The Development of a Link Model Consisting of in vitro Drug Release and Tablets Gastric Emptying Time: Application to Diclofenac Enteric Coated Tablets
I. Locatelli, M. Bogataj, A. Mrhar, I. Grabnar
Chair of Biopharmaceutics and Pharmacokinetics, Faculty of Pharmacy, University of Ljubljana, Ljubljana, Slovenia, EU
Objectives: The bioavailability of well permeable drugs administered as single-unit modified release tablet is markedly dependent on the gastric emptying time of such tablet, especially when taken in fasted stomach state. The influence of pH on drug release can be evaluated in vitro. On the other hand, the vector of tablets gastric emptying times is a random variable and can be generated from Weibull distribution function [1,2]. The purpose was to develop a link model consisting of the kinetics of drug release and the kinetics of gastric emptying of tablets under fasting conditions. With such model an in vivo drug release can be predicted.
Methods: In vitro release experiments were performed on enteric coated Voltaren tablets containing 50 mg of diclofenac sodium (Novartis Farma, Italy) using paddle method (USP Apparatus 2). The tablets were placed into simulated gastric fluid (SGF, pH = 1), which was replaced by simulated intestinal fluid (SIF, pH = 6.8) after 2, 10, 30, 50, 70, 90, 110, 130, 150, or 200 minutes in order to mimic several gastric residence times. The amount of diclofenac released was measured in 10 minutes intervals. Each in vitro experiment was performed in triplicates, resulting in total 30 diclofenac release profiles. Due to enteric coating and low diclofenac solubility in acid medium, the diclofenac release in SGF was extremely limited, whereas in SIF the release was rapid and complete. Modelling of diclofenac release data was performed in NONMEM IV.
Results: The diclofenac release data were described by the Weibull model with lag-time and inter-tablet variability on all model parameters. Additionally, a linear relationship between the tablets gastric residence time and the variability in lag-time parameter of the Weibull model was found to adequately describe a link between in vitro diclofenac release and gastric residence time. On the basis of Weibull distribution function with values of 70.2 min for shape parameter and 1.4 for scale parameter, 100 individual tablets gastric emptying times were generated in R environment. The shape and scale parameter values were previously estimated on the basis of the data collected from several studies evaluating the gastric emptying of tablets in healthy subjects under fasting conditions . The generated tablets gastric emptying times were applied to the developed link model using NONMEM simulation step. The results consisted of 100 individual predicted in vivo release profiles. Additionally, the mean predicted in vivo diclofenac release profile was compared to the in vivo diclofenac absorption profile obtained by deconvolution of mean diclofenac plasma concentration profile.
Conclusions: The link between diclofenac release from enteric coated tablets and tablets gastric emptying time was estimated allowing prediction of in vivo diclofenac release profiles. Additionally, a good correlation between the mean predicted in vivo release profile and mean absorption profile was established.
 Locatelli I, et al. Evaluation of dissolution profiles by the aid of a model of gastric emptying of tablets under fasting conditions. Conference abstract. Farm Vest, (Pharmaceutical Journal of Slovenia) 2008;59:185-186, Available on-line (15.3.2010): http://www.sfd.si/modules/catalog/products/prodfile/fv_posebna_stevilka.pdf
 Jamei M, et al. Population-based mechanistic prediction of oral drug absorption. AAPS J 2009;11(2):225-237.