Biomaterials 24 (2003) 1459–1468

Interpolymer complexes of poly(acrylic acid) and chitosan: influence

of the ionic hydrogel-forming medium
Paloma M. de la Torre, Susana Torrado, Santiago Torrado*
Department of Pharmaceutical Technology, School of Pharmacy, Complutense University, Avda. Computense S/N, Madrid 28040, Spain
Received 20 July 2002; accepted 27 October 2002

Abstract

Non-covalent polyionic complexes were developed for localized antibiotic delivery in the stomach. Freeze-dried interpolymer
complexes based on polyacrylic acid (PAA) and chitosan (CS) were prepared in a wide range of copolymer compositions by
dissolving both polymers in acidic conditions. The influence of hydrogel-forming medium on the swelling and drug release was
evaluated.
The properties of these complexes were investigated by using scanning electron microscopy, dynamic swelling/eroding and release
experiments in enzyme-free simulated gastric fluid (SGF). The electrostatic polymer/polymer interactions generate polyionic
complexes with different porous structures. In a low pH environment, the separation of both polymer chains augmented as the
amount of cationic and carboxilic groups increased within the network. However, the presence of higher amount of ions in the
hydrogel-forming medium produced a network collapse, decreasing the maximum swelling ratio in SGF. PAA:CS:A (1:2.5:2)—
1.75 m complexes released around 54% and 71% of the amoxicillin in 1 and 2 h, respectively, in acidic conditions. A faster drug
release from this interpolymer complex was observed when the ionic strength of the hydrogel-forming medium increased. Complexes
with a high amount of both polymer chains within the network, PAA:CS:A(2.5:5:2), showed a suitable amoxicillin release without
being affected by an increased amount of ions in the hydrogel-forming medium. These freeze-dried interpolymer complexes could
serve as potential candidates for amoxicillin delivery in an acidic enviroment.
r 2002 Elsevier Science Ltd. All rights reserved.

Keywords: Chitosan; Poly(acrylic acid); Polyionic complexes; Interpolymer complexes; Ionic strength; Amoxicillin

1. Introduction drug interactions generate hydrogels without the need

for covalent cross-linking [5–8].
Triple therapy, including two antibiotics and a proton The controlled drug release from these polyionic
pump inhibitor, is a rational approach to the treatment hydrogels could respond to changes in environmental
of Helicobacter pylori. Recently, it has been reported parameters such as pH or temperature. Thus, inter-
that these proton pump inhibitors increase the local polymer complexes as poly (acrylic acid/polyvinil
concentration of antibiotics in the stomach, which may alcohol) or poly (methacrylic-g-ethylene glycol) exhibit
contribute to the clinically proven synergic beneficial temperature- and pH-sensitive, reversible complexation
action in eradication therapy of H. pylori [1,2]. [9,10]. In the case of some polyelectrolyte hydrogels, the
Swelling polyionic hydrogels have been employed for equilibrium swelling and mesh sizes were mainly
localized delivery of antibiotics in the stomach [3,4]. governed by the percentage of the ionic component in
Only chitosan (CS) or in combination with several the polymer and the degree of ionization of the drug
anionic polymers has been used in many studies of [10–12]. For example, Peniche et al. [13] have verified the
polyionic hydrogels because of its suitable sustained complexation ability of salicylic acid acrylic derivatives
drug release properties and its good biocompatibility with polyvalent cationic ligands as CS. Also, in poly
[4,5]. The electrostatic polymer/polymer and polymer/ (methacrylic acid-g-ethylene glycol) copolymers, Bell
et al. [10] have reported that the presence of specific ions
*Corresponding author. Tel.: +34-1-3941727; fax: 34-1-3941736. modified the solute diffusitivity. Finally, Gabrielli et al.
E-mail address: torrado1@[Link] (S. Torrado). [14], studied the swelling properties of xylan/CS hydro-

0142-9612/02/$ - see front matter r 2002 Elsevier Science Ltd. All rights reserved.
PII: S 0 1 4 2 - 9 6 1 2 ( 0 2 ) 0 0 5 4 1 - 0
1460 P.M. de la Torre et al. / Biomaterials 24 (2003) 1459–1468

gels at various pH and salt concentrations (0.075 and The different hydrogel formulations after being freeze-
0.15 m). The effect of NaCl addition could be explained dried presented similar volumes. Samples were cut into
by the fact that increasing the total amount of ions 2
2 mm2 squares.
decreases the ratio of ions between the inside of the gel In the second part of this study, the effect of ionic
and the surrounding solution. Thus, in pH extreme strength of the hydrogel-forming medium employing
conditions (pH=3 and 12) the degree of swelling three different molarities of acetic acid: 1.75, 3.50, and
decreases when the ionic strength of the medium 5.25 m was evaluated. Hydrogels PAA:CS:A (1:2.5:2)
increases. and PAA:CS:A (2.5:5:2) were chosen to carry out those
Nevertheless, there are few studies about the influence experiments. Therefore, freeze-dried hydrogels obtained
of solutions, with different ionic strengths, employed in with different concentrations of acetic acid were:
the preparation of freeze-dried hydrogels. PAA:CS:A (1:2.5:2)—1.75 m, PAA:CS:A (2.5:5:2)—
The aim of this paper is to evaluate the different 1.75 m, PAA:CS:A (1:2.5:2)—3.50 m, PAA:CS:A
electrostatic interactions in polyionic hydrogels pro- (2.5:5:2)—3.50 m, PAA:CS:A (1:2.5:2)—5.25 m and
duced between the cationic chains from CS and the PAA:CS:A (2.5:5:2)—5.25 m.
carboxylic groups from poly(acrylic acid) chains (PAA). Polyionic complexes obtained with a 1.75 m acetic
The main emphasis was placed on the investigation of acid solution were prepared as described before. And the
(I) how the changes produced in the ionic groups of both hydrogels developed with 3.50 and 5.25 m acetic acid
polymeric chains affect drug release (II) the influence of solutions were prepared in the same manner, the
different ionic strengths, employed during the produc- difference being the concentration of acetic acid.
tion of these hydrogels.
2.3. Scanning electron microscopy (SEM) studies

2. Materials and methods Freeze-dried samples were mounted on an aluminum

sample mount. After coating with gold–palladium, the
2.1. Materials hydrogel samples were analyzed with a Jeols 6400
SEM. All micrographs were the product of secondary
CS (low molecular weight) was purchased from Fluka electron imaging used for surface morphology identifi-
Biochemika (MW 150000). Carbopols 974P NF (PAA) cation at different magnifications and an accelerating
was obtained from BF Goodrich (OH, USA). Amox- voltage of 20 kV.
icillin trihydrate BP/USP was obtained from Antibiotics
(Madrid, Spain). Glacial acetic acid PRS and hydro- 2.4. Swelling experiments
chloric acid PA 37% were purchased from Panreac
Qu!ımica S.A. (Barcelona, Spain). All chemical reagents Freeze-dried hydrogels were swollen in 450 ml 0.1 m
were of pharmaceutical grade or better. HCl of enzyme-free simulated gastric fluid (SGF, pH
1.2) at a temperature of 371C prepared by the procedure
2.2. Hydrogel preparation described in USP 24. The swelling studies were carried
out using testing apparatus 1 (Vankels dissolution
Hydrogels were prepared with CS and Carbopols apparatus). Empty dry baskets were weighted. The
974P NF (PAA). A constant proportion of 2.0 of basket’s rotation speed was set at 100 rpm. At 15 min
amoxicillin trihydrato (A) was added to all formula- time intervals, the samples were removed from the
tions. The ratios of freeze-dried PAA:CS:A—1.75 m swelling medium and blotted on a piece of filter paper
hydrogels with the least amount of CS were: 0:2.5:2, prior to weighing to remove excess surface moisture.
1:2.5:2, 2.5:2.5:2 and 5:2.5:2. The freeze-dried PAA:C- The swelling ratio (SW) was determined according to
S:A—1.75 m hydrogel with the highest amount of CS the following expression:
presented the ratios: 0:5:2, 1:5:2, 2.5:5:2 and 5:5:2.
SWðwtÞ ¼Ws =Wd ;
The procedure was as follows: PAA was dissolved in
25.0 ml of acetic acid 1.75 m. Next, 250.0 mg of where Ws is the weight of the swollen hydrogel and Wd
amoxicillin trihydrate was added to the gel gradually is the weight of the dried hydrogel. The data represent
while mixing, until a uniform dispersion of drug was mean7S.D. from three samples per formulation. The
formed. CS was added gradually and mixed to get a studies were carried out for 5 h.
homogeneous gel. The hydrogel was neutralized with
sodium hydroxide 3 m to reach a pH of 5.0 and the 2.5. Release experiments
volume was adjusted to 40.0 ml for each formulation.
The interpolymer complexes were cast onto petri-dishes The release studies of all formulations were performed
of 100.0 ml with 100.0 mm i.d. Finally, gels were freeze- using testing apparatus 1 of the USP 24 (Vankels
dried using a Liolabar 7 (Telstat Inc Madrid, Spain). dissolution apparatus), 100 rpm. The dissolution
 P.M. de la Torre et al. / Biomaterials 24 (2003) 1459–1468 1461

medium was 450 ml SGF (0.1 m HCl) at 371C. At these hydrogels to present a non-porous surface with
predetermined time points, 5 ml of SGF was removed flakes. However, when PAA is incorporated into the
and filtered using a 0.45 mm filter and assayed for formulation, the hydrogels present a highly porous
amoxicillin at 228.5 nm using a UV–VIS Beckman structure. Fig. 1B shows the polyionic complex PAA:C-
DUs-7 Spectrophotometer. The cumulative amount of S:A (5:2.5:2)—1.75 m (50
) where an irregular structure
amoxicillin trihydrate released from the hydrogel was is seen on its surface with a pore size of 60–95 mm. The
determined from the appropriate calibration curves. The formation of an open cell structure on its surface could
data represent mean7S.D. from three independent be attributed to the electrostatic interaction between the
experiments. anionic groups from PAA and the cationic groups from
To investigate more precisely the effect of interpoly- CS. This characteristic structure was confirmed by
mer complex formation on the release of amoxicillin, the examining the cross-section of the interpolymer com-
results were analyzed according to the following plex. Various authors observed a similar morphology in
equation [15]: freeze-dried CS-based hydrogels [3,5]. The electrostatic
Mt =M ¼ ktn ; interaction between both polymeric chains is shown
clearly in the interpolymer complex PAA:CS:A
where Mt =M is the amount of amoxicillin (%) released (2.5:5:2)—1.75 m (Fig. 1C). The pore size is generally
at time tðhÞ; n is a diffusional exponent and k is the regular with values of 25–50 mm. However, an ionic
apparent release rate (%/h). increase in the hydrogel-forming medium, as in the
interpolymer complex, PAA:CS:A (2.5:5:2)—5.25 m
(Fig. 1D), produces a change in its morphology.
3. Results and discussion The surface presents less quantity of pores with
values of 10–30 mm in diameter. A new electrostatic
3.1. Study of the influence of increasing PAA content on interaction between both polymeric chains is produced
the structure and behavior of CS-based hydrogels by ions of the hydrogel-forming medium. Similar
hydrogel structure was obtained in xylan/CS
Fig. 1 shows different scanning electron micrographs complexes [14].
of freeze-dried hydrogels with several ratios of PAA and In Fig. 2, swelling and drug release from the inter-
CS. The top-view of a freeze-dried hydrogel without polymer complexes of various ratios of PAA and CS
PAA (PAA:CS:A 0:5:2)—1.75 m at a high magnification were studied to investigate the effect of polyionic
(300
) is depicted in Fig. 1A. It is characteristic for complex formation.

Fig. 1. Scanning electron micrographs of freeze-dried CS-based hydrogels without or with PAA. The micrographs depict top view of the following
interpolymer complexes: PAA:CS:A (0:5:2)—1.75 m at a high magnification (300
) (A), PAA:CS:A (5:2.5:2))—1.75 m (50
) (B), PAA:CS:A
(2.5:5:2)—1.75 m (C), and PAA:CS:A (2.5:5:2)—5.25 m produced in different hydrogel-forming media (D).
1462 P.M. de la Torre et al. / Biomaterials 24 (2003) 1459–1468

Fig. 2. Swelling/eroding (A) and release kinetics (B) in SGF of freeze-dried polyionic complexes with a CS proportion of 2.5: PAA:CS:A (0:2.5:2)—
1.75 m (-E-), PAA:CS:A (1:2.5:2)—1.75 m (-’-), PAA:CS:A (2.5:2.5:2)—1.75 m (-m-), and PAA:CS:A (5:2.5:2)—1.75 m (-K-).

Fig. 2A depicts the swelling kinetics of these formula- where it was easy to get a fast breaking of the
tions. Freeze-dried hydrogels without PAA (PAA:CS:A interpolymer network, due to the greater separation
0:2.5:2)—1.75 m exhibited the lowest maximum Sw ratio among the polymeric chains.
at 5.78 in the first 15 min, the other formulations with Fig. 2B shows the release rates of these formulations.
PAA achieved in the same time similar Sw ratios (from The formulation with the lowest ratio of PAA:CS:A
9.85 to 10.04). Furthermore, when the PAA proportion (1:2.5:2)—1.75 m presented a suitable controlled release
was lower in the polyionic complexes, the time to reach profile, about 53.55% and 70.98% of amoxicillin was
the maximum Sw ratio changed, thus the formulation released after 1 and 2 h, respectively. These release
PAA:CS:A (1:2.5:2)—1.75 m presented the maximum Sw results will ensure maximum availability of the drug in
at 45 min. The regular porous nature of the freeze-dried the stomach [3,4]. The reason for this is based on a
polyionic complex and a stronger electrostatic interac- number of factors, which include: the rate of erosion, the
tion between both polymeric chains would facilitate a existence of a different electrostatic interaction within
better matrix–solvent interaction, allowing an efficient the network that controlled the drug release and the
and rapid swelling (see Scheme 1A and B). presence of different kinds of structures with pores of
Hydrogels, which are not covalently cross-linked, are different sizes [13,16,17]. The greater the PAA content in
not permanent and can dissolve when polymer–polymer the interpolymer complexes, the faster the release rate of
interactions dissociate. The addition of PAA to the amoxicillin was achieved; therefore the polyionic com-
hydrogel slowed down the rate of erosion (see Fig. 2A). plex with the highest PAA:CS:A ratio (5:2.5:2)—1.75 m
But in polyionic complexes with PAA, erosion rate attained a 94.54% of amoxicillin released in the first
increased as the amount of this polymer augmented in 15 min. This may be attributed to the fact that PAA
the hydrogel (PAA:CS:A (2.5:2.5:2)—1.75 m and presented a greater amount of free carboxylic pendant
(5:2.5:2)—1.75 m). Possibly, the presence of greater groups that would lead to a major expansion of the
PAA amounts within the complex produced a structure, polymeric chains, increasing the rate of movement of the
 P.M. de la Torre et al. / Biomaterials 24 (2003) 1459–1468 1463

Scheme 1. Schematic diagrams of swelling/eroding of polyionic complexes: PAA:CS:A (0:2.5:2)—1.75 m (A); PAA:CS:A (1:2.5:2)—1.75 m (B), and
PAA:CS:A (1:2.5:2)-5.25 m (C).

erosion front. Shin et al. [9] have reported that a similar In comparison to the formulations with 2.5 of CS
expansion of the polymeric chains was achieved when content, Fig. 3A shows that the presence of a higher
there was an increase of ionic groups in the formulation. amount of CS in the hydrogel caused an increase in the
To investigate more precisely the effect of interpoly- maximum Sw ratio. This could be attributed to the effect
mer complex formation on the release of amoxicillin, the of cationic materials, like CS, that contain pendent
results were analyzed according to the Peppas equation groups such as amines. These groups ionize in a low-pH
[15]. The hydrogel PAA:CS:A (5:2.5:2)—1.75 m did not environment (SGF), causing increased electrostatic
present an adequate fitting to this equation because it repulsions [4,18]. The Sw ratio was increased by the
showed a high initial burst effect. In other cases, the n presence of PAA, attaining a maximum Sw at 17.59 and
values were lower than 0.5 (Table 1). These non-Fickian 17.80 for PAA:CS:A ratios of (1:5:2)—1.75 m and
behaviors may suggest that the amoxicillin release from (2.5:5:2)—1.75 m, respectively. However, the polyionic
hydrogels was controlled by a combination of drug complex with the greatest amount of both components,
diffusion and swelling dissolution. PAA:CS:A (5:5:2)—1.75 m, presented a maximum Sw
Fig. 3 shows the swelling and release kinetics of ratio slightly lower than in the latter formulations,
different interpolymer complexes with increasing owing to the fact that this hydrogel was less porous and
amounts of PAA and a higher proportion of CS (5.0). the existence of a higher amount of PAA and CS chains
1464 P.M. de la Torre et al. / Biomaterials 24 (2003) 1459–1468

Table 1
Initial burst effects (%), kinetic constants (K), release exponents (n), determination coefficients (r2 ) following linear regression of release data of
amoxicillin CS–PAA interpolymer complexes produced in 1.75 m acetic acid medium

Polyionic complex PAA:CS:A Initial burst release (%) n k r2

(0:2.5:2)—1.75 m 46.03 0.32 0.17 0.9925

(1:2.5:2)—1.75 m 29.32 0.42 0.09 0.9992
(2.5:2.5:2)—1.75 m 61.51 0.16 0.38 0.9945
(0:5:2)—1.75 m 24.78 0.48 0.07 0.9983
(1:5:2)—1.75 m 54.04 0.18 0.99 0.9919
(5:2.5:2)—1.75 m 94.54 — — —
(2.5:5:2)—1.75 m 37.61 0.31 0.17 0.9966
(5:5:2)—1.75 m 74.26 0.11 0.55 0.9957

Fig. 3. Swelling/eroding (A) and release kinetics (B) in SGF of freeze-dried polyionic complexes with a CS proportion of 5.0: PAA:CS:A (0:5:2)—
1.75 m (-B-), PAA:CS:A (1:5:2)—1.75 m (-&-), PAA:CS:A (2.5:5:2)—1.75 m (-W-), and PAA:CS:A (5:5:2)—1.75 m (-J-).

within the interpolymer complex obstructed the uptake linked CS-based hydrogels obtained by different pro-
of solvent in the network [16]. cesses [5–7].
A greater number of electrostatic repulsions generated Fig. 3B shows that the existence of greater amounts of
a fast swelling and a slow down in the erosion profile for CS caused slower release profiles, PAA:CS:A (0:5:2)—
PAA:CS:A polyionic complexes with ratios of 1.75 m, and that it could be attributed to a greater
(2.5:5:2)—1.75 m and (5:5:2)—1.75 m, respectively number of bonds in the network structure that retarded
[4,18]. The electrostatic interaction between both poly- the amoxicillin diffusion.
meric chains in the network was not permanent, Drug release from these interpolymer complexes
originating a faster erosion rate than in other cross- followed a non-Fickian diffusion mechanism (see
 P.M. de la Torre et al. / Biomaterials 24 (2003) 1459–1468 1465

Fig. 4. Swelling/eroding profiles at 371C in SGF (pH=1.2) of freeze-dried polyionic complexes produced in different hydrogel-forming media. (A)
Formulations with a CS proportion of 2.5: PAA:CS:A (1:2.5:2)—1.75 m (-’-), PAA:CS:A (1:2.5:2)—3.50 m (-X-) and PAA:CS:A (1:2.5:2)—5.25 m
(-E-). (B) Formulations with a CS proportion of 5.0: PAA:CS:A (2.5:5:2)—1.75 m (-W-), PAA:CS:A (2.5:5:2)—3.50 m (-+-), and PAA:CS:A
(2.5:5:2)—5.25 m (-}-).

Table 1). Most of our hydrogels show that low n values respectively (Fig. 4B). There are several possible ex-
are related to a high initial burst release. The polyionic planations for the drastic decrease in maximum Sw
complex PAA:CS:A (2.5:5:2)—1.75 m showed an n value ratios of the polyionic complexes developed in high ionic
of 0.31, similar to the hydrogels containing only CS. strength hydrogel-forming media, firstly, absence of
However, this polyionic complex showed a suitable pores on the surface would result in decreased water rise
release in an early stage (37.61% initial burst effect by capillary action leading to decreased swelling,
percentage). These results suggested that the release was secondly, increasing the ionic strength would cause a
controlled by the diffusion of amoxicillin and a greater collapse in the structure of the network, there-
combination of swelling and dissolution of hydrogel. fore erosion was quick and extensive. Water uptake was
reduced by a lower swelling, thus, the repulsion between
3.2. Study of the influence of ionic strength in two amino groups was decreased hindering the uncoiling and
different interpolymer complexes expansion of polymer chains (see Scheme 1C). Similar
result with water uptake was obtained by Coviello et al.
The effect of the hydrogel-forming medium on [19] with a crosslinked system from scleroglucan
hydrogel swelling and eroding was studied (see Fig. 4). derivative, which was remarkably affected by ionic
Results showed that increasing the ionic strength of the strength.
medium greatly reduced the maximum Sw ratio. For- Fig. 5 shows the effect of the hydrogel-forming
mulations PAA:CS:A (1:2.5:2)—3.50 m and PAA:CS:A medium on amoxicillin release. Fig. 5A shows that
(1:2.5:2)—5.25 m had similar results at 3.03 and 3.72, polyionic complexes PAA:CS:A (1:2.5:2)—3.50 m and
respectively (Fig. 4A), and hydrogels PAA:CS:A PAA:CS:A (1:2.5:2)—5.25 m exhibited high drug release
(2.5:5:2)—3.50 m and PAA:CS:A (2.5:5:2)—5.25 m in the first hour at 95.26% and 92.77%, respectively,
achieved the maximum Sw ratio at 6.06 and 8.94, much faster than PAA:CS:A (1:2.5:2)—1.75 m. These
1466 P.M. de la Torre et al. / Biomaterials 24 (2003) 1459–1468

Fig. 5. Release profiles at 371C in SGF (pH=1.2) of freeze-dried polyionic complexes produced in different hydrogel-forming media. (A)
Formulations with a CS proportion of 2.5: PAA:CS:A (1:2.5:2)—1.75 m (-’-), PAA:CS:A (1:2.5:2)—3.50 m (-X-) and PAA:CS:A (1:2.5:2)—5.25 m
(-~-). (B) Formulations with a CS proportion of 5.0: PAA:CS:A (2.5:5:2)—1.75 m (-W-), PAA:CS:A (2.5:5:2)—3.50M (-+-), and PAA:CS:A
(2.5:5:2)—5.25 m (-}-).

Table 2
Initial burst effects (%), kinetic constants (K), release exponents (n), determination coefficients (r2 ) following linear regression of release data of
amoxicillin from CS–PAA interpolymer complexes produced in 3.50 and 5.25 m acetic acid media

Polyionic complex PAA:CS:A Initial burst release (%) n k r2

(1:2.5:2)—3.50 m 86 — — —
(1:2.5:2)—5.25 m 67.23 — — —
(2.5:5:2)—3.50 m 46.13 0.21 0.26 0.9966
(2.5:5:2)—5.25 m 40.30 0.24 0.21 0.9970

interpolymer complexes presented a high proportion of of the hydrogel-forming medium, achieving values from
drug released during the initial hydration. The release 58.10% to 61.31% in the first hour (see Fig. 5B).
from these hydrogels is too rapid for them to be useful Possibly, this could be attributed to the existence of a
for controlled drug delivery. A similar release by higher amount of both polymers within the hydrogel
increasing the ionic strength was obtained by Inukai that produced a greater entwining of both polymeric
et al. [20] in hyaluronate hydroxyethyl acrylate blend chains within the network. In these conditions, drug was
hydrogels. greatly entrapped into the mesh space of the network
However, the release rates from formulations with a structure. Tabata et al. [21] have reported similar results,
higher CS content were independent of the ions amount showing that an increased ionic strength in the solution
 P.M. de la Torre et al. / Biomaterials 24 (2003) 1459–1468 1467

did not affect the drug release, indicating the occurrence stomach for H. pylori treatment, due to its controlled
of coordinate bonding of drug to specific groups within release profile, porosity and good swelling-erosion
the hydrogel. kinetics.
Table 2 shows that interpolymer complexes developed
in the hydrogel-forming medium with greater ion
concentration (3.50 and 5.25 m). A good fitting of
polyionic complexes that contain a lower amount of Acknowledgements
CS (PAA:CS:A 1:2.5:2 3.50 and 1:2.5:2 5.25 m) was
difficult to perform due to the high initial burst release. This work was supported by FIS project no. PI
These could be related to a reduced swelling in an early 020418 and Complutense project no. PR78/02-11056.
stage and to a weak electrostatic interaction during a
late stage.
Table 2 also shows hydrogels with a higher CS References
content, PAA:CS:A (2.5:5:2)—3.5 m and PAA:CS:A
(2.5:5:2)—5.25 m, which presented a non-Fickian re- [1] Kees F, Holstege A, Ittner K, Zimmermann M, Lock G,
lease, with similar n values (0.21, 0.23, respectively). Scholmerich J, Grobecker H. Pharmacokinetic. Interaction
Similar non-Fickian release has been obtained by Anh between proton pump inhibitors and roxithromycin in volunteers.
et al. [15]. These results pointed out: firstly, that the Aliment Pharmacol Ther 2000;14:407–12.
[2] Atherton JC, Cullen DJE, Kirk G, Hawkey C, Spiller R.
ionic strength of hydrogel-forming medium had less Enhanced eradication of Helicobacter pylori by pre-versus post-
effect on amoxicillin release in these interpolymer prandial amoxycillin suspension with omeprazole: implications
complexes, and secondly, that swelling and dissolution for antibiotic delivery aliment. Pharmacol Ther 1996;10:631–5.
hydrogel kinetics were the predominant mechanisms of [3] Patel VR, Amiji MM. Preparation and characterization of freeze-
release in diffusion of amoxicillin. dried chitosan-poly (ethylene oxide) hydrogels for site-specific
antibiotic delivery in the stomach. Pharm Res 1996;13:588–93.
[4] Risbud MV, Hardikar AA, Bhat SV, Bhonde RR. PH-sensitive
freeze-dried chitosan-polyvinyl pirrolidone hidrogels as controlled
4. Conclusions release sytems for antibiotic delivery. J Control Rel 2000;68:
23–30.
These results lead to the conclusion that for a fixed [5] Noble L, Gray AI, Sadiq L, Uchegbu IF. A non-covalently cross-
linked chitosan based hydrogel. Int J Pharm 1999;192:173–82.
amount of CS, the presence of PAA in the formulation, [6] Chellat F, Tabrizian M, Dumitriu S, Chornet E, Rivard CH,
produced electrostatic polymer/polymer interactions Yahia L. Study of biodegradation behavior of chitosan-xanthan
within the network, between cationic and carboxilic microspheres in simulated physiological media. J Biomed Mater
chains of both polymers. In gastric simulated fluid, these Res 2000;53:592–9.
kind of interactions caused a greater swelling extent and [7] Bae YH, Kim SW. Hydrogel delivery systems based on polymer
blends, block-copolymers or interpenetrating networks. Adv
a slower eroding rate of these interpolymer complexes, Drug Del Rev 1993;11:109–35.
compared to freeze-dried hydrogels without PAA (see [8] Dumitriu S, Chornet E. Inclusion and release of proteins from
Scheme 1A and B). The presence of higher CS content in polysaccharide-based polyions complexes. Adv Drug Del Rev
the complexes, for example PAA:CS:A (2.5:5:2)— 1998;31:223–46.
[9] Shin HS, Kim SY, Lee YM. Indomethacin release behaviors from
1.75 m, generated a higher repulsion between the
pH and thermoresponsive poly (vinyl alcohol) and poly (acrylic
polymeric chains, therefore a further increase in its acid) IPN hydrogels for site-specific drug delivery. J Appl Polym
maximum Sw ratio and a more sustained erosion profile Sci 1997;65:685–93.
were obtained in the SGF. Finally, the presence of [10] Bell CL, Peppas NA. Modulation of drug permeation through
higher ion amount in the hydrogel-forming medium (see interpolymer complexed hydrogels for drug delivery applications.
Scheme 1C) produced a network collapse. These J Control Rel 1996;39:201–7.
[11] Khare AR, Peppas NA. Release behavior of bioactive agents from
changes of interactions caused a decrease in their Sw pH-sensitive hydrogels. J Biomater Sci Polym Ed 1993;4:275–89.
and a fast amoxicillin release for complexes with low [12] Bettini R, Colombo P, Peppas NA. Solubility effects on drug
amount of both polymers. However, increased ion transport through pH-sensitive, swelling-controlled release sys-
amount within the network did not affect the amox- tems: transport of theophylline and metoclopramide monohy-
icillin release in hydrogels with higher amount of both drochloride. J Control Rel 1995;37:105–11.
[13] Peniche C, Elvira C, San Roman J. Interpolymer complexes of
polymers. chitosan and polymethacrylic derivatives of salicylic acid:
The diffusion exponents (n) of these polyionic preparation characterization and modification by thermal treat-
complexes show that swelling and dissolution of the ment. Polymer 1998;39:6549–54.
hydrogel are the predominant release mechanisms on [14] Gabrielii I, Gatenholm P. Preparation and properties of hydrogels
based on hemicellulose. J Appl Polym Sci 1998;69:1661–7.
amoxicillin diffusion process.
[15] Ahn JS, Choi HK, Chun MK, Ryu JM, Jung JH, Kim YU, Cho
PAA:CS:A (1:2.5:2)—1.75 m and PAA:CS:A CS. Release of triamcenolone acetonide from mucoadhesive
(2.5:5:2)—1.75 m polyionic complexes could be the most polymer composed of chitosan and poly(acrylic acid) in vitro.
suitable hydrogels for site-specific drug delivery to the Biomaterials 2002;23:1411–6.
1468 P.M. de la Torre et al. / Biomaterials 24 (2003) 1459–1468

[16] Wang H, Li W, Lu Y, Wang Z. Studies on chitosan and poly [19] Coviello T, Grassi M, Rambone G, Alhaique FA. Crosslinked
(acrylic acid) interpolymer comples. I. Preparation structure pH- system from scleroglucan derivative; preparation and character-
sensitivity and salt sensitivity of complex-forming poly (acrylic ization. Biomaterials 2001;22:1899–909.
acid): chitosan semiinterprenetrating polymer network. J Appl [20] Inukai M, Jin Y, Yomota C, Yonese M. Preparation and
Polym Sci 1997;65:1445–50. characterization of hyaluronate–hydroxyethyl acrylate blend
[17] Munjeri O, Collet JH, Fell JT. Hydrogel beads based on amidated hydrogel for controlled release device. Chem Pharm Bull
pectins for colon-specific drug delivery: the role of chitosan in 2000;48:850–4.
modifying drug release. J Control Rel 1997;46:273–8. [21] Tabata Y, Matsui Y, Ikada Y. Growth factor release from
[18] Peppas N, Bures P, Leobandung W, Ichikawa H. Hydrogels in amylopectin hydrogel based on copper coordination. J Control
pharmaceutical formulations. Eur J Pharm Biopharm 2000;50:27–46. Rel 1998;56:135–48.