Kidney International, Vol. 65 (2004), pp. 2279–2289
`
`CLINICAL NEPHROLOGY – EPIDEMIOLOGY – CLINICAL TRIALS
`
`Oxidative stress and renal injury with intravenous iron in
`patients with chronic kidney disease
`
`RAJIV AGARWAL, NINA VASAVADA, NADINE G. SACHS, and SHAWN CHASE
`
`Division of Nephrology, Department of Medicine, Indiana University School of Medicine, Indianapolis, Indiana;
`and Richard L. Roudebush VA Medical Center, Indianapolis, Indiana
`
`Oxidative stress and renal injury with intravenous iron in pa-
`tients with chronic kidney disease.
`Background. Intravenous iron is widely prescribed in patients
`with chronic kidney disease (CKD) and can cause oxidative
`stress. The relationship of oxidative stress and renal injury in
`patients with CKD is unknown. Whether renal injury can occur
`at a time point when transferrin is incompletely saturated is also
`unclear.
`Methods. We conducted a randomized, open-label, parallel
`group trial to compare the oxidative stress induced by intra-
`venous administration of 100 mg iron sucrose over 5 minutes
`and its protection with N-acetylcysteine (NAC) in 20 sub-
`jects with stage 3 or 4 CKD. Transferrin saturation was
`measured with urea polyacrylamide gel electrophoresis, ox-
`idative stress by malondialdehyde (MDA) measurement by
`high-performance liquid chromatography, and renal
`injury
`by enzymuria and proteinuria. Reduced and oxidized glu-
`tathione and free radical scavengers as well as urinary monocyte
`chemoattractant protein-1 were also measured.
`Results. Parenteral iron increased plasma concentration and
`urinary excretion rate of MDA, a biomarker of lipid peroxida-
`tion, within 15 to 30 minutes of iron sucrose administration. This
`was accompanied by enzymuria and increase in proteinuria. In
`contrast, saturation of transferrin was not maximally seen until
`3 hours after the end of infusion. Oxidative stress, enzymuria
`and proteinuria were transient and were completely resolved in
`24 hours. NAC reduced acute generation of systemic oxidative
`stress but failed to abrogate proteinuria or enzymuria.
`Conclusion. Intravenous iron produces oxidative stress that
`is associated with transient proteinuria and tubular damage.
`The rapid production of oxidative stress even when transfer-
`rin is not completely saturation suggests free iron independent
`mechanism(s) to be operative in producing oxidative stress and
`transient renal injury. Long-term implications of these findings
`need further study.
`
`In the United States there are approximately 8 mil-
`lion individuals with moderate to severe chronic kid-
`
`Key words: Iron, anemia, malondialdehyde, oxidative stress, chronic
`kidney failure, randomized controlled trial.
`
`Received for publication October 25, 2003
`and in revised form November 22, 2003
`Accepted for publication December 22, 2003
`
`C(cid:1) 2004 by the International Society of Nephrology
`
`ney disease (CKD), not on dialysis [1]. Anemia sets in
`early during the course of renal disease, either as a re-
`sult of erythropoietin deficiency or due to insufficient ab-
`sorption of dietary iron [2]. Furthermore, a functional
`deficiency of iron may occur due to enhanced erythro-
`poiesis after therapy with recombinant human erythro-
`poietin that often necessitates therapy with intravenous
`iron [3]. The correction of anemia leads to improvement
`in morbidity, mortality, and quality of life [4]; however,
`potential toxicity due to the participation of elemen-
`tal iron in oxidation-reduction reactions has raised con-
`cern of potential risks of this commonly utilized therapy
`[5, 6].
`Oxidative stress plays an important role in the patho-
`genesis and progression of renal disease [7, 8]. Intra-
`venous iron has been shown to increase biologic markers
`of oxidative stress [9] in cell cultures [10], animal mod-
`els [11], and end-stage renal disease (ESRD) patients on
`hemodialysis [12–14]. Intravenous administration of iron
`sucrose in dialysis patients results in an increase in total
`peroxide, free iron, and markers of lipid peroxidation,
`and is significantly improved with the antioxidant, vita-
`min E [15]. Whether patients with CKD not on dialy-
`sis have a similar increase in oxidative stress and, above
`all, renal injury upon exposure to intravenous iron is un-
`known. This knowledge is of critical importance because
`oxidative stress and renal injury may lead to an acceler-
`ated course of renal [7] and cardiovascular disease [16,
`17]. Furthermore, it is unknown whether the acute ox-
`idative stress imposed by iron can be reduced with an-
`tioxidant therapy. In this context, animal and human data
`demonstrate beneficial effects of N-acetylcysteine (NAC)
`in models of oxidative stress such as contrast-induced
`acute renal failure [18].
`We hypothesized that in subjects with moderate to se-
`vere CKD, an infusion of intravenous iron would gen-
`erate oxidative stress and that this would cause tubular
`injury and increase in glomerular permeability. Further-
`more, the generation of oxidative stress, tubular injury,
`glomerular permeability, and renal inflammation would
`be mitigated by therapy with the antioxidant NAC.
`
`2279
`
`Luitpold Pharmaceuticals, Inc., Ex. 2011, p. 1
`Pharmacosmos A/S v. Luitpold Pharmaceuticals, Inc., IPR2015-01495
`
`

`
`2280
`
`Agarwal et al: Intravenous iron and oxidative stress
`
`METHODS
`Subjects
`Study participants were recruited from the Nephrol-
`ogy Clinic at the Richard L. Roudebush VA Medical
`Center. Subjects were at least 18 years in age, had an
`estimated glomerular filtration rate (GFR) ≤ 60 mL/min
`by the six-component Modified Diet in Renal Disease
`(MDRD) Study formula [19] but were not on renal
`replacement therapy and had iron deficiency anemia.
`Anemia was defined as blood hemoglobin concentration
`below the 95th percentile in the NHANES evaluation
`[20]. Iron deficiency was defined using National Kidney
`Foundation/Kidney Disease Outcome Quality Initiative
`(NKF-K/DOQI) guidelines as having a serum ferritin
`concentration of <100 ng/mL or serum transferrin satu-
`ration of <20% [21]. Subjects were excluded if they were
`hypersensitive to iron sucrose or NAC, were an organ
`transplant recipient or received therapy with immuno-
`suppressive agents, used an investigational drug 1 month
`prior to study, or demonstrated the presence of active
`asthma, human immunodeficiency virus (HIV), cancer,
`rheumatoid arthritis, alcoholism, or liver disease. Sub-
`jects were also excluded if they were anemic such that ery-
`throcyte transfusion was imminent (hemoglobin <8 g/dL,
`or active gastrointestinal bleeding), had acute renal fail-
`ure, defined as an increase in the baseline serum crea-
`tinine concentration of ≥0.5 mg/dL over 48 hours, were
`pregnant or breastfeeding, had received intravenous iron
`within 3 months of the study, demonstrated iron over-
`load (serum ferritin >800 ng/mL or transferrin satura-
`tion >50%), had anemia not caused by iron deficiency,
`had surgery or systemic or urinary tract infection within
`1 month of study, or a serum albumin <3.0 g/dL.
`The protocol was approved by the Institutional Review
`Board at Indiana University School of Medicine and the
`Research and Development Committee of the Richard
`L. Roudebush VA Medical Center. Written informed
`consent was obtained from each subject prior to study
`participation.
`
`Study design
`Subjects who satisfied the inclusion and exclusion crite-
`ria were assigned by a computer-generated randomized
`block randomization scheme to one of two open-label,
`1-week, parallel treatment arms: either no active in-
`tervention, or treatment with the antioxidant NAC
`(Mucomyst) (Apothecon, Inc., Princeton, NJ, USA) at
`a dose of 600 mg twice a day. The allocation scheme was
`concealed from investigators until time of subject ran-
`domization. Oral iron supplements were stopped for the
`duration of the study.
`The trial design is illustrated in Figure 1. Subjects were
`requested to be fasting before all visits. During the ini-
`tial study period, baseline blood and urine was collected
`
`and assayed for biomarkers of lipid peroxidation [plasma
`and urine malondialdehyde (MDA)], substrate for lipid
`peroxidation (lipid panel), evaluation of iron deficiency
`anemia [complete blood count, iron, total iron bind-
`ing capacity (TIBC), ferritin], redox state [plasma and
`erythrocyte oxidized and reduced glutathione (GSSG,
`GSH)], free radical scavengers [erythrocyte superox-
`ide dismutase (red blood cell SOD) and glutathione
`peroxidase (GSHPx)], and markers of renal tubular
`and glomerular damage and/or inflammation [urinary
`monocyte chemoattractant protein-1 (MCP-1), urine to-
`tal protein, and N-acetyl-b-D-glucosaminidase (NAG)].
`Subjects then received an intravenous dose of iron su-
`crose (Venofer) (American Regent Laboratories, Inc.,
`Shirley, NY, USA) at a dose of 100 mg infused over
`5 minutes. Blood and urine collections were obtained at
`baseline, 0.25, 0.5, 1, 2, 3, and 24 hours after iron dosing to
`assay for the above parameters. During the visits, the sub-
`jects completed the Kidney Disease and Quality of Life
`Short Form (KDQOL-SF, version 1.3). Subjects were ran-
`domized to take either no additional treatment or NAC
`600 mg twice a day for 1 week. The last dose of NAC was
`administered under direct supervision by the study nurse
`in the morning of the second infusion of intravenous iron.
`After 1 week, subjects had all study procedures repeated
`as on the initial study period. NAC-assigned subjects re-
`turned with remaining antioxidant doses to assess com-
`pliance. Subjects completed a KDQOL-SF modified to
`assess parameters over the 1-week intervening treatment
`period.
`
`Laboratory analysis
`Electrophoretic separation of transferrins. The iron
`forms of transferrin were separated using a urea poly-
`acrylamide gel (6% acrylamide gels with 6 mol/L urea)
`according to Williams, Evans, and Moreton [22]. This
`method separates transferrin into the apotransferrin,
`monoferric, and the diferric forms, according to their elec-
`trophoretic mobilities. Precipitation of transferrin from
`plasma was achieved through treatment with 2% 6,9-
`diamino-2-ethoxyacridine lactate monohydrate (Sigma
`Chemical Co., St. Louis, MO, USA) and subsequent
`centrifugation before being applied to the gel to pre-
`cipitate all serum proteins but b and c globulins [23].
`Electrophoresis was performed using a Criterion mini-gel
`system (Bio-Rad, Hercules, CA, USA). Fourteen plasma
`samples obtained from each subject over two visits were
`separated simultaneously on the same gel, along with con-
`trols of apotransferrin, monoferric transferrin(s), and di-
`ferric transferrin. Protein bands were visualized through
`staining with GelCode Blue stain reagent (Pierce,
`Rockford, IL, USA). Densitometric analysis was per-
`formed with a Gel Logic 100 apparatus and one-
`dimensional image analysis software (Kodak, Rochester,
`
`Luitpold Pharmaceuticals, Inc., Ex. 2011, p. 2
`Pharmacosmos A/S v. Luitpold Pharmaceuticals, Inc., IPR2015-01495
`
`

`
`Agarwal et al: Intravenous iron and oxidative stress
`
`2281
`
`Urine
`
`Single dose iron sucrose
`(100 mg/5 min IV push)
`
`Placebo group
`
`R
`
`Single dose iron sucrose
`(100 mg/5 min IV push)
`
`N-acetyl cysteine group
`(600 mg BID × 7d)
`
`Urine
`
`Fig. 1. Trial design. All patients with chronic kidney disease (CKD) had baseline urine and blood draws followed by administration of iron sucrose.
`Blood and urine specimens were collected at 0.25, 0.5, 1, 2, 3, and 24 hours after administration of the drug. Patients were then randomized to receive
`N-acetylcysteine (NAC) or nothing for 1 week in an open-label fashion. After 1 week, iron sucrose was again administered and tests performed as
`done at first visit.
`
`Urine
`
`NY, USA). Percent transferrin saturation was calculated
`as follows:
`Transferrin Saturation(%) =
`[Diferric transferrin+ (1/2 × monoferric transferrin(s))]
`[Apotransferrin + Monoferric transferrin(s)+ Diferric transferrin]
`Reduced and oxidized glutathione. Determination of
`reduced and oxidized glutathione concentrations was
`achieved through an isocratic high-performance liquid
`chromatography (HPLC) technique. Specimens were
`collected, processed, and analyzed as described by
`Paroni et al [24]. We modified the method by increas-
`ing the flow rate and changing the wash pattern of the
`column as described below.
`Briefly, the mobile phase consisted of 21 mmol/L pro-
`prionate buffer and acetonitrile, 95:5 (vol/vol) pumped
`at 2 mL/min. The column was a Merck LiChrospher
`RP-18, 5 l with guard placed in a column warmer
`◦
`set to 37
`C. Following three consecutive samples, the
`column was washed for 3 minutes with a solution of
`acetonitrile:methanol:water (55:25:20 vol/vol) followed
`by re-equilibration with mobile phase for 4 minutes.
`
`The chromatographic system consisted of a Hewlett-
`Packard Chromatographic Series 1100 autosampler and
`isocratic pump and Hewlett-Packard model 1046A pro-
`grammable fluorescence detector (Palo Alto, CA, USA).
`The autosampler was programmed to perform precolumn
`derivatization with o-phthalaldehyde with peak detection
`by fluorescence with an excitation wavelength of 340 nm
`and an emission wavelength 420 nm. Peak areas were
`determined using a ChromJet integrator (Thermo Sepa-
`ration Products, San Jose, CA, USA).
`Plasma and urinary MDA assay. MDA, a lipid hy-
`droperoxide,
`is formed by b-scission of peroxidized
`polyunsaturated fatty acids was measured by derivatiza-
`tion with thiobarbituric acid (TBA) as reported previ-
`ously [25]. Briefly, the mobile phase consisted of 40:60
`ratio (vol/vol) of methanol to 50 mmol/L potassium
`monobasic phosphate at pH 6.8, pumped at a rate of
`1.0 mL/min on a Hewlett-Packard Hypersil 5 l ODS
`100 × 4.6 mm placed in a column warmer set to 37
`◦
`C.
`Samples of plasma and urine were treated with the an-
`tioxidant, butylated hydroxytoluene (to prevent in vitro
`◦
`oxidation) and heat derivatized at 100
`C for 1 hour with
`
`Luitpold Pharmaceuticals, Inc., Ex. 2011, p. 3
`Pharmacosmos A/S v. Luitpold Pharmaceuticals, Inc., IPR2015-01495
`
`

`
`2282
`
`Agarwal et al: Intravenous iron and oxidative stress
`
`Table 1. Baseline characteristics of the study population
`Iron sucrose +
`Iron
`sucrose + N-acetylcysteine
`placebo
`(NAC)
`
`Overall
`
`TBA at an acid pH. Samples were extracted with n-
`butanol and 10 lL of the extract was injected at 1-minute
`intervals using an autosampler. The Hewlett-Packard
`model 1046A programmable fluorescence detector was
`set at excitation of 515 nm and emission of 553 nm. Re-
`tention time was 1.87 minutes; however, absence of in-
`terfering peaks, allows analysis to be carried out in incre-
`ments of 1 minute per sample. Within-day variability in
`estimation was between 8.6% and 10.3%. Between-days
`variability was 3.6% and 7.9%. Recovery was between
`88% and 101%.
`Urinary MCP-1 assay. MCP-1 was assayed in urine
`using a sandwich ELISA (Quantikine(cid:1) kit for Human
`MCP-1 Immunoassay) (R&D Systems, Minneapolis, MN,
`USA). A standard curve was generated using a four pa-
`rameter logistic curve-fit. Corrections were made for con-
`centration and values were expressed as pg MCP-1 per
`mg creatinine.
`SOD and GSHPx activity. SOD activity was measured
`in red blood cell lysate using an end point spectropho-
`tometric technique (Superoxide Dismutase Assay)
`(Cayman Chemical Co., Ann Arbor, MI, USA). GSHPx
`activity in red blood cell lysate was determined using an
`indirectly coupled reaction with glutathione reductase in
`a kinetic spectrophotometric assay (Glutathione Peroxi-
`dase Assay) Cayman Chemical Co.).
`Urine protein, creatinine, and NAG determination.
`Concentration of urinary protein was quantified through
`an end point spectrophotometric assay utilizing pyro-
`gallol red-molybdate (quanTtest red) (Quantimetrix,
`Redondo Beach, CA, USA). Urine creatinine concen-
`tration was determined using an end point spectrophoto-
`metric with an alkaline-picrate solution (Creatinine Kit)
`(Sigma Diagnostics, St. Louis, MO, USA). Urinary NAG
`was measured by colorimetric assay (Roche Diagnostics
`Corporation, Indianapolis, IN, USA). Other laboratory
`assays were performed using standard methods in the
`hospital laboratory.
`
`10
`71 ± 5.9
`10
`8/2
`93 ± 19.2
`30 ± 12.5
`
`2
`9
`6
`
`Number of patients
`Age years
`Male number
`Whites/Blacks number
`Weight kg
`Body mass index kg·m2
`Tobacco use number
`Current
`Former
`Coronary artery
`disease number
`Etiology of kidney disease number
`Hypertension
`3
`Diabetes mellitus
`6
`Other
`0
`Angiotensin-convering
`3
`enzyme (ACE) inhibitor
`therapy number
`Systolic blood pressure
`mm Hg
`Diastolic blood pressure
`mm Hg
`Heart rate beats/min
`Modification of Diet in
`Renal Disease Study
`glomerular filtration
`rate mL/min
`Angiotensin receptor
`blocker therapy number
`Erythropoietin therapy
`number
`Hemoglobin g/dL
`Iron lg/dL
`Total iron binding
`capacity lg/dL
`Ferritin ng/mL
`Serum albumin g/dL
`Total cholesterol mg/dL
`Low-density lipoprotein
`cholesterol mg/dL
`High-density lipoprotein
`cholesterol mg/dL
`Triglycerides mg/dL
`
`138 ± 29.0
`69 ± 10.5
`70 ± 10.1
`25 ± 7.1
`
`5
`
`4
`11.3 ± 1.1
`60.3 ± 28.8
`351 ± 92
`85 ± 75
`3.8 ± 0.4
`151 ± 53.5
`80 ± 37.1
`38 ± 11.8
`178 ± 112.4
`
`10
`75 ± 7.6
`10
`7/3
`97 ± 23.8
`32 ± 7.6
`
`20
`73 ± 7.0
`20
`15/5
`96 ± 21.3
`31 ± 10.1
`
`2
`8
`8
`
`3
`6
`1
`7
`
`4
`17
`14
`
`6
`12
`1
`10
`
`148 ± 23.4
`66 ± 11.7
`64 ± 10.7
`27 ± 7.4
`
`143 ± 26.1
`67 ± 11.0
`67 ± 10.6
`26 ± 7.2
`
`5
`
`10
`
`2
`11.1 ± 0.7
`59.4 ± 19.0
`322 ± 67
`127 ± 100
`4.0 ± 0.3
`160 ± 54.2
`74 ± 25.7
`39 ± 8.3
`224 ± 155.2
`
`6
`11.2 ± 0.9
`60.3 ± 28.8
`336 ± 80
`106 ± 89
`4.0 ± 0.4
`155 ± 52.6
`77 ± 31.2
`38 ± 9.9
`201 ± 134.0
`
`Statistical analysis
`The primary outcome variable of this study was the
`time course of urinary MDA, protein and NAG excre-
`tion rates, total amount excreted and their relationship
`to serum transferrin saturation. Other outcome vari-
`ables included the influence of NAC on these parameters.
`Treatment effects were analyzed using repeated measures
`analysis of variance with time and treatment as indepen-
`dent factors. Differences between means was tested by
`the least significant difference test. Scales of KDQOL
`were analyzed by repeated measures, one-way analysis
`of variance (ANOVA). All P values are two sided and
`significance set at <0.05. All statistical analyses were per-
`formed using Statistica for Windows, version 5.5 (Stat-
`Soft, Inc., Tulsa, OK, USA).
`
`RESULTS
`Study participants
`Actual subject study participation occurred between
`April 2003 and August 2003. Baseline and clinical char-
`acteristics are shown in Table 1. Subjects treated with in-
`travenous iron with and without NAC were similar with
`respect to demographic and clinical characteristics out-
`lined. There were no women reflecting the paucity of
`women with kidney disease among the veteran popula-
`tion. The distribution of the etiology of kidney disease,
`age, race, weight, and body mass index were similar. Base-
`line hematologic, lipid, and other laboratory characteris-
`tics were well matched.
`There was no increase in hemoglobin or reticulocytes
`over 1 week (Table 2). In fact, there was a small fall
`
`Luitpold Pharmaceuticals, Inc., Ex. 2011, p. 4
`Pharmacosmos A/S v. Luitpold Pharmaceuticals, Inc., IPR2015-01495
`
`

`
`Agarwal et al: Intravenous iron and oxidative stress
`
`2283
`
`Table 2. Baseline laboratory parameters before and after treatment
`with N-acetylcysteine (NAC) or no antioxidants
`Iron sucrose + placebo
`Iron sucrose + NAC
`Pre
`Post
`Pre
`Post
`8.0 ± 3.6
`7.8 ± 3.4
`7.3 ± 1.9
`6.7 ± 2.0
`White blood cells
`×1000/lL
`Hemoglobin g/dL 11.3 ± 1.1 11.3 ± 0.9
`11.1 ± 0.7
`10.8 ± 0.5
`84.0 ± 6.2 83.9 ± 6.7
`89.5 ± 5.6
`89.3 ± 5.5
`Mean corpuscular
`volume fL
`15.9 ± 1.6 15.8 ± 1.3
`14.6 ± 1.6
`14.5 ± 1.7
`RDW %
`33.5 ± 0.7 33.8 ± 0.7
`34.1 ± 0.8
`34.1 ± 0.8
`MCHC g/dL
`Platelets ×1000/lL 237 ± 95.6 243 ± 91.4 198 ± 40.6 192 ± 38.7
`1.9 ± 0.5
`1.8 ± 0.3
`1.4 ± 0.6
`1.4 ± 0.4
`Reticulocytes %
`59.4 ±28.8 51.9 ±14.7 59.4 ±19.0 54.1 ± 12.8
`Iron lg/dL
`351 ± 92.4 368 ± 91.9 322 ± 67.1 317 ± 44.2
`Total iron binding
`capacity lg/dL
`19 ± 6
`18 ± 10
`17 ± 3
`15 ± 5
`Transferrrin
`saturation %
`85 ± 75.4
`91 ± 57.2 127 ± 99.9 169 ± 107.3
`Ferritin ng/lL
`Abbreviations are: RDW, red cell distribution width; and MCHC, mean
`corpuscular hemoglobin concentration. Data shown are mean ± standard
`deviation. None of the changes are statistically significant except in hemoglobin
`(fall of 0.26 g/dL, (95% CI 0.5 to 0, P = 0.042) and in serum ferritin (increase of
`30.3 ng/mL, 95% CI 17.9 to 42.6 ng/mL, P < 0.001).
`
`in hemoglobin concentration of 0.26 g/dL (P = 0.042).
`Serum iron and TIBC were unchanged over 1 week, but
`serum ferritin increased by 30.3 ng/mL (P < 0.001).
`
`Time course of transferrin saturation, generation
`of oxidative stress, and renal injury
`Figure 2 shows the overall results of the trial. Plasma
`and urinary MDA (top two rows of plots) increased
`rapidly within 30 minutes of administration of intra-
`venous iron and were accompanied by an increase in pro-
`teinuria and enzymuria (third and fourth rows of plots). In
`contrast, transferrin saturation did not peak until 3 hours
`and returned to baseline by 24 hours (last row of plots). A
`significant effect of time was seen for all variables shown
`in Figure 1. Plasma MDA was increased at the 15 and
`30 minutes time point and was significantly different from
`the remaining levels (ANOVA, P < 0.0001). Urine MDA
`and NAG were maximal at 30 minutes and significantly
`different from rest of the excretion rates (ANOVA, P <
`0.0001). Proteinuria was also maximal at 30 minutes but
`the excretion rates at 15 and 60 minutes collection points
`were also similar to that at 30 minutes (ANOVA, P <
`0.0001). Since there was no effect of NAC and no dif-
`ference between the two visits, least square means from
`ANOVA and pooled standard errors of the means are
`shown in Table 3 to illustrate the time course of the ox-
`idative stress, iron saturation, and response.
`
`Effect of NAC on generation of oxidative stress
`and renal injury
`Compliance with intake of NAC assessed by measuring
`the volume of returned drug was 101% ± 8%. Because
`oxidative stress occurred within minutes, the area un-
`
`der the curve (AUC) of the plasma MDA concentration
`time curves were compared from baseline to 180 minutes.
`There was a significant improvement in plasma MDA
`generation with NAC (P = 0.048). The AUC between the
`two intravenous iron infusions using placebo were 184 ±
`244 lmol × min/L and 244 ± 51 lmol × min/L, whereas
`with NAC they were 217 ± 60 lmol × min/L and 199 ±
`45 lmol × min/L, respectively (Fig. 3, top panels). The
`redox ratio measured by plasma oxidized to reduced glu-
`tathione remained unaltered with treatment with NAC
`(Table 4). The total amount of MDA, protein, and NAC
`excreted in urine over the 24 hours in patients treated
`with NAC and placebo were unchanged. Although NAC
`did not change the 24-hour or the first 3-hour excretion of
`urinary MDA, Figure 3 shows that the amount of MDA
`excreted from 3 to 24 hours was reduced with NAC from
`4.24± 2.02 lg to 3.36± 1.47 lg compared to 3.50± 1.19 lg
`to 4.08 ± 1.88 lg following placebo (ANOVA for interac-
`tion, P = 0.041). The free radical scavenging enzymes in
`the red blood cells such as SOD and GSHPx remained un-
`changed with the administration of NAC (Table 4). NAC
`was unable to reduce the urinary MCP-1 levels that were
`unchanged 1 week after intravenous iron administration
`(Table 4).
`
`Changes in kidney disease quality of life
`There were no statistically significant changes in
`KDQOL, except in the pain scale (Table 5). Despite nee-
`dle sticks and intravenous catheters, subjects reported im-
`provement in pain (higher number is associated with less
`pain) on average 14.5% (95% CI, 0.06 to 28.9, P = 0.049).
`A trend toward improvement in physical functioning was
`seen 5.25% (95% CI, −1.1 to 11.6, P = 0.10).
`
`Safety and adverse events
`There were two serious adverse events seen. One pa-
`tient had sudden death the night after the day of infusion.
`He had underlying coronary artery disease and the event
`was not believed to be related to the study drug. Two pa-
`tients had diarrhea thought to be related to the study drug.
`One patient had diarrhea about 12 hours after the infu-
`sion that was accompanied by bloody stools. The other
`patient had diarrhea within 1 hour of administration of
`the study drug.
`
`DISCUSSION
`The major findings of our study are that oxidative stress
`occurs rapidly with infusion of intravenous iron sucrose
`in patients with CKD in doses recommended by the man-
`ufacturer. Oxidative stress so induced is accompanied
`by the occurrence of renal tubular damage and possibly
`increase in glomerular permeability as assessed by en-
`zymuria and increase in proteinuria, respectively. These
`
`Luitpold Pharmaceuticals, Inc., Ex. 2011, p. 5
`Pharmacosmos A/S v. Luitpold Pharmaceuticals, Inc., IPR2015-01495
`
`

`
`2284
`
`Agarwal et al: Intravenous iron and oxidative stress
`
`Iron sucrose + NAC
`Pre NAC
`Post NAC
`
`0
`
`200
`
`400
`
`600
`
`800 1400
`
`0
`
`200
`
`400
`
`600
`
`800
`
`2.0
`
`1.5
`
`1.0
`
`0.5
`
`0.0
`
`Plasma MDA, µmol/L
`
`20
`
`15
`
`10
`
`05
`
`40
`
`30
`
`20
`
`nmol/min
`
`Urinary MDA excretion rate,
`
`Iron sucrose – NAC
`Pre placebo
`Post placebo
`
`0
`
`200
`
`400
`
`600
`
`800 1400
`
`0
`
`200
`
`400
`
`600
`
`800
`
`2.0
`
`1.5
`
`1.0
`
`0.5
`
`0.0
`
`Plasma MDA, µmol/L
`
`20
`
`15
`
`10
`
`05
`
`40
`
`30
`
`20
`
`nmol/min
`
`Urinary MDA excretion rate,
`
`mU/min
`
`0
`
`200
`
`400
`
`600
`
`800
`
`0
`
`200
`
`400
`
`600
`
`800
`
`0
`
`200
`
`600
`400
`Time, minutes
`
`800 1400
`
`10
`
`0
`
`mU/min
`
`Urinary NAG excretion rate,
`
`5000
`4000
`3000
`2000
`1000
`0
`
`Urinary protein excretion rate,
`
`µg/min
`
`100
`
`80
`
`60
`
`40
`
`20
`
`0
`
`Transferrin saturation, %
`
`0
`
`200
`
`400
`
`600
`
`800
`
`0
`
`200
`
`400
`
`600
`
`800
`
`0
`
`200
`
`600
`400
`Time, minutes
`
`800 1400
`
`10
`
`0
`
`5000
`4000
`3000
`2000
`1000
`0
`
`Urinary NAG excretion rate,
`
`Urinary protein excretion rate,
`
`µg/min
`
`100
`
`80
`
`60
`
`40
`20
`
`0
`
`Transferrin saturation, %
`
`Fig. 2. Time course of biomarkers of oxidative stress, renal injury, and transferrin saturation. Time refers to the actual time of blood sampling or,
`in the case of urinary excretion rates, the midpoint of the sampling interval. Time 0 is the midpoint of injection of iron sucrose. Error bars reflect
`standard errors. Abbreviations are: NAC, N-acetylcysteine; MDA, malondialdehyde; NAG, N-acetyl-b-D-glucosaminidase.
`
`Luitpold Pharmaceuticals, Inc., Ex. 2011, p. 6
`Pharmacosmos A/S v. Luitpold Pharmaceuticals, Inc., IPR2015-01495
`
`

`
`Agarwal et al: Intravenous iron and oxidative stress
`
`2285
`
`Table 3. Time course of transferrin saturation (TSAT), oxidative stress, and renal injury markers
`
`Time
`minutes
`
`0
`15
`30
`60
`120
`180
`1440
`
`Urine N-acetyl-b-D-
`Urine
`Plasma
`glucosaminidase
`malondialdehyde
`malondialdehyde
`(NAG) mU/min
`(MDA) nmol/min
`(MDA) lmol/L
`TSAT %
`13.5 ± 1.5
`2.94 ± 0.24
`0.89 ± 0.14
`9.7 ± 2.2
`15.1 ± 2.5
`6.33 ± 0.88
`1.42 ± 0.12
`41.4 ± 5
`23.4 ± 5.6
`13.32 ± 1.81
`1.39 ± 0.2
`45.6 ± 5
`14.2 ± 3.3
`9.44 ± 1.79
`1.05 ± 0.11
`52 ± 5.1
`12.6 ± 2.5
`6.85 ± 0.78
`1.08 ± 0.16
`63.3 ± 4.8
`12.7 ± 3.8
`7.28 ± 1.01
`0.95 ± 0.13
`73 ± 4.8
`5.8 ± 1.3
`3.00 ± 0.39
`0.75 ± 0.1
`34.9 ± 7.3
`Influence of iron sucrose on TSAT, oxidative stress (plasma MDA, urine MDA) and renal injury (urine NAG, urine protein) markers in patients with chronic kidney
`disease (CKD). There was a significant treatment time interaction for all parameters. Values are least square mean ± pooled standard error of mean. Urine protein
`excretion rates represent geometric means and 95% confidence intervals.
`
`Midpoint of urine
`collection
`minutes
`−15
`7.5
`22.5
`45
`90
`150
`810
`
`Urine protein
`lg/min
`
`237 (61,914)
`312 (80,1219)
`337 (73,1566)
`251 (57,1099)
`203 (47,883)
`189 (39,916)
`100 (24,424)
`
`P = 0.048
`
`NS
`
`1400
`1200
`1000
`800
`600
`400
`200
`
`AUC plasma MDA,
`
`µmol × min/L
`
`Pre Post
`Iron sucrose
`– NAC
`
`Pre Post
`Iron sucrose
`+NAC
`
`P = 0.042
`
`Pre Post
`Iron sucrose
`– NAC
`
`Pre Post
`Iron sucrose
`+NAC
`
`Fig. 3. Area under curve (AUC) of plasma
`malondialdehyde (MDA) versus time curves
`from 0 to 3 hours and 3 to 24 hours show sig-
`nificant improvement with N-acetylcysteine
`(NAC). Amount of MDA excreted in urine
`over 3 to 24 hours was significantly improved
`with NAC. Error bars reflect the standard de-
`viations and the P values reflect the time-
`treatment interaction effect.
`
`02468
`
`Urine MDA excretion,
`
`µmol/L
`
`Pre Post
`Iron sucrose
`– NAC
`
`Pre Post
`Iron sucrose
`+NAC
`
`NS
`
`Pre Post
`Iron sucrose
`– NAC
`
`Pre Post
`Iron sucrose
`+NAC
`
`300
`
`250
`
`200
`
`150
`
`100
`
`AUC plasma MDA,
`
`µmol × min/L
`
`3.0
`2.5
`2.0
`1.5
`1.0
`0.5
`0.0
`
`Urine MDA excretion,
`
`µmol/L
`
`changes occur at a time when there is substantial fraction
`of transferrin available to bind to free iron. Furthermore,
`despite some improvement in oxidative stress markers
`with NAC, there is no protection from renal injury. Taken
`together, these data suggest direct damage to the kid-
`ney by the drug, independent of transferrin saturation.
`However, these changes are transient and resolve rapidly
`within 24 hours.
`The time course of transferrin saturation follows what
`has been reported by Parkkinen et al [26]. In hemodialy-
`sis patients, they reported transferrin saturation to be be-
`tween 80% and 90% at 31/2 hours following iron sucrose
`administration in dose and duration of infusion that were
`identical to our study. Transferrin saturation returned to
`normal in 48 hours. Free iron, measured as bleomycin-
`detectable iron, was not increased unless transferrin satu-
`ration was 80% or more. They reported growth of Staphy-
`
`lococcus epidermidis to be increased when incubated in
`serum collected at 31/2 hours and arrested by adding ex-
`ogenous apotransferrin strongly implicating the role of
`free iron in this process.
`In contrast to the findings of bacterial overgrowth in
`serum of patients with transferrin saturation, we found
`maximal oxidative stress, enzymuria and proteinuria to
`occur within 15 to 30 minutes of the infusion. These
`data suggest a novel mechanism of injury independent
`of transferrin saturation. No further increase in oxida-
`tive stress or renal damage was noted beyond the first
`30 minutes, despite progressive uptake of iron by trans-
`ferrin. The results of the antioxidant, NAC, administra-
`tion were mixed. Whereas NAC reduced acute systemic
`generation of oxidative stress and delayed renal genera-
`tion of oxidative stress, there was no improvement in ei-
`ther proteinuria or enzymuria. This further supports the
`
`Luitpold Pharmaceuticals, Inc., Ex. 2011, p. 7
`Pharmacosmos A/S v. Luitpold Pharmaceuticals, Inc., IPR2015-01495
`
`

`
`2286
`
`Agarwal et al: Intravenous iron and oxidative stress
`
`Table 4. Parameters related to redox ratio, free radical scavengers, and inflammation
`Iron sucrose +
`N-acetylcysteine (NAC)
`Baseline
`One week
`4580 ± 1108
`4589 ± 1204
`6.3 ± 0.5
`6.3 ± 0.7
`65 ± 13
`59 ± 15
`367 ± 424
`323 ± 317
`
`Iron sucrose + placebo
`Baseline
`One week
`4382 ± 1458
`4234 ± 1779
`6.6 ± 0.9
`6.6 ± 0.8
`48 ± 13
`49 ± 16
`422 ± 296
`416 ± 477
`
`Glutathione peroxidase U/mL red blood cells
`Superoxide dismutase (SOD) U/mL red blood cells
`Oxidized Glutathione %
`24-hour urinary monocyte chemoattractant protein-1
`(MCP-1) after iron infusion pg/mg creatinine
`
`Glutathione peroxidase, SOD, and plasma oxidized glutathione were measured before iron infusion. Urinary MCP-1 was measured in the 24-hour urine specimen
`following iron infusion. Data shown are mean ± SD. Baseline and 1 week refer to the time of measurements. These measurements were performed before iron infusion.
`
`Scale
`
`Table 5. Kidney disease quality of life before and after 1 week of iron administration
`Iron sucrose +
`Iron sucrose + placebo
`N-acetylcysteine (NAC)
`Baseline
`One week
`Baseline
`One week
`73 ± 14.9
`78 ± 13.3
`81 ± 12.4
`86 ± 10.8
`Symptom/problem list
`79 ± 20.9
`83 ± 12.4
`80 ± 16.0
`82 ± 10.9
`Effects of kidney disease
`74 ± 26.6
`75 ± 15.0
`69 ± 23.3
`72 ± 12.1
`Burden of kidney disease
`30 ± 35.0
`25 ± 35.4
`33 ± 35.4
`22 ± 36.3
`Work status
`85 ± 19.6
`91 ± 10.0
`89 ± 12.3
`91.1 ± 11.1
`Cognitive function
`81 ± 14.2
`82 ± 20.4
`85 ± 18.9
`86 ± 13.5
`Quality of social interaction
`55 ± 24.9
`68 ± 21.6
`65 ± 27.1
`72 ± 23.3
`Sleep
`92 ± 14.2
`93 ± 11.7
`92 ± 11.8
`83 ± 32.3
`Social support
`54 ± 23.2
`55 ± 22.2
`53 ± 22.9
`59 ± 20.3
`Overall health
`29 ± 26.3
`35 ± 28.1
`26 ± 22.0
`38 ± 27.9
`Physical functioning
`20 ± 32.9
`43 ± 45.7
`30 ± 35.0
`33 ± 35.4
`Role limitations—physical
`50 ± 28.1
`64 ± 30.1
`46 ± 21.2
`68 ± 25.1
`Pain
`40 ± 21.7
`41 ± 23.1
`43 ± 14.6
`38 ± 15.4
`General health
`79 ± 19.2
`85 ± 11.1
`83 ± 15.6
`78 ± 14.3
`Emotional well-being
`93 ± 14.1
`80 ± 35.8
`60 ± 46.6
`59 ± 32.4
`Role limitations—emotional
`64 ± 30.3
`76 ± 32.5
`68 ± 34.0
`78 ± 19.5
`Social function
`27 ± 20.0
`28 ± 21.1
`33 ± 16.0
`33 ± 21.5
`Energy/fatigue
`27 ± 6.0
`36 ± 7.7
`29 ± 8.6
`33 ± 10.9
`SF-12 Physical Health Composite
`54 ± 7.8
`53 ± 7.6
`53 ± 9.9
`48 ± 7.8
`SF-12 Mental Health Composite
`Data are mean ± standard deviation. No interaction effect between iron infusion and use of antioxidant was seen. There was significant improvement in pain subscale,
`but no other statistical change.
`
`hypothesis that iron sucrose caused damage not simply by
`increasing oxidative stress and/or free iron but also per-
`haps by a direct cytotoxic effect on the renal tubules. This
`may be due to direct effect of the drug on the glomerular
`permeability and tubular integrity. Danielson et al [27]
`