Journal of Controlled Release, 16 ( 199 1) 355-364 355

0 1991 Elsevier Science Publishers B.V. 0168-3659/91/$03.50

ADONIS016836599100037H

COREL 006 13

Release of dextromethorphan hydrobromide from freeze-dried

enzyme-degradable hydrogels
Waleed S.W. Shalaby, Garnet E. Peck and Kinam Park
Purdue University,School of Pharmacy, West Lafayette, Indiana, U.S.A.
(ReceivedNovember 1, 1990; accepted in revised form January 9, 199 1)

Dextromethorphan hydrobromide was incorporated into albumin-crosslinked polyvinylpyrrolidone

hydrogels by equilibrating the gels in a saturated drug solution followed by freeze drying to entrap the
drug in the network. Through freeze drying, highly porous hydrogels containing uniformly dispersed
drug were produced. Drug content in the gel increased as a function of the number of drug loading-
freeze drying cycles. Solvent penetration into freeze-dried gels resulted in an initial isotropic collapse
of the network followed by a gradual increase in gel size due to swelling. In the presence of pepsin,
freeze-dried networks degraded at a much faster rate than non-freeze-dried control samples. Drug re-
lease from freeze-dried hydrogels was degradation-independent and inconsistent with conventional
solute transport mechanisms through swellable low surface area devices. The rate of drug release was
dependent on the amount of drug loaded into the freeze-dried matrix. The potential use of these de-
vices for long-term oral drug delivery is discussed.

Keywords: Freeze drying; Hydrogel; Albumin; Pepsin; Enzymatic degradation; Solute transport; Drug
delivery

Introduction drug migration to the surface of the gel can result

depending on the properties of the solvent and
Polymeric devices have been used extensively drug [22]. When solute migration is apparent,
in the area of controlled drug delivery [ l-61. A both drug content and drug uniformity within the
variety of drug-loaded devices have been fabri- matrix become limited. In the latter case, when
cated using techniques involving solvent casting the drug is loaded during the formation of the
17,819 injection molding [ 9 1, compression network, events such as autoacceleration can lead
molding [ IO- 12 1, and microencapsulation [ 13- to the degradation of thermally labile drugs. In
18 ]. With respect to hydrogels, drug loading has addition, purification of the hydrogel network
primarily been limited to the equilibration of gels could result in unnecessary and even costly loss
in a drug-containing solvent followed by drying of drugs. Consequently, for the practical appli-
[ 19-241 or by incorporating the drug during cation of hydrogels as long-term drug delivery
synthesis of hydrogels [25-281. In the former devices, drug content, drug uniformity and drug
case, upon removal of the solvent, undesirable stability should be carefully considered to ensure
therapeutic activity and to minimize undesir-
Correspondence to: Kinam Park, Purdue University, School able release properties. Recently, albumin-cross-
of Pharmacy, West Lafayette, Indiana 47907, U.S.A. linked hydrogels have been developed as poten-
356

tial platforms for oral drug delivery [ 29-311. dried. A total of 4 drug loading cycles were car-
Since hydrogel retention was observed in the ried out on both air-dried and freeze-dried gels
canine stomach for up to 60 h [ 321, the use of by re-equilibrating drug-loaded gels in a drug
these systems for long-term oral drug delivery is saturated solution in the manner described
promising. With this in mind, the development above.
of an efficient means for drug loading into en-
zyme-digestible hydrogels using the process of Dynamic swelling studies
freeze drying was initiated. As a first step to-
wards this goal, the effects of freeze drying on To assess the effects of freeze drying on the
hydrogel swelling, hydrogel degradation, multi- swelling behavior, freeze-dried and air-dried gels
ple drug loading and drug release properties was were allowed to swell from the dried state in pep-
studied. sin-free simulated gastric fluid at 37’ C [ 33 1. Ho-
mogeneous gel swelling was achieved by placing a
Materials and Methods steel wire mesh in the swelling containers to pre-
vent freeze-dried gels from floating on the surface
Hydrogel preparation of the medium. At timed intervals the weight, di-
ameter, and thickness of the gels were recorded.
Albumin-crosslinked polyvinylpyrrolidone The diameter and the thickness were measured
(PVP) hydrogels were prepared in distilled using a ruler. The swelling ratio (Q) was then de-
deionized water using 1-vinyl-2-pyrrolidinone termined from the following relationship:
(Aldrich, Milwaukee, WI, U.S.A.) at 36%
Q= W/W
(w/v) ,2,2-azobis (2-methylpropionitrile) (East-
man Kodak, Rochester, NY, U.S.A.) as initiator where W and W are the weights of the swollen
at 0.5% (w/w) of the monomer, and 90% alky- and dry gels, respectively.
lated albumin as a crosslinking agent at 6.5% To characterize the degradable properties of
(w/w) of the monomer [ 29,301. The monomer freeze-dried hydrogels, freeze-dried and air-dried
solution was degassed and purged with nitrogen gels were added to pepsin-containing simulated
three times followed by polymerization at 60°C gastric fluid at 37’ C. The concentration of pepsin
for 18 h under nitrogen. Once polymerization was (Sigma, 2500 units/mg) used for all degradation
complete the gels were removed and cut into discs studies was 250 units/ml. This concentration was
( 12 mm diameterx 5 mm thickness). The gels selected to avoid concentration-dependent
were then purified and dried as described previ- changes in the degradation behavior [ 341. It was
ously [ 301. The swelling and degradation of found previously that at lower pepsin concentra-
freeze-dried gels were compared with those of air- tions gel degradation was concentration-depen-
dried gels. The gels were frozkn in the swollen dent. During the uptake of penetrants, the weight
state at -70°C and then freeze dried using a of the swollen gels was recorded until pepsin
model USM- 15 freeze dryer (Virtus Co., Gardi- digestion rendered them unsuitable for continued
ner, NY, U.S.A. ). measurements.

Drug loading Studies on the PVP release from degrading gels

Dextromethorphan hydrobromide (Hoff- Freeze-dried and air-dried gels were added to

mann-LaRoche, Nutley, NJ, U.S.A.) was loaded pepsin-containing simulated gastric fluid in the
into hydrogels by equilibrating the gels in a drug- USP II dissolution apparatus (VanKel Indus-
saturated solution (37.6 mg/ml) at 37°C for 48 tries, Inc. ) at 37’ C. The paddle speed was main-
h. Following drug incorporation, some hydrogels tained at 50 rpm. At timed intervals, lo-ml ali-
were air dried for 4 days followed by oven drying quots were removed from the 1 liter vessels and
at 60 oC for 24 h while other hydrogels were freeze replaced with 10 ml of fresh-pepsin-containing
 357

simulated gastric fluid. The samples were then

analyzed spectrophotometrically against appro-
priate blanks at 211 nm for PVP using an absorp-
tivity of 3.185~ 10m2 cm2M-’ that was deter-
mined in pepsin-free simulated gastric fluid. The Q
pepsin concentration was maintained at 250
units/ml.

Drag release studies

04
0 20 40 60 60 100 120 140 160
The release of dextromethorphan hydrobrom- time (h)
ide from freeze-dried hydrogels and air-dried hy-
drogels was examined using a USP II dissolution Fig. 1. Swelling kinetics of freeze-dried (0 ) and air-dried
apparatus. The paddle speed was maintained at ( A ) hydrogels in simulated gastric fluid without pepsin at
37°C. The absence of an error bar indicates that the standard
50 rpm. Release studies were performed at 37°C deviation of the data was smaller than the size of the symbol
using both pepsin-free simulated gastric fluid and (n=4).
pepsin-containing simulated gastric fluid (250
units/ml). Over timed intervals, lo-ml aliquots
were removed from the 1 liter vessels and re-
placed with 10 ml of fresh solvent. Drug release
was determined spectrophotometrically against
appropriate blanks at 278 nm using an absorptiv-
ity of 5.833 cm2mg-’ that was determined in pep-
sin-free simulated gastric fluid. The release ki-
netics of dextromethorphan hydrobromide were
analyzed using the following equation:
10
MJM, = kt” (1) 0 12 3 4 5 6 7 6
time (h)
where M, is the amount of drug released at time
t, M, is the total amount of drug released, k is a Fig. 2. Dynamic changes in gel diameter of freeze-dried ( 0 )
release constant, and n is the release exponent and air-dried ( A ) hydrogels resulting from solvent absorp-
[ 23,35,36]. tion. The absence of an error bar indicates that the standard
deviation of the data was smaller than the size of the symbol
(n=4).
Results

Dynamic swelling gels, however, was unique in that a substantial

isotropic collapse of the network occurred ini-
Through the use of freeze drying techniques, tially (Fig. 2). Within the first 30 lpin of solvent
non-glassy, highly porous, polymer networks were penetration, the average diameter of the freeze-
formed. Hydrogels prepared by the air-drying dried gels was reduced from 20 mm to 15 mm while
method showed glassy network structures while the average thickness was reduced from 8.3 mm
freeze-dried gels were highly porous and compa- to 6.8 mm. Over the next 8 h, the gel diameter
rable in size to the pre-freeze dried swollen state. increased back to its original diameter. Solvent
Dynamic swelling studies showed that there was penetration was characterized by a rubbery gel
no apparent difference in the dynamic swelling front moving through an air-entrapped porous
behavior over time, although freeze-dried and polymeric core. The rubbery outer core was found
non-freeze-dried gels were structurally distinct to resemble the pre-freeze-dried swollen state of
(Fig. 1). Solvent penetration into freeze-dried the hydrogel. The air-entrapped inner core was
358

made up of a combination of dry porous polymer

at the center and hydrated polymer near the rub-
bery front. After 72 h of swelling, the air bubbles
could no longer be visualized inside the gel. Be-
cause of the slow release of air bubbles from the
hydrogel, the rate of solvent penetration was be- 0

lieved to be reduced since the penetrant must first

displace the air from inside the gel for further
swelling to occur. Thus, the collapse of the gel 5
network suggests that the characteristic relaxa- 0
tion time of the polymer in response to solvent 0 20 40 80 80 100 120 140 160 180 2 0
(h)
absorption is much shorter than the characteris- time
tic diffusion time of the penetrant into the gel [ 371 Fig. 4. Swelling kinetics of freeze-dried (0 ) and air-dried
while the rate of swelling appears to be dependent ( A ) hydrogels in the presence of pepsin. The absence of an
error bar indicates that the standard deviation of the data
on the displacement of air from inside the gel.
was smaller than the size of the symbol ( n = 4).

Enzymatic degradation presence of the pepsin was attributed to bulk deg-

radation. Bulk degradation was characterized by
The rate and extent of PVP releasefrom freeze- an increase in swelling ratios as compared to non-
dried hydrogels was found to be significantly degrading control samples and a reduction in bulk
higher than that of air-dried control samples (Fig. integrity over time leading to complete gel disrup-
3). Over a 7&y period, freeze-dried gels lost up tion. Freeze-dried gels were found to swell to a
to 40% (w/w) of the total PVP content while lesser extent and degrade at a faster rate com-
control samples lost only 18% (w/w) of the total pared to air-dried gels (Fig. 4 ) . The earlier onset
PVP content. With respect to the air-dried sam- of gel disruption and the reduction in Q was largely
ples, an initial 12 h lag period was observed prior attributedto the more rapid loss of polymer chains
to PVP release. The possible mechanism under- from freeze-dried gels than from the air-dried
lying these distinctly different degradation pro- samples.
files is described below.
As seen in previous studies [ 30,311, swelling of Drug loading
both freeze-dried and air-dried hydrogels in the
Drug loading by the freeze-drying method re-
sulted in dosage forms with greater drug uniform-
ity as compared to the air-drying method. Al-
though the content of drug increased as the
loading cycle increased (Fig. 5 ) , drug uniformity
was not maintained by air drying (Fig. 6 ) . Drug
loading by the air-drying method resulted in drug
migration to the gel surface upon solvent re-
moval. Over multiple loading cycles, the amount
, ( * of drug contained on the surface of air dried sam-
0 25 50 75 100 125 150 175
ples increased significantly.As a result,the weight
time (h) and diameter of air-dried gels were comparable to
that by the freeze-drying method after multiple
Fig. 3. Release of PVP as a function of time from freeze-dried
(0 ) and air-dried ( A ) hydrogels in the presence of pepsin. drug loading cycles (Figs. 5 and 6). Thus, multi-
The absence of an error bar indicates that the standard devia- ple drug loading by the air-drying method could
tion of the data was smaller than the size of the symbol ( n = 4). not uniformly disperse the drug within the glassy
 04
0 5 10 15 20 25 30 : 5
loading cycle time (h)

Fig. 5. The content of dextromethorphan hydrobromide in Fig. 7. Percent drug release as a function of time from freeze-
freeze-dried ( 0 ) and air-dried ( A ) hydrogels as a function dried ( q ) and air-dried (A ) hydrogels in simulated gastric
of loading cycle. The absence of an error bar indicates that fluid without pepsin at 37°C. The absence of an error bar
the standard deviation of the data was smaller than the size indicates that the standard deviation of the data was smaller
ofthesymbol (n=4). than the size of the symbol ( n = 3 ) .

10 20 30
Fig. 6. Freeze-dried (A) and air-dried (B) hydrogels after 4
time (h)
drug-loading cycles. Dextromethorphan hydrobromide was
found to be more uniformly dispersed in the freeze-dried hy-
Fig. 8. Drug release from freeze-dried hydrogels as a function
drogel compared to the air-dried gel where the drug resided
of time in simulated gastric fluid without pepsin at 37°C. 0,
mainly on the surface.
one loading cycle; A, two loading cycles; 0, three loading
cycles; 0, four loading cycles. The absence of an error bar
gel. In freeze-dried gels, however, a large fraction indicates that the standard deviation of the data was smaller
of the drug was localized within the network after than the size of the symbol ( n = 3 ) .
single and multiple loading cycles. The surface of
the freeze-dried gels was much more uniform after of the total loaded drug was released from freeze-
drug loading, and no visual evidence of drug mi- dried gels. The more prolonged release of drug
gration to the surface was seen. from the freeze-dried gel is not surprising since
the drug was more uniformly dispersed in the
Drug release in the absence of pepsin freeze-dried gels compared to the air-dried gels
where the localization of large amounts of drug at
The process of freeze drying had a profound in- the surface was prominent. One may argue that
fluence on the release of dextromethorphan hy- differences in diffusional path lengths arising from
drobromide from hydrogels (Fig. 7). After a the initial size of freeze-dried gels versus air-dried
swelling time of 90 min, an average of 63.7% (w/ gels could account for the more sustained release
w) of the total loaded drug was released from air- of drug from freeze-dried gels. However, Fig. 2
dried gels while only an average of 38.6% (w/w) shows that the freeze-dried and air-dried gels are
360

comparable in size after the first hour of swelling.

Thus, differences in diffusional path lengths may
have a minimal contribution to the drug release
profiles observed in this study. Furthermore, if the
drug was uniformly dispersed within the air-dried
glassy gel, a more sustained release of drug would
result due to the presence of a glassy/rubbery
transition front which would restrict the release
of drug to the rubbery phase of the gel. In con-
trast, freeze-dried gels would be expected to re- 0 10 20 30 D
lease the drug more readily since the highly po- time (h)
rous network can facilitate drug release instead
of restricting it. Fig. 7 indicated that the rate of Fig. 9. Percent drug release as a function of time from freeze-
drug release from air-dried gels was more rapid dried hydrogels in the absence of pepsin ( A ) and in the pres-
ence of pepsin ( 0 ). The absence of an error bar indicates
than from freeze-dried gels. Thus, the lack of drug that the standard deviation of the data was smaller than the
uniformity within the air-dried network is fur- size of the symbol (n = 3 ) .
ther supported.
The exponential term (n) in eqn. 1 was used to from degrading hydrogels was not significantly
determine the mechanism of drug release for the different from that of non-degrading hydrogels
first 60% of the total drug released. Air-dried hy- (Fig. 9). Degradation-independent drug release
drogels were found to release drug by an anoma- may have arisen due to two factors. First, the high
lous transport mechanism (n~0.67). Drug re- solubility of dextromethorphan hydrobromide led
lease from the freeze-dried gel was not consistent to a rapid rate of drug dissolution and release upon
with either Fickian (n= 0.5)) anomalous solvent penetration. Second, the slower relative
(0.5<n<l.O), or case II (n=l.O) transport rate of hydrogel degradation did not significantly
mechanisms, since r~0.44. Although this phe- alter the drug-releasing environment with re-
nomenon requires further investigation, it is con- spect to non-degrading hydrogels. Degradation-
ceivable that the collapse of the freeze-dried net- dependent drug release would be expected if gel
work upon solvent penetration may have degradation was more prevalent. Further studies
influenced the kinetics of drug release are needed to support this concept.
significantly.
The release profiles of dextromethorphan hy
drobromide from multiple loaded freeze-dried hy-
Discussion
drogels were found to be consistent with drug re-
lease from related freeze-dried systems [ 16,171. A large amount of a hydrophilic drug was uni-
As the content of drug increased inside the hy- formly incorporated into enzyme-degradable hy-
drogel, an increase in the rate of drug release was drogels by freeze drying. The content of the loaded
observed (Fig. 8). The exponential term (n) for drug increased linearly as a function of the num-
drug release was found to range from n = 0.44 for ber of loading cycles. In the swollen state, since
gels loaded once to n=0.32 for gels loaded four 95% of the hydrogel is water, a large portion of
times. the solvent-occupied volume could be utilized for
further incorporation of drugs. Consequently, the
Drug release in the presence of pepsin use of freeze drying enables one to minimize the
amount of polymeric materials required for dos-
The release of dextromethorphan hydrobrom- age form preparation while maximizing the total
ide from both air-dried and freeze-dried hydro- drug content in the delivery system. In short, the
gels in the presence of pepsin was determined to three step process consisting of gel preparation,
be degradation-independent. The release of drug drug loading, and freeze drying may be a useful
 361

technique for uniformly loading hydrophilic drugs for PVP release from air-dried gels. In the case of
into hydrogels. the freeze-dried network, the absence of a glassy/
The swelling kinetics of both freeze-dried and rubbery transition front combined with the po-
air-dried hydrogels were comparable even though rous structure of the freeze-dried network served
structural changes arising from solvent penetra- to enhance the release of PVP. During solvent up-
tion were distinctly different. Solvent penetra- take, gel swelling was characterized by a rubbery
tion into freeze-dried hydrogels resulted in a 25% gel front moving through an air entrapped porous
reduction in gel size over the first 30 min. The polymeric core. Because a glassy/rubbery transi-
time required for the freeze-dried network to re- tion front does not exist, no significant barrier is
gain its original dry state dimensions was approx- present to constrain the oligopeptide segments
imately 8 h. It should be noted, however, that this during swelling. As a result, the average degree of
8-h time constraint is likely to vary depending on conformational constraint is largely reduced at an
the initial size and porosity of the network. earlier stage in swelling and thus the rate of gel
The release of PVP from freeze-dried hydrogels degradation is enhanced. Compared to air-dried
in the presence of pepsin was found to be more glassy polymers, the release of PVP from freeze-
prominent than that from air-dried gels. A 12-h dried networks was more rapid as seen in Fig. 3.
lag phase in PVP release was observed for air- Drug release from both air-dried and freeze-
dried gels while PVP release from freeze-dried gels dried hydrogels was found to be degradation-in-
was less than 1 h. Network digestibility may be dependent. Drug release from freeze-dried gels was
explained by considering the extent of confor- determined by a transport mechanism that may
mational constraints on oligopeptide segments be controlled by the collapse of the network and
that exist between polymer chains before and the presence of a porous/rubbery swelling front
during solvent uptake. When the extent or degree during solvent absorption. As the drug load in-
of conformational constraints is high, the rate of creased, however, only the rate of drug release was
enzymatic degradation is expected to be low due found to increase. The use of a hydrophilic drug
to the limited access to cleavable sites on the pro- in our study largely contributed to the limited
sustained release properties from the device. With
tein segments. When the degree of constraints
respect to long-term oral drug delivery, however,
become low as a result of swelling, enzymatic deg-
diffusion-controlled release may have a minimal
radation is expected to be enhanced since the pro-
effect on drug release from the stomach due to the
tein segments are more accessible to cleavage.
presence of gel surface erosion under fasted and
In the case of air-dried gels, the presence of a
fed conditions [ 321. Thus, it may be possible to
“glassy” polymer significantly reduces the prob-
utilize both erosion-controlled and diffusion-con-
ability of enzymatic degradation by limiting the
trolled mechanisms to provide sustained drug re-
conformational freedom of the oligopeptide seg-
lease even when the drug is hydrophilic.
ments. During solvent uptake, the degree of con-
In summary, the use of freeze drying represents
straint is reduced due to swelling. Consequently, a new means of homogeneously incorporating and
segment constraint is highest in the “glassy” dispersing hydrophilic drugs in enzyme-degrada-
phase and lowest in the rubbery phase. Since the ble hydrogels. The application of this process in
degree of conformational constraint is largely in- the development of long-term oral drug delivery
fluenced by the presence of a glassy core, gel deg- systems is promising.
radation is expected to be enhanced once the
glassy core has disappeared due to swelling. At
that point, large constraints on oligopeptide seg- Acknowledgements
ments are liberated and the hydrogel becomes
more susceptible to enzymatic degradation. Gel The authors wish to acknowledge the technical
degradation by the above mechanism is consis- assistance of Malisa Chen. This study was sup-
tent with the 12-h lag period observed in this study ported by the ICI Pharmaceuticals Group.
362

References 17 T. Sato, M. Kanke, H.G. Schroeder and P. Deluca, Po-

rous biodegradable microspheres for controlled deliv-
1 ery. I. Assessment of processing conditions and solvent
N.A. Peppas, Hydrogels in Medicine and Pharmacy, Vol.
removal techniques, Pharm. Res. 5, (1988) 21-30.
III, CRC Press, Boca Raton, Florida, 1987.
2 R.S. Langer and D.L. Wise, Medical Applications of 18 S. Benita, A. Barkai and Y.V. Pathak, Effect of drug
Controlled Release, Vol. I, CRC Press, Boca Raton, loading extent on in vitro release kinetic behavior of ni-
Florida, 1984. fedipine from polyacrylate microsphres, J. Controlled
3 R.S. Langer and D.L. Wise, Medical Applications of Release 12, (1990) 213-222.
Controlled Release, Vol. II, CRC Press, Boca Raton, 19 R.W. Korsmeyer, E.V. Meerwall and N.A. Peppas, Sol-
Florida, 1984. ute and penetrant diffusion in swellable polymers. II.
4 J.R. Robinson and V.H.L. Lee, Controlled Drug Deliv- verification of theoretical models, J. Polym. Sci. 24,
ery: Fundamentals and Applications, 2nd edn., Marcel (1986) 409-434.
Dekker Inc., New York, 1987. 20 C.C.R. Robert, P.A. Buri and N.A. Peppas, Influence of
5 D.A. Tirrel, L.G. Donaruma and A.B. Turek, Macrom- the drug solubility and dissolution medium on the re-
olecules as Drugs and Carriers for Biologically Active lease from poly ( 2-hydroxyethyl methacrylate ) micro-
Materials, Ann. NY Acad. Sci. 446, New York, 1985. spheres. J. Controlled Release 5, (1987) 151-157.
6 C.G. Pitt, The controlled parenteral delivery of poly- 21 N.B. Graham, M. Zultigar, B.B. MacDonald and M.E.
peptides and proteins, Int. J. Pharm., 59 (1990) 173- McNeil, Caffeine release from fully swollen
196. poly (ethylene oxide) hydrogels, J. Controlled Release
7 R. Langer and J. Folkman, Polymers for sustained re- 5, (1988) 243-252.
lease of proteins and other macromolecules, Nature 263, 22 R.A. Siegel, M. Falamarzian, B.A. Firestone and B.C.
(1976) 797-799. Moxley, pH-controlled dependent release from hydro-
8 C.G. Pitt, M.M. Gratzl, A.R. Jeffcoat, R. Zweidinger and phobic/polyelectrolyte copolymer hydrogels. J. Con-
A. Schindler, Sustained drug delivery systems II: factors trolled Release 8, (1988) 179-182.
affecting release rates from poly(e-caprolactone) and 23 L. Brannon-Peppas and N.A. Peppas, Solute and pene-
related biodegradable polyesters, J. Pharm. Sci., 68, trant diffusion in swellable polymers. IX. the mecha-
(1979) 1534-1538. nism of drug release from pH-sensitive swelling-con-
9 K.W. Leong, J. Kost, E. Mathiowitz and R. Langer, Po- trolled systems, J. Controlled Release 8, (1989) 267-
lyanhydrides for controlled release of bioactive agents, 274.
Biomaterials 7, (1986) 364-371. 24 S.H. Gehrke and PI. Lee, Hydrogels for drug delivery
10 R.L. Linhardt, H.B. Rosen and R. Laner, Bioerodible systems, in: P. Tyle, (Ed.), Specialized Drug Delivery
polyanhydrides for controlled drug delivery, Polym. Systems: Manufacturing and Production Technology,
Prepr. 24, (1983) 47-48. Marcel Dekker Inc., New York, 1990, pp. 333-392.
11 Y. Ikada, S.H. Hyon, K. Jamshidi, S. Higashi, T. Ya- 25 P. Bernfield and J. Wan, Antigens and enzymes made
mamuro, Y. Katutani and T. Kitsugi, Release of anti- insoluble by entrapping them into lattices of synthetic
biotic from composites of hydroxapatite and poly (lactic polymers, Science 142, (1963) 678-679.
acid), J. Controlled Release 2, (1985) 179-186.
26 B.K. Davis, Control of diabetes with polyacrylamide
12 R. Latter, D. Lund, K. Leong and J. Folkman, Con-
implants containing insulin, Experientia 28, ( 1972) 348.
trolled release of macromolecules: biological studies, J.
27 B.K. Davis, Diffusion in polymer gel implants, Proc. Nat.
Controlled Release 2, (1985) 331-341.
Acad. Sci. U.S.A. 71, (1974) 3120-3123.
13 L.M. Sanders, J.S. Kent, G.I. McRae, B.H. Vickery, T.R.
28 J. Heller, R.F. Helwing, R.W. Baker and M.E. Tuttle,
Tice and D.H. Lewis, Controlled release of luteinizing
Controlled release of water-soluble macromolecules from
hormone-releasing hormone anologue from poly (d, l-
bioerodible hydrogels, Biomatrials 4, (1983) 262-266.
lactide-co-glycolide) microspheres. J. Pharm. Sci. 73,
29 K. Park, Enzyme-digestible swelling hydrogels as plat-
(1984) 1294-1297.
14 G. Spenlehauer, M. Veillar and J.P. Benoit, Formation forms for long-term oral drug delivery: synthesis and
and characterization of cisplatin loaded poly(d,l-lac- characterization, Biomaterials 9, ( 1988) 435-44 1.
tide) microspheres for chemoembolization, J. Pharm. 30 W.S.W. Shalaby and K. Park, Biochemical and mechan-
Sci. 75 (1986) 750-755. ical characterization of enzyme-digestible hydrogels.
15 M. Cavalier, J.P. Benoit and C. Thies, The formation Pharm. Res. 7, (1990) 816-823.
and characterization of hydrocortisone-loaded poly (d, l- 31 W.S.W. Shalaby, W.E. Blevins and K. Park, Enzyme-di-
lactide) microspheres, J. Pharm. Pharacol. 38, ( 1986) gestible properties associated with albumin-crosslinked
249-253. hydrogels for long-term oral drug delivery, proc. Symp.
16 M.P. Redmon, A.J. Hickey and P. Deluca, Predniso- Controlled Release Bioact. Mater. 17, ( 1990) 134-l 35.
lone-2 l-acetate poly (glycolic acid) microsphres: influ- 32 W.S.W. Shalaby, W.E. Blevins and K. Park, Gastric re-
ence of matrix characteristics on release, J. Controlled tention of enzyme-digestible hydrogels in the canine
Release 9, (1989) 99-109. stomach under fasted and fed conditions: a preliminary
 363

analysis using new analytical techniques, ACS Sympo- 35 P.L. Ritger and N.A. Peppas, A simple equation for de-
sium Series, in press. scription of solute release. II. Fickian and anomalous re-
33 United States Pharmacopeia/National Formulary, USP lease from swellable devices. J. Controlled Release 5,
XXI/NF XVI, U.S.P. Convention Inc., 1985, p. 1424. (1987) 37-42.
34 C.K. Shim and K. Park, Examination of drug release 36 N.A. Peppas and J. J. Sahlin, Simple equation for the
from enzyme-digestible swelling hydrogels, Proc. Int. description of solute release. part 3. coupling of diffu-
Symp. Controlled Release Bioact. Mater. 16, (1989) sion and relaxation, Int. J. Pharm. 57, ( 1989) 169- 172.
219-220. 37 G.W.R. Davidson and N.A. Peppas, Solute and pene-
trant diffusion is swellable polymers, J. Controlled Re-
lease 3, (1986) 243-258.