Paweł S. Berezowicz, Barbara Adamik, Andrzej Kübler
Elimination of interleukin-6 and interleukin-8 during continuous veno-venous haemofiltration in intensive care patients
Chair of Anaesthesiology and Intensive Therapy,
Head: prof. A. Kübler. Medical University, Wrocław, Poland
We have assessed the serum and ultrafiltrate concentrations of interleukiens 6 and 8 in eighteen multiorgan failure patients, aged 18-72 yr, treated with continuous veno-venous haemofiltration and evaluated clinically with APACHE II, SAPS and SOFA scores. Assays were performed (ELISA, Quantikinine, R&D Systems, UL) at the forth hour after beginning of haemofiltration and subsequently every 24 hr. Mean serum and ultrafiltrate concentrations of IL-6 were 289.0 pg/ml and 20.5 pg/ml, respectively. Serum concentration of IL-8 was 708.2 pg/ml and it was lower than in the ultrafiltrare (751.9 pg/ml, 17 patients) The ultrafiltrare/serum concentration ratio varied from 1.094 to 10.87. There was no correlation between clinical scores and interleukin concentration with SOFA and IL-8 being only exception (R=0.39; p<0.05). The IL-8 serum concentration was inversely related to the leucocyte's count (R=-0.3321; p,0.005). In our study we could not eliminate significantly interleukines with the veno-venous haemofiltration.
Acute renal failure develops frequently as a part of multi-organ failure syndrome. Its pathogenesis is multifactoral, including vasodilatation and hypotension (leading to decrease of renal blood flow), endotoxins and nephrotoxic drugs . If no functional artificial substitution of the renal function is employed, acute renal failure invariably leads to death. Even in dialysed patients the mortality rate is high and exceeds 50% . A method of renal function substitution, matched to the ITU needs, is haemofiltration.
Recent years brought much information about the role of mediators of inflammatory reaction in the pathogenesis of sepsis and septic shock . Although there is no general agreement on the mechanisms leading to this potentially fatal condition, the role of mediators seems indisputable. Generalised infection leading to overproduction of inflammatory reaction mediators is the cause of endothelial lesions and microcirculation disorders, resulting in tissue hypoxia and irreversible damage. This kind of generalised inflammatory reaction leads to multiple organ failure and, not infrequently, to death .
In the eighties the greatest interest was attracted by cytokines: tumour necrosis factor alpha (TNF-α), interleukin 1 (IL-1) and interleukin 6 (IL-6) . Peripheral blood concentrations of these glycopeptides (well known to immunologists) correlated with the severity of the disease, morbidity and mortality. This was the base for investigating the potential benefit of elimination of these substances in the treatment of sepsis and septic shock. Unfortunately, none of the double-blind, multicentre clinical trials showed a beneficial effect of neutralisation/elimination of circulating cytokines [5, 6]. It was reported, however, that during haemofiltration, cytokines are eliminated from the blood and transferred to the filtration fluid. The present paper describes the effect of continuous veno-venous haemofiltration on the elimination of IL-6 and IL-8 in patients treated in the intensive care unit.
Eighteen patients, aged from 14 to 72 years (mean: 51) were treated by continuous veno-venous haemofiltration in the Department of Intensive Therapy of the Medical Academy of Wrocław during 1995-1996. Acute renal failure was a complication of diffuse peritonitis, septic shock, multiple organ failure, cardiac surgery (with the use of extracorporeal circulation) and extensive vascular surgical procedures. Patients under 14 years of age and those receiving immune- suppressive treatment, were excluded from the study. Another criterion of exclusion was duration of haemofiltration not exceeding 48 hours. Basic demographic data, diagnosis and bacteriological tests results are presented in table I.
Table I. Demographic and clinical data
|No.||Gender*||Age||Diagnosis||APACHE II on admission||Estimated risk of death (on admission)||SAPS II at start of haemofiltration||Outcome**||Bacteriological results***|
|1||M||49||Craniocerebral trauma, meningitis||34||0.7171||74||NS||O: Corynebacterium sp., Streptococcus agalactiae; PMR: cocci G+|
|2||M||44||Acute pancreatitis||27||0.6023||78||NS||O: Enterobacter cloacae, Enterococcus faecelis B:
S. haemolyticus, S. epidermidis|
|3||M||61||Colic tumour||27||0.6857||89||NS||DB: S. epidermidis;
O: Ps. aeruginosa|
|4||K||72||Colic tumour, ilaeus, peritonitis||25||0.6196||73||NS||B: E. coli, Kl. oxytoca, Morganella morgani, Streptococcus bovis, Bacteroides distasonis|
|5||M||42||Myocardial infarction, CA, pneumonia, sepsis||26||0.6620||54||NS||O: MRSA, Acinetobacter baumani|
|6||K||72||Retroperitoneal abscess||24||0.6106||42||NS||O: Proteus mirabilis, Ps. aeruginosa, Candida albicans;
U: Proteus mirabilis|
|7||M||40||Acute pancreatitis, CA||41||0.9494||88||NS||O: Ps. aeruginosa, Enterococcus faecalis, Micrococcus sp.; B: Candida albicans, Enterococcus faecalis|
|8||M||14||Abdominal trauma, DIC, acute renal failure||33||0.7724||64||NS||O: Candida albicans|
|9||M||47||Acute renal failure following cardiac surgery||17||0.0915||44||S||O: Ps. aeruginosa, Neisseria sp., Streptococcus sp.|
|10||M||42||Myocardial infarction, cardiogenic shock||17||0.2152||56||S||-|
|11||M||41||Haemorrhagic shock following renal transplantation||40||0.9041||74||NS||-|
|12||M||47||Acute renal failure following cardiac surgery||21||0.2571||47||NS||B: Acinetobacter baumani, Enterococcus faecalis, Staphylococcus lentus; K: Staphylococcus epidermidis|
|13||K||59||Acute renal failure following cardiac surgery||35||0.7915||45||NS||B: Enterococcus
faecalis, Ps. aeruginosa|
|14||M||65||Acute renal failure following abdominal aortic aneurysm surgery||32||0.4420||46||NS||B: Ps. aeruginosa, E. coli, Acinetobacter baumani|
|15||M||60||Acute renal failure following cardiac surgery, sepsis||13||0.0532||60||NS||
|16||M||55||Chronic renal failure, hypovolaemic shock,splenectomy ||34||0.9271||66||NS||K: E. coli|
|17||M||50||Myocardial infarction||24||0.4492|| ||NS||K: MRSA; O: E. coli, Morganelle morganii; U: E. coli, Wound smear: E. coli, Morganella morgani|
|18||M||58||Faecal peritonitis||22||0.6482||66||NS||B: E. coli, S. aureus, Enterococcus faecalis; Ps. aeruginosa; O: Ps. aeruginosa, Enterococcus faecalis, S. aureus, Acinetobacter baumani|
* M - M (male); K - F (female); ** NS - non survived; S - survived; *** K - blood; U - urine; O - bronchial secretions; B - peritoneal fluid; DB - abdominal drain(s); PMR - cerebrospinal fluid; MRSA - meticillin resistant Staph. aureus
All patients received treatment, according to their clinical condition. This consisted of mechanical ventilation, antibiotic treatment, according to bacteriological culture results, infusion of inotropic and vasopressor drugs, blood products and i.v. fluids, enteral or parenteral nutrition. The APACHE II score and estimated risk of death were calculated on admission . During haemofiltration treatment, the clinical condition of the patients was assessed daily according to the SAPS II scale , and retrospectively - to the SOFA scale .
Continuous veno-venous haemofiltration (CVVH) was carried out with an ADM-08 device and polysulphonate filters AV400 or AV600 (Fresenius, Bad Homburg, Germany). The venous access was assured by a double lumen catheter, introduced into the femoral, subclavian or internal jugular vein. The dose of heparin infused to the line was titrated on the basis of PTT values. As necessary, the circulating blood volume of the patient was supplemented with isotonic substitution fluid HF-01 (Fresenius, Bad Homburg, Germany). During 24h CVVH, 20 to 25 litres of ultrafiltrate were obtained.
Blood samples were drawn 4 hours after the beginning of the procedure and every 24h thereafter, to heparinised vials. Blood plasma, obtained by 15 minutes centrifuging at 3000 rpm. (temperature 4°C) was kept in temperature -70°C until laboratory tests were performed.
The concentrations of IL-6 and IL-8 in plasma and ultrafiltrate were assessed by ELISA immunoenzymatic tests, available commercially (Quantikine, R&D Systems Europe, Abingdon, Great Britain). These tests are specific for detection of natural and recombinant human interleukins 6 and 8 and do not react with other cytokines (TNF-a, IL-1, 2, 3, 4 or 7). The test sensitivity for IL-6 is 0.7 pg/ml and for IL-8 - 3 pg/ml in ultrafiltrate and 18 pg/ml in plasma. The variability between tests did not exceed 8% (IL-6) and 10% (IL-8).
In all plasma samples, both IL-6 and IL-8 were detected. Their concentrations correlated with each other (R=0.3866; p<0.01). Concentrations of IL-6 in plasma ranged from 8.0 to 936.4 pg/ml (mean 289.0; median: 171.3 pg/ml). The mean plasma concentration of IL-6 was higher in patients with clinical and bacteriological symptoms of sepsis, although the difference was not significant. IL-6 concentration in the ultrafiltrate ranged from 1.4 to 184.6 pg/ml (mean 20.5; median: 10.2 pg/ml) (Fig.1). Mean IL-6 concentrations in plasma and ultrafiltrate during CVVH are presented in figure 2. Differences between values are not statistically significant.
Plasma IL-8 concentration ranged from 9.6 to 14504.3 pg/ml (mean 708.2; median: 214.2 pg/ml) and respective values in the ultrafiltrate: 19.3; 6000; 751.9 and 201.2 pg/ml (Fig. 3). In 17 samples from 10 patients IL-8 concentration in the ultrafiltrate was higher than in plasma by a factor of 1.094 to 10.886. Mean values of IL-8 concentrations in plasma and ultrafiltrate during CVVH are presented in figure 4. The differences between values were not significant.
On admission to the ITU, the APACHE II score ranged from 13 to 41 (mean 27.64), and the estimated death risk from 0.0532 to 0.9494 (mean 0.5777). The SAPS II score averaged 63.97 ± 18.82 (range: 39-124), and the SOFA score 12.37 ± 2.97 (range: 7-18). Statistical significance was observed only between the SOFA score and plasma IL-8 concentration (R=0.3939; p<0.05). An inverse correlation was noted between plasma concentration of this cytokine and WBC count (R=-0.3321; p<0.05).
Multiple organ failure is a frequent and life-threatening complication of severe infections and trauma. Haemodynamic consequences of Gram-negative sepsis are the result of the action of endotoxins: lipopolysaccharide elements of the bacterial wall. Endotoxins do not act, however, directly on the host organism. Their mechanism of action consists of stimulating the release of inflammatory mediators from the white blood cells. An important role in this group is played by cytokines, expressing their proinflammatory activity. One of the most important factors in the pathogenesis of sepsis (but not the only one) may be the Tumour Necrosis Factor. The role of interleukins 1, 6 and 8 is also postulated . Many authors have reported the activity of different cytokines in experimental animals and in patients with sepsis of different origin, as well as following trauma and major surgery. A correlation between cytokine concentration and mortality has also been reported. Attempts to establish a prognostic value for different cytokines in sepsis have not yet given univocal results [11, 12, 13].
Fig. 1. IL-6 plasma and ultrafiltrate concentration
Fig. 2. Changes in IL-6 plasma and ultrafiltrate concentrations during CVVH
Fig. 3. IL-8 plasma and ultrafiltrate concentration
Fig. 4. Changes in IL-8 plasma and ultrafiltrate concentrations during CVVH
On the basis of the known pathogenic chain of sepsis (endotoxins -> TNF + other cytokines -> vasodilatation -> shock) several attempts were undertaken to neutralise one of these stages of septic shock by using monoclonal antibodies against selected cytokines. Clinical trials with anti-endotoxins, TNF receptor and IL-1 receptor antagonist did not meet the expectations [5, 6]. Moreover, the use of recombinant protein: TNF receptor (characterised by a very high affinity to TNF) increased the mortality rate in patients receiving it . It seems, though, that the cytokine system constitutes an integral part of the immune reaction, and attempts to eliminate one of these substances may be more harmful than beneficial .
Acute renal failure frequently complicates sepsis and septic shock. If no renal function substitution is introduced, it invariably leads to death. Evidence exists, that in intensive care patients haemofiltration is more useful than haemodialysis in eliminating of nitrogen metabolites and maintaining the correct volume of circulating blood. It does not destabilise circulatory homeostasis and does not provoke abrupt changes of osmolality and electrolyte concentration. It is also efficient in eliminating the so - called middle molecules (including cytokines) from the blood.
The first reports on the beneficial role of haemofiltration in ARDS and sepsis (both experimental and clinical) appeared in the eighties. In 1984, Gotloib et al. reported on encouraging results of continuous arterio-venous haemofiltration (CAVH) in the treatment of patients with ARDS . A few years later, Barzilay et al. compared different methods of treatment of renal failure in septic patients. In patients treated with CAVH, the mortality rate was lower, the most effective combination being haemofiltration and plasmapheresis. This could suggest a role of elimination of endotoxins and middle molecules in the treatment of sepsis . Papers assessing the influence of haemofiltration on respiratory and haemodynamic parameters were also published [17, 18]. The authors stated that this kind of treatment improved respiratory and circulatory function, as reflected by FiO2, PEEP, PAP, PaO2/FiO2 and VO2. Some authors postulate, that only "large ultrafiltrate volume" haemofiltration (about 100 litres per 24h) allows sufficient elimination of inflammatory mediators [19,20]. In view of the greater risk of complications and the high cost, the majority of the centres use lower flow rates: about 25 litres/24h.
One of the therapeutic mechanisms of haemofiltration is its efficacy in excess fluid elimination. Available data suggest that elimination of mediators by haemofiltration increase the survival rate in septic patients, so the concept of target-oriented treatment of septic states evokes hope and interest.
As opposed to haemodialysis, haemofiltration can eliminate from the blood water-soluble substances of molecular weight below 10 kDa, regardless of the size of the molecule. Theoretically, the elimination of water-soluble components depends on convection transport, leading to equalisation of concentration of a given substance on both sides of the filtration membrane. In reality, plasma proteins undergo adhesion to the membrane, which results in its polarisation and provokes a difference in ionic concentration on both sides of the membrane. Molecules of middle size are more readily eliminated by haemofiltration, than by haemodialysis. It is admitted that molecules of less than 50 kDa may be eliminated during haemofiltration, although this can depend on the physical characteristics of the filter used. In our patients polysulphonate filters were employed, permeable (according to the manufacturer) to particles up to 30 kDa molecular weight.
The molecular weight of IL-6 ranges from 23 to 26 kDa, and that of IL-8 from 8 to 9 kDa. These are, however, only approximate values, as the size of a cytokine molecule depends on its chemical characteristics. The polypeptide chains of the molecule can potentially accept 0- and N- glycosylation radicals and side oligosaccharide chains can also modify particle characteristics. Depending on the substitution of the side-chains, the existence of several forms of IL-6 is postulated (molecular weight ranging from 19 to 70 kDa) . Different forms may prevail in different clinical conditions, the biggest molecules being refractory to elimination by haemofiltration. Moreover, cytokines in the blood may exist as dimers or trimers; it is well - known that IL-8 exists as a homodimer .
The interpretation of the results presented in this paper suggests that no clinically important elimination of interleukins by haemofiltration could be achieved. The proportion of IL-6 concentration in the ultrafiltrate and in plasma did not exceed (in the majority of samples) 0.2; and in only 5 samples ranged from 0.2 to 0.5. The maximum elimination rate of IL-6 was estimated at about 4.5 mg/24h.
A somewhat different pattern was observed with regard to IL-8 elimination. In the majority of samples (28/45) the concentration ratio of plasma to ultrafiltrate did not exceed 0.5 (mean: 0.46). In 17 samples, however, it ranged from 1.09 to 10.89. A similar observation on the concentration of TNF has been reported by Tonnesen et al. . Some explanations of this fact are possible. First, blood cytokines may be bound to specific antibodies and/or soluble receptor proteins. Passage through the filter may liberate an active/detectable form of the cytokine. Second, a polysulphonate filtration membrane may, by itself, activate WBC's and local liberation of cytokines. Up to now, the precise influence of different types of membranes employed in medical devices on blood cells and proteins have not been recognised. Testing of dialysis and haemofiltration filters biocompatibility usually concentrate on their influence on activation of the coagulation system, complement cascade, and neutrophils as well as monocyte stimulation. The phenomenon of complement activation and neutropaenia following haemodialysis was described earlier. The influence of filtering membranes on immune system cells was not investigated until the last decade . In vitro studies demonstrate that polysulphonate filters minimally activate the complement system, elastase release and do not cause neutro- or thrombocytopaenia . Nevertheless, the problem of interaction between immune system cells and different types of filtration membranes needs further investigation.
The results obtained are consistent with those reported by other authors, stating that proinflammatory mediators can be eliminated by continuous veno-venous haemofiltration. The clinical importance of this finding, especially in terms of its significance in the treatment of sepsis, septic shock and multiple organ failure needs further multicentre clinical trials.
1.Cytokines are eliminated from the blood during continuous veno - venous haemofiltration.
2.The rate of elimination is unstable and relatively low.
3.Haemofiltration does not cause significant lowering of plasma cytokine activity.
4.The higher concentration of Il-8 in ultrafiltrate than in plasma may be the evidence of neutrophile activation by the filtrating system.
Originally published in Anestezjologia Intensywna Terapia 31; (3), 155-159, 1999.
1. Marino P. L.: Intensywna Terapia. Urban&Partner, Wrocław 1994.
2. Bone R. C.: The pathogenesis of sepsis. Annals of International Medicine 1991, 115, 457-69.
3. Goris R. J., te Boekhorst T. P., Nuytinck J. K., Gimbrere J. S.: Multipleorgan failure. Generalized autodestructive inflammation? Archives of Surgery 1985, 120, 1109-1115.
4. Blackwell T. S., Christman J. W.L.: Sepsis and current status. British Journal of Anaesthesia 1996, 77, 110-117.
5. Opal S. M., Fisher C. J. Jr.,Dhainaut J. F., Vincent J. L., Lowry S. F., Sadoff J. C., Slotman G. J.,Levy H., Balk R. A., Shelly M. P., Pribble J. P., LaBreque J. F., Lookabaugh J., Donovan H., Dubin H., Baughman R., Norman J., DeMaria E., Matzel K., Abraham E., Seneff M.: Confirmatory interleukin-1 receptor antagonist trial in serve sepsis: a phase III, randomized, double-blind, plcebocontrolled, multicenter trial. The Interleukin-1 Receptor antagonist Sepsis Investigator Group. Critical Care Medicine 25, 1115-1124.
6. Fisher C. J. Jr., Agosti J. M., Opal S. M., Lowry S. F., Balk R. A., Sadoff J. C., Abraham E., Schein R. M. H., Benjamin E.: Treatment of septic shock with the tumor necrosis factor receptor: Fc fusion protein. New England Journal of Medicine 1996, 334, 1697-1702.
7. Knaus W. A., Droper E. A., Wagner D. P., Zimmerman J. E.: APACHE II: a severity of disease classification system. Critical Care Medicine 1975, 13, 818-829.
8. Le Gall J. R., Lemeshow S., Saulnier F.: A new Slmplified Acute Physiology Score (SAPS II) based on European/ North American multicenter study. Journal of the American Medical Association 1993, 270, 2957-2963.
9. Vincent J. L., Moreno R., Takla J., Willatts S., De Mendonca A., Bruining H., Reinhart C. K., Suter P. M., Thijs L. G: The SOFA (Sepsisrelated Organ Failure Assessment) score to describe organ dysfunction/failure. On behalf of the Working Group on Sepsis-Related Problems of the European Society of Intensive Care Medicine.Intensive Care Medicine 1996, 22, 707-710.
10. Darville T., Giroir B., Jacobs R.: The systemic inflammatory response syndrome (SIRS): Immunology and potential immunotherapy. Infection 1993, 21, 279-290.
11. Casey L. C., Balk R. A., Bone R. C.: Plasma cytokine and endotoxin levels correlate with survival in patients with the sepsis syndrome. Annals of Internal Medicine 1993, 119, 771-778.
12. Bamas P., Ledoux D., Nys M., Vrindts Y., De Groote D., Franchimiont P., Lamy M.: Cytokine serum level during serve sepsis in human. II-6 as a marker of severity. Annals of Surgery 1992, 215, 356-362.
13. martin C., Saux P., Mege J.-L., Perrin G., Papazin L., Gouin F.: Prognostic values of serum cytokines in septic shock. Intensive Care Medicine 1994, 20, 272-277.
14. Bone R. C. A.: A critical evaluation of new agents for the treatment of sepsis. Journal of the American Medical Association 1991, 266, 1686-1691.
15. Gotloib L., Barzilay E., Shuctak A., Lev A.: Sequential hemofiltration in nonoliguric high capillary permeability pulmonary edema of serve sepsis. Critical Care Medicine 1984, 997-1000.
16. Barzilary E., Kessler D., Berlot G., Gullo A., Gerber D., Ben Zeev I.: Use of extracoporeal supportive techniques as additional treatment for septic-induced multiple organ failure patients. Critical Care Medicine 1989, 17, 636-637.
17. Garzia F., Todor R., Scalea T.: Continuous arteriovenous heamofiltration countercurrent dialysis (CAVH-D) in acute respiratory failure (ARDS). Journal of Trauma 1991, 31, 1277-1284.
18. Bashaw O. N., Anaes F. R., Hutchinson A.: Continuous arteriovenous haemofiltration and respiratory function in multiple organ systems failure. Intensive Care medicine 1992, 18, 334-338.
19. Grootendorst A. F., Bauman C. S. C., Hoeben K. H. N., van Saase JLCM, van Leengoed L. A. M. G.: The role of continous renal replacement therapy in sepsis and multiorgan failure. American Journal of Kidney Diseases 1996, 28, Suppl. 3, S50-S57.
20. Rogiers P., Zhang H., Vincent J.-L.: Hemofiltration is sepsis and septic shock; in: Yearbook of Intensive Care and Emergency Medicine (ed. Vincent J.-L.), Springer, Berlin, 1997, 133-139.
21.Sehgal P. b.: Interleukin-6: a regular of plasma protein gene expression in hepatic and non-hepatic tissues. Molecular Biology and Medicine 1990, 7, 117-130.
22. Clore G. M., Apella E., Yamada M., Matsushima K., Groneborn A. M: Determination of the secondary structure of interleukin-8 by nuclear magnetic resonance spectroscopy. Journal of Biological Chemistry 1989, 264, 18907-18911.
23. Tonnesen E., Hansen M. B., Hohndorf K., Diamant M., Bendtzen K., Wanscher M., Tofi P.: Cytokines in plasma and ultrafiltrate during continous arteriovenous haemofiltration. Anaesthesiology and Intensive Care 1993, 21, 752-758.
24. Hakim R. M.: Clinical implications of hemodialysis membrane biocompability. Kidney International 1993, 44, 484-494.
25. Schafer M. R., Heidland A., Horl H. W.: Relase of leukocyte elastase during hemodialysis. Contributions of Nephrology 1985, 45, 109-117.