JNeurolNeurosurgPsychiat194912287DEGENERATIONOFTHEGRANULARLAYE - PDF document

J.Neurol.Neurosurg.Psychiat.,1949,12,287.DEGENERATIONOFTHEGRANULARLAYEROFTHECEREBELLUMBYA.D.LEIGHandA.MEYERFromtheDepartmentofNeuropathology,InstituteofPsychiatry,MaudsleyHospital,LondonThefollowinginvestigationwasinitiatedbytheobservationbyoneofus(Meyer,1944),ofaselectivedegenerationofthegranularlayerofthecerebelluminacaseofaWernickesyndromeassociatedwithcarcinomaofthestomach.Thepossiblerelationtooneorothertypeofnutritionaldeficiency(notnecessarilythatofvitaminB1)wasdiscussed,butnoconsiderationwasgiventothefactthattheremightbeamoredirectcausalrelationtothecarcinoma.Therehasbeenafairnumberofreportsoncasesofvisceralcarcinomainwhichclinicaland/orpathologicalsignsofcerebellardegeneration(notassociatedwithsecondarydeposits)werediscovered.(Lhermitte,1922;Casper,1929;Scherer,1933;ParkerandKernohan,1933;Greenfield,1934;Kennard,1935;Ziilch,1936;CourvilleandFriedman,1940;Alessi,1940;Buchananandothers,1947.)Acausalrelationbetweenthecarcinomaandthecerebellarlesionwasbynomeansassumedinallofthesecommunications.IndeedinthecasesofCourvilleandFried-man,andBuchananandothers,itwasexpresslyrejectedinviewoftheclinicalhistory.IntheopinionofBrouwerandBiemond(1938),however,therapiddevelopmentofacerebellarlesionincasesofcarcinomawassuchastosuggesttheinfluenceofacarcino-toxicagent.Thegranularlayer,ifinvolved,wasnotalwaysselectivelyaffectedintheircases,butinthemostimportantobservationspublishedonthissubjectbyBertrandandGodet-Guillain(1942)thegranularlayerwasindeedselectivelyinvolved.Infourteenoutofsixteencasesofvisceralcarcinomatatheseauthorsde-scribedvaryingstagesofdegenerationofthegranularlayer,betrayingitselfevenonmacroscopicinspectionbyanappearancetheytermed"glace."Thesecerebellarchangesweremostmarkedinthedorsalpartofthevermisandwereusuallyaccom-paniedbydegenerativesignsintheinferiorolives,particularlyintheirdorsalportion.Thispredilec-tionforthedorsalaspectsofthecerebellumrecallsthatseeninthedelayedtypeofcerebellaratrophy(Foixandothers,1922),morerecentlyknownasparenchymatouscerebellaratrophy.Thisrelationship,whichisimportantinviewoftheusualageperiodofcarcinoma,hasinfactbeendiscussedintheliterature(Casper,1929;BrouwerandBiemond,1938).Thelatterauthorsconcluded,however,thattherewasnorelationshipbetweenthetwoconditions,astheyfoundadiffusedistribu-tionofthecerebellarchangesinthecarcinomacases,ascomparedwiththemorefocaldistributioninthedelayedatrophy.InthecasesdescribedbyBertrandandGodet-Guillainnoclinicalcerebellarsignswereobserved,andtheauthorsconsidered,therefore,thatthecerebellarlesionshaddevelopedrapidlyduringtheterminalcachecticphaseofthecarcinomatousprocess.Apartfromtheassociationwithvisceralcarcino-mata,aselectivedegenerationofthegranularlayerofthecerebellumhasbeenattributedtoavarietyofcauses.OneofthefirstdescriptionsoftheconditionwasgivenbyBielschowsky(1920)ininfantile,lateinfantile,andjuvenilecasesofamauroticidiocy.Sincehisdescription,this"cerebellopetal"typeofdegenerationhasbeenregardedbymanyasthepredominanttypeofcerebellarinvolvementincerebro-maculardegenera-tion.Norman(1940)showedthatselectivedegen-erationofthegranularlayermayalsooccurinmentaldefectivesothert

hanthosesufferingfromamauroticidiocy.Bothhispatientsdiedfrompulmonarytuberculosis,complicatedinonebyaterminaltuberculousenteritis.Inthispatienttheterminaldiseaseclearlyprecededtheonsetofcerebellarsigns.Wementionthisrelationshipbecausewastingdiseasesand,inparticular,severegastro-enteritishavebeengivensomeattention,especiallyintheolderliterature(Murri,1900;Rossi,1907;Archambault,1918).Senilecachexiaandmalnutritionhavealsobeendiscussed,particu-larlyinrelationtothedelayedcerebellaratrophiesofoldage(CourvilleandFriedman,1940).Diabetesmellitushasbeennotedintheliteratureasapossiblecause.Shinosaki(1926),describingcerebellarchangesinacaseofseverecongenitaldiabetes,clearlyillustratedaselectivedegenerationofthegranularlayer.Inkeepingwithviews287 A.D.LEIGHANDA.MEYERcurrentatthattimehediscussedthepossiblecausationofdiabetesbylesionsofthevermis,particularlyoftheuvula.OneofWinkelman's(1943)casesofgranularlayerdegenerationwasadiabeticwhodiedfollowinganoperationforacutegangreneoftheintestinaltractandhramorrhagicgastritis.ThemostimportantinvestigationinthisfieldwascarriedoutbyBertrandandTiffenau(1942),whofoundaselectivedegenerationofthegranulesinthedorsalpartofthecerebelluminfivecasesdyingindiabeticcoma,andinallcasestheappear-anceanddistributionweresimilar.Amoreorlessselectivedegenerationofthegranularlayerhasalsobeendescribedafterpro-longedexperimentalcarbon-monoxidepoisoning(FerraroandMorrison,1928),andinvarioustoxi-infectiousconditions,includinglead,carbonmon-oxide,andhistaminepoisoning,inanition,andsepticaemia(Williams,1934).Twopapersdescrib-ingmorerecentexperimentalworkrequirespecialmention.Upners(1939)administeredthiophen(C4H4S)todogsinsmalldosesoverlongperiods,andfoundaselectiveactiononthegranularlayer,resultinginitsdegeneration,usuallycombinedwithpathologicalchangesinthequadrigeminalregionresemblingthoseseeninWernicke'ssyn-drome.Inalltheanimalsfrequentvomitingandanorexiawereconspicuous.Likewisetheadminis-trationofmethylmercurycompoundstoratsandamonkey(Hunterandothers,1940)resultedinrapidlossofweight,followedbyspecificnervousdisturb-ances.Intheratsonwhichmultipleobservationswerepossibletheperipheralnervesandposteriorspinalrootswereaffectedfirst,theposteriorcolumnsandgranularlayerofthemiddlelobesofthecerebellumlater.Thisreviewoftheliterature,thoughprobablybynomeanscomplete,demonstratesthatcertainmetabolicfactorsmaybeofimportanceinthepathogenesis,classification,andetiologyofcertaincerebellardegenerations.Itwasforthisreasonthatweconsideredafurtherinvestigationoftheproblemwasjustified.MaterialOurmaterialconsistsof(1)fourteencasesofvisceralcarcinomawithorwithoutmetastasesinbrainandotherorgans.Eightoftheseshowedadegenerationofthegranularlayer.Thenegativecasescomprisedpatientssufferingfromavarietyofvisceralcarcinomata(ofrectum,stomach,ceso-phagus,gallbladder,bronchus,prostate).(2)Sixcasesdyingindiabeticorhypoglycemiccoma.Ofthesetwoweredefinitehyperglycaemicfatalities;inneitherwasthereanycerebellaraffection,althoughinoneacompleteolivarydegeneration,andsevereacutecellchanges(lysis)mostmarkedinthethirdlayerofthecerebralcortexandoftheparvo-cellularelementsofthecorpus

striatumwerepresent,similartothechangesdescribedbyBertrandandTiffenau(1942).Bothcasesareexcludedfromfulldescriptioninthispaper;theremainingfourincludeacaseofdoubt-fulhyperglycvmiccoma,adiabeticdyingfromirreversiblehypoglycemia,andtwoschizophrenicswithhypoglycaemiainducedfortherapeuticpurposesanddyinginirreversiblecoma.Thelastthreecasesofprovedhypoglycemiawerefoundinatotalmaterialofninecasesoffatalhypoglycxemia.(3)Fourcasesofamauroticidiocy,allshowingadegenerationofthegranularlayer.Inallcasesinvestigatedforthispaperafull,histologicalinvestigationofthebrainhasbeenmadebytheusualmethods.GroupI:VisceralCarcinomaCase1.-Amanaged62yearswasthesubjectofapreviousreportbyoneofus(Meyer,1944).Hediedafteranillnessofapproximatelytwelvemonths'duration,theclinicaldiagnosisbeinggastriccarcinoma.Thiswasconfirmedatautopsy.AutopsyFindings.-Macroscopicallytherewassomethickeningoftheleptomeningesatthebaseofthebrain,particularlywithintheinterpeduncularspace.Inthecerebellumtherewasnogeneralreductioninsize,butaglassyappearancewasnoticedinwhatwouldcorrespondtothegranularlayer.Histologically,therewerechangesinthehypothalamuscharacteristicofWernicke'sencephalopathy.Thesechanges,however,wereatypicalinsofarastheglio-mesodermalproliferationwasmoreconspicuousintheanteriorpartofthehypothalamusthaninthemammillarybodies,whichshowedsignsofaveryrecentaffectiononly.Inthecerebellumaselectivedegenerationofthegranularlayerwaspresent.InNisslpreparationsthePurkinje-celllayercontrastedprominentlywiththegranularlayer,thegranulesfailingtostainwithcresylviolet(Fig.1).Withhigh-powermagnification,insteadofthedarkly-stainednucleiofthegranules,withtheirwell-definedchromatin,onlytheshadowsofseverelydeformednucleiinvaryingdegreesofdisintegrationweretobeseen.Thecytoplasm,normallynotwellseen,appearedtomergewiththatofadjacentcells,thusproducingwhathasaptlybeentermed"conglutination."Thisprocesswasfoundthroughoutthegranularlayer,butwasnotquitesoadvancedintheventromedialsectorasintheremainingsectorsofthecerebellum.Inmyelinpreparationsalso,theventralandespeciallyventromedialareasshowedabetterstainingreactioninthegranularlayer,particularlynearthesummitoflobules.Thiswasduetonuclearstaining,whichwasinvaryingdegreesdeficientintheotherpartsofthegranularlayer.Thetranslucentappearancewasnotduetoadiminutionofthemyelinatedfibres,which,thoughconspicuouslybeaded,wereseentotraverse288 DEGENERATIONOFTHECEREBELLUM-AFIG.1.-Diffusedegenerationofgranularlayerinacaseofgastriccarcinoma(Case1).Nissl,x18.I..I..I.Nthisregioninnormalnumbers.Bielschowskysilverpreparationsdisclosednormalbasketcellsandprocesses,withnodiminutioninthefibresofthegranularlayer.TheBergmanngliacontainedanexcessoflipoids,andmuchfatwasalsoseeninthegliacellsandmesodermalelementsofthegranularlayer,andtoalesserdegreeinthewhitematter.Noappreciablemicroglialprolifera-tionwasseeninPenfieldpreparations,norwasanyfibrousgliosisrevealedbytheHolzertechnique.Thedentatenucleuswaslittleaffected,saveforanexcessoflipofuscininnervecellsandsomedegreeofpatchyglialproliferation.Intheinferiorolivesitwasimmediatelyapparentwithlowmagnificationthattherewasadiminutionofnervecells

,mostmarkedinthedorsomedialcornerofthenucleus.High-powerinspec-tionrevealedthatallnervecellsshowedsomepigmentdegeneration,butthatinthedorsomedialportionthecellshadbecomeopaquediscs,withthenucleieitherdarkandshrunkenorbreakingupintogranules.Inthisregiontherewasconsiderablegliomesodermalreaction,andinHolzerpreparationsamoderatediffuseglialfibrosiswasseen.Nodemyelinationwasnoted.Thebloodvesselsofbrainandmeningesshowedmoderatefibrosis,andinthecerebralwhitematterararefactionoftissueinclosevicinitytothethickenedvesselshadproducedthepictureofsmall"criblures."Apartfromthischange,however,nosignificantabnormalitywaspresentinthecerebralhemispheres.Case2.-Amanaged67wasadmittedtoKing'sCollegeHospitalwithasixweeks'historyofepigastricpainandlossofweight.Hisconditionrapidlydeteriorated,andhediedtwoweeksafteradmission.Thediagnosiswasofcarcinomaofthebladder,withmetastaticdepositsintheliver.AutopsyFindings.-Asquamous-celledcarcinomaofthebladder,withsecondarydepositsintheliverandtheabdominallymphglands,wasdiscoveredatnecropsy.Thebrainweighed1,440g.Themeningeswereopaqueoverthewholeofthevertex,andtherewasslightgeneralizedconvolutionalatrophy.Apartfromthesefindingsnogrossabnormalitywasnoted.Histologically,theonlypathologicalfindingsofimportancewereconfinedtothecerebellum.InNissl-stainedsectionsthePurkinjecelllayerandtheBergmannlayerwerewellpreserved.Thegranularlayer,however,showedvaryingdegreesofdegenerativechange.Somefoliawerenormalinappearance,whilstinothersanalmostcompletenecrosisofthegranularlayerhadoccurred.Eventhegranuleslessseverelyaffectedshowedsomelossofconfigurationofthenuclearchromatin,thenuclearmembranebeingnolongervisibleandtheshapenolongerquiterounded.Manycells,however,hadalmostentirelylosttheiraffinityforcresylviolet,amoreorlesshomogeneousandmarkedlydeformeddiskaloneremaining.Stillmoreadvanceddegenerationwasindicatedbyfaintshadowsofnucleardebris.TheGolgicellsofthislayerwerewellpreserved,standingoutinmarkedreliefamidsttheareasofnecrosis.Inmyelinpreparationsthegranularlayerwasonlylightlystained,butdidnotshowtheextreme"translucency"ofothercasesinthisseries.Therewasnolossofmyelinatedfibres,the"light"stainingresultingfromthepoorlystainedcellbodies.Silverpreparationsalsoshowednolossoffibresinthegranularlayer,althoughbeadingwaspresent.Basketcellsandfibreswerewellpreserved.Themole-cularlayershowednoabnormality.Thedentatenucleuswasintact,thenervecellsshowing289ii.9IIIIilI A.D.LEIGHANDA.MEYER4J~~~~~~Ka~~~~~A~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~4.e10~~~1v.~~~~~~~~~~~~~~I--w,s\*~~~~~~~~~~~~~~~tIf**w,*.%ta-4#.Cr'-.t,.:1...e00C-0cn_,tEI-AN1*0CNc*$.gcsUcn'C=ZoCXo^u8WC/2W'0gn.rOo066rZZbe0.0290.'i *-AN1#IRi**1w09.9d9#%4-'*ar2-%F,-4*Va,u'*~~~~*..*9.,C)v0._*-O)0CZCtCUCtCZ0~~~v7CZQ.'l�LXCZCXC-S;-0.-}-~-xD--l-iCO_._~CUS.VC-viiv~~~~C-44s."I.:....--*A*'_.,0^_,~~~~~~'VaJJ'|_:I4Iil.,I*r*1.t.1-40,...!"IIp..'tIIk...1.IL;lII.,!,,.'S.!'.-v.-,OF A.D.LEIGHANDA.MEYERalipofuscinchangecommensuratewiththeageofthepatient.Theolivesshowedageneraldiminutioninthenumberofnervecells,allofwhichshowedextremepigmentatrophy.InHolzerpreparationstherewasgliosisinthecere-bellarwhitematter(butnot

inthecortex)andintheregionofbothinferiorolives.Therewasnoabnormalityintheremainderofthebrainorinitsbloodvessels.Theremainingsixcasesshowedsubstantiallysimilarappearancesandwillnotbedescribedindetail,TherelevantdataappearinTableI.GroupII:HypoglyciemicandDiabeticComaInthisgroupwehadfourcases.Case12isdescribedbelowindetail,andtherelevantdataforthewholegroupareabstractedinTableII.Case12.-Awomanaged30,adiabeticof8years'duration,wasadmittedtoKing'sCollegeHospitalinhypoglyciemiccoma.Inspiteofvigoroustreatment,shediedafteraperiodofcomalastingseventeendays.AutopsyFindings.-WidespreadcavitationoftheSwiss-cheesetypewasseenwithinthecerebralwhitematterandthebasalganglia.Thiswasduetoaprofusegrowthofthegas-formingCl.Welchii,whichwasfoundBLEIGRANULARLAYERDEGENERATIONASSOCIATEDWITHVISCERALCARCINOMATAAge-DurationPahlGranularInferiorOtherC.N.S.abnormalitiesCaseatofillnessPtoogicalCach-SecondaryComplicationslayerolivedeath(mths.)diagnosisexiadepositsdegeneration6212Carcinomaof+-Wernicke'sGeneralized,SevereModeratefibrosisofcerebralstomachsyndrome;butmorechanges;andmeningeatbloodvessels,maniamarkedinoutfallofproducingtheappearanceofdorsalcere-nervecellssmall"criblures"inthebe1Iumindorso-cerebralwhitematter.(Fig.1)medialpor-tionofoli-varynucleus2676Carcinomaof+Hepatic-PatchySlightgeneralNilbladdernerve-cellreduction3603Carcinomaof+Hepatic_Patchy(Fig.2)SevereNilgall-bladderchanges;almostcom-pletedis-appearanceofnervecells4668Carcinomaof+-OralsignsofPatchy(Fig.3)SlightgeneralLipofuscinchangeinnervebladdervitaminBnerve-cellcellsofcerebralcortex,deficiencyreductionsubcorticalgreymasses,andbrainstemnuclei.Severepigmentatrophyofcellsofdentatenucleus.Moderategliosisinregionofdentatenucleusandincerebellarwhitematter,andinall;marginalzones.56612Carcinomaof+HepaticGlycosuriaGeneralized;NormalNilpancreassomereduc-tioninnum-berofPur-kinjecells6516Carcinomaof+CerebralMentallyGeneralizedSomegeneralNilbreastdefectivediminutionsincebirthofnervecells,withrarefactionofdorsalflexure7536Carcinomaof+Cerebral"Korsakow'sGeneralizedSlightgeneralNilbronchuspsychosis"nerve-cellreduction875UnknownCarcinomaof+__GeneralizedNormalGeneralpigmentatrophy,with.stomachsomereductionofthenervecellsofthecerebralcortex.Etatcribleinputamenandpallidum.Fibrosisofcerebralvesselswithvaryingdegreesofcerebralatheroma.292 DEGENERATIONOFTHECEREBELLUMTABLEII293DEGENERATIONOFTHEGRANULARLAYERASSOCIATEDWITHHYPOGLYCEMICANDDIABETICCOMAAgeDuration1ClinicalCach-CornlcaGranular01iarCaseatofdiagnosisCactplicalayernivaryOtherC.N.S.abnormalitiesdeathillnessdigoixatosdegenerationnce921Coma-Hypoglycxemic--DiffuseSomegeneralizedWidespreadseverenecrosisinfrontal,60hourscoma;schizo-reductioninparietal,andtemporalcortex,andphrenianervecellsinSommersectorandendfoliumoftheAmmonshorn.Dentatenucleus-generalcellularreductionwithpigmentatrophyofthosenervecellsremaining.1017Coma-Hypoglyciemic--Diffuse;Pur-NoabnormalityFociofcellularnecrosisinthefifth36hourscoma;schizo-kinjecellsandsixthlayersofthefrontal,phreniashowedacuteparietal,andtemporalcortex.ThecellchangeSommersectorandendfoliumofAmmonshornwerealmostdevoidofnervecells.1110Coma-?Hy

poglycxmic-EpilepticDiffuseNoabnormalitySectornecrosisofAmmonshorn,?dura-coma?hyper-attacksmarginalgliosisofthecerebraltiongIycsemicsincecortex,andsomeperivascularandcoma;dia-age6diffusegliosiswithinthecerebralbetesmellituswhitematter.1230Coma-HypoglycxemicDiffuse(Fig.4)NoabnormalitySeefullcasereport.17dayscoma;dia-betesmellitusinsidethebloodvesselsandtheadjacenttissues.Thereshowingcompletedisappearanceoftheirnuclei,andwasnoinflammatoryreactiontothebacillaryinvasion,staininghomogeneouslywithbasicdyes(Fig.4).Thissothatitmusthaveoccurredimmediatelybefore,orchangeinthegranularlayerwaspresentthroughoutjustafter,death.Histologicallytherewaswidespreadthecerebellarcortex.Inmyelinstainsthepalegreylossofnervecellsinthecerebralhemispheres,withaappearanceofthislayercontrastedwiththedarklycorrespondingmicroglialandneuroglialreaction.staininglayerofBergmannglia.SilverimpregnationThesepathologicalfindingshavebeenreportedinfullshowedpreservationofbasketcellsandfibres,withnoelsewhere(Lawrenceandothers,1942)andwerediminutioninthefibresofthegranularlayer.InconsideredtobetheresultoftheprolongedcerebralScharlachRsectionssomeincreaseinlipoidscouldbeanoxia.seeninthehistiocyticandmicroglialcells,inboththeThecerebellumshowedatypicalgranularlayerBergmannandgranularlayers.NofibrousgliosiswasdegenerationwhichhadnotbeennoticedatthetimeobservedwiththeHolzertechniqueinanyoftheoftheearlierpublication.InNisslpreparationsthecerebellarlayers.Purkinjecellswerewellstainedandinnowayabnormal.ThedentatenucleusandtheolivesofthisbrainwereThegranules,however,werethinnedout,manycellsnotavailable.TABLEIIIGRANULARLAYERDEGENERATIONINAMAUROTICIDIOCYAgeDurationClinicalCach-Complica-GranularOlivarmdaeathillesdiagnosisexiationslayernuleivrOtherC.N.S.abnormalitiesdeathillness~~~~~~~~degenerationnce1319SeveralAmauroticfamily+EpilepticfitsDiffuse;gliosisNotavailable"Ballooning"ofnervecellsyearsidiocypresentthroughoutthebraincharac-teristicofamauroticidiocy.1415SeveralAmauroticfamily+HypostaticDiffuse(Figs.Severelyaffected,"Ballooning"ofnervecellsyearsidiocypneumonia;5,6).Gliosislossofnervethroughoutthebraincharac-epiIepticpresentcellsparticularlyteristicofamauroticidiocy.fitsindorsomedialportionofnucleus15?SeveralAmauroticfamily+NoclinicalDiffuse;gliosisNoabnormality"Ballooning"ofnervecellsyearsidiocynotesavail-presentthroughoutthebraincharac-ableteristicofamauroticidiocy.1619SeveralAmauroticfamily+NoclinicalPatchy;severeNotavailable"Ballooning"ofnervecellsyearsidiocynotesavail-outfallofPur-throughoutthebraincharac-ablekinjecells;teristicofamauroticidiocy.gliosispresent A.D.LEIGHANDA.MEYERGroupIII:AmauroticIdiocyAgainfourcaseswereavailableforexamination.Case14isdescribedinfull,andTableIIIsummarisestheremainder.Case14.-Aboywasadmittedtohospitalattheageof11years.Hewasblindandsufferedfromepilepticattacks.Hedeterioratedslowlyforfouryears,becomingbedriddenandincontinent.Hefinallydiedattheageof15,asaresultofhypostaticpneumonia.AutopsyFindings.-Nomacroscopicabnormalitywasrecordedinthebrain.Histologicalexaminationdis-closedballooningandlipoidinfiltrationinthenervecellsthroughoutthenervoussystem.Inthecere

bellumagranularlayerdegenerationwaspresent.Withmyelintechniquesthecharacteristic"lightening"ofthegranularlayerwasclearlydemonstrated(Fig.5).InNisslsectionsthePurkinjecells,thoughglobularinshapeandinfiltratedwithlipoidmaterial,stainedwell,andshowednodiminutioninnumber.Incontrast,hardlyanormalgranulecellcouldbefoundthroughoutthewholecerebellarcortex(Fig.6).TheGolgicellsofthislayer,however,hadsurvivedintact.TheBergmanngliawasnotproliferated.InHolzerpreparationsaslightfibrousgliosiswasseeninthewhitematterofthelobulesandaroundthedentatenucleus.Inthegranularandmolecularlayerstherewasaconsiderableastrocyticgliosis.Bielschowskysilverpreparationsshowedalmostcompletedisappearanceofbasketcellsandtheirfibres(Fig.7).Theradialfibresofthemolecularlayerwereabsent.ThedendriticprocessesofthePurkinjecellsandtheiraxonesshowednoabnormality.Inthegranularlayertherewasgreatdiminutioninthenervefibres.Thecellsofthedentatenucleuswerenotdiminishedinnumber,butshowedwellmarked"ballooning"andlipoidinfiltration.Inthemedullatheinferiorolivesprovedtobeseverelyaffected.Withlowmagnificationageneraldiminutionincellswasatonceapparent,whilstinthedorsomedialportionofthenucleushardlyanervecellremained.Withhighermagnificationtheremainingcellsshowedmuchlipoidinfiltration,withdistortionoftheircellbodies.Holzerpreparationsdisclosedafibrousgliosisintheregionofbothinferiorolives.DiscussionInthefirstplaceitshouldbeemphasizedthatwehaveexcludedanycaseinwhichthepathologicalsignsofdegenerativechangesinthegranularcellswerequestionable.FaintstainingwiththeNisslmethodalone,withoutthephenomenaofconglutina-tion,deformation,orothernuclearorcytoplasmicsignsofcelldegeneration,wasneverconsideredaspathognomonic,asthismaybeastainingartefact,particularlyinbrainswhichhavebeenlonginformalin.Asithappens,mostofourmaterialwasembeddedsoonafterfixationinformalinwascom-pleted.Comparisonwithotherstainingmethods,especiallymyelinstains,washelpfulintheevaluationofthechanges.Even.aftertheapplicationofsuchrigidcriteria,ourresultsconstituteanimpressiveconfirmationoftheclaimsofBertrandandhisassociates,particu-larlywithregardtothecarcinomagroup,inwhicheightoutoffourteencasesshowedthistypeofdegeneration.InthediabeticandhypoglycrmicgroupourresultsdifferfromthoseofBertrandandhisassociates,forintwocasesofundoubteddiabeticcomathegranularlayerwasnormal.Inthreeoftheremainingfourcasesdeathwasduetoirreversiblehypoglycremiccoma;intwocasesitoccurredinnon-diabeticschizophrenicpatients.Thisis,toourknowledge,thefirstreportoftheoccurrenceofselectivedegenerationofthegranularlayerinirreversiblehypoglycxmiccoma.Thechangewasfoundinthreeofninefullyinvestigatedcasesofhypoglycaemiccoma.TheresultsinourfourcasesofamauroticidiocyconfirmthefindingfirstestablishedbyBielschowsky(1920),ofafrequentlyselectiveaffectionofthegranularlayerinthiscondition.Thehistologicaldetailofthedegenerationofthegranularlayerrequireslittlecomment.InallrespectsthisagreeswithpreviousdescriptionsandinparticularwiththatgivenbyBertrandandothers(1942),Williams(1934),Winkelman(1943),andMeyer(1944).Inallofourthreegroupsofcasesthechangesrangefromearlyconglutinationtoacompletenecros

isofthegranulecells.Thechangeswereeitherfocal,asinCase3(Fig.2),inwhichonlyafewlobuleswereaffected,ordiffuse(Fig.1),thelatterbeingthemorecommonfinding.Evenwithsuchadiffusechange,however,therewasanotinconsiderablevariationintheseverityofthelesionfromonelobuletotheother.Inviewoftheseobservationsitisdifficulttodrawasharpdividinglinebetweenfocalanddiffusevarietiesofcerebellaratrophy,aswasdonebyBrouwerandBiemond(1938).Theseauthorsregarddelayedcerebellaratrophyasanexampleoftheformer,andthecarcinomatousgroupasanexampleofthelatter.Itmaybethatthecerebellardegenerationoccurringwithcarcinomata,andwithotherdiseases,tendseventuallytobecomediffuse.Thesamereservationappliestothepredilectionofthedegenerationforthedorsalandmedialaspectsofthecerebellum,afeaturewhichhasbeenconsistentlyreportedinmanyexamplesofthedelayedMarie-Foix-Alajoua-ninevariety.Bertrandandhisassociatesalsoobservedthisselectiveincidenceinthemajorityoftheircasesassociatedwithcarcinomaanddiabetes,butinourmaterialarelativepreservationoftheventralpartswasseenononlytwooccasions.Thus,thereisnoabsoluteaffinityforthedorsalpartsof294 DEGENERATIONOFTHECEREBELLUMthecerebellumatleastinthecarcinomatousanddiabeticvarietyofcerebellarlesion;itseemsmorelikelythatsuchatransitionaldistributiontendstooccurduringtheevolutionoftheprocess.Likewise,whenwespeakofaselectivedegenera-tionofthegranularlayerthisdoesnotmeanthatthelesionswereabsolutelyconfinedtothispartofthecerebellum.Infact,inoneofthecarcinomagroup(Case5),andoneoftheamauroticgroup(Case13)thePurkinjecellswereslightlyreducedinnumber.TherewasamoreseverelossofPurkinjecellsinanotheroftheamauroticcases(Case16)and,lastly,inonehypoglycaemiccasethePurkinjecellsshowedverymarkedacutechromatolysis(Case10).Evidently,thepredilectionforthegranulesisnotabsolute.Thismayexplainwhyinsomeofthecarcinomacasesreportedintheliterature(forexample,Zuilch,1936;BrouwerandBiemond,1938)damagewasactuallymorecon-spicuousinthePurkinjecellsthaninthegranularlayer.In'thedelayedMarie-Foix-AlajouaninegroupthePurkinjecellaffectionisusuallythemainfeature,althoughGarcinandothers-(1940)reportedparticularlyseveredegenerationofthegranules.Astrikingfeatureinallourcasesassociatedwithcarcinoma,hyperglycemia,andhypoglycemia,istheabsenceofanappreciableglialproliferationwithinthegranularlayerorotherpartsofthecerebellarcortex.Insomecasestherewasamoderategliosisofthewhitematter.ThisabsenceofglialproliferationwasalsoremarkeduponbyBertrandandhiscolleagues.Forthisreason,andbecausenocerebellarsignswerenotedduringlife,theseauthorsconcludedthatthecerebellardegenera-tionoccurredintheterminalphaseofthedisease,forinstance,duringtheterminalcachexiaofthepatientssufferingfromcarcinoma.Itwillberememberedthatinnoneofourcarcinomaordiabetic-hypoglycaemiccaseswereclinicalsignsoldisturbedcerebellarfunctionnoted.Ifthesecaseswithoutglialfibrosisaretoberegardedascasesofrecentcerebellardegeneration,glialfibrosisshouldbemarkedincaseswhichhadexhibitedsignsofcerebellardysfunctionforsometime.Therewasinfactsuchaglialfibrosisinallourfouramauroticcases.Unfortunatelynoreliableclinicaldataareavailabletous,butinNorman's(1940)casesofmentaldef

ectcerebellarsignshadbeenpresentforatleastthelastyearsoflife,andtherewasveryextensiveglialsclerosisofalllayersofthecerebellarcortex.Atthepresentjunctureitisnotpossibletooffermorethantentativesuggestionsastothepossiblepathogenesisofthistypeofcerebellardegeneration.Itseemsthatafactorisinvolvedwhichmustbecommontosuchwidelydifferentconditionsassomaticcarcinoma,hyper-andhypoglycemiccoma,andlipoidosis.CourvilleandFriedman(1940)indiscussingthepathogenesisofthedelayed(paren-chymatous)cerebellaratrophies(intheirterminology"chronicprogressivedegenerationofthesuperiorcerebellarcortex")suggestedthatcirculatorydisturbancesproducedbyarachnoidalthickeningwereofimportance.Wehaveexaminedthebrainofapatient(notincludedinthisseries)where,inadditiontogeneralizedgranulardegenerationtherewasasevere,focalchangeintheneighbourhoodofasubarachnoidhaemorrhage.Quiteclearlyinthiscase,circulatorydisturbancesresultingfromthehxmorrhagehadplayedapartinproducingafocalgranulardegeneration.However,itisnotclearbywhatmechanismthefocallesionwasproduced,andindeedinallthecasesreportedheretherewasnoevidenceofadisorderofthebloodvessels.Bothvasomotordisturbancesanddefectiveutilizationofoxygenbythetissuesmaybeofimportance;indeed,itwasconsideredthatsuchmechanismswereintimatelyconcernedintheproductionofthecerebrallesionfoundinhypoglycemiccoma(Lawrenceandothers,1942).Itispossiblethathypoxiaarisesduringtheterminalphaseoftheillnessinpatientssufferingfromcarcinoma.How-ever,inviewofthechronicityofthelesion,itismoredifficulttopostulatethatcirculatorydisturb-ancesoranoxiaoccurredinamauroticidiocy.NorwoulditbeeasytoexplainwhythebruntofthedegenerationshouldhavefallenonthegranulesratherthanonthePurkinjecells,which,sinceSpielmeyer'sobservations(1922),havesooftenbeenshowntobeespeciallysusceptibletooxygenwant.Inafairlylargeseriesofcasesofalltypesofanoxicconditionsthathavebeeninvestigatedinthislaboratorythegranulesusuallyremainintact,whereasthePurkinjecellsfrequentlyshowthecharacteristichomogenizingchanges.Clearlyanoxiaaloneisnotafactorlikelytobeofoetiologicalsignificance.Itmightbeexpectedthatsuchacomparativelywell-definedmetabolicdisorderasamauroticfamilyidiocywouldthrowsomelightontheaetiologyofthegranularlayerchange.Infact,however,solittleisknownofthemetabolicprocessesthatoccurinthisconditionthatitwouldbespeculativetoascribethegranularlayerdegenerationtoanyspecificfactor.However,itmaybesaidthatinbothourcarcinomagroupandamauroticidiocygroup,cachexiawasamarkedfeatureasinothercasesreportedintheliterature(Lhermitte,1922;Zulch,1936;Norman,1940;Buchananandothers,1947).Enteritis,alcoholismandcirrhosishepatis,whichmightalsobeexpectedtoleadtoprofoundmetabolicdisorder,werealsoprominent295 A.D.LEIGHANDA.MEYERinthecasesreportedbyMurri(1900);Rossi(1907);Lhermitte(1922);Luithy(1930);Zuilch(1936);Norman(1940),andAlessi(1940).Whatisknownaboutthebiochemistryofthisterminalcachexiadoesnotlenditselftoevenatentativehypothesis.InCase1thecombinationofgranularlayerdegenerationwiththeWernicke'ssyndromesuggestedathiaminedeficiency,asdidUpner'sexperimentalfindings;butbothconditions,clearly,occurmorefrequentlyindependentlythanincom

bination.Therelationshipofolivarychangestothiaminedeficiencyappeared,however,tobemoreconsistent.Whiletheinferiorolivesareknowntobeveryreactivetoallkindsofnoxiousagencies,previousanalysisoftheliteratureandpersonalexperienceledoneofustoassumethatolivarydamagefrequentlyoccursinconditionsassociatedwithmetabolicand,inparticular,withnutritionaldisturbances(Meyer,1944).Particularattentionwas,therefore,paidtotheincidenceofolivarychangesinthepresentmaterial,but,again,althoughthetwolesionsfrequentlyoccurtogether,theyoccurtoooftenindependentlytoencouragetheassumptionofacommonspecificfactor.Pellagraalsocanbeexcluded;therewasnotonecaseinourhistologicallyverifiedPellagramaterialofabouttwentycaseswhichshoweddegenerationofthegranularlayeroroftheolives.Whilethereisthusthesuggestionthatadeficiencyfactorisconcernedintheactiologyofthegranularlayerdegeneration,itisclearthatafinalsolutionoftheproblemmustawaitfurtherinvestigations.SummarySelectivedegenerationofthegranularlayerofthecerebellumisdescribedineightcasesofvisceralcarcinoma,fourcasesofhypo-orhyperglycemiccoma,andfourcasesofamauroticidiocy.Thepathogenesisofthelesionisatpresentobscure,butattentionhasbeendirectedinthispapertothepossibilitythatsomedisturbanceofmetabolism,possiblyanutritionaldeficiency,isresponsibleforthedegeneration.Itisimpossibletomakeindividualacknowledgmentstothestaffofthevariousmentalhospitalsfromwhichwehavereceivedthebrainmaterial,butweshouldliketorecordourspecialgratitudetoProf.H.A.MagnusofKing'sCollegeHospital,whosupplieduswiththemajorityofthecarcinomaanddiabetescases.REFERENCESAlessi,D.(1940).Riv.Patol.Nerv.Ment.,55,148.Archambault,LaS.(1918).J.Nerv.Ment.Dis.,48,273.Bertrand,I.,andGodet-Guillain,J.(1942).C.R.Soc.Biol.Paris,136,664.andTiffeneau,R.(1942).Ibid.,136,500.Bielschowsky,M.(1920).J.Psychol.Neurol.,26,123.Brouwer,B.,andBiemond,A.J.(1938).J.BeigeNeurol.,38,691.Buchanan,A.R.,Overholt,L.C.,andNeubaerger,N.T.(1947).J.Neuropath.exp.Neuroi.,6,2,152.Casper,J.(1929).Z.ges.Neurol.Psychiat.,33,854,885-6.Courville,C.B.,andFriedman,A.P.(1940).Biull.LosAng.Neurol.Soc.,5,171.Ferraro,A.,andMorrison,R.(1928).Psychiat.Quart.,2,506.Foix,C.,Marie,P.,andAlajouanine,T.(1922).Rev.Neurol.,38,849,1082.Garcin,R.,Bertrand,I.,andGodet-Guillain,J.(1940).Ibid.,72,724.Greenfield,J.G.G.(1934).Brain,57,161.Hunter,D.,Bomford,R.B.,andRussell,D.S.(1940).Quart.J.Med.,n.s.9,193.Kennard,M.(1935).Proc.Kon.Akad.vanWetenschteAmsterdanm,38,544.Lawrence,R.D.,Meyer,A.,andNevin,S.(1942).Quart.J.Med.,11,181.Lhermitte,J.(1922).Rev.Neurol.,38,313.Liithy,F.(1930).Z.Neurol.,57,319.Mleyer,A.(1932).Z.ges.Neurol.Psychiat.,139,422.(1944).J.Neurol.,Neurosurg.Psychiat.,7,66.Murri,A.(1900).Riv.Crit.Clin.Med.,1,593.Norman,R.M.(1940).Brain,63,365.Parker,H.L.,andKernohan,J.W.(1933).Ibid.,56,191.Rossi,I.(1907).Nouv.Iconog.Saltpet.,20,66.Scherer,H.J.(1931).Z.ges.Neurol.Psychiat.,136,559.--(1933).Ibid.,145,335.Shinosaki,T.(1926).Ibid.,106,483.Spielmeyer,W.(1925).Ibid.,99,756.Upners,T.(1939).Ibid.,166,623.Williams,E.Y.(1934).Arch.Path.,17,206.Winkelman,N.W.(1943).J.Neuropath.exp.Neurol.,2,413.Zulch,K.J.(1936).Z.ges.Nfeurol.Psychiat.,156,493,57