Intravenous Iron Exacerbates Oxidative DNA Damage in Peripheral Blood Lymphocytes in Chronic Hemodialysis Patients

Abstract

Patients undergoing maintenance hemodialysis have elevated markers of oxidative stress, but the reasons for this are not fully understood. Intravenous administration of iron, which many of these patients receive, may provoke the generation of bioactive iron, which enhances oxidative stress and lipid peroxidation. In this study, 110 hemodialysis patients were randomly assigned to five groups that were administered single intravenous doses of iron sucrose, ranging from 20 to 500 mg. A time- and dosage-dependent rise in lymphocyte 8-hydroxy-2′-deoxyguanosine (8-OHdG) levels in lymphocyte DNA, a marker of oxidative DNA damage, with a significant increase at 2 h after intravenous iron of ≥200 mg (P < 0.05). Four weeks later, patients were randomly assigned to weekly iron sucrose (100 mg of elemental iron) or saline for 12 wk, and 89 patients completed the study. Mean lymphocyte 8-OHdG content was significantly higher in patients receiving intravenous iron compared with control subjects (P < 0.05), especially in those with ferritin levels >500 μg/L. In addition, flow cytometric techniques revealed increased production of reactive oxygen species in lymphocytes among those treated with intravenous iron. Treatment with intravenous iron but not saline was also associated with decreased plasma ascorbate and α-tocopherol levels and increased oxidized glutathione/reduced glutathione ratio (P < 0.05). In summary, intravenous iron sucrose provokes oxidative damage to peripheral blood lymphocyte DNA in hemodialysis patients, especially among those with high levels of ferritin.

Cumulative evidence indicates that oxidative stress frequently occurs in patients undergoing maintenance hemodialysis (HD) as a result of a decrease of antioxidant defenses and an overproduction of reactive oxygen species (ROS), such as superoxide anion, hydrogen peroxide, hydroxyl radical, and hypochlorous acid. ^1–4 Direct measurement of ROS is difficult because of its lower plasma levels and shorter half-life; however, measurement of oxidative byproducts combined with the assessment of antioxidant levels become more feasible. Oxidative damage to cellular constituents, such as membrane lipids, protein, and DNA, are fingerprints of oxidative byproducts. Recently, investigators disclosed that 8-hydroxy 2′-deoxyguanosine (8-OHdG) is one of the most abundant oxidative DNA products and a sensitive DNA damage marker because it can be detected by a HPLC electrochemical detection method in the femtomole range.^5,6 Our previous works identified peripheral blood leukocytes to be valuable for monitoring 8-OHdG level of cellular DNA in both HD and peritoneal dialysis patients.^7–9 Accordingly, measurement of 8-OHdG in leukocyte DNA is a useful tool to assess the oxidative DNA damage in long-term dialysis patients. ^7–9

The potential role of intravenous iron therapy in enhancing the efficacy of erythropoiesis-stimulating agents (ESA) in patients receiving regular dialysis has received increasing attention in recent years.¹⁰ The important reason for intravenous iron therapy is that it helps to reduce ESA requirements, allows dialysis patients to achieve increased hemoglobin levels, and derives more cost-effectiveness from ESA treatment^11–13; however, acute intravenous administration of high iron dosages may provoke the generation of bioactive iron (nontransferrin binding iron) and subsequently enhance oxidative stress with an increase in lipid peroxidation and acute impairment of forearm flow–mediated dilation in healthy individuals.^14,15 Drueke et al.¹⁶ disclosed an interrelation among the common carotid artery intima-media thickness, advanced oxidation products of proteins (AOPP), and the annual intravenous iron dosage administered in patients receiving regular HD. It is speculated that long-term repeated intravenous administration of iron might exaggerate oxidative stress and potentially do much harm to the human body. Thereafter, it deserves further appraisal of the potential hazards of enhanced oxidative stress in long-term HD patients receiving maintenance intravenous iron therapy. Increased concentration of malondialdehyde (MDA), one of the byproducts of the peroxidation of polyunsaturated fatty acids, has been reported in HD patients receiving intravenous iron therapy.^17,18 Other investigators further reported increased plasma levels of protein carbonyls and AOPP in long-term HD patients with intravenous iron supplementation.^16,19 In contrast to lipid, sugars, and proteins,^16–19 the reactions of DNA with oxidants associated with intravenous iron therapy have not been well studied in HD patients.

We therefore conducted a prospective, randomized, controlled study to explore whether intravenously administered iron could enhance oxidative damage to peripheral blood lymphocyte DNA among long-term HD patients. We first investigated the single dosage effect as well as the cumulative dosage effect of intravenous iron sucrose on lymphocyte 8-OHdG levels. To assess the effect of imbalance between ROS production and antioxidant defense on oxidative DNA damage, we not only measured the intracellular ROS production by circulating lymphocytes and plasma antioxidant levels in HD patients after intravenous iron therapy but also assessed their relations with the 8-OHdG generation. So far, the upper limit of serum ferritin at which too much iron supplementation is being administered is uncertain. Recently, Kidney Disease Outcomes Quality Initiative (KDOQI) guidelines had reduced the top level of ferritin for withholding iron therapy from 800 to 500 μg/L.²⁰ It is still debatable that iron depletion is unlikely and tissue iron overload common at ferritin levels >500 μg/L.^21,22 Finally, we aimed to appraise the safe upper levels of serum ferritin for intravenous iron therapy in terms of provocation of oxidative stress and DNA damage in HD patients.

RESULTS

Between July 2005 and June 2006, 110 patients were randomly allocated to five groups for assessment of the single-dose effect of intravenous iron sucrose. Four weeks later, all patients were randomly assigned to receive either normal saline (control group, n = 55) or iron sucrose (intravenous iron group, n = 55) for evaluation of cumulative dosage effect (Figure 1). The causes for chronic renal failure were glomerulonephritis (n = 30), interstitial nephritis (n = 11), diabetic nephropathy (n = 28), nephrosclerosis (n = 11), polycystic kidney disease (n = 10), and shrunken kidney with unknown etiology (n = 20). Their mean age was 60 ± 15 yr, and average duration on HD was 45 ± 43 mo.

Single-Dose Effect of Intravenous Iron

First, we assessed the changes in 8-OHdG contents in lymphocyte DNA for 4 h after a single dose of iron sucrose administration (Figure 1). Patients in the groups treated with 20, 50, 100, 200, and 500 mg of iron sucrose did not differ significantly from each other in terms of age, gender distribution, percentage of diabetes, HD vintage, basal lymphocyte 8-OHdG, and antioxidant concentrations in whole blood, hemoglobin values, weekly dosage of recombinant erythropoietin, and iron status (Table 1). Iron sucrose (200 mg) treatment resulted in a time-dependent rise of lymphocyte 8-OHdG levels, increasing to 27.1 ± 3.0/10⁶ and 25.8 ± 1.6/10⁶ Deoxyguanosine (dG) compared with the baselines at 2 and 4 h (P < 0.05), respectively (Figure 2A). There was also a dosage-dependent rise of lymphocyte 8-OHdG levels with a significant increase at 2 h after intravenous iron sucrose of 200 and 500 mg, respectively (Figure 2B). Electron microscopic assessment of lymphocytes at 2 h after intravenous iron dose of 200 mg manifested intracellular accumulation of electron-dense deposits within the cytoplasm, indicative of iron uptake (Figure 3). In contrast, there was a marginal effect on lymphocyte 8-OHdG levels at the dosages of 20 to 100 mg (P > 0.05).

Cumulative Dosage Effect of Intravenous Iron

In the randomized, controlled study, 21 patients (10 in intravenous iron group and 11 in control group) had withdrawn by the end of study as a result of cerebral infarct (n = 3), acute myocardial infarction (n = 1), gastrointestinal bleeding (n = 5), pneumonia (n = 3), arteriovenous access thrombosis (n = 4), and transfer to peritoneal dialysis (n = 3) and renal transplantation (n = 2). There was no significant difference in the basal ferritin and transferrin saturation (TSAT) distributions, ROS production, and plasma antioxidant levels between the dropout patients (n = 21) and those who completed the study (n = 89; Figure 1). At the end of study, patients in the intravenous iron (n = 45) and control (n = 44) groups did not differ significantly from each other in the demographic data and baseline biochemical parameters (Tables 2 and 3). Furthermore, both groups of patients were stratified into basal ferritin <500 μg/L and TSAT <20% (n = 8 in control group and n = 8 in intravenous iron group), basal ferritin <500 μg/L and TSAT ≥20% (n = 21 in control group and n = 23 in intravenous iron group), and basal ferritin of 501 to 800 μg/L (n = 15 in control group and n = 14 in intravenous iron group), respectively. Overall, supplementation with iron sucrose for 12 wk significantly increased serum ferritin concentrations from 380 (280 to 502) μg/L to 548 (375 to 683) μg/L and TSAT from 32 ± 10 to 49 ± 15% (P < 0.01); respectively. In contrast, there were no significant changes after treatment with normal saline (Table 3).

Intravenous Iron Aggravated 8-OHdG Levels of Lymphocyte DNA

Mean 8-OHdG contents in cellular DNA of lymphocytes significantly increased after 12 wk of supplementary iron sucrose in total 45 patients (18.6 ± 2.0 to 26.4 ± 2.8 /10⁶ dG; P < 0.05; Table 3). The increase was also noted in subgroups of patients with ferritin of 501 to 800 μg/L (19.1 ± 2.0 to 36.4 ± 4.1 /10⁶ dG; P < 0.01) after supplementation with iron sucrose (Figure 4), but the increase was modest only in intravenous iron patients with basal ferritin of <500 μg/L and TSAT of <20 or ≥20%. In contrast, there were no significant changes in mean 8-OHdG levels in total 44 control patients and in their subgroups as compared with their baseline values.

Intravenous Iron Enhanced Intracellular ROS Production of Lymphocytes

For evaluation of the activated oxygen metabolism of lymphocytes, the intracellular production of ROS in lymphocytes was analyzed by flow cytometry before and after 12 wk of supplementation with iron sucrose or normal saline. The 2′,7′-dichlorofluorescein (DCF) fluorescence intensity was determined for 30 min because the fluorescence increased almost linearly up to 30 min. Within 30 min, the spontaneous ROS production significantly increased above the initial level in unstimulated lymphocytes from patients with basal ferritin of 501 to 800 μg/L after 12 wk of supplementary iron sucrose (50 ± 21 to 90 ± 26%; P < 0.05; Figure 5A), but there was no significant difference after treatment with normal saline as compared with the baseline values (Figure 5A). Likewise, the phorbol-12-myristate-13-acetate (PMA)-stimulated production of ROS by lymphocytes significantly increased in intravenous iron patients with basal ferritin of 501 to 800 μg/L (221 ± 74 to 325 ± 84%; P < 0.05; Figure 5B). The increases in ROS production by lymphocytes after iron sucrose were marginal in patients with basal ferritin of <500 μg/L and TSAT of <20 or ≥20%. No significant change was noted in the PMA-stimulated ROS production by lymphocytes after treatment with normal saline (Figure 5B).

Intravenous Iron Reduced Antioxidant Capacity

There was a significant decrease in mean concentrations of plasma ascorbate or α-tocopherol adjusted for serum cholesterol and a rise in oxidized glutathione/reduced glutathione (GSSG/GSH) ratio after 12 wk of supplementary iron sucrose in total 45 patients (P < 0.05; Table 3). The decrease in plasma ascorbate or α-tocopherol adjusted for cholesterol and the increase in GSSG/GSH ratio were also noted in intravenous iron patients with basal ferritin of 501 to 800 μg/L (P < 0.05), but the changes were moderate in intravenous iron patients with basal ferritin of <500 μg/L and TSAT of <20 or ≥20% (Table 4). In contrast, there were no significant changes in mean plasma ascorbate or α-tocopherol adjusted for cholesterol and GSSG/GSH ratios in total 44 control patients (Table 3) and in their subgroups as compared with their baseline values (Table 4).

DISCUSSION

Iron is a cellular transition element, and its ionic forms are prone to participate in one-electron transfer reactions; however, the capacity also makes iron to generate free radicals. In the presence of iron metal, the reactive hydroxyl radicals can be formed via Fenton reaction and iron-catalyzed Haber-Weiss reaction.^23,24 The generated ROS subsequently initiate damage to cellular constituents. In vitro study by Zager et al.²⁵ showed that different iron preparations could trigger lipid peroxidation, mitochondrial toxicity, and even cytotoxicity in mouse proximal tubular segments and cultured human proximal tubular cells. Agarwal et al.²⁶ displayed the coherent findings that intravenous administration of 100 mg of iron sucrose over 5 min could transiently increase plasma MDA levels accompanied by enzymuria and proteinuria in patients with chronic kidney disease (CKD). A number of clinical studies further demonstrated that administration of intravenous iron to HD patients increased oxidative byproducts such as AOPP,²⁷ MDA,^17,28 carbonylated fibrinogen,¹⁹ and esterified F₂-isoprostanes,²⁹ as well as reduced plasma antioxidants such as superoxide dismutase and glutathione peroxidase,^17,28 but studies as to whether intravenous iron therapy could aggravate the oxidative damage to cellular DNA are rare.

DNA is particularly susceptible to oxidative injury in HD patients because of the increased ROS production.^7,8 8-OHdG is one of the most abundant oxidative DNA products among more than 20 different base modifications elicited by ROS.^30,31 Our previous work identified 8-OHdG contents in lymphocyte DNA as a reliable, steady-state marker for oxidative DNA damage in long-term HD patients.³²In vitro study by Takeuchi et al.³³ disclosed that cellular iron metabolism and hydroxyl radical production in human neutrophil-like cells were intimately associated with 8-OHdG levels. In healthy individuals, Nakano et al.³⁴ measured urinary excretion of 8-OHdG in 2507 Japanese and found that urine 8-OHdG levels positively correlated with serum ferritin levels. They assumed that body iron status might affect the in vivo generation of 8-OHdG. In patients with chronic hepatitis C, Kato et al.³⁵ demonstrated a close link between hepatic 8-OHdG and hepatic iron overload. They further revealed that reduction in iron stores with phlebotomy and low-iron diet resulted in greater repair of oxidative DNA damage in these patients. In our previous studies,^7,9 8-OHdG content in leukocyte DNA was positively correlated with serum ferritin and TSAT levels among patients receiving maintenance HD. Yoshimura et al.³⁶ also showed serum ferritin as a significant and independent determinant of serum 8-OHdG levels in patients on regular HD. These findings support the notion that the higher body iron stores, the more increased oxidative DNA damage; however, the association should not be interpreted as a causal relationship between intravenously administered iron and increased 8-OHdG levels.

On the contrary, our prospective controlled study demonstrated a time- and dosage-dependent rise in lymphocyte 8-OHdG levels with a significant increase at 2 h after a single intravenous iron administration of ≥200 mg. The most compelling finding is that repeated intravenous iron therapy significantly increased oxidative damage to lymphocyte DNA in long-term HD patients. 8-OHdG levels were shown to be the highest in long-term HD patients, followed by undialyzed patients with CKD and healthy individuals in our previous study.⁷ Taken collectively, DNA oxidation is a consequence of HD procedure and amplified by intravenous iron supplementation. Our study further corroborates the previous findings^{17,19,27–29} that intravenous iron therapy exhibits the pro-oxidant property in HD patients and leads to ROS overproduction and impairment in antioxidant defense. In this study, repeated intravenously administered iron not only increased spontaneous and PMA-induced ROS production in lymphocytes (Figure 4) but also reduced plasma levels of ascorbate and α-tocopherol adjusted for cholesterol (Table 3). We infer that oxidative DNA damage by intravenous administered iron may be further aggravated as a result of impaired oxidant scavenging system of plasma in HD patients as substantiated by decreased plasma levels of vitamins C and E and increased GSSG/GSH ratios.

The use of 8-OHdG as a dosimeter for oxidative stress might be debatable in light of overestimation of the 8-OHdG levels as a result of spurious oxidation of guanine during sample processing and storage, especially in the presence of substantial iron levels in plasma after intravenous iron therapy. In this study, rigorous procedures (the addition of antioxidants [2,2,6,6-tetramethylpiperidine-N-oxyl] during isolation and hydrolysis of DNA); avoiding the use of phenol and storing samples under nitrogen; and a randomized, controlled study design were used to minimize and control for sources of experimental error.^5,6 Smoking and systemic inflammatory disease can induce formation of 8-OHdG in peripheral leukocyte DNA.^37,38 angiotensin-converting enzyme inhibitors or angiotensin receptor blockers have also been shown to reduce DNA damage of lymphocytes in experimental chronic renal failure.³⁹ Therefore, to reduce oxidation as much as possible to identify the differences in lymphocyte 8-OHdG levels mainly as a result of intravenous iron therapy, it was of the utmost importance to exclude these confounders in this study.

A recent small, nonrandomized study also showed that 8-OHdG in serum levels significantly increased by repeated intravenous administration of 40 mg of ferric saccharate to seven patients after each HD session until hematocrit increased by 5%.⁴⁰ Furthermore, several lines of evidence indicate that 8-OHdG levels significantly increased in the DNA of animal kidney or liver sample after administration of the various iron preparations.^41,42 Those findings are complementary to our findings. Our study together with the study of Maruyama et al. suggest that these results are in fact due to iron and not just the iron sucrose.

The revised KDOQI guidelines state that there is insufficient evidence for routine intravenous iron administration when serum ferritin is >500 μg/L²⁰; however, the upper limit of serum ferritin at which intravenous iron supplementation should be withheld remains a matter of debate.⁴³ Lim et al.¹⁷ reported that after the single intravenous iron infusion, a significant decrease of plasma superoxide dismutase activity and a significant rise in plasma level of MDA occurred in HD patients with baseline ferritin >600 μg/L. To identify the safe upper levels of serum ferritin and TSAT for maintenance intravenous iron therapy in HD patients, we stratified our patients into three subgroups using combined measures of basal ferritin and TSAT. The extent of increased lymphocyte 8-OHdG level and intracellular ROS production or of decreased plasma vitamins C and E levels and increased GSSG/GSH ratios were only modest in the patients with basal ferritin of <500 μg/L. In contrast, in the patients with basal ferritin of >500 μg/L, these changes became statistically significant. Kalantar-Zadeh et al.⁴⁴ analyzed data from a cohort of 58,058 HD patients and found, after appropriate adjustment, that patients with serum ferritin between 200 and 1200 μg/L and TSAT between 30 and 50% were at the lowest risk for cardiovascular and all-cause mortality, compared with those for whom same iron parameters were outside these range. They inferred that the association of serum ferritin levels >500 μg/L with higher mortality rate in the unadjusted model was mainly due to inflammation and malnutrition. Nevertheless, a direct cardiovascular and toxicity of iron administration has not been completely excluded, because patients receiving >400 mg of intravenous iron monthly tended to have increased death rates as compared with those not receiving intravenous iron at all. There is mounting evidence for disordered oxidative and glycoxidative chemistry in uremia that may contribute to poor cardiovascular and global outcome. It is important, therefore, to understand whether prescriptive decisions in HD therapy, such as choice of membrane, modality, or intravenous iron therapy, affect oxidant stress. Our convincing results raise the issue of developing modified criteria for iron supplementation and support the opinions of KDOQI guidelines that decisions regarding intravenous iron administration should be weighed by ESA responsiveness, hemoglobin and TSAT levels, and the patient's clinical status when ferritin level is >500 μg/L.¹⁹

In summary, our randomized, controlled study demonstrated that repeated intravenous iron supplementation significantly aggravated the generation of 8-OHdG in peripheral blood lymphocytes of HD patients, especially in those with serum ferritin >500 μg/L. In addition, intravenously administered iron exhibits a pro-oxidant in vivo to enhance intracellular ROS production and impair antioxidant defenses in long-term HD patients. The long-term risks for intravenous iron–induced oxidative DNA damage remained to be determined, because our study was too short to assess long-term complications. From a pathophysiologic point of view, when considering the enhanced oxidant stress caused by intravenous iron, long-term, adequately sized prospective studies evaluating the safety of intravenous administered iron are needed to define better the appropriate clinical role, if any, of this therapy.

CONCISE METHODS

Patients and Study Protocol

To assess whether intravenous iron supplementation has an in vivo pro-oxidant effect on cellular DNA of peripheral blood lymphocytes, we screened a total of 232 patients undergoing long-term HD from the two dialysis centers of our affiliated hospital. Inclusion criteria were age between 20 and 80 yr and HD vintage >6 mo. Exclusion criteria were weekly dialysis time <12 h; urea Kt/V <1.2; serum ferritin >800 μg/L; habit of tobacco smoking; disorders such as malignancy, chronic inflammatory diseases, and acute infections; and treatment with iron supplements, angiotensin-converting enzyme inhibitors, or anti-inflammatory drugs 3 mo before enrollment. Dialytic procedure in the two dialysis centers was a standard bicarbonate session. Dialysis machines were sterilized daily, and water treatment circuits and tanks were sterilized weekly. The colony count of microorganisms in water used to prepare dialysis fluid did not exceed 200 colonies/ml. Endotoxin levels in dialysate, detected weekly by amoebocyte lysate test (LAL; Chromogenix, Charleston, SC), were <0.01 EU/ml.

Initially, all patients were randomly allocated to five groups. Patients in each group were supplied with a single dose of 20, 50, 100, 200, and 500 mg of iron sucrose (Nang Kuang Pharmaceutical Company, Tainan, Taiwan), respectively. Iron supply was administered intravenously for 60 min in the morning on a nondialysis day. Blood samples were taken for lymphocyte 8-OHdG measurements on four occasions: before, then 1, 2, and 4 h after supplementation in each group.

Four weeks after the single-dose test, an open-label, randomized, controlled study was carried out for 12 wk. In the 12-wk period, all patients were randomly assigned to receive supplementation with iron sucrose or normal saline. Iron sucrose (100 mg of elemental iron diluted in 250 ml of 0.9% saline) or 250 ml of 0.9% saline (control) was administered intravenously for 60 min after dialysis once a week for 12 wk. Blood samples were taken on two occasions: Before supplementation started and 1 wk after administration of the last iron dose at the 12th week. All patients fasted for 12 h immediately before the blood samples were taken. Any ongoing medications were continued without changes in dosage during the study. The protocol was approved by the Committee on Human Research at Taipei Veterans General Hospital. Informed consent was obtained from each of the study patients.

Laboratory Measurements

From each study patient, 14 ml of whole blood was collected by venipuncture before dialysis and divided into aliquots for the subsequent analyses to be performed. Two milliliters of blood was withdrawn into a heparinized Vacutainer tube for flow cytometric analysis of intracellular ROS production by lymphocytes. Of 12 ml of blood withdrawn into an EDTA-containing Vacutainer tube, 2 ml was immediately removed to determine the concentrations of reduced (GSH) and oxidized (GSSG) glutathiones. Lymphocytes were separated from 10 ml of whole blood on Ficoll-Hypaque density gradients, washed twice in PBS (pH 7.4), and centrifuged again. The cells recovered were confirmed to be 95% lymphocytes and frozen at −70°C until use for determination of 8-OHdG content in the cellular DNA.

Iron, cholesterol, and triglyceride in serum were determined using commercial kits by an autoanalyzer (Hitachi 736-60, Naka, Japan). Total iron binding capacity (TIBC) was measured by the TIBC Microtest (Daiichi, Tokyo, Japan) and serum ferritin by RIA (Incstar, Stillwater, MN). Percentage of TSAT was calculated by dividing serum iron concentration by TIBC × 100. Whole blood (0.5 ml) was immediately deproteinized to determine the GSH levels with an equal volume of 20% TCA after sampling. GSH was quantified as described by Beutler et al.⁴⁵ For derivatization of GSH, 0.5 ml of whole blood was treated with an equal volume of 12% perchloric acid containing 40 mmol/L N-ethylmaleimide and 2 mmol/L bathophenanthroline disulfonic acid. The derivatized glutathione was measured by an HPLC system similar to that developed by Asensi et al.⁴⁶ Plasma ascorbate was measured by a method described by Kyaw.⁴⁷ The concentration of α-tocopherol in plasma was determined using the procedure of Catignani and Bieri⁴⁸ with some modifications as described previously.^7–9 The concentration of α-tocopherol was calculated from a calibration curve constructed with internal standards. The value of α-tocopherol was adjusted for serum cholesterol. All assays were carried out on duplicate samples.

Measurement of 8-OHdG Content

Total DNA of lymphocytes was extracted by the pronase/ethanol method⁴⁹ with some modifications as described previously.^7–9,32 To block iron-catalyzed formation of 8-OHdG during sample processing, we added nitroxide 2,2,6,6-tetramethylpiperidine-N-oxyl in all steps of analysis of 8-OHdG.⁵⁰ The amount of 8-OHdG was measured by the method using HPLC equipped with an EC detector (Bioanalytical Systems, West Lafayette, IN) as described previously.^7–9,51 dG (Sigma Chemical Co., St. Louis, MO) and 8-OHdG (Cayman, Ann Arbor, MI) were used as standards. The 8-OHdG level is expressed as the number of 8-OHdG molecules per 10⁶ dG. Intra-assay coefficients of variation (CV) ranged from 3 to 7%, and interassay CV ranged from 4 to 9%; the lower numbers refer to the CV for the high standard, and the higher numbers refer to the CV for the low standard.

Flow Cytometric Analysis of Intracellular ROS Production

Two 100-μl aliquots of each sample were analyzed for ROS production, one at baseline and the other after activation with PMA. Leukocytes were harvested from blood samples after lysis of red blood cells by a lysis solution (0.15 M NH₄Cl, 10 mM NaHCO₃, and 10 mM EDTA [pH 7.4]). After centrifugation for 10 min at 350 × g and 4°C, leukocyte pellet was suspended in 1.5 ml of Hanks’ balanced salt solution (pH 7.3). Cell viability was determined by trypan blue exclusion. Leukocytes were stained with 2′,7′-dichlorofluorescin diacetate (DCF-DA; Molecular Probes, Eugene, OR). After labeling, leukocytes were incubated for 30 min at 37°C in the presence or absence of 100 ng/ml PMA (Sigma Chemical Co.) as described previously.³² Total leukocytes were subjected to flow cytometry analysis (FACSort; Becton-Dickinson, San Jose, CA) for measurement of intracellular production of ROS (O₂^−• and H₂O₂).⁵² The lymphocyte population was determined by gating on a forward scatter and side scatter dot plot as described previously.¹ Intracellular ROS production was expressed as mean fluorescence of DCF. ROS production was then monitored every 10 min on FACSort by measurement of the intensity of fluorescence emitted at 525 nm for DCF. Data for each sample were calculated by the CellQuest software (Becton-Dickinson) on a power Macintosh 6100/66 computer.

Statistical Analysis

Statistical analysis was performed using the computer software SSPS 12.0 (SPSS, Chicago, IL). Data are expressed as means ± SD. Values for serum ferritin were not normally distributed and are reported as medians with interquartile ranges. Data for more than two groups were analyzed using one-way ANOVA or Kruskal-Wallis test, as appropriate. Pearson χ² test was used for frequency measures. Wilcoxon singed ranks test was used for comparison of baselines and data after treatment with iron sucrose or normal saline. P < 0.05 is considered statistically significant.

DISCLOSURES

None.

Figure 1: Number of patients who entered the study, were assigned to five groups receiving a single dose of intravenously administered iron, 4 wk later randomly assigned to intravenous iron or control group for cumulative dosage effect, and completed the study.

Figure 2: Time- and dosage-dependent effects of intravenous iron sucrose on 8-OHdG content of peripheral blood lymphocytes in hemodialysis patients. (A and B) Four hours after intravenous administration of 200 mg of iron sucrose (A) and before (□) and 2 h (▪) in each group (n = 22) after a single dose of 20, 50, 100, 200, and 500 mg, respectively (B). *P < 0.05 versus 0 h by Wilcoxon signed ranks test.

Figure 3: (A through C) Electron micrographs of human lymphocytes before (A) and 2 h after intravenous administration of 200 mg of iron sucrose (B and C) to HD patients. (B) Lymphocytes treated with iron had distinct intracellular iron uptake (arrows). (C) In some cells, the iron deposits took on a tangled filamentous appearance (arrows).

Figure 4: 8-OHdG contents in cellular DNA of peripheral blood lymphocytes of HD patients in the randomized, controlled study. Patients in both groups were stratified into basal ferritin of <500 μg/L and TSAT of <20% (n = 8 in control group and n = 8 in intravenous iron group), basal ferritin of <500 μg/L and TSAT of ≥20% (n = 21 in control group and n = 23 in intravenous iron group), and basal ferritin of 501 to 800 μg/L (n = 15 in control group and n = 14 in intravenous iron group), respectively. **P < 0.01 versus 0 wk by Wilcoxon signed ranks test.

Figure 5: (A and B) Spontaneous (A) and PMA-induced (B) production of ROS in lymphocytes of HD patients in the randomized, controlled study. The percentages of increase in fluorescence above the resting level within 30 min are shown for all HD patients with basal ferritin of <500 μg/L and TSAT of <20% (n = 8 in control group and n = 8 in intravenous iron group), basal ferritin of <500 μg/L and TSAT of ≥20% (n = 21 in control group and n = 23 in intravenous iron group), and basal ferritin of 501 to 800 μg/L (n = 15 in control group and n = 14 in intravenous iron group), respectively. *P < 0.05 versus 0 wk by Wilcoxon signed ranks test.

Table 1: Baseline characteristics of long-term HD patients recruited in this studya

Table 2: Baseline characteristics of long-term HD patients in the two groups at the end of the randomized, controlled study

Table 3: Lymphocyte 8-OHdG levels, iron metabolism indices, and antioxidant concentrations before and 12 wk after intravenous iron or normal saline therapy in the randomized, controlled study

Table 4: Plasma ascorbate, α-tocopherol, and WB glutathione concentrations before and 12 wk after intravenous iron therapy in the randomized study

This study was supported by grants from the National Science Council (NSC 95-2314-B-010-077) and Taipei Veterans General Hospital (V97C1-093).

Part of this work was presented at the annual meeting of the American Society of Nephrology; October 31 through November 5, 2007; San Francisco, CA.

We are extremely grateful to P.C. Lee and Y.W. Chia for expert secretarial assistance and graphic design.