`
`Salt Selection and Optimisation Procedures for Pharmaceutical New Chemical
`Entities
`
`Richard J. Bastin,† Michael J. Bowker,*,†,§ and Bryan J. Slater‡
`Preformulation Department, Pharmaceutical Sciences, AVentis Pharma, Dagenham Research Centre (DRC), Rainham
`Road South, Dagenham, Essex RM10 7XS, UK, and World-Wide Physical Chemistry Department, DiscoVery Chemistry,
`AVentis Pharma, Dagenham Research Centre (DRC), Rainham Road South, Dagenham, Essex RM10 7XS, UK
`
`Abstract:
`Selection of an appropriate salt form for a new chemical entity
`provides the pharmaceutical chemist and formulation scientist
`with the opportunity to modify the characteristics of the
`potential drug substance and to permit the development of
`dosage forms with good bioavailability, stability, manufactur-
`ability, and patient compliance. Salts are most commonly
`employed for modifying aqueous solubility, however the salt
`form selected will influence a range of other properties such as
`melting point, hygroscopicity, chemical stability, dissolution
`rate, solution pH, crystal form, and mechanical properties.
`Where possible, a range of salts should be prepared for each
`new substance and their properties compared during a suitable
`preformulation program. Since it is normally possible to fully
`develop only one salt form, its properties should be appropriate
`to the primary route of administration and dosage form. An
`understanding of the influence of drug and salt properties on
`the finished product is essential to ensure selection of the best
`salt. The drug properties required for one dosage form may
`be quite different from those required for another. A well
`designed salt selection and optimisation study provides a sound
`base on which to build a rapid and economic product develop-
`ment programme.
`
`Introduction
`Modern drug discovery processes involve the screening
`of vast numbers of compounds that may have been made by
`the Company’s research laboratories over many years. Added
`to these may be the many thousands of compounds that have
`been manufactured as libraries of structurally related series
`by “combinatorial chemistry” techniques. All of these
`compounds are generally dissolved in dimethylsulphoxide
`(DMSO) solution and screened in an enzyme- or receptor-
`based assay system. If the number of “hits” produced is large,
`the numbers are usually refined by further screening and
`selection until a manageable number of “leads” is available.
`Many of these leads will show only weak or moderate
`activity and further refinement and optimisation is invariably
`necessary. These optimisation procedures usually involve
`numerous structural modifications, aided by computational
`techniques, until a small number (usually 1-5) of highly
`active “candidates” remain.
`
`* To whom correspondence should be sent.
`† Preformulation Department, Pharmaceutical Sciences.
`‡ World-Wide Physical Chemistry Department, Discovery Chemistry.
`§ Current address: M. J. Bowker Consulting Ltd., 36, Burses Way, Hutton,
`Brentwood, Essex CM13 2PS, UK
`
`These candidates are usually free bases, free acids, or
`neutral molecules, rather than their salts. Also, because of
`the generally higher molecular weights of modern drug
`substances and the increased use of DMSO solutions in the
`screening processes, it is becoming apparent that there is a
`tendency towards ever more lipophilic candidates being
`presented. Frequently, when first proposed as potential
`development candidates,
`they are often amorphous or
`partially crystalline as little effort has been made to
`investigate formal crystallisation procedures. The need for
`water-soluble candidates has been recognised1-4 for many
`years before the advent of ‘combinatorial chemistry.
`
`Investigations into the Possibilities of Salt Formation
`When first presented for initial preformulation investiga-
`tions, normally the amount of drug substance available from
`Discovery Chemistry rarely exceeds 1 g. To maximize the
`amount of data gained from such small quantities, semi-micro
`techniques have been developed and are used regularly within
`our groups. Invariably, the first information generated for
`each candidate is the calculated pKa value of each ionisable
`group in the molecule.5-8 This is quickly checked against
`the value determined experimentally on 1-2 mg of sample
`by potentiometric titration (e.g., Sirius Model GLpKa ap-
`paratus, Sirius Analytical Instruments Ltd.). Knowledge of
`the pKa value enables potential salt forming agents (coun-
`terions) to be selected, for each candidate, based on lists that
`are available in the literature.2,9-11 For the formation of a
`stable salt, it is widely accepted that there should be a
`minimum difference of about 3 units between the pKa value
`
`(1) Hirsch, C. A.; Messenger, R. J.; Brannon, J. L. J. Pharm. Sci. 1978, 67,
`231.
`(2) Gould, P. L. Int. J. Pharm. 1986, 33, 201.
`(3) Morris, K. R.; Fakes, M. G.; Thakur, A. B.; Newman, A. W.; Singh, A.
`K.; Venit, J. J.; Spagnuolo, C. J.; Serajuddin, A. T. M. Int. J. Pharm. 1994,
`105, 201.
`(4) Anderson, B. D.; Flora, K. P. In The Practice of Medicinal Chemistry;
`Wermuth, C. G., Ed.; Academic Press Ltd., 1996; Chapter 34.
`(5) Remington: The Science and Practice of Pharmacy, 19th ed.; Gennaro,
`A. R., Ed.; Mack Publishing Co.: Easton, Pennsylvania, 1993; Vol. II, p
`1456.
`(6) Hammett, L. P. Chem. ReV. 1935, 17, 125.
`(7) Perrin, D. D.; Dempsey, B.; Serjeant, E. P. pKa Prediction for Organic
`Acids and Bases; Chapman and Hall: London, 1981.
`(8) Albert, A. A.; Serjeant, E. P. Ionization Constants of Acids and Bases;
`Wiley: New York, 1984.
`(9) Wells, J. I. Pharmaceutical Preformulation, 2nd ed.; Ellis Horwood:
`Chichester, 1993; p 29.
`(10) Martindale, W. In The Extra Pharmacopoeia, 30th ed.; Reynolds, J. E. F.,
`Ed.; The Pharmaceutical Press: London, 1993.
`(11) Berge, S. M.; Bighley, L. D.; Monkhouse, D. C. J. Pharm. Sci. 1977, 66,
`1.
`
`10.1021/op000018u CCC: $19.00 © 2000 American Chemical Society and The Royal Society of Chemistry
`Published on Web 07/19/2000
`
`Vol. 4, No. 5, 2000 / Organic Process Research & Development
`
`¥
`
`427
`
`Downloaded by Scott Petersen at 09:18:50:551 on June 04, 2019
`
`from https://pubs.acs.org/doi/10.1021/op000018u.
`
`Mylan (IPR2020-00040) Ex. 1006, p. 001
`
`
`
`Table1.Classification of common pharmaceutical salts
`salt class
`
`examples
`
`inorganic acids
`sulfonic acids
`carboxylic acids
`anionic amino acids
`hydroxyacids
`fatty acids
`insoluble salts
`
`organic amines
`insoluble salts
`metallic
`cationic amino acids
`
`Anions
`hydrochloride, hydrobromide, sulfate, nitrate, phosphate
`mesylate,b esylate,c isethionate,d tosylate,e napsylate,f besylateg
`acetate, propionate, maleate, benzoate, salicylate, fumarate
`glutamate, aspartate
`Citrate, lactate, succinate, tartrate, glycollate
`hexanoate, octanoate, decanoate, oleate, stearate
`pamoate (embonate), polystyrene sulfonate (resinate)
`Cations
`triethylamine, ethanolamine, triethanolamine, meglumine, ethylenediamine, choline
`procaine, benzathine
`sodium, potassium, calcium, magnesium, zinc
`arginine, lysine, histidine
`
`a Based on data from various sources.9-11 b Methane sulfonate. c Ethane sulfonate. d 2-Hydroxyethane sulfonate. e Toluene sulfonate. f Naphthalene sulfonate. g Benzene
`sulfonate.
`
`of the group and that of its counterion, especially when the
`drug substance is a particularly weak acid or base. Occasion-
`ally, exceptions may be found where a salt has an acceptable
`stability, despite there being a smaller difference in the pKa
`values.
`A microplate technique has been developed for the
`screening of salts; this involves dissolving approximately 50
`mg of sample in a suitable, volatile solvent and adding a
`fixed volume of this solution, containing about 0.5 mg of
`sample, into each microplate well. Concentrated solutions
`of each potential counterion in equimolar proportion, or other
`appropriate stoichiometric ratio, are prepared and a few
`microlitres of each is added sequentially to each well. Thus,
`all of the wells in line 1 (x-direction) will contain the same
`combination of sample and counterion 1; all of the wells in
`line 2 contain the same combination of sample and counte-
`rion 2, etc. Different, potential crystallising solvents can be
`investigated methodically in the y-direction. The wells are
`inspected using an inverted microscope (Leica, Model
`DMIRB) at regular intervals for the appearance of crystals.
`Occasionally, crystallisation can be promoted by evaporation
`of any excess solvent in some wells using a slow stream of
`dry nitrogen gas.
`Once the combinations of counterion and solvent(s) are
`identified, studies at a slightly larger scale (usually 10-50
`mg, occasionally up to 500 mg) can be initiated to confirm
`the suitability and viability of the crystals. These studies can
`help identify problems with low melting points, determined
`by hot-stage microscopy, and hygroscopicity, if processed
`on a suitable apparatus (e.g., Dynamic Vapour Sorption
`Analyser, model DVS-1, Surface Measurement Systems
`Ltd.). Frequently these studies can also give preliminary
`information on the existence of solvates and hydrates,
`especially if differential scanning calorimetry (DSC, Mettler
`Toledo DSC, model 820),
`thermal gravimetric analysis
`(TGA, Mettler Toledo TGA, model 850) and hot-stage
`microscopy are also used in the evaluation process.
`In parallel with these studies, a preliminary high perfor-
`mance liquid chromatographic (HPLC) method is quickly
`developed to give an estimate of the purity of the sample,
`whilst infrared and other spectroscopic techniques may be
`
`428
`
`¥
`
`Vol. 4, No. 5, 2000 / Organic Process Research & Development
`
`used to define the salt and the stoichiometry. Knowledge of
`the approximate purity is important at this stage as the
`presence of high levels of some impurities can often hinder
`crystallisation or alter the polymorphic form obtained.
`Therefore, from these preliminary, small-scale studies, a
`range of potential salt formers and recrysallisation solvents
`can be quickly identified. Following further scale-up to gram
`quantities, more comprehensive data can be obtained to
`evaluate their suitability for use in formulations.
`
`Choice of the Salt Former
`Although the choice of salt is governed largely by the
`acidity or basicity of the ionisable group, safety of the
`counterion, drug indications, route of administration and the
`intended dosage form must also be considered. Toxicological
`and pharmacological implications of the selected salt former
`must be considered as well as the effects of the parent drug.
`Salt formers can be subdivided into a number of categories,
`depending upon their functionality and purpose. Some of the
`most frequently used examples are listed in Table 1.
`The vast majority of salts are developed to enhance the
`aqueous solubility of drug substances. For weakly basic drug
`substances, salts of an inorganic acid (e.g., hydrochloride,
`sulphate, or phosphate), a sulphonic acid (mesylate or
`isethionate), a carboxylic acid (acetate, maleate or fumarate),
`a hydroxyacid (citrate or tartrate), or possibly an amino acid
`(arginine or lysine) could be considered. Hydrochloride salts
`have often been the first choice for weakly basic drugs, since
`as a consequence of the low counterion pKa, salts can nearly
`always be formed, and recrystallisation from organic solvents
`is normally straightforward. However, the potential disad-
`vantages of hydrochloride salts may include unacceptably
`high acidity in formulations (e.g., parenteral products), the
`risk of corrosion, less than optimal solubility due to the risk
`of salting out and the potential for poor stability if the drug
`is acid labile and hygroscopic.2
`Occasionally, salts may be also prepared to decrease drug
`substance solubility for use in suspension formulations where
`very low solubility is necessary to prevent “Ostwald ripen-
`ing”, for taste-masking, or to prepare an extended release
`product. Embonate salts have been used in suspension
`
`Mylan (IPR2020-00040) Ex. 1006, p. 002
`
`
`
`Table2.a Preformulation studies that are normally considered for comparison of salt forms and parent compound for oral
`dosage forms
`
`test
`
`suitable techniques
`
`comments
`
`dissociation constant and
`basic physico-chemical
`properties
`melting point
`
`aqueous solubility
`
`pH of solution
`
`cosolvent solubility
`
`common ion effect on solubility
`
`hygroscopicity
`
`potentiometry, solubility,
`UV spectroscopy
`
`capillary m.pt., hot stage microscopy,
`differential scanning calorimetry
`overnight equilibration at 25 (cid:176) C; analysis
`by UV spectroscopy or HPLC
`
`overnight equilibration at 25 (cid:176) C, analysis
`by UV spectroscopy or HPLC
`overnight equilibration at 25 (cid:176) C
`in suitable media and analysis
`by UV spectroscopy or HPLC
`use DVS apparatus or expose
`to various RH values and
`measure weight gain after 1 week
`
`intrinsic dissolution rate
`
`use Wood’s apparatus14
`
`crystal shape and appearance
`
`SEM or optical microscopy
`
`particle size
`polymorphism/pseudopolymorphism
`powder properties
`stability
`
`SEM and laser diffraction
`recrystallizations, HSM, DSC, TGA
`bulk density measurement
`various
`
`determine pKa for parent drug
`
`perform on each salt and
`compare to parent
`Perform on each salt and
`compare to parent
`Examine pH of saturated solution
`if quantities permit.
`Determine solubilities in ethanol,
`poly(ethylene glycol), propylene glycol
`and glycerol and compare to parent.
`compare solubility in demineralized
`water with 1.2% NaCl for salts and parent
`
`perform at 53, 93, and 97% RH,
`and other values of interest;
`assign hygroscopicity
`classification to each salt13
`compare dissolution rates at various pHs
`(can provide data on wettability)
`Compare crystal habits and
`levels of agglomeration
`Examine particle size distributions.
`preliminary exploration
`determine Carr’s compressibility index
`perform on parent drug and undertake
`preliminary tests on appropriate salts
`
`formulations to increase the duration of action (e.g., chlor-
`promazine embonate). On some occasions, the selection of
`a salt with only modest aqueous solubility may be more
`suitable for use in tablet products prepared by wet granulation
`since the use of highly soluble salts can be detrimental to
`the granulation process. Depending on the dose required,
`aqueous solubilities in the range 0.1-1.0 mg/mL will
`normally be sufficient to satisfy the dissolution requirements
`for standard, solid, oral dosage forms of drugs with good to
`moderate potency. However, for parenteral solution products,
`higher solubilities, perhaps 10 mg/mL or greater, depending
`on the required dose and dose volume, may be required. For
`parenteral formulations, the pH of solution (normally within
`an acceptable range of 3-10 for intravenous solution) should
`be monitored to help ensure that the formulation will be well
`tolerated.
`Salts are also frequently prepared for the reasons other
`than solubility modification; it is frequently necessary to
`prepare a specific salt to either achieve adequate physical
`stability or for taste masking (e.g., dextropropoxyphene
`napsylate suspension). Manipulation of drug substance
`solubility by selection of salts may also be employed to
`modify the pharmacokinetic profile of the drug (e.g.,
`benzathine penicillin and insulin zinc complexes used in
`parenteral formulations). Salt formation may be also advan-
`tageous where the melting point of the active moiety is low,
`and it is necessary to mill or micronise the active ingredient
`to achieve adequate homogeneity. A suitably stable salt may
`have a melting point that is 50-100 (cid:176) C higher than the free
`acid or free base. Also, being more ionic, the crystals are
`
`likely to be less plastic and more easily deformed by brittle
`fracture.
`
`Scale-up of the Formation of Salts
`The information from the preliminary crystallisation
`studies is communicated to the Process Chemistry group,
`who by this time will have started their investigations into
`possible manufacturing routes for each of the candidates
`remaining. At this stage in the development process, Process
`Chemistry usually aim to quickly manufacture 50-200 g of
`the one or two candidates that may remain to progress them
`towards initial clinical evaluation. The manufacturing route
`may be the same as used by the Discovery Chemistry group
`but usually is significantly different. The aims of both the
`Process Chemistry and Preformulation groups for the fol-
`lowing 12-18 months is to collaborate extensively to ensure
`that, for the chosen candidate, there will be a viable synthetic
`route to the chosen form of the drug substance.
`A significant portion of this batch is destined for the
`preparation of 3-4 g of each of the salts that were thought
`to be viable from the smaller-scale studies. A similar sized
`portion of the free base/acid is also taken for comparison
`purposes. The combination of individual studies undertaken
`on each of these 3-4 g portions varies depending on the
`type(s) of dosage form ultimately required for marketing.
`Occasionally, it may be necessary to undertake a pharma-
`cokinetic evaluation of each salt in comparison with the free
`acid/base. The dosage forms most commonly used for the
`drug substances encountered during preliminary clinical
`investigations are tablets/capsules, inhalation dosage forms
`and injections.
`
`Vol. 4, No. 5, 2000 / Organic Process Research & Development
`
`¥
`
`429
`
`Mylan (IPR2020-00040) Ex. 1006, p. 003
`
`
`
`Table3.Tests to be considered for the evaluation of
`candidate salts
`
`test to be considered
`
`amount required, mg
`
`Structural Analysis
`
`mass spectroscopya
`1H NMRa
`13C NMRa
`Ir spectrum
`UV spectrum
`fluorescence spectruma
`elemental analysis
`Physicochemical Properties
`
`melting range
`pKa
`a
`C log P/log P a
`preliminary polymorphism study
`X-ray diffraction
`aqueous solubilityb
`pH - solubility profile
`cosolvent solubilitiesc
`propellant solubilityd
`
`Physical Properties
`
`hygroscopicity
`microscopy (SEM/optical)
`particle size (Malvern)
`size reduction (sonication)
`Impurities (hplc)
`
`related substancesa
`degradation productsa
`chiral puritya
`
`Stability Studies
`stability to hydrolysis (pH 2, 7, 10)a
`stability to oxidation (peroxide/peracid)a
`stability to photolysisa
`
`1
`5
`25
`1
`1
`1
`10
`
`2
`5
`5
`200-500
`20
`100
`500
`300
`500
`
`20
`10
`100
`300
`
`10
`10
`10
`
`15
`15
`15
`
`a Determined on free acid/base only. b Would include solubility in saline, 5%
`dextrose and some buffers c Also solubilities in complexing agents/surfactant
`systems where appropriate d Propellants and propellant/cosolvent systems for
`inhalation dosage forms.
`
`Tables 2 and 3 show the types of tests normally chosen,
`the information that they can produce and the amount of
`sample normally required for these common dosage forms.
`
`What to Develop: Salt or Free Acid/Base?
`The results obtained from each of these tests are tabulated
`for the free acid/base, together with each of the salts, and
`discussed in detail between the Formulation Scientists,
`Preformulation Analysts, Physical Chemists, Process Chem-
`ists, and occasionally Pharmacokineticists. The Preformu-
`lation Scientists assess the relative merits of each form for
`use in the proposed clinical formulations and whether the
`properties such as solubility are adequate to give the high
`concentrations required in the various pre-clinical formula-
`tions. Process Chemistry need to assess the likely yield of
`each salt, as salt formation creates an additional step in the
`manufacturing process. Usually, the decision-making process
`results in the proposal of a single salt for further study,
`although occasionally it is seen that none of the salts have
`optimum properties, and two different salts can be proposed
`for in-depth study. Also, it is occasionally found that the
`overall properties of the free acid/base are much better than
`any of the salts. This occurs more frequently where the
`
`430
`
`¥
`
`Vol. 4, No. 5, 2000 / Organic Process Research & Development
`
`candidate has a low pKa value and the resulting salts are
`less stable than required or when the salts are particularly
`hygroscopic or when they exhibit complex polymorphism/
`pseudopolymorphism (hydration or solvation).
`These relatively simple investigations give much useful
`information very quickly; it should be noted, however, that
`the preliminary polymorphism study is far from the in-depth
`study that is always undertaken later. This preliminary study
`uses a range of protic and aprotic solvents of widely differing
`polarity and will normally show the presence of a stable
`hydrate or solvate.
`Once a decision is agreed upon within the group, a
`document that gives a pre´cis of the discussions and the basis
`for the proposal is normally drafted for agreement by senior
`management. Examples of these salt selection studies are
`given below:
`
`Example No. 1 (RPR 111423)
`RPR 111423 is a candidate drug substance that has been
`evaluated for the treatment of symptoms related to infection
`by AIDS. It is a crystalline, very weak base with a pKa at
`4.25. A comprehensive screening of possible salts demon-
`strated only a monohydrochloride (RPR 111423A) and a
`mesylate (RPR 111423B) could be isolated as crystalline
`solids.
`It was decided that the free base should be taken through
`the simple evaluation process in comparison with these two
`salts. It was expected that the drug substance could be
`required in the form of tablets or capsules, with an injection
`form needed for some pre-clinical studies and for the
`determination of absolute bioavailability in man. Because
`of its high activity in screening studies, there was a possibility
`that very low dose oral formulations might be needed. This
`may require micronised drug substance to enable content
`uniformity requirements to be met; this micronised material
`would also be expected to enhance dissolution.
`The results from the relatively simple studies undertaken
`are given in Table 4. The two salts clearly demonstrated the
`predictable problems associated with a relatively low pKa
`value; the salts were quite weak and dissociated to liberate
`the free base in media with pH values below the pKa. The
`very low solubility of the free base resulted in immediate
`precipitation following dissociation. There was clear evidence
`for multiple polymorphism for each of the salts, and
`establishing the existence of a stable polymorph, or a suitable
`pseudopolymorph, may have been necessary before a deci-
`sion could be made on which of the two salts could be
`developed further.
`The corresponding results for the free base indicated that
`it appeared to be the better candidate; it showed no evidence
`of polymorphism, and it was not hygroscopic. The two major
`areas that required further investigation were whether it had
`sufficient solubility in gastrointestinal media and whether it
`could be micronised. Studies performed on samples of drug
`substance and on simple capsule formulations demonstrated
`that
`the dissolution rates of micronised free base were
`equivalent or superior to those of the salts under the same
`conditions.
`
`Mylan (IPR2020-00040) Ex. 1006, p. 004
`
`
`
`Table4.Comparison of some simple properties of RPR111423 and its two salts
`result for
`result for
`RPR 111423 (base)
`RPR 111423A (hydrochloride)
`
`test
`
`off-white to cream,
`crystalline powder
`10-100 (large
`rhombic crystals)
`241-244
`no other form detected
`
`pale yellow, highly
`agglomerated powder
`2 (cid:2) 1
`(microcrystalline laths)
`242
`at least four polymorphs detected;
`metastable forms revert to
`original on standing
`
`nothing detected
`at 25 (cid:176) C
`at 37 (cid:176) C
`
`loss of HCl detected
`at 110-120 (cid:176) C
`at 25 (cid:176) C
`
`at 37 (cid:176) C
`
`11.6
`0.71
`0.03
`0.01
`0.01
`0.01
`
`6.50
`
`14.7
`0.89
`0.05
`0.02
`0.02
`0.02
`
`25.7
`2.51
`0.05
`0.01
`0.01
`0.36
`
`2.43
`
`28.2
`4.58
`0.13
`0.02
`0.02
`0.99
`
`result for
`RPR 111423B (mesylate)
`
`cream to pale yellow, highly
`agglomerated powder
`7 (cid:2) 1
`(microcrystalline laths)
`210
`at least six polymorphs detected;
`phase changes detected on
`grinding or micronisation;
`reverts to original form on heating
`nothing detected
`at 25 (cid:176) C
`
`at 37 (cid:176) C
`
`131.4
`6.11
`0.01
`0.03
`0.01
`0.33
`
`2.74
`
`204.1
`8.91
`0.02
`0.34
`0.02
`0.50
`
`appearance
`
`particle size
`by microscopy, (cid:237)m
`melting range, (cid:176) C
`preliminary
`polymorphism study
`
`other thermal behavior
`
`aqueous solubility,
`mg/mL
`- at pH 1
`- at pH 2
`- at pH 4
`- at pH 6
`- at pH 6,8
`- in demineralized water
`pH of saturated
`solution, at 20 (cid:176) C,
`in water
`addition of
`water to concentrate
`- at pH 2
`- at pH 4
`hygroscopicity
`(hygrostat for 14 days)
`
`no changes detected
`no changes detected
`non-hygroscopic
`<0.2%w/w water
`uptake at any RH
`
`some precipitation of free base
`extensive precipitation of free base
`slightly hygroscopic
`2.3% w/w uptake at 53% RH
`22% w/w uptake at 97% RH
`
`some precipitation of free base
`extensive precipitation of free base
`moderately hygroscopic
`3.7% w/w uptake at 53% RH
`32% w/w uptake at 97% RH
`
`Example No. 2 (RPR 127963)
`RPR 127963 is a candidate drug substance that has been
`evaluated for the treatment of cardiovascular diseases; it is
`a crystalline, very weak base with a pKa at 4.10. In common
`with most similar drug substances intended for the treatment
`of cardiovascular disease, it was considered that a high-dose
`(up to 250 mg) solid, oral dosage form and a correspondingly
`high-dose (up to 50 mg/mL) injection would be ultimately
`required. In line with our standard protocol, a comprehensive
`evaluation of possible salts was undertaken, and this dem-
`onstrated that five crystalline salts (a hydrochloride, a
`mesylate, a citrate, a tartrate, and a sulphate) could be readily
`produced. It was decided to quickly profile each of these
`salts in comparison with the free base. The results of these
`studies are given in Table 5.
`When the anhydrous free base was evaluated, the exist-
`ence of an additional mono-, di-, and trihydrate was found
`quite rapidly. It was shown that all four of these forms could
`be interconverted under conditions that might be expected
`to be found in granulation processing. The other potential
`problem with the anhydrate was the low melting point. In
`considering the results obtained for the various salts, the
`solubilities of the citrate and the tartrate were much lower
`than required for an injectable form and lower than ideal
`for high dosage formulations. An additional problem for the
`tartrate salt was the high hygroscopicity. Both of these salts
`were rejected before completion of the full evaluation. The
`hydrochloride salt was also shown to have several problems
`such as lower than ideal solubility, probable multiple
`polymorphism, and the formation of hydrates.
`
`Thus, the mesylate and the sulphate were the two salts
`that remained; both had high melting points, excellent
`aqueous solubility, and were non-hygroscopic. The free base
`still remained a possible candidate, if a stable hydrate could
`be found. It was therefore decided to undertake some
`additional evaluations on these three forms; the results from
`these are presented in Table 6.
`These additional results demonstrate a slight advantage
`in favour of the sulphate salt because of its greater solubility
`in cosolvents. This would give the formulator a better chance
`of achieving a higher dose in an injectable formulation. It
`was considered that the sulphate salt (RPR 127963E) could
`be studied further in the more detailed evaluations that would
`follow over the next few months. The mesylate or the free
`base (if a suitably stable hydrate could be found) would
`provide a possible back-up, should unforeseen problems
`arise.
`
`Example No. 3 (RPR 200765)
`RPR200765 is a candidate drug substance proposed for
`the treatment of rheumatoid arthritis. It is another crystalline,
`weak base with a pKa of 5.3 which formed salts with a wide
`selection of counterions. It was expected that doses of 100-
`125 mg of RPR200765 in capsules would be required for
`clinical studies.
`Early studies suggested that RPR200765 free base was
`unacceptable for use in solid, oral dosage forms due to a
`very poor aqueous solubility of approximately 10 (cid:237)g/mL and
`poor bioavailability in animal models. However, RPR200765
`would form stable salts with hydrochloride, hydrobromide,
`
`Vol. 4, No. 5, 2000 / Organic Process Research & Development
`
`¥
`
`431
`
`Mylan (IPR2020-00040) Ex. 1006, p. 005
`
`
`
`Table5.Comparison of some simple properties of RPR127963 and its five salts
`
`result for
`free base
`(RPR 127963)
`
`result for
`HCl salt
`(RPR 127963A)
`
`result for
`mesylate salt
`(RPR 127963B)
`
`result for
`citrate salt
`(RPR 127963C)
`
`result for
`tartrate salt
`(RPR 127963D)
`
`result for
`sulfate salt
`(RPR 127963E)
`
`test
`
`appearance
`
`particle size
`(microscopy), (cid:237)m
`
`yellow,
`crystalline
`powder
`1-3 (cid:237)m
`(agglomerates
`of microcrystals)
`
`yellow,
`crystalline
`powder
`1-3 (cid:237)m
`(agglomerates
`of microcrystals)
`
`yellow,
`crystalline
`powder
`tightly packed
`spherulites of
`agglomerated
`microcrystals
`18 (cid:237)m diameter.
`280.9-282.2
`
`yellow,
`crystalline
`powder
`microcrystals
`(2-3 (cid:237)m)
`with some
`aggregates
`(70 (cid:237)m)
`130.2-134.3
`
`yellow,
`crystalline
`powder
`rounded
`agglomerates of
`microcrystals in
`domains (70 (cid:237)m)
`198.5-201.6
`
`yellow,
`crystalline
`powder
`aggregates of
`microcrystals
`(10-15 (cid:237)m)
`
`305.7-308.9
`
`no evidence
`of polymorphs
`
`stable
`hemihydrate
`detected
`
`unstable
`anhydrate
`
`no evidence of
`polymorphs
`
`166-191 (re-grows
`at about 166,
`recrystallizes at
`191, then melts at
`about 275
`two monohydrates
`and one anhydrate
`
`3.92
`5.2
`0.019
`2.84
`2.33
`non-hygroscopic
`
`108
`50.4
`0.022
`90
`1.76
`non-hygroscopic
`
`0.83
`n.d.
`n.d.
`n.d.
`2.49
`non-hygroscopic
`
`0.89
`n.d.
`n.d.
`n.d.
`2.56
`very
`hygroscopic
`
`(cid:24)50
`5.9
`0.018
`(cid:24)40
`1.32
`non-hygroscopic
`
`melting range, (cid:176) C
`
`119-123
`
`preliminary
`polymorphism study
`
`several hydrates
`detected
`
`aqueous solubility
`(25 (cid:176) C), mg/mL
`- in demineralized water
`- in 0.1 M HCl
`- in 0.1 M NaOH
`- in dextrose 5%w/v
`pH of saturated solution
`hygroscopicity
`
`n.d.a
`n.d.
`0.020
`n.d.
`n.d.
`n.d.
`
`a n.d. ) Not determined.
`
`Table6.Comparison of additional properties of RPR127963 (anhydrate), its mesylate (RPR 127963B) and sulfate (RPR
`12963E) salts
`
`test
`
`result for free base
`anhydrate (RPR 127963)
`
`result for mesylate
`salt (RPR 127963)
`
`result for sulfate
`salt (RPR 127963)
`
`solubility in
`cosolvents at
`25 (cid:176) C, mg/mL
`ethanol
`propylene glycol
`poly(ethylene glycol) 400
`dimethylsulphoxide
`N-methylpyrrolidone
`glycerol
`peanut oil
`intrinsic dissolution
`rate, mg(cid:226)min-1(cid:226)cm-2
`- in water
`- in 0.01 M HCl
`powder flow
`properties
`
`a n.d. ) Not determined.
`
`190
`35.4
`188
`> 500
`> 400
`42
`0.18
`
`0.01
`0.35
`n.d.
`
`0.6
`0.7
`0.2
`14
`4.4
`n.d.a
`none detected
`
`n.d.
`7.3
`Good, but becomes much worse
`with increasing humidity
`
`0.2
`1.7
`0.2
`110
`8.5
`2.7
`none detected
`
`n.d.
`7.7
`Sticks slightly
`
`methanesulfonate, and camphorsulfonate counterions. Aque-
`ous solubility, particle size and shape, powder properties,
`and polymorphism profile were considered to be the key
`properties to permit a choice of salt to be made. In addition,
`it was recognised that the use of some counterions with high
`molecular weights, would require a large excess of drug
`substance to achieve the required doses.
`Studies demonstrated that the solubility of RPR200765
`depended on the amount of drug substance used for the study.
`This occurred because the counterion reduced the pH of
`solution and enhanced solubility of the drug base. The
`mesylate salt consistently produced a higher solubility than
`any of the other salt forms. The higher solubility resulted in
`
`432
`
`¥
`
`Vol. 4, No. 5, 2000 / Organic Process Research & Development
`
`an enhanced dissolution rate of the mesylate salt compared
`to the other salt forms. The solubility and dissolution rate
`of the hydrobromide salt was particularly poor. Intrinsic
`dissolution rate studies on compressed disks could not be
`carried out because a good compact could not be obtained
`for most of the salts, and the studies were carried out using
`drug powder (equivalent to 50 mg free base) in capsules.
`Hygroscopicity studies demonstrated that the hydrochlo-
`ride and hydrobromide salts adsorbed large amounts of
`moisture on exposure to humidity, resulting in the formation
`of multiple hydrated forms. The methanesulfonate salt
`however, was a stable monohydrate form which lost moisture
`at very low humidity (<10% relative humidity (RH)) but
`
`Mylan (IPR2020-00040) Ex. 1006, p. 006
`
`
`
`Table7.Comparison of the physicochemical properties of RPR200765 salt forms
`result for mesylate salt
`result for camphorsulfonate
`result for hydrochloride
`(RPR 200765A)
`salt (RPR 200765C)
`salt (RPR 200765D)
`
`test
`
`appearance
`
`off-white to cream,
`free-flowing powder
`
`white to off-white, crystalline,
`free-flowing powder
`
`white, free-flowing
`powder
`
`result for hydrobromide
`(RPR200765E)
`
`white to off-white,
`crystalline, free-flowing
`powder
`569.43
`276-277
`3.29
`
`2.63
`
`hygroscopic with
`multiple hydrated
`forms
`loosely
`agglomerated,
`flaky material
`10-40 (cid:237)m
`particles
`
`566.61
`214
`39
`
`1.93