British Journal of Haematology, 2000, 110, 985–992
`
`Chelator-induced iron excretion in iron-overloaded marmosets
`
`Thomas Sergejew, Peter Forgiarini and Hans-Peter Schnebli Novartis Pharma AG, Basle, Switzerland
`
`Received 15 December 1999; accepted for publication 19 May 2000
`
`Summary. In order to test new orally active iron chelators in
`a predictive way, a primate model has been developed. This
`model makes use of
`the marmoset monkey (Callithrix
`jacchus) and its overall design is similar to a previously
`reported monkey model. However, this new model enables a
`higher compound throughput and requires lower amounts
`of test compound because the animals are much easier to
`handle and have much lower body weights. The marmosets
`were iron-overloaded by three intraperitoneal injections of
`iron (III) hydroxide polyisomaltose. For the iron-balance
`studies, the animals were kept in metabolic cages and were
`maintained on a low-iron diet in order to reduce faecal
`
`background. After compound administration, the excretion
`of iron in urine and faeces was followed for 2 d. A series of
`well-known chelators was tested for validation of the model.
`In particular, comparison of the iron-clearing properties of
`DFO, L1, CP94 and HBED in marmosets and humans
`demonstrated the predictive value of the model and justify
`our expectation that if iron chelators such as CGP65015,
`ICL670A and CGP75254A are active in marmosets, they
`will be active in humans as well.
`
`iron chelation, marmoset model,
`Keywords:
`CGP65015, CGP75254A.
`
`ICL670A,
`
`Excess tissue iron accumulates as a result of repeated blood
`transfusions in b-thalassaemia major, sickle cell disease and
`other transfusion-dependent anaemias, or as a result of
`excessive dietary iron uptake associated with anaemias or
`increased dietary iron uptake in hereditary haemochroma-
`tosis. Unless removed by chelator treatment or, in hereditary
`haemochromatosis patients, by phlebotomy,
`toxic and
`eventually lethal levels of iron are deposited mainly in the
`liver, endocrine organs and in the myocardium. The exact
`mechanism of iron damage to these tissues is unknown, but
`it is an established fact that organ failure correlates with
`iron burden.
`Iron chelators slowly mobilize these iron
`deposits, probably by an indirect process of continuously
`binding those amounts of soluble iron present in the ‘transit
`pool’ that are in equilibrium with the insoluble haemosi-
`derin. Solubilized, chelated iron is excreted in the urine and/
`or faeces.
`Since its introduction in 1962, desferrioxamine B (DFO,
`Desferal) has remained the only effective and safe treatment
`of
`iron overload (Sephton-Smith, 1962). Hundreds of
`published studies document its effectiveness in promoting
`iron excretion, both in animals and in man. DFO has been
`shown to reduce the body iron burden and thus iron-related
`morbidity (Brittenham et al, 1994) and mortality (Olivieri
`et al, 1994). However, its very short plasma half-life and
`poor oral bioavailability necessitate special modes of
`
`Correspondence: Dr Thomas Sergejew, Novartis Pharma AG,
`PO Box, CH-4002 Basle, Switzerland. E-mail: thomas.sergejew@
`pharma.novartis.com
`
`application (subcutaneous or intravenous infusion) that
`are impractical, can cause local reactions and are difficult to
`accept for many patients. Therefore, the need for novel,
`orally active iron chelators has been recognized for a long
`time (Hider & Hall, 1991; Hershko, 1992; Schnebli et al,
`1994).
`One of the difficulties in assessing the potential benefit of
`iron chelators for patients is the fact that rodents, the
`‘standard’
`laboratory animals, differ substantially from
`primates,
`including man,
`in their iron metabolism,
`i.e.
`uptake, pools,
`fluxes and excretion patterns (Finch &
`Huebers, 1982; Finch et al, 1978). This may explain why
`some compounds that were identified as orally active in
`rodents, later failed to be effective in non-human primates
`(Bergeron et al, 1998a) and in patients (Grady et al, 1994).
`To avoid this and in order to identify the best drug
`candidates for the expensive and time-consuming toxicity
`studies and clinical trials, a monkey model using iron-
`overloaded Cebus apella was developed a few years ago
`(Wolfe et al, 1989; Bergeron et al, 1991). This model was
`successfully used to compare compounds within several
`series of iron chelators (Bergeron et al, 1992,1993a; Peter
`et al, 1994). However, the size of these monkeys (about 4 kg
`body weight) necessitates a considerable infrastructure and
`amounts of test compounds not routinely available at early
`stages of research.
`We have therefore established a similar model using
`Callithrix jacchus, commonly known as the marmoset, which
`is much smaller (about 400 g body weight), is more easily
`handled and requires no more test compound than rats.
`
`q 2000 Blackwell Science Ltd
`
`985
`
`PGR2020-00009
`Pharmacosmos A/S v. American Regent, Inc.
`Petitioner Ex. 1078 - Page 1
`
`

`

`986
`
`T. Sergejew, P. Forgiarini and H.-P. Schnebli
`
`Table I. Iron chelators used in the present study.
`
`Code
`
`DFO
`
`L1
`CP94
`CP102
`CGP65015
`
`R-nDFT
`S-nDFT
`DnDFT
`DdnDFT
`ICL670A
`HBED
`
`DmHBED
`
`CGP75254A
`
`Parabactin
`CGP43902B
`
`Chemical name
`
`References to published Cebus results
`
`Desferrioxamine B (methanesulphonate salt); Desferal
`
`1,2-dimethyl-3-hydroxy-pyridin-4-one deferiprone, CP20, Kelfer
`1,2-diethyl-3-hydroxy-pyridin-4-one
`1-hydroxyethyl-2-ethyl-3-hydroxy-pyridin-4-one
`(^)-3-hydroxy-1-(2-hydroxyethyl)-2-(hydroxyphenyl-methyl)
`-1H-pyridin-4-one
`(R)-desmethyl desferrithiocin; l-norDFT
`(S)-desmethyl desferrithiocin; d-norDFT
`desazadesmethyl desferrithiocin
`desazadesmethyl-5,5-dimethyl desferrithiocin
`4-[3,5-bis-(2-hydroxy-phenyl)-[1,2,4]triazol-1-yl]-benzoic acid
`N,N0-bis(2-hydroxybenzyl)-ethylenediamine-N,N0-diacetic acid
`hydrochloride dihydrate
`dimethyl N,N0-bis(2-hydroxybenzyl)-ethylenediamine-N,N0-
`diacetate; dimethyl-HBED
`sodium N,N0-bis(2-hydroxybenzyl)-ethylenediamine-N-
`methoxycarbonylmethyl-N0-acetate; monoethyl-HBED
`Parabactin
`1-[N-ethoxycarbonylmethylpyridoxylidenium]-2-[20-pyrimidyl]
`-hydrazone bromide
`
`Bergeron et al (1991, 1992, 1993a, b,
`1998a, b); Peter et al (1994); Schnebli et al (1994)
`Bergeron et al (1992); Peter et al (1994)
`Bergeron et al (1992); Peter et al (1994)
`Singh (1994)
`
`Bergeron et al (1998b)
`Bergeron et al (1993a, 1998a); Peter et al (1994)
`Bergeron et al (1993a); Peter et al (1994)
`Bergeron et al (1993a); Peter et al (1994)
`
`Bergeron et al (1998a); Peter et al (1994)
`
`Peter et al (1994)
`
`Bergeron et al (1993b)
`Bergeron et al (1991)
`
`Here, we describe the details of this model and its utility by
`comparing the effects of a series of
`iron chelators in
`marmosets, Cebus apella and, where data is available, in
`man. Finally, the effect of three new, orally active iron
`chelators proprietary to Ciba-Geigy (now Novartis),
`CGP65015 (Lowther et al, 1999), ICL670A (Acklin et al,
`1998; Schnebli, 1998) and CGP75254A (Lowther et al,
`1998; Spanka et al, 1998), are presented.
`
`MATERIALS AND METHODS
`
`Animals. Adult (1- to 11-year-old) marmosets (Callithrix
`jacchus) of both sexes, weighing 300–450 g, were obtained
`from Ciba-Geigy’s husbandry unit. They were kept in pairs
`at 24–278C and a relative humidity of at least 60% in a 12-
`h light-dark cycle (lights on at 06·00 a.m., off at 06·00
`p.m.). The animals were maintained on a pellet diet (No.
`962 Nafag, Gossau, Switzerland) supplemented with fruit-
`vegetable mash, enriched with a vitamin-minerals concen-
`trate (No. 9628 Nafag, Gossau, Switzerland) and milk.
`Water was available ad libitum.
`Iron-overloading procedure. To approximate the iron status
`of thalassaemic patients in the marmosets, the animals were
`iron-overloaded by three i.p. injections of iron (III) hydroxide
`polyisomaltose (Anaemex, Ciba-Geigy, Basel, Switzerland) at
`14-d intervals (200 mg/kg twice and 100 mg/kg at the
`third injection). Prior to the first exposure to an iron
`chelator, the marmosets were rested for at least 8 weeks in
`order to allow definite distribution of the injected iron into
`all storage compartments.
`Chelators and compound administration. The structures of
`all test compounds have been published previously (see
`
`references in Table I). Parabactin was a gift from Professor
`R. J. Bergeron, University of Florida, Gainsville, FL, USA. All
`other compounds were synthesized in the laboratories of
`Ciba-Geigy (now Novartis), Basle, Switzerland. The com-
`pounds were dispersed in 40% aqueous Cremophor RH 40
`(BASF, Ludwigshafen, Germany)
`for administration. All
`dosages were expressed in terms of ‘iron binding equiva-
`lents’ (IBE) which take into account the stoichiometry of the
`chelator in the iron complex, i.e. 150 mmol IBE correspond
`to 150, 300 and 450 mmol of a mono-, bi- and tridentate
`chelator respectively. The standard dose was 150 mmol iron
`binding equivalents (IBE) per kg body weight. The applied
`volume was 5 ml/kg body weight.
`Iron excretion studies. In order to reduce the diet-derived
`faecal iron background, the monkeys received a low-iron
`diet 7 d before and throughout the experiments (4 d). The
`diet was prepared by mixing 628·9 g of a liquid diet for
`primates (Nafag no. 9694, Gossau, Switzerland), 83 g of the
`corresponding lipid mixture (Nafag no. 9696), 131·3 g of
`dietary fibre (Sanacell, Nafag) and 1·0 litre of distilled water.
`Each monkey received 30 g of this liquid diet mixed with
`20 g milk rice per day. The iron content of this food mixture
`was 3·1 mg/kg.
`The animals were placed into acrylic glass metabolic
`cages especially designed for marmosets (square area of
`0·2 m2 and 55 cm high), 48 h prior to drug administration.
`Marmosets live in pairs or families and react to separation
`from their mate. To keep the separation stress at a
`minimum, all animals from a family usually participated
`in the same experiment. By placing the animals from a
`social unit into adjacent metabolic cages, the animals,
`although separated, could see each other.
`
`q 2000 Blackwell Science Ltd, British Journal of Haematology 110: 985–992
`
`PGR2020-00009
`Pharmacosmos A/S v. American Regent, Inc.
`Petitioner Ex. 1078 - Page 2
`
`

`

`Chelator-induced Iron Excretion in Iron-overloaded Marmosets
`
`987
`
`Urine and faeces were collected in 24-h fractions to
`determine background iron excretion. After administration
`of the chelator, the marmosets were maintained in the same
`metabolic cage for another 48 h. They were frequently
`observed to note unusual reactions, and again, urine and
`faeces were collected in 24 h fractions. Thus, each animal
`served as its own control. Faecal samples were weighed and
`assayed for the presence of occult blood (haemo FEC,
`Boehringer, Rotkreuz, Switzerland) and the (infrequent)
`positive samples were rejected. Urine samples were weighed
`and checked with a dip stick (Combur10 Test, Boehringer,
`Rotkreuz, Switzerland) for pH and specific gravity, as well as
`to detect white blood cells, nitrite, protein, glucose, ketones,
`urobilinogen, bilirubin and blood.
`Determination of iron excretion. Urinary iron concentration
`was determined colorimetrically using the bathophenan-
`throline method (Smith et al, 1952). Two 100 ml aliquots of
`each sample were placed into separate wells of a conical-
`bottomed 96-well plate. To each well 100 ml of a 10%
`trichloroacetic acid solution in 3 mol/l HCl were added.
`After incubation at 658C for 4 h, the plate was centrifuged
`for 10 min at 1300 g. A 100-ml aliquot of each supernatant
`was transferred into the corresponding well of a flat-
`bottomed 96-well plate. Then, 100 ml of the iron indicator
`solution (6·32 ml of water, 80 ml of mercaptoacetic acid,
`12·4 mg of bathophenanthroline-disulphonic acid disodium
`salt, 3·16 ml of saturated sodium acetate and 4·22 ml of
`NaOH 5·9N) and blank solution (the same, but without
`bathophenanthroline) were added to the first and second
`aliquot of each sample respectively. The plate was agitated
`and the absorbance was measured after 30 min at 535 nm
`with a microplate reader (UVmax, Molecular Devices). The
`iron concentration of each sample was calculated using a
`standard curve (0–20 mg/ml) after subtraction of
`the
`individual blank.
`Faecal
`iron in each sample was determined by flame
`atomic absorption spectrometry after wet-ashing the entire
`(24 h) sample of faeces.
`All results are expressed as mg of iron excreted per kg
`body weight. For the calculation of chelator-induced iron
`excretion, the mean pretreatment values of each animal
`were subtracted from the post-treatment values.
`Other analytical procedures. Total serum iron and transfer-
`rin saturation were determined using a commercial kit
`obtained from Sigma, Buchs, Switzerland. The method was
`
`adapted to 96-well plates, which enabled the serum volume
`to be scaled down to 50 ml. This procedure is based on
`measuring serum iron and latent iron-binding capacity with
`ferrozine, a sulphonated derivative of diphenyltriazine
`(Stookey, 1970; Persijn et al, 1971). The absorbance was
`measured at 550 nm using a microplate reader (UVmax,
`Molecular Devices).
`Nontransferrin-bound serum iron (NTBI) was determined
`by a modification of the method of Zhang et al (1995). As
`already suggested by Singh et al (1990), the concentration
`of nitrilotriacetic acid (NTA) has to be at least 80 mmol/l in
`order to remove all NTBI from serum proteins. Fifty ml of
`800 mmol/l NTA (pH 7·0) were added to 450 ml serum and
`incubated at room temperature for 20 min. Then the
`solutions were ultrafiltered (10 000 g, 20 min) using
`Ultrafree-0·5 Centrifugal Filter Devices equipped with a
`Biomax-30 membrane, with a molecular weight cut-off of
`30 kDa (Millipore, Volketswil, Switzerland). Aliquots of
`200 ml of ultrafiltrate were transferred into wells of a 96-
`well microplate. Five microlitres of 0·5 mol/l bathophenan-
`throline-disulphonic acid disodium salt and 5 ml of 1 mol/l
`thioglycolic acid were added. The plate was agitated and the
`absorbance measured after 60 min at 535 nm with a
`microplate reader (UVmax, Molecular Devices). The NTBI
`concentration of each sample was calculated using a
`standard curve (0–20 mg/ml). Water was used as the blank.
`Total liver iron was determined by flame atomic absorp-
`tion spectrometry after wet-ashing a sample of approxi-
`mately 0·5 g of liver.
`
`RESULTS
`
`Iron status of the animals
`As depicted in Table II, the total serum iron levels, the NTBI
`levels and the transferrin saturation were somewhat higher
`in untreated marmoset monkeys than in normal humans.
`This may reflect a modest endogenous, probably diet-
`derived, iron-overload status in this marmoset population
`as described for marmosets in other laboratories (Miller et al,
`1997). The observation that older animals tended to have
`higher liver iron concentrations than younger ones supports
`this conclusion (data not shown). Ferritin levels could not
`be measured owing to lack of an anti-marmoset ferritin
`antibody.
`Injection of iron (III) hydroxide polyisomaltose resulted in
`
`Table II. Iron status in control and iron-overloaded marmosets.
`
`Control
`
`Iron-overloaded
`
`Parameter
`
`Unit
`
`Mean ^ SEM
`
`n
`
`Mean ^ SEM
`
`n
`
`Serum iron
`Transferrin saturation
`NTBI
`Liver iron
`
`mmol/l
`%
`mmol/l
`mg/g*
`
`24 ^ 3
`54 ^ 6
`1·0 ^ 0·3
`922 ^ 60
`
`7
`7
`7
`20
`
`45 ^ 1
`100 ^ 0
`3·6 ^ 0·3
`7378 ^ 293
`
`61
`61
`61
`20
`
`*Wet liver weight
`
`q 2000 Blackwell Science Ltd, British Journal of Haematology 110: 985–992
`
`PGR2020-00009
`Pharmacosmos A/S v. American Regent, Inc.
`Petitioner Ex. 1078 - Page 3
`
`

`

`988
`
`T. Sergejew, P. Forgiarini and H.-P. Schnebli
`
`a substantial further iron loading. At the end of the loading
`and equilibration period, all animals had fully saturated
`transferrin and their total serum iron and NTBI levels were
`approximately 2- and 3·5-fold higher than prior to loading
`(Table II). Liver iron increased by a factor of seven. In many
`marmosets,
`iron-overloading resulted in a loss of body
`weight that, however, reversed spontaneously.
`Marmosets, once iron-overloaded in this way, maintained
`their iron status for a long period of time, so that no (re-)
`loading became necessary. A series of 20 marmosets that
`were sacrificed more than 2 years after iron loading and
`after multiple single dose-chelator experiments, still had a
`very high total liver iron content (7378 ^ 293 compared
`with 922 ^ 60 mg/g). Histological examination of
`livers
`from iron-overloaded marmosets revealed that the iron was
`located in both hepatocytes and Kupffer cells. Stainable iron
`deposits were also found in the spleen and other organs
`including the heart.
`Maintenance of the iron-overload status of the marmosets
`was further ascertained by repeated measurements of
`chelator-induced iron excretion by the same compound in
`the same animals, in experiments spaced up to 18 months
`apart. As shown with selected test compounds, including
`DFO, no decrease of pharmacological response was experi-
`enced during this observation period (data not shown).
`
`Effects of iron chelators
`Untreated,
`iron-overloaded marmosets excrete very small
`amounts of
`iron in the urine (14 ^ 9 mg iron/kg body
`weight/day; n (cid:136) 100) and the bulk of the daily faecal iron
`(266 ^ 116 mg iron/kg body weight/day; n (cid:136) 100) repre-
`sents non-absorbed dietary iron.
`Figure 1 serves as an example of the iron elimination
`profile in urine and faeces that can be obtained with the
`model described here. When CGP65015 was applied (d 0),
`both urinary and faecal iron clearance increased dramati-
`cally above the pretreatment background during the first
`24 h after administration. With many chelators, excretion
`levels return to background during the subsequent 24- to
`
`iron excretion in iron-overloaded
`Fig 1. Urinary and faecal
`marmosets induced by a single dose of CGP65015 (150 mmol
`IBE/kg p.o. on d0); excretion profiles of four individual marmosets.
`
`like
`48-h collection period, while potent compounds
`CGP65015 still promote iron elimination during the second
`post-treatment day. In a separate experiment, it could be
`shown that CGP65015-induced iron excretion was negli-
`gible on the third day after compound application (data not
`shown).
`Chelator-induced iron excretion in marmosets was dose
`dependent with most compounds tested, as shown in Figs 2
`and 3 for DFO and Novartis’ three new chemical entities,
`CGP65015, ICL670A and CGP75254A. Note that DFO was
`given subcutaneously as oral application of DFO was
`practically ineffective in marmosets as is the case in
`thalassaemia patients. In both marmoset and Cebus, DFO
`was somewhat less effective than in man, probably because
`the compound had to be applied as a bolus s.c. injection
`because (for technical reasons) it could not be given as a
`continuous infusion as it is in patients.
`To further validate the marmoset model, a series of iron
`chelators were tested and the results obtained were
`compared with those generated in the Cebus (Fig 4). With
`DFO, at the standard dose of 150 mmol/kg, the marmoset
`excreted less iron than the Cebus, but at higher doses (Fig 2)
`the total iron excretion was similar in both monkeys. The
`more rapid metabolism of DFO in marmoset vs. Cebus in
`plasma (Steward et al, 1996) may help to explain this
`difference.
`including L1 (Kelfer), are bidentate
`Hydroxypyridones,
`iron chelators which form three to one complexes with iron.
`In both marmoset and Cebus, the rank order of potency was
`L1 , CP94 , CP102. With L1, the majority of the induced
`iron was excreted in the urine. In contrast, the more potent
`CP94 and CP102 excreted iron mainly in the faeces (data
`not shown). In the marmoset, CP102 was clearly more
`active than CP94, whereas
`this difference was not
`significant in the Cebus. CGP65015 was about four times
`more potent
`than CP102 and is
`the most effective
`hydroxypyridone in our hands.
`The Desferrithiocin (DFT) derivatives, (R)-nDFT, (S)-nDFT,
`dnDFT and ddnDFT, are tridentate iron chelators which
`form two to one complexes with iron. Interestingly, while
`the enantiomers of desmethyl-DFT (R)- and (S)-nDFT were
`equieffective in the marmoset, in the Cebus, the (R)-form
`was inactive, while the effect of the (S)-form was similar to
`that seen in the marmoset. Apart from these discrepancies,
`both the marmoset and Cebus models ranked the potency of
`the compound of this class of chelators similarly. As with
`CGP65015, at a dose of 150 mmol IBE, ICL670A excretes
`iron amounts of about 2500 mg iron/kg body weight and is
`one of the most potent orally active tridentate iron chelators
`ever tested in our laboratories.
`Both primate species responded similarly to the hexaden-
`tate chelator HBED: the compound was highly active after
`subcutaneous administration whereas, applied orally, it was
`poorly active or inactive in both monkeys. In an attempt to
`make the compound orally available, esters like dimethyl-
`HBED have been synthesized (Pitt et al, 1986). After
`absorption, such prodrugs have to be hydrolysed to the
`parent HBED, as the esters themselves have nearly no
`affinity for iron. Interestingly, dimethyl-HBED had a different
`
`q 2000 Blackwell Science Ltd, British Journal of Haematology 110: 985–992
`
`PGR2020-00009
`Pharmacosmos A/S v. American Regent, Inc.
`Petitioner Ex. 1078 - Page 4
`
`

`

`Chelator-induced Iron Excretion in Iron-overloaded Marmosets
`
`989
`
`Fig 2. Dose-dependent urinary and faecal
`iron excretion in iron-overloaded marmo-
`sets and Cebus monkeys induced by DFO;
`each bar represents the mean of four to six
`animals.
`
`Fig 3. Dose-dependent urinary and faecal
`iron excretion in iron-overloaded marmo-
`sets induced by CGP65015, ICL670A and
`CGP75254A; each bar represents the mean
`of three to eight monkeys.
`
`Fig 4. Comparison of total iron excretion
`in iron-overloaded marmosets (black
`columns) and Cebus monkeys (white
`columns). All iron chelators indicated were
`applied at 150 mmol IBE/kg; each bar
`represents the mean of three to six animals.
`
`the route of
`in the two species: regardless of
`effect
`administration (p.o. or s.c.), the drug strongly promoted
`iron excretion in the marmoset, while it was inactive in the
`Cebus. Apparently, unlike the marmoset, the Cebus cannot
`hydrolyse the diester.
`
`CGP75254A, the monoethyl ester of HBED, represents a
`new, interesting chemical entity that resulted from Novartis’
`internal iron-chelation research programme (Lowther et al,
`1998; Spanka et al, 1998). The compound is able to chelate
`iron and is demonstrably orally active. As shown in Fig 4,
`
`q 2000 Blackwell Science Ltd, British Journal of Haematology 110: 985–992
`
`PGR2020-00009
`Pharmacosmos A/S v. American Regent, Inc.
`Petitioner Ex. 1078 - Page 5
`
`

`

`990
`
`T. Sergejew, P. Forgiarini and H.-P. Schnebli
`
`the total iron excretion induced is nearly the same as that
`produced by an equimolar dose of subcutaneous HBED,
`suggesting that CGP75254A has high oral availability in
`the primate.
`In both monkey species, parabactin, a naturally occur-
`ring, hexadentate siderophore-like DFO, and the tridentate,
`pyridoxal
`isonicotinoyl
`hydrazone
`(PIH)-derivative,
`CGP43902B, were poorly active or inactive after oral
`administration. By contrast, when given subcutaneously,
`parabactin induced a marked iron excretion in both
`monkeys (data not shown).
`
`DISCUSSION
`
`The marmoset model described here has been adapted from
`the Cebus model used by Bergeron et al (1991). The main
`differences between the two models are in the metabolic
`cage design and the low-iron diet, both of which had to be
`adapted to the marmoset. The marmoset offers
`two
`advantages over the Cebus: the much lower compound
`requirement because of the lower body weight (one tenth)
`and the easier handling of the animal. In contrast to the
`Cebus, the marmoset can easily be restrained by a single
`experienced person, so that iron-loading injections and
`compound administration require no previous anaesthesia.
`Iron overloading of experimental animals with iron
`dextran is fairly standard and has been described in mice
`(Porter et al, 1990), rats (Porter et al, 1993), gerbils
`(Carthew et al, 1993) and Cebus monkeys (Bergeron et al,
`1991). The iron dextran is initially taken up by the
`reticuloendothelial system, but then equilibrates to the
`parenchymal system. This iron-overloading procedure also
`proved to be very effective in marmosets. The distribution of
`the excess iron in overloaded marmosets was similar to that
`of transfusional overload, i.e. stainable iron deposits were
`found not only in hepatocytes and Kupffer cells, but also in
`the spleen and other organs including the myocardium.
`Again similar to man, marmosets, once iron-overloaded,
`maintained their iron status for a long period of time. High
`liver iron contents were measured in marmosets, even after
`more than 2 years after iron loading and multiple experi-
`ments with a variety of chelators. Consistent with this, as in
`man and as previously observed in Cebus monkeys,
`spontaneous urinary iron excretion in iron-overloaded
`(but untreated) marmosets is minimal. The present marmo-
`set model
`thus appears to fairly well approximate the
`situation in transfusional haemosiderosis.
`In regularly transfused thalassaemia patients, chelation
`treatment is considered effective when 250–400 mg of iron
`are excreted per kg body weight per day (Gordeuk et al,
`1987). Chelators that are effective by this definition were
`easily detected in the marmoset model, even if the chelator-
`induced iron excretion was primarily into the faeces; by
`placing the animals on a low-iron diet several days prior to
`the experiments, the daily faecal background iron could be
`reduced to approximately 250 mg iron/kg/d. In addition,
`interindividual variability in faecal iron background was
`compensated by each animal serving as its own control.
`Obviously, detection of chelator-induced iron excretion in
`
`the urine is even easier as there is practically no urinary
`background excretion.
`The experience to date indicates that the marmoset is
`useful
`for assessing the effectiveness of various
`iron
`chelators. Marmosets clearly differentiate between com-
`pounds of different potencies and appear to serve as a good
`model for the comparison of these compounds. Although
`obvious differences were observed between the marmoset
`and the Cebus model, particularly with dimethyl-HBED,
`overall the results obtained with the two models were
`reasonably similar.
`A comparison of clinical results (where available) with the
`results obtained in marmosets, indicates that this model has
`a high predictive value for the potential usefulness of newly
`discovered chelators. In marmosets, and incidentally in the
`Cebus, orally applied DFO, HBED and L1 were only modestly
`active or inactive, reflecting the situation in the clinic,
`where orally applied DFO (Callender & Weatherall, 1980)
`and HBED (Grady et al, 1994) have been dismissed as
`therapeutic options and where oral L1 (Hershko et al, 1998)
`has shown a limited efficiency. On the positive side, s.c. DFO
`was active in both monkey models, although at lower doses
`this was less so in the marmoset than in the Cebus, probably
`as a result of the more rapid metabolism of the DFO in
`marmoset plasma. In both monkeys, the hydroxypyridone,
`CP94, was more effective than L1, which again is in
`agreement with the clinical findings (Porter, 1996).
`The utility of the marmoset model is further supported by
`the following observations. In order to design more potent
`hydroxypyridones (extending their duration of action by
`preventing extensive glucuronidation), 1-hydroxyalkyl deri-
`vatives, for example, CP102, have been developed by Hider
`(Singh, 1994).
`Indeed, as predicted from metabolism
`studies, CP102 turned out to be much more active than
`L1 and CP94 in the marmoset.
`It can be concluded that the marmoset model described
`here mimics to a large extent transfusional haemosiderosis
`and is
`therefore useful
`for
`the assessment of newly
`discovered chelators. Except for the two technical advan-
`tages of the marmoset (lower compound requirement, easier
`handling of the animals), the two monkey models appear to
`be equivalent, although it has not been established with
`certainty which of the two monkeys is phylogenetically
`more closely related to man. In summary, the results shown
`in the present report justify our expectation that if
`iron
`chelators are active in marmosets they will be active in
`humans as well.
`
`ACKNOWLEDGMENTS
`
`All Cebus results have been published before (Table I). The
`raw data were kindly provided by Professor R. J. Bergeron
`and J. Wiegand, University of Florida, Gainsville, FL, USA.
`1n order to compare the pharmacological effect between
`both monkey species, the excretion values were recalculated
`using the procedure described in Materials and methods.
`
`q 2000 Blackwell Science Ltd, British Journal of Haematology 110: 985–992
`
`PGR2020-00009
`Pharmacosmos A/S v. American Regent, Inc.
`Petitioner Ex. 1078 - Page 6
`
`

`

`Chelator-induced Iron Excretion in Iron-overloaded Marmosets
`
`991
`
`REFERENCES
`
`Acklin, P., Lattmann, R., Bu¨ hlmayer, P., Crowe, A., Faller, B., Jin, Y.,
`Nick, H.P., Sergejew, T., Spanka, C., Wong, A., Zbinden, P. &
`Schnebli, H.P. (1998) CGP72670, a potent, orally active iron
`chelator. Chimia, 52, 446.
`Bergeron, R.J., Streiff, R.R., Wiegand, J., Vinson, J.R.T., Luchetta, G.,
`Evans, K.M., Peter, H. & Jenny, H.B. (1991) A comparative
`evaluation of
`iron clearance models. Annals of the New York
`Academy of Sciences, 612, 378–393.
`Bergeron, R.J., Streiff, R.R., Wiegand, J., Luchetta, G., Creary, E.A. &
`Peter, H.H. (1992) A comparison of the iron-clearing properties of
`1,2-dimethyl-3-hydroxypyrid-4-one, 1,2-diethyl-3-hydroxypyrid-
`4-one, and deferoxamine. Blood, 79, 1882–1890.
`Bergeron, R.J., Streiff, R.R., Creary, E.A., Daniels, Jr, R.D., King, W.,
`Luchetta, G., Wiegand, J., Moerker, T. & Peter, H.H. (1993a) A
`comparative study of the iron-clearing properties of desferrithio-
`cin analogues with desferrioxamine B in a Cebus monkey model.
`Blood, 81, 2166–2173.
`Bergeron, R.J., Streiff, R.R., King, W., Daniels, Jr, R.D. & Wiegand, J.
`(1993b) A comparison of
`the iron-clearing properties of
`parabactin and desferrioxamine. Blood, 82, 2552–2557.
`Bergeron, R.J., Wiegand, J. & Brittenham, G.M. (1998a) HBED: a
`potential alternative to deferoxamine for iron-chelating therapy.
`Blood, 91, 1446–1452.
`Bergeron, R.J., Wiegand, J., Ratliff-Thompson, K. & Weimar, W.R.
`(1998b) The origin of the differences in (R)- and (S)-desmethyl-
`desferrithiocin. Annals of the New York Academy of Sciences, 850,
`202–216.
`Brittenham, G.M., Griffith, P.M., Nienhuis, A.W., McLaren, C.E.,
`Young, N.S., Tucker, E.E., Allen, C.J., Farrell, D.E. & Harris, J.W.
`(1994) Efficacy of deferoxamine in preventing complications of
`iron overload in patients with thalassemia major. New England
`Journal of Medicine, 331, 567–573.
`Callender, S.T. & Weatherall, D.J. (1980) Iron chelation with oral
`desferrioxamine. Lancet, 2, 689.
`Carthew, P., Dorman, B.M., Edwards, R.E., Francis, J.E. & Smith,
`A.G. (1993) A unique rodent model for both cardiotoxic and
`hepatotoxic effects of prolonged iron overload. Laboratory
`Investigation, 69, 217–222.
`Finch, C.A. & Huebers, H. (1982) Perspectives in iron metabolism.
`New England Journal of Medicine, 306, 1520–1528.
`Finch, C.A., Ragan, H.A., Dyer, I.A. & Cook, J.D. (1978) Body iron
`loss in animals. Proceedings of the Society for Experimental Biology
`and Medicine, 159, 335–338.
`Iron
`(1987)
`Gordeuk, V.R., Bacon, B.R. & Brittenham, G.M.
`overload: causes and consequences. Annual Review of Nutrition,
`7, 485–508.
`Grady, R.W., Salbe, A.D., Hilgartner, M.W. & Giardina, P.J. (1994)
`Results from a phase I clinical
`trial of HBED. Advances in
`Experimental Medicine and Biology, 356, 351–359.
`Hershko, C. (1992) Iron chelators in medicine. Molecular Aspects of
`Medicine, 13, 113–165.
`Hershko, C., Konijn, A.M. & Link, G. (1998) Iron chelators for
`thalassaemia. British Journal of Haematology, 101, 399–406.
`Hider, R.C. & Hall, A.D. (1991) Clinically useful chelators of
`tripositive elements. Progress in Medicinal Chemistry, 28, 41–173.
`Lowther, N., Tomlinson, B., Fox, R., Faller, B., Sergejew, T. &
`Donnelly, H. (1998) Caco-2 cell permeability of a new oral iron
`chelator: correlation with physicochemical properties and oral
`activity. Journal of Pharmaceutical Sciences, 87, 1041–1045.
`Lowther, N., Fox, R., Faller, B., Nick, H.P., Jin, Y., Sergejew, T.,
`Hirschberg, Y., Oberle, R. & Donnelly, H. (1999) In vitro and in
`situ permeability of a ‘second generation’ hydroxypyridinone oral
`
`iron chelator: correlation with physico-chemical properties and
`oral activity. Pharmaceutical Research, 16, 434–440.
`Miller, G.F., Barnard, D.E., Woodward, R.A., Flynn, B.M. & Bulte,
`J.W.M. (1997) Hepatic hemosiderosis in common marmosets,
`Callithrix jacchus: effect of diet on incidence and severity.
`Laboratory Animal Science, 47, 138–142.
`Olivieri, N.F., Nathan, D.G., MacMillan, J.H., Wayne, A.S., Liu, P.P.,
`McGee, A., Martin, M., Koren, G. & Kohen, A.R. (1994) Survival
`in medically treated patients with homozygous b-thalassemia.
`New England Journal of