Nephrol Dial Transplant (2005) 20: 1443–1449
`doi:10.1093/ndt/gfh820
`Advance Access publication 26 April 2005
`
`Original Article
`
`Hypersensitivity reactions and deaths associated with intravenous
`iron preparations
`
`George R. Bailie1,2,3, John A. Clark4, Christi E. Lane4 and Peter L. Lane5
`
`1Albany Nephrology Pharmacy (ANephRx) Group, Albany, NY, 2Nephrology Pharmacy Associates, Inc.,
`Ann Arbor, MI, 3Renal Research Institute, LLC, New York, NY, 4Galt Associates, Blue Bell, PA and
`5Luitpold Pharmaceuticals, Inc., Norristown, PA, USA
`
`Downloaded from
`
`http://ndt.oxfordjournals.org/
`
` by guest on March 11, 2016
`
`Abstract
`iron therapy is an accepted
`Background. Parenteral
`adjunctive management of anaemia in kidney disease.
`Newer agents may have fewer severe hypersensitivity
`adverse events (AE) compared with iron dextrans (ID).
`The rate of type 1 AE to iron sucrose (IS) and sodium
`ferric gluconate (SFG) relative to ID is unclear. We
`used the US Food and Drug Administration’s Freedom
`of Information (FOI) surveillance database to com-
`pare the type 1 AE profiles for the three intravenous
`iron preparations available in the United States.
`Methods. We tabulated reports received by the FOI
`database between January 1997 and September 2002,
`and calculated 100 mg dose equivalents for the treated
`population for each agent. We developed four clinical
`categories describing hypersensitivity AE (anaphylaxis,
`anaphylactoid reaction, urticaria and angioedema)
`and an algorithm describing anaphylaxis, for specific
`analyses.
`Results. All-event reporting rates were 29.2, 10.5 and
`4.2 reports/million 100 mg dose equivalents, while
`all-fatal-event reporting rates were 1.4, 0.6 and 0.0
`reports/million 100 mg dose equivalents for ID, SFG
`and IS, respectively. ID had the highest reporting rates
`in all four clinical categories and the anaphylaxis
`algorithm. SFG had intermediate reporting rates for
`urticaria, anaphylactoid reaction and the anaphylaxis
`algorithm, and a zero reporting rate for the anaphyl-
`axis clinical category. IS had either the lowest or a zero
`reporting rate in all clinical categories/algorithm.
`Conclusions. These findings confirm a higher risk for
`AE, especially serious type 1 reactions, with ID therapy
`than with newer intravenous iron products and also
`
`Correspondence and offprint requests to: George R. Bailie, PharmD,
`PhD, Albany College of Pharmacy, 106 New Scotland Avenue,
`Albany, NY 12208, USA. Email: bailieg@acp.edu
`The authors wish it to be known that P.L. Lane died before this
`manuscript was completed.
`
`suggest that IS carries the lowest risk for hypersensi-
`tivity reactions.
`
`Keywords: hypersensitivity reactions;
`intravenous iron; iron dextran; iron sucrose;
`sodium ferric gluconate; type 1 reactions
`
`Introduction
`
`Iron dextran has been available in the United States
`for over four decades [1] and in recent years, sales
`of intravenous iron therapy have increased steadily.
`The majority of
`this increased use has coincided
`with an increasing awareness of
`the need to use
`iron in combination with erythropoietic agents for
`optimal management of the anaemia of chronic kidney
`disease [K-DOQI update 2000, available at http://
`www.kidney.org/professionals/kdoqi/guidelines_updates/
`doqi_uptoc.html#an].
`Three intravenous iron preparations are currently
`approved for use in the US: iron dextran (InFeDÕ,
`Watson Pharma, Inc.; DexferrumÕ, American Regent,
`Inc.), sodium ferric gluconate complex in sucrose
`(FerrlecitÕ; Watson Pharma, Inc.) and iron sucrose
`(VenoferÕ; American Regent, Inc.). Serious type 1
`allergic reactions may occur more often after the
`administration of iron dextran than after the other
`two preparations, and are more often associated with
`fatal and life-threatening outcomes. The incidence of
`post-iron dextran immediate hypersensitivity reactions
`has been estimated as 1.1–3.2/100 treated population
`[2–5] while the case fatality proportion for post-iron
`dextran allergic episodes has been calculated as
`15.8% [1]. The risk for morbidity or mortality, plus
`an ongoing suboptimal management of anaemia in
`chronic kidney disease and end-stage renal disease, may
`have resulted in an inadequate therapeutic approach
`to anaemia. For example, recent data (2003 Annual
`
`ß The Author [2005]. Published by Oxford University Press on behalf of ERA-EDTA. All rights reserved.
`For Permissions, please email: journals.permissions@oupjournals.org
`
`Pharmacosmos A/S v. Luitpold Ex. Pharmaceuticals, Inc., IPR2015-01490
`
`Luitpold Pharmaceuticals, Inc., Ex. 2045, P. 1
`
`

`
`Downloaded from
`
`http://ndt.oxfordjournals.org/
`
` by guest on March 11, 2016
`
`1444
`
`Report) from the Centers for Medicare and Medicaid
`Services indicate that only 64% of haemodialysis
`patients receive intravenous iron [http://www.cms.hhs.
`gov/esrd/1.asp#9].
`Data to fully assess the safety profiles of intravenous
`iron products have been difficult to acquire. Clinical
`study designs are limited by exclusionary patient
`entry criteria, small numbers of exposed subjects and
`short durations of treatment [6]. Consequently, there
`are numerous examples of important adverse events
`(AE) that were not seen in clinical trials, but that
`were discovered in reporting systems during the post-
`marketing phase [7]. In addition to these issues, serious
`type 1 reactions are rare, which makes case ascertain-
`ment difficult even in large databases. Researchers
`who are presented with these kinds of methodological
`challenges frequently rely on surveillance databases
`rather than formal epidemiological studies to pro-
`vide risk clarification [1,8–10]. Although surveillance
`data are subject to various reporting biases, careful
`reviews of AE reporting trends that take into account
`existing biological
`and epidemiological
`evidence
`have become well-established methods in the field of
`pharmacovigilance [11].
`This study used a publicly available source, the
`Freedom of Information (FOI) surveillance data-
`base administered by the US Food and Drug
`Administration (FDA), together with market research
`data, to review the AE profiles of intravenous iron
`preparations available in the US. The objectives of the
`study were to describe the recent use of intravenous
`iron products in the US, to create definitions for
`analytical tools such as AE groupings, clinical cate-
`gories and an algorithm, and to examine AE reporting
`rates (RRs) and proportions for clinical relevance.
`
`Subjects and methods
`
`Data Sources
`
`The FOI Database is released to the public on a quarterly
`basis by the FDA and consists of two levels. The first contains
`electronic abstractions of individual patient adverse event
`(AE) reports that are forwarded to the FDA directly or via
`manufacturers following the approval of a product in the
`US [9]. The second FOI Database level contains the actual
`source documents (MedWatch forms) that were sent to the
`FDA. Data used for this study consisted entirely of the
`abstracted FOI electronic database and was obtained from
`a data vendor (Galt Associates, Sterling, VA, USA). All
`AE that are reported to the FOI Database are coded by
`the FDA using a standard AE terminology dictionary
`[Medical Dictionary for Regulatory Activities (MedDRAÕ)]
`composed of standardized descriptors called preferred terms
`[http://www.meddramsso.com/NewWeb2003/index.htm].
`MedDRAÕ terminology is the international medical termi-
`nology developed under the auspices of the International
`Conference on Harmonization of Technical Requirements for
`Registration of Pharmaceuticals for Human Use. MedDRAÕ
`is a registered trademark of the International Federation of
`Pharmaceutical Manufacturers Associations. Each report in
`
`G. R. Bailie et al.
`the FOI Database is associated with one or more MedDRAÕ
`preferred terms that are listed in a Reactions File. The FOI
`Database Reactions File can be linked to information about
`patient demographics, report sources, drug therapies, dates of
`therapeutic administration and patient outcomes [10].
`AE reports were included in this study if they (a) listed
`one or more of
`four intravenous iron trade names or
`their corresponding generic names as either a suspect or
`concomitant medication; (b) had FDA receipt dates between
`1 January 1997 and 30 September 2002 (the last available
`date at the time of study initiation); and (c) originated from
`a US healthcare practitioner. Searches for the study drugs
`(InFeDÕ, DexferrumÕ, FerrlecitÕ, VenoferÕ and their generic
`names) included exact spelling text strings as well as a variety
`of close misspellings.
`The exact indication for the use of parenteral iron, or the
`type of patient to whom it was administered, was not recorded
`in the FOI database.
`
`Methods used to obtain report counts
`The large number of preferred terms in the MedDRAÕ
`dictionary (currently over 15 000) can lead to such diffuse
`coding of clinically similar events that reporting patterns may
`be obscured. This effect can be addressed by identifying
`reports through the use of clinical categories that contain
`multiple MedDRAÕ preferred terms or through the use of
`single or grouped MedDRAÕ preferred terms that are applied
`in logical combinations (i.e. as clinical algorithms).
`Algorithms are particularly useful in locating reports of
`syndromes that are known to possess multiple clinical criteria
`in combination. For example, anaphylaxis could have been
`coded using only a single MedDRAÕ preferred term such as
`‘Anaphylaxis’, but could also have been coded using two
`MedDRAÕ preferred terms that referred to different body
`systems, such as ‘Hypotension’ and ‘Urticaria’. Thus, in addi-
`tion to the clinical category for anaphylaxis, we also defined
`an anaphylaxis algorithm that identified any report in which
`there was either a single MedDRAÕ preferred term indicative
`of anaphylaxis or a combination of the typical clinical conse-
`quences of anaphylaxis plus either of two skin indicators
`for histamine release (urticaria and skin angioedema).
`We used MedDRAÕ preferred-term coding to triage
`reports into seven analytical counts: (i) one clinical category
`that contained a single MedDRAÕ preferred term (anaphy-
`lactoid reaction); (ii) three clinical categories that contained
`between two and nine MedDRAÕ preferred terms (anaphy-
`laxis, upper airway angioedema and urticaria); (iii) one
`anaphylaxis algorithm that used several clinical categories
`in a logical sequence to identify reports of anaphylaxis; and
`(iv) two large summary categories of reports that contained
`any reported MedDRAÕ preferred term referring to a medical
`event (all reports and all fatal reports).
`Table 1 describes our clinical categories of type 1 reactions
`(anaphylactoid reaction, anaphylaxis, upper airway angio-
`edema and urticaria) and lists those MedDRA preferred
`terms used in the definitions. Table 2 describes the MedDRA
`preferred terms and rules used to define the anaphylaxis
`algorithm.
`This study followed the standard convention of totalling
`counts within a category or algorithm non-duplicatively
`(i.e. an AE report assigned to an analytical report count
`was always counted once for a given category, even if more
`
`Pharmacosmos A/S v. Luitpold Ex. Pharmaceuticals, Inc., IPR2015-01490
`
`Luitpold Pharmaceuticals, Inc., Ex. 2045, P. 2
`
`

`
`Hypersensitivity reactions to intravenous iron
`
`1445
`
`Table 1. Description and definition of clinical categories for type 1 reactions
`
`Clinical category
`
`Number of MedDRAÕ
`preferred terms
`
`Description
`
`Defined as the MedDRA preferred term ‘Anaphylactoid Reaction’
`Defined as the MedDRA preferred terms ‘Anaphylactic Reaction’ and
`‘Anaphylactic Shock’
`Definition includes oedema of the tongue, throat, pharynx and larynx
`Definition includes hives, urticaria and equivalent terms
`
`Downloaded from
`
`http://ndt.oxfordjournals.org/
`
` by guest on March 11, 2016
`
`Because of the difference in sizes of the units of supply for
`the three iron formulations, we normalized dosing to 100 mg
`dose equivalents. Thus, product exposure was defined as the
`number of 100 mg dose equivalents used in the US annually
`for each intravenous iron therapy. Data to perform this
`calculation was obtained from a market research vendor
`(IMS Health, Plymouth Meeting, PA, USA).
`
`Calculation of rates and proportions
`
`The US RR for the study interval was calculated for all
`events, all fatal events and each of the four clinical categories
`and anaphylaxis algorithm for each therapy by dividing
`the number of all reports, the number of fatal reports or
`the counts for each clinical category or algorithm by the
`number of 100 mg dose equivalents used in the study interval.
`The results were expressed as the number of AE reports/
`million 100 mg dose equivalents. The case fatality proportion
`for each clinical category and algorithm was calculated by
`dividing the fatal report count for that clinical category or
`algorithm by the total report count for that clinical category
`or algorithm.
`
`Results
`
`Exposure trend over time
`
`The total 100 mg dose equivalents in the US, per
`3 month period,
`for
`the three intravenous
`iron
`treatments increased steadily over the study interval,
`from 1.3 million in March 1997 to 3.6 million in
`March 2003 (Figure 1). Compared with 1997, 100 mg
`dose equivalents for all intravenous iron treatments
`for 2002 increased by 78.2%. There was an overall
`declining trend for iron dextran use since the fourth
`quarter of 1999, which coincided with the introduction
`of sodium ferric gluconate in mid-1999 and which
`accelerated following the introduction of iron sucrose
`into the US market in the fourth quarter of 2000.
`
`All-event and all-fatal-event reporting rates
`
`The all-event and all-fatal-event RRs for the three iron
`therapies are provided in Figure 2. The all-event RRs
`for iron dextran, sodium ferric gluconate and iron
`sucrose were 29.2, 10.5 and 4.2 reports/million 100 mg
`dose equivalents, respectively, while the all-fatal-event
`RRs were 1.4, 0.6 and 0.0 reports/million 100 mg dose
`equivalents, respectively.
`
`Anaphylactoid reaction
`Anaphylaxis
`
`Upper airway angioedema
`Urticaria
`
`1
`2
`
`5
`9
`
`Table 2. Anaphylaxis algorithm
`
`The anaphylaxis algorithm included reports with:
`A single code for anaphylaxis
`OR
`The combination of:
`A code for a clinical manifestation of systemic allergy
`Bronchospasm
`Circulatory collapse
`Dyspnoea or stridor
`Hypotension or decreased blood pressure
`Syncope or loss of consciousness
`Upper airway angioedema
`PLUS
`A skin indicator for histamine release
`Urticaria
`Angioedema
`
`than one MedDRAÕ preferred term that was used to define
`the category/algorithm was present in the report). In contrast,
`counts across categories or algorithms could be duplicative
`(i.e. a single AE report that was assigned to more than
`one category/algorithm was counted once for each such
`assignment).
`
`Estimation of exposure
`
`No sources are available that quantify the exact magnitude
`of parenteral iron doses administered to patients at any one
`time. Some data are available that suggest the magnitude of
`doses administered to haemodialysis patients. Intravenous
`iron tends to be used in one of two ways for the management
`of anaemia in chronic kidney disease and end-stage renal
`disease. In patients with iron deficiency, as defined by
`contemporary clinical practice guidelines [http://www.kidney.
`org/professionals/kdoqi/guidelines_updates/doqi_uptoc.
`html#an],
`it
`is recommended to administer intravenous
`iron in 1 g doses, repeated until the patient is deemed iron
`replete. The 1 g is given as ten 100 mg doses of iron dextran
`[12,13] or iron sucrose [14], or as eight 125 mg doses of sodium
`ferric gluconate [15],
`in each consecutive haemodialysis
`session. Following repletion, treated patients should receive
`maintenance iron. The amount given to each patient varies
`depending on iron utilization and ongoing iron losses,
`but typically averages 70 mg per week in haemodialysis
`patients [http://www.cms.hhs.gov/esrd/1.asp#9]. Of all intra-
`venous iron used, it is unclear how much is administered
`as repletion vs maintenance doses in end-stage renal disease
`and chronic kidney disease, or non-nephrology patients. The
`dosing data for other patient groups that receive parenteral
`iron is unknown. Thus, we arbitrarily attributed the same
`average dose to all groups of patients, whether haemodialysis
`or not.
`
`Pharmacosmos A/S v. Luitpold Ex. Pharmaceuticals, Inc., IPR2015-01490
`
`Luitpold Pharmaceuticals, Inc., Ex. 2045, P. 3
`
`

`
`Downloaded from
`
`http://ndt.oxfordjournals.org/
`
` by guest on March 11, 2016
`
`G. R. Bailie et al.
`
`Iron Sucrose
`Ferric Sodium Gluconate
`Iron Dextran
`All Injectable Iron Combined
`
`3500
`
`3000
`
`2500
`
`2000
`
`1500
`
`1000
`
`500
`
`0
`
`1446
`
`Sold (x 1,000)
`Number of Units
`
`Mar 03
`Dec 02
`Sep 02
`
`2
`n 0
`Ju
`
`Mar 02
`Dec 01
`Sep 01
`
`1
`n 0
`Ju
`
`Mar 01
`Dec 00
`Sep 00
`
`0
`n 0
`Ju
`
`Mar 00
`Dec 99
`Sep 99
`
`9
`n 9
`Ju
`
`Mar 99
`Dec 98
`Sep 98
`
`8
`n 9
`Ju
`
`Mar 98
`Dec 97
`Sep 97
`
`7
`n 9
`Ju
`
`Mar 97
`
`Quarter
`
`Fig. 1. Annual sales of 100 mg dose equivalents for three intravenous iron preparations in the US (January 1997–March 2003).
`
`All event reporting rate
`All fatal event reporting rate
`
`30
`
`25
`
`20
`
`15
`
`10
`
`5
`
`0
`
`dose equivalents
`
`Reports per million 100 mg
`
`Iron dextran
`
`Sodium ferric gluconate
`
`Iron sucrose
`
`Fig. 2. Six year all-event and all-fatal-event RRs for three intravenous iron preparations in the US (January 1997–December 2002).
`
`Clinical category reporting rates
`
`Algorithm reporting rate
`
`The RRs for four clinical categories indicative of
`type 1 allergy are presented in Figure 3. For the cate-
`gory ‘Urticaria’, the calculated RRs for iron dextran,
`sodium ferric gluconate and iron sucrose were 2.1,
`0.8 and 0.32 reports/million 100 mg dose equivalents,
`respectively. For the category ‘Anaphylactoid Reaction’,
`the corresponding RRs were 0.87, 0.46 and 0.0
`reports/million 100 mg dose equivalents, respectively.
`The RR for iron dextran for ‘Anaphylaxis’ (n¼ 20) and
`‘Upper Airway Angioedema’ (n¼ 17) were 0.6/million
`100 mg dose equivalents and 0.87/million 100 mg dose
`equivalents, respectively. Sodium ferric gluconate had
`RR of 0.46 and 0.0/million 100 mg dose equivalents
`for
`‘Anaphylactoid Reaction’ and ‘Anaphylaxis’,
`respectively. There were no reports in these latter
`categories for iron sucrose.
`
`The RR for the anaphylaxis algorithm is presented
`in Figure 4. The RR for intravenous iron dextran
`(3.1 reports/100 mg equivalents of therapy) was the
`highest, followed by that for sodium ferric gluconate
`(0.69 reports/100 mg equivalents of therapy) and iron
`sucrose (0.32/100 mg equivalents of therapy).
`
`Case fatality proportions
`
`Case fatality proportions were calculated for the
`four clinical categories and the anaphylaxis algorithm
`(Table 3). The iron dextran anaphylaxis category
`exhibited a case fatality proportion of 40.0%, while
`the proportion for the anaphylaxis algorithm was
`10.0%. Intravenous iron dextran also exhibited case
`fatality proportions of 10.7% for ‘Anaphylactoid
`
`Pharmacosmos A/S v. Luitpold Ex. Pharmaceuticals, Inc., IPR2015-01490
`
`Luitpold Pharmaceuticals, Inc., Ex. 2045, P. 4
`
`

`
`Downloaded from
`
`http://ndt.oxfordjournals.org/
`
` by guest on March 11, 2016
`
`Hypersensitivity reactions to intravenous iron
`2.5
`
`1447
`
`Iron dextran
`Sodium ferric gluconate
`Iron sucrose
`
`2.0
`
`1.5
`
`1.0
`
`0.5
`
`0.0
`
`Reports per million 100 mg dose equivalents
`
`Urticaria
`
`Upper Airway Angioedema
`
`Anaphylactoid Reaction
`
`Anaphylaxis
`
`Fig. 3. Clinical category RRs for three intravenous iron therapies for type 1 reactions.
`
`Clinical Category
`
`3.5
`
`3.0
`
`2.5
`
`2.0
`
`1.5
`
`1.0
`
`0.5
`
`0.0
`
`Reports per million 100 mg dose equivalents
`
`Iron dextran
`
`Sodium ferric gluconate
`
`Iron sucrose
`
`Fig. 4. Anaphylaxis algorithm RRs for three intravenous iron preparations.
`
`Reaction’ and 17.6% for ‘Upper Airway Angioedema’.
`The case fatality proportions for sodium ferric gluco-
`nate and iron sucrose were either zero or could not be
`calculated (i.e. there had been no fatal or non-fatal
`reports to the FDA).
`
`Discussion
`
`Sales of intravenous iron treatments available in the
`US increased substantially between 1997 and 2002, and
`
`reflect two opposing trends: an overall decrease in the
`use of intravenous iron dextran and the increasing use
`of the two newer preparations. The results of both this
`and prior studies suggest that these trends are at least
`partially attributable to the AE profile of iron dextran
`products. Our data can be compared, to some extent,
`with previous findings, even though it is impossible to
`determine if the data from the FDA are drawn from the
`general population or from dialysis patients or both.
`Faich and Strobos [1] used surveillance report data to
`investigate two major concerns that have been raised
`
`Pharmacosmos A/S v. Luitpold Ex. Pharmaceuticals, Inc., IPR2015-01490
`
`Luitpold Pharmaceuticals, Inc., Ex. 2045, P. 5
`
`

`
`Downloaded from
`
`http://ndt.oxfordjournals.org/
`
` by guest on March 11, 2016
`
`1448
`
`Table 3. Case fatality proportions for four clinical categories and
`the anaphylaxis algorithma
`
`Clinical category/
`algorithm
`
`Iron
`dextran (%)
`
`Sodium ferric
`gluconate (%)
`
`Iron
`sucrose (%)
`
`Clinical category
`Anaphylaxis
`Anaphylactoid
`reaction
`Upper airway
`angioedema
`Urticaria
`
`40.0
`10.7
`
`17.6
`
`0.0
`
`Anaphylaxis algorithm 10.0%
`
`N/A
`0.0
`
`N/A
`
`0.0
`
`0.0%
`
`N/A
`N/A
`
`N/A
`
`0.0
`
`0.0%
`
`aCategory or algorithm results for which no cases (fatal or non-
`fatal) were reported are indicated as ‘N/A’.
`
`about intravenous iron dextran preparations, namely
`an excessive incidence rate (or, in some instances, RR)
`for serious type 1 allergic reactions and an increased
`case fatality proportion among patients experiencing
`such reactions. They calculated a RR for key events of
`8.7 and 3.3 allergy episodes/million doses for iron
`dextran and sodium ferric gluconate, respectively (iron
`sucrose was not then available in the US). Our findings
`are consistent with those data, since each of the clinical
`categories for type 1 AE that we examined (anaphy-
`laxis, anaphylactoid reaction, urticaria and upper
`airway angioedema) showed a substantially higher
`RR for iron dextran than for either sodium ferric
`gluconate or
`iron sucrose.
`In addition, marked
`differences in the RRs for any AE and any fatal AE
`were also observed for iron dextran (about three to four
`times higher than the other two intravenous iron
`products combined). The same discrepancy was also
`seen with the anaphylaxis algorithm that was developed
`for this study. In these latter comparisons, the RR for
`iron dextran exceeded the average RR of the other two
`products by >6-fold.
`A more recent study also used data derived from
`the FDA to compare the reaction rates of high and
`low molecular weight formulations of iron dextran
`and sodium ferric gluconate [5]. That study examined
`all reported adverse effects, subsequently subdividing
`these into life-threatening and non-life-threatening
`events. Interestingly, the non-life-threatening group
`included some type 1 events, including allergic reac-
`tions, facial oedema, pruritis and urticaria, and some
`additional events that may be related to an allergic
`response,
`including hypotension and dyspnoea. The
`authors standardized the reaction rates per 100 mg
`dose equivalents. They reported that total AE were
`significantly increased with the use of high molecular
`weight iron dextran [odds ratio (OR): 5.5] and sodium
`ferric gluconate (OR: 6.2) compared with low molec-
`ular weight iron dextran and that the risk of death
`was substantially higher in the high molecular weight
`iron dextran recipients. An analysis of their data
`indicates that there were also substantial differences
`in life-threatening and non-life-threatening reactions
`between the products. For sodium ferric gluconate,
`
`G. R. Bailie et al.
`
`high and low molecular weight iron dextran, the life-
`threatening reaction rates were 3.69, 12.46 and 3.89 per
`million 100 mg dose equivalents, respectively, and the
`non-life-threatening reaction rates were 123.8, 114.1
`and 16.5 per million 100 mg doses, respectively. The
`reported death rates were 0.6 and 2.2 per million 100 mg
`doses of high and low molecular weight iron dextran,
`respectively (or 1.0 for the dextrans combined), and 0.92
`per million 100 mg doses of sodium ferric gluconate.
`While the Chertow et al. [5] study did not examine
`iron sucrose, nevertheless, some comparisons may
`be drawn between it and our current findings. In
`our study, the all-event RRs for iron dextran, sodium
`ferric gluconate and iron sucrose were 29.2, 10.5 and
`4.2 reports/100 mg equivalents of therapy, respectively,
`while the all-fatal-event RRs were 1.4, 0.6 and
`0.0 reports/100 mg equivalents of therapy, respectively.
`Thus, our death rates for iron dextran (1.4 vs 1.0) and
`sodium ferric gluconate (0.6 vs. 0.92) were very similar,
`which tends to provide some external validity to our
`data.
`The analysis of case fatality proportion is also
`consistent with previous studies [1,16]. Our study indi-
`cated that case fatality was confined to the iron dextran
`group. Using the anaphylaxis algorithm (which is more
`comparable to methods used in prior studies), the post-
`iron dextran case fatality proportion for anaphylaxis
`was calculated to be 10.0%, which is of the same order
`of magnitude as the 15.8% case fatality proportion that
`was calculated by Faich and Strobos [1].
`Although experience with iron sucrose was limited
`in this study to the 2 year interval from the fourth
`quarter of 2000 to the third quarter of 2002, it should
`be noted that it had a zero RR for all but one of the
`four allergic categories (the one non-zero category was
`‘Urticaria’; n¼ 1; RR ¼ 0.3 reports/100 mg equivalents
`of therapy). Iron sucrose also exhibited the lowest
`RR for the anaphylactic algorithm and demonstrated
`a lower RR compared with sodium ferric gluconate,
`except for those categories that showed a zero RR for
`both agents (upper airway angioedema and anaphy-
`laxis). Thus, we hypothesize that the rate of type 1
`allergic reactions is lower for iron sucrose than for
`sodium ferric gluconate and both are lower than for
`iron dextran, and that additional studies may be able
`to confirm that there is a potential difference in the
`incidence rates for allergic reactions between these two
`intravenous iron therapies.
`We also attempted to determine if disproportions
`exist among the AE profiles for the three intravenous
`iron agents that can be demonstrated through the
`categorization of all AE terminology into clinically
`relevant groupings.
`In general,
`the proportionate
`contribution of different kinds of AE to the overall
`reported AE profiles for these three therapies were
`similar. Although the results did indicate several
`possible areas of difference (e.g. superficial vascular
`reactivity), the clinical relevance of such observations
`is difficult to determine.
`Reported AE data are subject to a variety of biases
`for which adjustment is usually not possible. One major
`
`Pharmacosmos A/S v. Luitpold Ex. Pharmaceuticals, Inc., IPR2015-01490
`
`Luitpold Pharmaceuticals, Inc., Ex. 2045, P. 6
`
`

`
`Downloaded from
`
`http://ndt.oxfordjournals.org/
`
` by guest on March 11, 2016
`
`Hypersensitivity reactions to intravenous iron
`
`bias that affects many RR comparisons, the Weber
`(or new drug reporting) effect, occurs because new
`products receive more scrutiny and reporting activity
`within a few years following their introduction into a
`marketplace [17]. However, the Weber effect would be
`expected to produce the strongest upwards bias on the
`most recent entrants into a market (in our study, iron
`sucrose followed by sodium ferric gluconate). Since
`intravenous iron dextran was reintroduced into the
`US 5 years prior to the start of our study, the new drug
`reporting effect is an unlikely explanation for our
`results. For this same reason, it remains possible that
`iron dextran may have been subject to underreporting
`and that we have underestimated adverse reactions to
`iron dextran.
`Commonly encountered biases in surveillance data
`include differential use by indication, protopathic bias
`and publicity. Differences in the use of healthcare
`products in distinct target populations lead to differ-
`ential AE profiles as a result of the forwarding of
`background events that are related to patient char-
`acteristics. Protopathic bias results from the adminis-
`tration of an agent for premonitory symptoms, which
`then evolve over days to a few months into a serious,
`diagnosable event. Further, publicity can dispropor-
`tionately affect the RR of one product vs similar prod-
`ucts by stimulating product-specific reporter activity.
`However, we found no evidence to indicate that any of
`these biases could explain the marked and/or consistent
`increases in comparative RRs that were seen in the iron
`dextran and sodium ferric gluconate groups in our
`study.
`As have previous authors, we conclude that the
`magnitude of
`reporting discrepancy exhibited by
`iron dextran vis-a`-vis similarly used products is best
`explained as a comparative increase in post-treatment
`risk, which is manifested as an increased RR for all
`and fatal AE and all and fatal type 1 allergic reactions.
`The smaller, but consistent, differences between sodium
`ferric gluconate and iron sucrose in our study suggest
`that iron sucrose products may represent the best
`allergic profile seen to date for those intravenous iron
`products that have been studied. Additional incidence-
`based designs would be useful
`in addressing and
`clarifying this latter observation.
`
`Acknowledgements. Data from this
`study were presented in
`abstract form at the American Society of Nephrology meeting,
`San Diego, 15 November 2003. This study was supported by a
`grant from Luitpold Pharmaceuticals Inc and American Regent Inc.
`
`Conflict of interest statement. This study was funded by a grant
`from Luitpold Pharmaceuticals, Inc. G.R.B. is a member of the
`Speakers’ Bureau for American Regent, Inc. (which is affiliated
`with Luitpold Pharmaceuticals), which distributes iron sucrose in
`the United States.
`
`1449
`
`References
`
`1. Faich G, Strobos J. Sodium ferric gluconate complex in
`sucrose: safer intravenous iron therapy than iron dextrans.
`Am J Kidney Dis 1999; 33: 464–470
`Intravenous
`2. Hamstra RD, Block MH,
`Schocket AL.
`JAMA 1980;
`iron dextran in clinical medicine.
`243:
`1726–1731
`3. Woodman J, Shaw RJ, Shipman AJ et al. A surveil-
`lance programme on Imferon. Pharmaceut Med 1987; 1:
`289–296
`4. Fishbane S, Ungureanu V-D, Maesaka JK et al. The safety of
`intravenous iron dextran in hemodialysis patients. Am J Kidney
`Dis 1996; 28: 529–534
`5. Chertow GM, Mason PD, Vaage-Nilsen O, Ahlmen J. On the
`iron formulations. Nephrol Dial
`relative safety of parenteral
`Transplant 2004; 19: 1571–1575
`6. Faich GA. Adverse drug reaction surveillance. N Engl J Med
`1986; 314: 1589–1592
`7. Faich GA, Lawson DH, Tilson HH, Walker A. Clinical trials
`are not enough. Clin Res Drug Dev 1987; 1: 75–78
`8. Szarfman A, Machado SG, O’Neill RT. Use of screening
`algorithms and computer systems to efficiently signal higher-
`than-expected combinations of drugs and events in the US
`FDA’s Spontaneous Reports Database. Drug Safety 2002; 25:
`381–392
`reports on
`9. Omar MA, Wilson JP. FDA adverse event
`statin-associated rhabdomyolysis. Ann Pharmacother 2002; 36:
`288–295
`10. Woods SW, Martin A, Spector SG et al. Effects of
`development on olanzapine-associated adverse events. J Am
`Acad Child Adoles Psych 2002; 41: 1439–1446
`11. Clark JA, Klincewicz SL, Stang PE. Overview-spontaneous
`signaling. In: Mann RD, Andrews EB, eds. Pharmacovigilance.
`Wiley, Chichester: 2002; 259
`for InFeDÕ (iron dextran injection, USP),
`12. Package Insert
`February 2001
`13. Package Insert for DexferrumÕ (iron dextran injection, USP),
`1997
`14. Package Insert for VenoferÕ (iron sucrose injection, USP),
`May 2003
`15. Package Insert for FerrlecitÕ (sodium ferric gluconate complex
`in sucrose injection), November 2001
`16. Fletes R, Lazarus M, Gage J et al. Suspected iron dextran-
`related adverse drug effects in hemodialysis patients. Am J
`Kidney Dis 2001; 37: 743–749
`17. Weber JCP. Epidemiology of adverse reactions to nonsteroidal
`anti-inflammatory drugs. In: Rainslord KD, Velo EP, eds.
`Advances in Inflammation Research. Raven Press, New York:
`1984: 1–7
`
`Received for publication: 8.7.04
`Accepted in revised form: 11.3.05
`
`Pharmacosmos A/S v. Luitpold Ex. Pharmaceuticals, Inc., IPR2015-01490
`
`Luitpold Pharmaceuticals, Inc., Ex. 2045, P. 7