Hollow Microspheres for Use as a Floating Controlled Drug Delivery System in the Stomach Y. KAWASHIMA”, T. NIWA*,H. TAKEUCHI’, T. HINO*,AND Y . ITOH* Received June 1, 1990, from ‘Gifu.Pharmaceutical University, 5-6-1, Mitahora-hi ashi, Gifu 502, Japan, and SShowa Yakuhin Kako Company, Ltd., 41 1, Sakaue-cho, Seto-shr, Achi 489, Japan. Accepted for publication darch 21, 1991. Abstract 0 Hollow microspheres (microballoons).loaded with drug in their outer polymer shells, were prepared by a novel emulsion-solvent diffusion method. The ethano1:dichloromethanesolution of drug (tranilast or ibuprofen)and an enteric acrylic polymer were poured into an agitated aqueous solution of polyvinyl alcohol that was thermally controlled at 40 “C.The gas phase generated in the dispersed polymer droplet by the evaporation of dichloromethane formed an internal cavity in the microsphere of the polymer with the drug. The drugs incorporated in the solidified shell of the polymer were found to be partially or completely amorphous.The flowability and packabilityof the resultant microballoons were much improved compared with the raw crystals of drug. The microballoonsfloated continuously over the surface of acidic dissolution media containing surfactant for >12 h in vitro. The drug release behavior of the microballoonswas characterized as an enteric property, and drug release rates were drastically reduced depending on the polymer concentration at pH 6.8.
Orally administered drugs which are promptly absorbed in the alimentary canal and rapidly lost from the blood are generally film coated or microencapsulated to prolong the drug release period and drug action.13 However, most of these forms have several physiological limitations, such as the gastrointestinal (GI) transit time;4 incomplete drug release from devices or too short residence time of the pharmaceutical dosage forms in the upper position of small intestine (the absorption region of GI tract) leads to low bioavailability of the sustained-release dosage forms. Even if slow release of drug is attained, the drug released after passing the absorption site is not utilized, thus lowering the efficacy of the drug. To overcome this problem, several attempts have recently been made to extend the GI transit time of devices. One approach is bioadhesive systems that “stick dosage forms to the mucin-epithelial cell surface, providing longer transit time due to adhesion of the device to the gastric wall.5 Such adhesion may cause problems, such as irritation to the mucosa if an overdose of the drug occurs locally. Another suggestion is floating dosage forms. These have a specific density that is lower than that of gastric fluids; they remain buoyant in the stomach contents. Most of the floating systems are dominated by single-unit formulations [e.g., so-called hydrodynamically balanced systems (HBS)I.e A drawback of this system is the high variability of the GI transit time, due to its all-or-nothing emptying process.7.S Therefore, a multiple-unit floating system which can be distributed widely throughout the GI tract, providing a possibility of achieving a longer-lasting and more reliable release of drugs, has been sought. To achieve this goal, a novel method to prepare floating microspheres loaded with drug was developed as a modification of the emulsion-solvent diffusion method for preparation of spherical polymeric microsponges for a controlled drug delivery system previously devised by the present authors.9.10 The device prepared with the present technique is a polymeric 0022-3549/92’0200-0 7 35$02.50/0 0 1992. American Pharmaceutical Association
microsphere with a round cavity. This microsphere was termed “microballoon” due to its characteristic internal hollow structure and excellent floatability in vitro. Tranilast, an oral antiallergic agent, or ibuprofen, a n oral antiinflammatory agent, was loaded in the outer shell of the microballoon. In this study, the microballoon preparation process and the physicochemical properties of the microballoons, such as particle diameter, particle density, and crystalline form of drug, were investigated. Furthermore, the floating and the drug release behavior of the microballoons in vitro was clarified.
Experimental Section Preparation of Microballoons of Tranilast or Ibuprofen with Eudragit &The outline of the procedure for the preparation of microballoons was the same as the method described previously.9 Tranilast (Siratori Pharmaceutical Company, Ltd., Chiba, Japan; 0.5 g) or ibuprofen (Taito Koeki, Toyama, Japan; 0.25-0.5 g) and Eudragit S (Mhm Pharma GmbH, Germany; 1.0 g), a n enteric polymer soluble at pH >7.0, were dissolved in the ethano1:dichloromethane mixture (1:l v/v, 10 mL) a t room temperature. The drug solution was poured into 200 mL of water containing 0.75% (w/v) polyvinyl alcohol (PVA-120; Kuraray Company, Ltd., Tokyo, Japan) that was thermally controlled at 40 “C.Then, the solution was stirred with a propeller-type agitator at 300 rpm. The finely dispersed droplets of the polymer solution of drug were solidified in the aqueous phase via diffusion of the solvent. The dichloromethane that evaporated from the solidified droplet was removed by an aspirator, leaving the cavity of the microsphere filled with water. After agitating the system for 60 min, the microspheres were filtered, washed with water, and dried in a n oven (KCV-4D, Advantec, Tokyo, Japan) at 40 or 120°C for 2 h. During the drying procedure, a hollow cavity was formed inside the microsphere, resulting in the microballoon. Measurement of the Diffusion of Solvent from the Dispersed Droplets into Water-The diffusion behaviors of ethanol and dichloromethane from the dispersed droplets of drug solution into the aqueous phase were investigated during the preparation of microballoons containing tranilast. At appropriate intervals, 1mL of external aqueous phase was withdrawn by using a syringe connected with a filter. The ethanol and dichloromethane dissolved in the aqueous medium were detected by a gas chromatograph (model GC-l4A, Shimadzu, Japan) with a flame ionization detector. Model Experiment for the Formation of Microballoons-A large droplet of the ethano1:dichloromethane solution was poured through a pipette into water in a petri dish that was thermally controlled at 40°C, thereby dissolving the drug and polymer. The solidifying behavior of the droplet in water was investigated photographically. Identification of Crystalline Form of t h e Drug in Microballoons-The crystalline form of the drug dispersed in the crust of the microballoon was analyzed by X-ray powder difiactometry (RAD-lC, Rigaku, Tokyo, Japan) and differential scanning calorimetry (DSC; model CN808521, Rigaku, Tokyo, Japan) according to our previous report.11 Measurement of Micromeritic Properties of Microballoons The surface morphologies and the internal textures of the microballoons were observed by a scanning electron microscope (JSM-T330A, Nihon Denshi, Tokyo, Japan). The average diameter ofmicroballoons was represented by the geometric mean diameter obtained by a sieve Journal of Pharmaceutical Sciences / 135 Vol. 81, No 2, February 1992
method. The flov~ and packing properties were investigated by measuring the angle of repose and tapped density. represented by eq 1, of the microballoons was The porosity (4, determined by measuring the true (A) and particle densities (6) of massed microballoons:
-.25 I h
E
cn E
v
E =
(1 - pdpt) x 100
c
(1)
.-0
+J m L
* c
't 15
0 -*A*--_-_---*--------a-
_......-....................................................................
a,
(2)
V
c
0
where V and W are volume and weight of a microballoon, respectively. The true density was measured by using a helium-air pycnometer (model 1302, Micromeritics Instrument Company). The particle density was determined by measuring the mass volume and weight of the microballoons sampled. The mass volume was measured by both a photographic counting method and a water displacement method. Heywood's diameters ( ~ 0 1 2of a number (n) of photographs of microballoons were measured with an image analyzer (IBAS, KarlZeiss, Germany), and the mass volume was calculated by the following equation:
(3) The water displacement method measured the volume by using a pycnometer with an aqueous solution of Tween 20 (0.0296, W/V; polyoxyethylene sorbitan monolaurate; Kishida, Osaka, Japan) at 20 "C. Measurement of Drug Release Rate from Microballoone-The drug dissolution testa of microballoons were carried out by the paddle method specified in the U. S. Pharmacopeia XXI. Microballoons that had been fractionated and weighed to correspond to 50 mg of tranilast or 200 mg of ibuprofen were gently spread over the surface of 900 mL ofdissolution medium (disintegration test solution no. 2,with pH 6.8, and phosphate buffer, with pH 7.2, as specified in the Japanese Pharmacopeia XI and corresponding to the U. S. Pharmacopeia XXI), rotated at 100 rpm, and thermostatically controlled a t 37 "C. Perfect sink conditions prevailed during the drug diasolution testa. The sample was withdrawn at a suitable interval from the dissolution vessel and was assayed spectrophotometrically at 335 nm for tranilast and at 220 nm for ibuprofen (model 320, Hitachi Manufacturing, Tokyo, Japan).
Results and Discussion Mechanism of Formation of Microballoons by EmulsionSolvent Diffusion Method-The ethano1:dichloromethane solution with drug and polymer (Eudragit S), when poured into aqueous solution with stirring, was finely dispersed into discrete droplets, thereby forming an O N type emulsion. The concentrations of ethanol and dichloromethane contained in the external aqueous phase during the preparation of microballoons were monitored, as shown in Figure l.The dashed and dotted lines represent the equilibrium concentrations of ethanol (19.8 mg/mL) and dichloromethane (12.7 mg/mL) in the aqueous phase, respectively, when 10 mL of the twosolvent mixture ( l : l , v/v) was saturated with 200 mL of the aqueous solution of polyvinyl alcohol in a closed system (i.e., glass-stopped flask). Several seconds after the dispersion of the solvent mixture, all ethanol diffused out of the dispersed droplets into the aqueous medium, and then the equilibrium concentration of ethanol was retained during the preparation of the microballoons. In contrast, the dichloromethane did not diffuse thoroughly from the droplets into the aqueous phase but partly resided in the droplets. During the preparation, the concentration of dichloromethane in the aqueous medium decreased due to evaporation from the system. The rapid diffusion of ethanol (good solvent for the polymer) from the droplets into the aqueous phase might reduce the solubility of the polymer in the droplet because the polymer (Eudragit S) 130 / Journal of Pharmaceutical Sciences Vol. 87, No 2, February 1992
0
0
1 20 30 40 50 60 Agitation Time (rnin) Figure 1-Diffusion and evaporation profiles of ethanol (0)and dichloromethane (A)into and out of the external aqueous phase during the preparation of microballoons containing tranilast. Dashed and dotted lines represent the expected equilibrium concentrations of ethanol (19.8 mg/mL) and dichloromethane (12.7 mg/mL), respectively. 0
10
was insoluble in dichloromethane. The polymer precipitated instantly a t the surface of the droplet, forming a film-like shell enclosing the dichloromethane which was dissolving the drug. The mechanism of formation of the microballoon was photographically investigated with a model system in which the drug and polymer solution was dropped through a pipette into water in a petri dish that was thermally controlled at 40 "C, as shown in Figure 2. The poured droplet settled to the bottom of the dish and partly flattened due to gravity. The surface of the droplet became immediately opaque due to the formation of a gel-like film on the surface (Figure 2A). After a few minutes, the gel-like droplet was squeezed by a pincette, and a few bubbles of dichloromethane appeared, as shown in Figures 2B and 2C. This finding indicates that the gas phase of dichloromethane was generated in the droplet covered with gel-like film before the droplet was entirely solidified. The gas phaae should produce a cavity in the microsphere. The mechanism of formation of the microballoons is illustrated in Scheme I, based on the results in Figures 1 and 2. The mechanism proposed in the present study is clearly different from that of the emulsion-solvent evaporation method reported by Benita et al.,13 in which the emulsion droplet was solidified principally by the evaporation of solvent from the droplet. In the present system, the diffusion of ethanol (good solvent for Eudragit S)preceded the evaporation of dichloromethane (poor solvent for the polymer) from the emulsion droplet into the aqueous medium, drastically reducing the solubility of polymer in the droplet and forming a gel-like film on the surface. The mechanically strong A
B
C
2m
Figure 2-Microphotographs of the droplet of ethanokdichloromethane solution of Eudragit S and tranilast in water (40"C)in a petri dish, before (A) and after (B,C) the squeeze by a pincette.
0
om* F o r s a i l o n cf
enulsicr
9/W
emuisicr,
==+
3 Fzrma:;cn
she! I
Of
Generation o f 92s
s l o n of E t G H
Phase
Mlcroballoon
diffusioi o f CH2C12
Scheme I-Preparation procedure and mechanism of microballoon formation by the emulsion-solvent diffusion method.
solidified film (shell) produced at the surface of droplet with further depletion of ethanol prevented rupture and shrinkage of the microsphere during the evaporation of dichloromethane from the droplet; this was shown in the model experiment. The cavity produced by the gas phase was gradually filled with water due to the reduced pressure inside the droplet that was caused by evaporation of dichloromethane. The microballoon was prepared by removing water from the cavity of the microsphere with a drying process. The presence of totally hydrolyzed polyvinyl alcohol, employed as an emulsifying agent, significantly prevented aggregation of the droplets with solidified outer shells during the process. It was assumed that the polyvinyl alcohol, adsorbed to the interface between the droplets, and the aqueous medium might form a strong hydrated layer over the droplets, resulting in a stable dispersed system. Moreover, the rapid formation of film-like shell over the droplets, as shown in the model experiment, prevented the coalescence of the droplets. Therefore, the size distributions of microballoons might coincide with those of droplets formed at the pouring stage of drug and polymer solution. It was found that the main factors determining the sizes of microballoons were the concentrations of drug and polymer in the formulation, the pouring speed of drug and polymer solution, and the agitation speed of the system. In Figure 3, the average diameters and the recoveries of microballoons of tranilast are exhibited as a function of temperature of the aqueous dispersing medium. The recovery was represented by the weight percent of microballoons recovered having a diameter of c1410 pm, to 600 I
I100
I
exclude aggregates, versus the weights of drug and polymer loaded in the system. When the temperature was controlled in the range of 30 to 40 "C, the recovery of microballoonsreached 70-80%, indicating the optimized condition. Above 40 "C, the recovery dramatically decreased since the shell burst and the ruptured microspheres aggregated to a larger mass. At higher temperatures, the emulsion droplets were unstable due to the intense evaporation of dichloromethane. The recovery gradually decreased with decreasing temperature because of the increase in the number of aggregates due to the bridging action of polyvinyl alcohol, which coalesced the particles at the washing process.14 Therefore, the temperature of aqueous dispersing medium was maintained a t 40 "C. No free crystals of drug appeared in the aqueous medium, suggesting that the drug content in the microballoons coincided with that in the formulation. It was found that the drug content of the microballoons was >95% of the formulation value. Physicochemical Identification and Micromeritic P r o p erties of Microballoons-The X-ray powder diffraction patterns and DSC thermograms of microballoons, along with those of physical mixtures of raw crystals of drug and polymer, are shown in Figures 4 (tranilast)and 5 (ibuprofen). It was found that tranilast in the microballoons dried at 40 "C was changed in crystalline form (curves a and c in Figure 4). The endothermic peak due to the release of water appeared below 100 "C, indicating that tranilast in the microballoons that were dried a t 40 "C was the hydrate, which was converted to the anhydrate after dehydration a t -100 "C. The diffraction pattern and the thermal property of tranilast in the microballoons dried at 120 "C indicated that the original A
I
5
10
I
I
15 20 25 29 (degree)
I
30
35
B
4
Temperature ("C) Flgure &Average diameter and recovery of tranilast microballoonsas a function of temperature of the aqueous phase. Key: (0)average diameter of microballoons;(0)percent of -12 mesh (opening 1410 pm)
fraction of microballoons recovered.
100 200 T e m e r a t u r e ('C)
300
Figure +X-ray powder diffraction patterns (A) and DSC curves (6)of tranilast microballoons prepared by drying at (a) 40 "C and (b) 120 "C, and of (c) the physical mixture trani1ast:Eudragit S (1:2). Journal of Pharmaceutical Sciences I 137 Vol. 81, No 2, February 1992
A
r
R
5 213 (degree) Flgure +Scanning electron microphotographs of microballoons containing tranilast (A,B) and ibuprofen (C). Upper panels indicate the surface and lower panels are cross-sections.
B a
'! U
U
c
W
200
100
Temperature ('C) Flgure !&X-ray powder diffraction patterns (A) and DSC curves ( 8 )of microballoons containing ibuprofen and Eudragit S with ratios of (a) 1 :2, (b) 1:3 and (c) 1:4, and of (d) the physical mixture (1:4).
crystal (anhydrate) was present (curves b and c in Figure 4). The reduced diffraction intensity suggests that some drugs were dispersed uniformly in the molecular level like a solid dispersion in the polymer shell of the microballoons. Figure 5 reveals that ibuprofen existed completely in the amorphous state in the shell of microballoon, irrespective of the drug content, as indicated by characteristic peakless patterns of the X-ray diffraction and the DSC thermographs. It was also found that ibuprofen was solid dispersed in the acrylic polymer to a greater extent than tranilast by evaporating the organic solution of the drug and the polymer. Better compatibility of ibuprofen than tranilast with the acrylic polymer might explain the findings in the present system. The scanning electron microphotographs (SEM) of the surface of a microballoon and its cross-section are shown in
Figure 6, which indicates that the microballoon was a perfect sphere without pores on the surface, irrespective of the type of drug it contained. The characteristic internal structure of the microballoon, a spherical cavity enclosed with the rigid shell constructed with drug and polymer, was clearly different from the porous microsphere described by Benita et al.13 The microspheres prepared by the encapsulation method based on the solvent evaporation process (e.g., dich1oromethane:water system) generally had a porous sponge-like structure.15 On the surface of the microballoon containing tranilast, crystals of the drug appeared (Figures 6A and 6B), in contrast to the smooth surface of the microballoon containing ibuprofen (Figure 6C). This finding explains the fact that the crystallinity of the tranilast microballoon was higher than that of the ibuprofen microballoon, as shown in Figures 4 and 5, respectively. The particle densities measured by the present two methods were mostly smaller than unity (i.e., the density of gastric fluid; Table I). The data measured by the displacement method was more useful because the experimental conditions were similar to those inside the stomach. The intrusion of aqueous medium into the pores increased the density measured by the displacement method compared with that by the counting method. The microballoons (500-1000 pm) dried a t 120 "C floated continuously in the acid solution containing Tween 20 (0.02%)that was stirred with a magnetic stirrer for >12 h (Figure 7). The surface-active agent was added to the system to take into account the wetting effects of natural surface-active agents, such as phospholipids and bile salts, in the gastrointestinal tract. The excellent buoyancy shown in Figure 7 was due to the insolubilities of drug and polymer in the acid solution. The independency of the buoyancy from the acidity might be useful for floating the mi-
Table CPartlcle Density, Porosity, and Ratlo of Diameter to Shell Thlckness of Tranilast Mlcroballoons Measured by Countlng or Dlsplacement Methods
Size fractions,
Temperature, "C ~
~~~~
40 120
a
~~~~
rm
Densig, g/cm
250-297 297-500 500-1 000 250-297 297-500 500-1 OOO
0.876 0.664 0.642 0.799 0.712 0.634
Counting. Porosity, YO
DK"
Densig, gicm
Displacementb Porosity, YO
DK'
~
34.4 50.3 51.9 39.2 45.8 51.7
6.7 9.8 10.2 7.5 8.7 10.1
1.060 1.044 0.867 0.901
0.850 0.655
20.6 21 .a 35.1 31.4 35.3 50.1
4.9 5.0 6.8 6.2 6.8 9.7
Measured by photographic counting method. Measured by Tween 20 aqueous solution (0.02%) displacement method. D R = ratio of diameter
to shell thickness.
138 / Journal of Pharmaceutical Sciences Vol. 81, No 2, February 1992
_.......................
Dissolution Time ( h o u r ) Figure &Dissolution profiles of tranilast from microballoons prepared and 120 "C (0,D) in dissolution medium at pH by drying at 40 "C (0,O) 6.8 (0,0 )and pH 7.2 (0,D); trani1ast:Eudragit S = 1:2; size fractions: 500-1000 pm.
Flgure 7-Floating behavior of tranilast microballoons In the JPXl No. 1 solution containing Tween 20 (0.02%) at pH 1.2 after stirring for 12 h (120°C drying; 500-1OOO pn size fractions).
croballoons in patients with hyperacidity or anacidity. The microballoons with larger diameters were more floatable because their densities decreased with their increasing size. The ratio of diameter to shell thickness, calculated from the density data, agreed well with the SEM observations. The decreased density of the microballoon dried at 120"C,in contrast to that dried at 40 "C,resulted from dehydration and the hydrophobic property of anhydrated tranilast. The latter property might be a significant factor for the larger gap in the density data observed between 40 and 120 "Cby the displacement method compared with that seen with the counting method (Table I). The micromeritic properties of microballoonsare tabulated in Table 11. The flow properties, represented in terms of angle of repose of the microballoons,were much improved over those of the original crystals of tranilast. The smaller value of parameter "a" (in the equation of Kawakita and Liidde"9 for the microballoons indicates their higher packabilities. The larger values of parameters "b" (in the equation of Kawakita and Liiddele) and "k" (in Kuno's equationl') for the microballoons indicate that the rate of their packing process was much higher compared with that of the original crystals. The closest packing density in 10 mL of a measuring cylinder was equivalent for the microballoons and crystals. The improvements of micromeritic properties suggest that the microbal-
loons are easily handled and filled into a capsule. Therefore, capsules loaded with microballoons are recommended for a floating drug delivery system. Drug Release Behaviors of Microballoon-The dissolution profiles of tranilast and ibuprofen from microballoonsare shown in Figures 8 and 9, respectively. The dissolution rate of tranilast from the microballoons dried a t 120 "Cwas slower than that from microballoons dried at 40 "Cin all media; this was caused by the poor wettability of the microballoon dried a t 120 "C.The hydrated tranilast dried a t 40 "C had higher solubility compared with the anhydrate dried at 120 "C, as reported previously.11 The burst in release of tranilast from microballoons a t the initial stage resulted from the dissolution of drug crystals on the surface of microballoons,as shown in the SEMs (Figures 6A and 6B).With increasing polymer concentration, the release rate of ibuprofen from mimballoons decreased drastically (Figure 9). The drug release rate could be controlled by the polymer concentration. Eudragit S is an enteric polymer, being soluble above pH 7. Thus, the drug release rates from microballoons changed dramatically above and below pH 7. In acidic medium, the drug was never released due to the solubility limitations, and almost 90% of microballoons floated on the surface of test solution because the gas phase remained inside the balloons. The present systems were composed of the enteric, multiple-unit devices. Therefore, the subunits that lost buoyancy might successfully pass through the stomach and release the drug in significant
Table ICMlcromerltlc Propertles of Mlcroballoonr and Orlglnal Crystab of Tranllast Micromeritic Property
Microballoons, Microballoons, Original Dried at Dried at Crystals 40 "C 120 "C of Tranilast
29.7 Tranilast content, % Average diameter, pm 457 Angle of repose, 38.4 ad 0.200 b' 0.517 kb 0.0862 Closest packing density, g/cm3 0.617
-
31.3 423 39.9 0.226 0.138 0.0588
210 52.3 0.546 0.0259 0.00566
0.436
0.575
ParametersinKawakita'sequation:(n/c) = (llab) + @/a), C = (V, Vn)/V,, where n is the tap number, and V, and V , are the powder bed volumes at initial and nth tapped state, respectively. Parameter in Kuno's equation: pr - pn = (p, - p,)exp(-kn), where p,, pn, and po are the apparent densities at equilibrium, nt" tapped state, and initial state, respectively.
'0
1
2 3 4 5 6 7 Dissolution Time (hour)
8
Figure +Dissolution profiles of ibuprofen from microballoons prepared with various ratios of ibuprofen to Eudragit S in dissolution medium at pH 6.8 (opensymbols) and pH 7.2 (closed symbols). Key: (0.0) 1:2; (A,A) 1:3; (0,D) 1:4 (size fractions: 350-500 pm). Journal of Pharmaceutical Sciences / 139 Vol. 81, NO 2, February 7992
amounts in the upper gut, the absorption site. The multipleunit system proposed using microballoons may provide an ideal dosage form to prolong the residence time in stomach and to enhance the bioavailability of drug. In conclusion, microballoons of tranilast or ibuprofen with acrylic polymer, prepared by the emulsion-solvent diffusion method, are a novel floating dosage form that is adaptable to any intragastric condition. These microballoons can be used with a wide variety of drugs as a new drug delivery system; hydrophilic drugs can be used in this system, but a hydrophobic drug is generally preferable. Examples of such drugs, ibuprofen, ketoprofen, tranilast, 5-fluorouracil, tolbutamide, and indomethacin, etc., have been successfully encapsulated in the microballoons.1s
References and Notes 1. Goto, S.; Uchida, T.; Aoyama, T. J. Pharmucobio-Dyn. 1985,8, 270. 2. Suryakusuma, H.; Jun, H. W. J. Pharm. Pharmacol. 1984,36, 493 and 497. 3. McCinit James W.A ueous Pol meric Coatings for Pharmaceutical g o s u e Forms:barcel Detker: New York. 1989:D 81. 4. Davis, S.S.; fiardy, J. G.; Taylor, M. J.; Whalley, D. R.; Wilson, C. G. Int. J. Pharm. 1984.21.167 and 331. 5. Ch’ng, H. S.; Park, H.; Kelly, P.; Robinson, J. R. J.Pharm. Sci. 1985,74,399. 6. Sheth, P. R.;Tossounian, J. Drug Devel. Ind. Pharm. 1984,10, 313.
140 I Journal of Pharmaceutical Sciences Vol. 81, No 2, Februaty 1992
7. Kaniwa, N.; Aoyama, N.; Ogata, H.; Ejima, A. J.PharmacobioDyn. 1988,11,563and 571. 8. Bkhgaard, H.; Hansen, A. B.; Kofod, H. Optimization of Drug Delivery. Munksgaard Copenhagen, 1982;pp 67-79. 9. Kawashima, Y . ;Niwa, T.; Handa, T.; Takeuchi, H.; Iwamoto, T.; Ito, Y. J. Phurm. Sci. 1989,78,68. 10. Kawashima, Y .; Niwa, T.; Handa, T.; Takeuchi, H.; Iwamoto, T.; Ito, Y.Chem. Pharm. Bull. 1989,37,425. 11. Kawashima, Y.;Niwa, T.; Takeuchi, H.; Hino, T.; Ito, Y.; Furuyama, S.J. Pharm. Sci. 1991,80,472. 12. Heywood, H. Chem. Znd. 1937,56,149. 13. Benita, S.;Barkai, A.; Pathak, Y. V.J. Controlled Release 1990, 12,213. 14. Benita, S.;Benoit, J. P.; Puisieux, F.; Thies, C. J. Pharm. Sci. 1984, 73,1721. 15. Sato, T.; Kanke, M.; Schroeder, H. G.; DeLuca, P. P. Pharm. Res. 1988,5,21. 16. Kawakita, K.; Liidde, K. H. Powder Technol. 1970-71,4,61. 17. Kuno, H. In Powder (Theo and Application); Kubo, T.; Jinbo, G.; Saito, E.; Takahashi,%.; Hayakawa, S., Eds.; Maruzen: Tokyo, 1979. 18. Kawashima, Y.;Niwa, T.; Takeuchi, H.; Hino, T.; Itoh, Y., unpublished results.
Acknowledgments Part of the present research was supporkd b a Grant in Aid for General Scientific Research (No. A-01400005)gom the Ministry of Education, Science and Culture, Japan. The authors are grateful to Mr. Osamu Ikemizu and Mr. Shinji Ohmori for their technical assistance.
Orally administered drugs which are promptly absorbed in the alimentary canal and rapidly lost from the blood are generally film coated or microencapsulated to prolong the drug release period and drug action.13 However, most of these forms have several physiological limitations, such as the gastrointestinal (GI) transit time;4 incomplete drug release from devices or too short residence time of the pharmaceutical dosage forms in the upper position of small intestine (the absorption region of GI tract) leads to low bioavailability of the sustained-release dosage forms. Even if slow release of drug is attained, the drug released after passing the absorption site is not utilized, thus lowering the efficacy of the drug. To overcome this problem, several attempts have recently been made to extend the GI transit time of devices. One approach is bioadhesive systems that “stick dosage forms to the mucin-epithelial cell surface, providing longer transit time due to adhesion of the device to the gastric wall.5 Such adhesion may cause problems, such as irritation to the mucosa if an overdose of the drug occurs locally. Another suggestion is floating dosage forms. These have a specific density that is lower than that of gastric fluids; they remain buoyant in the stomach contents. Most of the floating systems are dominated by single-unit formulations [e.g., so-called hydrodynamically balanced systems (HBS)I.e A drawback of this system is the high variability of the GI transit time, due to its all-or-nothing emptying process.7.S Therefore, a multiple-unit floating system which can be distributed widely throughout the GI tract, providing a possibility of achieving a longer-lasting and more reliable release of drugs, has been sought. To achieve this goal, a novel method to prepare floating microspheres loaded with drug was developed as a modification of the emulsion-solvent diffusion method for preparation of spherical polymeric microsponges for a controlled drug delivery system previously devised by the present authors.9.10 The device prepared with the present technique is a polymeric 0022-3549/92’0200-0 7 35$02.50/0 0 1992. American Pharmaceutical Association
microsphere with a round cavity. This microsphere was termed “microballoon” due to its characteristic internal hollow structure and excellent floatability in vitro. Tranilast, an oral antiallergic agent, or ibuprofen, a n oral antiinflammatory agent, was loaded in the outer shell of the microballoon. In this study, the microballoon preparation process and the physicochemical properties of the microballoons, such as particle diameter, particle density, and crystalline form of drug, were investigated. Furthermore, the floating and the drug release behavior of the microballoons in vitro was clarified.
Experimental Section Preparation of Microballoons of Tranilast or Ibuprofen with Eudragit &The outline of the procedure for the preparation of microballoons was the same as the method described previously.9 Tranilast (Siratori Pharmaceutical Company, Ltd., Chiba, Japan; 0.5 g) or ibuprofen (Taito Koeki, Toyama, Japan; 0.25-0.5 g) and Eudragit S (Mhm Pharma GmbH, Germany; 1.0 g), a n enteric polymer soluble at pH >7.0, were dissolved in the ethano1:dichloromethane mixture (1:l v/v, 10 mL) a t room temperature. The drug solution was poured into 200 mL of water containing 0.75% (w/v) polyvinyl alcohol (PVA-120; Kuraray Company, Ltd., Tokyo, Japan) that was thermally controlled at 40 “C.Then, the solution was stirred with a propeller-type agitator at 300 rpm. The finely dispersed droplets of the polymer solution of drug were solidified in the aqueous phase via diffusion of the solvent. The dichloromethane that evaporated from the solidified droplet was removed by an aspirator, leaving the cavity of the microsphere filled with water. After agitating the system for 60 min, the microspheres were filtered, washed with water, and dried in a n oven (KCV-4D, Advantec, Tokyo, Japan) at 40 or 120°C for 2 h. During the drying procedure, a hollow cavity was formed inside the microsphere, resulting in the microballoon. Measurement of the Diffusion of Solvent from the Dispersed Droplets into Water-The diffusion behaviors of ethanol and dichloromethane from the dispersed droplets of drug solution into the aqueous phase were investigated during the preparation of microballoons containing tranilast. At appropriate intervals, 1mL of external aqueous phase was withdrawn by using a syringe connected with a filter. The ethanol and dichloromethane dissolved in the aqueous medium were detected by a gas chromatograph (model GC-l4A, Shimadzu, Japan) with a flame ionization detector. Model Experiment for the Formation of Microballoons-A large droplet of the ethano1:dichloromethane solution was poured through a pipette into water in a petri dish that was thermally controlled at 40°C, thereby dissolving the drug and polymer. The solidifying behavior of the droplet in water was investigated photographically. Identification of Crystalline Form of t h e Drug in Microballoons-The crystalline form of the drug dispersed in the crust of the microballoon was analyzed by X-ray powder difiactometry (RAD-lC, Rigaku, Tokyo, Japan) and differential scanning calorimetry (DSC; model CN808521, Rigaku, Tokyo, Japan) according to our previous report.11 Measurement of Micromeritic Properties of Microballoons The surface morphologies and the internal textures of the microballoons were observed by a scanning electron microscope (JSM-T330A, Nihon Denshi, Tokyo, Japan). The average diameter ofmicroballoons was represented by the geometric mean diameter obtained by a sieve Journal of Pharmaceutical Sciences / 135 Vol. 81, No 2, February 1992
method. The flov~ and packing properties were investigated by measuring the angle of repose and tapped density. represented by eq 1, of the microballoons was The porosity (4, determined by measuring the true (A) and particle densities (6) of massed microballoons:
-.25 I h
E
cn E
v
E =
(1 - pdpt) x 100
c
(1)
.-0
+J m L
* c
't 15
0 -*A*--_-_---*--------a-
_......-....................................................................
a,
(2)
V
c
0
where V and W are volume and weight of a microballoon, respectively. The true density was measured by using a helium-air pycnometer (model 1302, Micromeritics Instrument Company). The particle density was determined by measuring the mass volume and weight of the microballoons sampled. The mass volume was measured by both a photographic counting method and a water displacement method. Heywood's diameters ( ~ 0 1 2of a number (n) of photographs of microballoons were measured with an image analyzer (IBAS, KarlZeiss, Germany), and the mass volume was calculated by the following equation:
(3) The water displacement method measured the volume by using a pycnometer with an aqueous solution of Tween 20 (0.0296, W/V; polyoxyethylene sorbitan monolaurate; Kishida, Osaka, Japan) at 20 "C. Measurement of Drug Release Rate from Microballoone-The drug dissolution testa of microballoons were carried out by the paddle method specified in the U. S. Pharmacopeia XXI. Microballoons that had been fractionated and weighed to correspond to 50 mg of tranilast or 200 mg of ibuprofen were gently spread over the surface of 900 mL ofdissolution medium (disintegration test solution no. 2,with pH 6.8, and phosphate buffer, with pH 7.2, as specified in the Japanese Pharmacopeia XI and corresponding to the U. S. Pharmacopeia XXI), rotated at 100 rpm, and thermostatically controlled a t 37 "C. Perfect sink conditions prevailed during the drug diasolution testa. The sample was withdrawn at a suitable interval from the dissolution vessel and was assayed spectrophotometrically at 335 nm for tranilast and at 220 nm for ibuprofen (model 320, Hitachi Manufacturing, Tokyo, Japan).
Results and Discussion Mechanism of Formation of Microballoons by EmulsionSolvent Diffusion Method-The ethano1:dichloromethane solution with drug and polymer (Eudragit S), when poured into aqueous solution with stirring, was finely dispersed into discrete droplets, thereby forming an O N type emulsion. The concentrations of ethanol and dichloromethane contained in the external aqueous phase during the preparation of microballoons were monitored, as shown in Figure l.The dashed and dotted lines represent the equilibrium concentrations of ethanol (19.8 mg/mL) and dichloromethane (12.7 mg/mL) in the aqueous phase, respectively, when 10 mL of the twosolvent mixture ( l : l , v/v) was saturated with 200 mL of the aqueous solution of polyvinyl alcohol in a closed system (i.e., glass-stopped flask). Several seconds after the dispersion of the solvent mixture, all ethanol diffused out of the dispersed droplets into the aqueous medium, and then the equilibrium concentration of ethanol was retained during the preparation of the microballoons. In contrast, the dichloromethane did not diffuse thoroughly from the droplets into the aqueous phase but partly resided in the droplets. During the preparation, the concentration of dichloromethane in the aqueous medium decreased due to evaporation from the system. The rapid diffusion of ethanol (good solvent for the polymer) from the droplets into the aqueous phase might reduce the solubility of the polymer in the droplet because the polymer (Eudragit S) 130 / Journal of Pharmaceutical Sciences Vol. 87, No 2, February 1992
0
0
1 20 30 40 50 60 Agitation Time (rnin) Figure 1-Diffusion and evaporation profiles of ethanol (0)and dichloromethane (A)into and out of the external aqueous phase during the preparation of microballoons containing tranilast. Dashed and dotted lines represent the expected equilibrium concentrations of ethanol (19.8 mg/mL) and dichloromethane (12.7 mg/mL), respectively. 0
10
was insoluble in dichloromethane. The polymer precipitated instantly a t the surface of the droplet, forming a film-like shell enclosing the dichloromethane which was dissolving the drug. The mechanism of formation of the microballoon was photographically investigated with a model system in which the drug and polymer solution was dropped through a pipette into water in a petri dish that was thermally controlled at 40 "C, as shown in Figure 2. The poured droplet settled to the bottom of the dish and partly flattened due to gravity. The surface of the droplet became immediately opaque due to the formation of a gel-like film on the surface (Figure 2A). After a few minutes, the gel-like droplet was squeezed by a pincette, and a few bubbles of dichloromethane appeared, as shown in Figures 2B and 2C. This finding indicates that the gas phase of dichloromethane was generated in the droplet covered with gel-like film before the droplet was entirely solidified. The gas phaae should produce a cavity in the microsphere. The mechanism of formation of the microballoons is illustrated in Scheme I, based on the results in Figures 1 and 2. The mechanism proposed in the present study is clearly different from that of the emulsion-solvent evaporation method reported by Benita et al.,13 in which the emulsion droplet was solidified principally by the evaporation of solvent from the droplet. In the present system, the diffusion of ethanol (good solvent for Eudragit S)preceded the evaporation of dichloromethane (poor solvent for the polymer) from the emulsion droplet into the aqueous medium, drastically reducing the solubility of polymer in the droplet and forming a gel-like film on the surface. The mechanically strong A
B
C
2m
Figure 2-Microphotographs of the droplet of ethanokdichloromethane solution of Eudragit S and tranilast in water (40"C)in a petri dish, before (A) and after (B,C) the squeeze by a pincette.
0
om* F o r s a i l o n cf
enulsicr
9/W
emuisicr,
==+
3 Fzrma:;cn
she! I
Of
Generation o f 92s
s l o n of E t G H
Phase
Mlcroballoon
diffusioi o f CH2C12
Scheme I-Preparation procedure and mechanism of microballoon formation by the emulsion-solvent diffusion method.
solidified film (shell) produced at the surface of droplet with further depletion of ethanol prevented rupture and shrinkage of the microsphere during the evaporation of dichloromethane from the droplet; this was shown in the model experiment. The cavity produced by the gas phase was gradually filled with water due to the reduced pressure inside the droplet that was caused by evaporation of dichloromethane. The microballoon was prepared by removing water from the cavity of the microsphere with a drying process. The presence of totally hydrolyzed polyvinyl alcohol, employed as an emulsifying agent, significantly prevented aggregation of the droplets with solidified outer shells during the process. It was assumed that the polyvinyl alcohol, adsorbed to the interface between the droplets, and the aqueous medium might form a strong hydrated layer over the droplets, resulting in a stable dispersed system. Moreover, the rapid formation of film-like shell over the droplets, as shown in the model experiment, prevented the coalescence of the droplets. Therefore, the size distributions of microballoons might coincide with those of droplets formed at the pouring stage of drug and polymer solution. It was found that the main factors determining the sizes of microballoons were the concentrations of drug and polymer in the formulation, the pouring speed of drug and polymer solution, and the agitation speed of the system. In Figure 3, the average diameters and the recoveries of microballoons of tranilast are exhibited as a function of temperature of the aqueous dispersing medium. The recovery was represented by the weight percent of microballoons recovered having a diameter of c1410 pm, to 600 I
I100
I
exclude aggregates, versus the weights of drug and polymer loaded in the system. When the temperature was controlled in the range of 30 to 40 "C, the recovery of microballoonsreached 70-80%, indicating the optimized condition. Above 40 "C, the recovery dramatically decreased since the shell burst and the ruptured microspheres aggregated to a larger mass. At higher temperatures, the emulsion droplets were unstable due to the intense evaporation of dichloromethane. The recovery gradually decreased with decreasing temperature because of the increase in the number of aggregates due to the bridging action of polyvinyl alcohol, which coalesced the particles at the washing process.14 Therefore, the temperature of aqueous dispersing medium was maintained a t 40 "C. No free crystals of drug appeared in the aqueous medium, suggesting that the drug content in the microballoons coincided with that in the formulation. It was found that the drug content of the microballoons was >95% of the formulation value. Physicochemical Identification and Micromeritic P r o p erties of Microballoons-The X-ray powder diffraction patterns and DSC thermograms of microballoons, along with those of physical mixtures of raw crystals of drug and polymer, are shown in Figures 4 (tranilast)and 5 (ibuprofen). It was found that tranilast in the microballoons dried at 40 "C was changed in crystalline form (curves a and c in Figure 4). The endothermic peak due to the release of water appeared below 100 "C, indicating that tranilast in the microballoons that were dried a t 40 "C was the hydrate, which was converted to the anhydrate after dehydration a t -100 "C. The diffraction pattern and the thermal property of tranilast in the microballoons dried at 120 "C indicated that the original A
I
5
10
I
I
15 20 25 29 (degree)
I
30
35
B
4
Temperature ("C) Flgure &Average diameter and recovery of tranilast microballoonsas a function of temperature of the aqueous phase. Key: (0)average diameter of microballoons;(0)percent of -12 mesh (opening 1410 pm)
fraction of microballoons recovered.
100 200 T e m e r a t u r e ('C)
300
Figure +X-ray powder diffraction patterns (A) and DSC curves (6)of tranilast microballoons prepared by drying at (a) 40 "C and (b) 120 "C, and of (c) the physical mixture trani1ast:Eudragit S (1:2). Journal of Pharmaceutical Sciences I 137 Vol. 81, No 2, February 1992
A
r
R
5 213 (degree) Flgure +Scanning electron microphotographs of microballoons containing tranilast (A,B) and ibuprofen (C). Upper panels indicate the surface and lower panels are cross-sections.
B a
'! U
U
c
W
200
100
Temperature ('C) Flgure !&X-ray powder diffraction patterns (A) and DSC curves ( 8 )of microballoons containing ibuprofen and Eudragit S with ratios of (a) 1 :2, (b) 1:3 and (c) 1:4, and of (d) the physical mixture (1:4).
crystal (anhydrate) was present (curves b and c in Figure 4). The reduced diffraction intensity suggests that some drugs were dispersed uniformly in the molecular level like a solid dispersion in the polymer shell of the microballoons. Figure 5 reveals that ibuprofen existed completely in the amorphous state in the shell of microballoon, irrespective of the drug content, as indicated by characteristic peakless patterns of the X-ray diffraction and the DSC thermographs. It was also found that ibuprofen was solid dispersed in the acrylic polymer to a greater extent than tranilast by evaporating the organic solution of the drug and the polymer. Better compatibility of ibuprofen than tranilast with the acrylic polymer might explain the findings in the present system. The scanning electron microphotographs (SEM) of the surface of a microballoon and its cross-section are shown in
Figure 6, which indicates that the microballoon was a perfect sphere without pores on the surface, irrespective of the type of drug it contained. The characteristic internal structure of the microballoon, a spherical cavity enclosed with the rigid shell constructed with drug and polymer, was clearly different from the porous microsphere described by Benita et al.13 The microspheres prepared by the encapsulation method based on the solvent evaporation process (e.g., dich1oromethane:water system) generally had a porous sponge-like structure.15 On the surface of the microballoon containing tranilast, crystals of the drug appeared (Figures 6A and 6B), in contrast to the smooth surface of the microballoon containing ibuprofen (Figure 6C). This finding explains the fact that the crystallinity of the tranilast microballoon was higher than that of the ibuprofen microballoon, as shown in Figures 4 and 5, respectively. The particle densities measured by the present two methods were mostly smaller than unity (i.e., the density of gastric fluid; Table I). The data measured by the displacement method was more useful because the experimental conditions were similar to those inside the stomach. The intrusion of aqueous medium into the pores increased the density measured by the displacement method compared with that by the counting method. The microballoons (500-1000 pm) dried a t 120 "C floated continuously in the acid solution containing Tween 20 (0.02%)that was stirred with a magnetic stirrer for >12 h (Figure 7). The surface-active agent was added to the system to take into account the wetting effects of natural surface-active agents, such as phospholipids and bile salts, in the gastrointestinal tract. The excellent buoyancy shown in Figure 7 was due to the insolubilities of drug and polymer in the acid solution. The independency of the buoyancy from the acidity might be useful for floating the mi-
Table CPartlcle Density, Porosity, and Ratlo of Diameter to Shell Thlckness of Tranilast Mlcroballoons Measured by Countlng or Dlsplacement Methods
Size fractions,
Temperature, "C ~
~~~~
40 120
a
~~~~
rm
Densig, g/cm
250-297 297-500 500-1 000 250-297 297-500 500-1 OOO
0.876 0.664 0.642 0.799 0.712 0.634
Counting. Porosity, YO
DK"
Densig, gicm
Displacementb Porosity, YO
DK'
~
34.4 50.3 51.9 39.2 45.8 51.7
6.7 9.8 10.2 7.5 8.7 10.1
1.060 1.044 0.867 0.901
0.850 0.655
20.6 21 .a 35.1 31.4 35.3 50.1
4.9 5.0 6.8 6.2 6.8 9.7
Measured by photographic counting method. Measured by Tween 20 aqueous solution (0.02%) displacement method. D R = ratio of diameter
to shell thickness.
138 / Journal of Pharmaceutical Sciences Vol. 81, No 2, February 1992
_.......................
Dissolution Time ( h o u r ) Figure &Dissolution profiles of tranilast from microballoons prepared and 120 "C (0,D) in dissolution medium at pH by drying at 40 "C (0,O) 6.8 (0,0 )and pH 7.2 (0,D); trani1ast:Eudragit S = 1:2; size fractions: 500-1000 pm.
Flgure 7-Floating behavior of tranilast microballoons In the JPXl No. 1 solution containing Tween 20 (0.02%) at pH 1.2 after stirring for 12 h (120°C drying; 500-1OOO pn size fractions).
croballoons in patients with hyperacidity or anacidity. The microballoons with larger diameters were more floatable because their densities decreased with their increasing size. The ratio of diameter to shell thickness, calculated from the density data, agreed well with the SEM observations. The decreased density of the microballoon dried at 120"C,in contrast to that dried at 40 "C,resulted from dehydration and the hydrophobic property of anhydrated tranilast. The latter property might be a significant factor for the larger gap in the density data observed between 40 and 120 "Cby the displacement method compared with that seen with the counting method (Table I). The micromeritic properties of microballoonsare tabulated in Table 11. The flow properties, represented in terms of angle of repose of the microballoons,were much improved over those of the original crystals of tranilast. The smaller value of parameter "a" (in the equation of Kawakita and Liidde"9 for the microballoons indicates their higher packabilities. The larger values of parameters "b" (in the equation of Kawakita and Liiddele) and "k" (in Kuno's equationl') for the microballoons indicate that the rate of their packing process was much higher compared with that of the original crystals. The closest packing density in 10 mL of a measuring cylinder was equivalent for the microballoons and crystals. The improvements of micromeritic properties suggest that the microbal-
loons are easily handled and filled into a capsule. Therefore, capsules loaded with microballoons are recommended for a floating drug delivery system. Drug Release Behaviors of Microballoon-The dissolution profiles of tranilast and ibuprofen from microballoonsare shown in Figures 8 and 9, respectively. The dissolution rate of tranilast from the microballoons dried a t 120 "Cwas slower than that from microballoons dried at 40 "Cin all media; this was caused by the poor wettability of the microballoon dried a t 120 "C.The hydrated tranilast dried a t 40 "C had higher solubility compared with the anhydrate dried at 120 "C, as reported previously.11 The burst in release of tranilast from microballoons a t the initial stage resulted from the dissolution of drug crystals on the surface of microballoons,as shown in the SEMs (Figures 6A and 6B).With increasing polymer concentration, the release rate of ibuprofen from mimballoons decreased drastically (Figure 9). The drug release rate could be controlled by the polymer concentration. Eudragit S is an enteric polymer, being soluble above pH 7. Thus, the drug release rates from microballoons changed dramatically above and below pH 7. In acidic medium, the drug was never released due to the solubility limitations, and almost 90% of microballoons floated on the surface of test solution because the gas phase remained inside the balloons. The present systems were composed of the enteric, multiple-unit devices. Therefore, the subunits that lost buoyancy might successfully pass through the stomach and release the drug in significant
Table ICMlcromerltlc Propertles of Mlcroballoonr and Orlglnal Crystab of Tranllast Micromeritic Property
Microballoons, Microballoons, Original Dried at Dried at Crystals 40 "C 120 "C of Tranilast
29.7 Tranilast content, % Average diameter, pm 457 Angle of repose, 38.4 ad 0.200 b' 0.517 kb 0.0862 Closest packing density, g/cm3 0.617
-
31.3 423 39.9 0.226 0.138 0.0588
210 52.3 0.546 0.0259 0.00566
0.436
0.575
ParametersinKawakita'sequation:(n/c) = (llab) + @/a), C = (V, Vn)/V,, where n is the tap number, and V, and V , are the powder bed volumes at initial and nth tapped state, respectively. Parameter in Kuno's equation: pr - pn = (p, - p,)exp(-kn), where p,, pn, and po are the apparent densities at equilibrium, nt" tapped state, and initial state, respectively.
'0
1
2 3 4 5 6 7 Dissolution Time (hour)
8
Figure +Dissolution profiles of ibuprofen from microballoons prepared with various ratios of ibuprofen to Eudragit S in dissolution medium at pH 6.8 (opensymbols) and pH 7.2 (closed symbols). Key: (0.0) 1:2; (A,A) 1:3; (0,D) 1:4 (size fractions: 350-500 pm). Journal of Pharmaceutical Sciences / 139 Vol. 81, NO 2, February 7992
amounts in the upper gut, the absorption site. The multipleunit system proposed using microballoons may provide an ideal dosage form to prolong the residence time in stomach and to enhance the bioavailability of drug. In conclusion, microballoons of tranilast or ibuprofen with acrylic polymer, prepared by the emulsion-solvent diffusion method, are a novel floating dosage form that is adaptable to any intragastric condition. These microballoons can be used with a wide variety of drugs as a new drug delivery system; hydrophilic drugs can be used in this system, but a hydrophobic drug is generally preferable. Examples of such drugs, ibuprofen, ketoprofen, tranilast, 5-fluorouracil, tolbutamide, and indomethacin, etc., have been successfully encapsulated in the microballoons.1s
References and Notes 1. Goto, S.; Uchida, T.; Aoyama, T. J. Pharmucobio-Dyn. 1985,8, 270. 2. Suryakusuma, H.; Jun, H. W. J. Pharm. Pharmacol. 1984,36, 493 and 497. 3. McCinit James W.A ueous Pol meric Coatings for Pharmaceutical g o s u e Forms:barcel Detker: New York. 1989:D 81. 4. Davis, S.S.; fiardy, J. G.; Taylor, M. J.; Whalley, D. R.; Wilson, C. G. Int. J. Pharm. 1984.21.167 and 331. 5. Ch’ng, H. S.; Park, H.; Kelly, P.; Robinson, J. R. J.Pharm. Sci. 1985,74,399. 6. Sheth, P. R.;Tossounian, J. Drug Devel. Ind. Pharm. 1984,10, 313.
140 I Journal of Pharmaceutical Sciences Vol. 81, No 2, Februaty 1992
7. Kaniwa, N.; Aoyama, N.; Ogata, H.; Ejima, A. J.PharmacobioDyn. 1988,11,563and 571. 8. Bkhgaard, H.; Hansen, A. B.; Kofod, H. Optimization of Drug Delivery. Munksgaard Copenhagen, 1982;pp 67-79. 9. Kawashima, Y . ;Niwa, T.; Handa, T.; Takeuchi, H.; Iwamoto, T.; Ito, Y. J. Phurm. Sci. 1989,78,68. 10. Kawashima, Y .; Niwa, T.; Handa, T.; Takeuchi, H.; Iwamoto, T.; Ito, Y.Chem. Pharm. Bull. 1989,37,425. 11. Kawashima, Y.;Niwa, T.; Takeuchi, H.; Hino, T.; Ito, Y.; Furuyama, S.J. Pharm. Sci. 1991,80,472. 12. Heywood, H. Chem. Znd. 1937,56,149. 13. Benita, S.;Barkai, A.; Pathak, Y. V.J. Controlled Release 1990, 12,213. 14. Benita, S.;Benoit, J. P.; Puisieux, F.; Thies, C. J. Pharm. Sci. 1984, 73,1721. 15. Sato, T.; Kanke, M.; Schroeder, H. G.; DeLuca, P. P. Pharm. Res. 1988,5,21. 16. Kawakita, K.; Liidde, K. H. Powder Technol. 1970-71,4,61. 17. Kuno, H. In Powder (Theo and Application); Kubo, T.; Jinbo, G.; Saito, E.; Takahashi,%.; Hayakawa, S., Eds.; Maruzen: Tokyo, 1979. 18. Kawashima, Y.;Niwa, T.; Takeuchi, H.; Hino, T.; Itoh, Y., unpublished results.
Acknowledgments Part of the present research was supporkd b a Grant in Aid for General Scientific Research (No. A-01400005)gom the Ministry of Education, Science and Culture, Japan. The authors are grateful to Mr. Osamu Ikemizu and Mr. Shinji Ohmori for their technical assistance.
