The Environmental Monitoring ... - Microbiology Network

1 I,MICROBIOLOGYTPICSTheEnvironmentalMonit oringProgramInaGMPE nvironmentScottSutton"MicrobiologyTopics " Theroutineenvironmentalmonitoringprogram isacriticalaspectofdocumentingthestateof controlofthefacility <1116>"MicrobiologicalControlandMonitoringEnvi ron-mentsUsedfortheManufactureofHealthca reProducts"isprovided Explicitexamplesareprovidedfrompublicall y-availablesources(FDA-483observationsan dwarningletters)ofenforcementactivi-ties basedongoodmanufacturingpracticefailures intheenviron-mentalmonitoringprogram ~-~~-~-~-~~---~--=~-~~-----~e~~~__-'-~'- ~~--r""~--~~~~_IiScottSutton,CoordinahHI NTRODUCTIONT hequalification,orrequalification, especiallytrue: Cleaningstudies Contaminationcontrolplanning(1) Equipmentholdtimestudies(establishmentof cleananddirtyholdtimes~processholdtimesa reprocess-speCific) Selectionofsamplesitesforenvironmentalmo nitoring (EM)program,itssamplesites,frequencyofte sting, methodtoqualifyandjustifytheselectionof thesamplesiteswithina facilityusedforroutineenvironmentalmonit oringis ,thisdiscussiondoesnotincludesamplingoft hewatersystem,gasses, complexprogramsuchasenvironmentalmoni-to ring, , (2)is :"Inasepticprocessing, oneofthemostimportantlaboratorycontrolsi s ( ,whena givenbatchis beingmanufactured) ,allowingforimplementationofcorrec-tions beforeproductcontaminationoccurs( ).

2 " ,"Environmen-talmonitoringdatawillprovid einformationonthequalityof themanufacturingenvironment." (3)pointoutthatthemicrobiologicalassaysu sedhavelimitsofquantifi-cationhigherthan thecustomarycontrollevelsandsoaresubject toa greatdealof ,bytheirargument, ,thetrendofthedataisthecriticalaspect, ,pristineEMdataforanasepticprocessingfaC ilityspeakstothestateofcontrolofthatfaci lity,nottothe"sterility" subsequentarticle(4).Heobservedthatthere lationshipofEMdatatofinishedproductquali tywasanunproven, ,wecannotassumeitistrue,butthatitisunden iablethatthesedata(andparticularythetren dingof thesedata) , ,ofcourse, ,butvaluableasrawdataforthedetermination oftrendsinthefaCilityasa nottheonlyworkertopointoutthefundamental problemusing"rapid" (5) "better" \6 GYT6 PICS-desirable,onlythattheyarenota a schoolofthoughtthatbelievesthatsamplesit esfortheEMprogramshouldnotbedefined,that samplingfroma incorrectandcontrarytogoodmanufacturingp ractice(GMP).

3 Forexample,theFDAasepticprocessinggUidel ine(2, )states:"Itisimportantthatlocationsposin gthemostmicrobiologicalrisktotheproductb ea ,considerationshouldbegiventothepointsof contaminationriskina process,includingfactorssuchasdifficulty ofsetup,lengthofprocessingtime, "Allenvironmentalmonitoringlocationsshou ldbedescribedinSOPswithsufficientdetailt oallowforreproduciblesamplingofa , ( ,duringorat theconclusionofoperations), , ( ,surfacearea,airvolume), , , "Inotherwords,thesitesusedintheroutineEM program ," " 'swebsite(6).Thefollowing483observationd ealtwithsignificantis-suesinjustificatio noftheEMsamplesites(7):"Regardingtheincr easednon-routinesurveil-lancemonitoringp erformedtofurtherevaluatetheBuilding37 Flumanufacturingfacility,therewasnoplani nplacespecifyingthelocationstobetested,m ethodofsampling, (CFU)wereevaluatedformorphologicalcharac teristics,andonlycoloniesexhibitingGram- negativecharacteristicswereGramstainedan didentified The[redacted]methodusedforincreasedsur-v eillancemonitoringoftheenvironmenthasnot beenqualified.

4 "So,clearlyit is importanttohavea rationaleforthelocation, qualificationstudythatwillutilizemanymor esamplesitesthanwillbepresentintheroutin eprogram, (ISO)14644-1(8)describesa statesthatweshoulddeterminetheminimumnum berofsamplesitesbythefollowingequation:N =~ALwhereNLis theminimumnumberofsamplinglocations(roun deduptoa wholenumber)A is (whichistheintentandscope24 JournalofGXPC ompliance----'"'~~-.~""=------"-~,---~,S cottSutton,Coordinatl''u;of14644-1),butw ealsowishtoconsiderviableairsampling(bot hpassiveandactive) , ,bothviableandnon-viableactiveairsamplin gsitesshouldbedoneatthesamelocations(ora scloseaspracticaltoavoidcompro-misingthe othermeasureortheproductintegrity). ( ,settleplates)isa frequently-usedmeasureofcleanroom(orcont rolledzone) ,chiefamongthemtheabilitytoremaininconti nuousexposureforuptofourhours(fourhoursi scitedinEuropeanUnion[EU]2008gUidance[9] -extendedexposuretimesmustbedemonstrated viademonstrationof thegrowthpromotingcapabilitiesoftheageda ndexposedmedia).

5 Inaddition,passiveviablemonitoring(settl eplates)isnotdisruptivetotheimmediateenv ironmentandsomaypossiblysamplesitesveryn earproductexposurepoints(seereference10f ora discussionofthese,andother,advantages).I naddition,settleplatesarenotaspronetovar iationamongdifferentvendorsasareactivesa mplers(11).However, ,settleplatesmaybeparticularlysusceptibl etohan-dling,transport, ,currentregulatoryexpectationforairmonit oringincludestheiruseandthejustifica-tio nof prudentmeasureistousethesamenumberof (PIC/S),whichgenerallycanbecountedontopr ovidedetailsonalmosteverythingmicrobiolo gical,is silentonthispoint(12).Oddlyenough,eventh eParenteralDrugAssociation(PDA)'sTechnic alReport#13(13) (asrelevant).Eachsurfacewouldthenbetreat edasa ,thenumberofsurfacesamplingsitesforequip mentremainsunansweredandis ,quitefrankly, , ,FDA,EU,andtheUnitedStatesPharmacopeia(U SP) #13providesthefollowinggUidanceinthisreg ard:Summer2010 Volume14 Number325---I,MICROBIO[(;)GYl'O~ICS"Fact orstoconsiderinselectingsitesforroutines urveillanceare: statisticaldesign( ,followingthecalculationsinFederalStan-d ard209E)orshouldsiteselectionbemadeonthe basisofgridprofiling?]

6 Shouldsomesitesforroutinemonitoringberot ated?[Notefromauthor:As20gehasbeenwithdr awninfavorofISO14644,theansweris"No"] ,sanitize,ordisinfect? a givensitedisturbtheenvironmentsufficient lytocauseerroneousdatatobecollectedorcon taminateproduct?"(13).TheUSFoodandDrugAd ministrationTheFDAasepticprocessingguida ncedocument(2)statesinsectionIVA:"Airint heimmediateproximityofexposedsteril-ized containers/closuresandfilling/closingope rationswouldbeofappropriateparticlequali tywhenithasaper-cubic-meterparticlecount ofnomorethan3520ina representativelocationsnormallynotmoreth anonefootawayfromtheworksite,withintheai rflow, (ISO5).Werecommendthatmeasurementstoconf irmaircleanli-nessincriticalareasbetaken at siteswherethereismostpotentialrisktothee xposedsterilizedproduct,containers, "Someoperationscangeneratehighlevelsofpr oduct( ,powder)particlesthat,bytheirnature,dono tposea ,inthesecases, ,aircanbesampledinamannerthat,totheexten tpossible,characterizesthe truelevelofextrinsicparticlecontaminatio ntowhichtheproductis "Further, :"Sampletiming,frequency, "Itis importantthatlocationsposingthemostmicro biologicalrisktotheproductbea ,considerationshouldbegiventothepointsof contaminationriskina process,includingfactorssuchasdifficulty ofsetup,lengthofprocessingtime,andimpact of interventions.

7 "EuropeanUnionTheEDguidancedocumentManuf actureofSterileMedicinalProducts(9)provi dessomesiteselectionguid-ance:" ,volumetricair,andsurfacesampling( ,swabsandcontactplates).Samplingmethodsu sedinopera-tionshouldnotinterferewithzon eprotection."26 JournalofGXPC ompliance~"'~--UnitedStatesPharmacopeiaS imilarly,thefollowinggUidanceinthepropos edrevi-siontoUS?chapterd116>(14)is ofgeneralinterest:" "Otherareasofconcernrelativetointroducti onofcontaminationintocleanroomsareat ,areaswithinandarounddoorsandair-lockssh ouldbeincludedinthemonitoringscheme." roomisdetermined, writtenjustificationandshouldconsiderthe following: Contaminationvectors( ,handles,controlpanels,doors,etc.) Hightrafficareas Personnelflow Materialflow Wasteflow Surfacesthataredifficulttodisinfect HVAC returns Productrisk Extentofproductexposure Thetypeofactivityperformednearthatsite Interventionsandmanipulations ,ingeneral, a matterofsomeSGOUS alton,Coordina~()rdiscussionastherecomme ndationsinEurope(EUAnnex1),ISO,andtheUS( USP) ( ,samplingunderbothdynamicandat-restcondi tions) sitesthatshowconSistentlyhighercolonyfor mingunit(CFU)recovered, samplingbeaccelerated(particularlyinlowe rclass-controlledenvironments)duringtheq ualificationstudytoallowcollectionofsuff icientdatafromeachsitetoallthisdetermina tionof sitetestedona weeklybasismayrequirethreemonthsofdata(a tleast12-14datapoints)

8 (andnotwhatbestfitsthemandatedtimelinefo rfacilityqualification).SELECTIONOFROUTI NESITEST hequalificationstudyshouldincludesuffici entrep-licatesunderconditionsboth"atrest "and"dynamic"toallowidentificationof "usefulinformation"isnotmeanttodescribe" thosesitesthatgivethemostdesireablecount s"butratherthosesitesthateithergivethehi ghestcounts( ,serveasthemostsensitivemeasureof thestateof controlof theroom)orwereshowntobeappropriatelyplac edtoheralda 'IMUlRifIJSfOLEUU'~mPJCS->Thefollowingse ction( )fromtheFDAguid-ance(2)is relevantforconsideration:"Allenvironment almonitoringlocationsshouldbedescribedin SOPswithsufficientdetailtoallowforreprod uciblesamplingofa , ( ,duringorat theconclusionofoperations), , ( ,surfacearea,airvolume), , , "Thegenerationofrelevantdatato justifysamplesitesis (15).

9 "Forenvironmentalandpersonnelmonitoring: Youractiveair-samplingunitinoneasepticfi llingroomis notlocatedina "Youhavenotevaluatedthemicrobiologicalbu rdengeneratedfromthemanualasepticconnect ionfromthesourcevesseltotheXXXfillingves sel."It ,controllevels,andfrequencyofsamplingmus tbejusti-fiedbydataanda goodruleofthumbis thatthealertlevelshouldbeatthe95thpercen tileofobservedreadingsfora givenperiodoftime,theactionlevelatthe99t hpercentile(seethePDAT echnicalReport#13foranexcellentdiscussio nofsettingalertandactionlevels).Whilecom monindus-trypracticeis touncriticallyacceptregulatoryrecom-mend ationsforpredefinedcleanzones,thispracti ceis discouragedintheUS(seereference2, "EstablishingLevelsanda TrendingPro-gram").Thereis controversyovertheregulatoryguid-ancefor highlycontrolledareasaswellwithconcernth atcontrollevelssetsofarbelowthelevelofqu antifi-cationforplatecountassays(general ly25-30 CFUperplate,comparedwithregulatoryguidan cesettingalertandactionlevelsaslowassing ledigits).

10 ThisconcernledUSPtosuggesta frequencydistributionapproachfortheseare as(14).Aninterestingdiscussionofthisappr oachcanbefoundinCaputoandHuffman(16).Whi cheverapproachis chosentothedeterminationoftheinitalalert andactionlevels, failuretoinvestigaterepeatedexcursionsth atfiguredprominentlyin483observations(17 )andintheassociatedwarningletter(18).A secondexampleofthisenforcementstancecanb eseeninthewarningletterfrom2009regarding manufacturecontrolofa parenteralmedication(19)Collectionofdata andthenfailuretousethedatatodeterminethe "stateofcontrol"forthefacilityis describedintheFDAasepticprocessingguidan cedocument( )as:" ,mediafills,cleanroomqualification,andsa nitizationstudies,indevelopingmonitoring levels"(2)Thatthisis a long-standingexpectationofFDAisevidenced byanFDA-483observationfrom2001thatreads( inpart)," 'smicrobialalertandac-tionlimitsestablis hedfortheXXtoXXXmanufactur-ingareasareno tbasedonhistoricaldatatakenfromtheEMProg ram"(20).