these antibodies. Along these lines, isolated down- GABAergic deficit of the cerebellar pathways that can be beat nystagmus associated with GAD-Ab has recently been alleviated by a GABAergic drug, baclofen. reported.18 Although no attempt to treat the patient with The pathogenic effect of GAD-Ab in neurological mani- GABAergic drugs was mentioned in the publication, the festations is not yet fully accepted, specifically in progres- authors suggest GABAergic neurotransmission involve- sive late-onset cerebellar ataxia. The frequency of insulin- ment at the floccular level. Our patient also presented dependent diabetes mellitus and other organ-specific with cerebellar ataxia and cognitive impairment that did autoimmune manifestations observed in this group of pa- not respond to baclofen. This absence of a drug re- tients, as well as the high percentage of CSF inflamma- sponse might be due to the more complex GABAergic neu- tion, suggests that this cerebellar syndrome could have an ral circuitry for the control of ataxia and cognitive func- autoimmune-mediated pathogenesis.1,5,15 In this case, one tions. It could also be caused by permanent neuronal loss could suggest that autoimmunity directed against GAD and/or other mechanisms, such as a cellular-mediated au- would disturb GABA synthesis and GABAergic neuro- toimmune pathogenesis. transmission. Experimental data have shown that the serum or CSF of patients with stiff-person syndrome leads Accepted for Publication: July 12, 2004. to functional impairment of GABAergic synaptic trans- Correspondence: Caroline Tilikete, MD, Neuro- mission in vitro, suggesting a specific effect of GAD-Ab.16 Ophthalmology Unit, Hôpital Neurologique, Hospices This functional impairment of GABAergic transmission is Civils de Lyon, 59 Bd Pinel, 69 677 Bron CEDEX, France not found in the serum of patients with stiff-person syn- ([email protected]).

(REPRINTED) ARCH NEUROL / VOL 62, AUG 2005 WWW.ARCHNEUROL.COM 1302

©2005 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 10/01/2021 Author Contributions: Study concept and design: Tilik- of the World Federation of Neurology. International Cooperative Ataxia Rating ete, Vighetto, Trouillas, and Honnorat. Acquisition of data: Scale for pharmacological assessment of the cerebellar syndrome. J Neurol Sci. 1997;145:205-211. Tilikete. Analysis and interpretation of data: Tilikete, 8. Waespe W, Cohen B, Raphan T. Dynamic modification of the vestibulo-ocular Vighetto, and Honnorat. Drafting of the manuscript: Til- reflex by the nodulus and uvula. Science. 1985;228:199-202. ikete. Critical revision of the manuscript for important in- 9. Leigh RJ, Robinson DA, Zee DS. A hypothetical explanation for periodic alter- tellectual content: Vighetto, Trouillas, and Honnorat. Ad- nating nystagmus: instability in the optokinetic-vestibular system. AnnNYAcad ministrative, technical, and material support: Tilikete, Sci. 1981;374:619-635. 10. Halmagyi GM, Rudge P, Gresty MA, Leigh RJ, Zee DS. Treatment of periodic al- Vighetto, Trouillas, and Honnorat. Study supervision: ternating nystagmus. Ann Neurol. 1980;8:609-611. Honnorat. 11. Korres S, Balatsouras DG, Zournas C, Economou C, Gatsonis SD, Adamopoulos G. Periodic alternating nystagmus associated with Arnold-Chiari malformation. J Laryngol Otol. 2001;115:1001-1004. REFERENCES 12. Furman JM, Wall C III, Pang DL. Vestibular function in periodic alternating nystagmus. Brain. 1990;113:1425-1439. 1. Honnorat J, Saiz A, Giometto B, et al. Cerebellar ataxia with anti-glutamic acid 13. Dow ER, Anastasio TJ. Induction of periodic alternating nystagmus in intact gold- decarboxylase antibodies: study of 14 patients. Arch Neurol. 2001;58:225- fish by sinusoidal rotation. Neuroreport. 1997;8:2755-2759. 230. 14. Cohen B, Helwig D, Raphan T. Baclofen and velocity storage: a model of the ef- 2. Honnorat J, Trouillas P, Thivolet C, Aguera M, Belin MF. Autoantibodies to glu- fects of the drug on the vestibulo-ocular reflex in the rhesus monkey. J Physiol. tamate decarboxylase in a patient with cerebellar cortical atrophy, peripheral neu- 1987;393:703-725. ropathy, and slow eye movements. Arch Neurol. 1995;52:462-468. 15. Giometto B, Miotto D, Faresin F, Argentiero V, Scaravilli T, Tavolato B. Anti- 3. Abele M, Weller M, Mescheriakov S, Burk K, Dichgans J, Klockgether T. Cer- gabaergic neuron autoantibodies in a patient with stiff-man syndrome and ataxia. ebellar ataxia with glutamic acid decarboxylase autoantibodies. Neurology. 1999; J Neurol Sci. 1996;143:57-59. 52:857-859. 16. Dinkel K, Meinck HM, Jury KM, Karges W, Richter W. Inhibition of gamma- 4. Saiz A, Arpa J, Sagasta A, et al. Autoantibodies to glutamic acid decarboxylase aminobutyric acid synthesis by glutamic acid decarboxylase autoantibodies in in three patients with cerebellar ataxia, late-onset insulin-dependent diabetes melli- stiff-man syndrome. Ann Neurol. 1998;44:194-201. tus, and polyendocrine autoimmunity. Neurology. 1997;49:1026-1030. 17. Mitoma H, Song SY, Ishida K, Yamakuni T, Kobayashi T, Mizusawa H. Presyn- 5. Honnorat J, Trouillas P. Clinical presentation of immune mediated cerebellar ataxia aptic impairment of cerebellar inhibitory synapses by an autoantibody to gluta- [in French]. Rev Neurol (Paris). 2003;159:11-22. mate decarboxylase. J Neurol Sci. 2000;175:40-44. 6. Vianello M, Tavolato B, Giometto B. Glutamic acid decarboxylase autoantibod- 18. Antonini G, Nemni R, Giubilei F, et al. Autoantibodies to glutamic acid decarbox- ies and neurological disorders. Neurol Sci. 2002;23:145-151. ylase in downbeat nystagmus. J Neurol Neurosurg Psychiatry. 2003;74:998- 7. Trouillas P, Takayanagi T, Hallett M, et al; Ataxia Neuropharmacology Committee 999.

Trial Registration Required

As a member of the International Committee of Medi- cal Journal Editors (ICMJE), Archives of Neurology will require, as a condition of consideration for publication, registration of all trials in a public trials registry (such as http://ClinicalTrials.gov). Trials must be registered at or before the onset of patient enrollment. This policy ap- plies to any clinical trial starting enrollment after July 1, 2005. For trials that began enrollment before this date, registration will be required by September 13, 2005, before considering the trial for publication. The trial registration number should be supplied at the time of sub- mission. For details about this new policy, and for informa- tion on how the ICMJE defines a clinical trial, see the editorials by DeAngelis et al in the September 8, 2004 (2004;292:1363-1364) and June 15, 2005 (2005;293: 2927-2929) issues of JAMA. Also see the Instructions to Authors on our Web site: www.archneurol.com.

(REPRINTED) ARCH NEUROL / VOL 62, AUG 2005 WWW.ARCHNEUROL.COM 1303