w1 PhIb7fl
`139»
`N0.6
`v.19
`9.01 —————— ——SEQ= PIUFIOOOU
`r1: PHARMACEUTICAL kfiavaacw
`
`03/21/So
`
`Mylan v. Qualicaps, |PR2017—OO203
`QUALICAPS EX. 2014 — 1/12
`
`Mylan v. Qualicaps, IPR2017-00203
`QUALICAPS EX. 2014 - 1/12
`
`

`

`EDlTOR—EUROl’E
`Bernard Testa, University of Lausanne, Lausanne, Switzerland
`
`EDITOR—JAPAN
`Yuichi Sugiyama, University of Tokyo, Tokyo, Japan
`
`ASSOCIATE EDITORS
`William E. Evans, St. Jude Children’s Research Hospital, Memphis,
`Tennessee
`Kinam Park, Purdue University, West Lafayette, Indiana
`Bonnie L. Svarstad, University of Wisconsin, Madison, Wisconsin
`
`ASSISTANT EDITOR—EUROPE
`Joachim M. II. Mayer, University of Lausanne, Lausanne, Switzerland
`
`_______________%____,____/,
`_
`PHARMACEUTICAL RESEARCH
`An Official Journal of the American Association of Pharmaceutical Scientists
`ch areas
`_
`covered in
`‘
`65' Rcsear
`medlCll'l’dl
`Pharmaceutical Research publishes innovative basic research and technological advances in the pharmaceutical biomedical scienc
`Ill
`the journal include: pharmaceutics and drug delivery, pharmacokinetics and pharmacodynamics, drug metabolism pharmacology and toxicology:
`1
`. Veg
`'
`‘ca in
`-
`d cli
`chemistry, natural products chemistry, analytical chemistry, chemical kinetics and drug stability, biotechnology, pharmaceutical technology, an
`tigations, as well as articles on the social, economic, or management aspects of the pharmaccufica] Sciences
`Ken-ichi Inui, Kyoto University Hospital, Kyoto, Japan
`EDITOR-IN-CHIEF
`1.
`Myron K. Jacobson, University of Kentucky, Lexington, Kfimucky
`Rudy 14- Jllliano, University of North Carolina, Chapel Hill, North Caro ma
`Vincent H. L. Lee, Department of Pharmaceutical Sciences, University of
`Southern California, Los Angeles, California
`Hans E. Juiiginger, University of Leiden, Leiden, The Netherlands
`William J. Jusko, SUNYB School of Pharmacy, Amherst, New Y0
`.
`,
`Tetsuya Kamataki, Hokkaido University, Sapporo, Japan
`Neil Kaplowitz, University of Southern California, Los Angeles, Callfomla
`Ian W. Kellaway, Welsh School of Pharmacy, Cardiff, Wales
`Sung Wan Kim, University of Utah, Salt Lake City,
`Thomas Kisscl, University of Marburg, Marburg, Germany
`.
`Joachim Kohn, Rutgers University, Picataway, New Jersey
`Peter A. Kollman, University of California, San Francisco, California
`Jindrich Kopccek, University of Utah, Salt Lake City, Utah
`Thomas M. Lutlden, University of Nebraska Medical Center, Omaha,
`Nebraska
`Susan M. Lunte, University of Kansas, Lawrence, Kansas
`G°l'd0ll McKay, University of Saskatchewan, Saskatchewan, Canada
`Hans P. Merkle, Swiss Federal Institute of Technology, Zurich,
`Switzerland
`Kllmill K Midha, University of Saskatchewan, Saskatchewan, Canada
`Duane D. Miller, University of Tennessee, Memphis, Tennessee
`Randall J. Mrsny, Genentech MS#6, South San Francisco, California
`Bernd W. Muller, Christian-Albrechts-Universitat, Kiel, Germany
`Sliozo Muranishi, Kyoto Pharmaceutical University, Kyoto, Japan
`Tsuneji Nagai, Hoshi University, Tokyo, Japan
`Masahiro Nakano, Kumamoto University Hospital, Kumamoto, Japan
`Teruo Okano, Tokyo Women’s Medical College, Tokyo, Japan
`Mlchaftl J- Pikal, Eli Lilly Company, Indianapolis, Indiana
`Vcrnoique Preat, Universite Catholique de Louvain, Bruxelles, Belgium
`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
`Malcolm Rowland, University of Manchester, Manchester, England
`Wolfgang Sadéc, University of California, San Francisco, California
`Tomi K. Sawyer, Parke-Davis/Warner-Lambert, Ann Arbor, Michigan
`Wei-Chiang Shen, University of Southern California, Los Angeles,
`California
`Valentino J. Stella, University of Kansas, Lawrence, Kansas
`Andy Stcrgachis, University of Washington, Seattle, Washington
`Eric Tomlinson, GeneMedicine, The Woodlands, Texas
`University, Kanazawa, Japan
`of Limburg, Maastricht, The Netherlands
`'ty of Minnesota, Minneapolis, Minnesota
`Timothy S. Weidmann,
`Robert J. Wills, R. W. Johnson Pharmaceutical Research Institute,
`Raritan, New Jersey
`Keiji Yamamoto, Chiba University, Chiba, Japan
`
`EDITORIAL ADVISORY BOARD
`Gordon L. Amidon, University of Michigan, Ann Arbor, Michigan
`Per Artursson, Department of Pharmacy BMC, Uppsala, Sweden
`Jessie Lai-Sim Au, Ohio State University, Columbus, Ohio
`Shoji Awazu, Tokyo University of Pharmacy & Life Science, Tokyo, Japan
`Michael B. Bolger, University of Southern California, Los Angeles,
`California
`J. Lyle Bootman, University of Arizona, Tucson, Arizona
`Ronald T. Borchardt, University of Kansas, Lawrence, Kansas
`Ilarold G. Boxenbaum, Otsuka America Pharmaceutical, Rockville,
`Maryland
`D. Craig Brater, Indiana University, Indianapolis, Indiana
`Douwe D. Breimer, University of Leiden, Leiden, The Netherlands
`Alice Clark, University of Mississippi, University, Mississippi
`Patrick Couvreur, Universi
`d, Chatenay-Malabry, France
`Daan J. A. Crommelin, University
`ht, T116 Nlftlleflilllds
`Richard N. Dalby, UMAB School of Pharmacy, Baltimore, Maryland
`Stanley S. Davis, The University of Nottingham, Nottingham, England
`Jennifer B. Dressman, Johann Wolfgang Goethe-Universitat, Frankfurt,
`Germany
`Alexander T. Florence, University of London, London, England
`John G. Gambertoglio, University of California, San Francisco,
`California
`_
`Kathleen M. Giacomini, University of California at San Francisco,
`San Francisco, California
`Robert Gurny, Universite de Geneve, Geneve,'Switzerland
`_
`Jonathan lladgraft, University of Wales, Cardiff, Wales
`Abraham G. Hartzema, University of North Carolina, Chapel l-Iill,
`North Carolina
`Mitsuru Ilashida, Kyoto University, Kyoto, Japan
`Joel Ilay, University of Southern California, Los Angeles, California
`Susan Ilcrshenson, Amgen Inc., Thousand Oaks, California
`Brian B. Hoffman, VA Medical Center, Palo Alto, California
`Anton J. Hoplinger, University of Illinois, Chicago, Illinois
`Tatsuji lga, University of Tokyo Hospital, Tokyo, Japan
`
`INTERIM BOOK REVIEW EDITOR
`Kinam Park, Purdue University, School of Pharmacy, West Lafayette,
`Indiana 47097
`
`EDITORIAL ASSISTANTS
`Ruth Ellis-Ballard
`Elizahetli B. Gongora
`
`Phammccutical Research is published monthly by Plenum Publishing Cor
`Abstracts, Chemical Abstracts, Exccrpta Medica, Gazette dc l’AI’GI, In
`Plzqmzacculiczil Researcli participates in the Copyright Clearance Center (CCC) Transaction
`article in this journal indicates the copyright owner's consent that copies of the article may
`. The CCC is a nonpro
`the copier pay the flat flee of $9.50 per copy per article (no additional per-page fees) directly to the Copyright
`01923. for all copying not explicitly pc
`'
`107 or 103 of the U.S. Copy
`0724-8741/96$9.50
`'
`does not extend to other kinds of copying, such as copying for gencr
`libraries and other users registered with the CCC. There
`promotional purposes, for creating new collective works, or for resale, nor to the reprinting of figures, tables, and text excerpts.
`Advertising inquiries should be addressed to Edward J. McNeil and Rob Britt, McNeill Group, Inc., 301 Oxford Valley Road, Suite 80313, Yardley, Pennsylvania 19067-telephone (215)
`321-9662 and fax (215) 321-9636.
`bscription Department, Plenum Publishing Corporation, 233 Spring Street, New York, N.Y.
`10013 Of laxlld to the Subscription Department at its number (212) 807-1047, or may be telephoned to the Subscription Department's Journal Customer Service at (212)620-8468,
`Subscription inquiries and subscription orders should be addressed to the publisher at Su
`-8470, -8472, or -8082. Subscription rates:
`V°l“"‘° 13- 1995 (12 l55“°S) $575.00 (outside the U.S., $675.00). Prices for individual subscribers certifying that the journal is for their personal use, $129.00 (outside the U.S.,
`$151.00).
`S°c°“d‘°l355 P051326 paid at New York, N.Y., and at additional mailing offices. Postmaster: Send address changes to Plzamiacciltical Research, Plenum Publishing Corporation, 233
`Spring Street, New York, N.Y. 10013.
`Printed in the USA.
`
`Rosewood Drive, Danvers, Massachusetts
`photocopying fees by
`I
`r advertising or
`
`Mylan v. Qualicaps, |PR2017—OO203
`QUALICAPS EX. 2014 — 2/12
`
`Mylan v. Qualicaps, IPR2017-00203
`QUALICAPS EX. 2014 - 2/12
`
`

`

`PHARMACEUTICAL RESEARCH
`An Official Journal of the American Association of Pharmaceutical Scientists
`
`Volume 13 Number 6
`
`June 1996
`
`CONTENTS
`
`NEWS FROM THE NATIONAL INSTITUTE OF GENERAL MEDICAL SCIENCES (NIGMS)
`
`COMMENTARY
`
`Biomaterials Science at a Crossroads: Are Current Product Liability Laws in the United States Hampering
`Innovation and the Development of Safer Medical Implants?
`Joachim Kohn
`
`REVIEW
`
`Macromolecular Carrier Systems
`Biodistribution
`
`for Targeted Drug Delivery: Pharmacokinetic Considerations on
`
`Yoshinobu Takakura and Mitsuru Hashida
`
`REPORTS
`
`Modeling
`
`Population Pharmacokinetics of Terfenadine
`
`Richard L. Lalonde, Denis Lessard, and Jacques Gaudrealtlt
`
`Evaluation of Pharmacokinetic Studies: Is It Useful to Take into Account Concentrations Below the Limit
`of Quantification?
`
`Henri Humbert, Marie Daniele Cabiac, José Barradas, and Christopher Gerbeau
`
`High Variability in Drug Pharmacokinetics Complicates Determination of Bioequivalence: Experience
`with Verapamil
`
`Yll Chung Tsang, Radu Pop, Paul Gordon, John Hems, and Michael Spino
`
`Transdermal Drug Delivery
`
`Iontophoretic Delivery of a Telomeric Oligonucleotide
`Rhonda M. Brand and Patrick L. Iversen
`
`Analysis of Simultaneous Transport and Metabolism of Ethyl Nicotinate in Hairless Rat Skin
`
`855
`
`Kenji Stigibayashi, Teruaki Hayashi, Tomio Hatanaka, Masahiko Ogihara, and Yasunori Morimoto
`Mylan v. Qualicaps, |PR2017—OO203
`QUALICAPS EX. 2014 - 3/12
`
`Mylan v. Qualicaps, IPR2017-00203
`QUALICAPS EX. 2014 - 3/12
`
`

`

`
`
`Drug Targeting
`
`Targeted Delivery of Doxorubicin via Sterically Stabilized Immunoliposomes: Pharmacokinetics and
`Biodistribution in Tumor-bearing Mice
`
`Noam Emanuel, Eli Kedar, Elijah M. Bolotin, Nechama I. Smorodinsky, and Yechezkel Barenholz
`
`Erythrocytes as Carriers for Recombinant Human Erythropoietin
`M‘m”“‘]””“’c“l“d" Garmv R050‘/"1Wl)€l L(5P€Z» Silvia Stmz, Montserrat Pirzilla, and Jose’ Luque -
`
`Formulation of Polyiodinated Triglyceride Analogues in a Chylomicron Remnant—Like Liver—Selective
`Delivery Vehicle
`
`Marc A. Longino, Douglas A. Bakan, Jamey P. Weichert, and Raymond E_ Counsel]
`
`Development and Pharmacokinetics of Galactosylated Po1y—L—Glutamic Acid as a Biodegradable Carrier
`for Liver—Specific Drug Delivery
`
`Hideki Hirabayashi, Makiya Nishikawa, Yoshinobu Takakura, and Mitsuru Hashida
`
`Gastrointestinal Pharmacology
`
`Age—dependent Expression of P—Glycoprotein gpl70 in Caco-2 Cell Monolayers
`Ken-ichi Hosoya, Kwang-Jin Kim, and Vincent H. L. Lee
`
`In Vivo Assessment of Intestinal, Hepatic, and Pulmonary First Pass Metabolism of Propofol in the Rat
`Araz A. Raoofi Patrick F. Augustijns, and Roger K. Verheeck
`
`Enteral Absorption of Insulin in Rats from Mucoadhesive Chitosan—Coated Liposomes
`Hirofumi Takeuchi, Hiromitsu Yamamoto, Toshiyuki Niwa, Tomoaki Hino, and Yoshiaki Kawashima
`
`Contribution of Lysosomes to the Subcellular Distribution of Basic Drugs in the Rat Liver
`Jimko Ishizaki, Koichi Yokogawa, Masako Hirano, Emi Nakashima, Yoshimichi Sai, Shoji Ohkuma,
`Tohru Ohshima, and Fujio Ichimura
`
`Biophysics
`
`Solubilization of Retinoids by Bile Salt/Phospholipid Aggregates
`Ching—Yuan Li, Cheryl L. Zimmerman, and Timothy S. Wiedmann
`
`Permeation Behavior of Salbutamol Sulfate Through Hydrophilic and Hydrophobic Membranes Embedded
`by Thermo-responsive Cholesteryl Oleyl Carbonate
`Shan—Yang Lin, Yih-Yih Lin, and K0-Shao Chen
`
`The Effect of Hydrophobic Character of Drugs and Helix—Coil Transition of K-Carrageenan on the
`Polyelectrolyte—drug Interaction
`
`Ninus Caram—Lelham and Lars—0lof Sundelof
`
`Determination of Molecular Mobility of Lyophilized Bovine Serum Albumin and y-Globulin by Solid-State
`‘H NMR and Relation to Aggregation—Susceptibility
`
`Sumie Yoshioka, Yukio A30, and Shigeo Kojima
`
`Others
`
`Effects of Polyaminocarboxylate Metal Chelators on Iron-thiolate Induced Oxidation of Methionine— and
`Histidine-Containing Peptides
`
`931
`
`Fang Zhao, Jian Yang, and Christian Sch(')'neich
`This material was -to-piad
`an;h.er~tLMan.d maybe
`SL2 Eject US {>::-ps~jright Laws
`
`_
`Mylan V. Quallcaps, IPRZO17-00203
`A
`A I
`A
`A
`‘ -
`
`Mylan v. Qualicaps, IPR2017-00203
`QUALICAPS EX. 2014 - 4/12
`
`

`

`CONTENTS (Continued)
`
`Synthesis and Anti—Pr0liferative Activity of 2-Hydroxy—l,2—dihydroacronycine glycosides
`
`Sofia Mitaku, Alexios-Léandros Skaltsounis, Frangois Tillequin, Michel Koch, Yves Rolland, Alian
`Pierre’, and Ghanem Atassi
`
`Investigations into the Relationship Between Drug Properties, Filling, and the Release of Drugs from Hard
`Gelatin Capsules Using Multivariate Statistical Analysis
`James Hogan, Pei—Inn Shite, Fridrun Podczeck, and J. Michael Newton
`
`TECHNICAL NOTES
`
`Effect of the Spermicide, Nonoxynol 9, on Vaginal Permeability in Normal and Ovariectomized Rabbits
`
`Fiisim Acatiirk and Joseph R. Robinson
`
`Primary Cell Culture of the Rabbit Choroid Plexus: An Experimental System to Investigate Membrane
`Transport
`
`Vikram K. Ramanathan, Andrew C. Hui, Claire M. Brett, and Kathleen M. Giacomini
`
`The Effect of Iontophoresis on Skin Barrier Integrity: Non—invasive Evaluation by Impedance Spectroscopy
`and Transepidermal Water Loss
`Yogeshvar N. Kalia, Lourdes B. Nonato, and Richard H. Guy
`
`ERRATA
`
`Mylan v. Qualicaps, |PR2017—OO203
`QUALICAPS EX. 2014 — 5/12
`
`Mylan v. Qualicaps, IPR2017-00203
`QUALICAPS EX. 2014 - 5/12
`
`

`

`The New and Advanced
`VK 7000 Dissolution
`Testing Station
`The new VK 7000 both anticipates
`and exceeds your own and USP
`dissolution test requirements.
`Because capability is built-in, not
`added on, refinements such as
`one—touch operation, se|f—diagnos-
`tics, digital communications. and
`programmability are now available
`in a single stand—alone system!
`From every standpoint —— perfor-
`mance, accuracy or operator
`ergonomics — the new VK 7000
`simply defies comparison with all
`that ever was and is in dissolution
`testing. We invite you to see the
`VK 7000 now and prove this to
`your own satisfaction. Available in
`multiple formats, from 6 to 16
`spindles, in both Benchsaver and
`Easy Access models to preserve
`benchtop space and provide you
`with unimpeded access to all vessels.
`- Automation Ready.
`- Precision Heater/Circulator.
`- Leakproof Molded Waterbath.
`- Hard Copy Documentation.
`- Menu Driven Software.
`
`36 Meridian Road, Edison, NJ 08820
`Phone (800) 229-1108 - (908) 548-3616
`FAX (908) 548-9396
`VanKeI International Ltd. Hornchurch,
`Essex, England - (0708) 476162
`_ F fiS’fi‘8:8éf>3s7,2lPR2o17—oo2o3
`
`Mylan v. Qualicaps, IPR2017-00203
`QUALICAPS EX. 2014 - 6/12
`
`

`

`Pharmaceutical Research, Vol. 13, No. 6, 1996
`
`Report
`
`Investigations into the Relationship
`Between Drug Properties, Filling,
`and the Release of Drugs from Hard
`Gelatin Capsules Using Multivariate
`Statistical Analysis
`
`James Hogan,‘ Pei-Inn Shue,‘ Fridrun Podczeck,‘
`and J. Michael Newton”
`
`Received December 5, 1995; accepted March 4, 1996
`
`Purpose. The aim of the present work is to identify complex relation-
`ships between fonnulation variables and dosage form properties to aid
`the development of hard gelatin capsules.
`Methods. Multivariate statistical analysis was employed based on a
`statistical design, which considered drug solubility, particle size and
`concentration, type and concentration of filler and disintegrant, and
`concentration of standard lubricant and glidant as the main influence
`factors. Both the filling properties of the formulations and the
`disintegration/dissolution properties of the capsule content were
`studied.
`Results. From the two multivariate statistical methods used, nonpara-
`metric canonical analysis proved to be the superior method to deal
`with the complex information included in the data. While the filling
`performance of the formulation could clearly be attributed to the formu-
`lation variables such as drug particle size, type of filler, concentration
`of drug and glidant, the disintegration of the capsules and the dissolution
`of the drugs was not strongly related to the formulation variables
`chosen. In this respect as a trend, the drug solubility, and the type of
`disintegrant and filler appear to be more important factors.
`Conclusions. Based on an appropriate number of experiments, organi-
`sed in a statistical design, nonparametric canonical analysis can be
`used to identify relationships in a set of data that is grouped in influence
`and response variables to aid the development of a dosage form.
`
`KEY WORDS: hard gelatin capsule formulation; multivariate statisti-
`cal analysis; parametric and nonparametric canonical analysis; statistical
`design.
`
`INTRODUCTION
`
`The presentation of drugs in hard gelatin capsules as an
`oral dosage form has an historical background dating back to
`1834 (1). Currently, their output continues to increase and the
`number of formulations listed in, for example, Physicians’ Desk
`Reference (2), is 126. The basis of the formulation of powder-
`filled hard gelatin capsules is discussed by Cole (3). The objec-
`tive of formulations is to ensure that each capsule provides the
`dose of drug required by Pharrnacopoeial standards and that
`the drug should be released from the capsule to ensure drug
`bioavailability. The choice of type and quantity of ingredients
`to be incorporated to assist the formulation in terms of diluents,
`disintegrants, glidants, lubricants and wetting agents is part of
`the process of formulation and depends on the dose of drug
`
`' Department of Pharmaceutics, The School of Pharmacy, University
`of London, 29-39 Brunswick Square, London WCIN 1AX, England.
`2 To whom correspondence should be addressed.
`
`and the physical and chemical properties of the drug. Just how
`the drug properties are related to the formulations is not known.
`Hence an investigation into this relationship could be a valuable
`aid to capsule formulation.
`
`EXPERIMENTAL
`
`Experimental Design
`
`To relate drug properties to capsule performance is a com-
`plex task, hence there is a need for statistical design, which is
`appropriate for the use of multivariate statistical methods.
`Five drugs were chosen according to their solubility, which
`covers a range between 0.2 g1" and 200 gl" giving a factor
`of 3 on a logarithmic scale. The drugs are phenytoin (0.2gl"),‘~
`theophylline (8.0gl“), paracetamol (15.0gl‘), propranolol—HCl
`(5O.0gl"1) and aminophylline (200.0gl" 1). To describe the drug,
`if a relationship to the filling performance of the capsules is
`the target, the mean particle size has been determined, which
`was 26 mm for paracetamol and aminophylline, 57 pm for
`theophylline, 65 um for phenytoin and 122 um for proprano-
`lol-4HCl.
`
`Five fillers have been chosen for their relative solubility,
`which apparently increases in the following order: calcium
`phosphate < microcrystalline cellulose < maize starch < starch
`1500 < lactose monohydrate. Five disintegrants have been
`chosen randomly, and the swelling ability in water (22°C) has
`been measured as described by Podczeck and Révész (4). The
`disintegrants were ranked according to their relative swelling
`volume: Explotab (1680%) > AcDiSol (600%) > Amberlite
`(190%) > Polyplasdone XL (l50%)> maize starch (110%).
`In all cases, magnesium stearate was used as a lubricant, and
`Aerosil was incorporated as a glidant. In both cases, levels of
`0.0, 0.5, 1.0, 1.5 and 2.0% w/w have been used, and the midpoint
`of the experimental design was set to 1.0% in both cases. In
`the case of magnesium stearate,
`this is the widely accepted
`optimal
`lubricant concentration, whereas for Aerosil 0.5%
`appears the more usual concentration (5). However, from tab-
`letting it is known that the optimum concentration of Aerosil
`can vary between 0.2 and 2.0% depending on the formulation
`property of main concern. For example, with respect to a rapid
`dissolution rate 1.0% Aerosol is optimal (6), whereas 0.5%
`only is insufficient (7). At the extreme, 2.0% Aerosil has been
`shown to be optimal for a satisfactory filling and necessary
`compact strength (8,9). Finally, the optimal Aerosil concentra-
`tion has been reported to depend on the magensium stearate
`concentration and the way to incorporate both components into
`the powder mixture. Based on a statistical design, Staniforth
`et al. (10) found that at 1.0% magnesium stearate the coefficent
`of fill weight variation decreased with increased Aerosil concen-
`tration between 0.5 and 2.0%, again indicating that the Aerosil
`optimum might be above the commonly used 0.5%. Thus the
`use of the five levels of Aerosil in the experimental design for
`the current paper will be able to clarify this point, because both
`0.5% and 1.0% Aerosil are included in the design. Recently,
`Jones (11) published a survey of excipients used in capsule
`formulation, based on the marketed formulations in France,
`Germany and Italy. Quantitative information about excipients
`used in Italy revealed that the most commonly used Aerosil
`concentration in powder filled hard gelatin capsules is 1.5%,
`
`0724-874l/96/0600-0944$09.50/0 © I996 Plenum Publishing Corporation
`
`944
`
`Mylan v. Qualicaps, |PR2017—OO203
`QUALICAPS EX. 2014 — 7/12
`
`Mylan v. Qualicaps, IPR2017-00203
`QUALICAPS EX. 2014 - 7/12
`
`

`

`Multivariate Statistical Analysis and Capsule Formulation
`
`and that more than 75% of the formulations contain more
`
`than 0.9% Aerosil. Thus, the value of 0.5% (5) appears even
`more doubtful.
`Due to the nature of the data material, i.e., several influence
`factors and a variety of response variables, a multivariate analy-
`sis is required to identify relationships between these two groups
`of variables. The response variables (Y) are all of nominal
`(= numerical) nature, whereas the independent variables are
`nominal or ordinal depending on whether a rank number (ordi-
`nal) or an underlying variable (particle size, swelling volume)
`has been used to describe them. Hence, both parametric and
`nonparametric test procedures can be used. Such a parametric
`test procedure is the canonical analysis introduced by Hotelling
`(12). This method has been used for pharmaceutical problems,
`e.g., by Podczeck et al (13) and by Bohidar and Bohidar (14).
`The drug, disintegrant type and the concentrations of the excipi-
`ents used are described by their physical properties, but for the
`filler type a dummy variable has to be used. Table I shows the
`variable group X used in this kind of analysis. All formulation
`properties (response variables, Y) are used as their original
`
`Table 1. Variable Group X Used in the Classical (Parametric)
`Canonical Analysis
`
`EN D,(ps)
`
`d2(sol)
`
`ft
`
`fl
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`27
`28
`29
`30
`31
`32
`33
`
`26.0
`26.0
`26.0
`26.0
`26.0
`26.0
`26.0
`26.0
`26.0
`26.0
`26.0
`26.0
`65.0
`57.0
`26.0
`122.0"
`26.0
`26.0
`26.0
`26.0
`26.0
`26.0
`26.0
`26.0
`26.0
`26.0
`26.0
`26.0
`26.0
`26.0
`26.0
`26.0
`26.0
`
`15.0
`15.0
`15.0
`15.0
`15.0
`15.0
`15.0
`15.0
`15.0
`15.0
`15.0
`15.0
`0.2
`8.0
`15.0
`50.0
`200.0
`15.0
`15.0
`15.0
`15.0
`15.0
`15.0
`15.0
`15.0
`15.0
`15.0
`15.0
`15.0
`15.0
`15.0
`15.0
`15.0
`
`2
`2
`2
`2
`2
`2
`2
`2
`2
`2
`2
`2
`2
`2
`2
`2
`2
`2
`2
`2
`2
`1
`3
`4
`5
`2
`2
`2
`2
`1
`1
`5
`5
`
`44.0
`43.5
`42.5
`42.0
`44.0
`43.5
`42.5
`42.0
`48.0
`45.5
`40.5
`38.0
`43.0
`43.0
`43.0
`43.0
`43.0
`73.0
`58.0
`28.0
`13.0
`43.0
`43.0
`43.0
`43.0
`43.0
`43.0
`43.0
`43.0
`48.0
`38.0
`48.0
`38.0
`
`dt
`
`1680
`1680
`1680
`1680
`1680
`1680
`1680
`1680
`1680
`1680
`1680
`1680
`1680
`1680
`1680
`1680
`1680
`1680
`1680
`1680
`1680
`1680
`1680
`1680
`1680
`600
`190
`150
`110
`1680
`1680
`1680
`1680
`
`d1
`
`5.0
`5.0
`5.0 .
`5.0
`5.0
`5.0
`5.0
`5.0
`0.0
`2.5
`7.5
`10.0
`5.0
`5.0
`5.0
`5.0
`5.0
`5.0
`5.0
`5.0
`5.0
`5.0
`5.0
`5.0
`5.0
`5.0
`5.0
`5.0
`5.0
`0.0
`10.0
`0.0
`10.0
`
`11
`
`1.0
`1.0
`1.0
`1.0
`0.0
`0.5
`1.5
`2.0
`1.0
`1.0
`1.0
`1.0
`1.0
`1.0
`1.0
`1.0
`1.0
`1.0
`1.0
`1.0
`1.0
`1.0
`1.0
`1.0
`1.0
`1.0
`1.0
`1.0
`1.0
`1.0
`1.0
`1.0
`1.0
`
`gl
`
`0.0
`0.5
`1.5
`2.0
`1.0
`1.0
`1.0
`1.0
`1.0
`1.0
`1.0
`1.0
`1.0
`1.0
`1.0
`1.0
`1.0
`1.0
`1.0
`1.0
`1.0
`1.0
`1.0
`1.0
`1.0
`1.0
`1.0
`1.0
`1.0
`1.0
`1.0
`1.0
`1.0
`
`dc
`
`50.0
`50.0
`50.0
`50.0
`50.0
`50.0
`50.0
`50.0
`50.0
`50.0
`50.0
`50.0
`50.0
`50.0
`50.0
`50.0
`50.0
`20.0
`35.0
`65.0
`80.0
`50.0
`50.0
`50.0
`50.0
`50.0
`50.0
`50.0
`50.0
`50.0
`50.0
`50.0
`50.0
`
`945
`
`values (arithmetic mean, compare Materials and Methods) and
`presented in Table 2. Nonparametric canonical analysis (15) is
`the equivalent type of multivariate procedure if ordinal variables
`are to be used. The advantage compared to the classical canoni-
`cal analysis is its acceptance of nonlinear relationships. How-
`ever, all variables have to be transferred into ordinal data, which
`appears as a loss in information especially in the group of the
`response variables. Table 3 shows the data matrix of X using
`the rank of the physical properties of the excipients as ordinal
`data. Table 4 shows the classification of ordinal data for Y used
`
`in the nonparametric procedure.
`
`MATERIALS
`
`The five drugs used were of EP quality: aminophylline
`(Knoll AG, Germany), theophylline (BP-Knoll AG, Germany),
`propranolol
`hydrochloride
`(Becpharm U.K.),
`phenytoin
`(Recordati, Italy) and paracetamol (Becpharm, U.K.). The fillers
`employed included lactose monohydrate (Dairy Crest U.K.),
`maize starch (Beehive Industries, Holland), microcrystalline
`cellulose (Avicel PH102, FMC, USA), Starch 1500 (Colorcon
`Ltd., U.K.) and calcium phosphate (East Anglia Chemicals,
`U.K.), and were of EP quality. The disintegrating agents were
`maize starch (Beehive Industries, The Netherlands) (E.P.), pol-
`yplasdone XL (GAF Corporation, U.K.), Amberlite (Sigma
`Chemical, U.S.A.), Explotab (Forum Chemicals Ltd., U.K.) and
`Ac-Di-Sol (FMC, U.S.A.). Magnesium stearate (British Drug
`Houses, U.K.) and Aerosil 200 (Degussa, Belgium) were uti-
`lised as the lubricant and glidant respectively.
`All materials were used as received from the suppliers
`except Aerosil 200 and Theophylline, which were sieved
`through a 60 and 100 mesh screen respectively to facilitate
`blending. Batches of 1 kg were prepared in a Y-cone blender
`(Erweka, AR400, Copley, U.K.) rotating for 20 minutes at
`approximately 56 rpm.
`
`METHODS
`
`The minimum bulk density of the various powders was
`determined in a 100 ml measuring cylinder, inverting the cylin-
`der l0 times before measuring the volume occupied by the
`powder. The maximum bulk density was determined in accor-
`dance with BS 1440, 1967. The values reported represent the
`mean of 5 determinations.
`
`Preparation of Capsules
`
`The powder mixtures were filled into size no. 0 hard-
`gelatin capsules using an automatic capsule filling machine
`(Zanasi AZ-5, Italy). The dosator height, compression force
`and powder bed height were adjusted by trial and error to give
`the maximum bulk density of the formulation. At the desired
`settings, the machine was initially run until the powder bed
`came to an equilibrium by visual inspection, before approxi-
`mately l00 capsules were collected from each run. These cap-
`sules were stored in tied polythene bags for further studies.
`The fill weight of 20 individual capsules was determined as
`required by BP.
`
`Note: EN, experiment number; D,(ps), drug characterised by particle
`size; d2(sol), drug characterised by solubility; ft, filler type; fl, filler
`level; dt, disintegrant type; dl, disintegrant level; 11, lubricant level; gl,
`glidant level; dc, drug concentration.
`
`Disintegration Test
`
`Capsule disintegration times were measured in 800 ml of
`distilled water at 37 t 1°C using “BP Disintegration Test for
`Hard-Gelatin Capsules.”
`
`Mylan v. Qualicaps, |PR2017—OO203
`QUALICAPS EX. 2014 — 8/12
`
`Mylan v. Qualicaps, IPR2017-00203
`QUALICAPS EX. 2014 - 8/12
`
`

`

`946
`
`Hogan, Shue, Podczeck, and Newton
`
`Table 2. Variable Group Y Used in the Classical (Parametric) Canonical Analysis
`
`Packing and filling performance
`
`Drug release
`
`Vmin
`Vmax
`Carr
`CFV
`AUC
`MDT
`VDT
`DT
`
`EN
`[gcm’3]
`[gcm‘3]
`H
`[%]
`[%]
`[%min]
`[min]
`[min]
`[min]
`
`10.5
`107.6
`28.6
`2480
`15.09
`38.79
`1.63
`0.82
`0.50
`1
`8.6
`23.1
`15.1
`1268
`1.19
`36.71
`1.58
`0.79
`0.50
`2
`8.2
`14.0
`12.5
`957
`2.40
`36.81
`1.58
`0.72
`0.46
`3
`7.5
`42.4
`19.0
`1907
`1.28
`36.76
`1.58
`0.68
`0.43
`4
`6.9
`54.9
`18.0
`1994
`2.51
`34.21
`1.52
`0.76
`0.50
`5
`7.2
`33.0
`17.4
`1733
`1.86
`30.56
`1.44
`0.72
`0.50
`6
`8.4
`49.5
`18.8
`1869
`1.36
`32.39
`1.48
`0.71
`0.48
`7
`10.4
`45.5
`17.3
`1775
`1.18
`32.86
`1.49
`0.70
`0.47
`8
`9.0
`16.3
`11.0
`1069
`0.95
`31.94
`1.47
`0.72
`0.49
`9
`8.2
`72.4
`20.8
`2087
`0.79
`33.11
`1.49
`0.74
`0.50
`10
`7.7
`111.4
`25.5
`2719
`1.44
`30.28
`1.43
`0.71
`0.50
`11
`7.9
`55.9
`19.7
`1834
`4.06
`34.90
`1.54
`0.74
`0.48
`12
`6.6
`200000.0
`2000.0
`200000
`1.40
`30.57
`1.44
`0.78
`0.54
`13
`8.2
`70.6
`19.1
`1980
`0.72
`30.19
`1.43
`0.80
`0.56
`14
`9.9
`7.8
`9.0
`706
`0.75
`33.78
`1.51
`0.74
`0.49
`15
`7.6
`128.4
`28.3
`2993
`1.90
`24.70
`1.33
`0.83
`0.62
`16
`11.5
`41.3
`22.2
`1571
`0.90
`33.75
`1.51
`0.80
`0.53
`17
`7.5
`31.2
`16.4
`1361
`0.80
`29.49
`1.42
`0.78
`0.55
`18
`7.7
`45.6
`16.9
`1593
`0.98
`31.79
`1.47
`0.76
`0.52
`19
`7.4
`98.7
`22.6
`2486
`2.38
`33.58
`1.51
`0.67
`0.44
`20
`11.4
`74.8
`18.8
`2032
`3.57
`34.68
`1.53
`0.62
`0.40
`21
`10.8
`68.1
`19.0
`2044
`0.85
`32.50
`1.48
`0.80
`0.54
`22
`7.4
`147.8
`28.1
`3081
`1.25
`37.24
`1.59
`0.72
`0.46
`23
`7.0
`27.3
`12.7
`1261
`1.64
`31.30
`1.46
`0.58
`0.40
`24
`7.6
`56.6
`19.9
`2012
`18.52
`39.84
`1.66
`0.64
`0.38
`25
`7.7
`39.6
`16.3
`1546
`0.84
`39.04
`1.64
`0.73
`0.44
`26
`9.3
`172.8
`29.1
`3356
`1.12
`36.11
`1.57
`0.72
`0.46
`27
`9.8
`229.3
`29.6
`3683
`1.10
`36.50
`1.57
`0.68
`0.44
`28
`7.6
`29.4
`14.5
`1336
`0.96
`36.99
`1.59
`0.73
`0.46
`29
`12.1
`840.5
`70.4
`7121
`2.42
`34.34
`1.52
`0.83
`0.54
`30
`8.9
`117.1
`25.0
`2798
`1.32
`34.16
`1.52
`0.80
`0.53
`31
`10.0
`1 17068.5
`760.1
`82319
`20.67
`35.25
`1.54
`0.61
`0.40
`32
`
`
`
`
`
`
`
`
`
`0.42 0.65 1.57 36.15 5.94 2467 25.1 86.333 7.3
`
`Note: EN, experiment number; Vmin, minimum bulk density; Vmu, maximum bulk density H, Hausner’s ratio; Carr, Carr’s compressibility
`index; CFV, coefficient of fill weight variation; AUC, area under the dissolution curve; MDT, mean dissolution time; VDT, variance of the
`dissolution time; DT, disintegration time.
`
`Dissolution Test
`
`The dissolution rates of the drugs from the various formula-
`tions were determined by means of the B.P. Apparatus II
`method. The paddles were rotated at 50rprn in 1000 ml of
`distilled water maintained at 37 t 0.6°C. Six capsules from
`each batch were evaluated simultaneously using an automated
`dissolution apparatus (Pharma Test, PTWS, Germany) con-
`nected to a sample collector (Pharma Test, Type PTFC I, Ger-
`many). Ten or more samples were extracted from the dissolution
`medium of each capsule throughout its period of dissolution.
`Each sample was diluted 25 times and analysed by a uv-vis
`spectrophotometer (Perkin-Elmer 554, USA). The absorbance
`of the solution of paracetamol, theophylline, aminophylline and
`propranolol was determined at 242 nm, 271 nm, and 288 nm
`respectively. The absorbance values were transformed to con-
`centrations by reference to standard calibration curves obtained
`experimentally. The solubility of phenytoin is too low to ensure
`sink condition, hence a low percentage release was achieved. To
`allow quantitative comparisons with the other drugs, arbitrarily
`
`assigned values indicating poor dissolution were given to this
`formulation. The dissolution profiles were characterised by the
`area under the curve (AUC), the mean dissolution time (MDT)
`and the variance of dissolution time (VDT) (16).
`
`RESULTS AND DISCUSSION
`
`First, parametric canonical analysis has been undertaken
`to describe the relationship between the excipients used in
`the formulations and'the filling performance of the capsules
`characterised by the powder densities, powder flow and coeffi-
`cient of fill weight variation. The mathematical outcome is
`summarised in Table 5. The relationship between filling perfor-
`mance and the formulation components is significant. However,
`with this method only 27.8% (gzylu) of the variability of the
`filling properties can be explained, and therefore a prediction
`of filling properties from a given formulation appears to be
`impossible. Looking in detail at the interranging communalities
`(dz), it can be seen that the minimum bulk density of the powders
`is best described, whereas the Hausner’s ratio is clearly less
`
`Mylan v. Qualicaps, |PR2017—OO203
`QUALICAPS EX. 2014 — 9/12
`
`Mylan v. Qualicaps, IPR2017-00203
`QUALICAPS EX. 2014 - 9/12
`
`

`

`Multivariate Statistical Analysis and Capsule Formulation
`
`947
`
`Table 3. Variable Group X Used in the Nonparametric Canonical
`dependent on the formulation components. The major influence
`Analysis
`factors are probably the particle size of the drug, the amount
` of glidant used and the type of filler and disintegrant incorpo-
`EN
`D'(pS)
`(M501)
`ft
`fl
`dt
`d1
`H
`31
`dc
`rated into the formulation.
`1
`4
`3
`2
`6
`1
`3
`3
`1
`3
`Secondly, the same set of data was used in the nonparamet-
`2
`4
`3
`2
`6
`1
`3
`3
`2
`3
`ric canonical analysis (see Table 5). The test of significance
`3
`4
`3
`2
`4
`1
`3
`3
`4
`3
`already indicates that using this method the relationship between
`4
`4
`3
`2
`4
`1
`3
`3
`5
`3
`the two groups of variables can be identified in more detail,
`3
`4
`3
`2
`6
`1
`3
`1
`3
`3
`because nonlinear aspects are also evaluated. Furthermore, the
`:3]
`:
`3
`3
`2
`1
`3
`i
`3
`3
`ranking of the response variables to transfer them into ordinal
`8
`4
`3
`2
`4
`1
`3
`5