Red blood cell fatty acid profile of chronic renal failure patients receiving maintenance haemodialysis treatment

Prostaglandins, Leukotrienes and Essential FattyAcids (2002) 67(1),13^18 & 2002 Elsevier Science Ltd. All rights reserved. doi:10.1054/plef.375, available online at http://www.idealibrary.com on

Red blood cell fatty acid profile of chronic renal failure patients receiving maintenance haemodialysis treatment A. M. Koorts,1 M.Viljoen,1 M. C. Kruger2 1

Department of Physiology, University of Pretoria, Pretoria, South Africa Milk and Health Research Centre, Institute of Food, Nutrition and Human Health, Massey University, Palmerston North, New Zealand

2

Summary The fatty acid profile of chronic renal failure (CRF) patients on maintenance haemodialysis treatment (MHT) is abnormal and results point towards an essential fatty acid (EFA) deficiency. However, controversies still exist as to which of the essential fattyacids (EFAs) or EFA-products are decreased.In this study, the results of a comprehensive analysis of the fatty acids, performed on the red blood cells of14 CRF patients on MHT, are presented.The red blood cell membrane fatty acids determined in this study include a range of saturated fatty acids (SFAs), mono-unsaturated fatty acids (MUFAs) and poly-unsaturated fatty acids (PUFAs).Results confirmed the suggested presence ofan essential fattyacid deficiencyin CRF patients on MHT.It showed the total content of the n-6 fatty acids (31.6673.21vs 34.6772.05), as well as the total content of the PUFAs (37.2274.08 vs 40.9372.35), to be significantly decreased.The total MUFA content was, in contrast, significantly increased (16.8770.91vs15.4971.18).The EFA deficiencyprofile seeninthis studypointstowardsthat of a chronic inflammatory condition.Thisisborne out by the fact that all three precursors of the eicosanoidsFthe mediators of various inflammatory and immune responsesFwere reduced in the presence of an increase in MUFAs as well as SFAs.The possibility of the fatty acid profile of CRF patients on MHT being that of a chronic inflammatory condition is supported by the fact that continuous complement-dependent and complement-independent immune activation, due to bio-incompatibility between blood cells and the dialysis membranes, is known to occur during the dialysis process. & 2002 Elsevier Science Ltd. All rights reserved.

INTRODUCTION The human body contains four parent fatty acids, i.e., oleic acid, palmitoleic acid, linoleic acid and a-linolenic acid from which the various poly-unsaturated fatty acids are synthesised.1 Two of these fatty acids, linoleic acid and a-linolenic acid, are classified as essential fatty acids (EFAs) since the enzymes crucial for the synthesis of these fatty acids are not available in the human body.1,2 The

Received 9 July 2001 Accepted 4 March 2002 Correspondence to: M.C. Kruger, Milk and Health Research Centre, Institute for Food, Nutrition and Human Health, Massey University, Private Bag 11222, Palmerston North, New Zealand.Tel.: 64-6-350-5799X2936; Fax: 64-6-3505446; E-mail: [email protected] This work was funded by Research committee, University of Pretoria, and National Research Foundation, South Africa.

& 2002 Elsevier Science Ltd. All rights reserved.

only means by which adequate quantities can be obtained is by the consumption of food containing these EFAs. The EFAs are incorporated into the membranes of cells as part of the membrane phospholipids. They form the precursors of various other fatty acids, participate in the transport and oxidation of cholesterol and it is from these fatty acids that the lipid mediators; prostaglandins, thromboxanes and leukotrienes are synthesised.1,2 The EFAs, therefore, play an important role in various biological functions and an inadequate intake of these fatty acids can result in the clinical condition known as essential fatty acid deficiency (EFAD).1 The fatty acid profile of chronic renal failure (CRF) patients receiving maintenance haemodialysis treatment (MHT) are abnormal and possibly reflects the presence of an essential fatty acid deficiency.3–8 Various studies investigating the fatty acid profile of CRF patients on Prostaglandins, Leukotrienes and Essential FattyAcids (2002) 67(1), 13^18

14

Koorts et al.

maintenance haemodialysis treatment reported either an increase or a decrease in specific fatty acids. Although controversies still exist, it generally seems that these patients have an EFAD and therefore an increase in those fatty acids enzymatically produced in EFA deficiency. The possibility of an EFAD, suggested by studies on the composition of the membranes, is supported by clinical symptoms. CRF patients on maintenance haemodialysis treatment display many of the symptoms associated with EFAD. Such symptoms include increased erythrocyte fragility,3,7 platelet dysfunctions,3,9 dermatological problems such as dry and scaly skin,3,7,8 increased incidence of ischaemic heart disease due to cardiovascular autonomic dysfunction,6 cerebrovascular accidents,3,6 as well as a high risk of premature labour and retarded foetal growth in the few CRF patients who are not completely sterile.5 Susceptibility to infection due to impaired immune function is a further fatty acid deficiency associated condition also seen in CRF patients.3,7 Although various studies point towards the existence of EFAD and the effective treatment of EFAD symptoms by the proper EFA supplementation in MHT patients,8 controversies still exist as to which of the essential fatty acids and/or essential fatty acid products are decreased in dialysis patients. In order to obtain a more detailed picture of the membrane lipid abnormalities in CRF patients on maintenance haemodialysis treatment, we present the results of a comprehensive analysis of red blood cell membrane fatty acids of 14 CRF patients receiving maintenance haemodialysis treatment.

MATERIALS AND METHODS The study was approved by the Ethics Committee of the Faculty of Medicine, University of Pretoria (116/98). All patients and control subjects gave written informed consent. The blood was drawn pre-dialysis from the arterial line, before any mixing of blood and heparinised saline occurred. The procedure for the sampling of control subject blood was comparable with that of the patient. Ten millilitre of EDTA-anticoagulated blood was collected and kept on ice in the dark for red blood cell membrane fatty acid determinations.

Red blood cell preparation for membrane fatty acid determinations The blood was centrifuged at 1800g for 15 min at 41C. The supernatant and buffy coat was discarded, the cells were washed twice with cold 0.15 M NaCl, and centrifuged at 1800 g for 10 min at 41C. The washed red blood cells were stored at 701C. Prostaglandins, Leukotrienes and Essential FattyAcids (2002) 67(1), 13^18

Fatty acid analyses RBCs were extracted with chloroform/methanol (2:1 v/v) according to a modification of the method of Folch et al.10 All the extraction solvents contained 0.01% butylated hydroxytoluene (BHT) as an antioxidant. Heptadecanoic acid (C17:0) was used as internal standard to quantify the individual fatty acids. The lipids were transmethylated using 5% H2SO4/methanol at 701C for 2 h. After cooling, the resulting fatty acid methyl esters (FAMEs) were extracted with 1 ml of water and 2 ml of n-hexane. The top hexane layer was evaporated to dryness, redissolved in CS2 and analysed by GLC (Varian Model 3300 equipped with flame ionisation detection) using 30 m fused silica megabore DB-225 columns of 0.53 mm internal diameter ( J&W Scientific, Folsom, CA, USA). Gas flow rates were: hydrogen, 25 ml/min; air, 250 ml/min; and hydrogen (carrier gas), 5–8 ml/min. Temperature programming was linear at 41C/min, initial temperature 1601C, final temperature 2201C, injector temperature 2401C, and detector temperature 2501C. The FAMEs were identified by comparison of the retention times to those of a standard FAME mixture (Nu-Chek-Prep Inc., Elysian, MN, USA).

Statistical analyses The non-parametric Mann–Whitney test was applied for statistical comparisons between the patient group and the control group. A P-value less than 0.05 was taken as a significant difference and a P-value less than 0.1 and greater than 0.05 was taken as a non-significant difference.

RESULTS Fourteen CRF patients on MHT at the Pretoria Academic Hospital were suitable for inclusion in this study. Patients where doubt existed with regard to relevant medication, where the treatment regimen was altered within the time period of this study, those whose haematocrits were too low, or who were on the programme for less than 1 year were excluded from the study. Table 1 contains the relevant clinical characteristics of the patient group. The patient group consisted of 14 patients, 9 male, 5 female, with 6 caucasian and 8 black. The mean and the standard deviation for age for the 14 patients were 37.29 and 12.55 and the mean and standard deviation for the years on dialysis treatment were 5.82 and 6.10. The control group consisted of 10 individuals, 6 male, 4 female, 5 caucasian, 5 black. The mean and the standard deviation for age for the control subjects were 38.10 and 14.68. The red blood cell membrane fatty acids determined include saturated fatty acids (SFAs), mono-unsaturated fatty acids (MUFAs), and poly-unsaturated fatty acids & 2002 Elsevier Science Ltd. All rights reserved.

Red blood cell fatty acid profile

15

Table 1 Clinical information of the maintenance haemodialysis patients (MHT) Patient

Race

Sex

Age

Aetiology of CRF

Period on dialysis (years)

Dialysis protocol (h/week)

EPO

Ca channel blocker (mg)

1 2 3 4 5 6 7 8 9 10 11 12 13 14

Caucasian Caucasian Caucasian Caucasian Caucasian Black Black Black Black Black Black Black Caucasian Black

Male Female Female Female Female Male Male Male Male Male Female Male Male Male

61 39 54 16 36 27 21 36 38 47 34 30 32 51

Chronic glomerulonephritis Chronic pyelonephritas Polycystic kidney disease SLE Gassers Pancreatitis Hypertension Glomerulonephritis Hypertensive nephropathy Hypertensive nephropathy Glomerulonephritis Glomerulonephritis Trauma Hypertensive nephropathy

16 8.5 4 1 22.5 1.5 3.5 2.5 3 5 4 4 2 4

34 34 34 34 34 34 34 34 34 34 34 34 34 34

None None None None None 0.38 ml  3 week 0.55 ml  3 week 0.25 ml  3 week 0.41ml 1week 0.33 ml  3 week 0.43 ml  3 week 0.50 ml  3 week 0.35 ml  3 week 0.47 ml 1week

None None 10 None None 10 10 10 10 10 10 10 10 10

(PUFAs). The SFAs determined include; 14:0, 16:0, 18:0, 20:0 and 22:0. The MUFAs determined include; 16:1, 18:1n-9, 20:1n-9 and 22:1n-9. The PUFA’s determined include; 18:2n-6, 18:3n-6, 18:3n-3, 20:2n-6, 20:3n-6, 20:4n-6, 20:3n-3, 20:5n-3, 22:2n-6, 22:4n-6, 22:5n-6, 22:5n-3 and 22:6n-3. EFA content was expressed as wt%. Tables 2 and 3 contain the data and statistical comparison between the groups. A significant difference was detected between the patients and the control subjects for the content of 22:0 in the red blood cell membranes. It was increased in the CRF patients (  1.89, SD 0.39) compared to the control subjects (  1.542, SD 0.18), P-value 0.011. Several MUFAs were significantly different between the patients and the control subjects. The content of 18:1n-9 was higher in the patients (  13.341, SD 0.864) compared to the control subjects (  12.228, SD 0.999), P-value 0.0092. The content of 20:1n-9 was significantly higher in the patients (  0.3386, SD 0.131) compared to the control group (  0.222, SD 0.036), P-value 0.0034, and the content of 22:1n-9 was significantly higher in the patient group (  0.1171, SD 0.052) compared to the control group (  0.064, SD 0.033), P-value 0.0139. The total MUFA content of the red blood cell membranes of the patients was significantly higher compared to controls (Table 3) (Fig. 1). Three of the n-6 fatty acids were lower in the CRF patients compared to the control subjects, though not all differences were statistically significant. The content of 18:2n-6 was non-significantly reduced in the group of patients (  10.39, SD 2.205) compared to the control subjects (  12.111, SD 1.66), P-value 0.084. The content of 20:3n-6 was significantly reduced in the group of patients (  1.322, SD 0.373) compared to the control subjects (  1.624, SD 0.441), P-value 0.046 and the content of 20:4n-6 was non-significantly reduced in the group of patients (  14.808, SD 1.404) compared to & 2002 Elsevier Science Ltd. All rights reserved.

the control subjects (  15.923, SD 1.4697), P-value 0.1073. The total content of n-6 fatty acids was significantly decreased in the patients as compared to the controls (Table 3) (Fig. 2). Only one of the n-3 fatty acids (20:5n-3) was significantly decreased in the patients (  0.202, SD 0.070) compared to the control subjects (  0.282, SD 0.088), P-value 0.032. The total content of PUFAs was significantly reduced in the patients group compared to control (P = 0.037) (Table 3, Fig. 3). None of the patients were on vitamin supplementation. Analysis of the plasma vitamin C, E and A levels indicated high vitamin A levels in the renal dialysis patients and no difference in the vitamin E and C levels ( data not shown). DISCUSSION Various studies on the membrane fatty acid profile of CRF patients on MHT reported abnormalities in specific fatty acids.3–8 Although it is suggested that the EFA content is abnormal, no consistency exists as to the reported increase or decrease in specific fatty acids. In general, the profile would appear to be characterised by a decrease in the n-3 and n-6 series EFAs and EFA products such as dihomo-gamma linolenic acid (20:3n-6), arachidonic acid (20:4n-6), eicosapentanoic acid (20:5n-3) and docosahexanoic acid (22:6n-3). In addition to these deficiencies, the membrane fatty acid profile of CRF patients receiving MHT has been reported to display an increase in the n-9 series non-essential fatty acids including oleic acid (18:1n-9) and eicosatrienoic acid (20:3n-9).3,7 In the present study, the red blood cell membrane fatty acid composition of 14 CRF patients on MHT was investigated. A decrease in the content of the EFA, linoleic acid, and three of the EFA metabolic products, dihomo-gamma linolenic acid, arachidonic acid and Prostaglandins, Leukotrienes and Essential FattyAcids (2002) 67(1), 13^18

16

Koorts et al.

Table 2

Statistical comparison of the MHT patients (n=14) and control subjects (n=10). EFA content is expressed as wt%

Variable fatty acids F14:0 F16:0 F16:1 F18:0 F18:1n-9 F18:2n-6 F18:3n-6 F18:3n-3 F20:0 F20:1n-9 F20:2n-6 F20:3n-6 F20:4n-6 F20:3n-3 F 20:5n-3 F22:0 F22:1n-9 F22:2n-6 F22:4n-6 F22:5n-6 F22:5n-3 F22:6n-3 a b

mean SD mean SD mean SD mean SD mean SD mean SD mean SD mean SD mean SD mean SD mean SD mean SD mean SD mean SD mean SD mean SD mean SD mean SD mean SD mean SD mean SD mean SD

Patients

Controls

P-value Mann^Whitney

0.3457 0.1933 22.331 2.489 0.2 0.2005 17.352 1.6535 13.341 0.8647 10.393 2.2057 0.0457 0.0344 0.1679 0.0645 0.4771 0.203 0.3386 0.1316 0.4364 0.2035 1.3221 0.3736 14.808 1.404 0.0393 0.0385 0.2021 0.0707 1.89 0.3981 0.1171 0.0522 0.1129 0.0515 3.8714 0.6037 0.6729 0.4572 1.4057 0.2789 3.7436 0.8128

0.394 0.1097 21.209 0.8473 0.173 0.1008 16.638 1.1021 12.228 0.9999 12.111 1.6661 0.033 0.0216 0.127 0.0435 0.384 0.0617 0.222 0.0365 0.419 0.1303 1.624 0.4414 15.923 1.4697 0.03 0.0267 0.282 0.0883 1.542 0.1813 0.064 0.0334 0.113 0.0683 3.699 0.3949 0.748 0.2104 1.564 0.2228 4.26 0.6533

0.1977 0.1877 0.6819 0.7035 0.0092a 0.0841b 0.334 0.1688 0.2659 0.0034a 0.7697 0.0465 a 0.1073b 0.8836 0.0326a 0.0109a 0.0139a 0.6395 0.7474 0.5195 0.3641 0.2188

Significant difference. Non-significant difference.

eicosapentanoic acid was shown. The 18:2n-6 (linoleic acid) content was non-significantly reduced in the MHT patients compared to the control subjects while 20:3n-6 (dihomo-gamma linolenic acid) was significantly and 20:4n-6 (arachidonic acid) non-significantly reduced in the MHT patients, compared to the control subjects. Only one of the n-3 series fatty acids was significantly reduced in the MHT patients compared to controls, i.e., 20:5n-3 (eicosapentanoic acid) The non-significance of the decrease in linoleic acid and arachidonic acid contents could be the result of the small number of subjects, since all other results point towards EFA deficiency. Furthermore, in the present study the 18:1n-9 (oleic acid) content was significantly elevated in the patients, compared to the control subjects. An elevation in 20:3n-9 Prostaglandins, Leukotrienes and Essential FattyAcids (2002) 67(1), 13^18

content is also characteristic of EFA deficiency, but the content of 20:3n-9 was less than 0.05% of total fatty acid content and therefore could not be accurately quantified. However, in the present study an elevation in the content of two other n-9 fatty acids was found. The 20:1n-9 and 22:1n-9 fatty acid contents were significantly elevated in the MHT patients. The EFA linoleic acid and a-linolenic acid are, respectively, the building blocks for the n-3 series and n-6 series of PUFA, whereas oleic acid is the building block for the n-9 series PUFA. The same desaturase and elongase enzymes are responsible for the formation of the n-3, n-6 and n-9 series of metabolites from linoleic acid, a-linolenic acid and oleic acid but show a preference for the substrate containing the highest number of double & 2002 Elsevier Science Ltd. All rights reserved.

Red blood cell fatty acid profile

Table 3

17

Statistical comparison of the MHT patients (n=14) and control subjects (n=10). EFA content is expressed as wt%

Variable SFAs

mean SD mean SD mean SD mean SD mean SD mean SD mean SD

MUFAs PUFAs PUFAs/SFAs n-3 n-6 n-6/n-3

Patients

Controls

P-value Mann^Whitney

45.905 4.352 16.871 0.9126 37.221 4.0832 0.8247 0.1597 5.5586 1.0391 31.662 3.2149 5.8446 0.9996

43.577 1.6851 15.494 1.1897 40.933 2.351 0.9423 0.0884 6.263 0.8952 34.67 2.0548 5.6261 0.7728

0.4292 0.0065a 0.0376a 0.0371a 0.1514 0.0376a 0.93

SFAs=F14:0+F16:0+F18:0+F20:0+F22:0+F24:0 MUFAs=F16:1+F18:1+Fn-9 20:1+Fn-9 22:1+Fn-9 24:1 PUFAs=Fn-6 18:2+Fn-6 18:3+Fn-318:3+Fn-6 20:2+Fn-6 20:3+Fn-6 20:4+Fn-3 20:3+Fn-3 20:5+Fn-6 22:2+Fn-6 22:4+Fn-6 22:5+Fn-3 22:5+ Fn-3 22:6 n-3=Fn-318:3+Fn-3 20:3+Fn-3 20:5+Fn-3 22:5+Fn-3 22:6 n-6=Fn-6 18:2+Fn-6 18:3+Fn-6 20:2+Fn-6 20:3+Fn-6 20:4+Fn-6 22:2+ Fn-6 22: 4+Fn-6 22:5 a Significant difference.

45

19

*

17

40

15

35

13

30

patients

control

Fig. 1. Box and Whisker plot showing the difference in MUFA content of the cell membranes of the MHT patients as compared to the controls (wt %).*P=0.006. 40

* 35

30

25

patients

control

Fig. 2 Box and Whisker plot showing the difference in n-6 fatty acid content of the cellmembranes ofthe MHT patients as compared to the controls (wt %). *P = 0.04.

bonds with a-linolenic acid 4linoleic acid 4oleic acid. When a deficiency of the EFA exists this will result in an increase in the n-9 series fatty acids.8 Therefore, the increase in some of the n-9 series fatty acids further & 2002 Elsevier Science Ltd. All rights reserved.

*

patients

control

Fig. 3 Box and Whisker plot showing the difference in PUFA content of the cell membranes of the MHT patients as compared to the controls (wt%). *P=0.04.

supports the hypothesis of EFA deficiency in CRF patients on MHT. There are various possible causes that may result in an EFA deficiency profile such as seen in the MHT patients of this study. These include an inadequate intake of EFAs,1 an over-consumption of the EFA and their products such as with a chronic inflammatory condition,8 membrane PUFA oxidation due to an increased oxidative stress11 and an inappropriate rise in intracellular free calcium resulting in membrane breakdown.12 The most probable causes for the disturbed fatty acid profile seen in this study are the presence of a chronic inflammatory condition or of increased intracellular calcium concentrations. The two causes are of course not mutually exclusive as increased intracellular calcium activity forms part of the profile of a chronic inflammatory condition. The conclusion of a chronic inflammatory condition being present in CRF patients on MHT is supported by several observations of the fatty acid profile Prostaglandins, Leukotrienes and Essential FattyAcids (2002) 67(1), 13^18

18

Koorts et al.

seen in this study. Various fatty acids derived from the EFA are the precursors for the synthesis of the lipid mediators known as the eicosanoids.1 The eicosanoids are divided into three classes known as the series-1, series-2 and series-3 eicosanoids. The series-1 eicosanoids are derived from dihomo-gamma linolenic acid, a product of the desaturation and elongation of linoleic acid. The series-2 eicosanoids are derived from arachidonic acid, also a product of the desaturation and elongation of linoleic acid whereas the series-3 eicosanoids are derived from eicosapentaneoic acid, a product of the desaturation and elongation of a-linolenic acid.1,3 All three of these precursors were reduced in the present study. The eicosanoids are involved as mediators of various inflammatory and immune responses. It is known that in CRF patients receiving MHT, complement-dependent and complement-independent activation of various immune cells occur. This activation of the immune cells is said to be the result of the blood-membrane interaction during extracorporeal circulation, and include the activation of neutrophils, monocytes and platelets.13–15 This immune response is mostly the result of the bioincompatibility between the blood and the dialysis membranes. With such an activation of immune cells during dialysis the eicosanoids, leukotrienes and thromboxanes are generated and it has even been proposed that the increase in these lipid mediators is indicative of the bio-incompatibility of the dialysis membranes.7,16 The eicosanoids for which an increase was previously reported in MHT patients include the 5-lipoxygenase products leukotriene C4 and leukotriene B4, and the cyclo-oxygenase products thromboxane B2 and prostaglandin E2.9,13 Furthermore, it has also been shown that the production of the various eicosanoids is inadequate when the need arise due to the decreased fatty acid precursor content. Further support for deranged immunological functions in association with EFA disturbances is derived from the fact that supplementation of n-3 PUFA-deficient patients with a-linolenic acid or eicosapentaenoic acid may attenuate immunological reactivity, as indicated by the mitogen response of isolated lymphocytes.2 In conclusion it can be said that the results of this study confirms previous suggestions that end stage renal failure patients on haemodialysis treatment suffers from an EFA deficiency. It is suggested that this deficiency may be the result of a chronic inflammatory condition largely caused by immune activation as a result of blood-membrane bioincompatibility during transit of the blood through the extracorporeal circuit. This suggestion is supported, not only by the fatty acid profile of this study, which points towards an excessive consumption of EFAs in favour of eicosanoid synthesis, but also by the extracorporeal immune activation shown by other workers. Prostaglandins, Leukotrienes and Essential FattyAcids (2002) 67(1), 13^18

ACKNOWLEDGEMENTS The authors thank the Research committee, University of Pretoria, and National Research Foundation, South Africa, for funding. They also thank Marius Smuts, National Programme for Nutritional Intervention, Medical Research Council, South Africa, for fatty acid analyses.

REFERENCES 1. Sardesai V. M. The essential fatty acids. Nutr Clin Prac 1992; 7: 179–186. 2. Peck L. W., Monsen E. R., Ahmad S. Effect of three sources of longchain fatty acids on the plasma fatty acid profile, plasma prostaglandin E2 concentrations, and pruritus symptoms in HD patients. Am J Clin Nutr 1996; 64: 210–214. 3. Ahmad S., Dasgupta A., Kenny M. A. Fatty acid abnormalities in hemodialysis patients: effect of L-carnitine administration. Kidney Int 1989; 36(Suppl 27): S243–S246. 4. Kalofoutis A., Nicolaidou-Politis V., Bouloukos A. Significance of lymphocyte fatty acid changes in chronic renal failure. Nephron 1996; 73: 704–706. 5. Ramsay B., Cream J. J., Curtis J. R., Manku M. S., Stewart J. C. M. Erythrocyte membrane docosahexanoic acid in haemodialysis patients on erythropoetin. The Lancet 1992; 339: 1232–1233. 6. Christensen J. H., Aaroe J., Knudsen N., Dideriksen K., Kornerup H. J., Dyerberg J. Heart rate variability and n-3 fatty acids in patients with chronic renal failureFa pilot study. Clin Nephrol 1998; 49(2): 102–106. 7. Dasgupta A., Kenny M. A., Ahmad S. Abnormal fatty acid profile in chronic hemodialysis patients: possible deficiency of essential fatty acids, Clin Physiol Biochem 1990; 8: 238–243. 8. Peck L. W. Essential fatty acid deficiency in renal failure. Can supplements really help? J Am Dietary Assoc 1997; 97(10 Suppl 2): S150–S153. 9. Ylikorkala O., Huttunen K., Ja¨rvi J., Viinikka L. Prostacyclin and thromboxane in chronic uremia: effect of hemodialysis. Clin Nephrol 1982; 18(2): 83–87. 10. Folch J., Lees M., Sloane-Stanley G. H. A simple method for the isolation and purification of total lipids from animal tissues. J Biol Chem 1957; 226: 497–509. 11. Taccone-Gallucci M., Lubrano R., Clerico A., Meloni C., Morosetti M., Meschini L. Administration of GSH has no influence on the RBC membrane: oxidative damage to patients on hemodialysis. ASAIO J 1992; 38: 855–857. 12. Duncan C. J. Calcium, Oxygen Radicals and Cellular Damage. Cambridge, UK: Cambridge University Press, 1991. 13. Hornych A., Remy P., Luong N., Aumont J., Bariety J. Biocompatibility of haemodialysis membranes: heamodialysisrelated Leukotriene B4 and C4 generation. Nephron 1996; 74: 11–18. 14. Foidart J. B., Davin J.-C., Malaise M., Saint-Remy M., Mahieu P. Stimulation of platelet lipoxygenase during hemodialysis. Kidney Int 1988; 33(Suppl 24): S80–S81. 15. Viljoen M., de Oliveira A. A., Milne F. J. Physical properties of the red blood cells in chronic renal failure. Nephron 1991; 59: 271–278. 16. Schultze G., Heitz J., Krais T. H., Neumayer H.-H., Wagner K., Molzahn M. The impact of eicosanoids on compliance, cardiovascular performance, and coagulation during hemodialysis. Biomed Biochim Acta 1984; 43(8/9): S426–S429.

& 2002 Elsevier Science Ltd. All rights reserved.