J Am Soc Nephrol 11: 539–549, 2000 Vitamin E Attenuates Oxidative Stress Induced by Intravenous Iron in Patients on Hemodialysis JOHANNES M. ROOB,* GHOLAMALI KHOSCHSORUR,† ANDREAS TIRAN,‡ JORG¨ H. HORINA,* HERWIG HOLZER,* and BRIGITTE M. WINKLHOFER-ROOB§ *Division of Clinical Nephrology and Hemodialysis, Department of Internal Medicine, †Department of Laboratory Medicine I, ‡Department of Laboratory Medicine II, and §Institute of Biochemistry, Karl-Franzens University of Graz, Austria. Abstract. Intravenous iron application to anemic patients on cholesterol (5.88 Ϯ 1.09 mmol/mol) were normal, plasma hemodialysis leads to an “oversaturation” of transferrin. As a MDA concentrations were above normal (1.20 Ϯ 0.28 ␮mol/ result, non-transferrin-bound, redox-active iron might induce L), and bleomycin-detectable iron (BDI), indicating the pres- lipid peroxidation. To test the hypothesis that vitamin E atten- ence of redox-active iron, was not detectable. Upon iron infu- uates lipid peroxidation in patients receiving 100 mg of iro- sion, BDI and MDA concentrations increased significantly n(III) hydroxide sucrose complex intravenously during a he- (P Ͻ 0.001). BDI concentrations explained the increase over modialysis session, 22 patients were investigated in a baseline in MDA concentrations (MDA ϭ 1.29 ϩ 0.075 ϫ randomized cross-over design, either with or without a single BDI). Vitamin E supplementation, leading to a 68% increase in oral dose of 1200 IU of all-rac-␣-tocopheryl acetate taken 6 h plasma ␣-tocopherol concentrations, significantly reduced the ϭ before the hemodialysis session. Blood was drawn before and AUC0–180 min of MDA to cholesterol (P 0.004) and perox- 30, 60, 90, 135, and 180 min after the start of the iron infusion, ides to cholesterol (P ϭ 0.002). These data demonstrate that a and areas under the curve (AUC0–180 min) of ratios of plasma single oral dose of vitamin E attenuates lipid peroxidation in malondialdehyde (MDA) to cholesterol and plasma total per- patients on hemodialysis receiving intravenous iron. Given that oxides to cholesterol (two markers of lipid peroxidation) were intravenous iron is applied repeatedly to patients on hemodi- determined as the outcome variables. At baseline of the session alysis, this therapeutic approach may protect against oxidative without vitamin E supplementation, plasma ␣-tocopherol con- stress-related degenerative disease in the long term. centrations (27.6 Ϯ 1.8 ␮mol/L) and ratios of ␣-tocopherol to With the availability of recombinant human erythropoietin ferritin, with the protein ligand apoferritin replaced by sucrose, (rhEPO), treatment of patients with anemia associated with can be used. Iron in nonionic form as a water-soluble ferric chronic renal failure has changed substantially (1). Stimulation hydroxide complex is well tolerated and, being a large molec- of erythropoietic activity by rhEPO places enormous demands ular complex of approximately 43 kD (product characteriza- on an adequate iron supply. On average, 150 mg of iron is tion, Vifor International, St. Gallen, Switzerland), it is not needed for an increase in hemoglobin concentrations by 1 g/dl eliminated by the kidney or by hemodialysis treatment. (2). To replace iron losses and maintain adequate iron stores, Normally, iron is safely sequestered in transport proteins 1.5to2gofsupplemental iron per year is required for an such as transferrin and lactoferrin and stored in proteins such as individual patient on chronic hemodialysis. Given that oral ferritin and hemosiderin. In healthy subjects, transferrin satu- supplementation is frequently ineffective, intravenous iron ad- ration (TSAT), calculated from total serum iron and transferrin ministration has become the preferred clinical routine in many concentrations, is Յ45% (7). The doses recommended for iron hemodialysis centers (3,4). Subsequently, reduction of the supplementation in patients on chronic hemodialysis, i.e.,1to rhEPO dose led to significant cost reduction (5). Among other 4 mg iron/kg body weight or 100 to 200 mg iron, lead to an preparations (6), a polynuclear iron complex analogous to “oversaturation” of transferrin (8,9). Peak serum iron concen- trations depend not only on the dose, but also on the duration of the infusion: The higher the dose and the faster the appli- Received March 2, 1999. Accepted August 11, 1999. cation, the higher the peak iron concentrations (8). However, This work was presented in part at the 31st Annual Meeting of the American Society of Nephrology, Philadelphia, PA, October 1998, and has been pub- even infusion lasting 4 h led to an “oversaturation” of trans- lished in abstract form (J Am Soc Nephrol 9: 224A, 1998). ferrin (8). High percentage TSAT was found to be associated Correspondence to Dr. Brigitte M. Winklhofer-Roob, Institute of Biochemis- with the presence of non-transferrin-bound, potentially redox- try, Karl-Franzens University, Schubertstrasse 1, A-8010 Graz, Austria. Phone: active iron, and iron complexed with citrate or acetate, i.e., low ϩ43 316 380 5490; Fax: ϩ43 316 380 9857; E-mail: brigitte.winklhoferroob@ kfunigraz.ac.at molecular weight complexes, was shown to be redox-active 1046-6673/1103-0539 (10,11). Journal of the American Society of Nephrology Redox-active iron is a potent pro-oxidant (8,10,11). Hy- Copyright © 2000 by the American Society of Nephrology droxyl radical and lipid alkoxyl radical, formed by the Fenton 540 Journal of the American Society of Nephrology J Am Soc Nephrol 11: 539–549, 2000 reaction, represent the reactive oxygen species that trigger was administered intravenously during hemodialysis 2 to 3 times per iron-induced lipid peroxidation in the presence of hydrogen week at a weekly dose of 2,000 to 30,000 IU (mean Ϯ SD, 10,136 Ϯ peroxide or lipid hydroperoxides. These, like any other oxygen 7,266 IU). None took vitamin E supplements. Inclusion criteria were Ͻ ␮ Ͻ free radicals, can initiate the chain reaction of lipid peroxida- serum ferritin concentrations 100 g/L and/or TSAT 20%1mo tion by giving rise to the formation of a lipid radical from a before the start of this study. The study was approved by the Ethics Committee of the University Hospital and Faculty of Medicine, Uni- polyunsaturated fatty acid (PUFA). In different in vitro models versity of Graz, and informed consent was obtained from the patients. and in the intact animal, iron has been shown to initiate lipid peroxidation (12–16), the consequences of which are distur- Study Design bances of tissue and organ functions (17,18). Evidence has Study Aim A. All patients were investigated twice in a random- accumulated that oxidative modification of LDL is causally ized, two-period cross-over design, 7 d apart with and without sup- involved in atherogenesis (19). Vitamin E is a potent antioxi- plementation of a single oral dose of vitaminE6hbefore the start of dant that terminates the chain reaction of lipid peroxidation the hemodialysis session. All patients received iron(III) hydroxide (20). It has been demonstrated to inhibit lipid peroxidation in sucrose complex intravenously on each of the two occasions. The iron animals and human subjects (21) and to enhance the resistance infusion was started 30 min after the hemodialysis session had begun of LDL to copper(II) ion-induced oxidation both in healthy and lasted for 20 min. subjects (22) and patients with impaired vitamin E status (23). Study Aim B. To further explore the effect of iron treatment in The question of whether redox-active iron occurs as an the absence and presence of vitamin E supplementation, patients were immediate response to intravenous iron application with a also investigated during a hemodialysis session without iron applica- frequently used therapeutic dose and mode of application and tion, using the same variables as in the other sessions. This session what the effects are on in vivo lipid peroxidation have not been was performed 1 mo after the randomized cross-over trial, assuming that period effects are negligible. addressed before in patients on chronic hemodialysis. Given that redox-active iron causes lipid peroxidation, it could rep- resent the critical link between oversaturation of transferrin and Treatment ® lipid peroxidation. The “bleomycin assay,” developed by Gut- Iron. The iron preparation used (Venofer , Vifor, Inc., St. Gallen, Switzerland) was a solution of iron(III) hydroxide sucrose teridge et al. (24), allows quantification of bleomycin-detect- complex of approximately 43 kD containing 2% iron (20 mg Fe per able iron (BDI), a marker of non-transferrin-bound iron that milliliter of injectable solution, pH 10.5 to 11.0). A very small has the potency of becoming redox-active. In contrast, iron proportion (0.14%) of the iron was found not to be present as a high bound to proteins is not detected. Among different indexes for molecular weight complex when we separated iron according to assessing in vivo lipid peroxidation in human subjects, plasma molecular weight, using an ultrafiltration membrane (Vivaspin 500, malondialdehyde (MDA) concentrations are most frequently Vivascience Ltd., Binbrook, Lincoln, United Kingdom) that discrim- used (25). MDA is an end product of nonenzymatic, oxidative inates by molecular weight of 10 kD, and determined iron levels in the degeneration of PUFA containing three or more conjugated ultrafiltrate by the FerroZine method using a Hitachi analyzer. The double bonds (26,27). content of the ampoules (5 ml) was diluted with sterile 0.9% NaCl The purpose of this study was to test the hypothesis that a solution to give a total volume of 50 ml that was administered slowly single oral dose of vitamin E taken before intravenous iron over 20 min by infusion via the venous line of the extracorporeal circuit, using Pilote C from Fresenius Vial SA (Brezins, France). A application attenuates lipid peroxidation, which occurs in pa- dose of 100 mg was chosen because similar doses are frequently tients receiving iron(III) hydroxide sucrose complex intrave- applied to hemodialysis patients.