Differences between 2nd and 3rd generation seric parathormone determination methods on mortality in haemodialysis patients

(1)

RevistadelaSociedadEspañoladeNefrología www . r e v i s t a n e f r o l o g i a . c o m

Original article

Differences between 2nd and 3rd generation seric

parathormone determination methods on mortality in haemodialysis patients ^夽

Laura Rodríguez-Osorio

^a,b

, Concepción de la Piedra

^c

, Mercedes Rubert

^c

,

Marta Martín-Fernández

^c

, María Luisa González Casaus

^d

, Carolina Gracia-Iguacel

^a,b

, Jesús Egido

^a,b,e

, Ricardo Villa-Bellosta

^b

, Emilio González Parra

^a,b,e,^∗

aServiciodeNefrología,FundaciónJiménezDíaz,Madrid,Spain

bInstitutodeInvestigaciónSanitaria,FundaciónJiménezDíaz(IIS-FJD),Madrid,Spain

cServiciodeLaboratoriodeBioquímica,FundaciónJiménezDíaz,Madrid,Spain

dServiciodeAnálisisClínicos,HospitalGómezUlla,Madrid,Spain

eServiciodeNefrología,FundaciónJiménezDíaz,UniversidadAutónomadeMadrid,Madrid,Spain

a r t i c l e i n f o

Articlehistory:

Received13January2016 Accepted17November2016 Availableonline23August2017

Keywords:

Haemodialysis Parathormone PTHbio IntactPTH Mortality Cardiovascular

a bs t r a c t

Parathormoneplaysakeyroleincontrollingmineralmetabolism.PTHisconsideredaure- mictoxincausingcardiovasculardamageandcardiovascularmortalityindialysispatients.

TherearetwodifferentassaystomeasurePTHcalled2ndgenerationorintactPTH(iPTH) and3rdgenerationorbioPTH(PTHbio).

Objective:Toevaluatethedifferencesinmortalityofdialysispatientsbetweenbothassays tomeasurePTH,aswellasthepossibleprognosticroleofthePTHbio/iPTHratio.

Methods:145haemodialysispatientswereincludedwith2-yearmonitoringincludingbase- linelaboratorytestandannuallythereafter.

Results:21patientsdiedinthefirstyearand28inthesecond.Nocorrelationwasfound betweenPTH,PTHbioandPTHbio/iPTHratiowithmortality.BothPTHhaveaperfectcorrela- tionbetweenthemandcorrelatesimilarlywithothermoleculesofthemineralmetabolism.

TheextremebaselinevaluesofPTHarethoseofhighermortality.InsurvivalbyiPTHinter- vals(accordingtoguidelinesandCOSMOSstudy),aJcurveisobserved.WheniPTHincreases, the ratio decreases,possibly when increasingfragments no. 1–84. There is no greater

夽Pleasecitethisarticleas:Rodríguez-OsorioL,delaPiedraC,RubertM,Martín-FernándezM,GonzálezCasausML,Gracia-IguacelC, etal.Diferenciasentrelosmétodosdedeterminaciónde2.^ay3.^ageneracióndelaparathormonaséricasobrelamortalidadenelpaciente enhemodiálisis.Nefrología.2017;37:389–396.

∗ Correspondingauthor.

E-mailaddress:[email protected](E.GonzálezParra).

(2)

prognosticapproximationonmortalitywithPTHbiothanPTHi.Therewasalsonodifference inmortalitywhenprogressionratioPTHbio/PTHiwasanalysed.

Conclusions:Wedidn’tfindanyadvantagestousingbioPTHvs.PTHiasamarkerofmortality.

BioPTHlimitsofnormalitymustbereevaluatedbecauseitsrelationshipwithiPTHisnot consistent.Notknowingtheselimitsaffectsitsprognosticvalue.

by-nc-nd/4.0/).

Diferenciasentrelosmétodosdedeterminaciónde2.^ay3.^ageneraciónde laparathormonaséricasobrelamortalidadenelpacienteenhemodiálisis

Palabrasclave:

Hemodiálisis Parathormona PTHbio PTHintacta Mortalidad Cardiovascular

r e s u m en

Laparatohormonatieneunpapelfundamentalenelcontroldelmetabolismomineral.

Ademásesconsideradacomounatoxinaurémicaaloriginarda ñocardiovasculareinfluir enlamortalidadcardiovasculardelpacienteendiálisis.Existendosmétodosdemedición denominadosde2.âgeneraciónoPTHintacta(PTHi)yde3.âgeneraciónobioPTH(PTHbio).

Objetivo:Evaluarlasdiferenciasenlamortalidaddelpacienteendiálisisentreambasformas demedicióndePTH,asícomoelposiblepapelpronósticodesucociente.

Métodos:Seincluyeron145pacientesenhemodiálisisconunseguimientode2a ˜noscon determinaciónanalíticabasalyposteriormentedeformaanual.

Resultados:Veintiúnpacientesfallecieronelprimera ˜noy28elsegundo.Noseencontró correlaciónentrePTHi,PTHbioycocientePTHbio/PTHiconlamortalidad.AmbasPTHtienen unabuenacorrelaciónentreellasycorrelacionandemanerasimilarconotrasmoléculasdel metabolismomineral.LosvaloresbasalesdePTHextremossonlosdemayormortalidad.

EnlasupervivenciaportramosdePTHi(segúnguíasyestudioCOSMOS)seobservauna curvaenJ.AmayoraumentodePTHielcocientedesciende,posiblementealaumentarlos fragmentosno1-84.Noexisteunamayoraproximaciónpronósticasobremortalidadcon PTHbioqueconPTHi.Noseobservandiferenciasenelvalorpredictivodelcocientesobrela mortalidad.Tampocohubodiferenciasenmortalidadcuandoseanalizalaprogresióndel cocientePTHbio/PTHi.

Conclusiones:NoencontramosventajasenlautilizacióndePTHbiosobrelaPTHicomomar- cadordemortalidad.SedebenreevaluarloslímitesdelaPTHbiopuessurelaciónconla PTHinoesconstante.Elnoconoceresoslímitescondicionasuutilidadpronóstica.

by-nc-nd/4.0/).

Introduction

The Parathyroid Hormone (PTH) is a 84 amino acids pro- tein with a established role in the regulation of mineral metabolism; maintarget organs are the bone and kidneys amongothers.Inspiteofbeingahormone,inthecaseofthe renalpatient,itisclassifiedasatrueuremictoxinbecause itincreases progressivelywiththeprogressionchronickid- neydisease (CKD),and itresponsible formultiplesystemic effects.^1,2Itisincludedasoneoftheuremictoxinsthatcause cardiovasculardamage.^3,4

ThecontrolofPTHisimportantforthe adequate treat- ment of the abnormalities of mineral metabolismin CKD.

Thereforemuchofthetherapeuticeffortsaimedatcontrol- lingthese alterationsare dedicated toanadequate control ofPTH.Nevertheless,inhemodialysispatientsPTHisapoor markerof mortality. Onlylarge studies have revealed that high PTH values that are related to higher mortality. And

in thesestudies the results are variable.^4–8 However, other molecules suchas calcium and phosphorus haveshown a betterprognosticsignificanceonmortality,alongwithvita- min D, although withthe later the resultshave been less conclusive.^9,10

Circulating PTH includesa mixtureof peptides suchas the whole molecule 1–84 and smaller fragments resulting fromthecatabolismofPTH,thesearecallednon-1–84orcar- boxyterminalfragments.OnlyPTH1–84(wholeprotein)exert biological activity.Some of the carboxyterminal fragments haveaPTHantagonisticaction.PTHisclearedbythekidney andtheproportionofthesePTHpeptidesvariesaccordingto theCKDstage.ThemethodscurrentlyusedtomeasurePTH inclinical practiceare oftwocategories.Theso-called2nd generationthatmeasurestheintact1–84PTH(PTHi)andmul- titudeoffragmentsno1–84.The3rdgenerationmethod,also calledPTHbio,recentlymarketed,quantifythePTH1–84only, althoughitsuseisnotgeneralizedyet.^11,12

(3)

Inanattempttointegratetheinformationprovidedbay themeasurementofbothmolecules,someauthorshaveadvo- catedtheutilityoftheproportion(ratio)ofthesemoleculesby theratioPTHbio/PTHiasamarkeroftheboneremodeling^13,14 andmortality.¹⁵

In clinical practice, there are currently two criteria to define thenormalvalues ofPTH. Thefirst,and mostclas- sic,isbasedonhistomorphometryasdefinedbytheGuides KDOQI¹⁶ and theGuidelines SEN.¹⁷ Atthat time, basedon bonebiopsies,boneabnormalitieswerepresentifPTHiwas lowerthan150pg/mlorabove300pg/ml.Therangeofnormal- itywere establishedaccordingtorelationship ofiPTH with thelesionsobservedinthebonebiopsy.¹⁸Thesevaluesstill existinmostoftherecommendations.Theothercriterionis mortality.Thereareseveralstudiesthatfindincreasedmortal- itywithhighiPTHvalues;muchhigherthan300pg/ml.Thus, themorerecentKDIGOGuidelines¹⁹definedadequateiPTHas 2–9timesabovethenormalvalues.However,notallnephrol- ogistswerecompletelyconvincedwiththisrecommendation.

TherecentCOSMOSstudyhasfoundthatinalargecohort ofpatientsfollowed for5years,thePTH valuesassociated withlessmortalityarefintherangeof168–674pg/ml,²⁰very similartotheKDIGOrecommendations.Theauthorsofthe COSMOSstudyrecommendmodifyingthereferencevaluesof PTHibasedonbonehistomorphometricstudies,byPTHvalues basedonsurvival.Althoughtherearecontradictoryresults,it seemsthatthemortalityassociatedwithserumPTHdescribes aJshapecurve,withhighermortalityinpatientswithlower andhigherPTHvalues.Butthereisalsosignificantvariability betweenthedifferentPTH2ndgenerationcommercialkits²¹ whichmakesitdifficulttointerprettheresults.

ThevariabilityinPTHmeasurementsposesseveralprob- lems when interpreting their values, and deciding the targetvaluestobe achievedwhentreatingarenal patient.

What would be the normal values?, those based on his- tomorphometric studies?, the ones based on mortality?, both?; whichmethod use to measurethe PTH, 2ndor 3rd generation?

Thepresentstudyevaluatedthedifferencesinmortality predictionusingPTHvaluesobtainedwithmeasurementof 2ndand3rdgenerationinagroupofhemodialysispatients followedfor2years.Inaddition,wehaveevaluatedthePTH- bio/PTHiratioasapossiblemarkerofmortality,incasethe ratioofthecirculatingmoleculescouldbeamarker.Inaddi- tion,wehaveanalyzedtheassociationbetweenmortalityand serum iPTHlevelsrecommended byKDOQI(150–300pg/ml) and the range obtained by COSMOS (168–674pg/ml).

So the values of PTH that best predict mortality were determined.

Material and methods

Thisisatwo-yearprospectivestudyperformedunderdaily clinical practice conditions. The main objective of this studywastodeterminetheassociationofserumPTH with hemodialysispatientsurvival.Itincluded145hemodialysis patientsfromasinglecenter,whounderwentabaselineana- lyticaldetermination,followedbyafollow-upoftwoyears.We determinedclinicalparametersandbaselineanalysisinwhich

PTHwasdeterminedby2differentmethods,2ndand3rdgen- eration,aswellascalcium,phosphorus,alkalinephosphatase andbonealkalinephosphatase,FGF23,25(OH)-vitaminD,and boneremodelingmarkers.ThevaluesofPTH-intact(PTHi)or PTHof2ndgenerationweredeterminedbyanelectrochemi- luminescencemethodusingtheElecsysautoanalyzer(Roche);

this methoduses anantibodydirectedagainstaminoacids 26–32ofthe moleculeand anotheragainstthe aminoacids 55–64. Therefore it does not detect short carboxyterminal fragments, but this method measures PTH 1–84 and long PTHfragments7–84andamino-PTH11(intra-andinter-assay coefficients of variationare <2.5 and <3% respectively and sensitivityof1.2pg/mlmethod).ThePTHivalueswere nor- malizedtotheAllegroPTHiusingtheformulasgivenbythe SpanishSocietyofNephrology(SEN).ThevaluesofPTHof3rd generationorPTH-bioweredeterminedbyanelectrochemi- luminescence method also using the Elecsys autoanalyzer (Roche),whichusesanantibodydirectedagainstaminoacids 1–4 ofthemoleculeandanother againstaminoacids55–64, thus measuring the molecule 1–84 and the amino-PTH11 (coefficientsofintra-andinter-assayvariationare<3and<6%

respectivelyandthesensitivityofthemethodis5.50pg/ml).

NormalvaluesofPTHbweredefinedas50%ofthevalueof PTH.¹⁵

TheratioofPTHbio/PTHimoleculeswasalsocalculatedin ordertodeterminetheproportionofPTHfragments7–84.

Aregistryofdeceasedpatients(ofanycause)wasrecorded atyear1and2,andgeneralanalyticalparametersofmineral metabolism,inflammationandnutritionweredetermined.In boththeannualandthetwo-yearfollow-up,thePTHwasmea- suredusingthe2ndgenerationmethod.

TheADVIACENTAUR2400autoanalyzerwasusedforthe measurements of serum albumin (bromocresol green dye bindingtechnique;intra-andinter-assaycoefficientsofvari- ation were<1.3and <2%and thesensitivityofthemethod was1g/dl).Serumcalcium(arsenazoIIItechnique,coefficients ofintra-andinter-assayvariationare<1.2and<2%,andthe sensitivityofthemethodis0.5mg/dl),serumcreatinine(alka- line picratetechnique, intra- and inter-assay variationare

<3.1and<4%,andthesensitivityofthemethodis0.2mg/dl), serum phosphate (UV phosphomolybdatetechnique; intra- and interassay coefficients ofvariation are <2, 2 and <3%, and thesensitivityofthe methodis0.3mg/dl),total serum proteins(biurettechnique,intra-andinter-assaycoefficients ofvariationare<1.3and<2%,andsensitivityofthemethod is2g/dl),PCR(turbidimetrictechniqueenhancedwithlatex, coefficients of intra- and interassay variation are <4.9 and

<6%, andthe sensitivityofthe methodis0.003mg/dl), and serumalkalinephosphatase(hydrolysisp-nitrophenylphos- phate;Coefficientsofintra-andinterassayvariationare<1.9 and<2.4%,andthesensitivityofthemethodis5U/l).

Serumprealbuminvaluesweredeterminedbyturbidime- try in a SPA plus autoanalyzer, with intra- and interassay coefficientsofvariationbeing<4.4and<6%respectivelyanda sensitivityof0.006g/l.

The values of total vitamin D (25 [OH] vitamin D2 and D3) were determined by anelectrochemiluminescence methodperformedontheElecsysautoanalyzer(Roche)with coefficients ofintra- and inter-assay variation respectively

<7.5%and<8%andsensitivityof3ng/ml.

(4)

Plasma (C-terminal) fibroblast growth factor 23 (FGF-23) wasdeterminedbyanELISAmethod(Immutopics,USA)using 2polyclonalantibodiesdirectedtotheC-terminalfractionof FGF-23(coefficientsOfintra-andinter-assayvariationare<1.7 and<3.5%respectivelyandthesensitivityofthemethodis 1.5RU/ml).

Theboneisoenzymeofalkalinephosphatasewasdeter- minedinserumbyELISA(OSTASEBAP,IDS,UK)(coefficientsof intra-andinter-assayvariationare<4.5and<6.4%respectively andthesensitivityofthemethodis0.7Mg/l).

ThedeterminationoftheC-terminalpropeptideofprocol- lagentypeI(PINP)andthe␤-CrossLaps(␤-CTX)inserumwas performedbyelectrochemiluminescenceintheElecsysauto- analyzer(Roche)(forPINPcoefficientsofintra-andinterassay variationare<2.9and<3.7%respectivelyandthesensitivity ofthemethodis5␮g/l;for␤-CTX,coefficientsofintra-and interassayvariationare<4.6and<4.7%respectivelyandsen- sitivityofthemethodis0.07ng/ml).

Results

Aprospectivestudyincluding145hemodialysispatientswas conducted fortwo year period; 49% were women. Table 1 showsbaselineclinicalandbiochemistrydataofthepatients studied.85.1%ofthe patientswasonsometypevitaminD receptoractivator,and17.9%receivedcinacalcet.

Twentyonepatientsdiedduringthefirstyearfollow-up, and28inthesecondyear.PotentialClinicalandbiochemical variablesanalyticalthatcorrelatewithmortalityareshown inTable2.Theparametersrelatedtomortalitywerealbumin, prealbumin,C-reactiveprotein,ProBNP,creatinine,age,vas- cularaccesstype,andbonealkalinephosphatase.However, thePTHvaluesofboththe2nd(PTHi)andthe3rdgeneration (PTHbio)andtherationPTHi/PTHbioofthepatientswhodied werenotdifferentthaninpatientsthatremainedalive.

Themainobjectiveofthisstudyistodefinetheeffecton themortalityofPTHvaluesofboth2ndand3rdgeneration, astheratiobetweenboth.SecondgenerationPTH(PTHi)and thirdgenerationPTH(PTHbio)haveagoodcorrelationwithin eachother.ThecorrelationsofPTHfrom2ndandthe3rdgen- eration,withothermoleculesofthemineralmetabolismare similar(Table3).

The ratio PTHbio/PTHi decreases as la PTHi values increasessuggestingthatnon1–84fragmentsincreasesec- ondarytoadegradationofthePTH1–84(Fig.1).

Whenpatientswerestratified byPTH intervals,asignif- icant increase in mortality was observed in patients with extremePTHivalues: greater than 678pg/mland less than 150pg/ml, compared to the group of patients with PTH between150and300pg/ml(Fig.2);althoughthemortalityis higheramongpatientswithPTHibetween300and678pg/ml thaninthosewhosePTHiis150–300pg/ml(Fig.2),thisdif- ferencedidnotreachstatisticalsignificance.WiththePTHbio calculatedas50%ofthePTHithedifferencesarenotsignifi- cantinthesameintervals.

Comparison of mortality in the PTH range recom- mended by the current SEN and KDOQI PTHi guidelines (150–300pg/ml) and the valuesestablishedbythe COSMOS study(164–678pg/ml)(Table4),therelativeriskofmortality

Table1–Demographicandclinicalcharacteristics;and, baselinebiochemistryofthestudypopulation.

Variable

Age(years) 65.9±14.6

Dryweight(kg) 65.5±14.2

Vintage(years) Mean3(P25-2;P75-6)

Males(%) 51

Vascularaccess(%)

AVF 60.8

PTFE(politetrafluoroetileno) 19.2

Permanentcatheter 20

Hemodilalysistechnique(%)

Conventional 90

On-line 10

Hypertension(%) 86.9

Diabetesmellitus(%) 23.4

Dyslipemia(%) 35.2

M±SD

Hemoglobin(g/dl) 11.8±1.6

Hematocrit(%) 35.5±4.6

Urea(mg/dl) 117.7±36.5

Proteins(g/dl) 6.5±0.6

Albumin(g/dl) 3.6±0.4

Iron(␮g/dl) 65.4±28

Fesaturationindex(%) 31.7±14.3

Calcium(mg/dl) 9.2±0.7

Phosphate(mg/dl) 4.7±1.6

CO2 21.4±3.5

Totalcholesterol(mg/dl) 158±35

Prealbumin(mg/dl) 33±10

Ferritin(ng/dl) 406±242

Mean(P25;P75) Bonealkalinephosphatase(␮g/l) 29.6(20.7;41.9)

PINP(␮g/l) 255.3(161.2;486.6)

␤-CTX(ng/ml) 1.8(1.3;2.5)

PTH2.^agen(iPTH)(pg/ml) 205(116.5;386.1) PTH3.^agen(bioPTH)(pg/ml) 119.5(76.2;240

RatiobioPTH/iPTH 0.61(0.56;0.68)

25OH-D(ng/dl) 22.9(12.7;34.2)

Creatinine(mg/dl) 7.6(5.6;9.1)

Alkalinephosphatase(UI/l) 111(89;145)

Triglycerides(mg/dl) 146(102;193)

Creactiveprotein(mg/dl) 0.6(0.25;1.97)

ProBNP(pg/ml) 8325(3422;17450)

FGF-23(RU/ml) 788(384;3258)

isonlysignificantwithPTHgreaterthan300pg/mlandless than150pg/ml.ThisimpliesthatpatientswithPTHigreater than300havehighermortalityonlywhenpatientswithPTHi greaterthan678pg/mlareincluded.Inourstudy,iPTHshowed a better predictive value of mortality than bioPTH, taking 150–300 (75–150)pg/ml asa reference, but it is due to the patientswithhighestPTHvalues(Table4).

NodifferenceswereobservedinbioPTH/iPTHratiobetween patientswho diedandthosewho didnot.Likewise,nodif- ferences inmortalitywere observedwhenthe PTHbio/PTHi ratiowasanalyzedbyquintiles,sobioPTH/iPTHratiodoesnot appeartobeamarkerofmortality(Fig.3).ThebioPTH/iPTH valuehigherthan 0.8hasahighermortality,but giventhe smallgroupofpatientsthisdifferenceisnotsignificant.

(5)

Table2–Variablessignificantlyrelatedwithmortality.

Variable Noexitus Exitus pvalue

Age(years)M±DE 62.8±14;9 73.0±10.6 p<0.0001

VintageMe(P25,P75) 3(2;5.5) 4(3;7) p=0.027

Permanentcatheter(%) 9.9 36.4 p<0.0001

Albumin(g/dl)M±DE 3.72±0.48 3.51±0.42 p=0.0138

Prealbmin(mg/dl)M±DE 35.1±9.6 28.1±10 p=0.0008

Bonealkalinephosphatase(␮g/l)Me(P25,P75) 26.8(19.7;31.7) 31.4(22.4;45.5) p=0.037

Creatinine(mg/dl)Me(P25,P75) 8(5.8;9.8) 6.6(5.5;8.4) p=0.018

CRP(mg/dl)Me(P25,P75) 0.25(0.25;1.37) 1.3(0.25;3.4) p=0.0001

ProBNP(pg/ml)Me(P25,P75) 6590(2778;14450) 14,100(6580;34000) p=0.0001

Table3–CorrelationofPTH2.^ay3.^agenerationwithothermoleculesofmineralmetabolism.

Variable iPTH bioPTH

r p r p

25OH-D(ng/dl) −0.034 ns −0.033 ns

FGF-23(RU/ml) 0.296 0.0003 0.228 0.006

Alkalinephosphatase(UI/l) 0.268 0.0012 0.254 0.0022

Bonealkalinephosphatase(␮g/l) 0.342 0.0000 0.338 0.0000

Calcium(mg/dl) −0.001 ns −0.050 ns

Phosphate(mg/dl) 0.326 0.0001 0.295 0.0003

Table4–MortalityinpatientswithPTHwithinreferencevaluesestablishedbyGuidesSENandKDOQI(150–300pg/ml) andCOSMOS(168–674pg/ml)usingiPTHandbioPTHmethods.

Guides iPTH

Exitus No/yes

OR(IC95%) p Guides

bioPTH

Exitus No/yes

OR(IC95%) p

150–300 34/8 75–150 24/6

<150 27/19 2.99(1.14–7.88) 0.024 <75 29/19 2.62(0.90–7.60) ns

>300 27/20 3.15(1.20–8.25) 0.017 >150 35/23 2.63(0.93–7.42) ns

Cosmos iPTH

Exitus No/yes

OR(IC95%) p Cosmos

BioPTH

Exitus No/yes

OR(IC95%) p

168–674 47/22 84–337 56/27

<168 36/20 1.19(0.56–2.50) ns <84 25/15 1.24(0.57–2.74) ns

>674 5/6 2.56(0.71–9.32) ns >337 7/7 2.07(0.66–6.51) ns

Discussion

Inrenal patients, abnormalities ofmineralmetabolismare a risk for cardiovascular disease.²² However, among the

0.8

0.7

*

* **

***

0.6

0.5

0.4

< 75

iPTH (pg/ml)

bioPTH/iPTH

> 600 75−149 150−224

225−299 300−599

Fig.1–ChangesinPTHbio/PTHi,inrelationtotheiPTH values(2.^ageneration)(pg/ml).(*):p<0.05;(**):p<0.01;(***):

p<0.001.YaxisbioPTH/iPTH.XaxisiPTH(pg/ml).

parametersofmineralmetabolismPTHhasarelativelylow weightasamarkerofmorbidityandmortalityintherenal patient.InourstudyweobservedthatPTHisnotamarker ofmortalitywhenanalyzedinconjunctionwithotherfactors

60

* ns

*

ns ns 50 ns

40 30

Mortality (%)

PTH iPTH

* p < 0.05 (as compared with 150₋300) bioPTH 20

10

0 iPTH bioPTH

(0−150) (0−75)

(150−300) (75−150)

(300−678) (150−384)

> 678

> 384

Fig.2–MortalityaccordingtothePTHreferencevalues establishedbySENguidelines,KDOQIguidelinesandthe COSMOSstudy.

(6)

80 70 60 50

ns ns

40 30 20 10

0

(0.4−0.5) (0.5−0.6) (0.6−0.7) Ratio (bioPTH/iPTH)

Mortality (%)

(0.7−0.8) > 0.8

Fig.3–MortalityinrelationtothebioPTH/iPTHratio.Y axis:mortality(%);Xaxis

(0.4–0.5)(0.5–0.6)(0.6–0.7)(0.7–0.8)(0.8–0.9);

ratio(bioPTH/iPTH).

ofgreater influencesuchasage,yearsondialysis,albumin orvascularaccess.However,itisanimportantmodulatorof allalterations ofmineralmetabolism, soitscontrolshould notbeneglected.Asweobserved,PTHband PTHicorrelate verywellwithotherparameterssuchasphosphorus,FGF23, bonealkaline phosphatase and years of dialysis.All these parametersareassociatedwithmorbimortality.Infact,many researchgroupsfindagreaterinterestincombiningseveral parametersofmineralmetabolismasamarkerofmortality.⁹ ThemainobjectiveofthisworkhasbeentodefinePTHasa markerofmortalityandthereferencevaluesthatshouldused indailypractice.

Therelationship betweeniPTH and mortalityis contro- versial.AlthoughmoststudiesfindthatPTHiisamarkerof mortalitysome authors didnot.⁹ Theestablishedrange of 150–300pg/mlwerebasedonhistomorphometriccriteria,but thesevaluesarealsotheinflectionpointsinmortality.The meta-analysisofNatolietal.⁸describedPTHrelatedmortal- ity asaJcurve. WealsoaJ-shaped mortalitycurve where values with lower mortality are within the PTHi range of 150–300pg/ml.

Likeotherauthors,²³wefoundacorrelationbetweeniPTH andbioPTH.Thusonceadjusted,thevaluesofnormalitycould beusedinterchangeably.Inaddition,bothdeterminationscor- relatewiththesameclinicalandanalyticalparameters,soit appearsthatonemethoddoesnothaveanadvantageover theother. Itisknown thatastheglomerularfiltrationrate decreasesthereisanincreaseinnon-1–84fragments.²⁴Itis remarkabletoseethatasthe1–84PTHincreases, thefrag- mentsare alsoincreasedeventoagreater extentthanthe 1–84PTHmoleculeitself.Asnon-1–84fragmentsareantago- nistsof1–84PTHitcouldbeassumedthatatahigheriPTH,the systemiceffects,especiallybone,wouldbereducedbythese molecules.Thisaspectdeservestobeinvestigatedinorderto knowitsrelevance. Thechangesintheproportionofthese fragmentsareimportantwhendefiningthenormalvaluesof bioPTH.

However,inourstudywedidnotobservethatpatientswith alowPTHb/PTHIratiohadalowermortalityduetoanincrease

inthe7–84fragments,suggestingthatthePTH1–84molecule hasarelevantroleinthemortalityofthesepatients.

ThenormalvaluesofbioPTHaredefinedas54%ofiPTHi.¹⁵ However,inourstudy,it wasobservedthatthisproportion variesaccording tothe iPTHvalues.Thatis,athigher1–84 PTHthenon-1–84PTHfragmentsincreasetoagreaterextent thanthe1–84PTHitself.Therefore,thenormalvaluesof1–84 PTHmeasuredby3rdgenerationmethodsinthehemodialysis patientarenotknownatpresent,astheyarenotyetformally establishedinanyguidelinerecommendations.Accordingto ourresults,limitsofbioPTHshouldbenarrow,sinceelevations of1–84PTHareaccompaniedbyanevengreaterproportional increasenon1–84PTHfragments;thisshouldbeconsidered whenestablishingnormalvaluesforbioPTH.Itisnotknown how tointerpretthevaluesofbioPTHaftereliminatingthe distortionfactoroffragmentsnon1–84PTH,sincethereare nostudiestothatend.TheassumptionofbioPTHas50–54%

oftheiPTHdoesnotseemcorrect,asitdoesnotcorrespondto reality.Amulticenterstudyshouldbeperformedwithalarge numberofpatientstodefinethesevalues.

The COSMOSstudy findsthat a iPTH of674ng/ml may beestablishedastheupperlimitofthereferenceintervalin termsofmortality.Likeotherauthors²⁵wedidnotfind,that ahighermortalityinpatientswithiPTHbetween300and674 than inthosewithiPTHbetween150and 300ng/ml.These resultsshouldmakeuscautiousbeforeestablishingthevalue of678ng/mlastheupperreferencevalue.Mortalityissignifi- cantabove678pg/ml,butweshouldneverconsiderthatfigure asasafevalue.Statistical significanceshould notmakeus thinkthatanincreaseiniPTHiupto9timesnormalvalues issafe.SuchahighiPTHvaluemaycausealterationsofthe mineralmetabolismthatmayhavenegativerepercussionin therenalpatient.Thereisaconfidencezonebetween300and 678pg/mlinwhichmortalityislowbutwethinkthatbeing near300issaferthan678pg/ml.

The bioPTH (1–84)/PTHi(1–84 PTH+non 1–84 PTH) ratio hasbeenevaluatedasamarkerofbonehistomorphometric changesandasapredictorofmortality.Howevertheevidence isstillsparse.LikeinthestudybyMelamedetal.,¹⁵ourresults show thatthe ratio isnotassociatedwithmortality. Other authorshavefoundthatratiobioPTH/iPTHisagoodmarker ofmortalityinmenonhemodialysis.²⁶Thosepatientswith highervaluesofratiobioPTH/iPTHratiowouldhaveahigher mortality.

Ourstudyhasseverallimitations.Thesampleofpatients studiedissmall,sincealargergroupofextremePTHvalues isnecessary.Anotherlimitationofourstudyistohaveper- formedbioPTHatbaselineonlyand notbeingrepeatedin successivetimepoints.

Conclusions

AccordingtoourresultsthevaluesofPTHidefinedbyhisto- morphometriccriteriaarewell-establishedlimitsofmortality inhemodialysispatients,the300pg/mlvaluedefinesbetter thecutoffpointofmortalitythanthe674valuedefinedinthe COSMOSstudy,althoughthemortalitybetween150–300and 300–678pg/mlisthesame.The3rdgenerationofPTHmea- surementmethodthat definethe bioPTHdonotaddmore

(7)

precisiontodefinedefiningmortality.Thenormalvaluesof iPTHarenotwelldefined,sotheiPTHandbioPTHmaynot becomparedwiththevaluescurrentlystablished.Theratio bioPTH/iPTHratioisnotagoodmarkerofmortality.

Conflicts of interest

RocheSpainhassponsoredthemeasurementsofbioPTH,and hashadnoroleinthedesignorthepreparationofthispubli- cation.

Acknowledgements

OurthankstoMaribelVillarinoforthehelpwiththedevelop- mentofthestudy.L.R.-O.isaHealthProfessionalonResearch Training“RioHortegar”(CM13/00131),MinistryofEducation, GovernmentofSpain.R.V.B.isaprofessionalwithpostdoctoral contract“SaraBorrell”(CD14/00198)andaproject(SAF2014- 60699-JIN)oftheMinistryofEconomy(MINECO)andFEDER funds.PI14/00386.PI16/01298.FEDERfundsISCIII-RETICRED- inREN/RD06/0016,RD12/0021.

references

1. LevinA,BakrisGL,MolitchM,SmuldersM,TianJ,Williams LA,etal.PrevalenceofabnormalserumvitaminD,PTH, calcium,andphosphorusinpatientswithchronickidney disease:resultsofthestudytoevaluateearlykidneydisease.

KidneyInt.2007;71:31–8.

2. KovesdyCP,AhmadzadehS,AndersonJE,Kalantar-ZadehK.

Secondaryhyperparathyroidismisassociatedwithhigher mortalityinmenwithmoderatetoseverechronickidney disease.KidneyInt.2008;73:1296–302.

3. MoradiH,SicaDA,Kalantar-ZadehK.Cardiovascularburden associatedwithuremictoxinsinpatientswithchronickidney disease.AmJNephrol.2013;38:136–48.

4. NevesKR,GraciolliFG,dosReisLM,GraciolliRG,NevesCL, MagalhãesAO,etal.Vascularcalcification:contributionof parathyroidhormoneinrenalfailure.KidneyInt.

2007;71:1262–70.

5. BlockGA,KlassenPS,LazarusJM,OfsthunN,LowrieEG, ChertowGM.Mineralmetabolism,mortality,andmorbidityin maintenancehemodialysis.JAmSocNephrol.

2004;15:2208–18.

6. LertdumronglukP,LauWL,ParkJ,RheeCM,KovesdyCP, Kalantar-ZadehK.Impactofageonsurvivalpredictabilityof boneturnovermarkersinhemodialysispatients.Nephrol DialTransplant.2013;28:2535–45.

7. FloegeJ,KimJ,IrelandE,ChazotC,DruekeT,deFranciscoA, etal.,AROInvestigators.SerumiPTH,calciumandphosphate, andtheriskofmortalityinaEuropeanhaemodialysis population.NephrolDialTransplant.2011;26:1948–55.

8. NatoliJL,BoerR,NathansonBH,MillerRM,ChiroliS, GoodmanWG,etal.Isthereanassociationbetweenelevated orlowserumlevelsofphosphorus,parathyroidhormone,and calciumandmortalityinpatientswithendstagerenal disease?Ameta-analysis.BMCNephrol.2013;14:88.

9. StrejaE,WangHY,LauWL,MolnarMZ,KovesdyCP, Kalantar-ZadehK,etal.Mortalityofcombinedserum phosphorusandparathyroidhormoneconcentrationsand theirchangesovertimeinhemodialysispatients.Bone.

2014;61:201–7.

10.StevensLA,DjurdjevO,CardewS,CameronEC,LevinA.

Calcium,phosphate,andparathyroidhormonelevelsin combinationandasafunctionofdialysisdurationpredict mortality:evidenceforthecomplexityoftheassociation betweenmineralmetabolismandoutcomes.JAmSoc Nephrol.2004;15:770–9.

11.DelaPiedraC,FernándezE,GonzálezCasausML,González ParraE.Diferenciasenlafuncióndelospéptidos

paratiroideos.¿Quéestamosmidiendo?Nefrologia.

2008;28:123–8.

12.González-CasausML,González-ParraE,Sánchez-GonzálezC, AlbalateM,deLaPiedra-GordoC,etal.Lamenorproporción deparathormonacirculantebiológicamenteactivaendiálisis peritonealnopermiteelajusteintermétododeparathormona establecidaparahemodiálisis.Nefrologia.2014;34:330–40.

13.HerberthJ,BranscumAJ,MawadH,CantorT,Monier-Faugere MC,MallucheHH.IntactPTHcombinedwiththePTHratiofor diagnosisofboneturnoverindialysispatients:adiagnostic teststudy.AmJKidneyDis.2010;55:897–906.

14.KurajohM,InabaM,OkunoS,NagayamaH,YamadaS, ImanishiY,etal.ReductionofwholePTH/intactPTHratioas apredictorofbonemetabolismincinacalcettreatmentof hemodialysispatientswithsecondaryhyperparathyroidism.

OsteoporosInt.2011;22:923–30.

15.MelamedML,EustaceJA,PlantingaLC,JaarBG,FinkNE, ParekhRS,etal.Third-generationparathyroidhormone assaysandall-causemortalityinincidentdialysispatients:

theCHOICEstudy.NephrolDialTransplant.2008;23:1650–8.

16.NationalKidneyFoundation.K/DOQIclinicalpractice guidelines:bonemetabolismanddiseaseinchronickidney disease.AmJKidneyDis.2003;42Suppl.4:S1–201.

17.TorregrosaJV,CannataAndiaJ,BoverJ,CaravacaF,LorenzoV, MartíndeFranciscoAL,etal.Recomendacionesdela sociedadespa ˜noladeNefrologíaparaelmanejodelas alteracionesdelmetabolismoóseo-mineralenlospacientes conenfermedadrenalcrónica(SEN-MM).Nefrologia.2011;31 Suppl.1:1–32.

18.TorresA,LorenzoV,HernándezD,RodríguezJC,Concepción MT,RodríguezAP,etal.Bonediseaseinpredialysis, hemodialysis,andCAPDpatients:evidenceofabetterbone responsetoPTH.KidneyInt.1995;47:1434–42.

19.KidneyDisease:ImprovingGlobalOutcomes(KDIGO) CKD-MBDWorkGroup.KDIGOclinicalpracticeguidelinefor thediagnosis,evaluation,prevention,andtreatmentof ChronicKidneyDisease-MineralandBoneDisorder (CKD-MBD).KidneyIntSuppl.2009:S1–130.

20.Fernández-MartínJL,Martínez-CamblorP,DionisiMP,Floege J,KettelerM,LondonG,etal.,COSMOSgroup.Improvement ofmineralandbonemetabolismmarkersisassociatedwith bettersurvivalinhaemodialysispatients:theCOSMOSstudy.

NephrolDialTransplant.2015;30:1542–51.

21.IxJH,AndersonCA,SmitsG,PerskyMS,BlockGA.Effectof dietaryphosphateintakeonthecircadianrhythmofserum phosphateconcentrationsinchronickidneydisease:a crossoverstudy.AmJClinNutr.2014;100:1392–7.

22.MoeSM,DrüekeT,LameireN,EknoyanG.Chronickidney disease-mineral–bonedisorder:anewparadigm.AdvChronic KidneyDis.2007;14:3–12.

23.InabaM,NakatsukaK,ImanishiY,WatanabeM,MamiyaY, IshimuraE,etal.Technicalandclinicalcharacterizationof theBio-PTH(1-84)immunochemiluminometricassayand comparisonwithasecond-generationassayforparathyroid hormone.ClinChem.2004;50:385–90.

24.HerberthJ,Fahrleitner-PammerA,Obermayer-PietschB, KrisperP,HolzerH,MallucheH,etal.Changesintotal parathyroidhormone(PTH),PTH-(1-84)andlargeC-PTH fragmentsindifferentstagesofchronickidneydisease.Clin Nephrol.2006;65:328–34.

(8)

25.Kalantar-ZadehK,KuwaeN,RegidorDL,KovesdyCP, KilpatrickRD,ShinabergerCS,etal.Survivalpredictabilityof time-varyingindicatorsofbonediseaseinmaintenance hemodialysispatients.KidneyInt.2006;70:771–80.

26.InabaM,OkunoS,ImanishiY,IshimuraE,YamakawaT,Shoji S.IncreasedactivePTH(1-84)fractionasapredictorofpoor mortalityinmalehemodialysispatients.OsteoporosInt.

2013;24:2863–70.

Differences between 2nd and 3rd generation seric parathormone determination methods on mortality in haemodialysis patients

Original article