parisonofpredictedandmeasuredpurgefactorsforhydrazineandsubstitutedhydrazineintermediate(compound2)

identity/structureofGIofconcernstagedetails(H=100,reactivityM=10,L=1)(F=10,solubilityM=3,L=1)(H=10,volatilityM=3,L=1)totalpermultiplestagemeasuredlevelhydrazinestagehydrazide1:formation100101-nodistillation1000

stage nal2:intermediate10031300

stagecyclisation/isolation3:10031300

overallclearancecalculated9×108<0.1ppm(LOD)

identity/structureofGIofconcernstagedetails(H=reactivity100,L=1)M=10,(Fsolubility=10,M=3,(Hvolatility

L=1)=L10,=1)M=3,totalpermultiplestagemeasuredlevel

intermediate2

stage nal2:intermediate1001011000

stagecyclisation/isolation3:10031300

overallclearancecalculated3×105<1ppm(LOD)

impurity(I)notbeingdetected(<5ppm)(Table8).Asthisfactoris9×108.ThisisborneoutbytheexperimentalresultsrelatestoanimpuritywithinastartingmaterialthatitselfisthatshowlevelsintheAPItobenotdetected(LOD0.1ppm).introducedintotheoverallsyntheticprocessatanumberofWithrespecttothereactiveintermediate2,despiteitscloserstagesawayfromtheisolatedAPI,itisproposedthattheuseofproximitytothe nalAPI,thecalculatedpurgefactorisagainthisriskassessmenttoolcouldbeusedtoquicklyevaluatethehigh,3×105.Thisismirroredintheexperimentaldatathatriskofcarryover,inthisinstanceintoastartingmaterial,andshowed

thuseliminatetheneedforspeci canalyticaltesting.4.5.CaseStudy5:GIinaRegisteredStartingMaterial

(Example2).Inthisexamplethespeci cconcernrelatesagain

toapotentialGI(A)withinastartingmaterial(I).Inthis■levelstobe<1ppmatLODlevel(seeTable10).CONCLUSIONAthoroughevaluationoftheriskposedbyGTIsisnowainstancetheimpurityofconcern(A)isanitropyridylN-oxidecrucialpartoftheoverallprocessofevaluatingthequalityofderivative(Scheme5).medicines.SuchanevaluationinvolvesassessingthepotentialWhatisinterestingwiththisspeci cexampleisthattheforcarryoveroftheGTI(s)inquestionatlevelsofconcernintoimpurityofconcern(A)isunreactiveinbothstages1and2oftheAPIand/ordrugproduct.Thiscontributiondescribesatheprocess(seeTable9).Furthermore,asneitherstageisnumberofwaysinwhichsuchanevaluationcanbeconducted,isolated,thenthereisnoopportunitytopurgetheimpurityincludingthenovelapproachofpurgefactorcalculation.throughremovalbasedonsolubility,andtheimpurityisnotTheadvantageofsuchanapproachoverotherempiricalvolatile.Hence,thecalculatedpurgefactoris1;thatis,instagesapproachesisthatitprovidesaquantitativeassessmentofthe1and2oftheprocessthereisunlikelytobeanysigni cantrisk,basedonaneasilyunderstoodandstandardizedscale.NotreductioninthelevelofimpurityA.Thiswasmirroredonlycanthisapproachbeusedtoprovidecompellingevidenceexperimentallywhenaspikeof3000ppm(equivalenttothetobackupachemicalevaluation,itcanalso,asshownthroughspeci cationlimitof0.3%)wasmadeintostartingmaterial(I),therealcasestudyexamples,driveabetterunderstandingofanalysisatstage2withinthenonisolatedintermediate(IV)GTI-relatedrisksuchthatattentionisultimatelyfocusedonshowedlevelstohavereducedtoonly2000ppm,afactorofanalyticalcontrolofthosespeci cGTIsthatposeanactuallessthan2.Subsequentanalysisofthe nalisolatedratherthanatheoreticalriskofpresenceinthe naldrugintermediateshowedthatlevelshadbeenreducedto<1substance.

ppm;thiswasduetothehighsolubilityoftheimpuritywithinthesolventsystememployedforthisstage.Acomparisonoftheexperimentalresulttothepredictedpurgefactor(forthisstage■AUTHORINFORMATIONCorresponding

avalueof10)illustratestheconservativenatureofthesolubility*E-mail:andrew.teasdale@Author

termwhenassessingtheriskofcarryover.Notes

4.6.CaseStudy6:Hydrazine.Inthisexampletheriskof

carryoverofahighlyreactive(nonelectrophilic)reagent,hydrazine,andareactiveintermediate,substitutedhydrazine(2),wasassessed,asshowninScheme6.■Theauthorsdeclarenocompeting nancialinterest.REFERENCESThisexampleillustrateswelltheadvantageofthisfocusedQ3A(1)(R2):InternationalImpuritiesConferenceinNewDrugonSubstancesHarmonisation;October(ICH).2006.GuidelineapproachtocalculationofGI-relatedrisk.Hydrazine,ahighlyreactiveknowncarcinogen,isusedinthisprocess,threestagesQ3B(2)(R2):InternationalImpuritiesConferenceinNewDrugonProductsHarmonisation;2006.(ICH).GuidelinefromtheAPI.Aclassicalstage-countingapproachbasedonanassumedpurgeof10perstagewouldindicatea1000-foldQ3C(3)

reductioninthelevel.Incomparison,thecalculatedpurge(4)(R4):InternationalConferenceonHarmonisation(ICH),Guideline

EMEA/CHMP,Impurities:GuidelineGuidelinesonfortheResidualLimitsSolventsofGenotoxic;2009.Impurities,CPMP/SWP/5199/02;2006