《Q2（R2）：分析方法验证》英文版.docx

rcHJharmonisationforbetterhealthINTERNATIONALCOUNCILFORHARMONISATIONOFTECHNICALREQUIREMENTSFORPHARMACEUTICALSFORHUMANUSEICHHARMONISEDGUIDELINEValidationofAnalyticalProceduresQ2(R2)FinalVersionAdoptedon1November2023ThGisuidhIAsoednevelboytpheae(j)prObCr值CiXagDrkGi11ogau.dhabseesnubjtoccotnsubUahrOgnuIpaatrotirElIeCSc,or“4i'hdeCHProcAefe4.pfthPeroCtehfsegndarlai廿IeCOmmfeOnaddeodpttofAnergeuIahtodroiyf(sHegicns.Q2(R2)DocumentHistoryCodeHistoryDateQ2ApprovalbytheSteeringCommitteeunderSte2andreleaseforpublicconsultation.26October1993Q2AApprovalbytheSteeringCommitteeunderSte4andrecommendationforadoptiontothethreeICHregulatorybodies.27October1994Q2BApprovalbytheSteeringCommitteeunderSte2andreleaseforpublicconsultation.29November1995Q2BApprovalbytheSteeringCommitteeunderSte4andrecommendationforadoptiontothethreeICHregulatorybodies.6November1996Q2(R1)TheparentguidelineisnowrenamedQ2(R1)astheguidelineQ2Bonmethodologyhasbeenincorporatedtotheparentguideline.Thenewtitleis44ValidationofAnalyticalProcedures:TextandMethodology”.November2005Q2(R2)CompleterevisionofguidelinetoincludemorerecentapplicationofanalyticalproceduresandtoaligncontentwithQl.4EndorsementbytheMembersoftheICHAssemblyunderSt>ndreleaseforpublicconsultation.24March2022Q2(R2)AdoptionbytheRegulatoryMembersoftheICHAssemblyunderStd.p1November2023Q2(R2)ErrorCorrectiontoTable5:DissolutionwithHPLCasproductperformancetestforanimmediatereleasedosageformReportableRangeLinearityformulaeonpage25;CorrectiontoTables6-11onpages26-32.30November20231.egalnotice:Thisdocumentisprotectedbycopyrightandmay,with,theexceptionoftheICHIogo,beused,reproduced,incorporatedintootherworks,adapted,modified,translatedordistributedunderapubliclicenseprovidedthatICH'scopyrightinthedocumentisacknowledgedatalltimes.Incaseofanyadaption,modificationortranslationofthedocument,reasonablestepsmustbetakentoclearlylabel,demarcateorotherwiseidentifythatchangesweremadetoorbasedontheoriginaldocument.Anyimpressionthattheadaption,modificationortranslationoftheoriginaldocumentisendorsedorsponsoredbytheICHmustbeavoided.Thedocumentisprovided"asis"withoutwarrantyofanykind.InnoeventshalltheICHortheauthorsoftheoriginaldocumentbeliableforanyclaim,damagesorotherliabilityarisingfromtheuseofthedocument.Theabove-mentionedpermissionsdonotapplytocontentsuppliedbythirdparties.Therefore,fordocumentswherethecopyrightvestsinathirdparty,permissionforreproductionmustbeObtainedfrofnthiscopyrightholder.ICHHarmonisedGuidelineValidationofAnalyticalProceduresQ2(R2)ICHConsensusGuidelineTableofContents1 INTRODUCTION11.1 Objective11.2 Scope12 GENERALCONSIDERATIONSFORANALYTICALPROCEDUREVALIDATION12.1 AnalyticalProcedureValidationStudy22.2 ValidationDuringtheLifecycleofanAnalyticalProcedure42.3 ReportableRange42.4 DemonstrationofStability-IndicatingProperties52.5 ConsiderationsforMultivariateAnalyticalProcedures53 VALIDATIONTESTS,METHODOLOGYANDEVALUATION73.1 SpecificityZSelectivity73. /GjeneCroanlsid73.1.1.1 AbsenceOfInterference73.1.1.2 OrthogonalProcedureComparison73.1.1.3 TechnologyInherentJustification73.1 R.2ecommDenad.t.ae.d.73.1.1.1 Identification73.1.1.2 Assay,PurityandImpurityTest(s)83.2 Range83.2G.leneCroanlside.81.1.2.1 3.2R.2esp.o.n.s.e91.1.2.2 1.inearResponse91.1.2.3 Non-linearResponse91.1.2.4 MultivariateCalibration103.2 .V3aIidoafLtoiwoRenarnLiem.i.,.t.s103.2.1.1 BasedonVisualEvaluation103.2.1.2 BasedonSignakto-Noise103.2.1.3 BasedontheStandardDeviationofaLinearResponseandaSlope103.2.1.4 BasedonAccuracyandPrecisionatLowerRangeLimits113.2.1.5 RecommendedData113.3 AccuracyandPrecision113.3A.Iccur.a.c.y.113.3.1.1 ReferenceMaterialComparison123.3.1.2 SpHdngStudy123.3.1.3 OrthogonalProcedureComparison123.3.1.4 RecommendedData123.3.1.5 P.2rec.i.s.i.o.n.123.3.2.1 Repeatability133.3.2.2 IntermediatePrecision133.3.2.3 Reproducibility133.3.2.4 RecommendedData133.3.2.5 C.3oinbAinpepdrofdOcArhcecsuarnaPdcyreci,.,s.i.o.n133.3.3.1RecommendedData133.4Robustness144 GLOSSARY145 REFERENCES186 ANNEX1:SELECTIONOFVALIDATIONTESTS197 ANNEX2:ILLUSTRATIVEEXAMPLESFORANALYTICALTECHNIQUES201 INTRODUCTION1.1 ObjectiveThisguidelinepresentselementsforconsiderationduringthevalidationofClrlaIytjCaIproce,cludcdeaspartofregistrationapplications.Analyticalprocedurevalidationformsapartoftheanalyticalprocedurelifecycle,asdescribedwithinICU14nalFtrioccaeIdureDevelo.Q2n(Rt2)providesguidanceonselectionandevaluationofthevariousvalldtae/IoiiVaonsaiyticalprocedures.Thisguidelineincludesacollectionoftermsandtheirdefinitions,whicharemeanttobridgethedifferencesthatoftenexistbetweenvariouscompendiaanddocumentsoftheICHmemberregulatoryauthorities.Theobjectiveofvalidationofananalyticalprocedureistodemonstratethattheanalyticalprocedureisfitfortheintendedpurpose.FurthergeneralguidanceisprovidedonvcliClationstuM"aalticalprocedures.1.2 ScopeThisguidelineappliestoanalyticalproceduresusedforreleaseandstabilitytestingofcommercialdrugsubstancesandproducts,hereafterreferredtoasproducts,.Theguidelinecanalsobeappliedtootheranalyticalproceduresusedaspartoftheconstro水ftIOPharmaceuticalQualitySystem)followingarisk-basedapproach.Thescientificprinciplesdescribedinthisguidelinecanbeappliedinaphase-appropriatemannertoanalyticalproceduresusedduringclinicaldevelopment.Theguidelineisdirectedtocommonusesofanalyticalprocedures,suchasassay,potency,purity,impurity(quantitativeorlimittest),identityorotherquantitativeorqualitativemeasurements.2 GENERALCONSIDERATIONSFORANALYTICALPROCEDUREVALIDATIONThisguidelineindicatesthedatawhichshouldbepresentedinaregulatorysubmission.Analyticalprocedurevalidationdatashouldbesubmittedinthecorrespondingsectionsoftheapplication(ICHM4QTheCominoTnechnDioccaulmFeonifhRegistQftiOnPharmacuHiucinaaUIns).eRelevantdatacollectedduringvalidation(andanymethodologyusedforcalculatingvalidationresults)shouldbesubmittedtodemonstratethesuitabilityoftheprocedurefortheintendedpurpose.Suitabledataderivedfromdevelopmentstudies(seeICHQ14)canbeusedaspartofvalidationdata.WhenanestablishedPlatfQrmanalyptrioccei/SdUgedforanewpurpose,validationtestingcanbeabbreviated,ifscientificallyjustified.Approachesotherthanthosesetforthinthisguidelinemaybeapplicableandacceptablewithappropriatescience-basedjustification.Theapplicantisresponsiblefordesigningthevalidationstudiesandprotocolmostsuitablefortheirproduct.Refenneantce,efwDtherZsuitablycharacterisedmaterials,withdocumentedidentity,purity,oranyothercharacteristicsasnecessary,shouldbeusedinthevalidationstudy.Inpractice,theexperimentalworkcanbedesignedsothattheappropriateperformcncecharacareCrOnjSideIVediSeimsUItaneOUSIytoprovidesound,overallknowledgeoftheperformanceoftheanalyticalprocedure,forinstance:specifici,c/eran/deca,tivprecJVgr(hernporrtaan.bgleeAsdescribedinICHQ14,thesystseumitatbe,s*r8K7Rintegralpartofanalyticalproceduresandisgenerallyestablishedduringdevelopmentasaregularcheckofperformance.RobM5isrtypieakly5evaluatedaspartOfdevelopmentpriortotheexecutionoftheanalyticalprocedurevalidationstudy(ICHQ14).Finally,theanalDfiOCaevldaulriedsattr%snegydevelopedbasedonknowledgeoftheanalyticalprocedureandtheintendedpurpose.Thisincludestherequiredanalyticalprocedureperformancetoensurethequalityofthemeasuredresult(ICHQ14).Ifsuccessfullyexecuted,theanalyticalprocedurevalidationstrategywilldemonstratethattheanalyticalprocedureisfitfortheintendedpurpose.2.1 AnalyticalProcedureValidationStudyAvalidationstudyisdesignedtoprovidesufficientevidencethattheanalyticalproceduremeetsitsobjectives.TheseobjectivesaredescribedwithasuitablesetofperformancecharacteristicsandrelatedpeYfCrcnraitWheirChiCanvarydependingontheintendedpurposeoftheanalyticalprocedureandthespecifictechnologyselected.Section3''VALIDATIONTESTS,METHODOLOGYANDEVALUATION"summarisesthetypicalmethodologiesandvalidationteststhatcanbeused(seealsoFigure2inAnnex1onselectionofvalidationtests).Specificnon-bindingexamplesforcommontechniquesaregiveninAnnex2.Table1(below)providesthemeasuredqualityattributes,typicalperformancecharacteristicsandrelatedvalidationtests,whicharefurtherillustratedinAnnex1.Thevalidationstudyshouldbedocumented.Priortothevalidationstudy,avalidationprotocolshouldbegenerated.Theprotocolshouldcontaininformationabouttheintendedpurposeoftheanalyticalprocedure,theperformancecharacteristicstobevalidatedandtheassociatedcriteria.Incaseswherepriorknowledgeisused(e.jro.m,developmentorfrompreviousstudies),appropriatejustificationshouldbeprovided.Theresultsofthevalidationstudyshouldbesummarisedinavalidationreport.Theexperimentaldesignofthevalidationstudyshouldreflectthenumberofreplicatesusedinroutineanalysistogenerateareporrteas.bluf/jAestified,itmaybeacceptabletoperformsomevalidationtestsusingadifferentnumberofreplicatesortoadjustthenumberofreplicatesintheanalyticalprocedurebasedondatageneratedduringvalidation.Figure1showstheinter-relationshipbetweenICHQ2andICHQ14,andhowknowledgegeneratedduringanalyticalproceduredevelopmentasdescribedinICHQ14aidsthedesignofavalidationstudy.Table1:TypicalperformancecharacteristicsandrelatedvalidationtestsformeasuredqualityattributesMeasuredQuality.AttributeAnalyticalProcedurePerformanceCharacteristicstobe<Demonstrated(2)IdentityImpurity(Purity)Otherquantitativemeasurements(1)AssayContentorpotencyOtherquantitativemeasurements(1)QuantitativeTestLimitTestSpecificity(3)SpecificityTest+RangeResponse(CalibMraoJie1.owerRangeLimit-+QL+DL+Accuracy(4)AccuracyTest-+-+Precision(4)RepeatabilityTestIntermediatePrecisionTest-+(5)+(5)-signifiesthatthistestisnotnormallyconducted+signifiesthatthistestisnormallyconductedtinsomecomplexcasesDLmayalsobeevaluatedQL,DL:quantitationlimit,detectionlimit(1) otherquantitativemeasurementscanfollowtheschemeforimpurity,iftherangelimitisclosetotheDL/QL;otherquantitativemeasurementscanfollowtheschemeforassay(contentorpotency),iftherangelimitisnotclosetotheDL/QL(2) someperformancecharacteristicscanbesubstitutedwithtechnology-inherentjustificationinthecaseofcertainanalyticalproceduresforphysicochemicalproperties(3) lackofspecificityofoneanalyticalprocedureshouldbecompensatedbyoneormoreothersupportinganalyticalprocedures,unlessappropriatelyjustified(4) alternatively,acombinedapproachcanbeusedtoevaluateaccuracyandprecision(5) wherereproducibilityhasbeenperformedandintermediateprecisioncanbederivedfrom(hereproducibilitydataset,anindependentstudyforintermediateprecisionisnotrequiredFigure1:ValidationstudydesignandevaluationOb>cctveperformancecharactc%tcAnahticalprocedureIifecvclcAnalyticalproceduremanagementAppropriatedcvek*pfncnlhuDocumentvalidationresultsanddataEvaluationagainstacceptanceEIeriaorparameterningc、('onclusonsandacceptanceofanalyticalprocedureperformance*aldatonstrategyExaluationofporknouledge,includingavailabledevck>pmenOrvhtondauwithjustificationAdditionalexperunentsandevaluationaccording(oICHQ2me(bodologyorahcutveapproachwithJustiHcalionXaliiLfetHmte¼KandSrealuainofLLlLI2.2 ValidationDuringtheLifecycleofanAnalyticalProcedureChangesmayberequiredduringthelifecycleofavalidatedanalyticalprocedure.Insuchcases,partialorfullrevaIiBybenqUoiItod.Scienceandrisk-basedprinciplescanbeusedtojustifywhetherornotagivenperformancecharacteristicneedsrevalidation.Theextentofrevalidationdependsontheperformancecharacteristicsimpactedbythechange.TransferofavalidatedanalyticalprocedureshouldbeconsideredinthecontextofanalyticallifecyclechangesinlinewithICHQ14.Whentransferringanalyticalprocedurestoadifferentlaboratory,apartialorfullrevalidationoftheanalyticalprocedureperformancecharacteristicsand/orcomparativeanalysisofrepresentativesamplesshouldbeperformed.Justificationfornotperformingadditionaltransferexperimentsshouldbeprovidedifappropriate.CHUaliJcan心sendtodemonstratethattheanalyticalproceduremeetspredefinedperformancecriteriabyusingdatageneratedatmultiplesitesandcouldalsosatisfytherequirementsofanalyticalproceduretransferattheparticipatingsites.2.3 ReportableRangeTherequiredreportablerangeistypicallyderivedfromthespecificationanddependsontheintendeduseoftheprocedure.Thereportablerangeisconfirmedbydemonstratingthattheanalyticalprocedureprovidesresultswithacceptablereos,sacecuracyandprecision.Thereportablerangeshouldbeinclusiveoftheupperandlowerspecificationorreportinglimits,asapplicable.Table2exemplifiesrecommendedreportablerangesforcommonusesofanalyticalprocedures;otherrangesmaybeacceptableifjustified.Insomecases,e.*tLow,amounts,widerupperrangesmaybemorepractical.2.4 DemonstrationofStability-IndicatingPropertiesAvalidatedquantitativeanalyticalprocedurethatcandetectchangesinrelevantqualityattributesofaproductduringstorageisconsideredtobestability-indicating.TodemonstrateSpecificityZselectivityofastability-indicatingtest,samplescontainingrelevantdegradationproductsshouldbeincludedinthestudy.Thesecaninclude:samplesspikedwithtargetanalytesandknowninterferences;samplesthathavebeenexposedtovariousphysicalandchemicalstressconditions;andactualproductsamplesthatareeitheragedorhavebeenstoredunderstressedconditions.2.5 ConsiderationsforMultivariateAnalyticalProceduresForr11LiltivaanraiIayptteriocca,determinedthroughamultivariatecalibrationmodelutilisingmorethanoneinputvariable(e.ga.sp,ectrumwithmanywavelengthvariables).Themultivariatecalibrationmodelrelatestheinputdatatoavalueforthepropertyofinterest(i.e.mo,deloutput).Successfulvalidationofamultivariateprocedureshouldconsidercalibration,intetrensatUngandvalidation.Typically,developmentandvalidationareperformedintwophases.Inthefirstphase,modeldevelopmentconsistsofcalibrationandinternaltesting.Calibrationdataareusedtocreatethecalibrationmodel.Testdataareusedforinternaltestingandoptimisationofthemodel.Thetestdatacouldbeaseparatesetofdataorpartofthecalibsre®Sftdiinoanotationalmanner.Thisinternalteststepisusedtoobtainanestimateofthemodelperformanceandtofine-tuneanalgorithm'sparameters(e.Rhe.nu,mberofZctveanrti4ior/?pa/rtraLyleastsquares(PLS)toselectthemostsuitablemodelwithinagivensetofdata.Formoredetails,seeICHQ14.Inthesecondphase,nodvealUdaavtaiIoinds,aewttitihoindepesnadtnepnltesusedforvalidationofthemodel.Foridentificationlibraries,validationinvolvesanalysingsamples(Je,challengesamples)notrepresentedinthelibrarytodemonstratethediscriminativeabilityofthelibrarymodel.Table2:ExamplesofreportablerangesforcommonusesofanalyticalproceduresUseofanalyticalprocedureLowendofreportablerangeHighendofreportablerangeAssayofaproduct(1)80%ofdeclaredcontentor80%oflowerspecificationacceptancecriterion120%ofdeclaredcontentor120%oftheupperspecificationacceptancecriterionPotencyLowestspecificationacceptancecriterion-20%Highestspecificationacceptancecriterion÷2