AAN指南治療痛性糖尿病神經(jīng)病第1頁/共49頁Overview

BackgroundGapsincareAmericanAcademyofNeurology(AAN)guidelineprocessAnalysisofevidence,conclusions,recommendationsRecommendationsforfutureresearch第2頁/共49頁Background

Diabeticsensorimotorpolyneuropathyrepresentsadiffusesymmetricandlength-dependentinjurytoperipheralnervesthathasmajorimplicationsforQOL,morbidity,andcostfromapublichealthperspective.1,2

PDNaffects16%ofpatientswithdiabetes;itisfrequentlyunreported(12.5%)andmorefrequentlyuntreated(39%).3PDNpresentsanongoingmanagementproblemforpatients,caregivers,andphysicians.Manytreatmentoptionsareavailable,andarationalapproachtotreatingpatientswithPDNrequiresanunderstandingoftheevidenceforeachintervention.ThisguidelineaddressestheefficacyofpharmacologicandnonpharmacologictreatmentstoreducepainandimprovephysicalfunctionandQOLinpatientswithPDN.

第3頁/共49頁Background,cont.

PharmacologicAgents:Anticonvulsants,antidepressants,opioids,antiarrhythmics,cannabinoids,aldosereductaseinhibitors,proteinkinaseCbetainhibitors,antioxidants(α-lipoicacid),transketolaseactivators(thiaminesandallithiamines),topicalmedications(analgesicpatches,anestheticpatches,capsaicincream,clonidine),andothersNonpharmacologicModalities:Infraredtherapy,shoemagnets,exercise,acupuncture,externalstimulation(transcutaneouselectricalnervestimulation),spinalcordstimulation,biofeedbackandbehavioraltherapy,surgicaldecompression,andintrathecalbaclofen

第4頁/共49頁GapsinCare

Thechroniceffectofdrugtherapiesisnotknown(howlongtotreat,whenorwhethertowithdrawtreatment).Thereisaninsufficientnumberofcomparativestudiesamonghigh-qualitystudies(mostwereClassIIorlower).Thereisnouniformityinhowtomeasurepain,QOL,andfunctionacrossthestudiesexamined.Lackofcosteffectivenessisapparentinallofthestudies.Estimatednumbersneededtotreatareavailable,butnumbersneededtoharmarenotavailable.TheAANclassifiesstudiesbyqualityoftheevidence,notbycost.第5頁/共49頁GapsinCare,cont.

Practitionersdon’tidentifypainenoughinperipheralneuropathyordiabeticneuropathy.Patientswithdiabetesoftenaren’tawarethatnervepainisasymptom.Mostneuropathytherapiestreatpainbutnotnumbness.ThereisalackofattentiontoPDNasadiseaseentity.第6頁/共49頁AANGuidelineProcess

ClinicalQuestionEvidenceConclusionsRecommendations第7頁/共49頁ClinicalQuestions

Thefirststepindevelopingguidelinesistoclearlyformulatequestionstobeanswered.Questionsaddressareasofcontroversy,confusion,orvariationinpractice.Questionsmustbeanswerablewithdatafromtheliterature.Answeringthequestionmusthavethepotentialtoimprovecare/patientoutcomes.第8頁/共49頁

Literature

Search/Review

RelevantCompleteSearchReviewabstractsReviewfulltextSelectarticlesRigorous,Comprehensive,Transparent第9頁/共49頁AANClassificationof

Evidence

AllstudiesratedClassI,II,III,orIVFivedifferentclassificationsystems:TherapeuticRandomization,control,blindingDiagnosticComparisontogoldstandardPrognosticScreeningCausation第10頁/共49頁

AANLevelof

RecommendationsA=Establishedaseffective,ineffectiveorharmful(orestablishedasuseful/predictiveornotuseful/predictive)forthegivenconditioninthespecifiedpopulation.B=Probablyeffective,ineffectiveorharmful(orprobablyuseful/predictiveornotuseful/predictive)forthegivenconditioninthespecifiedpopulation.C=Possiblyeffective,ineffectiveorharmful(orpossiblyuseful/predictiveornotuseful/predictive)forthegivenconditioninthespecifiedpopulation.U=Datainadequateorconflicting;givencurrentknowledge,treatment(test,predictor)isunproven.

Notethatrecommendationscanbepositiveornegative.第11頁/共49頁TranslatingClasstoRecommendationsA=RequiresatleasttwoconsistentClassIstudies.*B=RequiresatleastoneClassIstudyortwoconsistentClassIIstudies.C=RequiresatleastoneClassIIstudyortwoconsistentClassIIIstudies.U=StudiesnotmeetingcriteriaforClassIthroughClassIII.第12頁/共49頁TranslatingClasstoRecommendations,cont.* Inexceptionalcases,oneconvincingClassIstudymaysufficeforan“A”recommendationif1)allcriteriaaremet,2)themagnitudeofeffectislarge(relativerateimprovedoutcome>5andthelowerlimitoftheconfidenceintervalis>2).第13頁/共49頁ApplyingThisProcess

totheIssue

Wewillnowturnourattentiontotheguidelines.第14頁/共49頁ClinicalQuestions1.InpatientswithPDN,whatistheefficacyofpharmacologicagentstoreducepainandimprovephysicalfunctionandQOL?InpatientswithPDN,whatistheefficacyofnonpharmacologicmodalitiestoreducepainandimprovephysicalfunctionandQOL?第15頁/共49頁MethodsMEDLINEandEMBASE1960toAugust2008Relevant,fullypublished,peer-reviewedarticlesMeSHterm“diabeticneuropathies”anditstextwordsynonymsandkeywordsforthetherapeuticinterventionsofinterest(seepublishedguidelineforfulllistofterms)第16頁/共49頁Methods,cont.Atleasttwoauthorsreviewedeacharticleforinclusion.Riskofbiaswasdeterminedusingtheclassificationofevidenceforeachstudy(ClassesI–IV).Strengthofpracticerecommendationswerelinkeddirectlytolevelsofevidence(LevelsA,B,C,andU).Conflictsofinterestweredisclosed.第17頁/共49頁LiteratureReview

79articles2234

abstractsInclusioncriteria:ArticlesdealingwithPDN,describingtheinterventionclearly,reportingstudycompletionrates,definingtheoutcomemeasuresclearlyExclusioncriteria:Casereportsandreviewpapers第18頁/共49頁AANClassificationofEvidence

forTherapeuticInterventionClassI:Arandomized,controlledclinicaltrialoftheinterventionofinterestwithmaskedorobjectiveoutcomeassessment,inarepresentativepopulation.Relevantbaselinecharacteristicsarepresentedandsubstantiallyequivalentamongtreatmentgroupsorthereisappropriatestatisticaladjustmentfordifferences.Thefollowingarealsorequired:concealedallocationprimaryoutcome(s)clearlydefinedexclusion/inclusioncriteriaclearlydefined第19頁/共49頁AANClassificationofEvidence

forTherapeuticIntervention,cont.adequateaccountingfordrop-outs(withatleast80%ofenrolledsubjectscompletingthestudy)andcross-overswithnumberssufficientlylowtohaveminimalpotentialforbias.Fornoninferiorityorequivalencetrialsclaimingtoproveefficacyforoneorbothdrugs,thefollowingarealsorequired**:Theauthorsexplicitlystatetheclinicallymeaningfuldifferencetobeexcludedbydefiningthethresholdforequivalenceornoninferiority.Thestandardtreatmentusedinthestudyissubstantiallysimilartothatusedinpreviousstudiesestablishingefficacyofthestandardtreatment.(e.g.foradrug,themodeofadministration,doseanddosageadjustmentsaresimilartothosepreviouslyshowntobeeffective).Theinclusionandexclusioncriteriaforpatientselectionandtheoutcomesofpatientsonthestandardtreatmentarecomparabletothoseofpreviousstudiesestablishingefficacyofthestandardtreatment.Theinterpretationoftheresultsofthestudyisbaseduponaperprotocolanalysisthattakesintoaccountdropoutsorcrossovers.第20頁/共49頁AANClassificationofEvidence

forTherapeuticIntervention,cont.ClassII:Arandomizedcontrolledclinicaltrialoftheinterventionofinterestinarepresentativepopulationwithmaskedorobjectiveoutcomeassessmentthatlacksonecriteriaaeaboveoraprospectivematchedcohortstudywithmaskedorobjectiveoutcomeassessmentinarepresentativepopulationthatmeetsbeabove.Relevantbaselinecharacteristicsarepresentedandsubstantiallyequivalentamongtreatmentgroupsorthereisappropriatestatisticaladjustmentfordifferences.第21頁/共49頁AANClassificationofEvidence

forTherapeuticIntervention,cont.ClassIII:Allothercontrolledtrials(includingwell-definednaturalhistorycontrolsorpatientsservingasowncontrols)inarepresentativepopulation,whereoutcomeisindependentlyassessed,orindependentlyderivedbyobjectiveoutcomemeasurement.**ClassIV:StudiesnotmeetingClassI,IIorIIIcriteriaincludingconsensusorexpertopinion.**Notethatnumbers13inClassI,item5arerequiredforClassIIinequivalencetrials.Ifanyoneofthethreearemissing,theclassisautomaticallydowngradedtoClassIII.第22頁/共49頁AnalysisofEvidence Question1:

InpatientswithPDN,whatistheefficacyofpharmacologicagentstoreducepainandimprovephysicalfunctionandQOL?

第23頁/共49頁Conclusions/Recommendation

Conclusions:BasedonconsistentClassIevidence,pregabalinisestablishedaseffectiveinlesseningthepainofPDN.PregabalinalsoimprovesQOLandlessenssleepinterference,thoughtheeffectsizeissmall.Recommendation:Ifclinicallyappropriate,pregabalinshouldbeofferedforthetreatmentofPDN(LevelA).

第24頁/共49頁Conclusions/Recommendation

Conclusions:Basedon1ClassIstudy,gabapentinisprobablyeffectiveinlesseningthepainofPDN.Basedon2ClassIIstudies,sodiumvalproateisprobablyeffectiveintreatingPDN.Recommendation:GabapentinandsodiumvalproateshouldbeconsideredforthetreatmentofPDN(Level

B).

第25頁/共49頁Conclusion/Recommendation

Conclusion:ThereisconflictingClassIIIevidencefortheeffectivenessoftopiramateintreatingPDN.Recommendation:ThereisinsufficientevidencetosupportorrefutetheuseoftopiramateforthetreatmentofPDN(LevelU).

第26頁/共49頁Conclusions/Recommendation

Conclusions:LamotrigineisprobablynoteffectiveintreatingPDN.BasedonClassIIevidence,oxcarbazepineisprobablynoteffectiveintreatingPDN.BasedonClassIIIevidence,lacosamideispossiblynoteffectiveintreatingPDN.Thedegreeofpainreliefaffordedbyanticonvulsantagentsisnotassociatedwithimprovedphysicalfunction.Recommendation:Oxcarbazepine,lamotrigine,andlacosamideshouldprobablynotbeconsideredforthetreatmentofPDN(LevelB).

第27頁/共49頁ClinicalContextAlthoughsodiumvalproatemaybeeffectiveintreatingPDN,itispotentiallyteratogenicandshouldbeavoidedindiabeticwomenofchildbearingage.

Duetopotentialadverseeffectssuchasweightgainandpotentialworseningofglycemiccontrol,thisdrugisunlikelytobethefirsttreatmentchoiceforPDN.

第28頁/共49頁Conclusions/Recommendations

Conclusions:Basedon3ClassIand5ClassIIstudies,theantidepressantsamitriptyline,venlafaxine,andduloxetineareprobablyeffectiveinlesseningthepainofPDN.VenlafaxineandduloxetinealsoimproveQOL.Venlafaxineissuperiortoplaceboinrelievingpainwhenaddedtogabapentin.Recommendations:Amitriptyline,venlafaxine,andduloxetineshouldbeconsideredforthetreatmentofPDN(LevelB).Dataareinsufficienttorecommendoneoftheseagentsovertheothers.Venlafaxinemaybeaddedtogabapentinforabetterresponse(LevelC).

第29頁/共49頁Conclusion/Recommendation

Conclusion:Thereisinsufficientevidencetodeterminewhetherdesipramine,imipramine,fluoxetine,orthecombinationofnortriptylineandfluphenazineareeffectiveforthetreatmentofPDN.Recommendation:Thereisinsufficientevidencetosupportorrefutetheuseofdesipramine,imipramine,fluoxetine,orthecombinationofnortriptylineandfluphenazineinthetreatmentofPDN(LevelU).

第30頁/共49頁Conclusions/Recommendation

Conclusions:BasedononeClassIstudy,dextromethorphanisprobablyeffectiveinlesseningthepainofPDNandimprovingQOL.BasedonClassIIevidence,morphinesulphate,tramadol,andoxycodonecontrolled-releaseareprobablyeffectiveinlesseningthepainofPDN.Dextromethorphan,tramadol,andoxycodonecontrolled-releasehavemoderateeffectsizes,reducingpainby27%comparedwithplacebo.Recommendation:Dextromethorphan,morphinesulphate,tramadol,andoxycodoneshouldbeconsideredforthetreatmentofPDN(LevelB).Dataareinsufficienttorecommendoneagentovertheother.

第31頁/共49頁ClinicalContextTheuseofopioidsforchronicnonmalignantpainhasgainedcredenceoverthelastdecadeduetothestudiesreviewedinthisarticle.Bothtramadolanddextromethorphanwereassociatedwithsubstantialadverseevents(e.g.,sedationin18%ontramadoland58%ondextromethorphan,nauseain23%ontramadol,andconstipationin21%ontramadol).Theuseofopioidscanbeassociatedwiththedevelopmentofnovelpainsyndromessuchasreboundheadache.Chronicuseofopioidsleadstotoleranceandfrequentescalationofdose.第32頁/共49頁Conclusions/Recommendation

Conclusions:BasedonClassIandClassIIevidence,capsaicincreamisprobablyeffectiveinlesseningthepainofPDN.BasedonClassIevidence,isosorbidedinitratesprayisprobablyeffectiveforthetreatmentofPDN.Recommendation:CapsaicinandisosorbidedinitratesprayshouldbeconsideredforthetreatmentofPDN(LevelB).

第33頁/共49頁Conclusions/Recommendation

Conclusions:BasedonClassIIIstudies,thereisinsufficientevidencetodetermineifIVlidocaineiseffectiveinlesseningthepainofPDN.BasedonClassIIIevidence,theLidodermpatchispossiblyeffectiveinlesseningthepainofPDN.Recommendation:TheLidodermpatchmaybeconsideredforthetreatmentofPDN(LevelC).

第34頁/共49頁Conclusions/Recommendation

Conclusions:BasedonClassIevidence,clonidineandpentoxifyllineareprobablynoteffectiveforthetreatmentofPDN.Theevidencefortheeffectivenessofmexiletineiscontradictory;however,theonlyClassIstudyofthisagentindicatesthatmexiletineisprobablyineffectiveforthetreatmentofPDN.Recommendation:Clonidine,pentoxifylline,andmexiletineshouldprobablynotbeconsideredforthetreatmentofPDN(LevelB).

第35頁/共49頁Conclusion/Recommendation

Conclusion:Thereisinsufficientevidencetodeterminewhethervitaminsandα-lipoicacidareeffectiveforthetreatmentofPDN.Recommendation:Thereisinsufficientevidencetosupportorrefutetheusefulnessofvitaminsandα-lipoicacidinthetreatmentofPDN(LevelU).

第36頁/共49頁ClinicalContextAlthoughcapsaicinhasbeeneffectiveinreducingpaininPDNclinicaltrials,manypatientsareintolerantofthesideeffects,mainlyburningpainoncontactwithwarm/hotwaterorinhotweather.

第37頁/共49頁AnalysisofEvidence Question2:InpatientswithPDN,whatistheefficacyofnonpharmacologicmodalitiestoreducepainandimprovephysicalfunctionandQOL?

第38頁/共49頁Conclusions/Recommendations

Conclusions:BasedonaClassIstudy,electricalstimulationisprobablyeffectiveinlesseningthepainofPDNandimprovingQOL.BasedonsingleClassIstudies,electromagneticfieldtreatment,low-intensitylasertreatment,andReikitherapyareprobablynoteffectiveforthetreatmentofPDN.Recommendations:PercutaneouselectricalnervestimulationshouldbeconsideredforthetreatmentofPDN(LevelB).Electromagneticfieldtreatment,low-intensitylasertreatment,andReikitherapyshouldprobablynotbeconsideredforthetreatmentofPDN(LevelB).

第39頁/共49頁Conclusion/Recommendation

Conclusion:ThereisnotenoughevidencetosupportorexcludeabenefitofamitriptylinepluselectrotherapyintreatingPDN.

Recommendation:EvidenceisinsufficienttosupportorrefutetheuseofamitriptylinepluselectrotherapyfortreatmentofPDN(LevelU).

第40頁/共49頁AnalysisofEvidence

ComparisonStudies:Studieswith2activetreatmentarmsandwithoutaplaceboarmwereconsideredseparatelyandgradedusingactivecontrolequivalencecriteria(seeappendixe-2andtablee-6ofthepublishedguideline).Weidentified6comparisonstudiesofagents(gabapentintoamitriptyline,4venlafaxinetocarbamazepine,nortriptyline+fluphenazinetocarbamazepine,capsaicintoamitriptyline,andbenfotiamine+cyanocobalaminwithconventionalvitaminB)butdidnotfindevidencetorecommendoneovertheother.5–10Noneofthestudiesdefinedthethresholdforequivalenceornoninferiority.

第41頁/共49頁ClinicalContextItisnotablethattheplaceboeffectvariedfrom0%to50%painreductioninthesestudies.AdjuvantanalgesicagentsaredrugsprimarilydevelopedforanindicationotherthanthetreatmentofPDN(e.g.,anticonvulsantsandantidepressants)thathavebeenfoundtolessenpainwhengiventopatientswithPDN.TheiruseinthetreatmentofPDNiscommon.11ThepanelrecognizesthatPDNisachronicdiseaseandthattherearenodataontheefficacyofthechronicuseofanytreatment,asmosttrialshavedurationsof2–20weeks.Itisimportanttonotethattheevidenceislimited,thedegreeofeffectivenesscanbeminor,thesideeffectscanbeintolerable,theimpactonimprovingphysicalfunctionislimited,andthecostishigh,particularlyfornovelagents.

第42頁/共49頁FutureResearchAformalizedprocessforratingpainscalesforuseinallclinicaltrialsshouldbedeveloped.ClinicaltrialsshouldbeexpandedtoincludeeffectsonQOLandphysicalfunctionwhenevaluatingefficacyofnewinterventionsforPDN;themeasuresshouldbestandardized.Futureclinicaltrialsshouldincludehead-to-headcomparisonsofdifferentmedicationsandcombinationsofmedications.BecausePDNisachronicdisease,trialsoflongerdurationshouldbedone.Standardmetricsforsideeffectstoqualifyeffectsizesofinterventionsneedtobedeveloped.Cost-effectivenessstudiesofdifferenttreatmentsshouldbedone.Themechanismofactionofelectricalstimulationisunknown;abetterunderstandingofitsrole,modeofapplication,andotheraspectsofitsuseshouldbestudied.第43頁/共49頁ReferencesBoultonAJ,VinikAI,ArezzoJC,etal.Diabeticneuropathies:astatementbytheAmericanDiabetesAssociation.DiabetesCare2005;28:956–962.GordoisA,ScuffhamP,ShearerA,OglesbyA,TobianJA.Thehealthcarecosts