Journal of Controlled Release, 16 ( 1991 ) 311-318
`© 1991 Elsevier Science Publishers B.V. 0168-3659/91 /$03.50
`ADONIS 016836599100072M
`
`311
`
`COREL00605
`
`Encapsulation of 5-aminosalicylic acid into Eudragit RS
`microspheres and modulation of their release characteristics by
`use of surfactants
`
`P.J. Watts, M.C. Davies and C.D. Melia
`Department of Pharmaceutical Sciences, University of Nottingham, Nottingham, U.K.
`(Received October 11, 1990; accepted in revised form November 29, 1990)
`
`An emulsification-solvent evaporation technique has been used to produce microspheres containing 5-
`aminosalicylic acid from the acrylate-methacrylate copolymer, Eudragit™ RS. Two emulsifiers were
`used in the preparation procedure, either sodium dodecyl sulphate (SOS) or Tween 20, at a range of
`concentrations. Microspheres could be produced with both surfactants, although the yield using SOS
`was very low. The rate of drug release was found to be dependent on the ~oncentration of emulsifier
`used to produce the microspheres. For example, microspheres produced using 0.25% w /v SOS released
`their encapsulated drug instantaneously on dissolution testing, whereas microspheres produced with(cid:173)
`out surfactant provided prolonged release of drug (ts0%> 360 min). Low temperature scanning elec(cid:173)
`tron microscopy revealed for these two systems that the differences in release resulted from different
`degrees of porosity in the microsphere drug-polymer matrix.
`
`Keywords: Microencapsulation; Eudragit™ RS; Colon delivery; 5-Aminosalicylic acid; Porosity
`
`Introduction
`
`Microencapsulation can be employed for the
`production of multiparticulate sustained or
`modified-release drug delivery systems. Several
`publications on the formulation and use of drug(cid:173)
`containing microspheres have utilised
`the
`EudragitTM (Rohm Pharma) series of polymers
`as the encapsulating materials [ 1-5]. The Eu(cid:173)
`dragits are a family of polymers based on acrylic
`and methacrylic acids suitable for use in orally(cid:173)
`administered drug delivery systems. They are
`available in a number of different grades pos-
`
`Correspondence to: M.C. Davies and C.D. Melia, Dept. of
`Phannaceutical Sciences, University of Nottingham, Not(cid:173)
`tingham NG7 2RD, U.K.
`
`sessing a range of physicochemical properties
`[ 6]. Some rapidly dissolve at clearly defined pH
`values whereas two grades, Eudragits RL and RS,
`are insoluble in aqueous media but permeable
`and as such have been shown to be suitable for
`use in sustained-release microencapsulated dos(cid:173)
`age forms [ 1-4]. The permeability of Eudragits
`RL and RS to aqueous solutions is due to the
`presence of quaternary ammonium groups in
`their structure; RL has the greater proportion of
`these groups and as such is more permeable than
`RS [7].
`As part of a research programme investigating
`the controlled delivery of drugs to the colon, we
`have produced Eudragit RS microspheres con(cid:173)
`taining 5-aminosalicylic acid ( 5-ASA), an agent
`active against inflammatory bowel disease [ 8 ] .
`Mylan Exhibit 1023
`
`

`
`312
`
`The colon-specific delivery of 5-ASA is currently
`receiving considerable research interest. Since 5-
`ASA largely is absorbed from the upper intes(cid:173)
`tine, selective delivery into the colon is required
`for it to be therapeutically effective [ 8 J. A num(cid:173)
`ber of approaches have been used to achieve this;
`for example, tablets coated with polymers which
`dissolve at colonic pH levels [ 9) and sustained
`release pellets whose release rate is maximal at
`colonic pH [ 10). Other methods of achieving
`colon-specific delivery exploit the metabolising
`properties of colonic bacteria. Thus delivery sys(cid:173)
`tems activated by bacterial cleavage of azo- bonds
`are under investigation: the azo- bonds have been
`incorporated into both polymers [ 11 , 12 J and
`prodrugs [ 8 ] . Prodrugs containing glycosidic
`linkages also have the potential to provide site(cid:173)
`specific release of drugs into the colon [ 13 ] . This
`paper describes the preparation, characterisa(cid:173)
`tion and release properties of 5-ASA-loaded Eu(cid:173)
`dragit RS microspheres, a potential sustained-re(cid:173)
`lease colonic drug-delivery system, and describes
`how the concentration and type of surfactant
`used for fabrication can modulate microsphere
`porosity and subsequent drug release. These
`microspheres do not have intrinsic colon-spe(cid:173)
`cific release characteristics and thus to maximise
`their therapeutic use they will require delivery
`within a colon-targeted device.
`
`Materials and Methods
`
`Materials
`
`Eudragit RS (Dumas~ U.K.., Tunbridge Wells),
`5-aminosalicylic acid (95-98% purity) (Sigma,
`Poole, U .K..), sodium dodecyl sulphate (Sigma),
`Tween 20 (Sigma), methylene chloride (GPR
`grade) (Rhone-Poulenc, Dagenham, U .K.). All
`reagents were used as received.
`
`Methods
`
`Microsphere preparation
`Microspheres were produced using an emulsi(cid:173)
`fication-solvent evaporation procedure [ 14].
`Eudragit RS was dissolved in methylene chloride
`to produce a l 0% w / v solution. 5-ASA was finely
`
`ground using a pestle and mortar to produce par(cid:173)
`ticles less than 20 µm (as determined by light
`microscopy) and added to the polymer solution
`to form a 5% w/v suspension. The dispersion of
`drug into the polymer solution was aided by son(cid:173)
`ication for 20-30 min. When necessary, extra
`cold water was added to the sonic bath to pre(cid:173)
`vent excessive warming of the drug-polymer so(cid:173)
`lution mixture. The drug-polymer-solvent phase
`was poured into an aqueous solution of either so(cid:173)
`dium dodecyl sulphate (SDS) or Tween 20 with
`agitation using an overhead paddle stirrer ( 200
`rpm). A range of surfactant concentrations were
`employed: 0, 0.05, 0.075, 0.10 and 0.25% w/v
`for SOS and 0.025, 0.05, 0.25, 0.5, and l % w /v
`for Tween 20. Each microsphere formulation was
`produced at least in duplicate. Stirring was con(cid:173)
`tinued at room temperature until evaporation of
`the methylene chloride was complete (typically
`4-5 h). In all cases, drug-polymer agglomerates
`were formed at the liquid surface as the solvent
`evaporation proceeded. This material was de(cid:173)
`canted from the sunken microspheres at the end
`of solvent evaporation. The resulting micro(cid:173)
`spheres were collected by filtration through a 20
`µm filter, washed with 200 ml of water and dried
`overnight in a freeze-drier. A 250- 500 µm size
`fraction of the dried microspheres was collected
`by sieving.
`
`Determination of drug content
`Microspheres were finely ground using a pes(cid:173)
`tle and mortar and accurately weighed quantities
`were transferred to 100 ml flasks. 100 ml of pH
`7 phosphate buffer ( 0.05 M) was added to each
`and the solution was sonicated to ensure disper(cid:173)
`sion of the powdered microspheres. After several
`hours of regular shaking, samples were with(cid:173)
`drawn, passed through a 1 µm membrane filter
`and the UV absorbance measured at 3 30 nm. The
`drug concentration was calculated with refer(cid:173)
`ence to a calibration curve of 5-ASA in pH 7
`phosphate buffer.
`
`Drug release studies
`Profiles of drug release from the microsphert=s
`were obtained using a USP (Apparatus 2) dis(cid:173)
`solution tester. Microsphere samples were placed
`
`

`
`in 500 ml of pH 7 phosphate buffer, containing
`0.02% w /v Tween 20 to aid wetting, and agita(cid:173)
`ted at 100 rpm. 10 ml samples were withdrawn
`from the dissolution flasks at regular intervals,
`passed through a 1 µm membrane filter and the
`UV absorbance at 330 nm measured. The drug
`concentration was calculated from the calibra(cid:173)
`tion curve of pure drug in pH 7 phosphate buffer.
`Any microspheres retained on the filter were re(cid:173)
`turned to the dissolution vessel by washing with
`10 ml of fresh buffer. Into each dissolution vessel
`was placed approximately 120--140 mg of micro(cid:173)
`spheres containing 40-45 mg of 5-ASA. We esti(cid:173)
`mated the solubility of 5-ASA to be at least 2 mg/
`ml at pH 7 and 37 °C, so the amount of drug in
`each dissolution vessel was well below saturation
`level. Dissolution profiles for equivalent concen(cid:173)
`trations of unencapsulated drug were obtained in
`the same way. UV scans of 5-ASA between 200
`and 400 nm remained unchanged over the pe(cid:173)
`riod of the release study. Two dissolution runs
`were carried out
`for each microsphere
`formulation.
`
`I~w temperature scanning electron microscopy
`Electron micrographs were obtained of two
`samples; those produced using no surfactant and
`those produced using 0.25% w /v SOS. The elec(cid:173)
`tron microscope stage was maintained at
`-
`l 80"C with liquid nitrogen in order to min(cid:173)
`imise damage to the samples from electron beam
`irradiation [ 15, 16] . Samples were mounted on
`a holder using carbon cement and then plunged
`into nitrogen slush. For fracturing, the holder was
`placed into the microscope preparation chamber
`and a cold knife run through the sample. Sam(cid:173)
`ples were then moved onto the microscope stage
`and examined at a low kV, while the temperature
`was raised to - 80 ° C to allow sublimation of any
`surface water. The sample was then moved back
`into the preparation chamber and sputter coated
`with gold, prior to return to the stage for imaging.
`
`Results and Discussion
`
`Microspbere production
`
`Microspheres containing 5-ASA could be pro(cid:173)
`duced successfully over the entire concentration
`
`313
`
`range of Tween 20 and SDS examined. In the
`complete absence of surfactant microspheres
`were produced, but the yield was low (see Table
`1 ): at the end of solvent evaporation 20 to 30%
`of the added mass of drug and polymer formed
`into microspheres, the remaining material hav(cid:173)
`ing formed agglomerates. With SDS, yields were
`little different, varying between 20 and 40% and
`were not related to surfactant concentration.
`Likewise for Tween 20, yields were similar at all
`surfactant concentrations although the effi(cid:173)
`ciency of the process was considerably higher
`with between 70 and 80% of the drug and poly(cid:173)
`mer formed into microspheres. The stirring speed
`was chosen to maximise the proportion of
`microspheres between 250 and 500 µm in diam(cid:173)
`eter. Typically at least 30% by weight of the
`microspheres fell within this size range. The
`emulsification procedure was clearly consider(cid:173)
`ably more efficient with Tween 20 as surfactant.
`The reason for the marked difference in yield
`seen between the two surfactants is unclear and
`is currently under investigation.
`Drug encapsulation efficiency was high (the(cid:173)
`oretical maximum loading 33% w /w) and the fi(cid:173)
`nal drug content of the microspheres did not ap-
`
`TABLE I
`
`Drug content and yield of 5-ASA: Eudrogit RS microsphere
`formulations
`
`Surfactant concentration Mean microsphere
`5-ASA content %
`used to produce
`microspheres (% w /v)
`w/ v (n = 3)
`
`No surfactant
`
`29.3± l.3
`
`sos
`0.05
`0.075
`0.10
`0.25
`
`Tween 20
`0.025
`0.05
`0.25
`0.5
`1.0
`
`29.2± 1.5
`30.1±1.5
`30.7± 1.0
`30.7± 1.4
`
`29.5 ± 1.3
`30.1±1.4
`28.9± 1.4
`29.2±0.4
`29.6±2.0
`
`Yield
`(%)
`
`20-30
`
`}
`
`20-40
`
`}
`
`70-80
`
`

`
`314
`
`pear to depend upon the concentration of
`surfactant (Table 1 ).
`An important issue to be considered when us(cid:173)
`ing solvent evaporation techniques for produc(cid:173)
`tion of microspheres is the potential problem of
`residual solvent. Since many of the solvents used,
`including methylene chloride, are toxic, it is im(cid:173)
`portant that levels remaining in the micro(cid:173)
`spheres at the end of production are minimal. It
`has recently been reported that methylene chlo(cid:173)
`ride levels in 50: 50 Eudragit RS: RL micro(cid:173)
`spheres produced by solvent evaporation and
`dried in air at 3 7 ° C were less than 0.1 o/o [ I 7].
`Although we did not measure the residual sol(cid:173)
`vent levels in our microspheres, it would seem
`reasonable to assume that they are of a similar
`order of magnitude.
`
`Drug release characteristics
`
`The drug release profiles from microspheres
`prepared using SDS and Tween 20 as surfactants
`are shown in Figs. I and 2 respectively, and the
`times for 50% of the encapsulated drug to be re(cid:173)
`leased ( ts0%) are given in Table 2. For both SDS
`and Tween 20, the rates of drug release were
`found to be dependent upon the concentration
`of surfactant used in the production procedure,
`with a progressive enhancement in the rate of
`drug release with increasing surfactant concen(cid:173)
`tration. The microspheres made without surfac(cid:173)
`tant showed the greatest retardation in drug
`availability with a t 50% for release greater than 360
`
`.... --.... ---·-----------------. -----------.
`
`100
`
`'O 80
`
`v • • v 60
`i
`~ 40
`10
`'# 20
`
`--·· <1na1.-e
`
`.m . .
`---- o.-
`- - o.n~
`----.-- 0.07ft
`
`60 120 180 240 300 S60
`nme lmlns)
`Fig. l. Effect of concentration Of sodium ~odecylsulphate
`(SOS) on the release rate of 5-ASA from Eudmgit RS
`microspheres.
`
`··-·· --,,.,,._.
`-- ...
`-- ·~
`---- ......
`
`o.._--+-~~~~~~~-
`
`o
`
`6o 120 180 240 300 360
`Time (mina)
`
`r"ig. 2. Effect of concentration of Tween 20 on the release
`rate of 5-ASA from Eudragit RS microspheres.
`
`TABLE 2
`
`Time for 50% of encapsulated 5-ASA to be released ( t SO'llo from
`Eudragit RS microsphere formulations
`
`Surfactant type
`and concentration
`(%w/v)
`sos
`0
`0.05
`0.075
`0.10
`0.25
`Tween 20
`0
`0.025
`0.05
`0.25
`0.5
`1.0
`
`I 5()% for drug
`release
`(min)
`
`>360
`300
`160
`140
`<30
`
`>360
`300
`120
`100
`30
`<30
`
`min. In contrast, microspheres made with 0.25%
`w /v SDS and Io/ow /v Tween 20 showed little or
`no retardation in availability over unencapsu(cid:173)
`lated 5-ASA. Such dramatic changes in drug re(cid:173)
`lease rates with changing SDS and Tween 20
`concentrations suggested that these surfactants
`may have a significant influence on the structure
`of the microsphere drug-polymer matrix.
`
`Low temperature SEM
`
`In order to investigate the influence of surfac(cid:173)
`tant on the structure of the micro:;pheres and
`hence the drug release kinetics, we used low tem(cid:173)
`perature scanning electron microscopy to exam-
`
`

`
`315
`
`Fig. 3. Electron micrographs of microspheres prepared without surfactant. Top left, overall view of microsphere population
`(magnification X50). Top right, detail ofmicrosphere surface (X460). Bottom left, freeze-fractured section ofmicrosphere
`( X260). Bottom right, high magnification view of bottom left x 2600).
`
`ine two microsphere formulations that exhibited
`extreme drug release profiles i.e. a rapidly releas(cid:173)
`ing system produced using 0.25% w/v SDS and
`a sustained-release system produced without
`surfactant.
`The surfactant-free system produced micro(cid:173)
`spheres that were smooth and spheroidal in ap(cid:173)
`pearance (Fig. 3a). Higher magnification micro(cid:173)
`graphs emphasised the smoothness and integrity
`of the surface with only a few small surface pores
`being visible (Fig. 3b ). In comparison, the
`microspheres produced with surfactant were
`roughly spheroidal in shape with an uneven outer
`surface (Fig. 4a). The surface was seen to con(cid:173)
`tain areas of high porosity (Fig. 4b). We esti-
`
`mated these pores to be up to 2 µm in diameter.
`When cryogenically fractured, more dramatic
`structural differences were revealed between the
`two systems. Fig. 3c shows a sectioned view of a
`representative microsphere from the surfactant(cid:173)
`free system. The drug-polymer matrix appeared
`generally dense with a number of large holes
`present. In a magnified view of this section (Fig.
`3d), there appear to be drug crystals visible. In
`contrast, the internal structure of the micro(cid:173)
`spheres produced using 0.25% w / v SOS re(cid:173)
`flected the surface topography. The drug-poly(cid:173)
`mer matrix was extremely open and porous
`resulting in a sponge-like appearance (Fig. 4c).
`Fig. 4d shows a crystal of 5-ASA suspended in
`
`

`
`316
`
`Fig. 4. Electron micrographs of microspheres prepared with 0.25% SDS as surfactant. Top left, overall view of microsphere
`population ( X 50). Top right, detail of microsphere surface ( X 330 ). Bottom left, freeze-fractured section of microsphere ( x 340 ).
`Bottom right, high magnification view of bottom left ( X3200).
`
`this open framework of polymer. There are fewer
`crystals visible in the micrographs of the 0.25%
`w /v SOS system because, although thew /w drug
`loading is equivalent to the surfactant-free sys(cid:173)
`tem, the bulk density is considerably lower due
`to their highly porous nature.
`The SEM work demonstrates that the ex(cid:173)
`tremely rapid release of drug from the 0.25%
`w /v SDS system is a consequence of a highly po(cid:173)
`rous surface and internal structure, which allows
`rapid entry of dissolution medium into and dis(cid:173)
`solved drug out of the microspheres. Conversely,
`in the surfactant-free system the 5-ASA is
`
`embedded in a dense, low porosity matrix. As a
`result, diffusion of aqueous medium into and
`dissolved drug out of the microspheres will be
`slow and to a greater extent through the hydrated
`polymer itself, rather than primarily through
`pores as in the case of the rapidly releasing
`system.
`
`Conclusions
`
`We have demonstrated that 5-ASA can be suc(cid:173)
`cessfully encapsulated into Eudragit RS to pro(cid:173)
`duce microspheres for potential sustained-re-
`
`

`
`lease oral drug-delivery applications. Moreover,
`the rate of drug release can be controlled by the
`concentration and type of surfactant used in the
`manufacturing process. For SDS, the enhance(cid:173)
`ment in release was seen to be due to surfactant(cid:173)
`induced changes in the drug-polymer matrix re(cid:173)
`sulting in increased microsphere porosity. Fur(cid:173)
`ther studies are examining the mechanism by
`which pores are formed in these systems, al(cid:173)
`though they probably arise from the deposition
`of the drug-polymer matrix around droplets of
`aqueous phase which have entered the forming
`microspheres as a result of the action of
`surfactant.
`There were also marked differences in the
`emulsification efficiency between the two sur(cid:173)
`factants, with SDS producing microsphere yields
`as low as 20%, compared to 70% with Tween 20.
`Clearly to be considered as an economically via(cid:173)
`ble process, microsphere yields as low as 20%
`would be unacceptable. However, in this work we
`have made no attempt to optimise the micros(cid:173)
`phere yield by other means. As one example, by
`including baffles in the mixing vessel and thus
`reducing vortex. formation, the yield of micro(cid:173)
`spheres in a solvent evaporation process could
`be increased from 50 to 90% [ 18 ] .
`There are several parameters that can be used
`to produce microsphere systems with a pro(cid:173)
`grammed rate of drug delivery, for example mi(cid:173)
`crosphere size or the level of drug-loading, but if
`there are fixed requirements for these, control
`over release needs to be exercised through addi(cid:173)
`tional parameters such as the nature of the en(cid:173)
`capsulating polymer or the inclusion of additives
`in the microsphere formulation e.g. plasticisers
`[ 14]. We have shown, in addition, that the con(cid:173)
`centration and type of surfactant can be used as
`a controlling factor to modulate the rate of drug
`release from Eudragit RS microspheres pro(cid:173)
`duced by the emulsification-solvent evaporation
`process.
`
`Acknowledgements
`
`We would like to thank Dr. J. Sargent for assis(cid:173)
`tance with the electron microscopy work and
`
`SERC/Reckitt and Colman Pharmaceuticals
`(Dr. G. Parr) for funding PJW.
`
`317
`
`References
`
`M. Kawata, N. Nakamura, S. Goto and T. Aoyama,
`Preparation and dissolution pattern of Eudragit RS mi(cid:173)
`crocapsules containing ketoprofen, Chem. Pharm. Bull.,
`34 ( 1986) 2618-2623.
`2 Y. Pongpaibul, J .C. Price and C.W. Whitworth, Prepa(cid:173)
`ration and evaluation of controlled-release indometha(cid:173)
`cin microspheres, Drug. Dev. Ind. Pharm., 10 ( 1984)
`1597- 1616.
`3 D. Babay, A. Hoffman and S. Benita, Design and release
`kinetic pattern evaluation of indomethacin micro(cid:173)
`spberes, Biomaterials, 9 ( 1988) 482-488.
`4 S. Benita, A. Hoffman and M. Donbrow, Microencap(cid:173)
`sulation of paracetamol using polyacrylate resins (Eu(cid:173)
`dragit Retard) , kinetics of drug release and evaluation
`of kinetic model, J . Phann. Phannacol., 37 ( 1985) 391-
`395.
`5 S. Goto, M. Kawata, M. Nakamura, M. Maekawa and
`T. Aoyama, Eudragit E, Land S (acrylic resins) micro(cid:173)
`capsules as pH-sensitive preparations of ketoprofen, J.
`Microencapsulation, 3 ( 1986) 305-316.
`6 Handbook of Pharmaceutical Excipients, Pharmaceut(cid:173)
`ical Society of Gt. Britain/ American Pharmaceutical
`Association, London/Washington, 1986, pp. 214-215.
`7 Eudragit RL and RS data sheets, Rohm Pharma GmbH,
`Darmstadt, F.R.G.
`8 G. Jame.rot, Newer 5-aminosalicylic acid based drugs in
`chronic inflammatory bowel disease, Drugs, 37 ( 1989)
`73- 86.
`9 M.J. Dew, R.E.J. Ryder, M. Evans, B.K. Evans and J.
`Rhodes, Colonic release of 5-aminosalicylic acid from
`an oral preparation in active ulcerative colitis, Br. J. Clin.
`Phannacol., 16 ( 1983) 185- 187.
`10 S.N. Rasmussen, S. Bondesen, E.F. Hvidberg, S.H.
`Hansen, V. Bonder et al., 5-aminosalicylic acid in a slow
`release preparation: bioavailability, plasma level and
`excretion in humans, Gastroenterology, 83 (1982)
`1062-1070.
`11 M. Saffran, G.S. Kumar, C. Savariar, J.C. Burnham, F.
`Williams and D.C. Neckers, A new approach to the oral
`administration of insulin and other Deptide drugs, Sci(cid:173)
`ence 233 ( 1986) 1081-1084.
`12 H. Brondsted and J. Kopecek, Hydrogels for site-spe(cid:173)
`cific oral drug delivery, Proc. Int. Symp. Control. Rel.
`Bioact. Mater., 17 ( 1990) 128-129.
`13 D.R. Friend and G.W. Chang, Drug glycosides: poten(cid:173)
`tial prodrugs for colon-specific drug delivery, J. Med.
`Chem., 28 ( 1985) 51-57.
`14 P.J. Watts, M.C. Davies and C.D. Melia, Microencap(cid:173)
`sulation using emulsification-solvent evaporation; an
`overview of techniques and applications, CRC Crit. Rev.
`Ther. Drug Carrier Sys., 7 ( 1990) 235- 259.
`
`

`
`318
`
`15
`
`J.A. Sargent, Low temperature scanning electron mi(cid:173)
`croscopy: advantages and limitations, Scanning Micros(cid:173)
`copy, 2 ( 1988) 835- 849.
`16 J .A. Sargent, Cryo-stabilisation of low melting-point
`materials for scanning electron microscopy, Proc. R.
`Microsc. Soc., 23 ( 1988) 275- 281.
`17 A. Barkai, Y. V. Pathak and S. Benita, Polyacrylate (Eu(cid:173)
`dragit Retard) microspheres for oral controlled release
`
`of nifedipine. I. Formulation design and process optim(cid:173)
`isation, Drug Dev. Ind. Pharm., 16 (1990) 2057-2075.
`18 R. Bodmeierand J .W. McGinity, Polylacticacid micro(cid:173)
`spheres containing quinidine base and quinidine sul(cid:173)
`phate prepared by the solvent evaporation technique. I.
`Methods and morphology, J. Microencapsulation, 4
`( 1987) 279-288.