Document N24OrD0zZpbKnXZa9Qmremkew

COVA-MCE'-*-\ THE DEVELOPMENT SERVICES COMPANY July1,1999 Covance LaboratorieIsnc. 9200 LeesburgPik* Vienna,Virgina22182 Tel:703/893-5400 Fax:703/759-6947 Paul Lieder,Ph.D. 3M CorporateToxicology Building220-2E-02 St.Paul,N4N 55144-1000 Dear Dr. Lieder: The expertreview,describedinthislettero,f themouse lyinphoma cellmutationtestingperformed by NOTOX on 3M product,T-6906, was conductedatmy requestby Dr. Brian Myhr, AssociateDirectorof Geneticand CellularToxicologyatCovance'sVienna,VA facilityD.r. Myhr has had over 20 yearsof e.xperiencceonductingand evaluatingdatafrom Mouse Lymphoma assays. Data from a dose range-finder, two trialosfthemutationassaywith and withoutratliverS9 activationa,nd thedescriptioonf thetechnical perfomiance of themutationassay,were reviewedas reportedina draftreportsignedApril25,1998,by theNOTOX studydirector. The testarticleT.-6906, ,vasevaluatedby theNOTOX studydirectoras noninutagenicintheabsence of S9 activatioannd mutagenic with S9 activation.Our assessmentofthe studyisthatitwas technically inadequate,and thereporteddataare insufficietnotreacha validconclusionregardingthemutagenic activitovfthetestarticlienthisassay. The studyshouldbe repeatedundercurrentlyacceptabletest conditionsto draw an%-validconclusionsregardingthemutagenic activitoyfT-6906. The specifidceficiencieosfthestudv(orreportinga)re listeads fol]onvs: 1. The mouse I-v-mphomaassav has been validatedspecificalifvorclone').7.2Cof L5178Y cells.The reportdoes not specif@t-he clone,and the issueisraisedifNCITOX has any karyotypicdocumentation that 3.7.2C cellsNvereused. This isverv importantto establishbecause the mutagenic responseof a different clonemav be greatlydiminished. 2. The S9 was preparedin-housefrom Wistar rats.S9 isusuativpreparedfrom Fischer344 or SpragueDawley rats,so one wonders abouttheefficacyoftheS9 preparationand any documentationof its enzvmatic activitN.I-n.fact.8% v/v S9 (80 pl/ml)was used intheassay;thiscontrastswiththeuse of 20 to .30gumi S9 concentrationussed routinelwyith commerciallyavailableS9 with documented highenzyme activityA.t 80 pl/mi.thisS9 would be highlytoxictomouse lvmphoma cells.While DMN gave a good responseintheassay.thereportdoes notgiveitsconcentrationinthecellculturess,o no comparison with literaturreesultsispossible. 3. EMS shouldnot be used as thepositivecontrolfornonactivatiotnestconditions.This compound inducesvery few smallmutant coloniesand thusprovidesno assurancethatthelaboratoryisableto detect thismajor classof mutants. PositivecontrolslikeMMS, which inducemany smallmutant colonies,are currentlyhighlyrecommended. (Note:theconcentratioonf EMS inthecultureswas not reported.) 4. The laborator-a%p,pears to have poor detectionof smallmutant colonies,which raisesthe issueof lack of definitivneegativeevaluationsoftestcompounds. The historicarlange of negativecontrolmutant frequencieswas givenas 7 to-i8x 10'.which ismuch too low,especialivforthemicrowellmethod. Mutant frequenciesbelow about 30 x 10-"ustaren'tacceptable.(By th@Nvay.thereportgivespositive THE AMERICAS EUROPE ASIA/PACIFIC AFRICA exponentsratherthannegativeexponentsformutant frequencies.) Inthedatatables,DMN shouldbe showing many more smallcoloniesd= largecolonies.InTable 3, in particularD,MA isshoam inducingprimarilylargecolonies.There are veryfew smallcoloniesand almost none in the concurrentnegativecontrol.This de@scribepsoor experimentalconditionsfordetectinginduced mutants. 5.The dose range-findintgestusing'3-hourtreatmentsisdescribedinTable 1. The toxicitovfthetest articlsehouldbe measured by multiplyingthetwo columns of responsesgiven(RCC and iCE -see our notationsinTable 1)tocalculattehe"survival"(RS) relativteo thesolventcontrol.When thisisdone,the survivaldecreasessteadilvwithincreasingdose down to23% at333 gg/ml withoutS9. (The increasein survivalat thetop dose of 1000 gg/n-disstrangeand suggestssome adsorptionof solubletestarticleto precipitatemdaterial.)One needsonlytotaketheassay to20% survival,so therejustwas no need to shift to24-hour treatmentsforthemutationassayas indicatedinthe rangefindingresultsectionon page 14 of thereport.Furthermore,themost common measure of toxicit(yRTG) includesthedecreaseingrowth rate over theexpressionperiod,so thetoxicitywould be even greaterdm indicatedinTable 1. The same would be trueforthe+S9 portionof thetestings,o thatthemutationassay with$9 was vervlikelycarried to highertestarticleconcentrationtshan needed. The precipitatinrgange could probably h@ve been avoided altogether. There .,.-caosnsensus agreement among mouse lymphoma assay expertsat the 1999 IWGTP workshop that RTG or RS shouldbe used forthetoxicitmyeasurement. Work isongoing tochoose one measure, probabivRTG. The choiceoftoxicitymeasurement iscriticatlothechoiceof propertestarticle concentrationrangeand thustotheevaluationoftestcompounds inthisassay.'Me method used inthe NOTOX studv .,,illeladto inappropriatelhyighconcentrationosf testarticlbeeingtested. 6. In both mutationassa,,t-rialst,heexpressionperiodwas 3 days afterthe3-hourtreatmentwithT-6906 with S9 metabolicactivation.This isjustnotacceptable.Itiswellknown thatthemaximum mutant detectionoccurs2 davs aftertreatment,and themutant frequencyfallsrapidlythereafterB.v waitingfor3 davs.thesensitivitoyftheassayisseriouslycompromised. The detectionof mutagenicactivitoyf T-6906 (ifreal.and Idoubt itis).%-oulbdecome much enhanced at2 days of expression. 7. The conditionsfordeterminingmutant frequencieswere notoptimalforcontrollinvgariabilityF.or the microNN-elmlethod. itisrecommended to use 1.6cells/-vvfeolrlCE and 2000 cells/welflormutant selectionx 4 platestogive0.768 x 106 cellsforanalysis.The NOTOX studyused I cell/welflorCE and 2000 cellsavelxl3 platesforonl%.0.576 x 10'cellsforanalysis.The softagar method of theassavanalyzes'3 x 106 CellS formutants. The smallernumbers used by NOTOX aredetrimentaltoprecisiona,nd consequent assa%-fluctuationcsould well be responsibleforspurious2-foldand 3-foldincreasesinmutant frequency, especiallNyvhenrelativteo low negativecontrolmutant frequencies.This islikelywhat occurredforthe Table 3 resultswith S9 thatwere evaluatedas showing mutagenicity. 8. Itisnot clearhow themutant freqtfencwyas calculated.The tablesshow totalnumber ofwellswith mutants.butthe Poissoncalculatiosnhouldcome from theproportionofwellswithno mutants.The text indicatesthis%%-asdone correctlvb,ut no verificatioinspossiblefrom the data provided. Total-wells evaluatedand totalempty .vells@hould be given foreach culture. 9. In additionto thereasonsgivenabove fornoteven making an evaluationatthispoint,theS9 activation resultshouldneverhave been evaluatedas positiveformutagenicity.Table 2 resultasrejustnot interpretableT.he testmaterialprecipitateadt 1000 gg/ml,and thesudden and largeincreaseinmutant frequencyiss-.-rnptomatoifcsome technicalproblem. There was no substantiailncreaseinsmallmutant coloniesrelativteo largemutantscolonies,which iswhat one would expectnearlvallthetime. And the toxicit-i,f-m.easured bN-RTG. would probablvhave been excessive(lessthan 10@/oRTG). IfRTG measurements had been used,the highestselecteddose probably would not have exceeded about 300 Ag/ml. Table 3 resultswith S9 activatioanreprobablyjustdemonstratingassay variabilityA.t 175 pg/n-d,where therewas a very toxicresponseat 17% survival,therewas no increaseinmutant frequency.This was apparentlythemost toxictreatmentinthetableand shouldhave shown the largestmutant frequency.The so-calledresponsesat250 and 750 pg/mi arethereforenotbelievablewithoutconfirmationunder better assayconditions.These were probablyexcess'welytoxictreatmentsby theRTG measurement oftoxicity. As an added note,thereisno reason by updated OECD guidelinesfor mouse lyrnphoma assay,adopted in July,1997,toperform 24-hourtreatments.Short-termtreatmentsof 3-5 hoursare sufficienwtithand withoutS9 activation.Ifnegativeresultasreobtainedand ifthetestmaterialisa pharmaceutical,then24hour treatmentsina second triawlithoutS9 arecurrentlyused to meet ICH guidelinesforFDA submission.Other submissionsto FDA, to CFSAN forexample, forfood additivesor food contact materialsw,illfollowtheupdatedOECD guidelinest;hereisno need for24-hour treatments.This informationwas justconfirmedby an FDA stafferinvolvedwith therevisionofthegenotoxicitgyuidelines intheFDA Redbook. Thank you fortheopportunitytoreviewthiswork. Should you have any additionaqluestions,pleasedo not hesitateto contactus. Sincerely, Steve R. Ha,@vorth,Ph.D. Vice President Geneticand CellularToxicology