`
`The Science of Dosage Form Design
`
`Edited by
`Michael E. Aulton BPharm PhD FAAPS MAPharmS
`Professor of Pharmaceutical Technology,
`School of Pharmacy,
`De Montfort University,
`Leicester, UK
`
`SECOND EDITION
`
`/,&\ CHURCHILL
`
`u
`" " " '~ LIVINGSTONE
`
`EDINBURGH LONDON NEW YORK OXFORD PHILADELPHIA ST LOUIS SYDNEY TORONTO 2002
`
`FRESENIUS EXHIBIT 1061
`Page 1 of 66
`
`
`
`CHU RCHILL LIVINGSTONE
`An imprint of Elsevier Science Limited
`
`© H arcourt Publishers Limited 2002
`© Elsevier Science Limited 2002. 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 othetwise,
`without either d1e prior permission of the publishers or a
`licence permitting resrricted copying in the Unired Kingd m
`issued by the Copyright Licensing Agency, 90 Tottenham
`Court Road, Lo:ndon \YI IT 4LP. Permissions may be sought
`directly from Els:cvie.r's Health Sciences Rights D epartment
`in Philadelphia, USA: phone: (+ I) 2 15 238 7869, fax : (+ I)
`21 5 238 2239, e-mail: healtl1permissions@elsevier.com. You
`may also complete your request on-line via the Elsevier
`Science homepage (http://www.elsevier.com), by electing
`' Customer Support' and then ' Obtaining Permissions'.
`
`First published I 988
`Second Edition 2002
`Reprinted 2002 twice
`
`Standard ed ition ISBN O 443 055 17 3
`
`Internationa l Student Edition ISBN O 443 05550 5
`Reprinted 2002 rwice
`
`British Library !Cataloguing in Publication D ata
`A catalogue record for this book is available from the British
`Library
`
`Library of Congiress Cataloging in P ublication Data
`A catalog record for this book is available from the Library of
`Congress
`
`Note
`Medical knowledge is constantly changing. As new
`information becomes available, changes in treatment,
`procedures, equipment and the use of d rugs become
`necessary. The editor, contributors and the publishers have
`taken ca re to ensure that the information given in this text is
`accurate and up to date. However, readers are strongly
`advised to confirm that the information, especially witl1
`regard to d rug usage, complies with the latest legislation and
`standards of practi1ce.
`
`MIMI !ournals and m~ltimedia
`
`p your source for books,
`
`on t:he health sC1ences
`www.elsevierhec1lth.com
`
`Printed in Spain
`
`The
`publisher's
`policy is to use
`paper manufac tured
`1rom sustalnable forests
`
`I
`
`FRESENIUS EXHIBIT 1061
`Page 2 of 66
`
`
`
`Contents
`
`What is 'Pharmaceutics'? xiii
`
`1. The design of dosage forms
`Peter York
`
`PART ONE
`Scientific principles of dosage form
`design 13
`2. Dissolution and solubility 15
`Michael Au/ton
`3. Properties of solutions 33
`Michael Aulron
`4. Rheology 4 1
`Chris Marrion
`
`5. Surface and interfacial phenomena 59
`John Fell
`6. Disperse systems 70
`David Anwood
`
`7. Kinetics and product stability 101
`John Pugh
`
`8. Pharmaceutical preformulation 11 3
`James Wells
`
`PART TWO
`Particle science and powder
`technology 139
`9. Solid-state properties 141
`Graham Buckton
`10. Particle-size analysis 152
`John Staniforrh
`11 . Particle-size reduction 166
`Joh11 Stamfonh
`12. Particle-size separation 174
`John Staniforth
`
`13. Mixing 181
`Andrew Twitchell
`
`14. Powder flow 197
`John Staniforth
`
`PART THREE
`Biopharmaceutical principles of drug
`delivery 211
`15. Introduction to biopharmaceutics 213
`Marianne Ashford
`
`16. The gastrointestinal tract - physiology and
`drug absorption 217
`Marianne Ashford
`
`17. Bioavailability - physicochemical and
`dosage form factors 234
`Marianne Ashford
`
`18. Assessment of biopharmaceutical
`properties 253
`Maria11ne Ashford
`19. Dosage regimens 275
`Smart Proudfoot, (updated by John Collett)
`20. Modified-release peroral dosage form 289
`John Collect, Chris M oreton
`
`PART FOUR
`Dosage form design and
`manufacture 307
`21. Solutions 309
`Michael Billany
`22. Clarification 323
`Andrew Twirchell
`
`23. Suspensions and emulsions 334
`Michael Bil/any
`
`FRESENIUS EXHIBIT 1061
`Page 3 of 66
`
`
`
`CONTENTS
`
`24. Powders and granules 360
`Malcolm Summers
`
`25. Granulation 364
`Malcolm Summers, Michael Aulron
`
`26. Drying 379
`M ichael Aulton
`
`27. Tablets and compaction 397
`Goran Alderbom
`
`28. Coating of tablets and
`multiparticulates 441
`John Hogan
`
`29. Hard gelatin capsules 449
`Brian Jones
`
`30. Soft gelatin capsules 461
`Keith Hutchison, Josephine Ferdinando
`
`31 . Pulmonary drug delivery 473
`Kevin Taylor
`
`32. Nasal drug delivery 489
`Peter Taylor
`
`33. Transdermal drug delivery 499
`Brian B any
`
`34. Rectal and vaginal drug delivery 534
`J osef Tukker
`
`35. Delivery of pharmaceutical proteins 544
`Daan Crommelin, Ewoud van Winden
`Albert M ekking
`
`36. Packs and packaging 554
`Dixie Dean
`
`37. Pharmaceutical plant design 571
`Michael Au/ion, Andrew Twitchell
`
`38. Heat transfer and the properties and use
`of steam 586
`A ndrew Twitchell
`
`PART FIVE
`Pharmaceutical microbiology 597
`39. Fundamentals of microbiology 599
`Geoff Hanlon
`
`40. Pharmaceutical applications of
`microbiological techniques 623
`No rman H odges
`
`41 . The action of physical and chemical agents
`on microorganisms 643
`Geoff H anlon, Norman Hodges
`
`42. Microbiological contamination and
`preservation of pharmaceutical
`products 658
`Malcolm Parker, Nonnan Hodges
`
`Index 669
`
`FRESENIUS EXHIBIT 1061
`Page 4 of 66
`
`
`
`8
`Pharmaceutical preformulation: the
`physicochemical properties of drug substances
`
`James Wells
`
`CHAPTER CONTENTS
`
`The concept of preformulation 114
`
`Spectroscopy 114
`
`Solubility 115
`Aqueous solubility 11 5
`Intrinsic solubility (C0)
`115
`pK. from solubility data 116
`Salts 116
`Solvents 118
`Partition coefficient (l(t>w)
`Solvent solubility 119
`Methodology and structure activity
`prediction 119
`Choice of non-aqueous solvent {oil) 119
`Structure-aclivity relationships 120
`Dissolution 122
`Intrinsic dissolution rate 122
`Measurement of intrinsic dissolution rate 123
`Common ion effect 123
`
`11 9
`
`Melting point 124
`Techniques 124
`Capillary melting 124
`Hot-stage microscopy 124
`Differenlial scanning calorimetry and thermal
`analysis 124
`Polymorphism 124
`Pseudopolymorphism (solvales) 125
`True polymorphism 126
`Crystal purity 126
`Solubility 126
`
`Assay development 127
`UV spectroscopy 128
`Molecular weight 128
`pKa 128
`Thin-layer chromatography 128
`High-performance liquid chromalography
`(HPLC) 128
`
`Normal-phase HPLC 129
`Reverse-phase HPLC 129
`
`Drug and product stability 129
`Temperature 130
`Order of reaction 130
`Hydrolysis 130
`The influence of pH
`Solvolysis 131
`Oxidation 131
`Chelating agents 131
`Photolysis 131
`Solid-state stability 132
`Hygroscopicity 132
`Stability assessment 132
`
`130
`
`Microscopy 132
`Crystal morphology 133
`Particle size analysis 133
`
`Powder flow properties 133
`Bulk density 133
`Angle of repose 134
`
`Compression properties 134
`Plastic material 134
`Fragmenlalion 135
`Elastic material 135
`Punch filming (sticking) 136
`
`Exclpient compatibility 136
`Method 136
`Interpretation 136
`
`Conclusions 1 38
`
`Reference$
`
`, 38
`
`Bibliography 138
`
`113
`
`FRESENIUS EXHIBIT 1061
`Page 5 of 66
`
`
`
`SCIENTIFIC PRINCIPLES O F DOSAGE FORM DESIGN
`
`THE CONCEPT OF PREFORMULATION
`
`Table 8.1 Frequency distribution of dosage form
`types manufactured In the UK
`
`Almost all new drugs are marketed as tablets, cap(cid:173)
`sules or both (Table 8.1). Although only a few ar e
`marketed as an injection (25% of those marketed as
`tablets) the intravenous route is always required
`during early toxicity, metabolic, bioavailabiliry and
`clinical studies to provide a precise drug and dose
`deposition. Other dosage forms may be required
`(Table 8. I ) but these are drug specific and depend to
`a large extent on the successful development of
`tablets, capsules and injections.
`Prior to the development of these three major
`dosage forms, it is essen tial that certain .fundamen(cid:173)
`tal physical and chemical properties of the drug
`molecule and other derived properties of the drug
`powder are d etermined . This information dictates
`many of the subsequent events and approach es in
`formulation development. This first learning phase is
`known as pref ormulation.
`A recommended list of the information required
`in preformulation is shown in Table 8.2. This is
`assembled, recognizing the relative importance and
`probable existence of only limited quantities of new
`bulk drug (mg rather than g). Invescigators must be
`pragmatic and generate data of immediate rele(cid:173)
`vance, especially if the likely d osage forms are
`known .
`Two fundamental p roperties are mandatory for a
`new compound:
`1. Intrinsic solubility (C0 ) ,
`2. Dissociation constant (pK.).
`
`Dosage form
`
`Frequency (%)
`
`Tablets
`
`Liquid oral
`
`Capsules
`
`Injections
`
`Suppositories and pessaries
`
`Topicals
`
`Eye p reparations
`
`Aerosols (inhalation)
`
`Others
`
`46
`
`16
`
`15
`
`13
`
`3
`
`3
`
`2
`
`1
`
`1
`
`Independent of this ph armaceutical profiling
`(Table 8.2), analysts will generate data (Table 8.3)
`to confirm str ucture and purity, and this should be
`used to complement and confirm pharmaceutical
`data. T heir greater training and knowledge in
`analysis will assist in the identification of suitable
`stability-indicating assays by high-performance
`liquid chromatography (HPLC).
`
`SPECTROSCOPY
`
`The first step in preformulation is to establish a
`simple analytical method. Most drugs absorb light in
`the ultraviolet wavelengths (I 90-390 run) as they are
`
`Tabla 8.2 Preformulatlon drug characterization
`
`Test
`
`Method/function/characterization
`
`Spectroscopy
`Solubility
`aqueous
`pK.
`salts
`solvents
`partition coeff Kbw
`dissolution
`Melting point
`Assay development
`Stability (in solution and solid state)
`Microscopy
`Powder flow
`bulk density
`angle of repose
`Compression propenies
`Excipieot compatibility
`
`11 4
`
`Simple UV assay
`Phase solub~ity. purity
`Intrinsic solubility, pH effects
`Solubility control. salt formation
`Solubility, hygroscopicity, stability
`Vehides. extraction
`Lipophilicity, structure activity
`BiOpharmacy
`DSC - poly morphism, hydratos. solvates
`UV. TLC , H PLC
`Thermal. hydrolysis, oxidation, photolysis, metal ions. pH.
`Morphology. panicle size
`Tablet and capsule formulation
`
`Tablet and capsule formation
`Excipient choice
`
`FRESENIUS EXHIBIT 1061
`Page 6 of 66
`
`
`
`Table 8.3 Analytical preformulatlon
`
`Attribute Test
`
`Identity Nuclear magne1ic resonance (NMR)
`Infra red spectroscopy (IR)
`Ullraviolel spectroscopy (UV)
`Thin-layer chromatography (TLC)
`Differential scanning calorimetry (DSC)
`Optical rotation. where applicable
`
`Purity
`
`Moislure (water and solvents)
`Inorganic elements
`Heavy metals
`Organic impurities
`Differential scanning calorimetry (DSC)
`
`Assay
`
`Titration
`Ultraviolet spectroscopy (UV)
`High-performance liquid chromatography (HPLC)
`Quality Appearance
`Odour
`Solution colour
`pH of slurry (sa1ura1ea solution)
`Melling poinl
`
`generally aromatic and contain double bonds. The
`acidic or basic nature of the molecule can be predicted
`from functional groups (Perrin et al 1981). Using the
`UV spectrum of the drug, it is possible to choose an
`analytical wavelength (often Amax) 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 (10 ) and the transmitted light (J) 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 ([) through which the
`light has passed. Equation 8.1 is the Beer- Lambert
`law, where e is the molar extinction coefficient.
`Absorbance (A) = log10 (// /) = eCl
`(8.1)
`In pharmacy it is usual to use the specific absorp(cid:173)
`~ion coefficient En m (El), where the pathlength
`1s l cm and the solution concentration is l % w/v
`(10 mg mL- 1), as doses of drugs and concentrations
`are generally in unit weights rather than molarity
`(El = lOe!MW').
`
`SOLUBILITY
`
`Aqueous solubility
`The availability of a drug is always limited and the
`preformulation scientist may only have 50 mg. As the
`
`PHARMACEUTICAL PREFORMULATION
`
`compound is new the quality is invariably poor, so
`that a large number of impurities may be present and
`often the first cr ystals come down as a metastable
`polymorph. Accordingly, as a minimum, the solubil(cid:173)
`ity and p.K;. must be determined. Solubility dictates
`the ease with which formulations for oral gavage and
`intravenous
`injection studies
`in animals are
`obtained. The pK. allows the informed use of pH to
`maintain solubility and to choose salts required to
`achieve good bioavailability from the solid state
`(Chapter 9) and improve stability (Chapter 7) and
`powder properties (Chapter 13 and 14).
`Kaplan (I 972) suggested that unless a compound
`has an aqueous solubility in excess of l % (1 O mg
`mL- 1) over the pH range 1-7 at 37°C, potential
`bioabsorption problems may occur. If the intrinsic
`d issolution rate was greater than 1 mg cm-2 min-1
`then absorption was unimpeded. Dissolution rates
`less than 0.1 mg cm-2 min- 1 were likely to give dis(cid:173)
`solution rate-limited absorption. T his tenfold differ(cid:173)
`ence in dissolution rate translates to a lower limit for
`solubility of 1 mg mL-1• Under sink conditions, dis(cid:173)
`solution rate and solubilities are proportional.
`A solubility of less than l mg mL- 1 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 (neutral mole(cid:173)
`cules, glycosides, steroids, alcohols, or where the p.K;.
`is less than 3 for a base or greater than l O for an
`acid) then liquid filJing in soft or hard gelatin cap(cid:173)
`sules may be necessary.
`
`Intrinsic solubility (C0 )
`An increase in solubility in acid compared to
`aqueous solubility suggests a weak base, and an
`increase in alkali a weak acid. In both cases a disso(cid:173)
`ciation constant (pK.) can be measured and salts
`should form. An increase in acidic and alkaline solu(cid:173)
`bility suggests either amphoteric or zwitterion
`behaviour. In this case there will be two pK.s, one
`acidic and one basic. No change in solubility sug(cid:173)
`gests a non-ionizable neutral molecule with no mea(cid:173)
`surable p.K;., and solubility manipulation 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 (C0 ) , ie. the fundemen(cid:173)
`tal solubility when completely unionized. The solu(cid:173)
`bility should
`ideally be measured at
`two
`temperatures:
`
`115
`
`FRESENIUS EXHIBIT 1061
`Page 7 of 66
`
`
`
`SCIENTIFIC PRINCIPLES OF DOSAGE FORM DESIGN
`
`1. 4°C to ensure physical stability and extend
`short-term storage and chemical stability until
`more definitive data are available. The maximum
`density of water occurs at 4°C. This leads to a
`minimum aqueous solubility.
`2. 37°C to support biopharmaceutical evaluation.
`
`However, as absolute purity is often in doubt it is
`more accurate to determine this crucial solubility by
`the use of a phase-solubility diagram (Fig. 8.1). 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 ind icative of
`impurities, which a h igher drug loading and its
`inherent impurities either promotes or suppresses
`solubility. In 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 an d the true intrinsic solubility of the
`drug. The United States Pharmacopoeia uses chis
`method to estimate the purity of mecamylamine
`hydrochloride.
`
`pKa from solubility data
`Seventy-five per cent of all drugs are weak bases;
`20% are weak acids and only 5% are non-ionic,
`amphoteric or alcohols. It is therefore appropriate to
`consider the Henderson-H asselbalch equations for
`weak bases and acids.
`
`For weak bases:
`pH= pK. + logto([B])/(BH+]) (S.2)
`and for weak acids: pH= pK. + log10([A-))/(HA)) (8.3)
`
`Equations 8 .2 and 8.3 can be used:
`l . to determine pK. by following changes in
`solubility
`2. to predict solubility at any pH, provided chat the
`intrinsic solubility (C0 ) and pK. are known
`3. to facilitate the selection of suitable salt-forming
`compounds and predict the solubility and pH
`properties of the salts.
`
`Albert and Serjeant (1 984) give a detailed account of
`how to obtain precise pK. values by potentil'.)metry,
`spectroscopy and conductivity.
`
`Salts
`
`A major improvement in solubility can be achieved
`by forming a salt. Acceptable pharmaceutical salt
`counter-ions are shown in Table 8.4. As an example,
`the consequence of changing chlordiazepoxide to
`various salt forms is shown in Table 8.5.
`In some cases, salts prepared from strong acids or
`bases are freely soluble but very hygroscopic. This
`does lead to instability in tablet or capsule formula(cid:173)
`tions, as some drug will dissolve in its own adsorbed
`films of moisture. It is often better to use a weaker
`acid or base to form the salt, provided any solubility
`requirements are met. A less soluble salt will gener-
`
`'j
`....J
`E
`
`C)
`
`..s
`&-
`:0
`:::,
`0
`Co~
`
`Self-association; complexation
`or solubilization b y impurities
`
`Pure - no interaction
`
`Suppression by common ion effect
`or salting out
`
`2
`
`4
`Drug / solvent phase rat io
`
`8
`
`Fig. 8.1 Effect of drug: solvent ratio on solubility when the drug Is Impure. Assuming the compound is a base and the estimate of its
`solubility in 0.1 M NaOH was 1 mg ml-1, four solutions of 3 ml should be set up containing 3, 6, 12 and 24 mg of drug. These give the
`phase ratios shown here. 3 ml is the smallest volume that can be manipulated for either centrifugation or filtration and dilution of UV
`analysis. The vials should be agitated continuously overnight and then the concentration in solution determined.
`
`116
`
`FRESENIUS EXHIBIT 1061
`Page 8 of 66
`
`
`
`PHARMACEUTICAL PREFORMULATION
`
`Table 8.4 Potential pharmaceutical salts
`
`Basic drugs
`
`Anion
`
`Hydrochloride
`
`Sulphate
`
`Mesylate
`
`Maleate
`
`Phosphate
`
`Saticylale
`
`Tart rate
`
`Lactate
`
`Citrate
`
`Succinate
`
`Acetate
`
`Others
`
`pK,,
`
`-6.10
`
`-3.00, + 1.96
`
`-1 .20
`1.92, 6.23
`
`2. 15. 7.20, 12.38
`
`3.00
`
`3 .00
`
`3 .10
`
`3.13, 4.76, 6.40
`
`4 .21. 5.64
`
`4 ,76
`
`-
`
`- - -
`•• Usage
`- - --
`43.0
`7,5
`
`2 .0
`3.0
`
`3.2
`
`0.9
`
`3.5
`
`0.8
`
`3.0
`
`0 .4
`
`1.3
`31 .4
`
`Acidic drugs
`
`Cation
`
`Potassium
`
`Sodium
`
`Lithium
`
`Calcium
`
`Magnesium
`
`Diethanotamine
`
`Zinc
`
`Choline
`
`Aluminium
`
`Others
`
`pK,
`
`16.00
`14 ,77
`
`13.82
`12.90
`
`11.42
`
`9.65
`
`8.96
`
`8.90
`
`5.00
`
`- - - - - -
`•• Usage
`- - - -
`10 .8
`
`62.0
`1.6
`
`10 .5
`
`1.3
`
`1.0
`
`3.0
`
`0.3
`
`07
`
`8.8
`
`Table 8 .5 Theoretical aolublllty and pH of salts of
`chlordiazepoxlde
`
`pK.
`
`Salt pH
`
`Solubility (mg mL- 1
`
`)
`
`Sall
`
`- - - -
`Base
`
`Hydrochloride
`
`Mateate
`
`Tanrate
`
`Benzoate
`
`4 .80
`-6.10
`
`1.92
`
`3.00
`
`4 .20
`
`4 .76
`
`830
`
`2 .53
`
`3.36
`
`2.0
`
`<165·'
`57. 1
`
`3.90
`4.50
`4,78
`
`17.9
`6.0
`4, 1
`
`Acetate 0
`- - - ----
`• Maximum solubility of chlordiazepoxide hydrochlonde.
`achieved at pH 2.89. is governed by crystal lattice energy
`and common ions.
`• Chlorctiazepoxide acetate m ay not form: pK. ol acetate
`too high a nd too close to that of drug ion.
`
`ally be less hygroscopic and form less acidic or basic
`solutions (Table 8.5). Injections shou ld ideally lie in
`the pH ran ge 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 as a propellant-acid reaction will
`corrode the canister.
`From Table 8.5, not only does the intrinsic pH of
`the base solution fall significantly if salt forms are
`produced but, as a consequence, the solubility
`
`increases exponentially (Eqns 8.2 and 8.3). T his has
`important implications in vivo. A weak base with an
`intrinsic solubility greater than l mg mL- 1 will be
`freely soluble in the gastrointestinal tract, especially
`in the stomach. However, it is usually better to for(cid:173)
`mulate with a salt, as it will control the pH of the dif(cid:173)
`fusion layer (the saturated solution immediately
`adjacent to the dissolving surface, known as the pH
`microenviroment). For example, although chlor(cid:173)
`diazepoxide base (C, = 2 mg mL- 1 at pH,., 8.3)
`meets the requirements for in vivo 'solubility'
`(Kaplan, 1972); commercial capsules contain chlor(cid:173)
`diazepoxide hydrochloride (C. = 165 mg mL-1 at
`pHsat 2.53).
`A weak base will have a high dissolution rate in the
`stomach, but as it moves down the gastrointestinal
`tract the pH rises and dissolution rate falls.
`Conversely, a weak acid has minimal dissolution in
`the stomach but becomes more soluble and dissolu (cid:173)
`tion rate increases down the gut. Paradoxically, as
`dissolution rate increases so absorption falls because
`the drug is ionized.
`The dissolution rate of a particular salt is usually
`much greater than that of the pa.rent drug. Sodium
`an d potassium salts of weak acids dissolve much
`more rapidly than do the parent acids, and some
`comparative data are shown in Table 8.6. O n the
`basis of bulk pH these salts would be expected to
`have
`lower dissolution rates in the scomach .
`However, the pH of the diffusion layer (found by
`measurin g the pH of a saturated bulk solution) is
`
`117
`
`FRESENIUS EXHIBIT 1061
`Page 9 of 66
`
`
`
`SCIENTIFIC PRINCIPLES OF DOSAGE FORM DESIGN
`
`higher than that of gastric fluid (which is approxi(cid:173)
`mately 1.5) because of its buffering action. The pH
`is the saturated unbuffered aqueous solution (calcu(cid:173)
`lated pH in Table 8.6) and the dissolution rate is gov(cid:173)
`erned by this pH and not the bulk medium 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.
`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 medium .
`With weak bases, their salts dissolve rapidly in the
`stomach but there is no absorption, as the drug is
`ionized and absorption is delayed until the intestine.
`Any undissolved drug, as salt, rapidly dissolves, as the
`higher diffusion layer pH compensates for the higher
`bulk pH, which would be extremely unfavourable to
`the free base. Data for chlordiazepoxide are shown in
`Table 8.5. The maleate salt has a predicted solubility
`of 57 mg mL-1 but, more importantly, reduces the
`pH by 5 units. By controlling diffusion layer pH the
`dissolution rate can increase manyfold, indepen(cid:173)
`dently of its position in the gastrointestinal tract. This
`is particularly important in the development of con(cid:173)
`trolled-release products.
`Different salts of a drug rarely change pharmacol(cid:173)
`ogy, but only physical properties. This statement has
`been qualified to acknowledge that salts do affect the
`intensity of response. However, the salt form does
`change the physiochemical properties of the drug.
`Changes in dissolu tion rate and solubility affect the
`rate and extent of absorption (bioavailability), and
`changes on hygroscopicity and stability influence
`formulation.
`Consequently each new drug candidate has to be
`examined to choose the most suitable salt, because
`
`each potential salt will behave differently and require
`separate preformulation screening. The regulat0ry
`authorities also treat each salt as a different chemical
`entity, par ticularly in the context of toxicity testing.
`
`Solvents
`
`It is generally necessary to formulate an m1ection
`even if there is no intention to market. T he first(cid:173)
`choice solvent is obviously water. However, although
`the drug may be freely soluble, it may be unstable in
`aqueous solution. Chlordiazepoxide HCl is such an
`example. Accordingly, water-miscible solvents are
`used:
`
`1. in formulations to improve solubility or stability
`2. in analysis to facilitate extraction and separation
`(e.g. 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 solvent solubility and stabil(cid:173)
`ity are unattainable Table 8. 7 shows a range of sol(cid:173)
`vents to fulfil these needs.
`Aqueous methanol is widely used in HPLC and is
`the standard solvent in sample extraction during
`analysis and stability testing. It is often made acidic
`or alkaline to increase solvent power and ensure con(cid:173)
`sistent ionic conditions for UV analysis. Other phar(cid:173)
`maceutical solvents are available but are generally
`only required in special cases. T he most acceptable
`non-aqueous solvents pharmaceutically are glycerol,
`propylene glycol and ethanol. Generally for a
`lipophilic drug (i.e. a partition coefficient (log P >I),
`solubility doubles through this series.
`Where bulk is limited an d the aqueous solubility is
`inadequate, it is better to measure the solubility in
`
`Table 8.6 Dissolution rates of weak acids and ttlelr sodium salts
`
`Drug
`
`pK,
`
`pH (at C,)
`
`Dissolution rate (mg cm-2 min-') x 102
`-
`-
`-
`- - - - -
`Dissolution media
`- - - - -
`Phosphate (pH 6.8)
`
`0.1 M HCI (pH 1.5)
`
`Saliqylic acid
`
`Sodium salicylate
`
`Benzoic acid
`
`Sodium benzoate
`
`Sulphathiazole
`
`Sodium sulpha1hiazole
`
`118
`
`J,0
`
`4.2
`
`7 .3
`
`2.'10
`
`8.78
`
`2.88
`9.35
`
`4.97
`
`10.75
`
`1.7
`
`1870
`
`2. 1
`
`980
`
`<0.1
`
`550
`
`27
`
`2500
`
`14
`
`1770
`
`05
`
`810
`
`FRESENIUS EXHIBIT 1061
`Page 10 of 66
`
`
`
`Table 8.7 Recommeoded solvents for preformulatlon screening
`
`PHARMACEUTICAL PREFORMULATION
`
`Solvent
`- - --
`Water
`
`Methanol
`
`Solubility parameter (.S)
`Application
`Dielectric constant ( E)
`- - - - - - -- - - - - - - - -- - - - - -- -- - ----- -
`80
`24.4
`A ll
`32
`
`14.7
`
`Extraction. separation
`
`-
`
`~·-
`
`-
`
`0.1 M HCI (pH 1.1)
`
`0. 1 M NaOH (pH 13.1)
`
`Buffer (pH 6-7)
`
`Ethanol
`
`Propylene glycol
`
`Glycerol
`
`PEG 300 or 400
`
`24
`32
`43
`35
`
`12.7
`
`12.6
`
`16.5
`
`aqueous solvent mixtures rather than in a pure
`organic solvent. Whereas solubilities at other levels
`and their mixtures can be predicted, the solubility in
`pure solvent is often inconsistent because of cosol(cid:173)
`vent effects. Furthermore, formulations rarely use
`pure non-aqueous solvent, particularly injections.
`For example, ethanol should only be used up to 10%
`in an injection to prevent haemolysis and pain at the
`injection site, and in clude isotonic salts.
`
`Partition coefficient ( K~)
`Partition coefficient (the solvent:water quotient of
`drug distribution) has a number of applications
`which are relevant to preformulation :
`
`I . Solubility: b oth aqueous and in mixed solvents
`2. Drug absorption in vivo: applied to a
`homologous series for structure activity
`relationships (SAR)
`3. Partition chromatography: choice of colum n
`(H PLC) or plate (TLC) and choice of mobile
`phase (eluant).
`
`Solvent solubility
`
`The relative polarities of solvents can be scaled using
`d ielectric constant (e), solubility parameter (8),
`interfacial ( -y) and hydrophilic- lipophilic balance
`(HLB). The best solvent in any given application is
`one whose polarity matches that of the solute; an
`ideal, fully compatible solution exists when 8,0._..,n, =
`c5,0 iu«· This can be ascertained by determining solu(cid:173)
`bility maxima, using a substituent contribution
`approach or the dielectric requirement of the system.
`T he most useful scale of polarity for a solute is ~
`(oil:water partition coefficient), as
`the other
`
`Dissolution (gas1ric). basic extraction
`
`Acidic extraction
`
`Dissolution (inlestinal)
`
`Formulation
`
`approaches do not allow easy estimates for the
`behaviour of crystalline solids. For a wide range of
`drugs it is possible to relate solvent solubility and the
`partition coefficient (log~ = log P). Yalkowsky and
`Roseman (1 981) derived the following expression
`for 48 drugs in propylene glycol:
`logC, = logCw + f (0.89 log P + 0.03)
`Equation 8.4 can be applied more generally by intro(cid:173)
`ducing a factor <I> to account for the relative solvent
`power of pharmaceutical solvents (see Table 8.8 for
`examples).
`For a wide range of solvents Eqn 8.4 now
`becomes:
`log c. = log C0 + J(log q> + 0.89 log P + 0.03)
`
`(8.4)
`
`(8.5)
`
`Methodology and structure activity prediction
`
`Choice of non-aqueous solvent (oil) T he oil:water
`partition (~) is a measure of the relative lipophilic(cid:173)
`ity (oil-loving) nature of a compound, usually in the
`unionized state (HA or B), between an aq ueous
`phase and an immiscible lipophilic solvent or oil.
`
`Tabla 8.8 Solvent power (cj,) of some pharmaceutical
`solvents
`
`- ------ ----- - - - -
`
`Relative solvent power(~,)
`-
`- -
`
`0.5
`
`Solvent
`
`Glycerol
`
`-
`
`Propylene glycol
`
`PEG 300 o r 400
`
`Ethanol
`
`ONA, DMF
`
`2
`
`4
`
`119
`
`FRESENIUS EXHIBIT 1061
`Page 11 of 66
`
`
`
`SCIENTIFIC PRINCIPLES OF DOSAGE FORM DESIGN
`
`Many partition solvents have been used. The largest
`database has been generated using n-octanol. The
`solubility parameter of octanol (o = 10.24) lies
`midway in the range for drugs (8- 12), although
`some non-polar (o < 7) and polar drugs (o > 13) are
`encountered. This allows measurable results between
`equal volumes of oil and aqueous phases.
`In the shake flask m.ethod the drug, dissolved in
`one of the phases, is shaken with the other partition(cid:173)
`ing 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- 1)
`is run off and centrifuged for 60 minutes at
`2000 rpm. The aqueous phase is assayed before (IC)
`and after partitioning (Cw) [the aqueous concentra(cid:173)
`tion] co give~= (~C - C.,,)/(Cw).
`If the transfer of solute to the oil phase is small,
`.1Cw is small, and any analytical error increases
`error in the estimate of I<!;,. Indeed, to encourage
`greater aqueous loss (>.:1Cw) a considerably more
`polar solvent, n-butanol, has been used. Where the
`partition coefficient is h igh, it is usual to reduce the
`ratio of the oil phase 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~= (10 IC - Cw)/C.,.).
`The partition of a polar solute between an inert
`non-polar hydrocarbon e.g. hexane, at'ld water is quite
`different from that of hydrogen bonding solvents such
`as octanol. The behaviour of the weak acid phenol
`(pK. = 10) and weak base nicotine (pK0 = 3.1) is
`worthy of note. For phenol, K::=•anol = 29.5, whereas
`[(!=• = 0.11. The acidic solvent chloroform sup(cid:173)
`presses partition (K;,':, = 2.239), whereas ethyl acetate
`and diethyl ether are more polar. The basic beha,..iour
`of the solvents give higher ~ - values. With solvents
`capable of both hydrogen donation and acceptance
`(octanol, nicrobenzene and oleyl alcohol),~ is inter(cid:173)
`mediate. For nicotine the behaviour is reversed, and
`the hydrogen donor (acidic) solvent chloroform parti(cid:173)
`tions most strongly ~ = 77.63), even though the
`neutral solvent nitrobenzene, which is marginally
`more lipophilic (log P = 1.87 against 1. 96 for chloro(cid:173)
`form), gives similar values for both phenol and nico(cid:173)
`tine. Clearly both solute and solvent characteristics
`are important.
`In general, polar solvents are advocated to corre(cid:173)
`late biological activity with physicochemical prop er(cid:173)
`ties. Solvents less polar than octanol, measured by
`water solvency, have been termed hyperdiscrirninat(cid:173)
`ing, whereas more polar solvents such as butanols
`and pentanols, are hypodiscriminating. This concept
`refers to the discriminating power of a partitioning
`solvent within a homologous series. With n-butanol
`
`120
`
`the values of log P tend to be close, whereas with
`heptane and other inert hydrocarbons the differences
`in solute lipophilicities are exaggerated. n-Octanol
`generally gives a range consistent with other physico(cid:173)
`chernical properties when compared to drug absorp(cid:173)
`tion in the GI tract. Hyperdiscriminating solvents
`reflect more closely the
`transport across the
`blood- brai