Used A_Formulation Optimization of Controlled Release Diclofenac Sodium_Gohel

8/12/2019 Used A_Formulation Optimization of Controlled Release Diclofenac Sodium_Gohel

1/8


2/8


3/8

M .C . Gohel , A.F . Amin / Journal of Controlled Release 51 (1998) 115 122 117

Table 133 full factorial design layout

Batch No. Variable level in t % Drug released80

coded form (min) Y Y Y 6 0 3 6 0 4 8 0

X X X (min) (min) (min)1 2 3

1 2 1 2 1 2 1 397 40.29 76.38 83.712 2 1 2 1 0 434 42.35 73.86 80.283 2 1 2 1 1 453 37.26 70.84 81.774 2 1 0 2 1 477 40.00 69.29 80.315 2 1 0 0 483 37.32 69.17 77.796 2 1 0 1 499 36.82 68.36 76.267 2 1 1 2 1 454 40.65 70.57 81.628 2 1 1 0 532 41.84 66.84 73.549 2 1 1 1 579 29.71 61.82 73.07

10 0 2 1 2 1 347 45.88 81.14 88.0211 0 2 1 0 375 40.72 76.80 85.0012 0 2 1 1 385 40.84 77.43 84.8613 0 0 2 1 395 42.14 76.58 89.6214 0 0 0 429 38.83 72.98 83.4215 0 0 1 448 37.48 71.27 82.2616 0 1 2 1 462 40.34 70.02 80.9817 0 1 0 475 39.98 69.41 80.4918 0 1 1 503 39.50 67.38 75.80

19 1 2 1 2 1 296 51.79 88.00 93.2520 1 2 1 0 307 48.13 86.15 93.5221 1 2 1 1 321 45.42 81.84 95.6322 1 0 2 1 348 40.32 79.25 91.7523 1 0 0 370 40.97 77.27 85.5224 1 0 1 382 41.02 77.77 85.2325 1 1 2 1 415 40.21 75.23 83.4626 1 1 0 436 42.25 73.69 80.0927 1 1 1 511 38.50 66.91 73.87

Coded Actual values

values X X X 1 2 3

2 1 500 5 00 1000 10 251 1500 15 50

X 5 Stirring speed in revolutions per minute (RPM).1 X 5 Concentration of CaCl .2 2 X 5 Percentage of heavy liquid parafn in dispersion medium.3

2.3. Assay triplicate (Absorbance 5 0.029 * Concentration 0.0062) [10].

Aqueous solutions of diclofenac sodium (3 to 30mg/ml) in phosphate buffer (pH 7.2) were prepared 2.4. Drug content and the absorbance was measured on a Hitachidouble beam U-2000 spectrophotometer at 276 nm To determine the efciency of entrapment, the[9]. An equation was generated by tting weighted microspheres were assayed for the drug content.linear regression model to the data obtained in Drug loaded microspheres (100 mg) of each batch


4/8


5/8


6/8

120 M .C . Gohel , A.F . Amin / Journal of Controlled Release 51 (1998) 115 122

spherical in shape and exhibited highest t amongst80all the batches. Moreover, about 30% of the drugwas released in the rst hour (loading dose) andthereafter the drug was released at a fairly controlled

rate. About 85% of the microspheres were found inthe size range of 250 to 500 m. The dissolution dataof microspheres of batch 9, a potential candidate for12 hour in-vitro release, was further analysed toascertain the mechanism of drug release [12]. Therelease prole tted best to Korsmeyer and Peppasequation [13] ( F 5 1.0334) giving the least residualsum of square as compared with Higuchi equation[14] (square root of time versus % drug released,F 5 3.1867) or Weibull equation [15] ( F 5 3.3925).This superiority is, however statistically insignicant

as shown by F -ratio test. The values of correlationcoefcient were found to be 0.9973, 0.9945, and0.9866 for Korsmeyer and Peppas, Higuchi andWeibull models respectively. For the Korsmeyermodel, the values of slope and intercept were foundto be 0.4240 and 2 1.2837 respectively. From thevalue of slope, it can be concluded that the drug isreleased by diffusion of anomalous type (non-Fickian).

Peck and co-workers [16] derived mathematicalFig. 1. Response surface plot.relationship for the expression of entire dissolution

prole from matrix tablets. An effort is made in thelarger microspheres. The signicance of X X inter- present investigation to derive similar type of rela-1 2action indicates that the cross-linking reaction can be tionship. A linear interactive model was generatedfavoured by appropriate selection of stirring speed. using data of percentage drug released at 60, 180,When a correct stirring speed is used intimate 300, 360, and 480 min from all the 27 batches. Thecontact between CaCl and sodium alginate can be Higuchi model tted well to the data set and hence2obtained. The signicance of X X interaction can be square root of time was chosen as an additional2 3explained by the fact, that the action of CaCl will independent variable. The multiple linear regression2be dependent on the viscosity of the dispersion analysis was performed using the actual values.

medium. Since both the X X and X X interactions Microsoft EXCEL was used to derive the equation.1 2 2 3are found to be signicant in nature, one has to use The derived equation describing the dissolution

Eq. (3) for the prediction of t . pattern is shown in Table 3 where, Y is the per-80 2The microspheres of batch 9 were found to be centage drug dissolved at time t . The R was found

Table 3Coefcients for mathematical models

2Response b b b b b b b b b R0 1 2 3 12 13 23 123 sqrt

Y 11.833 0.0144 2 0.0287 2 0.0297 2 0.0007 2 2.6E05 2 0.0073 4.99E 06 3.009 0.9979Y 32.347 0.0149 0.07282 0.1022 2 0.0012 2 0.0001 2 0.0245 2.52E05 0.73576 0Y 70.132 0.0155 2 0.2718 2 0.0405 2 0.0007 2 2.6E05 0.00497 2.11E06 0.95283 6 0Y 76.704 0.0145 0.0658 2 0.0733 2 0.0006 0.0001 2 0.00299 2 1.08E05 0.92124 8 0


7/8

M .C . Gohel , A.F . Amin / Journal of Controlled Release 51 (1998) 115 122 121

to be 0.9779, indicating a good t. The F - test wasfound to be signicant at P , 0.05. The derivedequation may be used for calculating percentage drugrelease from different batches within the factor

space. The residuals was chosen as one of theoptions while carrying out multiple linear regression

analysis in EXCEL and hence calculated percent-age drug release from all the 27 batches is obtainedin the output. The following constraints were chosenfor the selection of acceptance batches; 20% , Y ,6040% ; 50%, Y , 70% ; 65% , Y , 80%. The360 480batches 5, 6, 8, 9, 17 and 18 met the selectioncriteria. But, the batches 5 and 17 showed greater

Fig. 3. Contour curves for Y , Y and Y .60 360 480than 69% drug release after 360 min. Therefore, theyare considered as borderline cases. Hence, the nal

selection was done from batches 6, 8, 9 and 18. The 60, 360 and 480 min. The data are shown in Table 3.drug release rate from batches 6, 8, and 18 was Two-dimensional contour plots for each responsefound to be relatively slow in the terminal phase of were drawn separately and then overlapped as shownthe dissolution test (i.e. 4.5% per h in between 6 and in Fig. 3 for the selection of acceptable region8 h), whereas it was found to be 6.5% per hour from (ABCDA).batch 9. Therefore, a check-point ( X 5 550 RPM,1 X 5 14.5% CaCl solution and X 5 47.5% heavy2 2 3liquid parafn in the dispersion medium) was select- 4. Conclusioned close to the settings of batch 9 to validate thederived equation. The predicted and observed disso- This study demonstrates the use of factorial designlution prole for the check-point is depicted in Fig. for the preparation of controlled release diclofenac

2. The experimental release data compare quite well sodium microspheres. This statistical technique al-to the release prole predicted from the mathematical lows scientists to examine more than one indepen-model. To widen the scope of selection, co-efcients dent variable at a time. The desirable goals can bewere calculated for the percentage drug released at obtained by systematic formulation approach in

shortest possible time.

References

[1] P.A. Todd, E.M. Sorkin, Diclofenac sodium, a reappraisal of its pharmacodynamic and phamacokinetic properties, and

therapeutic efcacy. Drugs 35 (1988) 244285.[2] A.G. Gilman, T.W. Ralf, A.S. Nile. The Pharmacological

Basis of Therapeutics,: Volume I.B. Maxwell PublishingCorporation, 8th edition, 1991, p. 669.

[3] M. Rajaonarivoi, Y.C. Vauthier, G. Couarraze, F. Puisieux, P.Couvreur, Development of a new drug carrier made fromalginate. J. Pharm. Sci. 82(9) (1993) 912917.

[4] O. Skaugrud, Drug Delivery Systems with Alginate andChitosan, in: D.R. Karsa, R.A. Stephenson (Eds.), Excipientsand Delivery Systems for Pharmaceutical Formulations, TheRoyal Society of Chemistry, 1995, pp. 96107.

Fig. 2. Comparative observed and predicted dissolution proles [5] G.E.P. Box, K.B. Wilson, On the experimental attainment of for check point. optimum conditions,, J. R. Stat. Soc. Ser. B 13 (1951) 145.


8/8

122 M .C . Gohel , A.F . Amin / Journal of Controlled Release 51 (1998) 115 122

[6] O. Shirakura, M. Yamada, S. Hashimoto, Ishimaru, K. Second Course in Business Statistics: Regression Analysis,Takayama and T. Nagai, Particle size design using computer third edition, Dellen Publishing Co., California, 1989, pp.optimization technique, Drug Dev. Ind. Pharm. 17(4) (1991) 141226.471483. [12] M. Bamba, F. Puisieux, J.P. Marty, J.T. Carstensen, Int. J.

[7] R. Fassihi, J. Fabian, A.M. Sakr, Application of response Pharmaceutics 2 (1979) 307.surface methodology to design optimization in formulation [13] W.D. Lindner, B.C. Lippold, Drug release from hydrocolloidof a typical controlled release system, Pharm. Ind. 57(12) embeddings with high or low susceptibility to hydrodynamic(1995) 10391043. stress, Pharm. Res. 12(11) (1995) 17811785.

[8] Y.L. Loukas, Formulation optimization of novel multicom- [14] T. Higuchi, Mechanism of sustained-action medication, J.ponent photoprotective liposomes by using response surface Pharm. Sci. 52(12) (1963) 11451149.methodology, Analyst 121(3) (1996) 279284. [15] K.G. Nelson, L.Y. Wang, Determination of time course of

[9] K. Florey, Analytical proles of drug substances, Academic tablet disintegration II: Method using continuous functions,Press Inc., New York, vol. 19, 1990, p. 123. J. Pharm. Sci. 67(1) (1961) 8689.

[10] S. Bolton, Pharmaceutical statistics, 2nd Ed., Marcel Dekker, [16] G.E. Peck, A.D. Johnson, V.L. Anderson, A statisticalNew York, 1990, p. 234. approach for the development of an oral controlled-release

[11] W. Mendenhall, T. Sincich, Multiple Regression, in: A matrix tablet, Pharm. Res. 7 (1990) 10921097.

Documents

Used A_Formulation Optimization of Controlled Release Diclofenac Sodium_Gohel