m
`Polymor
`ph"
`Molecular Crystals
`
`JOEL BERNSTEIN
`
`Department of Chemist~-y
`Ben-Gurion University of the Negev
`
`CLARENDON PRESS ® OXFORD
`2002
`
`Lupin Ex. 1044 (Page 1 of 19)
`
`

`

`OXFORD
`
`UNIVEI(SITY PRESS
`Great Clarendon Street, Oxford ox2 6DP
`Oxford University Press is a department of the University of Oxford.
`It furthers the University’s objective of excellence in research, scholarship,
`and education by publishing worldwide in
`Oxford New York
`Auckland Bangkok Buenos Aires Cape Town Chennai
`Dar es Salaam Delhi Hong Kong Istanbul Karachi Kolkata
`Kuala Lumpur Madrid Melbourne Mexico City Mmnbai Nairobi
`Silo Paulo Shanghai Taipei Tokyo Toronto
`
`Oxford is a registered trade mark of Oxford University Press
`in the UI( and in certain other countries
`
`Published in the United States
`by Oxford University Press Inc., New York
`© Oxford University Press, 2002
`The moral rights of the author have been asserted
`Database right Oxford University Press (maker)
`
`First published 2002
`Reprinted 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,
`without the prior permission in w~’iting of Oxford University Press,
`or as expressly permitted by law, or under terms agreed with the appropriate
`reprographics rights organization. Enquiries concerning reproduction
`outside the scope of the above should be sent to the Rights Department,
`Oxford University Press, at the address above
`You must not circulate this book in any other binding or cover
`and you must impose this same condition on any acquirer
`A catalogue record for this book is available from the British Library
`Libral3, of Congress Cataloging in Publication Data
`Bernstein, Joel.
`Polymorphism in molecular crystals [ Joel Bernstein.
`(IUCr monographs on crystallography; 14)
`Includes index.
`1. Polymorphism (Crystallography) 2. Molecular ct3rstals. I. Title. II. International
`Union of Crystallography monographs on crystallography; 14.
`QD951.B57 2002 548’.3--dc21 2001047556
`ISBN 0 19 850605 8
`
`Printed in Great Britain
`on acid-fi’ee paper by
`TJ. International Ltd, Padstow
`
`Lupin Ex. 1044 (Page 2 of 19)
`
`

`

`7
`
`Polymorphism of pharmaceuticals
`
`After discovery of the first cases of polymorphism witfi dramatic differences in biological
`activity between two forms of tfie same drug.., no pharmaceutical manufacturer could ~eglect
`the problem. (Borka 1991)
`
`There are many mysteries of nature that we have not yet solved. Hurricanes, for example
`continue to occur and often cause massive devastation. Meteorologists cannot predict months
`in advance when and with what velocity a hurricane will strike a specific community. Poty-
`morphism is a parallel phenomenon. We know that it will probably happen. But not why or
`when. Unfortunately, there is nothing we can do today to prevent a hurricane from striking any
`community or polymorphism from striking any drug. (Sun 1998)
`
`7.1
`
`Introduction
`
`The increasing awareness and importance ofpolymorphism in the past 30 years or so is
`perhaps nowhere more evident than in the field of pharmaceuticals (Bavin 1989). The
`landmark chapters of Buerger and Bloom (t937) and McCrone (1965) did not place
`any special emphasis on pharmaceutical materials. One outstanding exception was
`the book of Kofler and Kofler (1954), whose authors were members of an academic
`department of pharmacognosy (Webster: the branch of pharmacology that treats or
`considers the natural and chemical history of unprepared medicines). The seminal
`paper on the subject of potymorphism of pharmaceuticals was the review of Haleblian
`and McCrone (1969), which set the scope and the standards for many subsequent
`works. The literature on the polymorphism of pharmaceuticals is now best described
`as vast. A multiauthored monograph has appeared covering various aspects of the
`subject (Brittain 1999b), along with major sections of other books (Byrn 1982; Byrn
`et a!. 1999), and an ever increasing number of reviews (Haleblian 1975; Kuhnert-
`Brandstiitter 1975; Bouche and Draguet-Brughmans 1977; Giron 1981; Burger 1983;
`Threlfall 1995; Streng 1997; Caira 1998; Yu et al. 1998; Winter 1999; Vippagunta
`et al. 2001) covering various aspects of polymorphism as related directly to problems
`in the pharmaceutical field and/or pharmaceutical compounds, including some of the
`economic and intellectural property implications (Henck et al. 1997). Therefore, it
`would be foolhardy to attempt to present a comprehensive review of the subject here.
`Rather, in keeping with the general philosophy of this book, the aim is to provide a
`general introduction to the subject, with sufficient examples to demonstrate the points
`raised, and commensurate relevant references for the reader to s~ek further details
`and information.
`
`Lupin Ex. 1044 (Page 3 of 19)
`
`

`

`OCCURRENCE OF POLYMORPHISM IN PHARMACEUTICALS
`
`241
`
`7.2 Occurrence of polymorphism in pharmaceuticals
`
`7.2.1 Drug substances
`
`The development of a new drug from a promising lead compound to a marketed prod-
`uct is a long and expensive process, with odds of success estimated at 1 in 10000
`(Yevich 1991). The strict quality control requirements and the intellectual property
`implications of the drug industry lead to thorough and intensive investigations of the
`formation and properties of solid substances intended for the use in pharmaceutical
`formulations, both active ingredients and excipients. These efforts, often extending
`over long periods of time and with many potential experimental and enviromnen-
`tal variables, can create conditions that can lead to the appearance of polymorphic
`forms, intentionally or serendipitously. While it may not be surprising that many
`pharmaceutically important materials have been found to be polymorphic, or that any
`particular compound may turn out to be polymorphic, every compound is essentially
`a new situation, and the state of our knowledge and understanding of the phenomenon
`of polymorphism is still such that we cannot predict with any degree of confidence
`if a compotmd will be polymorphic, prescribe how to make possible (unknown)
`polymorphs, or predict what their properties might be (Beyer et al. 200t).
`There have been a few attempts to compile instances of potymorphism in phar-
`maceutically important materials. Since even the definition of what comprises
`’pharmaceutically important materials’ is itself subject to debate it is difficult to judge
`how comprehensive such compilations might be. However, generally they serve as
`useful references and are given here. One of the first organized attempts at such a com-
`pilation for steroids, sulphonamides and barbiturates was by Kuhnert-Brandst~itter
`(1965) (see also Kuhnert-Brandst~itter and Martinek 1965). Much of those data may
`be found in her subsequent book (Kuhnert-Brandst~itter 1971), which also contains
`a compilation of many of the thermal studies of pharmaceutical compounds that
`revealed polymorphic behaviour. Numerous additional reports of studies by the Inns-
`bruck school of polymorphic pharlnaceutical compounds (using thermomicroscopy
`and IR spectroscopy) have appeared in the literature since the middle 1960s; many of
`those have been listed by Byrn et al. (1999). Borka and Haleblian (1990) compiled
`a list of over 500 references to reports of polymorphisln in over 470 pharmaceuti-
`cally important compounds.l This was shortly followed (Borka 1991) by a review
`of polymorphic substances included in Fasciculae 1-12 of the European Pharma-
`copoeia (EP), including a comparison of melting points in the EP and the original
`literature. The latter review was subsequently updated in 1995 by including EP entries
`for Fasciculae 13-19 (Borka 1995).
`Griesser and Burger (1999) compiled the information regarding 559 polymorphic
`forms, solvates (including hydrates) of drug solids at 25 °C in the 1997 edition of
`
`l Dr Borka has communicated with this author that he and Dr Haleblian did not receive galley proofs
`of this paper, which unfortunately contains ’numerous printing errors’. The list of errata actually contains
`48 of them. Even with that cautionary note, it is a useful compilation.
`
`Lupin Ex. 1044 (Page 4 of 19)
`
`

`

`242
`
`POLYMORPHISM OF PHARMACEUT|CALS
`
`the ER They also noted that of the 10 330 compounds in the 1997 edition Merck Index
`only 140 (1.4 per cent) are specifically noted as polymorphic, 540 (5 per cent) are
`noted as hydrates and 55 (0.5 per cent) have been specified as solvates. These numbers
`reflect a failure to report or to include these phenomena rather than representative
`statistics, and may suggest the current state of awareness of polymorphism on the
`part of compilers of such compendia and reference works.
`A survey by this author of the 1 October 2000 release of the CSD (,---225 000 entries)
`yielded 6353 hits for the qualifier ’form’, 1045 hits for the qualifier ’phase’ 528 hits
`for the qualifier ’polymorph’, 20i hits for the qualifier ~modification’, 28 342 hits for
`the qualifier ’solvate’ and 21 132 hits for the qualifier ’hydrate’. Of course, the last two
`numbers give no indication of whether the materials are polymorphic and none of these
`statistics indicate the instances for which the structures of more than one form have
`been determined. The data for hydrates and solvates from the CSD may be considered
`quite reliable, since molecules of solvation are usually positively identified in the
`course of a crystal structure determination. The frequency of polymorphs, however,
`is likely to be underestimated, since many crystal structures of polymorphic systems,
`or what later are discovered to be polymorphic systems, are reported without making
`note of that fact. if the structure of only one member of a polymorphic family has
`been reported, then there may not be any reference to the polymorphism in the CSD.
`References to drugs discovered prior to 1971 that form solvates have been compiled,
`along with a separate summary of the thermomicroscopic behaviour of drug hydrates
`by Byrn el al. (1999).
`Because of the rather select nature of the sample set and the distinct possibility that
`not all forms are always reported caution must be exercised in drawing conclusions
`from such statistical surveys. Nevertheless, according to Griesser and Burger (1999)
`there may be some apparent tendencies which should be monitored as data continue to
`accumulate: polymorphism seems to be more common for compounds with molecular
`weight below 350; polymorphism seems to be more common for compounds with
`low solubility in water; for organic salts, the formation of hydrates appears to be
`more cornmon among larger molecules; organic solvates appear to be more com~non
`among neutral compounds with higher molecular weights.
`
`7.2.2 Excipients
`
`Pharmaceutical formulations contain the active drug ingredient(s) as well as excip-
`ients that serve a variety of purposes: fillers, stabilizers, coatings, drying agents,
`etc. As solid materials excipients exhibit varying degrees of crystallinity, fl’om the
`highly crystalline calcium hydrogen phosphate to nem’ly amorphous derivatives of
`cellulose, These materials can also exhibit polymorphism which may influence their
`performance in the formulation. Giron (1995, 1997) has listed many of the excipients
`that are known to exhibit a number of forms (polymorphs, solvates and amorphous).
`They include many of those that are widely used: lactose, so]’bitol, glucose, sucrose,
`magnesimn stearate, various calcium phosphates and mannitol (Burger et al. 2000).
`
`Lupin Ex. 1044 (Page 5 of 19)
`
`

`

`IMPORTANCE OF POLYMORPHISM IN PHARMACEUTICALS
`
`243
`
`The solid nature of the excipient may influence the final physical form of the tablet
`(Byrn et al. 2001), such as a tendency to stick (Schmid et al. 2000), or may induce
`a polymorphic conversion of the active ingredient (Kitamura et al. 1994). Hence,
`there have been attempts to develop protocols for the selection of compatible active
`ingredient-excipient compositions (Serajuddin et al. 1999). For instance, nuclear
`lnagnetic resonance spectroscopy has been employed to study the structural changes
`in epichlorohydrin cross-linked high amylose starch excipient (Shiftan el al. 2000),
`and has also been used to discriminate between two polymorphs of prednisolone
`present in tablets with excipients, even at low concentrations (5 per cent w/w) of the
`active ingredient (Saindon et al. 1993). The characterization of excipients by thermal
`methods has also been reviewed by Giron (1997).
`
`7.3
`
`Importance of polymorphism in pharmaceuticals
`
`Polymorphism can influence every aspect of the solid state properties of a drug. Many
`of the examples given in preceding chapters on the preparation of different crystal
`modifications, on analytical methods to determine the existence of polylnorphs and to
`characterize them and to study structure/property relations, were taken from the phar-
`maceutical industry, in part because there is a vast and growing body of literature to
`provide examples. One of the important aspects of polymorphisln in pharmaceuticals
`is the possibility of interconversion among polymorphic forms, whether by design
`or happenstance. This topic has also been recently reviewed (Byrn et al. 1999, espe-
`cially Chapter 13) and will not be covered here. Rather, in this section, we will present
`some additional examples of the variation of properties relevant to the use, efficacy,
`stability, etc. of pharmaceutically important compounds that have been shown to vary
`among different crystal modifications.
`
`7.3.1 Dissolution rate and solubility
`
`The dissolution properties and solubility are often crucial factors in the choice of the
`crystalline form for forlnutation of a drug product (Carstensen 1977). In general, these
`two factors play a major, if not over-riding role in determining the bioavailability of
`the drug substance (see also Section 7.3.2). The physiological absorption of a solid
`dosage form usually involves the dissolution of the solid in the stomach and the
`rate and extent of that dissolution is often the rate-determining step in the overall
`absorption process. Since different crystalline forms can exhibit different dissolution
`kinetics and limits, these properties are routinely studied in great detail for any drug
`substance, whether polymorphic or not; clearly, characterization for polymorphic
`substances is even more critical. As a result, there is extensive literature covering
`such studies. An early review contained a compilation of references to many of the
`previously studied materials (Kuhnert-Brandst~itter 1973).
`The fundamentals of the measurements of solubility and dissolution properties of
`pharamceuticals are given elscwhere in considerable detail (Vachon and Grant 1987;
`
`Lupin Ex. 1044 (Page 6 of 19)
`
`

`

`244
`
`POLYMORPHISM OF PHARMACEUTICALS
`
`Byrn et al. 1999, esp. Chapter 6; Grant and Brittain 1995). Reviews of the physical
`principles of dissolution and solubility behaviour of organic materials (Grant and
`Higuchi 1990) and pharmaceuticals (Grant and Brittain 1995) have been given and,
`particularly relevant to this chapter, typical examples of the variation of these pro-
`cesses among crystalline forms of pharmaceuticals are comprehensively described by
`Brittain and Grant (1999).
`
`7.3.2 Bioavailability
`
`The rate and extent of the physiological absorption of an active drug substance are
`decisive factors in its overall efficacy (e,g. carbamazepine, Zannikos et al. 1991).
`These can vary among different crystal modilications, and they have become an
`important scientific and regulatory issue (Ahr et al. 2000).
`While a nmnber of studies of the connection between the crystal modification and
`bioavailability have been published, it is reasonable to assume that many more remain
`the intellectual property of pharmaceutical companies or in confidential documenta-
`tion submitted to regulatory agencies. Also many in vitro studies, especially those of
`extent or rate of dissolution, are used to extrapolate to expected bioavailabilites (e.g.
`Chikaraishi et al. 1995; Shah et al. !999), which is indeed proven in SOlne cases such
`as the tetramorphic tolbutamide (Kimura et al. 1999). We cite here a few additional
`examples from the open literature, limiting ourselves (except for one example below)
`to cases in which the bioavailability does differ among crystal modifications.
`The two polylnorphs of the barbiturate pentobarbital exhibit significantly differ-
`mat rates of absorption and area under the curve for oral administration (Fig. 7.1)
`(Draguet-Brughmans et al. 1979). Comparison of rectal absorption of supposito-
`ries containing the two polymorphs of indomethacin showed higher levels for the o~
`
`3,5
`
`3,0
`
`2,5
`
`0,5
`
`bI
`
`Time (rain)
`
`Fig. 7.1 Rate of absorption of the two polylnorphs of orally administered pentobarbital.
`a, Form 1; b, Form II. (From Draguet-Brughmalas et al. 1979, with permission.)
`
`Lupin Ex. 1044 (Page 7 of 19)
`
`

`

`IMPORTANCE OF POLYMORPHISM IN PHARMACEUTICALS
`
`245
`
`form than for the ?/ form (Yokoyama et al. 1979). For the trimorphic antileukemic
`mercaptopurine Form Ill was found to be 6-7 times more soluble than Form t, but
`the bioavailability (in rabbits) was approximately 1.5 times greater for Form III than
`for Form I (Yokoyama et al. 1980). The antianxiety agent nabilone has at least four
`polymorphs, designated A, B, C and D, which appear to be equally hydrophobic
`and insoluble. Forms B and D are bioavailable in dogs, while A and C are not. On
`prolonged storage, heating or grinding, all convert to the nonbioavailable Forln A
`(Thakar et al. 1977).
`The bioavailablity can also vary between crystalline and amorphous modifica-
`tions (Section 7.7), as well as among polymorphic forms. The amorphous form
`of the antibacterial azlocillin sodium has more activity than the crystalline form
`(Kalinkova and Stoeva 1996). Similarly, an antinelnatode drug, code named PF1022A
`exists in four modifications, designated o~ (amorphous) and crystalline I, II and tli.
`o~ and III have higher solubility and are more effective than I and II against tissue-
`dwelling nematodes (Kachi et al. 1998). An amorphous form of another code-name
`drug (YM022), generated by spray drying, has enhanced bioavailability over two
`crystalline forms (Yanu et al. 1996).
`The degree of hydration of different modifications also plays a role in the bioavail-
`ability. One of the most studied systems in this regard is the anhydrate/trihydrate of
`the antibiotic ampicillin, although the results have not always led to consistent con-
`clusions. Early in vitro solubilities (Poole et a!. 1968) and rates of dissolution (Poole
`and Bahai 1968) were shown to differ. In vivo the anhydrous form reaches a maxi-
`mum concentration, hz vivo bioavailability studies by Ali and Farouk (1972) (Fig. 7.2)
`
`Fig. 7.2 Urinary excretion rates foilowing administration of the anhydrate (~) and trihydrate
`(NI) forms of ampicillin. (From Brittain and Grant 1999, with permission.)
`
`Lupin Ex. 1044 (Page 8 of 19)
`
`

`

`246
`
`POLYMORPHISM OF PHARMACEUTICALS
`
`clearly indicated differences between the two. Other studies, however, indicated sim-
`ilar bioavailabilities (Cabana et al. 1969; Mayersohn and Endrenyi 1973; Hill et al.
`1975). Brittain and Grant (1999) have suggested that the differences reported by
`various groups indicate that the bioavailability in the case of arnpicillin is strongly
`influenced by the nature of the fornmlation of the dosage form.
`There are also some examples of solvates (as opposed to hydrates) that have been
`reported to exhibit differing bioavailability. For instance, for implants of the methanol
`solvate of predisolone tert-butyl acetate the mean absorption rate was found to be 4.7
`times that of the anhydrous form, which in turn was similar to the hemiacetone solvate
`(Ballard and Biles 1964). The same authors also reported that the mean absorption
`rate for all solvates of cortisot tert-butyl acetate were significantly different from that
`of the anhydrous form.
`While the rate of absorption may differ, the extent of absorption, may still be
`equivalent. Such is the case for sulphalneter (sulphamethoxydiazine). The different
`polymorphic forms have been shown to exhibit different equilibrium solubilities and
`dissolution rates (Moustafa et al. 1971). Form II is thermodynamically less stable
`and has an absorption rate about 1.4 times that of Form III, which is more stable in
`water (Khalil et al. 1972). This group determined that the two forms have different
`absorption, but using 72-hour excretion data showed that the extent of the absorption
`of the two forms was equivalent (Khalafallah el al. 1974).
`Perhaps the classic example of the dependence of bioavailability on polymorphic
`form is chloramphenicol pahnitate. Chloramphenicol is a broad spectrum antibiotic
`and antirickettsial which was developed in the 1960s and had a significant portion of
`the market until the appearance of side effects limited its use to topical application.
`The exceedingly bitter taste of the active chloramphenicol led to its formulation
`as an oral suspension of the tasteless 3-pahnitate (CAPP). The early physical and
`physiological studies on the material were SUlmnarized by Aguiar et al. (1967) (see
`also Mitra et al. 1993). There are three polymorphic forms (A, B and C) in addition to
`an amorphous form. The characterization of the various forms by melting point and
`IR analyses proved problematic, even inconclusive, due to polymorphic transitions
`during grinding for sample preparation (Borka and Backe-Hansen 1968).
`The A form is the most stable, but only the B and alnorphous forms are biologically
`active. Aguiar et al. determined the physiological absorption rate as a function of
`the A and B polymorphs, as shown in Fig. 7.3. The suspension containing only the
`metastable B forln gives higher blood levels following oral doses than those containing
`only Form A, by nearly an order of magnitude. Since particle size was shown to have
`little effect on blood levels, it was concluded that the structure of the solid plays
`an intimate role in determining the physiological absorption rate. As a result of this
`finding, the mechanism of this absorption and its connection with the polymorphism
`were investigated in considerable detail.
`CAPP is nearly insoluble in water; hence, it must be hydrolysed by enzymes in the
`small intestine before absorption can take place. According to one possible proposed
`mechanism (Aguiar el al. 1967), the first and rate determining step in the total process
`
`Lupin Ex. 1044 (Page 9 of 19)
`
`

`

`IMPORTANCE OF POLYMORPHISM IN PHARMACEUTICALS
`
`247
`
`24
`
`22
`
`2O
`
`14
`
`~2
`
`4
`
`2
`
`00% 13
`
`0% B
`
`I
`
`3
`
`7
`9
`5
`Time after dosing (h)
`
`1
`
`11
`
`Fig. 7.3 Peak blood serum levels of chloramphenicol following dosing for pure polymorphs
`A and B and various mixtures for a single oral dose equivalent to 1.5 g of chloramphenicol
`palmitate. (From Haleblian and McCrone 1969 (after Aguiar et al. 1967) with permission.)
`
`is a dissolution of the ester followed by enzymatic hydrolysis of CAPR Such a mech-
`anism is consistent with the generally accepted dissolution/absorption mechanism for
`most solid drugs. However, Andersgaard et al. (1974) proposed a second mechanism
`in which solid CAPP is enzymatically attacked in the slnall intestine. If dissolution
`is the first and rate determining step of the total process, then there should be a close
`relationship between the rates of dissolution and the rates of enzymatic hydrolysis of
`polymorphs A and B. On the other hand, no relationship of this sort is expected if
`CAPP is attacked in the undissolved state.
`The rates of dissolution on the one hand and in vitro hydrolysis of the solid by the
`enzyme pancreatic lipase on the other hand are given in Fig. 7.4. If dissolution is the
`first step in the total hydrolysis process, the reaction scheme may be written as
`
`undissolved CAPP ~ dissolved CAPP :=> hydrolyzed CAPP.
`
`Since the rate of the second step of this process must be the same for Forms A and
`B of CAPP, this model for the absorption process leads to the conclusion that any
`differences in the rate of formation of hydrolysed CAPP must be due to a difference
`in the rate of dissolution of the two polymorphs. The data presented in Fig. 7.4 are
`not compatible with the assumption that dissolution is the first and rate determining
`step, since the slopes at time zero are significantly different from those expected for
`such a mechanism. Rather, Andersgaard et al. claim that it is more reasonable to
`assume that CAPP is attacked in the undissolved state, probably by pancreatic lipase,
`which is known to act on substances insoluble in water (Waki 1970). Further studies
`on CAPP and the analogous stcaratc (Cameroni et al. 1976) apparently corroborate
`
`Lupin Ex. 1044 (Page 10 of 19)
`
`

`

`248
`
`POLYMORPHISM OF PHARMACEUTICALS
`
`12
`
`001
`801
`
`601
`40
`
`20
`
`10
`
`1 2 3 4 5 6
`Time (hours)
`
`1
`
`2
`Time (hours)
`
`3
`
`4
`
`Fig. 7.4 Left: Rates of dissolution of polylnorphs A and B of CAPP. Right: Rates of iJ~ vitro
`enzymatic hydrolysis by pancreatic Iipase of polymorphs A and B of CAPR (After Andersgaard
`et al. 1974, with permission.)
`
`this assulnption, indicating that there is a direct connection between the solid state
`structure and the enzymatic hydrolysis of chloramphenicol palmitate, it has been
`found that storage of the B form at the relatively high temperature of 75 °C leads
`to conversion to the inactive A form in a matter of hours, suggesting that extended
`storage at higher temperatures could lead to reduced efficacy of a formulated product
`(Devilliers et al. 1991 ).
`The effect of the transformation of crystal form after ingestion has been of increas-
`ing interest in the past decade (see also Chapter 10). In a recent study of the two
`anhydroug Forms (I and III) and a dihydrate of carbamazepine (Kobayashi et al.
`2000), it was shown that the initial in vitro dissolution rate was of the order III > I >
`dihydrate, with Form III being transformed more rapidly to dihydrate than Form I.
`The solubilities of the two anhydrous forms were 1.5-1.6 times that of the hydrated
`form. When dosed to dogs at 40 mg per body weight there were no significant differ-
`ences between for the area under the curve. However, at doses of 200 mg per body
`weight significant differences in plasma concentration versus time were observed. It
`was suggested that the difference was due to rapid transformation from Form III to
`dihydrate in the gastrointestinal fluids.
`Two recent examples involving commercially successful drugs indicate the diver-
`sity in bioavailability among different polymorphic systems. At one extreme, the two
`polymorphic forms of ranitidine hydrochloride (Glaxo SmithKline’s H2-antagonist
`Zantac®) have been shown to be bioequivalent, which is one of the reasons that
`ethical and generic companies were involved in litigations over the two forms (see
`Section ! 0.2). At the other end of the spectrum is Abbot’s protease inhibitor Norvir®,
`generically ritonavir. After approximately two years on the market a new thermo-
`dynamically stable polymorph began to precipitate out of the semisolid formulated
`product. This proved to have lower solubility with greatly reduced bioavailability,
`
`Lupin Ex. 1044 (Page 11 of 19)
`
`

`

`MICROSCOPY AND THERMOMICROSCOPY OF PHARMACEUTICALS 249
`
`resulting in removal of the drug from the market for almost a year, until a new gel
`capsule (i.e. ’solution’) formulation could be developed (Chemburkar et al. 2000).
`The apparently greater stability of a newly prepared crystal form of glibenclamide
`has also been attributed to reduction in dissolution and bioavailability of its tablets
`(Panagopoulou-Kaplani and Malamataris 2000).
`
`7.4 Microscopy and thcrmomicroscopy of pharmaceuticals
`
`In Chapter 4, we noted the efficiency, simplicity, and utility of microscopy and ther-
`momicroscopy in the characterization and study of polymorphs. Perhaps the most
`widespread and systematic application of these techniques has been in the field of
`pharmaceutical materials, as promoted particularly by the Institute of Pharlnacognosy
`at the University of Innsbruck, and summarized in the (now out of print) book by
`Kuhnert-Brandst~itter (197t). In addition to introductory chapters on techniques of
`hot stage microscopy, the book contains a compilation of the results of hot stage stud-
`ies on approximately 1000 pharmaceutically important compounds. Many of these
`results were reported in more detail in an extensive series of papers beginning in the
`1960s (Kuhnert-Brandst~itter 1962) and continuing through the 1980s by Kuhnert-
`Brandstfitter and her late successor, A. Burger. Later contributions often dealt with
`compounds actually listed in a pharmacopoeia (e.g. Burger et al. 1986).
`Recent descriptions of microscopy and thermomicroscopy applied to pharmaceu-
`tically important compounds, along with quite comprehensive references to standard
`texts on the fundamentals and apparatus of the technique may be found in Byrn et al.
`(1999) and Brittain (1999b).
`Prior to the development of routine and increasingly sophisticated analytical meth-
`ods, hot stage microscopy competed as one of the principal tools for polymorph
`characterizatiou and classification. As noted in Chapter 4, successful strategies for the
`investigation of polymorphs require the application of as many analytical techniques
`as possible, and hot stage microscopy should be considered as one of the first, if not the
`first to be employed in a comprehensive characterization of a compound (Morris et al.
`1998). In spite of our current ability to generate a great deal of precise analytical data,
`there is often no substitute for physically observing solid materials and their behaviour
`as a function of temperature, preferably on the polarizing microscope. Evidence for
`the success of this approach has been provided in a nulnber of recent studies.
`In some cases, new phases that may not be detectable by other methods may be
`detected optically (Chang et al. 1995). Solid state conversions and their monotropic
`(Burger et al. I997) or enantiotropic nature (Henck et aI. 2000), or the products
`of desolvations may be easily recognized (Schinzer et al. 1997). Intimate processes
`of polylnorphic behaviour, such as nucleation, crystal growth, habit transformation,
`sublimation and properties of the melt (e.g. degradation) may be readily observed and
`video recorded (de Wet et al. 199.8).
`As noted at the conclusion of Chapter 4, the amalgamation of a number of ana-
`lytical techniques into a single instrument considerably expands the possibilities for
`
`Lupin Ex. 1044 (Page 12 of 19)
`
`

`

`250
`
`POLYMORPHISM OF PHARMACEUTICALS
`
`detecting and characterizing polymorphs. This is a particularly powerful combination
`when optical/thennomicroscopy (preferably with video recording) is combined with
`other analytical methods. Thermomicroscopy combined with DTA led to the verifi-
`cation of a monotropic transformation between two forms of ibopamin (Laine el al.
`1995). Combining hot stage microscopy with Raman spectroscopy led to the direct
`characterization and correlation of the thermal and spectroscopic information on three
`polymorphs of paracetamol, as well as the first report of lufenuron, a chitin synthesis
`inhibitor used in pest control and crop protection, and the identity of the polymor-
`phic form in the marketed tablets (Szelagiewicz et al. 1999). FTIR was incorporated
`into a hot stage microscope to simultaneously obtain the visible and spectroscopic
`characterization of the three polymorphic forms of carbamazepine (Rustichelli et al.
`2000). Finally, microspectroscopic FTIR and FT-Raman were combined with hot
`stage microscopy to study the polymorphism in (R, S)-proxyphyilline, including the
`production and characterization of a new, kinetically very unstable form that most
`likely could not be detected or analysed by any other technique (Griesser et al. 2000).
`
`7.5 Thermal analysis of pharmaceuticals
`
`The fundamentals of the application of thermal analysis in the study and characteri-
`zation of polymorphs are given in Section 4.3, and many of the examples presented
`there are on pharmaceutically important compounds. The use of thermal analysis in
`the pharmaceutical area, including specific applications to polymorphic materials was
`reviewed in the early 1980s (Wollmann and Braun 1983; Giron-Forest 1984), the for-
`mer reference c