FMEA輝瑞企業(yè)培訓(xùn)資料

ManagingPharmaceuticalQuality:RiskorUncertaintyManagement?

AjazS.Hussain,Ph.D.OfficeofPharmaceuticalScienceCDER,FDAPQRIWorkshopFebruary1,2005WhatisQuality?Whatispharmaceuticalquality?consistentdeliveryofthelabelperformanceandlackofcontamination.operationalzedviaasetofpre-specifiedqualityattributes(e.g.,specifications,limits)andthroughtheCGMPregulations.FDA,initsqualitydefinition,isstandinginforthecustomer—anditisapparentthathealthcarepractitionersandpatientshighlyvalueanadditionaldrugattribute:productavailabilityGoodpharmaceuticalqualityrepresentsanacceptablylowriskoffailingtoachievethedesiredclinicalattributes.ManagementGoalsImprovingqualityandensuringavailabilityOptimaluseofourresourcesAsystemsapproachtoCMCreviewandCGMPinvestigationsBasedonknowledgeandprocessunderstandingAchieving“qualitybydesign”Demonstrating“scienceofdesign”Continuouslearningandimprovementthrough“manufacturingscience”AnApproachforQuality–RiskConnectionConceptofQualitybyDesign(QbD)Productandprocessperformancecharacteristicsarescientificallydesignedtomeetspecificobjectives,notmerelyempiricallyderivedfromperformanceoftestbatchesCharacteristicsimportanttodesiredperformancemustbederivedfromacombinationofpriorknowledgeandexperimentalassessmentduringproductdevelopment.Fromthisknowledgeanddata,amultivariatemodellinkingproductandprocessmeasurementsanddesiredattributesmaybeconstructed.Clinicalstudywouldthenbeviewedasconfirmatoryperformancetestingofthemodel.Woodcock,2004ASystemsApproachCMCReveiwCGMPInvestigationsScienceofDesignManufacturingScienceDeliverQualitybyDesignStateofControl&ContinuousImprovementQualitycannotbetestedintoaproduct;ithastobebydesign“MarketStandards”ScienceofDesign+ManufacturingScience=QualitybyDesign

Risk/BenefitandQualityHarmAcceptableRisk/BenefitQualityLabelNobenefit(placeboeffect)ManagingPharmaceuticalQualityQualityofanewmolecularentity(apotentialdrug)Intrinsicpharmacological&toxicologicalattributesIdentityComplexityArangeofuncertaintywithrespecttoidentityof“activemoiety”,purityandstabilityofmaterialsusedinevaluationofpharmacologicalandtoxicologicalattributes(ifamixture;variabilityaddsadditionaluncertainty)Variabilityintheextentandrateofdeliveryof“activemoiety”tothesitesofactionandvariabilityinthepharmacological&toxicologicalresponseandmeasurementsystemsfurtheraddsuncertaintyManagingPharmaceuticalQualityQualityofadrugproductForestablishingproposedtherapeuticclaim(label)DrugproductmanufacturedforclinicaltrialsAftersuccessfuldemonstrationoftherapeuticclaim(acceptablerisk-to-benefitratio)DrugproductmanufacturedforcommercialdistributionLifecycleoftheproduct(shelf-life,exclusivityperiod,genericcompetition,post-approvalchanges,…)Drugproductmanufacturedatmanydifferentfacilities,changesintheprocess,differentmanufactures,…Uncertainty,VariabilityandRiskQuality–ClinicalConnectionHowdoesaproductformulationanditsmanufacturingprocessimpactclinicalperformance?Withoutaclearunderstandingweareuncertain(lackofknowledge)Indecisionmakingtherearemanyadvantagesindistinguishingbetweenuncertainty,variability(randomvariation)andriskGoalsandCharacteristicsofaQualityDecisionSystem:ExampleGoal:expectedtohavethesameclinicaleffectandsafetyprofilewhenadministeredtopatientsundertheconditionsspecifiedinthelabelingCharacteristicsUncertaintyVariabilityRiskPharmaceuticalEquivalentSameactive,identicalamount,samedosageform,androuteofadministration.Identity,StrengthQuality,Purity.CompendialorotherstandardsPriorKnowledge(NDA)PostApproval:MonitoringprogramSuchasMedWatchConsumerComplaintsTherapeuticInequivalenceCoordinatingCommitteeNeedforBioequivalenceAssessmentDonotpresentaknownorpotentialbioequivalenceproblem.AcceptableinvitrostandardCompendialDissolutiontestmethodPresentaknownorpotentialbio-problem.

Appropriatebioequivalencestandard90%ConfidenceIntervalofTest/Refratioforrateandextentofabsorptionin80-125%rangeAdequatelyLabeledSimilaritywithreferencelabel,medicationerrors.,,Certaindifferencesduetochangesinthemanufacturer,distributor,pendingexclusivityissues,orothercharacteristicsManufacturedinconformancetoCGMP'sProcessValidationandQualitySystemDeviations,OutofSpecifications,...ANDAApplications:LimitedInformationContent(e.g.,IRCapsule)Generally1bio-batchBioequivalencegoalpost80-125%90%ConfidenceIntervalfortheTest/ReferenceratioforCmaxandAUCinbetweenthegoalpostNormalhealthysubjects,cross-overdesign,fasting(andfed)conditionsCommonforalloraldrugs–i.e.,procrusteanTocover““worstcase”scenariosIfmeanis100%and90%CIisoutside80-125say85-126.5?Executedbatchrecordandmasterbatchrecord(e.g.,10X)––applicationcommitmentPost-approvalprocessvalidationandstabilitycommitmentPostapprovalchangesbasedonSUPAC-IRDemonstrationof“qualitybydesign”?Analyticaldata+Executedbatchrecord+bio-study+processvalidationIQ,OQ,PQ,..PQ=3consecutivebatchesinconformanceReducedtesting––e.g.,compendialtestsForsimple,conventionalproductdesignsworksfinemostofthetime;qualitybydesignisthenthepriorknowledgeandwhateverdevelopmentdataisgenerated(heldatsite)Uncertainty,VariabilityandRiskUncertainty?Variability?Risk?Uncertainty,VariabilityandRiskProcrusteanstandardshavetoaddress“worstcase”scenariosUncertaintyisnotrisk,currentlywehavenochoicebuttoforcethisequalityUncertaintyisreducedbyimprovingknowledgeWelearnwhattocontrolandthedegreeofcontrolnecessarytominimizeriskForcontinuousqualityimprovementweshouldfocusonimprovinguncertaintymanagementprocessExampleofaCMCRegulatoryDecision:AcceptabilityofaPostApprovalManufacturingProcessChangeOriginalNDAorANDA=CMCQuality&Performance(“Insurance””)ContractForexampleinANDA’sRegulatorycommitments=ConditionsinexecutedbatchrecordsPriorApprovalSupplement*(PAS)ProductconformswithallestablishedspecificationsBut-““Specificationsdonottellthewholestory”E.g.,Shelf-lifeand/orbioavailabilitymayhavechangedand/oranewimpuritymaybeintroducedthatmaynotbedetectedwithestablishedanalyticalmethods,…sponsormaynotadequatelyqualifychanges(inspectionfrequencymaynotbesufficient),….*priorapprovalsupplementforprocessoptimizationandcontinuousimprovementeffortsCompanyX““GoesLean”“Cycle-timereductionsubgroupmembers,forexample,examineeachprocessfunction,forexample,dispensing,rollercompactionandcompression,todeterminehowtospeedupchangeoverandgetequipmenttorunfasterandmoreefficiently.””“Theteamsolicitsideasatregularmeetingsandviaemail.Theideasarethenratedfrom1to10basedon"bangforthebuck"toreducecycletime,andonhowdifficulttheywouldbetoachieve--e.g.,whethertheywillrequirevalidationorpriorFDAapproval.”P(pán)harmaMPostApprovalProcessChange(SUPACGuidance)“Within”(ChangeTargetsetting)“Outside”CurrentUncertaintyManagementAttheoperationallevelthemostefficientapproachformanaginguncertaintyis““demandmanagement””Strict““checkingthebox””processusingpre-specifiedrequirements(recommendations)andprocrusteanstandardsFDAguidancedocuments,483observations,..90%CI80-125%,in-processblenduniformitytests,……..SOP’s,…..CurrentDemandManagement:CharacteristicsForconventionalproductsandmanufacturingprocesses-easytoimplement,supervise,andmangeDecisionresponsibilityisdeferredtoasetof““procrustean””standards-liabilitydistributedtotheentirepharmaceuticalcommunity(e.g.,viaUSP,AAPS,etc.)Forinnovativeand/orcomplexproductsandprocessesnooneiswillingtotakeresponsibilityfordecisions(e.g.,developguidancedocument)––decisionliabilityisthenonthepersonwillingtotakeadecision.CurrentDemandManagement:CharacteristicsInnovationandcontinuousimprovementslowsdownandinefficiencyincreasesThelevelofqualityassuranceachievedisdifficulttomeasureandisburiedinhistoricalmindsetandclinicalvariabilityWithincreasingcomplexityamajorfailureisnecessarytosignalinadequaciesofthesystem––suchafailureisoftentheonlyapproachtointroducenewregulationsorimproveddecisioncriteriaChallengetoandalternateapproachestocurrentprocrusteanstandardsdifficulttoproveanddebatesdrainresourcesWithoutContinuous(Community)Learning:DemandManagementis““static”untilacrisisiscreated,itthenreactstoreplaceacurrentprocrusteanstandardwithanother.ContinuousImprovement:EnhancingCustomerSatisfaction-ReducingVariability“SpecialCause”or““CommonCause”Stable-Yes;Capable?UnstableCorrectiveActionsEliminate““SpecialCause””Reduce““CommonCause”VariabilityFrequent,MajorOOSMinor,OccasionalOOSStable&CapableOntheContinuousImprovementPathStateofControlImprovingUncertaintyManagementDemandmanagementSpecifiedandprocrusteanstandardsE.g.,90%CI80-125%,in-processblenduniformitytests,……..SOP’s,…..PassivemanagementQualitybyDesign,demonstrated“robustness”Canwebringasystems(CMCreviewandCGMPinvestigation)perspectivetobetterrecognizeacompany’sabilitytoachievequalitybydesignandreducetheneedforpriorapprovalsupplements?ActivemanagementContinuouslearningandleveragingknowledgetocreateflexibilityMovetowardsarisk-basedapproachContinuousimprovement(qualityandproductivity)OpportunitiesPATGuidancePATprovidesthepharmaceuticalcontextforLean,SixSigma!CPG7132c.08ComparabilityProtocolQualitySystemsApproachtoPharmaceuticalCGMP’sICHQ8,(9?),(10?)PATGuidanceOpensthedoortorealizethebenefitsofconnectingFisher––to-Shewart–to-DemingFocusonprocessunderstandingleadingtocontrolofprocessend-point!ResearchdataCTD-P2Sec.QbDDrugSubstanceorAPIIntendedUseRouteofadministrationPatientpopulation…..ProductDesignDesignSpecifications(Customerrequirements)P2.1and2.6P2.2,2.4,2.5,2.6DrugProductContainerClosureSystemMicrobiologicalAttributesCompatibility(e.g.,recon)ManufacturingProcessComponentsofdrugproductP2.3ManufacturingProcessDevelopment“DesignSpace”=f(IntendedUse*Design*Control)QualitySystemRiskClassificationProcessDesign&ControlSpecificationsProductDesignIntendedUseDesignRequirementsReliabilityToDeliverDesignRequirementsAssessmentBasedonICHQ8Information/KnowledgeJohnCBerridge,Q8Rapporteur(EFPIA).FDAManufacturingSubcommittee,July2004KnowledgeBasedDecisions:RequireScientificGeneralizeableKnowledge–the“SUPACGAP””SUPACChangeLevelsbasedonpriorknowledgefromthepharmaceuticalcommunity(AAPSSUPACWorkshops)+Research;Yetdifficulttogeneralizebecauseofmultifactorialaspects+lotofsubjectivityLimitedinformationinNDA/ANDAPriorknowledgewithinacompanyandamovetowardsmechanisticUnderstanding(ICHQ8isintendedtofillthisgap)Gap=UncertaintyUncertaintyManagement:QbD&FlexibilityTimeScale&ModeofResponseUncertaintyManagementSystemModificationQbDFlexibilityOperationalRootcauseinvestigation,Efficiency,etc.–LearningtoR&DControlofexcipientsandothersourcesof“commoncause”variabilityReduceCGMPRiskClassification–ContinuousImprovementofQualitySystemTacticalOn-linecontrol[DesignforManufacturability]CriticalControlPoints-Robustprocessend-pointRegulatorySpecifications“DesignSpace”Real–TimeRelease,ModularValidationReg.CMCApprovalStrategicScienceofDesign–Designtoreduce“Uncertainty”Sci.&Tech.Integration–ContinuousLearning&ImprovementRegulatoryCommunicationIntegrateSci-EnablingTechnologyPlatform–“Plug&Play”“TimetoMarket”+“ProductionEfficiency”ScienceofDesignOftendesignanddevelopmentactivitiesarecarriedoutbasedonexperientialknowledge,intuitionandroughguidelines––difficulttocommunicatetoindividualsfromdifferentbackgrounds(the“art””argument)Tolearnhowtorepresentdesignsatamuchhigherlevelthanthecurrentdescriptive“recipe””format(e.g.,executedbatchrecords,SOP’s)whilerigorouslydocumentingkeyconstrainsAValidatedSystemWehavebegunupdatingourcurrentthinkingonvalidationProcessValidationRequirementsforDrugProductsandActivePharmaceuticalIngredientsSubjecttoPre-MarketApproval(CPG7132c.08,Sec490.100).RationalexperimentaldesignandongoingevaluationofdataAchievingandmaintainingastateofcontrolforaprocessbeginsattheprocessdevelopmentphaseandcontinuesthroughoutthecommercialphaseofaproduct'slife-cycleRisk-basedapproaches-inspectionalscrutiny;useofadvancedtechnologies,andtheroleofconformancebatchesintheproductlife-cycle.Afocusonthreefull-scaleproductionbatcheswouldfailtorecognizethecompletestoryonvalidation.Packaging&LabelingProductionFacilities&EquipmentLaboratoryControlsMaterialsDraftGuidanceforIndustryQualitySystemsApproachtoPharmaceuticalCurrentGoodManufacturingPracticeRegulationsTraditionalgoalsNon-traditionalgoals(riskbased,flexibility,robustness,scalability,continuousimprovement,innovation,efficienc