RESEARCH"
`
`An Official Journal of the American Association of Pharmaceutical Scientists
`
`Lupin Ex. 1042 (Page 1 of 10)
`
`

`

`PHARMACEUTICAL RESEARCH
`An Official Journal of the American Association of Pharmaceutical Scientists
`
`Pharmaceutical Research publishes innovative basic research and technological advances ia the pharlnaceutical-biomediuai sciences: Research areas covered in
`the journaI include: pharlnaceutics and drug delivery, pharmacokineties and pharlnacodynamics, drug metabolism, pharlnaeology and toxicology, tnedioinaI
`chemistry, natural products chemistry, analytical chelnistry, chelnical kinetics and drug stability, biotechnology, pharmaceutical technology, and clinical inves-
`tigations, as well as articles on the social, econmnic, or lnanagement aspects of the pharmacenticaI sciences.
`
`EDITOR-IN-CHIEF
`Vincent H. L. Lee, Department of Pharmaceutical Sciences, University of
`Southern California, Los Angeles, California
`
`EDITOR EUROPE
`Dean Crommelin, Utrecht University, Utrecht, The Netherlands
`
`EDITOR--JAPAN
`Mitsuru HasMda, Kyot0 University, Kyoto, Japan
`
`ASSOCIATE EDITORS
`William E. Evans, St. Jude Children’s Research Hospital, Memphis,
`Tennessee
`Kinam Park, Purdue University, West Lafayette, ~[ndiana
`
`ASSOCIATE EDITORS--EUROPI~
`Meindert Danhof, Leiden University, Leiden, The Netherlands
`Josef J. Tukker, Utrecht University, Utrecht, The Netherlands
`
`EDITOR--SPECIAL FEATURES
`Randall J. Mrsny, Genentech, South San Francisco, California
`
`EDITORIAL ADVISORY BOARD
`Gordon L. Amidon, University of Michigan, Ann Arbor, Michigan
`Per Artursson, Uppsala UniverSity, Uppsala, Sweden
`Jessie L~-S. Au, Ohio State University, Columbus, Ohio
`Clare Bergson, Medical College of Georgia, Augusta. Georgia
`Ronald T. Borehardt, University of Kansas, Lawrence, Kansas
`Harry Brittain, Center for Pharmaceutical Physics, Milford, New Jersey
`John F. Carpenter; University of Colorado, Denver, Colorado
`Albert H. L.. Chow, Chinese University of Hong Kong, Shatin, Hong Kong
`Alice Clark~ University of Mississippi, University, Mississippi
`Paulo Columbo, University of Parma, Parma, Italy
`Michael Corbo, Bristol Meyers Squibb, Hillside, New Jersey
`Patrick Couvreur, Utflversit6 de Paris-Sud, Chatenay-Malabry, France
`Jennifer B. Dresslnan~ Universitat Frankfurt, Frankfurt, Germaoy
`Ronald B. Emeson, VanderbJlt University, Nashville, Tennessee
`Ho-Leung Funs, State University of New York-Buffalo, Amherst, New
`York
`Kathleen M. Glacominl, University of California at San Francisco, San
`Francisco, California,
`Robert Gurny, Universit6 de Geneve, Geneva, Switzerland
`Margareta Hammarlund-Udenaes, Uppsala University, Uppsala, Sweden
`Inn Haworth, University of Southern California, Los Angeles, California
`Masahiro Hayashi, Science University of Tokyo, Toky, Japan
`Susan Hershenson, A1ngen Inc., Thousand Oaks, California
`1htsuji Iga, University of Tokyo Hospital, Tokyo, Japan
`Lisbeth Ilium, West Pharmaceutical Service, Nottingham, United Kingdoln
`Ken-iehi Inul, Kyoto University Hospital, Kyoto, Japan
`Myron K. Jaeobson, University of Kentucky, Lexington, Kentucky
`Rudy L. Jnllano, University of North Carolina at Chapel Hill, Chapel Hill,
`North Carolina
`Hans E~ Junglnger~ University of Leiden, Leiden, The Netherlands
`Tetsuya Kamataki, Hokkaido University, Sapporo, Japan
`
`James Kamiuski, Schering-P1ough Corp., Kenilworth, New ,lesrsey
`Nell Kaplowitz, University of Southern Californa~ Los Angeles, Catifornia
`Chong-Kook Kim, Seoul National University, SeouI, Korea
`Kwang-Jin Kim, University of Southern Catifornia, Los Angeles, California
`Sung Wan Kim, University of Utah, Salt Lake City, Utah
`Young Ha Kim, Korea h~stitute of Science & Engineering, Seol, Korea
`Toshlklro Kinmra, Okayama University, Okayama, Japan
`Thomas Kissel~ Universit~it Marb~irg, Marbiirg, Gerlnany
`Joachim Kohn, Rutgers Uttiversity, Piscataway, New Jersey
`Jindrieh Kopecek, University of Utah, Salt Lake City, Utah
`Richard Laionde, Parke-Davis Pharmaceutical Research, Ann Arbor,
`Michigan
`Hat Bang Lee, Korea Research Institute of Chemical Technology, Tac-
`icon, Korea
`Thomas M. Ludden, GloboMax, Hanover, Maryland
`Susan Lunte, University of Kansas, Lawrence, Kansas
`Kristina Lnthman, Universitetet I Tromso, Tromso, Norway
`Panos Macheras, University of Athens, Athens, Greece
`Joachim Mayer, University of Lausanne, Lausanne, Switzerland
`Duane D. Milles, University of Tennessee, Mem’phis, Tennessee
`Tsnneji Nasal, Hoshi University, Tokyo., Japan
`Masahiro Nakano, Kumamoto University Hospital, Kumalnoto, Japan
`Ralpb Riven, Advanced Inhalation Research, Calnbridge, Massachusetts
`Teruo Okano, Tokyo Women’s Medical College, Tokyo, Japan
`Charles Pidegon, Purdue University, West Lafayette, Indiana
`Michael J. Pikal, University of Connecticut, Storrs, Connecticut
`Ronald E. Reid, University of British Columbia, Vancouver, Canada
`Jim E, Riviere, North Carolina State University, Raleigh, North Carolina
`joseph R. Robinson, University of Wisconsin, Madison, Wisconsin
`Wolfgang Sadie, University of California, San Francisco, Califln’nia
`Wei-Chiang Short, University of Southern California, Los Angeles,
`California
`Valentino J. Stella, University of Kansas, Lawrence; Kansas
`Ynichi Sugiyama, University of Tokyo, Tokyo, Japma
`Scan Sullivan, Valentis, Inc., The Woodlands, Texas
`Bonnie L. Svarstad, University of Wisconsin, Madison, Wisconsin
`Bernard Testa, University of Lausanne, Laussane, Switzerland
`Kenneth Thummel, University of Washington, Seattle, Washington
`Eric Tomlinson, GeneMedicine, Inc., The Woodlands, Texas
`Akira Tsuji, Kanazawa University, Kanazawa, Japan
`John Urquhart, Palo Alto, California
`Arto Urtti, University of Kuoplo, Kuopio, Finland
`Timothy S. Wiedmann, University of Minnesota, Minneapolis, MinnesOta
`Keiji Yamamato, Chiba University, Chiba, Japan
`
`BOOK REVIEW EDITOR
`Kinam Park, Purdue University, School of Pbm’macy, West Lafayette,
`Indiana 47097
`
`EDITORIAL ASSISTANTS
`Ruth Ellis-Ballard
`Elizabeth B. Gongora
`Hideo Ueda
`
`Pharmaceutical Research is published monthly by Kluwer Acadelnie/Plenum Publishers, 233 Spriag Street, New York, N.Y. 10013-1578. Pharmaaceuticat Research is abstracted or indexed
`in Bioselenoe Abstracts, Chemical Abstracts, Current Contents/Life Sciences, Exeerpta Mediea, Gazette de I’APGI, Index Medieus, International Pharmaceutical Abstracts, Medical
`Documentation Service/MEDLARS, Reference Update, Science Citation Index, Science Citation Index Expauded, and The ISI Alerting Services. ~ 2000 Plenum Publishing Corporation.
`Pharmaceutical Research participants iu the Copyright Clearance Center (CCC) Transactional Reporthlg Service. The appearance of a code line at the bottom of the first page of an article
`in this journal indicates the copyright owner’s consent that copies of the article may be made for personal or internal use. However, this consent is given on the condition that the copier
`pay the fiat flee of $18.00 per copy pet" article (no additional pet~page fees) directly to the Copyright Clearance Center, Inc, 222 Rosewood Drive, Danvers, Massachusetts 01923, for all
`copying not explicitly permitted by Sections 107 or 108 of the U.S, Copyright Law. Tl~e CCC is a nonprofit clearinghouse for the payment of photocopying fees by libraries and other
`users registered with the CCC. Therefore, this consent does not extend to other kinds of copying, such as copying for general distribution, for advertising or promotional purposes, for
`creating new collective works, or for resale, nor to the reprinting of figures, tables, and text excerpts. 0724-8741/00 $18.00.
`
`Advertising inquiries should be addressed to Advertising Sales, Pharmaceutical Research, McNeill Group, Inc., 301 Oxford Valley Road, Suite 903, Yardley, Pennsylvania 19067--telephone
`(215) 321-9662 and fax (215) 321-9636.
`
`/
`Subscription iuquiries and subseriptlou orders should be addressed to the publisher at Subscription Depariment, Kluwer Academic/Plennm Pebllshers, 233 Spring Street, New York, N.Y.
`10013-1578 or faxed to the Snbscriptlon Department at its number (212) 807-1047, or may be telephoned to the Subseriptlon Department’s Journal Customer Service at (212) 620-8035,
`-8404, -8466, -8472, or -8490, Subscription rates:
`Volume i7, 2000 (12 issues) $840.00 (outside the U.S., $985.00). Prices for individuals subscribers certifying that the, jourual is for their personal use, $I60.00 (outside the U.S,, $187.00).
`
`periodicals postage paid at New York, N.Y., and at additional mailing offices. Postmaster: Send address changes to Pharmaceutical Research, Kluwer Aeademie/Plenum Publishers, 233
`Spring Street, New York, N.Y. 10013-1578.
`
`Printed in the USA.
`
`Lupin Ex. 1042 (Page 2 of 10)
`
`

`

`phw..maceut~cal R~aearch, ~.gL .~7, No. 4, 2000
`
`Reseach Paper
`
`What is tlie True SolubilRy Advantage
`for Amorphous iPharmaceuticals?
`
`Bruno C. Hancock:’3 and Michael Parks~
`
`Reedved September 20, 1999; accepted December 19~ 1999
`
`Purpose. To evakmte the m~gnitude of d~.e solubi[ity advanlage
`amorpho~s pharmaceutical materials when compared to their crystal-
`line comaterparts,
`Methods. The thermaI properties of several &ugs in their amorphous
`and erystallhae slates were determh~ed using differential scanning calo-
`rimetry. From these properties the solubility advantage ~or the alnob
`pSous form was predicted as a ruction of temperature using a shnp/e
`thermalynamic analysis. ’Ilaese predictions were compared m the
`ms u lts of experimental measurements of the aqueous so]ubifities of the
`amoqphous and crystal~ne forms of the drugs at several temperatures.
`Resttlts. By treating each amorphous drug as either an equilibrium
`supra:cooled liquid or a pseudo-equifibrium glass, the soluNlity advan-
`tage eompar~l ~o the most stable cryst~Jline form was predicted to be
`between 10 and 1600 fold. The measured solubility advantage was
`usually conNderably less than ins. aul 1%r one compound sludied in
`detail its temperature dependence was also less than predicted. It was
`calculated that even for partially m~orp~ous materials the appare~x
`solubility enhancement (theoretical or measured) is likely m inflacnee
`m-vitro and in-,vivo dissolufl.on behavior,
`Conclusions. Amorphous phammceuttcals are markedly more soluble
`flaan their crystalline counterparu% however, their experhnentN solubil-
`ity advautage is tN?icaliy less than that preNcted from sunple thermody-
`namic considerations. This appeal3 to be a~e result of difficulties
`determining the solubility of amorphous materials under true equilib-
`rium col~dition~, Simple thermodynamic predictions can provide a u
`fffl indication of the theoretical maximum solubility advantage for
`amorphous pMrmaceuticals, which directly refl~s the drivin g three
`for ~hdr iNfial dissolution.
`
`KEY WORDS: amorphot~,~; crystal: solubility; dissolution.
`
`IN1 R )DUC 1. ION
`
`The existence of drugs m~d excipients in multipte physical
`forms (e.g., polymorphs, ~somers) provides phammcentical sci-
`entists with an opportunity to select the preferred form(s) of
`the materiNs used in a formulation. This ~s very useful since
`critical properties, such as particle morphology and solubility,
`frequently wu:y between the different physical forms of a mate-
`rial. The amorphm~s form of ph~macologica[ly active materials
`has received considerable attmation because in theery this
`represents the most eneNctic solid state of a material (Figure
`1), and thus it should provide the biggest advantage in terms
`of solubility and bioavailabilit¥ (1) Additionally, itmay provide
`significant changes from lhc usual crystalline form in terms of
`its mechanical properties, such as elastic modulus.
`For different crysta!line forms (e.g., polymorphs) the
`improved solubility of higher energy structures can be reliably
`estimated from a lmowledge of the the.rmody~tamic properties
`
`of the differet~.t forms (2). This is most simply achieved when
`data for the melting pofi~t, heat of [’us~on. a~.d heat capaciU of
`each form are available re.g,, (3)). In many eases it is also
`possible to directly measure the improvemems in solubility
`and biopharm aceutical per~k~rmance R~r such metastable crystal
`systems (4,5). A consideration of the data in the literat~re
`indicates that improvements i~ solubifity resulfiug from the use
`of alternate c~7stal forms can be expected to be as high as two
`fold (see later [br detai~s), an~ increases in maximum human
`plasma concentrations of up to s~x fold may be achieved (4)
`The measurement a~d estimation of the solubility and
`bioavaflability improvements that can be attained by using an
`amorphous form of a drug presents a more ag~fificant challe~ge
`becat~se of the far from equilibrium nature of the amorphous
`state. Thermodynan~c predictions of solubility enhancements
`have not been widely reported because of the difficulties
`involved in accurately characterizi~g amorphous drugs in terms
`of equifibrium thermody~amnic properties. Similarly, the deter-
`ruination o f meaningflfi experimental solubilities for amorphous
`pharmaceutical materials has been found to be extremeb~ diffi-
`cult because of the tendency for stlch materials to rapidly revert
`to the crystal.line state upon exposure to small quantities of
`solvet~ta (e.g., water vapor). Several reports in the literature
`indicate that the solubility advantage lot amoi’pb.ous &ug ~Brms
`may be qmte sig~.~ificant, fl~r example, 1.4 fold for indome~2hacin
`(6), 2 fold for cefalexin (7), 2.5 fold for tetracycline (8), and
`apNoximatdy 10 fold for a macrolide antibiotic (9) and novobi-
`ocin acid (l 0). Notably alm.ost all workers cite s~guificant ex.per-
`imeatal difficulties during solubilfi:y measurements due to
`crystalfization of the amorphous d~tg, and thus their reported
`experimental solubility catios are probably tmderestimates of
`the tree values for these materials. Otfly a few pbarmacokinefic
`investigations have beeu repo~ed (in animNs) (e.g, (l 1)), h.ow-
`ever these indicate that one should expec~ qui~e large improve-
`meres in the Nopharmaceuticai performance of amorphous
`drugs,
`In summary, m contrast to polymorphic crystalline drug
`forms, a simple method to estimate the theoretical maximum
`solubility of amorphous pharmaceuticals has not yet been pro-
`posed, nor has a consistent accurate method for assessing their
`apparent equilibrium solubilities been reposed. Thus, the objec-
`tive of the work reported herein was to use a simple thermody-
`uanlic approach to estimate the theoreticd maximum-sok~bil.ity
`improvement that can be achieved using m~mrphous compounds
`and to compare the resulting values with convenfionNly mea-
`sured solubility data. 11 was hoped that this approach would
`provide an estimate of the increased driving ff)rce for flee disso-
`lution of amorphous drug forms and indicate its relation to
`experimentally determined solubility values. To achieve this
`o~jective the thermal properties of several drags were measured
`using 4iff)renfial scanning cNorimelxy for use in the sok~bility
`calculations. Experimental solubility values were measured
`directly ~md/or collated ti’om the literature and then compared
`to (~e predicted values.
`
`Merck Frosst C~mada & Co., Kirk!and. Quebec, Canada
`Present address: Pfizer Inc. Gro~on. Connecticut 66340.
`To whom correspondence should be addressed.
`bruno, chaacock@groton,pfizer.com~,
`
`MATERIALS AND METHODS
`
`Materials
`
`(e-mail:
`
`lndomethacin, a hydrophobic poorly water soluble drug,
`was chosen for detailed cha~:acterization a~d study, Several
`
`397
`
`072&8741/00/0400-0397518,0010 @ 2000 PLenum Publisifing (?(u potation
`
`Lupin Ex. 1042 (Page 3 of 10)
`
`

`

`398
`
`Hancock and Parks
`
`"~\ glass
`
`.......... \ Tg
`orysf~l II .............
`
`liquid
`
`liquid
`
`Temperature
`Fig. 1. Schematic free-energy diagram for amorphous and crystalline
`materials (see text for explanation of abbreviations).
`
`AS~ = AH~/T}
`
`(5)
`
`This simple approach treats the amorphous form as a pseudo-
`equilibrium solid state at all temperatures below the melting
`point, and it is analogous to that which has been successfully
`used to estimate the relative solubilities of different crystalline
`polymorphs (2,14). In such instances the heat capacity differ-
`once between the two forms (ACp) is usually assumed to be
`constant, and has often been approximated by ACp ~o,~ 0 or
`ACv ~ ASf when experimental heat capacity data at the tempera-
`tures of interest are not available (15-17). In this study actual
`data for the heat capacity of the amorphous and crystalline
`forms of indomethacin (18) were used for the calculations, and
`comparisons were then made with results attained using the
`commonly applied approximations. Heat capacity differences
`between the glassy and equilibrium supercooled liquid forms
`measured at the glass transition (ACp@ were also available for
`each of the materials studied and were used foi" so,he of the
`solubility predictions.
`
`Solubility Measurements
`
`other drugs (i.e., glibenclamide, griseofulvin, hydrochlorthia-
`zide, polythiazide)were studied in less detail. All compounds
`were obtained in their thermodynamically most stable crystal-
`line form from Sigma Chemical Co., St. Louis, MO. The metas-
`table c~-polymorph of indomethacin was prepared by
`precipitation from a saturated methanol solution with water. The
`amorphous form of each compound was produced by quench
`cooling molten material in liquid nitrogen. The identity of the
`different drug forms was established using differential scanning
`calorimetry and powder X-ray diffraction experiments (see
`below). All solid samples were stored in a dry environment
`(over silica gel) and were presented for analysis as powders of
`less than 120 US mesh size (- 125 Ixm).
`
`Thermal Analysis "
`
`Powder samples of 5-10 mg were analyzed by differential
`scanning calorimetry (DSC) and thermogravimetric analysis
`(TGA) using a Seiko-220 thermal analysis system (Haake, Para-
`mus, NJ). Both TGA and DSC experiments were performed in
`a dry nitrogen atmosphere (60-100 ml/minute), heating the
`samples at a rate of 10°C/minute from ambient temperature to
`above their melting point(s). Calibration of the instruments with
`respect to temperature and/or enthalpy was achieved using high
`purity standards of indium, tin and gallium. Sample pans were
`made of alodined aluminum and were used with a vented cover:
`The mean results of triplicate determinations are reported,
`
`Powder X-ray Diffraction
`
`Powder x-ray diffraction measurements were used to con-
`firm the crystalline or amorphous nature of the starting materials
`and to identify the solids remaining in suspension at the end
`of the solubility experiments. A Scintag XDS-2000 instrument
`(Scintag, Cupertino, CA) with a nickel filtered copper radiation
`source was used and scans were taken between 2° and 70° 20.
`Samples were presented as lightly compacted powder disks.
`
`Solubility Predictions
`
`Predictions of the relative solubilities of the various crys-
`talline and amorphous forms of each drug were performed
`according to the method of Parks and co-workers (12,13). In
`this method the solubility ratio (cra/~r°) of the two forms (amor-
`phous = a; crystalline = c) being examined at any given temper-
`ature (T) is considered to be directly related to the free energy
`difference (AG) between those two forms (Fig. 1):
`
`AG~° = - R T in (o’~}/o-~-)
`
`(1)
`
`where R is the gas constant. The difference in free energy is
`estimated from the entropy (S) and enthalpy (H) differences
`between the two forms:
`
`AG~° = AH~° - (T AS~°)
`
`(2)
`
`Solubility measurements for indomethacin in deionized
`water were made using a closed, flat-bottomed, water-j acketed,
`glass vessel (70 mm height × 70 mm diameter) with an over-
`head 3-blade propeller stirrer operating at ~300 rpm. After
`equilibration at the desired temperature an excess of powdered
`drug was placed in the empty vessel, the stirrer started, and
`then two hundred milliliters of water were added to the vessel.
`At regular intervals a sample (~, 15 1TI1) of the liquid phase was
`withdrawn through a 0.22 ~m. filter and replaced with deionized
`water of the same temperature. Following dilution with a stan-
`dard solution of indomethacin in 50:50 methanol/water, the
`concentration of indomethacin in each sample was determined
`by UV-visible spectrometry at wavelengths of 266 and 318 nm.
`Solubility versus time profiles (over a 120 minute period) were
`determined at least four times for each form of the drug and
`at three different temperatures (5°C, 25°C, 45°C). The coeffi-
`cient of variation for replicate determinations was approxi-
`(4) mately five percent and rnean values are reported.
`
`and these enthalpy and entropy differences are calculated from
`the melting points (T}), enthalpy and entropy of fusion (AH~ &
`AS~), and isobaric heat capacities (C~, C~) as follows:
`
`AH~c = AH~ - (C~ - C~)(T} - T)
`
`AS~° = AS~ - (C; - C;)(ln (T}f/T))
`
`(3)
`
`Lupin Ex. 1042 (Page 4 of 10)
`
`

`

`Solubility Advantage for Amorphous Pharmaceuticals 399 ’
`
`Table 1, Therma! Properties of Different Forms of Indomethacin Mea-
`sured by Differential Scanning Calorimetry
`
`Tg
`(°C)
`
`--
`--
`42
`
`ACpTg
`(J/gK)
`
`--
`--
`0.41
`
`]~f
`(°C)
`
`162
`156
`--
`
`(Jig)
`
`102
`101
`--
`
`Form
`
`2t_Crystal
`~.Crystal
`Amorphous
`
`RESULTS
`
`Characterization of Raw Materials
`
`Tl~e experimentally determined thermal properties of the
`different forms of indomethacin are summarized in Table 1,
`and these results are in close agreement with those previously
`reported (19,20). The two polymorphic crystal forms differed in
`their melting point by approximately 6°C and were energetically
`very similar. The amorphous form was a glass at room tempera-
`ture and required moderate heating (to above 42°C) to attain
`the equilibrium supercooled liquid state. The identity of the
`various indomethacin forms was confirmed using X-ray powder
`diffraction experiments and comparison to reference data (19).
`The t~hErmal properties of the other drags studied were taken
`from the literature (3-5,12,13, 21) or measured by DSC. These
`results are presented in the footnote to Table 2.
`
`Solubility Predictions
`
`The predicted solubility ratios for the amorphous and ~-
`crystal forms of indomethacin relative to the ,/-crystal form are
`
`summarized in Table 2. A detailed analysis of these predictions
`will be included in the discussion section. The solubility ratios
`calculated for the other drugs considered are also summarized
`in Table 2. The magnitude of the predicted solubility advantage
`for different crystalline polymorphs ranged from 1.1 to 3.6 fold,
`whereas the predicted solubility ratio for the amorphous drug
`forms varied between 12 and 1652 fold.
`
`Solubility Measurements
`
`The experimentally determined solubility versus time pro-
`files for the various indomethacin forms are_shown in Figs. 2a,
`2b and 2c. At 5°C the enhanced solubility of the amorphous
`form relative to the ~/-crystal is clearly seen. A maximum solu-
`bility for the amorphous form occurred at approximately 10
`minutes and the solubility of the ,/-crystal form reached a
`constant value at approximately the same time in the experi-
`ment. At 25°C the m,’iximum in the solubility versus time profile
`for the amorphous form was more pronounced. The peak solu-
`bility occurred within the first 10 minutes of the experiment
`and the solubility of the amorphous form was consistently
`greater than that of the ~/-crystal form. At 45°C the peak solubil-
`ity for the amorphous form occurred very rapidly and declined
`equally quicldy. The (x-crystal polymorphic form also had a
`modestly improved solubility relative to the ~/-crystal form at
`45°C, The maximum solubility ratios attained at each tempera-
`ture for the indomethacin forms are smnmarized in Table 3,
`along with selected data for other drugs which have been
`reported in the literature. These literature data were chosen
`based on their apparent reliability and the possibility of being
`able to compare them with predicted values (i. e., both thermody-
`namic and solubility data were available). The experimental
`
`Table 2, Predicted Solubility Ratios for Indomethacin and Other Drug Compounds
`
`Compound
`
`Forms
`
`Solubility ratioa
`
`Co~mnent
`
`This work:
`Indomethacin
`Indomethacin
`
`Literature:
`Carbmnezapine (3)
`
`Chloramphenicol palmitate (4)
`Iopanoie acid (21)
`Mefenamic acid (5)
`Glibenclamide~
`Glucose (12,I3)
`Griseofulvinc
`Hydrochlorthiazide~/
`Iopanoic acid (21)
`Polythiazide~
`
`ot-crystal/~-crystal
`alnorphous/2t-crystal
`
`Ill-crystal/I-crystal
`
`A-crystal/B-erystal
`H-crystal/I-crystal
`I-crystal/H-crystal
`amorphous/crystal
`amorphous/crystal
`amorphous/crystal
`amorphous/crystal
`amorphous/I-crystal
`amorphous/crystal
`
`1.1- 1.2
`38 - 301
`25 - 104
`!6 ~ 41
`
`1.7 - 2.1
`1.7 - 2.0
`1.6 - 2.0
`1..6- 1.9
`1~6- 1.8
`1.5 - 1.7
`3.6
`2.3 - 2.8
`1.5
`112- 1652
`16 - 53
`38.- 441
`21 - 113
`12- 19
`48 - 455
`
`The range of values reflects the use of different ACp values for the calcnlations (see text for details).
`Glibenclamide: Tg = 58°C, ACpTg = 0.45 J/g/K, T~---- 177°C, AHf = 108 J/g,
`Griseofnlvin: Tg -- 91°C, ACpxg = 0.36 J/g/K, Tf = 221°C, AH~ = 107 J/g.
`Itydrochlorthiazide: Tg = 112°C, ACpTg = 0.31 Jig/K, Tr = 274°C, AH~ = 104 J/g.
`Polythiazide: Tg = 73°C, ACp.vg = 0.34J/g/K, Te = 220°C, AHf = 97 J/g.
`
`45°C
`5°C
`25°C
`45°C
`
`2°C
`12°C
`17°C
`26°C
`40°C
`58°C
`30°C
`37°C
`30°C
`23°C
`20°C
`21°C
`37°C
`37°C
`37°C
`
`Lupin Ex. 1042 (Page 5 of 10)
`
`

`

`400
`
`A
`
`3.0 ..... ¯
`
`DISCUSSION
`
`5°C
`
`Predicted Solubilities
`
`Hancock and Parks
`
`~ .0.5
`
`y-crystal
`~_ =~-~= = ~
`
`B
`
`0.0
`
`3,0
`
`~ 2.5
`~2,0
`
`C
`
`0.0
`
`3.0
`
`2.5
`
`1.5
`
`1.o
`
`0.0
`
`25%
`
`"t-crystal
`
`amorphous
`
`0 20 40 60 80 100 120
`Time (minutes)
`Fig. 2. Experimental aqueous solubility profiles for amorphous and
`crystalline indomethacin (. amorphous; ¯ ,/-crystal; ¯ a-crystal) (A)
`at 5°C.(B) at 25°C (C) at 45°C.
`
`solubility ratios varied between 1.1 and 4.0 for the crystalline
`polymorphs and 1.1 and 24 for the amorphous forms.
`
`Characterization of Undissolved Material
`
`Immediately after each experimental solubility determina-
`tion with the different forms of indomethacin the solid material
`suspended in the dissolution medium was recovered. This was
`achieved by fiItration and vacuum drying, and the recovered
`material was then analyzed by DSC, TGA and powder X-ray
`diffraction. The identities of the solid materials were ascertained
`by comparison to reference data (19) and are summarized in
`Table 4.
`
`The soh~bility ratios predicted for the amorphous indo-
`methacin from 0°C to the crystalline melting point using the
`different heat capacity approximations are shown in Fig. 3a.
`Because of the relatively large difference in free energy between
`the crystalline and amorphous forms of the drug the magnitude
`of the predicted solubility ratio is considerably higher than
`that recorded for the c~-crystal polymorph. The temperature
`dependence of the solubility ratio for the amorphous form is
`approximately logarithmic in all instances and the solubility
`ratio increases with decreasing temperature. Whereas for typical
`crystalline polymorphs the effects of changes in ACp are quite
`small (15-17) the effects on the predictions for the amorphous
`form are quite large. Comparison of the predictions for the
`amorphous indomethacin made using the approximations
`AC~’c ~-" 0, AC~’c ~ ~Cptg or AC~ic ~ AS~ with that made using
`experimentally measured heat capacity data reveals some inter-
`esting features. For the experimental heat capacity data the
`prediction follows the case where AC~’° ~ ACpTg above Tg. At
`Tg, where the heat capacity changes, there is a stepwise change
`in the predicted solubility ratio and the predicted values below
`Tg follow the case where AC~’e ~-- 0. The case where AC~’e ~
`AS~ provides an intermediate estimate of the solubility ratio
`at all temperatures. It can be clearly seen from this anatysis that
`there should be a significant stepwise increase in the solubility
`advantage at the glass transition temperature and the most
`appropriate ACp approximation is different above and below
`Tg. These features were not noted by Parks and co-workers in
`their studies of the solubility of amorphous glucose because
`all of their work was performed at temperatures above the glass
`transition temperature (12). The significance of these findings
`to the behavior of pharmaceutical dosage forms containing
`amorphous drug forms with glass transition temperatures which
`are close to physiological and/or ambient temperatures (e.g.,
`indomethacin) is obvious.
`
`Experimental Solubilities
`
`The experimentally measured solubilities of amorphous
`indomethacin were consistently greater than those of the ~!-
`crystalline form over a 40°C range in temperature and for a
`period of at least two hours following the start of the solubility
`determinations (Figs. 2a, 2b & 2c). The peak solubility for
`the amorphous indomethacin always occurred within the first
`10-15 minutes of the experiment and was as much as two fold
`greater than the ’steady-state’ solubility achieved at the end of
`the experiments. It appeared that the higher the temperature
`the more pronounced was the peak in the solubility versus time
`profile. Over the duration of the solubility experiments the
`amorphous starting material partially converted to the two most
`common crystalline polymorphs (Table 4). This change in phase
`may have been mediated by dissolution in and supersaturation
`of the aqueous dissolution media. Alternatively, it may only
`have been necessary to expose the solid amorphous indometha-
`cin ~o the solvent molecules in order to trigger crystallization
`in the solid state (22). Whatever the means of interconversion,
`it is’ cleat" that initially the amorphous drug was very highly
`soluble in the aqueous dissolution medium but that the maxi-
`mum level of supersaturation relative to the solubility of the
`
`Lupin Ex. 1042 (Page 6 of 10)
`
`

`

`Solubility Advantage for Amorphous Pharmaceuticals
`
`401
`
`Table 3. Experhnental Solubility Ratios for Indomethacin and Other Drug Compounds
`
`Compbund
`
`Forms
`
`Solubility ratio
`
`Comments
`
`This work:
`Indomethacin
`Indomethacin
`
`Literature:
`Carbamezapine (3)
`
`Chloramphenicol pah-nitate (4)
`Iopanoic acid (21)
`Mefenamic acid (5)
`Glibenclamide (30)
`Glucose (12,13)
`
`Griseofulvin (29)
`Hydrochlorthiazide (23)
`Iopanoic acid (21)
`Polythi_azide (23)
`
`a-crystal/y-crystal
`amorphous/y-crystal
`
`III-crysta!/I-cryStal
`
`A-crystal/B-crystal
`II-crystal/I- crystal
`I-crystal/H-crystal
`amorphous/crystal
`amorphous/crystal
`
`amorphous/crystal
`amorphous/crystal
`amorphous/I-crystal
`amorphous/crystal
`
`1.1
`4.4
`4.5
`2.8
`
`1.3
`1.4
`1.2
`1.2
`1.1
`1.1
`4.0
`1 ;6
`1.3
`14
`24
`21
`16
`1.4
`1.1
`3.7
`9.8
`
`45°C, water
`5°C, water
`25°C, water
`45°C, water
`
`2°C, 2-propanol
`12°C, 2-propanol
`17°C, 2-propanol
`26°C, 27propanoi
`40°C, 2-propanol
`58°C; 2-propanol
`30°C, 35% t-butanol (aq.)
`37°C, phosphate buffer (aq.)
`30°C, dodecyl alcohol
`23°C, buffer (aq.)
`20°C, methanol
`20°C, ethanol
`20°C, isopropyl alcohol .
`21°C, water
`37.°C, HC1 & PVP (aq.)
`37°C~ phosphate buffer (aq.)
`37°C, HC1 & PVP (aq.)
`
`~/-custal form could not be sustained. Thus the measured solu-
`bility of the amorphous drug declined to a nearly constant level
`within a period of about 20~60 minutes. Similar dissolution
`ldnetics have been previously reported for many metastable
`drug and excipient forms and they were not unexpected in this
`study. The correlation of the transformation kinetics in-vitro to
`the in-vivo situation is unknown, however it can be appreciated
`that it is likely that some level of in-vivo solubility enhancement
`would be achieved with the amorphous form of indometha-
`cin (11).
`The range of temperatures that was selected for the experi-
`menta[ solubility determinations is typical of that encountered
`by pharmaceutical products during their normal manufacture,
`packaging, storage, and use. Over this temperature range the
`expected increases in solubility with temperature were clearly
`seen for both the amorphous and ",/-crystal forms of indometha-
`cin (Figs. 2a, 2b, 2c; Table 3). Interestingly the steady-state
`solubility of the amorphous form at 5°C was approximately
`equal to that of the stable y-crystal at 450C. The alterations in
`solubility with temperature were of a similar magnitude for
`both forms and might have been amenable to a detailed thermo-
`dynamic analysis if data were available at a greater number
`