vitamin E-TPGS

European Journal of Pharmaceutical Sciences25(2005)445–453Enhanced oral paclitaxel absorption with vitamin E-TPGS:Effect on solubility and permeability in vitro,in situ and in vivoManthena V.S.Varma,Ramesh Panchagnula∗Department of Pharmaceutics,National Institute of Pharmaceutical Education and Research(NIPER),Phase X,SAS.Nagar,Mohali,Punjab160062,IndiaReceived21February2005;received in revised form7April2005;accepted11April2005Available online10May2005AbstractSolubility and permeability being important determinants of oral drug absorption,this study was aimed to investigate the effect of d-␣-tocopheryl polyethylene glycol1000succinate(TPGS)on the solubility and intestinal permeability of paclitaxel in vitro,in situ and in vivo, in order to estimate the absorption enhancement ability of TPGS.Aqueous solubility of paclitaxel is significantly enhanced by TPGS,where a linear increase was demonstrated above a TPGS concentration of0.1mg/ml.Paclitaxel demonstrated asymmetric transport across rat ileum with significantly greater(26-fold)basolateral-to-apical(B–A)permeability than that in apical-to-basolateral(A–B)direction.Presence of P-glycoprotein(P-gp)inhibitor,verapamil(200␮M),diminished asymmetric transport of paclitaxel suggesting the role of P-gp-mediated efflux. TPGS showed a concentration-dependent increase in A–B permeability and decreased B–A permeability.The maximum efflux inhibition activity was found at a minimum TPGS concentration of0.1mg/ml,however,further increase in TPGS concentration resulted in decreased A–B permeability with no change in B–A permeability.Thus,the maximum paclitaxel permeability attained with0.1mg/ml TPGS was attributed to the interplay between TPGS concentration dependent P-gp inhibition activity and miceller formation.In situ permeability studies in rats also demonstrated the role of efflux in limiting permeability of paclitaxel and inhibitory efficiency of TPGS.The plasma concentration of[14C]paclitaxel following oral administration(25mg/kg)was significantly increased by coadministration of TPGS at a dose of50mg/kg in rats.Bioavailability is enhanced about4.2-and6.3-fold when[14C]paclitaxel was administrated with verapamil(25mg/kg)and TPGS, respectively,as compared to[14C]paclitaxel administered alone.The effect of verapamil on oral bioavailability of[14C]paclitaxel was limited relative to the TPGS,consistent with the in vitro solubility and permeability enhancement ability of TPGS.In conclusion,the current data suggests that the coadministration of TPGS may improve the bioavailability of BCS class II–IV drugs with low solubility and/or less permeable as a result of significant P-gp-mediated efflux.©2005Elsevier B.V.All rights reserved.Keywords:Vitamin E-TPGS;P-glycoprotein;Oral absorption;Pharmacokinetics1.IntroductionPaclitaxel,an antimicrotubule anticancer drug used in wide variety of human cancers,is currently formulated with cremophor EL(polyethoxylated castor oil derivative)and de-hydrated alcohol(1:1),is administered through intravenous∗Corresponding author.Tel.:+91172214682;fax:+91172214692.E-mail address:panchagnula@(R.Panchagnula).infusion(Panchagnula,1998).Ethanol-cremophor EL vehi-cle required to solubilize paclitaxel in this formulation is toxic and also produces vasodilation,labored breathing,lethargy, and hypotension.In order to develop safer clinical formu-lations,many studies have been directed to novel oral for-mulations(Dhanikula and Panchagnula,1999;Mu and Feng, 2003;Feng et al.,2004;Yang et al.,2004;Win and Feng, 2005).Paclitaxel is very poorly absorbed on peroral admin-istration because of its low solubility and low permeabil-ity.Apart from its unfavorable physicochemical features for0928-0987/$–see front matter©2005Elsevier B.V.All rights reserved. doi:10.1016/j.ejps.2005.04.003446M.V.S.Varma,R.Panchagnula/European Journal of Pharmaceutical Sciences25(2005)445–453passive permeability,it is also believed that P-glycoprotein (P-gp)hinders the transport of paclitaxel from the gut(Varma et al.,2005a).An increasing number of drugs,including HIV protease inhibitors like indinavir,ritonavir,saquinavir and anti-cancer drugs like docetaxel,vinblastine,etc have been reported to be substrates for P-gp(Varma et al.,2003).In vivo studies confirmed that P-gp significantly limits oral bioavail-ability of several drugs,where intestinal permeability showed dose dependence with increased permeability as lumen con-centration increases(Williams and Sinko,1999;Malingre et al.,2001).Studies using mdr1a(−/−)mice showed direct evidences that P-gp strictly limits the uptake of orally administered paclitaxel(Sparreboom et al.,1997).Woo et al.(2003) demonstrated that about half of paclitaxel dose administered is extruded to the gut lumen by P-gp and only small amount of drug is lost by gut wall and liver metabolism.Thus,the oral bioavailability of paclitaxel can be significantly enhanced by effectively inhibiting P-gp-mediated efflux.Solubility and permeability of a drug are the fundamental determinants of its oral bioavailability(Varma et al.,2004). Surfactants are extensively used to increase the absorption of drugs from the intestine;as they show increased solubility of hydrophobic macromolecules,increased membranefluidity or disruption of tight junctions,interaction with metabolic enzymes and inhibition of efflux transporters(Nerurkar et al.,1997;Rege et al.,2002).Vitamin E-TPGS(TPGS) is non-ionic water soluble derivative of Vitamin E found to enhance the bioavailability of cyclosporin and amprenavir by enhancing solubility and/or permeability,or reducing intesti-nal metabolism(Sokol et al.,1991;Chang et al.,1996;Yu et al.,1999;Joshi et al.,2003).TPGS form micelles above the critical miceller concentration(CMC)and improve sol-ubility of lipophilic compounds.Previous reports suggested that coadministration of TPGS enhanced oral absorption of cyclosporin A due to improved solubilization by micelle for-mation(Sokol et al.,1991;Boudreaux et al.,1993).Chang et al.(1996)evaluated the effect of TPGS on the oral phar-macokinetics of cyclosporin A in healthy volunteers,and suggested that enhanced absorption,decreased counter trans-port by P-gp,or some unknown mechanism is responsible for the observed decrease in apparent oral ter on,it was demonstrated that TPGS enhanced the cytotox-icity of doxorubicin,vinblastine,paclitaxel and colchicine in the G185cells,by acting as reversing agent for P-gp-mediated multidrug resistance(Dintaman and Silverman, 1999).In the light of above discussion,the present work investi-gated the effect of TPGS on the solubility and permeability of a biopharmaceutic classification system(BCS)class IV drug, paclitaxel in vitro,in situ and in vivo.The functional role of P-gp in limiting the permeability of paclitaxel was determined along with the influence of miceller drug concentration on the effective permeability.Furthermore,we also studied the influence of TPGS on the oral bioavailability of paclitaxel in rats.2.Materials and methods2.1.ChemicalsPaclitaxel was gifted by Dabur India Ltd.(New Delhi,In-dia),and[14C]paclitaxel was purchased from Sigma–Aldrich Co.(MO,USA).[3H]Imipramine was purchased from NEN Bio(Boston,MA).Hydrochlorthiazide was received from Aristo Pharmaceuticals Ltd.(Daman,India),and propra-nolol HCl was from Sun Pharmaceutical Industries Ltd. (Mumbai,India).Frusemide was gifted by Dr.Reddys’Lab. (Hyderabad,India).l-Phenylalanine was purchased from Sisco Lab.(Mumbai,India).Other compounds verapamil, imipramine and d-glucose were purchased from Sigma Chemicals Co.(MO,USA).Lecithin and dodecane were purchased from Himedia Lab.Pvt.Ltd.(Mumbai,India). Vitamin E-TPGS was from Eastman,and DMSO was pro-cured from Sigma–Aldrich Co.(MO,USA).Solvents used for quantification were of HPLC grade(JT Baker,Mexico) and all other chemicals and reagents were of analytical grade.2.2.Animals and legal prerequisitesSprague–Dawley rats(270–350g)were used for in vitro transport,in situ single-pass perfusion and in vivo pharmacokinetic studies.Anesthesia,surgical and disposal procedures were justified in detail and were approved by the Institutional Animal Ethics Committee(IAEC,NIPER).The studies complied with local and federal requirements for an-imal studies.2.3.Solubility studiesEquilibrium solubility of paclitaxel in0–5mg/ml con-centration of TPGS was determined by shake-flask method (n=3).Excess amount of drug was added in TPGS solution in water and equilibrated at37±0.2◦C with vigorous shak-ing in shaker water bath(Julabo,Germany)for48h.Samples werefiltered using0.22␮mfilter(Millipore,USA).Aliquots of eachfiltrate were diluted appropriately and analyzed by RP-HPLC method.2.4.Measurement of apparent artificial membrane permeability(P PAMPA)Artificial membrane permeability studies were performed in the same manner described previously(Kansy et al., 1998;Bermejo et al.,2004).In brief,a96-well microplate (acceptor compartment)was completelyfilled with phos-phate buffer(pH7.4)containing5%DMSO.Eachfilter of the donor plate(Millipore Corp.,Bedford,MA)was im-pregnated with5␮l of10%(w/v)lecithin in dodecane.A 200␮l of10␮M[14C]paclitaxel(0.4␮Ci/ml)in different concentrations of TPGS solution(n=4for each TPGS con-centration studied)was added immediately and incubatedM.V.S.Varma,R.Panchagnula/European Journal of Pharmaceutical Sciences25(2005)445–453447 at room temperature for16h.A reference solution defin-ing equilibrium conditions was prepared as the sample so-lution with no membrane barrier.Thefilter surface waswetted with5␮l of60%(v/v)methanol/buffer solutionfor the reference.[14C]paclitaxel was quantified in the re-ceiver compartment by liquid scintillation counter(Wal-lac,Finland).Apparent artificial membrane permeability(P PAMPA)was calculated as previously reported(Ano et al.,2004).2.5.Bi-directional transport studiesPaclitaxel transport across rat ileum tissue was mea-sured by methods previously reported(Collett et al.,1999;Stephens et al.,2002).Non-fasting male Sprague–Dawleyrats were sacrificed by cervical dislocation after a mildanesthesia with thiopental(40mg/kg).Ileum segment wasimmediately removed,washed and mounted intact in mod-ified Ussing chambers.Intestinal mucosa was bathed with6ml each(donor and receptor chambers)of bicarbonate-buffered Ringer,pH7.4at37◦C,under continuous oxy-genation.After an equilibration period of30min,bathingmedium was replaced with TPGS solution(0–1mg/ml)containing labeled(0.2␮Ci/ml)and unlabelled pacli-taxel(5␮M)and imipramine(10␮M),from either the(A)pical or(B)asolateral chamber.P-gp inhibitor,vera-pamil(200␮M)was added to the apical chamber(inhibi-tion studies).Samples(500␮l)from the receiver chamberwere withdrawn every30min,for2h,and replaced withbicarbonate-buffered Ringer.Samples were analyzed by liq-uid scintillation counting.Imipramine,which is a passivepermeable compound,was used to assess the inherent pas-sive transcellular permeability and thus the intestinal tissueintegrity.Cumulative amount permeated was plotted against time tocalculate apparent permeability(P app)in cm/s,byP app=d Q/d tCA(1)where d Q/d t is the rate of appearance of compound in the re-ceiver chamber,C the substrate concentration in donor cham-ber,and A is the cross-section area(0.44cm2).P app valuesare reported as mean±S.D.of four tissues taken from inde-pendent rats.In order to evaluate the quantitative contribution of P-gp tolimit the intestinal absorption,we calculated the absorptionquotient(AQ)and secretory quotient(SQ)(Troutman andThakker,2003;Varma et al.,2005a):AQ=P P-gp,ABP PD,AB=P PD,BA−P app,ABP PD,BA(2)SQ=P P-gp,BAP PD,BA=P app,BA−P PD,BAP PD,BA(3)where P P-gp,AB and P P-gp,BA express the effect P-gp would have in attenuating absorption transport(AQ)and en-hancing secretory transport(SQ)of its substrates,respec-tively,and P PD,AB and P PD,BA are passive permeabil-ity in the absorption and secretory directions(equal to A–B and B–A permeabilities in the presence of vera-pamil(200␮M)).AQ quantifies the passive drug trans-port attenuation by P-gp across the intestinal enterocytes and provide information on the extent to which P-gp in-fluences intestinal absorption of P-gp substrates across the enterocytes.2.6.In situ permeability studiesThe surgical procedure and the in situ single-pass per-fusion experiments were performed according to the meth-ods described previously(Fagerholm et al.,1996;Hanafy et al.,2001;Varma et al.,2005b;Varma and Panchagnula, 2005)Rats were fasted for16–18h,prior to study,with tap water ad libidum.After anesthesia via intraperitoneal administration of thiopental sodium(50mg/kg),rats were placed on a heating pad to maintain body temperature at 37◦C.Ileum segment of8–10cm was isolated and cannu-lated with glass tubing.The segment was rinsed with phos-phate buffer saline(10ml)and the perfusate solution main-tained at37◦C was pumped at aflow rate of0.1ml/min using syringe pump(Harvard Apparatus PHD2000pump,MA, USA).The perfusion solution(pH7.4)consisted of NaCl 48mM,KCl5.4mM,Na2HPO42.8mM,NaH2PO44mM and d-glucose1g/l;contained paclitaxel(2␮M)with TPGS (0–1mg/ml)or verapamil(200␮M).Paclitaxel concentra-tion was low enough to avoid precipitation in the lumen dur-ing the course of the study.Single-pass perfusion procedure was followed to determine the permeability(Hanafy et al., 2001).Sampling was made every5min for a60min perfu-sion period with perfusion solution after20min equilibra-tion.Equilibration of20min prior to sampling was found to be sufficient for both washout and to reach an initial steady state(Hanafy et al.,2001).Waterflux was quanti-fied based on direct measurement of the volume at the out-let(Bermejo et al.,2004;Varma and Panchagnula,2005). To further keep a check on the intra-and inter-individual variability and presence of steady state,permeability of l-phenylalanine and propranolol(passive,highly permeable), frusemide and hydrochlorthiazide(passive,low permeable), and d-glucose were monitored,by co-perfusing along with drug solution without and with P-gp inhibitor.D-glucose was estimated using GOD-POD based assay kit(Autozyme,AC-CURex Biomedical Pvt.Ltd.,Mumbai,India).RP-HPLC with dual wavelength UV detector was used for simultaneous quantification of paclitaxel and other permeability markers (l-phenylalanine,propranolol,frusemide and hydrochlorthi-azide).In situ permeabilities without(P eff,control)and with (P eff,inh)P-gp inhibitor are calculated after correcting the outlet concentration for waterflux on the basis of the ra-tio of volume of perfusion solution collected and infused for each sampling point(5min)(Bermejo et al.,2004;Varma448M.V.S.Varma,R.Panchagnula/European Journal of Pharmaceutical Sciences25(2005)445–453 and Panchagnula,2005):P eff,control(or)P eff,inh=Q(C in/C out)−12Πrl(4)where Q is theflow rate,C in and C out the respective inlet and outlet concentration,r the radius of intestine(0.21cm) and l is the length of intestine measured after completion of perfusion.2.7.Bioanalysis by RP-HPLCHPLC was conducted on Waters(Waters Corp.,MA, USA)equipped with600controller pumps,Waters2487 UV–vis dualλabsorbance detector and configured to Millenium32software.Perfusion sample was loaded onto the column by means of717plus autosampler(Waters Corp., MA,USA).The column used for chromatographic sep-arations was4.6mm i.d.,250mm length and5␮m par-ticle size C18(Symmetry®,Waters,USA)attached to a guard column of C18(Waters,USA).For paclitaxel,pro-pranolol and frusemide,the mobile phase composed of70% methanol,26%pH5.0(20mM)acetate buffer and4%iso-propyl alcohol was pumped at aflow rate of0.6ml/min and chromatograms were recorded at220and230nm.In case of l-phenylalanine and hydrochlorthiazide,mobile phase con-sisting of52%methanol,41.7%pH5.0(20mM)acetate buffer and6.3%isopropyl alcohol were pumped at aflow rate of0.6ml/min and chromatograms were recorded at230and 275nm.2.8.Pharmacokinetic studyMale Sprague-Dawley rats were fasted,with water ad libi-tum,for16h before the oral administration.[14C]paclitaxel was dissolved in cremophor EL and ethanol(1:1,v/v) and diluted fourfold with saline,and thefinal solution of6.25mg/ml(0.2␮Ci/mg[14C]paclitaxel)was adminis-tered orally(25mg/kg of rat)with a blunt needle via the esophagus into stomach;or2mg/ml paclitaxel was injected intravenously(2mg/kg of rat),after cannulating femoral vein.The oral solution containing P-gp inhibitor included either verapamil(25mg/kg)or TPGS(50mg/kg). Subsequently,blood samples(200␮l)were collected under diethylether anesthesia,into heparinized micro-centrifuge tubes(100IU/ml blood),from retro-orbital vein(Zhang et al.,2000).An equivalent volume of dextrose saline is administered intra-peritoneally to maintain central com-partment(blood)volume.Blood samples were centrifuged for5min at5000×g and the[14C]paclitaxel concentra-tions in the plasma were determined by liquid scintillation counting.Area under the paclitaxel concentration-time curve(AUC) were calculated by trapezoidal rule.The apparent bioavail-ability(F oral)of orally administered paclitaxel was calculated byF=AUC oralAUC IVDose IVDose oral(5) 2.9.StatisticsDifference in pharmacokinetic parameters and difference in permeabilities of paclitaxel was evaluated by Student’s t-test(SigmaStat version2.03,SPSS Inc.,IL,USA)at P<0.05 and<0.01.3.Results3.1.Effect of TPGS on solubility and artificial membrane permeability(P PAMPA)of paclitaxelPaclitaxel is a non-ionic molecule and is practically in-soluble in water at37◦C.Equilibrium solubility is as low as 1.34±0.18␮g/ml and the presence of TPGS at concentration up to0.1mg/ml showed no significant change in its solubility. However,paclitaxel solubility is directly proportional to the TPGS concentration,above0.2mg/ml(Fig.1).At a TPGS concentration of5mg/ml,the solubility of paclitaxel in water was increased by about38-fold.The linear increase in pacli-taxel solubility above TPGS CMC(0.2mg/ml)(Yu et al., 1999),indicate the distribution of paclitaxel in the TPGS mi-celles.The total paclitaxel concentration(S total)above CMC of TPGS is given by Eq.(6):S total=S free[1+K a(SAA)m](6) where(SAA)m is the concentration of TPGS in micelle form, which is equal to the difference in the apparent TPGS con-centration and the CMC.Free drug concentration(S free)and the equilibrium distribution coefficient(K a)are calculated by fitting paclitaxel solubility and the TPGS miceller concentra-tion in Eq.(6).K a was found to be0.86mM−1.Aqueous solu-bility and K a being small,appreciable micellersolubilizationFig.1.TPGS concentration-dependent solubility of paclitaxel.Inset shows total solubility of paclitaxel vs miceller concentration of TPGS.Linear re-gression was obtained with Eq.(6)to estimate association constant(K a). Each bar represents mean±S.D.(n=3)of equilibrium solubility at48h.M.V.S.Varma,R.Panchagnula/European Journal of Pharmaceutical Sciences25(2005)445–453449Fig.2.Effect of TPGS concentration on artificial membrane permeability of paclitaxel(10␮M).Open points( )depict intrinsic permeability calculated using Eq.(7).Apparent permeability of paclitaxel decreased above CMC of TPGS,but the intrinsic permeability did not change.Values represent mean±S.D.(n=4).Artificial membrane constituted of hydrophobicfilter impregnated with5␮l of10%(w/v)lecithin in dodecane.of paclitaxel occurred at relatively large TPGS concentra-tions.Fig.2shows the effect of TPGS on the P PAMPA of pacli-taxel.P PAMPA was found to be0.86±0.1×10−6cm/s across artificial membrane formed with10%lecithin in dodecane. Presence of TPGS showed no effect on P PAMPA up to a con-centration of0.1mg/ml,above which,gradual decrease was observed.The decline in P PAMPA is attributable to the reduc-tion in availability of free drug due to TPGS micellization, above0.2mg/ml.To normalize for the free drug concentra-tion,the intrinsic permeability P PAMPA,int was calculated us-ing distribution constant(K a)and the(SAA)m byP PAMPA,int=P PAMPAF free=P PAMPA[1+K a(SAA)m](7)Intrinsic permeability values,in the presence of TPGS above0.2mg/ml,were found to be similar to the permeability of paclitaxel in the absence of TPGS(Fig.2).3.2.Effect of TPGS on bi-directional transport of paclitaxelPaclitaxel demonstrated polarized permeability in bi-directional transport studies across excised rat ileum.B–A permeability was found to be26-fold more then A–B perme-ability.P-gp inhibition by verapamil(200␮M)significantly increased A–B permeability(P<0.05),and decreased B–A permeability(P<0.05).AQ and SQ(calculated using A–B and B–A permeabilities in the presence of verapamil as pas-sive permeability of paclitaxel in the absorptive and secretory directions,respectively)indicated that about89%of passive permeability of paclitaxel is attenuated by P-gp-mediated ef-flux in A–B direction,and P-gp enhanced secretory perme-ability by1.18-fold.The secretory permeability decreased to66%in the presence of0.002mg/ml TPGS(from(9.24±0.59to Fig.3.Paclitaxel bi-directional transport across rat ileum,in the presence and absence of P-gp inhibitor verapamil(200␮M)or different concentrations of TPGS.The apparent permeability coefficient(P app)in A–B direction in-creased and the P app in B–A decreased as the TPGS concentration increased up to its CMC.AQ and SQ values are negligible when TPGS concentration is0.1–0.2mg/ml,indicating that TPGS completely inhibited P-gp efflux at a minimum concentration of0.1mg/ml.No change in B–A P app was found above CMC of TPGS(0.2mg/ml)while A–B P app reduced,consistent with TPGS micelle formation above its CMC.Each bar indicates mean±S.D. (n=4)(Ver,verapamil).Table1Bi-directional permeability of imipramine(10␮M),across rat ileum,in the presence of verapamil(200␮M)or different concentrations of TPGSP app(×106cm/s)a Efflux ratio bA–B B–AControl10.2±1.211.4±1.9 1.1+Verapamil(200␮M)11.3±0.911.8±0.9 1.0+TPGS(0.002mg/ml)10.9±1.59.2±1.10.8+TPGS(0.02mg/ml)12.1±1.28.4±1.60.7+TPGS(0.1mg/ml)12.4±2.113.6±1.0 1.1+TPGS(0.2mg/ml)9.9±1.612.4±1.8 1.2+TPGS(0.5mg/ml)10.1±1.89.1±1.20.9+TPGS(1.0mg/ml)12.9±2.010.7±0.80.8a Data are presented as mean±S.D.of four ileum segments.No statistical significance was found between permeabilities in absorptive and secretive directions.b Efflux ratio is the ratio of B–A permeability to A–B permeability.6.12±0.41)×10−6cm/s),whereas absorptive permeability increased4.7-fold(Fig.3).Increase in TPGS concentration up to0.2mg/ml diminished polarized transport by further in-creasing the absorptive transport and reducing the secretory transport.AQ and SQ decreased as a function of TPGS con-centration and was found to be negligible at0.1and0.2mg/ml indicating that,similar to verapamil,TPGS inhibit P-gp and the maximum inhibition was achieved at0.1mg/ml concen-tration.Further increase in TPGS concentration to0.5and 1mg/ml significantly decreased paclitaxel A–B permeabil-ity(P<0.05).AQ increased to0.45at TPGS concentration 1mg/ml,however,no change in secretory permeability was observed.Bi-directional transport of a passive permeability marker,imipramine,which was used to check the integrity of the biomembrane,showed no efflux and difference in its per-meability in the presence of various concentrations of TPGS (Table1).450M.V.S.Varma,R.Panchagnula/European Journal of Pharmaceutical Sciences25(2005)445–453Fig.4.Influence of luminal TPGS concentration on P-gp-mediated efflux and intestinal permeability of paclitaxel in rat ileum in situ.Open points ( )depict intrinsic permeability calculated using Eq.(7).Values represent mean±S.D.(n=4).∗P<0.05and**P<0.01,significantly different when compared to control.Permeability was measured in ileum by single-pass perfusion of drug solution in aqueous solutions of TPGS with and without verapamil(200␮M)(Ver,verapamil).3.3.Effect of TPGS on in situ permeability of paclitaxelIn situ single-pass perfusion in rat ileum also demon-strated functional role of P-gp in limiting paclitaxel oral absorption(Fig.4).Inhibition of P-gp by verapamil (200␮M),co-perfused with paclitaxel(2␮M),increased in situ permeability by 2.3-fold(from(0.08±0.01to 0.25±0.02)×10−4cm/s).Similarly,0.1mg/ml TPGS also increased in situ permeability,but to a lesser -bination of verapamil(200␮M)and0.1mg/ml TPGS ex-hibited permeability similar to that observed in the presence of only verapamil.As observed with bi-directional transport studies,1mg/ml TPGS demonstrated increased permeabil-ity over control,however,was significantly less than that of 0.1mg/ml TPGS,consistent to micelle formation reduces the free drug and free TPGS concentration above CMC,lead-ing to observed effects.Coperfusion of verapamil(200␮M) with paclitaxel and TPGS(1mg/ml)also did not significantly increase the permeability,substantiating the effect of mi-celles on permeability.Intrinsic permeability of paclitaxel in the presence of1mg/ml TPGS was found to be(0.20and 0.22)×10−4cm/s.TPGS did not show significant change in the transport of passive and carrier-mediated permeability markers in situ(Fig.5).3.4.Effect of TPGS on in vivo permeability of paclitaxelPlasma concentration-time profiles of[14C]paclitaxel ad-ministered intravenously(2mg/kg);given alone or in com-bination with verapamil(25mg/kg)or in combination with TPGS(50mg/kg),orally(25mg/kg)in Sprague–Dawley rats,are given in Fig.6.[14C]paclitaxel is rapidly distributed after intravenous administration and is very poorly absorbed after oral administration with apparent bioavailability of 4.7±0.5%.Coadministration of vera-pamil with[14C]paclitaxel resulted in a significantincrease Fig.5.Effect of TPGS on the rat in situ ileum permeability of high-andlow-permeable passive and carrier-mediated transport markers.Except for propranolol and hydrochlorthiazide,which showed reduced permeability in the presence of1mg/ml TPGS,no significant difference was found with the markers in the presence of verapamil(200␮M)and/or TPGS(P>0.05). Data points represent the mean±S.D.(n=4)(l-Phy,l-phenylalanine;Ppn, propranolol;Hct,hydrochlorthiazide;Fru,frusemide;Ver,verapamil). Fig.6.Plasma concentration–time profile of[14C]paclitaxel in rats af-ter(a)intravenous administration(2mg/kg);(b)after oral administration (25mg/ml)of[14C]paclitaxel alone and in combination with verapamilor TPGS.Data points represent mean and error bars show S.E.M.(n=4).∗P<0.05and**P<0.01,significantly different when compared to oral pacli-taxel alone.#P<0.05and##P<0.01,significantly different when compared to oral paclitaxel in combination with verapamil(Pcl,[14C]paclitaxel;Ver, verapamil).M.V.S.Varma,R.Panchagnula/European Journal of Pharmaceutical Sciences25(2005)445–453451 Table2Pharmacokinetic parameters of[14C]paclitaxel after intravenous(2mg/kg)or oral(25mg/kg)administration in ratsAdministrations C max(ng/ml)AUC0–24(ng h/ml)F oral a(%)i.v.[14C]paclitaxel1782.3±62.8Oral[14C]paclitaxel only82.1±7.51051.7±123.3 4.7±0.5 Oral[14C]paclitaxel+verapamil(25mg)210.7±24.6**4433.9±804.4**19.9±3.6** Oral[14C]paclitaxel+TPGS(50mg)254.9±18.9**6664.2±626.3**29.9±2.8** a Data are presented as mean±S.E.M.of four rats for each group.∗∗P<0.01,in comparison to[14C]paclitaxel administered alone orally.Statistical significance was assessed by Student’s t-test.in plasma concentration of[14C]paclitaxel.The maximum plasma concentration(C max)and apparent bioavailability of [14C]paclitaxel were2.6-and4.2-fold higher,when coad-ministrated with verapamil(Table2).Further increase in [14C]paclitaxel plasma concentration was observed after oral administration of[14C]paclitaxel with TPGS.C max and ap-parent bioavailability of[14C]paclitaxel administered with TPGS were3.1-and6.3-fold higher when[14C]paclitaxel was administrated alone;1.2-and1.5-fold higher when ad-ministrated with verapamil.4.DiscussionPaclitaxel,belonging to BCS class IV,has low solubility and permeability as a result of which this clinically potent molecule is orally inactive(Varma et al.,2005a).Presence of TPGS significantly improved the solubility by miceller sol-ubilization and enhanced permeability by P-gp efflux modu-lation.Below CMC,TPGS has no effect on the solubility of pacli-taxel,while above the CMC,solubility linearly increased with TPGS concentration,which is in agreement with the previ-ously reported results for cyclosporin(Ismailos et al.,1994) and amprenavir(Yu et al.,1999).To examine the effect of micelle formation on the passive permeability of paclitaxel, transport studies were carried out with artificial membranes (PAMPA)formed with lecithin in dodecane.Permeability was found to decline significantly when TPGS concentration in donor chamber is above its CMC.Intrinsic permeability was found to be similar to paclitaxel permeability below CMC and thus the decline in passive permeability is attributed to the presence of paclitaxel in TPGS micelles(Amidon et al., 1982).Paclitaxel demonstrated polarized transport across ileum segment,which is diminished in the presence of verapamil. AQ in the presence of P-gp-mediated efflux suggested that about89%of paclitaxel passive transport is attenuated by P-gp and thus may have a major effect on the efficiency of intestinal paclitaxel absorption in vivo(Varma and Pan-chagnula,2005).Below CMC,TPGS showed concentration-dependent increase in the net absorption of paclitaxel by decreasing efflux.Further increase in TPGS concentration (above CMC)lead to decrease in absorptive transport,which resulted in increased AQ.Decrease in absorptive permeabil-ity,which,however,is higher than that of control,could be due to micelle formation where the availability of free drug and free TPGS is low.In situ permeability studies also demon-strated the role of P-gp in limiting intestinal permeability of paclitaxel,effect of TPGS on P-gp-mediated efflux and the influence of micelle formation on the overall permeability of paclitaxel.A distinct difference in paclitaxel permeability was observed when coperfused with TPGS below and above its CMC.Transport markers were used in bi-directional and in situ permeability studies so as to check the integrity of the in-testinal tissue in the presence of TPGS.Only insignificant difference was found between permeability values in the ab-sence and presence of TPGS for transport markers,indicat-ing that changes in paclitaxel permeability in the presence of varying concentrations of TPGS are not due to compro-mise in membrane integrity.Because of the solubility and/or permeability limitations the absorption site for drugs belong-ing to BCS classes II–IV is shifted more towards the ileum, which has a transit time of∼140min(Kaus et al.,1999).The high P-gp expression levels in the lower intestine make the moderately absorbed P-gp substrates more susceptible to P-gp-mediated efflux(Mouly and Paine,2003;Siegmund et al., 2003).Rat being the best F a,human predictor animal model for passive permeability and further there exists a similar level of P-gp expression(Stephens et al.,2001)and overlapped sub-strate specificity with quantitatively same affinity for a large number of P-gp substrates in rat mdr1a and human MDR1 (Yamazaki et al.,2001),permeability studies in rat ileum pro-vides more meaningful forecast on in vivo absorption of P-gp substrates(Varma and Panchagnula,2005).A striking increase in the AUC of[14C]paclitaxel was observed in rats when the drug was administered orally in combination with TPGS.Absorption enhancement is sig-nificant given the reasons for poor pharmacokinetics of [14C]paclitaxel in vivo.It has been previously reported that systemic exposure to orally administered paclitaxel is sub-stantially enhanced by coadministration of P-gp inhibitor such as cyclosporin A,PSC833,verapamil,KR-30031and MS-209(Sokol et al.,1991;van Asperen et al.,1997,1998; Kimura et al.,2002;Woo et al.,2003).Combining with these P-gp inhibitors enhanced bioavailability of orally ad-ministered paclitaxel in normal mice to levels similar to those obtained in mdr1a(−/−)mice given paclitaxel only (van Asperen et al.,1997and1998).Thus the increase in。