Brain circuits implicated in psychogenic paralysis in conversion disorders and hypnosis

(1)

Article

Reference

Brain circuits implicated in psychogenic paralysis in conversion disorders and hypnosis

VUILLEUMIER, Patrik

Abstract

Conversion disorders are defined as neurological symptoms arising without organic damage to the nervous system, presumably in relation to various emotional stress factors, but the exact neural substrates of these symptoms and the mechanisms responsible for their production remain poorly understood. In the past 15 years, novel insights have been gained with the advent of functional neuroimaging studies in patients suffering from conversion disorders in both motor and non-motor (e.g. somatosensory, visual) domains. Several studies have also compared brain activation patterns in conversion to those observed during hypnosis, where similar functional losses can be evoked by suggestion. The current review summarizes these recent results and the main neurobiological hypotheses proposed to account for conversion symptoms, in particular motor deficits. An emerging model points to an important role of ventromedial prefrontal cortex (VMPFC), precuneus, and perhaps other limbic structures (including amygdala), all frequently found to be hyperactivated in conversion disorders in parallel to impaired recruitment of primary motor and/or [...]

VUILLEUMIER, Patrik. Brain circuits implicated in psychogenic paralysis in conversion disorders and hypnosis. Clinical Neurophysiology , 2014, vol. 44, no. 4, p. 323-327

DOI : 10.1016/j.neucli.2014.01.003 PMID : 25306073

Available at:

http://archive-ouverte.unige.ch/unige:40632

Disclaimer: layout of this document may differ from the published version.

(2)

Please cite this article in press as: Vuilleumier P. Brain circuits implicated in psychogenic paraly- Disponibleenlignesur

ScienceDirect

www.sciencedirect.com

REVIEW/MISE AU POINT

Brain circuits implicated in psychogenic paralysis in conversion disorders

and hypnosis

Circuits cérébraux des paralysies psychogènes liées aux troubles de conversion et à l’hypnose

P. Vuilleumier

LaboratoryforBehavioralNeurologyandImagingofCognition(LABNIC),DepartmentofNeuroscience (NEUFO),UniversityMedicalCenter(CMU),1,rueMicheliduCrest,1205Geneva,Switzerland Received27December2013;accepted12January2014

KEYWORDS Conversion;

Hysteria;

Hypnosis;

Paralysis;

Brain;

Neuroimaging;

Affect;

Self;

Precuneus;

VMPFC

Summary Conversiondisordersaredefinedasneurologicalsymptomsarisingwithoutorganic damagetothenervoussystem,presumablyinrelationtovariousemotionalstressfactors,but theexactneuralsubstratesofthesesymptomsandthemechanismsresponsiblefortheirpro- ductionremainpoorlyunderstood.Inthepast15years,novelinsightshavebeengainedwith theadventoffunctionalneuroimagingstudiesinpatientssufferingfromconversiondisordersin bothmotorandnon-motor(e.g.somatosensory,visual)domains.Severalstudieshavealsocom- paredbrainactivationpatternsinconversiontothoseobservedduringhypnosis,wheresimilar functionallossescanbeevokedby suggestion.Thecurrentreviewsummarizestheserecent resultsandthemainneurobiologicalhypothesesproposedtoaccountforconversionsymptoms, inparticularmotordeficits.Anemergingmodelpointstoanimportantroleofventromedialpre- frontalcortex(VMPFC),precuneus,andperhapsotherlimbicstructures(includingamygdala), allfrequentlyfoundtobehyperactivatedinconversiondisordersinparalleltoimpairedrecruit- mentofprimarymotorand/orsensorypathwaysatthecorticalorsubcortical(basalganglia) level.Thesefindingsareonlypartlysharedwithhypnosis,whereincreasesinprecuneuspre- dominate,togetherwithactivationofattentionalcontrolsystems,butwithoutanyactivationof VMPFC.BothVMPFCandprecuneusarekeyregionsforaccesstointernalrepresentationsabout theself,integratinginformationfrommemoryandimagerywithaffectiverelevance(inVMPFC) andsensoryoragencyrepresentations(inprecuneus).Itisthereforepostulatedthatconversion deficitsmightresultfromanalterationofconscioussensorimotorfunctionsandself-awareness undertheinfluenceofaffectiveandsensoryrepresentationsgeneratedintheseregions,which mightpromotecertainpatternsofbehaviorsinresponsetoself-relevantemotionalstates.

E-mailaddress:patrik.vuilleumier@unige.ch

http://dx.doi.org/10.1016/j.neucli.2014.01.003

(3)

Please cite this article in press as: Vuilleumier P. Brain circuits implicated in psychogenic paraly- MOTSCLÉS

Conversion; Hystérie; Hypnose; Paralysie; Cerveau; Neuroimagerie; Affect;

Soi; Précuneus; CPFVM

Résumé Lestroublesdeconversionsontdéfiniscommedessymptômesneurologiquesappa- raissant sans lésion organique du système nerveux, habituellement en relation avec des situationsdestressémotionnel,maislessubstratscérébrauxexactsdecessymptômesetles mécanismesresponsablesdeleurapparitionsontencoremalcompris.Aucoursdes15dernières années,denouvellesdonnéesontétéaccumuléesgrâceaudéveloppementdelaneuroimagerie fonctionnellechezdespatientssouffrantdetroublesdeconversionmoteursounonmoteurs (parexemple somatosensoriel,visuel).Plusieurs étudesontégalementcomparé l’activation cérébralelorsdeconversionàcelleobservéelorsdel’hypnose,oùdesdéficitsfonctionnels similairespeuvent êtreinduitsparsuggestion.Cetterevuerésume ces donnéesetlesprin- cipales hypothèses neurobiologiquesproposées pourexpliquerla genèse des symptômes de conversion,enparticulierdesdéficitsmoteurs.Unaspectémergentimportantconcernelerôle potentiel ducortex préfrontalventromédian(CPFVM), duprécuneus, et peut-être d’autres structureslimbiques(notammentlesamygdales),tousfréquemmentsuractivéslorsdetroubles deconversionetassociésàunealtérationde l’activationdesvoiesmotricesousensorielles élémentairesauniveaucorticalousous-cortical(notammentganglionsdelabase).Ceseffets nesontquepartiellementpartagésavecl’hypnose,oùl’activationduprécuneusprédomine, accompagnéed’uneactivationdessystèmesdecontrôledel’attention,maissansactivation duCPFVM. Le CPFVMetleprécuneus constituentdeux régionscrucialespourl’accèsàdes représentationsinternes du soi,intégrant des représentationsenmémoire avec leurvaleur affective(CPFVM)etdesinformationssensoriellesoud’agentivité(précuneus).Cesdonnées suggèrentquelessymptômesdeconversionpourraientrefléterunealtérationdes fonctions sensorimotricesetlaconsciencedesoisousl’influencedereprésentationsaffectivesetsenso- riellesgénéréesdanscesrégions,favorisantl’expressiondecertainscomportementsenréponse àdesétatsémotionnelsparticuliersetpertinentspourlesoi.

Introduction

Conversiondisordershavebeen knowntophysicians since ancienttimesandconsist ofneurologicaldysfunctionthat occurs in the absence of organic lesions within the ner- voussystem, but is distinct fromother major psychiatric symptomssuch aspsychosisor depression.Thus, patients maypresentwithaparalysisofoneor morelimbs,tremor orinvoluntarymovement,anesthesia, blindness,deafness, or various other deﬁcits, without any organic disease despitecarefulinvestigation andoftensigns suggestiveof anon-organic/non-physiologicdisruptionoffunction[6,92].

Importantly,thesesymptomsarenotintentionallyproduced inordertodeceivephysiciansorobtainsomedirectmate- rialgain,butratherappeartoresultfromagenuinebeliefor subjectiveexperienceofthepatientsthathe/sheissuffer- ingfromaparticulardisorder.Whilethesesymptomswere longlabeledas‘‘hysteria’’,thistermhasbeenreplacedby

‘‘conversion’’ in modern psychiatric classifications of the 20thcentury.Thischangeinterminologywaspartlydueto increasingrecognitionthatsuchproblemscouldaffectmen aswellaswomen,particularlyafterwar episodes[7],but alsoreflectedthemajorinfluence of SigmundFreud [30], whoproposedthatthesesymptomsresultfromatransfor- mation of a psychological conflict into symbolic physical manifestations due to repression in the unconscious (see other articles in this issue). However, although this diag- nostictermisuniqueinpsychiatrybydirectlyreferringtoa specificcausativemechanism,theexactcausesandmecha- nismsunderlyingsuchconversion[2]ofthepsycheintothe bodyactuallystillremainunclearandspeculative.Further- more,inthelatestversionofdiagnosticcriteriaestablished

byDSM-5in2013[4],theneedtoidentifyaparticularpsy- chologicalcausehasbeenreplacedbymoregeneralcriteria implyingsignificantdistressordisabilityinnormalactivities duetothesymptoms,achangethatwasmotivatedbythe factthatpsychologicalfactorsareoftendifficulttoidentify with certainty, or are found but with only weak or puta- tiveassociationwiththediagnosis.Inaddition,ithasbeen suggested thatrequiring theidentification of explicitpsy- chologicalfactorsactingonthepatient’sunconsciousisnot onlydifficultandunreliableinpractice,butalsoinherently maintainsaconceptualdualismbetweenmindandbodythat doesnotfitwiththecurrentneuroscientificview,andthat isgenerallynotadvocatedforothermentaldiseases[4].

Uncertainties regarding the definition and cause(s) of conversion disorders are also reflected by the fact that theyareclassifiedindistinctdiagnosticfamiliesintheDSM frameworkputforwardbytheAmericanPsychiatryAssoci- ation [2] and theICD taxonomy established by the World Health Organization [65], being listed among somatoform disordersintheformerbutamongdissociativedisordersin thelatter.Dissociationalsoreferstoadichotomybetween consciousandunconsciousprocessesdatingbacktotheideas of Janetinthe late19thcentury, whichprobablyinspired Freud’sconception ofconversion, butthisimplies distinct psychodynamicmechanisms[89].Inaddition,theroleofdis- sociationinconversiondisordersalsoremainshypothetical.

Nevertheless,inthecurrentstateofpsychiatricnosography, psychogenicmotororsensorylossesareincludedinthesame diseasecategory aspsychogenicamnesiaaccording tothe ICD(dissociativedisorders)butindistinctcategoriesaccord- ingtotheDSM.Finally,therelationofconversiondisorders to other major psychiatric conditions is unresolved but a

(4)

Please cite this article in press as: Vuilleumier P. Brain circuits implicated in psychogenic paraly- high rate of comorbidity with depression and anxiety has

longbeenrecognized[7,8,75],suggestingapotentialpartial overlapofriskfactorsattheindividualorenvironmentallev- els.Fortheseandotherreasons,ithasbeenproposedthat apurelydescriptiveanduncommittedtermof‘‘functional neurologicaldisorder’’shouldbepreferredtorefertothe diagnosisofconversionorpsychogenicsymptoms[60,95].

As a consequence of the persistent uncertainties and enduringcontroversies incurrent medicalperspectiveson hysteria or conversion, thiscondition remains very poorly understoodandstillrarelystudiedtoday.Thisisallthemore surprisingthathistorically,inthelate19th—early20thcen- tury,hystericaldisordersprovidedmajorbuildingblocksfor both thetheoreticalfoundationsandclinicalsemiologyat thebirthofmodernneurologyandpsychiatry.Yet,overthe last100years,thisconditionhasremainedinagreybound- aryzonebetweenthesetwodisciplinesandreceivedmuch less interest for scientific inquiry from both neurologists and psychiatrists, in contrasttoother diseases that were notevenclearlyidentifiedinthoseearlydays.Thus,while we nowadayshave a muchbetterknowledge of braincir- cuitsimplicatedincomplexneurologicalmovementdeficits, suchasParkinsonorHuntingtondisease,orthoseaffectedin complexmentaldisturbancessuchasobsessive-compulsive disorders or depression, our conceptual approachto psy- chogenicdeficitsisstillverymuchbasedonideasinitially formulatedbyCharcotorFreud,andourmechanisticunder- standingoftheirputativeneural substratesinthebrainis stillrudimentary. This state of affairs is evenmore strik- ingwhenconsidering that conversiondisordersareamong themostcommondiagnosisinneurology,thoughttorepre- sentatleast4%ofalladmissionsonageneralclinicalwards andthusequalinfrequencytomultiplesclerosisorepilepsy [93,96].

The currentarticlewillpresent aselectiveoverviewof recentattempts toinvestigate theneuralunderpinningof conversion disorders,particularly thanks tothe adventof functionalneuroimagingtechniques.Thisreviewwillfocus on motor deficits such as unilateral paresis or paralysis, whichhave mostoften beenstudied. Recentfindingscon- cerning positivesymptomssuchasmovement disordersor dystonia can be found elsewhere (see Voon etal. in this issue). A large variety of psychogenic neurological symptoms can be seen asmanifestations of conversion, which addsfurtherdifficultiestoasystematicinvestigationofthe neural bases of this condition and makes the few extant studiesoftendifficulttocompare(fordetailedreviewssee [13,104,105]).Inspiteofthis,someconvergencebetween studies is starting toemerge and, together with increas- ingneuroscientificknowledge aboutthebrainmechanisms ofbehaviorandemotion inhumans,providenovelinsights intopossiblepathwaysthroughwhichpsycheandbodymay interactinconversiondisorders.

Early neurobiological accounts of conversion disorders

Attempts to deﬁne clinical characteristics and putative neural mechanisms of conversion date back to Charcot and his students or co-workers, including Janet, Babin- ski,andFreud(forhistoricaloverview,seeCrommelinckin

thisissue). Charcot believed that conversion (then called hysteria) resulted from an alteration of the function of thenervoussystem, in the absenceof macroscopically or microscopicallyvisible damage.This dysfunctionwascon- ceived asa ‘‘dynamic cortical lesion’’ considered due to a sort of ‘‘invasion’’ of the corresponding brain pathways by a ‘‘psychological image’’, leading to hyper or hypo-excitabilityofthesepathways[15].Accordingtothis perspective,abnormalexcitability couldproducevirtually anysymptom,andthusmimicvariousorganicneurological disorderssuchasparalysis(reﬂectingasuppressionofthe normal‘‘mentalimage’’ofmovements)orseizure(reﬂect- inganoverexpressionofsuchimages).Thetriggeringfactors couldbegeneratedinternallybyideasorexternallybysug- gestion.Inparticular,Charcotunderscoredthesimilarities of hysteria with hypnotic phenomena, which he used to reproduceorabolishvarioussymptomsinhisclinicallessons.

Atthesametime,Janetemphasizedthenotionthat‘‘fixed ideas’’could act in an unconsciouspart of themind and thatconscious andunconsciouspartscouldbecome disso- ciatedinspecialcircumstances suchashypnosisor strong emotionalstates[42].Theseideaswerefurtherelaborated byFreudandBreuer[30]whotheorizedontheroleofuncon- sciousmotivesthat wereoften relatedtosexual conflicts andexpressedthroughsymbolicsomaticsymptomsdue to repressionmechanisms.Thispsychodynamicviewledtothe terminologyof‘‘conversion’’disorder,whichisstillinusein currentmedicalnosology.Duringthesameperiod,Babinski [7]putmucheffortandrigorintoidentifyingclinicalsigns thatcoulddistinguishconversionfromorganicneurological disorders,insistingontheexistenceof‘‘typicalmanifesta- tions’’allowing apositive diagnosis (unlikeadiagnosis by exclusionasoftenadvocatedinmodernpsychiatryandneu- rologytextbooks). Among the latter, Babinski emphasized theroleofsuggestion(leadinghimtoproposeanewterm of‘‘pithiatism’’,derivedfromtheGreek‘‘peithein’’mean- ing ‘‘persuasion’’) but also individual predispositions and particularkinds ofemotional experiences, whichhequal- ifiedas‘‘organizedaffectivestates’’(as opposedtomore

‘‘transient’’emotions). Accordingto him, theseaffective states mainly have a permissive role that can strengthen unconsciousideasand‘‘putthemintoaction’’butideation istheprimarycausalfactor(generatedbyunconsciouspro- cessesorsuggestion).Interestingly,Babinskialsonotedthat thecausalroleofparticularideasandparticularemotional experiencesisoftendifﬁculttoestablishinaretrospective manner.In all the latter accounts, however, the putative neuroanatomicalsubstratesofconversionwereonlyloosely speciﬁed.

Insubsequentdecades,changesinexcitabilityofsensory or motor pathways associated with hystericalconversions were attributed to various physiological mechanisms or anatomicalloci,suchasthefrontalcortex[67],thalamus [50], or cingulate cortex [88] (for more detailed review see [104,105]). Yet these hypotheses were not based on experimentalorobjectiveevidence,butfollowedspecula- tionsorextrapolationderivedfromtheoriesofthenervous system available at the time. For example, Pavlov (best knownfor hiswork onassociative conditioning) proposed thatemotionalarousalarisinginsubcorticalcentersrequires inhibitory control fromcortical centers (especially in the frontallobe),suchthatinsomesituationsofstrongarousal,

(5)

Please cite this article in press as: Vuilleumier P. Brain circuits implicated in psychogenic paraly- inhibitorycontrolmightbecomeexcessiveandoverﬂowto

otherneurological pathways,thus causing paralysis,anesthesia or blindness [67]. An active inhibition of motor, sensory,orevencognitiveprocessespreventingtheirinte- grationintoconsciousawarenesswasalsoproposedbymore recentauthors[10,64,78].

Other accounts have built on the notion that conversionsymptoms may more often affect the left than right hemibody, on the one hand, and evidence that the right hemispheremightbedominantforemotionprocessing,on theotherhand,tosuggestaroleforimpairedcommunica- tionorintegrationbetweenthetwohemispheres[29,34,90].

However,thisleft preponderancehasbeen questioned by systematicreviews[94]andnotallconversionsymptomsare strictlylateralized.

Finally,following theearliernotionthat hysteriamight betriggered bystrong emotions or traumaticexperiences [7,30,42],severalauthorsproposedthatconversion symp- tomsmightreflectanexpressionofstereotypedbehavioral responseslinkedtomoreprimitiveformsofdefense,suchas motorarrest,freezing,playingdead,agitation,andsoforth [47,61,62,110].Inkeepingwiththisview, itisknownthat specificmotor behaviorscanbe controlledbytopographi- callyorganizedcircuitsinbasalgangliaandperiaqueductal grayunderthecontrolofcortical(frontal)andsubcortical limbiccenters[12,104].However,thesemodelsdidnotrely onempiricalevidence and may noteasily accountfor all typesofconversionsymptoms.Nevertheless,likeotherearly accounts reviewed above,some of these ideas find novel echoes in recent research using functional neuroimaging methods.

Empirical investigations of motor conversion disorders

Systematic scientiﬁc investigations to elucidate the neural mechanisms by which conversion disorders may alter motororsensoryfunctionsbegantoappearinthelate60s and 70s, with the emergence of EEG recordings and the discoveryofspeciﬁcneurophysiologicalmarkersfordiffer- entstages of perceptualprocessing andresponsemaking.

While basic neurophysiological measures used in clinical assessment(suchvisual,somatosensory,auditory,ormotor evokedpotentials)aretypicallynormal,andmaythushelp toestablishthediagnosisofconversiondisorder,somealter- ationshavehoweverbeenreportedformoresophisticated measures associated with more complex task conditions or discriminative abilities (for more detailed review see [104,105]).Mostofthesefindingsresemblemodulationsalso knownto be governed by attention, motivation, habitua- tion,and/orexpectationfactors.Thus,inlinewithearlier neurobiological accounts of hysterical conversion, these alterationshave oftenbeen interpretedasreflectingmod- ulation (e.g. suppression) of afferent sensory or efferent motorprocessesundertheeffectofstressoremotionalsig- nals.However,greatvariabilityexistsbetweenstudiesand verylittleisknownaboutthespecificityofthesefindings.

Sincethe90s,theemergenceoffunctionalbrain-imaging techniques with higher anatomical resolution (including SPECT, PET and fMRI) has allowed more detailed investi- gationof theneural basesofseveralpsychiatricillnesses.

However,todate,relativelyfewstudieshavebeendevoted to the conversion disorders, in contrastto the wealth of studiesonotherconditionssuchasdepression,anxiety,or schizophrenia.Amongtheﬁrstpioneerimagingapproaches, Marshalletal.[56]performedPETscansinafemalepatient whohad hada paralysisof the leftlegfor 2years, which wasdiagnosedasaconversiondisorder. Brainactivitywas examinedwithPETduringthepreparationandexecution(or attemptedexecution) of movementswitheither theright or leftleg. Results showedthat, whereas motor areas of thelefthemispherewereactivatedduringmovementexe- cutedwiththerightleg(unaffected),movementsattempts made withthe leftleg (affected)produced no activity in the corresponding motor regions of the right hemisphere (in keeping withthe lack of movement), but alsoevoked an large increase in activity in more anteriorand medial prefrontalregions.Theseincreasesencompassedtheante- rior cingulate cortex (ACC), as well as the ventromedial (VMPFC)andorbitofrontalcortex(OFC).Astheseareasare part of limbic prefrontal cortices, this frontal activation was interpreted as reﬂecting a recruitment of emotional or motivationalprocesses that produced an active inhibi- tionofmotorcorticalareasactivationandhenceparalysis.

This interpretation thus accorded with previous accounts of conversion deficits in terms of active suppression of sensorimotor pathways undertheinfluence of other brain (e.g. subcortical) systems as proposed by Pavlov [67] or Ludwig [50]. In addition, the same authors subsequently reported a similarexperimental paradigm [37] conducted inasinglehealthyvolunteerinwhomleftlegparalysiswas inducedunderhypnosis.Again,movementattemptswiththe affectedlegproducednoincreaseinmotorcortex,butacti- vatedOFCandACC.However,sincetheseregionshavealso beenassociatedwithself-monitoring,errorprocessing,and conflictdetection[1,83],theircausalimplicationinactive inhibitionofmotoractivitycannotbefirmlydemonstratedin thisparadigm.Itisalsopossiblethatinstructingthepatient to move despite her paralysis induced conflict- or error- relatedactivityin medialprefrontal areas[76]duetothe taskratherthanduetoconversion.

In another early study using SPECT imaging [106], we avoidedtheseconfoundingfactorsbyprobingsensorimotor pathwaysinanindirectmanner,withoutexplicitinstructions to execute movements. Brain activity was measured in 7 patientswithaunilateralmotorconversiondeﬁcitinthree differentconditions:

• atrestwhiletheyhadmotorsymptoms;

• duringabilateralproprioceptivestimulationwithvibra- tion(expectedtoactivatethemotorpathways)whilethey hadmotorsymptoms;

• during the same stimulation a few months later after recoveryfromsymptoms.

Proprioceptive stimulation activated sensorimotor cortical areas in both hemispheres symmetrically, showing intactfunctionsdespitemotorsymptoms,butaselectivea decreaseinactivitywasobservedinthecortico-subcortical motorloopsinvolvingthebasalganglia(putamenandcau- datenucleus) andthalamusinthehemisphereoppositeto motorsymptoms.Theseabnormalitiesinbasalgangliaand thalamus disappeared when scanning was repeated after

(6)

Please cite this article in press as: Vuilleumier P. Brain circuits implicated in psychogenic paraly- Figure1 Illustration oftheGo/NoGo paradigmusedinCojanetal.[16,18,19].This allowedtestingfor brainactivityduring failuretomoveonGotrialswiththeparalyzedhandandcomparisonwithinhibitorycontrolactivityduringNoGotrials,inaddition tomotorpreparation.

recoveryfrommotorsymptoms.Further,afunctional network connectivity analysisbasedon principalcomponents suggestedthatchangesinbasalgangliaandthalamuswere coupledwiththeactivity levelin ventralandmedial prefrontal areas, speciﬁcally in scans obtained during motor symptoms,notafterrecovery.Conversely,hyperactivation in basal gangliawasfound in psychogenic dystonicmove- ments[86],without increasesin motor cortex, unlikethe pattern found in organic dystonia. Volumetric changes of thalamic nuclei have also been reported in patients with chronicmotorconversiondisorders[5,59].

The basal ganglia-thalamic loops arecrucial for motor execution and typically implicated in hypokinesia in Parkinson’sdisease. Decreased activity in the hemisphere contralateral tomotor conversion paralysis and increased activityinconversiondystoniamightreﬂectadysregulation ofmotorexecution processes,downstreamtomotorplan- ningincorticalareas,perhapsundertheinﬂuenceoflimbic prefrontal areasthat wereshown tobehyperactivatedin Marshalletal.’sstudy[56]orhypercoupledwithbasalgan- gliainourownstudy[106].Indeed,thelimbicventromedial prefrontalareasalsostronglyprojecttothebasalganglia.

Thelatterthusconstituteauniquecrossroadsinthebrain where motorsignalsconverge withmotivationalsignalsin ordertobeintegratedanddeterminebehavioralmotorout- puts. Interestingly, modulation of motor activity in basal gangliacircuitsisalsoknowntoberesponsiblefor stereo- typedmotorbehaviorsassociatedwithparticularaffective statesorimmobilityafterlimbinjury[23].Accordingtothis view,conversionsymptomsmightresultfromthegeneration ofparticularmotorpatterns(innateoracquired)ratherthan fromactiveinhibitionofmotorpathways.

A few other fMRIstudies however have found reduced activityinmotorcortexitselfduringconversionparalysisor weakness.Suchdecreaseswerereportedduringmovement attemptswiththeaffectedlimb[97],observationofhand movements in videos [11], or exposure to personal emo- tionalscenariospresumedtobeassociatedwithconversion

onset[45].Inthesestudies,additionalanomalieswerealso observedinotherbrainregions,butveryvariable,including decreasesinsupplementary motorareas andorbitofrontal corticesbut increasesin putamen and inferior prefrontal cortex[97],aswellasincreasesinhippocampusandamyg- dala[45]. These studies however included small samples ofpatients (n=1 to4)with heterogeneoussymptomsand psychiatriccomorbidities. Suchreduction in motor cortex activityduringmotorparalysishasgenerallybeenimputed toimpairedmotorreadinessandintentionalprogramming, in keeping withthe classic proposal of Charcot that psy- chogenicweaknessmightresultfromalackofactivationof the‘‘mentalimageofmotormovements’’.

In a more recent fMRI investigation, our group aimed at disentangling the role of impaired motor intention or active motor inhibition in conversion paralysis. To do so, wedesigneda modiﬁedGo/NoGo paradigmallowing usto probebrainactivityrelatedtovariousmotorcontrolstages includingmotor intention, execution, and inhibition. Two patients withleft motor conversion were examined using slightly different versions of the same paradigm [18,49], aswellasagroupof healthycontrols[18,19].In thistask (Fig.1),participantswereﬁrstpresentedwithavisualcue instructing them toprepare a movement with either the rightorlefthand(greyhandpicture),followedbyasecond cueinstructing toeither execute (greenhand, 75%of trials=Go)orcancelthepreparedmovement(redhand,25%

of trials=NoGo). Remarkably, motor preparation withthe left(paralyzed)handorright(intact)handproducednormal contralateralandsymmetricactivationsincontralateralpri- marymotorcortex(areaM1)forbothconversionpatients, justlikeinhealthyparticipants(Fig.2).Thisnormalpattern ofpreparatory motor activity suggestsa preserved ability togeneratemotorintentionsandimagery,despitethepsy- chogenicparalysis.As expected,however, alossof motor activationwasseenintherightmotorcortexatthetimeof theinstructedmovementexecutionwiththeleft/paralyzed hand (in accordance with the lack of movement, but in

(7)

Please cite this article in press as: Vuilleumier P. Brain circuits implicated in psychogenic paraly- Figure2 BrainactivityindifferentstagesoftheGo/NoGoparadigmusedinCojanetal.[18,19].A.Resultsinhealthycontrols, showingactivationincontralateralprimarymotorcortex(M1)duringpreparationandexecution(Go),andactivationinrightinferior frontalgyrus(IFG)duringinhibition(NoGo)forbothhands.B.ActivityinrightIFGduringnormalconditionandduringvoluntary simulationofleftparalysis,showingthatsimulationwasassociatedwithincreasesinrightIFGandproducedbyvoluntaryinhibition.

C.Resultsinapatientwithleftconversionparalysis,showingnormalactivationofcontralateralM1duringpreparationbutlack ofactivationofrightM1duringparalysis,togetherwithincreasesinprecuneusandventromedialprefrontalcortex(VMPFC).D.In thepatient,leftparalysiswasnotassociatedwithincreasedactivityinrightIFGasseenduringvoluntaryinhibitionofsimulation;

whereasactivationinVMPFCwasselectiveforpreparingorattemptingtomovetheleft/paralyzedhand.

AdaptedfromCojanetal.,[18,19].

contrast to left motor cortex activation during correctly executedrighthandmovement).Thus,intactmotorprepa- rationevidencedbycovertcorticalactivityinM1couldnot be transformed into actual movement in the conversion patients.Intact motor intentionat the corticallevel with impairedexecutionhasalsobeenassociatedwithdeﬁcient

‘‘energizing’’and‘‘patterning’’ofthemotorcommanddue tobasalgangliadysfunctioninParkinsondisease[68,81,99].

This dissociationbetween preparation andexecution may echothesubjectiveexperienceofconversionpatientswho typicallycomplainofaninabilitytomovedespitetheirwill.

OurGo/NoGoparadigmalsoenabledustodirectly test foraroleofactiveinhibitionandcomparisonofconversion disorder with voluntary simulation of paralysis. On NoGo trials, when a prepared movement had to be cancelled, healthyparticipantsshowedrobustincreasesinanetwork knowntobeassociatedwithmotorinhibitionandcognitive control,includingrightinferiorfrontalgyrus(rIFG)andright inferior parietal lobule (rIPL) [3,19,111]. Moreover, this rightfronto-parietalnetworkis similarlyrecruitedfor left orrighthandmotorinhibition[19].Patientswithleftmotor conversionparalysisshowedthesamepatternofincreases inrightinferiorfrontalandparietalcorticesonNoGotrials but,morecritically,nosuchactivationontheleftGotrials wheretheyfailedtomovetheleft/paralyzedhand[18,49].

Thissuggeststhatdespitepreservedmotorpreparation,the

lackofmotorexecutionwiththeirlefthandwasnotdueto activevoluntary inhibition.By contrast,whenwe asked a groupofhealthyvolunteerstosimulatealefthandparalysis (‘‘as iftheywerepatients withstroke whoeffortfullybut unsuccessfully tried to move their paralyzed limb’’), the rightinferiorfrontalgyrusshowedsigniﬁcantactivationon the leftgo trialswhen theymimicked paralysis (Fig.2B).

Nevertheless, motor cortex was symmetrically activated during covert motor preparation with either hand. These ﬁndingsindicatenotonlythatsimulatorsfollowedourtask instructionscorrectly,butalsothattheyactivelyinhibited thepreparedmotorcommandonleftGotrialsandtreated the latter condition like NoGo trials. This differs from conversion patients, in whomnoactive inhibition activity wasseeninprefrontal orparietal cortices,supporting the idea that their paralysis is not simulated voluntarily but rather results from a distinct kind of disturbance in the consciouscontrolofvoluntarymovements.

Remarkably,however,auniquepatternofactivationwas observed in our two conversion patients in anterior and posteriorbrainmidlinecorticalareas,theVMPFCandpre- cuneus[18,49].Thesetwoareasdidnotactivateinhealthy controls[19]norinsimulators[18]duringthesameGo/NoGo task,butwerespeciﬁcallyrecruitedduringlefthand conditions in conversion patients, particularly during the motorpreparationorexecutionphasesfortheVMPFC,and

(8)

Please cite this article in press as: Vuilleumier P. Brain circuits implicated in psychogenic paraly- Figure3 Brainactivitypatternsindifferentconversiondisorders.A.Inpatients withmotorloss(n=7),primarymotorcortex activatednormallyduringmentalimageryofhandrotationforboththeintactandimpairedsidebutwasaccompaniedbyincreased activityinventromedialprefrontalcortex(VMPFC)duringimplicitimageryforthelefthandonly.FromdeLangeetal.,[25].B.In anothergroupofpatientswithmotorloss(n=3),VMPFCwasactivatedduringmovementinitiationduringavisuomotormonitoring task(theright-sidedpanelshowstheoverlapforeachofpatientdepictedinadifferentcolor:red,green,yellow).FromSinanaj etal.,unpublished.C.Inpatientswithconversionanesthesia(n=4),painfulsensorystimulationproducedbothreducedactivation (blue-greencolors)insomatosensorycortexandincreasedactivation(red-yellowcolors)inVMPFCandanteriorcingulatecortex, relativetohealthysubjects.D.Inasinglepatientwithtransientconversionblindness,visualstimulievokedweakeractivationin occipitalareasassociatedwithgreateractivationandfunctionalcouplingofVMPFC.FromBeckeretal.,[9].

duringthemotorexecution(Go)orinhibition(NoGo)phases for the precuneus (Fig. 2C, D). Additional increase in a morelateralportionoftheright ventralprefrontalcortex wasalsoobservedduringattemptedmovementsonleftGo trialsinonepatient[18],butnotinthesecond[49].Both VMPFC and precuneus are implicated in representations of the self-including [63] autobiographical memory [82]

and self-reﬂexive imagery [14]. VMPFC is particularly recruited in tasks requiring access to affectively rele- vant information about the self,including personal traits judgments[21,43,84], introspectiononone’s ownfeelings [44,46,52,69]and retrievalofemotional informationfrom either past or prospective memory [20,91]. This region is also activated in response to incidental information related to the self without explicit task demands or conscious will [58]. On the other hand, the precuneus is more particularly recruited during autobiographical memory retrieval tasks involving self-centered perspective and sensory imagery [14], as well as judgment of self-agency in actions [27] or free motor choices [87].

Finally, more lateral regions of right ventral prefrontal cortex have been linked with motor inhibition processes promoted by threat signals [79], consistent with a more generalroleininhibitorycontrolandavoidanceinnegative emotional contexts [32,40,41]. Taken together, our ﬁnd- ings suggest that neural processes intimately associated

with access to self-related representations, as mediated by VMPFC and precuneus, may play an important role in disturbancesof movement control in patients withmotor conversion, together with other cortical limbic regions implicatedinemotionregulationandmotivatedbehavior.

Theseresults convergewiththoseofother recentfMRI studiesthatinvestigatedmotorimageryandself-monitoring in patients with unilateral motor conversion deﬁcits. de Lange etal. [24,25] examined a group of 7 patients dur- inganexplicitmotorimagerytask,wheretheywereasked toimagineactionswiththeirrightor lefthand,aswellas during an implicit motor task, where they were asked to make rotation judgments for right or left hand pictures.

Bothtasksareknowntoinducecovertmentalsimulationof handmovementsandactivateprimarymotorcorticalareas inhealthypeople[66].Inthesestudies, motorconversion patients showed a normal behavioral pattern in response timesandnormalactivationinmotorcorticesregardlessof handside[24,25],reminiscentofnormalactivationduring motorpreparationinourGo/NoGotask[18,49].Inaddition, deLangeetal. [24,25]alsofound a distinctiveactivation inVMPFCinconversionpatientsduringimageryconditions involvingtheaffectedhandside,althoughthiswasobserved onlyintheimplicittask(Fig.3A).ThisactivationofVMPFC wasattributedtoagreaterrecruitmentofself-monitoring processeswhenusingmotorprocessesassociatedwiththe

(9)

Please cite this article in press as: Vuilleumier P. Brain circuits implicated in psychogenic paraly- affectedlimbinanautomaticmanner,whilethedifference

betweenhandsdisappearedwhenpatientshadtomonitor their hand position in a more controlled/conscious manner during the explicit imagery condition. In accordance with these findings, in a recent (still ongoing) study, we alsoobservedincreasedactivationinVMPFCactivationdur- inga visuomotor taskin which conversion patients (n=3) had tomove a cursor on a computer screen using a joy- stickandmonitorthecursor’strajectoryinordertoreacha pre-specifiedvisualtargetonthescreen(Fig.3B).Thiswas observedeventhoughthesepatientsusedtheirnon-affected handandhadnosignificantparesisatthetimeoftesting.No suchactivationwasseeninVMPFCduringthistaskinacon- trolgroupof healthysubjects.Thus,VMPFC maynotonly activatefor theaffected handor during implicitmonitor- ingconditions,butappearstobemoregenerallyrecruited inmotor tasks in patients withmotor conversiondeficits.

Moreresearchisneededhowevertounderstandtheexact roleofthisregioninconversiondisorders.Actionmonitoring invariousmotororcognitivetasksisknowntorecruitmore dorsalregionsinACCandpre-SMA[1,83],ratherthanventral ACCandVMPFC.Hence, theroleofVMPFCaswellaspre- cuneusmightreﬂectamodulationofmotorbehaviordriven byaffect,imagery,and/ormemoryrelatedrepresentations, ratherthanjustresponsemonitoring.

Dissociation and hypnosis

Since the time of Charcot and Janet, conversion deﬁcits haveoftenbeenconsideredtoresemblesymptomsproduced byhypnoticsuggestions.Accordingly,conversionsymptoms may be changed or abolished under the effect of coun- teractinghypnoticsuggestions [48,57].Moreover,hypnotic suggestibilitytendstobehigherinindividualswithahistory ofconversionsymptoms[77].Ontheotherhand,patholog- icaldissociative states (e.g. post-trauma) have also been linkedwithchanges in activity in medialprefrontal areas and precuneus [22,31], and modulations of neural activ- ityinprecuneushaveoften been reportedunderhypnosis [26,28,51,70,72]. In further support of a functional link betweendeﬁcitsproducedbyconversionandthoseinduced byhypnosis,afewneuroimagingstudieshavereportedthat motorparalysisunder hypnosisisassociatedwithmodula- tionofbrainactivityinACCandventralprefrontalregions [37,108].Thus, somerecent accounts have proposedthat conversionsymptomsmightbeproducedbysimilarcognitive mechanismsas hypnoticsuggestion, implicating an active inhibitionoftheaccessofsensoryormotorrepresentation toconsciousawareness[10,64].

To clarify the putativelinks between these conditions, wealsoinvestigatedtheeffectofahypnoticallysuggested paralysisonbrainactivityduringthesameGo/NoGomotor task as described above (Fig. 1). As in healthy people andconversion patients,wefoundpreserved activationof contralateralprimarymotorcortexduringthemotorprepa- ration phase, for both the paralyzed and intact hands.

Critically,however,we found selectiveincrease in activa- tionintheprecuneusduringcovertmotorpreparationwith the paralyzedhand, suggesting intactimagery and inten- tionprocessesunderhypnosis[19].NomodulationofVMPFC or VLPFC was observed, unlike conversion. Furthermore,

the right fronto-parietalnetworks associated withexecu- tivecontrolandinhibitionswere globallyenhancedacross allconditions,presumablyreflectingenhancedattentional controlandfilteringunderhypnosis.Inparticulartheright inferiorfrontalgyrus(rIFG),typicallyrecruitedduringtasks requiring high attentional controland during inhibition of prepotentresponses[3,39,111],wasstronglyactivateddur- ing hypnotic paralysis, but equally so for both the NoGo trialsandGotrialsandforbothrightandlefthands.Similar increasesin rightIFG activitywerefound duringlefthyp- noticparalysisusingEEGrecording[16]inanothergroupof volunteersperformingthesameGo/NoGoparadigminnor- malandhypnoticstates.Thesefindingsaccordwiththeview thathypnoticsuggestionsactthroughmodulationofatten- tionalsystemsthatmayallowselectivefilteringofsensory, motor, orevenconceptual representations fromconscious awareness[18,38,64,73].

Takentogether,theseresultsthereforeonlypartlysup- portasimilaritybetweenconversionandhypnosis.Whereas recruitment of internal representations mediated by the precuneus appears to be a shared feature in both cases, perhaps reﬂecting accesstosensoryimagery andmemory information, more anterior brain regions in ventromedial and dorsolateral prefrontal systems appear differentially implicatedinconversionandhypnosis,respectively.

Modulation of brain network connectivity during motor conversion

Early imaging studies in conversion paralysis suggested that changes in activity of cortical motor areas [56] or subcortical-basal ganglia loops [106] were causally driven bymodulatoryinputsfromprefrontalareaspartlyoverlap- pingwithVMPFCandVLPFCthatwerefoundtobeenhanced in ourGo/NoGo paradigm[18]. Inthe latter,more recent fMRIstudy,wethereforealsotestedforanydifferencesin functional connectivityof the right andleft motorcortex depending on the nature of motor dysfunction. By using seeds in primary motor areas (M1) and computing temporal dynamiccorrelations overtimeduring theGo/NoGo task,wefoundthatinnormalconditions(healthycontrols), activity in M1 was selectively coupled with a network of regions implicated in motorcontrol includingdorsolateral premotor cortex(PMC),medialsupplementary motorarea (SMA),andposteriorparietallobule(SPL).Inpatientswith motorconversion,thesameconnectivitypatternwasfound for M1 contralateraltothe intacthand, whereasM1 con- tralateraltotheaffected/paralyzedhandshowedreduced functionalconnectivity withPMCbut increasedconnectiv- ity withVMPFC (Fig. 4). This suggests that motor control issomehowundertheinﬂuenceofinternalrepresentations relatedto self-relevantemotional informationor memory inpatientswithsymptomconversion,ratherthanunderthe inﬂuence of premotor systems normallyimplicated in the generationandsupervisionofvoluntarymovements.

Remarkably,adistinct patternof changesin functional connectivity was observed during hypnotic paralysis. M1 contralateraltotheparalyzedlimb wasdistinctivelymore coupledwiththeprecuneus,inposteriorandmedialparietal cortices,butlesscoupledwithpremotorareasinprefrontal cortices[19].Thissuggeststhathypnosismightactonthe

(10)

Please cite this article in press as: Vuilleumier P. Brain circuits implicated in psychogenic paraly- Figure 4 Functionalconnectivity of primary motor cortex.

Whilebothrightandleft primarymotor cortexaregenerally coupledwithipsilateralpremotorareas,leftconversionparal- ysisisassociatedwithaselectiveincreaseofconnectivitywith bothprecuneusandventromedialprefrontalcortex.Thisisnot seenduringvoluntarysimulationofleftparalysis.

FromCojanetal.,[18,19].

motorsystempredominantlythroughaninﬂuenceofinter- nal representations thatarerelatedtomnemonic sources orsensoryimagery,ratherthanlinkedtoemotionalmemory asinconversion.Notably,functionalconnectivityofmotor cortexduringsimulationofparalysisdidnotdifferfromthe normalsituationinhealthycontrols,conﬁrmingthatinthis caseparalysiswasproducedbyvoluntary(inhibitory)control ofmotorcommands[18].

As noted above, both the VMPFC and precuneus are crucially implicated in self-reﬂective processingand con- stitute core components of the so-called default mode network(DMN),whichistypicallymoreactiveatrestthan duringactivecognitive tasks[36,71].The DMN alsoshows ongoingcoherentactivityacrossdifferentbehavioralstates, evenduringconcomitanttaskperformance[100],and can be distinguished from task-related activity using various network connectivity measures [54]. We have explored the spatial and temporal organization of such intrinsic brain networksduring the Go/NoGo task described above using an independent component analysis (ICA) approach (for details see [17]). This analysisallowed us toidentify severaldistinctnetworkswithcoherentactivityduringthe task,includingtheDMN.Whenmeasuringthecontribution of different brain regions to each of these networks in patients with left motor conversion, relative to healthy controls andsimulators, wefound a selectivedecreasein contributiontotheDMN for theVMPFC/anteriorcingulate

and precuneus/posterior cingulate cortex (Fig. 5) during conversion only. Thus, in the patients, these two areas werelessengagedindefaultmodeactivityandpresumably moreengagedinadifferentnetwork.This convergeswith ourseed-basedfunctional connectivityanalysis(described above)suggestingthatVMPFCandprecuneusweredynam- ically coupled with motor cortex during conversion. In addition,ourICAanalysisalsoidentiﬁed amotor network includingprimarymotorcortex(M1),supplementarymotor area(SMA),andportionsofthebasalganglia.Interestingly, rightM1wasfoundtocontributelesstothisnetworkduring both conversion and simulation paralysis, whereas the contributionofSMAwasreducedduringsimulationonly.In addition,changesin connectivity were alsofound for the caudatenucleusinbothhemispheres,whichappearedless expressedinboththeDMNandthemotornetworksduring conversion as well as simulation, relative to its normal connectivity state (Fig. 5). The latter ﬁnding suggests a functional ‘‘disconnection’’of the caudate from the rest oftheDMNandthemotorcortexinbothtypesofparalysis.

These connectivity data converge with our previous results obtained with more conventional imaging analy- sesduringeither activemotorperformance [16—18,49]or passive proprioceptive stimulation [106], indicating that corticaland subcortical motor pathways may exhibitpar- ticular functional interactions with midline brain areas regulatingself-relatedandaffectiveprocessesduringmotor conversion disorders.Similar evidence has been gathered inrecent studies[101—103] concerningpatients withpsy- chogenic movement disorders such astremor or dystonia (i.e.,positive rathernegativemotorsymptoms), although pointing to the role of different circuits connecting the amygdala to SMA (see also review by Voon et al. in this issue).Futureresearchshouldusemoreadvancedconnec- tivitymeasuressuchasdynamiccausalmodelling(DCM)and comparedifferentconversionpatientgroups duringdiffer- entsymptomaticconditions(pre-vs.post-recovery)togain betterinsightsintosuchinteractionsbetweenmotorcontrol andemotion/motivationsystems.

Neuroimaging ﬁndings in non-motor conversion disorders

A few brain-imaging studies have examined conversion deﬁcits affecting other neurological functions such as somatosensoryloss[35,53],visualloss[9,85,109],ormem- ory loss [33,55,98]. In most cases, reduced activity was observedinbrainareasinvolvedintheimpairedfunctions, includingsomatosensorycortices,occipitallobe,ortempo- rallobe,respectively(forreviewsee[104,105]).Additional changes (hyper- or hypo-activation) were also generally foundinvariousotherregionssuchasthalamus,basalganglia, insula, anterior cingulate or prefrontal cortex, but withoutaclearsystematicpattern.Leftspatialneglecthas alsobeenobservedduringconversionsymptoms,withsym- metricparietalactivationbutincreasedrecruitmentofACC duringlinebisectionjudgments[80].

Of particular note, a recent fMRI study in a single patientwithtransientepisodeofpsychogenicblindnessand hallucinationsreportedpreservedresponsesinoccipitalcor- tex to simple visual stimulation but markedly decreased

(11)

Please cite this article in press as: Vuilleumier P. Brain circuits implicated in psychogenic paraly- Figure5 Functionalconnectivityusingindependentcomponentanalysis(ICA)duringaGo/NoGotask.Ventromedialprefrontal cortexshowedaselectivereductioninitsfunctionalconnectivity(i.e.temporalcorrelation)withanetworkofmidlinebrainareas correspondingtothedefaultmodenetwork;thesupplementarymotorareashowedaselectivereductionofconnectivitywitha networkofmotorandpremotorareasduringsimulation.

DatafromCojanetal.,[18,19],adaptedfromCojanandVuilleumier[17].

responsestomorecomplexpictures(faces),togetherwith selective increases in VMPFC (Fig. 3D), which occurred onlyduringblindnessepisodesandnot afterrecovery[9].

Functional connectivity analysis also revealed increased couplingbetween VPMFCandoccipital areasduringblind- ness episodes. Another fMRI study [109] conducted in 7 patientswithpsychogenicblindnessaffectingvariousparts ofthevisualfield(monocularor binocular)foundreduced activation of bilateral occipital areas to full field visual stimulation with checkerboards, again accompanied with increasedactivation inrightthalamus,striatum,andright VLPFC/lateral OFC. This pattern of sensorydecrease and fronto-limbicincreaseechoesfindingsin4patientssuffer- ingfrom unilateral anesthesiaand psychogenic pain[53], in whom somatosensory areas showed reduced response tonoxious stimuliwhile VMPCand ACC showed increased activation inthe same condition (Fig.3C), unlikehealthy participants.

Altogether, even though the existing results are still scarce and partly inconsistent across studies and conversion domains, recent neuroimaging data for non-motor deﬁcits show some convergence with those obtained in patients with motor conversion paralysis in several studies[18,24,25,56,106].Thus,anemergingpatternofﬁndings acrossdomainspointstonotonlyabnormalrecruitmentof brainareas involved in conscious control of movement or conscioussensory perception, butalso heightened activation of both cortical (VMPFC, precuneus) and subcortical

(amygdala, thalamus, striatum) areas thought to be normally involved in emotional regulation of behavior and memory.Itremainshowevertodeterminetheexactmecha- nismsthroughwhichthesesystemsmayinteract,aswellas theexactfactorsthatareresponsibleforthesealterations.

Conclusions

Overthelast15years,agrowingnumberoffunctionalbrain- imagingstudieshave investigatedtheneuralsubstrates of conversion symptoms, a clinical condition that has been recognized and much reﬂected upon for many centuries, and which remains puzzling today. These imaging results have begun to shed new light on the possible neurobiological underpinnings of conversion, although our current knowledgeisstillfarfrombeingabletoprovideacoherent framework.In particular,thereis nowclear neuroimaging evidence for changes in brain activity, affecting speciﬁc regionsinaccordancewiththenatureofthesymptomsand thetime-courseoftheirrecovery.Thus,morethan100years afterCharcotandFreud,itisnowpossibletodemonstrate thatconversionparalysisisassociatedwitha‘‘functionalor dynamic lesion’’in brain pathways, without organicdam- age, as thefounders of modernneurology and psychiatry had indeed postulated. Importantly, brain-imaging results alsoindicatethatfunctionalchangesmaynotonlyconcern thosebrainregionsdirectly associatedwiththesymptoms

(12)

Please cite this article in press as: Vuilleumier P. Brain circuits implicated in psychogenic paraly- (e.g. motor areas in paralysis, visual areas in blindness)

butalsoseveralothersinfrontal,parietal,andsubcortical areas, whose exact role in conversion still remains to be fullydetermined.Furthermore,byusingspeciﬁcexperimen- talparadigmsandvariousmethodologicalapproaches,ithas alsobecomepossibletotestmorespeciﬁchypothesesabout underlyingneuralandpsychodynamicmechanismsthathave been postulated in the past — such as the role of active inhibition,impairedaccesstoconsciousawarenessordisso- ciationfromconsciouswill,linkswithemotiondysregulation or traumatic memories,similarities withhypnosis,and so forth.

For motor conversion, work by our group and others suggest that motor imagery and motor intention are preserved and continue to normally activate primary motor cortex;whereasmovementexecutionisimpaired,presumably due to a suppressionor modulation of motor output pathwaysincortical-basalganglialoopsundertheinﬂuence of other brain regions (e.g. limbic prefrontal areas). Our resultsprovidenoevidenceforarecruitmentofinhibitory controlsystems,atleastasmediatedbypremotororlateral prefrontalcortices,suchastherightIFG[18,111].Further- more,theneuralchangesobservedinconversiondifferfrom those associated withsimulation, where voluntary inhibi- tionand/orattentionalmechanismsinsteadseemtoplaya role.Ourresultsalsorevealonlyapartialsimilaritybetween motor conversion disorders and hypnotic paralysis rather thanacompleteoverlap.

Taken together, these data lead to a new tentative functional neuroanatomy model of psychogenic paralysis [17,105,107]. On the one hand, in line with the classic viewsoriginatingfromCharcot,Janet,Babinskiandothers attheturnofthe20thcentury,brainactivityandconnec- tivity patternssuggest that both conversion and hypnosis might entailan apparent dominanceof internal represen- tationsfromimageryandmemoryonthecontrolofmotor behavior (Fig.6).In both conditions,the motorpathways (e.g. M1)appear functionally less coupled withpremotor corticalareasbutmorecoupledwithmidlinebrainregions typicallyassociatedwithself-related processingandmen- tal imagery, such as the precuneus. On the other hand, however, unlikethe classic view, conversion and hypnosis

appearto act on behavior and perception via partly distinct brain circuits. In conversion, modulations of motor and sensory awareness might primarily be driven by an increasedrecruitmentofregionsinVMPFC,implicatedinthe accesstoself-relevantaffectiverepresentationsandmem- ories[17,104,105,107],aswellasother subcorticallimbic regionssuchastheamygdala[45,79,101,102].Theseareas might inﬂuence movement, sensation, andself-awareness morebroadlybypromotingorsuppressingcertainpatterns ofbehaviorsthroughmodulatoryinputstomotororsensory pathways(asobservedformoreprimitiveadaptivereactions to threat, see [47,104,110]), and/or by imbuing behav- iorwithparticularaffective andself-relevantassociations though an integrationof sensorimotor activity withinter- nalrepresentationsretrievedfrommemoryorimagery.This modelechoes the long-established notion that conversion symptomsgenerallyariseundertheinﬂuenceofemotional factorsor pastemotionaltrauma,realor imagined[7,42], andtheFreudianviewthat‘‘hystericsprimarilysufferfrom reminiscences’’ [30], although without necessarily invok- ingtheroleofrepression,sexualorigin,andsymbolization aspostulatedby Freud.Conversely, underhypnosis,alter- ationsin conscious perception and movement might arise from interactionbetween imagery and sensorimotor pro- cessthroughamodulationimposedbyexecutiveattentional mechanismsmediated by lateral prefrontal cortical areas (e.g.rightIFG)andperhapsACC[16,19,73,74].

Nonetheless,moreresearchisneededtogobeyondthese general descriptivemodels focusing on the generation of speciﬁcsymptoms,inordertobetterunderstandtheexact mechanistic interplay between different brain systems involvedinconversion(andhypnosis),inordertoelucidate the causal factors responsible for triggering conversion episodesand associatedbrainpatterns, and toclarify the relationshipswithotherpsychiatricdiseasesreﬂectingmal- adaptive emotional adjustment andfrequently associated withconversion(such asdepression, anxiety,or posttrau- maticstressdisorder).Futureresearchshouldalsoestablish whethersimilarmechanismsunderlieconversiondisorders indifferentdomains(e.g.motor,visual,etc.),andwhether different types of conversion exist in different patients.

Thisknowledgewillbeinvaluableforimprovingtheclinical

Figure6 Schematicneuroanatomicalmodelofpsychogenicparalysis.Bothprecuneusandventromedialprefrontalcortex(VMPFC) arecorecomponentsofbrainnetworksmediatingaccesstoself-representationsandpersonalepisodicmemories,includingsensoryor agencyinformation(forprecuneus)andaffectiverelevance(forVMPFC).Frequentrecruitmentoftheseareasinconversionpatients andenhancedconnectivitywithmotororsensory areassuggestsanimportantrole inmodulatingbehavior andself-awareness, perhapsbymodulatingmotororsensoryprocesses(atcorticaland/orsubcorticallevel)undertheinﬂuenceofmemoryandaffective representationsunconsciouslygeneratedintheseregions.VMPFCactivityappearsspeciﬁctoconversion,whereasprecuneusisalso recruitedduringhypnosis.

(13)

Please cite this article in press as: Vuilleumier P. Brain circuits implicated in psychogenic paraly- diagnosis and management of these patients, for better

deﬁningtheirprognosis,butalsomoregenerallyforillumi- natingthe intricaterelationships betweenmind andbody andtheirpathologicaldistortioninconversiondisorders.

Disclosure of interest

Theauthordeclaresthathehasnoconﬂictsofinterestcon- cerningthisarticle.

Acknowledgments

This work has been supported by grants from the Swiss NationalScience Foundation(No127560 and 143764),the CogitoFoundation,andtheGenevaAcademicSociety(Fore- maneFund).TheauthorthanksYannCojan,LakshmiWeber, Indrit Sinanaj, Alain Forster, Nicole Cheseaux, and Selma Aybeckforcontributionstovariousprojectsandfruitfuldis- cussions.

References

[1]Alexander WH, Brown JW. Medial prefrontal cortex as an action-outcomepredictor.NatNeurosci2011;14(10):1338—44 [PubMedPMID:21926982.PubmedCentralPMCID:3183374].

[2]American PsychiatricAssociation.Diagnosticand Statistical ManualofMentalDisorders,DSMV.Washington,DC:American PsychiatricAssociation;2013.

[3]Aron AR, Fletcher PC, Bullmore ET, Sahakian BJ, Robbins TW.Stop-signalinhibitiondisruptedbydamagetorightinfe- riorfrontalgyrusinhumans.NatNeurosci2003;6(2):115—6 [PubMedPMID:12536210].

[4]StoneJ,LaFranceJrWC,BrownR,SpiegelD,LevensonJL, SharpeM.Conversiondisorder:currentproblemsandpoten- tialsolutionsforDSM-5.JPsychosomRes2011;71(6):369—76 [PubMedPMID:22118377].

[5]Atmaca M, Aydin A, Tezcan E, Poyraz AK, Kara B. Vol- umetric investigation of brain regions in patients with conversiondisorder.ProgNeuropsychopharmacolBiolPsychi- atry2006;30(4):708—13[PubMedPMID:16600450].

[6]AybekS,Hubschmid M,VuilleumierP,BurkhardPR,Berney A, Vingerhoets FJ. Hysteria:an historicalentity, a psychiatric condition or a neurological disease? Rev Med Suisse 2008;4(156):1151—6[PubMedPMID:18630169].

[7]BabinskiJ.Démembrementdel’hystérietraditionnelle:pithi- atisme.Paris:SemaineMédicale;1988.

[8]BallmaierM,SchmidtR.Conversiondisorderrevisited.Funct Neurol2005;20(3):105—13[PubMedPMID:16324233].

[9]BeckerB,ScheeleD,MoessnerR,MaierW,HurlemannR.Deci- phering theneuralsignature ofconversionblindness.AmJ Psychiatry2013;170(1):121—2[PubMedPMID:23288388].

[10]BellV,OakleyDA,HalliganPW,DeeleyQ.Dissociationinhys- teria and hypnosis:evidence from cognitive neuroscience.

J Neurol Neurosurg Psychiatry 2011;82(3):332—9 [PubMed PMID:20884677].

[11]BurgmerM,KonradC,JansenA, KugelH,SommerJ,Hein- del W, et al. Abnormal brainactivation during movement observationinpatientswithconversionparalysis.Neuroimage 2006;29(4):1336—43[PubMedPMID:16213162].

[12]CarriveP.Theperiaqueductalgrayanddefensive behavior:

functionalrepresentationandneuronalorganization.Behav BrainRes1993;58(1—2):27—47[PubMedPMID:8136048].

[13]Carson AJ, Brown R, David AS, Duncan R, Edwards MJ, Goldstein LH, et al. Functional (conversion) neurological

symptoms:researchsincethemillennium.JNeurolNeurosurg Psychiatry2012;83(8):842—50[PubMedPMID:22661497].

[14]CavannaAE,TrimbleMR.Theprecuneus:areviewofitsfunc- tionalanatomyandbehaviouralcorrelates.Brain2006;129(Pt 3):564—83[PubMedPMID:16399806].

[15]CharcotJM.Lec¸onsduMardi àlaSalpêtrière(1887—1888).

CEPL1974ed.Paris:BureauduProgrèsMédical;1892.

[16]CojanY,ArchimiA,CheseauxN,WaberL,VuilleumierP.Time- course of motor inhibition during hypnotic paralysis: EEG topographicalandsourceanalysis.Cortex2013;49(2):423—36 [PubMedPMID:23211547].

[17]Cojan Y, Vuilleumier P. Functional brain-imaging of psy- chogenicparalysisduringconversionandhypnosis.In:Hallett M,CloningerC,FahnS,HalliganP,JankovicJ,LangA,etal., editors.Psychogenicmovementdisordersandotherconver- siondisorders.Cambridge:CambridgeUniversityPress;2011.

p.143—59.

[18]Cojan Y, Waber L, Carruzzo A, Vuilleumier P. Motor inhibition in hysterical conversion paralysis. Neuroimage 2009;47(3):1026—37[PubMedPMID:19450695].

[19]CojanY,WaberL,SchwartzS,RossierL,ForsterA,Vuilleu- mierP.Thebrainunderself-control:modulationofinhibitory andmonitoringcorticalnetworksduringhypnoticparalysis.

Neuron2009;62(6):862—75[PubMedPMID:19555654].

[20]D’ArgembeauA,FeyersD,MajerusS,ColletteF,VanderLin- denM,MaquetP,etal.Self-reﬂectionacrosstime:cortical midline structuresdifferentiate between present and past selves.SocCognAffectNeurosci2008;3(3):244—52[PubMed PMID:19015116].

[21]D’ArgembeauA,RubyP,ColletteF,DegueldreC,BalteauE, LuxenA,etal.Distinctregionsofthemedialprefrontalcortex areassociatedwithself-referentialprocessingandperspec- tive taking. J Cogn Neurosci 2007;19(6):935—44 [PubMed PMID:17536964].

[22]DalenbergC,LoewensteinR,SpiegelD,BrewinC,LaniusR, FrankelS,etal.Scientiﬁcstudyofthedissociativedisorders.

PsychotherPsychosom2007;76(6):400—1[Authorreply1—3.

PubMedPMID:17917478].

[23]De Ceballos ML, Baker M, Rose S, Jenner P, Marsden CD.

Do enkephalins in basal ganglia mediate a physiological motor rest mechanism? Movement Disorders 1986;1:223—

333.

[24]de Lange FP, Roelofs K, Toni I. Increased self-monitoring during imagined movements in conversion paralysis. Neu- ropsychologia2007;45(9):2051—8[PubMedPMID:17367826].

[25]deLangeFP,RoelofsK,ToniI.Motorimagery:awindowinto themechanismsandalterationsofthemotorsystem.Cortex 2008;44(5):494—506[PubMedPMID:18387583].

[26]Del Casale A, Ferracuti S, Rapinesi C, Serata D, Sani G, SavojaV,etal.Neurocognitionunderhypnosis:ﬁndingsfrom recentfunctionalneuroimagingstudies.IntJClinExpHypn 2012;60(3):286—317[PubMedPMID:22681327].

[27]FarrerC,FrithCD.Experiencingoneselfvs.anotherperson as being the cause of anaction: the neural correlates of theexperienceof agency.Neuroimage2002;15(3):596—603 [PubMedPMID:11848702].

[28]FaymonvilleME,BolyM,LaureysS.Functionalneuroanatomy of thehypnotic state. J PhysiolParis 2006;99(4—6):463—9 [PubMedPMID:16750615].

[29]Flor-HenryP,Fromm-AuchD,TapperM,SchopﬂocherD.Aneu- ropsychologicalstudyofthestablesyndromeofhysteria.Biol Psychiatry1981;16:601—16.

[30]FreudS,BreuerJ.StudienüberHysterie.LeipzigandVienna:

Deutike;1895.

[31]FrewenP,LaneRD,NeufeldRW,DensmoreM,StevensT,Lanius R.Neural correlatesoflevelsofemotionalawareness dur- ingtrauma script-imagery inposttraumaticstress disorder.

PsychosomMed2008;70(1):27—31[PubMedPMID:18158370].