Comparison of cellulose diacetate and polysulfone membranes in the ...

Nephrol Dial Transplant (2000) 15: 224–230

Nephrology Dialysis Transplantation

Original Article

Comparison of cellulose diacetate and polysulfone membranes in the outcome of acute renal failure. A prospective randomized study Karine Gastaldello1, Christian Melot2, Robert-Jean Kahn2, Jean-Louis Vanherweghem1, Jean-Louis Vincent2 and Christian Tielemans1 1Department of Nephrology, Dialysis and Transplantation and 2Department of Intensive Care, Hoˆpital Erasme, Universite´ Libre de Bruxelles, Belgium

Abstract Background. Whether the nature of haemodialysis (HD) membranes can influence the outcome of acute renal failure (ARF ) remains debatable. Recent studies have suggested that dialysis with bioincompatible unsubstituted cellulosic membranes is associated with a less favourable patient outcome than dialysis with biocompatible synthetic membranes. Since we generally use a modified cellulosic membrane with substantially lower complement- and leukocyte-activating potential than cuprophane, for dialysis of patients with ARF, and because there are no data in the literature regarding the influence of modified cellulosic membranes on the outcome of patients with ARF, we compared the outcome of ARF patients dialysed either with cellulose diacetate or with a synthetic polysulfone membrane. We also investigated the potential role of permeability by comparing membranes with high-flux versus lowflux characteristics. Methods. This prospective, randomized, single centre study included 159 patients with ARF requiring HD. Patients were stratified according to age, gender, and APACHE II score and then randomized in chronological order to one of three dialysis membranes: lowflux polysulfone, high-flux polysulfone and meltspun cellulose diacetate. Results. Aetiologies of ARF and the prevalence of oliguria were similarly distributed among the three groups. There was no significant difference between the three groups for survival (multivariate Cox’s proportional hazards model, P=0.57), time necessary to recover renal function (P=0.82), and number of dialysis sessions required before recovery (P=0.86). Multivariate analysis showed that survival was significantly influenced only by the severity of the disease state (APACHE III score, P2.5 mg/dl ). No patients were excluded because of the aetiology of ARF, biochemical values, or the presence of concurrent conditions. All decisions regarding initiation, discontinuation of dialysis due to recovery of renal function, and technical conditions of the dialytic treatment were made by the attending clinical nephrologists who were usually not involved in this study. After a decision had been made to initiate HD, the eligible patients were first classified according to three covariables: age (under or over 50 years old), gender, and severity of disease assessed by the APACHE II score at the time of initiation of dialysis ( less or more than 20) [7]. The patients were then randomized by minimization to one of the three types of dialysis membranes which were then used for all subsequent dialysis sessions. The clinical staff were not blinded to the membrane assignment. No continuous renal support technique was used during the study. A total of 162 patients were included, but three patients were later excluded because of an error in the membrane used. All 159 patients were followed until death, discharge from the hospital and off dialysis, or 80 days after initiation of dialysis, whichever came first. In this latter category, if the patient was still requiring dialysis after 80 days of followup, it was considered that the patient would not recover renal function. Three types of single-use dialysis membranes were used with different characteristics of biocompatibility and permeability ( Table 1): high-flux polysulfone (HF PS) (Hemoflow F60S, Fresenius, Germany), low-flux polysulfone (LF PS) (Hemoflow F6S, Fresenius, Germany) and cellulose diacetate which is manufactured using the meltspinning technique (CDA, Altra Flux 140, Althin, USA). PS is a

225

synthetic membrane which does not induce complement activation and produces only a slight drop in white blood cell count. These membranes manufactured from synthetic polymers are considered as biocompatible because of their minimal inflammatory-type contact reactions with blood. Meltspun CDA membrane is a synthetically modified cellulose membrane with the same permeability characteristics as PS low-flux but a biocompatibility which is intermediate between the synthetic membranes and the regenerated unmodified cellulose [8]. In fact, CDA membranes induce a moderate complement activation, as well as a substantial drop in white blood cell count but to a lesser degree than cuprophane. All dialytic treatments were performed using monitors equipped with volumetric-control of ultrafiltration. In an attempt to deliver the same dose of dialysis to all patients, we used dialysers having similar hollow fibre configurations, clearance characteristics, and surface area. In addition, the dialysis schedule was adapted to each patient to maintain predialytic urea levels below 200 mg/dl (mainly 3 h daily). Blood flow rate was maintained between 200–250 ml/min. Dialysis water was obtained from reverse osmosis. Bicarbonate-based dialysate was used in all cases, and dialysate flow rate was 500 ml/min. Dialysate concentrations of potassium and calcium, ultrafiltration rate, and anticoagulant dose were adapted at each session by the nephrologist according to patient needs. All therapeutic decisions regarding mechanical ventilation, parenteral nutrition, inotropic support, and other specific treatments were made by the attending ICU physician.

Measurements of endotoxin levels Endotoxin levels in the dialysate were measured routinely using the chromogenic Limulus Amoebocyte specific assay LAL-CoatestA endotoxin (Chromogenix, Mo¨lndal, Sweden). Results are expressed as EU/ml. The sensitivity of the assay is 0.015 EU/ml in the dialysate.

Scoring of severity of illness Severity of illness was determined at initiation of dialysis treatment using both the APACHE II and APACHE III scores [7,9,10]. These scores represent the sum of values evaluating current physiologic condition, age, and previous chronic conditions. Since the APACHE II score is a major determinant of hospital mortality in ARF patients [11,12]: (provided it is recorded on dialysis initiation and not on ICU admission [7,11]) we used APACHE II scores for the stratification of patients. For probability analysis of survival (see statistical section), we preferred to use the APACHE III score in the multivariate regression analysis. We also recorded the existence of comorbid conditions (presence of hypertension, congestive heart failure, ischaemic cardiomyopathy, cerebral vascular disease, peripheral vascular disease, liver failure, respiratory insufficiency, neoplasm, immunocompromised status, diabetes), the presence of oliguria (as defined by urine output 38°C or 90 beats/min, white blood cell count >12 000/mm3 or 20/min.

226

K. Gastaldello et al.

Table 1. Characteristics of the haemodialysis membranes

Membrane material Creatinine clearance ml/mina Vitamin B12 clearance ml/mina Kuf ml/h/mmHg Sterilization technique Surface m2

CDA

LF PS

HF PS

meltspun cellulose diacetate 161 88 12.8 gamma rays 1.4

low-flux polysulfone

high-flux polysulfone

167 80 8.5 steam 1.3

172 118 40 steam 1.3

aData provided by the manufacturer (Blood and dialysate flow rate respectively at 200 and 500 ml/min; no net ultrafiltration). Kuf: ultrafiltration coefficient.

Outcome measurements Differences in the survival rate between the three groups were examined by multivariate analysis (Cox proportional hazards method ). Outcome measures included survival, cause of death, recovery of renal function (defined as discontinuation of dialysis because it was no longer required as measured by a spontaneous fall in serum creatinine without dialytic intervention), number of HD sessions up to 80 days. We also recorded in each patient confounding factors that may influence the outcome of ARF such as: administration of radiocontrast material, nephrotoxic drugs given after the initiation of dialysis, and the number of HD sessions with hypotension. Hypotension was defined as a decrease in mean arterial pressure of at least 20 mmHg requiring the discontinuation of ultrafiltration, the administration of fluid or the infusion, or increase in dosage, of vasopressive agents.

Statistical analysis Differences between groups were evaluated using the x2 test for discrete variables and analysis of variance (ANOVA) for continuous data. Probability of survival was analysed using a multivariate Cox’s proportional hazards model stratified by the type of dialysis membrane and adjusted for the following covariables: patient characteristics (age, gender), severity of illness (APACHE III score), oliguria, aetiology of ARF (hypotension, sepsis, toxic origin, rhabdomyolysis, chronic renal failure, other causes), comorbidities (hypertension, coronary artery disease, congestive heart failure, peripheral vascular disease, cerebral vascular disease, hepatic insufficiency, respiratory insufficiency, neoplasm, immunocompromised status, diabetes), organ support (inotropic agents, mechanical ventilation), and type of admission (non postoperative patient, elective postoperative patient, or emergency postoperative patient). P values PMMA+PS (26%) CU (58%)=PMMA (60%) CU (45%)>PS HF (18%)

Schiffl et al. 1994 Cosentino et al. 1994a Kurtal et al. 1995 Mehta et al. 1996b

4 27 5 24

Assouad et al. 1996b Liano et al. 1996a Neveu et al. 1996a

25 26 18

Simpson et al. 1993b Splendiani et al. 1996a Gastaldello et al. 1996b Himmelfarb et al. 1998

23 30 31 6

Jo¨rres et al. 1998b Schiffl et al. 1998

32 33

NS NS NS NS NS ND NS 0.03 0.003 NS 0.001

anot randomized for membrane evaluation; bAbstracts; ND: not done; NS: not significant; ATN: acute tubular necrosis; CU: cuprophane; CDA: meltspun cellulose diacetate; CA: cellulose acetate; HE: hemophane; PS: polysulfone; PA: polyamide; PMMA: polymethylmethacrylate; PAN: polyacrilonitrile.

between cuprophane and low-flux PMMA in the whole group of patients, but a better survival and recovery of renal function in a subgroup of non-oliguric patients. It must be emphasized that this subgroup was constructed a posteriori and that the distribution of the patients’ characteristics between the membranes within this subgroup was not reported. The second study by Schiffl et al. [4] found a marginally higher death rate, especially from sepsis, in patients dialysed on cuprophane as compared to AN69. However, PAN membranes differ from bioincompatible membranes not only in terms of biocompatibility but also in terms of permeability and adsorptive properties, so that the improved outcome may be due to a combination of the aforementioned factors. This study also presented several limitations which raised criticisms [19,21,22]. The third study by Kurtal et al. [5] found no difference between cuprophane and polyamide but causes of ARF were somewhat heterogeneously distributed in the two groups. Recently, a prospective multi-centre study by Himmelfarb et al. [6 ], which included patients from the Hakim study [3], shed some light on the role of the HD membrane in the outcome of patients with ARF requiring HD. This study, comparing bioincompatible unsubstituted cellulosic membranes with biocompatible synthetic membranes, demonstrated that biocompatibility is important in determining patient survival and recovery of renal function, particularly in patients who were non-oliguric at the time of initiation of dialysis. It should be noted that modified cellulosic membranes can exhibit more biocompatible patterns than unsubstituted cellulose so that conclusions drawn from the previously cited studies may not be applicable to all cellulosic membranes. Indeed, evidence suggests that modifications of cellulosic membranes reducing their complement activating potential may result in a patient survival rate comparable to synthetic biocom-

patible membranes. Hakim et al. [28] compared the relative risk of mortality of end-stage renal disease patients dialysed with synthetic, modified cellulosic or unsubstituted cellulosic membranes, and demonstrated that patient survival was greater with either synthetic or modified cellulosic membranes than with unsubstituted cellulosic membranes. Neveu et al. [18] reported similar results, i.e. a better survival rate in patients dialysed with synthetic or modified cellulosic membranes than in those dialysed with unsubstituted bioincompatible cellulosic membranes, although their study was not randomized for membrane allocation. Analysis of the other studies is impossible because substituted cellulosic membranes are sometimes regarded as the bioincompatible membrane [25] or, on the contrary, mixed with synthetic membranes [18,24]. Manufacturing conditions can also influence the performance and the biocompatibility of the membranes [29]. Hollow-fibre can be manufactured either by solution spinning or by solution meltspinning. Solution spinning processes involve synthetic polymers, solvents, pore formers, wetting agents and coagulation steps leading to toxic residuals. During the meltspinning process, heat is used to melt the polymer blend, nitrogen core gas is used to keep the lumen open during extrusion, and air cooling is employed to accomplish thermal phase inversion; the hollow fibres are further processed in hot water and glycerin baths into a semi-permeable membrane. In this study, we used meltspun CDA membranes, and the results may have been different with a cellulosic membrane prepared through another process. In conclusion, many studies have shown that the biocompatibility of the dialysis membrane plays an important role in the outcome of patients with ARF, but, no study has thus far investigated the possible role of modified cellulosic membrane in this issue. The

230

present controlled, randomized study demonstrated that the modified cellulosic membrane, CDA (which has less complement-activating potential ), is just as effective as synthetic PS membranes in the HD of patients with ARF, and at a lower cost. We also investigated the possible role of membrane permeability on the outcome of patients with ARF, since a previous study by Schiffl et al. [4] suggested that the difference in survival rate between patients dialysed with biocompatible or bioincompatible membranes may result from additive benefits of the permeability and the biocompatibility. We found there was no significant difference between high-flux and low-flux PS in patient outcome. Acknowledgements. The study was partially supported by the Fondation Erasme.

K. Gastaldello et al.

15. 16. 17. 18.

19. 20. 21.

References 1. Chew SL, Lins RL, Daelemans R, De Broe ME. Outcome in acute renal failure. Nephrol Dial Transplant 1993; 8: 101–107 2. Pascual M, Swinford RD, Tolkoff-Rubin N. Acute renal failure: role of dialysis membrane biocompatibility. Annu Rev Med 1997; 48: 467–476 3. Hakim RM, Wingard RL, Parker RA. Effect of the dialysis membrane in the treatment of patients with acute renal failure. N Engl J Med 1994; 331: 1338–1342 4. Schiffl H, Lang SM, Ko¨nig A, Strasser T, Haider MC, Held E. Biocompatible membranes in acute renal failure: prospective case-controlled study. Lancet 1994; 344: 570–572 5. Kurtal H, von Herrath D, Schaefer K. Is the choice of membrane important for patients with acute renal failure requiring hemodialysis? Artif Organs 1995; 19: 391–394 6. Himmelfarb J, Tolkoff NR, Chandran P, Parker RA, Wingard RL, Hakim R. A multicenter comparison of dialysis membranes in the treatment of acute renal failure requiring dialysis. J Am Soc Nephrol 1998; 9: 257–266 7. Damiano AM, Bergner M, Draper EA, Knaus WA, Wagner DP. Reliability of measure of severity of illness: acute physiology of chronic health evaluation-II. J Clin Epidemiol 1992; 45: 93–101 8. Klinkmann H, Vienken J. Membranes for dialysis. Nephrol Dial Transplant 1995; 10 [Suppl. 3]: 39–45 9. Knaus WA, Wagner DP, Draper EA et al. The APACHE III prognostic system: Risk prediction of hospital mortality for critically ill hospitalized adults. Chest 1991; 100: 1619–1636 10. Douma CE, Redekop WK, Van der Meulen JHP et al. Predicting mortality in intensive care patients with acute renal failure treated with dialysis. J Am Soc Nephrol 1997; 8: 111–117 11. Parker RA, Himmelfarb J, Tolkoff-Rubin N, Chandran P, Wingard RL, Hakim RM. Prognosis of patients with acute renal failure requiring hemodialysis: results of a multicenter study. Am J Kidney Dis 1998; 32: 432–443 12. Maher ER, Robinson KN, Scoble JE, Farrimond JG, Browne RG, Sweny P, Moorhead JF. Prognosis of critically ill patients with acute renal failure: APACHE II score and other predictive factors. Q J Med 1989; 72: 857–866 13. Solez K, Morel-Maroger L, Sraer JD. The morphology of ‘acute tubular necrosis’ in man: analysis of 57 renal biopsies and comparison with glycerol model. Medicine 1979; 58: 362–376 14. Linas SL, Whittenburg D, Parsons PE, Repine JE. Mild renal

22. 23. 24.

25.

26.

27. 28. 29. 30. 31.

32.

33.

ischemia activates primed neutrophils to cause acute renal failure. Kidney Int 1992; 42: 610–616 Schulman G, Fogo A, Gung A, Badr K, Hakim R. Complement activation retards resolution of acute ischemic renal failure in the rat. Kidney Int 1991; 40: 1069–1074 Walton DF, Cheung AK. Membrane biocompatibility. In: Nissenson AR, Fine RN, Gentle DE, eds Clinical Dialysis third edition, Appleton & Lange, Connecticut. 1995: 93–120 Kra¨nzlin B, Reuss A, Gretz N, Kirschfink M, Ryan CJ, Mujais SK. Recovery from ischemic acute renal failure: independence from dialysis membrane type. Nephron 1996; 73: 644–651 Neveu H, Kleinknecht D, Brivet F, Loirat Ph, Landais P, and The French Study Group on Acute Renal Failure. Prognostic factors in acute renal failure due to sepsis. Results of a prospective multicentre study. Nephrol Dial Transplant 1996; 11: 293–299 Shaldon S. Biocompatible membranes in acute renal failure: prospective case-controlled study. Nephrol Dial Transplant 1997; 12: 235–236 Mehta R, Farkas A, Pascual M, Fowler W. Effect of serial APACHE and organ failure scores on prediction of mortality. J Am Soc Nephrol 1995; 6: 550 [Abstract] Jacobs C. Membrane biocompatibility in the treatment of acute renal failure: what is the evidence in 1996? Nephrol Dial Transplant 1997; 12: 38–42 Mujais SK, Ivanovich P, Vidovich MI. Dialysis in acute renal failure: which membrane? Arti Organs 1994; 18: 470–472 Simpson K, Allison MEM. Dialysis and acute renal failure: can mortality be improved? Nephrol Dial Transplant 1993; 8: 946 [Abstract] Mehta R, McDonald B, Gabbai F, Pahl M, Farkas A, Pascual M, Fowler W and ARF Collaborative Study Group. Effect of biocompatible membranes on outcome of acute renal failure. J Am Soc Nephrol 1996; 7: 1457 [Abstract] Assouad M, Tseng S, Dunn K, Gonzalez J, Brennan S, Suki W. Biocompatibility of dialyzer membrane is important in the outcome of acute renal failure. J Am Soc Nephrol 1996; 7: 1437 [Abstract] Liano F, Pascual J, and the Madrid Acute Renal Failure Study Group. Epidemiology of acute renal failure: a prospective, multicenter, community-based study. Kidney Int 1996; 50: 811–818 Cosentino F, Chaff C, Piedmonte M. Risk factors influencing survival in ICU acute renal failure. Nephrol Dial Transplant 1994; 9 [Suppl. 4]: 179–182 Hakim RM, Held PJ, Stannard DC et al. Effect of the dialysis membrane on the mortality of chronic hemodialysis patients. Kidney Int 1996; 50: 566–570 Elsen R, Silva LK. Advantages of meltspun membrane technology compared to solution spinning. Contrib Nephrol 1993; 103: 125–138 Splendiani G, Mazzarella V. Artif Organs 1996; 20: 281 Gastaldello K, Melot JL, Vanherweghem JL, Kahn RJ, Tielemans C. Cellulose diacetate (CDA) does as well as polysulfone for the treatment of acute renal failure (ARF ) in the intensive care unit (ICU ). J Am Soc Nephrol 1996; 7: 1447 [Abstract] Jo¨rres A, Dobis C, Gahl GM et al. and the multicenter study group. Hemodialysis membrane biocompatibility does not impact on the mortality of patients with acute renal failure: a prospective randomized international multicenter trial. J Am Soc Nephrol 1998; 9: 212A–213A [Abstract] Schiffl H, Lang S, Haider M. Biocompatibility of dialyzer membranes may have a negative impact on outcome of acute renal failure, independent of the dose of dialysis delivered. A retrospective multicenter analysis. ASAIO J 1998; 44: M418–M422

Received for publication: 8.3.99 Accepted in revised form: 28.9.99