1Journal of Neurology, Neurosurgery, and Psychiatry 1997;62:16-21

Hallucinations and signs of parkinsonism helpdistinguish patients with dementia and corticalLewy bodies from patients with Alzheimer'sdisease at presentation: a clinicopathological study

T A Ala, K-H Yang, J H Sung, W H Frey II

AbstractObjectives-To compare, in a retrospec-tive clinicopathological study, the presen-tation features of patients with dementiaand cortical Lewy bodies (Lewy bodydementia) with those of patients withAlzheimer's disease.Methods-From a population of 426 casesfrom the dementia brain bank, 39 cases ofLewy body dementia and 61 cases ofAlzheimer's disease with presentationdetails were identified.Results-The Lewy body dementia grouphad significantly more frequent halluci-nations (23% v 3%, P = 0.006) and signsof parkinsonism (41% v 5%, P < 0-0001)than the Alzheimer's disease group. TheLewy body dimentia group also had agreater proportion of men (62% v 34%,P = 0.013).Conclusion-Hallucinations and signs ofparkinsonism help distinguish Lewy bodydementia from Alzheimer's disease atpresentation. These indicators may not bevery sensitive, because they were reportedfor less than half ofthe patients with Lewybody dementia.

(7 Neurol Neurosurg Psychiatry 1997;62: 16-21)

Alzheimer'sTreatment andResearch Center,Department ofNeurology, RamseyClinic/Health-Partners, University ofMinnesota, St Paul,Minnesota, USAT A AlaW H Frey II

Department ofPathology, Nam-WonMedical Center,Chunbuk, South KoreaK-H YangDepartment ofNeuropathology,University ofMinnesota,Minneapolis,Minnesota, USAJ H SungCorrespondence to:Dr T A Ala, Department ofNeurology, RamseyClinic/HealthPartners, 640Jackson Street, St Paul, MN55101-2595, USA.

Received 29 January 1996and in final revised form12 September 1996Accepted 12 September 1996

Keywords: Lewy body dementia; Alzheimer's disease;diagnosis

Interest in cortical Lewy bodies found on neu-

ropathological examination of the brains ofdemented patients has increased in recentyears. Even though there has not been muchagreement on the nosology of the condition,with different researchers using various termsranging from Lewy body dementia to theLewy body variant of Alzheimer's disease (forreviews, see Hansen and Galaskol andLennox2), there have been many reports thatindicate that dementia associated with corticalLewy bodies is relatively common, with 14%-20% of demented patients having corticalLewy bodies at necropsy.3 6By contrast with the presentation symptoms

and signs of Alzheimer's disease, theNewcastle research group7 has reported that intheir experience with cases of Lewy bodydementia confirmed at necropsy "80% presentwith a syndrome of fluctuating cognitiveimpairment associated with hallucinations(usually visual), and with parkinsonian fea-tures which are often precipitated by neurolep-

tic drugs". These symptoms and signs havebeen so characteristic that the Newcastlegroup' has included them as key features of a

proposed set of clinical criteria for the ante-mortem diagnosis of the condition which theyhave termed "senile dementia of Lewy bodytype".

In addition to the Newcastle criteria, at leasttwo other research groups have recently pro-posed antemortem clinical criteria for thediagnosis of Lewy body dementia. Both theNottingham Group for the Study ofNeurodegenerative Disease9 and the SanDiego group'° have presented criteria thatinclude the presence of multiple parkinsonianfeatures, with the second group indicating thatthe criteria should be applied to patients withmild to moderate dementia. Neither empha-sises the psychiatric features as prominently asthe Newcastle group.

Despite the claims of these research groupsregarding parkinsonian and psychiatric fea-tures at presentation, few independent serieshave investigated how often patients withLewy body dementia present with such signsor symptoms. Moreover, those that haveincluded presentation details have notdescribed as high a frequency of parkinsonianor psychiatric features. For example, in theirnecropsy series of 65 patients with clinicallydiagnosed Alzheimer's disease Forstl et al I I

reported that the early symptoms of the eightthat were found to have "the Lewy body variantof Alzheimer's disease" were "uncharacteris-tic" when compared with a matched groupwith Alzheimer's disease. No particular differ-ences in the presence of hallucinations or fluc-tuations in the course of illness were noted,and the only distinctive feature of the patientswith cortical Lewy bodies was the develop-ment of severe rigidity.

Another necropsy series that includeddescriptions of the first symptoms and signs ofpatients with Lewy body dementia wasreported by Gibb et al12; however, their seriesincluded both cases of dementia and cases ofparkinsonian disorders. Out of 140 cases,seven with "cortical Lewy body dementia"were identified, with two having early memoryimpairment associated with psychiatric fea-tures and two presenting with a parkinsoniansyndrome.One necropsy series with a relatively high

frequency of parkinsonian and psychiatric fea-tures at presentation was reported by theNottingham group.'3 Their study, however,was not limited to cases of dementia and also

16

on June 14, 2020 by guest. Protected by copyright.

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.62.1.16 on 1 January 1997. Dow

nloaded from

Hallucinations and signs ofparkinsonism help distinguish dementia and cortical Lery body patientsfrom those with Alzheimer's disease

included cases with extrapyramidal syn-dromes. Fifteen of 57 such cases were foundto have "diffuse Lewy body disease", with thepresentations of four of the 15 including psy-chiatric features such as hallucinations. Nineof the 15 presented with Parkinson's diseasewith or without mild cognitive impairnent,and the remaining two presented with cogni-tive impairment alone.

Considering the relative paucity of researchinto the presenting clinical features of patientswith Lewy body dementia, we conducted aretrospective study of the cases in our dementiabrain bank to identify those with cortical Lewybodies to review their presentations. Our pri-mary objective was to determine how often keyfeatures included in the recently proposedclinical criteria&I° were reported for patientswith Lewy body dementia relative to thosewith Alzheimer's disease. In addition to anoverall group of patients with cortical Lewybodies, we subcategorised our cases intogroups according to the concomitant presenceof Alzheimer's changes and the relative con-centration of cortical Lewy bodies to allowreaders to consider different pathological defi-nitions of Lewy body dementia. To minimiseconfounding factors such as infarcts, weincluded only cases that demonstrated corticalLewy bodies, Alzheimer's pathology, or both.

Patients and methodsWe retrospectively reviewed all 476 cases ofdementia received in the Ramsey Foundation-Alzheimer's Treatment and Research CenterBrain Bank from August 1988 to the end ofJuly 1992. Brain tissue from these cases hadbeen referred to us for the neuropathologicaldiagnosis of a demented patient's illness. Ourreferral sources were diverse and included pri-mary physicians, neurologists, hospitals, acade-mic centres, and nursing homes primarilylocated in the midwest of the United States.Most of the specimens comprised the entirebrain; a few consisted only of histopathologicalslides or selected tissue samples. If frozenspecimens had been received or if the speci-mens had been obtained soon after death, onecerebral hemisphere was removed and frozenfor research purposes. After each brain wasreceived, medical records were requested fordocumentation of the patient's medical his-tory.

All cases in this study had a neuropathologi-cal report detailing the gross and microscopi-cal examinations of at least one hemisphere,the brainstem, and the cerebellum, and med-ical records indicating that the patient had aprogressive dementing illness. Two patientpopulations were determined for comparisonthat were as devoid of other dementing pathol-ogy as possible, one with cortical Lewy bodiesand the other with Alzheimer's disease withoutcortical Lewy bodies. To focus our study onpatients that presented with dementia ratherthan on all patients with disease associatedwith Lewy bodies, we excluded patients withclinical Parkinson's disease, defined as thosewho were treated with medication for signs of

parkinsonism before the onset of dementia.The few patients in our brain bank with clinicalParkinson's disease were identified but notincluded in the data analysis. Otherwise, selec-tion was not influenced by clinical diagnosesor characteristics.

CLINICAL METHODSThe records for each case were reviewed todetermine if there was documentation of thepatient's presentation for dementing illness,defined as the first time that a patient's symp-toms or signs of cognitive impairment or asso-ciated psychiatric pathology were consideredby a physician. The physician's report at thetime of presentation must have included ahistory, physical examination, assessment,and plan considering the symptoms or signs.Parkinsonian or extrapyramidal features wereconsidered present if the examining physicianmentioned any sign of parkinsonism, includ-ing but not limited to such terms as shufflinggait, masked facies, bradykinesia, loss of pos-tural reflexes, increased muscle tone, or anytremor. The use of neuroleptic medication inassociation with the presentation was docu-mented, and any exaggerated or unexpectedresponse to neuroleptic drugs was recorded.Parkinsonian or extrapyramidal features notedwith the use of neuroleptic drugs were notincluded as signs of parkinsonism. Therecords were also reviewed for additional keyfeatures included in the Newcastle criteria8that were reported at presentation or to haveoccurred in the preceding year, includingprominent fluctuation in cognitive status orlevel of consciousness, hallucinations, and anyunexplained falls or episodes of syncope.Hallucinations were recorded only if the term"hallucination" was used. When available,presentation examinations by neurologistswere also analysed in a separate subcategory.

Although not specifically included in theircriteria, the Newcastle group8 reported thatdepression occurs significantly more often inLewy body dementia at presentation. Wetherefore recorded if patients were diagnosedwith and treated for depression in associationwith the dementia presentation or in the pre-ceding year.

Severity of disease at the time of presenta-tion was graded according to DSM-III-R cri-teria'4 on review of the records. Either a"mild", "moderate", or "severe" rating wasassigned, with mild indicating that the patientretained the capacity for independent living,moderate indicating that some degree ofsupervision was necessary, and severe indicat-ing that continuous supervision was required.If the available clinical information for a par-ticular case was insufficient to establish aseverity rating with confidence, the less severegrade was assigned.

PATHOLOGICAL METHODSThe tissue was fixed, sectioned, and grosslyinspected, and histopathological sections werethen systematically taken from the neocortexof the frontal (two sections), parietal, andoccipital lobes, and from the amygdala, hip-

17

on June 14, 2020 by guest. Protected by copyright.

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.62.1.16 on 1 January 1997. Dow

nloaded from

Ala, Yang, Sung, Frey II

pocampus, corpus striatum, thalamus, mid-brain including the substantia nigra (at leasttwo sections at different levels), pons,medulla, and cerebellar cortex and stainedwith haematoxylin and eosin andBielschowsky silver stains. The identificationof cortical and subcortical Lewy bodies wasbased on their characteristic size, location, andstaining properties, consistent with publisheddescriptions.'5 All cases were reviewed by bothneuropathologists (K-HY and JHS).

Cases in which Lewy bodies were noted insubcortical sections but not the cortex withhaematoxylin and eosin were further studiedby preparing a slide, using an antiubiquitinstain, of the medial temporal lobe includingthe hippocampus and parahippocampal gyrus.Cases in which no Lewy bodies were notedwith haematoxylin and eosin were likewise fur-ther studied with antiubiquitin stain if eithermoderate or more severe degeneration of thesubstantia nigra was noted or if there wereatypical clinical features.

Quantitative estimates of the density ofsenile plaques and neurofibrillary tangles weremade from the silver stained sections. Countsof Lewy bodies were made from sectionsstained with either haematoxylin and eosin orantiubiquitin.

Cases showing cortical Lewy bodies wereconsidered as a single overall group and werealso subcategorised according to the presenceor absence of Alzheimer's changes in the neo-cortex and the density of cortical Lewy bodies.The "cortical Lewy body only" subgroupcomprised cases with cortical Lewy bodies and< 25 senile plaques and < 10 neurofibrillarytangles per 100 x magnification field, the"cortical Lewy body and senile plaque" sub-group comprised cases with cortical Lewybodies and ) 25 senile plaques and < 10 neu-rofibrillary tangles, and the "cortical Lewybody and Alzheimer's disease" subgroup com-prised cases with cortical Lewy bodies and> 25 senile plaques and > 10 neurofibrillarytangles. A "high concentration of corticalLewy bodies" subgroup comprised cases witha mean cortical Lewy body count > 5 in atleast five 100 x fields irrespective of the pres-ence of senile plaques or neurofibrillary tan-gles. The Alzheimer's disease control groupincluded all cases received during the studyperiod with no cortical Lewy bodies and > 25senile plaques and >s 10 neurofibrillary tan-gles in the neocortex.We excluded cases that exhibited any gross

or microscopical dementing pathology otherthan Lewy bodies or Alzheimer's changes-such as infarcts of any size, haemorrhages,contusions, Pick bodies, neoplasms, and vas-cular malformations. Cases that had evidenceof extra-axial pathology, such as subduralhaematomas and meningiomas, were alsoexcluded. The presence of any degree of ather-osclerosis, arterioarteriolar sclerosis, rarefac-tion of the cerebral white matter, or amyloidinfiltration of the vessels was allowed, pro-vided that there were no gross or microscopi-cal infarcts. Ventricular dilatation was not areason for exclusion. Cases showing pathologi-

cal evidence of acute or subacute infarcts orhaemorrhages (having occurred within weeksof the patient's death) were not excluded.

STATISTICAL ANALYSISDifferences between patient groups wereanalysed with unpaired Student's t tests forcontinuous variables and Fisher's exact testsfor counted variables. Two sided P valueswere used for all comparisons, with signifi-cance defined at the P = 0 05 level. Data forsubgroups were analysed if the differencesbetween the overall cortical Lewy body groupand the Alzheimer's disease control groupwere significant.

ResultsOf the 476 cases received during the studyperiod, 50 were excluded due to inadequatetissue (19), inadequate records (29), orrecords not documenting that the patient wasdemented (two). Ninety (21 %) of the theremaining 426 were found to have corticalLewy bodies with or without other pathology.An additional four had Lewy bodies in thesubstantia nigra but no cortical Lewy bodies.Only one case was identified that had corticalLewy bodies but no Lewy bodies in the sub-stantia nigra in two sections. This case alsohad severe substantia nigra neuronal loss andgliosis. Twenty two of the 90 (24%) had highconcentrations of cortical Lewy bodies.

Forty of the 90 cases with cortical Lewybodies were excluded from further study dueto microscopical or gross infarcts (34), un-classified degeneration (two), a meningioma(one), a subdural haematoma (one), ametastatic neoplasm (one), and alcoholicdegeneration (one). One additional case wasexcluded due to treatment for clinicalParkinson's disease before the onset of demen-tia. The remaining 49 cases had cortical Lewybodies without other dementing pathologyexcept senile plaques or neurofibrillary tan-gles. Presentation details were available for 39of these 49, comprising the overall corticalLewy body group.

There were 83 cases identified out of the426 that had Alzheimer's disease withoutother dementing pathology. Presentationdetails were available for 61 of these, compris-ing the Alzheimer's disease control group.

Table 1 gives the presentation details for thecortical Lewy body and Alzheimer's diseasegroups. A significant difference was notedbetween the two groups in reports of halluci-nations, with these being reported for nine ofthe cortical Lewy body group (23%) and onlytwo of the Alzheimer's disease group (3%). Sixof the patients with cortical Lewy bodies hadvisual hallucinations; no descriptions of thehallucinations of the other three were men-tioned. Descriptions of the hallucinations ofthe patients in the Alzheimer's disease groupwere not mentioned, although one wasincluded due to hallucinations after a seizure.

Signs of parkinsonism at presentation werereported for less than half of the cortical Lewybody group (15 of 37 as reported by all physi-

18 on June 14, 2020 by guest. P

rotected by copyright.http://jnnp.bm

j.com/

J Neurol N

eurosurg Psychiatry: first published as 10.1136/jnnp.62.1.16 on 1 January 1997. D

ownloaded from

Hallucinations and signs ofparkinsonism help distinguish dementia and cortical Lewy body patients from those with Alzheimer's disease

Table 1 Key features reported at presentation

Symptoms and signs AD CLBat presentation group group P value

Any fall or syncopal 4 4 0-71episode in preceding year n = 60 n = 39

Prominent fluctuation 2 3 0-38n =60 n39=038

Hallucinations 2 9 0-006n=60 n =39 00

Treatment for depression 4 7 0 11n = 60 n = 39

Any sign of parkinsonism 3 15 < 0-0001n=59 n=37 <00

Two or more of above 0 10 < 0 0001symptoms or signs n = 60 n = 39

Any sign of parkinsonism 3 13 0 0011reported by neurologist n = 34 n = 28

Treatment for parkinsonism 0 1 0 39n =61 n=3903

Men/women 21/40 24/15 0-013Mean age of onset in years (SD) 680 (9-1) 69-7 (5 7) 0 30

n =61 n=39

Values are numbers of cases with features reported. Groups include patients of all dementiaseverity.AD = Alzheimer's disease; CLB = cortical Lewy bodies.

cians or 13 of 28 as reported by neurologists),but they were reported statistically more oftenthan for the Alzheimer's disease group (threeof 59). One patient from the cortical Lewybody group was being treated for parkinson-ism at the time of presentation to a physicianfor dementia, with the records documentingthat the onset of symptoms of dementia hadpreceded the onset of the parkinsonism.Ten of the 39 patients with cortical Lewy

bodies (26%) had two or more of the key fea-tures (fall or syncopal episode, fluctuation,hallucinations, depression, or parkinsoniansigns) at presentation by contrast with none ofthe patients with Alzheimer's disease. Fourpatients with cortical Lewy bodies had threekey features whereas there were none in thepatients with Alzheimer's disease. Eight of thecortical Lewy body group were treated withneuroleptic medication in association withtheir presentations, with no cases of exagger-ated or unexpected responses being noted.The cortical Lewy body group also had a

greater proportion ofmen 24 of 39 (62%) v 21of 61 (34%) for the Alzheimer's disease group.

Table 2 presents the data for the patientswith only mild or moderate severity of demen-tia. Consistent with the findings in table 1,

hallucinations, signs of parkinsonism, and twoor more features were statistically more fre-quent in the cortical Lewy body group than inthe Alzheimer's disease group. Three patientswith cortical Lewy bodies had three key fea-tures.

Table 2 also includes data categorisedaccording to the cortical Lewy body sub-groups. On the assumption that clinicianswould be more interested in presentationdetails of patients with mild or moderatedementia, excluding those with severelyadvanced conditions, the subgroup data areincluded in table 2 rather than table 1. Withthe exception of the cortical Lewy body andAlzheimer's disease subgroups, each of thesubgroups reflected essentially the same signif-icant differences relative to the Alzheimer'sdisease group as the overall mild-moderatecortical Lewy body group. A remarkable find-ing was that the cortical Lewy body only sub-group contributed relatively greater numbersof cases with hallucinations and cases with twoor more features than the other subgroups,with three of four (75%) having such reports.The cortical Lewy body and Alzheimer's dis-ease subgroup contributed only one or none ofthe key features, with no features reaching sig-nificance.

All 15 cases in the Alzheimer's diseasegroup for which key clinical features werereported were examined for cortical Lewybodies using antiubiquitin stain; no additionalcases with cortical Lewy bodies were identi-fied.

DiscussionAlbeit in a minority of cases, our results sup-port the claims of the Newcastle group7 thatmany patients with Lewy body dementia pre-sent with a syndrome very different from the"typical" patient with Alzheimer's disease.Twenty six per cent of those with Lewy bodydementia presented with two or more of thekey features, by contrast with none of thosewith Alzheimer's disease. Those with Lewybody dementia had hallucinations and signs ofparkinsonism reported more often than thosewith Alzheimer's disease, and they were notedto have a male predominance.

Table 2 Key features reported at presentation for patients with mild or moderate dementia

CLB subgroupsMild- Mild-moderate moderate High

Symptoms and signs AD CLB concentrationat presentation group group P value CLB only P value CLB+ SP P value CLB+AD P value CLB P value

Hallucinations 1 7 0 008 3 (2) 0 0012 4 (2) 0-038 0 1i. 4= 0 0027n =40 n=29 n=4 (2) n =20 (6) n=5 (1) n= 9 002Any sign of 1 10 0-0003 2 (2) 0-018 7 (2) 0-0008 1 (0) 0-21 4002parkinsonism n = 41 n°= 28 n= 4 (2) n 19 (6) n 5 (1) n=9 00025Two or more 0 8 0 0005 3 (2) 00003 5 (2) 0-0028 0 4 0-0006symptoms or signs n =40 n=29 n=4 (2) n =20 (6) n=5 (1) n=9 006

(see table 1)Any sign of 1 90-0093 2 (2) 0-057 6 (2) 0-029 1 (0) 0-17 4004parkinsonism n = 21 n = 22 n = 4 (2) n 16 (4) n 2 (0) n=6 0 004

reported byneurologist

Values are numbers of cases with features reported. The CLB only subgroup column includes cases with CLBs and < 25 senile plaques (SPs) and < 10 neurofibril-lary tangles (NFTs) per 100 x magnification field in the neocortex; the CLB and SP subgroup column includes cases with CLBs and > 25 SPs and < 10 NFTs; theCLB and AD subgroup column includes cases with CLBs and > 25 or more SPs and > 10 NFTs. Numbers in parentheses indicate number of cases with mean CLBcount greater than five in at least five 100 x magnification fields, summarised in the high concentration CLB subgroup column. P value is statistical comparison withthe AD column. P values in the high concentration CLB column were not adjusted for multiple comparisons.

19

on June 14, 2020 by guest. Protected by copyright.

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.62.1.16 on 1 January 1997. Dow

nloaded from

Ala, Yang, Sung, Frey II

On the other hand, our results do not sup-port the Newcastle group's report7 that 80% ofpatients with Lewy body dementia presentwith fluctuating cognitive impairment associ-ated with hallucinations and parkinsonian fea-tures. Whereas 23% were reported to have hadhallucinations at presentation, the other keysymptoms of falls, syncope, and prominentfluctuation were reported less often and withno greater frequency than for the patients withAlzheimer's disease. Less than 50% of thepatients had signs of parkinsonism docu-mented at presentation, whether reported byneurologists or by all the physicians irrespec-tive of specialty. Moreover, the Newcastleclaim that depression occurs more often inLewy body dementia at presentation8 was notsupported by our data.

This relatively low frequency of recognitionof parkinsonism at presentation also suggeststhat the clinical diagnostic criteria for Lewybody dementia proposed by the Nottingham9and San Diego'0 groups would not be verysensitive for distinguishing early disease, asboth require the presence of multiple parkin-sonian features.

It must be emphasised that our data collec-tion was retrospective, and the true frequencyof occurrence of many of the key symptomsand signs likely was higher. We acknowledgethat the absence of reports of falls, syncope,and fluctuation cannot be used to argue thatthey did not occur. Furthermore, our patientpopulation may not have been representativeof all cases of Lewy body dementia, as thosethat may have had a very rapid course andflorid psychiatric features could have beenneuropathologically diagnosed through otherchannels. Nevertheless, our data suggest thatsuch features were not prominent in the pre-sentations of most of the patients with Lewybody dementia.

Another explanation for the differencebetween our results and those of theNewcastle group8 is that our patient popula-tion was very different from theirs. Ourpatients were referred from a wide variety ofsources by contrast with the Newcastlepatients, most of whom were in psychogeri-atric units.4 By selecting similar patients fromour brain bank the frequency of reports of psy-chiatric symptoms would undoubtedly havebeen higher.

Even in a retrospective study such as this, inwhich respondents were not systematicallyasked if certain symptoms had occurred, thereports of hallucinations at presentation in23% of the patients with Lewy body dementiaare very remarkable. Whether patients of allseverities, only patients of mild or moderateseverity, or only patients in the cortical Lewybody subgroups are considered, hallucinationswere more often reported for patients withcortical Lewy bodies than for patients withAlzheimer's disease. Surprisingly, the onlyexception to this finding was in the subgroupof patients with Lewy body dementia thatincluded Alzheimer's disease (the corticalLewy body and Alzheimer's disease sub-group), in which no hallucinations were docu-

mented. It is of particular interest that visualhallucinations were specifically mentioned inmost of these reports. With the precaution thatthis study did not assess the occurrence of hal-lucinations in other dementing illnesses, thisfinding strongly suggests that visual hallucina-tions occurring early in a dementing illnessargue against Alzheimer's disease as being thesole aetiology.An incidental finding was the greater ten-

dency for Lewy body dementia to afflict men,by contrast with Alzheimer's disease. Althoughthis tendency has been mentioned by otherinvestigators,8 16 18 only Galasko et al'9 havereported it as significant relative toAlzheimer's disease.

That all of the cases were not examinedwith antiubiquitin stain is a potential weaknessin this study, raising the possibility that thepercentage of our dementia cases with corticalLewy bodies may actually have been higherthan 21 %. Because we found excellent corre-lation between the slides stained with antiu-biquitin and those stained with haematoxylinand eosin regarding the identification of corticalLewy bodies, we doubt that the percentagewould have been substantially higher.Furthermore, whether or not a more rigoroussearch for cortical Lewy bodies would haveincreased the frequency of key feature recogni-tion is not known, for many of the new casesof Lewy body dementia would most likelyhave been derived from the presentAlzheimer's disease group. More judiciousscrutiny of the cases with Alzheimer's diseasewith key features (with additional antiubiqui-tin slides) could have increased the frequencyif some were discovered to have cortical Lewybodies.We caution that our results are primarily

germane to the study of dementia and repre-sent only part of the total range of diseaseassociated with Lewy bodies. Although thenumbers were small, we excluded patients thatwere treated for Parkinson's disease beforeonset of the dementia. Similarly, our study didnot include patients who may have harbouredLewy bodies but were asymptomatic.20Our results suggest that clinical differences

may exist between the cortical Lewy body sub-groups, but the few cases in each group limitconvincing analysis. Previous studies thatmention presentation details are very difficultto compare in this regard because of the lackof standardisation of criteria and objectives. Asexamples, the Nottingham group'3 did notfully describe their neuropathological findingsin their report, and the study by the Newcastlegroup8 included cases with cortical Lewy bodiesand senile plaques, the study by Gibb et al12included cases with cortical Lewy bodies only,and the study by Forstl et all' included caseswith cortical Lewy bodies and senile plaquesor Alzheimer's disease. None of these studiesspecifically excluded cases with Lewy bodydementia with other neuropathology.

In summary, our results corroborate theadvice of the Newcastle group7 that "cliniciansshould be aware of the broad spectrum of clin-ical presentations of cortical Lewy body

20

on June 14, 2020 by guest. Protected by copyright.

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.62.1.16 on 1 January 1997. Dow

nloaded from

Hallucinations and signs ofparkinsonism help distinguish dementia and cortical Lewy body patients from those with Alzheimer's disease

disease". Many patients with Lewy bodydementia indeed presented with symptomsand signs different from the patient withAlzheimer's disease. In particular, the occur-rence of hallucinations in association withtheir presentations is very noteworthy. On theother hand, many did not seem to be differentfrom Alzheimer's disease, at least regardingthe features we studied. Furthermore, therecently proposed criteria8-'0 for the clinicaldiagnosis of Lewy body dementia would seemnot to be very sensitive in its early detection.Further work is needed to confirm our find-ings and to develop better clinical criteria forits early diagnosis.

The Ramsey Foundation-Alzheimer's Treatment and ResearchCenter Brain Bank is supported by the Ramsey Foundation, StPaul, Minnesota, by the Robert Wood Johnson 1962Charitable Trust, by the Unicare Foundation, and by a grantfrom the State of Minnesota. We are also grateful to Ms JudyJohnson and Ms Marilu LaVoie for the administration of thebrain bank.

1 Hansen LA, Galasko D. Lewy body disease. Curr OpinNeurol Neurosurg 1992;5:889-94.

2 Lennox G. Lewy body dementia. Bailliere's Clin Neurol1992;1:653-76.

3 Dickson DW, Crystal H, Mattiace LA, et al. Diffuse Lewybody disease: light and electron microscopic immunocy-tochemistry of senile plaques. Acta Neuropathol (Berl)1989;78:572-84.

4 Perry RH, Irving D, Blessed G, Fairbairn A, Perry EK.Senile dementia of Lewy body type. A clinically and neu-ropathologically distinct form of Lewy body dementia inthe elderly. JNeurol Sci 1990;95:119-39.

5 Ince PG, McArthur FK, Bjertness E, Torvik A, Candy JM,Edwardson JA. Neuropathological diagnoses in elderly

patients in Oslo: Alzheimer's disease, Lewy body disease,vascular lesions. Dementia 1995;6: 162-8.

6 Crystal HA, Dickson DW, Sliwinski MJ, et al. Pathologicalmarkers associated with normal aging and dementia inthe elderly. Ann Neurol 1993;34:566-73.

7 McKeith I, Fairbaim A, Briel R, Harrison R, Tasker N.Lewy bodies and Alzheimer's disease [letter]. Br JPsychiatry 1993;163:262-3.

8 McKeith IG, Perry RH, Fairbairn AF, Jabeen S, Perry EK.Operational criteria for senile dementia of Lewy bodytype (SDLT). PsycholMed 1992;22:911-22.

9 Byrne EJ, Lennox GG, Godwin-Austen RB, et al.Dementia associated with cortical Lewy bodies: proposedclinical diagnostic criteria [letter]. Dementia 1991;2:283-4.

10 Hansen L, Salmon D, Galasko D, et al. The Lewy bodyvariant of Alzheimer's disease: a clinical and pathologicentity. Neurology 1990;40:1-8.

11 Forstl H, Burns A, Luthert P, Cairns N, Levy R. TheLewy-body variant of Alzheimer's disease. Clinical andpathological findings. BrJ Psychiatry 1993;162:385-92.

12 Gibb WRG, Luthert PJ, Janota I, Lantos PL. CorticalLewy body dementia: clinical features and classification.J Neurol Neurosurg Psychiatry 1989;52: 185-92.

13 Byrne EJ, Lennox G, Lowe J, Godwin-Austen RB. DiffuseLewy body disease: clinical features in 15 cases. J NeurolNeurosurg Psychiatry 1989;52:709-17.

14 American Psychiatric Association. Organic mental syn-dromes and disorders. In: Diagnostic and statistical manualof mental disorders, 3rd ed rev. Washington, DC:American Psychiatric Association Press, 1987:97-163.

15 Pollanen MS, Dickson DW, Bergeron C. Pathology andbiology of the Lewy body. J Neuropathol Exp Neurol 1993;52:183-91.

16 Dickson DW. Lewy body variant [letter]. Neurology 1990;40:1147-8.

17 Kosaka K. Diffuse Lewy body disease in Japan. J Neurol1990;237: 197-204.

18 Kuzuhara S, Yoshimura M. Clinical and neuropathologicalaspects of diffuse Lewy body disease in the elderly. AdvNeurol 1993;60:464-9.

19 Galasko D, Hansen LA, Katzman R, et al. Clinical-neu-ropathological correlations in Alzheimer's disease andrelated dementias. Arch Neurol 1994;51:888-95.

20 Gibb WRG, Lees AJ. The relevance of the Lewy body tothe pathogenesis of idiopathic Parkinson's disease. JNeurol Neurosurg Psychiatry 1988;51 :745-52.

21

on June 14, 2020 by guest. Protected by copyright.

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.62.1.16 on 1 January 1997. Dow

nloaded from