doi:10.1006/cyto.2000.0701, available online at http://www.idealibrary.com on

SHORT COMMUNICATION

ROLE OF HAEMODIALYSIS AND HEPATITIS CVIRUS INFECTION IN SPONTANEOUS AND

INDUCED CYTOKINE PRODUCTION OFPATIENTS WITH CHRONIC RENAL DISEASE

Julio Martın,1,3* Patricia de Sequera,2 Juan Antonio Quiroga,1,3 Miguel Rico,1,3

Mario Fernandez,1,3 Carlos Arocena,1 Carlos Caramelo,2 Vicente Carreno1,3

Cytokines modulate general and virus infection-related host immune responses. We haveinvestigated cytokine responses in chronic renal disease patients with regard to haemodialysisand hepatitis C virus (HCV) infection. Compared with healthy subjects with normal renalfunction (n=15), non-dialyzed/renal disease individuals without HCV infection (n=11) showedincreased production of tumour necrosis factor (TNF)-�, interleukin (IL-)6, IL-10, interferon(IFN-)� and IL-12 by blood mononuclear cells (P<0.05). These inflammatory cytokineresponses were abolished in haemodialysis patients (n=37; P<0.05), except for IL-12. Thishyporesponsiveness in haemodialysis patients was more evident in stimulatory conditions, asshown by the consistent inhibition of IFN-� production, and the failure of exogenous IFN-� toprime for IL-12 inducibility (P<0.01). The disturbed cytokine response appeared to focus in theT-helper lymphocyte phenotype 1 (Th1) because the stimulation of IL-6 and IL-10 (Th2phenotype cytokines) was not impaired. The pattern of response was similar among haemodi-alysis patients with (n=24) or without (n=13) HCV infection. However, HCV-positivehaemodialysis patients had a blunted TNF-� response (P<0.05) and failed to increase thestimulated IFN-� and IL-12 production (P<0.01) compared with chronic hepatitis C patientswithout renal disease (n=25). On the contrary, IL-10 stimulation was higher in HCV-positivehaemodialysis patients (P<0.01). These results disclose the presence in haemodialysis patients ofmarkedly abnormal general and HCV infection-related cytokine responses; the inhibitoryalterations appear to affect predominantly the stimulated responses via the Th1 subset and itsrelationship with monocyte response with possible pathogenic and therapeutic implications.

� 2000 Academic Press

The prevalence of hepatitis C virus (HCV)infection in haemodialysis (HD) patients is particularlyhigh (from less than 10% to more than 50%).1–3

Long-term dialysis results in an alteration of the

1248

cellular immunity.4 The pattern of cytokine productionin HD patients in relation with chronic HCV infectionis unknown. HD patients infected by HCV will beunder the triple influence of uremia, the haemodialysisprocedure and the HCV infection itself. Accordingly,we have hypothesized: (i) that uremia, as occurs inend-stage renal failure, may induce relevant changes inthe cytokine response; moreover, the HD proceduremay introduce further changes in the pattern ofcytokines; and (ii) within the group of HD patients,HCV infection may modify the cytokine profile in theinfected versus the non-infected individuals. To analysethese hypotheses, we have studied in vitro the produc-tion by peripheral blood mononuclear cells (PBMC) ofrepresentative cytokines from the T helper (Th) pheno-types 1 (IFN-�, IL-12) and 2 (IL-6, IL-10),5 as well asthe monokine TNF-�.

From the Departments of 1Hepatology and 2Nephrology, FundacionJimenez Dıaz, Madrid; 3Fundacion para el Estudio de lasHepatitis Virales, Madrid, Spain. *Present address: Departmentof Neurology, University of Pennsylvania, Clinical ResearchBuilding, Philadelphia, PA 19104-6146, USA

Correspondence to: Dr Vicente Carreno, Fundacion para el estudiode las hepatitis virales, Guzman El Bueno 72, 28015 Madrid,Spain; E-mail: fehr@tdi.es

Received 20 July 1999; received in revised form 27 January 2000;accepted for publication 5 April 2000

� 2000 Academic Press1043–4666/00/081248+05 $35.00/0

KEY WORDS: cytokines/haemodialysis/chronic renal disease/hepatitis C virus infection

CYTOKINE, Vol. 12, No. 8 (August), 2000: pp 1248–1252

Haemodialysis, HCV infection and cytokines / 1249

RESULTS

Effects of uremia and dialysis on cytokineproduction

TNF-�, IL-6 and IL-10 productions were differentamong groups: HD resulted in diminished productionsof TNF-�, IL-6, IL-10, IFN-� and IL-12 comparedwith the non-HD/CRD individuals. Non-HD/CRDpatients produced more TNF-� (mean�SEM: 8.9�3.4 ng/ml) compared with HD groups (1.8�0.4 ng/ml)and donors (1.3�0.5 ng/ml) (P<0.05; Fig. 1A). Simi-larly, IL-10 production was higher in non-HD/CRD(879�177 pg/ml) vs HD patients (334�50 pg/ml) ordonors (400�60 pg/ml) (P<0.05; Fig. 1C). Also,IL-6 production was higher in non-HD/CRD (108�33 ng/ml) vs HD patients (52�9 pg/ml) or controls(75�11 pg/ml) (P<0.05; Fig. 1B). Again, non-HD/CRD patients produced more IFN� (6.4�4.9 IU/ml;Fig. 1D) and IL-12 (33.1�20.2 pg/ml; Fig. 2) com-pared with the HD individuals (0.9�0.7 IU/ml and3.1�0.5 pg/ml, respectively; P<0.05).

LPS stimulation increased (P<0.01) the produc-tion of TNF-�, IL-6 and IL-10 (Fig. 1) but differencespersisted in TNF-� and IL-10 among groups:stimulation-induced TNF-� production was counter-acted by HD (P<0.01; Fig. 1A). However, HD patientsshowed the greater augment in IL-10 productioncompared with non-HD/CRD or donors (P<0.01;Fig. 1C). IL-6 production was similar among groups.The stimulatory responses were abolished regardingIFN-� and IL-12. IFN-� production failed to increasein HD patients in response to LPS (Fig. 1D). IL-12production increased in non-HD/CRD (209.2�145.4 pg/ml) compared with HD patients (5.9�1.8 pg/ml; P<0.01) in whom a markedly impairedIL-12 inducibility was noted (Fig. 2).

The role of dialysis and HCV infection oncytokine responses

IL-10 inducibility was higher in HD/HCV-positive(1505�66 pg/ml) compared with HD/HCV-negative(1304�95 pg/ml; P<0.05) individuals (Fig. 1C). Con-versely, LPS-stimulated IL-6 production was slightlyhigher in the HD/HCV-negative patients (134�12 vs175�40 ng/ml; Fig. 1B). Thus, HCV does not appearto influence cytokine responses in HD patients. On adifferent level, however, HD/HCV-positive patientsshowed less LPS-stimulated TNF-� production(5.4�0.7 vs 11.9�1.7 ng/ml, P<0.05; Fig. 1A) butmore IL-10 secretion (1505�66 vs 1062�62 pg/ml,P<0.01; Fig. 1C) compared with chronic hepatitis C(CHC) patients. Finally, the HD/HCV-positive indi-viduals had a marked lack of responsiveness to IFN-�priming for IL-12 production compared with the non-uremic CHC patients (5.5�1.2 vs 117.4�37.8 pg/ml,P<0.01; Fig. 2).

DISCUSSION

In this work CRD induces increased spontaneousTNF-� and IL-12 overproductions that imply anupregulation of the inflammatory activity most likelydue to uremia. Patients on HD also suffer from achronic inflammation leading to oversecretion ofTNF-�, IL-1� and IL- 6.4,6–10 Indeed, our findingsfavour that HD results in a permanent activation ofmonocytes. However, HD patients showed diminishedcytokine production and some, albeit not all, of theircytokine responses to a secondary stimulus (LPS) wereconsistently inhibited compared with non-HD/CRDindividuals. This was particularly evident with IL-12inducibility, which was drastically impaired and wasalso evident with TNF-� and IFN-�, suggesting ageneralized disturbance of Th1 responses in HDpatients. Such alteration apparently does not affectthe Th2 response, as assessed by IL-10 and IL-6measurements.

Among HD patients, those HCV-infected pro-duced, both basally and after LPS-induction, moreIL-10 versus non-infected individuals. The behaviourof Th2-related cytokines (IL-10 and IL-6) is differentbecause their baseline higher PBMC production is nolonger evident in HD patients, indicating that CRDinduces a stimulated status, which is removed bydialysis. Unfortunately, we have been unable to recruita group of HCV-infected CRD patients not yet in HDwho would contribute to the clarification of whetherunstimulation of TNF-�, IL-12 and IFN-� in thesubset of HCV-positive individuals depends only onCRD or on CRD plus HD.

The marked differences in cytokine releasebetween CHC patients, controls and HD/HCV-positive individuals disclose the existence of additionaldifferences in the response to HCV infection dependingon the presence of CRD and HD. Our data illustratehow CHC primes PBMC to produce high amounts ofIL-12 and IFN-� upon stimulation11 and how thispriming mechanism is blocked in HD patients. HCVinfection promotes a predominant Th2-like (IL-10)over a Th1 (IFN-�) type response in HD. Thesecharacteristics may account for the finding that themanifestations of HCV-associated disease are milder inHD patients (e.g., normal or slightly increased ALTvalues)12 compared with CHC patients in whom Th1

responses are predominant.13 This phenomenon mightbe explained by the decreased Th1 activity shown inour study. Such dysregulated responses to HCV in HDopen other interesting aspects, such as the possiblerole of these cytokine responses when the HD indi-viduals are submitted to kidney transplantation andimmunosuppressive therapy.

1250 / Martín et al. CYTOKINE, Vol. 12, No. 8 (August, 2000: 1248–1252)

0LPS

20A

TN

F-α

(n

g/m

l)

Medium

4

8

12

16

0LPS

250B

IL-6

(n

g/m

l)

Medium

50

100

150

200

0LPS

2000C

IL-1

0 (p

g/m

l)

Medium

400

800

1200

1600

0LPS

20 D

IFN

-γ (

IU/m

l)

Medium

4

8

12

16

P < 0.01

P < 0.001

P < 0.01

P < 0.01

P < 0.001

P < 0.05

Figure 1. Spontaneous and LPS-stimulated TNF-� (A), IL-6 (B), IL-10 (C) and IFN-� (D) production by PBMC measured in culture supernatantsfrom: non-dialyzed (non-HD)/chronic renal disease (CRD; n=11, ); HD/HCV-negative (n=13, ); HD/HCV-positive (n=24, ); chronic hepatitisC (CHC) patients (n=25, ); and healthy individuals (n=15).

Results are expressed as mean value�standard error of the mean. Statistical analysis were performed by ANOVA (P values are shown).

Haemodialysis, HCV infection and cytokines / 1251

MATERIALS AND METHODS

PatientsForty-eight CRD patients were selected; among them,

six male and five female non-HD/CRD patients (HCV-negative) were recruited from the pre-dialysis outpatientclinic. The rest (24 male, 13 female) were included in aconventional HD programme (3–4 h, thrice weekly). Thediagnosis of renal disease has been reported elsewhere.14

Detection of serum HCV RNA was the criterium for HCVinfection. Accordingly, HD patients were grouped asHD/HCV-negative (n=13) and HD/HCV-positive (n=24).Demographics was comparable among CRD groups, as wasHD duration and membrane type between HD/HCV-positiveand -negative patients. Blood was always drawn before HD.Control groups were: (i) chronic hepatitis C patients (CHC,n=25) without known renal disease not having receivedantiviral or immunomodulating therapy; and (ii) healthyblood donors (HCV RNA/anti-HCV-negative; n=15). Noindividual had serological markers of hepatitis B or humanimmunodeficiency viruses infections. All patients gaveinformed consent to the study which conformed to theethical guidelines of the Declaration of Helsinki on HumanExperimentation.

Anti-HCV was assessed by immunoassay (OrthoDiagnostic Systems, Raritan, NJ, USA). Serum HCV RNAwas detected by reverse transcription and nested-polymerasechain reaction as described.15 Hepatitis B and humanimmunodeficiency virus markers were tested by immuno-assays (Abbott Labs., North Chicago, IL, USA).

PBMC culture and cytokine productionPBMC were isolated by gradient sedimentation from

heparinized venous blood, washed and suspended in RPMI1640 culture medium supplemented (Imperial Laboratories,Andover, UK).15 PBMC (2�106 viable cells/ml) were cul-tured in 12-well plates (Costar Corp., Cambridge, MA, USA)at 37�C with 5% CO2 and humidity for 48 h with mediumalone, or medium plus LPS (10 �g/ml; Sigma, St. Louis, MO,USA). For IL-12 expression PBMC were primed with 100 ng/ml IFN-� (Promega, Madison, WI, USA) for 16 h, followedby the addition of 10 �g/ml LPS until completion of 48 h. At

the culture end, supernatants were collected, centrifuged andstored at �80�C for measurement of TNF-�, IL-6, IL-10and IFN-� production by ELISA (Medgenix Diagnostics,Fleurus, Belgium). IL-12/p70 was assayed by ELISA asdescribed11 using reagents provided by M. K. Gately(Hoffmann-La Roche, Nutley, NJ, USA).

Statistical analysisDifferences between groups were studied by two-tailed

one-way analysis of variance and Duncan’s multiple rangemodified t-test using the SPSS statistical package (version6.0).

Acknowledgement

J.M. and P.S. contributed equally to this work.M.R. is a research fellow of Fundacion ConchitaRabago, Madrid. This study was supported byFundacion para el Estudio de las Hepatitis Virales andInstituto Reina Sofıa de Investigacion Renal, Madrid,Spain.

REFERENCES

1. Martin P, Friedman L (1995) Chronic viral hepatitis andthe management of chronic renal failure. Kidney Int 47:1231–1241.

2. Roth D (1995) Hepatitis C: the nephrologist’s view. Am JKidney Dis 25:3–16.

3. Pereira BJG, Levey AS (1997) Hepatitis C virus infection indialysis and renal transplantation. Kidney Int 51:981–999.

4. Descamps-Latscha B, Herbelin A (1993) Long-term dialysisand cellular immunity. Kidney Int 41(Suppl):S135–142.

5. Romagnani S, Parronchi P, D’Elios MM, Romagnani P,Annunziato F, Piccinni MP, Manetti R, Sampognaro S, Mavilia C,De Carli M, Maggi E, Del Prete GF (1997) An update on humanTh1 and Th2 cells. Int Arch Allergy Immunol 113:153–156.

6. Herbelin A, Nguyen AT, Zingraff J, Urena P, Descamps-Latscha B (1990) Influence of uremia and hemodialysis on circulat-ing interleukin-1 and tumor necrosis factor alpha. Kidney Int37:116–125.

7. Cavaillon JM, Poignet JL, Fitting C, Delons S (1992) Seruminterleukin-6 in long-term hemodialyzed patients. Nephron 60:307–313.

0.1IFN-γ/LPS

1000

IL-1

2 (p

g/m

l)

Medium

1

10

100

P < 0.05P < 0.05

Figure 2. Spontaneous and IFN-�-primed/LPS-stimulated IL-12 production by PBMC, measured in culture supernatants from non-HD/CRD(n=11, ); HD/HCV-negative (n=13, ); HD/HCV-positive (n=24, ); chronic hepatitis C (CHC) patients (n=25, ); and healthy individuals(n=15, ).

Results are expressed as mean value�standard error of the mean. Statistical analysis were performed by ANOVA (P values are shown).

1252 / Martín et al. CYTOKINE, Vol. 12, No. 8 (August, 2000: 1248–1252)

8. Mege JL, Olmer M, Purgus R, Bertocchio P, Farnarier C,Kaplanski S, Capo C, Bongrand P (1991) Haemodialysis membranesmodulate chronically the production of TNF alpha, IL1 beta andIL6. Nephrol Dial Transplant 6:868–875.

9. Chollet-Martin S, Stamatakis G, Bailly S, Mery JP,Gougerot-Pocidalo MA (1991) Induction of tumour necrosis factor-alpha during haemodialysis. Influence of the membrane type. ClinExp Immunol 83:329–332.

10. Girdnt M, Kohler H, Schiedhelm-Weick E, Schlaak JF,Meyer zum Buschenfelde KH, Fleischer B (1995) Production ofinterleukin-6, tumor necrosis factor � and interleukin-10 in vitrocorrelates with the clinical immune defect in chronic hemodialysispatients. Kidney Int 47:559–565.

11. Quiroga JA, Martın J, Navas S, Carreno V (1998)Induction of interleukin-12 production in chronic hepatitis C virusinfection correlates with the hepatocellular damage. J Infect Dis178:247–251.

12. Caramelo C, Ortiz A, Aguilera B, Porres JC, Navas S,Marriot E, Alberola ML, Alamo C, Galera A, Garron MP,Gonzalez-Parra E, Fernandez de Gabriel MV, Oliva H, Carreno V(1993) Liver disease patterns in hemodialysis patients with antibodiesto hepatitis C virus. Am J Kidney Dis 22:822–828.

13. Napoli J, Bishop GA, McGuinness PH, Painter DM,McCaughan GW (1996) Progressive liver injury in chronic hepatitisC infection correlates with increased intrahepatic expression ofTh1-associated cytokines. Hepatology 24:759–765.

14. Caramelo C, Alcazar R, Gallar P, Teruel JL, Velo M,Ortega O, Galera A, Da Silva M (1994) Choice of dialysis mem-brane does not influence the outcome of residual renal function inhaemodialysis patients. Nephrol Dial Transplant 9:675–677.

15. Martın J, Navas S, Quiroga JA, Carreno V (1996)Recombinant human granulocyte colony-stimulating factor reduceshepatitis C virus replication in mononuclear cells from chronichepatitis C patients. Cytokine 8:313–317.