A new approach to characterize postural deficits in ...

32
1 Preprint: Please note that this article has not completed peer review. A new approach to characterize postural deficits in chemotherapy-induced peripheral neuropathy and to analyze postural adaptions after an exercise intervention CURRENT STATUS: ACCEPTED Sarah Kneis Medical Center, University of Freiburg [email protected]Corresponding Author ORCiD: https://orcid.org/0000-0001-8350-8309 Anja Wehrle Medical Center, University of Freiburg Daniela Dalin Medical Center, University of Freiburg Isabella Katharina Wiesmeier Medical Center, University of Freiburg Johann Lambeck Medical Center, University of Freiburg Albert Gollhofer University of Freiburg Hartmut Bertz Medical Center, University of Freiburg Christoph Maurer Medical Center, University of Freiburg DOI: 10.21203/rs.2.16028/v1 SUBJECT AREAS Neurosurgery KEYWORDS

Transcript of A new approach to characterize postural deficits in ...

1

Preprint:Pleasenotethatthisarticlehasnotcompletedpeerreview.

Anewapproachtocharacterizeposturaldeficitsinchemotherapy-inducedperipheralneuropathyandtoanalyzeposturaladaptionsafteranexerciseinterventionCURRENTSTATUS:ACCEPTED

SarahKneisMedicalCenter,UniversityofFreiburg

[email protected]:https://orcid.org/0000-0001-8350-8309

AnjaWehrleMedicalCenter,UniversityofFreiburg

DanielaDalinMedicalCenter,UniversityofFreiburg

IsabellaKatharinaWiesmeierMedicalCenter,UniversityofFreiburg

JohannLambeckMedicalCenter,UniversityofFreiburg

AlbertGollhoferUniversityofFreiburg

HartmutBertzMedicalCenter,UniversityofFreiburg

ChristophMaurerMedicalCenter,UniversityofFreiburg

DOI:10.21203/rs.2.16028/v1

SUBJECTAREASNeurosurgery

KEYWORDS

2

Posturalstability,chemotherapy-inducedperipheralneuropathy,motorcontrol,sensoryweighting,model

3

AbstractBackgroundPosturalinstabilitypresentsacommonanddisablingconsequenceofchemotherapy-

inducedperipheralneuropathy(CIPN).However,knowledgeaboutposturalbehaviorofCIPNpatients

issparse.Withthispilotstudy,weusedanewapproachtoi)characterizeposturalimpairmentsas

comparedtohealthysubjects,ii)allocatepossibleabnormalitiestoasetofparametersdescribing

sensorimotorfunction,andiii)evaluatetheeffectsofabalance-basedexerciseintervention.

MethodsWeanalyzedspontaneousandexternallyperturbedposturalcontrolineightCIPNpatients

beforeandafterabalance-basedexerciseinterventionbyusingamodificationofanestablished

posturalcontrolmodel.Thesefindingswerecomparedto15matchedhealthysubjects.

ResultsSpontaneousswayamplitudeandvelocitywerelargerinCIPNpatientscomparedtohealthy

subjects.CIPNpatients’reactionstoexternalperturbationsweresmallercomparedtohealthy

subjects,indicatingthatpatientsfavorvestibularoverproprioceptivesensoryinformation.The

balance-basedexerciseinterventionup-weightedproprioceptiveinformationinpatients.

ConclusionsCIPNpatients’majorposturaldeficitmayrelatetounderuseofproprioceptiveinformation

thatresultsinalessaccurateposturecontrolasspontaneousswayresultsindicate.Thebalance-

basedexerciseinterventionisabletopartiallycorrectforthisabnormality.Ourstudycontributestoa

betterunderstandingofposturalimpairmentsinCIPNpatientsandsuggestsaneffectivetreatment

strategy.

1.IntroductionChemotherapy-inducedperipheralneuropathy(CIPN)isacommonandclinicallyrelevantside-effect

ofcancertreatment(1–3).CIPNcancausetreatmentdelaysanddosereductions,interferingwith

generaloutcomeorcompromisingsurvival(3–6).ConsequencesofCIPNcanleadtoexcessive

healthcarecostsandresourceuse(7).SymptomsofCIPNprimarilyincludeparaesthesia,dysesthesia,

numbnessandpainwithasymmetric,distal,length-dependent"gloveandstocking"distribution(3,8)

andlimitpatients’everydaylifeconsiderably.Additionally,CIPNpatientsoftensufferfrompostural

instability(9–17),contributingtoalowerqualityoflife(5,18),ahigherriskofmortality(19–22)and

increasedhealthcarecosts(23,24).

4

QuantitativereportsaboutCIPNpatients’posturalinstabilityarecurrentlyrising:CIPNhasbeen

associatedwithreducedgaitabilities(13,14)andchangesincenterofpressure(COP)displacements

(10,12,25).However,underlyingmechanismsaresparselydescribed(10,25):Wampleretal.(10)

assumedthatbesidessomatosensoryimpairmentsalsodiminishedvestibularfunctioncauses

increasedposturalswayinCIPNpatients.Furthermore,inanearlierstudywefoundchangesin

elicitabilityandsensitivityofspinalreflexcircuitryassociatedwithposturalinstabilityinCIPNpatients

(25).Morecomprehensiveknowledgeaboutneuropathy-inducedposturalinstabilityhasbeenderived

frompatientsdiagnosedwithdiabetes:Bonnetetal.(26)deducedlargerCOPdisplacements,which

weremorepronouncedwithvisualdisturbance.Diabeticneuropathypatientsseemtodelaypostural

reactions(27),shiftfromankletohipstrategy(26,28,29),andseemtousevestibularratherthan

proprioceptivecues(30).However,proprioceptionmaybeessentialforstabilityinbothquietstance

andduringunexpectedposturalperturbations(26,31–33),sinceitprovidesinformationaboutlower

limborientationwithrespecttothesupportbase(34,35).Theyreportaclearincreaseinpostural

swaywhenproprioceptivecuesaredeficient(10,26,32).Ourfirstaimhereistocharacterizethe

posturaldeficitsinCIPNandtoextractthesensorimotorabnormalitiesusingawell-establishedmodel

ofposturalcontrol(36–38).

Concerningtreatment,knowledgeaboutthemanagementofCIPN-inducedposturalinstabilityisstill

sparse(39–41).Generally,itisincreasinglysuggestedtofocusonstrengthandespeciallybalance

exercisesinordertoimprovephysicalfunctioningofCIPNpatients(11),whatwecouldconfirmina

randomizedcontrolledtrialbyourselves(42).Untilrecently,thereareonlytwootherinterventional

studiesshowingthatbalanceexercisesimprovedCIPN-relatedposturalcontroldeficits(39,41).

However,evidencefromdiabetesresearchonneuropathyfurthersupportsthisassumption(43–45).

Balancetrainingingeneralhasproventoenhanceposturalstabilitybyinducingneuronaladaptations

andimprovingmuscularoutput(46,47).Hence,weaimedtoimplementabalance-basedexercise

interventionforCIPNpatients.

Insum,thepresentstudywasundertakentoi)specifytheposturalabnormalitiesassociatedwith

CIPNduringspontaneousandexternallyperturbedstance,ii)toidentifytheunderlyingsensorimotor

5

malfunction,andiii)tomonitortheeffectofabalance-basedexerciseinterventioninapilotapproach.

2.Methods2.1.Patients

Thepresentpilotstudyprovidestwoapproaches:across-sectionalapproachtoidentifypostural-

controldifferencesbetweenCIPNpatientsandmatchedhealthycontrolsubjectsandaone-armed

longitudinalapproachtoevaluatetheeffectsofabalance-basedexerciseinterventiononCIPN-related

posturaldeficits.

Therefore,weexaminedeightcancerpatientswithdifferentcancerlocalizationsandtreatment

status,allreportingsevereneuropathysymptomsduetochemotherapy(CIPN).Thechemotherapies

appliedentailedtheneurotoxicagentsbortezomib,carboplatin,cisplatin,paclitaxal,docetaxaland

vincristine.Noneofthepatientshadanyneuropathysymptombeforetheapplicationofneurotoxic

agents.CIPNwasclinicallyandelectrophysiologicallyconfirmedinallpatients.Moreover,weassessed

patients’subjectiveCIPNsymptomsviatheneurotoxicitysubscale(NtxS)ofFACT&GOG(Functional

AssessmentofCancerTherapy/GynaecologyOncologyGroup)scoredfrom0–44(0=severe

symptoms;44=nosymptoms);Table1summarizesourpatients’clinicalinformation.

Weexcludedpatientswithotherpossiblesourcesofneuropathy(eghereditary,diabetes-oralcohol-

induced)andpatientssufferingfromadditionaldeficitsthatmightinteractwiththeirposturalcontrol

suchasarelevantreductionofmuscularstrengthorcertaincomorbidities(egosteolysis,severe

vertebraldegeneration,vestibulardeficits).Specifically,allpatientsunderwentdetailedvestibular

testingusingarotatingchair.Inaddition,patientsperformedanincrementalstresselectrocardiogram

onastationarybicycleintheInstituteforExercise-andOccupationalMedicine,MedicalCenter–

UniversityofFreiburginordertoexcludecardiovascularrisksduringexerciseandtodeterminethe

lactatethresholdforexercisecontrol.

Thecontrolgroupfortheposturalcontrolexperimentsconsistedof15healthysubjectsmatchedto

patients’age,weightandheight.Patientsunderwentassessmentsofposturecontroltwice(before

andafter12weeksofasupervisedexerciseintervention)whilehealthycontrolsubjectsunderwent

theassessmentonlyonce.

6

Patients'recruitmentanddatacollectiontookplaceintheClinicofInternalMedicineIandposture

analysesandclinicalassessmentstookplaceintheDepartmentofNeurologyandClinical

Neurophysiology,MedicalCenter–UniversityofFreiburg.

ThestudywasapprovedbytheEthicsCommissionofUniversityofFreiburg.Allsubjectsprovided

writteninformedconsenttotheexperimentalprocedureinaccordancewiththeDeclarationof

Helsinki.

2.2.Intervention

Theone-on-onetrainingsessionstookplaceinthedivisionofSportsOncologyintheClinicofInternal

MedicineI,twiceperweekover12weeks.Theinterventionprotocolincludedacardiovascularwarm-

upofupto20minutesonastationarybicyclewithanintensityof75–80%ofmaximumheartrate,

followedbythebalance-basedexercisesfor30minutesandmuscularendurancetrainingforthemain

musclegroups.Themainfocuswasonthebalancepartofthetraining.Balancetrainingprescription

includedaprogressiveincreaseovertheinterventionperiodintheexerciseamount(threetoeight

exerciseswiththreerepetitionseachà20–30s)anddifficultybyreducingthesupportsurfaceand

visualinput,addingmotor/cognitivetasksandinducinginstabilitytostimulatethesensorimotor

systemadequately(46,48).Wedocumentedvitalparameters,trainingprogress,andreasonsfor

missedsessions.

2.3.Procedureanddataanalysis

Forevaluatingposturalcontrol,spontaneousswayandperturbedstanceweremeasuredwitha

custom-builtmotionplatform(49,50)undertwovisualconditions,witheyesopenandwitheyes

closed.Eachtriallastedoneminute.Theparticipantsweretoldtostanduprightontheplatformin

comfortableshoes.Stancewidthwaspredeterminedwithinamarkedarea.Forsafetyreasons,

participantshadtoholdtworopeshangingfromtheceilinginacrossed-armspositionsothatthey

couldnotperceiveasomatosensoryspatialorientationsignal(Figure1D).

Dataanalysiswasconductedoff-linewithcustom-madesoftwareprogrammedinMATLAB®(The

MathWorksInc.,Natick,MA,USA).

Spontaneousswaywasmeasuredonthenon-movingplatform.Thecenterofpressure(COP)sway

7

pathwasdetectedwithaforcetransducingplatform(Figure1E-G,Kistlerplatformtype9286,

Winterthur,Switzerland).FromtheCOPexcursionsovertimeinanterior-posterior(ap)andmedio-

lateral(ml)swaydirections,wecalculatedtherootmeansquare(RMS)aroundthemeanCOP

position.Afterdifferentiatingthetimeseries,wecalculatedmeanvelocity(MV).Inaddition,center

frequency(CF)wasextractedfromthepowerspectrum(51,52).

Perturbedstancewasmeasuredonthemovingplatformtodifferentiatesensorycontributionsin

reactiontoexternaldisturbances.Weanalyzedrotationaltiltsinthesagittalplanewiththetiltaxis

passingthroughtheparticipant’sanklejoints.Platformrotationsweredesignedaspseudorandom

stimuli(PRTS,pseudorandomternarysequence,seeFigure1C)(53).ThePRTSstimulushasawide

spectralbandwidthwiththevelocitywaveformhavingspectralandstatisticalproperties

approximatingawhitenoisestimulus(53).Assuch,thisstimulusappearedtobeunpredictabletothe

testsubject.Weappliedtwopeakangulardisplacements(stimulusamplitude:0.5°and1°peak-to-

peak)andanalyzedatelevenstimulusfrequencies(0.05,0.15,0.3,0.4,0.55,0.7,0.9,1.1,1.35,1.75

and2.2Hz).

Angularexcursionsofthelower(hip-to-ankle:hipmovement)andupper(shoulder-to-hip:shoulder

movement)bodysegmentsandtheplatforminspaceweremeasuredusinganoptoelectronicmotion-

measuringdevicewithmarkersattachedtoshoulderandhip(Optotrak3020,Waterloo,Canada).

Eachmarkerconsistedofthreelight-emittingdiodes(LED)fixedtoarigidtriangle.Thetriangleswere

fixedtotheparticipant’shipsandshouldersandtoarigidbarontheplatform(Figure1D).3-DLED

positionsofthetriangleswereusedtocalculatemarkerpositions(Figure1A,B).Optotrak®and

Kistler®outputsignalsaswellasthestimulussignalsweresampledat100Hzusingananalogue-

digitalconverter.WerecordedalldatawithsoftwareprogrammedinLabView®(National

Instruments,Austin,Texas,USA).

Toanalyseposturalreactionsinrelationtoplatformstimuli,transferfunctionsfromstimulus-response

datawerecalculatedviaadiscreteFouriertransform.Fouriercoefficientsofstimulusandresponse

timeseriesareusedtodetermineGAINandPHASEwithrespecttostimulusfrequencies.GAIN

representsthesizeoftheposturalreactionasafunctionofstimulussize(platformangle),while

8

PHASEisrelatedtotherelativetimingbetweenposturalreactionandstimulus(54).

Furthermore,wecalculatedCOHERENCE,ameasureofreproducibilityoftheresponse.Whereasa

COHERENCEvalueofonereflectsperfectreproducibility,avalueofzeromeansnoreproducibilityat

all.

insertFigure1.

2.4.Parameteridentification

Transferfunctionsservedastheexperimentaldatabasisformodelsimulationsusingaspecific

versionofanestablishedposturalcontrolmodel(36,49,53,55–57)withactivetime-delayed

proportional,derivative,andintegralfeedbackaswellaspassivestiffnessanddampingtoextract

basicconstituentsofposturalcontrol.Thephysicalpartofthemodelisasingleinvertedpendulum

modelwithcorrectivetorqueappliedattheanklejoint.Themodelusedhereincludesanegative

feedbackloopthatrelatesbodyexcursiondetectedbyvisual,vestibular,andproprioceptivesensors

toacorrectivetorqueviaaneuralcontroller.Withthehelpofanautomatedoptimizationtool

(fmincon,MATLAB®,TheMathWorksInc.),whichminimizedthedifferencebetweenexperimentaland

simulatedGAINandPHASEcurves,weestimatedtheneuralcontroller’sparameterswithproportional

(Kp),derivative(Kd)andintegral(Ki)contributions(PDI-controller).Neuralcontrollergainsare,in

part,determinedbymassandheightofeachsubject’scenterofmass(COM).(53)Peterka,2002)

Becauseourcontrolgrouppresentedlowermassesandheightsthanpatients,wehadtocorrect

neuralcontrollergainsforthiseffect.Thatiswhyweprovidenumbersfor(Kp/mgh),(Kd/mgh),and

(Ki/mgh),where(mgh)representsthegravitationalpull(mass)x(gravitationalconstant)x(heightof

COM).Moreover,wederivedtimedelay(Td),proprioceptivesensoryweight(Wp),andbiomechanical

elasticity(Ppas)anddamping(Dpas)ofthemusclesandtendons.Wefittedmodelsimulationsto

experimentaltransferfunctionsunderdifferentstimulusamplitudesandvisualconditions.

2.5.Statistics

StatisticalanalyseswereperformedusingMicrosoftExcel,JMP®andStatview(SASInstituteInc.,

Cary,NC,USA).Weappliedparametricmethodsaftertestingthenormaldistributionand

homogeneityofvarianceswiththeKolmogorov-Smirnovtest.Duetotheexpecteddependency

9

betweenexperimentalconditionsandoutcomemeasures,statisticalsignificancewastestedbyan

analysisofvariance(ANOVA)forthecomparisonofhealthysubjectsandpatients.Visualcondition,

swaydirection,andbodysegment(hip,shoulder)werethewithin-subjects’factorsforspontaneous

sway.Forperturbedstance,weappliedvisualcondition,stimulusamplitude,stimulusfrequency,and

bodysegment(hip,shoulder)aswithin-subjects’factors.Fortheanalysisofthebalancebased

exerciseinterventioneffectonpatients,weusedamultivariateanalysisofvariance(MANOVA)witha

timeastherepeatedmeasurevariable,inaddition.Thelevelofstatisticalsignificancewassetat

p=0.05.

3.ResultsNoadverseeventswereobservedduringthestudyperiod.Theinterventioncomplianceintermsof

numberofsessionsperformedbythepatientswas70.1%,mainlyduetotheunderlyingdisease.

3.1.Spontaneoussway

ThepatientgroupbeforeinterventiondisplayedasignificantlylargerCOPRMSthancontrolsubjects

(Figures2A-BandTable2).Groupdesignationsignificantlyinteractedwithswaydirection,iethe

differencebetweencontrolsubjectsandCIPNpatientsislargerinapdirection.Moreover,group

designationsignificantlyinteractedwithvisualcondition,duetothelargeRMSinpatientswitheyes

closed.Afterintervention,thepatients’RMStendedtobesmaller(beforeintervention0.533cm;after

it0.501cm;F=2.98;p<0.088).

AswithRMS,thepre-interventionMVofthepatientgroupwassignificantlylargerthanincontrol

subjects(Figures2C-D,Table2).Thegroupdesignationsignificantlyinteractedwithvisualcondition

(seeTable2):MVvaluesdidnotdifferbetweengroupsintheeyes-opencondition,whereasthe

patients’MVwassignificantlylargerintheeyes-closedcondition.MVdidnotchangeafter

intervention.

CFdidnotdiffersignificantlybetweenpatientsandcontrolsubjects(Figures2E-F,Table2).

Furthermore,patientsdisplayednointerventioneffectsonCF.

[insertFigure2.]

3.2.Perturbedstance

10

ThetransferfunctionbetweenplatformtiltandbodyangulardisplacementischaracterizedbyGAIN

andPHASEbehavior.

Thedisturbance-inducedbodysway,ieGAINwassignificantlysmallerinpatientsbeforeintervention

(1.57)comparedtocontrolsubjects(1.87;F=62.3;p<0.0001;Fig.3A)andincreasedafter

intervention(1.63;F=18.0;p<0.0001;Fig.3A,Fig.4A-D).Furthermore,groupdesignationinteracted

significantlywithstimulusfrequency(F=3.70;p<0.0001),duetoadistortionofthetransferfunction

(Fig.3A,Fig.4A-D).Moreover,controlsubjects'GAINislargerwithclosedeyesthanopeneyes

whereaspatients’GAINwasalmostsimilarindependentofthevisualcondition:groupdesignation

significantlyinteractedwithvisualcondition(eyesopen:controlsubjects1.58;patientsbefore

intervention1.46;afterintervention1.51;eyesclosed:controlsubjects2.15;patientsbefore

intervention1.67;afterintervention1.74;visualcondition:F=25.6;p<0.0001,Fig.3D,Fig.4A-D).

Thedifferencebetweenshoulderandhipswayasafunctionofplatformtiltswasgreaterincontrol

subjectsthaninpatients(Fig.3E,Fig.4A-D):withasignificantinteractionbetweengroupdesignation

andbodysegment(F=2.85;p=0.022).Groupdesignationandstimulusamplitudedidnotinteract

significantlyastheeffectofstimulusamplitude(non-linearity)onGAINdidnotdifferbetweengroups.

ThePHASElaginthepatients’groupbeforeinterventionwasslightlybutsignificantlylesspronounced

thaninthecontrolgroup(controlsubjects-118.3;patients-107.6,F=10.3;p<0.0001;Fig.3B).After

intervention,PHASEchangedsignificantlyandfellintherangeofthecontrolsubjects’values(-121.3;

F=15.4;p<0.0001;Fig.3B).Groupdesignationsignificantlyinteractedwithvisualcondition(F=4.45,

p=0.01,Fig.3F),bodysegment(F=13.1,p<0.0001,Fig.3G),andstimulusamplitude(F=9.89,

p<0.0001).Theinteractionwithvisualconditionisbasedonthefactthatpatientswithopeneyes

displayedaPHASEadvanceof20degreeswithrespecttocontrolsubjects,whereastherewasno

significantPHASEdifferencebetweenpatientsandcontrolsubjectswitheyesclosed.Theinteraction

withbodysegmentisbasedonthefactthatthedifferencebetweenshoulderandhipPHASEwas

largerincontrolsubjectsthaninpatients.Theinteractionwithstimulusamplitudeisbasedona

pronouncedphasedifferencewithsmallstimulusamplitudes.Groupdesignationandstimulus

frequencydidnotinteractsignificantly(F=0.41;p=0.99):thePHASEeffectsweredistributedequally

11

acrossallfrequencies.

COHERENCEasameasureforthereproducibilityoftheresponsewassmallerinpatientsbefore

intervention(0.43)comparedtocontrolsubjects(0.50;F=103;p<0.0001;Fig.3C).COHERENCEdid

notchangeafterintervention(0.45).However,COHERENCEsignificantlyvariedwithstimulus

amplitude,frequency,visualcondition,andbodysegment,similarlyinbothgroups.

[insertFigure3and4.]

Thefollowingresultsarederivedfromthemodel-basedparameteridentificationprocedure(Fig.5D),

andpresenttherelevantparameterdifferencesbetweenpatientsandcontrolsubjects.

Therewasnosignificanteffectfortheintegralpartoftheneuralcontroller(Ki),aswellasforthe

proportional(Kp)andderivativepart(Kd)oftheneuralcontroller.

ThesensoryweightingfactorWpdifferedsignificantlybetweenpatientsandcontrolsubjects(F=9.89,

p=0.0001;Fig.5B).Whereaspatientsrelywithanaveragefactorof0.53onproprioceptivecuesand

hence0.47onspatialcues,controlsubjectsrelywithafactorof0.67onproprioceptiveand0.33on

spatialcues.Afterintervention,therewasasmallbutsignificantchangetowardscontrolsubjects(Wp

0.56,F=9.13,p=0.006).Groupdesignationdidnotinteractwithvisualconditionorstimulus

amplitude.

Thetimedelaybetweenstimulusandresponse(Td)didnotdiffersignificantlybetweenpatientsand

controlsubjects(F=1.10;p=0.34;Fig.5C).However,wedidnoteatendencytowardslargertime

delaysinpatientsbeforeintervention(175ms)comparedtocontrolsubjects(167ms)andpatients

afterintervention(168ms).

Parametersrelatedtopassivemuscleandtendonbehavior(PpasandDpas)didnotdiffersignificantly

betweengroupsandwerenotaffectedbytheintervention.

[insertFigure5.]

4.DiscussionAsposturalinstabilityisamomentoussymptomofCIPN(9–16),thefirstaimofthisstudywasto

assessthespecificsetofposturalcontroldeficitsassociatedwithCIPNcomparedtohealthysubjects.

Furthermore,sinceCIPNtreatmentoptionsareverylimitedsofar(3)andhintsintherecentliterature

12

indicatethatCIPNpatientsmightbenefitfromexercising(39,41,58),weevaluatedabalance-based

exerciseinterventionaimingtotreatpatients'functionalimpairmentsduetoCIPN.Whileformer

studiesmostlyinvestigatedspontaneousswaymeasures(displacement-,velocity-,andfrequency-

relatedmeasures),weaimedtodescribeCIPNpatients'sensorimotorbehaviorinmuchgreaterdetail

andtherefore,weadditionallygeneratedtransferfunctionsofpatients’behaviorasafunctionof

stanceperturbationwithpseudorandomstimulitoachieveourobjectives.

Concerningspontaneoussway,wefoundgreaterposturalswayinCIPNpatientssimilarlytoprevious

CIPNstudies(10,12,14,39).Additionally,ourfindingscorrespondtothatofothertypesofneuropathy.

Forexample,manyworkinggroups(59–63)reportincreasedRMSandMVinpatientswithdiabetic-

inducedneuropathy.Inourstudy,RMSandMVweresignificantlylargerinCIPNpatientsthanin

healthysubjects.Moreover,weobservedaspecificpreponderanceofdeficitsinanterior-posterior

direction(60)andamorepronouncedposturalswaywithclosedratherthanopeneyes(60,61).

Generally,closingtheireyescausessubjectstousevestibularandproprioceptivecuesforcontrolling

balance.Asproprioceptiveinformationareoftendeficientinneuropathypatients(26),itseems

reasonabletoassumethatpatientsmayprefervestibularoverproprioceptivecues.However,itiswell

knownthatthevestibularsignalcarriesalargeramountofnoisethantheproprioceptivesignal(64)

leadingtolessaccurateposturecontrol.Wespeculateatthispointthatthemainsourceforthelarger

RMSandMVisrelatedtoasensoryshifttowardsvestibularcues(seebelowsensoryweighting,and

Horaketal.(65)).Ifthatistrue,wemightbeabletorecoverthisfindingwhendissociating

proprioceptivefromvestibularframesoforientationusingplatformtilts.Whereasproprioceptivecues

maydragthebodyalongplatformmovements,quantifiedbyarelativelargerGAINastransfer

functionbetweenbodyexcursionsandplatformtilts,vestibularcueswouldstabilizethebodyin

space,quantifiedbyarelativesmallerGAIN.Infact,CIPNpatientspresentedsmallerGAINvalues

sincetheirreactiontoplatformtiltswerelesspronouncedthanthatofcontrolsubjects.Thus,they

mightratherusespacecoordinatesthanplatformmovementsforposturecontrol.Furthermore,GAIN

wassignificantlyaffectedbyvisualconditionandbodysegment.ThelargerGAINdifferencebetween

CIPNpatientsandcontrolsubjectsintheeyes-closedcomparedtotheeyes-openconditionsuggests

13

thatunder-usageofproprioceptionisdominantwhentherearelessadditionalorientingcues.

Moreover,thefindingofrelativelysmallGAINsinCIPNpatients’lowercomparedtotheirupperbody

segmentpointtoaslightlydifferentintersegmentalstrategy(29).OurPHASEfindings,representing

therelativetimingbetweenposturalreactionandstimulusalsopointtoadifferentintersegmental

behaviorintermsofupperwithrespecttolowerbodyangulardisplacements(26,28,29).Weassume

thatpatientsproactivelyorientatethemselves,especiallytheirupperbody,moretowardsspace

coordinates.ThisalsoindicatesanespeciallylowuseofproprioceptionaccordingtoourGAINresults.

Moreover,wespeculatethatCIPNpatientsproactivelyassumeasafetystrategythatmayfollowan

enhancedmuscleco-contraction(25),leadingtosmallerbodyexcursions.However,greaterco-

contractionlimitsone’sabilitytopreciselycontrolposture(66,67).

Toaddressthetransfer-functionabnormalitiesinCIPNpatients,wefittedthesubjects’dataviaa

simplefeedbacksystem(36,56,64,68).Usingthemodel-basedparameterestimation,weidentified

andquantifiedtheCIPNpatients’diminisheduseofproprioceptivecues:Thesensory-weightingfactor

forproprioception(Wp)issignificantlysmallerinCIPNpatientsthanincontrolsubjects.However,

patientsdidnotpresentadifferenterrorcorrectiongain(KpandKd)ofthefeedbackloop.

Furthermore,parametersrelatedtopassivemuscleandtendonbehavior(passivestiffnessand

damping,KpasandBpas)didalsonotdifferbetweenCIPNpatientsandcontrolsubjects.Thisseems

tobeinlinewiththenotionthatdifferencesinposturalcontrolbetweenCIPNpatientsandhealthy

subjectsmainlyrelyonthedifferentuseofsensorycues.Moreover,thisfindingmightindicatethat

musclesandtendonswerenotaffectedsupportingCIPN'sprimarilysensorycharacteristic.

HowdoestheexerciseinterventioninfluenceCIPNpatients’behavior?Interestingly,weobservedthat

CIPNpatients’mainabnormality(down-weightedproprioception)wasmodifiedbyexercise

intervention.CIPNpatients’GAINandPHASEvaluesapproachedthoseofhealthysubjectsduetothe

proprioceptiveup-weightingmentionedabove.Interestingly,theeffectofup-weightingproprioception

istheonlysignificantmodel-basedparametermodificationafterintervention.WhywouldCIPN

patientsprofitfromup-weightingproprioceptionwhilesufferingfromasupposedproprioceptive

deficit?Ourclinicalassessmentsdidnotsufficetoconclusivelyspecifyneurallesionsoridentify

14

CIPN'snervefibercontributionoccurringinourpatients.However,allpatientssufferedfromstrong

paresthesiaandreportedsignificantbalanceproblemsconfirmedbyourspontaneoussway

experiments.CIPNsensorysymptomsaredescribedtoreferto‘terminalarbordegeneration’andthe

preferentialdamageofmyelinatedprimaryafferentsensorynervefibers(2,69).Therefore,we

speculatethatpatients'peripheralinformationandsubsequentstimulusconductionarealtered,but

notcompletelydysfunctional.Thedown-weightingofproprioceptivecuescouldbeinterpretedasan

excessivecompensatorymechanism,whichletsCIPNpatientspre-interventionremaininasuboptimal

state.Theexerciseinterventionmaythuspartiallycorrectthisexcessandmaystimulatetheuseof

lessdamagedpathways.Asasideeffectofproprioceptiveup-weighting,theinterventionmaytrigger

down-weightingofvestibularcues,therebyreducingvestibularnoise.Asaconsequence,postural-

controlbehaviormightbemoreaccurateintermsoflessvariability.

Conclusively,wemaintainthatup-weightingproprioceptionandtherebydown-weightingvestibular

informationtowardsthebehaviorofhealthysubjectsrepresentsaclearbenefitforCIPNpatients.The

proprioceptivecueisconsideredtobemoreaccuratethanvestibularcuesforposturalstability(64).

Regardingtimedelay,patientsafterinterventiontendedtoimprovetheirreactiontimebetween

stimulusandresponse.WealsoidentifiedatendencyofspontaneousswayRMSvaluestobesmaller

afterintervention,beinginlinewithposturalswayfindingsafterbalancetraininginastudyof

Schwenketal.(39).Although,ourinterventioneffectsaresmall,weareconvincedthatCIPNpatients

benefitfromexercisingintermsofimprovedposturebehavoirthatsubstantiallycontributesto

patients'functionalstatus.Functionalperformanceisanacknowledgedprognosisfactorforcancer

survivor(70)whywestronglyproposetoverifyourinterventionresultsinagreaterrandomized

controlledtrial.Ourfindingsarelimitedbythesmallsamplesizeandthelackofpatientcontrol

group.Furthermore,expandingneurophysiologicalassessmentswouldprovideinsightsinadaptive

processesunexaminedinthisstudy.

5.ConclusionWebelievethatournewapproachcontributedtoadeeperunderstandingofCIPNpatients’postural

instability.Proprioceptivedown-weightingmightrepresentthemainposturaldeficitinCIPN.Our

15

exerciseinterventiontargetedspecificallythisabnormalitypresumablybyprimarilycorrectingthe

overactivecompensation,whichledtoasignificantimprovementinposturalstability.Webelievethat

abalance-basedexerciseinterventionisapromisingstrategytomanagefunctionalimpairmentsdue

toCIPNandthatitshouldthereforeberoutinelyintegratedwithinthetreatmentregimensofpatients

receivingneurotoxicagents.

6.Abbreviationsapanterior-posterior

CFcenterfrequency

CIPNchemotherapy-inducedperipheralneuropathy

COPcenterofpressure

Dpaspassivedamping

Kdderivativecontributionoftheneuralcontroller

Kiintegralcontributionoftheneuralcontroller

Kpproportionalcontributionoftheneuralcontroller

NtxSneurotoxicitysubscaleofFACT&GOG

mlmedio-lateral

MVmeanvelocity

Ppaspassivestiffness

RMSrootmeanssquare

Tdtimedelay

Wpproprioceptivesensoryweight

7.Declarations7.1.Ethicsapprovalandconsenttoparticipate

ThisstudywasapprovedbytheEthicsCommitteeoftheUniversityofFreiburg(478/11).All

proceduresperformedinstudiesinvolvinghumanparticipantswereinaccordancewiththeethical

standardsoftheinstitutionaland/ornationalresearchcommitteeandwiththe1964Helsinki

declarationanditslateramendmentsorcomparableethicalstandards.Writteninformedconsentwas

16

obtainedfromallindividualparticipantsincludedinthestudy.

7.2.Consentforpublication

Notapplicabl

7.3.Availabilityofdataandmaterial

Thedatasetsupportingtheconclusionsofthisarticleisincludedwithinthisarticle.Thedatathat

supportthefindingsofthisstudyareavailablefromthecorrespondingauthoruponreasonable

request.

7.4.Competinginterests

Theauthorsdeclarethattheresearchwasconductedintheabsenceofanycommercialorfinancial

relationshipsthatcouldbeconstruedasapotentialconflictofinterest.

7.5.Funding

SK,AWandCMcontributedtotheconceptionofthework,todataacquisition,analysisand

interpretation,wrotethefirstdraftandrevisedit.IW,DD,andJLcontributedtodataacquisitionand

analysisandrevisedtheworkcritically.AGandHBcontributedtothedesignoftheworkandrevised

itcritically.Allauthorsapprovedthefinalversionofthisarticleandagreetobeaccountableforall

aspectsoftheworkinensuringthatquestionsrelatedtotheaccuracyorintegrityofanypartofthe

workareappropriatelyinvestigatedandresolved.

7.6.Authors'contributions

SK,AWandCMcontributedtotheconceptionofthework,todataacquisition,analysisand

interpretation,wrotethefirstdraftandrevisedit.IW,DD,andJLcontributedtodataacquisitionand

analysisandrevisedtheworkcritically.AGandHBcontributedtothedesignoftheworkandrevised

itcritically.Allauthorsapprovedthefinalversionofthisarticleandagreetobeaccountableforall

aspectsoftheworkinensuringthatquestionsrelatedtotheaccuracyorintegrityofanypartofthe

workareappropriatelyinvestigatedandresolved.

7.7.Acknowledgements

Weacknowledgethecooperationandtrainingimplementationofsports-andphysiotherapistsofthe

SportsOncologyofDepartmentofMedicineI,MedicalCenter–UniversityofFreiburgandwethank

17

thepatientsandcontrolindividualsfortheircollaboration.

8.References1.ArgyriouAA,KyritsisAP,MakatsorisT,KalofonosHP.Chemotherapy-inducedperipheralneuropathy

inadults:acomprehensiveupdateoftheliterature.CancerManagRes.2014;6:135–47.

2.HanY,SmithMT.Pathobiologyofcancerchemotherapy-inducedperipheralneuropathy(CIPN).

FrontPharmacol.2013Dec18;4:156.

3.HershmanDL,LacchettiC,DworkinRH,SmithEML,BleekerJ,CavalettiG,etal.Preventionand

ManagementofChemotherapy-InducedPeripheralNeuropathyinSurvivorsofAdultCancers:

AmericanSocietyofClinicalOncologyClinicalPracticeGuideline.JClinOncol.2014Jun

20;32(18):1941–67.

4.Gutiérrez-GutiérrezG,SerenoM,MirallesA,Casado-SáenzE,Gutiérrez-RivasE.Chemotherapy-

inducedperipheralneuropathy:clinicalfeatures,diagnosis,preventionandtreatmentstrategies.Clin

TranslOncolOffPublFedSpanOncolSocNatlCancerInstMex.2010Feb;12(2):81–91.

5.StubblefieldMD,BursteinHJ,BurtonAW,CustodioCM,DengGE,HoM,etal.NCCNtaskforce

report:managementofneuropathyincancer.JNatlComprCancerNetwJNCCN.2009Sep;7Suppl

5:S1–26;quizS27-28.

6.WindebankAJ,GrisoldW.Chemotherapy-inducedneuropathy.JPeripherNervSyst.2008;13(1):27–

46.

7.PikeCT,BirnbaumHG,MuehlenbeinCE,PohlGM,NataleRB.Healthcarecostsandworklossburden

ofpatientswithchemotherapy-associatedperipheralneuropathyinbreast,ovarian,headandneck,

andnonsmallcelllungcancer.ChemotherResPract.2012;2012:913848.

8.AsburyA,ThomasP.PeripheralNerveDisorders2.Butterworth-Heinemann;1995.336p.

9.TofthagenC,OvercashJ,KipK.Fallsinpersonswithchemotherapy-inducedperipheralneuropathy.

SupportCareCancer.2012;20(3):583–9.

10.WamplerMA,ToppKS,MiaskowskiC,BylNN,RugoHS,HamelK.QuantitativeandClinical

DescriptionofPosturalInstabilityinWomenWithBreastCancerTreatedWithTaxaneChemotherapy.

ArchPhysMedRehabil.2007Aug;88(8):1002–8.

18

11.Winters-StoneKM,HorakF,JacobsPG,TrubowitzP,DieckmannNF,StoylesS,etal.Falls,

Functioning,andDisabilityAmongWomenWithPersistentSymptomsofChemotherapy-Induced

PeripheralNeuropathy.JClinOncol.2017Aug10;35(23):2604–12.

12.HermanHK,MonfortSM,PanXJ,ChaudhariAMW,LustbergMB.Effectofchemotherapy-induced

peripheralneuropathyonposturalcontrolincancersurvivors.JClinOncol.2017Feb

10;35(5_suppl):128–128.

13.MarshallTF,ZippGP,BattagliaF,MossR,BryanS.Chemotherapy-induced-peripheralneuropathy,

gaitandfallriskinolderadultsfollowingcancertreatment.JCancerResPract.2017Dec1;4(4):134–

8.

14.MonfortSM,PanX,PatrickR,RamaswamyB,WesolowskiR,NaughtonMJ,etal.Gait,balance,and

patient-reportedoutcomesduringtaxane-basedchemotherapyinearly-stagebreastcancerpatients.

BreastCancerResTreat.2017Jul1;164(1):69–77.

15.GewandterJS,FanL,MagnusonA,MustianK,PepponeL,HecklerC,etal.Fallsandfunctional

impairmentsincancersurvivorswithchemotherapy-inducedperipheralneuropathy(CIPN):a

UniversityofRochesterCCOPstudy.SupportCareCancer.2013Jul;21(7):2059–66.

16.BaoT,BasalC,SeluzickiC,LiSQ,SeidmanAD,MaoJJ.Long-termchemotherapy-induced

peripheralneuropathyamongbreastcancersurvivors:prevalence,riskfactors,andfallrisk.Breast

CancerResTreat.2016Sep;159(2):327–33.

17.KolbNA,SmithAG,SingletonJR,BeckSL,StoddardGJ,BrownS,etal.TheAssociationof

Chemotherapy-InducedPeripheralNeuropathySymptomsandtheRiskofFalling.JAMANeurol.2016

Jul1;73(7):860–6.

18.MolsF,BeijersT,VreugdenhilG,vandePoll-FranseL.Chemotherapy-inducedperipheral

neuropathyanditsassociationwithqualityoflife:asystematicreview.SupportCareCancer.2014

Aug;22(8):2261–9.

19.CesariM,KritchevskySB,NewmanAB,SimonsickEM,HarrisTB,PenninxBW,etal.Addedvalueof

physicalperformancemeasuresinpredictingadversehealth-relatedevents:resultsfromtheHealth,

AgingAndBodyCompositionStudy.JAmGeriatrSoc.2009;57(2):251–9.

19

20.CesariM,CerulloF,ZamboniV,PalmaRD,ScambiaG,BalducciL,etal.FunctionalStatusand

MortalityinOlderWomenWithGynecologicalCancer.JGerontolABiolSciMedSci.2013;68(9):1129–

33.

21.KlepinHD,GeigerAM,ToozeJA,NewmanAB,ColbertLH,BauerDC,etal.Physicalperformance

andsubsequentdisabilityandsurvivalinolderadultswithmalignancy:resultsfromthehealth,aging

andbodycompositionstudy.JAmGeriatrSoc.2010Jan;58(1):76–82.

22.OnderG,PenninxBWJH,FerrucciL,FriedLP,GuralnikJM,PahorM.Measuresofphysical

performanceandriskforprogressiveandcatastrophicdisability:resultsfromtheWomen’sHealth

andAgingStudy.JGerontolABiolSciMedSci.2005Jan;60(1):74–9.

23.AleknaV,StukasR,Tamulaitytė-MorozovienėI,ŠurkienėG,TamulaitienėM.Self-reported

consequencesandhealthcarecostsoffallsamongelderlywomen.MedKaunasLith.2015;51(1):57–

62.

24.AlexanderBH,RivaraFP,WolfME.Thecostandfrequencyofhospitalizationforfall-related

injuriesinolderadults.AmJPublicHealth.1992;82(7):1020–3.

25.KneisS,WehrleA,FreylerK,LehmannK,RudolphiB,HildenbrandB,etal.Balanceimpairments

andneuromuscularchangesinbreastcancerpatientswithchemotherapy-inducedperipheral

neuropathy.ClinNeurophysiol.2016Feb;127(2):1481–90.

26.BonnetC,CarelloC,TurveyMT.DiabetesandPosturalStability:ReviewandHypotheses.JMot

Behav.2009Mar;41:172–92.

27.DicksteinR,PeterkaRJ,HorakFB.Effectsoflightfingertiptouchonposturalresponsesinsubjects

withdiabeticneuropathy.JNeurolNeurosurgPsychiatry.2003Jan5;74(5):620–6.

28.DicksteinR,ShupertCL,HorakFB.Fingertiptouchimprovesposturalstabilityinpatientswith

peripheralneuropathy.GaitPosture.2001Dec;14(3):238–47.

29.SimmonsRW,RichardsonC,PozosR.Posturalstabilityofdiabeticpatientswithandwithout

cutaneoussensorydeficitinthefoot.DiabetesResClinPract.1997Jun;36(3):153–60.

30.HorakFB,HlavackaF.Somatosensorylossincreasesvestibulospinalsensitivity.JNeurophysiol.

2001Aug;86(2):575–85.

20

31.FitzpatrickR,McCloskeyDI.Proprioceptive,visualandvestibularthresholdsfortheperceptionof

swayduringstandinginhumans.JPhysiol.1994Jul1;478(Pt1):173–86.

32.NashnerLM,BlackFO,WallC.Adaptationtoalteredsupportandvisualconditionsduringstance:

patientswithvestibulardeficits.JNeurosci.1982May;2(5):536–44.

33.PeterkaRJ,BenolkenMS.Roleofsomatosensoryandvestibularcuesinattenuatingvisually

inducedhumanposturalsway.ExpBrainRes.1995;105(1):101–10.

34.CornaS,TarantolaJ,NardoneA,GiordanoA,SchieppatiM.Standingonacontinuouslymoving

platform:isbodyinertiacounteractedorexploited?ExpBrainRes.1999Feb;124(3):331–41.

35.SchieppatiM,GiordanoA,NardoneA.Variabilityinadynamicposturaltaskattestsample

flexibilityinbalancecontrolmechanisms.ExpBrainRes.2002May;144(2):200–10.

36.EngelhartD,PasmaJH,SchoutenAC,MeskersCGM,MaierAB,MergnerT,etal.Impairedstanding

balanceinelderly:anewengineeringmethodhelpstounravelcausesandeffects.JAmMedDir

Assoc.2014Mar;15(3):227.e1-6.

37.PasmaJH,EngelhartD,MaierAB,SchoutenAC,vanderKooijH,MeskersCGM.Changesinsensory

reweightingofproprioceptiveinformationduringstandingbalancewithageanddisease.J

Neurophysiol.2015Dec;114(6):3220–33.

38.WiesmeierIK,DalinD,MaurerC.ElderlyUseProprioceptionRatherthanVisualandVestibular

CuesforPosturalMotorControl.FrontAgingNeurosci.2015;7:97.

39.SchwenkM,GrewalGS,HollowayD,MuchnaA,GarlandL,NajafiB.InteractiveSensor-Based

BalanceTraininginOlderCancerPatientswithChemotherapy-InducedPeripheralNeuropathy:A

RandomizedControlledTrial.Gerontology.2016;62(5):553–63.

40.StreckmannF,ZopfEM,LehmannHC,MayK,RizzaJ,ZimmerP,etal.ExerciseIntervention

StudiesinPatientswithPeripheralNeuropathy:ASystematicReview.SportsMed.2014Jun14;44(9).

41.ZimmerP,TrebingS,Timmers-TrebingU,SchenkA,PaustR,BlochW,etal.Eight-week,

multimodalexercisecounteractsaprogressofchemotherapy-inducedperipheralneuropathyand

improvesbalanceandstrengthinmetastasizedcolorectalcancerpatients:arandomizedcontrolled

trial.SupportCareCancer.2017Sep30;26(2):615–24.

21

42.KneisS,WehrleA,MüllerJ,MaurerC,IhorstG,GollhoferA,etal.It’snevertoolate-balanceand

endurancetrainingimprovesfunctionalperformance,qualityoflife,andalleviatesneuropathic

symptomsincancersurvivorssufferingfromchemotherapy-inducedperipheralneuropathy:resultsof

arandomizedcontrolledtrial.BMCCancer.2019May2;19(1):414.

43.AlletL,ArmandS,deBieRA,GolayA,MonninD,AminianK,etal.Thegaitandbalanceofpatients

withdiabetescanbeimproved:arandomisedcontrolledtrial.Diabetologia.2010Mar;53(3):458–66.

44.LeeK,LeeS,SongC.Whole-bodyvibrationtrainingimprovesbalance,musclestrengthand

glycosylatedhemoglobininelderlypatientswithdiabeticneuropathy.TohokuJExpMed.

2013;231(4):305–14.

45.SongCH,PetrofskyJS,LeeSW,LeeKJ,YimJE.Effectsofanexerciseprogramonbalanceand

trunkproprioceptioninolderadultswithdiabeticneuropathies.DiabetesTechnolTher.2011

Aug;13(8):803–11.

46.TaubeW,GruberM,GollhoferA.Spinalandsupraspinaladaptationsassociatedwithbalance

trainingandtheirfunctionalrelevance.ActaPhysiolOxf.2008Jun;193(2):101–16.

47.ZechA,HübscherM,VogtL,BanzerW,HänselF,PfeiferK.Balancetrainingforneuromuscular

controlandperformanceenhancement:asystematicreview.JAthlTrain.2010Aug;45(4):392–403.

48.GranacherU,MuehlbauerT,TaubeW,GollhoferA,GruberM.SensorimotorTraining.In:Strength

andconditioning:biologicalprinciplesandpracticalapplications.SanFrancisco:Wiley-Blackwell;

2011.p.399.

49.CnyrimC,MergnerT,MaurerC.Potentialrolesofforcecuesinhumanstancecontrol.ExpBrain

Res.2009Feb14;194(3):419–33.

50.MaurerC,MergnerT,PeterkaRJ.Multisensorycontrolofhumanuprightstance.ExpBrainRes.

2006May;171(2):231–50.

51.MaurerC,PeterkaRJ.Anewinterpretationofspontaneousswaymeasuresbasedonasimple

modelofhumanposturalcontrol.JNeurophysiol.2005Jan;93(1):189–200.

52.PrietoTE,MyklebustJB,HoffmannRG,LovettEG,MyklebustBM.Measuresofposturalsteadiness:

differencesbetweenhealthyyoungandelderlyadults.IEEETransBiomedEng.1996Sep;43(9):956–

22

66.

53.PeterkaRJ.Sensorimotorintegrationinhumanposturalcontrol.JNeurophysiol.2002

Sep;88(3):1097–118.

54.WiesmeierIK,DalinD,WehrleA,GranacherU,MuehlbauerT,DietterleJ,etal.BalanceTraining

EnhancesVestibularFunctionandReducesOveractiveProprioceptiveFeedbackinElderly.FrontAging

Neurosci.2017;9:273.

55.KooijHvander,JacobsR,KoopmanB,HelmFvander.Anadaptivemodelofsensoryintegration

inadynamicenvironmentappliedtohumanstancecontrol.BiolCybern.2001Jan1;84(2):103–15.

56.MergnerT,MaurerC,PeterkaRJ.Sensorycontributionstothecontrolofstance:aposturecontrol

model.AdvExpMedBiol.2002;508:147–52.

57.WelchTDJ,TingLH.AFeedbackModelExplainstheDifferentialScalingofHumanPostural

ResponsestoPerturbationAccelerationandVelocity.JNeurophysiol.2009Jun;101(6):3294–309.

58.StreckmannF,KneisS,LeifertJA,BaumannFT,KleberM,IhorstG,etal.Exerciseprogram

improvestherapy-relatedside-effectsandqualityoflifeinlymphomapatientsundergoingtherapy.

AnnOncol.2014Jan2;25(2):493–9.

59.BoucherP,TeasdaleN,CourtemancheR,BardC,FleuryM.Posturalstabilityindiabetic

polyneuropathy.DiabetesCare.1995May;18(5):638–45.

60.CorriveauH,PrinceF,HébertR,RaîcheM,TessierD,MaheuxP,etal.EvaluationofPostural

StabilityinElderlywithDiabeticNeuropathy.DiabetesCare.2000Jan8;23(8):1187–91.

61.DiNardoW,GhirlandaG,CerconeS,PitoccoD,SoponaraC,CosenzaA,etal.Theuseofdynamic

posturographytodetectneurosensorialdisorderinIDDMwithoutclinicalneuropathy.JDiabetes

Complications.1999Apr;13(2):79–85.

62.SimoneauGG,UlbrechtJS,DerrJA,BeckerMB,CavanaghPR.Posturalinstabilityinpatientswith

diabeticsensoryneuropathy.DiabetesCare.1994Dec;17(12):1411–21.

63.UccioliL,GiacominiPG,MonticoneG,MagriniA,DurolaL,BrunoE,etal.Bodyswayindiabetic

neuropathy.DiabetesCare.1995Mar;18(3):339–44.

64.vanderKooijH,PeterkaRJ.Non-linearstimulus-responsebehaviorofthehumanstancecontrol

23

systemispredictedbyoptimizationofasystemwithsensoryandmotornoise.JComputNeurosci.

2011Jun;30(3):759–78.

65.HorakFB,DicksteinR,PeterkaRJ.Diabeticneuropathyandsurfacesway-referencingdisrupt

somatosensoryinformationforposturalstabilityinstance.SomatosensMotRes.2002Jan

1;19(4):316–26.

66.TuckerMG,KavanaghJJ,BarrettRS,MorrisonS.Age-relateddifferencesinposturalreactiontime

andcoordinationduringvoluntaryswaymovements.HumMovSci.2008Oct;27(5):728–37.

67.AllumJHJ,CarpenterMG,HoneggerF,AdkinAL,BloemBR.Age-dependentvariationsinthe

directionalsensitivityofbalancecorrectionsandcompensatoryarmmovementsinman.JPhysiol.

2002Jul15;542(Pt2):643–63.

68.MergnerT,MaurerC,PeterkaRJ.Amultisensoryposturecontrolmodelofhumanuprightstance.

ProgBrainRes.2003;142:189–201.

69.CataJP,WengHR,LeeBN,ReubenJM,DoughertyPM.Clinicalandexperimentalfindingsin

humansandanimalswithchemotherapy-inducedperipheralneuropathy.MinervaAnestesiol.2006

Mar;72(3):151–69.

70.BrownJC,HarhayMO,HarhayMN.Physicalfunctionasaprognosticbiomarkeramongcancer

survivors.BrJCancer.2015;112(1):194–8.

TablesTable1Subjects'characteristic

24

Pat Sex Height(cm)

Weight(kg)

Age(years)

Diagnosis Status MonthssinceED(n)

Cyclesofneurotoxicdrugs(n)

1 m 182 125 55 B-NHL PR 185 14

2 m 182 80 67 MM SD 58 34

3 f 161 86 60 MM Relapse 82 7

4 f 162 60 68 Breastcancer SD 66 19

5 m 180 80 64 MM SD 8 4

6 m 167 92 63 Breastcancer SD 195 25

7 m 176 64 67 Rectalcancer PD 118 79

8 m 188 107 42 Germcellcancer CR 66 4mean±SD

m:f6:2

174.8±10.1

86.8±21.5

60.8±8.7

97.3±64.8

23.3±24.9

hCon mean±SD

m:f9:6

17.0±9.1

78.0±6.3

59.6±10.0

Pat,CIPNpatients;SD,standarddeviation;hCon,healthycontrolsubjects;m,male;f,female;B-NHL,

B-NonHodgkinLymphoma;MM,multiplemyeloma;PR,partialremission;SD,stabledisease;PD,

progressivedisease;CR,completeremission;NCV,nerveconductionvelocity(normalvalues:

≥41m/s);APA,actionpotentialamplitude(normalvalues:≥6µV);ATR,Achillestendonreflex;PRT,

Patellatendonreflex;"-",noreflex;"(-)",reducedreflex;"+",normalreflex;NtxS,neurotoxicity

subscaleofFACT/GOG(FunctionalAssessmentofCancerTherapy/GynaecologyOncologyGroup)

scoredfrom0(severesymptoms)–44(nosymptoms)

Table2Spontaneousswaymeasures(RMS,MV,CF)withgroupeffects,andinteractionsbetween

groupsandvisualconditions/swaydirections.

25

hControl PATpre PATpost F-value p-value RMS(cm) 0.46±0.1810.926 <0.000

10.56±0.27

visualcondition eyesopeneyesclosed

0.45±0.180.56±0.19

5.06 0.007 0.49±0.230.63±0.30

swaydirection anterior-posteriormedio-lateral

0.52±0.160.35±0.17

4.06 0.018 0.67±0.280.41±0.19

MV(cm/s) 0.35±0.147.80 0.0005 0.69±0.57

visualcondition eyesopeneyesclosed

0.32±0.130.39±0.14

8.00 0.0004 0.45±0.310.92±0.84

swaydirection anterior-posteriormedio-lateral

0.41±0.130.27±0.12

n.s. 0.82±0.780.51±0.44

CF(Hz) 0.37±0.11 n.s. 0.46±0.18

hControl=healthycontrolsubjects,PATpre=patientsbeforeintervention,PATpost=patientsafter

intervention,RMS=rootmeanssquare,MV=meanvelocity,CF=centerfrequency,n.s.=not

significant.

Figures

26

Figure1

ExperimentalsetupPerturbedstance:PosturalreactionsofAupperbodyandBlowerbody

on1°peak-to-peakplatformrotation(Cstimulus)over20secondswitheyesopenina

representativehealthycontrolsubject(hCon)andapatientbefore(pre)andafter(post)

intervention.Bodyreaction(A/B)followstheplatformmovement(C).Schemeofasubject

standingontheplatforminanuprightpositionD.Spontaneoussway:COPdisplacementof

onerepresentativesubjectofthecontrolsubjects'group(E,hCon)andthepatients'groupF

before(pre)andGafter(post)intervention.Patients'COPdisplacement(F/G)ismore

pronouncedthancontrolsubjects'(E),butreducedafterintervention(G).deg,degrees,sec,

seconds,ap,anterior-posterior,ml,medio-lateral,cm,centimeter.

27

28

Figure2

SpontaneousswayMeanandstandarddeviationofA/Brootmeansquare(RMS),C/Dmean

velocity(MV)andE/Fcenterfrequency(CF)ofCOPswayinanterior-posterior(ap)and

medio-lateral(ml)directioneachfortheeyes-openandeyes-closedconditioninhealthy

controlsubjects(hCon),patientsbefore(pre)andafter(post)intervention.Significant

differencesofRMSandMVbetweencontrolsubjectsandpatientsaremorepronouncedin

theapdirectionandtheeyes-closedcondition.NotethatthelargeerrorbarsinDare

specificallyrelatedtohighfrequencycontentofpatients’sway.

29

Figure3

TransferfunctionsMeanandstandarddeviationofAGain,BPhaseandCCoherence

behaviorasafunctionoffrequency(f)andD/EmeanGainandF/GmeanPhasebehaviorfor

lowerbody(LB)andupperbody(UB)andforeyes-open(eo)andeyes-closed(ec)condition

inhealthycontrolsubjects(hCon),patientsbefore(pre)andafter(post)intervention.

Differencesbetweenhealthysubjectsandpatientsaremorepronouncedfortheupperbody

andwithclosedeyes.Postintervention,patientsshiftedtheirGain,PhaseandCoherence

valuestowardstherangeofcontrolsubjects'values.

30

Figure4

GainfactorMeanandstandarddeviationofA/BupperbodyandC/DlowerbodyGAINFACTOR

behaviorofpatientsbefore(pre)andafter(post)interventionasafunctionoffrequency(f).

GAINFACTORrepresentspatients’GAINvaluesnormalizedtovaluesofhealthycontrol

subjects(hCon)intheeyes-openandeyes-closedcondition.Differencesbetweenhealthy

subjectsandpatientsaremorepronouncedforupperbodyandwitheyesclosed.After

interventionpatientsshiftedtheirGAINandPHASEvaluestowardstherangeofcontrol

subjects'values.

31

Figure5

ModelparametersMeanandstandarddeviationofAtheneuralcontrollerwiththe

proportional(Kp/mghin1/°),derivative(Kd/mghins/°)andintegral(Ki/mghin1/s*°)

contributioncorrectedforsubjects’massesandheights,ofBtheproprioceptivesensory

weight(Wpin°/°)andCthelumpedtimedelay(Tdinseconds)forhealthycontrolsubjects

(hCon),patientsbefore(pre)andafter(post)intervention,eachshownintheeyes-open(eo)

andeyes-closed(ec)conditionandforBWpin0.5and1degree(deg)platformrotation.D

32

showsthemodifiedpostural-controlmodelusedtoidentifyabnormalposturalcontrol

parametersinCIPNpatientsviaanoptimizationprocedurewheredifferencesbetween

experimentaldataandmodelsimulationswereminimized:Themodelconsistsofabody

representedbyaninvertedpendulumwiththemassconcentratedatthecenterofmass

(COM)ofthebodyandthesensorsandneuromuscularsystemsincludingaNeural

Controller.θ,bodyswayangle;h,heightoftheCOMabovetheanklejoints;θref,external

stimulus;Kp,proportionalgain(stiffnessfactor),Kd,derivativegain(dampingfactor),Ki,

integralgainoftheNeuralController;Ppas,passivestiffnessfactor;Dpas,passivedamping

factor;Wp,proprioceptivesensoryweight;Td,feedbacktimedelay;T,controltorque;J,

momentofinertiaofthebody;mgh,bodymass*gravitationalconstant*heightoftheCOM

fromtheanklejoint;s,Laplacetransformvariable.