REVIEW ARTICLE
`
`Clin Pharmacokinet 2003; 42 (7): 665-685
`0312-5963/03/0007-0665/$30.00/0
`
`© Adis Data Information BV 2003. All rights reserved.
`
`Pharmacological Effects of
`Formulation Vehicles
`Implications for Cancer Chemotherapy
`
`Albert J. ten Tije,1 Jaap Verweij,1 Walter J. Loos1 and Alex Sparreboom1,2
`1 Department of Medical Oncology, Erasmus MC – Daniel den Hoed Cancer Center,
`Rotterdam, The Netherlands
`2 National Cancer Institute, Bethesda, Maryland, USA
`
`Contents
`Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 665
`1. Physicochemical Properties of Surfactants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 667
`2. Biological Properties of Surfactants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 668
`2.1 Acute Hypersensitivity Reactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 668
`2.2 Peripheral Neurotoxicity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 669
`2.3 Dyslipidaemia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 670
`2.4 Inhibition of P-Glycoprotein Activity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 670
`2.5 Intrinsic Antitumour Effects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 671
`3. Pharmacological Properties of Surfactants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 671
`3.1 Analytical Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 671
`3.2 Pharmacokinetics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 672
`4. Modulation of Drug Disposition Patterns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 673
`4.1 Intravenous Administration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 673
`4.2 Extravascular Routes of Administration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 677
`5. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 679
`
`Abstract
`
`The non-ionic surfactants Cremophor® EL (CrEL; polyoxyethyleneglycer-
`(Tween® 80; polyoxy-
`triricinoleate 35)
`and polysorbate 80
`ol
`ethylene-sorbitan-20-monooleate) are widely used as drug formulation vehicles,
`including for the taxane anticancer agents paclitaxel and docetaxel. A wealth of
`recent experimental data has indicated that both solubilisers are biologically and
`pharmacologically active compounds, and their use as drug formulation vehicles
`has been implicated in clinically important adverse effects, including acute
`hypersensitivity reactions and peripheral neuropathy.
`CrEL and Tween® 80 have also been demonstrated to influence the disposition
`of solubilised drugs that are administered intravenously. The overall resulting
`effect is a highly increased systemic drug exposure and a simultaneously
`decreased clearance, leading to alteration in the pharmacodynamic characteristics
`of the solubilised drug. Kinetic experiments revealed that this effect is primarily
`caused by reduced cellular uptake of the drug from large spherical micellar-like
`structures with a highly hydrophobic interior, which act as the principal carrier of
`circulating drug. Within the central blood compartment, this results in a profound
`
`AVENTIS EXHIBIT 2094
`Mylan v. Aventis, IPR2016-00712
`
`

`
`666
`
`ten Tije et al.
`
`alteration of drug accumulation in erythrocytes, thereby reducing the free drug
`fraction available for cellular partitioning and influencing drug distribution as
`well as elimination routes. The existence of CrEL and Tween® 80 in blood as
`large polar micelles has also raised additional complexities in the case of combi-
`nation chemotherapy regimens with taxanes, such that the disposition of several
`coadministered drugs, including anthracyclines and epipodophyllotoxins, is sig-
`nificantly altered. In contrast to the enhancing effects of Tween® 80, addition of
`CrEL to the formulation of oral drug preparations seems to result in significantly
`diminished drug uptake and reduced circulating concentrations.
`The drawbacks presented by the presence of CrEL or Tween® 80 in drug
`formulations have instigated extensive research to develop alternative delivery
`forms. Currently, several strategies are in progress to develop Tween® 80- and
`CrEL-free formulations of docetaxel and paclitaxel, which are based on pharma-
`ceutical (e.g. albumin nanoparticles, emulsions and liposomes), chemical (e.g.
`polyglutamates, analogues and prodrugs), or biological (e.g. oral drug administra-
`tion) strategies. These continued investigations should eventually lead to more
`rational and selective chemotherapeutic treatment.
`
`Paclitaxel and docetaxel are hydrophobic antine-
`oplastic agents demonstrating significant antitumour
`activity against a broad spectrum of human malig-
`nancies. After the identification of paclitaxel as the
`active ingredient in crude ethanolic extracts of the
`bark of the Pacific yew tree, Taxus brevifolia L, the
`development of this drug was suspended for over a
`decade because of problems in drug formulation.[1]
`After investigation of a large variety of excipients to
`enable parenteral administration of paclitaxel, the
`formulation approach using the polyoxyethylated
`castor oil derivative, Cremophor® EL1 (CrEL; poly-
`oxyethyleneglycerol triricinoleate 35), represented
`the most viable option.[2] Currently, paclitaxel is
`commercially available as vials containing 30mg of
`drug dissolved in 5mL of CrEL/dehydrated ethanol
`USP (1 : 1 by volume). CrEL is widely used as a
`vehicle for the solubilisation of a number of other
`hydrophobic drugs, including anaesthetics, vita-
`mins, sedatives, photosensitisers, immunosuppres-
`sives and (experimental) anticancer drugs (table I).
`The amount of CrEL per administration of paclitaxel
`is relatively high, and therefore its toxicological and
`pharmacological behaviour in the context of chemo-
`
`Table I. Examples of clinical drug preparations using Cremophor®
`EL or Tween® 80
`Agent
`
`Therapeutic class
`
`Amount
`administered (mL)a
`
`Cremophor® EL
`Kahalalide F
`Diazepam
`Aplidine
`Teniposide
`Didemnin B
`Cyclosporin
`C8KC
`Propofol
`Clanfenur
`BMS-247550
`DHA-paclitaxel
`Paclitaxel
`
`Antineoplastic
`Sedative
`Antineoplastic
`Antineoplastic
`Antineoplastic
`Immunosuppressive
`Photosensitiser
`Anaesthetic
`Antineoplastic
`Antineoplastic
`Antineoplastic
`Antineoplastic
`
`~0.5b
`1.5
`~1.5b
`1.5
`2.0
`3.5
`5.5
`7.0
`10.3
`~10b
`19.9
`25.8
`
`Tween® 80
`0.1
`Antineoplastic
`Carzelesin
`2.0
`Antineoplastic
`Docetaxel
`2.0
`Antineoplastic
`Etoposide
`a For an average patient with a body surface area of 1.77m2.
`b
`Investigational agent for which recommended dose has not
`yet been established.
`
`therapeutic treatment with paclitaxel is of major
`importance.[3]
`
`1 Use of tradenames is for product identification only and does not imply endorsement.
`
`© Adis Data Information BV 2003. All rights reserved.
`
`Clin Pharmacokinet 2003; 42 (7)
`
`

`
`Drug Disposition and Formulation Vehicles
`
`667
`
`1. Physicochemical Properties
`of Surfactants
`
`The structurally related taxane docetaxel is pre-
`pared by chemical manipulation of 10-deacetyl-bac-
`catin III, an inactive precursor isolated from the
`needles of the European yew tree, Taxus baccata
`CrEL is a white to off-white viscous liquid with
`L.[4] Like paclitaxel, it is a potent inhibitor of cell
`an approximate molecular weight of 3000Da and a
`specific gravity of 1.05–1.06. It is produced by the
`replication by stabilisation of the microtubule cytos-
`reaction of castor oil with ethylene oxide at a molar
`keleton. For clinical use, this slightly less hydropho-
`ratio of 1 : 35. Castor oil is a colourless or pale
`bic agent is formulated in another polyoxyethylated
`yellow fixed oil obtained from the seeds of Ricinus
`surfactant, polysorbate 80 (Tween® 80). The clini-
`communis, with an extremely high viscosity, and
`cally used formulation consists of 80mg of docetax-
`consists mainly of the glycerides of ricinoleic,
`el in 2mL of undiluted Tween® 80. This non-ionic
`isoricinoleic, stearic, dihydroxystearic and oleic
`surfactant is also used to solubilise several other
`acids. The non-ionic surfactant produced from cast-
`anticancer drugs, including etoposide and mi-
`or oil is usually of highly variable composition, with
`nor-groove-binding cyclopropylpyrroloindole ana-
`the major component (about 87%) identified as oxy-
`ethylated triglycerides of ricinoleic acid (figure 1).
`logues such as carzelesin (table I).
`As a result of the heterogeneous nature of castor oil
`In recent years, substantial evidence has been
`and its variable composition, the polyoxyethylated
`generated suggesting that CrEL and Tween® 80 are
`components of CrEL have been poorly character-
`biologically and pharmacologically active com-
`ised. Using fractionation by cyclodextrin-modified
`pounds. In this report, we will review the physico- micellar electrokinetic capillary chromatography
`chemical and biological properties of both non-ionic
`(CD-MEKC) and UV detection, in combination
`surfactants, with a focus on their effects on the with delayed extraction matrix-assisted laser de-
`disposition characteristics of the carried drugs and
`sorption/ionisation time of flight mass spectrometry
`that of other agents administered concomitantly.
`(DE-MALDITOF-MS), a more detailed structural
`
`H2C(CH2CH2O)xOCO(CH2)7CH=CHCH2CHOH(CH2)5CH3
`
`HC(CH2CH2O)YOCO(CH2)7CH=CHCH2CHOH(CH2)5CH3
`
`H2C(CH2CH2O)zOCO(CH2)7CH=CHCH2CHOH(CH2)5CH3
`
`(x + y + z ~ 35)
`
`a
`
`b
`
`HO(CH2CH2O)w
`
`(CH2CH2O)xOH
`
`(CH2CH2O)yOH
`
`O
`
`(W + X + Y + Z ~ 20)
`Fig. 1. Chemical structures of the primary constituents of (a) Cremophor® EL (polyoxyethyleneglycerol triricinoleate 35) and (b) Tween® 80
`(polyoxyethylene-20-monooleate).
`
`(CH2CH2O)zOCO(CH2)7CH=CH(CH2)7CH3
`
`© Adis Data Information BV 2003. All rights reserved.
`
`Clin Pharmacokinet 2003; 42 (7)
`
`

`
`668
`
`ten Tije et al.
`
`ated complement activation plays a significant role.
`elucidation and a semiquantitative analysis of CrEL
`components was achieved recently.[5] These inves-
`It has been postulated that due to binding of natural-
`tigations indicated that the elimination of water
`ly occurring anticholesterol antibodies to the hy-
`from
`ricinoleic acid during the synthesis of
`droxyl-rich surface of CrEL micelles, complement
`CrEL leads to various previously unidentified
`C3 is activated, leading to the clinical signs of
`species,
`including (glycerol-) polyoxyethylene-
`hypersensitivity reactions.[16] The CrEL-induced
`Δ9,11-didehydrostearate. It is noteworthy that equip-
`complement activation is clearly concentration de-
`ment used for intravenous administration of CrEL
`pendent, with a minimum CrEL concentration of
`approximately 2 μL/mL being required, a concentra-
`should be free of polyvinylchloride, since CrEL is
`capable of leaching phthalate-type plasticisers from tion readily achieved in plasma of cancer patients
`polyvinylchloride infusion bags and polyethylene-
`following standard doses of paclitaxel.[17] This ex-
`lined tubing sets, which can cause severe hepatic
`plains why slowing down the infusion rate of
`toxicity.[6,7]
`paclitaxel formulated with CrEL can alleviate hy-
`In contrast to CrEL, Tween® 80 is a relative
`persensitivity symptoms, and also explains the need
`homogenous and reproducible, amber-coloured, vis-
`for proper dissolution of CrEL-containing drugs to
`cous liquid (270–430 centistokes) with a molecular
`prevent large variations in CrEL infusion rate lead-
`weight of 1309.7Da and a density of 1.064 g/mL.
`ing to unpredictable reactions.[18] A recent investiga-
`The base chemical name of the major compo-
`tion into the structure-activity relationships of
`nent of Tween® 80 is polyoxyethylene-20-sorbitan
`surfactant-mediated complement activation has
`monooleate (figure 1), which is structurally similar
`shown that several analogues of CrEL have reduced
`to the polyethyleneglycols. Like most non-ionic
`ability to induce complement activation as measured
`surfactants, CrEL and Tween® 80 are capable of
`by a decrease in serum concentrations of the SC5b-9
`forming micelles in aqueous solution, with critical marker (figure 2). Additional clinical studies will be
`micellar concentrations of 0.009% (weight/volume)
`required to evaluate the clinical utility of some of
`and 0.01% (weight/volume), respectively, in prote-
`these substitute vehicles for CrEL-containing drugs.
`in-free aqueous solution.[8]
`In studies with dogs it was demonstrated that
`CrEL, mainly its minor free fatty acid constituents
`such as oleic acid, can cause histamine release.[20]
`Despite premedication with corticosteroids and his-
`tamine H1 and H2 blockers, minor reactions (e.g.
`flushing and rash) still occur in approximately 40%
`The most extensively described biological effect
`of all patients,[21-24] with major potentially life-
`of drugs formulated with CrEL is an acute hypersen-
`threatening reactions observed in 1.5–3% of treated
`sitivity reaction characterised by dyspnoea, flush-
`patients.[9]
`ing, rash, chest pain, tachycardia, hypotension,
`Oleic acid is also present in Tween® 80, and thus
`angioedema and generalised urticaria, and this reac-
`may be a cause of hypersensitivity reactions to
`tion has been attributed to CrEL.[9-12] Nevertheless,
`docetaxel therapy or other therapies using drugs
`allergic reactions to taxanes formulated without
`with Tween® 80 as a solvent. Patients allergic to
`CrEL have been reported as well,[13] suggesting that
`intravenously administered etoposide tolerated the
`some functionality of the taxane molecule contrib-
`oral formulation, which is devoid of Tween® 80,
`utes, in part, to the observed effect. Already in the
`very well.[25-28] The early clinical studies with
`1970s it was demonstrated that CrEL-containing
`docetaxel revealed an incidence of hypersensitivity
`drug preparations (e.g. rectal diazepam) can cause
`complement activation.[14,15] The mechanistic basis
`reactions ranging from 5–40%, with only a minority
`of more than grade 2 on the 4-point scale of the
`for this effect has not been fully elucidated, but a
`number of seminal studies indicate that CrEL-medi- National Cancer Institute common toxicity crite-
`
`2. Biological Properties of Surfactants
`
`2.1 Acute Hypersensitivity Reactions
`
`© Adis Data Information BV 2003. All rights reserved.
`
`Clin Pharmacokinet 2003; 42 (7)
`
`

`
`Drug Disposition and Formulation Vehicles
`
`669
`
`None
`
`Tween® 80
`
`Sorporol 230
`
`Sorporol 120Ex
`
`Cre m ophor® EL
`
`Riciporol 335
`
`Aceporol 345-T
`
`Aceporol 460
`
`20
`
`15
`
`10
`
`5
`
`0
`
`SC5b-9 (μg/mL)
`
`Fig. 2. Vehicle-mediated complement activation in human serum by Cremophor® EL, Tween® 80 and some structurally related analogues.
`Experiments were based on 50μL human serum incubations (45 minutes at 37°C) in the presence of each respective vehicle at a
`concentration of 10 μL/mL. The complement activation marker SC5b-9 was measured by enzyme-linked immunoassay. Data are presented
`as mean values ± SD of triplicate observations and were obtained from Loos et al.[19]
`
`2.2 Peripheral Neurotoxicity
`
`ria.[29-31] Hypersensitivity reactions to docetaxel
`degeneration and demyelinisation in rat dorsal root
`ganglion neurons.[39,40] The precise mechanism of
`therapy can be effectively ameliorated by premedi-
`cation with corticosteroids and antihistamines,[32]
`this CrEL-induced neurotoxicity remains unclear,
`consistent with a role of histamine in its aetiology. A but recent work has indicated that unsaturated fatty
`comparative evaluation of paclitaxel- and docetaxel-
`acids may cause neurotoxicity, possibly due to the
`mediated non-haematological toxicities, with the
`appearance of peroxidation products.[39,40] This sug-
`drugs given in an every 21-day schedule, is provided
`gests that the ethoxylated derivatives of castor oil
`in table II.
`probably account for most of the neuronal damage
`in addition to the presence of residual ethylene oxide
`residues.[41]
`A detailed investigation into neurological ad-
`verse effects associated with docetaxel chemothera-
`py was recently performed in a group of 186 pa-
`tients.[42] Twenty-one patients developed mild to
`moderate sensory neuropathy on treatment at a wide
`range of cumulative doses (50–750 mg/m2) and dose
`levels (10–115 mg/m2). Ten of these patients also
`developed weakness in proximal and distal extremi-
`ties of varying degree. Nine of the 21 patients had
`received neurotoxic chemotherapy before, and 16
`were treated with docetaxel at a dose level of
`100–115 mg/m2. This suggests that docetaxel pro-
`duces a mild and predominantly sensory neuropathy
`in a high proportion of treated patients. This adverse
`effect appears to be dose-dependent and may be
`severe and disabling at higher dose levels.[42-44] Cor-
`ticosteroid comedication does not prevent docetax-
`el-induced neuropathy.[45]
`
`A well-known adverse effect of agents formulat-
`ed in CrEL is peripheral neurotoxicity,[35] but it is
`less well acknowledged that CrEL may play an
`important causative role. In a study performed with
`radiolabelled paclitaxel
`in rats, no detectable
`paclitaxel could be demonstrated in the peripheral
`nerve fibres,[36] but electrophysiological studies in
`patients with neuropathy after treatment with
`paclitaxel have shown evidence of both axonal de-
`generation and demyelinisation.[37] In approximate-
`ly 25% of patients treated with cyclosporin, neuro-
`toxicity is noted.[38] This adverse effect is never
`induced by oral formulations of cyclosporin, which
`is consistent with observations that CrEL is not
`absorbed intact when given orally. Moreover, CrEL
`plasma concentrations achieved with therapeutic
`doses of intravenous paclitaxel or cyclosporin have
`been shown to produce axonal swelling, vesicular
`
`© Adis Data Information BV 2003. All rights reserved.
`
`Clin Pharmacokinet 2003; 42 (7)
`
`

`
`670
`
`ten Tije et al.
`
`Table II. Comparative nonhaematological toxicity of paclitaxel and
`docetaxela
`
`Adverse effect
`
`Incidence (%)
`paclitaxel
`(n = 812)
`
`docetaxel
`(n = 2045)
`
`41
`2
`
`0
`0
`
`2
`0
`
`60
`3
`
`15
`2
`
`64
`6.5
`
`31
`2.5
`
`49
`4
`
`Hypersensitivity reactionsb
`All
`Severe (at least grade 3)
`Fluid retentionb,c
`All
`Severe
`Nail changesd
`All
`Severe (at least grade 3)
`Peripheral neuropathye
`All
`Severe (at least grade 3)
`Skin toxicityf
`48
`2
`All
`5
`0
`Severe (at least grade 3)
`a Data represent overall incidence as percentage of patients
`with solid tumours treated with single-agent regimens
`containing either paclitaxel formulated in a mixture of
`Cremophor® EL and ethanol at doses of 135–300 mg/m2 or
`docetaxel formulated in Tween® 80 at a dose of 100 mg/m2,
`given every 21 days.[33,34]
`b All patients received a 3-day dexamethasone premedication
`(docetaxel, n = 92).
`c Characterised by one or more of the following events: poorly
`tolerated peripheral oedema, generalised oedema, pleural
`effusion requiring urgent drainage, dyspnoea at rest, cardiac
`tamponade, or pronounced abdominal distension (due to
`ascites).
`d Mostly changes in pigmentation or discoloration of the nail
`bed.
`e Mostly peripheral sensory (numbness, paraesthesias, loss of
`proprioception), axonal degeneration and secondary
`demyelination.
`Primarily involves pressure or trauma sites (e.g. hands, feet
`and elbows).
`
`f
`
`paclitaxel and docetaxel, with formulation vehicles
`contributing to the overall picture to a different
`extent.
`
`2.3 Dyslipidaemia
`
`In the mid-1970s, lipoprotein alterations caused
`by CrEL were mentioned for the first time.[48] Later,
`CrEL was found to alter the buoyant density of high-
`density lipoprotein (HDL) and shift the electropho-
`retic and density gradient HDL to low-density lipo-
`protein (LDL).[49-52] These authors demonstrated the
`strong affinity of paclitaxel for serum lipoprotein
`degradation products, potentially affecting
`the
`pharmacokinetics of the drug by altering protein
`binding characteristics. High concentrations of
`CrEL may also cause dyslipidaemia, possibly result-
`ing in rouleaux formation of erythrocytes.[53] Al-
`though cyclosporin is known for its atherosclerosis-
`inducing capacities, it remains unclear if the ob-
`served hyperlipidaemia after CrEL administration is
`contributing to this risk for vascular accidents. In
`vivo studies of the effects of cyclosporin on the de-
`endothelialised carotid artery of New Zealand White
`rabbits treated with therapeutic doses of cyclosporin
`(15 mg/kg/day) or with a vehicle control (CrEL)
`revealed intimal proliferation in both groups.[54]
`Mean plasma cholesterol levels were moderately
`increased in both groups. Although this may have
`contributed
`to
`foam cell
`formation
`in
`the
`cyclosporin-treated animals, it was not the sole de-
`terminant, as foam-cell-rich lesions were not ob-
`served in animals receiving only CrEL. In contrast,
`Tatou et al. observed significant adverse effects of
`CrEL on endothelial function and vascular muscle
`on isolated and perfused rat hearts, leading to a
`reduction of coronary flow and aortic output.[55] The
`potential clinical implications with respect to these
`CrEL-related phenomena remain unknown.
`
`Tween® 80 is capable of producing vesicular
`degeneration. This property depends upon the poly-
`ethylene substitutions produced by reaction of the
`polyol compound with ethylene oxide. However, the
`incidence of neurotoxicity during treatment with
`docetaxel is much lower as compared to that of
`P-glycoprotein is a drug transporting membrane
`paclitaxel (table II).[46,47] Furthermore, the Tween®
`protein, and its expression is increased in tumour
`80-containing epipodophyllotoxin etoposide is not
`cells having a multidrug resistance phenotype.[56,57]
`known to be neurotoxic. This suggests that the aeti-
`Several in vitro studies in the early 1990s observed
`ology of taxane-induced neuropathy is different for modulation of the activity of P-glycoprotein by
`
`2.4 Inhibition of P-Glycoprotein Activity
`
`© Adis Data Information BV 2003. All rights reserved.
`
`Clin Pharmacokinet 2003; 42 (7)
`
`

`
`Drug Disposition and Formulation Vehicles
`
`671
`
`3. Pharmacological Properties
`of Surfactants
`
`3.1 Analytical Methods
`
`CrEL.[58-60] Later, similar phenomena were observed
`for various other non-ionic surfactants, including
`Tween® 80,[61,62] Solutol HS 15[63] and Triton
`X-100.[64] However, in vivo studies never demon-
`strated reversal of multidrug resistance by any non-
`ionic surfactant, including CrEL and Tween®
`80.[65-67] The extremely low volume of distribution
`of CrEL and the rapid degradation of Tween® 80 in
`vivo are the likely explanations for this lack of in
`vivo efficacy (see section 3.2). Indeed, the volume of
`distribution of CrEL is approximately equal to the
`volume of the blood compartment, suggesting that
`concentrations necessary to affect reversal of mul-
`tidrug resistance in vitro are not reached in vivo in
`solid tumours.[68] However, it should be noted that
`the pharmacokinetic selectivity of CrEL for the cen-
`tral blood and bone marrow compartment can pro-
`vide an advantage to treatment of haematological
`malignancies with resistance
`to chemotherapy
`caused by elevated P-glycoprotein expression.[69]
`
`2.5 Intrinsic Antitumour Effects
`
`At present, a large variety of analytical proce-
`dures are available for clinical pharmacokinetic
`studies with CrEL and Tween® 80. The first assay
`developed for measurement of CrEL concentrations
`in patient material was based on the ability of
`this vehicle to modulate daunorubicin efflux in
`multidrug
`resistant T-cell
`leukaemia VLB100
`cells.[83] Alternatively, a more sensitive and reliable
`method was developed
`that required sample
`volumes of only 20μL.[84] This method is based on
`measurement of ricinoleic acid after base-in-
`duced hydrolysis (saponification) of CrEL fol-
`lowed by an acylchloride formation, precolumn
`derivatisation with naphthylamine, and reversed-
`phase high-performance liquid chromatography
`(HPLC) to detect N-ricinoleoyl-1-naphthylamine
`at 280nm. Because of the high costs and the time-
`consuming nature of both assays, a new method,
`based on a selective binding of CrEL to the Coomas-
`Cell-growth inhibitory properties of CrEL were
`sie Brilliant Blue G-250 dye in protein-free extracts
`first observed by Fj¨allskog et al. in doxorubicin- was developed for human plasma samples.[85,86] This
`resistant human breast cancer cell lines,[70,71] and method has also been used to measure Tween® 80
`were later confirmed in other malignant cell
`concentrations in murine and human plasma.[87]
`types.[72,73] The formation of free radicals by perox- More recently, a potentiometric titration method for
`idation of polyunsaturated fatty acids and/or a direct
`CrEL was developed for quantitative analysis in
`perturbing effect on the cell membrane are possible
`urine and plasma based on coated wire electrode as
`mechanisms responsible for this type of cell growth
`an
`end-point
`indicator with
`sodium
`te-
`inhibition.[74-76] Using in vitro clonogenic assays,
`traphenylborate at 20oC and pH 10.[88] Each of these
`however, it has been demonstrated that CrEL, at methods has its drawbacks and limitations, and the
`clinically achievable concentrations, can antagonise methodological differences between them probably
`the cytotoxicity of paclitaxel by a cell-cycle
`contribute to the variations in measured CrEL con-
`block.[77] Several reports also suggest that Tween®
`centrations.
`80 has intrinsic antitumour activity in animal mod-
`In addition to the Coomassie Brilliant Blue G-
`els,[78-80] which might be linked to the release of
`250 colourimetric dye-binding assay, various other
`analytical procedures are available for Tween® 80.
`oleic acid, a fatty acid known to interfere with
`malignant cell proliferation due to formation of per-
`Initially measurement of the polyoxyethylated por-
`oxides[81] and inhibition of angiogenesis.[82] The ex-
`tion of the molecule was used for quantification of
`act contribution of Tween® 80 to antitumour activi-
`Tween® 80 concentrations. The so-called polyol
`ty observed in patients treated with chemotherapeu- moiety is detectable by a wide variety of methods,
`tic drugs formulated in this vehicle substance has
`including a resorcinol-glucose precipitation, a
`not been clarified.
`colourimetric method using ammonium cobaltoth-
`
`© Adis Data Information BV 2003. All rights reserved.
`
`Clin Pharmacokinet 2003; 42 (7)
`
`

`
`ten Tije et al.
`
`1–3 and 24 hours, CrEL clearance increased from
`about 160 to 300 and 400 mL/h/m2, respectively
`(figure 3).[17] A recently developed population
`pharmacokinetic model revealed that the plasma
`concentration-time data of CrEL were best fitted to a
`three-compartment model with Michaelis-Menten
`elimination (table III).[98,99]
`It thus appears that CrEL shows schedule-depen-
`dent pharmacokinetics, possibly related to saturated
`elimination due to capacity-limited CrEL metabo-
`lism within the systemic circulation. This schedule
`dependency leads to an increase in systemic expo-
`sure, and thus an increase in CrEL-related biological
`effects, with shortening of the infusion duration. An
`example of this phenomenon is the apparent in-
`crease of allergic reactions in 1-hour versus 3- or
`24-hour infusions of paclitaxel,[9,100] as well as in-
`creased incidence of peripheral neuropathy with
`shorter paclitaxel infusions.[101,102] The observed
`changes in adverse effects as a function of paclitaxel
`infusion duration will need to be confirmed in larger
`comparative trials in order to provide recommenda-
`tions for treating clinicians.
`
`Table III. Population pharmacokinetic parameters of Cremophor®
`EL following paclitaxel administrationa
`
`Parameter
`V1 (L)
`Q2 (L/h)
`V2 (L)
`Q3 (L/h)
`V3 (L)
`Km (mL/L)
`Vmax (mL/h)
`
`Estimate
`2.59
`1.44
`1.81
`0.155
`1.61
`0.122
`0.193
`
`RSE (%)
`7
`24
`9
`22
`7
`61
`9
`
`Residual error
`34
`0.0951
`Additional (mL/L)
`8
`6.94
`Proportional (%)
`a Data are from patients treated with paclitaxel formulated in a
`mixture of Cremophor® EL and ethanol, and were obtained
`from Van den Bongard et al.[99] Determination of Cremophor®
`EL in plasma samples was performed by pre-column
`derivatisation and reversed-phase high-performance liquid
`chromatography, as described elsewhere.[84]
`Km = plasma concentration at half Vmax; Q2, Q3 =
`intercompartmental clearances from the central to the first or
`second peripheral compartments; RSE = relative standard error;
`Vmax = maximum elimination rate; V1, V2, V3 = volumes of the
`central, first peripheral and second peripheral compartments.
`
`135 mg/m2
`175 mg/m2
`225 mg/m2
`
`P = 0.03
`
`672
`
`750
`
`600
`
`450
`
`300
`
`150
`
`0
`
`CrEL clearance (mL/h/m2)
`
`1
`
`24
`
`3
`Infusion duration (h)
`Fig. 3. Effect of infusion duration on the clearance of Cremophor®
`EL (CrEL). Data are expressed as mean values ± SD and were
`obtained from patients treated with paclitaxel formulated in CrEL at
`dose levels of 135 mg/m2 (CrEL dose 11.3 mL/m2), 175 mg/m2
`(CrEL dose 14.6 mL/m2) or 225 mg/m2 (CrEL dose 18.8 mL/m2).[17]
`
`turbidimetric or gravimetric proce-
`iocyanate,
`dures, and complex
`formation with barium
`phosphomolybdic reagent.[89,90] The ammonium
`cobaltothiocyanate complexation has also been used
`in combination with HPLC and UV detection for
`analysis of Tween® 80 in urine and ascites fluid,
`using either post-column or on-line complexa-
`tion.[91-94] A less complex procedure that does not
`require complexation involves a one-step hydrolysis
`with sulphuric acid followed by HPLC with UV
`detection at 210nm.[95] Most recently, Tween® 80
`concentration in human plasma samples have been
`analysed by a liquid chromatographic assay with
`tandem mass-spectrometric detection, with a 60-
`fold increased sensitivity as compared with previous
`published assays.[96]
`
`3.2 Pharmacokinetics
`
`The various analytical methods described above
`have been used in different pharmacokinetic studies
`of CrEL, sometimes leading to conflicting results
`and conclusions. There have been no studies thus far
`comparing the different analytical methods. Initial
`pharmacokinetic analyses have indicated that CrEL
`shows linear pharmacokinetic behaviour.[97] How-
`ever, with prolongation of infusion duration from
`
`© Adis Data Information BV 2003. All rights reserved.
`
`Clin Pharmacokinet 2003; 42 (7)
`
`

`
`Drug Disposition and Formulation Vehicles
`
`673
`
`Cremophor® EL
`Tween® 80
`
`100
`
`10
`
`1
`
`Concentration (μL/mL)
`
`0.1
`
`0
`
`2
`
`4
`
`6
`
`8
`
`10
`
`Time (h)
`Fig. 4. Comparative plasma concentration-time profiles of
`Cremophor® EL and Tween® 80 in mice receiving 0.83 mL/kg of
`each vehicle by bolus injection. Data show mean values of four
`observations per time point and were obtained from Van Tellingen
`et al.[87]
`
`The terminal half-life of CrEL amounts to ap-
`proximately 80 hours with reported values ranging
`between 10 and 140 hours, depending on the sam-
`pling time period and the method used for CrEL
`analysis. Therefore, studies using sparse-sampling
`strategies with application of the bioassay method
`may lead to underestimation of the terminal half-
`life.[103] With the more sensitive colourimetric assay,
`detectable concentrations of CrEL were demonstrat-
`ed even 1 week after initial treatment.[68] Despite
`this relatively long terminal disposition phase of
`CrEL, long-term weekly administration of paclitax-
`el does not cause significant accumulation of CrEL
`although the vehicle is always detectable in pre-dose
`samples.[104] In all studies, the observed volume of
`distribution of CrEL was extremely small and al-
`most equal to the volume of the central blood com-
`centrations of Tween® 80 of 0.16 ± 0.05 μL/mL,
`partment. As outlined, this implies that tissue and
`consistent with more recent observations.[96,107] In
`tumour delivery of CrEL is insignificant.[68]
`vitro experiments have shown that this rapid elimi-
`Little is known about elimination routes of CrEL.
`nation is caused by a rapid carboxylesterase-medi-
`Pharmacokinetic studies in patients with hepatic
`ated hydrolysis in the systemic circulation, cleaving
`dysfunction treated with paclitaxel suggested that
`the oleic acid side chain from the molecule.[87] Earli-
`hepatobiliary elimination of CrEL is not of major
`er studies performed in rats and humans with the
`importance.[105] Despite its highly hydrophilic na-
`structurally related surfactants polysorbate 20 and
`ture, the renal elimination of CrEL accounts for less
`polysorbat