Design and Evaluation of Diclofenac Sodium Megaloporous Matrix System Aimed for Colonic Drug Delivery

Document Type : Research Paper


1 Department of Pharmacy, Faculty of Engineering and Technology, Annamalai University, Annamalai Nagar-608 002,

2 Department of Pharmacy, Faculty of Engineering and Technology, Annamalai University, Annamalai Nagar-608 002

3 KK College of Pharmacy, Gerugambakkam, Chennai-602 101

4 The Madras Pharmaceuticals, Karapakkam, Chennai-600 096, TN, India


     Megaloporous controlled release tablets of diclofenac sodium (DS) were prepared with two kinds of granules. One of them is the restraining-phase matrix granule (RMG) and it controls the release rate of the drug. The other one is the soluble housing-phase matrix granule (HMG) and controls liquid penetration into the system. Carnauba wax and Eudragit L 100 polymers were used to constitute the restraining and housing matrix phases, respectively. The prepared tablets were evaluated for various parameters. In vitro drug release study was carried out in simulated gastric fluid (pH 1.2) for the first 2 h and in phosphate buffer (pH 7.2) for the next 10 h following USP 25 paddle method. Two independent model methods, AUC and Lin Ju and Liaw's difference factor (ƒ1) and similarity factor (ƒ2) were used to compare various dissolution profiles. The fabricated megaloporous matrix tablets released only 3 to 5% of DS in pH 1.2 depending on the proportion of carnauba wax used in the RMG. Increase in polymer content/hardness value of the tablet resulted in a significant decrease in AUC0-2 h and AUC2-12 h values . The f1and f2analysis also confirms the discrimination between corresponding dissolutionpairs. The dissolution profiles of an ideal matrix formulation containing 15.77% carnauba wax and 6.76% Eudragit L100 was found to be comparable with the reference product (Voveran® SR) and theoretical release profile. The drug release from all fabricated products and reference product followed better Higuchi model than the zero order and first order kinetic models. Ritger-Peppas model analysis indicated that the DS release followed non-Fickian transport mechanism. From the above analysis, it is evident that the release mechanism of DS from matrix tablet is influenced by both hardness and polymer contents. The stability profiles indicate that the physico-chemical properties of the tablets are not affected on storage at 45°C /75% RH up to 6 months.


1. Introduction

     Diclofenac sodium (DS) is a well-known representative of non-steroidal anti-inflammatory drugs (NSAIDs), widely used to control pain and inflammation of rheumatic and non-rheumatic origin [1]. The conventional DS tablets make the drug immediately available for absorption in upper gastrointestinal (GI) tract resulting local GI toxicity varying from minor gastric discomfort to ulceration and bleeding of the mucosa [2]. It is well documented that the GI toxicity is not only caused by the inhibition of the prostaglandin synthesis, but is probably also due to direct contact of the drug with the mucosa [3]. In addition, due to the rapid systemic clearance of this drug, repeated daily dosing of 3 to 4 times a day is required in maintenance therapy that influences patient compliance. Colon targeted controlled release formulations are thus warranted to promote patient compliance and to reduce upper GI toxicity to some extent. DS is well adsorbed in the colon [4] and thus colon-specific release can be used for the treatment of widespread inflammatory bowel diseases.

     The colon-specific drug delivery may be designed based on one of the following mechanisms viz., prodrugs [5], pH-sensitive polymer [6], time-controlled dissolution [7], and microflora-activated drug release [8]. Geometric shape controlled matrix systems [9], inert matrix systems [10], megaloporous matrix systems [11, 12], lipid matrix systems [13] and hydrogel matrix systems [14] are the various options available to achieve controlled drug release in matrix systems.

      Principally, the megaloporous matrix system comprises two kinds of granules, namely restraining-phase matrix granule (RMG) and the soluble housing-phase matrix granule (HMG), which operate together to release the drug from the dosage form with a constant rate over an extended period of time. Usually, the RMG structure comprises the drug and rate controlling material and the other structure, HMG comprises the drug and a penetration rate controlling material. Previous studies on megaloporous matrix systems [11,12], reveals the use of substances such as carnauba wax, carbopol 934, Eudragit L, stearyl and cetyl alcohol, calcium hydrogen phosphate, talc and Encompress for RMG while polyethylene glycol 6000, lactose, magnesium stearate, calcium hydrogen phosphate and Encompress for HMG.

      In the present  work, megaloporous DS controlled release matrix system has been developed using carnauba wax polymer in RMG and Eudragit L100 polymer in HMG.


Figure 1: The effect of carnauba wax content and tablet hardness on the release profile of diclofenac sodium in simulated gastric fluid (pH 1.2) for the first 2 hours and in phosphate buffer (pH 7.2) for the next 10 hours. Values are given in mean±SD, (n=6).


2. Materials and methods

2.1. Materials

     Diclofenac sodium was obtained from HATTI drugs, Maharashtra. Eudragit L100 was procured from Rohm GmbH, Germany. Carnauba wax was purchased from Koster Keunen Inc. USA. Microcrystalline cellulose (Avicel® PH 101) was purchased from, FMC Corp., Philadelphia. Lactose I.P was obtained from Genuine Chemicals, Mumbai. Polyvinylpyrrolidone (PVP K30 with approximate molecular weight 50 000) was purchased from Merck, Germany. The commercially available DS sustained release tablets (Voveran® SR-75 mg) were procured from Novartis, Mumbai. All other reagents used were of analytical or pharmaceutical grade.

2.2. Preparation and characterization of matrix tablets

     The megaloporous matrix tablets were prepared from HMG and RMG as shown in Table 1. The HMG were prepared by wet mixing of DS, Eudragit L100 and Avicel® PH 101 in 101-capacity rapid mixer-granulator (Rotamix HSMG 10, Kevin Engineers, India) for 5 min. using isopropyl alcohol as granulating fluid. At semi-dried condition the granules were screened through sieve #14 and then dried at 40 °C in a tray drier (Bombay Engineering Works, India) until loss on drying became 2.0% in IR balance (Advanced Ltd. Mumbai, India). The dried granules were screened through sieve #14 and then sieve #80. The RMG were prepared by similar method as HMG employing DS, carnauba wax and lactose using 10% w/v PVP K30 and isopropyl alcohol as granulating agent. Finally, required portions of HMG and RMG were mixed, lubricated with magnesium stearate and compressed on a single punch tablet machine (Korsh, Germany) using 9.0 mm flat punches. The tablets were evaluated for their physical characters [15] like, weight variation, thickness and hardness. The drug content of each formulation was determined spectro-metrically at 276 nm, in a Shimadzu UV-160A spectrophotometer. Friability [16] was measured on a Roche friabilator (Hoffman la Roche, Basel).


Figure 2: Comparison of release profiles of an ideal formulation containing diclofenac sodium (F), reference product (Voveran ® SR) and theoretical release. Values are given in mean±SD, (n=6).


2.3. Plotting of theoretical release profile

     The theoretical release profile was designed from the pharmacokinetic parameters of the drug [17, 18] according to Krüger-Thiemer and Eriksen [19] and Robinson and Eriksen [20]. The theoretical release profile was proposed purposely designed for obtaining a controlled release profile characterized by an initial phase (2 h) of lag-time followed by a controlled release phase (12 h), according to zero order kinetics. The theoretical release profile was composed of initial and sustaining doses and the drug was considered to be released from both the portions simultaneously [21]. In order to select an ideal formulation among various formulations studied, the respective dissolution profiles were compared with the theoretical release profile for similarity.

2.4. In vitro release studies

     The in vitro release of DS from the formulated tablets was carried out in USP 25 dissolution test apparatus II [15] using 900 ml of dissolution medium maintained at 37.0±0.5°C and a stirring rate of 50 rpm. Six tablets from each formulation were tested individually in simulated gastric fluid (pH1.2) for the first 2 h and in phosphate buffer (pH 7.2) for the following 10 h. At every 1 h interval, samples of 5 ml were withdrawn from the dissolution medium and replaced with fresh medium to maintain the volume constant. After filtration and appropriate dilution, the amount of DS present in each sample was determined spectrophotometrical-ly at 276 nm.

2.5. Independent-model method (data analysis)

     In order to compare the dissolution profiles, two model-independent methods were used: Area under dissolution curve (AUC), obtained by trapezoidal rule [22] and Lin Ju and Liaw's [23] difference factor (ƒ1) and similarity factor (ƒ2).

In the present study, AUC was considered instead of mean dissolution time (MDT), because this parameter can probably provide a better indication of in vivo performance [24]. In order to assess the statistical significance between AUC values, a single factor analysis of variance (ANOVA) followed by Duncan's multiple range test (DMRT) [25] was carried out, at a 5% significance level.


The ƒ1  and ƒ2  factors provide a simple measure of similarity between pairs of dissolution profiles. The difference factor (ƒ1) is the percentage difference between two dissolution profiles at each time interval:


ƒ1 = {[Σ | Rt - Tt |] / Σ Rt} × 100 (1)

where, Rt indicates the released amount of drug of reference formulation; and Tt, the released amount of drug of test formulation. If the dissolution profiles are superimposed, ƒ1 reaches a value of 0, whereas the factor value increases when the differences between dissolution profiles also increase. From a practical point of view, values of ƒ1 between 0 and 15 can be considered as superimposed dissolution profiles. The similarity factor (ƒ2) can be calculated using the following expression:


ƒ2 = 50 × log {[1 / (1 + (Σ (Rt - Tt) 2) / N)]1/2 × 100} (2)


where, N indicates the number of experimental data. Values of ƒ2 between 50 and 100 can be considered as superimposed dissolution profiles.



Table 1.Formulation and characteristics of megaloporous tablets containing diclofenac sodium (Batch size-5000 tablets).






Formulation code



Composition (mg)

















Eudragit L100







Avicel® PH 101















Carnauba wax























Magnesium stearate






Physical characteristics







Weight (mg),







Drug content (mg), n=3






Hardness (kPa),







Thickness (mm),







Friability (%),







F1 to F4-fabricated formulations, HMG-housing matrix granules, RMG-restraining matrix granules.


2.6. Dependent-model method (data analysis)


In order to describe the DS release kinetics from individual tablet formulations, the corresponding dissolution data were fitted in various kinetic dissolution models [26]: Zero order, first order, and Higuchi [27], respectively.

Qt = Q0 + K0 t


where, Qt  is the amount of drug released at

time t; Q0 the amount of drug in the solution

at t = 0, (usually, Q0 = 0) and K0 the zero order

release constant.


logQt = logQα+ (K1 /2.303) t


Qα being  the  total  amount  of  drug  in  the

matrix and K1 the first order kinetic constant.

Qt = KH.  t ½





where, KH is the Higuchi rate constant.







mechanism of drug release from matrices, dissolution data were analyzed using the equation proposed by Harland et al. [28].

Q (t-l) /Qα = KK.(t-l)n (6)

where, Qt corresponds to the amount of drug released in time t, l is the lag time (l = 2 h), Qα is the total amount of drug that must be released at infinite time, KK a constant comprising the structural and geometric char-acteristics of the tablet, and ‘n’ is the release exponent indicating the type of drug release mechanism. To determine the exponent ‘n’, only the points in the release curves where Q(t-l)/Qα>0.6 were used. If ‘n’ approaches to 0.5, the release mechanism can be Fickian. If ‘n’ approaches to 1, the release mechanism can be zero order and on the other hand if 0.52). Further, the selected "best model" was confirmed by evaluating the statistical differences between R2 values of various kinetic models, using one way ANOVA followed by DMRT analysis. The values were considered statistically significant if the p < 0.05.


Table 2.Pharmacokinetic data and calculated parameters of diclofenac sodium for designing theoretical release profile.

Pharmacokinetic data

Reported values [16, 17]

Values taken




Elimination half life (t1/2)

1-2 h


1.5 h




Peak plasma conc. (Cp)

1 μg/ml


1 μg/ml




Peak plasma conc. time (Tp)

2-3 h


2.5 h




Volume of distribution (Vd)

0.17±0.11  l/kg


l/kg (or) 11.9






Bioavailability fraction (F)







Desired time of release (T)



12 h




Calculated parameters

Formula [18-20]






Elimination rate constant (Kel)



0.462 h-1




Zero-order release rate (K0)

Cp.Vd. Kel


5.50 mg/h




Initial dose (Db)



22.04 mg




Corrected initial dose (Di)

Db - (Tp.K0)


8.29 mg




Maintenance dose (Dm)



66 mg




Total dose (W)

Di + Dm


74.29 mg 75.00 mg


In vitro drug release (X)

X=Di + t. K0


For t=1




for any time t, (t=1 to12)



X=13.79 mg


*Based on average adult body weight, 70 kg.








2.7. Stability studies

     To assess the stability of DS megaloporous matrices, tablets of batch F4 were stored at 45°C/75% RH for 6 months [29]. The matrix tablets were observed for changes in physical appearance, color, drug content and dissolution characteristics after each month for 6 months. The assay of DS and the dissolution study followed the same procedure as previously described.


3. Results and discussion

     Theoretical release profile was worked out based on pharmacokinetic parameters of DS as shown in Table 2. The theoretical release profile designed in this study composed of initial and sustaining doses with

simultaneous release of the drug from both the portions [21]. The initial dose portion is to bring the blood level of DS immediately to therapeutic concentrations and the sustaining dose portion is to sustain this level for a period of 12 h. The order of release was theoretically set to be zero [30]. The obtained theoretical release values were as follows; dose required to give the desired blood level, when given in an immediately available form (Db) is 22.04 mg, zero-order release rate (K0) is 5.50 mg/h, corrected initial dose (Di) is 8.29 mg, maintenance dose (Dm) is 66 mg and total dose (W) is 74.29 mg, which is approximated to 75 mg.

3.1. Physico-chemical characteristics of DS matrices

     As summarised in Table 1, the evaluation of the prepared matrix tablets containing DS showed that the drug content of all formulations ranged from 99.85 to 101.49%, indicating an uniform amount of drug in the formulations. All formulations, except F2a yielded matrix tablets with a mean hardness value from 5.0 to 5.2 kPa. The formulation F2 and F2aare similar in composition but they differ in manufacturing in only hardness, with the difference of approximately 4 kPa units. All formulations passed the friability test (F < 1%).


3.2. Effect of pH of the dissolution medium

     The effect of pH of the dissolution medium on the release of DS is shown in Figure 1.

Insignificant amount of DS was released after 2 h in pH 1.2 solution, whereas in pH 7.2 dissolution medium, the release was reasonable. This initial lag period of 2 h may be due to (i) pH dependent solubility charac-teristics of both DS and Eudragit L100 (ii) hydrophobic nature of carnauba wax. DS, being a weak acid (pKa 4.0) and Eudragit L100, an enteric polymer (whose solubility begins only at pH>6.0) are practically insoluble in acidic solution, which limits the drug release from the matrix surface in pH 1.2 dissolution medium. On the other hand, carnauba wax is extremely hydrophobic, water repellant and insoluble in aqueous media irrespective of the pH. Owing to the high solubility of DS in pH 7.2 dissolution medium, large portion of the drug from the matrix surface may be released initially, generating many pores and cracks that facilitate further release of the drug. The release of DS was also attributed by the slow dissolution of Eudragit L100 in HMG providing a network of channels and pores that facilities the solvent front penetration and elevation of drug release from the internal core to be possible.Similar result was also reported by Billa et el [31].


 Table 3.AUC values for fabricated products containing diclofenac sodium (F1to F4), reference product (Voveran®SR) andtheoretical release profile.


Formulation code

AUC0-2 h

(% h)

AUC2-12 h

(% h)
















Voveran® SR



Theoretical release




Values are given in mean±SD, n=6. Values not sharing a common superscript letter differ significantly at P < 0.05 (DMRT). The AUC values should be compared within columns only.



3.3. Effect of hardness

     The effect of tablet hardness on the release rate was evaluated and the results compared to each other (Figure 1). Usually an increase in hardness of a tablet is accompanied by a decrease in release rate, due to a decrease in porosity of the tablet [32]. Thus F2a formulation, having higher hardness value of approximately 4 kPa units than F2, showed a significant decrease in the release profile.

     Significant decrease in AUC of F2a against F2 demonstrates the influence of tablet hardness on DS release (Table 3). The f1 and f2 analysis also showed a dissimilarity between the dissolution profiles of F2 and F2a (Table 4). This suggests that the formulation is slightly sensitive to moderate changes in hardness, with regard to the release rate. Since slight changes in hardness values cause nonuniformity of the release profile, hardness of the tablet should be taken into consideration during its commercial production.

3.4. Effect of carnauba wax and Avicel® PH 101 content

     The effect of carnauba wax and Avicel® PH 101 content on the release of DS is shown in Figure 1. The drug release rate decreased with increasing amount of carnauba wax, probably due to the extreme hydrophobic and water repellant nature of carnauba wax polymer [21]. The drug release rate increased with increasing amount of Avicel® PH 101, (replaced by the reduced amount of carnauba wax) where by its swelling behavior, allowed further penetration of the aqueous medium, resulting in rapid erosion of the polymer matrices. These two factors can be ascribed for the higher release rate of drug with formulations containing lower percentage of carnauba wax and higher percentage of Avicel® PH 101. Analogously, a significant decrease (p1 to F4 (Table 3).

     On the other hand, the difference between the AUC values of F4, reference product and theoretical release profile was found to be statistically insignificant (p>0.05). The f1 and f2analysis (Table 4) also suggest that the dissolution profile of F4 is superimposable with the reference product and theoretical release profile (Figure 2). Hence, F4 was adopted as an ideal formulation for further studies.


Table 4. Difference factor (f1) and similarity factor (f2) analysis of various dissolution pairs.


Dissolution pairs



F2 vs F2a



F4 vs Voveran® SR



F4 vs Theoretical release



F4 vs F4*




*After storage at 45°C/75% RH for 6 months.



3.5. Release kinetics

      The in vitro release data were fitted to various kinetic models and the respective drug release rates and release mechanisms were established (Table 5). Since the release of drug in acidic medium was negligible, the portions of the dissolution curves within the interval 3 to 12 h were only considered for the kinetic analysis. The formulation F1 was not considered for the analysis, because in this case more than 80% of the drug was released within the 5th h. As indicated by higher R2 value, the drug release from all formulations and reference product followed Higuchi model than the zero order and first order equations. This observation was further confirmed by ANOVA and DMRT studies, that the differences between R2 values of various kinetic models were statistically significant at p < 0.5.

     The analysis of experimental data in the light of the Harland et al equation, as well as the interpretation of the corresponding values of n, leads to a better understanding of the balance between these mechanisms. For formulations F2 to F4, n ranged from 0.573 to 0.718 (Table 5). This indicates that the release mechanism of DS from these matrices is non- Fickian transport, which suggests that both dissolution and diffusion of the drug in matrices and also its own erosion modulate drug release.


3.6. Stability studies

     The ideal batch F4 showed no changes in physical appearance, color, drug content and dissolution characteristics upon storage at 45 °C/75% RH for 6 months. The f1 and f2 analysis also showed a superimposable dissolution curves before and after the period of this storage (Table 4).


Table 5. Results of model fitting of diclofenac sodium controlled release tablets.












KH (%h-1/2)



K0 (%h-1)




K1 (%h-1)





































Formulation code













KK (h-n)
























































Values are given in mean±SD, n=6. Values not sharing a common superscript letter differ significantly at P < 0.05 (DMRT). The R2 values should be compared within rows only.




4. Conclusions

     In conclusion, the megaloporous tablets prepared with two simple granules could be used as an ideal dosage forms at the dosage intervals of every 12 h. The dissolution profiles of the ideal formulation F4 containing 15.77% carnauba wax and 6.76% Eudragit L100 was found to be superimposed with the theoretical and reference product release profile. The method developed for the megaloporous matrix preparation is simple and cheap and they do not need additional equipment and procedures for the industrial applications. Further, extrapolating the ideal formulation to commercial scale is also easily feasible by performing sufficient scaling up studies. Possibly, by focusing more attention in particular to ensure homogenous mixing of ingredients and consistent tablet hardness during industrial production, the similar dissolution profiles as the ideal formulation may be obtained.

     The analysis of the release profiles in the light of distinct kinetic models (zero-order, first-order, Higuchi and Harland et al.) led to the conclusion that, the drug release from all formulations followed Higuchi model and the release mechanism of DS from these matrices is anomalous (non-Fickian) transport. The ideal batch F4 showed no change in physical appearance, drug content and dissolution profile upon storage at 45°C/75% RH for 6 months.



      The authors wish to thank The Madras Pharmaceuticals (Chennai, India) for supporting this study.

[1]   Brogen RN, Heel RC, Pakes GE, Speight TM, Avery GS. Diclofenac sodium: A review of its pharmacological properties and therapeutic use in rheumatic diseases and pain of varying origin. Drugs 1980; 20: 24-48.
[2]    Carson J, Notis WM, Orris ES. Colonic ulceration and bleeding during diclofenac therapy. N Engl J Med 1989; 323: 135-7.
[3]    Bjarnasson I, Fehilly B, Smethurst P, Menzies IS, Levi AJ. Importance of local versus systemic effects of non-steroidal anti-inflammatory drugs in increasing small intestinal permeability in man. Gut 1991; 32: 275-7.
[4]    Gleiter CH, Antonin KH, Bieck P, Godbillon J, Schonleber W, Malchow H. Colonoscopy in the investigation of drug absorption in healthy volunteers. Gastrointest Endosc 1985; 31: 71-3.
[5]     Schacht E, Gevaert A, Kenawy ER, Molly K, Verstraete W, Adriaensens P, Carleer R, Gelan T. Polymers for colon specific drug delivery. J Control Rel 1996; 39: 327-38.
[6]      Khan MZI, Prebeg Z, Kurjakovic N. A pH-dependent colon targeted oral drug delivery system using methacrylic acid copolymers, I: manipulation of drug release using eudragit L100-55 and Eudragit S100 combinations. J Control Rel 1999; 58: 215-22.
[7]      Fukui E, Miyamura N, Uemura K, Kobayashi M. Preparation of enteric coated timed-release press-coated tablets and evaluation of their function by in vitro and in vivo tests for colon targeting. Int J Pharm 2000; 204: 7-15.
[8]      Prasad YVR, Krishnaiah YSR, Satyanarayana S. In vitro evaluation of guar gum as a carrier forcolon-specific drug delivery. J Conrol Rel 1998; 58: 281-7.
[9]      Karasulu HY, Ertan G. Different geometric shaped hydrogel theophylline tablets: Statistical approach for estimating drug release. IL Farmaco 2002; 57: 939-45.
[10]  Fassihi AR. Continuous matrix formation for controlled drug release: Compression of isotropic polymeric system. Int J Pharm 1986; 34: 169-72.
[11]  Ozgüney I, Ertan G, Güneri T. Dissolution char-acteristics of megaloporous tablets prepared with two kinds of matrix granules. IL Farmaco 2004; 59: 549-55.
[12]  Haan PD, Lerk CF. Parameters affecting the rate of exposure of the drug containing phase in the megaloporous system. Acta Pharm Technol 1988; 34: 106-7.
[13]  Kuttel S, Mezei J, Racz I. Formulation of a controlled-release iron preparation. Pharm Ind 1990; 52: 121-3.
[14]  Shah SS, Kulkarni MG, Mashelkar RA. pH dependent zero-order release from glassy hydrogels: Penetration vs. diffusion controlled. J Control Rel 1991; 15: 221-32.
[15]  Indian pharmacopoeia. Vol II. New Delhi:Controller of Publications, Government of India, 1996.
[16]  The United States Pharmacopoeial Convention, USP 25-NF-20. Rockville, 2002.
[17]  Benet LZ, Williams RL. Design and optimization of dosage regimens: Pharmacokinetic data. In: Goodman A, Gilman A, Rall TW, Nies AS, Taylor P, (editors). Goodman and Gillman's the pharma-cological basis of therapeutics. New York:Pergamon Press, 1990; pp.1650-736.
[18]   Benson MD, Aldo-Benson M, Brandt KD. Synovial fluid concentrations of diclofenac in patients with rheumatoid arthritis or osteoarthritis.Semin Arthritis Rheum 1985; 15: 65-7.
[19]  Krüger-Thiemer E, Eriksen SP. Mathematical model of sustained release preparations and its analysis. J Pharm Sci 1966; 55: 1249-53.
[20]  Robinson JR, Eriksen SP. Theoretical approach to sustained-release multiple-dose therapy: Noncumulative attainment of desired blood level. J Pharm Sci 1970; 59: 1796-800.
[21]  Lordi NG. Sustained release dosage forms. In: Lachman L, Lieberman HA, Kanig JL, (editors).The theory and practice of industrial pharmacy.Bombay: Varghese Publishing House, 1990; pp. 430-56.
[22]  Gibaldi M. Biopharmaceutics and clinical phar-macokinetics. Philadelphia: Lea and Febiger,1991; pp. 377-8.
[23]  Lin Ju H, Liaw SJ. On the assessment of similarity of drug dissolution profiles: A simulation study. Drug Inf J 1997; 31: 1273-89.
[24] Banakar UV. Dissolution & Bioavailability. In: Banakar UV, (editor). Pharmaceutical dissolution testing. New York: Marcel Dekker INC, 1992; pp.347-84.
[25] Duncan BD. Multiple range tests for correlated and heteroscedastic means. Biometrics 1957; 13: 359-64.
[26] Bamba M, Puisieux JP, Marty JP, Carstensen JT. Physical model for release of drug from gel forming sustained release preparations. Int J Pharm 1979; 3: 87-92.
[27] Higuchi T. Mechanism of sustained-action medication: Theoretical analysis of rate of release of solid drugs dispersed in solid matrices. J Pharm Sci 1963; 52: 1145-9.
[28]  Harland RS, Gazzaniga A, Sangalli ME, Colombo P, Peppas NA. Drug-polymer matrix swelling and dissolution. Pharm Res 1988; 5: 488-94.
[29] Carstensen JT. Solid state stability. In: Carstensen JT, Rhodes CT, (editors). Drug stability: Principles and practices. New York: MarcelDekker INC, 2000; pp. 145-89.
[30] Tarimci NM, Agabeyoglu I. Studies on sustained release. III. Matrix granules of sulfamethizole. Drug Dev Ind Pharm 1985; 11: 2043-56.
[31] Billa N, Yuen KH, Peh KK. Diclofenac release from eudragit containing matrices and effect of thermal treatment. Drug Dev Ind Pharm 1998; 24: 45-50.
[32] Shlieout G, Zessin G. Investigation of ethylcellulose as a matrix former and a new method to regard and evaluate the compaction data. Drug Dev Ind Pharm 1996; 22: 313-9.