Formulation of epichlorohydrin cross-linked starch microspheres G. HAMDI, G. PONCHEL* and D. DUCHE à NE Laboratoire de Pharmacotechnie et Biopharmacie, URA CNRS 1218 Faculte de Pharmacie, Universite de Paris-Sud, 5, rue J. B. CleÂment 92296 ChaÃtenay- Malabry, France (Received 22 December 1999; revised 2 April 2000; accepted 20 May 2000 ) The present work describes a water/oil emulsion technique for the production of microspheres by cross-linking soluble starch with epichlorohyrin, which is a very e
Transcript
Formulation of epichlorohydrin cross-linked starchmicrospheres
G. HAMDI, G. PONCHEL* and D. DUCHEÃ NE
Laboratoire de Pharmacotechnie et Biopharmacie, URA CNRS 1218 FaculteÂde Pharmacie, Universite de Paris-Sud, 5, rue J. B. Cle ment 92296 Chaà tenay-Malabry, France
(Received 22 December 1999; revised 2 April 2000; accepted 20 May 2000)
The present work describes a water/oil emulsion technique for the productionof microspheres by cross-linking soluble starch with epichlorohyrin, which is avery e� cient divalent cross-linking agent for starch. Because they are importantfeatures for potential applications, such as pulmonary administration, specialattention has been paid to control the mean particle size and size distribution.Microspheres ranging from 0.3±250 mm with narrow size distributions could beobtained. Due to the strongly basic nature of the aqueous phase, no stableemulsions could be obtained in the water/oil emulsion domains. In this context,the stirring rate during the emulsi®cation step was crucial for controlling theparticle size. Additionally, a high organic-to-aqueous phase ratio and thepresence of a surfactant agent helped to prevent the coalescence of the dropletsduring the formation of the microspheres. The process was not sensitive tomodi®cations of the chemical conditions, such as the cross-linking ratio, whichallows variation of the chemical nature of the polymer forming the core of themicrospheres without modifying their morphological characteristics.
Starch microspheres have been proposed for the controlled-delivery of drugsby the nasal route (BjoÈ rk and Edman 1988, 1990, Illum et al. 1990, Edman et al.1992) or for embolization (Lindberg et al. 1984). Starch microspheres have been
shown to be e� ective in the systemic delivery of peptides (BjoÈ rk and Edman 1988)after nasal administration; due to a combination of desirable properties, including:(i) adequate particle size for optimal deposition on the nasal mucosa after
administration, (ii) optimal swelling degree in contact of the mucous layer(Illum et al. 1987), (iii) controlled-delivery of the drug for a period of timecomparable to the duration of the contact of the microspheres with the mucosal
surface, and (iv) bioadhesion on the nasal ciliated mucosal surface (Illum 1987).Pharmaceutical applications of starch miscropsheres necessitate controlled
particle size, and generally narrow size distributions, because the localizationand distribution of the particles in the body depend on these parameters. Thereis an optimum particle size and distribution for each route of administration. For
j. microencapsulation, 2001, vol. 18, no. 3, 373±383
* To whom correspondence should be addressed. e-mail: [email protected]
Journal of Microencapsulation ISSN 0265±2048 print/ISSN 1464±5246 online # 2001 Taylor & Francis Ltdhttp://www.tandf.co.uk/journals
DOI: 10.1080/02652040010019505
Jour
nal o
f M
icro
enca
psul
atio
n D
ownl
oade
d fr
om in
form
ahea
lthca
re.c
om b
y T
echn
isch
e U
nive
rsite
it E
indh
oven
on
06/0
4/14
For
pers
onal
use
onl
y.
example, after intravenous administration, particles greater than 7 mm will becometrapped by mechanical ®ltration in the capillary bed in the lungs, whilst particlessmaller than 6 mm will be taken by the phagocytic cells (Rhodes et al. 1969,Yoshioka et al. 1981). After nasal administration, deposition of dry particles on thenasal mucosa is e� cient for particles between 40±100 mm (Schreier et al. 1993), butparticles smaller than 40 mm have a tendency to reach the respiratory tract directlyafter inspiration without impacting the nasal mucosa (Schreier et al. 1993).
Most studies based on the use of starch microspheres have been made withSpherex 1microspheres, which have been commercialized since 1994 by Pharma-cia. These particles were prepared by the action of epichlorohydrin on partiallyhydrolysed starch (Rothman and Lindberg 1977). They were drug loaded by theuser, generally by a soaking/lyophilization procedure. In this context, the aim ofthe present work was to propose an alternative method for the production ofbiodegradable starch microspheres. Because of the crucial importance of these
parameters for pharmaceutical applications, special attempts have been made tocontrol the mean size and size distribution during the preparation process. In thisrespect, the e� ect of formulation and processing parameters such as surfactantconcentration, organic phase volume, emulsi®cation process and stirring rate wereinvestigated.
Materials and methods
Materials
Soluble starch, i.e. partially hydrolysed starch (Glucidex 61) was supplied byRoquette FreÁ res (Lille, France). Starch molarities were expressed as anhydrogly-cose units (AGU, Mw ˆ 180 g/mol). Epichlorohydrin (ECH), solvents and re-agents were purchased from Prolabo (Paris, France). Sorbitan monooleate (Span80, HLB ˆ 4.7) was supplied by ICI.
Phase diagrams and stability studies
Ternary diagrams corresponding to di� erent systems were obtained by mixingthe di� erent ingradients in the corresponding proportions for a total volume of5 mL in glass tubes using a mechanical stirring device (Top-mix, Prolabo, France)for 3 min. The coalescence of the emulsions was estimated under static conditionsat 25 8C by measuring the respective heights of the di� erent phases, whichappeared as a function of time. The time necessary for observing the coalescenceof half of the initial height of the emulsion (T50%) was used for estimating thestability of the emulsion.
Microsphere preparation and study of the in¯uence of technological parameters
Cross-linked starch microspheres have been prepared according to a water-in-oil emulsion technique using epichlorohydrin as a crosslinking agent (Hamdi et al.1998). For a typical batch, the aqueous phase has been prepared by dissolving 8 gof soluble starch (Glucidex 61) in a 12 g alkaline solution (2 N) under mechanicalstirring which corresponds to 4 mol/L of starch, expressed as anhydroglucose units(AGU). The aqueous phase has been emulsi®ed in 100 ml of a cyclohexane±
chloroform mixture (4:1 v/v) containing varying amounts of sorbitan monooleate
374 G. Hamdi et al.
Jour
nal o
f M
icro
enca
psul
atio
n D
ownl
oade
d fr
om in
form
ahea
lthca
re.c
om b
y T
echn
isch
e U
nive
rsite
it E
indh
oven
on
06/0
4/14
For
pers
onal
use
onl
y.
(Span 80). The ratio of the volume of the organic phase volume to the mass of theaqueous phase mass has been varied. When not mentioned, it was kept constantand equal to 5/1. The mixture was homogenized for 3 min either with a laboratoryblade stirrer or a high-speed mechanical stirrer (Ultra Turrax T25, Janke etHunkel) for stirring higher than 1300 rpm. Varying amounts of epichlorophydrin,expressed as molar ratios of epichlorohydrin (ECH) to starch (expressed as AGU),were added under mechanical stirring at various stirring rates. The stirring wasmaintained for 18 h at 40 8C.
Microspheres were then isolated by centrifugation and washed twice withcyclohexane, and extensively with deionized water. Finally, the microspheres werelyophilized and kept in closed containers before use.
Size measurements
Size distribution analysis was performed using a Coulter Counter1 MultisizerII apparatus (Coulter Counter1, Margency, France). Measurements were madeon very dilute microsphere suspensions in an isotonic solution. Mean diameterswere calculated from the volume size distributions and corresponded to theswollen mean diameters of the miscropheres.
Morphology of the microspheres
A scanning electron miscroscope from Jeol instrument (JSM-840A) was usedto characterize the surface morphology and the shape of the microspheres. Drysamples of microspheres were deposited on double-faced adhesive and coated withpalladium/gold before observation.
Equilibrium swelling degree
Equilibrium swelling degree (ESD) was determined by swelling a suitablevolume of dried crosslinked startch microspheres in 5 ml phosphate bu� er salinepH ˆ 7.4 (PBS) overnight in a graduated glass. The volume of swollen micro-spheres was read directly. The ESD was expressed as the ratio of the swollenvolume VS to the mass of dried starch micropheres md.
Results and discussion
Numerous chemical modi®cations of starch have been described because thereis a need to prevent or to control the enzymatic degradation of starch and toimprove certain physico-chemical properties. Chemical modi®cations of starchhave been extensively studied from a chemical point of view (Kartha andSrivastava 1985, SÏ imkovic et al. 1996), for pharmaceutical applications (Bjorkand Edman 1990, Illum et al. 1990, Edman et al. 1992) as well as chromatographyanalysis (Weber et al. 1986, Chaudhari et al. 1989). About 20 years ago, a patentdescribed the preparation of starch microspheres using an inverse emulsion
technique (Rothman and Lindberg 1977). They were commercialized under thetrade name of Spherex1 and received considerable attention as drug carriers forvarious routes of administration. Since that time, and despite the considerable
literature concerning the preparation and characterization of microspheres, there
are only a few works describing the preparation of starch microspheres from non-chemically modi®ed starch (Chaudhari et al. 1989, Jiugao 1994).
The present work describes a technique which enables the production ofmicrospheres by cross-linking soluble starch with epichlorohydrin, which is avery e� cient divalent cross-linking agent for starch (Kartha and Srivastava 1985).
Special attention has been paid to control the mean particle size and size distri-bution, because they are important characteristics for pharmaceutical applications.
Phase diagrams and stability of emulsions
The water/oil emulsion used a sodium hydroxide solution of soluble starch asthe dispersed phase and an organic solvent containing a lipophilic surfactant as a
continuous phase. Phase diagrams were determined for di� erent ternary systems.Ternary diagrams were obtained for Isopar M/Span 80/Water and Isopar M/Span80/sodium hydroxide starch solutions, respectively. A wide region of stability wasobserved for the water/oil emulsion area corresponding either to stable emulsion or
to stable emulsion in the presence of an excess of aqueous phase (®gure 1).However, no stability was observed in the whole phase diagram when the aqueousphase was replaced by a solution of starch (4 moles AGU/L) in 2 N sodium
hydroxide, suggesting that sodium hydroxide had a negative e� ect on the stabilityof the emulsions.
In order to evaluate this phenomenon, the in¯uence of sodium hydroxideconcentration on the stability of the system has been studied. As illustrated in
®gure 2, the stability of the emulsion depended substantially on the concentrationof sodium hydroxide in the aqueous phase. The addition of 1 mole/L of sodiumhydroxide was su� cient to decrease the half-life of the droplets to a few minutes
only. This was probably due to a masking of the surface activity of the surfactant ina strongly basic environment. A salting-out e� ect has been proposed for explainingthis lack of stability. According to Shinoda and Takeda (1970), the addition of
sodium hydroxide to the aqueous phase would result in a decrease in the a� nity ofwater for the hydrophilic head of Span 80, leading to modi®cations of the surfacetension at the water±organic phase interface.
The e� cient cross-linking of starch by epichlorohydrin requires a concentra-tion of sodium hydroxide higher than 1 mole/L. Despite the fact that no stable
376 G. Hamdi et al.
Figure 1. Phase diagrams of the ternary systems for Isopar M/Span 80/water. (a) highstability of the emulsion; (b) high stability of the emulsion (with aqueous phase inexcess); and (c) unstable system.
Jour
nal o
f M
icro
enca
psul
atio
n D
ownl
oade
d fr
om in
form
ahea
lthca
re.c
om b
y T
echn
isch
e U
nive
rsite
it E
indh
oven
on
06/0
4/14
For
pers
onal
use
onl
y.
emulsions could be obtained for compositions corresponding to the ones required
by the cross-linking process, preliminary trials showed that it was feasible toprepare microspheres under condtions of instability of the emulsions. Never-
theless, the washing of the miscrospheres obtained by this technique was incom-plete because of the di� culty of removing completely the oily Isopar M from the
hydrophilic microspheres.As an alternative to Isopar M, a cyclohexane-chloroform 4:1 (v/v) mixture has
been selected. Figure 3 shows the ternary diagrams obtained for cyclohexane-chloroform (4:1)/Span 80/sodium hydroxide starch solution. The area of water/oil
emulsions corresponded to unstable emulsions, except a portion of this area whichcorresponded to an increase in the viscosity of the continuous phase due to high
concentrations in Span 80.
The possibility of preparing microspheres has been investigated for threedi� erent stability conditions by varying the amount of surfactant (table 1). This
was feasible for relatively low concentrations in Span 80 (2% Span; point A on thephase diagram and 10% Span 80; point B on the phase diagram) corresponding to
the area of instability. On the contrary, for high Span 80 concentration (30% Span
80, point C on the phase diagram) very small particles were obtained (smaller than
Figure 2. Stability of an emulsion composed of 20 g of sodium hydroxide solution in100 mL of Isopar M containing 2% of Span 80 as a function of the sodium hydroxideconcentration in the aqueous phase.
Figure 3. Phase diagrams of the ternary systems for cyclohexane-chloroform 4:1/Span 80/starch solution (4 moles AGU/L) in 2N sodium hydroxide.
Jour
nal o
f M
icro
enca
psul
atio
n D
ownl
oade
d fr
om in
form
ahea
lthca
re.c
om b
y T
echn
isch
e U
nive
rsite
it E
indh
oven
on
06/0
4/14
For
pers
onal
use
onl
y.
1 mm in size as observed from SEM microphotographs) but their aggregation made
their isolation impossible.No stable emulsions suitable for the preparation of starch microspheres could
be obtained either with Isopar M or cyclohexane/chloroform. As a consequence,for further studies, microspheres were prepared from instable emulsions.
In¯uence of emulsi®cation stirring rate before and during crosslinking
The aqueous phase was added under varying stirring rates to the organic phase
before adding the crosslinking agent (emulsi®cation step). After that, the cross-linking reaction was made under mechanical stirring (crosslinking step).
In order to verify the e� ect of the stirring rate before (emulsi®cation step) andduring the crosslinking step, starch microspheres have been made under the same
experimental conditions, but under varying stirring rates for each step.The widths of the size distribution as well as the mean diameter were
considerably in¯uenced by the stirring rate during the emulsi®cation step (®gure4). An increase in the stirring rate during the emulsi®cation step resulted in a
decrease in the mean diameter and in the narrowing of the size distributions. Meandiameters ranged from 138 to 3.6 mm for stirring rates ranging from 600±24 000 rpm. Broad distributions were obtained for 600 and 1800 rpm, whereasalmost monodisperse populations were observed when the stirring rate was24 000 rpm. Interestingly, very small particles (4 mm in the swollen state) could
be obtained when the stirring rate during the emulsi®cation step was 24 000 rpm.The stirring rate during the crosslinking step has almost no in¯uence on the
®nal size of the microspheres. The mean diameter of microspheres remainedpractically constant when the stirring rate was varied from 600 to 1800 rpm
(stirring rate during the emulsi®cation process: 24 000 rpm), suggesting that theaddition of epichlorohydrin in the medium was able to immediately rigidify thestarch-containing water droplets in the emulsion and to prevent them fromcoalescence and aggregation. As shown in many cases (Ishizaka et al. 1981,
Reddy et al. 1990, Sandraps and MoeÈ s 1993), the inverse relationship betweenthe size of the microspheres and the stirring rate during the emulsi®cation step wasdue to the mechanical breaking up of the aqueous phase into smaller sizeddroplets. As shown by the absence of e� ect of the stirring rate after addition of
epichlorohydrin, it was likely that the cross-linking of starch prevented the
378 G. Hamdi et al.
Table 1. Starch microspheres produced in di� erent stability regions of the ternarydiagram displayed in ®gure 3.
²a: Viscosity of the oil phase (mixture of cyclohexane/chloroform and Span 80).* Too high viscosity.Preparation parameters: starch concentration: 4 moles AGU/L; epichlorohydrin-to-
starch molar ratio [ECH/AGU]: 0.58 mol/mol; stirring rate during the cross-linking step:600 rpm, stirring rate during the emulsi®cation step: 9500 rpm.
Jour
nal o
f M
icro
enca
psul
atio
n D
ownl
oade
d fr
om in
form
ahea
lthca
re.c
om b
y T
echn
isch
e U
nive
rsite
it E
indh
oven
on
06/0
4/14
For
pers
onal
use
onl
y.
coalescence of the droplets. Therefore, it was only necessary to keep in suspension
the droplets already formed during the crosslinking process until the hardening of
these droplets by covalent crosslinking and the formation of the ®nal microspheres.Finally, it should be noted that the role of the surfactant in the break up of the
droplets was secondary.
SEM microphotographs show that microspheres were spherical in shape and
presented a smooth surface whatever the stirring rate (®gure 5). These micro-photographs con®rmed qualitatively the decrease in polydispersity when the
stirring rate was increased during the emulsi®cation phase.
E� ect of the organic to aqueous phase ratio
The e� ect of the organic-to-aqueous phase ratio on the mean diameter has been
studied from 1±12 without varying the concentrations of the polymer or surfactant.The stirring rate during the emulsi®cation step was ®xed to 9500 rpm. It can be
observed from ®gure 6 that the mean diameter decreased when the organic-to-
aqueous phase ratio was increased. The mean diameter (§SD) ranged from
4 § 0.5 mm for a 12/1 ratio to 60 § 6.3 mm for a 1/1 ratio. During the emulsi®cationstep, the size of the droplets and their distribution were the result of a dynamic
balance between coalescence and redispersion. The increase in the mean diameter
is a result of the coalescence of droplets due to the probability of a collisionbetween two or more droplets. The e� ect of the organic-to-aqueous phase ratio
was less sensitive when the ratio was higher than 5/1. This ratio was adopted for
further studies. As observed from optical microscopy, all the microspheres
obtained by varying the ratio of organic phase to aqueous phase were spherical
Figure 4. Size distributions of starch microspheres obtained at di� erent stirring rateduring the emulsi®cation step: (a) 600 rpm, (b) 1800 rpm, (c) 9500 rpm, and(d ) 24 000 rpm. Preparation parameters: starch concentration: 4 moles AGU/L; Span80: 2% (v/v of organic phase); organic-to-aqueous phase ratio: 5/1; epichlorohydrinmolar ratio: 0.58 ECH/AGU; stirring rate during the cross-linking step: 600 rpm.
Jour
nal o
f M
icro
enca
psul
atio
n D
ownl
oade
d fr
om in
form
ahea
lthca
re.c
om b
y T
echn
isch
e U
nive
rsite
it E
indh
oven
on
06/0
4/14
For
pers
onal
use
onl
y.
380 G. Hamdi et al.
Figure 5. SEM microphotographs of starch microspheres obtained at di� erent emulsi®ca-tion stirring rates: (a) 600 rpm, (b) 1800 rpm, (c) 9500 rpm, and (d ) 24 000rpm.Preparation parameters: starch concentration: 4 moles AGU; Span 80: 2% (v/volumeof organic phase); organic-to-aqueous phase ratio: 5/1; epichlorohydrin molar ratio:0.58 ECH/AGU; stirring rate during the cross-linking step: 600 rpm.
(a) (b)
(c) (d )
Figure 6. In¯uence of the organic phase to aqueous phase ratio on the mean diameter ofstarch microspheres (mean of three measurements § SD). Preparation parameters:starch concentration: 4 moles AGU; Span 80: 2% (v/v of organic phase); epichloro-hydrin molar ratio: 0.58 ECH/AGU; stirring rate during the emulsi®cation step:9500 rpm; stirring rate during the cross-linking step: 600 rpm.
Jour
nal o
f M
icro
enca
psul
atio
n D
ownl
oade
d fr
om in
form
ahea
lthca
re.c
om b
y T
echn
isch
e U
nive
rsite
it E
indh
oven
on
06/0
4/14
For
pers
onal
use
onl
y.
and exhibited smooth surfaces. However, some aggregations were noticed whenthe organic-to-aqueous phase ratio was close to 1.
In¯uence of the surfactant concentration
In preliminary experiments, it was observed that in the absence of surfactantsthe preparation of microspheres was not successful and resulted in aggregation. Inorder to eliminate these aggregates, it was necessary to add a minimal surfactantconcentration of 0.5%. The e� ect of the surfactant concentration (ranging from0.5±5%) on the mean diameter and the size distribution of starch microspheresprepared at emulsi®cation stirring rate of 9500 rpm has been determined. Anincrease in the surfactant concentration resulted in a slight decrease in the meandiameter of the microspheres, whilst a narrowing of the size distributions wasobserved. The mean diameter of starch microspheres prepared with an emulsi®ca-tion stirring rate of 9500 rpm, was decreased from 10.6 to 7.2 mm and the mode ofthe distributions was shifted to smaller particles from 13.5 to 3.7 mm, respectively,when the surfactant concentration was increased by 10 times (from 0.5 to 5%).However, as expected, the in¯uence of the surfactant was attenuated when thestirring rate during the emulsi®cation step was increased. In the present case, thesurfactant was not able to favour the formation of small droplets during theemulsi®cation process, as demonstrated by the lack of stability of the emulsions.More likely, it could form a ®lm around the droplets, which prevented theircoalescence possibly due to the Gibbs-Marangoni e� ect (Walstra 1993). As aresult, the size distributions were narrowed due to the reduced possibility ofcoalescence of the droplets during the formation of the miscrospheres.
E� ect of epichlorohydrin/starch molar ratio
The epichlorohydrin/starch molar ratio did not a� ect the mean diameter of theswollen starch microspheres, whilst the swelling volume degree of the micro-spheres could be considerably altered. An increase in the epichlorohydrin/starchratio from 0.5 to 2.33 (mol/mol) reduced the ability of the microspheres to swelland the swelling degree decreased from 13.6 to 2 mL/g (table 2).
ESD ˆ Vs=md…ml=g†
It is important to point out that chemical modi®cations could be made withoutin¯uencing the mechanism of formation of the microspheres. Changes in epi-chlorohydrin concentration did not a� ect the size of the swollen microspheres.
Table 2. In¯uence of epichlorohydrin/starch ratio on the mean diameter of the starchmiscrospheres. Preparation parameters: starch concentration: 4 moles of AGU/L;Span 80: 2% (v/volume of organic phase); organic-to-aqueous phase ratio: 5/1;stirring rate during the emulsi®cation step: 24 000 rpm; stirring rate during the cross-linking step: 600 rpm.
ECH/AGU Swelling volume degree Mean diameter § SD(mol/mol) (ml/g of dried microspheres) (mm)
However, as expected, a decrease in epichlorohydrin concentration lead to anincrease in the swelling degree of the starch microspheres. This result suggests thatthe nature of the polymeric network constituting the core of the microspherescould be adjusted by varying the chemical parameters but without modifying thepreparation process or inducing substantial changes in the morphological char-acteristics of the microspheres (Hamdi and Ponchel 1999).
Finally, it can be concluded that the present technique was suitable for thepreparation of microspheres by the cross-linking of starch by epichlorohydrin inan unstable inverse emulsion system. Because the method proposed allowsrelatively easy production of particles smaller than 5 mm in the swollen state,new applications such as pulmonary administration can be foreseen.
References
BjOÈ rk, E., and Edman, P., 1988, Degradable starch microspheres as a nasal delivery systemfor insulin. International Journal of Pharmaceutics, 47, 233±238.
BjOÈ rk, E., and Edman, P., 1990, Characterization of degradable starch microspheres as anasal delivery for drugs. International Journal of Pharmaceutics, 62, 187±192.
Chaudhari, M. R., Kamath, N. D., Bhide, S. V., and Kate, N. R., 1989, Preparation ofcrosslinked starch bead as a medium of gel ®ltration. Die St �arke, 27, 415±416.
Edman, P., BjOÈ rk, E., and RydEÂ n, L., 1992, Microspheres as a nasal system for peptidedrugs. Journal of Controlled Release, 21, 165±172.
Hamdi, G., and Ponchel, G., 1999. Enzymatic degradation of epichlorohydrin crosslinkedstarch microspheres by ¬-amylase. Pharmaceutical Research, 16, 867±875.
Hamdi, G., Ponchel, G., and DuchEÃ ne, D., 1998, An original method for studying in vitrothe enzymatic degradation of cross-linked starch microspheres. Journal of ControlledRelease, 55, 193.
Illum, L., 1987, Drug delivery systems for nasal application. S. T. P. Pharma Sciences, 3,594±598.
Illum, L., Farraj, N. F., Davis, S. S., Johansen, B. R., and O’Hagan, D. T., 1990,Investigation of the nasal absorption of biosynthetic human growth hormone insheep use of a bioadhesion microsphere delivery system. International Journal ofPharmaceutics, 63, 207±211.
Illum, L., Jérgensen, H., Bisgaard, H., Krogsgaard, O., and Rossing, N., 1987,Bioadhesive microspheres as a potential nasal drug delivery system. InternationalJournal of Pharmaceutics, 39, 189±199, 201.
Ishizaka, T., Endo, K., and Koishi, M., 1981, Preparation of egg albumin microcapsulesand microspheres. Journal of Pharmaceutical Science, 70, 358±363.
Jiugao, Y., and Jie, L., 1994, E� ect of suspension crosslinking reacting conditions on thesizes of starch microspheres. Die St �arke, 46, S252±S255.
Kartha, P. R., and Srivastava, A., 1985, Reaction of epichlorohydrin with carbohydratepolymers: Part II. Starch reaction mechanism and physicochemical properties ofmodi®ed starch. Die St �arke, 37, S297±S306.
Lindberg, B., Lote, K., and Teder, H., 1984, Biodegradable starch microspheresÐA newmedical tool. In Microspheres and Drug Therapy edited by S. S. Davis, L. Illum,J. G. McVie and E. Tomlinson (Amsterdam: Elsevier), pp. 153±188.
Reddy, B., Dorle, A. K., and Krishna, D. R., 1990, Albumin microspheres: e� ect ofprocess variables on size of distribution and in vitro release. Drug Developments inIndustrial Pharmacy, 16, 1791±1803.
Rhodes, B. A., Zolle, I., Buchanan, J., and Wagner, H. N. Jr., 1969, Preparation albuminmicrospheres for studies of the pulmonary circulation. Radiology, 92, 1453±1460.
Rothman, U. S. E., and Lindberg, B. J., 1977, Patent No 75 17632.Sandraps, P., and MoEÈ s, A. J., 1993, In¯uence of manufacturing parameters on the size
characteristics and the release pro®les of nifedipine from poly(d,l-lactide-co-glycolide) miscrospheres. International Journal of Pharmaceutics, 98, 157±164.
382 G. Hamdi et al.
Jour
nal o
f M
icro
enca
psul
atio
n D
ownl
oade
d fr
om in
form
ahea
lthca
re.c
om b
y T
echn
isch
e U
nive
rsite
it E
indh
oven
on
06/0
4/14
For
pers
onal
use
onl
y.
Schreier, H., Gonzalez-Rothin, R., and Stecenko, A., 1993, Pulmonary delivery ofliposomes. 209, 209±223.
Shinoda, K., and Takeda, H., 1970, The e� ect of added salts in water on the hydrophile-lipophile balance of non-ionic surfactant: the e� ect of added salts on the phaseinversion temperature of emulsions. Journal of Colloid and Interface Science, 32,642±646.
SÏ imkovic, I., Laszlo, J. A., and Thompson, A. R., 1996, Preparation of a weakly basic ionexchange by crosslinking starch with epichlorohydrin in the presence of NH4OH.Carbohydrate Polymers, 30, 25±30.
Walstra, P., 1993, Principles of emulsion formation. Chemical Engineering Science, 48, 333±349.
Weber, M., Colombel, C., Darzens, D., Foglietti, M.-J., and Percheron, F., 1986,Improved method for sequential puri®cation of polysaccharides by a� nitychromatography. Journal of Chromatography, 355, 456±462.
Yoshioka, T., Hashida, M., Muranishi, S., and Sezaki, H., 1981, Speci®c delivery ofmitomycin C to the liver, spleen and lung: nano- and microspherical carriers ofgelatine. International Journal of Pharmaceutics, 81, 618±620.
