The Application of the Box-Behnken Design Response Surface Methodology to Study Optimized Formulation Variables on the Drug Release Pattern of Benidipine Hydrochloride Extended-release Matrix Tablet

Background: In this research study, an attempt has been made using Box-Behnken design (BBD) Response Surface Methodology to find optimized formulation variables at 3 levels (low, medium, and high) to affect the dependent response, which is the % drug release pattern at different time intervals in extending drug release of Benidipine Hydrochloride(BH) matrix tablets. BH extended-release tablets reduce the side effects of multiple dosing used during conventional tablets. Methods: As in the preliminary work, we have found the most profound formulation factors for extending drug release of BH matrix tablets are Eudragit RS 100 amount (X1), HPMC K 100 M (X2), chitosan amount (X3), which we selected as independent factors at their low and high levels for this study considering % drug release at three different time intervals i.e., R1 (% of drug release in 2 hours), R2 (% of drug release in 15 hours) and R3 (% of drug release in 18 hours) as dependent variables using a dissolution media of phosphate buffer pH of 6.8 with 75rpm. Results: From the experimental runs of prepared tablets as predicted by the Design-Expert software, the model shows a quadratic equation due to less p-value, and a very less difference was observed between adjusted R² and predicted values R² in all selected responses, which we considered as the % drug release. Conclusion: Therefore, by using the graphical response surface plot of BBD software, the optimized formulation of BH extended-release tablet of Eudragit RS 100, HPMC K 100 M, and Chitosan containing an amount of 45mg, 105 mg, and 45.71 mg, respectively shows an extended drug release of more than 18 hours. To construct a satisfying fit of the model for the optimized formulation, a result analysis was carried out for the internally studentized residuals versus the experimental runs indicating all data points were placed within limits, and the values of the predicted and actual response of each run were normally distributed near a straight line.