Saudi Journal of Kidney Diseases and Transplantation

: 1994  |  Volume : 5  |  Issue : 2  |  Page : 163--167

Can Uremia and Hemodialysis Affect Plasma Levels of Circulating TNF-α

Mohsen Mohammed Hussein Al-Koussi1, Mohammed Ali Tohamy2, Inas Ahmed Reda El-Kareimy2, NAH Kamel2, Lubna Al-Menawy1, Iftikhar Ahmed Sheikh1, Faissal A.M Shaheen1,  
1 Jeddah Kidney Center, King Fahd Hospital, Jeddah, Saudi Arabia
2 Assyutt University Hospital, Assyutt, Egypt

Correspondence Address:
Faissal A.M Shaheen
Consultant Nephrologist and Director, Jeddah Kidney Center
Saudi Arabia


It is well known that uremia is associated with increased susceptibility to infection. In addition, patients on haemodialysis (HD) experience a variety of dialysis associated complications, both acute and chronic, many of them having features similar to acute phase response. Immunoregulatory cytokines such as tumor necrosis factor-a (TNF-a) have been implicated in the pathogenesis of immunological as well as inflammatory diseases. Thus, TNF-a levels could be expected to be high in uremic patients as well as in HD patients. We investigated the plasma levels of TNF-a in 17 patients with renal failure, seven patients with chronic renal failure (CRF) before commencement of HD and 10 patients maintained on regular HD. Eight age matched healthy subjects were studied as normal control. All CRF patients, who were not yet on dialysis, had high plasma levels of TNF-a (mean + SD 71.33 + 33.25 pg/ml). Out of the HD group, TNF-a plasma levels were not detectable in five patients and in the remaining five, TNF-a plasma level (mean + SD 21.06 + 7.72) were comparable to the normal controls (mean + SD 21 + 7.87). Our findings suggest that factors related to uremia, but not to HD, are responsible for high TNF-a plasma levels in these patients and that, HD probably has a beneficial effect by removal and/or neutralising of uremic toxins.

How to cite this article:
Al-Koussi MM, Tohamy MA, El-Kareimy IR, Kamel N, Al-Menawy L, Sheikh IA, Shaheen FA. Can Uremia and Hemodialysis Affect Plasma Levels of Circulating TNF-α.Saudi J Kidney Dis Transpl 1994;5:163-167

How to cite this URL:
Al-Koussi MM, Tohamy MA, El-Kareimy IR, Kamel N, Al-Menawy L, Sheikh IA, Shaheen FA. Can Uremia and Hemodialysis Affect Plasma Levels of Circulating TNF-α. Saudi J Kidney Dis Transpl [serial online] 1994 [cited 2020 Oct 20 ];5:163-167
Available from:

Full Text


Chronic renal failure (CRF) patients as well as the patients on haemodialysis is known to have an increased tendency of acquiring infectious diseases attributed to immunological deficiency [1],[2] . Chronic haemodialysis patients also experience a variety of distressing dialysis associated complications, both acute and chronic such as hypotension, fever, amyloidosis, and haemodialysis associated osteo-articular changes [3],[4],[5],[6] . In the recent past, considerable interest has been focused on the role of cytokines in the acute phase response of the inflammatory process. In this regard, three cytokines, interleukin-1 (IL-1), interleukin­6 (IL-6) and tumor necrosis factor-a (TNF­a) are of particular significance. Although they are different proteins transcribed from different genes, they share remarkable similarities in their biological properties [7],[8] . They all induce fever, hypotension, and inflammation when injected into the animals or human subjects [9] . Because of the similarity between the acute phase response and the post haemodialysis syn­drome, these cytokines have been impli­cated as the mediators of haemodialysis associated complications [10],[11],[12],[13],[14] . This concept has been proposed in the assumption that the human monocytes, which secrete TNF­a, are exposed to several inducers of cytokine release during haemodialysis such as membranes that are bio-incompatible and endotoxins transferred from the dialysate. Thus, TNF- α levels could be expected to be high in CRF patients due to the tendency for frequent infections as well as in HD patients due to the proposed monocyte activation during HD. This study, therefore, was designed to investigate the TNF- α levels both these group of patients.

 Materials and Methods

Seventeen patients with end stage renal failure (ESRF) were studied. None of these patients had clinical evidence of acute infection or any other disease which might contribute to alteration of TNF- α a levels. Drugs known to affect TNF- α levels such as steroids, antimicrobials, or non steroidal anti-inflammatory drugs [14] had not been given to any of them during the two weeks preceding the study.

Group I consisted of ten patients who were stabilised on haemodialysis (HD) for a period of at least 10 months (mean + SD 22.6 + SEM 16.1 months) prior to recruit­ment to the study [Table 1]. The duration of intermittent haemodialysis sessions ranged from 3.5 to 4 hours per session, thrice weekly. All were on dialysis with the same type of membrane, cuprammonium rayon, with surface area 1 m 2 sterilised with ethylene oxide gas. The dialysis baths for all these patients were acetate with bacterial colony counts less than 10 5 CFU/ml. Group II consisted of seven patients with end stage renal disease, not yet initiated on regular dialysis treatment. Blood samples were taken just before commencement of their first dialysis session. The control group consisted of eight healthy volunteers [Table 1].

TNF-α was determined using a kit purchased from Genzyme Corporation (Cambridge MA). The Predicta™ TNF-a enzyme immuno assay was containing 96­ well micro titer plate pre-coated with mono­clonal antibody to TNF -α . A measured volume of the sample was added to each test well and incubated to allow any TNF- a present to be bound by antibodies on the micro-titer plate. The wells were then washed and a biotin labelled polyclonal antibody to TNF- α was added which gets bound to the captured TNF-α. The wells were then washed and a peroxidase labelled sterptavidin reagent was added which gets attached to biotin in the immune complex on the plate. The wells were washed and substrate (peroxide) and chromagen were added which produces a blue color in the presence of peroxidase enzyme. The color reaction was then stopped by the addition of acid which changes the blue color to yellow. The intensity of the yellow color is in direct proportion to the amount of TNFa present in the samples. The absorbance was read at 450 nm on DIAMEDX-BP 96 ELISA reader.


The plasma levels of TNF-α in group I and II are shown in [Table 1]. The mean (+ SD) values of plasma TNF-a level in the eight healthy subjects was 21 (+ 7.87) pg/ml. The plasma levels of TNF- a in 7 uremic patients before commencement of their first dialysis session were significantly higher compared with the normal controls (p 0.05). However, when it was compared with the uremic patients not yet started on renal replacement therapy, it was highly significant (p [11],[16] as well as in vivo studies [17] on TNF-a levels in uremic non-dialyzed patients. The TNF- 0 elevation in this group of patients may be explained by monocyte activation due to factors related to uremia [11],[16] , although the need for adequate renal function for proper TNF-a metabolism should also be considered.

It has been claimed that the haemodialysis procedure might induce the release of TNF­α from monocytes and this stimulus could delayed and/or prolonged which explains the higher plasma TNF- α levels in pre­dialysis samples, as reported by some authors [18],[19] . The chronic release of TNF- a in haemodialysis patients could stimulate β- 2 -microglobulin production resulting is the genesis of dialysis associated amyloidosis [20] . The results of our study, however, did not support these hypothesis. Plasma TNF- α levels were not detected in five of our ten haemodialysis patients and as to the remaining five, it was comparable to the observed levels of normal controls. Several other reports also deny any role of the haemodialysis procedure in inducing TNF- α production [16],[21],[22],[23] .

Some investigators reported slight drop of TNF- α levels production after single haemodialysis session, but they suggested that the elevation of TNF- α levels occurring during or after dialysis is a response to activation of monocytes by membrane polymer which does not occur immediately [18] . They support their suggestion by the fact that values of TNF- α in continuous ambulatory peritoneal dialysis (CAPD) patients are comparable to the normal controls. However, our impression is that because of the fluctuation of the uremic state with haemodialysis procedure, there is also a consequent fluctuation in TNF- α levels which may be directly proportionate to the uremic state and its absence in CAPD might be related to the continuous removal of the uremic toxins.

The discrepancy between our results and reports showing high levels of TNF- α in HD patients may probably be attributed to the differences in the duration of HD treatment and in the dialysis procedure itself including the type and quality of dialysate used [24] and purity of water from bacterial contaminants. The link between TNF- α and post-haemodialysis complications may lot be accepted because of the continuous release of TNF- α antagonist TNF-binding protein-1 as well as the use of heparin during the haemodialysis procedure, which are known to bind TNF- α and block its action [9] .

In conclusion, it appears that the haemodialysis process has, in fact, a beneficial effect on ESRD patients by reducing or stopping TNF- α release from monocytes by removal of uremic toxins and by enhancing TNF- α antagonists release during the dialysis procedure which, at least in part, abolishes the deleterious effect of TNF- α in these patients.


1Ringoir S, Vanholder R. Phagocyte function in uremic patients. Verh K Acad Geneeskd Belg, 1991;53(3):189-203.
2Keane WF, Shapiro FL, Raij L. Incidence and type of infections occuring in 445 chronic haemodialysis patients. Trans Am Soc Artif Intern Organs 1977;23:41-7.
3Kleinman KS, Coburn JW. Amyloid syndromes associated with haemodialysis. Kidney Int 1989;35:567-75.
4DiRaimondo CR, Casey TT, DiRaimondo CV Stone WJ. Pathologic fractures associated with idiopathic amyloidosis of bone in chronic haemodialysis patients. Nephron 1986;43:22-7.
5Degoulet P, Reach I, Rozenbaum W, et al. Society of Nephrology, Computer Technology Commission. Dialysis computer program. VI - Survival and risk factors J Urol Nephrol Paris 1979;85:909-62.
6Weingast JA, VanDeKerkhove KM, Eiger SM, Kluger MJ, Port FK. Release of pyrogens during clinical haemodialysis. Trans Am Soc Artif Intern Organs 1985;31:359-62.
7Oppenheim JJ, Matshushima K, Yoshimura T, Leonard ED. The activities of cytokines are pleiotropic and interdependent. Immunol lett 1987;16:179-84.
8Le J, Vilcek J. Tumor necrosis factor and inter- leukin-1: Cytokines with multiple overlapping biological activities. Lab Invest 1987;56:234-48.
9Dinarello CA. Interleukin-1 and tumor necrosis factor and their naturally occuring antagonists during haemodialysis. Kidney Int 1992;supp38:S68- 77.
10Henderson LW, Koch KM, Dinarello CA, Shaldon S. Haemodialysis hypotension: the interleukin hypothesis. Blood purif 1993;l:3-8.
11Herbelin A, Urena P, Nguyen AT, Zingraff J. Influence of first and long term dialysis on uremia- associated increased basal production of interleukin-1 and tumour necrosis factor alpha by circulating monocytes. Nephrol Dial Transplant 1991;6(5):349-57.
12Pertosa G, Marfellsa C, Tarantino EA, et al. Involvement of peripheral blood monocytes in haemodialysis: in vivc induction of tumour necrosis factor alpha, interleukin 6 and beta 2­micro- globulin. Nephrol Dial Transplant 1991;6 suppl 2:18-23.
13Lonneman G, Bingel M, Koch KM, Shaldon S, Dinarello CA. Plasma interleukin-1 activity in humans undergoing haemodialysis with regenerated cellulosic membranes. Lymphokine Res 1987;6(2):63-70.
14Chollet Martin S, Stamatakis G, Bailly S, Mery JP, Gougerot-Pocidalo MA. Induction of tumour necrosis factor-alpha during haemodialysis. Influence of the membrane type. Clin Exp Immunol 1991;83(2):329-32.
15Fieren MW, Van den Bemd GJ, Efraim S, Bonta IL. Prostaglandin E2 inhibits the release of tumor necrosis factor-alpha, rather than interleukin 1 beta, from human macrophages Immunol lett 1992;31(l):85-90.
16Powell AC, Bland LA, Oettinger CW, et al. Enhanced release of TNF-alpha but not IL-1 beta, from uremic blood after endotoxin stimulation. Lymphokine Cytokine Res 1991;10(5):343-6.
17Peona C, Fortina A, Colombo P, et al. Levels of tumour necrosis factor and interleukin-1 in uremia and in haemodialysis. Minerva Urol Nefrol 1991;43(3):175-9.
18Ryan J, Beynon H, Rees AJ, Cassidy MJ.Evaluation of the in vitro production of tumour necrosis factor by monocytes in dialysis patients. Blood-Purif 1991 ;9(3):142-7.
19Annenkov AY, Strokov AG, Baranova FS. Alterations in mononuclear cell tumour necrosis factoralpha (TNF-alpha) response in patients on long term cuprophane haemodialysis. Clin Exp Immunol 1992;90(l):49-55.
20Ghysen J, De-Plaen JF, Van Ypersele de strihou C. The effect of membrane characteristics on tumour necrosis factor kinetics during haemodialysis. Nephrol Dial Transplant 1990;5(4):270-4.
21Caruana RJ, Lobel SA, Leffell MS, Campbell H, Cheek PL. Tumour necrosis factor, interleukin-1 and beta 2-microglobulin levels in choronic haemodialysis patients. Int J Artif Organs 1990;13(12):794-8.
22Powell AC, Bland LE, Oettinger CW, et al. Lack of plasma interleukin-1 beta or tumour necrosis factor alpha elevation during unfavourable haemodialysis conditions. J Am Soc Nephrol 1991;2(5):1007-13.
23Vaziri ND, Kaupke CJ, Yousefi S, Gonzales E, Cesario TC. Cytokine levels during haemodialysis. ASAIO Trans 1991;37(3):389-91.
24Anderson J, Briefel G, Jones JM, Ryu JH, McGuire M, Yun YP. Effects of acetate dialysate on transforming growth factor beta-1, interleukin, and beta 2-microglobulin plasma levels. Kidney Int 1991;40(6):1110-7.