In-vitro and in-vivo correlation for two gliclazide extended-release tablets

The aim of this study was to perform an in-vitro-in-vivo correlation (IVIVC) for two 60-mg gliclazide extended-release formulations (Fast and Slow release) given once a day and to compare their plasma concentrations over time. In-vitro release rate data were obtained for each formulation using the USP apparatus II, paddle stirrer at 50 and 100 rev min(-1) in 0.1 m HCl and pH 7.4 phosphate buffer. The similarity factor (f(2)) was used to analyse the dissolution data. Eighteen healthy subjects participated in the study, conducted according to a completely randomized, two-way crossover design. The formulations were compared using area under the plasma concentration-time curve, AUC(0-infinity) time to reach peak plasma concentration, T-max and peak plasma concentration C-max, while correlation was determined between in-vitro release and in-vivo absorption. A linear correlation model was developed using percent absorbed data versus percent dissolved data from the two formulations. Predicted gliclazide concentrations were obtained by use of a curve fitting equation. Prediction errors were estimated for C-max and area under the curve AUC(0-infinity) to determine the validity of the correlation. 0.1 m HCl at 50 rev min-1 was found to be the most discriminating dissolution method. Linear regression analysis of the mean percentage of dose absorbed versus the mean percentage of in-vitro release resulted in a significant correlation (r(2) > 0.98) for the two formulations. An average percent prediction error for C-max was 4.15% for Fast release and 3.99% for Slow release formulation whereas for AUC(0-infinity) it was 6.36% and 4.66% for Fast release and Slow release formulation, respectively.