Kidney International, Vol. 63 (2003), pp. 217–224

CLINICAL NEPHROLOGY – EPIDEMIOLOGY – CLINICAL TRIALS

Sodium ferric gluconate complex in hemodialysis patients.II. Adverse reactions in iron dextran-sensitiveand dextran-tolerant patients

DANIEL W. COYNE, N. FRANKLIN ADKINSON, JR., ALLEN R. NISSENSON, STEVEN FISHBANE,RAJIV AGARWAL, JOSEPH W. ESCHBACH, BECKIE MICHAEL, VAUGHN FOLKERT, DANIEL BATLLE,J. RICHARD TROUT, NAOMI DAHL, PAMELA MYIRSKI, JUR STROBOS, and DAVID G. WARNOCK,for the FERLECIT� INVESTIGATORS

Washington University School of Medicine, St. Louis, Missouri; Johns Hopkins University, Baltimore, Maryland; UCLA Schoolof Medicine, Los Angeles, California; Winthrop University Hospital, Mineola, New York; Indiana University School ofMedicine, Indianapolis, Indiana; Northwest Kidney Center, Seattle, Washington; Thomas Jefferson University, Philadelphia,Pennsylvania; Albert Einstein College of Medicine, New York, New York; Northwestern University, Chicago, Illinois; RutgersUniversity, Trenton, and Watson Laboratories, Morristown, New Jersey; Baltimore, Maryland; Washington, D.C.; andUniversity of Alabama at Birmingham, Birmingham, Alabama, USA

confirmed allergic events in the iron dextran-tolerant group.Sodium ferric gluconate complex in hemodialysis patients. II.Long-term exposure to SFGC in iron dextran-sensitive patientsAdverse reactions in iron dextran-sensitive and dextran-toler-resulted in intolerance in only one additional patient and noant patients.serious adverse events.Background. Iron dextran administration is associated with

Conclusions. Patients with a history of iron dextran sensitiv-a high incidence of adverse reactions including anaphylaxis anddeath. Although dextran, rather than iron, is believed to be the ity had approximately sevenfold higher rates of reaction tocause of these reactions, it is not known whether iron dextran- both placebo and SFGC compared to iron dextran tolerantsensitive patients can be safely administered another form of patients. However, logistic regression analysis, performed toparenteral iron, sodium ferric gluconate in sucrose (SFGC). account for the higher reaction rate to placebo, suggests that

Methods. In a 69 center, prospective, double-blind, con- this increased reactivity was not drug-specific nor immunologi-trolled trial of safety and tolerability of SFGC, the rate of cally mediated, but represented host idiosyncrasy. These resultsreactions to SFGC and placebo in 144 iron dextran-sensitive support the conclusions that reactions to SFGC can be attrib-patients was compared with 2194 patients who were previously uted to pseudoallergy, and that SFGC is not a true allergen.tolerant to iron dextran preparations. Serum tryptase levels, amarker of mast cell degranulation, also were measured.

Results. Among 143 iron dextran-sensitive patients exposedto SFGC, three (2.1%) were intolerant. All three had suspected The frequent blood loss associated with dialysis andallergic events to SFGC, including one patient with a serious its related procedures is estimated to result in 1 to 3 gramsreaction (0.7%). One dextran-sensitive patient (0.7%) had a of iron loss per year in chronic hemodialysis patients [1].suspected allergic reaction after placebo. In contrast, among

Regular administration of iron to hemodialysis patients2194 iron dextran-tolerant patients, reactions to SFGC weresignificantly less common, with SFGC intolerance seen in seven increases the percentage of patients in the target hemo-patients (0.3%; P � 0.020), including five (0.2%) who had globin range and reduces requirements for recombinantsuspected allergic events (P � 0.010), but none who had serious human erythropoietin (epoetin). Guidelines from bothevents (0.0%; P � 0.061). Two iron dextran-tolerant patients

American and European nephrology groups recommend(0.09%) had allergic-like reactions following placebo injec-administration of intravenous iron as both repletion andtions. Two of the three suspected allergic events in the iron

dextran-sensitive group were confirmed as mast cell dependent maintenance therapy in hemodialysis patients [2, 3] toby a 100% increase in serum tryptase, while there were no achieve target hemoglobin and ensure efficient use of

erythropoietin.For many years the only form of parenteral iron avail-Key words: allergy, drug safety, intravenous iron, renal failure, tryp-

tase, anemia, parenteral iron, hemoglobin. able in the United States was iron dextran. Review ofdrug reporting databases by Faich and Strobos identified

Received for publication February 28, 200231 deaths reportedly due to administration of iron dex-and in revised form July 16, 2002

Accepted for publication August 7, 2002 tran [4]. Allergic reactions due to iron dextran werereported in 8.7 per million doses per year but only in 3.3 2003 by the International Society of Nephrology

217

Coyne et al: Reactions to sodium ferric gluconate complex218

per million doses per year with sodium ferric gluconate based on information in the patient record and patientcomplex in sucrose (SFGC) [4]. In the same study there history. A predefined secondary end-point of the blindedwere 74 “allergic” events related to SFGC administration. study was to evaluate the cross sensitivity of SFGC andNone of these 74 events associated with SFGC were fatal. iron dextran by comparing patients who previously toler-In contrast, 31 of 196 (15.8%) of the allergic events ated iron dextran with those patients with prior ironreported for iron dextran resulted in death. While pseudo- dextran intolerance. The federal Food and Drug Admin-allergic reactions usually occur during the first exposure istration (FDA) approved the final study design.to iron dextran, anaphylactic reactions can follow multi-

Patient selectionple uneventful exposures [5]. Furthermore, iron dextranintolerance (which includes all forms of reactions) is Patients were enrolled in the double blind study at 69common, occurring in approximately 2.5% of iron dex- centers in the U.S. between August 1999 and Octobertran-naı̈ve patients [4]. The molecular weight of iron 2000, and all centers obtained approval from their re-dextran also appears to influence the incidence of ana- spective Institutional Review Boards. Patient eligibilityphylactic and intolerance reactions, with higher molecu- and exclusion criteria have been reported in detail pre-lar weight Dexferrum� being associated with a higher viously [7]. Briefly, patients with known sensitivity toincidence of reactions than InFeD� [5]. These data have iron dextran were limited in this trial to avoid over-called into question the wisdom of using iron dextran enrollment of atopic or multiple drug allergy patients,preparations at all, particularly if safer parenteral iron which could bias the overall study toward over-reactions.preparations are available [6]. Centers could enroll up to 10% of patients with a history

It is not known whether other forms of parenteral iron of iron dextran intolerance. Patients were excluded fromsuch as SFGC can be used safely in iron dextran-sensitive the study if they had prior treatment with SFGC, or werepatients. We have recently reported on a randomized undergoing acute or chronic therapy with antihistaminesdouble blind study of SFGC versus placebo in chronic or corticosteroids, which could mask possible allergicdialysis patients in the United States that confirmed the reactions. The first 1412 patients completing the double-safety of SFGC in a large number of patients, the vast blinded study without drug intolerance to SFGC thenmajority of whom had prior iron dextran tolerance [7]. entered an open-label study of SFGC, described below.Since SFGC does not contain any dextran moieties thatare thought to be important antigenic determinants, we Study designhypothesized that the incidence of drug intolerance and Patients who satisfied the inclusion criteria and signedadverse reactions to SFGC in those patients who did informed consent participated in the double-blind cross-have a history of iron dextran-sensitivity would not ex- over study that included four sequential hemodialysisceed that seen with placebo administration. sessions (HD). The first session consisted of screening

We report here the results of SFGC and placebo ad-procedures, medical history, review of medications and

ministration to the 144 patients from the randomizedallergy history, and physical exam. At the second session,

double blind study who had a history of prior iron dex-patients were randomized into one of two crossover

tran reactions, as well as the results of subsequent admin-treatment schedules: SFGC at HD session number 2 and

istration of multiple doses of SFGC to a subset of 92 ironplacebo at HD session number 3; or placebo at HD sessiondextran-sensitive patients who entered an open-labelnumber 2 and SFGC at HD session number 3. At HDstudy of SFGC. We also determined whether suspectedsession number 4, the patient was assessed for any ad-allergic reactions following SFGC and placebo were mastverse reactions after the third session. A single undilutedcell dependent by measuring serum tryptase, a markerdose of 125 mg of SFGC was administered intravenouslyof mast cell degranulation.over 10 minutes during the first hour of HD sessionnumber 2 or 3. Placebo was composed of 10 mL of

METHODS bacteriostatic saline (containing 9 mg/mL of benzyl alco-hol) and was administered at the same rate as SFGC atData were collected for patients with a history of ironthe other HD session. The placebo as well as study drugdextran allergies, known iron dextran tolerance, or sus-contained the preservative benzyl alcohol so that anypected allergic reactions to SFGC during a multicenter,reaction occurring after the study drug would be attribut-crossover, randomized, double-blind, placebo controlledable only to SFGC. Unblinded personnel prepared bothprospective trial of SFGC performed in 2534 hemodialy-placebo and SFGC; each syringe was covered with ansis patients [7]. Data also were collected from a subse-adhesive opaque paper and attached to a needle with aquent open label long-term study of SFGC in 1412 pa-steel hub to ensure blinding of personnel who adminis-tients successfully completing the double-blind study.tered the drug and who reported any adverse events.Prior dextran sensitivities were classified by each partici-

pating center as mild, drug intolerance, or anaphylactoid, Vital signs (blood pressure and pulse) were recorded at

Coyne et al: Reactions to sodium ferric gluconate complex 219

Table 1. Patient demographic data tryptase [8–10] levels determined during HD sessionnumber 2. By measuring tryptase levels, we defined theIron dextran Iron dextran

sensitive tolerant normal range for hemodialysis patients, which is notCharacteristics N � 144 N � 2194 different from the normal population (�10 ng/mL).Mean age 53.2 55.8 These results were used to define a significant increase% Male 45.1a 55.3

in post-event tryptase level compared to the baseline pre-Ethnicity N (%)White 23 (16.0) 474 (21.6) event tryptase level, being set at �2 standard deviationsBlack 87 (60.4) 1297 (59.2) from the mean, which was a doubling of the pre-dialysisHispanic 26 (18.1) 363 (16.5)

value. All patients in the double-blind study had bloodOther 8 (5.5) 60 (2.7)�2 non-iron drug allergies 28 (19.4)b 294 (13.4) samples obtained prior to HD sessions number 2 and 3

for possible baseline tryptase levels. If a suspected imme-aP � 0.02 compared to iron dextran tolerant groupbP � 0.12 compared to iron dextran tolerant group diate-type allergic reaction occurred, a second tryptase

level was drawn approximately one hour after the eventto evaluate possible drug-induced mast cell degranula-tion [8, 9]. If no reaction occurred, the baseline samplebaseline, five minutes into the injection, and at five andwas discarded.20 minutes after completion of the infusion [7].

The open label study began immediately following Statistical methodssuccessful completion of the double-blind crossover study.

Suspected and confirmed immediate-type allergicEach patient entering the open label study was assignedevents were compared between SFGC and placebo using20 vials each containing 62.5 mg of SFGC. The investiga-McNemar’s Test, which discounts patients with reactionstor and the primary attending nephrologist had full dis-to both drug and placebo. Logistic regression analysescretion on how to administer SFGC during the openwere performed on the incidence of all adverse eventslabel study. The patient completed the open label phaseto assess whether any cross-reactivity to iron dextranwhen all 1250 mg of SFGC had been administered orwas independently related to either SFGC or placebo.nine months had passed.Statistical significance was assessed between iron dextranexposed groups, with regards to the incidence of reac-Adverse events and allergic eventstions, by use of a two-sided Fisher’s Exact test. Statistical

Any adverse reaction that occurred during study drugsignificance was assessed between iron dextran intolerant

infusion was classified as an instantaneous reaction; oneand iron dextran tolerant groups by using a two-sided

that occurred during the remainder of the dialysis sessionFisher’s Exact test.

was considered an immediate reaction. Before the initia-tion of dialysis at the subsequent session, patients were

RESULTSreassessed for any intervening adverse events, which wereconsidered delayed reactions. Predefined adverse out- Among the 2534 patients enrolled in the double-blindcomes were classified by blinded personnel as: placebo-controlled trial, 2338 had previously received

iron dextran: 144 had prior iron dextran sensitivity and(1) Drug intolerance events, which precluded further 2194 were known to be iron dextran-tolerant. The demo-

drug administration; graphics of these two groups are shown in Table 1. Iron(2) Serious adverse events, which were any instanta- dextran-sensitive patients were less likely to be male (P �

neous or immediate reactions requiring extraordi- 0.02). Multiple drug allergies were more common in thenary treatment such as antihistamines, vasopres- iron dextran-sensitive group (P � 0.12), although thesors, or steroids; difference was not statistically significant.

(3) Suspected allergic events, which were any clinical Among the 144 patients with known iron dextran sen-events suggesting drug allergy, such as pruritic sitivity, 143 were exposed to SFGC during the placebo-rash, bronchospasm, or unexplained hypotension; controlled study. One patient had a reaction after pla-and cebo and withdrew consent before receiving SFGC. The

(4) Confirmed immediate-type allergic events, which previous reactions to iron dextran preparations werewere any suspected allergic reaction accompanied classified as mild, drug intolerance, or anaphylactoid.by a 100% increase in baseline serum tryptase level. The distributions of reactions by severity and iron dex-

tran preparations used at the time of the reaction areSerum tryptase determinations shown in Table 2. Among the 2338 patients exposed to

The initial 206 hemodialysis patients receiving SFGC any form of iron dextran, 1937 (82.8%) had receivedor placebo without incident at early selected sites had InFeD� only, 261 (11.2%) had received Dexferrum�

only, and 140 (6.0%) had received both drugs. The inci-baseline (immediate pre-dialysis) and post-dialysis total

Coyne et al: Reactions to sodium ferric gluconate complex220

Table 2. Summary of reactions in iron dextran sensitive patients a prior history of anaphylactoid reactions to both com-mercially available iron dextrans and allergic reactionsInFeD� Dexferrum� Totalto penicillin and cephalexin. The patient developed dys-Number exposed 2077 401 2338a

Any sensitivity 113 (5.4%)c 39 (9.7%)d 144 (6.2%)b pnea, hypotension, and wheezing which began four min-Type of sensitivity utes after completion of the SFGC injection. The patient

Mild 24 (1.2%) 11 (2.7%) 32 (1.4%)b

was given intravenous diphenhydramine, intravenousDrug intolerance 60 (2.9%) 21 (5.2%) 77 (3.3%)b

Anaphylactoid 28 (1.3%) 7 (1.7%) 34 (1.5%)b hydrocortisone, and subcutaneous epinephrine. The re-action resolved after 20 minutes, and the patient com-aSome patients were exposed to both agents, but are counted only once here

bSome patients were sensitive to both agents, but are counted only once here pleted the dialysis treatment and went home. The pa-cOne patient was dextran sensitive, but the type of sensitivity was not classified

tient’s baseline serum tryptase level of 11.7 ng/mL wasdP � 0.002 compared to InFed� incidence

approximately twofold higher than average of the con-trol group, consistent with constitutive, on-going mastcell degranulation. Tryptase decreased after the reaction

dence of Dexferrum� reactions was significantly higher to 10.8 ng/mL, and this serious adverse reaction was(39/401; 9.7%) than reactions to InFeD� (113/2077; 5.4%; therefore classified as a non-allergic event.P � 0.002). Among patients exposed to only one form Among the seven iron dextran-tolerant patients classi-of iron dextran, Dexferrum� reactions were still signifi- fied as intolerant to SFGC, two had reactions that werecantly higher (24/261; 9.2%) than reactions to InFeD� not consistent with an allergic event: isolated hypoten-(105/1937; 5.4%) (P � 0.021). sion (N � 1), and nausea (1). Four of these iron dextran-

Tryptase levels were obtained predialysis and one tolerant patients had suspected allergic reactions tohour after study drug administration in the first 206 pa- SFGC. These reactions were pruritus (2), chills (1), andtients at selected sites during HD session number 2 of dyspnea with chest pain (1). One of these patients alsothe double-blind trial. In this control group, baseline and experienced a delayed rash following placebo (discussedpost-dialysis tryptase levels were 4.45 � 3.44 and 4.62 � later in this section). Tryptase did not increase more4.06 ng/mL, respectively. These results demonstrated than 100% following any of these events, and thereforethat a tryptase value was detectable predialysis, was not all were classified as non-allergic events. The seventhstimulated by dialysis, and was not significantly different patient had a delayed rash following SFGC; because athan values in normal subjects [8]. A significant change post-event tryptase level could not be obtained this wasin tryptase levels was defined by these results as at least classified as an unconfirmed suspected allergic event.a 100% increase compared to the baseline level. SFGC intolerance due to suspected allergic events was

significantly more common (P � 0.010) in iron dextran-Intolerance reactions to SFGC sensitive patients (3/143; 2.1%) than iron dextran-toler-

Three of the 143 patients (2.1%) with prior reactions ant patients (5/2194; 0.2%).to iron dextran were considered intolerant to SFGC, while

Non-intolerance suspected allergic reactions to SFGCjust seven of 2194 (0.3%) iron dextran-tolerant patientswere considered intolerant to SFGC (P � 0.020). One Six other patients, and the patient described earlierpatient was considered intolerant to placebo and was iron with a delayed rash following placebo, had suspecteddextran tolerant. Results are summarized in Table 3. Over- allergic events that were not considered severe enough toall, patients with suspected allergic reactions did not have be classified as drug intolerance by the local investigator.a significant increase in tryptase level (baseline 6.13 � One of these patients had prior iron dextran sensitivity.3.76 ng/mL; post-treatment 6.95 � 3.43 ng/mL), and Three of these reactions occurred after placebo: nausea,these values were not significantly different than the dizziness, headache and vomiting in the patient withcontrol group’s baseline (4.45 � 3.44 ng/mL, P � 0.118) prior iron dextran sensitivity (1); flushing and malaiseand post-treatment tryptase levels (4.62 � 4.06 ng/mL, (1); and a delayed rash (1). The four reactions followingP � 0.063). SFGC were also possibly allergic, but were not consid-

Among the three intolerance reactions to SFGC in ered severe enough to warrant drug cessation. These reac-the dextran-sensitive patients that precluded further ad- tions were: nausea, unease and dry throat (1); nausea andministration of SFGC, one patient had back pain follow- dizziness (1); rash (1); and pruritus (1). None of theseing SFGC that was classified as mild by the local investi- reactions had �100% increase in tryptase post-event,gator. Tryptase levels rose from 3.8 to 7.8 ng/mL; by and therefore they were not classified as confirmed aller-definition a confirmed allergic reaction. A second patient gic events. Tryptase change from baseline could not bedeveloped instantaneous facial flushing on exposure to obtained in the delayed reaction.SFGC. Tryptase increased from 2.1 to 4.9 ng/mL and A total of 15 patients of 2338 with prior dextran expo-this response also was classified as a confirmed allergic sure had suspected allergic events during the double-

blind trial (Table 4). Twelve events occurred after SFGC,reaction. The third patient with SFGC intolerance had

Coyne et al: Reactions to sodium ferric gluconate complex 221

Table 3. Summary of SFGC intolerance by patient group

Iron dextran sensitive Iron dextran tolerant Total patients P valueb

End point N � 143 N � 2194 N � 2338 (Fisher’s Exact Test)

SFGC intolerance 3 (2.1%) 7 (0.3%) 10 (0.4%) 0.020Serious event 1 (0.7%) 0 (0.0%) 1 (0.04%) 0.061Suspected allergic events 3 (2.1%)a 5 (0.2%) 8 (0.3%) 0.010

aIncludes serious eventbIron dextran-sensitive group compared to Iron dextran-tolerant group

Table 4. All suspected allergic events by patient group

Iron dextran sensitive Iron dextran tolerant Total patientsEnd point N � 143 N � 2194 N � 2338

All suspected allergic events after SFGC 3 (2.1%)a 9 (0.4%)b 12 (0.5%)b

Delayedc 0 (0.0%) 1 (0.04%) 1 (0.04%)Immediate 3 (2.1%) 8 (0.3%) 11 (0.5%)Confirmed allergic event (tryptase increase �100%) 2 (1.4%) 0 (0.0%) 2 (0.08%)

Suspected allergic events after placebo 1 (0.7%) 2 (0.09%) 3 (0.12%)a P � 0.033 compared to iron dextran tolerant groupb Not all suspected allergic events were considered severe enough to warrant drug cessation. Four patients had mild immediate reactions not considered drug

intolerance. One patient had a suspected allergic reaction to both placebo and SFGC.c Delayed reactions were reported after the end of dialysis and immediate reactions were reported during the dialysis session

and four events after placebo, an event rate that is not SFGC-intolerant. Eight iron dextran-sensitive patientssignificantly different (P � 0.12). One patient had sus- received at least one dose of SFGC of greater than 125pected allergic events after both SFGC and placebo. Sus- mg during at least one treatment, and these eight eachpected allergic events following SFGC were significantly received an average cumulative dose of 1141 mg. Nonemore common (P � 0.033) in iron dextran-sensitive pa- had a serious event, a suspected allergic event, nor weretients (3/143; 2.1%) than iron dextran-tolerant patients considered intolerant to this higher dose of SFGC.(9/2194; 0.4%). Similarly, suspected allergic events after Among the eight patients who had suspected allergicplacebo were more common in iron dextran-sensitive pa- reactions to SFGC or placebo during the double-blindtients (1/143; 0.7%) than dextran-tolerant patients (2/2194; trial yet were not considered SFGC-intolerant, three en-0.09%), but this was not statistically significant. tered the open label study, and received multiple doses

Among all 2534 patients enrolled in the study, only of SFGC. None had a serious event, however, one patienttwo confirmed allergic reactions occurred; both were in complained of shortness of breath and a hot flash aroundiron dextran-sensitive patients following SFGC. Logistic his neck two days after receiving his third dose of SFGC.regression failed to identify cross-sensitivity between The investigator considered this reaction likely to beSFGC and prior iron dextran sensitivity in the incidence drug related and discontinued the patient from the study.of all adverse events (P � 0.69). The low occurrence ofsuspected allergic events did not permit direct logistic

DISCUSSIONregression analysis for these events.Administration of parenteral iron is essential to achieve

Repeated exposure to SFGC and maintain target hemoglobin in hemodialysis patients.Numerous studies have demonstrated that parenteralAmong the 2490 patients completing the double-blindiron can increase hemoglobin in epoetin-treated dialysistrial with demonstrated tolerance to SFGC, 1412 enteredpatients, and/or reduce the amount of epoetin neededan open label trial where 1250 mg of SFGC was adminis-to achieve target hemoglobin. Practice guidelines fromtered within nine months in any fashion at the localthe National Kidney Foundation and the European Bestinvestigators’ discretion. Of the 140 iron dextran-sensi-Practices Guidelines recommend repletion of iron storestive patients who were SFGC-tolerant, 92 entered thisin iron deficient hemodialysis patients, then maintenanceopen label trial. SFGC was administered as �125 mgof those stores by regular administration of parenteralper treatment in 84 of these patients. Each of theseiron [2, 3].patients received an average cumulative dose of 1155 �

In the United States, iron dextrans were the only ap-255 mg of SFGC. None had a serious event. One patientproved forms of parenteral iron available until 1998.received five doses of SFGC uneventfully, complained

of mild pruritus with the sixth dose, and was then labeled Retrospective and prospective studies have established

Coyne et al: Reactions to sodium ferric gluconate complex222

that iron dextrans can cause serious and life-threatening was not allergenic, based on the evidence that dextran,and not iron oxide, accounts for these serious and life-hypersensitivity reactions [4]. A meta-analysis of studies

of intravenous iron dextran preparations found a drug threatening hypersensitivity reactions. In this context itis important to note that drug reactions may result fromintolerance rate of 2.47 � 0.31%, and life threatening

reaction rate of 0.61 � 0.13% [7]. While all current intra- drug-specific immunological activity, or from “pseudo-allergy” [21]. Pseudoallergy results in reactions consis-venous iron products contain iron oxide, there are sig-

nificant chemical variations in the iron ligands. Iron dex- tent with drug allergy, but is due to host idiosyncrasyrather than a drug-specific immune response. There istran is composed of iron oxide crystals surrounded by a

matrix of cross-linked dextrans. Interestingly, our data no clear-cut method to distinguish drug immunoreactiv-ity and pseudoallergy short of detection of drug-specificfound a significantly higher prevalence of intolerance to

Dexferrum� (9.7%) than to InFeD� (5.4%). Even among immunoglobulins [22]. Various forms of pseudoallergyexist. Atopy increases risk of pseudoallergic events suchpatients exposed to only one brand of iron dextran, the

reaction rate was significantly higher following Dexfer- as radiocontrast reactions [23], probably due to fragileor unstable mast cells with increased “releaseability” onrum�. These results are consistent with a previous report

and the suggestion that Dexferrum� iron dextran may exposure to various stimuli [24]. Multiple drug allergypatients often have increased sensitivity to many chemi-be more allergenic than InFeD� [5].

In contrast to iron dextran, SFGC has a single ligand cally dissimilar drugs [25, 26]. Finally, some individualshave repeated reactions (including to placebo) presum-for each of the five available iron oxide coordinating

positions—one gluconate and four sucrose molecules— ably based on classically conditioned learned behavior[27]. We hypothesized that reactions following SFGCresulting in an iron oxide polymer rather than iron oxide

crystals. Additionally, SFGC contains no dextrans, which would reflect pseudoallergy and be characterized by alow incidence similar to placebo, insignificant mast cellmay well account for its favorable adverse effect profile

compared to iron dextran preparations. activation, and would segregate to patients with historiesof other drug allergies.In this study, SFGC was remarkably well tolerated,

with 140 of 143 iron dextran-sensitive patients receiving We first assessed whether SFGC is allergenic by blinded,subjective identification of suspected allergic reactionsSFGC without incident. Ninety-two of these patients en-

tered the open label trial and 91 received repeated doses based on a temporal relationship to drug administration.Among all patients, suspected allergic reactions follow-of SFGC without incident. Nevertheless, the incidence of

intolerance reactions to SFGC among the dextran-sensi- ing SFGC were low in incidence (0.5%) and not statisti-cally different from the rate following placebo (0.2%).tive patients was approximately sevenfold higher than

observed in the dextran-tolerant population (2.1 vs. These data both support the hypothesis and confirm thevalue of the placebo control study design.0.3%, P � 0.020). Similarly, the incidence of suspected

allergic reactions in the dextran-sensitive group was ap- Secondly, SFGC allergenicity was assessed by de-termining pre- and post-reaction serum mast cell-derivedproximately fivefold higher than observed in the dextran-

tolerant patients (2.1 vs. 0.4%, P � 0.010). The iron tryptase levels in patients with immediate-type suspectedallergic reactions. Tryptase, a marker of mast cell activa-dextran-sensitive patients also had a sevenfold higher

reaction rate following placebo administration (0.7 vs. tion, has not been used heretofore to assess drug aller-genicity in large scale studies. It appears that clinically-0.09%), suggesting that these patients may be predis-

posed to drug reactions. significant drug reactions are accompanied by at least a100%-rise in baseline serum tryptase levels [9, 28]. TheAlthough the proportion of iron dextran-intolerant

patients who were intolerant or experienced sensitivity appropriateness of this cut-off was confirmed by mea-surement of baseline and post-dialysis levels in 206 he-reactions with SFGC was statistically higher than iron

dextran-tolerant patients, the results of our study show modialysis patients as part of this study. Both patientswith significant mast cell degranulation following admin-that SFGC can be used safely in the vast majority of

these patients even if they have a prior history of iron istration of SFGC had a prior history of iron dextransensitivity. These reactions (back pain, and facial flush-dextran intolerance. It is important to note that while

hypersensitivity reactions following SFGC have been re- ing) could have resulted from pseudoallergy based onthe history of prior drug sensitivity, the lack of cross-ported anecdotally, serious reactions are extremely rare

and there have been no published reports of fatal reac- reactivity between SFGC and iron dextran, and the rare-ness of the events (0.08%). The single patient with ations [4, 7]. In contrast, at least 31 deaths apparently due

to hypersensitivity reactions to iron dextran occurred serious suspected anaphylactoid reaction to SFGC hadan elevated baseline tryptase level that did not increasebetween 1976 and 1996, and deaths also have been ob-

served rarely following iron sucrose injection [4, 11]. further after drug exposure, suggesting the possibility ofincreased total mast cell number or leakiness. This pa-Dextrans are known to elicit IgG and IgE antibody-

mediated reactions [12–20]. We hypothesized that SFGC tient also had a history of multiple drug sensitivities,

Coyne et al: Reactions to sodium ferric gluconate complex 223

including prior anaphylactoid reactions to iron dextran tions reflect pseudoallergy or host idiosyncrasy, and thatSFGC is not a true allergen. Therefore, caution shouldpreparations, which might be expected in a subject with

fragile mast cells. be used when administering SFGC or any drug to pa-tients with histories of multiple drug allergies.Lastly, we assessed SFGC allergenicity by use of logistic

regression analysis to identify any cross-reactivity betweenprior iron dextran sensitivity and reactions attributed to ACKNOWLEDGMENTSFGC. Patients with a history of iron dextran sensitivity This trial was supported by a grant from Watson Pharmaceuticals, Inc.had five- to tenfold higher rates of reaction to both pla-

Reprint requests to Daniel W. Coyne, M.D., Washington Universitycebo and SFGC. While the increase was statistically sig-School of Medicine, 660 South Euclid Avenue, Campus Box 8129, St.

nificant for SFGC, regression analysis did not identify Louis, Missouri 63110, USA.E-mail: Dcoyne@im.wustl.eduan interaction between SFGC and history of iron dextran

sensitivity. These apparently disparate findings addition-ally demonstrate the value of the placebo control study

APPENDIXdesign. The regression analysis takes into consideration

Additional participating investigatorsthat reactions following both placebo and SFGC were

Michael Austerlitz (Commerce, CA); Habib Azad (University ofsubstantially higher in patients with prior iron dextranCalifornia Irvine Medical Center); Raul Balagtas (Tampa, FL); Edwina

sensitivity. The regression analysis suggests that this in- Barnett (Charles Drew University); Mario Belledone (Rockville, MD);Jeffrey Berns (University of Pennsylvania); Anatole Besarab (Henrycreased reactivity is not drug-specific nor immunologi-Ford Health Systems); Jose Cangiano (San Juan Bautista School ofcally mediated, but represents host idiosyncrasy.Medicine, San Juan, PR); Michelle Crage (Woburn, MA); Laura Dember

Unexplained hypotension is commonly seen in dialysis (Boston University Medical Center); Devasmita Dev (Veterans Admin-istration North Texas Health Care); Jorge Diego (University of Miami,patients, and indeed was noted in our study followingMiami, FL); Albert Dreisbach (Tulane University School of Medicine);placebo infusions as well as after intravenous iron. SuchStephen Z. Fadem (Houston Kidney Center, Houston, TX); Danny

hypotensive episodes may represent hemodynamic insta- Fischer (Cincinnati, OH); Mary Gellens (St. Louis University MedicalCollege); Michael Germain (West Springfield, MA); Elias Ghafarybility during dialysis, an isolated manifestation of ana-(Gadsen, AL); Charles Graeber (Newington, CT); Mandeep Grewalphylaxis, or potentially evidence of iron toxicity. We(Chattanooga, TN); Lee Hamm (Tulane University School of Medi-

chose to study hypotension as a possible marker of ana- cine); Louisa Ho (Evanston, IL); Onyekachi Ifudu (State University ofNew York); Dennis Imperio (Sarasota, FL); Sam James (University ofphylaxis in our effort to catalog all possible adverseArizona Medical Center); Mark Kaplan (Nashville, TN); Ellie Kele-events that could be related to mast cell mediator release.pouris (Temple University School of Medicine); Ken Kleinman (Univer-

To our knowledge there is no evidence that iron toxicity sity of California at Los Angeles); Richard Lafayette (Stanford Univer-sity Medical Center); David Leehey (Loyola University Medical Center);leads to mast cell activation and thereby tryptase release.Susie Lew (George Washington University Medical Center); Jill Lind-The data from this study indicate that if intravenous ironberg (New Orleans, LA); Robert Lynn (Bronx Dialysis Center, New York

can induce mast cell activation, it must be a quite rare, City, NY); Thomas Marbury (Orlando, FL); Ronald Mars (University ofFlorida); Chamberlain Obialo (Morehouse School of Medicine, Atlanta,since tryptase levels did not change in 206 patients whoGA); Chika Oguagha (Nephrology Foundation of Brooklyn, New Yorkreceived SFGC without incident.City, NY); Jeffrey Petersen (Stanford University Medical Center); Wal-

The evaluation of drug safety in patients with serious ter Piering (Medical College of Wisconsin); Russell Pikus (Columbus,IN); Rasib Raja (Albert Einstein Medical Center); S. Noor Rahmanconcomitant illnesses and multiple medications is ideally(University of Texas at Houston, Houston, TX); Efrain Reisin (Louisi-performed using double-blind, placebo-controlled meth-ana State University Health Sciences Center); Steven Rosenblatt (San

ods. Hemodialysis patients frequently have hypotension, Antonio Kidney Diseases Center, San Antonio, TX); Jeff Sands (EmoryUniversity); Steve Schwab (Duke University Medical Center South);gastrointestinal symptoms, muscle cramps and other eventsAllan Schwartz (Hahnemann University Hospital); Andrea J. Shaerthat make determination of true study drug-related reac-(Medical University of South Carolina, Charleston, SC); Warren Sha-

tions exceedingly challenging. Additionally, the rates of piro (Brookdale, NY); Anupkumar Shetty (Henry Ford Health Sys-tems); Jerald Sigala (Orange, CA); Marcia Silver (Cleveland, OH);manifestation of allergic reactions, such as hypotension,Richard Sires (Ingelwood, CA); David Spiegel (University of Coloradopruritus, or rash, which occur in these patients during theHealth Sciences Center); Richard Swartz (University of Michigan); Kant

course of routine hemodialysis has not been established. Tucker (Simi Valley, CA); Nosratola Vaziri (University of CaliforniaIrvine Medical Center); Sergio Vega (West Palm Beach, FL); KevinOur prospective evaluation of SFGC in hemodialysisVitting (St. Joseph’s Hospital and Medical Center); Michael Walczykpatients in a double-blind, placebo-controlled study, fol-(Portland, OR); Harry Ward (Los Angeles, CA); Robert Weiss (Temple

lowed by continued open label use, provides a useful University); Duane Wombolt (Norfolk, VA); and Elizabeth Wrone(Palo Alto, CA).approach for assessing the true incidence, nature and

severity of reactions to SFGC.In summary, sodium ferric gluconate complex can be REFERENCES

safely administered to patients with prior iron dextran 1. Fishbane S, Wagner J: Sodium ferric gluconate complex in thetreatment of iron deficiency for patients on dialysis. Am J Kidneyallergies. Patients with prior iron dextran sensitivity andDis 37:879–883, 2001multiple drug allergies had a higher incidence of reac-

2. National Kidney Foundation: K/DOQI Clinical Practice Guide-tions to SFGC and placebo than iron dextran-tolerant lines for Anemia of Chronic Renal Disease, 2000. Am J Kidney

Dis 37(Supp 1):S182–S238, 2001patients. These results support the view that these reac-

Coyne et al: Reactions to sodium ferric gluconate complex224

3. Macdougall IC, Horl WH, Jacobs C, et al: European best practice 15. Paull J: A prospective study of dextran-induced anaphylactoidreactions in 5745 patients. Anaesth Intensive Care 15:163–167, 1987guidelines 6–8: Assessing and optimizing iron stores. Nephrol Dial

16. Vervloet D, Senft M, Dugue P, et al: Anaphylactic reactions toTransplant 15(Suppl 4):20–32, 2000modified fluid gelatins. J Allergy Clin Immunol 71:535–540, 19834. Faich G, Strobos J: Sodium ferric gluconate complex in sucrose:

17. Richter W: Hapten inhibition of passive antidextran dextran ana-Safer intravenous iron therapy than iron dextrans. Am J Kidneyphylaxis in guinea pigs. Role of molecular size in anaphylactogeni-Dis 33:464–470, 1999city and precipitability of dextran fractions. Int Arch Allergy Appl5. Fletes R, Lazarus JM, Gage J, Chertow GM: Suspected ironImmunol 41:826–844, 1971dextran-related adverse drug events in hemodialysis patients. Am

18. Ljungstrom KG, Renck H, Strandberg K, et al: Adverse reactionsJ Kidney Dis 37:743–749, 2001to dextran in Sweden 1970–1979. Acta Chir Scand 149:253–262,6. Horl WH: Should we still use iron dextran in hemodialysis pa-1983tients? Am J Kidney Dis 37:859–861, 2001

19. Hedin H, Richter W, Ring J: Dextran-induced anaphylactoid7. Michael B, Coyne DW, Fishbane S, et al: Sodium ferric gluconatereactions in man: Role of dextran reactive antibodies. Int Archcomplex in hemodialysis patients: adverse reactions compared toAllergy Appl Immunol 52:145–159, 1976placebo and iron dextran. Kidney Int 61:1830–1839, 2002

20. Renck H, Ljungstrom KG, Hedin H, Richter W: Prevention of8. Schwartz LB, Bradford TR, Rouse C, et al: Development ofdextran-induced anaphylactic reactions by hapten inhibition. III.a new, more sensitive immunoassay for human tryptase: Use inA Scandinavian multicenter study on the effects of 20 ml dextransystemic anaphylaxis. J Clin Immunol 14:190–204, 19941, 15%, administered before dextran 70 or dextran 40. Acta Chir9. Schwartz LB, Yunginger JW, Miller J, et al: Time course ofScand 149:355–360, 1983appearance and disappearance of human mast cell tryptase in the

21. Pradal M, Vervloet D: Drug Reactions, in Allergy and Allergiccirculation after anaphylaxis. J Clin Invest 83:1551–1555, 1989Diseases, edited by Kay A, Oxford, Blackwell Science, 1997, pp10. Schwartz LB: Laboratory assessment of immediate hypersensitiv-1671–1692ity and anaphylaxis: Utilization of tryptase as a marker of mast 22. Adkinson N: Drug Allergy, in Allergy: Principles and Practice (5thcell dependent events, in Immunology and Allergy Clinics of North ed), edited by Middelton E, St. Louis, Mosby, 1998, pp 1212–1224America: Diagnostic Laboratory Immunology, edited by Huston 23. Lieberman P: Anaphylactoid reactions to radiocontrast material.

D, Philadelphia, Saunders, 1994, pp 339–350 Clin Rev Allergy 9:319–338, 199111. U.S. Food and Drug Administration: Approval Package for 21– 24. Lichtenstein LM, MacGlashan DW Jr: The concept of basophil

135. Venofer (iron sucrose injection). Memorandum on Pre-45 releasability. J Allergy Clin Immunol 77:291–294, 1986Day Filing Meeting for NDA 21–135 for iron sucrose injection. 25. Asero R: Detection of patients with multiple drug allergy syn-Attachment 1. p. 44. Available on request from Freedom of Infor- drome by elective tolerance tests. Ann Allergy Asthma Immunolmation Officer, HFD-205, Center for Drug Evaluation and Re- 80:185–188, 1998search. 5600 Fishers Lane, Rockville, MD 20857, 2000 26. Sullivan TJ: Management of patients allergic to antimicrobial

12. Bailey G, Strub RL, Klein RC, Salvaggio J: Dextran-induced drugs. Allergy Proc 12:361–364, 1991anaphylaxis. JAMA 200:889–891, 1967 27. Siegel S: Multiple chemical sensitivity as a conditional response.

13. Ring J, Messmer K: Incidence and severity of anaphylactoid reac- Toxicol Ind Health 15:323–330, 1999tions to colloid volume substitutes. Lancet 1:466–469, 1977 28. Schwartz LB, Metcalfe DD, Miller JS, et al: Tryptase levels as

14. Ring J: Anaphylactoid reactions to intravenous solutions used for an indicator of mast-cell activation in systemic anaphylaxis andmastocytosis. N Engl J Med 316:1622–1626, 1987volume substitution. Clin Rev Allergy 9:397–414, 1991