BMJ - LETTERS TO THE EDITORLETTERS TO THE EDITOR Postictal psychosis related regional cerebral hyperperfusion Postictal psychosis is a known complication of complex partial seizure

LETTERS TOTHE EDITOR

Postictal psychosis related regionalcerebral hyperperfusion

Postictal psychosis is a known complicationof complex partial seizure in particulartemporal lobe epilepsy. It usually runs abenign and self limiting course. A postictalphenomenon with focal cerebral hypofunc-tion (similar to Todd’s palsy), rather thanongoing seizure activity, has beenpostulated.1 2 Surface EEG is either normalor showing non-specific slow waves.3 Hence,antipsychotic medications are prescribedinstead of antiepileptic drugs. Until recently,the pathogenic mechanisms have remainedunknown. In this communication, we reporton two patients with postictal psychosis, dur-ing which a cerebral SPECT study showed ahyperperfusion signal over the right temporallobe and contralateral basal ganglion. Ashyperperfusion in ictal cerebral SPECT isclosely linked to epileptic activities,4 our find-ings support a contrary explanation forpostictal psychosis.

Prolonged video-EEG telemetry study wasperformed in patients who underwent presur-

gical evaluation for epilepsy surgery. Antiepi-leptic drugs were withdrawn to facilitateseizure recording. A diagnosis of temporallobe epilepsy was based on analysis of theelectroclinical events and, if applicable, post-operative outcome after anterior temporallobectomy. Psychosis was diagnosed accord-ing to the fourth edition of the diagnosticsand statistical manual of mental disorders(DSM-IV) criteria of brief psychotic disorderswithout marked stressor. HMPAO-SPECT wasperformed during the psychotic period,which ranged from 2–4 days after the last sei-zure. Interictal cerebral SPECT, brain MRI,and a Wada test were performed as part ofpresurgical evaluation.

Patient 1 was a 34 year old Chinese womanwith complex partial seizures since the age of18. Her seizure control was suboptimal on acombination of antiepileptic drugs. BrainMRI showed a small hippocampus on theright. Interictal EEG showed bilateral tempo-ral sharp waves and ictal recordings con-firmed a right temporal epileptogenic focus.A Wada test confirmed right hippocampalmemory dysfunction. Six hours after her lastsecondary generalised tonic-clonic seizureafter video-EEG telemetry, she began todevelop emotional lability, talking nonsense,motor restlessness, and auditory hallucina-tion. A cerebral SPECT study was performedat day 4 after her last seizure. Her psychoticfeatures persisted although she was takingantipsychotic medication (pimozide). Cer-ebral SPECT showed a clear hyperperfusion

signal over the right lateral temporal neocor-tex and contralateral basal ganglion. Aninterictal cerebral SPECT study was repeatedat 4 weeks after postictal psychosis whichshowed a complete resolution of hyperper-fusion signal in the right temporal lobe andbasal ganglia. Anterior temporal lobectomywas performed and she became seizure freeafter surgery.

Patient 2 was a 44 year old man withintractable complex partial seizures since theage of 30. His seizures were intractable tomultiple antiepileptic drugs. Brain MRIshowed left hippocampal sclerosis. Interictalcerebral SPECT showed a relative hypoper-fusion area over the left hemisphere. Interic-tal surface EEG was non-lateralising but ictalEEG disclosed a right hemispheric onset. Onwithdrawal of antiepileptic drugs, sevencomplex partial seizures with secondary gen-eralised tonic clonic seizures were recordedwithin a period of 72 hours. His usualantiepileptic drugs were then restarted.Thirty hours after his last secondary general-ised tonic-clonic seizure; he began to developemotional lability, talking nonsense, restless-ness, auditory hallucination, persecutorydelusion, and delusion of superstition. Cer-ebral SPECT study, performed 2 days laterwhile his psychotic features persisted,showed two relative hyperperfused areas overthe right temporal neocortex and contralat-eral basal ganglion in addition to the originalhypoperfused area over the left hemisphere.An antipsychotic agent (thioridazine) was

Interictal SPECT and SPECT performed during postictal psychosis. (Top) A SPECT study of patient 1 showing areas of relative hyperperfusion over theright temporal neocortex (red arrows) and the left basal ganglia (blue and yellow arrows) during postictal psychosis. (Bottom) SPECT study of patient 2showing areas of hyperperfusion over the right temporal neocortex and the left basal ganglia. Arrows indicate areas of hyperperfusion.

J Neurol Neurosurg Psychiatry 2000;68:100–126100

on July 1, 2021 by guest. Protected by copyright.

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.68.1.112 on 1 January 2000. Dow

nloaded from

http://jnnp.bmj.com/

started after the cerebral SPECT. Hispsychotic symptoms resolved 2 weeks laterwith full recovery.

Cerebral SPECT performed during theinterictal period (IP) and during postictalpsychosis (PP) were analysed visually andareas of hyperperfusion were identified.Quantitaive data at regions of interest (ROIs)were measured on coronal and axial slidescontaining basal ganglia (BG), mesial (MT),and lateral (LT) temporal lobe structures.Asymmetry index (ASI) was calculated as((ROI focus−ROI contralateral)/ROIfocus+ROI contralateral))×200%. We set anarbitrary change of ASI >100% to be signifi-cant. As there were only two patients, statisti-cal testing was not performed.

Both patients showed postictal psychosisand had a regional increase in rCBF over theright temporal neocortex and the left basalganglion compared with their interictal study(figure). Quantitative analysis for patient 1showed changes of ASI during IP and PPover right MT was +75% (-6.64476 to-1.65289); over the right LT was +1167.8%(1.07527 to 12.55764); and over the left BGwas +206.8% (-2.07373 to 2.21574). Quan-titative analysis for patient 2 showed changesof ASI during IP and PP over right MT was−3.8% (13.14217 to 12.64158); over rightLT was +178.6% (10.4696 to 18.70027);and over left BG was +155.9% (−5.85556 to3.27522).

Postictal psychosis is a distinct clinicalevent associated with temporal lobe epilepsy.1

The diagnosis of postictal psychosis requiresa close temporal relation between bouts ofcomplex partial seizures and the onset of psy-chosis. The psychosis usually develops after acluster of complex partial seizures precipi-tated by abrupt withdrawal of antiepilepticdrugs.1 The cluster occurs in patients withpoor drug compliance or during video EEGtelemetry studies when antiepileptic drugsare withdrawn purposefully. The clinicalcourse of postictal psychosis is usually benignand predictable.1 5 In our patients, theduration of psychotic disturbances lastedfrom 10 to 14 days, which is in keeping withthe good prognosis. Antipsychotic drugs,such as haloperidol and fluphenazin are usu-ally prescribed.1

The underlying mechanism of postictalpsychosis is unknown. Postictal cerebralhypofunction has been postulated as an ana-logue to Todd’s paralysis after seizure.1 2

However, the presence of increased rCBFduring postictal psychosis, may suggest analternative explanation as ictal SPECT hasbeen shown to be highly sensitive and specificin demonstrating seizure foci.4

To conclude, our results are contradictoryto the hypofunction theory of Todd’s paraly-sis in postictal psychosis. We think that thesehyperperfusion areas are responsible for thepostictal psychosis. Further serial studies

with cerebral SPECT or PET may enhanceour understanding on the mechanism of pos-tictal psychosis.

G C Y FONGK Y FONG

W MAKK L TSANGK H CHAN

R T F CHEUNGS L HO

Division of Neurology, University Department ofMedicine, The University of Hong Kong

W Y HODepartment of Nuclear Medicine, Queen Mary

Hospital, Hong Kong

Correspondence to: Dr G C Y Fong, Departmentof Medicine, Queen Mary Hospital, Pokfulam Road,Hong Kong email cyfong.medicine @graduate.hku.hk

1 Savard G, Andermann F, Olivier A, et al. Postic-tal psychosis after partial complex seizures: amultiple case study. Epilepsia 1991;32:225–31.

2 Morell F. Memory loss as a Todd’s paresis.Epilpesia 1980;21:185.

3 Kanner AM, Stagno S, Kotagal P, et al. Postictalpsychiatric events during prolonged video-electro-encephalographic monitoring studies.Arch Neurol 1996;53:258–63.

4 Ho SS, Berkovic SF, Newton MR, et al. Parietallobe epilepsy: clinical features and seizurelocalization by ictal SPECT. Neurology 1994;44:2277–84.

5 Toone B. Psychoses and epilepsy. In: ReynoldsEH, Trimble MR, eds. Epilepsy and psychiatry.London: Churchill Livingstone, 1981:113–7.

Oncofetal matrix glycoproteins incerebral arteriovenous malformationsand neighbouring vessels

Cerebral arteriovenous malformations(AVMs) are thought to be congenital lesionsexhibiting features of either mature vascularwalls or embryonal anastomotic plexuses. It isgenerally assumed that changes in size aredependent on enlargement of the venouscompartment, organisation in the setting ofmicrohaemorrhages, and gliosis. However,recent findings are consistent with thehypothesis of ongoing angiogenesis.1 2

Previous research from this laboratorydisclosed that peculiar isoforms of fibronectin(FN) and tenascin (TN) typically occur infetal and neoplastic tissues.3–5 These isoformsare a blend of structurally diVerent glycopro-teins that result from alternative splicing of theprimary transcript and are mainly expressedin the extracellular matrix. Their expression isundetectable in normal adult tissues, with theexception of the vessels in the regeneratingendometrium. To gain further insight into thepathobiology of the AVMs the present reportsought to ascertain whether these lesions alsoexpress oncofetal FN and TN isoforms.

Tissue samples were obtained after neuro-surgical excisions of ruptured AVMs. All 10patients had experienced an intracerebralhaemorrhage as the first clinical manifesta-

tion of their disease. There was no drughistory before bleeding. Control specimensfrom two right gyri recti and one cerebellartonsil were obtained, respectively, from op-erations for ruptured aneurysms of the ante-rior communicating artery or for ArnoldChiari disease.

Immunohistochemical evaluations wereperformed on 5 µm thick cryostat sectionsusing a protocol reported previously.5 Owingto the limited amount of available material,only in a few cases was some fresh tissueretained to allow western blots. Distributionof FN and TN isoforms was investigatedusing three monoclonal antibodies (mAbs) ortwo Ab fragments, obtained by phage displaytechnology, respectively. These Abs, preparedin our laboratory, were found to work on freshfrozen material. According to the previouscharacterisations the BC-1 mAb and theTN-11 Ab fragments are specific for isoformsoccurring almost exclusively in fetal tissuesand in tumours, with the recognised TN iso-form being typically associated with anaplas-tic gliomas (table).

Control sections were processed identicallyto the other specimens, but the primaryantibody was substituted with a specificimmunoglobulin of recombinant antibodies.The antibodies were blocked using the specificantigens. The antigens were recombinant pro-tein containing the epitope produced in E Coli.For the mAb BC-1 we used the recombinantprotein containing the type-III repeats 7B-8–9.For the mAb IST-4 we used the recombinantprotein containing the type-III repeats 2–8.For the recombinant antibodies TN-11 andTN-12 the recombinant type-III repeat C andthe recombinant fragment containing theEGF-like repeats were used, respectively.

All 10 AVMs were found to contain largeamounts of FN and TN, as shown by intenseimmunostaining with the use of the IST-9 /IST-4 mAbs and the TN-12 Ab fragment.The staining was localised either in theendothelium or the subendothelial layer. Apositive response was found in several artery-like vessels and in a few vessels with thinnerwalls using the mAb BC-1. Staining with theTN-11 Ab fragment showed occurrence oftype III repeat C TN isoform in the innerlayers of the vascular components of thenidus, irrespective of their morphology.

Six out of the 10 examined specimens werefound to contain portions of cerebral tissuesurrounding the angiomatous nidus. In allthese cases the wall of several vesselsexhibited intense staining with the use of theTN-11 Ab fragment. Using the BC-1 mAbsome of these vessels exhibited some staining(figure). In the control specimens (brain andcerebellum) both the FN isoform containingthe ED-B sequence (ED-B+FN), and thetype III repeat C TN isoform were absent,despite the widespread distribution of totalFN and TN in the vascular walls.

Characterisation of the employed Abs and distribution of the recognized isoforms.

Anti-FN mAbs4 5 Anti-TN Ab fragments3

IST-4 IST-9 BC-1 TN-12 TN-11

Recognised isoforms Total FN Isoforms containingthe ED-A sequence

Isoform containing the ED-Bsequence

Total TN Type III repeat C Isoform

Distribution of the isoform (s) Widespread Widespread Absent in adult tissues (with theexception of the regeneratingendometrium)

Widespread Absent in adult tissues

Present in fetal tissues

Present in the vascular wall and thematrix of fetal tissues and tumours

Absent in several types ofmalignancies

Present in the vascular wallof anaplastic gliomas

J Neurol Neurosurg Psychiatry 2000;68:100–126 101


http://jnnp.bmj.com

/J N

eurol Neurosurg P


nloaded from


Previous findings showed that ED-B+FNpresents with conformational modifications inits central part and results from deregulation ofFN pre-mRNA.4 The distribution of thisisoform was found to be highly restricted innormal adult tissues. By contrast, ED-B+ FNexhibited widespread distribution in the vascu-lature of fetal tissues, including brain, and ofseveral types of malignancies. It was thereforeregarded as a marker of angiogenesis.5

Similarly, the type III repeat C TNisoform, recognised by the Ab fragmentTN-11, was found to occur in the vascularwalls of anaplastic gliomas. Northern blotanalysis showed that the mRNA of thisisoform was undetectable in normal tissuesand some malignancies, but was present inlarge amounts in fetal tissues, includingbrain, and in glioblastomas3

Recent advances in the pathology of cerebralAVMs suggest that these lesions might not bestatic. Tyrosine kinase, an endothelial cell spe-cific receptor upregulated in glioblastomas,was found to be highly expressed in bothAVMs and in the vessels of cerebral tissue bor-dering the malformations, by contrast with thedown regulation occurring in the vasculatureof the normal brain.1The pattern of distribu-tion of structural proteins was consistent withthe hypothesis of diVuse activation of angio-genesis, without specific relation to individualvessel types.2

Furthermore, use of the cell proliferationmarker MIB-1 showed endothelial prolifera-tion in arterioles, venules, and capillaries ofthe cerebral tissue neighbouring AVMs.1

The present findings indicate that aparticular FN isoform, mainly expressed bythe vasculature of fetal and tumorous tissues,as well as a TN isoform typically detected inthe walls of vessels in anaplastic gliomas, alsooccur in AVMs and in vessels of adjacent cer-ebral tissue, but that both isoforms are absentin normal brain. This evidence providesfurther support to the hypothesis of ongoingangiogenesis in and around these lesions.

The presence of angiogenic features inAVMs might result from maintenance of pro-liferating and remodelling potentials, or froma specific response to haemodynamic stress invascular structures subjected to increasedblood flow and pressure. Occurrence of thesefeatures also in vessels lying in areas periph-eral to the nidus might be related torecruitment of the neighbouring vasculature,possibly dependent on focal ischaemia in thesetting of arteriovenous shunting.1 2 However,the presence in apparently normal vascula-ture of molecules typically occurring in fetaltissues and malignancies indicate that cer-ebral AVMs may not be static lesions. Furtherstudies are needed to ascertain whether thisphenomenon results merely from haemody-namic stress or actually reflects an intrinsicgrowth potential. Should this second be thecase, current therapeutic strategies wouldpossibly require revision.

This study was partially supported by theNational Research Council (CNR), AIRCand the Ministry of University and ScientificResearch (MURST). We thank Sergio De-seri, EE, for his technical help and Mr. Tho-mas Wiley for manuscript revision.

ANTONIO PAUA DORCARATTO

G L VIALEDI S C A T Departement of Surgery, Division of

Neurosurgery, University of Genoa Medical School, SMartino Hospital, Pad 2, Largo Rosanna Benzi 10,

16132 Genova, Italy

P CASTELLANIA SIRI

L ZARDILaboratory of Cell Biology National Cancer Institute,

Genoa, Italy

Correspondence to: Dr A Pau

1 Hatva E, Jääskeläinen J, Hirvonen H, et al. Tieendothelial cell-specific receptor tyrosine ki-nase is upregulated in the vasculature of arter-ovenous malformations. J Neuropathol ExpNeurol 1996;55:1124–33.

2 Rothbart D, Awad IA, Jiyon L, et al. Expressionof angiogenic factors and structural proteins incentral nervous system vascular malforma-tions. Neurosurgery 1996;38:915–25.

3 Carnemolla B, Castellani P, Borsi L, et al. Iden-tification of a glioblastoma associated tn-c iso-form by a high aYnity recombinant antibody.Am J Pathol 1999;154:1345–52.

4 Carnemolla B, Leprini A, Allemanni G, et al.The Inclusion of type III repeat ED-B in thefibronectin molecule generates conformationalmodifications that unmask a cryptic sequence.J Biol Chem 1992;267:24689–92.

5 Castellani P, Viale G, Dorcaratto A, et al. Thefibronectin isoform containing the ED-B on-cofetal domain: a marker of angiogenesis. Int JCancer 1994;59:612–18.

Hashimoto’s encephalopathy presentingas “myxoedematous madness”

The neuropsychiatric sequelae of hypo-thyroidism range from lethargy and mentalslowing to the florid psychotic illness referredto as “myxoedematous madness”. The lastcondition is characterised by frank hypo-thyroidism accompanied by psychosis, andmay respond completely to thyroxine.1 Morerecently described is a syndrome of subacuteencephalopathy, associated with high titres ofthyroid autoantibodies, raised CSF protein,EEG abnormalities, and perfusion deficits inthe presence of normal structural neuro-imaging.2–4 In most cases, the encephalopathyoccurs without any gross change in circulat-ing concentrations of thyroid hormones, sug-gesting that an inflammatory process isresponsible for the cerebral dysfunction. Inthe absence of pathological data, the evidencefor a specific pathogenetic mechanism islargely circumstantial: a small vessel vasculitisand immune complex deposition have bothbeen suggested.3 4

Although none of the published cases ofHashimoto’s encephalopathy has describedpsychosis as a primary feature, it is possiblethat “myxoedematous madness”, a conditionfirst described in detail by Asher in 19491 liesin a range of encephalopathic phenomenamediated by autoimmune mechanisms. Thissuggestion would certainly be consistent withthe range of clinical presentations of otherautoimmune cerebral vasculitides.5 As au-toimmune thyroiditis is the commonest causeof thyroid failure in this country,6 it is likely tohave been present in at least some of Asher’soriginal 14 cases. Although most had floridmyxoedematous features at psychiatric pres-entation, this may simply reflect the diYcultyof diagnosing subclinical thyroid diseasebefore rapid laboratory assays became widelyavailable. Many features of the present case,however, favoured an endocrine rather thanan inflammatory mechanism, suggesting thatthe condition of “myxoedematous madness”,though rare, remains a valid diagnostic entity.

A 63 year old market stallholder withoutmedical or psychiatric history was brought toa local psychiatric hospital by the police. Hisbusiness had been in decline for severalmonths, and his family had noticed uncharac-teristic emotional lability. In the weekspreceding admission he had experienceddelusions and hallucinations, and exhibiteduncharacteristic behaviour. He had reporteda vision of the crucifixion, and hearing thevoice of his dead mother. He claimed that hishouse was occupied by the devil, drovearound aimlessly in his car, and appearedconstantly fearful and withdrawn. On the dayof admission he had made a bonfire in thegarden and burned his wife’s clothes, familyphotographs, furniture, and business papers.When his wife and son tried to intervene he

Immunostaining with the TN-11 Ab fragment or the BC-1 mAb shows the presence of the type IIIrepeat C TN-(A) and ED-B+ FN-(B) isoforms in angiomatous vessels. These isoforms are alsopresent in the wall of vessels of the cerebral tissue adjacent to the angiomatous nidus (TN: C; FN: D).Bar=10 µm.

102 J Neurol Neurosurg Psychiatry 2000;68:100–126


http://jnnp.bmj.com

/J N

eurol Neurosurg P


nloaded from


became aggressive and threatened them witha saw. The general practitioner was called andsuspected either an acute psychosis or asevere depressive illness. Police assistance wasrequested because of the patient’s continuingviolent behaviour.

On admission he was unkempt but coop-erative and apparently euthymic. He denieddepression, but displayed no insight into theirregularity of his behaviour. No psychoticfeatures were seen, although during theadmission he consistently rationalised allreported psychotic phenomena. He wasaggressive towards staV and made repeatedattempts to abscond. General physical exam-ination was unremarkable. Neurological ex-amination was normal except for spokenlanguage, which was fluent and grammatical,but contained word finding pauses, circum-locutions, and occasional semantic errors(for example, “I just want to get my feet backon the table”). Formal neuropsychologicaltesting, and a screen of laboratory tests forreversible causes of encephalopathy, wereperformed on admission, and results are pre-sented below (column A). Attention is drawnto his mild naming deficit, and poorperformance on the Rey figure, which wasdue to planning rather than visuospatialerrors, suggesting a predominantly “dysex-ecutive” pattern. CT and EEG were bothnormal, and SPECT disclosed widespreadbut mild cortical hypoperfusion. Trifluop-erazine (2 mg twice daily) was started onadmission, and thyroxine (75 µg once daily)added 1 week later. His mental state andbehaviour stabilised, leading to dischargeafter 2 months.

At 6 month follow up the patient hadstopped neuroleptic drugs, but continuedtaking thyroxine. He reported feeling “backto normal”, had bought a new house, and wasworking as a part time shop assistant. He stillhad subtle word finding diYculties, and wasreferred to the regional memory clinic forfurther evaluation, which took place 6months later. Behavioural assessment showedpersisting deficits in delayed recall of verbalmaterial, verbal fluency, and visuospatialfunction. Formal psychometric testing, bloodtests, and SPECT were repeated, 1 year afterthe original examinations. Laboratory andneuropsychological results are presented inthe table. It is of note that, whereas his nam-ing ability had improved, performance onfrontal executive tasks remained impaired.The appearance of the follow up SPECT dif-

fered minimally, if at all, from the first exam-ination.

In summary, therefore, this patient pre-sented in clear consciousness with a firstepisode of acute psychosis, and evidence ofsubtle executive and linguistic neuro-psychological disturbance, on the backgroundof gradual behavioural and aVective change.He was profoundly hypothyroid due to anautoimmune thyroiditis, but there was noclinical evidence of thyroid failure other thanthe abnormal mental state. The psychiatriccomponent of his illness recovered fully, andthe antithyroid microsomal antibody titre fellmarkedly after thyroxine replacement, al-though his mild neuropsychological deficitsremained unchanged. Corticosteroids werenot used at any stage.

The response to thyroxine does not, initself, imply that the cerebral illness had anendocrine origin; a recent report described apatient with a subacute encephalopathicillness and compensated hypothyroidism inthe presence of increased antimicrosomalantibodies, all of which responded to thyrox-ine replacement alone.4 In that case, however,both EEG and SPECT were abnormal, theSPECT showing multiple areas of severelyreduced perfusion, which normalised withtreatment. By contrast, in the present case theEEG was normal and the SPECT abnormal-ity was marginal and changed little, if at all,with treatment. The evidence for a significantvasculitic component to the illness is, there-fore, unconvincing.

The mild and relatively circumscribedneuropsychological deficits coupled withflorid psychotic phenomena, also contrastwith the profound global disturbance of cog-nition usually associated with Hashimoto’sencephalopathy.3 This distinction suggeststhat microvascular disruption and thyroidhormone depletion may emphasise diVerentaspects of the clinical range in Hashimoto’sencephalopathy. Although the present casewould support Asher’s conclusion that thepsychiatric features of Hashimoto’s encepha-litis typically respond to thyroid replacement,it additionally suggests that subtle neuro-psychological deficits may be apparent evenin the absence of obvious cerebral perfusiondeficits, and that these may not be fullyreversible.

P GARRARDJ R HODGES

University of Cambridge Neurology unit,Addenbrooke’s Hospital, Cambridge CB2 2QQ, UK

P J DE VRIESN HUNT

University of Cambridge Department of Psychiatry,Addenbrooke’s Hospital, Cambridge CB2 2QQ, UK

A CRAWFORDJ R HODGES

MRC Cognition and Brain Sciences Unit, 15 ChaucerRoad, Cambridge CB2 2EF, UK

K BALANDepartment of Nuclear Medicine, Addenbrooke’s

Hospital, Cambridge CB2 2QQ, UK

Correspondence to: Dr P Garrard, University ofCambridge Neurology Unit, Box 165, Adden-brooke’s Hospital, Cambridge CB2 2QQ, UKemail [email protected]

1 Asher R. Myxoedematous madness. BMJ1949;555–62.

2 Thrush DC, Boddie HG. Episodic encepha-lopathy associated with thyroid disorders. JNeurol Neurosurg Psychiatry 1974;37:696–700.

3 Shaw PJ, Walls TJ, Newman PK, et al.Hashimoto’s encephalopathy: a steroid respon-sive disorder associated with high anti-thyroidantibody titers: report of 5 cases. Neurology1991;41:228–33.

4 Forchetti CM, Katsamakis G, Garron DC.Autoimmune thyroiditis and a rapidly progres-sive dementia: global hypoperfusion onSPECT scanning suggests a possible mech-anism. Neurology 1997;49:623–6.

5 Scolding NJ, Jayne DR, Zajicek J, et al. Cerebralvasculitis: recognition, diagnosis and manage-ment. Q J Med 1997;90:61–73.

6 Dayan CM, Daniels GH. Chronic autoimmunethyroiditis. N Engl J Med 1996;335:99–107.

Alien hand sign in Creutzfeldt-Jakobdisease

The clinical picture of Creutzfeldt-Jakob dis-ease (CJD) includes various movement disor-ders such as myoclonus, parkinsonism,hemiballism, and dystonia. We report on apatient with CJD who manifested the alienhand sign. We suggest that CJD should beincluded in the diVerential diagnosis ofdiseases which present with an alien hand.

Creutztfeldt-Jakob disease, one of thehuman prion diseases, is characterised byrapidly progressive mental and motordeterioration.1 Involuntary movements occurin above 90% of the patients in the course ofthe disease, the most common beingmyoclonus.1 Other movement disordersrange from tremor to chorea, athetosis,dystonia, and hemiballism.1 We report on apatient with CJD who presented with an alienhand.

Alien hand is a rare and striking phenom-enon defined as “a patient’s failure to recog-nise the action of one of his hands as his own”.2

One of the patient’s hands acts as a stranger tothe body and is uncooperative. Thus, there isloss of feeling of ownership but not loss of sen-sation in the aVected hand. Originally de-scribed in callosal tumours,3 the aetiology ofalien hand also includes surgical callosotomy,4

infarction of the medial frontal cortex, occipi-totemporal lobe, and thalamus,1 5 infection,6

and corticobasal degeneration.5 7

A 70 year old, right handed Jewish manborn in Argentina, living in Israel for the past20 years, was admitted to the NeurologyDepartment. Until a month before hisadmission, he was apparently healthy andhelped in the accounting oYce of the villagewhere he lived. His neurological illness hadpresented insidiously during the past monthwith unsteadiness of gait and frequent falls.He also manifested behavioural changes,became aggressive, and had visual hallucina-tions, perceiving insects and mice movingthrough his visual field. Often, he expressedhis fear from seeing that the “ceiling was

Table 1 Laboratory and neuropsychological results at presentation (A) and at 12 month follow up (B)

Laboratory (units) A B

Full blood count Normal NormalErythocyte sedimentation rate 12 6Urea and electrolytes Normal NormalLiver function tests Normal NormalAntinuclear antibody Negative NegativeB12 and folate Normal Not testedVDRL Negative Not testedThyroid stimulating hormone (mU/l) 58.4 0.87Free T4 (pmol/l) 7.4 Not testedAntithyroid microsomal antibody titres 1:25600 1:1600Psychometric (normal/predicted range):

Folstein MMSE (>24) 25/30 25NART IQ 10th percentile 18th percentileWAIS-R (verbal) 13th percentile Not testedWAIS-R (performance) 27th percentile Not testedFAS verbal fluency (>30) 25 23Cognitive estimates test (15) 10/30 16/30Digit span forwards (>5) 7 6Rey-Osterreith complex figure (copy) (36) 25.5 24Rey-Osterreith complex figure (recall) ( 30%) Not tested 75%



http://jnnp.bmj.com

/J N

eurol Neurosurg P


nloaded from


falling over him”. His wife mentioned bizarre,useless movements of his left hand whichwere present from the beginning of thedisease.

On admission, he was awake, brady-phrenic, and partially collaborative. His con-versation was often disrupted by hallucina-tions. The aVect was sad and he had partialinsight for his mental dysfunction. He wasdisoriented for time, place, and situation. Hecould understand speech and was able to fol-low oral instructions involving two consecu-tive components. Naming was preserved.Prominent dysgraphia and dyscalculia werenoticed. Immediate recall and short termmemory were severely disturbed, whereaslong term memory, especially for personal lifeevents, was relatively spared. Abstract think-ing was severely aVected. Bimanual move-ments, such as clapping, were extremely diY-cult.

The cranial nerves were normal as wereocular fundi. The motor examination showednormal force. Deep reflexes were symmetricand plantar responses were flexor. The rightarm had a dystonic posture. His gait wasataxic on a wide base.

At times, the left arm would spontaneouslyrise in front of the patient during speaking orwhile using his right hand. He was unaware ofthese movements until they were brought tohis attention. When questioned about theirpurpose, the patient denied that they werevoluntary. No grasping of either hand or footwas found. The patient had no corticalsensory loss.

The laboratory data including blood chem-istry, haematology, and sedimentation ratewere normal, as were folic acid, vitamin B12concentrations, and thyroid function.Vene-real disease research laboratory and HIV testswere negative. The cerebrospinal fluid hadnormal content. Brain CT showed mildcerebral atrophy. An EEG showed severe dif-fuse slowing at admission. Within a week,repeated EEGs showed triphasic waves with aperiodic pattern of 1- 1.5 Hz.

During the next 2 weeks, the patient devel-oped myoclonic jerks. Severe dysphasia andcognitive decline were accompanied by con-fusion and aggression. He became grosslyataxic, and unable to walk and perform any ofhis daily activities even with help. Transferredto a chronic care hospital, he died few weekslater. Postmortem examination was not al-lowed.

This short fatal neurological disease mani-fested by fulminant dementia, myoclonicjerks, and extrapyramidal and cerebellar dys-function was strongly suggestive of CJD. Theperiodic EEG pattern reinforced this diagno-sis. Our patient’s alien hand was part of theotherwise characteristic clinical picture ofCJD, but it occurred early in the diseasecourse when no myoclonic jerks werepresent. We are aware of only one report ofalien hand in CJD. MacGowan et al8

described two patients with CJD with amyoclonic alien hand syndrome. In onepatient the left arm “was noted to havespontaneous movements which appearedpurposeful...wandered out of her view”. Inthe second, the alien limb performed com-plex actions such as unbuttoning her blouseand removing a hair pin. Although ourpatient had no myoclonus or pyramidal signswhen the alien hand appeared, in theirpatients it was associated with spontaneousor stimulus sensitive myoclonus, spastichemiparesis, and cortical sensory loss.

The literature seems to describe distinctforms of alien hand. All share the occurrenceof involuntary movements contrary to thepatient’s stated intent, but the types of move-ment diVer. In the callosal form, there arepurposeful movements of the non-dominanthand.9 In the frontal form, there is graspingand utilisation behaviour of the dominanthand.9 In the corticobasal degeneration, thereare aimless movements of either hand.5 7

When a consequence of tumorous or vascularpathology,9 alien hands can perform complexacts such as trying to tear clothes or undoingbuttons. The description by MacGowan et al8

has characteristics of the callosal form (espe-cially in patient 2). However, our casesuggests that the alien hand sign in CJD mayappear in a diVerent type, performing lesscomplex movements which resemble thosereported by Riley et al in corticobasaldegeneration.7 These authors described thealien limb as “ involuntarily rising and touch-ing the mouth and eyes” (patient 1). Thepatient thought that she “was powerless tostop this movement” and when directed tostop responded that “she can’t”. Anotherpatient’s left arm was at times “elevated infront of him”, while he was “unaware of thissituation until his attention was called to it”(patient 10).

Another related phenomenon coined as“arm levitation” was reported in progressivesupranuclear palsy. In these patients the arminvoluntarily raised and performed semi-purposeful movements.10

One common denominator between CJD,corticobasal degeneration, and progressivemultifocal leukoencephalopathy,6 in which analien hand sign has also been described, ismultifocality. In corticobasal degeneration, itwas proposed that more than one site isaVected or that a “release” phenomenonoccurs accounting for the aetiology of alienhand.7 In CJD, bilateral cortical damage tomotor areas might be the origin of their sub-sequent isolation and disconnection.

We suggest that CJD should be added tothe diVerential diagnosis of diseases present-ing with an alien hand with or without myo-clonus.

We are indebted to Professor Eran Zardel, Depart-ment of Physiology, University of California, LosAngeles, USA.

R INZELBERGP NISIPEANUS C BLUMENR L CARASSO

Department of Neurology, Hillel YaVe Medical Center,Hadera, Israel

Correspondence to: Dr Dr R Inzelberg, Depart-ment of Neurology, Hillel YaVe MedicalCenter, Hadera, 38100, Israelemail [email protected]

1 Brown P, Gibbs CJ, Rodgers-Johnson P, et al.Human spongiform encephalopathy: the Na-tional Institutes of Health series of 300 cases ofexperimentally transmitted disease. Ann Neurol1994;35:513–29.

2 Levine DN. The alien hand. In: Joseph AB,Young RR, eds. Movement disorders in neurologyand neuropsychiatry. Oxford: Blackwell, 1999:645–9.

3 Brion S, Jedynak CP. Troubles du transfertinterhemispherique. A propos de trois observa-tions de tumeurs du corps calleux. Le signe dela main etrangere. Rev Neurol 1972;126:257–66.

4 Bogen JE. The callosal syndromes. In: HeilmanKM, Valenstein E, eds. Clinical neuropsychology.2nd ed. New York: Oxford University Press,1985:295–338.

5 Doody RS, Jankovic J. The alien hand andrelated signs. J Neurol Neurosurg Psychiatry1992;55:806–10.

6 Berger JR, Concha M. Progressive multifocalleukoencephalopathy: the evolution of a diseaseonce considered rare. Journal of Neurovirology1995;1:5–18.

7 Riley DE, Lang AE, Lewis A, et al. Cortical-basal ganglionic degeneration. Neurology 1990;40:1203–12.

8 MacGowan DJL, Delanty N, Petito F, et al. Iso-lated myoclonic alien hand as the sole presen-tation of pathologically established Creutzfeldt-Jakob disease: a report of two patients. J NeurolNeurosurg Psychiatry 1997;63:404–7.

9 Feinberg TE, Schindler RJ, Gilson Flanagan N,et al. Two alien hand syndromes. Neurology1992;42:19–24.

10 Barclay CL, Bergeron C, Lang AE. Arm levita-tion in progressive supranuclear palsy. Neurol-ogy 1999;52:879–82.

Recurrent peripheral neuropathy in agirl with celiac disease

The involvement of the peripheral nervoussystem (PNS) in children with celiac diseaseis particularly rare. Furthermore, in bothchildren and adults with celiac disease,neurological complications are chronic andprogressive.1

We report on a 12 year old girl aVected byceliac disease, who on two separate occasionspresented with an acute peripheral neurologi-cal syndrome after accidental reintroductionof gluten in her diet.

This patient was born uneventfully tohealthy non-consanguineous parents with nofamily history of neurological or metabolicdiseases. At the age of 6 months she wasdiagnosed as having celiac disease accordingto the European Society of Paediatric Gastro-enterology and Nutrition (ESPGAN) crite-ria. Since then she was on a strict gluten freediet and was asymptomatic until the age of 10years when severe diarrhoea, vomiting, andabdominal pain manifested 6 days after theintake of corn flakes erroneously thought tobe gluten free. No previous infections hadbeen noticed. One week after the onset ofthese symptoms she experienced acute weak-ness and pins and needles sensation confinedto her legs. At that time her parents stoppedher intake of corn flakes on the suspicion thatthese were responsible for the symptoms.Despite this, symptoms worsened during thenext 2 days, confining her to bed.

At hospital admission, she was alert andmentally stable. Results of general physicalexamination were unremarkable. Neurologi-cal examination disclosed symmetric, pre-dominantly distal, weakness of the legs; theknee jerks and ankle reflexes were depressed;plantar reflexes were flexor. Distal stockingglove decreased in pin prick and temperaturewith sparing of propioception and lighttouch. Coordination tests were normal.

Laboratory investigations showed a whitecell count of 9300/mm3. The results of thefollowing investigations were within thenormal limits: haemogram, erythrocyte sedi-mentation rate, serum urea, nitrogen, electro-lytes, creatinine, glucose, transaminase, bi-lirubin, immunoglobulins (Igs), lead, iron,copper, urinalysis, urinary porphyrin, folicacid, and vitamins A, B1, B6, B12, and E. Anti-bodies to Campylobacter jejuni, neurotropicantivirus antibodies, specific and non-specificorgan autoantibodies, IgA and IgG antiglia-din antibodies (AGAs), IgA antiendomesiumantibodies (EMAs), and IgA antireticulumantibodies (ARA), assayed by enzyme linkedimmunoadsorbent assay (ELISA) and im-munofluorescence (IF) were also negative.Lumbar puncture was not performed. Anti-bodies against gangliosides GM1 and GQ1b,myelin associated glycoprotein and myelin



http://jnnp.bmj.com

/J N

eurol Neurosurg P


nloaded from


basic protein were not tested. Nerve conduc-tion studies were consistent with a predomi-nately motor demyelinating peripheral neu-ropathy (table). Her symptoms improvedspontaneously and she was discharged homeafter 2 weeks. For 2 years she was asympto-matic on a gluten free diet.

At the age of 12 she presented acutely withsevere abdominal pain 8 days after a weeklyintake of bread meant to be gluten free. Twoweeks later, due to persisting gastrointestinalsymptoms, her parents excluded the breadfrom her diet. After 2 further weeks, while theabdominal pain was gradually improving, shehad a new episode of acute weakness in thelower limbs and sensory abnormalities in-cluding burning paraesthesiae. On neurologi-cal examination the legs showed markeddiminution in muscle power; absent deeptendon reflexes, and a reduction in pain andtemperature; light touch, perception of posi-tion, and vibration were preserved. Walkingwas impaired and the patient was bedridden.Otherwise the examination was normal.

A haemogram showed white cell counts of9700/mm3. Laboratory investigations werewithin normal values as in the past. IgA andIgG AGA, IgA EMA, and IgA ARA assayedby ELISA and IF were again negative. Nerveconduction studies confirmed the presence ofa predominantly motor demyelinating neu-ropathy (table). The parents refused consentfor a lumbar puncture or nerve biopsy.

Over the next 2 weeks her neurological dis-abilities spontaneously improved until fullrecovery was complete. After 4 weeks, AGA,EMA, and ARA were still negative.

On her most recent admission, 1 year afterthe onset of her first neurological symptoms,she is still on a strict gluten free diet and hasno residual symptoms or signs.

The natural history of celiac disease is wellknown and the typical celiac enteropathy isoften associated with several other disorders.However, as celiac disease is a relatively com-mon and lifelong condition, it is likely thatsome of these associations may occur bychance.

This patient, who was diagnosed as havingfrank celiac disease at the age of 6 months,experienced two episodes of acute peripheralneuropathy, at the age of 10 and 12 years,respectively. Two major pieces of evidencestrongly support the assumption of a glutenderived disease: (1) the episodes occurred onboth occasions when gluten was accidentallyreintroduced in the diet; and (2) the responseto a gluten free diet was reasonably rapid,occurring within weeks.

The present case, however, diVers clinicallyfrom those with neurological involvement pre-viously reported. In the paediatric age group,

in fact, neurological complications of celiacdisease are rarely encountered and are mostlyconfined to the CNS2: to the best of ourknowledge, there are only two previouslyreported cases of PNS involvement in childrenwith celiac disease. In both cases, however,these were chronic axonal polyneuropathiespresenting during a gluten free diet.3 4

In both episodes in the present case neuro-physiology was strongly supportive of ademyelinating peripheral neuropathy, whichis most commonly attributed to a directimmune mediated attack to the myelin. Bycontrast, wallerian and axonal degenerationmay be caused by vasculitis, and nutritional,metabolic, and toxic factors.

An autoimmune pathogenesis in associ-ation with strong evidence of a geneticsusceptibility has been proposed for celiacdisease. Although it is well established thatAGA, EMA, and ARA are reliable indicatorsof sensitisation to gluten at least at the time ofdiagnosis, in the clinical practice at follow up,during a gluten challenge, pathological valuesof these antibodies may not be detected.5 Inthe present case the time course of the diseasemight be suggestive of an antibody mediatedresponse. However, we could not detectpathological concentrations of AGA, EMA,or ARA antibodies either during the course ofthe disease or at follow up.

It is known that in celiac disease manyimmunological perturbations can occur out-side the gastrointestinal tract. Crossing of theantigens through a damaged small intestinalmucosa, deposition of immune complexes intarget organs, a reduction in immune surveil-lance, mechanism of molecular mimicry, andactivated T cell response may contribute tothe pathogenesis of the diseases associatedwith celiac disease. Direct toxic eVects ofgliadin and vitamin deficiency are other pos-sible pathogenic mechanisms of damage tothe nervous system. Although we ruled out avitamin deficiency it is still questionablewhether a toxic neuropathy can be the case.

In conclusion, this case shows two majorissues: an acute polyneuropathy can be acomplication of celiac disease in childhoodand its benign course could help in theunderstanding of the underlying pathogenicmechanisms.

We are grateful to Professor Angela Vincent(Oxford) for her helpful suggestions in reviewing themanuscript.

AGATA POLIZZIMARIA FINOCCHIARO

ENRICO PARANOPIERO PAVONE

Division of Paediatric Neurology, Department ofPaediatrics, University of Catania

Catania, Italy

SALVATORE MUSUMECIDepartment of Paediatrics, University of Sassari,

Sassari, Italy

AGATA POLIZZINeurosciences Group, Institute of Molecular Medicine,

Department of Clinical Neurology, University ofOxford, Oxford, UK

Correspondence to: Dr Agata Polizzi, Division ofPaediatric Neurology, Department of Paediatrics,University of Catania, Viale A Doria 6, 95125Catania, Italy email: [email protected]

1 Cooke WT, Thomas Smith W. Neurologicaldisorders associated with adult coeliac disease.Brain 1966;89:683–722

2 Gobbi G, Bouquet F, Greco L, et al. Coeliacdisease, epilepsy and cerebral calcifications.Lancet 1992;340:439–43

3 Papadatou B, Di Capua M, Gambarara M, et al.Nervous system involvement in paediatric coe-liac patients. In: Mearin ML, Mulder CJJ, eds.Coeliac disease. Dordrecht: Kluwer Academic,1991:199–203.

4 Simonati A, Battistella PA, Guariso G, et al.Coeliac disease associated with peripheral neu-ropathy in a child: a case report. Neuropediatrics1998;29:155–8

5 Bottaro G, Sciacca A, Failla P, et al. Antigliadinantibodies in the various stages of coeliacdisease in children. Pediatr Med Chir 1988;10:409–13

Frontal release signs in older peoplewith peripheral vascular disease

A growing body of research examiningneurological aspects of clinically “silent” cer-ebrovascular disease suggests that neurologi-cal signs indicative of generalised organicbrain damage may occur in the absence ofcompleted stroke.1 These soft signs includeprimitive reflexes (frontal release signs), rep-resenting an anatomical and functional deaf-ferentation of cortical from subcortical struc-tures. Primitive reflexes are known to occur ina wide variety of dementias, includingAlzheimer’s disease2 and vascular dementia.3

It is likely that the presence of undetectedcerebrovascular disease accompanying pe-ripheral vascular disease is underestimated,as peripheral vascular disease is known to bea risk factor for transient ischaemic attacks. Astudy assessing 373 older patients withperipheral vascular disease found that 72 ofthe 144 patients who had not experienced atransient ischaemic attack or stroke werefound to have a degree of carotid stenosis ofbetween 60% and 99%.4

In the present study, the prevalence ofprimitive reflexes was examined in older peo-ple with peripheral vascular disease and anon-vascular control group. Independentpredictors of these reflexes were also exam-ined in peripheral vascular disease. Bothgroups were drawn from the same geographi-cal area. All were interviewed and examinedoutside hospital by myself. Interviewees werecommunity residents from the catchmentarea of an inner city London teaching hospi-tal.

Twenty five consecutive non-amputees onthe waiting list for femoropopliteal bypassoperation were compared with 25 postopera-tive patients who had undergone elective hipor knee replacement and a period of inpatientrehabilitation. All participants were aged 65and over at the time of interview. Patientswith peripheral vascular disease all had clini-cal and Doppler proved evidence of periph-eral ischaemia. Controls were interviewedbetween 6 months and 1 year after theiroperation. Both groups had no history ofstroke or transient ischaemic attack.

A more detailed description of instrumentsis provided elsewhere.1 All subjects were

Electrophysiological study suggestive in both episodes of an acute demyelinating peripheral neuropathyconfined to the lower limbs. Values were within normal limits in the upper limbs

1st Episode 2nd Episode

Peroneal LR TibialL

R PeronealL

R TibialL

R

MCV (ms) 26 27 22 2424 28 20 23

DL (ms) 7.3 8.0 7.2 8.87.5 8.4 7.0 9.0

F wave latency (ms) 70 72 83 84CMAP (µV) 3 2.7

Sural LR SuralL

R

SCV (ms) 38 4042 41

AMP (µV) 16.2 17.416.8 18

MVC=motor conduction velocity; DL=distal latency; CMAP=compound motor action potential;SCV=sensory conduction velocity; AMP=amplitude; L=left; R=right.



http://jnnp.bmj.com

/J N

eurol Neurosurg P


nloaded from


examined using a rating scale for theexamination of frontal release signs (FRSS),with nine operationally defined items, eachon a seven point semiquantitative scale. Thenine reflexes were paratonia and palmomen-tal, hand grasp, foot grasp, glabellar, rooting,snout, and visual/tactile sucking reflexes.Neuropsychological measures included theassessment of frontal lobe function (trailmak-ing tests A and B, behavioural dyscontrolscale, and the controlled word associationtest) and generalised cognitive impairment(CAMCOG). Depression was assessed usingthe Hamilton rating scale for depression, 15item geriatric depression scale, and diagnos-tic criteria for DSM IV major depressive dis-order. Family history of depression, wish todie, and suicidal ideation within the past yearwere also recorded, as were blood pressureand a checklist for chronic physical illness.

Total FRSS scores and scores on FRSSsubscales were compared between groupsusing the Mann-Whitney U test for inde-pendent samples. In the peripheral vasculardisease group, a correlation matrix for totalFRSS score against DSMIV depression,CAMCOG score, behavioural dyscontrolscale score, verbal fluency score (totalnumber of words beginning with F, A, and S)and trailmaking test times was examinedusing the Spearman correlation coeYcient,controlling for age, sex, blood pressure, andchronic physical illness. Behavioural dyscon-trol scale scores, trailmaking A/B test times,and verbal fluency scores were first convertedinto binary variables according to whetherthey were at/above or below the median valuefor the group. CAMCOG score was dividedinto subjects scoring 69 or above or less than69. Those associations with a two tailedsignificance of 0.1 or less were then enteredinto a linear regression equation using thestepwise method.

Patients with peripheral vascular diseasehad a higher mean score on the frontal releasesigns scale than controls (5.8 (SD 4.6) v 1.7(SD 1.9)) (Mann-Whitney U=144.500,Z=−3.33, two tailed p=001), as well as onglabellar and rooting reflexes (table). Onlyone variable (trailmaking B test time) wasentered into the equation; this accounted for23% of the variance in FRSS score (B=4.6,95% confidence interval (95% CI) (B)1.3–8.0, p=0.01).

In peripheral vascular disease, there is lim-ited information available concerning theintellectual and neurological sequelae ofcoexisting cerebrovascular disease. Phillips etal found greater impairment in psychomotorspeed and abstract reasoning in patients withperipheral vascular disease than age/sexmatched controls, with less significant diVer-ences between the groups in verbal fluency,concentration, abstract thought, perception,and constructional skills.5 Another study bythe same group found poorer performance inpatients with peripheral vascular disease thancontrols on visual memory, trailmaking Btest, and visuospatial skills. Patients withperipheral vascular disease were also equallyimpaired in these areas compared with amatched group of stroke patients.6

Small numbers of patients, which may alsohave obscured other significant findingsbetween the two groups, limit the presentstudy. However, there is some evidence thatclinically relevant cerebrovascular diseasemay accompany peripheral vascular diseaseand that concomitant disruption of frontal/subcortical brain function may not presentwith hard neurological signs. As it is possiblethat silent brain infarction was present inpatients with peripheral vascular disease, fur-ther studies incorporating brain imaging arerequired before there can be a clearer under-standing of the relation between peripheraland central vascular pathology.

I thank Dr Robert Howard for supervision of thisstudy and Professor Stephen Jackson and Mr PaulBaskerville for allowing me to interview patientsunder their care. The study was carried out as partof a University of London MD thesis.

RAHUL RAODepartment of Old Age Psychiatry, Maudsley Hospital

and Institute of Psychiatry, London

Correspondence to: Dr Rahul Rao, Department ofOld Age Psychiatry, Guy’s, King’s, and St ThomasMedical School, Job Ward, Thomas Guy House,Guy’s Hospital, St Thomas Street, London SE19RT, UK email [email protected]

1 Rao R, Jackson S, Howard R. Primitive reflexesin cerebrovascular disease: a community studyof older people with stroke and carotid stenosis.International Journal of Geriatric Psychiatry (inpress).

2 Burns A, Jacoby R, Levy R. Neurological signsin Alzheimer’s disease. Age Ageing 1991;20:45–51.

3 Vreeling FW, Houx PJ, Jolles J, et al. Primitivereflexes in Alzheimer’s disease and vasculardementia. Journal of Geriatric Psychiatry andNeurology 1995;8:111–17.

4 Alexandrova NA, Gibson WC, Norris JW, et al.Carotid artery stenosis in peripheral vasculardisease. J Vasc Surg 1996;23:645–9.

5 Phillips NA, Mate-Kole CC, Kirby RL. Neuro-psychological function in peripheral vasculardisease amputee patients. Arch Phys Med Reha-bil 1993;74:1309–14.

6 Phillips NA, Mate-Kole C. Cognitive deficits inperipheral vascular disease. A comparison ofmild stroke patients and normal controlsubjects. Stroke 1997;28:777–84.

Factitious clock drawing andconstructional apraxia

A 45 year old man presented with a 1 dayhistory of headache, possible seizures, andleft sided weakness. On the day of presenta-tion the patient’s wife had twice found him,inexplicably, on the floor. After the secondsuch episode she brought him to hospital forevaluation. Examination disclosed a com-plete left hemiplegia and hemianaesthesia,although muscle tone was documented to benormal and the plantar responses downgoingbilaterally. Brain CT was normal and routineblood examination was unremarkable. Therewere no further seizure-like episodes andthe patient was transferred to this hospital10 days later, hemiplegia unchanged, forpossible angiography and further investiga-tions.

He was an exsmoker with hypercholestero-laemia and peripheral vascular disease whichhad been treated by a left femoral angioplasty

5 years earlier. The angioplasty was compli-cated by the occurrence of a seizure, thoughtto be related to dye injection, and phenytoinhad been prescribed for a short time thereaf-ter. There was a remote history of heavy alco-hol use, but he had been abstinent for severalyears. His father had had a stroke at the age of65.

Six months earlier the patient had also col-lapsed at home and been taken to hospitalwith a left hemiplegia. Brain CT at that timewas normal, as were carotid Doppler studiesand an echocardiogram. During that admis-sion to hospital, several generalised seizure-like episodes were seen, some with retainedconsciousness, and he had again been startedon phenytoin therapy. A follow up outpatientbrain MRI was normal and it was concludedthat the hemiplegia was non-organic inorigin. He was described to have made agradual, near complete, recovery from thisfirst hemiplegic episode and was scheduledfor an imminent return to work at the time ofhis relapse.

On transfer to this hospital the patient wasalert, oriented, and cooperative. Although upto date on current aVairs and able to describethe investigations performed at the transfer-ring hospital, he scored only 23/30 on a minimental state examination, with absent threeword recall, impaired registration, and poorcopying of a two dimensional line drawing.Further bedside neuropsychological testingshowed other findings indicative of construc-tional apraxia and left hemineglect. Specifi-cally, when asked to draw a clock with thetime at 10 minutes to 2 o’clock, all the num-bers, and the clockhands, were placed on theright hand side of the clock outline (figure A).Copying of three dimensional line drawingswas also significantly impaired (figure B).When asked to bisect a line, however, thepatient did so only minimally to the right ofthe midpoint (58% of the distance from theleft side).

Cranial nerve examination suggested anincongruent and inconsistent left hemianop-sia to confrontation testing but was otherwisenormal, including bilaterally symmetric op-tokinetic nystagmus. Motor examinationshowed paralysis of the left arm and leg, withbilaterally symmetric bulk, tone, and deeptendon reflexes. The plantar response wasflexor bilaterally. Sensory examinationshowed decreased pinprick and absent lighttouch, joint position sense, and vibrationsense on the entire left side. There was alsoimpaired perception of a tuning fork’svibration on the left side of the forehead, witha distinct demarcation in the midline. Therest of the physical examination was unre-markable.

Brain CT and MRI, CSF examination, androutine EEG were normal. Routine haemato-logical and metabolic analyses plus erythro-cyte sedimentation rate, serum lactate, pro-thrombin time/partial thromboplastin time,fasting serum glucose, HbA1c, serum Ig sur-vey, and thyroid stimulating hormone were allwithin normal limits. A hypercoagulabilityprofile was negative. A lipid profile showedmild hyperlipidaemia with increased low

Table 1 Primitive reflexes in patients with peripheral vascular disease (n=25) and controls (n=25)

Hand grasp Foot grasp Glabellar Palmomental Paratonia Rooting Snout Sucking (tactile) Sucking (visual)

U 274.0 312.5 199.5 287.5 287.0 235.5 287.5 261.0 287.5pValue 0.15 1.0 0.001* 0.15 0.29 0.01* 0.44 0.08 0.30

*Higher mean score in people with peripheral vascular disease.



http://jnnp.bmj.com

/J N

eurol Neurosurg P


nloaded from


density lipoprotein (3.92 mmol/l) and triglyc-erides (4.30 mmol/l) and low high densitylipoprotein (0.73 mmol/l). Serum phenytoinconcentration was therapeutic at 74 µmol/l.An ECG was normal.

Ophthalmological consultation and formalvisual field testing demonstrated a concentri-cally constricted field of mild degree in theright eye and tunnel vision in the left eye.

The patient consented to overnight video-EEG monitoring and was seen on multipleoccasions to move his left arm and/or leg in anormal fashion, at one point using the leftarm to readjust his bed covers shortly afterarousal from sleep, before glancing briefly atthe video camera and completing the taskwith his right arm. The prolonged EEG wasnormal.

A formal neuropsychological assessmentperformed in hospital documented impairedattention, concentration, and workingmemory, as well as several atypical calcula-tion and spelling errors, the second involvingunusual “near miss” letter substitutions orreversals (for example, “anixety”, “excecu-tive”). The formal testing identified noconsistent evidence of visuospatial deficits orconstructional apraxia. The findings wereinterpreted as inconsistent with the patient’shistory but the possibility of a factitious aeti-ology was not specifically addressed—that is,tests designed to detect malingering duringneuropsychological testing1 2 were not admin-istered by the examiner, who had not beeninformed at the time of consultation of thepresumptive neurological diagnosis of malin-gering or factitious disorder.

No further investigations were performedand the patient was transferred via the origi-nal hospital to a rehabilitation facility andsubsequently discharged to home. Con-fronted with the findings of the videomonitoring the patient appeared sanguineand accepting of the evidence that he shouldbe able to move his left side. Six months laterhe was ambulatory but otherwise not signifi-cantly improved. He had been assessed by apsychiatrist but had refused psychiatric fol-low up, electing instead to be followed up bya psychologist. He understood his diagnosisto be “conversion disorder” and reported thathe was actively collecting information on thesubject via the internet.

Outpatient brain SPECT and visual andsomatosensory evoked potentials performed1 year after discharge demonstrated no hemi-spheric abnormalities. The patient remained

oV work and was receiving disability funding.He walked with a limp favouring his left sideand complained of persistent decreasedsensation on the left side. Forced choice sen-sory testing of finger and arm movement onthe left3 demonstrated performance to beworse than chance (68% wrong choices).Motor bulk, tone, and reflexes were symmet-ric and plantar responses downgoing. Hedrew a clock normally at the 1 year follow up.

The clinical and laboratory findings de-scribed above indicate beyond any doubt thenon-organic nature of this patient’s lefthemiplegia/hemianaesthesia. His seizure-likeepisodes at presentation are presumed tohave been non-epileptic in origin (as hadbeen suspected during his previous admissionto hospital) although this cannot be defini-tively proved.

The inability to copy line drawings or todraw a clock is, from a neurologist’s perspec-tive, typically associated with parietal lobedysfunction, usually of the non-dominanthemisphere, especially if associated with lefthemispatial neglect.4 To our knowledge, thisis the first reported case of factitious clockdrawing and constructional apraxia. Bedsidemental status testing also demonstrated themore common simulated deficits of impairedattention and absent three word recall.1 Inretrospect, the severe neglect on clockdrawing was perhaps “too good to be true”,especially in the light of the near normal linebisection demonstrated on the same day. Themirror image distortion of the house was alsovery unusual and, furthermore, the mirrorreversal itself is evidence of lack of clinicalneglect. The distortion of the cube, however,could easily be misinterpreted as evidence oforganic constructional impairment if seen inthe absence of the other relevent clinical andlaboratory information.

During follow up, the patient admitted tofeeling tremendous occupation relatedstresses, and described how he had come toboth fear and detest his job. Given the clearbenefit to the patient of removal from hiswork environment, the relapse of his symp-tomatology just as he was scheduled forreturn to work after his first non-organichemiplegic episode, and the intentionalityrequired to feign poor clock drawing andconstructional apraxia, there is much to sup-port a diagnosis of malingering.5

Nevertheless, classification as a factitious dis-order is at least as justifiable in view of the

patient’s willingness to undergo medicalinvestigations, including video monitoring.

It is unclear how or when the patientacquired the information needed to mimic aconstructional apraxia. Previous bedsideneuropsychological evaluations may haveserved to familiarise him with the format ofsuch testing, acting as an impetus to researchthe issue of stroke and focal brain deficits(which might also have occurred after hisfather’s stroke), much in the same way he isnow researching conversion disorder, therebydiscovering what expected answers shouldlook like. Despite repeated questioning, how-ever, no evidence could be gathered from thepatient to support this speculation.

I KHANI FAYAZ

Division of Neurology

J RIDGLEYDivision of Neuropsychology

R WENNBERGDepartment of Medicine, Division of Neurology, TheToronto Hospital, University of Toronto, Toronto, ON,

Canada

Correspondence to: Dr R Wennberg, EC8–022,The Toronto Hospital, 399 Bathurst Street, To-ronto, Ontario, Canada M5T 2S8. Telephone 001416 603 5402; fax 001 416 603 5768.

1 Bernard L, Houston W, Natoli L. Malingeringon neuropsychological memory tests: potentialobjective indicators. J Clin Psychol 1993;49:45–53.

2 Prigatano G, Amin K. Digit memory test:unequivocal cerebral dysfunction and sus-pected malingering. J Clin Exp Neuropsychol1993;15:537–46.

3 Pankratz L, Binder L, Wilcox L. Evaluation ofan exaggerated somatosensory deficit withsymptom validity testing. Arch Neurol 1987;44:798.

4 Strub R, Black W. Constructional ability. In:Strub R, Black W. The mental status examinationin neurology. Philadelphia: FA Davis, 1985:101–23.

5 American Psychiatric Association. Diagnosticand statistical manual of mental disorders. 4th ed,revised. Washington: American PsychiatricAssociation, 1994.

Anosognosia and mania associated withright thalamic haemorrhage

Both anosognosia and secondary mania areassociated with right hemispheric lesions.These two non-dominant syndromes, how-ever, are rarely described as occurringtogether. We present a patient with a rightthalamic haemorrhage giving rise to pro-found denial of hemiplegia and elated mood.This case suggests mechanisms for thecommon production of mania and anosogno-sia.

A 53 year old, right handed, black man,with a history of alcohol misuse and depend-ence and untreated hypertension, wasbrought to the emergency room a few hoursafter developing an intense headache and leftsided numbness and weakness.

On admission he was described as “bellig-erent,” “agitated,” and “confused.” Bloodpressure was 240/160. Neurological exam-ination disclosed left lower facial droop,decreased left corneal and gag reflexes, andleft hemiparesis with dense sensory deficits.With increasing obtundation, the patient wastransferred to the intensive care unit andintubated. Brain MRI showed a large, rightsided, hyperacute thalamic bleed with masseVect and oedema. The patient was extu-bated 2 days later and 4 days after the strokehe was described as being drowsy andinattentive, but was able to answer questions

(A) Asked to: “put all the numbers on a clock and make the time ten to two”. (B) Patient’s copies (atright) of three dimensional line drawings. Top: common constructional distortion of cube. Bottom:unusual mirror image representation of house.

A B



http://jnnp.bmj.com

/J N

eurol Neurosurg P


nloaded from


appropriately. Neurological examinationshowed contralateral gaze preference, supra-nuclear vertical gaze palsy, diYculty converg-ing, left sided flaccid hemiparesis, and dense,left sided hemianaesthesia. Deep tendonreflexes were absent on the left and Babinski’sreflex was present on the left. In addition,visual extinction and neglect were present.

At the time of onset of right sided weaknessthe patient insisted that he was “fine,” and anambulance was called over his objections.After being extubated, the patient acknowl-edged that he had had a stroke, but, despitehis hemiparesis, insisted that he was ready togo home and go back to work. His belief in hisability to walk led to near falls, and he wasmoved to a room nearer to the nurses’ stationfor closer observation. He told the nurses thatsomeone else’s arm was in his bed. On oneoccasion, holding up his left arm with hisright, he told the nurse to, “take it away; itkeeps scratching me.” That the left arm“smelled funny” was another reason hewanted the nurses to take it away.

Four weeks after the stroke he firstacknowledged that his left arm belonged tohim. He spontaneously recalled believingotherwise. By this time he had a moderatehemiplegia and recognised “a little weak-ness,” but continued to insist that he was welland able to return to work. By the 6th weekafter stroke the patient more consistentlyacknowledged that he was weak on the leftside of his body. A request for disabled hous-ing “so that I won’t be a burden to my family”seemed to indicate an appreciation of hisimpairment, but this insight was fleeting;within an hour of making such statements thepatient might insist that after a week’sexercise he would be ready to return to work.His awareness of his hemiplegia fluctuatedfor 8 weeks after stroke before becomingfixed, but remained shallow after 12 weeks; heno longer planned to return to work andapplied for social security disability insurance“because they say I’m disabled.”

The patient’s mood was remarkably cheer-ful and optimistic. A week after the stroke hewas noted to praise extravagantly the hospitalfood, and the nurses found him “talkative.”When he arrived on our ward 11 days afterthe stroke he was flirtatious with female staVand boasted of having fathered 64 children.His girlfriend was surprised when he kissedher in front of the staV because he had neverpublicly displayed aVection before. He re-ported excellent energy and expansivelyinvited all of the staV to his home for thanks-giving. Sleep was not disrupted or reducedand he had a good appetite. When beginningto acknowledge his left sided weakness, heremained blissfully unconcerned. He scored31 points on a mania rating scale,1 which waswell in the manic range. The mania resolvedgradually over a 10 week period after stroke.

Other than alcoholism, the patient had nohistory of psychiatric illness and there was nofamily history of psychiatric illness. Thepatient had not seen an physician in manyyears. Visual acuity was found to be reducedto 20/800 in both eyes on the basis of hyper-tensive retinopathy.

Evaluation 1 month after stroke showedmany deficits and a few strengths. Inattentionto the left hemispace was marked. By 2months after stroke he no longer extinguishedto double simultaneous stimulation, but,although he could see to the left, was stillmissing targets in his left visual hemifield.Visual integration, both with and without therequirement of construction, was severely

impaired. He was able to correctly recogniseand produce facial emotional information.Simple attention was intact, but attentionalcontrol (backward span and mental control)was impaired. Visuomotor tracking was slowand he had significant problems with concep-tual shifting (both auditory and visual). Lan-guage processing diYculties included verypoor reading ability, impaired confrontationnaming, and impaired performance on a ver-bal task of fluency and initiation. Auditorycomprehension was mildly impaired. Vo-cabulary scored formally in the borderlineimpaired range, as did abstract verbal reason-ing. On tests of praxis he demonstrated a ten-dency to use the hand as object. Memoryperformances were relatively intact. Hisinitial recall of two paragraphs scored for-mally within the low average range and after a30 minute delay, he was able to recall most ofthe information initially encoded, scoringformally within the average range.

Structural brain MRI on admission to theemergency room showed a large right tha-lamic hemorrhage with mass eVect andoedema, with oedema extending into the cer-ebral peduncle with internal susceptibilityconsistent with deoxyhaemoglobin. Alsopresent was increased T2 signal bilaterally infrontal areas consistent with ischaemicchanges. Brain CT 30 days after strokeshowed, in addition to the thalamic lesion,moderate cerebellar atrophy and mild tomoderate prominence of the frontal corticalsulci compatible with cerebral atrophy.

Structural MRI performed 44 days afterthe stroke showed a 2 cm right thalamic hae-matoma. Functional MRI2 performed thesame day demonstrated a 2 cm area of absentcerebral blood volume at the posteriormargin of the right thalamus without any evi-dence of decreased cerebral blood volumewithin the right parietal, frontal, or temporalcortex.

This is a case of anosognosia of hemiplegiaand mania co-occurring in a patient with alarge right thalamic haemorrhage. Althoughanosognosia and mania are not generallythought of as occurring together, whenBabinski3 introduced the term anosognosiahe used as one of his examples a case in whichthe patient, though not confused, was “a littleoverexcited,” and in a later paper he pre-sented a case4 in which there was “a certainagitation, which expresses itself by exagger-ated loquacity, a decrease in attention, and atendency to erotic ideas.” Weinstein andKahn5 noted that euphoria was common inpatients with anosognosia. Moreover, al-though Cutting6 emphasised that apathy isthe mood more usually associated withanosognosia, 10% of his patients with ano-sognosia were described as having “euphoricmood.”

Right sided thalamic lesions are known toproduce both anosognosia and mania, but therelation of each to the pathology is unclear.Only some of the patients with right hemi-spheric lesions are manic or agnosic. Thesetwo syndromes may be related to dysfunctionof diVerent neural networks and only occurtogether when a disease process aVects bothnetworks.

Another possibility is that these syndromesare aetiologically related. Could anosognosiabe a manifestation of mania? Although it iseasy to conceive how elevated mood mightfacilitate anosognosia of hemiplegia (or othertypes of anosognosia), it is diYcult to explainthe presence of denial of ownership anddislike of the left arm (other anosognosic

phenomena) on the basis of euphoria.Moreover, Starkstein et al,7 finding that simi-lar frequencies and severities of major andminor depression were present in patientswith and without anosognosia, suggest that aparticular mood state may not necessarilyinfluence insight.

Several explanations have been proposedto explain the phenomenon of anosognosia.8 9

All the models invoke dysfunction of the cer-ebral cortex, especially the parietal cortex. Itis interesting that in this case functional MRIfailed to demonstrate decreased CBV in theparietal lobe.

In summary, we present a case of maniaaccompanying anosognosia in a patient with aright thalamic haemorrhage. The coexistenceof mania and anosognosia may be more com-mon than previously appreciated. The associ-ation with anosognosia implies that themechanisms implicated in the pathogenesisof secondary mania may be similar to those ofanosognosia. The absence of evidence ofabnormal parietal, temporal, or frontal lobefunction by functional MRI in this case isintriguing.

ELIZABETH LIEBSONDepartment of Psychiatry, Tufts, New England

Medical Center, 750 Washington Street, Box 1007,Boston, MA 02111, USA. Telephone 001 617 636

1633; email [email protected]

1 Bech P, Kastrup M, Rafaelsen OJ. Mini-compendium of rating scales for states of anxi-ety, depression, mania, schizophrenia, withcorresponding DSM-III syndromes: maniascale. Acta Psychiatr Scand 1986;73(suppl236):29–31.

2 Belliveau J. Functional cerebral imaging by sus-ceptibility contrast NMR. Magn Reson Med1990;14:538–546.

3 Babinski J. Contribution a l’etude des troublesmentaux dans l’hemeplegie organique cereb-rale. Rev Neurol 1914;22:845–8.

4 Joltrain E. Un nouveau cas d’anosognosie. RevNeurol 1924:638–40.

5 Weinstein EA, Kahn RL. The syndrome of ano-sognosia. Arch Neurol Psychiatry 1950;64:772–91.

6 Cutting J. Study of anosognosia. J Neurol Neuro-surg Psychiatry 1978;41:548–55.

7 Starkstein SE, FedoroV JP, Price TR, et al. Ano-sognosia in patients with cerebrovascularlesions: a study of causative factors. Stroke1992;23:1446–53.

8 Ellis SJ, Small M. Denial of illness in stroke.Stroke 1993;24:757–9.

9 Heilman KM. Anosognosia: possible neuro-psychological mechanisms. In: Prigatano GP,Schacter DL, eds. Awareness of deficit after braininjury. New York: Oxford University Press,1991:53–62.

Epileptic cardiac asystole

A patient is reported on with habitualepisodes of collapse and loss of consciousnessassociated with EEG evidence of focalepileptiform discharges. Simultaneous ECGrecordings disclosed 25 seconds of cardiacventricular asystole occurring 24 secondsafter the onset of electrical seizure activity.After changes to antiepileptic medication andthe insertion of a permanent cardiac pace-maker he has had no further episodes. Incases of epileptic cardiac dysrhythmia, iso-lated EEG or ECG recording may proveinsuYcient and prolonged simultaneousEEG/ECG monitoring may be required.

Cardiac arrhythmias subsequent to epilep-tic seizures have been recognised for morethan 90 years. They provoke diagnosticconfusion and may be a mechanism ofsudden unexplained death in epilepsy.Whereas sinus tachycardia was noted toaccompany more than 90% of epilepticseizures, isolated bradycardia was seen much



http://jnnp.bmj.com

/J N

eurol Neurosurg P


nloaded from


less commonly (only 1 of 74 seizuresrecorded).1 A review in 1996 of the “ictalbradycardia syndrome” showed only 15documented cases in the literature of eitherbradycardia or asystole associated withseizures.2 Most patients had temporal lobeseizures. The longest duration of asystolepreviously reported is in a 17 year old manwith temporal lobe epilepsy who sustained a22 second pause in cardiac output. More

typically the asystolic periods in documentedcases are in the region of 5–10 seconds.2

Shorter duration asystole may not compro-mise cerebral function suYciently to causeloss of consciousness. Implantation of acardiac pacemaker is advocated but does notensure that lapses of consciousness are elimi-nated if these are directly related to theseizure rather than to the secondary asystole.We report on a patient with epileptic cardiac

asystole of 25 seconds duration demonstratedby prolonged simultaneous EEG/ECG moni-toring which responded well to pacemakerinsertion.

A previously well 34 year old right handedbuilder was referred with a 1 year history offortnightly episodes of loss of consciousness.There was no associated warning, aura, chestpain, or palpitations and the patient was onlyaware of the episode once consciousness was

16 Channel ictal EEG (eight channels illustrated with ECG) showing electrographic seizure onset and subsequent bradycardia and asystole.

Gain

Paper speed : 1.34 cm/s

Filter setting : Lf : 35 Hz

3

4

T5 T6

2

1

A

F8-T4

T4-T6

F7-T3

T3-T5

T4-C4

C4-Cz

Cz-C3

C3-T3

EKG

Time : 07:04:34

BF8-T4

T4-T6

F7-T3

T3-T5

T4-C4

C4-Cz

Cz-C3

C3-T3

EKG

Time : 07:04:58

C

F8-T4

T4-T6

F7-T3

T3-T5

T4-C4

C4-Cz

Cz-C3

C3-T3

EKG

Time : 07:05:21

T4C4CZC3T3

8 7 6 5

F8F7



http://jnnp.bmj.com

/J N

eurol Neurosurg P


nloaded from


restored and he found himself lying on thefloor. On recovery there was no confusion,drowsiness, dysphasia, or diuresis. Often,however, he sustained soft tissue injuries tohis face and scalp.

Witnesses reported that the patient would,without warning, suddenly collapse to theground where he would remain unrousable,inaccessible, and motionless for 90 to 120seconds. On two occasions he appeared con-fused and disorientated immediately before acollapse. During the period of unconscious-ness he would demonstrate no involuntarymovements, orofacial automatisms, or cyano-sis but he would become pale and “ashen”while staring straight ahead with a glazedlook. On resolution of the episode his colourwould return to normal and within 2 minuteshe would have fully recovered. Unusuallyduring one reported episode of unconscious-ness he was seen to briefly extend the fingersof both hands.

He was admitted to his local hospital andCT, MRI, interictal EEG, and 24 hour ECGwere normal. No episodes were witnessedwhile he was an inpatient but they werethought to be epileptic in origin and thereforehe was started on phenytoin, with no benefit.Carbamazepine was added, again with mini-mal eVect.

The patient was then referred to theEpilepsy Assessment Centre of The NationalSociety for Epilepsy and National Hospitalfor Neurology and Neurosurgery for furtherinvestigation and management.

Cardiovascular and neurological examina-tion was normal, as were MRI and routineinterictal EEG. Sixteen channel ambulatoryEEG using an Oxford Instruments digitalEEG receiver was performed continuously for340 hours before an episode was captured.Interictally rare spikes were seen over the rightfrontocentrotemporal region during sleep. Theonset of the episode was not witnessed and thepatient was found lying on the floor, regainingconsciousness at about 07:06. The event EEGshowed a short run of bilateral semirhythmic2–3 Hz activity at 07:04:34 (figure A), persist-ing for 8 seconds before being obscured bymuscle and movement artefact. Twenty fourseconds after the first EEG change, at07:04:58, the ECG changed from sinusrhythm at 90 bpm to a brief period of sinusbradycardia, followed by a period of asystolewith only very occasional ventricular com-plexes lasting 25–30 seconds (figure B). After afew seconds of bradycardia then tachycardia,sinus rhythm was restored. Throughout theepisode the QT interval on the ECG remainedwithin normal limits. The EEG became visibleagain 16 seconds into the asystolic period, atwhich time it was dominated by diVuse lowamplitude slow activity at 41)) nerves. There weremoderate decreases in the amplitude of com-pound action potentials in all the nervestested, and an amplitude reduction of 50%was detected across the cubital tunnel of theright ulnar nerve. Minimum F wave latencieswere prolonged in all the nerves tested. Thelatency in the right phrenic nerve was slightly



http://jnnp.bmj.com

/J N

eurol Neurosurg P


nloaded from


delayed (8.7 ms (normal

venous thrombosis is often asymptomatic, orpresents with non-specific pain, it is probablyunrecognised in many cases.2 Concurrentipsilateral spinal accessory neuropathy andinternal jugular venous thrombosis after CEAis expected to be rare, and this is underscoredby the lack of published cases. Despite thisapparent rarity, a common pathogeneticmechanism for postoperative spinal accessoryneuropathy and internal jugular venousthrombosis may well be present, at least insome cases, which may lead to the considera-tion of the possibility of both when either isdiscovered.

We report on a patient who developed rightspinal accessory neuropathy and internaljugular venous thrombosis after right CEA.

A 59 year old man underwent right CEAfor possibly symptomatic stenosis. Angio-graphy had shown 90% stenosis of the rightinternal carotid. The operation was doneunder general anaesthesia. The carotid bifur-cation was unusually distal, necessitating along dissection and high retraction. Noimmediate postoperative complications wereevident. The next day, the patient complainedof mild pain at the operative site, but he didnot notice any weakness. The pain spreadinto his right shoulder within several days; atthat time, he also noted diYculty raising hisright arm. His symptoms worsened further afew weeks later. The symptoms persisted, andhe presented for neurological evaluation 4months after CEA. At that time, he had someinduration along the incision site and apalpable cord within the right supraclavicularfossa. There was moderate atrophy of theright sternocleidomastoid and trapezius, withright shoulder drooping and minor rightscapular winging. Right arm abduction pro-duced more prominent scapular winging andwas limited to 90 degrees due to pain andweakness. Electrodiagnostic studies wereconsistent with partial right accessory neu-ropathy with minor denervation of the righttrapezius. Cervical ultrasonography and MRIdemonstrated right internal jugular venousthrombosis. The patient was treated with ashoulder support, analgesics, and low doseaspirin. There was no significant clinicalchange 1 year after CEA. Repeat electrodiag-nostic studies were consistent with severechronic right spinal accessory neuropathy,and repeat ultrasonography showed persist-ent right internal jugular venous thrombosis.

Spinal accessory neuropathy was firstreported as a complication of CEA in 1982.3

Since then, there have been several casereports and small series.1 4 A 1996 review ofreports of cranial neuropathy after CEAdisclosed only one patient with spinal acces-sory neuropathy in over 3000 cases.1 Al-though the authors did not include severalother reports3 4 which, taken together, mayseem to suggest a somewhat higher incidence,the overall small number of reported cases inproportion to the hundreds of thousands ofCEAs that have been done worldwidesuggests that clinically significant spinalaccessory neuropathy is a rare complication.Minor or transient spinal accessory neu-ropathy after CEA may be more frequent.The cause of spinal accessory neuropathyafter CEA is usually not well established, butintraoperative nerve stretching or compres-sion from retraction is most often invoked.4

Delayed onset (after 3 weeks) has been notedin some; for these patients, postoperativeinflammation and scarring seem more likelycauses. Spinal accessory nerve transection orischaemia/infarction (arterial or venous) are

other possibilities. As in our patient, highcarotid dissection and retraction have beenreported to precede spinal accessoryneuropathy.1 4

The spinal accessory nerve courses alongthe internal jugular vein and near the internalcarotid artery, typically well above the carotidbifurcation. It stands to reason that a highincision and retraction resulting from a highcarotid bifurcation would place the nerve atrisk. Whether this realisation may lead to anytechnical modification to decrease the risk ofspinal accessory neuropathy in those with ahigh bifurcation is unclear.

From our search, internal jugular venousthrombosis after CEA has been reported inonly one case.2 As Southcott et al noted,retraction of the internal jugular during CEAmay cause complete occlusion, leading tothrombosis from venous stasis or endothelialinjury. Other causes of internal jugularvenous thrombosis include jugular cannula-tion, blunt cervical trauma, and a hyperco-agulable state. Internal jugular venous throm-bosis may occur within a week after neckdissection, often with recanalisation after sev-eral months.5

The presence of induration about the inci-sion site and a palpable supraclavicular cordin our patient led us to suspect venousthrombosis. Internal jugular venous throm-bosis may often be asymptomatic. Potentialsymptoms of internal jugular venous throm-bosis include headache

Documents

BMJ - LETTERS TO THE EDITORLETTERS TO THE EDITOR Postictal psychosis related regional cerebral hyperperfusion Postictal psychosis is a known complication of complex partial seizure