Solid Dispersion of Poorly Water-Soluble Drugs: Early Promises,
`Subsequent Problems, and Recent Breakthroughs
`
`ABU T. M. SERAJUDDIN†
`
`Contribution from Pharmaceutics R & DDepartment, Bristol-Myers Squibb Pharmaceutical Research Institute,
`New Brunswick, New Jersey 08903-0191.
`
`Received October 7, 1998. Accepted for publication July 15, 1999.
`
`Abstract 0 Although there was a great interest in solid dispersion
`systems during the past four decades to increase dissolution rate
`and bioavailability of poorly water-soluble drugs, their commercial use
`has been very limited, primarily because of manufacturing difficulties
`and stability problems. Solid dispersions of drugs were generally
`produced by melt or solvent evaporation methods. The materials, which
`were usually semisolid and waxy in nature, were hardened by cooling
`to very low temperatures. They were then pulverized, sieved, mixed
`with relatively large amounts of excipients, and encapsulated into hard
`gelatin capsules or compressed into tablets. These operations were
`difficult to scale up for the manufacture of dosage forms. The situation
`has, however, been changing in recent years because of the availability
`of surface-active and self-emulsifying carriers and the development
`of technologies to encapsulate solid dispersions directly into hard
`gelatin capsules as melts. Solid plugs are formed inside the capsules
`when the melts are cooled to room temperature. Because of surface
`activity of carriers used, complete dissolution of drug from such solid
`dispersions can be obtained without the need for pulverization, sieving,
`mixing with excipients, etc. Equipment is available for large-scale
`manufacturing of such capsules. Some practical limitations of dosage
`form development might be the inadequate solubility of drugs in carriers
`and the instability of drugs and carriers at elevated temperatures
`necessary to manufacture capsules.
`
`Introduction
`The enhancement of oral bioavailability of poorly water-
`soluble drugs remains one of the most challenging aspects
`of drug development. Although salt formation, solubiliza-
`tion, and particle size reduction have commonly been used
`to increase dissolution rate and thereby oral absorption and
`bioavailability of such drugs,1 there are practical limita-
`tions of these techniques. The salt formation is not feasible
`for neutral compounds and the synthesis of appropriate salt
`forms of drugs that are weakly acidic or weakly basic may
`often not be practical. Even when salts can be prepared,
`an increased dissolution rate in the gastrointestinal tract
`may not be achieved in many cases because of the recon-
`version of salts into aggregates of their respective acid or
`base forms. The solubilization of drugs in organic solvents
`or in aqueous media by the use of surfactants and cosol-
`vents leads to liquid formulations that are usually undesir-
`able from the viewpoints of patient acceptability and
`commercialization. Although particle size reduction is
`commonly used to increase dissolution rate, there is a
`practical limit to how much size reduction can be achieved
`by such commonly used methods as controlled crystalliza-
`tion, grinding, etc. The use of very fine powders in a dosage
`
`† Present address: Novartis, 59 Route 10, Building 401, East
`Hanover, New Jersey 07936.
`
`Figure 1sA schematic representation of the bioavailability enhancement of
`a poorly water-soluble drug by solid dispersion compared with conventional
`tablet or capsule.
`
`form may also be problematic because of handling difficul-
`ties and poor wettability.
`In 1961, Sekiguchi and Obi2 developed a practical
`method whereby many of the limitations with the bioavail-
`ability enhancement of poorly water-soluble drugs just
`mentioned can be overcome. This method, which was later
`termed solid dispersion,3 involved the formation of eutectic
`mixtures of drugs with water-soluble carriers by the
`melting of their physical mixtures. Sekiguchi and Obi2
`suggested that the drug was present in a eutectic mixture
`in a microcrystalline state. Later, Goldberg et al.4,5 dem-
`onstrated that all the drug in a solid dispersion might not
`necessarily be present in a microcrystalline state; a certain
`fraction of the drug might be molecularly dispersed in the
`matrix, thereby forming a solid solution. In either case,
`once the solid dispersion was exposed to aqueous media
`and the carrier dissolved, the drug was released as very
`fine, colloidal particles. Because of greatly enhanced surface
`area obtained in this way, the dissolution rate and the
`bioavailability of poorly water-soluble drugs were expected
`to be high.
`The advantage of solid dispersion, compared with con-
`ventional capsule and tablet formulations, is shown sche-
`matically in Figure 1.6 From conventional capsules and
`tablets, the dissolution rate is limited by the size of the
`primary particles formed after the disintegration of dosage
`forms. In this case, an average particle size of 5 (cid:237)m is
`usually the lower limit, although higher particle sizes are
`preferred for ease of handling, formulation, and manufac-
`turing. On the other hand, if a solid dispersion or a solid
`solution is used, a portion of the drug dissolves immediately
`to saturate the gastrointestinal fluid, and the excess drug
`precipitates out as fine colloidal particles or oily globules
`of submicron size.
`Because of such early promises in the bioavailability
`enhancement of poorly water-soluble drugs, solid dispersion
`has become one of the most active areas of research in the
`pharmaceutical field. Numerous papers on various aspects
`
`1058 / Journal of Pharmaceutical Sciences
`Vol. 88, No. 10, October 1999
`
`10.1021/js980403l CCC: $18.00
`Published on Web 08/27/1999
`
`© 1999, American Chemical Society and
`American Pharmaceutical Association
`
`0001
`
`PSG2003
`Catalent Pharma Solutions v. Patheon Softgels
`IPR2018-00422
`
`

`

`of solid dispersion were published since 1961; Chiou and
`Riegelman3 and Ford7 reviewed the early research in this
`area. Despite an active research interest, the commercial
`application of solid dispersion in dosage form design has
`been very limited. Only two products, a griseofulvin-in-
`poly(ethylene glycol) solid dispersion (Gris-PEG, Novartis)
`and a nabilone-in-povidone solid dispersion (Cesamet, Lilly)
`were marketed during three decades following the initial
`work of Sekiguchi and Obi in 1961. The objectives of the
`present article are to critically review some of the limita-
`tions of solid dispersion that prevented its wider com-
`mercial application and to discuss how the situation is now
`changing because of the availability of new types of vehicles
`and the development of new manufacturing technologies.
`
`Limitations of Solid Dispersion Systems
`Problems limiting the commercial application of solid
`dispersion involve (a) its method of preparation, (b) repro-
`ducibility of its physicochemical properties, (c) its formula-
`tion into dosage forms, (d) the scale up of manufacturing
`processes, and (e) the physical and chemical stability of
`drug and vehicle. Some of the issues are discussed next.
`Method of PreparationsIn their pioneering study,
`Sekiguchi and Obi2 prepared solid dispersions of sulfathia-
`zole in such carriers as ascorbic acid, acetamide, nicotina-
`mide, nicotinic acid, succinimide, and urea by melting
`various drug-carrier mixtures. To minimize melting tem-
`peratures, eutectic mixtures of the drug with carriers were
`used. Yet, in all cases, except acetamide, the melting
`temperatures were >110 °C, which could chemically de-
`compose drugs and carriers.3 High temperatures (>100 °C)
`were also utilized by Goldberg et al. in preparing acetami-
`nophen-urea,4 griseofulvin-succinic acid,4 and chloram-
`phenicol-urea8 solid dispersions. After melting, the next
`difficult step in the preparation of solid dispersions was
`the hardening of melts so that they could be pulverized
`for subsequent formulation into powder-filled capsules or
`compressed tablets. Sekiguchi and Obi2 cooled the sul-
`fathiazole-urea melt rapidly in an ice bath with vigorous
`stirring until it solidified. Chiou and Riegelman9 facilitated
`hardening of the griseofulvin-PEG 6000 solid dispersion
`by blowing cold air after spreading it on a stainless steel
`plate and then storing the material in a desiccator for
`several days. In preparing primidone-citric acid solid
`dispersions, Summers and Enever10 spread the melt on
`Petri dishes, cooled it by storing the Petri dishes in a
`desiccator, and finally placed the desiccator at 60 °C for
`several days. Allen et al.11 prepared solid dispersions of
`corticosteroids in galactose, dextrose, and sucrose at 169,
`185, and 200 °C, respectively, and then placed them on
`aluminum boats over dry ice. Timko and Lordi12 also used
`blocks of dry ice to cool and solidify phenobarbital-citric
`acid mixtures that had previously been melted on a frying
`pan at 170 °C. The fusion method of preparing solid
`dispersion remained essentially similar over the period of
`time. More recently, Lin and Cham13 prepared nifedipine-
`PEG 6000 solid dispersions by blending physical mixtures
`of the drug and the carrier in a V-shaped blender and then
`heating the mixtures on a hot plate at 80-85 °C until they
`were completely melted. The melts were rapidly cooled by
`immersion in a freezing mixture of ice and sodium chloride,
`and the solids were stored for 24 h in a desiccator over silica
`gel before pulverization and sieving. Mura et al.14 solidified
`naproxen-PEG melts in an ice bath and the solids were
`then stored under reduced pressure in a desiccator for 48
`h before they were ground into powders with a mortar and
`pestle. In another study, Owusu-Ababio et al.15 prepared
`a mefenamic acid-PEG solid dispersion by heating the
`drug-carrier mixture on a hot plate to a temperature above
`
`the melting point of mefenamic acid (253 °C) and then
`cooling the melt to room temperature under a controlled
`environment.
`Another commonly used method of preparing a solid
`dispersion is the dissolution of drug and carrier in a
`common organic solvent, followed by the removal of solvent
`by evaporation.9,16,17 Because the drug used for solid
`dispersion is usually hydrophobic and the carrier is hy-
`drophilic, it is often difficult to identify a common solvent
`to dissolve both components. Large volumes of solvents as
`well as heating may be necessary to enable complete
`dissolution of both components. Chiou and Riegelman9 used
`500 mL of ethanol to dissolve 0.5 g of griseofulvin and 4.5
`g of PEG 6000. Although in most other reported studies
`the volumes of solvents necessary to prepare solid disper-
`sions were not specified, it is possible that they were
`similarly large. To minimize the volume of organic solvent
`necessary, Usui et al.18 dissolved a basic drug in a hydro
`alcoholic mixture of 1 N HCl and methanol, with drug-to-
`cosolvent ratios ranging from 1:48 to 1:20, because as a
`protonated species, the drug was more soluble in the acidic
`cosolvent system than in methanol alone. Some other
`investigators dissolved only the drug in the organic solvent,
`and the solutions were then added to the melted carriers.
`Vera et al.19 dissolved 1 g ofoxodipine per 150 mL of
`ethanol before mixing the solution with melted PEG 6000.
`In the preparation of piroxicam-PEG 4000 solid dispersion,
`Fernandez et al.20 dissolved the drug in chloroform and
`then mixed the solution with the melt of PEG 4000 at 70
`°C. Many different methods were used for the removal of
`organic solvents from solid dispersions. Simonelli et al.16
`evaporated ethanolic solvent on a steam bath and the
`residual solvent was then removed by applying reduced
`pressure. Chiou and Riegelman9 dried an ethanolic solution
`of griseofulvin and PEG 6000 in an oil bath at 115 °C until
`there was no evolution of ethanol bubbles. The viscous
`mass was then allowed to solidify by cooling in a stream of
`cold air. Other investigators used such techniques as
`vacuum-drying,20,21 spray-drying,22-25 spraying on sugar-
`beads using a fluidized bed-coating system,26 lyophiliza-
`tion,27 etc., for the removal of organic solvents from solid
`dispersions. None of the reports, however, addressed how
`much residual solvents were present in solid dispersions
`when different solvents, carriers, or drying techniques were
`used.
`Reproducibility of Physicochemical Propertiess
`In their pioneering studies, Sekiguchi and Obi2 observed
`that manufacturing conditions might greatly influence the
`physicochemical properties of solid dispersions formed.
`They cooled drug-carrier melts under vigorous stirring
`conditions to obtain fine and uniform drug particles in solid
`dispersions. Various investigators observed that heating
`rate, maximum temperature used, holding time at a high
`temperature, cooling method and rate, method of pulveri-
`zation, and particle size may greatly influence the proper-
`ties of solid dispersions prepared by the melt method.
`McGinity et al.28 prepared solid dispersions of tolbutamide
`in urea and PEG 6000 by flash cooling in a bath of dry ice
`and acetone or by gradual cooling over a period of several
`hours by immersion in an oil bath. The powder X-ray
`diffraction patterns of the tolbutamide-urea solid disper-
`sion differed markedly depending on the cooling rate. The
`slow-cooled solid dispersion of tolbutamide in urea dem-
`onstrated a complete lack of crystallinity for both the drug
`and urea, whereas the flash-cooled dispersions showed only
`the absence of drug crystallinity. In the powder X-ray
`diffraction patterns of tolbutamide-PEG 6000 solid disper-
`sions, peaks for both tolbutamide and PEG 6000 were
`observed; however, their degree of crystallinity in flash-
`cooled samples was less than that in the slow-cooled
`
`Journal of Pharmaceutical Sciences / 1059
`Vol. 88, No. 10, October 1999
`
`0002
`
`

`

`samples. In another study, a metastable amorphous form
`of nifedipine was formed in its solid dispersions in PEG
`4000 and PEG 6000 when the drug-carrier melts were
`cooled rapidly, whereas slow cooling of melts or powdering
`of solidified mass resulted in the crystallization of drug.29
`Gine´s et al.30 studied the effect of fusion temperature on
`oxazepam-PEG 4000 solid dispersions. Microscopic ex-
`amination revealed the presence of crystalline oxazepam
`and the spherulitic form of PEG 4000 in solid dispersions
`prepared by fusion at 100 °C. In contrast, a fusion tem-
`perature of 150 °C produced a solid dispersion with no
`crystalline form of the drug and the presence of PEG 4000
`in a hedritic form. Complete dissolution of drug in the
`carrier at 150 °C in contrast to 100 °C was reported to be
`responsible for such a difference in physicochemical prop-
`erties of the solid dispersions produced. Dordunoo et al.31
`also observed a change of triamterene and temazepam from
`crystalline to amorphous form in poly(ethylene glycol) solid
`dispersions when the fusion temperature was increased
`from 100 to 150 °C. Such changes in physical states of
`drugs in solid dispersions result into differences in drug
`dissolution rates in aqueous media.30 Drug-to-carrier ratio
`and particle size of solid dispersions were also reported to
`influence the dissolution rate of drug.32
`The properties of solid dispersions prepared by the
`solvent method may also vary depending on manufacturing
`conditions. The solvent method usually leads to amorphous
`forms of drugs. However, some crystallinity of drug may
`be observed depending on the drug-to-carrier ratio used.33
`Although no detailed studies were reported in the litera-
`ture, it is expected that the nature of solvent used, drug-
`to-solvent and carrier-to-solvent ratios, drying method, and
`drying rate may significantly influence the physicochemical
`properties of solid dispersions formed.
`Dosage Form DevelopmentsSolid dispersion must be
`developed into convenient dosage forms, such as capsules
`and tablets, for their clinical use and successful com-
`mercialization. As already mentioned, solid dispersions
`produced by the melt method are usually hardened at very
`low temperatures and then pulverized with mortars and
`pestles. Similarly, solid dispersions produced by the solvent
`method are also pulverized after solvent removal and
`hardening. Some of the challenges in the dosage form
`development of such materials are difficulty of pulveriza-
`tion and sifting of the dispersions, which are usually soft
`and tacky, poor flow and mixing properties of powders thus
`prepared, poor compressibility, drug-carrier incompat-
`ibility, and poor stability of dosage forms. However, there
`are very few reports in the literature addressing these
`important issues.7 Even the limited number of reports
`describing any dosage form developmental aspects of solid
`dispersions only confirm that the task of formulating solid
`dispersions into capsules or tablets may be a very complex
`and difficult one. In developing a tablet formulation for the
`indomethacin-PEG 6000 solid dispersion, Ford and Ru-
`binstein34 reported that the solid dispersion was not
`amenable to wet granulation because water could disrupt
`its physical structure. In addition, the dispersion was soft
`and tacky. To overcome these problems, the authors
`adopted an in situ dry granulation method where the
`excipients (calcium hydrogen phosphate and sodium starch
`glycolate) were preheated and rotated in a water-jacketed
`blender at 70 °C, and the indomethacin-PEG 6000 mixture
`that melted at 100 °C was then added to the moving
`powder. After mixing, the granules were passed through a
`20-mesh sieve and allowed to harden at 25 °C for 12 h.
`Then, the granules were mixed with a relatively high
`concentration of magnesium stearate (1%) and compressed
`into tablets. To process 100 mg of solid dispersion, 506 mg
`of other excipients were used, thus making the final weight
`
`1060 / Journal of Pharmaceutical Sciences
`Vol. 88, No. 10, October 1999
`
`of a 25-mg indomethacin tablet 606 mg. Yet, the tablet did
`not disintegrate in water despite the use of a large amount
`of excipients. It dissolved slowly by erosion, and the dis-
`solution rate decreased on aging of the tablet. In another
`study, the same investigators used an essentially similar
`in situ dry granulation method for the preparation of tablet
`dosage forms for a chlorpropamide-urea solid dispersion,
`where the drug, the carrier, and the excipients were mixed
`in a rotating flask on a water bath maintained at 100 °C.35
`The properties of these formulations also changed with
`time, and the authors concluded that aging could “limit
`their usefulness as prospective dosage forms”. During the
`development of a tablet formulation for a furesemide-poly-
`(vinylpyrrolidone) (PVP) solid dispersion, Akbuga et al.36
`observed that method of preparation, choice of disintegrant
`and particle size of solid dispersions were critical factors
`in determining the properties of tablets produced. Despite
`the use of relatively large amounts of disintegrants, the
`tablets did not disintegrate. Rather, they dissolved by
`erosion only, and the erosion rate varied depending on the
`disintegrant used. In addition, the dissolution rate of
`tablets prepared by double compression (slugging and
`recompression of dry granules) was much slower than that
`of the tablets prepared by single compression. The dissolu-
`tion rate of tablet was also dependent on the particle size
`of solid dispersion used; the rate decreased by a factor of 5
`when 100-mesh particles were used in place of 80-mesh
`particles. Also, the compressibility of solid dispersion
`decreased with a decrease in particle size. In another study,
`Sjo¨kvist and Nystrom 37 overcame the compression dif-
`ficulties due to sticking of griseofulvin-xylitol solid disper-
`sions to dies and punches by lubricating die wall and punch
`faces with 1% (w/w) magnesium stearate suspension before
`the compression of each tablet. The authors observed that
`the dissolution rate of tablet was highly sensitive to com-
`pression pressure. The sticking of solid dispersion to dies
`and punches might become so problematic that Kaur et
`al. 38 resorted to placing small pieces of grease-proof paper
`between metal surfaces and granules before the compres-
`sion of tablets.
`The lack of disintegration and the slow dissolution of
`tablets prepared from solid dispersions could be related to
`the soft and waxy nature of carriers used (e.g., PEG) in
`many of the reported studies. Such carriers essentially act
`as strong binders within tablets. During compression, the
`carriers could plasticize, soften, or melt, filling the pores
`within tablets and thus making them nondisintegrating.
`It is also possible that the softened and melted carriers coat
`the disintegrants and other ingredients used in tablets, and
`such a coating, along with the reduction of porosity of
`tablets, make the disintegrants ineffective. Use of a very
`high ratio of solid dispersion to added excipient might
`alleviate the problem. In one study,15 270 mg of microc-
`rystalline cellulose (Avicel) was used to formulate 30 mg
`of mefenamic acid-PEG solid dispersion into a tablet with
`good dissolution. The use of such a high ratio of added
`excipient would, however, greatly increase the size of tablet
`and might, therefore, be impractical in most formulations.
`Scale Up of Manufacturing ProcessessBecause very
`few solid dispersion products prepared by melt or solvent
`methods have been marketed, there are practically no
`reports on the scale up of such products. It is apparent from
`the discussion just presented that the scale up of the
`methods of preparation of solid dispersions and their
`dosage forms could be very challenging. In most of the
`studies reported in the literature, solid dispersions by the
`melt method were prepared in a small scale by heating
`drug-carrier mixtures in beakers, frying pans, etc. that
`were placed on hot plates and then cooling the melts in an
`ice bath, a dry ice-acetone mixture, etc.2-5,7-14 Because
`there could be condensation of moisture over solid disper-
`
`0003
`
`

`

`sions during cooling to low temperatures, strict protection
`from moisture was necessary in all cases. The scale up
`challenges may be illustrated with the example of the
`preparation of a phenytoin-PEG 4000 solid dispersion by
`Yakou et al.39 The drug-carrier mixture was heated at 250
`°C under constant stirring until a clear homogeneous melt
`was obtained, and the melt was air-cooled by spreading
`on stainless steel trays. The trays were stored in a
`desiccator for 3 days to enhance solidification of the solid
`dispersion. The resulting material was then crushed in a
`cutter mill, and the powders were sieved to collect a sieve
`fraction of 105-177-(cid:237)m particle size for use in the dosage
`form. The scale up of such a method would be difficult and
`it might even be impractical in many cases because of
`possible degradation of both drug and carrier at high
`temperatures used. The scale up might also necessitate a
`large capital investment because a chemical plant-like
`facility, rather than a common pharmaceutical dosage form
`manufacturing plant, would be required to process and
`manufacture the products. For scale up of the cooling
`process, Lefebvre et al.40 recommended such continuous
`operation as cooling on the surfaces of moving belts or
`rotating cylinders, and spray congealing. The practical
`application of the methods, however, was not demon-
`strated. Kennedy and Niebergall41 described a hot-melt
`fluid bed method whereby nonpareils could be coated with
`PEGs having molecular weights between 1450 and 4600.
`A similar method can possibly be used to deposit solid
`dispersions on nonpareils and might in the future find
`application in the manufacture and scale up of solid
`dispersion formulations.
`The physicochemical properties and stability of solid
`dispersions may also be affected by scale up because
`heating and cooling rates of solid dispersions under large-
`scale manufacturing conditions may differ greatly from that
`in small beakers.28,29 Drug-carrier compatibility at a high
`temperature also requires careful consideration. Dubois
`and Ford42 reported the chain scission of PEG 600 during
`fusion with disulfiram, furosemide, chlorthiazide, and
`chlorpropamide.
`The scale up of the solvent method of preparing solid
`dispersions may also be very challenging. A chemical plant
`environment would be necessary to evaporate hundreds
`and even thousands of liters of organic solvents necessary
`to prepare solid dispersions for kilogram quantities of
`drugs.17,20 The cost of recovery of these solvents may be
`very high. Removal of residual amounts of potentially toxic
`organic solvents such as chloroform and methanol from
`large masses of material may be difficult because the solid
`dispersions are usually amorphous and may exist in viscous
`and waxy forms. Solvates may also be formed with drugs
`and carriers. Because most dosage form manufacturing
`facilities are not equipped to handle large volumes of
`organic solvents, one way to resolve the issue might be the
`designation of solid dispersion as an active pharmaceutical
`ingredient or bulk drug substance. In that case, the
`responsibility of the manufacture of solid dispersion can
`be shifted to the chemical plant. It would be necessary to
`conduct all developmental activities using the solid disper-
`sion, so this approach might not be suitable for situations
`where active pharmaceutical ingredients have multiple
`uses (e.g., oral and parenteral).
`The final step in the manufacturing process, which is
`the conversion of solid dispersions into stable and market-
`able dosage forms, may be the most difficult one to scale
`up, optimize, and validate. Most of the commonly used solid
`dispersion vehicles are soft and sticky and, as a result, the
`pulverized forms of solid dispersions produced by such
`vehicles may not be amenable to processing by high-speed
`capsule or tablet filling machines.
`
`StabilitysThe physical instability of solid dispersions
`due to crystallization of drugs was the subject of most
`published reports in the literature.3,7 In a solid dispersion
`prepared by the melt method, a certain fraction of the drug
`may remain molecularly dispersed, depending on its solu-
`bility in the carrier used, thus forming a solid solution. How
`the excess drug exists may greatly depend on the method
`of manufacture of the system; it may, as a whole or in part,
`form a supersaturated solution, separate out as an amor-
`phous phase, or crystallize out. The supersaturated and
`amorphous forms may, in turn, crystallize out on aging.
`Similarly, certain carriers may also exist in thermody-
`namically unstable states in solid dispersions and undergo
`changes with time. Chiou43 reported that griseofulvin
`precipitated out in an amorphous form in a griseofulvin-
`PEG 6000 solid dispersion during the time of its prepara-
`tion. The amorphous material crystallized out on aging,
`except when the drug concentration in the dispersion was
`5% or less. Ford and Rubinstein44 attributed similar
`crystallization as the cause for a decrease in dissolution
`rate of drug from indomethacin-PEG 6000 solid disper-
`sions with time. The decrease in the dissolution rate of
`indomethacin was also dependent on drug concentration
`in the solid dispersion. The decrease was greater for a
`higher drug concentration because a larger fraction of drug
`crystallized out. In another study, Suzuki and Sunada45
`observed that on exposure of a nifedipine-nicotinamide-
`hydroxypropylmethylcellulose (HPMC) solid dispersion to
`60% RH at 30 °C or 75% RH at 40 °C for 1 month,
`nifedipine converted from the amorphous to the crystalline
`state, thus lowering the dissolution rate of nifedipine
`drastically. No conversion of nifedipine to the crystalline
`state was observed when the solid dispersion was stored
`at an elevated temperature in the absence of humidity.
`Although the presence of HPMC facilitated the conversion
`of nifedipine to an amorphous state during the cooling of
`drug-nicotinamide melt to room temperature at the time
`of manufacturing,
`it did not prevent the subsequent
`crystallization of drug under humid conditions. Pronounced
`decreases in dissolution rates due to drug crystallization
`were also reported for tablets prepared from solid disper-
`sions.34,35 No such decrease in dissolution rate on aging was
`observed by Khalil et al.46 in corticosteroid-PEG solid
`dispersions prepared with a drug-to-carrier ratio of 1:99,
`possibly because most of the drug was molecularly dis-
`persed in the carrier. The corticosteroid, however, exhibited
`chemical degradation due to oxidation by the peroxides
`present in PEG. The cooling rate of solid dispersions may
`also significantly influence their aging behavior. It has been
`reported that the crystallinity of drug in solid dispersions
`is less influenced by aging when a slow cooling rate is used
`because thermodynamically more stable systems are pro-
`duced during the time of preparation.47,48
`The conversion of drug to crystalline state is also the
`primary stability issue with solid dispersions prepared by
`the solvent method. PVP, which is commonly used as a
`carrier in such solid dispersions, is amorphous and does
`not convert to a crystalline state. However, certain other
`carriers may convert from their amorphous states to
`crystalline states in solid dispersions. Zografi and co-
`workers49,50 extensively studied the physicochemical prop-
`erties of the amorphous states of drugs and excipients and
`observed that the crystallization of amorphous materials
`is facilitated by moisture. This effect is why strict protection
`from moisture is necessary during the preparation and
`storage of most solid dispersions. Doherty and York51
`studied the stability of furosemide-PVP solid dispersion
`in the temperature range of 6 to 45 °C and 40% RH for up
`to 1 year. They did not observe any crystallization of
`furosemide and suggested that PVP may indeed act as a
`
`Journal of Pharmaceutical Sciences / 1061
`Vol. 88, No. 10, October 1999
`
`0004
`
`

`

`stabilizer in the solid dispersion by retarding crystallization
`of drug at a relatively low humidity. Rapid crystallization
`of furosemide in the solid dispersion was, however, evident
`when the humidity was raised to 75% RH. Similar obser-
`vations were also made by Guillaume et al.52 for an
`oxodipine-PVP solid dispersion where no crystallization
`of oxodipine was observed in 18 months when samples were
`stored under 55% RH at various temperatures, but the
`drug crystallized out at 80% RH. The stabilization of drugs
`in amorphous forms in solid dispersions is an active area
`of research in the pharmaceutical field. For an indometha-
`cin-PVP solid dispersion system, Taylor and Zografi53
`suggested that hydrogen bonding between the drug and
`PVP might offer an explanation for the absence of drug
`crystallization. Lu and Zografi54 recently demonstrated that
`indomethacin forms a completely miscible amorphous
`mixture with citric acid and PVP when the weight fraction
`of PVP in the ternary mixture exceeds 0.3 weight fraction.
`Thus, both the choice of carrier and the drug-to-carrier ratio
`are important considerations in the stabilization of solid
`dispersions.
`
`Breakthroughs in Solid Dispersion Technology
`Because of the various limitations just mentioned, it is
`not surprising that the solid dispersion system, despite its
`many potential advantages, has not been widely used in
`pharmaceutical dosage forms. Under the present health
`care economic climate, the goal of any drug development
`program in the pharmaceutical industry is to rapidly
`progress a new chemical entity from the discovery stage
`to clinical testing to determine whether it is safe and
`clinically effective. The limited supply of the bulk drug
`substance at the early drug development phase and the
`accelerated time line would not allow a formulator to
`address most of the challenges (vide supra) of a solid
`dispersion formulation. Most importantly, if a compound
`proves promising in early clinical testing, the scale up of
`complex manufacturing processes for the development of
`marketable dosage forms cannot be ensured.
`Two recent breakthroughs in the formulation of solid
`dispersion systems involve (1) the development of technolo-
`gies to fill solid dispersions directly into hard gelatin
`capsules and (2) the availability of surface-active and self-
`emulsifying carriers. As a result, there is renewed interest
`in such systems for use in commercial development of drug
`products.6,55
`Direct Capsule-FillingsAlthough the filling of semi-
`solid materials into hard gelatin capsules as melts, which
`solidify at room temperature, was first described by Fran-
`cois and Jones in 1978,56 it was not until much later that
`the potential application of the technique for solid disper-
`sions was fully realized. Chatham57 reported the possibility
`of preparing PEG-based solid dispersions by filling drug-
`PEG melts in hard gelatin capsules. By using PEG with
`molecular weights ranging from 1000 to 8000, Serajuddin
`et al.,