I!Jr@ Lt

Editorial

Lewy Body Dementia - Diagnosis and Treatment

IAN G. McKEITH,DOUGLASGALASKO,GORDONK. WILCOCKand E. JANE BYRNE

British Journal of Psychiatry (1995), 167, 709—717

In the foreword to a volume entitled RecentAdvances in Psychogeriatrics published in 1992,Tom Arie remarked upon the suddenappearanceofLewy body dementia (LBD) on the clinical scene.A diagnosis not previously found in the long list ofcauses of dementia was apparently being proclaimedas possibly the second most common cause ofdementia in the elderly. There has followedconsiderable debate as to how this new conceptshould be handled. Points of uncertainty ordisagreement include whether or not preciseneuropathological criteria of LBD have beenestablished, and, if so, whether such patients havea distinguishableclinicalsyndrome. If LBD isidentifiable ante-mortem, how should patients andtheir relatives be advised and what implications arethere for management?

At least part of the heterogeneity of opinion aboutLBD probably derives from the fact that patientsmay present in approximately equal numbers to oldage psychiatry services (cognitive impairment,psychosis and behavioural disturbance), geriatricmedicine (acute confusional states, syncope andfalls), or neurological services (Parkinsonism). Anauthority from each of these specialities wastherefore asked to write a brief review of the topicfrom their own particular perspective.

The neurologist

Neocortical Lewy bodies (LB), after escapingdetectionfor years, are now in sharp focusas majorcontributors to dementia and psychiatric symptomsin the elderly. In severalautopsy seriesof dementedpatients, LB rank as the second most common typeof degenerative pathology, occurring in 15-25% ofcases, exceeded only by the senile plaques (SP) andneurofibrillary tangles (NFF) of Alzheimer's disease(AD). This brief review, coloured by the experienceof a researchcentre conductinglongitudinal studiesin dementia, willhighlightthe overlapbetweenLBDand AD, as well as those features attributable to LB.

A burgeoning nomenclature has been used todescribe patients with LBD, reflecting the diversityof clinical and pathological findings. The term‘¿�diffuseLewy body disease' (DLBD) was originallycoined for demented patients whose brains showed

widespreadLB, found in the hippocampus,temporallobe, cingulate and neocortex, in addition to theirclassic sites in the substantia mgra and othersubcortical regions. It was assumed that the diffuselydistributed LB accounted for dementia anddistinguishedDLBD from idiopathicParkinson'sdisease (PD), in which LBD were thought to berestricted to subcortical areas. However, recentstudies have shown that virtually every PD patienthas at least some cortical LB.

What then differentiates LBD patients from nondemented PD patients? One factor is the number ofLB, particularly in the cortex, as there is a directcorrelation between the density of cortical LB andthe severityof dementia assessedclose to the timeof death. Alternatively, the presence of additionalpathology, such as that of AD, may add to the LBburden to produce dementia. To clarify therelationship between LB and cognitive or psychiatricsymptoms, more information is needed about thephysiologicalfunction and anatomical connectionsof the cortical neurons that are vulnerable to LBformation.

Many patients with cortical LB also show SP andNFF. In the hippocampus, the AD pathology maybe fairly severe, accounting for memory dysfunction.However, NFT counts usually are much lower inneocortical areas in LBD than in AD, and somepatients with LBD have minimal AD pathology. ADlesions and LB are both associated with old age, andtheir coincidence may be due to chance alone.It is interesting that LB and NFT share severalfeatures: both are cytoskeletal derangementscontaining neurofilaments or related proteins,which undergo phosphorylation and ubiquitination.These events may occur late in lesion formation, andstronger evidence is required to make a case thatsimilar mechanisms lead to the formation of LB andNFT.

Just as cortical LB may evade casual inspection,so can the clinical picture of LBD be overlooked.Since patients with LBD commonly have AD lesions,clinical features usually overlap with thosecharacteristic of AD, and include the gradual onsetof forgetfulness, followed by difficulty with wordfmding, calculation and other cognitive skills. Theage of onset of LBD is similar to that of AD, although

709

710 McKEITH ET AL

men are affected more often than women. In ourexperience, the pattern of neuropsychological deficitsmay help to distinguish LBD from ‘¿�pure'AD. Whenmatched with AD patients at a similar stage ofdementia, LBD patients show more severe impairmenton visuospatial tests such as block design, clockdrawing or copying figures, and on tests of ‘¿�executivefunction' and problem solving such as the WisconsinCard Sorting Test, the Trail-Making test, and oncategory fluency. These areas of impairmentresemble the types of neuropsychological deficitsreported in PD patients, and reinforce the argumentthat LB load is an important cause of dementia inPD. With the progression of dementia, the selectivityof this pattern may be lost. The clinical course maybe faster in LBD than in AD; occasionally patientsprogress to severe dementia or death within a yearor two after the onset of dementia.

Psychiatric symptoms are common in LBD, as isthe case for AD and for dementia associated withPD. In our experience, depression and delusions havea similar frequency in LBD, vascular dementia andAD. Hallucinations, however, seem to be a hallmarkof LBD, occurring in up to 50% of patients, oftenin the absence of an obvious precipitating illness ormetabolic disturbance. This is a higher rate than inAD or other dementias, and may result fromcholinergic depletion and LB formation in areas suchas the temporal lobe, as the hallucinations are almostalways visual, and consist of formed images ofpeople or animals. Some patients show fluctuatinglevels of cognition or alertness, which may last fromhours to days, and resemble delirium or transientcerebral ischaemia.

Extrapyramidal signs occur in over 50% ofpatients with LBD in the absence of neurolepticexposure, and are generally milder than thosefound in PD patients. They include bradykinesia,rigidity, hypophonic speech, masked facies, stoopedposture, and a slow and shuffling gait; the resting,‘¿�pill-rolling'tremor, one of the hallmarks of PD,is extremely rare. Parkinsoman findings are notunique to LBD, as they may arise in AD, inother degenerative dementias such as progressivesupranuclear palsy, or after treatment withneuroleptic medications. About 10-20% of clinicallydiagnosed AD patients show extrapyramidal signsat an early or intermediate stage of dementia.These patients have a higher prevalence ofpsychotic symptoms and a more rapid rate ofprogression than AD patients without extrapyramidal signs, and most likely have an overlapsyndrome with underlying neuropathology of ADand LB. However, not all EPS associated withAD can be explained by LB.

Unfortunately there are no clear guidelines fortreating patients with LBD. In fact, advice aboutmedication avoidance may be more germane. Thecombination of dopaminergic deficits withcholinergic depletion make these patients sensitiveto several classes of medications. Neurolepticsmay provoke dramatic deterioration of cognitiveand motor function, and should be used cautiouslyif required to control psychotic symptoms oragitation. Drugs with anticholinergic activity mayproduce mental cloudingor delirium.Althoughdopaminergic medications such as L-dopa, themainstay of treatment in idiopathic PD, are lessoften effective in LBD, they warrant a trial.Preliminary reports suggest that cholinergicenhancement with tacrine, a cholinesterase inhibitorapproved for the treatment of AD, may improvecognition and possibly psychiatric symptoms inLBD. Now that the clinical syndrome of LBDcan be diagnosed with reasonable confidence,systematic studies of treatment are both feasibleand desirable.

The geriatrician

ThetermLBD isconfusing,especiallytoclinicians,andisoftenusedindiscriminatelytoincludeseniledementiaofLewy bodytype(SDLT),DLBD, theLewy body variant of Alzheimer's disease (LBV),and so on. This is hardly surprising, however, as theliteratureinthisfieldisitselfconfusing.Asacliniciantrying to treat a patient, in the absence of any specifictreatmentforany of theLB variants,I wonderwhetheritreallymattersatthemoment aslongasonerememberstoavoidphenothiazinesandrelatedcompounds if possible. At a scientific level, however,itisan extremelychallengingand interestingdebatewhichonedaymustberesolved.Shouldonebea‘¿�lumper'ora ‘¿�splitter'?AreLB a distinctentityordotheymerelyrepresentthemicroscopicappearanceof the end points of differing pathological processeseven if they share certain neurochemical features?Certainly, LB in the cortex differ somewhat fromthose in the brainstem at a light microscopical level,and also biochemically. It has also been claimed thatLB from the brains of people with idiopathic PD aredifferent from those of so-called DLBD, but thisrequires confirmation. Until some order is introducedinto the neuropathological descriptions and this isgenerally agreed and applied, it will continue to bedifficult to identify discrete clinical syndromes thatmay be associated with a particular neuropathologicalsubstrate.

Whether or not there are two separate pathologicalprocesses, that is, LBD and AD, either of which can

LEWY BODYDEMENTIA 711

exist independently or together, one being moreadvancedthantheotherin a particularbrain,whether they are in fact manifestations of a singledisease process or whether both these possibilitiesexist has in my opinion still to be decided. This isfurther complicated by the description of a subsetof patients with three concomitant pathologies —¿�those of AD, diffuse LB and also localisedspongiform change predominantly affecting themedial temporal region in which the disease was nottransmissible or associated with accumulation orprion protein (i.e. not Creutzfeldt—Jakob disease,CJD) (Hansen eta!, 1989). Conceptually it is easiestto accept the presence of a spectrum of LB diseases,one of which, SDLT, appears neuropathologicallyto be commoner among elderly demented patients,characterised by moderate numbers of cortical LB,at least moderate numbers of SP, but fewer or absentNET, together with a number of subcorticalabnormalities.

Psychiatric and neurological features

There are no specific clinical features that reliablyindicate the presence of dementia associated withcortical LB. In some patients the clinical courseis indistinguishable from AD until quite late inthe disorder, while others with so-called typicalfeatures eventually prove to have other conditions.Nevertheless, there is a consensus emerging thata progressive dementing illness with markedfluctuation and hallucinations or extrapyramidalfeatures earlier than is usually noted in AD, shouldraise the possibility of a LBD, when organic illnesshas been excluded as a cause of the fluctuatingcognitive state and in the absence of evidence tosupport the diagnosis of vascular dementia (McKeithet a!, 1992a). However, these features are not alwayspresentinourexperience,or thatof others(e.g.Förstl et a!, 1993), although early rigidity isfrequently found.

It has also been suggested that such patientsshow background slowing on their electroencephalogram (EEG), associated with burst patternsthat are often frontally dominant at an earlystagein thediseasecomparedwiththenormalEEG or the mild abnormalities observed in early AD(Crystal et a!, 1990). The pattern of cognitiveimpairmenthas beensaidto affectvisuospatialfunction, attention and verbal fluency more thanis usually found in AD with a similar degreeof dementia (Hansen et a!, 1990). Psychiatricsymptoms are likely to bias referral towardspsychiatric services, while physical problems suchas extrapyramidal features may lead to older people

being referred more frequently to the geriatricservices. This would seem to be our experience.

Therapeutic issues

Until the clinical features of the differing entities aremore adequately resolved, some subjects withdementia, especially those who are elderly, willcontinue to be diagnosed incorrectly as having AD orone of the other dementias when in fact they havecortical LB, and vice versa. This has importantimplications, as has already become apparent, for theevaluation of new treatments for AD, and emphasisesthe need for post-mortem verification of the diagnosisin as many trial patients as possible. Accurate clinicalclassification is important if treatment regimes,anticholinesterase-based or otherwise, are to beevaluated specificallyin people thought to have a LBD.

As has been pointed out by others (McKeith eta!,1992b), there is little doubt that some patients withSDLT are particularly sensitiveto neuroleptics, even tothe extent of developing the neuroleptic malignantsyndrome. Certainly many of them appear to experiencesignificant side-effect profiles at quite low doses.Can thefeaturesofa LBD betreated?Thatsuch

patients have a significant cholinergic deficit is nowwell established (e.g. E. K. Perry eta!, 1990). Somepatients subsequently proven to have Lewy bodypathology as well as that of AD have been amongthose responding to an anticholinesterase such astacrine in clinical trials. This is therefore an avenueworthy of further exploration.

Anti-Parkinsonian treatment does appear to be ofmodest benefit,especiallyif combined withphysiotherapy, producing a subsequent improvementinmobility,althoughthisisoftenrelativelyshortlived.L-dopaandanticholinergicmedicationsdo,however, often appear to worsen the cognitive andbehaviouralfeatures,necessitatingtheuseoflowdoses of those drugs least likely to aggravateconfusion.Benefithasalsobeenreportedina pilotstudy (Williams eta!, 1993) in which the difficultiesofevaluatingtreatmentapproachesforDLBD werehelpfullyexplored.

It has also been reported that clozapine, a novelneuroleptic drug, was effective in treating psychoticsymptoms in one patient with DLBD. This remainsto be confirmed. Future studies should address thepotential risk:benefit ratio when prescribingneuroleptics for this type of disorder.

In conclusion

This is a challenging area of medicine and furtherwork needs to be undertaken before we will properlyunderstand its place in the spectrum of dementia

ReferencenSampleNo. (%)with180Fomo

& Langston(1988)260Consecutive

autopsies12(4.6)Lennox

et a!(1989)216Consecutive

autopsies15(6.9)Joachim

eta!(1988)150Ante-mortem

diagnosisof dementia26

(17.3)Dickson

eta!(1989)216Ante-mortem

diagnosisofdegenerativebraindisease27

(12.5)R.

H. Perryeta!(1990)93Psychiatric

hospitalpatientswith dementia20

(21.5)Bums

eta!(1990b)50Patients

fulfillingdiagnosticcriteriafor Alzheimer'sdisease6

(12)Galasko

eta!(1994)170Ante-mortem

diagnosisofpossibleor probableAlzheimer'sdisease42

(24.7)

712 McKEITH ET AL

disorders. Nevertheless, we are even now able todiagnose it clinically, albeit still with inadequatesensitivity and specificity, and are able to startexploringtherelevanttherapeuticissues.

The old age psychiatrist

“¿�Thegreatestdifficultyiswhenweaknessandstiffnessoflimbs come on without the usual attitude and facialexpression,andespeciallyifthereisalsochronicmentalfailure.I haveknownthenatureof sucha caseto beeithermistakenor mysteriousduringmanyyears.'(Gowers,1893)

This statement, applied by Gowers to paralysisagitans (PD), might aptly be applied, in terms of the“¿�mistakenor mysterious―nature, to LBD: mistakenbecause of diagnostic uncertainty, and mysteriousbecause of nosological uncertainty. What seems lessuncertain is that LBD is far from being rare,although a word of caution needs to be given:published case reports number only about 300, andstatements about the prevalence of LBD are based(mainly) on highly selected post-mortem series -wherein lies both the mistake and the mystery.

How common is LBD?

A definitive answer is not yet available. Statementsabout LBD being the second commonest cause ofthe dementia syndrome should, in my view, beviewed with caution. They are based on studies suchas those summarised in Table 1. The frequency ofLBDintheseseriesrangesbetween4.6%and21.5%.The lowest frequency is found in studies whichconsider all autopsies (Forno & Langston, 1988;

Table 1Prevalence of LBD in post-mortem series

Lennox et a!, 1989). I have previously reviewedevidence (Byrne eta!, 1991a) which indicates that thefrequency of the different aetiological causes ofthe dementia syndrome is influenced not only byhistological criteria but also by the referring speciality.Jellingereta!(1990)showed thatAD was foundmore frequently in the sample from psychiatrichospitals, whereas PD was more frequently found insamples from general medical and geriatric hospitalpatients. The prevalence in vivo is also not known.One study of consecutive psychogeriatric day-hospitalpatients found 14 of 60 patients with dementia (23.3%)fulfilled clinical operational diagnostic criteria(McKeith eta!, 1992a) for LBD (Ballard eta!, 1993).

Clinical features

Although the following description of the clinicalfeatures of LBD is drawn from samples from widelydifferent settings, a remarkably consistent ‘¿�core'symptom profile emerges. LBD is characterised bydementia, Parkinsonism and neuropsychiatricsymptoms; these are detailed in Table 2, togetherwith approximate frequencies. Age may modify thesefeatures. Kosaka (1990) described LBD in Japan, andfound a younger age of onset was associated witha higher frequency of marked Parkinsonian features.

Approximate% ofpatientsexperiencing

symptoms

Table 2Clinical features of LBD

Dementia‘¿�mild'at onsetmarkedearly visuo

spatialproblemsprominentattention

disorder

Fluctuationin cognitionwith episodesresembling

delirium

ParkinsonismParkinsonismas

presentingfeaturerigiditytremorbradykinesiagait disorderposturalabnormality

(mainlyflexed)often describedas mild

Neuropsychiatricvisualhallucinations,often

complexdepressionparanoiddelusions

100

80

90

20-258050405030

30-5025

1510

FeaturesLBDVDADCJDCognition

at onsetMildy impaired,prominentattentionalproblemsmay be normalPatchy

impairmentImpaired, prominentmemory problemsImpaired,

may befocalFluctuation

in cognitionEarly and sustainedMarked at nightrelative day-timelucidityRareNot

markedPlateau

of stablecognitionVery

rareCommonRareVeryrareParkinsonismAlmost

invariablymild inold, may be presentingfeature and resemblePDRarely

resemblesPD. Gait disorderprominentRarely

resemblesPD,rigidity prominentMay

occurPsychiatric

symptomsCommonDepression commonCommon but rarelyearly featuesNot

markedMyoclonusRelatively

rare and ‘¿�mild'Very rareMayoccurCommonEpilepsyNot

describedCommonQuite commonQuitecommonCourseVariableStepwise

orgradualGradualRapidDuration

(average)3—8 years2—5 years4—9 yearsY2 —¿�2yearsSex

ratio? Male>femaleMale >femaleFemale> maleNo difference

LEWY BODYDEMENTIA 713

Age also modifies features in PD: in late-onset PD(after70years),rigidityandbradykinesiaaremoresymmetrical, tremor is less marked and L-dopafluctuation is late and mild when compared with‘¿�classic'PD (Godwin-Austen & Lowe, 1987; Broe,1992).ThefluctuationincognitioninLBD whenitoccurs

earlyinthecourseoftheconditioncanbesaidtobe characteristic. The fluctuation can be observedwithin a day and from day to day (Williams et a!,1993), this may reflect the prominent attentionaldisorder of the condition. Superimposed on thisfluctuating level of cognitive function are acuteconfusional states (ACS), which are idiopathic,unrelated to the common causes of ACS in the elderly.A patient with dementia who is having frequent ACS(notlimitedtonocturnaloccurrence)forwhichnocause is found should be suspected of having LBD.

Similarly, although psychiatric features arecommon in AD (Burns eta!, 1990a), early prominenthallucinations (especially visual) or delusions shouldsuggestLBD. Lesscommon symptomsof LBDincludeautonomicdysfunction,myoclonus,supranuclear palsy, polydipsia, dyskinesia, chorea, lowermotorneuronesigns(fasciculation,musclewasting),dysarthria(common in patientswith markedParkinsonian features) and dysphagia.

There are now validated operational clinicaldiagnostic criteria for the diagnosis of LBD. TheNewcastle criteria (McKeith et a!, 1992a) requirefluctuationincognitivelevelwhichpersistsoveralong period (weeks or months) and is associated withone of: visual or auditory hallucinations, mildspontaneousextrapyramidalfeaturesorneurolepticsensitivity syndrome, repeated falls, and transientcloudingorlossofconsciousness.TheNottinghamcriteria(Byrneeta!,1991b)giveclinicalcriteriafortwo levels of probability: probable where dementiawithorwithoutmarkedfluctuationincognitionisassociated with unequivocal Parkinsonian symptomsatsomepointinthecauseofillness;possiblewheredementia is associated with equivocal or mildParkinsonian features. Both these criteria havesatisfactory reliability and specificity but are onlymoderately sensitive (Byrne, 1994; McKeith et a!,1994a). McKeith et a! (1994b) in their validationstudyprovidea usefuldiscussionofthepotentialreasons for misdiagnosis of LBD. Table 3summarisessome of the key featuresin thedifferential diagnosis of LBD from AD, vasculardementia (VD), and CJD.

Some believe that LBD is a subtype of AD (Hanseneta!, 1990; Förstleta!, 1993), and in some cases it canbe difficult to distinguish the two conditions.

Table3Differential diagnosis of LBD (revised from Byrne, 1992)

LBD, Lewy body dementia;VD. vasculardisease;AD, Alzheimer'sdisease;CJD, Creutzfeldt-Jakob disease;PD, Parkinson'sdisease.

714 McKEITH ET AL

Crystal eta! (1990)suggest that EEG may be helpful -in LBD, EEG change occurs early in the course,whereas EEC may be normal in the early stages of AD.

Even in ‘¿�Alzheimered'LBD (cases in whichAlzheimer histological change coexists with corticaland brainstem LB) neuroleptic sensitivity (McKeithet a!, 1992b) exists.UntiltheNINDS—AIRENS(Roman eta!,1993)

criteriaforthediagnosisof VD wereproposed,the Hachinski Ischaemia Scale (IS; Hachinski eta!,1975) was widely used as an aid to diagnosis.As McKeith et a! (1994a) have pointed out, thefluctuation and mild cognitive change (featuresof LBD) may lead to high scores on the IS inpatientswith LBD. The fluctuationin VD istypically nocturnal worsening with relative daytimelucidity (Wade & Hachinski, 1987) and the plateauxof stable cognition in VD are rare in LBD. SomecasesofLBD havemyoclonusandtriphasicwaveson the EEG associated with a rapidly progressiveclinical course (Burkhardt et a!, 1988; Byrne et a!,1989). The myoclonus in LBD is not as severeas in CJD, nor is there an exaggerated startle reflex.Ataxia is not a feature of LBD, whereas it iscommon in CJD.

There are many other causes of dementia withParkinsonism (see Lennox, 1992), most of which,such as Huntington's chorea or multiple systematrophy, have characteristic neurological symptoms.The clinical syndrome associated with progressivesupranuclear palsy may occasionally be due to LBD(Fearnley et a!, 1991).

Management

In cases where Parkinsonian features are severe anddisabling, anti-Parkinsonian agents may be given atherapeutic trial. There is always a trade-off betweenthe benefit of improved mobility, the release ofbehaviours that may ensue (such as wandering), andthe mental state, especially if psychotic symptomsare present.

One small pilot study of an open trial of the useof anti-Parkinsonian medication detailed thepotential problems but reported some benefit(Williamseta!,1993).Oldpeopledonotrequireasmuch L-dopaastheir

moreyouthfulcounterparts(Broe& Caird,1973).It is worth getting an estimate of the shifting baselinefor each individual patient - as motor, cognitive andpsychotic symptoms are variable over time —¿�beforecommencing any treatment. An empirical approachis recommended: start low and titrate the doseaccording to both motor and mental-state changes.Recommended starting doses include:

(a) Anti-Parkinsonian agents - L-dopa (100mg)and carbidopa (10 mg) as Sinemet 110 (half atablet twice a day), co-beneldopa (50 mg) andbenserazide (12.5 mg) as Madopar 62.5 (onetablet twice a day), and L-deprenyl (5 mgdaily); Broe & Caird (1973) described such alow titrated dose of L-dopa in elderly dementedpatients with Parkinsonism and reportbenefit.

(b) drugs to treat psychotic symptoms —¿�chlormethiazole (192-384mg daily), and lorazepam(1—2mgtwice a day, short-term use only)

(c)Neurolepticswith extreme caution—¿�thioridazine (10mg twice a day), sulpiride(100—200mg daily); neuroleptic sensitivityis a problem, but we have encountered itmuch less frequently than the Newcastlegroup (Byrne et a!, 1992), possibly becauseour routine dosage of neuroleptics is muchlower.

Thenovelantipsychotics(suchasclozapine)havebeen used with benefit in LBD (Abelskov &Torpdahl, 1993) but clozapine has to be carefullymonitored for adverse effects and is expensive. Somesuggest that chlormethiazole has a neuroprotectiveeffect (Crosseta!, 1991;Baldwin eta!, 1993). InLBDits use is empirical but was suggested by its efficacyin delirium tremens. Some patients experience breakthrough of psychotic symptoms on long-termtreatment.

Discussion

The primary aim of this article is to compare andcontrast the views of specialists in neurology,psychiatry of old age, and geriatric medicine. Ingeneral, the contributors are in broad agreementabout the clinical profile of patients with corticalLBD, although there are some differences betweenthem which are perplexing. Thus, for example,hallucination rates appear to be higher in neurologicalpractice (up to 50°lo)whereas Parkinsonism isseen in the majority (90°lo)of cases coming to oldage psychiatry —¿�although seldom as a presentingfeature, usually falling short of the classic triadof bradykinesia, rigidity and tremor, and probablyoften exacerbated by neuroleptics. These are certainlynot the skews in symptom rates which one wouldpredict on the basis of speciality referral bias. A likelypartial explanation is that all three contributors areexperts in the field of dementia, and their reviewsmay reflect this aspect of their practice rather morethan mirroring the general experience and knowledge

LEWY BODYDEMENTIA 715

within their main speciality. Standardised methodsfor eliciting and recording individual symptoms inthis group of patients have not yet been developed,and in their absence further interpretation of thedifferent reported symptom rates would be highlyspeculative.

Nevertheless, it is of particular interest to ourgroup in Newcastle that syncope and falls have notbeen generally identified in these reviews as primarycharacteristics of LBD. Up to one-third of ourautopsy-confirmed cases of SDLT had repeatedunexplained falls at the time of clinical presentation,and these were often a significant problem in thepatients' management. We have previously speculatedthat spontaneous fluctuations in levelof consciousnessand abnormalities in posture, balance and gaitsecondary to nigrostriatal dysfunction or centralbaroreceptor reflex dysfunction, may be contributorypatho-physiologicalmechanisms.Whether or notsyncope and falls are important manifestations ofLBD needs to be further clarified —¿�Ballard et a!(1993) found that 15 out of 16 (94%) dementedpatients with idiopathic clouding of consciousnesshad experienced frequent falls, and 14 of these metNewcastle criteria for SDLT.

Both Dr Galasko and Professor Wilcock havemade detailed reference to the overlap betweenAlzheimer-type and LB pathology in these patients.Until recently PD and AD were considered to beclinically separate disorders and the possibility thattheir pathologies may be inter-related poses somefundamentally challenging questions for clinicalclassification. Some light may be shed on thispathological relationship by the results ofapolipoprotein E genotyping which suggest thatpossession of one or more e4 alleles is a risk factor,not only for late-onset familial sporadic AD, but alsofor LBD, with an c4 allele frequency of 35—40%in both disorders compared with l4-17°lo incontrols (Benjamin et a!, 1994). ApolipoproteinE4 may exert its pathoplastic effect by decreasingthe solubility of j9-amyloid and predisposing to theformation of diffuse SP. The observation that PDpatients without dementia and without Alzheimertype pathology have normal e4 allele frequenciesraisesthe possibilitythat(3-amyloidosismay bea secondary feature of LBD and its expression atleast partially influenced by differential apolipoproteinmetabolism.

Returning to clinical issues, Dr Byrne hasemphasised the methods currently available forclinical diagnosis, including the role of investigations.Although cognitive fluctuation is undoubtedly acharacteristic of many of these patients, it is difficultto define, particularly in the later stages of illness,

when persistent and profound cognitive impairmentmay supervene. By contrast, the intermittent acuteconfusional states to which Dr Byrne makes referenceare usually more easily recognised. Ballard et a!(1993) found a question from the CAMDEX ratingscale - “¿�arethere episodes lasting days or weeks whenhis/her thinking seems quite clear and then becomesmuddled?―- useful in identifying fluctuation. Thereis as yet little systematic information available aboutthe role of investigative procedures in differentialdiagnosis —¿�our experience with EEG is that slowingof the dominant rhythm occurs earlier and moreoften in LBD than in AD.

Despite the nosological and diagnostic uncertainties,there do seem to be some general principles emergingabout the management of these patients. Therecognition of their unusual and fluctuatingsymptom profile needs to be explained to carers, andthe possibility of rapid decline and a worseprognosis borne in mind. Motor symptoms maywell be relieved by standard anti-Parkinsoniantreatments in appropriate low dose. Patients whoare not distressed by their hallucinations mayaccept them as the cost of improved mobility —¿�atherapeutically more satisfying outcome thanhaving to render a patient immobile in orderto diminish frightening or distressing psychoticsymptoms. Neuroleptic drugs are not absolutelycontraindicated in LBD, but recent Committeeon the Safety of Medicines guidelines suggestthat they should be used with extreme cautionbecause of their propensity to induce acute andlife-threatening deterioration in some cases.Several alternative strategies have been suggested —¿�recent case reports (Lee eta!, 1994; Allen eta!, 1995)describe gratifying remissions of psychotic symptomsin response to risperidone, which has high affinity5-HT2 and D2 antagonist properties. Extrapyramidalside-effects and severe neuroleptic sensitivity mayhowever also occur with risperidone and systematictrials are required to establish whether safe andefficacious dosing regimes are possible.

In conclusion, one must say that clinicians, be theyold age psychiatrists, geriatricians or neurologists,should not feel that they are being asked to deal witha new disorder. Review of our old autopsy materialsuggests that such patients were equally commonthree decades ago, but went unrecognised. What hashappened is that improved neuropathological methods(e.g. antiubiquitin immunocytochemical staining ofcortical LB) have revealed the disorder and led usto a point where we have the potential clinically todiagnose and treat demented patients more precisely.In order to proceed we must improve our methodsof defining and measuring relevant psychopathology

716

CRYSTAL, H. A., DICKSON, D. W., LIzARNI, J. E., ci a! (1990)Antemortemdiagnosisof diffuse Lewybody disease.Neurology,40, 1523—1528.

DICKSON,D. W., CRYSTAL,H., MATr1ACE, L. A., et a! (1989)Diffuse Lewy body disease:Light and electron microscopicimmunocytochemistryof senileplaques.Ada Neuropathologica,78, 572—584.

FEARNLEY,J. M., REvESz,T., BROOKS,D. J., eta! (1991) DiffuseLewy body disease presenting with a supranuclear gaze palsy.Journal of Neurology, Neurosurgery and Psychiatry, 54,159—161.

Fosuo, L. S. & LANOSTON,J. W. (1988) The amygdalapara-hippocampal region: a predilection site for Lewy bodies.Journal of Neuropathology and Experimental Neuro!ogy,47, 354.

FO@ri, H., BURNS, A., LUTHERT, P., et a! (1993) The Lewy-bodyvariant of Alzheimer'sdisease:clinicalandpathologicalfindings.British Journal of Psychiatry, 162, 385-392.

G@sKo, D., HANSEN, L. A., K@zt@wi, R., ci a! (1994) Clinicalneuropathologicalcorrelationsin Alzheüner'sdiseaseandrelateddementias.Archives of Neurology, 51, 888-895.

G0DwIN-AU5mN, R. B. & Lowa, J. S. (1987) The two types ofParkinson's disease.In Current Problemsin Neurology Vol. 6.Parkinson'sDisease- Clinical and ExperimentalAdvances(ed.F. Clifford-Rose), pp. 79—82.London: Libbey.

Gowens, W. R. (1893) A Manual of Diseasesof the NervousSystem(2nd edn) (ed. C. N. Darien), p.654. Hafner.

HACHINSKI, V. C., Ku's, L. D., Ziwv@, E., eta! (1975) Cerebral

bloodflow in dementia.Archivesof Neurology,32, 632—637.HANSEN, L. A., MASLIAH, E., TERRY, R. D., et a! (1989) A

neuropathological subset of Alzheimer's disease withconcomitant Lewy body disease and spongiform change. AdaNeuropathologica, 78, 194—201.

@ D., G@sKo, D., ci a! (1990) The Lewy bodyvariant of Alzheimer'sdisease:a clinical andpathologicalentity.Neurology, 40, 1-8.

JELLINOER, K., DMIIEWzYx, W., FIscmIR, P., et a! (1990)Clinicopathologicalanalysisof dementeddisordersin theelderly.Journal of Neurological Sciences,95, 239-258.

JOACHIM, C. L., MoERIS, J. H. & SwtoE, D. J. (1988) Clinically

diagnosed Alzheimer's disease: autopsy results in 150 cases.Annals of Neurology, 24, 50—56.

KoSAKA, K. (1990) Diffuse Lewy body disease in Japan. Journalof Neurology, 237, 197-204.

Len, H., COONEIr,J. M. & LAWLOR,B. A. (1994)Casereport - Theuseof risperidone,an atypicalneuroleptic,in Lewy bodydisease.International Journal of Geriatric Psychiatry, 9, 415-417.

Litt*iox, 0. (1992) Lewy body dementia. Bailiere's ClinicalNeurology, 1, 653—676.

—¿�, Lowa, J, LANDON, M., ci a! (1989) Diffuse Lewy body

disease, correlative neuropathology using anti-ubiquitinimmunocytochemistry.Journal of Neurology,NeurosurgeryandPsychiatry, 52, 1236—1247.

MCKErrH,I. 0., PERRY,R. H., FAIRBAIRN,A. F., et a! (1992a)Operational criteria for seniledementiaof the Lewy body type(SDLT). PsychologicalMedicine, 22, 911-922.

ci a! (1992b) Neuroleptic sensitivity inpatientswith seniledementiaof Lewybody type. BriiishMedicalJournal, 305, 673—674.

—¿�, FAIRBAIRN, A. F., Peawr, R. H., et a! (l994a) The

clinical diagnosis and misdiagnosis of senile dementia ofthe Lewy body type (SDLT) - do the diagnostic systemsfor dementianeedto berevised?BritishJournalof Psychiatry,165, 324—332.

—¿�, Bomwsu, R. A., et al(1994b) An evaluation of

the predictive validity and inter-rater reliability of clinicaldiagnostic criteria for senile dementia of Lewy body type.Neurology, 44, 872—877.

McKEITH ET AL

and neurological abnormality in demented patients.The terminology and tools of general psychiatry andneurology may need extensive modification if we areto succeed. Research and clinical collaboration acrossand between the specialities will undoubtedly helpto further this process.

References

ABELSKOV, K. E. & TORPDAHL, P. (1993) Lewy body demens - enny sygdomsenied. Ugeskr Laeger, 155, 457—459.

ALLEN, R. L. WALKER, Z., D'Am, P. J., et a! (1995) Risperidonefor psychoticand behavioural symptomsin Lewy Bodydementia.Lance:,346, 185.

BALDWIN, H. A., Jo@ms, J. A., CRoss, A. J., et a! (1993)Histological, biochemical and behavioural evidence for theneuroprotectiveaction of chiormethiazolefollowing prolongedcarotid artery occlusion. Neurodegeneration,2, 139—146.

BALLARD, C. 0., MOHAN, R. N. C., PATEL, A., et a! (1993)Idiopathic clouding of consciousness- do the patients havecortical Lewy body disease?International Journal of GeriatricPsychiatry, 8, 571—576.

BENJAMIN, R., LE@xE, A., EDWARDSON,J. A., ci a! (1994)Apolipoprotein E genes in Lewy body and Parkinson's disease.Lancet, 343, 1565.

BROE, G. (1992) Parkinson's disease and related disorders. InOxford Textbook of Geriatric Medicine (edsJ. Grimley Evans& T. F. Williams), pp. 546-557. Oxford: Oxford Medical.

& CMan, F. I. (1973)Levodopafor Parkinsonismin elderlyanddementedpatients.MedicalJournalof Australia, 1,630-635.

Bupjowuyr, C. R., Fuuiv, C. M., Ki irecH r-DeasAsrERs,B. K.,ci al (1988) Diffuse Lewy body and progressivedementia.Neurology, 3*, 1520-1528.

BURNS,A., JAcoBy, R. & LEvY, R. (1990a) Psychiatric phenomenain Alzheimer'sdisease.British Journalof Psychiatry,157,72-94.

LUTHERT, P., LEvY, R., ci al(l990b) Accuracy of clinicaldiagnosis of Alzheimer's disease. British Medical Journal, 301,1026.

Byiu'm,E. J. (1992)Diffuse Lewy body disease:disease,spectrumdisorderof variety of Alzheimer'sdisease.International Journalof Geriatric Psychiatry, 7, 229—234.

(1995) Lewy body disease: an alternative view. InDevelopments in Dementia and Functional Disorders in theElderly (eds R. Levy & R. Howard), pp. 21-30. WrightsonBiomedical.

LENNOX, 0., GODWIN-AUSTEN,R. B., ci a! (1989)Diffuse Lewy body disease: clinical features in 15 cases.Journal of Neurology, Neurosurgery and Psychiatry, 52,709—717.

S@iim, C. W. & ARIE, T. (1991a) The diagnosis ofdementia- 1.Clinicalandpathologicalcriteria:a reviewof theliterature. International Journal of Geriatric Psychiatry, 6,199—208.

LENNOX, 0., OoDwIN-AusrEN, R. B., et a! (1991b)Dementia associated with cortical Lewy bodies. Proposeddiagnosticcriteria.Dementia,2, 283—284.

Bums, A. & WAITE, J. (1992) Neuroleptic sensitivity indementiawith corticalLewy bodies.BritishMedicalJournal,305, 1158—1159.

COMMI1-rEEONSAFETYOFManiciHEs (1994) Neuroleptic sensitivityin patients with dementia. Current Problems in Pharmacovigilance, 20, 6.

CRoss, A. J., Jopms, J. A., BALDWIN,H. A., ci a! (1991)Neuroprotectiveactivity of chlormethiazolefollowing transientforebrain ischaemia in the gerbil. British Journal ofPharmacology, 104, 406-411.

LEWY BODY DEMENTIA 717

PERRY, E. K., MARSHALL, E., KERW1N, J., ci a! (1990) Evidence ofa monoaminergic—cholinergicimbalance related to visualhallucinations in Lewy body type dementia. Journal ofNeurochemistry, 55, 1454—1456.

PERRY, R. H., IRVING, D., BLESSED, 0., ci a! (1990) Senile dementiaof the Lewy body type. A clinically and neuropathologicallydistinct form of Lewy body dementiain the elderly. Journal oftheNeurologicalSciences,95, 139—199.

ROMAN 0. C., TATEMICH, I. T. K., ERKINJUNTrI, T., ci a! (1993)

Vascular dementia: diagnostic criteria for research studies.Report of the NINDS—AIREN international workshop.Neurology,43, 240—260.

W@w€,J. P. H. & HACHINEKI,V. C. (1987)Multi-infarctdementia.InDementia (ed. B. Pitt), pp. 209—228.Edinburgh:ChurchillLivingstone.

WIWAMS, S. W., BYRNE, E. J. & Sroiu's, P. (1993) The treatmentof diffuse Lewy body disease: a pilot study of five cases.InternationalJournalof GeriatricPsychiatry,8, 731-739.

Ian G. McKeith, MD,University Sub-Department of Old Age Psychiatry, Institute for the Health of theElderly, Newcastle General Hospital, Westgate Road, Newcastle upon Tyne NE4 6BE; Douglas Galasko,MD, Department of Neurosciences (Neurology), University of California, San Diego and Veterans AffairsMedical Center, San Diego, CA, USA; Gordon K. Wilcock, MD,Department of the Care of the Elderly,Frenchay Hospital, Bristol BS16 1LE; E. Jane Byrne, MRCPsych,School of Psychiatry and BehaviouralSciences, University of Manchester, Withmgton Hospital, West Didsbury, Manchester M20 8LR

Correspondence:Professor 1.0. McKeith,UniversitySub-Departmentof Old Age Psychiatry, Institute for Health ofthe Elderly,NewcastleGeneralHospital,WestgateRoad, Newcastleupon TyneNE4 6BE

(Received 21 March 1995, accepted 19 May 1995)