focal (segmental) dyshidrosis in syringomyelia · onset of current dyshidrosis and duration from...

SHORT REPORT

Focal (segmental) dyshidrosis in syringomyelia

Kazumasa Sudo, Naoto Fujiki, Sachiko Tsuji, Minoru Ajiki, Takuya Higashi,Masaaki Niino, Seiji Kikuchi, Fumio Moriwaka, Kunio Tashiro

AbstractThe features or mechanisms of dyshidro-sis have not been suYciently clarified.Neither has the diVerence between hyper-hidrosis and hypohidrosis. To clarify thefeatures and mechanisms of dyshidrosis(hyperhidrosis and hypohidrosis) in sy-ringomyelia, the clinical features focusingon hidrosis of 30 patients with syringomy-elia and Chiari malformation locatedfrom a syringomyelia database were pro-spectively analysed. The patients wereclassified into three groups: eight patients(26.7%) had segmental hypohidrosis, 10(33.3%) had segmental hyperhidrosis, and12 (40.0%) had normohidrosis. We foundthat the Karnofsky functional status forthe hyperhydrosis and normohidrosisgroups were significantly higher than forthe hypohidrosis group (p=0.0012), withno significant diVerences between thehyperhidrosis and normohidrosis groups.The duration from the onset of syringo-myelia to the current dyshidrosis was sig-nificantly longer in the hypohidrosisgroup than in the hyperhidrosis group(p=0.0027). A significant correlation wasidentified between the duration from theonset of syringomyelia to the time at studyand the performance score (r=−0.599,p=0.0003). The results substantiate previ-ous hypotheses that in its early stagesyringomyelia causes segmental hyperac-tivity of the sympathetic preganglionicneurons, and hyperactivity of these gradu-ally subsides as tissue damage progresses.Focal hyperhidrosis may be regarded as ahallmark of a relatively intact spinal cord,as well as normohidrosis.(J Neurol Neurosurg Psychiatry 1999;67:106–108)

Keywords: syringomyelia; Chiari malformation; dyshid-rosis; sympathetic preganglionic neuron

Syringomyelia is responsible for dyshidrosis aswell as other disorders of the autonomic nerv-ous system.1–4 Although the link between syrin-gomyelia and focal dyshidrosis (hyperhidrosisand hypohidrosis) has been described, noadequate notice has been taken of thephenomenon,1 5–8 nor has its mechanism beenproperly explained.1–3 9–11 Furthermore, no de-scriptions have been oVered of any relationbetween hyperhidrosis and hypohidrosis.

MethodsPATIENTS AND METHODS

We have stored clinical data of a consecutive 34patients (0.28%) with MRI confirmed syringo-myelia and Chiari malformation among the 11967 outpatients who visited our neurologyclinic from April 1989 to November 1996.Three quarters of the patients came withoutreferral, and a quarter came with referral. Ouroutpatient clinic is not only a primary but alsoa secondary and tertiary centre for diagnosisand treatment of all neurological disorders.The case records of each patient showed thatthere had been no referral bias to our clinic asa consequence of our interest in dyshidrosis insyringomyelia. The protocol includes items forthe autonomic nervous system as well as othersystems. The data were obtained prospectively,with the intention of avoiding any predetermi-nation bias, according to a protocol forexamining patients with syringomyelia, whichwe ourselves designed. At the time of regis-tration for this study, we obtained informedconsent from patients to enter their infor-mation into our study database. Four of the 34patients were dropped from the study becausethey refused to give us permission to employclinical information about themselves for anyclinical study; this left 30 patients. Twelve ofthese patients were operated on for syringomy-elia, and we completed entering their infor-mation at the time of the operation.

PROTOCOL FOR EXAMINING DYSHIDROSIS

We obtained detailed clinical information foreach patient, including their previous experi-ence of hidrosis. A structured protocol wasused to examine the state of hidrosis: step 1asking about the nature and distribution of thehidrosis and the eVects of room temperature,exercise, clothing, psychological burdens, food,etc; step 2 examining the state of hidrosis byobservation, manual examination, and drawinga metal spoon across the surface of the skin(spoon test12) either after adequate physicalexercise or when lying in a bed warmed inadvance by electric blankets. When furtherinvestigation was necessary to clarify the natureand distribution of dyshidrosis, the followingsteps were performed; step 3 taking a thermo-graph in a room at a temperature of 21–28°C;step 4 a hidrosis examination (the iodine-starchmethod) at a room temperature of 45–50°C, orin a bed warmed in advance. Consequently we

J Neurol Neurosurg Psychiatry 1999;67:106–108106

Department ofNeurology, HokkaidoUniversity School ofMedicine, Sapporo,JapanK SudoN FujikiS TsujiM AjikiT HigashiM NiinoS KikuchiF MoriwakaK Tashiro

Correspondence to:Dr Kazumasa Sudo,Department of Neurology,Hokkaido University Schoolof Medicine, Kita 14, Nishi5, Kita-Ku, Sapporo,060–8648 Japan. Telephone0081 11 716 1161, ext 6028;fax 0081 11 700 5356; [email protected]

Received 6 July 1998 and inrevised form7 January 1999Accepted 19 January 1999

on Decem

ber 23, 2020 by guest. Protected by copyright.

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.67.1.106 on 1 July 1999. Dow

nloaded from

on Decem


http://jnnp.bmj.com

/J N

eurol Neurosurg P


nloaded from

on Decem


http://jnnp.bmj.com

/J N

eurol Neurosurg P


nloaded from

on Decem


http://jnnp.bmj.com

/J N

eurol Neurosurg P


nloaded from

on Decem


http://jnnp.bmj.com

/J N

eurol Neurosurg P


nloaded from

on Decem


http://jnnp.bmj.com

/J N

eurol Neurosurg P


nloaded from

on Decem


http://jnnp.bmj.com

/J N

eurol Neurosurg P


nloaded from

on Decem


http://jnnp.bmj.com

/J N

eurol Neurosurg P


nloaded from

on Decem


http://jnnp.bmj.com

/J N

eurol Neurosurg P


nloaded from

on Decem


http://jnnp.bmj.com

/J N

eurol Neurosurg P


nloaded from

on Decem


http://jnnp.bmj.com

/J N

eurol Neurosurg P


nloaded from

on Decem


http://jnnp.bmj.com

/J N

eurol Neurosurg P


nloaded from

on Decem


http://jnnp.bmj.com

/J N

eurol Neurosurg P


nloaded from

http://jnnp.bmj.com/













performed thermography in 16 patients, andthe iodine-starch method in nine patients. Weassessed the performance status by Karnofskyperformance score (K score); this ranges from0 to 100, and the higher the score, the betterthe performance.13

CASE RECORDS

(See also our previous case records for threepatients with hyperhidrosis14).

Patient 18 (hyperhidrosis; 37 year oldwoman) had a 7 month history of persistenthyperhidrosis and pain in the left upper quad-rant of the body. There was no muscleweakness or atrophy. Thermography showedlow temperature in the left upper quadrant,which was consistent with hyperhidrosis of thearea (figure A). She underwent an operation (asyringosubarachnoid shunt) 7 months afteronset, after which hyperhidrosis and pain sub-sided within a week, as indicated by thermog-raphy (figure B).

STATISTICAL ANALYSIS

We performed a statistical analysis of fivefactors (age at time of study, age of onset ofsymptoms, duration from onset of symptoms

to time of study, K score, and duration offollow up period) for the three groups ofpatients (hyperhidrosis, normohidrosis, andhypohidrosis) by one way factorial analysis ofvariance (ANOVA) (Fisher’s PLSD method asa post hoc test), and for two factors (age ofonset of current dyshidrosis and duration fromonset of symptoms to current dyshidrosis)between the two groups of patients (hyperhid-rosis and hypohidrosis) by non-paired t test. Wethen obtained the Pearson’s correlation coef-ficient for duration from onset of symptoms totime of study, and K score for all 30 patients.

ResultsOf the 30 patients, eight (26.7%) had hypohid-rosis, 10 (33.3%) had segmental hyperhidrosis,and 12 (40%) had normohidrosis. In allpatients, the distribution of dyshidrosis corre-sponded with the location of the syrinx andother neurological manifestations; the syrinxwas located roughly in the region from the cen-tral canal to the unilateral (or sometimes bilat-eral with asymmetry) posterior angle of thespinal cord.

We have summarised the results of the statis-tical analyses in the table. Although we

Thermograph of patient 18 before (A) and 11 days after (B) surgical decompression of syrinx (anterior views). Temperature asymmetry has resolvedimmediately after decompression of the syrinx.

Table 1 Summary of evaluation of each item

Item Hypohidrosis Hyperhidrosis Normohidrosis p Value

Number of patients 8 10 12Age at time of study:

Range (y) 17-71 18-75 19-75Mean (SD) 45.4 (22.7) 39.2 (17.9) 42.9 (18.8) 0.7982*

Age of onset of symptoms:Range (y) 2-55 5-45 10-69Mean (SD) 21.5 (16.4) 23.8 (13.6) 33.3 (16.2) 0.2008*

Duration from onset of symptoms to study:Range (y) 8-45 1-33 1-32Mean (SD) 23.9 (13.3) 15.5 (10.1) 9.67 (9.61) 0.0273*{

Karnofsky performance score:Range 50-90 60-90 70-90Mean (SD) 65.0 (16.9) 85.0 (9.72) 83.3 (6.51) 0.0012*§¶

Follow up period:Range (y) 1-9 1-9 1-8Mean (SD) 3.38 (3.25) 5.00 (3.74) 2.83 (2.76) 0.2960*

Age of onset of current dyshidrosis:Range (y) 16-66 13.50 —Mean (SD) 42.3 (21.0) 29.2 (13.5) — 0.1287†

Duration of onset from symptoms to currentdyshidrosis:Range (y) 7-45 0-15 —Mean (SD) 12.6 (4.5) 5.50 (4.70) — 0.0027†§

*One way factorial ANOVA (Fisher’s PLSD method).†Non-paired t test.§Statistically significant.{p value for each pair of items: hyperhidrosis v hypohidrosis 0.1148; hypohidrosis v normohidrosis 0.0078; normohidrosis v hyper-hidrosis 0.2194.¶p value for each pair of items: hyperhidrosis v hyperhidrosis 0.0007; hypohidrosis v normohidrosis; 0.7282; normohidrosis v hypo-hidrosis 0.0012.

Focal (segmental) dyshidrosis in syringomyelia 107

on Decem


http://jnnp.bmj.com

/J N

eurol Neurosurg P


nloaded from


ourselves were not able to confirm any changein the features of hidrosis during the follow upperiod, the clinical histories of five patientsindicate that hidrosis decreased segmentally ordiVusely over a long period. Duration fromonset of symptoms to study was significantlylonger in the hypohidrosis group than the nor-mohidrosis group (ANOVA; F(2, 27)=4.127,p=0.0273) (table). The K scores were signifi-cantly higher in the hyperhidrosis and normo-hidrosis groups than in the hypohidrosis group(ANOVA; F(2, 27)=8.765, p=0.0012), with nosignificant diVerence between the hyperhidro-sis and normohidrosis groups (p=0.7282).Duration from onset of symptoms to currentdyshidrosis was significantly longer in thehypohidrosis than in the hyperhidrosis group(p=0.0027). There were no significant diVer-ences between hidrosis groups in the other fac-tors. A significant correlation between the twoitems (duration from onset of symptoms tostudy, and K score) was recognised (r=−0.599,n=30, p=0.0003).

DiscussionWe performed a MEDLINE search for publi-cations dealing with dyshidrosis in syringomy-elia between January 1966 and June 1997, witha language limitation of English, German, andFrench. This search confirmed the scant accu-mulation of information regarding dyshidrosisin syringomyelia; we were able to locate onlyfour cases accompanied by Chiari malforma-tion (three of which we have already reportedand are currently included in this study).10 14

We previously speculated that hyperhidrosisis caused by stimulation of sympathetic pregan-glionic neurons (SPGNs) rather than interfer-ence to the inhibitory tract.14 This is equivalentto the hypothesis regarding body hypertrophywhich we recently presented.4 Later, disinhibi-tion of the inhibitory local interneurons(ILINs), which are located in the vicinity ofSPGNs, was supposed to be the cause ofsegmental hyperactivity of the SPGNs.10 When,in the clinical course of syringomyelia, slowlyprogressive tissue damage around the syrinx3

reaches the lateral horn, it will segmentallyaVect the SPGNs or adjacent structures.14 Thistime our results have shown, from the view-point of sweating, that there evidently is ahyperactivity of the SPGNs as long as the dis-ability is mild; however, as the disabilityprogresses, the hyperactivity gradually de-creases and is replaced by hypoactivity (table).

We have a choice of two possibilities for themechanism responsible for the hyperactivity ofthe SPGNs so far; the first is that the SPGNsare stimulated directly by a minimal tissuedamage; the second is damage to the ILINs10

(preceding the damage to the SPGNs)—wehave recently acquired some knowledge ofthese ILINs.15 16 In either case, as the diseaseprogresses, hyperactivity will shift to normoac-tivity and finally to hypoactivity because ofprogressive and irreversible damage to the

SPGNs, consistent with the clinical history offive of our eight patients with hypohidrosis. Bycontrast, immediate resolution of hyperhidrosisafter decompression of the syrinx in patient 18,whose disability was minimal, showed that thedamage to the SPGNs was mild and tends tobe reversible in patients with mild disability(figure). Before we reached the above hypoth-esis for the mechanism of hyperhidrosis, wehad ruled out the possibility of interference tothe inhibitory tract that connects the uppercentre and the SPGNs as before because of thesegmental distribution of hyperhidrosis andbecause of the locational relation among thesyrinx, inhibitory tracts, and SPGNs.14 16

Current results substantiate previous specu-lation about the mechanism of and relationbetween hyperhidrosis and hypohidrosis insyringomyelia.3 9 14 We now think that, in syrin-gomyelia, focal hyperhidrosis can be regardedas a hallmark of a relatively intact—eventhough slightly damaged—spinal cord. We alsothink that, in syringomyelia, some part of nor-mohidrosis is associated with a considerableamount of spinal cord damage. We proposethat in diagnosing focal dyshidrosis, moreattention should be given to the possibility ofsyringomyelia.

We are indebted to Drs Kazutoshi Hida, Yoshinobu Iwasaki, andHiroshi Abe (Department of Neurosurgery, Hokkaido Univer-sity School of Medicine) for their help in providing us withclinical information regarding surgical treatment. This studywas supported by Research Grant (5B-3) for Nervous andMental Disorders, from the Japanese Ministry of Health andWelfare.

1 Spillane JD. Syringomyelia. An atlas of clinical neurology.London: Oxford University Press, 1968:254–63.

2 Foster JB, Hudgson P. The clinical features of communicat-ing syringomyelia. In: Barnett HJM, Foster JB, Hudgson P,eds. Syringomyelia. London: WB Saunders, 1973:16–29.

3 Schliep G. Syringomyelia and syringobulbia. In: Vinken PJ,Bruyn GW, eds. Handbook of clinical neurology. Amsterdam:North-Holland, 1978:255–327.

4 Sudo K, Owada Y, Yabe I, et al. Syringomyelia as a cause ofbody hypertrophy. Lancet 1996;347:1593–5.

5 Nudleman K, Andermann E, Andermann F, et al. The hemi3 syndrome. Hemihypertrophy, hemihypaesthesia, hemi-areflexia and scoliosis. Brain 1984;107:533–46.

6 Mancall EL. Syringomyelia. In: Rowland LP, ed. Merritt’stextbook of neurology. 8th ed. Philadelphia: Lea and Febiger,1989:687–91.

7 Adams RD, Victor M, Ropper AH. Syringomyelia. In:Adams RD, Victor M, Ropper AH, eds. Principles of neurol-ogy. 6th ed. New York: McGraw-Hill, 1997;1269–77.

8 Finlayson AI. Syringomyelia and related conditions. In:Joynt RJ, eds. Clinical neurology. Philadelphia: Lippincott-Raven, 1995;1–17.

9 Döring G. Zur Klinik vegetativer Störungen bei Syringomy-elie und über trophische Störungen im allgemeinen. DtschMed Wochenschr 1949;74:754–9.

10 Stovner LJ, Sjaastad O. Segmental hyperhidrosis in two sib-lings with Chiari type I malformation. Eur Neurol 1995;35:149–55.

11 Johnson RH, Lambie DG, Spalding JMK. Syringomyeliaand related conditions. In: Joynt RJ, ed. Clinical neurology.Philadelphia: Lippincott-Raven, 1995:1–17.

12 Pryse-Phillips W. Companion to clinical neurology. In:Pryse-Phillips W, eds. Boston: Little, Brown, 1995;812.

13 Milstein JM, Cohen ME, Sinks LF. The influence and reli-ability of neurologic assessment and Karnofsky perform-ance score on prognosis. Cancer 1985;56:1834–6.

14 Sudou K, Tashiro K. Segmental hyperhidrosis in syringomy-elia with Chiari malformation. J Neurol 1993;240:75–8.

15 Berry MM, Standring SM, Bannister LH. Autonomic nerv-ous system. In: Williams PL, ed. Gray’s anatomy. 38th ed.New York: Churchill Livingstone, 1995:1292–312.

16 Lewis DI, Coote JH. Chemical mediators of spinalinhibition of rat sympathetic neurones on stimulation in thenucleus tractus solitarii. J Physiol (London) 1995;486:483–94.

108 Sudo, Fujiki, Tsuji, et al

on Decem


http://jnnp.bmj.com

/J N

eurol Neurosurg P


nloaded from


LETTERS TOTHE EDITOR

Behavioural status during theintracarotid amobarbital procedure(Wada test): relevance for surgicalmanagement(J Neurol Neurosurg Psychiatry1999;67:549–559)

Presurgical evaluation in many epilepsyprogrammes often includes the intracarotidamobarbital procedure (IAP). Sodiumamytal is injected into the internal carotidartery to produce a temporary “pharmaco-logical paralysis” of hemispheric function.Traditionally, the IAP has been employed inpatients with refractory temporal lobe epi-lepsy being considered for anterior temporallobectomy. In these cases it is used todetermine cerebral dominance for language,1

to assess the risk of severe postsurgicalamnesia,2 and to predict postsurgical materialspecific memory changes.3 More recently, theuse of the IAP has been extended to compli-ment EEG localisation and radiological databy lateralising temporal lobe dysfunction.4

The IAP may have a hitherto unrecognisedrole in patients with refractory frontal lobeepilepsy being considered for frontal lobec-tomy. Specifically, observation of behaviouralfunction during the period of the ablationmay provide useful information about theintegrity of the contralateral frontal lobe.This is particularly relevant in those candi-dates with a history of cerebral trauma inwhom damage to the bifrontal lobe is knownor suspected. A review of the IAP studiesperformed on patients with temporal lobeepilepsy in our comprehensive epilepsy pro-gramme (1991–8) suggests that the emer-gence of frontal lobe behavioural features iscommon in patients in whom the aetiologyleads to the suspicion of bifrontal compro-mise (for example, a history of traumatichead injury). By contrast, these featuresrarely emerge in cases of non-traumatic aeti-ology, in which the integrity of frontal lobesystems is presumed. Although it remains anincidental finding in the context of determin-ing the suitability of a candidate for anteriortemporal lobectomy, this outcome may havepotential implications for the selection ofpatients for frontal lobectomy.

We report a case of frontal lobe epilepsy sec-ondary to a traumatic head injury. Out of con-cern for untoward postoperative behaviouralchange, we employed the IAP in an attempt topredict the risk of a frontal lobe syndrome.

A 39 year old man had a 23 year history ofsevere refractory epilepsy. The seizures post-dated a motor car accident at the age of 12years when he sustained a head injury with anill defined period of loss of consciousness. Sei-zures commenced within months of that injuryand, although initially well controlled, becamerefractory within a few years. The seizure typesincluded staring spells, violent tonic-clonicseizures, and atonic drop attacks. He had com-plications from his epilepsy including a frac-tured jaw, two episodes of severe burning dueto seizures while showering, multiple episodesof postictal confusion and probable postictalpsychosis, a lung abscess secondary to aspira-tion, and episodes of status epilepticus. Interic-tal EEG recordings showed bilateral general-ised spike and wave discharges at around 2

Hz-2.5 Hz with some mild increase in bilateralslow activity and no convincing evidence ofelectrographic focalisation. Video EEG moni-toring showed apparent generalised seizureswithout any focal onset on scalp EEG. BrainMRI disclosed a well defined atrophic lesioninvolving the left frontal pole considered likelyto be post-traumatic in origin. Interictal FDGPET and HMPO SPECT disclosed hypoper-fusion in the left anterior frontal regioncommensurate with the abnormality shown onMRI. Although his electroclinical pattern wassuggestive of symptomatic generalised epi-lepsy, because of the left frontal lesion, seizureonset from that region was considered likely.

On neuropsychological examination, hisgeneral cognitive function was normal. At abehavioural level, however, he presented asvery peurile in manner with a very rigid,inflexible cognitive style. The neuropsycho-logical opinion was of a mild frontal lobe syn-drome consistent with the history of trau-matic head injury. There was no currentevidence of psychiatric disorder. Althoughhaving successfully passed his final year ofsecondary school (together with severalcourses of advanced education), he hadremained unemployed due to his seizures. Hewas socially isolated and his interpersonalrelationships were limited.

He had severe life threatening epilepsy withthe surgical option the only remaining avenueof treatment. However, as surgical manage-ment would involve resection of the left fron-tal lobe against a background of traumatichead injury and the possibility of more gener-alised frontal lobe compromise, a left hemi-spheric IAP was performed. Sodium amytal(125 mg) was administered via a slow handinjection. Of relevance, no crossflow into thecontralateral anterior cerebral artery via theanterior communicating artery was present(as assessed by a separate injection of contrastmedium). The injection was accompanied bya dense right hemiplegia and global aphasicarrest. Resolution of language was character-ised by a dense perseveration of countingwhich could not be influenced by theexaminer. Despite normal comprehension, heshowed severely impaired capacity for motorregulation (go-no go paradigm), togetherwith marked behavioural disinhibition (agita-tion, swearing, verbosity, childishness). Al-though seemingly aware of some aspects ofhis behaviour (apologising for swearing), heseemed unable to modify his responses. Theoverall impression was of a pronounced fron-tal lobe syndrome, suggesting that the rightfrontal lobe had incurred some damagesecondary to the documented head traumaand that he must have been reliant on someleft frontal contribution.

On the basis of the IAP findings, a selectivecortical resection (as opposed to more exten-sive frontal lobectomy) restricted to the regionof damage was advised. Intraoperative electro-corticography showed active focal epileptiformdischarges maximal in the inferior frontal lobein the electrodes closest to the lesion. A corti-cal resection was performed with framelessstereotaxy guidance excision of the frontallesion. Histopathology on the resected tissueshowed an old post-traumatic cyst involvingthe cortex and white matter. His postoperativecourse was unremarkable. When reviewed 3months after surgery he was seizure free. Hisperformance on neuropsychological evalua-tion remained commensurate with presurgicalstatus. There were no novel subjective com-plaints. Mood, behaviour, and temperamentremained stable.

Despite its undoubted value in many indi-vidual cases of temporal lobe epilepsy, theIAP has remained a controversial assessmentinstrument.5 Amid this controversy its poten-tial usefulness in other patient groups seemsto have been overlooked. A primary criticismof its use in temporal lobe epilepsy has beenthe question of irrigation and whether themedial temporal lobe is adequately “disa-bled” during the procedure. This particularlimitation is not applicable when used in thepatient with frontal lobe epilepsy, as theregion of interest is clearly ablated via supplyfrom the carotid arterial system. Cautionmust, however, be exercised with respect topossible crossflow into the contralateral ante-rior cerebral artery via the anterior communi-cating artery. When such crossflow is present,the ability to assess validly the integrity ofcontralateral frontal lobe function will beconfounded by virtue of a pharmacologicallyinduced bilateral frontal lobe syndrome. Aswith the use in cases of temporal lobeepilepsy, only a restricted form of assessmentis possible with the frontal lobe patient duringthe period of ablation. An assessment focus-ing on issues of behavioural regulation wouldseem most useful.

It should be borne in mind that the degreeof frontal lobe dysfunction induced by theIAP represents the “worst case scenario” asthe entire frontal lobe is included in the abla-tion. There are likely to be few surgicalscenarios in which a comparable extensiveresection of tissue is likely to be considered,and results must be interpreted in thiscontext. This limitation not withstanding, theIAP does seem to have a role in separating outthose patients in whom more extensive fron-tal lobe resections could be considered asopposed to those in whom a more conserva-tive approach is warranted.

This case report forms only the basis for anovel hypothesis that clearly requires morerigorous scientific research before its clinicalutility can be reliably established. Nonethe-less, we think that it is worth drawing theattention of the epileptological community tothe potential application of the IAP in thesurgical management of extratemporal cases.

MARIE F O’SHEAMICHAEL M SALING

Department of Neuropsychology

SAMUEL F BERKOVICDepartment of Neurology, Austin and Repatriation

Medical Centre, Melbourne, Australia; andDepartment of Medicine, University of Melbourne,

Grattan Street, Parkville 3052, Australia.

Correspondence to: Dr Marie F O’Shea, Depart-ment of Neuropsychology, Austin and RepatriationMedical Centre (Austin Campus), Studley Road,Heidelberg, Victoria 3084, Australia. Telephone613 3 03 9496 5913; Fax 613 3 03 9457 2654.

1 Wada J, Rasmussen T. Intracarotid injection ofsodium amytal for the lateralization of cerebralspeech dominance: experimental and clinicalobservations. J Neurosurg 1960;17:266–82.

2 Rausch R, Silfveius H, Weiser H-G, et al.Intraarterial amobarbital procedures. In: EngelJ, ed, Surgical treatment of the epilepsies. 2nd ed.New York: Raven Press, 1993:341–57.

3 Kneebone AC, Chelune GJ, Dinner DS, et al.Intracarotid amobarbital procedure as a pre-dictor of material-specific memory change afteranterior temporal lobectomy. Epilepsia 1995;36:857–65.

4 Baxendale SA, Van Paesschen W, Thompson PJ,et al. The relation between quantitative MRImeasures of hippocampal structure and theintracarotid amobarbital procedure. Epilepsia1997;38:998–1007.

5 Jones-Gotman M, Barr WB, Dodrill CB, et al.Controversies concerning the use of intraarte-rial procedures. In: Engel J, ed. Surgicaltreatment of the epilepsies. 2nd ed. New York:Raven Press, 1993:445–9.

J Neurol Neurosurg Psychiatry 1999;67:549–559 549

Reversal of tetrabenazine induceddepression by selective noradrenaline(norepinephrine) reuptake inhibition

Tetrabenazine (TBZ), a synthetic benzoqui-nolizine, was first introduced as a neurolepticagent in 1960, and is now widely used in thetreatment of hyperkinetic movement disor-ders such as chorea, tics, or tardive dyski-nesia. The side eVect profile is mainlycharacterised by the triad of drowsiness/fatigue, parkinsonism, and depression; de-pression is found in about 15% of patientstreated with TBZ.1 We here report on therapid reversal of depressive symptoms in apatient treated with TBZ for orofacial dysto-nia by administering the new and highlyselective noradrenaline (norepinephrine) re-uptake inhibitor (SNRI) reboxetine.2

On admission, the 64 year old woman pre-sented with perioral and lingual hyperkinesiasas well as intermittent and involuntary move-ments of her lower jaw, which had lasted forabout 2 years, causing her a considerableimpairment of articulation. No history ofneuroleptic treatment or Parkinson’s diseasewas evident. Her cranial CT and bloodchemistry were normal. We diagnosed a seg-mental dystonia, which improved dramati-cally after initiation of a tetrabenazinemedication (60 mg a day). This successfultreatment response, however, was accompa-nied by a severe depressive syndrome, whichwas characterised by a mixed anxious-depressive mood, low self esteem, a completeloss of drive, and intermittent suicidalideations. After switching from TBZ totiapride, the patient recovered from depres-sion, but her neurological status worsenedsignificantly. The re-exposure to TBZ againameliorated hyperkinesia, but provoked adepressive relapse. A comedication withreboxetine (6 mg/day), a new and selectivenoradrenaline reuptake inhibitor, finally ledto a stable remission of the depressive symp-toms within a week, without any worsening ofthe dystonic syndrome.

Tetrabenanzine (TBZ) is known to act as amonoamine depleting and dopamine recep-tor blocking drug.1 In more detail, TBZbinds to and inhibits specifically the humanvesicular monoamine transporter isoform 2(hVMAT2). Whereas the indolamine serot-onin (5-HT) performs a similar aYnity forboth hVMAT1 and hVMAT2, catecho-lamines such as noradrenaline exhibit athreefold higher aYnity for hVMAT2.3 Asthese specific transporters are responsible forpackaging monoamine neurotransmittersinto presynaptic secretory vesicles for releaseby exocytosis, the inhibition of hVMAT2 bycompounds such as tetrabenazine thusresults in consecutive noradrenalinedepletion.4

Alterations of noradrenergic neurotrans-mission—that is, a neuronal noradrenalinedepletion—can therefore be postulated toform one major origin of TBZ induceddepression. In line with this assumption,brain-specific catecholaminergic activity en-hancers (CAEs) such as phenylethylaminehave been shown to antagonise TBZ induceddepression-like behaviour in rats.5 Modulat-ing this altered noradrenergic neurotransmis-sion pattern by the administration of selectivenoradrenaline reuptake inhibitors such asreboxetine may thus provide a new, specific,and fast acting tool in the management ofdepression caused by TBZ and related (neu-roleptic) compounds.

WOLFGANG SCHREIBERJÜRGEN-CHRISTIAN KRIEG

Department of Psychiatry and Psychotherapy,Philipps-University, Rudolf-Bultmann-Straâe 8,

D-35033 Marburg/Lahn, Germany

TOBIAS EICHHORNDepartment of Neurology, Philipps-University,

Rudolf-Bultmann-Straâe 8, D-35033 Marburg/Lahn,Germany

Correspondence to: Dr Wolfgang Schreiber, De-partment of Psychiatry and Psychotherapy,Philipps-University, Rudolf-Bultmann-Straâe 8,D-35033 Marburg/Lahn, Germany. Telephone0049 6421 286277; fax 0049 6421 285229; [email protected]

1 Jankovic J, Beach J. Long-term eVects oftetrabenazine in hyperkinetic movement disor-ders. Neurology 1997;48:358–62.

2 Montgomery S. Reboxetine: additional benefitsto the depressed patient. J Psychopharmacol1997;11(suppl 4):S9–15.

3 Erickson JD, Schäfer MK, Bonner TI, et al. Dis-tinct pharmacological properties and distribu-tion in neurons and endocrine cells of two iso-forms of the human vesicular monoaminetransporter. Proc Natl Acad Sci U S A 1996;93:5166–71.

4 HoVman BJ, Hansson SR, Mezey E, et al.Localization and dynamic regulation of bio-genic amine transporters in the mammaliancentral nervous system. Front Neuroendocrinol1998;19:187–231.

5 Knoll J, Miklya I, Knoll B, et al. Phenylethyl-amine and tyramine are mixed-acting sym-pathomimetic amines in the brain. Life Sci1996;58:2101–14.

Spinal sulcal artery syndrome due tospontaneous bilateral vertebral arterydissection

In young adults vertebral artery dissection(VAD) is an important cause of braininfarction.1 2 A known mechanism is micro-traumata due to abrupt head movements—for example, chiropractic manoeuvres. Inaddition a pathogenetic role of connective tis-sue diseases, cystic media necrosis, fibromus-cular dysplasia, migraine, and inflammatorydiseases has been postulated.3 In VAD initialneck pain is often reported, which may beslight. Lesions caused by VAD are cerebellaror brainstem infarcts, unilateral or bilateralthalamic infarcts (top of the basilar syn-drome), or infarctions in the posterior cer-ebral artery territory due to intra-arterialembolism or haemodynamic decompensationwhen collaterals are insuYcient.1 Lesions ofthe cervical spinal cord are rare because of itsgood collateral supply.4 5 We report on apatient with a syndrome of the spinal sulcalartery (incomplete Brown-Séquard syn-drome) caused by spontaneous bilateral VAD.

A 43 year old man with a history of arterialhypertension presented with left sided numb-ness sparing the face, which had evolved sud-denly while he was walking. In addition, hereported on dull right sided neck pain irradi-ating into the occiput, which had beeninitiated by a head rotation while he wasworking at a computer 2 weeks before. Theneck pain had spontaneously ceased 6 dayslater. Neurological examination discloseddissociated sensation defect on the left withan indistinct level around C4 to C6. Belowthis level on the left he had a markedhypalgesia and nearly a loss of temperaturesense. The right limbs were warmer than theleft ones. In addition, we found mild rightsided motor system deficits. Cranial nervefunction was intact, despite a right sidedHorner’s syndrome. According to chestradiography phrenic nerve function waspreserved. Routine laboratory findings in-cluding CSF analysis were normal. The

hemiparesis and the diVerent temperaturesensation in the limbs resolved completelywithin 3 weeks.

Tibial nerve somatosensory evoked poten-tials (SSEPs) had regular N22 and P40latencies and amplitudes. Central motor con-duction time (CMCT) after transcranialmagnetic stimulation was prolonged to theright abductor digiti minimi (9.2 ms) andtibialis anterior (23.1 ms). The CMCT to theleft target muscles was normal. Duplexsonography showed increased flow velocityon the level of the cervical vertebrae 3 to 5with a maximum of 214 cm/s in the right and197 cm/s in the left vertebral artery. Colourmode showed irregular narrowings of thelumen indicating dissections.

Cervical MRI showed a spinal cord infarc-tion at the level C2 (figure). The circumfer-ence and dorsal part of the cord were notaVected. In digital subtraction angiography(DSA) both vertebral arteries had string signsin the V1 and V2 segments with collateral flowto the distal V2–4 segments via the thyreocer-vical trunk (cervical ascendent artery) and thecostocervical trunk also. The anterior spinalartery was incompletely contrasted by unilat-eral spinal branches of the right vertebralartery. They originated at the level of dissec-tion. The intradural origins of the anterior spi-nal artery from the distal part of the vertebralarteries (V4 segment) were not visible.

Bilateral spontaneous VAD is not rare, butoften missed. In most cases, microtrauma pre-ceding the dissection can be recalled by thepatients. Due to the mild mechanical impact,the action of predisposing factors might bepostulated. Among these may be changing intype III collagen, migraine, fibromuscular dys-plasia, infections in the near past, and inflam-matory vasculopathy.2 Magnetic resonanceimaging with typical semilunar mural hae-matoma and in addition magnetic resonanceangiography (MRA) with complementarydocumentation of an irregular lumen or taper-ing occlusion have a high sensitivity andspecificity in cases of internal carotid arterydissection.1 By contrast, mural haematomas ofthe VA especially in the V1 and the V3segments are often not detectable by MRI. Incases of unclear non-invasive findings, DSA isstill the method of choice.1

In addition to consecutive brain infarc-tions, cervical spinal cord infarctions andnerve root compression syndromes mayoccur in cases of unilateral or bilateral VAD.Probably as a result of the pial collateral net-work and the dual posterior spinal artery, spi-

Coronal T2 weighted MRI: ventrolateralparamedian right sided medullary infarction.

550 Letters, Correspondence, Book reviews, Correction

nal cord infarction is often located in theanterior spinal artery territory with the greymatter of the anterior horns exhibiting thehighest vulnerability to ischaemia.4 5 Thismechanism may lead to a typical “snake eye”configuration of medullary infarction.3 Be-sides the supply via VA spinal branches,which is found in 19% only unilaterally,4

there are branches originating from theascendant cervical artery (thyreocervicaltrunk) and the costocervical trunk supplyingthe spinal cord.

DSA findings in the present case suggestthat spinal branches originating from theright V2 segment were dominant feeders ofthe anterior spinal artery whereas there wasno evidence of direct communication be-tween vertebral and spinal arteries from theV4 segment. The dissection involved the V2segment from which these spinal branchesoriginate. A transient occlusion of thesespinal branches is a likely consequence. Thisunusual type of arterial medullary supplymay explain why VAD causes spinal cord inf-arction. Contrary to Pullicino,5 who de-scribed upper limb atrophies due to cervicalspinal cord infarction involving the anteriorhorns, the present case shows a unilateralinvolvement of commissural, spinothalamic,pyramidal, and vasoconstrictor tracts. To ourknowledge sulcal spinal artery syndromecaused by bilateral spontaneous VAD has notyet been described. In conclusion, differentialdiagnosis of acute spinal symptoms in youngadults should include spontaneous unilateralor bilateral VAD with cervical spinal cordischaemia.

S WEIDAUERD CLAUS

Department of Neurology

M GARTENSCHLÄGERInstitute for Radiology, Klinikum Darmstadt,

Teaching Hospital University Frankfurt, Germany

Correspondence to: Professor D Claus, Depart-ment of Neurology, Klinikum Darmstadt, TeachingHospital University Frankfurt, Heidelberger Land-strasse 379, 64297 Darmstadt, Germany.

1 Auer AS, Felber S, Schmidauer C, et al.Magnetic resonance angiographic and clinicalfeatures of extracranial vertebral artery dissec-tion. J Neurol Neurosurg Psychiatry1998;64:474–81.

2 Caplan LR, Zarins CK, Hemmati M. Spontane-ous dissection of the extracranial vertebralarteries. Stroke 1985;16:1030–8.

3 Hundsberger T, Thömke F, Hopf HC, et al.Symmetrical infarction of the spinal cord dueto spontaneous bilateral vertebral artery dissec-tion. Stroke 1998;29:1742.

4 Kaneki M, Inoue K, Shimizu T, et al. Infarctionof the unilateral horn and the lateral column ofthe spinal cord with sparing of posteriorcolumns: demonstration by MRI. J NeurolNeurosurg Psychiatry 1994;57:629–31.

5 Pullicino P. Bilateral distal upper limb amyotro-phy and watershed infarcts from vertebraldissection. Stroke 1994;25:1870–2.

Spanish families with cavernousangiomas do not share the Hispano-American CCM1 haplotype

Cerebral cavernous malformations are vascu-lar malformations mostly located in the CNS.Their frequency is estimated close to 0.5% inthe general population.1 Cerebral cavernousmalformations occur as a sporadic or heredi-tary condition. From the Hispano-Americanpopulation, familial forms were reported witha high frequency.2 CCM1, a hitherto uniden-tified gene mapping on chromosome 7 wasshown to be involved in all families with cer-ebral cavernous malformations of Hispano-

American descent with a strong foundereVect.2 3 Around 50% of non-Hispano-American families showed linkage to CCM1but no common haplotype was found.4 5 Arecent study showed linkage of cerebralcavernous malformations to two additionalloci.5 No Spanish family with cerebralcavernous malformations has been analysedso far.

We report herein a genetic linkage analysisconducted on nine Spanish families withcerebral cavernous malformations. All proce-dures were approved by an ethics comittee.The families were unrelated and originatedfrom diVerent regions of Spain (south west(CVE2, 3, 4, 10, 17, 25), central (CVE24),

south east (CVE28), and north east (CVE29).Seventy seven subjects including 55 potentiallyinformative meioses and 12 spouses gave theirinformed consent. They were examined by aboard certified neurologist, underwent cer-ebral MRI, and blood samples were taken.Magnetic resonance imaging was used toestablish status for linkage analysis. Thirty fourmembers had MRI diagnosis of cavernomasand were considered as aVected. Among them,14 experienced neurological symptoms (cer-ebral haemorrhage n=6, seizures n=8). Nine-teen members with normal cerebral MRI wereconsidered as healthy. Twelve members with-out MRI investigation had an unknown status.Analysis of pedigrees was consistent with an

(A) Pedigrees of the nine families with cerebral cavernous malformations. Black symbols=symptomaticpatients with cavernous angiomas on MRI; half filled symbols=asymptomatic members with cavernousangiomas on MRI; empty symbols=asymptomatic members with normal MRI; questionmark=members with unknown status. (B) Comparison of the Hispano-American CCM1 haplotypeand the haplotypes segregating with the disease phenotype within Spanish families. Polymorphicmarkers are shown on the left. Numbers indicate the sizes in base pairs. Primers used to amplifyD7S2409 were diVerent from those in the Hispano-American families resulting in a diVerent size ofthe amplified fragment. M65B was not studied in the Hispano-American families. Family CVE24was not informative for D7S646. For families CVE17 and CVE29, the two haplotypes of the aVectedsiblings are indicated. ND=not determined.

CVE2 CVE3

?

??

CVE4

CVE17

CVE28

CVE24 CVE25

CVE29

CVE10

A

BMarker CVE2 CVE3 CVE4 CVE10 CVE24 CVE25 CVE28 CVE17 CVE29

D7S2410 273279 265 269 265 265 267 263 265 263 269263

D7S2409 221ND 219 215 221 219 219 223 219 223 219223

D7S1813 123137 127 127 127 125 127 131 125 127 127127

D7S1789 139137 133 133 129 131 133 129 129 133 133129

MS65B 135ND 133 131 133 135 133 129 ND ND 133137

D7S646 185185 185 187 197 183 185 181 187 197 201 185197

D7S558 107107 107 103 107 103 103 103 103 103 103103

D7S689 127129 125 129 127 127 139 127 125 127 127129

Hispanic-American

??

?

?

?

??

?

?

?

?

??

??

???

? ? ?

?

?

?

?

Letters, Correspondence, Book reviews, Correction 551

autosomal dominant pattern of inheritance(figure A).

Eight polymorphic microsatellite markersspanning the CCM1 interval were selected forlinkage analysis. Four were chosen from theGénéthon linkage map (D7S2410, D7S2409,D7S646, D7S689), and three from theCooperative Human Linkage Center(D7S1813, D7S1789, D7S558). The last one(M65B) was identified by SL based onsequencing data of a bacterial artificial chro-mosome (Genbank HSAC000065; BACRG085C05). The length of the genetic inter-val flanked by markers D7S2410 andD7S689 is 4 centimorgans (cM). Marker dis-tances between D7S2410/D7S2409,D7S1813/D7S1789/D7S646/D7S558, andD7S689 have been estimated to be 2.2 cM,and 1.8 cM, respectively.3 Oligonucleotidesequences are available through the GenomeData Bank (John Hopkins University, Balti-more). Genotyping and linkage analysis(LINKAGE package version 5.1) were per-formed as previously described.5

Lod scores were calculated in the fivefamilies having a suYcient number of poten-tially informative meioses—that is, CVE3(eight), CVE4 (16), CVE10 (seven), CVE25(five), and CVE28 (seven). Lod scores higherthan 1 were obtained for three families(CVE3, 4, and 28) for at least one marker.Due to incomplete informativity of threemarkers within family CVE4, lod scores didnot reach the level of 3. In family CVE10, lodscores were close to 1 for four markers(D7S2410, D7S1789, D7S558, D7S689).Family CVE25 showed a lod score close to 0for all markers. In this family, two aVectedand one asymptomatic sibling with normalstandard MRI inherited the same haplotypefrom their aVected father. When the data ofall examined families were pooled, a maxi-mum combined lod score of 5.92 wasobtained for marker D7S2410 at è=0.

In seven families (CVE2, 3, 4, 10, 24, 25,and 28), all aVected members inherited anhaplotype that was not shared by their healthyrelatives (figure B). In family CVE17, bothaVected siblings inherited a distinct haplo-type from their aVected mother. Although thelimited size of this family does not allow toformally conclude, this suggests geneticheterogeneity. In family CVE29, the twoaVected siblings inherited the same haplo-types from their mother and father whosestatus was unknown.

None of the families shared a common hap-lotype (figure B). In addition, the extendedHispano-American haplotype was not segre-gating with the disease phenotype in any of thenine families including the four families withsuggested linkage to CCM1. However, two outof nine families (CVE2 and 3), the D7S646(185bp) and D7S558 (107bp) alleles segregat-ing with the disease phenotype were identicalto the ones observed in the Hispano-Americanhaplotype. Consequently, we analysed the fre-quency of this combination of alleles within apanel of 80 haplotypes of 40 healthy whitesubjects. Frequency was 17% compared with22% in our Spanish sample. Therefore, thisfinding might be attributed to a random distri-bution of these alleles.

In conclusion, linkage analysis of Spanishfamilies with cerebral cavernous malforma-tions did not show any evidence for Hispano-American haplotype sharing or a foundereVect. Although our sample was limited insize and does therefore not formally excludethe presence of the Hispano-American haplo-type in additional Spanish families with

cerebral cavernous malformations, this hap-lotype is most likely not predominant inSpain, and the strong founder eVect seen inall published Hispano-American familieswith cerebral cavernous malformations mightbe specific for this population.

HJ is supported by the Schweizerische Stiftung fürmedizinisch-biologische Stipendien (Switzerland),SL by the Fonds de Recherche en Santé (Canada),PL by the Collège des Enseignants de Neurologieand ZENECA pharmaceutical group. The work wasfounded by INSERM, Ministère de l’EnseignementSuperieur et de la Recherche, CSIC, and the Fondode Investigacion de la Seguridad Social (Fiss:99/0407).

H H JUNGP LABAUGES LABERGE

E MARÉCHALE TOURNIER-LASSERVE

INSERM U25, Faculté de Médecine Necker, Paris,France

M LUCASLaboratorio de Biologia Molecular

J M GARCIA-MORENOM A GAMEROG IZQUIERDO

Servicio de Neurologia, Hospital Unversitario VirgenMacarena, Avenida Dr Fedriani, 41071 Sevilla, Spain

E TOURNIER-LASSERVEHôpital Lariboisière, Paris, France

Correspondence to: E Tournier-Lasserve,INSERM U25, Faculté de Médecine Necker, 156Rue de Vaugirard, 75730 Paris Cedex 15. France,Telephone 0033 1 45 67 25 97; fax 0033 1 40 56 0107; email: [email protected]

1 Otten P, Pizzolato GP, Rilliet B, et al. A proposde 131 cas d’angiomes caverneux (cavern-omes) du SNC, repérés par l’analyse rétrospec-tive de 24 535 autopsies. Neurochirurgie1989;35:82–3.

2 Günel M, Awad IA, Finberg K, et al. A foundermutation as a cause of cerebral cavernous mal-formation in hispanic Americans. N Engl J Med1996;334:946–51.

3 Johnson EW, Lyer LM, Rich SS, et al. Refinedlocalization of the cerebral cavernous malfor-mation gene (CCM1) to a 4-cM interval ofchromosome 7q contained in a well-definedYAC Contig. Genome Res 1995;5:368–80.

4 Labauge P, Laberge S, Brunereau L, et al.Hereditary cerebral cavernous angiomas: clini-cal and genetic features in 57 French families.Lancet 1998;352:1892–7.

5 Craig HD, Günel M, Cepeda O, et al. Multilo-cus linkage identifies two new loci for aMendelian form of stroke, cerebral cavernousmalformation, at 7p15–13 and 3q25.2–27.Hum Mol Gen 1998;7:1851–8.

Hydrocephalus caused by metastaticbrain lesions: treatment by thirdventriculostomy

Metastasis to the brain occurs in 20%–40%of cancer patients.1 About 20% of thesemetastases are located in the posterior fossa,cerebellum, and brainstem. Metastatic dis-ease to periventricular brain tissue canobstruct the flow of cerebrospinal fluid (CSF)produced in the ventricles to the subarach-noid space where it is normally absorbed byarachnoid granulations. This typically causesan obstructive or non-communication hydro-cephalus. A shunt has been customarilyplaced to drain CSF from a lateral ventriclethrough a pressure regulating valve and intothe atrium or peritoneal or pleural cavity.Even though this technique has been success-ful in relieving the hydrocephalus, it hasabout a 50% chance of infection or failurefrom blockage.2

Another option for the treatment ofobstructive hydrocephalus is third ventricu-lostomy, a minimal invasive endoscopic neu-

rosurgical procedure. In performing thirdventriculostomy, a hole is created in the floorof the third ventricle, allowing CSF inside theventricle to drain out to the CSF spacesurrounding the brain. Although third ven-triculostomy has a low operative morbidityand a high probability of success in secondaryhydrocephalus, it is only commonly used onpatients with aqueductal stenosis and thepediatric population. To avoid placing shuntsin patients with inoperable metastatic braintumours who typically have only a fewmonths to live, we have oVered the patientsthird ventriculostomy as a palliative proce-dure.

We performed third ventriculostomy onseven patients with hydrocephalus due tometastatic tumours of the posterior fossa orthalamus. They typically presented withsymptoms of acute hydrocephalus in additionto any local mass eVect of the tumour.Postoperatively, five patients were relieved ofhydrocephalic symptoms and follow up brainimaging studies disclosed decreased ventricu-lar size. These five patients had a medianhospital time of 6.5 days and median survivalof 9.5 weeks after the operation (table). Theirhospital stay was prolonged by care of theirprimary disease. However, most of ourpatients who underwent this operation forhydrocephalus caused by other diseases weredischarged from the hospital between 24 and48 hours from the procedure. There were nooperative complications. All five patients hadno evidence of redevelopment of hydrocepha-lus up to the last clinic visit.

Two patients had unsuccessful results fromtheir third ventriculostomy. One patient (case4) showed no change from his initialneurological exam after the procedure, buthis mental status deteriorated on post opera-tive day 6. Brain CT showed no change in thesize of his ventricles compared with the scanobtained on the day of admission. Thepatient’s family requested comfort care onlyand the patient died 2 days later. In thesecond case (case 6) the patient had improve-ment in his neurological examination andventricle size by CT scan immediately afterthe operation, but had recurrent symptoms ofhydrocephalus 11 days later. After placementof a ventriculoperitoneal shunt, his examina-tion returned to baseline.

Every patient except the person describedin pase 4 received brain radiation therapyafter the palliative procedure. One patient(case 3) underwent a course of radiationtreatment prior to the operation. Another(case 5) had radiation to her orbit in the dis-tant past after enucleation for retinoblast-oma. Even though previous radiotherapy maybe considered a contraindication for thirdventriculostomy by some authors, it did notseem to aVect the success of third ventricu-lostomy in our patients. Carcinomatous men-ingitis which could have caused a concomi-tant communicating hydrocephalus was notgrossly evident on examination, on any of thebrain imagings, or during endoscopy. How-ever, tumours in contact with CSF space canalso cause a communicating hydrocephalusby raising CSF protein which can obstructdistal CSF space and arachnoid granulations.

Our success rate of about 70% (five ofseven) for third ventriculostomy in periven-tricular metastatic disease is consistent withthe results obtained with third ventriculos-tomy for adult patients with secondaryhydrocephalus.3 This is comparable with thealternative shunting with an implanted cath-eter which has a first year revision rate as high


as 50%, with the highest failure rate in thefirst few months after shunt placement.2 Thecomplication rates for both procedures arelow. Third ventriculostomy and shunting canpotentially cause a stroke, bleeding, ventricu-litis, meningitis, a subdural haematoma, CSFleak, diabetes insipidus, and SIADH. How-ever shunting has additional risks of mechani-cal malfunction, complications associatedwith implanting a foreign body, and overd-rainage syndrome.4

Because third ventriculostomy restoresnear normal CSF dynamics,5 overdrainage isprevented. The procedure is also minimallyinvasive and safe. The procedure’s lowmorbidity, high eYcacy, and potentially shorthospital stay are well suited as a palliativetreatment of hydrocephalus for patients withan expected shortened life span. We proposethat third ventriculostomy should be oVeredas a first treatment to patients suVering fromobstructive hydrocephalus from unresectabletumours.

TIEN T NGUYENMARK V SMITH

GERARD S RODZIEWICZDepartment of Neurosurgery

SHEILA M LEMKEDepartment of Medicine, Division of Oncology, SUNY

Health Science Center, University Hospital, Syracuse,New York, USA

Correspondence to: Dr G S Rodziewicz, Depart-ment of Neurosurgery, 750 East Adam Street, Syra-cuse, NY 13210, USA. Telephone 001 315 4644470;fax 001 315 464 5520;[email protected]

1 Patchell RA. Brain metastases and carcinoma-tous meningitis. In: AbeloV MD, Armitage JO,Lichter AS, et al, eds. Clincal oncology. NewYork: Churchill Livingstone, 1995:629–41.

2 Borgbjerg BM, Gjerris F, Albeck MJ, et al. Fre-quency and causes of shunt revision in differentcerebrospinal fluid shunt types. Acta Neurochir(Wien) 1995;136:189–94.

3 Jones RFC, Kwok BCT, Stening WA, et al. Thecurrent status of endoscopic third ventriculos-tomy in the managment of non-communicating hydrocephalus. Minim InvasiveNeurosurg 1994;37:28–36.

4 Faulhauer K, Schmitz P. Overdrainage phenom-ena in shunt treated hydrocephalus. Acta Neu-rochir (Wien) 1978;45:89–101.

5 Frim DM, Goumnerova LC. Telemetric intra-ventricular pressure measurements after thirdventriculostomy in a patient with noncommu-nicating hydrocephalus. Neurosurgery 1997;4:1425–8.

Neuronal activity alters local blood flowin brain tumour adjacent to theactivating cortex

Characteristics of blood flow in brain tu-mours have been studied extensively1; thesestudies are important for diagnosis of malig-nancy and therapy monitoring. Our study isthe first to consider how activity dependentchanges of regional cerebral blood flow(rCBF) alter tumour blood flow in the braintumour adjacent to the activating cortex.

Such an interaction between cortical bloodflow and tumour blood flow may be of valuefor evaluating mechanisms of neurologicalsymptoms associated with brain tumours.

Neuronal activation causes an increase ofregional cerebral blood flow (rCBF) in theactivating cortical area.2 Near infrared spec-troscopy (NIRS) demonstrates the increasein rCBF during neuronal activity as increasesin oxygenated haemoglobin (oxy-Hb) andtotal haemoglobin (total-Hb) with a decreasein deoxyhaemoglobin (deoxy-Hb)3–5; NIRS isan optical method to measure concentrationchanges of oxy-Hb, deoxy-Hb, and total-Hb(oxy-Hb+deoxy-Hb) in cerebral vessels bymeans of the characteristic absorption spec-tra of haemoglobin in the near infrared range.

In the present study, we measured changesof oxygenation and haemodynamics in thebrain tumour adjacent to the activatingcortex by means of NIRS. We found transientdecreases in oxy-Hb and total-Hb in the

tumour during neuronal activation, suggest-ing that the local blood flow of the tumourwas decreased by a transient increase of rCBFinduced by neuronal activation.

The patient was a 35 year old right handedman who presented with complaints of head-ache and dizziness. A neurological examina-tion showed no abnormalities and a decline inlanguage functions. A postcontrast CTshowed a well defined large enhancingtumour (4×5 cm) compressing the left frontallobe. Computed tomographic angiographyshowed that the branches of the left middlecerebral artery supplied the tumour (figureA). The patient underwent a left frontalcraniotomy for removal of the tumour; thepathological diagnosis was meningioma. TheNIRS measurement was performed beforethe operation.

We measured haemodynamic changes inthe brain tumour during neuronal activationin the left frontal lobe induced by cognitive

Table 1 Clinical characteristics of patients who underwent third ventriculostomy for obstructive hydrocephalus

CaseNo

Age (y),Sex Diagnosis Result*

Postoperative stay inhospital(days)

Survival time(weeks)

1 70,M Lung mixed adenocarcinoma and squamous cancer metastasis to thalamus Improved 17 42 46,F Ovarian adenocarcinoma metastases to cerebrum and medulla Improved 9 133 38,F Breast ductal carcinoma metastases to brainstem and cerebellum Improved 3 84 75,M Rectal adenocarcinoma metastasis to cerebellum Failed 8 15 39,F Breast adenocarcinoma metastasis to cerebellum Improved 4 116 60,M Lung adenocarcinoma metastasis to thalamus Failed 6 6+†7 64,M Oesophageal carcinoma metastatic to cerebellum Improved 7+ 1+†

*Results are considered improved if the patient had resolution of symptoms and follow up imaging showed hydrocephalus improved or resolved.†Patient is currently alive.

1

0

–2

–1

–3

Co

nce

ntr

atio

n c

han

ges

(au

)

10 s

B

Oxy-HbDeoxy-HbTotal-Hb

A (A) CT angiography of the braintumour. Note that the tumour wassupplied by the branches of the leftmiddle cerebral artery. (B)Oxygenation changes in the braintumour during the naming taskmeasured by NIRS. The ordinatesindicate concentration changes ofoxy-Hb, deoxy-Hb, and total-Hb inarbitrary units (au). Horizontal thickbar indicates the period of the task.


tasks. We monitored concentration changesof oxy-Hb, deoxy-Hb, and total-Hb, using anNIRO-500 instrument (Hamamatsu Photon-ics KK, Japan). The optodes were placed atan interoptode distance of 3.5 cm on the leftforehead so that the centre of the two optodeswas placed on the centre of the tumour. Withan optode distance of 4 cm, correlations ofoxy-Hb and total-Hb measured by NIRS andrCBF measured by PET suggested that thereliable penetration depth of near infraredlight into brain tissue is about 1.3 cm3; thusthe present NIRS measurement area wasrestricted in the tumour. The patient wasseated and had his eyes open during theNIRS measurement. Informed consent wasobtained from the patient.

To activate the left frontal lobe, we used thefollowing four tasks: (1) semantic verbalfluency, which entails naming as many itemsin a semantic category (for example, animals)as possible; (2) confrontational naming,which involves naming ordinary items pre-sented by the tester; (3) backward digit span,a working memory task which involvesreporting of digits (2 to 8) in the reverseorder; (4) reading, which entails reading ashort descriptive passage aloud. The speechresponses of the patient to the tasks werenormal.

Figure B shows an example of changes inNIRS during the naming task. After thebeginning of the task, oxy-Hb and total-Hbdecreased to negative values during the task,and deoxy-Hb also decreased. These changesreturned to the control level gradually afterthe end of the task. The other tasks alsocaused similar changes of oxy-Hb, total-Hb,and deoxy-Hb.

The rCBF in the left frontal lobe is gener-ally increased by all the tasks used in thepresent study.3–5 Indeed, our NIRS activationstudy using the cognitive tasks showedincreases in oxy-Hb and total-Hb in the leftfrontal lobe in most normal adults—forexample, increases in oxy-Hb and total-Hb—were found in 92.3% of young adult subjects(mean (SD) 28.8 (4.4) years) during theword fluency task (unpublished data). There-fore, although we could not measure thechanges in rCBF in the left frontal lobe of thepatient, the evidence from our previous stud-ies strongly suggests that the tasks caused anincrease in rCBF in the left frontal lobe of thepatient.

The decrease in oxy-Hb and total-Hbrecorded from the brain tumour indicates adecrease of local blood flow in the tumourbecause the NIRS measurement area wasrestricted to the brain tumour.3 The de-creases in oxy-Hb and total-Hb were foundonly during the tasks; consequently, thesechanges were probably not due to changes insystemic blood pressure, which can altertumour blood flow.1 Based on these assump-tions, we suggest that the increase of rCBF inthe left frontal lobe induced by the tasks stolethe local blood flow of the brain tumourthrough the cortical branches, leading to thedecrease of local blood flow in the tumour.

The present report suggests that activitydependent increase in rCBF can steal bloodflow from the adjacent tissues including non-activating cortex. Recent NIRS activationstudies have shown that cognitive tasks causedecreases in oxy-Hb and total-Hb in the leftfrontal lobe in some normal subjects4 5; thesedecreases indicate a decrease in rCBF.5

Although the physiological mechanisms ofthe decrease in rCBF during neuronal activityhave not yet been elucidated, we hypothesise

that a stealing of blood flow is one of themechanisms.4 The present report supportsthis hypothesis.

KAORU SAKATANIHUANCONG ZUO

YENG WANGDepartment of Neurosurgery, China-Japan Friendship

Hospital, Beijing, China

WEMARA LICHTYGroup of Detection and Analysis of Human Body

Movement, Program of BME, Department of ElectricalEngineering, Tsinghua University, Japan

KIYOMI YABUDepartment of Rehabilitation, Takahashi

Neurosurgical Hospital, Japan

Correspondence to: Dr Kaoru Sakatani, Depart-ment of Neurosurgery, China-Japan FriendshipHospital, Yinghua East Rd., Hepingli, Beijing100029, People’s Republic of China. Telephone(fax) 0086 10 64203246; email [email protected]

1 Terada T, Miyamoto K, Hyotani G, et al. Localblood flow changes in malignant brain tumorsunder induced hypertension. Acta Neurochir(Wien) 1992;118:108–11.

2 Fox PT, Raichle ME. Focal physiologicaluncoupling of cerebral blood flow and oxida-tive metabolism during somatosensory stimu-lation in human subjects. Proc Natl Acad SciUSA. 1986;83:1140–4.

3 Hock C, Villringer K, Müller-Spahn F, et al.Decrease in parietal cerebral hemoglobinoxygenation during performance of a verbalfluency task in patients with Alzheimer’sdisease monitored by means of near-infraredspectroscopy (NIRS)-correlation with simulta-neous rCBF-PET measurements. Brain Res1997;755:293–303.

4 Sakatani K, Xie Y, Lichty W, et al. Language-activated cerebral blood oxygenation andhemodynamic changes of the left prefrontalcortex in poststroke aphasic patients: a nearinfrared spectroscopy study. Stroke 1998;29:1299–304

5 Hoshi Y, Onoe H, Watanabe Y, et al. Non-synchronous behavior of neuronal activity, oxi-dative metabolism and blood supply duringmental tasks in man. Neurosci Lett 1994;172:129–33.

Migraine aura masquerading as Balint’ssyndrome

Migraine is a common neurological disorderwith a prevalence of 0.5% to 2% in thegeneral population.1 In one fourth of totalmigraineurs, headache is preceded by anaura.2 We describe a patient with recurrentepisodes of migraine in whom headache waspreceded by a constellation of visual symp-tomatology which constituted salient compo-nents of Balint’s syndrome. This syndrome,consisting of a triad of simultagnosia, opticataxia, and oculomotor apraxia, is seen withbilateral lesions of occipitoparietal corticesaVecting connections between visual corticalregions and the frontal eye field.3

A 29 year old female teacher presentedwith an 8 year history of paroxysmal alternat-ing hemicranial and throbbing headachewhich was often associated with nausea andphotophobia. Patients fulfilled the requisitecriteria for establishing the diagnosis ofmigraine with aura as devised by theInternational Headache Society (1988).4 Sheused to have six to eight episodes of headachea month. There was no history of statusmigranosus during these years. On severaloccasions, headache was preceded by a pecu-liar constellation of visual symptomatologycomprising distortion of visual images fol-lowed by inability to perceive simultaneouslyobjects in the visual field and touch an objectunder direct visual guidance. However, shecould see the component parts of objectsduring the episode. These visual symptoms

lasted for about 10–25 minutes and were fol-lowed by a hemicranial, throbbing headachewhich was often associated with nausea, pho-tophobia, and occasionally vomiting. Head-ache used to last for about 4 to 18 hours andwould respond to either ergot drugs orsumatriptan, especially if taken at the begin-ning of the episode. Occasionally these visualsymptoms were not followed by headache.The patient would not lose contact with theenvironment during or after the visual symp-toms. Her mother and two younger sisterswere also having paroxysmal episodes ofcommon migraine.

Her general physical and neurologicalexamination in between the episodes wasunremarkable. Neurological examinationduring the aura symptoms disclosed that shewas unable to see simultaneously all theobjects in the visual field (simultagnosia). Shedid omit several words while reading aparagraph. However, she could comprehendand read each and every word individually.On being shown a complex picture compris-ing multiple subunits she was not able tocomprehend and perceive the entire picturebut was able to perceive all the components ofthe picture individually (seeing in piecemeal).These aforementioned features were consist-ent with simultagnosia. Besides simultagno-sia, she had optic ataxia as evidenced by herinability to coordinate hand and eye move-ments. Optic ataxia was tested as follows:each eye was tested separately and the handipsilateral to the eye being tested was used.The target stimulus was a 5 mm long pin witha white head placed at preselected locations.The patient was asked to touch this pin withher index finger without shifting her gazefrom the fixation point. The patient had diY-culty in performing this test but had no prob-lems in reaching out to her own body parts oran auditory stimulus with her eyes closed.These features were consistent with opticataxia. Moreover, gaze apraxia was evident byher inability to look at an object oncommand. However, she could do it sponta-neously. In addition, she had impairedsmooth pursuit and voluntary saccades in alldirections. Reflex eye movements were nor-mal. Visual acuity during the episode was 6/6bilaterally. Visual fields were also normalduring the episode as demonstrated by theconfrontation method. Ophthalmological ex-amination, including perimetry performedduring a symptom free period, was normal.There was no clinical evidence of Gerstmannsyndrome, prosopognosia, object agnosia, orcolour agnosia. Her cranial CT and magneticresonance angiography were unremarkable.

Electroencephalography was also non-contributory. The frequency of visual aurasymptoms and headache decreased consider-ably after the patient was started on flunar-izine at a daily dosage of 10 mg at bed time.

The visual impulses, after being received bythe primary visual cortex (Brodmann area17), are interpreted and integrated in visualassociation areas 18 and 19. Brodmann area19, in turn, is connected with the angulargyrus and frontal eye field by virtue of associ-ation fibres. Any lesion in the visual associ-ation areas or their connections would resultin impaired integration of visual impulsesdespite normal visual acuity.

The visual symptom complex in this casepossibly represents an aura of migraine. Thepathogenesis of migraine aura has been adebatable issue.5 in this case it is suggested thatthe pathophysiological process of migraineaura results in a disconnection syndrome by


involving visual association areas and theirassociation pathways. Optic ataxia, gazeapraxia, and simultagnosia seem to represent adissociation of visual information from thefrontal eye field and dorsal parietal regions.

PARVAIZ A SHAHA NAFEE

Division of Neurology, Department of Medicine,Government Medical College and Associated SMHSHospital, Srinagar, Kashmir, J and K 190001, India

Correspondence to: Dr Parvaiz A Shah, Firdousa-bad, Batmaloo, Srinager, Kashmir, J and K 190001,India. Telephone 0091 194 452379.

1 Ziegler DK. Headache: public health problem.Neurol Clin 1990:8:781–91.

2 Campbell JK. Manifestations of migraine. Neu-rol Clin 1990:8:841–55.

3 Damasio AR, Tranel D. Disorders of higherbrain function. In: Rosenberg RN, ed. Compre-hensive neurology. New York : Raven Press,1991:639–57.

4 Headache classification Committee or Inter-national Headache Society. Classification anddiagnostic criteria for headache disorders, cra-nial neuralgias and facial pain. Cephalalgia1988;3(suppl 7):1–96.

5 Blau JN. Migraine: theories of pathogenesis.Lancet 1992;339:1202–6.

“Can’t you use another vaccine”?postrabies vaccination encephalitis

A healthy 39 year old man was bitten on theankle by his own apparently normal dog. Afterthe incident the dog disappeared into theforest and was not seen again. Three days laterthe patient was seen at a provincial hospital inVietnam and started on an alternate day regi-men of suckling mouse brain postrabies expo-sure vaccination (SMBV). After the seconddose, he felt unusually lethargic although hewas still able to work. After the third dose, hebecame unrousable, and was transferred to theCentre for Tropical Diseases, Ho Chi MinhCity, the referral hospital for infectious dis-eases in southern Vietnam. On admission, hewas afebrile, confused, had slurred speech, andhis Glasgow coma score was 13. He had mildspastic weakness of his left face, left arm, andboth legs. Full blood count and results fromroutine biochemistry and chest radiographywere all normal. The CSF: blood glucose ratiowas 0.47 (63/140 mg%), the protein contentwas raised (78 mg/dl), and there was onelymphocyte/ml in the CSF. Screens for malariatoxoplasmosis, cryptococcus, and neurocyst-icercosis were negative, as was a CSF gramstain. The CSF was sterile after 2 weeks ofculture. Brain MRI (Access Toshiba LPT6.01p, 0.064 Tesla) showed areas of high signalthroughout the white matter, and cystic-likechange in the basal ganglion and right cerebel-lar hemisphere (figure A). These variably sizedlesions were bilateral, widely distributed,asymmetric, and showed no evidence ofhaemorrhage or mass eVect.

As paralytic rabies could not be excludedhe was managed conservatively and theSMBV course was continued. On the 4th dayafter admission he deteriorated with aGlasgow coma score of 10, and was inconti-nent of urine and faeces with generalisedspastic paraparesis. Methylprednisolone (500mg/ day) was given for 5 days followed by areducing course of prednisone for a presump-tive diagnosis of postvaccination encephalitis.The SMBV was stopped. Within 72 hours ofstarting steroids there was a dramatic im-provement in his neurological state. An MRIexamination performed 4 weeks later showeda marked decrease in both size and number ofbrain lesions and no new lesions (figure B).After 6 weeks he was discharged talking, eat-

ing, walking, and continent but with somepersistent emotional liability and mildmemory impairment. A follow up MRIexamination 5 weeks after discharge showedfurther improvement, apart from minorabnormalities in the basal ganglion, and gen-eralised increase in ventricular size, consist-ent with residual cerebral atrophy.

Rabies is caused by an RNA virus, a mem-ber of the Rhabdoviridae family, it infectsmammals and can be transmitted to humansby contact, generally from an animal excret-ing the virus in the saliva. Rabies manifests asan acute encephalomyelitis, the developmentof which is almost invariably fatal. Thedistinction between rabies and postvaccineencephalitis is diYcult and may be helped byantigen detection via a skin biopsy; however,this technique is not available in Vietnam.1

Paralytic rabies could not be excluded in thispatient and hence steroids were not used ini-tially. Steroids have been reported to increasemortality in experimental animals with ra-bies, and it has been suggested that they mayabrogate the immune response to the postex-posure vaccine, thus precipitating uncon-trolled rabies.2

There are three types of postexposure vac-cine in use world wide. The Semple type(STV) is obtained from inactivated virus pre-pared on adult animal nerve tissue; it is inex-pensive and relatively easy to produce. InIndia 3 million people receive postexposurecourses of STV (phenolised sheep brain)antirabies vaccine each year.1 These produce

neurological reactions, including postvacci-nation encephalomyelitis, in up to 1 in 220courses, with a 3% mortality.3 Clinical formsinclude a reversible mononeuritis multiplex,and meningoencephalitic and encephalomy-elitic reactions. Myelin basic protein andrelated neural proteins from the nervoustissue of the animal on which the virus wascultivated stimulate an autoimmune reactionin the human nervous system.

Tolerance has been improved by the devel-opment of the suckling mouse brain vaccine(SMBV). The attenuated virus is cultured onimmature mouse brain tissue, which containslittle myelin, thus reducing the risk of compli-cations. SMBV is inexpensive (US$1.5 pertreatment course) and easily manufacturedlocally; it is the most widely used postexposurevaccine in Vietnam. Rare neurological reac-tions do occur with SMBV, Complications ofthe CNS have been reported to occur aftervaccination with an incidence of 1:27000treated people, with a 22% mortality4 Themortality was particularly high (90%) if therewas extensive CNS involvement. The thirdtype of vaccine available is the human diploidcell tissue culture vaccine (HDCV), which isboth safe and eYcacious. However, the recom-mended regimen is not aVordable in mostdeveloping countries.

When we approached the Rabies Labora-tory, Ministry of Agriculture and Fisheries,United Kingdom for advice in this case theircomment was “why do you use the SMBV,can’t you use another vaccine”. Worldwideabout 10 million people each year receiverabies vaccine after exposure; at the Centrefor Tropical Diseases we treat 3000 peoplewith dog bites annually. The cost of anHDCV in Vietnam, administered in itspresent regimen (1ml given for 5 days on days0, 3, 7, 14, and 28 with an optional booster onday 90) is US$ 125, making the use of thisvaccine unaVordable.

This is the first report to show the demyeli-nating CNS lesions on MRI, and theirresolution after steroid therapy. It is relativelyrare for patients to survive if they developsevere CNS eVects after postexposure rabiesvaccination. Although the incidence of reac-tions to SMBV is very much lower than STV,this report confirms that it does still occur.Both SMBV and STV are widely usedthroughout the developing world, and wouldbe the vaccine administered to travellersexposed to animal bites in such countries.This case stresses the need for high dose ster-oids in postexposure vaccine encepahlitis andthe urgent need for the development anddeployment of a safe, and critically, aVord-able postrabies exposure vaccine regimen.The economic low dose multisite intradermalregimen using the HDCV provides an exam-ple of how this goal may be achieved althoughit is not yet widely accepted. Such a vaccineregimen (0.1 ml HDCV given at multisiteinjections on days 0, 7, 28, and 90) could bemade aVordable, and oVers excellent protec-tion without the risks of postexposure im-mune mediated encephalitis.5

N V V CHAUT T HIEN

Centre for Tropical Diseases, 190 Ben Ham Tu, District5, Ho Chi Minh City, Vietnam

R SELLARDepartment of Clinical Neurosciences, Western General

Hospital, Edinburgh, UK

R KNEENJ J FARRAR

Brain MRI in May 1997. (A) T2 weightedimage showing multiple areas of high signal inthe cerebral white matter. Bilateral subcorticaland periventricular lesions are seen. (B) BrainMRI in July 1997. T2 weighted image showsresolution of the white matter lesions.


Wellcome Trust Clinical Research Unit, Centre forTropical Diseases, 190 Ben Ham Tu, District 5, Ho

Chi Minh City, Vietnam

R KNEENJ J FARRAR

Centre for Tropical Medicine, NuYeld Department ofClinical Medicine, John RadcliVe Hospital, University

of Oxford, UK

Correspondence to: Dr J J Farrar, Wellcome TrustClinical Research Unit, Centre for Tropical Dis-eases, 190 Ben Ham Tu, District 5, Ho Chi MinhCity, Vietnam

1 Warrell MJ, Warrell DA. Rhabdoviruses: rabiesand rabies related viruses. In: Weatherall DJ,Ledingham JGG, Warrell DA, eds. Oxford text-book of medicine. 3rd ed. Oxford: OxfordUniversity Press, 1996:394–405.

2 Fishbein DB, Bernard KW. Rabies virus. In:Mandel GL, Douglas RG, Bennett JE, eds.Principle and practice of infectious diseases.4th ed.New York: Churchill Livingstone, 1995:1527–43.

3 Swaddiwudhipong W,Prayoonwiwat N, Kuna-sol P, et al. A high incidence of neurologicalcomplications following Semple anti-rabiesvaccine. Southeast Asian J Trop Med PublicHealth 1987;18:526–3.

4 Toro G; Vergara I; Roman G. Neuroparalyticaccidents of antirabies vaccination with suck-ling mouse brain vaccine. Clinical and patho-logic study of 21 cases. Arch Neurol 1977;34:694–700.

5 Warrell MJ, Nicholson KG, Warrell DA, et al.Economical multiple-site intradermal immuni-sation with human diploid cell strain vaccine iseVective for post-exposure rabies prophylaxis.Lancet 1985;i:1059–62.

Leukoencephalopathy associated withkhat misuse

The leaves of the tree Catha edulis, or khat(also qat and kat) are chewed by a large pro-portion of the adult population of the Yemen,and throughout Saharan and sub-SaharanAfrica. The leaves are also chewed bymembers of the Yemeni and Somali commu-nity in the United Kingdom.1 The psychoac-tive constituents of khat are cathin (d-norisoephedrine), cathidine, and cathinone(an alkaloid with a structure resemblingephedrine and amphetamine) and usersreport a mild euphoria similar to that ofamphetamine.1 Khat is acknowledged as aprecipitant of psychosis and has also beenreported to cause cognitive impairment.2 Wereport a case in which khat chewing has beenassociated with a severe and disabling neuro-logical illness.

A 56 year old Somali living in the UnitedKingdom for the past 18 years was admitted toa psychiatric hospital with a 5 week history ofprogressive confusion and agitation. His familyreported that he had been chewing khat, intheir opinion to excess, every day during thattime but had stopped 2 days before admission.There was one previous admission to hospital9 months previously with khat induced psy-chosis, from which he recovered without com-plications within 24 hours. On this occasion,shortly after admission, his conscious leveldeteriorated abruptly and he was referred forneurological opinion. He was apyrexial andgeneral medical examination was normal. Heopened his eyes spontaneously but there wasno verbal response and he did not obeycommands. He withdrew all four limbs topain. Upper and lower limbs were held in flex-ion with markedly increased tone. Reflexeswere brisk but equal. The right plantar wasextensor. There were bilateral palmomentaland grasp reflexes.

Full blood count, urea and electrolytes,glucose, liver function tests, thyroid functiontest, viral serology, and malaria screen all gave

normal results. Tests for HIV antibody,serum angiotensin converting enzyme, whitecell enzymes, and serum and urinary porphy-rins were negative. Erythrocyte sedimenta-tion rate on admission was 58 mm/h.

Examination of the CSF showed normalopening pressure, protein 0.27 g/l, glucose4.3 mmol/l (blood glucose 6.1 mmol/l), andno cells. His initial EEG was abnormal withdiVuse slow waves indicative of widespreadcerebral dysfunction.

A chest radiograph and ultrasound exam-ination of the abdomen were normal. CranialMRI, although contaminated by movementartefact, showed diVuse abnormality in thedeep cerebral white matter of both cerebralhemispheres. Fourteen days after admission hewas witnessed to have a single brief adversiveseizure with eye and head deviation to theright.

The patient was admitted to a rehabilita-tion unit. His mini mental state examinationscore and Barthel scores were zero. Feedingby percutaneous gastrostomy was started. Atrial of intravenous methylprednisolone (1 gon 3 consecutive days) gave no benefit.Repeated EEGs (on four occasions) showeddiVuse slow waves only. A second MRI(figure) 3 months after onset of symptomshowed the presence of a continuing diVuseextensive abnormal signal in the deep whitematter of both cerebral hemispheres withmarked cortical atrophy. Brain biopsy (viaright frontal craniotomy) was performed 3months after the onset of his illness. Therewas no evidence of acute inflammation,vasculitis, or infarction.

While undergoing rehabilitation there hasbeen slow improvement in his cognitive andlocomotor function. After 1 year he is able toopen and close his eyes, occasionally verbal-ise, localise pain, and obey simple commands.His plantars are flexor but he has persistentgrasp and palmomental reflexes. His nutri-tion is maintained by gastrostomy and he hasan indwelling catheter.

The clinical presentation, EEG, and MRIfindings suggest a rapidly progressive leu-koencephalopathy. There are no previousreports of leukoencephalopathy in associationwith khat or amphetamine misuse; it has,however, been reported in association withother recreational drugs taken by mouth orinhalation.3 4 An alternative for this man’s

presentation is a necrotising vasculitis, a welldescribed complication of oral amphetaminemisuse.5 The clinical features, MRI appear-ance, brain biopsy, absence of haemorrhage,and lack of response to steroids make thisunlikely.

The likely precipitant of this man’s illnessseems to be his use of khat. A drug screen onadmission was negative, and his family deniedmisuse of other drugs. It remains possiblethat the sample of khat chewed by this manwas contaminated. We are unaware of anyprevious reports of khat misuse with severeneurological deterioration; previous casesmay not have been investigated or reported.In reporting this case our intention is to alertothers to a possible complication of the mis-use of this drug. Evidence of other caseswould provide a powerful argument for therestriction of import and sale of khat.

P K MORRISHN NICOLAOU

P BRAKKENBERGP E M SMITH

Department of Neurology, University Hospital ofWales, Heath Park, CardiV CF4 4XN, UK

Correspondence to: Dr PK Morrish, Departmentof Neurology, University Hospital of Wales, HeathPark, CardiV CF4 4XN, UK. Telephone 0044 1222747747; fax 004 1222 744166; email:[email protected]

1 Pantelis C, Hindler CG, Taylor JC. Use andabuse of khat (Catha edulis): a review of thedistribution, pharmacology, side eVects and adescription of psychosis attributed to khatchewing. Psychol Med 1989;19:657–68.

2 Khattab NY, Galal A. Undetected neuropsycho-logical sequelae of khat chewing in standardaviation medical examination. Aviat SpaceEnviron Med 1995;66:739–44.

3 Celius EG, Andersson S. Leucoencephalopathyafter inhalation of heroin: a case report. J Neu-rol Neurosurg Psychiatry 1996;60:694.

4 Walters EC, van Wijngaarden GK, Stam FC, etal. Leucoencephalopathy after inhaling“heroin” pyrolysate. Lancet 1982;ii:1233–7.

5 Salanova V, Taubner R. Intracerebral haemor-rhage and vasculitis secondary to amphetamineuse. Postgrad Med J 1984;60:429–30.

Necrotising vasculitis with conductionblock in mononeuropathy multiplexwith cold agglutinins

Cold agglutinins are cold reactive autoanti-bodies that have haemolytic eVects on redblood cells mediated via complement fixa-tion. Neuropathy associated with cold agglu-tinins has been described,1–5 however, detailsof its pathomechanism are unclear. Here, wereport the clinical, electrophysiological, andpathological findings of a mononeuropathymultiplex in a patient with cold agglutinins,who responded very well to plasmapheresis.

A 72 year old man was admitted with a 1month history of progressing dysaesthesia andweakness of the limbs. He had no anaemia,jaundice, hepatosplenomegaly, or lymphaden-opathy. Cranial nerves and the cerebellumwere not involved. There was severe weaknessand atrophy of bilateral thenar, interossei, andplantar muscles with severe dysaesthesia ofboth palms and plantaris. Pin prick and lighttouch were reduced as well as position andvibratory sensation in both hands and feet.Deep tendon reflexes were hypoactive. Babin-ski’s sign was negative.

Laboratory investigation showed a raisederythrocyte sedimentation rate: 52 mm/hour(normal <10) and serum C reactive protein:1.8 mg/dl (normal; < 0.5). Blood cell countswere within normal limits. The followingwere normal or negative; IgG, IgA, IgE, IgM,

Cranial MRI 3 months after onset of symptomsshowing diVuse signal abnormality in the deepwhite matter of both cerebral hemispheres. Thereis also marked cortical atrophy.


M-protein, direct and indirect Coombs tests,cryoglobulin, antibodies to mycoplasma, my-elin associated glycoprotein, gangliosides(GM1, GD1b, asialo-GM1, GT1b, GQ1b,Gal-C), P-ANCA, and C-ANCA. The CSFwas normal. Titre of cold agglutinins wasdetectable at 1:1024 at 4°C (normal;<1:256). The patient’s serum agglutinatedadult group OI-red blood cells, but notOi-red blood cells or human cord red bloodcells, signifying cold agglutinins with Ispecificity. Immunoelectrophoresis of theeluate confirmed IgM composition.

The initial nerve conduction study showedsevere diminution or absence of compoundmuscle action potentials (CMAPs) withmildly diminished conduction velocities. Fwave latencies were mildly prolonged. Therewere no evoked sensory nerve action poten-tials (SNAPs) in median, ulnar, and suralnerves bilaterally. Electromyographic studiesof the aVected muscles showed moderateneurogenic changes, but there were no fibril-lation potentials except in the left anteriortibialis muscle. Sural nerve biopsy wasperformed. Epineurial vessels were sur-rounded by mononuclear cell infiltrates(figure A). Some vessels had focal necrosis oftheir wall. The small vessels in the endoneu-rium and epineurium showed slugging of redblood cells. The densities of large and smallmyelinated fibres were markedly decreased(diameter<5 µm: 1504/mm2, diameter >5µm:708/mm2, total: 2212/mm2)(figure B).Teased fibre analysis showed that 90% of thefibres were undergoing axonal degeneration.

Oral prednisolone (30–50 mg/day) for 4weeks reduced the erythrocyte sedimentationrate and C reactive protein, but not the serumtitre of cold agglutinins; neither was there anyimprovement of symptoms. He received mas-sive dose intravenous corticosteroid therapy.This moderately improved the muscle strengthand sensory disturbance. Follow up nerveconduction studies (71 days after the initialstudy) suggested conduction block of the rightmedian nerve on the forearm (CMAP, dura-tion at the wrist: 2.76 mV, 8.4 ms; CMAP,duration at the elbow: 1.87 mV, 8.8 ms),whereas CMAP could not be elicited in theinitial study. We adapted the following criteriato define conduction block: <15% change induration and >20% fall in negative peakamplitude between proximal and distal sites bypercutaneous supramaximal stimulation ofmotor nerves. As the conduction block mightdelay smooth recovery of symptoms, Doublefiltration plasmapheresis was performed fourtimes. After the second plasmapheresis, dys-

aesthesia and muscle strength improved re-markably.The titre of cold agglutinins wasreduced to 1:64. The motor nerve conductionvelocity (MCV) of the right median nerve like-wise improved (pretreatment; 40.0 m/s, post-treatment; 57.0 m/s). Double filtration plas-mapheresis was followed by oral azathioprine(50 mg/day) with tapering of steroid. He wasdischarged on prednisolone (20 mg/day). Inthe subsequent 4 years, he has had mild exac-erbation of dysaesthesia that responded tointermittent steroid therapy.

Characteristic features of the present caseare as follows: (1) subacute onset of monone-uropathy multiplex; (2) necrotising vasculitiswith marked loss of myelinated fibres; (3)probable conduction block in the mediannerve; (4) increased concentrations of serumtitres of cold agglutinin; and (5) markedresponse to plasmapheresis. Extensive inves-tigation for other causes of neuropathy wasnegative except for an increased serumconcentration of cold agglutinins, whichstrongly suggests that cold agglutinins mayplay an important part in the induction ofneuropathy in this case.

Six patients with neuropathy associated withcold agglutinins have been reported1–5 includ-ing our patient. Cold agglutinins are cold reac-tive autoantibodies that react with the anti-genic determinant termed I/i or Pr present onglycoproteins and glycolipids in erythrocytemembranes. Arai et al1 reported a case ofpolyneuropathy and IgMê M proteinemiawith anti-Pr2 CA activity. IgM M protein crossreacted with sialosyl paragloboside, GT1b,GD1a, GD1b, GM3, and GD3 present inmyelin and in endothelial cells of the periph-eral nervous system. It has been speculatedthat anti-Pr2 IgM protein induced immunemediated damage to vascular endothelium andperipheral nervous system myelin. A similarpathomechanism has been postulated in theother cases.2–3 However, necrotising vasculitishas never been reported in neuropathy withcold agglutinins. This is the first demonstra-tion of vasculitic neuropathy with cold aggluti-nins. Although the mechanism for neuropathywith cold agglutinins is unknown, mechanismssimilar to those in cryoglobulinaemic neu-ropathy have been postulated.4 The hypothesesare (1) immunologically mediated demyelina-tion; (2) ischaemic injury secondary to slug-ging or agglutination of red blood cells in thevasa nervorum; and (3) an associated vasculi-tis. In the present case, we have confirmed thenecrotising vasculitis and probable conductionblock. Pathophysiological explanations forassociation of vasculitis and conduction block

may be as follows. Firstly, conduction blockmay occur as a consequence of nerve ischae-mia due to small vessel occlusion. There havebeen reports of conduction block occurring invasculitic neuropathy which support this possi-bility. Secondly, humoral factors includingcold agglutinins may induce immune medi-ated demyelination in the peripheral nervoussystem. Taken together, neuropathy with coldagglutinins may involve immunologically me-diated demyelination, microcirculation occlu-sion, and vasa nervorum vasculitis. The diver-sity of pathomechanisms may come from thediVerence target antigens recognised by coldagglutinins. Plasmapheresis proved eVective inall cases. These findings strongly suggest thathumoral factors including cold agglutininsmay play an important part in the induction ofneuropathy with cold agglutinins. We re-commend plasmapheresis as first choice treat-ment for neuropathy associated with coldagglutinins.

We thank Dr Gerard Salazar for critical reading ofthe manuscript, Ms M Teshima and N Hirata fortheir technical assistance, Dr S Kusunoki (Depart-ment of Neurology, Institute for Brain research,University of Tokyo) for analyses of antibodies togangliosides, and Mr H Moug (Division of BloodTransfusion Medicine, University of Kagoshima)for characterization of cold agglutinin.

R OTSUKAF UMEHARAK ARIMURA

Y MARUYAMAY ARIMURA

M OSAMEThe Third Department of Internal Medicine,

Kagoshima University School of Medicine,Sakuragaoka 8–35–1 Kagoshima, Japan

Correspondence to: Dr R Otsuka, The ThirdDepartment of Internal Medicine, KagoshimaUniversity School of Medicine, Sakuragaoka8–35–1 Kagoshima, Japan. Telephone 0081 99 2755332; fax 0081 99 265 7164; [email protected]

1 Arai M, Yoshino H, Kusano Y, et al. Ataxicpolyneuropathy and anti-Pr2 IgMê M pro-teinemia. J Neurol 1992;239:147–51.

2 Willison HJ, Paterson G, Veitch J, et al. Periph-eral neuropathy assaciated with monoclonalIgM anti-Pr2 cold agglutinins. J Neurol Neuro-surg Psychiatry 1993;56:1178–83.

3 Herron B, Willison HJ, Veitch J, et al. Mono-clonal IgM cold agglutinins with anti-Pr1dspecificity in a patient with peripheral neu-ropathy. Vox Sang 1994;67:58–63.

4 Thomas TD, Donofrio PD, Angero J. Peripheralneuropathy in cold agglutinin disease. MuscleNerve 1991;14:331–4.

5 Valbonesi M, Guzzini F, Zerbi D, et al. Success-ful plasma exchange for a patient with chronicdemyelinating polyneuropathy and cold agglu-tinin disease due toanti-Pra. J Clin Apheresis1986;3:109–10.

(A) Sural nerve (toluidine blue staining) showing epineurial vessel surrounded by mononuclear cell infiltrates. Note fibrin deposition (arrow) and necrosisin media. (bar=20 µm). (B) Most of myelinated fibres are undergoing axonal degeneration. Many macrophages containing myelin debris infiltrate theendoneurium. (bar=30 µm).


CORRESPONDENCE

The cholinergic hypothesis ofAlzheimer’s disease: a review ofprogress

I read with interest the review of Francis et alregarding the progress of the cholinergichypothesis of Alzheimer’s disease.1 Theymentioned that donepezil produced improve-ment or no deterioration in more than 80% ofpatients, and that such responses should beviewed positively considering the progressive,degenerative nature of the disease. Variousdonepezil manufacturer’s medical repre-sentatives presenting data from a clinicalstudy2 also commonly use this statement.However, this only partially reveals the truth.In fact, the same study produced improve-ment or no deterioration in 59% patients onplacebo. I think that the beneficial eVect ofdonepezil in particular clinical trials shouldalways be critically reviewed in comparisonwith placebo. In addition, as both 24 weekplacebo controlled donepezil trials performedso far excluded patients with behavioural dis-turbances, my impression is that the positiveeVect of donepezil on the symptoms ofbehavioural disturbances still remains con-troversial. In fact there are reports thatdonepezil might induce behavioural distur-bances in patients with Alzheimer’s disease.3 4

Therefore I would be extremely cautiousabout prescribing donepezil to patients withAlzheimer’ s disease accompanied by behav-ioural disturbances.

Finally, donepezil was never investigated in a30 week randomised double blind study as wasmentioned in the review. The authors areprobably referring to the randomised 24 weekdouble blind placebo controlled trial with anadditional 6 week single blinded placebophase.

T BABICDepartment of Neurology, Medical School University

of Zagreb, Kisaticeva 12, 10000 Zagreb, Croatia.Telephone 00385 1 217280, fax 00385 1 217280,

email [email protected]

1 Francis PT, Palmer AM, Snape M, et al. Thecholinergic hypothesis of Alzheimer’s disease: areview of progress. J Neurol Neurosurg Psychia-try 1999;66:137–47.

2 Rogers SL, Farlow MR, Doody RS, et al. A24-week, double-blind, placebo-controlled trialof donepezil in patients with Alzheimer’sdisease. Donepezil Study Group. Neurology1998;50:136–45.

3 Wengel SP, Roccaforte WH, Burke WJ, et al.Behavioural complication associated withdonepezil. Am J Psychiatry 1998;155:1632–3.

4 Bouman WP, Pinner G. Violent behaviour-associated with donepezil. Am J Psychiatry1998;155:1626–7.

The authors reply:We thank Professor Babic for the letter, whichraises several interesting points. We agree thatit may be more helpful to put the resultsattributed to treatment with donepezil in thecontext of the placebo response. In general,looking at this as a class eVect in relation toseveral compounds, the picture emerging isthat about twice as many people obtain aresponse to active treatment as to that withplacebo. The high placebo response is a com-

mon factor in most studies in this field and isworthy of some explanation in its own right.Although it seems that these studies comparedrug treatment with that of a placebo (onetreatment against no treatment), the reality isthat it is a comparison of patients receivingtwo treatments against other patients who arereceiving one form of treatment. The addi-tional treatment regime is, of course, the careand attention that they receive by being partof the clinical study, which often seems tohave an impact, not just on the patient butalso on their main carer or carers.

As far as behavioural disturbances areconcerned, however, our review was makingthe point that evidence is emerging from clini-cal trials to suggest that cholinomimetic drugsas a whole may have a beneficial eVect on somenon-cognitive behavioural symptoms. This hasnow been reported for at least two cholinest-erase inhibitors, and two muscarinicagonists.1–5 In particular, a clear link is emerg-ing between psychotic symptoms and cholin-ergic dysfunction. Thus, Bodick et al2 haveshown that the M1/M4 agonist xanomelinecauses a dose dependent reduction in halluci-nations, agitation, and delusions in a 6 monthrandomised double blind placebo controlled,parallel group trial. In addition, Cummingsand Kaufer6 have shown that the cholinest-erase inhibitor tacrine is more eVective inreducing psychotic features than cognitive dis-turbances; tacrine also reduces or abolisheshallucinations in Parkinson’s disease.7 Anothercholinesterase inhibitor, metrifonate, was alsoshown to reduce the number of hallucinationsin a 26 week randomised, double blind,placebo controlled safety and eYcacy study inpatients with Alzheimer’s disease. Furthersupport for a link between acetylcholine andpsychosis derives from postmortem data show-ing that the activity of choline acetyltransferasein the temporal cortex of patients with Lewybody dementia was lower in those patientswith hallucinations than in patients withoutthis feature.8 Finally, in animals the partialM2/M4 agonist (5R,6R)6-(3-propylthio-1,2,5-thiadiazol-4-yl)-1-azabicyclo[3.2.1]octaneproduced a preclinical profile suggestive ofantipsychotic eYcacy9 and that the psychomi-metic NMDA receptor antagonist ketamine(when administered at subanaesthetic doses)reduced brain concentrations of acetylcho-line.10 Thus, on the basis of both clinical andpreclinical data, a clear rationale is emergingfor prescribing cholinomimetic agents fortreating the non-cognitive behavioural symp-toms associated with dementia, particularlypsychosis.

Professor Babic is also correct in identify-ing two of the studies referred to as the 30week randomised multicentre placebo con-trolled parallel group studies, which includeda 24 week double blinded treatment phase.

We are grateful to your correspondent for providingus with the opportunity to clarify these points.

PAUL T FRANCISNeuroscience Research Centre, GKT School of

Biomedical Science, King’s College London, LondonSE1 9RT, UK

ALAN M PALMER

MICHAEL SNAPECerebrus Pharmaceuticals Ltd, Winnersh, Wokingham,

RG41 5UA, UK

GORDON K WILCOCKDepartment of Care of the Elderly, Frenchay Hospital,

Bristol, BS16 2EW, UK

1 Kaufer DI, Cummings JL, Christine D. EVect oftacrine on behavioral symptoms in Alzheimer’sdisease: an open label study. J Geriatr PsychiatrNeurol 1996;9:1–6.

2 Bodick N, OVen W, Levey AI, et al. EVects ofxanomeline, a selective muscarinic agonist, oncognitive function and behavioral symptoms inAlzheimer’s disease. Arch Neurol 1997;54:465–73.

3 Cummings JL, Back C. The cholinergic hypo-thesis of neuropsychiatric symptoms inAlzheimer’s disease. Am J Geriatr Psychiatry1998;6:S64–78.

4 Kaufer DI, Catt K, Pollock BG, et al. Donepezilin Alzheimer’s disease: relative cognitive andneuropsychiatric responses and impact on care-giver distress. Neurology 1998;19(suppl 4):S182.

5 Morris JC, Cyrus PA, Orazem J, et al. Metri-fonate benefits cognitive, behavioral and globalfunction in patients with Alzheimer’s disease.Neurology 1998;50:1222–30.

6 Cummings JL, Kaufer D. Neuropsychiatricaspects of Alzheimer’s disease: the cholinergichypothesis revisited. Neurology 1996;47:867–83.

7 Hutchinson M, Fazzini E. Cholinesterase inhi-bition in Parkinson’s disease. J Neurol Neuro-surg Psychiatry 1986;61:324–5.

8 Perry EK, Marshall E, Kerwin J, et al. Evidenceof a monoaminergic-cholinergic imbalancerelated to visual hallucinations in Lewy bodydementia. J Neurochem 1990;55:1454–6.

9 Bymaster FP, Shannon HE, Rasmussen K, et al.Unexpected antipsychotic-like activity with themuscarinic receptor ligand (5R,6R)6-(3-propylthio-1,2,5-thiadiazol-4-yl)-1-azabicyclo[3.2.1]octane. Brain Res 1998;795:179–90.

10 Hong CH, Woo JI, Suh YH, et al. EVect ofketamine on the acetylcholine concentration ofvarious regions of rat brain. Seoul Journal ofMedicine 1987;28:347–51.

BOOK REVIEWS

Clinical Management of DiabeticNeuropathy. Edited by ARISTIDIS VEVES. (Pp348, US$125). Published by The HumanaPress, New Jersey, 1998. ISBN0-896-03528-X.

The neuropathies of diabetes are common (asthe chapters in this book repeatedly remindus) and can be very disagreeable. Symptom-less neuropathy underlies foot ulceration andsepsis as the commonest clinical consequenceof diabetic neuropathy but other extremelyunpleasant disorders range from exception-ally severe pain to the whole range ofproblems resulting from autonomic failure.This book comprehensively covers everyaspect of the subject, systematically (and attimes exhaustively) from its epidemiologyand pathogenesis (exhaustingly) to struc-tural, functional, and clinical problems andtheir treatment. Most of the authors are wellknown in the field and their accounts are upto date and authoritiative.

Unfortunately, struggle as they might, allauthorities have diYculty in defining whatthey mean by diabetic neuropathy and, in thisregard, understanding of this complicationboth in clinical and pathological terms, aswell as with regard to treatment, lags farbehind that of the other classic diabetic com-plications, nephropathy and retinopathy.Even its classification presents problems andattempts to do so are found in four diVerentchapters, describing four classifications. Rep-etition is an unfortunate feature of this bookand—quite apart from the confusion overclassification—aspects of pathogenesis, struc-tural changes, epidemiology, diagrams, andsome reference to treatment (for example,that of pain) appear repeatedly in diVerentchapters in greater or lesser detail.


This is certainly a book for the specialistand not at all (as the preface suggests) for thefamily practitioner. There are good reviews ofnerve structure, causation, and treatment ofpainful neuropathies and focal neuropathies.The comprehensive survey of the DiabetesControl and Complications Trial (DCCT)shows in detail the only treatment which istruly eVective (diabetic control); and thelengthy description of aldose reductase in-hibitor trials establishes that, even after morethan two decades of investigation, furthertrials are still needed.

Clinical evaluation of somatic and auto-nomic neuropathies are useful and also, tosome extent, comprehensive but lackspecificity—that is, normal values for simpletests are diYcult to find. The huge subject ofthe diabetic foot is covered in these chaptersand “the impact of micro and macrovasculardisease” is compressed into the last ninepages of the book.

The bibliography is important and oftenvery up to date with references ranging from33 to 283 per chapter.

If this book is at times confusing, this reflectsthe confusion regarding the nature and treat-ment of the diabetic neuropathies as much asthe overlap and repetition found in its differentchapters. It is a book of reference for thespecialist who will be well served by the com-prehensiveness of some of its reviews and theirassembly of the appropriate literature.

PETER WATKINS

Advanced Neurosurgical Navigation. ByEBEN ALEXANDER III and ROBERT J MACIUNAS.(Pp 605, DM398.00). Published by Thieme,New York, 1999. ISBN 3-13-115391 1.

The quest for a means of accurate localisationof structures during neurosurgery has taxedthe minds of clinicians from early in thehistory of the specialty, starting with Zernov’sencephalometer more than a century ago.Just as the solution to the mariners’ problemof determining longitude from which it partlytakes its name, neuronavigation ("the sur-geon’s sextant”) has relied on the advent ofnew technologies to provide solutions to anage old puzzle.

Advances In Neuronavigation begins by trac-ing the history of stereotaxis from a Cartesiancoordinate system devised by Clarke andHorsley at the beginning of this century,through ventriculography, stereotactic brainatlases, and CT/MR frame based stereotaxis.The final part of the first section discusses theroots of image guided frameless stereotaxisthrough the integration of high speed graph-ics computers, informatics, biotechnology,and robotics.

The remainder of the text is divided intofour sections. The first concerns the creationof maps from CT, MRI, MRA, PET, andvarious types of functional imaging. The fol-lowing section discusses clinical applicationsof stereotaxis, beginning with diVerent au-thors’ experiences of their own favouredframes, the biopsy of diYcult lesions such asthose in the brainstem or posterior fossa, andfinally experience with diVerent image guid-ance systems and their integration with theoperating microscope and endoscope. Therethen follows a series of chapters devoted to

radiosurgery, and to image guidance inepilepsy and functional surgery. The finalsection is entitled Frontiers in NeurosurgicalNavigation and considers, among other top-ics, intraoperative MRI, telepresence in neu-rosurgery, and robotics.

The incorporation of new technology islikely to alter surgical practice radically overthe coming decade and equipment thatseemed at the cutting edge of technology onlya few years ago, such as the mechanical arm,has already passed into near obsolescence at abewildering rate. This volume provides anexcellent account of the developments whichhave occurred in neuronavigation, and athought provoking insight into the widerapplications of equipment of which many ofus use only a fraction of the potential capabil-ity. The title of the book should perhaps haveincluded the word cranial, as there is almostno discussion of the impact that this technol-ogy has had in surgery of the spine. This asideit is an excellent book although, like the tech-nology it chronicles, one which is likely todate quite rapidly.

ROBERT MACFARLANE

Key Topics in Neurology. By PEM SMITH.(Pp 318, £22.95). Published by BIOSScientific Publishers Ltd, Oxford, 1998.ISBN 1-85996-261-0.

The title and back cover of the latest additionto Neurology Lite texts contains the usualproclamations. “Concise, key topics, revisionaid, essential, review”... the well trailedsoundbites demanded by the consumer in theincreasingly competitive market of “read less- learn more” books. This book, however, isunusual and distinct. Unlike many rivals it isnot an A5 facsimile of a superior parent A3reference tome. Brevity, so essential to thesuccess of an overview work, has sacrificedneither clarity nor clinical relevance. Thestrength of Key Topics in Neurology owes muchto the author’s ability to negotiate skilfully thecompromises necessary for a successfuldistillation of a large and complex field. Hehas not shied from wholesale culling ofneurological ballast. The allied ability to dis-tinguish and highlight the salient and relevantfrom the obscure and historical allows thissmall book to be surprisingly thorough in itscoverage and topicality. There is suYcient upto date information on most areas of neurol-ogy such that this book would be useful forspecialist registrars albeit without the detailor embellishment they seek. In terms of theaims of this book such observations must beregarded as complimentary.

My limited criticisms relate to details oflayout and presentation. I found the exclusivealphabetical arrangement of chapters mildlydisorientating in that, for example, Historytaking in Neurology is to be found at p 131.Similarly, the absence of diagrams and tablesis an unexpected omission as I would imaginethat this would have complemented the over-all style of the book. These are minor gripesof what in print largely matches the sleevehype and with a price tag of just £27-50 thebook will be welcomed by undergraduatesthrough to specialist registrars.

SIDDHARTHAN CHANDRAN

Readers may be interested in:

States of Mind. Edited by ROBERTA CONLAN.(Pp214, £19.99). Published by John WileyAnd Sons, Chichester, 1999. ISBN0-471-29963-4.

First Episode Psychosis. By K AITCHISON,KM MEEHAN, and R MURRAY. (Pp 152, £19.95).Published by Martin Dunitz, London, 1999.ISBN 1-85317-435-1.

Endoscopy in Neuro-otology. Edited By J

MAGNAN and M SANNA. (Pp 101, DM198).Published by Georg Thieme Verlag,Stuttgart, 1999. ISBN 313-113061-X.

Advances in Biological Psychiatry Vol 19:New Models For Depression. Edited by D

EBERT and K P EBMEIER. (Pp 204, US$170.50).Published by Karger, Basel, 1998. ISBN3-8055-6698-0.

Screening for Brain Dysfunction inPsychiatric Patients. By COOPER B HOLMES.(Pp 136, US$36.95). Published by Charles CThomas, Illinois, 1998. ISBN 0-398-06921-2.

Management of Depression. By GIN S

MALHI and PAUL K BRIDGES. (Pp 136, £19.95).Published by Martin, Dunitz Publishers,London 1998. ISBN 1-85317-547-1.

Clinical Research in Psychiatry. APractical Guide. Edited by STEPHEN

CURRAN AND CHRISTOPHER J WILLIAMS. (Pp185, £17.99). Published by ButterworthHeinemann, Oxford 1999. ISBN 075064073 1.

CORRECTION

K Sudo, N Fujiki, S Tsuji, M Ajiki, THigashi, M Niino, S Kikuchi, F Moriwaka, KTashiro.Focal (segmental) dyshidrosis in syringomy-elia. J Neurol Neurosurg Psychiatry1999;67:106-8. During the editorial processthe footnote to table 1(p 107) was wronglytranscribed.The last line—¶p value for eachpair of items: hyperhydrosis v hyperhydrosis0.0007; hypohydrosis v normohydrosis0.7282; normohydrosis v hypohydrosis0.0012 should read—¶p value for each pairof items: hyperhydrosis v hypohydrosis0.0007; hyperhydrosis v normohydrosis0.7282; normohydrosis v hypohydrosis0.0012.


focal (segmental) dyshidrosis in syringomyelia · onset of current dyshidrosis and duration from...

Documents