European Journal of Pharmaceutics and Biopharmaceutics 78 (2011) 480–491
`
`Contents lists available at ScienceDirect
`
`European Journal of Pharmaceutics and Biopharmaceutics
`
`j o u r n a l h o m e p a g e : w w w . e l s e v i e r . c o m / l o c a t e / e j p b
`
`Research paper
`A comparative study of the physicochemical properties of iron isomaltoside
`1000 (MonoferÒ), a new intravenous iron preparation and its clinical implications
`Markus R. Jahn a, Hans B. Andreasen b, Sören Fütterer a, Thomas Nawroth a, Volker Schünemann c, Ute Kolb d,
`Wolfgang Hofmeister e, Manuel Muñoz f, Klaus Bock g, Morten Meldal g, Peter Langguth a,⇑
`
`a Institute of Pharmacy and Biochemistry, Johannes-Gutenberg University, Mainz, Germany
`b Pharmacosmos A/S, Holbaek, Denmark
`c Department of Physics, Technical University Kaiserslautern, Kaiserslautern, Germany
`d Institute of Physical Chemistry, Johannes-Gutenberg University, Mainz, Germany
`e Institute of Geosciences, Johannes-Gutenberg University, Mainz, Germany
`f Transfusion Medicine, School of Medicine, University of Málaga, Málaga, Spain
`g Carlsberg Laboratory, Gamle Carlsberg Vej 10, Valby, Denmark
`
`a r t i c l e
`
`i n f o
`
`a b s t r a c t
`
`Article history:
`Received 14 December 2010
`Accepted in revised form 15 March 2011
`Available online 23 March 2011
`
`Keywords:
`Inorganic nanoparticles
`Colloids
`Stability
`Free iron
`Iron supplementation
`Release rate
`
`The treatment of iron deficiency anemia with polynuclear iron formulations is an established therapy in
`patients with chronic kidney disease but also in other disease areas like gastroenterology, cardiology,
`oncology, pre/post operatively and obstetrics’ and gynecology. Parenteral iron formulations represent
`colloidal systems in the lower nanometer size range which have traditionally been shown to consist of
`an iron core surrounded by a carbohydrate shell. In this publication, we for the first time describe the
`novel matrix structure of iron isomaltoside 1000 which differs from the traditional picture of an iron core
`surrounded by a carbohydrate. Despite some structural similarities between the different iron formula-
`tions, the products differ significantly in their physicochemical properties such as particle size, zeta
`potential, free and labile iron content, and release of iron in serum. This study compares the physiochem-
`ical properties of iron isomaltoside 1000 (MonoferÒ) with the currently available intravenous iron prep-
`arations and relates them to their biopharmaceutical properties and their approved clinical applications.
`The investigated products encompass low molecular weight iron dextran (CosmoFerÒ), sodium ferric glu-
`conate (FerrlecitÒ), iron sucrose (VenoferÒ), iron carboxymaltose (FerinjectÒ/InjectaferÒ), and ferumoxy-
`tol (FerahemeÒ) which are compared to iron isomaltoside 1000 (MonoferÒ). It is shown that significant
`and clinically relevant differences exist between sodium ferric gluconate and iron sucrose as labile iron
`formulations and iron dextran, iron carboxymaltose, ferumoxytol, and iron isomaltoside 1000 as stable
`polynuclear formulations. The differences exist in terms of their immunogenic potential, safety, and con-
`venience of use, the latter being expressed by the opportunity for high single-dose administration and
`short infusion times. Monofer is a new parenteral iron product with a very low immunogenic potential
`and a very low content of labile and free iron. This enables Monofer, as the only IV iron formulation,
`to be administered as a rapid high dose infusion in doses exceeding 1000 mg without the application
`of a test dose. This offers considerable dose flexibility, including the possibility of providing full iron
`repletion in a single infusion (one-dose iron repletion).
`
`Ó 2011 Elsevier B.V. All rights reserved.
`
`1. Introduction
`
`Parenteral iron therapy is today widely used for the treatment
`of iron deficiency anemia. Patients with chronic kidney disease
`(CKD) also frequently need treatment with parenteral iron prepa-
`rations in addition to erythropoietin stimulating agents [1]. For
`
`⇑ Corresponding author. Institute of Pharmacy and Biochemistry, Pharmaceutical
`
`Technology and Biopharmaceutics, Johannes-Gutenberg University Mainz, D-55099
`Mainz, Germany.Tel.: +4961313925746; fax: +49 6131 3925021.
`E-mail address: langguth@uni-mainz.de (P. Langguth).
`
`0939-6411/$ - see front matter Ó 2011 Elsevier B.V. All rights reserved.
`doi:10.1016/j.ejpb.2011.03.016
`
`renal failure patients on dialysis, the average iron requirements
`due to blood loss are equivalent to 1–3 g of elemental iron per year
`[2]. This can easily be accomplished by frequent low dose IV iron
`administrations, during the regular dialysis sessions.
`From initial, generalized use in nephrology parenteral iron ther-
`apy has spread in recent years to other disease areas; gastroenter-
`ology [3], cardiology [4,5], oncology [6], pre/post operatively [7],
`obstetrics’, and gynecology [8]. However, care providers in these
`segments have less frequent patient contact, resulting in an in-
`creased demand for convenient administration of large IV iron
`doses in one clinical session.
`
`PGR2020-00009
`Pharmacosmos A/S v. American Regent, Inc.
`Petitioner Ex. 1048 - Page 1
`
`

`

`M.R. Jahn et al. / European Journal of Pharmaceutics and Biopharmaceutics 78 (2011) 480–491
`
`481
`
`Historically, the first parenteral iron preparations were toxic,
`being administered as an iron oxyhydroxide complex. This prob-
`lem was circumvented with the introduction of compounds con-
`taining iron in a core surrounded by a carbohydrate shell [9]. The
`currently marketed parenteral iron preparations are considered
`equally efficacious but vary in molecular size, pharmacokinetics,
`and adverse reaction profiles. The intravenous iron agents cur-
`rently available include high molecular weight iron dextran
`low molecular weight iron dextran (CosmoferÒ,
`(DexferrumÒ),
`InfedÒ), sodium ferric gluconate (FerrlecitÒ), iron sucrose (VenoferÒ),
`iron carboxymaltose (FerinjectÒ/InjectaferÒ), and ferumoxytol (Fera-
`hemeÒ). High molecular weight iron dextran has been linked to an
`increased risk of anaphylaxis and anaphylactoid reactions, and it is
`not available in Europe [10–13]. Although this problem is very
`much reduced with low molecular weight iron dextran [10–13],
`there is still a test dose requirement and the infusion of larger
`doses is hampered by a 4–6 h infusion time. Sodium ferric gluco-
`nate and iron sucrose can only be used in moderate iron doses
`due to the relative weakness of the iron complex [14]. Two new
`parenteral iron compounds, iron carboxymaltose, and ferumoxytol
`were recently introduced in the EU and the US markets, respec-
`tively. The FDA failed to approve iron carboxymaltose for distribu-
`tion in the USA due to unexplained hypophosphatemia, an
`increased number of adverse cardiac events and an imbalance in
`death rates in the treatment arm compared to the control arm in
`different RCTs [15].
`Although more stable than sodium ferric gluconate and iron su-
`crose, the administration of iron carboxymaltose and ferumoxytol
`is still limited to a maximum total dose of 1000 mg and 510 mg,
`respectively.
`The newest IV iron agent Iron isomaltoside 1000 (MonoferÒ)
`(e.g., iron oligo isomaltoside (1000) as generic name) is developed
`and manufactured by Pharmacosmos in Denmark and was intro-
`duced in Europe in 2010. The carbohydrate isomaltoside 1000 is
`a pure linear chemical structure of repeating a1-6 linked glucose
`units, with an average size of 5.2 glucose units and an average
`molecular weight of 1000 Da, respectively. It is a nonbranched,
`nonanaphylactic carbohydrate [16,17], structurally different from
`branched polysaccharides used in iron dextran (Cosmofer).
`The production method and the short nonionic isomaltoside
`1000 allows for the construction of a special matrix-like structure
`with interchanging iron molecules and linear isomaltoside 1000
`oligomers. The resulting matrix contains about 10 iron molecules
`per one isomaltoside pentamer in a strongly bound structure that
`enables a controlled and slow release of bioavailable iron to iron-
`binding proteins with little risk of free iron toxicity [18,19]. This al-
`lows iron isomaltoside 1000 to be administered safely as a rapid
`high dose intravenous infusion or bolus injection offering consider-
`able dose flexibility, including the possibility of providing full iron
`repletion in a single infusion, the so-called one-dose iron repletion.
`This article introduces and compares physicochemical proper-
`ties of iron isomaltoside 1000 (MonoferÒ) with currently marketed
`iron formulations. In addition, this comparative study of polynu-
`clear iron formulations currently used in the treatment of anemic
`disorders includes perspectives on the relevance of these properties
`with respect to safety, efficacy, and convenience of administration.
`
`2. Materials and methods
`
`2.1. Materials
`
`mL in 2 mL ampoules; Teva, Mörfelden-Walldorf, Germany), iron
`isomaltoside 1000 (MonoferÒ, 100 mg Fe/mL in vials; Pharmacos-
`mos, Holbaek, Denmark), iron carboxymaltose (FerinjectÒ, 50 mg
`Fe/mL in 2 mL vials; Vifor, München, Germany), and ferumoxytol
`(FerahemeÒ, 30 mg Fe/mL, in 17 mL vials; AMAG Pharmaceuticals,
`Lexington, MA, USA) were obtained from a pharmacy or directly
`from the manufacturer. The FerrozineÒ reaction kit was purchased
`from Roche Diagnostics GmbH, Mannheim. All iron formulations
`were used immediately after opening the vial or kept at 4 °C under
`nitrogen. Solutions were made from double-distilled water.
`
`2.2. Gel permeation chromatography (GPC)
`
`The apparent average molecular weight was analyzed by gel
`permeation chromatography. Prior to sample analysis, the columns
`were calibrated using dextran standards. The dextran standards
`used for GPC calibration were the commercial available Pharma-
`cosmos standards and consisted of Dextran 25, 50, 80, 150, 270,
`and 410, respectively. The average molecular weights Mw and the
`peak average molecular weights MP were 23.000, 21.400; 48.600,
`43.500; 80.900, 66.700; 147.600, 123.600; 273.000, 196.300;
`409.800, 276.500 for Dextran 25, 50, 80, 150, 270, and 410, respec-
`tively. The standards have been evaluated against the Ph.EUR and
`USP dextran standards.
`The detector used in the GPC measurements is a VE 3580 RI
`detector (Viscotec). Data are collected and calculations are made
`using the Omnisec 4.1 software from Viscotec.
`The hydrodynamic diameter dh was calculated from the hydro-
`dynamic volume Vh = Mp |g|, where the intrinsic viscosity |g| is
`given by the Mark Houwink equation [20]
`jgj ¼ kMa
`v
`
`where Mav is the viscosity average molecular weight.
`
`2.3. Dynamic light scattering (DLS) and zeta potential
`
`The size distribution and zeta potential of the whole particle,
`which can include an iron hydroxide core plus a carbohydrate
`shell, was determined by DLS. The diluted samples (0.4 mg Fe/mL
`double-distilled and sterile filtered water) were measured using
`a Zetasizer Nano S (Malvern Instruments Ltd.; Worcestershire,
`UK) including a He–Ne Laser with a wavelength of k = 633 nm,
`which illuminated the samples and detects the scattering informa-
`tion at an angle of 173° (Noninvasive Back-scatter technology).
`Zeta potential measurements were performed at different pH val-
`ues by addition of 0.1 N HCl or NaOH, respectively. The data were
`analyzed with the firmware, Zetasizer Software DTSv612 yielding
`volume distribution data.
`
`2.4. Transmission electron microscopy (TEM)
`
`The dimension of the iron complex nanoparticle core was deter-
`mined with an EM420 transmission electron microscope (FEI/Phi-
`lips, Oregon, USA) at 120 kV. All preparations (1 mg Fe/mL,
`double-distilled water) were deposited onto a hydrophilized cup-
`ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiq
`
`per grid (300 mesh, Ø 3 mm) and were allowed to dry. The median
`
`of the geometrical diameter dg ¼
`
`ðd2s þ d2l Þ=2
`was determined
`(n = 50, ds = shortest dimension, dl = longest dimension).
`
`2.5. X-ray diffraction (XRD)
`
`Sodium ferric gluconate (FerrlecitÒ, 12.5 mg Fe/mL in 3.2 mL
`ampoules; Sanofi-Aventis, Frankfurt, Germany),
`iron sucrose
`(VenoferÒ, 20 mg Fe/mL in 5 mL ampoules; Vifor, München, Ger-
`many), low molecular weight iron dextran (CosmoFerÒ, 50 mg Fe/
`
`X-ray measurements of dried out solutions (30 °C) were per-
`formed with a XRD 3000 TT (Seifert, Ahrensburg, Germany) using
`Cu radiation (k = 1,54178 Å, 40 kV, 30 mA) in Bragg Brentano con-
`figuration (automatic divergence slit, angular rate 0,18°/min).
`
`PGR2020-00009
`Pharmacosmos A/S v. American Regent, Inc.
`Petitioner Ex. 1048 - Page 2
`
`

`

`482
`
`M.R. Jahn et al. / European Journal of Pharmaceutics and Biopharmaceutics 78 (2011) 480–491
`
`The particles mean diameter d was determined from the Scher-
`rer equation: d ¼ k
`b cos h, where b is the full width at half maximum
`of the peak at 36° 2h or 63° 2h.
`
`2.6. Mössbauer spectroscopy
`
`Mössbauer spectra of iron isomaltoside 1000 were recorded
`using a conventional spectrometer in the constant-acceleration
`mode. Isomer shifts are given relative to a-Fe at room temperature.
`The spectra were measured in a closed cycle cryostat (Cryo
`Industries of America, USA) at 150 K, equipped with permanent
`magnets. The magnetically split spectra were analyzed by least-
`square fits using Lorentzian line.
`
`2.7. Dialysable iron in buffer
`
`The amount of free iron was estimated using the dialysis tech-
`nique following pH adjustment of each iron dispersion to 7.5. A dis-
`persion volume containing 150 mg of iron (7.5 mL for LMW iron
`dextran, iron isomaltoside 1000, iron carboxymaltose and ferum-
`oxytol, respectively; 15.0 mL for sodium ferric gluconate, and
`11.25 mL for iron sucrose) was added resulting in concentrations
`of 20.0 mg Fe/mL for all iron products except for sodium ferric gluco-
`nate (10.0 mg Fe/mL) and iron sucrose (13.3 mg Fe/mL). Dilutions
`were made with water and 0.9% sodium chloride solution, respec-
`tively. The volumes were added inside the dialysis tubing (12,000–
`14,000 MWCO, Medicell, London, United Kingdom) and dialyzed
`for 24 h at 20 °C against 100 mL of water or sodium chloride solution,
`respectively. The total volume including the dialysis tube was
`107.5 mL. Dialysis of each iron agent was performed in duplicate.
`Iron in the surrounding solution was quantified using ICP-MS (induc-
`tively coupled plasma mass spectrometry). The ICP-MS instrument
`was a Thermo iCap 6000 ICP-OES (Thermo Scientific, Danmark).
`Iron is measured at 238,201 nm. The measurement is made ax-
`ial. Two-point (left–right) baseline correction and external linear
`calibration curve are used.
`The experiments were carried out at room temperature (20–
`24 °C). In order to evaluate the effect of pH on the level of dialysable,
`free iron above experiments were conducted also for the high dose IV
`iron formulations low molecular weight iron dextran, iron isomalt-
`ose, iron carboxymaltose, and ferumoxytol without pH adjustment.
`
`2.8. Acid soluble FeOOH
`
`The acidic hydrolysis of the FeOOH in [FeOOH]mLn was followed
`by quantifying the decreasing FeOOH concentration with UV-
`spectroscopy. The spectrometer used was a Lambda 20 (Perkin
`Elmer). Readings at 287.3 nm were made from a scan using data
`interval 1.0 nm, scan speed 249 nm/min, a slit with of 2.0 nm
`and a smooth with of 2.0 nm.
`The absorbance of iron agents (10 mg Fe/l, 10 mm path length)
`in 0.9% NaCl/0.2375 M HCl was measured at 287.3 nm from
`t = 0 min to t = 48 h, unless otherwise specified. Initial absorbance
`after dilution of the iron preparation at t  0 min was set to 1
`according to 100% undissolved FeOOH and all other measurements
`were normalized for this. Ln(normalized data) was plotted against
`time and fitted with a second degree polynomial (R2 > 0.990). Half-
`life t0.5 was calculated from fpolynomial(t0.5) = ln(0, 5).
`
`2.9. FerrozineÒ-detectable labile iron
`
`The dissolution of iron in serum was determined by the Ferro-
`zineÒ-method [21–24]. FerrozineÒ does not only detect the free
`iron but also the weakly bound iron in the complex and the trans-
`ferrin bound iron in serum, this determination allows one to quan-
`tify the in vitro labile iron pool of the investigated intravenous iron
`
`formulations. By this method, iron is detected in the ferrous as well
`as the ferric state as the ferric iron is reduced by ascorbate to fer-
`rous iron. Briefly, human serum was incubated with the iron prep-
`aration corresponding to theoretical doses of 200 mg and 500 mg
`leading to a serum concentration of 66.7 lg/mL and
`iron,
`166.7 lg/mL for a person with a body mass of 70 kg, respectively.
`These serum concentrations are consequences of a blood volume
`of 0.07 L per kg and a serum fraction of 60% of the blood volume,
`yielding a total serum volume of approx. 3 L [25]. The experiment
`was performed at room temperature (22 °C) in 1.5 mL Eppendorf-
`tubes. Incubations were done for 10 and 45 min, respectively.
`Thereafter, a 100 lL sample was analyzed by addition of 700 lL re-
`agent 1 containing thiourea (115 mM) and citric acid (200 mM),
`followed by addition of 350 lL of reagent 2 containing sodium
`ascorbate (150 mM) and FerrozineÒ (6 mM). Absorption of the
`complex was measured at 562 nm over approximately 60 min
`using a PERKIN ELMER Lambda 20 (Perkin Elmer Inc., Waltham,
`MA, USA) UV–Vis spectrometer. The obtained absorbance versus
`time curve was fitted to a second degree polynomial for each incu-
`bation period and the intercept with the ordinate was calculated to
`receive the comparable theoretical amount of FerrozineÒ-detect-
`able iron. The regression coefficient for the polynomial function
`was always better than 0.995. The labile iron pool was calculated
`by linear regression analysis of the obtained intercepts from curves
`at 10 min incubation and 45 min incubation.
`
`2.10. Elucidation of molecular structure of iron isomaltoside 1000
`
`Proton and carbon NMR spectra were obtained on a Bruker
`800 MHz NMR instrument as ca. 5% solutions in D2O at 300 K. Sig-
`nals were referenced to external dioxane.
`The iron isomaltoside 1000 formulation (10.3 mg) was dis-
`solved in D2O (600 lL). The sample was transferred to a 5 mm
`NMR-tube and the 13C NMR spectrum was recorded at 20 °C on a
`Bruker Avance 800 instrument at 201.12 MHz for carbon
`(799.96 MHz for proton), integrated and compared with the spec-
`trum of the oligosaccharide alone (7.3 mg) in D2O (600 lL) [26].
`Both samples were measured again and the signals integrated after
`addition of 2.24 mg of methyl b-maltoside as internal reference.
`Molecular modeling: First the isomaltodisaccharide was con-
`structed and an MD calculation using the modeling program,
`MOE (Molecular Operating Environment, Version 2009.10, Chemi-
`cal Computing Group Inc., Montreal, Canada), at 450 K, stepsize
`0,1 fs clearly showed a significant preference for the gt conforma-
`tion of C-5–C-6 bonds independent of the starting point. The O-1–
`C-6 of the glycosidic bond had a weak preference for a trans-
`arrangement and the orientation of the C-1–O-1 bond satisfied
`the exoanomeric effect.
`The isomaltoside pentamer (composed of 5 a1-6 linked glucose
`molecules) with glucitol at the reducing end was built from disac-
`charides in their preferred conformation and energy minimized.
`The molecule was soaked in water (eight layers) and molecular
`dynamics was performed at the above conditions corresponding
`to a period of 2 ns. The additive effect of the oligosaccharide repeat
`to stabilize the preferred conformation when compared to the
`disaccharide was significant. The resulting structure was re-soaked
`and was subjected to energy minimization in water.
`
`3. Results
`
`3.1. Overall particle size
`
`3.1.1. Gel permeation chromatography
`The distributions calculated from the GPC chromatograms of
`the iron preparations show homogenous distributions with the
`
`PGR2020-00009
`Pharmacosmos A/S v. American Regent, Inc.
`Petitioner Ex. 1048 - Page 3
`
`

`

`M.R. Jahn et al. / European Journal of Pharmaceutics and Biopharmaceutics 78 (2011) 480–491
`
`483
`
`exception of ferumoxytol and iron carboxymaltose which show
`additional smaller and larger diameter peaks (Fig. 1). The hydrody-
`namic diameters dh rise in the order iron sucrose < sodium ferric
`gluconate < iron isomaltoside 1000 < LMW iron dextran < iron
`carboxymaltose < ferumoxytol (Table 1). Ferumoxytol was eluted
`near the exclusion volume, indicating that both its diameter and
`molecular weight might be underestimated.
`
`3.1.2. Dynamic light scattering (DLS)
`The hydrodynamic diameter determined with DLS also mea-
`sures the carbohydrate shell of the IV iron agents and therefore
`is larger than iron oxide core diameters determined by TEM or
`XRD. In Fig. 2 narrow volume distributions of the whole particle
`diameters are shown. The medians of the hydrodynamic diameters
`rise from 8.3 to 23.6 nm in the order iron sucrose < sodium ferric
`gluconate < iron isomaltoside 1000 < LMW iron dextran  ferum-
`oxytol < iron carboxymaltose (Table 1). The zeta potentials of the
`iron preparations are shown in Table 2. Without pH adjustment,
`all iron preparations are negatively charged with the exception of
`iron carboxymaltose. The order of particle charges starting with
`the most negative iron preparation is ferumoxytol (43.2 mV) < ir-
`on gluconate  iron sucrose < iron isomaltoside 1000 < iron dex-
`tran < iron carboxymaltose
`(+3.7 mV). Acidification of
`the
`samples increased the zeta potential of iron carboxymaltose and
`decreased the negative zeta potential of all other compounds. At
`a pH value close to the physiological pH, all formulations showed
`a negative zeta potential, though that for iron carboxymaltose
`was much smaller.
`
`3.2. Size and structure of core
`
`3.2.1. Transmission electron microscopy (TEM)
`TEM images of IV iron agents are shown in Fig. 3. Dark, electron
`dense, beadlike structures present the cores of the iron oxide com-
`plexes, surrounded by a less electron dense matrix, which may be
`attributed to a carbohydrate fraction. The medians of the geomet-
`rical diameter of the core rise from 4.1 to 6.2 nm in the order so-
`dium ferric gluconate < iron sucrose < LMW iron dextran < iron
`isomaltoside 1000  ferumoxytol (Table 3). In case of iron carboxy-
`maltose cores tend to cluster and single cores are not definable.
`The median geometrical core diameter of
`these clusters is
`11.7 ± 4.4 nm.
`
`Fig. 1. Weight distribution vs. particle diameter as determined by gel permeation
`chromatography.
`
`Table 1
`Shell /Particle dimensions as determined by gel permeation chromatography (GPC)
`and dynamic light scattering (DLS).
`
`Iron complex
`
`MW (kDa)
`
`Calculated
`shell-Ø (nm)c
`
`Shell-Ø (nm)
`
`Sodium ferric gluconate
`Iron sucrose
`LMW iron dextran
`Iron isomaltoside 1000
`Iron carboxymaltose
`Ferumoxytol
`
`GPC
`
`164.1
`140.1
`165.0
`150.0
`233.1
`275.7
`
`20.3
`19.1
`20.7
`20.5
`23.8
`26.3
`
`DLS
`8.6a
`8.3a
`12.2a
`9.9a
`23.1a
`23.6a
`
`0.244b
`0.192b
`0.149b
`0.182b
`0.07b
`0.143b
`
`a Median-Ø.
`b Polydispersity index.
`c The most frequently found particle diameter in the distribution.
`
`Fig. 2. Volume distribution of the hydrodynamic diameter of intravenous iron preparations as determined by dynamic light scattering (DLS). Conditions: 0.4 mg Fe/ml.
`Vertical lines assign the median diameter.
`
`PGR2020-00009
`Pharmacosmos A/S v. American Regent, Inc.
`Petitioner Ex. 1048 - Page 4
`
`

`

`484
`
`M.R. Jahn et al. / European Journal of Pharmaceutics and Biopharmaceutics 78 (2011) 480–491
`
`Table 2
`Zeta potentials f of IV iron polynuclear complexes at different pH values.
`
`Iron gluconate
`
`Iron sucrose
`
`LMW iron dextran
`
`Ferumoxytol
`
`Iron isomalto-side 1000
`
`Iron carboxymaltose
`
`pH
`
`4.35
`7.4
`8.36a
`10.5
`
`f (mV)
`16.50
`29.70
`29.10
`29.60
`
`pH
`
`4.49
`7.43
`11.03a
`
`f (mV)
`14.25
`26.20
`28.15
`
`pH
`
`3.02
`6.4a
`7.31
`11.8
`
`f (mV)
`3.56
`15.30
`17.25
`15.75
`
`pH
`
`3.39
`6.6a
`7.36
`10.4
`
`f (mV)
`11.95
`43.20
`30.55
`34.40
`
`pH
`
`3.3
`6.3a
`7.35
`9.03
`11.5
`
`f (mV)
`3.98
`22.00
`21.05
`28.95
`26.40
`
`pH
`
`3.26
`5.36a
`7.26
`9.54
`
`f (mV)
`
`9.46
`3.68
`8.52
`16.35
`
`a pH in bidistilled sterile-filtrated water, without any pH adjustment.
`
`and 56° 2h could be a hint that also other structures are mixed in
`like akaganeite.
`The X-ray results of iron carboxymaltose indicate the akagane-
`ite structure with same intensities at same angles except for a min-
`or peak instead of a major peak at 12° 2h. LMW iron dextran and
`iron isomaltoside 1000 show a pattern which is similar to akagane-
`ite as well, but conformity is not as good (minor peaks instead of
`major peaks at 12° 2h and 35° 2h, in part missing minor peaks).
`The diffractogram of ferumoxytol, which is used as IV iron agent
`and contrast agent in magnetic resonance imaging as well, is close
`to pattern of magnetite and maghemite. Sharp peaks in the diffrac-
`togram belong to crystalline mannitol in the formulation.
`
`3.3. Ferrous iron content
`
`3.3.1. Mössbauer spectroscopy
`The mössbauer spectrum of iron isomaltoside 1000 shows a
`doublet with an isomer shift d = 0.44 mm/s and a quadrupole split-
`ting EQ = 0.78 mm/s (Fig. 5). Both parameters are characteristic for
`iron in the ferric state. There is no indication of iron in the ferrous
`state as characteristic isomer shifts and splittings are absent.
`
`3.4. Dialysable iron content
`
`3.4.1. Dialysis
`The results of the determination of the dialyzable ‘‘free’’ iron
`content are shown in Table 3. It appears that iron isomaltoside
`1000, iron carboxymaltose, and ferumoxytol yield very low free
`iron contents smaller than 0.002% of the total iron content. This
`was independent of the liquid used for the dilution and dialysis
`(water versus sodium chloride solution). Iron dextran yielded free
`iron contents of 0.1% and 0.2% in water and sodium chloride solu-
`tion, respectively. The highest free iron content was observed for
`sodium ferric gluconate yielding more than 1% in sodium chloride
`dilutions. However, the free iron content in the iron sucrose prep-
`aration (0.067% in NaCl and 0.057% in water) was lower than ex-
`pected. The experiment without pH adjustment showed that only
`iron carboxymaltose was affected by pH. As depicted in Fig. 6 the
`content of free iron in iron carboxymaltose increases from below
`the detection limit (<0.002%) at pH 7.5–0.262% when the experi-
`ment is conducted in nonbuffered 0.9% NaCl.
`
`3.5. Labile iron
`
`3.5.1. Acid soluble iron
`In acidic solution, FeOOH is dissociated: FeOOH þ 3HCl !
`Fe3þ þ 3Cl þ 2H2O. In this study, iron formulations [FeOOH]mLn
`with different carbohydrate ligands L were decomposed similarly:
`½FeOOHŠmLn þ 3mHCl ! mFe3þ þ 3mCl þ 2mH2O þ nL.
`As
`the
`molar extinction coefficient of
`the complex at 287.3 nm
`½FeOOHŠmLn  3000 M1 cm1)
`(e287:3nm
`is substantially higher
`than the
`½FeOOHŠmLn  580 M1 cm1) or
`extinction coefficient of Fe3+
`(e287:3nm
`carbohydrate (negligible), the decreasing FeOOH concentration is
`approximately proportional to the measured absorbance.
`
`Fig. 3. Transmission electron microscopy images of intravenous iron preparations.
`Conditions: 1 mg Fe/ml.
`
`3.2.2. X-ray diffraction (XRD)
`The particles mean diameters d of the cores were determined
`using the Scherrer equation and are presented in Table 3. The mean
`diameters of single core complexes are in the range of 3.3–6.4 nm
`and appear in accordance with diameters measured by TEM.
`In Fig. 4, X-ray diffractograms of IV iron agents (upper part of
`figure) are compared with diffraction data of standard iron oxides
`from the ICDD (lower part of figure, International Centre for Dif-
`fraction Data). Peaks belonging to the carbohydrate fraction are
`marked with arrows. With the exception of ferumoxytol the IV iron
`agents show broad regions of high intensities at in part similar an-
`gle values of diffraction with similar intensities.
`The patterns of iron sucrose and sodium ferric gluconate show a
`structure similar to 2-line ferrihydrite as there are just two major
`iron oxide peaks at 36° 2h and 62° 2h. Two others at 14° 2h and 22°
`2h belong to amorphous sucrose [27]. Small reflections at 40° 2h
`
`PGR2020-00009
`Pharmacosmos A/S v. American Regent, Inc.
`Petitioner Ex. 1048 - Page 5
`
`

`

`M.R. Jahn et al. / European Journal of Pharmaceutics and Biopharmaceutics 78 (2011) 480–491
`
`485
`
`Table 3
`Core dimensions as determined by transmission electron microscopy (TEM) and X-ray diffraction (XRD), acidic hydrolysis stability and dialysable iron with and without pH
`adjustment.
`
`Iron complex
`
`Sodium ferric gluconate
`Iron sucrose
`LMW iron dextran
`Iron isomaltoside 1000
`Iron carboxymaltose
`Ferumoxytol
`
`Core-Ø (nm)
`
`TEM
`4.1a ± 1.7
`5.0a ± 0.8
`5.6a ± 1.2
`6.3a ± 1.2
`11.7a,b ± 44
`6.2a ± 1.4
`
`t0.5 (h)
`
`XRD
`
`Acidic hydrolysis
`
`3.4
`3.3
`4.4
`4.2
`4.3
`6.4
`
`4.0 ± 0.1
`4.9 ± 0.1
`21.0 ± 1.2
`25.2 ± 1.2
`25.6 ± 1.6
`62.4 ± 0.4
`
`Dialysable irone (%)
`WFIf
`
`0.789 ± 0.048
`0.057d
`0.100 ± 0.0096
`<0.002c
`<0.002c
`<0.002c
`
`NaClf
`1.338d
`0.067d
`0.207 ± 0.0071
`<0.002c
`<0.002c
`<0.002c
`
`NaClg
`
`0.172h ± 0.0048
`0.014h ± 0.0029
`0.2621h ± 0
`0.005h ± 0.0047
`
`a Median-Ø.
`b Median-Ø of an agglomeration of several cores. Single cores are not definable.
`c Detection limit.
`d Only one sample.
`e Calculated from ICP-MS measurements.
`f Adjustment to pH 7.5.
`g Without pH adjustment.
`h Product approved for high dose administration.
`
`Fig. 5. Mössbauer spectrum of MonoFer as powder measured at 150 K. Measuring
`points are fitted with a line of Lorentz shape.
`Isomer shift d = 0.44 mm/s,
`quadrupole splitting EQ = 0.78 mm/s, line widths U = 0.69 mm/s.
`
`administered dose of 200 mg and 500 mg, respectively. The prod-
`ucts which show the highest fraction of labile, FerrozineÒ-detect-
`able iron are, by far, iron gluconate and iron sucrose (3.2 ± 0.4%
`for iron gluconate, 3.5 ± 0.2% for iron sucrose at a 200 mg dose,
`respectively). For the different compounds, the fraction of labile
`
`Fig. 4. X-ray spectra of intravenous iron preparations. At the bottom, the spectra
`are compared with diffraction data of standard iron oxides from the ICDD
`(International Centre for Diffraction Data). Thick reflexion lines of standard iron
`oxides: Intensity 70–100%. Thin reflexion lines of standard iron oxides: 30–69%
`intensity. Peaks belonging to carbohydrate fraction are assigned by arrows.
`
`The rate of hydrolysis is a measure of the relative stability of the
`FeOOH entity. In Fig. 7, it can be seen that the fraction of FeOOH
`remaining decreases with time, and in Table 3, the half-times of
`FeOOH decomposition of the various iron preparations are com-
`pared. The complex stability is increasing in the order of iron glu-
`conate < iron sucrose  iron dextran < iron carboxymaltose  iron
`isomaltoside 1000  ferumoxytol. There is some indication that
`the rate of degradation relates to the surface area of the iron com-
`plex formulation and decreasing with increasing particle size
`(Fig. 8).
`
`3.5.2. FerrozineÒ-detectable labile iron in human serum
`The results of the determination on detectable labile iron with
`the FerrozineÒ-method are shown in Fig. 9. The amount of the
`labile iron, measured by this test was nearly equivalent to the
`
`Fig. 6. Comparative free iron content in high dose IV iron products. The detection
`limit was 0.002%. Red bars indicate free iron content following adjustment of the
`diluted preparation to pH 7; blue bars indicate results obtained without pH
`adjustment. Star indicates concentrations below detection limit. SD’s are listed in
`Table 3.
`
`PGR2020-00009
`Pharmacosmos A/S v. American Regent, Inc.
`Petitioner Ex. 1048 - Page 6
`
`

`

`486
`
`M.R. Jahn et al. / European Journal of Pharmaceutics and Biopharmaceutics 78 (2011) 480–491
`
`sodium ferric gluconate
`iron sucrose
`LMW iron dextran
`iron isomaltoside
`iron carboxymaltose
`ferumoxytol
`20
`
`30
`
`Time [h]
`
`40
`
`50
`
`0
`
`10
`
`0.00
`
`-0.25
`
`-0.50
`
`-0.75
`
`-1.00
`
`-1.25
`
`-1.50
`
`-1.75
`
`-2.00
`
`ln(fraction FeOOH)
`
`Fig. 7. Acid soluble iron. Concentration: 10 lg Fe/ml, 0.2375 M HCl. At t = 0 min, the
`fraction of FeOOH is 1 according to 100% not hydrolyzed FeOOH. Each point
`represents the average of three measurements; error bars are sometimes smaller
`than symbols. Data were fitted with a second degree polynomial (R2 > 0.990). The
`solid line labels the half-time. SD’s are listed in Table 3. For further details refer to
`methods.
`
`Fig. 9. Comparative labile iron pools of parenteral iron products. Upper diagram:
`Concentration of the FerrozineÒ-detectable labile iron pool in lg/ml. The bars
`represent the average of at least four measurements. Lower diagram: FerrozineÒ-
`detectable labile iron in percentage of the total used dose. For each measurement,
`the iron complex was incubated in human serum for 10 and 45 min, respectively.
`Thereafter, the FerrozineÒ reaction was performed and in each case an intercept of a
`second degree polynomial regression function of the absorption versus time curve
`with the ordinate was calculated. These intercepts were extrapolated to an
`incubation time of t = 0 by linear regression, yielding the labile iron pool in serum
`for each intravenous iron product.
`
`200 MHz of isomaltoside 1000 confirmed the above-mention