• Aucun résultat trouvé

Cerebral responses and role of the prefrontal cortex in conditioned pain modulation: an fMRI study in healthy subjects

N/A
N/A
Protected

Academic year: 2021

Partager "Cerebral responses and role of the prefrontal cortex in conditioned pain modulation: an fMRI study in healthy subjects"

Copied!
12
0
0

Texte intégral

(1)

ContentslistsavailableatScienceDirect

Behavioural

Brain

Research

j o ur na l h o me p a g e :w w w . e l s e v i e r . c o m / l o c a t e / b b r

Research

report

Cerebral

responses

and

role

of

the

prefrontal

cortex

in

conditioned

pain

modulation:

an

fMRI

study

in

healthy

subjects

Volodymyr

B.

Bogdanov

a,b,∗

,

Alessandro

Viganò

b,c

,

Quentin

Noirhomme

d,e

,

Olena

V.

Bogdanova

f

,

Nathalie

Guy

g

,

Steven

Laureys

d,e

,

Perry

F.

Renshaw

f

,

Radhouane

Dallel

g,1

,

Christophe

Phillips

d,h,1

,

Jean

Schoenen

b,1

aINRA,NutritionetNeurobiologieIntégréeandBordeauxSegalenUniversity,UMR1286,146rueLéo-Saignat,BordeauxCedex,33076,France bHeadacheResearchUnit,DepartmentofNeurology,CHRCitadelle,UniversityofLiège,Liège,Belgium

cSAMILAL-Dept.Anatomy,Histology,ForensicMedicine,Orthopaedics-LaSapienza,UniversityofRome;Dept.ofNeurologyandPsychiatry-LaSapienza,

UniversityofRome

dCyclotronResearchCentre,UniversityofLiège,Liège,Belgium

eComaScienceGroup,DepartmentofNeurology,UniversityandUniversityHospitalofLiège,Liège,Belgium fUniversityofUtah,SaltLakeCity,USA

gClermontUniversité,Universitéd’Auvergne,BP10448,F-63000Clermont-Ferrand;Inserm,UMR1107,TrigeminalpainandMigraineF-63000

Clermont-Ferrand;CHU,Clermont-Ferrand

hDepartmentofelectricalengineeringandcomputerscience,UniversityofLiège,Liège,Belgium

h

i

g

h

l

i

g

h

t

s

•Themechanismsunderlyingconditionedpainmodulationaremultifaceted.

•Duringcold,application-specificcerebralactivationswerefoundinprecuneusandleftinsula. •Conditionedsuppressionofinsulatest-painresponsewasrelatedtohypoalgesia.

•Theinsularchangesarepredictedbyearlyprefrontalresponsetocoldconditioning. •Smallerearlyprefrontalresponsepredictsheterotopicnoxioushyperalgesia.

a

r

t

i

c

l

e

i

n

f

o

Articlehistory: Received27June2014

Receivedinrevisedform28October2014 Accepted18November2014

Availableonline25November2014 Keywords:

Brainimaging Prefrontalcortex

Conditionedpainmodulation

a

b

s

t

r

a

c

t

Themechanismsunderlyingconditionedpainmodulation(CPM)aremultifaceted.Wesearchedforalink betweenindividualdifferencesinprefrontalcortexactivityduringmulti-trialheterotopicnoxiouscold conditioningandmodulationofthecerebralresponsetophasicheatpain.

In24healthyfemalesubjects,weconditionedlaserheatstimulitothelefthandbyapplying alter-nativelyice-coldorlukewarmcompressestotherightfoot.Wecomparedpainratingswithcerebral fMRIBOLDresponses.WealsoanalyzedtherelationbetweenCPMandBOLDchangesproducedbythe heterotopiccoldconditioningitself,aswellastheimpactofanxietyandhabituationofcold-painratings. Specificcerebralactivationwasidentifiedinprecuneusandleftposteriorinsula/SII,respectively,during earlyandsustainedphasesofcoldapplication.Duringcoldconditioning,laserpaindecreased(n=7), increased(n=10)orstayedunchanged(n=7).Attheindividuallevel,thepsychophysicaleffectwas directlyproportionaltothecold-inducedmodulationofthelaser-inducedBOLDresponseinleftposterior insula/SII.ThelattercorrelatedwiththeBOLDresponserecorded80searlierduringtheinitial10-sphase ofcoldapplicationinanteriorcingulate,orbitofrontalandlateralprefrontalcortices.

Abbreviations: ACC,anteriorcingulatecortex;BOLD,bloodoxygenationlevel-dependentresponse;fMRI,functionalmagneticresonanceimaging;FWER,family-wise errorrating;NRS,numericratingscale;OFC,orbitofrontalcortex;PAG,periaqueductalgraymatter;PFC,prefrontalcortex;SMA,supplementarymotorarea;SIandSII, primaryandsecondarysomatosensorycortices;VOI,volumeofinterest.

∗ Correspondingauthorat:BordeauxSegalenUniversity,146rueLéo-Saignat,BordeauxCedex33076,France.Tel.:+33557571226;fax:+33557571227. E-mailaddress:vlabogd@yahoo.com(V.B.Bogdanov).

1 Lastauthorscontributedequallytothesupervisionofthestudy.

http://dx.doi.org/10.1016/j.bbr.2014.11.028

(2)

1. Introduction

Endogenouspaincontrolisacomplexmulti-stepprocess com-posedofseveralinteractingneuronal systemsand mechanisms: stimulation-inducedanalgesia[1],diffusenoxiousinhibitory con-trols[2],ascendingnociceptivecontrol[3]andplaceboanalgesia [4],distractionordisengagement[5,6],stress-inducedanalgesia [7],and“offset”analgesia[8,9].

Diffuse noxious inhibitory controls (DNIC) are a form of supraspinal endogenous analgesia where phasic test pain responses are attenuated by heterotopic noxious conditioning stimuli.

Theterm“conditionedpainmodulation”(CPM)wasproposed todefine thepsychophysicalparadigm totest DNIC [10,11]. In humans,theconditioningstimulusmodulatingtheresponsetothe teststimuluscanbepainfulornon-painfulandCPMcaninhibitor facilitatethisresponse[11].CPMcanbeassociatedwithchanges inneuronalactivityatmultiplesitesfromspinalcorddorsalhorn, medullarynucleusreticularisdorsalis,periaqueductalgraymatter [12–14]toprefrontalcortex[15,16]

UnderstandingthemechanismsinvolvedinCPMisimportant because it is impaired in several hyperalgesic conditions (see forareview[17]:tension-typeheadache[18],migraine[19,20], fibromyalgia[21,22],irritablebowelsyndrome[23],osteoarthritis [24]andchronicpost-surgicalpain[25].

Functionalneuroimagingstudieshaveshownthatactivationsof primarysomatosensory(SI),prefrontal(PFC)oranteriorcingulate cortices (ACC) are associated with conditioned pain modula-tion [15,16,26]. The PFC, for instance, is activated in healthy controls by heterotopic cold stimulation during rectal pain [23,27]. Sustained cold-induced responses in orbitofrontal cor-tex (OFC)are linked tocold-induced analgesia [15]. Functional couplingbetweenthesubgenualanteriorcingulatecortex(ACC) and PAG is related to stronger individual heterotopic noxious analgesia[16].

Themechanisms of endogenous analgesia are clearly multi-facetedanddespitethenumerousstudiesinhealthysubjectsand inpathologicalconditions,thepreciseunderlyingneuralprocesses arestilluncertain.Moreover,painisahighlysubjectiveexperience andcanbeinfluencedbydifferencesinindividualsusceptibilityas wellaspersonality.Theaimofourstudywasthereforetosearchfor alinkbetweenindividualdifferencesinactivityofcerebralregions ofinterestduringmulti-trialheterotopicnoxiousconditioningand modulationoftheresponsestotestpainstimuli.

Toassessbrainactivity,weanalyzedfMRIBOLDresponses pha-siclaserheat-inducedtestpainwithorwithoutsustainednoxious coldconditioningandtheircorrelationwithself-ratingsofpain lev-els.Wefocusedonthemodulationoftest-painresponsesinthe so-called“painmatrix”[28]andthelinksbetweentheindividual differencesinCPMandprecedingorconcurrentneuronalresponses tonoxiouscoldconditioninginprefrontalareas(ACC,lateralPFC andOFC).

Contrarytotheabove-mentionedstudies,weusedarepeated block paradigm of noxious cold conditioning alternating with lukewarm application, which allows identifyingmore precisely

responsesinPFCtonoxiouscoldstimulationandreducesnovelty effects.

We expected that continuous cold pain stimulation would reduceself-ratingsoflaserheat-inducedpainandassociated cere-bralresponsesinthepainmatrix.InthoseareasofPFCknownfrom previousstudiestobeinvolvedindescendinganalgesia(ACC, lat-eralPFCandOFC),wesearchedforcold-inducedactivation,butalso forpossibledeactivation,asreportedinACC[29]andPFCareas[30]. WehypothesizedthatindividualdifferencesinPFCresponsesto coldpainwouldcorrelatewithlevelsofCPM.

2. Materialandmethods 2.1. Subjects

Twenty-fourhealthyfemalevolunteers(meanage:31.3years, range:24–45years)tookpartinthemainexperimentandfour subjectsparticipatedinapilotstudy.

Noneofthesubjectswassufferingfrompain,neurologicalor psychiatricdisorders.Allsubjectssignedaninformedconsentform thatmentionedthegeneralobjective,i.e.tostudytheimpactofa coldstimulusonpainandphysiologicalbrainresponsesinducedby laserheatstimuli,andwerethusnaïveaboutourworking hypothe-sis.Theyreceivedafinancialcompensationfortravelexpensesand timespentinthelaboratory.TheEthicsCommitteeoftheUniversity ofLiègeapprovedthestudy.

2.2. Experimentalprotocol

Laserheatpulsesdeliveredtothedorsumofthelefthandserved aspainfulteststimuli,whereassustainednoxiouscoldappliedon therightfootwasusedastheconditioningstimulus.Tenblocks ofcoldstimulation(120s)alternatedwith11blocksoflukewarm stimulation(100s).Singlelaserstimuliweredeliveredonce40s beforetheendofeachblock(Fig.1).

Fortheconditioningstimulation,weusedanapproachsimilar tothatofSprengeretal.,whoused12smallbags(600goficeand 250mlofwater)tocoverthewholesurfaceofthelegandthefoot [16].Weappliedtwolargerbags(2L ofcrushediceand200ml ofwater)tothefootonly.Therewereseveralbagspreparedin advance.Theywerekeptinfoamboxeswhennotappliedtothe sub-ject.Besidescompressbagssomeadditionalicewasalwayspresent inthecontainer.Wecontrolledthetemperatureoftheicebags.In general,thetemperatureinCPMexperimentsisadjustedbya cir-culationcircuit(5–8◦C)[31,32]orastillicewaterbath(4–6◦C) [15,33]suchastobetolerable.Inourexperiment,thetemperature wasabout+2◦Canddidnotchangesignificantlythroughoutthe experiment.Thiswasachievedbyalargeproportionoficeinthe bagsandthestoragecontainer,whichmaintainedalow temper-atureduringthemeltingprocess.Ourmodeofcoldconditioning resultedinarelativelylesspowerfulnoxiousstimulation,butall oursubjectswereabletotoleraterepetitive2-minnoxiouscold applicationsduringtheentireexperimentalsession.

Thesameinvestigator(A.V.)manuallyappliedthecompresses overtherightfootofthesubjectsinthescannerroom.

(3)

Fig.1.Experimentalprotocol.Continuouslukewarm+35◦C(W)andcold2±2C(C)stimulationisappliedtotherightfootofthesubjects.Attheendofeachlukewarmand

coldblock,asinglelaserpulse(L)isappliedtothedorsumofthelefthand.

Laserpulses(testpain)weregeneratedbyathuliumYAGlaser (BaaselLasertech,Starnberg,Germany)thatemitscalibrated near-infraredradiation(wavelength1.96m;spotdiameter5mm;pulse duration1ms,energy600mJ).Tofamiliarizethesubjectswiththe procedure,threelaserpulseswereadministeredbeforethe begin-ningoftheexperiment.ThisexperimentwasrealizedusingCogent 2000(Cogent2000teamandCogentGraphicsdevelopedbyJohn RomayaattheWelcomeDepartmentofImagingNeuroscience). 2.3. Behavioraldataacquisition

BeforethefMRIstudy,thesubjectsfilledintheFrenchversionof Spielberger’straitanxietyinventory[34,35].Theywereinformed thattheyshouldbeattentivetothelevelofpainproducedbythe heatstimulusonthehandtheywouldhavetorateafterthe exper-imentalsession. Painratingswereperformedinpenandpencil formatonanumericratingscale(NRS)from0for‘nopain’to10for ‘worstimaginablepain’.Painratingsforlaserstimuliwereobtained separately during lukewarm and cold conditioning.“Behavioral CPM”wasdefinedasadifferenceofatleastonepoint between laser-painratingsduringcoldandlukewarmconditioning.

Subjectsratedcold-inducedpainforthe1standthelast con-ditioning, each time at onset and offset of the cold compress application.Theaverageofthesefourratingswastakenas mea-sureofcold-inducedpain.Habituationofcold-inducedpainwasthe differencebetweenratingsduringthe1standthelastapplication. 2.4. fMRIdataacquisitionandanalysis

MRIdatawereacquiredona3Thead-onlyscanner (Magne-tomAllegra,SiemensMedicalSolutions,Erlangen,Germany)using thestandard transmit-receivequadratureheadcoil.Astructural MRI was obtained in a separate session 20–30min beforethe functionalacquisition,usingahigh-resolutionT1-weightedimage [36]. Multislice T2*-weighted functional images were acquired with a gradient-echo echo-planar imaging sequence using an axial slice orientation and covering the whole brain (repeti-tion time=2040ms, echo time=30ms, flip angle=90◦, field of view=192×192mm,voxelsize3×3×3mm,34slices,25% inter-slicegap,matrixsize64×64×34).Thethreeinitialvolumeswere discardedtoavoidT1saturationeffects.Fieldmapdatawerealso acquiredtounwarpthefunctionalimages:acquisitiontime:TR/TE: 517/4.92&7.38ms, flip angle90◦, field of view: 220×220mm2, resolution:64×64,voxelsize:3.4×3.4×3mm,slices=32(3mm thick,30%gap;bandwidth:260Hz/Px).

ThefMRI imageswereprocessed usingthe“Statistical Para-metric Mapping” software (SPM8; Wellcome Trust Centre for Neuroimaging,UniversityCollegeLondon,UK;http//www.fil.ion. uce.ac.uk/spm)implementedinMATLAB(MathworksInc., Sher-born,MA,USA).Functionalimagetimeserieswerecorrectedfor

motionanddistortionusingthe“RealignandUnwarp”tool[37] togetherwiththeFieldmapToolbox[38].Thehigh-resolutionT1 image was co-registered with the functional images and seg-mentedintograymatter,whitematterandcerebrospinalfluid[39]. FunctionalimageswerespatiallynormalizedtotheMontreal Neu-rologicalInstitute(MNI)space(voxelsize:3×3×3mm)usingthe normalizationparametersobtainedfromthesegmentation pro-cedureand,subsequently,smoothedwithaGaussiankernelwith full-widthathalf-maximumof6mm.

2.5. Statistics

Dataanalysiswasperformedusingagenerallinearmodeland summarystatisticsapproach.

2.5.1. BOLDresponsestolaser-heatstimulationandtheir modulationbynoxiouscold

Atthefirstlevel,dataweresubjectedtohigh-passfilteringusing acut-offperiodof280s.Eachboxcarorimpulsestimulusfunction wasconvolvedwithacanonicalhemodynamicresponsefunction asimplementedinSPM8. Explicitmaskswerecreatedfromthe brainmaskimageavailableinSPM8.Thebinarymaskwascreated bythresholdingthebrainmaskimageatalevelof0.5intensity units.Thismaskfittedtheaverageanatomicalimageofoursubjects’ groupwell.

Preliminary experiments in four subjects with sustained cold/lukewarm stimulation onthe foot showed that the onset ofbothstimulationmodalitiesinducedhighamplitudebut tran-sientBOLDresponseswithmaximaindorsal SIcontralateral to thestimulatedside.Inordertodiscriminatethisearlyresponse, eachcoldconditionwasmodeledasthreesuccessiveblocks, last-ing10sfor“early”,10sfor“transitional”and100sfor“sustained” cold responses.For thesustainedlukewarm condition,weonly modeledthe1st10sfor“early”lukewarmandthe2nd10sfor “sec-ondary”lukewarmresponses.Thelaserresponsesweremodeledas instantaneouseventswithseparateregressorsforthoseoccurring duringcoldandthoseoccurringduringlukewarmstimulation.An autoregressivemodeloforder1,pluswhitenoise,accountedfor correlationoftheresiduals.

Aftermodelestimation,specificeffectsweretestedwith appro-priatelinearcontrastsoftheparameterestimates,resultingina contrastvaluefor eachvoxel.We computedthefollowing con-trasts:(1)earlynon-specificBOLDresponsestolukewarmorcold applications; (2) early cold pain responses vs early lukewarm responses; (3) sustainedcold responses; (4) laser pain-induced responsesduringlukewarmcondition,asa measureofbaseline testpain-inducedcerebralchanges,(5)cold-inducedmodulationof laserresponses,i.e.changeofthelaser-inducedresponsesduring coldapplicationascomparedtothoseinthelukewarmcondition.

(4)

Fig.2.Sequentialoutlineofthestatisticalanalysesusedinthestudyanddefinitionofterms.

The contrast imageswere included in a second-level group analysisusingarandom-effectsapproach, treatinginter-subject variabilityasarandomfactor.Group-leveleffectsweretestedusing one-sidedt-tests.

2.5.2. CorrelationbetweenbehavioralCPMandmodulationof laserpain-inducedBOLDresponses

Therelationbetweennoxiouscoldconditioningoflaser-evoked cerebralresponsesandbehavioralCPMassessedwiththeNRSwas furthertestedinthesecond-levelgroupanalysis.Themodulation oflaserresponses(contrast5)andthebehavioralanalgesiaratings werecomparedinabetween-subjectcorrelationanalysis(Fig.2). Thisallowedfindingbrainareas where cold-induced changeof thelaserpain-inducedresponsewasdirectlycorrelatedwithpain ratings,i.e.behavioralCPM.Thisregionalbehaviorallycorrelated modulationoflaser-induced neuronalresponseswasdefinedas “physiologicalCPM”.

2.5.3. Correlationbetweencold-inducedresponsesinprefrontal cortexandphysiologicalCPM

To verify our hypothesis on the relationship between cold-inducedphysiologicalCPMandcold-inducedresponsesinPFCareas, weanalyzedthecorrelationbetweenearly,sustained(negativeand positive)coldpainresponses(contrasts2and3)andphysiological CPM(seeFig.2).

Thevolumeofinterestwasderivedfromthestatisticalmapof physiologicalCPM.Clusterswereformedbythresholdingthevoxel statisticsatp<0.001uncorrected.Subsequently,thelargestcluster withinthe“painmatrix”wasbinarizedtocreateastudy-specific mask.ThismaskcoveredtheareaswherephysiologicalCPMwas observed.Theeigenvariatesofindividualcontrastimagesof cold-inducedmodulationoflaserresponses(contrast5)wereextracted fromabrainareaoutlinedwiththemaskofphysiologicalCPMand usedascovariateregressorsinfurthercorrelationanalyses,in par-ticular,inaone-samplet-testofindividualimagesofcontrasts2 and3(earlyandsustainedcoldresponses).

2.5.4. Significancelevelanddirectedsearchforpainresponses andpainmodulation

First,weusedanundirectedsearch,lookingforresponseswith p<0.05familywiseerrorrate(FWER)correctedforthewhole-brain volumebothatvoxelandclusterlevels.Fortheclusterlevel infer-ence,wefirstuseda cluster-formingthresholdofp<0.001(not corrected).Significantresultsfromthisundirectedsearchare indi-catedinthetextandinthetablesby*WB(wholebrain).Hereandin otheranalyses,wetookintoaccountbothpeak-andcluster-level significanceandlabeledthemas*WB,respectively,nexttopeak coordinatesandclustersize.

Second,weperformedadirectedsearchfocusingonvolumesof interest(VOI).Weapplieda“smallvolumecorrection”inorderto obtainFWER-correctedp-valuesatbothvoxeland clusterlevels (withacluster-formingthresholdofp<0.001).“Painmatrix”VOI includedbrainstem,thalamus,striatumandcorticalareasknown torespondtonoxiousstimulation(seeSupplementaryMaterial1). Significantresultsforlaserandcoldresponsesin“Painmatrix”VOI

werelabeledwith*PM.Whenalargeclusterencompassing multi-plebrainareaswasfound,itwasmaskedwiththepainmatrixto identifyspecificbrainareasincluded.WedefinedfivebilateralVOI includingPFCareasmostoftenreportedtobesourcesof descen-dingCPMeffects:medialOFC,lateralOFC,ACC,rostrallateralPFC andcaudallateralPFC(seeSupplementaryMaterial2).Thosefive masksofPFCVOIwereusedforcoldresponsesandthecorrelation analysisofcoldresponsesandphysiologicalCPM.Significantresults ofthisdirectedsearchwereindicatedwith*DS(directedsearch).

Finally,smallvolumecorrectionwasappliedtospheresaround physiologicalCPMpeakswitharadiusof10mm(seeTable2)to test whethertheindividualratingsofanxietyor across-session coldhabituationcouldinfluencetheabove-mentionedresponses. Significantresultsoftheseanalyseswerelabeled*SV(small vol-ume).Pearson’scorrelationwasestimatedfortherelationbetween behavioralandphysiologicalvariables.

3. Results

3.1. BehavioralCPM

Althoughbothlaserstimuliandcoldapplicationproducedpain, thelatterwasperceivedasmorepainfulontheNRSthantheformer (5.6±0.4vs3.8±0.3,respectively,mean±SEM)(Fig.3).

Wefoundnosignificantdifferenceinaverageratingsof laser-inducedpainbetweenthelukewarmandcoldconditions,e.g.no significantmean cold-induced behavioralCPM. Thiswasdue to thefactthat,attheindividuallevel,7subjectshadhypoalgesia duringcoldconditioning(NRSreductionof≥1),10subjectshad hyperalgesia(increase of ≥1) and 7 subjects had equal ratings oflaser-induced painduring thelukewarm and coldconditions (Fig.3).

3.2. CerebralBOLDresponses

3.2.1. Non-specificcerebralresponsestoconditioningstimuli Non-specific responses to the application on the right foot of eithercold orlukewarm compresseswere significant(FWER p<0.001voxellevel)in bilateral,but predominantly left,SI,SII, SMA,thalamus,insula, precuneus,ACC,MCCanddorsal premo-torcortex,aswellasin centralcerebellum,midbrain(including PAG),bilateralamygdala,predominantlyrighthippocampusand dorsolateralPFC.

Lesssignificantactivationwasalsofoundincaudatenucleus bilaterally, left putamen, frontal operculum and cuneus (FWER p<0.05voxellevel)(seeSupplementaryMaterial3).

3.2.2. Specificcerebralresponsestoearlycoldstimulation

Atthewholebrainlevel,whenonlytheresponsetothefirst10s ofcoldstimulationwasconsideredandcontrastedtothatduring thefirst10soflukewarmstimulation,wefoundsignificant activa-tioninalargeclusterincludingbilateralprecuneus,dorsalposterior cingulateandpredominantlyleftoccipitalcortex(1259voxelsor 34cm3,p<0.005FWERclusterlevel)(Table1AandFig.4).In addi-tion,threeothersignificantlylargeclusters(p<0.05FWERcluster

(5)

Fig.3.Distributionofscoresonanumericratingscale(NRS,0-10)fornoxiouspain anditsmodulationbythermalconditioning.Thetotalnumberofsubjects(abscissa) is24.(A)Coldpainscores.(B)Laserpainscoresduringlukewarmstimulation.(C) Variationoflaserpainscoresduringnoxiouscoldstimulationexpressedasthe differencebetweenindividuallaserpainscoresduringlukewarmandcold con-ditioning.Positivevaluesrepresentcold-inducedanalgesiawhilenegativevalues indicatecold-inducedhyperalgesia.

level)encompassedlefttemporo-occipitalcortex,rightdorsal pre-motor/motorcortexandrightputamen/amygdala.

UsingthedirectedsearchwithinVOIofthe“painmatrix”,the precuneussignalwasconfirmedtobesignificant(p<0.05FWER voxellevel)and significantclusterswereidentified inpremotor cortexandventralposteriorcingulate(bothp<0.05FWERcluster level).InprefrontalVOI,therewerethreesmall,butsignificant clus-tersinmedialOFCandrightcaudo-lateralPFC(Table1AandFig.4).

Table1

Brainareasshowingaresponsetonoxiousstimuli.

Clustersize(kE) andclusterlevel significance Peaklocation ([X,Y,Z]inMNI coordinates)and voxellevel significance 1A.Earlynoxiouscold-inducedBOLDresponses

Precuneus [–3–7349]*PM

Temporo-occipitalcortex 152*WB [–57–70–75]

Dorsalpremotor/motorcortex 124*WB(#65*PM) [45558]

Caudalputamen,amygdala, hippocampus

119*WB [27–4–8]

Dorsalposteriorcingulate #72*WB [6–4046]

Ventralposteriorcingulate #68*PM [3–4316]

MedialOFC 7*DS [1247–11]*DS

8*DS [–929–17]*DS

Caudo-lateralPFC 10*DS [482328]*DS

1B.Sustainednoxiouscold-inducedBOLDresponses Whitematter,caudalcorona

radiata

15*WB [–21–2525]*WB

Caudatetailandwhitematter adjacenttoit

5*WB [18–1628]*WB

1*WB [–18–125]*WB

1*WB [–18–1328]*WB

Whitematter,spleniumof corpuscallosum

3*WB [12–3713]*WB

Posteriorinsula 4*WB [–33–2219]*WB

Largecaudalwhitematter cluster

3350*WB [–21–2525]*WB

Rostro-dorsalthalamus #102*WB [–12–1919]

1C.Sustainednegativenoxiouscold-inducedBOLDresponse

MedialOFC 11*DS [–1229–17]*DS

7*DS [335–17]

1D.Laserpain-inducedresponses

Insula/frontaloperculum 514*WB [3617–2]*WB 79*WB [–3020–5]*WB 17*WB [45–1616]*WB 18*WB [–39–7-2]*WB 2*WB [–30147]*WB 1*WB [42–710]*WB Premotor #29*WB [481119]*WB Amygdala #10*WB [18–4–14]*WB #4*WB [–21–4–14]*WB ventralputamen 64*WB [–158–11]*WB anteriorthalamus 28*WB [924]*WB PCC 26*WB [3–2231]*WB SII 25*WB [–60–2822]*WB 1*WB [63–2222]*WB ACC 19*WB [02025]*WB angularcortex 18*WB [–60–4943]*WB 4*WB [33–4943]*WB 6*WB [60–4037]*WB Dorso-lateralPFC 9*WB [48387]*WB Hippocampus 2*WB [–15–34-8]*WB #48*PM [–18–7–14]*WB Caudo-ventralthalamus 1*WB [6–284]*WB Thalamus 1*WB [12–1010]*WB SI 1*WB [54–1016]*WB OFC 1*WB [2729–17]*WB Precuneus 287*WB [12–6431]*PM Rostro-lateralPFC 255*WB [–483819] 81*WB [365331] SMA #215*WB [61161]*PM midbrain #231*WB [0–31–5]*PM

Middletemporalgyrus 60*WB [–60–587]

Cerebellum 52*WB [–18–73–32]

57*WB [12–91–26]

Caudatenucleus #8 [–984]*PM

#24 [957]*PM

Note:WB,wholebrainvolumecorrectedundirectedsearch,*PM,“painmatrix”VOI directedsearch;*DS,directedsearchatfivesmallerPFCandACCVOI.Ifindicated nexttoclustersizeornexttopeakcoordinatepointsatclusterlevelandvoxellevel FWER<0.05significance;#,inclusiveanatomicalmaskwasappliedtostatistical imageinordertoseparatesignificantresponseinasmallbrainareawhichisapart ofalargercluster.

(6)

Fig.4. Cerebralresponsestoearly(shadedareaswithbackdots)andsustained(shadedareaswithwhitedots)cold-inducedpain.p<0.001uncorrectedmapsoverlaidon averageofindividualnormalizedstructuralT1images(n=24).Crosshairsarelocatedatpeaksofearlycold-painresponseinprecuneus(A),OFC(B),dorsolateralprefrontal cortex(C)andapeakofsustainedcoldpainresponseinposteriorinsula(D).VOIsofdirectedsearchareoutlinedwithgreyframes.

3.2.3. Specificcerebralresponsestosustainedcoldstimulation

SustainedcoldstimulationinducedsustainedBOLDresponses

inonlyafewclustersmostlyextendingintoposteriorwhite

mat-ter(Table1B).Responsesweresignificantinposteriorinsulaand tailofcaudate(FWER<0.05 voxellevel).Thosepeakswerepart ofalargesignificantwhitemattercluster(30cm3,p<0.001FWER clusterlevel)(Fig.4).Whenmaskingwiththe“painmatrix”VOI wasapplied,this30-cm3clusterincludedasignificantlargecluster inrostro-dorsalthalamus(p<0.01FWERclusterlevel)(Table1B). None of thedirected searches in prefrontal cortex VOI yielded significantsustained positiveresponsestocoldstimulation, but negativesustainedcold responses weredisclosedin mOFC VOI (Table1CandFig.5).

3.2.4. Laser-inducedcerebralresponsesduringlukewarm stimulation

Laserheatstimuliinducedasignificantresponseinmostareas ofthe“painmatrix”mask(seeSupplementaryMaterial1):insula, premotorcortex,amygdala,ventralputamen,anteriorthalamus, PCC,SII,ACC,and dorso-lateralPFC(peak levelFWERp<0.005) (Table1D).Outside of the“painmatrix”,there weresignificant laser-inducedresponsesbilaterallyinangulargyrusand cerebel-lum.

3.2.5. Changesinlaser-inducedcerebralresponsesduringcold stimulation

Theoveralldifferenceinresponsestolaserstimuliwithor with-outcontinuouscoldapplication(contrast 5)wasnotsignificant, whichwasduetothecontrastingCPMeffectbetweensubjects. Individualdifferencesincold-inducedbehavioralCPMscoreswere

infactproportionaltothecold-inducedchangeinlaser-induced responsesintherightanteriorinsularclusterpartiallyincluding whitematter,ventralpremotorcortexanddorsalstriatum.A sim-ilarrelationshipwasfoundinanotherclustercomprisingmostly

Fig.5. NegativeresponsestosustainednoxiouscoldstimulationinmedialOFC (out-linedwithwhite).Suprathreshodp<0.001uncorrectedT-maps.VOIsofdirected searchareoutlinedingrey.

(7)

Table2

PhysiologicalCPM:Coldmodulationoflaser-inducedresponsescorrelatedwith behavioralCPM.

Clustersize(kE) andclusterlevel significance Peaklocation ([X,Y,Z]inMNI coordinates)and voxellevel significance Anteriorinsula/white matter/premotor cortex 122*WB [302616]*WB Posteriorinsula/SII 90**WB [–36–1619]*WB LateralOFC 64*WB [2414–20] Fusiform gyrus/parahippocampal 84**WB [–21–52–11] Posteriorinsula/SII 58**PM [–36–1619]*PM

Note:seeTable1.

graymatterwithintheanatomicallydefined“painmatrix”mask: leftposteriorinsulaandSII.Thereweretwoothersignificantbut smallerclustersinlateralOFCandfusiformgyrus/parahippocampal cortex(Table2).However,whenwe appliedsmallvolume cor-rectionstothe“painmatrix”VOI,onlytheleftposteriorinsula/SII clustersurviveddirectedsearch(Table2andFig.6).

We used therefore the insula/SII cluster in the “pain matrix” (cluster size=58 voxels) to extract the eigenvariates of cold-inducedmodulationof laser responses(contrast 5). The cold-induced change of laser-induced responses in this cluster was directlycorrelated with individual behavioral CPM ratings. As mentioned, we called “physiological CPM” this concordant behavioral-BOLDresponsemodulation(Fig.2).

3.2.6. Correlationwithspecificcold-inducedprefrontalcortex responses

ThephysiologicalCPMparameterwasusedasacovariate regres-sorinaone-samplet-testofindividualimagesofcontrasts2and3 (earlyandsustainedcoldresponses).Thistestedthehypothesis thataspecificpatternofcold-inducedresponsesinPFCpreceeds subsequentphysiologicalCPMandhencemayexplaintheindividual differencesinCPM(seeFig.2fortheanalysisflow).When behav-ioralCPMratingswereincludedasacovariate,weobtainedsimilar

Fig.6.CerebralareasthatwereactivatedbytheLaser-teststimulationduring luke-warm,i.e.control,conditioning(shadedareawithblackdots)orsustainedcold conditioning(shadedareawithlight-greydots)andareaswherecold-induced inhi-bitionoflaserresponsescovariatedwithindividualanalgesiaratings(“physiological analgesia”,outlinedinwhite).p<0.001uncorrectedmapsareshownona sag-ittalsectionofaverageofindividualnormalizedstructuralT1images(n=24)at X=−36mmMNI.Crosshairlocationisatthemaximumofgreencontrast([–36–16 19]*PM).

Table3

Brainareaswherecold-inducedBOLDresponsesco-variatewithphysiologicalCPM (sourceofthephysiologicalCPMcovariateinleftinsula/SII).

Clustersizeand clusterlevel significance

Peaklocation ([X,Y,Z]inMNI coordinates) andvoxellevel significance 3A.Earlycold-inducedBOLDresponsesrelatedtophysiologicalCPM

Dorso-lateralPFC 29*WB [–155025]*WB 4*WB [–212531]*WB RostralACC 18*WB [94410]*WB 3*WB [–94116]*WB CaudalACC 12*WB [–122931]*WB Anterior-laterallingual/medial fusiformgyrus 11*WB [–24–55-2]*WB MedialPFC 8*WB [65616]*WB

Whitematter,caudalcorpus callosum

8*WB [–15–1331]*WB

Caudatehead 7*WB [–15201]*WB

Whitematterprecentralgyrus 4*WB [–36–728]*WB

Caudo-lateralOFC/rostralinsula 3*WB [–2723–14]*WB

1*WB [–3029–11]*WB 1*WB [–2411–17]WB Ventralstriatum/accumbens 2*WB [68–14]*WB SubgenualACC 2*WB [026–8]*WB Putamen 2*WB [2111–8]*WB 1*WB [27510]*WB Frontalpole 1*WB [186513]*WB 1*WB [–275910]*WB ACC 223*DS [94410]*DS Rostro-lateralPFC 178*DS [–155322]*DS MedialOFC 68*DS [635–2]*DS 11*DS [–944–8]*DS Caudo-lateralPFC 26*DS [452022]*DS LateralOFC 20*DS [–3029–11]*DS 4 [3929–17]*DS

3B.Positivesustainedcold-inducedBOLDresponsesrelatedto physiologicalCPM

Whitematter,internalcapsula, coronaradiata,putamen,insula

172*WB [–18114]

Cerebellum 146*WB [33–58–38]

Whitematteranteriorcorona radiata,adjacenttorostro-lateral PFC

117*WB [243816]

Rostro-lateralPFC 17*DS [274710]

3C.Negativesustainedcold-inducedBOLDresponsesrelatedto physiologicalCPM

LateralOFC,whitemattercorona radiata,insula,ACC,rostro-lateral PFC 229*WB [–3023–17] Cerebellum 83*WB [21–64–41] 64*WB [36–73–75] 73*WB [–27–64–41] Rostro-lateralPFC 41*DS [–244716]*DS LateralOFC 24*DS [–3023–17]*DS ACC 8*DS [–122919]*DS 5*DS [–124422] Caudo-lateralPFC 23*DS [453516]*DS

Note:seeTable1.

thoughlesssignificantresultsthatarenotpresentedbecausewe considerthemredundant.

EarlycoldresponsesindorsolateralandmedialPFC,ACC, lin-gual/fusiformgyrus,whitematteratprecentralgyrusandcaudal corpus callosum, caudo-lateral OFC/rostral insula, striatum and frontalpoleweredirectlycorrelatedwithphysiologicalCPM(FWER p<0.05 voxel level) (Table 2A and Fig.7). We also observed a large (64cm3) significant cluster (FWER p<0.05 cluster level) that included perigenual ACC, rostral bilateral ACC, left lateral OFC,frontaloperculum, DLPFC,posteriorACC and headof cau-dateaswellasrightdorso-medialPFCandfrontalpole(Table3A

andFig.7).

ThedirectedsearchwithinfiveVOIindicatedthatphysiological CPMwassignificantlypredictedbyearlycold-inducedresponses

(8)

Fig.7. Positiveresponsestoearly(outlinedwithwhiteframes)andnegativeresponsestosustainedcold-pain(greyshading)covariatingwith“physiologicalanalgesia”in posteriorinsula/SII.p<0.001uncorrectedmapsareshown.VOIsofdirectedsearchareoutlinedwithblackframes.Greatercold-inducedanalgesiaisrelatedtopositiveearly ornegativesustainedcold-painresponsesinACC,OFCandrostro-lateralPFC.

inallfiveVOI(medialandlateralOFC,ACC,leftrostralandright caudallateralPFC)(Table3A).

AsillustratedinFig.8formedialsubgenualACC,rostralACCand lateralOFCvoxels,BOLDresponseswerepositiveduringtheearly phaseofcoldconditioninginsubjectswhobecamehypoalgesicand hadreducedlaser-inducedactivationinleftposteriorinsula/SII,but negative(orlesspositive)inthosewhodevelopedhyperalgesia. Inthoseareastherewasaclearpositivecorrelationbetweenthe hemodynamicresponsesandphysiologicalCPM,asshowninFig.9 forthesubgenualACCVOI.

The correlationbetween the cerebralresponses during sus-tainedcoldconditioningandphysiologicalCPMweremorecomplex andvariablebetweenareas.Activationduringsustainedcold con-ditioning was positively related to physiological CPM in a few areasincludingcerebellum,coronaradiata,internalcapsula,dorsal putamenandinsula,andawhitematterclustercomprising ante-riorcoronaradiataandadjacentrostro-lateralPFC.(FWERp<0.05 clusterlevel).Thedirectedsearchconfirmeda singlesignificant cluster(FWERp<0.05clusterlevel)inrightrostro-lateralPFCVOI (Table3B).

By contrast, there was a negative correlation of sustained cold-inducedbrainactivitywithphysiologicalCPMinmostofthe prefrontalVOI(Table3C).Withthedirectedsearch,deactivation wassignificantlyrelatedtohypoalgesiaduringcoldconditioning inlateralOFC,ACC,leftrostro-lateralandrightcaudo-lateralPFC (Table3C).Atthewholebrainlevel,therewasasignificant clus-ter(6cm3,FWERp<0.05clusterlevel)includinglateralOFC,white matteranddorsolateralPFC(seeFig.7),aswellasthreesignificant clustersincerebellum.

3.3. Correlationswithanxietyandhabituation

Themean(trait)anxietyscoreinoursubjectswas36±1,range 26–51.Greatertraitanxietyscorescorrelatedoverallwith behav-ioralhyperalgesiaduringCPM(r=−0.46,p<0.05).

Accordingtotheretrospectivepainrating,therewasaslight decrease in cold pain during the application (from 6.0±0.5 to 5.1±0.4)andbetweenthefirstapplicationandthelastapplication (from6.7±0.4to4.5±0.4).CPMcorrelatedpositivelywith habit-uationofcoldpain,i.e.thedeclineofperceivedpaininthefootat theendoftheexperiment,(r=+0.49,p<0.05)andnegativelywith coldpainratingsattheendofthecompressapplication(r=−0.45, p<0.05).

Habituationofcoldpainratingscorrelatedwithcold-induced responsesinareasrelatedtophysiologicalCPM.Inparticular,like thelatter,habituationwasrelatedtoearlycold-inducedactivation inrostro-medialPFC[–66522],mOFC/sgACC[–1229–8]and cere-bellum[3–73–17](FWERp<0.05,clusterlevel)andothercluster inmOFCVOIconfirmedbydirectedsearch[632–14].Therewas,in addition,amedulla/caudalponscluster[6–40–35]wherenegative responsestosustainedcoldcorrelatedwithgreaterhabituationto cold(FWERp<0.05,clusterlevel).

4. Discussion

Theobjectiveofourstudywastobetterunderstandthe neu-ronalmechanismsof conditionedpainmodulation(CPM) using theheterotopicnoxious stimulationparadigm.Forthis purpose,

(9)

Fig.8.AveragetimeseriesofBOLDsignalsinprefrontalareaswhereearlycoldresponsescorrelatewith“physiologicalCPM”.VOIare6mmradiusspheresaroundthepeaks atmedialsubgenualanteriorcingulatecortex(sgACC[026−8]),anteriorcingulatecortex(ACC[94410])andlateralorbitofrontalcortex(lOFC[−3029−11])(Table3A). Eigenvariateswereextracted,normalizedtothemedianvalueofagiventimeseries,epochedandindividuallyaveragedforcoldleft)andlukewarm(right)conditionand thegroupaveragewasestimated.Thecriteriafordefining“physiologicalanalgesia”(dottedline)and“physiologicalhyperalgesia”(solidline)groupswereupperandlower quartilesofthephysiologicalanalgesiaparameterestimateextractedfromposteriorleftinsula/SIIVOI,anareawithbehaviorally-correlatedmodulationoflaser-induced neuronalresponses(seemethods2.5.2andresults3.2.5).

wesearchedwhetherindividualdifferencesbetweenconditioned modulationofbehavioralandcerebralresponsestothetestheat painwererelatedtothecerebralresponsesinducedbythecold conditioningstimulation.

4.1. Behavioralpainratings

Onaverage,overthe10consecutivetrials,therewasno signif-icantdecreaseinlaserpainratingsduringtheconditioningcold stimulation.FacilitatoryCPM(i.e.hyperalgesia)wasobservedin otherstudies,thoughoverallinasmallerproportionofsubjects. Inpre-operativepain-freepatients,facilitatoryCPMpredicts sub-sequentpost-surgerychronicpain[40]andhigheranalgesic con-sumption[41].ThelikelihoodtohavefacilitatoryCPMofaheattest stimulusbytourniquetischemicnoxiousconditioningwashigher insubjectswithlow5-HTTexpression[42].Inourstudy,only29% ofsubjectshadpainreduction,while29%hadnochangeand42% increasedlaserpain,clearlydemonstratinginter-individual differ-ences.ThiscontrastswithotherCPMstudies[15,16]andcouldhave severalnon-mutuallyexclusiveexplanations.

ThedegreeofCPMvariesbetweenstudiesanddependsonthe studydesign. For instance,CPM wasmarkedly greater(50%) in Pichéetal.[15]thaninSprengeretal.(20%)[16].Inourstudy,the conditioningcoldstimulationwasappliedrepeatedlycomparedto

a singlecounterstimulation inthetwopreviousstudies[15,16]. Individualdifferencesinhabituationtorepeatednoxiousstimuli maythusplayaroleinourparadigm,themoresothatwefound a positivecorrelationbetweenhabituation ofcold-inducedpain acrosssessionsand cold-inducedlaserpainreduction.Testpain producedbythetrainsofelectricshockslikeinPichéetal.’s pro-tocolmaybemoreamenabletoabehavioralCPMeffectthanpain inducedbyathermalstimuluswithsloweronsetandoffsetslopes usedinSprengeretal.’s[16]andinourstudy.Also,oursubjects ratedtheteststimulusaslesspainful(3.8/10NRS)comparedto thoseoftheabove-mentionedstudies.Thismayhavecontributed tothelackofconditionedhypoalgesiainsomesubjects,astheCPM effectcan beproportionaltothetest pain intensity[43]. How-ever,itisknownthatmoderatelypainfulteststimuli[15,44](about 40–50/100VAS),andevenonlyslightlypainfulones[31]are suffi-cienttoobtainreliableCPM.

Othermethodologicalissues arediscussedin Supplementary Material4.

Besides methodological aspects related to the experimental design, psychobehavioral factors might contribute to the large interindividualvariabilitywefoundforCPM.

Amongthem, apriori expectationis wellknowntostrongly correlate with drug-induced analgesia [45] and pain modula-tion[32,33,44].Inthecase ofpharmacologicalopioidanalgesia,

(10)

[33]. Stronger analgesia expectation and reduced stress during noxiouscoldconditioningwerelinkedtogreaterinhibitoryCPM [32].Althoughwekeptoursubjectsblindaboutthedirectionof expectedpainintensitychangesinducedbythecoldapplications, wecannotexcludethatexpectationplayedaroleinsomeofthem. Uncertainty about upcoming pain intensity also has a specific andpotenthyperalgesiceffect[46].Anxietyisknowntopartially mediatetheeffectofexpectationonthedirectionofCPM[44].Our findingthatanxietywasnegativelyrelatedwithbehavioralCPM supportsaroleofpsychologicalmodulators.

4.2. Cerebralresponsestotheconditioningcoldstimulation

Wefoundhighlysignificant(voxellevelFWERp<0.05) non-specificactivationsatonsetandoffsetofbothcoldandlukewarm stimulationsinmanyareasbelongingtotheso-called“painmatrix”. Themaximumofthisnon-specificresponsewasinleftSI,whichis consistentwithsomatosensoryresponsestothealternating sen-sorystimulationoftherightfoot.

By contrast, when early responses to cold were contrasted againstearlyresponsestolukewarm applicationtodetect pain-specificactivation,aclustercomprisingmainlybilateralprecuneus wassignificantly activatedbut not “pain matrix”areas suchas ACCandinsula.Theprecuneusisknowntorespondtomultimodal salientstimuli[47].

Wefoundnosignificantactivationinmostareasofthe“pain matrix”duringthesubsequentsustainedphaseofcoldapplication, withthenotableexceptionofposteriorinsula/SII.Thelatter activa-tionisconsistentwithacentralrolefortheoperculo-insularcortex inspecificpainprocessingandcodingofpainintensity[48–53]. Althoughourstudywasnotspecificallydesignedtoaddressthis question,itsresultsfavornonethelesstheconceptthatthe“pain matrix”ismostlya“saliencematrix”ofareasthatprocesssalient multimodalsensoryinputs[54]andthattheposteriorinsulamay bemorepain-specific[50,53].

Duringthesustainedphaseofcoldapplication,wefound deacti-vationinmedialOFC.Thisareaisknowntobeactivatedbyvaluation ofarewardoutcome[55]andwespeculatethatitsdeactivation couldbe related to the negative affective response due tothe sustainednoxiousstimulation.Theobservedclustersofsustained BOLDresponseinthewhitematterarediscussedinSupplementary Material5.

4.3. Cerebralresponsestothelaserteststimulusandtheir conditioningbycoldstimulation

Laserpain-inducedactivationsweresignificantinvariousareas of the “pain matrix” in line with a number of prior studies [47,56,57].Thisindicatesthattheintensityofthelaserstimuliwas sufficienttoinducecerebralresponsestypicallyreportedafter nox-iousstimulation.

Similartothebehaviorallaserpainratings(seeSection4.1),the laser-inducedBOLDresponsesdidnotdifferonaveragebetween cold and lukewarm conditioning.The inter-individual variation in CPM, however, that ranged from reduction to increase in painwascorrelatedwiththemodulationoftheBOLDresponse. Behavioralcold-inducedhypoalgesiawasassociatedwiththe sup-pression of laser-induced responses in posterior insula and SII contralateraltothecoldstimulation.Likesustainedcold-induced

Fig.9.RelationshipbetweenactivationofsgACCandbehavioralconditionedpain modulation.Eigenvariatesofbetaimagesforonlyearlycoldresponsewereextracted fromasphereof6mmradiusaroundthepeakofearlycoldvsearlywarmcontrast significantlycorrelatedto“physiologicalconditionedpainmodulation”(the eigen-variatesofthelaserwarmvslasercoldcontrastcorrelatedtobehavioralconditioned painmodulation)(Table3A).Thescatterpotisonlyforillustrationpurposes.

activation,thephysiologicalCPMthusinvolvedtheposteriorinsula, supportingagainitscriticalroleinpainaswellasCPMprocessing [50,53].

4.4. Correlationofcold-inducedcerebralresponseswith physiologicalCPM

In medial and lateral OFC, ACC, rostral and caudal lateral PFC,theearlycold-inducedBOLDresponsesweredirectlyrelated to subsequent CPM (see Figs. 8 and 9). These findings are in agreementwithotherstudies,where CPM wasassociated with activityevoked by thecold painstimulus inmPFC, lateralOFC [15,26]androstralACC[16].Bycontrast,BOLDactivationduring thelaterphaseofcoldstimulationwaslesspredictiveof physio-logicalCPM,asthecorrelationwassignificantonlyinrostro-lateral PFC.

Thecorrelationbetweenearlycold-inducedPFCresponsesand subsequentphysiological CPMcouldbe due tothefact thatthe conditioningstimulusisabletostrengthenconnectivitybetween corticalsourcesofdescendinganalgesiaandtheirtargetsin sub-cortical structures [58]. For instance, increased connectivity of subgenualACCwithmidbrainandamygdalawascorrelatedwith CPM[16].

Wealsofoundanunexpectedcorrelationbetween physiolog-icalCPManddeactivationinlateralOFC,ACC,lateralrostraland caudalPFCareasduringsustainedcoldapplication,whichwasnot reportedinothersstudies[15,16].Thismaybeduetothedifferent experimentaldesignofourstudy,butdifferencesinexpectation

(11)

orso-called“off-setanalgesia”couldbealternativeexplanations. Inamodelofvisceralpain,expectation-mediatedplacebo analge-siawasassociatedwithfMRIdeactivationinthedorsolateralPFC [59].NegativeBOLDresponsesinventro-medialPFC/OFCwerealso recordedduringlargereductionsofpainratinginducedbyaslight decreaseofa noxiousstimulation,i.e. off-setanalgesia[8].This paradigmcouldberelevantforourstudywheretherewas habit-uationofcoldpainratingsbetweenthe1standlastapplications, andhenceadecreaseinperceivedpain.

4.5. Correlationswithcoldpainhabituation

Habituationofcoldpainperceptioncorrelatedpositivelywith behavioral CPM, with early cold-induced lateral OFC activation thatisalsorelatedtophysiologicalCPM,andwithsustained cold-induceddeactivationinpons/cerebellum.Habituationofpainover severaldayswasfoundassociatedwithsubgenualACCactivation [60].Habituationthuslikelyinvolvescentralneuronalnetworks thatitsharesatleastpartiallywithCPM.Diffusenoxiousinhibitory control,non-noxiousinhibitorycontrolandhabituationareallpart ofendogenousanalgesiaandthoughttoberelatedtoeachother [43].

Toconclude,ourstudyconfirms that sustainednoxiouscold stimulationchieflyactivatesposteriorinsula/SII.Activationinother areasoftheso-called“painmatrix”appearstobemerely deter-minedbysaliencyandnoveltyratherthanbythenoxiouscharacter ofthestimulus.Wefindgreatinter-individualdifferencesinCMP thatrangesfromdecreasestoincreasesofthetestlaser-heatpain. LargerprefrontalpositiveBOLDresponsestoearlycold stimula-tionseemtopredictsubsequentconditionedanalgesia.Bycontrast, inadjacentprefrontalcorticesconditionedanalgesiaisassociated withdeactivationduringthesustainedconditioningcold stimula-tion.WeshowthatCPMisnegativelycorrelatedwithtraitanxiety, butpositivelywithhabituationofcoldpain.Theindividual differ-encesinprefrontalcortexactivityandCPMneedtobeconsidered infuturestudiesofdiseasedsubjects,includingmigraineurs. Acknowledgements

Thisstudywassupportedbytheresearchconvention3.4.650.09 fromtheNationalFundforScientificResearch–BelgiumtoJ.S.,and byaresearchgrantfromtheCHUofClermont-Ferrand–Franceto R.D.andJ.S.WearegratefultoDrsAndréLuxen,EricSalmonand PierreMaquetfororganizational andtoChristianDegueldrefor technicalassistance.

Theauthorsdeclarenoconflictsofinterest. AppendixA. Supplementarydata

Supplementarydataassociatedwiththisarticlecanbefound,in theonlineversion,athttp://dx.doi.org/10.1016/j.bbr.2014.11.028. References

[1]MayerDJ,LiebeskindJC.Painreductionbyfocalelectricalstimulationofthe brain:ananatomicalandbehavioralanalysis.BrainRes1974;68:73–93.

[2]LeBarsD,ChitourD,KrausE,ClotAM,DickensonAH,BessonJM.Theeffectof systemicmorphineupondiffusenoxiousinhibitorycontrols(DNIC)intherat: evidenceforaliftingofcertaindescendinginhibitorycontrolsofdorsalhorn convergentneurones.BrainRes1981;215:257–74.

[3]GearRW,AleyKO,LevineJD.Pain-inducedanalgesiamediatedbymesolimbic rewardcircuits.JNeurosci1999;19:7175–81.

[4]LevineJD,GordonNC,FieldsHL.Themechanismofplaceboanalgesia.Lancet 1978;2:654–7.

[5]ValetM,SprengerT,BoeckerH,WillochF,RummenyE,ConradB,etal. Dis-tractionmodulatesconnectivityofthecingulo-frontalcortexandthemidbrain duringpain—anfMRIanalysis.Pain2004;109:399–408.

[6]FreundW,KlugR,WeberF,StuberG,SchmitzB,WunderlichAP.Perception andsuppressionofthermallyinducedpain:afMRIstudy.SomatosensMotRes 2009;26:1–10.

[7]YilmazP,DiersM,DienerS,RanceM,WessaM,FlorH.Braincorrelatesof stress-inducedanalgesia.Pain2010;151:522–9.

[8]YelleMD,OshiroY,KraftRA,CoghillRC.Temporalfilteringofnociceptive informationbydynamicactivationofendogenouspainmodulatorysystems. JNeurosci2009;29:10264–71.

[9]LeknesS,BernaC,LeeMC,SnyderGD,BieleG,TraceyI.Theimportanceof context:whenrelativereliefrenderspainpleasant.Pain2013;154:402–10.

[10]Pud D, Granovsky Y, Yarnitsky D. The methodology of experimentally induceddiffusenoxiousinhibitorycontrol(DNIC)-likeeffectinhumans.Pain 2009;144:16–9.

[11]YarnitskyD,Arendt-NielsenL,BouhassiraD,EdwardsRR,FillingimRB, Gra-notM,etal.Recommendationsonterminologyandpracticeofpsychophysical DNICtesting.EurJPain2010;14:12.

[12]CarstensE,YokotaT,ZimmermannM.Inhibitionofspinalneuronalresponses tonoxiousskinheatingbystimulationofmesencephalicperiaqueductalgray inthecat.JNeurophysiol1979;42:558–68.

[13]Villanueva L, Le Bars D. The activation of bulbo-spinal controls by peripheralnociceptiveinputs:diffusenoxiousinhibitorycontrols.BiolRes 1995;28:113–25.

[14]SlukaKA,DeaconM,StibalA,StrisselS,TerpstraA.Spinalblockadeofopioid receptorspreventstheanalgesiaproducedbyTENSinarthriticrats.JPharmacol ExpTher1999;289:840–6.

[15]PicheM,ArsenaultM,RainvilleP.Cerebralandcerebrospinalprocesses under-lyingcounterirritationanalgesia.JNeurosci2009;29:14236–46.

[16]SprengerC,BingelU,BuchelC.Treatingpainwithpain:supraspinal mecha-nismsofendogenousanalgesiaelicitedbyheterotopicnoxiousconditioning stimulation.Pain2011;152:428–39.

[17]vanWijkG,VeldhuijzenDS.Perspectiveondiffusenoxiousinhibitorycontrols asamodelofendogenouspainmodulationinclinicalpainsyndromes.JPain 2010;11:408–19.

[18]PielstickerA,HaagG,ZaudigM,LautenbacherS.Impairmentofpaininhibition inchronictension-typeheadache.Pain2005;118:215–23.

[19]SandriniG,RossiP,MilanovI,SerraoM,CecchiniAP,NappiG.Abnormal modu-latoryinfluenceofdiffusenoxiousinhibitorycontrolsinmigraineandchronic tension-typeheadachepatients.Cephalalgia2006;26:782–9.

[20]de Tommaso M, Difruscolo O, Sardaro M,Libro G, Pecoraro C, Serpino C, et al. Effects of remote cutaneous pain on trigeminal laser-evoked potentials in migraine patients. J Headache Pain 2007;8: 167–74.

[21]StaudR,RobinsonME,VierckJrCJ,PriceDD.Diffusenoxiousinhibitorycontrols (DNIC)attenuatetemporalsummationofsecondpaininnormalmalesbutnot innormalfemalesorfibromyalgiapatients.Pain2003;101:167–74.

[22]NormandE,PotvinS,GaumondI,CloutierG,CorbinJF,MarchandS.Pain inhi-bitionisdeficientinchronicwidespreadpainbutnormalinmajordepressive disorder.JClinPsychiatry2010;72:219–24.

[23]Wilder-SmithCH,SchindlerD,LovbladK,RedmondSM,NirkkoA.Brain func-tionalmagneticresonanceimagingofrectalpainandactivationofendogenous inhibitorymechanismsinirritablebowelsyndromepatientsubgroupsand healthycontrols.Gut2004;53:1595–601.

[24]KosekE,OrdebergG.Lackofpressurepainmodulationbyheterotopicnoxious conditioningstimulationinpatientswithpainfulosteoarthritisbefore,butnot following,surgicalpainrelief.Pain2000;88:69–78.

[25]Wilder-SmithOH,SchreyerT,SchefferGJ,Arendt-NielsenL.Patientswith chronicpainafterabdominalsurgeryshowlesspreoperativeendogenouspain inhibitionandmorepostoperativehyperalgesia:apilotstudy.JPainPalliatCare Pharmacother2010;24:119–28.

[26]MoontR,CrispelY,LevR,PudD,YarnitskyD.Temporalchangesincortical activationduringconditionedpainmodulation(CPM),aLORETAstudy.Pain 2011;152:1469–77.

[27]SongGH,VenkatramanV,HoKY,CheeMW,YeohKG,Wilder-SmithCH.Cortical effectsofanticipationandendogenousmodulationofvisceralpainassessedby functionalbrainMRIinirritablebowelsyndromepatientsandhealthycontrols. Pain2006;126:79–90.

[28]Peyron R, Laurent B, Garcia-Larrea L. Functional imaging of brain responsestopain.Areviewandmeta-analysis.NeurophysiolClin2000;30: 263–88.

[29]HarperRM,MaceyPM,HendersonLA,WooMA,MaceyKE,FrysingerRC,etal. fMRIresponsestocoldpressorchallengesincontrolandobstructivesleep apneasubjects.JApplPhysiol2003;94:1583–95.

[30]MaceyPM, MaceyKE, Woo MA,Keens TG, HarperRM. Aberrantneural responsestocoldpressorchallengesincongenitalcentralhypoventilation syn-drome.PediatrRes2005;57:500–9.

[31]PlaghkiL,DelisleD,GodfraindJM.Heterotopicnociceptiveconditioningstimuli andmentaltaskmodulatedifferentlytheperceptionandphysiological corre-latesofshortCO2laserstimuli.Pain1994;57:181–92.

[32]Bjorkedal E, Flaten MA. Expectations of increased and decreased pain explaintheeffectofconditionedpainmodulation infemales.J PainRes 2012;5:289–300.

[33]GoffauxP,RedmondWJ,RainvilleP,MarchandS.Descendinganalgesia—when thespineechoeswhatthebrainexpects.Pain2007;130:137–43.

[34]SpielbergerCD,GoruchRL,LushenePR,VaggPR,JacobsGA.Manualforthe state-traitanxietyinventory(formY).PaloAlto,CA:ConsultingPsychologists Press;1983,36p.

(12)

2002;16:217–40.

[39]AshburnerJ,FristonKJ.Unifiedsegmentation.Neuroimage2005;26:839–51.

[40]YarnitskyD,CrispelY,EisenbergE,GranovskyY,Ben-NunA,SprecherE,etal. Predictionofchronicpost-operativepain:pre-operativeDNICtestingidentifies patientsatrisk.Pain2008;138:22–8,.Epub2008Jan8.

[41]GrosenK,VaseL,PilegaardHK,Pfeiffer-JensenM,DrewesAM.Conditionedpain modulationandsituationalpaincatastrophizingaspreoperativepredictorsof painfollowingchestwallsurgery:aprospectiveobservationalcohortstudy. PLoSOne2014;9:2014.

[42]LindstedtF,BerrebiJ,GreayerE,LonsdorfTB,SchallingM,IngvarM,etal. Con-ditionedpainmodulationisassociatedwithcommonpolymorphismsinthe serotonintransportergene.PLoSOne2011;6:0018252.

[43]TreisterR,EisenbergE,GershonE,HaddadM,PudD.Factorsaffecting—and rela-tionshipsbetween—differentmodesofendogenouspainmodulationinhealthy volunteers.EurJPain2010;14:608–14.

[44]CormierS,PicheM,RainvilleP.Expectationsmodulateheterotopicnoxious counter-stimulationanalgesia.JPain2013;14:114–25.

[45]BingelU,WanigasekeraV,WiechK,NiMhuircheartaighR,LeeMC,Ploner M, etal. Theeffect oftreatment expectation ondrug efficacy: imaging the analgesic benefit ofthe opioid remifentanil. SciTransl Med 2011;3, 70ra14.

[46]YoshidaW,SeymourB,KoltzenburgM,DolanRJ.Uncertaintyincreasespain: evidenceforanovelmechanismofpainmodulationinvolvingthe periaque-ductalgray.JNeurosci2013;33:5638–46.

[47]Mouraux A, Diukova A, Lee MC, Wise RG, Iannetti GD. A multisensory investigationofthefunctionalsignificanceofthepainmatrix.Neuroimage 2011;54:2237–49.

[52]FairhurstM,FairhurstK,BernaC,TraceyI.AnfMRIstudyexploringtheoverlap anddifferencesbetweenneuralrepresentationsofphysicalandrecalledpain. PLoSOne2012;7:e48711.

[53]PomaresFB,FaillenotI,BarralFG,PeyronR.The‘where’andthe‘when’ofthe BOLDresponsetopainintheinsularcortex.Discussiononamplitudesand latencies.Neuroimage2013;64:466–75.

[54]IannettiGD,MourauxA.Fromtheneuromatrixtothepainmatrix(andback). ExpBrainRes2010;205:1–12.

[55]O’DohertyJ,KringelbachML,RollsET,HornakJ,AndrewsC.Abstractrewardand punishmentrepresentationsinthehumanorbitofrontalcortex.NatNeurosci 2001;4:95–102.

[56]Mobascher A,BrinkmeyerJ, WarbrickT,MussoF, WittsackHJ,Stoermer R, et al.Fluctuations inelectrodermal activityreveal variationsinsingle trialbrainresponsestopainfullaserstimuli—afMRI/EEGstudy.Neuroimage 2009;44:1081–92.

[57]WatsonA,El-DeredyW,IannettiGD,LloydD,TraceyI,VogtBA,etal.Placebo conditioningandplaceboanalgesiamodulateacommonbrainnetworkduring painanticipationandperception.Pain2009;145:24–30.

[58]KongJ,JensenK,LoiotileR,CheethamA,WeyHY,TanY,etal.Functional con-nectivityofthefrontoparietalnetworkpredictscognitivemodulationofpain. Pain2013;154:459–67.

[59]ElsenbruchS,KotsisV,BensonS,RosenbergerC,ReidickD,SchedlowskiM,etal. Neuralmechanismsmediatingtheeffectsofexpectationinvisceralplacebo analgesia:anfMRIstudyinhealthyplaceborespondersandnonresponders. Pain2012;153:382–90.

[60]BingelU,SchoellE,HerkenW,BuchelC,MayA.Habituationtopainful stimu-lationinvolvestheantinociceptivesystem.Pain2007;131:21–30.

Figure

Fig. 1. Experimental protocol. Continuous lukewarm +35 ◦ C (W) and cold 2 ± 2 ◦ C (C) stimulation is applied to the right foot of the subjects
Fig. 2. Sequential outline of the statistical analyses used in the study and definition of terms.
Fig. 3. Distribution of scores on a numeric rating scale (NRS, 0-10) for noxious pain and its modulation by thermal conditioning
Fig. 4. Cerebral responses to early (shaded areas with back dots) and sustained (shaded areas with white dots) cold-induced pain
+5

Références

Documents relatifs

This paper reports the e€ects of temperature and pH on extraction of sweeteners from dried leaves, pretreatment of the extract and subse- quently e€ects of temperature on separation

fig. car leurs côtés ont deux à deux la même longueur.. Retenons : Une translation est une isométrie qui conserve l’orientation. C’est donc un déplacement.. dans le

Mais notre capitaine souffrant d’une forte fièvre, nous ne pûmes quitter le cap avant la fin du mois de mars.. Nous repartîmes alors et notre voyage se déroula

The findings of the innocuously-based thermal grill studies are important in this field of research in the sense that the revelation of the influence of dispositional

Dans ce travail, nous avons montré que l’inhibition pharmacologique des PI3Ks prévient la translocation nucléaire de l’EGFR en réponse à ses ligands et

L’archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des

La structuration génétique des populations de Neritina stumpffi Boettger, 1890 dans l’Indo-Pacifique et celle de Neritina canalis Sowerby, 1825 dans l’océan Pacifique, ont

Chaque fois qu’il va à la pêche, il revient avec de très gros poissons qu’il partage avec ses voisins.. Ses qualités ont fait de lui une