Loading...
[PDF] Toxicology vs Epidemiology
Toxicology vs Epidemiology Toxicology: • Uses lab experiments, typically on animals and with control groups • Role in risk analysis: identify dose-
[PDF] Chapter 2 Epidemiological and Toxicological Studies
Toxicological studies occur in a controlled laboratory Epidemiological studies usually involve a comparison between two groups of individuals
[PDF] Epidemiology and experimental toxicology-is there a meeting - NCBI
At first sight the question may seem to be a rhetorical one - we are so accustomed to look upon these two disciplines as complimentary for the assessment of
Toxicology and Epidemiology - Oxford Academic
18 juil 2011 · Toxicology and Epidemiology: Improving the Science with a Framework example, geographical differences or changes in disease
[PDF] An Applied Approach to Epidemiology and Toxicology for Engineers
UNIT III--INTRODUCTION TO TOXICOLOGY AND RISK ASSESSMENT Implicit in the practice of epidemiology is the need for the different dis-
[PDF] Introduction to toxicology - European Commission
Toxicant, toxin, and poison are often used interchangeably in the literature; however, there are subtle differences as indicated in the table Toxic substances
![[PDF] Epidemiology and experimental toxicology-is there a meeting - NCBI](https://pdfprof.com/EN_PDFV2/Docs/PDF_7/76823_7jrsocmed00147_0024.pdf.jpg "[PDF] Epidemiology and experimental toxicology-is there a meeting - NCBI")
76823_7jrsocmed00147_0024.pdf 470JournaloftheRoyalSocietyofMedicineVolume82August1989Epidemiologyandexperimentaltoxicology-isthereameetingground?DiscussionpaperPGrassoFRCPathToxicologyUnit,TheRobensInstitute,UniversityofSurrey,Guildford,Surrey,GU25XHKeywords:toxicology;epidemiology;animalmodels;interfaceAtfirstsightthequestionmayseemtobearhetoricalone-wearesoaccustomedtolookuponthesetwodisciplinesascomplimentaryfortheassessmentofhazardorriskthattheyappeartoustobecloselyinterlinked-butonreflectiononecouldaskwhyshouldtherebeameetingground?Thetwodisciplinesaresoverydifferent.Inepidemiologicalstudies,thespeciesinvestigatedisthedefinitivespecies(man)andtheconditionsofexposuretothechemical,althoughoftendifficulttodefine,arerealenvironmentalonesbothintermsofdoseandintermsofrouteofexposure.Inexperimentaltoxicology,testandcontrolgroupsareclearlydefinedandroutesoftreatmentanddoselevelswellestablished,butthespeciesuseddiffersubstantiallyfrommanintheirphysiology,anatomy,pharmacodynamicsandmetabolism.Infact,tomanypeople,experimentaltoxicologyisonlysuitedformakingtheworldasaferplaceforrats!Andyetdespitethesedifficultiesandreservations,therearemanysituationswherefindingsfromexperimentaltoxicologyandepidemiology,complementoneanothertotheadvantageofmankind.Animalmodels-usefulwhenavailableAfewexampleswillillustratethispoint.SoonafterWorldWarIIcaseswerebeingreportedofparesisandparalysisamongworkersusingcertaingluesandvarnishes.ThesereportsoriginatedfirstinJapan,andwerefollowedbyothersfromseveralcountries.Thesereportsledtoanumberofepidemiologicalstudieswhichconfirmedthattherewasanoccupationalassociationbetweentheuseofcertaingluesandvarnishesandthedisease'.Atfirst,itwasdifficulttolinktheparalysistoanydefinitivechemicalbutsuspicionsoonfellonn-hexane,whichhadbeenintroducedonalargescaleincertainindustries.Intensivelaboratoryinvestigationsoonestablishedthatn-hexaneproducedaperipheralneuropathyinanimalssimilartothatfoundinhumansandsomeindicationsofthemechanismofthenervedamagehasbeenobtainedfromtheanimalstudies.Thus,itisnowknownthatthereistheformationofatangleof10nmfibresintheaffectednervewhichinterfereswiththeflowofcytoplasmupanddowntheaxon.ThisinterferenceisseeminglyduetoablockageofthecytoplasmicflowatthenodesofRanvier.Laboratorystudieshavealsoidentifiedtheproximatemetaboliteresponsibleforthenervelesionandindoingsodemonstratedthatamuchlessfrequentlyusedsolvent,methylbutylketone(MBK)isalsoresponsiblefornerveparalysis2.Thisisaclearinstancewhereexperimentaltoxicologyhashelpedtounraveltheconnectionbetweenoccupationalexposureton-hexaneandaseriousoccupationaldisorderidentifiedepidemio-logically.Vinylchloride(VC)providedanotherinstancewhereexperimentaltoxicologywasofsomehelpinmakingalinkbetweenthetypeofcancerinducedandaputativereactivemetabolite.ThefirstindicationthatVCwaspotentiallydangerouscamefromtheclinicalobservationsofacroosteolysis(abonediseaseaffectingmainlythehands)inworkersexposedtoVC.ThisdiseasewasrecordedafterseveralyearsoftheuseofVCintheplasticsindustry.Toxicologistsrecognizedtheneedforin-vestigatingthesafetyofVCbecauseofthereportsofacroosteolysisandastudywasconductedbyViolaonthelong-termeffectsofVCexposure.Tumourswereproduced-inthisstudy.Itwasfollowedbyalargescaleinvestigationtoconfirmthisissue.Adose-relatedincidenceofhaemangio-sarcomatawereobservedinthisexperiment.Almostsimultaneously,similartumourswerereportedinworkerswhohadarepeatedhighlevelexposuretoVCfromcleaningthehugevetsusedforpoly-merizationofVCintheproductionofPVC.Theveryclosesimilarityofthetumoursfoundinmanandratprovidedexperimentaltoxicologistswithamodelforfurtherstudies.TheearlyhistologicallesionswerecharacterizedbyPopper,andGehringcarriedoutsomeexcellentstudiesonthemetabolismofVCbothathighlevelswhichledtotumourproductionandatlowlevelsofabout1ppmorthereaboutswhichwastheaverageexposurelevelofmenworkinginaPVCmanufacturingplant.ResultsshowedthatthemetabolismofVCathighlevelswasverydifferentfromthatatlowlevels-potentiallyreactivemetaboliteswereidentifiedatthehigherconcentrationswhilethemetabolicproductsidentifiedatlowexposureswereallthoughttorepresentadetoxificationreaction.Atthattime,(approximately15yearsago),therewasaconsiderabledebateonwhetherVCshouldnotbebannedaltogetherbecauseofthehazarditrepresentedtoworkersandtothosewholivedclosetoVCmanufacturingplants.Themetabolicdataprovidedasufflcientdegreeofreassuranceforthoseconcernedinthedecision-makingprocesstoestablish1ppmaspro-vidinga'deminimis'riskandVCwasnotpushedofftheindustrialscenealtogether.Ofcourse,therewereotherreasonsinfavourofkeepingVC(forexample,itsgreateconomicimportance)but,inmyview,thecontributionsofexperimentaltoxicologywereofvalueinaidingthedecision-makingprocess3.Otherexamplesofahappycooperationofthissortexistbuttherelationshipbetweenepidemi-ologyandexperimentaltoxicologyisnotalwayssoproductive.PaperreadtoSectionsofEpidemiology&CommunityMedicineandOccupationalMedicine,25-26October19880141-0768/89/080470-044$02.00/0o1989TheRoyalSocietyofMedicine
JournaloftheRoyalSocietyofMedicineVolume82August1989471DifficultiesexperiencedintheabsenceofanimalmodelsLetusnowlookatanothersituationwheretheepidemiologistcameupwithapositiveresultbuttheexperimentaltoxicologisthadfailedtocomeupwithamodel.Benzenewouldserveasanexample.Atonetime,benzenewasusedextensivelyasanindustrialsolventandarelativelyhighincidenceofleukaemiawasreportedamongworkmenexposedtohighlevelsofbenzene(150ppmandover).Thereareanumberofsuchreportswhich,tomymind,leavelittledoubtthatacauseandeffectrelationshipexistsbetweenbenzeneexposureandanexcessincidenceofleukaemia.Whenexperimentalanimalswereexposedtobenzene,noleukaemiawasinducedineitherratsormice-notconvincinglyatleast.Thisfailurecannotbeexplainedonthebasisthatlaboratoryrodentsarerefractorytothedevelopmentofthisdiseasesinceleukaemiaoccursnaturallyinthesespeciesandcanbereadilyinducedbyionisingradiation.Thereisafurthercomplicatingfactor.Benzeneisalsoabone-marrowpoisonandinterfereswiththeproductionofbloodcells,particularlythewhitecells.Themyelotoxiceffectoccursnotonlyinmanbutalsoinlaboratoryanimalsparticularlyinmicesothatitisevenmoredifflculttounderstandwhytheserodentsareseeminglytotallyresistanttothedevelopmentofleukaemiafrombenzene4.Yetanotherexampleisthecurrentconcerngener-atedbytheepidemiologicalstudiescarriedoutonoccupationalgroupsexposedtosolvents.Severalsuchepidemiologicalstudieshavebeencarriedoutandmostofthemshowsomeexcessofindividualswithaslightimpairmentofmentalfacultiesincertainoccupationalgroupsexposedtosolvents,particularlyinpainters-hencethename'painters'syndrome'.Theepidemiologicalstudiescarriedouthadseveraldeficienciesandtheconclusionsdrawnfromthesestudieshavebeenquestioned.Yettheyallappeartopointinthesamedirection,namelythatsomeintellectualorpersonalitydeficitexistsinpersonsexposedtolowlevelsofsolventvapour.ThispresentsoccupationalphysicianswithaseriousdilemmaIfthefindingsaretrue,thentheywouldneedtotakeenergeticstepstodealwiththesituation,butthequalityofthestudiesconducteddoesnotinspiremuchconfidencesothatmanyoccupationalphysiciansdonotfeeljustifiedinembarkingonexpensivepro-grammesofprevention.Unfortunately,behaviouralstudiesinanimalshavenotbeenfoundsuitableforinvestigatingtheso-calledpainters'syndromeandthecurrentexperimentalattemptsatanalysinglevelsofneurotransmitters(andotherchemicalsinthebrainthoughttobeimplicatedinCNSfunction)arenotlikelytoyieldanyusefulinformation.Intheabsenceofasuitablemodeltherewouldappearnowayforresolvingthedilemmathattheepidemiologicalstudiesonsolventexposedworkershaveposedforoccupationalhealthworkers2.Theepisodeofthe'toxicoilsyndrome'inSpainprovidesanotherexampleofthedifficultyexperiencedwhennoanimalmodelisavailabletostudythetoxiceffectsindetail.Theconnectionbetweenthetoxicoilandaseriousillnesswhichclaimedhundredsofdeathsisinnodoubt,butdespitethisclearcause/effectrelationshipwearestillignorantofthenatureofthechemicalwhichcausedsuchhavoc.Whetherthisisbecausetheoffendingoilwasdestroyedsothatitwasneveravailabletoexperimentalistsorbecausethelaboratoryanimalsarerefractorytotheeffectofthetoxicagentisimmaterial-thereisnoanimalmodelonwhichthetoxicologistscangotowork.Thisepisodeservestoemphasizetheneedforanimalmodelsinordertoarriveattherootofatoxicologicalproblemposedbyepidemiologists6.IndirectassistancefromanimalmodelsTheexperimentalistmayhelpinanindirectway.Anexampleofthistypeofsupportforepidemiologicalfindingsistheextensiveresearchdoneonthecarcinogenicityofcigarettesmokecondensate.Theepidemiologistestablishedbeyondanyreasonabledoubtthatcigarettesmokingincreasestheriskforthedevelopmentofpulmonarycancerinman.Theexperimentalistwasunabletorepeatthesefindings.Ratsandmiceareobligatorynose-breatherswhilemanisnot,sothattheconcentrationofthesmokeinthelungmaynotreachthesameconcentrationasinman.Evenwithgreatersophisticationforthegenerationofcigarettesmokebytheuseofspecializedmachinesnoonehasyetsucceededinproducinginratsormicethesametypeofpulmonarytumoursthatdevelopinmanfromcigarettesmoke,althoughcarcinomaofthelarynxhasbeenproducedinhamsters.Butscientistsdevelopedadifferentmodel-theyobtainedakindoftarfromcigarettesmokebycondensingitandthenpaintedthistar,suitablydiluted,ontheshavedskinofmice.Averylargecropoftumourswasproducedinadose-relatedmannerwhichappearedtoconfirmthefindingsoftheepidemiologists7.Thereisalessonhereforthosewhoinsistthattherouteofexposuremustbethesameasinmaniftheexperimentalmodelistobeofrelevance.Perhapsabitoflateralthinkingindevelopingsuchmodelsisnotoutofplacebutthishasitspitfalls!Ithinktheseexamplessufficetoindicatethatexperimentaltoxicologycanbeofgreatassistanceindefiningmechanismsofactiontherebyhelpingtoestablishtheriskorthehazardonsomeacceptablescientificbasis.Whereanimalmodelsarelackingconsiderableuncertaintyexistsonthelevelofriskforman.Predictivevalueofanimalstudies-needforcautionInmanyinstances,experimentalmodelsareoftenemployedinorderto'predict'apotentialhazardfromanewchemicalorinthediligentsearchforahazardfromachemicalthatiswidelyusedbutisasyetuntested.Itiscustomaryintheseinvestigationstoemployveryhighlevelsandprolongedadinistration.Undertheseconditionsitisnotsurprisingthatepidemiologicalresultsoftenfailtosupportthepredictionsfromanimalresults.Oneoftheclearestandmostrecentexampleofthissortofsituationistheminisagaonformaldehyde.Thiscompoundhadbeeninuseforseveralyearsasalaboratoryfixative,anembalmingagentandalsoindustriallyintheformationofseveralmanmadepolymericmaterials8.Ithadbeenknownforaslongasformaldehydehasbeeninusethatitirritatedthenasalmucosa-anyonewhoenteredahistologylaboratory10-11yearsagoiswellawareofthis,butsomehow,onegotusedtoit.Moreseriously,itinducedallergicphenomenainboththeupperrespiratorytractandinskin.Insomeindividualsitalsoproducedasthma.Thus,theadverseeffectsofformaldehydewereknownandsome
472JournaloftheRoyalSocietyofMedicineVolume82August1989measuresweretakentoreduceexposureandtoremovesusceptibleindividualsfromexposuretoitsfumes.NofurtherthoughtwasgiventothehazardsfromformaldehydeexposureuntiltheChemicalIndustryInstituteofToxicologycarriedoutacarcinogenicitytestbyinhalationroute,theonebywhichformaldehydeusuallycomesintocontactwithman.Theconcen-trationchosenwas,however,about7-15timesthatinhaledbymanandtheduration(5h/day,5days/weekfor108weeks)wasquitesubstantialandgreaterthananyhumanwouldwillinglyendure.Itwaswithsomesurprisethatwefirstlearntthattumoursweredevelopinginthenasalcavityoftreatedratsandthissurprisewasstillgreaterwhennewscamethroughthatalltheratstreatedwiththehighestdose(15ppm)developedsquamouscellcarcinoma8.Becauseofthewidespreaduseofformaldehyde(HCOH)thisnewscausedgreatconcern.SomeepidemiologicalstudiesontheeffectsofexposuretoHCOHhadalreadybeenconductedandtheywerenegative.Thecarcinogenicityresultsinratsledtoseveralotherstudiesonpopulationsexposedtoformaldehydevapour,e.g,morticians,laboratoryworkersetc,butresultswerealsonegative.Nevertheless,thesuspicionandconcernthatHCOHmaybecarcinogenictomancontinued,althoughtheywereconsiderablyreducedinthefaceofepidemiologicaldata.Theexperimentaliststhensetabouttoanswerthequestiontheythemselveshadposed.Theyshowedthatatthelevelswhichproducedtumours,severedamageoccurredinthemucosaoftheratresultinginsquamousmetaplasiaandahighturnoverofthesecells.Theyfurthershowedthatatconcentrationlevelsof1-3ppm-thesortoflevelstowhichmanislikelytobeexposed-themucosadidnotappeartobedamagedevenbyusingsensitivemethodstodetectit.Inaddition,itwasshownbiochemicallythatthemechanismforthedetoxificationofHCOHwasadequatetopreventanytissuedamagefromtheselowconcentrationsofformaldehydewithwhichmancomesintocontact8.Tomymind,theevidencefromepidemiologyisquiteclearthatat1-3ppmthereisvirtuallynoriskforthedevelopmentofcancerbyformaldehydeinmanbut,ofcourse,thisisnegativeevidenceandisnotfullyconvincing.Thepathologicalandbiochemicalevidencefromtestsdesignedtounderstandthereasonforthetissuedamagegivegoodscientificsupporttotheepidemiologicalfindings.Thisminisagaisnotatanendbecausetherearestillsomewhobelievethatthepositiveresultsintheratarethedominantfactortobetakenintoaccountandoverridestheepidemiologybutthereisalwayssomehopethatcommonsensemayprevail.Thiskindoffrustratingexperienceintheinter-relationshipbetweenepidemiologyandexperimentaltoxicologyisillustratedalsobythestoryofthecarcinogenicityofsaccharine.Saccharinewasintro-ducedasanartificialsweetenerabout90yearsagowhenitwasfoundthatdietarycarbohydratere-strictionwasanimportantmeaisforthecontrolofdiabetes.Despiteitsuseinmanycountriestherewasneveranyhintofanincreaseinbladdertumoursindiabetics.Nevertheless,whencyclamatewasintroducedasanartificialsweetenerandwasfoundtocomplementsaccharineitwasthoughtthatthemarketablesaccharine/cylamatecombinationneededtobetestedforcarcinogenicityinrats.Cyclamateitselfhadbeentestedinseverallong-termtestsinratswithanegativeresultbutthecombinationwasfoundtobecarcinogenic.Afairlyhighincidenceoftumoursofthebladderwerefoundinthetreatedrats.Atfirsttheculpritwasthoughttobethecyclamatebutseveralstudieswerecarriedoutonsaccharineitselfandinsomeexperiments,whereitwasgivenatadosagelevelwhichwas700or1000timesthatofthehumanintake,bladdertumourswereproduced.Thisledmanytointerpretthesefindingsasindicativeofcarcinogenicriskandtherewasconsiderablepressureontheauthoritiestobansaccharinewithoutceremonyjustascyclamateswereunceremoniouslybannedwhentheoriginalcarcinogenicitystudyofsaccharine/cyclamatecombinationwasfoundtobecarcinogenittotherat-bladder.However,wisercounselsprevailedandacallforepidemiologicalevidenceonthecarcinogenicityofsaccharinewasmade.Severalstudieshavebeencarriedoutandthesewereinthemain,negative.Saccharine,fortunatelyfortheweightwatchersanddiabetics,isstillwithusandseemstobeholdingitsownagainstthenewersweeteners.Thus,inthelasttwoexamples,theexperimentaltoxicologistdrewaredherringwhichsentepidemiologistsscurryingonafalsetrail9.Anotherexamplewhereexperimentaltoxicologyidentifiedahazardwhichwasnotsupportedbyepidemiologyistheepisodeofdioxintoxicity.Thischemicalisacontaminantofphenoxyaceticacidherbicides.AfterthepublicoutcryagainstthesetypesofherbicidesfollowingtheirextensiveuseinVietnamasdefoliatingagentstheyweresubjectedtoextensivestudiesinanimals.Theywerefoundtobebothhighlytoxicandcarcinogenicaswellasteratogenic.Thisactivitywassoontracedtoacontaminant-dioxin.Followingthisdiscoveryseveralepidemiologicalstudieswerecarriedout.Theonescarriedoutonemployeeswereinconclusive.ThestudiescarriedoutonVietnamveteranswho-hadahighexposuretocommercialphenoxyaceticacidscontainingdioxinwerenotanymoreinformative.Andyettheanimalevidencewassoconvincinglypositivethatthefailureofepidemiologicalstudiestoprovideapositiveanswerwasattributedtothepoorqualityofthestudies.ThencameanexplosioninachemicalfactorynearSevesoatowninNorthItalywheretherewasafairlyheavyhumanexposuretodioxin,althoughabriefone.Therewasconsiderableconcernbecauseitwasthoughtthatthishighexposurewouldresultinmanydeathsand,fetalmalformations.Fortunately,theseominousfearswereunfulfilledsinceapartfromchloracnetherewasnounusualincidenceofanyothertypeofdisease.Perhapsitwouldhavebeenbetteriftheexperimentshadnotbeencarriedout-aconsiderablenumberofpeopleatSevesowouldhavebeensparedagreatdealofanguishandapprehensionl0.PitfallsfrominappropriateanimalexperimentsLastlyIwouldliketodrawyourattentiontothedfficultiesexperimentaltxicologissbleadthemselvesintowhen-inappropriateexperimentaltestsareemployedandnotonlyconfusethemselvesbutconfisetheepidemiologistsas-well.The-exampleIliketoquoteisnickel.Someearlyepidemiologicalstudiesofworkersinnickelrefineriesdemonstratedclearlythattherewasanexcessincidenceofcancersofthenasalcavityandlunginthisgroupofpeople.Thiswaslaterconfirmed
JournaloftheRoyalSocietyofMedicineVolume82August1989473byfurtherepidemiologicalstudieswhichnotonlyimplicatedrefineriesbutalsootherprocessessuchasroastingandelectrolysis.Thechemistsidentifiedawidearrayofinorganicnickelcompoundsaswellasnickelmetalinthefumesinhaledbyworkersatthesefactoriesandawholeseriesofnickelcompoundswereinvestigatedexperimentally.Atthetimetheseearlyepidemiologicalresultswerebeingpublishedtherewasonesimpleand,accordingtomany,areliablewayoftestingforcarcinogenicactivity-namelyadministrationbythesubcutaneousorintramascularroute.Virtuallyallnickelcompoundsthatwerefoundinnickelprocessingplantsweretestedbythisrouteandallproducedsarcomas.Theresultsofthesubcutaneousrouteofadministrationwerethentakentoindicateasignificantcarcinogenicriskandnickelandallitssaltswerelabelledas'carcinogenic'inanimalsandpotentiallycarcinogenicforman.Theimplicationfromexperimentaltoxicologywasthatthenickelinprostheticimplantsorinthe'coininyourpocket',toborrowaphrase,waspotentiallycarcinogenic.Tomeitissurprisingthattheseexperimentalresultsattractedlittleattentionandthattherewasnooutrightbanonnickel.Perhapsitisaswellthatsuchdrasticmeasures-werenottakenbecausesubsequentexperiencewiththesubcutaneousandintramuscularrouterevealedthattheseroutesaremostunreliableasindicesofcarcinogenicity.Theentireproblemofnickelcarcinogenicityisnowbeingre-investigatedtoxicologicallybymoreappropriateroutes'1.Thisepisodeshowsquiteclearlyhowtheexperi-mentaltoxicologistcanmisleadhimselfandothersbyproducingfalsedatawhich,becauseoftheiremotiveimplications,aredifficulttobeplacedinaproperscientificperspective.Ihavetriedtoanswerthequestion-Istheremeetinggroundbetweenepidemiologyandexperi-mentaltoxicology?Ihavetriedtoindicatethattheanswertothisquestionisprobablyaguarded'yes'.Themeetingground,however,isnotsmoothanduniform.Itisquitefirminsomeinstances,andwherethisoccurs,thefindingsfromepidemiologycanbeconsiderablystrengthenedandsomelightisthrownonthemechanismofaction.I;Underthesecircumstances,itispossibletoarriveatarationalappraisalofhazardorrisk.Therearesituationswhereexperimentaltoxicologyfallsshortofprovidingamodelfortestingthehypothesisgeneratedbyepidemiologicalstudiesand,intheseinstances,theabsenceofameetinggroundleavesagapinourunderstandingoftheproblemoutlinedinthehypothesis.Thereareothersituationswhereexperimentaltoxicologysuggeststheexistenceofahazardwhichisnotconfirmedbyepidemiologicalfindings.Thismaybedueeithertotheuseofveryhighdoseswhichinduceconsiderablealterationinorganfunctionortoaninappropriateroutefortestingthematerial.Iwouldliketoconcludebysayingthat,althoughmostofourworkinexperimentaltoxicologymaywellbeonlyrelevanttomakethisplanetsafeforrats,someofitdoeshaveadirectrelevanceforman.References1CavanaghJB.Mechanismsoforganicsolventtoxicity:morphologicalchanges.In:Organicsolventsandthecentralnervoussystem.EnvironHealth5.Copenhagen:WHO,1985.2GrassoP.Neurotoxicandneurobehavioraleffectsoforganicsolventsonthenervoussystem.In:WeaverNK,ed.Occupationalmedicine-thepetroleumindustry.Philadelphia:HanleyandBelfusInc,19883ECETOC.Themutagenicityandcarcinogenicityofvinylchloride:ahistoricalreviewandssessment.ECETOC,Brussels,TechnicalReport.No31,19884VanRaalteHGS,GrassoP.HematologicmyelotoxicclastgeniccarcinogenicandleukemogeniceffectsofBenzene.RegulatoryToxicologyPharmacology1982;2:153-765IronsRD,NeptunDA.Effectsoftheprincipalhydroxy-metabolitesofbenzeneonmicrotubulepolymerisation.ArchToxicol1980;45:2976ToxicEpidemicSyndromeStudyGroup.ToxicepidemicsyndromeSpain1981.Lancet1982;ii:697-7027IARC.Monographsontheevaluationofcarcinogenicriskstohumans.Tobaccosmoking.Lyon:IARC,19868ClaryJJ,GibsonJ,WaritzRS,eds.Formaldehyde.Toxicology,epidemiology,mechanisms.NewYork:MarcelDekker,19839GrassoP.Carcinogensinfood.In:SearleCE,ed.Chemicalcarcinogens,2ndedn,vol2.-Monograph182.AmericanChemicalSociety,WashingtonDC:AmericanChemicalSociety,1984:122510LARC.Monographsontheevaluationofcarcinogenicriskofchemicalstohumans.IARCMonographs1982;Supplement4:23811[ARC.Monographsontheevaluationofthecarcinogenicriskof.chemicalstohumans.IARCMonographs1982;Supplement4:167(Accepted16February1989)