清潔驗證 華海

會計學1清潔驗證華海PRESENTATIONOverviewValidationApproachMonitoring&Revalidation第1頁/共36頁OverviewICHQ7:“Cleaningproceduresshouldnormallybevalidated.Ingeneral,cleaningvalidationshouldbedirectedtosituationsorprocessstepswherecontaminationorcarryoverofmaterialsposesthegreatestrisktoAPIquality.Forexample,inearlyproductionitmaybeunnecessarytovalidateequipmentcleaningprocedureswhereresiduesareremovedbysubsequentpurificationsteps.”第2頁/共36頁ValidationStudyProcessstepsContaminantsCleaningproceduresProcessEquipmentWorstCaseAcceptanceCriteriaReagents&SolutionsSamplingAnalyticalMethodsMonitoring&RevalidationPointstoconsider第3頁/共36頁ValidationStudyProcessSteps第4頁/共36頁ValidationStudyProcessStepsManufacturingProcessAPIA

ManufacturingProcessAPIB

0o1

0o1

0

1o2

1o2

2

2

2

RawMaterials

IntermediateA/3

IntermediateA/2

IntermediateA/1

CrudeA

RawMaterials

IntermediateB/3

IntermediateB/2

IntermediateB/1

CrudeB

APIA

APIB

第5頁/共36頁ValidationStudyContaminatsREAGENTS(RawMaterialsorpreviousintermediate)PRODUCT+

BYPRODUCTS(IntermediateorAPI)Solventusedascleaningagent:Isitpartofprocessornot?Note:IfpossibleonlyClass3solventsshouldbeused(ICHQ3C).MicrobialContamination: TheAPI,alsoitisnotsterileproduct,isusedtoproducesterilepharmaceuticalproducts?123第6頁/共36頁ValidationStudyCleaningproceduresThecleaningprocessisusuallyperformedbysolventsorsolutionwithchemical-physicalaction(i.e.detergents)orwithchemicalaction(i.e.oxidants;acids;bases;..;etc.)IntheAPIproductionplants:Oftenwatercannotbeused.Thecleaningprocessisdesignedforaspecificproduct第7頁/共36頁ValidationStudyCleaningproceduresVariousproductssameequipment:OnlyonecleaningprocedureforalltheproductsAdvantage: easyvalidationstudiesDisadvantages: it’snocommontousethesamecleaningprocessforalltheproducts.Ifitisinplacecouldbenecessarylongerprocesstimeandmoresolvent/solutionconsumption第8頁/共36頁ValidationStudyCleaningproceduresDifferentcleaningproceduresfordifferentproducts:Advantages: shortprocesstimeandminimumamountofsolvent/solution.Disadvantages: verylongvalidationstudy(allthecleaningproceduresshouldbevalidated).

第9頁/共36頁ValidationStudyCleaningproceduresItisnecessarytorationalizetheselectionandtheuseofthecleaningprocessesinordertoobtainalimitnumberofcleaningproceduresthatareabletoclean,withsufficientaccuracy,alltheprocessedproducts.Theproductshavebeengroupedintodifferentclassesaccordingtotheirchemicalcharacteristics(i.e.salts,chemicalstructure,chemicalproperties,…,etc).Theproductsthatarepartofthesameclassshouldbeclean,aspossible,withthesamecleaningprocess,usingthesamesolvent/solution第10頁/共36頁ValidationStudyCleaningproceduresGrouping第11頁/共36頁ValidationStudyEquipmentMULTIPURPOSEDEDICATEDTheproductcanbecontaminatedbydifferentproductsTheproductcanbecontaminatedbypreviousbatchesTheacceptancecriteriaisquantitativeNote: Alsoinamultipurposeplanttherearededicatedpartsofequipmentdifficulttoclean(i.e.centrifugebags;filters;…;etc.)Theacceptancecriteriaisqualitative

(i.e.visualinspection)第12頁/共36頁ValidationStudyEquipmentGrouping第13頁/共36頁ValidationStudyWorst–CaseWorst-case–ProductWorst-case–EquipmentWorst-case–CleaningprocedureInthemultipurposeplantsisnecessarytoselecttheworst-case第14頁/共36頁ValidationStudyWorst–CaseProductICHQ7:“….IfvariousAPIsorintermediatesaremanufacturedinthesameequipmentandtheequipmentiscleanedbythesameprocess,arepresentativeintermediateorAPIcanbeselectedforcleaningvalidation.Thisselectionshouldbebasedonthesolubilityanddifficultyofcleaningandthecalculationofresiduelimitsbasedonpotency,toxicity,andstability.…”第15頁/共36頁ValidationStudyWorst–CaseProduct–Grouping&MatrixingDifficultyofcleaning: 1.Low 2.Medium 3.HighSolubility: 1.Verysoluble 2.Soluble 3.VeryLowSolubleUsingaspecificcleaningSOP,theproductBistheworst-caseowingtoitisthemostdifficulttocleananditisthemostinsoluble第16頁/共36頁ValidationStudyWorst–CaseEquipmentTheequipment,foreachworkshop,canbegroupedbasedonitstypeandfunction(i.e.reactors;tanks;centrifuges;filters;granulators;…;etc.)TheequipmentofthesametypecanbegroupedbasedonitsoperatingprincipleAftertheequipmentofthesamegroupcanbeevaluatesbasedonitsconstructionmaterialsandontheeasilytodisassembledinordertohighlightthepossiblevisiblecontaminants第17頁/共36頁ValidationStudyWorst–CaseEquipmentForeachgroupofequipmenttheworst-casewillbetheequipmentconstructedwiththemostadsorbentmaterial,andmostdifficulttocleanandinwhichthepossiblevisiblecontaminantsaredifficulttosee.I.e. ● Equipment:Centrifuges ● Type:withhorizontalaxleandwithverticalaxle ● ConstructionMaterial:stainlesscoveredbyplasticmaterialWorst-CaseTheverticalaxlecentrifugesaremostdifficulttocleanofhorizontalaxleonesbecausetheyarecompletelydisassembled,sotheworst-casewillbeaverticalaxlecentrifugeand,withintheverticalaxlecentrifuges,theequipmentmostdifficulttocleanandwiththemostabsorbentconstructionmaterialwillbeselected第18頁/共36頁ValidationStudyWorst–CaseCleaningProceduresThecleaningSOPsshouldbeverifiedinaworst-casescenarioinordertoassurethetheyconsistentlyremovethecontaminantsfromtheequipmentsurface.I.e. Onthedirtyequipmentafterthemaximumholdingtime;Onthecleanedequipmentafterthemaximumholdingtime;Morestringentoperatingparameters(i.e.temperaturerange;cleaningvolume;mixingtime;…;etc.)第19頁/共36頁CaseStudySolubilityAccordingtoUSPPartsofsolventper1partofproductLevelVerysoluble<11Easilysolublefrom1to102Solublefrom10to303Partiallysolublefrom30to1004Slightlysolublefrom100to10005Lowsolublefrom1000to100006VeryLowsoluble/Insoluble>100007第20頁/共36頁CaseStudySolubilityAPI/IntermediatesSolubilityinWaterFosinoprilIntermediate1Level6Intermediate2Level5Intermediate3Level5Intermediate4Level2Intermediate5Level5Intermediate6Level4Intermediate7Levela5AztreonamIntermediateALevel5IntermediateBLevel5APILevel5StavudinaIntermediateXLevel3

IntermediateYLevel4第21頁/共36頁CaseStudyEquipmentThegroupedequipmenthavebeenevaluatedaccordingitsstructuralcharacteristicsas:Shape(easeofdisassembly),Dimension(volumeandsurface),Constructionmaterials(absorbent/notabsorbent),Utilities(i.e.mixer).Numberofoutputpipingthatiscleanedwiththeequipment.第22頁/共36頁CaseStudyProductsvsEquipmentAPIIntermediateSolubilityLevelTankCF009CF015CF020CF021CF030CF227+CC102CF241+CC099CF242+CC100AISI316AISI316AISI316AISI316AISI316LAISI316HastelloyVitrifiedSteel3300L330L4240L4240L4200L6500L8350L6580LFosinoprilSQ32,0345xxxxSQ32,0356xxxxxSQ28,3035xxxxxSQ28,3552xxxSQ27,6165xxxSQ29,2404xxSQ29,2425xxxAztreonamFATGAcido5xxKetoammide5xxxxxAztreonamNSA5xx第23頁/共36頁CaseStudyProductsvsEquipmentvsCleaningProceduresTankCleaningprocessProduct/IntermediateWorst-CaseEquipment

Material

Worst-CaseEquipmentNote1.Pre-wahing:CH2Cl2,2.Washing:NaOH5%,3.1°Rinsing:Processwater4.2°Rinsing:PurifiedwaterSQ…..CF020,CF021,CF226,CF267AISI316L,CF…..CF241Hastelloy,1.Pre-wahing:DMF+aceton2.Washing:NaOH5%,3.1°Rinsing:Processwater4.2°Rinsing:PurifiedwaterSQ…..CF009,CF030,CF226,CF227,AISI316LCF…..CF241Hastelloy,CF242VetrifiedSteel1.Pre-wahing:aceton2.Washing:NaOH5%,3.1°Rinsing:Processwater4.2°Rinsing:PurifiedwaterSQ…..CF020,CF021,CF226,CF227AISI316LCF…..CF241Hastelloy,CF242VetrifiedSteel1.Pre-wahing:CH2Cl2+aceton2.Washing:NaOH5%,3.1°Rinsing:Processwater4.2°Rinsing:PurifiedwaterSQ……CF009,CF226,AISI316LCF…..CF241Hastelloy,CF242,Vetrifiedsteel1.Washing:NaOH5%,2.1°Rinsing:Processwater3.2°Rinsing:PurifiedwaterSQ…..CF241,Hastelloy,CF…..CF242VetrifiedSteel第24頁/共36頁ValidationStudyAcceptanceCriteriaThe“classical”acceptancecriteriaisthefollowing:Visiblecontaminatsshouldbenotpresentafterthecleaningprocess(i.e.“spiking”studiesshouldbecarryoutinordertodefinethevisibleproductconcentrationforeachproduct)Nomorethan1/1000ofthenormaltherapeuticdoseshouldnotbeinthemaximumdailydoseofthefollowingproductNomorethan10ppmofproductshouldbeinthefollowingproduct第25頁/共36頁ValidationStudyAcceptanceCriteriaEachCompanyshouldbedefineandjustifyitsacceptancecriteriaonthebasisofitsproductcharacteristics.Example:1. ForanequipmenttheprocessesonlyAPIcouldbeusedtheMaximumAllowableCarryover(MAC)basedontherapeuticdoseMAC=DTAXMBSB/SFXDGBDTA = normaltherapeuticdoseofAPI(A)DGB = maximumdailydoseofthefollowingAPI(B)processedonthesameequipmentMBSB

= minimum“batchsize”ofthefollowingAPI(B)processedonthesameequipmentSF

= “safetyfactor”第26頁/共36頁ValidationStudyAcceptanceCriteriaOwingtotheintermediatesdon’thaveadailytherapeuticdosecouldbeusetheNoObservedEffectLevel(NOEL)tocalculateMACMAC=NOELAXMBSB/SFXDGBNOELA=LD50XP/2000 = “NoObservedEffectLevel”ofproduct(A)DGB = maximumdailydoseofthefollowingAPI(B)processedonthesameequipmentMBSB

= minimumbatchsizeofthefollowingAPI(B)processedonthesameequipmentSF = “safetyfactor”第27頁/共36頁ValidationStudyAcceptanceCriteriaIftheLD50isnotavailable,the10ppmapproachcouldbeused.(Thisapproachcouldbejustifiedbecausealimitof0,001%ofbatchsizeislessthantheacceptedunknownimpurities=0,1%)MAC=10ppmXMBSB10ppm = 10mgofAper1KgofthefollowingAPI(B)processedonthesameequipmentMBSB

= minimum“batchsize”(Kg)ofthefollowingAPI(B)processedonthesameequipment第28頁/共36頁ValidationStudyReagents/SolventsInwhichprocesssteptheyareused? (i.e.IfitisclosertotheAPIthegratercouldbetheimpactontheproductquality)Theresiduecouldcontaminatetheproduct? (i.e.solubleresidue;difficulttoclean;….;etc.)Whytheresiduecouldbecritical? (i.e.theresidualsolventcouldinterferewiththesolventusedforthefollowingprocess?;Theresiduecouldbetoxic?;….;etc.)REAGENTS(RawMaterialsorpreviousintermediate)PRODUCT+

BYPRODUCTS(IntermediateorAPI)第29頁/共36頁ValidationStudyReagents/SolventsVisibleinspectionQuantitativelimits(basedontoxicity;interferencegrade;…;etc.)第30頁/共36頁ValidationStudySamplingSwabAdvantagesdirectsamplingmethodanditcanremovedryresiduebymechanicalactioneconomiccanbeusedongreatvarietyofsurfacesDisadvantagestheresiduecouldbenotuniformontheequipmentsurfacesinvasivetheresultsarerelatedtothesamplingtechnique Requirementsthesamplingareashouldbemeasuredthecriticalsamplingpointshouldbeidentified第31頁/共36頁ValidationStudySamplingRinsingAdvantagesAlargeamountofequipmentsurfacescanbesampled(i.e.somepartsnotdisassembled)lessrelatedtothesamplingtechniqueaonlinecontrolsystemcanbeusedDisadvantagestheresiduecou