Journal of Controlled Release 103 (2005) 499 – 510 www.elsevier.com/locate/jconrel

Influence of micro-environmental pH on the gel layer behavior and release of a basic drug from various hydrophilic matrices Manthena V.S. Varma, Aditya M. Kaushal, Sanjay Garg* National Institute of Pharmaceutical Education and Research (NIPER), Sector 67, SAS Nagar, Punjab 160 062, India Received 29 May 2004; accepted 20 December 2004 Available online 25 January 2005

Abstract The purpose of this investigation is to understand the influence of gastrointestinal (GI) pH on the gel layer formation and its dynamics for various hydrophilic/swellable matrices, in the process of developing a pH-independent controlled release system for a basic drug, oxybutynin hydrochloride (OXB). Cylindrical matrices (8-mm diameter) without and with fumaric acid, were readily prepared by direct compression. Formulations were evaluated for in vitro drug release, and gel layer dynamics was studied by viscosity measurements and texture profiling analysis. In the in vitro drug release study, OXB, which shows pHdependent solubility, showed faster release from all the matrices in pH 1.2 medium. Release rates enhanced to a lesser extent with change of medium from pH 6.8 to pH 1.2, for HPMC polymer matrices. Anionic polymer matrices showed drastic differences in the release rates when medium was changed from pH 6.8 to pH 1.2. Addition of fumaric acid to matrices demonstrated pH-independent drug release, which was attributed to the micro-environmental pH manipulation within the hydrated gel layer. Viscosity and texture profiling studies revealed that saturation solubility of drug at swelling front play a major role in the pH-dependent drug release from HPMC matrices, while both saturation solubility and the altered gel consistency as a function of pH are involved with anionic polymer matrices. Presence of fumaric acid in HPMC matrices showed efficient retardation and pH-independent drug release. In conclusion, understanding the influence of GI physiological pH on the gel layer dynamics and manipulating the micro-environmental pH provides efficient and predictable in vivo performance from these swellable cylindrical matrices. D 2004 Elsevier B.V. All rights reserved. Keywords: Hydrophilic/swellable matrices; pH-independent drug release; Gel layer dynamics; Texture profiling; Oxybutynin hydrochloride

1. Introduction

* Corresponding author. Present affiliation: School of Pharmacy, The University of Auckland, Private Bag 92-019, Auckland, New Zealand. Tel.: +64 9 373 7599x82836; fax: +64 9 367 7192. E-mail address: [email protected] (S. Garg). 0168-3659/$ - see front matter D 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.jconrel.2004.12.015

The goal of controlled release (CR) delivery systems is to provide desirable delivery profile that can achieve predictable plasma levels. However, physiological variables in GI tract including GI pH, gastric residence time, intestinal motility and the GI

500

M.V.S. Varma et al. / Journal of Controlled Release 103 (2005) 499–510

contents, may alter the in vivo drug release performance of peroral CR systems [1]. Protic and ampholytic drugs demonstrate pH-dependent solubility profiles based on their ionizable groups. Thus, drug release rate from controlled release matrices, especially for these drugs, may vary as a function of their movement into various segments of GIT, leading to inefficient drug delivery and high inter-subject variability. Several attempts based on the incorporation of acidic excipients in the matrix, which keep the pH within the system and hence the solubility of ionizable drugs constant, irrespective of the dissolution medium, proved efficient drug release from matrices. Gabr showed pH-independent drug release of papaverine HCl from cellulose acetate and bees wax inert matrix tablets in the presence of an organic acid [2]. Streubel et al. studied the effect of fumaric acid, sorbic acid and adipic acid on HPMC and ethyl cellulose matrices and demonstrated pH-independent release of verapamil HCl with fumaric acid containing HPMC matrices [3]. Vinopocetine release from HPMC matrices showed a linear relationship between the release rates and the proportion of citric acid added to the matrix [4]. Tatavarti et al. modulated matrix micro-environmental pH by incorporating acidic polymers and thus enhanced the local solubility and demonstrated acidic polymer concentration dependent micro-environmental pH and the release of papaverine HCl [5]. Incorporation of acidic excipients is a more useful approach for basic drugs because, (i) solubility of basic drugs decreases exponentially with increase in pH and thus may precipitate within the matrix as it proceeds into lower part of intestine, leading to incomplete drug release, (ii) drug release being a function of drug solubility, release rates fall as the matrix moves to higher pH range, thereby introducing variability associated with GI transit. Swellable/hydrophilic matrices are characterized by the formation of gel layer on the matrix surface. Phenomena that govern gel layer formation and consequent drug release rate are water penetration, polymer swelling, drug dissolution, diffusion and matrix erosion. This gel layer formation and its stability that defines the kinetics of drug delivery from matrix systems, is controlled by the concentration, viscosity and chemical structure of the

polymer(s) [6]. Apart from the physicochemical properties of the drug (pK a and solubility), the polymer dissolution rate is also a dominant process in drug release, which in turn is dependent on the matrix shape, dimension, type of dissolution medium and the polymer [7,8]. A swelling and an erosion front form during the swelling process of hydrophilic matrices. The drug is soluble completely or to a certain extent in the gel layer based on the saturation solubility and the volume of medium available at the moving fronts. Thus, when a drug has a finite solubility, a layer of undissolved drug can be detected in the gel [9]. The position of this layer, separating the undissolved drug from dissolved drug gel region, called diffusion front, is thus highly dependent on the media conditions. Drugs demonstrating highly pH-dependent solubility thereby show complex gel layer behavior as the matrix formulations travel down the GIT. Difference in the gelation properties of the polymers as a function of media pH further complicates the drug release process and its predictability. The present investigation was an attempt to understand the influence of GI physiological pH on the gel layer formation and its dynamics; and on the drug release characteristics of a basic drug, OXB, from non-ionic and anionic hydrophilic matrices. We also studied the influence of manipulating microenvironmental pH of the hydrating swollen layer on the gel layer dynamics and ultimately the drug release characteristics. Two non-ionic polymers (hydroxypropyl methylcellulose, METHOCEL K4M and K15M) and two anionic polymers (carbopol 971P and xanthan gum) were compared with respect to change in mechanisms of drug release and the physical structure of matrix in GI physiological pH conditions based on drug release, viscosity and texture profiling studies.

2. Materials and methods 2.1. Materials Oxybutynin hydrochloride (USP) was received as gratis sample from Unichem Lab. Ltd. (Mumbai, India). Hydroxypropyl methylcellulose (HPMC) polymers were supplied by Dow Chemicals (Colorcon

M.V.S. Varma et al. / Journal of Controlled Release 103 (2005) 499–510

Asia Pvt. Ltd., Mumbai, India) as METHOCEL K4M and K15M having nominal viscosity of 4000 and 15,000 cps (2% aqueous solutions), respectively. Carbopol 971P (CP) (The BFGoodrich Co., Cleveland, OH) and xanthan gum (XG) (Xantural 75, Waterfield, UK) were received as gift samples from Signet, (Mumbai, India). Fumaric acid was procured from Sigma Chemicals (MO, USA). Magnesium stearate (Mallinckrodt, USA), Aerosil 200 (Degussa AG, Frankfurt), and directly compressible lactose (Flowlac 100, Meggle GMBH, Wasserburg) were obtained from Unichem Lab. (Baddi, India). Acetonitrile HPLC grade and methanol HPLC grade were obtained from JT Baker (Mexico) and Ranbaxy Fine Chem. Ltd. (SAS Nagar, India), respectively. All other chemicals and reagents used were of analytical grade. 2.2. Solubility of OXB Solubility of OXB was determined in simulated gastric fluid without enzymes (SGF pH 1.2), pH 6.8 phosphate buffer and in pH 10.2 phosphate buffer, by adding an excess of drug to the media (n=3). After equilibrating on shaker water bath (37F1 8C) for 24 h, samples were filtered through 0.22 Am filter and the drug concentration in filtrate was determined by HPLC. 2.3. Tablet preparation The tablets were made up of 5% OXB, polymers (varying proportions of K4M, K15M, CP and XG), fumaric acid (FA) (varying proportions), 1% magnesium stearate, 0.25% aerosil and lactose q.s. as filler. Blending of the components was carried out in a laboratory model drum blender (Kalweka, Mumbai, India) and the blend was compressed using 8 mm diameter, flat faced die and punches (CMS-15, Cadmach tablet press, India). Hardness for all the formulations was adjusted in the range of 11–14 Kp (TBH20, Erweka Instruments, CT, USA) and the weight of each tablet was 200F4 mg. Batch size for each formulation was 100 g (500 tablets).

501

apparatus (Electrolab, Mumbai, India). The operating speed of the dissolution basket was 100 rpm, media used were 500 ml of SGF (pH 1.2) or pH 6.8 phosphate buffer maintained at 37F0.5 8C. Progress of the dissolution was monitored by withdrawing samples at predetermined intervals and assayed by HPLC. The data were indicated as meanFS.D. of three replicates for each dissolution. The drug release data was analyzed and interpreted with the simple power law expression, which can best describe the kinetics of drug release from swellable hydrophilic matrices [10,11]. Mt ¼ kt n Ml

ð1Þ

where M t /M l, is the fraction of drug release at time t, k, the kinetic constant and n, the release exponent that characterizes the mechanism of drug release. For a cylindrical matrix, values of n=0.45 indicate Fickian drug release, 0.45bnb0.89 indicate an anomalous (non-Fickian) release, whereas values of n=0.89 indicate case II transport kinetics. Fickian diffusional release occurs by molecular diffusion of the drug due to a chemical potential gradient. Case II relaxational release is the drug transport mechanism associated with stresses and state transition in hydrophilic glassy polymers, which swell in water or biological fluids. The two phenomena controlling drug release from swellable matrices are considered additive [12,13]. Mt ¼ k1 t m þ k2 t 2m Ml

ð2Þ

where the first term is contribution of Fickian diffusion and the second term is the case II relaxational contribution. m value of 0.45 was taken based on the aspect ratio of the matrices. The ratio of relaxational over Fickian contribution is given by [10,12,14] R k2 ¼ tm : F k1

ð3Þ

2.4. Dissolution studies

To describe the shape of release profile, data were also fitted to Weibull distribution [15].
M ¼ 1  exp  at b ð4Þ

The dissolution behavior of OXB from various formulations was carried out on USP rotating basket

where, M is accumulated fraction of drug in solution at time t. a is a scale parameter given as k1/(1h) and

502

M.V.S. Varma et al. / Journal of Controlled Release 103 (2005) 499–510

b is shape parameter given by (1h). k1 is a constant not dependent on time with units (time)h1 and h is a pure number.

test speed and post-test speed 0.2 mm s1; maximum compression force, 40 N; and auto trigger force 0.005 N. Data were captured at a rate of 200 points s1 by Texture Expertk for Windows software.

2.5. Viscosity determinations 2.7. HPLC analysis Polymer solutions of K4M, K15M, CP and XG were prepared by thorough dispersion of 1%w/v polymer in buffers of pH 1.2, 3.8, 5.2 and 6.8, under agitation. Solutions were maintained at 37F1 8C for 48 h to ensure complete hydration. Viscosity was performed on a cup and bob viscometer (model DVIII+, Brookfield, MA). All measurements were carried out at room temperature (24F2 8C). Apparent viscosities were measured at a shear rate of 189 S1 after allowing the solutions to equilibrate for 10 min. The mean of not less than three replicates was taken to represent viscosity. To study the effect of fumaric acid on the pH and viscosity of polymer solutions, 0.1%w/v of fumaric acid was added to polymer solutions (in pH 6.8) and pH and viscosity were recorded after equilibrating for 48 h.

Chromatography was carried out on Shimadzu HPLC equipped with Class VP software for data processing. Samples were analyzed at 25 8C on a 4.6mm250-mm 5 Am cyano column (Hypersil BDS CPS) attached to a Hypersil BDS CPS precolumn. UV–Vis detector (SPD-10A VP) was set at 203 nm and mobile phase was pumped at a flow rate of 1.5 ml/min. Mobile phase consisting of Solvent A and acetonitrile in the proportion of 60:40 was prepared, where Solvent A is a mixture of water, methanol and triethylamine in the proportion of 3200 ml, 800 ml and 0.9 ml, respectively, with pH 3.5 (adjusted with phosphoric acid). Mobile phase was filtered through 0.45 Am nylon filter (Millipore) and deaerated in ultrasonic bath (Branson sonicator). Sample injections of 50 Al each were made automatically using autosampler (SIL-10AD VP).

2.6. Texture Profiling on hydrated matrices Gel layer formation and its dynamics as a function of time was evaluated by texture profiling analysis (Texture Analyzer, TA XT2i, Stable Micros Systems, UK). Texture profiling for hydrated matrices of four different polymers with and without FA in different dissolution media were carried by the procedure reported earlier, with minor modifications [16,17]. One planar base of each tablet was sealed off by fixing the tablet on to a glass coverslip using a fixer solution (Fevikwik, Fevi India). Each sealed tablet was then placed in the basket assembly and proceeded as mentioned in dissolution studies. This allows medium ingression from all the directions and simulates the actual process of gel dynamics that occurs in the dissolution studies, as the tablets freely move in the basket. At predetermined time intervals, tablets (n=3) were removed and subjected to texture analysis. A flat-tipped steel probe, 2 mm in diameter, was connected to a force transducer that measured the force of resistance encountered by the probe during advancement into the sample. Test parameters fixed for all the samples included, pre-test speed, 1 mm s1;

3. Results and discussion 3.1. pH-dependent dissolution of OXB OXB, a monoprotic base with pK a 8.04, showed pH-dependent solubility in the GI pH range (Fig. 1). The solubility of OXB is 754F25.2 mg ml1 at pH 1.2, 1.24F0.58 mg ml 1 at pH 6.8, and 0.021F0.0015 mg ml1 at pH 10.2. Based on the pK a and the intrinsic solubility of the drug (saturation solubility above pK a), theoretical solubility profile was generated [18]. Theoretical and experimental data, which is in agreement to the literature values [19], indicated exponential decline in the solubility of OXB from pH 1.2 to pH 6.8. A remarkable difference in the drug release from all the hydrophilic matrices was observed in pH 1.2 SGF and pH 6.8 phosphate buffer. Fig. 2 shows that K4M, K15M, CP and XG at 75%, 50%, 15% and 10%w/w proportions, respectively, controlled OXB release in pH 6.8 phosphate buffer. These polymer concentrations were optimized in the preliminary

M.V.S. Varma et al. / Journal of Controlled Release 103 (2005) 499–510

Solubility of OXB (mg/ml)

100000

formulations but was variable from polymer-topolymer matrices. The enhanced release in pH 1.2 could be due to (i) high solubility of OXB at lower pH and/or (ii) pH-dependent swelling and gel layer dynamics of hydrophilic polymers. HPMC polymers (K4M and K15M) showed relatively less hastening of release kinetics than shown by CP and XG. In case of HPMC matrices (K4M, 50%w/w), approximately 34% and 10% of the drug was released after 2 h in SGF and pH 6.8 phosphate buffer, respectively. After, 6 h, 64.2% of the drug was released in pH 1.2 medium, while only 27.6% was released in phosphate buffer. However, CP and XG polymeric matrices showed quite a faster release in SGF over pH 6.8 phosphate buffer. The reason for this could be the ionic nature of these polymers where pH may influence the swelling behavior and the gel layer dynamics, apart from pH-dependent solubility of OXB. Fig. 3 illustrates the R/F ratio values versus fraction released from the matrices for all four hydrophilic polymers, in both pH 6.8 and pH 1.2 media. Model was

10000 1000 100 10 pKa = 8.04

1 0.1 0.01 0

2

4

6

8

10

12

14

pH

Fig. 1. Theoretical (E) and experimental (5) pH solubility profiles of OXB HCl at 37 8C. Theoretical profile was generated by the equation S/S o=10+pK apH+1, where S and S o are solubilities at test pH and at any pH above pK a, respectively. S o was determined experimentally at pH 10.2 and pK a value of 8.04 was taken from literature [19]. Experimental data indicate meanFS.D. of three independent determinations.

studies, to obtain near zero-order release however, with incomplete release (data not shown). OXB release in pH 1.2 was relatively faster from all the

a)

50% K15M (pH 1.2) 50% K15M + 10% FA (pH 6.8) 50% K15M + 5% FA (pH 6.8) 50% K15M (pH 6.8)

100 80

% Release

80

% Release

b)

75% K4M (pH 1.2) 75% K4M + 10 % FA (pH 6.8) 75% K4M + 5% FA (pH 6.8) 75% K4M (pH 6.8)

100

60 40

60 40 20

20 0

0 0

2

4

6

8

10

0

12

2

4

Time (hrs)

d) 15% CP (pH1.2) 15% CP + 10% FA (pH 6.8) 15% CP + 5% FA (pH 6.8) 15% CP (pH 6.8)

100

8

10

12

8

10

12

10% XG (pH 1.2) 10% XG + 10% FA (pH 6.8) 10% XG + 5% FA (pH 6.8) 10% XG (pH 6.8)

100 80

% Release

80

6

Time (hrs)

c)

% Release

503

60 40

60 40 20

20

0

0 0

2

4

6

Time (hrs)

8

10

12

0

2

4

6

Time (hrs)

Fig. 2. Effect of dissolution medium pH and the presence of fumaric acid in the matrices on OXB release from (a) 75% HPMC K4M, (b) 50% HPMC K15M, (c) 15% carbopol 971P, and (d) 10% xanthan gum matrices.

504

M.V.S. Varma et al. / Journal of Controlled Release 103 (2005) 499–510 K4M (pH 6.8) K4M (pH 1.2) K15M (pH 6.8) K15M (pH 1.2) CP (pH 6.8) CP (pH 1.2) XG (pH 6.8) XG (pH 1.2)

3.0 2.5

R/F

2.0 1.5 1.0 0.5 0.0 0

0.1

0.2

0.3

0.4

0.5

0.6

Fraction Released Fig. 3. Ratio of relaxational over Fickian contribution versus fraction released for non-ionic (K4M, K15M) and anionic (CP, XG) polymer matrices in pH 6.8 and pH 1.2 dissolution media.

fitted up to 60% of drug release. HPMC polymers showed no significant difference in terms of mechanism with change in media pH. Relaxational process predominated for most of the drug release over Fickian diffusion with K15M matrices, while K4M matrices showed equal contribution of both the mechanisms (R/F c1). CP matrices in pH 6.8 showed equal contribution of both the phenomenon but diffusional release predominated in the acidic medium. XG polymer matrices showed high relaxational contribution in pH 1.2 but Fickian diffusion in pH 6.8. XG matrices showed a more homogenous release (b=1, h=0) in pH 1.2 than in pH 6.8, while K4M matrices showed homogenous release in 6.8 pH. K4M matrices showed an exponential release (bb1, 0bhb1) with a steeper initial burst in pH 1.2 (Table 1). This can be attributed to high solubility and thus fast dissolution of the drug from the matrix surface, well before the formation of a matured gel layer. The rate constant k1 increased in pH 1.2 from all the matrices, however, large differences in the rate constant at pH 1.2 and 6.8 were observed for CP and XG matrices. 3.2. Influence of fumaric acid on OXB release: modulation of micro-environmental pH Addition of organic acids to the formulation could create a constant acidic micro-environment inside the gel layer, irrespective of the surrounding dissolution medium. The pH inside the gel layer is expected to be acidic and thus the solubility of basic drug will be

high leading to faster and usually complete release. Ideally, organic acids with high acidic strength (low pK a values) and relatively low solubility in lower pH range are suitable as they can provide low pH in the matrix for longer periods even at low proportions. Because of their high solubility in high pH range, organic acids can act as pore formers at high pH values, thus low proportions are always desirable. With these fundamental requirements, fumaric acid was selected for our studies, which has a pK a1 3.03 and pK a2 4.54 (25 8C) and aqueous solubility of 6.3 mg ml1 (25 8C) [20]. Addition of fumaric acid to hydrophilic matrices significantly increased the drug release in pH 6.8 phosphate buffer (Fig. 2). The increase in release from each was further found to be dependent on the amount

Table 1 Mathematical modeling and drug release kinetics from hydrophilic matrices without (in different media) and with fumaric acid, fitted to Weibull distributiona Formulation (dissolution medium)

a

b

hb

K1

R2

75% K4M (pH 6.8) 75% K4M+5% FA (pH 6.8) 75% K4M+10% FA (pH 6.8) 75% K4M (pH 1.2) 50% K15M (pH 6.8) 50% K15M+5% FA (pH 6.8) 50% K15M+10% FA (pH 6.8) 50% K15M (pH 1.2) 15% CP (pH 6.8) 15% CP+5% FA (pH 6.8) 15% CP+10% FA (pH 6.8) 15% CP (pH 1.2) 10% XG (pH 6.8) 10% XG+5% FA (pH 6.8) 10% XG+10% FA (pH 6.8) 10% XG (pH 1.2)

0.056 0.108

1.038 0.896

0.038 0.104

0.058 0.097

0.997 0.997

0.182

0.801

0.199

0.146

0.999

0.201 0.067 0.098

0.809 0.880 0.934

0.191 0.120 0.066

0.163 0.059 0.092

0.997 0.997 0.990

0.206

0.870

0.130

0.179

1.000

0.202 0.100 0.125

0.942 0.854 1.018

0.058 0.146 0.018

0.190 0.085 0.127

0.996 0.988 0.988

0.258

0.868

0.132

0.224

0.986

0.350 0.068 0.151

0.780 0.832 0.930

0.220 0.168 0.070

0.273 0.056 0.140

0.978 0.982 0.997

0.318

0.996

0.004

0.317

0.997

0.340

1.010

0.010

0.343

1.000

a

Drug release data was fitted to Eq. (4) using Sigmastat 2.03 software (SPSS, USA). b 0bhb1, rate of release is monotonically decreasing with time; hb0, increasing initially and decreasing afterwards; hc1, homogeneous release [15].

M.V.S. Varma et al. / Journal of Controlled Release 103 (2005) 499–510

acid. These results once again substantiate the hypothesis that low pH influences the formation of gel layer and its mechanical stability. 3.3. Viscosity of gelatinous layer To further confirm the hypothesis of pH-dependent gel layer behaviour, the viscosity studies were carried out for 1%w/v solution of each polymer in media with pH 1.2, 3.8, 5.2 and 6.8. When hydrophilic matrices come into contact with water, they absorb water and swell to form a gel, which serves as a barrier to drug diffusion. Since movement of the drug solute through the matrix system is diffusion controlled, it may be expected from the Stokes–Einstein equation that the drug release process will be inversely proportional to the mechanical strength/viscosity of the gel layer. Since gel layer behavior is both dynamic and gradient, viscosity measurement could be difficult. Thus an assumption was made that the bulk viscosities of the diluted polymer solutions are indicative of the diffusion resistance experienced by the solute. The apparent viscosities of the polymer solutions at various pH are presented in Fig. 4. Viscosities of HPMC solutions were constant irrespective of solution pH. HPMC polymers are non-ionic and therefore the solubility and swelling behavior are not influenced by pH. K15M solution (1%w/v) showed higher viscosity 350

Brookefield Viscosity (CPs)

of fumaric acid. However, resulting release profiles of matrices with 10% fumaric acid (in pH 6.8 phosphate buffer) almost overlapped with the release of corresponding matrices without fumaric acid, in pH 1.2 SGF. This is in good agreement with the above discussed hypothesis that organic acids maintain a constant micro-environmental pH of the matrices/gel layer. The reason for lesser release of drug from matrices with 5%w/w fumaric acid could be that the amount of fumaric acid available as the buffer ingresses into the formulation may not be sufficient to maintain the micro-environmental pH to a level as provided by 10%w/w. It should be noted that pH lowering by fumaric acid is also influenced by the buffering capacity of the medium. Streubel et al. reported fumaric acid concentration independent drug release of verapamil hydrochloride from HPMC matrices, however the concentrations used in the study were much higher than those used in the present study [3]. Further, to monitor the micro-environmental pH, dissolution studies of HPMC K15M matrices with fumaric acid (5% and 10%w/w) were carried out in pH 6.8 phosphate buffer containing either thymol blue (0.001%w/v) or methyl red (0.001%w/v). Thymol blue is red (pHb2.8) or yellow (pHN2.8) at different pH ranges, while methyl red is red at acidic pH and yellow above pH 5.8. In case of K15M matrices with 10% fumaric acid, matrix core remained red in both media and yellowish from the surface. It was also observed that red and yellowish colored layers move towards the center, as a function of time and the matrix core remained red until 8 h in medium containing thymol blue. In case of matrices with only 5%w/w fumaric acid, tablet core remained red in only methyl red medium at 1, 2, 4 and 6 h. Furthermore, monitoring of pH of dissolution medium during dissolution showed no deviation from pH 6.8F0.2. Concisely, this set of experiments, indicated that 10%w/w fumaric acid maintained the micro-environmental pH below 2.8 for at least 8 h while 5%w/w fumaric acid maintained pH between 2.8 and 5.8 for at least 6 h. Increase in release with further increase in fumaric acid proportion, may occur not due to influence of micro-environmental pH, but because of the leaching of fumaric acid from the formulations [21,22]. As was observed with the differences in release kinetics with change of pH from 6.8 to 1.2, the drug release rate increased to large extent for matrices with CP and XG, when added with fumaric

505

K4M K15M

300

CP971P XG

250 200 150 100 50 0 0

2

4

6

8

pH Fig. 4. pH viscosity profiles of 1%w/v solutions of non-ionic and anionic hydrophilic polymers, after equilibration for 48 h (37 8C). Open points indicate the data of pH and viscosities of corresponding polymer solutions in pH 6.8 buffer, after addition of 0.1%w/v of fumaric acid and equilibrating for 48 h (37 8C). Each data point indicates meanFS.D. of three independent determinations.

M.V.S. Varma et al. / Journal of Controlled Release 103 (2005) 499–510

significant correlation between DE or MDT and viscosity of polymer solutions in the corresponding conditions was observed (Fig. 5). However, a trend was seen for CP and XG polymers, where DE is high at low viscosity and MDT is more when the viscosity was high. Choeng et al. showed good correlation between viscosities of HPMC solution and the T 50%, at various polymer proportions [26]. However, HPMC polymers showed no relation between viscosity and the release behavior of basic drug from their matrices in different pH conditions. These results indicate that the enhanced OXB release from HPMC polymers at

a)

80 1 K4M

2

K15M

CP

XG

2

1

1

60

DE

than K4M solution, which correlates to faster release of OXB from matrices with K4M (75%w/w) versus K15M (50%w/w), at pH 6.8 (Fig. 2). HPMC polymers are giant macromolecules and have a great affinity for water. When a polymer chain comes into contact with water, polymer water interaction replaces polymer– polymer attractions, and their hydrodynamic volumes also increase [23,24]. High viscosity grade HPMCs are made up of larger macromolecules and adopt more extended configuration and therefore produce highly viscous gel layer. The viscosity of CP and XG were found to be highly pH dependent with least viscosity at pH 1.2 and higher at pH 6.8. However, the viscosities of both the polymers at pH 6.8 were higher than the viscosities of HPMC polymers. This can be correlated to greater retardation of OXB release with 10% and 15%w/w of XG and CP, respectively, compared to 50%w/w of K15M and 75%w/w of K4M. CP and XG possessing carboxyl acid functional groups, have a difference in the process of gel layer formation and its mechanical stability as a function of pH. They are unionized at acidic pH and thus have less of intermolecular hydrogen-bonding. Viscosity rise was very abrupt in case of CP, while for XG a gradual increase in viscosity was observed from pH 1.2 to 6.8. The stability of XG to change in pH is relatively high and can be attributed to the rod shape of the molecule, its intermolecular hydrogen-bonding and to its pK a. Addition of 0.1%w/v fumaric acid to the polymer solutions in pH 6.8 buffer reduced the pH of all polymer solutions to ~2, after 48 h of equilibration. Apparent viscosities determined showed that fumaric acid presence has no effect on the viscosity of HPMC polymers, while viscosities of CP and XG drastically reduced and matched the viscosities of simple polymer solutions in the corresponding pH (Fig. 4). These results indicate that the manipulation of microenvironmental pH in HPMC matrices show no influence on the gel layer dynamics. However, gel layer of CP and XG matrices will be highly influenced by the pH of the medium or the presence of organic acids in the matrix. The dissolution efficiency (DE) and the mean dissolution time (MDT), determined from the drug release studies of matrices with and without 10%w/w fumaric acid, were compared to analyze the influence of viscosity on the drug release profile [11,25,26]. No

2

1

2

40

3 3

3

20

3

0 0

100

200

300

Viscosity of solution (Cps)

b)

6.0

3

3

5.0

MDT

506

3

3

2

4.0

1 1

1

2

2

2

3.0 1

2.0 0

100

200

300

Viscosity of solution (Cps) Fig. 5. Relationship between (a) dissolution efficiency and the apparent viscosity, and (b) mean dissolution time and the apparent viscosity of the polymer solutions. Key: 1, Viscosity in pH 1.2 and drug release from control matrix (without fumaric acid) in pH 1.2 medium; 2, Viscosity in the presence of 0.1%w/v fumaric acid and drug release of 10%w/w fumaric acid containing matrix in pH 6.8 medium; 3, Viscosity in pH 6.8 and drug release from control matrix (without fumaric acid) in pH 6.8 medium.

M.V.S. Varma et al. / Journal of Controlled Release 103 (2005) 499–510

pH 1.2 or from the matrix containing 10%w/w fumaric acid was mainly due to the saturation solubility difference at the swelling front and not due to the difference in the gel layer viscosity at different pH. On the other hand, in the case of CP and XG polymers, both the saturation solubility difference at the swelling front and the difference in the gelation of matrix and its stability leads to high release rates at low pH range. 3.4. Texture profiling and gel layer dynamics of swollen matrices Yang et al. [16] and Pillay and Fassihi [17] demonstrated the potential and application of texture analysis in characterization of the swelling behavior of hydrophilic matrices. The advantage of this technique over other techniques (e.g. imaging analysis) is its simultaneous interpretation of gel layer growth and the mechanical stability/consistency of the gel layer formed for any drug–polymer matrix [9,27,28]. Fig. 6 shows the typical force–displacement (F–D) profiles of three hydrophilic matrices (without and with fumaric acid) after 1.5 h of hydration in pH 6.8 medium. It is obvious from the figure that the three polymeric matrices have different degree of hydration and swelling, and presence of acidic microenvironment pH showed altered swelling behavior. F–D profiles are quantitative indications of the polymer viscosity/resistance within the gel layer as a function of distance. Gel layer thickness and the work done against the resistance offered by the gel layer, as determined from the F–D profiles, are represented as a function of time in Fig. 7. In the control matrices (without fumaric acid), K15M matrices showed consistent increase in the gel layer thickness and the work done with time in 6.8 pH medium. It showed an increase of gel layer thickness from 1.76F0.11 mm at 0.5 h to 4.93F0.54 mm at 2 h and to 6.66F0.65 mm at 3 h. Similarly, work done, which is an indicative of gel consistency changed from 11.88F0.56 N mm at 0.5 h to 45.43F3.05 N mm at 3 h. CP and XG matrices showed gel layer thickness of 2.74F0.19 mm and 3.8F0.23 mm by the end of 2 h. XG matrices completely hydrated after 2 h, and K15M and CP after 3 h. For anionic polymer matrices, the change in gel layer dynamics was found to be less, especially

507

a) Force (N) 45.00 CP

XG

K15M

40.00 35.00 30.00 25.00 20.00 15.00 10.00 5.00 0.00 0.0 -5.00

1.0

2.0

3.0 4.0 Distance (mm)

b) Force (N) 45.00 CP XG

40.00

K15M

35.00 30.00 25.00 20.00 15.00 10.00 5.00 0.00 0.0 -5.00

1.0

2.0

3.0 4.0 Distance (mm)

Fig. 6. Representative force–distance plots of hydrated gel layer of polymeric matrices (K15M 50%w/w, CP 15%w/w and XG 10%w/ w) (a) without fumaric acid and (b) with 10% fumaric acid, after 1.5 h of exposure to pH 6.8 phosphate buffer.

with CP matrices. Gel layer thickness (i.e. diffusional path length) and the stability of gel (depicted by work done) are two main components controlling drug release. However, with XG and CP matrices, even though both these parameters are less, showed more retardation in drug release with respect to HPMC polymeric matrix. Initial growth in the gel layer thickness occurred fairly fast with anionic polymers and not much change was observed as a function of time. Gel consistency showed a similar trend, indicating that these polymers showed faster hydration and thus faster matured gel layer.

508

M.V.S. Varma et al. / Journal of Controlled Release 103 (2005) 499–510

(b) 8

K15M

7

XG

6

CP

Gel layer thickness (mm)

Gel layer thickness (mm)

(a)

5 4 3 2 1

8

K15M+FA

7

XG + FA

6

CP + FA

5 4 3 2 1 0

0 0

0.5

1

1.5

2

2.5

0

3

0.5

1

Time (hrs)

(c)

2

2.5

3

(d) 70

70

K15M+FA

K15M

60

Workdone [N.mm]

Workdone [N.mm]

1.5

Time (hrs)

CP

50

XG

40 30 20 10

60

CP + FA

50

XG + FA

40 30 20 10 0

0 0.5

1

1.5

2

3

Time (hrs)

0.5

1

1.5

2

3

Time (hrs)

Fig. 7. Dynamic changes in the gel layer thickness (a, b) and the total work performed against the gel resistance (c, d) as a function of time for HPMC (K15M), carbopol (CP) and xanthan gum (XG) matrices without (a, c) and with (b, d) fumaric acid. Work done against the resistance offered by the gel layer was obtained as area under the F–D profiles. Data indicates meanFS.D. of three independent determinations.

The other reason for efficient retardation by CP and XG could be the contribution of ionic interactions between drug and the anionic polymers. The presence of carboxyl functional groups may lead to OXB HCl– anioinic polymer interactions which are further dependent on ionization state of the functional groups [29,30]. In acidic environment, the carboxyl groups of these polymers are unionized and may not show any interaction [31]. However, further studies are required to confirm this type of interactions for OXB. Presence of fumaric acid in HPMC matrices showed a change in overall gel layer thickness and the work done as compared to gel layer behavior of matrices without fumaric acid. It is evident that initial gel formation was delayed, and a significant difference in gel layer thickness and work done was observed till 1.5 h. The probable reason could be due to fast release of drug from the surface because of its high solubility in acidic environment created by fumaric acid. Initial fast release of drug and excipients drives the medium into matrix, and increases the water

volume fraction in the gel layer because of which the gel layer consistency decrease to certain extent with increase in surface erosion [21]. In case of anionic polymer matrices, presence of fumaric acid, showed a drastic effect on gel layer dynamics, where both the gel layer thickness and the work done against gel resistance were reduced, when compared with control matrices. Gel layer thickness of XG matrices was 3.8F0.24 mm, while it was just 2.24F0.33 mm for XG matrices containing fumaric acid. From the results, it is also obvious that the gel layer thickness showed no growth with time, and further there was not much change in gel consistency. Thus, front synchronization seems to be achieved. Anionic polymers are unionized in acidic pH and show less solubility. It can be concluded that presence of fumaric acid created acidic micro-environmental pH in the matrices leading to low polymer solubility and enhanced erosion process. Drug release is controlled by the volume fractions (concentration) of water, polymer and the

M.V.S. Varma et al. / Journal of Controlled Release 103 (2005) 499–510

drug across the gel layer formed. Due to the pHdependent solubility, concentration gradient across the gel layer will vary as a function of medium pH. When dynamic swelling/dissolution conditions are established in a swellable/hydrophilic matrix, a drug volume fraction gradient is established in the region between the swelling and erosion fronts. As the drug starts dissolving at the swelling front, volume fraction at diffusion front V ds , is given by V ds=C sd V w/q d, where, C s is solubility (g cm3); V w, water volume fraction (cm3 water cm3 gel); q d, drug density [9,28]. It must be noted that for a non-ionizable drug, C s is independent of medium pH and thus only V w can change as a function of time. However, for ionizable drugs C s is highly dependent on pH that exist in the gel layer and thus both C s and V w factors vary as the matrix system progress from stomach through intestine, increasing the complexity of drug release process. Basic drugs like OXB, and other weak bases with pK a 5–7 show exponential change in the C s as a function of pH with negligible solubility above pK a, ultimately changing the diffusion front position and the V ds. HPMC matrices gel layer characteristics and viscosity are pH independent, and thus the gel layer dynamics should remain constant as the formulation progress in GIT. However, for basic drugs as the C s is high at lower pH, drug releases fast from the immature gel layer at low pH range, which further leads to either surface erosion of the matrix or increased V w at the swelling front. Diffusion front position changes in different pH, thus keeping the micro-environmental pH constant throughout the release and producing medium independent drug release. Our data also indicated that CP and XG polymers which showed pH-dependent viscosity and gel layer behavior, shows a sudden release of drugs in lower pH or in the presence of organic acids. It can be concluded that CP and XG show pH-dependent drug release for not only protic molecules but also non-ionic drugs. This provokes the concept that incorporation of basic excipients into matrices will not only keep the solubility of the highly soluble basic drugs low in the low pH range and thus provide prolonged drug release without burst effect but will also improve the gel layer consistency of anionic polymers, irrespective of the medium.

509

4. Conclusions This study has demonstrated the influence of pH on the gel layer behavior and the drug release characteristics of non-ionic and anionic hydrophilic/swellable matrices. Addition of 10%w/w fumaric acid to matrices of various polymers showed pH-independent drug release. However, most efficient and pH-independent release was shown only by HPMC polymer matrices, especially HPMC K15M (50%w/w). Through this approach, it was possible to design a pH-independent directly compressible monolithic system for OXB. Viscosity and texture profiling studies correlated well with the gel layer behavior under different dissolution conditions, and to the OXB release from both the nonionic and anionic polymer matrices.

Acknowledgements The authors would like to thank Unichem Lab. (Mumbai, India) for kindly supplying Oxybutynin hydrochloride. The authors also wish to thank Mr. Gunjan for his skillful technical assistance in viscosity and texture profiling studies, and Mr. Sateesh Kandivilli for valuable comments and inputs in mathematical modeling.

References [1] J.S. Grundy, R.T. Foster, The nifedipine gastrointestinal therapeutic system (GITS). Evaluation of pharmaceutical, pharmacokinetic and pharmacological properties, Clin. Pharmacokinet. 30 (1996) 28 – 51. [2] K.E. Gabr, Effect of organic acids on the release patterns of weakly basic drugs from inert sustained release matrix tablets, Eur. J. Pharm. Biopharm. 38 (1992) 199 – 202. [3] A. Streubel, J. Siepmann, A. Dashevsky, R. Bodmeier, pHindependent release of a weakly basic drug from waterinsoluble and -soluble matrix tablets, J. Control. Release 67 (2000) 101 – 110. [4] S. Nie, W. Pan, X. Li, X. Wu, The effect of citric acid added to hydroxypropyl methylcellulose (HPMC) matrix tablets on the release profile of vinpocetine, Drug Dev. Ind. Pharm. 30 (2004) 627 – 635. [5] A.S. Tataverti, K.A. Mehta, L.L. Augsberger, S.W. Hoag, Influence of methacrylic and acrylic acid polymers on the release performance of weakly basic drugs from sustained release hydrophilic matrices, J. Pharm. Sci. 93 (2004) 2319 – 2331.

510

M.V.S. Varma et al. / Journal of Controlled Release 103 (2005) 499–510

[6] M.V.S. Varma, A.M. Kaushal, A. Garg, S. Garg, Factors affecting mechanism and kinetics of drug release from matrixbased oral controlled drug delivery systems, Am. J. Drug Deliv. 2 (2004) 43 – 57. [7] J. Siepmann, H. Kranz, N.A. Peppas, R. Bodmeier, Calculation of the required size and shape of hydroxypropyl methylcellulose matrices to achieve desired drug release profiles, Int. J. Pharm. 201 (2000) 151 – 164. [8] T.D. Reynolds, S.A. Mitchell, K.M. Balwinski, Investigation of the effect of tablet surface area/volume on drug release from hydroxypropylmethylcellulose controlled-release matrix tablets, Drug Dev. Ind. Pharm. 28 (2002) 457 – 466. [9] P. Colombo, R. Bettini, N.A. Peppas, Observation of swelling process and diffusion front position during swelling in hydroxypropyl methyl cellulose (HPMC) matrices containing a soluble drug, J. Control. Release 61 (1999) 83 – 91. [10] C. Chebli, I. Moussa, S. Buczkowski, L. Cartilier, Substituted amylose as a matrix for sustained drug release, Pharm. Res. 16 (1999) 1436 – 1440. [11] P. Costa, J.M. Sousa Lobo, Modeling and comparison of dissolution profiles, Eur. J. Pharm. Sci. 13 (2001) 123 – 133. [12] N.A. Peppas, J.J. Sahlin, A simple equation for the description of solute release: III. Coupling of diffusion and relaxation, Int. J. Pharm. 57 (1989) 169 – 172. [13] H. Kim, R. Fassihi, A new ternary polymeric matrix system for controlled drug delivery of highly soluble drugs: I. Diltiazem hydrochloride, Pharm. Res. 14 (1997) 1415 – 1421. [14] R. Bettini, P. Colombo, G. Massimo, P.L. Catellani, T. Vitali, Swelling and drug release in hydrogel matrices: polymer viscosity and matrix porosity effects, Eur. J. Pharm. Sci. 2 (1994) 213 – 219. [15] P. Macheras, A. Dokoumetzidis, On the heterogeneity of drug dissolution and release, Pharm. Res. 17 (2000) 108 – 112. [16] L. Yang, B. Johnson, R. Fassihi, Determination of continuous changes in the gel layer thickness of poly(ethylene oxide) and HPMC tablets undergoing hydration: a texture analysis study, Pharm. Res. 15 (1998) 1902 – 1906. [17] V. Pillay, R. Fassihi, A novel approach for constant rate delivery of highly soluble bioactives from a simple monolithic system, J. Control. Release 67 (2000) 67 – 78. [18] A. Avdeef, Physicochemical profiling (solubility, permeability and charge state), Curr. Top. Med. Chem. 1 (2001) 277 – 351.

[19] E. Miyamoto, S. Kawashima, Y. Murata, Y. Yamada, Y. Demizu, Physicochemical properties of oxybutynin, Analyst 199 (1994) 1489 – 1492. [20] The Merck index, vol. XII, Merck & Co., Inc, NJ, 1996, p. 726. [21] R. Espinoza, E. Hong, L.V. Robles, Influence of admixed citric acid on the release profile of pelanserin hydrochloride from HPMC matrix tablets, Int. J. Pharm. 201 (2000) 173 – 265. [22] I.M. Gonzalez, L.V. Robles, Influence of enteric citric acid on the release profile of 4-aminopyridine from HPMC matrix tablets, Int. J. Pharm. 251 (2003) 183 – 193. [23] M.J. Vazquez, B. Perez-Marcos, J.L. Gomez-Amoza, R. Martinez-Pacheco, C. Souto, A. Concheiro, Influence of technological variables on release of drugs from hydrophilic matrices, Drug Dev. Ind. Pharm. 18 (1992) 1355 – 1375. [24] M.J. Vazquez, M. Casalderrey, R. Duro, J.L. Gomez-Amoza, R. Martinez-Pacheco, C. Souto, A. Concheiro, Atenolol release from hydrophilic matrix tablets with hydroxypropyl methylcellulose (HPMC) mixtures as gelling agent: effects of the viscosity of the HPMC mixtures, Eur. J. Pharm. Sci. 4 (1996) 39 – 48. [25] M.C. Bonferoni, C. Caramella, M.E. Sangalli, U. Conte, R.M. Hernandez, J.L. Pedraz, Rheological behavior of hydrophilic polymers and drug release from erodible matrices, J. Control. Release 18 (1992) 205 – 212. [26] L.W. Cheong, P.W. Heng, L.F. Wong, Relationship between polymer viscosity and drug release from a matrix system, Pharm. Res. 9 (1992) 1510 – 1514. [27] P. Gao, R.H. Meury, Swelling of hydroxypropyl methylcellulose matrix tablets: 1. Characterization of swelling using a novel optical imaging method, J. Pharm. Sci. 85 (1996) 725 – 731. [28] P. Colombo, R. Bettini, P.L. Catellani, P. Santi, N.A. Peppas, Drug volume fraction profile in the gel phase and drug release kinetics in hydroxypropylmethyl cellulose matrices containing a soluble drug, Eur. J. Pharm. Sci. 9 (1999) 33 – 40. [29] S. Takka, S. Rajbhandari, A. Sakr, Effect of anionic polymers on the release of propranolol hydrochloride from matrix tablets, Eur. J. Pharm. Biopharm. 52 (2001) 75 – 82. [30] M.A. Dabbagh, J.L. Ford, M.H. Rubinstein, J.E. Hogan, A.R. Rajabi Siahboomi, Release of propranol hydrochloride from matrix tablets containing sodium carboxymethylcellulose, Pharm. Dev. Technol. 4 (1999) 313 – 324. [31] N. Sarisuta, E.L. Parrott, Relationship of dissolution rate to viscosity of polymeric solutions, J. Pharm. Sci. 71 (1982) 1375 – 1380.