Clinical relevance of cytokine production in hemodialysis

1. Kidney International, Vol. 58, Suppl. 76 (2000), pp. S-104–S-111
Clinical relevance of cytokine production in hemodialysis
GIOVANNI PERTOSA, GIUSEPPE GRANDALIANO, LORETO GESUALDO,
and FRANCESCO PAOLO SCHENA
Division of Nephrology, Department of Emergency and Transplantation, University of Bari, Policlinico, Bari, Italy
Clinical relevance of cytokine production in hemodialysis. ates a complex of acute and chronic side effects also
Blood–dialyzer interaction in hemodialysis has the potential known as “bioincompatibility phenomena” [4, 5].
to activate mononuclear cells leading to the production of
Many aspects regarding morbidity and mortality of
inflammatory cytokines. The extent of activation is dependent
dialysis patients may be related to these cellular events
on the dialyzer material used and is considered an index of
and, in particular, to the production of cytokines by pe-
biocompatibility. Cytokines, such as interleukin-1␤ (IL-1␤),
tumor necrosis factor-␣ (TNF-␣), and IL-6, may induce an ripheral blood mononuclear cells (PBMCs) [6, 7]. Cyto-
inflammatory state and are believed to play a significant role in kines are a family of pleiotropic polypeptides with a
dialysis-related morbidity. The interleukin hypothesis suggests
molecular weight ranging from 10 to 45 kD that, pro-
that the release of proinflammatory cytokines acts as an under-
duced by different cells in response to inflammatory stim-
lying pathophysiologic event in hemodialysis-related acute
uli, may modulate a variety of functions not only in
manifestations, such as fever and hypotension. Nevertheless,
a cytokine overproduction may alter sleep pattern in chronic circulating immune cells, but also in mesenchymal, endo-
hemodialyzed patients, thus explaining the presence of sleep thelial, and epithelial cells [8]. There is an increasing
disorders in these patients. A potential role of cytokines in
body of evidence that the interaction between blood
chronic-related morbidity has also been suggested. High levels
and dialytic membranes induces the release of several
of some inflammatory cytokines are often associated with ane-
mia caused by hyporesponsiveness to erythropoietin. Cytokine cytokines from circulating mononuclear cells, such as
production may also play a relevant role in bone remodeling by interleukin-1 (IL-1), IL-6, IL-8, tumor necrosis factor-␣
regulating osteoblast/osteoclast cell functions and parathyroid
(TNF-␣), and monocyte chemotactic factor-1 (MCP-1).
hormone (PTH). Finally, cytokine release may have a long-
The specific action of any of these monocyte-derived
term deleterious effect on mortality of uremic patients by alter-
cytokines may be relevant in the pathogenesis of clinical
ing immune response and increasing susceptibility to infections.
Bioincompatibility of dialytic membranes may also contribute manifestations often observed in end-stage renal disease
to malnutrition in dialysis patients by increasing the monocyte (ESRD) patients undergoing chronic hemodialysis [9, 10].
release of catabolic cytokines such as TNF-␣ and IL-6. Bioin-
compatible dialytic treatment may induce an inappropriate
monocyte activation and cytokine production, which, in turn, MECHANISMS INVOLVED IN CYTOKINE
may mediate some of the immune and metabolic dysfunction
PRODUCTION DURING HEMODIALYSIS
associated with hemodialysis. The use of biocompatible dialytic
membranes appears to reduce the monocyte activation and to In vitro and in vivo data support the hypothesis that
improve the survival of hemodialysis patients. cytokine transcription and/or production during hemodi-
alysis are mainly caused by (1) direct contact of PBMCs
with dialysis membrane, (2) active complement frag-
During hemodialysis, blood contact with a foreign sur- ments (C3a, C5a, C5b-9) generated during hemodialysis,
face, such as a complement-activating dialytic mem- and (3) backtransport of bacterial-derived material [for
brane, promotes a variety of complex and interrelated example, lipopolysaccharide (LPS)] from the dialysate
events, leading to an acute inflammatory response. In to the blood compartment.
particular, activation of mononuclear cells and concomi- The dialyzer used seems to play an important role,
tant complement activation induce the release of an both directly and directly, in cytokine induction during
array of inflammatory mediators into the extracellular hemodialysis through complement activation. It is well
environment, including cytokines, reactive oxygen spe-
known that in the absence of any other stimulus, adher-
cies (ROS), and nitric oxide (NO) [1–3]. Thus, hemodial-
ence to dialysis membrane induces selective mRNA ex-
ysis imbalances several homeostasis systems and gener-
pression of monocyte mediators and proto-oncogenes.
Indeed, Betz et al have demonstrated that cuprophan
Key words: biocompatibility, cytokines, hemodialysis complications. membranes stimulate IL-1 expression in monocytes in
the absence of complement [11]. On the other hand,
 2000 by the International Society of Nephrology
S-104

2. Pertosa et al: Hemodialysis and cytokine production S-105
cellulosic membranes can activate, through the alterna- soluble TNF receptors, has further complicated this
cloudy issue [21]. The quality of the antibodies and the
tive pathway, the complement cascade and can generate
active fragments able to stimulate cytokine gene expres- enzyme-linked immunosorbent assay (ELISA) or radio-
immunoassay kits (RIA) used were not always highly
sion and secretion by monocytes [12, 13]. Furthermore,
several studies support the hypothesis that terminal com- reliable. Studies using bioassays (considered useful in
the past), in light of the discovery of cytokine-specific
plement complex (TCC) generation may have a potential
role in activating mononuclear cells [14]. Sublethal TCC inhibitors, are now questionable. The introduction of
molecular biological techniques to the study of cytokines
doses can affect cell metabolism and constitute a potent
signal for activation of monocytes to produce inflamma- production has partly overcome some of these difficulties
[22]. Unlike bioassays, these techniques are highly spe-
tory mediators such as TNF-␣ and IL-6 [13]. The third
possible pathogenic factor to consider in the increased cific and are not influenced by cytokine-binding proteins
and inhibitors. Using appropriate probes, these tech-
cytokine production during hemodialysis is the LPS frag-
ments contaminating the dialysate and able to cross the niques allow the researcher to detect specific cytokine
mRNAs and their cellular source. By using these tech-
membrane [15]. We have demonstrated that the basal
release of TNF-␣ and IL-6 during hemodialysis is inde- niques, we recently evaluated the role of different hemo-
dialysis membranes in regulating spontaneous IL-6, IL-8,
pendent of the biocompatibility features of the mem-
brane used, while it is considerably influenced by the and MCP-1 gene expression and protein synthesis by
PBMCs isolated from chronic hemodialyzed patients.
endotoxin content of the dialysate [16]. Indeed, mono-
cytes, isolated from uremic patients treated with a dialy- Our data demonstrated the independent modulation of
cytokines, gene expression, and protein secretion in un-
sate characterized by high endotoxin levels, spontane-
ously released a significantly greater amount of TNF-␣ stimulated PBMCs by different hemodialysis mem-
branes. Indeed, the transcriptional activation of these
and IL-6 compared with healthy controls and nondia-
lyzed uremic patients. In contrast, the use of a dialysate cytokines occurred in unstimulated PBMCs following
with low endotoxin concentration significantly reduced the use of cellulosic membranes, whereas their protein
the cytokine production. synthesis was seemingly impaired [23]. By contrast, we
The contact with the dialysis membrane as well as the observed that long-term hemodialysis with synthetic
interactions with complement fractions, although able high-flux membranes, such as polymethylmethacrylate
to induce a selective cytokine gene transcription in (PMMA) and polyamide, down-regulated cytokine gene
monocytes, does not always automatically stimulate the expression and improved the ability of PBMCs to secrete
translation of the specific proteins. We have previously cytokines in culture. Thus, molecular biology techniques
shown that IL-6 gene expression is strikingly increased can help to identify the mechanisms leading to cytokine
during hemodialysis with cuprophan membrane, but its transcription during hemodialysis. Overall, the applica-
protein secretion is clearly down-regulated [17]. Interest- tion of these techniques may shed light on the study
ingly, Schindler et al have demonstrated that recombi- of bioincompatibility of different dialyzers and dialysis
nant C5a in vitro stimulates transcription rather than procedures.
translation of IL-1 and that LPS or IL-1 itself is required
as a translational signal [12]. In support of this hypothe-
CLINICAL RELEVANCE OF CYTOKINE
sis, we and others have demonstrated that patients
PRODUCTION IN HEMODIALYSIS:
treated with cuprophan have higher mRNA levels for
ACUTE AND CHRONIC EFFECTS
different cytokines, but in order to obtain a higher trans-
More than 15 years ago, Henderson et al first proposed
lation, a second stimulus, such as LPS, is required [12, 16].
the “interleukin hypothesis,” incriminating IL-1 produced
In contrast, some authors have reported an impaired
during dialysis as the cause of different acute responses
endotoxin-induced IL-1 and IL-6 release from PBMCs
observed in patients on hemodialysis [24]. The better
of hemodialysis patients, suggesting a reduced ability of
understanding of the biological effects of proinflamma-
the immune cells of these patients to respond to patho-
tory cytokines gained over the last decade strongly sup-
logic stimuli [18, 19]. Finally, other authors did not see
ports the hypothesis that these soluble mediators may
evidence of any production or release of different cyto-
be involved in the pathogenesis of both acute and chronic
kines (TNF-␣, IL-6, and mRNA coding for IL-1␤) by
complications of dialysis treatment, including fever, hy-
monocytes during high-flux bicarbonate hemodialysis,
potension, sleep disorders, dialysis-related amyloidosis,
neither with complement-activating membranes nor with
impaired immunity, bone disease, malnutrition, and ane-
contaminated dialysate [20]. Because of the great vari-
mia (Table 1).
ability of the experimental design and techniques used,
Herein, we describe the main pathophysiologic mecha-
these conflicting results can easily generate confusion.
nisms underlying dialysis-adverse clinical events poten-
Moreover, in the last few years, the discovery of cyto-
kine-specific inhibitory proteins, such as IL-1RA and tially related to an altered cytokine production.

3. Pertosa et al: Hemodialysis and cytokine production
S-106
Table 1. Clinical relevance of cytokine production of sleep disturbance or daytime sleepiness in 45 of 54
in hemodialysis patients
hemodialysis patients (83.3%) evaluated, and causes
Acute Chronic were secondary to delayed sleep onset and frequent
Fever Anemia awakening in 46.3 and 35.2% of patients, respectively
Sleep disorders Bone disease [29]. Moreover, symptoms of mild or severe restless legs
Hypotension Malnutrition
were reported by 57.4% of patients. The data reviewed
Immunological dysfunction
in this study support the hypothesis that immune-active
molecules, such as cytokines, may induce profound alter-
ations in several neurotransmitters in the brain. The fre-
quent sleeping during hemodialysis and the abnormal
FEVER
sleep patterns of patients are rather speculative corre-
Fever and chills commonly occur during or after dial-
lates of increased cytokine production. It has been dem-
ysis in the presence of endotoxin-contaminated dialysate.
onstrated that IL-1␤ and TNF-␣ are involved in physio-
This specific endotoxin-induced effect may be potenti-
logic sleep regulation [30] and may induce slow-wave
ated by the use of a bioincompatible complement-acti-
sleep. There is a daily rhythm of TNF-␣ and IL-1␤
vating membrane. However, a slight increase in body
mRNA expression in the central nervous system, with
temperature can also be observed in patients undergoing
the highest levels occurring during peak sleep periods.
hemodialysis with an uncontaminated dialysate.
Moreover, IL-1␤ and TNF-␣ are part of a larger bio-
Fever is a coordinate endocrine, autonomic, and be-
chemical cascade involved in sleep regulation; other
havioral response organized by the brain as a reaction
somnogenic substances in this cascade include growth
to an inflammatory stimulus. The classic model of fever
hormone-releasing hormone and NO. Thus, we may hy-
pathogenesis suggests that IL-1, IL-6, or TNF-␣ pro-
pothesize that an alteration in cytokine production (for
duced by PBMCs in the bloodstream may be recognized
example, overproduction) directly or via an increased
as pyrogenic signal by specific centers within the central
synthesis of other somnogenic substances, such as NO,
nervous system [25]. At this level, they can induce the
may explain the altered sleep pattern observed in chronic
synthesis of prostaglandins that represent the central
hemodialyzed patients, particularly in those treated with
mediators of the coordinated response leading to fever.
bioincompatible membranes [31].
Luheshi demonstrated that local administration of LPS
into a subcutaneous rat air pouch elicits marked fever,
accompanied by an increase in TNF-␣, IL-1, and IL-6 HYPOTENSION
levels in the pouch, but only IL-6 in the plasma [26].
The incidence of a symptomatic reduction in blood
The author suggests that TNF-␣ and IL-1 probably act
pressure during hemodialysis ranges from 15 to 50% of
locally to stimulate the release of one or more secondary
dialysis sessions [32].
circulating mediator(s) (for example, IL-6) that can in-
Hypotension and cardiovascular instability are the most
teract with the brain. By contrast, Kozak et al demon-
frequent side effects of dialysis, occurring both chroni-
strated that the injection of high dose of LPS can induce
cally in long-term hemodialysis patients and acutely dur-
fever in mice with a null mutation of IL-6 gene, but
ing dialytic sessions, and have been related to induction
not in mice with the disruption of IL-1␤ gene [27]. The
of IL-1 and TNF-␣ synthesis in monocytes [24]. More-
authors concluded that IL-1␤ may have a critical role
over, it has been proposed that hypotensive responses
in inducing fever during systemic inflammation. Indeed,
to hemodialysis may result from cytokine-induced syn-
several lines of evidence suggest that IL-1 may act di-
thesis of NO, a potent vasodilator in vitro and in vivo,
rectly on specific regulatory sites within the hypothala-
within vascular smooth muscle cells and endothelial cells
mus, turning up the set point of the physiologic thermo-
[33]. Particularly, it has been demonstrated that acetate
stat and causing an increase in body temperature via the
can directly activate human monocytes to produce IL-1
normal thermoregulatory pathways. This hypothesis is
[34], and Amore et al found that acetate-containing dial-
strongly supported by Hammond et al’s recent observa-
ysate up-regulated inducible NO synthase (iNOS) gene
tion that type I IL-1 receptor is expressed in human
expression and NO production in endothelial cells as
hypothalamus [28].
compared with acetate-free buffers [35]. In contrast,
Noris et al demonstrated that plasma from patients dia-
SLEEP DISORDERS lyzed with acetate-free biofiltration (AFB), a technique
using a buffer-free dialysate, postdilution with a sterile
Sleep disorders have been reported to be high among
bicarbonate solution, and a polyacrylonitrile dialyzer,
uremic patients on hemodialysis treatment, likely con-
did not stimulate endothelial NO synthesis as compared
tributing to the impaired quality of life experienced by
with plasma from patients treated with acetate dialysate
many of these patients. By using a sleep questionnaire,
Walker, Fine, and Kryger demonstrated the presence [36]. Moreover, plasma IL-1␤ was greater after acetate

4. Pertosa et al: Hemodialysis and cytokine production S-107
dialysis than after AFB, corresponding to a more pro-
nounced intradialytic decrease in systolic blood pressure
in patients treated with acetate buffer than in those
treated with AFB. These data are consistent with the
possibility that cytokines, released in excessive amounts
during acetate dialysis, may contribute to hemodynamic
instability, via an increased production of NO by mono-
cytes and endothelial cells. Moreover, the interaction
between bioincompatible complement-activating mem-
branes and PBMCs can induce iNOS expression and
enzyme activity, most likely mediated by an increase
production of IL-1␤ and TNF-␣, contributing to the de-
velopment of hypotension in dialyzed patients [3].
ANEMIA
Cytokines are well-known regulatory factors for eryth-
ropoiesis, especially in pathological conditions. Chronic
inflammatory diseases, characterized by high cytokine cir-
culating levels, are often associated with anemia caused
Fig. 1. Intracellular cross-talk between cytokine and erythropoietin
by a hyporesponsiveness to erythropoietin [37]. A sig- signaling pathways.
nificant percentage of uremic patients present erythro-
poietin resistance even in the absence of comorbid condi-
tions such as iron depletion, hyperparathyroidism, or
(Fig. 1). Tyrosine-phosphorylated STATs dimerize and
aluminum overload [38]. The presence of a deranged
migrate to the nucleus, where they can induce the expres-
cytokine production in dialysis may suggest a role for
sion of a variety of proinflammatory and cell-activating
these soluble mediators in the reduced response to eryth-
genes (Fig. 1) [41]. On the other hand, these transcription
ropoietin in this patient population. Goicoechea et al
factors can cause the up-regulation of a growing family of
demonstrated in uremic patients undergoing chronic he-
genes encoding proteins known as suppressor of cytokine
modialysis a significant and direct correlation between
signaling (SOCS), which can interfere with JAK–STAT
the erythropoietin dose and the IL-6 and TNF-␣ produc-
interaction [42]. SOCS may thus inhibit the cytokine-
tion from stimulated and unstimulated cultured PBMCs
activating signal, turning on a negative feedback loop
[39]. Allen et al, using normal and uremic bone marrow
(Fig. 1). Noteworthy, SOCS activated from one cytokine
to test erythropoietin response in the presence of uremic
can inhibit the signal induced by a second cytokine acting
serum, recently reported a more direct and causal rela-
on the same cell [43]. Interestingly, erythropoietin induces
tionship between cytokine production and erythropoie-
erythroid cell proliferation interacting with a specific cell
tin resistance in hemodialysis [40]. They did not observe
surface receptor belonging to the cytokine receptor su-
any difference in erythropoietin response between nor-
perfamily and inducing the JAK-STAT pathway [44].
mal and uremic bone marrow. However, when bone mar-
Therefore, it is conceivable that SOCS expressed in re-
row was cultured with uremic serum, the erythropoietin
sponse to IL-1, TNF-␣, or IL-6 in erythroid cells can
effect on erythroid colony formation was clearly inhib-
switch off erythropoietin signaling and inhibit its prolif-
ited. The addition of specific anti–TNF-␣ and interferon-␥ erative effect on this cell line (Fig. 1).
(IFN-␥) antibodies to this system almost completely re-
stored erythropoietin response.
BONE DISEASE
The increasing knowledge on intracellular events in-
duced by cytokines suggests that mechanisms underlying Patients with end-stage renal disease present with vari-
their ability to inhibit the erythropoietin effect are cur- ous debilitating forms of osteodystrophy characterized
rently present. Several of these inflammatory mediators either by high or low bone turnover. Alterations in para-
utilize Janus kinases (JAK) and the signal transducers thyroid hormone (PTH) and calcitriol production do not
and activators of transcription (STAT) proteins to modu- completely account for the observed abnormalities in
late gene expression in their target cells [41]. The differ- bone resorption/formation. Recent reports stress the role
ent JAK isoforms are activated upon receptor dimeri- of cytokines and their inhibitors in the process of bone
remodeling directly or modulating the expression and/or
zation and phosphorylate one or more STAT proteins

5. Pertosa et al: Hemodialysis and cytokine production
S-108
the effects of PTH. It is well established that marrow cells and the utilization of exogenously administered nutrients
[58]. IL-1 is well known to act directly on the hypothala-
can modulate bone remodeling through local cytokine
release. Moreover, there is an increasing body of evi- mus, causing anorexia [58]. The role of TNF-␣ in neopla-
sia-induced cachexia is well established, and there is now
dence suggesting several cytokines as autocrine factors
regulating osteoblast/osteoclast cell functions. IL-6 is evidence of the involvement of TNF-␣, formerly known
as cachectin, in uremia-associated malnutrition. Indeed,
physiologically produced by osteoblasts in response to
PTH and, interestingly, may induce osteoclastogenesis its plasma concentration correlates with biochemical
signs of protein catabolism in hemodialysis patients
and bone resorption [45, 46]. In mice, both acute neutral-
ization and chronic deficiency of IL-6 are associated with [57, 58]. Finally, Kaizu et al reported that hemodialyzed
patients with a high plasma IL-6 concentration presented
markedly lower levels of biochemical markers of bone
resorption in response to PTH infusion when compared a lower albumin levels and a significantly higher weight
loss over a three-year period than patients with low
with animals with normal IL-6 levels [47]. In addition to
IL-6, IL-1 and TNF-␣ may induce directly bone resorption plasma IL-6 [59]. Moreover, the circulating IL-6 concen-
tration was inversely correlated with serum albumin,
by stimulating the development of osteclast-like multinu-
cleated cells and by increasing the bone-resorbing activ- cholinesterase, and midarm muscle area. Furthermore,
a direct correlation between cell content of IL-1 receptor
ity of formed osteoclast [48]. Moreover, these two cyto-
kines may modulate the actions of calciotropic hormones antagonist and some nutritional parameters, such as
body mass index, anthropometry-derived arm muscle
on osteoblast by inhibiting intracellular calcium release
and inositol trisphosphate production in a tyrosine ki- area, serum cholesterol, and triglycerides, was found in
16 patients dialyzed with reprocessed cellulose dialyzers
nase-dependent manner [49].
Beside the direct effect on bone cells, cytokine may [60]. These findings suggest a direct correlation between
nutrition and cytokine production and that malnutrition
modulate PTH production. Parathyroid cells express
IL-8 type B receptor and respond to IL-8 incubation could depress cytokine production and potentially con-
tribute to reduced immune responsiveness in patients on
with a marked increase in PTH expression [50]. On the
other hand, IL-1 can induce an up-regulation of extracel- chronic hemodialysis.
lular calcium-sensing receptor mRNA while inhibiting
PTH secretion in cultured parathyroid tissue slices [51].
IMMUNOLOGIC DYSFUNCTION
In addition to the two classic variants of renal osteodys-
There is an increasing body of evidence that uremic
trophy, amyloid bone disease is the third form of hemodi-
patients on dialysis present an increased susceptibility to
alysis-related bone pathology. The incidence of amyloid
infections. Particularly the use of cellulosic membranes is
bone disease is much greater in patients dialyzed with
associated with dysfunction of phagocytic cells, natural
cellulosic membranes than with biocompatible filters
killer cells, and other immunologic alterations, including
[52]. Cellulosic membrane may indeed induce an in-
altered cytokine production and complement system ac-
creased synthesis of ␤2-microglobulin via complement
tivation [61]. Indeed, bacterial infections are the most
system activation and cytokine release. Particularly IL-1,
common cause for hospitalization and the second most
TNF-␣, and IL-6 have been shown to stimulate ␤2-micro-
common cause for death [62]. The presence of a pro-
globulin release by leukocyte and endothelial cells [53].
found defect in lymphocyte and monocyte function in
uremia is further suggested by the observation of an
MALNUTRITION extended survival of skin allografts, a marked decreased
cutaneous responsiveness to a broad panel of antigens
Malnutrition is often present in hemodialyzed patients,
and several reports have demonstrated the adverse effect and, finally, a reduced seroconversion after vaccination
[63]. A growing interest has been recently focused on
of malnutrition on their morbidity and mortality [54].
Patients undergoing chronic hemodialysis show evidence the relationships between cytokine plasma levels and/or
production by immune cells and the dysfunction of
of accelerated protein catabolism primarily because of
a significant loss of amino acid induced by the dialysis phagocytes, natural killers and T lymphocytes often ob-
served in hemodialysis patients [21]. Although the physi-
procedure [55]. Experimental data suggest that the dial-
ysis procedure per se leads to enhanced catabolism, as ologic role of circulating cytokines is unknown, it is con-
ceivable that a proinflammatory milieu characterized by
well as a direct loss of plasma amino acids and proteins
into the dialysate [56]. In addition, several studies suggest high cytokine plasma levels may influence specific and
tightly regulated cellular processes in leukocytes, induc-
a role for cellulosic membranes in enhancing active ca-
tabolism, most likely through an increased release of ing an inappropriate cellular and/or humoral immune
response. Since individual cytokines have multiple syner-
proinflammatory cytokines [57]. Indeed, cytokines such
as IL-1, TNF-␣, and particularly IL-6 appear to play a gistic and antagonistic actions on different cellular tar-
gets, cytokine combinations may exert immunologic ef-
central role in both the loss of skeletal muscle proteins

6. Pertosa et al: Hemodialysis and cytokine production S-109
fects not foreseeable based on the actions of the single altered PBMC IL-12 production may determine an im-
cytokine. Moreover, the presence of high circulating lev- munodeficiency state in hemodialyzed patients. Indeed,
els of soluble receptors or binding proteins, reported in the authors hypothesize that an altered IL-12 release
patients on hemodialysis, may further complicate the may contribute to a depressed cell-mediated immune
scenario [21]. However, it should be considered that cyto- response in these patients, by inducing a shift toward a
kines exert their major physiologic and pathophysiologic Th2 cell response [65]. Thus, the use of poor biocompati-
effects as autocrine or paracrine factors. Therefore, the ble membranes, by recurrently activating mononuclear
altered intracellular processing and local release of a cells or altering Th1/Th2 balance, may contribute to
cytokine may be more relevant in the pathogenesis of the down-regulation of the synthesis and release of different
uremic immune dysfunction than its increased circulating immunoregulatory cytokines and may play a role in cell-
levels. Insufficient or delayed cytokine release may de- mediated immunodeficiency of dialyzed uremic patients.
crease the immune reaction and, consequently, increases
the risk of infection. We and others have recently re-
CONCLUSIONS
ported that contact of circulating mononuclear cells with
In summary, bioincompatible dialytic treatment may
complement-activating membranes results in the in-
induce an inappropriate monocyte activation and cyto-
creased gene expression for an array of different cyto-
kine production, which, in turn, mediate some of the
kines, including IL-6, MCP-1, and IL-8, without a corre-
immune and metabolic dysfunction associated with he-
sponding increased translation into protein [23]. As a
modialysis. Activation of immunocompetent cells during
consequence, mononuclear cells of uremic patients on
dialysis are chronically activated, although they cannot the hemodialysis session results in acute and long-term
completely demonstrate their activated phenotype. There- adverse effects. Clinical alterations resulting from cyto-
fore, uremic monocytes may easily become exhausted kine production and release include fever, cardiovascular
and subsequently refractory to any further stimulation. instability, sleep disorders, and increased muscle protein
This hypothesis is further supported by the observation catabolism. Altered cytokine release also contributes to
of an impaired endotoxin-induced IL-1, TNF-␣, and IL-6 the immunodeficiency of dialysis patients and increases
release from PBMCs of hemodialysis patients. By exam- the morbidity and mortality of these patients [67]. Recent
ining the Th1 and Th2 cytokine profiles in 22 stable studies support the hypothesis that hemodialysis with
hemodialyzed patients and 22 healthy controls, Daichou synthetic and less complement-activating membranes
et al demonstrated that the T-cell activity was signifi- may normalize monocyte function [23] and improve im-
cantly retarded in uremic patients as compared with nor- munologic parameters by reducing the circulating levels
mal controls [64]. Furthermore, they showed that the of proinflammatory cytokines, thus contributing to ame-
production of IL-2, which is involved in cell-mediated lioration of the survival of hemodialysis patients [68].
immune responses, and IL-4 and IL-10, which affect hu-
moral immunity, were significantly lower in patients than ACKNOWLEDGMENTS
in controls. Finally, an increased production of IL-12 by
We thank Terumo (Japan) and Bieffe (Italy) for technical and
macrophages and IFN-␥ by Th1 cells, as well as elevated financial support.
plasma levels of soluble IL-2 receptor, was observed in
Reprint requests to Giovanni Pertosa, M.D., Division of Nephrology,
uremic patients as compared with controls. They con-
Department of Emergency and Transplantation, University of Bari,
cluded that the balance of Th1- and Th2-type responses Policlinico, Bari, Italy.
is altered in hemodialyzed patients, contributing to the E-mail: g.pertosa@nephro.uniba.it
immune dysfunction in these patients. Other authors
also demonstrated an alteration of Th1 and Th2 cells REFERENCES
network. Memoli et al demonstrated that cultured
1. Memoli B: Cytokine production in hemodialysis. Blood Purif
PBMCs harvested from uremic patients dialyzed with a 17:149–158, 1999
cuprophan membrane spontaneously release a greater 2. Cristol JP, Canaud B, Rabesandratana H, Gaillard I, Serre
A, Mion C: Enhancement of reactive oxygen species production
amount of IL-12 than PBMCs from normal controls and
and cell surface markers expression due to hemodialysis. Nephrol
nondialyzed uremic patients, whereas under mitogen Dial Transplant 9:389–394, 1994
stimulation, IL-12 supernatant levels were significantly 3. Amore A, Bonaudo R, Ghigo D, Arese M, Costamagna C, Cirina
P, Gianoglio B, Perugini L, Coppo R: Enhanced production of
lower than controls [65]. In contrast, the use of a less com-
nitric oxide by blood-dialysis membrane interaction. J Am Soc
plement-activating membrane, such as PMMA mem- Nephrol 6:1278–1283, 1995
brane, normalized IL-12 release from PBMCs. IL-12 pro- 4. Gesualdo L, Pertosa G, Grandaliano G, Schena FP: Cytokines
and bioincompatibility. Nephrol Dial Transplant 13:1622–1626,
motes the differentiation of cytokine-producing Th1
1998
cells, likely by inducing the IFN-␥ gene expression in
5. Hakim RM: Clinical implications of hemodialysis membrane bio-
both T cells and natural killer cells and IFN-␥ plays an compatibility. Kidney Int 44:484–494, 1993
6. Tetta C, Camussi G, Turello E, Salomone M, Aimo G, Priolo
important role in the resistance to infection [66]. An

7. Pertosa et al: Hemodialysis and cytokine production
S-110
G, Segoloni G, Vercellone A: Production of cytokines in hemodi- LR, Zheng H: IL-6 and IL-1 beta in fever: Studies using cytokine-
deficient (knockout) mice. Ann NY Acad Sci 856:33–47, 1998
alysis. Blood Purif 8:337–346, 1990
28. Hammond EA, Smart D, Toulmond S, Suman-Chauhan N,
7. Shaldon S, Lonnemann G, Koch KM: Cytokine relevance in
Hughes J, Hall MD: The interleukin-1 type I receptor is expressed
biocompatibility. Contrib Nephrol 79:227–236, 1987
in human hypothalamus. Brain 122:1697–1707, 1999
8. Nathan CF: Secretory products of macrophages. J Clin Invest
29. Walker S, Fine A, Kryger MH: Sleep complaints are common
79:319–326, 1987
in a dialysis unit. Am J Kidney Dis 26:751–756, 1995
9. Dinarello CA: Cytokines: Agents provocateurs in hemodialysis?
30. Krueger JM, Fang J, Taishi P, Chen Z, Kushikata T, Gardi J:
Kidney Int 41:683–694, 1992
Sleep: A physiologic role for IL-1 beta and TNF alpha. Ann NY
10. Vilcek J, Le J: Immunology of cytokines: An introduction, in The
Acad Sci 856:148–159, 1998
Cytokine Handbook, edited by Thomson A, San Diego, Academic
31. Takahashi S, Kapas L, Fang J, Krueger JM: Somnogenic relation-
Press, 1991, pp 1–17
ships between tumor necrosis factor and interleukin-1. Am J Phys-
11. Betz M, Haensch GM, Rauterberg EW, Bommer J, Ritz E:
iol 276:R1132–R1140, 1999
Cuprammonium membranes stimulate interleukin-1 release and
32. Orofino L, Marcen R, Quereda C, Villafruela JJ, Sabater J,
arachidonic acid metabolism monocytes in the absence of comple-
Matesanz R, Pascual J, Ortuno J: Epidemiology of symptomatic
ment. Kidney Int 34:67–73, 1988
hypotension in hemodialysis: Is cool dialysate beneficial for all
12. Schindler R, Lonnemann G, Shaldon S, Koch KM, Dinarello
patients? Am J Nephrol 10:177–180, 1990
CA: Transcription, not synthesis, of interleukin 1 and tumor necro-
33. Yokokawa K, Kohno M, Yoshkawa J: Nitric oxide mediates the
sis factor by complement. Kidney Int 37:85–93, 1990
cardiovascular instability of hemodialysis patients. Curr Opin
13. Pertosa G, Tarantino EA, Gesualdo L, Montinaro V: C5b-9
Nephrol Hypertens 5:359–363, 1996
generation and cytokine production in hemodialyzed patients. Kid-
34. Bingel M, Koch KM, Lonnemann G, Dinarello CA, Shaldon
ney Int 43(Suppl 41):S221–S225, 1993
S: Enhancement of in-vitro human interleukin-1 production by
14. Hansch GM, Seitz M, Betz M: Effect of late complement compo-
sodium acetate. Lancet 1:14–16, 1987
nents C5b-9 on human monocytes: Release of prostanoids, oxygen
35. Amore A, Cirina P, Mitola S, Peruzzi L, Bonaudo R, Gianoglio
radicals and a factor inducing cell proliferation. Int Arch Allergy
B, Coppo R: Acetate intolerance is mediated by enhanced synthesis
Appl Immunol 82:317–320, 1987
of nitric oxide (NO) by endothelial cells. J Am Soc Nephrol 8:1431–
15. Lonnemann G, Behme TC, Lenzner B, Floege J, Schulze M,
1436, 1997
Colton CK, Koch KM, Shaldon S: Permeability of dialyzer mem-
36. Noris M, Todeschini M, Casiraghi F, Roccatello D, Martino
branes to TNF␣-inducing substances derived from water bacteria.
G, Minetti L, Imberti B, Gaspari F, Atti M, Remuzzi G: Effect
Kidney Int 42:61–68, 1992
of acetate, bicarbonate dialysis and acetate-free biofiltration on
16. Pertosa G, Gesualdo L, Bottalico D, Schena FP: Endotoxins
nitric oxide synthesis: Implications for dialysis hypotension. Am J
modulate chronically tumour necrosis factor alpha and interleukin
Kidney Dis 32:115–124, 1998
6 release by uraemic monocytes. Nephrol Dial Transplant 10:328–
37. Jongen-Lavrencic M, Peeters HR, Wognum A, Vreugdenhil
333, 1995
G, Breedveld FC, Sweak AJ: Elevated levels of inflammatory
17. Pertosa G, Gesualdo L, Tarantino EA, Ranieri E, Bottalico
cytokines in bone marrow of patients with rheumatoid arthritis
D, Schena FP: Influence of hemodialysis on interleukin-6 produc-
and anemia of chronic disease. J Rheumatol 24:1504–1509, 1997
tion and gene expression by peripheral blood mononuclear cells.
38. Tarng DC, Huang TP, Chen TW, Yang WC: Erythropoietin
Kidney Int 43(Suppl 39):S149–S153, 1993 hyporesponsiveness: From iron deficiency to iron overload. Kidney
18. Lonnemann G, Barndt I, Kaever V, Haubitz M, Schindler R, Int 55(Suppl 69):S107–S109, 1999
Shaldon S, Koch KM: Impaired endotoxin-induced interleukin-1␤ 39. Goicoechea M, Martin J, de Sequera P, Quiroga JA, Ortiz A,
secretion, not total production, of mononuclear cells from ESRD Carreno V, Caramelo C: Role of cytokines in the response to
patients. Kidney Int 47:1158–1167, 1995 erythropoietin in hemodialysis patients. Kidney Int 54:1337–1343,
19. Paczek L, Schaefer RM, Heidland A: Dialysis membranes de- 1998
crease immunoglobulin and interleukin-6 production by peripheral 40. Allen DA, Breen C, Yaqoob MM, Macdougall IC: Inhibition
blood mononuclear cells in vitro. Nephrol Dial Transplant 3:41–44, of CFU-E colony formation in uremic patients with inflammatory
1991 disease: Role of IFN-gamma and TNF-alpha. J Invest Med 47:204–
20. Grooteman MPC, Nubé MJ, Daha MR, Van Limbeek J, Van 211, 1999
Deuren M, Schoorl M, Bet PM, Van Houte AJ: Cytokine profiles 41. Ihle JN: STATs: Signal transducers and activators of transcription.
during clinical high-flux dialysis: No evidence for cytokine genera- Cell 84:331–334, 1996
tion by circulating monocytes. J Am Soc Nephrol 8:1745–1754, 42. Kovanen PE, Leonard WJ: Inhibitors keep cytokines in check.
1997 Curr Biol 9:R899–R902, 1999
21. Descamps-Latscha B, Herbelin A, Nguyen AT, Roux-Lombard 43. Dickensheets HL, Venkataraman C, Schindler U, Donnelly
P, Zingraff J, Moynot A, Verger C, Dahmane D, de Groote D, RP: Interferons inhibit activation of STAT6 by interleukin 4 in
Jungers P: Balance between IL-1 beta, TNF-alpha and their spe- human monocytes by inducing SOCS-1 gene expression. Proc Natl
cific inhibitors in chronic renal failure and maintenance dialysis: Acad Sci USA 96:10800–10805, 1999
Relationship with activation markers of T cells, B cells and mono- 44. Kirito K, Uchida M, Yamada M, Miura Y, Komatsu N: A distinct
cytes. J Immunol 154:882–892, 1995 function of STAT proteins in erythropoietin signal transduction.
22. Schindler R, Koch KM: The application of molecular biology J Biol Chem 272:16507–16513, 1997
techniques in the study of biocompatibility in hemodialysis. 45. Monier-Faugere MC, Malluche HH: Role of cytokines in renal
Nephrosciences 5:25–35, 1995 osteodystrophy. Curr Opin Nephrol Hypertens 6:327–332, 1997
23. Pertosa G, Grandaliano G, Gesualdo L, Ranieri E, Monno R, 46. Ferreira A, Simon P, Di-Beke TB, Descamps-Latscha B: Poten-
Schena FP: Interleukin-6, interleukin-8 and monocyte chemotactic tial role of cytokines in renal osteodystrophy. Nephrol Dial Trans-
peptide-1 gene expression and protein synthesis are independently plant 11:399–400, 1996
modulated by hemodialysis membranes. Kidney Int 54:570–579, 47. Grey A, Mitnick MA, Masiukiewicz U, Sun BH, Rudikoff S,
1998 Jilka RL, Manolagas SC, Insogna K: A role for interleukin-6 in
24. Henderson LW, Koch KM, Dinarello CA, Shaldon S: Hemodi- parathyroid hormone-induced bone resorption in vivo. Endocrinol-
alysis hypotension: The interleukin-1 hypothesis. Blood Purif ogy 140:4683–4690, 1999
1:3–8, 1983 48. Tani-Ishii N, Tsunoda A, Teranaka T, Umemoto T: Autocrine
25. Dinarello CA: Cytokines as endogenous pyrogens. J Infect Dis regulation of osteoclast formation and bone resorption by IL-1,
179(Suppl 2):S294–S304, 1999 TNF-␣. J Dent Res 78:1617–1623, 1999
26. Luheshi GN: Cytokines and fever: Mechanisms and sites of action. 49. Tam VK, Schtland S, Green J: Inflammatory cytokines (IL-1
Ann NY Acad Sci 856:83–89, 1998 alpha, TNF-alpha) and LPS modulate the Ca2⫹ signaling pathway
in osteoblasts. Am J Physiol 274:C1686–C1698, 1998
27. Kozak W, Kluger MJ, Soszynski D, Conn CA, Rudolph K, Leon

8. Pertosa et al: Hemodialysis and cytokine production S-111
50. Angeletti RH, D’Amico T, Ashok S, Russell J: The chemokine tional support and the cachexia syndrome: Interactions and thera-
peutic options. Cancer 79:1828–1839, 1997
interleukin 8 regulates parathyroid secretion. J Bone Miner Res
59. Kaizu Y, Kimura M, Yoneyama T, Miyagi K, Hibi I, Kumagai
13:1232–1237, 1998
H: Interleukin-6 may mediate malnutrition in chronic hemodialysis
51. Nielsen PK, Rasmussen AK, Butters R, Feldt-Rasmussen U,
patients. Am J Kidney Dis 31:93–100, 1998
Bendtzen K, Diaz R, Brown EM, Olgaard K: Inhibition of PTH
60. King AJ, Kehayias JJ, Roubenoff R, Schmid CH, Pereira BJ:
secretion by interleukin-1 beta in bovine parathyroid glands in
Cytokine production and nutritional status in hemodialysis pa-
vitro is associated with an up-regulation of the calcium-sensing
tients. Int J Artif Organs 21:4–11, 1998
receptor mRNA. Biochem Biophys Res Commun 238:880–885,
61. Vanholder R, Ringoir S: Infectious morbidity and defects of
1997
phagocytic function in end-stage renal disease: A review. J Am
52. Ferreira A, Urena P, Ang KS, Simon P, Morieux C, de Vemejoul
Soc Nephrol 3:1541–1554, 1993
MC, Drueke T: Relationship between serum ␤2-microglobulin,
62. Bloembergen WE, Port FK: Epidemiological perspective on infec-
bone histology and dialysis membranes in uraemic patients.
tions in chronic dialysis. Adv Ren Replace Ther 3:201–207, 1996
Nephrol Dial Transplant 10:1701–1707, 1995
63. Cohen G, Haag-Weber M, Horl WH: Immune dysfunction in
53. Zaoui P, Stone WJ, Hakim RM: Effects of dialysis membrane on
uremia. Kidney Int 52(Suppl 62):S79–S82, 1997
␤2m production and cellular expression. Kidney Int 38:962–968,
64. Daichou Y, Kurashige S, Hashimoto S, Suzuki S: Characteristic
1990
cytokine products of Th1 and Th2 cells in hemodialysis patients.
54. Bergstrom J: Effect of in vivo contact between blood and dialysis
Nephron 83:237–245, 1999
membranes on protein catabolism in humans. Kidney Int 38:487–
65. Memoli B, Marzano L, Bisesti V, Andreucci M, Guida B: Hemo-
494, 1990
dialysis-related lymphomonuclear release of interleukin-12 in pa-
55. Guarnieri G, Toigo G, Fiotti N, Ciocchi B, Situlin R, Giansante tients with end-stage renal disease. J Am Soc Nephrol 10:2171–
C, Vasile A, Carraro M, Faccini L, Biolo G: Mechanisms of 2176, 1999
malnutrition in uremia. Kidney Int 52(Suppl 62):S41–S44, 1997 66. Boehm U, Klamp T, Groot M, Howard JC: Cellular responses
56. Guiterrez A, Alvestrand A, Wahren J, Bergstrom J: Effect of to interferon-␥. Ann Rev Immunol 15:749–795, 1997
in vivo contact between blood and dialysis membranes on protein 67. Hakim RM, Held FJ, Stannard DC, Wolfe RA, Port FK, Daug-
catabolism in humans. Kidney Int 38:487–494, 1990 irdas JT, Agodoa L: Effects of dialysis membrane on mortality
57. Flores EA, Bistrian BR, Pomposelli JJ, Dinarello CA, Black- of chronic hemodialysis patients. Kidney Int 50:566–570, 1996
burn GL, Istfan NW: Infusion of tumor necrosis factor: Cachectin 68. Kimmel PL, Phillips TM, Simmens SJ, Peterson RA, Weihs KL,
promotes muscle catabolism in the rat. J Clin Invest 83:1614–1622, Alleyne S, Cruz I, Yanovski JA, Veis JH: Immunological func-
1989 tion and survival in hemodialysis patients. Kidney Int 54:236–244,
1998
58. Moldawer LL, Copeland EM: Proinflammatory cytokines, nutri-