E ectsofPhlebotomyontheGrowthofFerricNitrilotriacetate-InducedRenalCellCarcinoma

AkikoMizote,AkiraIHida,MutsumiHosako,MasayoshiFujisawa,MikaKamekawa,andShigeruOkada

DepartmentofPathologicalResearch,FacultyofMedicine,OkayamaUniversityGraduateSchoolofMedicineandDentistry,Okayama700-8558,Japan

Theferricnitrilotriacetate-inducedcarcinogenesismodelisuniqueinthatreactiveoxygenspecies-freeradicalsareinvolvedinthecarcinogenicprocess.Buttheeectsofiron-withdrawalintheprogressionofrenalcellcarcinomaarenotwellunderstood.Weperformedrepeatedphlebotomiesonanimalsthathadbeenadministeredferricnitrilotriacetateintheinitiationstageofrenalcellcarcinoma(phlebotomygroup),andcomparedthedevelopmentofrenaltumorswiththosenotreceivingrepeatedphlebotomies(non-phlebotomygroup).Ferricnitrilotriacetate-treatedmaleWistarratswererandomlydividedinto2groups:aphlebotomygroup(21rats)andanon-phlebotomygroup(17rats).Tenage-adjustednormalratswerealsoobservedasanormalgroup.Hematocritwasmaintainedunder25 inthephlebotomygroup.Hematocritlevelsinthenormalgroupandinthenon-phlebotomygroupwerenotsignificantlydierent.Asaresult,theincidenceofrenalcellcarcinomawasnotsignificantlydierentbetweenphlebotomyandnon-phlebotomyanimals.However,thetotalweightoftherenalcellcarcinomawassignificantlyheavierintheanimalsfromnon-phlebotomygroupthaninthosefromthephlebotomygroup(23.64g±18.54 .54.40g±42.40, ＜0.05).Thepresentstudydemonstratedthatphlebotomyaftertheadministrationofferricnitrilotriacetatedidnotreducetheincidenceofrenalcellcarcinoma.Inaddition,weshowedthatironwithdrawalatthepromotionstageofcarcinogenesiswillretardtumorgrowth.

Keywords:ferricnitrilotriacetate,renalcellcarcinoma,phlebotomy

I ntraperitonealinjectionofferricnitrilotriacetate(Fe-NTA)inducesrenalproximaltubulardamage,

andrepeatedinjectionsultimatelyleadtoahighincidenceofrenalcellcarcinoma(RCC)inratsandmice(reviewedbyOkada［1］).TheFe-NTA-inducedcarcinogenesisisauniqueanimalmodelcharacterizedbythefollowingfeatures:1)thekidneyisthesoletargetorgan;2)thereisahighincidenceoftumors(30-90 ),oftenwith

pulmonarymetastasesandperitonealinvasioninmalerodents;and3)reactiveoxygenspecies-freeradicalsareinvolvedinthecarcinogenicprocess.OurstudyandotherpreviousreportshaveshownthatafterFe-NTAtreat-mentinthekidney,anincreaseinoxidativeDNAbasemodificationscantakeplace,givingsuchproductsas8-hydroxydeoxyguanosine［2-5］, thymine-tyrosinecross-links［6］,thiobarbituricacid-reactivesubstance［7,8,9］,saturatedandunsaturatedmutagenicaldehydese.g.,4-hydroxy-2-nonenal(HNE)andmalondialdehyde(MDA)［3,10］,andHNE-,MDA-modifiedproteins［7,9,10,11］.SupplementalvitaminEandother

ReceivedFebruary21,2002;acceptedMarch18,2002.Correspondingauthor.Phone:＋81-86-235-7141;Fax:＋81-86-235-7148E-mail:okadas＠cc.okayama-u.ac.jp(S.Okada)

http://www.lib.okayama-u.ac.jp/www/acta/

ActaMed.Okayama,2002Vol.56,No.4,pp.199-204

OriginalArticle

Copyrightｃ2002byOkayamaUniversityMedicalSchool.

radicalscavengerspresentinfoodsuppressedtheinci-denceofRCC［5,12］.Fe-NTA-inducedRCCwasshowntohavenomutationsinH-,K-andN-rasoncogenes,andalowincidenceofmutationinthep53tumorsuppressorgene［13,14］.Recentdatasuggestthatinactivationofp15 andp16 genescouldbeoneofthemajorpathwaysresponsibleforFe-NTA-inducedcarcinogenesis［15］.Inhumansaswellasanimals,thereisnoactiveiron

excretionmechanism(reviewedbyOkada［16］).There-fore,theironinjectedduringFe-NTAcancerinductionstudiesismostlyretained,andtheexperimentalanimalsaretherebycharacteristicallyoverloadedwithiron.Theeectsofiron-overloadinthegrowthofRCCarenotknown.WeperformedrepeatedphlebotomiesinanimalshavingreceivedFe-NTA,andcomparedgrowthofRCCwiththoseobservedinnon-phlebotomizedanimals.

MaterialsandMethods

Threeweek-oldmaleWistarratswerepurchasedfromCharlesRiver,Japan.Theywerefednormalratchow(OrientalYeast,Tokyo,Japan).AnimalexperimentswerestrictlyperformedinaccordancewiththeGuidelinesforAnimalExperimentsbyOkayamaUniversityMedicalSchool.Afteraweekofacclimation,42ratsreceivedFe-NTA,and10ratswereleftuntreatedthroughout.Fe-NTAwasadministered2timesaweekfor3months(5-7mg/kgbodyweight)byintraperitonealinjection.Fe-NTAtreat-edratsreceivedtotalof27-54mgofiron.Fe-NTAwaspreparedbythemethoddescribedby

Awaietal［17］.Inbrief,nitrilotriaceticaciddisodiumsalt(NTA,NacalaiTesque,Kyoto,Japan)andferricnitrate9HO(Wako,Osaka,Japan)weredissolvedindistilledwater,andtheferricsolutionwasmixedwiththeNTAsolution.ThepHwasadjustedto7.0withsodiumbicarbonate(Ishizu,Osaka,Japan).ThemolarratioFetoNTAwas1:4.AttheterminationofFe-NTAtreatment,38outof

42ratstreatedwithFe-NTAremainedalive,andalloftheanimalswereswitchedtoalow-irondiet(OrientalYeast,Tokyo,Japan).Theanimalswererandomlydividedinto2groups,namely,aphlebotomygroup(21rats)andanon-phlebotomygroup(17rats).Thephlebotomygroupunderwentexsanguinationsof5mlorlessatatimeunderetheranesthesia,2timesaweek,fromaretro-orbitalvenousinplexusthroughahepariniz-

edhematocrittube.Hematocritwasmaintainedunder25 .EighteenmonthsfromthefinalFe-NTAinjection,all

animalsweresacrificedbyexsanguinationunderetheranesthesia.Serumwascollectedaftercentrifugation(10min,3000rpm).Theweightoftheorgansandtumors,serumiron,andtotalironbindingcapacity(TIBC)wasmeasured.TheobservationoftumorhistologywasperformedbyroutineHematoxylin-Eosinstaining.Sol-ubleironfromtheliverandkidneywasmeasuredasdescribedbelow.

SolubleironwasmeasuredbythemethodofBrueckmannandZondek［18］.Toextractsolubleiron,0.1goflivertissuewasgroundinaglasshomogenizer,0.5mlofsaturatedsodium pyrophosphate,andonemlof10 trichlo-roaceticacid.Then,thehomogenatewasquantitativelytransferredtoawideglasscentrifugetube.Thetubewasthenheatedinaboilingwaterbathforexactly7min,andimmediatelycentrifuged.Theseextractscontainedlowmolecularweightironandferritin.IronwasmeasuredbyanFe-testkit(Wako,Osaka,Japan).Allglasswarewaswashedwith0.1NHClandrinsedwithiron-freewater.

Student’s t-testwasusedtoforthestatisticalanalyses.

Results

ChangesinbodyweightduringtheexperimentareshowninFig.1.Inthephlebotomygroup,4ofthe21ratsdiedduringphlebotomyorduringanesthesia;oneofthe17non-phlebotomyratsdiedduetotumorgrowth

Mizoteetal. ActaMed.Okayama Vol.56,No.4200

Fig.1 Changesinbodyweightduringtheexperiment.(▲)phlebotomygroup(n＝21);(□)non-phlebotomygroup(n＝17);(●)untreatedgroup(n＝10).(average±S.D.).

duringtheobservationperiod.Serumironwasbetween25and30μg/dlinthe

phlebotomygroupduringtheobservationperiod.TIBCincreasedfromthefirst12months,butdecreasedwithage.Hematocritremainedunder25 (Fig.2).ThefinalironconcentrationattheendoftheexperimentisshowninTable1.Animalsinthephlebotomygroupwereslightlyhypoferrinemic.However,theTIBC ofthenon-phlebotomygroupwasnotsignificantlydierentfromthatofthenormalgroup.Hematocritdidnotsignificantlydierbetweentheuntreatedandnon-phlebotomygroups(26 -38 ).Solubleironwassignificantlydecreasedinthekidneysofthephlebotomygroup(P＜0.05vs.untreatedandnon-phlebotomy).Solubleironinliverwasnotsignificantlydierentamonggroups(Fig.3).TheincidenceofRCCwasnotsignificantlydierent

betweenthephlebotomyandnon-phlebotomyanimals.However,thetotalweightofRCtumorswassignificantlyheavierinanimalsofthenon-phlebotomygroupthanin

thoseofthephlebotomygroup(Table2).MacroscopicandmicroscopicpicturesofFe-NTA-

inducedRCCinanon-phlebotomizedratareshowninFig.4.Microscopically,thetumorswerecomposedofgranularorclearcells,and/orrevealedsarcomatoidsubtypes(Fig.4b,c).Fe-NTA-inducedRCCinaphlebotomizedratisshowninFig.5.Generally,thetumorsfromphlebotomizedratsweresmallerthanthosetumorsfromnon-phlebotomizedrats.Themicroscopicpictureswerenotdierentamongtheindividualsofeachgroup(Fig.5b).

Discussion

Fe-NTA,wheninjectedintraperitoneally,isabsorbedfromtheperitoneumintotheportalvein.Aftersomeclearanceofironintheliver,Fe-NTAflowsintothesystemiccirculation.Fe-NTAexistsasfreeironintheserumforatleastafewhoursafterserumtransferrin

EectsofPhlebotomyonIron-inducedRCCAugust2002

Table1 Dierencesinserumiron,TIBC,andhematocritbetweenthe3groupsattheendoftheexperiment

groupsSerumiron(μg/dl)

TIBC(μg/dl)

Hematocrit(%)

Normal 136.50±66.03 634.83±65.11 36.76±3.75Phlebotomy 22.67± 2.89 896.67±82.70 20.33±2.60Non-phlebotomy 77.00±40.89 670.00±88.41 32.59±4.79

P＜0.01vs.untreatedornon-phlebotomygroups. P＜0.05vs.normalornon-phlebotomygroups.

Fig.2 Changesinserumiron,totalironbindingcapacity(TIBC),andhematocritinthephlebotomygroup.(▲)hematocrit;(□)TIBC;(●)serumiron.(average±S.D.).statisticallysignificant(P＜0.01)fromtheinitiationofphlebotomy.

Table2 Incidenceofrenalcellcarcinomaanddierencesintumorweightinanimalsofthephlebotomyandnon-phlebotomygroups

groupsNumberofratsused

Numberofeectiverats

IncidenceofRCC(%)

Tumorweights(g)

Normal 10 10 0(0)Phlebotomy 21 17 9(53) 23.64±18.54Non-phlebotomy 17 16 9(56) 54.40±42.40

P＜0.05vs.non-phlebotomy.

Fig.3 Solubleironinliverandkidneyattheendoftheexperi-ment.Filledcolumn,kidney;opencolumn,liver. P＜0.05vs.untreatedornon-phlebotomygroup.

201

moleculesaresaturated;subsequentexcretionofironbythekidneythentakesplace［19］.SincethemolecularweightofFe-NTAislow,itisfilteredthroughglomerulitotheluminaoftherenalproximaltubules,wherethelipidperoxidationreactionisinduced［20,21,22］.Forlipidperoxidationtooccur,thefollowingconditionsmustbemetintheluminalenvironmentintheproximaltubules:1)anabsenceofalbuminthatworksasnon-specificscavengerintheurine;2)abundantcysteineasanironreductantbythepresenceoftheGSHcycle;3)neutral

pHandaverageionicstrength.ThisenvironmentappearstobeoptimalfortheFenton-likereactiontooccurinvivo［23］.Whenreactiveoxygenspecies-free-radicalsaregener-

atedincloseproximitytoDNA,theycancausepointmutations,DNAcross-linking,andDNAstrandbreaks,andthuspossessthecapacitytoalternormalcells［24］.Recently,Toyokuni’sgroupreportedthespecificalleliclossofthep16 tumorsuppressorgeneasoneoftheearlyeventsleadingtoratrenalcarcinogenesisdueto

a

b cFig.4 RenalcellcarcinomainducedbyFe-NTAinnon-phlebotomizedrats.(a)Alargeright-sidedrenaltumorthatwasobserved18monthsafterthefinalFe-NTAtreatment,andthehistologyoftherenalcellcarcinoma,(b)granularcellcarcinoma,(c)sarcomatoidcarcinoma.

Mizoteetal. ActaMed.Okayama Vol.56,No.4202

iron-mediatedoxidativedamage［25］.Freeradicalscanalsoactivateproteinkinasesandgrowthfactorsandtheirreceptors,whicharebothessentialfortumorpromotion［26］.Although6outof12renalcellcarcinomacaseshadeitherlocalordistantmetastasesaccordingtoprevi-ouslongitudinalreports［14］,wedidnotobserveanymetastasesinthepresentstudy.Thisdiscrepancymightreflectoneormoreofthefollowing4explanations:theuseofdierentWistarstrains(slc:Wistarinaprevious

studyvs.crj:Wistarinthepresentstudy),earlytermi-nationofthepresentstudyat18months,ahigherirondoseinthepreviousstudy,oralowerirondietaftertheinitialFe-NTAtreatmentinthepresentstudy.Freeironisastrongmediatoroffreeradicalproduc-

tion［1,24］.Thepotencyofironasacarcinogenissummarizedinarecentreview［16］.WhenFe-NTAwasincubatedwithprimaryculturerenalcells,invitrotrans-formationoftheculturedcellswasobserved.Thisstrong-lysuggeststhattheeectsofironontargetcellsaredirect［27］.Weinberghasarguedthathostironstatusmayinfluencetheincidenceandintensityofneoplasia,andthatiron-withholdingrepresentsadefenseagainstinfectionandneoplasia［28,29］.Weinberghasfurthermorestressedthegrowthadvantageofneoplasticcells,andalsothesuppressionofthehostdefensesysteminthemilieuofiron overload. Ourresultswerein accord withWeinberg’shypothesis,andgrowthretardationwasobservedunderaniron-restrictedenvironment.Neoplas-ticcellgrowthretardationisalsoexpectedunderiron-restrictedenvironment.Intheculturesystem oftheFe-NTA-inducedRK532cellline［27］;wewillattempttodemonstratethisprocessinournextexperiment.Thereremainssomeconcernthatthephlebotomytreat-mentitselfmightaectvariousmetabolicpathways,andthatthegrowthretardationoftumorsasregardssizecouldbeinterpretedasageneralgrowthretardationaectingbodymassduetothedeclineofthehost’svitalstatus.WhenweconsiderthebodyweightcurveshowninFig.1,itisapparentthattherewasnogrowthretardationamongtheanimalsinourphlebotomygroup;therefore,inthisregards,nogrossnegativeeectcouldhavebeenattributedtothephlebotomyprocedureitself.Recentsuccessfultrials ofiron withholding andphlebotomyinpatientswithchronichepatitisChavebeenencouraging,andthereisnoevidencethatvitalstatuswasnegativelyaectedbyphlebotomy［30,31］.Wefedanimalsaniron-freedietaftertheinitiation

stage.Solubleironintheliverwasnotsignificantlydierentbetweenthephlebotomyandnon-phlebotomygroups,butsolubleironinthekidneywassignificantlydierentbetweenthe2groups. Thismay meanmetabolicallyactiveironintheliverwasusedduringbodygrowth(Fig.1),andwaskeptunderhomeostaticcontrolthroughtheproductionoftransferrin.Solubleironinthekidneymayhavebeenpresentinalocalenvironment,andwassubjecttotheironstatusoftheentirebody.Anon-phlebotomized,iron-richenvironmentisbeneficialfor

a

bFig.5 Renalcellcarcinoma induced byFe-NTA in non-phlebotomizedrats.(a)Asmallright-sidedrenaltumorthatwasobserved18monthsafterthefinalFe-NTAtreatment,andthemicroscopicpictureofarenalcellcarcinoma(b).Thehistologicalpicturesdidnotdierfromthoseobtainedfromnon-phlebotomizedrattumors.Thepicturehereshowsagranularcelltypetumorwithapapillarygrowthpattern.

203EectsofPhlebotomyonIron-inducedRCCAugust2002

tumorgrowth［32］.Therefore,weconcludethatironwithdrawalatthepromotionstageofcarcinogenesiswillretardtumorgrowth.

Acknowledgement.WearegratefultoMr.H.Sakaiforhisrespectfulcareofthelaboratoryanimals.

Reference

1. OkadaS:Iron-inducedtissuedamageandcancer:Theroleofreactiveoxygenspecies-freeradicals.［Review］PatholInt(1996) ,311-332.

2. UmemuraT,SaiK,TakagiA,HasegawaRandKurokawaY:OxidativeDNAdamage,lipidperoxidationandnephrotoxicityinducedintheratkidneyafterferricnitrilotriacetateadministration.CancerLett(1990) ,95-100.

3. ToyokuniS,TanakaT,NishiyamaY,OkamotoK,NakashimaY,HamazakiS,OkadaSandHiaiH:InductionofrenalcellcarcinomainmaleWistarratstreatedwithcupricnitrilotriacetate.LabInvest(1996),239-248.

4. YamaguchiR,HiranoT,AsamiS,SugitaAandKasaiH:Increaseinthe8-hydroxyguaninerepairactivityintheratkidneyaftertheadminis-trationofarenalcarcinogen,ferricnitrilotriacetate.EnvironHealthPerspect(1996) (Suppl3),651-653.

5. ZhangD,OkadaS,YuY,ZhengP,YamaguchiRandKasaiH:VitaminEinhibitsapoptosis,DNAmodification,andcancerincidenceinducedbyiron-mediatedperoxidationinWistarratkidney.CancerRes(1997) ,2410-2414.

6. ToyokuniS,MoriT,HiaiHandDizdarogluM:TreatmentofWistarratswitharenalcarcinogen,ferricnitrilotriacetate,causesDNA-proteincross-linkingbetweenthymineandtyrosineintheirrenalchromatin.IntJCancer(1995) ,309-313.

7. HamazakiS,OkadaS,EbinaY,LiJLandMidorikawaO:EectofdietaryvitaminEonferricnitrilotriacetate-inducednephrotoxicityinrats.ToxicolApplPharmacol(1988) ,500-506.

8. ToyokuniS,OkadaS,HamazakiS,MinamiyamaY,YamadaY,LiangP,FukunagaYandMidorikawaO:Combinedhistochemicalandbiochemicalanalysis ofsex hormone dependence offerricnitrilotriacetate-inducedrenallipidperoxidationinddYmice.CancerRes(1990) ,5574-5580.

9. ToyokuniS,UchidaK,OkamotoK,Hattori-NakakukiY,HiaiHandStadtmanER:Formationof4-hydroxy-2-nonenal-modifiedproteinsintherenalproximaltubulesofratstreatedwitharenalcarcinogen,ferricnitrilotriacetate.ProcNatlAcadSciUSA(1994) ,2616-20.

10. ToyokuniS,LuoXP,TanakaT,UchidaK,HiaiHandLehotayDC:InductionofawiderangeofC(2-12)aldehydesandC(7-12)acyloinsinthekidneyofWistarratsaftertreatmentwitharenalcarcinogen,ferricnitrilotriacetate.FreeRadBiolMed(1997) ,1019-1027.

11. UchidaK,ItakuraK,KawakishiS,HiaiH,ToyokuniSandStadtmanER:Characterizationofepitopesrecognizedby4-hydroxy-2-nonenalspecificantibodies.ArchBiochemBiophys(1995) ,241-248.

12. KimotoT,KoyaS,HinoK,YamamotoY,NomuraY,MicallefMJ,HanayaT,AraiS,IkedaMandKurimotoM:Renalcarcinogenesisinducedbyferricnitrilotriacetateinmice,andprotectionfromitbyBrazilianpropolisandartepillinC.PatholInt(2000) ,679-689.

13. AkiyamaT,HamazakiSandOkadaS:Absenceofrasmutationsandlowincidenceofp53mutationsinrenalcellcarcinomasinducedbyferricnitrilotriacetate.JpnJCancerRes(1995) ,1143-1149.

14. NishiyamaY,SuwaH,OkamotoK,FukumotoM,HiaiHandToyokuniS:LowincidenceofpointmutationsinH-,K-andN-rasoncogenesandp53tumorsuppressorgeneinrenalcellcarcinomaandperitoneal

mesotheliomaofWistarratsinducedbyferricnitrilotriacetate.JpnJCancerRes(1995) ,1150-1158.

15. TanakaT,IwasaY,KondoS,HiaiHandToyokuniS:Highincidenceofalleliclossonchromosome5andinactivationofp15 andp16 tumorsuppressorgenesinoxystress-inducedrenalcellcarcinomaofrats.Oncogene(1999) ,3793-3797.

16. OkadaS:Ironandcarcinogenesisinlaboratoryanimalsandhumans:amechanisticconsiderationandreviewofliterature.IntJClinOncol(1998),191-203.

17. AwaiM,NarasakiM,YamanoiYandSenoS:Inductionofdiabetesinanimalsbyparenteraladministrationofferricnitrilotriacetate.Amodelofexperimentalhemochromatosis.AmJPathol(1979) ,663-673.

18. BrueckmannGandZondekSG:Animprovedmethodforthedetermi-nationofnon-hemiron.JBiolChem(1940) ,23-30.

19. LiuM,OkadaSandKawabataT:Radical-promoting“free”ironlevelintheserumofratstreatedwithferricnitrilotriacetate:Comparisonwithotherironchelatecomplexes.ActaMedOkayama(1991) ,401-408.

20. HamazakiS,OkadaS,ToyokuniSandMidorikawaO:Thiobarbituricacid-reactivesubstanceformationofratkidneybrushbordermem-branevesiclesinducedbyferricnitrilotriacetate.ArchBiochemBiophys(1989) ,348-354.

21. OkadaS,MinamiyamaY,HamazakiS,ToyokuniSandSotomatsuA:Glutathionecycledependencyofferricnitrilotriacetate-inducedlipidperoxidationinmouseproximalrenaltubules.ArchBiochemBiophys(1993) ,138-42.

22. KawabataT,MaY-X,YamadoriIandOkadaS:Iron-inducedapoptosisinmouserenalproximaltubulesafteraninjectionofarenalcarcinogen,iron-nitrilotriacetate.Carcinogenesis(1997) ,1389-

1394.23. ToyokuniSandSagripantiJL:Inductionofoxidativesingle-and

double-strandbreaksinDNAbyferriccitrate.FreeRadicBiolMed(1993) ,117-23.

24. HalliwellBandAruomaOI:DNAandFreeRadicals.EllisHorwood,WestSussex(1993)

25. HiroyasuM,OzekiM,KohdaH,EchizenyaM,TanakaT,HiaiHandToyokuniS:Specificalleliclossofp16 tumorsuppressorgeneafterweeksofiron-mediatedoxidativedamageduringratrenalcarcinogenesis.AmJPathol(2002) ,419-424.

26. DreherDandJunodAF:Roleofoxygenfreeradicalsincancerdevelopment.EurJCancer(1996) A,30-38.

27. KakehashiC,MoriM,KawabataTandOkadaS:Invitrotransforma-tionofratrenalcellsbytreatmentwithferricnitrilotriacetate.ActaMedOkayama(2001) ,97-103.

28. WeinbergED:Ironwithholding:Adefenseagainstinfectionandneoplasia.PhysiolRev(1984) ,65-102.

29. WeinbergED:Rolesofironinneoplasia.Promotion,prevention,andtherapy.BiolTraceElemRes(1992) ,123-140.

30. HayashiH,TakikawaT,NishimuraN,YanoM,IsomuraTandSakamotoN:ImprovementofserumaminotransferaselevelsafterphlebotomyinpatientswithchronicactivehepatitisCandexcesshepaticiron.AmJGastroenterol(1994) ,986-988.

31. KatoJ,KobuneM,NakamuraT,KuroiwaG,TakadaK,TakimotoR,SatoY,FujikawaK,TakahashiM,TakayamaT,IkedaTandNiitsuY:Normalizationofelevatedhepatic8-hydroxy-2’-deoxyguanosinelevelsinchronichepatitisCpatientsbyphlebotomyandlowirondiet.CancerRes(2001) ,8697-8702.

32. WeinbergED:Theroleofironincancer.EurJCancerPrev(1996),19-36.

Mizoteetal. ActaMed.Okayama Vol.56,No.4204