physicochemical and pharmacokinetic characterization of

Regular Article

Physicochemical and Pharmacokinetic Characterizationof Amorphous Solid Dispersion of Tranilast with Enhanced Solubility

in Gastric Fluid and Improved Oral Bioavailability

Satomi ONOUE1,*, Yoshiki KOJO1, Yosuke AOKI1, Yohei KAWABATA1,Yukinori YAMAUCHI2 and Shizuo YAMADA1

1Department of Pharmacokinetics and Pharmacodynamics and Global Center of Excellence (COE) Program,School of Pharmaceutical Sciences, University of Shizuoka, Shizuoka, Japan

2Department of Pharmaceutical Physical Chemistry, College of Pharmaceutical Sciences,Matsuyama University, Matsuyama, Japan

Full text of this paper is available at http://www.jstage.jst.go.jp/browse/dmpk

Summary: In the present study, amorphous solid dispersion (ASD) formulations of tranilast (TL) with 8hydrophilic polymers were prepared by a solvent evaporation method with the aim of improving dissolutionbehavior in gastric fluid and thereby enhancing oral bioavailability. The physicochemical properties werecharacterized with a focus on morphology, crystallinity, thermal behavior, dissolution, drug-polymerinteraction, and stability. Of all TL formulations, ASD formulation with Eudragit EPO exhibited the highestimprovement in dissolution behavior with a 3,000-fold increase in the first-order dissolution rate underacidic conditions (pH 1.2). Spectroscopic studies using infrared and near-infrared analyses revealedthe drug-polymer interaction in the Eudragit EPO-based ASD formulation. On the basis of dissolution,crystallinity, and stability data, the maximum allowable drug load in the Eudragit EPO-based ASD formula-tion was deduced to be ca. 50%. Pharmacokinetic profiling of orally dosed TL formulations in rats wasalso carried out using UPLC/ESI-MS. After oral administration of the Eudragit EPO-based ASD formulationin rats, enhanced TL exposure was observed with an increase of oral bioavailability by 19-fold, and thevariation of AUC was ca. 4 times lower than that with crystalline TL. With these data, the ASD approachcould be a viable formulation strategy for enhancing the wettability and oral bioavailability of TL, resulting inimproved therapeutic potential of TL for the treatment of inflammatory diseases.

Keywords: tranilast; amorphous; solid dispersion; dissolution; bioavailability

Introduction

Tranilast ¤TL¥ ªN-¤3,4-dimethoxycinnamoyl¥ anthranilicacid« inhibits the release of inflammatory mediators from mastcells and basophils,1,2¥ so it has been clinically used for thetreatment of inflammatory diseases such as bronchial asthma,atopic rhinitis, and atopic dermatitis.3¥ Recently, attention hasbeen drawn to the therapeutic potential of TL for variouschronic diseases, including hepatic fibrosis,4,5¥ pulmonaryfibrosis,6,7¥ and cancer.8®10¥ In spite of the attractive biologicalfunctions of TL as an anti-allergic agent, its bioavailabilityafter oral administration is low with high variability,11¥ pos-

sibly leading to limited clinical outcomes. The solubilities ofTL in water and acidic medium ¤buffer solution of pH 1.2¥were found to be 14.5 and 0.7 µg/mL, respectively,12¥ sug-gesting poor dissolution behavior of TL in the gastrointestinaltract, particularly the stomach. Generally, drug release isa crucial and limiting step for oral drug bioavailability,particularly for biopharmaceutical classification system ¤BCS¥class II drugs with low gastrointestinal solubility and highpermeability. Therefore, improvement of the dissolutioncharacteristics in gastric fluid could be a promising approachto enhance the oral bioavailability and therapeutic potentialof TL for clinical treatment of inflammatory diseases.

Received August 29, 2011; Accepted December 29, 2011J-STAGE Advance Published Date: January 13, 2012, doi:10.2133/dmpk.DMPK-11-RG-101*To whom correspondence should be addressed: Satomi ONOUE, Ph.D. Department of Pharmacokinetics and Pharmacodynamics and Global Centerof Excellence (COE) Program, School of Pharmaceutical Sciences, University of Shizuoka, 52-1 Yada, Suruga-ku, Shizuoka 422-8526, Japan.Tel. +81-54-264-5633, Fax. +81-54-264-5635, E-mail: [email protected] work was supported in part by a Grant-in-Aid for Young Scientists (B) (No. 22790043; S. Onoue) from the Ministry of Education, Culture,Sports, Science and Technology in Japan.

Drug Metab. Pharmacokinet. 27 (4): 379387 (2012). Copyright © 2012 by the Japanese Society for the Study of Xenobiotics (JSSX)

379

http://www.jstage.jst.go.jp/browse/dmpk

http://dx.doi.org/10.2133/dmpk.DMPK-11-RG-101

To improve the aqueous solubility of poorly solubledrugs, several types of approaches have been proposed,which include solubilization using good solvents or co-solvent mixtures, solid state manipulation, particle engineer-ing, solubilization by lipids, and solubilization with complex-ing agents. In particular, dissolution and solubility enhance-ment can be achieved by dispersing the poorly soluble drugin a solid matrix carrier, either as a eutectic or phase-separated mixture, or as an amorphous solid dispersion¤ASD¥. Previously, our group developed a crystalline soliddispersion ¤CSD¥ system of TL employing wet millingtechnologies.11¥ The CSD formulation of TL exhibited rapiddissolution behavior in water and improved oral bio-availability compared with those of crystalline TL. How-ever, recent work demonstrated that ASD formulation oftacrolimus, an immunosuppressive agent, exceeded itsCSD formulation in terms of both aqueous solubility anddissolution behavior.13¥ These observations prompted us todevelop new ASD formulations of TL ¤ASD-TL¥ usingvarious polymers and apply them to oral delivery systems,with the aim of enhancing the dissolution properties andoral bioavailability of TL with low variation.In the present investigation, new ASD-TLs with eight

hydrophilic polymers were prepared by a solvent evapo-ration method. Various ASD-TLs with drug loading of10% were characterized by scanning electron microscopy¤SEM¥ for morphology, polarized light microscopy ¤PLM¥and powder X-ray diffraction ¤PXRD¥ for crystallinity,differential scanning calorimetry ¤DSC¥ for thermal behav-ior, and dissolution testing. Possible interaction betweenTL and the polymer was assessed by Fourier transforminfrared ¤FT-IR¥ and near-infrared ¤NIR¥ spectral analyses.The formulation was optimized on the basis of results fromdissolution and stability tests on the ASD-TL with vari-ous drug loads ¤10®90%¥. Pharmacokinetic ¤PK¥ profilingof TL after the oral administration of the optimized ASD-TL in rats was conducted using ultra-performance liquidchromatography equipped with electrospray ionization massspectrometry ¤UPLC/ESI-MS¥.

Materials and Methods

Chemicals: Crystalline TL was supplied from KisseiPharmaceutical Co., Ltd. ¤Nagano, Japan¥. Amorphous TLwas obtained by amorphization of crystalline TL using asolvent evaporation method. Briefly, 1mg of crystalline TLwas dissolved in 10mL of 90% ¤v/v¥ 1,4-dioxane solution,and the solution was freeze-dried using an FD-81 freeze-dryer ¤Tokyo Rikakikai, Tokyo, Japan¥. Amorphization waschecked by PXRD and DSC. Hydroxypropyl methylcellu-lose ¤HPMC¥ was provided by Shin-Etsu Chemical ¤Tokyo,Japan¥, and Kollidon VA64 and polyvinylpyrrolidone ¤PVP¥were provided by BASF Japan ¤Tokyo, Japan¥. EudragitEPO was supplied by Degussa Japan ¤Tokyo, Japan¥. Potas-sium bromide ¤KBr¥, 1,4-dioxane, hydrochloric acid ¤HCl¥,hydroxypropyl cellulose ¤HPC¥, and ammonium acetate

were purchased from Wako Pure Chemical Industries¤Osaka, Japan¥. Methanol ¤liquid chromatography grade¥ waspurchased from Kanto Chemical ¤Tokyo, Japan¥. All otherchemicals were purchased from commercial sources.ASD formulation of TL:

PreparationA solvent evaporation method was used for preparation

of ASD formulation containing 10% ¤w/w¥ TL. Briefly,crystalline TL ¤30mg¥ and hydrophilic polymer ¤270mg¥were dissolved in 17mL of 90% ¤v/v¥ 1,4-dioxane solution.The solution was mixed well and frozen at %80ôC. Thesample was freeze-dried using an FD-81 freeze-dryer. Afterdrying, the sample was sieved with two different mesh sizesof sieves ¤1 and 2mm¥.UPLC/ESI-MS analysis of TLThe amount of TL in the obtained dry powder was

determined by an internal standard method using aWaters Acquity UPLC system ¤Waters, Milford, MA,USA¥, which included a binary solvent manager, samplemanager, column compartment, and SQD connected withMassLynx software. An Acquity UPLC BEH C 18 column¤particle size: 1.7 µm, column size: 2.1mm ' 50mm;Waters¥ was used, and the column temperature wasmaintained at 50ôC. The standard ¤diclofenac¥ and sampleswere separated using a gradient mobile phase consisting ofmethanol ¤A¥ and 5mM ammonium acetate ¤B¥ with a flowrate of 0.25mL/min. The gradient conditions of the mobilephase were 0®0.5min, 30% A; 0.5®3.5min, 30®70% A;3.5®4.0min, 90% A; and 4.0®4.5min, 30% A. Peaks forTL and the internal standard were detected at retentiontimes of 2.56 and 3.45min, respectively. Analysis wascarried out using selected ion recording ¤SIR¥ for specificm/z 326 and 294 for TL ªM®H«% and diclofenac ªM®H«%,respectively.Powder X-ray diffraction ¤PXRD¥: A PXRD

pattern was collected using D8 ADVANCE ¤Bruker AXSGmbH, Karlsruhe, Germany¥ with CuKÆ radiation gen-erated at 40mA and 35 kV. Data were obtained from 4ô to40ô ¤2Í¥ at a step size of 0.014ô and scanning speed of4ô/min.Thermal analysis: DSC was performed using a DSC

Q1000 ¤TA Instruments, New Castle, DE, USA¥. The DSCthermograms were collected in an aluminum closed-pansystem using a sample weight of ca. 3mg and a heatingrate of 5ôC/min with a nitrogen purge at 70mL/min. Thetemperature axis was calibrated with indium ¤ca. 5mg,99.999% pure, onset at 156.6ôC¥.Scanning electron microscopy ¤SEM¥: Represen-

tative SEM images of TL samples were taken using ascanning electron microscope, VE-7800 ¤Keyence Corpo-ration, Osaka, Japan¥, without Au or Pt coating. For theSEM observations, each sample was fixed on an aluminumsample holder using double-sided carbon tape.Polarized light microscopy ¤PLM¥: Representative

PLM images of TL samples were taken using a CX41

Satomi ONOUE, et al.380

Copyright © 2012 by the Japanese Society for the Study of Xenobiotics (JSSX)

microscope ¤Olympus Co. Ltd., Tokyo, Japan¥. TL sampleswere examined under various conditions including differ-ential interference contrast and slightly uncrossed polars, andusing a red-wave compensator.Dissolution test: Dissolution testing on TL and ASD

formulations of TL ¤10%, w/w¥ with various polymers wascarried out for 120min in 900mL of HCl solution ¤pH 1.2¥with constant stirring of 50 rpm in a dissolution testapparatus NTR 6100A ¤Toyama Sangyo, Osaka, Japan¥ at37ôC. Each powder sample ¤450 µg of TL¥ was weighed inthe dissolution vessel ¤final concentration of TL: 0.5 µg/mL¥. Dissolution testing on ASD formulations of TL ¤10®70%, w/w¥ with Eudragit EPO was conducted undera supersaturated condition. Briefly, each powder sample¤3.15mg of TL¥ was weighed in 900mL of HCl solution¤pH 1.2¥, and final concentration of TL in the dissolutionvessel was 3.5 µg/mL, corresponding to ca. 5 times higherconcentration than the equilibrium solubility of TL ¤0.7 µg/mL¥ at pH 1.2. Analyzed samples ¤0.6mL¥ were collected atthe indicated times ¤5, 10, 20, 40 and 60min for dissolutiontesting under sink conditions; and 6, 10, 30, 60, 90 and120min for dissolution testing under supersaturated condi-tions¥ with a Toyama W-PAS-615 auto-sampler ¤ToyamaSangyo¥. The initial volume of dissolution medium wasmaintained by adding 0.6mL of deionized water. Thecollected samples were diluted with an equal volume ofmethanol, and filtered through a 0.2-µm membrane filter¤Millex LG, Millipore, Billerica, MA, USA¥. The concen-trations of TL were determined by Waters UPLC/ESI-MSas described in Materials and Methods ¤UPLC/ESI-MS analysisof TL¥.Spectroscopic analyses:

Fourier transform infrared ¤FT-IR¥ spectroscopyFT-IR analysis was conducted to evaluate drug-polymer

interaction. Briefly, powder samples were prepared bymixing approximately 2 to 3mg of TL and the physicalmixture or ASD formulation with approximately 300mgof KBr, and the mixture was pressed for preparation of theKBr disk. Spectra were recorded on IR Prestige-21 with IRsolution software ¤Shimadzu, Kyoto, Japan¥, and 40 scanswere performed with a resolution of 4 cm%1.Near-infrared ¤NIR¥ spectroscopyNIR analysis was also carried out for further character-

ization of possible interaction between TL and polymer.Briefly, powder samples of TL and the physical mixture orASD formulation were placed in cylindrical glass vials, andNIR spectroscopic analysis ¤32 scans¥ using a Bruker MPAsystem with a diffuse-reflectance integrating-sphere ¤BrukerOptics, Ettlingen, Germany¥ was performed in reflectancemode with a resolution of 2 nm over a range of 800 to2,800 nm.Stability testing: Stability study of the TL samples was

carried out at 40 + 2ôC/75 + 5% relative humidity ¤RH¥or 60 + 2ôC for 4 weeks in a stability chamber ¤LabcarePvt. Ltd., Mumbai, India¥. Samples after storage were

subjected to the UPLC/ESI-MS, XRPD, DSC, and PLMexperiments.Pharmacokinetic studies:

AnimalsMale Sprague®Dawley rats, weighing ca. 400 g ¤11 weeks

of age; Japan SLC, Shizuoka, Japan¥, were housed two percage in the laboratory with free access to food and water,and maintained on a 12-h dark/light cycle in a room withcontrolled temperature ¤24 + 1ôC¥ and humidity ¤55 + 5%¥.All procedures used in the present study were conductedin accordance with the guidelines approved by the Institu-tional Animal Care and Ethical Committee of the Universityof Shizuoka.Pharmacokinetic studyAfter intravenous administration of TL ¤0.5mg/kg¥

dissolved in DMSO, or oral administration of TL samples¤10mg TL/kg¥ suspended in 1mL of distilled water, bloodsamples were obtained at a volume of 400 µL from the tailvein of unanesthetized rats at the indicated times ¤0.25,0.5, 1, 3, 6, 12 and 24 h¥. Each blood sample ¤400 µL¥ wascentrifuged at 10,000'g to prepare serum samples. Theserum samples were kept frozen at below %20ôC until theywere analyzed. TL concentrations in serum were determinedby UPLC/ESI-MS. In brief, 100 µL of methanol was addedto 50 µL of serum sample, and the solution was centrifugedat 3,000 rpm for 10min. The supernatant was filtratedthrough the 0.2-µm filter, and then the filtrate was analyzedby UPLC/ESI-MS, as described in Materials and Methods¤UPLC/ESI-MS analysis of TL¥. The pharmacokinetic param-eters for TL were calculated by means of noncompartmentalmethods using the WinNonlin¬ program ¤Ver. 4.1, Phar-sight Corporation, Mountain View, CA, USA¥.Statistical analysis: For statistical comparisons, one-

way analysis of variance ¤ANOVA¥ with pairwise comparisonby Fisherös least significant difference procedure was used.A p value of less than 0.05 was considered significant forall analyses.

Results and Discussion

Physicochemical characterization of ASD formu-lations: In the present study, ASD formulations of TL wereprepared by a solvent evaporation method with the use ofvarious polymers, including HPC¤SSL¥, HPC¤L¥, HPC¤H¥,Eudragit EPO, HPMC, PVP¤K30¥, PVP¤K90¥, and KollidonVA64 ¤Table 1¥. An ASD formulation can be defined asa distribution of active ingredients in amorphous formsurrounded by inert carriers.14¥ To clarify amorphization ofTL in the SD formulations, the physical state of eachformulation was evaluated by PXRD and DSC ¤Fig. 1¥. Asshown in Figure 1A, several intense peaks were observed inthe PXRD pattern of crystalline TL, and they were indicativeof the most stable anhydrous form ¤Form I¥.15¥ In contrast, allthe ASD formulations of TL exhibited a halo diffractionpattern ¤Fig. 1A and Table 1¥, and typical peaks for Form Ior other known crystalline forms were negligible. In addition

Amorphous Solid Dispersion of Tranilast with Improved Solubility and Bioavailability 381


to PXRD analysis, the thermal behaviors of ASD formula-tions were studied for characterization of the physicochemicalstatus of TL in these formulations. According to the DSCthermograms of several forms of TL ¤Fig. 1B and Table 1¥,crystalline TL produced a melting endotherm at 212ôC,the thermal event of which was almost identical to thatof Form I as reported previously.15¥ None of the preparedASD formulations of TL displayed any thermal events in theexamined temperature range. The lack of endothermal peaksof TL in ASD formulations suggested that TL in theseformulations might exist in a high-energy amorphous state.ASD formulation can be classified according to the molecularinteraction of drug and carriers in solid solutions, solidsuspensions, or a mixture of both.16¥ In particular, drug andcarrier are totally miscible and soluble in amorphous solidsolutions, developing a homogeneous molecular interactionbetween them. On the basis of formulation process andphysicochemical data, the ASD formulations of TL that weprepared might be amorphous solid solutions.For further characterization, SEM and PLM experiments

were carried out for ASD formulations of TL ¤Fig. 2¥. In theSEM images, crystalline TL showed particles that werepredominantly dispersed and irregularly shaped, with sizesranging over about 10®100 µm ¤Fig. 2A-I¥, whereas theASD formulation of TL had the appearance of typical, flakyfreeze-dried material ¤Figs. 2A-II, 2A-III and 2A-IV¥.Similar morphological transitions were also observed inother SD formulations with various polymers ¤data notshown¥. Thus, SEM micrographs of crystalline TL and ASDformulations revealed clear changes in the morphology ofthe powder particles after lyophilization owing to theevident formation of SD. There seemed to be a markedincrease in the surface area of the drug substance when

Table 1. Physicochemical properties of TL formulations

Molecular weightof polymers ¤Da¥ PXPD DSC

¤endotherm¥ PLM Initial dissolution rate¤h%1; mean + 95% CI¥

Crystalline TL ¯ Form I 212ôC Intensebirefringence

0.0042 + 0.0020¤0.907¥

Solid dispersion of TL withHydroxypropylcellulose ¤HPC-SSL¥ 15,000®30,000 Halo Pattern Negligible Weak

birefringence6.8 + 1.4¤0.982¥

Hydroxypropylcellulose ¤HPC-L¥ 55,000®70,000 Halo Pattern Negligible Weakbirefringence

5.4 + 1.5¤0.956¥

Hydroxypropylcellulose ¤HPC-H¥ 250,000®400,000 Halo Pattern Negligible Weakbirefringence

0.80 + 0.10¤0.996¥

Hydroxypropylmethylcellulose ¤HPMC¥ ca. 86,000 Halo Pattern Negligible Weakbirefringence

4.1 + 1.6¤0.999¥

Polyvinylpyrrolidone ¤PVP K30¥ ca. 40,000 Halo Pattern Negligible Weakbirefringence

3.1 + 0.53¤0.917¥

Polyvinylpyrrolidone ¤PVP K90¥ ca. 360,000 Halo Pattern Negligible Weakbirefringence

2.7 + 0.28¤0.972¥

Polyvinylpyrrolidone-polyvinylacetate copolymer ¤KollidonVA64¥ 45,000®70,000 Halo Pattern Negligible Weakbirefringence

13 + 2.7¤0.949¥

Polyacrylates and polymethacrylates ¤Eudragit EPO¥ca. 150,000 Halo Pattern Negligible No

birefringence16 + 3.0¤0.911¥

*Figures in parentheses are correlation coefficient values.

Fig. 1. Physicochemical characterization of TL formulationsusing (A) powder X-ray diffraction and (B) differential scanningcalorimetry(I) Crystalline TL, (II) ASD formulation of TL with Kollidone VA64(ASD-TL/KOL), (III) ASD formulation with HPC (SSL) [ASD-TL/HPC(SSL)], and (IV) ASD formulation with Eudragit EPO (ASD-TL/EUD).



compared with that of crystalline TL alone. Although TLin the new formulations was expected to be amorphous,PLM appearance revealed that TL molecules in most ASDformulations might be slightly recrystallized during thelyophilizing process as evidenced by weak birefringence¤Table 1 and Fig. 2B¥. Interestingly, only ASD formulationof TL with Eudragit EPO ¤ASD-TL/EUD¥ showed nobirefringence ¤Fig. 2B-IV¥. In general, detection sensitivityof PLM for crystallinity assessment is much higher than thatof other analytical techniques such as PXRD and DSC. Thiscould partially explain the data discrepancy. Only a verysmall amount of crystalline TL might be contained in mostASD formulations, whereas TL in the ASD-TL/EUD could

be amorphized more completely. The enthalpy, entropy,and free energy of an amorphous drug can be much higherthan those of its crystalline counterpart,17¥ so that the excessfree energy of an amorphous TL may result in improvementin its dissolution behavior.Dissolution behavior of ASD formulations: To

clarify possible enhancement of dissolution behavior by ASDapproaches, dissolution testing for ASD formulations wascarried out in HCl solution ¤pH 1.2¥ to simulate gastricconditions ¤Fig. 3¥. Initial dissolution of TL from thepowders was likely to follow first-order kinetics, and thecorresponding first-order dissolution rates are given inTable 1. Poor dissolution behaviors were seen in crystallineTL at a first-order dissolution rate of 0.004 h%1, and thephysical mixture of TL and polymer also exhibited limiteddissolution behavior in the same manner as crystalline TL¤data not shown¥. There was marked improvement indissolution behavior in most ASD formulations of TL.Among all the ASD formulations tested, ASD formulation ofTL with Eudragit EPO and Kollidon VA64 showed thehighest concentration of dissolved TL compared with that ofcrystalline TL, the first-order dissolution rates of whichwere calculated to be ca. 3,000-fold higher than that ofcrystalline TL. These observations were consistent withprevious reports, showing that the formulation of poorlysoluble drugs as SD could lead to a marked improvementin the dissolution properties.18¥ Typical mechanisms for theimprovement of dissolution characteristics of drugs by theASD approach are reduction in particle size, the absence ofcrystallinity, and improved wettability.19¥ On the basis ofthe present findings, taken together with the preparationprocess, the presence of all the hydrophilic polymers testedmight have a beneficial effect on wettability of TL, possiblyby a homogeneous molecular interaction of drug andpolymers.As observed in the PXRD and DSC analyses, most TL

could adopt an amorphous state in the ASD powders,

Fig. 2. Morphological observations of TL samples using(A) scanning electron microscopy and (B) polarized lightmicroscopy(I) Crystalline TL, (II) ASD formulation of TL with Kollidone VA64(ASD-TL/KOL), (III) ASD formulation with HPC (SSL) [ASD-TL/HPC(SSL)], and (IV) ASD formulation with Eudragit EPO (ASD-TL/EUD).Each bar represents 100 µm.

Fig. 3. Dissolution profiles of TL formulation in acidic solution(pH 1.2)�, crystalline TL; , ASD formulation of TL with HPC (SSL); , withHPC (L); , with HPC (H); , with PVP (K30); , with PVP (K90);, with Eudragit EPO; , with HPMC; and , with Kollidon VA64.

Each bar represents mean + SE of 3 independent experiments.



although small particles of crystalline TL were visible in theASD-TL with several polymers according to PLM experi-ments. The remaining nano- or micro-crystals dispersed inthe amorphous matrix could serve in part as seeds forcrystallization,20¥ eventually leading to reduced wettabilityand limited clinical outcomes. From these results, ASD-TL/EUD was chosen as a promising ASD formulation of TLand subjected to further formulation optimization andphysicochemical/pharmacokinetic characterizations.Interaction between TL and polymer: Theoret-

ically, amorphous TL could be molecularly dispersed in thematrix carrier in an ASD formulation, and the interactionbetween these molecules might result in better amorphiza-tion of TL. To evaluate possible interaction between TL andpolymer molecules, spectroscopic characterization on ASD-TL/EUD was carried out by FT-IR and NIR spectroscopicanalyses ¤Fig. 4¥. There were no significant differences in IRspectral patterns between crystalline TL and the physicalmixture of TL and Eudragit EPO ¤data not shown¥. Incontrast, a slight transition of the bands at 1,650®1,700cm%1 was observed in ASD-TL/EUD ¤Fig. 4A¥, corre-sponding to the C©O stretching band. The variations inIR spectral patterns of TL samples are considered to beindicative of different molecular environments, especially

around the carboxylate moiety of TL samples. For furtherelucidation, second derivative NIR spectra of TL sampleswere superimposed ¤Fig. 4B¥. As observed in FT-IR spectralanalysis, crystalline/amorphous TL and the physical mixtureof TL and copolymer exhibited similar spectral patterns inthe bands at 1,880®1,920 nm, representing the secondovertones of the carboxylic acid C©O stretch.21¥ Interest-ingly, a clear transition was observed in the ASD-TL/EUDwith an intense negative peak at 1,910 nm and a positivepeak at 1,885 nm. These data provided further evidence forthe interaction between TL and the copolymer. EudragitEPO is a cationic copolymer based on dimethylaminoethylmethacrylate and neutral methacrylic esters, so the cationicfunctional group in the copolymer can electrostaticallyinteract with a carboxylic acid of the TL molecule in theASD formulation. The interaction with the polymer mightbe attributable to the improved dissolution behavior of TLand stabilization of the amorphous state. In addition to thesolid-state physicochemical properties, the cationic copoly-mer can dissolve in solution of pH g 5; therefore, thecopolymer tends to dissolve fast in gastric fluid. Dissolvedcopolymer might also interact with TL molecule in amedium, and the electrostatic interaction could lead toprevention of precipitation or re-crystallization.Preferable loading amount of TL in a Eudragit

EPO-based ASD formulation: Despite the developmentof extensive expertise with ASD formulations, they are notwidely used in commercial products, mainly because thereis a possibility that the amorphous state may undergocrystallization during processing or storage.16¥ In particular,drugs at higher concentration are often present in thecrystalline form within ASD formulations or re-crystallizeover time, resulting in unstable formulations with lowerdissolution rates. In the present study, several ASD-TL/EUDs with various drug concentrations ranging from 10to 90% were prepared for comparison. Preliminary PLMexperiments demonstrated that partial re-crystallizationoccurred in the Eudragit EPO-based formulations of TLwith drug concentrations of 80% or higher ¤data not shown¥.For further optimization, dissolution testing under super-saturated conditions ¤ca. 5 times higher than the equilibriumsolubility of TL at pH 1.2¥ was carried out for ASD-TL/EUDs with various drug concentrations under 70% ¤Fig. 5¥.All the ASD-TL/EUDs exhibited rapid dissolution behavior,and complete supersaturation was achieved with the 2 htesting period. The ASD-TL/EUDs with a drug concen-tration of under 50% showed a tremendous increase indissolution rate as evidenced by ca. 4®5-fold greater concen-tration of the dissolved TL than equilibrium solubility ¤Ceq¥of TL at 5min. Interestingly, reduced dissolution behaviorwas seen in the ASD-TL/EUDs with drug concentrationsof over 60% in a drug load-dependent manner. The muchhigher supersaturated concentration and its duration havethe potential to enhance the oral bioavailability of this poorlywater-soluble drug.

Fig. 4. Spectroscopic analyses of ASD formulation withEudragit EPO (ASD-TL/EUD)(A) Baseline-corrected and normalized IR and (B) superimposedsecond derivative NIR spectra. Black line, crystalline TL; yellow line,amorphous TL without polymer; blue line, physical mixture of TL andEudragit EPO; red line, ASD-TL/EUD; and green line, Eudragit EPO.



Temperature and humidity stresses sometimes affect thestorage stability of amorphous drugs since they may increasedrug mobility and promote drug crystallization. Since thesephysicochemical transitions may lead to a decreased solu-bility and dissolution rate of an ASD formulation of TL, asolid-state stability study was conducted for the ASD-TL/EUDs with a drug concentration under 50% ¤Table 2¥. TheASD-TL/EUD was stored at 60ôC or 40ôC/75% RH for 2 or4 weeks, followed by PLM experiments to evaluate possibletransition to a crystalline state. There were no significantchanges in PLM observations of ASD formulations stored at60ôC, suggesting no re-crystallization during storage. Thesefindings were consistent with the results from PXRD andDSC analyses on these formulations ¤data not shown¥. Incontrast, weak birefringence was observed in ASD-TL/EUDs with a drug concentration of 50% stored at 40ôC/75%RH for 4 weeks, although the physicochemical transition wasnegligible in both PXRD and DSC analyses. The very slightre-crystallization of TL might be caused only in an ASDformulation with a relatively high drug load. The EudragitEPO used in the present study can absorb moisture, whichmay result in phase separation, crystal growth, or conversionfrom the amorphous to the crystalline state during long-termstorage. Since slight re-crystallization was observed in theASD-TL/EUD with a drug concentration of 50% afterstorage at 40ôC/75% RH for 4 weeks, dissolution testing inacidic solution ¤pH 1.2¥ under supersaturated conditions wasalso carried out to clarify any possible reduction in dissolu-tion behavior. Although the aged ASD-TL/EUDs with drugconcentrations of 50% still exhibited rapid dissolutionin acidic solution, the degree of supersaturation ¤C/Ceq¥decreased to ca. 2.6 ¤data not shown¥. Thus, there was ca.42% reduction in maximum supersaturation of TL afterstorage at 40ôC/75% RH for 4 weeks, and this finding was

consistent with the PLM observation on the aged ASD-TL/EUDs with a drug concentration of 50%. From the presentresults, moisture protection might be needed to preventphysicochemical transition of ASD formulations of TL.Currently, TL is administered orally at a dose of 300mg/

day in Japan for the clinical treatment of several inflam-matory diseases. In general, poorly water-soluble drugs haveto be administered at a high dose; therefore, only a smallamount of excipients can be added to the formulation inorder to preclude difficulties relating to patient compliance.Thus, it is important to be able to produce SD formulationscontaining as much of the active pharmaceutical ingredient aspossible. On the basis of the present findings on crystallinity,dissolution behavior, and solid-state stability, the maximumload with TL for Eudragit EPO was deduced to be as muchas 50%, although moisture protection might be needed forlong-term stability.Pharmacokinetic profiling of the Eudragit EPO-

based ASD formulation: The observations on theimproved dissolution properties of ASD-TL prompted us toclarify the possible improvement in the oral bioavailability ofTL, so the pharmacokinetic behaviors of crystalline TL andASD-TL/EUD with a drug load of 50% were assessed inrats. The blood concentration®time profiles of TL in ratsafter oral administration of crystalline TL and ASD-TL/EUD ¤10mg TL/kg¥ are shown in Figure 6, and relevantpharmacokinetic parameters including Cmax, Tmax, AUC0®inf,and absolute bioavailability are listed in Table 3. SerumTL levels in the blood were found to be very low whencrystalline TL was administered orally, and the Cmax andAUC0®inf values were 0.1 µg/mL and 0.8 µg&h/mL, respec-tively. The blood concentration of TL decreased graduallywith apparent elimination kinetics of 0.17 h%1. On the basisof the AUC0®inf value ¤3.29 µg&h/mL¥ of intravenouslyadministered TL ¤0.5mg/kg¥, absolute bioavailability ofcrystalline TL was calculated to be as low as 1.2%. Incontrast, ASD-TL/EUD showed improved pharmacokineticbehavior compared with crystalline TL. Oral administrationof ASD-TL/EUD resulted in rapid elevation of TL bloodlevels up to Cmax 4.6 µg/mL, and the AUC0®inf value was

Fig. 5. Supersaturation in acidic solution (pH 1.2) from disso-lution of ASD formulation with various concentrations of TL(10–70%), ASD formulation of TL with drug load of 10%; , 40%; , 50%; ,

60%; and , 70%. Each value is expressed as the measuredconcentration of dissolved TL (C) vs. equilibrium solubility of TL(Ceq), and each bar represents mean + SE of 3 independentexperiments.

Table 2. Stability test on ASD formulation with Eudragit EPO(ASD-TL/EUD)

Storage conditionsPLM observations on ASD-TL/EUD

Loading amount10% 30% 40% 50%

Initial ®No

birefringenceNo

birefringenceNo

birefringenceNo

birefringence

2weeks

40ôC/75% RH N.C. N.C. N.C. N.C.

60ôC N.C. N.C. N.C. N.C.

4weeks

40ôC/75% RH N.C. N.C. N.C. Weakbirefringence

60ôC N.C. N.C. N.C. N.C.

N.C.: not changed.



calculated to be 14.6 µg&h/mL. Absolute bioavailability ofASD-TL/EUD was calculated to be ca. 22%. Thus, thereappeared to be ca. 19-fold enhancement in the oral bio-availability of TL with the use of the ASD approach. Thesefindings were relatively consistent with the results from thedissolution test, demonstrating the accelerated dissolutionbehavior of ASD-TL/EUD. Interestingly, there was ca. 70%reduction of inter-individual variation in AUC from ASD-TL/EUD ªcoefficient of variation ¤CV¥: 13%« compared withthat from crystalline TL ¤CV: 50%¥. Since solubility of TLis variable depending on the pH of the solvent medium,the individual variability in gastric condition or gastric dys-functions may affect the drug release in gastric fluid andthereby cause inconsistent absorption. However, ASD for-mulations exhibited a marked increase in solubility of TL atacidic pH, and this might be attributable to improved andconsistent absorption with low variation in bioavailability.Previously, our group developed a CSD formulation

of TL, in which fine crystalline particles of TL were dis-persed in solid water-soluble matrices.11¥ There was markedimprovement in the dissolution behavior of the CSD formu-lation compared with that of crystalline TL, and pharmaco-kinetic data on the CSD formulation were indicative of highsystemic exposure with an increase of oral bioavailabilityby ca. 31-fold for crystalline TL. In this context, the ASDapproach was slightly less effective for improving oral bio-availability than the CSD formulation strategy, althoughclinical application of the ASD formulation of TL mightalso provide better clinical outcomes than current TL-basedmedication for inflammatory or fibrotic diseases. In spite ofits pharmaceutical usefulness, there would be difficulty inscaling-up for the manufacture of the CSD formulation, sothe development of new optimized manufacturing techniqueswith high scalability have been attempted in academic andindustrial research. In addition, the CSD approach employ-ing wet-milling techniques might be unsuitable for pharma-ceutical substances with low melting points since the

generation of friction heat would result in partial amor-phization during the wet-milling process. In contrast, potentscalable manufacturing processes to obtain ASD formulationshave been developed, which include the melting method,the solvent-evaporation method, and the solvent-wettingmethod.16¥ These formulation techniques would be appli-cable to poorly water-soluble drugs with a low melting pointas an alternative to the CSD approach. On the basis ofphysicochemical and pharmacokinetic properties, takentogether with their manufacturability, ASD approaches canbe a viable dosage option for improving the therapeuticpotential of TL through enhancement of its wettability andoral bioavailability.

Conclusion

In the present study, several ASD formulations of TLwere prepared by a solvent evaporation method. Therewas significant improvement in the dissolution behavior ofmost ASD formulations, especially the ASD formulationwith Eudragit EPO, which exhibited the highest dissolutionbehavior with better amorphization. Spectroscopic analyseson the ASD formulation suggested the interaction of TLwith the copolymer, and the maximum drug load wasdeduced to be ca. 50% on the basis of the results fromdissolution and stability studies. According to the pharmaco-kinetic profiles of ASD-TL/EUD, there was significantimprovement in the systemic exposure of TL with increasesin oral bioavailability by ca. 19-fold. From these observa-tions of improved dissolution and pharmacokinetic behav-iors, the ASD approach could be a promising formulationstrategy for enhancing the bioavailability of TL, possiblyleading to better clinical outcomes for the treatment ofinflammatory and fibrotic diseases.

Acknowledgments: The authors are grateful to KisseiPharmaceutical Co., Ltd. ¤Nagano, Japan¥, for kindly pro-viding tranilast.

References

1¥ Koda, A., Nagai, H., Watanabe, S., Yanagihara, Y. and Sakamoto,K.: Inhibition of hypersensitivity reactions by a new drug, N¤3$,4$-dimethoxycinnamoyl¥ anthranilic acid ¤N-5$¥. J. Allergy Clin.Immunol., 57: 396®407 ¤1976¥.

Fig. 6. Blood TL concentrations in rats after intravenous or oraladministration of TL formulation, TL solution (0.5mg/kg, i.v.); �, crystalline TL (10mg/kg, p.o.);

and , ASD of TL with Eudragit EPO (ASD-TL/EUD, 10mg TL/kg,p.o.). Data represent mean + SE of 4–6 experiments.

Table 3. Pharmacokinetic parameters of TL formulationsfollowing oral administration

Cmax¤µg/mL¥

Tmax¤h¥

AUC0®inf

¤µg&h/mL¥BA¤%¥

Crystalline TL¤10mg/kg¥ 0.1 + 0.0 1.8 + 0.7 0.8 + 0.4 1.2

ASD-TL/EPO¤10mg TL/kg¥ 4.6 + 0.3 0.54 + 0.15 14.6 + 1.9 22.2

Cmax: maximum concentration; Tmax: time to maximum concentration; AUC0®inf:area under the curve of blood concentration vs. time from t © 0 to t © W afteradministration. Values are expressed as mean + SE of 4®6 experiments.



http://www.ncbi.nlm.nih.gov/sites/entrez?cmd=Retrieve&db=PubMed&list_uids=131140&dopt=Abstract


2¥ Suzawa, H., Kikuchi, S., Ichikawa, K. and Koda, A.: Inhibitoryaction of tranilast, an anti-allergic drug, on the release of cytokinesand PGE2 from human monocytes-macrophages. Jpn. J. Pharmacol.,60: 85®90 ¤1992¥.

3¥ Konneh, M.: Tranilast Kissei Pharmaceutical. IDrugs, 1: 141®146¤1998¥.

4¥ Ikeda, H., Inao, M. and Fujiwara, K.: Inhibitory effect of tranilaston activation and transforming growth factor beta 1 expression incultured rat stellate cells. Biochem. Biophys. Res. Commun., 227: 322®327 ¤1996¥.

5¥ Uno, M., Kurita, S., Misu, H., Ando, H., Ota, T., Matsuzawa-Nagata, N., Kita, Y., Nabemoto, S., Akahori, H., Zen, Y.,Nakanuma, Y., Kaneko, S. and Takamura, T.: Tranilast, anantifibrogenic agent, ameliorates a dietary rat model of nonalcoholicsteatohepatitis. Hepatology, 48: 109®118 ¤2008¥.

6¥ Mori, H., Kawada, K., Zhang, P., Uesugi, Y., Sakamoto, O.and Koda, A.: Bleomycin-induced pulmonary fibrosis in geneticallymast cell-deficient WBB6F1-W/Wv mice and mechanism ofthe suppressive effect of tranilast, an antiallergic drug inhibitingmediator release from mast cells, on fibrosis. Int. Arch. Allergy Appl.Immunol., 95: 195®201 ¤1991¥.

7¥ Mori, H., Tanaka, H., Kawada, K., Nagai, H. and Koda, A.:Suppressive effects of tranilast on pulmonary fibrosis and activationof alveolar macrophages in mice treated with bleomycin: role ofalveolar macrophages in the fibrosis. Jpn. J. Pharmacol., 67: 279®289¤1995¥.

8¥ Izumi, K., Mizokami, A., Li, Y. Q., Narimoto, K., Sugimoto, K.,Kadono, Y., Kitagawa, Y., Konaka, H., Koh, E., Keller, E. T. andNamiki, M.: Tranilast inhibits hormone refractory prostate cancercell proliferation and suppresses transforming growth factor beta1-associated osteoblastic changes. Prostate, 69: 1222®1234 ¤2009¥.

9¥ Nie, L., Oishi, Y., Doi, I., Shibata, H. and Kojima, I.: Inhibition ofproliferation of MCF-7 breast cancer cells by a blocker of Ca¤2¦¥-permeable channel. Cell Calcium, 22: 75®82 ¤1997¥.

10¥ Noguchi, N., Kawashiri, S., Tanaka, A., Kato, K. and Nakaya, H.:Effects of fibroblast growth inhibitor on proliferation and metastasisof oral squamous cell carcinoma. Oral Oncol., 39: 240®247 ¤2003¥.

11¥ Kawabata, Y., Yamamoto, K., Debari, K., Onoue, S. and Yamada,S.: Novel crystalline solid dispersion of tranilast with high

photostability and improved oral bioavailability. Eur. J. Pharm.Sci., 39: 256®262 ¤2010¥.

12¥ The Society of Japanese Pharmacopoeia: Tranilast. Japanese OrangeBook No. 12 ¤Approved drug products with therapeutic equivalenceevaluations¥, Tokyo, Yakuji Nippo Ltd., 2002, p 171.

13¥ Joe, J. H., Lee, W. M., Park, Y. J., Joe, K. H., Oh, D. H., Seo,Y. G., Woo, J. S., Yong, C. S. and Choi, H. G.: Effect of thesolid-dispersion method on the solubility and crystalline propertyof tacrolimus. Int. J. Pharm., 395: 161®166 ¤2010¥.

14¥ Chiou, W. L. and Riegelman, S.: Pharmaceutical applications ofsolid dispersion systems. J. Pharm. Sci., 60: 1281®1302 ¤1971¥.

15¥ Vogt, F. G., Cohen, D. E., Bowman, J. D., Spoors, G. P., Zuber,G. E., Trescher, G. A., Dellöorco, P. C., Katrincic, L. M.,Debrosse, C. W. and Curtis Haltiwanger, R.: Structural analysis ofpolymorphism and solvation in tranilast. J. Pharm. Sci., 94: 651®665¤2005¥.

16¥ Vasconcelos, T., Sarmento, B. and Costa, P.: Solid dispersions asstrategy to improve oral bioavailability of poor water soluble drugs.Drug Discov. Today, 12: 1068®1075 ¤2007¥.

17¥ Friesen, D. T., Shanker, R., Crew, M., Smithey, D. T., Curatolo,W. J. and Nightingale, J. A.: Hydroxypropyl methylcelluloseacetate succinate-based spray-dried dispersions: an overview. Mol.Pharm., 5: 1003®1019 ¤2008¥.

18¥ Ren, F., Jing, Q., Tang, Y., Shen, Y., Chen, J., Gao, F. andCui, J.: Characteristics of bicalutamide solid dispersions andimprovement of the dissolution. Drug Dev. Ind. Pharm., 32: 967®972¤2006¥.

19¥ Leuner, C. and Dressman, J.: Improving drug solubility for oraldelivery using solid dispersions. Eur. J. Pharm. Biopharm., 50: 47®60¤2000¥.

20¥ Yada, S., Ohya, M., Ohuchi, Y., Hamaura, T., Wakiyama, N.,Usui, F., Kusai, A. and Yamamoto, K.: Solid phase transition ofCS-891 enantiotropes during grinding. Int. J. Pharm., 255: 69®79¤2003¥.

21¥ Savolainen, M., Heinz, A., Strachan, C., Gordon, K. C., Yliruusi,J., Rades, T. and Sandler, N.: Screening for differences in theamorphous state of indomethacin using multivariate visualization.Eur. J. Pharm. Sci., 30: 113®123 ¤2007¥.

































physicochemical and pharmacokinetic characterization of

Documents