Adequate iron chelation therapy for at least six months improves survival in transfusion-dependent patients wih lower risk myelodysplastic syndromes

ContentslistsavailableatScienceDirect

Leukemia

Research

j o u r n al ho me p a g e :w w w . e l s e v i e r . c o m / l o c a t e / l e u k r e s

Adequate

iron

chelation

therapy

for

at

least

six

months

improves

survival

in

transfusion-dependent

patients

with

lower

risk

myelodysplastic

syndromes

Michel

Delforge

_,

_Dominik

_Selleslag

_,

_Yves

_Beguin

_,

_Agnès

_Triffet

_,

_Philippe

_Mineur

_,

Koen

Theunissen

_,

_Carlos

_Graux

_,

_Fabienne

_Trullemans

_,

_Dominique

_Boulet

_,

Koen

Van

Eygen

_,

_Lucien

_Noens

_,

_Steven

_Van

_Steenweghen

_,

_Jan

_Lemmens

_,

Pascal

Pierre

_,

_Randal

_D’hondt

_,

_Augustin

_Ferrant

_,

_Dries

_Deeren

_,

_Ann

_Van

_De

_Velde

_,

Wim

Wynendaele

_,

_Marc

_André

_,

_Robrecht

_De

_Bock

_,

_André

_Efira

_,

_Dimitri

_Breems

_,

Anne

Deweweire

_,

_Kurt

_Geldhof

_,

_Wim

_Pluymers

_,

_Amanda

_Harrington

_,

Karen

MacDonald

_,

_Ivo

_Abraham

_,

_Christophe

_Ravoet

a_{UniversitairZiekenhuisGasthuisberg,Leuven,Belgium} b_{AlgemeenZiekenhuisSint-Jan,Brugge,Belgium}

c_{CentreHospitalierUniversitairedeLiègeandUniversitédeLiège,Liège,Belgium} d_{CentreHospitalierUniversitairedeCharleroi,Charleroi,Belgium}

e_{GrandHôpitaldeCharleroi,Charleroi,Belgium} f_{JessaZiekenhuis,Hasselt,Belgium}

g_{CliniquesUniversitairesUCLdeMont-Godinne,Mont-Godinne,Belgium} h_{UniversitairZiekenhuisBrussel,Brussel,Belgium}

i_{CentreHospitalierRégionalCliniqueSt.Joseph,Mons,Belgium} j_{AlgemeenZiekenhuisGroeninge,Kortrijk,Belgium}

k_{UniversitairZiekenhuisGent,Gent,Belgium}

l_{CentreHospitalierRégionaldelaCitadelle,Liège,Belgium} m_{AlgemeenZiekenhuisSt-Augustinus,Wilrijk,Belgium} n_{CliniqueduSudLuxembourg,Arlon,Belgium} o_{AlgemeenZiekenhuisDamiaan,Oostende,Belgium} p_{CliniquesUniversitairesUCLSt.Luc,Woluwe,Belgium} q_{HeiligHartZiekenhuis,Roeselare,Belgium}

r_{UniversitairZiekenhuisAntwerpen,Edegem,Belgium} s_{AlgemeenZiekenhuisImelda,Bonheiden,Belgium}

t_{ZiekenhuisNetwerkAntwerpenMiddelheim,Antwerpen,Belgium} u_{CentreHospitalierUniversitaireBrugmann,Bruxelles,Belgium} v_{ZiekenhuisNetwerkAntwerpenStuivenberg,Antwerpen,Belgium} w_{RéseauHospitalierdeMédecineSocialeBaudour,Baudour,Belgium} x_{JanYpermanZiekenhuis,Ieper,Belgium}

y_{NovartisPharma,Vilvoorde,Belgium} z_{Matrix45,Tucson,AZ,USA}

A_{CenterforHealthOutcomesandPharmacoeconomicResearch,UniversityofArizona,Tucson,AZ,USA} B_{CentreHospitalierdeJolimont,Jolimont,Belgium}

C_{InstitutJulesBordet,Bruxelles,Belgium}

a

r

t

i

c

l

e

i

n

f

o

Articlehistory:

Received17November2013

Receivedinrevisedform3February2014 Accepted5February2014

Availableonline14February2014 Keywords:

Myelodysplasticsyndromes

a

b

s

t

r

a

c

t

Background:Mostpatientswithmyelodysplasticsyndromes(MDS)requiretransfusionsattheriskofiron overloadandassociatedorgandamage,anddeath.Emergingevidenceindicatesthatironchelation ther-apy(ICT)couldreducemortalityandimprovesurvivalintransfusion-dependentMDSpatients,especially thoseclassifiedasInternationalPrognosticScoringSystem(IPSS)LoworIntermediate-1(Low/Int-1). Methods:Follow-upofaretrospectivestudy.Sampleincluded127Low/Int-1MDSpatientsfrom28centers inBelgium.Statisticalanalysisstratifiedbyduration(≥6versus<6months)andqualityofchelation (adequateversusweak).

∗ Correspondingauthorat:IvoAbraham,6159WestSunsetRoad,Tucson,AZ85743,USA.Tel.:+12024873982. E-mailaddresses:[email protected],[email protected](I.Abraham).

http://dx.doi.org/10.1016/j.leukres.2014.02.003

Transfusion

Transfusiondependency Ironoverload

Chelationtherapy Cardiacdeath

Results:Crudechelationratewas63%but88%amongpatientswithserumferritin≥1000␮g/L.Ofthe80 chelatedpatients,70%werechelatedadequatelymainlywithdeferasirox(26%)ordeferasiroxfollowing deferoxamine(39%).Mortalitywas70%amongnon-chelated,40%amongchelated,32%amongpatients chelated≥6m, and30%amongpatientschelatedadequately;withatrendtowardreducedcardiac mortalityinchelatedpatients.Overall,medianoverallsurvival(OS)was10.2yearsforchelatedand3.1 yearsfornon-chelatedpatients(p<0.001).Forpatientschelated≥6morpatientsclassifiedasadequately chelated,medianOSwas10.5years.Mortalityincreasedasafunctionofaveragemonthlytransfusion intensity(HR=1.08,p=0.04)butwaslowerinpatientsreceivingadequatechelationorchelation≥6m (HR=0.24,p<0.001).

Conclusion:SixormoremonthsofadequateICTisassociatedwithmarkedlybetteroverallsurvival.This suggestsapossiblesurvivalbenefitofICTintransfusion-dependentpatientswithlower-riskMDS.

©2014ElsevierLtd.Allrightsreserved.

1. Introduction

Myelodysplasticsyndromes(MDS)areaheterogeneousgroup of hematopoietic stem cell disorders impairing the production of normal mature blood cells and characterizedby cytopenias, cytologicdysplasia,and riskoftransformationtoacutemyeloid leukemia(AML)[1–6].From60%towellover90%ofMDSpatients developanemia[3,7–10],withupto60%ofthesepatients pre-senting with severe anemia (hemoglobin level<10g/dL). The majorityofpatientswillrequireredbloodcell(RBC)transfusions [3] to manage the anemia, prevent anemia-related comorbid-ity,maintain physicalperformance, and improve quality of life [11]. In 40% of patients transfusion is the sole therapeutic option.

Transfusiondependenceisassociated withincreasedrisksof disease progression [1,3,6,7] and death [1,6,7,12]. After about tentotwentytransfusions[5,13,14],transfusion-dependentMDS patientsinvariably develop iron overload at a rate of approxi-mately0.5mg/kg/day[15].Forevery500␮g/Lincreaseinserum ferritinabove1000␮g/L,mortalityriskrisesby30%[6]. Observa-tionalstudieshaveshownthatironchelationtherapy(ICT)lowers ironburden[16–20]andisassociatedwithbettersurvival[21–23]. Inonestudy,mediansurvivaltimeforpatientswithInternational PrognosticScoringSystem(IPSS)LoworIntermediate-1 (Low/Int-1)riskchelatedforatleast6monthswas124monthssincefirst transfusioncomparedto53monthsforthosenotchelated[21]. AstudyincludingalsohigherriskMDSpatientsreportedmedian survivaltimesof74and49months,respectively[22].Inathird study,80%ofpatientsreceivingICTwerestillaliveafter4years comparedto44%ofpatientsnotreceivingICT[23].ICTis recom-mendedintransfusion-dependentpatientsandshouldbeinitiated prophylactically before clinically significant iron accumulation occurs[13]. Patientswithlower-risk MDS andhightransfusion requirementsmay bemorelikely tobenefit fromICT[1,24,25]. The recently published European LeukemiaNet guidelines rec-ommendedICT should be consideredin transfusion-dependent patientswithrefractoryanemia,refractoryanemiawithring sider-oblasts,orMDSwithisolated5qdeletionandaserumferritinlevel higherthan1000ng/mLafterapproximately25unitsofredcells [26].

Werecentlyreportedonaretrospectivestudy(CICL670ABE02, hereafterABE02)evaluatingironstatusanditsmanagementin193 transfusion-dependentMDSpatientsfrom29centersinBelgium [27].Afollow-onstudy(CICL670ABE03,hereafterABE03)aimed to investigate the effect of ICT versus no such treatment on overall survival. While we analyzed the data for all patients, wereport hereonthesubsample ofMDS patientsclassifiedas Low/Int-1atdiagnosisasthesepatientsmaybenefitmostfromICT [1,24,25].

2. Methods

2.1. Designandprocedures

Asafollow-oninvestigationtothepriorABE02study,thepresentABE03study wasamulticenter,retrospective,observational,non-interventionalstudy[27]. Ret-rospectivechartreviewsfortheABE02studywereperformedbetweenSeptember andNovember2008andyielded193patientrecordsfrom29centers.Asavailable frommedicalrecords,datawererecordedforeachpatientfromthetimeof diagno-sisofMDS.Thisperiodrangedfrom0to32yearswithamediantimesincediagnosis of2yearsandamean±SDof3.4±4.0years.

In2010,28ofthesecentersrepresenting186patientsagreedtoparticipatein theABE03follow-onstudy(netlossof7patients).Chartreviewswereconducted betweenOctober2010andMarch2011andcoveredthetwo-yearperiodsincethe endoftheABE02study.Forpatientswhodiedorwerelosttofollow-upduringthe observationalstudyperiod,thechartreviewwasconducteduptotheirlastrecorded timepoint.

Theprimaryendpointwasoverallsurvival(OS),definedastimebetweenMDS diagnosisanddeath(fromanycause),losttofollowup(censored),orendofstudy. ThestudywasapprovedbytheEthicalCommitteeofeachparticipatingcenter. ThestudywasconductedinaccordancewiththeHelsinkiDeclarationoftheWorld MedicalAssociationassubsequentlyamended.

2.2. Ironchelationtherapy

Patientswereconsideredtohavereceivedchelationtherapyiftheyweretreated witheitherdeferoxamineordeferasirox.Nopatientsreceiveddeferiprone.We dif-ferentiatedbetweenpatientschelatedforatleastsixmonthsandthosechelatedfor ashorterperiod.Further,inordertoallowcomparisonbetweenchelationtherapies, andinanalogywithpreviouspublications[21],patientsreceivingICTwere classi-fiedaseither“adequatelychelated”or“weaklychelated”.Thedefinitionsforthese termswereadoptedfromRoseetal.[21]todiscriminatebetweenpatientsreceiving theironchelatorinanoptimal(or“adequate”)versussuboptimal(“or“weak”)way, andthisindependentoftheadministrationrouteorthenoveltyofthedrug.Hence, slowsubcutaneousdeferoxamineinfusionsadministeredonmultipledays/weekor deferasiroxatanydosewasconsidered“adequatechelation”.Alternately, deferox-aminebyanyothermethodofadministration(intravenouslyonceafterorduring eachtransfusionorsubcutaneouslyasabolusinfusionorother)wasconsidered “weakchelation”.

2.3. Variables

Dataasavailablefromthemedicalrecordandusedinthestatisticalanalysis included:patientdemographics(age andgender);currentmedical statusand comorbidities(diabetes,signsofheartfailure,othercardiovasculardisease, abnor-mallivertests,andotherseverecomorbiditieswithpotentialimpactonsurvival)at entryintostudyABE02;historyofMDS(dateofdiagnosis,French-British-American [FAB]and/orWorldHealthOrganization[WHO]classificationatdiagnosis,andIPSS prognosticclassificationatdiagnosis);cytogeneticriskgroupatdiagnosis (conven-tionalkaryotypeorfluorescenceinsituhybridization);previousMDStreatments (growthfactors,immunomodulatoryagents,chemotherapy,andallogeneic trans-plantation);mostrecentredbloodcellhematology(hemoglobin[Hb],transferrin saturation[TSAT]andserumferritinlevels);bloodtransfusionhistory(numberof redbloodcell[RBC]unitssincediagnosis,numberofRBCunitsinpast4months); progressiontoeitheracutemyeloidleukemia(AML)ormoresevereMDS;andiron chelationtherapy(ifany,agent,androuteofadministration).Averagetransfusion intensitywascalculatedastheratiobetweentotalnumberofRBCtransfusions receivedandthedifferenceinmonthsbetweenfirstRBCtransfusionandlastdate recorded.

2.4. Statisticalanalysis

Descriptivestatisticsforcontinuousvariablesincludedmeasuresofcentral ten-dencyanddispersion;fordiscretevariables,counts,percentages,andproportions werecalculated.TestsofsignificanceincludedStudent’st-testforindependent samplesandanalysisofvarianceforrespectively2and≥3subgroupcomparisons oncontinuousvariables;and2_{-basedcontingencyanalysiswith,asapplicable,}

Fisher’sexacttestfordiscretevariables.Bonferroni-classcorrectionswereusedto correctformultiplicity.

ActuarialprobabilityofOSwasapproximatedusingtheKaplan–Meierproduct limitmethod,andcomparisonsbetween/amongKaplan–Meiercurveswere per-formedusingthelog-ranktest.Forthosepatientsaliveattheendofthestudy,data werecensoredatthetimeoftheirlastrecordedvisit.MultivariateCox Proportional-Hazard(PH)analysiswasusedtoidentifyindependentpredictorsofOS,while adjustingforthepotentialconfoundingeffectofage.Covariatessatisfyingthe pro-portionalhazardsassumptionwereincludedinthisanalysis.Statisticalsignificance wassetat˛=0.05two-tailed.StatisticalanalysiswasperformedusingSPSS®_v.20

[28].

3. Results

3.1. Dataavailability

Ofthe186patients,127(88%)wereclassifiedasLow/Int-1at diagnosis.Thesepatientsconstitutedtheanalysisset.Themean timedifferencebetweenpatients’lastdatapointinABE02andthe follow-updatapointinABE03was25.4±1.3months.Ofthe127 patients,55(43%)werestillaliveatthetimeofdatarecordingfor theABE03study,65(51%)haddied,and7(6%)werelostto follow-up.

3.2. CharacteristicsoftheLow/Int-1patients

Table1summarizespatientcharacteristicsandclinicalstatus atentryintoABE02forall127Low/Int-1patients,andcomparing patientswhowere(n=80,63%)andwhowerenotchelated(n=47, 37%);and,amongthe80chelatedpatients,thosechelatedforat least6months(n=62,49%).Themean(±SD)ageatthetimeof MDSdiagnosiswas72(±9.2)years(range40–95)withpatients chelatedfor6monthsormorebeingslightlyyoungerthan non-chelatedpatients(p=0.04).Thedifferenceingenderproportions wasnotstatisticallysignificant.

SomecentersusedtheFAB,sometheWHO,andsomeboth clas-sificationmethods.Perdataavailability,themostcommontypes ofMDSwererefractoryanemiawithoutorwithringed siderob-lasts(57%perWHOand90%perFAB).Similarly,71(85%)ofthe 84patientswithreportedcytogeneticresultsfellinthefavorable riskcategory.Therewerenostatisticallysignificantdifferencesby chelationstatusforWHO,FAB,andkaryotypeclassification.Most patients(77%)hadatleastonecomorbidityofinterestwith cardio-vasculardisease(includingsignsofheartfailure)beingthemost common(76%),withnosignificantdifferencesbychelationstatus. Aboutone-fifth(27or21%)ofpatientshadneverreceiveda spe-cifictreatmentfortheirMDS.Theproportionofuntreatedpatients didnotdiffersignificantly betweenchelated(28%)comparedto non-chelated(17%)patients(p=notsignificant[ns]). Erythropoi-eticgrowthfactorswerethemostcommontreatment(55%)with nostatisticallysignificantdifferencesbychelationstatus(allp=ns); neitherwerethereanystatisticallysignificantdifferencesinthe proportionsofotherMDStreatmentsbychelationstatus(allp=ns), orreceivingnoMDStreatmentatall(allp=ns).

3.3. Anemia,transfusion,andironstatus

Mean±SD Hb concentrations were 9.1±1.9g/dL (range 3.2–13.9) at enrollment into ABE02 and 8.7±1.9g/dL (range 4.8–15.9)attheendofABE03(Table1).Thedifferencesbetween non-chelatedand,respectively,chelated(p=0.004)andchelated ≥6months(p=0.01)patients’meanlastrecordedHblevelwere

Fig.1.Classificationchartofpatientsbychelationstatus,withsubclassificationby durationandqualityofironchelationtherapy.

statisticallysignificant.Onaverage,patientshadreceived13.3±9.5 RBCtransfusionsinthepast4months,withnon-chelatedpatients onaveragehavingbeengivenfewer transfusionsthan chelated patients(p=0.02)orpatientschelated≥6months(≥6m;p=0.04). Thedistributionoftransfusions(0–10,11–20,and21–50)trended toward proportionately fewertransfusions amongnon-chelated versus chelated and chelated ≥6m (both p=0.01). Hence the mean number of units of RBCs transfused was lower for non-chelatedcomparedtochelatedpatients(p=0.001)andespecially patientschelated≥6m(p<0.001).Averagetransfusionintensity permonthwas2.8±2.9,withnosignificantdifferencesbetween non-chelated,chelated,andchelated≥6mpatients(allp=ns).The mostrecentserumferritin levelsaveraged2925±2998and the highestrecordedlevelaveraged3984±4292withnostatistically significantdifferencesbychelationstatus(allp=ns).

3.4. Ironchelationtherapy

The crude chelation rate was 63%, but 88% among patients withserumferritinconcentrations≥1000␮g/L.Thirty-fivepercent receiveddeferoxamine;26.3%weretreatedwithdeferasirox;and for39%deferasiroxfollowingpriordeferoxaminetherapy.

Fig.1presentsaclassificationchartofpatientswithLow/Int-1 MDSbychelationstatus;andwithinthechelatedcategory,by dura-tionandqualityofchelation.Ofthe80chelatedpatients,56(70%) werechelatedadequatelywhile24(30%)werechelatedweakly(see alsoTable2).Intermsofduration,62(78%)weretreatedfor≥6m, ofwhich50(81%)receivedadequatechelation.

3.5. Survivaloutcomes

Intotal,65patients(52%)haddiedsincetheendoftheABE02 study,butproportionatelymoresoamongthenon-chelated(70%) patientscomparedtothechelatedpatients(40%;p=0.002)and thosechelated≥6m(32%)(p=0.001)(Table2).Inaddition,there weresignificantlyfewerpatientdeathsamongadequatelychelated patients(30%)relativetothosechelatedweakly(63%)(p=0.01). Therewasatrendtowardreducedcardiacdeathsinchelatedversus non-chelatedpatients(p=0.05).Themortalityratefrom progres-sionofMDSdiseasedidnotdifferbychelationstatus(allp=ns).

Otherprimarycausesof deatharelistedhereindescending order of frequency among the 127 patients (note that some patientshadmorethanoneidentifiedcauseofdeathhencetotals mayexceedthe totalnumber of deceased patientsin Table2): infection/septicemia(n=7),respiratory(n=6),vascular(n=4), pro-gressivedeterioration(incl.oldage)(n=4),unknowncause(incl.

Table1

CharacteristicsandclinicalstatusofpatientsclassifiedasIPSSLoworIntermediate-1bychelationstatus.

Allpatients (N=127) Non-chelated (N=47) Chelated (N=80) Chelated≥6 months(N=62) P* _P**

AgeinyearsatMDSdiagnosis,mean(SD) 72(9.2) 73(9.0) 71(9.3) 70(9.4) 0.11 0.04 Gender,n(%)

Male 56(44) 20(42) 36(45) 30(48) 0.85 0.57 Female 71(56) 27(57) 44(55) 32(52)

Mostrecenthemoglobinlevel,mean(SD) 8.7(1.9) 9.4(2.2) 8.4(1.6) 8.4(1.7) 0.008 0.02

Missing,n 1 1 0 0

RBCtransfusionsinpast4months,mean (SD) 13.3(9.5) 10.7(8.4) 14.9(9.8) 14.2(8.7) 0.02 0.04 0–10transfusions,n(%) 45(39) 24(56) 21(28) 15(27) 11–20transfusions,n(%) 52(44) 13(30) 39(53) 34(61) 0.01 0.01 21–50transfusions,n(%) 20(17) 6(14) 14(19) 7(13) Missing,n 10 4 6 6

TotalnumberofRBCunitstransfused, mean(SD)

105(92) 70(90) 127(87) 144(91) 0.001 <0.001

Missing,n 4 0 4 4

Averagetransfusionintensitypermonth,a

mean(SD)

2.8(2.9) 2.6(3.5) 2.9(2.5) 2.7(1.2) 0.58 0.85

Missing,n 5 0 5 5

Mostrecentserumferritinlevel,mean(SD) 2925(2998) 3025(2755) 2868(3144) 3027(3455) 0.77 1.00

Missing,n 3 2 1 1

Highestrecordedserumferritinlevel, mean(SD) 3984(4292) 4220(5091) 3862(3849) 3865(3997) 0.69 0.71 <1000␮g/L,n(%) 19(16) 9(22) 10(13) 8(13) 1000–1999␮g/L,n(%) 20(17) 4(10) 16(20) 14(23) 2000–2999␮g/L,n(%) 29(24) 12(29) 17(21) 11(18) 0.27 0.15 3000–3999␮g/L,n(%) 15(12) 6(15) 9(11) 6(10) ≥4000␮g/L,n(%) 38(31) 10(24) 28(35) 23(37) Missing,n 6 6 0 0

IPSSclassificationatdiagnosis,nb

Low 54 18 36 28 0.58 0.56

Intermediate-1 73 29 44 34

WHOclassificationatdiagnosis,nb

RA 27 10 17 14 0.11 0.12 RARS 24 5 19 17 RCMD 14 7 7 5 RSCMD 1 1 0 0 RAEB-I 5 1 4 3 RAEB-II 1 0 1 1 5q-syndrome 15 9 6 5 Unclassified 3 0 3 2

FABclassificationatdiagnosis,nb

RA 60 25 35 26 0.48 0.19 RARS 34 8 26 22 RAEB 7 2 5 2 RAEB-T 1 0 1 1 CMML 2 1 1 0 Karyotype,nb

Favorable(normal,isolated5q-,isolated 20q-,ordeletionY)

71 25 46 36 0.09 0.11

Poor(anychromosomesevenanomaly or≥3aberrations)

7 4 3 2

Intermediate(allotheranomalies) 6 0 6 4 MDStreatmentseverreceived,nc

Erythropoieticgrowthfactors 70 22 48 38 0.20 0.17

Immunomodulatoryagents 32 13 19 17 0.68 1.00

ATG±cyclosporin 3 0 3 3 0.30 0.26

Lowdosechemotherapy 19 5 14 11 0.44 0.41

Intensivechemotherapy 3 1 2 2 1.00 1.00

Allogeneictransplantation 4 2 2 2 0.63 1.00

Hypomethylatingagents 7 5 2 2 0.10 0.24

Othergrowthfactors 14 4 10 8 0.57 0.55

Othertreatments 14 3 11 11 0.25 0.09

Notreatment 27 13 14 10 0.19 0.16

a_{AveragetransfusionintensitywascalculatedastheratiobetweentotalnumberofRBCtransfusionsreceivedandthedifferenceinmonthsbetweenthefirstRBCtransfusion}

andthelaststudydaterecorded.

b _{Onavailabledataandasreportedbyinvestigators.} c _{Patientsmayreceivemorethanonetreatment.}

AML=acute myeloid leukemia; CMML=chronic myelomonocytic leukemia; FAB=French-American-British; IPSS=International Prognostic Scoring System; MDS=myelodysplastic syndrome;RA=refractory anemia; RAEB-I=refractory anemiawith excessof blasts-I; RAEB-II=refractory anemiawith excessof blasts-II; RAEB-T=refractoryanemiawithexcessofblastsintransformation;RARS=refractoryanemiawithringedsideroblasts;RCMD=refractorycytopeniawithmultilineage dysplasia;RSCMD=refractorysideroblasticcytopeniawithmultilineagedysplasia;SD=standarddeviation;WHO=WorldHealthOrganization.

* _{p-Valuefordifferencesbetweenchelatedandnon-chelatedpatients.}

Table2

Ironchelationtherapyandoutcomes.

Ironchelationtherapy,n(%)

Non-chelated 47(37) Chelated 80(63) Weaklychelated 24(30) Adequatelychelated 56(70) Deferoxamine 28(35) Deferasirox 21(26)

Deferoxamineanddeferasirox 31(39) Outcomes,n(%)

AllPatients,n(%)

All(N=127) Non-chelated(N=47) Chelated(N=80) P

Death 65(51) 33(70) 32(40) 0.002

Causeofdeath

Cardiac 8(6) 6(13) 2(3) 0.05

MDSprogression 22(17) 7(15) 15(19) 0.64

Othercausesofdeath 40(32) 21(45) 19(24) 0.02

Patientswhowereeithernotchelatedorchelatedatleast6months,n(%)

All(N=109) Non-chelated(N=47) Chelated>6months(N=62) P

Death 53(49) 33(70) 20(32) <0.001

Causeofdeath

Cardiac 8(7) 6(13) 2(3) 0.07

MDSprogression 16(15) 7(15) 9(14) 1.00

patientswhodiedoutsideofthehospital)(n=3),andtwoeachfor

traumaorgastro-intestinalcauses,deathrelatedtochemotherapy,

ordeathduetoothermalignancies.

TheKaplan–MeiercurvesandassociatedstatisticalresultsforOS

byvariouschelationstratifications(chelatedversusnon-chelated;

adequatelyversusweaklyversusnon-chelated;andnon-chelated

versuschelated≥6m)arepresentedinFigs.2–4.MedianOSwas

10.2 years for chelatedand 3.1years for non-chelated patients (p<0.001). For patients chelated ≥6mor patientsclassified as adequatelychelated,medianOSwas10.5years.MedianOSwas higheramongpatientschelated≥6mversusnon-chelatedpatients (p<0.001); among patients chelated _≥6m versus patients not chelated (p<0.001); among patients adequately versus weakly chelated (p=0.001); and among patients adequately chelated

Fig.2.Overallsurvival(OS)amongnon-chelated(n=47)versuschelatedpatients (n=80).Median(SE)OSwas3.1(0.4)monthsamongnon-chelatedand10.2(1.4) yearsamongchelatedpatients(p<0.001).

Fig.3.Overallsurvival(OS)amongnon-chelated(n=47),weaklychelated(n=24), andadequatelychelated(n=56)patients.Median(SE)OSwas3.1(0.4)yearsamong non-chelated,4.3(0.8)yearsamongweaklychelated,and10.5(2.0)yearsamong adequatelychelatedpatients.

versusthosenotchelated(p<0.001).Mediansurvivaldidnotdiffer significantlybetweenweaklyandnon-chelatedpatients(p=ns).

Preliminary unadjusted univariate analyses identified four parameters witha statisticallysignificantassociation with sur-vival:RBCunitsreceivedintheprecedingfourmonths(p=0.049); averagemonthlytransfusionintensity(p=0.023);adequate chela-tion (p=0.002); and chelated ≥6m (p<0.001). Because of the

Table3

CoxProportional-Hazardsmodelforoverallsurvival.

Parameters Hazardratio 95%CI P

Malegendera _{1.91 1.10–3.31 0.02}

Adequatechelationb _{0.22 0.12–0.41 <0.001}

a_{Referenceisfemalegender.} b_{Referencecategoryisnochelation.}

Fig. 4.Overall survival (OS) among non-chelated (n=47) versus patients chelated>6months(n=62).Median(SE)OSwas3.1(0.4)yearsamongnon-chelated and10.5(2.2)yearsamongpatientschelated>6months(p<0.001).

convergenceofadequatechelationandchelationof≥6minmost patients, the former variable was retained. In the multivariate analysis(Table3),whilecontrollingforage(andthusage-related mortality), male gender was associated with higher mortality risk(HR=1.91,p=0.02)whereasadequatechelationhadastrong mitigatingeffectonmortality(HR=0.22,p<0.001).Similarresults were obtainedwhen chelation ≥6m was substituted by being adequatelychelated(datanotshown).

4. Discussion

In this sample of 127 transfusion-dependent patients with lower-riskMDSwedemonstratethatironchelationtherapy, espe-cially withan adequateregimen and for 6months or more, is associatedwithprolongedsurvival.Patientswhoreceived appro-priateironchelationtendedtoliveinexcessoftenyearsfollowing theirdiagnosisofMDS,comparedtoabout3yearsfornon-chelated patients.Themortalityratewas70%amongnon-chelatedpatients asopposedto32%amongthosechelatedforatleast6months. Mod-elingrevealedthatthesurvivaleffectofadequatechelationwas pronouncedbutalsothatmalepatientstendedtobeathigherrisk forshorterOS.Importantly,therewasatrendforreducedcardiac mortalityinchelatedpatients.

OurfindingsbuilduponandextendthosereportedbyRoseetal. inFrench[21]andmorerecentlybyLyonsetal.[29]inUScenters.In theFrenchstudy[21]ofLow/Int1patients,55%ofpatientsreceived chelationtherapy,allforatleast6months;whereasintheUSstudy oflowerriskMDSpatients[29]48%wherechelated(including32% for6monthsormore).Thesecruderateswerelowerthanthe60% rateobservedinoursample.However,allthreestudiesare consis-tentintheirobservationthatpatientstreatedwithICThavealonger survivaltime.At10.2yearsinourcohortand10.3yearsintheRose study,medianoverall survivaltimes forchelatedpatientswere nearlyidentical,yethigherthanthe8.3yearsreportedbyLyons etal.[29].Thistrendwasalsoevidencedamongpatientschelated foratleastsixmonths:10.5yearsinourstudy,10.3yearsintheRose etal.[21]study,and8.7yearsintheLyonsetal.study[29].These survivaltimesarelongercomparedtothe6.3yearsfroma matched-pairanalysisfrom94 chelatedpatientsreportedbyNeukirchen etal.[22].Althoughthepatientcohortsaredifferent,allfour stud-iesshowbetteroverallsurvivalinchelatedpatients,indicatingthat adequateICTmaybebeneficialintheLow/Int-1MDSpopulation.

Our data arein agreementwith consensus-basedguidelines [23–26,30,31], identifyingtransfusion-dependent Low/Int-1risk MDSpatientsasanimportanttargetpopulationforICT.An addi-tionalobservationwasmaderegardingthedosinganddurationof ICT.OnlypatientsreceivingadequatedosingofICTandtreatedfor atleast6monthshadasignificantoverallsurvivalbenefit. Further-more,despitethehighertransfusionneedinthechelatedpatients serumferritin levelsat theendof theobservationperiodwere similarinthechelatedandthenon-chelatedpatientssupporting theefficiencyofICTinreversingtransfusionalironoverload. How-ever,oneshouldalsotakeintoaccountthecontributiontoiron overloadofincreasedironabsorptionduetoineffective erythro-poiesis,particularlyamonganemicpatients.

Onelimitationofourstudyisapotentialbiasinpatientselection. Forinstance,physiciansmaydecidetoprovideICTonlytopatients withareasonablelifeexpectancy.However,patientgroupswere balanced forMDS subtypes, IPSS,and anti-MDStreatment.The cohortofchelatedpatientshadalowermeanhemoglobinleveland ahighertransfusionneed,diseasecharacteristicsthatareknown riskfactorsforreducedsurvivalinMDS[1,6,7].Onemight prefer-entiallyselectthesepatientsasanimportanttargetpopulationfor ICT.

Despitetheseconcernsandrecognizingthatcausalitycannotbe inferredfromourstudy,theobservationofamorethanthreefold highermediansurvivalofpatientstreatedwithICTinthisparticular cohortofpatientsmakesthevalueofICTevenmorecompelling.Our studythereforeaddsadditionalevidencetothealreadypublished studiesshowingthatadequatechelationoftransfusion-dependent Low/Int-1patientsdoesmorethanjustreducingserumferritin lev-els,butis associatedwithlongersurvivalof patients.However, asobservationaleffectivenessstudiesdonotprovidecausal evi-dence,randomizedtrialsofICTremainanecessity.Inthatregard theresultsoftheongoingTELESTOtrial(NCT00940602)areawaited asprospectiveevidencethatICTreducesmortalityin transfusion-dependentLow/Int-1patients.

Mortalityamongpatientschelatedfor6monthsormorewas 32%comparedto70%amongnon-chelatedpatients.However,this differenceinmortalitywasnotrelatedtoMDSdiseaseprogression. Incontrast,thecrudecardiacmortalityratewas4timeshigherin thenon-chelatedcohort.

AswithABE02,ABE03wasaretrospectiveobservationalstudyin oneEuropeancountry,notaprospectivepopulation-basedcohort studywithrandomsubjectselection.Thestudydidnotexamine thecomparativeeffectivenessofICTregimensandserumferritin wastheonlymarkerofironstatus.Futurestudiesmaybenefitfrom moresensitivemarkerslikenontransferrin-bound ironor labile plasmairon,buttheseareasofyetnotgenerallyestablished mark-ers.

Insummary,wehaveconfirmedthatadequateICT,preferably forsixmonthsorlonger,isassociatedwithmarkedlyprolonged overallsurvivalintransfusion-dependentpatientswithlowerrisk MDS.Thissuggeststhat,intheLow/Int-1risksubpopulationofMDS patients,ICTmayhaveapatientsurvivalbenefit.Patient conve-nienceandcomfortassociatedwithdeferasiroxtherapymustbe consideredinclinicaldecision-making.

Funding

The study was sponsored by Novartis Pharma (Vilvoorde, Belgium).Thestudyconceptwasdevelopedbythesponsorin con-sultationwithexperts.Statisticalanalysiswasdoneindependent fromthesponsor,thoughsponsorwasallowedtoprovideinput. The manuscript was drafted by Abraham and Harrington and subsequently reviewed and revised by Delforge and Selleslag. Sponsorhadrightofreviewandcommentbutalleditorialdecisions

weremadeby theexternal authors.Manuscriptpreparation by AbrahamandHarringtonwasdoneunderaresearchcontractfrom sponsortoMatrix45andtherewasnoadditionalmedicaleditorial assistance.

Conflictofinterest

WP is employed by Novartis Pharma. IA, AH, and KM are employees of Matrix45. By company policy, they are prohib-itedfromowningequity inclientorganizations(exceptthrough mutualfundsorotherindependentcollectiveinvestment instru-ments) or contracting independently withclient organizations. Matrix45 provides similar services to other biopharmaceutical companiesonanon-exclusivitybasis. MD,DS,YB,LN, WW,KT, DD,CR,AF,CGandFThaveservedonadvisoryboardsforNovartis Pharma.Allotherauthorsdeclarenocompetingfinancial inter-ests.

Contributors: Study concept and design done by Abraham, Delforge,MacDonald,Pluymers,Selleslag,Ravoet.Thestudywas enrolledandchartreviewedbyAndré,Beguin,Boulet,Breems,De Bock,Deeren,Delforge,Deweweire,D’hondt,Efira,Ferrant, Geld-hof, Graux, Lemmens, Mineur, Noens, Pierre, Ravoet, Selleslag, Theunissen, Triffet, Trullemans, Van De Velde, Van Eygen, Van Steenweghen,Wynendaele.ThestudyimplementedbyAbraham, MacDonald, Pluymers. Data collected by Harrington, MacDon-ald.StatisticalanalysisdonebyAbraham,Harrington,MacDonald. Manuscript drafted by Abraham, Beguin, Delforge, Harrington, Selleslag.ManuscriptforintellectualcontentreviewedbyAndré, Boulet,Breems,DeBock,Deeren,Deweweire,D’hondt,Efira, Fer-rant,Geldhof,Graux,Lemmens,Mineur,Noens,Pierre,Theunissen, Triffet,Trullemans,VanDeVelde,VanEygen,VanSteenweghen, Wynendaele.

References

[1]Malcovati L.Impactof transfusiondependencyandsecondaryiron over-loadonthesurvivalofpatientswithmyelodysplasticsyndromes.LeukRes 2007;31:S2–6.

[2]SekeresMA.Themyelodysplasticsyndromes.ExpOpinBiolTher2007;7(March (3)):369–77.

[3]JabbourE,KantarjianHM,KollerC,TaherA.Redbloodcelltransfusionsandiron overloadinthetreatmentofpatientswithmyelodysplasticsyndromes.Cancer 2008;112(5):1089–95.

[4]TefferiA.Myelodysplasticsyndromes—many newdrugs,littletherapeutic progress.MayoClinProc2010;85:1042–5.

[5]List AF. Identifying best practices in the monitoring and treatment of ironoverloadinmyelodysplasticsyndrome.HematolOncolClinNorthAm 2006;20(Supplement1):8–14.

[6]MalcovatiL,PortaMG,PascuttoC,InvernizziR,BoniM,TravaglinoE,etal. Prognosticfactorsandlifeexpectancyinmyelodysplasticsyndromesclassified accordingtoWHOcriteria:abasisforclinicaldecisionmaking.JClinOncol 2005;23:7594–603.

[7]BennettJM.Consensusstatementonironoverloadinmyelodysplastic syn-dromes.AmJHematol2008;83(11):858–61.

[8]GreenbergP,CoxC,LeBeauMM,FenauxP,MorelP,SanzG,etal.International scoringsystemforevaluatingprognosisinmyelodysplasticsyndromes.Blood 1997;89:2079–88.

[9]AlessandrinoEP,AmadoriS,BarosiG,CazzolaM,GrossiA,LiberatoLN,etal. Evidence-andconsensus-basedpracticeguidelinesforthetherapyofprimary myelodysplasticsyndromes.AstatementfromtheItalianSocietyof Hemato-logy.Haematologica2002;87:1286–306.

[10]GreenbergPL.Themyelodysplasticsyndromes.In:HoffmanR,BenzE,ShattilS, CohenH,editors.Hematology:basicprinciplesandpractice.3rded.NewYork, NY:ChurchillLivingstone;2000.p.1106–29.

[11]KasnerMT,LugerSM.Updateonthetherapyformyelodysplasticsyndrome. AmJHematol2009;84:177–86.

[12]CazzolaM,MalcovatiL.Myelodysplasticsyndromes—copingwithineffective hematopoiesis.NEnglJMed2005;352:536–8.

[13]BrittenhamGM.Iron-chelatingtherapyfortransfusionalironoverload.NEngl JMed2011;364:146–56.

[14]Porter JB. Practical management of iron overload. Br J Haematol 2001;115:239–52.

[15]HoffbrandAV,TaherA,CappelliniMD.HowItreattransfusionalironoverload. Blood2012;120:3657–9.

[16]GattermannN,FinelliC, PortaMD,FenauxP,GanserA,Guerci-BreslerA, etal.Deferasiroxiniron-overloaded patientswithtransfusion-dependent myelodysplasticsyndromes:resultsfromthelarge1-yearEPICstudy.LeukRes 2010;34:1143–50.

[17]GattermannN,JarischA,SchlagR,BlumenstengelK,GoebelerM,Groschek M, et al. Deferasirox treatment of iron-overloaded chelation-naïve and prechelatedpatientswithmyelodysplasticsyndromesinmedicalpractice: resultsfromtheobservationalstudieseXtendandeXjange.EurJHaematol 2012;88:260–8.

[18]GreenbergPL,KollerCA,CabantchikZI,WarsiG,GlynosT,PaleyC,etal. Prospectiveassessmentofeffectsoniron-overloadparametersofdeferasirox therapyin patients with myelodysplastic syndromes.LeukRes 2010;34: 1560–5.

[19]PorterJ,GalanelloR,SaglioG,NeufeldEJ,VichinskyE,CappelliniMD,etal. Relativeresponseof patientswith myelodysplasticsyndromes and other transfusion-dependentanaemiastodeferasirox(ICL670):a1-yrprospective study.EurJHaematol2008;80:168–76.

[20]CappelliniMD,PorterJ,El-BeshlawyA,LiCK,SeymourJF,ElalfyM,etal. Tail-oringironchelationbyironintakeandserumferritin:theprospectiveEPIC studyofdeferasiroxin1744patientswithtransfusion-dependentanemias. Haematologica2010;95:557–66.

[21]RoseC,BrechignacS,VassiliefD,PascalL,StamatoullasA,GuerciA,etal. Doesironchelationtherapyimprovesurvivalinregularlytransfusedlower riskMDSpatients?Amulticenterstudy bythe GFM.LeukRes2010;34: 864–70.

[22]NeukirchenJ, Fox F, Kündgen A,NachtkampK, StruppC, HaasR, et al. ImprovedsurvivalinMDSpatientsreceivingironchelationtherapy—amatched pairanalysisof188patientsfromtheDüsseldorfMDSregistry.LeukRes 2012;36:1067–70.

[23]LeitchHA.Improvingclinicaloutcomeinpatientswithmyelodysplastic syn-dromeandironoverloadusingironchelationtherapy.LeukRes2007;31(Suppl. 3):S7–9.

[24]GattermannN.Overviewofguidelinesonironchelationtherapyinpatients withmyelodysplasticsyndromesandtransfusionalironoverload.IntJHematol 2008;88:24–9.

[25]SantiniV,AlessandrinoP,AngelucciE,BarosiG,BillioA,DiMaioM,etal. Clin-icalmanagementofmyelodysplasticsyndromes:updateofSIE,SIES,GITMO practiceguidelines.LeukRes2010;34:1576–88.

[26]Malcovati L, Hellström-Lindberg E, Bowen D, Adès L, Cermak J, del Ca ˜nizo, et al. Diagnosis and treatment of primary myelodysplastic syn-dromesinadults:recommendationsfromtheEuropeanLeukemiaNet.Blood 2013,http://dx.doi.org/10.1182/blood-2013-03-492884(prepublishedonline 26.08.13).

[27]DelforgeM,SelleslagD,TriffetA,MineurP,BriesG,GrauxC,etal.Ironstatusand treatmentmodalitiesintransfusion-dependentpatientswithmyelodysplastic syndromes.AnnHematol2011;90:655–66.

[28]IBMCorp.IBMSPSSStatisticsforMac.20.0;2011.

[29]LyonsRM,MarekBJ,PaleyC,EspositoJ,GarboL,DiBellaN,Garcia-ManeroG. Comparisonof24-monthoutcomesinchelatedandnon-chelatedlower-risk patientswithmyelodysplasticsyndromesinaprospectiveregistry.LeukRes 2014;38:149–54.

[30]BowenD,CulliganD,JowittS,KelseyS,MuftiG,OscierD,etal.Guidelinesfor thediagnosisandtherapyofadultmyelodysplasticsyndromes.BrJHaematol 2003;120:187–200.

[31]National Comprehensive Cancer Network. NCCN clinical practice guide-linesinoncology–v2.2010.Myelodysplasticsyndromes; 2010.Available at:https://subscriptions.nccn.org/gllogin.aspx?ReturnURL=http://www.nccn. org/professionals/physiciangls/PDF/mds.pdf[accessed18.07.13].