European Journal of Cancer 37 (2001) 1590–1598
`
`www.ejconline.com
`
`Cremophor EL: the drawbacks and advantages of
`vehicle selection for drug formulation
`
`H. Gelderblom*, J. Verweij, K. Nooter, A. Sparreboom
`
`Department of Medical Oncology, Rotterdam Cancer Institute (Daniel den Hoed Kliniek) and University Hospital Rotterdam,
`3075 EA Rotterdam, The Netherlands
`
`Received 26 February 2001; received in revised form 26 March 2001; accepted 3 May 2001
`
`Abstract
`
`Cremophor EL (CrEL) is a formulation vehicle used for various poorly-water soluble drugs, including the anticancer agent
`paclitaxel (Taxol). In contrast to earlier reports, CrEL is not an inert vehicle, but exerts a range of biological effects, some of which
`have important clinical implications. Its use has been associated with severe anaphylactoid hypersensitivity reactions, hyperlipi-
`daemia, abnormal lipoprotein patterns, aggregation of erythrocytes and peripheral neuropathy. The pharmacokinetic behaviour of
`CrEL is dose-independent, although its clearance is highly influenced by duration of the infusion. This is particularly important
`since CrEL can affect the disposition of various drugs by changing the unbound drug concentration through micellar encapsulation.
`In addition, it has been shown that CrEL, as an integral component of paclitaxel chemotherapy, modifies the toxicity profile of
`certain anticancer agents given concomitantly, by mechanisms other than kinetic interference. A clear understanding of the biolo-
`gical and pharmacological role of CrEL is essential to help oncologists avoid side-effects associated with the use of paclitaxel or
`other agents using this vehicle. With the present development of various new anticancer agents, it is recommended that alternative
`formulation approaches should be pursued to allow a better control of the toxicity of the treatment and the pharmacological
`interactions related to the use of CrEL. # 2001 Elsevier Science Ltd. All rights reserved.
`
`Keywords: Cremophor EL; Formulation vehicles; Paclitaxel; Pharmacokinetics; Pharmacodynamics
`
`1. Introduction
`
`The choice of a suitable pharmaceutical formulation
`is an essential step in anticancer drug development. This
`development starts with the acquisition of a chemical
`entity from either natural sources or entirely synthetic
`routes. Subsequently, the compound is screened for
`cytotoxic activity in vitro and in vivo. Once the screening
`process has been completed the compound should be
`properly pharmaceutically-formulated and produced
`before entering animal toxicology and pharmacokinetic
`studies and subsequently human phase I, II and III
`studies [1,2]. In our opinion, pharmaceutical formula-
`tion is a seriously underrated aspect of anticancer drug
`development.
`With only a few exceptions, most new anticancer
`compounds are initially developed for intravenous (i.v.)
`
`* Corresponding author. Tel.: +31-10-439-1754; fax: +31-10-439-
`1003.
`E-mail address: gelderblom@onch.azr.nl (H. Gelderblom).
`
`use, despite some drawbacks such as the morbidity
`associated with gaining i.v. access, risk of i.v. catheter-
`related infection, thrombosis and extravasation, and
`patients’ preference for oral
`therapy when equally
`effective [3]. Important reasons for choosing i.v. use for
`initial drug development are the fact that usually less
`gastrointestinal
`toxicity occurs,
`there is
`immediate
`100% bioavailability and instantaneous pharmacody-
`namic effects and there is a possibility to modify the
`dosing rate or even halt the infusion if necessary. Solu-
`bility of the compound is a specific demand for i.v.
`administration, even for the newer chemotherapeutic
`agents which are known to be poorly water-soluble.
`Classical solubility approaches, which will be discussed
`subsequently, include the use of colloidal systems, pro-
`drug development or solubilisation techniques.
`Colloidal systems such as liposomes, microcapsules,
`microspheres, nanoparticles or macromolecule com-
`plexes may protect the anticancer drug from premature
`degradation or (chemical) inactivation within the sys-
`temic circulation. Prodrugs are inactive derivatives that
`
`0959-8049/01/$ - see front matter # 2001 Elsevier Science Ltd. All rights reserved.
`P I I : S 0 9 5 9 - 8 0 4 9 ( 0 1 ) 0 0 1 7 1 - X
`
`Abraxis EX2090
`Apotex Inc. and Apotex Corp. v. Abraxis Bioscience, LLC
`IPR2018-00151; IPR2018-00152; IPR2018-00153
`
`

`

`H. Gelderblom et al. / European Journal of Cancer 37 (2001) 1590–1598
`
`1591
`
`release the active drug following spontaneous degrada-
`tion or enzymatic reactions. Solubilisation is the process
`of uptake of drugs through complex formation into for,
`e.g. oligomers of dextrose and fatty acids, through co-
`solvent systems (such as ethanol, polyethyleneglycol and
`glycerol), or through surfactant systems. The surfactant
`systems consist of either amphoteric compounds (e.g.
`lecithin or gelatin), ionic surfactants (e.g. sodium pal-
`mitate) or non-ionic surfactants (e.g. Tween 80 and
`Cremophor EL (CrEL)) [4,5]. This review focuses on
`biological and pharmacological properties of CrEL,
`which is the formulation vehicle of various hydrophobic
`drugs, including the anticancer agent paclitaxel.
`
`2. Paclitaxel formulation
`
`After the identification of paclitaxel as the active
`ingredient in crude ethanolic extracts of the bark of the
`Western Yew tree, Taxus brevifolia, against several
`murine tumours [6], the development of the drug was
`suspended for more than a decade due to problems
`associated with the solubilisation of the drug. Paclitaxel
`is insoluble in water (less than 0.03 mg/ml), slightly
`soluble in octanol, propylene glycol and butanol, solu-
`ble in CrEL, ethanol, methanol, chloroform, acetone
`and ether, and freely soluble in dimethyl acetamide. The
`formulation approach using 50% CrEL and 50% dehy-
`drated ethanol United States Pharmacopeia (USP) was
`chosen for further development [7]. The pharmaceutical
`formulation of paclitaxel
`(paclitaxel; Bristol-Myers
`Squibb) contains 30 mg paclitaxel dissolved in 5 ml of
`this (1:1, v/v) mixture.
`
`The heterogeneous non-ionic surfactant CrEL is a
`white to off-white viscous liquid with an approximate
`molecular weight of  3 kDa and a specific gravity
`(25C/25C) of 1.05–1.06, and it is produced by the
`reaction of castor oil with ethylene oxide at a molar
`ratio of 1:35 [4]. Castor oil is a colourless or pale yellow
`fixed oil obtained from the seeds of Ricinus communis,
`with an extremely high viscosity, and consists mainly of
`the glycerides of ricinoleic,
`isoricinoleic, stearic and
`dihydroxystearic acids. CrEL is usually of highly vari-
`able composition, with the major component identified
`as oxylated triglycerides of ricinoleic acid (i.e. poly-
`oxyethylene glycerol triricinoleate 35) (Fig. 1). Polyvinyl
`chloride (PVC)-free equipment for CrEL administration
`is obligatory, since CrEL is known to leach plasticizers
`from PVC infusion bags and polyethylene-lined tubing
`sets which can cause severe hepatic toxicity [5].
`CrEL is being used as a vehicle for the solubilisation
`of a wide variety of hydrophobic drugs,
`including
`anaesthetics, photosensitisers, sedatives,
`immunosup-
`pressive agents and (experimental) anticancer drugs
`(Table 1). The amount of CrEL administered with these
`drugs averages 5 ml (range, 1.5–10.3 ml), although
`paclitaxel is an exception as the amount of CrEL is
`much higher per administration, approximately 26 ml.
`Therefore, it is important to understand the biological
`and pharmacological behaviour of CrEL, especially in
`the formulation of paclitaxel.
`
`3. Biological effects of Cremophor EL (CrEL)
`
`3.1. Anaphylactic hypersensitivity reactions
`
`The most well known biological effect of paclitaxel
`formulated with CrEL is a clinical acute hypersensitivity
`reaction, characterised by dyspnoea, flushing, rash,
`chest pain, tachycardia, hypotension, angio-oedema,
`and generalised urticaria. Despite premedication, con-
`sisting of high-dose corticosteroids, H1 and H2 antago-
`nists, minor reactions (flushing and rash) still occur in
`
`Table 1
`Examples of surfactant systems using CrEL
`
`Agent
`
`Therapeutic class
`
`Aplidine
`C8KC
`Clanfenur
`Cyclosporin A
`Diazepam
`Didemnin B
`Paclitaxel
`Propofol
`Teniposide
`
`Antineoplastics
`Photosensitisers
`Antineoplastics
`Immunosuppressives
`Sedatives
`Antineoplastics
`Antineoplastics
`Anaesthetics
`Antineoplastics
`
`Amount administereda (ml)
` 1.5
`5.5
`10.3
`3.5
`1.5
`2.0
`25.8
` 7.0
`1.5
`
`structures of paclitaxel
`Fig. 1. Chemical
`oxyethyleneglycerol triricinoleate 35) (b).
`
`(a) of CrEL (poly-
`
`CrEL, Cremophor EL.
`a For an average patient for a single administration of dose.
`
`Abraxis EX2090
`Apotex Inc. and Apotex Corp. v. Abraxis Bioscience, LLC
`IPR2018-00151; IPR2018-00152; IPR2018-00153
`
`

`

`1592
`
`H. Gelderblom et al. / European Journal of Cancer 37 (2001) 1590–1598
`
`41–44% of all patients and major, potentially life-
`threatening, reactions in 1.5–3% [8–10]. Mostly, the
`hypersensitivity reaction occurs within the first two
`courses of paclitaxel and can be prevented by reducing
`the infusion rate. Various (pre)clinical observations,
`discussed below, point to CrEL as the main contributor
`to the hypersensitivity reactions:
`
`and a change in shape of leucocytes in blood smears
`[24]. This suggests that manual blood-cell count analysis
`is warranted after the administration of preparations
`containing CrEL. Whether the observed hyperlipidae-
`mia after CrEL administration increases the risk of
`vascular accidents is, as yet, unknown.
`
`(a) Using an elegant series of in vitro studies, com-
`plement activation by CrEL was found to cause
`the hypersensitivity reaction associated with
`paclitaxel chemotherapy [11], as well as reactions
`seen to other drugs where CrEL was used as a
`vehicle [12,13]. Recently, it was shown that the
`CrEL-induced complement activation in human
`serum was clearly concentration-dependent with
`a minimum activating CrEL level in the order of
`2 ml/ml, a concentration readily achieved clini-
`cally in the plasma following standard doses of
`paclitaxel [14].
`(b) Histamine release in dogs by CrEL was mainly
`caused by one of its (minor) constituents, oleic
`acid [15], whereas the cardiac toxicity attributed
`to paclitaxel, mainly asymptomatic rhythm dis-
`turbances, might also be caused by CrEL
`through a mechanism of histamine release [16].
`(c) Improper mixing of high-dose cyclosporin A
`infusions caused (non-solubilised) CrEL to sink
`to the bottom of vials, producing anaphylactoid
`responses because of highly concentrated CrEL
`at the initial i.v. bolus [17]. All patients allergic to
`i.v. cyclosporin A, tolerated the CrEL-free oral
`formulation [18].
`(d) Finally, the fact that CrEL concentrations are
`lower with prolonged paclitaxel infusion schemes
`(see below) may be an explanation for the lower
`incidence of hypersensitivity reactions with these
`schemes. Collectively, these findings indicate that
`CrEL plays a crucial role in the occurrence of
`hypersensitivity reactions of paclitaxel and other
`drugs using CrEL as a formulation vehicle.
`
`3.2. Lipoprotein patterns and hyperlipidaemia
`
`Lipoprotein alterations accompanying the adminis-
`tration of miconazole formulated with CrEL were
`reported as early as 1977 by Bagnarello and colleagues
`[19]. Later, CrEL was found to alter the buoyant density
`of high-density lipoproteins (HDL) [20] and shift the
`electrophoretic and density gradient HDL to low den-
`sity lipoproteins (LDLs) [21–23]. These authors showed
`that paclitaxel had a strong affinity for the serum lipo-
`protein dissociation products, potentially affecting the
`biodistribution and clearance of the drug. High con-
`centrations of CrEL may also cause hyperlipidaemia,
`possibly resulting in Roulaux formation of erythrocytes
`
`3.3. Neurotoxicity
`
`Axonal degeneration and demyelination, one of the
`principal side-effects of paclitaxel resulting in peripheral
`neuropathy, is also supposedly a biological effect caused
`by CrEL. The plasma levels of CrEL achieved after
`therapeutic doses of paclitaxel and i.v. cyclosporin A
`(also formulated in CrEL), have been shown to produce
`axonal swelling, vesicular degeneration and demyelina-
`tion in rat dorsal ganglion neurones [25]. Interestingly,
`in rats treated with CrEL-free [3H]-paclitaxel, paclitaxel
`was not detectable in the peripheral nervous system,
`indicating that the anticancer drug itself might not be
`responsible for the observed toxicity [26]. The hypoth-
`esis of CrEL-induced neurotoxicity is further supported
`by the fact that i.v. cyclosporin A causes neurotoxicity
`in approximately 25% of patients [27], while this side-
`effect is rarely seen with oral administration. This is also
`consistent with a previous finding that CrEL is not
`absorbed when given orally as a result of intestinal
`degradation [28]. It is also noteworthy that neurological
`symptoms are 10 times less common after treatment
`with docetaxel, a semi-synthetic taxane chemically simi-
`lar to paclitaxel, than with paclitaxel [29]. This could be
`because unlike paclitaxel, docetaxel
`is formulated in
`Tween 80 (i.e. polyoxyethylenesorbitan monooletate),
`again rendering CrEL as the likely cause of the clinical
`observations of neuropathy. The neurotoxic properties
`of CrEL are most likely induced by residual unsaturated
`fatty acids, possibly due to the appearance of peroxida-
`tion products [30]. Therefore, it is suggested that the
`ethoxylated derivatives of castor oil account for most of
`the neuronal damage observed [31].
`
`3.4. Reversal of P-glycoprotein activity
`
`In the early 1990s, several groups independently
`observed that CrEL was able in vitro to modulate the
`activity of P-glycoprotein, a drug-transporting mem-
`brane protein that is elevated in tumour cells having a
`multidrug resistance phenotype [32–35]. More recently,
`similar phenomena have also been described for various
`other non-ionic surfactants, including Tween 80 [36],
`Solutol HS 15 [37] and Triton X-100 [38]. Surprisingly,
`however, multidrug resistance has never been success-
`fully modified in vivo by any non-ionic surfactant,
`including CrEL [39–41]. A possible explanation for this
`lack of in vivo efficacy is the extremely low volume of
`distribution of CrEL, approximately equal
`to the
`
`Abraxis EX2090
`Apotex Inc. and Apotex Corp. v. Abraxis Bioscience, LLC
`IPR2018-00151; IPR2018-00152; IPR2018-00153
`
`

`

`H. Gelderblom et al. / European Journal of Cancer 37 (2001) 1590–1598
`
`1593
`
`volume of the blood compartment, suggesting that con-
`centrations necessary to affect reversal of multidrug
`resistance in vitro are probably not attained in vivo in
`(solid) tumours [42]. Recent pharmacokinetic experi-
`ments conducted in mdr1a P-glycoprotein knockout
`mice support this lack of efficacy, despite high peak
`plasma levels of CrEL [28]. In contrast to treatment in
`solid tumours, the pharmacokinetic selectivity of CrEL
`for the central blood/bone marrow compartment can be
`an advantage in the treatment of haematological malig-
`nancies, in which the expression of P-glycoprotein is
`known to be a principal factor contributing to resistance
`to chemotherapy [43].
`
`3.5. In vitro cytotoxicity
`
`Cytotoxic properties of CrEL in doxorubicin-resistant
`human breast-cancer cell lines were first reported by
`Fja¨ llskog and colleagues [44], and confirmed in various
`human tumour samples [45,46]. It was postulated that
`formation of free radicals by peroxidation of poly-
`unsaturated fatty acids and/or a direct perturbing effect
`in the cell membrane causing fluidity and leakage are
`possible mechanisms contributing to this type of cyto-
`toxicity [47–49]. Using clonogenic assays, however, it
`has been demonstrated that CrEL can antagonise the
`cytotoxicity of paclitaxel by blocking the cell-cycle that
`results in the inhibition of cytokinesis [50]. Thus,
`although CrEL in itself might have some potential to
`affect cell survival, the concentrations required to mod-
`ulate cell growth will also change paclitaxel-mediated
`(and overall) cytotoxicity. In addition, the pharmacoki-
`netic selectivity of CrEL most likely precludes any
`vehicle-mediated change in (solid) tumour cell kill in
`vivo.
`
`4. Pharmacokinetics of CrEL
`
`4.1. Analytical methods
`
`In view of the contribution of CrEL to clinically
`observed effects, and to enable further assessment of the
`impact of its use on paclitaxel pharmacology, the kinetic
`behaviour of CrEL has been studied extensively in
`recent years. For this purpose, a variety of analytical
`methods have been developed. The first assay developed
`for the measurement of CrEL concentrations in plasma
`was based on the ability of CrEL to modulate daunor-
`ubicin efflux in multidrug resistant T-cell
`leukaemia
`VLB100 cells [51]. Later, a more sensitive and reliable
`method was developed, which required only micro-
`volumes (20 ml) of plasma [52]. This method is based on
`the measurement of ricinoleic acid after saponification
`of CrEL followed by precolumn derivatisation and
`reversed-phase high-performance
`liquid chromato-
`
`graphy. Because of the high costs, and time consuming
`nature of both assays a new method, based on a selec-
`tive binding of CrEL to the Coomassie brilliant blue G-
`250 dye in protein-free extracts, was developed [53,54].
`Most recently, an electrochemical detection method was
`developed [55]. Large-scale pharmacokinetic studies
`have only recently been possible with the development
`of these newer methods.
`
`4.2. CrEL Disposition
`
`Clinical pharmacokinetic studies with CrEL following
`3-h paclitaxel
`infusions indicate a dose-independent
`behaviour with a terminal disposition half-life of
`approximately 80 h, with a large range depending on the
`method used for measurement [56,57]. Interestingly,
`with prolongation of the duration of infusion from 1 to
`3 and to 24 h, the CrEL clearance increased from
`approximately 160 to 300 to 400 ml/h/m2, respectively
`[14] (Fig. 2). It thus appears that CrEL shows a linear
`and dose-independent, but schedule-dependent phar-
`macokinetic behaviour, possibly related to the satura-
`tion of serum esterase-mediated metabolic degradation.
`This schedule dependency leads to an increase in sys-
`temic exposure, and thus an increase in the possible
`CrEL-related biological (side-)effects with a shortening
`of the infusion duration. An example of this phenom-
`enon is the higher risk of allergic reactions in the 1-h
`versus 3-h or 24-h infusions of paclitaxel.
`As mentioned earlier, the volume of distribution of
`CrEL is extremely low, implying that the tissue (and
`tumour) delivery of CrEL is probably insignificant. This
`is in line with observations that CrEL levels in normal
`and tumour tissue were not detectable in mice [58].
`Not much is known about the elimination routes of
`CrEL. CrEL may be largely degraded in the blood
`compartment
`by
`serum carboxylesterase-induced
`degradation, similar to that described for Tween 80 [59],
`causing a gradual release of the ricinoleic acid residues
`
`Fig. 2. Schedule-dependency for CrEL clearance (CL) as a function of
`the duration of infusion in cancer patients treated with paclitaxel at
`dose of 135 (CrEL: 11.3 ml/m2), 175 (CrEL: 14.6 ml/m2) and 225
`(CrEL: 18.8 ml/m2) mg/m2 [14].
`
`Abraxis EX2090
`Apotex Inc. and Apotex Corp. v. Abraxis Bioscience, LLC
`IPR2018-00151; IPR2018-00152; IPR2018-00153
`
`

`

`1594
`
`H. Gelderblom et al. / European Journal of Cancer 37 (2001) 1590–1598
`
`attached to the triglyceride structure. It has been shown
`that hepatobiliary elimination of CrEL is a minor elim-
`ination pathway [60]. In addition, the urinary excretion
`of CrEL accounted for less than 0.1% of the adminis-
`tered dose, in spite of its relatively hydrophilic nature
`[61].
`
`4.3. Effects of CrEL on drug disposition
`
`Various studies have shown that CrEL alters the
`pharmacokinetics of many drugs including cyclosporin
`A, etoposide, doxorubicin, a number of photosensitisers
`(e.g. C8KC) and paclitaxel. Initially, the effect of CrEL
`on the disposition of paclitaxel was studied in mice that
`received the drug by i.v. injection at doses of 2, 10 and
`20 mg/kg in the presence of various amounts of CrEL
`[58] (Fig. 3). The paclitaxel clearance of 2.4 l/h/kg at the
`lowest dose level was reduced to 0.33 and 0.15 l/h/kg at
`the 10 and 20 mg/kg dose levels. It was also shown that
`the area under the concentration time curve (AUC) of
`paclitaxel is higher when it is formulated in CrEL com-
`pared with formulation in Tween 80, suggesting that
`CrEL is responsible for the non-linearity of paclitaxel
`disposition. Despite the fact that much higher plasma
`levels of paclitaxel are reached when given in the CrEL-
`containing formulation, the tissue levels of paclitaxel
`were essentially similar with all tested preparations,
`indicating that the profound influence of CrEL is only
`taking place in the central blood compartment. In sub-
`sequent years, numerous causes of this apparent non-
`linear pharmacokinetic behaviour were proposed. It has
`been suggested that CrEL might interfere with P-glyco-
`protein-mediated biliary secretion,
`thereby reducing
`paclitaxel elimination [62]. In the isolated perfused rat
`liver, CrEL inhibited the hepatic elimination of pacli-
`taxel, primarily preventing the drug from reaching the
`sites of metabolism and excretion [63]. However, recent
`studies indicate that drug-transporting P-glycoproteins
`are not essential per se for normal hepatobiliary secre-
`tion of paclitaxel [28,64] and that, as discussed, the dis-
`
`position of CrEL itself limits the potential to modulate
`P-glycoprotein activity in vivo [42].
`More recently, it has been proposed that the effect of
`CrEL on paclitaxel pharmacokinetics is associated with
`encapsulation of the drug within CrEL micelles, causing
`(concentration-dependent) changes in cellular partition-
`ing and blood:plasma concentration ratios of paclitaxel
`[64] (Fig. 4). It was shown that the affinity of paclitaxel
`was
`(in decreasing order) CrEL>plasma>human
`serum albumin, with CrEL present above the critical
`micellar concentration (i.e.  0.01%). Since this effect
`was also observed in the absence of plasma proteins, it
`could not have been caused by altered protein binding
`or by an increased affinity of paclitaxel for protein dis-
`sociation products that are produced by the action of
`CrEL on native lipoproteins [22]. These findings are
`consistent with the hypothesis that paclitaxel can be
`entrapped within micelles (composed primarily of poly-
`ethyleneglycerol triricinoleate) and that these micelles
`act as the principal carrier of paclitaxel in the systemic
`circulation. The percentage of total paclitaxel trapped in
`micelles increases disproportionately with higher doses
`of CrEL administered.
`The hypothesis that the non-linear pharmacokinetics
`of paclitaxel is related to time-varying CrEL concentra-
`tions was recently confirmed in a group of cancer
`patients all receiving increasing doses of 135, 175 and
`225 mg/m2 [65]. Again, the plasma clearance of pacli-
`taxel turned out to be dose-dependent with the slowest
`clearance at the highest dose level (Fig. 5). In line with
`the in vitro data, the non-linear disposition of paclitaxel
`in the plasma appeared to be an artifact caused by dose-
`related levels of CrEL in blood [65]. Therefore, the non-
`micellar bound or unbound fraction of paclitaxel
`in
`plasma might be a better pharmacokinetic parameter to
`predict toxicity and to guide dosing of paclitaxel, since it
`is generally acknowledged that the unbound fraction of
`a drug is capable of diffusing across biological barriers
`and interacting with essential structures in (tumour) tis-
`sues [66].
`
`Fig. 3. Effect of the formulation vehicle (CrEL, Tween 80 or dime-
`thylacetamide (DMA)) on paclitaxel concentration in female FVB
`mice receiving paclitaxel at a dose of 10 mg/kg [58].
`
`Fig. 4. Effect of the CrEL concentration on the fraction of unbound
`paclitaxel (fu) in human plasma [56].
`
`Abraxis EX2090
`Apotex Inc. and Apotex Corp. v. Abraxis Bioscience, LLC
`IPR2018-00151; IPR2018-00152; IPR2018-00153
`
`

`

`H. Gelderblom et al. / European Journal of Cancer 37 (2001) 1590–1598
`
`1595
`
`4.4. Effects of CrEL on anticancer drug
`pharmacodynamics
`
`Besides its effect on (anticancer) drug disposition, the
`use of CrEL has also been associated with alterations in
`the pharmacodynamic outcome of a number of drugs
`co-administered with paclitaxel. For example, clinical
`combination chemotherapy studies with paclitaxel and
`cisplatin revealed important sequence-dependent differ-
`ences in toxicity, with less myelotoxicity when the tax-
`ane was given before cisplatin, that could not be linked
`to consistent changes
`in pharmacokinetics
`[67,68].
`Interestingly, no toxicity- or pharmacokinetic-sequence
`interaction was seen in the combination of carboplatin
`and paclitaxel [69]. One of the mechanisms underlying
`these clinically important interactions appears to be
`related to a selective inhibition of cisplatin accumula-
`tion in peripheral blood and bone marrow cells by
`CrEL, without affecting antitumour activity [70–72]
`(Fig. 6). In addition,
`it has been demonstrated that
`CrEL markedly reduced the level of serum haemato-
`poietic inhibitory activity, which resulted in a decrease
`in femoral bone marrow cellularity and an upregulation
`of B and T cells, as well as a transiently elevated inci-
`dence of both primitive and committed haematopoietic
`progenitor cells within several hours after injection [73].
`Taken together, these findings suggest that the unusual
`clinical observations noted with cisplatin-paclitaxel
`combination therapy are due to a (myelo)protective
`interaction with cisplatin by CrEL. Since CrEL doses of
`up to 30 ml/m2 as a 3-h infusion can be safely admi-
`nistered [74], it is anticipated that an improvement in
`the therapeutic window for cisplatin could be obtained
`by re-formulation of this agent with CrEL. This con-
`cept is currently under further investigation (data not
`shown).
`
`In contrast to the potential benefits obtained from the
`use of CrEL in the case of cisplatin chemotherapy,
`combination studies conducted with the combination of
`paclitaxel and doxorubicin have shown that CrEL can
`be linked to clinically important kinetic and dynamic
`interactions that greatly impact upon the overall toxi-
`city profiles. Two independent studies have shown pro-
`found decreases in doxorubicin clearance, particularly
`when paclitaxel was given prior to the anthracycline
`[62,75], resulting in substantially worse mucositis and
`haematological
`toxicity. Therefore,
`it
`is advised to
`administer doxorubicin first in this combination. In
`addition, it is expected that even minor modifications in
`the duration of the infusion or dose may lead to unpre-
`dictable pharmacodynamic consequences. Similar inter-
`actions with CrEL have also been described for
`etoposide, epirubicin and the irinotecan metabolite SN-
`38 [76]. Clearly, the magnitude of these interactions
`depend largely on the combination drug involved, the
`paclitaxel and CrEL dose administered, and the dura-
`tion of infusion administration sequence used for the
`combined agents.
`
`5. Conclusions and future perspectives
`
`Recent investigations have revealed that CrEL, a
`widely used formulation vehicle, is a biologically and
`pharmacologically active ingredient of various com-
`mercially available drugs. For example, when used in
`paclitaxel
`administrations,
`an exceptionally
`large
`amount of CrEL is inevitably co-administered with the
`i.v. infusions, causing important biological events that
`can lead to serious acute hypersensitivity reactions and
`neurological toxicity, depending on the dose and dura-
`tion of infusion. In addition, the substantial effects of
`
`Fig. 5. Effect of dose on paclitaxel plasma (circles) and whole blood
`area under the concentration curve (AUC) (triangles)
`in cancer
`patients treated with paclitaxel at three different dose levels [65]. Please
`note that the higher the AUC, the slower the clearance of a drug.
`
`Fig. 6. Influence of CrEL on anticancer drug pharmacodynamics.
`Bars indicate the amount of drug accumulated in peripheral blood
`leucocytes (PBL) (expressed as percentage relative to a 100% control)
`for 2-h incubations with cisplatin (CDDP), carboplatin (Carbo),
`doxorubicin (dox), and epirubicin (Epi) ([70], data not shown). NS,
`non significant.
`
`Abraxis EX2090
`Apotex Inc. and Apotex Corp. v. Abraxis Bioscience, LLC
`IPR2018-00151; IPR2018-00152; IPR2018-00153
`
`

`

`1596
`
`H. Gelderblom et al. / European Journal of Cancer 37 (2001) 1590–1598
`
`CrEL on the disposition of other co-administered drugs
`(e.g. anthracyclines) with a narrow therapeutic window
`in poly-chemotherapeutic regimens also can be poten-
`tially hazardous to the patient. In contrast, the in vitro
`and in vivo observations of myeloprotective effects rela-
`ted to the use of CrEL in combination with some
`agents, such as cisplatin, might be used to re-formulate
`such agents with CrEL in order to achieve an optimisa-
`tion of their therapeutic window.
`In view of the inherent problems associated with the
`use of CrEL, it can be anticipated that there may be a
`therapeutic advantage from using paclitaxel formula-
`tions in which CrEL is absent. Such new formulation
`approaches should clearly be pursued to allow a better
`control of the systemic (CrEL-mediated) toxicity and
`pharmacokinetic interactions observed with numerous
`agents given in combination with the taxane. Obviously,
`alternative paclitaxel
`formulations should allow the
`drug to be delivered at adequate doses, and the pre-
`paration should be stable for several hours to allow
`handling in the clinical setting. Currently, a large variety
`of new (CrEL-free) formulation vehicles for paclitaxel
`are in (pre)clinical development,
`including co-solvent
`systems (Tween 80/ethanol/Pluronic L64), water-soluble
`polymers (e.g. polyethylene glycols), emulsions (e.g.
`triacetin), liposomes, cyclodextrines, nanocapsules and
`microspheres (reviewed in Ref. [76]). Although one of
`these preparations might eventually replace the current
`paclitaxel formulation, for the coming years we will
`have to cope with the drawbacks of CrEL in the clinical
`setting. It is also important to keep in mind that CrEL is
`widely used not only by oncologists, but also by anaes-
`thetists and in transplantation medicine, and that few
`physicians are aware of the biological and pharmacolo-
`gical activity of the compound. It is thus essential to
`find other ways to solubilise the drugs formulated in
`CrEL but also, in some cases, to explore the potential
`positive effects of this castor oil derivative. It is of note
`that a proper understanding of all the biological effects
`of the formulation would presumably have avoided the
`major delays in development of paclitaxel as an antic-
`ancer agent. This underscores the importance of the
`choice of vehicle in drug development in general.
`
`References
`
`1. Jonkman-de Vries JD, Flora KP, Bult A, Beijnen JH. Pharma-
`ceutical development of (investigational) anticancer agents for
`parenteral use — a review. ECC Newsletter 1994, 3, 11–14.
`2. Pinedo HM. Review: Development of new anticancer drugs. Med
`Oncol Tumor Pharmacother 1986, 3, 63–69.
`3. Liu G, Fransen E, Fitch MI, Warner El. Patient preferences for
`oral versus intravenous palliative chemotherapy. J Clin Oncol
`1997, 15, 110–115.
`4. Hoffman H. Polyoxythylenglycerol triricinoleat 35 DAC 1979.
`Pharm Zeit 1984, 129, 1730–1733.
`
`5. Goldspiel BR. Guidelines for administration. In: McGuire WP,
`Rowinsky EK, eds. Paclitaxel in Cancer Treatment. New York,
`Marcel Dekker, 1995, 175–186.
`6. Wani MC, Taylor HL, Wall ME, Coggon P, McPhail AT. Plant
`antitumor agents. VI: the isolation and structure of Taxol, a
`novel antileukemic and antitumor agent from Taxus brevifolia. J
`Am Chem Soc 1971, 93, 2325–2327.
`7. Adams JD, Flora KP, Goldspiel BR, Wilson JW, Arbuck SGI.
`Taxol: a history of pharmaceutical development and current
`pharmaceutical concerns. J Natl Cancer Inst Monogr 1993, 15,
`141–147.
`8. Rowinsky EK, Eisenhauer EA, Chaudhry V, Chaudrhy V,
`Arbuck SG. Clinical
`toxicities encountered with paclitaxel
`(Taxol). Semin Oncol 1993, 20, 1–15.
`9. Weiss RB, Donehower RC, Wiernik PH, et al. Hypersensitivity
`reactions from Taxol. J Clin Oncol 1990, 8, 1263–1268.
`10. Eisenhauer EA, ten Bokkel Huinink WW, Swenerton KD, et al.
`European-Canadian randomized trial of paclitaxel in relapsed
`ovarian cancer: high dose versus low-dose and long versus short
`infusion. J Clin Oncol 1994, 12, 2654–2666.
`11. Szebeni J, Muggia FM, Alving CR. Complement activation by
`Cremophor EL as a possible contributer to hypersensitivity to
`paclitaxel: an in vitro study. J Natl Cancer Inst 1998, 90, 300–306.
`12. Watkins J, Ward AM, Appleyard TN. Adverse reactions to
`intravenous anaesthetic induction agents. Br Med J 1997, 2,
`1084–1085.
`13. Hu¨ ttel MS, Schou Olesen A, Stoffersen E. Complement-mediated
`reactions to diazepam with Cremophor as solvent (stesolid MR).
`Br J Anaesth 1980, 52, 77–79.
`14. van Zuylen L, Gianni L, Verweij J, et al. Interrelationships of
`paclitaxel disposition,
`infusion duration and Cremophor EL
`kinetics in cancer