resistance to recombinant human erythropoietin therapy in ......elísio manuel de sousa costa | v |...

Resistance to recombinant human erythopoietin therapy in haemodialysis patients

Elísio Manuel de Sousa Costa | � |

Resistance to recombinant human erythropoietin

therapy in haemodialysis patients

Elísio Manuel de Sousa Costa

Faculdade de Farmácia

Universidade do Porto

2008


Elísio Manuel de Sousa Costa | �� |


Elísio Manuel de Sousa Costa | �� |

Resistance to recombinant human erythropoietin

therapy in haemodialysis patients

Resistência à terapêutica com eritropoietina humana

recombinante em doentes hemodializados

Dissertation submitted to the Faculdade de Farmácia, Universidade do Porto,

for the degree of Doctor of Philosophy

Dissertação de candidatura ao grau de Doutor apresentada à

Faculdade de Farmácia da Universidade do Porto

Supervisors/Orientadores

Professora Doutora Maria Alice dos Santos Silva G. Martins

Professor Doutor Alexandre Tiedtke Quintanilha

Professor Doutor Luís Filipe Barbosa Amado Belo


Elísio Manuel de Sousa Costa | �V |

Título Resistance to recombinant human erythopoietin therapy in

haemodialysis patients

Autor Elísio Manuel de Sousa Costa

Edição Do Autor

Produção Gráfica MED�SA - Edições e Divulgações Científicas, Lda

�SBN 978-989-20-1447-0

Ano 2008


Elísio Manuel de Sousa Costa | V |

A forma de fazer é ser.

Lao-Tzu


Elísio Manuel de Sousa Costa | V� |


Elísio Manuel de Sousa Costa | V�� |

Declaração

Ao abrigo do artigo 8º do Decreto-Lei nº 388/70, declara-se que fazem parte inte-

grante desta dissertação os seguintes trabalhos já publicados ou em publicação:

Artigos em revistas de circulação internacional com arbitragem científica:

Costa E, Rocha S, Rocha-Pereira P, Castro E, Miranda V, do Sameiro Faria M,

Loureiro A, Quintanilha A, Belo L, Santos-Silva A. (2008). Altered Erythrocyte

Membrane Protein Composition in Chronic Kidney Disease Stage 5 Patients un-

der Haemodialysis and Recombinant Human Erythropoietin Therapy. Blood Purif

26:267-273.

Costa E, Pereira BJ, Rocha-Pereira P, Rocha S, Reis F, Castro E, Teixeira F,

Miranda V, do Sameiro Faria M, Loureiro A, Quintanilha A, Belo L, Santos-Sil-

va A. (2008). Role of Prohepcidin, �nflammatory Markers and �ron Status in Re-

sistance to rhEPO Therapy in Hemodialysis Patients. Am J Nephrol 28:677-683.

Costa E, Lima M, Alves JM, Rocha S, Rocha-Pereira P, Castro E, Miranda V,

Faria MD, Loureiro A, Quintanilha A, Belo L, Santos-Silva A. (2008). �nflam-

mation, T-Cell Phenotype, and �nflammatory Cytokines in Chronic Kidney Dis-

ease Patients Under Hemodialysis and its Relationship to Resistance to Recombi-

nant Human Erythropoietin Therapy. J Clin Immunol 28:268-275.

Costa E, Lima M, Rocha S, Rocha-Pereira P, Castro E, Miranda V, Faria MS,

Loureiro A, Quintanilha A, Belo L, Santos-Silva A. (2008). �L-7 serum levels

and lymphopenia in hemodialysis patients, non-responders to recombinant hu-

man erythropoietin therapy. Blood Cells Mol Dis 41:134-135.


Elísio Manuel de Sousa Costa | V�� |

Costa E, Rocha S, Rocha-Pereira P, Nascimento H, Castro E, Miranda V, Faria MS,

Loureiro A, Quintanilha A, Belo L, Santos-Silva A. (2008). Neutrophil activa-

tion and resistance to recombinant human erythropoietin therapy in hemodi-

alysis patients. Am J Nephrol 28:935-940.

Costa E, Rocha S, Rocha-Pereira P, Reis F, Castro E, Teixeira F, Miranda V,

Sameiro Faria M, Loureiro A, Quintanilha A, Belo L, Santos-Silva A. (2008).

DMT1 (NRAMP2/DCT1) genetic variability and resistance to rhEPO therapy in

chronic kidney disease patients under haemodialysis. Acta Haematologica 120:11-13.

Costa E, Rocha S, Rocha-Pereira P, Castro E, Miranda V, Sameiro Faria M,

Loureiro A, Quintanilha A, Belo L, Santos-Silva A. (2008). Band 3 profile as a

marker of erythrocyte changes in chronic kidney disease patients. The Open Clini-

cal Chemistry Journal 1:57-63.


Sameiro Faria M, Loureiro A, Quintanilha A, Belo L, Santos-Silva A. Effect of

hemodialysis procedure in prohepcidin serum levels in chronic kidney disease

patients. Clin Nephrol, in press.


Loureiro A, Quintanilha A, Belo L, Santos-Silva A. Changes in red blood cells

membrane protein composition during haemodialysis procedure. Renal Failure,

in press.


Sameiro Faria M, Loureiro A, Quintanilha A, Belo L, Santos-Silva A. Cross-talk

between inflammation, coagulation/fibrinolysis and vascular access in Chronic

Kidney Disease Patients under Haemodialysis and recombinant human erythro-

poietin therapy. J Vasc Access, in press.


Elísio Manuel de Sousa Costa | �X |

Comunicações em actas de encontros científicos com arbitragem científica:

Costa E, Rocha S, Rocha-Pereira P, Castro EB, Miranda V, Faria MS, Loureiro A,

Quintanilha A, Belo L, Santos-Silva A (2007). Neutrophil activation markers

in chronic renal failure patients under haemodialysis and recombinant human

eryhtropoietin. Haematologica Supp 92(1):290.


Quintanilha A, Belo L, Santos-Silva A. (2007). Band 3 as a marker of erythro-

cyte changes in chronic renal failure. Haematologica Supp 92(1):295.


Quintanilha A, Belo L, Santos-Silva A. (2007). Fibrinolytic activity in chronic

renal failure patients under haemodialysis and its relationship to recombinant

human erythropoietin resistance. Haematologica Supp 92(1):429-430.

Costa E, Rocha S, Rocha-Pereira P, Castro E, Reis F, Teixeira F, Miranda V,

Faria MS, Loureiro A, Quintanilha A, Belo L, Santos-Silva A. (2007). Altered

erythrocyte membrane band 3 profile in chronic renal failure patients under

haemodialysis. Boletim SPHM 22(3): 20-21.


Faria MS, Loureiro A, Quintanilha A, Belo L, Santos-Silva A. (2007). Fibrino-

lytic activity and vascular access in chronic renal failure patients under haemo-

dialysis. Boletim SPHM 22(3): 22-23.


Faria MS, Loureiro A, Quintanilha A, Belo L, Santos-Silva A. (2008). Higher

fibrinolytic and inflammatory markers are associated with centrel venous cath-

eters use in chronic kidney disease patients under hemodialysis. Port J Nephrol

Hypert Supp 22(1):85.


Elísio Manuel de Sousa Costa | X |

Faria MS, Costa E, Lima M, Alves JM, Rocha S, Rocha-Pereira P, Castro EB,

Miranda V, Loureiro A, Quintanilha A, Belo L, Santos-Silva A. (2008). T-Cell

phenotype and inflammation in chronic kidney disease patients under haemo-

dialysis and its relationship to resistance to recombinant humam erythropoietin

therapy. Port J Nephrol Hypert Supp 22(1):87.

Costa E, Rocha s, Rocha-Pereira P, Reis F, Castro E, Teixeira F, Miranda V,

Sameiro-Faria M, Loureiro A, Quintanilha A, Belo L, Santos-Silva A (2008). DMT1

(NRAMP2/DCT1) genetic variability and resistance to rhepo therapy in chronic

kidney disease patients under haemodialysis. NDT Plus 1. (Supp 2): ii154.

Costa E, Pereira BJG, Rocha-Pereira P, Rocha S, Reis F, Castro E, Teixeira F,

Miranda V, Sameiro FariaM, Loureiro A, Quintanilha A, Belo L, Santos-Silva

A. (2008). Prohepcidin and resistance to recombinant human erythropoietin

therapy in chronic kidney disease patients on hemodialysis. NDT Plus 1. (Supp

2): ii154-ii155.

Costa E, Rocha s, Rocha-Pereira P, Castro E, Miranda V, Sameiro-Faria M,

Loureiro A, Quintanilha A, Belo L, Santos-Silva A (2008). Erythrocyte membrane

protein composition in chronic kidney disease patients under haemodialysis and

recombinant human erythropoietin therapy. NDT Plus 1. (Supp 2): ii155.

Costa E, Lima M, Rocha S, Rocha-Pereira P, Castro E, Miranda V, Sameiro Faria M,

Loureiro A, Quintanilha A, Belo L, Santos-Silva A. (2008). �L-7 serum levels and

lymphopenia in hemodialysis patients, non-responders to recombinant human

erythropoietin therapy. Haematologica 93(S1):313.


Loureiro A, Quintanilha A, Belo L, Santos-Silva A. (2008). Changes in red

blood cells membrane protein composition during haemodialysis procedure.

Haematologica 93(S1):564-565.


Elísio Manuel de Sousa Costa | X� |

Contents/Índice

Acknowledgements/Agradecimentos

Resumo

Abstract

Resumé

Abbreviations/Abreviaturas

Part I - Introduction

1- The kidney

1.1 - Anatomy

1.2 - Kidney function

2- Chronic kidney disease (CKD)

2.1 - Associated pathologies and CKD stages

2.2 - Laboratory evaluation of kidney function

2.3 - Vascular access

2.4 - Haemodialysis procedure

3- CKD patients under haemodialysis

3.1 - Anaemia

3.1.1 - Normal erythropoiesis

3.1.2 - Erythropoietin

3.1.3 - Anaemia and therapeutic use of recombinant

human EPO (rhEPO) therapy in CKD

3.1.4 - Potential therapeutic use of rhEPO

3.2 - �nflammation and leukocyte activation

3.3 - �ron metabolism

3.4 - Oxidative stress

3.5 - Erythrocyte damage

4- Resistance to rhEPO therapy

XV

X�X

XX��

XXV��

XXX�

1

3

3

4

6

6

6

8

9

13

13

13

16

19

21

23

28

32

33

36


Elísio Manuel de Sousa Costa | X�� |

Part II - Aims

Part III – Publications

Paper I

“�nflammation, T-Cell Phenotype, and �nflammatory Cy-

tokines in Chronic Kidney Disease Patients Under He-

modialysis and its Relationship to Resistance to Recom-

binant Human Erythropoietin Therapy” J Clin Immunol

2008; 28:268-275.

Paper II

“�L-7 serum levels and lymphopenia in hemodialysis pa-

tients, non-responders to recombinant human erythro-

poietin therapy” Blood Cells Mol Dis 2008; 41:134-135.

Paper III

“Neutrophil activation and resistance to recombinant hu-

man erythropoietin therapy in hemodialysis patients.”

Am J Nephrol 2008;28:935-940.

Paper IV

“Role of Prohepcidin, �nflammatory Markers and �ron

Status in Resistance to rhEPO Therapy in Hemodialysis

Patients.” Am J Nephrol 2008;28:677-683.

Paper V

DMT1 (NRAMP2/DCT1) genetic variability and resistance

to rhEPO therapy in chronic kidney disease patients un-

der haemodialysis. Acta Haematologica 2008;120:11-13.

39

47

49

59

63

71

81


Elísio Manuel de Sousa Costa | X�� |

Paper VI

Effect of hemodialysis procedure in prohepcidin serum

levels in chronic kidney disease patients. Clin Nephrol, in

press.

Paper VII

“Altered Erythrocyte Membrane Protein Composition in

Chronic Kidney Disease Stage 5 Patients under Haemodi-

alysis and Recombinant Human Erythropoietin Therapy”

Blood Purif 2008;26:267-273.

Paper VIII

“Changes in red blood cells membrane protein compo-

sition during haemodialysis procedure” Renal Failure, in

press.

Paper IX

“Band 3 profile as a marker of erythrocyte changes in

chronic kidney disease patients” The Open Clinical Chemis-

try Journal 2008;1:57-63.

Paper X

“Cross-talk between inflammation, coagulation/fibri-

nolysis and vascular access in Chronic Kidney Disease

Patients under Haemodialysis and recombinant human

erythropoietin therapy” J Vasc Access, in press.

Part IV - Discussion and conclusions

Part V - References

85

91

101

109

119

127

143


Elísio Manuel de Sousa Costa | X�V |


Elísio Manuel de Sousa Costa | XV |

Acknowledgements/Agradecimentos

This work was performed in the Departamento de Bioquímica of

the Faculdade de Farmácia of Universidade do Porto and in �nstituto de

Biologia Molecular e Celular (�BMC) of Universidade do Porto.

� would like to express my gratitude to a number of people for

their contribution to my work:

- As minhas primeiras palavras de agradecimento são natural-

mente endereçadas à Professora Doutora Alice Santos-Silva, orienta-

dora desta dissertação, para lhe agradecer o facto de me ter dado a

previlégio de trabalhar com ela, e também por tudo aquilo que me

ensinou. Gostaria ainda de salientar a minha profunda admiração e

reconhecimento pela forma como me orientou, pelo entusiasmo e in-

teresse que sempre demonstrou, pela sua disponibilidade, pelo seu pro-

fissionalismo e capacidade de crítica construtiva. Para além de toda a

sua amizade e carinho, que sempre demonstrou nos momentos mais

difíceis.

- Ao Professor Doutor Luis Belo, gostaria de lhe agradecer a for-

ma pronta e simpática com que aceitou ser meu co-orientador. Saliento

ainda, com profundo agradecimento, a sua disponibilidade, os conhe-

cimentos que me transmitiu e a forma entusiasmada com que o fez. A

sua motivação contagiante foi sempre algo muito apreciado por mim.


Elísio Manuel de Sousa Costa | XV� |

- Ao Professor Doutor Alexandre Quintanilha, gostaria de lhe

agradecer o previlégio de o poder ter como meu co-orientador. Saliento

ainda o apoio científico, o espírito crítico com que sempre analisou os

resultados obtidos e a disponibilidade sempre demonstrada.

- À Dr. Maria do Sameiro-Faria, ao Dr. Vasco Miranda e ao Dr.

Alfredo Loureiro, pelo apoio clínico que deram no desenvolvimento

deste trabalho, sem o qual este trabalho não teria sido possível. Gos-

taria ainda de sublinhar o entusiasmo demonstrado e os incentivos

constantes.

- À Dr. Susana Rocha, gostaria de lhe agradecer não só a

amizade, como os conhecimentos que me transmitiu, bem como toda

a ajuda laboratorial, sem os quais o meu trabalho teria sido muito mais

difícil.

- À Doutora Margariga Lima, gostaria em primeiro lugar de lhe

agradecer ter-me dado a oportunidade de trabalhar com ela no Labo-

ratório de Citometria de Fluxo do Hospital Geral de Santo António,

assim como, todos os conhecimentos que me trasmitiu.

- À Professora Doutora Petronila Rocha-Pereira, gostaria de lhe

agradecer todo o apoio laboratorial que me proporcionou.

- Ao Professor Doutor Fávio Reis e ao Professor Doutor Frederico

Teixeira, gostaria de lhes agradecer o entusiamo e apoio que me deram

ao longo de todo o trabalho.


Elísio Manuel de Sousa Costa | XV�� |

- Às professoras do Departamento de Bioquímica da Faculdade

de Farmácia da Universidade do Porto, nomeadamente às Professoras

Doutoras Natércia Teixeira, Elisabeth Castro e Elsa Broze da Rocha,

pelo constante incentivo.

- A todos os colegas, colaboradores e funcionários do Laboratório

de Bioquímica da Faculdade de Farmácia da Universidade do Porto em

geral, com quem trabalhei no Laboratório, gostaria de lhes agradecer o

ambiente descontraído e ciêntificamente estimulante que me propor-

cionaram.

- A todos os elementos da equipa de enfermagem das clínicas de

hemodiálise envolvidas, pelo apoio prestado na colheita de amostras,

em particular à Enfermeira Leonilde por toda a sua paciência.

- À Escola Superior de Saúde do �nstituto Politécnico de Bra-

gança, gostaria de agradecer os dias disponíveis que me proporcionou,

sem os quais não teria sido possível desenvolver esta dissertação.

- Ao Prof. Doutor Henrique Lecour, pelo incentivo que me deu

na fase final de elaboração desta dissertação.

- A todos os meus amigos e familiares, gostaria de lhes agradecer

a compreensão e apoio ao longo destes anos de trabalho.

- Agradeço ainda à Fundação para a Ciência e Tecnologia a bolsa

de doutoramento que me atribuiu (SFRH/BD/27688/2006).

- Por último gostaria ainda de agradecer à Roche Farmacêutica e

à Amgen a ajuda financeira disponibilizada.


Elísio Manuel de Sousa Costa | XV�� |


Elísio Manuel de Sousa Costa | X�X |

RESUMO

A anemia é uma das complicações mais frequentes nos doentes

renais crónicos (DRC) em hemodiálise. Esta anemia está associada com

défice de produção de eritropoietina (EPO) pelos rins, que condiciona

uma diminuição na produção de eritrócitos. Nas últimas duas décadas,

o uso de eritropoietina humana recombinante (EPOhr) no tratamento

destes doentes, permitiu corrigir a anemia, reduzir as complicações as-

sociadas com a anemia e melhorar significativamente a sua qualidade

de vida. No entanto, existe uma grande variabilidade na resposta à ter-

apêutica com EPOhr. Cerca de 25% dos doentes necessitam de doses

elevadas e 5 a 10% não respondem a esta terapêutica. O mecanismo

subjacente a esta resistência à terapêutica com EPOhr é ainda contro-

verso e pouco definido.

Com o objectivo de clarificar o mecanismo de resistência à

terapêutica com EPOhr em doentes hemodialisados, estudámos as

alterações associadas, com particular interesse na inflamação, acti-

vação leucocitária, ciclo do ferro, stress oxidativo e lesão eritrocitária.

Foram estudados 63 DRC em hemodiálise e terapêutica com EPOhr

(32 respondedores e 31 não respondedores à terapêutica com EPOhr)

e 26 indivíduos controlo. Em 20 dos DRC (10 respondedores e 10 não

respondedores à terapêutica com EPOhr), foram também colhidas

amostras de sangue imediatamente após a hemodiálise para estudar os

efeitos deste procedimento.

Quando comparados com os controlos, os DRC em hemodiálise


Elísio Manuel de Sousa Costa | XX |

apresentaram linfocitopenia, resultante de uma diminuição da conta-

gem dos linfócitos CD3+ e em que ambos os subtipos de linfócitos T

CD4+ e CD8+ se encontravam diminuídos. Estes linfócitos apresenta-

vam marcadores celulares de estimulação continuada aumentados e

capacidade aumentada para produzir citoquinas associadas com a res-

posta imune do tipo Th1. Adicionalmente, estes doentes apresentavam

marcadores inflamatórios elevados, e aumento na activação dos neu-

trófilos. No que se refere ao estudo do ciclo do ferro, os DRC apresen-

tavam aumento dos níveis séricos de ferritina e prohepcidina, e uma

diminuição na transferrina. Adicionalmente, foram também encon-

tradas alterações na composição proteica da membrana dos eritrócitos

e no perfil da banda 3, sendo a diminuição da espectrina a alteração

mais significativa. Aumento na capacidade antioxidante total (TAS),

na peroxidação lipidica (TBA) e na razão TBA/TAS foram também ob-

servados.

Quando comparamos os dois grupos de DRC, verificamos

que os não respondedores à terapêutica com EPOhr apresentavam

diminuição no número total de linfócitos e nos linfócitos T CD4+, e

aumento nos marcadores inflamatórios e na activação dos neutrófilos.

Não encontramos diferenças significativas nos parâmetros relacionados

com o ciclo do ferro, com excepção do receptor solúvel da transferrina,

que se encontrava aumentado nos não respondedores. Os níveis séricos

de próhepcidina encontravam-se diminuídos nos não respondedores;

no entanto, encontravam-se mais elevados que no grupo controlo. Uma

diminuição acentuada no conteúdo em espectrina, alterações no perfil

de banda 3 [diminuição fragmentos proteolíticos da banda 3 (Pfrag) e

na razão Pfrag/monómero de banda 3] e uma tendência para valores

aumentados de hemoglobina ligada à membrana foram também en-

contrados nos DRC não respondedores à terapêutica com EPOhr.


Elísio Manuel de Sousa Costa | XX� |

Em conclusão, apesar da etiologia à resistência à terapêutica com

EPOhr não estar ainda completamente esclarecida, os nossos resulta-

dos confirmam que a inflamação parece ter um papel muito impor-

tante. Observamoa ainda, uma relação importante entre resistência à

terapêutica com EPOhr com défice funcional em ferro, linfocitopenia

e linfocitopenia T CD4+, níveis plasmáticos aumentados de elastase,

níveis séricos aumentados de interleucina-7, e alterações na estrutu-

ra das proteínas de membrana do eritrócito e no perfil de banda 3.

Mais estudos serão necessários para se entender a associação entre a

inflamação, e resistência à terapêutica com EPOhr e diminuição na

disponibilidade em ferro.


Elísio Manuel de Sousa Costa | XX�� |


Elísio Manuel de Sousa Costa | XX�� |

ABSTRACT

Anaemia is a common complication in haemodialysis patients.

This anaemia is associated to a decreased bone marrow production of

erythrocytes, due to the inability of the failing kidneys to secrete eryth-

ropoietin (EPO). The introduction of recombinant human EPO (rhE-

PO) therapy led to a significant reduction in anaemia and improved

patients’ quality of life. However, there is a marked variability in the

sensitivity to rhEPO, with up to 10-fold variability in dose require-

ments to achieve correction of anaemia. Approximately 5-10% of the

patients show a marked resistance to rhEPO therapy.

To better clarify the mechanism of resistance to rhEPO thera-

py in haemodialysis patients, we studied systemic changes associat-

ed with resistance to rhEPO therapy in haemodialysis patients under

rhEPO therapies, with particular interest on inflammation, leukocyte

activation, iron status, oxidative stress and erythrocyte damage. We

studied 63 chronic kidney disease (CKD) patients under haemodialysis

and rhEPO therapies (32 responders and 31 non-responders to rhEPO

therapy) and 26 healthy volunteers. �n 20 of the CKD patients (10

responders and 10 non-responders to rhEPO therapy), blood samples

were also collected immediately after dialysis to study the effect of the

haemodialysis procedure.

When compared to controls, haemodialysis patients presented

lymphopenia, which results, at least in part, from a decrease in to-

tal circulating CD3+ T-lymphocytes and affect both the CD4+ and the


Elísio Manuel de Sousa Costa | XX�V |

CD8+ T-cell subsets. These lymphocytes presented markers of enhanced

continuous activation state and enhanced ability to produce Th1 re-

lated cytokines. Furthermore, haemodialysis patients presented raised

markers of an inflammatory process, and of an enhanced neutrophil

activation, as showed by the high serum levels of elastase. Concerning to

iron status, patients showed increased ferritin and prohepcidin serum

levels, and a decrease in transferrin. Furthermore, some changes were

observed in erythrocyte membrane protein composition and in band

3 profile, being the decrease in spectrin the most significant change.

Higher plasma levels of total antioxidant status (TAS), lipidic peroxida-

tion (TBA) and TBA/TAS ratio were also found.

When comparing the two groups of patients, we found that

non-responders presented a significant decrease in total lymphocyte

and CD4+ T-cell counts, a more accentuated inflammatory process

and indicators of enhanced neutrophil activation. No significant dif-

ferences were found in serum iron status markers between the two

groups of patients, except for the soluble transferrin receptor, which

was higher among non-responders. Prohepcidin serum levels were

significantly lower in non-responders, but were higher than those in

the control group. An accentuated decrease in erythrocyte membrane

spectrin, alterations in band 3 profile [decrease in band 3 proteolytic

fragments (Pfrag) and in Pfrag/band 3 monomer ratio], and a trend to

higher values of membrane bound haemoglobin were also found in

non-responders patients.

�n conclusion, although the etiology of resistance to rhEPO ther-

apy is still unknown, our work confirms that inflammation seems to

have an important role in its pathophysiology. We also showed that

resistance to rhEPO therapy is associated with “functional” iron defi-


Elísio Manuel de Sousa Costa | XXV |

ciency, lymphopenia and CD4+ lymphopenia, higher elastase plasma

levels, increased interleukin-7 serum levels, and alterations in erythro-

cyte membrane protein structure and in band 3 profile. Further studies

are needed to understand the rise in inflammation with the associated

need in higher doses of rhEPO and the reduced iron availability.


Elísio Manuel de Sousa Costa | XXV� |


Elísio Manuel de Sousa Costa | XXV�� |

RESUMÉ

L’anémie est une des complications les plus fréquentes dans les

malades insuffisants rénales chroniques (�RC) que font hémodialyse.

Cette anémie est associée au déficit de production d’érythropoïétine

(EPO) par les reins fait qui conditionne la diminution de la produc-

tion d’érythrocytes. Dans les deux dernières décennies, l’utilisation

de l’EPO humaine recombinante (EPOhr) dans le traitement de ces

malades, a permis de corriger l’anémie, de réduire les complications as-

sociées à l’anémie et d’améliorer significativement la qualité de vie de

ces malades. Néanmoins, il y a une grande variabilité dans la réponse

à la thérapeutique avec EPOhr, dans laquelle 25% des malades ont be-

soin de doses élevées et entre 5 à 10% de ces malades ne répondent pas

à cette thérapeutique. Le mécanisme qui est associé à cette résistance

reste encore controverse et peut défini.

Avec l’objectif de clarifier le mécanisme de la résistance à la théra-

peutique avec EPOhr dans les malades �RC que font l’hémodialyse, on

a étudié les modifications associées á cette résistance. On a étudié, en

particulier, celles qui sont associées à l’inflammation, à l’activation

leucocytaire, au cycle du fer, au stress oxydant et à la blessure éryth-

rocytaire. Dans ce travail, on a étudié 63 malades �RC que font de

l´hémodialyse et de la thérapeutique avec EPOhr (32 répondants et

31 non répondants à la thérapeutique avec EPOhr) et 26 individus

composaient le group control. Dans un group de 20 malades �RC (10

répondants et 10 non répondants à la thérapeutique avec EPOhr),


Elísio Manuel de Sousa Costa | XXV�� |

on a récolté aussi des échantillons de sang immédiatement après

l’hémodialyse afin d’étudier les effets de cette procédure.

La comparaison entre le group control et les malades �RC que

font de l’hémodialyse a permis démontrer que ces derniers présen-

tent une diminution du nombre de lymphocytes. Cette diminuition

est le résultat d’une diminution du comptage des lymphocytes CD3+

des deux sub-types de lymphocytes T CD4+ et CD8+ qui se trouvent

diminués. Ces lymphocytes présentaient des marqueurs cellulaires de

stimulation continue augmentés et aussi une capacité de produire des

citoquines associées avec la réponse immunitaire type Th1 augmentée.

Supplémentairement, ces malades présentaient des marqueurs inflam-

matoires, et de l’augmentation dans l’activation des neutrophiles. En

ce qui concerne à l’étude du cycle du fer, les malades �RC présentaient

une augmentation des niveaux sériques de la ferritine et de la pro-

hepcidine, et une diminution dans la transferrine. On a aussi trouvé

des modifications dans la composition protéique de la membrane des

erythrocytes et aussi dans le profil de la bande 3 (la diminution de

l’espectrine est la modification la plus significative qui on a trouvé).

L’augmentation de la capacité antioxydante totale (TAS), de la peroxy-

dation lipidique (TBA) et de la raison TBA/TAS ont aussi été démon-

trées.

Si on compare les deux groupes de malades �RC, on peut vérifier

que les non répondants à la thérapeutique avec EPOhr présentaient

une diminution du nombre total de lymphocytes et du nombre de lym-

phocytes T CD4+, et une augmentation dans les marqueurs inflamma-

toires et dans l’activation des neutrophiles. On n’a pas trouvé des dif-

férences significatives dans les paramètres du cycle du fer, à l’exception

du récepteur soluble de la transferrine, qui se trouvait augmentée dans


Elísio Manuel de Sousa Costa | XX�X |

le cas des non répondants. Les niveaux sériques de prohepcidine se

trouvaient diminués dans les non répondants; néanmoins ils se trou-

vaient plus élevés que ceux du group control. La diminution accentuée

du contenu de l’espectrine, les modifications du profil de la bande 3

[diminution des fragments protéolytiques de la bande 3 (Pfrag) et de la

raison Pfrag/monomère de la bande 3], et une tendance pour présenter

des valeurs augmentées d’hémoglobine liée à la membrane ont aussi

été trouvées dans les malades �RC non répondants à la thérapeutique

avec EPOhr.

En conclusion, l’étiologie associée à la résistance à la thérapeu-

tique avec EPOhr ne reste pas encore complètement éclaircie. On a

cependant confirmé que l’inflammation semble avoir un rôle très im-

portant dans la problématique de cette résistance. On a aussi trouvé

une relation entre la résistance à la thérapeutique EPOhr avec le dé-

ficit fonctionnel du fer, la diminution des lymphocytes et des lympho-

cytes T CD4+, l’augmentation des niveaux plasmatiques de l´elastase,

l’augmentation des niveaux sériques d’interleucina-7, et la modifi-

cation des protéines de la membrane de l’érythrocyte et dans le pro-

fil de la bande 3. La complète compréhension de l’association entre

l’inflammation et la résistance à cette thérapeutique et à la diminu-

tion dans la disponibilité en fer, demeure un champ d’investigation

actuel.


Elísio Manuel de Sousa Costa | XXX |


Elísio Manuel de Sousa Costa | XXX� |

Abbreviations/Abreviaturas

AVF Arterious Venous Fistula

BFU-E Burst-forming unit-erythroid

CFU-E Colony-forming unit-erythroid

CKD Chronic kidney disease

CVC Central Venous Catheter

CRP C-reactive protein

DNA Deoxyribonucleic acid

EPO Erythropoietin

EPOR Erythropoietin receptor

F�H Factor inhibiting H�F

GFR Glomerular filtration rate

GM-CSF Granulocyte/monocyte colony-stimulating factor

GPx Glutathione peroxidase

HFE Hemachromatosis

H�F Hipoxia inducible factor

H�V Human immunodeficiency virus

HMWAg High molecular weight aggregates

�FN �nterferon

�L �nterleukin

JAK2 Janus Kinase 2

MBH Membrane bound haemoglobin

MCH Mean cell haemoglobin

MCHC Mean cell haemoglobin concentration


Elísio Manuel de Sousa Costa | XXX�� |

MCV Mean cell volume

MHC Major histocompatibility complex

MW Molecular weight

Pfrag Proteolytic fragments

PHD Prolyl hydroxylase domain

PTFE Polytetrafluoroethylene

RDW Red cell distribution width

rhEPO Recombinant human erythropoietin

RNA Ribonucleic acid

ROS Reactive oxygen species

RP� Reticulocyte production index

SCF Stem cell factor

SOD Superoxide dismutase

STAT5 Signal transducer and activator of transcription 5

s-TfR Soluble transferrin receptor

TAS Total antioxidant status

TBA Thiobarbituric acid

TGF Transforming growth factor

TNF Tumour necrosis factor

VHL Von Hippel-Lindau tumour suppresso protein


Elísio Manuel de Sousa Costa | 1 |

Part I - Introduction



1 - The Kidney

1.1 - Anatomy

The kidneys are located on the posterior wall of the abdomen,

one on each side of the vertebral column at the level of the T12-L3

vertebrae. The right kidney usually lies slightly inferior to the left kid-

ney, probably owing to its relationship with the liver. Each kidney has

roughly 10 cm in length, 5 cm in width and 2.5 cm in thickness. The

kidneys are connected with the diaphragm, which separates them from

the pleural cavities. More inferiorly, the posterior surfaces of the kid-

ney are related to the quadratus lumborum muscle. The right kidney

is separated from the liver by the hepatorenal recess. The left kidney is

connected to the stomach, spleen, pancreas, jejunum, and descending

colon (Fox, 2008; Brenner, 2000).

The kidneys are encased in a fibrous capsule called the renal

capsule (Fox, 2008), and are composed by two regions, the renal cor-

tex and medulla (Brenner, 2000) (Fig. 1).

Fig. 1- Kidney structure (adapted from Fox, 2008).



The basic structure of the kidneys is the nephron. There are about

one million of these microscopic structures in the kidney, which are

responsible for filtering the blood to remove waste products (Brenner,

2000). Each nephron is constituted by a glomerulus and a tubule (Fig.

2). The renal cortex is where the glomerolus, proximal tubules, and

distal tubules are found. �n the renal medulla we can found the loop of

Henle, and the collecting tubules (Fox, 2008).

Fig. 2 - Basic structure of the kidneys – nephron (adapted from Fox, 2008).

1.2 - Kidney function

The kidney has a considerable number of functions. The most

important are toxin and drug removal, fluid balance, electrolyte and

acid-base balance, blood pressure control and endocrine function.

The kidneys remove excess water, salts, and wastes of protein

metabolism from the blood, while returning nutrients and chemicals to

the blood. Blood enters and exits the glomerulus through the afferent

and the efferent arterioles. While in the glomerulus, the blood pressure



conditions water and various dissolved substances to be filtered out to

the glomerular capillaries into the glomerular capsule as the glomeru-

lar filtrate. The total rate of glomerular filtration, typically, is about 180

liters per day (Brenner, 2000). Most of this volume is returned to the

bloodstream by a process referred as reabsorption. Reabsorbed sub-

stances include water, glucose and other nutrients, sodium and other

ions. Reabsorption begins in the proximal convoluted tubules and con-

tinues in the loop of Henle, distal convoluted tubules and collecting

tubules. Another process called renal secretion is responsible for the

movement of substances and fluids into the distal and collecting tu-

bules, from blood in the capillaries around these tubules. Urine from

the various collecting ducts drains into the renal pelvis, ureter, and

bladder (Fox, 2008).

The waste products descarted by the kidney are generated from

normal metabolic processes, including the breakdown of active tissues,

ingested foods, and other substances (Brenner, 2000).

The kidney also plays an important role in the regulation of the

blood levels of various minerals, such as calcium, sodium, and potas-

sium, and in the production of hormones and enzymes, namely 1,25

(OH)2 – vitamin D3, erythropoietin (EPO) and renin, all playing im-

portant functions in the body (Levin and Li, 2005; Klemmer, 2005).

The suprarenal glands also have endocrine functions. The adrenal cor-

tex secretes corticosteroids and androgens, and the adrenal medulla

secretes catecholamines (mostly epinephrine).



2- Chronic kidney disease (CKD)

2.1 - Associated pathologies and CKD stages

CKD is a gradual and usually permanent loss of kidney func-

tion over time, usually months to years. CKD can result from primary

diseases of the kidneys themselves. However, the major causes are sec-

ondary to other diseases, as diabetes, high blood pressure, inflamma-

tion, post infectious conditions, systemic lupus erythematosus, use of

analgesics over long periods of time, human immunodeficiency virus

(H�V) infection, sickle cell disease, heroin abuse, amyloidosis, chronic

kidney infections, and certain cancers (Lenz and Fornoni, 2006). CKD

is usually divided in five stages according to kidney function (Table �).

Table I: Stages of chronic kidney disease, based on kidney function.

Stage Description GFR* (mL/min/1.73m2)

1 Slight kidney damage with normal More than 90

or increased filtration

2 Mild decrease in kidney function 60-89

3 Moderate decrease in kidney function 30-59

4 Severe decrease in kidney function 15-29

5 Kidney failure requiring dialysis or transplantation Less than 15

*GFR: glomerular filtration rate.

2.2 - Laboratory evaluation of kidney function

CKD can cause no symptoms in its early stages. However, even

in these stages the disease can be detected by routinely laboratory tests.

The presence in urine of proteins, such as albumin, red and white blood

cells, casts and crystals (solids), and increased creatinine levels, can be

used as good indicators of CKD.



Creatinine and urea concentrations in the blood are the most

commonly used blood tests for CKD screening, and for disease moni-

torization. Creatinine is a product of normal muscle breakdown and

urea is the waste product of aminoacid and basic amine breakdown.

The amount of creatinine and urea excreted in the urine together with

their blood levels can be used to calculate the level of kidney func-

tion and the glomerular filtration rate (GFR). The normal GFR is about

100-140 mL/min in men and 85-115 mL/min in women (Lenz and

Fornoni, 2006; Taskapan, 2008). Based in the GFR, the patients are

classified in one of the five stages of CKD (Table �). The blood levels of

these substances rises as kidney function worsens.

CKD is also associated to umbalances in electrolytes, especially

in potassium, phosphorus, and calcium. Hyperkalemia is of particular

concern due to the higher risk of cardiac dysrhythmias. The acid-base

balance of the blood is usually disrupted as well. The low levels of cal-

cium in the blood results from the decreased production of the active

form of vitamin D, produced by hydroxylation of hydroxyvitamin D in

the kidney (Klemmer, 2005). The incapacity for phosphorus excretion

leads to increased levels in blood (Taskapan, 2008).

The presence of anaemia is also a common finding in these pa-

tients due, mainly, to a deficient EPO production. This condition can

also be associated with iron deficiency and with other nutritional defi-

ciencies, namely in vitamin B12 and folates (Foley, 1996).

Ultrasound and kidney biopsy are often used in the diagnosis of

CKD. The first one is usually used to detect alterations in kidney size,

which can be shrunken or enlarged in these patients, and to detect the

presence of urinary obstruction. Kidney biopsy is used essentially in

cases in which the cause of the kidney disease is unclear (Fox, 2008).



2.3 - Vascular access

A successful haemodialysis procedure requires a functional vas-

cular access. Unfortunately, no major advances in the field of haemo-

dialysis vascular access have been observed for the past three decades.

This, has probably contributed to the fact that vascular access dysfunc-

tion is one of the most important causes of morbidity in the haemodi-

alysis population (Santoro, 2000). Access-related problems are respon-

sible for 50% of the hospitalizations of haemodialysis patients (�fudu,

1996).

Nowadays, there are three main forms of haemodialysis vascu-

lar access: arteriovenous fistula (AVF), polytetrafluoroethylene (PTFE)

graft, and central venous catheter (CVC). Each of these forms has its

own specific advantages and problems. AVF is the first choice for vas-

cular access in patients under chronic haemodialysis, considering the

low associated risk of infection and thrombosis. For this type of vascu-

lar access two major complications are more frequently referred: initial

failure to mature (primary nonfunction) and a later venous stenosis

followed by thrombosis (Schwab, 1999). PTFE graft is easy to place and

ready to use; however, extremely high rates of stenosis cases, throm-

bosis and infection are reported for this vascular access (Miller, 2000).

CVC is the least desirable method of haemodialysis access. �n a sense,

every CVC placement represents a failure, namely, in preparing a na-

tive AVF or in creating a suitable alternative. Nevertheless, catheters

are an unavoidable need for many patients who do not have func-

tional AVF access for any unknown reason. �n the recent years, the use

of CVC increased in both acute and chronic uremic patients, and this

may underlie the increased morbidity and mortality observed in these

patients. The type of insertion and the management of CVC may lead

to several complications, namely infections and thrombosis (Mandolfo,

2002).



2.4 - Haemodialysis procedure

The failure of renal clearance of both small solutes and large

molecular weight compounds is associated to a progressive renal de-

terioration, and may lead to multiple potentially lifethreatening com-

plications, namely blood volume overload, hiperkalemia, metabolic

acidosis, fatigue, encephalopathy and malnutrition.

The goals of haemodialysis are to treat uremic symptoms by re-

moving toxic metabolites, to correct acid-base and electrolyte distur-

bances and to maintain blood volume status. This procedure improves

patient’s quality of life, lowers morbidity and mortality rates, and main-

tains nutritional stability of CKD patients. To reach these goals, stage

5 CKD patients with GFR less than 10 to 15 mL/min undergo thrice-

weekly haemodialysis. Other options for chronic renal replacement

therapy include peritoneal dialysis and renal transplantation (National

Kidney Fundation/K-DOQ1, 1997).

Georg Haas performed the first human dialysis in a patient with

acute renal failure, in 1924. Afterwards, haemodialysis was used spo-

radically in the treatment of patients with acute renal failure (Gott-

schalk and Fellner, 1997; Kolff, 2002). By that time, the most impor-

tant limitations of the haemodialysis therapy included clotting of blood

in the extracorporeal circuit and difficulty in obtaining unfailing vas-

cular access. The purification of heparin, in 1943, and the development

of a teflon arteriovenous shunt, in 1960 (Blagg, 1999; Scribner, 2002),

led to a general applicability of haemodialysis to patients with CKD.

Haemodialysis procedure involves the movement of solutes and

solvent (water) across a semipermeable membrane (dialyser). Three

major mechanisms govern the movement of molecules in this proce-

dure: diffusion, ultrafiltration and convection. The description of these



mechanisms is summarized in Table �� (Ronco and Clark, 2001; Bowry,

2002; Locatelli, 2002; Spalding and Farrington, 2003).

Table II- Mechanisms associated to solute and solvent removal during haemo-

dialysis.

Mechanism Description

Diffusion Solute movement, across the semipermeable membrane, from the

side of higher concentration to the side of lower concentration.

The diffusion level is dependent on size, shape and charge of the

molecules.

�t is efficient in removing small molecular weight species, such as

electrolytes.

Ultrafiltration Movement of solvent across the semipermeable membrane, from

high to low pressure.

�t is responsible for the removal of the excess of body water.

Convection As the solvent moves down under a pressure gradient, the dissolved

solutes are dragged across the membrane.

Removal is dependent on the sieving coefficient of the membrane.

�t is responsible for the removal of small and middle molecular

weight species.

�n the haemodialysis circuit (Fig. 3), the blood from the patient

flows on one side and the dialysate flows on the other side of the di-

alysis membrane, where the exchange of molecules occurs. The di-

alyser consists of thousands of hollow capillary fibres, composed of

a biocompatible synthetic semipermeable membrane, that create a

large contact surface area (up to 2 m2). This membrane acts as a filter

and prevents bacteria from entering in blood (dialysate is not sterile).



However, smaller molecular weight contaminants in the water system

can pass into the patient’s blood, namely, copper, aluminium, chora-

mines, and endotoxins (Hoenich and Levin, 2003; Association for the

advancement of medical instrumentation, 2004).

�nitially, haemodialysis membranes were cellulose based and

were developed as a variation of this, with a varying predominant

number of acetate molecules per hexose repeat. Subsequently, new

membranes were developed using more synthetic compounds, such as

polyacrilonitryl and/or polysulphone. The cellulosic membranes were

predominantly low-flux, a term referring to the porosity of the mem-

branes, and presented a molecular weight (MW) cut-off below 5KDa.

Fig. 3 - Diagram of a traditional haemodialysis circuit. During haemodialysis, blood

flows through the arteriovenous access from the patient to the dialyser. Along the circuit,

arterial and venous pressure monitorization is performed. A blood pump moves blood

throughout the circuit, and an air detector ensures that no air enters in the circuit (adapted

from Rosner, 2005).



The synthetic membranes could be manufactured in a low-flux or a

high-flux format (European Best Practice Guidelines, 2002). The latter

allows the clearance of larger molecules, such as beta-2 microglobulin

(MW 11.8 KDa).

The dialysis machine circuit also includes a rotatory blood pump,

pressure monitors, air bubble detectors, a pump for heparin dosing,

and various other safety monitors. During the haemodialysis proce-

dure, the patients can be exposed to as much as 100L of dialysate (Ros-

ner, 2005).

Haemodialysis, usually, takes place in clinicals and hospitals,

lasts for three to five hours, and is performed three times a week.

There is the possibility to perform haemodialysis at home; long noctur-

nal (overnight) haemodialysis or short daily haemodialysis, as tradi-

tionally, three times a week (Rosner, 2005).

The clearance, reflecting the volume of blood or plasma that is

cleared of a specific solute in a unit of time, is used to determine the

dialysis adequacy. Usually, blood urea nitrogen serves as the marker of

clearance, since this solute is easy to measure, and its levels correlate

with uremic symptoms (Depner, 1991). Blood urea nitrogen is also an

indicator of protein intake and a marker of nutritional adequacy (Gotch

and Sargent, 1985). However, as blood urea nitrogen is not specific for

dialysis adequacy, two measures are usually calculated considering the

decrease in blood urea nitrogen that occurs during haemodialysis: urea

reduction rate and KT/V (K: clearance of urea during the treatment, T:

duration of the dialysis session, and V: volume of distribution of urea

or total body water) (Gotch and Sargent, 1985; Depner, 1991). Other

signs and symptoms are indicative of inadequate haemodialysis, name-

ly, low serum albumin levels, nausea, vomiting, declining functional



status, encephalopathy, peripheral neuropathy, pericarditis, persistent

volume overload and hypertension (Daugirdas, 2003; Barros, 2006).

Haemodialysis increases longevity of patients with end-stage re-

nal disease, by removing the metabolic end products and excess of wa-

ter. Despite the technologic development of haemodialysis procedures

and of medical support in the last years, the mortality and morbidity

of these patients remain high, about 10 to 20 times higher than that

found in general population (Foley, 1998).

3- CKD patients under haemodialysis

There are several complications associated to CKD patients un-

der haemodialysis, namely, anaemia, inflammation, leukocyte activa-

tion, oxidative stress, disturbances in iron metabolism and erythrocyte

damage.

3.1 – Anaemia

�n CKD, anaemia results, mainly, from an insufficient renal pro-

duction of EPO. However, other factors may also contribute to the de-

velopment of anaemia, by disturbing erythopoiesis or the mobilization

of erythropoietic nutrients.

For a better understanding of the mechanisms that may underlie

the development of the anaemia, a brief presentation about erythro-

poiesis will be presented.

3.1.1 - Normal erythropoiesis

The mean lifespan of a normal human erythrocyte is about 120

days. Erythrocytes are involved in transporting carbon dioxide and



nitric oxide, but their principal function is to deliver oxygen from the

lungs to the other tissues of the body. The amount of oxygen delivered

to the tissues depends on the amount of haemoglobin in the circulating

erythrocytes.

�n normal adults, about 200 billion of the oldest erythrocytes

(about 1% of the total number) are replaced every day by an equal

number of newly formed erythrocytes. �n situations in which the

erythrocytes are abnormally lost from the circulation by bleeding or

by increased destruction, the rate of new erythrocyte production can

exceed one trillion per day.

Erythropoiesis is the process by which the haematopoietic tissue

of the bone marrow produces erythrocytes (Fig. 4). �t is a continu-

ous process of proliferation, differentiation and maturation, beginning

Fig. 4 - Schematic diagram showing normal erythropoiesis (adapted from Dianzani,

1996). SCF - stem cell factor; G-CSF - granulocyte colony-stimulating factor, IL- interleu-

kin, GM-CSF – granulocyte-macrophage colony-stimulating factor, CFU - colony-form-

ing unit, EPO – erythropoietin.



with the haematopoietic stem cell and ending with the erythrocyte (�s-

cove, 1977; Gregory and Eaves, 1978). Haematopoietic stem cells are

rare, less than one in ten thousand nucleated cells of the bone marrow,

and they can self-renew or differentiate into all types of blood cells.

The earliest stage of progenitor cell differentiation that is committed

to the erythroid lineage is the burst-forming unit-erythroid (BFU-E).

Human BFU-Es are defined by their ability to form large “bursts” of

erythroblast colonies or one very large colony of erythroblasts, after

two to three weeks in semisolid tissue culture. Erythroid “bursts” can

contain more than one thousand erythroblasts and, thus, a single BFU-

E and its progeny can have ten or more rounds of cell division before

reaching the terminal postmitotic stages of differentiation. The next

defined stage is the colony-forming unit-erythroid (CFU-E). Human

CFU-Es require one week to form single colonies of up to 64 eryth-

roblasts in tissue culture. Thus, CFU-Es and their progeny have six

or fewer rounds of cell division (Gregory and Eaves, 1978; Emerson,

1985). The erythropoietic stages subsequent to CFU-E (proerythro-

blast, basophilic erythroblast, orthochromatic erythroblast and reticu-

locyte) are defined by their light microscopic appearance in stained

preparations. The percentage of cells in active cell cycle is greatest in

the CFU-E and proerythroblast stages, and cell division ceases at the

polychromatophilic stage.

To ensure a constant production of erythrocytes, several nutri-

ents are required, including iron for haemoglobin synthesis, folic acid

and vitamin B12 for DNA synthesis (Koury and Ponka, 2004). Sev-

eral growth factors and cytokines, namely, trombopoietin, interleukin

(�L)-3, �L-6, �L-8, �L-9, �L-11; granulocyte/monocyte colony-stimulat-

ing factor (GM-CSF) increase differentiation and proliferation ofear-

ly erythroid progenitors, stimulating the production of erythrocytes.



They act on erythroid progenitors to prevent apoptosis, to induce pro-

liferation and to promote or delay/inhibit differentiation (Birkmann,

1997). Delay in differentiation may play an important role by allowing

self-renewal and expansion of erythroid progenitors before terminal

maturation. Steroid and glucocorticoids hormones may also enhance

erythrocyte production. There is also evidence for a role of the renin-

angiotensin system in the regulation of erythropoiesis (Birkmann,

1997). None of these factors is critical for in vivo erythroid develop-

ment. However, genetic and biochemical approaches have demonstrat-

ed that stem cell factor (SCF) and EPO, which act on their respective

receptors C-kit and EPO receptor (EPOR) are absolutely required for

erythroid cell proliferation (Broudy, 1997).

There is no evidence that an excess of erythroid progenitors or

of mature erythrocytes regulate SCF production by stromal cells. How-

ever, EPO concentration and EPOR activation are the major param-

eters that determine erythrocyte production. EPO production is well

controlled by renal oxygenation. A sufficient number of erythrocyte

and adequate oxygenation results in decrease of EPO production by

the kidney, through degradation of hypoxia inducible factor (H�F) (Fig.

5). As a consequence, the less EPO sensitive erythroid progenitors and

precursors will die of apoptosis (Birkmann, 1997; Broudy, 1997).

3.1.2 – Erythropoietin (EPO)

The involvement of a humoral factor (named as haemopoietin)

in the regulation of haematopoiesis, was firstly described in literature

in 1906 (Carnot and Deflandre, 1906). However, only 40 years later

a linkage between EPO and erythropoiesis was described (Bondsdorff

and Jalavisto, 1948), and only in the 1950s was established that the



kidney is the main site of production of EPO (Jacobson, 1957). �n 1977,

EPO was purified from urine collected from patients suffering from

aplastic anaemia (Miyake, 1977). The nucleotide sequence of human

EPO gene was determined in 1985 and the cloning and expression of

the gene led to the production of recombinant human EPO (rhEPO)

Fig. 5 - Schematic representation of the PHD-VHL-HIF axis. The hypoxia-inducible

factor (HIF) - α subunit is synthesized continuously but is rapidly destroyed in the

presence of oxygen and iron. Oxygen- and iron-dependent prolyl hydroxylase domain

(PHD) enzymes hydroxylate specific praline residues in HIF-α, increasing its affinity

for the von Hippel-Lindau tumour suppressor protein (VHL). The binding of VHL to

hydroxylated HIF-α then targets HIF- α for destruction by a multiprotein ubiquitin

ligase (denoted “ligase”) that mediates proteosomal degradation of HIF- α subunits.

Another oxygen- and iron- dependent enzyme, factor inhibiting HIF (FIH), hydrox-

ylates an asparagines residue in HIF-α, reducing its ability to activate transcription

by inhibiting binding of the transcriptional coactivator complex p300/CBT. Under

hypoxic conditions, the hydroxylation of HIF-α by PHDs and FIH is inhibited and

proteosomal degradation is thus slowed. HIF-α rapidly accumulates and dimerises

with HIF-β, which is expressed constitutively and is present in excess. The p300/CBP

coactivator is recruited, and the DNA binding complex subsequently up-regulates

hypoxia-responsive genes. OH denotes hydroxyl group. (Smith, 2008).



(Lin, 1985; Jacobs, 1985). The first clinical trial using rhEPO in the

treatment of the anaemia of end-stage renal failure was published in

1987 (Eschbach, 1987). Nowadays, rhEPO is currently used for the

treatment of anaemia of CKD, and also in other clinical situations, as

shown in table �� (Kimel, 2008; Obladen, 2000). This therapy allowed

a significant reduction in the anaemia and in the associated adverse

effects, improving patient’s quality of life.

Table III - Most important clinical applications of rhEPO*

Replacement therapy (low endogenous EPO)

- CKD

- Anaemia of prematurity

- Non-myeloid malignancy

- H�V infection

Supportive therapy

- Post chemotherapy

- Post transplantation

- Sport (abuse by athletes)

- Alternative to blood transfusion

Anaemia associated with

- Haemoglobinopathy

- Autoimmune disease

- Acute renal failure

Possible future applications

- Neuroprotetive agent

- Diabetic nephropathy

- Anaemia of congestive cardiac failure

*Adapted from Kimel, 2008.



EPO is an endogenous cytokine that is essential in erythropoiesis

regulation. This glycoprotein has a molecular weight of 30-35 kDa, 165

amino acids and is heavily glycosylated, with the carbohydrate moiety

comprising approximately 40% of its weight. There are three N-ter-

minal glycosylation sites at aspartate residues 24, 38 and 83, and one

O-linked acidic oligonucleotide side-chain at serine 126. Human EPO

has two disulphide bridges, between cysteines 7 and 161, and between

cysteines 29 and 33, which are important in maintaining in vivo its bio-

activity and the correct shape for binding to the EPOR (Lai, 1986).

The regulation of EPO gene expression occurs essentially at the

transcriptional level by DNA-dependent mRNA synthesis and gene ac-

tivation. �n kidneys, hypoxia gives rise to increased EPO expression,

stimulated by the DNA binding protein, H�F, which binds to the 3´

flanking region of EPO gene (Wang and Semenza, 1993). EPO is secret-

ed into the plasma and, within the bone marrow, binds to EPOR in the

surface of erythroid progenitor and precursor cells. EPOR activation

follows a sequential dimerization activation mechanism involving the

Janus kinase 2 (JAK2), and phosphorylation and nuclear translocation

of signal transducer and activator of transcription 5 (STAT5) pathways

(Fig. 5).

3.1.3 - Anaemia and therapeutic use of recombinant human EPO

(rhEPO) in CKD

Anaemia, as referred, is a frequent complication associated with

CKD, and is mainly due to insufficient erythropoietin renal produc-

tion. Accumulation of uremic toxins, excessive toxic storage of alumin-

ium in the bone marrow (Miyoshi, 2006), blood loss (either iatrogenic,

from the puncture sites of the vascular access and blood sampling, or



from other sources, such as the gastrointestinal tract), and premature

erythrocyte destruction have been also frequently associated with

anaemia in CKD patients (Medina, 1994; Pisoni, 2004).

�n CKD, the anaemia develops gradually, with the progressive de-

cline of renal function (Bárány 2001; Locatelli, 2004a; Smrzova, 2005)

and EPO synthesis (Bárány 2001; Smrzova, 2005). Anaemia can lead

to adverse clinical effects, namely, reduction in tissue oxygenation, in-

crease in cardiac output, left ventricular hypertrophy, congestive heart

disease, fatigue, reduction in exercise capacity, and immunodeficiency

(Foley, 1996; Locatelli, 1998).

Until 18 years ago, the treatment of anaemia of CKD was blood

transfusion; since them, the management of this anaemia has been

improved by the introduction of rhEPO therapy. This therapy allowed

a significant reduction in the associated adverse effects of anaemia and

improved patient’s quality of life.

Until recently, only two forms of rhEPO were commercially

available for clinical use, epoetin alfa (Eprex®) and epoetin beta (Ne-

oRecormon®), which are structurally very similar. Subcutaneous or

intravenous administration of these epoietins can be used once daily,

two or three times a week, depending on the clinical status of the pa-

tient. The rhEPO dose is adjusted to achieve target haemoglobin levels

between 11 and 12 g/dL (Locatelli, 2004b).

Several different types of erythropoietic agents have been devel-

oped recently, namely darbopoietin alfa (Aranesp®), which contains

five N-linked glycosylation sites (only three for EPO), resulting in a

longer plasma half-life. Treatment of anaemia of CKD with darbopoi-

etin alfa allows a less frequent administration, as compared to the oth-

ers epoetins (Osterborg, 2004).

�n spite of the strike benefits to CKD patients, we must refer



that the use of rhEPO has been associated with some complications,

namely, hypertension, thrombosis, allergic reactions and pure red cell

aplasia (Locatelli, 2001; Casadevall, 2002).

3.1.4 - Potential therapeutic use of rhEPO

Currently, the main indication for rhEPO is the treatment of

anaemia. However, an increasingly growing body of evidence indicates

that therapeutic benefits of rhEPO could be far beyond correction of

anaemia, namely in cardiovascular and nervous systems (Bogoyevitch,

2004; Bahlmann, 2004; Brines & Cerami, 2005). Some of the principal

studies regarding the protective effect of rhEPO in rat models are pre-

sented in table �V.

Table IV – Studies in rat model on acute protective effects of rhEPO.

Reference Experimental model Mechanism of

tissue protection

Abdelrahman, 2004 Haemorrhagic shock Reduced caspase activity

(reduced apoptosis)

Bagnis, 2001 Cisplatin nephrotoxicity �ncreased regeneration of renal

tubular cells

Yang, 2003 �schemia/reperfusion Heat shock protein 70 activa-

tion, reduced caspase activity

Vesey, 2004 �schemia/reperfusion Reduced apoptosis, increased

regeneration of renal tubular cells

Johnson, 2006 �schemia/reperfusion Reduced apoptosis, increased

regeneration of renal tubular cells



Considering the wide distribution of EPOR expression in body

cells and tissues (Dame, 2000), it would be surprising if no effect were

observed outside the erythroid cells. The central nervous system is

known to be a site of both EPO and EPOR expression, in fetus and

adults (Marti, 1996; Morishita, 1997; Dame, 2000). �t is not clear

whether EPO passage across the blood/brain barrier is a process facili-

tated by a specific mechanism, but neural cells are stimulated to grow

in vitro in response to EPO (Bueni, 2000; Brines, 2000; Cerami, 2001).

This data have lead to speculate that rhEPO may offer a therapeutic

advantage in various neurological injuries – speculation born out by

preclinical experience (Cerami, 2001). EPO and EPOR are expressed in

peripheral nerves before injury and show a differential regulation after

crush injury, suggesting that EPO may have peripheral effects in the

nervous system (Campana, 2001).

The effect of EPO on vascular function, angiogenesis, fatigue,

preadaptation to altitude, haemoglobin oxygen dissociation, and car-

diac performance has also been documented. The effects of EPO on an-

giogenesis both in culture and in intact endometrium have suggested

a role in neovascularization (Carlini, 1995; Kapiteijn, 2001). Whether

these findings have implications in the vascularization of neoplastic

tissues is unknown. Direct effects upon the vascular tissues have been

shown in vitro, as well as changes in blood pressure after longer term

rhEPO administration. Furthermore, effects of EPO in fatigue (sepa-

rate from the changes mediated by haemoglobin increase) have been

suggested but have been difficult to substantiate (Carlini, 1995; Vaziri,

2001).

EPO receptors are also present in kidney tissue, especially on tu-

bular cells (Westenfelder, 1999). �t has been described that EPO ac-



tivates different intracellular pathways in kidney tissue, such as the

P�3K/Akt pathway and the heat shock protein 70, suggesting that the

administration of rhEPO has a protective effect in acute renal failure,

such as ischemia injury (Bagnis, 2001; Yang, 2003; Abdelrahman,

2004; Vesey, 2004; Johnson, 2006).

3.2 - Inflammation and leukocyte activation

�nflammation is the physiological response to a variety of nox-

ious stimuli, such as tissue injury caused by infection or physical dam-

age. �t is a complex process that involves the participation of several

cells and molecules, and may present different intensities and dura-

tion.

�nflammation usually refers to a localised process. However, if

the noxious stimulus is severe enough, distant systemic changes may

also occur, and these changes are referred as “acute phase response”,

which is accompanied by signs and symptoms such as fever, anorexia,

and somnolence. This acute phase response may include neuroendo-

crine, metabolic and haematopoietic changes, as well as changes in

non-protein plasma constituents (Ceciliani, 2002). The haematopoietic

response includes leukocytosis and leukocyte activation, thrombocy-

tosis, and anaemia secondary to erythrocyte damage and/or decreased

erythropoiesis (Trey and Kushner, 1995).

�nflammatory stimuli induces the release of cytokines, includ-

ing tumour necrosis factor (TNF)-α, �L-1, �L-6, and interferon (�FN)-γ,

which may be produced by several cells, including leukocytes, fibro-

blasts and endothelial cells (Kushner and Rzewnicki, 1999). This re-

lease of cytokines causes many systemic changes, including increased

synthesis and release of positive acute-phase proteins, such as C-reac-



tive protein (CRP) and fibrinogen, as well as the suppression of nega-

tive acute-phase proteins, such as albumin and transferrin (Macdou-

gal, 1995; Cooper, 2003; Smrzova, 2005).

The causes for the inflammatory response in haemodialysis pa-

tients are not well clarified. There are several potential sources, includ-

ing bacterial contamination of the dialyser, incompatibility with the

dialyser membrane and infection of the vascular access. However, the

dialysis procedure may only be partially responsible for the inflamma-

tory response, because even patients with renal insufficiency who are

not yet on dialysis present raised inflammatory markers, which rise

further after starting regular haemodialysis treatment, suggesting that

the disease per se triggers an inflammatory response (Gunnell, 1999;

Macdougall and Cooper, 2002; Schindler, 2002).

Along an inflammatory response, the iron from the erythropoi-

esis traffic is mobilised to storage sites within the reticuloendothelial

system, inhibiting erythroid progenitor proliferation and differentia-

tion, and blunting, therefore, the response to EPO (endogenous and/

or exogenous). An erythropoiesis-suppressing effect has been also at-

tributed to increased activity of pro-inflammatory cytokines reported

in this inflammatory condition, and this relationship has been pro-

posed as a potential factor associated to rhEPO therapy resistance (Des-

camps-Latscha, 1995; Gunnell, 1999; Macdougall and Cooper, 2002;

Schindler, 2002; Cooper, 2003). Actually, it was reported that pro-in-

flammatory cytokines, such as �L-1, �L-2, �L-4, �L-6, TNF-α and �NF-γ

diminish BFU-E and CFU-E cells, resulting in suppression of eryth-

ropoiesis (Means and Krantz, 1996; Allen, 1999; Macdougall, 2002).

Moreover, it was reported (Macdougall, 2002) that serum derived

from haemodialysed patients suppresses erythroid colony-forming re-



sponse to EPO, in a manner that can be inhibited by antibodies against

TNF-α and �NF-γ. These data suggest a key role in the rhEPO response

for these inflammatory mediators (Waltzer, 1984; Foley, 1996; Meyer,

2002; Cooper, 2003).

Leukocytosis and recruitment of circulating leukocytes into the

affected areas are hallmarks of inflammation. Leukocytes are chimio-

attracted to inflammatory regions and their transmigration from blood

to the injured tissue is primarily mediated by the expression of cell-

adhesion molecules in the endothelium, which interact with surface

receptors on leukocytes (Muller, 1999; Sullivan, 2000). This leukocyte-

endothelial interaction is regulated by a cascade of molecular steps that

correspond to the morphological changes that accompany adhesion.

This adhesion cascade has been divided into sequential steps based on

visual assessment of the post-capillary venules during the early stages

of acute inflammation. �n the absence of inflammation, leukocytes are

“rolling” along the vessel wall. After the inflammatory stimulus is ap-

plied, leukocytes roll along the post-capillary venules (but not arte-

rioles or small arteries) at velocities distinctly below that of flowing

blood. Some rolling cells can be seen to arrest and, after a few minu-

tes, change, their shape, in an apparent response to local chemotactic

stimuli. Extravasation into the extravascular tissue follows. Each of

these steps requires either upregulation or activation of distinct sets of

adhesion molecules. At the inflammatory site, leukocytes release their

granulation products and may exert their phagocytic capacities.

�n acute inflammation, the leukocyte infiltration is predomi-

nantly of neutrophils, whereas in chronic inflammation a mononucle-

ar cell infiltration (predominantly macrophages and lymphocytes) is

observed. Although leukocyte-endothelial cell interaction is important



for leukocyte extravasation and trafficking in physiological situations,

there is increasing evidence that altered leukocyte-endothelial inter-

actions are implicated in the pathogenesis of diseases associated with

inflammation, possibly by damaging the endothelium or altering endo-

thelial function (Harlan, 1985; Ley, 2007).

Leukocytosis is essential as the primary host defence, and neu-

trophils, the major leukocyte population of blood in adults, play a

primordial role. �t is well known that neutrophils have mechanisms

that are used to destroy invading microorganisms. These cells use an

extraordinary array of oxygen-dependent and oxygen-independent

microbicidal weapons to destroy and remove infectious agents (Witko-

Sarsat, 2000). Oxygen-dependent mechanisms involve the production

of reactive oxygen species (ROS), which can be microbicidal (Roos,

2003), and lead to the development of oxidative stress. Oxygen-inde-

pendent mechanisms include chemotaxis, phagocytosis and degranu-

lation. The generation of microbicidal oxidants by neutrophils results

from the activation of a multiprotein enzyme complex known as the

reduced nicotinamide adenine dinucleotide phosphate oxidase, which

catalyzes the formation of superoxide anion (O2

-).

Activated neutrophils also undergo degranulation, with the re-

lease of several components, namely, proteases (such as elastase) and

cationic proteins (such as lactoferrin).

Elastase is a member of the chymotrypsin superfamily of ser-

ine proteinases, expressed in monocytes and mast cells, but mainly

expressed in neutrophils, where it is compartmentalized in the prima-

ry azurophil granules. The intracellular function of this enzyme is the

degradation of foreign microorganisms that are phagocytosed by the

neutrophil (Brinkmann, 2004). Elastase can also degrade local extra-



cellular matrix proteins (such as elastin), remodel damaged tissue, and

facilitate neutrophil migration into or through tissues. Moreover, elas-

tase also modulates cytokine expression at epithelial and endothelial

surfaces, up-regulating the production of cytokines, such as �L-6, �L-8,

TGF-β (transforming growth factor β) and GM-CSF; it also promotes

the degradation of cytokines, such as �L-1, TNF-α and �L-2. There is

evidence in literature that high levels of elastase are one of the major

pathological factors in the development of several chronic inflamma-

tory lung conditions (Fitch, 2006).

Plasma lactoferrin is predominantly neutrophil derived and its

presence in the specific granules is often used to identify these types

of granules. Lactoferrin is also found in other granules, in the tertia-

ry granules, though in lower concentrations (Olofsson, 1977; Baynes

1986; Halliwell and Gutteridge, 1990; Saito, 1993). Lactoferrin is a

multifunctional iron glycoprotein, which is known to exert a broad-

spectrum primary defence activity against bacteria, fungi, protozoa and

viruses. �t can bind to large amounts of free iron. The iron bound to

lactoferrin is taken up by activated macrophages, which express specific

lactoferrin receptors. During inflammation, this causes iron deprivation

of the erythroid precursors, which fail to express lactoferrin receptors

(Bárány, 2001). Other mechanisms in which lactoferrin is implicated

include a growth regulatory function in normal cells, coagulation, and

perhaps cellular adhesion modulation (Levay and Viljoen, 1995).

End-renal failure induces also a clinical state of immunodefi-

ciency that probably underlies the incidence of infections and the high

mortality due to infectious complications. This immunodeficiency is

characterized by a deficient response to some vaccinations, namely for

hepatitis B virus. The mechanisms responsible for this immune defect

are still unknown. However, it is known that lymphopenia is associat-



ed to end-renal failure, occurring in the B and T lymphocyte compart-

ment, that could be related to an increased turnover of lymphocytes,

to a disturbance in lymphocyte homeostasis due to uraemia, and/or

to increased peripheral lymphocyte apoptosis associated with an ac-

tivation stimulus (Bender, 1984; Waltzer, 1984; Sester, 2000; Libetta,

2001; Meyer, 2002; Litjens, 2006).

3.3 - Iron metabolism

�ron is an essential trace element that is required for growth and

development of living organisms, but excess of free iron is toxic for

the cell (Arth, 1999; Atanasio, 2006; Deicher and Horl, 2006). Mam-

mals lack a regulatory pathway for iron excretion, and iron balance

is maintained by the tight regulation of iron absorption from the in-

testine (Park, 2001; Atanasio, 2006). The intestinal iron absorption is

regulated by the level of body iron stores and by the amount of iron

needed for erythropoiesis (Arth, 1999; Park, 2001; Nemeth, 2003; Ata-

nasio, 2006).

�n haemodialysis patients, iron absorption is similar to that found

in healthy individuals; however, when under rhEPO therapy, the ab-

sorption of iron increases as much as 5 times (Skikne and Cook; 1992).

This increased iron absorption is not sufficient to compensate for the

iron lost during the haemodialysis procedure, and with the frequent

blood draws performed on these patients. For this reason, intravenous

iron administration has become a standard therapy for most patients

receiving rhEPO. To avoid iron overload, with potentially harmful con-

sequences, there is a need to monitor iron therapy by performing regu-

lar blood tests reflecting body’s iron stores.

Recently, a complex regulatory network that governs iron traf-

fic emerged, and points to hepcidin as a major evolutionary conserved



regulator of iron distribution (Nicolas, 2002; Kemma, 2005; Nemeth,

2006). This small hormone, produced by the mammalian liver, has

been proposed as a central mediator of dietary iron absorption. High

levels of hepcidin were found to be associated with a decrease in both

iron uptake from the small intestine and in iron release from macro-

phages; a decreased placental iron transport was also observed (Fig. 6)

(Kulaksiz, 2004).

The synthesis of hepcidin is stimulated by inflammation and

iron overload. �t is synthesized as preprohepcidin, a protein with 84

amino acids. This peptide is cleaved leading to prohepcidin with 60

aminoacids, which is further processed, giving rise to the 25 amino-

acids protein, hepcidin (Dallalio, 2003; Hsu, 2006).

Fig. 6 - Key sites of iron metabolism. There are 4 sites with special functions in iron han-

dling: enterocytes (A), erythroid precursors (B), macrophages (C), and hepatocytes (D)

(Swinkels, 2006).



Hepcidin was reported to bind to the transmembrane iron export

ferroportin, which is present on macrophages, on the basolateral site

of enterocytes, and also on hepatocytes. �t has been demonstrated in

vitro that hepcidin induces the internalization and degradation of ferro-

portin, a crucial protein for cellular iron export (Domenico, 2007). By

diminishing the effective number of iron exporters on the membrane

of enterocytes and of macrophages, hepcidin suppresses iron uptake

and release, respectively. This is the phenotype of ferroportin disease,

where iron accumulation is observed mainly in macrophages and often

combined with anaemia (Njajou, 2002).

The in vitro stimulation of fresh human hepatocytes by pro-in-

flammatory �L-6 showed strong induction of hepcidin mRNA, indi-

cating that this cytokine may be an important mediator of hepcidin

induction in inflammation (Fleming and Sly, 2001; Nemeth, 2004;

Wrighting & Andrews, 2006). Upon an inflammatory stimulus, �L-6 is

released and binds to a complex of �L-6 receptor α and gp130. The �L-6

ligand-receptor interaction results in the activation of JAKs pathways

that phosphorylate STAT proteins, predominantly STAT3. Upon phos-

phorylation at tyrosine residue 705, STAT3 translocates into the nucle-

olus where it regulates transcription of many target genes (Carbia-Na-

gashima and Arzt, 2004), including the hepcidin related gene (Fig. 7).

This increased hepcidin expression during inflammation ex-

plains sequestration of iron in the macrophages and inhibition of intes-

tinal iron absorption, the two hallmarks of the anaemia of inflamma-

tion, which is normocytic or microcytic iron-refractory (Nicolas, 2002;

Dallalio, 2003; Kulaksiz, 2004; Hsu, 2006). This decreased availability

in iron may be a host defence mechanism against invading microorga-

nisms.



The biological importance of the precursor molecule of hepci-

din, named prohepcidin, in regulating iron metabolism, is still unde-

termined. However, there is some evidence that prohepcidin levels are

a reliable indicator of hepcidin levels and activity (Dallalio, 2003; Ku-

laksiz, 2004; Taes, 2004; Hsu, 2006).

Hepcidin synthesis is regulated by inflammation, a common

finding in haemodialysis patients. Actually, increased levels of prohep-

cidin have been reported in CKD patients under haemodialysis (Nico-

las, 2002; Dallalio, 2003; Kulaksiz, 2004; Locatelli, 2004; Taes, 2004;

Hsu, 2006; Malyszko, 2006).

�ron uptake from plasma to erythroid cells is regulated by dif-

ferent proteins, namely, transferrin receptor, haemochromatosis (HFE)

protein and DMT1 (NRAMP2/DCT1). �t is found in literature, that al-

terations in transferrin receptor gene are associated to HFE (Wallace

and Subramaniam, 2007), and HFE gene mutations are associated to a

reduction in the amount of rhEPO necessary to support erythropoiesis

in haemodialysis patients (Valenti, 2008). On the other hand, muta-

Fig. 7– IL-6 binds to a complex of the IL-6 receptor α and gp130, which results in the

activation of JAK and STAT3 proteins, resulting in hepcidin gene expression.



tions in DMT1 gene, in mice, are associated to an inhibition of intes-

tinal iron absorption and to a decrease in erythroid cell precursor’s

iron uptake, resulting in hypocromic and microcytic anaemia (Flem-

ing, 1997).

3.4 - Oxidative stress

Oxidative stress and its related biological effects have a patholog-

ic relevance in many disease states, namely, in those having an associ-

ated inflammatory component (Sommerburg, 1998; Spittle, 2001; Pu-

pim, 2004). Oxidative stress has been defined as an umbalance between

oxidants and antioxidants, resulting in an overall increase in cellular

levels of ROS. �t can be produced by both endogenous and exogenous

sources. Potential endogenous sources of ROS include the oxidative

phosphorylation, P450 metabolism, and inflammatory cell activation.

During mitochondrial oxidative metabolism, the majority of the

oxygen consumed is reduced to water; however, an estimated 4% to

5% of molecular oxygen is converted to ROS, primarily superoxide an-

ion, formed by an initial one-electron reduction of molecular oxygen.

Superoxide can be dismutated by superoxide dismutase (SOD) to yield

hydrogen peroxide, less reactive and with a lower oxidant power. �n

the presence of partially reduced metal ions, in particular iron, hydro-

gen peroxide is subsequently converted through Fenton and Haber-

Weiss reactions to a hydroxyl radical. The hydroxyl radical is highly re-

active and can interact with nucleic acids, lipids, and proteins (Barber,

1994; Betteridge, 2000).

Neutrophils and macrophages are an additional endogenous

source and are major contributors to the cellular ROS. Activated neu-

trophils and macrophages, through the respiratory burst, elicit a rapid



but transient increase in oxygen uptake that gives rise to a variety of

ROS, including superoxide anion, hydrogen peroxide, and nitric oxide

(Vuillaume, 1987; Witz, 1991). Under normal physiological conditions,

cells are capable of counterbalance the production of ROS with an-

tioxidants. Endogenous cellular antioxidant defences are mainly en-

zymatic and include SOD, glutathione peroxidase (GPx), and catalase

(�manishi, 1986). Nonenzymatic antioxidants such as vitamin E, vita-

min C, β-carotene, glutathione, and coenzyme Q function to reduce

ROS (Vuillaume, 1987; Witz, 1991). When the redox balance is shifted

in favour of cellular oxidants, oxidative damage to nucleic acids, lipids,

or proteins can result and produce modifications in cell function and

cell viability (�manishi, 1986; Vuillaume, 1987; Witz, 1991).

There is increasing evidence that ROS are involved in the pro-

gression of renal damage and uraemic symptoms. The oxidative stress

appears to be multifactorial and is thought to be caused only in part by

the impaired kidney function (Sommerburg, 1998; Spittle, 2001; Pu-

pim, 2004). The several factors that have been implicated in the devel-

opment of oxidative stress in CKD patients undergoing haemodialysis

include the haemodialysis procedure, uraemia, inflammation, distur-

bances in iron homeostasis, and reduction in the antioxidant defence

systems (Lucchi, 2000; Fryer, 2000; Sezer, 2007).

3.5 - Erythrocyte damage

The erythrocyte, presenting a limited biosynthesis capacity, suf-

fers and accumulates physical and/or chemical changes, which become

more pronounced with cell aging, and whenever an unusual physical

or chemical stress develops (Locatelli, 2004a).

The erythrocyte membrane is a complex structure comprising a



lipidic bilayer, integral proteins and the skeleton. Spectrin is the major

protein of the cytoskeleton, and, therefore, the major responsible for

erythrocyte shape, integrity and deformability. �t links the cytoskeleton

to the lipid bilayer, by vertical protein interactions with the transmem-

brane proteins, band 3 and glicophorin C (Lucchi, 2000). �n the vertical

protein interaction of spectrin with band 3 are also involved ankyrin

(also known as band 2.1) and protein 4.2. A normal linkage of spectrin

with the other proteins of the cytoskeleton assures normal horizontal

protein interactions (Fig. 8).

Fig 8 - A schematic representation of the red blood cell membrane. The membrane is a

complex structure in which a plasma membrane envelope composed of amphiphilic lipid

molecules is anchored to a two dimensional elastic network of skeletal proteins through

tethering sites (transmembrane proteins) embedded in the lipid bilayer. RhAG: Rh blood

group antigen related; GPA: glicophorin A; GPC: glicophorin C. (An and Mohandas,

2008).



Erythrocytes are physically stressed during the haemodialy-

sis process, and metabolically stressed by the unfavourable plasmatic

environment, due to metabolite accumulation, and by the high rate

of haemoglobin autoxidation, due to the increase in haemoglobin

turnover, a physiologic compensation mechanism triggered in case of

anaemia (Lucchi, 2000; Stoya, 2002). The erythrocytes are, therefore,

continuously challenged to sustain haemoglobin in its reduced func-

tional form, as well as to maintain the integrity and deformability of

the membrane.

When haemoglobin is denatured, it links to the cytoplasmic

domain of band 3, triggering its aggregation and leading to the forma-

tion of strictly lipidic portions of the membrane, poorly linked to the

cytoskeleton. These cells are probably more prone to undergo vesicula-

tion (loss of poorly linked membrane portions) whenever they have to

circulate through the haemodialysis membranes or the microvascula-

ture. Vesiculation may, therefore, lead to modifications in the erythro-

cyte membrane of CKD patients under haemodialysis (Reliene, 2002;

Gallagher, 2005; Rocha, 2005).

Erythrocytes that develop intracellular defects earlier during

their life span are removed prematurely from circulation (Santos-Sil-

va, 1998; Rocha-Pereira, 2004). The removal of senescent or damaged

erythrocytes seems to involve the development of a senescent neoanti-

gen on the plasma membrane surface, marking the cell for death. This

neoantigen is immunologically related to band 3 (Kay, 1994). The de-

terioration of the erythrocyte metabolism and/or of its antioxidant de-

fences may lead to the development of oxidative stress within the cell,

allowing oxidation and linkage of haemoglobin to the cytoplasmatic

domain of band 3, promoting its aggregation, the binding of natural



antiband 3 autoantibodies and complement activation, marking the

erythrocyte for death.

An abnormal band 3 profile [high molecular weight aggregates

(HMWAg), band 3 monomer and proteolytic fragments (Pfrag)], has

been associated to younger, damaged and/or senescent erythrocytes.

Older and damaged erythrocytes present with higher HMWAg and

lower Pfrag. Younger erythrocytes show reduced HMWAg and higher

Pfrag (Santos-Silva, 1998). Moreover, several diseases, known as in-

flammatory conditions, presented abnormal band 3 profiles, sugges-

tive of oxidative stress development (Santos-Silva, 1998; Belo, 2002;

Rocha-Pereira, 2004).

4 - Resistance to rhEPO therapy

Anaemia is a common complication that contributes to the bur-

den of haemodialysis patients. �t has also a negative impact on cardio-

vascular system, cognitive function, exercise capacity and quality of

life, resulting in a significant morbidity and mortality in these patients.

The introduction of rhEPO therapy in the early 1990s for treatment of

anaemia of CKD has led to a significant reduction in anaemia and to

an improvement in patients’ quality of life (Locatelli, 1998; Bárámy,

2001; Locatelli, 2004a; Smrzova, 2005). There is, however, a marked

variability in the sensitivity to rhEPO, with up to 10-fold variability in

dose requirements to achieve correction of the anaemia. Furthermore,

around 5-10% of the patients show a marked resistance to rhEPO

therapy (Bárámy, 2001; Macdougall, 2002; Schindler, 2002; Smrzo-

va, 2005). The European Best Practice Guidelines define resistance to

rhEPO therapy as a failure to achieve target haemoglobin levels (be-



tween 11 and 12 g/dL) with maintained doses of rhEPO higher than

300 �U/Kg/week of epoetin or of higher doses than 1.5 μg/Kg/week of

darbopoietin-alfa (Locatelli, 2004b).

The reasons for this variability in rhEPO response are unclear

(Foley, 1996; Spittle, 2001; Drueke and Eckardt, 2002; Cooper, 2003;

Himmelfarb, 2004; Smrzova, 2005). There are a lot of conditions re-

ported as associated with rhEPO resistance, namely, inflammation, oxi-

dative stress and iron deficiency, as major causes (Foley, 1996; Gunnell,

1999; Spittle, 2001; Drueke and Eckardt, 2002; Cooper, 2003; Himmel-

farb, 2004; Pupim, 2004; Smrzova, 2005), and blood loss, hyperpara-

thyroidism, aluminium toxicity and vitamin B12 or folate deficiency,

as minor causes. However, exclusion of these factors does not eliminate

the marked variability in sensitivity to rhEPO (Macdougall and Cooper,

2002). Furthermore, most of the published studies investigated only,

a few number of parameters and included a very limited number of

patients. Therapeutic and haemodialysis procedure have changed in

the last few years. The value of those changes concerning the haemo-

dialysis membrane biocompatibility and vascular access, need also to

be clarified in what concernes to resistance to rhEPO therapy.



Part II - Aims



Considering the knowledge presented previously and given

the implications of rhEPO resistance in morbidity and mortality in

haemodialysis patients, the major aim of this work was to contribute

to a better understanding of the resistance to rhEPO therapy observed

in some haemodialysis patients.

To better clarify the mechanisms of resistance to rhEPO therapy,

we studied systemic changes associated with resistance to rhEPO ther-

apy in haemodialysis patients (responders and non-responders), with

particular interest on inflammation, leukocyte activation, iron status,

oxidative stress and erythrocyte damage.

- �nflammatory stimuli induces the release of cytokines, which in

turn cause many systemic changes, including increased synthesis and

release of positive and negative acute-phase proteins. Some of these

pro-inflammatory cytokines, such as interleukin �L-1, �L-2, �L-4, �L-6,

TNF-α and �NF-γ, are described to have an erythropoiesis-suppressing

effect, the reason why we hypothesized that an enhanced inflammato-

ry stimulus could explain the variability in the therapeutic response to

rhEPO therapy. This hypothesis was studied as presented in paper I:

“Inflammation, T-Cell Phenotype, and Inflammatory Cytokines in

Chronic Kidney Disease Patients Under Hemodialysis and its Relationship to

Resistance to Recombinant Human Erythropoietin Therapy”. J Clin Immunol

2008; 28:268-275.



- The pro-inflammatory cytokines are produced by several type

of cells, namely, by T-cells. As enhanced T-cell cytokine production

is associated to T-cell activation, we hypothesized that an increase in

T-cell activation and cytokine production could have a role in rhEPO

resistance. This hypothesis was also assessed in paper I.

- Additionally, end-renal failure induces a clinical state of im-

munodeficiency. The mechanism responsible for this immune defect is

still unknown. However, it is known that lymphopenia is associated to

end-renal failure, occurring in the B and T lymphocyte compartment,

probably due to increased apoptosis. As �L-7 has recently emerged as

a key cytokine involved in controlling the homeostatic turnover and

survival of peripheral resting memory CD4+ T cells, we hypothesized

that the serum levels of this interleukin could be related to the de-

creased number in total lymphocytes and in CD4+ T-cells, that we had

observed in haemodialysis patients (as presented in paper �). These

cells were particularly decreased in non-responders to rhEPO therapy.

This hypothesis was assessed in paper II:

“IL-7 serum levels and lymphopenia in hemodialysis patients, non-re-

sponders to recombinant human erythropoietin therapy”. Blood Cells Mol Dis

2008; 41:134-135.

- Activation of polymorphonuclear leukocytes leads to degranu-

lation, with release of several components, namely elastase and lac-

toferrin. The major role of lactoferrin in iron metabolism is to control

iron availability. Elastase modulates cytokine expression at epithelial

and endothelial surfaces, up-regulating cytokines, such as the pro-in-

flammatory �L-6, suggesting that elastase is one of the major pathologi-



cal factors in the development of several chronic inflammatory diseas-

es. Moreover, elastase, as a serinoprotease may modulate pathogenic

changes in plasmatic and cellular components. We hypothesized that

neutrophil activation as part of an inflammatory condition, could play

a role in iron availability for erythropoiesis, and in the enhancement

of inflammation. We have also hypothesized that neutrophil activation

may also result from their interaction with the nonbiological materi-

als of the extracorporeal circuit during the haemodialysis procedure.

These hypotheses were assessed in paper III:

“Neutrophil activation and resistance to recombinant human erythro-

poietin therapy in hemodialysis patients”. Am J Nephrol 2008; 28: 935-940.

- �t is known that the inflammatory stimulus has an important

impact in iron metabolism, by mobilizing iron from erythropoiesis traf-

fic to storage sites within the reticuloendothelial system, inhibiting

erythroid progenitor proliferation and differentiation. This change in

iron mobilization may lead to an iron depleted erythropoiesis. Further-

more, recently, a complex regulatory network that governs iron traffic

emerged, and points to hepcidin as a major evolutionary conserved

regulator of iron distribution. The synthesis of hepcidin is stimulated by

inflammation, and has been reported to be increased in haemodialysis

patients. Given the importance of iron mobilization for erythropoiesis,

we hypothesized that alterations in iron traffic (deficiency is avoided

by iron supplementation in haemodialysis patients) could be related to

resistance to rhEPO therapy. Several important factors in iron traffic

were studied and were reported in papers IV and V:

“Role of Prohepcidin, Inflammatory Markers and Iron Status in Resis-

tance to rhEPO Therapy in Hemodialysis Patients”. Am J Nephrol 2008;28:677-

683.



“Effect of hemodialysis procedure in prohepcidin serum levels in chronic

kidney disease patients”. Clin Nephrol, in press.

- �ron uptake from plasma to erythroid precursor cells is regu-

lated by different proteins, namely, transferrin receptor, HFE protein

and DMT1. �n literature, alterations in transferrin receptor gene have

been associated to HFE, and HFE gene mutations are associated to a

reduction of the amount of rhEPO necessary to support erythropoiesis

in haemodialysis patients. However, these two proteins are unlikely to

be the main causes of the resistance to rhEPO therapy in CKD patients

under haemodialysis, as HFE is characterized by increases in serum

iron parameters, namely transferrin saturation, not found in non-re-

sponder CKD patients. On the other hand, the DMT1 gene could be a

good candidate to underlie the variability of response to rhEPO in CKD

patients. �ndeed, the few described mutation cases in the DMT1 gene

are associated to an inhibition of intestinal iron absorption and to a

decrease in erythroid cell precursor’s iron uptake, resulting in hypocro-

mic and microcytic anaemia. This anaemia is similar to that observed

in non-responder CKD patients, resulting of a functional iron depleted

erythropoiesis in the presence of adequate serum iron levels. We tested

the hypotheses that alterations in DMT1 can be related with resistance

to rhEPO therapy, as presented in paper VI:

“DMT1 (NRAMP2/DCT1) genetic variability and resistance to rhEPO

therapy in chronic kidney disease patients under haemodialysis“. Acta Haema-

tologica 2008;120: 11-13.

- Besides the reduction in erythropoietin production (as the main

cause), premature erythrocyte removal is frequently associated to the

anaemia in haemodialysis patients. This premature damage could be



related to alterations in the erythrocyte membrane due to the plasmat-

ic changes occurring in these patients and/or to changes resulting from

the haemodialysis procedure per se. These hypotheses were studied and

reported in papers VII and VIII:

“Altered Erythrocyte Membrane Protein Composition in Chronic Kidney

Disease Stage 5 Patients under Haemodialysis and Recombinant Human Eryth-

ropoietin Therapy”. Blood Purification 2008; 26: 267-273.

“Changes in red blood cell membrane protein composition during hae-

modialysis procedure”. Renal Failure, in press.

- As previously referred, leukocyte activation is part of an in-

flammatory response, and is an important source of ROS and prote-

ases, both of which may impose oxidative and proteolytic damages to

erythrocyte and plasma constituents. Actually, oxidative stress has been

reported to occur in haemodialysis patients and has been proposed as a

significant factor in haemodialysis-related shortened erythrocyte sur-

vival. We hypothesized that non-responders patients could have an

enhanced oxidative stress status and enhanced erythrocyte damage

and/or senescence. This hypothese was assessed in paper IX:

“Band 3 profile as a marker of erythrocyte changes in chronic kidney

disease patients”. The Open Clinical Chemistry Journal 2008; 1: 57-63.

- Different intensities in the inflammatory process and fibrinolyt-

ic/endothelial cell dysfunction have been found for the different types

of vascular access. Furthermore, there is some evidence that the use

of central vascular catheter increases the morbidity and mortality of

haemodialysis patients, by increasing the risk of infection and throm-

bosis. We hypothesized that the type of vascular access could modu-



late the inflammatory response and the endothelial (dys)function and

therefore, it could be a valuable factor in the response to rhEPO thera-

py. This hypothesis was assessed in paper X:

“Cross-talk between inflammation, coagulation/fibrinolysis and vascu-

lar access in Chronic Kidney Disease Patients under Haemodialysis”. J Vasc

Access, in press.

Part III - Publications





Paper I

“Inflammation, T-Cell Phenotype, and Inflammatory Cy-

tokines in Chronic Kidney Disease Patients Under Hemodialy-

sis and its Relationship to Resistance to Recombinant Human

Erythropoietin Therapy”

J Clin Immunol 2008; 28: 268-275.



Inflammation, T-Cell Phenotype, and InflammatoryCytokines in Chronic Kidney Disease PatientsUnder Hemodialysis and its Relationship to Resistanceto Recombinant Human Erythropoietin Therapy

Elísio Costa & Margarida Lima & João Moura Alves &

Susana Rocha & Petronila Rocha-Pereira &

Elisabeth Castro & Vasco Miranda &

Maria do Sameiro Faria & Alfredo Loureiro &

Alexandre Quintanilha & Luís Belo & Alice Santos-Silva

Received: 9 November 2007 /Accepted: 28 December 2007 /Published online: 18 January 2008# Springer Science + Business Media, LLC 2008

AbstractBackground Resistance to recombinant human erythropoi-etin (rhEPO) occurs in some chronic kidney disease (CKD)patients, which may be due to enhanced systemic inflam-matory response and to the erythropoiesis-suppressingeffect of pro-inflammatory cytokines, some of which areproduced by T cells.Aim of study The aim of this study was to investigate therelationship between resistance to rhEPO therapy inhemodialysis CKD patients and inflammatory markers [C-reactive protein (CRP), soluble interleukin (IL)-2 receptor(sIL2R), and serum albumin levels], blood cell counts, T-

cell phenotype, cytokine production by T cells, and serumcytokine levels.Materials and Methods We studied 50 hemodialysis CKDpatients, 25 responders and 25 nonresponders to rhEPO,and compared them to each other and with 25 healthycontrols. When compared to controls, CKD patientsshowed increased serum levels of CRP, IL-6, and sIL2Rand a T-cell lymphopenia, due to decreased numbers ofboth CD4+ and CD8+ T cells. T cells from CKD patientshad an immunophenotype compatible with chronic T-cellstimulation as shown by the increased percentage ofCD28−, CD57+, HLA-DR+, CD28−HLA-DR+, and CD57+

J Clin Immunol (2008) 28:268–275DOI 10.1007/s10875-007-9168-x

E. Costa (*) : S. Rocha : E. Castro : L. Belo :A. Santos-SilvaServiço de Bioquímica,Faculdade de Farmácia da Universidade do Porto,Rua Aníbal Cunha, 164,4099-030 Porto, Portugale-mail: [email protected]

E. Costa : S. Rocha : P. Rocha-Pereira : E. Castro :A. Quintanilha : L. Belo :A. Santos-SilvaInstituto de Biologia Molecular e Celular (IBMC),Universidade do Porto,Porto, Portugal

E. CostaDepartamento de Tecnologias de Diagnóstico e Terapêutica,Escola Superior de Saúde, Instituto Politécnico de Bragança,Porto, Portugal

M. Lima : J. M. AlvesLaboratório de Citometria, Hospital Geral Santo António, HGSA,Porto, Portugal

P. Rocha-PereiraUniversidade Beira Interior,Covilhã, Portugal

V. Miranda :M. d. S. FariaFresenius Medical Center, Dinefro-Diálises e Nefrologia, SA,Porto, Portugal

A. LoureiroUninefro-Sociedade Prestadora de Cuidados Médicos e de Diálise,SA,Porto, Portugal

A. QuintanilhaInstituto Ciências Biomédicas Abel Salazar, ICBAS,Universidade do Porto,Porto, Portugal



HLA-DR+ T cells and produce higher levels of IL-2, INF-γ,and TNF-α after short-term in vitro stimulation, althoughTh1 cytokines were not detectable in serum. Statisticallysignificant differences were found between responders andnonresponders to rhEPO therapy for total lymphocyte andCD4+ T-lymphocyte counts, albumin (lower in nonrespond-ers) and CRP (higher in nonresponders) levels.Conclusion CKD patients under hemodialysis present withraised inflammatory markers and decrease of total lympho-cyte and CD4+ T-lymphocyte counts when compared withcontrols. Some of those markers are even further enhancedin nonresponders to rhEPO therapy patients, but resistanceto this therapy cannot be justified by a Th1 polarized T-cellresponse.

Keywords rhEPO . Lymphocytes . Inflammation .

Resistance to rhEPO therapy

Introduction

Despite the technological advances in hemodialysis proce-dures and medical support in the last years, the mortalityand morbidity of patients with chronic kidney disease(CKD) remain 10 to 20 times higher than that found ingeneral population, and anemia is still an independent riskfactor [1–6].

In the last two decades, the institution of recombinanthuman erythropoietin (rhEPO) therapy allowed the correc-tion of the anemia in the majority of CKD patients,consequently reducing its associated complications andimproving significantly the quality of life. However, morethan 25% of these patients require high doses of rhEPO,and about 5% to 10% of them do not respond to the therapy[4, 5–8]. Many factors were described to be associated withrhEPO therapy resistance, namely, iron, vitamin B12, orfolate deficiencies, blood loss, hyperparathyroidism, oxida-tive stress, aluminum toxicity, infection, and inflammatoryconditions [3, 6–11].

Inflammatory stimuli induce the release of cytokines,which in turn causes many systemic changes, includingincreased synthesis and release of positive acute-phaseproteins, such as C-reactive protein (CRP), as well assuppression of negative ones, such as albumin andtransferrin [6–11]. The causes for the inflammatory re-sponse in hemodialysis patients are not obvious. There areseveral potential sources, including bacterial contaminationof the dialyser, dialyser membrane incompatibility, andinfections of the vascular access. However, the dialysisprocedure may only be partially responsible because evenpatients with renal insufficiency who are not yet on dialysishave raised inflammatory markers that rise further afterstarting regular hemodialysis treatment [3, 12, 13]. This

inflammatory response may mobilize iron from erythropoi-esis traffic to store sites within the reticuloendothelialsystem, inhibit erythroid progenitor proliferation anddifferentiation, blunt the response to erythropoietin, andaccelerate removal of erythrocytes coated with immunecomplexes or immunoglobulins. The erythropoiesis-sup-pressing effect of the increased activity of pro-inflammatorycytokines associated to this inflammatory condition hasbeen proposed as an important factor associated to rhEPOtherapy resistance [3, 6, 7, 12, 14, 15]. Pro-inflammatorycytokines such as interleukin (IL)-1, IL-2, IL-4, IL-6, tumornecrosis factor (TNF)-α, and interferon (INF)-γ diminishcolony formation of burst-forming (BFU-Es) and colony-forming (CFU-Es) unit-erythroid cells, suggesting thatthese cytokines may cause suppression of erythropoiesis[3, 16, 17].

Besides inflammation, end-renal failure induces a clin-ical state of immunodeficiency associated with a higherincidence of infections and a higher mortality due toinfectious complications. This immunodeficiency is char-acterized by a deficient response to some vaccinations,namely, hepatitis B virus. The mechanism responsible forthis immune defect is still unknown. However, it is knownthat lymphopenia is associated to end-renal failure, occur-ring in the B- and T-lymphocyte compartment, probablydue to increased apoptosis [18–20].

There are a considerable number of papers regarding theassociations between inflammation and resistance to rhEPOtherapy. However, only a few investigated at the same timeinflammatory markers, T-cell phenotype, T-cell cytokineproduction, and cytokine serum levels, and most of theseincluded a very limited number of patients. Furthermore,therapeutic and hemodialysis procedure approacheschanged in the last few years, namely, concerning hemo-dialysis membrane biocompatibility, and these changescould influence the causes of rhEPO resistance.

The aim of this study was to investigate the relationshipbetween resistance to rhEPO therapy and inflammatorymarkers, T-cell activation-related phenotype, cytokine pro-duction by T cells, and serum cytokine levels in hemodi-alysis CKD patients.

Materials and Methods

Subjects and Samples

Seventy-five individuals were included in this study: 50hemodialysis CKD patients (25 responders and 25 non-responders to rhEPO therapy) and 25 healthy controls. TherhEPO maintenance dose for responder patients was 7.51±6.52 U kg−1 week−1 Hb−1 and for nonresponders was 59.40±23.31 U kg−1 week−1 Hb−1.

J Clin Immunol (2008) 28:268–275 269269



Blood samples were collected before hemodialysis, withand without anticoagulant [ethylenediamine tetraacetic acid(EDTA) and sodium/lithium heparine] to obtain wholeblood, plasma, and serum.

Classification of CKD patients as responders or non-responders was performed in accordance with the EuropeanBest Practice Guidelines [21], which defines resistance torhEPO as a failure to achieve target hemoglobin levels (11–12 g/dl) with doses of rhEPO higher than 300 IU kg−1

week−1 of epoetin or 1.5 μg kg−1 week−1 of darbopoietin-α.Besides rhEPO therapy, all patients were under iron and

folate prophylactic therapies, in accordance to the recom-mendations of European Best Practice Guidelines [21] toavoid nutrient erythropoietic deficiencies.

The causes of renal failure in patient population were asfollows: diabetic nephropathy (n=16), chronic glomerulone-phritis (n=6), polycystic kidney disease (n=5), hypertensivenephrosclerosis (n=3), obstructive nephropathy (n=3), py-elonephritis associated with neurogenic bladder (n=1),nephrolithiasis (n=1), chronic interstitial nephritis (n=1),Alport syndrome (n=1), renal vascular disease due topolyarteritis (n=1), and uncertain etiology (n=12). Allpatients used high-flux polysulfone FX-class dialysers(Fresenius Medical Care, Bad Homburg, Germany). Nostatistically significant differences where found betweenresponders and nonresponders among CKD patientsconcerning to age, gender, body weight, body mass index,time on dialysis, urea reduction ratio, Kt/V and parathyroidhormone serum levels.

Patients with malignancy, hematological disorders, sys-temic autoimmune diseases, and acute infections wereexcluded. None of the patients were under vitamin-Dtherapy. All patients gave their informed consent toparticipate in this study.

Healthy volunteers, with normal hematological andbiochemical values, without any history of renal orinflammatory disease, were used as normal controls and,as far as possible, were age- and gender-matched with CKDpatients.

Assays

Hemoglobin (Hb) and white blood cell (WBC) count wasmeasured using an automatic counter (Sysmex K1000,Hamburg, Germany), and leukocyte differential countswere evaluated in Wright-stained blood films. Serum CRPwas measured by immunoturbidimetry (CRP latex HSRoche kit, Roche Diagnostics). Enzyme-linked immuno-sorbent assay was used for measurement of serum solubleinterleukin 2 receptor (sIL2R; Human IL-2 SRα, R&Dsystems, Minnesota, USA). Serum albumin levels weremeasured using a colorimetric assay end-point method(Albumin Plus; Roche GmbH, Mannheim, Germany), and

the immunoglobulins were quantified by immunoturbidim-etry (Tina-quant IgA, Tina-quant IgG, and Tina-quant IgM,Roche GmbH, Mannheim, Germany).

Flow Cytometry

T-cell cytokine production Heparinised blood cells werecultured for 4 h in Roswell Park Memorial Institute (RPMI)1640 medium at 37°C in a 5%CO2 and 95% humidity sterileenvironment in the presence of 25 ng/ml of phorbol-12myristate 13-acetate, 1 μg/ml of ionomycin, and 10 μg/mlof brefeldin A (stimulated samples) or only with brefeldinA (unstimulated samples). Immediately after the incubationperiod, cells were stained with allophycocyanin (APC)conjugated anti-CD3 and fluorescein isothiocyanate (FITC)conjugated anti-CD8 mouse anti-human monoclonal anti-bodies [MAbs; Becton Dickinson, Biosciences (BDB), SanJose, CA, USA] for 15 min in the dark at room temperature.After incubation, cells were washed once in 2 ml ofphosphate-buffered saline (PBS). After discarding thesupernatant, cells were fixed, permeabilized, and stainedwith phycoerythrin (PE) conjugated MAbs directed againsthuman cytokines (IL-2, TNF-α, and INF-γ) or PE-conjugated isotype-matched MAbs reagents (negative con-trols; Pharmingen, San Diego, CA, USA). For this purpose,the Fix & Perm reagent (Caltag, San Francisco, CA, USA)was used, strictly following the recommendations of themanufacturer. Once stained, cells were washed once in 2 mlof PBS, suspended in 0.5 ml of PBS, and analyzed in theflow cytometer.

T-cell phenotype Cell surface markers were evaluated byflow cytometry with a whole-blood stain-lyse-and-then-wash method, using the florescence-activated cell sorting(FACS) lysing solution (BDB). Cells were stained withanti-CD8 APC or anti-CD4 APC, anti-CD28 PE-cyanin 5(PC5), anti-HLA-DR PE, and anti-CD57 FITC MAbs(BDB). HLA-DR and CD57 expression were used toevaluate early and late T-cell activation, and absence ofCD28 expression on T lymphocytes was used to quantifythe fraction of memory effector T cells/large granularlymphocytes [22–25].

Serum cytokine levels IL-2, IL-4, IL-6, IL-10, TNF-α, andINF-γ serum levels were quantified using the BD™Cytometric Bead Array Human Th1/Th2 Cytokine Kit II(BDB).

Data acquisition and analysis Data acquisition wasperformed in a FACSCalibur flow cytometer, equippedwith a 488-nm argon ion and a 635-nm red diode laser,using the CellQUEST™ software program (BDB). For T-cell phenotype and T-cell cytokine production analysis,

270 J Clin Immunol (2008) 28:268–275



the Paint-A-Gate PRO software program (BDB) wasused. To quantify the expression of costimulatorymolecules and activation-related markers on T cells, thepercentage of HLA-DR+, CD28+, and CD57+ cells wererecorded after gating for CD8+ and CD4+ T lymphocytes.Evaluation of cytokine production by CD8+ and CD4+ Tcells was based on the percentage of cytokine+ cells withinCD3+ CD8+ and CD3+ CD8− lymphocytes, respectively.Serum cytokine levels analysis was performed using theBDTM CBA Software.

Statistical Analysis

For statistical analysis, we used the Statistical Package forSocial Sciences, version 14.0. Kolmogorov–Smirnov sta-tistics was used to evaluate sample normality distribution.Multiple comparisons between groups were performed byone-way analysis of variance supplemented with Tukey’shonestly significant difference (HSD) post-hoc test. Forsingle comparisons, we used the Student’s t test wheneverthe parameters presented a Gaussian distribution and the

Table I Blood Cell Count and Immunoglobulin Serum Levels

Healthy Controls(n=25)

CKD Patients

Total (n=50) rhEPO Responders (n=25) rhEPO Nonresponders (n=25)

Hb (g/dl) 14.1±1.3 (12.0–17.0) 11.1±1.8 (7.1–15.5)* 11.9±1.5 (9.5–15.5)* 10.3±1.7 (7.1–11.9)*,***White blood cells (109/l) 5.8±1.6 (3.5–9.6) 6.2±2.2 (3.0–12.3) 6.4±1.9 (4.2–11.5) 6.0±2.4 (3.0–12.3)Lymphocytes (109/l) 2.2±0.7 (1.1–3.5) 1.4±0.6 (0.4–3.1)* 1.6±0.5 (0.5–3.0)** 1.2±0.6 (0.4–3.1)*,***CD3+ T cells (109/l) 1.6±0.7 (0.5–3.0) 1.0±0.5 (0.2–2.4)** 1.2±0.6 (0.2–2.4) 0.9±0.5 (0.2–2.2)**,***CD4+ T cells (109/l) 1.1±0.5 (0.3–2.2) 0.7±0.3 (0.1–1.16)** 0.8±0.4 (0.1–1.6)** 0.6±0.3 (0.1–1.13)**,***CD8+ T cells (109/l) 0.5±0.2 (0.2–0.8) 0.3±0.2 (0.1–0.9)** 0.3±0.2 (0.1–0.7) 0.3±0.2 (0.1–0.9)**CD4 /CD8 ratio 2.7±1.0 (1.3–4.2) 3.0±1.4 (0.8–7.9) 2.9±1.7 (1.2–7.9) 2.6±1.3 (0.8–6.0)Monocytes (109/l) 0.2±0.1 (0.1–0.4) 0.4±0.2 (0.2–0.9)** 0.4±0.1 (0.2–0.7)* 0.3±0.2 (0.2–0.9)Neutrophils (109/l) 3.0±1.0 (1.5–5.2) 4.1±1.7 (1.8–8.9)** 4.1±1.7 (2.0–8.9)** 4.1±1.8 (1.8–8.3)**Eosinophils (109/l) 0.2±0.3 (0.0–1.6) 0.2±0.2 (0.0–1.1) 0.2±0.2 (0.0–1.1) 0.2±0.1 (0.0–0.5)Neutrophil/Lymphocyte ratio 1.37±0.51 (0.56–2.66) 3.34±1.92 (0.96–8.25)* 2.80±1.14 (1.28–8.19)* 3.88±2.05 (0.96–8.25)*,***IgG (mg/dl) 977±161 (688–1279) 1078±353 (509–2163) 1043±315 (633–2163) 1112±392 (509–1986)IgA (mg/dl) 215±86 (77–387) 275±126 (91–653) 275±119 (91–542) 275±136 (104–653)IgM (mg/dl) 122±65 (33–277) 106±78 (29–518) 100±57 (29–276) 113±95 (30–518)

Results are presented as mean±one standard deviation (minimum–maximum)*p<0.0001 vs controls**p<0.05 vs controls***p<0.05 vs responders

Table II Expression of HLA-DR, CD57 and CD28 Molecules on Blood CD4+ and CD8+ T Cells

Healthy Controls (n=12) CKD Patients


% CD4+ T cellsHLA-DR+ 8.7±4.2 (4.7–18.6) 17.4±7.4 (6.8–41.3)* 16.4±9.1 (6.8- 41.3)* 18.8±4.2 (12.8–25.8)*CD57+ 2.4±3.4 (0.4–12.6) 9.0±7.7 (0.6–32.0)* 8.2±8.9 (0.7–32.0)* 10.8±6.7 (0.6–22.1)*CD28− 0.9±0.9 (0.3–3.1) 7.0±7.0 (0.1–23.5)* 6.6±5.5 (0.1–17.0)* 8.6±8.5 (0.1–23.5)*CD57+ HLA-DR+ 23.7±13.6 (0.3–44.2) 44.9±19.8 (13.9–8.3)* 43.3±17.8 (19.7–68.3)* 47.2±23.7 (13.9–67.4)*CD28− HLA-DR+ 25.3±23.2 (9.7–72.2) 43.6±19.5 (16.7–77.0)* 42.8±20.4 (19.9–69.1)* 44.7±19.8 (16.7–77.0)*% CD8+ T cellsHLA-DR+ 25.1±12.8 (10.4–44.4) 45.7±16.5 (17.3–73.5)* 41.4±17.4 (17.3–65.2)* 51.5±14.1 (32.1–73.5)*CD57+ 21.1±13.5 (4.9–43.4) 38.3±15.9 (14.2–60.0)* 34.8±11.2 (16.4–52.9)* 44.7±19.4 (14.2–60.0)*CD28− 28.4±13.7 (11.8–57.2) 44.0±22.9 (7.5–85.7)* 41.3±19.0 (15.0–64.2)* 49.5±26.9 (7.5–85.7)*CD28− HLA-DR+ 33.7±18.7 (0.1–61.3) 54.8±19.3 (22.6–86.0)* 50.4±18.0 (22.6–73.4)* 58.9±20.7 (31.4–86.0)*CD57+ HLA-DR+ 35.1±21.2 (0.1–66.2) 54.9±18.2 (18.0–81.7)* 50.2±16.2 (18.0–72.1) 58.5±19.8 (27.2–81.7)*

Results are presented as mean±one standard deviation (minimum–maximum)*p<0.05 vs controls

J Clin Immunol (2008) 28:268–275 271271



Mann–Whitney U test in the case of a non-Gaussiandistribution. Significance was accepted at p less than 0.05.

Results

The hematological characteristics of the three groups studiedare summarized in Table I. Hemodialysis CKD patients weremore anemic than controls, and nonresponder patients wereeven more anemic than responder patients. No differencewas found between the three groups of patients concerningthe WBC count. However, CKD patients showed lympho-penia as well as increased monocyte and neutrophil countsand increased neutrophil/lymphocyte ratio, as compared tocontrols. Lymphopenia observed in CKD patients results atleast in part from a decrease in total circulating CD3+ Tlymphocytes and affects both the CD4+ and the CD8+ T-cellsubsets (p<0.05). Statistically significant differences were

found between responders and nonresponders to rhEPOtherapy, concerning total lymphocyte and CD4+ T-cell counts(lower for nonresponders; p<0.05), and to neutrophil/lymphocyte ratio (higher for nonresponders; p<0.05). Nostatistically significant differences were found between thethree groups of individuals concerning immunoglobulinserum levels (Table I).

CKD patients showed a statistically significant increase(p<0.05) in the proportion of CD28−, CD57+, HLA-DR+,CD28−HLA-DR+ and CD57+ HLA-DR+ T cells whencompared to controls, both in the CD4+ and in the CD8+

T-cell compartments (Table II). No statistically significantdifferences were found between responders and nonrespond-ers to rhEPO concerning expression of CD28 co-stimulatory,and CD57 and HLA-DR cell-surface activation-relatedmolecules.

CKD patients showed a statistically significant increasein serum levels of CRP, sIL2R, and IL-6 as compared to

Table III Inflammatory Markers and Serum Cytokine Levels of Studied Subjects

Healthy Controls (n=25) CKD patients


Albumin (g/dl) NM 3.8±0.4 (2.7–4.5) 4.0±0.4 (2.9–4.5) 3.7±0.4 (2.7–4.2)***CPR (mg/dl) 1.75 (0.76–4.47) 5.75 (1.90–14.01)* 3.20 (1.73–7.23)** 10.14 (3.82–38.99)*,***sIL2R (nmol/l) 758±235 (324–1290) 4199±1762 (1795–8377)* 4005±1835 (1795–8377)* 4394±1701 (2232–8265)*IL-2 (pg/ml) 2.5±3.0 (0.0–9.3) ND ND NDIL-4 (pg/ml) ND ND ND NDIL-6 (pg/ml) 1.90 (0–3.75) 7.80 (3.85–15.05)* 5.75 (3.83–13.95)* 8.80 (4.55 – 21.30)*IL-10 (pg/ml) 1.20 (0–2.80) 1.10 (0–2.15) 0.0 (0–1.6) 1.40 (0–2.25)***TNF-α (pg/ml) 1.40 (0–3.55) ND ND NDINF-γ (pg/ml) ND ND ND ND

Results are presented as mean±one standard deviation (minimum–maximum) or as median values (inter-quartile range). Zero values representsamples in which serum levels of the mentioned cytokine were undetectable: IL-2 (7 out of 18 controls’ and all patients’ samples); IL-6 (7 out of18 controls’ samples); TNF-α (8 out of 18 controls’ and all patients’ samples); IL-10 (8 out of 18 controls’ and 25 out of 50 patients’ samples; 17out of 25 responders and 8 out of 25 nonresponders)NM Not made; ND not detected (limit of detection <1 pg/ml)*p<0.0001, vs controls**p<0.05, vs controls***p<0.05 vs responders

Fig. 1 Illustrative dot plots showing the IL-2, INF-γ, and TNF-α expression on CD8+ and CD8− cells, after gating on CD3+ T lymphocytes. Reddots CD3+ CD8− Cytokine−; yellow dots CD3+ CD8+ Cytokine−; violet dots CD3+ CD8− Cytokine+; black dots CD3+ CD8+ Cytokine+ T cells

272 J Clin Immunol (2008) 28:268–275



controls (p<0.0001; Table III). In contrast, serum levels ofIL-2 and TNF-α were lower in CKD patients, these Th1cytokines being detected in 11 and 10 out of 18 controlsamples, respectively, whereas in none of the CKDpatients’ samples. Concerning IL-10, no differences werefound between CKD patients and controls, this cytokinebeing detected in 10 out of 18 controls’ and in 25 out of 50CKD patients’ samples. Serum levels of IL-4 and INF-γwere undetectable in both control’s and patient’s samples.

T cells from CKD patients produced higher amounts ofIL-2, TNF-α, and IFN-γ after short-term in vitro stimula-tion as compared to controls (Figs. 1 and 2). Statisticallysignificant differences (p<0.05) were observed for IL-2,INF-γ, and TNF-α in the case of CD8+ T cells but only forIL-2 and TNF-α in the case of CD4+ T cells.

When comparing responders and nonresponders CKDpatients, we found that nonresponders showed decreasedserum levels of albumin (p<0.01) and increased CRP (p<0.05) and IL-10 (p<0.05) serum levels, this cytokine beingdetected in 17 of the 25 analyzed samples, as compared to8 out of the 25 responder’s samples (Table III). Nodifferences were observed between responders and non-responders in the levels of the other soluble mediators,neither on the ability to produce Th1 cytokines after short-term in vitro stimulation.

Discussion

In this study, we demonstrated that CKD patients underhemodialysis treatment show neutrophilia, increased neu-trophil/lymphocyte ratio, and higher levels of IL-6, sIL2R,and CRP pro-inflammatory molecules in the serum, anddecreased albumin serum levels, confirming the presence ofan inflammatory process. The fact that nonresponders hadhigher CRP and lower albumin serum levels as compared toresponders would suggest a relationship between resistanceto rhEPO therapy and the magnitude of the inflammatoryresponse, already described in literature [13, 14, 26–28].

In addition, this study demonstrated that CKD patientshave lower number of total lymphocytes, CD4+ and CD8+

T lymphocytes, when compared to controls, althoughstatistically significant differences between responders andnonresponders to rhEPO therapy were found only for totallymphocytes and CD4+ T cells, which were lower innonresponders. T and B lymphopenia have been previouslydescribed in literature in CKD patients, associated withrenal-replacement therapy [29, 30]. B lymphopenia hasbeen associated with stage 5 of CKD and T lymphopenia toa selective depletion of naïve CD4+ and CD8+ T lympho-cytes and CD4+ central memory population [31]. There aresome possible explanations for lymphocyte depletion inCKD patients, namely, increased turnover, disturbance of

lymphocyte homeostasis due to uremia and increasedperipheral lymphocyte apoptosis associated with activationstimulus [14, 31–35]. This lymphocyte depletion could be

Fig. 2 Percentage of Cytokine+ cells among CD4+ (CD3+ CD8−; whitebars) and CD8+ (CD3+ CD8+; grey bars) T cells after short-term invitro stimulation with phorbol-12 myristate 13-acetate plus Ionomycinin controls, as well as in responder and nonresponder CKD patients.Boxplot shows median value (horizontal line in box) and first and thirdquartiles (inferior and superior lines of the box, respectively)

J Clin Immunol (2008) 28:268–275 273273



exacerbated in nonresponders, justifying the differencefound between the two groups of patients concerning totallymphocytes and CD4+ T lymphocytes.

Increased proportions of both early (HLA-DR) and late-activation (CD57) markers on both CD4+ and CD8+ T-cellsubsets are compatible with an enhanced continuousactivation state in CKD patients [22]. This probably resultsfrom persistent antigen stimulation/chronic inflammationassociated with hemodialysis and/or chronic renal failure,which probably also justifies the increase in the fraction ofmemory effector T cells/large granular lymphocytes, asshown by the higher percentage of CD4+/CD28− and CD8+/CD28− T cells observed in hemodialysis CKD patientswhen compared to controls [25]. Curiously, higher levels ofexpression of HLA-DR and CD57 were observed not onlyon total CD4+ and CD8+ T cells, but also on CD4+/CD28−

and CD8+/CD28− T-cell subsets. On literature, we foundonly one paper [14] describing cytokine and CD28 T-cellexpression in CKD patients, and the authors showedstatistically significant differences between responders andnonresponders to rhEPO therapy. However, in our study, nodifferences were found between the two groups of patients.These differences could be due to the difference betweenstudied CKD patient population, namely, concerning theexistence of other rhEPO resistance-related factors.

In accordance to their enhanced activation state, T cellsfrom CKD patients under hemodialysis treatment show anenhanced ability to produce Th1-related cytokines (IL-2,INF-γ, and TNF-α) after short-term in vitro stimulation,although these cytokines were undetectable in serum. Thisincreased capacity to produce Th1 cytokines could justify, atleast in part, the anemia found in CKD patients. In fact, thesecytokines are described to be associated to an inhibitoryeffect to the formation of CFU-Es and BFU-Es [12–14, 16].However, this enhanced capacity to produce cytokines is notassociated to the refractoriness to rhEPO therapy, aspreviously described in nonresponder CKD patients [7].

In conclusion, our results show that CKD patients havean enhanced inflammatory response as well as evidence ofTh1 polarized T-cell activation process and suggest thatresistance to rhEPO therapy is associated to inflammationmarkers and CD4+ lymphopenia but cannot be ascribed toan enhanced T-cell activation state neither to a mediatedTh1 response. Further studies are required to understand themechanism of lymphocyte loss and its consequences in theresponse to rhEPO therapy.

Acknowledgments The authors are grateful to Amgen for financialsupport and to the nurses of Fresenius Medical Center, Dinefro-Diálises e Nefrologia, SA and Uninefro-Sociedade Prestadora deCuidados Médicos e de Diálise, SA, for technical support. This studywas supported by a PhD grant (SFRH/BD/27688/2006) attributed toE. Costa by Fundação para a Ciencia e Tecnologia (FCT) and FundoSocial Europeu.

References

1. Kerr PG. Renal anaemia: recent developments, innovativeapproaches and future directions for improved management.Nephrology. 2006;11:542–8.

2. Macdougall IC. Present and future strategies in the treatment ofrenal anaemia. Nephrol Dial Transpl. 2001;16(suppl 5):50–5.

3. Macdougall IC, Cooper AC. Erythropoietin resistance: the role ofinflammation and pro-inflammatory cytokines. Nephrol DialTransplant. 2002;17(Suppl 11):39–43.

4. Foley RN, Parfrey PS, Sarnak MJ. Clinical epidemiology ofcardiovascular disease in chronic renal failure. Am J Kidney Dis.1998;32(suppl 3):S112–9.

5. Foley RN, Parfrey PS, Harnett JD, Kent GM, Murray DC, BarrePE. The impact of anemia on cardiomyopathy morbidity andmortality in end-stage renal disease. Am J Kidney Dis. 1996;28:53–61.

6. Smrzova J, Balla J, Bárány P. Inflammation and resistance toerythropoiesis-stimulting agents—what do we know and what needsto be clarified. Nephrol Dial Transpant. 2005;20(Suppl 8):viii2–7.

7. Cooper AC, Mikhail A, Lethbridge MW, Kemeny DM, MacdougallIC. Increased expression of erythropoiesis inhibiting cytokines(IFN-γ, TNF-α, IL-10, and IL-13) by T cells in patients exhibitinga poor response to erythropoietin therapy. J Am Soc Nephrol.2003;14:1776–84.

8. Macdougal IC. Poor response to erythropoietin: practical guide-lines on investigation and management. Nephrol Dial Transplant.1995;10:607–14.

9. Drueke TB. R-HuEPO hyporesponsiveness: who and why.Nephrol Dial Transplant. 1995;10(suppl 2):S62–8.

10. Danielson B. R-HuEPO hyporesponsiveness: who and why.Nephol Dial Transplant. 1995;10(suppl 2):S69–73.

11. Fishbane S. Hyporesponsiveness to recombinant human erythro-poietin in dialysis patients. Dial Transplant. 2000;29:545–8.

12. Schindler R, Senf R, Frei U. Influencing the inflammatoryresponse of haemodialysis patients by cytokine elimination usinglarge-pore membranes. Nephrol Dial Transplant. 2002;17:17–9.

13. Gunnell J, Yeun JY, Depner TA, Kaysen GA. Acute-phaseresponse predicts erythropoietin resistance in hemodialysis andperitoneal dialysis patients. Am J Kidney Dis. 1999;33:63–72.

14. Cooper AC, Breen CP, Vyas B, Ochola J, Kemeny DM,Macdougall IC. Poor response to recombinant erythropoitin isassociated with loss of T-lymphocyte CD28 expression andaltered interleukin-10 production. Nephrol Dial Transplant.2003;18:133–40.

15. Descamps-Latscha B, Herbelin A, Nguyen AT, Roux-Lombard P,Zingraff J, Moynot A, Verger C, Dahmane D, Groote D, JungersP, Dayer JM. Balance between IL-1β, TNF-α, and their specificinhibitors in chronic renal failure and maintenance dialysis. JImmunology. 1995;154:882–92.

16. Means RT Jr, Krantz SB. Inhibition of human erythroid colonyforming units by interferons α and β: differing mechanismsdespite shared receptor. Exp Haematol. 1996;24:204–8.

17. Allen DA, Breen C, Yaqoob MM, Macdougall IC. Inhibition ofCFU-E colony formation in uremic patients with inflammatorydisease: role of INF-γ and TNF-α. J Invest Med. 1999;47:204–11.

18. Meier P, Dayer E, Blanc E, Wauters JP. Early T cell activationcorrelates with expression of apoptosis markers in patients withend-stage renal disease. J Am Soc Nephol. 2002;13:204–12.

19. Waltzer WC, Bachvaroff RJ, Raisbeck AP, Egelandsdal B, PullisC, Shen L, Rapaport FT. Immunological monitoring in patientswith end-stage renal disease. J Clin Immunol. 1984;4:364–8.

20. Litjens NHR, van Druninger CJ, Betjes MGH. Progressive loss ofrenal function is associated with activation and depletion of naïveT lymphocytes. Clinical Immunology. 2006;118:83–91.

274 J Clin Immunol (2008) 28:268–275



21. Locatelli F, Aljama P, Barany P, European Best Practice Guide-lines Working Group, et al. Revised European best practiceguidelines for the management of anaemia in patients with chronicrenal failure. Nephrol Dial Transplant. 2004;19(Suppl 2):ii1–47.

22. d’Angeac AD, Monier S, Pilling D, Travaglio-Encinoza A, RemeT, Salmon M. CD57+ T lymphocytes are derived from CD57-precursors by differentiation occurring in late immune responses.Eur J Immunol. 1997;24:1503–11.

23. Lima M, Teixeira Mdos A, Queiros ML, Santos AH, GoncalvesC, Correia J, Farinha F, Mendonca F, Soares JM, Almeida J, OrfaoA, Justica B. Immunophenotype and TCR-Vbeta repertoire ofperipheral blood T-cells in acute infectious mononucleosis. BloodCells Mol Dis. 2003;30:1–12.

24. Lima M, Almeida J, Teixeira MA, Santos AH, Queiros ML,Fonseca S, Moura J, Goncalves M, Orfao A, Pinto Ribeiro AC.Reactive phenotypes after acute and chronic NK-cell activation. JBiol Regul Homeost Agents. 2004;18:331–4.

25. Romero P, Zippelius A, Kurth I, Pittet MJ, Touvrey C, Iancu EM,Corthesy P, Devevre E, Speiser DE, Rufer N. Four functionallydistinct populations of human effector-memory CD8+ T lympho-cytes. J Immunol. 2007;178:4112–9.

26. Bárány P. Inflammation, serum C-reactive protein, and erythro-poietin resistance. Nephrol Dial Transplant. 2001;16:224–7.

27. Spittle MA, Hoenich NA, Handelman GJ, Adhikarla R, Homel P,Levin NW. Oxidative stress and inflammation in hemodialysispatients. Am J Kidney Dis. 2001;38:1408–13.

28. Reddan DN, Klassen PS, Szczech LA, Coladonato JA, O’Shea S,Owen WF Jr, Lowrie EG. White blood cells as a novel mortalitypredictor in haemodialysis patients. Nephrol Dial Transplant.2003;18:1167–73.

29. Fernandez-Fresnedo G, Ramos MA, Gonzalez-Pardo MC, deFrancisco AL, Lopez-Hoyos M, Aris M. B lymphopenia in uremiais related to an accelerated in vitro apoptosis and dysregulation ofBcl-2. Nephrol Dial Transplant. 2000;15:502–10.

30. Kurz P, Kohler H, Meuer S, Hutteroth T, Meyer zumBuschenfelde KH. Impaired cellular immune response inchronic renal failure: evidence for a T cell defect. Kidney Int.1986;29:1209–14.

31. Litjens NHR, van Druningen CJ, Betjes MGH. Progressiveloss of renal function is associated with activation anddepletion of naïve T lymphocytes. Clinical Immunology.2006;118:83–91.

32. Kang I, Hong MS, Nolasco H, Park SH, Dan JM, Choi JY, Craft J.Age-associated change in the frequency of memory CD4+ T cellsimpairs long term CD4+ T cell responses to influenza vaccine. JImmunol. 2004;173:673–81.

33. Napolitano LA, Grant RM, Dreeks SG, Schmidt D, De Rosa SC,Herzenberg LA, Herndier BG, Andersson J, McCune JM.Increased production of IL-7 accompanies HIV-1-mediated T-celldepletion: implications for T-cell homeostasis. Nat Med.2001;7:73–9.

34. Moser B, Roth G, Brunner M, Lilaj T, Deicher R, Wolner E,Kovarik J, Boltz-Nitulescu G, Vychytil A, Ankersmit HJ.Aberrant T cell activation and heightened apoptotic turnover inend-stage renal failure patients: a comparative evaluation betweennon-dialysis, haemodialysis, and peritoneal dialysis. BiochemBiophys Res Commun. 2003;308:581–5.

35. Jiang Q, Li WQ, Aiello FB, Mazzucchelli B, Asefa B, KhaledAR, Durum SK. Cell biology of IL-7, a key lymphotrophim.Cytokine Growth Factor Rev. 2005;16:513–33.

J Clin Immunol (2008) 28:268–275 275275



Paper II

“IL-7 serum levels and lymphopenia in hemodialysis pa-

tients, non-responders to recombinant human erythropoietin

therapy”

Blood Cells Mol Dis 2008; 41: 134-135.



Letter to the Editor

IL-7 serum levels and lymphopenia in hemodialysis patients,non-responders to recombinant human erythropoietin therapy

There is a marked variability in the sensitivity to recombinanthuman erythropoietin (rhEPO) therapy in hemodialysis patients, withup to 10-fold variability in dose requirements to achieve correction ofanemia [1]. Approximately 5 to 10% of patients show a markedresistance to rhEPO therapy [2]. This variability however remainspoorly clarified. We showed that hemodialysis patients have lymphope-nia [3], which results at least in part from a decrease in total circulatingCD3+ T-lymphocytes and affects both the CD4+ and the CD8+ T-cellsubsets.Wealsoobserved thatnon-responders to rhEPO therapypatientspresentwith a lower number of total lymphocyte and CD4+ T-cell counts,when compared with responder patients. The lymphocyte depletionfound in hemodialysis patients, particularly in non-responders to rhEPOtherapy, could be related to an increased turnover of lymphocytes, to adisturbance in lymphocyte homeostasis due to uraemia, and/or toincreased peripheral lymphocyte apoptosis associated with an activa-tion stimulus.

Interleukin (IL)-7 has emerged as a key cytokine involved in con-trolling the homeostatic turnover and the survival of resting memoryCD4+ T cells [4]. We hypothesized that the serum levels of this inter-leukin could be related to the decreased number of total lymphocyteand CD4+ T-cell counts that we have found in hemodialysis patients,especially in non-responders to rhEPO therapy. In order to test thishypothesis we selected 63 hemodialysis patients (36 males, 27females; mean age 62.1±15.7) under rhEPO treatment, for a mediantime period of 36 months. The hemodialysis patients included 32responders and 31 non-responders to rhEPO therapy. Classification ofhemodialysis patients, as responders or non-responders, was per-formed in accordance with the European Best Practice Guidelines [5].The rhEPO maintenance dose for responder patients was 90±58 U/kgper week and for non-responders was 573±194 U/kg per week. Thetwo groups of patients were matched for age, gender, weight, bodymass index, mean time on hemodialysis, urea reduction ratio, ureaKtv, and serum parathyroid hormone levels. Patients with auto-immune disease, malignancy, hematological disorders, and acute orchronic infection were excluded from the study. Healthy volunteers(n=26), with normal hematological and biochemical values, withoutany history of renal or inflammatory disease, were used as normalcontrols. They werematched as far as possible for age and gender withhemodialysis patients.

Serum levels of IL-7 were determined by the Quantikine highsensitivity immunoassay (R & D Systems, Minneapolic, Minnesota,USA) according to the manufacturer's recommendations.

No statistically difference was found between patients andcontrols in total white blood cell count; however, hemodialysispatients were lymphopenic [1.5±0.6 vs 2.2±0.7×109/L, pb0.05),

which was the result, at least in part, of a decrease in totalcirculating CD3+ T-lymphocytes [1.0 (0.7–1.3×109/L) vs 1.6 (1.0–2.0×109/L), pb0.05] and both the CD4+ [0.7 (0.4–0.9×109/L) vs 1.0(0.8–1.6×109/L), pb0.05] and the CD8+ [0.3 (0.1–0.4×109/L) vs 0.5(0.3–0.7×109/L), pb0.05] T-cell subsets. When comparing respon-ders and non-responders to rhEPO therapy, statistically significantdifferences were found for total lymphocyte [responders: 1.6±0.5×109/L vs non-responders: 1.4×0.7×109/L, pb0.05] and CD4+

T-cell counts [responders: 0.8 (0.5–0.9×109/L) vs non-responders:0.5 (0.4–0.9×109/L), pb0.05].

No significant difference was found in serum IL-7 levels (pN0.05)between hemodialysis patients [10.5 (7.4–13.5 pg/mL)] and healthycontrols [11.2 (7.2–16.0 pg/mL)], suggesting that the lymphopeniafound in these patients is not associated with increased serum levelsof IL-7. Moreover, no correlation between IL-7 serum levels and totallymphocyte count (r=−0.21; p=0.11) was found. However, among thetwo groups of patients, non-responders showed significantly higherIL-7 serum levels [non-responders: 12.0 (7.6–17.8 pg/mL) vs respon-ders: 9.6 (6.1–11.5 pg/mL), pb0.05] (Fig. 1), suggesting a relationshipbetween the increased levels of this cytokine and decreased numberof total lymphocytes and CD4+ T-cell count in the non-responderpatients.

Increased IL-7 serum levels have also been described in otherclinical settings associated with lymphopenia, namely marrowtransplantation, HIV infection, chemotherapy treatment for cancer,and idiopathic CD4+ lymphopenia [6]. The increased IL-7 serum levelsfound in our lymphopenic patients could be related to T-cell depletion

Blood Cells, Molecules, and Diseases 41 (2008) 134–135

Fig. 1. Serum IL-7 levels among controls and hemodialysis patients according to theresponse to rhEPO therapy. Boxplot shows median value (horizontal line in box) andfirst and third quartiles (inferior and superior line of the box, respectively).

1079-9796/$ – see front matter © 2008 Elsevier Inc. All rights reserved.doi:10.1016/j.bcmd.2008.02.007

Contents lists available at ScienceDirect

Blood Cells, Molecules, and Diseases

j ourna l homepage: www.e lsev ie r.com/ locate /ybcmd



[6]. Elevated serum IL-7 concentration may predict a lessened res-ponse to rhEPO therapy.

Acknowledgments

The authors are grateful to Amgen for the financial support, and tothe nurses of Fresenius Medical Center, Dinefro – Diálises e Nefrologia,SA and Uninefro – Sociedade Prestadora de Cuidados Médicos e deDiálise, SA, for their technical support. This study was supported by aPhD grant (SFRH/BD/27688/2006) attributed to E. Costa by Fundaçãopara a Ciencia e Tecnologia (FCT) and Fundo Social Europeu.

References

[1] J. Smrzova, J. Balla, P. Bárány, Inflammation and resistance to erythropoiesis-stimulting agents—what dowe knowandwhat needs to be clarified? Nephrol. Dial.Transpant. 20 (Suppl 8) (2005) viii2–viii7.

[2] P. Bárány, Inflamation, serum C-reactive protein, and erythropoietin resistance,Nephrol. Dial. Transplant. 16 (2001) 224–227.

[3] E. Costa, M. Lima, J.M. Alves, S. Rocha, P. Rocha-Pereira, E. Castro, V. Miranda, M.Sameiro-Faria, A. Loureiro, A. Quintanilha, L. Belo, A. Santos-Silva, Inflammation, T-cell phenotype and inflammatory cytokines in chronic kidney disease patientsunder haemodialysis and its relationship to resistance to recombinant humanerythropoietin therapy. J Clin Immunol. (in press).

[4] L.A. Napolitano, R.M. Grant, S.G. Deeks, D. Schmidt, S.C. De Rosa, L.A. Herzenberg, B.G.Herndier, J. Andersson, J.M. McCune, Increased production of IL-7 accompanies HIV-1-mediated T-cell depletion: implications for T-cell homeostasis, Nat. Med. 7 (2001)73–79.

[5] F. Locatelli, P. Aljama, P. Barany, et al., European Best Practice Guidelines WorkingGroup. Revised European best practice guidelines for the management of anaemiain patients with chronic renal failure, Nephrol. Dial. Transplant. 19 (Suppl 2) (2004)ii1–ii47.

[6] T.J. Fry, C.L. Mackall, The many faces of IL-7: from lymphopoiesis to peripheral T cellmaintenance, J. Immunol. 174 (2005) 6571–6576.

Elísio Costa⁎Susana Rocha

Elisabeth CastroLuís Belo

Alice Santos-SilvaFaculdade Farmácia, Serviço de Bioquímica,

Universidade do Porto, Porto, PortugalE-mail addresses: [email protected], [email protected].

⁎Corresponding author. Faculdade Farmácia, Serviço de Bioquímica,Universidade do Porto, Rua Aníbal Cunha,

164, 4099-030 Porto, Portugal

Alexandre QuintanilhaElísio Costa

Susana RochaElisabeth Castro

Luís BeloAlice Santos-Silva

Petronila Rocha-PereiraInstituto de Biologia Molecular e Celular (IBMC),

Universidade do Porto, Porto, Portugal

Elísio CostaDepartamento de Tecnologias de Diagnóstico e Terapêutica,Escola Superior de Saúde, Instituto Politécnico de Bragança,

Porto, Portugal

Margarida LimaLaboratório de Citometria,

Hospital Geral Santo António, HGSA, Porto, Portugal

Petronila Rocha-PereiraCentro de Investigação em Ciências da Saúde (CICS),

Universidade Beira Interior, Covilhã, Portugal

Vasco MirandaMaria do Sameiro Faria

Fresenius Medical Center, Dinefro – Diálises e Nefrologia,SA, Portugal

Alfredo LoureiroUninefro – Sociedade Prestadora de Cuidados Médicos e de Diálise,

SA, Portugal

Alexandre QuintanilhaInstituto Ciências Biomédicas Abel Salazar (ICBAS),

Universidade do Porto, Porto, Portugal

22 February 2008

135Letter to the Editor



Paper III

“Neutrophil activation and resistance to recombinant

human erythropoietin therapy in hemodialysis patients.”

Am J Nephrol 2008; 28: 935-940.



Fax +41 61 306 12 34E-Mail [email protected]

Original Report: Laboratory Investigation

Am J Nephrol 2008;28:935–940 DOI: 10.1159/000142147

Neutrophil Activation and Resistance to Recombinant Human Erythropoietin Therapy in Hemodialysis Patients

Elísio Costa a–c Susana Rocha a, b Petronila Rocha-Pereira b, d

Henrique Nascimento a, b Elisabeth Castro a, b Vasco Miranda e

Maria do Sameiro Faria e Alfredo Loureiro f Alexandre Quintanilha b, g

Luís Belo a, b Alice Santos-Silva a, b

a Faculdade Farmácia, Serviço de Bioquímica, Universidade do Porto, Porto ; b Instituto Biologia Molecular eCelular, Universidade do Porto, Porto ; c Instituto de Ciências da Saúde, Centro Regional do Porto da Universidade Católica Portuguesa, Porto; d Centro de Investigação em Ciências da Saúde, Universidade Beira Interior, Covilhã ; e FMC, Dinefro – Diálises e Nefrologia, SA, Maia ; f Uninefro – Sociedade Prestadora de Cuidados Médicos e de Diálise, SA, Santo Tirso ; g Instituto Ciências Biomédicas Abel Salazar, Universidade do Porto, Porto , Portugal

pared with controls, CKD patients presented with signifi-cantly higher CRP, neutrophil and elastase levels. When we compared the 2 groups of patients, we found that non- responders presented statistically significantly higher elas-tase plasma levels. A positive significant correlation was found between elastase levels and weekly rhEPO dose and CRP serum levels. After the hemodialysis procedure, a statis-tically significant rise in elastase, lactoferrin and, elastase/neutrophil and lactoferrin/neutrophil ratios were found. Conclusions: Our data show that CKD patients under hemo-dialysis present higher elastase levels (particularly in non-responding patients), which could be related to the rise in neutrophils, and to be part of the enhanced inflammatory process found in these patients.

Copyright © 2008 S. Karger AG, Basel

Key Words rhEPO therapy, resistance � Neutrophil activation � Elastase � Lactoferrin

Abstract Aim: The aim of this work was to evaluate the neutrophil activation state in chronic kidney disease (CKD) patients un-der hemodialysis, and its linkage with resistance to recombi-nant human erythropoietin (rhEPO) therapy. Methods: We studied 63 CKD patients under hemodialysis and rhEPO treatment (32 responders and 31 non-responders to rhEPO therapy). In 20 of the CKD patients (10 responders and 10 non-responders to rhEPO therapy), blood samples were also collected immediately after dialysis. Twenty-six healthy vol-unteers were included in a control group. Hemoglobin lev-els, total and differential leukocyte counts, and circulating levels of C-reactive protein (CRP), elastase and lactoferrin were measured in all patients and controls. Results: Com-

Received: April 4, 2008 Accepted: May 19, 2008 Published online: June 30, 2008

NephrologyAmerican Journal of

Elísio Costa Serviço de Bioquímica, Faculdade de Farmácia da Universidade do Porto Rua Aníbal Cunha, 164PT–4099-030 Porto (Portugal) Tel. +351 222 003 977, Fax +351 222 078 900, E-Mail [email protected]

© 2008 S. Karger AG, Basel0250–8095/08/0286–0935$24.50/0

Accessible online at:www.karger.com/ajn



Costa et al.

Am J Nephrol 2008;28:935–940936

Introduction

Despite the technologic developments in hemodialysis procedures and medical support over the last years, the mortality and morbidity of chronic kidney disease (CKD) patients under hemodialysis remains high, about 10–20 times higher than that found in the general population. Anemia is the most frequent complication associated with hemodialysis [1–3] .

The management of anemia was improved by the in-troduction of recombinant human erythropoietin (rhE-PO) therapy, allowing a significant correction of anemia. However, there is marked variability in the sensitivity to rhEPO, with an up to 10-fold variability in dose require-ments to achieve correction of anemia, and around 5–10% of the patients showing a marked resistance to rhE-PO therapy [4–6] .

The interaction of blood with non-biological materials of the extracorporeal circuit during hemodialysis leads to the activation of several non-cellular and cellular sys-tems, including polymorphonuclear leukocytes (PMN) [7, 8] . The activation of PMN may lead to degranulation, with release of several components, namely, cationic pro-teins and proteases and the production of oxygen me-tabolites. Reports on this subject [7, 9, 10] , though in a limited number, have shown that neutrophils are acti-vated during the hemodialysis procedure, but none of those reports studied the relationship between resistance to rhEPO therapy and neutrophil activation markers. Ac-tually, it has been reported that rhEPO resistance seems to be an associated inflammatory feature. Moreover, none of those works calculated the ratios of neutrophil activation products per number of neutrophils, a more accurate measure of leukocyte activation.

The aim of this work was to evaluate the neutrophil activation state in CKD patients under hemodialysis, and its linkage with resistance to rhEPO therapy by measur-ing circulating levels of elastase and lactoferrin. These two substances are contained within primary and sec-ondary neutrophil granules, respectively, and are fre-quently used as indirect markers of neutrophil activation in vivo.

Subjects and Methods

We performed a cross-sectional study by evaluating 63 CKD patients under hemodialysis (36 males, 27 females; mean age 8 SD 62.1 8 15.7 years) and rhEPO treatment. The CKD patients included 32 responders and 31 non-responders to rhEPO therapy. Classification of the patients as responders or non-responders was

performed in accordance with the European Best Practice Guide-lines [11] which defines resistance to rhEPO as failure to achieve target hemoglobin levels (between 11 and 12 g/dl) with mainte-nance doses of rhEPO of 1 300 IU/kg/week of epoetin or 1.5 � g/kg/week of darbepoietin- � . The rhEPO maintenance dose for re-sponding patients was 89.65 8 57.62 U/kg/week and for non-re-sponders 572.99 8 193.84 U/kg/week. The 2 groups of patients were matched for age (responders 62.3 8 16.8 years vs. non-re-sponders 62.9 8 15.1 years, p 1 0.05), gender (responders 56.25% males vs. non-responders 58.06% males, p 1 0.05), weight (re-sponders 63.9 8 11.7 kg vs. non-responders 58.6 8 13.1 kg, p 1 0.05), body mass index (responders 23.7 8 3.4 vs. non-responders 21.8 8 4.7, p 1 0.05), mean time on hemodialysis (responders 68.0 8 65.7 months vs. non-responders 62.88 8 51.2 months, p 1 0.05), urea reduction ratio (responders 25.58 8 7.71% vs. non-re-sponders 27.8 8 8.6%, p 1 0.05), urea Ktv (responders 1.5 8 0.3 vs. non-responders 1.63 8 0.3, p 1 0.05) and parathyroid hor-mone serum levels (responders 294.5 8 282.9 pg/ml vs. non-re-sponders 282.2 8 244.7 pg/ml, p 1 0.05).

Twenty of the CKD patients (10 responders and 10 non-re-sponders to rhEPO therapy) were also evaluated before and im-mediately after hemodialysis.

CKD patients were under therapeutic hemodialysis three times per week, for 3–5 h, for a median of 36 months. All patients used the high-flux polysulfone FX-class dialyzers of Fresenius: 34 with FX60; 27 with FX80, and 2 with FX100 dialyzer type. The causes of renal failure in the patient population were as follows: diabetic nephropathy (n = 19); chronic glomerulonephritis (n = 11); polycystic kidney disease (n = 3); hypertensive nephrosclero-sis (n = 3); obstructive nephropathy (n = 3); pyelonephritis associ-ated with neurogenic bladder (n = 1); nephrolithiasis (n = 1); chronic interstitial nephritis (n = 1); Alport syndrome (n = 1); re-nal vascular disease due to polyarteritis (n = 1), and chronic renal failure of uncertain etiology (n = 19). Patients with autoimmune diseases, malignancy, hematological disorders, and acute or chronic infection were excluded. Intravenous iron supplementa-tion was based on the European Best Practice Guidelines for the management of anemia in patients with hemodialysis [11] . All pa-tients gave their informed consent to participate in this study.

Healthy volunteer were selected as controls based on normal hematological and biochemical values, and no history of kidney or inflammatory disease. They were also matched as far as pos-sible for age and gender with the CKD patients (8 males, 17 fe-males; mean 8 SD age 47.81 8 14.69 years).

Blood Sample Assays Blood samples were collected with and without EDTA as an-

ticoagulant in order to obtain whole blood, serum and plasma. In the cross-sectional study, blood samples were collected from CKD patients before starting hemodialysis. To evaluate the effect of hemodialysis in the parameters studied, blood samples were also collected immediately after hemodialysis.

The hemoglobin concentration and white blood cell count were measured using an automatic counter (Sysmex K1000, Ham-burg, Germany) and leukocyte differential counts were evaluated in Wright-stained blood films. Plasma levels of elastase and lac-toferrin were evaluated by enzyme immunoassays (human PMN Elastase ELISA, Bender MedSystems; Lactoferrin ELISA Kit, Cal-biochem, respectively). Serum C-reactive protein (CRP) was de-termined by immunoturbidimetry (CRP latex HS Roche kit,



Neutrophil Activation and Resistance to rhEPO Therapy

Am J Nephrol 2008;28:935–940 937

Roche Diagnostics). Serum albumin levels were measured using a colorimetric assay end-point method (Albumin Plus; Roche GmbH, Mannheim, Germany). The serum iron concentration was determined using a colorimetric method (Randox Laborato-ries Ltd., Northern Ireland, UK), whereas serum ferritin (Randox Laboratories Ltd.) and serum transferrin (Randox Laboratories Ltd.) were measured by immunoturbidimetry. Transferrin satu-ration (TS) was calculated using the formula: TS (%) = 70.9 ! serum iron concentration ( � g/dl)/serum transferrin concentra-tion (mg/dl).

Data Analysis For statistical analysis, we used the Statistical Package for So-

cial Sciences (SPSS version 15.0 for Windows, SPSS Inc., Chicago, Ill., USA). Kolmogorov Smirnov statistics were used to evaluate sample normality distribution. Multiple comparisons between groups were performed by one-way ANOVA supplemented with Turkey’s HSD post hoc test. For single comparisons, we used the Student’s t test whenever the parameters presented a Gaussian distribution and the Mann-Whitney U test in the case of a non-Gaussian distribution. To compare data before and after hemodi-alysis, we used paired-samples t test or Wilcoxon test. Spearman’s rank correlation coefficient was used to evaluate relationships be-tween sets of data. Significance was accepted at p ! 0.05.

Results

The hematological characteristics, iron status and the inflammatory markers of the 3 groups studied (cross-sectional study) are summarized in table 1 . Hemodialysis CKD patients were anemic, with non-responding pa-tients being more anemic than responders. No difference was found between controls and patients concerning the white blood cell count. However, CKD patients showed

lymphopenia, as well as higher monocyte and neutrophil counts as compared to controls. When comparing the 2 groups of patients, non-responders presented a signifi-cantly lower lymphocyte count than responders.

Compared with controls, CKD patients presented sig-nificantly higher ferritin, CRP and elastase plasma levels ( table 2 ). However, no statistically significant differences were found for the elastase/neutrophil ratio. No statisti-cally significant differences were also found between CKD patients and controls for plasma levels of lactoferrin and lactoferrin/neutrophil ratio. When we compared the 2 groups of patients, we found that non-responders showed decreased serum levels of albumin, and higher CRP and elastase plasma levels. Statistically significantly positive correlations were found between elastase plasma levels and weekly rhEPO dose and CRP ( fig. 1 ).

Analyzing the results for hematological data, and for neutrophil activation markers, before and immediately after the hemodialysis procedure ( table 2 ), we found a sta-tistically significant rise in hemoglobin, elastase, lacto-ferrin and, elastase/neutrophil and lactoferrin/neutro-phil ratios. No statistically significant changes were found between responders and non-responders for any of the parameters studied.

Discussion

As far as we know, we are the first group to assess the association between neutrophil activation and resistance to rhEPO therapy in CKD patients under hemodialysis.

Table 1. Hematological data, iron status and neutrophil activation markers for controls and CKD patients

Controls (n = 26) All patients (n = 63) Responders (n = 32) Non-responders (n = 31)

Hb, g/dl 13.90 (13.2–15.00) 10.90 (10.30–12.30)a 11.70 (10.83–12.68)a 10.4 (9.00–11.30)a, b

White cell counts, !109/l 5.7881.59 6.2382.10 6.4281.96 6.0482.26Lymphocytes, !109/l 2.3580.75 1.4780.60a 1.5880.49a 1.3680.69a, b

Monocytes, !109/l 0.2580.08 0.3880.16a 0.4080.13a 0.3580.17a

Neutrophils, !109/l 3.0381.02 4.1481.79a 4.1781.87a 4.1181.73a

Ferritin, ng/ml 85.10 (40.90–123.00) 383.80 (188.20–546.25)b 399.00 (294.68–545.25)b 342.00 (163.00–580.05)b

Tranferrin saturation, % 19.44 (17.28–27.94) 21.80 (17.70–30.91) 24.20 (19.96–33.29) 22.10 (17.79–29.44)Albumin, g/dl NM 3.880.4 4.080.4 3.780.4b

CRP, mg/dl 1.75 (0.76–4.70) 5.75 (1.90–14.01)a 3.20 (1.73–7.23)a 10.14 (3.82–38.99)a, b

Elastase, �g/l 28.29 (26.03–34.74) 36.11 (29.69–50.65)a 34.13 (28.76–39.16)a 39.75 (31.15–64.84)a, b

Elastase/neutrophil ratio 10.86 (7.44–12.12) 8.91 (7.43–13.78) 8.70 (7.32–11.42) 10.25 (7.56–17.41)Lactoferrin, �g/l 236.56 (193.56–295.03) 239.35 (165.64–332.60) 239.60 (170.90–332.70) 228.90 (160.33–341.50)Lactoferrin/neutrophil ratio 72.11 (55.52–111.83) 60.32 (42.82–99.45) 64.55 (44.31–101.49) 58.78 (41.50–102.99)

NM = Not made. Results are presented as mean 8 standard deviation and as median (interquartile ranges).a p < 0.05 vs. controls; b p < 0.05 vs. responders.



Costa et al.

Am J Nephrol 2008;28:935–940938

We were able to demonstrate that elastase, a neutrophil protease released by degranulation, is significantly high-er in the plasma of CKD patients under hemodialysis, be-ing particularly high in non-responding CKD patients. This is probably part of the enhanced inflammatory sta-tus reported for these CKD patients.

In the cross-sectional study, performed before hemo-dialysis, we assessed chronic inflammation associated with CKD and with resistance to rhEPO therapy. We found that CKD patients under hemodialysis treatment show a significant rise in neutrophils and increased CRP levels in serum, confirming the presence of an inflamma-

tory process in these patients. The fact that non-respond-ers had higher CRP levels suggests that inflammation is related to resistance to rhEPO therapy, in agreement with previous studies [12, 13] . The lymphopenia found in he-modialysis patients is particularly enhanced in non-re-sponders. This feature has already been described in the literature [13] and results, at least in part, from a decrease in total circulating CD3+ T lymphocytes and affects both the CD4+ and the CD8+ T-cell subsets.

We used lactoferrin and elastase plasma levels as neu-trophil activation markers. Elastase levels were signifi-cantly higher in hemodialysis patients, particularly in

Hemodialysis patients (n = 20)

before after

Hb, g/dl 12.10 (10.95–12.80) 13.20 (11.15–14.60)a

White cell counts, !109/l 5.8681.51 5.9382.19Neutrophils, !109/l 3.8281.24 3.9781.77Monocytes, !109/l 0.2480.38 0.1780.12Lymphocytes, !109/l 1.6480.69 1.6680.64Elastase, �g/l 36.16 (29.71–47.13) 51.69 (40.08–71.68)a

Elastase/neutrophil ratio 10.66 (7.32–13.54) 14.66 (13.34–18.95)a

Lactoferrin, �g/l 198.61 (137.81–216.97) 236.56 (171.28–363.63)a

Lactoferrin/neutrophil ratio 48.33 (33.88–64.31) 60.72 (51.81–94.81)a

CRP, mg/dl 3.06 (1.39–5.22) 3.53 (1.54–5.56)

Results are presented as mean 8 standard deviation and as median (interquartile ranges). a p < 0.05 vs. before hemodialysis.

Table 2. Hematological data and neutrophil activation markers forCKD patients, before and after thehemodialysis procedure

0

50

100

150E

last

ase

(µg

/l)

0 200 400 600 800 1,000 1,200

Weekly rhEPO (IU/kg dose)a

r = 0.257; p = 0.047

0

50

100

150

Ela

sta

se(µ

g/l

)

0 10 20 30 40 50

CRP (mg/dl)b

r = 0.359; p = 0.014

Fig. 1. Correlation between plasma elastase levels and weekly rhEPO dose ( a ) and CRP ( b ) in hemodialysis patients.



Neutrophil Activation and Resistance to rhEPO Therapy

Am J Nephrol 2008;28:935–940 939

non-responding patients. However, no statistical differ-ences were found for the elastase/neutrophil ratio be-tween controls and patient groups, suggesting that the higher elastase levels are mainly associated with the in-creased neutrophil count found in these patients. Even so, it is important to highlight that the elastase/neutrophil ratio presented a trend to higher values in non-respond-ers. Moreover, we found statistically significant positive correlations between elastase levels and CRP, suggesting that the rise in elastase levels is part of the inflammatory process found in hemodialysis patients and particularly enhanced in non-responders. The statistically significant correlation between elastase levels and weekly rhEPO doses also corroborate this hypothesis; in fact, non-re-sponders to rhEPO therapy who require higher weekly rhEPO doses to achieve target hemoglobin levels are also associated with an increased inflammatory process.

On the other hand, plasma levels of lactoferrin and the lactoferrin/neutrophil ratio did not differ between groups. A plausible justification for this result is that, although the major circulating lactoferrin is neutrophil-derived, it may also be produced by other cells [15] . Contrary to the evaluation of elastase, the ELISA kit used is not specific for the evaluation of PMN leukocyte lactoferrin levels. Thus, in our study, elastase is a more specific marker of neutrophil activation. Moreover, we can also consider the hypothesis that lactoferrin is being used or cleared from the circulation in a pronounced way. Indeed, lactoferrin is involved in iron metabolism, which is known to be al-tered in these patients [16] .

By evaluating CKD patients before and after hemodi-alysis, we assessed whether our patients respond in same manner to the hemodialysis process. After hemodialysis, a higher hemoglobin level was found, and has been at-

tributed to a translocation of erythrocytes from the splanchnic circulation (and possibly from the spleen) in order to compensate for the hypovolemic stress during dialysis ultrafiltration [17, 18] . After hemodialysis we also found a statistically significant increase in both elastase and lactoferrin concentrations (and in their ratios per neutrophil), suggesting a neutrophilic activation process with degranulation during the hemodialysis procedure. We also found that neutrophils from non-responding pa-tients to rhEPO therapy do not respond any differently than those found in responding patients. It is likely that the higher elastase levels found in non-responders (cross-sectional study) is likely to result from a chronic but not acute stimulus of the hemodialysis process.

In conclusion, our data show that CKD patients under hemodialysis present higher elastase levels (particularly in non-responding patients), which could be related to the rise in neutrophils and could be part of the enhanced inflammatory process found in these patients. Moreover, neutrophil activation is triggered by the hemodialysis procedure, but does not explain the higher elastase levels found in non-responding patients. Elastase, but not lac-toferrin, may prove to be a good marker of resistance to rhEPO therapy in CKD patients under hemodialysis.

Acknowledgements

We are very grateful to the nurses of FMC, Dinefro – Diálises e Nefrologia, SA, and Uninefro – Sociedade Prestadora de Cuida-dos Médicos e de Diálise, SA, for technical support. We are also very grateful to Amgen and to Roche Farmacêutica Química, Lda for financial support. This study was supported by a PhD grant (SFRH/BD/27688/2006) to E. Costa from FCT and FSE.

References

1 Foley RN, Parfrey PS, Sarnak MJ: Clinical epidemiology of cardiovascular disease in chronic renal failure. Am J Kidney Dis 1998; 32(suppl 3):S112–S119.

2 Foley RN, Parfrey PS, Harnett JD, Kent GM, Murray DC, Barre PE: The impact of anemia on cardiomyopathy morbidity and mortality in end-stage renal disease. Am J Kidney Dis 1996; 28: 53–61.

3 Locatelli F, Conte F, Marcelli D: The impact of hematocrit levels and erythropoietin treatment on overall and cardiovascular mortality and morbidity – the experience of the Lombardy dialysis registry. Nephrol Dial Transplant 1998; 13: 1642–1644.

4 Spittle MA, Hoenich NA, Handelman GJ, Adhikarla R, Homel P, Levin NW: Oxidative stress and inflammation in hemodialysis pa-tients. Am J Kidney Dis 2001; 38: 1408–1413.

5 Pupim LB, Himmelfarb J, McMonagle E, Shyr Y, Ikizler TA: Influence of initiation of maintenance hemodialysis on biomarkers of inflammation and oxidative stress. Kidney Int 2004; 65: 2371–2379.

6 Sommerburg O, Grune T, Hampl H, Riedel E, van Kuijk FJ, Ehrich JHH, Siems WG: Does long-term treatment of renal anemia with recombinant erythropoietin influence oxidative stress in haemodialysed patients? Nephrol Dial Transplant 1998; 13: 2583–2587.

7 Lucchi L, Bergamini S, Botti B, Rapanà R, Ciuffreda A, Ruggiero P, Ballestri M, Tomasi A, Albertazzi A: Influence of different he-modialysis membranes on red blood cell sus-ceptibility to oxidative stress. Artif Organs 2000; 24: 1–6.

8 Valentini J, Schmitt GC, Grotto D, Santa Ma-ria LD, Boeira SP, Piva SJ, Brucker N, Bohrer D, Pomblum VJ, Emanuelli T, Garcia SC: Human erythrocyte delta-aminolevulinate dehydratase activity and oxidative stress in hemodialysis patients. Clin Biochem 2007; 40: 591–594.



Costa et al.

Am J Nephrol 2008;28:935–940940

9 Skoutelis AT, Kaleridis VE, Goumenos DS, Athanassiou GM, Missirlis YF, Vlachojannis JG, Bassaris HP: Polymorphonuclear leuko-cyte rigidity is defective in patients with chronic renal failure. Nephrol Dial Trans-plant 2000; 15: 1788–1793.

10 Leitienne P, Fouque D, Rigal D, Adeleine P, Trzeciak MC, Laville M: Heparins and blood polymorphonuclear stimulation in haemo-dialysis: an expansion of the biocompatibil-ity concept. Nephrol Dial Transplant 2000; 15: 1631–1637.

11 Locatelli F, Aljama P, Barany P, Canaud B, Carrera F, Eckardt KU, Horl WH, Macdou-gal IC, Macleod A, Wiecek A, Cameron S; European Best Practice Guidelines Working Group: Revised European best practice guidelines for the management of anemia in patients with chronic renal failure. Nephrol Dial Transplant 2004; 19(suppl 2):ii1–ii47.

12 Bárány P: Inflammation, serum C-reactive protein, and erythropoietin resistance. Nephrol Dial Transplant 2001; 16: 224–227.

13 Costa E, Lima M, Alves JM, Rocha S, Rocha-Pereira P, Castro E, Miranda V, Sameiro-Faria M, Loureiro A, Quintanilha A, Belo L, Santos-Silva A: Inflammation, T-cell pheno-type and inflammatory cytokines in chronic kidney disease patients under haemodialysis and its relationship to resistance to recombi-nant human erythropoietin therapy. J Clin Immunol 2008; 28: 268–275.

14 Belo L, Santos-Silva A, Caslake M, Cooney J, Pereira-Leite L, Quintanilha A, Rebelo I: Neutrophil activation and C-reactive pro-tein concentration in preeclampsia. Hyper-tens Pregnancy 2003; 22: 129–141.

15 Levay PF, Viljoen M: Lactoferrin: a general review. Haematologica 1995; 80: 252–267.

16 Costa E, Pereira BJG, Rocha-Pereira P, Rocha S, Reis F, Castro E, Teixeira F, Miranda V, Sameiro Faria M, Loureiro A, Quintanilha A, Belo L, Santos-Silva A: Role of prohepci-din, inflammatory markers and iron status in resistance to rhEPO therapy in hemodi-alysis patients. Am J Nephrol 2008; 28: 677–683.

17 Dasselaar JJ, Hooge MNL, Pruim J, Nijnuis H, Wiersum A, Jong PE, Huisman RM, Franssen CFM: Relative blood volume changes underestimated total blood volume changes during hemodialysis. Clin J Am Soc Nephrol 2007; 2: 669–674.

18 Yu AW, Nawab ZM, Barnes WE, Lai KN, Ing TS, Daugirdas JT: Splanchnic erythrocyte content decreases during hemodialysis: a new compensatory mechanism for hypovo-lemia. Kidney Int 1997; 51: 1986–1990.



Paper IV

“Role of Prohepcidin, Inflammatory Markers and Iron

Status in Resistance to rhEPO Therapy in Hemodialysis Pa-

tients”

Am J Nephrol 2008; 28: 677-683.




Original Report: Patient-Oriented, Translational Research

Am J Nephrol 2008;28:677–683 DOI: 10.1159/000121478

Role of Prohepcidin, Inflammatory Markers and Iron Status in Resistance to rhEPO Therapy in Hemodialysis Patients

Elísio Costa a, d, e Brian J.G. Pereira i Petronila Rocha-Pereira b Susana Rocha d, e

Flávio Reis c Elisabeth Castro d, e Frederico Teixeira c Vasco Miranda h

Maria do Sameiro Faria h Alfredo Loureiro g Alexandre Quintanilha e, f

Luís Belo d, e Alice Santos-Silva d, e

a Escola Superior de Saúde, Instituto Politécnico de Bragança, Bragança , b Universidade da Beira Interior, Covilhã , c Instituto de Farmacologia e Terapêutica Experimental, Universidade de Coimbra, Coimbra , d Faculdade de Farmácia, Serviço de Bioquímica, e Instituto de Biologia Molecular e Celular e f Instituto de Ciências Biomédicas Abel Salazar, Universidade do Porto, g Uninefro – Sociedade Prestadora de Cuidados Médicos e de Diálise, SA, e h FMC, Dinefro – Diálises e Nefrologia, SA, Porto , Portugal; i AMAG Pharmaceuticals, Inc., Cambridge, Mass. , USA

Prohepcidin levels correlated negatively with s-TfR and re-ticulocyte count. The weekly rhEPO/kg dose was found to be positively correlated with CRP, hemoglobin and s-TfR. In con-clusion, our data show that a close interaction exists be-tween inflammation, iron status and prohepcidin serum lev-els that ultimately regulate intracellular iron availability. Prohepcidin and s-TfR, together with CRP, may prove to be good markers of resistance to rhEPO therapy in HD pa-tients. Copyright © 2008 S. Karger AG, Basel

Introduction

Despite significant advances in dialysis technology and the development of innovative therapies to prevent or attenuate the complications of hemodialysis (HD), mortality and morbidity of patients on dialysis remain high, about 10–20 times higher than that found in the general population [1] . Anemia is among the most fre-quent complications associated with HD, and can lead to a reduction in tissue oxygenation, increase in cardiac out-

Key Words Erythropoietin � Hepcidin � Iron � Iron metabolism � Prohepcidin � Resistance � rhEPO

Abstract The aim of our study was to assess possible relations be-tween prohepcidin, iron status and inflammatory markers in hemodialysis (HD) patients, as well as its association with re-sistance to recombinant human erythropoietin (rhEPO) ther-apy. Fifty HD patients and 25 healthy controls were enrolled in the study. Among HD patients, 25 were non-responders and 25 were responders to rhEPO therapy. Complete blood cell count, reticulocyte count, and circulating levels of ferri-tin, iron, transferrin saturation, C-reactive protein (CRP), sol-uble interleukin (IL)-2 receptor (s-IL2R), soluble transferrin receptor (s-TfR), IL-6 and prohepcidin were measured in all patients and controls. HD patients showed higher circulat-ing levels of ferritin, s-TfR, CRP, IL-6, s-IL2R and prohepcidin, and lower levels of transferrin compared to healthy controls. Higher levels of s-TfR, CRP and lower levels prohepcidin were observed among non-responders compared to responders.

Received: October 17, 2007 Accepted: February 14, 2008 Published online: March 20, 2008

NephrologyAmerican Journal of

Elísio Costa Serviço de Bioquímica, Faculdade de Farmácia Universidade do Porto, Rua Aníbal Cunha, 164 PT–4099-030 Porto (Portugal) Tel. +35 1222 003 977, Fax +35 1918 653 244, E-Mail [email protected]

© 2008 S. Karger AG, Basel0250–8095/08/0284–0677$24.50/0

Accessible online at:www.karger.com/ajn



Costa et al.

Am J Nephrol 2008;28:677–683678

put, left ventricular hypertrophy, congestive heart dis-ease, fatigue, reduction in exercise capacity and immuno-deficiency [2, 3] .

The introduction of recombinant human erythropoi-etin (rhEPO) therapy in the early 1990s [4–6] has led to a significant reduction in anemia and improved patients’ quality of life. However, there is marked variability in the sensitivity to rhEPO, with up to 10-fold variability in dose requirements to achieve correction of anemia [4–10] . Ap-proximately 5–10% of the patients show a marked resis-tance to rhEPO therapy [4–6, 9–17] . This variability how-ever remains relatively unexplained. rhEPO resistance is associated with several conditions including iron defi-ciency, oxidative stress, infection, inflammation, blood loss, hyperparathyroidism, aluminum toxicity and vita-min B 12 and folate deficiencies [9–17] . However, exclusion of these factors does not eliminate the marked variability in the sensitivity to rhEPO therapy [9–11] .

Alteration in iron hemostasis is one of the most stud-ied causes of resistance to rhEPO therapy. Iron is an es-sential trace element that is required for growth and de-velopment of living organisms, but excess of free iron is toxic for the cell [18–20] . Mammals lack a regulatory pathway for iron excretion, and iron balance is main-tained by the tight regulation of iron absorption from the intestine [20, 21] . Intestinal iron absorption is regulated by the level of body iron stores and to the amount of iron needed for erythropoiesis [18–22] . In HD patients, iron absorption is similar to that found in healthy individuals, however, during rhEPO therapy the absorption of iron increases as much as 5 times [23, 24] . This increased iron absorption is not sufficient to compensate for the iron loss associated with the HD procedure, and frequent blood sampling and procedures performed on these pa-tients. For this reason, intravenous iron administration has become the standard therapy for most patients re-ceiving rhEPO therapy. To avoid iron overload, with po-tentially harmful consequences, there is a need to moni-tor iron therapy by performing regular blood tests that reflect the body’s iron stores.

Recently, a complex regulatory network that governs iron traffic has emerged, pointing to hepcidin as a major evolutionary conserved regulator of iron distribution [18–22, 25–30] . This small hormone produced by the mammalian liver has been proposed as a central media-tor of dietary iron absorption. Hepcidin was found to be associated with decreases in both iron uptake from the small intestine and release of iron from macrophages, as well as decreased placental iron transport [28, 29] . The synthesis of hepcidin is stimulated by inflammation and

iron overload. It is synthesized as a preprohepcidin of 84 amino acids. The signal peptide is cleaved, leading to the 60-amino acid prohepcidin, which is further processed giving rise to the 25-amino acid hepcidin [30, 31] .

Hepcidin synthesis is regulated by inflammation, a common finding in HD patients, being enhanced in non-responders compared with responders. Development or worsening of anemia is also common in rhEPO resis-tance due to higher doses of rhEPO and intravenous iron supplementation. Moreover, increased levels of prohep-cidin have been reported in HD patients [27–34] , and the relationship between prohepcidin serum levels and rhEPO resistance in HD patients has not been adequate-ly defined.

We sought, therefore, to study the relationship be-tween prohepcidin (a hepcidin precursor molecule), iron status and inflammatory markers in HD patients, as well as its association with resistance to rhEPO therapy.

Patients and Methods

Subjects Fifty HD patients and 25 healthy controls were enrolled in the

study. The HD patients included 25 rhEPO responders (16 males and 9 females; mean age 63.32 8 17.05 years) and 25 non-re-sponders (16 males and 9 females; mean age 62.90 8 15.08 years). Classification of HD patients as responders or non-responders was performed in accordance with the European Best Practice Guidelines [34] , which defines resistance to rhEPO as a failure to achieve target hemoglobin (Hb) levels with doses of rhEPO 1 300 IU/kg/week of epoetin or 1.5 � g/kg/week of darbepoetin- � . Healthy volunteers were selected based on normal hematological and biochemical values, and no history of kidney or inflamma-tory disease. They were also matched (as far as possible) for age and gender with HD patients (8 males and 17 females; mean age 47.81 8 14.69 years). The two groups of patients were also matched for age, gender, weight, body mass index, mean time on HD, urea reduction ratio, urea Ktv and serum levels of parathyroid hor-mone. Patients with autoimmune disease, malignancy, hemato-logical disorders and acute or chronic infection were excluded from the study. All subjects gave their informed consent to par-ticipate in this study. Intravenous iron supplementation was based on the European Best Practice Guidelines for the Management of Anemia in patients with HD [34] .

Assays Blood samples were collected with and without anticoagulant

ethylenediamine tetraacetic acid, in order to obtain whole blood, plasma and serum. Blood sample collection was performed in all patients immediately before the start of HD. Red blood cell (RBC) count, hematocrit, Hb concentration, hematological indices [mean cell volume (MCV), mean cell Hb (MCH) and mean cell Hb concentration (MCHC)] and red cell distribution width (RDW) were measured using an automatic blood cell counter (Sysmex K1000; Sysmex, Hamburg, Germany). Reticulocyte



Prohepcidin and Resistance to rhEPO Therapy

Am J Nephrol 2008;28:677–683 679

count was measured by microscopic counting on blood smears after vital staining with new methylene blue (reticulocyte stain; Sigma, St. Louis, Mo., USA). The reticulocyte production index was calculated to measure the effective RBC production [35] .

Serum iron concentration was determined using a colorimet-ric method (iron, Randox Laboratories, Crumlin, UK), whereas serum ferritin and serum transferrin were measured by immu-noturbidimetry (ferritin and transferrin, Randox Laborato-ries). Transferrin saturation (TS) was calculated by the formula: TS (%) = 70.9 ! serum iron concentration (in � g/dl)/serum transferrin concentration (in mg/dl). Enzyme-linked immuno-sorbent assays were used to measure plasma soluble transferrin receptor (s-TfR; human sTfR immunoassay, R&D Systems, Min-neapolis, Minn., USA), soluble interleukin (IL)-2 receptor (s-IL2R; human IL-2 SR � , R&D Systems) and serum prohepcidin concentrations (Hepcidin Prohormone ELISA, IBL, Hamburg, Germany). Serum C-reactive protein (CRP) was determined by immunoturbidimetry (CRP latex HS Roche kit, Roche Diagnos-tics, Basel, Switzerland). Serum IL-6 levels were quantified using the BD Cytometric Bead Array (CBA) Human Th1/Th2 Cytokine Kit II (BD Biosciences, San Diego, Calif., USA,) and analyzed us-ing the BD CBA Software.

Statistical Analysis For statistical analysis, the Statistical Package for Social Sci-

ences, version 14.0, was used. Kolmogorov-Smirnov statistics were used to evaluate normal sample distribution. Measurements are presented as means 8 SD for parametric values or as medians (interquartile range) for non-parametric values. Multiple com-parisons between groups were performed by one-way ANOVA supplemented with Tukey’s HSD post hoc test. Whenever the pa-rameters presented a non-Gaussian distribution, the Kruskal-Wallis and the Mann-Whitney U test were performed. Spear-man’s rank correlation coefficient was used to evaluate relation-ships between sets of data. Significance was accepted at p ! 0.05. For correlations, 1 � g of darbepoetin- � was considered equiva-lent to 200 IU of epoetin.

Results

HD patients were dialyzed three times per week for 3–5 h, for a median of 36 months. All patients used the high-flux polysulfone FX-class dialyzers (Fresenius Med-ical Care, Bad Homburg, Germany). The causes of renal failure were diabetic nephropathy in 16, chronic glomer-ulonephritis in 6, polycystic kidney disease in 5, hyper-tensive nephrosclerosis in 3, obstructive nephropathy in 3, pyelonephritis associated with neurogenic bladder, nephrolithiasis, chronic interstitial nephritis, Alport’s syndrome and polyarteritis in one each, and uncertain in 12 patients.

Tables 1 and 2 show hematological and biochemical data, iron status and inflammatory markers for controls and HD patients. Data are broken down based on the patient’s status as responders or non-responders. The rhEPO maintenance dose was 7.51 8 6.52 U/g/week/Hb (84.10 8 57.72 U/kg/week) for responders and 59.40 8 23.31 U/kg/week/Hb (590.92 8 189.63 U/kg/week) for non-responders. No statistically significant difference was found in total iron supplementation given during the last year between responder and non-responder patients (data not shown).

Compared to controls, HD patients presented with a significantly lower RBC count, hemoglobin and hemato-crit, and no difference in RBC indices. HD patients also had a significantly higher RDW, reticulocyte count and reticulocyte production index ( table 1 ). Among HD pa-tients, non-responders had lower Hb, hematocrit, MCH and MCHC, and a higher RDW, despite higher doses of rhEPO.

Table 1. Hematological indices among healthy controls and HD patients

Healthy controls(n = 25)

All HD patients(n = 50)

rhEPO responders(n = 25)

rhEPO non-responders(n = 25)

Hb, g/dl 14.1281.27 11.0681.77b 11.8581.50b 10.2681.69b, c

Hematocrit, % 43.5084.28 33.9385.01b 35.8284.10b 32.0485.18b, c

RBC, !1012/l 4.7280.59 3.6980.56b 3.8180.45b 3.5880.65b

MCV, fl 92.00 (90.00–94.00) 93.75 (89.00–99.28) 95.50 (92.55–98.40)a 92.30 (85.2–100.8)MCH, pg 29.80 (28.90–30.50) 30.35 (27.65–32.22) 31.50 (29.90–32.50)a 28.90 (25.70–32.10)c

MCHC, g/dl 32.4780.58 31.7582.55 33.0981.97 30.4182.38b, d

RDW, % 12.70 (12.40–13.20) 16.05 (14.53–17.58)b 14.60 (13.65–15.30)b 17.40 (16.40–18.40)b, d

Reticulocytes, !109/l 33.57822.78 60.85831.44b 54.94830.40a 66.75831.95b

RPI 0.4680.31 1.0480.55b 1.1580.62a 0.9380.47b

a p < 0.05, b p < 0.0001, vs. controls; c p < 0.05, d p < 0.0001, vs. responders.RPI = Reticulocyte production index. Results are presented as means 8 SD and as medians (interquartile ranges).



Costa et al.

Am J Nephrol 2008;28:677–683680

Compared to healthy controls, HD patients had sig-nificantly higher serum ferritin and s-TfR, and signif-icantly lower serum transferrin levels ( table 2 ). CRP, s-IL2R, IL-6 and prohepcidin levels were significantly higher in HD patients compared to healthy controls ( ta-ble 2 ).

Among HD patients, s-TfR as well as CRP levels were significantly higher in non-responders than responders

( table 2 ; fig. 1 ). Prohepcidin serum levels were signifi-cantly lower in non-responders than responders, but were higher than those in the control group ( table 2 ; fig. 2 ).

When the results are analyzed by gender, we found that female HD patients showed statistically significant-ly lower prohepcidin serum levels compared with male patients (128.41 8 42.68 vs. 179.62 8 88.89 ng/ml, p = 0.009), particularly in non-responding female patients

Table 2. Serum indices of iron status and inflammatory markers among healthy controls and HD patients

Healthy controls(n = 25)

All HD patients(n = 50)



Iron, �g/dl 73.42825.24 55.32826.52a 60.24822.97 50.40829.27a


Transferrin, mg/dl 231.50 (205.00–268.00) 170.50 (143.50–187.00)b 173.00 (152.50–186.00)b 161.00 (139.00–211.00)b

TS, % 21.8387.97 22.89811.06 25.0589.69 20.73812.09s-TfR, nmol/l 20.8588.56 26.84811.87a 19.5686.83 34.13811.42d

Prohepcidin, ng/ml 92.11818.28 151.74843.52b 165.72836.69b 137.77846.03b, c

CRP, mg/dl 1.75 (0.76–4.70) 5.75 (1.90–14.01)b 3.20 (1.73–7.23)a 10.14 (3.82–38.99)b, c

s-IL2R, nmol/l 758.838234.95 4,199.9481,762.81b 4,005.7181,835.70b 4,394.1781,701.80b

IL-6, pg/ml 1.90 (0–3.75) 7.80 (3.85–15.05)b 5.75 (3.83–13.95)b 8.80 (4.55–21.30)b

a p < 0.05, b p < 0.0001, vs. controls; c p < 0.05, d p < 0.0001, vs. responders.Results are presented as means 8 SD and as medians (interquartile ranges).

10

20

30

40

50

60

70

0 200 400 600 800 1,000 1,200

Weekly rhEPO (IU)/kg dose

s-T

fR(n

mo

l/l)

Fig. 1, 2. Distribution of plasma levels of s-TfR ( 1 ) and prohepcidin ( 2 ) in HD patients based on weekly rhEPO dose. Circles correspond to responders and triangles to non-responding HD patients. Uninterrupted and hatched lines correspond to the median and standard deviations, respectively, of s-TfR ( 1 ) and prohepcidin ( 2 ) among healthy controls. 1 A higher proportion of non-responders (96%) had s-TfR levels 1 20.85 nmol/l (mean among healthy controls), compared with responders (32%, p ! 0.05).

50

100

150

200

250

300

0 200 400 600 800 1,000 1,200

Weekly rhEPO (IU)/kg dose

Pro

he

pci

din

(ng

/ml)

1 2




Am J Nephrol 2008;28:677–683 681

( fig. 3 ). No statistically significant differences in iron sta-tus and inflammatory markers were found between male and female HD patients (data not shown).

Prohepcidin correlated negatively with reticulocyte count (r = –0.324, p = 0.023) and s-TfR (r = –0.359, p = 0.010). However, no statistically significant correlation was found between weekly rhEPO/kg dose and prohep-cidin. In HD patients, statistically significant correla-tions were noted between weekly rhEPO/kg dose with CRP (r = 0.400, p = 0.004) and s-TfR (r = 0.588, p ! 0.0001), and between CRP with MCV (r = –0.322, p = 0.023), MCH (r = –0.393, p = 0.005), iron (r = –0.582, p ! 0.0001) and TS (r = –0.456, p = 0.001).

Discussion

Hepcidin, a liver-derived peptide stimulated by in-flammatory cytokines, acts as a negative regulator of in-testinal iron absorption and its release from macrophages; it is believed to represent the molecular link between chronic inflammation, the iron status of the body and anemia. Excessive hepcidin production occurs in patients with inflammatory and infectious diseases, resulting in anemia of inflammation [18–23] . Furthermore, in vitro stimulation of fresh human hepatocytes by proinflam-matory IL-6 showed strong induction of hepcidin mRNA,

indicating that this cytokine may be the mediator of hep-cidin induction in inflammation [36–38] . This increased hepcidin expression during inflammation and infection explains sequestration of iron in the macrophages and inhibition of intestinal iron absorption, the two hall-marks of the anemia of inflammation, which is a normo-cytic or microcytic iron-refractory anemia [27–32] . This decreased availability of iron may be a host defense mech-anism against invading microorganisms.

The biological importance of the precursor molecule of hepcidin, named prohepcidin, in regulating iron me-tabolism has so far remained undetermined. However, there is some evidence that prohepcidin levels are a reli-able indicator of hepcidin levels and activity [30–33] . Due to the lack of available laboratory methods for quantify-ing circulating hepcidin in clinical samples due to the small peptide size, serum prohepcidin levels are used as an indicator of hepcidin levels, which are easily measured with the widely available ELISA kits.

We found that HD patients had higher serum levels of prohepcidin as well as higher levels of markers of chron-ic inflammation, e.g. ferritin, CRP, IL-6 and s-IL2R, and markers of reduced iron mobilization [36–38] , e.g. lower transferrin levels. These results suggest that the high lev-els of prohepcidin found in HD patients could be related to an underlying chronic inflammation. However, no correlations between prohepcidin and these inflamma-tory markers were found, suggesting that other factors than just inflammation may regulate serum prohepcidin levels.

Among HD patients, non-responders had higher lev-els of CRP than responders, but prohepcidin serum levels were lower. A recent report [39] showed that erythropoi-etin downregulates liver hepcidin expression, acting, therefore, as a hepcidin-inhibitory hormone [22, 28] . Since non-responders were treated with much higher doses of rhEPO compared with responders, the lower prohepcidin levels among non-responders could be ex-plained by this mechanism. On a broader note, despite the treatment with rhEPO, HD patients showed higher levels of prohepcidin. This could be explained by the fact that the stimulus of inflammation for prohepcidin syn-thesis in HD patients is stronger than the inhibitory effect of rhEPO. Moreover, the use of large amounts of rhEPO may result in increased iron utilization by the bone mar-row, resulting in depletion of iron stores and ultimately decreased intracellular iron availability, which could also decrease prohepcidin levels. Therefore, an additional ex-planation for the lower prohepcidin levels in rhEPO non-responding patients could be lower intracellular iron

50

100

150

200

250

300

Controls Responders Non-responders

Pro

he

pci

din

(ng

/ml)

Male

Femalep < 0.05

Fig. 3. Serum prohepcidin levels among healthy controls and HD patients according to gender and response to rhEPO therapy.



Costa et al.

Am J Nephrol 2008;28:677–683682

availability (as reflected by the increased s-TfR observed in the non-responders).

Female HD patients showed a lower serum level of prohepcidin compared with male patients. This sex-spe-cific level of prohepcidin in HD patients has not previ-ously been described in the literature. However, in wild-type mice a slight reduction in Hamp1 mRNA levels was documented [40] and also in hemojuvelin-mutant female mice [41] , in which a more accentuated reduction was found. The reason for these sex differences is presently unknown, but may be attributed to hormonal and/or iron status differences between the two patient groups. No differences in iron status were found between male and female HD patients (data not shown), suggesting that the hormonal gender differences could be the reason for low-er prohepcidin serum levels found in female patients.

The negative correlation between prohepcidin serum levels and reticulocyte count could be an indicator that higher prohepcidin levels have an inhibitory effect on erythropoiesis. Among HD patients, the non-responders to rhEPO therapy presented with more severe anemia, as shown by the significantly lower Hb levels compared to responders ( table 1 ). However, no statistically significant differences were found in serum iron status markers be-tween the patient groups, except for s-TfR, which was higher among non-responders. Moreover, in responders s-TfR levels were similar to those in healthy controls. The levels of this soluble receptor could be increased in two clinical settings: increased erythropoietic activity or iron deficiency [42, 43] . In our patients, the positive correla-tion between s-TfR and weekly rhEPO/kg doses suggest-ed that s-TfR was an indicator of the erythropoietic effect of the rhEPO administered and not an indicator of iron body deficiency. However, considering that responders and non-responders presented similar iron stores, the significantly higher levels of s-TfR observed in non-re-sponders also indicate a disturbance in mobilizing the

iron transferrin linked into precursor erythroid cells, ex-plaining also the lower levels of MCH and MCHC ob-served in non-responders. Altogether, these data support a ‘functional iron deficiency’ in non-responding patients. The higher proportion of non-responding HD patients presenting with s-TfR levels higher than the mean of the healthy control group ( fig. 1 ) suggest that this parameter could be used as an indicator of resistance to rhEPO ther-apy in HD patients.

As previously described [9–14] , CRP is higher in HD patients who are non-responders to rhEPO therapy, indi-cating that inflammation is related to resistance to this therapy. Additionally, CRP serum levels were positively correlated with the weekly rhEPO/kg dose, suggesting that CRP is a good predictor of resistance to rhEPO ther-apy in HD patients. The inverse correlations found be-tween CRP and MCV, MCH, serum iron and TS are in agreement with the observations that chronic inflamma-tion leads to iron trapping in macrophages and reduced serum iron levels. This interaction between iron metabo-lism and inflammatory markers leads to a difficulty in distinguishing iron deficiency, which usually responds to iron therapy, from inflammatory iron block.

In conclusion, our data show that a close interaction exists between inflammation, iron status and prohep-cidin serum levels that ultimately regulate intracellu-lar iron availability. Prohepcidin and s-TfR, together with CRP, may prove to be good markers of resistance to rhEPO therapy in HD patients.

Acknowledgments

We are very grateful to the nurses of FMC, Dinefro – Diálises e Nefrologia, SA, and Uninefro – Sociedade Prestadora de Cuida-dos Médicos e de Diálise, SA, for technical support. This study was supported by a PhD grant (SFRH/BD/27688/2006) attributed to E. Costa by FCT and FSE.

References


2 Foley RN, Parfrey PS, Harnett JD, Kent GM, Murray DC, Barre PE: The impact of anemia on cardiomyopathy morbidity and mortality in end-stage renal disease. Am J Kidney Dis 1996; 28: 53–61.


4 Smrzova J, Balla J, Bárány P: Inflammation and resistance to erythropoiesis-stimulating agents – what do we know and what needs to be clarified? Nephrol Dial Transpant 2005; 20(suppl 8):viii2–viii7.

5 Bárány P: Inflammation, serum C-reactive protein, and erythropoietin resistance. Ne-phrol Dial Transplant 2001; 16: 224–227.

6 Locatelli F, Pisoni RL, Combe C, Bommer J, Andreucci VE, Piera L, Greenwood R, Feld-man HI, Port FK, Held PJ: Anemia in haemo-dialysis patients of five European countries: association with morbidity and mortality in the Dialysis Outcomes and Practice Patterns Study (DOPPS). Nephrol Dial Transplant 2004; 19: 121–132.




Am J Nephrol 2008;28:677–683 683

7 Fisher JW: Erythropoietin: physiology and pharmacology update. Exp Biol Med 2003; 228: 1–14.

8 Weiss MJ: New insights into erythropoietin and epoetin alfa: mechanisms of action, tar-get tissues, and clinical applications. Oncol-ogist 2003; 8(suppl 3):18–29.

9 Macdougall IC, Cooper AC: Erythropoietin resistance: the role of inflammation and pro-inflammatory cytokines. Nephrol Dial Transplant 2002; 17(suppl 11):39–43.

10 Schindler R, Senf R, Frei U: Influencing the inflammatory response of hemodialysis pa-tients by cytokine elimination using large-pore membranes. Nephrol Dial Transplant 2002; 17: 17–19.

11 Gunnell J, Yeun JY, Depner TA, Kaysen GA: Acute-phase response predicts erythropoie-tin resistance in hemodialysis and peritoneal dialysis patients. Am J Kidney Dis 1999; 33: 63–72.

12 Cooper AC, Mikhail A, Lethbridge MW, Ke-meny DM, Macdougall IC: Increased expres-sion of erythropoiesis inhibiting cytokines (IFN- � , TNF- � , IL10, and IL-13) by T cells in patients exhibiting a poor response to eryth-ropoietin therapy. J Am Soc Nephrol 2003; 14: 1776–1784.



15 Sommerburg O, Grune T, Hampl H, Riedel E, van Kuijk FJ, Ehrich JHH, Siems WG: Does long-term treatment of renal anemia with recombinant erythropoietin influence oxidative stress in haemodialysed patients? Nephrol Dial Transplant 1998; 13: 2583–2587.

16 Himmelfarb J: Linking oxidative stress and inflammation in kidney disease: which is the chicken and which is the egg? Semin Dial 2004; 17: 449–454.

17 Drueke TB, Eckardt KU: Role of secondary hyperparathyroidism in erythropoietin re-sistance of chronic renal failure patients. Nephrol Dial Transplant 2002; 17(suppl 5): 28–31.

18 Deicher R, Horl WH: New insights into the regulation of iron homeostasis. Eur J Clin In-vest 2006; 36: 301–309.

19 Arth RH: Iron metabolism in end-stage re-nal disease. Semin Dial 1999; 12: 224–230.

20 Atanasiu V, Manolescu B, Stoian I: Hepcidin – central regulator. Eur J Haematol 2006; 78: 1–10.

21 Park CH, Valore AJ, Waring AJ, Ganz T: Hepcidin, a urinary antimicrobial peptide synthesized in the liver. J Biol Chem 2001; 276: 7806–7810.

22 Nemeth E, Valore EV, Territo M, Schiller G, Lichtenstein A, Ganz T: Hepcidin, a putative mediator of anemia of inflammation, is type II acute-phase protein. Blood 2003; 101: 2461–2463.

23 Skikne B, Ahluwalia N, Fergusson B, Chonko A, Cook J: Effects of recombinant human erythropoietin on iron absorption in chron-ic renal failure (abstract). Blood 1998; 92: 24B.

24 Skikne BS, Cook JD: Effect of enhanced erythropoiesis on iron absorption. J Lab Clin Med 1992; 120: 746–751.

25 Kemna E, Tjalsma H, Laarakkers C, Nemeth E, Willems H, Swinkels D: Novel urine hep-cidin assay by mass spectrometry. Blood 2005; 106: 3268–3270.

26 Nemeth E, Preza GC, Jun CL, Kaplan J, War-ing AJ, Ganz T: The N-terminus of hepcidin is essential for its interaction with ferropor-tin: structure-function study. Blood 2006; 107: 328–333.

27 Nicolas G, Chauvet C, Viatte L, Danan JL, Bigard X, Devaux I, Beaumont C, Kahn A, Vaulont S: The gene encoding the iron regu-latory peptide hepcidin is regulated by ane-mia, hypoxia, and inflammation. J Clin In-vest 2002; 110: 1037–1044.

28 Nicolas G, Viatte L, Bennoun M, Beaumont C, Kahn A, Vaulont S: Hepcidin, a new iron regulatory peptide. Blood Cells Mol Dis 2002; 29: 327–353.

29 Kulaksiz H, Gehrke SG, Janetzko A, Rost D, Bruckner T, Kallinowski B, Stremmel W: Pro-hepcidin: expression and cell specific localisation in the liver and its regulation in hereditary haemochromatosis, chronic renal insufficiency, and renal anaemia. Gut 2004; 53: 735–743.

30 Dallalio G, Fleury T, Means RT: Serum hep-cidin in clinical specimens. Br J Haematol 2003; 122: 996–1000.

31 Hsu SP, Ching CK, Chien CT, Hung KY: Plasma prohepcidin positively correlates with haematocrit in chronic hemodialysis patients. Blood Purif 2006; 24: 311–316.

32 Malyszko J, Malyszko JS, Pawlak K, Mysli-wiec M: Hepcidin, iron status, and renal function in chronic renal failure, kidney transplantation, and hemodialysis. Am J Haematol 2006; 81: 832–837.

33 Taes YEC, Wuyts B, Boelaert JR, Vriese AS, Delanghe JR: Prohepcidin accumulates in renal insufficiency. Clin Chem Lab Med 2004; 42: 387–389.

34 Locatelli F, Aljama P, Bárány P, Canaud B, Carrera F, Eckardt KU, Hörl WH, Macdou-gal IC, Macleod A, Wiecek A, Cameron S, European Best Practice Guidelines Working Group: Revised European best practice guidelines for the management of anemia in patients with chronic renal failure. Nephrol Dial Transplant 2004; 19(suppl 2):ii1–ii47.

35 Hillman RS, Ault KA: Clinical approach to anemia; in Hillman RS, Ault KA, Rinder HM (eds): Hematology in Clinical Practice, ed 3. New York, McGraw-Hill, 2002, pp 12–26.

36 Wrighting DM, Andrews NC: Interleukin-6 induces hepcidin expression through STAT3. Blood 2006; 108: 3204–3209.

37 Fleming RE, Sly WS: Hepcidin: a putative iron-regulatory hormone relevant to heredi-tary hemochromatosis and the anemia of chronic disease. Proc Natl Acad Sci USA 2001; 98: 8160–8162.

38 Nemeth E, Rivera S, Gabayan V, Keller C, Taudorf S, Pedersen BK, Ganz T: IL-6 medi-ates hypoferremia of inflammation by in-ducing the synthesis of the iron regulatory hormone hepcidin. J Clin Invest 2004; 113: 1271–1276.

39 Vokurka M, Krijt J, Sulc K, Necas E: Hepci-din mRNA levels in mouse liver respond to inhibition of erythropoiesis. Physiol Res 2006; 55: 667–674.

40 Courselaud B, Troadec MB, Fruchon S, et al: Strain and gender modulate hepatic hepci-din 1 and 2 mRNA expression in mice. Blood Cells Mol Dis 2004; 32: 283–289.

41 Krijt J, Niederkofler V, Salie R, Sefc L, Peli-chovská T, Vokurka M, Necas E: Effect of phlebotomy on hepcidin expression in he-mojuvelin-mutant mice. Blood Cells Mol Dis 2007; 39: 92–95.

42 Furaso M, Munaretto G, Spinello M, Rebe-schini M, Amici G, Gallieni M, Picoli A: Sol-uble transferrin receptors and reticulocyte hemoglobin concentration in the assessment of iron deficiency in hemodialysis patients. J Nephrol 2005; 18: 72–79.

43 Tarng D, Huang T: Determinants of circulat-ing soluble transferring receptor level in chronic hemodialysis patients. Nephrol Dial Transplant 2002; 17: 1063–1069.



Paper V

“Effect of hemodialysis procedure in prohepcidin serum

levels in chronic kidney disease patients”.

Clin Nephrol, in press.



LETTER TO THE EDITOR

Effect of hemodialysisprocedure in prohepcidinserum levels in regularhemodialysis patients

E. Costa1,2,3, S. Rocha1,2,

P. Rocha-Pereira2,4, F. Reis5,

E. Castro1,2, F. Teixeira5, V. Miranda6,

M. do Sameiro Faria6, A. Loureiro7,

A. Quintanilha2,8, L. Belo1,2 and

A. Santos-Silva1,2

1Faculdade de Farmácia, Serviço de

Bioquímica, 2Instituto de Biologia

Molecular e Celular, Universidade do

Porto, 3Instituto de Ciências da Saúde,

Universidade Católica Portuguesa,4Centro de Investigaçao Ciências da

Saúde, Universidade da Beira Interior,

Covilha, 5Instituto de Farmacologia

e Terapêutica Experimental,

Universidade de Coimbra,6FMC, Dinefro-Diálises e Nefrologia,

SA, 7Uninefro-Sociedade Prestadora de

cuidados Médicos e de Diálise, SA,8Instituto de Ciências Biomédicas Abel

Salazar, Universidade do Porto

Sir –, A complex regulatory net work that

governs iron traffic has emerged, and point to

hepcidin as a major evo lutionary con served

reg u la tor of iron dis tri bu tion [Arth 1999,

Atanasiu et al. 2006]. This small hor mone

produced by the mammalian liver has been

proposed as a central mediator of dietary iron

ab sorp tion. Hepcidin seems to down-regulate

both iron up take from the small in testine and

release of iron from macrophages, as well as

pla cen tal iron trans port [Kulaksiz et al. 2004,

Nicolas et al. 2002].

The syn the sis of hepcidin is stim ulated by

inflammation and iron over load. It is synthe-

sized as a preprohepcidin of 84 amino acids.

The signal peptide is cleaved leading to the 60

amino ac ids prohepcidin, which is fur ther

processed to the 25 amino acids hepcidin

[Dallalio et al. 2003, Hsu et al. 2006].

The bi o log i cal im por tance of the pre cur -

sor mol e cule of hepcidin (prohepcidin) in

reg u lat ing iron me tab o lism has so far re-

mained un de ter mined. How ever, there is

some ev i dence that prohepcidin lev els are a

re li able in di ca tor of hepcidin lev els and ac tiv-

ity [Dallalio et al. 2003, Hsu et al. 2006]. Due

to the lack of available lab oratory methods for

quan ti fy ing cir cu lat ing hepcidin in clinical

sam ples, se rum prohepcidin lev els which are

easily measured with the widely avail able

ELISA kits, are used as an indicator of hep-

cidin levels.

Hemodialysis (HD) patients present high

se rum lev els of prohepcidin, as well as of

markers of inflammation and of re duced iron

mobilization [Costa et al. 2008]. Although in-

flammation could be the cause of the high lev-

els of prohepcidin in HD patients, no relevant

cor re la tions be tween prohepcidin and in flam-

ma tory mark ers were re ported, sug gest ing

that other fac tors than just in flammation may

reg u late se rum prohepcidin lev els in HD pa-

tients. No data ex ist about the role of HD pro -

cedure up on prohepcidin se rum lev els in

these pa tients. We won der if prohepcidin,

presenting a low molecular weight (10 kDa)

is removed dur ing the HD pro cedure.

We studied 20 regular HD pa tients (8

males, 12 females, mean age 60.3 ± 16.7

years), un der re com bi nant hu man eryth ro poi-

e tin (rhEPO) treat ment. Pa tients were under

therapeutic HD 3 times per week, for 3 – 5 h,

for a median period of 41 months. All pa tients

used the high-flux polysulfone FX-class

dialyzers of Fresenius (12 the FX60 and 8 the

FX80 dialyzer type). The causes of re nal fail -

ure in the patient pop ulation were as fol lows:

diabetic nephropathy (n = 5), chronic glome-

rulonephritis (n = 3), hy pertensive nephro -

sclerosis (n = 2), obstructive nephropathy

(n = 1), nephrolithiasis (n = 1), chronic in ter-

stitial ne phritis (n = 1) and chronic re nal fail-

ure of un certain e tiology (n = 7). Pa tients with

au to im mune dis ease, ma lig nancy, he ma to -

logical dis orders and acute or chronic in fec-

tion were excluded. All par ticipants gave

their in formed con sent to participate in this

study. A con trol group was also studied and

in cluded 26 healthy vol un teers pre sent ing

nor mal he ma to log i cal and bio chem i cal val-

ues, with no history of re nal or inflammatory

diseases, and, as far as possible, age and gen-

der matched with HD patients.

Blood samples were drawn from fasting

controls and from HD pa tients be fore and im-

Clinical Nephrology, Vol. 70 – No. /2007 – Letters to the editor 1

CN - 6172 / 10.10.2008



me di ately af ter the sec ond di al y sis ses sion of

the week. Se rum lev els of prohepcidin were

determined by an en zyme-linked immun sor -

bent as say (Hepcidin Prohormone ELISA,

IBL, Hamburg, Ger many) ac cording to the

man u fac turer’s recommendations.

As ex pected, before HD, the pa tients

showed sig nif i cantly higher prohepcidin

lev els (164.97 (150.48 – 218.42 ng/ml)) vs.

88.67 (79.21 – 104.22 ng/ml), p < 0.001),

when compared to healthy con trols (more

than 2-fold the con trol value). Af ter the dialy-

sis session, HD pa tients showed a sta tistically

sig nif i cant de crease in prohepcidin se rum

levels (Fig ure 1). Even so, the levels of pro -

hepcidin af ter HD were still sig nificantly

higher than that found in healthy controls.

The me dian de crease of prohepcidin se rum

levels af ter the HD procedure was 11.1%.

These re sults dem onstrate for the first time

that HD pro cedure re duces prohepcidin se -

rum lev els. Con sidering that our pa tients in -

cluded 10 re sponders and 10 nonresponders

to rhEPO ther apy and, there fore, pa tients

re ceiv ing dif fer ent doses of rhEPO, we

searched for a cor relation be tween weekly

rhEPO/kg dose and the re duction ob served in

prohepcidin se rum lev els af ter HD, and no

cor re la tion was found.

In con clu sion, our re sults dem on strate

that HD pro cedure is implicated in the reg ula-

tion of the prohepcidin se rum lev els, by re -

mov ing prohepcidin from cir cu la tion. More-

over, this re moval could have an ad ditional

ben e fit ef fect. The re duc tion in prohepcidin

serum lev els would diminish the down-regu -

lation ef fect upon iron mo bi li za tion that

would be ex pected to occur with the pro-

hepcidin se rum lev els observed before the

HD procedure.

Acknowledgments

The authors are grateful to the nurses of

Fresenius Med i cal Cen ter, Dinefro-Diálises e

Nefrologia, SA and Uninefro-Sociedade Pre-

stadora de Cuidados Médicos e de Diálise,

SA, for technical sup port. This study was

supported by a PhD grant (SFRH/BD/276-

88/2006) attributed to E. Costa by Fundaçao

para a Ciencia e Tecnologia (FCT) and Fundo

So cial Europeu.

References

Arth RH. Iron metabolism in end-stage re nal disease.

Semin Dial. 1999; 12: 224-230.

Atanasiu V, Manolescu B, Stoian I. Hepcidin – cen tral

reg u la tor. Eur J Haematol. 2006; 78: 1-10.

Nicolas G, Viatte L, Bennoun M, Beaumont C, Kahn A,

Vaulont S. Hepcidin, a new iron reg ulatory pep tide.

Blood Cell Mol Dis. 2002; 29: 327-353.

Costa E, Pereira BJG, Rocha-Pereira P, Rocha S, Reis F,

Cas tro E, Teixeira F, Miranda V, Sameiro Faria M,

Loureiro A, Quintanilha A, Belo L, Santos-Silva A.

Role of prohepcidin, in flam ma tory mark ers and iron

sta tus in re sis tance to rhEPO therapy in hemodialysis

patients. Am J Nephrol. 2008; 28: 677-683.

Dallalio G, Fleu ry T, Means RT. Se rum hepcidin in clin i-

cal spec imens. Br J Haematol. 2003; 122: 996-1000.

Hsu SP, Ching CK, Chien CT, Hung KY. Plasma pro hep -

cidin pos itively cor relates with hematocrit in chronic

hemodialysis pa tients. Blood Purif. 2006; 24: 311-

316.

Kulaksiz H, Gehrke SG, Janetzko A, Rost D, Bruckner T,

Kallinowski B, Stremmel W. Prohepcidin: ex pres sion

and cell-specific lo cal isa tion in the liver and its reg u-

lation in hereditary hemochromatosis, chronic renal

in suf fi ciency, and re nal ane mia. Gut. 2004; 53: 735-

743.

Cor re spon dence to

E. Costa

Serviço de Bioquímica,

Faculdade de Farmácia,

Universidade do Porto,

Rua Aníbal Cunha, 164,

4099-030 Porto, Portugal

[email protected]

Clinical Nephrology, Vol. 70 – No. /2007 – Letters to the editor 2

CN - 6172 / 10.10.2008

Fig ure 1. Se rum prohepcidin lev els from con trols

and HD patients (be fore and im mediately after HD).

Boxplot shows me dian value (hor izontal line in box)

and first and third quartiles (in fe rior and su pe rior line

of the box, respectively).



Paper VI

“DMT1 (NRAMP2/DCT1) genetic variability and resis-

tance to rhEPO therapy in chronic kidney disease patients un-

der haemodialysis”.

Acta Haematologica 2008; 120: 11-13.




Brief Communication

Acta Haematol 2008;120:11–13 DOI: 10.1159/000149496

DMT1 (NRAMP2/DCT1) Genetic Variability and Resistance to Recombinant Human Erythropoietin Therapy in Chronic Kidney Disease Patients under Haemodialysis

Elísio Costa a, b, d Susana Rocha a, b Petronila Rocha-Pereira b, e Flávio Reis f

Elisabeth Castro a, b Frederico Teixeira f Vasco Miranda g Maria do Sameiro Faria g

Alfredo Loureiro h Alexandre Quintanilha b, c Luís Belo a, b Alice Santos-Silva a, b

a Faculdade de Farmácia, Serviço de Bioquímica, b

Instituto de Biologia Molecular e Celular, and c

Instituto de Ciências Biomédicas Abel Salazar, Universidade do Porto, Porto , d Instituto de Ciências da Saúde,

Universidade Católica Portuguesa, Porto , e Centro de Investigação de Ciências da Saúde da Universidade da

Beira Interior, Covilhã , f Instituto de Farmacologia e Terapêutica Experimental, Universidade de Coimbra, Coimbra , g

FMC, Dinefro – Diálises e Nefrologia, SA, Maia , h Uninefro – Sociedade Prestadora de Cuidados Médicos e de Diálise,

SA, Guimarães , Portugal

ceptor, haemochromatosis (HFE) protein and DMT1 (NRAMP2/DCT1). In literature, alterations in the trans-ferrin receptor gene have been associated with HFE [5] , and HFE gene mutations are associated with a reduction in the amount of rhEPO necessary to support erythro-poiesis in haemodialysis patients [6] . However, these 2 proteins are unlikely to be the main causes of the resis-tance to rhEPO therapy in CKD patients under haemo-dialysis, as HFE is characterized by increases in serum iron parameters, namely transferrin saturation, not found in non-responder CKD patients. On the other hand, the DMT1 gene could be a good candidate to un-derlie the variability of response to rhEPO in CKD pa-tients. Indeed, the few described mutation cases in the DMT1 gene are associated with an inhibition of intesti-nal iron absorption and a decrease in erythroid cell pre-cursor iron uptake, resulting in hypochromic and mi-crocytic anaemia [7] . This type of anaemia is similar to that observed in non-responder CKD patients, resulting of a functional iron-depleted erythropoiesis in the pres-ence of adequate serum iron levels [4] . As far as we know, there are no data regarding the association of DMT1 gene

The management of anaemia in chronic kidney dis-ease (CKD) was changed by the introduction of recombi-nant human erythropoietin (rhEPO) therapy, allowing a significant correction of anaemia; however, around 5–10% of the patients show a marked resistance to rhEPO therapy [1, 2] . Several conditions were reported as being associated with rhEPO resistance, but changes in iron ho-meostasis is one of the most studied causes [3] .

CKD patients who do not respond to rhEPO therapy seem to present a ‘functional’ iron deficiency, character-ized by the presence of adequate iron stores as defined by conventional criteria, but apparently with an inability to sufficiently mobilize this iron to adequately support erythropoiesis [4] . We recently found that compared with responder patients, non-responders presented no statisti-cal differences for iron, transferrin saturation and ferri-tin values, but had a significantly lower haemoglobin concentration and mean cell haemoglobin, as well as sig-nificantly higher plasma levels of the soluble transferrin receptor (s-TfR) [4] .

Iron uptake from plasma to erythroid precursor cells is regulated by different proteins, namely transferrin re-

Received: March 27, 2008 Accepted after revision: May 23, 2008 Published online: August 1, 2008

Elísio Costa Serviço de Bioquímica, Faculdade de Farmácia, Universidade do Porto Rua Aníbal Cunha, 164 4099-030 Porto (Portugal) Tel. +351 222 078 900, Fax +351 222 003 977, E-Mail [email protected]

© 2008 S. Karger AG, Basel 0001–5792/08/1201–0011$24.50/0

Accessible online at: www.karger.com/aha



Costa et al.

Acta Haematol 2008;120:11–13 12

mutations/polymorphisms in patients with ‘functional’ iron deficiency.

In this context, we studied 63 CKD patients under haemodialysis (36 males, 27 females; age 62.1 8 15.7 years) and rhEPO treatment (10 with darbopoietin- � and 53 with epoetin), including 32 responders and 31 non-re-sponders to rhEPO therapy [8] . The rhEPO maintenance dose for responders was 89.65 8 57.62 U/kg/week and for non-responders 572.99 8 193.84 U/kg/week. Patients with autoimmune disease, malignancy, haematological disorders and acute or chronic infection were excluded. Intravenous iron supplementation (iron sucrose) was based on the European Best Practice Guidelines for the Management of Anemia in CKD patients under haemo-dialysis [8] . No statistically significant difference was found in total iron supplement given during the last year, between responder and non-responder patients. We also studied a control group with 26 healthy volunteers, pre-senting normal haematological and biochemical values, with no history of renal or inflammatory diseases, and, as far as possible, age and gender matched with the CKD patients (8 males, 17 females; mean age 47.81 8 14.69). All patients and controls gave their informed consent to participate in this study.

In all participants (patients and controls), haemato-logical counts were performed (Sysmex K1000; Sysmex, Hamburg, Germany) as well as the evaluation of serum iron levels (Randox Laboratories Ltd., UK), ferritin (Ran-dox Laboratories Ltd.) and transferrin (Randox Labora-tories Ltd.) and of plasma s-TfR (human sTfR immunoas-say, R&D Systems, Minneapolis, Minn., USA). Genotype of the c.309+44C ] A, c.1254T ] C, c.1303T ] C and c.1629–16C ] G DMT1 polymorphisms/mutations was performed by restriction fragment length analysis, using the conditions previously described [9] , with minor mod-ifications. Mutation nomenclature was used according to the recommendations of the Human Genome Variation Society (2005; http://www.hgvs.org/mutnomen).

The frequencies of the DMT1 gene polymorphisms/mutations tested are shown in table 1 . No statistically sig-nificant differences were found in allele frequency be-tween the controls and CKD patients, nor when compar-ing the 2 groups of patients. These frequencies were sim-ilar to those previously described in literature [9] .

Estimations of the haplotype frequencies were per-formed using the program Arlequin version 3.11 [10] . This program estimates the maximum likelihood haplo-type frequencies from observed data using an expecta-tion-maximization algorithm for multilocus genotype data when the genetic phase is unknown. The most fre-quent haplotypes found were: CCTC (c.309+44C, c.1254C, c.1303T, c.1629–16C), ACCC (c.309+44A, c.1254C, c.1303C, c.1629–16C) and ACTG (c.309+44A, c.1254C, c.1303T, c.1629–16G) ( table 2 ). No statistically significant differences were found in the frequencies of the CCTC and ACCC haplotypes between controls and CKD patients, nor between responders and non-respond-er patients. The frequencies of these 2 haplotypes are very similar to those found in other populations [9] . However, the ACTG haplotype was statistically more frequent in CKD patients than in the control group (p ! 0.05); nev-ertheless, no statistical differences were found between

Table 1. Frequency of studied DMT1 alleles in CKD patients and controls (%)

c.309+44C ] A c.1254T ] C c.1303T ] C c.1629–16C ] G

C allele A allele T allele C allele T allele C allele C allele G allele

CKD patients (n = 63) 76.98 23.02 84.13 15.87 99.21 0.79 92.06 7.94 rhEPO responders (n = 32) 73.44 26.56 82.81 17.19 100 0 92.19 7.81 rhEPO non-responders (n = 31) 80.65 19.35 85.48 14.52 98.39 1.61 91.94 8.06 Controls (n = 26) 76.92 23.08 80.77 19.23 100 0 94.23 5.77

Table 2. Haplotype frequencies in CKD patients and controls de-termined by Arlequin version 3.11 (%)

CCTC ACCC ACTG

CKD patients (n = 63) 65.87 18.25 6.35 rhEPO responders (n = 32) 64.06 21.88 7.81 rhEPO non-responders (n = 31) 67.74 14.51 4.83 Controls (n = 26) 71.15 5.76 0

CCTC: c.309+44C; c.1254C; c.1303T; c.1629–16C. ACCC: c.309+44A; c.1254C; c.1303C; c.1629–16C. ACTG: c.309+44A; c.1254C; c.1303T; c.1629–16G.



Resistance to rhEPO Therapy in CKD Patients under Haemodialysis

Acta Haematol 2008;120:11–13 13

responder and non-responder CKD patients. The fre-quency of this ACTG haplotype in CKD patients was also very similar to that found in literature; however, in the control group, the prevalence of this haplotype is signifi-cantly lower than that found in other previously described control populations [9] . These differences can result from the small sample size of our control group.

When we compared rhEPO doses, haematological data and iron status (iron, ferritin and transferrin satura-tion) of CKD patients under haemodialysis by haplotype, no statistically significant differences were found (data not shown), suggesting that DMT1 gene haplotypes are not associated with changes in haematological data and iron status in CKD patients, nor in rhEPO doses required to achieve target haemoglobin levels.

CKD patients under haemodialysis are treated with intravenous iron; thus, the main factors limiting iron availability for erythropoiesis could be a deficiency in iron uptake to erythroblast and/or iron trapping into macrophages. The latter is associated with a failure in re-cycling/export mainly due to a decreased activity of the iron export ferroportin-1 (related to increased hepcidin expression associated with the inflammatory process found in haemodialysis patients). As DMT1 protein is re-

lated to both mechanisms, it is reasonable to hypothesize that mutations/polymorphisms in this gene may decrease the iron availability for erythroblast iron metabolism and haem biosynthesis in CKD non-responder patients to rhEPO therapy, even in the presence of adequate serum iron levels. However, our study failed to demonstrate a link between mutations/polymorphisms in the DMT1 gene and resistance to rhEPO therapy. Moreover, no influence of DMT1 gene mutations/polymorphisms on rhEPO doses, haematological data and iron status was demonstrated in our CKD patients under haemodialysis. Further studies are warranted, namely using a higher number of CKD patients, full screening of mutations in DMT1 gene and/or other genes related to iron metabo-lism, to clarify the mechanism underlying rhEPO resis-tance.

Acknowledgements

We are very grateful to FMC, Dinefro – Diálises e Nefrologia, SA, and Uninefro – Sociedade Prestadora de Cuidados Médicos e de Diálise, SA, and to their nurses for the technical support. This study was supported by a PhD grant (SFRH/BD/27688/2006) at-tributed to E. Costa by FCT and FSE.

References



3 Deicher R, Horl WH: New insights into the regulation of iron homeostasis. Eur J Clin In-vest 2006; 36: 301–309.

4 Costa E, Pereira BJG, Rocha-Pereira P, Rocha S, Reis F, Castro E, Teixeira F, Miranda V, Sameiro-Faria M, Loureiro A, Quintanilha A, Belo L, Santos-Silva A: Role of prohepci-din, inflammatory markers and iron status in resistance to rhEPO therapy in hemodi-alysis patients. Am J Nephrol 2008; 28: 677–683.

5 Wallace DF, Subramaniam VN: Non-HFE haemochromatosis. World J Gastroenterol 2007; 13: 4690–4698.

6 Valenti L, Valenti G, Como G, Santorelli G, Dongiovanni P, Rametta R, Fracanzani AL, Tavazzi D, Messa PG, Fargion S: HFE geno-type influences erythropoiesis support re-quirement in hemodialysis patients: a pro-spective study. Am J Nephrol 2008; 28: 311–316.

7 Fleming MD, Trenor CC, Su MA, Foernzler D, Beier DR, Dietrich WF, Andrews NC: Mi-crocytic anaemia mice have a mutation in Nramp2, a candidate iron transporter gene. Nat Genet 1997; 16: 383–386.

8 Locatelli F, Aljama P, Barany P, Canaud B, Carrera F, Eckardt KU, Horl WH, Macdou-gal IC, Macleod A, Wiecek A, Cameron S, European Best Practice Guidelines Working Group: Revised European best practice guidelines for the management of anemia in patients with chronic renal failure. Nephrol Dial Transplant 2004; 19(suppl 2):ii1–ii47.

9 Kelleher T, Ryan E, Barrett S, O’Keane C, Crowe J: DMT1 genetic variability is not re-sponsible for phenotype variability in hered-itary HFE. Blood Cells Mol Dis 2004; 33: 35–39.

10 Excoffier L, Laval LG, Schneider S: Arlequin ver 3.0: an integral software package for pop-ulation genetics data analysis. Evol Bioinfor-matics Online 2005; 1: 47–50.



Paper VII

“Altered Erythrocyte Membrane Protein Composition in

Chronic Kidney Disease Stage 5 Patients under Haemodialysis

and Recombinant Human Erythropoietin Therapy”

Blood Purif 2008; 26: 267-273.




Original Paper

Blood Purif 2008;26:267–273 DOI: 10.1159/000126922

Altered Erythrocyte Membrane Protein Composition in Chronic Kidney Disease Stage 5 Patients under Haemodialysis and Recombinant Human Erythropoietin Therapy

Elísio Costa a, b, d Susana Rocha a, b Petronila Rocha-Pereira e

Elisabeth Castro a, b Vasco Miranda f Maria do Sameiro Faria f Alfredo Loureiro g

Alexandre Quintanilha b, c Luís Belo a, b Alice Santos-Silva a, b

a Serviço de Bioquímica, Faculdade de Farmácia, Universidade do Porto, b Instituto de Biologia Molecular e Celular, Universidade do Porto ; c Instituto de Ciências Biomédicas Abel Salazar (ICBAS), Universidade do Porto, Porto , d Departamento das Tecnologias de Diagnóstico e Terapêutica, Escola Superior de Saúde, Instituto Politécnico de Bragança, Bragança , e Universidade da Beira Interior, Covilhã , f FMC, Dinefro – Diálises e Nefrologia, SA, Maia , g Uninefro – Sociedade Prestadora de Cuidados Médicos e de Diálise , SA, Guimarães , Portugal

more microcytic and anisocytic RBCs, than responders; sig-nificantly altered ankyrin/band 3 and spectrin/ankyrin ratios were also observed. CKD stage 5 patients under HD are as-sociated with an altered protein membrane structure, which seems to the disease itself and/or to the interaction with HD membranes.

Copyright © 2008 S. Karger AG, Basel

Introduction

Anaemia is a common complication that contributes to the burden of chronic kidney disease (CKD) stage 5. It has also a negative impact on cardiovascular system, cog-nitive function, exercise capacity and quality of life, re-sulting in a significant mortality and morbidity in these patients [1–3] .

This anaemia can result from a decreased bone mar-row production of red blood cells (RBC), due to the in-ability of the failing kidneys to secrete erythropoietin (EPO) [4] , to the accumulation of uraemic toxins, or to

Key Words Chronic kidney disease � Chronic renal failure � Haemodialysis � Spectrin � Erythrocyte membrane proteins � Resistance to rhEPO

Abstract Our aim was to evaluate red blood cell (RBC) membrane pro-tein composition in chronic kidney disease (CKD) stage 5 pa-tients under haemodialysis (HD) and recombinant human erythropoietin (rhEPO) therapy, and its linkage to rhEPO hy-poresponsiveness. We evaluated in 63 CKD stage 5 patients (32 responders and 31 non-responders to rhEPO therapy) and in 26 healthy controls RBC count, haematocrit, haemo-globin concentration, haematimetric indices, reticulocyte count, reticulocyte production index, RBC osmotic fragility test and membrane protein analyses. CKD stage 5 patients presented significant changes in membrane protein compo-sition, namely a reduction in spectrin, associated to altered protein 4.1/spectrin and spectrin/band 3 ratios. Non-re-sponder CKD stage 5 patients were more anaemic, with

Received: July 25, 2007 Accepted: December 12, 2007 Published online: April 17, 2008

Alice Santos-Silva Serviço de Bioquímica, Faculdade de Farmácia Universidade do Porto, Rua Aníbal Cunha, 164 PT–4099-030 Porto (Portugal) Tel. +351 22 207 8900, Fax +351 22 200 3977, E-Mail [email protected]

© 2008 S. Karger AG, Basel0253–5068/08/0263–0267$24.50/0

Accessible online at:www.karger.com/bpu



Costa et al.

Blood Purif 2008;26:267–273268

excessive toxic storage of aluminium in the bone marrow [5] . Blood loss and premature RBC destruction are also frequently associated with anaemia in CKD stage 5 pa-tients [6–9] .

Modifications in RBC membrane protein composi-tion may account for changes in the deformability of the cell, compromising its circulation in the microvascula-ture and its survival. The RBC membrane is a complex structure comprising a lipidic bilayer, integral proteins and the skeleton. Spectrin is the major protein of the cy-toskeleton, and, therefore, the major responsible for RBC shape, integrity and deformability. It links the cytoskel-eton to the lipid bilayer, by vertical protein interactions with the transmembrane proteins, band 3 and glycopho-rin C [10] . In the vertical protein interaction of spectrin with band 3 are also involved ankyrin (also known as band 2.1) and protein 4.2. A normal linkage of spectrin with the other proteins of the cytoskeleton assures nor-mal horizontal protein interactions.

RBCs are physically stressed during the haemodialysis (HD) process, and metabolically stressed by the unfa-vourable plasmatic environment, due to metabolite accu-mulation, and by the high levels of haemoglobin (Hb) au-toxidation, due to the increase in Hb turnover, a physio-logic compensation mechanism triggered in case of anaemia [11, 12] . The RBCs are, therefore, continuously challenged to sustain Hb in its reduced functional form, as well as to maintain the integrity and deformability of the membrane.

When Hb is denatured, it links to the cytoplasmic pole of band 3, triggering its aggregation and leading to the formation of strictly lipidic portions of the membrane, poorly linked to the cytoskeleton. These cells are proba-bly more prone to undergo vesiculation (loss of poorly linked membrane portions) whenever they have to circu-late through the HD membranes or the microvascula-ture. Vesiculation may, therefore, lead to the development of deficiencies in the erythrocyte membrane of CKD stage 5 patients [10, 13, 14] .

The development of recombinant human EPO(rhEPO) therapy represents a major advance in the treat-ment of renal anaemia, reducing its associated complica-tions and improving significantly the quality of life of CKD stage 5 patients [15, 16] . However, more than 25%of these patients do not respond to the usual doses ofrhEPO, requiring higher doses to avoid anaemia, and about 5–10% of them do not respond to rhEPO therapy [10–19] . There are several factors that may contribute to the development of this rhEPO resistance [15–20] . In spite of the several proposed factors for rhEPO resistance,

the marked variability in individual sensitivity to the drug is still unexplained. In these cases of resistance to rhEPO therapy, it is often difficult to get target levels of Hb (between 11 and 12 g/dl) as RBCs are probably even more metabolically and physically stressed.

The aim of our work was to evaluate the changes in RBC membrane protein composition in CKD stage 5 pa-tients under HD and rhEPO therapy, and to compare RBC membrane protein composition in responder and non-responder patients, to rhEPO therapy.

Materials and Methods

Subjects We studied 63 CKD stage 5 patients (36 males, 27 females;

mean age 62.1 8 15.7) under rhEPO treatment, for a median pe-riod of time of 36 months, 10 with darbopoietin- � and 53 with epoetin. The CKD stage 5 patients included 32 responders and 31 non-responders to rhEPO therapy. Classification of the patients, as responders or non-responders, was performed in accordance with the European Best Practice Guidelines [21] , which defines resistance to rhEPO as a failure to achieve target Hb levels (be-tween 11 and 12 g/dl) with maintained doses of rhEPO 1 300 IU/kg/week of epoetin or 1.5 � g/kg/week of darbopoietin- � . The two groups of patients were matched for age, gender, weight, body mass index, mean time under HD, urea reduction ratio, urea KtV and parathyroid hormone serum levels.

CKD stage 5 patients were under therapeutic HD three times per week, for 3–5 h, for a median period of time of 36 months. All patients used the high-flux polysulfone FX-class dialysers of Fre-senius, 34 with FX60, 27 with FX80 and 2 with FX100 dialyser type. The causes of renal failure in patient’s population were as follows: diabetic nephropathy (n = 19), chronic glomerulonephri-tis (n = 11), polycystic kidney disease (n = 3), hypertensive neph-rosclerosis (n = 3), obstructive nephropathy (n = 3), pyelonephri-tis associated with neurogenic bladder (n = 1), nephrolithiasis(n = 1), chronic interstitial nephritis (n = 1), Alport’s syndrome(n = 1), renal vascular disease due to polyarteritis (n = 1) and chronic renal failure of uncertain aetiology (n = 19). Patients with autoimmune disease, malignancy, haematological disorders, and acute or chronic infection were excluded. All patients gave their informed consent to participate in this study.

Besides rhEPO therapy, all patients were under iron and folate prophylactic therapies, in accordance to the recommendations of European Best Practice Guidelines [21] , to avoid nutrient eryth-ropoietic deficiencies.

Erythrocytes from diabetic patients were associated [22] with a higher anionic membrane charge and with an altered deform-ability, though no changes in membrane protein composition were reported. To exclude this possible confounding factor we further analysed the results by performing statistical analysis with and without exclusion of the 19 CKD stage 5 patients with a diabetic nephropathy.

The control group included 26 healthy volunteers presenting normal haematological and biochemical values, with no history of renal or inflammatory diseases, and, as far as possible, age- and



RBC Membrane Proteins in HD Patients Blood Purif 2008;26:267–273 269

gender-matched with CKD stage 5 patients. Controls did not re-ceive any medication known to interfere with the studied vari-ables.

Assays Blood samples (using EDTA and heparin as anticoagulants)

were drawn from fasting controls or before the second dialysis session of the week in HD patients.

RBC count, haematocrit, Hb concentration, haematimetric indices [mean cell volume (MCV), mean cell Hb (MCH) and mean cell Hb concentration (MCHC)] and red cell distribution width (RDW) were measured by using an automatic blood cell counter (Sysmex K1000; Sysmex, Hamburg, Germany). Reticulo-cyte count was made by microscopic counting on blood smears after vital staining with new methylene blue (reticulocyte stain; Sigma, St. Louis, Mo., USA). The reticulocyte production index (RPI) was calculated as an appropriate way to measure the effec-tive RBC production, by correcting for both changes in haemato-crit (degree of anaemia) and for premature reticulocyte releasing from the bone marrow [23] .

Osmotic fragility test (OFT) was performed (heparin as anti-coagulant) with fresh blood and after 24 h of blood incubation at 37 ° C [24] . Blood incubation will trigger a metabolic stress within the RBCs due to a progressive reduction in plasmatic glucose, and, therefore, it will turn the assay more sensitive, as the number of older RBCs, that are more susceptible to metabolic stress, will in-crease along the incubation period [24] .

Serum iron concentration was determined using a colorimet-ric method (Randox Laboratories Ltd, Northern Ireland), where-as serum ferritin and serum transferrin were measured by im-munoturbidimetry (Randox Laboratories Ltd, Northern Ire-land). Transferrin saturation (TS) was calculated by the formula: TS (%) = 70.9 ! serum iron concentration in ( � g/dl)/serum transferrin concentration in (mg/dl). The vitamin B 12 serum lev-els were determined by electrochemiluminescence immunoassay (Roche Laboratories, Mannheim, Germany).

To prepare the RBC membranes, plasma and leukocytes were isolated and discarded after centrifugation on a double density gradient (Histopaque 1.077 and 1.119, Sigma). RBCs were washed in saline solution and then lysed by hypotonic lysis, according to Dodge et al. [25] . The obtained membrane suspensions were washed in Dodge buffer, adding phenylmethylsulphonyl fluoride as a protease inhibitor in the first two washes, with a final con-centration of 0.1 m M . The protein concentration of the RBC membrane suspension was determined by Bradford’s method [26] .

The electrophoretic analysis of the RBC membrane proteins was carried out on a discontinuous system of polyacrylamide in the presence of sodium dodecyl sulphate (SDS-PAGE), using a 5–15% linear acrylamide gradient gel and a 3.5–17% exponential acrylamide gradient gel, according to the Laemmli and Fair-banks methods, respectively [27, 28] . The proteins were stained with Coomassie brilliant blue and scanned (Darkroom CN UV/wl, BioCaptMW Version 99; Vilber Lourmat, Marne-la-Vallée, France). The relative amount of each major RBC membrane pro-tein (spectrin, ankyrin, band 3, 4.1, 4.2, 5, 6 and 7 proteins) was quantified by densitometry (Bio1D++ Version 99; Vilber Lour-mat, France). The electrophoretic analysis for each RBC mem-brane sample was performed in duplicate gels, and, in each gel, duplicates of each sample were loaded. Moreover, some ratios

between membrane protein values, for patients and controls, were evaluated to further analyse membrane protein interac-tion.

Data Analysis Measurements are presented as mean 8 SD or as median val-

ues (interquartile range). For statistical analysis, we used the Sta-tistical Package for Social Sciences (SPSS Version 15.0 for Win-dows; SPSS, Inc., Chicago, Ill., USA). Kolmogorov-Smirnov sta-tistics were used to evaluate sample normality distribution. Multiple comparisons between groups were performed by one-way ANOVA supplemented with Turkey’s HSD post-hoc test. For data presenting a non-Gaussian distribution, we used the Krus-kal-Wallis test and Mann-Whitney U test. Spearman’s rank cor-relation coefficient was used to evaluate relationships between sets of data. Significance was accepted at p ! 0.05.

Results

The results were analysed in two ways, one in order to study the differences between healthy controls and CKD stage 5 patients, and secondly to study the differences be-tween responder and non-responder patients.

In tables 1 and 2 we present the weekly rhEPO dose, haematological data, OFT values and RBC membrane protein profile for controls and CKD stage 5 patients. The patients presented anaemia, slightly regenerative, as shown by the significantly decreased RBC count, Hb and haematocrit, alongside with a significantly increased re-ticulocyte count, RPI and RDW values ( table 1 ). No sta-tistical differences were found in OFT before incubation; however, after incubation, a significant decrease in OFT was observed. Significantly higher ferritin and vitamin B 12 serum levels were also found in CKD stage 5 pa-tients.

A statistically significant decrease in spectrin was found in CKD stage 5 patients, associated with a statisti-cally significant increase in bands 6 and 7 ( table 2 ). The ratio protein 4.1/spectrin was increased and the ratio spectrin/band 3 was decreased in CKD stage 5 patients. An inverse correlation between spectrin and RDW was also observed in CKD stage 5 patients ( fig. 1 ).

When comparing the haematological data between the two groups of CKD stage 5 patients, we found that non-responder patients were more anaemic, as shown by the significant decrease in Hb and haematocrit, and that RBCs were more hypochromic, as reflected by signifi-cantly lower values of MCH and MCHC; a trend to a low-er value of MCV and a significantly higher RDW, reflect-ing anisocytosis, was also observed in non-responders ( table 1 ). Even though the significant worsening of the



Costa et al.

Blood Purif 2008;26:267–273270

anaemia in non-responders, we did not find a statisti-cally significant increase for reticulocyte count and RPI in this group of CKD stage 5 patients. The OFT presented only significantly different values when performed after 24 h of incubation at 37 ° C, being decreased in non-re-sponder patients.

The erythrocyte membrane protein profile showed significantly lower values for ankyrin and ankyrin/band 3 ratio in responder CKD stage 5 patients. A trend to low-er values for spectrin ( table 2 ) and a statistically signifi-cant lower value for spectrin/ankyrin ratio was found in non-responder patients.

Table 1. Weekly rhEPO dose, haematological data, iron status and vitamin B12 serum levels for controls and CKD stage 5 patients

Controls(n = 26)

CKD stage 5 patients(n = 63)



Weekly rhEPO, IU/kg dose – 327.488281.37 89.65857.62 572.998193.84c

Hb, g/dl 14.1281.27 11.0881.67b 11.7781.41b 10.3781.63b, c

Haematocrit, % 43.5084.28 33.9084.76b 35.5183.92b 32.2380.50b, c

RBC, !1012/l 4.7280.59 3.6880.54b 3.7680.42b 3.5880.64b

MCV, fl 92.00 (90.00–94.00) 93.80 (90.00–98.20)a 95.80 (92.48–98.08)a 92.30 (85.40–100.30)MCH, pg 29.8381.39 30.1583.04 31.2981.53b 28.9783.73c

MCHC, g/dl 32.4780.58 32.0382.37 33.1681.77 30.8582.35a, c

RDW, % 12.7980.52 15.9282.56b 14.5681.23b 17.3282.83b, c

Reticulocytes, !109/l 33.57822.78 61.03831.36b 55.12830.98a 67.14831.06b

RPI 0.42 (0.19–0.66) 0.98 (0.58–1.40)b 1.08 (0.72–1.51)b 0.92 (0.52–1.24)a

OFT 0.458 (0.439–0.470) 0.452 (0.421–0.474) 0.458 (0.420–0.474) 0.438 (0.422–0.469)OFT with incubation 0.55380.051 0.51580.054a 0.53280.037 0.49880.063b, c


Transferrin saturation, % 19.44 (17.28–27.94) 21.80 (17.70–30.91) 24.20 (19.96–33.29) 22.10 (17.79–29.44)Vitamin B12, pg/ml 528.50 (429.20–664.50) 845.30 (651.60–1,134.00)b 851.60 (707.55–1,009.18)b 769.90 (571.70–1,166.00)b

Hb = Haemoglobin concentration; RBC = red blood cell count; MCV = mean cell volume; MCH = mean cell haemoglobin; MCHC = mean cell hae-moglobin concentration; RDW = red cell distribution width; RPI = reticulocyte production index; OFT = osmotic fragility test. Results are presented as mean 8 SD and as median (interquartile ranges).

a p < 0.05 vs. controls; b p < 0.001 vs. controls; c p < 0.05 vs. responders.

Table 2. RBC membrane protein profile for controls and CKD stage 5 patients

Controls(n = 26)

CKD stage 5 patients(n = 63)



Spectrin, % 27.63 (26.41–28.79) 24.27 (19.39–26.13)b 24.75 (22.38–26.63)a 22.35 (18.95–25.92)b

Ankyrin, % 6.9781.62 6.5381.90 6.0982.07 6.9781.60c

Band 3, % 38.5783.99 39.2984.03 39.9284.03 38.6583.70Protein 4.1, % 7.5681.45 7.2481.49 7.1881.33 7.3181.63Protein 4.2, % 5.5180.72 5.4481.44 5.5481.57 5.3581.29Band 5, % 6.8280.86 6.8781.03 6.7081.02 7.0481.00Band 6, % 5.1981.04 6.9881.37b 6.6181.30a 7.3781.32 b

Band 7, % 2.2080.65 3.3281.24b 3.1680.98 a 3.4981.43 b

Protein 4.1/spectrin 0.27680.624 0.33080.120a 0.31080.105 0.34080.130a

Protein 4.1/band 3 0.192 (0.154–0.227) 0.183 (0.155–0.208) 0.183 (0.154–0.208) 0.183 (0.159–0.205)Protein 4.2/band 3 0.149 (0.125–0.162) 0.138 (0.110–0.163) 0.135 (0.110–0.169) 0.142 (0.110–0.161)Spectrin/band 3 0.707 (0.649–0.822) 0.569 (0.512–0.686)b 0.572 (0.541–0.685)b 0.544 (0.486–0.687)b

Ankyrin/band 3 0.18580.585 0.16980.057 0.15580.060 0.18380.052c

Spectrin/ankyrin 4.1881.07 3.7781.84 4.4482.25 3.1080.94b, c

Results are presented as mean 8 SD and as median (interquartile ranges).a p < 0.05 vs. controls; b p < 0.001 vs. controls; c p < 0.05 vs. responders.




We found similar statistic differences between groups when CKD stage 5 patients with diabetic nephropathy were removed from the analyses. Actually, the values that we found for CKD stage 5 patients, with and without dia-betic nephropathy, were very similar (p 1 0.1 for all vari-ables).

Discussion

The severity of the anaemia presented by CKD stage 5 patients is mainly determined by the rate of renal insuf-ficiency and, therefore, by the reduction in the renal se-cretion of EPO.

It is accepted that the HD procedure is able to promote a complex biological response when the patient’s blood interacts with the artificial HD membranes [29–31] . A few reports also showed HD-induced changes in RBCs, namely in membrane protein composition and in eryth-rocyte osmotic fragility [32–34] , though most of these works quantified only some of the major RBC membrane proteins and the OFT was performed without blood in-cubation.

Besides the regular interaction of RBC with HD mem-branes, a sustained interaction of blood cells with higher concentrations of several circulating metabolites may also play a role in RBC survival and in leukocyte activa-tion. There is increasing evidence that oxygen metabo-

lites are involved in the progression of renal damage and uraemic symptoms [9, 11, 12] . The activation of leuko-cytes leads to the production and release of proteases and oxygen metabolites, both of which are able to induce ox-idative and proteolytic changes to RBCs [35, 36] .

In the present study, despite rhEPO therapy, anaemia was a consistent finding in our CKD stage 5 patients un-der HD ( table 1 ). This anaemia was associated with a sta-tistically significant increase in circulating reticulocytes, RPI, and RDW, reflecting the erythropoietic stimulus triggered by rhEPO therapy.

Some changes were observed in RBC membrane pro-tein composition of our CKD stage 5 patients, being the decrease in spectrin the most significant change ( table 2 ). This spectrin deficiency ( table 2 ) may account for a poor linkage of the cytoskeleton to the membrane, favouring membrane vesiculation, and, probably, a reduction in the RBC lifespan of CKD stage 5 patients [10, 13] . Significant increases were also observed in protein bands 6 and 7, which may further reflect an altered membrane protein interaction and destabilization of membrane structure. A primary protein deficiency, as occurs in hereditary sphe-rocytosis, may occur isolated or impose secondary defi-ciencies in one or more proteins [37] , being the type and amount of the secondary protein deficiencies involved in the haematological and clinical outcome of the disease [14] . It was reported that ‘minor’ RBC membrane proteins may play a role in membrane vesiculation, especially when associated to deficiencies in major RBC membrane proteins [14] .

The data suggestive of an altered membrane protein structure in CKD stage 5 patients was further strength-ened by the statistically significant changes observed for band 4.1/spectrin and spectrin/band 3 ratios ( table 2 ) and by the statistically significant negative correlation be-tween spectrin and RDW ( fig. 1 ).

Analyzing the results presented by the two studied groups of CKD stage 5 patients, responders and non-re-sponders to rhEPO therapy, we observed that the latter group presented a more severe anaemia ( table 1 ). The worsening of the anaemia was associated to a significant decrease in MCH and MCHC, showing a hypochromic RBC change. Considering that both groups were under iron and folate prophylactic therapies, and that iron sta-tus and vitamin B 12 serum levels were normal, these changes cannot be attributable to a deficiency in these erythropoietic nutrients. The lower MCHC observed in non-responders may explain the significant decrease ob-served for OFT with incubation.

10

15

20

25

30

35

12 14 16 18 20 22

RDW (%)

r = –0.364; p = 0.003

Sp

ec

trin

(%)

Fig. 1. Correlation of spectrin values with RDW in CKD stage 5 patients. RBC membrane proteins in HD patients.



Costa et al.

Blood Purif 2008;26:267–273272

A statistically significant increase in RDW was also observed in non-responders, which could reflect reticu-locytosis and/or anisocytosis due to morphologic chang-es in RBCs. As responders and non-responders presented with a similar reticulocyte value, the increase in RDW seems to mainly result from anisocytosis, which could reflect an altered and destabilized membrane structure. Non-responders presented with a decrease in spectrin, though without statistically significance, when compared to responders ( table 2 ). This spectrin deficiency seems to be linked to the production of hypochromic and aniso-cytic RBCs, and this anisocytosis could reflect an even more enhanced change in membrane protein interac-tions.

The large overlapping of spectrin values found in the two groups of patients suggests that the disease itself and/or to the interaction with HD membranes can be involved in RBC membrane protein alterations. However, non-re-sponders have more accentuated RBC membrane protein alterations, which may be due to a higher RBC metabolic stress. Inflammation could be one of the metabolic fac-tors involved; in fact, our non-responders CKD stage5 patients presented significantly increased levels of C-reactive protein (data not shown). We should also notexclude the hypothesis that the much higher doses of rhEPO used in non-responders can also be involved in RBC membrane protein alterations.

The decrease for ankyrin and ankyrin/band 3 ratio found in responders patients, as compared with non-re-sponders and controls, suggests that probably all CKD stage 5 patients have low ankyrin levels. This reduction was not detectable in non-responder patients probably because they presented a higher number of reticulocytes, though without statistical significance. Actually, it is known that reticulocytes present higher levels of ankyrin in their membranes, masking, therefore, the reduction in ankyrin [38] .

As far as we know, there are no reports in the literature regarding RBC membrane protein analysis associated to the use of high-flux polysulfone FX-class dialysers of Fre-senius. Studies using cuprophane and polyacrylonitrile dialysis membranes reported a reduction in both spec-trin and band 3, and an isolated reduction in band 3, re-spectively [32] . These alterations were associated with the type of dialysis membranes used and to a loss of RBC membrane during HD procedure, due to mechanical or chemical damages [32, 39] . In the present study, we can-not exclude a possible interaction of RBCs with the dialy-sis membranes, though the enhancement of the RBC membrane destabilization observed in non-responders, undergoing the same HD protocol, suggests an involve-ment of the pathology in the RBC membrane changes. Or even the involvement of the higher dose of rhEPO used in non-responder patients.

In conclusion, our data suggest that CKD stage 5 is as-sociated with an altered protein membrane structure, which may be due to the disease itself and/or the interac-tion with HD membranes. Further studies are warranted, namely in CKD stage 5 patients not yet on hemodialysis and/or on rhEPO therapy, to clarify the mechanism un-derlying rhEPO resistance, which seems to be linked to an even more destabilized RBC membrane structure. Moreover, studies about membrane vesiculation would be helpful in clarifying this mechanism.

Acknowledgements

We are very grateful to FMC, Dinefro – Diálises e Nefrologia, SA and Uninefro – Sociedade Prestadora de Cuidados Médicos e de Diálise, SA, and to their nurses for the technical support. This study was supported by a PhD grant (SFRH/BD/27688/2006) at-tributed to E. Costa by FCT and FSE.

References


2 Foley RN, Parfrey PS, Harnett JD, Kent GM, Murray DC, Barre PE: The impact of anemia on cardiomyopathy morbidity and mortality in end-stage renal disease.Am J Kidney Dis 1996; 28: 53–61.


4 Nissenson AR, Nimer SD, Wolcott DL: Re-combinant human erythropoietin and renal anemia: molecular biology, clinical efficacy, and nervous system effects. Ann Intern Med 1991; 114: 402–416.

5 Miyoshi I, Saito T, Bandobashi K, Ohtsuki Y, Taguchi H Concomitant deposition of alu-minum and iron in bone marrow trabeculae. Intern Med 2006; 45: 117–118.

6 Pisoni RL, Bragg-Gresham JL, Young EW, Akizawa T, Asano Y, Locatelli F, Bommer J, Cruz JM, Kerr PJ, Mendelssohn DC, Held PJ, Port FK: Anemia management and outcomes from 12 countries in the dialysis outcomes and practice patterns study (DOPPS). Am J Kidney Dis 2004; 44: 94–111.




7 Medina A, Ellis C, Levitt MD: Use of alveolar carbon monoxide measurements to assess red blood cell survival in hemodialysis pa-tients. Am J Hematol 1994; 46: 91–94.

8 Bonomini M, Sirolli V, Reale M, Arduini A: Involvement of phosphatidylserine exposure in the recognition and phagocytosis of ure-mic erythrocytes. Am J Kidney Dis 2001; 37: 807–814.

9 Scott Brimble K, McFarlane A, Winegard N, Crowther M, Churchill DN: Effect of chron-ic kidney disease on red blood cell rheology. Clin Hemorheol Microcirc 2006; 34: 411–420.

10 Reliene R, Marini M, Zanella A, Reinhart WH, Ribeiro ML, del Giudice EM, Perrotta S, Ionoscon A, Eber S, Lutz HU: Splenectomy prolongs in vivo survival of erythrocytes dif-ferently in spectrin/ankyrin- and band 3-de-ficient hereditary spherocytosis. Blood 2002; 100: 2208–2215.

11 Lucchi L, Bergamini S, Iannone A, Perrone S, Stipo L, Olmeda F, Caruso F, Tomasi A, Albertazzi A: Erythrocyte susceptibility to oxidative stress in chronic renal failure pa-tients under different substitutive treat-ments. Artif Organs 2005; 29: 67–72.

12 Stoya G, Klemm A, Baumann E, Vogelsang H, Ott U, Linss W, Stein G: Determination of autofluorescence of red blood cells (RbCs) in uremic patients as a marker of oxidative damage. Clin Nephrol 2002; 58: 198–204.

13 Gallagher PG: Red cell membrane disorders. Hematology Am Soc Hematol Educ Pro-gram 2005: 13–18.

14 Rocha S, Rebelo I, Costa E, Catarino C, Belo L, Castro EMB, Cabeda JM, Barbot J, Quin-tanilha A, Santos-Silva A: Protein deficiency balance as a predictor of clinical outcome in hereditary spherocytosis. Eur J Haematol 2005; 74: 374–380.

15 Smrzova J, Balla J, Bárány P: Inflammation and resistance to erythropoiesis-stimulting agents – what do we know and what needs to be clarified? Nephrol Dial Transplant 2005; 20(suppl 8):viii2–viii7.

16 Locatelli F, Pisoni RL, Combe C, Bommer J, Andreucci VE, Piera L, Greenwood R, Feld-man HI, Port FK, Held PJ: Anaemia in hae-modialysis patients of five European coun-tries: association with morbidity and mortality in the Dialysis – outcomes and Pracice Patterns Study (DOPPS). Nephrol Dial Transplant 2004; 19: 121–132.

17 Macdougall IC, Cooper AC: Erythropoietin resistance: the role of inflammation andpro-inflammatory cytokines. Nephrol Dial Transplant 2002; 17(suppl 11):39–43.



20 Sommerburg O, Grune T, Hampl H, Riedel E, van Kuijk FJ, Ehrich JHH, Siems WG: Does long-term treatment of renal anaemia with recombinant erythropoietin influence oxidative stress in haemodialysed patients? Nephrol Dial Transplant 1998; 13: 2583–2587.

21 Locatelli F, Aljama P, Barany P, Canaud B, Carrera F, Eckardt KU, Horl WH, Macdou-gal IC, Macleod A, Wiecek A, Cameron S; European Best Practice Guidelines Working Group. Revised European best practice guidelines for the management of anaemia in patients with chronic renal failure. Nephrol Dial Transplant 2004; 19(suppl 2):ii1–ii47.

22 Brown CD, Ghali HS, Zhao Z, Thomas LL, Friedman EA: Association of reduced red blood cell deformability and diabetic ne-phropathy. Kidney Int 2005; 67: 295–300

23 Hillman RS, Ault KA: Clinical approach to anemia; in: Hematology in Clinical Practice, ed 3. New York, McGraw-Hill, 2002, pp 12–26.

24 Roper D, Layton M, Mitchell Lewis S: Inves-tigation of the hereditary haemolytic anemi-as: membrane and enzyme abnormalities; in Lewis SM, Bain BJ, Bates I (eds): Dacie and Lewis Practical Haematology. London, Churchill Livingstone, 2001, pp 167–198.

25 Dodge JT, Mitchell C, Hanahan DJ: The preparation and chemical characteristics of haemoglobin-free ghosts of human erythro-cytes. Arch Biochem Biophys 1963; 100: 119–130.

26 Bradford MM: A rapid and sensitive method for the quantification of microgram quanti-ties od protein utilizing the principle of the protein-dye binding. Anal Biochem 1976; 72: 248–254.

27 Laemmli UK: Cleavage of structural pro-teins during the assembly of the head of the bacteriophage T. Nature 1970; 227: 680–685.

28 Fairbanks G, Steck TL, Wallach DFH: Elec-trophoresis of the major polypeptides of the human erythrocyte membrane. Biochemis-try 1971; 10: 2606–2616.

29 Levin RD, Kwaan HC, Ivanivich P: Changes in platelet functions during hemodialysis. J Lab Clin Med 1978; 92: 779–786.

30 Craddock PR, Fehr J, Brigham KL, Kronem-berg RS, Jacob HS: Complement and leuko-cyte mediated pulmonary dysfunction in he-modialysis. N Eng J Med 1977; 296: 769–774.

31 Craddock PR, Hammerschmidt DE, White JG, Dalmasso AP, Jacob HS: Complement (C5a)-induced granulocyte aggregation in vitro: a possible mechanism of complement-mediated leukostasis and leukopenia. J Clin Invest 1977; 60: 260–264.

32 Martos MR, Hendry BM, Rodrígues-Puyol M, Dwight J, Díez-Marqués ML, Rodríguez-Puyol D: Haemodialyser biocompatibility and erythrocyte struture and function. Clin Chim Acta 1997; 265: 235–246.

33 Ibrahim FF, Ghannam MM, Ali FM: Effect of dialysis on erythrocyte membrane of chronically hemodialyzed patients. Renal failure 2002; 24: 779–790.

34 Wu SG, Jeng FR, Wei SY, Su CZ, Chung TC, Chang WJ, Chang HW: Red blood cell os-motic fragility in chronically hemodialyed. Nephron 1998; 78: 28–32.

35 Santos-Silva A, Castro EMB, Teixeira NA, Guerra FC, Quintanilha A: Erythrocyte membrane band 3 profile imposed by cellu-lar aging, by activated neutrophils and by neutrophilic elastase. Clin Chim Acta 1998; 275: 185–196.

36 Santos-Silva A, Castro EM, Teixeira NA, Guerra FC, Quintanilha A: Erythrocyte membrane band 3 profile imposed by cellu-lar aging, by activated neutrophils and by neutrophilic elastase. Clin Chim Acta 1998; 275: 185–196.

37 Margetis P, Antonelou M, Karababa F, Loutradi A, Margaritis L, Papassideri I: Physiologically important secondary modi-fications of red cell membrane in hereditary spherocytosis – evidence for in vivo oxida-tion and lipid rafts protein variations. Blood Cells Mol Dis 2007; 38: 210–220.

38 Giudice EM, Francese M, Polito R, Nobili B: Apparently Normal Ankyrin Content in Un-splenectomized Hereditary Spherocytosis Patients with the Inactivation of one An-kyrin (ANK1) Allele. Haematologica 1997; 82: 332–333.

39 Sevillano G, Rodrígues-Puyol M, Martos R, Duque I, Lamas S, Diez-Marques ML, Lúcio J, Rodriguez-Puvol D: Cellulose acetato membrane improves some aspects of red blood cell function in hemodialysis patients. Nephrol Dial Transplant 1990; 5: 497–499.



Paper VIII

“Changes in red blood cells membrane protein composi-

tion during haemodialysis procedure”

Renal Failure, in press.



Renal Failure, 30:1–5, 2008 Copyright © Informa Healthcare USA, Inc.ISSN: 0886-022X print / 1525-6049 onlineDOI: 10.1080/08860220802422036

1

LRNFCLINICAL STUDY

Changes in Red Blood Cells Membrane Protein Composition during Hemodialysis Procedure

RBC Membrane Protein Alterations during HemodialysisElísio CostaFaculdade de Farmácia, Serviço de Bioquímica, Universidade do Porto, Porto, Portugal; Instituto de Biologia Molecular e Celular, Universidade do Porto, Porto, Portugal; Escola Superior de Saúde, Instituto Politécnico de Bragança

Susana RochaFaculdade de Farmácia, Serviço de Bioquímica, Universidade do Porto, Porto, Portugal; Instituto de Biologia Molecular e Celular, Universidade do Porto, Porto, Portugal

Petronila Rocha-PereiraUniversidade da Beira Interior, Covilhã

Elisabeth CastroFaculdade de Farmácia, Serviço de Bioquímica, Universidade do Porto, Porto, Portugal; Instituto de Biologia Molecular e Celular, Universidade do Porto, Porto, Portugal

Vasco Miranda and Maria do Sameiro FariaFMC, Dinefro—Diálises e Nefrologia, SA

Alfredo LoureiroUninefro—Sociedade Prestadora de cuidados Médicos e de Diálise, SA

Alexandre QuintanilhaInstituto de Biologia Molecular e Celular, Universidade do Porto, Porto, Portugal; Instituto de Ciências Biomédicas Abel Salazar (ICBAS), Universidade do Porto, Porto, Portugal

Luís Belo and Alice Santos-SilvaFaculdade de Farmácia, Serviço de Bioquímica, Universidade do Porto, Porto, Portugal; Instituto de Biologia Molecular e Celular, Universidade do Porto, Porto, Portugal

Our aim was to evaluate the influence of the hemodialysis(HD) procedure in red blood cells (RBC) membrane proteincomposition. We evaluated hematological data (RBC count,hemoglobin concentration, and hematimetric indices) and RBCmembrane protein composition (linear and exponential gradient

polyacrylamide gel electrophoresis in the presence of sodiumdodecylsulfate [SDS-PAGE] followed by densitometry analysisof RBC membrane proteins) before and immediately after theHD procedure in 20 patients (10 responders and 10 non-respondersto recombinant human erythropoietin therapy [rhEPO]) and26 healthy controls. Before HD, patients presented anaemia andsignificant changes in membrane protein composition, namely, astatistically significant reduction in spectrin associated with asignificant increase in bands 6, as well as an altered membraneprotein interaction (protein 4.1/spectrin, protein 4.1/band 3, protein4.2/band 3 and spectrin/band 3). After HD, we found that patientsshowed a statistically significant increase in RBC count andhemoglobin, a further and statistically significant decrease inspectrin, an increase in band 3, and an altered spectrin/band 3 ratio.

Received 20 May 2008; revised 13 August 2008; accepted20 August 2008.

Address correspondence to Elísio Costa, Serviço de Bioquímica,Faculdade de Farmácia, Universidade do Porto, Rua Aníbal Cunha,164, 4099-030 Porto, Portugal; E-mail: elisio_ [email protected]

1

5

5

10

15

20

2530

35

40



2 E. Costa et al.

When comparing responders and non-responders patients afterHD, we found that the non-responders presented a trend to ahigher reduction in spectrin. Our data suggest that HD procedureseems to contribute to a reduction in spectrin, which is normallyassociated with a reduction in RBC deformability, being thatreduction in spectrin is higher in non-responder patients.

Keywords chronic kidney disease, chronic renal failure,hemodialysis, spectrin, erythrocyte membraneproteins, resistance to rhEPO

INTRODUCTION

The red blood cell (RBC) membrane is a complexstructure comprising a lipidic bilayer, transmembrane orintegral proteins, and peripheral proteins, including thecytoskeleton proteins. Modifications in RBC membraneprotein composition may account for changes in thedeformability of the cell, compromising its circulation inthe microvasculature and its survival.[1–3]

Spectrin is the major protein of the cytoskeleton, andtherefore most responsible for RBC shape, integrity, anddeformability. It links the cytoskeleton to the lipid bilayerby vertical protein interactions with the transmembraneproteins, band 3, and glycophorin C.[2,3] In the verticalprotein interaction of spectrin with band 3 are alsoinvolved ankyrin (also known as band 2.1) and protein 4.2.A normal linkage of spectrin with the other proteins of thecytoskeleton assures normal horizontal protein interac-tions, and its linkage with the transmembrane proteinsassures normal vertical interactions.[2,3]

Anaemia is a common complication in hemodialysis(HD) patients, due mainly to a failure in erythropoietinkidney production.[4] Moreover, the lifespan of RBCs ofHD patients being shortened is an additional cause ofanaemia in these patients.[5,6]

We recently reported[7] that HD patients presentchanges in RBC membrane protein composition, thedecrease in spectrin being the most significant change.This spectrin deficiency may account for a poorer link-age of the cytoskeleton to the membrane, favoringmembrane vesiculation and probably a reduction in theRBC lifespan of HD patients. We also found that HDpatient non-responders to recombinant human erythro-poietin (rhEPO) therapy, receiving high doses of rhEPOcompared to responder’s patients, presented a trend tolower values of spectrin, though without statisticallysignificance, and a significant increase in ankyrin thatresulted in a significant disturbance in horizontal andvertical protein interactions, as suggested by the alteredankyrin/band 3 and spectrin/ankyrin ratios. We havehypothesized that alterations in RBC protein membrane

structure in HD patients could be related to the chronicrenal failure and/or to the HD procedure itself, and thatin non-responders, it could be also related to the higherdoses of rhEPO received by these patients. The aim ofour work was to evaluate the effect of HD procedure inRBC membrane protein composition in HD patients,including responders and non-responders to rhEPOtherapy.

MATERIALS AND METHODS

Subjects

We studied 20 HD patients (8 males, 12 females;mean age 60.3 ± 16.7 years) under rhEPO treatment,including 10 responders and 10 non-responders to rhEPOtherapy. Classification of the patients, as responders ornon-responders, was performed in accordance with theEuropean Best Practice Guidelines,[8] which definesresistance to rhEPO as a failure to achieve targethemoglobin (Hb) levels (between 11 and 12 g/dL) withmaintained doses of rhEPO higher than 300 IU/Kg/weekof epoetin or 1.5 μg/Kg/week of darbepoetin-α. Thetwo groups of patients were matched for age, gender,weight, body mass index, mean time under HD, ureareduction ratio, urea Ktv, and parathyroid hormoneserum levels.

HD patients were under therapeutic HD three timesper week, for 3 to 5 h, for a median period of time of41 months. All patients used the high-flux polysulfoneFX-class dialyzers of Fresenius, 12 with FX60 and 8 withFX80 dialyzer type. The causes of renal failure in patient’spopulation were as follows: diabetic nephropathy (n = 5),chronic glomerulonephritis (n = 3), hypertensive nephro-sclerosis (n = 2), obstructive nephropathy (n = 1), neph-rolithiasis (n = 1), chronic interstitial nephritis (n = 1), andchronic renal failure of uncertain etiology (n = 7). Patientswith autoimmune disease, malignancy, hematologicaldisorders, and acute or chronic infection were excluded.All patients gave their informed consent to participate inthis study.

Besides rhEPO therapy, all patients were under ironand folate prophylactic therapies, in accordance to therecommendations of European Best Practice Guidelines,[8]

to avoid nutrient erythropoietic deficiencies.The control group included 26 healthy volunteers

presenting normal hematological and biochemical values,with no history of renal or inflammatory diseases, and, asfar as possible, age- and gender-matched with HDpatients. Controls did not receive any medication knownto interfere with the studied variables.

50

55

60

65

70

75

80

85

90

95

100

105

110

115

120

125

130

135

140

45



RBC Membrane Protein Alterations during Hemodialysis 3

Assays

Blood samples (using EDTA as anticoagulants) weredrawn from fasting controls or before and immediately afterthe second dialysis session of the week in HD patients. Weevaluated hematological data (RBC count, Hb concentra-tion, and hematimetric indices) by using an automaticblood cell counter (Sysmex K1000; Sysmex, Hamburg,Germany) and RBC membrane protein composition usinglinear and exponential gradient polyacrylamide gel elec-trophoresis in the presence of sodium dodecylsulfate(SDS-PAGE) followed by development with Coomassiebrilliant blue R-250 (Sigma) and densitometric analysis, asreported previously.[7]

Data Analysis

For statistical analysis, we used the Statistical Packagefor Social Sciences (SPSS version 15.0 for Windows,SPSS, Inc., Chicago, Illinois, USA). Kolmogorov Smirnovstatistics were used to evaluate sample normality distribution.

Multiple comparisons between groups were performedby one-way ANOVA supplemented with Turkey’s HSDpost-hoc test. For comparing data, before and after HDprocedure, paired-samples T-test or Wilcoxon test wereused. Significance was accepted at p less than 0.05.

RESULTS

The results were analyzed in order to study thedifferences between healthy controls and HD patients,to study changes imposed by the HD procedure, andto study the differences between responders and non-responders HD patients. In Table 1, we present thehematological data and RBC membrane protein profilefor controls and HD patients (before and immediatelyafter HD).

Before HD, the patients presented anaemia, as shownby the significantly decreased RBC count and Hb concen-tration, with a significant increase in the mean cell volume(MCV), mean cell Hb (MCH), and RBC distributionwidth (RDW); statistically significant changes in membrane

Table 1 Hematological data and RBC membrane protein profile for controls and HD patients

Controls (n = 26)

HD Patients (n = 20)

Before HD After HD

Spectrin/ankirin 4.18 ± 1.07 4.48 ± 1.361 4.45 ± 1.49Hb (g/dL) 14.12 ± 1.27 11.94 ± 1.67* 12.98 ± 2.20*†

RBC (×1012/L) 4.72 ± 0.59 3.78 ± 0.61* 4.13 ± 0.84†

MCV (fl) 92.00 (90.00–94.00) 93.00 (90.25–98.00)* 93.35 (93.00–98.00)*MCH (pg) 29.83 ± 1.39 31.83 ± 2.05* 31.60 ±1.94*MCHC (g/dL) 32.47 ± 0.58 32.39 ± 0.71 32.13 ± 0.79RDW (%) 12.79 ± 0.52 14.84 ± 1.11* 14.77 ± 1.19*Spectrin (%) 27.63 (26.41–28.79) 25.58 (24.10–27.07)* 24.47 (22.31–26.95)*†

Ankyrin (%) 6.97 ± 1.62 6.39 ± 1.55 6.23 ± 1.28Band 3 (%) 38.57 ± 3.99 38.10 ± 3.78 41.13 ± 2.44*†

Protein 4.1 (%) 7.56 ± 1.45 6.48 ± 1.60 6.39 ± 1.69Protein 4.2 (%) 5.51 ± 0.72 4.34 ± 0.99 4.84 ± 1.04Band 5 (%) 6.82 ± 0.86 6.56 ± 0.91 6.71 ± 0.59Band 6 (%) 5.19 ± 1.04 6.46 ± 0.87* 6.17 ± 1.15*Band 7 (%) 2.20 ± 0.65 2.09 ± 0.43 2.37 ± 0.34Protein 4.1/spectrin 0.276 ± 0.624 0.243 ± 0.070* 0.251 ± 0.081*Protein 4.1/band 3 0.192 (0.154–0.227) 0.170 (0.138–0.206)* 0.163 (0.121–0.202)*Protein 4.2/band 3 0.149 (0.125–0.162) 0.114 (0.101–0.133)* 0.118 (0.101–0.147)*Spectrin/band 3 0.707 (0.649–0.822) 0.685 (0.626–0.796)* 0.647 (0.566–0.689)*†

Ankyrin/band 3 0.185 ± 0.585 0.171 ± 0.049 0.152 ± 0.330*

*p < 0.05 vs. controls, †p < 0.05 vs. before HD. Results are presented as mean ± standard deviation and asmedian (interquartile ranges).

Abbreviations: Hb = hemoglobin concentration, RBC = red blood cell count, MCV = mean cell volume,MCH = mean cell hemoglobin, MCHC = mean cell hemoglobin concentration, RDW = red cell distribution width.

145

150

155

160

165

170

175

180



4 E. Costa et al.

protein composition were observed, namely, a reduction inspectrin and an increase in band 6. Though only slightchanges were observed in the other proteins, the ratiosbetween them, reflecting altered horizontal and verticalprotein interactions—protein 4.1/spectrin, protein 4.1/band 3,protein 4.2/band 3 and spectrin/band 3—were observed.

After dialysis, HD patients showed a statistical signifi-cant increase in RBC count and Hb concentration; nostatistical significant changes were found in the other hemato-logical parameters (see Table 1). Concerning the RBC mem-brane protein profile, a statistically significant decrease inspectrin, an increase in band 3, and an altered spectrin/band3 ratio were found after HD procedure. We also observedafter HD some trends toward the control profile (band 6 andband 7) and some trends in opposition to control profile(a further decrease in ankyrin, protein 4.1, protein 4.2, andband 5). We must notice that band 3 increased and reached astatistically significantly higher value than that presented bythe control. Most of the ratios, reflecting protein interactions,presented values that were still different from those of thecontrol.

When comparing the two groups of HD patients(responders and non-responders), we did not find signifi-cant changes in membrane protein profile, excepting thatnon-responders presented a trend (p = 0.124) to a higherreduction in spectrin membrane content (see Figure 1).

DISCUSSION

It is accepted that the HD procedure is able to pro-mote a complex biological response when the patient’sblood interacts with the artificial HD membranes.[9–13]

However, the effect of HD procedure in RBC membraneprotein content has not been adequately defined.

In the present study, despite rhEPO therapy,anaemia was a consistent finding in our HD patients.This anaemia was associated with a statistically signifi-cant increase in RDW, suggesting the presence ofanisocytic RBCs. Higher Hb levels and RBC countswere found after HD. This increase in circulating RBCshas been described[14,15] to be associated with a translo-cation of RBCs from the splanchnic circulation (andpossibly from the splenic circulation) in order tocompensate the hypovolemic stress during dialysisultrafiltration.

The changes observed in the RBC membrane proteincomposition of HD patients (see Table 1), when comparedto controls, were similar to those we have recentlyreported,[7] namely, a statistically significant reduction inspectrin and in band 6 (no significant change was observedfor band 7). The lower number of patients enrolled inthe present study may explain the slighter differences inprotein profile and in the ratios reflecting membraneprotein interactions. We still found in the present studydisturbed horizontal (protein 4.1/spectrin) and vertical(protein 4.1/band 3; protein 4.2/band 3; spectrin/band 3)interactions, suggesting a disturbance in membranedeformability and integrity. Spectrin deficiency mayaccount for a poorer linkage of the cytoskeleton to themembrane, favoring membrane vesiculation and probablya reduction in the RBC lifespan of HD patients.[2,16] Theincrease in protein band 6 may further reflect an alteredmembrane protein interaction and destabilization of mem-brane structure.

When studying the differences in membrane proteincomposition imposed by the HD procedure, the resultsshowed some trends toward a control pattern profile forsome of the membrane proteins—bands 3, 6, and 7. Spec-trin showed an even lower value after HD, and ankyrin,protein 4.1, protein 4.2, and band 5 also present a trendtoward decreasing. The analysis of the ratios reflectingthe protein interactions remained altered, when comparedto controls; however, when comparing the ratios beforeand after HD, only the ratio spectrin/band 3 showed astatistically significant value, reflecting a vertical membraneprotein disturbance. However, no statistically significantcorrelation between the RBC membrane protein alterationsand the RBC counts and hemoglobin levels were found,which can be due to the lower patients number included inthis study.

Moreover, we also found that this spectrin reductionis enhanced in the non-responder group (see Figure 1),which can be associated with an enhanced inflammationprocess. Indeed, results from our and other groups[17]

have demonstrated that HD and particularly resistance to

Figure 1. Spectrin reduction (%) in responders and non-responders HD patients.

185

190

195

200

205

210

215

220

225

230

235

240

245

250

255

260



RBC Membrane Protein Alterations during Hemodialysis 5

rhEPO therapy are associated with a pronounced inflam-matory response, demonstrated by higher levels of C-reactiveprotein (data not shown). However, no statistically signif-icant differences were found between C-reactive proteinlevels and RBC membrane protein modifications foundduring HD procedure. We also should not exclude thepossibility that the much higher doses of rhEPO used innon-responders, can be also involved in RBC membraneprotein alterations.

In conclusion, the HD procedure seems to contributeto a reduction in spectrin, which is normally associatedwith a reduction in RBC deformability, being enhanced innon-responder patients. Therefore, we cannot exclude thatthe inflammatory product, being enhanced in HD patients(especially in non-responders) may play a role in RBCmembrane protein disturbance.

ACKNOWLEDGMENTS

We are very grateful to FMC, Dinefro—Diálises eNefrologia, SA, and Uninefro— Sociedade Prestadora deCuidados Médicos e de Diálise, SA, and to their nurses forthe technical support. This study was supported by a PhDgrant (SFRH/BD/27688/2006) attributed to E. Costa byFCT and FSE.

DECLARATION OF INTEREST

The authors report no conflicts of interest. Theauthors alone are responsible for the content and writing ofthe paper.

REFERENCES

1. Reliene R, Marini M, Zanella A, Reinhart WH, Ribeiro ML,del Giudice EM, Perrotta S, Ionoscon A, Eber S, Lutz HU.Splenectomy prolongs in vivo survival of erythrocytes dif-ferently in spectrin/ankyrin- and band 3-deficient hereditaryspherocytosis. Blood. 2002;100:2208–2215.

2. Gallagher PG. Red cell membrane disorders: Hematology.Am Soc Hematol Educ Program. 2005:13–18.

3. Rocha S, Rebelo I, Costa E, Catarino C, Belo L, CastroEMB, Cabeda JM, Barbot J, Quintanilha A, Santos-Silva A.Protein deficiency balance as a predictor of clinical out-come in hereditary spherocytosis. Eur J Hematol. 2005;74:374–380.

4. Locatelli F, Conte F, Marcelli D. The impact of hematocritlevels and erythropoietin treatment on overall and cardiovas-cular mortality and morbidity—the experience of the Lom-bardy dialysis registry. Nephrol Dial Transplant. 1998;13:1642–1644.

5. Lucchi L, Bergamini S, Iannone A, Perrone S, Stipo L,Olmeda F, Caruso F, Tomasi A, Albertazzi A. Erythrocytesusceptibility to oxidative stress in chronic renal failurepatients under different substitutive treatments. Artif Organs.2005;29:67–72.

6. Stoya G, Klemm A, Baumann E, Vogelsang H, Ott U, LinssW, Stein G. Determination of autofluorescence of red bloodcells (RbCs) in uremic patients as a marker of oxidativedamage. Clin Nephrol. 2002;58:198–204.

7. Costa E, Rocha S, Rocha-Pereira P, Castro E, Miranda V,Sameiro-Faria M, Loureiro A, Quintanilha A, Belo L,Santos-Silva A. Altered erythrocyte membrane proteincomposition in chronic kidney disease stage 5 patients underhemodialysis and recombinant human erythropoietin therapy.Blood Purif. 2008;26:267–273.

8. Locatelli F, Aljama P, Barany P, Canaud B, Carrera F,Eckardt KU, Horl WH, Macdougal IC, Macleod A, Wiecek A,Cameron S; European Best Practice Guidelines WorkingGroup. Revised European best practice guidelines for the man-agement of anaemia in patients with chronic renal failure. Neph-rol Dial Transplant. 2004;19 (Suppl. 2): ii1–ii47.

9. Levin RD, Kwaan HC, Ivanivich P. Changes in platelet func-tions during hemodialysis. J Lab Clin Med. 1978;92: 779–786.

10. Craddock PR, Fehr J, Brigham KL, Kronemberg RS, JacobHS. Complement and leukocyte mediated pulmonary dys-function in hemodialysis. N Eng J Med. 1977;296:769–774.

11. Craddock PR, Hammerschmidt DE, White JG, Dalmasso AP,Jacob HS. Complement (C5a)-induced granulocyte aggregationin vitro: A possible mechanism of complement-mediated leuko-stasis and leukopenia. J Clin Invest. 1977;60: 260–264.

12. Dolegowska B, Kwiatkowska E, Wesolowska T, Bober J,Chlubek D, Ciechanowski K. Effect of hemodialysis on the con-tent of fatty acids in monolayers of erythrocyte membranes inpatients with chronic renal failure. Ren Fail. 2007; 29:447–452.

13. Brimble KS, McFarlane A, Winegard N, Crowther M,Churchill DN. Effect of chronic kidney disease on red blood cellrheology. Clin Hemorheol Microcirculation. 2006;34: 411–420.

14. Dasselaar JJ, Hooge MNL, Pruim J, Nijnuis H, Wiersum A,Jong PE, Huisman RM, Franssen CFM. Relative bloodvolume changes underestimated total blood volumechanges during hemodialysis. Clin J Am Soc Nephrol.2007; 2:669–674.

15. Yu AW, Nawab ZM, Barnes WE, Lai KN, Ing TS,Daugirdas JT. Splanchnic erythrocyte content decreasesduring hemodialysis: A new compensatory mechanism forhypovolemia. Kidney Int. 1997;51:1986–1990.

16. Reliene R, Marini M, Zanella A, Reinhart WH, Ribeiro ML,del Giudice EM, Perrotta S, Ionoscon A, Eber S, Lutz HU.Splenectomy prolongs in vivo survival of erythrocytesdifferently in spectrin/ankyrin- and band 3-deficient hereditaryspherocytosis. Blood. 2002;100:2208–2215.

17. Costa E, Lima M, Alves JM, Rocha S, Rocha-Pereira P,Castro E, Miranda V, Sameiro-Faria M, Loureiro A,Quintanilha A, Belo L, Santos-Silva A. Inflammation, T-cellphenotype and inflammatory cytokines in chronic kidneydisease patients under hemodialysis and its relationship toresistance to recombinant human erythropoietin therapy. JClin Immunol., in press.

5

3 4

2

265

270

275

280

285

290

295

300

305

310

315

320

325

330

335

340

345

350

355

360



Paper IX

“Band 3 profile as a marker of erythrocyte changes in

chronic kidney disease patients”

The Open Clinical Chemistry Journal 2008; 1: 57-63.



The Open Clinical Chemistry Journal, 2008, 1, 57-63 57

1874-2416/08 2008 Bentham Open

Open Access

Band 3 Profile as a Marker of Erythrocyte Changes in Chronic Kidney Disease Patients

Elísio Costa1,2,3,*, Susana Rocha

2,3, Petronila Rocha-Pereira

3,4, Elisabeth Castro

2,3, Vasco Miranda

5,

Maria do Sameiro Faria5, Alfredo Loureiro

6, Alexandre Quintanilha

3,7, Luís Belo

2,3 and Alice

Santos-Silva2,3

1Instituto de Ciências da Saúde, Universidade Católica Portuguesa;

2Faculdade de Farmácia da Universidade do

Porto; 3Instituto de Biologia Molecular e Celular da Universidade do Porto (IBMC);

4Centro Investigação Ciências

Saúde, Universidade Beira Interior, Covilhã; 5FMC, Dinefro - Diálises e Nefrologia, SA;

6Uninefro � Sociedade

Prestadora de Cuidados Médicos e de Diálise, SA and 7Instituto de Ciências Biomédicas Abel Salazar da Universidade

do Porto (ICBAS)

Abstract: Our aim was to study changes in red blood cell (RBC) membrane band 3 profile, as a cumulative marker of

RBC changes, in chronic kidney disease (CKD) patients under haemodialysis and recombinant human erythropoietin

(rhEPO) therapy and its linkage with resistance to this therapy.

We studied 63 CKD patients, 32 responders and 31 non-responders to rhEPO therapy, and 26 healthy individuals. We

evaluated the band 3 profile [% of band 3 monomer, high molecular weight aggregates (HMWAg), and proteolytic frag-

ments (Pfrag)], membrane-bound haemoglobin (MBH), haematological data, total serum bilirubin, glutathione peroxidase

(GPx) and superoxide dismutase activities, total antioxidant status (TAS) and plasma lipid peroxidation (TBA). Compared

to controls, band 3 profile presented by CKD patients showed statistically significant lower HMWAg and Pfrag values

and a significant higher value in band 3 monomer. GPx, TBA and TAS activities, and TBA/TAS ratio were also signifi-

cantly higher in CKD patients. Comparing responders to non-responders CKD patients, significantly lower value in Pfrag

and a trend for a higher value in MBH were found in non-responders.

Our data suggest that CKD patients present younger RBC population, which could be related to the rhEPO therapy. The

adverse plasma environment associated to CKD patients under hemodialysis imposes changes in band 3 profile, particu-

larly in non-responders, suggesting that resistance to rhEPO therapy in CKD patients seems to be associated to an increase

in RBC damage.

Key Words: Band 3 protein, membrane bound haemoglobin, oxidative stress, RBC, haemodialysis, resitance to rhEPO therapy.

INTRODUCTION

Haemodialysis increases longevity of patients with end-stage renal disease by removing the metabolic end products and excess of water. Despite the technologic developments of haemodialysis procedures and of medical support in the last years, the morbidity and mortality of these patients re-main high, about 10 to 20 times higher than that found in general population [1-3]. Anaemia is a frequent complication associated with chronic kidney disease (CKD), and is mainly due to insufficient erythropoietin renal production.

Until 18 years ago, the treatment of anaemia of CKD was blood transfusion; however, the management of this anaemia has been improved by the introduction of recombinant hu-man EPO (rhEPO) therapy. This therapy allowed a signifi-cant reduction in the associated adverse effects of anaemia and improved patient�s quality of life. There is, however, a

*Address correspondence to this author at the Serviço de Bioquímica,

Faculdade de Farmácia da Universidade do Porto, Rua Aníbal Cunha, 164,

4099-030 Porto, Portugal; E-mail: [email protected]

marked variability in the sensitivity to rhEPO, with up to 10-fold variability in dose requirements to achieve correction of the anaemia [4-10].

Although oxygen is essential to human life, paradoxi-cally, it can also cause great harm via oxidation/reduction reactions and formation of reactive oxygen species (ROS). Oxidative stress and its related biological effects have a pathologic relevance in many disease states, namely, in those having an associated inflammatory component [6-8]. There are increasing evidence that ROS are involved in the pro-gression of renal damage and uraemic symptoms. The oxida-tive stress appears to be multifactorial and is thought to be caused only in part by the impaired kidney function [6-8]. The several factors that have been implicated in the devel-opment of oxidative stress in CKD patients undergoing haemodialysis include the haemodialysis procedure, urae-mia, inflammation, disturbances in iron homeostasis, and reduction in the antioxidant defence systems. The later in-cludes enzymatic systems involved in red blood cell (RBC) antioxidant defence, such as superoxide dismutase (SOD) and glutathione peroxidase (GPx). These enzymes are impor-



58 The Open Clinical Chemistry Journal, 2008, Volume 1 Costa et al.

tant to maintain higher levels of reduced glutathione, which is important in detoxifying the cell from ROS, and, therefore, in protecting the cell from their deleterious effects. Most of these factors are described also as independent factors of resistance to rhEPO therapy, particularly inflammation [11-15].

The RBC, presenting a limited biosynthesis capacity, suffers and accumulates physical and/or chemical changes, which become more pronounced with cell aging, and when-ever an unusual physical or chemical stress develops [16]. RBCs that develop intracellular defects earlier during their life span are removed prematurely from circulation [16-20]. The removal of senescent or damaged RBCs seems to in-volve the binding of a senescent neoantigen the RBC mem-brane surface, marking the cell for death. This neoantigen is immunologically related to band 3, a RBC transmembrane protein [18]. The degradation of the RBC metabolism and/or of its antioxidant defences may lead to the development of oxidative stress within the cell, allowing oxidation of hae-moglobin (Hb) and its linkage to the cytoplasmatic domain of band 3, promoting its aggregation. This aggregation con-ditioned the binding of natural antiband 3 autoantibodies and complement activation, marking RBC for death.

An abnormal band 3 profile [high molecular weight ag-gregates (HMWAg), band 3 monomer, and proteolytic frag-ments (Pfrag)], has been associated to younger, damage and/or senescence RBCs. Older and damaged RBCs pre-sented higher band 3 aggregation and lower fragmentation. The younger RBCs showed reduced aggregation and higher fragmentation [16-20].

Our aim was to study RBC damage and oxidative stress status occurring in CKD patients under haemodialysis and rhEPO therapy. The study was performed in a healthy con-trol group and in two groups of CKD patients (responders and non-responders). Some of the CKD patients were also evaluated before and immediately after haemodialysis to study the effect of the dialysis procedure per se. We per-formed a basic RBC study, by evaluating RBC count, Hb concentration, haematocrit (Ht) and the haematimetric in-dexes; the band 3 profile was evaluated as a cumulative marker of RBC damage; reticulocyte count, reticulocyte pro-duction index (RPI) and bilirubin levels, as markers of RBC production/removal; SOD and GPx activities, as markers of RBC antioxidant capacity; membrane bound Hb (MBH), as a marker of oxidative Hb damage; plasma total antioxidant status (TAS) and lipidic peroxidation (TBA), as well as the ratio TBA/TAS were studied to evaluate the oxidative stress in the plasma RBC environment.

MATERIAL AND METHODS

Subjects

We studied 89 individuals including 63 CKD patients, 32 responders and 31 non-responders to rhEPO therapy, and 26 healthy controls. The rhEPO maintenance dose for re-sponder�s patients was 90 ± 58 U/Kg/week and for non-responders was 573 ± 194 U/Kg/week. The two groups of patients were matched for age, gender, weight, body mass index, mean time under haemodialysis, urea reduction ratio, urea Ktv and parathyroid hormone serum levels.

No analytical indicators of iron deficiency and/or vitamin B12 and folate deficiencies were found in CKD patients and controls (data not shown).

The causes of renal failure in patient�s population were as follows: diabetic nephropathy (n=19), chronic glomeru-lonephritis (n=11), polycystic kidney disease (n=3), hyper-tensive nephrosclerosis (n=3), obstructive nephropathy (n=3), pyelonephritis associated with neurogenic bladder (n=1), nephrolithiasis (n=1), chronic interstitial nephritis (n=1), Alport syndrome (n=1), renal vascular disease due to polyarteritis (n=1) and chronic renal failure of uncertain ae-tiology (n=19).

Patients with autoimmune diseases, malignancies, hae-matological disorders, inflammatory disorders and acute or chronic infections, were excluded. All subjects gave their informed consent to participate in this study. Classification of CKD patients as responders or non-responders was per-formed in accordance with the European Best Practice Guidelines [21] that defines resistance to rhEPO as a failure to achieve target Hb levels (11-12 g/dL) with doses of epo-etin more than 300 U/Kg /week or 1.5 mg/Kg /week of dar-bopoietin-alfa.

Age and gender-matched individuals, with normal hae-matological and biochemical values, without any history of renal or inflammatory diseases were used as controls.

ASSAYS

Blood Samples

Blood samples were collected from CKD patients before starting haemodialysis. In 20 of the CKD patients (10 re-sponders and 10 non-responders to rhEPO therapy), blood samples were also collected immediately after the dialysis session.

Blood samples were collected with and without EDTA as anticoagulant, in order to obtain whole blood and serum. Whole blood was used for hematological procedures. To access GPx and SOD activities, whole blood and isolated and washed RBC aliquots, respectively, were made and stored at 70°C, until assayed. Plasma and serum aliquots were also made and stored at 70°C, until assayed.

Haematological and Biochemical Studies

RBC count, Ht, Hb concentration, haematimetric indices [mean cell volume (MCV), mean cell Hb (MCH) and mean cell Hb concentration (MCHC)] and red cell distribution width (RDW) were measured by using an automatic blood cell counter (Sysmex K1000; Sysmex, Hamburg, Germany).

Reticulocyte count was performed by microscopic count-ing on blood smears, after vital staining with New methylene blue (reticulocyte stain; Sigma, St Louis, MO, USA). The reticulocyte production index (RPI) was calculated as an appropriate way to measure the effective RBC production, by considering both changes in Ht (degree of anaemia) and in the premature reticulocyte release from the bone marrow [22].

Erythrocyte GPx and SOD activities were evaluated by using commercially available kits (RANSEL and RANSOD,



Band 3 Profile in CKD Patients The Open Clinical Chemistry Journal, 2008, Volume 1 59

Randox, UK, respectively). Total antioxidant status (TAS)

was evaluated in serum by a commercial colorimetric assay

(TAS, Randox). Plasma lipidic peroxidation was estimated

by TBA assay [23]. Moreover, the ratio TBA/TAS was

evaluated.

Serum total bilirubin levels were evaluated in an auto-

analyser (Cobas Mira S, Roche), using a commercially

available kit (Bilirubin total, ABX Diagnostics). Serum C-

reactive protein (CRP) was determined by immunoturbidi-

metry (CRP latex HS Roche kit, Roche Diagnostics).

Preparation of Erythrocyte Membranes

Plasma and white blood cells were isolated from RBCs and discarded after centrifugation on a double density gradient (Histopaque 1.077 and 1.119, Sigma). RBCs were washed in saline, and immediately lysed, according to Dodge method [24]. The membranes were carefully washed in Dodge buffer (the first two washes used phenylmethylsulfo- nyl fluoride, a protease inhibitor, at a final concentration of 0.1 M). Protein concentration of the membrane suspensions was determined [25].

Membrane-Bound Hb (MBH)

MBH was measured spectrophotometrically at 415 nm, after protein dissociation with Triton X-100 (5% in Dodge buffer). The background was read at 700 nm to correct the absorbance at 415 nm; the obtained value and the protein concentration were used to estimate the percentage of MBH.

Band 3 Profile

RBC membranes were treated with an equal volume of a solubilisation buffer containing 0.125 M Tris HCl pH 6.8, 4% sodium dodecil sulphate (SDS), 20% glycerol, and 10% 2-mercaptoethanol, heat-denatured and submitted to polyacrylamide gel electrophoresis (SDS-PAGE) (20 gprotein/lane), using the discontinuous Laemmli system [26] (a 9% separating gel and a 4.5% stacking gel). Membrane proteins were electrophoretically transferred from gels to a nitrocellulose sheet with a porosity of 0.2 m (Sigma) [27]. Additional reactive sites on the nitrocellulose were blocked by incubation in 5% of low fat dry milk and 0.1% Triton-X 100 in PBS (phosphate-buffered saline), pH 7.0, for 1 h at room temperature and under gentle rotation. Band 3 immunoblot was performed [18]; monoclonal antibodies antihuman band 3, produced in mouse, recognizing an epitope located in the cytoplasmic pole of the band 3 molecule [28] (Sigma), were added (dilution 1:3000) and incubated for 4 h; the washing of the nitrocellulose was followed by the addition and incubation with antimouse Ig peroxidase linked (Sigma) for 1 h (dilution 1:4000). The incubations were carried out at room temperature; the dilutions of the antibodies were prepared with PBS pH 7.0 containing 0.1% Triton-X 100 and 5% of low fat dry milk. In the washes, the same buffer without dry milk was used. Hydrogen peroxide and -cloronaphtol were used to develop the immunoblot. The band 3 profile was quantified by densitometry (Bio1D++ version 99; Vilber Lourmat, France). We evaluated the percentage of HMWAg, of band 3 monomer and of Pfrag.

DATA ANALYSIS

Statistical analyses were carried out using the Statistical Package for Social Sciences (SPSS). Multiple comparisons between groups were performed by one-way ANOVA sup-plemented with Tukey´s honestly significant difference (HSD) post hoc test. For data not normally distributed, dif-ferences between the three groups were evaluated by the Kruskal-Wallis test; for single comparisons (two groups), the Mann-Whitney U test was used. To assess the influence of the haemodialysis procedure per se on the evaluated parame-ters, paired t-test or Wilcoxon test were used. Significance was accepted at p less than 0.05.

Diabetic patients have been associated to an increased oxidative stress [29,30]. To exclude this possible confound-ing factor we further analysed the results, by performing statistical analysis with and without exclusion of the 19 CKD patients with a diabetic nephropathy.

RESULTS

The results were analysed in order to study the differ-ences between healthy controls and CKD patients; to study the differences between responders and non-responders CKD patients (cross-sectional study), and to study changes im-posed by the haemodialysis procedure.

We found that most of the CKD patients (all patients vscontrols) presented anaemia with significantly lower RBC count (3.68 ± 0.54 x 10

12/L vs 4.72 ± 0.59 x 10

12/L, p<0.05),

Hb [10.90 g/dL (10.30-12.30 g/dL) vs 13.90 g/dL (13.20-15.00 g/dL), p<0.05)] and Ht [34.20 % (30.60-37.10 %) vs43.10% (40.10-46.70%), p<0.05)]; higher reticulocyte count (61.03 ± 31.36 x 10

9/L vs 33.57 ± 22.78 x 10

9/L, p<0.05), RPI

[0.98 (0.58-1.40) vs 0.42 (0.19-0.66), p<0.05)] and RDW (15.92 ± 2.56% vs 12.79 ± 0.52 %, p<0.05) values were also observed. CKD patients showed (Fig. 1) statistically signifi-cant lower HMWAg [15.23 % (13.38-19.40 %) vs 19.90 % (15.42-21.12%), p<0.05)] and Pfrag (22.70 ± 6.01% vs 26.29 ± 4.78 %, p<0.05) values, and higher Band 3 monomer val-ues [61.84 % (56.87-64.41 %) vs 55.28 % (53.39-57.41%), p<0.05)]. The ratios HMWAg/band 3 monomer (0.27 ± 0.07 vs 0.33 ± 0.07, p<0.05) and Pfrag/band 3 monomer (0.38 ± 0.13 vs 0.48 ± 0.13, p<0.05) were also lower. GPx activity (45.82 ± 13.69 IU/g Hb vs 35.62 ± 8.83 IU/g Hb, p<0.05), TBA (1.49 ± 0.31 x10

-3 mmol/L vs 0.84 ± 0.27 x10

-3 mmol/L

p<0.05), TAS (1.79 ± 0.24 mmol/L vs 1.41 ± 0.29 mmol/L, p<0.05) and CRP levels [5.76 mg/dL (1.90-14.01 mg/dL) vs1.75 mg/dL (0.76-4.70 mg/dL), p<0.05)], and TBA/TAS ratio (0.85 ± 0.18 vs 0.60 ± 0.28, p<0.05) were higher in CKD patients. Similar values for the activity of SOD and for MBH, between controls and patients, were observed.

In Table 1, we present the haematological data, for con-trols and CKD patients. When comparing the haematological data between the two groups of CKD patients, we found that non-responders patients were all anaemic and showed sig-nificantly lower values of Hb and Ht; their RBCs were more hypocromic, as reflected by the significantly lower values of MCH and MCHC; a significantly higher RDW was also ob-served (Table 1). Statistically significant decreases in Pfrag and in the Pfrag/band 3 monomer ratio, and increase CRP were also found in non-responders CKD patients (Table 2).




Analysing the results for the studied parameters, before and immediately after the dialysis procedure, in 20 of the studied CKD patients, we found a statistically significant rise in Hb [12.10 (10.95-12.80 g/dL) vs 13.20 (11.15-14.60 g/dL), p<0.05] and significant decreases in MBH [75.00 (49.00-109.50 x10

-4%) vs 24.50 (13.00-64.50 x10

-4%),

p<0.05], in plasma TBA (1.52 ± 0.30 x10-3

mmol/L vs 1.36 ±

0.26 x10-3

mmol/L, p<0.05) and in plasma TAS (1.58 ± 0.38 mmol/L vs 1.26 ± 0.35 mmol/L, p<0.05) after haemodialysis. The other studied parameters remain unchanged after haemodialysis. No statistically significant differences were found between responders and non-responders patients, for the studied changes after the haemodialysis procedure.

Fig. (1). Examples of densitometer tracing of immunoblots for band 3 profile. A- Control; B- Responder patient; C- Non-responder patient.

Table 1. Haematological Data for Controls and CKD Patients Under Haemodialysis and rhEPO Therapies

Controls

(n=26)

Responders

(n=32)

Non-Responders

(n=31)

Hb (g/dL) 13.90 (13.2-15.00) 11.70 (10.83-12.68) * 10.4 (9.00-11.30) * §

Haematocrit (%) 43.10 (40.10-46.70) 35.15 (32.25-38.35) * 31.10 (27.70-35.20) * §

RBC (x1012/L) 4.72 ± 0.59 3.76 ± 0.42 b) 3.58 ± 0.64 *

MCV (fL) 92.00 (90.00-94.00) 95.80 (92.48-98.08) * 92.30 (85.40-100.30)

MCH (pg) 29.83 ± 1.39 31.29 ± 1.53 b) 28.97 ± 3.73 §

MCHC (g/dL) 32.48 ± 0.58 33.16 ± 1.77 30.85 ± 2.35 * §

RDW (%) 12.79 ± 0.52 14.56 ± 1.23 * 17.32 ± 2.83 * §

Reticulocytes (x109/L) 33.57 ± 22.78 55.12 ± 30.98 * 67.14 ± 31.06 *

RPI 0.42 (0.19-0.66) 1.08 (0.72-1.51) * 0.92 (0.52-1.24) *

Bilirubin (mg/dL) 0.62 ± 0.25 0.61 ± 0.23 0.62 ± 0.24

* p<0.05, vs controls; § p<0.05 vs responders.

Hb: haemoglobin concentration; RBC: red blood cell count; MCV: mean cell volume; MCH: mean cell haemoglobin; MCHC: mean cell haemoglobin concentration; RDW: red blood cell distribution width; RPI: reticulocyte production indice. Results are presented as mean ± standard deviation or as median (interquartile ranges).




We found similar statistical differences between groups when CKD patients with diabetic nephropathy were removed from the analyses, to exclude diabetes as a confounding fac-tor (as previously referred). Actually, the values that we found for CKD patients, with and without diabetic nephropa-thy, were very similar.

DISCUSSION

The treatment of anaemia, the most frequent haematological alteration found in CKD patients, has been improved in the last years by rhEPO therapy. This therapy increases RBC production, providing an increased number of younger and more powerful RBCs in terms of functionality. However, it is known that in this kind of patients RBCs have to circulate in an adverse plasma environment and that they have to undergo a haemodialysis procedure regularly. Both conditions may induce metabolical and/or physical RBC damage that may lead to premature senescence and removal of RBCs.

As previously described [1-3], in the present study, de-spite rhEPO therapy, anaemia was a common finding in our CKD patients under haemodialysis (Table 1). The reduction in Hb concentration was associated with a statistically sig-nificant increase in circulating reticulocytes and RPI (twice the value of controls), reflecting the erythropoietic stimulus triggered by rhEPO therapy. The higher RDW observed in CKD patients, may reflect the rise in reticulocyte count and/or a higher RBC damage, with anisocytosis develop-ment.

To evaluate oxidative and proteolytic RBC damage we evaluated the band 3 profile and MBH. As Hb denatures, it binds to intracellular surface of the membrane, leading to

band 3 aggregation [31]. Previous studies in our lab showed that oxidative stress conditions [16-20] are associated to an increase in band 3 HMWAg and a reduction in Pfrag. Moreover, we have also reported that younger RBCs present a decreased band 3 HMWAg and a rise in Pfrag [17].

We report, for the first time, changes in the erythrocyte membrane band 3 profile in CKD patients. CKD patients presented a decrease in HMWAg and in HMWAg/band 3 monomer ratio. These changes seem to reflect a younger RBC population. In fact, erythrocyte membranes from younger RBCs are poorer in HMWAg and richer in Pfrag. However, our CKD also presented a decrease in Pfrag and in Pfrag/band 3 monomer ratio, which are associated to a rise in RBC damage. Thus, it seems that the band 3 profile observed in CKD patients is associated both to an increase in younger RBCs and to an increase in damaged RBCs. The same results were previously described to be associated with pregnancy [20].

Haemodialysis patients exhibited higher plasma levels of TBA and also of TAS. A higher TBA/TAS ratio reflecting a higher rise in TBA than in TAS, is in favour of an oxidative stress. No difference between controls and CKD patients was found for RBC SOD activity; this, has been related to an enzymatic inactivation due to hyperproduction of ROS [32]; the increase in RBC GPx activity is probably due to the higher levels of circulating reticulocytes (almost twice the value of controls). Actually, it has been reported that reticulocytes present a higher GPx activity when compared to mature RBCs. Data of about antioxidant RBC enzymes activities in haemodialysis patients are controversial, as authors reported decreased, increased or even unchanged enzyme activities [33-35].

Table 2. RBC Damage, Enzymatic Activities of Erythrocyte SOD and GPx, Plasma Lipid Peroxidation, Antioxidant Status and

CRP, for Controls and CKD Patients

Controls

(n=26)

Responders

(n=32)

Non-Responders

(n=31)

HMWAg (%) 19.90 (15.42-21.12) 14.86 (11.30-20.19) * 15.92 (14.28-18.68) *

Band 3 monomer (%) 55.28 (53.39-57.41) 61.26 (56.08-65.06) * 62.17 (58.01-64.29) *

Pfrag (%) 26.29 ± 4.78 24.01 ± 6.03 21.34 ± 5.78 * §

HMWAg/ Band 3 monomer 0.33 ± 0.07 0.26 ± 0.09 * 0.27 ± 0.06 *

Pfrag/ Band 3 monomer 0.48 ± 0.11 0.41 ± 0.14 * 0.35 ± 0.13 * §

MBH (x10-4 %) 53.00 (37.75-89.75) 45.50 (25.25-80.75) 58.50 (30.50-100.75)

GPx (IU/g Hb) 35.62 ± 8.83 48.73±13.46 * 43.11±13.87

SOD (IU/g Hb) 1039.8 (737.4-1331.6) 858.97 (662.4-1256.5) 1074.76 (581.6-2638.7)

TBA (x10-3 mmol/L) 0.84 ± 0.27 1.49 ± 0.32 * 1.48 ± 0.30 *

TAS (mmol/L) 1.41 ± 0.29 1.81 ± 0.24 * 1.76 ± 0.24 *

TBA/TAS 0.60 ± 0.28 0.83 ± 0.18 * 0.86 ± 0.18 *

CRP (mg/dL) 1.75 (0.76-4.70) 3.20 (1.73-7.23) * 10.14 (3.82-38.99) * §

* p<0.05, vs controls; §

p<0.05 vs responders.

HMWAg; high molecular weight aggregates; Pfrag: proteolytic fragments; SOD: superoxide dismutase; GPx: glutathione peroxidase; MBH: membrane bound haemoglobin; TAS: total antioxidant status; TBA: plasma lipid peroxidation; CRP: C-reactive protein.

Results are presented as mean ± standard deviation or as median (interquartile ranges).




The increased erythropoietic stimulation triggered by rhEPO therapy, seems to provide a younger RBC population that has to face an adverse plasma oxidative environment as shown by higher values of plasma lipid peroxidation and by the band 3 profile showing changes denoting a younger but also a damaged RBC population.

Analyzing the haematological results presented by the two groups of CKD patients, responders and non-responders to rhEPO therapy, we found that the non-responders pre-sented anaemia, showing a statistically significant decrease in Hb, Ht and RBC count (Table 1). These changes were associated to significantly lower MCH and MCHC values, showing a hypocromic RBC change, which was not due to erythropoietic deficiencies in iron, folic acid or vitamin B12 (data not shown). A statistically significant rise in RDW was also observed in non-responders that could reflect reticulocy-tosis and/or anisocytosis due to a higher RBC damage. As responders and non-responders presented with a similar re-ticulocyte value, though presenting non-responders a trend to higher reticulocyte count, the increase in RDW seems to mainly result from anisocytosis, reflecting a higher RBC damage.

Non-responders CKD patients, compared to responders, though presenting a similar reticulocyte and RBC count, showed a decrease in Pfrag and in Pfrag/band 3 monomer ratio, and a trend to higher values of MBH (Table 2), sug-gesting that they present a higher RBC damage that may result from a more adverse plasma microenvironment. Resis-tance to rhEPO therapy in CKD patients seems, therefore, to be associated to an increase in RBC damage. No alterations were found in non-responders patients in the parameter re-lated to RBC antioxidant capacity; TAS and TBA.

This study also shows that the haemodialysis procedure per se does not lead to an increase in the studied markers RBC damage. Actually, no differences were found after haemodialysis, in band 3 profile and, besides that, a decrease in MBH, TBA and TAS were observed, probably reflecting the improvement in the plasma environment and/or the re-moval of damaged RBCs.

It is known that the linkage of denatured Hb to RBC membrane at the cytoplasmic pole of the protein band 3 fa-vours the aggregation of this transmembrane protein [36] and turns the cell membrane more rigid. In our study, the MBH decreased after the haemodialysis procedure, suggesting that damaged RBCs may have been removed from circulation. Actually, damaged RBCs may be more susceptible to haemolyse, when submitted to mechanical stress as may oc-cur during haemodialysis. Another possibility is that the small portions of membrane with bound Hb and band 3 ag-gregates might be lost by a process of membrane vesicula-tion, as occurs in splenic sinusoids, where RBCs have also to deform, to go through that microvasculature.

The increased indicators of RBC damage found in our CKD patients, particularly in non-responders (decrease in Pfrag and a trend to increase of MBH), was also associated to an increase in CRP. There are evidences that a rise in proinflammatory cytokines levels (especially interleukine-6) and in the acute phase reactant CRP, are among the most powerful independent predictors of resistance to rhEPO ther-apy, and are associated with cardiovascular morbidity and

mortality in CKD patients [4-7]. There are also some evi-dence of a bidirectional and synergistic linkage between in-flammation and oxidative stress. However, further studies are required targeting specific oxidative pathways, as well as specific inflammatory pathways to elucidate this biological linkage [6,7]. The increase RBC damage found in CKD pa-tients, particularly in non-responders, can also be associated to the proteolytic effect of elastase released from neutrophils during haemodialysis procedure. In fact, recently higher plasma levels of elastase were demonstrated by us in non-responders patients [37].

Our data suggest that CKD patient�s present younger RBC population, which could be related to the rhEPO ther-apy. The adverse plasma environment associated to CKD patients under hemodialysis imposes changes in band 3 pro-file, particularly in non-responders.

ACKNOWLEDGEMENTS

We are very grateful to FMC, Dinefro � Diálises e Ne-frologia, SA and Uninefro � Sociedade Prestadora de Cui-dados Médicos e de Diálise, SA, and to their nurses for the technical support. This study was supported by a PhD grant (SFRH/BD/27688/2006) attributed to E. Costa by Fundação para a Ciência e Tecnologia and Fundo Social Europeu.

REFERENCES

[1] Foley, R.N.; Parfrey, P.S.; Sarnak, M.J. Clinical epidemiology of

cardiovascular disease in chronic renal failure. Am. J. Kidney Dis., 1998, 32(suppl. 3), S112-S119.

[2] Foley, R.N.; Parfrey, P.S.; Harnett, J.D.; Kent, G.M.; Murray, D.C.; Barre, P.E. The impact of anemia on cardiomyopathy morbidity

and mortality in end-stage renal disease. Am. J. Kidney Dis., 1996,28, 53-61.

[3] Locatelli, F.; Conte, F.; Marcelli, D. The impact of hematocrit levels and erythropoietin treatment on overall and cardiovascular

mortality and morbidity - the experience of the Lombardy dialysis registry. Nephrol. Dial. Transplant., 1998, 13, 1642-1644.

[4] Smrzova, J.; Balla, J.; Bárány, P. Inflammation and resistance to erythropoiesis-stimulting agents-what do we know and what needs

to be clarified? Nephrol. Dial. Transplant., 2005, 20(Suppl. 8), viii2-viii7.

[5] Bárány, P. Inflamation, serum C-reactive protein, and erythro-poietin resistance. Nephrol. Dial. Transplant., 2001, 16, 224-227.

[6] Spittle, M.A.; Hoenich, N.A.; Handelman, G.J.; Adhikarla, R.; Homel, P.; Levin, N.W. Oxidative stress and inflammation in

hemodialysis patients. Am. J. Kidney Dis., 2001, 38, 1408-1413. [7] Pupim, L.B.; Himmelfarb, J.; McMonagle, E.; Shyr, Y.; Ikizler,

T.A. Influence of initiation of maintenance hemodialysis on bio-markers of inflammation and oxidative stress. Kidney Int., 2004,

65, 2371-2379. [8] Sommerburg, O.; Grune, T.; Hampl, H.; Riedel, E.; van Kuijk, F.J.;

Ehrich, J.H.H.; Siems, W.G. Does long-term treatment of renal anaemia with recombinant erythropoietin influence oxidative stress

in haemodialysed patients? Nephrol. Dial. Transplant., 1998, 13,2583-2587.

[9] Himmelfarb, J. Linking oxidative stress and inflammation in kid-ney disease: which is the chicken and which is the egg? Semin. Di-

al., 2004, 17, 449-454. [10] Drueke, T.B.; Eckardt, K.U. Role of secondary hyperparathyroid-

ism in erythropoietin resistance of chronic renal failure patients. Nephrol. Dial. Transplant., 2002, 17(suppl 5), 28-31.

[11] Lucchi, L.; Bergamini, S.; Botti, B.; Rapanà, R.; Ciuffreda, A.; Ruggiero, P.; Ballestri, M.; Tomasi, A.; Albertazzi, A. Influence of

different hemodialysis membrane on red blood cell susceptibility of oxidative stress. Artif. Organs, 2000, 24, 1-6.

[12] Fryer M.J. Vitamin E as a protective antioxidant in progressive renal failure. Nephrology, 2000, 5, 1-7.

[13] Sezer, M.T.; Akin, H.; Demir, M.; Erturk, J.; Aydin, Z.D.; Savik, E.; Tunc, N. The effect of serum albumin levels on iron-induced




oxidative stress in chronic renal failure patients. J. Nephrol., 2007,

20, 196-203. [14] Costa, E.; Lima, M.; Alves, J.M.; Rocha, S.; Rocha-Pereira, P.;

Castro, E.; Miranda, V.; Sameiro-Faria, M.; Loureiro, A.; Quin-tanilha, A.; Belo, L.; Santos- Silva, A. Inflammation, T-cell pheno-

type and inflammatory cytokines in chronic kidney disease patients under haemodialysis and its relationship to resistance to recombi-

nant human erythropoietin therapy. J. Clin. Immunol., 2008, 28,268-275.

[15] Imanishi, H.; Nakai, T.; Abe, T.; Takino, T. Glutathione-linked enzyme activities in red cell aging. Clin. Chim. Acta, 1986, 159,

73-76. [16] Rocha-Pereira, P.; Santos-Silva, A.; Rebelo, I.; Figueiredo, A.;

Quintanilha, A.; Teixeira, F. Erythrocyte damage in mild and severe psoriasis. Br. J .Dermatol., 2004, 150, 232-244.

[17] Santos-Silva, A.; Castro, E.M.B.; Teixeira, N.A.; Guerra, F.C.; Quintanilha, A. Erythrocyte membrane band 3 profile imposed by

cellular aging, by activated neutrophils and by neutrophilic elastase. Clin. Chim. Acta, 1998, 275, 185-196.

[18] Santos-Silva, A.; Castro, E.M.B.; Teixeira, N.A.; Guerra, F.C.; Quintanilha, A. Altered erythrocyte membrane band 3 profile as a

marker in patients at risk for cardiovascular disease. Atherosclerosis, 1995, 116, 199-209.

[19] Santos-Silva, A.; Rebelo, I.; Castro, E.M.B.; Belo, L.; Catarino, C.; Monteiro, I.; Almeida, M.D.; Quintanilha, A. Erythrocyte damage

and leukocyte activation in ischemic stroke. Clin. Chim. Acta,2002, 320, 29-35.

[20] Belo, L.; Rebelo, I.; Castro, E.M.B.; Catarino, C.; Pereira-Leite, L.; Quintanilha, A.; Santos-Silva, A. Band 3 as a marker of erythrocyte

changes in pregnancy. Eur. J. Haematol., 2002, 69, 145-151. [21] Locatelli, F.; Aljama, P.; Barany, P.; Canaud, B.; Carrera, F.;

Eckardt, K.U.; Horl, W.H.; Macdougal, I.C.; Macleod, A.; Wiecek, A.; Cameron, S. European Best Practice Guidelines Working

Group. Revised European best practice guidelines for the manage-ment of anaemia in patients with chronic renal failure. Nephrol.

Dial. Transplant., 2004, 19 (Suppl 2), ii1-ii47. [22] Hillman, R.S.; Ault, K.A. Clinical approach to anemia In: Hema-

tology in Clinical Practice, McGraw-Hill: New-York, 2002.[23] Niehaus, W.G.; Samuelsson, B. Formation of malonaldehyde from

phospholipids arachidonate during microsomal lipid peroxidation. Eur. J. Biochem., 1968, 6, 126-130.

[24] Dodge, J.T.; Mitchell, C.; Hanahan, D.J. The preparation and chemical characteristics of hemoglobin-free ghosts of human

erythrocytes. Arch. Biochem. Biophys., 1963, 100, 119-130.

[25] Bradford, M.M. A rapid and sensitive method for the quantitation

of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem., 1976, 72, 248-254.

[26] Laemmli, U.K. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature, 1970, 227, 680-685.

[27] Towbin, H.; Staehelin, T.; Gordon, J. Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets:

procedure and some applications. Proc. Natl. Acad. Sci. USA, 1979,76, 4350-4354.

[28] Czerwinski, M.; Wasniowska, K.; Steuden, I.; Duk, M.; Wiedlocha, A.; Lisowska, E. Degradation of the human erythrocyte membrane

band 3 studied with monoclonal antibody directed against an epitope on the cytoplasmic fragment of band 3. Eur. J. Biochem.,

1988, 174, 647-654. [29] Philips, M.; Cataneo, R.N.; Cheema, T.; Greenberg, J. Increased

breath biomarkers of oxidative stress in diabetes mellitus. Clin. Chim. Acta, 2004, 344, 189-194.

[30] Maritim, A.C.; Sanders, R.A.; Watkins, J.B. Diabetes, oxidative stress, and antioxidants: a review. J. Biochem. Mol. Toxicol., 2003,

17, 24-38. [31] Lutz, H.U. Naturally occurring anti-band 3 antibodies. Transfus.

Med. Rev., 1992, 6, 201-211. [32] Delmas-Beauvieux, M.C.; Combe, C.; Peuchant, E.; Carbonneau,

M.A.; Dubourg, L.; de Précigout, V.; Aparicio, M.; Clerc, M. Evaluation of red blood cell lipoperoxidation in hemodialysed pa-

tients during erythropoietin therapy supplemented or not with iron. Nephron, 1995, 69, 404-410.

[33] Chen, C.K.; Liaw, J.M.; Juang, J.G.; Lin, T.H. Antioxidant en-zymes and trace elements in hemodialyzed patients. Biol. Trace

Elem. Res., 1997, 58, 149-157. [34] Túri, S.; Németh, I.; Vargha, I.; Matkovics, B.; Dobos, E. Erythro-

cyte defense mechanisms against free oxygen radicals in haemo-dialysed uraemic children. Pediatr. Nephrol., 1991, 5, 179-183.

[35] Paul, J.L.; Sall, N.D.; Soni, T.; Poignet, J.L.; Lindenbaum, A.; Man, N.K.; Moatti, N.; Raichvarg, D. Lipid peroxidation abnor-

malities in hemodialyzed patients. Nephron, 1993, 64, 106-109. [36] Low, P.S. Role of hemoglobin desnaturation and band 3 clustering

in initiating red cell removal. In: Red Blood Cell Aging, Plenum Press:New York, 1991.

[37] Costa, E.; Rocha, S.; Rocha-Pereira, P.; Nascimento, H.; Castro, E.; Miranda, V.; Sameiro-Faria, M.; Loureiro, A.; Quintanilha, A.;

Belo, L.; Santos-Silva, A. Neutrophil activation and resistance to recombinant human erythropoietin therapy in hemodialysis pa-

tients. Am. J. Nephrol., 2008, 28, 935-940.

Received: July 18, 2008 Revised: July 31, 2008 Accepted: September 3, 2008

© Costa et al.; Licensee Bentham Open.

This is an open access article licensed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/

by-nc/3.0/) which permits unrestricted, non-commercial use, distribution and reproduction in any medium, provided the work is properly cited.



Paper X

“Cross-talk between inflammation, coagulation/fibrino-

lysis and vascular access in Chronic Kidney Disease Patients

under Haemodialysis”

J Vasc Access, in press.



1129-7298/001-04$25.00/0

The Journal of Vascular Access 2008; 9: 00-00

Cross-talk between inflammation,coagulation/fibrinolysis and vascular access in hemodialysis patients

E. COSTA1,2,3, S. ROCHA2,3, P. ROCHA-PEREIRA3,4, E. CASTRO2,3, F. REIS5, F. TEIXEIRA5, V. MIRANDA6, M. DOSAMEIRO FARIA6, A. LOUREIRO7, A. QUINTANILHA2,8, L. BELO2,3, A. SANTOS-SILVA2,3

1Instituto de Ciências da Saúde da Universidade Católica Portuguesa, Porto - Portugal2Faculdade Farmácia, Serviço de Bioquímica, Universidade do Porto - Portugal3Instituto Biologia Molecular e Celular (IBMC), Universidade do Porto - Portugal4Centro Investigação Ciências Saúde, Universidade Beira Interior, Covilhã- Portugal5Instituto de Farmacologia e Terapêutica Experimental, Faculdade Medicina, Universidade Coimbra - Portugal6FMC, Dinefro – Diálises e Nefrologia, SA - Portugal7Uninefro – Sociedade Prestadora de cuidados Médicos e de Diálise, SA - Portugal8Instituto Ciências Biomédicas Abel Salazar (ICBAS), Universidade do Porto - Portugal

Abstract: This work aimed to study the association between fibrinolytic/endothelial cell function and inflamma-tory markers in chronic kidney disease (CKD) patients undergoing hemodialysis (HD) and recombinant humanerythropoietin (rhEPO) therapies, and its relationship with the type of vascular access (VA) used for the HD pro-cedure. As fibrinolytic/endothelial cell function markers we evaluated plasminogen activator inhibitor type-1 (PAI-1), tissue plasminogen activator (tPA) and D-dimers, and as inflammatory markers; C-reactive protein (CRP),soluble interleukin (IL)-2 receptor (sIL2R), IL-6 and serum albumin levels. The study was performed in 50 CKDpatients undergoing regular HD, 11 with a central venous dialysis catheter (CVC) and 39 with an arteriovenous fi-stula (AVF), and in 25 healthy controls. Compared to controls, CKD patients presented with significantly higherlevels of CRP, s-IL2R, IL-6 and D-dimers, and significantly lower levels of PAI-1. The tPA/PAI-1 ratio was signifi-cantly higher in CKD patients. We also found statistical significant correlations in CKD patients between D-dimerslevels and inflammatory markers: CRP, albumin, s-IL2R and IL-6. When comparing the two groups of CKD pa-tients, we found that those with a CVC presented statistically significant lower levels of hemoglobin concentrationand albumin, and higher levels of CRP, IL-6, D-dimers and tPA. Our results showed an association between fibri-nolytic/endothelial cell function and increased inflammatory markers in CKD patients. The increased levels of D-dimer, tPA and inflammatory markers in CKD patients using a CVC, led us to propose a relationship between thetype of VA chosen for HD, and the risk of thrombogenesis. (J Vasc Access 2008; 9: 0-0)

Key words: Fibrinolytic activity, Chronic renal failure, Vascular access, D-dimers

© Wichtig Editore, 2008

INTRODUCTION

A successful hemodialysis (HD) procedure requiresa functional vascular access (VA). Unfortunately, nomajor advances in the field of HD VA have been ob-served for the past 30 yrs, and this has probablycontributed to VA dysfunction as being one of themost important causes of morbidity in the HD pop-ulation (1). Access-related problems are responsiblefor 50% of the hospitalizations of HD patients (2).Currently, there are three main forms of HD VA: ar-teriovenous fistula (AVF), polytetrafluoroethylene

(PTFE) graft, and central venous catheter (CVC).Each of these forms has its own specific problems.AVF is the first choice for VA in patients undergoingchronic HD due to its relative low risk of infectionand thrombosis. However, this type of VA has two ma-jor complications: initial failure to mature (primarynon-function) and a later venous stenosis followed bythrombosis (3). A PTFE graft is relatively easy toplaceand ready to use; however, extremely high ratesof stenosis, thrombosis, and infection are reportedfor this VA (4). CVC is the least desirable method ofHD access. In a sense, every CVC placement repre-



2

Inflammation, coagulation/fibrinolysis and vascular access in CKD patients

sents a failure: failure to prepare a native AVF inadvance of initiating dialysis; failure to detect fail-ing AV access and maintain its function, or createa suitable alternative. Nevertheless, catheters arean unavoidable necessity for many patients whodo not have functional AV access for any reason.In recent years, the use of CVCs has increased inboth acute and chronic uremic patients, and thiscould underlie the increased morbidity and mor-tality observed in these patients. The type of in-sertion and the management of a CVC can lead toseveral complications, namely infections andthrombosis (5). An inflammatory stimulus triggers the synthesis ofacute phase proteins, namely C-reactive protein(CRP), a prominent marker of inflammatory re-sponse in the general population and in chronickidney disease (CKD) patients (6-8). CRP increasessignificantly in hemodialyzed patients, as comparedwith healthy controls (8, 9) and this rise appears tobe a common feature in hemodialyzed patients.Chronic inflammation is also associated with ather-osclerotic cardiovascular disease (CVD). CRP hasbeen proposed as a new risk marker for CVD events;the high rate of morbidity and mortality observed inaemodialyzed patients (10, 11) could reflect the over-lapping of these proinflammatory conditions and/orthe inflammatory process in CKD patients favoringCVD events. CKD patients undergoing HD have been also associ-ated with complex hemostatic disorders, clinically ex-pressed with both bleeding tendency and hypercoag-ulability, though its etiology is poorly understood. Adisturbance in fibrinolysis was reported in CKD pa-tients; however, studies of different fibrinolysis pa-rameters in regularly dialyzed patients have yieldedconflicting results, with some indicating suppressedfibrinolysis and others showing increased fibrinolysis(12). Several hemostatic proteins are also consideredendothelial cell function markers, namely plasmino-gen activator inhibitor 1 (PAI-1) and tissue plasmino-gen activator (tPA), being both secreted by endothe-lial cells.There are a considerable number of studies re-garding the association of the type of VA with in-flammatory markers and fibrinolytic/endothelialcell dysfunction (13, 14). However, only a few haveinvestigated simultaneously inflammatory and fib-rinolytic/endothelial cell dysfunction markers. This work aimed to study the association between fib-rinolytic/endothelial cell function and inflamma-tory markers in CKD patients undergoing HD andrecombinant human erythropoietin (rhEPO) ther-apies, and its relationship with the type of VA usedfor the HD procedure.

MATERIALS AND METHODS

Subjects

We studied 50 CKD patients undergoing regular HD(32 males, 18 females; mean age 64.5 ± 15.4 yrs)and rhEPO therapy. These patients were recruited,after giving informed consent, from an HD clinicduring a 1-month period. In 11 patients, the VA usedfor HD was a tunneled CVC, and in 39 patients it wasan AVF. The CVC were used for a median period of36 months (12-84 months). The incidence of VA in-fection in the unit is 5.6%.CKD patients were undergoing therapeutic HD threetimes per week, for 3-5 hr, for a median period of 36months. All patients used the high-flux polysulfoneFX-class dialyzers of Fresenius, 25 with FX60, 23 withFX80 and 2 with the FX100 dialyzer type. Patients with autoimmune disease, malignancy,hematological disorders, and acute or chronic in-fection were excluded. We also excluded patientsthat presented, at the time of analysis with a positiveblood culture result.The control group included 25 healthy volunteers (8males, 17 females; mean age 47.81 ± 14.69 yrs) pre-senting normal hematological and biochemical val-ues, with no history of renal or inflammatory dis-eases, and, as far as possible, age and gendermatched with CKD patients.

Assays

Blood samples were drawn from fasting controls andin the case of the hemodialyzed patients before thesecond dialysis session of the week. Blood sampleswere collected with and without anticoagulant [eth-ylenediamine tetraacetic acid (EDTA)], in order toobtain whole blood, plasma and serum. Hemoglobin (Hb) concentration and white bloodcell (WBC) count were measured using an auto-matic counter (Sysmex K1000, Hamburg, Germany);differential leukocyte counts were evaluated inWright-stained blood films. Serum CRP was meas-ured by immunoturbidimetry (CRP latex HS Rochekit, Roche Diagnostics) and an enzyme-linked im-munoabsorbent assay was used for measurement ofserum soluble interleukin 2 receptor (s-IL2R) (hu-man IL-2 SRa, R&D systems, MN, USA). Serum al-bumin levels were measured using a colorimetric as-say end-point method (Albumin Plus; Roche GmbH,Mannheim, Germany). Serum IL-6 levels were quan-tified using the BDTM Cytometric Bead Array Hu-man Th1/Th2 Cytokine Kit II (BD Biosciences, USA,San Diego), and analyzed using the BDTM CBA soft-ware. Plasma levels of PAI-1, tPA and D-dimers were



3

Costa et al

evaluated by enzyme-linked immunoabsorbent as-says (TintElize PAI-1, TintElize tPA and TintElize D-dimer, Biopool-Trinity Biotech Company, respec-tively). Serum iron concentration was determinedusing a colorimetric method (Iron, Randox Labora-tories Ltd, Northern Ireland, UK), whereas serumferritin and serum transferrin were measured by im-munoturbidimetry (Ferritin, Laboratories Ltd,Northern Ireland, UK; Transferrin, LaboratoriesLtd, Northern Ireland, UK). Transferrin saturation(TS) was calculated by the formula: TS (%) = 70.9 xserum iron concentration in (µg/dL) / serum trans-ferrin concentration in (mg/dL).

Data analysis

For statistical analysis, we used the Statistical Packagefor Social Sciences, version 14.0. KolmogorovSmirnov statistics were used to evaluate sample nor-mality distribution. Comparisons between groupswere performed using the Kruskal-Wallis test andthe Mann-Whitney U test (data with a non-Gaussiandistribution) or one-way ANOVA supplemented withTukey’s HSD post-hoc test (data with a Gaussian dis-tribution). Spearman’s rank correlation coefficientwas used to evaluate relationships between sets ofdata. Adjustment of statistical differences for con-founding factors (comorbidity, age, sex and bodymass index) was performed by analysis of covari-ance, after log transformation of variavel[AQ: vari-

ance?] (when necessary). Multiple regression analy-sis using the stepwise method was used to determineindependent factors affecting D-dimers levels. Sig-nificance was accepted at p<0.05.

RESULTS

Tables I and II show rhEPO doses, hematologicaldata, fibrinolytic and inflammatory cell markers,dialysis adequacy parameters, and nutritional andiron status, for controls and CKD patients.Compared to controls, CKD patients presented withsignificantly higher counts of WBC and neutrophils,and higher levels of ferritin, CRP, s-IL2R, IL-6 and D-dimers. CKD patients also presented with signifi-cantly lower levels of Hb and PAI-1. The tPA/PAI-1ratio was significantly higher for CKD patients.When comparing the two groups of CKD patients, wefound that those with a CVC VA presented with sta-tistically significant lower levels of Hb, TS and albu-min, and higher levels of neutrophils, CRP, IL-6, D-dimers and tPA. Statistical significance remainedafter adjustment for the confounding factors. Atrend to higher rhEPO doses to achieve target Hblevels (11-12 g/dL) was found in CKD patients usinga CVC as VA (Tab. I). We also found in CKD patients, statistically significantcorrelations between D-dimer levels and inflamma-tory markers: CRP, albumin, s-IL2R and IL-6 (Fig. 1).

TABLE I - HEMATOLOGICAL DATA, FIBRINOLYTIC AND INFLAMMATORY CELL MARKERS, DIALYSIS ADEQUACY PARAMETERS, AND NUTRITIONAL AND IRON STATUS IN CONTROLS AND CKD PATIENTS

Healthy Controls All CKD Patients(n=25) (n=50)

Weekly rhEPO (IU/Kg) dose – 271.29 (79.22-577.29)Hb (g/dL) 14.12 ± 1.27 11.06 ± 1.77 a)WBC (x 109/L) 5.80 ± 1.60 10.03 ± 4.54 a)Neutrophils (x 109/L) 3.00 ± 1.00 6.08 ± 3.01 a)Platelets (x 109/L) 242.12 ± 78.57 212.48 ± 98.39CRP (mg/dL) 1.75 (0.76-4.70) 5.75 (1.90-14.00) a)s-IL2R (nmol/L) 758.83 ± 234.95 4199.94 ± 1762.82 a)IL-6 (pg/mL) 1.90 (0-3.75) 7.8 (3.85-15.05) a)Albumin (g/dl) – 3.84 ± 0.42D-dimer (ng/mL) 63.50 (51.94-77.41) 157.26 (97.72-417.09) a)tPA (ng/mL) 4.72 ± 2.95 4.60 ± 3.57PAI-1 (ng/mL) 58.17 ± 24.25 42.85 ± 23.53 a)tPA/PAI-1 ratio 0.09 ± 0.06 0.14 ± 0.11 a)Transferrin saturation (%) 21.83 ± 7.97 22.89 ± 11.06Ferritin (ng/mL) 85.10 (37.88-123.95) 389.65 (186.58-604.90)a)Ktv – 1.59 ± 0.29Urea reduction (%) – 26.98 ± 8.53Protein catabolic rate (g/Kg/day) – 0.85 ± 0.04

*Results are presented as mean ± one standard deviation or as median values (inter-quartile range). a) p<0.05, vs. controls



4


TABLE II - HEMATOLOGICAL DATA, FIBRINOLYTIC AND INFLAMMATORY CELL MARKERS, DIALYSIS ADEQUACY PARAMETERS, AND NUTRITIONAL AND IRON STATUS ACCORDING TO THE TYPE OF VASCULAR ACCESS

CKD patients with AVF CKD patients with CVC(n=39) (n=11)

Weekly rhEPO (IU/Kg) dose 156.01 (56.82-582.52) 456.12 (102.56-575.54)Hb (g/dL) 11.27 ± 1.86 10.30 ± 1.18 a)WBC (x 109/L) 10.11 ± 4.57 9.77 ± 4.64 Neutrophils (x 109/L) 5.90 ± 2.96 6.71 ± 3.27 a)Platelets (x 109/L) 203.17 ± 85.19 245.45 ± 135.36CRP (mg/dL) 4.10 (1.80-11.02) 6.46 (4.10-39.70) a)s-IL2R (nmol/L) 4177.78 ± 1745.00 4278.51 ± 1909.72 IL-6 (pg/mL) 5.95 (3.80-12.40) 14.00 (5.70-30.20) a) Albumin (g/dl) 3.94 ± 0.36 3.51 ± 0.46 a)D-dimer (ng/mL) 141.63 (89.13-214.75) 632.88 (339.75-967.88) a)tPA (ng/mL) 3.94 ± 2.21 6.99 ± 5.99 a)PAI-1 (ng/mL) 40.74 ± 22.37 50.37 ± 27.06tPA/PAI-1 ratio 0.14 ± 0.12 0.15 ± 0.08 Transferrin saturation (%) 25.19 ± 10.84 14.70 ± 7.63 a)Ferritin (ng/mL) 454.66 (246.00-634.00) 374.00 (120.00-452.00)Ktv 1.63 ± 0.27 1.50 ± 0.36Urea reduction (%) 25.78 ± 7.01 31.12 ± 11.97Protein catabolic rate (g/Kg/day) 0.87 ± 0.04 0.81 ± 0.03

*Results are presented as mean ± one standard deviation or as median values (inter-quartile range). a) p<0.05 vs. CKD patientswith an AVF

Fig. 1 - Correlation betweenD-dimer levels and inflam-matory markers in CKD pa-tients: CRP (A), albumin(B), s-IL2R (C) and IL-6(D). Circles represent pa-tients with an AVF and tri-angle patients with a CVC.



5

Costa et al

Multiple regression analysis identified the weight(b=-0.367; p=0.03), neutrophil count (b=0.399;p=0.001) and s-IL2R (b=0.279; p=0.022) as the in-dependent variables significantly associated with D-dimers (R2=0.427).

DISCUSSION

CKD patients showed anemia, which, as reported inthe literature (15), is mainly determined by the rateof renal insufficiency; and therefore, by the reduc-tion in the renal secretion of erythropoietin. Addi-tionally, CKD patients undergoing HD treatmentpresented with leukocytosis, neutrophilia and in-creased levels of IL-6, s-IL2R and CRP inflammatorymolecules in the serum, confirming the presence ofan inflammatory process.In our CKD patients, we found low levels of PAI-1 andnormal tPA plasma levels, suggesting that endothe-lial and fibrinolytic functions are not apparently al-tered in these patients (16). However, the tPA/PAI-1 ratio was increased in our CKD patients, and theincrease in this ratio has been described as a novelrisk marker for recurrent myocardial infarction (17).We also found an increase in D-dimers in our pa-tients; however, it is uncertain if the higher fibrinfragment D-dimer value in CKD patients is a result ofa higher activation of coagulation or of fibrinolysis. The higher tPA/PAI-1 ratio in CKD patients under-going HD could suggest a higher fibrinolysis in thesepatients because only PAI-1 presented a significantlylower value. However, we did not measure fibrinolyticactivity. On the other hand, an interesting cross-talkhas been proposed to exist between inflammationand coagulation (18). The higher inflammatory sta-tus observed in CKD patients could actually under-lie coagulation activation. For instance, CRP isknown to promote tissue factor synthesis (19), apotent procoagulant, by monocytes. The correla-tions and the multiple regression analysis that wefound between D-dimer levels and inflammatorymarkers, underlie this hypothesis. From that per-spective, the higher D-dimer levels in the CKD pa-tients can also be observed as a marker of the in-flammatory process in these patients. When comparing our results to that previously de-scribed, we found some conflicting results, namelyfor PAI-1 that was been described to be increased(20, 21) or normal (22). For tPA, we found descrip-tions of normal values, as we found in this work,and other describing higher levels (20). These dif-ferences could be due to the difference between thestudied CKD patient population and due to the typeof kit used in the laboratorial evaluation. Increase D-

dimer and tPA/PAI-1 ratio is the most consistentfinding in most of the studies. When we analyzed the results of the studied param-eters according to the type of VA used, the CVC pa-tients showed lower albumin and Hb values, andhigher tPA, D-dimers, CRP and IL-6 levels as com-pared to those with an AVF, suggesting that these pa-tients have an even higher inflammatory process.Our results also showed in CKD patients using aCVC as VA that they seem to require a higher rhEPOdose to achieve the target Hb levels, suggesting alsocross-talk of erythropoiesis with inflammation. It is important to note that we measured tPA antigen,which reflects both active free tPA and inert tPAbound to inhibitor(s) (23). Therefore, higher tPAvalues are not synonymous with higher activated fib-rinolysis; higher t-PA values could result from en-dothelial dysfunction (23). Although tPA was higherin CVC patients, the tPA/PAI-1 ratio remained un-changed suggesting a similar activation of coagula-tion and fibrinolysis. The higher D-dimer values aremore likely to result from enhanced coagulationprobably linked to a higher inflammatory stimulus,than to a higher fibrinolysis, as only a trend to higherPAI-1 levels was observed in CVC patients.In summary, our results showed an altered and across-talk[AQ: correct?] between hemostasis and in-creased inflammatory markers in CKD patients. Theincreased levels of D-dimer, tPA and inflammatorymarkers in CKD patients using a CVC, led us to pro-pose a relationship between the type of VA chosenfor HD, and the risk of thrombogenesis. It seems rea-sonable to assume that these patients could presentwith an increased risk for CVD events.

ACKNOWLEDGEMENTS

We are very grateful to FMC, Dinefro - Diálises e Nefrol-ogia, SA and Uninefro - Sociedade Prestadora de Cuida-dos Médicos e de Diálise, SA, and to their nurses for thetechnical support.

This study was supported by a PhD grant (SFRH/BD/27688/2006) attributed to E. Costa by FCT.

Conflict of interest statement: none declared.

Address for correspondence:Elísio CostaServiço de Bioquímica Faculdade de Farmácia da Universidade do PortoRua Aníbal Cunha, 1644099-030 PortoPortugal [email protected] or [email protected]



6


REFERENCES

1. Santoro A. Confounding factors in the assessment ofdelivered hemodialysis dose. Kidney Int 2000; 77(suppl): S19-27.

2. Ifudu O, Mayers JD, Cohen LS, et al. Correlates of vas-cular access and nonvascular access-related hospital-izations in hemodialysis patients. Am J Nephrol 1996;16: 118-23.

3. Schwab SJ, Harrington JT, Singh A, et al. Vascular ac-cess for hemodialysis. Kidney Int 1999; 55: 2078-90.

4. Miller PE, Carlton D, Deierhoi MH, Redden DT,Allon M. Natural history of arteriovenous grafts in he-modialysis patients. Am J Kidney Dis 2000; 36: 68-74.

5. Mandolfo S, Piazza W, Galli F. Central venous catheterand the hemodialysis patient: a difficult symbiosis. JVasc Access 2002; 3: 64-73.

6. Schindler R, Senf R, Frei U. Influencing the inflam-matory response of haemodialysis patients by cytokineelimination using large-pore membranes. NephrolDial Transplant 2002; 17: 17-9.

7. Caglar K, Peng Y, Pupim LB, et al. Inflammatory sig-nals associated with hemodialysis. Kidney Int2002; 62:1408-16.

8. Fishbane S. Hyporesponsiveness to recombinanthuman erythropoietin in dialysis patients. NephrolDial Transplant 2000; 29: 545-8.

9. Costa E, Lima M, Alves JM, et al. Inflammation, T-cellphenotype and inflammatory cytokines in chronickidney disease patients under hemodialysis and its re-lationship to resistance to recombinant human ery-thropoietin therapy. J Clin Immunol 2008; 28: 268-75.

10. Hung C, Chen Y, Chou C, Yang C. Nutritional and in-flammatory markers in the prediction of mortality inChinese hemodialysis patients. Nephron Clin Pract2005; 100: c20-6.

11. Wanner C, Zimmermann J, Schwedler S, Metzger T.Inflammation and cardiovascular risk in dialysis pa-tients. Kidney Int 2002; 61 (suppl): S99-102.

12. Agras PI, Baskin E, Cengiz N, Kirazli S, Saatci U,Ozbek N. Global fibrinolytic capacity in children ondialysis. Thromb Res 2005; 115: 185-9.

13. Yu A, Egberg N, Jacobson SH. Haemostatic compli-cations in haemodialysis patients: effect of type of vas-cular access and dialysis filter. Scand J Clin Lab Invest2003; 63: 127-34.

14. Kanno Y, Kobayashi K, Takane H, et al. Elevation of

plasma D-dimer is closely associated with venousthrombosis produced by double-lumen catheter inpre-dialysis patients. Nephrol Dial Transplant 2007;22: 1224-7.

15. Iseki K, Kohagura K. Anemia as a risk factor forchronic kidney disease. Kidney Int Suppl 2007; 107:S4-9.

16. Shireman PK, McCarthy WJ, Pearce WH, Shively VP,Cipollone M, Kwaan HC. Elevations of tissue-typeplasminogen activator and differential expression ofurokinase-type plasminogen activator in diseasedaorta. J Vasc Surg 1997; 25: 157-64.

17. Wiman B, Andersson T, Hallqvist J, Reuterwall C,Ahlbom A, deFaire U. Plasma levels of tissue plas-minogen activator/plasminogen activator inhibitor-1 complex and von Willebrand factor are significantrisk markers for recurrent myocardial infarction inthe Stockholm Heart Epidemiology Program(SHEEP) study. Arterioscler Thromb Vasc Biol 2000;20: 2019-23.

18. Schoenmakers S, Reitsma P, Spek C. Blood coagula-tion factors as inflammatory mediators. Blood CellsMol Dis 2005; 34: 30-7.

19. Carter AM. Inflammation, thrombosis and acute coro-nary syndromes. Diab Vasc Dis Res 2005; 2: 113-21.

20. Segarra A, Chacón P, Martinez-Eyarre C, et al. Circu-lating levels of plasminogen activator, and thrombo-modulin in hemodialysis patients: biochemicalcorrelations and role as independent predictors ofcoronary artery stenosis. J Am Soc Nephrol 2001; 12:1255-63.

21. Molino D, De Santo NG, Marotta R, Anastasio P,Mosavat M, De Lucia D. Plasma levels of plasminogenactivator inhibitor type 1, factor VIII, prothrombin ac-tivation fragment 1+2, anticardiolipin, and antipro-thrombin antibodies are risk factors for thrombosisin hemodialysis patients. Semin Nephrol 2004; 24:495-501.

22. Pawlak K, Zolbach K, Borawski J, et al. Chronic viralhepatitis C, oxidative stress and the coagulation/fib-rinolysis system in haemodialysis patients. ThrombRes 2008, doi:10.1016/j.thromres.2008.02.012.

23. Pradhan AD, LaCroix AZ, Langer RD, et al. Tissueplasminogen activator antigen and d-dimer asmarkers for atherothrombotic risk among healthypostmenopausal women. Circulation 2004; 110:292-300.



Part IV - Discussion and Conclusions



We investigated systemic changes associated with resistance

to rhEPO therapy in CKD patients under haemodialysis and rhEPO

therapies, with particular interest on inflammation, iron status, oxida-

tive stress and erythrocyte damage. The studies were performed in two

groups of haemodialysis patients, one classified as responder and the

other as non-responder to rhEPO therapy. The studied parameters in

CKD patients were also compared with those presented by a healthy

control group.

Patients were selected from two Haemodialysis Clinics (FMC,

Dinefro – Diálises e Nefrologia, SA, and Uninefro – Sociedade Presta-

dora de Cuidados Médicos e de Diálise, SA). Non-responders were

identified as those patients who required high doses of rhEPO (higher

than 300 �U/Kg/week of epoetin or 1.5 μg/Kg/week of darbopoietin-

α) to achieve target haemoglobin levels, between 11 and 12 g/dL. Re-

sponder patients were subsequently selected, in order to match with

the non-responder patients for age, gender, weight, body mass index,

mean time under haemodialysis, urea reduction ratio, urea Ktv and

parathyroid hormone serum levels. Blood samples were collected in

the second haemodialysis session of the week.

�n the last 10 responders and 10 non-responders patients se-

lected, we also made an evaluation before and after the haemodialysis

procedure, to assess the effect of haemodialysis in some of the evalu-

ated parameters.

The control group included healthy volunteers, presenting nor-

mal haematological and biochemical values, with no history of renal or



inflammatory diseases, and, as far as possible, age and gender matched

with the haemodialysis patients. We found some difficulties in select-

ing a control group exactly with the same age of the patient’s group, as

most of the patients presented an advanced age.

�n table V, we summarise the most important findings observed

for the haemodialysis patients (all patients vs controls) and for non-

responders patients (non-responders vs responders).

Table V - Most important findings in haemodialysis patients (as compared to

control) and in non-responders patients (as compared to responders).

All patients Non-responders

vs controls vs responders

Anaemia + +

Microcytosis - +

Hypochromia - +

Anisocytosis + +

Reticulocytosis + -

Lymphopenia + +

T-cell activation + -

Production of Th1 cytokines + -

Neutrophil activation + +

�nflammation + +

�ron metabolism disturbance + +

Alterations in erythrocyte

membrane protein composition + +

Brand 3 profile changes + +

(-) no differences; (+) statistically significant differences



Haemodialysis patients

To compensate the anaemia of CKD patients under haemodi-

alysis, mainly due to renal failure in EPO production, administration

of rhEPO is nowadays included in the therapeutic protocols for these

patients. However, the dose chosen to be used should only targets to

partially correct the anaemia. �n fact, the pre-dialysis levels of haemo-

globin higher than the target levels (11-12 g/dL) are not desirable, due

to the risk associated with the effects arising from post-dialysis haemo-

concentration and its related complications (Strippoli, 2003). �n our

patients, we found an increase of about 11% in haemoglobin concen-

tration after haemodialysis. This increase is associated to a transloca-

tion of erythrocytes from the splanchnic circulation (and possibly from

the splenic circulation), in order to compensate the hypovolemic stress

during dialysis ultrafiltration (Yu, 1997; Dasselaar, 2007).

Our CKD patients under haemodialysis and rhEPO therapies pre-

sented lower haemoglobin concentration, haematocrit and erythrocyte

count when compared with controls (paper V��). These reductions are

regenerative, as showed by the increase in reticulocyte count and in re-

ticulocyte production index (RP�), reflecting the erythropoietic stimu-

lus triggered by rhEPO therapy; these changes are also associated with

anisocytosis, as suggested by the high red blood cell distribution width

(RDW) values. This anisocytosis could result from the observed reticu-

locytosis and/or from an erythrocyte damage induced by the haemodi-

alysis procedure or even from changes induced by plasma metabolites

and/or leukocyte activation products. This last hypothesis was later

tested by evaluating the erythrocyte changes and leukocyte activation

induced by the haemodialysis procedure (paper ��).

The haemodialysis patients also presented lymphopenia, which



results, at least in part, from a decrease in total circulating CD3+ T-lym-

phocytes and affect both the CD4+ and the CD8+ T-cell subsets (pa-

per �). Furthermore, the lymphocytes presented markers of enhanced

continuous activation [increased proportions of both early (HLA-DR)

and late-activation (CD57) markers on both CD4+ and CD8+ T-cell

subsets]. This, probably, results from persistent antigen stimulation/

chronic inflammation associated with haemodialysis and/or chronic

renal failure, which probably also justifies the increase of the fraction of

memory effectors T-cells / large granular lymphocytes, as shown by the

higher percentage of CD4+/CD28- and CD8+/CD28- T-cells observed.

Moreover, T-cells from haemodialysis patients showed an enhanced

ability to produce Th1 related cytokines (�L-2, �NF-γ and TNF-α) after

short term in vitro stimulation. This increased capacity to produce Th1

cytokines could justify, at least in part, the anaemia found in these

patients. �n fact these cytokines are described as inhibitors of erythro-

poiesis (Cooper, 2003).

The presence of an inflammatory process (higher neutrophil

count, increased neutrophil/lymphocyte ratio, higher levels of �L-6,

soluble �L-2 receptor and CRP, and decreased albumin serum levels)

was a consistent finding in our haemodialysis patients (paper �V). The

causes for this inflammatory response are not well clarified. There are

several potential sources, including bacterial contamination of the di-

alyser, incompatibility with the dialyser membrane and the type of the

vascular access.

Our haemodialysis patients using CVC showed a higher inflam-

matory process, as they presented with lower albumin and higher CRP

and �L-6 levels as compared to those with AVF. However, the use of

CVC could not be the only factor associated to this increase in the in-



flammatory process, because the patients using the AVF also presented

increased inflammatory markers, when compared with controls.

Besides the higher neutrophil count, as compared to controls,

haemodialysis patients also presented with increased plasma levels of

neutrophilic elastase (paper ��). However, the elastase/neutrophil ratio

was similar to that found in controls, suggesting that the rise in elastase

levels could be related to the increased neutrophil count and be part of

the inflammatory process.

�t was reported (Brinkmann, 2004) that elastase modulates cy-

tokine expression at epithelial and endothelial surfaces, up-regulating

cytokines such as the pro-inflammatory �L-6, and, therefore, elastase

may be one of the major factors involved in the development of sev-

eral chronic inflammatory diseases, although it is difficult to conclude

if increased elastase plasma levels are the cause or the consequence

of an inflammatory process. We figured, therefore, that the potential

role for the high levels of elastase observed in CKD patients should be

considered and clarified its origin.

During the haemodialysis procedure a neutrophilic activation

process occurs, as shown by the increase in elastase and lactoferrin

concentrations (paper ��) observed immediately after the haemodialy-

sis procedure (and in their ratios per neutrophil). However, in contrast

to elastase, plasma levels of lactoferrin were not increased before the

haemodialysis procedure, probably because lactoferrin could be used

or cleared from circulation in a pronounced way. �ndeed, lactoferrin is

involved in iron metabolism, and could be involved in the decreased

iron availability for erythropoiesis. Moreover, elastase, as a serinopro-

tease, may modulate pathogenic changes in plasmatic constituents and

in cells, namely, in erythrocytes.



The inflammatory stimulus has an important impact in iron me-

tabolism, by mobilizing iron from erythropoiesis traffic to storage sites

within the reticuloendothelial system, inhibiting erythroid progenitor

proliferation and differentiation (Jacober, 2007). This change in iron

mobilization leads to an iron depleted erythropoiesis. �n our haemo-

dialysis patients, we found an increase in ferritin serum levels and a

decrease in transferrin. These results suggest an increase of iron storage

and a reduction in iron mobilization for erythropoiesis. An increased

level of prohepcidin, the precursor molecule of hepcidin, was also

found in haemodialysis patients. �L-6 strongly induces hepcidin mRNA

expression (Flemming and Sly, 2001; Wrighting and Andrews, 2006),

indicating that the high levels of prohepcidin found in haemodialy-

sis patients could be related to the underlying chronic inflammation.

The increased hepcidin expression during inflammation and infection

explains sequestration of iron in the macrophages and inhibition of

intestinal iron absorption, the two hallmarks of the anaemia of inflam-

mation, which is normocytic or microcytic and hypocromic (Dallalio,

2003; Kulaksiz, 2004; Hsu, 2006). However, no correlations between

prohepcidin and the studied inflammatory markers were found, sug-

gesting that other factors than just inflammation regulate prohepcidin

serum levels. We identified the haemodialysis procedure, which re-

moves approximately 10% of prohepcidin from circulation, as of one

of these factors (paper V).

As previously referred, the high RDW values (anisocytosis)

could result from reticulocyosis or erythrocyte damage. Actually, some

changes were observed in erythrocyte membrane protein composi-

tion of our haemodialysis patients, being the decrease in spectrin the

most significant change (papers V�� and V��). This spectrin deficiency



may account for a poor linkage of the cytoskeleton to the membrane,

favouring membrane vesiculation, and, probably, a reduction in the

erythrocyte lifespan of haemodialysis patients (Reliene, 2002; Galla-

gher, 2005). Other alterations were also found, namely, increases of

protein bands 6 and 7, and in band 4.1/spectrin and spectrin/band 3

ratios which may further reflect an altered membrane protein interac-

tion and destabilization of membrane structure.

A primary protein deficiency, as occurs in Hereditary Sphero-

cytosis, may occur isolated or impose secondary deficiencies in one or

more proteins (Margetis, 2007), being the type and amount of the sec-

ondary protein deficiencies involved in the haematological and clinical

outcome of the disease (Rocha, 2005). Moreover, minor erythrocyte

membrane proteins may play a role in membrane vesiculation, espe-

cially when associated to deficiencies in major erythrocyte membrane

proteins (Rocha, 2005). The erythrocyte membrane protein alterations

that we observed could be related to the interaction with higher con-

centrations of several circulating metabolites, with leukocyte activation

products (Lucchi, 2000; Reliene, 2002; Stoya, 2002) and/or with the

haemodialysis procedure. The increased neutrophil activation products

that we found prior and after the haemodialysis procedure are able to

induce oxidative and proteolytic changes in erythrocytes, as reported

elsewhere (Santos-Silva, 1998). �ndeed, we found a different eryth-

rocyte membrane band 3 profile in haemodialysis patients, namely a

decrease in HMWAg, Pfrag and in Pfrag/band 3 monomer and HM-

WAg/band 3 monomer ratios (paper �X). This profile presents changes

reflecting the presence of younger erythrocytes, but also of damaged

erythrocytes. �ndeed, a decrease in HMWAg, accompanied by an in-

crease in Pfrag, is found in younger erythrocytes; however, the latter

change, in Pfrag, was not observed in haemodialysis patients.



�nflammatory conditions have been associated with oxida-

tive stress, leukocyte activation and oxidative changes in erythrocyte,

namely in band 3 profile (Santos-Silva, 1998; Belo, 2002).

Haemodialysis patients also exhibited higher plasma levels of to-

tal antioxidant status (TAS), lipidic peroxidation (TBA) and TBA/TAS

ratio (higher rise in TBA than in TAS), demonstrating an increased oxi-

dative stress in haemodialysis patients (paper �X). We also showed that

the haemodialysis procedure per se does not lead to an increase in the

studied markers of oxidative erythrocyte damage (paper �X). Actually,

no differences were found after haemodialysis, in band 3 profile and,

besides that, a decrease in membrane bound haemoglobin (MBH),

TBA and TAS were observed, probably reflecting the improvement in

the plasmatic environment and/or in the removal of oxidatively damaged

erythrocytes.

Alterations associated with resistance to rhEPO therapy

CKD patient’s non-responders to rhEPO therapy presented a

mild to moderate anaemia, even with the administration of higher

rhEPO doses (papers �,��,��,�V,V��). This anaemia is hypochromic [de-

creased mean cell haemoglobin (MCH) and mean cell haemoglobin

concentration (MCHC)], and was associated with a more accentuated

anisocytosis than in responder patients. Considering that all patients

were under iron and folate prophylactic therapies, and, that iron status

and vitamin B12 serum levels were normal, these changes can not be

attributable to a deficiency in these erythropoietic nutrients.

The CKD patients non-responders to rhEPO therapy seem

to present a “functional” iron deficiency, characterized by the pres-

ence of adequate iron stores as defined by conventional criteria, but



apparently with an inability to sufficiently mobilize iron to adequately

support erythropoiesis. �n fact, no statistically significant differences

were found in serum iron status markers between the groups of pa-

tients (responders and non-responders), except for the soluble transfer-

rin receptor (s-TfR), which was higher among non-responders (paper

�V). This “functional” iron deficiency is reflected by a decrease in MCH

and in MCHC, and by a more accentuated anisocytosis (high RDW) in

non-responders patients. Furthermore, the inverse correlation found

between CRP and mean cell volume, MCH, serum iron and transferrin

saturation, are in agreement with the idea that the chronic inflamma-

tion enhanced in non-responders, may lead to trapping of iron within

the macrophages and to a reduction in the availlability of iron. Ad-

ditionally, CKD patients non-responders to rhEPO therapy presented

lower prohepcidin serum levels when compared with responders (pa-

per �V). As previously referred, hepcidin is up-regulated by inflamma-

tion; on the other hand, erythropoietin downregulates liver hepcidin

expression (Vokurka, 2006). As non-responders patients were treat-

ed with much higher doses of rhEPO compared with responders, the

lower prohepcidin levels found in those patients could be explained by

this mechanism.

The inability of non-responders patients to sufficiently mobi-

lize iron to adequately support erythropoiesis, led us to hypothesize

that mutations/polymorphisms in DMT1 gene could be related to a

decrease in erythroid cell precursor’s iron uptake, which could result

in hypocromic and microcytic anaemia (Fleming, 1997). However, our

study failed to demonstrate a link between mutations/polymorphisms

in DMT1 gene and resistance to rhEPO therapy (paper V�). Moreover,

no influence of DMT1 gene mutations/polymorphisms in rhEPO doses,



haematological data or iron status was demonstrated in our haemodi-

alysis patients.

A more accentuated decrease in erythrocyte membrane spectrin,

though without statistically significance, was found in non-responders

CKD patients (paper V�). This spectrin deficiency was accompanied by

the production of hypocromic and anisocytic erythrocytes, and this an-

isocytosis could reflect an even more enhanced change in membrane

protein interactions in non-responders (as referred, a significant re-

duction in spectrin was observed when compared to control values).

However, a large overlapping of spectrin values was found in the two

groups of patients (responders and non-responders) confirming that the

disease itself and/or the interaction with haemodialysis membranes

could be involved in the observed erythrocyte membrane protein al-

terations (Sevillano, 1990; Martos, 1997). Non-responders patients also

showed a decrease in Pfrag and in Pfrag/band 3 monomer ratio and a

trend to higher values of MBH, suggesting that they present a higher

erythrocyte damage that may result from an even more adverse plas-

matic microenvironment (paper �X). When studying the differences

in membrane protein composition before and after the haemodialysis

procedure, we found that this spectrin reduction was enhanced in the

non-responder group (paper V��); moreover, we observed that this

could be associated to an enhanced inflammation process and/or to

the interaction with proteases released by neutrophils as we also found

significant rises in elastase and lactoferrin, both products of leukocyte

activation. We should not exclude the hypothesis that the much higher

doses of rhEPO used in non-responders can be involved in erythrocyte

membrane protein alterations.

Resistance to rhEPO therapy in haemodialysis patients seems,



therefore, to be associated to an increase in erythrocyte damage and

the previous studies on erythrocytes strongly suggest an important

role for the inflammatory process on such damage. Non-responders

patients presented an enhanced inflammatory process as demonstrated

by the higher CRP and lower albumin, as compared to responders (pa-

pers �V and X). This enhanced inflammatory stimulus induces the re-

lease of cytokines, including pro-inflammatory cytokines that can have

an erythropoiesis-suppressing effect. Moreover, CRP serum levels were

positively correlated with weekly rhEPO/Kg dose, suggesting that CRP

should be further studied as a biomarker of resistance to rhEPO ther-

apy in haemodialysis patients. Furthermore, these results confirmed

that resistance to rhEPO therapy is associated with an enhanced in-

flammatory state. To further clarify the role of inflammation in the

pathophysiologic process of resistance to rhEPO therapy we studied

the potential contribution of lymphocytes and neutrophils.

Statistically significant decreases in total lymphocytes and CD4+

T-cell counts were found in non-responder patients (paper �). There

are some possible explanations for this lymphocyte depletion, namely,

increased turnover, disturbances of lymphocyte homeostasis due to

uraemia, and increased peripheral lymphocyte apoptosis associated

with the activation stimulus (Bender, 1984; Waltzer, 1984; Sester,

2000; Libetta, 2001; Meyer, 2002; Litjens, 2006). The exact mecha-

nism of this T-cell depletion is still unclear. As �L-7 emerged recently as

a key cytokine involved in controlling the homeostatic turnover and

the survival of peripheral resting memory CD4+ T cells ( Litjens 2006),

we hypothesized that the serum levels of �L-7 could be related with

this decreased number of total lymphocytes and CD4+ T-cells. �n fact,

our results showed a relationship between increased levels of �L-7 and



decreased number of total lymphocytes and CD4+ T-cell count, sug-

gesting that the increased �L-7 serum levels could be related with on-

going T-cell depletion (paper ��). However, further studies are needed

to understand the mechanism of lymphocyte loss and its consequences

and/or involvement in the resistance to rhEPO therapy, as a higher �L-

7 concentration seems to predict a poorer response to rhEPO therapy.

Concerning neutrophils, we found slightly higher values for

non-responders, and plasma levels of elastase were enhanced in non-

responders (paper ��). The elastase/neutrophil ratio was similar to that

found in controls, suggesting that the rise in elastase levels could be

related to increased neutrophil count and be part of the inflammatory

process. Furthermore, a statistically significant positive correlation was

found between elastase levels and CRP, and the weekly rhEPO doses,

corroborating that hypothesis. However, no differences were found be-

tween responders and non-responders patients in neutrophilic activa-

tion during haemodialysis procedure, suggesting that the higher elas-

tase levels found in non-responders is likely to result from a chronic

and not acute stimulus triggered by the haemodialysis process, or to

the increased number of neutrophils found in these patients. Our re-

sults show that elastase, but not lactoferrin, may prove to be a good

marker of resistance to rhEPO therapy in haemodialysis patients.

In summary, although the etiology of resistance to rhEPO ther-

apy is still unknown, our work confirms that inflammation seems to

have an important role in its pathophysiology (Fig. 9). We also demon-

strated that resistance to rhEPO therapy is associated with “functional”

iron deficiency, lymphopenia and CD4+ lymphopenia, higher elastase

plasma levels, increased �L-7 serum levels, and alterations in erythro-

cyte membrane protein structure and in band 3 profile.



The exact origins of the inflammatory process remain unclear.

We speculate that the release of elastase during the haemodialysis pro-

cedure could have an important role in amplifying the inflammatory

process in haemodialysis patients, particularly in non-responders. Fur-

thermore, this elastase release could also have an important role in the

alterations found in erythrocyte membrane protein structure and in

band 3 profile, further contributing to worsening the anaemia. The in-

flammatory process, the rhEPO doses administrated and the lactoferrin

release during haemodialysis could have an important role in iron up-

Fig. 9 - Potential mechanisms involved in the pathophysiology of resis-

tance to rhEPO therapy. Black lines represents responders patients and red

lines non-responders patients; (+) represents activation and (-) inhibition.



take from the small intestine, in the release of iron from macrophages

and, finally, in the availability of iron for erythropoiesis, as proposed

in Fig. 11.

Our data suggest that further studies are needed to understand

the pathophysiology of resistance to rhEPO therapy.

�t will be important to clarify the mechanism of lymphocyte loss,

by studying the role of �L-7 levels and/or the peripheral lymphocyte

apoptosis; the effect of higher levels of elastase in the inflammatory

process, and in the alterations in erythrocyte membrane protein com-

position and in band 3 profile; to perform a full screening of mutations

in DMT1 gene and/or other genes related to iron metabolism.

�t would be also important to perform studies involving CKD

patients not yet under haemodialysis and/or rhEPO therapy. Of great

importance would be also the development of animal studies, such as

a rat model of chronic renal failure, as some histological and molecu-

lar studies can not be performed in humans. �t will be also important

to evaluate the importance of the type of vascular access in induc-

ing inflammation and endothelial dysfunction, and its association with

thrombotic risk factors and with higher risk for cardiovascular events

in haemodialysis patients.



Part V - References



Abdelrahman M, Sharples EJ, McDonald MC, Collin M, Patel NS, Yaqoob

MM, Thiemermann C. (2004). Erythropoietin attenuates the tissue injury associ-

ated with hemmorrhagic shock and myocardial ischemia. Shock 22:63-9.

Allen DA, Breen C, Yaqoob MM, Macdougall IC. (1999). �nhibition of CFU-

E colony formation in uremic patients with inflammatory disease: role of �NF-γ

and TNF-α. J Invest Med 47:204-11.

An X, Mohandas N. (2008). Disorders of red cell membrane. Br J Haematol

141:367-75.

Arth RH. (1999). �ron metabolism in end-stage renal disease. Seminars in Dialysis

12:224-30.

Association for the advancement of medical instrumentation. (2004).

ANS�/AAM� RD:2004: Dialysate for Hemodialysis. Arlington, VA.

Atanasio V, Manolescu B, Stoian I. (2006). Hepcidin – central regulator. Eur

J Haematol 78:1-10.

Bagnis C, Beaufils H, Jacquiud C, Adabra Y, Jouanneau C, Le Nahour

G, Jaudon MC, Bourbouze R, Jacobs C, Deray G. (2001). Erythropoietin

enhances recovery after cisplatin-induced acute renal failure in the rat. Nephrol

Dial Transplant 16:932-8.

Bahlmann FH, De Groot K, Haller H, Fliser D. (2004). Erythropoetin – is it

more than correcting anemia? Nephrol Dial Transplant 19:20-2.

Bárány P. (2001). �nflammation, serum C-reactive protein, and erythropoietin

resistance. Nephrol Dial Transplant 16:224-7.

Barber DA, Harris SR. (1994). Oxygen free radicals and antioxidants: a review.

Am Pharm NS34:26-35.



Barros E, Manfro RC, Thomé FS, Gonçalves LF. (2006). Nefrologia rotinas,

diagnóstico e tratamento. Artmed Editoras SA, Porto Alegre RS.

Baynes R, Bezwoda W, Bothwell T, Khan Q, Mansoor N. (1986). The non

immune inflammatory response: serial changes in plasma iron, iron binding ca-

pacity, lactoferrin, ferritin and C reactive protein. Scan J Clin Lab Invest 46:695-

704.

Belo L, Rebelo I, Castro EMB, Catarino C, Pereira-Leite L, Quintanilha

A, Santos-Silva A. (2002). Band 3 as a marker of erythrocyte changes in preg-

nancy. Eur J Haematol 69:145-51.

Bender BS, Curtis JL, Nagel JE, Chrest FJ, Kraus ES, Briefel GR, W.H.

Adler WH. (1984). Analysis of immune status of hemodialyzed adults: associa-

tion with prior transfusions. Kidney Int 26: 436-43.

Betteridge DJ. (2000). What is oxidative stress? Metabolism 49:3-8.

Birkmann J, Oez S, Smetak M, Kaiser G, Kappauf H, Gallmeier WM.

(1997). Effects of recombinant human thrombopoietin alone and in combina-

tion with erythropoietin and early-acting cytokines on human mobilized purified

CD34+ progenitor cells cultured in serum-depleted medium. Stem Cells 15:18-32.

Blagg CR. (1999). The early years of chronic dialysis: the seatle contribution.

Am J Nephrol 19:350-4.

Bogoyevitch MA. (2004). An update on the cardiac effects of erythopoietin

cardioprotetion by erythropoietin and the lessons learnt from studies in neuro-

protection. Cardiovasc Res 63:208-16.

Bowry SK. (2002). Dialysis membranes today. Int J Artif Organs 25:447-60.

Bondsdorff E, Jalavisto E. (1948). A humoral mechanism in anoxic erythro-

cytosis. Acta Physiol Scand 16:150-70.

Brenner BM. (2000). The kidney. 6th ed. W.B. Saunders, Philadelphia.

Brines M, Cerami A. (2005). Emerging biological roles for erythropoietin in the

nervous system. Nat Rev Neurosci 6:484-94.



Brines ML, Ghezzi P, Keenan S, Agnello D, de Lanerolle NC, Cerami C, Itri

LM, Cerami A. (2000). Erythropoietin crosses the blood-brain barrier to protect

against experimental brain injury. Proc Natl Acad Sci USA 97: 10526-31.

Brinkmann V, Reichard U, Goosmann C, Fauler B, Uhlemann Y, Weiss

DS, Weinrauch Y, Zychlinsky A. (2004). Neutrophil extracellular traps kill

bacteria. Science 303:1532-5.

Broudy VC. (1997). Stem cell factor and hematopoiesis. Blood 90:1345-64.

Buemi M, Allegra A, Corica F, Floccari F, D’Avella D, Aloisi C, Calapai

G, Iacopino G, Frisina N. (2000). �ntravenous recombinant erythropoietin

does not lead to an increase in cerebrospinal fluid erythropoietin concentration.

Nephrol Dial Transplant 15: 422-3.

Campana WM, Myers RR. (2001). Erythropoietin and erythropoietin recep-

tors in the peripheral nervous system: Changes after nerve injury. FASEB J 15:

1804-6.

Carbia-Nagashima A, Arzt E. (2004). �ntracellular proteins and mechanisms

involved in the control of gp130/JAK/STAT cytokine signaling. IUBMB Life 56:83-8.

Carlini RG, Reyes AA, Rothstein M. (1995). Recombinant human erythro-

poietin stimulates angiogenesis in vitro. Kidney Int 47: 740-5.

Carnot P, Deflandre C. (1906). Sur l´activite hemopoietique du serum au cours

de la regeneration du sang. C R Acad Sci (Paris) 143:384-6.

Casadevall N, Nataf J, Viron B, Kolta A, Kiladjian JJ, Martin-Dupont P,

Michaud P, Papo T, Ugo V, Teyssandier I, Varet B, Mayeux P. (2002). Pure

red-cell aplasia and antierythropoietin antibodies in patients treated with recom-

binant erythropoietin. N Engl J Med 346:469-75.

Ceciliani F, Giordano A, Spagnolo V. (2002). The systemic reaction during

inflammation: the acute-phase proteins. Curr Pharm Des 9:211-23.

Cerami A, Brines ML, Ghezzi P, Cerami CJ. (2001). Effects of epoetin alfa on

the central nervous system. Semin Oncol 28: 66-70.



Cooper AC, Mikhail A, Lethbridge MW, Kemeny DM, Macdougall IC.

(2003). �ncreased expression of erythropoiesis inhibiting cytokines (�FN-γ,

TNF-α, �L-10, and �L-13) by T cells in patients exhibiting a poor response to

erythropoietin therapy. J Am Soc Nephrol 14:1776-84.

Cooper AC, Breen CP, Vyas B, Ochola J, Kemeny DM, Macdougall IC.

(2003). Poor response to recombinant erythropoitin is associated with loss of

T-lymphocyte CD28 expression and altered interleukin-10 production. Nephrol

Dial Transplant 18:133-40.

Craddock PR, Fehr J, Brigham KL, Kronemberg RS, Jacob HS. (1977a).

Complement and leukocyte mediated pulmonary dysfunction in hemodialysis. N

Eng J Med 296:769-74.

Craddock PR, Hammerschmidt DE, White JG, Dalmasso AP, Jacob HS.

(1977b). Complement (C5a)-induced granulocyte aggregation in vitro: a pos-

sible mechanism of complement-mdiated leukostasis and leukopenia. J Clin Invest

60:260-4.

Dallalio G, Fleury T, Means RT. (2003). Serum hepcidin in clinical specimens.

Br J Haematol 122:996-1000.

Dame C, Bartmann P, Wolber E, Fahnenstich H, Hofmann D, Fandrey J.

(2000). Erythropoietin gene expression in different areas of the developing hu-

man central nervous system. Dev Brain Res 125: 69-74

Dasselaar JJ, Hooge MNL, Pruim J, Nijnuis H, Wiersum A, Jong PE, Huis-

man RM, Franssen CFM. (2007). Relative blood volume changes underesti-

mated total blood volume changes during hemodialysis. Clin J Am Soc Nephrol

2:669-74.

Daugirdas JT, Blake PG, Ing TS. (2003). Manual de diálisis. Masson SA,

Barcelona.

Deicher R, Horl WH. (2006). New insights into the regulation of iron homeo-

stasis. Eur J Clin Invest 36:301-9.



Depner TA. (1991). Prescribing hemodialysis: a guide to urea modeling. MA,

Kluwer Academic, Boston.

Descamps-Latscha B, Herbelin A, Nguyen AT, Roux-Lombard P, Zingraff

J, Moynot A, Verger C, Dahmane D, Groote D, Jungers P, Dayer JM. (1995).

Balance between �L-1β, TNF-α, and their specific inhibitors in chronic renal failure

and maintenance dialysis. J Immunol 154:882-92.

Dianzani I, Garelli E, Ramenghi U. (1996). Diamond-Blackfan anemia: a con-

genital defect in erythropoiesis. Haematologica 81:560-72.

Domenico I, Ward DM, Langelier C, Vaughn MB, Nemeth E, Sundquist

WI, Ganz T, Musci G, Kaplan J. (2007). The molecular mechanism of hepcidin-

mediated ferroportin down-regulation. Mol Biol Cell 18:2569-78.

Drueke TB, Eckardt KU. (2002). Role of secondary hyperparathyroidism in

erythropoietin resistance of chronic renal failure patients. Nephrol Dial Transplant

17(suppl 5):28-31.

Donahue RE, Emerson SG, Wang EA, Wong GG, Clark SC, Nathan DG.

(1985). Demonstration of burst-promoting activity of recombinant human GM-

CSF on circulating erythroid progenitors using an assay involving the delayed

addition of erythropoietin. Blood 66:1479-81.

Eschbach JW, Egrie JC, Downing MR, Browne JK, Adamson JW. (1987).

Correction of the anemia of end-stage renal disease with recombinant human

erythropoietin. Results of a combined phase � and �� clinical trial. N Engl J Med

316:73-8.

European Best Practice Guidelines. (2002). Section ��. Biocompatibility.

Nephrol Dial Transplant 17(suppl. 7): S32-S44.

Fitch PM, Roghanian A, Howie SEM, Sallenave JM. (2006). Human neu-

trophil elastase inhibitors in innate and adaptive immunity. Biochem Soc Transact

34:279-82.



Fleming MD, Trenor CC, Su MA, Foernzler D, Beier DR, Dietrich WF,

Andrews NC. (1997). Microcytic anaemia mice have a mutation in Nramp2, a

candidate iron transporter gene. Nature Genet 16:383-386.

Fleming RE, Sly WS. (2001). Hepcidin: a putative iron-regulatory hormone

relevant to hereditary hemochromatosis and the anemia of chronic disease. Proc

Nath Acad USA 98:8160-2.

Foley RN, Parfrey PS, Harnett JD, Kent GM, Murray DC, Barre PE. (1996).

The impact of anemia on cardiomyopathy morbidity and mortality in end-stage

renal disease. Am J Kidney Dis 28:53-61.

Foley RN, Parfrey PS, Sarnak MJ. (1998). Clinical epidemiology ofcardiovas-

cular disease in chronic renal failure. Am J Kidney Dis 32(suppl 3): S112-S119.

Fox SI. (2008). Human Physiology - 10th Edition. McGraw-Hill, Boston.

Fryer MJ. (2000). Vitamin E as a protective antioxidant in progressive renal

failure. Nephrology 5:1-7.

Furaso M, Munaretto G, Spinello M, Rebeschini M, Amici G, Gallieni M,

Picoli A. (2005). Soluble transferrin receptors and reticulocyte hemoglobin con-

centration in the assessment of iron deficiency in hemodialysis patients. J Nephrol

18:72-9.

Gallagher PG. (2005). Red cell membrane disorders. Hematology. Am Soc Hema-

tol Educ Program:13-8.

Gotch FA, Sargent JA. (1985). A mechanistic analysis of the National Coopera-

tive Dialysis Study (NCDS). Kidney Int 28:526-34.

Gottschalk CW, Fellner SK. (1997). History of the science of dialysis. Am J

Nephrol 17:289-98.

Gregory CJ, Eaves AC. (1978). Three stages of erythropoietic progenitor cell

differentiation distinguished by a number of physical and biologic properties.

Blood 51:527-37.



Gunnell J, Yeun JY, Depner TA, Kaysen GA. (1999). Acute-phase response

predicts erythropoietin resistance in hemodialysis and peritoneal dialysis pa-

tients. Am J Kidney Dis 33:63-72.

Harlan JM. (1985). Leukocyte-endothelial interactions. Blood 65:513-25.

Halliwell B, Gutteridge JMC. (1990). The antioxidants of human extracelular

fluids. Arch Biochem Biophys 280:1-8.

Himmelfarb J. (2004). Linking oxidative stress and inflammation in kidney dis-

ease: which is the chicken and which is the egg? Seminars in Dialysis 17: 449-54.

Hoenich NA, Levin R. (2003). The implications of water quality in hemodialy-

sis. Semin Dial 16:492-97.

Hsu SP, Ching CK, Chien CT, Hung KY. (2006). Plasma Prohepcidin posi-

tively correlates with haematocrit in chronic hemodialysis patients. Blood Purif

24: 311-6.

Ibrahim FF, Ghannam MM, Ali FM. (2002). Effect of dialysis on erythrocyte

membrane of chronically hemodialyzed patients. Renal Failure 24: 779-90.

Ifudu O, Mayers JD, Cohen LS, Paul H, Brezsnyak WF, Avram MM, Her-

man AI, Friedman EA. (1996). Correlates of vascular access and nonvascular

access-related hospitalizations in hemodialysis patients. Am J Nephrol 16: 118-23.

Imanishi H, Nakai T, Abe T, Takino T. (1986). Glutathione-linked enzyme

activities in red cell aging. Clin Chim Acta 159: 73-6.

Iscove NN. (1977). The role of erythropoietin in regulation of population size

and cell cycling of early and late erythroid precursors in mouse boné marrow. Cell

Tissue Kinet 10: 323-34.

Jacober ML, Mamoni RL, Lima CS, Dos Anjos BL, Grotto HZ. (2007).

Anaemia in patients with cancer: role of inflammatory activity on iron metabo-

lism and severity of anaemia. Med Oncol 24: 323-9.

Jacobson LO, Goldwasser E, Fried W, Ptzak L. (1957). Role of the kidney in

erythropoiesis. Nature 179: 633-4.



Jacobs K, Shoemaker C, Rundersforf R, Neill SD, Kaufman RJ, Mufson

A, Seehra J, Jones SS, Hewick R, Fritsch EF, et al.. (1985). �solation and

characterization of genomic and cDNA clones of human erythropoietin. Nature

313: 806-10.

Johnson DW, Pat B, Vesey DA, Guan Z, Endre Z, Gobe GC. (2006). Delayed

administration od darbepoetin or erythropoietin protects against ischemic acute

renal injury and failure. Kidney Int 69: 1806-13.

Kapiteijn K, Koolwijk P, Van Der WR, Helmerhorst F, Kooistra T, Van

Hinsbergh VW. (2001). Steroids and cytokines in endometrial angiogenesis. An-

ticancer Res 21: 4231-42.

Kay MM, Wyant T, Goodman J. (1994). Autoantibodies to band 3 during ag-

ing and disease and aging interventions. Ann N Y Acad Sci 719: 419-47.

Kemma E, Tjalsma H, Laarakkers C, Nemeth E, Willems H, Swinkels D.

(2005). Novel urine hepcidin assay by mass spectrometry. Blood 106:3268-70.

Kimel M, Leidy NK, Mannix S, Dixon J. (2008). Does epoetin alfa improve

health-related quality of life in chronically ill patients with anemia? Summary of

trials of cancer, H�V/A�DS, and chronic kidney disease. Value Health 11: 57-75.

Klemmer PJ. (2005). Calcium loading, calcium accumulation, and associated

cardiovascular risks in dialysis patients. Blood Purif 23 Suppl 1: 12-9.

Kolff WJ. (2002). Lasker clinical medical research award: the artificial kidney and its

effect on the development of other artificial organs. Nat Med 8: 1063-65.

Koury MJ, Ponka P (2004). New insights into erythropoiesis: the roles of fo-

late, vitamin B12, and iron. Annu Rev Nutr 24: 105-31.

Kulaksiz H, Gehrke SG, Janetzko A, Rost D, Bruckner T, Kallinowski B,

Stremmel W. (2004). Pro-hepcidin: expression and cell specific localisation in

the liver and its regulation in hereditary haemochromatosis, chronic renal insuf-

ficiency, and renal anemia. Gut 53: 735-43.



Kushner I, Rzewnicki D. (1999). Acute phase response, pp. 317-29. �n Gallin J�,

Snyderman R, Fearon DT, Haynes BF, Nathan C [eds.], �nflammation: basic prin-

ciples and clinical correlates, 3rd ed. Lippincott Williams & Wilkins, Philadelphia.

Lai PH, Everett R, Wang FF, Arakawa T, Goldwasser E. (1986). Structural

characterization of human erythropoietin. J Biol Chem 261: 3116-21.

Lenz O, Fornoni A. (2006). Chronic kidney disease care delivered by US fam-

ily medicine and internal medicine trainees: results from an online survey. BMC

Medicine 4: 30-9.

Levay PF, Viljoen M. (1995). Lactoferrin: a general review. Haematologica 80:

252-67.

Levin RD, Kwaan HC, Ivanivich P. (1978). Changes in platelet functions dur-

ing hemodialysis. J Lab Clin Med 92: 779-86.

Levin A, Li YC. (2005). Vitamin D and its analogues: do they protect against cardio-

vascular disease in patients with kidney disease? Kidney Int 68: 1973-81.

Ley K, Laudanna C, Cybulsky MI, Nourshargh S. (2007). Getting to the site

of inflammation: the leukocyte adhesion cascade updated. Nature Rev Immunol

7: 678-89.

Libetta C, Rampino T, Dal Canton A. (2001). Polarization of T-helper lympho-

cytes toward the Th2 phenotype in uremic patients. Am J Kidney Dis 38: 286-95.

Lin FK, Suggs S, Lin CH, Browne JK, Smalling R, Egrie JC, Chen KK, Fox

GM, Martin F, Stabinsky Z, Badrawi SM, Lai P, Goldwassert E. (1985).

Cloning and expression of the human erythropoietin gene. Proc Natl Acad Sci USA

82:7580-4.

Litjens NHR, van Druninger CJ, Betjes MGH. (2006). Progressive loss of

renal function is associated with activation and depletion of naïve T lymphocytes.

Clin Immunol 118: 83-91.



Locatelli F, Conte F, Marcelli D. (1998). The impact of hematocrit levels and

erythropoietin treatment on overall and cardiovascular mortality and morbidity

– the experience of the Lombardy dialysis registry. Nephrol Dial Transplant 13:

1642-4.

Locatelli F, Olivares J, Walker R, Wilkie M, Jenkins B, Dewey C, Gray SJ;

European/Australian NESP 980202 Study Group. (2001). Novel erythropoi-

esis stimulating protein for treatment of anemia in chronic renal insufficiency.

Kidney Int 60: 741-7.

Locatelli F, Manzoni C, Di Filippo S. (2002). The importance of convective

transport. Kidney Int Suppl 80: 115-20.

Locatelli F, Pisoni RL, Combe C, Bommer J, Andreucci VE, Piera L,

Greenwood R, Feldman HI, Port FK, Held PJ. (2004a). Anaemia in hae-

modialysis patients of five European countries: association with morbidity and

mortality in the Dialysis Outcomes and Practice Patterns Study (DOPPS). Nephrol

Dial Transplant 19: 121-32.

Locatelli F, Aljama P, Barany P, Canaud B, Carrera F, Eckardt KU, Horl

WH, Macdougal IC, Macleod A, Wiecek A, Cameron S. (2004b). Euro-

pean Best Practice Guidelines Working Group. Revised European best practice

guidelines for the management of anemia in patients with chronic renal failure.

Nephrol Dial Transplant 19 [Suppl 2]: ii1-ii47.

Lucchi L, Bergamini S, Botti B, Rapanà R, Ciuffreda A, Ruggiero P, Ballestri

M, Tomasi A, Albertazzi A. (2000). �nfluence of different hemodialysis membrane

on red blood cell susceptibility of oxidative stress. Artif Organs 24:1-6.

Macdougal IC. (1995). Poor response to erythropoietin: practical guidelines on

investigation and management. Nephrol Dial Transplant 10: 607-14.

Macdougall IC, Cooper AC. (2002). Erythropoietin resistance: the role of in-

flammation and pro-inflammatory cytokines. Nephrol Dial Transplant 17 (Suppl

11): 39-43.



Malyszko J, Malyszko JS, Pawlak K, Mysliwiec M. (2006). Hepcidin, iron

status, and renal function in chronic renal failure, kidney transplantation, and

hemodialysis. Am J Haematol 81: 832-7.

Mandolfo S, Piazza W, Galli F. (2002). Central venous catheter and the hemo-

dialysis patient: a difficult symbiosis. J Vasc Access 3: 64-73.

Margetis P, Antonelou M, karababa F, Loutradi A, Margaritis L, Papas-

sideri I. (2007). Physiologically important secondary modifications of red cell

membrane in hereditary spherocytosis – evidence for in vivo oxidation and lipid

raft protein modifications. Blood Cells Mol Dis 38: 210-20.

Martos MR, Hendry BM, Rodrígues-Puyol M, Dwight J, Díez-Marqués

ML, Rodríguez-Puyol D. (1997). Haemodialyser biocompatibility and erythro-

cyte struture and function. Clin Chim Acta 265: 235-46.

Marsden JT. (2006). Erythropoietin - measurement and clinical applications.

Ann Clin Biochem 43: 97-104.

Marti HH,Wenger RH, Rivas LA, Straumann U, Digicaylioglu M, Henn

V,Yonekawa Y, Bauer C, Gassmann M. (1996). Erythropoietin gene expres-

sion in human, monkey and murine brain. Eur J Neurosci 8: 666-76

Means RT Jr, Krantz SB. (1996). �nhibition of human erythroid colony form-

ing units by interferons α and β: differing mechanisms despite shared receptor.

Exp Haematol 24: 204-8.

Medina A, Ellis C, Levitt MD. (1994). Use of alveolar carbon monoxide mea-

surements to assess red blood cell survival in hemodialysis patients. Am J Hematol

46: 91-4.

Meier P, Dayer E, Blanc E, Wauters JP. (2002). Early T cell activation cor-

relates with expression of apoptosis markers in patients with end-stage renal

disease. J Am Soc Nephrol 13: 204-12.

Mellman I, Steinman RM. (2001). Dendritic cells: specialized and regulated

antigen processing machines. Cell 106: 255-8.



Miller PE, Carlton D, Deierhoi MH, Redden DT, Allon M. (2000). Natural

history of arteriovenous grafts in hemodialysis patients. Am J Kidney Dis 36: 68-

74.

Miyoshi I, Saito T, Bandobashi K, Ohtsuki Y, Taguchi H. (2006). Concomi-

tant deposition of aluminum and iron in bone marrow trabeculae. Intern Med

45: 117-8.

Miyake T, Kung CKH, Goldwasser E. (1977). Purification of human erythro-

poietin. J Biol Chem 252: 5558-64.

Morishita E, Masuda S, Nagao M,Yasuda Y, Sasaki R. (1997). Erythropoi-

etin receptor is expressed in rat hippocampal and cerebral cortical neurons, and

erythropoietin prevents in vitro glutamateinduced neuronal death. Neuroscience

76: 105-16

Muller WA. (1999). Leukocyte-endothelial cell adhesion molecules in transen-

dothelial migration, pp. 585-592. In Gallin J�, Snyderman R, Fearon DT, Haynes

BF and Nathan C [eds.], �nflammation: basic principles and clinical correlates, 3rd

ed. Lippincott Williams & Wilkins, Philadelphia.

National Kidney Fundation/K-DOQ1. (1997). Clinical practice guidelines:

hemodialysis adequacy: peritoneal dialysis adequacy. Am J Kidney Dis 30 (Suppl

2): S1-S136.

Nemeth E, Valore EV, Territo M, Schiller G, Lichtenstein A, Ganz T. (2003).

Hepcidin, a putative mediator of anemia of inflammation, is type �� acute-phase

protein. Blood 101: 2461-3.

Nemeth E, Rivera S, Gabayan V, Keller C, Taudorf S, Pedersen BK, Ganz

T. (2004). �L-6 mediates hypoferremia of inflammation by inducing the synthesis

of the iron regulatory hormone hepcidin. J Clin Invest 113: 1271-6.

Nemeth E, Preza GC, Jun CL, Kaplan J, Waring AJ, Ganz T. (2006). The

N-terminus of hepcidin is essential for its interation with ferroportin: struture-

function study. Blood 107: 328-33.



Nicolas G, Chauvet C, ViattebL, Danan JL, Bigard X, Devaux I, Beau-

mont C, Kahn A, Vaulont S. (2002). The gene encoding the iron regulatory

peptide hepcidin is regulated by anemia, hypoxia, and inflammation. J Clin Invest

110: 1037-44.

Njajou OT, de Jong G, Berghuis B, Vaessen N, Snijders PJ, Goossens JP,

Wilson JH, Breuning MH, Oostra BA, Heutink P, Sandkuijl LA, van Duijn

CM. (2002). Dominant hemochromatosis due to N144H mutation of SLC11A3:

clinical and biological characteristics. Blood Cells Mol Dis 29: 439-43.

Obladen M, Diepold K, Maier RF. (2000). Venous and arterial hematologic

profiles of very low birth weight infants. European Multicenter rhEPO Study

Group. Pediatrics 106: 707-11.

Olofsson T, Olsson I, Venge P, Elgefors B. (1977). Serum myeloperoxidase

and lactoferrin in neutropenia. Scand J Haematol 18: 73-80.

Osterborg A. (2004). New erythropoietic proteins: rationale and clinical data.

Semin Oncol 34: 12.

Palmer E. (2003). Negative selection-clearing out the bad apples from the T-cell

repertoire. Nat Rev Immunol 3: 383-91.

Park CH, Valore AJ, Waring AJ, Ganz T. (2001). Hepcidin, a urinary antimi-

crobial peptide synthesized in the liver. J Biol Chem 276: 7806-10.

Pisoni RL, Bragg-Gresham JL, Young EW, Akizawa T, Asano Y, Locatelli F,

Bommer J, Cruz JM, Kerr PJ, Mendelssohn DC, Held PJ, Port FK. (2004).

Anemia management and outcomes from 12 countries in the dialysis outcomes

and practice patterns study (DOPPS). Am J Kidney Dis 44: 94-111.

Pupim LB, Himmelfarb J, McMonagle E, Shyr Y, Ikizler TA. (2004). �nflu-

ence of initiation of maintenance hemodialysis on biomarkers of inflammation

and oxidative stress. Kidney Int 65: 2371-9.



Reliene R, Marini M, Zanella A, Reinhart WH, Ribeiro ML, del Giudice

EM, Perrotta S, Ionoscon A, Eber S, Lutz HU. (2002). Splenectomy prolongs

in vivo survival of erythrocytes differently in spectrin/ankyrin- and band 3-defi-

cient hereditary spherocytosis. Blood 100: 2208-15.

Rocha-Pereira P, Santos-Silva A, Rebelo I, Figueiredo A, Quintanilha

A, Teixeira F. (2004). Erythrocyte damage in mild and severe psoriasis. Br J

Dermatol 150: 232-44.

Rocha S, Rebelo I, Costa E, Catarino C, Belo L, Castro EMB, Cabeda JM,

Barbot J, Quintanilha A, Santos-Silva A. (2005). Protein deficiency balance

as a predictor of clinical outcome in hereditary spherocytosis. Eur J Haematol 74:

374-80.

Ronco C, Clark W. (2001). Factors affecting hemodialysis and peritoneal dialy-

sis efficiency. Semin Dial 14: 257-62.

Roos D, Van Bruggen R, Meischl C. (2003). Oxidative killing of microbes by

neutrophils. Microbes Infect 5, 1307-15.

Rosner MH. (2005). Hemodialysis for non-nephrologist. Southern Med J 98: 785-

91.

Rothstein DM, Sayegh MH. (2003). T-cell costimulatory pathways in allograft

rejection and tolerance. Immunol Rev 196: 85-108.

Saito N, Takemori N, Hirai K, Onodera R, Watanabe S, Naiki M. (1993).

Ultrastructural localization of lactoferrin in the granules other than typical sec-

ondary granules of human neutrophils. Human Cell 6: 42-8.

Santana MA, Rosenstein Y. (2003). What it takes to become an effector T cell:

the process, the cells involved, and the mechanisms. J Cell Physiol 195: 392-401.

Santoro A. (2000). Confounding factors in the assessment of delivered hemodi-

alysis dose. Kidney Int 58 [Suppl. 76]: S19-S27.

Santos-Silva A, Castro EMB, Teixeira NA, Guerra FC, Quintanilha A.

(1995). Altered erythrocyte membrane band 3 profile as a marker in patients at

risk for cardiovascular disease. Atherosclerosis 116: 199-209.



Santos-Silva A, Castro EMB, Teixeira NA, Guerra FC, Quintanilha A. (1998).

Erythrocyte membrane band 3 profile imposed by cellular aging, by activated neu-

trophils and by neutrophilic elastase. Clin Chim Acta 275: 185-96.

Schindler R, Senf R, Frei U. (2002). �nfluencing the inflammatory response

of hemodialysis patients by cytokine elimination using large-pore membranes.


Schwab SJ, Harrington JT, Singh A, Roher R, Shohaib S, Perrone R, Mey-

er K, Beasley D. (1999). Vascular access for hemodialysis. Kidney Int 55: 2078-90.

Scribner BH. (2002). Lasker clinical medicine research award: medical dilem-

mas: the old is new. Nat Med 8: 1066-7.

Sester U, Sester M, Hauk M, Kaul H, Kohler H, Girndt M. (2000). T-cell

activation follows Th1 rather than Th2 pattern in haemodialysis patients. Nephrol


Sevillano G, Rodrígues-Puyol M, Martos R, Duque I, Lamas S, Diez-

Marques ML, Lúcio J, Rodriguez-Puvol D. (1990). Cellulose acetato mem-

brane improves some aspects of red blood cell function in hemodialysis patients.


Sezer MT, Akin H, Demir M, Erturk J, Aydin ZD, Savik E, Tunc N. (2007).

The effect of serum albumin levels on iron-induced oxidative stress in chronic

renal failure patients. J Nephrol 20: 196-203.

Smith TG, Robbins PA, Ratcliffe PJ. (2008). The human side of hypoxia-in-

ducible factor. Br J Haematol 141:325-34.

Spalding E, Farrington K. (2003). Hemodiafiltration: current status. Nephron

Clin Pract 9: c87-c96.

Skikne BS, Cook JD. (1992). Effect of enhanced erythropoiesis on iron absor-

tion. J Lab Clin Med 120: 746-51.

Smrzova J, Balla J, Bárány P. (2005). �nflammation and resistance to eryth-

ropoiesis-stimulating agents- what do we know and what needs to be clarified?

Nephrol Dial Transplant 20 (Suppl 8): viii2-7.



Sommerburg O, Grune T, Hampl H, Riedel E, van Kuijk FJ, Ehrich JHH,

Siems WG. (1998). Does long-term treatment of renal anaemia with recombi-

nant erythropoietin influence oxidative stress in haemodialysed patients? Nephrol


Spittle MA, Hoenich NA, Handelman GJ, Adhikarla R, Homel P, Levin

NW. (2001). Oxidative stress and inflammation in hemodialysis patients. Am J

Kidney Dis 38:1408-13.

Stoya G, Klemm A, Baumann E, Vogelsang H, Ott U, Linss W, Stein G.

(2002). Determination of autofluorescence of red blood cells (RbCs) in uremic

patients as a marker of oxidative damage. Clin Nephrol 58: 198-204.

Strippoli GF, Manno C, Schena FP, Craig JC. (2003). Haemoglobin and hae-

matocrit targets for the anaemia of chronic renal disease (Cochrane review). Co-

chrane Database Syst Rev 1: CD003967.

Sullivan GW, Sarembock IJ, Linden J. (2000). The role of inflammation in

vascular diseases. J Leukoc Biol 67: 591-602.

Swinkels DW, Janssen MCH, Bergmans J, Marx JJM. (2006). Hereditary

hemochromatosis: genetic complexity and new diagnostic approaches. Clin Chem

52: 950-68

Taes YEC, Wuyts B, Boelaert JR, Vriese AS, Delanghe JR. (2004). Prohepci-

din accumulates in renal insufficiency. Clin Chem Lab Med 42: 387-9.

Taskapan H, Tam P, Au V, Chow S, Fung J, Nagai G, Roscoe J, Ng P, Sika-

neta T, Ting R, Oreopoulos DG. (2008). �mprovement in eGFR in patients with

chronic kidney disease attending a nephrology clinic. Int Urol Nephrol, in press.

Trey JE, Kushner I. (1995). The acute phase response and the hematopoietic

system: the role of cytokines. Crit Rev Oncol Hematol 21: 1-18.

Valenti L, Valenti G, Como G, Santorelli G, Dongiovanni P, Rametta R,

Fracanzani AL, Tavazzi D, Messa PG, Fargion S. (2008). HFE Genotype

�nfluences Erythropoiesis Support Requirement in Hemodialysis Patients: A Pro-

spective Study. Am J Nephrol 28: 311-316.



Vaziri ND. (2001). Cardiovascular effects of erythropoietin and anemia correc-

tion. Curr Opin Nephrol Hypertens 10: 633-7

Vesey DA, Cheung C, Pat B, Endre Z, Gobe G, Johnson DW. (2004). Eryth-

ropoietin protects against ischaemic acute renal injury. Nephrol Dial Transplant

19: 348-55.

Vokurka M, Krijt J, Sulc K, Necas E. (2006). Hepcidin mRNA levels in mouse

liver respond to inhibition of erythropoiesis. Physiol Res 55:667-74.

Vuillaume M. (1987). Reduced oxygen species, mutation, induction and câncer

initiation. Mutat Res 186: 43-72.

Wallace DF, Subramaniam VN. (2007). Non-HFE haemochromatosis. World J

Gastroenterol 13: 4690-8.

Waltzer WC, Bachvaroff RJ, Raisbeck AP, Egelandsdal B, Pullis C, Shen

L, Rapaport FT. (1984). �mmunological monitoring in patients with end-stage

renal disease. J Clin Immunol 4: 364-8.

Wang GL, Semenza GL. (1993). General involvement of hypoxia-inducible fac-

tor 1 in transcriptional response to hypoxia. Proc Natl Acad Sci USA 90: 4304-8.

Westenfelder C, Biddle DL, Baranowski RL. (1999). Human, rat, and mouse

kidney cells express functional erythropoietin receptors. Kidney Int 55: 808-20.

Witko-Sarsat V, Rieu P, Descamps-Latscha B, Lesavre P, Halbwachs-Mec-

arelli L. (2000). Neutrophils: molecules, functions and pathophysiological as-

pects. Lab Invest 80, 617-53.

Witz G. (1991). Active oxygen species as factors in multistage carcinogenesis.

Proc Soc Exp Biol Med 198: 675-82

Wrighting DM, Andrews NC. (2006). �nterleukin-6 induces hepcidin expres-

sion trough STAT3. Blood 108: 3204-9.

Wu SG, Jeng FR, Wei SY, Su CZ, Chung TC, Chang WJ, Chang HW. (1998).

Red blood cell osmotic fragility in chronically hemodialyed. Nephron 78: 28-32.



Yang CW, Li C, Jung JY, Shin SJ, Choi BS, Lim SW, Sun BK, Kim YS,

Kim J, Chang YS, Bang B. (2003). Preconditioning with erythropoietin pro-

tects against subsequent ischemia-reperfusion injury in rat kidney. FASEB J 17:

1754-5.

Yu AW, Nawab ZM, Barnes WE, Lai KN, Ing TS, Daugirdas JT. (1997).

Splanchnic erythrocyte content decreases during hemodialysis: a new compensa-

tory mechanism for hypovolemia. Kidney Int 51: 1986-90.

Zimmermann J, Herrlinger S, Pruy A, Metzger T, Wanner C. (1999). �n-

flammation enhances cardiovascular risk and mortality in hemodialysis patients.

Kidney Int 55: 648-58.

resistance to recombinant human erythropoietin therapy in ......elísio manuel de sousa costa | v |...

Documents