Pharmaceutics
`"IE Sliiflllllfl [ll llllflflflfl [lll‘lll llflSiflll
`
`,
`
`;
`
`; Ediftédby'M E AuIth ,
`
`>
`
`I
`
`i.
`
`1.
`
`AstraZeneca Exhibit 2094 p. 1
`InnoPharma Licensing LLC V. AstraZeneca AB IPR2017-00904
`Fresenius-Kabi USA LLC V. AstraZeneca AB IPR2017-01910
`
`

`

`CHURCHILL LIVINGSTONE
`
`Medical Division of Longmém Group UK Limited
`Distributed in the United States of America by
`Churchill Livingstone 1110., 650 Avenue of the Americas,
`New York, 10011, and associated companies, branches
`and representatives throughout the world.
`
`© Michael Aulton 1988
`
`All rights reserved. No part of this publication may
`be reproduced, stored in a retrieval system, or
`transmitted in any form or by any means, electronic,
`mechanical, photocopying, recording, or otherwise,
`without the prior permission of the publishers
`(Churchill Livingstone, Robert Stevenson House, 1-3
`Baxter’s Place, Leith Walk, Edinburgh EH1 3AF'), or
`a Licence permitting restricted copying in the United
`Kingdom issued by the Copyright LiCensing Agency Ltd,
`90 Tottenham Court Road, London, WlP 9HE.
`
`First published 1988
`Reprinted 1989
`Reprinted 1990
`Reprinted 1991
`Reprinted 1992
`
`ISBN D-LiLiB-UBELiB-fi
`
`British Library Cataloguing in Publication Data
`Pharmaceutics: the science of dosage form
`design.
`1. Pharmaceutics
`I. Aulton, Michael E.
`615’.19
`RS403
`
`Library of Congress Cataloging in Publication Data
`Pharmaceutics: the science of dosage form design.
`Replaces: Cooper and Gunn’s tutorial pharmacy.
`6th ed. 1972.
`
`2. Drugs
`
`Includes bibliographies and index.
`1. Drugs — Design of delivery systems.
`— Dosage forms. 3. Biopharmaceutics.
`4. Pharmaceutical technology.
`5. Chemistry,
`Pharmaceutical.
`6. Microbiology, Pharmaceutical.
`I. Aulton, Michael E.
`2. Chemistry,
`[DNLM: l. Biopharmaceutics.
`Pharmaceutical.
`3. Dosage Forms.
`4. Technology,
`Pharmaceutical.
`5. Microbiology, Pharmaceutical.
`QV 785 P5366]
`RS420.P48
`1987
`
`615.5’8
`
`86—25888
`
`Printed in Hong Kong
`CPP/OS
`
`The
`publisher’s
`policy is to use
`paper manufactured
`from sustainable forests
`
`AstraZeneca Exhibit 2094 p. 2
`
`

`

`Contents
`
`
`
`1
`Preface
`Contributors
`Acknowledgements
`About this book
`.
`1 The demgn Of dosage forms
`PART ONE Physicochemical
`principles of pharmaceutics
`2 Rheology and the flow of fluids
`3 Solutions and their properties
`4 Surface and interfacial phenomena
`5 Solubility and dissolution rate
`6 Disperse systems
`7 Kinetics and Stablhty testing
`PART TWO Biopharmaceutics
`8 Introduction to biopharmaceutics
`9 Factors influencing bioavailability
`10 Assessment of bioavailabilit
`11 Dosage regimens
`"
`PART THREE Drug delivery systems
`.
`12 Packs for pharmaceutical products
`.
`13 Preformulatlon
`.
`14 Solutions
`.
`15 SuspenSlons
`.
`16 EmuISions
`17 Powders and granules
`18 Tablets
`19 Capsules
`.
`20 Therapeutic aerosols
`21 Parenteral products
`22 Topical preparations
`23 Suppositories and pessaries
`
`-
`
`1
`
`15
`17
`38
`50
`62
`81
`119
`129
`131
`135
`174
`191
`213
`215
`223
`254
`269
`282
`_ 300
`304-
`322
`341
`359 '
`381
`412
`
`PART FOUR Pharmaceutical
`vii
`ix microbiology
`Xi
`24 Fundamentals of microbiology
`xiii
`’25 The action of physical and chemical
`agents on micro-organisms
`26 Principles of sterilization
`27 Microbiological Contamination and
`preservation of pharmaceutical
`preparations
`28 Pharmaceutical applications of
`microbiological techniques
`PART FIVE Pharmaceutical
`technology
`29 Materials of fabrication and corrosion
`30 Heat transfer and the properties of
`steam
`31 Filtration
`32 Mixing
`33 Particle size analysis
`34 Particle Size “dual?“
`35 Particle Size separation
`36 Powder flow
`.
`37 Granulation
`_
`.
`.38 Drying
`.
`39 Tableting
`.
`40 Tablet coating
`.
`41 Encapsulation
`.
`.
`42 DeSlgn and operation of clean rooms
`.
`.
`.
`.
`,
`43 Sterilization practice
`,
`44 P k .
`t hnol
`ac aging ec
`Index
`
`,
`
`.
`
`ogy
`
`423
`425
`
`452
`472
`
`479
`
`49-1
`
`509
`511
`
`525
`538
`550
`564
`581
`591
`600
`
`616
`
`629
`647
`669
`678
`
`686
`700
`712
`
`725
`
`AstraZeneca Exhibit 2094 p. 3
`
`

`

`l3
`5‘ I Wells and M E Aulton
`
`
`Preformulation
`
`__________________—_.__———————-—————-———
`
`THE CONCEPT OF PREFORMULATION
`
`1 SPECTROSCOPY
`
`'
`
`2 SOLUBILITY
`
`Intrinsic solubility (Co)
`pKa from solubility data
`Salts
`
`Solvents
`
`Partition coefficient (Kw)°
`Cosolvent solubility
`Methodology and structure activity prediction
`Choice of non-aqueous solvent (oil)
`Structure-activity relationships
`Dissolution
`
`Intrinsic dissolution rate
`
`Measurement of intrinsic dissolution rate
`
`Common ion effect
`
`3 MELTING POINT
`
`Capillary melting
`Hot stage microscopy
`Differential scanning calorimetry and differential
`thermal analysis
`Polymorphism
`Pseudopolymorphism
`True polymorphism
`Crystal purity
`Solubility
`
`4 ASSAY DEVELOPMENT
`
`U.v. spectroscopy
`Molecular weight
`
`pKa
`Mixtures
`
`Thin layer chromatography (t.l.c.)
`High performance liquid chromatography
`(h.p.l.c.)
`
`Normal phase h.p.l.c.
`Reserve phase h.p.l.c.
`
`5 DRUG AND PRODUCT STABILITY
`
`Temperature
`Order of reaction
`
`Hydrolysis
`The influence of pH
`
`Solvolysis
`Oxidation
`
`Chelating agents
`Photolysis
`Solid-state stability
`
`Hygroscopicity
`Stability assessment
`
`6 MICROSCOPY
`
`Crystal morphology
`Particle size analysis
`
`7 POWDER FLOW PROPERTIES
`
`Bulk density
`Angle of repose
`
`8 COMPRESSION PROPERTIES
`
`Interpretation of the results from the compression
`scheme
`
`Plastic material
`
`Fragmenting material
`Elastic material
`
`Punch filming (sticking)
`
`9 EXCIPIENT COMPATABILITY
`
`Method
`
`Interpretation
`
`10 CONCLUSIONS
`
`223
`
`AstraZeneca Exhibit 2094 p. 4
`
`

`

`224 DRUG DELIVERY SYSTEMS
`
`THE CONCEPT OF PREFORMULATION
`
`Almost all new drugs which are active orally are
`marketed as tablets, capsules or both (Table 13.1).
`Although only a few drug compounds are eventu-
`ally marketed as an injection (25% of those drugs
`marketed as tablets), an injection, particularly by
`the intravenous route,
`iS always required during
`early toxicity, metabolic, bioavailability and
`clinical studies in order to guarantee precise drug
`and dose deposition. Other dosage forms may be
`required (Table 13.1) but these are usually drug
`specific andoften depend to a large extent on the
`successful development of tablets, capsules and
`injections.
`Prior to the development of these three major
`dosage forms with a new drug candidate,
`it
`is
`
`Table 13.1 Frequency distribution of dosage form types
`manufactured in the UK (Data Sheet Compendium)M
`
`
`
`Dosage form Frequency (%)H
`
`45.8
`Tablets
`16.0
`Liquid oral
`15.0
`Injections
`13.0
`Capsules
`3.3
`Suppositories and pessaries
`3.0
`Topicals
`1.8
`Eye preparations
`1.2
`Aerosols (inhalation)
`0.3
`Others
`——————_——_*_._
`
`that certain fundamental physical and
`essential
`chemical properties of the drug molecule and
`other derived properties of the drug powder are
`determined. This information will dictate many of
`the subsequent events and possible approaches in
`formulation development. This
`first
`learning
`phase is known as preformulation.
`A suggested list of information required in
`preformulation is
`shown in Table 13.2.
`It
`is
`assembled in a logical order recognizing the relative
`importance of the data and probable existence of
`only limited quantities of bulk drug (mg rather
`than g) at this stage of its development. Investi-
`gators Should also be pragmatic and only generate
`data which is of immediate relevance, especially
`if the likely dosage forms are known.
`However, a knowledge of two fundamental
`properties is mandatory for a new compound:
`
`intrinsic solubility (C0),
`1
`2 dissociation constant (pKa).
`
`These will immediately determine (a) the need and
`(b) the possibility of making more soluble salts of
`the drug to eliminate solubility-related poor bio-
`availability, particularly from solid dosage forms.
`Independent of this pharmaceutical profiling
`(Table 13.2), analysts will need to generate data
`(Table 13.3) to support the assay of existing and
`
`Table 13.2 Preformulation drug characterization
`
`M T
`
`Method/function/characterization ‘
`est
`References/BibliographyW
`l Spectroscopy
`Simple u.v. assay
`Dalglish (1969)
`
`2 Solubility
`aqueous
`pKa
`salts
`solvents
`partition coeff K3,
`dissolution
`
`3 Melting point
`
`Phase solubility, purity
`Intrinsic solubility, pH effects
`Solubility control, salt formation
`Solubility, hygroscopicity, Stability
`Vehicles, extraction
`Lipophilicity, structure activity
`Biopharmacy
`
`DSC ——- polymorphism, hydrates, solvates
`
`4 Assay development
`
`u.v., t.l.c., h.p.l.c.
`
`5 Stability (in solution
`and in solid State)
`
`Thermal, hydrolysis, oxidation, photolysis,
`metal ions, pH
`
`Mader (1954); Higuchi and Connors (1965)
`
`Albert and Serjeant (1984)
`Berge, et al. (1977)
`YalkOwski and Roseman (1981)
`Leo, et al. (1971)
`Swarbrick (1970)
`
`Wendlandt (1974); Haleblian (1975),
`Haleblian and McCrone (1969)
`
`Jaffe and Orchin (1962); Bristow (1976)
`
`Mollica et al. (1978); Connors et al. (1979)
`
`6 Microscopy
`7 Powder flow
`
`Morphology, particle Size
`
`McCrone et al. (1978)
`
`(a) bulk density
`(b) angle of repose
`
`Tablet and capsule formulation
`Tablet and capsule formulation
`
`8 Compression properties
`
`Aids excipient choice
`
`Neumann (1967)
`Neumann (1967)
`
`DeBoer et a1. (1978); Jones (1981)
`
`9 Excipient compatibility
`
`Preliminary screening by DSC, confirmation by
`t.l.c.
`M
`
`Smith (1982)
`
`AstraZeneca Exhibit 2094 p. 5
`
`

`

`Table 13.3 Analytical preformulation
`
`
`
`Attribute Test
`
`Identity
`
`Purity
`
`Assay
`
`Nuclear magnetic resonance (n.m.r.)
`Infra red spectroscopy (i.r.)
`Ultraviolet spectroscopy (u.v.)
`Thin layer chromatography (t.l.c.)
`Differential scanning calorimetry (DSC)
`Optical rotation, where applicable
`
`Moisture (water and solvents)
`Inorganic elements
`Heavy metals
`Organic impurities
`Differential scanning calorimetry (DSC)
`Titration
`
`Ultraviolet spectroscopy (u.v.)
`High performance liquid chromatography
`(h.p.l.c.)
`
`Quality
`
`Appearance
`Odour
`Solution colour
`
`pH of slurry (saturated solution)
`
`Melting point
`
`new bulk material. Although their data meet a
`different need, it can often be. used to complement
`and confirm pharmaceutical data. Their greater
`training and knowledge in analysis will assist, for
`example, in the identification of suitable stability-
`indicating assays by ultraviolet spectroscopy (u.v.)
`or high performance liquid chromatography
`(h.p.l.c.) and the screening of incompatibilities by
`thin layer chromatography (t.l.c.).
`
`1 SPECTROSCOPY
`
`The first step in preformulation is to establish a
`simple. analytical method so
`that
`all
`future
`measurements can be quantitative. Most .drugs
`absorb light
`in the ultraviolet wavelengths
`(190—390 nm) since they are generally aromatic
`and/or contain double bonds. Confirmation of
`synthetic structure is, of course, the responsibility
`of the synthetic chemist. The acidic or basic
`nature of the molecule can be predicted from the
`functional groups in its structure (Perrin et al.,
`1981). This will
`indicate suitable solvents
`to
`ensure solution of either the ionized or undis—
`sociated species. This is important since the ionic
`status of a drug can alter the shape of the u. v.
`spectrum by increasing absorption or changing the
`wavelengths (bathochromic (red) or hypsochromic
`(blue) shifts) at which maxima, minima or both
`occur.
`
`PREFORMULATION
`
`225
`
`the new drug
`spectrum of
`Once the u.v.
`molecule is established, it is possible to choose an
`analytical wavelength (often kmax) that is suitable
`to quantify the amount of drug in a particular
`solution. Excitation of the molecule in solution
`
`causes a loss in light energy and the net change
`from the intensity of the incident light (Io) and the
`transmitted light
`(I)
`can be measured. The
`amount of light absorbed by a solution of drug is
`proportional to the concentration (C) and the path
`length of the solution (I) through which the light
`source has passed. Equation 13.1 is
`the well-
`known Beer—Lambert
`law where e is the molar
`
`extinction coefficient.
`
`Absorbance (A) = loglo (IQ/I) = eCl
`
`(13.1)
`
`in pharmacy to quote the
`is usual
`it
`However,
`specific absorption coefficient (E } 22/3,, or E i for short)
`where the pathlength is
`1 cm and the solution
`concentration is 1% w/v (10 mg ml'l) since doses
`and concentrations are generally in metric units.
`
`2 SOLUBILITY
`
`the
`When a preformulation programme begins,
`availability of drug is always limited and the
`preformulation scientist may only have 50 mg.
`Thus,
`it
`is important that
`the best use of this
`limited bulk is made to support the continuing
`efforts to the synthetic chemists and the biologists
`pursuing activity and toxicity screens. Further-
`more, because the compound is new, the quality
`is invariably poor,
`so that a large number of
`impurities may be present and often the first crys-
`tals come down as a metastable polymorph (see
`Section 3 of this chapter and Chapter 5). Accord-
`ingly, if nothing else is measured, the solubility
`and pKa must be determined. These control all
`future work. The solubility dictates the ease with
`which formulations
`for
`intravenous injection
`studies in animals are obtained. The pKa allows
`the informed use of pH to maintain solubility and
`to choose salts should they be required to achieve
`good bioavailability from the solid state (Chapter
`9) and improve stability (Chapter 7) and powder
`properties (Chapter 36).
`Kaplan (1972) suggested that unless acompound
`has
`an aqueous
`solubility in excess of
`1%
`(10 mg mlrl) over the pH range 1—7 at 37 °C,
`
`AstraZeneca Exhibit 2094 p. 6
`
`

`

`226
`
`DRUG DELIVERY SYSTEMS
`
`then potential bioabsorption problems may occur.
`He also found that if the intrinsic dissolution rate
`was greater than 1mg cm—2 min—1 then absorp-
`tion was unimpeded. However, dissolution rates
`of less than 0.1 mg cm‘2 min*1 were likely to give
`dissolution rate-limited absorption. This ten-fold
`difference in dissolution rate translates into a
`lower
`limit
`for solubility of
`1 mg ml‘1 since,
`under sink conditions, dissolution rate and solu-
`bilities are proportional (Hamlin er al., 1965).
`A solubility of less than 1 mg ml“l indicates the
`need for a salt, particularly if the drug will be
`formulated as a tablet or capsule. In the range
`1—10 mg ml—1
`serious consideration should be
`given to salt formation. When the solubility of the
`drug cannot be manipulated in this way (as in the
`case of neutral molecules, glycosides, steroids,
`alcohols or where the pK, is less than 3 for a base
`or greater than 10 for an acid) then liquid filling
`in soft gelatin capsules (in a solution in PEG 400,
`glyceryl triacetate or fractionated coconut oil) or
`as
`a paste or
`semisolid (dissolved in oil or
`triglyceride) in a hard gelatin capsule may be
`necessary.
`
`Intrinsic solubility (Co)
`
`An increase in solubility of the new drug in an
`acidic solution compared with its aqueous solu-
`
`bility suggests a weak base, and an increase in
`alkali, a weak acid. In both cases a dissociation
`constant (pK,) will be measurable and salts should
`form. An increase in both acidic and alkaline solu—
`bility suggests either amphoteric or zwitterion
`behaviour; in this case there will be two pKas,-one
`acidic and one basic. No change in solubility
`suggests a non—ionizable, neutral molecule with no
`measurable pKa. Here solubility manipulations
`will require either solvents or complexation.
`When the purity of the drug sample can be
`assured, the solubility value obtained in acid for
`a weak acid or alkali for a weak base can be
`assumed to be the intrinsic solubility (CD). The
`solubility should ideally be measured at
`two
`temperatures:
`(a) 4 or
`5 “C to ensure good
`physical stability and to extend short-term storage
`and chemical stability until more definitive data
`is available and (b) 37 °C to support biopharma—
`ceutical evaluation.
`However, since absolute purity is often in doubt
`for the first few batches of new drug, it is more
`accurate to determine this crucial solubility by use
`of a phase—solubility diagram (Fig. 13.1). The data
`are obtained from a series of experiments in which
`the ratio of the amount of drug to the amount of
`dissolving solvent is varied. Any deviation from
`the horizontal is indicative of impurities which a
`higher drug loading and its inherent impurities
`
`
`
`lSoluhiiity(mgml“)
`
`9
`
`
`Self-association: complexation
`
`or soiubilization by impurities
`
`
`
`
`
`
`Pure—no interaction
`
`Suppression by common ion effect
`or salting out
`
`
`
`1
`
`2
`
`4
`
`Drug /Solvent Phase Ratio
`
`8
`
`Fig". 13.] Effect of drug: solvent ratio on solubility when the drug is iulpure. Assuming the compound is a base and the
`estimate of its solubility in 0.1 M NaOH was 1 mg ml‘l, four solutions of 3 ml should be set up containing 3, 6, 12 and
`Mam-g of drug. These give the phase ratios shown here. 3 ml is the smallest volume which can be manipulated for either
`cenmfugauon or filtration and dilution for u.v. analysis. The vials should be agitated continuously overnight and then the
`concentration in solution determined
`
`AstraZeneca Exhibit 2094 p. 7
`
`

`

`
`Table 13.4 Potential pharmaceutical salts
`
`
`Basic drugs
`Acidic drugs
`
`
`Anion
`pK"
`“/n Usage“
`Cation
`13K,
`“/0 Usage"
`
`PREFORMULATION 227
`
`Hydrochloride
`Sulphate
`Tosylate
`Mesylate
`Napsylate
`Besylate
`Maleate
`Phosphate
`Salicylate
`Tartrate
`Lactate
`Citrate
`Benzoate
`Succinate
`Acetate
`Others
`
`—-6. 10
`—3.00, +1.96
`— 1.34
`— 1. 20
`0.17 ‘
`0.70
`1.92, 6.23
`2.15, 7.20, 12.38
`3.00
`3.00
`3.10
`3.13, 4.76, 6.40
`4.20
`4.21, 5.64
`4 . 76
`
`Potassium
`Sodium
`Lithium
`Calcium
`Magnesium
`Diethanolamine
`Zinc
`Choline
`Aluminium
`Others
`
`43.0
`7.5
`0.1
`2.0
`0.3
`0.3
`3.0
`3.2
`0.9
`3.5
`0.8
`3.0
`0.5
`0.4
`l .3
`30.2
`
`16.00
`14.77
`13. 82
`12.90
`11.42
`9.65
`8.96
`8.90
`5.00
`
`10.8
`62.0
`1.6
`10.5
`1.3
`1.0
`3.0
`0.3
`0.7
`8.8
`
`* Martindale (1982), The Extra Pharmacopoeia, 28th edition. The Pharmaceutical Press, London.
`
`either promotes or suppresses solubility. In the
`cases where the observed result changes with the
`amount of solvent, the line is extrapolated to zero
`phase ratio, where solubility will be independent
`of solvent level and a true estimate of the intrinsic
`
`the drug. The United States
`solubility of
`Pharmacopoeia uses this method to estimate the
`purity of mecamylamine hydrochloride.
`
`pJKa from solubility data
`
`Seventy—five per cent of all drugs are weak bases
`(20% are weak acids and the remaining 5% are
`non—ionic, amphoteric or alcohols). It is therefore
`appropriate to consider the Henderson—Hasselbalch
`equations for weak bases and acids. These have
`been discussed in Chapter 3 and their bioavail~
`ability consequences in Chapter 9.
`For weak bases
`
`DH = pKa + logloflB] / [BH+])
`
`(13.2)
`
`and for weak acids
`
`PH Z pKa + 10gm([A‘] / [HAD '(13-3)
`
`Equations (13.2) and (13.3) are used:
`
`(a)
`
`to determine the pK, by following changes in
`solubility,
`(b) to allow the prediction of solubility at any pH
`provided that the intrinsic solubility (Co) and
`pK, are known, and
`to facilitate the selection of suitable salt-
`
`(c)
`
`forming cornpounds and predict the solubility
`and pH properties of the salts.
`
`Albert and Serjeant (1984) give a detailed account
`of how to obtain precise pKa values by potentio-
`metry, spectroscopy and conductivity and there is
`further discussion in Chapter 3 of this book.
`
`Salts
`
`in solubility can be
`improvement
`A major
`achieved by selection of a salt. Acceptable phar—
`maceutical salts are shown in Table 13.4 which
`
`also includes their corresponding pK, values (see
`Chapter 3 for further discussion). As an example,
`the consequences of changing chloridiazepoxide to
`various salt forms is shown in Table 13.5.
`
`Table 13.5 Theoretical solubility and pH of salts of
`chlordiazepoxide
`
`Salt
`
`Base
`
`pKa
`
`Salt pH Solubility
`(mg ml")
`
`4. 80
`
`8.30
`
`2.0
`
`Hydrochloride
`Sulphate
`Besylate
`Maleate
`Tartrate
`Benzoate
`Acetate”)
`
`<165‘”
`2.53
`—6.10
`Freely soluble
`2.53
`—3.00
`Freely soluble
`2.53
`0.70
`57.1
`3.36
`1.92
`17.9
`3.90
`3.00
`6.0
`4.50
`4.20
`
`4.76 ——4rl 4.78
`
`
`(1) Maximum solubility of chlordiazepoxide hydrochloride,
`achieved at pH 2.39, is governed by crystal lattice energy
`and common ions.
`
`(23 Chlordiazepoxide acetate may not form; pKI too high and
`close to drug.
`
`AstraZeneca Exhibit 2094 p. 8
`
`

`

`228
`
`DRUG DELIVERY SYSTEMS
`
`In some cases, salts prepared from strong acids
`or bases are freely soluble but also very hygro-
`scopic. This can lead to instability in tablet or
`capsule formulations since some drug will dissolve
`in its own adsorbed films of moisture (the usual
`prerequisite for breakdown) and in the case of a
`weak base, a strongly acid solution may be auto-
`catalytic. Accordingly it is often better to use a
`weak acid or base to form the salt provided any
`solubility requirements are met. A salt which is
`less soluble will also be generally less hygroscopic
`and form less acidic or basic solutions (Table
`13.5). This may also be important in physiological
`terms. Injections should lie in the pH range 3—9
`to prevent vessel or tissue damage and pain at the
`injection site. Oral syrups should not be too acidic
`to enhance palatability. Packaging may also be
`susceptible; undue alkalinity will attack glass, and
`hydrochloride salts should not be used in aerosol
`cans since a propellant-acid reaction will corrode
`the canister.
`
`It is also clear from Table 13.5 that not only
`does the intrinsic pH of the base solution, in this
`example chlordiazepoxide, fall significantly from
`pH 8.3 if salt
`forms are produced but, as
`a
`consequence, the solubility increases exponentially
`(Eqns 13.2 and 13.3). This has important impli-
`cations in viva. A weak base with an intrinsic solu—
`bility greater than 1 mg ml"1 will be freely soluble
`in the gastrointestinal
`tract, especially in the
`stomach. None the less it
`is usually better to
`formulate with a salt since it will control the pH
`of
`the diffusion layer
`(the saturated solution
`immediately adjacent
`to the dissolving surface).
`For example, although chlordiazepoxide base (C8
`= 2 mg ml‘1 at pH 8.3) meets the requirements
`for in viva ‘solubility’ (Kaplan, 1972); commercial
`capsules contain chlordiazepoxide hydrochloride
`(CS = 165 mg ml“1 at pH 2.53).
`A weak base will have a high dissolution rate
`in the stomach, but as it moves down the gastro-
`intestinal tract the pH rises and dissolution rate
`falls. Conversely a weak acid
`has minimal
`dissolution in the stomach but it becomes more
`soluble and its dissolution rate increases as it
`moves down the gut. Paradoxically as dissolution
`rate increases, so absorption falls because the drug
`is ionized.
`
`than the parent drug.
`usually much greater
`Sodium and potassium salts of weak acids dissolve
`much more rapidly than the parent acid and some
`comparative data are shown in Table 13.6. On the
`basis of bulk pH these salts would be expected to
`have lower dissolution rates
`in the stomach.
`However, the pH of the diffusion layer (found by
`measuring the pH of a saturated bulk solution) is
`higher than the pH of gastric fluid (which is
`usually approximately pH 1.5) because of their
`buffering action. The pH approximates
`to a
`saturated unbuffered aqueous solution (calculated
`pH in Table 13.6) and the dissolution rate is
`governed by this pH and not the bulk media pH.
`In the intestine, the salt does not depress the pH,
`unlike the acid which is neutralized, and the
`diffusion layer pH is again raised to promote
`dissolution.
`Since ‘ solubility
`is
`exponentially
`dependent on pH (Eqns 13.2 and 13.3) there is
`a significant increase in dissolution rate over the
`free acid. Providing that the acid forming the salt
`is strong, the pH of the solution adjacent to the
`dissolving surface will be that of the salt, whereas
`for the dissolving free base, it will be the pH of
`the bulk dissolving media. With weak bases, their
`salts dissolve rapidly in the stomach but there is
`no absorption since the drug is
`ionized and
`absorption is delayed until the intestine. There,
`any undissolved drug, as salt, rapidly dissolves
`since the higher diffusion layer pH compensates
`for the higher bulk pH which would be extremely
`unfavourable to the free base. Data for chlordi-
`azepoxide are shown in Table 13.5. The maleate
`
`-
`
`Table 13.6 Dissolution rates of weak acids and their sodium
`salts
`
`
`Drug
`
`pKa
`
`pH
`(at C5)
`
`Dissolution rate
`
`(mg cm'Z min”) X 102
`Dissolution media
`
`
`phosphate
`0.1M HCl
`
`(pH 1.5) (pH 6.8)a“
`
`Salicylic acid
`Sodium salicylate
`
`Benzoic acid
`Sodium benzoate
`
`Sulphathiazole
`Sodium
`
`3.0
`
`4.2
`
`7.3
`
`2.40
`8.78
`
`2.88
`9.35
`
`4.97
`
`1.7
`1870
`
`2.1
`
`980
`
`27
`2500
`
`14
`1770
`
`<0.1
`
`0.5
`
`
`
`10.75 5 50sulphathiazole 810
`
`
`
`
`
`The dissolution rate of a particular
`
`salt
`
`is
`
`Dissolution rate data from Nelson (1958).
`
`AstraZeneca Exhibit 2094 p. 9
`
`

`

`PREFORMULATION
`
`229'
`
`salt has a predicted solubility of 57 mg 1111‘1 but
`more importantly reduces the pH by 5 units. By
`controlling diffusion layer pH, the dissolution rate
`can increase many fold,
`independently of
`its
`position in the gastrointestinal tract.
`Miller and Holland (1960) stated that different
`salts of a drug rarely change its pharmacology,
`only its physical properties. Wagner (1961) qual-
`ified this statement to acknowledge that salts do
`effect the intensity of the response. The salt form
`does change the physicochemical properties of the
`drug. Changes in dissolution rate and solubility
`affect the rate and extent of absorption (bioavail-
`ability),
`and changes
`in hygroscopicity
`and
`stability influence formulation.
`Consequently each new drug candidate must be
`examined extensively to choose the most suitable
`salt for formulation and because each potential salt
`will behave differently, each requires separate
`preformulation screening.
`
`Solvents
`
`As mentioned in the introduction, it is generally
`necessary to anticipate the formulation of an injec-
`tion even if there is no intention of actually
`marketing such a product. The first choice for a
`solvent
`is obviously water. However, although
`the drug may be freely soluble, some are unstable
`in aqueous solution. Accordingly water-miscible
`solvents must be used (a) as cosolvents in formu-
`lations to improve solubility or stability and (b) in
`analysis to facilitate extraction and separation (e.g.
`in chromatography).
`Oils are used in emulsions, topicals (creams and
`ointments), intramuscular injections and liquid-fill
`oral preparations (soft and hard gelatin capsules)
`when aqueous pH and cosolvent solubility and
`
`stability are unattainable. Table 13.7 shows a
`range of solvents to fulfil
`these needs. Aqueous
`methanol is widely used in h. p.l.c. and is the stan-
`dard solvent in sample extraction during analysis
`and stability testing. It is often made acidic or
`alkaline to increase solvent power and ensure
`consistent ionic conditions for u.v. analysis. Other
`pharmaceutical solvents are available (see Spiegel
`and Noseworthy, 1963) but are generally only
`required in special cases. The most acceptable
`non-aqueous
`solvents pharmaceutically are
`glycerol, propylene glycol and ethanol. Generally
`for a lipophilic drug (Le. one with a partition
`coefficient
`(log P) greater
`than 1),
`solubility
`doubles through this series.
`Where bulk is limited and when the aqueous
`solubility is inadequate, it is useful to measure the
`solubility in aqueous cosolvent mixtures rather
`than in a pure organic solvent. Whereas solubili-
`ties at other levels and their mixtures can be
`predicted, the solubility in pure solvent is often
`inconsistent. Furthermore,
`formulations
`rarely
`demand pure non-aqueous solvent, particularly
`injections. For example, ethanol should only be
`used up to 10% in an injection to prevent haemo-
`lysis and pain at the injection site (Cadwallader,
`1978).
`The reader will find more details on the prop-
`erties of solutions in Chapter 3 and on the formu-
`lation of solutions in Chapter 14.
`
`Partition coefficient (K3,)
`
`(solventzwater quotient of
`Partition coefficient
`drug distribution) has a number of applications
`which are relevant to preformulation.
`(a) Solubility: both aqueous and in mixed
`solvents.
`
`Table 13.7 Recommended solvents for preformulation screening
`Salami
`Dielectric constant (e)
`Solubility parameter (6)
`
`
`
`Application
`
`80
`32
`
`24.4
`14.7
`
`All
`Water
`Extraction, separation
`Methanol
`Dissolution (gastric), basic extraction
`0.1 M HCl (pH 1.07)
`Acidic extraction
`0.1 M NaOH (pH 13.1)
`Dissolution (intestinal)
`Buffer (pH 7)
`Formulation, extraction
`12.7
`24
`Ethanol
`Formulation
`12.6
`32
`Propylene glycol
`Formulation
`16.5
`43
`Glycerol
`
`
`
`35PEG 300 or 400 Formulation
`
`Solvents are considered further in the chapter on the formulation of solutions (Chapter 14).
`
`AstraZeneca Exhibit 2094 p. 10
`
`

`

`230
`
`DRUG DELIVERY SYSTEMS
`
`(b) Drug absorption in viva: applied to a homo-
`logous drug series
`for Structure
`activity
`relationships (SAR).
`(c) Partition chromatography: choice of column
`(h.p.l.c.) or plate (t,.l.c.) and choice of mobile
`phase (eluant), see Section 4.
`The measurement of K3, and its use in cosolvent
`
`solubility and structure activity relationships are
`pertinent here, while application in aqueous solu—
`bility prediction is discussed under melting
`point
`(Section 3) and in chromatography in
`Section 4.
`
`Cosoloent solubility
`
`The relative polarities of solvents have been scaled
`using dielectric constant (e), solubility parameter
`(6), interfacial tension (y) and hydrophilic—lipophilic
`balance (HLB), the latter normally being applied
`to non-ionic surfactants in emulsion technology.
`The best solvent
`in any given situation is one
`whose polarity matches that of the solute; an ideal,
`fully compatible solution exists when SW. =
`550mm. This can be ascertained by determining
`solubility maxima, using a substituent contri-
`bution approach (Fedors, 1974). An alternative
`measure is to evaluate the dielectic requirement of
`the system (Paruta & others, 1965).
`However the most practically useful scale of
`polarity or lipophilicity for a solute is its K8, oil:
`water partition coefficient,
`since
`the other
`approaches do not allow easy estimates for the
`behaviour of crystalline solids (solutes).
`For a .wide range of drugs it is possible to relate
`solvent
`solubility and the partition coefficient
`(log K3, # log P). Yalkowsky and Roseman (1981)
`derived the following expression for 48 drugs in
`propylene glycol
`
`log C, = log Cw + f(0.89 log P + 0.03)
`
`(13.4)
`
`Equation 13.4 can be applied more generally by
`introducing a factor (1) to account for the relative
`solvent power of pharmaceutical solvents found in
`practice (Table 13.8), and predicted by comparison
`of interfacial tension against a liquid hydrocarbon,
`tetradecane (Yalkowsky et al., 1975).
`Equation 13.4 now becomes, for a wide range of
`solvents:
`
`Table 13.8 Solvent power ((1)) of some pharmaceutical
`solvents
`
`Solvent
`Relative solvent
`power
`
`
`0.5
`Glycerol
`1
`Propylene glycol (PEG 300 and 400)
`2
`Ethanol
`
`DMA, DMF 4
`
`log C8 = log C0 + f(log q) + 0.89 log P + 0.03)
`(13.5)
`
`Methodology and structure activity prediction
`
`Choice of non-aqueous solvent (oil) The oil:
`water partition (K3,) is a measure of the relative
`lipophilicity (oil-loving) nature of a compound,
`usually in the unionized State (HA or B), between
`an aqueous phase and an immiscible lipophilic
`solvent or oil. Many partition solvents have been
`used (Leo et al., 1971), but the largest data base
`has been generated using n-octanol. This is, aside
`from any scientific argument,
`the major justifi-
`cation to continue its use in preformulation. The
`solubility parameter of octanol (5 = 10.24) lies
`midway in the range for the majority of 'drugs (5
`= between 8 and 12) although some non-polar
`(6 < 7) and polar drugs (6 > 13) are encountered.
`This allows more easily measurable results than in
`inert
`solvents
`(e.g. hydrocarbons) Since
`it
`is
`convenient to partition between equal volumes of
`oil and aqueous phases.
`A typical technique is the shake flask method
`whereby the drug, dissolved in one of the phases,
`is shaken with the other partitioning solvent for
`30 minutes, allowed to stand for 5 minutes and
`then the majority of the lower aqueous phase
`(density of octanol = 0.8258 g ml‘l) is run off
`and centrifuged for 60 minutes at 2000 rpm. The
`aqueous phase is assayed before (2C) and after
`(Cw,
`aqueous
`solubility) partitioning to give
`K3; 2 (2C _ Cw)/Cw)
`Clearly if the transfer of solute to the oil phase
`is small, AC“,
`is small and any analytical error
`increases error in the estimate of K8,.
`Indeed
`to encourage greater aqueous
`loss
`(>ACW)
`a
`considerably more polar solvent, n-butanol, has
`been used. Where the partition coefficient is high,
`it
`is usual
`to reduce the ratio of the oil phase
`
`AstraZeneca Exhibit 2094 p. 11
`
`

`

`PREFORMULATION
`
`231
`
`from 1:1 to 1:4 or 1:9 in order to increase the
`aqueous concentration (Cw) to a measurable level.
`For a 1:9 oil:water ratio K8, = (10 EC — Cw)/
`
`Cw.
`The partition of a polar solute between an inert
`non-polar hydrocarbon e.g. hexane, heptane etc.,
`is quite different to hydrogen bonding solvents
`like octanol
`(Hansch and Dunn,
`1972). The
`behaviour of the weak acid phenol (pK, = 10) and
`weak base nicotine (pK, = 3.1) is worthy of note.
`For phenol, {(3:3‘3“GI = 29.5 whereas Kim“ 2
`0.11. The acidic solvent, chloroform, suppresses
`partition (K3,. = 2.239) whereas although ethyl
`acetate and diethyl ether are more polar, the basic
`behaviour of the solvents give the highest K3,.
`With solvents capable of both hydrogen donation
`and acceptable (octanol,